Commun. Math. Phys. 294, 1–19 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0920-3
Communications in
Mathematical Physics
Some Remarks about Semiclassical Trace Invariants and Quantum Normal Forms Victor Guillemin1, , Thierry Paul2 1 Department of Mathematics, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139, USA. E-mail:
[email protected]
2 CNRS and Département de Mathématiques et Applications, École Normale Supérieure,
45, Rue d’Ulm, F-75730 Paris Cedex 05, France. E-mail:
[email protected] Received: 23 January 2009 / Accepted: 15 June 2009 Published online: 11 September 2009 – © Springer-Verlag 2009
Abstract: In this paper we explore the connection between semi-classical and quantum Birkhoff canonical forms (BCF) for Schrödinger operators. In particular we give a “non-symbolic” operator theoretic derivation of the quantum Birkhoff canonical form and provide an explicit recipe for expressing the quantum BCF in terms of the semiclassical BCF. 1. Introduction Let X be a compact manifold and H : L 2 (X ) → L 2 (X ) a self-adjoint first order elliptic pseudodifferential operator with leading symbol H (x, ξ ). From the wave trace eit E k , (1.1) E k ∈Spec(H )
one can read off many properties of the “classical dynamical system” associated with H , i.e. the flow generated by the vector field ξH =
∂H ∂ ∂H ∂ − . ∂ξi ∂ xi ∂ xi ∂ξi
(1.2)
For instance it was observed in the ‘70’s’ by Colin de Verdière, Chazarain and Duistermaat-Guillemin that (1.1) determines the period spectrum of (1.2) and the linear Poincaré map about a non-degenerate periodic trajectory, γ , of (1.2) ([2–4]). More recently it was shown by one of us [5] that (1.1) determines the entire Poincaré map about γ , i.e. determines, up to isomorphism, the classical dynamical system associated with H in a formal neighborhood of γ . The proof of this result involved a microlocal Birkhoff canonical form for H in a formal neighborhood of γ and an algorithm First author supported by NSF grant DMS 890771.
2
V. Guillemin, T. Paul
for computing the wave trace invariants associated with γ from the microlocal Birkhoff canonical form. Subsequently a more compact and elegant algorithm for computing these invariants from the Birkhoff canonical form was discovered by Zelditch [11,12] making the computation of these local trace invariants extremely simple and explicit. In this paper we will discuss some semiclassical analogues of these results. In our set-up H can either be the Schrödinger operator on Rn − 2 + V with V → ∞ as x tends to infinity, or more generally a self-adjoint semiclassical elliptic pseudodifferential operator H (x, Dx ) whose symbol, H (x, ξ ), is proper (as a map from T ∗ X into R). Let E be a regular value of H and γ a non-degenerate periodic trajectory of period Tγ lying on the energy surface H = E 1. Consider the Gutzwiller trace (see [6]) E − Ei ψ , (1.3) where ψ is a C ∞ function whose Fourier transform is compactly supported with support in a small neighborhood of Tγ and is identically one in a still smaller neighborhood. As shown in [8,9] (1.3) has an asymptotic expansion ei
Sγ
+σγ
∞
a k k ,
(1.4)
k=0
and we will show below how to compute the terms of this expansion to all orders in terms of a microlocal Birkhoff canonical form for H in a formal neighborhood of γ by means of a Zelditch-type algorithm 2 . If γ is non-degenerate so are all its iterates γ r . Then, for each of these iterates one gets an expansion of (1.3) similar to (1.4), ei
Sγ r
+σγ r
∞
ak,r k ,
(1.5)
k=0 1 For simplicity we will consider periodic trajectories of elliptic type in this paper, however our results are true for non-degenerate periodic of all types, hyperbolic, mixed elliptic hyperbolic, focus-focus, etc. Unfortunately however the Zelditch algorithm depends upon the type of the trajectory and in dimension n there are roughly as many types of trajectories as there are Cartan subalgebras of Sp(2n) (see for instance [1]) i.e. the number of types can be quite large. 2 For elliptic trajectories non-degeneracy means that the numbers
θ1 , . . . , θn , 2π are linearly independent over the rationals, eiθκ , κ = 1, . . . , n being the eigenvalues of the Poincaré map about γ . The results above are true to order O(r ) providing (κ1 θ1 + · · · + κn θ )n + l2π = 0 for all | κ1 | + · · · + | κn |≤ r , i.e. providing there are no resonances of order ≤ r .
Semiclassical Trace Invariants
3
and for these expansions as well the coefficients ak,r can be computed from the microlocal Birkhoff canonical form theorem for H in a formal neighborhood of γ . Conversely one can show Theorem 1.1. The constants ak,r , κ, r = 0, 1, . . . determine the microlocal Birkhoff canonical form for H in a formal neighborhood of γ (and hence, a fortiori, determine the classical Birkhoff canonical form). One of the main goals of this paper will be to give a proof of this result. Our proof, in Sects. 2, 3 and 6 is, with semiclassical modifications, more or less the same as the proof of the Guillemin-Zelditch results [5,11,12] alluded to above. An alternative proof based on Grushun reductions, flux norms and trace formulas for monodromy operators can be found in [7]. Another main goal of this paper is to develop a purely quantum mechanical approach to the theory of Birkhoff canonical forms in which symbolic expansions get replaced by operator theoretic expansions and estimates involving Hermite functions. This can be seen as a “local” version of the Rayleigh-Schrödinger perturbation formalism where no “ ” parameter is involved. The virtue of this approach is that the dependence of the normal form is an intrinsic part of the theory, and avoids any additional semiclassical computation. This approach is developed in Sect. 4 and the connection of this to the symbolic approach of Sects. 2–3 is described in Sect. 5. To conclude these prefatory remarks we would like to thank Cyrille Heriveuax for his perusal of the first draft of the our manuscript and we would also like to express our gratitude to the referee for his careful line-by-line reading of the manuscript and his many helpful suggestions. 2. The Classical Birkhoff Canonical Form Theorem Let M be a 2n + 2 dimensional symplectic manifold, H a C ∞ function and ξH =
∂H ∂ ∂H ∂ − ∂ξi ∂ xi ∂ xi ∂ξi
(2.1)
the Hamiltonian vector field associated with H . Let E be a regular value of H and γ a non-degenerate elliptic periodic trajectory of ξ H lying on the energy surface, H = E. Without loss of generality one can assume that the period of γ is 2π . In this section we will review the statement (and give a brief sketch of the proof) of the classical Birkhoff canonical form theorem for the pair (H, γ ). Let (x, ξ, t, τ ) be the standard cotangent coordinates on T ∗ (Rn × S 1 ) and let √ pi = xi2 + ξi2 and qi = arg (xi + −1 yi ). (2.2) Theorem 2.1. There exists a symplectomorphism, ϕ, of a neighborhood of γ in M onto a neighborhood of p = τ = 0 such that ϕ o γ (t) = (0, 0, t, 0) and ϕ ∗ H = H1 ( p, τ ) + H2 (x, ξ, t, τ ), H2 vanishing to infinite order at p = τ = 0. We break the proof of this up into the following five steps.
(2.3)
4
V. Guillemin, T. Paul
Step 1. For small there exists a periodic trajectory, γ , on the energy surface, H = E + , which depends smoothly on and is equal to γ for = 0. The union of these trajectories is a 2 dimensional symplectic submanifold , , of M which is invariant under the flow of ξ H . Using the Weinstein tubular neighborhood theorem one can map a neighborhood of γ symplectically onto a neighborhood of p = τ = 0 in T ∗ (Rn × S 1 ) such that gets mapped onto p = 0 and ϕ o γ (t) = (0, 0, t, 0). Thus we can henceforth assume that M = T ∗ (Rn × S 1 ) and is the set, p = 0. Step 2. We can assume without loss of generality that the restriction of H to is a function of τ alone, i.e. H = E + h(τ ) on . With this normalization, H = E + h(τ ) +
θi (τ ) pi + O( p 2 ),
(2.4)
where h(τ ) = τ + O(τ 2 ) and θi = θi (0), i = 1, . . . , n
(2.5)
are the rotation angles associated with γ . Since γ is non- degenerate, θ1 , . . . , θn , 2π are linearly independent over the rationals. Step 3. Theorem 2.1 can be deduced from the following result (which will also be the main ingredient in our proof of the “microlocal” Birkhoff canonical theorem in the next section). Theorem 2.2. Given a neighborhood, U, of p = τ = 0 and G = G(x, ξ, t, τ ) ∈ C ∞ (U), there exist functions F, G 1 , R ∈ C ∞ (U) with the properties i. G 1 = G 1 ( p, τ ), ii. {H, F} = G + G 1 + R, iii. R vanishes to infinite order on p = τ = 0. Moreover, if G vanishes to order κ on p = τ = 0, one can choose F to have this property as well. Proof of the assertion. Theorem 2.2 ⇒ Theorem 2.1. By induction one can assume that H is of the form, H = H0 ( p, τ ) + G(x, ξ, t, τ ), where G vanishes to order κ on p = τ = 0. We will show that H can be conjugated to a Hamiltonian of the same form with G vanishing to order κ + 1 on p = τ = 0. By Theorem 2.2 there exists an F, G and R such that F vanishes to order κ and R to order ∞ on p = τ = 0, G 1 = G 1 ( p, τ ) and {H, F} = G + G 1 + R. Thus 1 {F, {F, H }} + · · · 2! = H0 ( p, τ ) − G 1 ( p, τ ) + · · · ,
(expξ F )∗ H = H + {F, H } +
the “dots” indicating terms which vanish to order κ + 1 on p = τ = 0. Step 4. Theorem 2.2 follows (by induction on κ) from the following slightly weaker result:
Semiclassical Trace Invariants
5
Lemma 2.3. Given a neighborhood, U, of p = τ = 0 and a function, G ∈ C ∞ (U), which vanishes to order κ on p = τ = 0, there exists functions F, G 1 , R ∈ C ∞ (U) such that i. G 1 = G 1 ( p, τ ), ii. {H, F} = G + G 1 + R, iii. F vanishes to order κ and R to order κ + 1 on p = τ = 0. Step 5. Proof of Lemma 2.3. In proving Lemma 2.3 we can replace H by the Hamiltonian θi pi , H0 = E + τ + since H ( p, q, t, τ ) − H0 ( p, q, t, τ ) vanishes to second order in τ, p. Consider now the identity {H0 , F} = G + G 1 ( p, τ ) + O( p ∞ ). √ √ Introducing the complex coordinates, z = x + −1ξ , and z = x − −1ξ , this can be written as n √ ∂F ∂ ∂ F+ −1 θi z i − zi = G + G 1 + O( p ∞ ). ∂z i ∂z i ∂t i=1
Expanding F, G and G 1 in Fourier-Taylor series about z = z = 0: aµ,ν,m (τ )z µ z ν e2πimt , F= µ=ν
G= G1 =
bµ,ν,m (τ )z µ z ν e2πimt , cµ (τ )z µ z µ ,
µ
one can rewrite this as the system of equations n √ −1 θi (µi − νi ) + 2π m aµ,ν,m (τ ) = bµ,ν,m (τ )
(2.6)
i=1
for µ = ν or µ = ν and m = 0, and − cµ (τ ) = bµ,µ,0 (τ )
(2.7)
for µ = ν and m = 0. By assumption the numbers, θ1 , . . . , θn , 2π, are linearly independent over the rationals, so this system has a unique solution. Moreover, for µ and ν fixed, bµ,ν,m (τ )e2πimt is the (µ, ν) Taylor coefficient of G(z, z, t, τ ) about z = z = 0; so, with µ and ν fixed and j >> 0, | bµ,ν,m (τ ) |≤ Cµ,ν, j m − j
6
V. Guillemin, T. Paul
for all m. Hence, by (2.6), − j−1 | aµ,ν,m (τ ) |≤ Cµ,ν, jm
for all m. Thus aµ,ν (t, τ ) =
aµ,ν,m (τ )e2πimt
is a C ∞ function of t and τ . Now let F(z, z, t, τ ) and G 1 ( p, τ ) be C ∞ functions with Taylor expansion:
aµ,ν (t, τ )z µ z ν
µ=ν
and
cµ (τ )z µ z µ
µ
about z = z = 0. Note, by the way, that if G vanishes to order κ on p = τ = 0, so does F and G; so we have proved Theorem 2.2 (and, a fortiori Lemma 2.3) with H replaced by H0 . 3. The Semiclassical Version of the Birkhoff Canonical Form Theorem Let X be an (n + 1)-dimensional manifold and H : C0∞ (X ) → C ∞ (X ) a semiclassical elliptic pseudo-differential operator with leading symbol, H (x, ξ ), and let γ be a periodic trajectory of the bicharacteristic vector field (2.1). As in Sect. 1 we will assume that γ is elliptic and non-degenerate, with rotation numbers (2.4). Let Pi and Dt be the differential operators on Rn × S 1 associated with the symbols (2.2) and τ , i.e. Pi = −2 ∂x2i + xi2 and Dt = −i∂t . We will prove below the following semiclassical version of Theorem 2.1 ∞ Theorem n 3.1.1 There exists a semiclassical Fourier integral operator Aϕ : C0 (X ) → ∞ C R × S implementing the symplectomorphism (2.3) such that microlocally on a neighborhood, U, of p = τ = 0,
A∗ϕ = A−1 ϕ
(3.1)
Aϕ H A−1 ϕ = H (P1 , . . . , Pn , Dt , ) + H ,
(3.2)
and
the symbol of H vanishing to infinite order on p = τ = 0.
Semiclassical Trace Invariants
7
Proof. Let Bϕ be any Fourier integral operator implementing ϕ and having the property (3.1). Then, by Theorem 2.1, the leading symbol of Bϕ H Bϕ−1 is of the form H0 ( p, τ ) + H0 ( p, q, t, τ ),
(3.3)
H0 ( p, q, t, τ ) being a function which vanishes to infinite order on p = τ = 0. Thus the symbol,H0 , of Bϕ H Bϕ−1 is of the form H0 ( p, τ ) + H0 ( p, q, t, τ ) + H1 ( p, q, t, τ ) + O(2 ).
(3.4)
By Theorem 2.2 there exists a function, F( p, q, t, τ ), with the property {H0 , F} = H1 ( p, q, t, τ ) + H1 ( p, τ ) + H1 ( p, q, t, τ ),
(3.5)
where H1 vanishes to infinite order on p = τ = 0. Let Q be a self-adjoint pseudo-differential operator with leading symbol F and consider the unitary pseudo-differential operator Us = eis Q . Let −1 Hs = Us Bϕ H Us Bϕ = Us Bϕ H Bϕ−1 U−s . Then ∂ Hs = i[Q, Hs ], ∂s
(3.6)
∂ so ∂s Hs is of order −1, and hence the leading symbol of Hs is independent of s. In par∂ Hs is equal, by (3.6) to the leading symbol of i[Q, Hs ] ticular the leading symbol of ∂s which, by (3.5), is: − H1 ( p, q, t, τ ) + H1 ( p, τ ) + H1 ( p, q, t, τ ) .
Thus by (3.4) and (3.5) the symbol of −1 U1 Bϕ H U1 Bϕ = Bϕ H Bϕ−1 +
1 0
∂ Hs ds ∂s
is of the form H0 ( p, τ ) + H1 ( p, τ ) + H0 + H1 + O(2 ),
(3.7)
the term in parentheses being a term which vanishes to infinite order on p = τ = 0. By repeating the argument one can successively replace the terms of order 2 , . . . , r , etc. in (3.7) by expressions of the form r Hr ( p, τ ) + Hr ( p, q, t, τ ) with Hr vanishing to infinite order on p = τ = 0.
8
V. Guillemin, T. Paul
4. A Direct Construction of the Quantum Birkhoff Form In this section we present a “quantum” construction of the quantum Birkhoff normal form which is in a sense algebraically equivalent to the classical one of Sect. 2. To do this we will need to define for operators the equivalent of “a Taylor expansion which vanishes at a given order”. We will first start in the L 2 (Rn × S 1 ) setting, and show at the end of the section the link with Theorem 3.1. Definition 4.1. Let us consider on L 2 (Rn × S 1 , d xdt) the following operators: • ai− = • ai+ = • Dt =
√1 (x i + ∂x ) i 2 √1 (x i − ∂x ) i 2 ∂ −i ∂t
We will say that an operator A on L 2 (Rn × S 1 ) is an “ordered polynomial of order greater than p ∈ N” (OPOG( p)) if there exists P ∈ N such that: i
A=
[2] P
j
αi j (t, )Dt
i= p j=0
i−2 j
bl
(4.1)
l=1
with, ∀l, bl ∈ {a1− , a1+ , . . . , an− , an+ } and αi j ∈ C ∞ (S1 × [0, 1[). i−2 j In (4.1) bl is meant to be the ordered product b1 . . . bi−2 j . l=1
The meaning of this definition is clarified by the following: basis of L 2 (Rn × S 1 ) defined by Hµ (x, t) = −n/4 Lemma√ 4.2. Let Hµ denote √ the iµ t , where the h are the (normalized) Hermite funcn+1 h µ1 (x1 / ) . . . h µn (xn / )e j
tions. Let us define moreover |µ| := µ2 2 . Let A be an OPOG(p). Then: ∀M < +∞ (microlocal “cut-off”), ∃ C = C(A, M) such that, p
||AHµ || L 2 ≤ C|µ| 2 , ∀µ ∈ Nn+1 s.t |µ| ≤ M. Proof. The proof follows immediately from the two well known facts (expressed here in one dimension): a ± Hµ =
(µ ± 1)Hµ±1
and Dt eimt = meimt .
For the rest of this section we will need the following collection of results.
Semiclassical Trace Invariants
9
Proposition 4.3. Let A be a (Weyl) pseudodifferential operator on L 2 (Rn × S 1 ) with symbol of type S1,0 . Then, ∀L ∈ N and ∀M < +∞, there exists an OPOG(1) A L and a constant C = C(A, L , M) such that, ||(A − A L )Hµ || L 2 ≤ C|µ|
L+1 2
), ∀µ ∈ Nn+1 s.t. |µ| ≤ M.
Moreover, if the principal symbol of A is of the form: a0 (x1 , ξ1 , . . . , xn , ξn , t, τ ) = θi (xi2 + ξi2 ) + τ + h.o.t., (or is any function whose symbol vanishes to first order at x = ξ = τ = 0) then A L is an OPOG(2). Proof. Let us take the L th order Taylor expansion of the (total) symbol of A in the variables x, ξ, τ, near the origin. Noticing that a pseudodifferential operator with polynomial symbol in x, ξ, τ, is an OPOG, we just have to estimate the action, on Hµ , of a pseudo-differential operator whose symbol vanishes at the origin to order L in the variable x, ξ, τ . The result is easily obtained for the τ part, as the “t” part of Hµ is an exponential. For the µ part we will prove this result in one dimension, the extension to n dimensions being straightforward. Let us define a coherent state at (q, p) to be a function of the form ψqap (x) := px √ −1/4 a x−q ei , for a in the Schwartz class and ||a|| L 2 = 1. Let us also set ϕq p = ψqap
for a(η) = π −1/2 e−η /2 . It is well known, and easy to check using the generating function of the Hermite polynomials, that
1 t Hµ = − 4 e−i 2 ϕq(t) p(t) dt, 2
S1
where q(t) + i p(t) = eit (q + i p) , q 2 + p 2 = (µ + 21 ). Therefore, for any operator A, ||AHµ || = O(
sup p 2 +q 2 =(µ+ 12 )
1
− 4 ||Aϕq p ||).
(4.2)
Lemma 4.4. let H a pseudodifferential operator whose (total) Weyl symbol vanishes at the origin to order M. Then, if q 2 +p2 = O(1): M ||H ψqap || = O ( p 2 + q 2 ) 2 . Before proving the lemma we observe that the proof of the proposition follows easily from the lemma using (4.2). Proof. An easy computation shows that, if h is the (pseudodifferential) symbol of H , then H ψqap = ψqbp with
√ √ b(η) = h(q + η, p + ν)eiην a(ν)dν, ˆ (4.3) R
where aˆ is the ( independent) Fourier transform of a.
10
V. Guillemin, T. Paul
k k+1 bk 2 2 ), where Developing (4.3) we get that H ψqap = k=K k=M Dk h(q, p)ψq p + O( bk ∈ S and Dk is an homogeneous differential operator of order k. It is easy to conclude, thanks to the hypothesis q 2 +p2 = O(1), that k
2 (q 2 + p 2 )
M−k 2
M
= O((q 2 + p 2 ) 2 ),
M+1 2
M
= O((q 2 + p 2 ) 2 ).
This proposition is crucial for the rest of this section, as it allows us to reduce all computations to the polynomial setting. For example A may have a symbol bounded at infinity (class S(1), an assumption which we will need for the application below of Egorov’s Theorem in the proof of Theorem 4.9), but, with respect to the algebraic equations we will have to solve, one can consider it as a “OPOG” (see Theorem 4.9 below). Lemma 4.5. Let A be a OPOG(1) on L 2 (Rn × S 1 ). Let us suppose that A is a symmetric operator. For P ∈ N (large), let A P := A + (|Dθ |2 + |x|2 + |Dx |2 ) P .
(4.4)
Then A P is an elliptic selfadjoint pseudo-differential operator. Therefore eis family of unitary Fourier integral operators.
AP
is a
Proof. It is enough to observe that A P is, defined on the domain of |Dθ |2 + |x|2 + |Dx |2 , a selfadjoint pseudodifferential operator with symbol of type S1,0 .
Lemma 4.6. Let H0 be the operator H0 =
n
θi ai− ai+ + Dt ,
1
then, if W is an OPOG(r ), so is
[H0 ,W ] i .
d is H0 / e W eis H0 /|s=0 which, since H0 is quadratic, is the same polyProof. [H0i ,W ] = ds nomial as W modulo the substitution ai− → eis ai− , ai+ → e−is ai and shifting of the coefficients in t by s. Therefore the result is immediate.
More generally: Lemma 4.7. For any H and W of type OPOG(m) and OPOG(r ) respectively, an OPOG(m + r − 2).
[H,W ] i
is
The proof is immediate noting that [ai− , a +j ] = δi j and that, for any C ∞ function a(t), [Dt , a] = ia . We can now state the main result of this section: Theorem 4.8. Let H be a (Weyl) pseudo-differential operator on L 2 (Rn × S 1 ) whose principal symbol is of the form: H0 (x, ξ ; t, τ ) =
n 1
θi (xi2 + ξi2 ) + τ + H2 ,
Semiclassical Trace Invariants
11
where H2 vanishes to third order at x = ξ = τ = 0 and θ1 , . . . , θn , 2π are line2 arly independent over the rationals. Let us define, as before, Pi = −2 ∂∂x 2 + xi2 and i
Dt = −i ∂t∂ . Then, ∀M < +∞, there exists a family of unitary operators (U L ) L=3... and constants (C L ) L=3... , and a C ∞ function h( pi , . . . , pn , τ, ) such that: L+1 || U L HU L−1 −h(P1 , . . . , Pn , Dt , ) Hµ || L 2 (Rn ×S 1 ) ≤ C L |µ| 2 ∀µ ∈ Nn+1 s.t |µ| ≤ M.
Proof. The proof of Theorem 4.8 will be a consequence of the following: Theorem 4.9. Let H be as before, and let G be an OPOG(3). Then there exists a function G 1 ( p1 , . . . , pn , τ, ), an OPOG F and an operator R such that: i.
[H,F] i
= G + G 1 + R, L+1
ii. R satisfies: ||R Hµ || = O(|µ| 2 ), ∀µ ∈ Nn+1 , |µ| = O(1) and ∀L ∈ N, iii. if G is an OPOG(κ) so is F, iv. if G is a symmetric operator, so is F and G 1 is real. Let us first prove that Theorem 4.9 implies Theorem 4.8: By induction, as in the “classical” case and thanks to Proposition 4.3, one can assume that H is of the form H = H0 + G, where G is an OPOG(κ). Let us consider the operators ei FP
FP
H e−i
FP
and
FP
H (s) := eis H e−is , where F satisfies Theorem 4.9 and FP is defined by (4.4) for P large enough. Since we are in an iterative perturbative setting, it is easy to check by taking P large FP
F
enough that we can omit the subscript P in H (s) and let e±i stand for e±i in the rest of the computation. We have: [F,H ] F F [F, H ] [F, i ] [F, [F, [F, 10 t0 s0 H (u)dudsdt]/i]/i] ei H e−i = H + + + i i i [F,H ] [F, H ] [F, i ] ˜ + +R = H0 + G + i i ] [F, [F,H i ] ˜ = H0 − G 1 + R + + R. (4.5) i Since we are interested in letting all the operators acting on the Hµ for |µ| = O(1), we can microlocalize near x = ξ = τ = 0 and replace F and H by their microlocalF˜ ized versions F˜ and H˜ . ei is a Fourier integral operator and, byEgorov’s Theorem, H˜ (s) is a family of pseudodifferential operators, and so is 10 t0 s0 H˜ (u)dudsdt. By Proposition 4.3, Lemma 4.7 and Lemma 4.2 we have, since G is an OPOG(κ), || R˜ Hµ || = O(|µ|κ+1 ). [F,H ]
[F, i ] By the same argument, satisfies the same estimate. Developing R˜ by the i Lagrange formula (4.5) to arbitrary order, we get, thanks to Lemma 4.7, R˜ = G˜ + R, where G˜ is an OPOG(κ + 1) and
||R Hµ || = O(|µ|
L+1 2
).
12
Therefore, letting G =
V. Guillemin, T. Paul ] [F, [F,H i ] i
ei
FP
˜ we have: + G,
H e−i
FP
= H0 + G 1 + G + R,
with G an OPOG(κ + 1). By induction Theorem 4.8 follows. Proof of Theorem 4.9. Let us first prove the following Lemma 4.10. Let H0 be as before and let G be an OPOG(r ). Then there exists a OPOG(r ) F and G 1 = G 1 ( p1 , . . . , pn , Dt , ), such that [H0 , F] = G + G1. i
(4.6)
Proof. By Lemma 4.6, if F is an OPOG, it must be an OPOG(r ), since the left-hand side of (4.6) is an OPOG(r ). Let us take the matrix elements of (4.6) relating to the Hµ s. We get: −i.(µ − ν) < µ|F|ν >=< µ|G + G 1 |ν > + < µ|R|ν >, where .(µ − ν) := n1 θi µi + µn+1 and < µ|.|ν >= Hµ , .Hν . We get immediately that G 1 (µ, ) = − < µ|G|µ >. Moreover, let us define F by: < µ|F|ν >:=
< µ|G + G 1 |ν > , −i.(µ − ν)
which exists by the non-resonance condition. To show that F is an OPOG one just j has to decompose G = G l in monomial OPOGs G l = α(t)Dt b1 . . . bm , bi ∈ + + {a1 , a1 , . . . , an , an }. Then, for each ν there is only one µ for which < µ|G +G 1 |ν >= 0 and the difference µ − ν depends obviously only on G l , not on ν. Let us call this difference ρG l . Then F is given by the sum: F=
1 Gl . −i.ρG l
It is easy to check that one can pass from Lemma 4.10 to Theorem 4.9 by induction, writing [H, F + F ] = [H, F] + [H0 , F ] + [H − H0 , F] + [H − H0 , F ].
We will show finally that Theorems 4.8 and 3.1 are equivalent. Once again we can start by considering an Hamiltonian on L 2 (Rn × S 1 ) since any Fourier integral operator Bϕ , as defined in the beginning of the proof of Theorem 3.1, intertwines the original Hamiltonian H : C0∞ (X ) → C ∞ (X ) of Sect. 3 with a pseudodifferential operator on L 2 (Rn × S 1 ) satisfying the hypothesis of Theorem 4.8. W3
W4
Let us remark first of all that if U L = ei ei . . . ei integral operators, then so is U L . Secondly we have
WL
, all ei
Wl
being Fourier
Proposition 4.11. Let A be a pseudodifferential operator of total Weyl symbol a(x, ξ, t, τ, ). Then a vanishes to infinite order at p = τ = 0 if and only if ||AHµ || L 2 (Rn ×S 1 ) = O(|µ|∞ ).
Semiclassical Trace Invariants
13
Proof. The “if” part is exactly Proposition 4.3. For the “only if” part let us observe that, if the total symbol didn’t vanish to infinite order, then it would contain terms of the form αkmnr (t)k (x + iξ )m (x − iξ )n τ r . Let us prove this can’t happen in dimension 1, the extension to dimension n being straightforward. Each term of the form (x + i Dx )m (x − i Dx )n = a m (a + )n gives rise to an operator Am,n such that: Am,n Hµ =
|m+n| 2
∼ |µ| Therefore
(µ + 1) . . . (µ + n)(µ + n − 1) . . . (µ + n − m)Hµ+m−n
m+n 2
Hµ+m−n .
cmn Am,n Hµ = ||
cm,m−l Hµ+l ∼
cmn Am,n Hµ ||2 ∼
cm,m−l |µ|
2m−l 2
Hµ+l . In particular:
|µ|2m−l ,
so || cmn Amn Hµ || = O(|µ|∞ ) implies Cmn = 0. It is easy to check that the same argument is also valid for any ordered product of a’s and a + ’s.
In the next section we will show how the functions H of Theorem 4.8 and h of Theorem 3.1 are related. 5. Link Between the Two Quantum Constructions Consider a symbol (on R2n ) of the form h( p1 , . . . , pn ) ξ 2 +x 2
with pi = i 2 i . There are several ways of quantizing h: one of them consists in associating to h, by the spectral theorem, the operator h(P1 , . . . , Pn ) = h(P), −2 ∂x2 +x 2
i i where Pi = . Another one is the Weyl quantization procedure. 2 In this section we want to compute the Weyl symbol h we of h(P1 , . . . , Pn ) and apply the result to the situation of the preceding sections. By the metaplectic invariance of the Weyl quantization and the fact that h(P1 , . . . , Pn ) commutes with all the Pi ’s, we know that h we has the form
h we ( p1 , . . . , pn ) = h we ( p), that is, is a function of the classical harmonic oscillators pi := ξi2 + xi2 . To see how this h we is related to the h above we note that H is diagonal on the Hermite basis h j . Therefore
x+y xξ d xdξ 1 h(( j + )) =< h j , H h j >= h we ( )2 + ξ 2 ei h j (x)h j (ξ ) n/2 . 2 2 We now claim
14
V. Guillemin, T. Paul
Proposition 5.1. Let h be either in the Schwartz class, or a polynomial function. Let ˆ h(s) = (2π1 )n h( p)e−is. p dp be the Fourier transform of h. Then
2i tan(s/2). p we ˆ h ( p) = h(s)e (s)ds, (5.1) n where tan(s/2). p stands for i tan(si /2) pi and (s) = i=1 (1 − 2i tan(si /2)), and where (5.1) has to be interpreted in the sense of distribution, that is, for each ϕ in the Schwartz class of R,
2i tan(s/2). p ˆ h we ( p)ϕ( p)dp = h(s)e (s)dsϕ( p)dp
2i tan(s/2) ˆ = h(s)(s)ϕˆ ds. Finally, as → 0, h we ∼ h +
∞
cl 2l .
(5.2)
l=1
is.P ds, where eis.P is a zeroth order semiclassical pseudoˆ Proof. Let h(P) = h(s)e differential operator whose Weyl symbol will be computed from its Wick symbol (see n eisi .Pi , it is 5.5 below for the definition). Let us first remark that since eis.P = i=1 enough to prove the theorem in the one-dimensional case. Let ϕxξ be a coherent state at (x, ξ ), that is ξy
1
ϕxξ (y) = (π )− 4 ei e− Let z =
ξ√ +i x , z 2
=
ξ √ +i x 2
and z(t) =
ξ(t)+i √ x(t) . 2
(y−x)2 2
.
A straightforward computation gives
2zz −|z|2 −|z |2 2 ϕxξ , ϕx ξ = e .
(5.3)
Moreover decomposing ϕxξ on the Hermite basis leads to eis P ϕxξ = ei 2 ϕx(s)ξ(s) , s
(5.4)
−2 ∂x2 +x 2
where P = and z(t) = eit z. 2 The Wick symbol of eis P is defined as
σ wi (eis P )(x, ξ ) := ϕxξ , eis P ϕxξ
(5.5)
which, by (5.3) and (5.4), is equal to −is
e
− 1−e
x 2 +ξ 2 2
+i 2s
.
Moreover, using the Weyl quantization formula, it is immediate to see that the Weyl and Wick symbols are related by
where = −
∂2 ∂x2
∂2 + ∂ξ 2 .
σ wi = e−
4
σ we ,
Semiclassical Trace Invariants
15
It is a standard fact that the Wick symbol determines the operator: indeed the function
−2zz +|z|2 +|z |2 2
e (ϕxξ , eis P ϕx ξ ) obviously determines eis P . Moreover it is easily seen to be analytic in z and z . Therefore it is determined by its values on the diagonal z = z i.e., precisely, the Wick symbol of eis P . A straightforward calculation shows that, for (2k+1)π s , k ∈ Z, 2 2 = (1 − 2i tan(s/2))e
(2k+1)π , 2
This shows that, for 2s = σ
we
(e
− 4
e
2i tan(s/2)
x 2 +ξ 2 2
−is
=e
x 2 +ξ 2 2
+i 2s
.
(5.6)
k ∈ Z, we have
)( p) = (1 − 2i tan(s/2))e
is P
− 1−e
2i tan(s/2)
x 2 +ξ 2 2
.
Let us now take ϕ in the Schwartz class of R, and let Bϕ be the operator of (total) Weyl symbol ϕ( x
2 +ξ 2
2
). Let
f (s) := 2π
σ we (eis P )( p)ϕ( p) pdp = Trace[eis P Bϕ ].
Lemma 5.2. f ∈ C ∞ (R). Proof. By metaplectic invariance we know that Bϕ is diagonal on the Hermite basis. Therefore, ∀k ∈ N, (−i)k
1 dk 1 f (s) := Trace[eis P P k Bϕ ] = < h j , Bϕ h j > (( j + ))k eis( j+ 2 ). k ds 2
Since h j is microlocalized on the circle of radius ( j + 21 ) and ϕ is in the Schwartz class, the sum is absolutely convergent for each k.
2i tan(s/2)
x 2 +ξ 2 2
Therefore f (s) = 2π (1 − 2i tan(s/2))e ϕ( p) pdp and (5.6) is valid in the sense of distribution (in the variable p) for all s ∈ R. This expression gives (5.1) immediately for h in the Schwartz class. When h is a polynomial function it is straightforward to check that, since hˆ is a sum of derivatives of the Dirac mass and eis P is a Weyl operator whose symbol is C ∞ with respect to s, the formula also holds in this case. The asymptotic expansion (5.2) is obtained by expanding e
2itg(s/2)
x 2 +ξ 2 2
near eis
x 2 +ξ 2 2
.
Formula (5.1) shows clearly that h we depends only on the 2π periodization of ˆh(s)ei s2 , therefore Corollary 5.3. h we depends only on the values h (k + 21 ) , k ∈ N.
16
V. Guillemin, T. Paul
We mention one application of formula (5.1). Let us suppose first that we have computed the quantum normal form at order K , that is h K ( p) = ck p k := ck p1k1 . . . pnkn , |k|=k1 +···kn ≤K
|k|=k1 +···kn ≤K
and let us define h we K as the Weyl symbol of h K (P). Corollary 5.4. h we K ( p) =
ck
|k|=k1 +···kn ≤K
:=
|k|=k1 +···kn ≤K
ck
2i tan(s/2) p ∂K (s)e |s=0 ∂sk ∂K ∂sk11 . . . ∂ kn sn
(s)e2i
tan(s1 /2) p1 +···+tan(sn /2) pn
|s=0 .
Let us come back now to the comparison between the two constructions of Sects. 2 and 3. Clearly the “θ ” part doesn’t play any role, as the Weyl quantization of any function f (τ ) is exactly f (Dθ ). Therefore we have the following Theorem 5.5. The functions H of Theorem 3.1 and h of Theorem 4.8 are related by the formula
2i tan(s/2). p ˆ Dt , )e H (P1 , . . . , Pn , Dt , ) = h(s, (s)ds, where hˆ is the Fourier transform of h with respect to the variables pi . In particular H − h = O(2 ). Proof. The proof follows immediately from Proposition 5.1, and the unicity of the (quantum) Birkhoff normal form.
6. The Computation of the Semiclassical Birkhoff Canonical Form from the Asymptotics of the Trace Formula In this section we will abandon the quantum approach to Birkhoff canonical forms developed in Sects. 4–5 and revert to the symbolic approach of Sects. 2–3. Using this approach we will prove that the wave trace data coming from the Gutzwiller formula determine the Quantum Birkhoff canonical form constructed in Sect. 3. Our goal will be by “mimicking” (with semiclassical modifications) the proof of this result by Zelditch in [11,12] and in particular avoid the method of “Grushin reduction” used in [7] to equate the trace formula of [6,8,9] with the trace formula for a monodromy operator. Warning. The Aϕ in display 6.3 below is not the family of U L ’s figuring in Theorem 4.8 but is the “symbolic” Aϕ figuring in Theorem 3.1. In particular the estimates in Theorems 4.8 and 4.9 will not play any role in this proof. Let X and H be as in the Introduction. Let γ be a periodic trajectory of the vector field (2.1) of period 2π .
Semiclassical Trace Invariants
17
For l ∈ Z let ψl be a Schwartz function on the real line whose Fourier transform ψˆl is supported in a neighborhood of 2πl containing no other period The semiclassical of (2.1). of the form: trace formula gives an asymptotic expansion for Trace ψl H −E
Trace ψl
H−E
∼
∞
dlm m ,
(6.1)
m=0
where the dl ’s are distributions acting on ψˆl with support concentrated at {2πl}. We will show that the knowledge of the dl s determine the quantum semiclassical Birkhoff form of Sect. 2, and therefore the classical one. Let us first rewrite the l.h.s of (6.1) as
it H −E ˆ Trace ϕψ(t)e dt . (6.2) Since ψˆ is supported near a single period of (2.1) we know from the general theory of Fourier integral operators that one can microlocalize (6.1) near γ . Therefore we can conjugate (6.2) by the semiclassical Fourier integral operator Aϕ of Theorem 3.1. This leads to the computation of
it H −E −1 ˆ Trace Aϕ ψ(t)e dt Aϕ
H (P1 ,...,Pn ,Dt ,)+H −E it ˆ = Tr ψ(t)ρ(P1 , . . . , Pn , Dt )e dt , (6.3) where ρ ∈ C0∞ (Rn+1 ) with ρ = 1 in a neighborhood of p = τ = 0 and Tr stands for the Trace in L 2 (Rn × S 1 ). Let us note that, as is standard in the proof on trace formulas, by the independence condition of the Poincaré angles (see footnote (2)), γ is isolated on its energy shell {H = E}. By standard stationary phase techniques this is enough to show that the contribution of H in (6.3) is of order O(∞ ). Let us write H (P1 , . . . , Pn , Dt , ) as E + Dt + θi Pi + cr,s ()P r Dts . (6.4) r ∈Nn ,s∈Z
We will first prove l (t, θ ) be the function defined by Proposition 6.1. Let gr,s it θ1 +···+θn 2 e ∂ s ∂ r l ˆ gr,s (t, θ ) = −i −i t ψ(t) . t∂θ ∂t i (1 − eitθi )
(6.5)
Let us fix l ∈ Z. Then the knowledge of all the dlm s for m < M in (6.1) determines the following quantities: l cr,s ()gr,s (2πl, θ ) (6.6) |r |+s=m
for all m < M.
18
V. Guillemin, T. Paul
Proof. The r.h.s. of (6.3) can be computed thanks to (6.4) using 1 spectrum Pi = {(µi + ), µi ∈ N}, 2 spectrum Dt = {ν, n ∈ Z}. Thus the r.h.s of (6.3) can be written as
∞ k 1 it ν+θ.(µ+ 12 ) (it) ψˆl (t) ρ (µ + ), ν e 2 k! µ,ν k=0 k 1 r s |µ|+s−1 × cr,s () µ + ν dt, 2 r,s
(6.7)
since the support of ψˆ l contains only one period, and therefore the trace can be microlocalized infinitely close to the periodic trajectory, making the role of H inessential. Using the following remark of S. Zelditch:
1 µ+ 2
r
r
∂ ν = −i t∂θ s
∂ −i ∂t
s e
it ν+θ.(µ+ 12 )
we get, mod(∞ ),
ψˆl (t)
∞ (it)k k=0
k!
r,s
r s k ∂ ∂ it ν+θ.(µ+ 12 ) |µ|+s−1 −i cr,s () −i e dt. t∂θ ∂t µ,ν
(6.8) Since
ν∈Z
eitν
= 2π
l δ(t − 2πl) , and
µ∈Nn
e
itθ. µ+ 21
=
θ1 +···+θn
eit 2 , i 1−eitθi
together
with the fact that ψˆ is supported near 2πl, we get that (6.8) is equal to ⎡ 2π ⎣
∞ (i)k k=0
k!
×
t k ψˆl (t)
|µ|+s−1
r,s θ1 +···+θn
∂ cr,s () −i t∂θ
eit 2 i 1 − eitθi
.
r k ∂ s −i ∂t (6.9)
t=2πl
Rearranging terms in increasing powers of shows that the quantities (6.6) can be computed recursively.
The fact that one can compute the cr,s () from the quantities (6.6) is an easy consequence of the rational independence of the θi s and the Kronecker theorem, and is exactly the same as in [5].
Semiclassical Trace Invariants
19
References 1. Bridges, T.J., Cushman, R.H., MacKay, R.S.: Dynamics near an irrational collision of eigenvalues for symplectic mappings. Lamgford, W.F. (ed.) In: Normal Forms and Homoclinic Chaos. Fields Inst. Commun. 4, Providence, RI: Amer. Math. Soc., 1995, pp. 61–79 2. Chazarain, J.: Formule de Poisson pour les variétés Riemanniennes. Invent. Math. 24, 65–82 (1974) 3. Colin de Verdière, Y.: Spectre du Laplacien et longueurs des géodésiques périodiques. Compos. Math. 27, 83–106 (1973) 4. Duistermaat, J.J., Guillemin, V.: The spectrum of positive elliptic operators and periodic bicharacteristics. Inv. Math. 29, 39–79 (1975) 5. Guillemin, V.: Wave-trace invariants. Duke Math. J. 83, 287–352 (1976) 6. Gutzwiller, M.: Periodic orbits and classical quantization conditions. J. Math. Phys. 12, 343–358 (1971) 7. Iantchenko, A., Sjöstrand, J., Zworski, M.: Birkhoff normal forms in semi-classical inverse problems. Math. Res. Lett. 9, 337–362 (2002) 8. Paul, T., Uribe, A.: Sur la formule semi-classique des traces. C.R. Acad. Sci Paris 313, I, 217–222 (1991) 9. Paul, T., Uribe, A.: The semi-classical trace formula and propagation of wave packets. J. Funct. Anal. 132, 192–249 (1995) 10. Robert, D.: Autour de l’approximation semi-classique. Basel-Boston: Birkhäuser, 1987 11. Zelditch, S.: Wave invariants at elliptic closed geodesics. Geom. Funct. Anal. 7, 145–213 (1997) 12. Zelditch, S.: Wave invariants for non-degenerate closed geodesics. Geom. Funct. Anal. 8, 179–217 (1998) Communicated by B. Simon
Commun. Math. Phys. 294, 21–60 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0899-9
Communications in
Mathematical Physics
On Dilation Symmetries Arising from Scaling Limits Henning Bostelmann1,2, , Claudio D’Antoni2, , Gerardo Morsella2, 1 University of York, Department of Mathematics, Heslington, York YO10 5DD, United Kingdom.
E-mail:
[email protected]
2 Università di Roma “Tor Vergata”, Dipartimento di Matematica, Via della Ricerca Scientifica,
I-00133 Roma, Italy. E-mail:
[email protected],
[email protected] Received: 24 January 2009 / Accepted: 18 March 2009 Published online: 30 August 2009 – © Springer-Verlag 2009
Abstract: Quantum field theories, at short scales, can be approximated by a scaling limit theory. In this approximation, an additional symmetry is gained, namely dilation covariance. To understand the structure of this dilation symmetry, we investigate it in a nonperturbative, model independent context. To that end, it turns out to be necessary to consider non-pure vacuum states in the limit. These can be decomposed into an integral of pure states; we investigate how the symmetries and observables of the theory behave under this decomposition. In particular, we consider several natural conditions of increasing strength that yield restrictions on the decomposed dilation symmetry. 1. Introduction In the analysis of quantum field theories, the notion of scaling limits plays an important role. The physical picture underlying this mathematical concept is as follows: One considers measurements in smaller and smaller space-time regions, at the same time increasing the energy content of the states involved, so that the characteristic action scale remains constant. Passing to the limit of infinitesimal scales, one obtains a new quantum field theory, the scaling limit of the original model. The scaling limit theory can be seen as an approximation of the full theory in the short-distance regime. However, it may differ significantly from the full theory in fundamental aspects, for example regarding its charge structure: In quantum chromodynamics, it is expected that confined charges (color) appear in the limit theory, but are not visible as such in the full theory. The virtues of the scaling limit theory include that it is typically simpler than the original one. In fact, in relevant examples, one expects it to be interactionless (asymptotic freedom). But even where this is not the case, the limit theory should possess Supported in part by the EU network “Noncommutative Geometry” (MRTN-CT-2006-0031962) Supported in part by PRIN-MIUR and GNAMPA-INDAM Supported in part by PRIN-MIUR, GNAMPA-INDAM and the Scuola Normale Superiore
22
H. Bostelmann, C. D’Antoni, G. Morsella
an additional symmetry: It should be dilation covariant, since any finite masses in the original model can be neglected in the limit of large energies. On the mathematical side, a very natural description of scaling limits has been given by Buchholz and Verch [BV95]. This description, formulated in the C∗ algebraic framework of local quantum physics [Haa96], originates directly from the physical notions, and avoids any additional input motivated merely on the technical side, such as a rescaling of coupling constants or mass parameters, or the choice of renormalization factors for quantum fields. This has the advantage of allowing an intrinsic, model-independent description of the short distance properties of the theory at hand. In particular, it has been successfully applied to the analysis of the charge structure of the theory in the scaling limit and to the intrinsic characterization of charge confinement [Buc96b,DMV04]. While the framework of Buchholz and Verch seems rather abstract at first, it has recently been shown that it reproduces the usual picture of multiplicative field renormalization in typical cases [BDM09]. The approach of [BV95] is based on the notion of the scaling algebra A, which consists—roughly speaking—of sequences of observables λ → Aλ at varying scale λ, uniformly bounded in norm, and subject to certain continuity conditions. (We shall recall the precise definition in Sec. 2.1.) The task of passing to the scaling limit is then reduced to finding a suitable state ω0 on the C∗ algebra A that represents the vacuum of the limit theory; it is constructed as a limit of vacuum states at finite scales. The limit theory itself is then obtained by a standard GNS construction with respect to ω0 . It should be easy in this context to describe the additional dilation symmetry that arises in the scaling limit. In fact, the scaling algebra A carries a very natural representation µ → δ µ of the dilation group, which acts by shifting the argument of the functions λ → Aλ : δ µ (A)λ = Aµλ . However, things turn out to be more involved: The limit states ω0 described in [BV95] are not invariant under this group action, and thus one does not obtain a canonical group representation in the limit Hilbert space. In [BDM09], generalized limit states have been introduced, some of which are invariant under dilations, and give rise to a unitary implementation of the dilation group in the limit theory. However, these dilation invariant limit states are never pure; rather they arise as a mixture of states of the Buchholz-Verch type, which are pure in 2+1 or more space-time dimensions. The object of the present paper is to analyze this generalized class of limit states in more detail, in order to describe the structure of the dilation symmetry associated with the dilation invariant ones. In particular we will show that, as briefly mentioned in [BDM09], the decomposition of these states in pure (Buchholz-Verch type) states gives rise to a direct integral decomposition of the limit Hilbert space, which also induces a decomposition of observables and of Poincaré symmetries. It should be noted here that the entire construction is complicated by the fact that uncountably many extremal states are involved in this decomposition, and that the measure space underlying the direct integral is of a very general nature. Because of this, we need to use a notion of direct integral of Hilbert spaces which is more general than the one previously employed in the quantum field theory literature [DS85]. It is also of interest to discuss how the special but physically important class of theories with a unique scaling limit, as defined in [BV95], fits into our generalized framework. It turns out that, up to some technical conditions, uniqueness of the scaling limit in the Buchholz-Verch framework is equivalent to the factorization of our generalized scaling limit into a tensor product of an irreducible scaling limit theory and a commutative part, which is just the image under the scaling limit representation of the center of the
On Dilation Symmetries Arising from Scaling Limits
23
scaling algebra. In particular we show that such factorization holds for a restricted class of theories, those with a convergent scaling limit. This class includes in particular dilation invariant theories and free field models. The technical conditions referred to above consist in a suitable separability requirement of the scaling limit Hilbert space, which is needed in order to be able to employ the full power of direct integrals theory. As a matter of fact, such separability condition is a consequence of a refined version of the Haag-Swieca compactness condition. With these results at hand, it is possible to discuss the structure of the unitarily implemented dilation symmetry in dilation invariant scaling limit states. The outcome is that in general the dilations do not decompose, not even in the factorizing situation. Rather, the dilations intertwine in a suitable sense the different pure limit states that occur in the direct integral decomposition. A complete factorization of the dilation symmetry is however obtained in the convergent scaling limit case. For such theories, therefore, one gets a unitary implementation of the dilation symmetry in the pure limit theory. The remainder of this paper is organized as follows: First, in Sec. 2, we recall the notion of scaling limits in the algebraic approach to quantum field theory, and generalize some fundamental results of [BV95] to our situation. In Sec. 3, we establish the direct integral decomposition mentioned above, including a decomposition of local observables and Poincaré symmetries. Section 4 contains a discussion of unique scaling limits as a special case. We define several conditions that generalize the notion from [BV95], and discuss relations between them. Then, in Sec. 5, we analyze the structure of dilation symmetries in the limit Hilbert space, and their decomposition along the direct integral, on different levels of generality. In Sec. 6 we propose a stronger version of the HaagSwieca compactness condition and we show that it implies the separability property used in the analysis of Sec. 4. Section 7 discusses some simple models as examples, showing in particular that these fulfill all of our conditions proposed in Sec. 4 and 6. We conclude with a brief outlook in Sec. 8. The Appendix reviews the concept of direct integrals of Hilbert spaces, which we need in a more general variant than covered in the standard literature. 2. Definitions and General Results We shall first recall the definition of scaling limits in the algebraic approach to quantum field theory, and prove some fundamental results regarding uniqueness of the limit vacuum state and geometric modular action. 2.1. The setting. We consider quantum field theory on (s + 1) dimensional Minkowski space. For our analysis, we work entirely within the framework of algebraic quantum field theory [Haa96], where observables localized in an open bounded subset O ⊂ Rs+1 of spacetime are described by the selfadjoint elements of a C∗ algebra A(O) in such a way that if O1 ⊂ O2 then A(O1 ) ⊂ A(O2 ). The correspondence A : O → A(O) thus defined is called a net of algebras. The inductive limit C∗ algebra of O → A(O) is denoted again by A and is called the quasilocal algebra. Let us repeat the formal definition of a quantum field theoretical model in this context. Definition 2.1. Let G be a Lie group of point transformations of Minkowski space that includes the translation group. A local net of algebras with symmetry group G is a net of algebras A together with a representation g → αg of G as automorphisms of A, such that
24
H. Bostelmann, C. D’Antoni, G. Morsella
(i) [A1 , A2 ] = 0 if O1 , O2 are two spacelike separated regions, and Ai ∈ A(Oi ); (ii) αg A(O) = A(g.O) for all O, g. We call A a net in a positive energy representation if, in addition, the A(O) are W ∗ algebras acting on a common Hilbert space H, and (iii) there is a strongly continuous unitary representation g → U (g) of G on H such that αg = ad U (g); (iv) the joint spectrum of the generators of translations U (x) lies in the closed forward light cone V¯ + ; (v) there exists a vector Ω ∈ H which is invariant under all U (g) and cyclic for A. We call A a net in the vacuum sector if, in addition, (vi) the vector Ω is unique (up to scalar factors) as an invariant vector for the translation group. Our approach is to start from a local net A in the vacuum sector, with the Poincaré ↑ group P+ as its symmetry group; this net A will be kept fixed in all that follows. Our aim is to describe the short-distance scaling limit of A. Following [BV95], we define B to be the set of bounded functions B : R+ → B(H), λ → B λ . Equipped with pointwise addition, multiplication, and ∗ operation, and with the norm B = supλ B λ , the set B becomes a C∗ algebra. Let G be the group formed by Poincaré transformations and dilations; we will write G g = (µ, x, Λ) with µ ∈ R+ , x ∈ Rs+1 , and Λ a Lorentz matrix. G acts on B via a representation α, given by (α g B)λ = αλµx,Λ (B λµ ) for g = (µ, x, Λ) ∈ G, B ∈ B,
(2.1)
where α is the Poincaré group representation on A. Note the rescaling of translations with the scale parameter λ. We now define new local algebras as subsets of B: A(O) := A ∈ B | Aλ ∈ A(λO) for all λ > 0; g → α g (A) is norm continuous . (2.2) This is a net of local algebras in the sense of Def. 2.1, with the enlarged symmetry group G [BDM09]. We denote by A the associated quasilocal algebra, i.e. the inductive limit of A(O) as O Rs+1 . This A is called the scaling algebra. Note that A has a large center Z(A), consisting of all operators A of the form Aλ = f (λ)1, where f : R+ → C is a bounded uniformly continuous function on R+ as a group under multiplication. We often identify A ∈ Z(A) with the function f without further notice. For a description of the scaling limit, we first consider states on Z(A). Let m be a mean on the bounded uniformly continuous functions1 on R+ , i.e., a positive normalized linear functional on the commutative C∗ algebra Z(A). We say that m is asymptotic if m( f ) = limλ→0 f (λ) whenever the limit on the right-hand side exists; or, equivalently, if m( f ) = 0 whenever f (λ) = 0 for small λ. Asymptotic means are, in this sense, generalizations of the limit λ → 0. Further we consider two important classes of means: (i) m is called multiplicative if m( f g) = m( f )m(g) for all functions f, g. (ii) m is called invariant if m( f µ ) = m( f ) for all functions f and all µ > 0, where f µ = f (µ · ). 1 In contrast to [BDM09], we do not consider means on the bounded functions on R , but rather on the + bounded uniformly continuous functions. While all of them can be extended to the bounded functions, these extensions do not play a role in our current investigation.
On Dilation Symmetries Arising from Scaling Limits
25
It is an important fact that (i) and (ii) are mutually exclusive; there are no multiplicative invariant means in our situation (cf. [Mit66]). We now extend these “generalized limits” of functions to a limit of operator sequences, using a projection technique. Let ω = (Ω| · |Ω) be the vacuum state of A. This state induces a projector (or conditional expectation) in A onto Z(A), which we denote by the same symbol: ω : A → Z(A), (ω(A))λ = ω(Aλ )1.
(2.3)
Using this projector, any mean m defines a state ωm on A by ωm := m ◦ ω. If here m is asymptotic, we call ωm a limit state, and typically denote it by ω0 . These are the states that correspond to scaling limits of the quantum field theory. Since there is a one-to-one correspondence between asymptotic means and limit states, we will usually work with the state ω0 only, and not refer to the mean m explicitly. A limit state ω0 will be called multiplicative2 or invariant if the corresponding mean has this property. Multiplicative limit states correspond to those considered by Buchholz and Verch in [BV95]. Every other limit state arises from these by convex combinations and weak∗ limits; this follows directly from the property of states on the commutative algebra Z(A). Given a limit state ω0 , we can obtain the limit theory via a GNS construction: Let π0 be the GNS representation of A with respect to ω0 , and H0 the representation space, with GNS vector Ω0 . Denoting by G0 the subgroup of G under which ω0 is invariant, we canonically obtain a strongly continuous unitary representation of G0 on H0 by setting U0 (g)π0 (A)Ω0 := π0 (α g (A))Ω0 , g ∈ G0 . The subgroup G0 contains the Poincaré group; and if ω0 is invariant, then G0 = G. The translation part of U0 fulfills the spectrum condition [BDM09]. Setting A0 (O) := π0 (A(O)) , one obtains a local net A0 with symmetry group G0 in a positive energy representation: the limit theory. 2.2. Multiplicity of the vacuum state. If ω0 is a multiplicative limit state, its restriction to Z(A) is pure. It has been shown in [BV95] that in the case s ≥ 2, this property extends to the entire theory: ω0 is a pure vacuum state on A, and π0 is an irreducible representation. On the other hand, if ω0 is not multiplicative, the same must be false, since already ω0 Z(A) is non-pure. However, we shall show that this property of the center is the only “source” of reducibility: namely one has π0 (A) = π0 (Z(A)) . We need some preparations to prove this. In the following, set Z0 := π0 (Z(A)) , and let HZ := clos(Z0 Ω0 ) ⊂ H0 be the representation space of the commutative algebra. Lemma 2.2. Let PZ ∈ B(H0 ) be the orthogonal projector onto HZ. If s ≥ 2, then PZ ∈ π0 (A) , and HZ is the space of all translation-invariant vectors in H0 . Proof. As a consequence of the spectrum condition in the theory A0 , it is known [Ara64] that the translation operators U0 (x) are contained in π0 (A) . Now let U∞ be an ultraweak cluster point of U0 (x) as x goes to spacelike infinity on some fixed sequence within the time-0 plane. (Such cluster points exist by the Alaoglu-Bourbaki theorem.) Then U∞ ∈ π0 (A) ; we will show U∞ = PZ. To that end, we first note that ω0 (A B) = ω0 (ω(A) B) for all A ∈ A, B ∈ Z(A),
(2.4)
2 For clarity, we note that a multiplicative limit state, by this definition, is not a multiplicative functional on A, but is multiplicative only on the center Z(A).
26
H. Bostelmann, C. D’Antoni, G. Morsella
which follows directly from the definition of ω0 . Now we make use of the cluster property of the vacuum at finite scales. As in [BV95, Lemma 4.3], one can obtain the following norm estimate in the algebra A: ω(A α x B) − ω(A)ω(B) ≤ c
rs ˙ + AB ˙ A B |x|s−1
(2.5)
for fixed r > 0, x in the time-0 plane with |x| > 3r , and for A, B chosen from some norm-dense subset of A(Or ), with Or being the standard double cone of radius r around the origin. Here c > 0 is some constant, and the dot denotes the time derivative. This implies that as |x| → ∞, lim ω0 (A α x B) = ω0 (ω(A)ω(B)) x
(2.6)
for these A, B. Now it follows from Eq. (2.4) – with ω(B) in place of B – that (π0 (A)Ω0 |U∞ π0 (B)Ω0 ) = lim ω0 (A∗ α x B) = (π0 (A)Ω0 |π0 (ω(B))Ω0 ). (2.7) x
Continuing this relation from the dense sets chosen, this means U∞ π0 (B)Ω0 = π0 (ω(B))Ω0 for all B ∈ A.
(2.8)
2 = U , and img U This shows that U∞ ∞ ∞ = HZ. Also, again applying Eq. (2.4), one ∗ = U . Thus U is the unique orthogonal projector onto H . For the last obtains U∞ ∞ ∞ Z part, note that translations act trivially on HZ, and that U∞ leaves all translation-invariant vectors unchanged; so HZ is the space of all translation-invariant vectors.
We are now ready to prove the announced result about the commutant of π0 (A). Theorem 2.3. Let s ≥ 2. Let ω0 be a limit state, and let π0 be the corresponding GNS representation. Then π0 (A) = π0 (Z(A)) . Proof. Let B ∈ π0 (A) . By Lemma 2.2, B commutes with PZ; hence BHZ ⊂ HZ, and B HZ ∈ B(HZ) is well-defined. As Z0 ⊂ π0 (A) , we know that [B HZ, C HZ] = 0 for all C ∈ Z0
(2.9)
as an equation in B(HZ). Since Z0 HZ is a maximal abelian algebra in B(HZ) [BR79, Lemma 4.3.15], there exists C ∈ Z0 with B HZ = C HZ. Now for any A ∈ π0 (A) , we can compute B AΩ0 = ABΩ0 = ACΩ0 = C AΩ0 . Since Ω0 is cyclic for π0 , this implies B = C. Thus π0 is trivial.
(A)
(2.10)
⊂ Z0 . The reverse inclusion
It should be noted that the same theorem does not hold in 1+1 space-time dimensions. In this case, it is known even in free field theory [BV98, Sec. 4] that the algebra π0 (A) has a large center, even if π0 (Z(A)) = C1. We can now easily reproduce the known results for multiplicative limit states. In this case, the GNS representation of the abelian algebra Z(A) for the state ω0 must be irreducible; thus Z0 = C1, and dim HZ = 1. The above results imply: Corollary 2.4. Let ω0 be a multiplicative limit state, and let s ≥ 2. Then Ω0 is unique up to a scalar factor as an invariant vector for the translations U0 (x), and the representation π0 is irreducible. A0 is a net in the vacuum sector in the sense of Def. 2.1.
On Dilation Symmetries Arising from Scaling Limits
27
2.3. Wedge algebras and geometric modular action. While we have defined the scaling limit in terms of local algebras for bounded regions, it is also worthwhile to consider algebras associated with unbounded, in particular wedge-shaped regions. This is particularly important in the context of charge analysis for the limit theory [DMV04,DM06]. While we do not enter this topic here, and do not build on it in the following, we wish to discuss briefly how wedge algebras and the condition of geometric modular action fit into our context. Again, this transfers results of [BV95] to our generalized class of limit states. Let W be a wedge region, i.e. W is a Poincaré transform of the right wedge W+ = −W− = {x ∈ R4 | x · e± < 0}, where e± := (±1, 1, 0, 0).
(2.11)
Note that (W + ) = W− . We introduce the one-parameter group (Λt )t∈R of Lorentz boosts leaving W+ invariant, fixed by Λt e± = exp(±t)e± , and acting as the identity on the edge (e± )⊥ of W+ . Let furthermore j be the inversion with respect to the edge of W+ , i.e. je± = −e± and j = 1 on (e± )⊥ . Note that j 2 = 1 and jW+ = W− . For a local net of algebras (resp. for a net of algebras in a positive energy representation) O → A(O), we define the algebra A(W) associated to the wedge W as the C∗ -algebra (resp. W∗ -algebra) generated by the algebras A(O), where O is any double cone whose closure is contained in W (O ⊂⊂ W in symbols). With these definitions, we can adapt the arguments in [BV95, Lemma 6.1], which do not depend on irreducibility of the net. It is then straightforward to verify that, for a net A in a positive energy representation, the vacuum vector Ω is cyclic and separating for all wedge algebras A(W). This allows us to introduce the notion of geometric modular action. Definition 2.5. Let A be a local net in a positive energy representation, and denote by ∆, J the modular objects associated to A(W+ ), Ω. The net A is said to satisfy the condition of geometric modular action if there holds ∆it = U (Λ2π t ),
t ∈ R, ↑
J U (x, Λ)J = U ( j x, jΛj), (x, Λ) ∈ P+ , J A(O)J = A( jO).
(*)
If A satisfies the condition of geometric modular action, then it also satisfies wedge duality, since, according to Tomita-Takesaki theory and equation (*), A(W+ ) = J A(W+ )J = A(W− ).
(2.12)
This also implies that A satisfies essential Haag duality, i.e. that the dual net Ad of A, defined on double cones O as A(W), (2.13) Ad (O) := W ⊃O
is local and such that A(O) ⊂ Ad (O) for each double cone O. From now on, let A be a net in the vacuum sector, and ω0 a scaling limit state, with π0 the corresponding scaling limit representation. It holds that π0 (A(W)) = A0 (W), since clearly π0 (A(W)) =
O⊂⊂W
π0 (A(O))
·
,
(2.14)
28
H. Bostelmann, C. D’Antoni, G. Morsella
and therefore
π0 (A(W)) =
π0 (A(O)) =
O⊂⊂W
A0 (O) = A0 (W) .
(2.15)
O⊂⊂W
Proposition 2.6. Assume that A satisfies the condition of geometric modular action. Then for each limit state ω0 , the corresponding limit theory A0 also satisfies the condition of geometric modular action. Proof. It’s a straightforward adaptation of the proofs of Lemma 6.2 and Proposition 6.3 of [BV95]. The only point which is worth mentioning is the proof that ω0 is a KMS state (at inverse temperature 2π ) for the algebra A(W+ ) with respect to the one-parameter group of automorphisms (α Λt )t∈R , which goes as follows. Let m be the mean which induces ω0 . Then m is a weak∗ limit of convex combinations of multiplicative means, and therefore ω0 is a weak∗ limit of convex combinations of multiplicative limit states. For such states, the arguments in [BV95, Lemma 6.2] show that they are KMS on A(W+ ), and therefore, the set of KMS states at a fixed inverse temperature being convex and weak∗ closed [BR81, Thm. 5.3.30], this holds also for ω0 . 3. Decomposition Theory Our aim is now to decompose an arbitrary limit state ω0 into “simple” limit states of the Buchholz-Verch type, and to obtain corresponding decompositions of the relevant objects in the limit theory. We start by proving an integral decomposition which is a consequence of standard results. Proposition 3.1. Let ω0 be a limit state. There exists a compact Hausdorff space Z, a regular Borel probability measure ν on Z, and for each z ∈ Z a multiplicative limit state ω z , such that ω0 (A) = dν(z) ω z (A) for all A ∈ A. Z
Further, the map Z(A) → C(Z), C → (z → ω z (C)) is surjective. Proof. Let π0 be the GNS representation of A for ω0 . Consider the C∗ algebra π0 (Z(A)). It is well known that this commutative algebra is isomorphic to C(Z) for a compact Hausdorff space Z, with the isomorphism being given by π0 (C) → (z → ρz (π0 (C))), where the ρz are multiplicative functionals. Now by the Riesz representation theorem, the GNS state (Ω0 | · |Ω0 ) on π0 (Z(A)) ∼ = C(Z) is given by a regular Borel measure ν on Z. Explicitly, one has for all C ∈ Z(A), ω0 (C) = (Ω0 |π0 (C)|Ω0 ) = dν(z) ρz ◦ π0 (C). (3.1) Z
It is clear that ν(Z) = 1. In the above expression, mz := ρz ◦ π0 are multiplicative means; they are asymptotic, since π0 (A) = 0 whenever Aλ vanishes for small λ. Thus, setting ω z = ρz ◦ π0 ◦ ω as usual, we obtain multiplicative limit states ω z on A such that ω0 (A) = dν(z) ω z (A) for all A ∈ A. (3.2) Z
As a last point, the map Z(A) → C(Z), C → (z → ω z (C)) = (z → ρz (π0 (C))) is surjective by construction.
On Dilation Symmetries Arising from Scaling Limits
29
We have thus decomposed a general limit state ω0 into multiplicative limit states ω z . In the case s ≥ 2, this will also be a decomposition into pure states; but the above result does not depend on that. Also, we emphasize that our aim is not a decomposition of the von Neumann algebra π0 (A) along its center; rather we work on the C∗ algebraic side only. We would now like to interpret the above decomposition in the sense of decomposing the limit Hilbert space H0 as a direct integral. This is complicated by the fact that our measure spaces (Z, ν) can be of a very general nature, making the limit Hilbert space nonseparable. In fact, if ω0 is an invariant limit state, one finds that all vectors of the form π0 (C)Ω0 are mutually orthogonal if C λ = χ (λ)1, where χ is a character on R+ . Since there are clearly uncountably many characters—just take χ (λ) = λık with k ∈ R—the limit Hilbert space H0 cannot be separable in this case. The theory of direct integrals of Hilbert spaces in the absence of separability assumptions is nonstandard and only partially complete; we give a brief review in Appendix A. Here we note that the notion of a direct integral over Z, with fiber spaces Hz , crucially depends on the specification of a fundamental family Γ ⊂ z∈Z Hz . This Γ is a vector space with certain extra conditions (see Def. A.1) that serves to define which Hilbert space valued functions are considered measurable. Indeed, using the exact notions, we prove: Theorem 3.2. Let ω0 be a limit state, and Z, ν, ω z as in Proposition 3.1. Let πz , Hz , Ωz be the GNS representation objects corresponding to ω z . Then, Hz Γ := {z → πz (A)Ωz | A ∈ A} ⊂ z∈Z
is a fundamental family. With respect to this family, it holds that Γ ∼ H0 = dν(z) Hz , Z
where the isomorphism is given by π0 (A)Ω0 →
Γ
Z
dν(z) πz (A)Ωz , A ∈ A.
Proof. It is clear that Γ is a linear space; and per Prop. 3.1, the function z → πz (A)Ωz 2 = ω z (A∗ A) is integrable for any A ∈ A. Thus Γ is a fundamental family per Definition A.1. The map W : H0 → Hz , π0 (A)Ω0 → (z → πz (A)Ωz ) (3.3) z∈Z
is clearly linear and isometric when A ranges through A; thus W can in fact be extended to a well-defined isometric map from H0 into Γ¯ . It remains to show that W is surjective. In fact, since L ∞ (Z) · Γ is total in the direct integral space, it suffices to show that all vectors of the form Γ dν(z) f (z) πz (A)Ωz , f ∈ L ∞ (Z), A ∈ A, (3.4) Z
can be approximated in norm with vectors of the form W π0 (B)Ω0 , B ∈ A.
30
H. Bostelmann, C. D’Antoni, G. Morsella
To that end, let f ∈ L ∞ (Z) and A ∈ A be fixed. We first note that, as a simple consequence of Lusin’s theorem, there exist functions f n ∈ C(Z) such that f n ∞ ≤ f ∞ and limn→∞ f n (z) = f (z) for almost every z ∈ Z. On the other hand, per Proposition 3.1 there exist C n ∈ Z(A) such that ω z (C n ) = f n (z) for all z ∈ Z, which implies πz (C n ) = f n (z)1. Therefore we have lim
n→∞ Z
( f (z)πz (A) − πz (C n A))Ωz 2 dν(z) = 0
by an application of the dominated convergence theorem.
(3.5)
In the following, we will usually not denote the above isomorphism explicitly, but rather identify H0 with its direct integral representation. In this way, the subspace HZ ⊂ H0 is isomorphic to the function space L 2 (Z, ν), where f ∈ L 2 (Z, ν) is identified
Γ with Z dν(z) f (z)Ωz ∈ H0 . The next corollary follows directly from the proof above, since a decomposition of operators needs to be checked on the fundamental family only (Lemma A.2). Corollary 3.3. With respect to the direct integral decomposition in Theorem 3.2, all
Γ operators π0 (A), A ∈ A are decomposable, and one has π0 = Z dν(z)πz . If A ∈ Z(A),
Γ then π0 (A) is diagonal, with π0 (A) = Z dν(z) ω z (A)1. Finally, we remark that Lorentz symmetries U0 (x, Λ) in the limit theory are decomposable. ↑
Proposition 3.4. Let g → Uz (g) be the implementation of P+ on the limit Hilbert space Hz corresponding to ω z . Then, one has U0 (g) =
Γ
Z
↑
dν(z)Uz (g) for all g ∈ P+ .
Proof. Again, it suffices to verify this on vectors from Γ . With W being the isomorphism ↑ introduced in the proof of Theorem 3.2, one obtains for all g ∈ P+ and A ∈ A, W U0 (g)π0 (A)Ω0 = W π0 (α g A)Ω0 = =
Γ
Z
dν(z)Uz (g)πz (A)Ωz .
This proves the proposition.
Γ
Z
dν(z)πz (α g A)Ωz (3.6)
It should be remarked that the same simple structure cannot be expected for dilations, if they exist as a symmetry of the limit. For even if ω0 ◦ α µ = ω0 , the multiplicative limit states ω z cannot be invariant under α µ , not even when restricted to Z(A). Thus, the unitaries U0 (µ) will not commute with π0 (Z(A)), and can therefore not be decomposable. In special situations, there may be a generalized sense in which the dilation unitaries can be decomposed; we will investigate this in more detail in Sec. 5.
On Dilation Symmetries Arising from Scaling Limits
31
4. Unique and Factorizing Scaling Limits The limit theory on the Hilbert space H0 is composed, as discussed in the previous section, of simpler components that live on the “fibre” Hilbert spaces Hz of the direct integral. It is natural to ask whether the theories on these spaces Hz , or more precisely, the nets of algebras Az (O) = πz (A(O)) , are similar or identical in a certain sense. While no models have been explicitly constructed for which the limit theories substantially depend on the choice of a (multiplicative) limit state,3 it does not seem to be excluded that measurable properties, such as the mass spectrum or charge structure of Az , can depend on z. For most applications in physics, however, one expects that the situation is simpler, and that the limit theory does not depend substantially on the choice of ω z . Here it would be much too strict to require that the representations πz are unitarily equivalent. [In fact, for s ≥ 2, the πz are irreducible per Thm. 2.3, and since they do not agree on Z(A), they are even pairwise disjoint.] Rather one can expect that their images, the algebras Az (O), are unique as sets, up to unitaries that identify the different Hilbert spaces Hz ; see Def. 4.1 below. This is the situation of a unique scaling limit in the sense of Buchholz and Verch. In the present section, we want to elaborate how the situation of unique scaling limits, originally formulated for multiplicative limit states, fits into our generalized context. To that end, we will formulate several conditions on the limit theory that roughly correspond to unique limits, and discuss their mutual dependencies. 4.1. Definitions. We shall first motivate and define the conditions to be considered; the proofs of their interrelations are deferred to sections further below. We start by recalling the condition of a unique scaling limit in the sense of [BV95], with some slight modifications. Definition 4.1. The theory A is said to have a unique scaling limit if there exists a local Poincaré covariant net (Au , Hu , Ωu , Uu ) in the vacuum sector such that the following holds. For every multiplicative limit state ω0 , there exists a unitary V : H0 → Hu such ↑ that V Ω0 = Ωu , V U0 (g)V ∗ = Uu (g) for all g ∈ P+ , and V A0 (O)V ∗ ⊂ Au (O) for all open bounded regions O. This includes the aspect of a “unique vacuum structure”. Compared with [BV95], we have somewhat weakened the condition, since we require only inclusion of V A0 (O)V ∗ in Au (O), not equality. This is for the following reason. Supposing that both A and Au fulfill the condition of geometric modular action (Definition 2.5), such that the Haagdualized nets of A0 and Au are well-defined, our condition precisely implies that these dualized nets agree for all multiplicative limit states. Since for many applications, particularly charge analysis [DMV04], the dualized limit nets are seen as the fundamental objects, we think that this is a reasonable generalization of the condition. For a general, not necessarily multiplicative limit state ω0 , we obtain a decomposition ω0 = Z dν(z)ω z into multiplicative states, as discussed in Sec. 3, and thus obtain from Def. 4.1 corresponding unitaries Vz for every z. Due to the very general nature of the measure space Z, and due to a possible arbitrariness in the choice of Vz , particularly if Au possesses inner symmetries, an analysis of ω0 seems impossible in this generality. Rather we will often make use of a regularity condition, which is formulated as follows. 3 See however [Buc96a, Sec. 5] for some ideas to that end.
32
H. Bostelmann, C. D’Antoni, G. Morsella
Definition 4.2. Suppose that the theory A has a unique scaling limit. We say that a limit state ω0 is regular if there is a choice of the unitaries Vz such that for any A ∈ A, the function ϕ A : Z → Hu , z → Vz πz (A)Ωz is Lusin measurable [i.e., is contained in L 2 (Z, ν, Hu )]. We shall later give a sufficient condition for the above regularity, which actually implies that the functions ϕ A can be chosen constant in generic cases. Our concepts so far refer to multiplicative limit states mostly. We will now give a generalization of Def. 4.1 that involves generalized limit states directly, and that seems natural in our context. It is based on the picture that the limit Hilbert space should have a tensor product structure, H0 ∼ = HZ ⊗ Hu , where Hu is the unique representation space associated with multiplicative limit states, and HZ is the representation space of Z(A) under π0 . All objects of the theory—local algebras, Poincaré symmetries, and the vacuum vector—should factorize along this tensor product. We now formulate this in detail. Definition 4.3. The theory A is said to have a factorizing scaling limit if there exists a local Poincaré covariant net (Au , Hu , Ωu , Uu ) in the vacuum sector such that the following holds. For every limit state ω0 , there exists a decomposable unitary V :
Γ,⊕ H0 → L 2 (Z, ν, Hu ), V = Z dν(z)Vz with unitaries Vz : Hz → Hu , such that ↑ V Ω0 = ΩZ ⊗ Ωu , V U0 (g)V ∗ = 1 ⊗ Uu (g) for all g ∈ P+ , and Vz Az (O)Vz∗ ⊂ Au (O) for all open bounded regions O and all z ∈ Z. Here ΩZ ∈ HZ denotes the GNS vector of the commutative algebra. The conditions on local algebras are deliberately chosen quite strict. We require Vz Az (O)Vz∗ ⊂ Au (O) ¯ u (O). This serves to for every z, rather than the weaker condition V A0 (O)V ∗ ⊂ Z0 ⊗A avoid countability problems; see Sec. 4.3 for further discussion. In subsequent sections, we will show that the notion of a unique scaling limit and a factorizing scaling limit are cum grano salis identical, up to the extra regularity condition in Definition 4.2 that we have to assume. We also consider a stronger condition, which is easier to check in models. Our ansatz is to require a sufficiently large subset Aconv ⊂ A such that for each A ∈ Aconv , the function λ → ω(Aλ ) is convergent as λ → 0. Consider the following definition: Definition 4.4. The theory A is said to have a convergent scaling limit if there exists an α-invariant C∗ subalgebra Aconv ⊂ A with the following properties: (i) For each A ∈ Aconv , the function λ → ω(Aλ ) converges as λ → 0. (ii) If ω0 is a multiplicative limit state, then π0 (A(O)∩Aconv ) is weakly dense in A0 (O) for every open bounded region O. It follows directly from (ii) that also π0 (Aconv )Ω0 is dense in H0 . The condition roughly says that “convergent scaling functions” are sufficient for describing the limit theory—considering nonconvergent sequences is only required for technical consistency of our formalism, for describing the image of Z(A), which does not directly relate to quantum theory. This is heuristically expected in many physical models: In usual renormalization approaches in formal perturbation theory, the selection of subsequences or filters to enforce convergence seems not to be widespread, and sequences of pointlike fields can be chosen to converge in matrix elements. We will show that the above condition is sufficient for the scaling limit to be unique, and all limit states to be regular. In fact, we shall see later that also the structure of dilations simplifies.
On Dilation Symmetries Arising from Scaling Limits
33
Unique limit
Factorizing limit
Convergent limit
Regularity condition
Fig. 1. Implications between the conditions on the limit theory. Arrows marked with ∗ are only proven under additional separability assumptions
We also mention that this condition has been employed in [CM08] in order to discuss some functoriality properties of the scaling limit with respect to the formation of subsystems of the observable net. Figure 1 summarizes the different conditions we introduced, and shows the implications we briefly mentioned. We will now go ahead and prove that the individual arrows are indeed correct. However, in order to avoid problems with the direct integral spaces involved, we shall make certain separability assumptions in most cases. Let us comment on these. For multiplicative limit states, it seems a reasonable assumption that the limit Hilbert space H0 is separable. This would follow, from example, from the Haag-Swieca compactness condition; cf. [Buc96a]. For general limit states, in particular if these are invariant, H0 cannot be separable since already HZ ∼ = L 2 (Z, ν) is nonseparable, as discussed in Sec. 3. We can however reasonably assume that H0 fulfills a condition which we call uniform separability; cf. Def. A.3 in the Appendix. This means that a countable
Γ set {χ j } ⊂ H0 = Z dν(z)Hz exists such that {χ j (z)} is dense in every Hz . As we shall see in Sec. 6, uniform separability follows from a sharpened version of the Haag-Swieca compactness condition; and we will show in Sec. 7 that this compactness condition is indeed fulfilled in relevant examples. 4.2. Unique limit ⇒ factorizing limit. In the following, we suppose that A has a unique scaling limit. We fix a regular limit state ω0 , and denote the associated objects Z, ν, H0 , π0 , Ω0 , Hz , πz , Ωz , Vz as before. In order to prove that the scaling limit factorizes, we have to construct a unitary V : H0 → L 2 (Z, ν, Hu ) with appropriate properties. In fact,
Γ,⊕ this V is intuitively given by V = Z dν(z)Vz ; and the key question turns out to be whether this V is surjective. We will prove this only under separability assumptions. Proposition 4.5. Let A have a unique scaling limit; let ω0 be a regular limit state; and suppose that H0 is uniformly separable. Then, Γ,⊕ V : H0 → L 2 (Z, ν, Hu ), V = dν(z)Vz Z
defines a unitary operator. Proof. First, it is clear that if H0 is uniformly separable, then all Hz , and in particular Hu , are separable. Hence L 2 (Z, ν, Hu ) is uniformly separable. Now note that V is well-defined precisely by the regularity condition. Further, writing explicitly V π0 (A)Ω0 = z → Vz πz (A)Ωz , A ∈ A, (4.1)
34
H. Bostelmann, C. D’Antoni, G. Morsella
one has
V π0 (A)Ω0 2 = dν(z) Vz πz (A)Ωz 2 = dν(z) πz (A)Ωz 2 = π0 (A)Ω0 2 , Z
Z
(4.2) so V is isometric. It remains to show that V is surjective. To that end, let P be the orthogonal projector onto img V . Since
⊕ V commutes with all diagonal operators, so does P; thus P is decomposable: P = Z dν(z)P(z). Now compute 0 = (1 − P)V =
Γ,⊕
Z
dν(z)(1 − P(z))Vz .
(4.3)
Using uniform separability of both spaces involved, we obtain that (1 − P(z))Vz = 0 a. e. Since the Vz are surjective onto Hu , this means P(z) = 1 a. e. This implies P = 1, so V is surjective. It is clear that V Ω0 = ΩZ ⊗ Ωu ; and we can also verify from the properties of the Vz with respect to Poincaré symmetries that ↑
V U0 (g)V ∗ = 1 ⊗ Uu (g) for all g ∈ P+ .
(4.4)
Also, by the definition of the unique scaling limit, it must hold that Vz Az (O)Vz∗ ⊂ Au (O) for all z. Summarizing the results of this section, we have shown: Theorem 4.6. Suppose that A has a unique scaling limit, that every limit state ω0 is regular, and that the limit spaces H0 are uniformly separable. Then the scaling limit of A is factorizing. 4.3. Factorizing limit ⇒ unique limit. Now reversing the arrow, we start from a theory with factorizing scaling limit, and want to show that the scaling limit is unique in the sense of Buchholz and Verch, and that the limit states are regular. At first glance, this implication seems to be apparent from the definitions. A detailed investigation however reveals some subtleties, which again lead us to making separability assumptions. Theorem 4.7. Assume that A has a factorizing scaling limit. Then the scaling limit is unique. If the space Hu is separable, all limit states ω0 are regular. Proof. It is clear that the scaling limit is unique by Def. 4.1, specializing the conditions of Def. 4.3 to the case where ω0 is multiplicative, and Z consists of a single point. Now
Γ,⊕ let ω0 be a limit state; we need to show it is regular. Let V = Z dν(z)Vz be the unitary guaranteed by Def. 4.3. By definition, the map z → Vz πz (A)Ωz is measurable for any A ∈ A. But we have to show that each Vz fulfills the conditions of Def. 4.1; in fact, we will have to modify the Vz on a null set. First, we have V Ω0 = ΩZ ⊗ Ωu by assumption. On the other hand, V Ω0 =
⊕ Z dν(z) Vz Ωz , so that Vz Ωz = Ωu for z ∈ Z\NΩ , where NΩ is a null set. Next we consider Poincaré transformations. Starting from Def. 4.3, we know that: ↑
V U0 (g)V ∗ = 1 ⊗ Uu (g) for all g ∈ P+ .
(4.5)
On Dilation Symmetries Arising from Scaling Limits
35
Since U0 (g) factorizes by Prop. 3.4, we can rewrite this equation as
⊕
Z
dν(z) Vz Uz (g)Vz∗
=
⊕
Z
dν(z) Uu (g).
(4.6)
Now if Hu is separable, and thus L 2 (Z, ν, Hu ) uniformly separable, we can conclude that Vz Uz (g)Vz∗ = Uu (g) for all z ∈ Z\Ng , with a null set Ng depending on g. We pick ↑ a countable dense subset Pc of P+ , and consider the null set N := NΩ ∪ (∪g∈Pc Ng ). Our results so far are that Vz Ωz = Ωu , Vz Uz (g)Vz∗ = Uu (g) for all z ∈ Z\N , g ∈ Pc .
(4.7)
↑ Indeed, by continuity of the representations, the same holds for all g ∈ P+ . Now let Vˆz be those unitaries obtained by evaluating Def. 4.3 for the multiplicative limit states ω z . We set Vz for z ∈ Z\N , Wz := ˆ (4.8) Vz for z ∈ N .
Γ,⊕ Then we have V = Z dν(z) Wz , and the Wz fulfill the relations in Eq. (4.7) for all ↑ z ∈ Z and g ∈ P+ . As a last point, Wz Az (O)Wz∗ ⊂ Au (O) holds for every z, since both Vz and Vˆz have this property. Thus ω0 is regular. Let us add some comments on the conditions required for Vz in Def. 4.3, regarding Poincaré transformations and local algebras. We could choose stricter conditions on Vz , requiring that ↑
Vz Uz (g)Vz∗ = Uu (g) for all z ∈ Z and g ∈ P+ .
(4.9)
In this case, the countability problem in the proof above does not occur, and Thm. 4.7 holds without the requirement that Hu is separable. On the other hand, it does not seem reasonable to weaken the conditions on Vz with respect to local algebras, requiring only that ¯ u (O) for all O. V A0 (O)V ∗ ⊂ Z0 ⊗A
(4.10)
(We shall show below that this relation is implied by the chosen conditions on Vz .) For if we require only (4.10), and we wish to apply the techniques used in the proof of Thm. 4.7, it becomes necessary not only to require separability of Hu —which seems reasonable for applications in physics—but also separability of the algebras Au (O). That would however be too strict for our purposes, since the local algebras are expected to be isomorphic to the hyperfinite type III1 factor [BDF87]. We now show that Eq. (4.10) follows from Def. 4.3 as given. Proposition 4.8. Let A have a factorizing scaling limit. With V the unitary of Def. 4.3, ¯ u (O) for any bounded open region O. one has V A0 (O)V ∗ ⊂ Z0 ⊗A
36
H. Bostelmann, C. D’Antoni, G. Morsella
Proof. Let A ∈ A(O), and A ∈ Au (O) . We compute the commutator [1 ⊗ A , V π0 (A)V ∗ ] as a direct integral: ⊕ dν(z) [A , Vz πz (A)Vz∗ ]. (4.11) [1 ⊗ A , V π0 (A)V ∗ ] = Z
Now by our requirements on the Vz , we have Vz πz (A)Vz∗ ∈ Au (O) for all z, hence the commutator under the integral vanishes. Since A ∈ A(O) was arbitrary, this means ¯ u (O). V π0 (A(O))V ∗ ⊂ (1 ⊗ Au (O) ) = Z0 ⊗A By weak closure, this inclusion extends to V A0
(O)V ∗ .
(4.12)
4.4. Convergent limit ⇒ unique limit. We now assume that the theory has a convergent scaling limit, and show that our other conditions follow. The main simplification in the convergent case is as follows: For every A ∈ Aconv , the function λ → ω(Aλ ) converges to a finite limit as λ → 0; so all asymptotic means applied to this function yield the same value. Hence the value of ω0 (A) is the same for all limit states ω0 , multiplicative or not. Theorem 4.9. If the scaling limit of A is convergent, then it is unique. If further a multiplicative limit state exists such that the associated limit space H0 is separable, then all limit states are regular, and H0 is uniformly separable for any limit state. Proof. We pick a fixed multiplicative limit state ωu and denote the corresponding representation objects as Hu , πu , Uu , Ωu . Given any other multiplicative limit state ω0 , we define a map V by V : H0 → Hu , π0 (A)Ω0 → πu (A)Ωu for all A ∈ Aconv .
(4.13)
The convergence property of A ∈ Aconv implies π0 (A)Ω0 2 = ω0 (A∗ A) = ωu (A∗ A) = πu (A)Ωu 2 ,
(4.14)
so the linear map V is both well-defined and isometric. It is also densely defined and surjective by assumption (Def. 4.4). Hence V extends to a unitary. Using the α-invariance of ↑ Aconv , one checks by direct computation that V U0 (g)V ∗ = Uu (g) for all g ∈ P+ . Also, ∗ V Ω0 = Ωu is clear. Further, if A ∈ A(O) ∩ Aconv , it is clear that V π0 (A)V = πu (A). By weak density, this means V A0 (O)V ∗ = Au (O). Thus the scaling limit is unique. Now let ω0 not necessarily be multiplicative. Decomposing it into multiplicative states ω z as in Prop. 3.1, the above construction gives us unitaries Vz : Hz → Hu for every z. In fact, the functions z → Vz πz (A)Ωz = πu (A)Ωu are constant for all A ∈ Aconv , in particular measurable. Now let χ ∈ Hu and B ∈ A. We can choose a sequence (An )n∈N in Aconv such that πu (An )Ωu → χ in norm. Noticing that (Vz πz (B)Ωz |χ ) = lim (Vz πz (B)Ωz |πu (An )Ωu ) = lim ω z (B ∗ An ), n→∞
n→∞
(4.15)
we see that the left-hand side, as a function of z, is the pointwise limit of continuous functions, and hence measurable. Thus z → Vz πz (B)Ωz is weakly measurable. Now if Hu was chosen separable, which is possible by assumption, weak measurability implies Lusin measurability of the function (cf. Appendix). Thus ω0 is regular. Finally, in the separable case, we remark that we can pick a countable subset of Acount ⊂ Aconv such that πu (Acount )Ωu is dense in Hu . Then π0 (Acount )Ω0 becomes a fundamental sequence in H0 , so that this space is uniformly separable.
On Dilation Symmetries Arising from Scaling Limits
37
Of course, it follows as a corollary to the preceding sections that the limit is also factorizing. Let us spell this out more explicitly. Proposition 4.10. Suppose that A has a convergent scaling limit, and that there exists a multiplicative limit state ωu for which the representation space Hu is separable. Let ω0 be
Γ,⊕ any scaling limit state. There exists a unitary V = Z dν(z) Vz : H0 → L 2 (Z, ν, Hu ) such that V π0 (A C)Ω0 = π0 (C)ΩZ ⊗ πu (A)Ωu for all A ∈ Aconv , C ∈ Z(A), and such that the Vz fulfill all requirements of Def. 4.3. Proof. We use notation as in the proof of Thm. 4.9. Let Vz : Hz → Hu be the unitaries constructed there. Then, z → Vz∗ is a measurable family of operators. Namely, for any A ∈ Aconv , we find Vz∗ πu (A)Ωu = πz (A)Ωz
(4.16)
which is in Γ ; hence measurability is checked on the fundamental family (cf. Lemma A.2). So the operator ⊕,Γ ∗ V := dν(z) Vz∗ (4.17) Z
is well-defined. Domain and range of V ∗ are both uniformly separable, see Thm. 4.9.
Γ,⊕ Thus also the adjoint of V ∗ , denoted as V , is decomposable with V = Z dν(z) Vz . It is then clear that V is unitary. Also, we have for A ∈ Aconv and C ∈ Z(A), Γ ⊕ dν(z) πz (C)πz (A)Ωz = dν(z) πz (C)Vz πz (A)Ωz V π0 (A C)Ω0 = V Z Z ⊕ = dν(z) πz (C)πu (A)Ωu = (π0 (C)ΩZ) ⊗ (πu (A)Ωu ). (4.18) Z
As a direct consequence of the discussion following Eq. (4.14), the Vz have all the properties required in Def. 4.3 regarding vacuum vector, symmetries, and local algebras. 5. Dilation Covariance in the Limit Our next aim is to analyze the structure of dilation symmetries in the limit theory. To that end, we consider a scaling limit state ω0 which is invariant under δ µ . As shown in [BDM09, Sec. 2], the associated limit theory is covariant with respect to a strongly continuous unitary representation g ∈ G → U0 (g) of the extended symmetry group G, including both Poincaré symmetries and dilations. Our interest is how the dilation unitaries U0 (µ) relate to decomposition theory in Sec. 3, and how they behave in the more specific situations analyzed in Sec. 4. We will consider three cases of decreasing scope: first, the general situation; second, the factorizing scaling limit; third, the convergent scaling limit. We first consider a general theory as in Sec. 3, and analyze the decomposition of the dilation operators corresponding to the direct integral decomposition of H0 introduced
38
H. Bostelmann, C. D’Antoni, G. Morsella
in Thm. 3.2. To this end, we first note that δ µ leaves Z(A) invariant; thus we have a representation of the dilations UZ(µ) := U0 (µ) HZ on HZ. Identifying HZ with L 2 (Z, ν) as before, the UZ(µ) act on a function space. This action, and its extension to the entire Hilbert space, can be described in more detail. Proposition 5.1. Let ω0 be an invariant limit state. There exist an action of the dilations through homeomorphisms z → µ.z of Z, and unitary operators Uz (µ) : Hz → Hµ.z for µ ∈ R+ , z ∈ Z, such that: (i) the measure ν is invariant under the transformation z → µ.z; (ii) UZ(µ)χ (z) = χ (µ−1 .z) for all χ ∈ L 2 (Z, ν), as an equation in the L 2 sense; (iii) Uz (1) = 1, Uz (µ)∗ = Uµ.z (µ−1 ), Uµ.z (µ )Uz (µ) = Uz (µ µ) for all z ∈ Z, µ, µ ∈ R+ ; ↑ (iv) Uz (µ)Uz (x, Λ) = Uµ.z (µx, Λ)Uz (µ) for all z ∈ Z, (x, Λ) ∈ P+ ;
Γ
Γ −1 (v) U0 (µ)χ = Z dν(z) Uµ−1 .z (µ)χ (µ .z) for all χ ∈ Z dν(z) Hz . Proof. Recalling that Z is the spectrum of the commutative C∗ algebra π0 (Z(A)), we define the homeomorphism z → µ.z as the one induced by the automorphism ad UZ(µ−1 ) of π0 (Z(A)). For C ∈ Z(A), we know that π0 (C)Ω0 ∈ HZ corresponds to the function z → χC (z) = ω z (C), precisely the image of π0 (C) in the Gelfand isomorphism. Applying U0 (µ−1 ) to this vector, one obtains (UZ(µ−1 )χC )(z) = χC (µ.z);
(5.1)
thus (ii) holds for all χ ∈ C(Z). Taking the scalar product of Eq. (5.1) with Ω0 , one sees that Z dν(z)χ (z) = Z dν(z)χ (µ.z) for all µ and χ ∈ C(Z), so (i) follows. Now for general χ ∈ L 2 (Z, ν), statement (ii) follows by density. Expressing the action of z → µ.z on the level of algebras, it is easy to see that ωµ.z ◦ δ µ Z(A) = ω z Z(A).
(5.2)
Since however δ µ commutes with the projector ω : A → Z(A), the same equation holds on all of A. Therefore, the maps Uz (µ) : Hz → Hµ.z given by Uz (µ)πz (A)Ωz := πµ.z (δ µ (A))Ωµ.z
(5.3)
are well-defined and unitary. The properties of Uz (µ) listed in (iii) and (iv) then follow from this definition by easy computations.
Γ Now for (v): As before, we identify H0 with Z dν(z) Hz . Then we have, for all A ∈ A, U0 (µ)π0 (A)Ω0 = π0 (δ µ (A))Ω0 = =
Γ
Z
Γ
Z
dν(z) πz (δ µ (A))Ωz
dν(z) Uµ−1 .z (µ)πµ−1 .z (A)Ωµ−1 .z .
(5.4)
Γ Given now a vector χ ∈ Z dν(z) Hz , we can find a sequence (π0 (An )Ω0 )n∈N converging in norm to χ . Passing to a subsequence, we can also assume that πz (An )Ωz → χ (z)
On Dilation Symmetries Arising from Scaling Limits
39
in norm for almost every z ∈ Z. Hence, using the dominated convergence theorem and (i), we see that Γ Γ lim dν(z) Uµ−1 .z (µ)πµ−1 .z (An )Ωµ−1 .z = dν(z) Uµ−1 .z (µ)χ (µ−1 .z), n→+∞ Z
Z
which gives (v).
Thus dilations act between the fibers of the direct integral decomposition by unitaries Uz (µ), which depend on the fiber. They fulfill the cocycle-type composition rule Uµ.z (µ )Uz (µ) = Uz (µ µ) that one would naively expect; cf. also the theory of equivariant disintegrations for separable C ∗ algebras [Tak02, Ch. X §3]. We shall now further restrict to the situation of a factorizing scaling limit, as in Def. 4.3, in which the fiber spaces Hz are all identified with a unique space Hu . By this identification, we can regard the unitaries Uz (µ) as endomorphisms Uˆ z (µ) of Hu . Our result for these endomorphisms is as follows. Proposition 5.2. Let ω0 be an invariant limit state. Suppose that the scaling limit of A is factorizing, and let V = dν(z)Vz be the unitary of Def. 4.3. Then, the unitary operators Uˆ z (µ) : Hu → Hu , Uˆ z (µ) = Vµ.z Uz (µ)Vz∗ fulfill for any z ∈ Z, µ, µ ∈ R+ the relations Uˆ z (1) = 1, Uˆ z (µ)∗ = Uˆ µ.z (µ−1 ), Uˆ µ.z (µ )Uˆ z (µ) = Uˆ z (µ µ). If H0 is uniformly separable, one has
∗
V U0 (µ)V = (UZ(µ) ⊗ 1)
⊕
Z
dν(z)Uˆ z (µ); ↑
and for every µ > 0, there is a null set N ⊂ Z such that for any (x, Λ) ∈ P+ and any z ∈ Z\N , Uˆ z (µ)Uu (x, Λ) = Uu (µx, Λ)Uˆ z (µ). Proof. It is clear that Uˆ z (µ), defined as above, are unitary, and their composition relations follow from Prop. 5.1 (iii). Now let H0 be uniformly separable. Then, together with V , also V ∗ is decomposable. By a short computation, one finds for any χ ∈ L 2 (Z, ν, Hu ): ⊕ dν(z) Vz Uµ−1 .z (µ)Vµ∗−1 .z χ (µ−1 .z). (5.5) V U0 (µ)V ∗ χ = Z
Now, following Prop. 5.1 (ii), the operator UZ(µ) ⊗ 1 acts on vectors χ via ⊕ (UZ(µ) ⊗ 1)χ = dν(z) χ (µ−1 .z). Z
(5.6)
Together with Eq. (5.5), this entails V U0 (µ)V ∗ χ = (UZ(µ) ⊗ 1)
⊕
Z
dν(z) Vµ.z Uz (µ)Vz∗ χ (z),
(5.7)
40
H. Bostelmann, C. D’Antoni, G. Morsella
of which the second assertion follows. Further, one computes from V U0 (x, Λ)V ∗ = 1 ⊗ Uu (x, Λ) and from Eq. (5.7) that ⊕ ⊕ ˆ dν(z)Uz (µ)Uu (x, Λ) = dν(z)Uu (µx, Λ)Uˆ z (µ). (5.8) Z
Z
Uniform separability implies that the integrands agree except on a null set. This null set may depend on x, Λ. However, we can choose it uniformly on a countable dense set of the group, and hence, by continuity, uniformly for all Poincaré group elements. This shows that the dilation symmetries factorize into a central part, UZ(µ)⊗1, which
⊕ “mixes” the fibers of the direct integral, and a decomposable part, Z dν(z)Uˆ z (µ). The unitaries Uˆ z (µ) will generally depend on z; and like the Uz (µ) before, they do not necessarily fulfill a group relation, but a cocycle equation Uˆ µ.z (µ )Uˆ z (µ) = Uˆ z (µ µ),
(5.9)
as shown above, where µ.z can in general not be replaced with z. However, using the commutation relations with the other parts of the symmetry group, one sees that Uˆ z (µ )Uˆ z (µ)Uˆ z (µ µ)∗ is (a. e.) an inner symmetry of the theory Au . On the other hand, this representation property “up to an inner symmetry” cannot be avoided if such symmetries exist in the theory at all; for they might be multiplied to Vz in a virtually arbitrary fashion at any point z. In this respect, we encounter a similar situation with respect to dilation symmetries as Buchholz and Verch [BV95]. In the present context, however, it seems more transparent how this cocycle arises. Under somewhat stricter assumptions, we can prove a stronger result that avoids the ambiguities discussed above. Let us consider the case of a convergent scaling limit, per Def. 4.4. In this case, we shall see that the Uz (µ) can actually be chosen independent of z, and yield a group representation in the usual sense. Proposition 5.3. Let A have a convergent scaling limit, and let ωu be a multiplicative limit state with separable representation space Hu . Then the Poincaré group representation Uu on Hu extends to a representation of the extended symmetry group G. For any invariant limit state ω0 with associated representation U0 of G, one has V U0 (µ)V ∗ = (U0 (µ) HZ) ⊗ Uu (µ), where V is the unitary introduced in Proposition 4.10. Proof. With ωu , also every ωu ◦ δ µ is a scaling limit state. Thanks to the invariance of Aconv under dilations, we thus have for each A ∈ Aconv , πu (δ µ (A))Ωu 2 = ωu ◦ δ µ (A∗ A) = ωu (A∗ A) = πu (A)Ωu 2 .
(5.10)
This yields the existence of a unitary strongly continuous representation µ → Uu (µ) on Hu such that Uu (µ)πu (A)Ωu = πu (δ µ (A))Ωu ,
A ∈ Aconv .
(5.11)
That also implies Uu (µ)Uu (x, Λ)πu (A)Ωu = πu (α µ,x,Λ (A))Ωu ,
A ∈ Aconv ,
(5.12)
On Dilation Symmetries Arising from Scaling Limits
41
which shows that (µ, Λ, x) → Uu (µ)Uu (Λ, x) is a unitary representation of G on Hu , extending the representation of the Poincaré group. Now if V : H0 → HZ ⊗ Hu is the unitary of Prop. 4.10, a calculation shows that V U0 (µ)V ∗ π0 (C)Ω0 ⊗ πu (A)Ωu = π0 (δ µ (C))Ω0 ⊗ πu (δ µ (A))Ωu , C ∈ Z(A), A ∈ Aconv , (5.13) which entails that V U0 (µ)V ∗ = (U0 (µ) HZ) ⊗ Uu (µ).
Thus, the limit theory is “dilation covariant” in the usual sense, with a unitary acting on Hu . Considering the unitaries 1 ⊗ Uu (g), we actually get a unitary representation in any limit theory, even corresponding to multiplicative states. Only for compatibility with the scaling limit representation π0 it is necessary to consider invariant means, and to take U0 (µ) HZ into account. 6. Phase Space Properties In this section, we wish to investigate how the notion of phase space conditions, specifically the (quite weak) Haag-Swieca compactness condition [HS65], fits into our context, and how it transfers to the limit theory. An important aspect here is that Haag-Swieca compactness of a quantum field theory guarantees that the corresponding Hilbert space is separable; this property transfers to multiplicative limit states in certain circumstances [Buc96a]. We shall give a strengthened version of the compactness condition that guarantees our general limit spaces to be uniformly separable, a property that turned out to be valuable in the previous sections. We need some extra structures to that end. First, we consider “properly rescaled” vector-valued functions χ : R+ → H. Specifically, for A ∈ A, let AΩ denote the function λ → Aλ Ω. We set H = clos{AΩ | A ∈ A},
(6.1)
where the closure is taken in the supremum norm χ = supλ χ λ . Then H is a Banach space, in fact a Banach module over Z(A) in a natural way. Given a limit state, we transfer the limit representation π0 to vector-valued functions. To that end, consider the space C(Γ ) of Γ -continuous vector fields, as defined in the Appendix. We define η0 : H → C(Γ ) on a dense set by η0 (AΩ) := π0 (A)Ω0 .
(6.2)
This is well-defined, since one computes 1/2 π0 (A)Ω0 ∞ = sup πz (A)Ωz = sup ω z (A∗ A) z∈Z
∗
≤ ω(A A)
1/2
z∈Z
= sup Aλ Ω2 λ>0
1/2
= AΩ.
(6.3)
That also shows η0 ≤ 1. Note that η0 fulfills η0 (Cχ ) = π0 (C)η0 (χ ) for all C ∈ Z(A), χ ∈ H,
(6.4)
42
H. Bostelmann, C. D’Antoni, G. Morsella
this easily being checked for χ = AΩ. So η0 preserves the module structure in this sense. Further, η0 : H → C(Γ ) clearly has dense range. It is important in our context that H is left invariant under multiplication with suitably rescaled functions of the Hamiltonian. More precisely, we denote these functions as f (H ) for f ∈ S([0, +∞)); they are defined as elements of B by f (H )λ = f (λH ), with norm f (H ) ≤ f ∞ . They act on H by pointwise multiplication. The following lemma generalizes an observation in [Buc96a]. Lemma 6.1. Let f ∈ S([0, +∞)). Then, for each χ ∈ H, we have f (H )χ ∈ H. There exists a test function g ∈ S(R) such that for all A ∈ A,
f (H )AΩ = α g AΩ := dt g(t) α t A Ω. Proof. We continue f to a test function fˆ ∈ S(R), and choose g as the Fourier transform of fˆ. One finds by spectral analysis of H that for any A ∈ A, ∞ fˆ(λE)d P(E)Aλ Ω = dt g(t)eıλH t Aλ Ω = (α g A)λ Ω. (6.5) f (λH )Aλ Ω = 0
This shows that f (H )AΩ has the proposed form, and is an element of H. Since f (λH ) ≤ f ∞ uniformly in λ, we may pass to limits in AΩ and obtain that f (H )χ ∈ H for all χ ∈ H. As a next step towards phase space conditions, let us explain a notion of compact maps adapted to our context. To that end, let E be a Banach space and F a Banach module over the commutative Banach algebra R. We say that a linear map ψ : E → F is of uniform rank 1 if it is of the form ψ = e( · ) f with e : E → R linear and continuous, and f ∈ F. Sums of n such terms are called of uniform rank n.4 We say that ψ is uniformly compact if it is an infinite sum of terms of uniform rank 1, ψ = ∞ j=0 e j ( · ) f j , where the sum converges in the Banach norm. For R = C and F a Hilbert space, these definitions reduce to the usual notions of compact or finite-rank maps. We are now in the position to consider Haag-Swieca compactness. We fix, once and < for all, an element C < ∈ Z(A) with C < ≤ 1, C < λ = 0 for λ > 1, and C λ = 1 for λ < 1/2. For a given β > 0 and any bounded region O, we consider the map Θ (β,O) : A(O) → H,
A → e−β H C < AΩ.
(6.6)
This is indeed well-defined due to Lemma 6.1. Our variant of the Haag-Swieca compactness condition, uniform at small scales, is then as follows. Definition 6.2. A quantum field theory fulfills the uniform Haag-Swieca compactness condition if, for each bounded region O, there is β > 0 such that the map Θ (β,O) is uniformly compact. We note that this property is independent of the choice of C < ; the role of that factor is to ensure that we restrict our attention to the short-distance rather than the long-distance regime. We do not discuss relations of uniform Haag-Swieca compactness with other 4 Note that the “uniform rank” is rather an upper estimate, in the sense that a map of uniform rank n may at the same time be of uniform rank n − 1.
On Dilation Symmetries Arising from Scaling Limits
43
versions of phase space conditions here. Rather, we show in Sec. 7 that the condition is fulfilled in some simple models. We now investigate how the compactness property transfers to the scaling limit. To that end, we consider the corresponding phase space map in the limit theory, (β,O )
Θ0
A → e−β H0 AΩ0 .
: A0 (O) → H0 ,
(6.7)
Its relation to Θ (β,O) is rather direct. (β,O )
Proposition 6.3. For any fixed O and β > 0, one has η0 ◦ Θ (β,O) = Θ0 (β,O ) If Θ (β,O) is uniformly compact, so is Θ0 ◦ π0 .
◦ π0 .
Proof. Given β, we choose a function gβ relating to f β (E) = exp(−β E) per Lemma 6.1. For any A ∈ A(O), we compute η0 Θ (β,O) (A) = η0 (C < α gβ AΩ) = π0 (C < )π0 (α gβ A)Ω0 (β,O )
= α0,gβ π0 (A)Ω0 = Θ0
π0 (A).
(6.8)
(β,O )
◦ π0 as proposed. Now let Θ (β,O) be uniformly comThus η0 ◦ Θ (β,O) = Θ0 pact, Θ (β,O) = j e j ( · ) f j . Then η0 can be exchanged with the infinite sum due to continuity, which yields (β,O ) Θ0 ◦ π0 = η0 (e j ( · ) f j ) = (π0 ◦ e j ( · ))(η0 f j ), (6.9) j (β,O )
using Eq. (6.4). Thus Θ0
j
◦ π0 is uniformly compact.
(β,O )
◦ π0 ⊂ C(Γ ). Since we can The above results show in particular that img Θ0 write (β,O ) Θ0 ◦ π0 (A) = dν(z) Θz(β,O) ◦ πz (A) (6.10) (β,O )
with the obvious definition of Θz , the above proposition establishes a rather strong form of compactness in the limit theory, uniform in z; note that the sum in Eq. (6.9) converges with respect to the supremum norm of C(Γ ). We now come to the main result of the section, showing that compactness in the above form implies uniform separability of the limit Hilbert space. Theorem 6.4. Suppose that the theory A fulfils uniform Haag-Swieca compactness. Then, for any limit state ω0 , the representation space H0 is uniformly separable, where the fundamental sequence can be chosen from C(Γ ). Proof. We choose a sequence of regions Ok such that Ok Rs+1 , and a sequence (βk )k∈N in R+ such that all Θ (βk ,Ok ) are uniformly compact. By Prop. 6.3 above, also (β ,O ) (k) Θ0 k k ◦ π0 are uniformly compact. Explicitly, choose e j : A(Ok ) → C(Z) and f j(k) ∈ C(Γ ) such that
(βk ,Ok )
Θ0
◦ π0 =
j
(k) e(k) j (·) fj .
(6.11)
44
H. Bostelmann, C. D’Antoni, G. Morsella (k)
We will construct a fundamental sequence using the f j . To that end, let A ∈ A(O) for some O. For k large enough, we know that (k) (β ,O ) (k) e−βk H0 π0 (A)Ω0 = Θ0 k k (π0 (A)) = e j (A) f j . (6.12) j
The sum converges in the supremum norm, i.e., uniformly at all points z. Let us choose a fixed z. Then, it is clear that (k) 2 2 (k) e−Hz πz (A)Ωz = e j (A) (z) e−Hz +βk Hz f j (z), (6.13) j (k)
noting that exp(−Hz2 +βk Hz ) is a bounded operator. Observe that (e j (A))(z) are merely numerical factors. Since A and O were arbitrary, and ∪O πz (A(O))Ωz is dense in Hz , this means (k)
e−Hz Hz ⊂ clos span{ e−Hz +βk Hz f j (z)| j, k ∈ N}. 2
2
(6.14)
Now exp(−Hz2 ) is a selfadjoint operator with trivial kernel, thus its image is dense. Hence the exp(−Hz2 + βk Hz ) f j(k) (z) are total in Hz . This holds for all z, thus {exp(−H02 + (k)
βk H0 ) f j | j, k ∈ N} is a fundamental sequence. Applying Lemma 6.1 to f (E) = exp(−E 2 + βk E), we find that the elements of the fundamental sequence lie in C(Γ ). 7. Examples We are now going to investigate the structures discussed in simple models. Particularly, we wish to show that our conditions on “convergent scaling limits” (Def. 4.4) and “uniform Haag-Swieca compactness” (Def. 6.2) can be fulfilled at least in simple situations. To that end, we first consider the situation where the theory A “at finite scales” is equipped with a dilation symmetry. Then, we investigate the real scalar free field as a concrete example. 7.1. Dilation covariant theories. We now consider the case where the net A, which our investigation starts from, is already dilation covariant. One expects that the scaling limit construction reproduces the theory A in this case, and that the dilation symmetry obtained from the scaling algebra coincides with the original one. We shall show that this is indeed the case under a mild phase space condition, and also that this implies the stronger phase space condition in Def. 6.2. This extends a discussion in [BV95, Sec. 5]. Technically, we will assume in the following that A is a local net in the vacuum sector with symmetry group G, which is generated by the Poincaré group and the dilation group. We shall denote the corresponding unitaries as U (µ, x, Λ) = U (µ)U (x, Λ). The mild phase space condition referred to is the Haag-Swieca compactness condition for the original theory: We assume that for each bounded region O in Minkowski space, there exists β > 0 such that the map Θ (β,O) : A(O) → H, A → exp(−β H )AΩ is compact. (This is equivalent to a formulation where the factor exp(−β H ) is replaced with a sharp energy cutoff, as used in [HS65].)
On Dilation Symmetries Arising from Scaling Limits
45
Theorem 7.1. Let A be a dilation covariant net in the vacuum sector which satisfies the Haag-Swieca compactness condition. Then A has a convergent scaling limit. ˆ Proof. For each O, we introduce the C∗ -subalgebra A(O) ⊂ A(O) of those elements A ∈ A(O) for which g → αg (A) is norm continuous. Since the symmetries are impleˆ mented by continuous unitary groups, A(O) is strongly dense in A(O). We then define a C∗ -subalgebra of the scaling algebra A(O), ˆ Aconv (O) := {λ → U (λ)AU (λ)∗ | A ∈ A(O)},
(7.1)
and the α-invariant algebra Aconv ⊂ A is defined as the C∗ -inductive limit of the Aconv (O). It is evident that condition (i) in Def. 4.4 is fulfilled by Aconv , as the functions λ → ω(Aλ ) are constant in the present case. Now let ω0 be a multiplicative limit state. With similar arguments5 as in [BV95, Prop. 5.1], using the Haag-Swieca compactness condition, we can construct a net isomorphism φ from A0 to A, which has the property ˆ that if Aλ = U (λ)AU (λ)∗ with A ∈ A(O), then φ(π0 (A)) = A. From this, and from ˆ the strong density of A(O) in A(O), it follows that π0 (Aconv (O)) is strongly dense in A0 (O). Thus condition (ii) in Def. 4.4 is satisfied as well. Since the isomorphism φ above can be shown to intertwine the respective vacuum states, it is actually the adjoint action of a unitary W : H0 → H. We remark that H, and then also H0 , is separable due to the Haag-Swieca compactness condition. Then as a consequence of Prop. 5.3, A has a factorizing scaling limit and the representation of the symmetry group G0 factorizes too. It is also clear from the proof above and from that of Thm. 4.9, that Au is unitarily equivalent to A through the operator W , taken here for πu in place of π0 . Furthermore, this W also intertwines the dilations in the scaling limit with those of the underlying theory. Corollary 7.2. Under the hypothesis of Thm. 7.1, there holds W Uu (µ)W ∗ = U (µ). ˆ Proof. It is sufficient to verify the relation on vectors of the form AΩ with A ∈ A(O). For such vectors it follows by noting that Aλ = U (λ)AU (λ)∗ is an element of Aconv (O), ˆ and that δ µ (A)λ = U (λ)αµ (A)U (λ)∗ with αµ (A) ∈ A(µO). For showing the consistency of our definitions, we now prove that the Haag-Swieca compactness condition at finite scales, together with dilation covariance, implies our uniform compactness condition of Def. 6.2. Proposition 7.3. If the dilation covariant local net A fulfills the Haag-Swieca compactness condition, then it also fulfills uniform Haag-Swieca compactness. Proof. Let O be fixed, and let β > 0 such that Θ (β,O) is compact; Θ (β,O) =
∞
e j ( · ) f j with e j ∈ A(O)∗ , f j ∈ H.
(7.2)
j=1 5 Since in contrast to [BV95], we here take the A (O) to be W∗ algebras, we need to amend the argument 0 in step (d) of [BV95, Prop. 5.1] slightly: We first construct the isomorphism φ on the C∗ algebra π0 (A(O)), and then continue it to the weak closure; cf. [KR97, Lemma 10.1.10].
46
H. Bostelmann, C. D’Antoni, G. Morsella
Taking the normal part, we can in fact arrange that e j ∈ A(O)∗ . (See [BDF87, Lemma 2.2] for a similar argument.) Now define e j : A(O) → Z(A) by e j (A)λ = e j U (λ)∗ C < (7.3) λ Aλ U (λ) . That the image is indeed in Z(A), i.e., continuous under δ µ , is seen as follows. We compute for λ, µ > 0, e (A)λµ − e (A)λ = e j U (λµ)∗ (C < A)λµ U (λµ) − U (λ)∗ (C < A)λ U (λ) j j ≤ e j δ µ (C < A)−C < A + e j U (µ)∗ · U (µ) −e j C < A. (7.4) Now as µ → 0, the first summand vanishes due to norm continuity of δ µ on A, and the second due to strong continuity of U (µ); both limits are uniform in λ. Thus δ µ acts continuously on e j (A). Further, we define f j ∈ H by f
jλ
= U (λ) f j .
(7.5)
ˆ for suitable O This is indeed an element of H: Namely, given > 0, choose A ∈ A(O) ˆ is as in the proof of Thm. 7.1. Then, Aλ = U (λ)AU (λ)∗ such that AΩ− f j < ; here A defines an element of A, and AΩ − f j ≤ AΩ − f j < . Hence f j is contained in the closure of AΩ. Now we are in the position to show that Θ (β,O) = j e j f j . Let J ∈ N be fixed. It is straightforward to compute that for any A ∈ A(O) and λ > 0, J J Θ (β,O) (A) − e j (A) f j = U (λ) Θ (β,O) (Bλ ) − e j Bλ ) f j , j=1
λ
j=1
where Bλ = U (λ)∗ C < λ Aλ U (λ).
(7.6)
Note here that Bλ ∈ A(O) for any λ. This entails J J (β,O) Θ − e j ( · ) f j ≤ Θ (β,O) − e j ( · ) f j A. j=1
(7.7)
j=1
The right-hand side vanishes as J → ∞, as a consequence of the compactness condition at finite scales. This shows that Θ (β,O) is uniformly compact. 7.2. The scaling limit of a free field. We now show in a simple, concrete example from free field theory that the model has a convergent scaling limit in the sense of Def. 4.4. Specifically, we consider a real scalar free field of mass m > 0, in 2+1 or 3+1 spacetime dimensions. The algebraic scaling limit of this model is the massless real scalar field; this was as already discussed in [BV98], and in parts we rely on the arguments given there. However, we need to consider several aspects that were not handled in that work, in particular continuity aspects of Poincaré and dilation transformations. Also, as mentioned before, in contrast to [BV98] we deal with weakly closed local algebras at fixed scales and in the limit theory.
On Dilation Symmetries Arising from Scaling Limits
47
We start by recalling, for convenience, the necessary notations and definitions from [BV98]. We consider the Weyl algebra W over D(Rs ), s = 2, 3: ı W ( f )W (g) = e− 2 σ ( f,g) W ( f + g), σ ( f, g) = Im dx f (x)g(x). (7.8) ↑
Then, we define a mass dependent automorphic action of P+ on W by (m)
(m)
αx,Λ (W ( f )) = W (τx,Λ f ),
(7.9)
↑
where the action τ (m) of P+ on D(Rs ) is defined by the following formulas. In those, we write f˜(p) = (2π )−s/2 dx f (x)e−ixp for the Fourier transform of f, which we split into f˜ = f˜R + ı f˜I , where f R = Re f and f I = Im f ; also, ωm (p) := m 2 + |p|2 . (τx(m) f )(y) := f (y − x), (m)
˜ ˜ f )∼ R (p) := cos(tωm (p)) f R (p) − ωm (p) sin(tωm (p)) f I (p),
(m)
−1 ˜ f )∼ sin(tωm (p)) f˜R (p), I (p) := cos(tωm (p)) f I (p) + ωm (p)
(τt (τt
(m)
(τΛ f )∼ R (p) := ϕΛ (ωm (p), p),
(7.11)
f
−1 (τΛ(m) f )∼ I (p) := ωm (p) ψΛ (ωm (p), p). f
f
(7.10)
(7.12)
f
Here the functions ϕΛ , ψΛ : Rs+1 → C are defined by 1 1 ˜ f f R (Λ−1 p) + f˜R (ΛT p) + (Λ−1 p)0 f˜I (Λ−1 p) − (ΛT p)0 f˜I (ΛT p) , ϕΛ ( p) := 2 2ı 1 1˜ f −1 T f R (Λ p)− f˜R (Λ p) + (Λ−1 p)0 f˜I (Λ−1 p) + (ΛT p)0 f˜I (ΛT p) , ψΛ ( p) := − 2ı 2 (7.13) where we use the notation Λp = ((Λp)0 , Λp) for Λ ∈ L, p ∈ Rs+1 . One verifies that all the above expressions are even in ωm (p), which, due to the analytic properties of f˜, (m) implies that τx,Λ D(Rs ) ⊂ D(Rs ). We also introduce the action σ of dilations on W by σλ (W ( f )) = W (δλ f ),
(7.14)
(δλ f )(x) := λ−(s+1)/2 ( f R )(λ−1 x) + ıλ−(s−1)/2 ( f I )(λ−1 x).
(7.15)
with
(m)
(λm)
It holds that αλx,Λ ◦ σλ = σλ ◦ αx,Λ . Finally we define the vacuum state of mass m ≥ 0 on W as 1
ω(m) (W ( f )) = e− 2 f m , where f 2m :=
2
2 1 dpωm (p)−1/2 f˜R (p) + ı ωm (p)1/2 f˜I (p) . 2 Rs
(7.16)
(7.17)
48
H. Bostelmann, C. D’Antoni, G. Morsella (m)
There holds clearly ω(m) ◦ αx,Λ = ω(m) , ω(m) ◦ σλ = ω(λm) . Proceeding now along the lines of [BV98], we consider the GNS representation (π (0) , H(0) , Ω (0) ) of W induced by the massless vacuum state ω(0) . For each m ≥ 0, we define a net O → A(m) (O) of von Neumann algebras on H(0) as (m) A(m) (ΛO B + x) := {π (0) αx,Λ (W (g)) : supp g ⊂ B} , (7.18) where O B is any double cone with base the open ball B in the time t = 0 plane. For other open regions we can define the algebras by taking unions, but this will not be relevant for the following discussion. Due to the local normality of the different states ω(m) , m ≥ 0, with respect to each other [EF74], these nets are isomorphic to the nets generated by the free scalar field of mass m on the respective Fock spaces. From now on, we will identify elements of W and of A(m) , and therefore we will drop the indication of the representation π (0) . We denote the (dilation and Poincaré covariant) scaling algebra associated to A(m) by A(m) . The next lemma generalizes the results of [BV98, Lemma 3.2] to the present situation. ↑
Lemma 7.4. Let a > 1 and h D ∈ D((1/a, a)), h P ∈ D(P+ ), f ∈ D(Rs ), and consider the function W : R+ → A(m) given by dµ (m) W λ := d x dΛ h D (µ)h P (x, Λ)αµλx,Λ ◦ σµλ (W ( f )), ↑ µ R + × P+ where dΛ is the left-invariant Haar measure on the Lorentz group and the integral is to be understood in the weak sense. Then: (i) there exists a double cone O such that W ∈ A(m) (O); (ii) there holds in the strong operator topology, dµ (0) −1 d x dΛ h D (µ)h P (x, Λ)αµx,Λ lim σλ (W λ ) = ◦ σµ (W ( f )) =: W0 ; ↑ µ λ→0+ R + × P+ (iii) the span of the operators W0 of the form above, with W ∈ A(m) (O) for fixed O, is strongly dense in A(0) (O). (m)
Proof. Since W is the convolution, with respect to the action (µ, x, Λ) → α µ,x,Λ , of the function h D ⊗ h P with the bounded function λ → σλ (W ( f )), and thanks to the support properties of h D , h P and f , (i) follows. In order to prove (ii), we start by observing that, for each vector χ ∈ H(0) , dµ σλ−1 (W λ ) − W0 χ ≤ d x dΛ |h D (µ)h P (x, Λ)| ↑ µ R + × P+ (µλm) (0) (7.19) × W δµ τx,Λ f − W δµ τx,Λ f χ . Now f → W ( f ) is known to be continuous with respect to · 0 on the initial space and the strong operator topology on the target space [BR81, Prop. 5.2.4]. Since the norm ↑ · 0 is δµ -invariant, it therefore suffices to show that for each fixed (x, Λ) ∈ P+ , (m) (0) lim + τx,Λ f − τx,Λ f 0 = 0; (7.20) m→0
On Dilation Symmetries Arising from Scaling Limits
49
for (ii) then follows from the dominated convergence theorem. In order to show Eq. (7.20), we introduce the following family of functions f (m) (p) of two arguments: F = f : [0, 1] × Rs → C f (m) ( · ) ∈ D(Rs ) for each fixed m ∈ [0, 1]; lim f (m) (p) = f (0) (p) for each fixed p ∈ Rs ;
m→0
f (m) (p)| ≤ g(p) . ∃g ∈ S(Rs ) : ∀m ∈ [0, 1] ∀p ∈ Rs : |
(7.21)
It is clear that for f ∈ F, one has f (m) − f (0) 0 → 0 as m → 0 per dominated convergence. Also, each f ∈ D(Rs ), with trivial dependence on m, falls into F. So it (·) leaves F invariant, where remains to show that the (naturally defined) action of τx,Λ (·)
it suffices to check this for a set of generating subgroups. Indeed, τx,Λ F ⊂ F is clear for spatial translations and rotations. For time translations and boosts, it was already remarked that D(Rs ) is invariant under these at fixed m, and pointwise convergence as m → 0 is clear. Further, from Eqs. (7.11) and (7.13), one sees that f˜ is modified by at most polynomially growing functions, uniform in m ≤ 1, hence uniform S-bounds (m) (m) f as well. (Again, it enters here that all expressions are even in ωm , for hold for τx,Λ which it is needed that f˜ is smooth.) This completes the proof of (ii). Finally, (iii) follows from the observation that as h D and h P converge to delta functions, W0 converges strongly to W ( f ) thanks to the strong continuity of the function (0) (µ, x, Λ) → αµx,Λ ◦ σµ (W ( f )); and of course the span of the Weyl operators with supp f ⊂⊂ O is strongly dense in A(0) (O). Using the above lemma, we can prove the following. Theorem 7.5. The theory of the massive real scalar free field in s = 2, 3 spatial dimensions has a convergent scaling limit. (m)
Proof. Consider the C∗ -subalgebra Aconv (O) of A(m) (O) which is generated by the ele(m) ments W ∈ A(m) (O) defined in the previous lemma, and let Aconv be the corresponding (m) quasi-local algebra. Since α µ,x,Λ (W ) is again an element of the same form, just with (m)
shifted function h D ⊗ h P , the algebra Aconv is α (m) invariant. In order to verify that (m) , we start by observing that, λ → ω(m) (Aλ ) has a limit, as λ → 0, for all A ∈ Aconv thanks to Lemma 7.4 (ii) and to the fact that σλ is unitarily implemented on H(0) , for each such A there exists limλ→0+ σλ−1 (Aλ ) =: A in the strong operator topology. Then (m) (O) there holds the inequality if A ∈ Aconv |ω(m) (Aλ ) − ω(0) (A)| ≤ (ω(m) − ω(0) ) A(0) (λO)A + |ω(0) (σλ−1 (Aλ )) − ω(0) (A)|.
(7.22) Together with the fact that limλ→0+ (ω(m) − ω(0) ) A(0) (λO) = 0 as a consequence of the local normality of ω(m) with respect to ω(0) , this implies that limλ→0+ ω(m) (Aλ ) = (m) (m) ω(0) (A) for all A in some local algebra Aconv (O). This then extends to all of Aconv by density.
50
H. Bostelmann, C. D’Antoni, G. Morsella (m)
It remains to show that for multiplicative limit states, π0 (Aconv ∩ A(m) (O)) is weakly (m) dense in A0 (O) = π0 (A(m) (O)) for any O. To that end, we use similar methods as in Thm. 7.1. With O fixed and U the ultrafilter that underlies the limit state, we define φ : π0 (A(m) (O)) → A(0) (O), π0 (A) → lim σλ−1 (Aλ ),
(7.23)
U
with the limit understood in the weak operator topology. Using methods as in [BV98, Sec. 3], one can show that φ is indeed a well-defined isometric ∗ homomorphism, which further satisfies ω0 = ω(0) ◦ φ on the domain of φ. Hence φ is given by the adjoint action of a partial isometry, and can be continued by weak closure to a ∗ homomorphism (m) φ − : A0(m) (O) → A(0) (O). On the other hand, for W ∈ Aconv (O) as in Lemma 7.4, one finds φ(π0 (W )) = W0 by (ii) of that lemma. However, the double commutant of those W0 is all of A(0) (O), see (iii) of the same lemma. So φ − is in fact an isomorphism; and inverting φ − , one obtains the proposed density. 7.3. Phase space conditions in the free field. Our aim in this section is to prove the uniform compactness condition of Sec. 6 in the case of a real scalar free field, again of mass m ≥ 0, in 3 + 1 or higher dimensions. To that end, we will use a short-distance expansion of local operators, very similar to the method used in the Appendix of [Bos05b], however in a refined formulation. In this section, we will consider a fixed mass m ≥ 0 throughout, and therefore we drop the label (m) from the local algebras, the vacuum state, and the Hilbert space norm for simplicity. We rewrite the Weyl operators of Eq. (7.8) in terms of the familiar free field φ and its time derivative ∂0 φ in the time-0 plane, (7.24) W ( f ) = exp ı φ(Re f ) − ∂0 φ(Im f ) , f ∈ D(Rs ). Also, we need to introduce some multi-index notation. Given n ∈ N0 , we consider multi-indexes ν = (ν1 , . . . , νn ) ∈ ({0, 1} × Ns0 )n ; that is, each ν j has the form ν j = (ν j0 , ν j1 , . . . , ν js ) with ν j0 ∈ {0, 1}, ν jk ∈ N0 for 1 ≤ k ≤ s. These indices will be ν ν used for labeling derivatives in configuration space, ∂ ν j = ∂0 j0 . . . ∂s js . We denote νj! =
s
ν jk ! , ν! =
k=0
n
ν j ! , |ν j | =
j=1
s k=0
ν jk , |ν| =
n
|ν j |.
(7.25)
j=1
Now we can define the following local fields as quadratic forms on a dense domain: φn,ν = :
n
∂ ν j φ: (0).
(7.26)
j=1
These will form a basis in the space of local fields at x = 0. Further, for given r > 0, we choose a test function h ∈ D(Rs ) which is equal to 1 for |x| ≤ r ; then we set ν h ν j (x) = sk=1 xk jk h(x). This is used to define the following functionals on A(Or ).
ı n (−1) j ν j0 (1−ν10 ) Ω [∂0 φ(h ν1 ), [. . . [∂0(1−νn0 ) φ(h νn ), A] . . .]Ω . (7.27) σn,ν (A) = n! ν! One sees that this definition is independent of the choice of h. We can therefore consistently consider σn,ν as a functional on ∪O A(O), though its norm may increase as O grows large. The significance of these functionals becomes clear in the following lemma.
On Dilation Symmetries Arising from Scaling Limits
51
Lemma 7.6. We have for all Weyl operators A = W ( f ) with f ∈ D(Rs ), A=
∞
σn,ν (A) φn,ν
n=0 ν
in the sense of matrix elements between vectors of finite energy and finite particle number. Proof. We indicate only briefly how this combinatorial formula can be obtained; see also [Bos00, Sec. 7.2] and [Bos05b, Appendix]. Using Wick ordering, we rewrite Eq. (7.24) for the Weyl operators as 2 W ( f ) = e− f /2 : exp ı φ(Re f ) − ∂0 φ(Im f ) : = e− f
2 /2
∞ n n ı : φ(Re f ) − ∂0 φ(Im f ) : . n!
(7.28)
n=0
Now, in each factor of the n-fold product :(. . .)n:, we expand both Re f and Im f into a Taylor series in momentum space. Note that this is justified, since those functions have compact support in configuration space, since they are evaluated in scalar products with functions of compact support in momentum space, and since the sum over n is finite in matrix elements. The Taylor expansion in momentum space corresponds to an expansion in derivatives of δ-functions in configuration space, and this is what produces the fields φn,ν localized at 0. We then need to identify the remaining factors with σn,ν (A), which is done using the known commutation relations of W ( f ) with φ and ∂0 φ. Our main task will now be to extend the above formula to more general states and more observables, by showing that the sum converges in a suitable norm. To that end, we need estimates of the fields and functionals involved. Lemma 7.7. Given s ≥ 2, m ≥ 0, and r0 > 0, there exists a constant c such that the following holds for any n, ν: √ e−β H φn,ν Ω ≤ cn (n!)1/2 ν! (β/2 s)−|ν|−n(s−1)/2 for any β > 0, (a) P(E)φn,ν P(E) ≤ cn E |ν|+n(s−1)/2 σn,ν A(Or ) ≤ c (n!) n
−1/2
(ν!)
−1
for any E > 0, provided s ≥ 3, (3r )
|ν|+n(s−1)/2
for any r ≤ r0 .
(b) (c)
Proof. One has n n √ e−β H φn,ν Ω = a ∗ (e−βω p ν j ) Ω ≤ n! e−βω p ν j . j=1
(7.29)
j=1
For the single-particle vectors on the right-hand side, one uses scaling arguments to obtain the estimate √ e−βω p ν j ≤ c1 ν j ! (β/2 s)−|ν j |−(s−1)/2 for β > 0, (7.30) where c1 is a constant (depending on s and m). This implies (a). For (b), we use energy bounds for creation operators a ∗ ( f ), similar to [BP90, Sec. 3.3]. One finds for single 1/2 particle space functions f 1 , . . . , f k in the domain of ωm , P(E)a ∗ (ωm f 1 ) . . . a ∗ (ωm f k ) ≤ E k/2 Q(E) f 1 . . . Q(E) f k , 1/2
1/2
(7.31)
52
H. Bostelmann, C. D’Antoni, G. Morsella
with Q(E) being the energy projector for energy E in single particle space. This leads us to P(E)φn,ν P(E) ≤ 2n E n/2
n
−1/2 ν j
ωm
p χ E ,
(7.32)
j=1
ν where p ν j = sk=0 pk jk , and χ E is the characteristic function of ωm (p) ≤ E. For the single-particle functions, one obtains −1/2 ν j
ωm
p χ E ≤ c2 E |ν j |+(s−2)/2
(7.33)
with a constant c2 , which implies (b). Now consider the functional σn,ν . We choose a real-valued test function h 1 ∈ D(R+ ) such that h 1 (x) ≤ 1 for all x, h 1 (x) = 1 on [0, 1], and h 1 (x) = 0 for x ≥ 2. Then, h r (x) = h 1 (|x|/r ) is a valid choice for the test function used in the definition of σn,ν A(Or ), see Eq. (7.27). Expressing the fields φ there in annihilation and creation operators, and writing each commutator as a sum of two terms, we obtain n 4n (1−ν j0 ) σn,ν A(Or ) ≤ √ ωm h r,ν j . n! ν! j=1
(7.34)
Again, we use scaling arguments for the single-particle space functions and obtain for r ≤ r0 , (1−ν j0 )
ωm
h r,ν j ≤ c3 (3r )|ν j |+(s−1)/2
with a constant c3 that depends on r0 . This yields (c).
(7.35)
We now define the “scale-covariant” objects that will allow us to expand the maps Θ (β,O) in a series. They are constructed of the fields φn,ν and the functionals σn,ν by multiplication with appropriate powers of λ. We begin with the quantum fields. Proposition 7.8. For any n, ν and β > 0, the function χ n,ν,β : λ → λ|ν|+n(s−1)/2 e−βλH φn,ν Ω √ defines an element of H, with norm estimate χ n,ν,β ≤ cn (n!)1/2 ν!(β/2 s)−|ν|−n(s−1)/2. Proof. We use techniques from [BDM09], and adopt the notation introduced there. In particular, R denotes the function R λ = (1 + λH )−1 , and we write A() = supλ R λ Aλ R λ , where Aλ may be unbounded quadratic forms. Let n, ν be fixed in the following. We set φ λ = λ|ν|+n(s−1)/2 φn,ν .
(7.36)
From Lemma 7.7, one sees that P(E/λ)φ λ P(E/λ) is bounded uniformly in λ. Hence, applying [BDM09, Lemma 2.6], we obtain φ() < ∞ for sufficiently large . Also, the action g → α g φ of the symmetry group on φ (which extends canonically from bounded operators to quadratic forms) is continuous in some · () : This is clear for
On Dilation Symmetries Arising from Scaling Limits
53
translations by the energy-damping factor; for dilations it is immediate from the definition; and for Lorentz transformations it holds since they act by a finite-dimensional matrix representation on φn,ν . Thus, φ is an element of the space Φ defined in [BDM09, Eq. (2.39)]. Moreover, each φ λ is clearly an element of the Fredenhagen-Hertel field content ΦFH . Thus, [BDM09, Thm. 3.8] provides us with a sequence (An ) in A(O), with O a fixed neighborhood of zero, such that An − φ() → 0 as n → ∞. Now since exp(−β H )R − < ∞, we obtain exp(−β H )(An − φ)Ω → 0. Note here that exp(−β H )An Ω ∈ H by Lemma 6.1. Hence exp(−β H )φΩ = χ n,ν,β lies in H, since this space is closed in norm. The estimate for χ n,ν,β follows directly from Lemma 7.7(a). Next, we rephrase the functionals σn,ν as maps from the scaling algebra A to its center. Lemma 7.9. For any n, ν, the definition (σ n,ν (A))λ = λ−|ν|−n(s−1)/2 σn,ν ((C < A)λ ) yields a linear map σ n,ν : ∪O A(O) → Z(A), with its norm bounded by √ σ n,ν A(Or ) ≤ cn ( n! ν!)−1 (3r )|ν|+n(s−1)/2 for r ≤ r0 ; here r0 , c are the constants of Lemma 7.7. Proof. The norm estimate is a consequence of Lemma 7.7(c), where one notes that (C < A)λ ∈ A(Oλr ) for λ ≤ 1, and (C < A)λ = 0 for λ > 1, so that these operators are always contained in A(Or0 ). It remains to show that σ n,ν (A) ∈ Z(A), i.e., that µ → δ µ (σ n,ν (A)) is continuous. But this follows from continuity of µ → δ µ A and the definition of σ n,ν . We are now in the position to prove that n,ν σ n,ν χ n,ν,β is a norm convergent expansion of the map Θ (β,O) . Theorem 7.10. Let s ≥ 3. For each r > 0, there exists β > 0 such that Θ (β,Or ) =
∞
σ n,ν ( · )χ n,ν,β .
n=0 ν
Proof. We will show below that the series in the statement converges absolutely in the Banach space B(A(Or ), H), i.e. that ∞
σ n,ν A(Or )χ n,ν,β < ∞.
(7.37)
n=0 ν
Once this has been established, the assertion of the theorem is obtained as follows. From Lemma 7.6, we know that ∞ n=0 ν
λ−|ν|−n(s−1)/2 σn,ν (A)(χ n,ν,β )λ = e−λβ H AΩ
(7.38)
54
H. Bostelmann, C. D’Antoni, G. Morsella
at each fixed λ, whenever A is a linear combination of Weyl operators, and when evaluated in scalar products with vectors from a dense set. But (7.37) also shows that the left hand side of (7.38) converges in B(A(λOr ), H), and it is therefore strongly continuous for A in norm bounded parts of A(λOr ). Then by Kaplansky’s theorem (7.38) holds for in H. Finally, this entails for all A ∈ A(Or ) that any A ∈ A(λOr ) as an equality (β,Or ) (A) at each fixed λ > 0, i.e. the statement. σ (A) (χ ) = Θ λ n,ν,β λ λ n,ν n,ν We √ now prove Eq. (7.37). Let r0 > 0 be fixed in the following, and r < r0 . Set a := 6 s. From Prop. 7.8 and Lemma 7.9, we obtain the estimate σ n,ν A(Or )χ n,ν,β ≤ c2n (ar/β)|ν|+n(s−1)/2 s n n = c2 (ar/β)(s−1)/2 (ar/β)ν jk . (ar/β)ν j0 j=1
(7.39)
k=1
Factorizing the sum over multi-indexes ν accordingly, we obtain at fixed n, ν
(ar/β)(s−1)/2 n σ n,ν A(Or )χ n,ν,β ≤ 2c2 , (1 − ar/β)s
(7.40)
where we suppose ar/β < 1, and where each sum over ν j0 ∈ {0, 1} has been estimated by introducing a factor of 2. Now if we choose r/β small enough, we can certainly achieve that the expression in Eq. (7.40) is summable over n as a geometric series, and hence the series in Eq. (7.37) converges. This establishes the phase space condition of Def. 6.2 in our context. Corollary 7.11. The theory of a real scalar free field of mass m ≥ 0 in 3 + 1 or more space-time dimensions fulfills the uniform Haag-Swieca compactness condition. While our goal was to show that the maps Θ (β,O) are uniformly compact, it follows from Eq. (7.37) that they are actually uniformly nuclear at all scales, or by a slightly modified argument, even uniformly p-nuclear for any 0 < p ≤ 1. So we can generalize the somewhat stronger Buchholz-Wichmann condition [BW86] to our context. Several other types of phase space conditions can be derived with similar methods as in Thm. 7.10 as well. Particularly, one can show for s ≥ 3 that the sum n,ν σn,ν φn,ν converges in norm under a cutoff in energy E and restriction to a fixed local algebra A(Or ), with estimates uniform in E · r , where this product is small. This implies that Phase Space Condition II of [BDM09], which guarantees a regular behavior of pointlike fields under scaling, is fulfilled for those models. 8. Conclusions In this paper, we have considered short-distance scaling limits in the model independent framework of Buchholz and Verch [BV95]. In order to describe the dilation symmetry that arises in the limit theory, we passed to a generalized class of limit states, which includes states invariant under scaling. The essential results of [BV95] carry over to this generalization, including the structure of Poincaré symmetries and geometric modular action. However, the dilation invariant limit states are not pure. Rather, they can be decomposed into states of the Buchholz-Verch type, which are pure in two or more spatial
On Dilation Symmetries Arising from Scaling Limits
55
dimensions. This decomposition gives rise to a direct integral decomposition of the Hilbert space of the limit theory, under which local observables, Poincaré symmetries, and other relevant objects of the theory can be decomposed—except for dilations. The dilation symmetry has a more intricate structure, intertwining the pure components of the limit state. The situation simplifies if we consider the situation of a “unique limit” in the classification of [BV95]; our condition of a “factorizing limit” turned out to be equivalent modulo technicalities. Under this restriction, the dilation unitaries in the limit are, up to a central part, decomposable operators. The decomposed components do not necessarily fulfill a group relation though, but a somewhat weaker cocycle relation. Only under stronger assumptions (“convergent scaling limits”) we were able to show that the dilation symmetry factorizes into a tensor product of unitary group representations. It is unknown at present which type of models would make the generalized decomposition formulas necessary. In this paper, we have only considered very simple examples, which all turned out to fall into the more restricted class of convergent scaling limits. However, thinking e.g. of infinite tensor products of free fields with increasing masses as suggested in [Buc96a, Sec. 5], it may well be that some models violate the condition of uniqueness of the scaling limit, or even exhibit massive particles in the limit theory. In this case, the direct integral decomposition would be needed to obtain a reasonable description of dilation symmetry in the limit, with the symmetry operators intertwining fibers of the direct integral that correspond to different masses. As a next step in the analysis of dilation symmetries in the limit theory, one would like to investigate further the deviation of the theory at finite scales from the idealized dilation covariant limit theory; so to speak, the next-order term in the approximation λ → 0. This would be interesting e.g. for applications to deep inelastic scattering, which can currently only be treated in formal perturbation theory. It is expected that the dilation symmetry in the limit also contains some information about these next-order terms. Our formalism, however, does at the moment not capture these next-order approximations, and would need to be generalized considerably. Further, it would be interesting to see whether the dilation symmetry we analyzed can be used to obtain restrictions on the type of interaction in the limit theory, possibly leading to criteria for asymptotic freedom. Here we do not refer to restrictions on the form of the Lagrangian, a concept that is not visible in our framework. Rather, we think that dilation symmetries should manifest themselves in the coefficients of the operator product expansion [Bos05a,BDM09,BF08] or in the general structure of local observables [BF77]. In this context, it seems worthwhile to investigate simplified low-dimensional interacting models, such as the 1+1 dimensional massive models with factorizing S matrix that have recently been rigorously constructed in the algebraic framework [Lec08]. By abstract arguments, these models possess a scaling limit theory in our context. Just as in the Schwinger model [Buc96b,BV98], one expects here that even the limit theory for multiplicative states has a nontrivial center. This may be seen as a peculiarity of the 1+1 dimensional situation; we have not specifically dealt with this problem in the present paper. But neglecting these aspects, one would expect that the limit theory corresponds to a massless (and dilation covariant) model with factorizing S matrix, although it would probably not have an interpretation in the usual terms of scattering theory. Such models of “massless scattering” have indeed been considered in the physics literature; see [FS94] for a review. Their mathematical status as quantum field theories, however,
56
H. Bostelmann, C. D’Antoni, G. Morsella
remains largely unclear at this time. Nevertheless, one should be able to treat them with our methods. In fact, these examples may give a hint to the restrictions on interaction that the dilation symmetry implies. At least in a certain class of two-particle S matrices—those which tend to 1 at large momenta—one expects that the limit theory is chiral, i.e., factors into a tensor product of two models living on the left and right light ray, respectively. On the other hand, the theories we consider are always local; and for chiral local models, dilation covariance is the essential property that guarantees conformal covariance [GLW98]. So if those models have a nontrivial scaling limit at all, they underly quite rigid restrictions, since local conformal chiral nets are—at least partially—classifiable in a discrete series [KL04]. The detailed investigation of these aspects of factorizing S matrix models is however the subject of ongoing research, and some surprises are likely to turn up. Acknowledgements. The authors are obliged to Laszlo Zsido and Michael Müger for helpful discussions. They also profited from financial support by the Erwin Schrödinger Institute, Vienna, and from the friendly atmosphere there. HB further wishes to thank the II. Institut für Theoretische Physik, Hamburg, for their hospitality.
A. Direct Integrals of Hilbert Spaces In our investigation, we make use of the concept of direct integrals of Hilbert spaces, H = Z dν(z)Hz , where the integral is defined on some measure space (Z, ν). Due to difficult measure-theoretic problems, the standard literature treats these direct integrals only under separability assumptions on the Hilbert spaces involved; see e.g. [KR97]. These are however not a priori implied in our analysis; and even where we make such assumptions, we need to apply them with care. While we can often reasonably assume the “fiber spaces” Hz to be separable, the measure space Z will, in our applications, be of a very general nature, and even L 2 (Z, ν) is known to be non-separable in some situations. The concept of direct integrals can be generalized to that case. Since however the literature on that topic6 is somewhat scattered and not easily accessible, we give here a brief review for the convenience of the reader, restricted to the case that concerns us. In the following, let Z be a compact topological space and ν a finite regular Borel measure on Z. For each z ∈ Z, we consider a Hilbert space Hz with scalar product · | · "z and associated norm · z . Elements χ ∈ z∈Z Hz will be called vector fields and alternatively denoted as maps, z → χ (z). Direct integrals of this field of Hilbert spaces Hz over Z are not unique, but depend on the choice of a fundamental family. Definition A.1. A fundamental family is a linear subspace Γ ⊂ z∈Z Hz such that for every χ ∈ Γ , the function z → χ (z)2z is ν-integrable. If the same function is always continuous, we say that Γ is a continuous fundamental family. The continuity aspect will be discussed further below, for the moment we focus on measurability. Eachfundamental family Γ uniquely extends to a minimal vector space Γ¯ , with Γ ⊂ Γ¯ ⊂ z∈Z Hz , which has the following properties [Wil70, Corollary 2.3]: 6 In the general case, we largely follow Wils [Wil70], however with some changes in notation. Other, somewhat stronger notions of direct integrals exist, e.g. [God51, Ch. III], [Seg51]; see [Mar69] for a comparison. Under separability assumptions (Definition A.3), all these notions agree, and we are in the case described in [Tak79, Ch. IV.8], [Dix81, Part II Ch. 1].
On Dilation Symmetries Arising from Scaling Limits
(i) (ii) (iii) (iv)
57
z → χ (z)2z is ν-integrable for all χ ∈ Γ¯ . If for χ ∈ z Hz , there exists χˆ ∈ Γ¯ such that χ (z) = χˆ (z) a.e., then χ ∈ Γ¯ . If χ ∈ Γ¯ and f ∈ L ∞ (Z, ν), then f χ ∈ Γ¯ , where ( f χ )(z) := f (z)χ (z). Γ¯ is complete with respect to the seminorm χ = ( Z dν(z)χ (z)2 )1/2 .
Such Γ¯ is called an integrable family. It is obtained from Γ by multiplication with L ∞ functions and closure in · . The elements of Γ¯ are called Γ -measurable functions; they are in fact analogues of square-integrable functions, and the usual measure theoretic results hold for them: Egoroff’s theorem; the dominated convergence theorem; any norm-convergent sequence (χn ) in Γ¯ has a subsequence on which χn (z) converges pointwise a.e.; and if (χn ) is a sequence in Γ¯ that converges pointwise a.e., the limit function χ is in Γ¯ . (Cf. Propositions 1.3 and 1.5 of [Mar69].) After dividing out vectors of zero norm (which we do not denote explicitly), Γ¯ becomes a Hilbert space, which we call the direct integral of the Hz with respect to Γ , and denote it as Γ H= dν(z)Hz , with scalar product (χ |χ) ˆ = dν(z) χ (z)|χ(z)" ˆ z . (A.1) Z
Z
Γ Correspondingly, the elements χ ∈ H are denoted as χ = Z χ (z)dν(z). We also consider bounded operators between such direct integral spaces. Let Hz , Hˆ z be two fields of Hilbert spaces over the same measure space Z, and let Γ, Γˆ be associated fundamental families. We call B ∈ z∈Z B(Hz , Hˆ z ) a measurable field of operators if ess supz B(z) < ∞, and if for every χ ∈ Γ¯ , the vector field z → B(z)χ (z) is Γˆ -measurable, i.e. an element of Γ¯ˆ . In fact it suffices to check the measurability condition on the fundamental family Γ . Lemma A.2. Let B ∈ z∈Z B(Hz , Hˆ z ) such that ess supz B(z) < ∞, and such that z → B(z)χ (z) is Γˆ -measurable for every χ ∈ Γ . Then B is a measurable field of operators. Proof. Evidently, z → B(z)χ (z) is also Γˆ -measurable if χ is taken from L ∞ (Z, ν) · Γ or from its linear span. This span is however dense in H. So let χ ∈ H. There exists a sequence χn ∈ span L ∞ (Z, ν) · Γ such that χn → χ in norm; by the remarks after Def. A.1, we can assume that χn (z) → χ (z) a. e. But then B(z)χn (z) → B(z)χ (z) a. e., due to continuity of each B(z). So B(z)χ (z) is a pointwise a. e. limit of functions in Γ¯ˆ . This implies (z → B(z)χ (z)) ∈ Γ¯ˆ , which was to be shown. A measurable field of operators B now defines a bounded linear operator H → Hˆ
Γ,Γˆ ˆ of this form are which we denote as B = Z dν(z)B(z). Operators in B(H, H) called decomposable. Their decomposition need not be unique however, not even a. e. If here Γ = Γˆ , and if B(z) = f (z)1Hz with a function f ∈ L ∞ (Z , ν), then B is called a diagonalizable operator. We sometimes write this multiplication operator as
Γ M f = Z dν(z) f (z)1. Let A be a C∗ algebra, and let for each z ∈ Z a representation πz of A on Hz be given, such that z → πz (A) is a measurable field of operators for any A ∈ A. Then,
Γ π(A) = Z dν(z)πz (A) defines a new representation π of A on H, which we formally
Γ denote as π = Z dν(z)πz .
58
H. Bostelmann, C. D’Antoni, G. Morsella
In many cases, obtaining useful results regarding decomposable operators requires additional separability assumptions. The following property will usually be general enough for us.
Γ Definition A.3. A fundamental sequence in H = Z dν(z)Hz is a sequence (χ j ) j∈N in Γ¯ such that for every z ∈ Z, the set {χ j (z) | j ∈ N} is total in Hz . If such a fundamental sequence exists for H, we say that H is uniformly separable. This more restrictive situation agrees with the setting of [Tak79,Dix81]; cf. [Mar69, Prop. 1.13]. Note that this property implies that the fiber spaces Hz are separable, but the integral space H does not need to be separable if Z is sufficiently general. Under the above separability assumption, additional desirable properties of decomposable operators hold true.
Γ
Γˆ Theorem A.4. Let H = Z dν(z)Hz and Hˆ = Z dν(z)Hˆ z both be uniformly sepa Γ,Γˆ rable. Then, for each decomposable operator B = Z dν(z)B(z), also B ∗ is decom Γˆ ,Γ posable, with B ∗ = Z dν(z)B(z)∗ . One has B = B ∗ = ess supz B(z).
Γ,Γˆ Decompositions of operators are unique in the following sense: If Z dν(z)B(z) =
Γ,Γˆ ˆ ˆ dν(z) B(z), then B(z) = B(z) for almost every z. Z
For the proof methods, see e.g. [God51, Ch. III Sec. 13]. Note that the theorem is false if the separability assumption is dropped; see Example 7.6 and Remark 7.11 of [Tak79, Ch. IV]. We also obtain an important characterization of decomposable operators. ˆ is decomTheorem A.5. Let H, Hˆ be uniformly separable. An operator B ∈ B(H, H) posable if and only if it commutes with all diagonalizable operators; i.e. M f B = B Mˆ f for all f ∈ L ∞ (Z , ν). ˆ A proof can be found in [Dix81, Ch. II §2 Sec. 5 Thm. 1]. In particular, if H = H, we know that both the decomposable operators and the diagonalizable operators form W∗ algebras, which are their mutual commutants. Note that the “if” part of the theorem is known to be false for sufficiently general direct integrals, violating the separability assumption [Sch90]. We now discuss the case of a continuous fundamental family Γ ; cf. [God51, Ch. III Sec. 2]. In this case, we can consider the space of Γ -continuous functions, denoted C(Γ ), and defined as the closed span of C(Z) · Γ in the supremum norm, χ ∞ = supz∈Z χ (z)z . With this norm, C(Γ ) is a Banach space, in fact a Banach module over the commutative C∗ algebra C(Z). We have C(Γ ) ⊂ Γ¯ in a natural way, and this inclusion is dense, but it is important to note that different norms are used in these two spaces. A simple but particularly important example for direct integrals arises as follows [Tak79, Ch. IV.7]. Let Hu be a fixed Hilbert space, and Z a measure space as above. For each z ∈ Z, set Hz = Hu . Then the set Γ of constant functions Z → Hu is a continuous fundamental family; and the associated integrable family Γ¯ is precisely the space of all square-integrable, Lusin-measurable functions χ : Z → Hu . We denote ⊕ the corresponding direct integral space as L 2 (Z, ν, Hu ) = Z dν(z)Hu (with reference to the “canonical” fundamental family). This space is isomorphic to L 2 (Z, ν) ⊗ Hu ; the canonical isomorphism, which we do not denote explicitly, maps f ⊗ χ to the
On Dilation Symmetries Arising from Scaling Limits
59
function z → f (z)χ . In this way, the algebra of diagonal operators is identified with L ∞ (Z, ν) ⊗ 1. If here Hu is separable, then L 2 (Z, ν, Hu ) is clearly uniformly separable. In this case, a simple criterion identifies the elements of the integral space: A function χ : Z → Hu is Lusin measurable if and only if it is weakly measurable, i.e. if z → χ (z)|η" is measurable for any fixed η ∈ Hu . Also, the algebra of decomposable operators is iso¯ morphic to L ∞ (Z, ν)⊗B(H u ). References [Ara64] [BDF87] [BDM09] [BF77] [BF08] [Bos00] [Bos05a] [Bos05b] [BP90] [BR79] [BR81] [Buc96a] [Buc96b] [BV95] [BV98] [BW86] [CM08] [Dix81] [DM06] [DMV04] [DS85] [EF74] [FS94]
[GLW98] [God51] [Haa96]
Araki, H.: On the algebra of all local observables. Prog. Theor. Phys. 32, 844–854 (1964) Buchholz, D., D’Antoni, C., Fredenhagen, K.: The universal structure of local algebras. Commun. Math. Phys. 111, 123–135 (1987) Bostelmann, H., D’Antoni, C., Morsella, G.: Scaling algebras and pointlike fields. A nonperturbative approach to renormalization. Commun. Math. Phys. 285, 763–798 (2009) Buchholz, D., Fredenhagen, K.: Dilations and interactions. J. Math. Phys. 18, 1107–1111 (1977) Bostelmann, H., Fewster, C.J.: Quantum inequalities from operator product expansions. Commun. Math. Phys. (2009). doi:10.1007/s00220-009-0853-x Bostelmann, H.: Lokale Algebren und Operatorprodukte am Punkt. Thesis, Universität Göttingen, 2000. Available online at http://webdoc.sub.gwdg.de/diss/2000/bostelmann/ Bostelmann, H.: Operator product expansions as a consequence of phase space properties. J. Math. Phys. 46, 082304 (2005) Bostelmann, H.: Phase space properties and the short distance structure in quantum field theory. J. Math. Phys. 46, 052301 (2005) Buchholz, D., Porrmann, M.: How small is the phase space in quantum field theory? Ann Inst. H. Poincaré 52, 237–257 (1990) Bratteli, O., Robinson, D.W.: Operator Algebras and Quantum Statistical Mechanics, Volume I. New York: Springer, 1979 Bratteli, O., Robinson, D.W.: Operator Algebras and Quantum Statistical Mechanics, Volume II. New York: Springer, 1981 Buchholz, D.: Phase space properties of local observables and structure of scaling limits. Ann. Inst. H. Poincaré 64, 433–459 (1996) Buchholz, D.: Quarks, gluons, colour: facts or fiction? Nucl Phys. B 469, 333–353 (1996) Buchholz, D., Verch, R.: Scaling algebras and renormalization group in algebraic quantum field theory. Rev. Math. Phys. 7, 1195–1239 (1995) Buchholz, D., Verch, R.: Scaling algebras and renormalization group in algebraic quantum field theory. II. Instructive examples. Rev. Math. Phys. 10, 775–800 (1998) Buchholz, D., Wichmann, E.H.: Causal independence and the energy-level density of states in local quantum field theory. Commun. Math. Phys. 106, 321–344 (1986) Conti, R., Morsella, G.: Scaling limit for subsystems and Doplicher-Roberts reconstruction. Ann. H. Poincaré 10, 485–511 (2009) Dixmier, J.: Von Neumann Algebras. Amsterdam: North-Holland, 1981 D’Antoni, C., Morsella, G.: Scaling algebras and superselection sectors: Study of a class of models. Rev. Math. Phys. 18, 565–594 (2006) D’Antoni, C., Morsella, G., Verch, R.: Scaling algebras for charged fields and short-distance analysis for localizable and topological charges. Ann. H. Poincaré 5, 809–870 (2004) Driessler, W., Summers, S.J.: Central decomposition of Poincaré-invariant nets of local field algebras and absence of spontaneous breaking of the Lorentz group. Ann. Inst. H. Poincaré Phys. Theor. 43, 147–166 (1985) Eckmann, J.P., Fröhlich, J.: Unitary equivalence of local algebras in the quasifree representation. Ann. Inst. H. Poincaré Sect. A (N.S.) 20, 201–209 (1974) Fendley, P., Saleur, H.: Massless integrable quantum field theories and massless scattering in 1+1 dimensions. In: Gava, E., Masiero, A., Nariain, K.S., Randjbar-Daemi, S., Shafi, Q. (eds.), Proceedings of the 1993 Summer School on High Energy Physics and Cosmology, Volume 10 of ICTP Series in Theoretical Physics, Singapore: World Scientific, 1994 Guido, D., Longo, R., Wiesbrock, H.W.: Extensions of conformal nets and superselection structures. Commun. Math. Phys. 192, 217–244 (1998) Godement, R.: Sur la théorie des représentations unitaires. Ann. of Math. 53, 68–124 (1951) Haag, R.: Local Quantum Physics. Berlin: Springer, 2nd edition, 1996
60
[HS65] [KL04] [KR97] [Lec08] [Mar69] [Mit66] [Sch90] [Seg51] [Tak79] [Tak02] [Wil70]
H. Bostelmann, C. D’Antoni, G. Morsella
Haag, R., Swieca, J.A.: When does a quantum field theory describe particles? Commun. Math. Phys. 1, 308–320 (1965) Kawahigashi, Y., Longo, R.: Classification of local conformal nets. Case c < 1. Ann. Math. 160, 493–522 (2004) Kadison, R.V., Ringrose, J.R.: Fundamentals of the Theory of Operator Algebras, Volume II: Advanced Theory. Orlando FL: Academic Press, 1997 Lechner, G.: Construction of quantum field theories with factorizing S-matrices. Commun. Math. Phys. 277, 821–860 (2008) Maréchal, O.: Champs mesurables d’espaces hilbertiens. Bull. Sci. Math. 93, 113–143 (1969) Mitchell, T.: Fixed points and multiplicative left invariant means. Trans. Amer. Math. Soc. 122, 195–202 (1966) Schaflitzel, R.: The algebra of decomposable operators in direct integrals of not necessarily separable Hilbert spaces. Proc. Amer. Math. Soc. 110, 983–987 (1990) Segal, I.: Decompositions of Operator Algebras I, II, Volume 9 of Mem. Amer. Math. Soc. Providence, RI: Amer. Math. Soc., 1951 Takesaki, M.: Theory of Operator Algebras I. Berlin: Springer, 1979 Takesaki, M.: Theory of Operator Algebras II. Berlin: Springer, 2002 Wils, W.: Direct integrals of Hilbert spaces I. Math. Scand. 26, 73–88 (1970)
Communicated by Y. Kawahigashi
Commun. Math. Phys. 294, 61–72 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0934-x
Communications in
Mathematical Physics
Blowup of Jang’s Equation at Outermost Marginally Trapped Surfaces Jan Metzger Albert-Einstein-Institut, Am Mühlenberg 1, D-14476 Potsdam, Germany. E-mail:
[email protected] Received: 30 January 2009 / Accepted: 24 July 2009 Published online: 8 October 2009 – © The Author(s) 2009. This article is published with open access at Springerlink.com
Abstract: The aim of this paper is to accurately describe the blowup of Jang’s equation. First, we discuss how to construct solutions that blow up at an outermost MOTS. Second, we exclude the possibility that there are extra blowup surfaces in data sets with non-positive mean curvature. Then we investigate the rate of convergence of the blowup to a cylinder near a strictly stable MOTS and show exponential convergence with an identifiable rate near a strictly stable MOTS. 1. Introduction This paper is concerned with the examination of the relation of Jang’s equation to marginally outer trapped surfaces (MOTS). To set the perspective, we consider Cauchy data (M, g, K ) for the Einstein equations. Such data sets are 3-manifolds M equipped with a Riemannian metric together with a symmetric bilinear form K representing the second fundamental form of the time slice M in space-time. A marginally outer trapped surface is a surface with θ + = H + P = 0, where H is the mean curvature of in M and P = tr K − K (ν, ν) for the normal ν to . In the paper [AM07], inspired by an idea of Schoen [Sch04], we constructed MOTS in the presence of barrier surfaces by inducing a blow-up of Jang’s equation. In this context, Jang’s equation [SY81,Jan78] is an equation of prescribed mean curvature for the graph of a function in M × R. For details we refer to Sect. 2. In this note, we take a slightly different perspective. Consider a data set (M, g, K ) with a non-empty outer boundary ∂ + M and assume that we are given the outermost MOTS in (M, g, K ). Here, outermost means that there is no other MOTS on the outside of . From [AM07] it follows that (M, g, K ) always contains a unique such surface, or does not contain outer trapped surfaces at all, under the assumption that ∂ M is outer untrapped. As stated in Theorem 3.1, we show that there exists a solution f Research on this project started while the author was supported in part by a Feodor-Lynen Fellowship of the Humboldt Foundation.
62
J. Metzger
to Jang’s equation that actually blows up at , assuming that ∂ M is inner and outer untrapped. By blow-up we mean that outside from the function f is such that graph f is a smooth submanifold of M × R with a cylindrical end converging to × R. There is however a catch, as f may blow up at other surfaces, too. These surfaces are marginally inner trapped. In Theorem 3.4 we show that the other blow-up surfaces can not occur if the data set has non-positive mean curvature. To put the result in perspective note that if the dominant energy condition holds, the graph of f is of non-negative Yamabe type and thus can be equipped with a (singular) metric of zero scalar curvature. This was used by Schoen and Yau in [SY81] to prove the positive mass theorem. Later Bray [Bra01] proposed to use Jang’s equation to relate the Penrose conjecture in its general setting to the Riemannian Penrose inequality [HI01,Bra01] on a manifold constructed from Jang’s graph. One of the main questions in this program is whether or not Jang’s equation can be made to blow up at a specific MOTS. This question was raised in the literature, cf. for example [MÓM04] where this is discussed in the rotationally symmetric case. Here we give the positive answer that blow-up solutions exist at outermost MOTS. The author recently learned that the existence of the blow-up solution is used in [Khu09] to prove a Penrose-like inequality. With the blow-up constructed, we can turn to the asymptotic behavior of the blowup itself. It has been shown in [SY81] that such a blow-up must be asymptotic to a cylinder over the outermost MOTS. In Theorems 4.2 and 4.4 we show that under the assumption of strict stability the convergence rate is exponential with a power directly related to the principal eigenvalue of the MOTS. The general idea is to show the existence of a super-solution with at most logarithmic blow-up of the desired rate. Turning the picture sideways yields exponential decay, when writing the solution as a graph over the cylinder in question. Furthermore, we show that beyond a certain decay rate, the solution must be trivial, thus exhibiting the actual rate. We expect that the knowledge of these asymptotics is tied to the question whether the blow-up solution is unique. Furthermore note that the constant in the Penrose-like inequality in [Khu09] depends on the geometry of the solution. We thus expect that the value on this constant is related to the asymptotic behavior near the blow-up cylinder. Before turning to these results, we introduce some notation in Sect. 2. Section 3 proceeds with the construction of the the blow-up. We will not go into details here, but emphasize the general idea and point to the results needed from the paper [AM07]. In Sect. 4, we perform the calculation of the asymptotics. 2. Preliminaries Let (M, g, K ) be an initial data set for the Einstein equations. That is M is a 3-dimensional manifold, g a Riemannian metric on M and K a symmetric 2-tensor. We do not require any energy condition to hold. Assume that ∂ M is the disjoint union ∂ M = ∂ − M ∪ ∂ + M, where ∂ ± M are smooth surfaces without boundary. We refer to ∂ − M as the inner boundary and endow it with the normal vector ν pointing into M. The outer boundary ∂ + M is endowed with the normal ν pointing out of M. We denote by H [∂ M] the mean curvature of ∂ M with respect to the normal vector field ν, and by P[∂ M] = tr ∂ M K the trace of the tensor K restricted to the 2-dimensional surface ∂ M. Then the inward and outward expansions of ∂ M are defined by θ ± [∂ M] = P[∂ M] ± H [∂ M]. Assume that θ + [∂ − M] = 0, and that θ + [∂ + M] > 0 and θ − [∂ + M] < 0.
Blowup of Jang’s Equation at Outermost Marginally Trapped Surfaces
63
If ⊂ M is a smooth, embedded surface homologous to ∂ + M, then bounds a region together with ∂ + M. In this case, we define θ ± [] as above, where H is computed with respect to the normal vector field pointing into (that is in direction of ∂ + M). is called a marginally outer trapped surface (MOTS), if θ + [] = 0. We say that ∂ M is an outermost MOTS, if there is no other MOTS in M, which is homologous to ∂ + M. In [AM07] it is proved that for any initial data set (M, g, K ) which contains a MOTS, there is also an outermost MOTS surrounding it. Let ⊂ M be a MOTS and consider a normal variation of in M, that is a map ∂ F : × (−ε, ε) → M such that F(·, 0) = id and ∂s F( p, s) = f ν, where f is s=0 a function on and ν is the normal of . Then the change of θ + is given by ∂θ + [F(, s)] = L M f, ∂ds s=0 where L M is a quasi-linear elliptic operator of second order along . It is given by 1 L M f = − f + 2S(∇ f ) + f div S − |χ + |2 − |S|2 + Sc − µ − J (ν) . 2 In this expression ∇, div and denote the gradient, divergence and Laplace-Beltrami operator tangential to . The tangential 1-form S is given by S = K (·, ν)T , χ + is the bilinear form χ + = A + K , where A is the second fundamental form of in M and K is the projection of K to T × T . Furthermore, Sc denotes the scalar curvature of , µ = 21 ( M Sc − |K |2 + (tr K )2 ), and J = M divK − d tr K . For a more detailed investigation of this operator we refer to [AMS05] and [AMS07]. The facts we will need here are that L M has a principal eigenvalue λ, which is real and has a one-dimensional eigenspace which is spanned by a positive function. If λ is non-negative is called stable, and if λ is positive, is called strictly stable. In particular, if is strictly stable as a MOTS, there exists an outward deformation strictly increasing θ + . In M¯ = M × R, we consider Jang’s equation [Jan78,SY81] for the graph of a function f : M → R. Let N := graph f = {(x, z) : z = f (x)}. The mean curvature H[ f ] of N with respect to the downward normal is given by
∇f
H[ f ] = div 1 + |∇ f |2
.
Define K¯ on M¯ by K¯ (x,z) (X, Y ) = K x (π X, π Y ), where π : T M¯ → T M denotes the orthogonal projection onto the horizontal tangent vectors. Let P[ f ] = tr N K¯ . Then Jang’s equation becomes J [ f ] = H[ f ] − P[ f ] = 0.
(2.1)
64
J. Metzger
Fig. 1. The situation in Theorem 3.1. All of the shaded region belongs to M, whereas f is only defined in 0
3. The Blowup The main result of this section is that we can construct a solution to Jang’s equation which blows up at the outermost MOTS in (M, g, K ) and has zero Dirichlet boundary data at ∂ + M. In fact, we chose the assumptions on the outer boundary ∂ + M so that we can prescribe more general Dirichlet data there. The focus here lies on the blow-up in the interior, so that we do not investigate the optimal conditions for ∂ + M. Theorem 3.1. If (M, g, K ) is an initial data set with ∂ M = ∂ − M ∪ ∂ + M such that ∂ − M is an outermost MOTS, θ + [∂ + M] > 0 and θ − [∂ + M] < 0, then there exists an open set 0 ⊂ M and a function f : 0 → R such that 1. 2. 3. 4. 5. 6.
M\0 does not intersect ∂ M, θ − [∂0 ] = 0 with respect to the normal vector pointing into 0 , J [ f ] = 0, N + = graph f ∩ M × R+ is asymptotic to the cylinder ∂ − M × R+ , N − = graph f ∩ M × R− is asymptotic to the cylinder ∂0 × R− , and f |∂ + M = 0.
For data sets (M, g, K ) which do not contain surfaces with θ − = 0, the above theorem implies the following result. Corollary 3.2. If (M, g, K ) is as in Theorem 3.1, and in addition there are no subsets ⊂ M with θ − [∂] = 0 with respect to the normal pointing out of , then there exists a function f : M → R such that 1. J [ f ] = 0, 2. N = graph f is asymptotic to the cylinder ∂ − M × R+ , 3. f |∂ + M = 0. Remark 3.3. Analogous results hold if (M, g, K ) is asymptotically flat with appropriate decay of g and K instead of having an outer boundary ∂ + M. Then the assertion f |∂ + M = 0 in Theorem 3.1 has to be replaced by f (x) → 0 as x → ∞. The proof of Theorem 3.1 is largely based on the tools developed in [SY81 and AM07]. Thus we will not include all details here, but provide a summary, which facts will have to be used.
Blowup of Jang’s Equation at Outermost Marginally Trapped Surfaces
65
Proof. We will assume that (M, g, K ) is embedded into (M , g , K ) which extends M beyond the boundary ∂ − M such that ∂ − M lies in the interior of M , without further requirements. N where the are the connected components of ∂ M. As ∂ M is Let ∂ − M = ∪i=1 i i an outermost MOTS, each of the i is stable [AM07, Cor. 5.3]. Following the proof of [AM07, Th. 5.1], we deform ∂ − M to a surface s by pushing the components i out of M, into the extension M . To this end, let φi > 0 be the principal eigenfunction of the stability operator of i and extend the vector field X i = −φi νi to a neighborhood of i in M . Flowing i by X i yields a family of surfaces is , s ∈ [0, ε) so that the is form a smooth foliation for small enough ε with is ∈ M \ M. If i is strictly stable then ∂ θ + [s ] = −λφ < 0, ∂s s=0 where λ is the principal eigenvalue of i . Thus, for small enough ε, we have θ + [is ] < 0 for all s ∈ (0, ε). If i has principal eigenvalue λ = 0, then the is satisfy ∂ θ + [s ] = 0. ∂s s=0 In this case it is possible to change the data K on is as follows: K˜ = K − 21 ψ(s)γs ,
(3.1)
where γs is the metric on s and ψ is a smooth function ψ : [0, ε] → R. The operator θ˜ + , which means θ + computed with respect to the data K˜ instead of K , satisfies θ˜ + [is ] = θ + [is ] − ψ(s). It is clear from Eq. (3.1) that ψ can be chosen such that ψ(0) = ψ (0) = 0 and θ˜ + [is ] < 0 for all s ∈ (0, ε) provided ε is small enough. Then K˜ is C 1,1 when extended by K to the rest of M. Replace each original boundary component i of M by a surface iε as constructed above, and replace K with K˜ , such that the following properties are satisfied. Let M˜ denote the manifold with boundary components iε resulting from this procedure. Thus ˜ g , K˜ ) with the following properties: we construct from (M, g, K ) a data set ( M, ˜ 1. M ⊂ M˜ with g | M = g, K˜ | M = K , and ∂ + M = ∂ + M, + − ˜ 2. θ [∂ M] < 0, and 3. the region M˜ \ M is foliated by surfaces s with θ + (s ) < 0. The method developed in Sect. 3.2 in [AM07] now allows the modification of the data ˜ g, ˜ g , K˜ ) to a new data set, which we also denote by ( M, ˜ K˜ ), although K˜ changes ( M, in this step. This data set has the following properties: ˜ 1. M ⊂ M˜ with g | M = g, K˜ | M = K , and ∂ + M = ∂ + M, ˜ < 0, 2. θ + [∂ − M] ˜ > 0, where H is the mean curvature of ∂ − M with respect to the normal 3. H [∂ − M] ˜ pointing out of ∂ − M, 4. the region M˜ \ M is foliated by surfaces s with θ + (s ) < 0.
66
J. Metzger
By Sect. 3.3 in [AM07] this enables us to solve the boundary value problem ⎧ ⎪ ⎨J [ f τ ] = τ f τ in M˜ f τ = 2τδ on ∂ − M˜ ⎪ ⎩f =0 on ∂ + M˜ τ
(3.2)
where δ is a lower bound for H on ∂ − M. The solvability of this equation follows, provided an estimate for the gradient at the boundary can be found. The barrier construction at ∂ − M˜ was carried out in detail in [AM07], whereas the barrier construction at ∂ + M˜ is standard due to the stronger requirement that θ + [∂ + M] > 0 and θ − [∂ + M] < 0. The solution f τ to Eq. (3.2) satisfies an estimate of the form sup | f τ | + sup |∇ f τ | ≤ M˜
M˜
C , τ
(3.3)
˜ g, where C is a constant depending only on the data ( M, ˜ K˜ ) but not on τ . The gradient estimate implies in particular that there exists an ε > 0 independent of τ such that f τ (x) ≥
δ 4τ
˜ ∀x with dist(x, ∂ − M).
The graphs Nτ have uniformly bounded curvature in M˜ × R away from the boundary. This allows the extraction of a sequence τi → 0 such that the Nτi converge to a manifold N , cf. [AM07, Prop. 3.8], [SY81, Sect. 4]. This convergence determines three open ˜ subsets of M: − := {x ∈ M : f τi (x) → −∞ locally uniformly as i → ∞}, 0 := {x ∈ M : lim sup | f τi (x)| < ∞}, i→∞
+ := {x ∈ M : f τi (x) → ∞ locally uniformly as i → ∞}. ˜ we have that + = ∅ and + contains a From the fact that the f τ blow up near ∂ − M, ˜ As already noted in [SY81] ∂+ \ ∂ M˜ consists of MOTS. As the neighborhood of ∂ − . region M˜ \ M is foliated by surfaces with θ + < 0, we must have that + ⊃ ( M˜ \ M) and hence ∂+ is a MOTS in M. As ∂ − M was assumed to be an outermost MOTS in M, we conclude that the closure of + is M˜ \ M. The barriers near ∂ + M are so that they imply that the f τ are uniformly bounded near + ∂ M. Thus 0 contains a neighborhood of ∂ + M and 0 ⊂ M. The limit manifold N over 0 is a graph satisfying J [ f τ ] = 0, and has the desired asymptotics. We will now discuss a geometric condition to assert that the resulting graph is nonsingular on M, i.e., M = 0 in Theorem 3.1. Theorem 3.4. Let (M, g, K ) be as in Theorem 3.1 with tr K ≤ 0. Then in the assertion of Theorem 3.1 we have that 0 = M, that is f is defined on M and has no other blow-up than near ∂ − M.
Blowup of Jang’s Equation at Outermost Marginally Trapped Surfaces
67
Proof. This follows from a simple argument using the maximum principle. Let f τ be a solution to the regularized problem H[ f τ ] − P[ f τ ] − τ f τ = 0
(3.4)
˜ as in the proof of Theorem 3.1. We claim that f τ can not have a negative minimum in M, in the region where the data is unmodified. Assume that x ∈ M is such a minimum. There we have H[ f τ ] ≥ 0, and since graph f is horizontal at x we have that P[ f τ ] = tr K ≤ 0, thus the right-hand side of (3.4) is non-negative, whereas τ f τ is assumed to be negative, a contradiction. Since we know that in the limit τ → 0, the functions f τ must blow-up in the modified region which lies in + , we infer a lower bound for f τ from the above argument. Thus 0 = M as claimed. 4. Asymptotic Behavior Here, we discuss a refinement of [SY81, Cor. 2], which says that N = graph f converges uniformly in C 2 to the cylinder ∂ − M × R for large values of f . A barrier construction allows us to determine the asymptotics of this convergence. Before we present our result, recall the statement of [SY81, Cor. 2]: Theorem 4.1. Let N = graph f be the manifold constructed in the proof of Theorem 3.1 and let be a connected component of ∂ − M. Let U be a neighborhood of with positive distance to ∂ − M \ . Then for all ε > 0 there exists z¯ = z¯ (ε), depending also on the geometry of (M, g, K ), such that N ∩ U × [¯z , ∞) can be written as the graph of a function u over C z¯ := × [¯z , ∞), so that |u( p, z)| + | C z¯ ∇ u( p, z)| + | C z¯ ∇ 2 u( p, z)| < ε for all ( p, z) ∈ C z¯ . Here,
C z¯
∇ denotes covariant differentiation along C z¯ .
If is strictly stable, we can in fact say more about u. Theorem 4.2. Assume the situation of Theorem 4.1. If in addition is strictly stable √ with principal eigenvalue λ > 0, we have that for all δ < λ there exists c = c(δ) depending only on the data (M, g, K ) and δ such that |u( p, z)| + | C z¯ ∇ u( p, z)| + | C z¯ ∇ 2 u( p, z)| ≤ c exp(−δz). Proof. Denote by β > 0 the eigenfunction to the principal eigenvalue λ on normalized such that max β = 1. We denote by ν the normal vector field of pointing into M. Consider the map : × [0, s¯ ] → M : ( p, s) → exp M p (sβν).
(4.1)
Given ε > 0 we can choose s¯ > 0 small enough such that the surfaces s = (, s) with s ∈ [0, s¯ ] form a local foliation near with lapse β such that θ + [s ] ≥ λ(1 − ε)βs.
68
J. Metzger
Denote the region swiped out by these s by Us¯ . Note that ∂Us¯ = ∪ s¯ and dist(s¯ , ) > 0. We can assume that dist(s¯ , ∂ M) > 0. On Us¯ we consider functions w of the form w = φ(s). For such functions Jang’s operator can be computed as follows: φ + φ φ θ − 1+ P − σ −2 K (ν, ν) + 2 3 , J [w] = βσ βσ β σ where σ 2 = 1+β −2 (φ )2 , and φ denotes the derivative of φ with respect to s, cf. [AM07]. The quantities θ + , K (ν, ν) and P are computed on the respective s . Note that with our normalization σ −2 ≤ β 2 |φ |−2 ≤ |φ |−2 , and if we assume that φ ≥ µ for a large µ = µ(β) we have 1 + φ ≤ 2|φ |−2 . βσ Furthermore, φ |φ | |φ | |φ | β 2 σ 3 = β 2 (1 + β −2 φ 2 )3/2 ≤ β 2 (β −2 φ 2 )3/2 = β |φ |3 . On the other hand, increasing µ = µ(β, ε) if necessary, we have |φ | ≥ 1 − ε, βσ if |φ | ≥ µ. In combination we find that J [w] ≤ −λ(1 − ε)βs +
c1 |φ | + β , |φ |2 |φ |3
(4.2)
with c > 0 depending on ε and the data (M, g, K ), provided |φ | ≥ µ and φ < 0. Choosing φ(s) = a log s with a = (1 − ε)−1 λ−1/2 , we calculate that φ (s) =
a a , φ (s) = − 2 , s s
so that 1 s2 = = (1 − ε)2 λs 2 , |φ |2 a2
φ s = 2 = (1 − ε)2 λs. 3 |φ | a
Thus we can choose s¯ so small that |φ | ≥ µ(β, ε) and the estimate in (4.2) holds. We can then decrease s¯ further, so that s¯ ≤ εβ/(c1 (1−ε)). This choice makes the right-hand side of (4.2) non-positive, that is J [w] ≤ 0. Hence, we obtain a super-solution w with Jτ w ≤ 0 at least where w ≥ 0, that is near .
Blowup of Jang’s Equation at Outermost Marginally Trapped Surfaces
69
As w blows up near the horizon, and the f τ are bounded uniformly in τ on s¯ , we can translate w vertically to w¯ = w + b with a suitable b > 0 so that ¯ s¯ f τ |s¯ ≤ w| for all τ > 0. Then the maximum principle implies that f τ ≤ w¯ for all τ > 0 in Us¯ and consequently the function f constructed in Theorem 3.1 also satisfies f ≤ w. ¯ Near , the graph of w¯ can be written as the graph of a function v¯ over × (¯z , ∞), where v decays exponentially in z. This is due to the fact that by the assumptions on β, the parameter s is comparable to the distance to . By the above construction u ≤ v, where u is the function from Theorem 4.1. Thus we find the claimed estimate for u. Getting the desired estimates for the derivatives of u is then a standard procedure, but as it is a little work to set the stage, we briefly indicate how to proceed. We choose coordinates of a neighborhood × R in a slightly different manner as ¯ : × (−ε, ε) → M be the map above. Let ¯ : × (−ε, ε) × R → M × R : (x, s, z) → expx (sν), z . For a function h on C z¯ we let graph¯ h be the set ¯ graph¯ h = {(x, h(x), z) : (x, z) ∈ × R}. From Theorem 4.1, it is clear that for large enough z¯ the set N ∩ M × [¯z , ∞) can be written as graph¯ h, where h decays exponentially by the above reasoning. We can compute the value of Jang’s operator for h as follows: ¯ ( H¯ − P)[N ] = J h, where J is a quasi-linear elliptic operator of mean curvature type. To be more precise, J h has the form ij
ij
J h = ∂z2 h + γh(x,z) ∇i,2 j h − 2γh(x,z) ∂i (h)K (∂s , ∂ j ) − θ + [h(x,z) ] + Q(h,
C z¯
∇ h,
C z¯
(4.3)
∇ 2 h),
where γs is the metric on s and Q is of the form Q(h,
C z¯
∇ h,
C z¯
∇ 2 h) = h ∗
C z¯
∇h +
C z¯
∇h ∗
C z¯
∇h +
C z¯
∇h ∗
C z¯
∇h ∗
C z¯
∇ 2 h,
where ∗ denotes some contraction with a bounded tensor. Furthermore, the vectors ∂i , i = 1, 2 denote directions tangential to and ∂z the direction along the R-factor in C z¯ . By freezing coefficients, we therefore conclude that h satisfies a linear, uniformly elliptic equation of the form a i j ∂i ∂ j h + b,
C z¯
∇ h − θ + [h(x,z) ] = 0.
By construction we have that |θ + [s ]| ≤ κs for some fixed κ. Thus θ + [h(x,z) ] decays exponentially in z. Now we are in the position to use standard interior estimates for linear elliptic equations to conclude the decay of higher derivatives of h. This decay translates back into the decay of the first and second derivatives of u as the coordinate transformation is smooth and controlled by the geometry of (M, g, K ).
70
J. Metzger
Remark 4.3. If is not strictly stable, but has positive k th variation, we find that the foliation near satisfies θ + [s ] ≥ κs k . Then a function of the form φ(s) = as − p with large a and p = k−1 2 yields a super-solution. This super-solution can be used to prove that |u| ≤ C z 2/(1−k) as above. We can get even more information about the decay rate. A closer look at Eq. (4.3) yields that the expression for J h on C z¯ can also be written as follows: J h = (∂z2 − L M )h + Q (h,
C z¯
∇ h,
C z¯
∇ 2 h),
since θs+ = s L M 1 + O(s 2 ), ij
γh(x,z) ∇i,2 j h = h + Q 1 (h, ∇h, ∇ 2 h), and ij
γh(x,z) ∂i h K (∂s , ∂ j ) = S(∇h) + Q 2 (h, ∇h), where the differential operators ∇ and are with respect to . Then note that L M h = h L M 1 − h + 2S(∇h). Further investigation of the structure of Q yields that |Q (h,
C z¯
∇ h,
C z¯
∇ 2 h)| ≤ C |h|2 + | C z¯ ∇ h|2 + |h|| C z¯ ∇ 2 h| + | C z¯ ∇ h|2 | C z¯ ∇ 2 h| ,
so that in view of the differential Harnack estimate | C z¯ ∇ h| ≤ c|h| for positive solutions of linear elliptic equations we have that in fact |Q (h,
C z¯
∇ h,
C z¯
∇ 2 h)| ≤ c|h| |h| + | C z¯ ∇ h| + | C z¯ ∇ 2 h| ,
provided |h| ≤ C. By projecting the equation J h = 0 to the one-dimensional eigenspace of L M it is now a somewhat standard ODE argument to show the following result. Theorem 4.4. Under the assumptions of Theorem 4.2 there are no solutions h : × [0, ∞) → R to the equation Jh = 0 with decay |h( p, z)| + | C z¯ ∇ h( p, z)| + | C z¯ ∇ h( p, z)| ≤ C exp(−δz) such that δ >
√ λ and h > 0.
(4.4)
Blowup of Jang’s Equation at Outermost Marginally Trapped Surfaces
71
Proof. Assume that h > 0 is such a solution. We derive a contradiction as follows. Let λ be the principal eigenvalue and φ be the corresponding eigenfunction of L M as before. Let L ∗M be the (formal) adjoint of L M on L 2 () and denote by φ ∗ > 0 its principal eigenfunction, normalized such that φφ ∗ dµ = 1. Then the operator Pu =
φ ∗ u dµ φ
is a projection onto the eigenspace spanned by φ and moreover commutes with L M . We interpret h(z) as a family of functions on , that is h(z)( p) = h( p, z) for p ∈ . Choose α(z) such that Ph(z) = α(z)φ, and β(z) accordingly, β(z)φ = P Q (h(·, z),
C z¯
∇ h(·, z),
C z¯
∇ 2 h(·, p)) .
Then Eq. (4.4) and the fact that P commutes with L M and ∂z imply α (z) − λα(z) = β(z). Using φ ∗ > 0 and h > 0 yields α(z) > 0 and we can furthermore estimate that
φ ∗ |h( p, z)| |h( p, z)| + |∇h( p, z)| + |∇ 2 h( p, z)| dµ ≤ c exp(−δz) φ ∗ |h( p, z)| dµ
|β(z)| ≤ c
≤ c exp(−δz)α(z). Thus, we conclude that on [˜z , 0) the function α > 0 satisfies a differential inequality of the form α (z) − λα ≤ εα, √ where ε > 0 can be chosen arbitrarily small by choosing z˜ large enough. If λ + ε < δ this ODE has no solutions with decay exp(−δz) other than the trivial solution. Thus α ≡ 0 and we arrive at the desired contradiction. Acknowledgement. The author thanks the Mittag-Leffler-Institute, Djursholm, Sweden for hospitality and support during the program Geometry, Analysis, and General Relativity in Fall 2008. Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
72
J. Metzger
References [AM07] [AMS05] [AMS07] [Bra01] [HI01] [Jan78] [Khu09] [MÓM04] [Sch04] [SY81]
Andersson, L., Metzger, J.: The area of horizons and the trapped region. Commun. Math. Phys. 290, 941–972 (2009) Andersson, L., Mars, M., Simon, W.: Local existence of dynamical and trapping horizons. Phys. Rev. Lett. 95, 111102 (2005) Andersson, L., Mars, M., Simon, W.: Stability of marginally outer trapped surfaces and existence of marginally outer trapped tubes. http://arxiv.org/abs/0704.2889v2[gr-qc], 2007 Bray, H.L.: Proof of the riemannian penrose inequality using the positive mass theorem. J. Diff. Geom. 59(2), 177–267 (2001) Huisken, G., Ilmanen, T.: The inverse mean curvature flow and the riemannian penrose inequality. J. Diff. Geom. 59(3), 353–437 (2001) Jang, P.S.: On the positivity of energy in general relativity. J. Math. Phys. 19, 1152–1155 (1978) Khuri, M.: A penrose-like inequality for general initial data sets. Commun. Math. Phys. 290(2), 779–788 (2009) Malec, E., Murchadha, N.Ó.: The Jang equation, apparent horizons and the Penrose inequality. Class. Quant. Grav. 21(24), 5777–5787 (2004) Schoen, R.: Talk Given at the Miami Waves Conference, January 2004 Schoen, R., Yau, S.-T.: Proof of the positive mass theorem. II. Commun. Math. Phys. 79(2), 231–260 (1981)
Communicated by P. T. Chru´sciel
Commun. Math. Phys. 294, 73–95 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0954-6
Communications in
Mathematical Physics
Incompressible Limits and Propagation of Acoustic Waves in Large Domains with Boundaries Eduard Feireisl Institute of Mathematics of the Academy of Sciences of the Czech Republic, Žitná 25, 115 67 Praha 1, Czech Republic. E-mail:
[email protected] Received: 3 February 2009 / Accepted: 26 August 2009 Published online: 19 November 2009 – © Springer-Verlag 2009
Abstract: We study the incompressible limit for the full Navier-Stokes-Fourier system on unbounded domains with boundaries, supplemented with the complete slip boundary condition for the velocity field. Using an abstract result of Tosio Kato we show that the energy of acoustic waves decays to zero on any compact subset of the physical space. This in turn implies strong convergence of the velocity field to its limit in the incompressible regime.
1. Introduction Propagation and attenuation of acoustic waves plays an important role in the analysis of fluid flows in the low Mach number regime, in particular in the so-called incompressible limit, when the speed of sound in the material becomes infinite. Recently, several studies have been devoted to a rigorous justification of the low Mach number limit for the complete Navier-Stokes-Fourier system. Alazard [1,2] studied the problem for a quite general class of initial data that give rise to sufficiently regular solutions defined, however, only on a short time interval. In this purely “hyperbolic” approach proposed in the seminal paper by Klainerman and Majda [14], the presence of viscosity in the Navier-Stokes system plays only a marginal role. A different technique, based on the concept of weak solutions, was used by Lions and Masmoudi [22,23], and further developed in a series of papers by Desjardins and Grenier [7], Desjardins et al. [8], Masmoudi [24–26], among others. These results concern the Navier-Stokes system describing a compressible barotropic fluid flow, where the basic framework is provided by the existence theory developed by Lions [21]. The work of E.F. was supported by Grant 201/08/0315 of GA CR ˇ as a part of the general research programme of the Academy of Sciences of the Czech Republic, Institutional Research Plan AV0Z10190503.
74
E. Feireisl
The same strategy, based on global-in-time weak solutions, was later adapted to the complete Navier-Stokes-Fourier system in [10]. This approach leans on energy-entropy estimates, where the presence of viscosity is indispensable. As frequently observed in practical as well as numerical experiments, the influence of acoustic waves on the fluid motion is negligible in the low Mach number limit (cf. Klein et al. [17]). In terms of the mathematical theory, the acoustic waves are supported by the gradient part of the fluid velocity u, specifically H⊥ [u], where u = H[u] + H⊥ [u] and H is the Helmholtz projection onto the space of solenoidal functions. In the (hypothetical) incompressible limit, the mass density tends to a constant, the speed of sound becomes infinite whereas the gradient component H⊥ “disappears” since the limit velocity field is solenoidal. However, if the initial data are ill prepared, meaning the fluctuations of the fluid density and temperature are of the same order as the Mach number, and if the fluid is contained in a bounded spatial domain with an acoustically hard boundary (see Wilcox [37]), the gradient component of the velocity H⊥ [u] develops fast time oscillations with frequencies inversely proportional to the Mach number (cf. Lions and Masmoudi [22], Schochet [30,31]). Accordingly, the gradient part H⊥ [u] tends to zero only weakly, meaning in the sense of integral averages, with respect to time - a phenomenon that may destabilize certain numerical schemes when applied to the original system. A rather heuristic argument why the acoustic waves can be neglected in the low Mach number flows encountered in the real world applications arising in meteorology, oceanography, or astrophysics, is usually based on the fact that the underlying physical space is unbounded or, more correctly, sufficiently large when compared to the sound speed in the material in question (cf. Klein [15,16]). In accordance with the fundamental observation of Lighthill [20], propagation of acoustic waves in the low Mach number regime may be described by a simple linear wave equation with a source term - Lighthill’s tensor - containing the remaining quantities appearing in the complete Navier-Stokes system. The expected local decay of the acoustic energy then follows immediately from the dispersive estimates. Desjardins and Grenier [7] exploited this idea combined with the non-trivial Strichartz estimates for the acoustic equation in order to show strong (pointwise) convergence of the velocity field in the low Mach number limit for a barotropic fluid flow in the whole physical space R3 . A similar approach was adapted in [11] to the complete Navier-Stokes-Fourier system considered on “large” spatial domains, on which the Strichartz estimates were replaced by global integrability of the local energy established by Burq [4], Smith and Sogge [33]. Note that the concept of the so-called radiation boundary conditions, amply used in numerical analysis, is based on the same physical principle (see Engquist and Majda [9]). In contrast with the simple geometry of the whole space R3 , where the efficient mathematical tools based on Fourier analysis are at hand, any real problem of wave propagation inevitably includes the influence of the boundary representing a wall in the physical space. As is well-known, the Strichartz estimates become much more delicate and usually require severe geometrical restrictions to be imposed on the boundary. For instance, if the fluid domain is exterior to a compact obstacle, the latter must be starshaped or at least non-trapping (see Burq [4], Metcalfe [27], Smith and Sogge [33]), and the references cited therein). In this paper, we propose a simple method that may be used to establish strong (pointwise a.a.) convergence of acoustic waves in the low Mach number limit for fluid flows
Study of Incompressible Limit for Navier – Stokes – Fourier System
75
with ill-prepared data for a sufficiently vast class of physical domains ⊂ R3 . Pursuing the philosophy that any real physical space is always bounded but possibly “large” with respect to the speed of sound in the medium, we consider a family of bounded domains {ε }ε>0 ⊂ R3 such that ε ≈ in a certain sense as ε → 0. More specifically, we suppose that ⊂ R3 is an unbounded domain with a compact boundary ∂,
(1.1)
ε = Br (ε) ∩ ,
(1.2)
and set
where Br (ε) is a ball centered at zero with a radius r (ε), with r (ε) → ∞. Our approach is based on a nowadays classical result of Kato [13] (cf. also Burq et al. [5]) concerning weighted L 2 space-time estimates for a class of abstract operators in a Hilbert space: Theorem 1.1. [ Reed and Simon [29, Theorem XIII.25 and Corollary] ]. Let A be a closed densely defined linear operator and H a self-adjoint densely defined linear operator in a Hilbert space X . For λ ∈ / R, let R H [λ] = (H − λId)−1 denote the resolvent of H . Suppose that =
sup
λ∈ / R, v∈D (A∗ ), v X =1
A ◦ R H [λ] ◦ A∗ [v] X < ∞.
(1.3)
Then π sup w∈X, w X =1 2
∞ −∞
A exp(−it H )[w] 2X dt ≤ 2 .
The paper is organized as follows. In Sect. 2, we introduce a scaled Navier-StokesFourier system and formulate the problem of the incompressible limit for vanishing Mach number. Following the idea of Lighthill [20], we then rewrite the Navier-StokesFourier system as a wave (acoustic) equation, where all “non-hyperbolic” components are considered as a source term (see Sect. 3). In Sect. 4, we collect the necessary uniform bounds, independent of the Mach number, resulting from the total dissipation balance. This step is now well understood, and the results are taken over from [10] without proofs. Section 5 is central and contains the bulk of the analysis of acoustic waves. Assuming the boundary of the limit domain is acoustically hard, meaning the fluid velocity satisfies the complete slip or Navier-like boundary conditions, we show that Theorem 1.1 may be applied in order to deduce a uniform local decay of the acoustic energy. A remarkable feature of the present approach is that we do not need any kind of non-trapping condition to be imposed on the boundary. As a matter of fact, our technique applies whenever the spatial domain admits the limiting absorption principle for the corresponding wave operator (see Vainberg [35, Chap. VIII.2]).
76
E. Feireisl
2. Primitive System and its Incompressible Limit 2.1. Navier-Stokes-Fourier system. We consider a scaled Navier-Stokes-Fourier system in the form: ∂t + divx (u) = 0, 1 ∂t (u) + divx (u ⊗ u) + 2 ∇x p(, ϑ) = divx S(ϑ, ∇x u), ε q(ϑ, ∇x ϑ) = σε , ∂t (s(, ϑ)) + divx (s(, ϑ)u) + divx ϑ supplemented with the total energy balance 2 ε d |u|2 + e(, ϑ) (t, ·) dx = 0, dt ε 2
(2.1) (2.2) (2.3)
(2.4)
where = (t, x) is the density, u = u(t, x) the velocity field, ϑ the temperature, and p = p(, ϑ), e = e(, ϑ), s = s(, ϑ) denote the pressure, the (specific) internal energy, the (specific) entropy, respectively, obeying Gibbs’ relation 1 ϑ Ds(, ϑ) = De(, ϑ) + p(, ϑ)D . (2.5) In addition, the viscous stress tensor S is given by Newton’s rheological law 2 t S(ϑ, ∇x u) = µ(ϑ) ∇x u + ∇x u − Idivx u + η(ϑ)I divx u, 3
(2.6)
while the heat flux q(ϑ, ∇x ϑ) satisfies Fourier’s law q(ϑ, ∇x ϑ) = −κ(ϑ)∇x ϑ.
(2.7)
Finally, in virtue of the Second Law of Thermodynamics, the entropy production rate σε satisfies 1 κ(ϑ) 2 2 ε S : ∇x u + |∇x ϑ| ≥ 0. (2.8) σε ≥ ϑ ϑ Relation (2.8) reflects the fact that the weak solutions considered in the present paper may hypothetically produce “more” entropy than given by the classical formula 1 κ(ϑ) 2 2 (2.9) ε S : ∇x u + |∇x ϑ| . σε = ϑ ϑ Note, however, that (2.8) reduces to (2.9) for any weak solution of problem (2.1–2.8) as soon as this solution is smooth (see [10]). The small parameter ε can be interpreted as the Mach number. When ε → 0, the speed of sound becomes infinite, and, accordingly, the fluid can be considered as incompressible in the asymptotic limit ε → 0. Note that there are different scalings of the primitive system leading to the same mathematical problems, a typical example is provided by a long-time, small-velocity, small-viscosity setting (see Klein et al. [17]).
Study of Incompressible Limit for Navier – Stokes – Fourier System
77
2.2. Energetically insulating boundary conditions. In accordance with the total energy conservation imposed though (2.4), the system is supplemented with conservative boundary conditions u · n|∂ε = 0, [Sn] × n|∂ε = 0, q · n|∂ε = 0.
(2.10) (2.11)
The complete slip boundary condition (2.10) implies, in particular, that the boundary is acoustically hard (cf. Sect. 3 below). Note that the more conventional no-slip boundary condition u|∂ε = 0 results in a fast decay of acoustic waves in a generic class of bounded domains because of creation of a boundary layer effect (see Desjardins et al. [8]). On the other hand, our approach applies if (2.10) is replaced by a more general stipulation of Navier’s type u · n|∂ε = 0, β[u]tan + [S[ϑ, ∇x u]n]tan |∂ε = 0. 2.3. Ill-prepared initial data. The initial state of the system is determined by the following conditions: 1 1 , ϑ(0, ·) = ϑ0,ε = ϑ + εϑ0,ε , (0, ·) = 0,ε = + ε0,ε
where
, ϑ > 0,
(2.12)
ε
1 0,ε dx =
ε
1 ϑ0,ε dx = 0 for all ε > 0,
(2.13)
and 1 1 {0,ε }ε>0 , {ϑ0,ε }ε>0 are bounded in L 2 ∩ L ∞ ().
(2.14)
u(0, ·) = u0,ε ,
(2.15)
{u0,ε }ε>0 is bounded in L 2 ∩ L ∞ (; R3 ).
(2.16)
In addition
where
2.4. Incompressible limit. Let {ε , uε , ϑε }ε>0 be a family of weak solutions to the Navier-Stokes-Fourier system (2.1–2.8) supplemented with the boundary conditions (2.10), (2.11) and the initial condition (2.12), (2.15). The precise meaning of the concept of weak solution will become clear in Sect. 3 in the context of the acoustic equation. Here, we only point out that, in general, the entropy production σε may be interpreted as a non-negative measure satisfying (2.3), (2.8) in the sense of distributions (see also [10, Chap. 2]). Our main goal is to show strong (pointwise a.a.) convergence ⎧ ⎫ ⎨ ε → ⎬ a.a. in (0, T ) × , (2.17) ⎩ ⎭ ϑε → ϑ
78
E. Feireisl
and uε → U a.a. in (0, T ) ×
(2.18)
at least for suitable subsequences. In other words, the convergence imposed on the initial data through (2.12–2.16) “propagates” in time. This is not surprising for ε , ϑε , but far less obvious for the velocity uε . The piece of information contained in (2.17), (2.18) is clearly sufficient to identify the limit problem represented by the incompressible Navier-Stokes system divx U = 0,
(∂t U + divx (U ⊗ U)) + ∇x = divx µ(ϑ) ∇x U + ∇xt U , supplemented with the conventional heat equation c p (∂t + divx (U)) − divx (κ(ϑ)∇x ) = 0, where c p > 0 denotes the specific heat at constant pressure, and stands for the (relative) temperature identified as a weak limit of (ϑε − ϑ)/ε. If, in addition, a driving force is imposed, more sophisticated models as the Oberbeck-Boussinesq approximation may be obtained (see [10, Chap. 5]). As we will see in Sect. 4, the pointwise convergence of the density and the temperature claimed in (2.17) follows easily from the uniform bounds established in Sect. 4 below. On the other hand, the strong convergence of the velocity (2.18) is far more delicate and intimately related to propagation and attenuation of acoustic waves. As a matter of fact, (2.18) is not expected to hold on bounded domains with acoustically hard boundary, where large amplitude rapidly oscillating waves are generated in the limit ε → 0 (see, for instance, Lions and Masmoudi [22], or Schochet [31] ). Accordingly, for (2.18) to hold it is necessary that the target domain be unbounded, more specifically, the two closely related properties must be satisfied: • the point spectrum of the associated wave operator must be empty; • the local acoustic energy decays in time (cf. Morawetz [28], Walker [36]). 3. Lighthill’s Acoustic Equation We begin by introducing a “time lifting” ε of the measure σε through formula ε ; ϕ = σε ; I [ϕ], where we have set ε ; ϕ = σε ; I [ϕ], I [ϕ](t, x) t ϕ(z, x) dz for any ϕ ∈ L 1 (0, T ; C(ε )). =
(3.1)
0
+ It is easy to check that ε can be identified with an abstract function ε ∈ L ∞ weak (0, T ; M (ε )), where
ε (τ ), ϕ = lim σε , ψδ ϕ, δ→0+
Study of Incompressible Limit for Navier – Stokes – Fourier System
with
ψδ (t) =
⎧ 0 ⎪ ⎪ ⎪ ⎨ 1
δ ⎪ ⎪ ⎪ ⎩ 1
79
for t ∈ [0, τ ), (t − τ ), for t ∈ (τ, τ + δ), for t ≥ τ + δ,
in particular, the measure ε is well-defined for any τ ∈ [0, T ), and the mapping τ → ε is non-increasing in the sense of measures. Here the subscript in L ∞ weak stands for “weakly measurable”. Following the idea of Lighthill [20], we rewrite the Navier-Stokes-Fourier system (2.1–2.3) in the form: ε∂t Z ε + divx Vε = εdivx Fε1 , A ε∂t Vε + ω∇x Z ε = ε divx F2ε + ∇x Fε3 + 2 ∇x ε , ε ω
(3.2) (3.3)
supplemented with the homogeneous Neumann boundary conditions Vε · n|∂ε = 0, where
s(ε , ϑε ) − s(, ϑ) ε − A A + ε ε , Vε = ε uε , Zε = + ε ω ε εω
s(ε , ϑε ) − s(, ϑ) A A κ∇x ϑε 1 Fε = ε , uε + ω ε ω εϑε F2ε = Sε − ε uε ⊗ uε ,
and
ε − Fε3 = ω ε2
+ Aε
s(ε , ϑε ) − s(, ϑ) p(ε , ϑε ) − p(, ϑ) − . ε2 ε2
(3.4)
(3.5) (3.6) (3.7)
(3.8)
Here the constants A and ω are chosen to eliminate the first order term in the (formal) asymptotic expansion of (3.8) in terms of the quantities (ε − )/ε, (ϑε − ϑ)/ε , namely A
∂ p(, ϑ) ∂s(, ϑ) ∂ p(, ϑ) ∂s(, ϑ) = , ω+ A = . ∂ϑ ∂ϑ ∂ ∂
(3.9)
In order to guarantee the wave speed ω to be strictly positive, we impose the hypothesis of thermodynamic stability in the form ∂ p(, ϑ) ∂e(, ϑ) > 0, > 0 for all , ϑ > 0 ∂ ∂ϑ
(3.10)
(see Bechtel et al. [3]). Indeed (3.10), together with Gibbs’ relation (2.5), yield ω > 0, in particular, Eqs. (3.2), (3.3) form a hyperbolic system provided the right-hand is considered as given. Relation (3.10) plays a crucial role in the uniform estimates presented in Sect. 4 below.
80
E. Feireisl
System (3.2), (3.3) can be viewed as a variant of Lighthill’s acoustic analogy supplemented with acoustically hard boundary condition (3.4) (cf. Lighthill [19]). We assume that Eqs. (3.2), (3.3) as well as the boundary condition (3.4) are satisfied in a weak sense, more precisely, the integral identity
T
ε
0
[ε Z ε ∂t ϕ + Vε · ∇x ϕ] dx dt =
T 0
ε
εFε1 · ∇x ϕ dx dt
(3.11)
holds for any test function ϕ ∈ Cc∞ ((0, T ) × ε ), and T [εVε · ∂t ϕ + ωZ ε divx ϕ] dx dt 0
=
ε T
A ε ; divx ϕ εF2ε : ∇x ϕ + εFε3 divx ϕ dx dt + εω ε
0
(3.12)
is satisfied for any ϕ ∈ Cc∞ ((0, T ) × ε ; R3 ), ϕ · n|∂ε = 0,
where we have identified ε (τ )ψ dx =< ε (τ ); ψ >, ψ ∈ C(ε ). ε
4. Total Dissipation Balance, Uniform Bounds 4.1. Total dissipation balance. Besides the acoustic equation specified in (3.11), (3.12), any weak solution {ε , uε , ϑε } of the Navier-Stokes-Fourier system (2.1–2.4) satisfies the total dissipation balance
1 1 ε |uε |2 + 2 Hϑ (ε , ϑε )−∂ Hϑ (, ϑ)(ε −)− Hϑ (, ϑ) (τ, ·) dx ε ε 2 ϑ + 2 σε [0, τ ] × ε ε 1 1 0,ε |u0,ε |2 + 2 Hϑ (0,ε , ϑ0,ε )−∂ Hϑ (, ϑ)(0,ε −)− Hϑ (, ϑ) dx = ε ε 2 (4.1)
for a.a. τ ∈ [0, T ],
Study of Incompressible Limit for Navier – Stokes – Fourier System
81
where we have introduced the Helmholtz function Hϑ (, ϑ) = e(, ϑ) − ϑs(, ϑ) (see [10, Chap. 5.2.2]). As a direct consequence of Gibbs’ relation (2.5) and hypothesis of thermodynamic stability (3.10), the function Hϑ enjoys the following remarkable properties: • the function → Hϑ (, ϑ) is strictly convex; • the function ϑ → Hϑ (, ϑ) is decreasing for ϑ < ϑ and increasing for ϑ > ϑ for any fixed > 0. Moreover, we have ⎫
⎧ Hϑ (, ϑ) − ∂ Hϑ (, ϑ)( − ) − Hϑ (, ϑ) ≥ c | − |2 + |ϑ − ϑ|2 ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎬ whenever ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎩ ⎭ /2 < < 2, ϑ/2 < ϑ < 2ϑ,
(4.2)
and
Hϑ (, ϑ) − ∂ Hϑ (, ϑ)( − ) − Hϑ (, ϑ) ≥ c e(, ϑ) + ϑ|s(, ϑ)| otherwise (4.3) (see [10, Chap. 5, Lemma 5.1]).
4.2. Uniform estimates. Total dissipation balance (4.1), together with the structural properties of the Helmholtz function Hϑ specified in (4.2), (4.3), can be used to deduce uniform bounds on the family of the weak solutions {ε , uε , ϑε }ε>0 . Indeed hypotheses (2.12–2.16) imply that the integral on the right-hand side of (4.1) is bounded, therefore the quantities on the left-hand side are bounded uniformly with respect to ε → 0. Furthermore, in order to exploit (4.3), certain technical assumptions must be imposed on the structural properties of the thermodynamic functions p, e, and s. Motivated by the existence theory developed in [10, Chap. 3], we assume that the pressure p is given through formula a p(, ϑ) = ϑ 5/2 P + ϑ 4 , a > 0, (4.4) ϑ 3/2 3 where P ∈ C 1 [0, ∞), P(0) = 0, P (Z ) > 0 for all Z ≥ 0,
(4.5)
and 0<
5 3 P(Z ) −
P (Z )Z
Z lim
Z →∞
< c for all Z > 0,
P(Z ) = p∞ > 0. Z 5/3
(4.6) (4.7)
82
E. Feireisl
Moreover, in accordance with Gibbs’ relation (2.5), we suppose that e(, ϑ) =
ϑ4 3 5/2 ϑ P +a , 3/2 2 ϑ
(4.8)
and, finally, the transport coefficients satisfy 0 < c1 (1 + ϑ) ≤ µ(ϑ) ≤ c2 (1 + ϑ), 0 ≤ η(ϑ) ≤ c2 (1 + ϑ) 0 < c1 (1 + ϑ 3 ) ≤ κ(ϑ) ≤ c2 (1 + ϑ 3 )
(4.9) (4.10)
for all ϑ > 0 (see [10, Chaps. 1-3] for the physical background and further discussion concerning the structural hypotheses introduced above). Note that (4.6) is a direct consequence of the hypothesis of thermodynamic stability stated in (3.10). In view of (4.2), (4.3), it is convenient to introduce the essential and residual parts of a function h as h = [h]ess + [h]res , [h]ess = (ε , ϑε )h, [h]res = (1 − (ε , ϑε )) h, where ∈ Cc∞ (0, ∞)2 , 0 ≤ ≤ 1, ≡ 1 in an open neighborhood of the point [, ϑ]. The total dissipation balance established in (4.1), together with the structural properties of the Helmholtz function stated in (4.2), (4.3), and the restrictions imposed through hypotheses (4.4–4.10), give rise to uniform estimates on the quantities appearing in the acoustic equation (3.11), (3.12) independent of ε. Very roughly indeed, we may say that the essential components are bounded in the Lebesgue space L 2 , while the residual parts vanish in the asymptotic limit in the L 1 −norm. We state the resulting list of estimates referring to [10, Chap. 8.2] for the detailed proofs. To begin, we write Z ε = Z ε1 + Z ε2 , with ε − = + ε res ε − 2 Zε = + ε ess
Z ε1
s(ε , ϑε ) − s(, ϑ) A A ε , + ε ω ε εω res s(ε , ϑε ) − s(, ϑ) A ε ω ε ess
(cf. (3.5), where ess sup Z ε1 M(ε ) ≤ εc, ess sup Z ε2 L 2 (ε ) ≤ c. t∈(0,T )
t∈(0,T )
Similarly, Vε = Vε1 + Vε2 ,
(4.11)
Study of Incompressible Limit for Navier – Stokes – Fourier System
83
where ⎫ ⎧ 1 2 ⎨ ess supt∈(0,T ) Vε L 1 (ε ;R3 ) ≤ εc, ess supt∈(0,T ) Vε L 2 (ε ;R3 ) ≤ c, ⎬ ⎩
⎭
ess supt∈(0,T ) Vε1 L 5/4 (ε ;R3 ) ≤ c.
(4.12)
The constants in (4.11), (4.12) are independent of ε. The driving forces in (3.11), (3.12) admit similar bounds, namely Fε1 = Fε1,1 + Fε1,2 , where
T 0
Fε1,1 2L 1 ( ;R3 ) + Fε1,2 2L 2 ( ;R3 ) dt ≤ c, ε ε F2ε
=
F2,1 ε
(4.13)
+ F2,2 ε ,
where
T 0
2 1,2 2 F1,1 ε L 1 (ε ;R3×3 ) + Fε L 2 (ε ;R3×3 )
dt ≤ c
(4.14)
and, finally, Fε3 +
A ε2 ω
ε = Fε3,1
with ess sup Fε3,1 M(ε ) ≤ c. t∈(0,T )
(4.15)
5. Analysis of Acoustic Waves Having collected all the necessary preliminaries, we are in a position to formulate rigorously the main result of this paper. To simplify the forthcoming analysis, we assume that ⊂ R3 is an unbounded domain with compact boundary. Generalization to a larger class of spatial domains satisfying the so-called limiting absorption principle is straightforward. Theorem 5.1. Let ⊂ R3 be an unbounded domain with a compact boundary of class C 2+ν , ν > 0. Assume that the thermodynamic functions p, e, s satisfy Gibbs’ equation (2.5), together with the structural restrictions (4.4–4.8). In addition, let the transport coefficients µ, η, and κ obey (4.9), (4.10). Let {ε , uε , ϑε }ε>0 be a family of (weak) solutions to the acoustic equation (3.11), (3.12) in (0, T ) × ε satisfying the total dissipation balance (4.1), where • the initial data 0,ε , ϑ0,ε , u0,ε obey (2.12–2.16); • ε = ∩ Br (ε) , where lim εr (ε) = ∞.
ε→0
(5.1)
84
E. Feireisl
Then, passing to a subsequence as the case may be, we have ε → , ϑε → ϑ in L 1 ((0, T ) × K ), and uε → U in L 1 ((0, T ) × K ; R3 )
(5.2)
for any compact K ⊂ . Remark 5.1. The hypotheses of Theorem 5.1 are obviously satisfied provided the trio {ε , uε , ϑε }ε>0 represents a weak solution of Navier-Stokes-Fourier system (2.1–2.11) in (0, T ) × ε in the sense specified in [10, Chap. 3]. The existence of such a solution for a large class of initial data including those specified in (2.12–2.16) was established in [10, Chap. 3, Theorem 3.1]. Remark 5.2. The balls Br (ε) in the definition of ε may be replaced by general bounded domains B˜ ε , namely ε = ∩ B˜ ε , with Br (ε) ⊂ B˜ ε . Hypothesis (5.1) means the distance to ∂ Br (ε) dominates the speed of sound proportional to 1/ε. In particular, the acoustic waves cannot reach the outer boundary ∂ Br (ε) and return to a fixed compact set K ⊂ within the time interval (0, T ). Remark 5.3. The convergence result stated in (5.2) is not optimal with respect to the space variable, where the velocity field enjoys higher regularity, however, the main issue in the proof of Theorem 5.1 is to eliminate fast oscillations of acoustic waves in time. The remaining part of this section is devoted to the proof of Theorem 5.1. Since ε − ϑε − ϑ ϑε − ϑ ε − ε − ϑε − ϑ = = + , + , ε ε ε ε ε ε ess res ess
res
where, by virtue of (4.1), (4.7), and (4.8), ϑ −ϑ ε − ε esssupt∈(0,T ) ≤ c, esssupt∈(0,T ) ε ε ess L 2 (ε )
≤ c,
ess L 2 (ε )
and ε − esssupt∈(0,T ) ε
res L 1 (ε )
ϑ −ϑ ε ≤ εc, esssupt∈(0,T ) ε
≤ εc,
res L 1 (ε )
the proof of Theorem 5.1 reduces to showing the strong convergence of the velocity field stated in (5.2). Moreover, we claim that for (5.2) to hold it is enough to show t → uε (t, ·) · w dx → t → U(t, ·) · w dx in L 1 (0, T ) (5.3)
Study of Incompressible Limit for Navier – Stokes – Fourier System
85
for any fixed w ∈ Cc∞ (K ; R3 ), K ⊂ a given ball. Indeed, by virtue of the dissipation balance (4.1) and Korn’s inequality, we get 0
T
uε 2W 1,2 ( ;R3 ) dt ≤ c ε
(see [10, Chap. 8.2.2] for details), in particular, extending uε outside ε we may infer that uε → U weakly in L 2 (0, T ; W 1,2 (; R3 )). As W 1,2 (; R3 ) is compactly imbedded into L 2 (K ) for any bounded K , it is easy to see that (5.3) yields (5.2). Finally, since [uε ]res → 0 in, say, L 1 ((0, T ) × K ), it is enough to show (5.3) with uε replaced by [uε ]ess , which is equivalent to t → Vε (t, ·) · w dx → t → V(t, ·) · w dx in L 1 (0, T )
(5.4)
for any fixed w ∈ Cc∞ (K ; R3 ), where Vε = ε uε appears in the acoustic equation (3.11), (3.12), and V = U. 5.1. Regularization. To begin, it is useful to observe that we may assume, without loss of generality, that all quantities appearing in system (3.11), (3.12) are smooth. To this end, we deduce from (3.11), (3.12), using the uniform bounds established in (4.11), (4.12), that Z ε ∈ Cweak−(∗) ([0, T ]; M(ε )), Vε ∈ Cweak ([0, T ]; L 5/4 (ε ; R3 )), in particular, the initial values Z ε (0, ·) = Z 0,ε ∈ M(ε ), V(0, ·) = V0,ε = 0,ε u0,ε ∈ L 2 (ε ; R3 ) are well defined. Moreover, in accordance with (4.11), (4.12), 1 2 Z 0,ε = Z 0,ε + Z 0,ε ,
where 1 2 Z 0,ε M() + Z 0,ε L 2 (ε ) + V0,ε L 2 (ε ;R3 ) + ≤ c.
(5.5)
i }δ>0 ⊂ Cc∞ (ε ), For a fixed ε > 0, there exist families of smooth functions {Z 0,ε,δ 3 ∞ i = 1, 2, {V0,ε,δ }δ>0 ⊂ Cc (ε ; R ), such that 1 2 {Z 0,ε,δ }ε,δ>0 is bounded in L 1 (), {Z 0,ε,δ }ε,δ>0 is bounded in L 2 (),
(5.6)
{V0,ε,δ }ε,δ>0 is bounded in L (; R ),
(5.7)
2
3
86
E. Feireisl
and, in addition, 1 1 2 2 Z 0,ε,δ ϕ dx → Z 0,ε ; ϕ, Z 0,ε,δ ϕ dx → Z 0,ε ϕ dx for any ϕ ∈ Cc∞ (ε ), ε V0,ε,δ · ϕ dx → V0,ε · ϕ dx for any ϕ ∈ Cc∞ (ε ; R3 ),
ε
as δ → 0. Similarly, we can find ⎧ 1 ⎫ 1,1 1,2 1,i Fε,δ = Fε,δ + Fε,δ , Fε,δ ∈ Cc∞ ((0, T ) × ε ; R3 ), i = 1, 2, ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎬ 2,1 2,2 2,i 2 3×3 ∞ Fε,δ = Fε,δ + Fε,δ , Fε,δ ∈ Cc ((0, T ) × ε ; R ), i = 1, 2, ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎩ ⎭ 3,1 Fε,δ ∈ Cc∞ ((0, T ) × ε ) such that 1 2 Fε,δ → Fε1 in L 2 (0, T ; L 1 (ε ; R3 )), Fε,δ → Fε2 in L 2 (0, T ; L 2 (ε ; R3 )),
F1ε,δ → F1ε
in L (0, T ; L (ε ; R 2
1
3×3
)),
F2ε,δ → F2ε
in L (0, T ; L (ε ; R 2
2
3×3
(5.8) )), (5.9)
and 3,1 sup Fε,δ L 1 (ε ) ≤ c,
t∈[0,T ]
T 0
ε
3,1 Fε,δ ϕ dx dt →
0
T
< Fε3,1 ; ϕ > dt
(5.10)
for any ϕ ∈ Cc∞ ([0, T ] × ε ), as δ → 0. Now, consider the (unique) solution Z ε,δ , Vε,δ of the initial-boundary value problem 1 ε∂t Z ε,δ + divx Vε,δ = εdivx Fε,δ in(0, T ) × ε ,
(5.11)
ε∂t Vε,δ + ω∇x Z ε,δ = in (0, T ) × ε , Vε,δ · n|∂ε = 0, Z ε,δ (0, ·) = Z 0,ε,δ , Vε,δ (0, ·) = V0,ε,δ .
(5.12) (5.13) (5.14)
εdivx F2ε,δ
3 + ε∇x Fε,δ
Keeping ε > 0 fixed and letting δ → 0, we easily check that
Vε − Vε,δ (t, ·) · w dx → 0 as δ → 0 ess sup t∈(0,T )
for any w ∈ Cc∞ (K ; R3 ) as in (5.4). Accordingly, it is enough to show (5.4) with Vε replaced by Vε,δ(ε) for δ(ε) small enough. In what follows, we drop the subscript δ and replace the weak formulation of the acoustic equation (3.11), (3.12) by its classical counterpart (5.11), (5.12), supplemented by (5.13), (5.14). The data appearing in (5.11–5.14) are smooth and satisfy the bounds established in (4.13–4.15) uniformly for ε → 0.
Study of Incompressible Limit for Navier – Stokes – Fourier System
87
5.2. Finite speed of propagation - extension to the set . System (5.11), (5.12) admits a √ finite speed of propagation of order ω/ε. This can be easily seen multiplying Eq. (5.11) by Z ε,δ , taking the scalar product of (5.12) with Vε,δ , and integrating the resulting expression over the set √ ω (t, x) t ∈ [0, τ ], x ∈ ε , |x| < r − t . ε Consequently, by virtue of hypothesis (5.1), we may assume, extending the data in (5.11–5.14) to be zero outside ε , that Z ε,δ , Vε,δ are smooth, compactly supported in the set [0, T ] × , and solve the acoustic equation (5.11–5.14) in (0, T ) × . Thus the resulting problem reads as follows: Show that the family
t →
Vε (t, ·) · w dx
is precompact in L 1 (0, T )
(5.15)
for any w ∈ Cc∞ (K ; R3 ), K ⊂ K ⊂ a bounded ball, provided that ε∂t Z ε + divx Vε = εdivx Fε1 in (0, T ) × , εdivx F2ε
in (0, T ) × , ε∂t Vε + ω∇x Z ε = Vε · n|∂ = 0, Z ε (0, ·) = Z 0,ε , Vε (0, ·) = V0,ε in ,
(5.16) (5.17) (5.18) (5.19)
where 1 2 Z 0,ε = Z 0,ε + Z 0,ε ,
i Z 0,ε ∈ Cc∞ (), i = 1, 2,
V0,ε ∈ Cc∞ (; R3 ), and Fε1 = Fε1,1 + Fε1,2 , Fε1,i ∈ Cc∞ ((0, T ) × ; R3 ), i = 1, 2,
2,2 2,i ∞ 3×3 ), i = 1, 2, F2ε = F2,1 ε + Fε , Fε ∈ C c ((0, T ) × ; R
with 1 2 {Z 0,ε }ε>0 bounded in L 1 (), {Z 0,ε }ε>0 bounded in L 2 (),
(5.20)
{V0,ε }ε>0 bounded in L (; R ), ⎫ ⎧ 1,1 ⎨ {Fε }ε>0 bounded in L 2 (0, T ; L 1 (; R3 )), ⎬
(5.21)
⎭ ⎩ 1,2 {Fε,0 }ε>0 bounded in L 2 (0, T ; L 2 (; R3 )), ⎫ ⎧ 2,1 ⎨ {Fε }ε>0 bounded in L 2 (0, T ; L 1 (; R3×3 )), ⎬
(5.22)
⎭
(5.23)
2
⎩
3
3×3 2 2 {F2,2 )). ε,0 }ε>0 bounded in L (0, T ; L (; R
88
E. Feireisl
5.3. Compactness of the solenoidal part. Consider ψ ∈ W 1,2 ∩W 1,∞ (; R3 ), divx ψ = 0, ψ · n|∂ = 0. Multiplying Eq. (5.17) on ψ and integrating by parts, we obtain d Vε · ψ dx = − F2ε : ∇x ψ dx, Vε (0, ·) · ψ dx = V0,ε · ψ dx, dt in particular the family t → Vε · ψ dx is precompact in C[0, T ].
In other words, introducing the Helmholtz decomposition v = H[v] + H⊥ [v], H⊥ = ∇x , where is the unique solution of the Neumann problem = divx v, ∇x · n|∂ = v · n|∂
(5.24)
such that ∇x ∈ L 2 (; R3 ), ∈ L 6 () whenever v ∈ L 2 (; R3 ), we may infer that t → Vε (t, ·) · H[ψ] dx is precompact in C[0, T ],
(5.25)
provided ψ ∈ Cc∞ (K ; R3 ) is the same as in (5.15). 5.4. Local decay of the gradient component. In light of the previous arguments, it is enough to show (5.15) for the gradient part H⊥ [Vε ], H⊥ [Vε ] = ∇x ε , where ε is uniquely determined through (5.24). Accordingly, problem (5.16–5.19) can be interpreted in terms of ε as follows: ε∂t Z ε + ε = εdivx Fε1 ,
(5.26)
ε∂t ε + ωZ ε = ∇x ε · n|∂ = 0, Z ε (0, ·) = Z 0,ε , ε (0, ·) = 0,ε = −1 N [div x V0,ε ].
(5.27) (5.28) (5.29)
2 ε−1 N [div x div x Fε ],
The symbol N denotes the Laplace operator on the domain endowed with the homogeneous Neumann boundary condition. More precisely, − N is a non-negative self-adjoint operator on the space L 2 () with a domain of definition D(− N ) = v ∈ L 2 () ∇x v ∈ W 1,2 (; R3 ), v ∈ L 2 (), 1,2 (− N )[v]w dx = ∇x v · ∇x w dx for any w ∈ W () = v ∈ W 2,2 () ∇x v · n|∂ = 0 .
Study of Incompressible Limit for Navier – Stokes – Fourier System
89
We denote by {Pλ }λ≥0 the spectral resolution associated to − N - a system of orthogonal projections in L 2 () such that ∞ (− N )[v]w dx = λ d Pλ [v]w dx for any v ∈ D(− N ), w ∈ L 2 ().
0
Using {Pλ }λ≥0 we can define G(− N ) for any (possibly complex valued) Borel function G through formula ∞ G(− N )[v]w dx = G(λ) d Pλ [v]w dx .
0
5.5. Homogeneous equation. To simplify notation, we will assume hereafter that ω = 1. Our goal is to express solutions of problem (5.26–5.29) by means of Duhamel’s formula. To this end, we examine first the associated homogeneous equation ∂t Z + = 0, ∂t + Z = 0 in (0, ∞) × ,
(5.30)
supplemented with the Neumann boundary condition ∇x · n|∂ = 0,
(5.31)
Z (0, ·) = Z 0 , (0, ·) = 0 in .
(5.32)
and the initial conditions √ The (unique) solution of (5.30–5.32) can be written in terms of − N as 1 i 0 + √ [Z 0 ] (t, ·) = exp it − N 2 2 − N 1 i 0 − √ [Z 0 ] , + exp −it − N 2 2 − N √ 1 d i − N Z0 − [0 ] Z (t, ·) = − (t, ·) = exp it − N dt 2 2 √ 1 i − N Z0 + [0 ] . + exp −it − N 2 2
(5.33)
(5.34)
In particular, problem (5.30–5.32) generates a group in the associated energy space (Z , ) ∈ L 2 () × H 1,2 (), where H 1,2 () denotes the homogeneous Sobolev space, H 1,2 () = {v | v ∈ L 6 (), ∇x v ∈ L 2 (; R3 )}. At this stage, we apply Theorem 1.1 taking • X = L√2 (), • H = − N , • A = ϕG(− N ), ϕ ∈ Cc∞ (), G ∈ Cc∞ (0, ∞), for a suitable non-negative function G.
90
E. Feireisl
Since A ◦ R H [λ] ◦ A∗ = ϕG(− N ) √
1 G(− N )ϕ, − N − λ
it is enough to verify hypothesis (1.3) of Theorem 1.1 for the values of the parameter λ belonging to a bounded set Q of the complex plane, namely λ ∈ Q = {z ∈ C | Re[z] ∈ [a, b], 0 < |Im[z]| < d}, where 0 < a < b < ∞, supp[G] ⊂ (a 2 , b2 ). and d > 0. Indeed if λ ∈ / Q, then 1 G(− N ) √ G(− N ) − N − λ is a bounded linear operator, with a norm bounded in terms of the parameters a, b, d. Thus we can rewrite A ◦ R H [λ] ◦ A∗ = ϕ with
M(− N , λ) ϕ, (− N ) − λ2
M(− N , λ) = G(− N )( − N + λ)G(− N )
– a bounded linear operator in L 2 () for λ ∈ Q. At this stage, we recall that the operator − N satisfies the limiting absorption principle, namely 1 sup V ◦ (− ) − µ ◦ V 2 2 ≤ c(α, β, ϕ) < ∞ (5.35) N µ∈C;α
V(x) = (1 + |x|2 )− 2 , s > 1. It is known that the limiting absorption principle holds for N provided is an exterior domain (see Vainberg [34, Chap. VIII.2], Leis [18, Chap. 4.6, Theorem 4.36]). Several extensions to other classes of unbounded domains and more general elliptic operators are available (see Dermenjian and Guillot [6], Shimizu [32] ). Consequently, in order to verify hypotheses of Theorem 1.1, it is enough to show it is enough to show that −1 ≤c (5.36) V ◦ H ( − N ) ◦ ϕ 2 2 L[L ();L ()]
for any H ∈
Cc∞ (0, ∞).
Following Isozaki [12] we write ∞ exp i − N t H˜ (t) dt, H ( − N ) = −∞
(5.37)
Study of Incompressible Limit for Navier – Stokes – Fourier System
91
where H˜ is the Fourier transform of H . On the other hand, 2 2 2 s/2 = (1 + |x|2 )s exp i − N t [ϕg] dx. (1 + |x| ) exp i − N t [ϕg] 2 L ()
√ However, since w = exp i − N t [ϕg] solves the wave equation ∂t2 w − N w = 0 that admits a finite speed of propagation of order 1, we have 2 (1 + |x|2 )s exp i − N t [ϕg] dx 2 = (1 + |x|2 )s exp i − N t [ϕg] dx, |x|≤t+r
where r is the radius of support of ϕ. Thus we may infer that 2 2 (1+|x|2 )s exp i − N t [ϕg] dx ≤ c(1+t 2s ) exp i N t [ϕg] 2
L ()
|x|≤r +t
,
which, together with (5.37), yields (5.36). Therefore hypothesis (1.3) is satisfied as a direct consequence of the limiting absorption principle stated in (5.35). Applying Theorem 1.1 we may infer that ∞ 2 dt ϕG(− N )H (− N ) exp i − N t [g] 2 L ()
−∞
≤
c H (− N )[g] 2L 2 ()
for any β ∈ R, ϕ ∈ Cc∞ (), and H a Borel function defined in [0, ∞). Consequently, ∞ 2 dt ϕG(− N ) exp i − N t [g] 2 L ()
−∞
≤
cG,H H (− N )[g] 2L 2 ()
(5.38)
for any H ∈ C ∞ (0, ∞) bounded below away from zero on any compact subset of (0, ∞). 5.6. Application of Duhamel’s formula. In accordance with (5.33), the −component of the (unique) solution of the acoustic equation may be written by means of Duhamel’s formula in the form 1 t i ε (t, ·) = exp i 0,ε + √ − N [Z 0,ε ] ε 2 2 − N 1 t i 0,ε − √ + exp −i − N [Z 0,ε ] ε 2 2 − N t 1 1 t −s i 2 1 exp i − N divx divx Fε + √ [divx Fε ] ds + ε 2 N 2 − N 0 t 1 1 t −s i 2 1 + − N exp −i divx divx Fε − √ [divx Fε ] ds. ε 2 N 2 − N 0 (5.39)
92
E. Feireisl
5.6.1. Uniform bounds. To begin, we estimate the expressions in the square brackets in (5.39) in terms of the uniform bounds (5.20–5.23). As 0,ε is given by (5.29), it follows from (5.21) that − N 0,ε L 2 () ≤ c1 V0,ε L 2 (;R3 ) ≤ c2 (5.40) uniformly for ε → 0. Similarly, in accordance with (5.20) and the standard elliptic estimates, 1 2 (− N + Id)−1 [Z 0,ε ] L 2 () ≤ c1 Z 0,ε L 1 () + Z 0,ε L 2 () ≤ c2 . (5.41) Next, we write 1 1 [divx divx F2ε (t, ·)]ϕ dx = F2ε (t, ·) : ∇x2 [ϕ] dx. N N We claim that the mapping assigning to a function ϕ the integral on the right-hand side represents a bounded linear form on the Hilbert space D(− N ) ∩ D((− N )−1/2 ), the norm of which can be estimated in terms of F2ε (L 2 ⊕L 1 )(;R3×3 ) . Indeed, denoting h=
1 [ϕ] meaning N [h] = ϕ, N
we have 2N [h] ∈ L 2 () and
− N [h] ∈ L 2 (),
in other words, by means of the standard elliptic theory, h ∈ L 6 (), ∇x2 h ∈ L 2 ∩ L ∞ (; R3×3 ). 1 As a similar bound can be obtained for √− divx Fε1 , we may infer, using the standard N Riesz representation theorem, that 1 i 2 1 [divx divx Fε (t, ·)] ± √ [divx Fε (t, ·)] ϕ dx − N N 1 1 2 Hε (t, ·) N [ϕ] ± Hε (t, ·) √ = [ϕ] dx − N 1 for any ϕ ∈ D( N ) ∩ D √ (5.42) − N for a.a. t ∈ (0, T ), with
{Hε1 }ε>0 , {Hε2 }ε>0 bounded in L 2 (0, T ; L 2 (; C)). In accordance with relations (5.40–5.42), formula (5.39) can be recast in the form
t 1 1 N [h 1ε ]+ √ ε (t, ·) = exp ±i [h 2ε ] ± i N [h 3ε ]+ √ [h 4ε ] − N ε − N − N t
t −s exp ±i − N + ε 0
1 N [Hε1 ]+ √
− N
[Hε2 ] ± i
1 N [Hε3 ]+ √
− N
[Hε4 ]
ds,
(5.43) where {h iε }ε>0 is bounded in L 2 (), and {Hεi }ε>0 is bounded in L 2 ((0, T ) × ), i = 1, . . . , 4. (5.44)
Study of Incompressible Limit for Navier – Stokes – Fourier System
5.6.2. Convergence. Taking H (ξ ) =
1 ξ
93
in (5.38) we get
2 ϕG(− N ) exp ±i t − N [ N [h i ]] dt ε ε −∞ L 2 () ∞ 2 =ε dt ≤ εcG h iε L 2 () , i = 1, 3, ϕG(− N ) exp ±it − N [ N [h iε ]] 2 ∞
L ()
−∞
and, for H (ξ ) =
√
ξ,
2 ϕG(− N ) exp ±i t − N ( − N )−1 [h i ] dt ε ε −∞ L 2 () ∞ 2 =ε ϕG(− N ) exp ±it − N ( − N )−1 [h iε ] 2 ∞
L ()
−∞
dt
≤ εcG h iε L 2 () , i = 2, 4. Similarly, 2 ϕG(− N ) exp ±i t − s − N N H i ds dt ε ε 0 0 L 2 () ∞ T 2 ϕG(− N ) exp ±it − N exp i −s − N N H i ds dt ≤εc ε ε −∞ 0 L 2 () 2 T T −s i 2 i exp i H − ds=εc ds, i = 1, 3. ≤εcG Hε 2 N G ε L () ε 2
T
T
L ()
0
0
(5.45) Finally, 0
2 1 i ϕG(− N ) exp ±i t − s − N Hε ds dt √ ε − N 0 L 2 () ∞ T −s ≤ εc ϕG(− N ) exp ±it − N exp i ε − N −∞ 0 2 1 × √ Hεi ds dt 2 − N L () 2 T T −s i 2 i exp i H ≤εcG − ds = εc ds, i = 2, 4. Hε 2 N G ε L () ε 0 0 L 2 () (5.46)
T
T
Thus we conclude that ϕG(− N )[ε ] → 0 in L 2 ((0, T ) × ) as ε → 0 for any G ∈ Cc∞ (0, ∞) and ϕ ∈ Cc∞ ().
(5.47)
94
E. Feireisl
In order to establish (5.4), we take ϕ ∈ Cc∞ () such that ϕ ≡ 1 on the set K containing the support of w and write ∇x ε · w dx = − ε divx w dx = − ϕε divx w dx =− ϕG(− N )[ε ]divx w dx+ ϕ(G(− N )−Id)[ε ]divx w dx,
where, in accordance with (5.47), the first integral on the right-hand side tends to zero for ε → 0 for any fixed G. On the other hand, we can take a family of functions G(λ) 1, in particular, (G(− N ) − Id)[h] → 0 for any fixed h ∈ L 2 (). Consequently, writing ϕ(G(− N ) − Id)[ε ]divx w dx = ϕ(G(− N ) − Id)[ε ]divx w dx ε (G(− N ) − Id)[divx w] dx,
we can deduce (5.4) from (5.43), (5.47) as soon as we observe that 1 N [divx w], √ [divx w] ∈ L 2 (). − N Thus we have proved (5.4), and therefore Theorem 5.1. References 1. Alazard, T.: Low Mach number flows and combustion. SIAM J. Math. Anal. 38(4), 1186–1213 (electronic) (2006) 2. Alazard, T.: Low Mach number limit of the full Navier-Stokes equations. Arch. Rat. Mech. Anal. 180, 1–73 (2006) 3. Bechtel, S.E., Rooney, F.J., Forest, M.G.: Connection between stability, convexity of internal energy, and the second law for compressible Newtonian fuids. J. Appl. Mech. 72, 299–300 (2005) 4. Burq, N.: Global Strichartz estimates for nontrapping geometries: about an article by H. F. Smith and C. D. Sogge: “Global Strichartz estimates for nontrapping perturbations of the Laplacian”. Comm. Part. Diff. Eqs. 28(9–10), 1675–1683 (2003) 5. Burq, N., Planchon, F., Stalker, J.G., Tahvildar-Zadeh, A.S.: Strichartz estimates for the wave and Schrödinger equations with potentials of critical decay. Indiana Univ. Math. J. 53(6), 1665–1680 (2004) 6. Dermejian, Y., Guillot, J.-C.: Théorie spectrale de la propagation des ondes acoustiques dans un milieu stratifié perturbé. J. Diff. Eqs. 62, 357–409 (1986) 7. Desjardins, B., Grenier, E.: Low Mach number limit of viscous compressible flows in the whole space. R. Soc. Lond. Proc. Ser. A Math. Phys. Eng. Sci. 455(1986), 2271–2279 (1999) 8. Desjardins, B., Grenier, E., Lions, P.-L., Masmoudi, N.: Incompressible limit for solutions of the isentropic Navier-Stokes equations with Dirichlet boundary conditions. J. Math. Pures Appl. 78, 461–471 (1999) 9. Engquist, B., Majda, A.: Radiation boundary conditions for acoustic and elastic wave calculations. Comm. Pure Appl. Math. 32(3), 314–358 (1979) 10. Feireisl, E., Novotný, A.: Singular Limits in Thermodynamics of Viscous Fluids. Basel: Birkhäuser-Verlag, 2009 11. Feireisl, E., Poul, L.: On compactness of the velicity field in the incompressible limit of the full NavierStokes-Fourier system on large domains. Math. Meth. Appl. Sci. 32, 1269–1286 (2009) 12. Isozaki, H.: Singular limits for the compressible Euler equation in an exterior domain. J. Reine Angew. Math. 381, 1–36 (1987)
Study of Incompressible Limit for Navier – Stokes – Fourier System
95
13. Kato, T.: Wave operators and similarity for some non-selfadjoint operators. Math. Ann. 162, 258–279, (1965/1966) 14. Klainerman, S., Majda, A.: Singular limits of quasilinear hyperbolic systems with large parameters and the incompressible limit of compressible fluids. Comm. Pure Appl. Math. 34, 481–524 (1981) 15. Klein, R.: Asymptotic analyses for atmospheric flows and the construction of asymptotically adaptive numerical methods. Z. Angw. Math. Mech. 80, 765–777 (2000) 16. Klein, R.: Multiple spatial scales in engineering and atmospheric low Mach number flows. ESAIM: Math. Mod. Numer. Anal. 39, 537–559 (2005) 17. Klein, R., Botta, N., Schneider, T., Munz, C.D., Roller, S., Meister, A., Hoffmann, L., Sonar, T.: Asymptotic adaptive methods for multi-scale problems in fluid mechanics. J. Engrg. Math. 39, 261–343 (2001) 18. Leis, R.: Initial-boundary Value Problems in Mathematical Physics. Stuttgart: B. G. Teubner, 1986 19. Lighthill, J.: On sound generated aerodynamically I. General theory. Proc. of the Royal Society of London A 211, 564–587 (1952) 20. Lighthill, J.: Waves in Fluids. Cambridge: Cambridge University Press, 1978 21. Lions, P.-L.: Mathematical Topics in Fluid Dynamics, Vol. 2, Compressible Models. Oxford: Oxford Science Publication, 1998 22. Lions, P.-L., Masmoudi, N.: Incompressible limit for a viscous compressible fluid. J. Math. Pures Appl. 77, 585–627 (1998) 23. Lions, P.-L., Masmoudi, N.: Une approche locale de la limite incompressible. C.R. Acad. Sci. Paris Sér. I Math. 329(5), 387–392 (1999) 24. Masmoudi, N.: Asymptotic problems and compressible and incompressible limits. In: Advances in Mathematical Fluid Mechanics, edited by Málek, J., Neˇcas, J., Rokyta, M., Berlin: Springer-Verlag, 2000, pp. 119–158 25. Masmoudi, N.: Examples of singular limits in hydrodynamics. In: Handbook of Differential Equations, III, Dafermos, C., Feireisl, E., eds., Amsterdam: Elsevier, 2006 26. Masmoudi, N.: Rigorous derivation of the anelastic approximation. J. Math. Pures Appl. 88, 230– 240 (2007) 27. Metcalfe, J.L.: Global Strichartz estimates for solutions to the wave equation exterior to a convex obstacle. Trans. Amer. Math. Soc. 356(12), 4839–4855 (electronic) (2004) 28. Morawetz, C.S.: Decay for solutions of the exterior problem for the wave equation. Comm. Pure Appl. Math. 28, 229–264 (1975) 29. Reed, M., Simon, B.: Methods of Modern Mathematical Physics. IV. Analysis of Operators. New York: Academic Press [Harcourt Brace Jovanovich Publishers], 1978 30. Schochet, S.: Fast singular limits of hyperbolic PDE’s. J. Diff. Eqs. 114, 476–512 (1994) 31. Schochet, S.: The mathematical theory of low Mach number flows. M2 AN Math. Model Numer. Anal. 39, 441–458 (2005) 32. Shimizu, S.: The limiting absorption principle. Math. Meth. Appl. Sci. 19, 187–215 (1996) 33. Smith, H.F., Sogge, C.D.: Global Strichartz estimates for nontrapping perturbations of the Laplacian. Comm. Part. Diff. Eqs. 25(11–12), 2171–2183 (2000) 34. Vaigant, V.A.: An example of the nonexistence with respect to time of the global solutions of NavierStokes equations for a compressible viscous barotropic fluid (in Russian). Dokl. Akad. Nauk 339(2), 155– 156 (1994) 35. Va˘ınberg, B.R.: Asimptoticheskie metody v uravneniyakh matematicheskoi fiziki. Moscow: Moskov. Gos. Univ., 1982 36. Walker, H.F.: Some remarks on the local energy decay of solutions of the initial-boundary value problem for the wave equation in unbounded domains. J. Diff. Eqs. 23(3), 459–471 (1977) 37. Wilcox, C.H.: Sound Propagation in Stratified Fluids. Appl. Math. Ser. 50. Berlin: Springer-Verlag, 1984 Communicated by P. Constantin
Commun. Math. Phys. 294, 97–119 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0917-y
Communications in
Mathematical Physics
On q -Deformed gl+1 -Whittaker Function I Anton Gerasimov1,2,3 , Dimitri Lebedev1 , Sergey Oblezin1 1 Institute for Theoretical and Experimental Physics,
117259, Moscow, Russia. E-mail:
[email protected];
[email protected]
2 School of Mathematics, Trinity College, Dublin 2, Ireland.
E-mail:
[email protected]
3 Hamilton Mathematics Institute, TCD, Dublin 2, Ireland
Received: 4 February 2009 / Accepted: 25 June 2009 Published online: 26 September 2009 – © Springer-Verlag 2009
Abstract: We propose a new explicit form of q-deformed Whittaker functions solving q-deformed gl+1 -Toda chains. In the limit q → 1 the constructed solutions reduce to the classical gl+1 -Whittaker functions of class one in the form proposed by Givental. An important property of the proposed expression for the q-deformed gl+1 -Whittaker function is that it can be represented as a character of C∗ × GL+1 . This provides a q-version of the Shintani-Casselman-Shalika formula for the p-adic Whittaker function. The Shintani-Casselman-Shalika formula is recovered in the limit q → 0 when the q-deformed Whittaker function is reduced to a character of a finite-dimensional representation of gl+1 expressed through the Gelfand-Zetlin basis. Introduction Whittaker functions corresponding to semisimple finite-dimensional Lie algebras arise in various parts of modern mathematics. In particular, these functions appear in representation theory as matrix elements of infinite-dimensional representations, in the theory of quantum integrable systems as a common eigenfunction of Toda chain quantum Hamiltonians, in string theory as generating functions of correlators in Type A topological string theory on flag manifolds, and in number theory in the description of local Archimedean L-factors corresponding to automorphic representations. Although much studied, Whittaker functions seem to have some deep properties that are not yet fully revealed. In this paper we study the q-deformed gl+1 -Whittaker functions. The q-deformed Whittaker functions can be identified with the common eigenfunctions of a set of commuting q-deformed Toda chain Hamiltonians. This q-deformed Toda chain (also known as the relativistic Toda chain [Ru]) was discussed in terms of representation theory of quantum groups in [Se1,Et,Se2] and an integral representation for the q-deformed gl+1 -Whittaker function was constructed in [KLS]. Recently the q-deformed Toda chain attracted a special interest due to its connection with quantum K -theory of flag manifolds
98
A. Gerasimov, D. Lebedev, S. Oblezin
[GiL]. In this paper we pursue another direction. Our principal motivation to study q-deformed Whittaker functions is that in this more general setting, some important hidden properties of classical Whittaker functions become visible. The main result of the paper is given by Theorem 2.1 where a new expression for the q-deformed gl+1 -Whittaker function (for q < 1) is introduced. As a simple corollary of Theorem 2.1, the q-deformed gl+1 -Whittaker function can be represented as a character of C∗ × GL+1 . In the limit q → 1 this leads to a similar representation of the classical gl+1 -Whittaker function. This representation is not easy to perceive looking directly at the classical Whittaker functions. The importance of this representation of a (q-deformed) gl+1 -Whittaker function becomes obvious if we notice that in the limit q → 0 the constructed q-deformed Whittaker function reduces to the p-adic Whittaker function. In this limit, the representation as a character reduces to the well-known Shintani-Casselman-Shalika representation of the p-adic G L +1 -Whittaker function as a character of a finite-dimensional representation of G L +1 [Sh,CS]. Thus, the representation of a (q-deformed) gl+1 -Whittaker function as a character can be considered as a q-version of the Shintani-Casselman-Shalika representation. For instance, the constructed q-deformed Whittaker function vanishes outside a dominant weight cone of gl+1 similarly to the Shintani-Casselman-Shalika p-adic Whittaker function. We expect that the representation of the classical Whittaker function as a character should provide important insights into the arithmetic geometry at an infinite place of Spec(Z). Let us also remark that taking into account the results of [CS] one should expect that in the case of an arbitrary semisimple Lie algebra g, q-deformed g-Whittaker function should be given by a character of C∗ × L G(C), where Lie( L G) = L g is the Langlands dual Lie algebra. It is worth mentioning that the q → 1 limit of the explicit expression of the q-deformed Whittaker function proposed in this paper reduces to the integral representations for classical Whittaker functions introduced by Givental [Gi,GKLO]. We consider this as a sign of an “arithmetic nature” of this integral representation. On the other hand, the explicit solution has an obvious relation with the Gelfand-Zetlin parametrization of finite-dimensional representations of gl+1 (and precisely reproduces to the Gelfand-Zetlin formula for characters of finite-dimensional representations in the limit q → 0). The duality of Gelfand-Zetlin and Givental representations was already noticed in [GLO]. Let us comment on our approach to the derivation of explicit expressions for q-deformed Whittaker functions. It is known [Et] that in a certain limit defining difference equations for Macdonald polynomials are transformed into the eigenfunction equations for the q-deformed Toda chain. This is a simple generalization of the Inozemtsev limit [I], which transforms the Calogero-Sutherland integrable model into the standard Toda chain. The other ingredient we use is a recursive construction of Macdonald polynomials (analogous to the recursive construction for (q-deformed) Toda chain eigenfunctions [KL1,KLS]). We combine these results to obtain a recursive expression for the q-deformed gl+1 -Whittaker functions. The explicit form of the q-deformed Whittaker function implies various interesting interpretations. These include connections with representation theory (via characters of Demazure modules), geometry of quiver varieties, quantum cohomology of flag manifolds and will be discussed elsewhere [GLO2]. Finally, note that eigenfunctions of the q-deformed Toda chain were discussed previously (e.g. [KLS,GKL1,BF and FFJMM]). The relation of these constructions with the one proposed in this paper is an interesting question which deserves further considerations.
On q-Deformed gl+1 -Whittaker Function I
99
The paper is organized as follows. In Sect. 1 we recall a system of mutually commuting Macdonald-Ruijsenaars difference operators and a recursive construction of their common eigenfunctions. In Sect. 2 we derive a recursive expression for solutions of the q-deformed gl+1 -Toda chain. In Sect. 3 various limiting cases elucidating the construction of the q-deformed gl+1 -Whittaker function are discussed. In Sect. 4, details of the proof of Theorem 2.1 are given. 1. Macdonald-Ruijsenaars Difference Operators In this section we recall some relevant facts from the theory of Macdonald polynomials (see e.g. [Mac,Kir,AOS]). Consider symmetric polynomials in variables (x1 , . . . , x+1 ) over the field Q(q, t) of rational functions in q, t. Given a partition = (0 ≤ 1 ≤ 2 ≤ · · · ≤ +1 ), denote by the same symbol the Young diagram containing + 1 rows with k boxes in the k th row, and the upper row having the maximal length +1 . Let m and π be the following two bases in the space of symmetric polynomials indexed by partitions : m =
σ ∈S+1
+1 1 2 xσ(1) xσ(2) · . . . · xσ(+1) ,
π = π1 π2 · . . . · π+1 ,
πn =
+1
xkn ,
k=1
where S+1 is the permutation group. Define a scalar product , q,t on the space of symmetric functions over Q(q, t) as follows: π , π q,t = δ, · z (q, t), where z (q, t) =
nmn m n ! ·
n≥1
N 1 − q k , 1 − t k
m n = |{k| k = n}|.
k=1
In the following we always assume that q, t ∈ R>0 , 0 < q < 1. The following remarkable theorem was proved by Macdonald [Mac]. gl
gl
Theorem 1.1 (Macdonald). There is a unique basis P +1 = P +1 (x; q, t) in the ring of symmetric polynomial function over Q(q, t) such that gl
P +1 = m +
u m ,
<
with u ∈ Q(q, t), and for = , gl gl P +1 , P +1
q,t
= 0.
100
A. Gerasimov, D. Lebedev, S. Oblezin
The elements of the basis constructed in the theorem above are called Macdonald polynomials. Macdonald polynomials are eigenfunctions of the set of mutually commuting Macdonald-Ruijsenaars difference operators [Mac,Ru], t xi − x j gl Hr +1 = t r (r −1)/2 Tm , r = 1, . . . , + 1, (1.1) xi − x j i∈Ir , j ∈I / r
Ir
m∈Ir
where the sum is taken over ordered subsets Ir = {i 1 < i 2 < · · · < ir } ⊂ {1, 2, . . . , + 1}, and operators Tm are defined by the relations Tm xn = q δm,n xn Tm .
Tm Tn = Tm Tn ,
The simplest operator of this kind is given by gl+1
H1
=
+1 t xi − x j Tj . xi − x j i=1 j, j =i
The eigenvalues of Hr are given by (see e.g. [EK]) gl+1
Hr
gl
gl
r P +1 (x; q, t) = c P +1 (x; q, t),
where gl+1
r = χr c
r = 1, . . . , + 1,
+1 +1 q i=1 Ei,i i t i=1 Ei,i (+1−i) = q i t +1−i . Ir i∈Ir
Here E i, j are the standard generators of gl+1 , Ir = (i 1 < i 2 < · · · < ir ) ⊂ {1, 2, . . . , + gl 1} and χr +1 (g) are the characters of fundamental representations Vr = r C+1 of gl+1 . In terms of the generating series H gl+1 (ξ ) =
+1
gl+1
ξ +1−r Hr
,
H0 = 1,
r =0
we have gl
H gl+1 (ξ ) P +1 (x; q, t) =
+1
gl
(ξ + t +1−i q i ) P +1 (x; q, t).
i=1
In the following we consider 0 < t < 1. Define (x|q, t) as (x|q, t) =
∞ 1 − x x −1 q n i j i = j n=0
n 1 − t xi x −1 j q
,
and introduce the following scalar product on the space of symmetric functions in ( + 1) variables x1 , . . . , x+1 : = f, gq,t
1 ( + 1)!
+1 ı d xi f (x −1 ) g(x) (x|q, t), 2π xi i=1
(1.2)
On q-Deformed gl+1 -Whittaker Function I
101
where the integration domain
is such that each xi encircles xi = ∞, the orientation is chosen in a such way that x=∞ xi−1 d xi = −2πı and the variables xi are in the region gl+1
defined by inequalities t < |xi /x j | < t −1 , i < j. Then, the difference operators Hr : are self-adjoint with respect to , q,t gl+1
f, Hr
gl+1
gq,t = Hr
f, gq,t .
For Macdonald polynomials one has an analog of the Cauchy-Littlewood formula, C+1,m+1 (x, y|q, t) =
P (x; q, t) P (y; q, t) b (q, t),
m ≤ ,
where the sum is over all Young diagrams of glm+1 and
C+1,m+1 (x, y|q, t) =
+1 m+1 ∞ 1 − t xi y j q n , 1 − xi y j q n
(1.3)
i=1 j=1 n=0
b (q, t) =
gl
gl
P , P q,t =
1 gl P +1 ,
gl
P +1 q,t
k −k+1
i=1 k=i
n=1
,
1 − t k−i q i −k+1 +1−n , 1 − t k+1−i q i −k+1 −n
where +1 = 0 is assumed in the last formula. Proposition 1.1 ([AOS]). The following relations hold: 1. gl P +1 (x; q, t)
gl
=
gl
P , P q,t gl
gl
!P , P q,t
×C+1, (x, y
−1
ı dyi 2π yi i=1
gl
|q, t)P (y; q, t) (y|q, t),
(1.4)
where the integration domain is as in (1.2) with the additional conditions |xi y −1 j |< 1, i = 1, . . . , + 1, j = 1, . . . , .
102
A. Gerasimov, D. Lebedev, S. Oblezin
2. gl
gl
P +1 , P +1 q,t
gl
P +1 (x; q, t) =
gl
gl
( + 1)!P +1 , P +1 q,t
×C+1,+1 (x, y
−1
+1 ı dyi 2π yi
i=1 gl+1 |q, t)P (y; q, t) (y|q, t),
(1.5)
where the integration domain is as in (1.2) with the additional conditions |xi y −1 j |< 1, i = 1, . . . , + 1, j = 1, . . . , + 1. 3.
⎛ gl
+1 ⎝ P+(+1) k (x; q, t) =
+1
⎞ gl
x kj ⎠ P +1 (x; q, t).
(1.6)
j=1
Here +(+1)k is a Young diagram obtained from by a substitution j → j +k and gl gl P +1 , P +1 q,t
∞ 1 − t j−i q i − j +n 1 − t j−i q i − j +n+1 = · . 1 − t j−i+1 q i − j +n 1 − t j−i−1 q i − j +1 1≤i< j≤+1 n=0
These relations provide a recursive construction of Macdonald polynomials corresponding to arbitrary Young diagrams. Remark 1.1. Relations (1.4) and (1.5) are analogous to the action of recursion and Baxter operators on Whittaker functions considered in [GLO]. The above considerations imply immediately the following proposition. gl
Proposition 1.2. Let Hk +1 (x) and C+1, (x, y|q, t) be given by (1.1) and (1.3) respectively. The following intertwining relations hold gl gl gl Hk +1 (x)C+1, (x, y|q, t) = t k−1 Hk−1 (y) + t k Hk (y) C+1, (x, y|q, t), (1.7) for k = 1, . . . , + 1. 2. q-Deformed gl+1 -Whittaker Function Quantum Hamiltonians of the q-deformed gl+1 -Toda chain can be considered as a particular degeneration of Macdonald-Ruijsenaars difference operators [Et]. This is an analog of the Inozemtsev limit [I], which produces Toda chains from Calogero-Sutherland models. In this section we give an explicit expression for the q-deformed gl+1 -Whittaker function obtained using a degeneration of recursive operators for Macdonald polynomials discussed in the previous section. We also consider certain interesting features of the obtained expressions. Details of the proof will be given in the last section. The quantum q-deformed gl+1 -Toda chain is defined by a set of + 1 mutually gl commuting functionally independent quantum Hamiltonians Hr +1 , r = 1, . . . , + 1, 1−δi , 1 1−δir −ir −1 , 1 1−δi −i , 1 gl X i 1 1 · X i 2 2 1 · . . . · X ir Ti1 · . . . · Tir , (2.1) Hr +1 (x) = Ir
On q-Deformed gl+1 -Whittaker Function I
103
where the summation goes over all ordered subsets Ir = {i 1 < i 2 < . . . < ir } of −1 , 1 < i ≤ + 1 with X 1 = 1, Ti = q xi ∂xi . {1, 2, . . . , + 1} and X i (x) := 1 − xi xi−1 The simplest operator is given by gl H1 +1 (x)
= T1 +
(1 − xi+1 xi−1 )Ti+1 .
i=1
Common eigenfunctions of these Hamiltonians are given by q-deformed gl+1 -Whittaker functions [Et] ⎛ ⎞ gl gl gl ⎝ Hr +1 (x) λ1+1 q λi ⎠ λ1+1 ,...,λ+1 (x 1 , . . . , x +1 ) = ,··· ,λ+1 (x 1 , . . . , x +1 ), Ir i∈Ir
(2.2) where λi , i = 1, . . . + 1 are real numbers. The eigenvalues in the r.h.s. of (2.2) can be identified with the characters of fundamental representations Vωr = r C+1 of G L +1 , χr (z) = z i 1 · · · z ir , r = 1, . . . , + 1, (2.3) Ir
where z = (z 1 , . . . , z +1 ) and we assume z j = q λ j . Note that the set of Eqs. (2.2) has an infinite-dimensional linear space of solutions (due to the possibility of multiplying any solution by an arbitrary function f (x1 , . . . , x+1 ) periodic with respect to the shifts xi → xi + m i , m i ∈ Z). To obtain a finite-dimensional space of solutions we restrict the variables to the lattice Z+1 as follows x j = q p+1, j + j−1 ,
p+1, j ∈ Z, j = 1, . . . , + 1.
We shall use the following notation: p +1 = ( p+1,1 , . . . , p+1,+1 ). The complete set of commuting Hamiltonians (2.1) can be restricted to the lattice Z+1 using the substitution X i ( p +1 ) = 1 − q p+1,i − p+1,i−1 +1 , X 1 ( p +1 ) = 1 and Ti f ( p +1 ) = f ( p +1 ) with p+1,k = p+1,k + δk,i . Thus the first non-trivial Hamiltonian is given by: gl
H1 +1 ( p +1 ) = T1 +
1 − q p+1,i+1 − p+1,i +1 Ti+1 . i=1
We shall be interested in a solution of the eigenvalue problem of the q-deformed Toda chain on the lattice Z+1 : ⎞ ⎛ gl gl gl q λi ⎠ λ +1 ( p +1 ), (2.4) Hr +1 ( p +1 ) λ +1 ( p +1 ) = ⎝ Ir i∈Ir
where λ = (λ1 , . . . , λ+1 ). Let P (+1) be a set of collections of integers pi, j ∈ Z , i = 1, . . . , + 1 , j = 1, . . . , i satisfying the conditions pi+1, j ≤ pi, j ≤ pi+1, j+1 with fixed p+1,i , i = 1, . . . , + 1 . Thus P (+1) is a set of Gelfand-Zetlin patterns corresponding to an irreducible finite-dimensional representation of G L +1 (C) (see e.g. [ZS]). We denote by P+1, ⊂ P (+1) the subset p+1,i ≤ p,i ≤ p+1,i+1 , i = 1, . . . , of P (+1) .
104
A. Gerasimov, D. Lebedev, S. Oblezin
Theorem 2.1. The following function is a solution of the eigenfunction problem (2.4) gl+1
λ
( p +1 ) =
+1
pk,i ∈P (+1)
k=1
q
λk
k−1
k i=1
k−1 pk,i − i=1 pk−1,i
( pk,i+1 − pk,i )q !
k=2 i=1
×
k
,
(2.5)
( pk,i − pk+1,i )q ! ( pk+1,i+1 − pk,i )q !
k=1 i=1
p+1,1 ≤ · · · ≤ p+1,+1 , gl+1
λ
( p +1 ) = 0,
otherwise.
Here we set (n)q ! = (1 − q) . . . (1 − q n ). The proof of the theorem will be given in Sect. 4. Example 2.1. Let g = gl2 and ( p2,1 , p2,2 ) ∈ Z2 . The function
gl
λ12,λ2 ( p2,1 , p2,2 ) =
p2,1 ≤ p1,1 ≤ p2,2
q λ1 p1,1 q λ2 ( p2,1 + p2,2 − p1,1 ) , ( p1,1 − p2,1 )q !( p2,2 − p1,1 )q !
gl
λ12,λ2 ( p2,1 , p2,2 ) = 0,
p2,1 ≤ p2,2 ,
p2,1 > p2,2 ,
is a common eigenfunction of commuting Hamiltonians gl
gl
H1 2 = T1 + (1 − q p2,2 − p2,1 +1 )T2 ,
H2 2 = T1 T2 .
Note that the formula (2.5) can be easily rewritten in the recursive form. Corollary 2.1. The following recursive relation holds gl λ +1 ( p +1 )
=
p,i ∈P+1,
( p ) q
λ+1
+1 i=1
p+1,i − i=1 p,i
gl
×Q+1, ( p +1 , p |q) λ ( p ),
(2.6)
where Q+1, ( p +1 , p |q) =
1 i=1
,
( p,i − p+1,i )q ! ( p+1,i+1 − p,i )q !
and ( p ) =
−1
( p,i+1 − p,i )q !,
i=1
where the notations λ = (λ1 , . . . , λ+1 ), λ = (λ1 , . . . , λ ) are used.
(2.7)
On q-Deformed gl+1 -Whittaker Function I
105
Remark 2.1. The solution (2.5) is a q-analog of Givental’s integral representation of gl+1 -Whittaker function [Gi,GKLO]: gl ψλ +1 (x1 , . . . , x+1 )
=
R
gl
i=1
gl Q gl+1 (t +1 , t |λ+1 )
gl
dt,i Q gl+1 (t +1 , t |λ+1 ) λ1,...,λ (t ),
= exp ıλ+1
+1
(2.8)
t+1,i
i=1
−
t,i
t −t t,i −t+1,i+1 +1,i ,i , e +e −
i=1
i=1
where λ = (λ1 , . . . , λ+1 ), t k = (tk1 , . . . , tkk ), xi := t+1,i , i = 1, . . . , + 1 and we gl assume that Q gl1 (t11 |λ1 ) = eıλ1 t1,1 . 0
Proposition 2.1. There exists a C∗ × G L +1 (C) module V such that the common eigenfunction (2.5) of the q-deformed Toda chain allows the following representation for p+1,1 ≤ p+1,2 ≤ . . . p+1,+1 : gl+1
λ
( p +1 ) = Tr V q L 0
+1
q λi Hi ,
(2.9)
i=1
where Hi := E i,i , i = 1, . . . , + 1 are Cartan generators of gl+1 = Lie(G L +1 ) and L 0 is a generator of Lie(C∗ ). Proof. It is useful to rewrite (2.5) in the following form: gl+1
λ
( p +1 ) = ( p +1 )−1
q λk+1 (
i
pk+1,i − j pk,i )
pk,i ∈P (+1) k=1
k pk+1, i+1 − pk+1, i × , pk, i − pk+1, i q i=1
( p +1 ) = where
( p+1, j+1 − p+1, j )q !,
j=1
(n)q ! n . = k q (n − k)q ! (k)q !
Now taking into account the identities n n−1 n−1 = + qk , k q k q k−1 q
∞
1 = q kn , n (1 − q ) k=0
106
A. Gerasimov, D. Lebedev, S. Oblezin gl
one obtains an expansion of the function λ +1 ( p +1 ) into the sum of terms q N q m 1 λ1 · · · q m +1 λ+1 with positive integral coefficients K N ,m 1 ,...,m +1 . Let G L 1 ×···×G L 1 ⊂ G L +1 be the diagonal subgroup, then let us define the following C∗ ×G L 1 ×···×G L 1 -module:
V =
VN ,m 1 ,···,m +1 ,
VN ,m 1 ,···,m +1 = C K N ,m 1 ,...,m +1 .
N , m 1 ,...,m +1
Each VN ,m 1 ,···m +1 is acted on by the factor C∗ via multiplication by q N ; the torus (G L 1 )+1 acts on each VN ,m 1 ,···m +1 by multiplication by q m 1 λ1 +...+m +1 λ+1 . Let us note that the q-Toda chain eigenfunction problem (2.2) contains the variables z i := q λi only due to the eigenvalues given by the central functions χr (z) (2.3). Since the initial conditions for the eigenfunction (2.5) can also be expressed through χr (z), therefore we can extend the structure group of the module V from the torus C∗ × (G L 1 )+1 to the whole group C∗ × G L +1 . Thus we obtain the representation (2.9) and complete the proof.
Remark 2.2. There exists a finite-dimensional C∗ × G L +1 (C) module V f such that the following representation holds for p+1,1 ≤ p+1,2 ≤ . . . p+1,+1 : gl+1 ( p ) = ( p ) gl+1 ( p ) = Tr V f q L 0 λ λ +1 +1 +1
+1
q λi Hi .
(2.10)
i=1
The module V entering (2.9) and the module V f entering (2.10) have a structure of modules under the action of (quantum) affine Lie algebras which will be discussed [GLO2]. See however Proposition 3.4 for an explicit description of V f .
3. Various Limits Besides the limit q → 1, which recovers the classical gl+1 -Whittaker function as a solution of the gl+1 -Toda chain, there are other interesting limits elucidating the meaning of the q-deformed Toda chain equations. In a limit q → 0, the q-deformed gl+1 -Whittaker functions are given by the characters of irreducible representations of gl+1 . This will allow us to identify the Whittaker functions with p-adic Whittaker functions according to Shintani-Casselman-Shalika formula [Sh,CS]. There is also another q → 1 limit which clarifies the recursive structure of q-deformed gl+1 -Whittaker functions. 3.1. A limit q → 0. In this subsection we discuss a limit q → 0 of the constructed q-deformed Whittaker function (we restrict the Whittaker function to the domain { p+1,1 ≤ . . . ≤ p+1,+1 }, where it is non-trivial). We will show that in the domain { p+1,1 ≤ . . . ≤ p+1,+1 } the system of equations for common eigenfunctions of q-deformed Toda chain Hamiltonians reduces to the Pieri formulas (a particular case of Littlewood-Richardson rules) for the decomposition of the tensor product of an arbitrary finite-dimensional representation and a fundamental representation of gl+1 .
On q-Deformed gl+1 -Whittaker Function I
107
Let us rewrite the q-deformed Whittaker function (2.5) using the variables z i = q λ+1,i ,
gl+1
( p +1 |z) =
+1
pk,i ∈P (+1)
k=1
(
k
zk
i=1
k−1 pk,i − i=1 pk−1,i )
k−1 ( pk,i+1 − pk,i )q !
×
k=2 i=1 k
,
( pk,i − pk+1,i )q ! ( pk+1,i+1 − pk,i )q !
k=1 i=1
p+1,1 ≤ · · · ≤ p+1,+1 ,
(3.1)
where z = (z 1 , . . . , z +1 ). Proposition 3.1. 1. In the limit q → 0, the eigenfunction (3.1) is given in the domain p+1,1 ≤ · · · ≤ p+1,+1 by gl
χ p +1 (z) := gl+1 ( p +1 |z)|q=0 = +1
+1
(
k
zk
i=1
k−1 pk,i − i=1 pk−1,i )
.
(3.2)
pk,i ∈P +1 k=1
gl
2. The functions χ p +1 (z) satisfy the following set of difference equations: +1
gl+1
χr gl+1
where χr
gl
(z) χ p +1 (z) = +1
Ir
gl
χ p +1+Ir (z), r = 1, . . . , + 1,
(3.3)
+1
(z) are the characters of fundamental representations Vωr = gl χr +1 (z) = z i 1 · · · z ir , r = 1, . . . , + 1,
r
C+1 :
Ir
and Ir = (i 1 < i 2 < · · · < ir ) ⊆ {1, 2, . . . , + 1}. gl 3. The functions χ p +1 (z) can be identified with characters of irreducible finite+1 dimensional representations of G L +1 corresponding to partitions p+1,1 ≤ · · · ≤ p+1,+1 . Proof. The relations (3.2) and (3.3) follow directly from the similar relations for generic q. To prove the last statement note that (3.2) can be identified with the expression for characters of irreducible finite-dimensional representations of G L +1 obtained using the Gelfand-Zetlin bases (see e.g. [ZS]). Let { pi j }, i = 1, . . . , + 1, j = 1, . . . , i be a Gelfand-Zetlin (GZ) pattern P (+1) , that is the integers pi, j should satisfy the conditions pi+1, j ≤ pi, j ≤ pi+1, j+1 . An irreducible finite-dimensional representation can be realized in a vector space with the basis v p parametrized by GZ patterns { pi j } with fixed p+1,i . The action of the Cartan generators on v p is given by s+1 −s +1,+1 z 1E 11 z 2E 22 · · · z +1 v p = z 1s1 z 2s2 −s1 · · · z +1 vp, E
sk =
k i=1
pki .
(3.4)
108
A. Gerasimov, D. Lebedev, S. Oblezin
Thus, we have for the character gl
+1 χ p+1,1 ,..., p+1,+1 (z 1 , . . . , z +1 ) =
pk,i
+1
∈P (+1)
k=1
(
k
zk
i=1
k−1 pk,i − i=1 pk−1,i )
.
(3.5)
Remark 3.1. The second identity in Proposition 3.1 is known as the Pieri formula (see e.g. [FH], Appendix A). Thus, the q-deformed Toda chain equations can be considered as q-deformations of the Pieri formulas. There is a generalization of q-Toda chain equations providing a q-version of a general Littlewood-Richardson rule. The GZ representation of the characters (3.2) has an obvious recursive structure. Namely, one should introduce variables z i = q λi , i = 1, . . . , + 1 and then take the limit q → 0 in (2.6). This leads to the following. Corollary 3.1. Characters satisfy the following recursive relation: gl
+1 χ p+1,1 ,..., p+1,+1 (z 1 , . . . , z +1 ) +1 p − p gl z +1i=1 +1,i i=1 ,i χ p,1 ,..., p, (z 1 , . . . , z ), =
(3.6)
p,i ∈P+1,
where the sum runs over p = ( p,1 , . . . , p, ) satisfying the GZ conditions p+1,i ≤ p,i ≤ p+1,i+1 . Note that these recursive relations can be derived using the classical Cauchy-Littlewood formula C+1,m+1 (x, y) =
+1 m+1
gl 1 gl = χ +1 (x) χ m+1 (y), 1 − xi y j
m ≤ ,
(3.7)
i=1 j=1
gl
gl
where the sum runs over Young diagrams of glm+1 and χ +1 (x) = χ +1 (x1 , . . . , x+1 ) are the characters of the irreducible finite-dimensional representation of G L +1 corresponding to Young diagram . gl
Proposition 3.2. The following integral relations for the characters χ +1 (x) hold: gl χ +1 (x)
gl χ +1 (x)
=
y1 =∞
···
=
y1 =∞
···
ı dyi gl C+1, (x, y −1 )χ (y) (y|0, 0)), (3.8) 2π yi
y+1 =∞ i=1
+1 ı dyi gl C+1,+1 (x, y −1 )χ +1 (y) (y|0, 0), 2π yi y+1 =∞ i=1
⎛ gl+1 ⎝ χ+(+1) k (x) =
+1 j=1
⎞ gl
x kj ⎠ χ +1 (x).
On q-Deformed gl+1 -Whittaker Function I
109
The relations above provide the character of the irreducible finite-dimensional representation of G L +1 corresponding to any Young diagram . Remark 3.2. The relations (3.8) can be obtained from similar relations for Macdonald polynomials in the limit t → 0, q → 0. These recursion relations are analogs of the Mellin-Barnes recursion relations for the classical Whittaker functions (see [KL1,GKL, GLO] for details). According to Shintani-Casselman-Shalika formula, the p-adic Whittaker function corresponding to an algebraic reductive group G is equal to the character of the Langlands dual Lie group L G 0 acting in an irreducible finite-dimensional representation [Sh,CS]. Thus according to Proposition 3.2 we can consider gl+1 -Whittaker functions at q → 0 as an incarnation of p-adic Whittaker functions (this is in complete agreement with the results of [GLO3]). Moreover, taking into account Proposition 2.1 one can consider the main result of this paper as a generalization of the Shintani-CasselmanShalika formula to the q-deformed case which includes a limiting case of classical gl+1 Whittaker functions. This interpretation of classical Whittaker functions evidently deserves further attention. 3.2. The limit q → 1. In this subsection we consider a modified limit q → 1 leading to a very simple degeneration of the q-deformed Toda chain. In this limit the q-deformed Toda chain can be easily solved. Moreover the form of the solution makes the recursive expressions (2.6) for the q-deformed Toda chain solution very natural. Let us redefine the q-deformed Toda chain Hamiltonians and their common eigenfunctions as follows (we assume p+1,1 ≤ · · · ≤ p+1,+1 below): gl+1
Jr
gl+1
= ( p +1 ) Hr
gl+1 ( p |z) = ( p ) · gl+1 ( p |z), +1 +1 +1
( p +1 )−1 ,
( p +1 ) =
( p+1, j+1 − p+1, j )q !,
j=1
where gl+1 ( p +1 |z) is given by (3.1). Explicitly, we have that 1−δi −i , 1 1−δir −i , 1 1−δi −i , 1 gl X i1 2 1 · · · · · Ti1 · · · · · Tir , (3.9) Jr +1 = X ir −1 r −1 · X ir r +1 r Ir
X i = 1 − q p+1,i+1 − p+1, i . Let us now take the limit where we assume ir +1 = + 2 and q → 1: gl+1 ( p |z) = lim gl+1 ( p |z), ψ +1 +1 q→1
gl+1
hr
gl +1
= lim Jr q→1
.
We have limq→1 (1 − q n ) = 0 and, therefore, we obtain from (2.1) that gl+1
hr
= T+2−r · · · · · T+1 ,
where Ti acts on the functions of p+1, j as follows: p +1 ), p+1,k = p+1,k + δk,i , i, k = 1, . . . , + 1. Ti f ( p +1 ) = f ( Now the eigenvalue problem is easily solved.
110
A. Gerasimov, D. Lebedev, S. Oblezin
Proposition 3.3. 1. The function gl+1 ( p |z) ψ +1
p+1, 1 p+1, i+1 − p+1, i gl gl+1 = χ+1+1 (z) χ+1−i (z) ,
r = 1, . . . , + 1, (3.10)
i=1
is an eigenfunction of the family of mutually commuting difference operators gl+1 gl+1
hr gl+1
where χr
ψ
gl+1
( p +1 |z) = χr
gl+1 ( p |z), (z) ψ +1
(z) is the character of the fundamental representation Vωr = gl+1
χr
(z) =
z i1 · · · z ir , z i = q λi ,
(3.11) r
C+1 :
i = 1, . . . , + 1,
Ir
Ir = {i 1 < i 2 < · · · < ir } ⊂ {1, 2, . . . , + 1} and gl+1
hr
= T+2−r · · · · · T+1 , r = 1, . . . , + 1.
2. In the domain p+1,1 ≤ · · · ≤ p+1,+1 , the following recursive relation holds: gl+1 ( p |z) = ψ +1
+1
p,i ∈P+1,
z +1i=1
p+1,i − i=1 p,i
p+1, i+1 − p+1, i gl ( p |z ), ·ψ × p, i − p+1, i
(3.12)
i=1
where z = (z 1 , . . . , z ). Proof. The identity (3.11) follows from the construction. Let us prove that (3.12) follows from (3.11). Using the relation gl+1
χr
gl
gl
(z) = χr (z ) + z +1 · χr −1 (z ),
r = 1, . . . , + 1,
we have
gl
χ+1+1 (z)
p+1, 1 p+1, i+1 − p+1, i gl+1 χ+1−i (z) i=1
=
p+1,1 gl z +1 χ (z )
p+1,i+1
i=1
p,i = p+1,i
gl
z +1 χ−i (z )
p,i − p+1,i p gl +1,i+1 − p+1, i · χ+1−i (z ) · p, i − p+1, i
p+1,i+1 − p,i
On q-Deformed gl+1 -Whittaker Function I
=
p,i ∈P+1,
z +1i
111
p+1,i+1 − i p,i
p+1,i+1 − p+1, i p, i − p+1, i i=1
p, 1 −1 p, i+1 − p, i gl gl · χ (z ) χ−i (z ) i=1
=
p,i ∈P+1,
z +1i
p+1,i+1 − i p,i
p+1,i+1 − p+1, i gl ( p ). ψ λ p, i − p+1, i
(3.13)
i=1
This completes the proof. Remark 3.3. The functions
gl+1 ( p |z), ψ gl+1 ( p +1 |z) = −1 ( p +1 )ψ +1 satisfy the following recursive relations: p − ψ gl+1 ( p +1 |z) = z +1i +1,i i
p,i
p,i ∈P+1,
×
−1 i=1
( p,i
( p,i+1 − p,i )! ψ gl ( p |z ). − p+1,i )! ( p+1,i+1 − p,i )!
(3.14)
This makes the formula (2.6) for the solution of q-deformed Toda chain slightly less mysterious. Proposition 3.4. The following representation holds: gl+1 ( p |z) = Tr V f ψ +1
+1
q λi Hi ,
(3.15)
i=1
where ⊗( p+1,+1 − p+1, )
V f = Vω1
⊗( p+1,2 − p+1,1 )
⊗ · · · ⊗ Vω
⊗p
⊗ Vω+1+1,1 ,
(3.16)
and Vωn = ∧n C are the fundamental representations of G L +1 . Proof. This is an obvious consequence of Proposition 3.3.
The module V f entering (2.10) is isomorphic to (3.16) as the G L +1 -module but has a more refined structure under the action of quantum affine Lie algebras and will be discussed in [GLO2]. 4. Proof of Theorem 2.1 In this section we provide a proof of Theorem 2.1. To derive an explicit expression (2.5) for q-deformed gl+1 -Whittaker function we take as a motivation the limit t → ∞ of recursive relations (1.4) for Macdonald polynomials. Note that the recursion relations in Proposition 1.1 were defined for 0 < t < 1 and thus taking the limit t → ∞ needs some care. In particular we define the analog of the pairing (1.2) from scratch. We start with some useful relations that will be used in the proof of Theorem 2.1.
112
A. Gerasimov, D. Lebedev, S. Oblezin
4.1. q-deformed Toda chain from Macdonald-Ruijsenaars system. In this subsection we demonstrate that quantum Hamiltonians of the q-deformed Toda chain arise as a limit of Macdonald-Ruijsenaars operators when t → ∞. Let us take t = q −k . Proposition 4.1. The following relations hold: gl+1
Hr
gl
= lim Hr,k+1 k→∞ 1−δi , 1 1−δir −ir −1 , 1 1−δi −i , 1 X i 1 1 · X i 2 2 1 · · · · · X ir Ti1 · · · · · Tir , = Ir
where gl
gl+1
Hr,k+1 = D(x) Hr
(xi q k i ) D(x)−1 ,
D(x) =
+1
−k(+1−i)
xi
,
i=1
and the sum is taken over all subsets Ir = {i 1 < i 2 < · · · < ir } ⊂ {1, 2, . . . , + 1}. We −1 take X i = 1 − xi xi−1 , i = 2, . . . , + 1 , with X 1 = 1 and Ti x j = q δi, j x j Ti . Proof. Make a change of variables xi : xi −→ xi t −i , i = 1, . . . , + 1. Then for any i and any Ir , containing i we have: ⎛
⎞ ⎛ t xi − x j xi − x j t i−1− j xi − xi−1 ⎝ ⎠ −→ ⎝t br,i × i− j x − xj xi − x j t xi t −1 − xi−1 j ∈I / r i j>i ⎞ xi t j+1−i − x j ⎠, × xi t j−i − x j
(4.1)
j
/ Ir | j > i}|. Making a substitution t = q −k and conjugating the where br,i = |{ j ∈ gl+1 Hamiltonians Hr by D(x) =
+1
−ki
xi
,
i=1
ki , leads to the multiplication of each term (4.1) in the sum (1.1) by i∈Ir q i := + 1 − i. Taking into account that for any i and for any subset Ir containing i one has i∈Ir i − br,i = r (r − 1)/2, we obtain in the limit k → ∞, gl+1
Hr
=
1−δi 1 , 1
X i1
1−δi 2 −i 1 , 1
· X i2
1−δir −ir −1 , 1
· · · · · X ir
Ir −1 , i = 2, . . . , + 1 and X 1 = 1 where X i = 1 − xi xi−1
Ti1 · · · · · Tir ,
On q-Deformed gl+1 -Whittaker Function I
113
4.2. Recursive kernel Q +1, (x, y|q) for the q-deformed Whittaker function. In this subsection by taking an appropriate limit of the Cauchy-Littlewood kernel for Macdonald polynomials we derive its analog for q-deformed Whittaker functions and verify the intertwining relations with q-deformed Toda chain Hamiltonians. Let t = q −k , i = + 1 − i. Given the Cauchy-Littlewood kernel C+1, (x, y|q, t) (1.3), define a new kernel by −ki −k Q +1, (x, y|q) = lim (xi y+1−i ) · Rk (q) · C+1, (x, y|q, q ) , (4.2) k→∞
i=1
where Rk (q) =
k
−q a j , (1 − q j )2
j=1
aj =
( + 1) 2
j+
−1 k 3
.
Proposition 4.2. The following explicit expression for Q +1, (x, y|q) defined by (4.2) holds: Q +1, (x, y|q) =
∞ ∞ 1 − (xi yi )−1 q n 1 − xi+1 yi q −1 q n · . 1 − qn 1 − qn i=1 n=1
(4.3)
i=1 n=1
Proof. Making the substitution xi → xi t −i , yi → yi t i in C+1, (x, y|q, t) and taking t = q −k we have C+1, (x, y|q, t) =
∞ 1 − xi yi q n−k 1 − xi+1 yi q n 1 − xi yi q n 1 − xi+1 yi q n+k
n=0 i=1
×
+1 i−2 1 − xi y j q n+(i− j−1)k 1 − x+2−i y+1− j q n+( j−i)k . 1 − xi y j q n+(i− j)k 1 − x+2−i y+1− j q n+( j+1−i)k i=3 j=1
One encounters four types of factors which can be rewritten as k ∞ k 1 − x yq n−k −j k = (1 − x yq ) = (x y) (−q − j ) 1 − (x y)−1 q j , n 1 − x yq
n=0
j=1
j=1
∞ k−1 k 1 − x yq n n −1 j 1 − x yq , = (1 − x yq ) = q 1 − x yq n+k
n=0
n=0
j=1
∞ 2k 2k 1 − x yq n−(m+1)k −j k −j −1 j , = (1 − x yq ) = (x y) (−q ) 1 − (x y) q 1 − x yq n−mk
n=0
j=k+1
∞ n=0
j=k+1
2k 1 − x yq n+mk −1 j 1 − x yq . = q 1 − x yq n+(m+1)k j=k+1
Now it is easy to take the limit k → ∞ and obtain (4.3)
114
A. Gerasimov, D. Lebedev, S. Oblezin
Let us introduce a set of slightly modified mutually commuting Hamiltonians 1−δi −i , 1 1−δir −i , 1 1−δi −i , 1 rgl (y) = H Yi1 2 1 · · · · · Yir −1 r −1 · Yir r +1 r Ti1 · · · · · Tir , (4.4) Ir
where −1 Yi (y) = 1 − yi yi+1 ,
1 ≤ i < ,
Y = 1.
We assume here Ir = (i 1 < i 2 < · · · < ir ) ⊂ {1, 2, . . . , } and we set ir +1 = + 1. Proposition 4.3. The following intertwining relations hold: gl gl (y) + H gl (y) Q +1, (x, y|q), Hk +1 (x) Q +1, (x, y|q) = H k−1 k gl+1
where k = 1, . . . , + 1. Here Q +1, (x, y|q) , Hk (4.3), (2.1) and (4.4) respectively.
(4.5)
gl (y) are defined by (x) and H k
Proof. Direct calculation similar to the one used in the proof of Proposition 4.1.
Let us introduce a function Q+1, ( p +1 , p |q) on the lattice Z+1 × Z as follows: Q+1, ( p +1 , p |q) = Q +1, (q p+1,i +i−1 , q − p,i −i+1 |q). Corollary 4.1. The following explicit expression for Q+1, ( p +1 , p |q) holds:
Q+1, ( p +1 , p |q) = i=1
( p,i − p+1,i )( p+1,i+1 − p,i )
i=1 ( p,i
− p+1,i )q ! ( p+1,i+1 − p,i )q !
,
where (n) = 1 when n ≥ 0 and (n) = 0 otherwise. Proposition 4.4. For any k = 1, . . . , + 1 the following intertwining relations hold: gl+1
Hk
( p +1 )Q+1, ( p +1 , p |q) gl (− p ) + H gl (− p ) Q+1, ( p , p |q). = H k−1 k +1
Proof. Follows from Proposition 4.3.
(4.6)
4.3. Pairing. Define the pairing: f, gq =
ı dyi (y) f (y −1 )g(y), 2π yi 0
(4.7)
i=1
where (y) =
−1 i=1
∞ ∞
n=1 (1 − q
n)
−1 n n=0 (1 − yi+1 yi q )
,
f (y −1 ) := f (y1−1 , . . . , y−1 ).
(4.8)
The integration domain 0 is such that each yi goes around yi = ∞ and is in the region defined by inequalities |yi+1 yi−1 | ≤ q, i = 1, . . . , .
On q-Deformed gl+1 -Whittaker Function I
115
gl rgl (y) are adjoint with respect to the Proposition 4.5. Hamiltonians Hr (y) and H pairing (4.7): gl gl f, gq , f, Hk gq = H k
k = 1, . . . , .
Proof. Let us adopt the following notations: gl rgl (y) = A Ir (y) TIr , H B Ir (y) TIr , Hr (y) = Ir
Ir
where TIr := Ti1 Ti2 · . . . · Tir . One should prove
dyi (y) f (y −1 ) TIr · TI−1 A Ir (y)TIr g(y) r 0 i=1 2πı yi Ir ⎛ ⎞ −1 T (y)T dyi I r I · TI−1 = (y) ⎝ TI−1 A Ir (y)TIr · r f (y −1 )⎠ g(y). r r 2πı yi (y) 0 i=1
Ir
Let us first prove the following lemma: Lemma 4.1. For any Ir = (i 1 < i 2 < · · · < ir ) ⊂ {1, 2, . . . , } the following relation holds: ∗ B Ir (y) = Ir (y) · TI−1 A (y)T , I I r r r where Ir (y) = ((y))−1 TI−1 (y) TIr , r for all i ∈ Ir . Here for a function f (y) we define f ∗ (y) := f (y −1 ). Proof. By direct calculation one derives 1−δi −i , 1 r k k−1 −1 ∗ −1 yi k −1 1−q , (TIr A Ir (y)TIr ) = yik k=1
and yik 1−δik+1 −ik , 1 1 − r yik +1 ∗Ir (y) = 1−δi −i , 1 , k k−1 k=1 1 − q −1 yi k −1 yik where we set i 0 := 0, ir +1 := + 1. In this way we obtain 1−δi 2 −i 1 , 1
(TI−1 A Ir (y)TIr )∗ · ∗Ir (y) = Yi1 r
1−δir −ir −1 , 1
· · · · · Yir −1
1−δir +1 −ir , 1
· Yir
.
116
A. Gerasimov, D. Lebedev, S. Oblezin
Using the lemma one arrives at the following identity
dyi (y) f (y −1 ) TIr · TI−1 A Ir (y)TIr g(y) r 2πı y i 0 i=1
=
Ir
Ir
∗ dyi (y) B Ir (y)Tr f (y −1 )g(y), 2πı y i 0 (Ir ) i=1
where the integration domains 0 (Ir ) differ from the original integration domain 0 by multiplying the variables yi , i ∈ Ir by q in the definition of 0 . By the lemma the poles of the integrand are inside the circles |yi+1 yi−1 | < 1, and therefore the contours 0 (Ir ) can be deformed to 0 without encountering poles of the integrand. This proves (4.9) and, therefore, the proposition.
To construct recursive formulas for q-deformed Whittaker functions one should introduce a pairing on functions defined on the lattice {yi = q p,i +i−1 ; i = 1, . . . , ; p,i ∈ Z } with appropriate decay at infinities. Let us define the following analog of (4.7): f, glat = ( p ) f (− p )g( p ), (4.9) p ∈Z
where ( p ) =
−1
( p,i+1 − p,i ) ( p,i+1 − p,i )q !.
(4.10)
i=1
Thus ( p ) provides an extension of the measure ( p ) defined in the dominant domain p ∈ P+1, (see (2.7)) to the lattice Z . Let us note that the variables
yi = q p,i +i−1 for p ∈ P+1, satisfy conditions |yi+1 yi−1 | ≤ q entering the definition of the domain of integration 0 . The following proposition can be easily proved by mimicking the proof of Proposition 4.5.
gl rgl ( p ) are adjoint with respect to the Proposition 4.6. Hamiltonians Hr ( p ) and H pairing (4.9), gl gl f, glat , f, Hk glat = H k
k = 1, . . . , .
(4.11)
4.4. Proof of Theorem 2.1. Now we are ready to prove Theorem 2.1. We use recursion gl over the rank of glk . Set λ11 ( p11 ) = q λ1 p11 and assume that gl
gl
gl
gl
Hr ( p ) · λ1,...,λ ( p ) = χr (q i λi Ei,i ) λ1,...,λ ( p ), gl χr (q i λi Ei,i ) = z i 1 z i 2 · · · · · z ir , z i = q λi .
(4.12)
()
Ir
Here E i, j are the standard generators of gl , Ir() = (i 1 < i 2 < · · · < ir ) ⊂ {1, 2, . . . , } gl and χr (g) are characters of fundamental representations Vr = r C of gl .
On q-Deformed gl+1 -Whittaker Function I
117
gl
Let us define the function λ1+1 ,...,λ+1 ( p +1 ) as follows: gl
λ1+1 ,...,λ+1 ( p +1 ) =
p
∈Z
·q λ+1 (
( p ) Q+1, ( p +1 , p ) +1
p+1,i − i=1 p,i )
i=1
gl
λ1,...,λ ( p ),
(4.13)
where Q+1, ( p +1 , p ) =
( p,i − p+1,i ) ( p+1,i+1 − p,i ) , ( p,i − p+1,i )q ! ( p+1,i+1 − p,i )q ! i=1
and ( p ) =
−1
( p,i+1 − p,i ) ( p,i+1 − p,i )q !.
i=1
One should verify the relations: gl+1
Hr
gl
gl
gl
+1 ( p +1 ) · λ1+1 (q i λi Ei,i ) λ1+1 ,...,λ+1 ( p +1 ) = χr ,...,λ+1 ( p +1 ), gl z i 1 z i 2 · · · · · z ir , z i = q λi , χr +1 (q i λi Ei,i ) =
(4.14)
(+1)
Ir (+1)
where Ir = {i 1 < i 2 < . . . < ir } ⊂ (1, 2, . . . , + 1). gl Applying Hamiltonians Hr +1 ( p +1 ) to (4.13) and using the intertwining relation given in Proposition 4.4, one obtains
gl+1
( p )Q+1, ( p +1 , p ) q λ+1 ( i p+1,i − k p,k ) +1 gl (− p ) + H rgl (− p ) Q+1, ( p , p ) q λ+1 ( i = q λ+1 H r −1 +1
Hr
p+1,i − k p,k )
Now using (4.11), one obtains gl+1
gl
( p +1 ) λ1+1 ,...,λ+1 ( p +1 ) gl = ( p ) Hr +1 ( p +1 )Q+1, ( p +1 , p |q)
Hr
p ∈Z
× q λ+1 (
p+1,i − k p,k )
gl
λ1,...,λ ( p ) gl (− p ) + H rgl (− p ) = ( p ) q λ+1 H r −1 p ∈Z
× Q+1, ( p +1 , p ) q λ+1 (
i
i
p+1,i − k p,k )
gl
λ1,...,λ ( p )
.
118
A. Gerasimov, D. Lebedev, S. Oblezin
=
p ∈Z
( p ) Q+1, ( p +1 , p ) q λ+1 (
i
p+1,i − k p,k )
gl gl gl × q λ+1 Hr −1 ( p ) + Hr ( p ) λ1,...,λ ( p ) ⎞ ⎛ ⎟ gl ⎜ = ⎝q λ+1 q λi + q λi ⎠ λ1+1 ,...,λ+1 ( p +1 ) ()
()
Ir −1 i∈Ir −1
⎛ ⎜ =⎝ (+1)
Ir
(+1)
()
⎞
Ir
()
i∈Ir
⎟ gl q λi ⎠ λ1+1 ,...,λ+1 ( p +1 ),
i∈Ir
()
(+1)
= {i 1 < i 2 < · · · < ir } ⊂ where Ir = {i 1 < i 2 < · · · < ir } ⊂ (1, 2, . . . , ) and Ir (1, 2, . . . , + 1). In the last equality above we use the following relation: gl+1
χr
gl
gl
(z) = z +1 χr −1 (z ) + χr (z ),
where z = (z 1 , z 2 , . . . , z ) for z i = q λi . This completes the proof of Theorem 2.1.
Acknowledgements. The research of AG was partly supported by SFI Research Frontier Programme and Marie Curie RTN Forces Universe from EU. The research of SO is partially supported by RF President Grant MK-134.2007.1.
References [AOS] [BF] [CS] [Ch] [CK] [Et] [EK] [I] [FH] [FFJMM] [GKL] [GKL1] [GKLO] [GLO]
Awata, H., Odake, S., Shiraishi, J.: Integral representations of the Macdonald symmetric functions. Commun. Math. Phys. 179, 647–666 (1996) Braverman, A., Finkelberg, M.: Finite-difference quantum Toda lattice via equivariant K-theory. Trans. Groups 10, 363–386 (2005) Casselman, W., Shalika, J.: The unramified principal series of p-adic groups II. The Whittaker Function. Comp. Math. 41, 207–231 (1980) Cherednik, I.V.: Quantum groups as hidden symmetries of classic representation theory. In Differential Geometric Methods in Theoretical Physics (Chester,1988), Teaneck, NJ: World Sci. Publishing, 1989, pp. 47–54 Cheung, P., Kac, V.: Quantum Calculus. Berlin-Heidelberg-New York: Springer, 2001 Etingof, P.: Whittaker functions on quantum groups and q-deformed Toda operators. Amer. Math. Soc. Transl. Ser.2, Vol. 194, Providence, RI: Amer. Math. Soc., 1999, pp. 9–25 Etingof, P.I., Kirillov, A.A. Jr..: Macdonald’s polynomials and representations of quantum groups. Math. Res. Let. 1, 279–296 (1994) Inozemtsev, V.I.: Finite Toda lattice. Commun. Math. Phys. 121, 629–638 (1989) Fulton, W., Harris, J.: Representation Theory. A First Course. Berlin-Heidelberg-New York: Springer, 1991 3 subspaces and quantum Feigin, B., Feigin, E., Jimbo, M., Miwa, T., Mukhin, E.: Principal sl Toda hamiltonians. http://arxiv.org/abs/0707.1635v2[math.QA], 2007 Gerasimov, A., Kharchev, S., Lebedev, D.: Representation Theory and Quantum Inverse Scattering Method: Open Toda Chain and Hyperbolic Sutherland Model, Int. Math. Res. Notes, No.17, 823–854 (2004) Gerasimov, A., Kharchev, S., Lebedev, D.: Representation Theory and Quantum Integrability. Progress in Math. 237, Basel: Birkhäuser, 2005, pp. 133–156 Gerasimov, A., Kharchev, S., Lebedev, D., Oblezin, S.: On a Gauss-Givental representation of quantum Toda chain wave function. Int. Math. Res. Notices, ArticleID 96489, 23 pages, 2006 Gerasimov, A., Lebedev, D., Oblezin, S.: Baxter operator and archimedean Hecke algebra. Commun. Math. Phys. 284(3), 867–896 (2008)
On q-Deformed gl+1 -Whittaker Function I [GLO1] [GLO2] [GLO3] [GK] [Gi] [GiL] [Kir] [KL1] [KLS] [Mac] [Ru] [Se1] [Se2] [Sh] [ZS]
119
Gerasimov, A., Lebedev, D., Oblezin, S.: New Integral Representations of Whittaker Functions for Classical Lie Groups. http://arxiv.org/abs/0705.2886v1[math.RT], 2007 Gerasimov, A., Lebedev, D., Oblezin, S.: On q-deformed gl+1 -Whittaker functions II. Commun. Math. Phys., to apper (this issue), 2009, http://arxiv.org/abs/0805.3754v2[math.RT], 2008 Gerasimov, A., Lebedev, D., Oblezin, S.: Baxter q-operators and their arithmetic implications. Lett. Math. Phys. 88(1–3), 3–30 (2009) Givental, A., Kim, B.: Quantum cohomology of flag manifolds and Toda lattices. Commun. Math. Phys. 168, 609–641 (1995) Givental, A.: Stationary Phase Integrals, Quantum Toda Lattices, Flag Manifolds and the Mirror Conjecture. In: Topics in Singularity Theory, Amer. Math. Soc. Transl. Ser., 2 Vol. 180, Providence, RI: Amer.Math.Soc., 1997, pp. 103–115 Givental, A., Lee, Y.-P.: Quantum K-theory on flag manifolds, finite-difference Toda lattices and quantum groups. Invent. Math. 151, 193–219 (2003) Kirillov, A. Jr.: Traces of intertwining operators and Macdonald’s polynomials. PhD Thesis, Yale University, May 1995, available at http://arxiv.org/abs/q-alg/9503012v1, 1995 Kharchev, S., Lebedev, D.: Eigenfunctions of G L(n, r ) Toda chain: the Mellin-Barnes representation. JETP Lett. 71, 235–238 (2000) Kharchev, S., Lebedev, D., Semenov-Tian-Shansky, M.: Unitary representations of u q (sl(2, r )), the modular double and the multiparticle q-deformed Toda chains. Commun. Math. Phys. 225, 573–609 (2002) Macdonald, I.G.: A New Class of Symmetric Functions. Publ. I.R.M.A. Strasbourg, 372/S-20, Actes 20e Séminaire Lotharingien de Combinatoire, 131–171 (1988) Ruijsenaars, S.: The relativistic Toda systems. Commun. Math. Phys. 133, 217–247 (1990) Sevostyanov, A.: Regular nilpotent elements and quantum groups. Commun. Math. Phys. 204, 1–16 (1999) Sevostyanov, A.: Quantum deformation of Whittaker modules and Toda lattice. Duke Math. J. 105, 211–238 (2000) Shintani, T.: On an explicit formula for class 1 Whittaker functions on G L n over p-adic fields. Proc. Japan Acad. 52, 180–182 (1976) Zhelobenko, D., Shtern, A.: Representations of Lie Groups. Moscow: Nauka, 1983
Communicated by Y. Kawahigashi
Commun. Math. Phys. 294, 121–143 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0919-9
Communications in
Mathematical Physics
On q -Deformed gl+1 -Whittaker Function II Anton Gerasimov1,2,3 , Dimitri Lebedev1 , Sergey Oblezin1 1 Institute for Theoretical and Experimental Physics, 117259 Moscow, Russia.
E-mail:
[email protected];
[email protected]
2 School of Mathematics, Trinity College Dublin, Dublin 2, Ireland.
E-mail:
[email protected]
3 Hamilton Mathematical Institute TCD, Dublin 2, Ireland
Received: 4 February 2009 / Accepted: 25 June 2009 Published online: 17 September 2009 – © Springer-Verlag 2009
Abstract: A representation of a specialization of a q-deformed class one lattice gl+1 -Whittaker function in terms of cohomology groups of line bundles on the space QMd (P ) of quasi-maps P1 → P of degree d is proposed. For = 1, this provides an interpretation of the non-specialized q-deformed gl2 -Whittaker function in terms of QMd (P1 ). In particular the (q-version of the) Mellin-Barnes representation of the gl2 -Whittaker function is realized as a semi-infinite period map. The explicit form of the period map manifests an important role of q-version of -function as a topological genus in semi-infinite geometry. A relation with the Givental-Lee universal solution (J -function) of q-deformed gl2 -Toda chain is also discussed.
Introduction In our work [GLO1] (which is the first in the series of papers [GLO1,GLO2]) we have proposed an explicit representation of a q-deformed class one lattice gl+1 -Whittaker function defined as a common eigenfunction of a complete set of commuting quantum Hamiltonians of a q-deformed gl+1 -Toda chain. Here “class one” means that the Whittaker function is non-zero only in the dominant domain. The case = 1 was discussed previously in [GLO3] (for related results in this direction see [KLS, GiL,GKL1,BF,FFJMM]). A special feature of the proposed representation is that the gl q-deformed class one gl+1 -Whittaker function z +1 ( p) with z = (z 1 , . . . , z +1 ) and p = ( p1 , . . . , p+1 ) ∈ Z+1 , is given by a character of a C∗ × G L +1 (C)-module V p . The expression in terms of a character can be considered as a q-version of ShintaniCasselman-Shalika representation of class one p-adic Whittaker functions [Sh,CS]. Indeed our representation of a q-deformed gl+1 -Whittaker function reduces, in a certain
122
A. Gerasimov, D. Lebedev, S. Oblezin
limit, to the Shintani-Casselman-Shalika representation of a p-adic Whittaker function. Note that the representation of a q-deformed Whittaker function as a character is a q-analog of the Givental integral representation [Gi2,GKLO] of the classical gl+1 Whittaker function. The main objective of this paper is to better understand the representation of the q-deformed gl+1 -Whittaker function as a character. Below we consider a specialization of the q-deformed Whittaker function given by the trace over a certain C∗ × G L +1 (C)module Vn,k (in the case = 1 there is actually no specialization). Our main result is presented in Theorem 3.1. We provide a description of the C∗ × G L +1 (C)-module Vn,k as a zero degree cohomology group of a line bundle on an algebraic version LP+ of a semi-infinite cycle LP + in a universal covering LP of the space of loops in P . We define LP+ as an appropriate limit d → ∞ of the space QMd (P ) of degree d quasi-maps of P1 to P [Gi1,CJS]. In particular for = 1 this provides a description of the q-deformed gl2 -Whittaker function in terms of cohomology of line bundles over LP1+ . A universal solution of the q-deformed gl+1 -Toda chain [GiL] was given in terms of cohomology groups of line bundles over QMd (X ), X = G/B for finite d. We demonstrate how our interpretation of the q-deformed gl+1 -Whittaker function is reconciled with the results of [GiL]. Using Theorem 3.1, we interpret a q-version of the Mellin-Barnes integral representation of the specialized q-Whittaker function as an instance of the Riemann-RochHirzebruch theorem in a semi-infinite setting. The corresponding Todd class is expressed in terms of a q-version of the -function. Analogously, the classical -function appears in a description of the fundamental class of semi-infinite homology theory and enters the Mellin-Barnes integral representation of the classical Whittaker function. Let us stress that the C∗ × G L +1 (C)-module Vn,k arising in the description of the q-deformed gl+1 -Whittaker function is not irreducible. It would be natural to look for an interpretation of Vn,k as an irreducible module of a quantum affine Lie group. A relation of the geometry of semi-infinite flags to representation theory of affine Lie algebras was ∞ proposed in [FF]. The semi-infinite flag space is defined as X 2 = G(K)/H (O)N (K), where K = C((t)), O = C[[t]], B = H N is a Borel subgroup of G, N is its unipotent radical and H is the associated Cartan subgroup. The semi-infinite flag spaces are not easy to deal with. An interesting approach to the semi-infinite geometry was proposed by Drinfeld. He introduced a space of quasi-maps QMd (P1 , G/B) that should be con∞ sidered as a finite-dimensional substitute for the semi-infinite flag space X 2 (see e.g. [FM,FFM,Bra]). Thus, taking into account the constructions proposed in this paper one can expect that (q-deformed) gl+1 -Whittaker functions (encoding the Gromov-Witten invariants and their K -theory generalizations) can be expressed in terms of representation theory of affine Lie algebras (see [GiL] for a related conjecture and [FFJMM] for recent progress in this direction). The paper [GLO2] establishes a connection of our results to the representation theory of (quantum) affine Lie groups. The paper is organized as follows. In Sect. 1, explicit solutions of the q-deformed gl+1 -Toda chain (q-versions of Whittaker functions) are recalled. In Sect. 2, we derive integral expressions for the counting of holomorphic sections of line bundles on the space of quasi-maps. In Sect. 3 we derive a representation of the specialized q-Whittaker functions in terms of cohomology of holomorphic line bundles on the space of quasi-maps of P1 to P . We propose an interpretation of the q-Whittaker functions as semi-infinite periods. In Sect. 4 the analogous interpretation of the classical Whittaker functions is discussed. In Sect. 5, we clarify the connection of our interpretation of the q-deformed gl+1 -Whittaker function with the results of [GiL].
On q-Deformed gl+1 -Whittaker Function II
123
1. q-Deformed gl+1 -Whittaker Function In this section we recall a construction [GLO1] of the q-deformed gl+1 -Whittaker funcgl tion z +1 ( p +1 ) defined on the lattice p +1 = ( p+1,1 , . . . , p+1,+1 ) ∈ Z+1 . We will consider only class one Whittaker functions, i.e. Whittaker functions satisfying the condition gl+1
z
( p +1 ) = 0
(1.1)
outside the dominant domain p+1,1 ≥ · · · ≥ p+1,+1 . The q-deformed gl+1 -Whittaker functions are common eigenfunctions of q-deformed gl+1 -Toda chain Hamiltonians: 1−δi −i , 1 1−δir −i , 1 1−δi −i , 1 gl Hr +1 ( p +1 ) = X i1 2 1 · . . . · Ti1 · . . . · Tir , X ir −1 r −1 · X ir r +1 r Ir
(1.2) where the sum is over ordered subsets Ir = {i 1 < i 2 < . . . < ir } ⊂ {1, 2, · · · , + 1} and we assume ir +1 = + 2. In (1.2) we use the following notations Ti f ( p +1 ) = f ( p +1 ), p+1,k = p+1,k + δk,i , i, k = 1, . . . , + 1, X i = 1 − q p+1,i − p+1,i+1 +1 , i = 1, . . . , , and X +1 = 1. We assume q ∈ R, 0 < q < 1. For example, the first nontrivial Hamiltonian has the following form: gl
H1 +1 ( p +1 ) =
(1 − q p+1,i − p+1,i+1 +1 )Ti + T+1 .
(1.3)
i=1
The main result of [GLO1] is a construction of common eigenfunctions of quantum Hamiltonians (1.2): gl gl+1 gl+1 Hr +1 ( p +1 )z 1 ,...,z ( p ) = ( z i ) z 1 ,...,z (1.4) +1 +1 ( p +1 ). +1 Ir i∈Ir
Denote by P (+1) ⊂ Z(+1)/2 a subset of integers pn,i , n = 1, . . . , + 1, i = 1, . . . , n satisfying the Gelfand-Zetlin conditions pk+1,i ≥ pk,i ≥ pk+1,i+1 for k = 1, . . . , . In the following we use the standard notation (n)q ! = (1 − q)...(1 − q n ). gl
+1 Theorem 1.1. Let z 1 ,...,z +1 ( p +1 ) be a function given in the dominant domain p+1,1 ≥ . . . ≥ p+1,+1 by
gl
+1 z 1 ,...,z +1 ( p +1 ) =
+1
zk
i
pk,i − i pk−1,i
pk,i ∈P (+1) k=1 k−1 ( pk,i − pk,i+1 )q !
×
k=2 i=1 k k=1 i=1
( pk+1,i − pk,i )q ! ( pk,i − pk+1,i+1 )q !
,
(1.5)
124
A. Gerasimov, D. Lebedev, S. Oblezin gl
+1 and zero otherwise. Then, z 1 ,...,z +1 ( p +1 ) is a common solution of the eigenvalue problem (1.4).
Formula (1.5) can be written also in the recursive form. Corollary 1.1. Let P+1, be a set of p = ( p,1 , . . . , p, ) satisfying the conditions p+1,i ≥ p,i ≥ p+1,i+1 . The following recursive relation holds: gl
+1 z 1 ,...,z +1 ( p +1 ) =
( p ) z +1i
p+1,i − i p,i
p ∈P+1,
gl
Q +1, ( p +1 , p |q)z 1 ,...,z ( p ),
(1.6) where 1
Q +1, ( p +1 , p |q) =
i=1 −1
( p ) =
,
( p+1,i − p,i )q ! ( p,i − p+1,i+1 )q ! (1.7)
( p,i − p,i+1 )q !.
i=1
Remark 1.1. The representation (1.6) is a q-analog of Givental’s integral representation of the classical gl+1 -Whittaker function [Gi2,JK]: gl+1
ψλ
(x1 , . . . , x+1 ) =
R i=1
gl
gl
dt,i Q gl+1 (t +1 , t |λ+1 )ψλ1,...,λ (t ),
(1.8)
gl
Q gl+1 (t +1 , t |λ+1 )
+1 t −t
t,i −t+1,i+1 +1,i ,i = exp ıλ+1 e , t+1,i − t,i − +e i=1
i=1
i=1
where λ = (λ1 , . . . , λ+1 ), t k = (tk1 , . . . , tkk ), xi := t+1,i , i = 1, . . . , + 1 and gl z i = q γi , λi = γi log q and we assume that Q gl1 (t1,1 |λ1 ) = eıλ1 t1,1 . For the represen0 tation theory derivation of this integral representation of gl+1 -Whittaker function see [GKLO]. The representation (1.5) of the q-Whittaker function turns into representation (1.8) of the classical Whittaker function in the appropriate limit. As an example consider g = gl2 . Let p1 := p2,1 ∈ Z, p2 := p2,2 ∈ Z and p := p1,1 ∈ Z. Then the function
gl
2 z 1 ,z 2 ( p1 , p2 ) =
p2 ≤ p≤ p1
gl
2 z 1 ,z 2 ( p1 , p2 ) = 0,
p p +p −p
z1 z2 1 2 , ( p1 − p)q !( p − p2 )q !
p1 < p2 ,
p1 ≥ p2 , (1.9)
On q-Deformed gl+1 -Whittaker Function II
125
is a solution of the system of equations: gl2 gl2 (1 − q p1 − p2 +1 )T1 + T2 z 1 ,z 2 ( p1 , p2 ) = (z 1 + z 2 ) z 1 ,z 2 ( p1 , p2 ), gl
gl
(1.10)
2 2 T1 T2 z 1 ,z 2 ( p1 , p2 ) = z 1 z 2 z 1 ,z 2 ( p1 , p2 ).
Let us consider the following specialization of the q-deformed gl+1 -Whittaker function: gl
gl
+1 +1 z 1 ,...,z +1 (n, k) := z 1 ,...,z +1 (n + k, k, . . . , k).
(1.11)
Let Pn,k be a Gelfand-Zetlin pattern such that ( p+1,1 , . . . , p+1,+1 ) = (n +k, k, . . . , k). Then, the relations p+1,i ≥ p,i ≥ p+1,i+1 for the elements of a Gelfand-Zetlin pattern imply pk,i=1 = 0 and we have that
+1 n+k− p p −p p −k z +1 ,1 z ,1 −1,1 z 1 1,1 gl+1 k ··· zi z 1 ,...,z +1 (n, k) = (n + k − p,1 )q ! ( p,1 − p−1,1 )q ! ( p1,1 − k)q ! i=1 Pn,k
+1 n +1 z +1 z n1 = z ik ··· 1 . (1.12) (n +1 )q ! (n 1 )q ! n +···+n =n i=1
+1
1
gl
+1 Theorem 1.2. z 1 ,...,z +1 (n, k) satisfies following difference equation: +1 gl+1 −1 n gl+1 (1 − z i T ) z 1 ,...,z +1 (n, k) = q z 1 ,...,z +1 (n, k),
(1.13)
i=1
where T · f (n) = f (n + 1). Proof. The proof is based on the explicit expression (1.5). Introduce the generating function gl+1 z 1 ,...,z +1 (t, k)
=
t
n
gl+1 z 1 ,...,z +1 (n, k)
=
n∈Z
+1
z ik , m m=0 (1 − t z i q )
∞
j=1
where we use the identity ∞
xn 1 . = m (n)q ! m=0 (1 − xq )
∞
n=0
gl
gl
+1 +1 Due to the fact that z 1 ,...,z +1 (n, k) = 0 for n < 0, the generating function z 1 ,...,z +1 (t, k) is regular at t = 0. It is easy to check now the following identity:
+1
gl
gl
+1 +1 (1 − t z i ) z 1 ,...,z +1 (t, k) = z 1 ,...,z +1 (qt, k).
j=1
Expanding the latter relation in powers of t, we obtain (1.13) for the coefficients of gl+1 z 1 ,...,z
+1 (t, k).
126
A. Gerasimov, D. Lebedev, S. Oblezin
Remark 1.2. The difference equation (1.13) for the specialized q-Whittaker function gl+1 z 1 ,...,z +1 (n, k) can be derived directly from the system of equations (1.4) for the nongl
+1 specialized q-deformed Whittaker function z 1 ,...,z +1 ( p1 , p2 , . . . , p+1 ) and the condition
gl
+1 z 1 ,...,z +1 ( p1 , p2 , . . . , p+1 ) = 0
outside the dominant domain p1 ≥ · · · ≥ p+1 . Lemma 1.1. The following integral representation for the specialized q-deformed gl+1 Whittaker functions holds:
+1 +1 dt −n gl +1 t z (n, k) = z ik q (z i t), (1.14) t=0 2πı t i=1
i=1
where q (x) =
∞ n=0
1 . 1 − qnx
Proof. Using the identity ∞ n=0
one obtains, for n ≥ 0, that gl z +1 (n, k)
=
∞ 1 xm = , 1 − xq n (m)q !
+1
m=0
z ik
i=1
=
+1
+1 ∞ dt −n 1 t 1 − z i tq m t=02πı t i=1 m=0
z ik
z 1n 1
n 1 +...+n +1 =n
i=1
(n 1 )q !
· ··· ·
n +1 z +1 . (n +1 )q !
(1.15)
gl
For n < 0, we obviously have that z +1 (n, k) = 0.
The corresponding integral representation for the classical gl2 -Whittaker function is given by the Mellin-Barnes representation for the gl2 -Whittaker function, ıσ +∞ ı ı λ − λ2 λ − λ1 gl2 (λ1 +λ2 )x2 λ(x1 −x2 ) , ψλ1 ,λ2 (x1 , x2 ) = e dλ e ı ı iσ −∞ (1.16) where σ > max{Im λ j , j = 1, . . . , + 1}. Remark 1.3. The expression
+1 gl+1 z 1 ,...,z +1 (n, k) = z ik i=1
gl
+1 z 1 ,...,z +1 (n, k) = 0,
n 1 +···+n +1 =n
n<0
z n +1 z 1n 1 · . . . · +1 , (n 1 )q ! (n +1 )q !
n ≥ 0, (1.17)
On q-Deformed gl+1 -Whittaker Function II
127
is a q-analog of the Givental integral representation for the equivariant Gromov-Witten invariants of X = P [Gi3], f λ (T ) =
k=1
dtk,1 e F (t) , 1
(1.18)
where is an appropriate integration domain, λ = (λ1 , . . . , λ+1 ), T := t+1,1 and F(t) = ıλ1 t11 +
ıλk+1 (tk+1,1 − tk,1 ) − et11 −
k=1
etk+1,1 −tk,1 .
k=1
The representation (1.17) for specialized q-Whittaker function turns into (1.18) in the appropriate limit. 2. Counting Holomorphic Sections In this section we are going to provide an interpretation of the explicit expressions (1.12) for q-deformed specialized gl+1 -Whittaker functions in terms of traces of operators acting on the spaces of holomorphic sections of line bundles on infinite-dimensional manifolds. For this aim, we first consider an auxiliary problem of counting holomorphic sections on finite-dimensional manifolds approximating the infinite-dimensional ones. The relevant finite-dimensional manifolds are spaces of the quasi-maps of P1 to G L +1 (C)-homogeneous spaces. 2.1. Space of quasi-maps. Let us start with recalling the general construction of the quasi-map compactification of the space of holomorphic maps of P1 to the partial flag spaces of the complex Lie group G L +1 due to Drinfeld. Let αi , i = 1, . . . , , be a set of simple roots of the complex Lie algebra gl+1 . To any ordered subset of simple roots {αi1 , . . . , αir } indexed by an ordered subset I P = {i 1 < · · · < ir } ⊂ {1, . . . , }, one can associate a parabolic subgroup P ⊂ G L +1 . Namely, let B ⊂ G L +1 be the subgroup of upper-triangular matrices generated by Cartan torus and one-parameter unipotent subgroups corresponding to positive simple roots. Then a parabolic subgroup P is generated by B and one-parameter unipotent subgroups corresponding to negative roots (−αi ) such that i ∈ / I P . In particular, when r = one gets P = B, and the corresponding homogeneous space G L +1 /B coincides with the full flag space. On the other hand for a parabolic subgroup P0 ⊂ G L +1 associated to the first simple root (i.e. I P0 = {1} ⊂ {1, 2, . . . , }), the corresponding homogeneous space G L +1 /P0 is isomorphic to the projective space P . Partial flag spaces G L +1 /P possess canonical projective embeddings Pn j −1 , n j = ( + 1)!/j! ( + 1 − j)!. (2.1) π : G L +1 /P → = j∈I P
The group H 2 (G L +1 /P, Z) = Zr is naturally isomorphic to a sublattice of the weight lattice of sl+1 and is spanned by the weights ωi indexed by I P . Let L j , j = 1, . . . , r , be the line bundles on G L +1 /P obtained as pull backs of O(1) from the direct factors Pn j −1 in the right-hand side (r.h.s.) of (2.1). The lattice H 2 (G L +1 /P, Z) = Zr is generated by the first Chern classes c1 (Li ).
128
A. Gerasimov, D. Lebedev, S. Oblezin
Let Md (G L +1 /P) be a non-compact space of holomorphic maps of P1 of multidegree d ∈ H 2 (G L +1 /P, Z) to the flag space G L +1 /P. Due to (2.1), Md (G L +1 /P) is a subspace of the product of space Md j (Pn j −1 ). Explicitly, each Md j (Pn j −1 ) can be described as a set of collections of n j , relatively prime polynomials of degree d j , up to a common constant factor. The space Md j (Pn j −1 ) allows for a compactification by the space of quasi-maps QMd j (Pn j −1 ) defined as a set of collections of n j , polynomials of degree d j , up to a common constant factor. The space ofquasi-maps QMd (G L +1 /P) is then constructed as a closure of Md (G L +1 /P) in j QMd j (Pn j ). Thus defined QMd (G L +1 /P) is an (in general singular) irreducible projective variety. A small resolution of this space is known due to [La,Ku]. On the space of holomorphic maps Md (G L +1 /P) of P1 to G L +1 /P, there is a natural action of the group C∗ × G L +1 (and, thus, of its maximal compact subgroup S 1 × U+1 ). Here, the action of G L +1 is induced by the standard action on flag spaces and the action of C∗ is induced by the action of C∗ on P1 given by (y1 , y2 ) → (ξ y1 , y2 ) in homogeneous coordinates (y1 , y2 ) on P1 . This action of C∗ × G L +1 can be extended to an action on the space QMd (G L +1 /P) of quasi-maps. In the following we consider a parabolic subgroup P0 ⊂ G L +1 associated to the first simple root (and thus I P0 = {1} ⊂ {1, 2, . . . , }). The corresponding homogeneous space G L +1 /P0 is a projective space P . The space of quasi-maps1 QMd (P ) is a non-singular projective variety P(+1)(d+1)−1 . A quasi-map φ ∈ QMd (P ) is given by a collection (a0 (y) : a1 (y) : . . . a (y)) , of homogeneous polynomials ai (y) in variables y = (y1 , y2 ) of degree d ak (y) =
d
j d− j
ak, j y1 y2
, k = 0, . . . , ,
j=0
considered up to the multiplication of all ai (y)’s by a nonzero complex number. The action of (ξ, g) ∈ C∗ × G L +1 on QMd (P ) is given by
ξ : (a0 (y) : a1 (y) : . . . a (y)) −→ a0 (y ξ ) : a1 (y ξ ) : . . . : a (y ξ ) ,
+1 +1 g : (a0 (y) : a1 (y) : . . . : a (y)) −→ g1,k ak−1 (y) : . . . : g+1,k ak−1 (y) , k=1
k=1
(2.2) where g = gi j and y ξ = (ξ y1 , y2 ). 2.2. Generating functions of holomorphic sections. Let O(1) be a standard line bundle on P(+1)(d+1)−1 . The space of sections of the line bundle O(n) := O(1)⊗n on QMd (P ) is naturally a C∗ × G L +1 -module. We are interested in calculating the corresponding character. Let T ∈ G L +1 be a Cartan torus, H1 , . . . , H+1 be a basis in Lie(T ), and L 0 be a generator of Lie(C∗ ). 1 The compactification of M (P ) by the space QM (P ) of quasi-maps arises naturally in the linear d d sigma-model description of Gromov-Witten invariants of projective spaces [W].
On q-Deformed gl+1 -Whittaker Function II
129
Let Lk be a one-dimensional G L +1 -module such that Hi Lk = kLk , for i = 1, . . . , + 1. Cohomology groups H ∗ (QMd (P ), O(n) ⊗ Lk ) have a natural structure of the C∗ × G L +1 (C)-module. We denote by Lk (n) = O(n) ⊗ Lk the line bundles O(n) twisted by the one-dimensional G L +1 -module Lk . (d) Let An,k (z, q), be the character of the C∗ ×G L +1 -module Vn,k,d = H 0 (QMd (P ), Lk (n)), n ≥ 0,
(d)
An,k (z, q) = Tr Vn,k,d q L 0 e λi Hi , where we assume that q ∈ R, 0 < q < 1. This character can be straightforwardly calculated as follows. The space Vn,k,d = H 0 (QMd (P ), Lk (n)) can be identified with the space of degree n homogeneous polynomials in ( + 1)(d + 1) variables ar,i , for r = 0, . . . , and i = 0, . . . , d. Define Vk,d = ⊕∞ n=0 Vn,k,d , and the grading on Vk,d is defined by the eigenvalue decomposition with respect to the action of an operator D, t D : Vn,k,d → t n Vn,k,d , t ∈ C∗ . The action of the subgroup (C∗ × T ) ⊂ G(C) = C∗ × G L +1 is given by
e λi Hi : (a0 (y) : a1 (y) : . . . a (y)) −→ eλ1 a0 (y) : eλ2 a1 (y) : . . . : eλ+1 a (y) , (2.3) where ar (y) =
d
j d− j
ar, j y1 y2
, r = 0, . . . , .
j=0
The action of the generator L 0 of C∗ is as follows: q L0 :
ar, j −→ q j ar, j .
(2.4)
Proposition 2.1. For the C∗ × G L +1 -character of the module Vn,k,d , the following integral representation holds:
+1 +1 d dt 1 (d) λi Hi L0 An,k (z, q) = Tr Vn,k,d q e , = z ik n+1 (1 − tq j z m ) t=0 2πı t m=1 j=0
i=1
(2.5) where z = (z 1 , . . . , z +1 ) and z m =
eλm .
Proof. A simple calculation gives us that (d) Ak (z, t, q)
= Tr Vk,d t D q L 0 e
λi Hi
=
+1 i=1
z ik
d +1 m=1 j=0
1 . (2.6) (1 − tq j z m )
The projection on the subspace of Vk,d of the grading n with respect to D can be realized by taking a residue, dt λi Hi λi Hi L0 D L0 Tr Vn,k,d q e = Tr t q e , (2.7) V n+1 k,d t=0 2πı t where t=0 denotes the contour integral around a small circle around t = 0. This gives us the integral expression (2.5).
130
A. Gerasimov, D. Lebedev, S. Oblezin
2.3. Equivariant Euler characteristic of line bundles on QMd (P ). Characters (2.5) of the space of holomorphic sections can be related to equivariant holomorphic Euler characteristics of line bundles on QMd (P ). First we recall the standard facts about line bundles on projective spaces. Line bundles O(n) on projective spaces P N are equivariant with respect to the standard action of U N +1 on P N . The U N +1 -equivariant Euler characteristic of O(n) is defined by the character χU N +1 (P N , O(n)) =
N
(−1)m Tr H m (P N ,O(n)) e
λi Hi
, e
λi Hi
∈ U N +1 .
(2.8)
m=0
Cohomology groups of O(n) on projective spaces P N have the following properties (see e.g. [OSS]): dim H m (P N , O(n)) = 0, m = 0, N , dim H N (P N , O(n)) = 0, n ≥ 0, dim H 0 (P N , O(n)) = 0, n < 0.
(2.9)
Taking into account (2.9) the expression (2.8) reduces to χU N +1 (P N , O(n)) = Tr H 0 (P N , O(n)) e
λi Hi
+ (−1) N Tr H N (P N , O(n)) e
λi Hi
. (2.10)
(d)
We have QMd (P ) = P(+1)(d+1)−1 and, thus, for n ≥ 0, we can identify An,k (z, q) with equivariant holomorphic Euler characteristic of Lk (n),
L 0 + A(d) n,k (z, q) = Tr H 0 (QMd (P ),Lk (n)) e
λi Hi
= χG (QMd (P ), Lk (n)), n ≥ 0,
where G = S 1 × U+1 . The equivariant Euler characteristic of a holomorphic vector bundle on the projective space possesses a canonical integral representation. The equivariant cohomology of a point with respect to the maximal compact subgroup G = S 1 × U+1 of C∗ × G L +1 can be described as HG∗ (pt, C) = C[λ1 , . . . , λ+1 ]S+1 ⊗ C[], where λ1 , . . . , λ+1 and are associated with the generators H1 , . . . , H+1 and L 0 respectively. According to the Riemann-Roch-Hirzebruch (RRH) theorem, one can express the U N +1 -equivariant holomorphic Euler characteristic of a U N +1 -equivariant vector bundle E on P N as follows: χU N +1 (P N , E) =
N
(−1)m Tr H m (P N ,E ) e
λi Hi
m=0
= ChU N +1 (E) TdU N +1 (T P N ), [P N ],
(2.11)
where Hi are generators of the Lie algebra gl N +1 , T P N is the tangent bundle to P N , ChU N +1 (E) is a U N +1 -equivariant Chern character of E, and TdU N +1 (E) is a U N +1 -equivariant Todd genus of E [A]. The tangent bundle T P N to the projective space P N is U N +1 -equivariantly stableequivalent to O(1)⊕(N +1) as the following lemma shows.
On q-Deformed gl+1 -Whittaker Function II
131
Lemma 2.1. The following relation holds in U N +1 -equivariant topological K -theory K U N +1 (P N ): [T P N ] ⊕ [O] = [O(1)]⊕(N +1) ,
(2.12)
where [E] is a class of a vector bundle E in K U N +1 (P N ). Proof. For a tangent sheaf to the complex projective space P N we have the Euler exact sequence (see e.g. [GH]) 0 −→ O −→ O(1)⊕(N +1) −→ T P N −→ 0.
(2.13)
The maps (2.13) are explicitly U N +1 -equivariant and, thus, we obtain the relation (2.12) in U N +1 -equivariant K -groups of P N .
Lemma 2.1 and the fact that the Todd class depends only on the stable equivalence class of a vector bundle allows us to rewrite RRH-theory on projective spaces as follows: χU N +1 (P N , E) = ChU N +1 (E) TdU N +1 (O(1)⊕(N +1) , [P N ].
(2.14)
In the following we will consider only the case of line bundles and thus we take E = O(n), n ∈ Z. The pairing of the cohomology classes with the fundamental class entering the formulation of the RRH-theorem can be expressed explicitly using a particular model for the equivariant cohomology ring HU∗ N +1 (P N , C). The cohomology ring H ∗ (P N , C) is generated by an element x ∈ H 2 (P N , C) with a single relation x N +1 = 0, H ∗ (P N , C) = C[x]/x N +1 .
(2.15)
The U N +1 -equivariant analog of (2.15) is given by HU∗ N +1 (P N , C) = C[x] ⊗ C[λ1 , . . . , λ N +1 ]SN +1
⎛ ⎝
N +1
⎞ (x − λ j )⎠ ,
j=1
which is naturally a module over HU∗ N +1 (pt, C) = C[λ1 , . . . , λ N +1 ]SN +1 , where S N +1 is the permutation group of a set of N + 1 elements. The pairing of an element of HU∗ N +1 (P N , C) represented by P(x, λ) with a U N +1 -equivariant fundamental cycle [P N ] can be expressed in terms of the integral 1 P(x, λ) , dx N +1 P(λ), [P N ] = 2πı C0 j=1 (x − λ j ) where the integration contour C0 encircles the poles x = λ j , j = 1, . . . , (N + 1). The pairing for H ∗ (P N , C) is obtained by a specialization λ j = 0, j = 1, . . . , (N + 1). The equivariant Chern character and Todd class can be written in terms of this model of HU∗ N +1 (P N , C) as follows (see e.g. [BGV]): N +1 ChU N +1 (O(n)) = enx , TdU N +1 O(1)⊕(N +1) = j=1
(x − λ j ) 1 − e−(x−λ j )
.
132
A. Gerasimov, D. Lebedev, S. Oblezin
Therefore we have the following integral representation of the equivariant holomorphic Euler characteristic (t = e−x , z i = eλi , i = 1, . . . , N + 1): χU N +1 (z) = ChU N +1 (O(n)) TdU N +1 O(1)⊕(N +1) , [P N ] =
1 2πı
=−
1 2πı
dx
C0
N +1 i=1
(x − λi )
N +1 dt
t n+1
C0
i=1
enx
N +1 i=1
(x − λi ) (1 − e−(x−λi ) )
1 , 1 − t zi
(2.16)
(2.17)
where in the last expression the integration contour C0 encircles the poles t = z −1 j , j = 1, . . . , + 1. The integral representation (2.17) can be obtained directly using a particular realization of (U N +1 -equivariant) K -theory on P N (see e.g. [A]). The K -group K (P N ) is generated by a class t of the line bundle O(1) satisfying the relation (1 − t) N +1 = 0. We have the following isomorphisms for (U N +1 -equivariant) K -groups of projective spaces: K (P N ) = C[t, t −1 ]/(1 − t) N +1 ,
K U N +1 (P N ) = C[t, t −1 , z, z −1 ]
+1 N
(1 − t z j ).
j=1
(2.18) The equivariant analog of the pairing with the fundamental class of P N in K -theory is given by R, [P N ] K = −
1 2πı
C0
R(t) dt , N +1 t j=1 (1 − t z j )
(2.19)
where R(t) is a rational function representing an element of K U N +1 (P N ) and the integration contour C0 encircles the poles t = z −1 j , j = 1, . . . , (N +1). Using the representation of the pairing (2.19), one can represent the RRH expression for the Euler characteristic as 1 χU N +1 (z) = [O(n)], [P ] K = − 2πı
N
C0
N +1 dt
t n+1
i=1
1 . 1 − t zi
(2.20)
This reproduces the representation (2.17). Now we would like to apply the integral representation for the equivariant Euler characteristic to the S 1 × U+1 -equivariant line bundle Lk (n) on QMd (P ). Consider the S 1 × U+1 -equivariant cohomology of the projective space P(V(+1)(d+1) ), where the d 1 vector space V(+1)(d+1) = ⊕+1 j=1 ⊕m=0 V j,m has the structure of an S -module with an action given by eıθ : V j,m → eımθ V j,m , dim V j,m = 1, θ ∈ S 1 ,
On q-Deformed gl+1 -Whittaker Function II
133
and each Vm = V1,m ⊕ V2,m ⊕ . . . V+1,m is isomorphic to the standard U+1 -module. Then for the G = S 1 × U+1 -equivariant cohomology of P(V(+1)(d+1) ) we have an isomorphism HS∗1 ×U (P(V(+1)(d+1) )) +1
= C[x, λ, ]
+1 d
(x − λ j − m),
(2.21)
j=1 m=0
where x is a generator of H ∗ (P(V(+1)(d+1) ), C). The pairing with the S 1 × U+1 -equivariant fundamental cycle [P(V(+1)(d+1) )] can be expressed in the form of the contour integral P(λ, ), [P(V(+1)(d+1) )] =
1 2πı
C
P(x, λ, ) dx , +1 d j=1 m=0 (x − λ j − m)
(2.22)
where the integration contour C encircles the poles x = λ j + m, j = 1, . . . , + 1, m = 0, 1, . . . , d. Specializing the previous discussion to the action of G = S 1 ×U+1 on QMd (P ) (+1)(d+1)−1 P described in Sect. 2.1 we obtain Ch G (Lk (n)) = enx+k(λ1 +...+λ+1 ) , Td G (T P(+1)(d+1)−1 ) =
+1 d x − m − λi . 1 − eλi +m −x i=1 m=0
Let q = e, t = e−x , and z i = eλi , 1 ≤ i ≤ + 1, then χG (QMd (P ), Lk (n)) = Ch G (Lk (n)) Td G (T P(+1)(d+1)−1 ), [P ] +1 d d +1 1 (x − λi − m) 1 nx+k(λ1 +...+λ+1 ) e dx = 2πı C (x − λi − m) (1 − e−(x−λi −m ) ) i=1 m=0 i=1 n=0
+1 +1 d dt 1 . (2.23) = − z ik n+1 2πı t (1 − t zi q m ) C i=1 m=0
i=1
For n ≥ 0 one has the identity χG (QMd (P ), Lk (n)) = Tr Vn,k,d q L 0 e
i
λi Hi
.
Deforming the contour for n ≥ 0 we obtain the following integral representation for the character: Tr Vn,k,d q
L0
e
i
λi Hi
=
+1 i=1
which coincides with (2.5).
z ik
t=0
+1 d dt 1 , 2πı t n+1 (1 − t z i q m ) i=1 m=0
134
A. Gerasimov, D. Lebedev, S. Oblezin
Remark 2.1. Without a restriction n ≥ 0, the integral representation for the equivariant Euler characteristic can be represented as a difference of two terms:
+1 +1 d dt 1 k χG (QMd (P ), Lk (n)) = zi n+1 2πı t (1 − t zi q m ) t=0 i=1 i=1 m=0
+1 +1 d dt 1 z ik + . n+1 2πı t (1 − t zi q m ) t=∞ i=1 m=0
i=1
This decomposition corresponds to the decomposition (2.10), χG (QMd (P ), Lk (n)) = Tr H 0 (QMd (P ),Lk (n)) q L 0 e
i
λi Hi
+ (−1)(+1)(d+1)−1 Tr H (+1)(d+1)−1 (QMd (P ),Lk (n)) q L 0 e
i
λi Hi
. (0)
Remark 2.2. In the limit q → 0 one has an integral representation for a character χ(n,k) of an irreducible finite-dimensional representation Vn,k,0 = Symn C+1 ⊗ Lk of G L +1 :
+1 +1 dt 1 (0) λ1 H1 +...+λ+1 H+1 = z ik , (2.24) χ(n,k) (z) = tr Vn,k,0 e n+1 2π ı t 1 − t zi i=1
t=0
i=1
where z i = exp λi and the G L( + 1)-module Vn,k,0 , n ≥ 0 is realized as a zero-degree cohomology space H 0 (P , Lk (n)). 3. K-Theory of LP+ and q-Whittaker Functions In this section we establish a direct connection between the q-deformed class one specialized gl+1 -Whittaker functions and geometry of the space LP+ understood as an appropriate limit of QMd (P ) when d → +∞. Geometrically LP+ should be considered as a space of maps of algebraic disks to P (see [Gi1] for details). In general, let L X = Map(S 1 , X ) be the space of free contractible loops in a compact Kähler manifold X . There is a natural action of S 1 on L X by loop rotations. The universal covering LX can be defined as a space of maps D → X of the disk D considered up to a homotopy preserving the image of the boundary loop S 1 ⊂ D. The group of covering transformations of the universal cover L X → L X is isomorphic to π2 (X ). Let L X+ ⊂ L X be a semi-infinite cycle of loops that are boundaries of holomorphic maps D → X . For X = P define an algebraic version LP+ of LP + as a set of collections of the regular series ai (z) = ai,0 + ai,1 z + ai,2 z 2 + · · · , 0 ≤ i ≤ , modulo the action of C∗ . The topology on this space should be defined by considering LP+ as a limit of QMd (P ) when d → ∞. This space inherits the action of G = S 1 × U ( + 1) defined previously on QMd (P ). In the following we do not define appropriate topology rigorously leaving this for another occasion. Instead we define the limit d → +∞ on the level of cohomology of QMd (P ) and of the space of
On q-Deformed gl+1 -Whittaker Function II
135
holomorphic sections of line bundles on QMd (P ). Let us take the limit d → +∞ of (d) the character A(d) n,k (z, q) given by the integral expression (2.5). The limit of An,k (z, q) can be interpreted as a character of a C∗ × G L +1 -module Vn,k,∞ defined as follows. Let Vk,∞ be a linear space of polynomials of an infinite number of variables ai,m , i = 0, . . . , , m ∈ Z≥0 . Let L 0 be a generator of Lie(C∗ ), T ∈ G L +1 be a Cartan torus, and H1 , . . . , H+1 be a basis in Lie(T ). Define actions of L 0 and H j on the generators ai,m as follows: L 0 : ai,m −→ m ai,m
e
j
λj Hj
: ai,m −→ eλi ai,m .
This supplies Vk,∞ with a structure of a C∗ × G L +1 -module. Now the linear subspace Vn,k,∞ ⊂ Vk,∞ is defined as a subspace of polynomials of the variables ai,m , i = 0, . . . , , m ∈ Z≥0 of the total degree n. gl
Theorem 3.1. Let z +1 (n, k) be a specialization (1.11) of the solution of the q-deformed Toda lattice defined in Theorem 1.1. Then the following holds: gl+1
z
(n, k) = χn,k (z),
(3.1)
where L0 χn,k (z) = lim A(d) e n,k (z, q) = Tr Vn,k,∞ q d→∞
i
λi Hi
, z i = eλi .
(d)
Proof. For the function χn,k (z) = limd→∞ An,k (z, q) the following integral representation holds:
+1 +1 ∞ dt 1 χn,k (z) = z ik n+1 (1 − tq j z m ) t=0 2πı t m=1 j=0 i=1
+1 +1 dt = z ik q (t z m ). (3.2) n+1 t=0 2πı t m=1
i=1
The relations (3.1) follow directly from the explicit integral expression (3.2) and Lemma 1.1. The representation in terms of the trace over Vn,k,∞ follows from the statement of Proposition 2.1 with obvious modifications for d → +∞.
For n ≥ 0 we can identify the character A(d) with the equivariant Euler characteristic n,k expressed through the Riemann-Roch-Hirzebruch formula Ch G (Lk (n)) Td G (T QMd (P )). (3.3) χG (QMd (P ), Lk (n)) = QMd (P )
Note that the left and right hand sides of (3.3) depend on the equivariant parameters λ = (λ1 , . . . , λ+1 ). In the following we imply a continuation of the expressions entering (3.3) to λ in a complex domain. Taking the limit d → +∞ we should obtain a semi-infinite analog of the RiemannRoch-Hirzebruch formula for χG (LP+ , Lk (n)). Using the description (2.18), (2.19) of
136
A. Gerasimov, D. Lebedev, S. Oblezin
the equivariant K-groups of projective spaces and taking the limit d → +∞ in the integral representation of the Euler characteristic (2.23) one obtains the following integral representation for the equivariant Euler characteristic of line bundles on LP+ : ⎛ ⎞ +1 +1 ∞ dt 1 z kj ⎠ χG (LP+ , Lk (n)) = − ⎝ n+1 2πı t (1 − tq j z i ) C j=1 i=1 j=0 ⎛ ⎞ +1 +1 dt = −⎝ z kj ⎠ q (t z i ), (3.4) n+1 C 2π ı t j=1
i=1
where z i = eλi and the integration contour C encircles all poles except t = 0 and q (y) =
∞ n=0
1 . 1 − yq n
Problem 3.1. Define an equivariant (co)homology theory for LP+ in such a way that Chern and Todd classes Ch G (Lk (n)), Td G (T LP+ ) make sense and the expression Ch G (Lk (n)) Td G (T LP+ ), LP+
is well-defined and is equal to ⎛ gl z +1 (n, k)
= −⎝
+1 j=1
⎞ z kj ⎠
C
+1 dt q (t z i ). 2π ı t n+1
(3.5)
i=1
Remark 3.1. The conjectural relation above provides a description of the specialized q-deformed gl+1 -Whittaker function as a semi-infinite period, gl+1 Ch G (Lk (n)) Td G (T LP+ ), n ≥ 0. (3.6) z (n, k) = LP+
The K -theory of the semi-infinite spaces LP+ is closely connected with a quantum version of K -theory of projective spaces proposed in [GiL]. The generating function F(n, z, q) of the correlation functions in the K -theory version of Gromov-Witten theory with the target space P obeys the following difference equation [GiL]: +1 −1 (1 − z i T ) · F(n, z, q) = q n F(n, z, q),
(3.7)
i=1
where T · f (n) = f (n + 1). The specialized q-deformed gl+1 -Whittaker function satisfies the same equation (3.7) (see Lemma 1.1 and relation (1.13)). Therefore the Whittaker function can be considered as a correlation function of some special operator singled out gl by the class one condition (i.e. the condition z +1 (n, k) = 0 for n < 0). We provide some information on this operator in Sect. 5.
On q-Deformed gl+1 -Whittaker Function II
137
4. Quantum Cohomology and Whittaker Function In the previous section we proposed a description of the q-deformed class one specialized gl+1 -Whittaker function in terms of a semi-infinite version of RiemannRoch-Hirzebruch theorem. This expresses the q-Whittaker function as a semi-infinite period. Its classical (i.e. non-deformed) counterpart can be also expressed in terms of a semi-infinite period. In this section we provide this conjectural representation 2 . We start by recalling the notion of quantum cohomology. The quantum cohomology Q H ∗ (X ) of a compact symplectic manifold X can be defined in terms of semi-infinite geometry of a universal cover L X of the loop space L X . One of the descriptions is given by a Morse-Smale-Bott-Novikov-Floer complex constructed in terms of critical points of an area functional on L X . Its cohomology groups (interpreted as Floer cohomology groups F H ∗ ( L X ) of L X ) are isomorphic to the semi-infinite cohomology H ∞/2+∗ (L X ) arising naturally in the Hamiltonian formalism of a topological two-dimensional sigma model with the target space X . In the following we will use an equivariant version of quantum cohomology Q H ∗ (P ) of projective spaces considered in [Gi1] (see also [CJS] for a non-equivariant version). We have defined the universal covering L X of the loop space L X as a space of maps D → X of the disk D considered up to a homotopy preserving the image of the boundary loop S 1 ⊂ D. The group of covering transformations of the universal cover L X → L X is isomorphic to the image ⊂ H2 (X ) of the Hurewicz homomorphism π2 (X ) → H2 (X ), where H2 (X ) denotes integral homology modulo torsion. Let r be ±1 the rank of and C [] C[u ±1 1 , . . . , u r ] be its group algebra. As a vector space ∗ the quantum cohomology Q H (X ) of X is isomorphic to the ordinary cohomology ±1 ±1 ±1 1 H ∗ (X, C[u ±1 1 , . . . , u r ]), over the group algebra C [] C[u 1 , . . . , u r ]. Let S act 1 on the loop space L X by loop rotations. For the corresponding S -equivariant quantum cohomology we have the following isomorphism of vector spaces: ±1 Q HS∗1 (X ) = H ∗ X, C[u ±1 1 , . . . , u r ]() , where we use the identification HS∗1 (pt, C) = H ∗ (B S 1 , C) = C[], and the standard localization of the equivariant cohomology HS∗1 (pt) with respect to the maximal ideal generated by is implied. The quantum cohomology space Q HS∗1 (X ) has a natural structure of a module over an algebra D generated by u i = exp τi , vi = −∂/∂τi , i = 1, . . . , r . More precisely, Q HS∗1 (X ) as a linear space over C() is generated by solutions of the system of linear differential equations f (τ ) = ( f 1 (τ ), f 2 (τ ), . . . , f n (τ )), n = dim H ∗ (X ), (4.1) r where the flat connection ∇ = i=1 dτi ∇i provides an action of vi on Q HS∗1 (X ). The D-module Q HS∗1 (P ) is of rank one. It is generated by an element f ∗ satisfying the relation (v +1 − u) f ∗ = 0, i.e. the quantum cohomology can be represented as ∇i f (τ ) = 0,
2 Whittaker functions naturally arise in the description of Gromov-Witten invariants of flag spaces. In the mirror dual description they are expressed in terms of periods of top-dimensional holomorphic forms on non-compact Calabi-Yau spaces [Gi2]. Thus, the possibility to express Whittaker functions as semi-infinite periods leads to a formulation of the mirror symmetry as an identification of two period maps - semi-infinite and finite ones.
138
A. Gerasimov, D. Lebedev, S. Oblezin
Q H ∗ (P ) D/(v +1 −u). Explicitly we have the differential equation for the generator f ∗ (τ ), ∂ +1 τ −e (4.2) f ∗ (τ, ) = 0. − ∂τ The representation (4.1) arises after transformation of (4.2) to the matrix differential equation of the first order. The (S 1 × U+1 )-equivariant analog of quantum cohomology Q H ∗ (P ) allows for a similar representation with the differential equation (4.2) replaced by +1 ∂ τ − + λk − e (4.3) f ∗ (τ, λ, ) = 0. ∂τ k=1
Proposition 4.1. The general solution of (4.3) is given by a linear combination of the integrals f
(a)
(τ, λ, ) =
γa
dλ e
+1
λτ
λk −λ
k=1
λk − λ , a = 1, · · · , n
(4.4)
with a suitable choice of a bases of the integration contours γa . Proof. Note that the function Q(λ, λ) =
+1
λk −λ
k=1
λk − λ ,
(4.5)
obeys the difference equation +1
(λ − λk ) Q(λ, λ) = (−1)+1 Q(λ − , λ).
(4.6)
k=1
Therefore, the function f (τ, λ) =
γ
λτ
dλ e Q(λ, λ, )
(4.7)
satisfies (4.3) provided the choice of the contour γ allows for an integration by parts. The contours can be chosen in such a way that the total derivatives do not give a contribution into the integral (4.4).
A particular choice of γ in (4.7) gives us a special solution of Eq. (4.3), f ∗ (τ, λ, ) =
σ +ı∞
σ −ı∞
dλ e
λτ
+1 k=1
λk −λ
λk − λ ,
(4.8)
where σ is such that σ < min { Re λ j , j = 1, . . . , + 1}. This is a unique solution of (4.3) exponentially decaying when τ → +∞.
On q-Deformed gl+1 -Whittaker Function II
139
Remark 4.1. In the case of = 1, the differential equation (4.2) is equivalent to an eigenvalue problem for the quadratic Hamiltonian of the sl2 -Toda chain. The solution given in the integral form (4.8) coincides in this case with the Mellin-Barnes representation of the gl2 -Whittaker function (1.16) for x2 = 0 and x1 = τ . Replacing formally -functions by infinite products over their poles one has for f ∗ (τ, λ, ) the following expression:
dx eτ x/
+1 ∞ j=1 n=0
1 . x − λ j − n
(4.9)
This formal representation can be interpreted using the model for cohomology of QMd (P ) discussed in Sect. 2.3. Naively (4.9) can be considered as an integral over LP+ of exp(τ ω/), where ω is an element of the second S 1 × U+1 -equivariant cohomology of LP+ . Recall that we define LP+ on the level of cohomology as a limit of QMd (P ) when d → +∞. However a correct regularization for (4.9) is given by (4.8) and, thus, a geometric interpretation of (4.8) implies some modification of LP+ . We attribute the difference between (4.9) and (4.8) to the fact that the proper interpretation of LP+ as d → +∞ limit deserves more care in this case and does not coincide with a straightforward limit d → +∞ on the level of cohomology. Let us denote the corresponding hypothetical modified limit by LP+ . Problem 4.1. Find the space LP+ and construct the equivariant (co)homology for LP+ in such a way that the integral eτ ω/, ω ∈ HS21 ×U (LP+ , C) +1
LP+
is well-defined and is equal to f ∗ (τ, λ, ) given by (4.8). 5. S1 -localization In this section we clarify a relation between our construction of q-Whittaker functions and the results of [GiL]. To do this we calculate the equivariant Euler characteristic χG (QMd (P ), Lk (n)) for G = S 1 × U+1 using Borel localization for S 1 -action. The character (2.23) can be calculated using an equivariant localization as follows. We have a compact Lie group G = S 1 × U+1 acting on the projective space X = QMd (P ) ≡ P(+1)(d+1)−1 . Recall that QMd (P ) is defined as a set of ( + 1) polynomials each of degree d considered up to a common constant factor (a0 (y), a1 (y), . . . , a (y)) ∼ (ρ a0 (y), ρ a1 (y), . . . , ρ a (y)) , ρ ∈ C∗ , where ar (y) =
d
ar,i y1i y2d−i , r = 0, . . . , .
i=0
The action of an element q = eıθ of S 1 on ar, j is given by q : ar, j → q j ar, j , r = 0, . . . , .
140
A. Gerasimov, D. Lebedev, S. Oblezin 1
The line bundles Lk (n) are equivariant with respect to the action of S 1 . Let X S ⊂ X be a set of S 1 -fixed points. It is a union of smooth components Yi . The Bott localization formula gives the following expression for the equivariant Euler characteristic (2.23) (see e.g. [BGV]): Ch G (Lk (n)|Y )Td G (T Yi ) i χG (QMd (P ), Lk (n)) = , (5.1) E (N G Yi ) Y i 1 Yi ∈X S
1
where the sum runs over all components Yi in X S , NYi is the normal bundle of Yi in X , Ch G (Lk (n)), Td G (Yi ) are equivariant Chern character and Todd class and EG (NYi ) is the equivariant Euler class of NYi . It is easy to infer that the subvarieties Yi , i = 0, . . . , in QMd (P ) are isomorphic to the projective spaces P and are defined by the equations Yi = {ak, j = 0, j = i}, i = 0, . . . , . To calculate the action of S 1 on the normal bundle to Yi we consider the intersection of Yi with open subsets Uak,i = {ak,i = 0} of QMd (P ). Natural coordinates on Uak,i are ξr, j = ar, j /ak,i , (r, j) = (k, i), and the intersections Yi ∩ Uak,i = {ak,i = 0} are defined by the equations ξr,i = 0, r = k. Thus one can take a collection of coordinates ξr, j , j = i as a local section of the dual to the normal bundle NYi . The action of S 1 on NYi can then be found by considering the action on section ξr, j : ξr, j → q j−i ξr, j . Similarly, one can show that q ∈ S 1 acts on the restriction of the line bundle O(n) on Yi by multiplication on q n i . The fixed point formula (5.1) reduces to the following explicit expression: χG (QMd (P ), Lk (n)) ⎞ ⎛ +1 d z kj ⎠ = −⎝ j=1
i=0
C0
dt q ni 1 , +1 d +1 n+1 m−i 2πıt ) j=1 (1 − t z j ) j=1 m=0,m=i (1 − t z j q
where the integration contour C0 encircles the ( + 1) poles defined by the equations t = z −1 j , j = 1, . . . , + 1. Lemma 5.1. The following identity holds for n ≥ 0: dt 1 n+1 +1 d m 2πıt C j=1 m=0 (1 − t z j q ) =
d i=0
C0
1 dt q ni , +1 d m−i 2πıt n+1 +1 ) j=1 (1 − t z j ) j=1 m=0,m=i (1 − t z j q
(5.2)
−i where the integration contour C encircles poles defined by the equations t = z −1 j q , j = 1, . . . , + 1, i = 0, . . . , d and the integration contour C0 encircles ( + 1) poles defined by the equations t = z −1 j , j = 1, . . . , + 1.
On q-Deformed gl+1 -Whittaker Function II
141
Proof. We have that C
d dt 1 dt 1 = , n+1 n+1 +1 d +1 d m m 2πıt Ci 2πıt j=1 m=0 (1 − t z j q ) j=1 m=0 (1 − t z j q ) i=0
where the integration contour Ci encircles ( + 1) poles defined by the equations t = −i −i in the r.h.s., we z −1 j q , j = 1, . . . , + 1. Making the change of variables t → tq obtain that dt 1 n+1 +1 d m 2πıt C j=1 m=0 (1 − t z j q ) d dt 1 = n+1 +1 d m 2πıt j=1 m=0 (1 − t z j q ) i=0 Ci d dt 1 ni = . q n+1 +1 d m−i ) 2πıt C0 j=1 m=0 (1 − t z j q i=0
We are going to consider a continuation of the expression (5.2) to q ∈ C∗ , |q| < 1 and the limit d → ∞. Proposition 5.1. The specialization (1.11), (1.14) of the q-deformed gl+1 -Whittaker function can be written in the following form gl+1
z
˜ (n, k) = In,k (z) L(z), [P ] K ,
(5.3)
where +1 1 ˜ t) = = q (qt z j ), L(z, +1 ∞ k j=1 k=1 (1 − t z j q ) j=1
⎛ In,k (z, t) = ⎝
+1 j=1
⎞ z kj ⎠ t −n
∞ i=0
q ni +1 i j=1
(5.4)
1
m=1 (1 − t z j q
−m )
,
(5.5)
and the pairing , K is the standard pairing (2.19) on K U+1 (P ) taking values in K U+1 ( pt). The representation of the Whittaker function given in Proposition 5.1 establishes a direct connection with the results of Givental-Lie [GiL]. In [GiL] the function (5.5) was interpreted as a universal solution of the reduction (1.13) of the q-deformed gl+1 -Toda chain. Indeed, the function In,k (z, t) satisfies the eigenvalue problem +1 j=1
(1 − z i T −1 ) In,k (z, t) = q n In,k (z, t),
(5.6)
142
A. Gerasimov, D. Lebedev, S. Oblezin
modulo the relation +1 j=1 (1−t z j ) = 0 holding in K U+1 (P ) and is uniquely determined by the normalization condition ⎛ ⎞ +1 In,k (z, t)|q=0 = ⎝ z kj ⎠ t −n , n ≥ 0. j=1
The solution In,k (z, t) is universal in the sense that taking the pairing dt In,k (z, t) f (t) 1 , In,k (z), f K = − +1 2π ı C0 t j=1 (1 − z j t)
(5.7)
with arbitrary f ∈ K U+1 (P ) one obtains a solution (5.7) of the q-deformed reduced ˜ gl+1 -Toda chain (1.13). Taking f = L(z) given by (5.4) we recover the specialized q-deformed gl+1 -Whittaker function constructed in this paper. Acknowledgement. We are grateful to A. Rosly for useful discussions. The research of AG was partly supported by SFI Research Frontier Programme and Marie Curie RTN Forces Universe from EU. The research of SO is partially supported by RF President Grant MK-134.2007.1. The research was also partially supported by Grant RFBR-08-01-00931-a.
References [A] [BGV]
Atiyah, M.F.: K-theory. New York: W. A. Benjamin, 1967 Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators, Berlin-Heidelberg-New York: Springer, 1992 [Bra] Braverman, A.: Spaces of Quasi-Maps into the Flag Varieties and their Applications. Proc. ICM 2006, Madrid, Zurich: Eur. Math. Soc., 2006 [BF] Braverman, A., Finkelberg, M.: Finite-difference quantum Toda lattice via equivariant K-theory. Trans. Groups 10, 363–386 (2005) [CS] Casselman, W., Shalika, J.: The unramified principal series of p-adic groups II. The Whittaker function. Comp. Math. 41, 207–231 (1980) [CJS] Cohen, R.L., Jones, J.D.S., Segal, G.B.: Floer’s Infinite Dimensional Morse theory and Homotopy theory. Floer Memorial Volume, Birkhauser Verlag, Prog. in Math. 133, Basel: Birlchäuser, 1995, pp. 287–325 [FF] Feigin, B., Frenkel, E.: Affine Kac-Moody algebras and semi-infinite flag manifolds. Commun. Math. Phys. 128, 161–189 (1990) 3 subspaces and quantum [FFJMM] Feigin, B., Feigin, E., Jimbo, M., Miwa, T., Mukhin, E.: Principal sl Toda hamiltonians. http://arxiv.org/abs/0707.1635v2[math.QA], 2007 [FM] Finkelberg, M., Mirkovi´c, I.: Semiinfinite Flags. I. Case of a global curve P1 . In: Differential Topology, Infinite-Dimensional Lie Algebras, and Applications. Amer. Math. Soc. Transl. Ser. 2, 194, Providence, RI: Amer. Math. Soc., 1999, pp. 81–112 [FFM] Feigin, B., Finkelberg, M., Mirkovi´c, I.: Semiinfinite Flags. II. Local and Global Intersection Cohomology of Quasimaps’ Spaces. In: Differential Topology, Infinite-dimensional Lie Algebras, and Applications. Amer. Math. Soc. Transl. Ser. 2, 194, Providence, RI: Amer. Math. Soc., 1999, pp. 113–148 [GKL1] Gerasimov, A., Kharchev, S., Lebedev, D.: Representation Theory and Quantum Integrability, Progress in Math. 237 Basel: Birlchäuser, 2005, pp. 133–156 [GKLO] Gerasimov, A., Kharchev, S., Lebedev, D., Oblezin, S.: On a Gauss-Givental Representation of Quantum Toda Chain Wave Function. Int. Math. Res. Notices 2006, Aricle ID 96489, 23 pages (2006) [GLO1] Gerasimov, A., Lebedev, D., Oblezin, S.: On q-deformed gl+1 -Whittaker Functions I. Commun. Math. Phys., to appear, this issue [GLO2] Gerasimov, A., Lebedev, D., Oblezin, S.: On q-deformed gl+1 -Whittaker Functions III. http:// arxiv.org/abs/0805.3754v2[math.RT], 2008
On q-Deformed gl+1 -Whittaker Function II [GLO3] [GLO4] [Gi1] [Gi2] [Gi3] [GiL] [GH] [JK] [KLS] [Ku] [La] [OSS] [Sh] [W]
143
Gerasimov, A., Lebedev, D., Oblezin, S.: Baxter Q-operators and their arithmetic implications, Lett. Math. Phys. 88(1–3), 3–30 (2009) Gerasimov, A., Lebedev, D., Oblezin, S.: Baxter operator and Archimedean Hecke algebra, Commun. Math. Phys. 284(3), 867–896 (2008) Givental, A.: Homological geometry I. Projective hypersurfaces. Selecta Mathematica, New Series Volume 1, 2 (1995), 325–345 Givental, A.: Stationary phase integrals, quantum Toda lattices, flag manifolds and the mirror conjecture. In: Topics in Singularity Theory, Amer. Math. Soc. Transl. Ser., 2 Vol. 180, Providence, RI: Amer.Math.Soc., 1997, pp. 103–115 Givental, A.: A mirror theorem for toric complete intersections. In: Topological Field Theory, Primitive Forms and Related Topics, ed. M. Kashiwara, A. Matsuo, K. Saito, I. Satake, Boston: Birkhäuser, 1998, p.141 Givental, A., Lee, Y.-P.: Quantum K-theory on flag manifolds, finite-difference Toda lattices and quantum groups. Invent. Math. 151, 193–219 (2003) Griffiths, Ph., Harris, J.: Principles of Algebraic Geometry. New York: Wiley-Interscience, 1978 Joe, D., Kim, B.: Equivariant mirrors and the Virasoro conjecture for flag manifolds. Int. Math. Res. Notices 2003(15), 859–882, (2003) Kharchev, S., Lebedev, D., Semenov-Tian-Shansky, M.: Unitary representations of u q (sl(2, r )), the modular double and the multiparticle q-deformed Toda chains. Commun. Math. Phys. 225, 573–609 (2002) Kuznetsov, A.: Laumon’s resolution of Drinfeld’s compactification is small. Math. Res. Lett. 4, 349–364 (1997) Laumon, G.: Faisceaux automorphes liés aux séries d’Eisenstein. Persp. Math 10, 227–281 (1990) Okonek, Ch., Schneider, M., Spindler, H.: Vector Bundles on Complex Projective Spaces. Progress in Math. 3, Boston, MA: Birkhäuser, 1980 Shintani, T.: On an explicit formula for class 1 Whittaker functions on gln over p-adic fields. Proc. Japan Acad. 52, 180–182 (1976) Witten, E.: Phases of N = 2 theories in two dimensions. Nucl. Phys. B403, 159–222, (1993)
Communicated by Y. Kawahigashi
Commun. Math. Phys. 294, 145–168 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0855-8
Communications in
Mathematical Physics
Nonlinear Diffusion of Dislocation Density and Self-Similar Solutions Piotr Biler1 , Grzegorz Karch1 , Régis Monneau2 1 Instytut Matematyczny, Uniwersytet Wrocławski, pl. Grunwaldzki 2/4,
50-384 Wrocław, Poland. E-mail:
[email protected];
[email protected] URL: http://www.math.uni.wroc.pl/~karch 2 CERMICS, École Nationale des Ponts et Chaussées, 6 et 8 Avenue Blaise Pascal, Cité Descartes, Champs-sur-Marne, 77455 Marne-la-Vallée Cedex 2, France. E-mail:
[email protected] URL: http://cermics.enpc.fr/~monneau Received: 8 February 2009 / Accepted: 25 March 2009 Published online: 25 June 2009 – © Springer-Verlag 2009
Abstract: We study a nonlinear pseudodifferential equation describing the dynamics of dislocations in crystals. The long time asymptotics of solutions is described by the self-similar profiles. 1. Introduction Dislocation dynamics. Dislocations are line defects in crystals whose typical length is ∼10−6 m and their thickness is ∼10−9 m. When the material is submitted to shear stresses, these lines can move in the crystallographic planes and this dynamics can be observed using electron microscopy. The elementary mechanisms at the origin of the deformation of monocrystals are rather well understood, however, many questions concerning the plastic behavior of materials containing a high density of defects are still open. Hence, in recent years, new physical theories describing the collective behavior of dislocations have been developed and numerical simulations of dislocations have been performed. We refer the reader to the recent publications [1,23] for the comprehensive references about modeling of dislocation dynamics. One possible (simplified) model of the dislocation dynamics is given by the system of ODEs, y˙i = F − V0 (yi ) − V (yi − y j ) for i = 1, . . . , N , (1.1) j∈{1,...,N }\{i}
where F is a given constant force, V0 is a given potential and V is a potential of twobody interactions. One can think of yi as the position of dislocation straight lines. In this model, dislocations can be of two types, + or −, depending on the sign of their Burgers vector (see the book by Hirth and Lothe [20] for a physical definition of the Burgers vector).
146
P. Biler, G. Karch, R. Monneau
Self-similar solutions (i.e. solutions of the form yi (t) = g(t)Yi with constant Yi ) to system (1.1) with the particular potential V (z) = − 1z as well as their role in the asymptotic behavior of other solutions of (1.1) were studied by Head in [16,17]. More recently, Forcadel et al. showed in [14, Th. 8.1] that, under suitable assumptions on V0 and V in (1.1), the rescaled “cumulative distribution function” N t 1 ρ ε (x, t) = ε − + H x − εyi (1.2) 2 ε i=1
(where H is the Heaviside function) satisfies (as a discontinuous viscosity solution) the following nonlocal eikonal equation: ε ρ (·, t) x + Mε (x) |ρxε (x, t)| (1.3) ρtε (x, t) = c ε ε for (x, t) ∈ R × (0, +∞), with c(y) = V0 (y) − F. Here, M ε is the nonlocal operator defined by
ε M (U )(x) = J (z) E U (x + εz) − U (x) dz, (1.4) R
V (z)
where J (z) = on R\{0} and E is the modification of the integer part: E(r ) = k + 1/2 if k ≤ r < k + 1. Note that the nonlocal operator M ε describes the interactions between dislocation lines, hence, interactions are completely characterized by the kernel J . Next, under the assumption that the kernel J is a sufficiently smooth, even, nonnegative function with the following behavior at infinity J (z) =
1 for all |z| ≥ R0 |z|2
(1.5)
and for some R0 > 0, the rescaled cumulative distribution function ρ ε , defined in (1.2), is proved to converge (cf. [14, Th. 2.5]) toward the unique solution of the corresponding initial value problem for nonlinear diffusion equation (−u, u x ), ut = H
(1.6)
is a continuous function and is a Lévy operator of order 1. where the Hamiltonian H It is defined for any function U ∈ Cb2 (R) and for r > 0 by the formula 1 − U (x) = C(1) U (x + z) − U (x) − zU (x)1{|z|≤r } dz (1.7) |z|2 R with a constant C(1) > 0. Finally, in the particular case of c ≡ 0 in (1.3), we have (L , p) = L| p| (cf. [14, Th. 2.6]) which allows us to rewrite Eq. (1.6) in the form H u t + |u x |u = 0.
(1.8)
One can show that the definition of is independent of r > 0, hence, we fix r = 1.
1/2 In fact, for suitably chosen C(1), = 1 = −∂ 2 /∂ x 2 is the pseudodifferential 1 w)(ξ ) = |ξ | operator defined in the Fourier variables by ( w(ξ ) (cf. formulae (2.3) and (2.4) below). In this particular case, Eq. (1.8) is an integrated form of a model studied
Nonlinear Diffusion of Dislocation Density
147
by Head [18] for the self-dynamics of a dislocation density represented by u x . Indeed, denoting v = u x we may rewrite Eq. (1.8) as vt + (|v|Hv)x = 0,
(1.9)
where H is the Hilbert transform defined by ) = − i sgn(ξ ) (Hv)(ξ v (ξ ). Let us recall two well known properties of this transform (cf. [30]) 1 v(y) Hv(x) = P.V. dy and 1 v = Hvx . π x − y R
(1.10)
(1.11)
Head [18] called (1.9) the equation of motion of the dislocation continuum and constructed an explicit self-similar solution. Numerical studies of solutions to this equation were performed in [12]. Quasi-geostrophic equations. Let us recall completely different physical motivations which also lead to Eq. (1.9). The 2D quasi-geostrophic equations (QG), modeling the dynamics of the mixture of cold and hot air in a thin layer and the fronts between them, are of the form θt + (u · ∇)θ = 0, u = ∇ ⊥ ψ, θ = −(− )1/2 ψ R2
(1.12)
∇⊥
for x ∈ and t > 0, where = (−∂x2 , ∂x1 ). Here, θ (x, t) represents the air temperature. Pioneering studies concerning a finite time blow up criterion of solutions to (1.12) are due to Constantin et al. [9]. Much earlier, Constantin et al. [8] proposed a one-dimensional version of the Euler equations, namely, ωt + ωHω = 0, and studied formation of singularities of its solutions. Those results motivated the authors of [6,7,10] to consider another simplified model derived from (1.12) in the following way. First, one should write system (1.12) in another equivalent form. From the second and third equation in (1.12) we have the representation u = −∇ ⊥ (− )−1/2 θ = −R ⊥ θ,
(1.13)
where we have used the notation R ⊥ θ = (−R2 θ, R1 θ ), with the Riesz transforms defined by (see e.g. [30]) (x j − y j )θ (y, t) 1 R j (θ )(x, t) = P.V. dy. 2 2π |x − y|3 R Using Eq. (1.13), we find that (1.12) can be transformed into θt − div ((R ⊥ θ )θ ) = 0,
(1.14)
because div (R ⊥ θ ) = 0. To construct the 1D model, the authors of [6,7,10] considered the unknown function θ = θ (x, t) for x ∈ R and t > 0, and replaced the Riesz transform −R ⊥ in (1.14), by the Hilbert transform H (cf. (1.10)–(1.11)). Then Eq. (1.14) is converted into the model equation θt + (θ Hθ )x = 0 for x ∈ R and t > 0.
(1.15)
148
P. Biler, G. Karch, R. Monneau
Obviously, for θ ≥ 0, both models (1.9) and (1.15) are identical. However, in the case of Eq. (1.15), it is possible to show that the complex valued function z(x, t) = Hθ (x, t) − iθ (x, t) satisfies the inviscid Burgers equation z t + zz x = 0. This property of solutions has been systematically used in [6,7] to study the existence, the regularity and the blow up in finite time of solutions to Eq. (1.15). We refer the reader to those publications for additional references concerning Eq. (1.15). Below, see Remarks 2.6, 2.10 and 7.7, we explain how our results on Eq. (1.9) and its generalizations contribute to the theory developed for model (1.15). Organization of the paper. In the next section, we state the initial value problem considered in this paper and we formulate our main results. In Sect. 3, we construct explicitly the self-similar solution. In Sect. 4, we recall the necessary material about viscosity solutions, which will be used systematically in the remainder of the paper. In Sect. 5, we prove the uniqueness of the self-similar solution. Under the additional assumption that the solution is confined between its boundary values at infinity, we prove the stability of the self-similar solution, namely Theorem 2.5. In Sect. 6, we prove further decay properties of a solution with compact support. Applying these estimates, we finish the proof of Theorem 2.5 in the general case. In Sect. 7, we introduce an ε-regularized equation, for which we prove both the global existence of a smooth solution and the corresponding gradient estimates. Finally in Sect. 8, we deduce the gradient estimate in the limit case ε = 0, namely Theorem 2.7, using the corresponding estimates for the approximate ε-problem. 2. Main Results Motivated by physics described above, we study the following initial value problem for the nonlinear and nonlocal equation involving u = u(x, t), u t = −|u x | α u on R × (0, +∞), u(x, 0) = u 0 (x) for x ∈ R,
(2.1) (2.2)
where the assumptions on the initial datum u 0 will be made precise later. Here, for
α/2 α ∈ (0, 2), α = −∂ 2 /∂ x 2 is the pseudodifferential operator defined via the Fourier transform α w)(ξ ) = |ξ |α w ( (ξ ).
(2.3)
Recall that the operator α has the Lévy–Khintchine integral representation for every α ∈ (0, 2),
dz − α w(x) = C(α) , (2.4) w(x + z) − w(x) − zw (x)1{|z|≤1} |z|1+α R where C(α) > 0 is a constant. This formula (discussed in, e.g., [13, Th. 1] for functions w in the Schwartz space) allows us to extend the definition of α to functions which are bounded and sufficiently smooth, however, not necessarily decaying at infinity. As we have already explained (cf. Eq. (1.8)), in the particular case α = 1, Eq. (2.1) is a mean field model that has been derived rigorously in [14] as the limit of a system of particles in interactions (cf. (1.1)) with forces V (z) = − 1z . Here, the density u x means the positive density |u x | of dislocations of type of the sign of u x . Moreover, the
Nonlinear Diffusion of Dislocation Density
149
occurrence of the absolute value |u x | in the equation allows the vanishing of dislocation particles of the opposite sign. In the present paper, we study the general case α ∈ (0, 2) that could be seen as a mean field model of particles modeled by system (1.1) with repulsive interactions V (z) = − z1α . Here, we would like also to keep in mind that (2.1) is the simplest nonlinear anomalous diffusion model (described by the Lévy operator α ) which degenerates for u x = 0. First note that Eq. (2.1) is invariant under the scaling u λ (x, t) = u(λx, λα+1 t) u(x, t) is a solution to (2.1), then u λ
(2.5) u λ (x, t)
for each λ > 0 which means that if u = = is so. Hence, our first goal is to construct self-similar solutions of Eq. (2.1), i.e. solutions which are invariant under the scaling (2.5). By a standard argument, any self-similar solution should have the following form: x u α (x, t) = α (y) with y = 1/(α+1) , (2.6) t where the self-similar profile α has to satisfy the following equation: − (α + 1)−1 y α (y) = −(α α (y)) α (y) for all y ∈ R.
(2.7)
In our first theorem, we construct solutions to Eq. (2.7). Theorem 2.1 (Existence of self-similar profile). Let α ∈ (0, 2). There exists a nondecreasing function α of the regularity C 1+α/2 at each point and analytic on the interval (−yα , yα ) for some yα > 0, which satisfies
0 on (−∞, −yα ),
α = 1 on (yα , +∞), and (α α )(y) =
y for all y ∈ (−yα , yα ). α+1
Remark 2.2. We can obtain the self-similar solutions corresponding to different boundary values at infinity, simply considering for any γ > 0 and b ∈ R the profiles
γ α γ −1/(α+1) y + b which are also solutions of Eq. (2.7). Remark 2.3. The fact that ∂ y α has compact support reveals a finite velocity propagation of the support of the solution which is typical for solutions the porous medium equation, cf. Remark 2.8 below. At least formally, the function α is the solution of (2.7), and the self-similar function u α given by (2.6) is a solution of Eq. (2.1) with the initial datum being the Heaviside function
0 if x < 0, (2.8) u 0 (x) = H (x) = 1 if x > 0. In order to check that u α given by (2.6) solves (2.1), we introduce a suitable notion of viscosity solutions to the initial value problem (2.1)–(2.2), see Sect. 4. In this setting, we show in Theorem 4.7 the existence and the uniqueness of a solution for any initial condition u 0 in BU C(R), i.e. the space of bounded and uniformly continuous functions on R. Although the initial datum (2.8) is not continuous, we have the following result.
150
P. Biler, G. Karch, R. Monneau
Theorem 2.4 (Uniqueness of self-similar solution). Let α ∈ (0, 2). Then the function u α defined in (2.6) with the profile α constructed in Theorem 2.1 is the unique viscosity solution of Eq. (2.1) with the initial datum (2.8). In Theorem 2.4, the uniqueness holds in the sense that if u is another viscosity solution to (2.1), (2.8), then u = u α on (R × [0, +∞))\{(0, 0)}. The self-similar solutions are not only unique, but are also stable in this framework of viscosity solutions, as the following result shows. Theorem 2.5 (Stability of the self-similar solution). Let α ∈ (0, 2). For any initial data u 0 ∈ BU C(R) satisfying lim u 0 (x) = 0 and
x→−∞
lim u 0 (x) = 1,
x→+∞
(2.9)
let us consider the unique viscosity solution u = u(x, t) of (2.1)–(2.2) and, for each λ > 1, its rescaled version u λ = u λ (x, t) given by Eq. (2.5). Then, for any compact set K ⊂ (R × [0, +∞)) \ {(0, 0)}, we have x u λ (x, t) → α 1/(α+1) in L ∞ (K ) as λ → +∞. (2.10) t We stress the fact that Theorem 2.5 contains a result on the long time behaviour of solution because, first, choosing t = 1 in (2.10) and, next, substituting λ = t 1/(α+1) we 1/(α+1) obtain the convergence of u xt , t toward the self-similar profile α (x). On the other hand, convergence (2.10) can be seen as a stability result when we consider initial data which are perturbations of the Heaviside function. This is a nonstandard stability result in the framework of discontinuous viscosity solutions. It shows that the approach by viscosity solutions is a good one in the sense of Hadamard, even if we consider here initial conditions which are perturbations of the Heaviside function. Remark 2.6. In the particular case of α = 1, the nonnegative function U (x, t) = t −1/2 1 (xt −1/2 ), with 1 — the self-similar profile provided by Theorem 2.1, is the compactly supported self-similar solution of (1.15). This function attracts other nonnegative solutions to (1.15) in the sense stated in Theorem 2.5. Finally, we have the following result of independent interest. Theorem 2.7 (Optimal decay estimates). Let α ∈ (0, 1]. For any initial condition u 0 ∈ BU C(R) such that u 0,x ∈ L 1 (R), the unique viscosity solution u of (2.1)–(2.2) satisfies u(·, t) ∞ ≤ u 0 ∞ and u x (·, t) ∞ ≤ u 0,x ∞ for any t > 0. Moreover, for every p ∈ [1, +∞) we have pα+1
u x (·, t) p ≤ C p,α u 0,x 1p(α+1) t
( p−1) − p(α+1)
for any t > 0,
(2.11)
with some constant C p,α > 0 depending only on p and α. The decay given in (2.11) is optimal in the sense that the self-similar solution satisfies (u α )x (·, t) p = ( α ) y (·) p t
( p−1) − p(α+1)
.
Nonlinear Diffusion of Dislocation Density
151
Remark 2.8. The equation satisfied by v = u x , vt = −(|v|α−1 Hv)x
(2.12)
(with the Hilbert transform H defined in (1.10)) can be treated as a nonlocal counterpart of the porous medium equation.
Indeed, for α = 2 and for nonnegative v, Eq. (2.12) reduces to vt = (vvx )x = v 2 /2 x x . As in the case of the porous medium equation (see e.g. [32] and the references therein), estimates (2.11) show a regularizing effect created by the equation, even for the anomalous diffusion: if v0 ∈ L 1 (R) then v ∈ L p (R) for each p > 1. Observealso that Eq. (2.12) has compactly supported self-similar solution 1 1 − α+1 α+1 , where the profile α was constructed in Theorem 2.1. v(x, t) = t
α x/t This function for α = 2 corresponds to the well known Barenblatt–Prattle solution of the porous medium equation. Remark 2.9. As we have already mentioned, see Theorem 4.7 below, the initial value problem (2.1)–(2.2) has the unique global-in-time viscosity solution for any initial datum u 0 ∈ BU C(R). Under the additional assumption u 0,x ∈ L p (R), the corresponding solution satisfies u x (·, t) ∈ L p (R) for all t > 0. Indeed, this is an immediate consequence of the L p -inequalities stated in Remark 7.7 and of a limit argument analogous to that in Step 6 of the proof of Theorem 2.7. Remark 2.10. For any positive, sufficiently regular and vanishing at infinity initial condition v0 ∈ L 2 (R), the corresponding solution v = v(x, t) of (1.15) is global-in-time and analytic, see [6, Th. 2.1]. Since v = u x ≥ 0, this result holds true for solutions of problem (2.1)–(2.2) with α = 1. On the other hand, the nonexistence of global-in-time solutions to the initial value problem for Eq. (1.15) has been always proved assuming that the initial datum is (smooth enough and) negative at some point, see [7, Th. 2.1 and Remark 2.3], [6, Th. 3.1 and 4.8] and [10]. Those arguments cannot be applied to Eq. (2.1) with α = 1 due to the factor |v| (= |u x |) in the nonlinearity and lower regularity of the data. Remark 2.11. For α ∈ (1, 2), we do not know how to define the product |u x | (α u) in the sense of distributions, which is an obstacle for us to prove the result of Theorem 2.7 in this case, see Sect. 7. Note, however, that the inequalities from Theorem 2.7 are valid for α ∈ (1, 2] as well, provided the solution u = u(x, t) is sufficiently regular. 3. Construction of Self-Similar Solutions Proof of Theorem 2.1. The crucial role in the construction of the self-similar profile α is played by the function
2 α/2 for |x| < 1, (3.1) v(x) = K (α) 1 − |x| 0 for |x| ≥ 1, with K (α) = (1/2) [2α (1 + α/2)((1 + α)/2)]−1 . This function (together with its multidimensional counterparts) has an important probabilistic interpretation. Indeed, if {X (t)}t≥0 denotes the symmetric α-stable process in R of order α ∈ (0, 2] and if T = inf{t : |X (t)| > 1} is the first passage time of the process to the exterior of the segment {x : |x| ≤ 1}, Getoor [15] proved that Ex (T ) = v(x), where Ex denotes the expectation under the condition X (0) = x.
152
P. Biler, G. Karch, R. Monneau
In particular, it was computed in [15, Th. 5.2] using a purely analytical argument (based on definition (2.3) and on properties of the Fourier transform) that α v ∈ L 1 (R) and α v(x) = 1 for |x| < 1.
(3.2)
Now, for the function v, we define the bounded, nondecreasing, C 1+α/2 -function x u(x) = v(y) dy, 0
which obviously satisfies u(x) = M(α) for all x ≥ 1 and u(x) = −M(α) for x ≤ −1 with 1 α/2 π M(α) = K (α) 1 − |y|2 dy =
2 . 0 2α (α + 1) 1+α 2 Then, for any ϕ ∈ Cc∞ (R), we can introduce the following duality:
α u, ϕ = u(y)(α ϕ)(y) dy. R
This defines α u as a distribution, because we can check (using the Lévy–Khintchine formula (2.4)) that there exists a constant C > 0 such that |(α ϕ)(x)| ≤
C ϕ W 2,∞ (R) 1 + |x|1+α
.
If, moreover, supp ϕ ⊂ (−1, 1), it is easy to check using the properties of the function v = v(x) that
∂x (α u), ϕ = − u, α (∂x ϕ) = − u, ∂x (α ϕ) = α (∂x u), ϕ = 1, ϕ, where the last inequality is a consequence of (3.2). From the symmetry of v, we deduce the antisymmetry of u, and then (α u)(−x) = −(α u)(x). Therefore, we get the equality (α u)(x) = x in D (−1, 1), and thus by [21, Cor. 3.1.5], in the classical sense for each y ∈ (−1, 1), too. Finally, we define the nonnegative function γ −1/(α+1) 2M(α) u γ .
α (y) = y + M(α) with γ −1 = α+1 α+1 Now, for yα = γ 1/(α+1) = [2M(α)]−1/(α+1) , we can check easily that α is exactly as stated in Theorem 2.1, which ends the proof. Let us note that we will not use in the sequel the explicit form of the function α , but only its properties listed in Theorem 2.1. Remark 3.1. It has been known since the work of Head and Louat [19] (see also [18]) that
1/2 the function v(x) = K 1 − |x|2 (with a suitably chosen constant K = K (1) > 0) is the solution of the equation (1 v)(x) = 1 on (−1, 1). This result is a consequence of an inversion theorem due to Muskhelishvili, see either [28, p. 251] or [31, Sect. 4.3].
Nonlinear Diffusion of Dislocation Density
153
4. Notion of Viscosity Solutions Here, we consider Eq. (2.1) and its vanishing viscosity approximation, i.e. the following initial value problem for α ∈ (0, 2) and η ≥ 0, u t = ηu x x − |u x | α u on R × (0, +∞), u(x, 0) = u 0 (x) for x ∈ R.
(4.1) (4.2)
In this section, we present the framework of viscosity solutions to problem (4.1)– (4.2). To this end, we recall briefly the necessary material, which can be either found in the literature or is essentially a standard adaptation of those results. We also refer the reader to Crandall et al. [11] for a classical text on viscosity solutions to local (i.e. partial differential) equations. Let us first recall the definition of relaxed lower semi-continuous (lsc, for short) and upper semi-continuous (usc, for short) limits of a family of functions u ε which is locally bounded uniformly with respect to ε, lim sup ∗ u ε (x, t) = lim sup u ε (y, s) and ε→0
ε→0 y→x,s→t
lim inf ∗ u ε (x, t) = lim inf u ε (y, s). ε→0
ε→0 y→x,s→t
If the family consists of a single element, we recognize the usc envelope and the lsc envelope of a locally bounded function u, u ∗ (x, t) = lim sup u(y, s) y→x,s→t
and
u ∗ (x, t) = lim inf u(y, s). y→x,s→t
Now, we recall the definition of a viscosity solution for (4.1)–(4.2). Here, the difficulty is caused by the measure |z|−1−α dz appearing in the Lévy–Khintchine formula (2.4) which is singular at the origin and, consequently, the function has to be at least C 1,1 in space in order that α u(·, t) makes sense (especially for α close to 2). We refer the reader, for instance, to [4,25,29] for the stationary case, and to [23,24] for the evolution equation where this question is discussed in detail. Now, we are in a position to define viscosity solutions. Definition 4.1 (Viscosity solution/subsolution/supersolution). A bounded usc (resp. lsc) function u : R × R+ → R is a viscosity subsolution (resp. supersolution) of Eq. (4.1) on R × (0, +∞) if for any point (x0 , t0 ) with t0 > 0, any τ ∈ (0, t0 ), and any test function φ belonging to C 2 (R × (0, +∞)) ∩ L ∞ (R × (0, +∞)) such that (u − φ) attains a maximum (resp. minimum) at the point (x0 , t0 ) on the cylinder Q τ (x0 , t0 ) := R × (t0 − τ, t0 + τ ), we have ∂t φ(x0 , t0 ) − ηφx x (x0 , t0 ) + |φx (x0 , t0 )| (α φ(·, t0 ))(x0 ) ≤ 0 (resp. ≥ 0), where (α φ(·, t0 ))(x0 ) is given by the Lévy–Khintchine formula (2.4). We say that u is a viscosity subsolution (resp. supersolution) of problem (4.1)–(4.2) on R × [0, +∞), if it satisfies moreover at time t = 0, u(·, 0) ≤ u ∗0 (resp. u(·, 0) ≥ (u 0 )∗ ). A function u : R × R+ → R is a viscosity solution of (4.1) on R × (0, +∞) (resp. R × [0, +∞)) if u ∗ is a viscosity subsolution and u ∗ is a viscosity supersolution of the equation on R × (0, +∞) (resp. R × [0, +∞)). Other equivalent definitions are also natural, see for instance [4].
154
P. Biler, G. Karch, R. Monneau
Remark 4.2. Any bounded function u ∈ C 1+β (with some β > max{0, α − 1}) which satisfies pointwisely (using the Lévy–Khintchine formula (2.4)) Eq. (4.1) with η = 0, is indeed a viscosity solution. Theorem 4.3 (Comparison principle). Consider a bounded usc subsolution u and a bounded lsc supersolution v of (4.1)–(4.2). If u(x, 0) ≤ u 0 (x) ≤ v(x, 0) for some u 0 ∈ BU C(R), then u ≤ v on R × [0, +∞). Proof. Recall that in [23, Th. 5], the comparison principle is proved for α = 1 and η = 0 under the additional assumption that u 0 ∈ W 1,∞ (R). Looking at the proof of that result, the regularity of the initial data u 0 is only used to show that
sup (u 0 )ε (x) − (u 0 )ε (x) → 0 as ε → 0, (4.3) x∈R
where (u 0 )ε and (u 0 )ε are respectively sup and inf-convolutions. It is easy (and classical) to check that (4.3) is still true for u 0 ∈ BU C(R). The general case can be done either considering a variation of the proof of [23] taking into account the additional Laplace operator, or applying the “maximum principle” from [25], or following, for instance, the lines of [4]. We skip here the detail of this adaptation. This finishes the proof. Theorem 4.4 (Stability). Let {u ε }ε>0 be a sequence of viscosity subsolutions (resp. supersolutions) of Eq. (4.1) which are locally bounded, uniformly in ε. Then u = lim sup∗ u ε (resp. u = lim inf ∗ u ε ) is a subsolution (resp. supersolution) of (4.1) on R × (0, +∞). Proof. A counterpart of Theorem 4.4 is proved in [4, Th.1]. Here, the result for the time dependent problem is again a classical adaptation of that argument, so we skip details. Remark 4.5. One can generalize directly Theorem 4.4 assuming that {u ε }ε>0 are solutions to the sequence of Eq. (4.1) with η = ε. Then, in the limit ε → 0+ , we obtain viscosity subsolutions (resp. supersolutions) of Eq. (2.1). We use this property in the proof of Theorem 2.7. Remark 4.6. In Theorem 4.4, we only claim that the limit u is a supersolution on R × (0, +∞), but not on R × [0, +∞). In other words, we do not claim that u satisfies the initial condition. Without further properties of the initial data u 0 , it may happen that u(·, 0) ≤ u ∗0 is not true. Theorem 4.7 (Existence). Consider u 0 ∈ BU C(R). Then there exists the unique bounded continuous viscosity solution u of (4.1)–(4.2). Proof. Applying the argument of [22] (already adapted from the classical arguments), we can construct a solution by the Perron method, if we are able to construct suitable barriers. Case 1. First, assume that u 0 ∈ W 2,∞ (R). Then the following functions u ± (x, t) = u 0 (x) ± Ct
(4.4)
are barriers for C > 0 large enough (depending on the norm u 0 W 2,∞ (R) ), and we get the existence of solutions by the Perron method.
Nonlinear Diffusion of Dislocation Density
155
Case 2. Let u 0 ∈ BU C(R). For any ε > 0, we can regularize u 0 by a convolution, and get a function u ε0 ∈ W 2,∞ (R) which satisfies, moreover, |u ε0 − u 0 | ≤ ε.
(4.5) u ε0
uε
the solution of (4.1)–(4.2) with the initial condition instead of u 0 . Let us call Then, from the fact that the equation does not see the constants and from the comparison principle Theorem 4.3, we have for any ε, δ > 0, |u ε − u δ | ≤ ε + δ. Therefore, {u ε }ε>0 is the Cauchy sequence which converges in L ∞ (R × [0, +∞)) to some continuous function u (because all the functions u ε are continuous). By the stability result Theorem 4.4, we see that u is a viscosity solution of Eq. (4.1) on R × (0, +∞). To recover the initial boundary condition, we simply remark that u ε (x, 0) = u ε0 (x) satisfies (4.5), and then passing to the limit, we get u(x, 0) = u 0 (x). This shows that u is a viscosity solution of problem (4.1)–(4.2) on R × [0, +∞), and concludes the proof of Theorem 4.7. 5. Uniqueness and Stability of the Self-similar Solution Lemma 5.1 (Comparison with the self-similar solution). Let v be a subsolution (resp. a supersolution) of Eq. (2.1) with the Heaviside initial datum given in (2.8). Then we have v ∗ ≤ (u α )∗ (resp. (u α )∗ ≤ v∗ ). Proof. Using Remark 4.2 and properties of α gathered in Theorem 2.1, it is straightforward to check that the self-similar solution u α (x, t) given in (2.6) is a viscosity solution of Eqs. (4.1)–(4.2) with the initial condition (2.8). Now, we show the inequality (u α )∗ ≤ v∗ . Let v be a viscosity supersolution of (4.1)– (4.2) with the Heaviside initial datum (2.8). Given a > 0 and v a (x, t) = v(a + x, t), we have (u α )∗ (x, 0) ≤ (u 0 )∗ (x) ≤ (u 0 )∗ (a + x) ≤ v a (x, 0). Because of the translation invariance of Eq. (2.1), we see that v a is still a supersolution. Moreover, for any a > 0, we can always find an initial condition u a ∈ BU C(R) such that u α (x, 0) ≤ u a (x) ≤ v a (x, 0). Therefore, applying the comparison principle (Theorem 4.3), we deduce that u α ≤ va . Because this is true for any a > 0, we can take the limit as a → 0 and get (u α )∗ ≤ v∗ . For a subsolution v, we proceed similarly to obtain v ∗ ≤ (u α )∗ . This finishes the proof of Lemma 5.1. Proof of Theorem 2.4. We consider a viscosity solution v of Eq. (2.1) with the Heaviside initial datum (2.8). Using both inequalities of Lemma 5.1, and the fact that (u α )∗ = (u α )∗ on (R × [0, +∞))\ {(0, 0)}, we deduce the equality v = u α on (R × [0, +∞))\ {(0, 0)}, which ends the proof of Theorem 2.4. We will now prove the following weaker version of Theorem 2.5.
156
P. Biler, G. Karch, R. Monneau
Theorem 5.2 (Convergence for suitable initial data). The convergence in Theorem 2.5 holds true under the following additional assumption lim u 0 (y) = 0 ≤ u 0 (x) ≤ 1 = lim u 0 (y). y→+∞
y→−∞
(5.1)
Proof. Step 1. Limits after rescaling of the solution. Consider a solution u of (2.1)–(2.2) with an initial condition u 0 satisfying (5.1). Recall that for any λ > 0, the rescaled solution is given by u λ (x, t) = u(λx, λα+1 t). Let us define u = lim sup ∗ u λ and u = lim inf ∗ u λ . λ→+∞
λ→+∞
From the stability result Theorem 4.4, we know that u (resp. u) is a subsolution (resp. supersolution) of (2.1) on R × (0, +∞). Step 2. The initial condition. We now want to prove that u(x, 0) = u(x, 0) = H (x) for x ∈ R\ {0},
(5.2)
where H is the Heaviside function. To this end, we remark that u 0 satisfies for some γ > 0 the inequality |u 0 (x)| ≤ γ (note that γ = 1 under assumption (5.1)), and for each ε > 0, there exists M > 0 such that |u 0 (x)| < ε for x ≤ −M. In particular, we get u 0 (x) < ε + γ H (x + M), and then from the comparison principle, we deduce u(x, t) ≤ ε + (u γα )∗ (x + M, t) with u γα (x, t) = γα
x
t 1/(α+1)
and γα (y) = γ α γ −1/(α+1) y .
(5.3)
(5.4)
γ
Here α is the self-similar profile solution of (2.7) with the boundary conditions 0 and γ γ at infinity. Moreover, because u α is continuous off the origin, we can simply drop the star ∗ , when we are interested in points different from the origin. This implies x + Mλ−1 u λ (x, t) ≤ ε + γα , t 1/(α+1) and then u(x, t) ≤ ε + γα
x t 1/(α+1)
.
Therefore, for every x < 0 we have u(x, 0) ≤ ε + γα (−∞) = ε. Because this is true for every ε > 0, we get u(x, 0) ≤ 0 for every x < 0. We get the other inequalities similarly, and finally conclude that (5.2) is valid.
Nonlinear Diffusion of Dislocation Density
157
Step 3. Initial condition at the origin, using assumption (5.1). We now make use of (5.1) to identify the initial values of the limits u and u. We deduce from the comparison principle that 0 ≤ u(x, 0) ≤ u(x, 0) ≤ 1, and then for every x ∈ R we have u(x, 0) ≤ H ∗ (x) and u(x, 0) ≥ H∗ (x). Step 4. Identification of the limits after rescaling. From Lemma 5.1, we obtain u ≤ (u α )∗ = (u α )∗ ≤ u on (R × [0, +∞))\ {(0, 0)}. We have by the construction u ≤ u, hence we infer u = u = u α on (R × [0, +∞))\ {(0, 0)}. Step 5. Conclusion for the convergence. Then for any compact K ⊂ (R × [0, +∞))\ {(0, 0)}, we can easily deduce that sup |u λ (x, t) − u α (x, t)| → 0 as λ → +∞,
(x,t)∈K
which finishes the proof of Theorem 5.2.
6. Further Decay Properties and the End of the Proof of Theorem 2.5 Theorem 6.1 (Decay of a solution with compact support). Let u be the solution to (2.1)– (2.2) with the initial datum u 0 ∈ BU C(R) satisfying for some A > 0, u 0 (x) ≤ 0 for |x| ≥ A,
(6.1)
and u 0 (x) ≤ γ for some γ > 0 and all x ∈ R. Then, there exist on α, but independent of A, γ ) such that
β, β
> 0 (depending
u(x, t) ≤ Ct −β , and u(x, t) ≤ 0 for |x| ≥ C t β
with some constants C = C(α, A, γ ) and C = C (α, A, γ ). First, we need the following Lemma 6.2 (Decay after the first interaction). Consider α and yα defined in Theorem 2.1. Let ν ∈ (1/2, 1) and ξν ∈ (0, yα ) be such that α (ξν ) = ν. Let T > 0 be defined by A = ξν . γ 1/(α+1) T 1/(α+1)
(6.2)
Then, under the assumptions of Theorem 6.1, we have u(x, t) ≤ νγ for all t ≥ T, x ∈ R, and
yα u(x, t) ≤ 0 for all 0 ≤ t ≤ T and |x| ≥ A 1 + ξν
(6.3) .
(6.4)
158
P. Biler, G. Karch, R. Monneau
γ Proof. Let us denote α (y) = γ α γ −1/(α+1) y . Then we have γ H (x + A) ≥ u 0 (x) for x ∈ R, where γ H (x + A) = lim+ γα t→0
x+A t 1/(α+1)
for x + A = 0.
Now, we apply the comparison principle to deduce that x+A
γα 1/(α+1) ≥ u(x, t) for (x, t) ∈ R × (0, +∞). t Thisargument can be made rigorous by simply replacing the function γ H (x + A) by γ 1/(α+1)
α (x + A + δ)/(tε ) with δ > 0 and some sequence tε → 0+ , and then taking the limit δ → 0+ . Therefore we have x+A ≥ u(x, t) for (x, t) ∈ R × (0, +∞). γ α γ 1/(α+1) t 1/(α+1) From the properties of the support of α , we also deduce that u(x, t) ≤ 0 for x ≤ − A + yα (γ t)1/(α+1) , and then, by symmetry, u(x, t) ≤ 0 for |x| ≥ A + yα (γ t)1/(α+1) . Moreover, it follows from the monotonicity of α that A ≥ u(x, t) γ α (γ t)1/(α+1) for x ≤ 0, and by symmetry we can prove the same property for x ≥ 0. Then for T > 0 defined in (6.2) we easily deduce (6.3) and (6.4). This ends the proof of Lemma (6.2). Proof of Theorem 6.1. We apply recurrently Lemma 6.2. Define A0 = A, γ0 = γ , and yα An , γn+1 = νγn , and An+1 = An 1 + = ξν . ξν (γn Tn )1/(α+1) This gives yα n An = A0 1 + , γn = ν n γ0 , Tn = K µn , ξν with K =
1 γ0
A0 ξν
α+1
, 1<µ=
yα α+1 1 1+ , ν ξν
Nonlinear Diffusion of Dislocation Density
159
and therefore u(x, t) ≤ γn for t ≥ T0 + · · · + Tn−1 = K
µn − 1 . µ−1
In particular, we get for any n ∈ N, u(x, t) ≤ γ0 ν n for t ≥ K 0 µn with K 0 = K /(µ − 1). This implies β
u(x, t) ≤ γ0 K 0 t −β for any t > 0, x ∈ R, with β=−
log ν > 0. log µ
Similarly, we have u(x, t) ≤ 0 for |x| ≥ An if t ≤ T0 + · · · + Tn−1 = K
µn − 1 . µ−1
In particular, we get for any n ∈ N\ {0},
yα n , if t ≤ K 0 µn u(x, t) ≤ 0 for |x| ≥ A0 1 + ξν
with K 0 = K /µ. This implies
u(x, t) ≤ 0 for |x| ≥ A0 (K 0 )−β t β with β = This ends the proof of Theorem 6.1.
log 1 + log µ
yα ξν
for t ≥ 0,
> 0.
As a corollary, we can now remove assumption (5.1) in Theorem 5.2 and complete the proof of Theorem 2.5. Proof of Theorem 2.5. We simply repeat Step 3 of the proof of Theorem 5.2, but here without assuming (5.1). Then, for any ε > 0 there exists A > 0 such that u 0 (x) ≤ 1 + ε for |x| ≥ A. By Theorem 6.1 applied to the solution u(x, t) − 1 − ε, this implies that there exists a constant C > 0 (depending on ε) such that u(x, t) ≤ 1 + ε + Ct −β . Therefore, for any for λ > 0 the following inequality u λ (x, t) ≤ 1 + ε + Ct −β λ−β
160
P. Biler, G. Karch, R. Monneau
holds true, which implies that u = lim sup ∗ u λ satisfies λ→+∞
u(x, t) ≤ 1 + ε for (x, t) ∈ R × (0, +∞). Since this is true for any ε > 0, we deduce that u(x, t) ≤ 1 for (x, t) ∈ R × (0, +∞). Let us now define u˜ = min (1, u) . By the construction, ˜ u(x, t) = u(x, t) for (x, t) ∈ R × [0, +∞)\ {(0, 0)}, ˜ 0) ≤ H ∗ (x) for all x ∈ R. Therefore, u˜ is a subsolution of and, by (5.2), we have u(x, (2.1)–(2.2) on R × [0, +∞) with the initial datum being the Heaviside function. Similarly, we can show that u = lim sup ∗ u λ satisfies λ→+∞
u ≥ 0 for (x, t) ∈ R × (0, +∞).
Hence, the function u˜ = max 0, u , which is a supersolution of (2.1)–(2.2) on R × [0, +∞) with the Heaviside initial datum. Finally, the conclusion of the proof is the same as in the proof of Theorem 5.2 where u (resp. u) is replaced by u˜ (resp. u). ˜ This finishes the proof of Theorem 2.5. 7. Approximate Equation and Gradient Estimates In this section, in order to prove our gradient estimates for viscosity solutions stated in Theorem 2.7, we replace Eq. (2.1) by an approximate equation for which smooth solutions do exist. Indeed, with ε > 0, we consider the following initial value problem: u t = εu x x − |u x |α u on R × (0, +∞), u(x, 0) = u 0 (x) for x ∈ R.
(7.1) (7.2)
We have added to Eq. (2.1) an auxiliary viscosity term which is stronger than α u and u x . In the case α ∈ (0, 1], we will see later (in Sect. 8) that it is possible to pass to the limit ε → 0+ in L ∞ (R), which is the required convergence for the framework of viscosity solutions. The difficulty in the case α ∈ (1, 2) comes from the fact that, for the limit equation with ε = 0, we are not able to give a meaning to the product |u x | (α u) in the sense of distributions, while it is possible when α ∈ (0, 1]. Our results on qualitative properties of solutions to the regularized problem (7.1)– (7.2) are stated in the following two theorems. Theorem 7.1 (Approximate equation – existence of solutions). Let α ∈ (0, 1] and ε > 0. Given any initial datum u 0 ∈ C 2 (R) such that u 0,x ∈ L 1 (R) ∩ L ∞ (R), there exists a unique solution u ∈ C(R × [0, +∞)) ∩ C 2,1 (R × (0, +∞)) of (7.1)–(7.2). This solution satisfies u x ∈ C([0, T ], L p (R)) ∩ C((0, T ]; W 1, p (R)) ∩ C 1 ((0, T ], L p (R)) for every p ∈ (1, ∞) and each T > 0.
(7.3)
Nonlinear Diffusion of Dislocation Density
161
Theorem 7.2 (Approximate equation – decay estimates). Under the assumptions of Theorem 7.1, the solution u = u(x, t) of (7.1)–(7.2) satisfies u(·, t) ∞ ≤ u 0 ∞ , u x (·, t) ∞ ≤ u 0,x ∞ ,
(7.4)
and pα+1
u x (·, t) p ≤ C p,α u 0,x 1p(α+1) t
1 − α+1 1− 1p
,
(7.5)
for every p ∈ [1, ∞), all t > 0, and constants C p,α > 0, see (7.20) below), independent of ε > 0, t > 0 and u 0 . Existence theory. First, we construct solutions to the initial value problem for the regularized equation. Proof of Theorem 7.1. Note first that α u = α−1 Hu x ,
(7.6)
where H denotes the Hilbert transform, see (1.10). We recall that the Hilbert transform is bounded on the L p -space for any p ∈ (1, +∞) (see [30, Ch. 2, Th. 1]), i.e. it satisfies for any function v ∈ L p (R) the following inequality Hv p ≤ C p v p
(7.7)
with a constant C p independent of v. For α ∈ (0, 1), the operator α−1 defined analogously as in (2.3) corresponds to the convolution with the Riesz potential α−1 v = Cα | · |−α ∗ v. Hence, by [30, Ch. 5, Th. 1], for any p > 1/α with α ∈ (0, 1] and any function v ∈ L q (R), we have α−1 v p ≤ C p,α v q with
1 1 = + 1 − α. q p
(7.8)
Now, if u = u(x, t) is a solution to (7.1)–(7.2), using identity (7.6), we write the initial value problem for v = u x , vt = εvx x − (|v|α−1 Hv)x on R × (0, +∞), v(·, 0) = v0 = u 0,x ∈ L 1 (R) ∩ L ∞ (R), as well as its equivalent integral formulation t v(t) = G(εt) ∗ v0 − ∂x G(ε(t − τ )) ∗ (|v|α−1 Hv) dτ,
(7.9) (7.10)
(7.11)
0
with the Gauss–Weierstrass kernel G(x, t) = (4π t)−1/2 exp(−x 2 /(4t)). The next step is completely standard and consists in applying the Banach contraction principle to Eq. (7.11) in a ball in the Banach space XT = C([0, T ]; L 1 (R) ∩ L ∞ (R)) endowed with the usual norm v T = supt∈[0,T ] ( v(t) 1 + v(t) ∞ ). Using well known estimates of the heat semigroup and inequalities (7.7)–(7.8) combined with the imbedding L 1 (R) ∩ L ∞ (R) ⊂ L p (R) for each p ∈ [1, ∞], we obtain a solution v = v(x, t)
162
P. Biler, G. Karch, R. Monneau
to Eq. (7.11) in the space XT provided T > 0 is sufficiently small. We refer the reader to, e.g., [2,5] for examples of such a reasoning. This solution satisfies (7.3) for every p ∈ (1, ∞) and each T > 0, by standard regularity estimates of solutions to parabolic equations. Moreover, following the reasoning in [2], one can show that the solution is regular. Finally, this local-in-time solution can be extended to global-in-time (i.e. for all T > 0) because of the estimates v(t) p ≤ v0 p for every p ∈ [1, ∞] being the immediate consequence of inequalities (7.17), (7.18), and (7.21) below. Gradient estimates. In the proof of the decay estimates of u x , we shall require several properties of the operator α . First, we recall the Nash inequality for the operator α . Lemma 7.3 (Nash inequality). Let α > 0. There exists a constant C N > 0 such that 2(1+α)
w 2
≤ C N α/2 w 22 w 2α 1
(7.12)
for all functions w satisfying w ∈ L 1 (R) and α/2 w ∈ L 2 (R). The proof of inequality (7.12) is given, e.g., in [26, Lemma 2.2]. Our next tool is the, so-called, Stroock–Varopoulos inequality. Lemma 7.4 (Stroock–Varopoulos inequality). Let 0 ≤ α ≤ 2. For every p > 1, we have p 2 α 4( p − 1) α p−2 2 |w| 2 ( w)|w| w dx ≥ dx (7.13) p2 R R for all w ∈ L p (R) such that α w ∈ L p (R). If α w ∈ L 1 (R), we obtain (α w) sgn w dx ≥ 0.
(7.14)
Moreover, if w, α w ∈ L 2 (R), it follows that (α w)w + dx ≥ 0 and (α w)w − dx ≤ 0,
(7.15)
R
R
R
where w + = max{0, w} and w − = max{0, −w}. Inequality (7.13) is well known in the theory of sub-Markovian operators and its statement and the proof is given, e.g., in [27, Th. 2.1, combined with the Beurling–Deny condition (1.7)]. Inequality (7.14), called the (generalized) Kato inequality, is used, e.g., in [13] to construct entropy solutions of conservation laws with a Lévy diffusion. It can be easily deduced from [13, Lemma 1] by an approximation argument. The proof of (7.15) can be found, for example, in [27, Prop. 1.6]. Remark 7.5. Remark that inequality (7.14) appears to be a limit case of (7.13) for p = 1. Inequality (7.15) for w + follows easily from (7.14) by a comparison argument if, for instance, w ∈ Cc∞ (R). Finally, remark that the constant appearing in (7.13) is the same as for the Laplace operator ∂ 2 /∂ x 2 = −2 . Our proof of the decay of v(t) = u x (t) is based on the following Gagliardo–Nirenberg type inequality:
Nonlinear Diffusion of Dislocation Density
163
Lemma 7.6 (Gagliardo–Nirenberg type inequality). Assume that p ∈ (1, ∞) and α > 0 are fixed and arbitrary. For all v ∈ L 1 (R) such that α/2 |v|( p+1)/2 ∈ L 2 (R), the following inequality is valid: 2 v ap ≤ C N α/2 |v|( p+1)/2 v b1 , (7.16) 2
where a=
p( p + α) pα + 1 , b= , p−1 p−1
and C N is the constant from the Nash inequality (7.12). Proof. Without loss of generality, we can assume that v 1 = 0. Substituting w = |v|( p+1)/2 in the Nash inequality (7.12) we obtain 2 ( p+1)(1+α) α( p+1) ≤ C N α/2 |v|( p+1)/2 v ( p+1)/2 . v p+1 2
Next, it suffices to apply two particular cases of the Hölder inequality
v p 1/ p 2 v 1
p2 /( p2 −1) p/( p+1)
≤ v p+1 as well as v ( p+1)/2 ≤ v p
1/( p+1)
v 1
,
and compute carefully all the exponents which appear on both sides of the resulting inequality. Proof of Theorem 7.2. The first inequality in (7.4) is an immediate consequence of the comparison principle from Theorem 4.3, because classical solutions are viscosity solutions, as well. The maximum principle and an argument based on inequalities (7.15) (cf. [26] for more detail) lead to the second inequality in (7.4). We also discuss this inequality in Remark 7.7 below. For the proof of the L 1 -estimate u x (t) 1 ≤ u 0,x 1
(7.17)
(i.e. (7.5) with p = 1 and C p,α = 1), we multiply Eq. (7.9) by sgn v = sgn u x and we integrate with respect to x to obtain d (α−1 Hv)|v| sgn v dx. |v| dx = ε vx x sgn v dx − x dt R R R The first term on the right-hand side is nonpositive by the Kato inequality (i.e. (7.14) with α = 2) hence we skip it. Remark that (formally) α−1 ( Hv)|v| sgn v dx = (α−1 Hv)vx (sgn v)2 + (α−1 Hvx )v dx x R R = (α−1 Hv)v dx = 0. R
x
√ Now, approximating the sign function in a standard way by sgnδ (z) = z/ z 2 + δ, integrating by parts, and passing to the limit δ → 0+ , one can show rigorously that the
164
P. Biler, G. Karch, R. Monneau
second term on right-hand side of the above inequality is nonpositive. This completes the proof of (7.17) with p = 1. Next, we multiply the equation in (7.9) by |v| p−2 v with p > 1 to get 1 d p p−2 (α−1 Hv)|v| |v| p−2 v dx. |v| dx = ε vx x |v| v dx − x p dt R R R We drop the first term on the right-hand side, because it is nonpositive by (7.13) with α = 2. Integrating by parts and using the elementary identity p − 1 p−1 |v| |v| p−2 v = |v| v , x x p we transform the second quantity on the right-hand side as follows: p−1 α−1 p−2 − ( Hv)|v| |v| v dx = − (α v)|v| p−1 v dx. x p R R Consequently, by the Stroock–Varopoulos inequality (7.13) (with the exponent p replaced by p + 1), we obtain d 4 p( p − 1) α/2 ( p+1)/2 2 p |v| (7.18) v(t) p ≤ − . 2 dt ( p + 1)2 Hence, the interpolation inequality (7.16) combined with (7.17) lead to the following p inequality for v(t) p : 4 p( p − 1) d p ( pα+1)/( p−1) −1 p ( p+α)/( p−1) v(t) p ≤ − C N v0 1 v(t) p . (7.19) 2 dt ( p + 1) Recall now that if a nonnegative (sufficiently smooth function) f = f (t) satisfies, for all t > 0, the inequality f (t) ≤ −K f (t)β with constants K > 0 and β > 1, then f (t) ≤ (K (β − 1)t)−1/(β−1) . Applying this simple result to the differential inequality (7.19), we complete the proof of the L p -decay estimate (7.5) with the constant C p,α
1 4 p(α + 1) − α+1 = C N−1 ( p + 1)2
1− 1p
where C N is the constant from the Nash inequality (7.12).
,
(7.20)
Remark 7.7. Note that, for every fixed α, we have lim p→∞ C p,α = +∞. By this reason, we are not allowed to pass directly to the limit p → +∞ in inequalities (7.5) (as was done in, e.g., [26, Th. 2.3]) in order to obtain a decay estimate of v(t) in the L ∞ -norm. Nevertheless, using (7.19) we immediately deduce the inequality v(·, t) p ≤ v0 (·) p valid for every p ∈ (1, ∞). Hence, passing to the limit p → +∞ we get v(·, t) ∞ ≤ v0 (·) ∞ .
(7.21)
It is natural to expect that, under the assumptions of Theorem 7.2, the quantity v(·, t) ∞ should decay at the rate t −1/(α+1) . For a proof, one might follow an idea from [6, Lemma 4.7] where the decay estimates of vx (·, t) were obtained for solutions of a certain regularization of Eq. (1.15). Here, however, we did not try to go this way, because our main goal was to study decay estimates for the problem (2.1)–(2.2) whose viscosity solutions are not regular enough a priori to handle decay properties of u x x (x, t) = vx (x, t).
Nonlinear Diffusion of Dislocation Density
165
8. Passage to the Limit and the Proof of Theorem 2.7 Now, we are in a position to complete the proof of the gradient estimates (2.11). First, we show that from the sequence {u ε }ε>0 of solutions to the approximate problem (7.1)–(7.2) one can extract, via the Ascoli–Arzelà theorem, a subsequence converging uniformly. Theorem 4.4 on the stability and Remark 4.5 imply that the limit function is a viscosity solution to (2.1)–(2.2). Passing to the limit ε → 0+ in inequalities (7.4) and (7.5) we complete our reasoning. Proof of Theorem 2.7. First, let us suppose that u 0 ∈ C ∞ (R) ∩ W 2,∞ (R) with u 0,x ∈ L 1 (R)∩ L ∞ (R). Denote by u ε = u ε (x, t) the corresponding solution to the approximate problem with ε > 0. Step 1. Modulus of continuity in space. Under this additional assumption, we have u εx (·, t) p ≤ C p t −γ p
(8.1)
pα+1 1 1 − 1p . The Sobolev imbedding theorem with C p = C p,α u 0,x 1p(α+1) and γ p = α+1 implies that there exist some β ∈ (0, 1) and C0 > 0 such that
|u ε (x + h, t) − u ε (x, t)| ≤ |h|β C0 C p t −γ p .
(8.2)
∞ Step 2. Modulus of continuity in time. Let us consider a nonnegative function ϕ ∈ C (R) with supp ϕ ⊂ [−1, 1] such that R ϕ(x) dx = 1, and for any δ > 0 set ϕδ (x) = δ −1 ϕ(δ −1 x). Then, multiplying (7.1) by ϕδ and integrating in space, we get d ε u (x, t)ϕδ (x) dx = ε u ε (·, t), (ϕδ )x x dt R − ϕδ (x) |u εx (x, t)|(H α−1 u εx (x, t)) dx,
R
+ 1/ p
= 1, and then with 1/ p d ε ≤ ε u ε (·, t) ∞ (ϕδ )x x 1 u (x, t)ϕ (x) dx δ dt R
+ ϕδ ∞ u εx (·, t) p H α−1 u εx (·, t) p .
(8.3)
Here, we have used relation (7.6). Combining inequalities (7.7) and (7.8) with estimate (8.1), we get for p > 1/α, H α−1 u εx (·, t) p ≤ C p C p ,α Cq t −γq . Then for any bounded time interval I ⊂ (0, +∞) there exists a constant C I,δ such that for all t ∈ I , we have for any ε ∈ (0, 1], d ε ≤ C I,δ . u (x, t)ϕ (x) dx δ dt R Now, for any t, t + s ∈ I , we get ε u ε (x, t + s)ϕδ (x) dx − u (x, t)ϕδ (x) dx ≤ |s|C I,δ . R
R
166
P. Biler, G. Karch, R. Monneau
Therefore, the following estimate: |u ε (0, t + s) − u ε (0, t)| ϕδ (x) dx ≤ |s|C I,δ + R
× sup |u ε (x, t + s) − u ε (0, t + s)| + |u ε (x, t) − u ε (0, t)| x∈[−δ,δ]
holds true. Using the Hölder inequality (8.2), we deduce that there exists a constant C I depending on I , but independent of δ and of ε ∈ (0, 1], such that |u ε (0, t + s) − u ε (0, t)| ≤ |s|C I,δ + C I δ β . Since the above inequality is true for any δ, this shows the existence of a modulus of continuity ω I satisfying |u ε (0, t + s) − u ε (0, t)| ≤ ω I (|s|) for any t, t + s ∈ I. By the translation invariance of the problem, this estimate is indeed true for any x ∈ R, i.e. |u ε (x, t + s) − u ε (x, t)| ≤ ω I (|s|) for any t, t + s ∈ I, x ∈ R.
(8.4)
From estimates (8.2) and (8.4), and using the Step 3. Convergence as ε → Ascoli–Arzelà theorem and the Cantor diagonal argument, we deduce that there exists a subsequence (still denoted {u ε }ε ) which converges to a limit u ∈ C(R × (0, +∞)). By the stability result in Theorem 4.4 (see also Remark 4.5), we have that u is a viscosity solution of (2.1) on R × (0, +∞). 0+ .
Step 4. Checking the initial conditions for u 0 smooth. Remark that for u 0 ∈ W 2,∞ we can use the barriers given in (4.4) with some constant C > 0 uniform in ε ∈ (0, 1]. This ensures that u is continuous up to t = 0 and satisfies u(·, 0) = u 0 , so this proves the result under additional assumptions. Step 5. General case. The proof in the case of less regular initial conditions simply follows by an approximation argument as in the proof of Theorem 4.7. Step 6. Gradient estimates. To pass to the limit ε → 0+ in estimates (7.5), we use the inequality h −1 u ε (· + h, t) − u ε (·, t) p ≤ u εx (·, t) p
(8.5)
with fixed h > 0. Hence, by the Fatou lemma combined with the pointwise convergence of u ε toward u, we deduce from (8.5) and (7.5) that pα+1
h −1 u(· + h, t) − u(·, t) p ≤ C p,α u 0,x 1p(α+1) t
1 − α+1 1− 1p
for all h > 0. For every fixed t > 0, the sequence {h −1 (u(·+h, t)−u(·, t))}h>0 is bounded in L p (R) and converges (up to a subsequence) weakly in L p (R) toward u x (·, t) (see, e.g., [30, Ch. V, Prop. 3]). Using the well known property of a weak convergence in Banach spaces we conclude h u x (·, t) p ≤ lim inf + h→0
−1
pα+1 p(α+1)
u(· + h, t) − u(·, t) p ≤ C p,α u 0,x 1
This finishes the proof of Theorem 2.7.
t
1 − α+1 1− 1p
.
Nonlinear Diffusion of Dislocation Density
167
Acknowledgements. The third author is indebted to Cyril Imbert for stimulating and enlightening discussions on the subject of this paper. The authors thank the anonymous referee for calling their attention to the onedimensional counterpart of the quasi-geostrophic equation (1.15). This work was supported by the contract ANR MICA (2006–2009), by the European Commission Marie Curie Host Fellowship for the Transfer of Knowledge “Harmonic Analysis, Nonlinear Analysis and Probability” MTKD-CT-2004-013389, and by the Polish Ministry of Science grant N201 022 32/0902.
References 1. Alvarez, O., Hoch, P., Le Bouar, Y., Monneau, R.: Dislocation dynamics: short time existence and uniqueness of the solution. Arch. Rat. Mech. Anal. 181, 449–504 (2006) 2. Amour, L., Ben-Artzi, M.: Global existence and decay for viscous Hamilton-Jacobi equations. Nonlinear Anal. 31, 621–628 (1998) 3. Barles, G., Chasseigne, E., Imbert, C.: On the Dirichlet problem for second-order elliptic integro-differential equations. Indiana Univ. Math. J. 57, 213–246 (2008) 4. Barles, G., Imbert, C.: Second-order elliptic integro-differential equations: viscosity solutions’ theory revisited. Ann. I.H.P., Anal Non-Lin. 25, 567–585 (2008) 5. Ben-Artzi, M., Souplet, Ph., Weissler, F.B.: The local theory for viscous Hamilton-Jacobi equations in Lebesgue spaces. J. Math. Pures Appl. 81, 343–378 (2002) 6. Castro, A., Córdoba, D.: Global existence, singularities and ill-posedness for a nonlocal flux. Adv. Math. 219, 1916–1936 (2008) 7. Chae, D., Córdoba, A., Córdoba, D., Fontelos, M.A.: Finite time singularities in a 1D model of the quasi-geostrophic equation. Adv. Math. 194, 203–223 (2005) 8. Constantin, P., Lax, P., Majda, A.: A simple one-dimensional model for the three dimensional vorticity. Comm. Pure Appl. Math. 38, 715–724 (1985) 9. Constantin, P., Majda, A., Tabak, E.: Formation of strong fronts in the 2-D quasi-geostrophic thermal active scalar. Nonlinearity 7, 1495–1533 (1994) 10. Córdoba, A., Córdoba, D., Fontelos, M.A.: Formation of singularities for a transport equation with nonlocal velocity. Ann. Math. 162, 1377–1389 (2005) 11. Crandall, M.G., Ishii, H., Lions, P.-L.: User’s guide to viscosity solutions of second order partial differential equations. Bull. Amer. Math. Soc. (N.S.), 27, 1–67 (1992) 12. Deslippe, J., Tedstrom, R., Daw, M.S., Chrzan, D., Neeraj, T., Mills, M.: Dynamics scaling in a simple one-dimensional model of dislocation activity. Phil. Mag. 84, 2445–2454 (2004) 13. Droniou, J., Imbert, C.: Fractal first order partial differential equations. Arch. Rat. Mech. Anal. 182, 299–331 (2006) 14. Forcadel, N., Imbert, C., Monneau, R.: Homogenization of the dislocation dynamics and of some particle systems with two-body interactions. Disc. Contin. Dyn. Syst. Ser. A 23, 785–826 (2009) 15. Getoor, R.K.: First passage times for symmetric stable processes in space. Trans. Amer. Math. Soc. 101, 75–90 (1961) 16. Head, A.K.: Dislocation group dynamics I. Similarity solutions od the n-body problem. Phil. Mag. 26, 43–53 (1972) 17. Head, A.K.: Dislocation group dynamics II. General solutions of the n-body problem. Phil. Mag. 26, 55–63 (1972) 18. Head, A.K.: Dislocation group dynamics III. Similarity solutions of the continuum approximation. Phil. Mag. 26, 65–72 (1972) 19. Head, A.K., Louat, N.: The distribution of dislocations in linear arrays. Austral. J. Phys. 8, 1–7 (1955) 20. Hirth, J.R., Lothe, L.: Theory of Dislocations. Second Ed., Malabar, FL: Krieger, 1992 21. Hörmander,: The Analysis of Linear Partial Differential Operators. Vol. 1, New York: Springer-Verlag, 1990 22. Imbert, C.: A non-local regularization of first order Hamilton-Jacobi equations. J. Differ. Eq. 211, 214–246 (2005) 23. Imbert, C., Monneau, R., Rouy, E.: Homogenization of first order equations, with (u/ε)-periodic Hamiltonians. Part II: application to dislocations dynamics. Comm. Part. Diff. Eq. 33, 479–516 (2008) 24. Jakobsen, E.R., Karlsen, K.H.: Continuous dependence estimates for viscosity solutions of integro-PDEs. J. Differ. Eq. 212, 278–318 (2005) 25. Jakobsen, E.R., Karlsen, K.H.: A maximum principle for semicontinuous functions applicable to integro-partial differential equations. NoDEA Nonlin. Differ. Eqs. Appl. 13, 137–165 (2006) 26. Karch, G., Miao, C., Xu, X.: On the convergence of solutions of fractal Burgers equation toward rarefaction waves. SIAM J. Math. Anal. 39, 1536–1549 (2008)
168
P. Biler, G. Karch, R. Monneau
27. Liskevich, V.A., Semenov, Yu.A.: Some problems on Markov semigroups. In: Schrödinger Operators, Markov Semigroups, Wavelet Analysis, Operator Algebras, Math. Top. 11, Berlin: Akademie Verlag, 1996, pp. 163–217 28. Muskhelishvili, N.I.: Singular Integral Equations. Groningen: P. Noordhoff, N. V., 1953 29. Sayah, A.: Équations d’Hamilton-Jacobi du premier ordre avec termes intégro-différentiels. I Unicité des solutions de viscosité, II Existence de solutions de viscosité. Comm. Part. Diff. Eq. 16, 1057–1093 (1991) 30. Stein, E.M.: Singular Integrals and Differentiability Properties of Functions. Princeton Mathematical Series 30, Princeton, NJ: Princeton University Press, 1970 31. Tricomi, F.G.: Integral Equations. New York-London: Interscience Publ., 1957 32. Vázquez, J.L.: Smoothing and Decay Estimates for Nonlinear Diffusion Equations. Equations of Porous Medium Type. Oxford Lecture Series in Mathematics and its Applications 33, Oxford: Oxford University Press, 2006 Communicated by P. Constantin
Commun. Math. Phys. 294, 169–197 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0935-9
Communications in
Mathematical Physics
On the Massive Wave Equation on Slowly Rotating Kerr-AdS Spacetimes Gustav Holzegel Department of Mathematics, Fine Hall, Princeton University, Washington Road, Princeton, NJ 08544, United States. E-mail:
[email protected] Received: 27 February 2009 / Accepted: 29 July 2009 Published online: 2 October 2009 – © Springer-Verlag 2009
Abstract: The massive wave equation g ψ − α 3 ψ = 0 is studied on a fixed Kerr-anti de Sitter background M, g M,a, . We first prove that in the Schwarzschild case (a = 0), ψ remains uniformly bounded on the black hole exterior provided that α < 49 , i.e. the Breitenlohner-Freedman bound holds. Our proof is based on vectorfield multipliers and commutators: The total flux of the usual energy current arising from the timelike Killing vector field T (which fails to be non-negative pointwise) is shown to be non-negative with the help of a Hardy inequality after integration over a spacelike slice. In addition to T , we construct a vectorfield whose energy identity captures the redshift producing good estimates close to the horizon. The argument is finally generalized to slowly rotating Kerr-AdS backgrounds. This is achieved by replacing the Killing vectorfield T = ∂t with K = ∂t + λ∂φ for an appropriate λ ∼ a, which is also Killing and–in contrast to the asymptotically flat case–everywhere causal on the black hole exterior. The separability properties of the wave equation on Kerr-AdS are not used. As a consequence, the theorem also applies to spacetimes sufficiently close to the Kerr-AdS spacetime, as long as they admit a causal Killing field K which is null on the horizon.
1. Introduction The study of linear wave equations on black hole spacetimes has acquired a prominent role within the subject of general relativity. The main reason is the expectation that understanding the mechanisms responsible for the decay of linear waves on black hole exteriors in a sufficiently robust setting provides important insights for the non-linear black hole stability problem [7]. The mathematical analysis of linear waves in this context was initiated by the pioneering work of Kay and Wald establishing boundedness (up to and including the event horizon) for φ satisfying g φ = 0 on Schwarzschild spacetimes [14,20]. Since then considerable progress has been achieved, especially in the last few years. Most of these
170
G. Holzegel
recent decay and boundedness theorems for linear waves concern black hole spacetimes satisfying the vacuum Einstein equations Rµν −
1 Rgµν + gµν = 0 2
(1)
with = 0 and, motivated by cosmological considerations, > 0. In particular, by now polynomial decay rates have been established for g φ = 0 on Schwarzschild [4,8,16] and more recently, Kerr spacetimes [6,7,19]. In the course of work on the decay problem, a much more robust understanding of boundedness on both Schwarzschild and Kerr spacetimes was also obtained [6,7] allowing one to prove boundedness for a large class of spacetimes, which are not exactly Schwarzschild or Kerr but only assumed to be sufficiently close Many of the above results have been extended to the cosmological case, > 0. Here (much stronger) decay rates have been established for the wave equation on Schwarzschild-de Sitter spacetimes [5,13,17]. For further discussion we refer the reader to the lecture notes [7], which among other things provide an account of previous work on these problems (Sect. 4.4, 5.5, 6.3 ibidem) and a comparison with results that have been obtained in the heuristic tradition (Sect. 4.6). In contrast to the case of a positive cosmological constant, the choice < 0 in (1) has remained relatively unexplored. While this problem certainly deserves mathematical attention in its own right, there is also considerable interest from high energy physics, see [9,18]. In this paper, we study the equation g ψ − α
ψ =0 3
(2)
on a class of spacetimes, which will include slowly rotating Kerr anti-de Sitter spacetimes [3]. These spacetimes generalize the well-known Kerr solution (the latter being the unique two-parameter family of stationary, axisymmetric asymptotically flat black hole solutions to (1) with = 0.). They are axisymmetric, stationary solutions of (1) with < 0, parametrized by their mass M and angular momentum per unit mass a = MJ . Before we comment further on their geometry, let us discuss Eq. (2). The main motivation to include the zeroth order term in (2) is the case α = 2, the conformally invariant case. In pure AdS the Green’s function for (2) with α = 2 is supported purely on the light cone, which makes it a natural analogue of the equation ψ = 0 in asymptotically flat space (and also explains the adjective “massless” which is sometimes used in the physics literature in connection with this choice of α). The case α = 2 also occurs naturally in classical general relativity when studying a Maxwell field or linear gravitational perturbations in AdS [12]. Again for the case of pure AdS, it is well known ([1,2,12]) that (2) is only well-posed for α < 54 , the so-called second Breitenlohner-Freedman bound. While no solutions exist for α ≥ 49 , one has an infinite number of solutions depending on boundary conditions for α in the range 45 ≤ α < 49 , the latter bound being the first Breitenlohner-Freedman bound. For general asymptotically AdS spacetimes, however, the wellposedness of (2) has not yet been established explicitly. In this context it is essential to notice that asymptotically AdS spaces are not globally hyperbolic. To make the dynamics of (2) well-posed suitable boundary conditions will have to be imposed on the timelike boundary of the spacetime. We will address this boundary initial value problem in detail in a separate
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
171
paper. For the purpose of the present paper, we will assume that α < 94 and that we are given a solution to (2) which decays suitably near the AdS boundary. This pointwise “radial” decay depends on how close α is to the Breitenlohner Freedman bound and ensures in particular that there is no energy flux through the timelike boundary I. It is precisely the decay suggested by the mode analysis of [2] in pure AdS and expected to hold for all asymptotically AdS spacetimes. The global geometry of the Kerr-AdS background spacetimes is quite different from their asymptotically flat counterparts. Most notably perhaps, null-infinity is now timelike, entailing the non-globally hyperbolic nature of the spacetime mentioned above. With the boundary conditions imposed there will be no radiation flux through infinity and the only possible decay mechanism is provided by an energy flux through the horizon. Another geometric feature, which we will exploit to a great extent in the present paper, is the existence of an everywhere causal Killing vectorfield on the black hole exterior of slowly rotating Kerr-AdS spacetimes.1 This vectorfield was used previously by Hawking and Reall [10] to obtain a positive conserved energy for any matter fields satisfying the dominant energy condition. The scalar field (2) in the subcase α ≤ 0 provides an example. In particular, the argument of [10] excluded a negative energy flux through the horizon and hence superradiance as a mechanism of instability, at least if 2 with r |a| − 3 < rhoz hoz being the location of the event horizon.
For 0 < α < 49 , however, the energy momentum tensor associated with ψ does not satisfy the dominant energy condition and the energy current associated with the causal Killing field fails to be positive pointwise. In particular, as noted for instance in [15], the positivity argument of [10] breaks down for 0 < α < 94 , including the most interesting case α = 2. In this paper we present a simple resolution of this problem: Using a Hardy inequality we show that the energy flux arising from the Killing field is still positive in an integrated sense. With the existence of the everywhere causal Killing field and its associated globally positive energy, superradiance is eliminated as an obstacle to stability.2 This makes the problem much easier to deal with than the asymptotically flat case, where such a vectorfield is not available. In particular, we can avoid the intricate bootstrap and harmonic analysis techniques of [6], which were necessary to deal with superradiant phenomena and trapping (see also [19]).3 It is well-known that the notion of positive energy outlined above is not sufficient to prevent the scalar field from blowing up on the horizon in evolution. This issue was first addressed and resolved for Schwarzschild in the celebrated work of Kay and Wald [14,20] exploiting the special symmetry properties of the background spacetime. With the recent work of Dafermos and Rodnianski, in particular their mathematical understanding of the celebrated redshift, there is now a much more stable argument available, which does not hinge on the discrete symmetries of Schwarzschild and is in fact applicable to any black hole event horizon with positive surface gravity [7]. This geometric understanding of the role of the event horizon for the boundedness and decay mechanism 1 This is crucially different from the asymptotically flat case, where any linear combination of the two available Killing fields K = ∂t + λ∂φ (for some constant λ) is somewhere spacelike. 2 As expected perhaps, the restriction on a (which can be computed explicitly) becomes tighter compared to the case α ≤ 0 of Hawking and Reall. 3 Superradiance is induced by the existence of an ergosphere [i.e. an effect which is not present in Schwarzschild] arising from the fact that the energy density associated with the Killing field ∂t can be negative inside the ergosphere. This allows for a negative energy flux through the horizon and hence an amplification of the amplitude for backscattered waves. See [21] for a nice discussion and also [7] for a detailed mathematical treatment.
172
G. Holzegel
was first developed in [8] and plays a crucial role in the recent proof of boundedness [6] and polynomial decay [7] of scalar waves on slowly rotating Kerr spacetimes. The boundedness statement of [6] holds for a much more general class of spacetimes nearby Kerr, while the decay statement of [7] requires the background to be exactly Kerr. For our considerations at the horizon we will adapt the ideas developed by Dafermos and Rodnianski. A vectorfield Y is constructed whose energy identity, if coupled to the timelike Killing vectorfield K in an appropriate way, provides control over Sobolev norms whose weights do not degenerate at the horizon. An additional complication compared to the asymptotically-flat massless case is that due to the zeroth order term in the wave equation (2), a zeroth order flux-term of the redshift vectorfield on the horizon has the wrong sign. However, this term can, after a computation, be absorbed by the “good” positive terms at our disposal. With Y and K at hand, boundedness is shown adapting the argument of [6], taking care of the different weights in r which appear due to the asymptotically hyperbolic nature of the background space.4 It should be emphasized that the proof does not require the construction of a globally positive spacetime integral arising from a virial vectorfield but that it suffices to use the timelike Killing field and the redshift vectorfield alone. The fact that the more elementary statement of boundedness can be obtained from the use of these two vectorfields alone in the non-superradiant regime was observed by Dafermos and Rodnianski in [6] in the asymptotically-flat case. Previously, even the understanding of boundedness was intriniscially tied to the understanding of a globally positive spacetime term and hence decay, see [8] and also Sect. 3.4 of [7]. Finally, here is an outline of the paper. We introduce the AdS-Schwarzschild and Kerr backgrounds equipped with regular coordinate systems on the black hole exterior (defining in particular the spacelike slices we are going to work with) in Sect. 2. In the following section the class of solutions we wish to consider is defined and the notion of vectorfield multipliers discussed. For reasons of presentation we first state and prove the boundedness theorem for Schwarzschild-AdS in Sect. 4, before we turn to the generalization to Kerr-AdS in Sect. 5. The paper concludes with some final remarks and future directions. 2. The Black Hole Backgrounds 2.1. Schwarzschild-AdS. In the familiar (t, r ) coordinates, the Schwarzschild AdS metric reads −1 2M r 2 2M r 2 g =− 1− + 2 dt 2 + 1 − + 2 dr 2 + r 2 dω2 , r l r l
(3)
where = − l32 is the cosmological constant. This coordinate system is not well-behaved at the zeros of 2M r 2 + 2 . 1−µ= 1− r l
(4)
4 Since angular momentum operators do not commute with the wave operator for a = 0, we have to commute the wave equation with T and the redshift vectorfield Y to obtain L 2 control of certain derivatives,
which leads to control over all derivatives using elliptic estimates on the asymptotically hyperbolic spacelike slices .
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
173
Let us define ⎛
p = ⎝ Ml 2 +
⎛
⎞1 3 6 l 2 4 ⎠ M l + 27
and q = ⎝ Ml 2 −
⎞1 3 6 l 2 4 ⎠ M l + . 27
(5)
2
Clearly, p > 0, q < 0 with pq = − l3 . The expression (4) has a single real root at rhoz = p + q > 0,
(6)
the location of the black hole event horizon. Note that for l → ∞ we have rhoz → 2M. In general we have the estimate 2 − 3 pq (r − rhoz ) r 2 + rrhoz + rhoz r 3 + l 2 r − 2Ml 2 = 1−µ = 2 rl 2 3 rl 2 3 2 2 r − rhoz (r − rhoz ) r + rrhoz + rhoz + l = ≥ , (7) rl 2 rl 2 which will be useful later. Here is a coordinate system which is well behaved everywhere on the black hole exterior and the horizon: It arises from the coordinate transformation r t = t + r (r ) − l arctan , (8) l where r (r ) is a solution of the differential equation 1 dr
= 2M dr 1− r +
r2 l2
=
1 1−µ
and r (3M) = 0.
(9)
The variable r is often called the tortoise coordinate. In the new (t , r ) coordinates the metric reads 2
r 1 + 2M 2 4M r + l2
g = − (1 − µ) dt + dr + dr 2 + r 2 dω2 , dt 2 2 r2 r 1 + rl 2 1 + l2
(10)
which is clearly regular on the horizon. For notational convenience let us agree on the shorthand notation k± = 1 ±
2M r 2 + 2 r l
and
k0 = 1 +
r2 . l2
(11)
Slices τ of constant t will play a prominent role for stating energy identities in the paper. Their normal 2
r 1 + 2M 2M r + l2 ∂r ∂ − − ∇t = 2 t 2 2 r 1 + rl 2 1 + rl 2
(12)
174
G. Holzegel
is everywhere timelike: k+
g ∇t , ∇t = g t t = − 2 . k0
(13)
We denote the unit-normal by
√ k+ ∇t
2M n = − √ ∂t − √ ∂r . =
k0 −g (∇t , ∇t ) r k+
(14)
We also define the (Killing) vectorfields i (i = 1, 2, 3) to be a basis of generators of the Lie-algebra of S O (3) corresponding to the spherical symmetry of the Schwarzschild / to denote the gradient of the metric induced on the metric. Moreover, we will write ∇ S O (3)-orbits. Finally, a Penrose diagram of (the black hole exterior of) Schwarzschild-AdS spacetime with two slices of constant t is depicted below.
H+ τ I
0
2.2. Kerr-AdS. The Kerr-AdS metric in Boyer Lindquist coordinates reads 2
θ r 2 + a 2 − − a 2 sin2 θ 2 2 sin2 θ d φ˜ 2 dr + dθ + g=
−
θ 2
θ r 2 + a 2 − −
− θ a 2 sin2 θ 2 ˜ − − a sin2 θ d φdt dt −2 with the identifications = r 2 + a 2 cos2 θ, r2
± = r 2 + a 2 1 + 2 ± 2Mr, l a2 cos2 θ, l2 a2 = 1− 2. l
θ = 1 −
(15)
(16) (17) (18) (19)
2
Once again, let k0 = 1 + rl 2 . A coordinate system which is regular on the horizon is obtained by the transformations t = t + A (r )
and
φ = φ˜ + B (r ) ,
(20)
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
175
where 2Mr dA = dr
− 1 +
r2 l2
a dB = . dr
−
and
(21)
The new metric coefficients become , gt t = gtt ,
θ = gφ˜ φ˜ , gt φ = gt φ˜ , 1 2 2
, = − a sin θ k + + 0 2 l2 r2 1 + l2
gθθ =
(22)
gφφ
(23)
grr
a sin2 θ [ + 2Mr ] , k0 a2 1 2Mr − 2 sin2 θ . = k0 l
(24)
gφr = −
(25)
gt r
(26)
Note that the angular momentum term in the last expression grows faster in r than the mass term.5 For a = 0 the metric reduces to the Schwarzschild-AdS metric in t , r coordinates, cf. (10). The inverse components are
gt
t
=−
+ + a 2 sin2 θ −k0 + k02 θ
2Mr l2
,
θ 2 a
, g φφ = , , gt φ = k 0 l 2 θ
θ sin2 θ
− a 2Mr
, g φr = , gt r = . = k0
(27)
g θθ =
(28)
grr
(29)
The unit normal of a constant t slice is n =
−g t
t
gt r gt φ
∂t − ∂ − ∂ , g(n , n ) = −1, r
φ −g t t −g t t
(30)
with the determinant of the metric induced on constant t slices being √
det h =
det gt =const
= sin θ
t
t . −g
(31)
5 Hence the Kerr-AdS metric is not uniformly close to the Schwarzschild metric in these coordinates! It follows that for the statement that “Schwarzschild-AdS is close to Kerr-AdS” one has to use both a regular coordinate patch for r ≤ R and a Boyer Lindquist patch for r > R (for some R away from the horizon). For the asymptotically-flat case this is not necessary.
176
G. Holzegel
3. The Dynamics 3.1. The class of solutions. Having defined a proper coordinate system on the black hole backgrounds in the previous section we can turn to the wave equation we would like to study (recall = − l32 ): g ψ +
α ψ = 0, l2
(32)
for α ∈ R. The choice α = 2 in (32) corresponds to the conformally invariant case, cf. the discussion in the Introduction. Generally we will assume α < 49 , the upper Breitenlohner-Freedman bound. Before we state a theorem on the global dynamics of the above scalar field, let us address the issue of well-posedness of (32). It turns out that this has not yet been proven explicitly for asymptotically anti de Sitter spacetimes, nor has it for the special cases of Schwarzschild and Kerr-AdS. For pure AdS (M = a = 0), however, it can be deduced from [1]. The peculiarity of this problem arises from the fact that due to the timelike nature of I asymptotically AdS spacetimes are not globally hyperbolic. Hence in general “appropriate” boundary conditions have to be imposed on I to make the dynamics well-posed. We will formulate and prove the precise well-posedness statement for asymptotically AdS spacetimes in a separate paper. For the purpose of this paper, we content ourselves with considering a class of ψ defined as follows: Definition 3.1. Fix α < 49 , a Kerr-AdS spacetime M, g M,a, and a constant t slice 0 in D = J + (I) ∩ J − (I), as well as an integer k ≥ 0. We say that the function ψ is k if in J + a solution of class Cdec ( 0 ) ∩ J − (I) , • ψ is C k , • ψ satisfies (32), √ • for all δ < 21 9 − 4α 3
lim |r 2 +n+δ ∂rn ψ| = 0
r →∞
holds for n = 0, 1, . . . , k.
(33)
k there is no In particular, the decay (33) ensures that for a solution of class Cdec energy flux (cf. Sect. 4.1) through the AdS boundary. The decay (33) is precisely the one expected from the AdS case [1] and also strongly suggested from an asymptotic expansion in r of Eq. (32) (performed in [2] for pure AdS). The existence of a large class of solutions with the properties of Definition 3.1, which arise from appropriate initial data prescribed on 0 , would follow from a general well-posedness statement phrased in terms of weighted Sobolev norms (cf. also the Appendix). We emphasize again that uniqueness is only expected for α < 45 .
3.2. Vectorfield multipliers and commutators. We will obtain estimates for the field ψ via vectorfield multipliers (and eventually commutators). Since the general technique is well known and reviewed in detail in [7] we will only give a brief summary. The starting point is the energy momentum tensor of the above scalar field α 1 Tµν = ∂µ ψ∂ν ψ − gµν ∂β ψ∂ β ψ − 2 ψ 2 . 2 l
(34)
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
177
It satisfies ∇ µ Tµν = 0,
(35)
provided that (32) holds. Consider a vectorfield X on spacetime. We define its associated currents JµX = Tµν X ν , K = Tµν X
(X ) µν
π
(36) ,
(37)
where (X ) π µν = 21 (∇ µ X ν + ∇ ν X µ ) is the deformation tensor of the vectorfield X . One has (using (35)) the identity (38) ∇ µ JµX = ∇ µ Tµν X ν = K X . Note that π vanishes if X is Killing. Integrating (38) over regions of spacetime relates boundary and volume terms via Stokes’ theorem. In particular, for background Killing vectorfields we obtain conservation laws. In writing out the aforementioned integral identities we will sometimes not spell out the measure explicitly (e.g. Eq. (72)), it being implicit that the measure is the one induced on the slices (cf. Eq. (31)) or the spacetime measure respectively. Besides using vectorfields as multipliers we will also use them as commutators. If X is a vectorfield and ψ satisfies (2) then X (ψ) satisfies (cf. the Appendix of [7]) α g X (ψ) + 2 X (ψ) = −2(X ) π γβ ∇γ ∇β ψ − 2 2 ∇ γ (X ) πγ µ − ∇µ (X ) πγγ ∇ µ ψ. l Note that if X is Killing the right-hand side vanishes. In general one may apply multipliers to the commuted equation to derive estimates for higher order derivatives. 4. Boundedness in the Schwarzschild Case Here is our boundedness theorem for Schwarzschild-anti de Sitter:
Theorem 4.1. Fix a Schwarzschild-anti de Sitter spacetime M, g M>0, and 0 = τ0 a slice of constant t = τ0 in D = J + (I) ∩ J − (I).6 Let α < 49 and ψ n+1 with k ψ ∈ C n+1−k for k = 0, ..., n, where n ≥ 0 is be a solution to (32) of class Cdec i dec an integer. If n 2 2 1 k 2 k k 2 /
r 2 dr dω < ∞, ∂ ∂ ψ + r ψ + | ∇ ψ| (39) t r 2 r 0 k=0
then
2 2 1 k 2 k k 2 /
r 2 dr dω ∂ ∂ ψ + r ψ + | ∇ ψ| t r 2 r k=0 τ n 2 2 1 k 2 k k 2 2 / ψ| r dr dω ∂t ψ + r ∂r ψ + |∇ ≤C 2 0 r
n
(40)
k=0
for a constant C which just depends on M, l and α. Here τ denotes any constant t
slice to the future of 0 and restricted to r ≥ rhoz . 6 Note that such slices satisfy in particular H− ∩ = ∅. 0
178
G. Holzegel
By Sobolev embedding on S 2 we immediately obtain Corollary 4.1. The pointwise bound 2 2 2 1 k / k ψ|2 r 2 dr dω + r 2 ∂r k ψ + |∇ k=0 0 r 2 ∂t ψ |ψ| ≤ C (41) 3 r2 holds in the exterior J − I + ∩ J + ( 0 ) for a constant C just depending on the initial data, M, l and α. The remainder of this section is spent proving the above theorem. In the following we denote by B and b constants which just depend on the fixed parameters M, l and α. We also define R (τ1 , τ2 ) = ∪τ1 ≤τ ≤τ2 τ
(42)
to be the region enclosed by the slices τ1 and τ2 , a piece of I, and the horizon piece (43) H (τ1 , τ2 ) = H ∩ J + τ1 ∩ J − τ2 . Compare the figure above. 4.1. Positivity of energy. The first step is to obtain a positive energy arising from the Killing vectorfield T = ∂t . For this we apply the vectorfield identity (38) in the region R t1 , t2 . For the energy flux through a slice we obtain ∞
T (∂t , n ) g¯ dr dω E t = 1 = 2
rhoz
∞
rhoz
S2
S2
(∂t ψ)
2
k+ k02
α 2 2 2 / − 2ψ + (∂r ψ) (1 − µ) + |∇ψ| r dr dω. (44) l 2
The flux through the horizon is E H[t1 ,t2 ] =
H(t1 ,t2 )
(∂t ψ)2 r 2 dt dω,
(45)
hence in particular non-negative. Finally, the flux through I, the AdS boundary, vanishes because of the boundary conditions imposed. Combining these facts we obtain the energy identity E t2 = E t1 + E H[t1 ,t2 ] , (46) stating in particular that E (t ) is non-increasing. Next we show that the energy flux through the slices is positive: Lemma 4.1. We have 1 E t ≥ 2
∞
S2
rhoz
/ +|∇ψ|
(∂ ψ) t
2
r 2 dr dω.
2
k+ k02
4 + 1 − α (∂r ψ)2 (1 − µ) 9 (47)
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
179
Proof.
∞
rhoz
ψ r dr = 2 2
3 r 3 − rhoz
3
∞ ψ 2 rhoz
2 − 3
∞
rhoz
3 dr. ψψr r 3 − rhoz
(48)
The boundary term vanishes because of the decay of ψ at infinity and we obtain the Hardy inequality 3 2 ∞ 3 r − rhoz 4 ∞ 4 ∞ 2 2 2 ψ r dr ≤ dr ≤ (∂r ψ) (∂r ψ)2 r 2 l 2 (1 − µ) dr, 9 rhoz r2 9 rhoz rhoz (49) 3 ≤ r 3 in the last step. Hence where we used (7) and that r 3 − rhoz ∞ ∞ α 4 2 2 ψ r dr ≤ α (∂r ψ)2 r 2 (1 − µ) dr. l 2 rhoz 9 rhoz
Inserting this into (44) yields the result.
(50)
The lemma clearly reveals the relevance of the Breitenlohner-Freedman bound in ensuring a positive energy. We emphasize that one has positivity of energy only in an integrated sense! Lemma 4.1 also answers a question posed in [15] on whether one can construct a positive energy for the range 0 < α < 49 . Finally, we note that a similar Hardy inequality was used previously in the Appendix of [1], where the massive wave equation is studied on pure AdS. Having established positivity of the energy we only have to deal with the fact that the control over the (∂r ψ)2 term degenerates at the horizon. This is achieved using a so-called “redshift vectorfield”. See [8] for its first appearance in the context of asymptotically flat Schwarzschild black holes and the recent [7] for a version applicable to all non-extremal black holes. An additional difficulty in the present context lies in the fact that the energy momentum tensor does not satisfy the dominant energy condition. 4.2. The redshift. Define γ 1−µ k+ γ Y = ∂t + − + β ∂r . +β 2k0 2k0 2 2
(51)
Here γ ≥ 0 and β ≥ 0 are both functions of r only, which we will define below. We compute the current 2 3 k2
k2
ψ) (∂ t µ JµY n = T (Y, n ) = k+ 0 + β k+2 0 γ 2 2 2k04
(∂t ψ) (∂r ψ) − k+ γ + β k+ k− 2k0 2k β k− (∂r ψ)2 γ 0 k+ k0 + √ + 2k0 2 2 k+ +
+
/ 2− |∇ψ| 2
α 2 ψ l2
γ β k+ . √ + 2 k+ 2
(52)
180
G. Holzegel
Next consider the current associated with N = T + eY for some small constant e (depending on α): µ JµN n
3 k2
k2 (∂t ψ)2 3 0 2 0 = T (N , n ) = k+ k0 + eγ k+ + eβ k+ 2 2 2k04
(∂t ψ) (∂r ψ) − k+ eγ + eβ k+ k− + 2k0 2k k02 γ β k− (∂r ψ)2 0 k+ k0 + e √ + √ k− + e 2k0 2 2 k+ k+ / 2 − α2 ψ 2 k0 |∇ψ| γ β l k+ . + √ +e √ +e 2 2 k+ 2 k+
(53)
The bulk term of the vectorfield N reads 1 N αβ K = πY Tαβ e 2M 2r k+ (∂t ψ)2 2 γ − γ + k = + k β + 2 − µ) − γ (β (1 ) ,r − ,r + r2 l2 r 2k02
2M 2r 8M (∂t ψ) (∂r ψ) −2 γ − γ +2β + k k k − + (β (1−µ)−γ ) ,r − ,r − + 2k0 r2 l2 r2 2 γ,r k− k− k− M r 2 + (∂r ψ) γ − + β,r − + (β (1 − µ) − γ ) r 2 l2 2 2 r k− β,r γ,r β 2M 2r 2 / − − + |∇ψ| + 2 + 2 2 r2 l 2 γ,r α 2 1 γ k− β,r β 2M 2r βk− − + + . (54) + 2ψ + 2 − l r r 2 2 r2 l 2 KY =
Let us define the functions γ and β. Set γ = ξ (r ) (1 + (1 − µ)) and β = 1δ (1 − µ) ξ (r ). Here ξ is a smooth positive function equal to 1 in r ≤ r0 and identically zero for r ≥ r1 . The quantity δ is a small parameter. (Note that β,r is positive in r ≤ r0 ). Let us regard ξ as being fixed. We choose r0 and δ such that the following conditions hold in r ≤ r0 : 2M 2r k+ γ + 2 − γ,r k− + k+2 β,r + 2 (β (1 − µ) − γ ) ≥ b, r2 l r 2 γ,r k− k− k− M r − + + β,r − γ (β (1 − µ) − γ ) ≥ b, r 2 l2 2 2 r γ,r β 2M 2r β,r k− + − ≥ b, + − 2 2 r2 l2 2
(55)
(56)
(57)
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
181
and
2M 2r 8M (∂t ψ) (∂r ψ) −2 γ + 2 − γ,r k− + 2β,r k− k+ − 2 (β (1 − µ) − γ ) 2k0 r2 l r 2 1 (∂t ψ) 2M 2r k+ 2 ≤ γ + 2 − γ,r k− + k+ β,r + 2 (β (1 − µ) − γ ) 2 r2 l r 2k02 2 γ,r k− k− k− M r 2 + (∂r ψ) γ − + β,r − + (58) (β (1 − µ) − γ ) . r 2 l2 2 2 r
It is easily seen all these inequalities can be achieved by inserting the expressions for γ and β, then choosing δ sufficiently small and finally r0 sufficiently close to rhoz to exploit factors of k− = (1 − µ). With γ and β being determined choose e so small that eγ ≤
1 2k0
and, in case that 0 < α < 49 , also 9 1 − 49 α 1 e sup (γ + βk+ ) < 2k0 8α r
(59)
(60)
hold for all r . 4.2.1. The N bulk term. Let us denote by K0N the expression for K N with the zeroth order term removed. Lemma 4.2. The quantities δ, r0 and r1 can be chosen such that / 2 K0N ≥ b (∂t ψ)2 + (∂r ψ)2 + |∇ψ|
(61)
pointwise in r ≤ r0 . On the other hand, / 2 − K0N ≤ B (∂t ψ)2 + (∂r ψ)2 + |∇ψ|
(62)
pointwise in r0 ≤ r ≤ r1 . Proof. The second inequality is immediate since we are away from the horizon. The first is a consequence of the inequalities (55)-(58). 4.2.2. The N boundary terms. Let us investigate what the current J N actually controls. We can write √ √ √ k+ k+ eγ k+ (∂t ψ)2 (∂r ψ)2 k0 N µ − eγ 2 + Jµ n = T (N , n ) = √ k− + 2 k0 2 4 k+ 2k0 √ 2 γ k+ ∂r ψ 2 k− ∂t ψ β +e k+ 2 − k+ ∂t ψ + √ ∂r ψ +e 4 k0 2 4 k0 k+ α 2 2 / − l2 ψ |∇ψ| k0 γ β (63) + k+ √ +e √ +e 2 2 k+ 2 k+
182
G. Holzegel
with the term in the penultimate line being manifestly non-negative. From (59) one easily obtains a lower bound for the square bracket multiplying the (∂t ψ)2 -term. Inequality (59) in turn allows us to control the bad (for 0 < α < 49 !) zeroth order term by borrowing from the r -derivative term using the Hardy inequality (50). Namely, if 0 < α < 49 , then: e α 2 ψ 1+ (γ + βk+ ) r 2 dr 2 l 2k 0 rhoz 4 ∞ α 9 1 − 9α ≤ 1+ ψ 2 r 2 dr 2 8α rhoz l ∞ 4 1 1+ α ≤ (∂r ψ)2 (1 − µ) r 2 dr. 2 9 rhoz
∞
(64)
Hence finally
∞
√
k+ 2 r dr dω k02 rhoz √ ∞ k+ (∂t ψ)2 ≥ 2 2 2k 0 rhoz S √ 4 γ k+ 2 (∂r ψ)2 1 1 − α (1 − µ) + e r dr dω + 2 2 9 4 ∞ / 2 2 |∇ψ| r dr dω, + 2 rhoz S 2 S2
µ JµN n
(65)
(66)
and we have established the following Lemma 4.3. ≥b
∞
rhoz
S2
µ JµN n
=
∞
S2
rhoz
µ JµN n
√
k+ 2 r dr dω k0
(∂t ψ) / 2 r 2 dr dω. + (∂r ψ)2 r 2 + |∇ψ| r2 2
(67)
Note that the degeneration of the (∂r ψ)2 -term which occurred in the T -energy has disappeared.
4.3. The boundedness. Proposition 4.1. There exist constants B and b such that for any τ2 ≥ τ1 ≥ 0, τ2 µ N µ Jµ n H+ + b dτ JµN n − H(τ1 ,τ2 ) τ τ12 τ µ ≤ KN + B dτ JµT n . R(τ1 ,τ2 )
τ1
τ
(68)
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
183
Proof. We first note that all zeroth order terms can be absorbed by the last term in (68) using inequality (50). Hence it suffices to prove the inequality for the first order terms. We have τ2 M α 2 N µ 2 2 /
eγ |∇ψ| − 2 ψ + (∂t ψ) r 2 dt dω. Jµ n H+ = 2r k0 l H(τ1 ,τ2 ) τ1 S2 (69) For α ≤ 0 the term has a good sign while for 0 < α < 49 , τ2 r1 Mr α 2 N µ
− Jµ n H+ ≤ dt dω dr ∂r −eγ ψ 2k0 l 2 H(τ1 ,τ2 ) τ1 S2 rhoz Bα 2 η (∂r ψ)2 + ψ dt dr dω ≤e η l2 R(τ1 ,τ2 )∩{r ≤r1 }
(70)
for any η > 0 using that γ is supported for r < r1 only. So we only need a little bit of the ∂r term which we can borrow from the good bulk term K N in r ≤ r0 (by Lemma 4.2) and from the T -energy term in the remaining region. We have τ2 Bα µ −K N + eη (∂r ψ)2 + e 2 ψ 2 +b dτ JµN n η l R(τ1 ,τ2 )∩{r ≤r1 } τ1 τ τ2 µ T ≤ B˜ dτ Jµ n (71) τ1
τ
for a small constant b and large constants B and B˜ as a consequence of Lemma 4.2. Combining this inequality with (70) yields (68). Remark. We have discarded good terms (the t and angular derivative) on the horizon in the proof of the proposition as they are not needed for the following argument. Later, when we start commuting the equation with the redshift vectorfield we will need to keep those positive terms in order to estimate certain errorterms (cf. Sec. 5.3). Using the previous proposition we can prove boundedness as follows. The N identity reads µ µ µ JµN n + JµN n H+ + KN = JµN n 0 . (72) H+
τ
R(0,τ )
An application of Proposition 4.1 yields the inequality τ2 τ2 µ µ JµN n + b dτ JµN n ≤ B dτ τ2
τ1
τ
τ1
τ
µ
τ
JµT n +
τ1
µ
JµN n τ . 1
(73)
Using (46) (in particular the fact that the T -energy from initial data) is non-increasing µ µ and setting f (τ ) = τ JµN n as well as D = 0 JµT n we arrive at τ2 f (τ2 ) + b f (τ ) dτ ≤ B D (τ2 − τ1 ) + f (τ1 ) (74) τ1
for any τ2 ≥ τ1 ≥ 0 from which boundedness of f (τ ) follows from a pigeonhole argument, cf. [7].
184
G. Holzegel
Due to the spherical symmetry of the background we can commute Eq. (32) with angular momentum operators and obtain boundedness of µ JµN k ψ n (75)
for any integer k assuming such a bound on the data. Theorem 4.1 follows. 4.4. Pointwise bounds. For completeness we give the proof of Corollary 4.1. We have ∞ ∞
1 |∂r ψ t , r, ω |dr ≤ √ 3 |∂r ψ (t , r, ω) |2 r 4 dr , (76) |ψ t , r, ω | ≤ r r 3r 2 and by Sobolev embedding on S 2 ,
∞
|∂r ψ t , r, ω |2 r 4 dr ≤ C˜
r
∞
r
≤ C˜
2
2 |∂r k (ψ) t , r, ω |2 r 4 dr dω S 2
S 2 k=0
k=0
µ JµN k ψ n ,
(77)
from which the statement of the corollary follows.
4.5. Higher order quantities. Clearly, via commuting with T and angular momentum operators, we obtain control over certain higher order energies as well. Using just this collection of vectorfields however, we will still not be able to estimate the derivative transverse to the horizon. This problem was resolved for the asymptotically-flat Kerr case in [6,7], roughly speaking by commuting with the redshift vectorfield as well. We will adapt their argument to the present asymptotically hyperbolic case in Sect. 5.3, when we deal with the Kerr-AdS metric. There the argument is unavoidable even to obtain a pointwise bound on ψ since one can no longer (trivially) commute with angular momentum operators! 5. The Kerr-AdS Case To generalize the argument to include the case of Kerr-AdS we have to circumvent several difficulties. First of all, the timelike Killing field T = ∂t is no longer timelike everywhere on the black hole exterior due to the presence of an ergoregion close to the horizon. Hence the T -identity alone will not produce positive boundary terms. The resolution is to consider the Killing field K = T + λ
(78)
for an appropriate constant λ and = ∂φ being the Killing field corresponding to the axisymmetry of the Kerr-AdS metric. K is everywhere timelike on the black hole exte2 . This rior (it is null on the horizon, coinciding with the generators) as long as |a|l < rhoz
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
185
is a consequence of the asymptotically AdS nature of the background7 : Note that the analogous vectorfield in asymptotically flat space would turn spacelike near infinity! These properties of K were first exploited by Hawking and Reall [10] to obtain a pointwise positive energy for fields whose energy momentum tensor satisfies the dominant energy condition, see the discussion in the Introduction. Not surprisingly, we will show below that for sufficiently small a the energy identity associated to K produces boundary terms with manifestly non-negative first order terms. The zeroth order term however still has the wrong sign for 0 < α < 94 (as in Schwarzschild-AdS), a consequence of the dominant energy condition being violated in this range. We resolve this issue by generalizing the Hardy inequality (50) to the Kerr-AdS case allowing us to again control the zeroth order term from the derivative term in the K energy. Because we have to “borrow” from the r -derivative term for the Hardy inequality, we have to impose stronger 2 . restrictions on the smallness of a than just |a|l < rhoz Since the argument close to the horizon involving the vectorfield Y is stable in itself (it applies to any black hole event horizon with positive surface gravity by Theorem 7.1 of [7]) our previous argument using the vectorfield N = K + eY immediately yields boundedness of the L 2 -norm of all derivatives. In particular, no further restrictions on the size of a arise. The only remaining difficulty is that we can no longer commute with angular momentum operators to obtain pointwise bounds because the spherical symmetry of Schwarzschild has been broken. The way around this is to commute (32) with K and Y and to derive elliptic estimates on the asymptotically hyperbolic slices . This is carried out in Sect. 5.3.
5.1. The vectorfield K : Positive energy. We first compute the currents associated with the Killing vectors T = ∂t and = ∂φ ,
− 1 1 α 1
−g t t (∂t ψ)2 + ψ2 (∂r ψ)2 − 2 2 2 −g t t
l 2 −g t t
2 1 φφ θθ 2 rφ g + ψ + g ψ) + 2g ψ) ∂ ψ (79) ∂ (∂ (∂ φ θ r φ
2 −g t t
T (∂t , n ) =
and t r t φ 2 g g
T ∂φ , n = −g t t . (80) ∂φ ψ (∂t φ) + t t ∂φ ψ (∂r φ) + t t ∂φ ψ g g We observe that we cannot control the (∂r ψ) ∂φ ψ term in the T -energy, because the (∂r ψ)2 term from which we have to borrow degenerates on the horizon. However, consider the vectorfield K = T + λ
with λ =
7 More precisely, the fact that g ∼ g ∼ g 2 φφ ∼ r . t t t φ
a + a2
2 rhoz
(81)
186
G. Holzegel
and its current on constant t slices α
2 −g t t · T (K , n ) = − 2 ψ 2 + g θθ (∂θ ψ)2 l 2a
∂φ ψ (∂t ψ) + −g t t (∂t ψ)2 − g t t 2 2 rhoz + a 2 2a
+ g φφ − g t φ 2 ψ ∂ φ rhoz + a 2
− a
+ . (∂r ψ)2 + 2 (∂r ψ) ∂φ ψ gr φ − g t r 2 rhoz + a 2
(82)
We choose a so small that 1 φφ 2a 2 2a 2 2
g − gt φ 2 = ≥ 0. − 2 + a2 k l 2 4 rhoz rhoz + a 2 4 θ sin2 θ 0 θ
(83)
This is possible because both terms decay like r12 in r . In the asymptotically flat case
the g t φ -term will eventually dominate because k0 = 1 in the asymptotically flat case 2 (k0 = 1 + rl 2 in AdS!). We also choose a small enough so that 2 1 t t 2a 1 φφ
ψ) + −g g ψ ψ ≥ 0. ∂ ∂ (∂t ψ)2 − g t t 2 (∂ φ t φ 2 8 rhoz + a 2
(84)
This is easily achieved since again all terms decay like r12 at infinity and the cross-term has a factor of a. Of the first order terms it remains to control the r φ cross-term. For this we note (cf. Lemma 5.1) a a 2 2 k = + a − 2Mr r 0 hoz 2 + a2 2 + a2 rhoz k0 rhoz a ( − − k0 (rhoz + r ) (r − rhoz )) = 2 k0 rhoz + a 2 l2 a (r − rhoz ) r− . (85) = k0 l2 rhoz The reason for the above choice of K is now obvious: The (∂r ψ) ∂φ ψ -term has acquired a weight which degenerates on the horizon. Hence we can borrow from the (∂r ψ)2 term whose weight also degenerates. Moreover, the decay of (85) in r is strong, which will be exploited soon. Before we continue we obtain rather precise control over how the quantity − deteriorates on the horizon: gr φ − g t
r
Lemma 5.1. We have a 2l 2 1 3 2 2 2 2
− = 2 (r − rhoz ) r + r rhoz + r rhoz + l + a − l rhoz
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
187
and 3 2 3 + 3a 2 (r − r r − rhoz hoz )
− − = f, r 2 + a2 l 2
(86)
where 1 a 2l 2 3 2 (r − rhoz ) r 3 l 2 − 4a 2 + r 2 rhoz f = 2 − + a r hoz rhoz r + a2 l 2 2 a 4 l 2 4 2 2 + r rhoz . − 8a 4 + a 2 rhoz + l 2 + rhoz rhoz + 3a 2 − rhoz
(87)
Proof. Direct computation. Corollary 5.1. For sufficiently small a we have f ≥ 0. Proof. All coefficients of the polynomial in the square bracket of (87) are positive for 2 is seen to be sufficient). small enough a (the condition |a| < 21 l in addition to |a|l < rhoz Corollary 5.2. For sufficiently small a we have 16 g φφ
2 a (r − rhoz ) l2 r− < f. k0 l2 rhoz
(88)
Proof. Direct computation. Of course we recover our previous result (7) for a = 0. Note also that f grows slower (like ∼ r 2 ) than − (which grows like ∼ r 4 ). Going back to (85) we have the estimate rφ
2 (∂r φ) ∂φ ψ g − g t r
a 2 + a2 rhoz 2 2 8 a (r − rhoz ) l2 1 r− ≤ (∂r ψ)2 + g φφ ∂φ ψ φφ 2 g k0 l rhoz 8 ≤
2 f (∂r φ)2 1 φφ + g ∂φ ψ 2 8
(89)
for sufficiently small a using Corollary 5.2. It is here where the good decay of (85) is exploited, allowing us to borrow from f (and not the full − ) only. Now that positivity has been obtained for all first order terms we can turn to the zeroth order term and generalize the Hardy inequality (50). We note α l2
dθ dφ S2
α ≤ 2 l
S2
∞
2
ψ t , r, θ, φ rhoz ∞ dθ dφ sin θ dr r 2 + a 2 ψ 2 dr sin θ
rhoz
(90)
188
G. Holzegel
and estimate as previously, 2 ∞ 3 3 + 3a 2 (r − r r − rhoz α ∞ 2 4 hoz ) 2 2 dr r + a ψ ≤ α dr (∂r ψ)2 l 2 rhoz 9 rhoz r 2 + a2 l 2 ∞ 4 = α dr ( − − f ) (∂r ψ)2 (91) 9 rhoz using Lemma 5.1. We have obtained the Hardy inequality ∞ 2 α
ψ t , r, θ, φ dθ dφ dr sin θ 2 2 l S rhoz ∞ 4 sin θ ≤ α dθ dφ dr ( − − f ) (∂r ψ)2 . 9 S2 rhoz
(92)
Inserting the estimates (83), (84), (89) and (92) into (82) we finally have Proposition 5.1. For sufficiently small a the estimate ∞ µ µ
K Jµ (ψ) n = dθ dφ dr sin θ −g t t JµK n 2 S rhoz ∞ 1
− ≥b dθ dφ dr sin θ 2 (∂t ψ)2 + 2 (∂r ψ)2 r r rhoz S2 2 θ + ∂φ ψ + (∂θ ψ)2 2
θ sin θ
(93)
holds for a constant b which just depends on M, l and α < 49 . How large is a allowed to be? Note that in the case α ≥ 0 our estimates reduce to the simple geometric condition that K is everywhere timelike outside the black hole. 2 , the This in turn translates into a single inequality for a, thereby recovering |a|l < rhoz 9 result of [10]. If on the other hand 0 < α < 4 , the additional restriction of Corollary 5.2 has to be imposed. The reason is that the − (∂r φ)2 -term in the K -identity has to control both the mixed term (89) and the zeroth order term (92). We used a coefficient f (independent of α) to control the former and ( − − f ) to control the latter establishing that a uniform a can be chosen for α < 49 . Finally we compute the flux term on the horizon, where the normal is proportional to K : 2 a T (K , K ) = ∂φ ψ > 0. (94) ∂t ψ + 2 rhoz + a 2 H H 5.2. The vectorfield N. Now that we again have a positive, non-increasing K -energy we can invoke the identical argument as in the Schwarzschild case close to the horizon, involving the vectorfields Y and N = K + eY . This will eventually produce a bound ∞ 2 1 dθ dφ dr sin θ 2 (∂t ψ)2 + r 2 (∂r ψ)2 + ∂φ ψ 2 r
θ sin θ S2 rhoz
θ µ µ 2 N ≤B + Jµ (ψ) n ≤ B JµN (ψ) n 0 , (95) (∂θ ψ) 0
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
189
where the “bad” − -weight from the r -derivative term has now disappeared. Note that no further smallness restrictions on a arise as long as the horizon has positive surface gravity.
5.3. Higher order energies and pointwise bounds. We finally address the issue of pointwise bounds. Unfortunately, we can no longer exploit the commutation with the angular derivatives because of the broken spherical symmetry. However, we can certainly commute with T (or K respectively) to obtain integral bounds for certain higher order m the inequality energies. Given a solution of class Cdec
µ JµN K k ψ n ≤ B
0
µ JµN K k ψ n
(96)
for any non-negative integer k < m is an automatic consequence of the fact that K commutes with the wave operator. We write the wave equation as √ 1 √ ∂i g i j g∂ j ψ g = −g t
t
√ 1 α
(∂t ∂t ψ) − 2g t i ∂t ∂i ψ − √ ∂r g t r g (∂t ψ) − 2 ψ, g l
(97)
where i = 1, 2, 3 (or r, θ, φ respectively). Using both the decay assumed for ψ at infinity and that we control the right-hand side by (96), we obtain for any r0 > rhoz the elliptic estimate 1
i j kl
sin θ h h ∇ ψ) ∇ ∇ ψ −g t t dr dθ dφ (∇ i k j l
−g t t ∩{r ≥r0 >rhoz } µ µ
sin θ −g t t dr dθ dφ (98) JµN (ψ) n + JµN (K ψ) n ≤ C (r0 ) ∩{r ≥rhoz } away from the horizon, where h i j is the inverse of the induced metric on . Note that the argument breaks down at the horizon because of the degenerating weight of grr (whereas h rr is well-behaved there).8 To obtain good estimates close to the horizon we again adapt the ideas of [6,7]. Their resolution is to commute the equation with a version of the redshift vector field 1 1 Yˆ = ∂t − ∂r 2k0 2
(99)
which yields the equation 2 α 4r Yˆ ψ = + − Yˆ Yˆ (ψ) − Yˆ T (ψ) + λ1 (∂r ψ) + λ2 (∂t ψ) + 2 Yˆ (ψ), l
(100)
8 Actually, the estimate (98) remains true if one inserts an additional weight of r n for any n < √ max 2, 9 − 4α into the integrand on the left-hand side. This improvement is outlined in the Appendix where stronger weighted Sobolev norms are derived.
190
G. Holzegel
where λ1 =
1 1
∼ , 2 − r 0
λ2 = −
(101)
1 a 2 + 5r 2 ∼ 2, 2 l k0 r
(102)
and a prime denotes taking a derivative with respect to r . What will be crucial is that the coefficient of the first term on the right-hand side of (100) has a (good) sign close to the horizon. There is a geometric reason for this which was observed in [7]: Computing the surface gravity κ defined by (cf. [21]) ∂b K a K a = −2κ K b (103) r =rhoz
r =rhoz
we obtain κ=
− (rhoz ) 2 2 rhoz + a2
(104)
for the Kerr-AdS black hole under consideration. This means that on the horizon the coefficient of the Yˆ Yˆ (ψ)-term on the right-hand side of (100) is proportional to the surface gravity of the black hole, which is positive for a non-degenerate horizon! Remarkably, this observation is not bound to the special properties of the Kerr metric but a general fact about black hole event horizons with positive surface gravity. For details the reader may consult [7]. Let us investigate why this sign allows us to derive good estimates close to the horizon. The first step is to apply the identity (38) with the vectorfield multiplier N to Eq. (100): µ µ N ˆ N ˆ Jµ Y ψ n + Jµ Y ψ n H + K N Yˆ ψ τ H+ (0,τ ) R(0,τ )∩{r ≤r0 } µ =− K N Yˆ ψ + JµN Yˆ ψ n 0 R(0,τ )∩{r0 ≤r ≤r1 } 0 (105) + E N Yˆ ψ + E N Yˆ ψ , R(0,τ )∩{rhoz ≤r ≤r0 }
R(0,τ )∩{r >r0 }
where 2 2
− N − eYˆ Yˆ ψ Yˆ Yˆ (ψ) E N Yˆ ψ = −2e − Yˆ Yˆ (ψ) − 4r ˆ ˆ + N Y ψ Y T ψ − λ1 (∂r ψ) N Yˆ ψ − λ2 (∂t ψ) N Yˆ ψ α (106) − 2 N Yˆ ψ (Y (ψ)). l Let us start with the horizon term in (105). As previously (cf. (69)) this flux has a good sign except for the lowest order term when 0 < α < 49 .
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
191
This latter term is estimated as previously (cf. (70)) borrowing a bit from the good K N term: eγ M α ˆ 2 Yψ 2 H+ (0,τ ) 2r k0 l 1 τ N ˆ ≤ K Y ψ + B dt JµN (ψ) n µ ≤ K N Yˆ ψ + B Dτ, (107) 0 τ with D=
0
µ µ JµN (ψ) n 0 + JµN (K ψ) n 0 .
(108)
We turn to the E N -terms in (105). Note that N − eYˆ = K on the horizon. Hence in r ≤ r0 (where weights in r don’t matter) we have −
2 − N − eYˆ Yˆ ψ Yˆ Yˆ (ψ) ≤ B K + (1 − µ) Yˆ Yˆ ψ Yˆ Yˆ (ψ) 2 2 1 ˆ ≤ Y K ψ + Yˆ Yˆ (ψ) (109)
by choosing r0 sufficiently close to the horizon to exploit the (1 − µ)-term as a smallness factor. We are going to choose small enough to borrow from the good first term in (106). Similarly, 2 2 2 4r ˆ ˆ N Y ψ Y T ψ ≤ Yˆ Yˆ (ψ) + B Yˆ K ψ + B|λ| · ∂φ Yˆ (ψ) . (110) Using several integrations by parts and the elliptic estimate (119) we derive the following estimate for the last term: 2 B|λ| ∂φ Yˆ (ψ) ≤ K N Yˆ ψ + B D (τ + 1) R(0,τ ) R(0,τ )∩{r ≤r0 } 1 1 1 µ µ µ + JµN Yˆ ψ )n H+ + JµN Yˆ ψ n + JµN Yˆ ψ n . (111) 8 H(0,τ ) 8 τ 8 0 Let us denote the three boundary-terms in the last line collectively by P1 . The other terms of E N involve lower order terms and are estimated via Cauchy’s inequality, putting a small weight on the Yˆ Yˆ -term so that one can borrow again from the good term. Hence finally R(0,τ )∩{r ≤r0 }
E N Yˆ ψ ≤
R(0,τ )∩{r ≤r0 }
K N Yˆ ψ
2 1 1 ˆ N Y K ψ + K (ψ) + B D (τ + 1) + P1 + R(0,τ )∩{r ≤r0 } ≤ K N Yˆ ψ + B D (τ + 1) + P1 .
R(0,τ )∩{r ≤r0 }
(112)
192
G. Holzegel
The penultimate term in the second line deals in particular with the first order terms (which have the wrong sign if α > 0) arising in the K N terms. For r ≥ r0 we only have to be careful with the weights in r : E N Yˆ ψ
E N Yˆ ψ ≤ dt . (113)
−g t t R(0,τ )∩{r ≥r0 } τ
In the Ad S case g t t ∼ r12 , so in comparison with the asymptotically flat case (where this quantity approaches 1 at infinity) one loses a power of r . However, the decay of N E Yˆ ψ in r is easily seen to be strong enough to have
E N Yˆ ψ
E N Yˆ ψ ≤ dt
−g t t R(0,τ )∩{r ≥r0 } τ 0 τ µ JµN (ψ) n + JµN (K ψ) ≤ B Dτ, ≤ dt
τ
τ
0
(114)
using the elliptic estimate (98) above. For K N we have the analogous estimates to Lemma 4.2. In fact the lowest (first) order terms of the wrong sign can now be simply controlled τ by adding 0 dt τ JµN (ψ) n µ ≤ B Dτ to the right-hand side – a Hardy inequality is no longer necessary. Putting these estimates together the identity (105) turns into the estimate τ µ µ N ˆ Jµ Y ψ n + b dτ JµN Yˆ ψ n τ τ 0 µ ≤ B D (τ + 1) + 2 JµN Yˆ ψ n 0 . (115) 0
This is the analogue of (74) and hence µ N ˆ Jµ Y ψ n ≤ B D + τ
0
JµN
µ ˆ Y ψ n 0
(116)
follows as previously. We have finally obtained control over the rr derivative at the horizon. We now show how to control the second derivatives on the topological twospheres close to the horizon defined by constant (t , r ). For this we employ a second elliptic estimate. Write the wave equation as 1 1
∂θ (sin θ θ ∂θ ψ) + g φφ − 2λg t φ + λ2 g t t ∂φ2 ψ sin θ
= −g t t ∂t (K ψ) − 2g t r ∂r (K ψ) + g t t λ − 2g t φ ∂φ (K ψ) − grr ∂r ∂r ψ + ξ (r − rhoz ) ∂r ∂φ ψ + lower order terms,
(117)
where ξ is some bounded function. Note also the degenerating weight of grr on the horizon. From (117) we derive µ µ µ / 2 ψ|2 ≤ B JµN (ψ) n + JµN (K ψ) n + JµN Yˆ ψ n (118) |∇ Sr
Sr
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
193
on a sphere of radius r . Close to the horizon we obtain µ / 2 ψ|2 ≤ B JµN (ψ) n + JµN (K ψ) + · JµN Yˆ ψ , (119) |∇ τ ∩{r ≤r0 }
τ ∩{r ≤r0 }
with arising from choosing r0 small enough to exploit the degenerating weights in (117). Together with the previous elliptic estimate finally 1 µ µ i j kl N N
h h (∇i ∇k ψ) ∇ j ∇l ψ + J (K ψ) n + J (ψ) n
−g t t µ µ µ JµN (ψ) n + JµN (K ψ) n + JµN Yˆ ψ n ≤B τ µ µ µ ≤B JµN (ψ) n + JµN (K ψ) n 0 + JµN Yˆ ψ n 0 . (120) 0
To write the estimate in a more geometric form we observe 1 µ ij
h ∇i ψ∇ j ψ ≤ B JµN (ψ) n
−g t t and, from (92),
1
−g t t
The factor of √ r2
ψ2 ≤ B
∞
S2
rhoz
∼ r arises because
1
−g t t
r 2 ψ 2 dr dω ≤ B √
(121)
µ
JµN (ψ) n .
(122)
h ∼ r for the hyperbolic metric (instead of
in the asymptotically flat case). In the other direction we have µ JµN Yˆ ψ n 1 µ µ N N i j kl ≤B h h (∇i ∇k ψ) ∇ j ∇l ψ , (123) J (K ψ) n + J (ψ) n +
−g t t
which is seen by direct computation taking care of the weights of r . Hence the inequality (120) becomes µ µ ψ J N (K ψ) n + J N (ψ) n + 2 Hw ( ) µ µ ≤ B ψ 2 J N (K ψ) n 0 + J N (ψ) n 0 (124) + Hw ( 0 )
with Hw2 denoting the n ψ
0
√ 1 -weighted −g t t
H⊥1
:=
≤ B ψ
Hw2 ( )
1
H 2 norm of . Finally, we define 2 r 2 ∇ · n ψ + (n ψ)2
−g t t µ µ J N (K ψ) n + J N (ψ) n +
and state the boundedness theorem for the massive wave equation on Kerr-AdS:
(125)
194
G. Holzegel
Theorem 5.1. Fix a cosmological constant = − l32 , a mass M > 0 and some α < 49 . There exists an amax > 0, depending only on , M and α, such that the following statement is true for all a with |a| < amax . Let 0 = τ0 bea slice of constant t = τ0 in D = J + (I) ∩ J − (I) of the Kerr-anti 2 . If de Sitter spacetime M, g M,a, . Let ψ be a solution to (32) of class Cdec µ µ JµK (ψ) n 0 + JµK (K ψ) n 0 < ∞, (126) 0
0
then ψ
Hw2 ( )
+ n ψ
≤ C ψ˜
Hw2 ( 0 )
H⊥1 ( )
+
1 m=0
+ n ψ˜
H⊥1 ( 0 )
µ JµK K m ψ n +
1 m=0 0
µ JµK K m ψ n 0
(127)
holds on J + ( 0 ) ∩ D for a uniform constant C depending only on the parameters M, a, l and α. Here denotes any constant t -slice to the future of 0 . We remark that in contrast to the statement in the asymptotically flat case [7] we need the K boundary term in this estimate. This is because the weighted L 2 norm of the second time-derivative that one obtains from the current J K is stronger than what can be derived from the wave equation in combination with the boundedness of |ψ| Hw2 ( ) and |n ψ| H 1 ( ) alone. ⊥
r2
As mentioned previously, if α ≤ 0 then amax = hoz l since all that was needed in the proof is that the vectorfield K is timelike on the exterior (cf. [10]). In general, the restriction on amax depends on α and becomes tighter as α approaches the Breitenlohner Freedman bound, α → 49 . However, there is a uniform lower bound on amax (i.e. amax ≥ auni f or m > 0 for any α < 49 ) which can be worked out explicitly from the estimates of Sect. 5.1. The Sobolev embedding theorem for asymptotically hyperbolic space yields Corollary 5.3. We have the pointwise bound 1 |ψ| ≤ C ψ˜ 2 + n ψ˜ 1 + Hw ( 0 )
H⊥ ( 0 )
m=0 0
µ JµK K m ψ n 0
(128)
on J + ( 0 ) ∩ D. To prove the corollary we rely on the following general Sobolev embedding theorem (cf. Theorem 3.4 in [11]) Theorem 5.2. Let (N , h) be a smooth complete Riemannian 3-manifold with Ricci curvature bounded from below and positive injectivity radius and u ∈ H 2 (N ) a function on N . Then 2 2 sup |u| ≤ B |∇ j u|2 dv (h). (129) N
j=0
N
Remark. Our τ is only complete with respect to the asymptotically hyperbolic end, but it is straightforward to incorporate the boundary at r = rhoz > 0.
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
195
6. Final Comments As mentioned in the abstract of the paper, the result does not make use of the separability properties of (2) with respect to the Kerr background. In fact it does not make use of the axisymmetry either! All that was needed was the causal Killing vectorfield K on the black hole exterior. In view of this fact, Theorem 5.1 can be stated in the following generalized setting: Fix the Killing vectorfield K = T + λ of a slowly rotating Kerr-AdS background as in Sect. 5. Perturb the metric such that it stays C 1 -close to the Kerr-AdS metric and such that K remains both Killing and null on the horizon.9 Then Theorem 5.1 remains true for such spacetimes. The main motivation for generalizations of this type are non-linear situations, in which the metric is not known explicitly a-priori but is itself dynamical. In view of this one should use as less quantitative assumptions on the metric as possible to obtain bounds on the fields. Compare [7] for a more detailed discussion. As a further generalization one may assume only an approximate causal Killing field, i.e. a vectorfield whose deformation tensor decays sufficiently fast in t. Treating the latter as a decaying error term in the estimates one can prove boundedness of (2) for all spacetimes approaching a spacetime that is C 1 -close in the sense above10 to a slowly rotating Kerr-AdS solution. The question whether ψ decays in time and if so, at what rate remains open. Acknowledgements. I would like to thank Mihalis Dafermos and Igor Rodnianski for stimulating discussions and useful comments. I am also grateful to an anonymous referee for his careful reading of the manuscript and many insightful remarks and suggestions.
A. Radial Decay In this Appendix we outline how – assuming the boundary conditions (33) – one can establish boundedness of appropriate higher weighted Sobolev-norms. The idea is the 3 , let T k (ψ), denote the application of k times following. Given a solution ψ of class Cdec the vectorfield T to ψ. We know that the T -energy associated with T k (ψ) is conserved for k = 0, 1, 2. If we now revisit the wave equation to do elliptic estimates, this will introduce natural weighted Sobolev norms whose r -weights depend on 0 < α < 49 . These optimized norms are expected to play a crucial role for the local existence theorem and are hence presented here. Recall from (97) that (for Kerr-AdS) one may write the wave equation in the form √ α 1 √ ∂i g i j g∂ j ψ + 2 ψ g l √ 1
= −g t t (∂t ∂t ψ) − 2g t i ∂t ∂i ψ − √ ∂r g t r g (∂t ψ). (130) g 2
The weighted of the right-hand side decays very fast in view of the boundedness L norm µ µ of JµK T k ψ n + JµK (ψ) n . In particular, denoting the right-hand side by f we have ∞ µ µ (131) dr dωr 4 f 2 < B JµK (T ψ) n + JµK (ψ) n . rhoz
S2
9 The C 1 regularity is necessary because the surface gravity, whose positivity was essential for the argument, is C 1 in the metric. 10 I.e. in particular admitting a timelike Killing field K on the black hole exterior which becomes null on the horizon.
196
G. Holzegel
Let now χ = χ (r ) be a cut-off function which is equal to 1 for r > R for √ some fixed 2n R > rhoz and is zero close to the horizon. Set σ = n+3 for some n < 9 − 4α and multiply (130) by √ 1 α √ n ij g gχ ∂ ψ − χ σ ψ gr ∂ √ i j g l2 α √ n √ = χ f + (1 − σ )) 2 ψ gr + χ,r gr j ∂ j φ gr n . l Integrating over the slice , we observe (in view of (131)) that we can bound the right hand side as long as n ≤ 2 (provided we can borrow an of the term r 2+n ψ 2 dr dω from the left). The radial derivatives11 on the left-hand side can after several integrations by parts be estimated by ∞ dr dωr 6+n (∂r ∂r ψ)2 S2 R ∞ + dr dω (16 − 4 (5 + n) − α + ασ ) r 4+n (∂r ψ)2 S2 R ∞ (132) 1−σ 2+n 2 dr dωα −σ α + n (n + 3) r ψ 2 S2 R µ µ K K ≤B Jµ (T ψ) n + Jµ (ψ) n .
The boundary terms arising in the computations all vanish: on the left because of the √ cut-off function, and on the right in view of (33) (the restriction n < 9 − 4α is imposed in order to make these boundary terms vanish). Note also that the error-terms introduced by the derivatives of the cut-off function are well in the interior (bounded r ) and hence unproblematic. The coefficient of the zeroth order term in (132)√is positive if n(n+3) 2n n+3 = σ < 2α+n(n+3) , which an easy calculation reveals to be true for n < 9 − 4α. (In particular we have room to borrow the aforementioned of this term for the right-hand side.) The second term on the left of (132) is negative. How large can we allow n to be to still absorb this term by the first term using a Hardy inequality? As one easily checks, the Hardy condition (4 + 4n + α (1 − σ ))
4 (n + 5)2
<1
(133)
is satisfied for n<
√
9 − 4α.
We conclude that away from the horizon rn
h i j h kl (∇i ∇k ψ) ∇ j ∇l ψ
−g t t ∩{r ≥R} µ µ JµK (ψ) n + JµK (T ψ) n . ≤ C (R, n) ∩{r ≥rhoz }
11 These derivatives are the problematic ones.
(134)
(135)
Massive Wave Equation on Slowly Rotating Kerr AdS Spacetimes
197
√ √ for any n < min 2, 9 − 4α . In case that 9 − 4α > 2 one simply repeats the argument after another commutation with T . In particular, (131) holds with a weight µ r 6 replacing r 4 on the left, if one adds the term JµK (T T ψ) n on the right-hand √ 12 From this one derives estimate (135) with any weight n < side. 9 − 4α adding K (T T ψ) n µ on the right-hand side. J µ References 1. Bachelot, A.: The Dirac system on the anti-de Sitter universe. Commun. Math. Phys. 283, 127–167 (2008) 2. Breitenlohner, P., Freedman, D.Z.: Stability in gauged extended supergravity. Ann. Phys. 144, 249 (1982) 3. Carter, B.: Hamilton-Jacobi and Schroedinger separable solutions of Einstein’s equations. Commun. Math. Phys. 10, 280 (1968) 4. Dafermos, M., Rodnianski, I.: A proof of Price’s law for the collapse of a self-gravitating scalar field. Invent. Math. 162, 381–457 (2005) 5. Dafermos, M., Rodnianski, I.: The wave equation on Schwarzschild-de Sitter spacetimes. http://arxiv. org/abs/0709.27661[gr-qc], 2007 6. Dafermos, M., Rodnianski, I.: A proof of the uniform boundedness of solutions to the wave equation on slowly rotating Kerr backgrounds. http://arxiv.org/abs/0805.43091[gr-qc], 2008 7. Dafermos, M., Rodnianski, I.: Lectures on black holes and linear waves. http://arxiv.org/abs/0811. 03541[gr-qc], 2008 8. Dafermos, M., Rodnianski, I.: The red-shift effect and radiation decay on black hole spacetimes. Comm. Pure Appl. Math. 62, 859–919 (2009) 9. Gubser, S.S.: Breaking an Abelian gauge symmetry near a black hole horizon. Phys. Rev. D 78, 065034 (2008) 10. Hawking, S.W., Reall, H.S.: Charged and rotating AdS black holes and their CFT duals. Phys. Rev. D 61, 024014 (2000) 11. Hebey, E.: Nonlinear Analysis on Manifolds: Sobolev Spaces and Inequalities. Courant Lecture Notes in Mathematics 5, New York: Courant Institute, 1999 12. Ishibashi, A., Wald, R.M.: Dynamics in non-globally hyperbolic static spacetimes. III: anti-de Sitter spacetime. Class. Quant. Grav. 21, 2981–3014 (2004) 13. Bony, J.-F., Häfner, D.: Decay and non-decay of the local energy for the wave equation in the de SitterSchwarzschild metric. Commun. Math. Phys. 282, 697–719 (2008) 14. Kay, B.S., Wald, R.M.: Linear stability of Schwarzschild under perturbations which are nonvanishing on the bifurcation two sphere. Class. Quant. Grav. 4, 893–898 (1987) 15. Kunduri, H.K., Lucietti, J., Reall, H.S.: Gravitational perturbations of higher dimensional rotating black holes: Tensor Perturbations. Phys. Rev. D 74, 084021 (2006) 16. Blue, P., Sterbenz, J.: Uniform decay of local energy and the semi-linear wave equation on Schwarzschild space. Commun. Math. Phys. 268(2), 481–504 (2006) 17. Melrose, R., Sá, Barreto, A., Vasy, A.: Asymptotics of solutions of the wave equation on de Sitter-Schwarzschild space. http://arxiv.org/abs/:0811.22291[math.Ap], 2008 18. Sonner, J.: A rotating holographic superconductor. http://arxiv.org/abs/0903.06272[hep-tu], 2009 19. Tataru, D., Tohaneanu, M.: Local energy estimate on Kerr black hole backgrounds. http://arxiv.org/abs/ 0810.57662[math.AP], 2008 20. Wald, R.M.: Note on the stability of the Schwarzschild metric. J. Math. Phys. 20, 1056–1058 (1979) 21. Wald, R.M.: General Relativity. Chicago, IL: The University of Chicago Press, 1984 Communicated by P. T. Chru´sciel
12 This in turn is a consequence of the Hardy inequality
2 r (∂t ∂t ψ)2 dr ≤ 49 r 4 (∂r ∂t ∂t ψ)2 dr .
Commun. Math. Phys. 294, 199–228 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0918-x
Communications in
Mathematical Physics
Ergodicity and Energy Distributions for Some Boundary Driven Integrable Hamiltonian Chains Peter Balint1, , Kevin K. Lin2 , Lai-Sang Young3, 1 Institute of Mathematics, Budapest University of Technology and Economics,
Egry József u. 1, H-1111 Budapest, Hungary. E-mail:
[email protected]
2 Department of Mathematics, University of Arizona, Tucson, AZ 85721,
USA. E-mail:
[email protected]
3 Courant Institute of Mathematical Sciences, New York University,
New York, NY 10012, USA. E-mail:
[email protected] Received: 3 March 2009 / Accepted: 17 July 2009 Published online: 15 September 2009 – © Springer-Verlag 2009
Abstract: We consider systems of moving particles in 1-dimensional space interacting through energy storage sites. The ends of the systems are coupled to heat baths, and resulting steady states are studied. When the two heat baths are equal, an explicit formula for the (unique) equilibrium distribution is given. The bulk of the paper concerns nonequilibrium steady states, i.e., when the chain is coupled to two unequal heat baths. Rigorous results including ergodicity are proved. Numerical studies are carried out for two types of bath distributions. For chains driven by exponential baths, our main finding is that the system does not approach local thermodynamic equilibrium as system size tends to infinity. For bath distributions that are sharply peaked Gaussians, in spite of the near-integrable dynamics, transport properties are found to be more normal than expected. Introduction This paper concerns a simple model of energy transport along a chain of Hamiltonian systems with conservative, nearest-neighbor coupling. The ends of the chain are in contact with unequal heat baths, and of interest to us are properties of the resulting nonequilibrium steady state. In this paper, we address rigorously questions of invariant densities and ergodicity. This is followed by a numerical investigation of energy profiles along the chain and marginal distributions at specific locations as the length of the chain tends to infinity. The setup above is used in heat conduction studies, as are the questions raised [3,11, 21]. We propose to look at these questions in a broader theoretical framework, as transport of conserved quantities by dynamical processes in general and particle systems in particular. As an example of this extended viewpoint, we will sometimes consider energy Research partially supported by OTKA (Hungarian National Research Fund) grants: F 60206 and K 71693; and by the Bolyai scholarship of the Hungarian Academy of Sciences. Research partially supported by NSF Grant DMS-0600974.
200
P. Balint, K. K. Lin, L.-S. Young
sources that are not necessarily Gibbsian in distribution, even though we will continue to call them “heat baths”. Except for system-bath interactions, the class of models considered here obey purely deterministic (as opposed to stochastic) microscopic rules.1 Transport properties naturally depend on these dynamical rules. It is generally thought — though far from proved — that “chaotic” local dynamics lead to diffusion-like behavior on the macroscopic level, while “integrable” dynamics, which lie at the opposite end of the spectrum, operate differently. An early example that contributes to this understanding is a chain of coupled harmonic oscillators [28]; see [21] for other examples and more detailed discussion. The class of models considered in this paper is closer to the integrable end.
Description of model. The constituent subsystems in our chains are as simple as can be: each consists of a point mass moving freely back and forth along a unit length interval. We order these intervals linearly. Energy is stored at the sites where adjacent intervals meet; when a particle reaches a site, it exchanges energy with the site, then turns around and moves away at a speed equal to the square root of its new (kinetic) energy. The left and right ends of the chain are coupled to heat baths, which function like other sites except that the energies they emit are random; we assume they are i.i.d. with respect to certain prescribed distributions. This is a complete description of the class of models studied in this paper. If we replace the two baths by two regular sites, then the resulting closed chain is a very simple, totally non-chaotic dynamical system: it has many independent conserved quantities (besides the total energy) and is highly non-ergodic. For this reason, we will sometimes refer to it (loosely) as an “integrable Hamiltonian chain”. Any ergodic property of the open chain must come solely from the randomness of the baths, and this randomness must find a way to penetrate the entire chain, which can be arbitrarily long.
Preview of results. This paper has two separate parts: a rigorous part, and a numerical part. Our main rigorous result is ergodicity; under some conditions which require that the bath distributions be sufficiently spread out, we prove that there is at most one invariant probability measure, and that this measure has a density. In particular, we prove that our system admits no singular invariant probability measures. We also give the explicit form of the invariant density when the system is in equilibrium, i.e., when the two bath distributions are equal. Numerical studies of nonequilibrium steady states were carried out for two types of bath distributions: exponential distributions and sharply peaked Gaussians. When in equilibrium, exponential bath distributions give rise to Gibbs distributions for the moving particles in the chain. A finding that was a little surprising to us is that when forced out of equilibrium, our chain produces local marginal densities that are mixtures of Gibbs distributions (which are therefore not Gibbs). This is discussed in Section 6.2. Sharply peaked Gaussian bath distributions were chosen to accentuate the near-integrability of the dynamics. Here we discover that injected energies move about in a hybrid ballisticdiffusive motion, resembling a random walk with “momentum”, meaning a tendency to continue in the same direction; see Section 7.2. 1 There is a small exception to this statement: a stochastic rule is used to “break the tie” in the unlikely event of same-site double collisions; see the end of Section 1.1.
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
201
Related works. With regard to ergodicity type results for boundary driven Hamiltonian chains, the only previous works we are aware of are for anharmonic chains [12,15,27] and for certain particle models in which rotating disks are used as energy storage devices [13]. The latter was preceded by [20,25], which contain numerical studies of similar models, and by [16], which contains a simplification of the model in [20] and a partial analysis. A similar class of models, with paddles in the place of rotating disks, has also been studied [22]. These models have to some degree motivated the idea of energies at “sites” in the present paper. Related works in systems defined by dynamical local rules include [4–7,17]. For stochastic models, the collection of results is much larger, and we mention only some that are closer to this paper in spirit: [1,2,8–10,16,18,19,23,24,26,29]. 1. Model Description 1.1. Particle systems. We consider a model with N sites, denoted {1, 2, . . . , N }, connected to two baths which we think of as located at sites 0 and N + 1. The physical space of the system is the line segment [0, N + 1]. In this system, there are 2N + 1 particles, N of which are “stationary” and the rest are in perpetual motion. The i th stationary particle is confined to site i, while the i th moving particle, i = 0, 1, . . . , N , moves back and forth along the segment [i, i + 1]. Each particle carries an energy. The energy of the i th stationary particle √ is denoted by ξi , while the i th moving particle carries energy ηi and moves with speed ηi . The phase variables of this system are therefore X = ((x0 , σ0 , η0 ), ξ1 , (x1 , σ1 , η1 ), ξ2 , (· · · ), · · · ξ N , (x N , σ N , η N )). Here we let xi (t) ∈ [0, 1] denote the location at time t of the i th moving particle and σi (t) = ± its direction of travel, identifying the segments [i, i + 1] with [0, 1]. The variables ηi , ξi ∈ (0, ∞) are energies as defined earlier. Because the moving particles are trapped in gaps between integer sites, we will also refer to the ηi as “gap energies.” The rules for updating this continuous-time process are as follows. When a moving particle reaches the boundary of the segment to which it is confined, a collision occurs, and the post-collision dynamics are defined as follows: Suppose for definiteness that at time t0 , xi (t0− ) = 0, σi (t0− ) = − ;
i = 0 ,
i.e., at time t0 , the i th particle collides with site i, which we assume is not the left bath. We assume further that xi−1 (t0− ) = 1, i.e., it is not a double collision at site i. Then an exchange of energy between the particle at site i and the i th moving particle takes place, and the latter reverses its direction, i.e., ξi (t0+ ) = ηi (t0− ), ηi (t0+ ) = ξi (t0− ), xi (t0+ ) = 0, and σi (t0+ ) = +. An analogous rule holds when xi (t0− ) = 1 and σi− (t0 ) = + provided that i = N . When a particle reaches a bath, it bounces back with a random energy. More precisely, we think of the left and right baths as having energy distributions with densities L(ξ ) and R(ξ ) respectively. Each time the 0th moving particle reaches the left bath, its energy is replaced by one drawn randomly and independently from the distribution L(ξ ); an analogous statement holds when the N th moving particle reaches the right bath. Emissions from the left and right baths are independent. We say a multiple collision occurs at time t0 if at that time, two or more collisions occur simultaneously. These collisions may involve the same site, or different sites, or
202
P. Balint, K. K. Lin, L.-S. Young
some combination of sites and baths. The post-collision rule above leaves no ambiguity as to how to continue the dynamics after a multiple collision except when there is a “same-site double collision”, i.e., when two particles arrive, from the left and the right, to one of the sites at exactly the same time. There is no natural way to continue the dynamics after such a collision. We fix (arbitrarily) the following rule: equal probability is assigned to the two scenarios corresponding to one of the particles arriving slightly ahead of the other; if collisions of this type occur simultaneously at multiple sites, then the probabilities are assigned independently at each site. 1.2. Markov jump processes with very degenerate transitions. The process described in Section 1.1 is a Markov jump process. One may represent such a process as a unit speed flow built over the “jump section”. Let M+ denote the set of configurations X immediately following a collision. For each X ∈ M+ , let r (X ) denote the time to the next collision, and let M := {(X, s) : X ∈ M+ and s ∈ [0, r (X )]}. Transition probabilities P t for the Markov process on M are defined as follows: t – for t < r (X ), P(X,0) = δ(X,t) , the point mass at (X, t); t – for t = r (X ), a jump occurs, and P(X,0) is the probability measure supported on + M × {0} given by the rules in the last subsection.
Geometrically, each point in M+ × {0} moves “up” in the second coordinate at unit speed until it reaches the pre-jump section {(X, s) : s = r (X )}, at which time its image is identified, instantaneously, with a distribution on M+ × {0}. This distribution is a point mass a vast majority of the time. It has a density in one direction if the collision involves exactly one of the baths, in two directions if both baths are involved, and it is supported on a finite set of points following a same-site double collision. Transition probabilities at other points in M are defined in the obvious way. The continuous-time Markov process above induces a discrete-time Markov chain on M+ defined as follows: For X ∈ M+ , the (time-1) transition probability starting r (X ) from X , denoted p X , is given by p X (A) = P(X,0) (A × {0}) for all A ⊂ M+ . A Borel probability measure µ on M+ is said to be invariant under the discrete-time Markov chain p n if for all Borel sets A ⊂ M+ , µ(A) = p X (A) dµ(X ). Invariant measures for the continuous-time process P t are defined similarly. The invariant measures of these two processes are related via the following elementary lemma, the proof of which we omit. Let λk denote Lebesgue measure on Rk . n Lemma 1.1. (a) For every Borel probability measure µ on M+ invariant under p , µˆ := (µ × λ1 )|M is invariant under P t ; µ(M) ˆ < ∞ if and only if r dµ < ∞. (b) Conversely, each P t -invariant probability measure µˆ gives rise to a p n -invariant measure µ defined by
µ(A) = lim
ε→0
1 µ({(X, ˆ s) : X ∈ A, s ∈ [0, ε]}. ε
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
203
1.3. The roaming particles viewpoint. In Section 1.1, we have described our model as consisting of N + 1 moving particles each confined to a unit interval, exchanging energy with the particles confined to sites when they meet. The following entirely equivalent way of viewing this model is sometimes convenient and is used in many of the arguments: Think of each particle as carrying an energy which remains unchanged throughout, i.e., there are no energy exchanges. Instead, the particles exchange positions when they meet. More precisely, a particle may be moving or at rest; when at rest, it is at one of the sites, and is the sole occupant of that site. When a moving particle reaches site i, it exchanges positions with the particle sitting at site i, and waits there until it is relieved by the next particle to reach this site. Each particle is permitted to roam the length of the chain until it exits at one of the two ends. Formally, to define an equivalence with the model described in Section 1.1, one should view all particles carrying the same energy as indistinguishable. Notice that this does not prevent one from following, with no ambiguity, the trajectory of a particle from the moment it enters the chain to when it leaves – except in the case of same-site double collisions involving 3 particles with the same energy. In such a collision, one cannot deduce from the subsequent dynamics which one of the approaching particles is assumed to have arrived first. In this roaming particles description of the model, there is also the induced discretetime process, which is defined in the obvious way. Part I. Rigorous Results 2. Statement of Results Our results pertain to invariant measures of the processes defined in Section 1. When the bath distributions L(ξ ) and R(ξ ) are equal, we regard the system as being in equilibrium, and view its invariant measures as equilibrium distributions. When L(ξ ) = R(ξ ), we view the system as forced out of equilibrium, and its invariant measures as nonequilibrium steady states. Our first result applies to situations both in and out of equilibrium. As noted in the introduction, the closed chain corresponding to our model is integrable and highly nonergodic, so that ergodicity of the open system, if true, is brought about solely by the randomness from the baths. Intuitively, the more spread out the injected energies, the more likely ergodicity will be achieved. Theorem 1. Assume L(ξ ), R(ξ ) > 0 for all ξ ∈ (0, ∞). Then the following hold for all N -chains for both the continuous-time process P t and its associated discrete-time process p n : (a) There is at most one invariant probability measure (which is therefore ergodic). (b) This measure has a density with respect to Lebesgue measure. The conditions above can be weakened as follows: There is a function φ(N ) = O( N12 ) such that (a) and (b) hold for the N -chain if L(ξ ) and R(ξ ) are strictly positive on open sets I L and I R , and there exist ξ ∈ I L ∪ I R and ξ ∈ I L ∩ I R such that ξ = φ(N )ξ . 1 . The form of our assumption An example of φ(N ) that is sufficient is φ(N ) = (18N )2 involving φ(N ) is quite possibly an artifact of our proof, but it is essential for L(ξ ) and R(ξ ) to have some spread. Our results imply, in particular, that the processes we consider have no singular invariant probability measures.
204
P. Balint, K. K. Lin, L.-S. Young
We do not prove the existence of invariant measures in this paper. For existence, the two issues are: (1) tightness, which requires that one controls the dynamics of super-fast and super-slow particles. (2) Discontinuities of transition probabilities p X as a function of X due to same-site double collisions. Techniques for treating (1) in nonequilibrium situations are generally lacking, and (2) is likely to involve very technical arguments. We have elected to leave these problems for future work. Our next result gives an explicit formula for the equilibrium distribution in terms of bath distributions. In general, no such explicit expressions exist for nonequilibrium steady states, and our models appear to be no exception. Theorem 2. Let L(ξ ) = R(ξ ) = ρ(ξ ) be a probability density on (0, ∞) satisfying (i) 1 √ ρ(ξ )dξ < ∞ and (ii) ρ(ξ ) = O(ξ −κ ) for some κ > 3/2 as ξ → ∞. Then the ξ following measure is invariant for the (continuous-time) process defined in Section 1.1: N N 1 dξ1 · · · dξ N · ρ(ξi ) × dη0 · · · dη N · √ ρ(ηi ) × λ N +1 × ω N +1 , ηi i=1
where λ N +1 = Lebesgue measure on configurations in {(σ1 , · · · , σ N +1 )}.
i=0
[0, 1] N +1
and ω N +1 assigns equal weight to the
The following corollary suggests some examples to which our results apply: Corollary 1. Any finite chain with bath injections L(ξ ) = R(ξ ) given by either (a) an exponential distribution or (b) a Gaussian truncated at 0 and normalized has a unique (hence ergodic) equilibrium distribution, and it is given by the expression in Theorem 2. 3. Density and Ergodicity: Outline of Proof For definiteness, we will work with the discrete-time Markov chain on M+ . By Lemma 1.1, the assertions in Theorem 1 for p n imply the corresponding assertions for P t . Proving uniqueness of invariant measures or ergodicity requires, roughly speaking, that we be able to steer a trajectory from one location of the phase space to another. If the transition probabilities have densities on open sets, then one needs only to do so in an approximate way. Our model, unfortunately, has highly degenerate transition probabilities. One must, therefore, tackle hand in hand the problems of (i) acquisition of densities for p nX and (ii) steering of trajectories. The purpose of this section is to outline how we plan to do this. Section 3.1 introduces definitions and ideas that will be used. In Section 3.2, we formulate three propositions to which the proof of Theorem 1 will be reduced. 3.1. Basic ingredients of the proof. (A) Acquiring densities. Given a finite Borel measure ν on M+ , we let ν = ν⊥ + νac denote the decomposition of ν into a singular and an absolutely continuous part with respect to Lebesgue measure, and say ν has an absolutely continuous component when νac = 0. We say X ∈ M+ eventually acquires a density if for some n > 0, p nX has an absolutely continuous component, i.e. ( p nX )ac = 0.
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
205
Lemma 3.1. If every X ∈ M+ acquires a density eventually, then every invariant probability measure of p n has a density. Proof. For any measure ν, if ν is absolutely continuous with respect to Lebesgue, then so is νn := p nX dν(X ). This is because under the dynamics, Lebesgue measure is carried to a measure equivalent to Lebesgue, followed, at certain steps, by a diffusion in one or two directions corresponding to bath injections. Thus (νn )ac (M+ ) ≥ νac (M+ ) for all n ≥ 1. The hypothesis of this lemma implies that this inequality is strict for some n unless ν⊥ = 0. Now let µ be an invariant probability measure for p n . Since µn = µ, it follows that µ⊥ = 0.
Recall that I L = {L > 0} and I R = {R > 0}. We call a finite or infinite sequence of points X 0 , X 1 , . . . in M+ a sample path if such a sequence can, in principle, occur. In particular, if X n−1 is followed by a collision with the right bath, then the ξ N -coordinate of X n must lie in I R , and similarly with the left bath. Let X = (X 0 , X 1 , . . . , X n ) be a sample path obtained by injecting into the system the energy sequence ε = (ε1 , . . . , εm ) in the order shown. (One does not specify whether the injection is from the left or the right; it is forced by the sequence.) We write
X 0 (ε) = X n . Suppose X has no multiple collisions. It is easy to see that there is a small neighborhood E of ε in Rm such that for all ε ∈ E, (i) ε is a feasible sequence of injections, i.e., if the sequence dictates that εj be injected from the left (right) bath, then εj ∈ I L (resp. I R ), and (ii) the sample path X = (X 0 , X 1 , . . . , X n ) produced by injecting ε has the exact same sequence of collisions as X . In particular, it also has no multiple collisions. We may therefore extend = X 0 to a mapping from E to M+ with (ε ) = X n . As such, is clearly continuous. Lemma 3.2. : E → M+ is continuously differentiable. In the proof below, it will be useful to adopt the viewpoint expressed in Section 1.3, i.e. to track the movements of the injected energies through time. Notice that for a sample path with no multiple collisions, there is no ambiguity whatsoever about the trajectory of an injected energy. ∂ xi i ∂ξi Proof. We verify that ∂η δε j , δε j and δε j exist and are continuous on E. For ηi and ξi , it is easy: either ε j is carried by the particle in question, or it is not. If it is, then the partial derivative is = 1; if not, then it is = 0. For xi , consider first the case where ηi carries the injection εk , k = j. We let p be the total number of times ε j crosses the interval [i, i + 1] (in either direction) before X n . ∂ xi = 0. Suppose not. Then perturbing (only) ε j to ε j + δε, the time If p = 0, then δε j
gained per crossing (with a sign) will be ∂ xi √ = lim p εk · δε→0 δε j
√1 εj
√1 εj
−√
−√
1 . ε j +δε
1 ε j +δε
δε
The case where ε j is carried by ηi is left to the reader.
Consequently
=
p 21 − 32 ε ε . 2 k j
In the setting above, if D (ε) is onto as a linear map, then by the implicit function theorem, there is an open set E ⊂ E with ε ∈ E such that carries Lebesgue measure on E to a measure in M+ with a strictly positive density on a neighborhood of (ε). This implies in particular ( p nX 0 )ac > 0. We summarize as follows:
206
P. Balint, K. K. Lin, L.-S. Young
Definition 3.1. We say X ∈ M+ has full rank2 if there is a sequence of injections ε = (ε1 , · · · , εm ) leading to a sample path X = (X 0 , X 1 , . . . , X n ) with X 0 = X such that (i) X has no multiple collisions, and (ii) D X 0 (ε) is onto. Corollary 3.1. If X ∈ M+ has full rank and X is as above, then p nX has strictly positive densities on an open set containing X n . Having full rank is obviously an open condition, meaning if X has full rank, then so does Y for all Y sufficiently close to X . (B) Ergodic components. One way to force two points to be in the same ergodic component (in the sense to be made precise) is to show that they have “overlapping futures”. This motivates the following relation: Definition 3.2. For X, Y ∈ M+ , we write X ∼ Y if there exist a positive Lebesgue n measure set A = A(X, Y ) ⊂ M+ and m, n ∈ Z+ such that ( p m X )ac |A and ( pY )ac |A have strictly positive densities. Here is how this condition will be used: Suppose µ and ν are ergodic measures, and there is a positive µ-measure set Aµ and a positive ν-measure set Aν such that X ∼ Y for all X ∈ Aµ and Y ∈ Aν . Then the ergodic theorem tells us that µ = ν since they have the same ergodic averages along positive measure sets of sample paths. As an immediate corollary of the ideas in Part (A), we obtain Corollary 3.2. If X ∈ M+ has full rank, then there is a neighborhood N of X such that Y ∼ Z for all Y, Z ∈ N . (C) Constant energy configurations. For fixed e ∈ (0, ∞), let Qe = {X ∈ M+ : ηi = ξi = e for all i}. If we start from X ∈ Qe and inject only energies having value e, then the resulting sample path(s) will remain in the set Qe . (Obviously, this is feasible only if e ∈ I L ∩ I R .) We call these constant-energy sample paths, and say X ∈ Qe has no multiple collisions if that is true of its constant-energy sample path. Needless to say, constant-energy sample paths occur with probability zero. However, they have very simple dynamics, and perturbations are relatively easy to control. Our plan is to exploit these facts by driving all sample paths to some Qe and to work from there. 3.2. Intermediate propositions. We claim that Theorem 1 follows readily from the following two propositions: Proposition 3.1. Every X ∈ M+ eventually acquires a density. Proposition 3.2. X ∼ Y for a.e. X, Y ∈ M+ with respect to Lebesgue measure. We remark that to rule out the presence of singular invariant measures, we need Proposition 3.1 to hold for every X , not just almost every X . 2 This property has the flavor of Hörmander’s condition for hypoellipticity for SDEs – in a setting that is largely deterministic and has discontinuities.
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
207
Proof of Theorem 1 assuming Propositions 3.1 and 3.2. Proposition 3.1 and Lemma 3.1 together imply that every invariant probability measure has a density, proving part (b). To prove part (a), let µ and ν be ergodic measures. Since they have densities, Proposition 3.2 implies, as noted in Paragraph (B), that µ = ν.
Next we identify three (more concrete) conditions that will imply Propositions 3.1 and 3.2. The following shorthand is convenient: For X, Y ∈ M+ , we write X ⇒ Y if given any neighborhood N of Y , there exists n such that p nX (N ) > 0, X Y if given any neighborhood N of Y , there is a neighborhood N of X such that X ⇒ Y for all X ∈ N . Proposition 3.3. Given X ∈ M+ and e ∈ I L ∩ I R , there exists Z = Z (X ) ∈ Qe such that X ⇒ Z . Proposition 3.4. Z Z for all Z , Z ∈ Qe , any e ∈ I L ∩ I R . Proposition 3.5. Every Z ∈ Qe with no multiple collisions has full rank. Proof of Propositions 3.1–3.2 assuming Propositions 3.3–3.5. To prove Proposition 3.1, we concatenate Propositions 3.3 and 3.4 to show that for every X ∈ M+ , X ⇒ Z for some Z = Z (X ) ∈ Qe with no multiple collisions. The assertion is proved if there is a neighborhood N Z of Z such that every Z ∈ N Z eventually acquires a density. This follows from Proposition 3.5, Corollary 3.1 and the fact that full rank is an open condition. To prove Proposition 3.2, it is necessary to produce a common Z = Z (X, Y ) ∈ Qe with no multiple collisions such that X, Y ⇒ Z . To this end, we first apply Proposition 3.3 to produce Z (X ) ∈ Qe with X ⇒ Z (X ) and Z (Y ) ∈ Qe with Y ⇒ Z (Y ). We then fix an (arbitrary) Z ∈ Qe with no multiple collisions. Since Z (X ), Z (Y ) Z (Proposition 3.4), we have X, Y ⇒ Z as before. Proposition 3.5 and Corollary 3.2 then give the desired result.
Proving Theorem 1 has thus been reduced to proving Propositions 3.3–3.5. 4. Proofs of Propositions 3.3–3.5 In Sections 4.1–4.2, we give an algorithm for driving sample paths from given initial conditions to constant-energy surfaces. This can be viewed as changing the energies in a configuration. In Sections 4.3–4.4, we focus on changing the relative positions of the moving particles. 4.1. Sample paths from X to Qe (“typical” initial conditions). The hypothesis of Theorem 1 guarantees the existence of some ξ ∈ I L ∩ I R . We may assume, without loss of generality, that ξ = 1. Given X ∈ M+ , we seek sample paths that lead to the constant energy surface Q1 . Consider first the following specific question: By injecting only particles with energy 1, will we eventually “flush out” all the energies in X , replacing them with new particles having energy 1? To study this question, we propose to suppress some information, to focus on the following evolution of arrays of energies: To each X ∈ M+ , we associate the array E(X ) = (η0 , ξ1 , η1 , ξ2 , . . . , ξ N , η N ), where ηi and ξi are the energies of X and they are arranged in the order shown. Then corresponding to each sample path X = (X 0 , X 1 , . . .)
208
P. Balint, K. K. Lin, L.-S. Young
is the sequence of moves E(X 0 ) → E(X 1 ) → E(X 2 ) → · · ·. For example, collision between site i and the particle on its right corresponds to swapping the (2i + 1)st entry in the array with the (2i)th ; the rightmost particle exiting the system and an energy of value ξ R entering corresponds to replacing the (2N + 1)st entry by ξ R , and so on. For as long as there are no same-site double collisions, we can trace the movements of energies in (E(X 0 ), E(X 1 ), · · · ) as discussed in Section 1.3. We define the exit time of the j th entry in E(X 0 ) to be the number of moves before this energy exits the system, and define T (X ) to be the last exit time of all the elements in E(X 0 ). A priori, T (X ) ≤ ∞. Lemma 4.1. Given X 0 ∈ M+ , suppose that injecting all 1s gives rise to a sample path X with no same-site double collisions. Then T (X ) < ∞. At each step, we will refer to the original energies of X 0 that remain as the “old energies”, and the ones that are injected as “new energies”. In particular, all new energies have value 1 (some old energies may also have value 1). The lemma asserts that in finite time, the number of old energies remaining will decrease to 0. Proof. For each n, we let E(X n ) = (η0 , ξ1 , η1 , ξ2 , · · · , ξ N , η N ), and say X n is in state (·, i) if the leftmost old energy is ξi , (+, i) if the leftmost old energy is ηi and σi = +, (−, i) if the leftmost old energy is ηi and σi = −. The terminology above is not intended to suggest that we are working with a reduced system. No such system is defined; nevertheless it makes sense to discuss which transitions among these states are permissible in the underlying Markov dynamics. Notice first that independent of the state of a system, it will change eventually. This is because all energies are nonzero, so a collision involving the leftmost old energy is guaranteed to occur at some point. We list below all the transitions between states that are feasible, skipping over (many) steps in the Markov chain that do not involve the leftmost old energy: (1) Suppose the system is in state (−, i). If i = 0, then the only possible transition is (−, i) → (·, i). If i = 0, then the leftmost old energy exits the system, and the new state is determined by the next leftmost old energy (if one remains). (2) Suppose the system is in state (+, i). If i = N , then T (X ) is reached as this last remaining old energy exits the system. If not, we claim the only two possibilities are (+, i) → (·, i + 1) and (+, i) → (−, i). The first case corresponds to the energy originally at site i + 1 being new, the second case old. (3) Finally, consider the case where the system is in state (·, i). If the next collision is with the particle from the left, then (·, i) → (−, i − 1). If it is with the particle from the right, then there are two possibilities corresponding to the approaching energy being new or old, namely (·, i) → (+, i) or (·, i) → (·, i). Notice that the system cannot remain in state (·, i) forever, since the particle from the left will arrive sooner or later, causing the state to change. For this reason, let us agree not to count (·, i) → (·, i) as a transition. To summarize, the only possible transitions are (−, i) → (·, i), (·, i) → (−, i − 1), (+, i), (+, i) → (·, i + 1), (−, i) except where the transition leads to an old energy exiting the system. When that happens, either there is no old energy left, or the system can start again in any state. The following observation is crucial:
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
209
Sublemma 4.1. The transition sequence (−, i) → (·, i) → (+, i) is forbidden. We first complete the proof of Lemma 4.1 assuming the result in this sublemma: Observe first that if one starts from (+, i), either the evolution is (+, i) → (·, i + 1) → (+, i + 1) → (·, i + 2) → · · · leading to an exit at the right, or a state of the form (−, j) is reached; a similar assertion holds if one starts from (·, i). On the other hand, starting from (−, i), the only possible sequence permitted by the sublemma is (−, i) → (·, i) → (−, i − 1) → (·, i − 1) → · · · leading to an exit at the left. This proves that in finite time (meaning a finite number of steps with respect to the Markov chain p n ), the number of old energies remaining will decrease by one.
Proof of Sublemma. Suppose the transition (−, i) → (·, i) takes place at t = 0. Since all the energies to the left of site i are new and therefore have speed one, a particle from the left is guaranteed to reach site i at time t = t0 < 2, since the previous collision with the site from the left took place strictly before t = 0 (no same-site double collisions). We will argue that when this particle arrives, it will find an old energy at site i, that in fact the state of the system has not changed between t = 0 and t0 . Thus the next transition has to be (·, i) → (−, i − 1). Here are the events that may transpire between t = 0 and t0 on the segment [i, i + 1]: Observe that the transition at t = 0 is necessarily between an old and a new energy (otherwise the state at t = 0− could not be (−, i)). This will result in a new energy leaving site i for site i + 1 at t = 0+ , and arriving at t = 1. The only way (·, i) → (+, i) can happen is for a new energy to move leftward on [i, i + 1], and to arrive at site i before time t0 . This cannot happen, since new energies travel at unit speed. (Notice that the old energy at site i may change between t = 1 and t0 , but the state of the system does not.)
Partial proof of Proposition 3.3. We prove the result for X = X 0 under the additional assumption that injecting 1s gives rise to a sample path with no same-site double collisions. By Lemma 4.1, there is a sample path X = (X 0 , . . . , X n ) with X n ∈ Q1 . Let N be a neighborhood of X n . If X has no multiple collisions, then all nearby sample paths will end in N as discussed in Section 3.1. If X has multiple (but not same-site) collisions,3 then injecting a slightly perturbed energy sequence may – is likely to, in fact – desynchronize the simultaneous jumps. If, for example, two collisions at sites i and i occur simultaneously at step j, then for perturbed injected energies, X j may be replaced by X j and X j corresponding to two collisions that happen in quick succession. This aside, the situation is similar to that with no multiple collisions, and we still have p kX (N ) > 0 but possibly for some k > n.
4.2. Sample paths from X to Qe (“exceptional” cases). As noted in Section 3.2, to rule out the existence of singular invariant measures, it is necessary to show that every X ∈ M+ eventually acquires a density. Our strategy is to inject energy 1s into the system and argue that this produces a sample path X that leads to a point Z ∈ Q1 . To prove X ⇒ Z , however, requires more than that: it requires that a positive measure set of sample paths starting from X follow X . This is not a problem if X has no same-site double collisions, for in the absence of such collisions, the dynamics are essentially continuous (as explained above). Yet it is unavoidable that for some X ∈ M+ , injecting 1s will lead 3 These points are discontinuities only for the (discrete-time) Markov chain, affecting the number of steps. They are not discontinuities at all for the continuous-time jump process.
210
P. Balint, K. K. Lin, L.-S. Young
to same-site double collisions. It is an exceptional situation, but one that we must deal with if we are to follow the same route of proof. We now address the potential problems. Recall that at each same-site double collision, there are two ways to continue the sample path. Partitioning nearby sample paths according to itineraries. Let X = (X 0 , X 1 , . . . , X n ) be a sample path obtained by injecting a sequence of m energies all of which are 1s and making a specific choice at each same-site double collision. As in Section 3.1(A), we consider energies ε = (ε1 , . . . , εm ) near 1 = (1, . . . , 1), and ask the following question: Which injection sequence ε will give rise to a sample path X (ε) that has the same collision sequence as X ? To focus on the issues at hand, we will ignore multiple collisions that do not involve same-site double collisions, for they are harmless as explained earlier. The computation in Lemma 3.2 motivates the following coordinate change, which is not essential but simplifies the notation: Let ψ(ε) = √1ε − 1, and let (ε1 , . . . , εm ) = (ψ(ε1 ), . . . , ψ(εm )). Then maps a neighborhood of 1 diffeomorphically onto a neighborhood of 0 = (0, . . . , 0). We assume that X (ε) has the same itinerary as X through step i − 1, and that at step i, X has a double collision at site j. To determine whether the left or right particle will arrive first at site j for X (ε) , we use X as the point of reference, and let t j (ε) and t j−1 (ε) be the times gained (with a sign) by the particles approaching site j from the right and left respectively. Then t j (ε) = p1 ψ(ε1 ) + · · · + pm ψ(εm ) and t j−1 (ε) = q1 ψ(ε1 ) + · · · + qm ψ(εm ), where pk and qk are the numbers of times – up to and including the approach to site j – the injected energy εk has passed through the intervals [ j, j + 1] and [ j − 1, j] respectively. Thus for X (ε) , the right particle arrives first if and only if t j (ε) > t j−1 (ε). Thus given X = (X 0 , X 1 , . . . , X n ) as above, there is a decreasing sequence of subsets Rm = V0 ⊃ V1 ⊃ V2 ⊃ · · · ⊃ Vn defined as follows: Vi = Vi−1 except where a same-site double collision occurs; when a same-site double collision occurs, say at site j, we let Vi = Vi−1 ∩ H , where H = {ε ∈ Rm : t j (ε) ≥ t j−1 (ε)} or {t j (ε) ≤ t j−1 (ε)} depending on whether we have chosen to let the right or left particle arrive first in X . If there are multiple same-site double collisions, then we intersect with the half-spaces corresponding to all of them. The sequence of Vi so obtained will have the property that all ε close enough to 1 with (ε) ∈ Vn give sample paths that have the same collision sequence as X . In particular, if Vn is nontrivial, meaning it has interior, then the set of ε for which X (ε) shadows X will have positive measure. That is to say, X 0 ⇒ X n . In general, there is no guarantee that Vn is nontrivial. Supposing Vn for a sample path is nontrivial, we say the choices made at step n + 1 are viable if they lead to a nontrivial Vn+1 . Observe that inductively, a viable choice can be made at each step, since one of the two half-spaces must intersect nontrivially Vn from the previous step. We first treat a special situation involving the same-site collision of three energies all of which are 1s. The setting is as above. Sublemma 4.2. Suppose Vn is nontrivial for a sample path X , and at step n + 1, the energy εk , which is = 1, is involved in a same-site double collision with two other energies both of which are 1s. Then the choice to have εk arrive first is viable if εk has the following history:
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
211
(i) its movement up to the collision above is monotonically from left to right; (ii) all of its previous same-site double collisions are with two other energy 1s. Proof. By (i), we have that in the collision at step n + 1, pk = 0 and qk = 1. Choosing to have εk arrive first corresponds to choosing the half-space {(εk ) ≥ G(ε)}, where G(ε) is a linear combination of (εi ), i = k. By (ii), all the other half-spaces in the definition of Vn either do not involve εk or are of the same type as the one above.
Completing the proof of Proposition 3.3. Given X 0 ∈ M+ , we now inject 1s and make a choice at every same-site double collision with the aim of obtaining a sample path X 0 , X 1 , . . . , X n with X n ∈ Q1 and X 0 ⇒ X n . Our plan is to follow the scheme in Section 4.1 to reach Q1 , and to make viable choices along the way. The proof of Lemma 4.1 is based on the following observation: the leftmost old energy either moves right monotonically until it exits, or it turns around and starts to go left, and by Sublemma 4.1, once that happens it must move left monotonically until it exits. Revisiting the arguments, we see that with the exception of Sublemma 4.1, all statements in the proof of Lemma 4.1 hold for any sample path, with whatever choices are made at double collisions. Moreover, the only way Sublemma 4.1 can fail is that at t = 0, a double collision occurs at site i, we choose to have the left particle arrive first, and this is followed by another double collision at t = 2 at which time we choose to have the right particle arrive first. It is only through these “two bad decisions” that the leftmost old energy can turn around and head right again. The choices at same-site double collisions are arbitrary provided the following two conditions are met: (i) The choice must be viable. (ii) In the setting of Sublemma 4.2, we choose to have εk arrive first. Following these rules, it suffices to produce X n ∈ Q1 for some n; the nontriviality of Vn follows from (i). Suppose, to derive a contradiction, that no X n ever reaches Q1 , i.e., a set of old energies is trapped in the chain forever. Let i 0 be the leftmost site that it visits infinitely often. That means from some step n 0 on, the leftmost energy never ventures to the left of site i 0 , but it returns to site i 0 infinitely many times, each time repeating the “two bad decisions” scenario above. The situation to the left of site i 0 is as follows: From time n 0 on, all the energies strictly to the left of this site have speed one. Thus by rule (ii), after a new energy enters from the left bath, it will march monotonically to the right until at least site i 0 , whether or not it is involved in any same-site double collisions. To complete the proof, consider a moment after time n 0 when the leftmost old energy arrives at site i 0 . From the discussion above we know it has to be involved in a samesite double collision, and that the energy approaching from the left is new. By (ii), we choose to have the new energy arrive first, making the first “bad decision”. But then in the collision at the same site 2 units of time later, again we choose to have the energy from the left arrive first, causing the old energy to move to site i 0 − 1 and contradicting the definition of site i 0 .
4.3. Moving about on constant energy surfaces. This subsection focuses on getting from Z ∈ Qe to Z ∈ Qe for e ∈ I L ∩ I R . In general, this cannot be accomplished by considering sample paths that lie on Qe alone, since such sample paths are periodic as we will see momentarily. Instead we will make controlled excursions from Qe aimed at returning to specific target points. As before, we assume 1 ∈ I L ∩ I R and work with Q1 .
212
P. Balint, K. K. Lin, L.-S. Young
Circular tracks between sites. It is often useful to represent the pair (xi , σi ) by a single coordinate z i and to view the particle as making laps in a circle of length 2. More precisely, if we represent this circle as [0, 2]/ ∼ with end points identified, then z i = xi for σi = + and z i = 2 − xi when σi = −. A continuous-time sample path Z (t), t ≥ 0, for the process described in Section 1 then gives rise to a curve Z ∗ (t) = (z 0 (t), z 1 (t), . . . , z N (t)) ∈ T N +1 . (We write T N +1 even though each factor has length 2.) For each i, z i (t) goes around the circle in uniform motion, changing speed only at z i = 0 and 1. Accordingly, we denote by ∂(T N +1 ) the set of points such that z i = 0 or z i = 1 for (at least) one i. Asterisks will be used to signify the use of circular-track notation: A discrete-time sample path Z = (Z 0 , Z 1 , · · · ) in the notation of previous sections corresponds to Z ∗ = (Z 0∗ , Z 1∗ , · · · ) with Z i∗ ∈ ∂(T N +1 ). Returns of Z to M+0 = M+ ∩ {x0 = 0} translates into returns of Z ∗ to ∂(T N +1 ) ∩ {z 0 = 0}. We now identify two features of constant energy configurations that make them easy to work with, beginning with the simplest case: Lemma 4.2. Suppose we start from Z ∈ Q1 with no multiple collisions. (a) By injecting all 1s, the sample path returns to Z in 2N + 2 steps. (b) If all the injected energies are 1s except possibly for one, which we call e, and if we assume the resulting sample path has no same-site double collisions, then the energy e moves monotonically through the system until it exits, after which the system rejoins the periodic sample path in (a). Proof. (a) This is especially easy to see in continuous time and in circular-track notation: each z i (t) is periodic with period 2. Thus the continuous-time sample path Z (t) is periodic with period 2, and in every two units of time, there are exactly 2N +2 collisions. (b) Suppose the energy e has just arrived at site j from the left. Then the next collision at site j is with the particle from the right, because all cycles take 2 units of time to complete and the right one has a (strict) headstart. This proves the monotonic movement of the energy e. After it leaves the system, the relative positions between z i and z j remain unchanged for all i, j, since the passage of the energy e through the chain leads to identical time gains for z i and z j .
Another nice property of constant energy sample paths is that the process is continuous along these paths. We make precise this idea: Definition 4.1. Given a sample path X = (X 0 , X 1 , . . . , X n ), we say the process is continuous at X if the following holds: Let ε = (ε1 , . . . , εm ) be the sequence in injections. Then given any neighborhood N of X n , there are neighborhoods N0 of X 0 and U of ε such that for all X 0 ∈ N0 and ε ∈ U, all the sample paths generated follow X into N . (We do not require that it reaches N in exactly n steps; it is likely to take more than n steps if X has multiple collisions.) Lemma 4.3. Let X = (X 0 , . . . , X n ) be a sample path. (a) If X either has no same-site double collisions or every such collision involves three identical energies, then the process is continuous at X . (b) If the process is continuous at X , then X 0 X n .
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
213
The ideas behind part (a) are the same as those discussed in Sections 4.1 and 4.2 and will not be repeated here. Part (b) is immediate, for “” requires less than continuity. The following lemma is the analog of Lemma 4.2(b) without the assumption of “no multiple collisions”: Lemma 4.4. Given any Z ∈ Q1 , let all the energies injected be 1s except possibly for one which has value e. Then (i) there is a sample path in which the energy e moves through the system monotonically; (ii) except for a discrete set of values e, it will do so without being involved in any same-site double collisions. Proof. The monotonic motion of e is clear if it is not involved in any same-site double collisions. If it is, choose to have it arrive first every time. This proves (i). To prove (ii), let J be a finite interval of possible energies to be injected at t = 0, and suppose e is on its way from site 0 to site 1. In circular-tracks notation, z 1 is periodic with period 2, and except for a finite subset of e ∈ J , z 0 = 1 at exactly the same time that z 1 = 0. We avoid this “bad set” of e, ensuring that the special energy will not be in a double collision at site 1. Now this energy leaves site 1 exactly when z 1 = 0, and at such a moment, the configuration to the right of site 1 is identical (independent of the value of e in use). The same argument as before says that to avoid a double collision at site 2, another finite subset of J may have to be removed.
Injecting a single energy different than 1 will not produce the desired positional variations. We now show that injecting a second energy – both appropriately chosen and injected at appropriate times – will do the trick. The hypothesis of Theorem 1 guarantees, after scaling to put 1 ∈ I L ∩ I R , that there exists e ∈ I L ∪ I R with e = o(1/N 2 ). (Lemma 4.5 below is the only part of the proof in which this hypothesis is used.) Let us assume for definiteness that e ∈ I L , and consider initial conditions Z ∈ M+0 = M+ ∩ {x0 = 0}. Lemma 4.5. There exists a0 > 0 for which the following hold: Given a with |a| < a0 , Z 0 ∈ Q1 ∩ M+0 and k ∈ {1, . . . , N }, there is a sample path Z = (Z 0 , . . . , Z n ) with Z n ∈ Q1 ∩ M+0 such that if Z 0∗ = (0, z 1 , . . . , z N ), then Z n∗ = (0, z 1 , . . . , z k−1 , z k − a, z k+1 − a, . . . , z N − a). Moreover, the sample path Z may be chosen so that it has no same-site double collisions except where all the energies involved are 1s. Proof. First we focus on constructing a sample path that leads to the desired Z n without attempting to avoid double collisions. In the construction to follow, all the injections except for two will be 1s. At t = 0, we inject energy e = o(1/N 2 ) from the left bath, letting it move monotonically from left to right. When it is about halfway between sites k and k + 1, we inject from the left an energy e = e(a) ≈ 1 (the relation between e and a will be clarified later). The energy e also moves monotonically from left to right until it reaches the k th site, which takes O(N ) units of time. Since the slow particle takes longer than O(N ) units of time to reach site k + 1, the e-energy waiting at site k is met by a particle from the left. So it turns around and moves monotonically left, eventually exiting at the left end. As for the slow energy, it eventually reaches site k + 1, and continues its way monotonically to the right until it exits the chain. Let Z n be the first return to M+0 after the exit of e (the e-energy exits much earlier). By definition, Z n ∈ Q1 . It remains to investigate its z i -coordinates. Reasoning as in
214
P. Balint, K. K. Lin, L.-S. Young
the proof of Lemma 4.2, we see that the e -energy does not cause a shift in the relative positions of the z i because its effects on all of the z i are identical. The e-energy, on the other hand, has made one full lap (first half on its way to site k and second half on its way back) on the segments [ j, j + 1] for j = 0, 1, . . . , k − 1 but not for j ≥ k. Let 1 a = 2(1 − e− 2 ) be the (signed) time gain per lap for a particle with energy e over a particle with energy 1. Then the position of z 0 relative to z j for j < k is maintained while relative to z j for j ≥ k, z 0 is ahead by distance a, leading to the coordinates of Z n as claimed. We now see that the bound imposed on a is that e(a) must be in I L ∩ I R . This completes the proof of the lemma except for the claim regarding double collisions. From Lemma 4.4(b), we know that for e outside of a discrete set of values, the particle carrying this special energy will not be involved in any same-site double collisions. It remains to arrange for the energy e to avoid such collisions. Here the situation is different: since a is given, e is fixed, but notice that increasing e is tantamount to delaying the injection of the energy e. Thus an argument similar to that in Lemma 4.4(b) can be used to show that by avoiding a further discrete set of values for e (depending on Z 0 , e and k), the energy e will not be involved in same-site double collisions. This completes the proof.
Remark. The following two facts make this result very useful: (a) The dynamical events described in Lemma 4.5 require no pre-conditions on Z 0 other than Z 0 ∈ Q1 ∩ M+0 . (b) The range of admissible a is independent of Z 0 . Suppose a > 0, and for some j > k, the j th moving particle in Z 0 is such that x j < a and σ j = +. Then it will follow that σ j = − in Z n , since it is distance a behind its original position. This is very natural when one thinks about it in terms of circular tracks, but it is a jump in the phase space topology in Section 1.1.
4.4. Proofs of Propositions 3.4 and 3.5. Proof of Proposition 3.4. Given Z , Z ∈ Q1 , we will construct a sample path Z = (Z 0 , Z 1 , · · · , Z n ) with Z 0 = Z , Z n = Z and Z 0 Z n . Since all constant energy sample paths pass through M+0 , we may assume Z , Z ∈ M+0 . Let Z ∗ = (0, z 1 , . . . , z N ) and (Z )∗ = (0, z 1 , . . . , z N ). Our plan is to apply Lemma 4.5 as many times as needed to nudge each z i toward z i one i at a time beginning with i = 1. Write z 1 − z 1 = ja for some j ∈ Z+ and a small enough for Lemma 4.5. Applying Lemma 4.5 j times with k = 1, we produce Z n 1 ∈ Q1 ∩ M+0 with Z n∗1 = (0, z 1 , z 2 + (z 1 − z 1 ), . . . , z n + (z 1 − z 1 )). We then apply Lemma 4.5 to Z n 1 with k = 2 repeatedly to change it to Z n 2 with Z n∗2 = (0, z 1 , z 2 , z 3 + (z 2 − z 2 − (z 1 − z 1 )), . . . , z n + (z 2 − z 2 − (z 1 − z 1 ))), and so on until Z n = Z n N is reached. Notice that the sample path Z so obtained has the property that it has no same-site double collisions other than those that involve three identical energies (all of which 1s). Lemma 4.3 says that the dynamics are continuous at such sample paths. Hence Z Z .
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
215
Proof of Proposition 3.5. Given Z ∈ Q1 with no multiple collisions, our aim here is to produce a sample path Z 0 = Z , Z 1 , . . . , Z n , also free of multiple collisions, by injecting a sequence of energies ε = (ε1 , . . . , εm ) so that in the notation of Section 3.1, the map
has the property that D (ε) is onto. We first give the algorithm, with explanations to follow: (1) We inject 1s until n 0 , the first time Z n 0 ∈ M+0 . (2) At step n 0 , we inject into the system an energy e = o(1/N 2 ) chosen to avoid multiple collisions, followed by all 1s, until the first time the sample path returns to Q1 ∩ M+0 ; call this step n 1 . (3) By Lemma 4.1, T (Z n 1 ) < ∞. Let n be the smallest integer ≥ n 1 + T (Z n 1 ) for which Z n ∈ M+0 . We now prove that D (ε) has full rank: Let δηi denote an infinitesimal displacement in the ηi variable at Z n . To see that δηi is in the range of D (ε), we trace this energy backwards (in the sense of Section 4.1) to locate its point of origin. (3) above ensures that it was injected at step j for some n 1 < j ≤ n. Varying ε j , therefore, leads directly to variations in ηi . The same argument applies to δξi . Next we consider displacements in xi . Assume for definiteness that e enters the system from the left. We fix k ∈ {1, . . . , N }, and let ε j , j = j (k), be one of the energies injected from the left when e is about halfway between sites k and k + 1. By Lemma 4.5, varying ε j leads to a variation of the form δk = δxk + . . . + δx N in Z n 1 . By Lemma 4.2, this displacement is retained between Z n 1 and Z n since only 1s are injected. The vectors {δk , k = 1, 2, . . . , N } span the subspace corresponding to positional variations. The proof of Proposition 3.5 is complete.
5. Equilibrium Distributions 5.1. Systems with a single site. This subsection treats exclusively the case N = 1. As we will see, nearly all of the ideas in a complete proof of Theorem 2 (for general N ) show up already in this very simple situation. Following the notation in Section 1.1, we consider phase variables X = (η0 , ξ, η1 ; x0 , x1 ; σ0 , σ1 ) ∈ (0, ∞)3 × [0, 1]2 × {−, +}2 ; the phase space has 4 components corresponding to the 4 elements of {−, +}2 . Define ∞ 1 ρ(η) ρ(η) , where Z = τ (η) = √ √ dη . Z η η 0 Theorem 2 asserts the invariance of the probability measure µ = τ (η0 )ρ(ξ )τ (η1 )I[0,1]2 (x0 , x1 ) dη0 dξ dη1 d x0 d x1 × w4 , where I(·) is the indicator function and w4 gives equal weight to the 4 points. Let µt = PXt dµ(X ) be the distribution at time t with initial distribution µ. To prove Theorem 2, we need to show µt = µ for all t > 0, equivalently for all small enough t > 0. We say a system undergoes a simple change in configuration between times 0 and t if during this period at most one collision occurs and this collision involves only a single
216
P. Balint, K. K. Lin, L.-S. Young
moving particle. We will show in a sense to be made precise that for any system (independent of size), in short enough time intervals it suffices to consider simple changes in configuration. This observation simplifies the proof for N = 1. It is crucial in the analysis of N -chains; without it, the complexity of the situation gets out of hand quickly. In stochastic processes, analogous ideas are provided by independence and exponential clocks. The dynamics being largely deterministic here, we will have to argue it “by hand”, deducing it from tail assumptions on ρ. In what follows, |ν| denotes the total variation norm of a finite signed measure ν, and µ = ν + O(t) means |µ − ν| = O(t). For notational simplicity, we assume ρ(η) = O(η−2 ) as η → ∞. Proof of Theorem 2 for the case N = 1. Step 1. Reduction to simple changes in configuration. We will prove |µt − µ| = O(t 1+δ ) for some δ > 0 as t → 0. To see that this implies the invariance of µ, fix s > 0. By a repeated application of the estimate above with step size ns , we obtain µs = µ + n · O(( ns )1+δ ), which tends to µ as n → ∞. In what follows, we focus on the component M (+,+) = {(σ0 , σ1 ) = (+, +)}; other components are analyzed similarly. Instead of comparing µ and µt directly on all of M (+,+) , we will compare µ with some µˆ t ≈ µt (to be defined) on a large subset t ⊂ M (+,+) (to be defined). (A) We consider in the place of µt the measure µˆ t = PXt d µ(X ˆ ), where µˆ is the restriction of µ to the set {η0 , ξ, η1 < t −2 }. Notice that if all particles in a system have energy < t −2 , then their speeds are < t −1 , so no moving particle colliding with a site or a bath can set off a second collision to take place within t units of time. We claim that confusing µt with µˆ t leads to an error of a size we can tolerate, since ∞ ∞ ρ(ξ )dξ = O(t 2 ) and τ (η)dη = O(t 3 ). t −2
t −2
(B) Next, we permit at most one√moving particle to have a collision. For a particle at x ∈ (0, 1) with√σ = + and speed η, a collision occurred in the previous t units of time if and only if t η > x. Let √ √ t = {X ∈ M (+,+) : (i) η0 , ξ, η1 < t −2 and (ii) t η0 < x0 or t η1 < x1 }. 2 √ We claim that µ(t ) = 1− O(t 1+δ ). This is because if t η > x, then either (a) η > t − 3 , ∞ 1 2 or (b) x < t · t − 3 = t 3 . Now (a) occurs with probability − 2 τ (η)dη = O(t), and (b) t 3 2 2 4 √ √ occurs with probability t 3 . Thus µ{t η0 > x0 and t η1 > x1 } = [O(t 3 )]2 = O(t 3 ). The implications of (A) and (B) above are as follows: Suppose we show µˆ t |t = µ|t + O(t 1+δ ) for some δ with 0 < δ ≤
1 3.
(∗)
Then by (B), we have
µ = µ · It + O(t 1+δ ) = µˆ t · It + O(t 1+δ ). Now we know from (A) that |µˆ t | = 1 − O(t 1+δ ). This together with the equality above implies µˆ t (tc ) = O(t 1+δ ), so that µˆ t · It = µˆ t + O(t 1+δ ). Thus µ = µˆ t + O(t 1+δ ) = µt + O(t 1+δ ) , the second equality following from (A). This is what we seek.
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
217
Step 2. Analysis of simple changes in configuration. Let ϕ and ϕˆt be the densities of µ and µˆ t respectively, and fix X¯ = (η¯ 0 , ξ¯ , η¯ 1 ; x¯0 , x¯1 ; +, +) ∈ t . We seek to compare ϕ( X¯ ) and ϕˆt ( X¯ ), remembering that to obtain µˆ t , one considers only initial conditions −2 in which √ all energies are < √t . Case 1: t η¯ 0 < √ x¯0 and t η¯ 1√< x¯1 . The only way to reach X¯ in time t is to start from (η¯ 0 , ξ¯ , η¯ 1 ;√x¯0 − t η¯ 0 , x¯1 −√t η¯ 1 ; +, +). No√collision occurs, and ϕ( X¯ ) = ϕˆt ( X¯ ). Case 2: t η¯ 0 < x¯0 and t η¯ 1 > x¯1 . Let s η¯ 1 √ = x¯1 . We claim that the only way to reach X¯ in time t is to start from ( η ¯ , ξ, η ; x ¯ − t η¯ 0 , x1 ; +, −), where ξ = η¯ 1 , η1 = ξ¯ 0 1 0 √ and x1 = (t − s) η1 . Starting here, one reaches X¯ after a single exchange of energy between the right particle and the site. To compute the densities at X¯ , we let ε > 0 be a verysmall number. Initial conditions with x1 ∈ ((t − s) ξ¯ − , (t − s) ξ¯ + ) and √ / ξ¯ = ε/ η¯ 1 will reach the target interval (x¯1 − ε, x¯1 + ε) at time t. Since ρ(η¯ 1 )τ (ξ¯ ) = ρ(η¯ 1 )
1 ρ(ξ¯ ) 1 ρ(ξ¯ ) = ρ(η¯ 1 ) √ ε = ρ(ξ¯ )τ (η¯ 1 )ε , Z ξ¯ Z η¯ 1
¯ we conclude ϕˆt ( X¯ ). √ that ϕ( X ) = √ √ Case 3: t η¯ 0 > x¯0 and t η¯ 1 < x¯1 . Here we let s be such that s η¯ 0 = x¯0 . The only √ way to reach X¯ is for the left particle to start at x0 = (t − s) η0 < 1 if η0 is its initial energy, go left, reach the left bath at time t − s, and have the newly emitted energy η¯ 0 reach x¯0 at time t. To compare densities, again√fix a small target interval (√ x¯0 − ε, x¯0√ + ε). √ This forces x0 ∈ ((t − s) η0 − (η0 ), (t − s) η0 + (η0 )) with (η0 )/ η0 = ε/ η¯ 0 . Thus −2 √ t η0 1 ϕˆt ( X¯ ) = ρ(ξ¯ )τ (η¯ 1 ) · τ (η0 ) √ dη0 · ρ(η¯ 0 ) 4 η¯ 0 0 t −2 1 ρ(η¯ 0 ) 1 ρ(η0 )dη0 = ϕ( X¯ ) · (1 − O(t 2 )). = ρ(ξ¯ )τ (η¯ 1 ) · √ 4 Z η¯ 0 0 √ √ Since t η¯ 0 > x¯0 and t η¯ 1 > x¯1 is not permitted for X¯ ∈ t , we have exhausted all viable cases, completing the proof of (∗).
5.2. Proof of Theorem 2. We follow closely Section 5.1, adapting the ideas there to chains with N sites. We begin with the analogous set of reductions. As in Section 5.1, we seek to show µt = µ + O(t 1+δ ) for some δ > 0, focusing on a fixed component of the phase space M (σ¯ ) = {(σ0 , · · · , σ N ) = σ¯ }. Instead of µt , we consider µˆ t = PXt d µ(X ˆ ) where µˆ is the restriction of µ to {ηi , ξi < t −2 , all i}. It follows from the estimates in Section 5.1 that (for fixed N ) this introduces an error within the tolerable range: ∞ ∞ |µt − µˆ t | < N ρ + (N + 1) τ = O(t 2 ). t −2
(σ¯ )
We define t = t
t −2
to be
{X ∈ M (σ¯ ) : (i) ξi , ηi < t −2 for all i; √ √ (ii) t ηi < xi if σi = +, t ηi < 1 − xi if σi = − for all except at most one i} ,
218
P. Balint, K. K. Lin, L.-S. Young
√ the idea being that when the above relation between t, ηi and xi holds, no collision occurs in the t units of time prior to arrival in that configuration. The same argument as 4 before shows that µ(t ) = 1 − O(t 3 ): the set on which this relation is violated by two 4 of the moving particles has measure O(t 3 ); the set on which it is violated by more than two particles is smaller. As in Section 5.1, we seek to compare ϕˆt and ϕ on the set t . As before, all changes in configuration involved are simple: there are 3 cases corresponding to no collisions, a collision with a site, and a collision with a bath. The analysis is identical to that in Section 5.1.
Part II. Numerical Results We report here some results on nonequilibrium steady states (NESS) obtained by studying numerically two families of models: the first has exponential bath injections and the second, which we call the “two-energy model”, is chosen for the transparency of the role played by the integrable dynamics.
6. Exponential Baths We consider in this section processes defined in Section 1.1 for which the bath injections have exponential distributions, i.e., L(ξ ) = β L e−β L ξ and R(ξ ) = β R e−β R ξ , where, as usual, β L = TL−1 and β R = TR−1 are to be thought of as inverse temperatures. We are interested in steady states of systems that are in contact with two baths at unequal temperatures, i.e., where TL = TR . It follows from Theorem 1 that all invariant distributions that arise from these bath injections are ergodic. In Section 6.1, we demonstrate that mean energy profiles are well defined, both for finite N and as N → ∞. In Section 6.2, we focus on specific points along the chain, and investigate marginals of the NESS on (very) short segments. In the simulations shown, mean bath temperatures are TL = 1 and TR = 10, and chains of various lengths up to N = 1600 are used. 6.1. Macroscopic energy profiles. For a chain with N sites, we let E[ξi ] denote the mean energy at site i, “mean” being taken with respect to the unique steady state distribution. Let f N : [0, 1] → R be the function which linearly interpolates f N ( Ni+1 ) = E[ξi ], i = 1, · · · , N . If as N increases, f N converges (pointwise) to a function f on [0, 1], we will call f the site-energy profile of this model. Similarly, we let f N be the function that interpolates f N ( Ni+1 +2 ) = E[ηi ], i = 0, 1, . . . , N , and call f = lim N f N the gap-energy profile. Profiles of gap energies conditioned on σi being + or − are denoted f N+ and f N− . Figure 1 shows plots of f N , f N± , and f N . Finite-chain profiles are found to vary little as N goes from 100 to 1600, so we assume the limit profile will not be too different. Convergence time to steady state is slow, and increases with N as expected. (See numerical details in the caption.) 6.2. Local equilibrium properties. This subsection is about local properties of NESS, by which we refer to marginals of the form µˆ x,,N , where x ∈ (0, 1), N is the length of
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
219
5
4
Mean gap energy
Mean site energy
8
6
4
2
3
2
1
0
0 0.2
0.4
0.6
0.8
x (a) Site energies, N = 1600
0
0.2
0.4
0.6
0.8
1
x (b) Gap energies, N =200
Fig. 1. Mean energy profiles for the exponential baths model. The bath temperatures are TL = 1 and T R = 10. Panel (a) shows the mean site energies f N for N = 1600. In (b), we have superimposed the gap energy profiles f N+ , f N− , and f N for N = 200. Numerical details: We impose a constraint that injected energies are ≥ 0.01 but do not impose a high energy cutoff. We simulate the system for 10k events (an “event” being one collision anywhere in the chain), increasing k until the computed profile stabilizes. This occurred after 1013 events in (a), 4 × 1011 events in (b)
the chain, N , and µˆ x,,N is the marginal distribution of the NESS on the -chain centered at site [x N ]. For fixed (unequal) bath distributions, since the energy gradient on the -chain tends to zero as N → ∞, one might expect the µˆ x,,N to resemble equilibrium distributions, i.e., invariant measures on chains with equal bath injections. Theorem 2 gives an explicit formula for the equilibrium distribution µρ,N of the N -chain when the bath distributions are L = R = ρ. This result is valid for very general ρ. Notice that when specialized to the case ρ(ξ ) = βe−βξ , the density of µρ,N has the √ familiar form Z1 e−β H . More precisely, let us use (xi , vi ), vi = σi ηi , as coordinates in the gaps instead of (ηi , xi , σi ). Then −(N +1) −β i vi2 e i d xi dvi . dµρ,N = β N e−β i ξi i dξi × Z β
(1)
Returning to the situation of unequal exponential baths, one way to define local thermodynamic equilibrium (LTE) is to require that for every x ∈ (0, 1) and ∈ Z+ , as N → ∞, the marginals µˆ x,,N tend to a probability measure having the form in (1) with N replaced by and β = β(x) for some β(x) > 0. For a more physical notion of LTE, one sometimes considers only -chains for 1 N . A. Single-site and single-gap marginals. We consider marginal distributions of the NESS at single sites and single gaps at [x N ], where x ∈ (0, 1) is fixed and N is varied. To show that the system does not tend to LTE (by any definition), it suffices to show that these marginals are not Gibbsian. This is because if marginals on -chains tended to µˆ x,,N , projecting onto the single site or gap at [x N ] one would again obtain a distribution of the form in (1). Figure 2 shows two examples of single-site marginal densities at x = 0.25 and 0.6 for a chain of length N = 1600. The plots are log-linear, so that exponential functions
220
P. Balint, K. K. Lin, L.-S. Young
10
Probability density
Probability density
-0.5 10 -1 10 -1.5 10 -2 10
-0.75 10
10
-1.25
10 10
-1.5
-1.75 10
-2.5 10
10
0
10
-1
-2
-2.25
0
20
10
Energy
20
Energy
(b) x = 0.6
(a) x = 0.25
Fig. 2. Site energy distributions for the exponential baths model in log-linear scales. The parameters are TL = 1, T R = 10, N = 1600. Empirical data are plotted in open squares; the solid curve represents the mixture of left and right bath distributions that best fits the data (in the total variation norm)
would appear as straight lines. Without a doubt both of the distributions shown are very far from exponential. These distributions are compared to mixtures of the bath distributions, meaning distributions of the form a L(ξ ) + (1 − a)R(ξ ), a = a(x) ∈ (0, 1). In Fig. 2, open squares represent empirical data (from simulations), while the solid curves are given by the formulas for mixtures with a chosen in each of the two cases to fit the data best. Given the numerical cutoffs, etc., we think the fits between empirical data and the mixtures curves are excellent. Plots (not shown) at other locations evenly spaced along the chain show the same phenomenon, with the “knee” moving steadily to the left as x increases. Likewise, single-gap marginals are found to be mixtures: At gaps adjacent to [x N ] 2 for fixed x, instead of Z β−1 e−βv for some β = β(x), we find numerically the marginals to be of the form a·
e−β L v Z βL
2
+ (1 − a) ·
e−β R v Z βR
2
for some a = a(x), in clear violation of the prescription of Gibbs for Hamiltonian systems in equilibrium. We explain how we arrived at the idea of “mixtures”: First, the energies in a chain should reflect those injected, and if there is a nontrivial discrepancy between β L and β R , it is difficult to imagine having an abundance of the energies in the “middle” to constitute all the exponential distributions along the chain. We are also influenced by the following stochastic model, which can be thought of as a “zeroth-order” approximation to our Hamiltonian systems.
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
221
The random swaps model. This is a stochastic model defined by N random variables ξ1 , . . . , ξ N , to be thought of as energies, located at sites 1, . . . , N . As usual there are two baths, situated at sites 0 and N +1. At bonds between sites, i and i +1 for i ∈ {0, 1, . . . , N } are exponential clocks which ring independently at rate 1. When a clock goes off, energies between the 2 sites are swapped. That is to say, for i = 0, N , when the clock between sites i and i + 1 rings, the values of ξi and ξi+1 are interchanged. Swapping energy with the left bath means that ξ1 is replaced by an energy drawn randomly from the left bath distribution L(ξ ), and similarly for ξ N and the right bath, which has distribution R(ξ ). The energies emitted by each bath are assumed to be i.i.d. and independent of emissions by the other bath. A model along similar lines was studied in [10]. Proposition 6.1. In the random swaps model, for each i ∈ {1, . . . , N }, the marginal distribution ρi of the unique NESS is given by
i N +1−i L+ R ρi = N +1 N +1 for arbitrary bath distributions L and R. Sketch of proof. First, we distinguish only between whether an energy is “from the left” or “from the right”. In this reduced system, it is easy to see that there is a unique steady state, the single-site marginals of which are, by definition, mixtures of these two kinds of energies. Since the trajectory of each energy, from the time it enters the system to when it leaves, is that of a simple unbiased random walk, the weights in the formula for ρi follow from basic recursive relations. Finally, each of the energies in this reduced system can be assigned independently any one of the allowed values, so that the marginals are really mixtures of L and R.
Since Proposition 6.1 holds for arbitrary bath distributions L and R, it is natural to ask if a similar result holds for the Hamiltonian chain with non-exponential baths. We investigated this question and found the answer to be negative. For example, when L and R are uniform distributions, marginal distributions are far from mixtures of uniform distributions, see Fig. 3. When the densities of L and R have shapes closer to exponential distributions, single-site marginals are closer to mixtures, but noticeable differences were seen in each of the half-dozen or so cases tested. The numerical results above raise the following questions: (A) In the case of exponential baths, are single-site marginals as N → ∞ genuinely mixtures of the two bath distributions, or are they simply close to mixtures? (B) If the mixtures result here is exact, are exponential baths the only distributions that have this property, and if so, what are the underlying reasons? Independent of the answer to (A), we believe we have shown very definitively that local marginal distributions do not have the form in (1). Hence the concept of LTE does not apply to this class of Hamiltonian chains. B. Vanishing of spatial correlations. We investigate next if, as N → ∞, µˆ x,,N → µρ, where ρ is the mixture found in Paragraph A. Since spatial correlations are expected to be largest between adjacent sites and gaps, we verify only that (i) marginal distributions on two adjacent sites are product measures; (ii) marginal distributions on two adjacent gaps are product measures, and the directions of travel are independent;
P. Balint, K. K. Lin, L.-S. Young
Probability density
222
0.2
0.1
0 0
2.5
5
7.5
10
Energy
Fig. 3. An example to show that site marginals are not necessarily mixtures. Bath distributions here are uniform distributions: L(ξ ) is supported on [1,5], R(ξ ) on [8,12]; N = 2000. Empirical data are plotted in open squares as above 0.3
Prob( ξ [Nx]+1 | ξ [Nx] )
Prob( ξ [Nx]+1 | ξ [Nx] )
0.3
0.2
0.1
0
0.2
0.1
0 0
10
20
ξ [Nx]
(a) N = 100
0
10
20
ξ [Nx]
(b) N = 1600
Fig. 4. Conditional probabilities in adjacent sites
for the exponential-baths model. The 4 curves in each panel show P ξ[N x]+1 ∈ I | ξ[N x] ∈ [2k, 2(k + 1)] for k = 0, 1, 2, · · · , 14 and I = [0, 2], [2, 4], [4, 6], and [6, 8] (top to bottom). In both panels, x = 0.6. Bath parameters are TL = 1, T R = 10
(iii) marginal distributions on a√site and its adjacent gap are product measures, with gap density = const·ρ(η)/ η, where site density = ρ. A sample of simulation results in support of (i) is shown in Fig. 4. In the horizontal axis are values of ξi for i = [0.6N ], and plotted are conditional distributions of ξi+1 given the various values of ξi . Here we group the values of ξi+1 into intervals (0, 2), (2, 4), (4, 6), and (6, 8), the 4 graphs representing the conditional probabilities of ξi+1 being in each one of these intervals given the value of ξi in the horizontal axis. The two plots, for N = 100 and 1600 respectively, show weak dependence of ξi and ξi+1 in the first and close to zero dependence for the longer chain.
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
223
Table 1. Joint distributions of gap energies conditioned on directions of travel. Here, N = 1600 and i = [0.6 N ]. The numbers are raw probabilities of the specified events
A small sample of data is shown in Table 1 to illustrate the combinations of joint densities that need to be checked to verify (ii). Here we focus on two adjacent gaps at x = 0.6, and distinguish only between “high” and “low” energies, referring to energies η > 3 as “high” and η < 3 as “low”. The fractions of time in all configurations of high-low and (σi , σi+1 ), referring to the directions of travel in these gaps, are tabulated. Finally, we have computed energy distributions for adjacent sites and gaps at various sites along the chain. The data confirm property (iii) above. Remark. The “local product structure” finding above may be valid beyond the case of exponential baths. Our simulations show that it holds both for the uniform distribution models in Fig. 3 and for the “two-energy models” discussed in the next section. Since the latter are rather “extreme” from many points of view, it gives reason to conjecture that this phenomenon – which is very natural – may hold widely. 7. Two-Energy Models The bath distributions in this section are sharply peaked Gaussians truncated at 0. In the simulations shown, these Gaussians have means 1 and 5, and their standard deviations are chosen to be 0.2, so that the probability of being within ±0.5 of the means is roughly 99%. Even though these systems are small perturbations of integrable models, we know from Theorem 1 that all of their invariant measures are ergodic. 7.1. Macroscopic energy profiles and local distributions. Figure 5(a) shows site-energy profiles f N for a range of N . Unlike the exponential case, these profiles vary substantially with N : For N = 10 (not shown), the profile is essentially flat; as N increases, it acquires a gradient and appears to be stabilizing, but even at N = 3200, the profile is still moving a little. From these profiles, one suspects – correctly – that for (very) small N , most energies move monotonically along the chain, entering from one end and exiting at the other. As N increases, some of the energies turn around, some doing so a number of times before exiting, creating a gradient in the profile. Figure 5(b) shows profiles for gap energies with specified directions, showing that for the chain with N = 200, f N+ and f N− are substantially different, with more energy
P. Balint, K. K. Lin, L.-S. Young 5
5
4
4
Mean gap energy
Mean site energy
224
3
2
1
3
2
1
0
0 0
0.2
0.4
0.6
x (a) Mea n site en er gies
0.8
1
0
0.2
0.4
0.6
0.8
1
x
(b) Mean gap energies
Fig. 5. Mean energy profiles for two-energy model. Panel (a) shows mean site energies; the curves are N =40, 200, and 3200, in order of increasing values at x = 1. Panel (b) shows mean gap energies f N− , f N , and f N+ (top to bottom) for N = 200. The rules used to determine when a steady state is reached are as Fig. 1. See text for the bath distributions
1s moving to the right and 5s moving to the left. This is to be contrasted with Fig. 1(b), which shows that for exponential baths, this left–right discrepancy has by and large vanished by N = 200. Indeed in the present model, f N+ and f N− remain quite far apart in some of the gaps even at N = 3200, leaving open the question whether or not the distributions conditioned on σi = + and − will eventually equalize. A follow-up investigation is discussed in Section 7.2. With regard to local distributions, computations on site-to-site, gap-to-gap, and siteto-gap correlations similar to those in Section 6 were carried out. Independence was found to be achieved already at relatively small N . Figure 6 captures the two main points: The two curves representing P(ξ[0.6N ]+1 = 1|ξ[0.6N ] = k), k = 1 and 5, are virtually on top of each other for N ≥ 400, illustrating the rapid vanishing of correlations between adjacent sites, while both curves continue to have a slightly negative slope at N = 3200, illustrating the slow convergence of mean site energies. 7.2. Ballistic transport vs. diffusion: a phenomenological explanation. The purpose of this subsection is to examine more closely the way in which energy is transported from one end of the chain to the other under near-integrable dynamics. Since the vast majority of the energies in this system are very close to 1 or 5, let us assume there are only two kinds of energies in the system. For definiteness, we adopt the view in Section 1.3, and focus on the movement of 1s on the right side of the chain. We will demonstrate that modulo certain time changes, the statistics generated by the movements of these energies resemble those of a particle undergoing a 1-D diffusion with a spatially-varying diffusion coefficient. We begin by introducing an especially simple model which will be used for comparison purposes: Model A. Consider a Markov chain with state space {0, 1, · · · , N + 1} defined as follows: Starting from 1, one performs a simple, unbiased random walk until either 0 or N + 1 is reached, then returns to 1 to start over again. For 1 ≤ i ≤ N − 1, let n (i) and rn (i) denote the number of left and right crossings respectively between sites i and i + 1
Prob( ξ[Nx]+1 | ξ[Nx] )
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
225
0.2
0.1
0
0
1000
2000
3000
N Fig. 6. Nearest-neighbor conditional probabilities in the two-energy model. This plot shows P(ξ[N x]+1 < 3 | ξ[N x] < 3) (open squares) and P(ξ[N x]+1 < 3 | ξ[N x] > 3) (solid discs) as functions of N . The location is x = 0.6; N ranges from 100 to 3200
in the first n steps, and let E N ,i = limn→∞ E N [n (i)]/E N [rn (i)]. Here are some easy facts: (i) For all N and i, 0 < E N ,i < 1. (ii) For each N , E N ,i decreases as i increases. (iii) For each x ∈ (0, 1), E N ,[x N ] increases to 1 as N → ∞. These facts follow from the following observations: Fix i and N , and focus only on left and right crossings between sites i and i + 1. Then a left crossing is necessarily followed by a right crossing, while a right crossing can be followed by either. The probability that a right crossing is followed by another right crossing is equal to the probability of starting from site i + 1 and reaching site N + 1 before site i in the random walk. The numbers E N ,i can be explicitly estimated if one so desires. To bring this simple model closer to our Hamiltonian chains, we first identify some relevant features of the chains. The following idea is used in the rigorous part of this paper (Lemmas 4.2 and 4.4): Consider a scenario in which an energy 1 is in the midst of many energy 5s, i.e. in some segment of the chain, ξ j = η j = 5 for all j except for a ˆ where η ˆ = 1. Then the energy 1 will move monotonically in some direction single j, j until the pattern is disrupted by the approach of another (oncoming) energy 1. After such a “collision” a variety of things can happen; the resulting motion of the energies depends on the details of the interaction. Since the energy profiles have nonzero gradient (see Fig. 5), 1s are more sparse on the right side of the chain. Thus the closer to the right end, the larger the “mean free distance” between “collisions” of energy 1s. With regard to the ratio of left to right crossings (and not the actual number of crossings per unit time), the observations above suggest the following modification of Model A: Model B. Let λ : (0, 1) → (0, ∞) be a function that monotonically increases from 1 to ∞ on some subinterval (x∗ , 1) ⊂ (0, 1). We consider a process similar to that in Model A but with transition rules modified as follows: when at site i = [x N ], go X sites to the
226
P. Balint, K. K. Lin, L.-S. Young
(a) 1
10
0.8
Prob(-) / Prob(+)
Probability
-1.5 10 -2
-2.5 10
10
0.6 0.4 0.2
-3
0 10
20
30
40
50
60
70
0
Run length r
(b)
0.2
0.4
0.6
0.8
1
x
(c)
/
Fig. 7. Statistics on lengths of runs. (a) Mean lengths of runs in different intervals [a, b]. Note that runs that reach the boundary of [a, b] are artificially cut short by our numerical procedure (see “numerical details” below); we do not consider averages based on too many (> 10%) such runs meaningful and have omitted those averages. (*) marks averages for which 5 − 10% of the runs were cut short. (b) Log-linear plot of the histogram of run-lengths in the interval [0.6, 0.8], for N = 3200. (c) The ratio P(ηi ≈ 1, σ = −)/P(ηi ≈ 1, σ = +) as a function of x = i/N . The curves are, from bottom to top, N =100, 400, 1600, 3200. Numerical details for (a) and (b): For a given interval [a, b], we begin with a suitably prepared system and wait until an energy ≈ 1 hits the site at x = (a + b)/2. The energy is tracked until it hits either x = a or x = b; the lengths of runs executed by this energy are recorded. This process is repeated (after a delay to ensure that the energies are not too correlated). Each data point in Table (a) based on 104 – 106 runs collected this way
left/right with probability 21 / 21 , where X is an exponential random variable with mean λ(x), and return as before to 1 to start over once 0 or N + 1 is reached. First we confirm numerically that in terms of the statistics generated, Model B gives a good approximation of movements of energy 1s on the right side of the real chain. (Obviously, we do not claim that the two models are equivalent.) We study the lengths of runs of randomly picked 1s in specified segments of the chain, a run being defined to be a consecutive sequence of moves in the same direction. Some mean length-of-runs are tabulated in Fig. 7(a). The numbers in each row increase as one moves to the right, a trend consistent with the environment becoming increasingly dominated by energy 5s. As N increases, these means stabilize, as one would expect them to when the local marginals tend to a distribution of the form µρ, . Histograms of lengths of runs within specified intervals are plotted and found to have roughly exponential distributions for large N ; one such plot is shown in Fig. 7(b). Returning to Models A and B, observe that they are qualitatively similar if we imagine that the sites in Model B are “packed closer and closer” as x → 1. More precisely, the ratio of left to right crossings at site [x N ] in Model B should resemble that at [y N ] in Model A for some y > x, with (1 − y)/(1 − x) → 0 as x → 1. Properties (i-iii) in
Ergodicity and Energy Distributions for Integrable Hamiltonian Chains
227
Model A therefore pass to Model B. Assuming they pass from there to our Hamiltonian chain, one would conclude that for our chain: – left-right traffic will equalize everywhere as N → ∞; – this equalization occurs more slowly than at corresponding sites in Model A, the discrepancy increasing as x → 1. Simulations to compute the ratios of left/right crossings were performed. The results, shown in Fig. 7(c), are very much consistent with the predictions above: for each N , the x → Prob(σ[N x] = −)/Prob(σ[N x] = +) curves are decreasing, and at each x ∈ (0, 1), this ratio increases with N and appears to head toward 1. Remarks on connection with theory. We have used the fact, proven rigorously in Section 4.3, that in regions of the chain occupied by significantly more energies of one kind than the other, the ones with low density will, on average, have relatively long runs. The same is true, in fact, for the energies with high density. Our rigorous work leaves untreated the situation where the two energies occur in roughly comparable proportions. For these regions, we have seen via simulations that the mean lengths of runs tend still to be greater than 2, the value for simple unbiased random walks. For example, when TL = 1 and TR = 5, the region with the most even mix of 1s and 5s is [0.2, 0.4], and there the mean run-length for 1s is > 5 (see Fig. 7(a)). Thus the phenomena discussed in this subsection appear to be valid for both energies on a good part of the chain.
Summary and Conclusion We considered a Hamiltonian chain with many conserved quantities and studied its nonequilibrium steady states when the two ends of the chain are put in contact with unequal heat baths. Our main findings are: Rigorous results. First, under mild restrictions on the bath distributions L(ξ ) and R(ξ ), we proved ergodicity of the invariant measure (assuming existence). Second, we identified a class of equilibrium measures {µρ,N }, where µρ,N is the unique invariant probability on the N -chain with L = R = ρ; the measures µρ,N are product measures. Numerical results. Simulations were performed for chains with two kinds of bath distributions: exponential distributions and sharply peaked Gaussians, the latter giving rise to what is called the “two-energy model”. 1. We demonstrated numerically that (a) NESS exist for finite N and mean energy profiles converge as N → ∞; (b) as N → ∞ local marginals at x ∈ (0, 1) tend to measures of the form µρ, for some distribution ρ = ρ(x). 2. For exponential bath distributions, the limits of local marginals are definitively not Gibbs measures, i.e., this chain violates the concept of LTE. The marginals appear to be weighted averages of Gibbs measures. 3. For the 2-energy model, the paths traced out by energies resemble the sample paths of a random walk with a bias in favor of continuing in the same direction, with this bias increasing as x → 0 or 1. We conclude that the resulting transport behavior is more normal, i.e., more diffusive, than one might have expected given the integrable character of the dynamics.
228
P. Balint, K. K. Lin, L.-S. Young
References 1. Bernardin, C., Olla, S.: Fourier’s law for a microscopic model of heat conduction. J. Stat. Phys. 118, 271–289 (2005) 2. Bertini, L., De Sole, A., Gabrielli, D., Jona-Lasinio, G., Landim, C.: Macroscopic fluctuation theory for stationary non-equilibrium states. J. Statist. Phys. 107, 635–675 (2002) 3. Bonetto, F., Lebowitz, J., Rey-Bellet, L.: Fourier law: a challenge to theorists. In: Mathematical Physics 2000, edited by Fokas, A., Grigoryan, A., Kibble, T., Zegarlinski, B., London: Imp. Coll. Press, 2000 4. Bricmont, J., Kupiainen, A.: Fourier’s law from closure equations. Phys. Rev. Lett. 98, 214301 (2007) 5. Bricmont, J., Kupiainen, A.: Towards a derivation of Fourier’s law for coupled anharmonic oscillators. Commun. Math. Phys. 274, 555–626 (2007) 6. Collet, P., Eckmann, J.-P.: A model of heat conduction. Preprint, http://arxiv.org/abs/0804.3025v1[mathph], 2008 7. Collet, P., Eckmann, J.-P., Mejía-Monasterio, C.: Superdiffusive heat transport in a class of deterministic one-dimensional many-particle Lorentz gases. Preprint, http://arxiv.org/abs/0810.4461v1[cond-mat. stat.-mech], 2008 8. Derrida, B.: Non-equilibrium steady states: fluctuations and large deviations of the density and of the current. J. Stat. Mech. (2007) P07023, doi:10.1088/1742-5468/2007/07/p02023, July 2007 9. Derrida, B., Lebowitz, J.L., Speer, E.R.: Large deviation of the density profile in the steady state of the open symmetric simple exclusion process. J. Stat. Phys. 107, 599–634 (2002) 10. Dhar, A., Dhar, D.: Absence of local thermal equilibrium in two models of heat conduction. Phys. Rev. Lett. 82, 480–483 (1999) 11. de Groot, S.R., Mazur, P.: Non-equilibrium Thermodynamics. Amsterdam: North-Holland, 1962 12. Eckmann, J.-P., Hairer, M.: Non-equilibrium statistical mechanics of strongly anharmonic chains of oscillators. Commun. Math. Phys. 212, 105–164 (2000) 13. Eckmann, J.-P., Jacquet, P.: Controllability for chains of dynamical scatterers. Nonlinearity 20, 1601– 1617 (2007) 14. Eckmann, J.-P., Mejía-Monasterio, C., Zabey, E.: Memory effects in nonequilibrium transport for deterministic Hamiltonian systems. to appear in J. Stat. Phys., 2006 15. Eckmann, J.-P., Pillet, C.-A., Rey-Bellet, L.: Non-equilibrium statistical mechanics of anharmonic chains coupled to two heat baths at different temperatures. Commun. Math. Phys. 201, 657–697 (1999) 16. Eckmann, J.-P., Young, L.-S.: Nonequilibrium energy profiles for a class of 1-D models. Commun. Math. Phys. 262, 237–267 (2006) 17. Gaspard, P., Gilbert, T.: Heat conduction and Fourier’s law in a class of many particle dispersing billiards. New J. Phys. 10, 103004 (2008) 18. Kipnis, C., Landim, C.: Scaling Limits of Interacting Particle Systems. Berlin: Springer-Verlag, 1999 19. Kipnis, C., Marchioro, C., Presutti, E.: Heat flow in an exactly solvable model. J. Stat. Phys. 27, 65–74 (1982) 20. Larralde, H., Leyvraz, F., Mejía-Monasterio, C.: Transport properties of a modified Lorentz gas. J. Stat. Phys. 113, 197–231 (2003) 21. Lepri, S., Livi, R., Politi, A.: Thermal conduction in classical low-dimensional lattices. Phys. Rep. 377(1), 1–80 (2003) 22. Li, B., Casati, G., Wang, J., Prosen, T.: Fourier Law in the alternate-mass hard-core potential chain. Phys. Rev. Lett. 92, 254301 (2004) 23. Lin, K.K., Young, L.-S.: Correlations in nonequilibrium steady states of random-halves models. J. Stat. Phys. 128, 607–639 (2007) 24. Olla, S., Varadhan, S.R.S., Yau, H.T.: Hydrodynamical limit for a Hamiltonian system with weak noise. Commun. Math. Phys. 155, 523–560 (1993) 25. Rateitschak, K., Klages, R., Nicolis, G.: Thermostating by deterministic scattering: the periodic Lorentz gas. J. Stat. Phys. 99, 1339–1364 (2000) 26. Ravishankar, K., Young, L.-S.: Local thermodynamic equilibrium for some stochastic models of Hamiltonian origin. J Stat. Phys. 128, 641–665 (2007) 27. Rey-Bellet, L., Thomas, L.E.: Exponential convergence to non-equilibrium stationary states in classical statistical mechanics. Commun. Math. Phys. 225, 305–329 (2002) 28. Rieder, Z., Lebowitz, J.L., Lieb, E.: Properties of a harmonic crystal in a stationary nonequilibrium state. J. Math. Phys. 8, 1073 (1967) 29. Spohn, H.: Long range correlations for stochastic lattice gases in a nonequilibrium steady state. J. Phys. A 16, 4275–4291 (1983) Communicated by A. Kupiainen
Commun. Math. Phys. 294, 229–249 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0869-2
Communications in
Mathematical Physics
Ground State Energy of Large Atoms in a Self-Generated Magnetic Field László Erd˝os1, , Jan Philip Solovej2, 1 Institute of Mathematics, University of Munich, Theresienstr. 39,
D-80333 Munich, Germany. E-mail:
[email protected]
2 Department of Mathematics, University of Copenhagen, Universitetsparken 5,
DK-2100 Copenhagen, Denmark. E-mail:
[email protected] Received: 11 March 2009 / Accepted: 16 April 2009 Published online: 3 July 2009 – © Springer-Verlag 2009
Abstract: We consider a large atom with nuclear charge Z described by non-relativistic quantum mechanics with classical or quantized electromagnetic field. We prove that the absolute ground state energy, allowing for minimizing over all possible self-generated electromagnetic fields, is given by the non-magnetic Thomas-Fermi theory to leading order in the simultaneous Z → ∞, α → 0 limit if Z α 2 ≤ κ for some universal κ, where α is the fine structure constant. 1. Introduction The ground state energy of non-relativistic atoms and molecules with large nuclear charge Z can be described by Thomas-Fermi theory to leading order in the Z → ∞ limit [L,LS]. Magnetic fields in this context were taken into account only as an external field, either a homogeneous one [LSY1,LSY2] or an inhomogeneous one [ES] but subject to certain regularity conditions. Self-generated magnetic fields, obtained from Maxwell’s equation are not known to satisfy these conditions. In this paper we extend the validity of Thomas-Fermi theory by allowing a self-generated magnetic field that interacts with the electrons. This means we look for the absolute ground state of the system, after minimizing for both the electron wave function and for the magnetic field and we show that the additional magnetic field does not change the leading order Thomas-Fermi energy. Apart from finite energy, no other assumption is assumed on the magnetic field. The nonrelativistic model of an atom in three spatial dimensions with nuclear charge Z ≥ 1 and with N electrons in a classical magnetic field is given by the Hamiltonian N 1 1 Z cl H N ,Z (A) = T j (A) − + + B2 (1.1) |x j | |xi − x j | 8π α 2 R3 j=1
i< j
Partially supported by SFB-TR12 of the German Science Foundation. Work partially supported by the Danish Natural Science Research Council and by a Mercator Guest Professorship from the German Science Foundation.
230
L. Erd˝os, J. P. Solovej
acting on the space of antisymmetric functions 1N H with a single particle Hilbert space H = L 2 (R3 ) ⊗ C2 . The coordinates of the N electrons are denoted by x = (x1 , x2 , . . . , x N ). The vector potential A : R3 → R3 generates the magnetic field B = ∇ × A and it can be chosen divergence-free, ∇ · A = 0. The last term in (1.1) is the energy of the magnetic field. The kinetic energy of an electron is given by the Pauli operator T (A) = [σ · (p + A)]2 = (p + A)2 + σ · B,
p = −i∇x .
Here σ is the vector of Pauli matrices. We use the convention that for any one-body operator T , the subscript in T j indicates that the operator acts on the j th variable, i.e. T j (A) = [σ j · (−i∇x j + A(x j ))]2 . The term −Z |x j |−1 describes the attraction of the j th electron to the nucleus located at the origin and the term |xi − x j |−1 is the electrostatic repulsion between the i th and j th electron. Our units are 2 (2me2 )−1 for the length, 2me4 −2 for the energy and 2mec−1 for the magnetic vector potential, where m is the electron mass, e is the electron charge and is the Planck constant. In these units, the only physical parameter that appears in (1.1) is the dimensionless fine structure constant α = e2 (c)−1 . We will assume that Z α 2 ≤ κ with some sufficiently small universal constant κ ≤ 1 and we will investigate the simultaneous limit Z → ∞, α → 0. Note that the field energy is added to the total energy of the system and by the condition ∇ · A = 0 we have B2 = |∇ ⊗ A|2 , (1.2) R3
R3
where ∇ ⊗ A denotes the 3 × 3 matrix of all derivatives ∂i A j and |∇ ⊗ A|2 = 3 2 i, j=1 |∂i A j | . We will always assume that the vector potential belongs to the space of divergence free H 1 -vector fields A := {A ∈ H 1 (R3 , R3 ), ∇ · A = 0}. In the analogous nonrelativistic model of quantum electrodynamics, the electromagnetic vector potential is quantized. In the Coulomb gauge it is given by A (x) = A(x) = A− (x) + A− (x)∗ with A− (x) =
α 1/2 2π
R3
g(k) aλ (k)eλ (k)eik·x dk. √ |k| λ=±
Here g(k) is a cutoff function, satisfying |g(k)| ≤ 1 and supp g ⊂ {k ∈ R3 : |k| ≤ } with a constant < ∞ (ultraviolet cutoff). The field operators A(x) depend on the cutoff function g(k) whose precise form is unimportant; the only relevant parameter is . For each k, the two polarization vectors e− (k), e+ (k) ∈ R3 are chosen such that together with the direction of propagation k/|k| they are orthonormal. The operators aλ (k), aλ (k)∗ are annihilation and creation operators acting on the bosonic Fock space F over L 2 (R3 ) and satisfying the canonical commutation relations [aλ (k), aλ (k )] = [aλ (k)∗ , aλ (k )∗ ] = 0, [aλ (k), aλ (k )∗ ] = δλλ δ(k − k ).
Energy of Atoms in a Self-Generated Field
The field energy is given by Hf = α The total Hamiltonian is qed H N ,Z
=
N j=1
−1
231
R3
|k|
aλ (k)∗ aλ (k)dk.
λ=±
1 Z T j (A ) − + + Hf , |x j | |xi − x j |
(1.3)
i< j
and it acts on ( 1N H) ⊗ F. The stability of atoms in a classical magnetic field [F,FLL,LL,LLS] implies that the operator (1.1) is bounded from below uniformly in A, if Z α 2 is small enough. It is known [LY,ES2] that stability fails if Z α 2 is too large. The analogous stability result for quantized field [BFG] states that (1.3) is bounded from below if Z α 2 is small. In particular, we can for each fixed A define the operators in (1.1) and (1.3) as the Friedrichs extensions of these operators defined on smooth functions with compact support. The ground state energy of the operator with a classical field is given by N
cl cl ∞ 3 2 E N ,Z (A) = inf , H N ,Z (A) : ∈ C0 (R ) ⊗ C , = 1 , 1
and after minimizing in A we set cl E cl N ,Z = inf E N ,Z (A). A∈A
We note that it is sufficient to minimize over all A ∈ A0 , where A0 = Hc1 (R3 , R3 ) denotes the space of compactly supported H 1 vector fields. It is easy to see that the Euler-Lagrange equations for the above minimizations in and A correspond to the stationary version of the coupled Maxwell-Pauli system, i.e., the eigenvalue problem 2 H Ncl,Z (A) = E cl N ,Z together with the Maxwell equation ∇ × B = 4π α J , where J is the current of the wave function . It is for this reason that it is natural to refer to B as a self-generated magnetic field in this context. In the case of the quantized field, we define N
qed qed ∞ 3 2 C0 (R ) ⊗ C ⊗ F, = 1 . E N ,Z = inf , H N ,Z : ∈ 1 qed
The stability results of [F,FLL,LL,LLS,BFG] imply that E cl N ,Z > −∞ and E N ,Z > −∞ 2 if Z α is small enough. Finally, we define the ground state energy with no magnetic field as cl E nf N ,Z := E N ,Z (A = 0).
In all three cases we define E #Z := inf E #N ,Z , N ∈N
# ∈ {cl, qed, nf},
for the absolute (grand canonical) ground state energy. The main result of this paper states that the magnetic field does not change the leading term of the absolute ground state energy of a large atom in the Z → ∞ limit. In particular, Thomas-Fermi theory is correct to leading order even with including self-generated magnetic field.
232
L. Erd˝os, J. P. Solovej
Theorem 1.1. There exists a positive constant κ such that if Z α 2 ≤ κ, then 7
1
cl nf 3 − 63 . E nf Z ≥ EZ ≥ EZ − C Z
(1.4)
For the quantized case, if we additionally assume 1
7
1
≤ κ 4 Z 12 −γ α − 4 with some 0 ≤ γ ≤
1 63 ,
(1.5)
then qed
2 E nf Z + C Z α ≥ E Z
7
3 −γ . ≥ E nf Z − CZ
(1.6)
We note that Z α2 Z 7/3 if κ −1/4 Z 11/12 in the Z → ∞ limit. Remark 1. The leading term asymptotics of the non-magnetic problem is given by the 7 2 3 Thomas-Fermi theory and E nf Z = −cTF Z + O(Z ) as Z → ∞, where cTF = 3.678 · 2
(3π 2 ) 3 is the Thomas-Fermi constant. The leading order asymptotics was established in [LS] (see also [L]). The correction to order Z 2 is known as the Scott correction and was established in [H,SW1] and for molecules in [IS] (see also [SW2,SW3,SS]). The next term in the expansion of order Z 5/3 was rigorously established for atoms in [FS]. Remark 2. The exponents in the error terms are far from being optimal. They can be improved by strengthening our general semiclassical result Lemma 1.3 for special Coulomb like potentials using multi-scale analysis. Remark 3. For simplicity, we state and prove our results for atoms, but the same proofs work for molecules as well; if the number of nuclei K is fixed, each has a charge Z , and assume that the nuclei centers {R1 , . . . , R K } are at least at distance Z −1/3 away, i.e. |Ri − R j | ≥ cZ −1/3 , i = j. Remark 4. Theorem 1.1 holds for the magnetic Schrödinger operator as well, i.e. if we replace the Pauli operator T (A) = [σ · (p + A)]2 by T (A) = (p + A)2 everywhere. The proof is a trivial modification of the Pauli case. The argument is in fact even easier; instead of the magnetic Lieb-Thirring inequality for the Pauli operator one uses the usual Lieb-Thirring inequality that holds for the magnetic Schrödinger operator uniformly in the magnetic field. We leave the details to the reader. Note that although the condition Z α 2 ≤ κ is not needed in order to ensure stability in the Schrödinger case, we still need it in the statement in Theorem 1.1. In the Schrödinger case this condition is not optimal. The upper bound in (1.4) is trivial by using a non-magnetic trial state. The upper bound in (1.6) is obtained by a trial state that is the tensor product of a non-magnetic electronic trial function with the vacuum | of F. The field energy H f and all terms that are linear in A give zero expectation value in the vacuum. The only effect of the quantized field is in the nonlinear term A2 . A simple calculation shows that |A2 | ≤ Cα2 . The main task is to prove the lower bounds. Using the results from [BFG], the result for the quantized field (1.6) will directly follow from an analogous result for a slightly modified Hamiltonian with a classical field. Let N 1 1 Z + + Ti (A)− |∇ ⊗ A|2 H N ,Z (A) = H N ,Z ,α (A) = |xi | |xi − x j | 8π α 2 |x|≤3r i=1
i< j
(1.7)
Energy of Atoms in a Self-Generated Field
233
with some r = D Z −1/3 with D ≥ 1.
(1.8)
Note that instead of the local field energy, the total local H 1 -norm of A is added in (1.7). By (1.2), we have H Ncl,Z (A) ≥ H N ,Z (A)
(1.9)
for any A ∈ A. We define the ground state energy of the modified Hamiltonian (1.7), N
∞ 3 2 E N ,Z (A) := inf , H N ,Z (A) : ∈ C0 (R ) ⊗ C , = 1 , 1
and set E N ,Z := inf E N ,Z (A), A∈A
E Z := inf E N ,Z , N
where the infimum for A ∈ A can again be restricted to compactly supported vector potentials A ∈ A0 . For the modified classical Hamiltonian we have the following theorem: Theorem 1.2. Let Z α 2 ≤ κ and assume that r = D Z −1/3 with 1 ≤ D ≤ Z 1/63 . Then nf 7/3 −1 E nf D . Z ≥ EZ ≥ EZ − C Z
(1.10)
Taking into account (1.9), Theorem 1.2 immediately implies the lower bound in (1.4). The proof of the lower bound in (1.6) follows from Theorem 1.2 adapting an argument in [BFG] that we will review in Sect. 6 for completeness. One of the key ingredients of the proof of Theorem 1.2 is the following semiclassical statement that is of interest in itself. The first version is formulated under general conditions but without an effective error term. In our proof we actually use the second version that has a quantitative error term. Theorem 1.3. Let Th (A) = [σ · (hp + A)]2 or Th (A) = (hp + A)2 , h ≤ 1, and V ≥ 0. 1) If V ∈ L 5/2 (R3 ) ∩ L 4 (R3 ), then
Tr [Th (A) − V ]−+ h −2 B2 ≥ Tr −h 2 − V +o(h −3 ) as h → 0. (1.11) R3
−
2) Assume that V ∞ ≤ K with some 1 ≤ K ≤ Ch −2 and consider the operators with Dirichlet boundary condition on ⊂ R3 . Let B R√denote the ball of radius R about the origin and let √h := + B√h denote the h-neighborhood of the set . We set | √h | for the Lebesgue measure of √h . Then B2 Tr [(Th (A) − V ) ]− + h −2 3 R
≥ Tr (−h 2 − V ) −
1/2
1/2 −3 5/2 √ 3/2 3/2 1 + hK . (1.12) −Ch K | h | h K
234
L. Erd˝os, J. P. Solovej
Remark. Despite the electrons being confined to , their motion generates a magnetic field in the whole R3 , so the magnetic field energy in (1.12) is given by integration over R3 . We use the convention that letters C, c denote positive universal constants whose values may change from line to line. 2. Reduction to the Main Lemmas Proof of Theorem 1.2. We focus on the lower bound, the upper bound is trivial. We start with two localizations, one on scale r ≥ Z −1/3 and the other one on scale d ≤ Z −1/3 . The first one is designed to address the difficulty that the H 1 -norm of A is available only locally around the nucleus. This step would not be needed for the direct proof of (1.4). The second localization removes the “Coulomb tooth”, i.e. the Coulomb singularity near the nucleus. In this section we reduce the proof of the lower bound in Theorem 1.2 to two lemmas. Lemma 2.1 will show that the Coulomb tooth is indeed negligible. Lemma 2.2 shows that the magnetic field cannot substantially lower the energy for the problem without the Coulomb tooth. In the proof of Lemma 2.2 we will use Theorem 1.3. Recall that B R denotes the ball of radius R about the origin. We construct a pair of smooth cutoff functions satisfying the following conditions: θ02 + θ12 ≡ 1, supp θ1 ⊂ B2d , θ1 ≡ 1 on Bd , |∇θ0 |, |∇θ1 | ≤ Cd −1 . We will choose d = δ Z −1/3
(2.13)
with some δ ≤ 1, in particular d ≤ r . We split the Hamiltonian as H N ,Z (A) = H N0 ,Z (A) + H N1 ,Z (A) with H N0 ,Z (A)
N
Z 2 2 θ0 T (A) − − |∇θ0 | + |∇θ1 | θ0 = |x| i i=1 1 1 + + |∇ ⊗ A|2 , |xi − x j | 16π α 2 B3r
(2.14)
i< j
H N1 ,Z (A) =
N
Z θ1 T (A) − − |∇θ0 |2 + |∇θ1 |2 θ1 |x| i i=1 1 + |∇ ⊗ A|2 , 16π α 2 B3r
where we used the IMS localization formula that is valid for the Pauli operator as well as for the Schrödinger operator. In Sect. 3 we deal with H N1 ,Z , to prove that it is negligible:
Energy of Atoms in a Self-Generated Field
235
Lemma 2.1. There is a positive universal constant κ such that for any Z , α with Z α 2 ≤ κ we have inf inf H N1 ,Z (A) ≥ −C Z 7/3 δ 1/2 − Z 2/3 δ −2 N A∈A0
if C Z −2/3 ≤ δ ≤ D with a sufficiently large constant C. Starting Sect. 4 we will treat H N0 ,Z (A) and we prove the following: Lemma 2.2. There is a positive universal constant κ such that for any Z , α with Z α 2 ≤ κ we have
inf inf H N0 ,Z (A) ≥ −cTF Z 7/3 − C Z 7/3 Z −1/30 + D −1 (2.15) N
A
with a sufficiently large constant C if Z −1/6 ≤ δ ≤ 1 and D ≤ Z 1/24 δ 13/16 . The main ingredient in the proof is Theorem 1.3 that will be proven in Sect. 5. The proof of the lower bound in Theorem 1.2 then follows from Lemmas 2.1 and 2.2 after choosing δ = Z −2/63 . 3. Estimating the Coulomb Tooth Proof of the Lemma 2.1. Let χ 0 be a smooth cutoff function supported on B3r such that |∇ χ0 | ≤ Cr −1 and χ 0 ≡ 1 on B2r . Let A := |B3r |−1 B3r A. We define A0 := (A − A ) χ0 ,
B0 := ∇ × A0 ,
(3.16)
0 ∇ ⊗ A + (A − A ) ⊗ ∇ χ0 . Clearly then ∇ ⊗ A0 = χ 2 2 2 2 −2 B0 ≤ |∇ ⊗ A0 | ≤ 2 χ 0 |∇ ⊗ A| + Cr (A − A )2 R3 R3 R3 B3r ≤ C1 |∇ ⊗ A|2 (3.17) B3r
for some universal constant C1 , where in the last step we used the Poincaré inequality. Let ϕ be a real phase such that ∇ϕ = A . Since χ 0 ≡ 1 on the support of θ1 by D ≥ δ, we have θ1 T (A)θ1 = θ1 e−iϕ T (A − A )eiϕ θ1 = θ1 e−iϕ T (A0 )eiϕ θ1 . After these localizations, we have H N1 ,Z (A) ≥
N
θ1 e−iϕ (T (A0 ) − W (x)) eiϕ θ1 + j
j=1
with
W (x) =
Z + Cd −2 1(|x| ≤ 2d). |x|
1 2C1 α 2
B20
(3.18)
236
L. Erd˝os, J. P. Solovej
Now we use the “running energy scale” argument in [LLS]. ∞ N
θ1 e−iϕ [T (A0 ) − W ] eiϕ θ1 ≥ − N−e (T (A0 ) − W )de j=1
j
µ
≥−
0 µ
≥− 0
N−e (T (A0 ) − W )de −
0
∞
µ ∞
N0
N−e (T (A0 ) − W )de −
µ
µ
N0
e
T (A0 ) − W + e de
e2 e T (A0 ) − W + µ µ
de,
(3.19)
where N−e (A) denotes the number of eigenvalues of a self-adjoint operator A below −e. In the first term we use the bound T (A0 ) ≥ (p + A0 )2 − |B0 | and the CLR bound: µ µ 3/2 N−e (T (A0 ) − W )de ≤ C de (W + |B0 | − e)+ R3 0 0 µ 3/2 ≤C de (W − e/2)+ 3 0 µ R 3/2 +C de (|B0 | − e2 /2µ)+ 3 R 0 5/2 ≤C W + Cµ1/2 B20 R3 R3 5/2 1/2 −2 1/2 = C Z d + Cd + Cµ B20 . (3.20) R3
In the second term of (3.19) we use 1 e 2eZ 2 (p + A0 ) − 1(|x| ≤ 2d) T (A0 ) − W ≥ µ 2 µ|x| 1 Ce + (p + A0 )2 − |B0 | − 1(|x| ≤ 2d), 2 µd 2 and that (p + A0 )2 − i.e. T (A0 ) −
2eZ 4eZ 2eZ 2 , 1(|x| ≤ 2d) ≥ (p + A0 )2 − ≥− µ|x| µ|x| µ
(3.21)
2 1 e eZ Ce W ≥ (p + A0 )2 − 2 − |B0 | − 1(|x| ≤ 2d). µ 2 µ µd 2
We choose µ = 4Z 2 , then using Ce/µd 2 ≤ e2 /4µ for µ ≤ e (i.e. C ≤ (δ Z 2/3 )2 ), we get ∞ ∞ 1 e2 e e2 2 de ≤ (p + A0 ) − |B0 | + de N0 T (A0 ) − W + N0 µ µ 2 4µ µ µ µ 3/2 ≤C de (|B0 | − e2 /4µ)+ 3 0 R 1/2 ≤ Cµ B20 . (3.22) R3
Energy of Atoms in a Self-Generated Field
237
Note that if Z α 2 ≤ κ with some sufficiently small universal constant κ, then the magnetic energy terms in (3.20) and (3.22) can be controlled by the corresponding term in (3.18). Combining the estimates (3.18), (3.19), (3.20) and (3.22) we obtain H N1 ,Z (A) ≥ −C Z 5/2 d 1/2 − Cd −2 and Lemma 2.1 follows.
(3.23)
4. Removing the Magnetic Field Proof of Lemma 2.2. We start with two preparations. In Sect. 4.1 we give an upper bound for the number of electrons N in the truncated model described by H N0 ,Z (A). In Sect. 4.2 we then reduce the problem to a one-body semiclassical statement on boxes. The semiclassical problem will be investigated in Sect. 5 and this will complete the proof of Lemma 2.2. 4.1. Upper bound on the number of electrons N . Let N
E 0N ,Z (A) := inf , H N0 ,Z (A) : ∈ C0∞ (R3 ) ⊗ C2 , = 1 1
be the ground state energy of the truncated Hamiltonian H N0 ,Z (A) defined in (2.14). The following lemma shows that we can assume N ≤ C Z when taking the infimum over N in (2.15). The proof is a slight modification of the proof of the Ruskai-Sigal theorem as presented in [CFKS]. We note that the original proof was given for the non-magnetic case and it can be trivially extended to the Schrödinger operator with a magnetic field but not to the Pauli operator. This is because a key element of the proof, the standard lower bound on the hydrogen atom, − − Z /|x| ≥ −Z 2 /4, is valid if − = p2 replaced by (p + A)2 but there is no lower bound for the ground state energy of the hydrogen atom with the Pauli kinetic energy that is independent of the magnetic field. However, for the truncated Coulomb potential the trivial lower bound can be used. Lemma 4.1. There exist universal constants c and C such that for any fixed A ∈ A0 and Z we have E 0N ,Z (A) = E 0N −1,Z (A) whenever N ≥ C Z and Z −1/6 ≤ δ ≤ c. In particular inf inf E 0N ,Z (A) = inf N A∈A0
inf E 0N ,Z (A)
N ≤C Z A∈A0
(4.24)
if Z −1/6 ≤ δ ≤ c. Proof. We mostly follow the proof of Theorem 3.15 in [CFKS] and we will indicate only the necessary changes. For any x = (x1 , x2 , . . . x N ) ∈ R3N we define x∞ (x) := max{|xi |, : i = 1, 2, . . . , N }, A0 := {x : |x j | < ∀ j = 1, 2, . . . N }, Ai := x : |xi | ≥ (1 − ζ )x∞ (x), x∞ (x) > 2
238
L. Erd˝os, J. P. Solovej
for some fixed positive and ζ < 1/2 to be chosen later. According to Lemma 3.16 in [CFKS], there is partition of unity {Ji }i=0,1,...N , with i Ji2 ≡ 1, supp Ji ⊂ Ai such that the gradient estimates L(x) =
N
|∇ Ji (x)|2 ≤
C N 1/2 2
|∇ Ji (x)|2 ≤
C N 1/2 x∞ (x)
i=0
L(x) =
N i=0
if x ∈ A0 , if x ∈ A j , j ≥ 1
hold with a suitable universal constant C. Moreover, J0 is symmetric in all variables, while Ji , i ≥ 1, is symmetric in all variables except xi . We subtract the local field energy that is an irrelevant constant, i.e. define 1 0 H N := H N ,Z (A) − |∇ ⊗ A|2 16π α 2 B3r and E N = inf Spec H N . We will show that E N = E N −1 for N ≥ C Z . By removing one electron to infinity, clearly E N ≤ E N −1 ≤ 0; here we used the fact that A is compactly supported. By the IMS localization H N = J0 (H N − L)J0 +
N
Ji (H N − L)Ji .
(4.25)
i=1
In the first term we use that on the support of θ0 we have −Z |x|−1 ≥ −Z d −1 . Hence N (N − 1) C N 1/2 −1 −2 J0 . (4.26) − J0 (H N − L)J0 ≥ J0 −C Z N d − C N d + 4 2 Choosing = 8d we see that J0 (H N − L)J0 ≥ 0 if N ≥ C Z with a constant C if δ ≥ C Z −2/3 . To estimate the terms Ji (H N − L)Ji for i = 0, we define H N(i)−1
N
Z 1 2 2 θ0 T (A) − − |∇θ0 | + |∇θ1 | . θ0 + := |x| |x − xj| k j j=1 k< j j =i
k, j =i
On the support of Ji we have |xi | ≥ /4 = 2d, so ∇θ0 and ∇θ1 vanish. Then we can estimate N −1 C N 1/2 Z (i) + − Ji Ji (H N − L)Ji ≥ Ji H N −1 − |xi | 2x∞ (x) x∞ (x) C N 1/2 Z 1/3 1 N −1 (1 − ζ ) − Z − Ji ≥ Ji E N −1 + |xi | 2 δ ≥ Ji E N −1 Ji (4.27) if N ≥ C Z and N is large. Thus we conclude from (4.25), (4.26) and (4.27) that E N ≥ E N −1 if N ≥ C Z .
Energy of Atoms in a Self-Generated Field
239
4.2. Reduction to a one-body problem. We start by presenting an abstract lemma whose proof is given in Appendix A. Lemma 4.2. Let h be a one-particle operator on H = L 2 (R3 ) and let W be a twoparticle operator defined on H ∧ H. We assume that the domains of h and W include the C0∞ functions. Let θ ∈ C ∞ (R3 ) with compact support := supp θ . Then ⎧ ⎡ ⎫ ⎤ N N ⎨ ⎬ inf , ⎣ θi hi θi + θi θ j Wi j θ j θi ⎦ : ∈ C0∞ (R N ), = 1 ⎩ ⎭ 1 i=1 1≤i< j≤N ⎧ ⎛ ⎫ ⎞ n n ⎨ ⎬ ≥ inf inf , ⎝ hi + Wi j ⎠ : ∈ C0∞ ( ), = 1 , ⎩ ⎭ n≤N i=1
1
1≤i< j≤n
(4.28) where hi denotes the operator h acting on the component of the tensor product, and similar convention is used for the two-particle operators. The same result holds with obvious changes if H = L 2 (R3 ) ⊗ C2 . i th
To continue the proof of Lemma 2.2, we first localize H N0 ,Z (A) onto a ball Br of radius r = D Z −1/3 (see (1.8)) and we also localize the magnetic field as in Sect. 3. We introduce smooth cutoff functions χ0 and χ1 with χ02 + χ12 ≡ 1, supp χ0 ⊂ B2r , χ0 ≡ 1 on Br , |∇χ0 |, |∇χ1 | ≤ Cr −1 . We get H N0 ,Z (A)
≥
N
θ0 χ0 e
i=1
1 + 2 Cα
−iϕ0
1 Z Ti (A0 ) − eiϕ0 χ0 θ0 + |xi | |xi − x j | i< j
R3
B20 − C N d −2 − C N Zr −1
(4.29)
using that the new localization error |∇χ1 |2 + |∇χ0 |2 ≤ Cr −2 ≤ Cd −2 and that −Z /|x| ≥ −Zr −1 on the support of χ1 . We also used (3.17). Let Ad,r = {x : d ≤ |x| ≤ r } ⊂ R3 . Using (4.24), the positivity of the Coulomb repulsion |xi − x j |−1 > 0 and Lemma 4.2 with θ := θ0 χ0 we obtain ⎧ ⎪ ⎨
⎡
inf inf H N0 ,Z (A) ≥ inf inf inf , ⎣ N A∈A0 N ≤C Z A0 ∈A0 ⎪ ⎩
+
⎫ ⎪ ⎬
⎤ 1 Z ⎦ T (A0 ) − + |x| i |x − x j | i=1 i< j i Ad,r
N
1 B2 − C Z 5/3 δ −2 − C Z 7/3 D −1 , Cα 2 R3 0 ⎪ ⎭
(4.30)
where the infimum is over all antisymmetric wave functions ∈ 1N C0∞ (Ad,r ) ⊗ C2 with 2 = 1. The notation [H ] Q indicates the N -particle operator H with Dirichlet boundary condition on the domain Q N ⊂ R3N . We define f (x)g(y) 1 dxdy. D( f, g) := 2 R3 ×R3 |x − y|
240
L. Erd˝os, J. P. Solovej
Lemma 4.3. There is a universal constant C0 > 0 such that for any ∈ 1N C0∞ (R3 )⊗ C2 with 2 = 1, for any nonnegative function : R3 → R with D(, ) < ∞, for any A ∈ A0 , and for any ε > 0 we have
⎡
⎤
1 ⎦ + C0 Ti (A) + B2 , ⎣ε 3 |xi − x j | R i=1 i< j N
∗ |xi |−1 − Cε−1 N . ≥ −D(, ) + , N
(4.31)
i=1
Proof. By the Lieb-Oxford inequality [LO] and by the positivity of the quadratic form D(·, ·), 1 4/3 ≥ D( , ) − C , 3 |xi − x j | R i< j N 4/3 ≥ −D(, ) + , ( ∗ |xi |−1 ) − C ,
(4.32)
R3
i=1
where (x) is the one-particle density of . The error term is controlled by the following kinetic energy inequality for the Pauli operator
% ,
N
&
T (A)i ≥ c
i=1
5/3
R3
4/3
min{ , γ } − γ
B2
(4.33)
R3
with some positive universal constant c and for any γ > 0. For the proof of (4.33) use the magnetic Lieb-Thirring inequality
%
& N , [T (A) − U ]i ≥ −C i=1
U R3
5/2
− Cγ
−3
U −γ 4
R3
R3
B2 .
With the choice U = β min{ , γ } we can ensure that 21 U ≥ CU 5/2 +Cγ −3 U 4 if β is sufficiently small (independent of γ ) and this proves (4.33). Thus 4/3 5/3 4/3 ≤ γ −1 min{ , γ } + γ R3 R3 R3 % N & −1 ≤ (cγ ) T (A)i + c−1 B2 + γ N (4.34) , 2/3
1/3
i=1
R3
so choosing γ = Cε−1 with a sufficiently large constant C, we obtain (4.31).
Energy of Atoms in a Self-Generated Field
241
Using Lemma 4.3 we can continue the estimate (4.30) (writing A instead of A0 in the infimum) as inf inf H N0 ,Z (A) N
A
⎧ % & N ⎨ [T (A) + W ]i ≥ (1 − ε) inf inf inf , N ≤C Z A∈A0 ⎩ i=1
Ad,r
1 + Cα 2
B2 R3
−D(, ) − Cε−1 Z − C Z 5/3 δ −2 − C Z 7/3 D −1 with W (x) :=
⎫ ⎬ ⎭
(4.35)
Z 1 − + ∗ |x|−1 1−ε |x|
and assuming that α ≤ α0 with some small universal α0 . We now perform a rescaling: x = Z −1/3 X , p = Z 1/3 P and A(X ) = Z −2/3 A(Z −1/3 X ),
B(X ) = ∇ × A(X ) = Z −1 B(Z −1/3 X ).
A) := [(hP + A) · σ ]2 , we obtain that the kinetic energy Introducing h = Z −1/3 and Th ( changes as A) · σ ]2 = Z 4/3 Th ( A) [(p + A) · σ ]2 = Z 4/3 [(Z −1/3 P + and the field energy changes as 2 B (x)dx = Z R3
R3
B2 (X )dX.
The new potential energy is (X ) = Z −4/3 W (Z −1/3 X ) = W
1 1 − + ∗ |X |−1 , 1−ε |X |
where (X ) = Z −2 (Z −1/3 X ) and D( , ) = Z −7/3 D(, ). After rescaling, we get from (4.35), inf inf H N0 ,Z (A) N A∈A0
⎧ % & N ⎨ 4/3 ]i ≥ (1 − ε)Z inf inf inf , [Th ( A) + W N ≤C Z A∈A0 ⎩ i=1
+ Aδ,D
−Z 7/3 D( , ) − Cε−1 Z − C Z 5/3 δ −2 − C Z 7/3 D −1 ,
h −2 C Z α2
B2
R3
⎫ ⎬ ⎭
(4.36)
where Aδ,D = {X : δ ≤ |X | ≤ D} and inf denotes the infimum over all normalized antisymmetric functions. Using (1.12) from Theorem 1.3 and the fact that Z α 2 ≤ κ, we get
) Aδ,D − C Z 13/6 D 3 δ −13/4 inf inf H N0 ,Z (A) ≥ (1 − ε)Z 4/3 Tr (−h 2 + W N
A
−
−Z 7/3 D( , ) − Cε−1 Z − C Z 5/3 δ −2 − C Z 7/3 D −1 , (4.37)
242
L. Erd˝os, J. P. Solovej
assuming δ ≥ Z −2/9 . By a standard semiclassical result for Coulomb-like potentials (see e.g. the result in Sect. V.2 of [L]):
2 2 −3 −5/2 − Ch −3+1/10 , ≥ Tr −h + W ≥ −Csc h Tr (−h + W ) Aδ,D W −
−
R3
(4.38) 1 where Csc = 2/(15π 2 ) is the Weyl constant in semiclassics. The 10 exponent in the error term is far from being optimal; the methods developed to prove the Scott correction can yield an exponent up to one (see Remark 1 after Theorem 1.1). Taking the optimal to be the Thomas-Fermi density for Z = 1 = TF (see, e.g. Sect. II of [L]) and defining the Thomas-Fermi constant as
5/2 1 −1 , − + TF ∗ |X | |X | R3 −
cTF := D(TF , TF ) + C SC we get ' inf inf N
A
H N0 ,Z (A)
≥ (1 − ε)
−3/2
Z
7/3
−D( , ) − Csc
5/2 ( 1 −1 − + ∗ |X | |X | R3 −
−C Z 7/3−1/30 − C Z 13/6 D 3 δ −13/4 − Cε−1 Z − C Z 5/3 δ −2 − C Z 7/3 D −1 ≥ −(1 − ε)−3/2 cTF Z 7/3 − C Z 7/3−1/30 − Cε−1 Z − C Z 5/3 δ −2 − C Z 7/3 D −1
(4.39) ≥ −cTF Z 7/3 − C Z 7/3 Z −1/30 + D −1 , where we optimized for ε and we used that D ≤ Z 1/24 δ 13/16 and Z −1/6 ≤ δ ≤ 1. This completes the proof of Lemma 2.2. 5. Semiclassics: Proof of Theorem 1.3 We present the Schrödinger and Pauli cases in parallel. We prove the statement with Dirichlet boundary conditions (1.12) in detail and in Sect. 5.4 we comment on the necessary changes for the proof of the (1.11). The potential V is defined only on , but we extend it to be zero on R3 \ ; we continue to denote by V its extension. 5.1. Localization onto boxes. We choose a length L with h ≤ L ≤ 13 h 1/2 . Let L = + B L be the L-neighborhood of . Let Q k = {y ∈ R3 : y − k∞ < L/2} with k ∈ (LZ)3 ∩ L denote a non-overlapping covering of with boxes of size L. In this sectionthe index k will always run over the set (LZ)3 ∩ L . Let ξk be a partition of unity, k ξk2 ≡ 1, subordinated to the collection of boxes Q k , such that supp ξk ⊂ (2Q)k ,
|∇ξk | ≤ C L −1 ,
where (2Q)k denotes the cube of side-length 2L with center k. Let ξk be a cutoff funck := (3Q)k and tion such that ξk ≡ 1 on (2Q)k (i.e. on the support of ξk ), supp ξk ⊂ Q |∇ ξk | ≤ C L −1 .
Energy of Atoms in a Self-Generated Field
k |−1 Let A k = | Q Bk := ∇ × Ak , then
k Q
243
A. Similarly to (3.16), we define Ak := (A − A k ) ξk and
R3
B2k ≤ C
k Q
|∇ ⊗ A|2
(5.40)
as in (3.17). From the IMS localization with ψk satisfying h∇ψk = Ak we have ∗ −2 2 −2 Tr [[Th (A) − V ] ]− + h B = inf Tr (γ [Th (A) − V ]) + h |∇ ⊗ A|2 R3
γ
≥ inf γ
R3
∗
Ek (γ )
k∈(LZ)3 ∩
(5.41)
L
with
Ek (γ ) := Tr γ ξk e−iψk [Th (A − A k ) − V ]eiψk ξk − γ |h∇ξk |2 + c0 h −2
k Q
|∇ ⊗ A|2
with some universal constant c0 . Here inf ∗γ denotes infimum over all density matrices 0 ≤ γ ≤ 1 that are supported on , i.e. they are operators on L 2 ( ) ⊗ C2 . We also used R3 B2 = R3 |∇ ⊗ A|2 and we reallocated the second integral. We introduce the notation −2 Fk := c0 h |∇ ⊗ A|2 . k Q
5.2. A priori bound on the local field energy. In case of the Pauli operator, for each fixed k we apply the magnetic Lieb-Thirring inequality [LLS] together with (5.40) and box Q with the bound V ∞ ≤ K to obtain that for any density matrix γ ,
Ek (γ ) ≥ Tr [Th (Ak ) − V − Ch 2 L −2 ] Qk + Fk − ≥ −Ch −3 [V + Ch 2 L −2 ]5/2 k Q
−C
k Q
[V + Ch 2 L −2 ]4
1/4
h −2
3/4
k Q
B2k
+ Fk
c 0 −2 h ≥ −C h −3 K 5/2 L 3 + h 2 L −2 + K 4 L 3 + h 8 L −5 − |∇ ⊗ A|2 + Fk 2 k Q 1 ≥ −Ch −3 K 5/2 L 3 + Fk , (5.42) 2 using h ≤ L and 1 ≤ K ≤ Ch −2 . In the Schrödinger case we use the usual LiebThirring inequality [LT] that holds with a magnetic field as well. The estimate (5.42) is then valid even without the third term in the second line. Let S ⊂ (LZ)3 ∩ L denote the set of those k indices such that Fk ≤ Ch −3 K 5/2 L 3
(5.43)
holds with some large constant C. In particular Ek (γ ) ≥ 0, for all k ∈ S and for any γ .
(5.44)
244
L. Erd˝os, J. P. Solovej
5.3. Improved bound. We use the Schwarz inequality in the form Th (A − A k ) ≥ −(1 − εk )h 2 − Cεk−1 (A − A k )2 , with some 0 < εk < 13 . We have for any γ supported on that
Ek (γ ) ≥ Tr 1 ξk [−(1 − 2εk )h 2 − V − Ch 2 L −2 ]ξk 1 −
−1 2 2 +Tr 1 Qk [−εk h − Cεk (A − A k ) ]1 Qk + Fk .
(5.45)
We will show at the end of the section that
Tr 1 ξk [−(1 − 2εk )h 2 − V − Ch 2 L −2 ]ξk 1 −
2 −3 5/2 k |. ≥ Tr 1 ξk (−h − V )ξk 1 − Ch K (εk + h 2 L −2 )| Q
(5.46)
−
−
Using (5.44) and (5.46), ∗ ∗ inf Ek (γ ) ≥ inf Ek (γ ) γ
γ
k
≥
k
≥
k∈S
k
Tr 1 ξk [−h 2 − V ]ξk 1 + Dk −
k∈S
inf Tr ξk γk ξk 1 [−h 2 − V ]1 + Dk γk
≥ Tr (−h 2 − V ) + Dk −
k∈S
(5.47)
k∈S
with
k |(εk + h 2 L −2 ) + Fk . Dk := Tr [−εk h 2 − Cεk−1 (A − A k )2 ] Qk − Ch −3 K 5/2 | Q −
(5.48) In the last step in (5.47) we used that for any collection of density matrices γk , the density matrix k 1 ξk γk ξk 1 is admissible in the variational principle
Tr (−h 2 − V ) = inf Tr γ −h 2 − V : 0 ≤ γ ≤ 1, supp γ ⊂ . −
(5.49) We estimate Dk for k ∈ S as follows: −4 −3 k |(εk + h 2 L −2 ) + Fk Dk ≥ −Cεk h (A − A k )5 − Ch −3 K 5/2 | Q k Q
≥ Fk − Cεk−4 h 2 L 1/2 Fk
5/2
k |(εk + h 2 L −2 ). − Ch −3 K 5/2 | Q
(5.50)
In the first step we used the Lieb-Thirring inequality, in the second step the Hölder and Sobolev inequalities in the form 5/2 5 1/2 2 (A − A k ) ≤ C L |∇ ⊗ A| . k Q
k Q
Energy of Atoms in a Self-Generated Field
245
We choose εk = h L −1/2 K −1/2 Fk , 1/2
and using the a priori bound (5.43), we see that εk ≤ Ch −1/2 L K 3/4 . Thus, assuming L ≤ ch 1/2 K −3/4
(5.51)
with a sufficiently small constant c, we get εk ≤ 1/3. With this choice of εk , and recalling k | = 9|Q k | = 9L 3 , we have |Q Dk ≥ Fk − Ch −2 L 5/2 K 2 Fk − Ch −3 K 5/2 L 3 h 2 L −2
≥ −Ch −3 L 3 K 5/2 h −1 L 2 K 3/2 + h 2 L −2 . 1/2
(5.52)
If we choose L = h 3/4 K −3/8 , then
1/2 Dk ≥ −Ch −3 L 3 K 5/2 h K 3/2 . This choice is allowed by (5.51) if K ≤ ch −2/3 . If ch −2/3 ≤ K ≤ h −2 , then we choose L = ch 1/2 K −3/4 and we get from (5.52), Dk ≥ −Ch −3 L 3 K 5/2 (1 + h K 3/2 ). Combining these two inequalities, we get that
1/2
1/2 Dk ≥ −Ch −3 L 3 K 5/2 h K 3/2 1 + h K 3/2
(5.53)
always holds. Summing up (5.53) for all k and using that L 3 ≤ C| 3L | ≤ C| √h | k∈(LZ)3 ∩ L
(recall that √h is a and (5.53), ∗
inf γ
√
k
h-neighborhood of and 3L ≤ h 1/2 ), we obtain from (5.47)
Ek (γ ) ≥ Tr (−h 2 − V )
−
1/2
1/2 1 + h K 3/2 , −Ch −3 K 5/2 | √h | h K 3/2
(5.54)
and this proves (1.12). Finally, we prove (5.46). Let γ be a trial density matrix for the left hand side of (5.46). We can assume that
0 ≥ Tr γ 1 ξk [−(1 − 2εk )h 2 − V − Ch 2 L −2 ]ξk 1 .
246
L. Erd˝os, J. P. Solovej
Then
1 1 0 ≥ Tr γ 1 ξk [− h 2 + K ]ξk 1 + Tr γ 1 ξk [− h 2 −V −Ch 2 L −2 − K ]ξk 1 6 6 1 2 ≥ Tr γ 1 ξk [− h + K ]ξk 1 − Ch −3 [V + K + Ch 2 L −2 ]5/2 , (5.55) 6 k Q
where we used the Lieb-Thirring inequality. Thus, using |V | ≤ K , h ≤ L and K ≥ 1, we have 1 2 k |. Tr γ 1 ξk [− h + K ]ξk 1 ≤ Ch −3 K 5/2 | Q 6 Therefore
Tr γ 1 ξk [−(1 − 2εk )h 2 − V − Ch 2 L −2 ]ξk 1
k |. ≥ Tr γ 1 ξk (−h 2 − V )ξk 1 − Ch −3 K 5/2 (εk + h 2 L −2 )| Q
Now (5.46) follows by variational principle.
(5.56)
5.4. Reduction of (1.11) to (1.12). We approximate V ∈ L 5/2 ∩ L 4 by a bounded , V ∞ ≤ K , that is supported on a ball B R/2 and V ≤ V . By choosing K potential V and R sufficiently large, we can make V − V 5/2 +V − V 4 arbitrarily small. We choose a cutoff function χ R that is supported on B R , χ R ≡ 1 on B R/2 and |∇χ R | ≤ C R −1 and R2 ≡ 1. let χ R satisfy χ R2 + χ Borrowing a small part of the kinetic energy, by IMS localization we have ]χ R Th (A) − V ≥ (1 − ε)χ R [Th (A) − V ) − |∇χ R |2 − |∇ +εTh (A) − (V − (1 − ε)V χ R |2 .
(5.57)
Using the magnetic Lieb-Thirring inequality [LLS] to estimate the second term, we get ) B R ]− Tr [Th (A) − V ]− ≥ (1 − ε)Tr [(Th (A) − V −3/2 −3 5/2 −Cε h |U | − C
1 |U | − h −2 3 2 R
R3
4
B2 (5.58) R3
with ) + |∇χ R |2 + |∇ U := (V − (1 − ε)V χ R |2 . For 2the first term in (5.58) we use (1.12) (and that it holds even with a 1/2 in front of B ) and the fact that
) B R Tr (−h 2 − V ≥ Tr −h 2 − V −
−
≤ V . The second and the third terms in (5.58) can be made arbiby monotonicity, V trarily small compared with h −3 for any fixed ε if R and K are sufficiently large and h is small. Finally, choosing ε sufficiently small, we proved (1.11).
Energy of Atoms in a Self-Generated Field
247
6. Proof of the Quantized Field Case For the proof of the lower bound in (1.6), we follow the argument of [BFG] to reduce the problem to the classical bound (1.10). We set |g(k)|2 |k| aλ (k)∗ aλ (k)dk Hg = α −1 R3
λ=±
to be the cutoff field energy, then H f ≥ Hg and only the modes appearing in Hg interact with the electron. By Lemma 3 of [BFG], for any real function f ∈ L 1 (R3 ) ∩ L ∞ (R3 ) we have 1 f (x)|∇ ⊗ A(x)|2 dx ≤ α 2 f ∞ Hg + Cα4 f 1 . 8π R3 Applying it with f being the characteristic function of the ball B3r with r = D Z −1/3 (the radius of the ball is here chosen differently from [BFG]) and using Z α 2 ≤ κ we get 2 Zα Z H f ≥ Hg ≥ Hg ≥ |∇ ⊗ A(x)|2 dx − Cκ −1 α4 D 3 . κ 8π κ B3r Setting α = (κ/Z )1/2 , i.e. Z α 2 = κ, we have for κ sufficiently small, −1 4 3 E N ,Z ≥ E N ,Z , α − Cκ α D , qed
where E N ,Z , α is the ground state energy of the Hamiltonian (1.7) with fine structure constant α . Applying (1.10) to this Hamiltonian, we get qed
EZ
7
−1 3 ≥ E nf − Cκ −1 α4 D 3 Z − CZ D
1
whenever 1 ≤ D ≤ Z 63 . Writing D = Z γ and applying the upper bound (1.5) on , we obtain the lower bound in (1.6). A. Proof of Lemma 4.2 Let the function χ (x) ∈ C ∞ (R3 ) be defined such that θ 2 (x)+χ 2 (x) ≡ 1. For any subset α ⊂ {1, 2, . . . , N } we denote by xα the collection of variables {xi : i ∈ α} and define ) ) α = α (xα ) := θ (xi ), α = α (xα ) := χ (xi ). i∈α
i∈α
We set the notation α c = {1, 2, . . . , N } \ α for the complement of the set α and set n := {1, 2, . . . , n}. Let |α| denote the cardinality of the set α. For an arbitrary function ∈ 1N C0∞ (R3 ), = 1, and for 0 ≤ n ≤ N we define * + , n := n Tr n c n c | |n c n , where Tr n c denotes taking the partial trace for the xn+1 , xn+2 , . . . , x N variables. Define .N the fermionic Fock space as F = n=0 Hn with Hn := n H and we define a density matrix N N n |α| on F. := = n α⊂{1,2,...,N }
n=0
248
L. Erd˝os, J. P. Solovej
We first prove that ≤ I on F. It is sufficient to show that ≤ I on the n-particle sectors for each n. Let n ≤ N , choose ∈ Hn , and compute , |α| = dxα dxα (xα ) n (xα , xα )(xα ) α⊂{1,2,...N } |α|=n
=
α⊂{1,2,...N } |α|=n
dxα dxα dyα c (xα )α (xα )α c (yα c ) (xα , yα c )
α
× (xα , yα c )α c (yα c )α (xα )(xα ) ≤ dxα dxα dyα c 2α (xα )2α c (yα c )| (xα , yα c )|2 |(xα )|2 α
=
22
dx| (x)|2
α
= 22 ,
2α (xα )2α c (xα c )
/N
(A.59)
using Schwarz inequality and that 1 ≡ j=1 [θ 2 (x j )+χ 2 (x j )] = α 2α (xα )2α c (xα c ). Second, for a fixed n ≤ N , we compute % & % & N N n n 0 N n Tr F Tr F hi = hi n n=0 i=1 n=0 i=1 = dxα dxα c (xα , xα c )α (xα )α c (xα c )hi (α (xα )α c (xα c ) (xα , xα c )) α
=
N
i∈α
i=1 α : i∈α
=
N
dx (xα , xα c )2α\{i} (xα\{i} )2α c (xα c )θ (xi )hi (θ (xi ) (xα , xα c ))
, θi hi θi ,
(A.60)
i=1
where the trace on the left/hand side is computed on F. In the last step we used that for 2 2 , where the summation any fixed i, we have 1 ≡ j=i [θ 2 (x j ) + χ 2 (x j )] = 1 α 1 α 1 αc c is over all 1 α ⊂ {1, 2, . . . , N }\{i} and 1 α = {1, 2, . . . , N }\{i}\α. A similar calculation for the two-body potential shows that ⎡ ⎤ N 0 2 3 Tr F ⎣ , θi θ j Wi j θ j θi . Wi j ⎦ = n=0 1≤i< j≤n
1≤i< j≤N
Thus, by the variational principle, ⎛ N , ⎝ θi hi θi + i=1
⎞
n=0
i=1
θi θ j Wi j θ j θi ⎠
1≤i< j≤N
⎛ N n 0 ⎝ ≥ inf Tr F ⎣ hi +
⎡
⎞⎤ Wi j ⎠⎦.
1≤i< j≤n
Since is a density matrix supported on , we obtain (4.28).
Energy of Atoms in a Self-Generated Field
249
References [BFG] [CFKS] [ES]
[ES2] [F] [FLL] [FS] [H] [IS] [L] [LL] [LLS] [LO] [LS] [LSY1] [LSY2] [LT]
[LY] [SW1] [SW2] [SW3] [SS]
Bugliaro, L., Fröhlich, J., Graf, G.M.: Stability of quantum electrodynamics with nonrelativistic matter. Phys. Rev. Lett. 77, 3494–3497 (1996) Cycon, H.L., Froese, R.G., Kirsch, W., Simon, B.: Schrödinger Operators with Application to Quantum Mechanics and Global Geometry. Berlin-Heidelberg-New York: Springer-Verlag, 1987 Erd˝os, L., Solovej, J.P.: Semiclassical eigenvalue estimates for the pauli operator with strong non-homogeneous magnetic fields. II. leading order asymptotic estimates. Commun. Math. Phys. 188, 599–656 (1997) Erd˝os, L., Solovej, J.P.: The kernel of dirac operators on S3 and R3 . Rev. Math. Phys. 13, 1247– 1280 (2001) Fefferman, C.: Stability of coulomb systems in a magnetic field. Proc. Nat. Acad. Sci. USA 92, 5006–5007 (1995) Fröhlich, J., Lieb, E.H., Loss, M.: Stability of coulomb systems with magnetic fields. I. the oneelectron atom. Commun. Math. Phys. 104, 251–270 (1986) Fefferman, C., Seco, L.A.: On the energy of a large atom. Bull. AMS 23(2), 525–530 (1990) Hughes, W.: An atomic energy bound that gives scott’s correction. Adv. Math. 79, 213–270 (1990) Ivrii, V.I., Sigal, I.M.: Asymptotics of the ground state energies of large coulomb systems. Ann. of Math. (2) 138, 243–335 (1993) Lieb, E.H.: Thomas-fermi and related theories of atoms and molecules. Rev. Mod. Phys. 65(4), 603–641 (1981) Lieb, E.H., Loss, M.: Stability of coulomb systems with magnetic fields. II. the many-electron atom and the one-electron molecule. Commun. Math. Phys. 104, 271–282 (1986) Lieb, E.H., Loss, M., Solovej, J.P.: Stability of matter in magnetic fields. Phys. Rev. Lett. 75, 985–989 (1995) Lieb, E.H., Oxford, S.: Improved lower bound on the indirect coulomb energy. Int. J. Quant. Chem. 19, 427–439 (1981) Lieb, E.H., Simon, B.: The thomas-fermi theory of atoms. molecules and solids. Adv. Math. 23, 22–116 (1977) Lieb, E.H., Solovej, J.P., Yngvason, J.: Asymptotics of heavy atoms in high magnetic fields: I. lowest landau band region. Commun. Pure Appl. Math. 47, 513–591 (1994) Lieb, E.H., Solovej, J.P., Yngvason, J.: Asymptotics of heavy atoms in high magnetic fields: II. semiclassical regions. Commun. Math. Phys. 161, 77–124 (1994) Lieb, E.H., Thirring, W.: A bound on the moments of the eigenvalues of the Schrödinger Hamiltonian and their relation to Sobolev inequalities. In: Studies in Mathematical Physics: Essays in Honor of Valentine Bargmann. E. H. Lieb, B. Simon, A. Wightman, eds., Princeton, NJ: Princeton University Press, 1976, pp. 269–303 Loss, M., Yau, H.T.: Stabilty of coulomb systems with magnetic fields. III. zero energy bound states of the pauli operator. Commun. Math. Phys. 104, 283–290 (1986) Siedentop, H., Weikard, R.: On the leading energy correction for the statistical model of an atom: interacting case. Commun. Math. Phys. 112, 471–490 (1987) Siedentop, H., Weikard, R.: On the leading correction of the thomas-fermi model: lower bound. Invent. Math. 97, 159–193 (1990) Siedentop, H., Weikard, R.: A new phase space localization technique with application to the sum of negative eigenvalues of schrödinger operators. Ann. Sci. École Norm. Sup. (4) 24(2), 215–225 (1991) Solovej, J.P., Spitzer, W.L.: A new coherent states approach to semiclassics which gives scott’s correction. Commun. Math. Phys. 241, 383–420 (2003)
Communicated by I. M. Sigal
Commun. Math. Phys. 294, 251–272 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0903-4
Communications in
Mathematical Physics
The Hermitian Laplace Operator on Nearly Kähler Manifolds Andrei Moroianu1 , Uwe Semmelmann2 1 CMLS, École Polytechnique, UMR 7640 du CNRS, 91128 Palaiseau, France.
E-mail:
[email protected]
2 Mathematisches Institut, Universität zu Köln, Weyertal 86-90, D-50931 Köln, Germany.
E-mail:
[email protected] Received: 23 March 2009 / Accepted: 11 May 2009 Published online: 9 August 2009 – © Springer-Verlag 2009
Abstract: The moduli space N K of infinitesimal deformations of a nearly Kähler structure on a compact 6-dimensional manifold is described by a certain eigenspace of the Laplace operator acting on co-closed primitive (1, 1) forms (cf. Moroianu et al. in Pacific J Math 235:57–72, 2008). Using the Hermitian Laplace operator and some representation theory, we compute the space N K on all 6-dimensional homogeneous nearly Kähler manifolds. It turns out that the nearly Kähler structure is rigid except for the flag manifold F(1, 2) = SU3 /T 2 , which carries an 8-dimensional moduli space of infinitesimal nearly Kähler deformations, modeled on the Lie algebra su3 of the isometry group. 1. Introduction Nearly Kähler manifolds were introduced in the 70’s by A. Gray [8] in the context of weak holonomy. More recently, 6-dimensional nearly Kähler manifolds turned out to be related to a multitude of topics among which we mention: Spin manifolds with Killing spinors (Grunewald), SU3 -structures, geometries with torsion (Cleyton, Swann), stable forms (Hitchin), or super-symmetric models in theoretical physics (Friedrich, Ivanov). Up to now, the only sources of compact examples are the naturally reductive 3-symmetric spaces, classified by Gray and Wolf [13], and the twistor spaces over positive quaternion-Kähler manifolds, equipped with the non-integrable almost complex structure. Based on previous work by R. Cleyton and A. Swann [6], P.-A. Nagy has shown in 2002 that every simply connected nearly Kähler manifold is a Riemannian product of factors which are either of one of these two types, or 6-dimensional [12]. Moreover, J.-B. Butruille has shown [5] that every homogeneous 6-dimensional nearly Kähler manifold is a 3-symmetric space G/K , more precisely isometric with S 6 = G 2 /SU3 , S 3 × S 3 = SU2 × SU2 × SU2 /SU2 , CP3 = SO5 /U2 × S 1 or F(1, 2) = SU3 /T 2 , all endowed with the metric defined by the Killing form of G. This work was supported by the French-German cooperation project Procope no. 17825PG.
252
A. Moroianu, U. Semmelmann
A method of finding new examples is to take some homogeneous nearly Kähler manifold and try to deform its structure. In [10] we have studied the deformation problem for 6-dimensional nearly Kähler manifolds (M 6 , g) and proved that if M is compact, and has normalized scalar curvature scalg = 30, then the space N K of infinitesimal deformations of the nearly Kähler structure is isomorphic to the eigenspace for the eigenvalue 12 of the restriction of the Laplace operator g to the space of co-closed (1,1) primitive (1, 1)-forms 0 M. It is thus natural to investigate the Laplace operator on the known 3-symmetric examples (besides the sphere S 6 , whose space of nearly Kähler structures is well-understood, and isomorphic to SO7 /G 2 ∼ = RP7 , see [7] or [5, Prop. 7.2]). Recall that the spectrum of the Laplace operator on symmetric spaces can be computed in terms of Casimir eigenvalues using the Peter-Weyl formalism. It turns out that a similar method can be applied ¯ (called the Hermitian in order to compute the spectrum of a modified Laplace operator Laplace operator) on 3-symmetric spaces. This operator is SU3 -equivariant and coincides with the usual Laplace operator on co-closed primitive (1, 1)-forms. The space of infinitesimal nearly Kähler deformations is thus identified with the space of co-closed ¯ = 12α}. Our main result is that the forms in 0(1,1) (12) := {α ∈ C ∞ (0(1,1) M) | α nearly Kähler structure is rigid on S 3 × S 3 and CP3 , and that the space of infinitesimal nearly Kähler deformations of the flag manifold F(1, 2) is eight-dimensional. The paper is organized as follows. After some preliminaries on nearly Kähler man(1,1) ifolds, we give two general procedures for constructing elements in 0 (12) out of Killing vector fields or eigenfunctions of the Laplace operator for the eigenvalue 12 (Corollary 4.5 and Proposition 4.11). We show that these elements can not be co-closed, thus obtaining an upper bound for the dimension of the space of infinitesimal nearly Kähler deformations (Proposition 4.12). We then compute this upper bound explicitly on the 3-symmetric examples and find that it vanishes for S 3 × S 3 and CP3 , which therefore have no infinitesimal nearly Kähler deformation. This upper bound is equal to 8 on the flag manifold F(1, 2) = SU3 /T 2 and in the last section we construct an explicit isomorphism between the Lie algebra of the isometry group su3 and the space of infinitesimal nearly Kähler deformations on F(1, 2). In addition, our explicit computations (in Sect. 5) of the spectrum of the Hermitian Laplace operator on the 3-symmetric spaces, together with the results in [11] show that every infinitesimal Einstein deformation on a 3-symmetric space is automatically an infinitesimal nearly Kähler deformation. 2. Preliminaries on Nearly Kähler Manifolds An almost Hermitian manifold (M 2m , g, J ) is called nearly Kähler if (∇ X J )(X ) = 0,
∀ X ∈ T M,
(1)
where ∇ denotes the Levi-Civita connection of g. The canonical Hermitian connection ¯ defined by ∇, ∇¯ X Y := ∇ X Y − 21 J (∇ X J )Y,
∀ X ∈ T M, ∀ Y ∈ C ∞ (M)
(2)
¯ = 0 and ∇¯ J = 0) with torsion T¯X Y = −J (∇ X J )Y . is a Um connection on M (i.e. ∇g A fundamental observation, which goes back to Gray, is the fact that ∇¯ T¯ = 0 on every nearly Kähler manifold (see [2]).
The Hermitian Laplace Operator on Nearly Kähler Manifolds
253
We denote the Kähler form of M by ω := g(J., .). The tensor + := ∇ω is totally skew-symmetric and of type (3, 0) + (0, 3) by (1). From now on we assume that the dimension of M is 2m = 6 and that the nearly Kähler structure is strict, i.e. (M, g, J ) is not Kähler. It is well-known that M is Einstein in this case. We will always normalize the scalar curvature of M to scal = 30, in which case we also have | + |2 = 4 point¯ wise. The form + can be seen as the real part of a ∇-parallel complex volume form + + i − on M, where − = ∗ + is the Hodge dual of + . Thus M carries a SU3 ¯ Notice that Hitchin has shown structure whose minimal connection (cf. [6]) is exactly ∇. + − that a SU3 structure (ω, , ) is nearly Kähler if and only if the following exterior system holds: dω = 3 + (3) d − = −2ω ∧ ω. Let A ∈ 1 M ⊗ EndM denote the tensor A X := J (∇ X J ) = − +J X , where Y+ denotes the endomorphism associated to Y + via the metric. Since for every unit vector X , A X defines a complex structure on the 4-dimensional space X ⊥ ∩ (J X )⊥ , we easily get in a local orthonormal basis {ei } the formulas |A X |2 = 2|X |2 , Aei Aei (X ) = −4X,
∀X ∈ T M, ∀X ∈ T M,
(4) (5)
where here and henceforth, we use Einstein’s summation convention on repeating subscripts. The following algebraic relations are satisfied for every SU3 structure (ω, + ) on T M (notice that we identify vectors and 1-forms via the metric): A X ei ∧ ei + X − + (X ) ∧ + (J X + ) ∧ ω
= −2X ∧ ω, ∀X ∈ T M, + = −J X , ∀X ∈ T M, = X ∧ ω2 , ∀X ∈ T M, = X ∧ +, ∀X ∈ T M.
(6) (7) (8) (9)
The Hodge operator satisfies ∗2 = (−1) p on p M and moreover ∗ (X ∧ + ) = J X + , ∗(φ ∧ ω) = −φ,
∀φ ∈
∗(J X ∧ ω ) = −2X, 2
∀X ∈ T M,
(10)
(1,1) 0 M,
(11)
∀X ∈ T M.
(12)
From now on we assume that (M, g) is compact 6-dimensional not isometric to the round sphere (S 6 , can). It is well-known that every Killing vector field ξ on M is an automorphism of the whole nearly Kähler structure (see [10]). In particular, L ξ ω = 0,
L ξ + = 0,
L ξ − = 0.
(13)
¯ Then the formula (cf. [1]) Let now R and R¯ denote the curvature tensors of ∇ and ∇. RW X Y Z = R¯ W X Y Z − 41 g(Y, W )g(X, Z ) + 41 g(X, Y )g(Z , W ) + 43 g(Y, J W )g(J X, Z ) − 43 g(Y, J X )g(J W, Z ) − 21 g(X, J W )g(J Y, Z ) may be rewritten as R X Y = − X ∧ Y + R CY XY
254
A. Moroianu, U. Semmelmann
and R¯ X Y = − 43 (X ∧ Y + J X ∧ J Y − 23 ω(X, Y )J ) + R CY XY , where R CY X Y is a curvature tensor of Calabi-Yau type. We will recall the definition of the curvature endomorphism q(R) (cf. [10]). Let E M be the vector bundle associated to the bundle of orthonormal frames via a representation π : SO(n) → Aut(E). The Levi-Civita connection of M induces a connection on E M, whose curvature satisfies R XE YM = π∗ (R X Y ) = π∗ (R(X ∧ Y )), where we denote with π∗ the differential of π and identify the Lie algebra of S O(n), i.e. the skew-symmetric endomorphisms, with 2 . In order to keep notations as simple as possible, we introduce the notation π∗ (A) = A∗ . The curvature endomorphism q(R) ∈ End(E M) is defined as q(R) = 21 (ei ∧ e j )∗ R(ei ∧ e j )∗
(14)
for any local orthonormal frame {ei }. In particular, q(R) = Ric on T M. By the same formula we may define for any curvature tensor S, or more generally any endomorphism S of 2 T M, a bundle morphism q(S). In any point q : R → q(R) defines an equivariant map from the space of algebraic curvature tensors to the space of endomorphisms of E. Since a Calabi-Yau algebraic curvature tensor has vanishing Ricci curvature, q(R CY ) = 0 holds on T M. Let R 0X Y be defined by R 0X Y = X ∧ Y + J X ∧ J Y − 23 ω(X, Y )J . Then a direct calculation gives (ei ∧ e j )∗ (ei ∧ e j )∗ + 21 (ei ∧ e j )∗ (J ei ∧ J e j )∗ − 23 ω∗ ω∗ . q(R 0 ) = 21 We apply this formula on T M. The first summand is exactly the SO(n)-Casimir, which acts as −5id. The third summand is easily seen to be 23 id, whereas the second summand acts as −id (cf. [11]). Altogether we obtain q(R 0 ) = − 16 3 id, which gives the following ¯ acting on T M: expression for q( R) ¯ T M = 4 id T M . q( R)|
(15)
3. The Hermitian Laplace Operator In the next two sections (M 6 , g, J ) will be a compact nearly Kähler manifold with scalar curvature normalized to scalg = 30. We denote as usual by the Laplace operator = d ∗ d + dd ∗ = ∇ ∗ ∇ + q(R) on differential forms. We introduce the Hermitian Laplace operator ¯ ¯ = ∇¯ ∗ ∇¯ + q( R),
(16)
which can be defined on any associated bundle E M. In [11] we have computed the ¯ on a primitive (1, 1)-form φ: difference of the operators and ¯ ( − )φ = (J d ∗ φ) + .
(17)
¯ coincide on co-closed primitive (1, 1)-forms. We now compute In particular, and ¯ on 1-forms. Using the calculation in [11] (or directly from (15)) we the difference −
The Hermitian Laplace Operator on Nearly Kähler Manifolds
255
¯ = id on T M. It remains to compute the operator P = ∇ ∗ ∇ − ∇¯ ∗ ∇¯ have q(R) − q( R) on T M. A direct calculation using (5) gives for every 1-form θ , P(θ ) = − 41 Aei Aei θ − Aei ∇¯ ei θ = θ − Aei ∇¯ ei θ = θ + 21 Aei Aei θ − Aei ∇ei θ = −θ − Aei ∇ei θ. In order to compute the last term, we introduce the metric adjoint α : 2 M → T M of the bundle homomorphism X ∈ T M → X + ∈ 2 M. It is easy to check that α(X + ) = 2X (cf. [10]). Keeping in mind that A is totally skew-symmetric, we compute for an arbitrary vector X ∈ T M,
Aei (∇ei θ ), X = A X ei , ∇ei θ = A X , ei ∧ ∇ei θ = A X , dθ = − +J X , dθ = − J X, α(dθ ) = J α(dθ ), X , whence Aei (∇ei θ ) = J α(dθ ). Summarizing our calculations we have proved the following Proposition 3.1. Let (M 6 , g, J ) be a nearly Kähler manifold with scalar curvature normalized to scalg = 30. Then for any 1-form θ it holds that ¯ = −J α(dθ ). ( − )θ The next result is a formula for the commutator of J and α ◦ d on 1-forms. Lemma 3.2. For all 1-forms θ , the following formula holds: α(dθ ) = 4J θ + J α(d J θ ). Proof. Differentiating the identity θ ∧ + = J θ ∧ − gives dθ ∧ + = d J θ ∧ − +2J θ ∧ ω2 . With respect to the SU3 -invariant decomposition 2 M = (1,1) M ⊕ (2,0)+(0,2) M, we can write dθ = (dθ )(1,1) + 21 α(dθ ) + and d J θ = (d J θ )(1,1) + 21 α(d J θ ) + . Since the wedge product of forms of type (1, 1) and (3, 0) vanishes we derive the equation + + 1 2 (α(dθ ) ) ∧
= 21 (α(d J θ ) + ) ∧ − + 2J θ ∧ ω2 .
Using (8) and (9) we obtain 2 1 2 α(dθ ) ∧ ω
= 21 J α(d J θ ) ∧ ω2 + 2J θ ∧ ω2 .
Taking the Hodge dual of this equation and using (12) gives J α(dθ ) = −α(d J θ ) − 4θ, which proves the lemma. Finally we note two interesting consequences of Proposition 3.1 and Lemma 3.2. Corollary 3.3. For any closed 1-form θ it holds that ¯ = 0, ( − )θ
¯ θ = 4J θ. ( − )J
¯ coincide on θ . Proof. For a closed 1-form θ Lemma 3.1 directly implies that and For the second equation we use Proposition 3.1 together with Lemma 3.2 to conclude ¯ θ = −J α(d J θ ) = 4J θ − α(dθ ) = 4J θ, ( − )J since θ is closed. This completes the proof of the corollary.
256
A. Moroianu, U. Semmelmann
¯ 4. Special -Eigenforms on Nearly Kähler Manifolds In this section we assume moreover that (M, g) is not isometric to the standard sphere ¯ (S 6 , can). In the first part of this section we will show how to construct -eigenforms on M starting from Killing vector fields. Let ξ be a non-trivial Killing vector field on (M, g), which in particular implies d ∗ ξ = 0 and ξ = 2Ric(ξ ) = 10ξ . As an immediate consequence of the Cartan formula and (13) we obtain d J ξ = L ξ ω − ξ dω = −3ξ ψ +
(18)
so by (4), the square norm of d J ξ (as a 2-form) is |d J ξ |2 = 18|ξ |2 .
(19)
In [9] we showed already that the vector field J ξ is co-closed if ξ is a Killing vector field and has unit length. However it turns out that this also holds more generally. Proposition 4.1. Let ξ be a Killing vector field on M. Then d ∗ J ξ = 0. Proof. Let dv denote the volume form of (M, g). We start with computing the L 2 -norm of d ∗ J ξ . ∗ 2 ∗ ∗ d J ξ L 2 =
d J ξ, d J ξ dv = [ J ξ, J ξ − d ∗ d J ξ, J ξ ]dv
M
[ ∇ ∗ ∇ J ξ, J ξ + 5|J ξ |2 − |d J ξ |2 ]dv
=
M
M
=
[|∇ J ξ |2 + 5|ξ |2 − |d J ξ |2 ]dv =
M
[|∇ J ξ |2 − 13|ξ |2 ]dv. M
Here we used the well-known Bochner formula for 1-forms, i.e. θ = ∇ ∗ ∇θ + Ric(θ ), with Ric(θ ) = 5θ in our case. Next we consider the decomposition of ∇ J ξ into its symmetric and skew-symmetric parts 2∇ J ξ = d J ξ + L J ξ g, which together with (19) leads to |∇ J ξ |2 = 41 (|d J ξ |2 + |L J ξ g|2 ) = 9|ξ |2 + 14 |L J ξ g|2 .
(20)
(Recall that the endomorphism square norm of a 2-form is twice its square norm as a form.) In order to compute the last norm, we express L J ξ g as follows: L J ξ g(X, Y ) = g(∇ X J ξ, Y ) + g(X, ∇Y J ξ ) = g(J ∇ X ξ, Y ) + g(X, J ∇Y ξ ) + + (X, ξ, Y ) + + (Y, ξ, X ) = −g(∇ X ξ, J Y ) − g(J X, ∇Y ξ ) = −dξ (1,1) (X, J Y ), whence L J ξ g2L 2 = 2dξ (1,1) 2L 2 .
(21)
On the other hand, as an application of Lemma 3.2 together with Eq. (18) we get α(dξ ) = 4J ξ + J α(d J ξ ) = −2J ξ , so dξ (2,0) = −J ξ + .
(22)
The Hermitian Laplace Operator on Nearly Kähler Manifolds
257
Moreover, ξ = 10ξ since ξ is a Killing vector field, which yields dξ (1,1) 2L 2 = dξ 2L 2 − dξ (2,0) 2L 2 = 10ξ 2L 2 − 2ξ 2L 2 = 8ξ 2L 2 . This last equation, together with (20) and (21) gives ∇ J ξ 2L 2 = 13ξ 2L 2 . Substituting this into the first equation proves that d ∗ J ξ has vanishing L 2 -norm and thus that J ξ is co-closed. Proposition 4.2. Let ξ be a Killing vector field on M. Then ξ = 10ξ,
and
J ξ = 18J ξ.
In particular, J ξ can never be a Killing vector field. Proof. The first equation holds for every Killing vector field on an Einstein manifold with Ric = 5id. From (18) we know d J ξ = −3ξ + . Hence the second assertion follows from: (10) (12) d ∗ d J ξ = − ∗ d ∗ d J ξ = −3 ∗ d(J ξ ∧ + ) = 9 ∗ (ξ ∧ ω2 ) = 18J ξ. Since the differential d commutes with the Laplace operator , every Killing vector field ξ defines two -eigenforms of degree 2: d J ξ = 18d J ξ
and
dξ = 10dξ.
As a direct consequence of Proposition 4.2, together with formulas (18), (22), and Proposition 3.1 we get: Corollary 4.3. Every Killing vector field on M satisfies ¯ = 12ξ, ξ
¯ ξ = 12J ξ. J
¯ -eigenform. By (22) we have Our next goal is to show that the (1, 1)-part of dξ is a dξ = φ − J ξ + ,
(23)
for some (1, 1)-form φ. Using Proposition 4.1, we can write in a local orthonormal basis {ei }:
dξ, ω = 21 dξ, ei ∧ J ei = ∇ei ξ, J ei = d ∗ J ξ = 0, thus showing that φ is primitive. The differential of φ can be computed from the Cartan formula: (23) (7) dφ = d(J ξ + + dξ ) = −d(ξ − ) (13) = −L ξ − + ξ d − = −2ξ ω2 = −4J ξ ∧ ω. From here we obtain ∗dφ = −4 ∗ (J ξ ∧ ω) = 4ξ ∧ ω,
(24)
258
A. Moroianu, U. Semmelmann
whence (23) d ∗ dφ = 4dξ ∧ ω − 12ξ ∧ + = 4φ ∧ ω − 4(J ξ + ) ∧ ω − 12ξ ∧ + (9) = 4φ ∧ ω − 16ξ ∧ + . Using (10) and (11), we thus get d ∗ dφ = − ∗ d ∗ dφ = 4φ + 16J ξ + . On the other hand, (11) (24) (12) d ∗ φ = − ∗ d ∗ φ = ∗d(φ ∧ ω) = X (−4J ξ ∧ ω2 + 3φ ∧ + ) = 8ξ and finally dd ∗ φ = 8dξ = 8φ − 8J ξ + . The calculations above thus prove the following proposition Proposition 4.4. Let (M 6 , g, J ) be a compact nearly Kähler manifold with scalar curvature scalg = 30, not isometric to the standard sphere. Let ξ be a Killing vector field on (1,1) M and let φ be the (1, 1)-part of dξ . Then φ is primitive, i.e. φ = (dξ )0 . Moreover ∗ + d φ = 8ξ and φ = 12φ + 8J ξ . Corollary 4.5. The primitive (1, 1)-form ϕ satisfies ¯ = 12φ. φ Proof. From (17) and the proposition above we get ¯ = φ − ( − )φ ¯ φ = 12φ + 8J ξ + − (J d ∗ φ) + = 12φ. In the second part of this section we will present another way of obtaining primi¯ tive -eigenforms of type (1, 1), starting from eigenfunctions of the Laplace operator. Let f be such an eigenfunction, i.e. f = λ f . We consider the primitive (1, 1)-form (1,1) η := (d J d f )0 . Lemma 4.6. The form η is explicitly given by η = d J d f + 2d f + +
λ 3
f ω.
Proof. According to the decomposition of 2 M into irreducible SU3 -summands, we can write d J d f = η + γ + + hω for some vector field γ and function h. From Lemma 3.2 we get 2γ = α(d J d f ) = −4d f . In order to compute h, we write (12) 6h dv = hω ∧ ω2 = d J d f ∧ ω2 = d(J d f ∧ ω2 ) = 2d ∗ d f = 2λ f dv.
The Hermitian Laplace Operator on Nearly Kähler Manifolds
259
We will now compute the Laplacian of the three summands of η separately. First, we ¯ f = λd f . Since ¯ commutes with J , we have d f = λd f and Corollary 3.3 yields d ¯ also have J d f = λJ d f and from the second equation in Corollary 3.3 we obtain ¯ d f + ( − )J ¯ d f = (λ + 4)J d f. J d f = J Hence, d J d f is a -eigenform for the eigenvalue λ + 4. Lemma 4.7. The co-differential of the (1, 1)-form η is given by d ∗ η = 2λ 3 − 4 J d f. Proof. Notice that d ∗ ( f ω) = −d f ω and that d ∗ J d f = − ∗ d ∗ J d f = − 21 ∗ d(d f ∧ ω2 ) = 0, since dω2 = 0. Using this we obtain d ∗ η = J d f + 2d ∗ (d f + ) − λ3 d f ω = (λ + 4)J d f − 2 ∗ d(d f ∧ − ) − λ3 J d f (12) = (λ + 4 − λ3 )J d f − 4 ∗ (d f ∧ ω2 ) = ( 2λ 3 − 4)J d f. In order to compute of the second summand of η we need three additional formulas Lemma 4.8. ¯ ) + . ¯ + ) = (X (X ¯ Since + is ∇-parallel ¯ ¯ = ∇¯ ∗ ∇¯ + q( R). Proof. Recall that we immediately obtain ¯ + ) = −∇¯ ei ∇¯ ei (X + ) = −(∇¯ ei ∇¯ ei X ) + . ∇¯ ∗ ∇(X The map A → A∗ + is a SU3 -equivariant map from 2 to 3 . But since 3 does not (1,1) contain the representation 0 as an irreducible summand, it follows that A∗ + = 0 for any skew-symmetric endomorphism A corresponding to some primitive (1, 1)-form. Hence we conclude ¯ ¯ ) + , ¯ i )∗ (X + ) = (ωi∗ R(ω ¯ i )∗ X ) + = (q( R)X q( R)(X + ) = ωi∗ R(ω where, since the holonomy of ∇¯ is included in SU3 , the sum goes over some ortho(1,1) ¯ + ) = normal basis {ωi } of 0 M. Combining these two formulas we obtain (X ¯ ) + . (X Lemma 4.9. ¯ ( − )(d f + ) = 6(d f + ) −
4λ 3
f ω − 2η.
Proof. From Proposition 3.4 in [11] we have ¯ ¯ ¯ ( − )(d f + ) = (∇ ∗ ∇ − ∇¯ ∗ ∇)(d f + ) f + ) + (q(R) − q( R))(d ¯ = (∇ ∗ ∇ − ∇¯ ∗ ∇)(d f + ) + 4d f + . The first part of the right hand side reads ¯ (∇ ∗ ∇ − ∇¯ ∗ ∇)(d f + ) = − 41 Aei ∗ Aei ∗ d f + − Aei ∗ ∇¯ ei (d f + ).
(25)
260
A. Moroianu, U. Semmelmann
From (5) we get Aei ∗ Aei ∗ d f + = Aei ∗ (Aei ek ∧ + (d f, ek , ·)) = Aei Aei ek ∧ + (d f, ek , ·) + Aei ek ∧ Aei + (d f, ek , ·) = −4ek ∧ ek d+f + Aei ek ∧ Aei e j + (d f, ek , e j ) = −8d+f , where we used the vanishing of the expression E = Aei ek ∧ Aei e j + (d f, ek , e j ): E = A J ei ek ∧ A J ei e j + (d f, ek , e j ) = Aei J ek ∧ Aei J e j + (d f, ek , e j ) = Aei ek ∧ Aei e j + (d f, J ek , J e j ) = −E. It remains to compute the second term in (25). We notice that by Schur’s Lemma, every SU3 -equivariant map from the space of symmetric tensors Sym2 M to T M vanishes, so in particular (since ∇d f is symmetric), one has Aei ∇ei d f = 0. We then compute Aei ∗ ∇¯ ei d+f = Aei ∗ ((∇¯ ei d f ) + ) = (Aei ∇¯ ei d f ) + + (∇¯ ei d f )Aei ∗ + (6) = (Aei ∇ei d f ) + − 21 (Aei Aei d f ) + − 2(∇¯ ei d f )(ei ∧ ω) = 2d+f + 2d ∗ d f ω + Aei d f, ei ω + 2ei ∧ J ∇¯ ei d f = 2d+f +2λ f ω + 2ei ∧ ∇¯ ei J d f = 2d+f +2λ f ω + 2d J d f − ei ∧ Aei J d f = 2d+f + 2λ f ω + 2d J d f + 2 A J d f = 4d+f + 2λ f ω + 2d J d f. Plugging back what we obtained into (25) yields ¯ f + ) = −(2d+f + 2λ f ω + 2d J d f ), (∇ ∗ ∇ − ∇¯ ∗ ∇)(d which together with Lemma 4.6 and the first equation prove the desired formula.
Lemma 4.10. f ω = (λ + 12) f ω − 2(d f + ). Proof. Since d ∗ ( f ω) = −d f ω = −J d f we have dd ∗ ( f ω) = −d J d f . For the second summand of ( f ω) we first compute d( f ω) = d f ∧ ω + 3 f + . Since d ∗ + = 1 ∗ ∗ + + ∗ + + 3 d dω = 4ω, we get d f = −d f + f d = −d f + 4 f ω. Moreover d ∗ (d f ∧ ω) = − ∗ d(J d f ∧ ω) = − ∗ (d J d f ∧ ω − 3J d f ∧ + ) = − ∗ ([η − 2d f + −
λ 3
f ω] ∧ ω) + 3 ∗ (J d f ∧ + )
= η + 2 ∗ ((d f + ) ∧ ω) + = η + 2d f + + Recalling that η = d J d f + 2d f + +
2λ 3
λ 3
2λ 3
f ω − 3d f +
f ω − 3d f + .
f ω, we obtain
f ω = −d J d f − 3d f + + 12 f ω + η − d f + +
2λ 3
f ω = (λ + 12) f ω − 2d f + .
The Hermitian Laplace Operator on Nearly Kähler Manifolds
261
Applying these three lemmas we conclude ¯ f + ) + ( − )(d ¯ f + ) = (λ + 6)(d f + ) − (d f + ) = (d
4λ 3
f ω − 2η,
and thus η = (λ + 4)d J d f + (2λ + 12)(d f + ) − + = λη + 4 − 2λ 3 (d f ).
8λ 3
f ω − 4η + λ3 (λ + 12) f ω −
2λ + 3 (d f )
¯ on Finally we have once again to apply the formula for the difference of and primitive (1, 1)-forms. We obtain ¯ = η − J d ∗ η + = η + 2λ − 4 (d f + ) = λη. η 3 Summarizing our calculations we obtain the following result. Proposition 4.11. Let f be an -eigenfunction with f = λ f Then the primitive (1, 1)form η := (d J d f )0(1,1) satisfies ¯ = λη and d ∗ η = 2λ − 4 J d f. η 3 ¯ ¯ = Let 0 (12) ⊂ C ∞ (M) be the -eigenspace for the eigenvalue 12 (notice that (1,1) ¯ on functions) and let 0 (12) denote the space of primitive (1, 1)-eigenforms of corresponding to the eigenvalue 12. Summarizing Corollary 4.5 and Proposition 4.11, we have constructed a linear mapping (1,1)
: i(M) → 0
(1,1)
(12),
(ξ ) := dξ0 (1,1)
from the space of Killing vector fields into 0 : 0 (12) → 0(1,1) (12),
(12) and a linear mapping
( f ) := (d J d f )0(1,1) .
Let moreover N K ⊂ 0(1,1) (12) denote the space of nearly Kähler deformations, which (1,1) by [10] is just the space of co-closed forms in 0 (12). Proposition 4.12. The linear mappings and defined above are injective and the sum Im() + Im() + N K ⊂ 0(1,1) (12) is a direct sum. In particular, dim(N K) ≤ dim(0(1,1) (12)) − dim(i(M)) − dim(0 (12)).
(26)
Proof. It is enough to show that if ξ ∈ i(M), f ∈ 0 (12) and α ∈ N K satisfy (1,1)
dξ0
(1,1)
+ (d J d f )0
+ α = 0,
(27)
then ξ = 0 and f = 0. We apply d ∗ to (27). Using Propositions 4.4 and 4.11 to express the co-differentials of the first two terms we get 8ξ + 8J d f = 0.
(28)
Since J ξ is co-closed (Proposition 4.1), formula (28) implies 0 = d ∗ J ξ = d ∗ d f = 12 f , i.e. f = 0. Plugging back into (28) yields ξ = 0 too.
262
A. Moroianu, U. Semmelmann
5. The Homogeneous Laplace Operator on Reductive Homogeneous Spaces 5.1. The Peter-Weyl formalism. Let M = G/K be a homogeneous space with compact Lie groups K ⊂ G and let π : K → Aut(E) be a representation of K . We denote by E M := G ×π E the associated vector bundle over M. The Peter-Weyl theorem and the Frobenius reciprocity yield the following isomorphism of G-representations: Vγ ⊗ Hom K (Vγ , E), (29) L 2 (E M) ∼ = γ ∈Gˆ
where Gˆ is the set of (non-isomorphic) irreducible G-representations. If not otherwise stated we will consider only complex representations. Recall that the space of smooth sections C ∞ (E M) can be identified with the space C ∞ (G; E) K of K -invariant E-valued functions, i.e. functions f : G → E with f (gk) = π(k)−1 f (g). This space is made into a G-representation by the left-regular representation , defined by ((g) f )(a) = f (g −1 a). Let v ∈ Vγ and A ∈ Hom K (Vγ , E), then the invariant E-valued function corresponding to v ⊗ A is defined by g → A(g −1 v). In particular, each summand in the Hilbert space direct sum (29) is a subset of C ∞ (E M) ⊂ L 2 (E M). Let g be the Lie algebra of G. We denote by B the Killing form of g, B(X, Y ) := tr(ad X ◦ adY ). The Killing form is non-degenerate and negative definite if G is compact and semi-simple, which will be the case in all examples below. If π : G → Aut(E) is a G-representation, the Casimir operator of (G, π ) acts on E by the formula CasπG = (π∗ X i )2 , (30) where {X i } is a (−B)-orthonormal basis of g and π∗ : g → End(E) denotes the differential of the representation π . Remark 5.1. Notice that the Casimir operator is divided by k if one uses the scalar product −k B instead of −B. If G is simple, the adjoint representation ad on the complexification gC is irreducible, so, by Schur’s Lemma, its Casimir operator acts as a scalar. Taking the trace in (30) for G = −1. π = ad yields the useful formula Casad Let Vγ be an irreducible G-representation of highest weight γ . By Freudenthal’s formula the Casimir operator acts on Vγ by scalar multiplication with ρ2 − ρ + γ 2 , where ρ denotes the half-sum of the positive roots and · is the norm induced by −B on the dual of the Lie algebra of the maximal torus of G. Notice that these scalars are always non-positive. Indeed ρ2 − ρ + γ 2 = − γ , γ + 2ρ B and γ , ρ ≥ 0, since γ is a dominant weight, i.e. it is in the the closure of the fixed Weyl chamber, whereas ρ is the half-sum of positive weights and thus by definition has a non-negative scalar product with γ . 5.2. The homogeneous Laplace operator. We denote by ∇¯ the canonical homogeneous connection on M = G/K . It coincides with the Levi-Civita connection only in the case that G/K is a symmetric space. A crucial observation is that the canonical homogeneous connection coincides with the canonical Hermitian connection on naturally reduc¯ ∈ tive 3-symmetric spaces (see below). We define the curvature endomorphism q( R) ¯ ¯ ¯ π = ∇¯ ∗ ∇+q( End(E M) as in (14) and introduce as in (16) the second order operator R) acting on sections of the associated bundle E M := G ×π E.
The Hermitian Laplace Operator on Nearly Kähler Manifolds
263
Lemma 5.2. Let G be a compact semi-simple Lie group, K ⊂ G a compact subgroup, and let M = G/K be the naturally reductive homogeneous space equipped with the Riemannian metric induced by −B. For every K -representation π on E, let E M := ¯ acts G ×π E be the associated vector bundle over M. Then the endomorphism q( R) ¯ = −CasπK . Moreover the differential operator ¯ acts on the fibre-wise on E M as q( R) space of sections of E M, considered as G-representation via the left-regular represen¯ = −CasG . tation, as Proof. Consider the Ad(K )-invariant decomposition g = k ⊕ p. For any vector X ∈ g we write X = X k + X p, with X k ∈ k and X p ∈ p. The canonical homogeneous connection is the left-invariant connection in the principal K -fibre bundle G → G/K corresponding to the projection X → X k. It follows that one can do for the canonical homogeneous connection on G/K the same identifications as for the Levi Civita connection on Riemannian symmetric spaces. In particular, the covariant derivative of a section φ ∈ (E M) with respect to ˆ of the corresponding function φˆ ∈ some X ∈ p translates into the derivative X (φ) ∞ K ˆ = C (G; E) , which is minus the differential of the left-regular representation X (φ) ∗ ¯ ¯ ˆ −∗ (X )φ. Hence, if {eµ } is an orthonormal basis in p, the rough Laplacian ∇ ∇ trans¯ it remains ¯ = ∇¯ ∗ ∇¯ +q( R) lates into the sum −∗ (eµ )∗ (eµ ) = (−CasG +CasK ). Since K K ¯ = −Cas = −Casπ in order to complete the proof of the lemma. to show that q( R) We claim that the differential i ∗ : k → so(p) ∼ = 2 p of the isotropy representation i : K → SO(p) is given by i ∗ (A) = − 21 eµ ∧ [A, eµ ] for any A ∈ k. Indeed ( 21 eµ ∧ [A, eµ ])∗ X = − 21 B(eµ , X )[A, eµ ] + 21 B([A, eµ ], X )eµ = −[A, X ]. Next we recall that for X, Y ∈ p the curvature R¯ X,Y of the canonical connection acts by −π∗ ([X, Y ]k) on every associated vector bundle E M, defined by the representation π . Hence the curvature operator R¯ can be written for any X, Y ∈ p as ¯ ∧ Y ) = 1 eµ ∧ R¯ X,Y eµ = − 1 eµ ∧ [[X, Y ]k, eµ ] = i ∗ ([X, Y ]k). R(X 2 2 Let PSO(p) = G ×i SO(p) be the bundle of orthonormal frames of M = G/K . Then any SO(p)-representation π˜ defines a K -representation by π = π˜ ◦ i. Moreover any vector bundle E M associated to PSO(p) via π˜ can be written as a vector bundle associated via π to the K -principle bundle G → G/K , i.e. E M = PSO(p) ×π˜ E = G ×π E. ¯ we have Let { f α } be an orthonormal basis of k. Then by the definition of q( R) ¯ µ ∧ eν )) = 1 π˜ ∗ (eµ ∧ eν ) π∗ ([eµ , eν ]k) ¯ = 1 π˜ ∗ (eµ ∧ eν ) π˜ ∗ ( R(e q( R) 2 2 = −21 B([eµ , eν ], f α )π˜ ∗ (eµ ∧eν ) π∗ ( f α ) = − 21 B(eν , [ f α , eµ ])π˜ ∗ (eµ ∧ eν ) π∗ ( f α ) = 21 π˜ ∗ (eµ ∧ [ f α , eµ ]) π∗ ( f α ) = −π∗ ( f α ) π∗ ( f α ) = −CasπK . ¯ ∈ End(E M) acts fibre-wise as −CasπK . Let Z ∈ k and We have shown that q( R) f ∈ C ∞ (G; E) K , then the K -invariance of f implies π∗ (Z ) f = −Z ( f ) = ∗ (Z ) f and also CasπK = CasK , which concludes the proof of the lemma.
264
A. Moroianu, U. Semmelmann
¯ on sections of E M is the set It follows from this lemma that the spectrum of of numbers λγ = ρ + γ 2 − ρ2 , where γ is the highest weight of an irreducible G-representation Vγ such that Hom K (Vγ , E) = 0, i.e. such that the decomposition of Vγ , considered as a K -representation, contains components of the K -representation E. 5.3. Nearly Kähler deformations and Laplace eigenvalues. Let (M, g, J ) be a compact simply connected 6-dimensional nearly Kähler manifold not isometric to the round sphere, with scalar curvature normalized to scalg = 30. Recall the following result from [10]: Theorem 5.3. The Laplace operator coincides with the Hermitian Laplace operator ¯ on co-closed primitive (1, 1)-forms. The space N K of infinitesimal deformations of the nearly Kähler structure of M is isomorphic to the eigenspace for the eigenvalue 12 ¯ to the space of co-closed primitive (1, 1)-forms on M. of the restriction of (or ) Assume from now on that M is a 6-dimensional naturally reductive 3-symmetric space G/K in the list of Gray and Wolf, i.e. SU2 × SU2 × SU2 /SU2 , SO5 /U2 or SU3 /T 2 . As was noticed before, the canonical homogeneous and the canonical Hermitian connection coincide, since for the later it can be shown that its torsion and its curvature are parallel, a property, which by the Ambrose-Singer-Theorem characterizes the canonical homogeneous connection (cf. [5]). In order to determine the space N K on M we thus ¯ need to apply the previous calculations to compute the -eigenspace for the eigenvalue 12 on primitive (1, 1)-forms and decide which of these eigenforms are co-closed. According to Lemma 5.2 and the decomposition (29) we have to carry out three 1,1 steps: first to determine the K -representation 1,1 0 p defining the bundle 0 T M, then to compute the Casimir eigenvalues with the Freudenthal formula, which gives all pos¯ sible -eigenvalues and finally to check whether the G-representation Vγ realizing the eigenvalue 12 satisfies Hom K (Vγ , 1,1 0 p) = {0} and thus really appears as eigenspace. Before going on, we make the following useful observation Lemma 5.4. Let (G/K , g) be a 6-dimensional homogeneous strict nearly Kähler manifold of scalar curvature scalg = 30. Then the homogeneous metric g is induced from 1 B, where B is the Killing form of G. − 12 Proof. Let G/K be a 6-dimensional homogeneous strict nearly Kähler manifold. Then the metric is induced from a multiple of the Killing form, i.e. G/K is a normal homogeneous space with Ad(K )-invariant decomposition g = k ⊕ p. The scalar curvature of the metric h induced by −B may be computed as (cf. [3]) scalh =
3 2
− 3CasλK ,
where λ : K → so(p) is the isotropy representation. From Lemma 5.2 we know that ¯ which on the tangent bundle was computed in Lemma 15 as q( R) ¯ = CasλK = −q( R), 2scalh 3 2 5 15 id. Hence we obtain the equation scalh = 2 + 5 scalh and it follows scalh = 2 , i.e. 1 the metric g corresponding to − 12 B has scalar curvature scalg = 30. ¯ 5.4. The -spectrum on S 3 × S 3 . Let K = SU2 with Lie algebra k = su2 and G = K × K × K with Lie algebra g = k ⊕ k ⊕ k. We consider the 6-dimensional manifold
The Hermitian Laplace Operator on Nearly Kähler Manifolds
265
M = G/K , where K is diagonally embedded. The tangent space at o = eK can be identified with p = {(X, Y, Z ) ∈ k ⊕ k ⊕ k | X + Y + Z = 0}. 1 Let B be the Killing form of k and define B0 = − 12 B. Then it follows from Lemma 5.4 that the invariant scalar product
B0 ((X, Y, Z ), (X, Y, Z )) = B0 (X, X ) + B0 (Y, Y ) + B0 (Z , Z ) defines a normal metric, which is the homogeneous nearly Kähler metric g of scalar curvature scalg = 30. The canonical almost complex structure on the 3-symmetric space M, corresponding to the 3rd order G-automorphism σ , with σ (k1 , k2 , k3 ) = (k2 , k3 , k1 ), is defined as J (X, Y, Z ) =
√2 (Z , 3
X, Y ) +
√1 (X, Y, 3
Z ).
The (1, 0)-subspace p1,0 of pC defined by J is isomorphic to the complexified adjoint 2 representation of SU2 on suC 2 . Let E = C denote the standard representation of SU2 ∼ ∼ ¯ (notice that E = E because every SU2 = Sp1 representation is quaternionic). (1,1)
Lemma 5.5. The SU2 -representation defining the bundle 0 ducible summands Sym4 E and Sym2 E.
T M splits into the irre-
Proof. The defining SU2 -representation of (1,1) T M is p1,0 ⊗ p0,1 ∼ = Sym2 E ⊗ 4 E ⊕ Sym 2 E ⊕ Sym 0 E from the Clebsch-Gordan formula. Since we Sym2 E ∼ Sym = are interested in primitive (1, 1)-forms, we still have to delete the trivial summand Sym0 E ∼ = C. Since G = SU2 × SU2 × SU2 , every irreducible G-representation is isomorphic to one of the representations Va,b,c = Syma E ⊗Symb E ⊗Symc E. The Casimir operator of the SU2 -representation Symk E (with respect to B) is − 18 k(k + 2) and the Casimir operator ¯ of G is the sum of the three SU2 -Casimir operators. Hence all possible -eigenvalues with respect to the metric B0 are of the form 3 2 (a(a
+ 2) + b(b + 2) + c(c + 2))
(31)
for non-negative integers a, b, c. It is easy to check that the eigenvalue 12 is obtained only for (a, b, c) equal to (2, 0, 0), (0, 2, 0) or (0, 0, 2). The restrictions to SU2 (diagonally embedded in G) of the three corresponding G-representations are all equal to the SU2 -representation Sym2 E, thus dim HomSU2 (V2,0,0 , 0(1,1) p) = 1, and similarly ¯ on primitive (1, 1)-forms for for the two other summands. Hence the eigenspace of the eigenvalue 12 is isomorphic to V2,0,0 ⊕ V0,2,0 ⊕ V0,0,2 and its dimension, i.e. the multiplicity of the eigenvalue 12, is equal to 9. Since the isometry group of the nearly Kähler manifold M = SU2 ×SU2 ×SU2 /SU2 has dimension 9, the inequality (26) yields (1,1)
dim(N K) ≤ dim(0
(12)) − dim(i(M)) − dim(0 (12)) = − dim(0 (12)) ≤ 0.
266
A. Moroianu, U. Semmelmann
We thus have obtained the following Theorem 5.6. The homogeneous nearly Kähler structure on S 3 × S 3 does not admit any infinitesimal nearly Kähler deformations. Finally we remark that there are also no infinitesimal Einstein deformations either. In [11] we showed that the space of infinitesimal Einstein deformations of a nearly Kähler metric g, with normalized scalar curvature scalg = 30, is isomorphic to the direct sum ¯ of -eigenspaces of primitive co-closed (1, 1)-forms for the eigenvalues 2, 6 and 12. It ¯ is clear from (31) that neither 2 nor 6 can be realized as -eigenvalues. Corollary 5.7. The homogeneous nearly Kähler metric on S 3 × S 3 does not admit any infinitesimal Einstein deformations. ¯ 5.5. The -spectrum on CP 3 . In this section we consider the complex projective space CP 3 = SO5 /U2 , where U2 is embedded by U2 ⊂ SO4 ⊂ SO5 . Let G = SO5 with Lie algebra g and K = U2 with Lie algebra k. We denote the Killing form of G with B. Then we have the B-orthogonal decomposition g = k ⊕ p, where p can be identified with the tangent space in o = eK . The space p splits as p = m ⊕ n, where m resp. n can be identified with the horizontal resp. vertical tangent space at o of the twistor space 1 fibration SO5 /U2 → SO5 /SO4 = S 4 . We know from Lemma 5.4 that B0 = − 12 B defines the homogeneous nearly Kähler metric g of scalar curvature scalg = 30. Let {ε1 , ε2 } denote the canonical basis of R2 . Then the positive roots of SO5 are α1 = ε1 , α2 = ε2 , α3 = ε1 + ε2 , α4 = ε1 − ε2 , with ρ = 23 ε1 + 21 ε2 . Let gα ⊂ gC be the root space corresponding to the root α. Then mC = gα1 ⊕ g−α1 ⊕ gα2 ⊕ g−α2 ,
nC = gα3 ⊕ g−α3 .
The invariant almost complex structure J may be defined by specifying the (1, 0)subspace p1,0 of pC : p1,0 = {X − i J X | X ∈ p} = gα1 ⊕ gα2 ⊕ g−α3 . It follows that J is not integrable, since the restricted root system {α1 , α2 , −α3 } is not closed under addition (cf. [4]). We note that replacing −α3 by α3 yields an integrable almost complex structure. This corresponds to the well-known fact that on the twistor space the non integrable almost complex structure J is transformed into the integrable one by replacing J with −J on the vertical tangent space. Let Ck denote the U1 -representation on C defined by (z, v) → z k v, for v ∈ C and z ∈ U1 ∼ = C∗ . Then, since U2 = (SU2 × U1 )/Z2 , any irreducible U2 -representation is of the form E a,b = Syma E ⊗ Cb , with a ∈ N, b ∈ Z and a ≡ b mod 2. As usual let E = C2 denote the standard representation of SU2 . With this notation we obtain the following decomposition of p1,0 considered as a U2 -representation: p1,0 ∼ = E 0,−2 ⊕ E 1,1
E 0,−2 ∼ = g−α3 and E 1,1 ∼ = gα1 ⊕ gα2 . (32) Since p0,1 is obtained from p1,0 by conjugation we have p0,1 ∼ = E 0,2 ⊕ E 1,−1 . The definwith
ing U2 -representation of (1,1) T M is p1,0 ⊗ p0,1 , which obviously decomposes into 5 irreducible summands, among which, two are isomorphic to the trivial representation E 0,0 . Considering only primitive (1, 1)-forms we still have to delete one of the trivial summands and obtain
The Hermitian Laplace Operator on Nearly Kähler Manifolds
267 (1,1)
Lemma 5.8. The U2 -representation defining the bundle 0 decomposition into irreducible summands: (1,1)
0
T M has the following
p = E 0,0 ⊕ E 1,3 ⊕ E 1,−3 ⊕ E 2,0 .
Let Va,b be an irreducible SO5 -representation of highest weight γ = (a, b) with a, b ∈ N and a ≥ b ≥ 0, e.g. V1,0 = 1 and V1,1 = 2 . The scalar product induced by the Killing form B on the dual t∗ ∼ = R2 of the maximal torus of SO5 is − 16 times the Euclidean scalar product. By the Freudenthal formula we thus get CasVa,b = γ , γ + 2ρ B = − 16 (a(a + 3) + b(b + 1)).
(33)
G Notice that we have V1,1 = soC 5 and Cas V1,1 = −1, which is consistent with Casad = −1. ¯ It follows (cf. Remark 5.1) that all possible -eigenvalues with respect to the metric induced by B0 are of the form 2(a(a + 3) + b(b + 1)). The eigenvalue 12 is realized if and only if (a, b) = (1, 1). We still have to decide whether the SO5 -representation V1,1 actually appears in the decomposition (29) of L 2 (1,1 0 T M). However this follows from
Lemma 5.9. The SO5 -representation V1,1 restricted to U2 ⊂ SO5 has the following decomposition as U2 -representation: V1,1 ∼ = (E 0,0 ⊕ E 2,0 ) ⊕ (E 0,−2 ⊕ E 1,1 ⊕ E 0,2 ⊕ E 1,−1 ) and in particular C dim HomU2 (V1,1 , 1,1 0 p )=2
and
dim HomU2 (V1,1 , C) = 1.
Proof. We know already that V1,1 = soC 5 is the complexified adjoint representation and C ⊕(p1,0 ⊕p0,1 ). The decomposition of the last two summands is contained = u that soC 2 5 in (32). Hence it remains to make explicit the adjoint representation of U2 on uC 2 . It is clear that its restriction to U1 acts trivially, whereas its restriction to SU2 decomposes C ∼ into C ⊕ suC 2 , i.e. u2 = E 0,0 ⊕ E 2,0 . ¯ on primitive (1, 1)-forms for the eigenvalue 12 is thus isomorThe eigenspace of phic to the sum of two copies of V1,1 , i.e. the eigenvalue 12 has multiplicity 2 · 10 = 20. It is now easy to calculate the smallest eigenvalue and the corresponding eigenspace ¯ which coincides of the Laplace operator on non-constant functions. We do this for , (1,1) with on functions. Then we have to replace 0 p in the calculations above with the trivial representation C and to look for SO5 -representations Va,b containing the zero weight. It follows from Lemma 5.9 and (33) that the -eigenspace on functions 0 (12) is isomorphic to V1,1 and is thus 10-dimensional. Since the dimension of the isometry group of the nearly Kähler manifold SO(5)/U2 is 10, the inequality (26) shows that (1,1)
dim(N K) ≤ dim(0
(12)) − dim(i(M)) − dim(0 (12)) = 20 − 10 − 10 = 0,
so there are no infinitesimal nearly Kähler deformations in this case either. Finally, we remark like before that there are also no other infinitesimal Einstein defor¯ on mations, since by (33), the eigenvalues 2 and 6 do not occur in the spectrum of (1,1) 0 M. Summarizing, we have obtained the following: Theorem 5.10. The homogeneous nearly Kähler structure on CP3 = SO5 /U2 does not admit any infinitesimal nearly Kähler or Einstein deformations.
268
A. Moroianu, U. Semmelmann
¯ 5.6. The -spectrum on the flag manifold F(1, 2). In this section we consider the flag manifold M = SU3 /T 2 , where T 2 ⊂ SU3 is the maximal torus. Let g = su3 and let k = t, the Lie algebra of T 2 . We have the decomposition g=k⊕p
and
p = m ⊕ n.
Denoting by E i j , Si j the “real and imaginary” part of the projection of the vector X i j ∈ gl3 (equal to 1 on i th row and j th column and 0 elsewhere) onto su3 : E i j = X i j − X ji
Si j = i(X i j + X ji ),
the subspaces m and n are explicitly given by m = span{E 12 , S12 , E 13 , S13 } = span{e1 , e2 , e3 , e4 }, n = span{E 23 , S23 } = span{e5 , e6 }. The dual of the Lie algebra t of the maximal torus T 2 can be identified with t∗ ∼ = {(λ1 , λ2 , λ3 ) ∈ R3 | λ1 + λ2 + λ3 = 0}. If {εi } denotes the canonical basis in R3 then the set of positive roots is given as φ + = {αi j = εi − ε j | 1 ≤ i < j ≤ 3} and the half-sum of the positive roots is ρ = ε1 − ε3 . 1 B defines the Let B denote the Killing form of SU3 . By Lemma 5.4, B0 = − 12 homogeneous nearly Kähler metric g of scalar curvature scalg = 30. The almost complex structure J is explicitly defined on p by J (e1 ) = e2 ,
J (e3 ) = −e4 ,
J (e5 ) = e6 .
Alternatively we may define the (1, 0)-subspace of pC : p1,0 = gα12 ⊕ gα31 ⊕ gα23 = span{X 12 , X 31 , X 23 }, where gα is the root space for α. It follows that J is not integrable, since the restricted root system {α12 , α31 , α23 } is not closed under addition (cf. [4]). ¯ Let E = C3 be the standard representation of SU3 with conjugate representation E. Any irreducible representations of SU3 is isomorphic to one of the representations ¯ 0, Vk,l := (Symk E ⊗ Syml E) where the right-hand side denotes the kernel of the contraction map ¯ Symk E ⊗ Syml E¯ → Symk−1 E ⊗ Syml−1 E, ¯ The weights of Symk E are i.e. Vk,l is the Cartan summand in Symk E ⊗ Syml E. aε1 + bε2 + cε3 ,
with a, b, c ≥ 0, a + b + c = k.
If v1 , v2 , v3 are the weight vectors of E, then these weights correspond to the weight vectors v1a · v2b · v3c in Symk E. Since the weights of Syml E¯ are just minus the weights of Syml E, we see that the weights of Vk,l are (a − a )ε1 + (b − b )ε2 + (c − c )ε3 , a, b, c, a , b , c ≥ 0, a + b + c = k, a + b + c = l.
(34)
The Hermitian Laplace Operator on Nearly Kähler Manifolds
269
From the given definition of the almost complex structure J it is clear that the T 2 -representation on p1,0 splits in three one-dimensional T 2 -representations with the weights α12 , α31 , α23 . Since the weights of a tensor product representation are the sums of weights of each factor and since ε1 + ε2 + ε3 = 0 on the Lie algebra of the maximal torus of SU3 , we immediately obtain Corollary 5.11. The weights of the T 2 -representation on 1,1 p ∼ = p1,0 ⊗ p0,1 are ±3ε1 , ±3ε2 , ±3ε3 , and 0. It remains to compute the Casimir operator of the irreducible SU3 -representations Vk,l . The highest weight of Vk,l is γ = kε1 − lε3 and ρ = ε1 − ε3 , thus CasVk,l = γ , γ + 2ρ B = − 16 (k(k + 2) + l(l + 2)).
(35)
Here we use again the Freudenthal formula and the fact that the Killing form B induces − 16 times the Euclidean scalar product on t∗ ⊂ R3 (easy calculation). Notice that we G have V1,1 = suC 3 and CasV1,1 = −1, which is consistent with Casad = −1 as in the previous cases. ¯ It follows that all possible -eigenvalues (with respect to the metric B0 ) are of the form 2(k(k + 2) + l(l + 2)). Obviously the eigenvalue 12 can only be obtained for k = l = 1. Moreover, the restriction of the SU3 -representation V1,1 contains the zero weight space. In fact, from (34), the zero weight appears in Vk,l if and only if there exist a, b, c, a , b , c ≥ 0, a + b + c = k, a + b + c = l such that (a − a )ε1 + (b − b )ε2 + (1,1) (c − c )ε3 = 0, which is equivalent to k = l. We see that dim Hom T 2 (V1,1 , 0 p) = 2 · 2 = 4. ¯ on primitive (1, 1)-forms for the eigenvalue 12 is isomorHence the eigenspace of phic to the sum of four copies of V1,1 , i.e. the eigenvalue 12 has multiplicity 4 · 8 = 32. Computing the smallest eigenvalue and the corresponding eigenspace of the Laplace operator on non-constant functions we find V0,0 for the eigenvalue 0 and V1,1 for the eigenvalue 12. All other possible representations give a larger eigenvalue. Hence, the -eigenspace on functions 0 (12) is isomorphic to two copies of V1,1 , i.e. the eigenvalue 12 has multiplicity 8 · 2 = 16. Since the dimension of the isometry group of the nearly Kähler manifold SU3 /T 2 is 8, we obtain from (26) (1,1)
dim(N K) ≤ dim(0
(12)) − dim(i(M)) − dim(0 (12)) = 8.
(36)
In the next section we will show by an explicit construction that actually the equality holds, so the flag manifold has an 8-dimensional space of infinitesimal nearly Kähler deformations. Before describing this construction we note that there are no infinitesimal Einstein deformations other than the nearly Kähler deformations. It follows from (35) that the ¯ on (1,1) M. The eigenvalue 6 could eigenvalue 2 does not occur in the spectrum of 0 be realized on the SU3 -representations V = V1,0 or V = V0,1 . However it is easy to (1,1) check that Hom T 2 (V, 0 p) = {0}. Corollary 5.12. Every infinitesimal Einstein deformation of the homogeneous nearly Kähler metric on F(1, 2) = SU3 /T 2 is an infinitesimal nearly Kähler deformation.
270
A. Moroianu, U. Semmelmann
6. The Infinitesimal Nearly Kähler Deformations on SU3 / T 2 In this section we describe by explicit computation the space of infinitesimal nearly Kähler deformations of the flag manifold F(1, 2) = SU3 /T 2 . The Lie algebra u3 is spanned by {h 1 , h 2 , h 3 , e1 , . . . , e6 }, where h 1 = i E 11 , h 2 = i E 22 , h 3 = i E 33 , e1 = E 12 − E 21 , e3 = E 13 − E 31 , e5 = E 23 − E 32 , e2 = i(E 12 + E 21 ), e4 = i(E 13 + E 31 ), e6 = i(E 23 + E 32 ). We consider the bi-invariant metric g on SU3 induced by −B/12, where B denotes the Killing form of su3 . It is easy to check that |ei |2 =1 and |h i − h j |2 = 1 with respect √ to g. We extend this metric to U3 in the obvious way which makes the frame {ei , 2h j } orthonormal. This defines a metric, also denoted by g, on the manifold M = F(1, 2). From now on we identify vectors and 1-forms using this metric and use the notation ei j = ei ∧ e j , etc. An easy explicit commutator calculation yields the exterior derivative of the leftinvariant 1-forms ei on U3 : de1 de2 de3 de4 de5 de6
= = = = = =
−2e2 ∧ (h 1 − h 2 ) + e35 + e46 , 2e1 ∧ (h 1 − h 2 ) + e45 − e36 , 2e4 ∧ (h 3 − h 1 ) − e15 + e26 , −2e3 ∧ (h 3 − h 1 ) − e25 − e16 , −2e6 ∧ (h 2 − h 3 ) + e13 + e24 , 2e5 ∧ (h 2 − h 3 ) + e14 − e23 .
(37)
Let J denote the almost complex structure on M = F(1, 2) whose Kähler form is ω = e12 − e34 + e56 . (It is easy to check that ω, which a priori is a left-invariant 2-form on U3 , projects to M because L h i ω = 0.) J induces an orientation on M with volume form −e123456 . Let + + i − denote the associated complex volume form on M defined by the adT 3 -invariant form (e2 + i J e2 ) ∧ (e4 + i J e4 ) ∧ (e6 + i J e6 ). Explicitly, + = e136 + e246 + e235 − e145 ,
− = e236 − e146 − e135 − e245 .
Using (37) we readily obtain d(e12 ) = −d(e34 ) = d(e56 ) = + ,
(38)
so dω = 3 + ,
and
d − = −2ω2 .
The pair (g, J ) thus defines a nearly Kähler structure on M (a fact which we already knew). We fix now an element ξ ∈ su3 ⊂ u3 , and denote by X the right-invariant vector field on U3 defined by ξ . Consider the functions xi = g(X, ei ),
vi = g(X, h i ).
(39)
The functions vi are projectable to M and clearly v1 + v2 + v3 = 0. Let us introduce the vector fields on U3 , a1 = x6 e5 − x5 e6 ,
a2 = x3 e4 − x4 e3 ,
a3 = x2 e1 − x1 e2 .
The Hermitian Laplace Operator on Nearly Kähler Manifolds
271
One can check that they project to M. Of course, one has J a1 = x5 e5 + x6 e6 ,
J a2 = x3 e3 + x4 e4 ,
J a3 = x1 e1 + x2 e2 .
The commutator relations in SU3 yield dv1 = a2 − a3 ,
dv2 = a3 − a1 ,
dv3 = a1 − a2 .
(40)
Using (37) and some straightforward computations we obtain d(J a1 ) = (−a1 + a2 + a3 ) + + 4(v2 − v3 )e56 , d(J a2 ) = (a1 − a2 + a3 ) + + 4(v1 − v3 )e34 , d(J a3 ) = (a1 + a2 − a3 ) + + 4(v1 − v2 )e12 .
(41)
We claim that the 2-form ϕ = v1 e56 − v2 e34 + v3 e12
(42)
on M is of type (1,1), primitive, co-closed, and satisfies ϕ = 12ϕ. The first two assertions are obvious (recall that v1 + v2 + v3 = 0). In order to prove that ϕ is co-closed, it is enough to prove that dϕ ∧ ω = 0. Using (38) and (40) we compute: dϕ ∧ ω = [(a2 − a3 ) ∧ e56 − (a3 − a1 ) ∧ e34 + (a1 − a2 ) ∧ e12 ] ∧ (e12 − e34 + e56 ) = (a1 − a2 ) ∧ e1256 − (a3 − a2 ) ∧ e1234 + (a1 − a2 ) ∧ e3456 = 0. Finally, using (41), we get ϕ = d ∗ dϕ = − ∗ d ∗ [(a2 − a3 ) ∧ e56 − (a3 − a1 ) ∧ e34 + (a1 − a2 ) ∧ e12 ] = −∗ d[J a2 ∧ e12 + J a3 ∧ e34 + J a3 ∧ e56 − J a1 ∧ e12 − J a1 ∧ e34 − J a2 ∧ e56 ] = − ∗ [d(J a2 ) ∧ (e12 − e56 ) + d(J a3 ) ∧ (e34 + e56 ) − d(J a1 ) ∧ (e12 + e34 )] = − ∗ [(a1 + a2 + a3 ) + ∧ (e12 − e56 + e34 + e56 − e12 − e34 ) −2 (a2 + ) ∧ (e12 − e56 )−2(a3 + ) ∧ (e34 +e56 )+2(a1 + ) ∧ (e12 + e34 ) + 4(v1 − v3 )e34 ∧ (e12 − e56 ) + 4(v1 − v2 )e12 ∧ (e34 + e56 ) − 4(v2 − v3 )e56 ∧ (e12 + e34 )] = − ∗ [4(2v1 − v2 − v3 )e1234 + 4(v1 + v3 − 2v2 )e1256 + 4(2v3 − v1 − v2 )e3456 ] = − ∗ [12v1 e1234 − 12v2 e1256 + 12v3 e3456 ] = 12ϕ. Taking into account the inequality (36), we deduce at once the following Corollary 6.1. The space of infinitesimal nearly Kähler deformations of the nearly Kähler structure on F(1, 2) is isomorphic to the Lie algebra of SU3 . More precisely, every right-invariant vector field X on SU3 defines an element ϕ ∈ N K via the formulas (39) and (42). Acknowledgements. We are grateful to Gregor Weingart for helpful discussions and in particular for suggesting the statement of Lemma 5.4.
272
A. Moroianu, U. Semmelmann
References 1. Baum, H., Friedrich, Th., Grunewald, R., Kath, I.: Twistor and Killing Spinors on Riemannian Manifolds. Stuttgart–Leipzig: Teubner–Verlag, 1991 2. Belgun, F., Moroianu, A.: Nearly Kähler 6-manifolds with reduced holonomy. Ann. Global Anal. Geom. 19, 307–319 (2001) 3. Besse, A.: Einstein manifolds. Ergebnisse der Mathematik und ihrer Grenzgebiete (3) 10. Berlin: SpringerVerlag, 1987 4. Borel, A., Hirzebruch, F.: Characteristic classes and homogeneous spaces I. 80, 458–538 (1958) 5. Butruille, J.-B.: Classification des variétés approximativement kähleriennes homogènes. Ann. Global Anal. Geom. 27, 201–225 (2005) 6. Cleyton, R., Swann, A.: Einstein metrics via intrinsic or parallel torsion. Math. Z. 247, 513–528 (2004) 7. Friedrich, Th.: Nearly Kähler and nearly parallel G 2 -structures on spheres. Arch. Math. (Brno) 42, 241– 243 (2006) 8. Gray, A.: The structure of nearly Kähler manifolds. Math. Ann. 223, 233–248 (1976) 9. Moroianu, A., Nagy, P.-A., Semmelmann, U.: Unit Killing Vector Fields on Nearly Kähler Manifolds. Internat. J. Math. 16, 281–301 (2005) 10. Moroianu, A., Nagy, P.-A., Semmelmann, U.: Deformations of Nearly Kähler Structures. Pacific J. Math. 235, 57–72 (2008) 11. Moroianu, A., Semmelmann, U.: Infinitesimal Einstein Deformations of Nearly Kähler Metrics. to appear in Trans. Amer. Math. Soc., 2009 12. Nagy, P.-A.: Nearly Kähler geometry and Riemannian foliations. Asian J. Math. 3, 481–504 (2002) 13. Wolf, J., Gray, A.: Homogeneous spaces defined by Lie group automorphisms I, II. J. Differ. Geom. 2, 77–114, 115–159 (1968) Communicated by A. Connes
Commun. Math. Phys. 294, 273–301 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0933-y
Communications in
Mathematical Physics
Second-Order Corrections to Mean Field Evolution of Weakly Interacting Bosons. I. Manoussos G. Grillakis 1 , Matei Machedon 1 , Dionisios Margetis 1,2,3 1 Department of Mathematics, University of Maryland, College Park,
MD 20742, USA. E-mail:
[email protected]
2 Institute for Physical Science and Technology, University of Maryland,
College Park, MD 20742, USA
3 Center for Scientific Computation and Mathematical Modeling, University of Maryland,
College Park, MD 20742, USA Received: 31 March 2009 / Accepted: 30 July 2009 Published online: 2 October 2009 – © Springer-Verlag 2009
Abstract: Inspired by the works of Rodnianski and Schlein [31] and Wu [34,35], we derive a new nonlinear Schrödinger equation that describes a second-order correction to the usual tensor product (mean-field) approximation for the Hamiltonian evolution of a many-particle system in Bose-Einstein condensation. We show that our new equation, if it has solutions with appropriate smoothness and decay properties, implies a new Fock space estimate. We also show that for an interaction potential v(x) = χ (x)|x|−1 , where is sufficiently small and χ ∈ C0∞ even, our program can be easily implemented locally in time. We leave global in time issues, more singular potentials and sophisticated estimates for a subsequent part (Part II) of this paper. 1. Introduction An advance in physics in 1995 was the first experimental observation of atoms with integer spin (Bosons) occupying a macroscopic quantum state (condensate) in a dilute gas at very low temperatures [1,4]. This phenomenon of Bose-Einstein condensation has been observed in many similar experiments since. These observations have rekindled interest in the quantum theory of large Boson systems. For recent reviews, see e.g. [23,29]. A system of N interacting Bosons at zero temperature is described by a symmetric wave function satisfying the N -body Schrödinger equation. For large N , this description is impractical. It is thus desirable to replace the many-body evolution by effective (in an appropriate sense) partial differential equations for wave functions in much lower space dimensions. This approach has led to “mean-field” approximations in which the single particle wave function for the condensate satisfies nonlinear Schrödinger equations (in 3 + 1 dimensions). Under this approximation, the N -body wave function is viewed simply as a tensor product of one-particle states. For early related works, see the papers by Gross [15,16], Pitaevskii [28] and Wu [34,35]. In particular, Wu [34,35] introduced a second-order approximation for the Boson many-body wave function in terms of the pair-excitation function, a suitable kernel that describes the scattering of atom pairs
274
M. G. Grillakis, M. Machedon, D. Margetis
from the condensate to other states. Wu’s formulation forms a nontrivial extension of works by Lee, Huang and Yang [21] for the periodic Boson system. Approximations carried out for pair excitations [21,34,35] make use of quantized fields in the Fock space. (The Fock space formalism and Wu’s formulation are reviewed in Sects. 1.1 and 1.3, respectively.) Connecting mean-field approaches to the actual many-particle Hamiltonian evolution raises fundamental questions. One question is the rigorous derivation and interpretation of the mean field limit. Elgart, Erd˝os, Schlein and Yau [6–11] showed rigorously how mean-field limits for Bosons can be extracted in the limit N → ∞ by using Bogoliubov-Born-Green-Kirkwood-Yvon (BBGKY) hierarchies for reduced density matrices. Another issue concerns the convergence of the microscopic evolution towards the mean field dynamics. Recently, Rodnianski and Schlein [31] provided estimates for the rate of convergence in the case with Hartree dynamics by invoking the formalism of Fock space. In this paper, inspired by the works of Rodnianski and Schlein [31] and Wu [34,35], we derive a new nonlinear Schrödinger equation describing an improved approximation for the evolution of the Boson system. This approximation offers a second-order correction to the usual tensor product (mean field limit) for the many-body wave function. Our equation yields a corresponding new estimate in Fock space, which complements nicely the previous estimate [31]. The static version of the many-body problem is not studied here. The energy spectrum was addressed by Dyson [5] and by Lee, Huang and Yang [21]. A mathematical proof of the Bose-Einstein condensation for the time-independent case was provided recently by Lieb, Seiringer, Solovej and Yngvanson [22–25]. 1.1. Fock space formalism. Next, we review the Fock space F over L 2 (R3 ), following Rodnianski and Schlein [31]. The elements of F are vectors of the form ψ = (ψ0 , ψ1 (x1 ), ψ2 (x1 , x2 ), . . .), where ψ0 ∈ C and ψn ∈ L 2s (R3n ) are symmetric in x1 , . . . , xn . The Hilbert space structure of F is given by (φ, ψ) = n φn ψn d x. For f ∈ L 2 (R3 ) the (unbounded, closed, densely defined) creation operator a ∗ ( f ) : F → F and annihilation operator a( f¯) : F → F are defined by n ∗ 1 a ( f )ψn−1 (x1 , x2 , . . . , xn ) = √ f (x j )ψn−1 (x1 , . . . , x j−1 , x j+1 , . . . xn ), n j=1 √ a( f )ψn+1 (x1 , x2 , . . . , xn ) = n + 1 ψ(n+1) (x, x1 , . . . , xn ) f (x) d x.
The operator valued distributions ax∗ and ax defined by ∗ f (x)ax∗ d x, a (f) = a( f ) = f (x) ax d x. These distributions satisfy the canonical commutation relations [ax , a ∗y ] = δ(x − y), [ax , a y ] = [ax∗ , a ∗y ] = 0.
(1)
Second-Order Corrections for Weakly Interacting Bosons I
275
Let N be a fixed integer (the total number of particles), and v(x) be an even potential. Consider the Fock space Hamiltonian H N : F → F defined by 1 HN = ax∗ ∆ax d x + v(x − y)ax∗ a ∗y ax a y d x d y 2N 1 (2) =: H0 + V. N This H N is a diagonal operator which acts on each ψn in correspondence to the Hamiltonian H N ,n =
n
∆x j +
j=1
n 1 v(xi − x j ). 2N i, j=1
In the particular case n = N , this is the mean field Hamiltonian. Except for the Introduction, this paper deals only with the Fock space Hamiltonian. The reader is alerted that “PDE” Hamiltonians such as H N ,n will always have two subscripts. The sign of v will not play a role in our analysis. However, the reader is alerted that due to our sign convention, v ≤ 0 is the “good” sign. The time evolution in the coordinate space for Bose-Einstein condensation deals with the function eitHn,n ψ0
(3)
for tensor product initial data, i.e., if ψ0 (x1 , x2 , . . . , xn ) = φ0 (x1 )φ0 (x2 ) . . . φ0 (xn ), where φ0 L 2 (R3 ) = 1. This approach has been highly successful, even for very singular potentials, in the work of Elgart, Erd˝os, Schlein and Yau [6–11]. In this context, the convergence of evolution to the appropriate mean field limit (tensor product) as N → ∞ is established at the level of marginal density matrices γi(N ) in the trace norm topology. The density matrices are defined as (N ) γi (t, x1 , . . . , xi ; x1 , . . . xi ) = ψ(t, x1 , . . . , x N )ψ(t, x1 , . . . , x N ) d xi+1 · · · d x N .
1.2. Coherent states. There are alternative approaches, due to Hepp [17], Ginibre and Velo [13], and, most recently, Rodnianski and Schlein [31] which can treat Coulomb potentials v. These approaches rely on studying the Fock space evolution eit HN ψ 0 , where the initial data ψ 0 is a coherent state, ψ 0 = (c0 , c1 φ0 (x1 ), c2 φ0 (x1 )φ0 (x2 ), · · · ); see (4) below. The evolution (3) can then be extracted as a “Fourier coefficient” from the Fock space evolution, see [31]. Under the assumption that v is a Coulomb potential, this approach leads to strong L 2 -convergence, still at the level of the density matrices (N ) γi , as we will briefly explain below. To clarify the issues involved, let us consider the one-particle wave function φ(t, x) (to be determined later as the solution of a Hartree equation), satisfying the initial condition φ(0, x) = φ0 (x). Define the skew-Hermitian unbounded operator A(φ) = a(φ) − a ∗ (φ)
276
M. G. Grillakis, M. Machedon, D. Margetis
and the vacuum state Ω = (1, 0, 0, . . .) ∈ F. Accordingly, consider the operator √
W (φ) = e−
N A(φ)
,
which is the Weyl operator used by Rodnianski and Schlein [31]. The coherent state for the initial data φ0 is √
ψ 0 = W (φ0 )Ω = e− =e
−N φ2 /2
N A(φ0 )
1, . . . ,
Ω n 1/2
N n!
φ0 (x1 ) . . . φ0 (xn ), · · ·
.
(4)
Hence, the top candidate approximation for eitHN ψ 0 reads √
ψ tensor (t) = e−
N A(φ(t,·))
Ω.
(5)
Rodnianski and Schlein [31] showed that this approximation works (under suitable assumptions on v), in the sense that √ √ 1 itHN e ψ 0 , a ∗y ax eit HN ψ 0 − e− N A(φ(t,·)) Ω, a ∗y ax e− N A(φ(t,·)) Ω Tr N eCt ) N → ∞; = O( N the symbol Tr here stands for the trace norm in x ∈ R3 and y ∈ R3 . The first term in the last relation, including N1 , is essentially the density matrix γ1(N ) (t, x, y). For the precise statement of the problem and details of the proof, see Theorem 3.1 of Rodnianski and Schlein [31]. Our goal here is to find an explicit approximation for the evolution in the Fock space. For this purpose, we adopt an idea germane to Wu’s second-order approximation for the N -body wave function in Fock space [34,35].
1.3. Wu’s approach. We first comment on the case with periodic boundary conditions, when the condensate is the zero-momentum state. For this setting, Lee, Huang and Yang [21] studied systematically the scattering of atoms from the condensate to states of opposite momenta. By diagonalizing an approximation for the Hamiltonian in Fock space, these authors derived a formula for the N -particle wave function that deviates from the usual tensor product, as it expresses excitation of particles from zero monentum to pairs of opposite momenta. For non-periodic settings, Wu [34,35] invokes the splitting ax = a0 (t)φ(t, x) + ax,1 (t), where a0 corresponds to the condensate, [a0 , a0∗ ] = 1, and ax,1 corresponds to ∗ ]. Wu applies the following states orthogonal to the condensate, [a0 , ax,1 ] = 0 = [a0 , ax,1 ansatz for the N -body wave function in Fock space: N (t) eP [K 0 ] ψ N0 (t),
(6)
where ψ N0 (t) describes the tensor product, N (t) is a normalization factor, and P[K 0 ] is an operator that averages out in space the excitation of particles from the condensate φ to
Second-Order Corrections for Weakly Interacting Bosons I
277
other states with the effective kernel (pair excitation function) K 0 . An explicit formula for P[K 0 ] is ∗ P[K 0 ] = [2N0 (t)]−1 ax,1 a ∗y,1 K 0 (t, x, y) a0 (t)2 , (7) where N0 is the expectation value of particle number at the condensate. This K 0 is not a-priori known (in contrast to the case of the classical Boltzmann gas) but is determined by means consistent with the many-body dynamics. In the periodic case, (6) reduces to the many-body wave function of Lee, Huang and Yang [21]. Wu derives a coupled system of dispersive hyperbolic partial differential equations for (φ, K 0 ) via an approximation for the N -body Hamiltonian that is consistent with ansatz (6). A feature of this system is the spatially nonlocal couplings induced by K 0 . Observable quantities such that the depletion of the condensate can be computed directly from solutions of this PDE system. This system has been solved only in a limited number of cases [26,27,35].
1.4. Scope and outline. Our objective in this work is to find an explicit approximation for the evolution eitHN ψ 0 in the Fock space norm, where ψ 0 is the coherent state (4). This would imply an approximation for the evolution eitHN ,N ψ0 in L 2 (R3N ) as N → ∞. To the best of our knowledge, no such approximation is available in the mathematics or physics literature. In particular, the tensor product type approximation (5) for φ satisfying a Hartree equation, as in [31], is not known to be such a Fock space approximation (nor do we expect it to be). To accomplish our goal, we propose to modify (5) in two ways. One minor correction is the multiplication by an oscillatory term. A second correction is a composition with a second-order “Weyl operator”. Both corrections are inspired by the work of Wu [34,35]; see also [26,27]. However, our set-up and derived equation are essentially different from these works. We proceed to describe the second order correction. Let k(t, x, y) = k(t, y, x) be a function (or kernel) to be determined later, with k(0, x, y) = 0. The minimum regularity expected of k is k ∈ L 2 (d x d y) for a.e. t. We define the operator 1 (8) B= k(t, x, y)ax a y − k(t, x, y)ax∗ a ∗y d x d y. 2 Notice that B is skew-Hermitian, i.e., iB is self-adjoint. The operator e B could be defined by the spectral theorem; see [30]. However, we prefer the more direct approach of defining it first on the dense subset of vectors with finitely many non-zero components, where it can be defined by a convergent Taylor series if k L 2 (d xd y) is sufficiently small. Indeed, B restricted to the subspace of vectors with all entries past the first N identically zero has norm ≤ C N k L 2 . Then e B is extended to F as a unitary operator.
278
M. G. Grillakis, M. Machedon, D. Margetis
Now we have described all ingredients needed to state our results and derivations. The remainder of the paper is organized as follows. In Section 2 we state our main result and outline its proof. In Sect. 3 we study implications of the Hartree equation satisfied by the one-particle wave function φ(t, x). In Sect. 4 we develop bookkeeping tools of Lie algebra for computing requisite operators containing B. In Sect. 5 we study the evolution equation for a matrix K that involves the kernel k. In Sect. 6 we develop an argument for the existence of solution to the equation for the kernel k. In Sect. 7 we find conditions under which terms involved in the error term e B V e−B are bounded. In Sect. 8 we study similarly the error term e B [A, V ]e−B . In Sect. 9 we show that we can control traces needed in derivations. 2. Statement of Main Result and Outline of Proof In this section we state our strategy for general potentials satisfying certain properties. Later in the paper we show that all assumptions of the related theorem are satisfied , : sufficiently small, and χ ∈ C0∞ : even. locally in time for v(x) = χ (x) |x| Theorem 1. Suppose that v is an even potential. Let φ be a smooth solution of the Hartree equation i
∂φ + ∆φ + (v ∗ |φ|2 )φ = 0 ∂t
(9)
with initial conditions φ0 , and assume the three conditions listed below: 1. Assume that we have k(t, x, y) ∈ L 2 (d xd y) for a.e. t, where k is symmetric, and solves (iut + ug T + gu − (1 + p)m) = (i pt + [g, p] + um)(1 + p)−1 u,
(10)
where all products in (10) are interpreted as spatial compositions of kernels, “1” is the identity operator, and u(t, x, y) := sh(k) := k +
1 kkk + · · · , 3!
1 kk + · · · , (11) 2! g(t, x, y) := −∆x δ(x − y) − v(x − y)φ(t, x)φ(t, y) − (v ∗ |φ|2 )(t, x)δ(x − y),
δ(x − y) + p(t, x, y) := ch(k) := δ(x − y) + m(t, x, y) := v(x − y)φ(t, x)φ(t, y). 2. Also, assume that the functions
f (t) := e B [A, V ]e−B ΩF and g(t) := e B V e−B ΩF are locally integrable (V is defined in (2)).
Second-Order Corrections for Weakly Interacting Bosons I
3. Finally, assume that
279
d(t, x, x) d x is locally integrable in time, where d(t, x, y) = ish(k)t + sh(k)g T + gsh(k) sh(k) − (ich(k)t + [g, ch(k)]) ch(k) −sh(k)mch(k) − ch(k)msh(k).
Then, there exist real functions χ0 , χ1 such that √
e−
N A(t) −B(t) −i
t
≤
0
e
e t
f (s)ds + √ N
t
0
0 (N χ0 (s)+χ1 (s))ds
Ω − eitHN ψ 0 F
g(s)ds . N
(12)
Recall that we defined (see Sect. 1) √
ψ 0 = e−
N A(0)
Ω an arbitrary coherent state (initial data),
A(t) = a(φ(t, ·)) − a ∗ (φ(t, ·)), 1 B(t) = k(t, x, y)ax a y − k(t, x, y)ax∗ a ∗y d x d y. 2 A few remarks on Theorem 1 are in order. Remark 1. Written explicitly, the left-hand side of (10) equals ∂ iut + ug T + gu − (1 + p)m = i − ∆x − ∆ y u(t, x, y) ∂t −φ(t, x) v(x − z)φ(t, z)u(t, z, y) dz − φ(t, y) u(t, x, z)v(z − y)φ(t, z) dz −(v ∗ |φ|2 )(t, x)u(t, x, y) − (v ∗ |φ|2 )(t, y)u(t, x, y) −v(x − y)φ(t, x)φ(t, y) −φ(t, y) (1 + p)(t, x, z)v(z − y)φ(t, z) dz. The main term in the right-hand side equals ∂ i pt + [g, p] + um = i p(t, x, y) + −∆x + ∆ y p(t, x, y) ∂t −φ(t, x) v(x − z)φ(t, z) p(t, z, y) dz +φ(t, y) p(t, x, z)v(z − y)φ(t, z) dz −(v ∗ |φ|2 )(t, x) p(t, x, y) + (v ∗ |φ|2 )(t, y) p(t, x, y) + u(t, x, z)v(z − y)φ(t, z)φ(t, x) dz. Remark 2. The algebra, as well as the local analysis presented in this paper do not depend on the sign of v. However, the global in time analysis of our equations would require v to be non-positive.
280
M. G. Grillakis, M. Machedon, D. Margetis
Remark 3. Our √techniques would allow us to consider more general initial data of the form ψ 0 = e− N A(0) e−B(0) Ω. For convenience, we only consider the case of tensor products (B(0) = 0) in this paper. √
Proof. Since e ei
NA
and e B are unitary, the left-hand side of (12) equals
t
0 (N χ0 (s)+χ1 (s))ds
e B(t) e
√
√ N A(t) itH N − N A(0)
e
e
Ω − ΩF .
Define Ψ (t) = e B(t) e
√
√ N A(t) it H − N A(0)
e
e
Ω.
In Corollary 1 of Sect. 5 we show that our equations for φ, k insure that 1 ∂ Ψ = LΨ, i ∂t where L = L − N χ0 − χ1 for some L: Hermitian, i.e. L = L ∗ , where L commutes with functions of time, χ0 , χ1 are real functions of time, and, most importantly (see Corollary 1 of Sect. 5 and the remark following it), LΩF ≤ N −1/2 e B [A, V ]e−B ΩF + N −1 e B V e−B ΩF .
(13)
We apply energy estimates to
t 1 ∂
− L (ei 0 (N χ0 (s)+χ1 (s))ds Ψ − Ω) = LΩ. i ∂t
Explicitly, ∂ i t (N χ0 (s)+χ1 )ds (e 0 Ψ − Ω)2F ∂t t ∂ i t (N χ0 (s)+χ1 )ds i 0 (N χ0 (s)+χ1 )ds 0 = 2 (e Ψ − Ω), e Ψ −Ω ∂t t t ∂ i 0 (N χ0 (s)+χ1 )ds i 0 (N χ0 (s)+χ1 )ds
− i L (e Ψ − Ω), e Ψ −Ω = 2 ∂t t = 2 i LΩ, ei 0 (N χ0 (s)+χ1 )ds Ψ − Ω t ≤ 2 N−1/2 e B [A, V ]e−B ΩF + N−1 e B V e−B ΩF (ei 0 (N χ0 (s)+χ1 )ds Ψ −Ω)F . Thus t ∂ (ei 0 (N χ0 (s)+χ1 )ds Ψ − Ω) ≤ N −1/2 e B [A, V ]e−B ΩF + N −1 e B V e−B ΩF ∂t
and (12) holds. This concludes the proof.
Second-Order Corrections for Weakly Interacting Bosons I
281
3. The Hartree Equation In this section we see how far we can go by using only the Hartree equation for the one-particle wave function φ. Lemma 1. The following commutation relations hold (where the t dependence is suppressed, A denotes A(φ) and V is defined by formula (2)): [A, V ] = v(x − y) φ(y)ax∗ ax a y + φ(y)ax∗ a ∗y ax d x d y, [A, [A, V ]] = v(x − y) φ(y)φ(x)ax a y + φ(y)φ(x)ax∗ a ∗y + 2φ(y)φ(x)ax∗ a y d x d y, +2 v ∗ |φ 2 | (x)ax∗ ax d x, A, A, [A, V ] = 6 v ∗ |φ 2 | (x) φ(x)ax∗ + φ(x)ax d x,
A, A, [ A, [A, V ]]
= 12 v ∗ |φ 2 | (x)|φ(x)|2 d x.
(14)
Proof. This is an elementary calculation and is left to the interested reader. √
√
Now, we consider Ψ1 (t) = e N A(t) eit H e− N A(0) Ω for which we have the basic calculation in the spirit of Hepp [17], Ginibre-Velo [13], and Rodnianski-Schlein [31]; see Eq. (3.7) in [31]. Proposition 1. If φ satisfies the Hartree equation i
∂φ + ∆φ + (v ∗ |φ|2 )φ = 0 ∂t
while Ψ1 (t) = e
√
√ N A(t) it H − N A(0)
e
e
Ω,
then Ψ1 (t) satisfies 1 1 ∂ Ψ1 (t) = H0 + [A, [A, V ]] i ∂t 2 N +N −1/2 [A, V ] + N −1 V − v(x − y)|φ(t, x)|2 |φ(t, y)|2 d x d y Ψ1 (t). 2 Proof. Recall the formulas ∂ C(t) −C(t) 1 ˙ + ··· ˙ + 1 C, [C, C] e = C˙ + [C, C] e ∂t 2! 3! and eC H e−C = H + [C, H ] +
1 [C, [C, H ]] + · · · 2!
282
M. G. Grillakis, M. Machedon, D. Margetis
Applying these relations to C =
√
N A we get
1 ∂ ψ1 (t) = L 1 ψ1 , i ∂t
(15)
where
√ √ 1 ∂ √ N A(t) −√ N A(t) e e + e N A(t) H e− N A(t) i ∂t 1 N ˙ + H + N 1/2 [A, H0 ] N 1/2 A˙ + [A, A] = i 2 N +N −1/2 [A, V ] + [A, [A, H0 ]] 2 N 1 N 1/2 A, A, [A, V ] + A, A, [ A, [A, V ]] . × [A, [A, V ]] + 2 3! 4! √ Eliminating the terms with a weight of N , or setting 1 1 ˙ A, A, [A, V ] = 0, (16) A + [A, H0 ] + i 3! L1 =
is exactly equivalent to the Hartree equation (9). By taking an additional bracket with A in (16), we have 1 1 ˙ [A, A] + [A, [A, H0 ]] + A, A, [ A, [A, V ]] = 0, i 3! and thus simplify (15) to 1 ∂ 1 ψ1 (t) = H0 + [A, [A, V ]] i ∂t 2 +N This concludes the proof.
−1/2
[A, V ] + N
−1
1 A, A, [A, [A, V ]] ψ1 . V−N 4!
on the right-hand side are the main ones. The next two terms are two terms The first 1 1 and O N . The last term equals O √ N
−
N 2
v(x − y)|φ(t, x)|2 |φ(t, y)|2 d x d y := −N χ0 .
Notice that L 1 (Ω) is not small because of the presence of ax∗ a ∗y in [A, [A, V ]]. In order to eliminate these terms, we introduce B (see (8)) and take ψ = e B ψ1 . Accordingly, we compute 1 ∂ ψ = Lψ, i ∂t
Second-Order Corrections for Weakly Interacting Bosons I
where 1 L= i
283
∂ B −B e e + e B L 1 e−B ∂t
= L Q + N −1/2 e B [A, V ]e−B + N −1 e B V e−B − N χ0 , and LQ
1 = i
∂ B −B 1 B H0 + [A, [A, V ]] e−B e e +e ∂t 2
(17)
contains all quadratics in the operators a, a ∗ . Equation (10) for k turns out to be equivalent to the requirement that L has no terms of the form a ∗ a ∗ . Terms of the form aa ∗ will occur, and will be converted to a ∗ a at the expense of χ1 . In other words, we require that L Q have no terms of the form a ∗ a ∗ . For a similar argument (but for a different set-up), see Wu [35]. 4. The Lie Algebra of “Symplectic Matrices” In this section we describe the bookkeeping tools needed to compute L Q of (17) in closed form. The results of this section are essentially standard, but they are included here for the sake of completeness. We start with the remark that [a( f 1 ) + a ∗ (g1 ), a( f 2 ) + a ∗ (g2 )] = f 1 g2 − f 2 g1 f = − f 1 g1 J 2 , (18) g2 where
0 −δ(x − y) J= . δ(x − y) 0
This observation explains why we have to invoke symplectic linear algebra. We thus consider the infinite-dimensional Lie algebra sp of “matrices” of the form d k S(d, k, l) = l −d T for symmetric kernels k = k(t, x, y) and l = l(t, x, y), and arbitrary kernel d(t, x, y). (The dependence on t will be suppressed when not needed.) This situation is analogous to the Lie algebra of the finite-dimensional complex symplectic group, with x, y playing the role of i and j. We also consider the Lie algebra Quad of quadratics of the form ∗ d k 1 −a y ∗ ax ax Q(d, k, l) := l −d T ay 2 ∗ ∗ ax a y + a y ax 1 = − d(x, y) dx dy + k(x, y)ax a y d x d y 2 2 1 − (19) l(x, y)ax∗ a ∗y d x d y 2
284
M. G. Grillakis, M. Machedon, D. Margetis
(k, l and d as before). Furthermore, we agree to identify operators which differ (formally) by a scalar operator. Thus, d(x, y)ax a ∗y is considered equivalent to d(x, y)a ∗y ax . We recall the following result related to the metaplectic representation (see, e.g. [12]). Theorem 2. Let S = S(d, k, l), Q = Q(d, k, l) related as above. Let f , g be functions (or distributions). Denote f := f (x)ax + g(x)ax∗ d x. (ax , ax∗ ) g We have the following commutation relation: f f ∗ ∗ = (ax , ax )S , Q, (ax , ax ) g g where products are interpreted as compositions. We also have f f Q ∗ −Q ∗ S , e e (ax , ax ) = (ax , ax )e g g
(20)
(21)
provided that e Q makes sense as a unitary operator (Q: skew-Hermitian). Proof. The commutation relation (20) can be easily checked directly, but we point out that it follows from (18). In fact, using (18), for any rank one quadratic we have a( f 1 ) + a ∗ (g1 ) a( f 2 ) + a ∗ (g2 ) , a( f ) + a ∗ (g) f1 f2 f ∗ f 1 g1 + f 2 g2 J = − ax ax . g g2 g1 Thus, for any R we have a f ax ax∗ R ∗y , a( f ) + a ∗ (g) = − ax ax∗ R + R T J . ay g Now specialize to R = 21 S J , S ∈ sp, and use S T = J S J to complete the proof. The second part, Eq. (21), follows from the identity e Q Ce−Q = C + [Q, C] +
1 [Q, [Q, C]] + · · · , 2!
or, in the language of adjoint representations, Ad(e Q )(C) = ead(Q) (C), which is applied to C = a( f ) + a ∗ (g).
A closely related result is provided by the following theorem. Theorem 3. 1. The linear map I : sp → Quad defined by S(d, k, l) → Q(d, k, l) is a Lie algebra isomorphism. 2. Moreover, if S = S(t), Q = Q(t) and I(S(t)) = Q(t) is skew-Hermitian, so that e Q is well defined, we have ∂ Q −Q ∂ S −S = e . e e e (22) I ∂t ∂t
Second-Order Corrections for Weakly Interacting Bosons I
285
3. Also, if R ∈ sp, we have I e S Re−S = e Q I(R)e−Q .
(23)
Remark 4. In the finite-dimensional case, this is (closely related to) the “infinitesimal metaplectic representation”; see p. 186 in [12] . In the infinite dimensional case, we must be careful, as some of our operators are not of trace class. For instance, ax ax∗ does not make sense. Proof. First, we point out that (21) implies (23), at least in the case where R is the “rank one” matrix f hi . R= g Notice that (21) can also be written as ax −Q S T ax f g eQ f g e . = e ax∗ ax∗ In conclusion, we find ∗ −a y −Q ∗ e ax ax R e ay f a Q ∗ h i a a J a ∗y e−Q =e x x g y f a Q −Q Q ∗ a h i a e e J a ∗y e−Q =e x x g y f J S J ay h i Je = ax ax∗ e S a ∗y g ∗ −a y , = ax ax∗ e S Re−S ay Q
since S T = J S J if S ∈ sp, and J e J S J = e−S J . We now give a direct proof that (19) preserves Lie brackets. Denote the quadratic 1 ∗ ∗ ∗ ∗ ∗ building blocks by Q x y = ax a y , Q x y = ax a y , N x y = 2 ax a y + a y ax . One can verify the following commutation relations, which will be also needed below: Q x y , Q ∗zw = δ(x − z)N yw + δ(x − w)N yz + δ(y − z)N xw + δ(y − w)N x z , (24) (25) Q x y , Nzw = δ(x − w)Q yz + δ(y − w)Q x z , ∗ ∗ (26) N x y , Q zw = δ(x − z)Q yw + δ(x − w)Q yz , N x y , Nzw = δ(x − w)Nzy − δ(y − z)N xw . (27) Using (24) we compute 1 1 k(x, y)ax a y d xd y, − l(x, y)ax∗ a ∗y d xd y = − (kl)(x, y)N x y d x d y, 2 2
286
M. G. Grillakis, M. Machedon, D. Margetis
which corresponds to the relation 0 0 k , l 0 0
0 0
=
kl 0 . 0 −lk
The other three cases are similar. To prove (22), expand both the left-hand side and the right-hand side as ∂ S −S I e e ∂t 1 ˙ ˙ = I S + [S, S] + · · · 2 1 ˙ + ··· = Q˙ + [Q, Q] 2 ∂ Q −Q e . = e ∂t The proof of (23) is along the same lines.
Remark 5. Note on rigor: All the Lie algebra results that we have used are standard in the finite-dimensional case. In our applications, S will be K , where K is a matrix of the form (29), see below, and Q will be B = I(K ). The unbounded operator B is skew-Hermitian and e B ψ is defined by a convergent Taylor series if ψ ∈ F has only finitely many non-zero components, provided k(t, ·, ·) L 2 (d x d y) is small. We then extend e B to all F as a unitary operator. The norm k(t, ·, ·) L 2 (d x d y) iterates under compositions; thus, the kernel e K is well defined by its convergent Taylor expansion. In the expression e B Pe−B = P + [B, P] + · · ·
(28)
for P, a first- or second-order polynomial in a, a ∗ , we point out that the right-hand side stays a polynomial of the same degree, and converges when applied to a Fock space vector with finitely many non-zero components. For our application, we need to know if (28) is true when applied to Ω. The same comment applies to the series ∂ B −B 1 ˙ + ··· . e e = B˙ + [B, B] ∂t 2 5. Equation for Kernel k Now apply the isomorphism of the previous section to the operator B = I(K ) for
0 k(t, x, y) . k(t, x, y) 0
K =
(29)
This agrees to the letter with the isomorphism (19). The next two isomorphisms, (30) and (31), require special treatment because aa ∗ terms mirroring the a ∗ a terms are missing
Second-Order Corrections for Weakly Interacting Bosons I
287
in (2), (14). However, the discrepancy only happens on the diagonal. Once the relevant terms are commuted with B, they fit the pattern exactly. It isn’t quite true that −(∆δ)(x − y) 0 H0 = I 0 (∆δ)(x − y) −∆ 0 =I (30) 0 ∆ since, strictly speaking, ∗ ax ∆ax + ax ∆ax∗ −(∆δ)(x − y) 0 = I dx 0 (∆δ)(x − y) 2 is undefined. However, one can compute directly that [∆x ax , a ∗y ] = (∆δ)(x − y). Using that, we compute 1 [B, H0 ] = (∆x + ∆ y )k(x, y)ax a y + (∆x + ∆ y )k(x, y)ax∗ a ∗y d x d y. 2 This commutator is in agreement with (29), (30), and the result can be represented in accordance with (19), namely 0k −(∆δ)(x − y) 0 [B, H0 ] = I , . 0 (∆δ)(x − y) k0 We also have e B H0 e−B − H0 0 −(∆δ)(x − y) 0 K −(∆δ)(x − y) −K e =I e − , 0 (∆δ)(x − y) 0 (∆δ)(x − y) since e B H0 e−B − H0 = [B, H0 ] + 21 [B, [B, H0 ]] + · · ·. The same comment applies to the diagonal part of 1 v12φ 1 φ 2 −v12 φ 1 φ2 − v ∗ |φ|2 δ12 , (31) [A, [A, V ]] = I −v12 φ1 φ2 v12 φ1 φ 2 + v ∗ |φ|2 δ12 2 where v12 φ1 φ2 is an abbreviation for the product v(x − y)φ(x)φ(y), etc. Formula (31) isn’t quite true either, but becomes true after commuting with B. To apply our isomorphism, we quarantine the “bad” terms in (30) and the diagonal part of (31). Define g 0 0 m , G= and M= −m 0 0 −g T where g = −∆δ12 − v12 φ 1 φ2 − (v ∗ |φ|2 )δ12 , m = v12 φ 1 φ 2 , and split H0 +
1 [A, [A, V ]] = HG + I(M), 2
288
M. G. Grillakis, M. Machedon, D. Margetis
where
HG = H0 + v(x − y)φ(y)φ(x)ax∗ a y d x d y + v ∗ |φ 2 | (x)ax∗ ax d x.
(32)
By the above discussion we have [B, HG ] = I([K , G]) and B −B K [e , HG ]e = I([e , G]e−K ). Write
1 ∂ B −B 1 e e + e B H0 + [A, [A, V ]] e−B i ∂t 2 1 ∂ B −B e e + HG + [e B , HG ]e−B + e B I(M)e−B = i ∂t 1 ∂ K −K e e = HG + I + [e K , G]e−K + e K Me−K i ∂t = HG + I(M1 + M2 + M3 ).
LQ =
(33)
Notice that if K is given by (29), then ch(k) sh(k) K e = , sh(k) ch(k) where 1 1 ch(k) = I + kk + kkkk + · · · , (34) 2 4! and similarly for sh(k). Products are interpreted, of course, as compositions of operators. We compute 1 ch(k)t sh(k)t ch(k) −sh(k) M1 = −sh(k) ch(k) i sh(k)t ch(k)t 1 ch(k)t ch(k) − sh(k)t sh(k) −ch(k)t sh(k) + sh(k)t ch(k) = ∗ ∗ i T [ch(k), g] −sh(k)g − gsh(k) [e K , G] = ∗ ∗ and M2 = [e K , G]e−K =
[ch, g] ch + (shg T + gsh)sh −[ch, g]sh − (shg T + gsh)ch , ∗ ∗
where sh is an abbreviation for sh(k), etc, and −shm ch − chmsh shmsh + chmch . M3 = e K Me−K = ∗ ∗ Now define M = M1 + M2 + M3 . We have proved the following theorem.
Second-Order Corrections for Weakly Interacting Bosons I
289
Theorem 4. Recall the isomorphism (19) of Theorem 3. 1. If L Q is given by (17), then
L Q = H0 + v(x − y)φ(y)φ(x)ax∗ a y d x d y + v ∗ |φ 2 | (x)ax∗ ax d x + I (M) .
(35)
2. The coefficient of ax a y in I (M) is −M12 or (ish(k)t + sh(k)g T + gsh(k))ch(k) − (ich(k)t − [ch(k), g])sh(k) −sh(k)msh(k) − ch(k)mch(k). 3. The coefficient of ax∗ a ∗y equals minus the complex conjugate of the coefficient of ax a y . 4. The coefficient of −
ax a ∗y + a ∗y ax 2
is M11 , or d(t, x, y) = ish(k)t + sh(k)g T + gsh(k) sh(k) − (ich(k)t + [g, ch(k)]) ch(k) −sh(k)mch(k) − ch(k)msh(k).
(36)
Corollary 1. If φ and k satisfy (9) and (10) of Theorem 1, then the coefficients of ax a y and ax∗ a ∗y drop out and L Q becomes ∗ v ∗ |φ 2 | (x)ax∗ ax d x L Q = H0 + v(x − y)φ(t, y)φ(t, x)ax a y d x d y + ax a ∗y + a ∗y ax − d(t, x, y) d x d y, 2 where d is given by (36) and the full operator reads ∗ v ∗ |φ 2 | (x)ax∗ ax d x L = H0 + v(x − y)φ(y)φ(t, x)ax a y d xd y + − d(t, x, y)a ∗y ax d x + N −1/2 e B [A, V ]e−B + N −1 e B V e−B − N χ0 − χ1 := L − N χ 0 − χ1 , and 1 χ0 = 2
v(x − y)|φ(t, x)|2 |φ(t, y)|2 d x d y, 1 χ1 (t) = − d(t, x, x)d x. 2 Remark 6. Notice that
LΩ = N −1/2 e B [A, V ]e−B + N −1 e B V e−B Ω,
and therefore we can derive the bound LΩ ≤ N −1/2 e B [A, V ]e−B Ω + N −1 e B V e−B Ω.
290
M. G. Grillakis, M. Machedon, D. Margetis
Also, L is (formally) self-adjoint by construction. The kernel d(t, x, y), being the sum of the (1,1) entry of the self-adjoint matrices 1i ∂t∂ e K e−K , [e K , G]e−K = e K Ge−K − G and the visibly self-adjoint term −sh(k)mch(k) − ch(k)msh(k), is self-adjoint; thus, it has a real trace. Hence, L is also self-adjoint. In the remainder of this paper, we check that the hypotheses of our main theorem are satisfied, locally in time, for the potential v(x) = χ (x) |x| . 6. Solutions to Equation 10 0 0 Theorem 5. Let 0 be sufficiently small and assume that v(x) = |x| , or v(x) = χ (x) |x| for χ ∈ C0∞ (R3 ). Assume that φ is a smooth solution to the Hartree equation (16), φ L 2 (d x) = 1. Then there exists k ∈ L ∞ ([0, 1])L 2 (d xd y) solving (10) with initial conditions k(0, x, y) = 0 for 0 ≤ t ≤ 1. The solution k satisfies the following additional properties:
1.
2.
3.
∂ i − ∆x − ∆ y k L ∞ [0,1]L 2 (d xd y) ≤ C, ∂t ∂ i − ∆x − ∆ y sh(k) L ∞ [0,1]L 2 (d xd y) ≤ C, ∂t ∂ i − ∆x + ∆ y p L ∞ [0,1]L 2 (d xd y) ≤ C. ∂t
4. The kernel k agrees on [0, 1] with a kernel k for which k
1 1
X 2,2+
≤ C;
see (38) for the definition of the space X s,δ and, of course, 21 + denotes a fixed number slightly bigger than 21 . Proof. We first establish some notation. Let S denote the Schrödinger operator S=i
∂ − ∆x − ∆ y ∂t
and let T be the transport operator T =i
∂ − ∆x + ∆ y . ∂t
Let : L 2 (d xd y) → L 2 (d xd y) denote schematically any linear operator of operator norm ≤ C0 , where C is a “universal constant”. In practice, will be (composition with)
Second-Order Corrections for Weakly Interacting Bosons I
291
a kernel of the type φ(t, x)φ(t, y)v(x − y), or multiplication by v ∗ |φ|2 . Also, recall the inhomogeneous term m(t, x, y) = v(x − y)φ(t, x)φ(t, y). Then, Eq. (10), written explicitly, becomes Sk = m + S(k − u) + (u) + ( p) + (T p + ( p) + (u))(1 + p)−1 u.
(37)
Note that ch(k)2 − sh(k)sh(k) = 1; thus, 1 + p = ch(k) ≥ 1 as an operator and (1 + p)−1 is bounded from L 2 to L 2 . We plan to iterate in the norm N (k) = k L ∞ [0,1]L 2 (d xd y) + Sk L ∞ [0,1]L 2 (d xd y) . Notice that m L 2 (d xd y) ≤ C0 . Now solve Sk0 = m with initial conditions k0 (0, ·, ·) = 0, where N (k0 ) ≤ C0 . Define u 0 , p0 corresponding to k0 . For the next iterate, solve Sk1 = m + S(k0 − u 0 ) + (u 0 ) + ( p0 ) + (T p0 + ( p0 ) + (u 0 ))(1 + p0 )−1 u 0 ; the non-linear terms satisfy S(u 0 − k0 ) L ∞ [0,1]L 2 (d xd y) 1 (Sk0 )k 0 k0 − k0 (Sk0 )k0 + k0 k 0 Sk0 + · · · L ∞ [0,1]L 2 (d xd y) = 3! = O(N (k0 )3 ). Also, recalling that p0 = ch(k0 ) − 1, we have 1 T ( p0 ) L ∞ [0,1]L 2 (d xd y) = (Sk0 )k 0 − k0 (Sk0 ) + · · · L ∞ [0,1]L 2 (d xd y) 2 = O N (k0 )2 . Thus, N (k1 ) ≤ C0 + C02 . Continuing this way, we obtain a fixed point solution in this space which satisfies the first three requirements of Theorem 5. N a N a In fact, we can apply the same argument to ∂t∂ D k, since ∂t∂ D m ∈ L ∞ [0, 1] 1 2 L (d x d y) for 0 ≤ a < 2 . However, we cannot repeat the argument for D 1/2 k. We would like to have S D 1/2 k L ∞ [0,1]L 2 (d x d y) finite. Unfortunately, this misses “logarithmically” because of the singularity of v. Fortunately, we can use the well-known X s,δ spaces (see [2,18,20]) to show that |S|s D 1/2 u L 2 (dt)L 2 (d x d y) is finite locally in time for (all) 1 > s > 21 . This assertion will be sufficient for our purposes. Recall the definition of X s,δ : δ u L 2 (dτ dξ ) := u X s,δ . (38) |ξ |s |τ − |ξ |2 | + 1 Going back to (37), we write S(k) = m + F,
292
M. G. Grillakis, M. Machedon, D. Margetis
where we define the expression F(k) := S(k − u) − (u) + pm + (T ( p) + ( p) + um) (1 + p)−1 u. The idea is to localize in time on the right-hand side: S( k) = χ (t) (m + F), k = k on [0, 1]. where χ ∈ C0∞ (R), χ = 1 on [0, 1]. Then, N a D k L 2 [0,1]L 2 (d x d y) ≤ C for As we already pointed out, we can estimate S ∂t∂ 1
0 ≤ a < 2 . We can further localize k in time to insure that these relations hold globally in time. By using the triangle inequality |τ −|ξ |2 |+|τ | ≥ |ξ |2 , we immediately conclude that 1+ 3 2 |ξ | 2 − |τ − |ξ |2 | + 1 kχ L 2 (dτ dξ ) ≤ C.
7. Error Term e B V e−B The goal of this section is to list explicitly all terms in e B V e−B and to find conditions under which these terms are bounded. Recall that V is defined by V = v(x0 − y0 )Q ∗x0 y0 Q x0 y0 d x0 dy0 . For simplicity, shb(k) denotes either sh(k) or sh(k), and chb(k) denotes either ch(k) or ch(k). Let x0 = y0 ; we obtain e B Q ∗x0 y0 Q x0 y0 e−B = e B Q ∗x0 y0 e−B e B Q x0 y0 e−B . According to the isomorphism (19), we have Q ∗x0 y0 = I
0 −2δ(x − x0 )δ(y − y0 )
0 , 0
where the operator e
B
Q ∗x0 y0 e−B
ch(k) −sh(k) ch(k) sh(k) 0 0 =I −2δ(x − x0 )δ(y − y0 ) 0 −sh(k) ch(k) sh(k) ch(k)
is a linear combination of the terms chb(k)(x, x0 )chb(k)(y0 , y)Q ∗x y d x d y, shb(k)(x, x0 )chb(k)(y0 , y)N x y d x d y, shb(k)(x, x0 )shb(k)(y0 , y)Q x y d x d y.
(39)
Second-Order Corrections for Weakly Interacting Bosons I
A similar calculation shows that e B Q x0 y0 e−B is a linear combination of chb(k)(x, x0 )chb(k)(y0 , y)Q x y d x d y, shb(k)(x, x0 )chb(k)(y0 , y)N x y d x d y, shb(k)(x, x0 )shb(k)(y0 , y)Q ∗x y d x d y.
293
(40)
Thus, e B Q ∗x0 y0 Q x0 y0 e−B is a linear combination of the nine possible terms obtained by combining the above. Now we list all terms in e B V e−B Ω. Terms in e B V e−B ending in Q x y are automatically discarded because they contribute nothing when applied to Ω. The remaining six terms are listed below. chb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x2 , x0 )chb(k)(y0 , y2 )
v(x0 − y0 )Q ∗x1 y1 N x2 y2 Ωd x1 dy1 d x2 dy2 d x0 dy0 ,
(41)
chb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x2 , x0 )shb(k)(y0 , y2 )
v(x0 − y0 )Q ∗x1 y1 Q ∗x2 y2 Ωd x1 dy1 d x2 dy2 d x0 dy0 ,
(42)
shb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x2 , x0 )chb(k)(y0 , y2 )
v(x0 − y0 )N x1 y1 N x2 y2 Ωd x1 dy1 d x2 dy2 d x0 dy0 ,
(43)
shb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x2 , x0 )shb(k)(y0 , y2 )
v(x0 − y0 )N x1 y1 Q ∗x2 y2 Ωd x1 dy1 d x2 dy2 d x0 dy0 ,
(44)
shb(k)(x1 , x0 )shb(k)(y0 , y1 )shb(k)(x2 , x0 )chb(k)(y0 , y2 )
v(x0 − y0 )Q x1 y1 N x2 y2 Ωd x1 dy1 d x2 dy2 d x0 dy0 ,
(45)
shb(k)(x1 , x0 )shb(k)(y0 , y1 )shb(k)(x2 , x0 )shb(k)(y0 , y2 ) v(x0 − y0 )Q x1 y1 Q ∗x2 y2 Ωd x1 dy1 d x2 dy2 d x0 dy0 .
(46)
To compute the above six terms,recall (24) through (27) as well as (1). In general, N x y Ω = 1/2δ(x − y)Ω, while f (x, y)Q ∗x y d xd yΩ = (0, 0, f (x, y), 0, . . .) up to symmetrization and normalization. The resulting contributions (neglecting symmetrization and normalization) follow. From (41): ψ(x1 , y1 ) = chb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x2 , x0 ) ×chb(k)(y0 , x2 )v(x0 − y0 )d x2 d x0 dy0 . From (42): ψ(x1 , y1 , x2 , y2 ) =
(47)
chb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x2 , x0 ) ×shb(k)(y0 , y2 )v(x0 − y0 )d x0 dy0 .
(48)
294
M. G. Grillakis, M. Machedon, D. Margetis
From (43): ψ=
shb(k)(x1 , x0 )chb(k)(y0 , x1 )shb(k)(x2 , x0 ) ×chb(k)(y0 , x2 )v(x0 − y0 )d x1 d x2 d x0 dy0 .
(49)
From (44), with the N and Q ∗ reversed, we get ψ(x2 , y2 ) =
shb(k)(x1 , x0 )chb(k)(y0 , x1 )shb(k)(x2 , x0 ) ×shb(k)(y0 , y2 )v(x0 − y0 )d x1 d x0 dy0 ,
(50)
as well as the contribution from [N , Q ∗ ], i.e. ψ(y1 , y2 ) =
shb(k)(x1 , x0 )chb(k)(y0 , y1 )shb(k)(x1 , x0 ) ×shb(k)(y0 , y2 )v(x0 − y0 )d x1 d x0 dy0 .
(51)
The contribution of (45) is zero, and, finally, the contribution of (46), using (24), consists of four numbers, which can be represented by the two formulas ψ=
shb(k)(x1 , x0 )shb(k)(y0 , x1 )shb(k)(x2 , x0 )
(52)
×shb(k)(y0 , x2 )v(x0 − y0 )d x1 d x2 d x0 dy0 and |shb(k)|2 (x1 , x0 )|shb(k)|2 (y0 , y1 )v(x0 − y0 )d x1 dy1 d x0 dy0 .
ψ=
(53)
We can now state the following proposition. Proposition 2. The state e B V e−B Ω has entries on the zeroth, second and fourth slot of a Fock space vector of the form given above. In addition, if ∂ i − ∆x − ∆ y sh(k) L 1 [0,T ]L 2 (d xd y) ≤ C1 , ∂t ∂ i − ∆x + ∆ y p L 1 [0,T ]L 2 (d xd y) ≤ C2 , ∂t and v(x) =
1 |x| ,
1 or v(x) = χ (x) |x| , then
T 0
e B V e−B Ω2F dt ≤ C,
where C only depends on C1 and C2 .
Second-Order Corrections for Weakly Interacting Bosons I
295
Proof. This follows by writing ch(k) = δ(x − y)+ p and applying Cauchy-Schwartz and local smoothing estimates as in the work of Sjölin [32], Vega [33]; see also Constantin and Saut [3]. In fact, we need the following slight generalization (see Lemma 2 below): If ∂ i − ∆x1 − ∆x2 ± ∆x3 · · · ± ∆xn f (t, x1 , . . . xn ) L 1 [0,T ]L 2 (dtd x) ≤ C, ∂t with initial conditions 0, then
f (t, x1 , x2 , . . .) L 2 [0,T ]L 2 (d xd y) ≤ C. |x1 − x2 |
(54)
We will check a typical term, (48). This amounts to proving the following three terms are in L 2 . 1. ψ pp (t, x1 , y1 , x2 , y2 ) = p(t, x1 , x0 ) p(t, y0 , y1 )shb(k)(t, x2 , x0 )shb(k)(t, y0 , y2 )v(x0 − y0 ) d x0 dy0 . We use Cauchy-Schwartz in x0 , y0 to get
T
|ψ pp |2 dt d x1 d x2 dy1 dy2 0 ≤ sup | p(t, x1 , x0 ) p(t, y0 , y1 )|2 d x1 d x0 dy1 dy0 t
×
T
|shb(k)(t, x2 , x0 )shb(k)(t, y0 , y2 )v(x0 − y0 )|2 dt d x2 d x0 dy2 dy0 ≤ C.
0
The first term is estimated by energy, and the second one is an application of (54) with f = shb(k)shb(k). Notice that, because of the absolute value, we can choose either sh(k) or sh(k) to insure that the Laplacians in x0 , y0 have the same signs. 2. ψ pδ (t, x1 , y1 , x2 , y2 ) = p(t, x1 , x0 )shb(k)(t, x2 , x0 )shb(k)(t, y1 , y2 )v(x0 − y1 ) d x0 . Here, we use Cauchy-Schwartz in x0 to estimate, in a similar fashion,
T
|ψ pδ |2 dt d x1 d x2 dy1 dy2 ≤ sup | p(t, x1 , x0 )|2 d x1 d x0
0
t
×
T 0
|shb(k)(t, x2 , x0 )shb(k)(t, y1 , y2 )v(x0 − y1 )|2 dt d x2 d x0 dy2 dy0 ≤ C.
296
M. G. Grillakis, M. Machedon, D. Margetis
3. ψδδ (x1 , y1 , x2 , y2 ) = shb(k)(t, x2 , x1 )shb(k)(t, y1 , y2 )v(x1 − y1 ), which is just a direct application of (54). All other terms are similar.
We have to sketch the proof of the local smoothing estimate that we used above. Lemma 2. If f : R3n+1 → C satisfies ∂ i − ∆x1 − ∆x2 ± ∆x3 · · · ± ∆xn f (t, x1 , . . . xn ) L 1 [0,T ]L 2 (d xd y) ≤ C ∂t with initial conditions f (0, · · · ) = 0, then
f (t, x1 , x2 , . . .) L 2 [0,T ]L 2 (d x) ≤ C. |x1 − x2 |
Proof. We follow the general outline of Sjölin, [32]. Using Duhamel’s principle, it suffices to assume that ∂ i − ∆x1 − ∆x2 ± ∆x3 · · · ± ∆xn f (t, x1 , · · · xn ) = 0 (55) ∂t with initial conditions f (0, · · · ) = f 0 ∈ L 2 . Furthermore, after the change of variables x1 → x1√+ x2 , x2 → x2√−x1 , it suffices to prove that 2
2
f (t, x1 , x2 , . . .) L 2 [0,T ]L 2 (d x) ≤ C, |x1 |
where f satisfies the same equation (55). Changing notation, denote x = (x2 , x3 , . . .) and let < ξ >2 be the relevant expression ±|ξ2 |2 ± |ξ3 |2 . . .. Write 2 2 f (t, x1 , x) = eit (|ξ1 | +<ξ > ) ei x1 ·ξ1 +i x·ξ f 0 (ξ1 , ξ ) dξ1 dξ. Thus, we obtain | f (t, x1 , x)|2 dtd x1 d x |x1 |2 i x ·(ξ −η )+i x·(ξ −η) 2 2 2 2 e 1 1 1 = eit (|ξ1 | −|η1 | +<ξ > −<η> ) |x1 |2 × f (ξ , ξ ) f 0 (η1 , η)dξ1 dξ dη1 dηdt d x d x1 0 1 1 = c δ(|ξ1 |2 − |η1 |2 ) f 0 (η1 , ξ )dξ1 dη1 dξ f 0 (ξ1 , ξ ) |ξ1 − η1 | ≤ | f 0 (ξ1 , ξ )|2 d x1 dξ, because one can easily check that sup δ(|ξ1 |2 − |η1 |2 ) ξ1
1 dη1 ≤ C. |ξ1 − η1 |
1 is bounded from L 2 (dη1 ) to L 2 (dξ1 ). Thus, the kernel δ(|ξ1 |2 − |η1 |2 ) |ξ1 −η 1|
Second-Order Corrections for Weakly Interacting Bosons I
297
8. Error Terms e B [ A, V ]e−B We proceed to check the operator e B [A, V ]e−B . The calculations of this section are similar to those of the preceding section with the notable exception of (60)–(63). Recall the calculations of Lemma 1 and write B −B e [A, V ]e = v(x − y) φ(y)e B ax∗ e−B e B ax a y e−B (56) +φ(y)e B ax∗ a ∗y e−B e B ax e−B d x d y. Now fix x0 . We start with the term (56). According to Theorem 2, we have B ∗ −B e ax0 e = sh(k)(x, x0 )ax + ch(k)(x, x0 )ax∗ d x, while e B ax0 a y0 e−B has been computed in (40). The relevant terms are and shb(k)(x, x0 )chb(k)(y0 , y)N x y d x d y shb(k)(x, x0 )shb(k)(y0 , y)Q ∗x y d x d y. Combining these two terms, there are three non-zero terms (which will act on Ω): 1.
v(x0 − y0 )φ(y0 )shb(k)(x1 , x0 )shb(k)(x2 , x0 ) ×shb(k)(y0 , y2 )ax1 Q ∗x2 y2 Ωd x1 d x2 dy2 d x0 dy0 .
(57)
This term contributes terms of the form ψ(t, y2 ) = v(x0 − y0 )φ(t, y0 )(shb(k)(t, x1 , x0 ))2 shb(k)(t, y0 , y2 )d x1 d x0 dy0 (58) as well as the term ψ(t, x2 ) =
v(x0 − y0 )φ(t, y0 )shb(k)(t, x1 , x0 )shb(k)(t, x2 , x0 ) ×shb(k)(t, y0 , x1 )d x1 d x0 dy0 ,
(59)
which we know how to estimate. The second contribution is: 2.
v(x0 − y0 )φ(y0 )chb(k)(x1 , x0 )shb(k)(x2 , x0 )Ω ×chb(k)(y0 , y2 )ax∗1 N x2 y2 d x1 d x2 dy2 d x0 dy0 .
(60)
ax∗1
with ax2 , we find that (60) contributes ψ(t, y2 ) = v(x0 − y0 )φ(t, y0 )chb(k)(t, x1 , x0 )shb(k)(t, x1 , x0 )
Commuting
×chb(k)(t, y0 , y2 )d x1 d x0 dy0 .
(61)
298
M. G. Grillakis, M. Machedon, D. Margetis
We expand chb(k)(t, x1 , x0 ) = δ(x1 − x0 ) + p(k)(t, x1 − x0 ). The contributions of p are similar to previous terms, but δ(x1 − x0 ) presents a new type of term, which will be addressed in Lemma 3. These contributions are ψδp (t, y2 ) = v(x1 − y0 )φ(t, y0 )shb(k)(t, x1 , x1 )p(k)(t, y0 , y2 ) d x1 dy0 (62) and
ψδδ (t, y2 ) = φ(t, y2 )
v(x1 − y2 )shb(k)(t, x1 , x1 )d x1 .
(63)
The last contribution of (56) is: 3.
v(x0 − y0 )φ(y0 )chb(k)(x1 , x0 )shb(k)(x2 , x0 )shb(k)(y0 , y2 )ax∗1 Q ∗x2 y2 Ω ×d x1 d x2 dy2 d x0 dy0 ∼ ψ(x1 , x2 , y2 ),
where
ψ(t, x1 , x2 , y2 ) = v(x0 − y0 )φ(t, y0 ) ×chb(k)(t, x1 , x0 )shb(k)(t, x2 , x0 )shb(k)(t, y0 , y2 )d x0 dy0 ,
modulo normalization and symmetrization. This term, as well as all the terms in (57), are similar to previous ones and are omitted. We can now state the following proposition: Proposition 3. The state e B [A, V ]e−B Ω has entries in the first and third slot of a Fock space vector of the form given above. In addition, if ∂ i − ∆x − ∆ y sh(k) L 1 [0,T ]L 2 (d xd y) ≤ C1 , ∂t ∂ i − ∆x + ∆ y p L 1 [0,T ]L 2 (d xd y) ≤ C1 ∂t and shb(k)(t, x, x) L 2 ([0,T ]L 2 (d x)) ≤ C3 , and v(x) =
χ (x) |x|
(64)
for χ a C0∞ cut-off function, then
T 0
e B [A, V ]e−B Ω2F ≤ C,
where C only depends on C1 , C2 , C3 . Proof. The proof is similar to that of Proposition 2, the only exception being the terms (62), (63). It is only for the purpose of handling these terms that the Coulomb potential has to be truncated, since the convolution of the Coulomb potential with the L 2 function shb(k)(x, x) does not make sense. If v is truncated to be in L 1 (d x), then we estimate the convolution in L 2 (d x), and take φ ∈ L ∞ (dydt).
Second-Order Corrections for Weakly Interacting Bosons I
299
To apply this proposition, we need the following lemma. 1 1
Lemma 3. Let u ∈ X 2 , 2 + . Then, u(t, x, x) L 2 (dt d x) ≤ Cu
1 1
X 2,2+
.
Proof. As it is well known, it suffices to prove the result for u satisfying ∂ i − ∆x − ∆ y u(t, x, y) = 0 ∂t 1
with initial conditions u(0, x, y) = u 0 (x, y) ∈ H 2 . This can be proved as a “Morawetz estimate”, see [14], or as a space-time estimate as in [19]. Following the second approach, the space-time Fourier transform of u (evaluated at 2ξ rather than ξ for neatness) is u 0 (ξ − η, ξ + η)dη
u (τ, 2ξ ) = c δ(τ − |ξ − η|2 − |ξ + η|2 ) δ(τ − |ξ − η|2 − |ξ + η|2 ) F(ξ − η, ξ + η)dη, =c (|ξ − η| + |ξ + η|)1/2 where F(ξ − η, ξ + η) = (|ξ − η| + |ξ + η|)1/2 u 0 (ξ − η, ξ + η). By Plancherel’s theorem, it suffices to show that u L 2 (dτ dξ ) ≤ CF L 2 (dξ dη) . This, in turn, follows from the pointwise estimate (Cauchy-Schwartz with measures) δ(τ − |ξ − η|2 − |ξ + η|2 ) dη | u (τ, 2ξ )|2 ≤ c |ξ − η| + |ξ + η| × δ(τ − |ξ − η|2 − |ξ + η|2 )|F(ξ − η, ξ + η)|2 dη, and the remark that
δ(τ − |ξ − η|2 − |ξ + η|2 ) dη ≤ C. |ξ − η| + |ξ + η|
9. The Trace
d(t, x, x)d x
This section addresses the control of traces involved in derivations. Recall that d(t, x, y) = ish(k)t + sh(k)g T + gsh(k) sh(k) − (ich(k)t + [g, ch(k)]) ch(k) −sh(k)mch(k) − ch(k)msh(k). Notice that if k1 (x, y) ∈ L 2 (d x d y) and k2 (x, y) ∈ L 2 (d x d y) then |k1 k2 |(x, x)d x ≤ |k1 (x, y)||k2 (y, x)|d y d x ≤ k1 L 2 k2 L 2 .
300
M. G. Grillakis, M. Machedon, D. Margetis
Recall from Theorem 5 that if v(x) = |x| or v(x) = χ (x) |x| then ish(k)t + sh(k)g T + gsh(k), ich(k)t + [g, ch(k)] and sh(k) are in L ∞ ([0, 1])L 2 (d xd y). This allows us to control all traces except the contribution of δ(x − y) to the second term. But, in fact, we have ich(k)t + [g, ch(k)] = ikt − ∆x k − ∆ y k k − k ikt − ∆x k − ∆ y k + · · · ,
which has bounded trace, uniformly in [0, 1]. Acknowledgements. The first two authors thank William Goldman and John Millson for discussions related to the Lie algebra of the symplectic group, and Sergiu Klainerman for the interest shown for this work. The third author is grateful to Tai Tsun Wu for useful discussions on the physics of the Boson system. The third author’s research was partially supported by the NSF-MRSEC grant DMR-0520471 at the University of Maryland, and by the Maryland NanoCenter.
References 1. Anderson, M.H., Ensher, J.R., Matthews, M.R., Wieman, C.E., Cornell, E.A.: Observation of BoseEinstein condensation in a dilute atomic vapor. Science 269, 198–201 (1995) 2. Bourgain, J.: Fourier transform restriction phenomena for certain lattice subsets and applications to nonlinear evolution equations I, II. Geom. Funct. Anal. 3, 107–156 (1993) and 202–262 (1993) 3. Constantin, P., Saut, S.: Local smoothing properties of dispersive equations. JAMS 1, 431–439 (1988) 4. Davis, K.B., Mewes, M.-O., Andrews, M.R., van Druten, N.J., Durfee, D.S., Kurn, D.M., Ketterle, W.: Bose-Einstein condensation in a gas of sodium atoms. Phys. Rev. Lett. 75, 3969–3973 (1995) 5. Dyson, F.J.: Ground-state energy of a hard sphere gas. Phys. Rev. 106, 20–26 (1957) 6. Elgart, A., Erd˝os, L., Schlein, B., Yau, H.-T.: Gross-Pitaevskii equation as the mean field limit of weakly coupled bosons. Arch. Rat. Mech. Anal. 179, 265–283 (2006) 7. Erd˝os, L., Yau, H.-T.: Derivation of the non-linear Schrödinger equation from a many-body Coulomb system. Adv. Theor. Math. Phys. 5, 1169–1205 (2001) 8. Erd˝os, L., Schlein, B., Yau, H.-T.: Derivation of the Gross-Pitaevskii hierarchy for the dynamics of Bose-Einstein condensate. Commun. Pure Appl. Math. 59, 1659–1741 (2006) 9. Erd˝os, L., Schlein, B., Yau, H.-T.: Derivation of the cubic non-linear Schrödinger equation from quantum dynamics of many-body systems. Invent. Math. 167, 515–614 (2007) 10. Erd˝os, L., Schlein, B., Yau, H.-T.: Rigorous derivation of the Gross-Pitaevskii equation. Phys. Rev. Lett. 98, 040404 (2007) 11. Erd˝os, L., Schlein, B., Yau, H.-T.: Derivation of the Gross-Pitaevskii equation for the dynamics of Bose-Einstein condensate. To appear in Annals Math 12. Folland, G.B.: Harmonic Analysis in Phase Space. Annals of Math Studies, Vol. 122 Princeton Univerity Press, Princeton, NJ, 1989 13. Ginibre, J., Velo, G.: The classical field limit of scattering theory for non-relativistic many-boson systems, I and II. Commun. Math. Phys. 66, 37–76 (1979) and 68, 45–68 (1979) 14. Grillakis, M.G., Margetis, D.: A priori estimates for many-body Hamiltonian evolution of interacting Boson system. J. Hyperb. Diff. Eqs. 5, 857–883 (2008) 15. Gross, E.P.: Structure of a quantized vortex in boson systems. Nuovo Cim. 20, 454–477 (1961) 16. Gross, E.P.: Hydrodynamics of a superfluid condensate. J. Math. Phys. 4, 195–207 (1963) 17. Hepp, K.: The classical limit for quantum mechanical correlation functions. Commun. Math. Phys. 35, 265–277 (1974) 18. Kenig, C., Ponce, G., Vega, L.: The Cauchy problem for the K-dV equation in Sobolev spaces with negative indices. Duke Math. J. 71, 1–21 (1994) 19. Klainerman, S., Machedon, M.: Space-time estimates for null forms and the local existence theorem. Commun. Pure Appl. Math. 46, 1221–1268 (1993) 20. Klainerman, S., Machedon, M.: Smoothing estimates for null forms and applications. Duke Math. J. 81, 99–103 (1995) 21. Lee, T.D., Huang, K., Yang, C.N.: Eigenvalues and eigenfunctions of a Bose system of hard spheres and its low-temperature properties. Phys. Rev. 106, 1135–1145 (1957) 22. Lieb, E.H., Seiringer, R.: Derivation of the Gross-Pitaevskii Equation for rotating Bose gases. Commun. Math. Phys. 264, 505–537 (2006)
Second-Order Corrections for Weakly Interacting Bosons I
301
23. Lieb, E.H., Seiringer, R., Solovej, J.P., Yngvanson, J.: The Mathematics of the Bose Gas and its Condensation. Birkhaüser Verlag, Basel, 2005 24. Lieb, E.H., Seiringer, R., Yngvanson, J.: Bosons in a trap: a rigorous derivation of the Gross-Pitaevskii energy functional. Phys. Rev. A 61, 043602 (2006) 25. Lieb, E.H., Seiringer, R., Yngvason, J.: A rigorous derivation of the Gross-Pitaevskii energy functional for a two-dimensional Bose gas. Commun. Math. Phys. 224, 17–31 (2001) 26. Margetis, D.: Studies in Classical Electromagnetic Radiation and Bose-Einstein Condensation. Ph.D. thesis, Harvard University, 1999 27. Margetis, D.: Solvable model for pair excitation in trapped Boson gas at zero temperature. J. Phys. A: Math. Theor. 41, 235004 (2008); Corrigendum. J. Phys. A: Math. Theor. 41, 459801 (2008) 28. Pitaevskii, L.P.: Vortex lines in an imperfect Bose gas. Soviet Phys. JETP 13, 451–454 (1961) 29. Pitaevskii, L.P., Stringari, S.: Bose-Einstein Condensation. Oxford: Oxford University Press, 2003 30. Riesz, F., Nagy, B.: Functional analysis. New York: Frederick Ungar Publishing, 1955 31. Rodnianski, I., Schlein, B.: Quantum fluctuations and rate of convergence towards mean field dynamics. Commun. Math. Phys. 291(2), 31–61 (2009) 32. Sjölin, P.: Regularity of solutions to the Schrödinger equation. Duke Math. J. 55, 699–715 (1987) 33. Vega, L.: Schrödinger equations: Pointwise convergence to the initial data. Proc. AMS 102, 874–878 (1988) 34. Wu, T.T.: Some nonequilibrium properties of a Bose system of hard spheres at extremely low temperatures. J. Math. Phys. 2, 105–123 (1961) 35. Wu, T.T.: Bose-Einstein condensation in an external potential at zero temperature: General theory. Phys. Rev. A 58, 1465–1474 (1998) Communicated by H.-T. Yau
Commun. Math. Phys. 294, 303–342 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0952-8
Communications in
Mathematical Physics
Entropic Bounds on Semiclassical Measures for Quantized One-Dimensional Maps Boris Gutkin Fachbereich Physik, Universität Duisburg-Essen, Lotharstrasse 1, 47048 Duisburg, Germany. E-mail:
[email protected] Received: 15 April 2008 / Accepted: 16 September 2009 Published online: 24 November 2009 – © Springer-Verlag 2009
Abstract: Quantum ergodicity asserts that almost all infinite sequences of eigenstates of quantized ergodic Hamiltonian systems are equidistributed in phase space. This, however, does not prohibit existence of exceptional sequences which might converge to different (non-Liouville) classical invariant measures. It has been recently shown by N. Anantharaman and S. Nonnenmacher in [20,21] (with H. Koch) that for Anosov geodesic flows the metric entropy of any semiclassical measure µ must satisfy a certain bound. This remarkable result seems to be optimal for manifolds of constant negative curvature, but not in the general case, where it might become even trivial if the (negative) curvature of the Riemannian manifold varies a lot. It has been conjectured by the same authors, that in fact, a stronger bound (valid in the general case) should hold. In the present work we consider such entropic bounds using the model of quantized piecewise linear one-dimensional maps. For a certain class of maps with non-uniform expansion rates we prove the Anantharaman-Nonnenmacher conjecture. Furthermore, for these maps we are able to construct some explicit sequences of eigenstates which saturate the bound. This demonstrates that the conjectured bound is actually optimal in that case. 1. Introduction The theory of quantum chaos deals with quantum systems whose classical limit is chaotic. It is assumed in general that chaotic dynamics induce certain characteristic patterns. For instance, the Random Matrix conjecture predicts that statistical distribution of high-lying eigenvalues in a chaotic system is the same as in certain ensembles of random matrices and depends only on symmetries of the system [1]. In the same spirit, it is believed that eigenstates of chaotic systems are delocalized over the entire available part of the phase space [2,3] which is totally different from the case of quasi-integrable systems, where eigenstates are known to concentrate near KAM tori [4]. The rigorous implementation of that idea is known as Quantum Ergodicity Theorem. It was first proven by
304
B. Gutkin
A. I. Schnirelman for Laplacians on surfaces of negative curvature [5] and later generalized [6,7] and extended to other systems e.g., ergodic billiards [8,9], quantized maps [10] and general Hamiltonians [11]. Very generally, the Quantum Ergodicity Theorem states that for a classically ergodic system “almost all” eigenstates ψk become uniformly distributed over the phase space in the semiclassical limit k → ∞. To give a more precise meaning of this statement it is convenient to use the notion of measure. Given some Hamiltonian system, let ψk , k = 1, 2, . . . be a sequence of normalized eigenstates of the corresponding quantum Hamiltonian operator. With any ψk one can associate the distribution µk , such that µk ( f ) = ψk Opk ( f ), ψk ,
f ∈ Cc∞ (X ),
where f (q, p) is a classical observable on the phase space X of the system and Opk ( f ) is the corresponding quantum observable. Here Opk ( · ) is a quantization procedure set at the scale k which in turn is fixed by the state ψk under consideration. As k → ∞, k → 0 and we are looking for possible semiclassical limits of the measure µk . Although the exact form of µk depends on the quantization procedure Opk ( · ) (e.g., Weyl, AntiWick quantization, etc.), a weak limit of µk as k → ∞, in the distributional topology does not depend on the choice of the quantization. Any such limit µ (called semiclassical measure) is a probability measure which is, in addition, invariant under the classical flow. The Quantum Ergodicity theorem asserts that for “almost all” sequences of eigenstates the corresponding semiclassical measure is actually the Liouville measure. Since the Quantum Ergodicity theorem does not exclude the possibility that exceptional sequences of eigenstates produce non-Liouville classically invariant measures, it makes sense to ask whether such measures might actually appear. In the context of Anosov geodesic flows on surfaces of negative curvature it was conjectured [12] that a typical system possesses Quantum Unique Ergodicity property, meaning that all sequences of eigenstates converge to the Liouville measure. However, so far there have been only a limited number of rigorous results supporting this conjecture. One of the most significant results in that direction was obtained by E. Lindenstrauss in [13], where he proved that all Hecke eigenstates of the Laplacian on some compact arithmetic surfaces are equidistributed. If (as widely believed) all the Laplacian eigenstates are non-degenerate, this result would amount to the proof of Quantum Unique Ergodicity for the arithmetic case. On the other hand, it is known that exceptional sequences actually do appear in some quantum systems. For instance, as has been recently shown by A. Hassell [14], Quantum Unique Ergodicity fails for “almost all” stadia billiards. Another example is provided by quantized hyperbolic symplectomorphisms of two-dimensional torus (known as “cat maps” in physics literature) [15,16]. Here the semiclassical measures induced by exceptional sequences of eigenstates can be, for instance, composed of two ergodic components: µ = aµL + (1 − a)µD ,
1 ≥ a ≥ 1/2,
(1)
where the first part µL is the Liouville measure equidistributed over the entire phase space and the second part µD is the Dirac peak concentrated on a single unstable periodic orbit. Similar sequences of eigenstates have been also constructed for the “Walsh quantization” of the baker’s map [17]. For quantized hyperbolic symplectomorphisms of higher-dimensional tori there exists a different type of semiclassical measures which are Lebesgue measures on some invariant co-isotropic subspaces of the torus [18]. As we know that non-Liouville semiclassical measures do appear (at least) in some systems, it would be of great interest to understand which kind might exist in the general
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
305
case. Quite recently, it has been proven by N. Anantharaman and S. Nonnenmacher [19– 21] (with H. Koch) that for the Laplacian on a compact Riemannian manifold X with Anosov geodesic flow on the unit cotangent bundle S ∗ X the metric (Kolmogorov-Sinai) entropy HKS (µ) of any semiclassical measure µ on S ∗ X must satisfy a certain bound. Particularly, in the two-dimensional case the following result holds [21]: 1 HKS (µ) ≥ | log J u (x)|dµ − λmax , (2) 2 S∗ X where J u (x) is the unstable Jacobian of the flow at the point x ∈ S ∗ X and λmax is the maximum expansion rate of the flow. If the maximum expansion rate is close to its average value, this remarkable bound gives valuable information on µ itself. In particular, for surfaces with a constant negative curvature it implies that the fraction of any semiclassical measure concentrated on periodic orbits cannot exceed “half” of the total measure of the phase space. On the other hand, if the expansion rate varies a lot, the above bound might provide little information, as the right hand side of (2) can even become negative. Thus, it is natural to expect that (2) is not an optimal result, and a stronger bound might exist in a general case. Such a bound has been conjectured in [17,19]. The conjecture states that for an Anosov canonical map (resp. Hamiltonian flow) on a compact symplectic manifold (resp. on a compact energy shell) any semiclassical measure must satisfy: 1 HKS (µ) ≥ (3) | log J u (x)|dµ. 2 Assuming that the bound is true, it provides certain restriction on the class of possible semiclassical measures in the general case. Suppose, for instance, that a semiclassical measure takes the form (1). (Note that for Anosov geodesic flows we don’t know whether such semiclassical measures actually exist.) Then the bound (3) implies that the LiouD ville part should be always present and its proportion satisfy a ≥ λavλ+λ , where λav is D the average Lyapunov exponent (with respect to the Liouville measure) and λD is the Lyapunov exponent for the periodic orbit where µD is localized. 2. Model and Statement of the Main Results The central purpose of this paper is to provide support for the conjectured bound (3) using the model of quantized one-dimensional piecewise linear maps. A procedure for quantization of one-dimensional linear maps was originally introduced in [22] in order to generate families of quantum graphs with some special properties. Being much simpler on the technical level, these models still exhibit characteristic properties of typical quantum chaotic Hamiltonian systems. Most importantly, it turns out that the quantum evolution here follows the classical evolution till the (Ehrenfest) time which grows logarithmically with the dimension N of the “quantum” Hilbert space.1 Note that in such a model N is always finite and N −1 plays the role of the Planck’s constant. As will be shown in the body of the paper, the construction of quantized one-dimensional maps is also closely related to the Walsh quantized baker’s maps in [17]. We consider a class of piecewise linear maps T : [0, 1) → [0, 1) =: I which preserve the Lebesgue measure on I . More specifically, let {I j , j = 1, . . . l} be a partition of 1 As we deal in the present paper with a discrete time evolution, the term “time” stands here and after for the number of iterations of either classical or quantum maps.
306
B. Gutkin
the unit interval I = ∪lj=1 I j into l subintervals I j = β− (I j ), β+ (I j ) , j = 1, . . . l, where β− (I j ) and β+ (I j ) denote the left and right endpoints of I j respectively. At each subinterval I j , T is then defined as a simple linear map T : I j → I : T (x) = j (x − β− (I j )),
j = 1/|I j |,
for x ∈ I j , j = 1, . . . l,
(4)
with |I j | standing for the length of I j . Furthermore, we will assume that the slopes j are positive integers satisfying l
−1 j = 1,
j > 1, j = 1, . . . l.
j=1
These conditions guarantee that the map is both Lebesgue-measure preserving and chaotic. Note that each such map T is uniquely determined by the ordered set of its slopes = {1 , . . . l }, so the notation T will be sometimes used to define the corresponding map. We will now briefly describe the procedure introduced by P. Pako´nski et al [22] for quantization of such maps. Let M = {E i , i = 1, . . . N } be a partition of I¯ = [0, 1] into N intervals E i = [(i − 1)/N , i/N ], i = 1, . . . N of equal lengths. For each interval E i we will denote by β+ (E i ) (β− (E i )) the right (resp. left) endpoint of E i and by N β (E ) the set of all endpoints of the partition M. Obviously both M β(M) = ∪i=1 ± i and β(M) are uniquely determined by the size N of the partition. In what follows we will consider an increasingly refined sequence of the above partitions Mk whose sizes Nk , k = 1, . . . ∞ grow exponentially. Conditions 1. Given a map T of the form (4) we impose the following conditions on the sequence of Mk : • Each partition Mk is a refinement of the previous one. That means for each k ≥ 1, Nk+1 /Nk is an integer number greater than one. • The set of the endpoints of the initial partition M1 must include all singular points of T , i.e., β(M1 ) ⊇ β(Ii ) for all i = 1, . . . l. For a sequence of partitions Mk , k = 1, . . . ∞ satisfying Conditions 1 consider the corresponding sequence of the transfer (Frobenius-Perron) operators given by Nk × Nk doubly stochastic matrices Bk , whose elements read as: −1 |E i ∩ T −1 E j | Ei if E i ∩ T −1 E j = ∅ Bk (i, j) = = (5) 0 otherwise, |E i | where Ei is the slope of T at the partition element E i . We will call a piecewise linear map T quantizable if there exists a sequence of partitions Mk , k = 1, . . . ∞ such that for each matrix Bk one can find a unitary matrix Uk of the same dimension satisfying Bk ( j, i) = |Uk (i, j)|2
(6)
for each matrix element ( j, i); j, i ∈ {1, . . . Nk }.2 For quantizable maps the matrices Uk are regarded as “quantizations” of Bk and play the role of quantum evolution operators acting on the Nk -dimensional Hilbert space Hk C Nk equipped with the Nk ¯ ψ(i)ϕ(i) for vectors ψ = (ψ(1), . . . ψ(Nk )), standard scalar product: ψ, ϕ = i=1 ϕ = (ϕ(1), . . . ϕ(Nk )). As an example, consider the following linear map (see Fig. 1): 2 Note that our definition for U matrix corresponds to the adjoint of the corresponding quantum evolution in [22,23].
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
1
1
307
1
1
1/2
0
1
0
1/2
1
0
1/2
1
0
1
Fig. 1. Linear maps with uniform (left) and non-uniform slopes (middle) which allow “tensorial” quantization. On the right is shown the asymmetric Baker map corresponding to the linear map in the middle
T (x) = 2x
mod 1,
x ∈ I.
(7)
Here for the sequence of partitions Mk of the unit interval into Nk = 2k equal pieces, the matrix elements Bk (i, j) of the classical transfer operators take the values 1/2 if j = 2i, j = 2i − 1, j + Nk = 2i, j + Nk = 2i − 1 and 0 otherwise: ⎞ ⎛ 1 1 0 0 1 1 1 1 ⎜0 0 1 1⎟ ··· . , B4 = ⎝ B2 = ⎠, 2 1 1 2 1 1 0 0 0 0 1 1 It is worth mentioning that the structure of Bk , actually, resembles the structure of the map T (rotated clockwise by π/2). It is also easy to see that the map (7) is quantizable. By a permutation of rows Bk can be brought into the block diagonal form such that every block is 2×2 matrix B2 whose elements are all 1/2. Thus the question of the quantization of T reduces to finding a unitary 2 × 2 matrix U satisfying |U(l, m)|2 = 1/2 for all its elements. An appropriate choice for U is provided, for instance, by the two-dimensional matrix F2 , where F p is the p-dimensional discrete Fourier transform: 2πi 1 (l − 1)(m − 1) , l, m = 1, . . . p. (8) F p (l, m) = √ exp − p p This construction can be straightforwardly generalized to all other maps with uniform slopes. For a general piecewise linear map it becomes, however, a non-trivial problem to determine whether the corresponding doubly stochastic matrices Bk allow the representation (6) in terms of unitary matrices Uk (see [22,24] and references there). Nevertheless, we show in the Appendix to the paper that the class of quantizable piecewise linear maps is quite wide and contains many maps with non-uniform slopes. Notice that the above quantization of one-dimensional piecewise linear maps is just a formal procedure for generation of unitary matrices Uk . To turn it to a “meaningful” quantization one needs, in addition, to make a connection between classical observables on the unit interval and the corresponding quantum observables on the Hilbert space Hk . Such a quantization procedure has been introduced in [23]. With a classical observable f ∈ L 2 [0, 1] one associates the sequence of the quantum observables Opk ( f ), defined by the diagonal matrices of the dimension Nk whose components Opk ( f ) j, j equal the average value of f at j’s element of the partition Mk . The key observation making the above quantization interesting is the existence of the semiclassical correspondence (Egorov property) between evolutions of classical and quantum observables. Precisely, for a Lipschitz continuous observable f (x) one has [23]:
308
B. Gutkin
Uk∗ Opk ( f )Uk − Opk ( f ◦ T ) = O
1 Nk
.
(9)
Recall that the size Nk of the partition plays here the role of the inverse Planck constant −1 k and the semiclassical limit corresponds to k → ∞. Equipped with the above quantization procedure we can define now the sequence of the semiclassical measures associated with the eigenstates of Uk . For any normalized eigenstate ψk = (ψ(1), . . . ψ(Nk )) ∈ Hk , Uk ψk = eiθk ψk , let µk be the associated probability measure, whose density is given by the piecewise constant function ρk (x) = Nk |ψ(i)|2 , for x ∈ E i , i = 1, . . . Nk . By this definition, the quantum average of any observable f ∈ L 2 [0, 1] can be written as: ψk Opk ( f )ψk = f (x) ρk (x)d x ≡ f (x) dµk . (10) I
I
We will be concerned with the possible weak limits of µk as k → ∞ and call any such limiting measure µ a semiclassical measure. Speaking informally µ characterizes the possible sets of the localization on I for the eigenstates of quantized maps. (Alternatively (see [23]), one can think that µ determines the limiting eigenstates distribution on the sequence of quantum graphs provided by Uk ’s.) An immediate consequence of the Egorov property is that any semiclassical measure µ must be invariant under the map T . Indeed, since ψk is an eigenstate of Uk : 1 , (11) f (x) dµk (x) = ψk Uk∗ Opk ( f )Uk ψk = f (T (x)) dµk (x) + O N k I I and the invariance of µ follows immediately after taking the limit k → ∞. As there exist many classical measures preserved by T , the invariance alone does not determine all possible outcomes for the semiclassical measures. Similarly to the case of Hamiltonian systems, using the Egorov property one can show by standard methods (see e.g., [26]) that almost any sequence of the eigenstates gives rise to the Lebesgue measure in the semiclassical limit (this was proved in [23] by a somewhat different method). Theorem 1 (Quantum Ergodicity [23, Thm. 2]). Let T be a quantizable piecewise linear map of the form (4) and let Uk , k = 1, . . . ∞ be a sequence of its quantizations (i) with eigenstates ψk , i = 1, . . . Nk . Then for each k there exists a subsequence of Nk (i N )
eigenstates: k := {ψk(i1 ) , . . . ψk k } such that limk→∞ Nk /Nk and for any sequence of eigenstates ψk j ∈ k j , j = 1, . . . ∞ and any Lipschitz continuous function f one has: lim ψk j Opk j ( f )ψk j = f (x) d x. (12) j→∞
I
In the present paper we go beyond the Quantum Ergodicity and ask about the possible exceptional semiclassical measures. In what follows we will restrict our treatment to a special subclass of the piecewise linear maps (4): Definition 1. Let I = ∪li=1 I j be a partition of I into a number of subintervals. We define T p as a piecewise linear map (4) whose precise form at each subinterval is given by: T p (x) = j x
mod 1,
for x ∈ I j , j = 1, 2 . . . l.
(13)
Furthermore, the slopes j = p n j , j = 1, 2 . . . l are given by positive powers n j ∈ N of an integer p > 1, satisfying condition: lj=1 p −n j = 1.
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
309
Using the method developed in [17,20,21] we can easily prove for these maps an analog of the bound (2): Theorem 2 Let Uk , k = 1, . . . ∞ be a sequence of quantizations of T p and let µ be a semiclassical measure induced by a subsequence ψk , k = 1, . . . ∞ of the eigenstates of Uk ’s. Then, the following bound holds for the metric entropy of µ:
1 1 log (x) dµ(x) − log max = µ(I j ) log j − log max , 2 2 l
HKS (T p , µ) ≥ I
j=1
(14) where max := max1≤ j≤l j and µ(I j ) are the measures of the intervals I j . As the right hand side of (14) can in principle be negative, one might suspect that this bound is not optimal for the maps T p with non-uniform slopes. The main result of the present paper is a proof of a stronger bound on the metric entropy of semiclassical measures. Namely, in the body of the paper we show that the maps T p allow a special type of “tensorial” quantizations. For the maps T p quantized in that way we prove the precise analog of the Anantharaman-Nonnenmacher conjecture (3). Theorem 3 Let T p be a map as in Definition 1 and let Uk , k = 1, . . . ∞ be a sequence of “tensorial” quantization of T p . Then for any sequence of eigenstates ψk of Uk , k = 1, . . . ∞ the corresponding semiclassical measure µ satisfies: 1 µ(I j ) log j . 2 l
HKS (T p , µ) ≥
(15)
j=1
Furthermore, for certain “tensorial” quantizations of T p it becomes possible to construct explicit subsequences of eigenstates of Uk . Using these eigenstates we obtain a set of semiclassical measures which can be subsequently analyzed to test (15). It turns out that some of these semiclassical measures, in fact, saturate the bound implying that the result is sharp. Remark 1. In what follows we will distinguish the class of “tensorial” quantizations (which will be introduced in the next section) from general quantizations of one dimensional maps which only need to satisfy the condition (6). Some of the results in the paper are specific for “tensorial” quantizations and it will be always explicitly mentioned that we deal with such class of quantization whenever it is relevant. Let us also notice that the results of the present paper can in fact be proven for a more general class of one dimensional linear maps. For instance, the bound (14) can be proven for all quantizable one dimensional maps, while the main result (15) can be straightforwardly extended to all maps (with general quantizations) whose slopes are given by integer powers of an integer number [33]. However, the restriction of the discussion to the class of maps T p provided by Definition 1 and “tensorial” quantizations, allows to represent the main ideas and results in a particularly transparent way, reducing the technical details to the minimum.
310
B. Gutkin
The paper is organized as follows. In Sect. 3 we introduce a special class of “tensorial” quantizations for the maps T p and demonstrate their connection with the Walsh quantized Baker maps. In Sect. 4 we review the construction in [23] for quantization of observables and prove the Egorov property up to the Ehrenfest time. In Sect. 5 we connect metric entropy of semiclassical measures with a certain type of quantum entropy functions. Based on the method of [20] we then prove Theorem 2 in Sect. 6 using the Entropic Uncertainty Principle. In the next two sections we deal with “tensorial” quantizations of T p . Section 7 is devoted to the proof of Theorem 3. In Sect. 8 we explicitly construct a certain class of semiclassical measures for the maps T p and test the bound (15). The concluding remarks are presented in Sect. 9. 3. “Tensorial” Quantizations of One-Dimensional Piecewise Linear Maps We will consider now in detail the quantization procedure for the piecewise linear maps T p given by Definition 1. As we show below, these maps allow a special type of “tensorial” quantizations mimicking the action of the corresponding classical shift map. This quantization procedure is basically restricted to the class of maps T p . For the sake of completeness, we also consider in Appendix A a different approach suitable for quantization of a more general class of maps (4). Maps with uniform slopes. We will first consider the piecewise linear maps T¯ p := T{ p,... p} with the uniform slope j ≡ p ∈ N i.e, the maps: T¯ p (x) = px mod 1,
x ∈ I.
(16)
(Here and after we will use the bar symbol to distinguish the above uniform maps from non-uniform ones.) For any point x ∈ I it will be convenient to use a p-base numeral system: x = 0.x1 x2 x3 . . ., xi ∈ {0, . . . , p − 1} to represent x. Obviously, each point is then encoded by an infinite sequence (not necessarily unique) of symbols x1 , x2 , x3 . . .. With such representation for the points of I the action of T¯ p becomes equivalent to the shift map: T¯ p : x1 x2 x3 x4 · · · → x2 x3 x4 x5 . . . .
(17)
In the following we will use the symbol x = x1 x2 x3 . . . xm for both finite and infinite sequences with the notation |x| := m reserved for the length of the sequence. So for x with |x| = ∞ the symbol x will stand for the corresponding point x = 0.x in the interval I . For a sequence x, with finite |x| = m we will use notation x to denote the corresponding cylinder set. A point x belongs to x if the first m digits of x after the point coincide with x1 , x2 , . . . xm . For any map T¯ p , there exists a sequence of natural Markov partitions Mk into Nk = p k cylinder sets of the length k: {E x = x, |x| = k}. The corresponding transfer operator is then given by the matrix Bk , whose elements: −1 , i = 1, . . . k − 1 if xi = xi+1 p (18) Bk (x, x ) = 0 otherwise, give the transition probabilities for reaching E x , x = x1 x2 . . . xk starting from E x , x = x1 x2 . . . xk after one step of the classical evolution. These matrices can be now
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
311
“quantized” as follows. Let H C p , be a vector space of the dimension p with a scalar product ·, · and an orthonormal basis {| j, j ∈ {0 . . . p − 1}}. Take U to be a unitary transformation on H such that in the basis above: |Ui, j |2 = 1/ p,
Ui, j := i|U| j.
(19)
(One possible choice for the matrix Ui, j is provided by the p-dimensional discrete Fourier transform.) With each partition Mk we now associate Nk -dimensional Hilbert space: Hk = H ⊗ H ⊗ · · · ⊗ H . k
Using the orthonormal basis in Hk given by the vectors: |x := |x1 ⊗ |x2 ⊗ · · · ⊗ |xk ,
x = x1 . . . xk , xi ∈ {0 . . . p − 1},
one defines the unitary transformation U¯ k as: U¯ k |x = |x2 ⊗ |x3 ⊗ · · · ⊗ |xk ⊗ U|x1 ,
(20)
and the corresponding adjoint: U¯ k∗ |x = U∗ |xk ⊗ |x1 ⊗ |x2 ⊗ · · · ⊗ |xk−1 .
(21)
The action of U¯ k basically mimics the action of the shift map. From this and property (19) of the U matrix it follows immediately that U¯ k satisfies (6) and therefore, indeed, a quantization of Bk . Note that if U∗ is given by the discrete Fourier transform F p , the matrix U¯ k coincides with the evolution operator of the Walsh-quantized Baker map in [17]. In that case U¯ k4 = 1 and the spectrum of U¯ k is highly degenerate. Note also that the U matrix in the definition (20) of U¯ k should not necessarily be constant. More general construction is obtained if one takes U in the form U(x) = exp(iφ(x))U (x2 , x3 . . . xk ), where φ(x) is a real function of x and U (x2 , x3 . . . xk ) is a unitary matrix depending on x2 , x3 . . . xk and satisfying (19). Maps with non-uniform slopes. Let us consider now maps T p given by Definition 1 such that not all n j are equal. For a given p we will use exactly the same representation x = x1 x2 x3 x4 . . ., xi ∈ {0, 1 . . . p − 1} for the point x = 0.x, and the same set of the partitions Mk , k ≥ n max := max j∈{1,...l} n j as for the map T¯ p with the uniform expansion rate. The action of T p in that representation is given by the shift map, where the size of the shift depends on the point itself: T p : x1 x2 x3 x4 · · · → xn j +1 xn j +2 xn j +3 . . . ,
if 0.x ∈ I j , j = 1, 2 . . . l.
(22)
The corresponding classical evolution matrix for the partition Mk is then given by −n p i if x ⊆ Ii and xj = xn i + j , j = 1, . . . k − n i (23) Bk (x, x ) = 0 otherwise.
312
B. Gutkin
Remark 2. It should be emphasized that for the uniform map T¯ p the above p-base encoding (for the points on the interval I ) provides a standard symbolic representation of the dynamical system defined with respect to the initial Markov partition lj=1 I j . That means the sequence x = x1 x2 x3 . . .; xi ∈ {0, . . . p − 1} describes the future history of the point 0.x regarding that partition under the action of T¯ p . On the other hand, for maps T p with non-uniform slopes the above representation possesses no such a dynamical significance. In particular, T p does not act as a simple shift map on the sequences x. Later we will also use the dynamical symbolic representation for the points on I which is defined with respect to the initial Markov partition lj=1 I j . In that case the points on I are represented by sequences of symbols ε = ε0 ε1 ε2 . . .; εi ∈ {1, . . . l} which encode their future histories under the action of T p . In that representation T p acts as the standard shift map on ε. To distinguish between two different representation systems we will always use x and ε letters to denote the corresponding symbols. Note that with this notation ε0 ε1 . . . εn and x1 x2 . . . xn refer, in general, to different types of sets. While the first set is a cylinder defined with respect to the action of the map T p , the second set is a cylinder for the corresponding uniform map T¯ p (and its own Markov partition). Now it is not difficult to “quantize” the matrices (23) using exactly the same Hilbert space as in the uniform case. For each state |x, x = x1 . . . xk such that x ⊆ I j , define the action of Uk on |x by: Uk |x = |xn j +1 ⊗ · · · ⊗ |xk ⊗ Un j |xn j ⊗ Un j −1 |xn j −1 ⊗ · · · ⊗ U1 |x1 , (24) where all the matrices Ui , i = 1, . . . n j satisfy (19). Note that the unitarity of Uk requires the flip in the order of the shifted p-bits on the right hand side of (24). It follows straightforwardly from the definition that Uk is indeed unitary and fulfills (6), thereby it provides a “quantization” of Bk . As for the maps with uniform slopes, the matrices Ui , i = 1, . . . n max should not necessarily be constant and can, for instance, depend on xn max +1 , xn max +2 . . . xk , as well. Example. As an example of the above quantization construction consider the map T2 = T{2,4,4} (see Fig. 1) which will be a principle model for us in what follows. Explicitly, for x = x1 x2 x3 . . ., xi ∈ {0, 1} the action of T2 on x = 0.x is given by 2x mod 1 if 0 ≤ x < 1/2 T2 (x) = (25) 4x mod 1 if 1/2 ≤ x < 1. For the vector space Hk = H ⊗ · · · ⊗ H (k times), H C2 the corresponding quantum evolution acts on |x ∈ Hk as: |x2 ⊗ |x3 ⊗ · · · ⊗ |xk ⊗ U1 |x1 if x1 = 0 (26) Uk |x = |x3 ⊗ · · · ⊗ U2 |x2 ⊗ U1 |x1 if x1 = 1, where U1,2 are constant matrices whose elements satisfy: |U1,2 (i, j)| =
√
1/2.
Connection with Walsh quantized Baker maps. Although one-dimensional maps are not Hamiltonian systems, there exists a close connection between their quantizations and Walsh quantized Baker maps. This is clear for the maps with uniform expansion rates, since in that case Uk defined by Eq. (20) with U∗ = F p , gives precisely the evolution operator of the Walsh quantized standard Baker map. Let us show now that in a similar way some natural quantizations of non-uniformly expanding maps provide
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
313
the time evolution operator for à la Walsh quantized asymmetric Baker maps. For the sake of concreteness consider the map (25). The corresponding asymmetric Baker map T2 : (q, p) → (q, p), q, p ∈ [0, 1) (see Fig. 1) is then defined by: if q ∈ [0, 1/2) (T2 (q), p + 21 ) (27) T2 (q, p) = (T2 (q), p + 41 4q) if q ∈ [1/2, 1). Using the binary representation q = 0.x1 x2 . . ., p = 0.x−1 x−2 . . ., xi ∈ {0, 1} for the space and momentum coordinates of the phase space one can assign to the point x = (q, p) the bi-sequence x = . . . x−2 x−1 · x1 x2 . . .. In this representation the action of the map T2 on the point x takes a simple form . . . x−2 x−1 x1 · x2 x3 . . . if x1 = 0 T2 (. . . x−2 x−1 · x1 x2 . . . ) = (28) . . . x−1 x2 x1 · x3 x4 . . . if x1 = 1. To quantize such a map one can apply the same procedure as in [17]. With each cylinder (rectangle) set x− · x+ m , x− = x−(k−m) . . . x−2 x−1 , x+ = x1 x2 . . . xm , m ∈ {0, . . . k} in the phase space one associates the following coherent state in Hk : |x− · x+ m := |x1 ⊗ |x2 ⊗ · · · ⊗ |xm ⊗ F2∗ |x−(k−m) ⊗ · · · ⊗ F2∗ |x−2 ⊗ F2∗ |x−1 . Note that each coherent state |x− · x+ m is strictly localized in the rectangle x− · x+ m . One then uses these states in order to quantize observables by means of the Anti-Wick quantization procedure (see [17] for details). It is a simple observation that the operator Uk defined by Eq. (26) with U1 = U2 = F2∗ acts on the coherent states in accordance with the action of the classical map (28): Uk |x− · x+ m = |T2 (x− · x+ )m− ,
Uk∗ |x− · x+ m = |T2−1 (x− · x+ )m+ , m ∈ {1, . . . k − 2},
where equals either 1 or 2. Thereby, the operator Uk plays the role of the quantum evolution operator for à la Walsh quantized map T2 . In an analogous way the quantization (24) of the map T p with Ui = F p∗ , i = 1, . . . n max supplies the quantum evolution operator for à la Walsh quantized asymmetric Baker map T p whose expansion rates are given by the powers of p. As both quantizations provide the same set of eigenfunctions, one can utilize the results of the present paper in order to deduce the bound (3) for semiclassical measures of à la Walsh quantized T p . 4. Quantization of Observables We recall now the procedure for the quantization of observables introduced in [23]. Let Mk be the partition of the unit interval I¯ into Nk intervals {E i = [(i − 1)Nk−1 , i Nk−1 ], i = 1, . . . Nk } and let Hk C Nk denote the corresponding Hilbert space. For each function f ∈ L 2 ( I¯) the corresponding quantum observable Op( f ) is given by the matrix, whose elements are Op( f )i, j := δi, j Nk f (x) d x, i, j = 1, . . . Nk . (29) Ei
Let Ic be the circle corresponding to I = [0, 1), where the endpoints 0 and 1 are identified. It will be assumed that Ic is equipped with the standard Euclidean metric
314
B. Gutkin
coming from R. In particular, the distance d(x, y) between two points x, y ∈ Ic is defined by d(x, y) := min{|x − y|, |x − y − 1|}. Below we will deal with a class of observables f ∈ Lip(Ic ) which are Lipschitz continuous on Ic . Recall that the space Lip(Ic ) is equipped with the Lipschitz norm: f Lip = sup | f (x)| + sup x∈I
x = y∈I
| f (x) − f (y)| d(x, y)
(30)
and f ∈ Lip(Ic ) iff f Lip is finite. The definition (29) is strongly motivated by the existence of the correspondence between classical and quantum evolutions of observables (Egorov property). In the context of quantized one-dimensional maps the Egorov property was proved in [23, Thm. 3] for Lipschitz continuous observables undergoing one step evolution. The following theorem is a straightforward extension of that result to larger times. Theorem 4 Let U = Uk be a quantum evolution operator (satisfying (6)) for a quantizable one-dimensional map T of the form (4) and let f be a Lipschitz continuous function on Ic , then U −n Op( f )U n − Op( f ◦ T n ) ≤ D(T ) f Lip
nmax , Nk
(31)
where D(T ) is a constant independent of n and Nk . Proof. For n = 1 the following bound was proved in [23]: U −1 Op( f )U − Op( f ◦ T ) ≤ f Lip
D(T ) . Nk
(32)
From this one immediately gets for n iterations: U −n Op( f )U n − Op( f ◦ T n ) n U −i Op( f ◦ T n−i )U i − U 1−i Op( f ◦ T n−i+1 )U i−1 ≤ i=1 n D(T ) n ≤ f ◦ T i−1 Lip ≤ D(T ) f Lip max , Nk Nk
(33)
i=1
where we used the fact that f ◦ T i ∈ Lip(Ic ) and f ◦ T i Lip ≤ imax f Lip . The inequality (31) implies that the quantum evolution follows the classical one up to the time: n E := log Nk / log max which plays the role of the Ehrenfest time for the model. (Here and after y denotes the largest integer smaller than y.) It is worth noticing that for a certain class of observables the Egorov property turns out to be exact. Let x1 , x2 be two points on the lattice β(Mk ), then with an interval X = [x1 , x2 ] ⊂ I we can associate the projection operator PX := Op(χ X ), where χ X is the characteristic function on the set X . For such operators one has the following result.
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
315
Proposition 1. Let X ⊂ I be an interval (or union of intervals) such that all the endpoints β(X ) and β(T −1 X ) belong to β(Mk ), then U −1 PX U = PT −1 X . Proof. Written in matrix form the left side of (34) is given by (U j,l )∗ U j,m , (U ∗ PX U )l,m =
(34)
(35)
{ j|E j ⊆X }
where E j denotes j’s element of the partition Mk . Observe that when E j ⊆ X , the elements (U j,l )∗ = 0, (U j,m = 0) only if T (El ) ⊆ X (resp. T (E m ) ⊆ X ). On the other hand, if the last condition holds, one can extend the summation in (35) to all values of j. By the unitarity of U it gives the right side of (34). For the class of maps T p the proposition above implies the exact correspondence between classical and quantum evolutions of some projection operators up to the times of order nE . Corollary 1. Let T p , be a map of the form (13). Denote U a general quantization of T p (satisfying (6)) acting on the vector space Hk of the dimension Nk = p k . For a cylinder x of the length |x| ≤ k the evolution of the corresponding projection operator Px is given by U −n Px U n = PT p−n x
for all n + |x|/n max ≤ n E ,
(36)
where max = p n max is the maximum slope of T p . 5. Metric Entropy of Semiclassical Measures Let Uk : Hk → Hk , k = 1, · · · ∞ be a sequence of unitary quantizations of a quantizable map T of the form (4). For a given sequence of eigenstates: ψk ∈ Hk , Uk ψk = eiθk ψk , the corresponding measures µk , k = 1, · · · ∞ are defined by Eq. (10) through the Riesz representation theorem. We will be concerned with the possible outcome for semiclassical T -invariant measures µ = limk→∞ µk . Following the approach of [17,19,20] we will consider the metric entropy HKS (T, µ) of µ. Below we recall some basic properties of classical entropies and connect them to a certain type of quantum entropies. s Let π = i=1 Ii be a certain partition of I into s intervals. Given a measure µ on I the entropy function of µ with respect to the partition π is defined by h π (µ) := −
s
µ(Ii ) log(µ(Ii )).
i=1
More generally, one can consider the pressure function: pπ,v (µ) := −
s
µ(Ii ) log(vi2 µ(Ii )),
i=1
where the weights v = {vi |i = 1, . . . s} are given by a set of real numbers fixed for a given partition. Obviously, if all vi equal one, then pπ,v is just the entropy defined
316
B. Gutkin
s above. An important feature of h π (µ) is its subadditivity property. If π = i=1 Ii and s τ = i=1 Ji are two partitions, then for the partition π ∨ τ consisting of the elements Ii ∩ J j and a measure µ one has: h π ∨τ (µ) ≤ h π (µ) + h τ (µ).
(37)
Now consider dynamically generated refinements of π . Define ε = ε0 ε1 . . . εn−1 , and let it be a sequence of the elements εi ∈ {1, . . . s} of the length |ε| = n. For any n ≥ 1 set partition: π (n) = |ε|=n ε of I is the collection of the sets: ε := T −(n−1) Iεn−1 ∩ T −(n−2) Iεn−2 ∩ . . . Iε0 . Each cylinder ε has a simple meaning as the set of the points with the same “ε-future” up to n iteration. One is interested in the entropies of T -invariant measures µ with respect to the partitions π (n) : h n (µ) := h π (n) (µ) = − µ(ε) log(µ(ε)). |ε|=n
If µ is T -invariant, it follows (see e.g., [25]) by the subadditivity (37) that: h n+m (µ) ≤ h n (µ) + h m (µ).
(38)
For the entropy function this implies the existence of the limit: 1 h n (µ). n→∞ n
Hπ (T, µ) = lim
(39)
The metric (Kolmogorov-Sinai) entropy is then defined as the supremum over all finite measurable initial partitions π : HKS (T, µ) = sup Hπ (T, µ). π
In the quantum mechanical framework one needs to define a quantum entropy (preslimit. Note that a measure of sure) reproducing h n (µ) (resp. pn,v (µ)) in the semiclassical each set Ii can be written as the average µ(Ii ) = χIi (x) dµ over the classical observable χIi (x) which is the characteristic function of the set Ii . The quantum observable corresponding to χIi is then simply the projection operatorPi := PIi = Op(χIi ) on the set Ii . Now we need to “quantize” the refined partitions |ε|=n ε. The most straightforward approach would be considering quantization of the observables χε . A different scheme was suggested in [20]. Instead of taking classically refined observables χε and then quantizing them, one considers a natural quantum dynamical refinement of the initial quantum partition. We will say that a sequence of operators πˆ = {πˆ i , i = 1 . . . s} defines a quantum partition of Hk if they resolve the unity operator: 1Hk =
s
πˆ i∗ πˆ i .
i=1
For a quantum partition πˆ the entropy (resp. pressure) of a state ψ ∈ Hk is given by hˆ πˆ (ψ) := −
s i=1
πˆ i ψ2 log(πˆ i ψ2 ),
pˆ π,v ˆ (ψ) := −
s i=1
πˆ i ψ2 log(πˆ i ψ2 vi2 ).
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
317
Now with each set ε of π (n) one associates the operator defined by: Pεi ( p) = U − p Pεi U p .
Pε := Pεn−1 (n − 1) . . . Pε1 (1)Pε0 (0),
(40)
As follows immediately from the definition of Pε , the sets of the operators πˆ (n) = {Pε , |ε| = n}, πˆ ∗(n) = {Pε∗ , |ε| = n} define quantum partitions of 1Hk . Note that Pε∗ and Pε differ only by the order of the components Pεi (i) and both πˆ (n) , πˆ ∗(n) correspond to the same classical partition π (n) . For an eigenfunction ψk ∈ Hk of the operator Uk let hˆ πˆ (n) (ψk ), hˆ πˆ ∗(n) (ψk ) be the corresponding entropies. After introducing the weight functions: µˆ ∗k (ε) := Pε∗ ψk 2
µˆ k (ε) := Pε ψk 2 ,
(41)
for the elements ε of the corresponding classical partition π (n) , the quantum entropies of ψk can be written with a slight abuse of notation (in principle, µˆ k , µˆ ∗k are not measures but merely positive weight functions defined only on the elements of the partitions) as the classical entropy function of µˆ k , µˆ ∗k : hˆ πˆ (n) (ψk ) = h n (µˆ k ),
h n (µˆ k ) = −
hˆ πˆ ∗(n) (ψk ) = h n (µˆ ∗k ),
h n (µˆ ∗k ) = −
µˆ k (ε) log µˆ k (ε),
|ε|=n
|ε|=n
µˆ ∗k (ε) log µˆ ∗k (ε).
(42)
Note that the weight function µˆ k satisfies an important compatibility condition: µˆ k (ε0 . . . εn−1 ) =
µˆ k (ε0 . . . εn ).
εn ∈{1...s}
As Pεi (i), i ≤ n = |ε| approximately commute with each other for a finite n by virtue of the Egorov property, the same property also holds (up to semiclassically small errors) for µˆ ∗k . Furthermore, for a finite n the Egorov property guarantees that both µˆ k (ε) and µˆ ∗k (ε) equal (up to semiclassically small errors) the measure µk (ε) induced by the eigenstate ψk . Hence in the semiclassical limit: lim h n (µˆ k ) = lim h n (µˆ ∗k ) = h n (µ),
k→∞
k→∞
(43)
where µ = limk→∞ µk is the corresponding semiclassical measure. To extract from h n (µ) the metric entropy HKS (T, µ) of the measure µ it is necessary to apply the classical limit (39). In complete analogy, the quantum pressures of ψk : pˆ πˆ (n) ,v (ψk ) = pn,v (µˆ k ),
pn,v (µˆ k ) = −
pˆ πˆ ∗(n) ,v (ψk ) = pn,v (µˆ ∗k ),
pn,v (µˆ ∗k ) = −
|ε|=n
|ε|=n
µˆ k (ε) log µˆ k (ε)vε2 , µˆ ∗k (ε) log µˆ ∗k (ε)vε2 (44)
converge in the limit k → ∞ to the classical pressure pn,v (µ) of µ.
318
B. Gutkin
6. Bound on Metric Entropy The main purpose of this section is to prove the bound (14) on the possible values of HKS (T p , µ). In what follows we will closely follow the approach developed in [20,21] for Anosov geodesic flows. The main technical tool is a variant of the entropic uncertainty relation first proposed in [27,28] and later generalized and proved in [29]. Here we will make use of a particular case of the statement appearing in [20,21]. s , τˆ = Theorem 5 (Entropic Uncertainty Principle [20, Thm. 6.5]). Let πˆ = {πˆ i }i=1 s , be two partitions of the unity operator 1 on a complex Hilbert space (H, ., .) {τˆi }i=1 H s , w = {w }s be the families of the associated weights. For any and let v = {vi }i=1 i i=1 normalized ψ ∈ H and any isometry U on H the corresponding pressures satisfy:
pˆ π,v ˆ ∗j ). ˆ (ψ) + pˆ τˆ ,w (Uψ) ≥ −2 log(sup v j wk τˆk U π
(45)
j,k
Let T p be a map satisfying Definition 1 and let Uk be its quantization satisfying (6) which acts on the Hilbert space Hk . In what follows we will use Theorem 5 for the Hilbert space Hk , quantum partitions πˆ = {Pε∗ , |ε| = n}, τˆ = {Pε , |ε| = n}, defined by (40) as “quantizations” of n-times refinement π (n) of the classical partition π = li=1 Ii . For each ε-element of the partitions τˆ , πˆ the corresponding weights are then defined by n−1 1/2 vε = wε = i=0 εi , ε = ε0 . . . εn−1 , where εi is the expansion rate of the map T p at the interval Iεi , εi = 1, . . . l. Finally, the isometry U will be the unitary transformation (Uk )n and the normalized state ψ will be an eigenstate ψk of Uk . With such a choice the left side of (45) reads: µˆ k (ε) log(µˆ k (ε)vε2 ) + µˆ ∗k (ε) log(µˆ ∗k (ε)vε2 ). pn,v (µˆ k ) + pn,v (µˆ ∗k ) = − |ε|=n
Thus, in order to bound pn,v (µk ) from below we need an estimation on the right hand side of (45). This amounts to the control over the elements: Pε U n Pε = U Pεn−1 U Pεn−2 . . . U Pε0 U Pεn−1 . . . U Pε0 ,
where U = Uk . The following proposition gives the required estimation. Proposition 2. Let π = li=1 Ii be the classical partition of I and let πˆ = {Pi := Op(χ Ii ), i = 1 . . . l} be the corresponding quantum partition. Then for any sequence ε = ε0 . . . εn−1 , n > 0, the product Pε = U Pεn−1 U Pεn−2 . . . U Pε0 , satisfies the bound: 1/2
Pε ≤ Nk
n−1
−1/2 . εi
(46)
i=0
Proof. It is easy to understand the source of (46). Since the structure of U basically mimics the structure of the corresponding classical map, each time when U Pε j acts on a −1/2
vector from the Hilbert space, it “decreases” its components by the factor ε j . More precisely, for any v ∈ Hk , the absolute value of j’s component of the vector v = U Pεi v satisfies the bound |v j | ≤ (−1/2 ) εi
max
m=1,...Nk
|vm |.
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
319
Applying this inequality n times one gets for the components of the vector v (n) = Pε v: n−1 −1/2 (n) |v j | ≤ εi max |vm |, j = 1, . . . Nk . m=1,...Nk
i=0
From this the desired estimation follows immediately, since |v j | ≤ v for all j and 1/2 v (n) ≤ Nk max |v (n) j |. The entropic uncertainty principle together with Proposition 2 give now the bound on the pressure of ψk : 1/2 , (47) pn,v (µˆ k ) + pn,v (µˆ ∗k ) ≥ −2 log Nk which can be also written as h n (µˆ k ) + h n (µˆ ∗k ) ≥
1/2 . µˆ k (ε) + µˆ ∗k (ε) log vε2 − 2 log Nk
(48)
|ε|=n
n 1/2 Note that such a bound becomes nontrivial only for times n when vε = i=1 εi is 1/2 comparable with Nk . In other words, n should be of the same order as the Ehrenfest time n E . For shorter times (48) would only imply that h n (µˆ k ) + h n (µˆ ∗k ) > C0 , where C0 < 0 (which is completely redundant as h n is a positive function). It is now tempting to use the inequality (48) for n = n E to get a bound on the metric entropy. Recall, however, that in such a case the relevant partition used to define h n E is of the quantum size Nk−1 . On the other hand, the correct order of the semiclassical and classical limits in the definition of HKS (T, µ) requires a bound on the entropy function for partitions of a finite (classical) size, independent of k. Thus in order to extract useful information from (47,48) it is necessary to connect the pressure pn E ,v (µˆ k ) for the quantum time n E with the pressure pn,v (µˆ k ) for an arbitrary classical time n (independent of k). To this end it has been suggested in [17] to make use of the subadditivity of the metric entropy. More specifically, for a classical invariant measure µ the subadditivity of the entropy function implies: pn+m,v (µ) ≤ pn,v (µ) + pm,v (µ), ⇒ pm,v (µ) ≤ qpn,v (µ) + pr,v (µ), m = qn + r.
(49)
The subadditivity property (49) cannot be utilized straightforwardly, as the weights µˆ ∗k , µˆ k are not T -invariant, in general. However, as follows from the following lemma, by virtue of the Egorov property the measures µk (ε) of cylinders ε turn out to be invariant for sufficiently short times. Lemma 1. Let ε be a cylinder of the size |ε| = m ≤ n E , and let µk (ε) be its measure as defined by Eq. (10). Then µk (ε) = µˆ ∗k (ε) = µˆ k (ε), where µˆ ∗k (ε), µˆ k (ε) as in Eq. (41). Furthermore, for any integer 0 ≤ n such that n + m ≤ n E the measures remain invariant under the action of T p−n : µk (ε) = µk (T p−n ε).
(50)
320
B. Gutkin
Proof. Note that any set ε of the size |ε| = m ≤ n E can be written in the x-representation (see Remark 2) as x1 x2 . . . xm , where m ≤ k. The proof of the lemma then straightforwardly follows from Corollary 1 and the fact that all Px commute with each other for |x| ≤ k. From this lemma the desired connection between the pressures for partitions of classical and quantum sizes immediately follows. Proposition 3. Let Uk , k = 1, . . . ∞ be a sequence of unitary quantizations of a map T p (satisfying Definition 1), and let {ψk } be a sequence of their eigenstates with the corresponding measures µk , as defined by eq. (10). Set the pressure for each µk to be pn,v (µk ) := − µk (ε) log µk (ε)vε2 . |ε|=n
Then for n E = qn + r , q, n, r ∈ N, 0 ≤ r < n we have: pn E ,v (µk ) ≤ qpn,v (µk ) + pr,v (µk ).
(51)
Proof. Straightforwardly follows from the subadditivity of h n and (50). Equipped with the above proposition we can prove now the bound (14) on the metric entropy for maps T p . Theorem 6 Let Uk , k = 1, . . . ∞ be a sequence of unitary quantizations of a map T p , and let {ψk } be a sequence of their eigenstates. Assuming that the corresponding limiting invariant measure µ = limk→∞ µk exists, µ must satisfy the following bound: 1 HKS (T p , µ) ≥ µ(I j ) log j − log max . (52) 2 j
Proof. From the bound (48) and Proposition 3 it follows that the pressure for the partition of an arbitrary fixed size 0 < n < n E satisfies the inequality: pn,v (µk ) 1 pr,v (µk ) r pn,v (µk ) ≥ − log max − − . n 2 nE n nE
(53)
Because r , pr,v are bounded for a fixed n, the last two terms in the right hand side of (53) vanish when k → ∞ and one gets: pn,v (µ) 1 ≥ − log max . n 2 To complete the proof it remains to notice that n µ(ε) log εi , pn,v (µ) = h n (µ) − |ε|=n
and
µ(ε) log
|ε|=n
since µ is an invariant measure.
n i=1
i=1
εi
(54)
=
j
µ(I j ) log j ,
(55)
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
321
7. Proof of the Anantharaman-Nonnenmacher Conjecture for T p Maps As we have shown in the previous section, the method of N. Anantharaman and S. Nonnenmacher can be employed for the proof of the bound (14). However, exactly as for Anosov geodesics flows, such an approach does not allow to prove a stronger result (15). Very roughly, the reason for this can be explained in the following way. For a generic map the entropy function h n (µk ) is a “non-homogeneous” quantity which contains contributions from the cylinders ε with different “expansion rates” ε . The domain of validity for subadditivity of the entropy function is determined by an entry (cylinder) with the largest expansion rate and thus, restricted to the times n ≤ n E . On the other max , hand, the bound (47) becomes informative for times n ≥ n, ¯ where n¯ = n E log 2 log l log = i=1 µ(Ii ) log i . When the expansion rate is highly non-uniform one is unable to match long “quantum” times n > n¯ with short “classical” times n < n E , see Fig. 2. This results in the bound (14) which is, in general, weaker than (15) for maps with non-uniform slopes. Below we formulate a certain modification of the original strategy to overcome the problem.
7.1. General idea. Speaking informally, the basic idea here is to “homogenize” the original system, making it uniformly expanding first and only then apply the method used in the previous section. More specifically, we consider the class of maps T = T p , given by Definition 1. In what follows we adopt the tower construction widely used in the theory of dynamical systems (see e.g., [30]). As we show in the next subsection, T can be regarded as the first return map for a certain uniformly expanding dynamical system. Namely, the action of T on I turns out to be equivalent to the action of the : so-called tower map T I → I on a subset (“zero level”) of the tower phase space I. By a standard construction for first return maps, any invariant measure µ for T induces . The corresponding metric entropies HKS (T , a measure µ on I invariant under T µ), HKS (T, µ) are then related to each other by Abramov’s formula and the entropic bound (15) turns out to be equivalent to: 1 , µ) ≥ log p. HKS (T 2
(56)
Thus, in order to prove the conjecture of S. Nonnenmacher and N. Anantharaman for maps T p one needs to show (56) for the measure µ. It turns out that a pure classical construction above can be “lifted” to the quantum level. Recall that µ is a semiclassical measure generated by eigenstates of a sequence {Uk } of unitary quantizations of T . The crucial observation is that µ is actually a semi k } of quantizations of T . In Subsect. 7.3 we show classical measure for a sequence {U that for each sequence {ψk } of the eigenstates of {Uk } generating in the semiclassical k } generating the limit the measure µ there exists a sequence { k } of eigenstates of {U measure µ. This is schematically depicted by the following diagram: µψk k→∞
µ = µ ◦ T −1
Quantum
+3 µ k k→∞
Classical3+ −1 µ= µ◦T
(57)
322
B. Gutkin
n
classical
quantum
nE classical
I1
n
I0 [0]
quantum
[1]
nE Fig. 2. On the left down (up) the case when (14) provides non-trivial (resp. trivial) bound on the metric entropy HKS (T, µ) is shown. On the right is depicted the tower for the map T{2,4,4}
is a map with a uniform expansion rate one can apply the method used in Since T the previous section in order to prove (56). From this the metric bound (15) follows immediately. As we would like to keep the exposition and notation below as simple as possible, we will first consider in detail the map T2 = T{2,4,4} defined in (25). The results can then be straightforwardly extended to all other maps T p . 7.2. Classical towers. In what follows we construct the tower dynamical system corresponding to the map T := T{2,4,4} as defined by Eq. (25). To this end let us double the original phase space and consider the set I := I × {0, 1}. We will refer to the sets I0 := I × {0} = {(x, 0), x ∈ I }, I1 := I × {1} = {(x, 1), x ∈ I } as zero and first levels of the tower I = I0 ∪ I1 respectively. Let D0 := [0, 1/2), D1 := [1/2, 1). The tower : map T I → I is then defined by: ¯ (x, η) = (T (x), 0) if η = 0, x ∈ D0 or η = 1 and any x, (58) T (T¯ (x), 1) if η = 0, x ∈ D1 where T¯ := T{2,2} is the uniformly expanding map corresponding to T . Consider now I on the set I on the first return map T I0 . It is straightforward to see that the action of T 0 0 I0 ∼ = I coincides with the action of T on I . In other words, T can be regarded as the first return map for the lowest level of the tower (see Fig. 2). I ) one can construct (using Given an invariant measure µ for T (equivalently for T 0 a standard procedure, see e.g., [25,31]) the probability measure µ which is invariant . Precisely, for a set A ⊆ I one defines the measures of the sets under the tower map T (A × {0}), (A × {1}) by µ(A × {0}) = −1 µ(A),
µ(A × {1}) = −1 µ(T¯ −1 A ∩ [1/2, 1]),
with the normalization constant = 1+µ([1/2, 1]). If A = x is a cylinder set (defined with respect to the action of T¯ ) this can be rewritten as: µ(x × {0}) = −1 µ(x),
µ(x × {1}) = −1 µ(1x).
(59)
it makes sense to consider the corresponding metric entropy Since µ is invariant under T , , HKS (T µ). An important observation is that HKS (T µ) is related to HKS (T, µ). As T is −1 the first return map for I0 , and µ( I0 ) = , by Abramov’s formula (see e.g., [25]): , µ). HKS (T, µ) = HKS (T
(60)
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
323
Having an invariant measure µ on I it is possible in turn to construct a measure µ¯ on I which is invariant under the homogeneous map T¯ . Let π I : I → I be a natural projection on the tower: π I (x, η) = x, for all x ∈ I , η ∈ {0, 1}. As = T¯ ◦ π I , πI ◦ T it follows immediately that the measure µ¯ := µ ◦ π ∗I
(61)
is invariant under T¯ . Furthermore, the metric entropy of µ¯ turns out to be equal to the metric entropy of µ: , HKS (T¯ , µ) ¯ = HKS (T µ). (62) This equality can be deduced, from a version of the Abramov-Rokhlin relative entropy formula in [32]. For the sake of completeness we give a simple proof of (62) in Appendix B. The above construction allows a straightforward extension to the case of an arbitrary map T p satisfying Definition 1. The tower phase space here is defined as lt = n max (n max = max j=1,...l {n j }) copies of I : I = I × {0, 1, . . . , lt − 1} ∼ =
l t −1
Ij,
(63)
j=0
p : I → I where the set I j = I × { j} stands for j’s level of the tower. The tower map T is then defined with the help of the uniformly expanding map T¯ p given by Eq. (16). For each level η ∈ {0, 1, . . . , lt − 1} define the set: Ij. Dη := { j|n j =η+1}
p is given by: Then the action of the map T ¯ if x ∈ T¯ η Dη p (x, η) = (T p (x), 0) T (T¯ p (x), η + 1) if x ∈ T¯ η Dη .
(64)
p a point x ∈ Dη × {0} climbs η Such a definition implies that under the action of T steps upstairs in the tower phase space, then it “jumps” downstairs to zero level and the process is repeated. It is now straightforward to see that the map T p coincides with the first return map p for zero level of T I0 of the tower. As a result, starting from an invariant measure µ p . For a set µ for the tower map T for T p one can easily construct the invariant measure A × {η} ⊆ I , with A ⊆ I and level η ∈ {0, . . . lt − 1} the corresponding measure is given by µ(A × {η}) = −1
{k|n k ≥η}
µ(T¯ p−η (A) ∩ Ik ),
=
l
n j µ(I j ),
(65)
j=1
where is the average return time to zero level of the tower. Precisely as for the map T{2,4,4} , one can also construct the measure µ¯ invariant under the action of T¯ p . The corresponding metric entropies are then related by: p , HKS (T p , µ) = HKS (T¯ p , µ) ¯ = HKS (T µ). (66)
324
B. Gutkin
7.3. Quantum towers. We are going now to consider the quantum analog of the above tower construction. Construction. Let U = Uk be a tensorial quantization of the map T = T{2,4,4} , acting on the Hilbert space Hk of the dimension 2k = dim(Hk ). We will assume that U is of the form (26). In that case U allows an obvious decomposition: U = U¯ P0 + V¯ U¯ P1 ,
V¯ = σ U¯ .
(67)
Here U¯ : Hk → Hk , U¯ : Hk → Hk stand for tensorial quantization (20) of the uniformly expanding map T¯ = T{2,2} with U given by U1 and U2 respectively, and σ stands for the unitary transformation reversing the order of the last two symbols in |x1 ⊗ · · · ⊗ |xk−1 ⊗ |xk ∈ Hk : σ |x1 ⊗ · · · ⊗ |xk−1 ⊗ |xk = |x1 ⊗ · · · ⊗ |xk ⊗ |xk−1 . In addition to P0 , P1 it will be also convenient to use the projection operator: P1 = U¯ P1 U¯ ∗ .
(68)
Explicitly its action on the basis states of Hk is given by: P1 |x1 ⊗ · · · ⊗ |xk−1 ⊗ |xk = |x1 ⊗ · · · ⊗ |xk−1 ⊗ U1 P1 U1∗ |xk , where P1 |i = δi,1 |i, i = 0, 1. It worth to notice that P1 commutes with V¯ : V¯ P1 = P1 V¯ .
(69)
=H 0 ⊕ H 1 , dim(H) = 2k + 2k−1 with We define now the “tower” Hilbert space H 0 := Hk , and H 1 := U¯ P1 Hk ≡ P Hk , H 1
(70)
0 , H 1 with the corresponding to zero and the first levels of the tower. As we identify H 0 (resp. φ ∈ H 1 ) can also be regarded space (resp. the subspace of) Hk , any vector φ ∈ H as a vector from Hk . By this identification, operations defined on Hk can be “lifted” to 0 , H 1 , as well. In particular, we can define the scalar product on H the Hilbert spaces H using the scalar product on Hk . Namely for = (φ0 , φ1 ) ∈ H, = (φ0 , φ1 ) ∈ H, 0 and φ1 , φ ∈ H 1 : with φ0 , φ0 ∈ H 1 ( , ) = φ0 , φ0 + φ1 , φ1 . can be easily constructed from an orthonormal basis in Hk . An orthonormal basis in H A convenient choice is provided by the vectors: E(x,0) := (|x1 ⊗ · · · ⊗ |xk−1 ⊗ |xk , 0), x = x1 . . . xk−1 xk , E(x,1) := (0, |x1 ⊗ · · · ⊗ |xk−1 ⊗ |1 ),
x = x1 . . . xk−2 xk−1 ,
(71)
where |0 := U1 |0, |1 := U1 |1 and xi , i = 1, . . . k (resp. i = 1, . . . k − 1) run over all possible sequences of {0, 1}.
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
325
In what follows we will consider a one-parameter family of tower evolution operators →H defined in the following way. For any = (φ0 , φ1 ) ∈ H, with φ0 ∈ H 0 θ : H U and φ1 ∈ H1 : θ := (V¯ P φ1 + U¯ P0 φ0 , eiθ U¯ P1 φ0 ). U 1
(72)
→H is given by ∗ : H Correspondingly, the adjoint operation U θ θ∗ = (e−iθ P1 U¯ ∗ φ1 + P0 U¯ ∗ φ0 , P V¯ ∗ φ0 ). U 1
(73)
and U ∗ ∈ H θ is a unitary θ , U Main properties. It is straightforward to see that U θ operation on H: ∗ be as above, then θ , U Proposition 4. Let U θ θ U θ∗ U θ∗ = U θ = 1. U ∗ U Proof. In the block matrix representation the product U θ θ takes the form P0 + P1 P0 U¯ ∗ V¯ P1 P0 U¯ ∗ e−iθ P1 U¯ ∗ U¯ P0 V¯ P1 = . P1 V¯ ∗ 0 P V¯ ∗ U¯ P0 P eiθ U¯ P1 0 1
1
(74) Using Eqs. (68, 69) it is now straightforward to see that the off-diagonal terms in the right side of (74) vanish. θ . Specifically the short Below we demonstrate that the Egorov property holds for U time evolution of projection operators is prescribed by the classical evolution of the corresponding tower map. Proposition 5. Let x ⊆ I be a cylinder of the length m = |x| < k − 1, then: θ = (P0 P ¯ −1 ⊕ P P ¯ −1 ), θ∗ (Px ⊕ 0) U U T x 1 T x ∗ Uθ (0 ⊕ P1 Px ) Uθ = (P1 PT¯ −1 x ⊕ 0). Proof. In the matrix representation the left side of (75) reads: ¯ P0 U¯ ∗ e−iθ P1 U¯ ∗ U P0 V¯ P1 Px 0 . P1 V¯ ∗ 0 0 0 eiθ U¯ P1 0
(75) (76)
(77)
By Eqs. (68, 69) the off diagonal terms of the above product are zeros, and the diagonal part: P0 U¯ ∗ Px U¯ P0 0 0 P1 V¯ ∗ Px V¯ P1 coincides with the right side of Eq. (75), as follows from Corollary 1 and the equality: σ Px σ = Px . Equation (76) is then proved analogously.
326
B. Gutkin
Let X be a subset of I and let PX be the corresponding projection operator. The projection operator corresponding to the subset X × {0} ∪ X × {1} of the tower, is then defined as: X = PX ⊕ P PX . P 1 The following corollary follows immediately from Proposition 5. Corollary 2. For a given cylinder x ⊂ I of the length |x| = m and an integer n such ¯ −n be the projection operator corresponding to the subset that n + m < k − 1, let P T x −n −n ¯ ¯ T x × {0} ∪ T x × {1} of the tower. Then: ¯ −n U θ∗ P U T x θ = PT¯ −n−1 x .
(78)
Eigenfunctions and semiclassical measures. Given an eigenfunction ψ of the original θ = eiθ evolution operator U , U ψ = eiθ ψ we can construct the eigenfunction , U of the tower evolution operator. Precisely, −1 = ψ, U¯ P1 ψ ψ 2 ,
ψ = 1 + ψ, P1 ψ,
(79)
θ . Indeed, this is so, since by (68), is the normalized eigenstate of the operator U 1 1 θ = (U¯ P0 + U¯ 1 P U¯ P1 )ψ, eiθ U¯ P1 ψ − 2 = U ψ, eiθ U¯ P1 ψ − 2 , U ψ ψ 1 and ψ is the eigenstate of U . It is also easy to verify that ( , )) = 1. For any sequence of quantizations {Uk } of T and their eigenstates {ψk }, let us con θk } of quantizations of the tower map T determined sider the corresponding sequence {U by Eq. (72). (Note that these quantizations depend on the eigenvalues eiθk of ψk ’s.) θk } can be constructed applying Eq. (79). As a result, Then the eigenstates { k } of {U a sequence of semiclassical measures µk on I induces the sequence of semiclassical measures µk on I . For a cylinder x ⊂ I , |x| ≤ k the measures µk of the tower sets x × {0}, x × {1} are defined as: µk (x × {0}) = ( k , (Px ⊕ 0) k) ,
µk (x × {1}) = ( k , (0 ⊕ P1 Px ) k) .
By Eq. (79) these measures are related to the measure µk of the set x: µk (x × {0}) = k−1 µk (x),
µk (x × {1}) = k−1 µk (1x),
(80)
where we set k = ψk . Note that after taking the limit k → ∞ in (80) one obtains Eqs. (59), where µ = limk→∞ µk is precisely the measure of the classical tower obtained from the semiclassical measure µ = limk→∞ µk by the procedure from the previous section. Also, defining the measure µ¯ k on I by x k) = −1 µk (x) + µk (1x) , µ¯ k (x) := ( k , P (81) k one reveals in the semiclassical limit the measure µ¯ = limk→∞ µ¯ k related to µ by Eq. (61).
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
327
Extension to other maps T p . The above construction of quantum towers can be straightforwardly extended to other maps T p . Let U be a quantization of T p acting on the Hilbert space Hk , as defined by Eq. (24). Note that U can be cast in the form: U=
l t −1
U¯ (0) = 1, U¯ (η) := U¯ η U¯ η−1 . . . U¯ 1 for η > 0,
ση U¯ (η+1) PDη ,
η=0
where U¯ i , i = 1, . . . lt are quantizations of the uniformly expanding map T¯ p , σ j is the operator which reverse the order of the last j elements in the products |x1 ⊗· · ·⊗|xk−1 , lt is the height of the corresponding classical tower and PDη is the projection operator corresponding to the set Dη as defined in Sect. 7.2. Mimicking the construction of the classical towers one defines then the tower Hilbert space as the direct sum: = H
lt −1
η , H
η=0
η ∼ H = Pη + P¯η Hk ,
with the projection operators Pη , P¯η defined by Pη := U¯ (η) PDη (U¯ (η) )∗ ,
P¯η := U¯ (η) P¯Dη (U¯ (η) )∗ ,
P¯Dη :=
PD j .
j>η
θ whose action on the states = φ0 , φ1 , . . . Now take the tower evolution operator U is given by: φlt −1 ∈ H ⎛
l t −1
θ = ⎝ U
η=0
⎞ ση U¯ η+1 Pη φη , eiθ U¯ 1 P¯0 φ0 , eiθ U¯ 2 P¯1 φ1 , . . . , eiθ U¯ lt P¯lt −1 φlt −1 ⎠ . (82)
θ follow from It can be easily checked that the unitarity and the Egorov property for U the same properties for the operator U . Finally, if U ψ = eiθ ψ then = ψ, U¯ (1) P¯D0 ψ, U¯ (2) P¯D1 ψ, . . . , U¯ (lt −1) P¯Dlt −2 ψ / ψ
(83)
θ with the eigenvalue eiθ and the normalization constant given by is the eigenstate of U ψ = 1 +
l t −2 j=0
ψ, P¯D j ψ =
l
ψ, PI j ψn j .
j=1
This shows that any semiclassical measure µ = limk→∞ µk on I induced by a sequence of eigenstates {ψk } of Uk , generates through Eq. (83) the semiclassical measures: µ= µk on I and µ¯ = limk→∞ µk ◦ π ∗I on I which are invariant under the action limk→∞ p and T¯ p , respectively. Note also that the corresponding metric entropies are related of T to each other by Eq. (66).
328
B. Gutkin
7.4. Proof of Theorem 3. Let us now prove the bound (15) for the map T{2,4,4} . Theorem 7 Let {Uk }∞ k=1 be a sequence of tensorial quantizations of T = T{2,4,4} . For ∞ a sequence {ψk }k=1 of eigenstates Uk ψk = eiθk ψk let µ = limk→∞ µk be the corresponding semiclassical measure, then HKS (T, µ) ≥
µ(0) + 2µ(1) log 2. 2
(84)
Proof. To prove (84) it is possible, in principle, to follow precisely the scheme described , µ) for the corresponding in the beginning of the section i.e., to prove the bound on HKS (T semiclassical measure µ on the tower and then deduce the bound (84) using Abramov’s formula. From the technical point of view, however, it turns out to be easier to prove an equivalent bound for the metric entropy HKS (T¯ , µ) ¯ of the measure µ. ¯ Let { k }∞ be the sequence of the tower eigenstates corresponding to the sequence k=1 ˆ of ψk ’s, and let h n ( k ) ≡ h n (µ¯ k ) be the entropy function for the corresponding measures µ¯ k : x k 2 log( P x k 2 ). (85) h n (µ¯ k ) = − µ¯ k (x) log µ¯ k (x) = − P |x|=n
|x|=n
Then the metric entropy HKS (T¯ , µ) ¯ is obtained after first applying the semiclassical limit: h n (µ) ¯ = lim h n (µ¯ k ),
(86)
1 HKS (T¯ , µ) ¯ = lim h n (µ). ¯ n→∞ n
(87)
k→∞
and then the classical limit:
¯ corresponds to the “probing” of towers with parAs can be seen from Eq. (85), HKS (T¯ , µ) x ). This explains titions made of “vertical rectangles” (represented by the projections P , µ). ¯ and HKS (T ˜ To prove the source of the equality between metric entropies HKS (T¯ , µ) the bound on HKS (T¯ , µ) ¯ we will make use of the same scheme as in [17]. The first step is to get the bound on the entropy function, when n is of the same order as k. This is provided by the following proposition. Proposition 6. Let h n (µ¯ k ) be as in (85) and set n = k − 1, then k−1 − 1 log 2. h k−1 (µ¯ k ) ≥ 2
(88)
Proof. We will use the Uncertainty Entropic principle (Theorem 5) for the partitions: y , |y| = k − 1}, weights: vy = wy ≡ 1 and the isometry operation π = τ = {P k −1 θk ) . Since k is an eigenstate of U θk it follows immediately from (45): U = (U h k−1 ( k ) ≥ − log(
sup
|y|=|y |=k−1
y (U θk )k−1 P y ). P
(89)
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
329
y (U θk )k−1 P y . To Thus one needs to estimate the norm of the operator C(y , y) = P this end let us calculate the matrix elements of C(y , y): ( E(x ,i ) , C(y , y)E(x,i)) , in the basis of orthogonal states (71) with the parameters: i, i ∈ {0, 1}, |x| = k, (|x | = k) if i = 0 (resp. i = 0) and |x| = k − 1, (|x | = k − 1) if i = 1 (resp. i = 1). The action of the projection operator on the basis states is given by ⎛ y E(x,i) = E(x,i) ⎝ P
k −1
⎞ δxm ,ym ⎠ .
(90)
m=1
Hence for each pair of y, y there exist at most two values of x and two values of x such that the matrix elements ( E(x ,i ) , C(y , y)E(x,i)) are not zeros. From that follows: C(y , y) ≤ 2
max
(x,i),(x ,i )
=2
max
|((E(x ,i ) , C(y , y)E(x,i)) |
(x,i),(x ,i )
θk )k−1 E(x,i)) |. |((E(x ,i ) , (U
(91)
θk )k−1 in the basis of Therefore, it remains to estimate the elements of the operator (U θk on {E(x,i) } up to times k closely {E(x,i) }. To this end, let us notice that the action of U on the sets x × {i} of connected to the action of the corresponding tower map T I. k −1 Specifically, let E = (Uθk ) E(x,i) . Then, as follows from Eq. (72), depending on x, i the state E takes the values (e, 0) or (0, e), where e is of the form: e=
ei Qθk |xk ⊗ Ui1 |xi1 ⊗ Ui2 |xi2 ⊗ · · · ⊗ Uik−1 |xik−1 if i = 0, ei Qθk Ui1 |xi1 ⊗ |1 ⊗ Ui2 |xi2 ⊗ · · · ⊗ Uik−1 |xik−1 if i = 1.
(92)
Here Uim is either U1 or U2 , xi1 , xi2 . . . xik−1 is some permutation of the original sequence √ x1 , x2 . . . xk−1 and Q is an integer number. Since |x j , U1,2 xi | = 1/ 2 for any pair xi , x j ∈ {0, 1}, θk ) |((E(x ,i ) , (U
k−1
E(x,i)) | = |((E(x ,i ) , E))| ≤ 2
− k−1 2
.
Together with (89) and (91) this gives the proof of the proposition. The second necessary step is to connect values h k−1 (µ¯ k ) of the entropy at quantum times of order k to its values h n (µ¯ k ) at short fixed classical times n. Proposition 7. Let h n (µ¯ k ) be as in (85), and let k − 1 = qn + r , r < n, where n < k − 1 is a fixed (classical) time and q, r are integers, then 1 n log 2 1 h n (µ¯ k ) ≥ h k−1 (µ¯ k ) − . n k−1 k−1
(93)
330
B. Gutkin
Proof. The prove of (93) is analogous to the proof of Proposition 3. One makes use of the fact that the measure µ¯ k is invariant under the transformation T¯ j for “short” times. From the definition of µ¯ k and Eq. (78) it follows that for any cylinder x of length |x| = m: (94) µ¯ k (x) = µ¯ k (T¯ −n x), for m + n ≤ k − 1. Let n, q, r be as in the conditions of the proposition. Then the subadditivity property of the entropy function implies: h k−1 (µ¯ k ) ≤ qh n (µ¯ k ) + h r (µ¯ k ).
(95)
Since |h r (µ¯ k )| is bounded from above by n log 2, one immediately obtains the inequality (93). End of the proof of Theorem 7: The final step is to combine Propositions 6 and 7: 1 log 2 (n + 1) log 2 h n (µ¯ k ) ≥ − , for all n < k. n 2 k−1 Taking in (96) first limit k → ∞ and then n → ∞ gives: log 2 HKS (T¯ , µ) , ¯ ≥ 2 which by (62, 60) implies the bound: log 2 HKS (T, µ) ≥ . 2 Since = µ(0) + 2µ(1), this gives the bound (84).
(96)
(97)
Sketch of proof of Theorem 3. All the ingredients of the above construction can be straightforwardly extended from the map T{2,4,4} to a general map T p satisfying Definition 1. Specificaly, given a sequence of eigenstates Uk ψk = eiθk ψk , k = 1, . . . ∞, one k }∞ of eigenstates for quantizations {U θk }∞ of k ∈ H first constructs the sequence { k=1 k=1 the tower map T p . These sets of eigenstates induce then two sequences of related measures (defined as in Eq. (81)): {µk }∞ ¯ k }∞ k=1 , {µ k=1 on I . Assuming that in the semiclassical limit µk ’s converge to an invariant measure µ of T p , the sequence of the measures µ¯ k must converge to the measure µ¯ which remains invariant under the action of the corresponding uniformly expanding map T¯ p . Repeating the same steps as in the proof of Proposition 6 it is strightforward to get the metric bound on the measures µ¯ k : k + 1 − lt h k+1−lt (µ¯ k ) ≥ (98) + 1 − lt log p. 2 Furtheremore, the subadditivity property of h n together with Eq. (98) then imply that the limiting measure µ¯ = limk→∞ µ¯ k must satisfy the bound: log p . HKS (T¯ p , µ) ¯ ≥ 2 Since the metric entropies HKS (T¯ p , µ), ¯ HKS (T p , µ) are connected to each other by Eq. (66) one immediately gets log p HKS (T p , µ) ≥ , (99) 2 where is as in Eq. (65). Finally, it remains to check that the right side of (99) coincides with the right side of (15).
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
331
8. Explicit Sequences of “Non-ergodic” Eigenstates Below we construct some explicit sequences of eigenstates for maps T¯ p , T p quantized as in Sect. 3. Having such sequences we can calculate the induced semiclassical measures and test the bound (15) for the corresponding metric entropies. 8.1. Maps with uniform slopes. Let us first consider the map T¯ p whose quantization is given by Eq. (20). Note that if U∗ is given by the discrete Fourier transform F p , the evolution operator U¯ k and the corresponding eigenstates are precisely the same as for the Walsh-quantized baker’s map treated in [17]. As we show below, the same construction can be carried out for a general U. Similarly to the case of the Walsh-quantized baker’s map, the resulting semiclassical measures on I turn out to be Bernoulli measures. Let w ∈ H be an eigenstate of U; then it is easy to see that (w)
ψk
(w)
= w ⊗ w ⊗ · · · ⊗ w,
ψk
k
∈ Hk = H ⊗ H ⊗ · · · ⊗ H
(100)
k
is the eigenstate of U¯ k . It is now straightforward to compute the semiclassical measure p−1 (w) µw corresponding to the sequence ψk , k = 1, . . . ∞. Assuming that w = i=0 wi |i, where {|i, i = 0, . . . p−1} is an orthonormal basis in H, the µw -measure of the cylinder set x, x = x1 . . . xm , xi ∈ {0, . . . p − 1} is given by: (w)
(w)
µw (x) = lim ψk Px ψk = k→∞
m
|wxi |2 .
(101)
i=1
As this is a simple product measure (i.e., a measure µ which factorizes: µ(x) = m i=1 µ(xi ) with respect to the corresponding Markov partition), one gets for the metric entropy: HKS (T¯ p , µw ) = −
p−1
|wi |2 log(|wi |2 ).
(102)
i=0
Now let us show that for some quantizations of T¯ p there also exist semiclassical measures which are linear combinations of Bernoulli measures constructed in the previous example. Assume that U is such that there exists an eigenstate w(0) ∈ H of Ud : Ud w (0) = eiθ w (0) for some integer d > 0, while w (0) is not an eigenstate of Ui for any integer i < d. (This, for instance, is possible if Ud = 1 for some d > 1, as in the case: U = F p∗ ). Taking then such normalized state w (0) one can form the corresponding sequence w( j) := e−iθ j/d U j w (0) , j = 0, . . . d − 1 of cyclically related states: e−iθ/d Uw ( j) = w ( j+1 mod d) , with |w ( j) |w (i) | < 1 if i = j. Define w0 := w (0) ⊗ w (1) · · · ⊗ w (d−1) and let w := {w0 , w1 , . . . wd−1 } be the sequence of the states obtained from w0 by the cyclic permutation of its components: w j := w ( j
mod d)
⊗ w (1+ j
mod d)
⊗ · · · ⊗ w (d−1+ j
mod d)
,
j = 0, . . . d − 1.
For each k satisfying: k mod d = 1 one looks for normalized eigenstates of U¯ k in the form (w)
ψk
=
d−1 j=0
(k)
Cj
w j ⊗ w j ⊗ · · · ⊗ w j ⊗w j , (k−1)/d
(w)
ψk
∈ Hk .
(103)
332
B. Gutkin (k)
Equation (103) then defines an eigenstate of U¯ k if all C j are equal to some constant C (k) (w)
fixed by the normalisation condition ψk = 1. (Note that the eigenstates provided by Eq. (100) could be seen as a particular case of (103) when d = 1.) The corresponding semiclassical measure is then given by (w)
(w)
µw (x) = lim ψk Px ψk k→∞ = lim
k→∞
d−1
|C (k) |2 w j ⊗ · · · ⊗ w j Px wi ⊗ · · · ⊗ wi .
i, j=0
Since |w ( j) |w (i) | < 1 if i = j, only the diagonal terms survive the limit k → ∞ and one has: d−1 1 ( j) µw (x), µw (x) = d
( j) µw (x)
j=0
=
m
|wx( kj+k−1
| ,
mod d) 2
(104)
k=1
( j)
where wi is i’s component of the vector w( j) in the basis {|i, i = 0, . . . p − 1}. Although µw is not a simple product measure, it is still possible to calculate the metric entropy by noticing that the product measures entering in (104) are cyclically related to ( j+1 mod d) ( j) each other: µw = T¯ p∗ ◦ µw and remain invariant under the action of (T¯ p )d = T¯ pd . Since the metric entropy is an affine function of the measure, one has: 1 HKS (T¯ p , µw ) = HKS (T¯ pd , µw ) d d−1 d−1 p−1 1 1 ( j) 2 ( j) ( j) = 2 HKS (T¯ pd , µw ) = − |wi | log(|wi |2 ), d d j=0
j=0 i=0
(105) ( j)
where we have used the fact that each µw is a simple product measure and all ( j) HKS (T¯ pd , µw ) are equal to each other. Furthermore, by a simple application of the Uncertainty Entropic Principle one has for all j = 0, . . . d − 1, ( j) ˆ ( j) ) + h(Uw ˆ h(w ) ≥ −2 log(max |Uk,m |) = log p, k,m
ˆ ( j) ) = h(w
p−1
( j)
( j)
|wi |2 log(|wi |2 ).
i=0
As e−iθ j/d Uw ( j) = w ( j+1 mod d) , it follows immediately by (105) that HKS (T¯ p , µw ) ≥ 1 2 log p which is precisely the bound (14) (equivalent to (15) in that case). It is worth mentioning that for U = F p∗ and d = 1 there exist vectors w such that the measures µw saturate this bound [17]. ( j)
Remark 3. Note that if all wi = 0 the measures above are supported on the whole I . As has been shown in [17] in the case when U = F p∗ , it is also possible to construct an entirely different class of exceptional sequences of eigenstates, where parts of the
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
333
corresponding semiclassical measures are localized on some periodic trajectories. This construction uses the fact that U¯ kn = 1 for a “short” time n, meaning the spectrum of U¯ k becomes highly degenerate. Since no such degeneracies are expected for quantized maps with non-uniform slopes, it seems that this type of semiclassical measures can be constructed only for the maps T¯ p . We refer the reader to [15, 17] for the details of the construction.
8.2. Maps with non-uniform slopes. For the maps T p we will look for sequences of eigenstates exactly in the same form (103), as in the uniform case. As we will show, for certain quantizations of T p one can, indeed, construct such sequences by choosing the (k) constants Ci in an appropriate way. Rather than consider a general case, we will provide below several concrete examples of such a construction for the map T2 (see Eq. (25)) whose quantization is given by (26) with some choice for U1 , U2 . In order to calculate the metric entropy of a semiclassical measure µ we need, in general, a knowledge of the measures µ(ε0 . . . εn−1 ), where ε0 . . . εn−1 , εi ∈ {1, 2, 3} are cylinder sets defined with respect to the Markov partition j=1,2,3 I j . On the other hand, because of the eigenstates, structure, it turns out to be easier to calculate the measures µ(x1 . . . xn ) of the “binary” cylinders, where xi ∈ {0, 1} (as defined in Sect. 3). The transition between two representations is straightforward. Given some cylinder ε0 . . . εn−1 in ε-representation its binary representation can be obtained by switching every element εi , i = 0, . . . n − 1 to 0 if εi = 1, to 10 if εi = 2 and to 11 if εi = 3 respectively. (For instance the cylinder 123 in the binary representation becomes 01011.) Thus, in order to obtain the measures of the sets ε0 . . . εn−1 , it is convenient first to transform them into the binary form using the above procedure and then find their measures as in the previous subsection. Example 1. Below we construct a semiclassical measure which is totally concentrated on a part of I , where T2 has a uniform slope. Let U1 = U2 = U be a two-by-two √ matrix satisfying U2 = 1, |U(i, j)| = 1/ 2, e.g., the discrete Fourier transform. Let U|1 =: |e+ . Since U|e+ = |1 it can be easily seen that for even k, (1)
ψk = |1 ⊗ |e+ ⊗ · · · ⊗ |1 ⊗ |e+
(106)
k
(1)
is an eigenstate of Uk . For the sequence of states ψk the induced semiclassical measure µ(1) has entire support on the Cantor set: {x = 0.x1 x2 x3 · · · ∈ I, x2i+1 = 1}. Correspondingly, µ(1) (ε0 . . . εn−1 ) = 1/2n if εi ∈ {2, 3} for all i, and µ(1) (ε0 . . . εn−1 ) = 0 otherwise. The metric entropy of µ(1) can be easily calculated and it is given by: HKS (T2 , µ(1) ) = log 2. This saturates the bound (15) coinciding in that case with (14). Let us mention that since µ(1) is supported on the interval [1/2, 1] (where the slope of T2 is 4), it is also an eigenmeasure of T¯4 map, with the same metric entropy as for T2 . Example 2. Let us show now that for some quantisations of T2 one can construct sequences of eigenstates precisely in the form (100). As a result, the corresponding semiclassical measure µ(2) is given by Eq. (101). Although it remains invariant under the action of both T2 and T¯2 , the metric entropies of µ(2) with respect to T2 and T¯2 turn out to be different.
334
B. Gutkin
Consider the following quantization of T2 . Let U be an arbitrary unitary matrix whose √ elements have modules 1/ 2 and let w be one of its eigenvectors with the eigenvalue eiγ . We now fix U1 , U2 by the conditions U2 = e−iγ U, U1 = U. The state (2)
ψk = w ⊗ w ⊗ · · · ⊗ w, k
is then an eigenstate of Uk . (Note that this example allows a straightforward general(2) ization to all maps T p .) Denote µw the corresponding semiclassical measure. Unlike (2) the previous example, in general, µw is supported over all I . For a given state w = w0 |0 + w1 |1 the measures of the sets ε0 , ε0 = {1, 2, 3} are: ⎧ ⎨ p for ε0 = 1 (2) (2) pq for ε0 = 2 µ(2) w (ε0 ) = lim ψk Pε0 ψk = ⎩ q 2 for ε = 3, k→∞ 0 (2)
where we introduced notation |w0 |2 = p, |w1 |2 = q. Since µw is a simple product measure the corresponding metric entropy is given by: (2) µ(2) HKS (T2 , µ(2) w )=− w (ε0 ) log µw (ε0 ) ε0 ={1,2,3}
= −( p log p + pq log( pq) + q 2 log q 2 ). This should be compared to the bound given by the right side of (15). For the measure (2) µw this bound is equal to 21 (1 + q) log 2. Recall that w is an eigenvector of a unitary matrix whose entries have the same This restricts the possible values of q, √ modules. √ p = 1 − q to the interval [(2 − 2)/4, (2 + 2)/4]. As can be easily checked, for all values of q, p in this interval the strict inequality (15) holds. It is worth to mention (2) that µw is, in fact, an eigenmeasure of the map T¯2 , as well. The corresponding metric entropy HKS (T¯2 , µ(2) w ), however, is given by Eq. (102) and it is obviously different from (2) the expression for HKS (T2 , µw ). Example 3. Unlike two previous examples, here we construct semiclassical measures which somewhat differ from the semiclassical measures obtained for the maps with uniform slopes. Most importantly, we demonstrate that for a certain choice of parameters the bound (15) is actually saturated. To construct such semiclassical measures we are looking for normalized eigenstates of Uk in the form of two state products: (3)
(k)
(1) (2) (1) (2) ⊗ w (2) ⊗ w (1) ⊗ w ψk = C1 w ⊗ · · · ⊗ w ⊗ w
k
(2) + C2(k) w
⊗w
(1)
⊗w
(2)
(1) (2) (1) ⊗w ⊗ · · · ⊗ w ⊗ w .
k
Take U2 = U1 = U,
1 U= √ 2
1 eiα , e−iα −1
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
335
HKS 0.7 0.6 0.5 0.4
0
0.5
1
1.5
Re z 2
Fig. 3. Metric entropy (green, solid line) and the corresponding bound (blue, dashed line) (15) for the semiclassical measure from Example 3 as functions of Re(z) when Im(z) = 0 (k)
and for a given z ∈ C set C1
√ (k) = zC2 , c = 1 + |z 2 − 1|2 ,
√ w (1) = c−1/2 |0 + e−iα (z 2 − 1)|1 ,
√ w (2) = c−1/2 z|0 + e−iα ( 2 − z)|1 .
With such a choice (U)2 = 1 and Uw (1) = w (2) , Uw (2) = w (1) . It is easy to check that ψk(3) is an eigenstate of Uk for any z ∈ C. The resulting semiclassical measure (k)
(3)
µz = (C1 )2 µ1z + (C2 )2 µ2z , Ci = limk→∞ |Ci |, i = 1, 2 is a linear combination of two simple product measures µ1z , µ2z = µ1z −1 defined in Eq. (104). Note that this measure has the same structure as in the case of maps with uniform slopes. However, here the con(2) (2) (1) stants C1 , C2 are not equal, in general. Denote p1 = |w0 |2 , p2 = |w1 |2 , q1 = |w0 |2 , (1) 2 (i) (i) q2 = |w1 | , where w0 , w1 are the first and second component of the vectors w(i) , i = 1, 2. As will be shown below, the metric entropy of µ(3) z can be explicitly calculated and it is given by HKS (T2 , µ(3) z )=−
pk log pk + qk log qk , 2 k=1,2
where = 2(µ([10]) + µ([11])) + µ([0]) = 1 + (C1 )2 p1 + (C2 )2 p2 , |z| . C1 = |z|C2 = # |z|2 + 1 The plot in Fig. 3 shows the metric entropy versus the bound of (15) as functions of the real part of z when Im(z) = 0.
2
log 2 given by right side
Some special cases: 1) |z| = 1. In this case p1 = p2 , q1 = q2 and the resulting measures are of a simple product type. Furthermore, both w(1) and w (2) are the eigenvectors of the same unitary matrix whose elements have equal modulus. Thus one actually, (2) gets the measures of the same type as for one vector product states ψk in Example 2. 2) z = 0, z = ∞. In that case either C2 or C1 vanishes and we get the states considered
336
B. Gutkin
√ √ in Example 1. 3) z = 2 (or z −1 = 2). In such a case p1 = q1 = 1/2, p2 = 0, q2 = 1 √ ) = 2 log 2 saturates the bound (15). and the metric entropy HKS (T2 , µ(3) 3 2 Calculation of metric entropy. Using the same approach as in the last example, one can, in principle, construct d-state product eigenstates of the type (103) for other maps T p . For such a construction it is necessary to define the quantum evolution operator Uk by Eq. (24) with Ui = U, i = 1, . . . n max , where Ud = 1 and the vectors w ( j) , j = 0 . . . d − 1 are cyclically related to each other by U (precisely as in Sect. 8.1). (k) Assume that by an appropriate choice of constants Ci one can obtain a sequence of (w) eigenstates ψk of Uk ’s in the form (103). The corresponding semiclassical measures ( j)
µw then have the form of a linear combination of the components µw from Eq. (104). Let us show now how the metric entropy of µw can be computed. First, note that ψk(w) being an eigenstate of Uk , is in addition, an eigenstate for the operator (U¯ k )d , where U¯ k is the quantization (20) of the map T¯ p with the uniform slope p. Since (U¯ k )d is also a quantization of the map T¯ pd , the semiclassical measure µw and all its components ( j) µw turn out to be invariant under T¯ pd . Furthermore, as has been shown in Sect. 7.2, using measure µw one can construct the corresponding measure µ¯ w invariant under the action of T¯ p . An important observation is that this measure has the same structure as µw . Namely, it is a linear combination of the simple product measures given by Eq. (104): µ¯ w =
d−1 j=0
( j)
α j µw ,
d−1
α j = 1, α j ∈ R.
j=0
It is now easy to calculate HKS (T¯ p , µ¯ w ) using the affineness of the metric entropy and ( j) the fact that HKS (T¯ pd , µw ) is equal to HKS (T¯ pd , µw ) for each j: d−1 1 1 1 ( j) HKS (T¯ p , µ¯ w ) = HKS (T¯ pd , µ¯ w ) = α j HKS (T¯ pd , µw ) = HKS (T¯ pd , µw ). d d d j=0
(107) Using now connection (66) between the metric entropies of µw , µ¯ w and expression (105) for HKS (T¯ pd , µw ) one obtains: d−1 p−1 ( j) 2 ( j) ¯ HKS (T p , µw ) = HKS (T pd , µw ) = − |wi | log(|wi |2 ). d d
(108)
j=0 i=0
Note that as the right side of (15) amounts to log2 p the proof of the AnantharamanNonnenmacher conjecture for the measure µw amounts to the proof of the bound log p d , HKS (T¯ pd , µw ) ≥ 2
(109)
for the uniformly expanding map. This bound has been shown already in Sect. 8.1. Note also that the bound (15) saturates if (109) saturates for the corresponding uniformly expanding map T¯ pd .
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
337
Finally, it is instructive to see how the formula (108) can be understood in an intuitive way. Recall that each point ζ ∈ I can be encoded by sequences ε(ζ ) = ε0 ε1 . . ., εi ∈ {1, . . . l} in accordance with the dynamical (forward) “history” of ζ regarding the action of T p . Furthermore, this sequence generates the set of cylinders: G n (ζ ) := {ε0 . . . εn−1 , n = 1, 2, . . . } corresponding to the “partial histories” of the point evolution. Assuming that µw is ergodic, the Shannon-McMillan-Breiman theorem asserts that for almost every (with respect to µw ) point ζ ∈ I the metric entropy of T p is given by: 1 log µw (G n (ζ )). (110) n Speaking informally, this means that on average the measures of cylinders G n = ε0 . . . εn−1 decay exponentially as functions of n: µw (G n ) ∼ exp(−n HKS (T p , µw )) with the exponent given by HKS (T p , µw ). Analogously, one can argue that HKS (T¯ pd , µw ) determines the measures of G n = x1 . . . xm n in the x-representation: µw (G n ) ∼ exp(− mdn HKS (T¯ pd , µw )). On the other hand, by Birkhoff’s ergodic theorem the lengths of G n in both representations can be easily connected. Namely, for almost every ζ ∈ I (with respect to µw ), HKS (T p , µw ) = − lim
n→∞
lim m n /n =
n→∞
l
n i µw (Ii ) = ,
i=1
where m n , n is the length of G n (ζ ) in the x and ε-representation, respectively. Comparing the asymptotics of µw (G n ) in both representations gives Eq. (108). 9. Conclusions and Outlook In the current paper we have proved the Anantharaman-Nonnenmacher conjecture for the “tensorial” quantizations of the one-dimensional piecewise linear maps T p . It should be emphasized that we deal here with the “tensorial” quantizations mostly for the sake of convenience, as these quantizations allow very explicit treatment. Actually the current method with minimal adjustments can be used to prove the result for general quantizations of the maps T p . On the other hand, the present approach is clearly restricted to the class of maps whose slopes are all powers of an integer p > 1, since only such maps can be represented as first return maps for towers with uniform expansion rates. To prove the conjecture for a more general class of maps (or other chaotic systems) a more flexible tower construction is needed. We believe that such a modified tower construction is in fact possible and can be utilized to prove the result for all quantizable piecewise linear (chaotic) maps T , [33]. Note also that in the recent preprint [34] of G. Riviere a related approach has been used to prove the Anantharaman-Nonnenmacher conjecture for geodesics flows on two-dimensional Riemannian manifolds of non-positive curvature. The present application demonstrates that quantized one-dimensional maps (or similarly à la Walsh quantized asymmetric Baker maps) can be useful as toy models for understanding general features of quantum chaotic systems. On the technical level these models are much simpler than generic chaotic Hamiltonian systems, but still exhibit their most important features. A quite rare opportunity (for chaotic systems) to construct explicit sequences of eigenstates make them potentially useful as test systems. Another possibility is to use one dimensional maps as models for scattering systems. By opening a “gap” in the unit interval one can produce quantized one-dimensional maps with
338
B. Gutkin
an “absorption” (in complete analogy with the open Walsh-Baker maps introduced in [35]). Finally, since we know already that various exceptional semiclassical measures appear for the “tensorial” quantizations of the maps T p it would be of interest to identify an opposite class of quantizations for which quantum unique ergodicity holds i.e, no exceptional sequences of eigenstates are present. Acknowledgements. I would like to thank Andreas Knauf and Christoph Schumacher for numerous fruitful discussions. I am particularly grateful to Stephane Nonnenmacher for reading the preliminary version of the manuscript and very helpful comments. Most of the present work was accomplished during my pleasant stay in Erlangen-Nuremberg University. I am grateful to all my colleagues at the Mathematical Department for the hospitality extended to me. The financial support of the Minerva Foundation and SFB/TR12 of the Deutsche Forschungsgemainschaft is acknowledged.
Appendix A: Quantization of General Maps We have shown in Sect. 3 that the maps T p can be quantized by means of the “tensorial” quantization procedure. Here we discuss how a more general class of maps T , = {1 , . . . l } can be quantized. Let i1 , . . . i , > 1 be the maximal set of different slopes in , i.e., in = im for n = m. Assuming that each slope ik has a multiplicity m k ≥ 1, the Lebesgue measure preservation condition mk = 1, ik
(111)
k=1
imposes certain restrictions on the values of ik , m k . In particular, it is clear that the set ik , k = 1, . . . must have a greatest common divisor p larger than one. This means ¯ k, ik = p
¯ k ∈ N for k ∈ {1, . . . }.
¯ i ’s are relatively prime. Then it follows immediately from (111) Assume now that all ¯ k , m¯ k ∈ N, k ∈ {1, . . . l}, where lk=1 m¯ k = p. that m k ’s are of the form m k = m¯ k We are going now to show that the maps T whose slopes satisfy the above conditions are quantizable. ¯ i , p ∈ N, ¯ i+1 ≥ ¯i Proposition 8. Let T be a map (4) with the slopes i = p ¯ i ’s are relatively prime integers, then T is having multiplicities m i , i ∈ {1, . . . l}. If “quantizable”. Proof. As the first step notice that T can be represented as a composition of the uniformly expanding map T¯ p and the “block diagonal” map TBD , whose slopes are uniform at each block. Lemma 2. Let T be a map as defined above, then T = T¯ p ◦ TBD , where T¯ p (x) = px mod 1 and mj for x ∈ [bi , bi+1 ], bi = . TBD (x) = (i x mod 1) / p + bi , j j
Proof. Straightforward calculation.
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
1
1
339
1
= 0
1/2
1
0
1/2
1
0
1/2
1
Fig. 4. A “generic” map (112) and its decomposition into the uniformly expanding and the “block diagonal” parts
The parameters entering into the definition of TBD have the following simple meaning. m The points bi , bi+1 mark the position of i’s block which is the square of the size jj . Inside of each such block the map TBD acts as a piecewise linear map with the uniform ¯ i. expansion rate Example. To illustrate the above lemma consider as an example the map with the slopes 6 and 4: 6x mod 1 if x ∈ [0, 1/2) T (x) = (112) 4x mod 1 if x ∈ [1/2, 1). As shown in Fig. 4, it can be decomposed into the uniformly expanding map T¯2 = 2x mod 1 and the “block diagonal” map: TBD (x) =
(6x mod 1)/2 if x ∈ [0, 1/2) (4x mod 1)/2 + 1/2 if x ∈ [1/2, 1).
Let us now define a set of partitions Mk of I by setting their sizes Nk . Take N0 = ¯ i , then Nk = (N0 )k for k = 1, . . . ∞. It is clear that these partitions satisfy p li=1 Conditions 1. For each partition Mk denote by B¯ k , BkBD the corresponding evolution operators for the map T¯ p and TBD respectively. Note that both B¯ k and BkBD are quantizable i.e., one can find unitary matrices U¯ k , UkBD satisfying (6). Indeed, this is completely obvious for B¯ k as T¯ p has the uniform slope. Since BkBD is of the block diagonal form, the corresponding quantum evolution UkBD can be defined as the block diagonal matrix of the same structure where each block is quantized with the help of the discrete Fourier transform. Given matrices B¯ k , BkBD , and the quantizations U¯ k , UkBD one can easily construct the transfer operator for the composition map T = T¯ p ◦ TBD and the corresponding quantization. Lemma 3. Let T , Mk be the map and partition as above and let Bk be the corresponding evolution operator, then Bk = BkBD B¯ k and the matrix Uk = U¯ k UkBD satisfies (6). Proof. Straightforward check. From this the proof of the theorem follows immediately.
340
B. Gutkin
It worth mentioning that using the above quantization procedure it is also possible to obtain the tensorial quantization of the maps T p . For instance, the map T{2,4,4} can be decomposed into the uniformly expanding T¯2 and the map TBD consisting of two “blocks” with the slope two and one, respectively. Correspondingly, the tensorial quantization (67) ¯ BD , where U¯ , U BD = P0 + U¯ ∗ U¯ 1 U¯ P1 of T{2,4,4} can be cast in the form: U = UU give the “uniform” and the “block-diagonal” components.
Appendix B: Proof of Eq. (62) Let T , T¯ be as in Sect. 7.2 and T˜ be the corresponding tower map given by (58). From the Markov partition of I : {ε0 , ε0 ∈ {1, 2}} one can easily construct the Markov partition of I˜: {ε0 × {η}, ε0 ∈ {1, 2} and η ∈ {0, 1}}. The corresponding n-times refined (with respect to T˜ ) partition is given then by the set of cylinders: {˜ε, ε˜ = ε˜ 0 . . . ε˜ n−1 }, where ε˜ i = (εi , ηi ), εi ∈ {1, 2} and ηi ∈ {0, 1}}. The metric entropy HKS (T˜ , µ) ˜ is determined by the corresponding limit of the entropy function: ˜ =− h n (µ)
µ(˜ ˜ ε) log µ(˜ ˜ ε).
(113)
|˜ε |=n
For a cylinder ˜ε let ε = π I ˜ε be the corresponding cylinder in I containing exactly the same sequence of ε as in ε˜ . Note that the time evolution of any point ζ˜ ∈ I˜ is completely determined by the sequence ε and the initial level η0 . Therefore, for a given ε there are precisely two non-empty cylinders ˜ε, ˜ε such that π I ˜ε = π I ˜ε = ε. Furthermore, µ(˜ ˜ ε) = −1 µ(ε), µ(˜ ˜ ε ) = −1 µ(1ε) and h n (µ) ˜ can be rewritten as: h n (µ) ˜ = − −1
µ(ε) log µ(ε) −1 + µ(1ε) log µ(1ε) −1 .
|ε|=n
On the other hand, the entropy of the measure µ¯ is given by h n (µ) ¯ = −
−1
µ(ε) ¯ + µ(1ε) ¯ . µ(ε) ¯ + µ(1ε) ¯ log
|ε|=n
It remains to see that two limits limn→∞ h n (µ)/n, ˜ limn→∞ h n (µ)/n ¯ coincide. By the convexity of the entropy function ¯ ≥ h n (µ) ˜ + log 2. h n (µ)
(114)
Since, log(x + y) ≥ log x one also has: ¯ ≤ h n (µ). ˜ h n (µ) From (114, 115) the claim immediately follows.
(115)
Entropic Bounds on Semiclassical Measures for Quantized 1-D Maps
341
References 1. Bohigas, O.: Random matrix theory and chaotic dynamics. In: Giannoni, M.J., Voros, A., Zinn-Justin, J., eds., Chaos et physique quantique, (École d’été des Houches, Session LII, 1989), Amsterdam: North Holland, 1991 2. Berry, M.V.: Regular and irregular semiclassical wave functions. J. Phys. A 10, 2083–2091 (1977) 3. Voros, A.: Semiclassical ergodicity of quantum eigenstates in the Wigner representation. In: Stochastic Behavior in Classical and Quantum Hamiltonian Systems, Casati, G., Ford, J., eds., Proceedings of the Volta Memorial Conference, Como, Italy, 1977, Lecture Notes in Phys. 93, Berlin: Springer, 1979, pp. 326–333 4. Lazutkin, V.F.: Semiclassical asymptotics of eigenfunctions. In: Partial Differential Equations V, Berlin: Springer, 1999 5. Schnirelman, A.I.: Ergodic properties of eigenfunctions. Usp. Mat. Nauk 29, no. 6 (180), 181–182 (1974) 6. Zelditch, S.: Uniform distribution of the eigenfunctions on compact hyperbolic surfaces. Duke Math. J. 55, 919–941 (1987) 7. Colinde Verdière, Y.: Ergodicité et fonctions propres du laplacien. Commun. Math. Phys. 102, 497–502 (1985) 8. Gérard, P., Leichtnam, É.: Ergodic properties of eigenfunctions for the Dirichlet problem. Duke Math. J. 71(2), 559–607 (1993) 9. Zworski, M., Zelditch, S.: Ergodicity of eigenfunctions for ergodic billiards. Commun. Math. Phys. 175, 673–682 (1996) 10. Bouzouina, A., De Bièvre, S.: Equipartition of the eigenfunctions of quantized ergodic maps on the torus. Commun. Math. Phys. 178, 83–105 (1996) 11. Helffer, B., Martinez, A., Robert, D.: Ergodicité et limite semi-classique. Commun. Math. Phys. 109, 313–326 (1987) 12. Rudnick, Z., Sarnak, P.: The behavior of eigenstates of arithmetic hyperbolic manifolds. Commun. Math. Phys. 161, 195–213 (1994) 13. Lindenstrauss, E.: Invariant measures and arithmetic quantum unique ergodicity. Ann. Math. 163, 165–219 (2006) 14. Hassell, A.: Ergodic billiards that are not quantum unique ergodic, with an appendix by A. Hassell, L. Hillairet. Preprint (2008) http://arxiv.org/abs/0807.0666v3[math,AP], 2008, to appear in Ann. of Math 15. Faure, F., Nonnenmacher, S., De Bièvre, S.: Scarred eigenstates for quantum cat maps of minimal periods. Commun. Math. Phys. 239, 449–492 (2003) 16. Faure, F., Nonnenmacher, S.: On the maximal scarring for quantum cat map eigenstates. Commun. Math. Phys. 245, 201–214 (2004) 17. Anantharaman, N., Nonnenmacher, S.: Entropy of semiclassical measures of the Walsh-quantized baker’s map. Ann. H. Poincaré 8, 37–74 (2007) 18. Kelmer, D.: Arithmetic quantum unique ergodicity for symplectic linear maps of the multidimensional torus. Preprint (2005), http://arxiv.org/abs/math-ph/0510079v5, 2007, to appear in Ann. of Math. 19. Anantharaman, N.: Entropy and the localization of eigenfunctions. Ann. of Math. 168(2), 435–475 (2008) 20. Anantharaman, N., Nonnenmacher, S.: Half–delocalization of eigenfunctions of the Laplacian on an Anosov manifold. Ann. de l’Inst. Fourier 57(7), 2465–2523 (2007) 21. Anantharaman, N., Nonnenmacher, S., Koch, H.: Entropy of eigenfunctions. http://arXiv.org/abs/0704. 1564v1[math-ph], 2007 ˙ 22. Pako´nski, P., Zyczkowski, K., Ku´s, M.: Classical 1D maps, quantum graphs and ensembles of unitary matrices. J. Phys. A 34(43), 9303–9317 (2001) 23. Berkolaiko, G., Keating, J.K., Smilansky, U.: Quantum Ergodicity for Graphs Related to Interval Maps. Commun. Math. Phys. 273, 137–159 (2007) 24. Zyczkowski, K., Ku´s, M., Słomczy´nski, W., Sommers, H.-J.: Random unistochastic matrices. J. Phys. A 36(12), 3425–3450 (2003) 25. Keller, G.: Equilibrium States in Ergodic Theory. London Mathematical Society Student Texts 42 Cambridge: Cambridge University Press, 1998 26. De Bièvre, S.: Quantum chaos: a brief first visit. In: Second Summer School in Analysis and Mathematical Physics (Cuernavaca, 2000), Vol. 289 of Contemp. Math., Providence, RI: Amer. Math. Soc., 2001, pp. 161–218 27. Deutsch, D.: Uncertainty in quantum measurements. Phys. Rev. Lett. 50, 631–633 (1983) 28. Kraus, K.: Complementary observables and uncertainty relations. Phys. Rev. D 35, 3070–3075 (1987) 29. Maassen, H., Uffink, J.B.M.: Generalized entropic uncertainty relations. Phys. Rev. Lett. 60, 1103–1106 (1988) 30. Baladi, V.: Positive Transfer Operators and Decay of Correlations. Advanced Series in Nonlinear Dynamics, Vol 16, Singapore: World Scientific, 2000
342
B. Gutkin
31. Luzzatto, S.: Stochastic-like behavior in non-uniformly expanding maps. In: Handbook of Dynamical Systems, Vol. 1B, B. Hasselblatt and A. Katok (eds.), London: Elsevier, 2006, pp. 265–326 32. Denker, M., Holzmann, H.: Markov partitions for fibre expanding systems. Colloq. Math. 110, 485–492 (2008) 33. Gutkin, B.: Quantum towers and entropic bounds on semiclassical measures. In preparation 34. Riviere, G.: Entropy of semiclassical measures in dimension 2. http://arXiv.org/abs/0809.0230v2[mathph], 2008 35. Nonnenmacher, S., Rubin, M.: Resonant eigenstates for a quantized chaotic system. Nonlinearity 20, 1387–1420 (2007) Communicated by S. Zelditch
Commun. Math. Phys. 294, 343–352 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0967-1
Communications in
Mathematical Physics
The Vanishing of Two-Point Functions for Three-Loop Superstring Scattering Amplitudes Samuel Grushevsky1 , Riccardo Salvati Manni2 1 Mathematics Department, Princeton University, Fine Hall, Washington Road,
Princeton, NJ 08544, USA. E-mail:
[email protected]
2 Dipartimento di Matematica, Università “La Sapienza”,
Piazzale A. Moro 2, Roma, I 00185, Italy. E-mail:
[email protected] Received: 16 June 2008 / Accepted: 18 December 2008 Published online: 12 December 2009 – © Springer-Verlag 2009
Abstract: In this paper we show that the two-point function for the three-loop chiral superstring measure ansatz proposed by Cacciatori, Dalla Piazza, and van Geemen [2] vanishes. Our proof uses the reformulation of the ansatz given in [8], theta functions, and specifically the theory of the 00 linear system, introduced by van Geemen and van der Geer [6], on Jacobians. At the two-loop level, where the amplitudes were computed by D’Hoker and Phong [11–14,17,18], we give a new proof of the vanishing of the two-point function (which was proven by them). We also discuss the possible approaches to proving the vanishing of the two-point function for the proposed ansatz in higher genera [3,8,25]. 1. Introduction An investigation of the problem of computing the superstring measure explicitly for arbitrary genus of the worldsheet was begun by the work of Green and Schwarz [7], who gave an explicit formula in genus 1 using operator methods. D’Hoker and Phong in a series of papers [11–14] introduced a gauge-fixing procedure and computed from first principles the genus 2 superstring measure, verifying that it satisfied the physical constraints, e.g. the vanishing of the 1,2,3-point functions. They also proposed in [15,16] to search for an ansatz for the superstring measure in arbitrary genus as the product of the bosonic measure and a modular form. The ansatz for three-loop measure in this form was then proposed by Cacciatori, Dalla Piazza, and van Geemen in [2]. The genus g ≤ 3 ansatze were reformulated in terms of syzygetic subspaces by the first author in [8], where an ansatz for general genus was proposed, under the assumption on holomorphicity of certain 2r -roots. Cacciatori, Dalla Piazza, and van Geemen in [3] give the genus 4 ansatz in terms of quadrics in the theta constants. The second author in [25] showed that the proposed ansatz is holomorphic in genus 5. Dalla Piazza and van Geemen in [4] proved the uniqueness of the Research is supported in part by National Science Foundation under the grant DMS-05-55867.
344
S. Grushevsky, R. Salvati Manni
modular form in genus 3 satisfying the factorization constraints. Morozov in [23] surveyed this work and gave an alternative proof that factorization constraints are satisfied for the ansatz; in [24] he has also investigated the 1,2,3-point functions of the proposed ansatz, proving under certain non-trivial mathematical assumption that they vanish on the hyperelliptic locus. In this paper we use the techniques of theta functions, and especially the 00 sublinear system of the linear system |2| introduced by van Geemen and van der Geer [6] to prove the vanishing of the 2-point function in genus 3. We also obtain a new proof of the vanishing of the 2-point function in genus 2. 2. Notations and Definitions We denote by Ag the moduli space of complex principally polarized abelian varieties (ppav for short) of dimension g, and by Hg the Siegel upper half-space of symmetric complex matrices with positive-definite imaginary part, called period matrices. The space Hg is the universal cover of Ag , with the deck group Sp(2g, Z), so that we have Ag = Hg / Sp(2g, Z) for a certain action of the symplectic group. A function f : Hg → C is called a (scalar) modular form of weight k with respect to a subgroup ⊂ Sp(2g, Z) if f (γ ◦ τ ) = det(Cτ + D)k f (τ )
∀γ ∈ , ∀τ ∈ Hg ,
where C and D are the lower blocks if we write γ as four g × g blocks. For a period matrix τ ∈ Hg the principal polarization τ on the abelian variety Aτ := Cg /(Zg + τ Zg ) is the divisor of the theta function exp(πi(n t τ n + 2n t z)). θ (τ, z) := n∈Zg
Notice that for fixed τ theta is a function of z ∈ Cg , and its automorphy properties under the lattice Zg + τ Zg define the bundle τ . Given a point of order two on Aτ , which can be uniquely represented as τ ε+δ 2 for g ε, δ ∈ Z2 (where Z2 = {0, 1} is the additive group), the associated theta function with characteristic is ε θ (τ, z) := exp(πi((n + ε)t τ (n + ε) + 2(n + ε)t (z + δ)). δ n∈Zg
ε is odd or even depending on whether the scalar product δ ε · δ ∈ Z2 is equal to 1 or 0, respectively. The theta function with characteristic is the generator of the space of sections of the bundle τ + τ ε+δ 2 (where we have implicitly identified the principally polarized abelian variety with its dual, and think of points as bundles of degree 0). Thus the square of any theta function with characteristic is a section of 2τ , and the basis for the space of sections of this bundle is given by theta functions of the second order ε (2τ, 2z) [ε](τ, z) := θ 0 As a function of z, θ
The 2-Point Function for 3-Loop Superstrings
345
g
for all ε ∈ Z2 . Riemann’s addition formula is an explicit expression of the squares of theta functions with characteristics in this basis: ε θ (τ, z)2 = (−1)δ·σ [σ ](τ, 0)[σ + ε](τ, z). (1) δ g σ ∈Z2
Theta constants are restrictions of theta functions to z = 0; thus all theta constants with odd characteristics vanish identically in τ , while theta constants with even characteristics and all theta constants of the second order do not vanish identically. All theta constants with characteristics are modular forms of weight one half with respect to a certain normal subgroup of finite index (4, 8) ⊂ Sp(2g, Z), while all theta constants of the second order are modular forms of weight one half with respect to a bigger normal subgroup (2, 4) ⊃ (4, 8). Theta constants with characteristics are not algebraically independent, and satisfy a host of algebraic identities, some of which follow from Riemann’s addition formula. However, the theta constants of the second order are algebraically independent for g = 1, 2, and the only relation among them in genus 3 is of degree 16, and has been known classically. It is discussed in detail in [6] — here we give the explicit formula for easy reference. Indeed, a special case of Riemann’s quartic addition theorem in genus 3 is the following identity for theta constants (where we suppress the argument τ ): θ
000 000 000 000 θ θ θ 110 010 100 000 001 001 001 000 000 000 000 001 θ θ . =θ θ θ θ +θ θ 100 000 111 011 101 001 110 010
If we denote the three terms in this relation by ri , so that the relation is r1 = r2 + r3 , then multiplying the 4 “conjugate” relations r1 = ±r2 ± r3 yields the identity F := r14 + r24 + r34 − 2r12 r22 − 2r22 r32 − 2r32 r12 = 0.
(2)
Notice that F is a polynomial of degree 8 in the squares of theta constants with characteristics, and thus by applying Riemann’s addition formula (1) F can be rewritten as a polynomial of degree 16 in theta constants of the second order. We refer to [1,19] for details on theta functions and modular forms, and the current knowledge about the ideal of relations among theta constants of the second order for g > 3 (which is not known completely even for g = 4). 3. The Linear System 00 In this section we review the definition and some facts about the linear system 00 ⊂ |2| introduced and studied in [6]. We refer to that paper for details, as well as to [5,9,20] for a review and results on the importance of the linear system 00 for the Schottky problem of characterizing Jacobians. The linear system 00 ⊂ |2| is defined to consist of all sections vanishing to order at least four at the origin. Since all sections of 2 are even, this is equivalent to the value and all the second derivatives ∂zi ∂z j vanishing at zero. These conditions turn out be
346
S. Grushevsky, R. Salvati Manni
independent, when (Aτ , ) is an indecomposable ppav (i.e. not isomorphic to a product of lower-dimensional ppavs). In this case the matrix ⎛ ⎞ ](τ,0) ](τ,0) [ε1 ](τ, 0) ∂[ε∂τ111 . . . ∂[ε∂τ1gg ⎜ ⎟ .. .. .. .. ⎜ ⎟ . . . . ⎝ ⎠ ∂[ε2g ](τ,0) ∂[ε2g ](τ,0) g [ε2 ](τ, 0) ... ∂τ11 ∂τgg has rank
g(g+1) 2
+ 1, cf. [26] and thus ([6], Prop. 1.1)
dim 00 = dim |2| − 1 −
1 = 2g − 1 −
1≤i≤ j≤g
g(g + 1) . 2
(3)
Thus the linear system 00 is zero for g ≤ 2, has dimension 1 for g = 3, and higher dimension for all other genera. The above description leads to a simple construction of a basis for the space 00 . Proposition 1. Let τ0 be an irreducible point of Hg (i.e. corresponding to indecomposg able ppav). Denote N := 1 + g(g+1) 2 , and choose ε1 , . . . , ε N ∈ Z2 such that the modular form ⎞ ⎛ ](τ,0) ](τ,0) . . . ∂[ε∂τ1gg [ε1 ](τ, 0) ∂[ε∂τ111 ⎟ ⎜ .. .. .. ⎟ gε1 ,...,ε N (τ ) := det ⎜ . . . ⎠ ⎝ ∂[ε N ](τ,0) ∂[ε N ](τ,0) ... [ε N ](τ, 0) ∂τ11 ∂τgg does not vanish at τ0 . Then the sections ⎛ [ε1 ](τ0 , z) [ε1 ](τ0 , 0) ⎜ .. .. ⎜ ⎜ . . f ε (τ0 , z) := det ⎜ ⎜[ε N ](τ0 , z) [ε N ](τ0 , 0) ⎝ [ε](τ0 , z) [ε](τ0 , 0)
∂[ε1 ](τ0 ,0) ∂τ11
.. .
∂[ε N ](τ0 ,0) ∂τ0 τ11 ∂[ε](τ0 ,0) ∂τ11
... .. . ... ...
∂[ε1 ](τ0 ,0) ∂τgg
⎞
⎟ ⎟ ⎟ , ∂[ε N ](τ0 ,0) ⎟ ⎟ ∂τgg ⎠ ∂[ε](τ0 ,0) ∂τgg
g
for ε ∈ Z2 \ {ε1 , . . . , ε N } form a basis of 00 ⊂ |2τ0 |. Proof. The proof is a simple linear algebra argument that we recall for completeness. First note that f ε (τ0 , z) belongs to 00 , as the determinant and all the second z-derivatives (equal to the first τ -derivatives by the heat equation) vanish for z = 0, as two of the columns of the matrix become identical. It thus remains to show that the functions f ε for various ε are linearly independent. Indeed, recall that theta functions of the second order form a basis of sections of 2, and now note that the basis element [ε](τ0 , z) enters only the expression of f ε (τ0 , z), and that with non-zero coefficient gε1 ,...,ε N (τ0 ). Remark 2. It can be shown that on the open sets {gε1 ,...,ε N (τ ) = 0} the coefficients of the basis vectors are in fact modular forms of weight g + 1 + N /2, see [10]. In particular when g = 3 we have a global expression of the unique section f (τ, z) of the space 00 . Remark 3. Observe that if the period matrix τ is decomposable, then the dimension of 00 increases; however, a basis can still be constructed by using the same method.
The 2-Point Function for 3-Loop Superstrings
347
There exists another method for constructing elements of 00 — it is described in [6], and is as follows. Suppose I is an algebraic relation among theta constants of the second order (in genus g). This is to say, suppose I ∈ C[x0...0 , . . . , x1...1 ] is a polynomial in 2g variables such that for any τ ∈ Hg we have I ([ε](τ )) = 0. Then the function f I (z) :=
∂I ([0 . . . 0](τ, 0), . . . , [1 . . . 1](τ, 0)) [ε](τ, z) g ∂xε
ε∈Z2
lies in 00 ⊂ |2τ |. Indeed, since I vanishes identically on Hg , by Euler’s formula we have f I (0) = 0. Moreover, by the heat equation 2πi(1 + δ j,k )
∂ I ∂[ε](τ ) ∂2 fI ∂ I ([0 . . . 0], . . . [1 . . . 1]) |z=0 = = , ∂z j ∂z k ∂x ∂τ ∂τ jk ε jk g ε∈Z2
which is zero since I vanishes identically on Hg , and thus its derivative in any direction is also zero. In [6], Prop. 1.2 it is shown that as I ranges over the ideal of relations among theta constants, the functions f I generate the linear system 00 . Since for g ≥ 4 the ideal of algebraic relations among theta constants of the second order is not completely known, for g ≥ 4 this method does not yield a complete description of the basis of 00 . However, the geometry of these relations is intriguing, and this method produces elements of 00 with coefficients algebraic in theta constants, rather than involving their derivatives as well. 4. The Proposed Ansatz for the Superstring Measure An ansatz for the 3-loop superstring measure was proposed in [2]. The reformulation of this ansatz in terms of products of theta constants with characteristics in a syzygetic subspace given in [8] is as follows. For any i = 0 . . . g define (g) Gi
ε (τ ) := δ
2g V ⊂Z2 ; dim V =i
α β
4−i ε+α θ (τ )2 . δ+β
(4)
∈V
Noticethat since any i-dimensional linear subspace contains zero, all products will conε . Since all odd theta constants vanish identically, it is enough to sum over the tain θ δ even cosets of syzygetic i-dimensional subspaces containing [ε, δ], see [8,25]. 2g To simplify notations, we write m := [ε, δ] ∈ Z2 for characteristics and similarly ε . Then the proposed ansatz for the superstring measure is the product write θm := θ δ of the bosonic measure (which is a form on Mg ) and, for any even characteristic m, the expression (g) m :=
g i(i−1) (g) (−1)i 2 2 G i [m] i=0
(5)
348
S. Grushevsky, R. Salvati Manni
which is a modular form of weight 8 with respect to a subgroup of Sp(2g, Z) conjugate to (1, 2). In particular for genus 3 we have (3)
(3)
(3)
(3)
(3) m := G 0 [m] − G 1 [m] + 2G 2 [m] − 8G 3 [m].
(3) In [25] it is shown that the sum m m is a non-zero multiple of the modular form F given by (2), and thus vanishes identically on H3 . (g) (g) From definition (4) of the summands G i [m] of the measure m it follows that (g) G i [m] is a polynomial in the squares of theta constants with characteristics for i ≤ 3, divisible by θm2 (τ ). Since this is the only kind of summands appearing in the definition (g) of m for g ≤ 3, by applying Riemann’s addition formula (1) we get (g)
(g)
Proposition 4. For g ≤ 3 the modular form m defined by (5) and the ratio m /θm2 (τ, 0) are both polynomials in theta constants of the second order, of degrees 16 and 14, respectively. 5. The Vanishing of the 2-Point Function We recall (see [12] for explicit formulas) that the vanishing of the cosmological constant
(g) reduces to the identity m m = 0 (proven for the proposed ansatz for g ≤ 4 in [25]), and this also implies the vanishing of the 1-point function, while as shown in [18] the vanishing of the two-point function is equivalent to the vanishing of 2 (g) m Sm (a, b) m
for any points a, b on the Riemann surface (thought of as embedded into its Jacobian), where Sm is the Szeg˝o kernel Sm (a, b) :=
θm (a − b) , θm (0)E(a, b)
with E being the prime form on the Riemann surface. Since the prime form does not depend on m, it is a common factor in all summands above, and thus does not matter for the vanishing of the 2-point function, so the vanishing of the 2-point function is equivalent to the vanishing of X 2 (a, b) :=
(g) m (τ ) θ (τ, a − b)2 , 2 (τ, 0) m θ m m
where τ is the period matrix of the Jacobian J ac(C) of a Riemann surface C, and a, b ∈ C ⊂ J ac(C) are arbitrary. We will now relate the vanishing of the 2-point function and the 00 linear system. Set m (τ ) θ (τ, z)2 (6) X 2 (τ, z) := 2 (τ, 0) m θ m m and note that for τ fixed this function is a section of |2τ |. The vanishing of the 2-point function is then equivalent to X 2 (z) vanishing along the surface C − C ⊂ J ac(C). By Proposition 2.1 in [6] and the subsequent remark, a section of |2| vanishes along the surface C − C if and only if it lies in 00 . We thus get
The 2-Point Function for 3-Loop Superstrings
349
Theorem 5. The 2-point function for the proposed superstring measure ansatz vanishes (for genus g) if and only if for the period matrix τ of any Jacobian of a Riemann surface of genus g the section X 2 (τ, z) of 2τ defined above lies in 00 . Since for g = 1, 2 the linear system 00 is zero, the vanishing of the two-point function is equivalent to X 2 (τ, z) vanishing identically in z and τ for g ≤ 2. If we write out X 2 (τ, z) as a linear combination of the basis for sections of 2τ given by theta functions of the second order X 2 (τ, z) = cε (τ )[ε](τ, z), (7) then X 2 vanishes identically if and only if each cε (τ ) vanishes identically. This allows us to recover the result of D’Hoker and Phong in genus 2. Proposition 6. The 2-point function for the proposed superstring ansatz vanishes identically for g ≤ 2. Proof. Let us apply Riemann’s addition formula (1) to the definition (6) of the two-point function to rewrite it in terms of theta functions of the second order (notice that the only term depending on z is θm (τ, z)2 , and we apply the addition formula to it as well). Notice that by Proposition 4 the coefficients cε in (7) obtained in this way are explicit polynomials of degree 15 in theta constants of the second order, and since there are no algebraic relations among theta constants of the second order for g ≤ 2, one needs to verify that all polynomials cε are zero. This can be done on a computer (we used Maple). Note that the computation can be made easier by noting that since X 2 (τ, z) has a transformation formula with respect to the entire symplectic group, (i.e. it is a Jacobi form) and the coefficients cε (τ ) are permuted under the action of a suitable subgroup Sp(2g, Z) that acts monomially on the theta constants of the second order, it is enough to check that just one of cε is the zero polynomial. In the case of g = 3, recall from (3) that the space 00 is one-dimensional, and by the results of [6] we know that it is generated by F2 :=
∂F ([000](τ, 0), . . . , [111](τ, 0)) [ε](τ, z), ∂xε 3
ε∈Z2
where we recall that F, given by (2), is the only polynomial relation of degree 16 among the 8 theta constants of the second order for g = 3. Proposition 7. For any τ ∈ H3 the sections F2 and X 2 of 2τ are proportional; more precisely F2 = − 14 5 X 2. Proof. We have explicit expressions for F2 and X 2 as linear combinations of the basis of the sections of 2 given by the second order theta functions. Thus what we need to verify is that the coefficient in 5F2 + 14X 2 of any [ε](z) is equal to zero. This coefficient is a polynomial of degree 15 in theta constants of the second order and can be verified to be zero using Maple (since the only relation among theta constants of the second order is of degree 16, a polynomial in theta constants of the second order of degree 15 vanishes identically only if it is zero). Notice that by modularity it is again enough to verify that the coefficient of [000](z) in 5F2 + 14X 2 is equal to zero.
350
S. Grushevsky, R. Salvati Manni
Theorem 8. The 2-point function for the proposed ansatz for the 3-loop superstring measure vanishes identically. Proof. By the above proposition we see that for any τ ∈ H3 the function X 2 , being a constant multiple of F2 , lies in the linear system 00 ⊂ |2τ |. By Theorem 5 this is equivalent to the identical vanishing of the two-point function. Remark 9. We note that the global section F2 is proportional also to the global section f (τ, z). Really we have that F2 (τ, z) = c(τ ) f (τ, z) for any irreducible τ ∈ Hg . Moreover c(τ ) results to be a modular function with respect to Sp(6, Z) that is regular on the set of irreducible point, so it is regular everywhere (modular form) and hence it is a non-zero constant. This identity produces eight nontrivial identities expressing each jacobian determinant gε1 ,...,ε7 (τ ) as a polynomial of degree 15 in the theta constants [σ ](τ, 0). 6. Conclusion There are two generalizations that it is natural to try to prove. First, one could ask whether the vanishing of the 2-point function can be obtained for the proposed in [8] ansatz in higher genera. For genus 4 the ansatz is also given in [3] and is manifestly holomorphic in either formulation. The holomorphicity of the ansatz in genus 5 was proven in [25], and thus it is natural to ask whether the 2-point function vanishes for g ≤ 5. By Theorem 5 we know that this is equivalent to X 2 lying in the linear system 00 . However, already for genus 4 the geometry of the situation is much more complicated: instead of just one relation F in genus 3 the ideal of relations among theta constants of the second order in genus 4 is unknown, and an explicit basis for 00 is unknown for g = 4. Moreover, it could be that here the fact that we are working on the moduli space of curves M4 rather than A4 plays a role — the geometry of 00 depends on this, see [20]. Second, one could try to prove the vanishing of the 3-point function. As shown in [18] for genus 2, this is equivalent to proving that the sum (g) m Sm (a, b)Sm (b, c)Sm (c, a) m
vanishes. Using the explicit formula for the Szeg˝o kernel and canceling the m-independent factor, this is equivalent to the function X 3 (a, b, c) :=
m (τ ) θm (a − b)θm (b − c)θm (c − a) θm3 (τ, 0) m
vanishing identically for a, b, c ∈ C. However, in this case we do not know a natural function on J ac(C)×n of which X 3 is a restriction, and there is no analog of the theory of the 00 for more points. It seems that the identity among the third order theta functions obtained by Krichever in his proof of the trisecant conjecture ([21], formula (1.18)) may potentially be useful in relating the 3-point and 2-point functions, but so far we have not been able to find an explicit way to do this.
The 2-Point Function for 3-Loop Superstrings
351
Added in proof. The recent paper of Matone and Volpato [22], written after the current manuscript was submitted for publication, uses other identities relating theta functions and abelian differentials to show that the expression X 3 above in fact does not vanish identically in genus 3. Note, however, that as shown by D’Hoker and Phong in their series of papers, the full expression for N -point function involves many more terms resulting from gauge-fixing, and thus the result of [22] suggests that these terms have a non-zero contribution starting from the 3-point function in genus 3. Acknowledgements. We are grateful to Eric D’Hoker and Duong Phong for introducing us to questions about the superstring scattering amplitudes and explanations regarding the conjectured properties of N -point functions. The computations for this paper were done using Maplesoft’s Maple© software. We would like to thank the referee.
References 1. Birkenhake, Ch., Lange, H.: Complex abelian varieties: second, augmented edition. Grundlehren der mathematischen Wissenschaften 302, Berlin: Springer-Verlag, 2004 2. Cacciatori, S.L., Dalla Piazza, F., van Geemen, B.: Modular forms and three loop superstring amplitudes. Nucl. Phys. B. 800, 565–590 (2008) 3. Cacciatori, S.L., Dalla Piazza, F., van Geemen, B.: Genus four superstring measures. Lett. Math. Phys. 85, 185–193 (2008) 4. Dalla Piazza, F., van Geemen, B.: Siegel modular forms and finite symplectic groups. http://arXiv.org/ abs/0804.3769v2[math.AG], 2008 5. van Geemen, B.: The Schottky problem and second order theta functions. In: Taller de variedades abelianas y funciones theta, Sociedad Matemática Mexicana, Aportaciones Matemáticas, Investigación 13, 41–84 (1998) 6. van Geemen, B., van der Geer, G.: Kummer varieties and the moduli spaces of abelian varieties. Amer. J. of Math. 108, 615–642 (1986) 7. Green, M.B., Schwarz, J.H.: Supersymmetrical string theories. Phys. Lett. B 109, 444–448 (1982) 8. Grushevsky, S.: Superstring scattering amplitudes in higher genus. Commun. Math. Phys. 287, 749–767 (2009) 9. Grushevsky, S.: A special case of the 00 conjecture. http://arXiv.org/abs/0804.0525v2[math.AG], 2009 10. Grushevsky, S., Salvati Manni, R.: Two generalizations of Jacobi’s derivative formula. Math. Res. Lett. 12(5-6), 921–932 (2005) 11. D’Hoker, E., Phong, D.H.: Two-loop superstrings I, main formulas. Phys. Lett. B 529, 241–255 (2002) 12. D’Hoker, E., Phong, D.H.: Two-loop superstrings II, the chiral measure on moduli space. Nucl. Phys. B 636, 3–60 (2002) 13. D’Hoker, E., Phong, D.H.: Two-loop superstrings III, slice independence and absence of ambiguities. Nucl. Phys. B 636, 61–79 (2002) 14. D’Hoker, E., Phong, D.H.: Two-loop superstrings IV, The cosmological constant and modular forms. Nucl. Phys. B 639, 129–181 (2002) 15. D’Hoker, E., Phong, D.H.: Asyzygies, modular forms, and the superstring measure I. Nucl. Phys. B 710, 58–82 (2005) 16. D’Hoker, E., Phong, D.H.: Asyzygies, modular forms, and the superstring measure. II. Nucl. Phys. B 710, 83–116 (2005) 17. D’Hoker, E., Phong, D.H.: Two-loop superstrings V, Gauge slice independence of the N-point function. Nucl. Phys. B 715, 91–119 (2005) 18. D’Hoker, E., Phong, D.H.: Two-loop superstrings VI, non-renormalization theorems and the 4-point function. Nucl. Phys. B 715, 3–90 (2005) 19. Igusa, J.-I.: Theta functions. Die Grundlehren der mathematischen Wissenschaften, Band 194. New YorkHeidelberg: Springer-Verlag, 1972 20. Izadi, E.: The geometric structure of A4 , the structure of the Prym map, double solids and 00 -divisors. J. Reine Angew. Math. 462, 93–158 (1995) 21. Krichever, I.: Characterizing Jacobians via trisecants of the Kummer Variety. Ann. of Math., to appear, http://arXiv.org/abs/math/0605625v4[math.AG], 2008 22. Matone, M., Volpato, R.: Superstring measure and non-renormalization of the three-point amplitude. Nucl. Phys. B 806, 735–747 (2009)
352
S. Grushevsky, R. Salvati Manni
23. Morozov, A.: NSR superstring measures revisited. JHEP 0805, 086 (2008) 24. Morozov, A.: NSR measures on hyperelliptic locus and non-renormalization of 1,2,3-point functions. Phys. Lett. B 664, 116–122 (2008) 25. Salvati Manni, R.: Remarks on Superstring amplitudes in higher genus. Nucl. Phys. B 801, 163–173 (2008) 26. Sasaki, R.: Modular forms vanishing at the reducible points of the Siegel upper-half space. J. Reine Angew. Math. 345, 111–121 (1983) Communicated by N.A. Nekrasov
Commun. Math. Phys. 294, 353–388 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0941-y
Communications in
Mathematical Physics
Escape Rates and Physically Relevant Measures for Billiards with Small Holes Mark Demers1, , Paul Wright2, , Lai-Sang Young3, 1 Department of Mathematics and Computer Science, Fairfield University,
Fairfield, USA. E-mail:
[email protected]
2 Department of Mathematics, University of Maryland,
College Park, USA. E-mail:
[email protected]
3 Courant Institute of Mathematical Sciences, New York University,
New York, USA. E-mail:
[email protected] Received: 4 November 2008 / Accepted: 3 April 2009 Published online: 25 November 2009 – © Springer-Verlag 2009
Abstract: We study the billiard map corresponding to a periodic Lorentz gas in 2-dimensions in the presence of small holes in the table. We allow holes in the form of open sets away from the scatterers as well as segments on the boundaries of the scatterers. For a large class of smooth initial distributions, we establish the existence of a common escape rate and normalized limiting distribution. This limiting distribution is conditionally invariant and is the natural analogue of the SRB measure of a closed system. Finally, we prove that as the size of the hole tends to zero, the limiting distribution converges to the smooth invariant measure of the billiard map. This paper is about leaky dynamical systems, or dynamical systems with holes. Consider a dynamical system defined by a map or a flow on a phase space M, and let H ⊂ M be a hole through which orbits escape, that is to say, once an orbit enters H , we stop considering it from that point on. Starting from an initial probability distribution µ0 on M, mass will leak out of the system as it evolves. Let µn denote the distribution remaining at time n. The most basic question one can ask about a leaky system is its rate of escape, i.e. whether µn (M) ∼ ϑ n for some ϑ. Another important question concerns the nature of the remaining distribution. One way to formulate that is to normalize µn , and to inquire about properties of µn /µn (M) as n tends to infinity. Such limiting distributions, when they exist, are not invariant; they are conditionally invariant, meaning they are invariant up to a normalization. Comparisons of systems with small holes with the corresponding closed systems, i.e. systems for which the holes have been plugged, are also natural. These are some of the questions we will address in this paper. We do not consider these questions in the abstract, however; for a review paper in this direction, see [DY]. Our context here is that of billiard systems with small holes.
This research is partially supported by NSF grant DMS-0801139. This research is partially supported by an NSF Mathematical Sciences Postdoctoral Research Fellowship. This research is partially supported by a grant from the NSF.
354
M. Demers, P. Wright, L.-S. Young
Specifically, we carry out our analysis for the collision map of a 2-dimensional periodic Lorentz gas, and expect our results to be extendable to other dispersing billiards. Our holes are “physical” holes, in the sense that they are derived from holes in the physical domain of the system, i.e., the billiard table: we consider both convex holes away from the scatterers and holes that live on the boundaries of the scatterers. The holes considered in this paper are very small, but their placements are immaterial. For these leaky systems, we prove that there is a common rate of escape and a common limiting distribution for a large class of natural initial distributions including those with densities with respect to Liouville measure. These conditionally invariant measures, therefore, can be viewed as characteristic of the leaky systems in question, in a way that is analogous to physical measures or SRB measures for closed systems. We show, in fact, that as hole size tends to zero, these measures tend to the natural invariant measure of the corresponding closed billiard system. Our proof involves constructing a Markov tower extension with a special property over the billiard map, the new requirement being that it respects the hole. Let us backtrack a little for readers not already familiar with these ideas: In much the same way that Markov partitions have proved to be very useful in the study of Anosov and Axiom A diffeomorphisms, it was shown, beginning with [Y] and continued in a number of other papers, that many systems with sufficiently strong hyperbolic properties (but which are not necessarily uniformly hyperbolic) admit countable Markov extensions. Roughly speaking, these extensions behave like countable state Markov chains “with nonlinearity”; they have considerably simpler structures than the original dynamical system. The idea behind this work is that escape dynamics are much simpler in a Markov setting when the hole corresponds to a collection of “states”; this is what we mean by the Markov extension “respecting the hole.” All this is not for free, however. We pay a price with a somewhat elaborate construction of the tower, and again when we pass the information back to the billiard system, in exchange for having a Markov structure to work with in the treatment of the hole. There are advantages to this route of proof: First, once a Markov extension is constructed for a system, it can be used many times over for entirely different purposes. For the billiard maps studied here, these extensions were constructed in [Y]; our main task is to adapt them to holes. Second, once results on escape dynamics are established on towers, they apply to all Markov extensions. Here, the desired results are already known in a special case, namely expanding towers [BDM]; we need to extend them to the general, hyperbolic setting. What we propose here is a unified, generic approach for dealing with holes in dynamical systems, one that can, in principle, be carried out for all systems that admit Markov towers. Such systems include logistic maps, rank one attractors including the Hénon family, piecewise hyperbolic maps and other dispersing billiards in 2 or more dimensions. Conditionally invariant measures were first introduced in probabilistic settings, namely countable state Markov chains and topological Markov chains, beginning with [V] and more recently in [FKMP and CMS3]. In this setting, such measures are called quasi-stationary distributions and the existence of a Yaglom limit corresponds to the limit µn /µn (M), which we use here to identify a physical conditionally invariant measure for the leaky system. The first works to study deterministic systems with holes took advantage of finite Markov partitions. These include: Expanding maps on Rn with holes which are elements of a finite Markov partition [PY,CMS1,CMS2]; Smale horseshoes [C1,C2]; Anosov diffeomorphisms [CM1,CM2,CMT1,CMT2]; billiards with convex scatterers satisfying a
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
355
non-eclipsing condition [LM,R] and large parameter logistic maps whose critical point maps out of the interval [HY]. In the latter two, the holes are chosen in such a way that the surviving dynamics are uniformly expanding or hyperbolic with Markov partitions. First results which drop Markov requirements on the map include piecewise expanding maps of the interval [BaK,CV,LiM,D1,BDM]; Misiurewicz [D2] and Collet-Eckmann [BDM] maps with generic holes; and piecewise uniformly hyperbolic maps [DL]. The tower construction is used in the one-dimensional studies [D1,D2,BDM]. Typically a restriction on the size of the hole is introduced in order to control the dynamics when a finite Markov partition is absent. General conditions ensuring the existence of conditionally invariant measures are first given in [CMM]. The physical relevance of such measures, however, is unclear without further qualifications. As noted in [DY], under very weak assumptions on the dynamical system, many such measures exist: for any prescribed rate of escape, one can construct infinitely many conditionally invariant densities. This is the reason for the emphasis placed in this paper on the limit µn /µn (M), which identifies a unique, physically relevant conditionally invariant measure. This paper is organized as follows: Our results are formulated in Sect. 1. In Sects. 2 and 3, the geometry of billiard maps and holes are looked at carefully as we modify previous constructions to give a generalized horseshoe that respects the hole. Out of this horseshoe, a Markov tower extension is constructed and results on escape dynamics on it proved; this is carried out in Sects. 4 and 5. These results are passed back to the billiard system in Sect. 6, where the remaining theorems are also proved. 1. Formulation of Results 1.1. Basic definitions. We consider a closed dynamical system defined by a self-map f of a manifold M, and let H ⊂ M be a hole through which orbits escape, i.e., we stop considering an orbit once it enters H . In this paper we are primarily concerned with holes that are open subsets of the phase space; they are not too large and generally not f -invariant. We will refer to the triplet ( f, M, H ) as a leaky system. First we introduce some notation. Let M˚ = M\H . At least to begin with, let us make ˚ : M˚ ∩ f −1 M˚ → M, ˚ and a formal distinction between f and f˚ = f |( M˚ ∩ f −1 M) n ˚ Let η be a probability measure on M. ˚ We define f˚∗ η to write f˚n = f n |( i=0 f −i M). ˚ If η be the measure on M˚ defined by ( f˚∗ η)(A) = η( f˚−1 A) for each Borel set A ⊂ M. (n) n n ˚ ˚ ˚ is an initial distribution on M, then η := f ∗ η/| f ∗ η| is the normalized distribution of points remaining in M˚ after n units of time. Given an initial distribution η, the most basic question is the rate at which mass is leaked out of the system. We define the escape rate starting from η to be − log ϑ(η), where n 1 −i ˚ log ϑ(η) = lim log η assuming such a limit exists. f M n→∞ n i=0
Another basic object is the limiting distribution η(∞) defined to be η(∞) = limn→∞ η(n) if this weak limit exists. Of particular interest is when there is a number ϑ∗ and a probability measure µ∗ with the property that for all η in a large class of natural initial distributions (such as those having densities with respect to Lebesgue measure), we have ϑ(η) = ϑ∗ and η(∞) = µ∗ . In such a situation, µ∗ can be thought of as a physical measure for the leaky system ( f, M, H ), in analogy with the idea of physical measures for closed systems.
356
M. Demers, P. Wright, L.-S. Young
A Borel probability measure η on M is said to be conditionally invariant if it satisfies f˚∗ η = ϑη for some ϑ ∈ (0, 1]. Clearly, the escape rate of a conditionally invariant measure η is well defined and is equal to − log ϑ. Most leaky dynamical systems admit many conditionally invariant measures; see [DY]. In particular, limiting distributions, when they exist, are often conditionally invariant; they are among the more important conditionally invariant measures from an observational point of view. Finally, when a physical measure η for a leaky system ( f, M, H ) has absolutely continuous conditional measures on the unstable manifolds of the underlying closed system ( f, M), we will call it an SRB measure for the leaky system, in analogy with the idea of SRB measures for closed systems. 1.2. Setting of present work. The underlying closed dynamical system here is the billiard map associated with a 2-dimensional periodic Lorentz gas. Let {i : i = 1, · · · , d} be pairwise disjoint C 3 simply-connected curves on T2 with strictly positive curvature, and consider the billiard flow on the “table” X = T2 \ i {interiori }. We assume the “finite horizon” condition, which imposes an upper bound on the number of consecutive tangential collisions with ∪i . The phase space of the unit-speed billiard flow is M = (X × S1 )/ ∼ with suitable identifications at the boundary. Let M = ∪i i × [− π2 , π2 ] ⊂ M be the cross-section to the billiard flow corresponding to collision with the scatterers, and let f : M → M be the Poincaré map. The coordinates on M are denoted by (r, ϕ), where r ∈ ∪i is parametrized by arc length and ϕ is the angle a unit tangent vector at r makes with the normal pointing into the domain X . We denote by ν the invariant probability measure induced on M by Liouville measure on M, i.e., dν = c cos ϕdr dϕ, where c is the normalizing constant. We consider the following two types of holes: Holes of Type I. In the table X , a hole σ of this type is an open interval in the boundary of a scatterer. When q0 ∈ ∪i , we refer to {q0 } as an infinitesimal hole, and let h (q0 ) denote the collection of all open intervals σ ⊂ ∪i in the h-neighborhood of q0 . A hole σ in X of this type corresponds to a set Hσ ⊂ M of the form (a, b) × [− π2 , π2 ]. Holes of Type II. A hole σ of this type is an open convex subset of X away from ∪i i and bounded by a C 3 simple closed curve with strictly positive curvature. As above, we regard {q0 } ⊂ X \ ∪i as an infinitesimal hole, and use h (q0 ) to denote the set of all σ in the h-neighborhood of q0 . In this case, σ ⊂ X does not correspond directly to a set in M. Rather, σ corresponds directly to a set in M, the phase space for the billiard flow, and we must make a choice as to which set in the cross section M will represent the hole for the billiard map. There is a well defined set Bσ ⊂ M consisting of all (r, ϕ) whose trajectories under the billiard flow on M will enter σ × S1 before reaching M again. Thus Hσ = f (Bσ ) is a natural candidate for the hole in M representing σ , and will be taken as such in this work. However, it would also have been possible to take Bσ as the representative set. The geometry of Bσ and Hσ in phase space will be discussed in detail in Sect. 3.1. Also, we note that the requirement that ∂σ be a C 3 simple closed curve with strictly positive curvature can be considerably relaxed. It is even possible to allow some holes σ that are not convex. See the remark at the end of Sect. 3.1. 1.3. Statement of results. Let G = G(Hσ ) denote the set of finite Borel measures η on M that are absolutely continuous with respect to ν with dη/dν being (i) Lipschitz on ∞ f −i M. ˚ Notice that each connected component of M and (ii) strictly positive on ∩i=0 measures on M with Lipschitz dη/dν correspond to measures on M having a Lipschitz density with respect to Liouville measure.
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
357
Standing hypotheses for Theorems 1–3. We assume (1) f : M → M is the billiard map defined in Sect. 1.2, (2) {q0 } is an infinitesimal hole of either Type I or Type II, and (3) σ ∈ h (q0 ), where h > 0 is assumed to be sufficiently small. Theorem 1 (Common escape rate). All initial distributions η ∈ G have a common escape rate − log ϑ∗ for some ϑ∗ < 1; more precisely, for all η ∈ G, ϑ(η) is well defined and is equal to ϑ∗ . Theorem 2 (Common limiting distribution). (a) For all η ∈ G, the normalized surviving distributions f˚∗n η/| f˚∗n η| converge weakly to a common conditionally invariant distribution µ∗ with ϑ(µ∗ ) = ϑ∗ . (b) In fact, for all η ∈ G, there is a constant c(η) > 0 s.t. ϑ∗−n f˚∗n η converges weakly to c(η)µ∗ . Thus from an observational point of view, − log ϑ∗ is the escape rate and µ∗ the physical measure for the leaky system ( f, M, Hσ ). Theorem 3 (Geometry of limiting distribution). (a) µ∗ is singular with respect to ν; (b) µ∗ has strictly positive conditional densities on local unstable manifolds. The precise meaning of the statement in part (b) of Theorem 3 is that there are countably many “patches” (Vi , µi ), i = 1, 2, . . ., where for each i, (i) Vi ⊂ M is the union of a continuous family of unstable curves {γ u }; (ii) µi is a measure on Vi whose conditional measures on {γ u } have strictly positive densities with respect to the Riemannian measures on γ u ; (iii) µi ≤ µ∗ for each i, and i µi ≥ µ∗ . This justifies viewing µ∗ as the SRB measure for the leaky system ( f, M, Hσ ). Our final result can be interpreted as a kind of stability for the natural invariant measure ν of the billiard map without holes. Theorem 4 (Small-hole limit). We assume (1) and (2) in the Standing Hypotheses above. Let σh ∈ h (q0 ), h > 0, be an arbitrary family of holes, and let − log ϑ∗ (σh ) and µ∗ (σh ) be the escape rate and physical measure for the leaky system ( f, M, Hσh ). Then ϑ∗ (σh ) → 1 and µ∗ (σh ) → ν as h → 0. Some straightforward generalizations: Our proofs continue to hold under the more general conditions below, but we have elected not to discuss them (or to include them formally in the statement of our theorems) because keeping track of an increased number of objects will necessitate more cumbersome notation. 1. Holes. Our results apply to more general classes of holes than those described above. For example, we could fix a finite number of infinitesimal holes {q0 }, . . . , {qk } and consider σ = ∪i σi with σi ∈ h (qi ). In fact, we may take more than one σi in each h (qi ) for as long as the total number of holes is uniformly bounded. See Sect. 3.4 for further generalizations on the types of holes allowed. 2. Initial distributions. Theorems 1 and 2 (and consequently Theorems 3 and 4) remain true with G replaced by a broader class of measures. For example, we use only the Lipschitz property of dη/dν along unstable leaves, and it is sufficient for dη/dν to be strictly positive on large enough open sets (see Remark 6.3). Moreover, dη/dν need not be bounded provided it blows up sufficiently slowly near the singularity set for f .
358
M. Demers, P. Wright, L.-S. Young
Finally, we remark that Theorem 2(b) continues to hold without requiring that dη/dν be strictly positive anywhere, except that now c(η) might be 0. 2. Relevant Dynamical Structures Our plan is to show that the billiard maps described in Sect. 1.2 admit certain structures called “generalized horseshoes” which can be arranged to “respect the holes.” The main results are summarized in Proposition 2.2 in Sect. 2.2 and proved in Sect. 3. 2.1. Generalized horseshoes. We begin by recalling the idea of a horseshoe with infinitely many branches and variable return times introduced in [Y] for general dynamical systems without holes. These objects will be referred to in this paper as “generalized horseshoes”. Following the notation in Sect. 1.1 of [Y], we consider a smooth or piecewise smooth invertible map f : M → M, and let µ and µγ denote respectively the Riemannian measure on M and on γ where γ ⊂ M is a submanifold. We say the pair ( , R) defines a generalized horseshoe if (P1)–(P5) below hold (see [Y] for precise formulation): (P1) is a compact subset of M with a hyperbolic product structure, i.e., = (∪ u ) ∩ (∪ s ), where s and u are continuous families of local stable and unstable manifolds, and µγ {γ ∩ } > 0 for every γ ∈ u . (P2) R : → Z+ is a return time function to . Modulo a set of µ-measure zero, is the disjoint union of s-subsets j , j = 1, 2, . . . , with the property that for each j, R| j = R j ∈ Z+ and f R j ( j ) is a u-subset of . There is a notion of separation time s0 (·, ·), depending only on the unstable coordinate, defined for pairs of points in , and there are numbers C > 0 and α < 1 such that the following hold for all x, y ∈ : (P3) For y ∈ γ s (x), d( f n x, f n y) ≤ Cα n for all n ≥ 0. (P4) For y ∈ γ u (x) and 0 ≤ k ≤ n < s0 (x, y), (a) d( f n x, f n y) ≤ Cα s0 (x,y)−n ; n det D f u ( f i x) ≤ Cα s0 (x,y)−n . (b) log i=k det D f u ( f i y)
∞ det D f ( f x) ≤ Cα n for all n ≥ 0. (P5) (a) For y ∈ γ s (x), log i=n det D f u ( f i y) u (b) For γ , γ ∈ , if : γ ∩ → γ ∩ is defined by (x) = γ s (x) ∩ γ , then u
is absolutely continuous and
i
d(−1 ∗ µγ ) (x) dµγ
∞ det D f ( f x) . = i=0 det D f u ( f i x) u
i
The meanings of the last three conditions are as follows: Orbits that have not “separated” are related by local hyperbolic estimates; they also have comparable derivatives. Specifically, (P3) and (P4)(a) are (nonuniform) hyperbolic conditions on orbits starting from . (P4)(b) and (P5) treat more refined properties such as distortion and absolute continuity of s , conditions that are known to hold for C 1+ε hyperbolic systems. We say the generalized horseshoe ( , R) has exponential return times if there exist C0 > 0 and θ0 > 0 such that for all γ ∈ u , µγ {R > n} ≤ C0 θ0n for all n ≥ 0. The setting described above is that of [Y]; it does not involve holes. In this setting, we now identify a set H ⊂ M (to be regarded later as the hole) and introduce a few relevant terminologies. Let ( , R) be a generalized horseshoe for f with ⊂ (M \ H ). We say ( , R) respects H if for every i and every with 0 ≤ ≤ Ri , f ( i ) either does not intersect H or is completely contained in H .
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
359
The following definitions of “mixing” are motivated by Markov-chain considerations: Let s ⊂ be an s-subset. We say s makes a full return to at time n if there are numR +···+Ri j ( s ) ⊂ i j+1 bers i 0 , i 1 , . . . , i k with n = Ri0 + · · · + Rik such that s ⊂ i0 , f i0 n s for j < k, and f ( ) is a u-subset of . (i) We say the horseshoe ( , R) is mixing if there exists N such that for every n ≥ N , some s-subset s (n) makes a full return at time n. (ii) If ( , R) respects H , then when we treat H as a hole, we say the surviving dynamics are mixing if in addition to the condition in (i), we require that f s (n) ∩ H = ∅ for all with 0 ≤ ≤ n. This is equivalent to requiring that s (n) makes a full return to at time n under the dynamics of f˚, where f˚ is the map defined in Sect. 1.1. We note that the mixing of f in the usual sense of ergodic theory does not imply that any generalized horseshoe constructed is necessarily mixing in the sense of the last paragraph, nor does mixing of the horseshoe imply that of its surviving dynamics. 2.2. Main Proposition for billiards with holes. With these general ideas out of the way, we now return to the setting of the present paper. From here on, f : M → M is the billiard map of the 2-D Lorentz gas as in Sect. 1.2. The following result lies at the heart of the approach taken in this paper: Proposition 2.1 (Theorem 6(a) of [Y]). The map f admits a generalized horseshoe with exponential return times. A few more definitions are needed before we are equipped to state our main proposition: We call Q ⊂ M a rectangular region if ∂ Q = ∂ u Q ∪ ∂ s Q, where ∂ u Q consists of two unstable curves and ∂ s Q two stable curves. We let Q( ) denote the smallest rectangular region containing , and define µu ( ) := inf γ ∈ u µγ ( ∩ γ ). Finally, for a generalized horseshoe ( , R) respecting a hole H , we define n( , R; H ) = sup{n ∈ Z+ : no point in falls into H in the first n iterates}. In the rest of this paper, C and α will be the constants in (P3)–(P5) for the closed system f . All notation is as in Sect. 1.2. Proposition 2.2. Given an infinitesimal hole {q0 } of Type I or II, there exist C0 , κ > 0, θ0 ∈ (0, 1), and a rectangular region Q such that for all small enough h we have the following: (a) For each σ ∈ h (q0 ), (i) f admits a generalized horseshoe ( (σ ) , R (σ ) ) respecting Hσ ; (ii) both ( (σ ) , R (σ ) ) and the corresponding surviving dynamics are mixing. (b) All σ ∈ h (q0 ) have the following uniform properties: (i) Q( (σ ) ) ≈ Q 1 , and µu ( (σ ) ) ≥ κ; (ii) µγ {R (σ ) > n} < C0 θ0n for all n ≥ 0; (iii) (P3)–(P5) hold with the constants C and α. ¯ → ∞ as h → 0. Moreover, if n(h) ¯ = inf σ ∈ h (q0 ) n( , R; Hσ ), then n(h) Clarification: 1. Here and in Sect. 3, there is a set, namely Hσ , that is identified to be “the hole,” and a horseshoe is constructed to respect it. Notice that the construction is continued after a set enters Hσ . For reasons to become clear in Sect. 6, we cannot simply disregard those parts of the phase space that lie in the forward images of Hσ . 1 By Q( (σ ) ) ≈ Q, we only wish to convey that both rectangular regions are located in roughly the same region of the phase space, M, and not anything technical in the sense of convergence.
360
M. Demers, P. Wright, L.-S. Young
2. Proposition 2.2 treats only small h, i.e. small holes. The smallness of the holes and the uniformness of the estimates in part (b) are needed for the spectral arguments in Sect. 4 to apply. Without any restriction on h, all the conclusions of Proposition 2.2 remain true except for the following: (a)(ii), where for large holes the surviving dynamics need not be mixing, (b)(i), and (b)(ii), where C0 and θ0 may be σ -dependent. The assertions for large h will be evident from our proofs; no separate arguments will be provided. A proof of Proposition 2.2 will require that we repeat the construction in the proof of Proposition 2.1 – and along the way, to carry out a treatment of holes and related issues. We believe it is more illuminating conceptually (and more efficient in terms of journal pages) to focus on what is new rather than to provide a proof written from scratch. We will, therefore, proceed as follows: The rest of this section contains a review of all the arguments used in the proof of Proposition 2.1, with technical estimates omitted and specific references given in their place. A proof of Proposition 2.2 is given in Sect. 3. There we go through the same arguments point by point, explain where modifications are needed and treat new issues that arise. For readers willing to skip more technical aspects of the analysis not related to holes, we expect that they will get a clear idea of the proof from this paper alone. For readers who wish to see all detail, we ask that they read this proof alongside the papers referenced. 2.3. Outline of construction in [Y]. In this subsection, the setting and notation are both identical to that in Sect. 8 of [Y]. Referring the reader to [Y] for detail, we identify below 7 main ideas that form the crux of the proof of Proposition 2.1. We will point out the use of billiard properties and other geometric facts that may potentially be impacted by the presence of holes. Holes are not discussed explicitly, however, until Sect. 3. Notation and conventions. In [Y], S0 and ∂ M were used interchangeably. Here we use exclusively ∂ M. Clearly, f −1 ∂ M is the discontinuity set of f . (i) u- and s-curves. Invariant cones C u and C s are fixed at each point, and curves all of whose tangent vectors are in C u (resp. C s ) are called u-curves (resp. s-curves). (ii) The p-metric. Euclidean distance on M is denoted by d(·, ·). Unless declared otherwise, distances and derivatives along u- and s-curves are measured with respect to a semi-metric called the p-metric defined by cos ϕdr . These two metrics are 1 related by cp(x, y) ≤ d(x, y) ≤ p(x, y) 2 . By Wδu (x), we refer to the piece of local unstable curve of p-length 2δ centered at x. (P3)–(P5) in Sect. 2.1 hold with respect to the p-metric. See Sect. 8.3 in [Y] for details. (iii) Derivative bounds. With respect to the p-metric, there is a number λ > 1 so that all vectors in C u are expanded by ≥ λ and all vectors in C s contracted by ≤ λ−1 . Furthermore, derivatives at x along u-curves are ∼ d(x, ∂ M)−1 . For purposes of distortion control, homogeneity strips of the form 1 1 π π , k ≥ k0 , Ik = (r, ϕ) : − 2 < ϕ < − 2 k 2 (k + 1)2 are used, with {I−k } defined similarly in a neighborhood of ϕ = − π2 . For convenience, we will refer to M \ (∪|k|≥k0 Ik ) as one of the “Ik ”. Important Geometric Facts (†). The following facts are used many times in the proof: (a) the discontinuity set f −1 ∂ M is the union of a finite number of compact piecewise smooth decreasing curves, each of which stretches from {ϕ = π/2} to {ϕ = −π/2};
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
361
(b) u-curves are uniformly transversal (with angles bounded away from zero) to ∂ M and to f −1 ∂ M. 1. Local stable and unstable manifolds. Only homogeneous local stable and unstable curves are considered. Homogeneity for Wδu , for example, means that for all n ≥ 0, f −n Wδu lies in no more than 3 contiguous Ik . Let δ1 > 0 be a small number to be 1
chosen. We let λ1 = λ 4 , δ = δ14 , and define
for all n ≥ 0}, Bλ+1 ,δ1 = {x ∈ M : d( f n x, ∂ M ∪ f −1 (∂ M)) ≥ δ1 λ−n 1 Bλ−1 ,δ1 = {x ∈ M : d( f −n x, ∂ M ∪ f (∂ M)) ≥ δ1 λ−n for all n ≥ 0}. 1
We require d( f n x, f −1 (∂ M)) ≥ δ1 λ−n 1 to ensure the existence of a local unstable curve through x, while the requirement on d( f n x, ∂ M) is to ensure its homogeneity.2 Similar s (x) is well defined reasons apply to stable curves. Observe that (i) for all x ∈ Bλ+1 ,δ1 , W10δ and homogeneous (this is straightforward since δ << δ1 and λ1 is closer to 1 than λ); and (ii) as δ1 → 0, ν(Bλ+1 ,δ1 ) → 1 (this follows from a standard Borel-Cantelli type argument). Analogous statements hold for Bλ−1 ,δ1 . 2. Construction of the Cantor set . The choice of is, in fact, quite arbitrary. We pick a density point x1 of Bλ+1 ,2δ1 ∩ Bλ−1 ,2δ1 at least 2δ1 away from f −1 (∂ M) ∪ ∂ M ∪ f (∂ M), and let = Wδu (x1 ).3 For each n, we define n = {y ∈ : d( f i y, f −1 (∂ M)) ≥ δ1 λ−i 1 for 0 ≤ i ≤ n}, and let ∞ = ∩n n . Then ∞ ⊂ Bλ+1 ,δ1 , by the footnote in item 1 above and our choice of x1 far from ∂ M. Let s consist of all Wδs (y), y ∈ ∞ , and let u be the set of all u curves that meet every γ s ∈ s and which extend by a distance > δ homogeneous Wloc on both sides of the curves in s . The set , which is defined to be (∪ u ) ∩ (∪ s ), clearly has a hyperbolic product structure. (P5)(b) is standard. This together with the choice of x1 guarantees µγ {γ ∩ } > 0 for γ ∈ u , completing the proof of (P1). A natural definition of separation time for x, y ∈ γ u is as follows: Let [x, y] be the subsegment of γ u connecting x and y. Then f n x and f n y are “not yet separated,” i.e. s0 (x, y) ≥ n, if for all i ≤ n, f i [x, y] is connected and is contained in at most 3 contiguous Ik . With this definition of s0 (·, ·), (P3)–(P5)(a) are checked using previously known billiard estimates. 3. The return map f R : → . We point out that there is some flexibility in choosing the return map f R : Certain conditions have to be met when a return takes place, but when these conditions are met, we are not obligated to call it a return; in particular, R is not necessarily the first time an s-subrectangle of Q u-crosses Q, where Q = Q( ). ˜ n = n \ {R ≤ n}. On ˜ n is a partition P˜ n whose We first define f R on ∞ . Let elements are segments representing distinct trajectories. The rules are different before and after a certain time R1 , a lower bound for which is determined by λ1 , δ1 and the derivative of f .4 2 In fact, provided δ is chosen sufficiently small, one can verify that d( f n x, f −1 (∂ M)) ≥ δ λ−n implies 1 1 1 −(n+1) for all n ≥ 0. This fact, which was not used in [Y], will be used in item 2 that d( f n+1 x, ∂ M) ≥ δ1 λ1
below to simplify our presentation. 3 Later we will impose one further technical condition on the choice of x . See the very end of Sect. 2.4. 1 4 In [Y], properties of R are used in 4 places: (I)(i) in Sect. 3.2, Sublemma 3 in Sect. 7.3, the paragraph 1 following (**) in Sect. 8.4, and a requirement in Sect. 8.3 that stable manifolds pushed forward more than R1 times are sufficiently contracted.
362
M. Demers, P. Wright, L.-S. Young
(a) For n < R1 , P˜ n is constructed from the results of the previous step5 as follows: Let ω ∈ P˜ n−1 , and let ω be a component of ω ∩ n . Inserting cut-points only where necessary, we divide ω into subsegments ωi with the property that f n (ωi ) is homogeneous. These are the elements of P˜ n . No point returns before time R1 . (b) For n ≥ R1 , we proceed as in (a) to obtain ωi . If f n (ωi ) u-crosses the middle of Q with ≥ 1.5δ sticking out on each side, then we declare that R = n on ωi ∩ f −n , and the elements of P˜ n |ωi ∩˜ n are the connected components of ωi \ f −n . Otherwise put ωi ∈ P˜ n as before. This defines R on a subset of ∞ (which we do not know yet has full measure); the defis -curves. nition is extended to the associated s-subset of by making R constant on Wloc −n The s-subsets associated with ωi ∩ f in (b) above are the j in (P2). It remains to check that f R ( j ) is in fact a u-subset of . This is called the “matching of Cantor sets” in [Y] and is a consequence of the fact that ∞ is dynamically defined and that R1 is chosen sufficiently large. It remains to prove that p{R ≥ n} decays exponentially with n. Paragraphs 4, 5 and 6 contain the 3 main ingredients of the proof, with the final count given in 7. 4. Growth of u-curves to “long” segments. This is probably the single most important point, so we include a few more details. We first give the main idea before adapting it to the form it is used. Let ε0 > 0 be a number the significance of which we will explain later. Here we think of a u-curve whose p-length exceeds ε0 > 0 as “long”. Consider a u-curve ω. We introduce a stopping time T on ω as follows. For n = 1, 2, . . ., we divide f n ω into homogeneous segments representing distinguishable trajectories. For x ∈ ω, let T (x) = inf{n > 0 : the segment of f n ω containing f n x has p−length > ε0 }. Lemma 2.3. There exist D1 > 0 and θ1 < 1 such that for any u-curve ω, p(ω \ {T ≤ n}) < D1 θ1n
for all n ≥ 1.
This lemma relies on the following important geometric property of the class of billiards in question. This choice of ε0 > 0 is closely connected to this property: n f −i (∂ M) passing through or (*) ([BSC1], Lemma 8.4) The number of curves in ∪i=1 ending in any one point in M is ≤ K 0 n, where K 0 is a constant depending only on the “table” X . 1 Let α0 := 2 ∞ k=k0 k 2 , where {Ik , |k| ≥ k0 } are the homogeneity strips, and assume 1
that λ−1 + α0 < 1. Choose m large enough that θ1 := (K 0 m + 1) m (λ−1 + α0 ) < 1. u -curve of p-length ≤ ε We may then fix ε0 < δ to be small enough that every Wloc 0 −i has the property that it intersects ≤ K 0 m smooth segments of ∪m 1 f (∂ M), so that the u -curve has ≤ (K m + 1) connected components. f m -image of such a Wloc 0 The proof of Lemma 2.3, which follows [BSC2], goes as follows: Consider a large n, which we may assume is a multiple of m. (Once Lemma 2.3 is proved for multiples of m, the estimate can be extended to intermediate values by enlarging the constant D1 .) 5 In [Y], it was sufficient to allow returns to at times that were multiples of a large fixed integer m. Not only is this not necessary (see Paragraph 4), here it is essential that we avoid such periodic behavior to ensure mixing. Thus we take m = 1 when choosing return times in Paragraph 3. This is the only substantial departure we make from the construction in [Y].
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
363
We label distinguishable trajectories by their Ik -itineraries. Notice that because f i ω is the union of a number of (disconnected) u-curves, it is possible for many distinguishable trajectories to have the same Ik -itinerary. Specifically, by (*), each trajectory of length jm, j ∈ Z+ , gives birth to at most (K 0 m + 1) trajectories of length ( j + 1)m with the same Ik -itinerary. To estimate p(ω \ {T ≤ n}), we assume the worst case scenario, in which the f n -images of subsegments of ω corresponding to all distinguishable trajectories have length ≤ ε0 . We then sum over all possible itineraries using bounds on D f along u-curves in Ik . We now adapt Lemma 2.3 to the form in which it will be used. Let ω = f k ω for some ω ∈ P˜ k in the construction in Paragraph 3. As we continue to evolve ω, f n ω is not just chopped up by the discontinuity set, bits of it that go near f −1 (∂ M) will be lost by intersecting with f k+n k+n , and we need to estimate p(ωn \ {T ≤ n}), where ωn := ω ∩ f k (k+n ) takes into consideration these intersections and T is redefined accordingly. A priori this may require a larger bound than that given in Lemma 2.3: it is conceivable that there are segments that will grow to length ε0 without losing these “bits” but which do not now reach this reference length. We claim that all such segments have been counted, because (i) the deletion procedure does not create new connected components; it merely trims the ends of segments adjacent to cut-points; and (ii) the combinatorics in Lemma 2.1 count all possible itineraries (and not just those that lead to “short” segments). This yields the desired estimate on p(ωn \ {T ≤ n}), which is Sublemma 2 in Sect. 8.4 of [Y]. 5. Growth of “gaps” of . Let ω be the subsegment of some γ u ∈ u connecting the two s-boundaries of Q. We think of this as a return in the construction outlined in Paragraph 3, with the connected components ω of ωc = ω \ being f k -images of elements of P˜ k . We define a stopping time T on ωc by considering one ω at a time and defining on it the stopping time in Paragraph 4. Lemma 2.4. There exist D2 > 0 and θ2 < 1 independent of ω such that p(ωnc \ {T ≤ n}) < D2 θ2n
for all n ≥ 1.
The idea of the proof is as follows. We may identify ω with (see Paragraph 2), so that the collection of ω is precisely the collection of gaps in . We say ω is of generation q if this is the first time a part of ω is removed in the construction of ∞ . There are two separate estimates: (I ) := p(ω ); (I I ) := p(ωn \{T ≤ n}). q>εn gen(ω )=q
q≤εn gen(ω )=q
(I) has exponentially small p-measure: this follows from a comparison of the growth rate of D f along u-curves versus the rate at which these curves get cut (see Paragraph 4). (II) is bounded above by
q≤εn gen(ω )=q
C p(ω ) n−q−1 · D1 θ1 . p( f q−1 ω )
This is obtained by applying the modified version of Lemma 2.3 to f q−1 ω . A lower bound on p( f q−1 ω ) can be estimated as these curves have not been cut by f −1 (∂ M) (though they may have been shortened to maintain homogeneity), reducing the estimate to q gen(ω )=q p(ω ), which is ≤ p(ω).
364
M. Demers, P. Wright, L.-S. Young
6. Return of “long” segments. This concerns the evolution of unstable curves after they have grown “long”, where “long” has the same meaning as in Paragraph 4. The following geometric fact from [BSC2] is used: u -curve ω with p(ω) > ε and (**) Given ε0 > 0, ∃n 0 s.t. for every homogeneous Wloc 0 q every q ≥ n 0 , f ω contains a homogeneous segment which u-crosses the middle half of Q with > 2δ sticking out from each side.
We choose ε0 > 0 as explained in Paragraph 4 above, and apply (**) with q = n 0 to the segments that arise in Paragraphs 4 and 5 when the stopping time T is reached. For example, ω here may be equal to f n ω , where ω is a subsegment of the ω in the last paragraph of Paragraph 4 with T |ω = n. We claim that a fixed fraction of such a segment will make a return within n 0 iterates. To guarantee that, two other facts need to be established: (i) The small bits deleted by intersecting with f n+k n+k before the return still leave a segment which u-crosses the middle half of Q with > 1.5δ sticking out from each side; this is easily checked. (ii) For q ≤ n 0 , ( f q ) is uniformly bounded on f −q -images of homogeneous segments that u-cross Q. This is true because a segment contained in Ik for too large a k cannot grow to length δ in n 0 iterates. 7. Tail estimate of return time. We now prove p{R ≥ n} ≤ C0 θ0n for some θ0 < 1. On , introduce a sequence of stopping times T1 < T2 < · · · as follows: A stopping time T of the type in Paragraph 4 or 5 is initiated on a segment as soon as Tk is reached, and Tk+1 is set equal to Tk + T . In this process, we stop considering points that are lost to deletions or have returned to . The desired bound follows immediately from the following two estimates: (i) There exists ε > 0, D3 ≥ 1, and θ3 < 1 such that p(T[ε n] > n) < D3 θ3n . (ii) There exists ε1 > 0 such that if Tk |ω = n, then p(ω∩{R > n +n 0 }) ≤ (1−ε1 ) p(ω), where n 0 is as in (**) in Paragraph 6. (ii) is explained in Paragraph 6. To prove (i), we let p = [ε n], decompose into sets of the form A(k1 , . . . , k p ) = {x ∈ : T1 (x), . . . , T p (x) are defined with Ti = ki }, apply Lemmas 2.1 and 2.2 to each set and recombine the results. The argument here is combinatorial, and does not use further geometric information about the system. 2.4. Sketch of proof of (**) following [BSC2]. Property (**) is a weaker version of Theorem 3.13 in [BSC2]. We refer the reader to [BSC2] for detail, but include an outline of its proof because a modified version of the argument will be needed in the proof of Proposition 2.2. We omit the proof of the following elementary fact, which relies on the geometry of the discontinuity set including Property (*): Sublemma A. Given any u-curve γ , through µγ -a.e. x ∈ γ passes a homogeneous s (x) for some δ(x) > 0. The analogous statement holds for s-curves. Wδ(x) u -curve as required in (**), the problem is reduced Instead of considering every Wloc to a finite number of “mixing boxes” U1 , U2 , . . . , Uk with the following properties:
(i) U j is a hyperbolic product set defined by (homogeneous) families u (U j ) and s (U j ); located in the middle third of U j is an s-subset U˜ j with ν(U˜ j ) > 0; (ii) ∪ u (U j ) fills up nearly 100% of the measure of Q(U j ); and u -curve ω with p(ω) > ε passes through the middle third of one of the (iii) every Wloc 0 Q(U j ) in the manner shown in Fig. 1 (left).
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
365
Fig. 1. Left: A mixing box U j . Right: The target box U0
That (i) and (ii) can be arranged follows from Sublemma A. That a finite number of U j suffices for (iii) follows from a compactness argument. Next we choose a suitable subset U˜ 0 ⊂ to be used in the mixing. To do that, first pick a hyperbolic product set U0 related to Q( ) as shown in Fig. 1 (right). We require that it meet Q( ) in a set of positive measure, that it sticks out of Q( ) in the u-direction by more than 2δ, and that the curves in u (U0 ) fill up nearly 100% of Q(U0 ). Let 0 > 0 be a small number, and let U˜ 0 ⊂ U0 consist of those density points of U0 ∩ Q( ) with the additional property that if a homogeneous stable curve γ s with p(γ s ) < 0 meets such a point, then p(γ s ∩ U0 )/ p(γ s ) ≈ 1. For 0 small enough, ν(U˜ 0 ) > 0 because the u -curves is absolutely continuous. foliation into Wloc By the mixing property of ( f, ν), there exists n 0 such that for all q ≥ n 0 , ν( f q (U˜ j ) ∩ U˜ 0 ) > 0 for every U˜ j . We may assume also that n 0 is so large that for q ≥ n 0 , if x ∈ U˜ j is such that f q x ∈ U˜ 0 , then p( f q (γ s (x))) < 0 , where γ s (x) is the stable curve in s (U˜ j ) passing through x. Let q ≥ n 0 and j be fixed, and let x ∈ U˜ j be as above. From the high density of unstable curves in both U j and U0 , we are guaranteed that there are two elements γ1u , γ2u ∈ u (U j ) sandwiching the middle third of Q(U j ) such that for each i, a subsegment of γiu containing γ s (x) ∩ γiu is mapped under f q onto some γˆiu ∈ u (U0 ). Let Q ∗ = Q ∗ (q, j) be the u-subrectangle of Q(U0 ) with ∂ u Q ∗ = γˆ1u ∪ γˆ2u . Sublemma B. f −q | Q ∗ is continuous, equivalently, Q ∗ ∩ (∪0 f i (∂ M)) = ∅. q
Sublemma B is an immediate consequence of the geometry of the discontinuity set: q By the choice of x1 in item 2 of Sect. 2.3, Q ∗ ∩∂ M = ∅. Suppose Q ∗ ∩(∪1 f i (∂ M)) = ∅. q i Since ∪1 f (∂ M) is the union of finitely many piecewise smooth (increasing) u-curves each connected component of which stretches from {ϕ = −π/2} to {ϕ = π/2}, and q these curves cannot touch ∂ u Q ∗ , a piecewise smooth segment from ∪1 f i (∂ M) that ∗ s ∗ enters Q through one component of ∂ Q must exit through the other. In particular, it must cross f q γ s (x), which is a contradiction. u -curve with p(ω) > ε . We pick U so that ω passes To prove (**), let ω be a Wloc 0 j through the middle third of U j as in (iii) above. Sublemma B then guarantees that f q (ω∩ f −q Q ∗ ) connects the two components of ∂ s Q ∗ . This completes the proof of (**), except that we have not yet verified that f q (ω ∩ f −q Q ∗ ) is homogeneous. To finish this last point, we modify the above argument as follows: First, we define a u curve γ to be strictly homogeneous if for all n ≥ 0, f −n γ is contained inside one Wloc s curves is defined analogously. The homogeneity strip Ik (n). Strict homogeneity for Wloc conclusions of Sublemma A remain valid if, in its statement, the word “homogeneous” is replaced by “strictly homogeneous.” Thus the mixing boxes U1 , . . . , Uk can be chosen so that their defining families are comprised entirely of strictly homogeneous local manifolds. Furthermore, if x1 is also chosen as a density point of points with sufficiently long strictly homogeneous unstable curves, u (U0 ) can be chosen to be comprised entirely u -curves. Having done this, an argument very similar to of strictly homogeneous Wloc
366
M. Demers, P. Wright, L.-S. Young
the proof of Sublemma B shows that f −i Q ∗ ∩ (∪k ∂ Ik ) = ∅ for 0 ≤ i ≤ q, and this completes the proof of (**). 3. Horseshoes Respecting Holes for Billiard Maps 3.1. Geometry of holes in phase space. We summarize here some relevant geometric properties and explain how we plan to incorporate holes into our horseshoe construction. Holes of Type I. Recall from Sect. 1.2 that for q0 ∈ ∪i and σ ∈ h (q0 ), Hσ ⊂ M is a rectangle of the form (a, b) × [− π2 , π2 ]. We define ∂ Hσ := {a, b} × [− π2 , π2 ], i.e. ∂ Hσ is the boundary of Hσ viewed as a subset of M. It will also be convenient to let H0 ⊂ M denote the vertical line {q0 } × [− π2 , π2 ]. To construct a horseshoe respecting Hσ , it is necessary to view two nearby points as having separated when they lie on opposite sides of ∂ Hσ or on opposite sides of Hσ in M \ Hσ . Thus it is convenient to view f −1 (∂ Hσ ) as part of the discontinuity set of f . For simplicity, consider first the case where q0 does not lie on a line in the table X tangent to more than one scatterer. Then f −1 (∂ Hσ ) is a finite union of pairs of roughly parallel, smooth s-curves. (Recall that s-curves are negatively sloped, with slopes uniformly bounded away from 0 and −∞.) Each of the curves comprising f −1 (∂ Hσ ) begins and ends in ∂ M ∪ f −1 (∂ M), that is to say, the geometric properties of f −1 (∂ Hσ ) ∪ f −1 (∂ M) are similar to those of f −1 (∂ M). Likewise, f (∂ Hσ ) is a finite union of pairs of (increasing) u-curves that begin and end in ∂ M ∪ f (∂ M), and it will be convenient to regard that as part of the discontinuity set of f −1 . Let Nε (·) denote the ε-neighborhood of a set. We will need the following lemma. Lemma 3.1. For each ε > 0 there is an h > 0 such that for each σ ∈ h , Hσ ⊂ Nε (H0 ), f Hσ ⊂ Nε ( f H0 ), and f −1 Hσ ⊂ Nε ( f −1 H0 ). As f is discontinuous, Lemma 3.1 is not immediate. However, it can be easily verified, and we leave the proof to the reader. Points q0 that lie on lines in X with multiple tangencies to scatterers lead to slightly more complicated geometries, and special care is needed when defining what is meant by f H0 and f −1 H0 . For example, consider the case where q0 ∈ 3 lies on a line that
Fig. 2. An infinitesmal hole aligned with multiple tangencies. Left: q0 lies on a line segment in the billiard table X that is tangent to two scatterers. Right: Induced singularity curves in the subset 1 × [−π/2, π/2] of the phase space M
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
367
is tangent to 1 and 2 , but which is not tangent to any other scatterer including 3 . Suppose further that r1 ∈ 1 , r2 ∈ 2 are the points of tangency, that r2 is closer to q0 than r1 is, that no other scatterer touches the line segment [q0 , r1 ], and that 1 and 2 both lie on the same side of [q0 , r1 ]; see Fig. 2 (left). Let σ be a small hole of Type I with q0 ∈ σ . Then in 2 × [−π/2, π/2], f −1 (∂ Hσ ) appears as described above. However, 2 “obstructs” the view of σ from 1 , and so in 1 × [−π/2, π/2], f −1 (Hσ ) is a small triangular region whose three sides are composed of a segment from 1 × {π/2}, a segment from f −1 (2 × {π/2}), and a single segment from f −1 (∂ Hσ ). See Fig. 2 (right). As a consequence, when we write f −1 H0 , we include in this set not just (r2 , π/2), but also f −1 (r2 , π/2) = (r1 , π/2). This is necessary in order for Lemma 3.1 to continue to hold. Aside from such minor modifications, the case of multiple tangencies is no different than when they are not present, and we leave further details to the reader. Holes of Type II. For simplicity, consider first the case where q0 does not lie on a line in the “table” X tangent to more than one scatterer. Recall from Sect. 1.2 that “the hole” Hσ here is taken to be f (Bσ ), where Bσ consists of points in M which enter σ ×S1 under the billiard flow before returning to the section M. As with holes of Type I, we define ∂ Hσ to be the boundary of Hσ viewed as a subset of M. The set Bσ as a subset of M has similar geometric properties as f −1 Hσ for Type I holes, i.e., f −1 (∂ Hσ )\(∂ M ∪ f −1 (∂ M)) consists of pairs of negatively sloped curves ending in ∂ M ∪ f −1 (∂ M). The slopes of these curves are uniformly bounded (independent of σ ) away from −∞ and 0. For the reasons discussed, it will be convenient to view this set as part of the discontinuity set of f . The infinitesimal hole H0 ⊂ M is defined in the natural way, and the analog of Lemma 3.1 can be verified. We will say more about the geometry of Hσ in Sect. 3.3. Points q0 that lie on multiple tangencies lead to slightly more complicated geometries, and special care is needed when defining what is meant by the sets f −1 H0 , H0 , and f H0 as in the case of Type I holes. Further generalizations on holes of Type II: In addition to the generalizations discussed in Sect. 1.3, sufficient conditions on the holes allowed in h for Prop. 2.2 to remain true are the following, as can be seen from our proofs: (1) There exist N and L for which the following hold for all sufficiently small h: (a) f −1 (∂ Hσ ), ∂ Hσ , and f (∂ Hσ ) each consist of no more than N smooth curves, all of which have length no greater than L. (b) For each σ ∈ h , f −1 (∂ Hσ )\(∂ M ∪ f −1 (∂ M)) consists of piecewise smooth, negatively sloped curves (with slopes uniformly bounded away from −∞ and 0), and the end points of these curves must lie on ∂ M ∪ f −1 (∂ M). (2) The analog of Lemma 3.1 holds. Thus it would be permissible to allow a convex hole σ to be in h that did not have a C 3 simple closed curve with strictly positive curvature as its boundary. For example, conditions (a) and (b) above hold if ∂σ is a piecewise C 3 simple closed curve which consists of finitely many smooth segments that are either strictly positively curved or flat. As another generalization, consider the case when any line segment in the table X with its endpoints on two scatterers that passes through the convex hull of σ also intersects σ . Then it is no loss of generality to replace σ by its convex hull. Using this, one can often verify that the set Hσ that arises satisfies properties (a) and (b) above, even if σ is not itself convex. See Fig. 3. In Sect. 3.2, the discussion is for holes of Type I with a single interval deleted. The proof follows mutatis mutandis for holes of Type II, with the necessary minor modifications discussed in Sect. 3.3.
368
M. Demers, P. Wright, L.-S. Young
Fig. 3. Examples of Type II holes that are permissible
3.2. Proof of Proposition 2.2 (for holes of Type I). The idea of the proof is as follows. First we construct a horseshoe ( (0) , R (0) ) with the desired properties for the infinitesimal hole {q0 }. Then we construct ( (σ ) , R (σ ) ) for all σ ∈ h (q0 ), and show that with (σ ) sufficiently close to (0) in a sense to be made precise, ( (σ ) , R (σ ) ) will inherit the desired properties with essentially the same bounds. To ensure that (σ ) can be taken “close enough” to (0) , we decrease the size of the hole, i.e., we let h → 0. Now the constructions of ( (0) , R (0) ) and ( (σ ) , R (σ ) ) are essentially identical. To avoid repeating ourselves more than needed, we will carry out the two constructions simultaneously. It is useful to keep in mind, however, that logically, the case of the infinitesimal hole is treated first, and some of the information so obtained is used to guide the arguments for positive-size holes. As explained in Sect. 3.1, to ensure that the horseshoe respects the hole, it is convenient to include f −1 (∂ Hσ ) as part of the discontinuity set for f . Since Hσ will be viewed as a perturbation of H0 , we include f −1 (H0 ) in this set as well. The following convention will be adopted when we consider a system with hole Hσ : (a) Suppose for definiteness q0 ∈ 1 . The new phase space Mσ is obtained from M by cutting 1 × [− π2 , π2 ] along the lines comprising H0 ∪ ∂ Hσ , splitting it into three connected components. (b) As a consequence, the new discontinuity set of f is f −1 (∂ Mσ ), and the new discontinuity set of f −1 is f (∂ Mσ ). We use the notation “σ = 0” for the infinitesimal hole, so that M0 is obtained from M by cutting along H0 . Notice immediately that this changes the definitions of stable and unstable curves, in the sense that if γ was a stable curve for the system without holes, then γ continues to be a stable curve if and only if (i) γ ∩ ∂ Mσ = ∅, and (ii) f n (γ ) ∩ f −1 ∂ Mσ = ∅ for all n ≥ 0; a similar characterization holds for unstable curves. All objects constructed below will be σ -dependent, but we will suppress mention of σ except where it is necessary. Observe also that the Important Geometric Facts (†) in Sect. 2.3 with f −1 ∂ Mσ instead of f −1 ∂ M as the new discontinuity set remains valid. We now follow sequentially the 7 points outlined in Sect. 2.3 and discuss the modifications needed. These modifications, along with two additional points (8 and 9) form a complete proof of Proposition 2.2. We believe we have prepared ourselves adequately in Sects. 2.3 and 2.4 so that the discussion to follow can be understood on its own, but encourage readers who wish to see proofs complete with all technical detail to read the rest of this section alongside the relevant parts of [Y and BSC2]. The notation within each item below is as in Sect. 2.3.
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
369 (σ )±
1. The relationships λ = λ41 and δ = δ14 are as before, and the sets Bλ1 ,δ1 are defined in a manner similar to that in Sect. 2.3. For example, Bλ(σ1 ,δ)+1 = {x ∈ Mσ : d(x, ∂ Mσ ) ≥ δ1 and d( f n x, f −1 ∂ Mσ ) ≥ δ1 λ−n 1 ∀n ≥ 0}. As in Sect. 2.3, the condition on d( f n x, f −1 ∂ Mσ ) is to ensure the existence of stable curves, and the necessity for x to be away from ∂ Mσ is obvious (cf. footnote in item 1 of Sect. 2.3). Properties (i) and (ii) continue to hold for each σ given the geometry of the new discontinuity set. With regard to the choice of δ1 , we let δ1 be as in [Y], (0)+ (0)− and shrink it if necessary to ensure that Bλ1 ,2δ1 ∩ Bλ1 ,2δ1 has positive ν-measure away from f −1 (∂ M0 ) ∪ ∂ M0 ∪ f (∂ M0 ). This is where the sets (σ ) will be located (see Paragraph 2). (σ )± (0)± The following lemma relates Bλ1 ,δ1 and Bλ1 ,δ1 : (σ )±
(0)±
Lemma 3.2. (i) For all σ ∈ h , we have Bλ1 ,δ1 ⊂ Bλ1 ,δ1 . (ii) As h → 0, sup ν(Bλ(0)+ \ Bλ(σ1 ,δ)+1 ) → 0, 1 ,δ1
sup ν(Bλ(0)− \ Bλ(σ1 ,δ)−1 ) → 0. 1 ,δ1
σ ∈ h
σ ∈ h
Proof. (i) follows immediately from ∂ Mσ ⊃ ∂ M0 . As for (ii), let ε > 0 be given. Recall that Nα (·) denotes the α-neighborhood of a set. By Lemma 3.1 we may choose h small enough that for all σ ∈ h , ∂ Hσ ∈ Nε (H0 ) and f −1 (∂ Hσ ) ∈ Nε ( f −1 H0 ). Then if (0)+ (σ )+ x ∈ Bλ1 ,δ1 \ Bλ1 ,δ1 , either x ∈ Nδ1 (∂ Hσ ) \ Nδ1 (H0 ), or x ∈ ∪n≥0 f −n (Nδ1 λ−n ( f −1 ∂ Hσ ) \ Nδ1 λ−n ( f −1 H0 )). 1
1
We estimate the ν-measure of the right side separately for ∪n≥n ε and ∪n
)
ε n ε = inf{n ≥ 0 : δ1 λ−n 1 ≤ ε} ≈ ln λ1 . For n ≥ n ε , the measure in question is ≤ ν(Nδ1 λ−n ( f −1 (∂ Hσ )) ≤ const · δ1 λ−n 1 ≤ const · ε.
ln(
1
n≥n ε
n≥n ε
f −1 (∂ Hσ ) consists of a finite number of smooth compact curves
Here we have used that the total length of which is bounded independent of σ . Adding to this that these curves are within a distance ε of the curves in f −1 H0 , we see that for each n < n ε , ν(Nδ1 λ−n ( f −1 (∂ Hσ )) \ Nδ1 λ−n ( f −1 H0 )) ≤ const · ε. 1
1
Similarly, ν(Nδ1 (∂ Hσ ) \ Nδ1 (H0 )) ≤ const · ε. Hence (0)+ ν(Bλ1 ,δ1
\
(σ )+ Bλ1 ,δ1 )
which tends to 0 as ε → 0.
δ1 ≤ const · (ε + (n ε + 1)ε) ≤ const · ε ln ε
,
(0)
(0)+
(0)−
2. To construct the Cantor sets, we first pick x1 as a density point of Bλ1 ,2δ1 ∩ Bλ1 ,2δ1 at least 2δ1 away from f −1 (∂ M0 )∪∂ M0 ∪ f (∂ M0 ) and the boundaries of the homogeneity (0) strips, and begin to construct (0) with = Wδu (x1 ). We then do the same for each σ , (σ ) (σ )+ (σ )− i.e., pick x1 as a density point of Bλ1 ,2δ1 ∩ Bλ1 ,2δ1 and begin to construct (σ ) centered at x1(σ ) – except that for reasons to become clear, we will want d(x1(σ ) , x1(0) ) < δ2 , where δ2 > 0 is determined by properties of ( (0) , R (0) ) (requirements will appear below, and
370
M. Demers, P. Wright, L.-S. Young
in items 6 and 9). Suffice it to say here that however small δ2 may be, Lemma 3.2 guarantees that this can be done by shrinking h. Once x1(σ ) is chosen, we set = Wδu (x1(σ ) ) and n = {y ∈ : d( f i y, f −1 (∂ Mσ )) ≥ δ1 λ−i 1 for 0 ≤ i ≤ n}. Then the sets ∞ , s , u and (σ ) are constructed as before. (σ ) (0) That Q( (σ ) ) ≈ Q( (0) ) follows immediately from the proximity of x1 to x1 . u (σ ) u (0) Since δ is fixed, µ ( ) ≈ µ ( ) > 0 can be arranged by taking δ2 sufficiently small and using Lemma 3.2 with h sufficiently small. This proves Proposition 2.2(b)(i). With the separation time happening sooner due to the enlarged discontinuity set, (P3)– (P5) remain true with the same C and α for the closed system; in other words, Proposition 2.2(b)(iii) requires no further work. 3. To arrange for mixing properties (not done in [Y]), we will need to delay the return times to by forbidding returns before time R2 for some R2 ≥ R1 determined by ( (0) , R (0) ); see Lemma 3.4. This aside, the construction of f R is as before. The matching of Cantor sets argument should be looked at again since the Cantor sets are different, but the proof goes through as before because the sets are dynamically defined. Notice that for ω ∈ P˜ n , f i ω is either entirely in the hole or outside of the hole, as is i f ( s ), where s is the s-subset of associated with ω, for 0 ≤ i ≤ n; this is a direct consequence of our taking the boundary of the hole into consideration in our definition of the discontinuity set. Together with the fact that is away from ∂ Mσ , it ensures that the generalized horseshoe we are constructing respects the hole. 4. This is where one of the more substantial modifications occurs: Lemma 2.3, which is based largely on the competition between expansion along u-curves and the rate at which they are cut, is clearly affected by the additional cutting due to our enlarged discontinuity set. The condition (*) in Sect. 2.3 must now be replaced by Lemma 3.3. There exists K 1 such that for any m ∈ Z+ , there exists ε0 > 0 with the property that for any u-curve with p(ω) < ε0 , f m (ω) has ≤ (K 1 m 2 + 4) connected components with respect to the enlarged discontinuity set. Proof. Let m ∈ Z+ be given. As in Sect. 2.3, choose ε0 > 0 small enough such that if ω is a u-curve with p(ω) < ε0 , f m (ω) has ≤ (K 0 m + 1) connected components with respect to the original discontinuity set f −1 S0 . Let ω j be the f −m -image of one of these connected components. This means that for 0 ≤ k ≤ m, f k (ω j ) is, in reality, a connected u-curve even though it may not be connected with respect to our enlarged discontinuity set. Since f k ω j is an (increasing) u-curve, it can meet the three vertical lines making up (∂ Hσ ) ∪ H0 in no more than three points. (As the slopes dϕ/dr of u-curves are never less than the curvature of i at r , connected u-curves cannot wrap around the cylinder i × [−π/2, π/2] and meet (∂ Hσ ) ∪ H0 more than once.) Hence the cardinality −k ((∂ H ) ∪ H )} is ≤ 3(m + 1), and as (∂ H ) ∪ H is the additional of {ω j ∩ m σ 0 σ 0 k=0 f set added to ∂ M to create ∂ Mσ , it follows that f m ω has ≤ (K 0 m + 1) · (3(m + 1) + 1) connected components with respect to the enlarged discontinuity set. Using Lemma 3.3, one adapts easily the proof of Lemma 2.3 to the present setup with 1 θ1 = (K 1 m 2 + 4) m (λ−1 + α0 ), where m is chosen large enough so that this number is < 1. The constant D1 depends only on the properties of D f and is unchanged. Hence Lemma 2.3 is valid with D1 and θ1 modified but independent of σ . As in Sect. 2.3, these estimates can then be adapted to estimate p(ωn \{T ≤ n}).
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
371
5. Lemma 2.4 remains valid with modified constants which are independent of σ . Returning to the sketch of the proof provided in Sect. 2.3, we see that both sets of estimates boil down to the geometry of the new discontinuity set and the rates of growth versus cutting, which has been taken care of for the enlarged discontinuity set in Paragraph 4 above. 6. We need to show that there exist n 1 and ε1 > 0 independent of σ such that for every homogeneous u-curve with p-length > ε0 , a fraction ≥ ε1 of ω returns within the next n 1 steps. Before we enlarged the discontinuity set, this property followed from property (**) in Sect. 2.3. We replace (**) here with the following: Lemma 3.4. Given ε0 > 0, provided h and δ2 are sufficiently small, there exists n 1 u -curve ω such that the following holds for each σ ∈ h : for every homogeneous Wloc with p(ω) > ε0 and each q ∈ {n 1 , n 1 + 1}, f q ω contains a homogeneous segment that u-crosses the middle half of Q( (σ ) ) with greater than 2δ sticking out from each side. Once Lemma 3.4 is proved, the fact that a fraction ε1 (independent of σ ) has the desired properties follows from derivative estimates as in Sect. 2.3 and our uniform lower bound on µu ( (σ ) ). The reason we want q to take two consecutive values in the statement of Lemma 3.4 has to do with the mixing property in item 9 below. Proof. Fix ε0 > 0. We first prove the following for the case σ = 0: u -curve ω with p(ω) > (**)’ For σ = 0, there exists n 1 such that any homogeneous Wloc q ε0 and every q ≥ n 1 , f ω contains a homogeneous segment that u-crosses the middle fourth of Q( (0) ) with greater than 4δ sticking out from each side.
The proof of (**)’ is completely analogous to the proof of (**) outlined in Sect. 2.4. Sublemmas A and B continue to hold due to the similar geometry of the discontinuity set. Notice that unlike (**)’, the assertion in Lemma 3.4 is only for q = n 1 and n 1 + 1, so that its proof involves only a finite number of mixing boxes U j and a finite number of iterates. This will be important in the perturbative argument to follow. Consider now σ = 0, and consider a homogeneous unstable curve ω with p(ω) > ε0 . First, ω continues to be an unstable curve with respect to the discontinuity set f −1 ∂ M0 , so by the proof of (**)’, for q ∈ {n 1 , n 1 + 1} and every j, there is a rectangular region Q ∗ = Q ∗ (q, j) such that (i) Q ∗ u-crosses the middle fourth of Q( (0) ) with > 4δ sticking out, (ii) f −q Q ∗ is an s-subrectangle in the middle third of Q(U j ), and (iii) for i = 0, 1, · · · , q, f −i Q ∗ stays clear of f −1 ∂ M0 by some amount. Lemma 3.1 ensures that for h small enough, (iii) continues to hold with f −1 ∂ M0 replaced by f −1 ∂ Mσ . Finally, provided δ2 is small enough, (i) holds for Q( (σ ) ) with > 2δ sticking out on each side. 7. Once steps 4, 5 and 6 have been completed, the argument here is unchanged (as it is largely combinatorial), guaranteeing constants C0 and θ0 independent of σ with p{R ≥ n} ≤ C0 θ0n . This completes the proof of Proposition 2.2(a)(i) and (b)(ii). We have reached the end of the 7 steps outlined in Sect. 2.3. Two items remain: 8. That n(h) ¯ → ∞ as h → 0 is easy: Orbits from (σ ) start away from H0 and cannot −1 approach f H0 faster than a fixed rate. Thus using Lemma 3.1, we can arrange for orbits starting from (σ ) to stay out of Hσ for as long as we wish by taking h small. 9. The mixing of ( (σ ) , R (σ ) ) follows from Lemma 3.5. There exists R2 ≥ R1 (independent of σ ) such that for small enough h, the construction in Step 3 can be modified to give the following:
372
M. Demers, P. Wright, L.-S. Young
(i) no returns are allowed before time R2 , and (ii) at both times R2 and R2 + 1, there are s-subsets of (σ ) making full returns. Proof. Again we first consider the case σ = 0. Here R2 is chosen as follows: Without allowing any returns, let R1 be the smallest time greater than or equal to R1 such that there exists ω ∈ P˜ R1 with p( f R1 ω) > ε0 > 0. With ε0 chosen as before, we take n 1 from Lemma 3.4 and set R2 = R1 + n 1 . Using Lemma 3.4, we find two subsegments ω and ω ⊂ ω such that f R2 ω and f R2 +1 ω are both homogeneous segments that u-cross the middle half of Q( (0) ) with greater than 2δ sticking out from each side. We may suppose that ω and ω are disjoint since f has no fixed points. They give rise to two s-subsets of (0) with the properties in (ii). From time R2 on, returns to (0) are allowed as before. When σ = 0, we follow the same procedure as above to ensure the mixing of ( (σ ) , R (σ ) ). The only concern is that R1 = R1 (σ ) (and hence also R2 = R1 + n 1 ) might not be independent of σ . This is not a problem as the construction above involves only a finite number of steps: With h and δ2 sufficiently small, the elements of P˜ n(σ ) can (0) be defined in such a way that they are in a one-to-one correspondence with those of P˜ n for n ≤ R1 (0). Finally, mixing of the surviving dynamics is ensured by choosing h small enough that n(h) ¯ > R2 + 1. This ensures that the s-subsets s that make full returns at times R2 and R2 + 1 cannot fall into the hole prior to returning. The proof of Proposition 2.2 for holes of Type I is now complete. 3.3. Modifications needed for holes of Type II. The proof for Type II holes is very similar to that for Type I holes. There are, however, some differences due to the more complicated geometry of ∂ Hσ . In the discussion below, we assume the infinitesimal hole {q0 } does not lie on any segment in the table tangent to more than one scatterer. The general situation is left to the reader. From the discussion of the geometry of Type II holes in Sect. 3.1, we see that the Important Geometric Facts (†) in Sect. 2.3 continue to hold with Mσ in the place of M, except that u-curves need not be transversal to the ∂ Hσ ∪ H0 part of ∂ Mσ . Potential problems that may arise are discussed below. The discontinuity set of f , i.e. f −1 ∂ Mσ , has the same geometric properties as before. We now go through the 9 points in Sect. 3.2. No modifications are needed in items 1–3. As expected, item 4 is where the most substantial modifications occur: Modifications in Item 4. Lemma 3.3 is still true as stated, but the geometry is different. In the discussion below related to this lemma, the discontinuity set refers to f −1 ∂ M, not the enlarged discontinuity set f −1 ∂ Mσ , and unstable curves are defined accordingly. For Type I holes, the proof relies on the fact that any (increasing) connected u-curve ω meets ∂ Hσ ∪ H0 , which is the union of three vertical lines, in at most three points. Lemma 3.6. Any unstable curve ω meets ∂ Hσ ∪ H0 in at most three points. Even though Lemma 3.3 is stated for u-curves, we need it only for unstable curves (and the argument here is slightly simpler for unstable curves). Proof. Let us distinguish between two different types of curves that comprise ∂ Hσ : Primary segments, which are the forward images of curves in ∂ Bσ \ f −1 (∂ M),
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
373
Fig. 4. Representative examples of the geometry of Type II holes. Top set: On the left is a configuration on the billiard table X , while on the right is the resulting configuration in the subset 1 × [− π2 , π2 ] of phase space. In this subset, H0 consists of a single primary segment whose endpoints lie on f (2 × { π2 }) and 1 × { π2 }. For σ = 0, ∂ Hσ contains two primary segments and a single secondary segment that lies on f (2 × { π2 }). (Recall that by convention ∂ Hσ does not include subsegments of ∂ M.) Bottom set: The analogous situation when the view of 1 from q0 is obstructed by two scatterers, instead of just one. Observe that now ∂ Hσ contains two secondary segments in 1 × [− π2 , π2 ]. The situation when the view of 1 from q0 is unobstructed by other scatterers is simple and is left to the reader
and secondary segments, which are subsegments of f (∂ M). For examples, see Fig. 4. In general, when q0 does not lie on a line segment with multiple tangencies to the scatterers, secondary segments are absent in H0 , while each component of H0 gives rise to two primary segments in ∂ Hσ for σ = 0. To prove the lemma, observe first that H0 can have no more than one component in any connected component of M \ f (∂ M). Second, ω must also be entirely contained inside one connected component of M \ f (∂ M). This is because unstable curves for f cannot cross the discontinuity set of f −1 . As a consequence, ω also cannot cross any secondary segment as secondary segments of ∂ Hσ are contained in f (∂ M). It remains to show that ω can meet each primary segment in at most one point. Although primary segments are increasing, their tangent vectors lie outside of unstable cones (except at ∂ M where the unstable cone is degenerate). This is because the curves in ∂ Bσ \ f −1 ∂ M are decreasing, while the unstable cones are defined to be the forward images of {0 ≤ dϕ dr ≤ ∞} under D f . Hence primary segments have greater slopes than ω.
374
M. Demers, P. Wright, L.-S. Young
As pointed out in Sect. 2.3, item 4, Lemma 3.3 must be modified to account for the deletions that arise from intersections with forward images of n , and one might be concerned about the absence of uniform estimates on transversality in (†) between ∂ Hσ and unstable curves. This, in fact, is not a problem, because such deletions occur only in neighborhoods of f −1 ∂ Mσ , which are decreasing curves and hence uniformly transversal to u-curves. This completes the modifications associated with item 4. No modifications are required for items 5, 7, 8 and 9. Modifications in Item 6. In the proof of (**)’, the argument needs to be modified, again q due to the difference in geometry: In order to prove that Q ∗ ∩ (∪0 f i (∂ M0 )) = ∅ (Subq i lemma B), in the case of Type I holes we use that ∪1 f (∂ M0 ) is the union of finitely many piecewise smooth increasing curves that stretch from {ϕ = − π2 } to {ϕ = π2 }. For Type II holes, this is not true. However, it can be arranged that Sublemma B will continue to hold as we now explain: First,
q q q−1 ∪1 f i (∂ M0 ) ⊂ (∪1 f i (∂ M)) ∪ (∪0 f i (H0 )) ∪ f q (H0 ). If we write the right side as A ∪ f q (H0 ), then A has the desired geometry, i.e. it is the union of finitely many piecewise smooth increasing curves that stretch from {ϕ = − π2 } to {ϕ = π2 }. Thus the same argument as before shows that this set is disjoint from Q ∗ . One way to ensure that Q ∗ ∩ f q (H0 ) = ∅ is to choose the mixing boxes U j disjoint from H0 , which can easily be arranged given the geometry of primary segments discussed above. This completes the proof of Proposition 2.2 for Type II holes. 4. Escape Dynamics on Markov Towers In this section and the next, we lift the problems from the billiard systems in question to their Markov tower extensions, and solve the problems there. In Sect. 4, we review relevant works and formulate results on towers. Proofs are given in Sect. 5. 4.1. From generalized horseshoes to Markov towers (review). It is shown in [Y] that given a map f : M → M with a generalized horseshoe ( , R) as defined in Sect. 2.1, one can associate a Markov extension F : → which focuses on the return dynamics to (and suppresses details between returns). We first recall some facts about this very general construction, taking the opportunity to introduce some notation. Let = {(x, n) ∈ × N : n < R(x)}, and define F : → as follows: For < R(x) − 1, we let F(x, ) = (x, + 1), and define F(x, R(x) − 1) = ( f R(x) (x), 0). Equivalently, one can view as the disjoint union ∪≥0 , where , the th level of the tower, is a copy of {x ∈ : R(x) > }. This is the representation we will use. There is a natural projection π : → M such that π ◦ F = f ◦ π . In general, π is not one-to-one, but for each ≥ 0, it maps bijectively onto f ( ∩ {R ≥ }). In the construction of ( , R), one usually introduces an increasing sequence of partitions of into s-subsets representing distinguishable itineraries in the first n steps. ˜ .) These partitions (In Sects. 2.3 and 3.2, these partitions were given by P˜ of
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
375
induce a partition {, j } of which is finite on each level and and is a (countable) Markov partition for F. We define a separation time s(x, y) ≤ s0 (x, y) by inf{n > 0 : F n x, F n y lie in different , j }. We borrow the following language from ( , R) for use on : For each , j, recall that s (π(, j )) and u (π(, j )) are the stable and unstable families defining the hyperbolic product set π(, j ). We will say γ˜ ⊂ , j is an unstable leaf of , j if π(γ˜ ) = γ ∩ π(, j ) for some γ ∈ u (π(, j )), and use u (, j ) to denote the set of all such γ˜ . Let u () = ∪, j u (, j ) be the set of all unstable leaves of . Stable leaves of , j and the families s (, j ) and s () are defined similarly. Associated with F : → , which we may think of as a “hyperbolic tower”, is its quotient “expanding tower” obtained by collapsing stable leaves to points. Topologically, = /∼, where for x, y ∈ , x ∼ y if and only if y ∈ γ (x) for some γ ∈ s (). Let π : → be the projection defined by ∼, and let F : → be the induced map on satisfying F ◦ π = π ◦ F. We will use the notation = π ( ), , j = π(, j ), and so on. It is shown in [Y] that there is a well defined differential structure on preserved by F. Recall that µγ is the Riemannian measure on γ , and for γ , γ ∈ u ( ), γ ,γ : γ ∩ → γ ∩ is the holonomy map obtained by sliding along stable curves, i.e. γ ,γ (x) = γ s (x) ∩ γ . We introduce the following notation: For x ∈ i ∩ γ , let γ be such that f Ri (γ ) ⊂ γ . Then J u ( f R )(x) = Jm γ ,m γ ( f Ri |(γ ∩ i ))(x) is the Jacobian of f R with respect to the measures m γ and m γ . Lemma 1 of [Y], which we recall below, is key to the differential structure on . Lemma 4.1. There is a function u : → R such that for each γ ∈ u ( ), if m γ is the measure whose density with respect to µγ is eu Iγ ∩ , then we have the following: (1) For all γ , γ ∈ u ( ), (γ ,γ )∗ m γ = m γ . (2) J u ( f R )(x) = J u ( f R )(y) for all y ∈ γ s (x). (3) ∃C1 > 0 (depending on C and α) such that for each i and all x, y ∈ i ∩ γ , u R J ( f )(x) s( f R x, f R y)/2 . J u ( f R )(y) − 1 ≤ C1 α
(1)
1
The properties of u include |u| ≤ C and |u(x) − u(y)| ≤ 4Cα 2 s(x,y) on each γ . (1) and (2) together imply that there is a natural measure m on with respect to which the Jacobian of F, J F, is well defined: First, identify 0 with γ ∩ for any γ ∈ u ( ), and let m|0 be the measure that corresponds to m γ . (1) says that m so R
defined is independent of γ , and (2) says that with respect to m, J F (x) = J u ( f R )(y) for any y ∈ γ s (x). We then extend m to ∪>0 in such a way that J F ≡ 1 on all of −1 \ F ( 0 ). In the rest of Sect. 4.1 we will assume m{R > n} < C0 θ0n for some C0 ≥ 1 and θ0 < 1.6 One of the reasons for passing from the hyperbolic tower to the expanding tower is that the spectral properties of the transfer√operator associated with the latter can be leveraged. We fix β with 1 > β > max{θ0 , α}, and define a symbolic metric on by √ dβ (x, y) = β s(x,y) . Since β > α, Lemma 4.1(3) implies that J F is log-Lipshitz with 6 Our default rule is to use the same symbol for corresponding objects for f, F and F when no ambiguity can arise given context. Thus R is the name of the return time function on , 0 and 0 .
376
M. Demers, P. Wright, L.-S. Young
respect to this metric. A natural function space on is B = {ρ ∈ L 1 ( , m) : ρ < ∞}, where ρ = ρ∞ + ρLip and ρ∞ = sup sup |ρ(x)|β ,
ρLip = sup Lip(ρ| , j )β .
, j x∈ , j
, j
Lip(·) above is with respect to the symbolic metric dβ . The weights β provide the needed contraction from one level to the next, and β > θ0 is needed to maintain exponential tail estimates. 4.2. Towers with Markov holes. Now consider a leaky system ( f, M, H ) as defined in Sect. 2.1, and suppose ( , R) is a generalized horseshoe respecting the hole H . Let F : → be the associated tower map with π : → M, and let H˜ = π −1 (H ). Then (F, , H˜ ) is a leaky system in itself. With the horseshoe respecting H , we have that H˜ is the union of a collection of , j , usually an infinite number of them; we refer to holes of this type as “Markov holes”. The notation H := H˜ ∩ will be used. Projecting and letting H = π ( H˜ ), we obtain the quotient leaky system (F, , H ). Let us say (F, , H˜ ) and (F, , H ) are mixing if the surviving dynamics of the horseshoe that gives rise to these towers are mixing; see Sect. 2.1. ˚ = \ H˜ , we introduce the notation Letting n ˚ = {x ∈ : F i x ∈ n = ∩i=0 F −1 / H˜ for 0 ≤ i ≤ n},
˚ = 0 . Corresponding objects for (F, , H ) are denoted by n . so that in particular 4.2.1. What is known: Spectral properties of expanding towers. Expanding towers (that are not necessarily quotients of hyperbolic towers) with Markov holes were studied in [D1 and BDM]. The following theorem summarizes several results proved in [BDM, Proposition 2.4, Corollary 2.5], under some conditions on the tower that are easily satisfied here. We refer the reader to [BDM] for detail, and state their results in our context of (F, , H ). Let B˚ = {ρ ∈ L 1 ( 0 , m) : ρ < ∞}, where ρ is as above, and let L denote the ˚ i.e., for ρ ∈ B˚ and x ∈ 0 , transfer operator associated with F| 1 defined on B, Lρ(x) = ρ(y)(J F(y))−1 . y∈ 0 ∩F
−1
x
Theorem 4.2 [BDM]. Let (F, , H ) be such that (i) (F, ) has exponential return times and (ii) (F, , H ) is mixing. Assume the following condition on hole size: ≥1
β −(−1) m(H ) <
(1 − β)m( 0 ) . 1 + C1
(2)
Then the following hold: modulus; ϑ∗ is real (1) L is quasi-compact with a unique eigenvalue ϑ∗ of maximum and > β, and it has a unique eigenfunction h ∗ ∈ B˚ with h ∗ dm = 1. In addition, ˚ there exist constants D > 0 and τ < 1 such that for all ρ ∈ B,
Escape Rates and Physically Relevant Measures for Billiards with Small Holes n
ϑ∗−n L ρ − d(ρ)h ∗ ≤ Dρτ n ,
where d(ρ) = lim λ−n n→∞
(2) The eigenvalue ϑ∗ satisfies ϑ∗ > 1 −
1+C1 m( 0 )
≥1 β
377
n
ρ dm < ∞.
−(−1) m(H ).
The spectral property of L as described in Theorem 4.2(1) implies that all ρ except for those in a codimension 1 subspace have d(ρ) = 0. Given the pivotal role played by the base 0 of the tower , one would guess that for a density ρ, if ρ > 0 on 0 , then d(ρ) = 0. A slightly more general condition is given in Corollary 4.3 below. We call , j a surviving element of the tower if some part of , j returns to 0 before entering H . Corollary 4.3 [BDM]. Let ρ ∈ B˚ be a nonnegative function that is > 0 on a surviving , j . Then d(ρ) > 0. 4.2.2. What is desired: Results for hyperbolic towers. Here we formulate a set of results for the hyperbolic tower that connect the results in Sect. 4.2.1 to the stated theorems for billiards. Let G˜ be the class of measures η on with the following properties: (i) η has absolutely continuous conditional measures on unstable leaves; and (ii) π ∗ η = ρdm for some ρ ∈ B˚ with d(ρ) > 0. Let ( (σ ) , R (σ ) ) be a generalized horseshoe with the properties in Proposition 2.2, and let (F, ) be its associated tower. Let n(, H˜ ) := sup{ : H = ∅}, i.e., n(, H˜ ) = n( (σ ) , R (σ ) ; Hσ ) as defined in Sect. 2.2. Theorem 4.4. Assume that n(, H˜ ) is large enough that
β −(−1) m( ) <
≥n(, H˜ )
(1 − β)m( 0 ) . 1 + C1
(3)
Then the following hold: ˜ (a) There exists ϑ∗ < 1 such that for all η ∈ G, log ϑ∗ = lim
n→∞
1 log η(n ). n
(b) There exists a conditionally invariant distribution µ˜ ∗ ∈ G˜ with escape rate − log ϑ∗ ˜ if ρ is the density of π ∗ η and d(ρ) is as for which the following hold: For all η ∈ G, in Theorem 4.2, then lim
n→∞
F˚∗n η = µ˜ ∗ ˚ F˚∗n η()
and
lim ϑ −n F˚∗n η n→∞ ∗
= d(ρ) · µ˜ ∗ ,
where the convergence is in the weak* topology. (c) µ˜ ∗ has absolutely continuous conditional measures on unstable leaves. Remark 4.5. In Sect. 5.1 we show that π ∗ µ˜ ∗ = h ∗ m. Thus µ˜ ∗ ∈ G˜ and the ϑ∗ of Theorem 4.2 is the same as the ϑ∗ of Theorem 4.4. Theorem 4.4 treats one hole at a time. The following uniform bounds are also needed, mostly for purposes of proving Theorem 4. Proposition 4.6. Consider all (F, , H˜ ) arising from any ( (σ ) , R (σ ) ) in Proposition 2.2 for which the hole condition in (3) is met. Let µ˜ ∗ and ϑ∗ be as in Theorem 4.4. Then there are constants C2 , K > 0 such that
378
M. Demers, P. Wright, L.-S. Young
(i) the conditional densities ργ of µ˜ ∗ | with respect to µγ on unstable leaves satisfy C2−1 ϑ∗− ≤ ργ ≤ C2 ϑ∗− ; (ii) µ˜ ∗ (∪>L ) ≤ Kβ −L θ0L ; and (iii) ϑ∗ → 1 as n(, H˜ ) → ∞. 5. Proofs of Theorems on the Tower The following notational abbreviations are used only in this section: – We will sometimes drop the ˜ used to distinguish between objects on M and corresponding objects on ; there can be no ambiguity as long as we restrict ourselves to the towers. ˚ Specifically, F∗n η is to be interpreted as F˚∗n η, and – We will at times drop the ˚ in F. n F ∗ η is to be interpreted the same way. We focus on the stable direction, since that is what lies between Theorem 4.2 and The˚ that are Lipschitz in the stable orem 4.4. The following is a class of test functions on s s s direction. For γ ∈ () and x, y ∈ γ , we denote by d s (x, y) the distance between π(x) and π(y) according to the p-metric, so that d s (F n x, F n y) ≤ λ−n d s (x, y) for some λ > 1 (see Sect. 2.3). Let Fb be the set of bounded, measurable functions on ˚ For ϕ ∈ Fb , we define |ϕ|sLip to be the Lipshitz constant of ϕ restricted to stable . leaves, i.e. |ϕ|sLip =
sup
sup
˚ x,y∈γ s γ s ∈ s ()
ϕ(x) − ϕ(y) , ds (x, y)
˚ = {ϕ ∈ Fb : |ϕ|sLip < ∞}. and let Lips () 5.1. Proof of Theorem 4.4. A. Escape rates. Theorem 4.4(a) follows easily from Theorem 4.2 as (F, , H ) and (F, , H ) have the same escape rate. In more detail, let η ∈ G˜ and notice that since H is a union of , j , we have, for each n, η(n ) = η( n ), ˜ dη = ρ ∈ B˚ with d(ρ) > 0. Theorem 4.2(1) then where η = π ∗ η. By definition of G, dm n implies that ϑ∗−n L ρ converges to d(ρ)h ∗ . Since the convergence is in the · -norm, n we may integrate with respect to m. Noting that L ρ dm = n ρ dm = η( n ), we have lim ϑ −n η(n ) n→∞ ∗
= lim ϑ∗−n η( n ) = d(ρ). n→∞
(4)
Thus − log ϑ∗ , where ϑ∗ is the eigenvalue in Theorem 4.2, is the common escape rate ˜ of (F, , H ) for initial distributions in G. B. Uniqueness of limiting distributions. We first prove uniqueness postponing the proof of existence of limiting distributions. ˜ we define a measure ηs on u (), i.e. a measure transverse to unstaGiven η ∈ G, ble leaves, as follows: Set ηs ( u (, j )) = 0 if η(, j ) = 0. If η(, j ) = 0, then ηs | u (, j ) is the factor measure of η|, j normalized, and {ρdm γ , γ ∈ u (, j )} is the disintegration of η into measures on unstable leaves. We will use the convention that ηs (, j ) = 1, and ρ|γ is the density with respect to m γ , so that ρ|γ dηs (γ ) = ρ, where dπ ∗ η = ρdm.
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
379
˜ Suppose for i = 1, 2, there exists µi∗ such that Lemma 5.1. Let η1 and η2 ∈ G. lim ϑ −n F∗n ηi n→∞ ∗
= d(ρ i )µi∗ ,
where ρ i is the density of π ∗ ηi Then µ1∗ = µ2∗ . The crux of the argument for Lemma 5.1 is contained in ˚ Lemma 5.2. Let η1 and η2 be as above, and assume ρ 1 = ρ 2 . Then for all ϕ ∈ Lips (), −n n n ϑ∗ |F∗ η1 (ϕ) − F∗ η2 (ϕ)| → 0 exponentially fast as n → ∞. Proof. For i = 1, 2, let ηis and ρi be the (normalized) factor measure and (unnormalized) densities on γ ∈ u () of ηi as described above. We consider functions which are constant along stable leaves to be defined on both ˚ and 0 and do not distinguish between the two versions of such functions. For each , j , let γˆ ∈ u (, j ) be a representative leaf. Then |F∗n η1 (ϕ) − F∗n η2 (ϕ)| ≤
, j
γˆ ∩n, j
dm γˆ
γs
ρ1 ϕ ◦ F n dη1s −
γs
ρ2 ϕ ◦ F n dη2s . (5)
Next fix x ∈ γˆ ∩ n and estimate the integrals on γ s (x). Define ϕ n = Then, n s n s s ρ1 ϕ ◦ F dη1 − s ρ2 ϕ ◦ F dη2 γ γ n s n s ≤ ρ1 (ϕ ◦ F − ϕ n )dη1 + ρ2 (ϕ ◦ F − ϕ n )dη2 γs γs + ϕ n ρ1 dη1s − ϕ n ρ2 dη2s . γs
γs
ϕ ◦ F n dη1s .
γs
Since ϕ n is constant on γ s and ρ 1 = ρ 2 , the third term above is 0. For the first two terms, we note that for each y ∈ γ s (x), |ϕ n (y) − ϕ ◦ F n (y)| ≤ |ϕ|sLip λ−n . Thus ϑ∗−n |F∗n µ1 (ϕ) − F∗n µ2 (ϕ)| ≤ ϑ∗−n = n
n, j
2ρ 1 dm |ϕ|sLip λ−n
, j −n n 2ϑ∗ |L ρ 1 |1 |ϕ|sLip λ−n ,
which proves the lemma since ϑ∗−n |L ρ 1 | → d(ρ 1 ) by Theorem 4.2.
(6)
Remark 5.3. We have used in the proof above a property of the billiard maps, namely d s (F n x, F n y) ≤ λ−n d s (x, y). For general towers, one has only the contraction guaranteed by (P3) which is nonuniform. It is not hard to see that the lemma holds in the n more general case with the exponential rate given by max{α 2 , β −n θ0n } in the place of λ−n ; we leave the proof to the interested reader.
380
M. Demers, P. Wright, L.-S. Young
Proof of Lemma 5.1. Let µ∗ = h ∗ m be the conditionally invariant measure given by Theorem 4.2. For i = 1, 2, we have, on the one hand, n lim ϑ −n F ∗ ηi n→∞ ∗
= d(ρ i )µ∗ ,
which follows from Theorem 4.2, and on the other, lim ϑ −n π ∗ F∗n ηi n→∞ ∗
= d(ρ i )π ∗ µi∗ , n
which follows from the hypothesis of the lemma. Since π ∗ F∗n ηi = F ∗ π ∗ ηi for each n ≥ 0, we have π ∗ µ1∗ = µ∗ = π ∗ µ2∗ . Thus ϑ∗−n |F∗n µ1∗ − F∗n µ2∗ | → 0 as n → ∞ by Lemma 5.2. But ϑ∗−n F∗n µi∗ = µi∗ since µi∗ is conditionally invariant. Hence µ1∗ = µ2∗ . ˚ C. Convergence to conditionally invariant measure. For a probability measure η on , n n n −n n ˜ |F∗ η| = η( ) = π ∗ η( ). So for η ∈ G, (4) implies limn→∞ ϑ∗ |F∗ η| = d(ρ) > 0, where ρ is the density of η = π ∗ η. More than that is true: Lemma 5.4. ϑ∗−n F∗n η/d(ρ) converges weakly to a conditionally invariant probability measure µ∗ as n → ∞. This is half of Theorem 4.4(b). Once we have this, it will follow immediately that F∗n η ϑ∗−n F∗n η = lim = µ∗ , n→∞ |F∗n η| n→∞ ϑ∗−n |F n η| ∗ lim
(7)
which is the other half. We will use the following algorithm to “lift” measures from to : Fix a measure µs on u () with µs ( u (, j )) = 1. Given η on with density ρ, we define π −1 ∗ η to η decomposes be the measure on with the property that restricted to each , j , π −1 ∗ into the factor measure µs and leaf measures {ρdm γ }, where ρ|π −1 (x) ≡ ρ(x). Notice that π ∗ π −1 ∗ η = η. ˚ and show that ϑ∗−n F∗n η(ϕ) Proof of Lemma 5.4. Our first step is to fix ϕ ∈ Lips () s is a Cauchy sequence. For a fixed µ as above, let ϕ n (x) = γ s (x) ϕ ◦ F˚ n dµs . Define ˜ η has density ρ ∈ B˚ with d(ρ) > 0. Then by definition of π −1 η = π ∗ η. Since η ∈ G, ∗ , ˚n (π −1 dµs (γ ) ϕ ◦ F˚ n ρ dm γ ∗ π ∗ η)(ϕ ◦ F ) = =
, j
u (, j )
, j
, j
γu
ρ ϕ n dm = π ∗ η(ϕ n ).
(8)
For n, k1 , k2 ≥ 0, write |ϑ∗−n−k1 F∗n+k1 η(ϕ) − ϑ∗−n−k2 F∗n+k2 η(ϕ)
k1 ≤ ϑ∗−n−k1 |F∗n+k1 η(ϕ) − F∗n π −1 ∗ π ∗ F∗ η(ϕ)|
k1 −n−k2 n −1 F∗ π ∗ π ∗ F∗k2 η(ϕ)| +|ϑ∗−n−k1 F∗n π −1 ∗ π ∗ F∗ η(ϕ) − ϑ∗ k2 n+k2 +ϑ∗−n−k2 |F∗n π −1 η(ϕ)|. ∗ π ∗ F∗ η(ϕ)η(ϕ) − F∗
(9)
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
381
The first and third terms of (9) are estimated using Lemma 5.2 since π ∗ (ϑ∗−ki F∗ki η) = ki π ∗ (ϑ∗−ki π −1 ∗ π ∗ F∗ η) for i = 1, 2. Thus by Lemma 5.2, ki s n ϑ∗−n−ki |F∗n+ki η(ϕ) − F∗n π −1 ∗ π ∗ F∗ η(ϕ)| ≤ C d(ρ)(|ϕ|Lip + |ϕ|∞ )ζ
for some C > 0 and ζ < 1. We now fix n and estimate the second term of (9). Due to (8), for any k ≥ 0 we have k −n−k −1 ϑ∗−n−k F∗n π −1 π ∗ π ∗ F∗k η(ϕ ◦ F n · 1˚ n ) = ϑ∗−n−k π ∗ F∗k η(ϕ n · 1 n ) ∗ π ∗ F∗ η(ϕ) = ϑ∗ k k ϕ n · L ρ dm. = ϑ∗−n−k F ∗ η(ϕ n · 1 n ) = ϑ∗−n−k n
˜ we estimate Recalling that ρ ∈ B˚ and d(ρ) > 0 since η ∈ G, k1 −n−k2 n −1 |ϑ∗−n−k1 F∗n π −1 F∗ π ∗ π ∗ F∗k2 η(ϕ)| ∗ π ∗ F∗ η(ϕ) − ϑ∗ −k1 k1 ≤ |ϕ|∞ ϑ∗−n ϑ∗ L ρ − d(ρ)h ∗ dm n −k2 k2 −n +|ϕ|∞ ϑ∗ ϑ∗ L ρ − d(ρ)h ∗ dm.
(10)
n
Both terms of ϑ∗−n
n
(10) are small: By Theorem 4.2(1), −k k −n −k k dm L ρ − d(ρ)h ≤ ϑ L ρ − d(ρ)h ϑ∗ ∗ ∗ ∗ ϑ∗ n
≤ ϑ∗−n |L 1β |1 Dρτ k
n
1β dm
˚ ϑ −n |Ln 1β |1 converges ˚ . Since 1β ∈ B, for k = ki , where 1β (x) = β − for x ∈ ∗ to d(1β ) as n → ∞. Thus (10) can be made arbitrarily small by choosing k1 and k2 sufficiently large. We have shown that ϑ∗−n F∗n η(ϕ)/d(ρ) is a Cauchy sequence and therefore converges to a number Q(ϕ). The functional Q(ϕ) := limn→∞ ϑ∗−n F∗n η(ϕ)/d(ρ) is clearly linear in ϕ, positive and satisfies Q(1) = 1. Also |Q(ϕ)| ≤ |ϕ|∞ Q(1) so that Q extends to a ˚ the set of bounded functions which are continuous bounded linear functional on Cb0 (), ˚ on each , j . By the Riesz representation theorem, there exists a unique Borel probability measure ˚ [H, Sect. 56]. Also, µ∗ satisfying µ∗ (ϕ) = Q(ϕ) for each ϕ ∈ Cb0 () ˚ = lim ϑ∗−n F∗n η(ϕ ◦ F) ˚ = ϑ∗ lim ϑ∗−n−1 F∗n+1 η(ϕ) = ϑ∗ d(ρ)µ∗ (ϕ), d(ρ)µ∗ (ϕ ◦ F) n→∞
n→∞
so that µ∗ is a conditionally invariant measure for F˚ with escape rate − log ϑ∗ .
This completes the proof of parts (a) and (b) of Theorem 4.4. To prove part (c), we must show that µ∗ has absolutely continuous conditional measures on unstable leaves. Proof of a stronger version of this fact is contained in the proof of Proposition 4.6(i) below. Notice that from the proof of Lemma 5.1, we have π ∗ µ∗ = µ∗ so that the density of ˜ π ∗ µ∗ is precisely h ∗ . Since h ∗ ∈ B˚ and d(h ∗ ) = 1 > 0, we conclude µ∗ ∈ G.
382
M. Demers, P. Wright, L.-S. Young
5.2. Proof of Proposition 4.6. Consider the set of holes h (q0 ) for fixed q0 where h is small enough as required in Proposition 2.2. For σ ∈ h (q0 ), let (σ ) be the tower with ) holes induced by the generalized horseshoe, and let µ(σ ∗ be the conditionally invariant measure given by Theorem 4.4. Proof of (i). We drop the superscript (σ ) in what follows and point out that the constants we use are uniform for all σ ∈ h (q0 ) and h sufficiently small. Choose γ0 ∈ u (0 ) and let η0 be the measure supported on γ0 with uniform den˜ It is immediate that π ∗ η0 has density sity with respect to µγ . We claim that η0 ∈ G. −u ˚ ρ = e |γ0 with respect to m, which is in B by Lemma 4.1. To see that d(ρ) > 0, notice that the mixing assumption on (F, ) implies that 0 is necessarily a surviving partition element. Since ρ > 0 on 0 , Corollary 4.3 implies that d(ρ) > 0. (n) By Theorem 4.4(b), η(n) := F∗n η0 /|F∗n η0 | converges to µ∗ . Let ργ denote the density of η(n) with respect to µγ on γ ∈ u (). Notice that inverse branches of F˚ n on γ are well defined. For any x1 , x2 ∈ γ , treating one branch at a time and summing over all branches, we obtain that −1 (n) ˚n Jµγ F˚ n (y2 ) ργ (x1 ) y1 ∈ F˚ −n x1 (Jµγ F (y1 )) = ≤ sup ≤ eC (n) n (y ))−1 n (y ) ˚ ˚ F F (J J −n ργ (x2 ) ˚ −n µ 2 µ 1 ˚ y ∈F x γ γ y2 ∈ F
x2
1
1
by Property (P4)(b), where Jµγ F˚ n is the Jacobian of F˚ n with respect to µγ . Since by Proposition 2.2 the constant C is independent of σ , x and n, we have e−C ≤
(n)
supx∈γ ργ (x) inf x∈γ ργ(n) (x)
≤ eC .
(11)
This estimate plus the minimum length κ of µu ( ) given by Proposition 2.2 yields the desired uniform upper and lower bounds on the conditional densities of η(n) with respect to µγ (and hence to m γ ) on 0 . The uniformity of these bounds in n implies that they pass to the conditional densities of µ∗ in the limit as n → ∞. Since µ∗ is conditionally invariant, µ∗ |˚ = ϑ∗−1 µ∗ | F˚ −1 ˚ . The required bounds on the densities extend easily to ˚ for > 0. Proof of (ii). We decompose µ∗ into a normalized factor measure µs∗ on u ( ) and densities ργ with respect to m γ on γ ∈ u ( ). Then (σ ) µ∗ (∪≥L ) = dµs∗ ργ dm γ ≤ C2 ϑ∗− m( ) ≤ C2 C0 θ0 β − . (σ )
u ≥L ( )
γ
≥L
≥L
Here we have used Proposition 4.6(i) to estimate ργ , Proposition 2.2 and Lemma 4.1 for the uniformity of C0 and θ0 , and the fact that ϑ∗ > β. The sum can be made arbitrarily small since β > θ0 . Proof of (iii). Notice that n(, H˜ ) ≥ n(h) ¯ by definition of n(h) ¯ in Sect. 2.2. From Theorem 4.2(2), we know that the escape rate − log ϑ∗ satisfies 1 + C1 −1 1 + C1 −1 ϑ∗ > 1 − β m(H ∩ ) > 1 − β C0 θ0 . κ κ ≥1
By Proposition 2.2, n(h) ¯ → ∞ as h → 0, so that ϑ∗ → 1.
≥n(h) ¯
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
383
6. Proofs of Theorems for Billiards In the Proofs of Theorems 1–3, we fix a hole σ that is acceptable with respect to Proposition 2.2 and for which n( (σ ) , R (σ ) , Hσ ) is large enough to meet the condition in Theorem 4.4. We suppress mention of σ , and let (F, , H˜ ) be the tower constructed n ˚ M ∞ = ∩n≥0 M n . from ( , R, H ). Define M n = ∩i=0 f −n M, 6.1. Proof of Theorems 1 and 2. The first order of business is to show that each η ∈ G ˚ with can be lifted to a measure η˜ ∈ G˜ in such a way that the escape dynamics on initial distribution η˜ reflect those on M˚ with initial distribution η. Recall that the natural invariant probability measure for the closed billiard system f : M → M is denoted by ν. In [Y, Sect. 2], it is shown that there is a unique invariant probability measure ν˜ for the tower map F : → with absolutely continuous conditional measures on unstable leaves, and this measure has the property π∗ ν˜ = ν. Given η ∈ G, we define η˜ on as follows: By definition, every η ∈ G is absolutely continuous with respect to ν. Let ˜ ˜ where ψ˜ = ψ ◦ π . This ψ = dη dν . We take η˜ to be the measure given by d η˜ = ψd ν, implies in particular that π∗ η˜ = η. ˜ Lemma 6.1. If η ∈ G, then η˜ ∈ G. As before, let Fb denote the set of bounded functions on . For ϕ ∈ Fb and γ ∈ u (), we let Lipu (ϕ|γ ) be the Lipschitz constant of ϕ|γ with respect to the dβ -metric (notice that dβ , the symbolic metric defined on , can be thought of as a metric on unstable leaves). Let |ϕ|uLip =
sup
γ ∈ u ()
Lipu (ϕ|γ ),
and Lipu () = {ϕ ∈ Fb : |ϕ|uLip < ∞}. The first step toward proving Lemma 6.1 is ˜ uLip ≤ Lemma 6.2. Let ϕ : M → R be Lipschitz. Then ϕ˜ := ϕ ◦ π ∈ Lipu () with |ϕ| CLip(ϕ). Proof. Recall that for x, y ∈ M lying in a piece of local unstable manifold, we have d(x, y) ≤ p(x, y)1/2 , where p(·, ·) is the p-metric (see Sect. 2.3). Now for γ ∈ u () and x, y ∈ γ , we have 1
|ϕ(x) ˜ − ϕ(y)| ˜ = |ϕ(π x) − ϕ(π y)| ≤ Lip(ϕ)d(π x, π y) ≤ Lip(ϕ) p(π x, π y) 2 . 1
By (P4)(a), p(π x, π y) 2 ≤ Cα s(π x,π y)/2 , which is ≤ Cdβ (x, y) since s ≤ s0 and √ β ≥ α. ˚ Let ψ = Proof of Lemma 6.1. (i) First we show π ∗ η˜ = ρ m with ρ ∈ B. s u disintegrating η˜ into η˜ and {ργ dm γ , γ ∈ ()}, we obtain ργ := ψ˜ ·
Then
d ν˜ dµγ · . dµγ dm γ
Now ψ˜ is bounded by assumption and is ∈ Lipu () by Lemma 6.2, Lipu ()
dη dν .
dµγ dm γ
d ν˜ dµγ
is bounded
([Y], Sect. 2), as is (Lemma 4.1). Thus we conclude that ργ ∈ and is ∈ u Lip () and is bounded. Recall that ρ(x) = γ s (x) ργ d η˜ s . It follows immediately that |ρ|∞ ≤ supγ |ργ |∞ and Lip(ρ) ≤ supγ Lipu (ργ ).
384
M. Demers, P. Wright, L.-S. Young
(ii) It remains to show d(ρ) > 0. By definition of G, ψ > 0 on M ∞ , the set of points −u are ˚ so ψ˜ > 0 on ∞ . The fact that d ν/dµ which never escape from M, ˜ γ and e ∞ strictly positive implies that ρ > 0 on ; hence it is > 0 on a surviving cylinder set, k i.e. a set E k such that F maps E k onto a surviving , j before any part of it enters the k n hole. By Corollary 4.3, d(L ρ) > 0. Since n g dm = L g dm for each n ≥ 0 and k
g ∈ L 1 (m), we have d(L ρ) = ϑ∗k d(ρ) so that d(ρ) > 0 as well.
Proof of Theorems 1 and 2. Given η ∈ G, let η˜ be as defined earlier. Then η˜ ∈ G˜ by Lemma 6.1. For ϕ ∈ C 0 (M), let ϕ˜ = ϕ ◦ π . Then ϕ˜ ∈ Cb0 () and for n ≥ 0 we have, f˚∗n η(ϕ) = η(ϕ ◦ f n · 1 M n ) = η( ˜ ϕ˜ ◦ F n · 1n ) = F˚∗n η( ˜ ϕ). ˜
(12)
˚ = F˚∗n η( ˚ = η( Setting ϕ ≡ 1 in (12), we have η(M n ) = f˚∗n η( M) ˜ ) ˜ n ) for n > 0, so lim
n→∞
1 1 log η(M n ) = lim log η( ˜ n ) = log ϑ∗ n→∞ n n
by Theorem 4.4(a). This proves Theorem 1. Let µ∗ = π∗ µ˜ ∗ , where µ˜ ∗ is given by Theorem 4.4. Then µ∗ (ϕ) = µ˜ ∗ (ϕ), ˜ and f˚∗ µ∗ (ϕ) = F˚∗ µ˜ ∗ (ϕ) ˜ = ϑ∗ µ˜ ∗ (ϕ) ˜ = ϑ∗ µ∗ (ϕ), proving µ∗ is conditionally invariant. Using (12) again, the fact that the normalizations are equal, and Theorem 4.4(b), we obtain lim
n→∞
f˚∗n η(ϕ) = lim n→∞ ˚ f˚∗n η( M)
F˚∗n η( ˜ ϕ) ˜ ˜ = µ∗ (ϕ). = µ˜ ∗ (ϕ) n ˚ ˚ F∗ η()
Thus f˚∗n η/η(M n ) → µ∗ weakly. Finally, lim ϑ −n f˚∗n η(ϕ) n→∞ ∗
= lim ϑ∗−n F˚∗n η( ˜ ϕ) ˜ = d(ρ) · µ˜ ∗ (ϕ) ˜ = d(ρ) · µ∗ (ϕ), n→∞
˜ This completes the proof of Theorem 2. where d(ρ) > 0 since η˜ ∈ G.
Remark 6.3. In the proof of Lemma 6.1, step (i) holds for any η that has Lipschitz densities on unstable leaves. Thus for this class of measures, Theorem 2(b) holds (with c(η) possibly equal to zero). It is also clear from step (ii) that to show d(ρ) > 0, it suffices to assume ψ > 0 on M ∞ ∩ , or on M ∞ ∩π(, j ), where π (, j ) is any surviving element. 6.2. Proof of Theorem 3. Let µ∗ = π∗ µ˜ ∗ be as above. (a) Since f˚∗ µ∗ = ϑ∗ µ∗ , it follows that µ∗ is supported on M \ ∪n≥0 f n (H ), where H = Hσ . This set has Lebesgue measure zero since by the ergodicity of f , ∪n≥0 f n (H ) has full Lebesgue measure. Thus µ∗ is singular with respect to Lebesgue measure. (b) First, we argue that µ∗ has absolutely continuous conditional measures on unstable leaves (without claiming that the densities are strictly positive). This is true because for each , j, µ˜ ∗ |, j has absolutely continuous conditional measures on γ ∈ u (, j ), and π |, j , which is one-to-one, identifies each γ with a positive Lebesgue measure subset of a local unstable manifold of f .
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
385
The rest of the proof is concerned with showing that the conditional densities of µ∗ are strictly positive. To do that, it is not productive to view µ∗ as π∗ µ˜ ∗ . Instead, we will view µ∗ as the weak limit of ν (n) := f˚∗n ν/| f˚∗n ν| as n → ∞, where ν is the natural invariant measure for f . This convergence of ν (n) is guaranteed by Theorem 2. We will prove that µ∗ has the properties immediately following the statement of Theorem 3 in Sect. 1.3. Step 1: Our first patch is built on V = ∪{γ u : γ u ∈ u ( )}, where u = u ( ) is the defining family of unstable curves for . To understand the geometric properties of ν (n) |V , observe that in backward time, each γ u ∈ u either falls into the hole completely or stays out completely. This is because f (∂ H ) is regarded as part of the discontinuity set for f −1 when we constructed the horseshoe (see Sect. 3.2). Thus there is a decreasing sequence of sets Un = ∪{γ u ∈ u : f −i γ u ∩ H = ∅ for all 0 ≤ i ≤ n} ⊂ V consisting of whole γ u -curves. Assuming ν(Un ) > 0 for now, we have ν (n) |V = cn ν|Un for some constant cn > 0 as ν is f -invariant. Let ζ be a limit point of ν (n) |V , i.e., ζ = limn k ν (n k ) |V . Assuming ζ (V ) > 0, lower bounds for conditional probability densities of ν (n k ) |V , equivalently those of ν|Un , are passed to ζ , and these bounds are strictly positive. To see that ζ (V ) > 0, recall that ν = π∗ ν˜ for some ν˜ on the tower , so that ν (n) = π∗ ν˜ (n) , where ν˜ (n) = F˚∗n ν˜ /| F˚∗n ν˜ |. Since π(0 ) ⊂ V , we have ζ = lim ν (n k ) |V ≥ lim π∗ (˜ν (n k ) |0 ) = π∗ (µ˜ ∗ |0 ). nk
nk
We have written an inequality (as opposed to equality) above because parts of for ≥ 1 may get mapped into V as well. Clearly, µ˜ ∗ (0 ) > 0, thereby ensuring ζ (V ) > 0, hence ν(Un ) > 0 and the strictly positive conditional densities property above. This together with ζ ≤ µ∗ |V (equality is not claimed because it is possible for part of ν (n k ) from outside of V to leak into V in the limit) proves that (V, ζ ) is an acceptable patch. Step 2: Next we use (V, ζ ) to build patches (V, j , ζ, j ) corresponding to partition elements , j of the tower with > 0 and µ˜ ∗ (, j ) > 0. From Sects. 3 and 4, we know that π(, j ) is a hyperbolic product set, and π(, j ) = f ( s ) for some s-subset s ⊂ . Moreover, f i ( s ) ∩ H = ∅ for all 0 < i ≤ . Thus we may assume V, j = ∪{γ u : γ u ∈ u (π(, j ))} ⊂ f (V ). Let ζ, j = ϑ∗− ( f ∗ ζ )|V, j . Then ζ, j has strictly positive conditional densities on unstable curves because ζ does, and ζ, j ≤ µ∗ |V, j as µ∗ satisfies f˚∗ µ∗ = ϑ∗ µ∗ . Finally, since ζ, j ≥ π∗ (µ˜ ∗ |, j ) for each , j, it follows that , j ζ, j ≥ µ∗ , completing the proof of Theorem 3. 6.3. Proof of Theorem 4. Suppose h n is a sequence of numbers tending to 0, σh n ∈ h n (q0 ) is a sequence of holes in the billiard table, and Hn = Hσh n the corresponding holes in M. For each n, let ϑn be the escape rate and µn the physical measure for the leaky system ( f, M, Hn ) given by Theorem 2. By Proposition 4.6(iii), we have ϑn → 1 as n → ∞. To prove µn → ν, we will assume, having passed to a subsequence, that µn converges weakly to some µ∞ , and show that (i) µ∞ is f -invariant, and (ii) it has absolutely continuous conditional measures on unstable leaves. These two properties together uniquely characterize ν. The following notation will be used: (n) is the generalized horseshoe respecting the hole Hn , (n) = ∪ (n) is the corresponding tower, Fn : (n) → n is the tower
386
M. Demers, P. Wright, L.-S. Young
map, πn : (n) → M is the projection, and µ˜ n is the conditionally invariant measure on (n) that projects to µn . (i) Proof of f -invariance: Let S = ∂ M ∪ f −1 ∂ M. Lemma 6.4. µ∞ (S) = 0 Proof. Let δ1 and λ1 be as in Sect. 2.3, and let Nε (S) denote the ε-neighborhood of S. We claim that there exist constants C3 , ς > 0 such that for ε < δ1 and for all n, µn (Nε (S)) ≤ C3 ες . By the construction of = (n), any n, d( f ( ), S) ≥ δ1 λ− 1 . Thus f ( ) ∩ Nε (S) = ∅ for all ≤ − log(ε/δ1 )/ log λ1 . Hence µn (Nε (S)) ≤ µ˜ n ( (n)), >− log(ε/δ1 )/ log λ1
which by Proposition 4.6(ii) is ≤ K (β −1 θ0 )− log(ε/δ1 )/ log λ1 , proving the claim above with C3 = K /δ1 and ς = log(βθ0−1 )/ log λ1 . Since C3 and ς are independent of n, these bounds pass to µ∞ , implying µ∞ (S) = 0. Having established that f is well defined µ∞ -a.e., we now verify that µ∞ is f -invariant: Let ϕ : M → R be a continuous function. Then (ϕ ◦ f )dµ∞ = lim (ϕ ◦ f )dµn = lim ϕ d( f ∗ µn ), n→∞
and
n→∞
ϕ d( f ∗ µn ) =
M\Hn
ϕ d( f ∗ µn ) +
Hn
ϕ d( f ∗ µn ).
(13)
Since( f ∗ µn )| M\Hn = f˚∗ µn = ϑn µn , the first integral on the right side of (13) is equal to ϑn ϕ dµn , while the absolute value of the second is bounded by (1 − ϑn )|ϕ|∞ . Since ϑn → 1 as n → ∞, the right side of (13) tends to ϕ dµ∞ . (ii) Absolutely continuous conditional measures on unstable leaves: Since the measures µ˜ n do not live on the same space for different n, a first task here is to find common domains in M on which (πn )∗ µ˜ n can be compared. In the constructions to follow, the discontinuity set refers to the real discontinuity set of f , not the ones that include boundaries of holes (as was done in Sect. 3). We choose a rectangular region Qˇ slightly larger than Q in Proposition 2.2, large enough that Qˇ ⊃ (n) for all n, and let ˇ u denote the set of all homogeneous unstable ˇ Let Vˇ = ∪{γ u ∈ ˇ u }. Then (n) ⊂ Vˇ curves connecting the two components of ∂ s Q. u u ˇ ˇ for all n, for γ ∩ Q ∈ for every γ ∈ ( (n)) (defined using the enlarged discontinuity set). Now for all n, (πn )∗ (µ˜ n |0 (n) ) is a sequence of measures on Vˇ with absolutely continuous conditional measures on the elements of ˇ u . Moreover, the conditional densities are uniformly bounded from above with a bound independent of n (Proposition 4.6(i)). Let µ∞,0 be a limit point of (πn )∗ (µ˜ n |0 (n) ). Assuming µ∞,0 (Vˇ ) > 0, these density bounds are inherited by µ∞,0 . To show µ∞,0 (Vˇ ) > 0, we will argue there exists b > 0 such that µ˜ n (0 (n)) > b for all n, and that is true because the µ˜ n are probability measures, there is a uniform lower bound on µ˜ n (∪
Escape Rates and Physically Relevant Measures for Billiards with Small Holes
387
For > 0, we define Qˇ to be the finite union of s-subrectangles of Qˇ retained in steps in the construction of when f has no holes, i.e. roughly speaking, Qˇ consists of points that stay away from S = ∂ M ∪ f −1 ∂ M by a distance ≥ δ1 λ−i 1 at step i. Let − ˚ ˇ ˇ ˇ ˇ V = V ∩ Q . Then πn ( Fn (n)) ⊂ V for all n. In fact, for each j, πn ( F˚n− , j (n)) is contained in a connected component of Qˇ . The argument for µ∞,0 can now be repeated to conclude the existence of a limit point of (πn ◦ F˚n− )∗ (µ˜ n | (n) ) with absolutely continuous conditional measures on unstable leaves. Pushing all measures forward by f ∗ , this gives a limit point µ∞, of (πn )∗ (µ˜ n | (n) ) as n → ∞ with the same property. To proceed systematically, we perform a Cantor diagonal argument, choosing a single subsequence n k with the property that for each ≥ 0, (πn k )∗ (µ˜ n k | (n k ) ) converges to a measure µ∞, on f Vˇ . Finally, to conclude µ∞ = µ∞, , we need a tightness condition as the towers are noncompact. This is given by Proposition 4.6(ii). The proof of Theorem 4 is now complete. Acknowledgements. The second author (P.W.) would like to thank The Courant Institute of Mathematical Sciences, New York University, where he was affiliated when this project began. The authors thank MSRI, Berkeley, and ESI, Vienna, where part of this work was carried out.
References [BaK] [BDM] [BSC1] [BSC2] [B] [C1] [C2] [CM1] [CM2] [CM3] [CMT1] [CMT2] [CMM] [CMS1] [CMS2] [CMS3] [CV]
Baladi, V., Keller, G.: Zeta functions and transfer operators for piecewise monotonic transformations. Commun. Math. Phys. 127, 459–477 (1990) Bruin, H., Demers, M., Melbourne, I.: Convergence properties and an equilibrium principle for certain dynamical systems with holes. To appear in Ergod. Th. and Dynam. Sys. Bunimovich, L.A., Sina˘ı, Ya.G., Chernov, N.I.: Markov partitions for two-dimensional hyperbolic billiards. Russ. Math. Surv. 45(3), 105–152 (1990) Bunimovich, L.A., Sina˘ı, Ya.G., Chernov, N.I.: Statistical properties of two-dimensional hyperbolic billiards. Russ. Math. Surv. 46(4), 47–106 (1991) Buzzi, J.: Markov extensions for multidimensional dynamical systems. Israel J. of Math. 112, 357–380 (1999) Cenvoca, N.N.: A natural invariant measure on smale’s horseshoe. Soviet Math. Dokl. 23, 87–91 (1981) Cenvoca, N.N.: Statistical properties of smooth Smale horseshoes. In: Mathematical Problems of Statistical Mechanics and Dynamics, R.L. Dobrushin, ed. Dordrecht: Reidel, 1986, pp. 199–256 Chernov, N., Markarian, R.: Ergodic properties of anosov maps with rectangular holes. Bol. Soc. Bras. Mat. 28, 271–314 (1997) Chernov, N., Markarian, R.: Anosov maps with rectangular holes. Nonergodic Cases. Bol. Soc. Bras. Mat. 28, 315–342 (1997) Chernov, N., Markarian, R.: Chaotic Billiards. Number 127 in Mathematical Surveys and Monographs, Providence, RI: Amer. Math. Soc., 2006 Chernov, N., Markarian, R., Troubetzkoy, S.: Conditionally invariant measures for anosov maps with small holes. Ergod. Th. and Dynam. Sys. 18, 1049–1073 (1998) Chernov, N., Markarian, R., Troubetzkoy, S.: Invariant measures for anosov maps with small holes. Ergod. Th. and Dynam. Sys. 20, 1007–1044 (2000) Collet, P., Martínez, S., Maume-Deschamps, V.: On the existence of conditionally invariant probability measures in dynamical systems. Nonlinearity 13, 1263–1274 (2000) Collet, P., Martínez, S., Schmitt, B.: The Yorke-Pianigiani measure and the asymptotic law on the limit cantor set of expanding systems. Nonlinearity 7, 1437–1443 (1994) Collet, P., Martínez, S., Schmitt, B.: Quasi-stationary distribution and Gibbs measure of expanding systems. In: Instabilities and Nonequilibrium Structures. V. E. Tirapegui, W. Zeller, eds. Dordrecht: Kluwer, 1996, pp. 205–219 Collet, P., Martínez, S., Schmitt, B.: The Pianigiani-Yorke measure for topological Markov chains. Israel J. Math. 97, 61–70 (1997) Chernov, N., van dem Bedem, H.: Expanding maps of an interval with holes. Ergod. Th. and Dynam. Sys. 22, 637–654 (2002)
388
[D1] [D2] [DL] [DY] [FKMP] [H] [HY] [KS] [LiM] [LM] [PY] [R] [S] [V] [W] [Y]
M. Demers, P. Wright, L.-S. Young
Demers, M.: Markov extensions for dynamical systems with holes: an application to expanding maps of the interval. Israel J. of Math. 146, 189–221 (2005) Demers, M.: Markov extensions and conditionally invariant measures for certain logistic maps with small holes. Ergod. Th. and Dynam. Sys. 25(4), 1139–1171 (2005) Demers, M., Liverani, C.: Stability of statistical properties in two-dimensional piecewise hyperbolic maps. Trans. Amer. Math. Soc. 360(9), 4777–4814 (2008) Demers, M., Young, L.-S.: Escape rates and conditionally invariant measures. Nonlinearity 19, 377–397 (2006) Ferrari, P.A., Kesten, H., Martínez, S., Picco, P.: Existence of quasi-stationary distributions. A Renewal Dynamical Approach. Annals of Prob. 23(2), 501–521 (1995) Halmos, P.R.: Measure Theory. University Series in Higher Mathematics, Princeton, NJ: D. Van Nostrand Co., Inc., 1950, 304 p. Homburg, A., Young, T.: Intermittency in families of unimodal maps. Ergod. Th. and Dynam. Sys. 22(1), 203–225 (2002) Katok, A., Strelcyn, J.M.: Invariant Manifolds, Entropy and Billiards; Smooth Maps with Singularities. Volume 1222, Springer Lecture Notes in Math., Berlin-Heidelberg-NewYork: Springer, 1986 Liverani, C., Maume-Deschamps, V.: Lasota-Yorke maps with holes: conditionally invariant probability measures and invariant probability measures on the survivor set. Annales de l’Institut Henri Poincaré Probability and Statistics 39, 385–412 (2003) Lopes, A., Markarian, R.: Open billiards: cantor sets, invariant and conditionally invariant probabilities. SIAM J. Appl. Math. 56, 651–680 (1996) Pianigiani, G., Yorke, J.: Expanding maps on sets which are almost invariant: decay and chaos. Trans. Amer. Math. Soc. 252, 351–366 (1979) Richardson, P.A., Jr.: Natural Measures on the Unstable and Invariant Manifolds of Open Billiard Dynamical Systems. Doctoral Dissertation, Department of Mathematics, University of North Texas, 1999 Sina˘ı, Ya.G.: Dynamical systems with elastic collisions. Ergodic Properties of Dispersing Billiards. Usp. Mat. Nauk 25(2), 141–192 (1970) Vere-Jones, D.: Geometric ergodicity in denumerable Markov chains. Quart. J. Math. 13, 7–28 (1962) Wojtkowski, M.: Invariant families of cones and Lyapunov exponents. Ergod. Th. Dynam. Sys. 5(1), 145–161 (1985) Young, L.-S.: Statistical properties of dynamical systems with some hyperbolicity. Annals of Math. 147(3), 585–650 (1998)
Communicated by G. Gallavotti
Commun. Math. Phys. 294, 389–410 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0938-6
Communications in
Mathematical Physics
Argyres-Seiberg Duality and the Higgs Branch Davide Gaiotto1 , Andrew Neitzke2 , Yuji Tachikawa1 1 School of Natural Sciences, Institute for Advanced Study, Princeton,
New Jersey 08540, USA. E-mail:
[email protected]
2 Jefferson Physical Laboratory, Harvard University, Cambridge, MA 02138, USA
Received: 10 November 2008 / Accepted: 20 August 2009 Published online: 1 December 2009 – © Springer-Verlag 2009
Abstract: We demonstrate the agreement between the Higgs branches of two N = 2 theories proposed by Argyres and Seiberg to be S-dual, namely the SU(3) gauge theory with six quarks, and the SU(2) gauge theory with one pair of quarks coupled to the superconformal theory with E 6 flavor symmetry. In mathematical terms, we demonstrate the equivalence between a hyperkähler quotient of a linear space and another hyperkähler quotient involving the minimal nilpotent orbit of E 6 , modulo the identification of the twistor lines. Contents 1. 2. 3. 4.
5.
6. 7.
Introduction . . . . . . . . . . . . . . . 1.1 Argyres-Seiberg duality . . . . . . . 1.2 Higgs branch . . . . . . . . . . . . Rudiments of Hyperkähler Cones . . . . Geometry of the Minimal Nilpotent Orbit SU(3) Side . . . . . . . . . . . . . . . . 4.1 Poisson brackets . . . . . . . . . . . 4.2 Conjugation . . . . . . . . . . . . . 4.3 Constraints . . . . . . . . . . . . . Exceptional Side . . . . . . . . . . . . . 5.1 Poisson brackets . . . . . . . . . . . 5.2 Conjugation . . . . . . . . . . . . . 5.3 Constraints . . . . . . . . . . . . . 5.4 Gauge invariant operators . . . . . . Comparison . . . . . . . . . . . . . . . 6.1 Identification of operators . . . . . . 6.2 Constraints . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
390 390 391 394 396 398 398 399 399 400 400 401 401 402 402 402 403 404
390
D. Gaiotto, A. Neitzke, Y. Tachikawa
A. Conventions . . . . . . . . . . . . . B. Twistor Spaces of Hyperkähler Cones C. Comparison of the Kähler Potential . C.1 Exceptional side . . . . . . . . . C.2 SU(3) side . . . . . . . . . . . . C.3 Comparison . . . . . . . . . . . D. Mathematical Summary . . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
404 405 406 406 407 407 408
1. Introduction 1.1. Argyres-Seiberg duality. In a remarkable paper [1], a new type of strong-weak duality of four-dimensional N = 2 theories was introduced. Consider an N = 2 supersymmetric SU(3) gauge theory with six quarks in the fundamental representation. This theory has vanishing one-loop beta function, and the gauge coupling constant τ=
θ 8πi + 2 π g
(1.1)
is exactly marginal. Argyres and Seiberg carried out a detailed study of the behavior of the Seiberg-Witten curve close to the point τ → 1, where the theory is infinitely strongly-coupled, and were led to conjecture a dual description involving an SU(2) group with gauge coupling τ =
1 . 1−τ
(1.2)
To understand the matter content of the dual theory, one first needs to recall the interacting superconformal field theory (SCFT) with flavor symmetry E 6 first described by [2]. This theory has one-dimensional Coulomb branch parametrized by u whose scaling dimension is 3, and is realized as the low-energy limit of the worldvolume theory on a D3-brane probing the transverse geometry of an F-theory 7-brane with E 6 gauge group. The gauge group on the 7-brane then manifests as a flavor symmetry from the point of view of the D3-brane. We denote this theory by SCFT[E 6 ] following [1]. Now, the theory Argyres and Seiberg proposed as the dual of the SU(3) gauge theory with six quarks consists of the SU(2) gauge bosons, coupled to one hypermultiplet in the doublet representation, and also to a subgroup SU(2) ⊂ E 6 of SCFT[E 6 ]. The SU(2) subgroup is chosen so that the raising operator of SU(2) maps to the raising operator for the highest root of E 6 . In the following, we refer to two sides of the duality as the SU(3) side and the exceptional side, respectively. Argyres and Seiberg provided a few compelling pieces of evidence for this duality. First, the flavor symmetry agrees. On the SU(3) side, there is a U(6) = U(1) × SU(6) symmetry which rotates the six quarks. On the exceptional side, there is an SO(2) symmetry which rotates a pair of quarks in the doublet representation, which can be identified with the U(1) part of U(6). Then, the flavor symmetry of the SCFT with E 6 is broken down to the maximal subgroup commuting with SU(2) ⊂ E 6 , which is SU(6). Second, the scaling dimensions of Coulomb-branch operators agree. Indeed, on the SU(3) side one has tr φ 2 and tr φ 3 , where φ is the adjoint chiral multiplet of SU(3). The dimensions are thus 2 and 3. On the exceptional side, one has tr ϕ 2 (where ϕ is the adjoint chiral multiplet of SU(2)), which has dimension 2, and the Coulomb-branch operator u of SCFT[E 6 ], which has dimension 3.
Argyres-Seiberg Duality and the Higgs Branch
391
Third, Argyres and Seiberg studied in detail the deformation of the Seiberg-Witten curve under the SU(6) mass deformation, and found remarkable agreement. Fourth, they computed the current algebra central charge of the SU(6) flavor symmetry on the SU(3) side, which agreed with the central charge of the E 6 symmetry on the exceptional side, inferred from the fact that the beta function of the SU(2) gauge group coupling is zero. This is as it should be, because SU(6) arises as a subgroup of E 6 on the exceptional side. This provided a prediction of the current central charge of SCFT[E 6 ] for the first time, which was later reproduced holographically by [3]. There are generalizations to similar duality pairs involving SCFTs with flavor symmetries other than E 6 [1,4]. Our aim in this note is to present further convincing evidence for this duality, by showing that the Higgs branches of the two sides of the duality are equivalent as hyperkähler cones. Mathematically speaking, we will show the agreement of their twistor spaces as complex varieties with real structure, but we have not been able to prove that they share the same family of twistor lines. Instead we give numerical evidence that their Kähler potentials agree in Appendix C. 1.2. Higgs branch. On the SU(3) side, let us denote the squark fields by Q ia ,
Q˜ ia ,
(1.3)
where i = 1, . . . , 6 are the flavor indices and a = 1, 2, 3 the color indices. The Higgs branch is the locus where the F-term and the D-term both vanish, divided by the action of the gauge group SU(3). As is well known, this space can also be obtained by setting F = 0 without setting D = 0, and dividing by the complexified gauge group SL(3, C). Thus the Higgs branch is parametrized by gauge invariant composite operators M i j = Q ia Q˜ aj ,
j
B i jk = abc Q ia Q b Q kc ,
B˜ i jk = abc Q˜ ia Q˜ bj Q˜ ck
(1.4)
which satisfy various constraints, e.g. B [i jk M l] m = 0
(1.5)
to which we will come back later. The fields Q ia , Q˜ ia have 36 complex components, while the F-term condition imposes 8 complex constraints. The quotient by SL(3, C) reduces the complex dimension further by 8, so the Higgs branch has complex dimension 2 × 3 × 6 − 8 − 8 = 20.
(1.6)
Our problem is to understand how this structure of the Higgs branch is realized on the exceptional side. Firstly, we have one hypermultiplet in the doublet representation, which we denote as vα , v˜α in N = 1 superfield notation. Here α = 1, 2 is the doublet index. We also have the Higgs branch of SCFT[E 6 ], the structure of which is known through the F-theoretic construction of the SCFT. Recall that this theory is the worldvolume theory on one D3-brane probing a F-theory 7-brane of type E 6 . Say the D3-brane extends along the directions 0123, and the 7-brane along the directions 01234567. The onedimensional Coulomb branch of this theory is identified with the transverse directions 89 to the 7-brane. The theory becomes superconformal when the D3-brane hits the 7-brane, at which point the Higgs branch emanates. This is identified as the process where a D3-brane is absorbed into the worldvolume of the 7-brane as an E 6 instanton
392
D. Gaiotto, A. Neitzke, Y. Tachikawa
along the directions 4567. The real dimension of the N -instanton moduli space of E 6 is 4h E 6 N with the dual Coxeter number h E 6 = 12. The center-of-mass motion of the instanton corresponds to a decoupled free hypermultiplet, and thus the genuine moduli space is the so-called ‘centered’ one-instanton moduli space without the center-of-mass motion, which has complex dimension 22. The SU(2) gauge group couples to the quark fields vα , v˜α , and this instanton moduli space. Imposing the F-term condition and dividing by the complexified gauge group, we find the complexified dimension of the Higgs branch as 2 × 2 + 22 − 3 − 3 = 20,
(1.7)
which correctly reproduces the dimension of the Higgs branch on the SU(3) side. We would like to perform more detailed checks, and for that purpose one needs to have a concrete description of the instanton moduli. It is well known that the ADHM description is available for classical gauge groups, but how can we proceed for exceptional groups? Luckily, there is another description of the 1-instanton moduli spaces, applicable to any group G, which identifies the centered 1-instanton moduli space with the minimal nilpotent orbit of G [5]. Let us now define the minimal nilpotent orbit. G C acts on the complexified Lie algebra gC , which has the Cartan generators H i and the raising/lowering operators E ±ρ for roots ρ. G C also acts on the dual vector space g∗C of gC via the coadjoint action,1 and the minimal nilpotent orbit Omin (G) of G is the orbit of (E θ )∗ , where θ denotes the highest root: Omin (G) = G C · (E θ )∗ .
(1.8)
The minimal nilpotent orbit is known to have polynomial defining equations. Moreover, they can be chosen to be quadratic, transforming covariantly under G C . These relations are known under the name of the Joseph ideal [6]. The simplest example is the case G = SU(2). In this case gC is three-dimensional; denote its three coordinates by a, b and c, which transform as a triplet of SU(2). The minimal nilpotent orbit is then given by a 2 + b2 + c2 = 0
(1.9)
which describes the space C2 /Z2 , and as is well-known, the centered one-instanton moduli space of SU(2) is exactly this orbifold. Let us come back to the case of E 6 . We fix an SU(2) subalgebra generated by E ±θ . The maximal commuting subalgebra is then SU(6). The E 6 algebra can be decomposed under the subgroup SU(2) × SU(6) into Xi j,
Yα[i jk] ,
Z αβ ,
(1.10)
where i, j, k = 1, . . . , 6 are the SU(6) indices, α, β = 1, 2 those for SU(2). Here X ij i jk
and Z αβ are adjoints of SU(6) and SU(2) respectively, and Yα transforms as the threeindex anti-symmetric tensor of SU(6) times the doublet of SU(2). The minimal nilpotent orbit is then given by the simultaneous zero locus of quadratic equations in X , Y and Z which we describe in detail later. 1 One can of course identify g∗ and g using the Killing form, but it is more mathematically natural to C C use the coadjoint representation here.
Argyres-Seiberg Duality and the Higgs Branch
393
For now let us see what are the gauge-invariant coordinates of the Higgs branch of the exceptional side. The SU(2) gauge group is identified to the SU(2) ⊂ E 6 just chosen above, i.e. the SU(2) gauge bosons couple to the current of this SU(2) subgroup of the E 6 symmetry. We also have the quarks vα and v˜α in addition to the fields X , Y and Z , and we need to make SU(2)-invariant combinations of them. Moreover, we need to impose the F-term equation, which is Z αβ + v(α v˜β) = 0
(1.11)
as we argue later. Thus, any appearance of Z inside a composite operator can be eliminated in favor of v and v. ˜ Therefore we have the following natural gauge-invariant composites, from which all gauge-invariant operators can be generated as will be shown in Sec. 5.4: (v v), ˜
Xi j,
(Y i jk v),
(Yi jk v). ˜
(1.12)
Here we defined (uw) ≡ u α wβ αβ
(1.13) i jk
for two doublets u α and wα , and Yi jk,α is defined by lowering the indices of Yα by the epsilon tensor, see Appendix A. This suggests the following identifications between the operators on the two sides of the duality: tr M ↔ (v v), ˜ B
i jk
↔ (Y
i jk
v),
Mˆ i j ↔ X i j , B˜ i jk ↔ (Yi jk v), ˜
(1.14) (1.15)
where Mˆ ij is the traceless part of M ij . The identifications preserve the dimensions of the operators if we assign dimensions 2 to the fields X , Y and Z . The SU(6) transformation nicely agrees. The U(1) part of the flavor symmetry can be matched if one assigns charge ˜ and charge ±3 to v, v. ±1 to Q, Q, ˜ This factor of 3 was predicted in the original paper [1] from a totally different point of view, by demanding that the two-point function of two U(1) currents should agree under the duality. Let us quickly recall how it was derived there. The form of the twopoint function of the U(1) current jµ is strongly constrained by the conservation and the conformal symmetry, and we have jµ (x) jν (0) ∝ k
x 2 gµν − 2xµ xν + ··· . x8
(1.16)
k is called the central charge, and · · · stands for less singular terms. Let us normalize k such that one hypermultiplet of charge q contributes q 2 to k. Assign Q, Q˜ the charge ±1, and let the charge of v, v˜ be ±q. Then k calculated from the SU(3) side is 6 × 3 = 18, while k determined from the exceptional side is 2q 2 . Equating these, Argyres and Seiberg concluded that the charge of v, v˜ should be q = ±3. The agreement is already impressive at this stage, but we would like to see how the constraints are mapped. We would also like to study how the hyperkähler structures agree, because so far we considered the Higgs branch only as a complex manifold. For that purpose we need to recall more about the hyperkähler cone.
394
D. Gaiotto, A. Neitzke, Y. Tachikawa
The structure of the rest of the paper is as follows: We discuss in Sec. 2 what data are mathematically necessary to show the equivalence of the Higgs branches. Section 3 is devoted to the description of the minimal nilpotent orbit, i.e. the 1-instanton moduli space, as a hyperkähler space. Sections 4 and 5 will be spent in calculating the necessary data on the SU(3) side and the exceptional side, respectively. Then they are compared in Sec. 6 which shows remarkable agreement. We conclude in Sec. 7. We have four Appendices: Appendix A collects our conventions, Appendix B gathers the machinery of twistor spaces required to show the equivalence of hyperkähler cones, and Appendix C compares the Kähler potentials of the duality pair. Appendix D is a summary for mathematicians. 2. Rudiments of Hyperkähler Cones Here we collect the basics of the hyperkähler cones in a physics language. Mathematically precise formulation can be found in [7,8]. The Higgs branch M of an N = 2 gauge theory is a hyperkähler manifold, i.e. one has three complex structures J 1,2,3 satisfying J 1 J 2 = J 3 , compatible with the metric g, and the associated two-forms ω1,2,3 are all closed. We choose a particular N = 1 supersymmetry subgroup of the N = 2 supersymmetry group, which distinguishes one of the complex structures, say J ≡ J 3 . M is then thought of as a Kähler manifold with the Kähler form ω = ω3 . = ω1 + iω2 is a closed (2, 0)-form on M which then defines a holomorphic symplectic structure on M. Physically this means that the N = 1 chiral ring, i.e. the ring of holomorphic functions on M, has a natural holomorphic Poisson bracket [ f 1 , f 2 ] = ( −1 )i j ∂i f 1 ∂ j f 2
(2.1)
for two holomorphic functions f 1,2 on M. Second, we are dealing with the Higgs branch of an N = 2 superconformal theory, which has the dilation and the SU(2) R symmetry built in the symmetry algebra. The dilation makes M into a cone with the metric 2 2 dsM = dr 2 + r 2 dsbase ,
(2.2)
and SU(2) R symmetry acts on the base of the cone as an isometry, rotating the three complex structures as a triplet. These two conditions make M into a hyperkähler cone. K = r 2 is a Kähler potential with respect to any of the complex structures J 1,2,3 , and is called the hyperkähler potential in the mathematical literature. The dilatation assigns the scaling dimensions, or equivalently the weights, to the chiral operators on M. Let us consider an element of SU(2) R which acts on the three complex structures as (J 1 , J 2 , J 3 ) → (J 1 , −J 2 , −J 3 ). This element defines an anti-holomorphic involution σ : M → M because it reverses J ≡ J 3 . This induces an operation σ ∗ on holomorphic functions on M via (σ ∗ ( f ))(x) ≡ f (σ (x)). σ ∗ maps holomorphic functions to anti-holomorphic functions, but is a linear operation, not a conjugate-linear operation. We call this operation the conjugation. As will be detailed in Appendix B, the space M as a complex manifold, with the Poisson brackets, the scaling weights and the conjugation, almost suffices to reconstruct the hyperkähler metric on M. Therefore, our main task in checking the agreement of the Higgs branches of the duality pair is to identify them as complex manifolds, and to show that the extra data defined on them also coincide. In order to complete the proof we need to show that the families of the twistor lines coincide, which we have not been
Argyres-Seiberg Duality and the Higgs Branch
395
able to do. Instead we will give numerical support by calculating the Kähler potential directly on both sides in Appendix C. The Higgs branches M that we treat here are gauge theory moduli spaces. They can be described by the hyperkähler quotient construction [9], which we now review. Let us start with an N = 2 gauge theory with the gauge group G, whose hypermultiplets take value in the hyperkähler manifold X . The action of G on X preserves three Kähler structures, and thus there are three moment maps µas (s = 1, 2, 3; a = 1, . . . , dim G) which satisfy dµas = ιξ a ωs ,
(2.3)
where ξ a is the Killing vector associated to the a th generator of G. The Higgs branch of the gauge theory, in the absence of any non-zero Fayet-Iliopoulos parameter, is then given by M ≡ X ///G ≡ {x ∈ X µas (x) = 0}/G. (2.4) With one complex structure J = J 3 chosen, it is convenient to call D a = µa3 ,
F a = µa1 + iµa2 .
(2.5)
M = {x ∈ X F a = 0}/G C .
(2.6)
Then, as a complex manifold,
It is instructive to note that F a is exactly the Hamiltonian which generates the G action on the chiral ring of X , under the Poisson bracket associated to = ω1 + iω2 . The conjugation σ ∗ and the Poisson bracket [·, ·] on the quotient M are given by the restriction of the corresponding operations on X . It is instructive to see why the Poisson bracket of the quotient is well-defined: two G-invariant holomorphic functions f 1,2 on X lead to the same function on M if and only if f 1 = f 2 + u a F a with holomorphic functions u a . Then we have, for a G-invariant holomorphic function h, [ f 1 , h] − [ f 2 , h] = [u a F a , h] = [u a , h]F a + u a [F a , h]
(2.7)
on X . The first term in the right hand side is zero on M because we set F a = 0, while the second term is zero because h is G-invariant. Therefore [ f 1 , h] and [ f 2 , h] determine the same holomorphic function on M. The Kähler potential of M is similarly the restriction of that of X to the zero locus of the moment maps in our situation, as discussed in Sec. 2B of [9]. To illustrate the procedure, let us consider an N = 1 supersymmetric U(1) gauge theory coupled to chiral fields i of charge qi whose Lagrangian is 4 ∗ 2qi V L= d θ i e i + ξ V , (2.8) i
where ξ is the Fayet-Iliopoulos parameter. The moduli space can be determined by taking the gauge coupling to be formally infinite, i.e. treating the linear superfield V as an auxiliary field. Then V is determined via its equation of motion qi i∗ e2qi V i + ξ = 0, (2.9) i
396
D. Gaiotto, A. Neitzke, Y. Tachikawa
i.e. i = eqi V i solve the usual D-term equation. The Kähler potential of the moduli space is then given by plugging the solution to (2.9) into (2.8). It can be generalized to any gauge group, and the result agrees with the mathematical formula given in Sec. 3.1 of [10]. This shows that the Kähler potential is given just by the restriction of the original one if ξ = 0. This analysis does not incorporate quantum corrections, but it is wellknown that for N = 2 theories the quantum effect does not modify the hyperkähler structure, see Sec. 3 of [11]. 3. Geometry of the Minimal Nilpotent Orbit Here we gather the relevant information on the hyperkähler geometry of the minimal nilpotent orbit of any simple group G, which coincides with the centered moduli space of single instantons with gauge group G [5,8]. We hope this section might be useful for anyone who wants to deal with the one-instanton moduli space. In the following G stands for a compact simple Lie group, gR its Lie algebra. We let G C and gC be complexifications of G and gR respectively. The existence of a uniform description of the one-instanton moduli space applicable to any G might be understood as follows: we can construct a one-instanton configuration easily by taking a BPST instanton of SU(2) and regard it as an instanton of G via a group embedding SU(2) ⊂ G. It is known that any one-instanton of G arises in this manner [12]. The one-instanton moduli space is then parameterized by the position, the size, and the gauge orientation of the BPST instanton inside G. This description realizes the one-instanton moduli space as a cone over a homogeneous manifold G/H , where H is the maximal subgroup of G which commutes with the SU(2) used in the embedding. It is however not directly suitable for the analysis of its complex structure. For that purpose we use another realization of the one-instanton moduli space as the minimal nilpotent orbit Omin of G [5]. Let us define Omin . First we decompose gC into the Cartan generators H i and the raising/lowering operators E ±ρ for roots ρ. The minimal nilpotent orbit Omin (G) is then the orbit of (E θ )∗ in g∗C , where θ denotes the highest root: Omin (G) = G C · (E θ )∗ ⊂ g∗C .
(3.1)
We will write Omin without explicitly writing G for the sake of simplicity when there is no confusion. We think of elements of gC as holomorphic functions on Omin , i.e. we have holomorphic functions2 Xa (a = 1, . . . , dim G ) on Omin . The defining equations of Omin are a set of quadratic equations which we call the Joseph relations [6].3 These relations can be studied using a theorem of Kostant [13]: Let V (α) denote the representation space of a semisimple group G with the highest weight α, and let v ∈ V (α)∗ be a vector in the highest weight space. The orbit G C · v is then an affine algebraic variety whose defining ideal I is generated by its degree-two part I2 . Furthermore, I2 is given by the relation Sym2 V (α) = V (2α) ⊕ I2 ,
(3.2)
∗ 2 More mathematically, one has a natural holomorphic g∗ -valued function X : O min → gC given by C the embedding. Then every element t ∈ gC gives a holomorphic function (X, t) on Omin via x ∈ Omin → (X(x), t). Our Xa is (X, T a ) for a generator T a of gC . We take a real basis of gC , so in fact T a ∈ gR ⊂ gC . 3 Strictly speaking, the Joseph ideal is a two-sided ideal in the universal enveloping algebra of g , and C what we use below is its associated ideal in the polynomial algebra.
Argyres-Seiberg Duality and the Higgs Branch
397
where we identify Sym2 V (α) as the space of degree-two polynomials on V (α)∗ . The minimal nilpotent orbit is exactly of this form where V (α) is the adjoint representation, i.e. Omin = {X ∈ g∗C (X ⊗ X)|I2 = 0}. (3.3) For practice, let us apply this to the case G = SU(2). There, V (α) is the triplet representation, so by (3.2) I2 is the singlet representation. Therefore, if we parameterize su(2) by (a, b, c), the minimal nilpotent orbit is given by the equation a 2 + b2 + c2 = 0,
(3.4)
which is C2 /Z2 as it should be. Now that we have given Omin as a complex manifold, let us describe its hyperkähler structure. The main fact we use is that G acts isometrically on Omin , preserving the hyperkähler structure. There is a triplet of moment maps µas for this action where a = 1, . . . , dim G and s = 1, 2, 3. The functions Xa are the holomorphic moment maps of the G action, i.e. Xa = µa1 + iµa2 . It follows that their Poisson bracket is [Xa , Xb ] = f ab c Xc ,
(3.5)
where f ab c are the structure constants of G. Phrased differently, the holomorphic symplectic structure underlying the hyperkähler structure of the nilpotent orbit is the standard Kirilov–Kostant–Souriau symplectic form on the coadjoint orbit [5,8]. The conjugation is given by the SU(2) R action, which sends (µ1 , µ2 , µ3 ) to (µ1 , −µ2 , −µ3 ). Therefore σ ∗ (Xa ) = (Xa )∗ .
(3.6)
The scaling dimension of X is fixed to be two, as it should be for the F-term in an N = 2 supersymmetric theory. Let us next describe a Kähler potential for Omin , which was determined in [14]. The derivation boils down to the following: G acts on Omin with cohomogeneity one, and by averaging over this action we can consider K to be G-invariant; so K is a √ function of tr XX∗ . K should be of scaling dimension two, so that K is proportional to tr XX∗ up to a constant. The constant factor can be fixed by considering a particular element on Omin . For this purpose we again turn to the minimal nilpotent orbit of SU(2), which is C2 /Z2 . The normalization of the Kähler potential of the minimal nilpotent orbit of a general group can then be determined because it contains the minimal nilpotent orbit of SU(2) as a subspace. We parameterize C2 by (u, u) ˜ and divide by the multiplication by −1. We define our conventions for the holomorphic Poisson bracket and the Kähler potential of a flat H as follows: K = |u|2 + |u| ˜ 2,
[u, u] ˜ = 1.
(3.7)
Now, C2 /Z2 is parametrized by Z 11 = u 2 /2,
Z 12 = Z 21 = u u/2, ˜
Z 22 = u˜ 2 /2
(3.8)
which satisfy Z 11 Z 22 = Z 12 2 .
(3.9)
398
D. Gaiotto, A. Neitzke, Y. Tachikawa
The Kähler potential is now K = 2 |Z 11 |2 + |Z 22 |2 + 2|Z 12 |2 = 2 Z αβ Z¯ αβ .
(3.10)
Then, the moment map associated to the generator J3 of non-R SU(2) acting on C2 /Z2 can be explicitly calculated, with the result F = Z 12 ,
D=
2 (|Z 11 |2 − |Z 22 |2 ). K
(3.11)
Now that the preparation is done, we move on to the calculation of the Higgs branch on both sides of the duality. 4. SU(3) Side The theory has six quarks in the fundamental representation, Q ia ,
Q˜ ia ,
(4.1)
where a = 1, . . . , 6 and i = 1, 2, 3. As is well known, any SU(3)-invariant polynomial constructed out of these fields is a polynomial in the operators [11]: M i j = Q ia Q˜ aj ,
j
B i jk = abc Q ia Q b Q kc ,
B˜ i jk = abc Q˜ ia Q˜ bj Q˜ ck .
(4.2)
In the following we study the Poisson brackets, the action of the conjugation, and the constraints in turn. 4.1. Poisson brackets. The Poisson bracket of the basic fields is given by [Q ia , Q˜ bj ] = δ i j δ b a .
(4.3)
[M ij , Q ak ] = −δ kj Q ia ,
(4.4)
Then we have, for example,
i.e. M ij is the generator of U(6). We define tr M to be the trace of M i j , and 1 Mˆ ij = M ij − δ ij tr M 6
(4.5)
is its traceless part. Mˆ i j is the SU(6) generator and tr M the U(1) generator. We define the U(1) charge q of an operator O to be given by [tr M, O] = −qO.
(4.6)
The most complicated bracket is [B i jk , B˜ lmn ] = 18M [i [l M j m δ k] n] 1 = 18 Mˆ [i [l Mˆ j m δ k] n] + 6(tr M) Mˆ [i [l δ j m δ k] n] + (tr M)2 δ [i [l δ j m δ k] n] . 2 (4.7)
Argyres-Seiberg Duality and the Higgs Branch
399
4.2. Conjugation. We choose the involution on the elementary fields to be σ ∗ (Q ia ) = ( Q˜ ia )∗ ,
σ ∗ ( Q˜ ia ) = −(Q ia )∗ .
(4.8)
Then the transformation of the composites are σ ∗ (M ij ) = −(M j i )∗ ,
σ ∗ (tr M) = −(tr M)∗ ,
(4.9)
σ ∗ (B i jk ) = ( B˜ i jk )∗ ,
σ ∗ ( B˜ i jk ) = −(B i jk )∗ .
(4.10)
4.3. Constraints. The constraints were studied in [11]. Those which come before imposing the F-term constraint are B i jk B˜ lmn = 6M [i l M j m M k] n , B i j[k B lmn] = 0, B˜ i j[k B˜ lmn] = 0, M [i j B klm] = 0, M i [ j B˜ klm] = 0.
(4.11) (4.12) (4.13)
The F-term constraint 1 ˜ =0 Q ia Q˜ ib − δab (Q Q) 3
(4.14)
Mˆ ij B jkl = 16 (tr M)B ikl ,
(4.15)
Mˆ ij B˜ ikl = 16 (tr M) B˜ jkl ,
(4.16)
Mˆ ij M j k
(4.17)
further imposes
=
1 6 (tr
M)Mki .
We will find it convenient later to have constraints in terms of irreducible representations (irreps) of SU(6). We use the Dynkin labels to distinguish the irreps in the following. The M B = 0 relations (4.13), (4.15), (4.16) give Mˆ {i l B [ jk]}l = 0, Mˆ l {i B[ jk]}l = 0,
Mˆ {i l B˜ [ jk]}l = 0, Mˆ l {i B˜ [ jk]}l = 0.
(4.18) (4.19)
Here we defined the projector from a tensor with the structure Ai[ jk] to the irrep (1, 1, 0, 0, 0) by A{i[ jk]} ≡ Ai[ jk] − A[i[ jk]] . We also have 1 Mˆ [i l B jk]l = (tr M)B i jk , 6
1 Mˆ l [i B˜ jk]l = (tr M) B˜ i jk . 6
(4.20)
The M M = 0 relation (4.17) gives 1 j Mˆ ij Mˆ k = δki Mˆ nm Mˆ mn , 6 1 j i Mˆ j Mˆ i = (tr M)2 . 6
(4.21) (4.22)
The B B = 0 relation (4.12) gives B ikl B jkl = 0,
B˜ ikl B˜ jkl = 0.
(4.23)
400
D. Gaiotto, A. Neitzke, Y. Tachikawa
Finally, the decomposition of the B B˜ = M M M relation gives, using (4.21) and (4.22) repeatedly, 2 (tr M)3 , 9 2 = (tr M)2 Mˆ ij , 9 2 = (tr M) Mˆ [i [k Mˆ j] l] 0,1,0,1,0 , 3 j = 6 Mˆ [i l Mˆ m Mˆ k] n 0,0,2,0,0 .
B i jk B˜ i jk = B ikl B˜ jkl adj B i jm B˜ klm 0,1,0,1,0 B i jk B˜ lmn
0,0,2,0,0
(4.24) (4.25) (4.26) (4.27)
5. Exceptional Side 5.1. Poisson brackets. We have chiral fields Xa which transform in the adjoint of E 6 , and satisfy the quadratic Joseph identities. We decompose Xa under the subgroup SU(2) × SU(6) ⊂ E 6 . It gives X ij , Yα[i jk] ,
Z αβ ,
(5.1) i jk
where X ij and Z αβ are the adjoints of SU(6) and SU(2) respectively, and Yα is in the doublet of SU(2) and in the representation (0, 0, 1, 0, 0), i.e. the three-index antisymmetric tensor, of SU(6). The Poisson brackets of the fields X , Y and Z are exactly the Lie brackets as explained above, which we take to be
[Z αβ , Z γ δ ] =
[X ij , X lk ] = δli X kj − δ kj X li ,
(5.2)
1 (αγ Z βδ + βγ Z αδ + αδ Z βγ + βδ Z αγ ) 2
(5.3)
and 1 [X ij , Yαklm ] = −3δ [k j Yαlm]i + δ ij Yαklm , 2 i jk [Z αβ , Yγi jk ] = Y(α β)γ ,
(5.4) (5.5)
and finally [Yαi jk , Yβlmn ] = i jklmn Z αβ −
3 αβ (X [i p jk]lmnp + X [l p mn]i jkp ). 2
(5.6)
The final commutation relation can also be written as [Yαi jk , Ylmn β ] = −18X [i [l δ j m δ k] n] − 6Z αβ δ [i [l δ j m δ k] n] .
(5.7)
As we explained above, X , Y and Z are the holomorphic moment maps of the E 6 action. Therefore the contribution from Omin to the F-term constraint for the SU(2) gauge group is given just by Z αβ . We take the bracket of v and v˜ to be [vα , v˜β ] = αβ .
(5.8)
Argyres-Seiberg Duality and the Higgs Branch
401
Then we have [v(α v˜β) , vγ ] = v(α β)γ ,
(5.9)
and [(v v), ˜ vα ] = vα ,
[(v v), ˜ v˜α ] = −v˜α .
(5.10)
Recall that we define (uw) ≡ u α wβ αβ for two doublets u α and wα . It is straightforward to check that v(α v˜β) is the moment map of the SU(2) action on v and v. ˜ Thus the F-term condition is v(α v˜β) + Z αβ = 0.
(5.11)
5.2. Conjugation. We take the conjugation on the variables v, v˜ to be σ ∗ (vα ) = (v˜β )∗ αβ ,
σ ∗ (v˜α ) = (vβ )∗ αβ .
(5.12)
i jk
In terms of our variables (X ij , Yα , Z αβ ), the conjugation acts as follows: σ ∗ (X ij ) = −(X j i )∗ , σ ∗ (Yαi jk ) = (Yi jk β )∗ αβ , ∗
(5.13)
σ ∗ (Yi jk α ) = −(Yβ )∗ αβ , i jk
∗
σ (Z αβ ) = (Z γ δ ) αγ βδ .
(5.14) (5.15)
5.3. Constraints. As explained in Sec. 3, the Joseph relations are given by (X ⊗ X)|I2 = 0,
(5.16)
Sym2 V (adj) = V (2adj) ⊕ I2 .
(5.17)
where I2 is given by the relation
Here, V (adj) is the adjoint representation of E 6 whose Dynkin label is adj = 0 0 01 0 0 . We then have (5.18) I2 = V 1 0 00 0 1 ⊕ V 0 0 00 0 0 . The representations which appear in I2 , decomposed under SU(2) × SU(6), are summarized in Table 1. The table reads as follows: e.g. for relation 4, the fourth column tells us there is one Joseph identity transforming as a doublet in SU(2) and as (0, 0, 1, 0, 0) under SU(6), but the fifth column says one can construct two objects in i jk this representation from bilinears in X ij , Yα and Z αβ . This means the identity has the form 0 = Yαi jk Z βγ αβ + c4 X [i l Yγjk]l ,
(5.19)
where c4 needs to be fixed, which can be done e.g. by explicitly evaluating the right hand side on a few elements on the nilpotent orbit. Elements on the nilpotent orbit can be readily generated, because one knows that the point X ij = 0,
Yαi jk = 0,
Z 11 = 1,
Z 12 = Z 22 = 0
(5.20)
402
D. Gaiotto, A. Neitzke, Y. Tachikawa Table 1. Decomposition of I2 in terms of SU(2) × SU(6) ⊂ E 6
1. 2. 3. 4. 5. 6. 7.
SU(2)
SU(6)
in I2
in Sym2 V (adj)
3 2 2 2 1 1 1
(1,0,0,0,1) (1,1,0,0,0) (0,0,0,1,1) (0,0,1,0,0) (0,1,0,1,0) (1,0,0,0,1) (0,0,0,0,0)
1 1 1 1 1 1 2
2 1 1 2 2 1 3
is on the nilpotent orbit by definition. Then the rest of the points can be generated by the coadjoint action of E 6 , which can be obtained by exponentiating the structure constants. Carrying out this program, we obtain the following full set of Joseph identities: 1. 2. 3.
1 ikl 0 = X ij Z αβ + Y(α Y jklβ) , 4 0 = X l {i Y[ jk]}lα , 0=
4.
0=
5.
0=
6. 7. 7’.
X l Yα[ jk]}l , Yαi jk Z βγ αβ
(5.21) (5.22)
{i
(5.23) [i
+ X l Yγjk]l , (Yαi jm Yklmβ αβ − 4X [i [k X j] l] ) 0,1,0,1,0 ,
1 0 = X ki X k j − δ ij X k l X l k , 6 0 = Yαi jk Yi jkβ αβ + 24Z αβ Z γ δ αγ βδ , 0=
X ij X j i
+ 3Z αβ Z γ δ
αγ βδ
.
(5.24) (5.25) (5.26) (5.27) (5.28)
5.4. Gauge invariant operators. Let us enumerate the generators of the SU(2)-invariant i jk operators constructed out of vα , v˜α , and X ij , Yα , Z αβ , using the F-term equation (5.11) and the Joseph identities (5.21) ∼ (5.28). Suppose we have a monomial constructed from those fields. We first replace every appearance of Z αβ by −v(α v˜β) . All the SU(2) indices are contracted by epsilon tensors of SU(2). Therefore the monomial is a product of X ij , (v v), ˜ (Y i jk v), (Y i jk v) ˜ and (Y i jk Y lmn ). The last of these can be eliminated using the Joseph identities. Indeed, the combination of the relations (5.25), (5.27) and (5.28) gives a Joseph identity of the form Yαi jk Ylmn β αβ = 18X [i [l X j m δ k] n] − 3Z αβ Z γ δ αγ βδ δ [i [l δ j m δ k] n] .
(5.29)
We conclude that any SU(2)-invariant polynomial is a polynomial in ˜ (Y i jk v), and (Y i jk v). ˜ X ij , (v v),
(5.30)
6. Comparison 6.1. Identification of operators. Let us now proceed to the comparison of the structures we studied in Sec. 4 and in Sec. 5. We first make the following identification: Mˆ ij = X ij ,
tr M = −3(v v). ˜
(6.1)
Argyres-Seiberg Duality and the Higgs Branch
403
These are the moment maps of the flavor symmetries SU(6) and U(1), so the identification is fixed including the coefficients, and then the Poisson brackets involving either Mˆ ˜ also agrees with that or tr M automatically agree. The conjugation acting on X ij , (v v) i ˆ on M j and tr M. We then set B i jk = c(Y i jk v),
B˜ i jk = c(Y ˜ i jk v). ˜
(6.2)
One has σ ((Y i jk v)) = (Yi jk v) ˜ ∗.
(6.3)
To be consistent with (4.10), we need to have c˜ = c∗ .
(6.4)
Let us then calculate the Poisson bracket of (Y i jk v) and (Ylmn v) ˜ using (5.29). We have 9 ˜ 2 . (6.5) ˜ = −18X [i [l X j m δ k] n] + 18(v v)X ˜ [i [l X j m δ k] n] − (v v) [(Y i jk v), (Ylmn v)] 2 Comparing with the bracket [B i jk , B˜ lmn ] calculated in (4.7), we find they indeed agree if cc˜ = −1. Thus we conclude c = c˜ = i, i.e. B i jk = i(Y i jk v),
B˜ i jk = i(Yi jk v). ˜
(6.6)
6.2. Constraints. Now, let us check using the Joseph relations that the constraints on the SU(3) side, listed in Eqs. (4.18) ∼ (4.27), can be correctly reproduced on the exceptional side. • • • • • •
(4.18): Contract v or v˜ to the relation 2, (5.22). (4.19): Contract v or v˜ to the relation 3, (5.23). (4.20): Contract v or v˜ to the relation 4, (5.24). (4.21): This is exactly the relation 6, (5.26). (4.22): This is exactly the relation 7’, (5.28). (4.23): Contract vα vβ or v˜α v˜β to the relation 1 (5.21).
As for the relation of the type B B˜ = M M M, • • • •
(4.24): The singlet part. Contract vα v˜β to the relation 7 (5.27). (4.25): The adjoint part. Contract vα v˜β to the relation 1 (5.21). (4.26): The (0, 1, 0, 1, 0) part. Contract vα v˜β to the relation 5 (5.25). (4.27):This is the (0, 0, 2, 0, 0) part and is slightly trickier, but it follows from a cubic Joseph identity 0 = αγ βδ Z αβ Yγi jk Ylmn,δ 0,0,2,0,0 −6X [i l X j m X k] n 0,0,2,0,0 (6.7) upon replacing Z αβ with v(α v˜β) . This cubic Joseph identity itself can be derived from the quadratic Joseph identities, as it should be. First, we use the relation 4 (5.24) to show αγ βδ Z αβ Yγi jk Ylmn δ 0,0,2,0,0 ∝ X [i p Yαjk] p Ylmn β αβ . (6.8)
404
D. Gaiotto, A. Neitzke, Y. Tachikawa
Now, the antisymmetric product of two Y ’s contain both the singlet and the (0, 1, 0, 1, 0) part. One sees the singlet drops out inside the projector to the (0, 0, 2, 0, 0) part, so we have ∝ X [i p (Y jk] p Ylmn ) 0,1,0,1,0 0,0,2,0,0 . (6.9) Then we use the relation 5 (5.25) to transform this to ∝ X li X j m X k n 0,0,2,0,0 .
(6.10)
The proportionality constant can be fixed, e.g. by evaluating on a few points on the orbit. This concludes the comparison of the constraints. 7. Conclusions In the previous three sections, we determined the Higgs branches both on the SU(3) side and on the exceptional side. We demonstrated that their defining equations agree, and furthermore exhibited that the Poisson bracket and the conjugation are the same on both sides. As was stated in Sec. 2 and will be detailed in Appendix B, these are (almost) sufficient to conclude that they are the same as hyperkähler manifolds. To remove any remaining doubts, we compare the Kähler potentials of the two sides in Appendix C. Again, they show remarkable agreement with one another. Thus we definitely showed the agreement of the Higgs branches of the new S-duality pair proposed by Argyres and Seiberg in [1], which provides a convincing check of their conjecture. In this paper we only dealt with the example involving E 6 , but there are more examples of similar dualities in [1 and 4]. It would be interesting to carry out the same analysis of the Higgs branches to those examples. A pressing issue is to understand the Argyres-Seiberg duality more fully. For example, it would be nicer to have an embedding of this duality in string/M-theory. We hope to revisit these problems in the future. Acknowledgements. The authors thank Alfred D. Shapere for collaboration at an early stage of the project. They would also like to thank S. Cherkis, C. R. LeBrun, H. Nakajima for helpful discussions. They also relied heavily on the softwares LiE4 and Mathematica. DG is supported in part by the DOE grant DE-FG0290ER40542 and in part by the Roger Dashen membership in the Institute for Advanced Study. AN was supported in part by the Martin A. and Helen Chooljian Membership at the Institute for Advanced Study, and in part by the NSF under grant numbers PHY-0503584 and PHY-0804450. YT is supported in part by the NSF grant PHY-0503584, and in part by the Marvin L. Goldberger membership in the Institute for Advanced Study.
A. Conventions Greek indices α, β are for the doublets of SU(2), a, b, c, . . . for the triplets of SU(3) and i, j, k, . . . for the sextets of SU(6). We define (uw) ≡ u α wβ αβ for two doublets u α and wα , 4 It can be downloaded from http://www-math.univ-poitiers.fr/~maavl/LiE/.
(A.1)
Argyres-Seiberg Duality and the Higgs Branch
405
We use the following sign conventions for the epsilon tensors of SU(2), SU(3) and SU(6): αβ = −αβ , abc = abc , i jklmn = i jklmn .
(A.2)
We normalize the antisymmetrizer [abc...] and the symmetrizer (abc...) so that they are projectors, i.e. T i jk = T [i jk]
(A.3)
for the antisymmetric tensors T i jk , etc. We raise and lower three antisymmetrized indices of SU(6) via the following rule: T i jk =
1 i jklmn Tlmn , 6
Tlmn =
1 i jk T i jklmn . 6
(A.4)
Our convention for the placement of the indices of the complex conjugate is e.g. Z¯ αβ ≡ (Z αβ )∗ ,
(A.5)
i.e. the complex conjugation is always accompanied by the exchange of subscripts and superscripts, as is suitable for the action of SU groups. We take the Kähler potential of a flat C parameterized by z with the standard metric to be K = |z|2 .
(A.6)
B. Twistor Spaces of Hyperkähler Cones Recall that a hyperkähler manifold M admits a continuous family of complex structures Jζ , parameterized by ζ ∈ CP1 . The full information in the hyperkähler metric is captured by this family of complex structures and their Poisson brackets. It can be encoded into purely holomorphic data on a complex manifold Z, the twistor space of M, as we now review. Topologically Z = M × CP1 . Its complex structure can be specified by specifying which functions on Z are holomorphic: they are f (x, ζ ) which are holomorphic in ζ for fixed x ∈ M, and also holomorphic in x with respect to complex structure Jζ for fixed ζ . Hence we may view Z as a holomorphic fiber bundle over CP1 , where the fiber over ζ is just a copy of M, equipped with complex structure Jζ . The Poisson brackets on the holomorphic functions in each fiber glue together globally to give a bracket operation on Z. This bracket operation is globally twisted by the line bundle O(−2): i.e. given local holomorphic functions f 1 , f 2 we get a local section { f 1 , f 2 } of O(−2), and more generally if f 1 , f 2 are sections of O(d1 ), O(d2 ) then { f 1 , f 2 } is a section of O(d1 + d2 − 2). Finally there is an involution σ on Z, simply defined by (x, ζ ) → (x, −1/ζ¯ ). This is an antiholomorphic involution, since the complex structure Jζ is opposite to J−1/ζ¯ . As a complex manifold Z is a fibration over CP1 , and (x, ζ ) with x fixed gives a holomorphic section of this fibration, which is invariant under σ . The normal bundle to this section is isomorphic to the line bundle O(1)⊕n , where n is the complex dimension of M. Conversely, a holomorphic section of Z which is invariant under σ and whose normal bundle is isomorphic to O(1)⊕n is called a twistor line. Therefore, the points on M give rise to a n-dimensional family of twistor lines on Z.
406
D. Gaiotto, A. Neitzke, Y. Tachikawa
It was shown in [9] that given Z, together with its Poisson brackets and antiholomorphic involution, one can canonically reconstruct a hyperkähler metric on the space of twistor lines. Therefore, to check that our two hyperkähler cones are the same is essentially to check that their twistor spaces Z are the same. Now, the twistor space of a hyperkähler cone can be constructed from the data we described in Sec. 2, i.e. the Poisson bracket, the dilatation and the conjugation on M. We pick one complex structure induced from the hyperkähler structure, and regard M as a complex manifold. We then form Z as a complex manifold as Z = ((C2 \ (0, 0)) × M)/C× ,
(B.1)
where C× acts on the first factor by multiplication, and on the second factor as the natural
complexification of the action of the dilatation. Then the Poisson bracket on M naturally induces one on Z. We define σ on Z to send (z, w, x) ∈ C2 ×M to (−w, ¯ z¯ , σ (x)). Then it is straightforward to check that this Z is the twistor space of M, using the SU(2) R action on M rotating three complex structures. There is a subtle problem remaining, however. Namely, the theorem in [9] asserts that there is a component of the space of the twistor lines of Z which agrees metrically with the original hyperkähler manifold M, but does not exclude the possibility that the space of twistor lines has many components, each of which is a hyperkähler manifold with the same complex structure but with a different metric. Mathematicians the authors consulted know no concrete example where this latter possibility is realized, so the authors think it quite unlikely that our two hyperkähler manifolds are the same as holomorphic symplectic manifolds but not as hyperkähler manifolds. To dispel this last possibility, in the next appendix we directly compare the Kähler potential of our two hyperkähler manifolds. C. Comparison of the Kähler Potential In this Appendix, we describe the method to calculate and compare the Kähler potential of the Higgs branches on the two sides of the duality. C.1. Exceptional side. The invariant norm of E 6 in our notation is 1 Z αβ Z¯ αβ + Yαi jk Y¯iαjk + X ij X¯ j i . 6 Therefore the correctly normalized Kähler potential is
1 i jk K E 6 = 2 Z αβ Z¯ αβ + Yα Y¯iαjk + X ij X¯ j i , 6
(C.1)
(C.2)
and the D-term for the SU(2) ⊂ E 6 is 2 1 i jk ¯ γ i jk ¯ γ (E 6 ) γδ γδ ¯ ¯ Z αγ Z δβ + Z βγ Z δα + (Yα Yi jk γβ + Yβ Yi jk γ α ) . Dαβ = K E6 12 (C.3) We also have quarks vα , v˜α which have (|vα |2 + |v˜α |2 ) K v,v˜ = α
(C.4)
Argyres-Seiberg Duality and the Higgs Branch
407
and (v,v) ˜
Dαβ
=
1 (vα v¯ γ γβ + vβ v¯ γ γ α + v˜α v¯˜ γ γβ + v˜β v¯˜ γ γ α ). 2
(C.5)
The Kähler potential of the exceptional side is thus given by K v,v˜ + K E 6
(C.6)
restricted to the locus v(α v˜β) + Z αβ = 0,
(v,v) ˜
Dαβ
(E )
+ Dαβ 6 = 0
(C.7)
expressed as a function of M ij , B i jk , B˜ i jk and their complex conjugates. C.2. SU(3) side. We start from the Kähler potential |Q ia |2 + | Q˜ a |2 . K =
(C.8)
i
i,a
Using the analysis in [11], the Kähler potential on the quotient was determined in [15] as
ν2 (C.9) K =2 m i2 + , 4 i=1,2,3
where (m 21 , m 22 , m 23 , 0, 0, 0) are the eigenvalues of M ij M¯ j k , and ν is defined by 3ν =
|Q ia |2 − | Q˜ ia |2 ,
(C.10)
i,a
i.e. 1/3 of the U(1) D-term. In terms of gauge invariants we have ⎛ ⎞ 2 ν ν ⎝ m2 + + ⎠ = 16 B i jk B¯ i jk , i 4 2 i=1,2,3 ⎛ ⎞ 2 ν ν ⎝ m2 + − ⎠ = 16 B˜ i jk B¯˜ i jk . i 4 2
(C.11)
(C.12)
i=1,2,3
C.3. Comparison. Now, the Kähler potentials of the two sides, (C.6) and (C.9) should ˜ but we have not been able to check that analytically. agree as functions of M, B and B, Instead, one can check it numerically on as many points on the quotient as computer time allows. The algorithm is as follows: i jk
1. Generate a point X = (X ij , Yα , Z αβ ) on the nilpotent orbit of E 6 , by applying an i jk
element of the group E 6 to the point (Z 11 , Z 12 , Z 22 ) = (1, 0, 0), X ij = Yα
= 0.
408
D. Gaiotto, A. Neitzke, Y. Tachikawa
2. Find vα , v˜α which satisfy v(α v˜β) + Z αβ = 0.
(C.13)
˜ This is more or less unique up to C× action on v, v. 3. Apply SL(2, C) action to (v, v, ˜ X) to find the solution of the D-term equation, (v,v) ˜
Dαβ
(E )
+ Dαβ 6 = 0.
(C.14)
This is equivalent to the minimization of ˜ + K E6 (g(X)), K v,v˜ (g(v), g(v))
(C.15)
where g is an SL(2, C) action. 4. Form M, B, B˜ from v, v˜ and X thus obtained, and calculate ν and m i . At this point, two checks of the sanity of the calculation are possible. One is to see that three eigenvalues of M M¯ are zero. Another is to see that ν determined from (C.11), (C.12) is equal to ν= |vα |2 − |v˜α |2 . (C.16) α
The latter fact follows from the identification of ν as 1/3 of the U(1) moment map on the quotient. 5. Evaluate the Kähler potential of the SU(3) side using (C.9) and compare it to that of the exceptional side (C.6). We implemented the algorithm above in Mathematica, and found that the value of the Kähler potential at any points agrees on both sides of the duality to arbitrary accuracy.5 An analytic proof of the agreement of the Kähler potential will be welcomed. D. Mathematical Summary Let us summarize briefly in the language of mathematics what was done in this paper. Let M(m, n) be M(m, n) = Hom(V, W ) ⊕ Hom(W, V )
where V = Cm , W = Cn
(D.1)
which is a flat hyperkähler space of quaternionic dimension mn. It has a natural triholomorphic action of U(m) × U(n) induced from its action on V and W . Let N (m, n) be the flat hyperkähler space N (m, n) = Rm ⊗R Hn
(D.2)
of quaternionic dimension mn, which has a natural triholomorphic action of SO(m) × Sp(n). One then defines a hyperkähler quotient A by A1 = M(6, 3)///SU(3). 5 We thank H. Elvang for improvement of the accuracy in the calculation.
(D.3)
Argyres-Seiberg Duality and the Higgs Branch
409
We consider another hyperkähler quotient A2 = (N (2, 1) × Omin (E 6 ))///Sp(1)
(D.4)
where Omin (G) is the minimal nilpotent orbit of the group G, and the Sp(1) action on Omin (E 6 ) is given by considering the maximally compact subgroup Sp(1)×SU(6) ⊂ E 6 . One sees easily that A1,2 are both of quaternionic dimension 10, both carry a natural triholomorphic action of SU(6) × U(1). Our claim is that A1 = A2 as hyperkähler cones. We demonstrated that A1 and A2 match as holomorphic symplectic varieties by explicitly showing that their defining equations and the holomorphic symplectic forms are the same. We also found that the twistor spaces of A1 and A2 are the same as complex manifolds with antiholomorphic involution, but could not show that A1 and A2 correspond to the same family of twistor lines. Instead we directly compared the Kähler potentials of A1 and A2 . Again we could not rigorously prove the equivalence, but we performed numerical calculations of the Kähler potential which convinced us that they agree. The equivalence of A1,2 was suggested by the analysis of a new type of S-duality in four-dimensional N = 2 supersymmetric gauge theories in [1]. In [1,4], more examples of the same type of duality were described, of which we record two more here. Now consider B1 = N (12, 2)///Sp(2)
(D.5)
B2 = Omin (E 7 )///Sp(1).
(D.6)
and
Here Sp(1) acts on Omin (E 7 ) through the maximal subgroup Sp(1)×SO(12) ⊂ E 7 . The quaternionic dimension of B1,2 is 14, and both have triholomorphic actions of SO(12). We believe B1 = B2 as hyperkähler cones. For an example which involves Omin (E 8 ), consider C1 = (Z ⊕ N (11, 3))///Sp(3).
(D.7)
Here Z is a pseudoreal irreducible representation of Sp(3) of quaternionic dimension 7, which arises as ∧3C X = Z ⊕ X,
(D.8)
where X = C6 is the defining representation of Sp(3). Let us take another hyperkähler quotient C2 = Omin (E 8 )///SO(5),
(D.9)
where SO(5) acts via embedding SO(5) × SO(11) ⊂ SO(16) ⊂ E 8 .
(D.10)
It is easy to check that C1,2 are both of quaternionic dimension 19, and SO(11) acts triholomorphically on both C1 and C2 . We predict that C1 = C2 as hyperkähler cones.
410
D. Gaiotto, A. Neitzke, Y. Tachikawa
References 1. Argyres, P.C., Seiberg, N.: S-duality in N = 2 supersymmetric gauge theories. JHEP 0712, 088 (2007) 2. Minahan, J.A., Nemeschansky, D.: An N = 2 superconformal fixed point with E 6 global symmetry. Nucl. Phys. B 482, 142 (1996) 3. Aharony, O., Tachikawa, Y.: A holographic computation of the central charges of d = 4, N = 2 SCFTs. JHEP 0801, 037 (2008) 4. Argyres, P.C., Wittig, J.R.: Infinite coupling duals of N = 2 gauge theories and new rank 1 superconformal field theories. JHEP 0801, 074 (2008) 5. Kronheimer, P.B.: Instantons and the geometry of the nilpotent variety. J. Diff. Geom. 32, 473 (1990) 6. Joseph, A.: The minimal orbit in a simple Lie algebra and its associated maximal ideal. Ann. Sci. École Norm. Sup. Ser. 4 9, 1 (1976) 7. Swann, A.: Hyperkähler and quaternionic Käher geometry. Math. Ann. 289, 421 (1991) 8. Brylinski, R.: Instantons and Kähler geometry of nilpotent orbits. In: Representation Theories and Algebraic Geometry. NATO Adv. Sci. Inst. Ser. C Math. Phys. Sci. 514, Dordrecht: Kluwer, 1998, pp. 85–125 9. Hitchin, N.J., Karlhede, A., Lindström, U., Roˇcek, M.: Hyperkähler metrics and supersymmetry. Commun. Math. Phys. 108, 535 (1987) 10. Biquard, O., Gauduchon, P.: Hyper-Kähler metrics on cotangent bundles of Hermitian symmetric spaces. In: Lecture Notes in Pure and Appl. Math. 184, Newyork: Dekker, 1997, pp. 287–298 11. Argyres, P.C., Plesser, M.R., Seiberg, N.: The Moduli Space of N = 2 SUSY QCD and Duality in N = 1 SUSY QCD. Nucl. Phys. B 471, 159 (1996) 12. Vainshtein, A.I., Zakharov, V.I., Novikov, V.A., Shifman, M.A.: ABC of instantons. Sov. Phys. Usp. 25, 195 (1982) [Usp. Fiz. Nauk 136, 553 (1982)] 13. Garfinkle, D.: A new construction of the Joseph ideal. MIT thesis, 1982. Available on-line at the service ‘MIT Theses in DSpace.’ http://dspace.mit.edu/handle/1721.1/15620, 1982 (see chap. III) 14. Kobak, P., Swann, A.: The hyperkähler geometry associated to Wolf spaces. Boll. Unione Mat. Ital. Serie 8, Sez. B Artic. Ric. Mat. 4, 587 (2001) 15. Antoniadis, I., Pioline, B.: Higgs branch, hyperkähler quotient and duality in SUSY N = 2 Yang-Mills theories. Int. J. Mod. Phys. A 12, 4907 (1997) Communicated by A. Kapustin
Commun. Math. Phys. 294, 411–437 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0955-5
Communications in
Mathematical Physics
Isometric Immersions and Compensated Compactness Gui-Qiang Chen1 , Marshall Slemrod2 , Dehua Wang3 1 Department of Mathematics, Northwestern University, Evanston, IL 60208, USA.
E-mail:
[email protected]
2 Department of Mathematics, University of Wisconsin, Madison, WI 53706, USA.
E-mail:
[email protected]
3 Department of Mathematics, University of Pittsburgh, Pittsburgh, PA 15260, USA.
E-mail:
[email protected] Received: 27 December 2008 / Accepted: 10 September 2009 Published online: 27 November 2009 – © Springer-Verlag 2009
Abstract: A fundamental problem in differential geometry is to characterize intrinsic metrics on a two-dimensional Riemannian manifold M2 which can be realized as isometric immersions into R3 . This problem can be formulated as initial and/or boundary value problems for a system of nonlinear partial differential equations of mixed elliptic-hyperbolic type whose mathematical theory is largely incomplete. In this paper, we develop a general approach, which combines a fluid dynamic formulation of balance laws for the Gauss-Codazzi system with a compensated compactness framework, to deal with the initial and/or boundary value problems for isometric immersions in R3 . The compensated compactness framework formed here is a natural formulation to ensure the weak continuity of the Gauss-Codazzi system for approximate solutions, which yields the isometric realization of two-dimensional surfaces in R3 . As a first application of this approach, we study the isometric immersion problem for two-dimensional Riemannian manifolds with strictly negative Gauss curvature. We prove that there exists a C 1,1 isometric immersion of the two-dimensional manifold in R3 satisfying our prescribed initial conditions. To achieve this, we introduce a vanishing viscosity method depending on the features of initial value problems for isometric immersions and present a technique to make the a priori estimates including the L ∞ control and H −1 –compactness for the viscous approximate solutions. This yields the weak convergence of the vanishing viscosity approximate solutions and the weak continuity of the Gauss-Codazzi system for the approximate solutions, hence the existence of an isometric immersion of the manifold into R3 satisfying our initial conditions. The theory is applied to a specific example of the metric associated with the catenoid. 1. Introduction A fundamental problem in differential geometry is to characterize intrinsic metrics on a two-dimensional Riemannian manifold M2 which can be realized as isometric immersions into R3 (cf. Yau [40]; also see [21,34,36]). Important results have been achieved
412
G.-Q. Chen, M. Slemrod, D. Wang
for the embedding of surfaces with positive Gauss curvature which can be formulated as an elliptic boundary value problem (cf. [21]). For the case of surfaces of negative Gauss curvature where the underlying partial differential equations are hyperbolic, the complimentary problem would be an initial or initial-boundary value problem. Hong in [23] first proved that complete negatively curved surfaces can be isometrically immersed in R3 if the Gauss curvature decays at a certain rate in the time-like direction. In fact, a crucial lemma in Hong [23] (also see Lemma 10.2.9 in [21]) shows that, for such a decay rate of the negative Gauss curvature, there exists a unique global smooth, small solution forward in time for prescribed smooth, small initial data. Our main theorem, Theorem 5.1(i), indicates that in fact we can solve the corresponding problem for a class of large non-smooth initial data. Possible implication of our approach may be in existence theorems for equilibrium configurations of a catenoidal shell as detailed in Vaziri-Mahedevan [39]. When the Gauss curvature changes sign, the immersion problem then becomes an initial-boundary value problem of mixed elliptic-hyperbolic type, which is still under investigation. The purpose of this paper is to introduce a general approach, which combines a fluid dynamic formulation of balance laws with a compensated compactness framework, to deal with the isometric immersion problem in R3 (even when the Gauss curvature changes sign). In Sect. 2, we formulate the isometric immersion problem for two-dimensional Riemannian manifolds in R3 via solvability of the Gauss-Codazzi system. In Sect. 3, we introduce a fluid dynamic formulation of balance laws for the Gauss-Codazzi system for isometric immersions. Then, in Sect. 4, we provide a compensated compactness framework and present one of our main observations that this framework is a natural formulation to ensure the weak continuity of the Gauss-Codazzi system for approximate solutions, which yields the isometric realization of two-dimensional surfaces in R3 . A generalization of this approach to higher dimensional immersions has been given in Chen-Slemrod-Wang [8]. In Sect. 5, as a first application of this approach, we focus on the isometric immersion problem of two-dimensional Riemannian manifolds with strictly negative Gauss curvature. Since the local existence of smooth solutions follows from the standard hyperbolic theory, we are concerned here with the global existence of solutions of the initial value problem with large initial data. The metrics gi j we study have special structures and forms usually associated with the catenoid of revolution when g11 = g22 = cosh(x) and g12 = 0. For these cases, while Hong’s theorem [23] applies to obtain the existence of a solution for small smooth initial data, our result yields a large-data existence theorem for a C 1,1 isometric immersion. To achieve this, we introduce a vanishing viscosity method depending on the features of the initial value problem for isometric immersions and present a technique to make the a priori estimates including the L ∞ control and H −1 –compactness for the viscous approximate solutions. This yields the weak convergence of the vanishing viscosity approximate solutions and the weak continuity of the Gauss-Codazzi system for the approximate solutions, hence the existence of a C 1,1 –isometric immersion of the manifold into R3 with prescribed initial conditions. We remark in passing that, for the fundamental ideas and early applications of compensated compactness, see the classical papers by Tartar [38] and Murat [32]. For applications to the theory of hyperbolic conservation laws, see for example [4,10,13,18,37]. In particular, the compensated compactness approach has been applied in [3,6,11,12, 25,26] to the one-dimensional Euler equations for unsteady isentropic flow, allowing for cavitation, in Morawetz [29,30] and Chen-Slemrod-Wang [7] for two-dimensional
Isometric Immersions and Compensated Compactness
413
steady transonic flow away from stagnation points, and in Chen-Dafermos-SlemrodWang [5] for subsonic-sonic flows. 2. The Isometric Immersion Problem for Two-Dimensional Riemannian Manifolds in R3 In this section, we formulate the isometric immersion problem for two-dimensional Riemannian manifolds in R3 via solvability of the Gauss-Codazzi system. Let ⊂ R2 be an open set. Consider a map r : → R3 so that, for (x, y) ∈ , the two vectors {∂x r, ∂y r} in R3 span the tangent plane at r(x, y) of the surface r() ⊂ R3 . Then n=
∂x r × ∂y r |∂x r × ∂y r|
is the unit normal of the surface r() ⊂ R3 . The metric on the surface in R3 is ds 2 = dr · dr
(2.1)
ds 2 = (∂x r · ∂x r) (dx)2 + 2(∂x r · ∂y r) dxdy + (∂y r · ∂y r) (dy)2 .
(2.2)
or, in local (x, y)–coordinates,
Let gi j , i, j = 1, 2, be the given metric of a two-dimensional Riemannian manifold M parameterized on . The first fundamental form I for M on is I := g11 (dx)2 + 2g12 dxdy + g22 (dy)2 .
(2.3)
Then the isometric immersion problem is to seek a map r : → R3 such that dr · dr = I, that is, ∂x r · ∂x r = g11 , ∂x r · ∂y r = g12 , ∂y r · ∂y r = g22 ,
(2.4)
so that {∂x r, ∂y r} in R3 are linearly independent. The equations in (2.4) are three nonlinear partial differential equations for the three components of r(x, y). The corresponding second fundamental form is II := −dn · dr = h 11 (dx)2 + 2h 12 dxdy + h 22 (dy)2 ,
(2.5)
and (h i j )1≤i, j≤2 is the orthogonality of n to the tangent plane. Since n · dr = 0, then d(n · dr) = 0 implies −II + n · d 2 r = 0, i.e.,
2 II = (n · ∂x2 r) (dx)2 + 2(n · ∂xy r) dxdy + (n · ∂y2 r) (dy)2 .
The fundamental theorem of surface theory (cf. [14,21]) indicates that there exists a surface in R3 whose first and second fundamental forms are I and II if the smooth coefficients (gi j ) and (h i j ) of the two given quadratic forms I and II with (gi j ) > 0 satisfy the Gauss-Codazzi system. It is indicated in Mardare [28] (Theorem 9; also see [27]) that this theorem holds even when (h i j ) is only in L ∞ for given (gi j ) in C 1,1 , for
414
G.-Q. Chen, M. Slemrod, D. Wang
which the immersion surface is C 1,1 . This shows that, for the realization of a two-dimensional Riemannian manifold in R3 with given metric (gi j ) > 0, it suffices to solve for (h i j ) ∈ L ∞ determined by the Gauss-Codazzi system to recover r a posteriori. The simplest way to write the Gauss-Codazzi system (cf. [14,21]) is as (2)
(2)
(2)
∂x M − ∂y L = 22 L − 212 M + 11 N , (1)
(1)
(1)
(2.6)
∂x N − ∂y M = −22 L + 212 M − 11 N , with L N − M 2 = κ.
(2.7)
Here h 11 L=√ , |g|
h 12 M=√ , |g|
h 22 N=√ , |g|
2 , κ(x, y) is the Gauss curvature that is determined by the |g| = det (gi j ) = g11 g22 − g12 relation: R1212 (m) (m) (n) (m) (n) (m) κ(x, y) = , Ri jkl = glm ∂k i j − ∂ j ik + i j nk − ik n j , |g|
Ri jkl is the curvature tensor and depends on (gi j ) and its first and second derivatives, and (k)
i j =
1 kl g ∂ j gil + ∂i g jl − ∂l gi j 2
is the Christoffel symbol and depends on the first derivatives of (gi j ), where the summation convention is used, (g kl ) denotes the inverse of (gi j ), and (∂1 , ∂2 ) = (∂x , ∂y ). Therefore, given a positive definite metric (gi j ) ∈ C 1,1 , the Gauss-Codazzi system gives us three equations for the three unknowns (L , M, N ) determining the second fundamental form II . Note that, although (gi j ) is positive definite, R1212 may change sign and so does the Gauss curvature κ. Thus, as we will discuss in Sect. 3, the Gauss-Codazzi system (2.6)–(2.7) generically is of mixed hyperbolic-elliptic type, as in transonic flow (cf. [2,7,9,31]). In §3–4, we introduce a general approach to deal with the isometric immersion problem involving nonlinear partial differential equations of mixed hyperbolic-elliptic type by combining a fluid dynamic formulation of balance laws in §3 with a compensated compactness framework in §4. As an example of direct applications of this approach, in §5 we show how this approach can be applied to establish an isometric immersion of a two-dimensional Riemannian manifold with negative Gauss curvature in R3 . 3. Fluid Dynamic Formulation for the Gauss-Codazzi System From the viewpoint of geometry, the constraint condition (2.7) is a Monge-Ampère equation and the equations in (2.6) are integrability relations. However, our goal here is to put the problem into a fluid dynamic formulation so that the isometric immersion problem may be solved via the approaches that have shown to be useful in fluid dynamics
Isometric Immersions and Compensated Compactness
415
for solving nonlinear systems of balance laws. To achieve this, we formulate the isometric immersion problem via solvability of the Gauss-Codazzi system (2.6)–(2.7), that is, solving first for h i j , i, j = 1, 2, via (2.6) with constraint (2.7) and then recovering r a posteriori. To do this, we set L = ρv 2 + p,
M = −ρuv,
N = ρu 2 + p,
and set q 2 = u 2 + v 2 as usual. Then the Codazzi equations in (2.6) become the familiar balance laws of momentum: (2)
(2)
(2)
(1)
(1)
(1)
∂x (ρuv) + ∂y (ρv 2 + p) = −(ρv 2 + p)22 − 2ρuv12 − (ρu 2 + p)11 ,
(3.1)
∂x (ρu 2 + p) + ∂y (ρuv) = −(ρv 2 + p)22 − 2ρuv12 − (ρu 2 + p)11 , and the Gauss equation (2.7) becomes ρpq 2 + p 2 = κ.
(3.2)
From this, we can see that, if the Gauss curvature κ is allowed to be both positive and negative, the “pressure” p cannot be restricted to be positive. Our simple choice for p is the Chaplygin-type gas: 1 p=− . ρ Then, from (3.2), we find −q 2 +
1 = κ, ρ2
and hence we have the “Bernoulli” relation: 1 ρ= . 2 q +κ
(3.3)
This yields p = − q 2 + κ,
(3.4)
and the formulas for u 2 and v 2 : u 2 = p( p − M),
v 2 = p( p − L),
M 2 = (N − p)(L − p).
The last relation for M 2 gives the relation for p in terms of (L , M, N ), and then the first two give the relations for (u, v) in terms of (L , M, N ). We rewrite (3.1) as ∂x (ρuv) + ∂y (ρv 2 + p) = R1 , ∂x (ρu 2 + p) + ∂y (ρuv) = R2 , where R1 and R2 denote the right-hand sides of (3.1).
(3.5)
416
G.-Q. Chen, M. Slemrod, D. Wang
We now find the corresponding “geometric rotationality–continuity equations”. Multiplying the first equation of (3.5) by v and the second by u, and setting ∂x v − ∂y u = −σ, we see v R1 1 div(ρu, ρv) − ∂y κ = + σ u, ρ 2 ρ R2 1 u div(ρu, ρv) − ∂x κ = − σ v, ρ 2 ρ and hence 1ρ ∂y κ + 2v 1ρ div(ρu, ρv) = ∂x κ + 2u
div(ρu, ρv) =
R1 ρuσ + , v v R2 ρvσ − . u u
Thus, the right hand sides of (3.6) are equal, which gives a formula for σ : 1 1 1 σ = v − u . ρ∂ ρ∂ κ + R κ + R x 2 y 1 ρq 2 2 2
(3.6)
(3.7)
If we substitute this formula for σ into (3.6), we can write down our “rotationality-continuity equations” as 1 1 1 ∂x v − ∂y u = u − v , (3.8) κ + R κ + R ρ∂ ρ∂ y 1 x 2 ρq 2 2 2 1 ρu 1 ρv v u ∂x (ρu) + ∂y (ρv) = ∂x κ + ∂y κ + 2 R 1 + 2 R 2 . (3.9) 2 2 2q 2q q q In summary, the Gauss-Codazzi system (2.6)–(2.7), the momentum equations (3.1)– (3.4), and the rotationality-continuity equations (3.3) and (3.8)–(3.9) are all formally equivalent. However, for weak solutions, we know from our experience with gas dynamics that this equivalence breaks down. In Chen-Dafermos-Slemrod-Wang [5], the decision was made (as is standard in gas dynamics) to solve the rotationality-continuity equations and view the momentum equations as “entropy” equalities which may become inequalities for weak solutions. In geometry, this situation is just the reverse. It is the Gauss-Codazzi system that must be solved exactly and hence the rotationality-continuity equations will become “entropy” inequalities for weak solutions. The above issue becomes apparent when we set up “viscous” regularization that preserves the “divergence” form of the equations, which will be introduced in §5.3. This is crucial since we need to solve (3.8)–(3.9) exactly, as we have noted. To continue further our analogy, let us define the “sound” speed: c2 = p (ρ),
(3.10)
which in our case gives c2 =
1 . ρ2
(3.11)
Isometric Immersions and Compensated Compactness
417
Since our “Bernoulli” relation is (3.3), we see c2 = q 2 + κ.
(3.12)
Hence, under this formulation, (i) when κ > 0, the “flow” is subsonic, i.e., q < c, and system (3.1)–(3.2) is elliptic; (ii) when κ < 0, the “flow” is supersonic, i.e., q > c, and system (3.1)–(3.2) is hyperbolic; (iii) when κ = 0, the “flow” is sonic, i.e., q = c, and system (3.1)–(3.2) is degenerate. In general, system (3.1)–(3.2) is of mixed hyperbolic-elliptic type. Thus, the isometric immersion problem involves the existence of solutions to nonlinear partial differential equations of mixed hyperbolic-elliptic type. 4. Compensated Compactness Framework for Isometric Immersions In this section, we provide a compensated compactness framework and present our new observation that this framework is a natural formulation to ensure the weak continuity of the Gauss-Codazzi system for approximate solutions. Let a sequence of functions (L ε , M ε , N ε )(x, y), defined on an open subset ⊂ R2 , satisfy the following Framework (A): (A.1) |(L ε , M ε , N ε )(x, y)| ≤ C a.e. (x, y) ∈ , for some C > 0 independent of ε; −1 (); (A.2) ∂x M ε − ∂y L ε and ∂x N ε − ∂y M ε are confined in a compact set in Hloc (A.3) There exist oεj (1), j = 1, 2, 3, with oεj (1) → 0 in the sense of distributions as ε → 0 such that (2) ε (2) ε (2) ε ∂x M ε − ∂y L ε = 22 L − 212 M + 11 N + o1ε (1), (1)
(1)
(1)
∂x N ε − ∂y M ε = −22 L ε + 212 M ε − 11 N ε + o2ε (1),
(4.1)
and L ε N ε − (M ε )2 = κ + o3ε (1).
(4.2)
Then we have Theorem 4.1. (Compensated compactness framework). Let a sequence of functions (L ε , M ε , N ε )(x, y) satisfy Framework (A). Then there exists a subsequence (still labeled) ¯ M, ¯ N¯ ) as ε → 0 such (L ε , M ε , N ε )(x, y) that converges weak-star in L ∞ () to ( L, that ¯ M, ¯ N¯ )(x, y)| ≤ C a.e. (x, y) ∈ ; (i) |( L, (ii) the Gauss equation (2.7) is weakly continuous with respect to the subsequence ¯ M, ¯ N¯ ) as ε → 0; (L ε , M ε , N ε )(x, y) that converges weak-star in L ∞ () to ( L, (iii) the Codazzi equations in (2.6) hold. ¯ M, ¯ N¯ ) is a bounded weak solution to the Gauss-Codazzi system Specifically the limit ( L, (2.6)–(2.7), which yields an isometric realization of the corresponding two-dimensional Riemannian manifold in R3 .
418
G.-Q. Chen, M. Slemrod, D. Wang
Proof. By the div-curl lemma of Tartar-Murat [38,32] and the Young measure representation theorem for a uniformly bounded sequence of functions (cf. Tartar [38]), we employ (A.1)–(A.2) to conclude that there exist a family of Young measures {νx,y }(x,y)∈ and a subsequence (still labeled) (L ε , M ε , N ε )(x, y) that converges weak-star in L ∞ () ¯ M, ¯ N¯ ) as ε → 0 such that to ( L, ¯ M, ¯ N¯ )(x, y) = (νx,y , L , νx,y , M , νx,y , N ) a.e. (x, y) ∈ ; (a) ( L, ¯ ¯ N¯ )(x, y)| ≤ C (b) |( L, M, a.e. (x, y) ∈ ; (c) the following commutation identity holds: ¯ 2 − L¯ N¯ . νx,y , M 2 − L N = νx,y , M 2 − νx,y , L νx,y , N = ( M)
(4.3)
¯ M, ¯ N¯ ) also Since the equations in (4.1) are linear in (L ε , M ε , N ε ), then the limit ( L, satisfies the equations in (2.6) in the sense of distributions. Furthermore, condition (4.2) yields that νx,y , L N − M 2 = κ(x, y)
a.e. (x, y) ∈ .
(4.4)
The combination (4.3) with (4.4) yields the weak continuity of the Gauss equation with ¯ M, ¯ N¯ ) respect to the sequence (L ε , M ε , N ε ) that converges weak-star in L ∞ () to ( L, as ε → 0: ¯ 2 = κ. L¯ N¯ − ( M) ¯ M, ¯ N¯ ) is a bounded weak solution of the Gauss-Codazzi system (2.6)– Therefore, ( L, (2.7). Then the fundamental theorem of surface theory [27,28] implies an isometric realization of the corresponding two-dimensional Riemannian manifold in R3 . This completes the proof.
Remark 4.1. In the compensated compactness framework, Condition (A.1) can be relaxed to the following condition: For p > 2, (A.1)’ (L ε , M ε , N ε ) L p () ≤ C, for some C > 0 independent of ε. Then all the arguments for Theorem 4.1 follow only with the weak convergence in L p (), p > 2, replacing the weak-star convergence in L ∞ (), with the aid of the Young measure representation theorem for a uniformly L p bounded sequence of functions (cf. Ball [1]). There are various ways to construct approximate solutions by either analytical methods, such as vanishing viscosity and relaxation, or numerical methods, such as finite difference schemes and finite elements. Even though the solution to the Gauss-Codazzi system may eventually turn out to be more regular, especially in the region of positive Gauss curvature κ > 0, the point of considering weak solutions here is to demonstrate that such solutions may be constructed by merely using very crude estimates. Such estimates are available in a variety of approximating methods through basic energy-type estimates, besides the L ∞ estimate. On the other hand, in the region of negative Gauss curvature κ < 0, discontinuous solutions are expected so that the estimates can be improved at most up to BV in general. The compensated compactness framework (Theorem 4.1) indicates that, in order to find an isometric immersion, it suffices to construct a sequence of approximate solutions ¯ M, ¯ N¯ ) (L ε , M ε , N ε )(x, y) satisfying Framework (A), which yields its weak limit ( L, to be an isometric immersion. To achieve this through the fluid dynamic formulation
Isometric Immersions and Compensated Compactness
419
(3.1) and (3.3) (or (3.4)), a uniform L ∞ estimate of (u ε , v ε ) is required such that the sequence (L ε , M ε , N ε ) = (ρ ε (v ε )2 + p ε , −ρ ε u ε v ε , ρ ε (u ε )2 + p ε ) with pε = −
1 = (u ε )2 + (v ε )2 + κ ε ρ
satisfies Framework (A). The fluid dynamic formulation, (3.1) and (3.3) (or (3.4)), and the compensated compactness framework (Theorem 4.1) provide a unified approach to deal with various isometric immersion problems even for the case when the Gauss curvature changes sign, that is, for the equations of mixed elliptic-hyperbolic type.
5. Isometric Immersions of Two-Dimensional Riemannian Manifolds with Negative Gauss Curvature As a first example, in this section, we show how this approach can be applied to establish an isometric immersion of a two-dimensional Riemannian manifold with negative Gauss curvature in R3 . 5.1. Reformulation. In this case, κ < 0 in and, more specifically, κ = −γ 2 , γ > 0 in . For convenience, we assume γ ∈ C 1 in this section and rescale (L , M, N ) in this case as L L˜ = , γ
M M˜ = , γ
N N˜ = , γ
so that (2.7) becomes L˜ N˜ − M˜ 2 = −1. Then, without ambiguity, we redefine the “fluid variables” via L˜ = ρv 2 + p,
M˜ = −ρuv,
N˜ = ρu 2 + p,
and set q 2 = u 2 + v 2 , where we have still used (u, v, p, ρ) as the scaled variables and will use them hereafter (although they are different from those in §2–§4). Then the equations in (2.6) possess the same form of balance laws of momentum: ∂x (ρuv) + ∂y (ρv 2 + p) = R1 , ∂x (ρu 2 + p) + ∂y (ρuv) = R2 ,
(5.1)
420
G.-Q. Chen, M. Slemrod, D. Wang
where (2) (2) (2) R1 := −(ρv 2 + p)˜ 22 − 2ρuv ˜ 12 − (ρu 2 + p)˜ 11 , (1) (1) R2 := −(ρv + p)˜ 22 − 2ρuv ˜ 12 − (ρu 2 γy γx (1) (1) (1) (1) ˜ 11 = 11 + , ˜ 12 = 12 + , 2
γ
(2) (2) ˜ 11 = 11 ,
(2) ˜ 12
γx (2) , = 12 + 2γ
2γ
(5.2)
(1) + p)˜ 11 , (1) (1) ˜ 22 = 22 ,
(2) (2) ˜ 22 = 22 +
(5.3)
γy . γ
Furthermore, the constraint L˜ N˜ − M˜ 2 = −1 becomes ρpq 2 + p 2 = −1.
(5.4)
From p = − ρ1 and (5.4), we have the “Bernoulli” relation: 1 ρ= or p = − q 2 − 1, q2 − 1
(5.5)
˜ ˜ 2 = ( N˜ − p)( L˜ − p). ( M) u 2 = p( p − N˜ ), v 2 = p( p − L),
(5.6)
which yields
˜ M, ˜ N˜ ), and the first Then the last relation in (5.6) gives the relation for p in terms of ( L, ˜ ˜ ˜ two give the relations for (u, v) in terms of ( L, M, N ). Similar to the calculation in §3, we can write down our “rotationality–continuity equations” as 1 (v R2 − u R1 ) =: S1 , ρq 2 v u ∂x (ρu) + ∂y (ρv) = 2 R1 + 2 R2 =: S2 . q q
∂x v − ∂y u = −
(5.7) (5.8)
Under the new scaling, the “sound” speed is c2 = p (ρ) =
1 > 0. ρ2
(5.9)
Then the “Bernoulli” relation (3.3) yields c2 = q 2 − 1.
(5.10)
Therefore, q > c, and the “flow” is always supersonic, i.e., the system is purely hyperbolic.
Isometric Immersions and Compensated Compactness
421
5.2. Riemann invariants. In polar coordinates (u, v) = (q cos θ, q sin θ ), we use q 2 = 1 + ρ12 and ρq 2 = ρ + ρ1 from (5.5) to obtain (2) (2) (2) (2) (2) R1 = −ρq 2 sin2 θ ˜ 22 − 2ρq 2 sin θ cos θ ˜ 12 − ρq 2 cos2 θ ˜ 11 − p ˜ 22 + ˜ 11 1 1 (2) (2) (2) 2 2 ˜ ˜ sin θ 22 − 2ρq sin θ cos θ 12 − ρ + cos2 θ ˜ 11 =− ρ+ ρ ρ 1 (2) (2) ˜ 22 + ˜ 11 + ρ 1 (2) (2) (2) = −ρ sin2 θ ˜ 22 + cos2 θ ˜ 22 − 2ρq 2 sin θ cos θ ˜ 12 ρ 1 (2) (2) − ρ cos2 θ ˜ 11 + sin2 θ ˜ 11 ρ 1 (2) (2) = −ρ 1 − cos2 θ ˜ 22 + cos2 θ ˜ 22 ρ 1 (2) (2) (2) − 2ρq 2 sin θ cos θ ˜ 12 − ρ 1 − sin2 θ ˜ 11 + sin2 θ ˜ 11 ρ 1 1 (2) (2) (2) 2 2 ˜ ˜ = ρ 1 + 2 cos θ 22 − 2ρq sin θ cos θ 12 + ρ 1 + 2 sin2 θ ˜ 11 ρ ρ (2) (2) − ρ ˜ + ˜ 22
=
11
(2) ρq cos θ ˜ 22 − 2ρq 2 sin θ (2) (2) , − ρ ˜ 22 + ˜ 11 2
2
(2) (2) cos θ ˜ 12 + ρq 2 sin2 θ ˜ 11
and similarly, (1) (1) (1) (1) (1) R2 = ρq 2 cos2 θ ˜ 22 − 2ρq 2 sin θ cos θ ˜ 12 + ρq 2 sin2 θ ˜ 11 − ρ ˜ 22 + ˜ 11 . Then (5.7) and (5.8) become sin θ ∂x q + q cos θ ∂x θ − cos θ ∂y q + q sin θ ∂y θ = S1 ,
q2 − 1 sin θ cos θ ∂ ∂ S2 . q + sin θ ∂ θ + q − cos θ ∂ θ = − x x y y q(q 2 − 1) q(q 2 − 1) q
(5.11) (5.12)
That is, as a first-order system, (5.7) and (5.8) can be written as
− cos θ q sin θ sin θ q cos θ q q √ S1 1 cos θ sin θ ∂x θ + q(q 21−1) sin θ − cos θ ∂y θ = − q 2 −1 S2 . q(q 2 −1) q
(5.13) One of our main observations is that, under this reformation, the two coefficient matrices in (5.13) actually commute, which guarantees that they have common eigenvectors. The eigenvalues of the first and second matrices are cos θ λ± = sin θ ± , q2 − 1
sin θ µ± = − cos θ ± , q2 − 1
422
G.-Q. Chen, M. Slemrod, D. Wang
and the common left eigenvectors of the two coefficient matrices are 1 (± , 1). q q2 − 1 Thus, we may define the Riemann invariants W± = W± (θ, q) as 1 ∂θ W± = 1, ∂q W± = ± , q q2 − 1
(5.14)
which yields W± = θ ± arccos
1 . q
(5.15)
Now multiplication (5.13) by (∂q W± , ∂θ W± ) from the left yields q2 − 1 λ+ S2 , q (5.16) q2 − 1 λ− ∂q W− ∂x q + ∂θ W− ∂x θ + µ− ∂q W− ∂y q + ∂θ W− ∂y θ = S1 ∂q W− − S2 . q (5.17)
∂q W+ ∂x q + ∂θ W+ ∂x θ + µ+ ∂q W+ ∂y q + ∂θ W+ ∂y θ = S1 ∂q W+ −
From (5.15), ∂x W± = ∂q W± ∂x q + ∂θ W± ∂x θ,
∂y W± = ∂q W± ∂y q + ∂θ W± ∂y θ,
and we can write (5.16) and (5.17) as q2 − 1 S2 , S1 − λ+ ∂x W+ + µ+ ∂y W+ = q q q2 − 1 q2 − 1 1 S2 . λ− ∂x W− + µ− ∂y W− = − S1 − q q q2 − 1 1
(5.18) (5.19)
5.3. Vanishing viscosity method via parabolic regularization. Now we introduce a vanishing viscosity method via parabolic regularization to obtain the uniform L ∞ estimate by identifying invariant regions for the approximate solutions. First, if R˜1 and R˜2 denote the additional terms that should be added to the right-hand side of the Gauss-Codazzi system (5.1), our choice is R˜1 = ε∂y2 (ρv),
R˜2 = ε∂y2 (ρu),
(5.20)
which gives us the system of “viscous” parabolic regularization: ∂x (ρuv) + ∂y (ρv 2 + p) = R1 + ε∂y2 (ρv) = R1 + R˜ 1 , ∂x (ρu 2 + p) + ∂y (ρuv) = R2 + ε∂y2 (ρu) = R2 + R˜ 2 .
(5.21)
Isometric Immersions and Compensated Compactness
423
From Eqs. (3.8) and (3.9), we see 1 1 S˜1 = − 2 v R˜2 − u R˜1 , S˜2 = 2 v R˜1 + u R˜2 ρq q
(5.22)
should be added to S1 and S2 on the right-hand side of (3.8) and (3.9). In polar coordinates (u, v) = (q cos θ, q sin θ ), (5.31) becomes 2 S˜1 = ε ∂y θ ∂y (ρq) + ε∂y2 θ, ρq
1 S˜2 = ε ∂y2 (ρq) − ερ(∂y θ )2 . q
Note the identity 1 = ε∂y2 (arccsc(ρq)) ε∂y2 arccos q
1 = −ε∂y ∂y (ρq) − ρq ρ 2 q 2 − 1
1 = −ε∂y ∂y (ρq) − ρq ρ 2 q 2 − 1 Then S˜2 1 2 arccos − ε∂y = −ε∂ y ρ2 q
ρq
ρ2q
∂y2 (ρq)
S˜2 ε − (∂y θ )2 . ρ2 ρ
1 ρ2q 2
ε
−1
(5.23)
∂y (ρq) −
ε (∂y θ )2 , ρ
and thus 1 S˜2 2ε 2 2 2 ˜ ˜ ˜ S1 − (q − 1) S2 = S1 − 2 = ∂y θ ∂y (ρq) + ε∂y θ + ε∂y arccos ρ ρq q
1 ε + ε∂y ∂y (ρq) + (∂y θ )2 . 2 2 ρ ρq ρ q − 1 Since ∂y θ = ∂y W+ +
∂y (ρq) , ρq ρ 2 q 2 − 1
then 2εq ∂y W+ ∂y (ρq) + ε(∂y W+ )2 . S˜1 − (q 2 − 1) S˜2 = ε∂y2 W+ + ρ Similarly, using ∂y θ = ∂y W− −
∂y (ρq) , ρq ρ 2 q 2 − 1
we have − S˜1 − (q 2 − 1) S˜2 = −ε∂y2 W− −
2εq ∂y W− ∂y (ρq) + ε(∂y W− )2 . ρ
424
G.-Q. Chen, M. Slemrod, D. Wang
Thus, if we add the above S˜1 and S˜2 to the original S1 and S2 , (5.18) and (5.19) become ∂ W+ ∂ W+ q q 2 − 1 λ+ + µ+ ∂x ∂y (5.24) 2εq ∂y W+ ∂y (ρq) + ε(∂y W+ )2 + S1 − (q 2 − 1)S2 , = ε∂y2 W+ + ρ ∂ W− ∂ W− q q 2 − 1 λ− + µ− ∂x ∂y 2εq ∂y W− ∂y (ρq) + ε(∂y W− )2 − S1 − (q 2 − 1)S2 . = −ε∂y2 W− − ρ
(5.25)
Substitution of R1 and R2 into S1 and S2 yields S1 ± (q 2 − 1)S2 1 (1) (1) (1) (1) (1) = −q sin θ ˜ 22 cos2 θ − 2˜ 12 sin θ cos θ + ˜ 11 sin2 θ − 2 ˜ 22 + ˜ 11 q 1 (2) (2) (2) (2) (2) − q cos θ −˜ 22 cos2 θ + 2˜ 12 sin θ cos θ − ˜ 11 sin2 θ + 2 ˜ 22 + ˜ 11 q 1 1 (1) (1) (1) (1) (1) q cos θ ˜ 22 cos2 θ − 2˜ 12 sin θ cos θ + ˜ 11 sin2 θ − 2 ˜ 22 + ˜ 11 ± ρ q 1 (2) (2) (2) (2) (2) . +q sin θ ˜ 22 cos2 θ − 2˜ 12 sin θ cos θ + ˜ 11 sin2 θ − 2 ˜ 22 + ˜ 11 q (5.26) Then system (5.24)–(5.25) is parabolic when λ+ > 0 and λ− < 0. Furthermore, setting (E, F, G) = (g11 , g12 , g22 ), we recall the following classical identities: 2G Fy − GG x − F G x G E x −2F Fx +F E y (1) (1) 11 = , , 22 = 2(E G−F 2 ) 2(E G − F 2 ) E G y − 2F Fy + F G x 2E Fx −E E y −F E x (2) (2) 11 = , , 22 = 2(E G−F 2 ) 2(E G − F 2 ) E G x − F Ey G E −F G (1) (2) , 12 = 2(EyG−F 2x) , 12 = 2(E G − F 2 ) ⎡ ⎢ (E G − F 2 )2 κ = det ⎣
− 21 E yy + Fxy − 21 G xx
1 2 Ex
Fx − 21 Fy
Fx − 21 G x
E
F
1 2 Gy
F
G
⎡
0
⎢ − det ⎣ 21 E y 1 2 Gx
and γ 2 = −κ.
1 2 Ey
⎤ 1 2 Gx
E
⎥ F ⎦,
F
G
⎤ ⎥ ⎦
Isometric Immersions and Compensated Compactness
425
(b)
(a) Fig. 1. Level sets W± = C±
5.4. L ∞ –estimate for the viscous approximate solutions. Based on the calculation above for the Riemann invariants, we now introduce an approach to make the L ∞ estimate. First we need to sketch the graphs of the level sets of W± . In the u − v plane, on W± = C± , we have u cos C± + v sin C± = 1, which shows that the level sets W± = C± are straight lines as in Fig. 1. As an example, we now focus on the case: F = 0,
E(x) = G(x).
(5.27)
Then we see (1)
11 =
E E E (1) (1) (2) (2) (2) , 12 = 0, 22 = − ; 11 = 0, 12 = , 22 = 0. 2E 2E 2E
Therefore, we have (1) ˜ 11 =
E γ E E γ (1) (1) (2) (2) (2) + , ˜ 12 = 0, ˜ 22 = − ; ˜ 11 = 0, ˜ 12 = + , ˜ 22 = 0, 2E γ 2E 2E 2γ
and (5.26) becomes κ 1 κ 1 2 E 2E − γ sin θ ∓ − + qγ cos θ q 2γ 2 ρ 2 q E 2γ 2 ρ q E (5.28) q E 1 κ q E 1 κ =− + 2 2 sin θ ∓ + 2 cos θ. 2 E ρ q κ 2ρ E q κ
S1 ± (q 2 − 1)S2 =
Define θ± (q) by
E − EE + q12 κκ E + =∓ tan θ± = ± ρ EE + ρ 21q 2 κκ ρ EE +
1 κ q2 κ 1
ρ2q2
κ κ
.
(5.29)
426
G.-Q. Chen, M. Slemrod, D. Wang
Then, at the critical points of W+ , (5.24) becomes 1 κ q E 2 + (tan θ− − tan θ ) cos θ = 0, ε∂y W+ + 2 E ρ2q 2 κ and, at the critical points of W− , (5.25) becomes q E 1 κ (tan θ − tan θ+ ) cos θ = 0. − ε∂y2 W− + + 2 2 2 E ρ q κ
(5.30)
(5.31)
If we fix the intersection point of θ± (q) at θ = 0, q = β, where β > 1 is a constant, then the above ordinary differential equation (5.29) becomes 1 κ (x) E (x) + = 0, β 2 κ(x) E(x)
(5.32)
i.e., 1 d ln(|κ(x)| β 2 E(x)) = 0. dx
Thus, we find 1
|κ(x)| β 2 E(x) = const. > 0. Since κ(x) < 0, then κ(x) = −κ0 E(x)−β , 2
where κ0 > 0 is a constant, and Eq. (5.29) becomes q 2 − 1(β 2 − q 2 ) . tan θ± = ± 2 β − (β 2 − 1)q 2
(5.33)
In (5.33), if β 2 − (β 2 − 1)q 2 = 0, we define the associated root as q = q0 :=
β2 β2 − 1
21
.
2 2 2 The denominator √ β − (β − 1)q < 0 if q > q0 . We also desire q0 < β, which is ensured if β > 2. The curves θ± (q) are independent of (x, y) and look like the sketch in Fig. 2. Assume that we have a solution E(x) to (5.32). It is easy to check that the term E 1 κ E + ρ 2 q 2 κ in (5.30) and (5.31) has the sign of −E (x) when q > q0 , since
E 1 κ −E 2 (β − 1)(q 2 − q02 ). + 2 2 = E ρ q κ Eq 2 For the moment, we assume −E > 0. Fix another constant α with q0 < α < β, and sketch the W± lines and the diamond-shaped region as in Fig. 3.
Isometric Immersions and Compensated Compactness
Fig. 2. Graphs of θ± for β >
427
√ 2
Fig. 3. Invariant regions
Since cos θ > 0 for θ ∈ (− π2 , π2 ), Eqs. (5.30) and (5.31) imply that, at any point where ∇W+ = 0 (respectively ∇W− = 0), >0 for θ > θ− (q), 2 ε∂y W+ <0 for θ < θ− (q),
428
G.-Q. Chen, M. Slemrod, D. Wang
ε∂y2 W−
>0 <0
for θ > θ+ (q), for θ < θ+ (q).
Hence, when λ+ > 0 and λ− < 0, by the parabolic maximum principle (cf. [19,35]), W+ has no internal maximum for θ > θ− (q), W+ has no internal minimum for θ < θ− (q), W− has no internal maximum for θ > θ+ (q), W− has no internal minimum for θ < θ+ (q). We will be able to conclude that the diamond-shaped region in Fig. 3 is an invariant region if we can show that it is contained in the strip where λ+ > 0 and λ− < 0. We will do this as follows. Recall that ∂θ W± = 1, then W± are increasing with increasing θ on any circle q = constant. In Fig. 3, we require that the line with constant W− through u = α, v = 0 intersect the curve θ+ , which is ensured if α is close to β due to the following calculations. The slope of curve θ+ at u = β, v = 0 can be computed from q 2 − 1(β 2 − q 2 ) v tan θ+ = = 2 , u β − (β 2 − 1)q 2 which implies
and thus
u q2 − 1 (β 2 − q 2 ), v= 2 β − (β 2 − 1)q 2 2 β2 − 1 dv >0 = du u=β,v=0 β2 − 2
for β >
√ 2.
The slope of the line W− through u = β, v = 0 can be computed from the equation of the level set: 1 = W− (0, β) = − arccos(β −1 ), θ − arccos q that is, d du which gives us
v 1 arctan − arccos = 0, u q dv 1 = . 2 du u=β,v=0 β −1
So the intersection will occur if
1 2 β2 − 1 > , 2 β2 − 2 β −1
Isometric Immersions and Compensated Compactness
429
Fig. 4. Graphs of λ± = 0
i.e., 2(β 2 − 1) > β 2 − 2, which always holds for nonzero β. Thus, locally near β, the desired intersection occurs. A similar argument shows that the line with constant W+ intersects the curve θ− locally near β. From the definition of λ± : cos θ λ± = sin θ ± , q2 − 1 we easily see that the lines λ± = 0 are v = ±1, u > 0, as sketched in Fig. 4. Notice that λ+ > 0 and λ− < 0 are in the region above λ+ = 0 and below λ− = 0. If we now super-impose Fig. 3 on top of Fig. 4 and choose appropriate α and β, we see that there is a region where the four-sided (diamond-shaped) region of Fig. 3 is entirely confined above λ+ = 0 and below λ− = 0 in Fig. 4. Hence, the parabolic maximum/minimum principles apply and the four-sided region is an invariant region. All these arguments yield the uniform L ∞ bounds for (u ε , v ε , p ε , ρ ε ), which implies |(L ε , M ε , N ε )| ≤ C, for some constant C > 0 depending only on the data and γ L ∞ . Finally, let us examine the following examples: 2
Example 5.1. Catenoid: E(x) = (cosh(cx)) β 2 −1 , κ(x) = −κ0 E(x)−β , where c = 0 and κ0 > 0 are two constants. Note that, in this example, E (x) > 0 for x > 0 and E (x) < 0 for x < 0. So our hypothesis for the invariant region −E (x) > 0 will hold when x < 0. For the relevant parabolic equation for x time-like in the half plane = {(x, y) : −x0 ≤ x ≤ 0, y ∈ R}, where x0 > 0 is arbitrary, with periodic initial data (q(−x0 , y), θ (−x0 , y)) in the four sided region, the parabolic maximum/minimum principles yield the invariant region for the y-periodic solution. These arguments yield the uniform L ∞ bounds for (u ε , v ε , p ε , ρ ε ) which implies that |(L ε , M ε , N ε )| ≤ C for some constant C > 0 depending only on the initial data and of course the given metric (E, F, G) with E(x) = G(x), F ≡ 0. We must give the initial data on the line x = −x0 < 0. Example 5.2. Helicoid: The metric associated with the helicoid is ds 2 = E(d X )2 + (dY )2
2
430
G.-Q. Chen, M. Slemrod, D. Wang
with E(Y ) = λ2 + Y 2 and the Gauss curvature κ=−
λ2 , (λ2 + Y 2 )2
where λ > 0 is a constant. We make a change of variables to rewrite the helicoid metric in isothermal coordinates, i.e., allow X, Y to depend on x, y so that ds 2 = (E X x2 + Yx2 )dx2 + 2(E X x X y + Yx Yy )dxdy + (E X y2 + Yy2 )dy 2 . Hence, if we set √ √ Yx = − E X y , Yy = E X x , then ds 2 = E(X x2 + X y2 )(dx2 + dy 2 ), which gives the metric in isothermal coordinates. The above equations for X and Y may be rewritten as Yx − √ = Xy, E and with
φ(Y ) =
Yy √ = X x, E
√ dY = ln(Y + λ2 + Y 2 ). √ λ2 + Y 2
Then we have −φx = X y , φy = X x , i.e., the Cauchy-Riemann equations. A convenient solution is given by φ = −x,
X = y,
which yields −x = ln(Y +
√
λ2 + Y 2 ),
that is, 1 Y = − (λ2 ex − e−x ). 2 Thus, in the new (x, y)-coordinates, 1 1 −λ2 −λ2 E = λ2 + Y 2 = λ2 + (λ4 e2x + e−2x ), κ = 2 = , 1 2 1 4 2x 2 4 (λ + Y 2 )2 −2x ) 2 2 λ + 4 (λ e + e and
ds 2 =
1 2 1 4 2x λ + (λ e + e−2x ) (dx2 + dy 2 ). 2 4
Isometric Immersions and Compensated Compactness
431
Hence, we have κ (x) 2E (x) = , E(x) κ(x) √ and so relation (5.32) is satisfied with β = 2, thus tan θ± = ± q 2 − 1, −
the graphs of θ± never intersect, and thus our method of invariant regions does not apply to this example. Example 5.3. Torus: The metric for the torus is usually written as ds 2 = Ed X 2 + b2 dY 2 with E = (a + b cos Y )2 , κ(Y ) =
cos Y , b(a + b cos Y )
where a > b > 0 are constants. The same argument as given in Example 5.2 above yields the metric in isothermal coordinates as ds 2 = E(dx2 + dy 2 ) with E = (a + b cos Y )2 = (a + b cos(φ −1 (x))2 ), where
√ a 2 − b2 sin Y φ(Y ) = √ arctan , b + a cos Y a 2 − b2 b
κ(x) =
cos(φ −1 (x)) cos Y . = b(a + b cos Y ) b a + b cos(φ −1 (x))
A direct calculation yields a(φ −1 (x)) tan(φ −1 (x)) κ (x) =− , κ(x) a + b cos(φ −1 (x)) and the ratio κ (x) κ(x)
E (x) b(φ −1 (x)) sin(φ −1 (x)) =− , E(x) a + b cos(φ −1 (x))
E (x) a = E(x) b cos(φ −1 (x))
is not a constant. So (5.32) does not hold and unfortunately our method of invariant regions does not apply.
432
G.-Q. Chen, M. Slemrod, D. Wang
−1 −1 5.5. Hloc –compactness. We now show how the Hloc –compactness can be achieved for the viscous periodic approximate solutions via parabolic regularization. In §5.4, Example 5.1, for any initial data in the four-sided region which is periodic in y with period P, we have a uniform L ∞ estimate on (u ε , v ε , p ε , ρ ε ) as the periodic solution to the viscous equations (5.21). From the equations in (5.23), we have
v u 1 ∂2 R + R + ε (ρq) − ερ(∂y θ )2 1 2 q2 q2 q ∂y 2 1 = B(x, y) + ε ∂y2 (ρq) − ερ(∂y θ )2 , q
∂x (ρu) + ∂y (ρv) =
(5.34)
where ρ 1 1 (2) (2) (2) sin θ − ρq 2 sin2 θ − ˜ 22 − ρq 2 sin(2θ)˜ 12 − ρq 2 cos2 θ − ˜ 11 q ρ ρ ρ 1 1 (1) (1) (1) ˜ 22 −ρq 2 sin(2θ)˜ 12 − ρq 2 cos2 θ − ˜ 11 . + cos θ − ρq 2 sin2 θ − q ρ ρ
B(x, y) =
Our L ∞ estimate in §5.4, Example 5.1, guarantees that B(x, y) is uniformly bounded with respect to ε. Using the periodicity, we find that, for any fixed x0 > 0,
0 P ∂y q 1 2 ∂y (ρq)dydx ∂y (ρq)dydx = q2 −x0 0 q −x0 0 0 P 3 0 P ∂y q 3 ρ (∂y q)2 −ρ ∂ q dydx = − dydx. = y 2 q q2 −x0 0 −x0 0 0
P
Now integrating both sides of (5.34) over = {(x, y) : −x0 ≤ x ≤ 0, 0 ≤ y ≤ P}, we find 0 P 3 ρ (∂y q)2 2 ε + ρ(∂y θ ) dydx q2 −x0 0 0 P P = B(x, y) dydx − ((ρu)(0, y) − (ρu)(x0 , y)) dy −x0
0
0
≤ C,
where C > 0 is independent of ε, but may depend on x0 and P. This implies that √
ε∂y θ,
√ ε∂y q
2 are in L loc () uniformly in ε.
Therefore, we have Proposition 5.1. Consider the viscous system (5.21) in = {(x, y) : −x0 ≤ x ≤ 0, y ∈ R} with periodic initial data (q, θ )|x=−x0 = (q0 (y), θ0 (y)). Then √
ε∂y q,
√ εθy
2 are in L loc () uniformly in ε.
Isometric Immersions and Compensated Compactness
433
Using Proposition 5.1 and the viscous system (5.21), we conclude that −1 (). ∂x M˜ ε − ∂y L˜ ε , ∂x N˜ ε − ∂y M˜ ε are compact in Hloc
Since γ ∈ C 1 , we conclude that (L ε , M ε , N ε )(x, y) = γ (x)( L˜ ε , M˜ ε , N˜ ε )(x, y) satisfies Framework (A) in §4. Then the compensated compactness framework (Theorem 4.1) implies that there is a subsequence (still labeled) (L ε , M ε , N ε )(x, y) that converges ¯ M, ¯ N¯ ) as ε → 0 such that the limit ( L, ¯ M, ¯ N¯ ) is a bounded, periodic weak-star to ( L, ¯ M, ¯ N¯ ) is a weak weak solution to the Gauss-Codazzi system (2.6)–(2.7). Therefore, ( L, solution of (2.6)–(2.7). We summarize this as Proposition 5.2. Proposition 5.2. For the initial value problem of Proposition 5.1, (L ε , M ε , N ε ) possesses a weak-star convergent subsequence that converges to a periodic weak solution of the associated initial value problem for the Gauss-Codazzi system (2.6)–(2.7) when ε → 0. 5.6. Existence of isometric immersions: Main theorem and examples. We now focus on the case (5.27): F = 0, and E = G depends only on x to state an existence result for isometric immersions and analyze examples for this case. Let us look for a special solution: (θ, q) = (0, β)
(constant state),
for the Gauss-Codazzi system for the case (5.27). In this case, (1) ˜ 11 =
E γ E E γ (1) (1) (2) (2) (2) + , ˜ 12 = 0, ˜ 22 = − ; ˜ 11 = 0, ˜ 12 = + , ˜ 22 = 0, 2E γ 2E 2E 2γ
and the Gauss-Codazzi system (5.1) becomes 1 E E E γ γ = −p + (ρq 2 + p)( + ) = ρq 2 + (ρq 2 + p) ∂x ρ 2E 2E γ 2E γ γ E 1 +ρ . ρ= = ρq 2 2 2E γ q −1 When q(x) ≡ β, this reduces to 1 γ E = − , β2 γ 2E or 1 κ (x) E = − . β 2 κ(x) E
(5.35)
Hence, q(x) = β becomes an exact solution precisely in this special case. Our theorem given below shows that, in fact, the prescribed initial conditions in this special case can be satisfied and that, for this choice of E, F, G, there exists a weak solution for arbitrary bounded data in our diamond-shaped region (see Figs. 3–4) (especially when α ∈ (1, β) is sufficiently close to β).
434
G.-Q. Chen, M. Slemrod, D. Wang 2
β 2 −1 , Theorem 5.1. For the catenoid √ with the metric E(x) = G(x) = (cosh(cx)) F(x) = 0, c = 0, and β > 2 in Example 5.1, assume that the initial data
(q, θ )|x=−x0 = (q0 (y), θ0 (y)) := (q(−x0 , y), θ (−x0 , y))
(5.36)
is L ∞ and lies in the diamond-shaped region of Figs. 3–4. Then the Gauss-Codazzi system (2.6)–(2.7) has a weak solution in = {(x, y) : −x0 ≤ x ≤ 0, y ∈ R} with the initial data (5.36). Proof. Step 1. For the initial data (q0 (y), θ0 (y)), we can find (q0P (y), θ0P (y)) for P > 0 such that (i) q0P , θ0P ∈ C 1 (R), q0P , θ P are periodic with period P; (ii) q0P → q0 , θ0P → θ0 a.e. in R and weakly in L ∞ (R) as P → ∞. In particular, the functions q0P and θ0P are bounded in L ∞ (R), and (q0P , θ0P ) converges p to (q0 , θ0 ) in L loc (R), p ∈ [1, ∞), as P → ∞. This can be achieved by the standard symmetric mollification procedure: First truncate the initial data (W−,0 , W+,0 ) = (W− (q0 , θ0 ), W+ (q0 , θ0 )) in the interval − P2 ≤ y ≤ P2 and make the periodic extension to the whole space y ∈ R, and then take the standard symmetric mollification approximation to get the C ∞ approximate sequence P , W P ) that yields the corresponding C ∞ approximate sequence (W−,0 +,0 (q0P , θ0P )
=
W P − W P −1 W P + W P +,0 −,0 +,0 −,0 cos , 2 2 −W
converging to (q0 , θ0 ) = ((cos( +,0 2 −,0 ))−1 , +,0 2 −,0 ) a.e. in R as P → ∞. Since the standard symmetric mollification is an average-smoothing operator, then the approximate sequence (q0P , θ0P ) still lies in our diamond-shaped region of Figs. 3–4. W
W
+W
Step 2. Following the arguments in Sect. 5.4, we can establish the uniform L ∞ a priori estimates for the corresponding viscous solutions in two parameters > 0 and P > 0. For fixed P, then we can show that there exists a unique periodic viscous solution with period P to the parabolic system (5.21), which can be achieved by combining the stan−1 dard local existence theorem with the L ∞ estimates. The Hloc –compactness follows from the argument in §5.5. For fixed P, letting the viscous coefficient tend 0, we employ Proposition 5.2 to obtain the global periodic weak solution (L P , M P , N P ) of the Gauss-Codazzi system (2.6)–(2.7), periodic in y with period P, in the region = {(x, y) : −x0 ≤ x ≤ 0, y ∈ R}. Step 3. Since the sequence of periodic solutions (L P , M P , N P ) still stays in the invariant region, which yields the uniform L ∞ bound in P as P → ∞. This uniform bound also −1 –compactness of yields the Hloc ∂x M˜ P − ∂y L˜ P , ∂x N˜ P − ∂y M˜ P . Using the compensated compactness framework (Theorem 4.1) again and letting P → ∞, we obtain a global weak solution (L , M, N ) of the Gauss-Codazzi system (2.6)–(2.7) in the region = {(x, y) : −x0 ≤ x ≤ 0, y ∈ R}.
Isometric Immersions and Compensated Compactness
435
Fig. 5. C 1,1 -catenoid in the (x, y, z)–coordinates
By the even symmetry of the problem, we can obtain again a weak solution on the domain 0 ≤ x ≤ x0 , y ∈ R. Together, they form a weak solution in −x0 ≤ x ≤ x0 , y ∈ R if the same data is prescribed at x = x0 and x = −x0 . As before, the associated immersion is in C 1,1 . Theorem 5.2. Assume that identical initial data (5.36) is prescribed at x = x0 and x = −x0 , where x0 > 0 is arbitrary, and the data is in L ∞ and lies in the diamondshaped region of Figs. 3–4 for the case of the catenoid (as given in Theorem 5.1), then the initial value problem (2.6)–(2.7) and (5.36) has a weak solution in L ∞ ([−x0 , x0 ] × R). This yields a C 1,1 ([−x0 , x0 ] × R) immersion of the Riemannian manifold into R3 . Remark 5.1. The catenoid with circular cross-section is sketched in Fig. 5. Our theorem asserts the existence of a C 1,1 -surface for the associated metric for a class of identical non-circular cross-sections prescribed at x = x0 and x = −x0 . Since we give periodic cross sections z = z(y) at x = −x0 , our theorem yields a sequence of catenoid-like surfaces. Acknowledgements. This paper was completed when the authors attended the “Workshop on Nonlinear PDEs of Mixed Type Arising in Mechanics and Geometry”, which was held at the American Institute of Mathematics, Palo Alto, California, March 17–21, 2008. Gui-Qiang Chen’s research was supported in part by the National Science Foundation under Grants DMS-0807551, DMS-0720925, and DMS-0505473. Marshall Slemrod’s research was supported in part by the National Science Foundation under Grant DMS-0647554. Dehua Wang’s research was supported in part by the National Science Foundation under Grant DMS-0604362, and by the Office of Naval Research under Grant N00014-07-1-0668.
References 1. Ball, J.M.: A version of the fundamental theorem for Young measures. Lecture Notes in Phys. 344, Berlin: Springer, 1989, pp. 207–215 2. Bers, L.: Mathematical Aspects of Subsonic and Transonic Gas Dynamics. New York: John Wiley & Sons, Inc., London: Chapman & Hall, Ltd., 1958 3. Chen, G.-Q.: Convergence of the Lax-Friedrichs scheme for isentropic gas dynamics (III). Acta Math. Sci. 6, 75–120 (1986) (in English); 8, 243–276 (1988) (in Chinese)
436
G.-Q. Chen, M. Slemrod, D. Wang
4. Chen, G.-Q.: Euler equations and related hyperbolic conservation laws. In: Handbook of Differential Equations: Evolutionary Differential Equations, Vol. 2, Eds. Dafermos, C.M., Feireisl, E., Amsterdam: Elsevier Science B.V, 2005 pp. 1–104 5. Chen, G.-Q., Dafermos, C.M., Slemrod, M., Wang, D.: On two-dimensional sonic-subsonic flow. Commun. Math. Phys. 271, 635–647 (2007) 6. Chen, G.-Q., LeFloch, Ph.: Compressible Euler equations with general pressure law. Arch. Rat. Mech. Anal. 153, 221–259 (2000); Existence theory for the isentropic Euler equations. Arch. Rat. Mech. Anal. 166, 81–98 (2003) 7. Chen, G.-Q., Slemrod, M., Wang, D.: Vanishing viscosity method for transonic flow. Arch. Rat. Mech. Anal. 189, 159–188 (2008) 8. Chen, G.-Q., Slemrod, M., Wang, D.: Weak continuity of the Gauss-Codazzi-Ricci system for isometric embedding. Proc. Amer. Math. Soc. 2009 (to appear) 9. Courant, R., Friedrichs, K.O.: Supersonic Flow and Shock Waves. New York: Springer, 1948 10. Dafermos, C.M.: Hyperbolic Conservation Laws in Continuum Physics. 2nd edition, Berlin: SpringerVerlag, 2005 11. Ding, X., Chen, G.-Q., Luo, P.: Convergence of the Lax-Friedrichs scheme for isentropic gas dynamics (I)-(II). Acta Math. Sci. 5, 483–500, 501–540 (1985) (in English); 7, 467–480 (1987), 8, 61–94 (1988) (in Chinese) 12. DiPerna, R.J.: Convergence of viscosity method for isentropic gas dynamics. Commun. Math. Phys. 91, 1–30 (1983) 13. DiPerna, R.J.: Compensated compactness and general systems of conservation laws. Trans. Amer. Math. Soc. 292, 383–420 (1985) 14. do Carmo, M.P.: Riemannian Geometry. Transl. by F. Flaherty, Boston, MA: Birkhäuser, 1992 15. Dong, G.-C.: The semi-global isometric imbedding in R 3 of two-dimensional Riemannian manifolds with Gaussian curvature changing sign cleanly. J. Part. Diff. Eqs. 6, 62–79 (1993) 16. Efimov, N.V.: The impossibility in Euclidean 3-space of a complete regular surface with a negative upper bound of the Gaussian curvature. Dokl. Akad. Nauk SSSR (N.S.), 150, 1206–1209 (1963); Sov. Math. Dokl. 4, 843–846 (1963) 17. Efimov, N.V.: Surfaces with slowly varying negative curvature. Russ. Math. Surv. 21, 1–55 (1966) 18. Evans, L.C.: Weak Convergence Methods for Nonlinear Partial Differential Equations. CBMS-RCSM, 74, Providence, RI: Amer. Math. Soc., 1990 19. Evans, L.C.: Partial Differential Equations. Providence, RI: Amer. Math. Soc., 1998 20. Gromov, M.: Partial Differential Relations. Berlin: Springer-Verlag, 1986 21. Han, Q., Hong, J.-X.: Isometric embedding of Riemannian manifolds in Euclidean spaces. Providence, RI: Amer. Math. Soc., 2006 22. Hilbert, D.: Ueber flächen von constanter Gausscher Krümmung. Trans. Amer. Math. Soc. 2, 87–99 (1901) 23. Hong, J.-X.: Realization in R 3 of complete Riemannian manifolds with negative curvature. Commun. Anal. Geom. 1, 487–514 (1993) 24. Hong, J.-X.: Recent developments of realization of surfaces in R3 . AMS/IP Stud. Adv. Math. 20, Providence, RI: Amer. Math. Soc., 2001 25. Lions, P.-L., Perthame, B., Souganidis, P.: Existence and stability of entropy solutions for the hyperbolic systems of isentropic gas dynamics in Eulerian and Lagrangian coordinates. Comm. Pure Appl. Math. 49, 599–638 (1996) 26. Lions, P.-L., Perthame, B., Tadmor, E.: Kinetic formulation of the isentropic gas dynamics and p-systems. Commun. Math. Phys. 163, 169–172 (1994) 27. Mardare, S.: The fundamental theorem of surface theory for surfaces with little regularity. J. Elasticity 73, 251–290 (2003) 28. Mardare, S.: On Pfaff systems with L p coefficients and their applications in differential geometry. J. Math. Pure Appl. 84, 1659–1692 (2005) 29. Morawetz, C.S.: On a weak solution for a transonic flow problem. Comm. Pure Appl. Math. 38, 797– 818 (1985) 30. Morawetz, C.S.: On steady transonic flow by compensated compactness. Methods Appl. Anal. 2, 257–268 (1995) 31. Morawetz, C.S.: Mixed equations and transonic flow. J. Hyper. Diff. Eqs. 1, 1–26 (2004) 32. Murat, F.: Compacite par compensation. Ann. Suola Norm. Pisa (4) 5, 489–507 (1978) 33. Nash, J.: The imbedding problem for Riemannian manifolds. Ann. Math. (2) 63, 20–63 (1956) 34. Poznyak, È. G., Shikin, E.V.: Small parameters in the theory of isometric imbeddings of two-dimensional Riemannian manifolds in Euclidean spaces. In: Some Questions of Differential Geometry in the Large, Amer. Math. Soc. Transl. Ser. 2, 176, Providence, RI: Amer. Math. Soc., 1996, pp. 151–192 35. Protter, M.H., Weinberger, H.F.L.: Maximum Principles in Differential Equations. Berlin-HeidelbergNew York: Springer, 1984
Isometric Immersions and Compensated Compactness
437
36. Rozendorn, È. R.: Surfaces of negative curvature. In: Geometry, III, Encyclopaedia Math. Sci. 48, Berlin: Springer, 1992, pp. 87–178, 251–256 37. Serre, D.: Systems of Conservation Laws. Vols. 1–2, Cambridge: Cambridge University Press, 1999, 2000 38. Tartar, L.: Compensated compactness and applications to partial differential equations. In: Nonlinear Analysis and Mechanics. Heriot-Watt Symposium IV, Res. Notes in Math. 39, Boston-London: Pitman, 1979, pp. 136–212 39. Vaziri, A., Mahedevan, L.: Localized and extended deformations of elastic shells. Proc. National Acad. Sci. USA 105(23), 7913–7918 (2008) 40. Yau, S.-T.: Review of geometry and analysis. In: Mathematics: Frontiers and Perspectives. International Mathematics Union, Eds. V. Arnold, M. Atiyah, P. Lax, B. Mazur, Providence, RI: Amer. Math. Soc., 2000, pp. 353–401 Communicated by P. Constantin
Commun. Math. Phys. 294, 439–470 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0960-8
Communications in
Mathematical Physics
Renormalized Electron Mass in Nonrelativistic QED Jürg Fröhlich1, , Alessandro Pizzo2 1 Institute of Theoretical Physics, ETH Zürich, CH-8093 Zürich,
Switzerland. E-mail:
[email protected]
2 Department of Mathematics, University of California Davis,
One Shields Avenue, Davis, California 95616, USA. E-mail:
[email protected] Received: 5 February 2009 / Accepted: 27 July 2009 Published online: 5 December 2009 – © The Author(s) 2009. This article is published with open access at Springerlink.com
Abstract: Within the framework of nonrelativistic QED, we prove that, for small values of the coupling constant, the energy function, E P , of a dressed electron is twice differen∂2 E
tiable in the momentum P in a neighborhood of P = 0. Furthermore, P2 is bounded (∂| P|) from below by a constant larger than zero. Our results are proven with the help of iterative analytic perturbation theory.
I. Description of the Problem, Definition of the Model, and Outline of the Proof In this paper, we study problems connected with the renormalized electron mass in a model of quantum electrodynamics (QED) with nonrelativistic matter. We are interested in rigorously controlling radiative corrections to the electron mass caused by the interaction of the electron with the soft modes of the quantized electromagnetic field. The model describing interactions between nonrelativistic, quantum-mechanical charged matter and the quantized radiation field at low energies (i.e., energies smaller than the rest energy of an electron) is the “standard model”, see [7]. In this paper, we consider a system consisting of a single spinless electron, described as a nonrelativistic particle that is minimally coupled to the quantized radiation field, and photons. Electron spin can easily be included in our description without substantial complications. The physical system studied in this paper exhibits space translations invariance. The Hamiltonian, H , generating the time evolution, commutes with the vector operator, P, representing the total momentum of the system, which generates space translations. If an infrared regularization, e.g., an infrared cutoff σ on the photon frequency, is imposed on the interaction Hamiltonian, there exist single-electron or dressed one-electron states, as long as their momentum is smaller than the bare electron mass, m, of the electron. This means that a notion of mass shell in the energy momentum spectrum is meaningful for Also at IHES, Bures-sur-Yvette.
440
J. Fröhlich, A. Pizzo
velocities | P|/m smaller than the speed of light c; (with c ≡ m ≡ 1 in our units). Vectors { σ } describing dressed one-electron states are normalizable vectors in the Hilbert space H of pure states of the system. They are characterized as solutions of the equation < 1, H σ σ = E σP σ , | P|
(I.1)
where H σ is the Hamiltonian with an infrared cutoff σ in the interaction term and E σ , P If in the the energy of a dressed electron, is a function of the momentum operator P. joint spectrum of the components of P the support of the vector σ is contained in a ball centered at the origin and of radius less than 1 ≡ mc, then Eq. (I.1) has solutions; see = 0, Eq. (I.1) can be studied for the fiber vectors, σ , corresponding [6]. Since [H, P] ¯ P
of the total momentum (both the total momentum operator and points in to a value, P, its spectrum will henceforth be denoted by P – without danger of confusion). Thus we consider the equation H Pσ Pσ = E σP Pσ ,
(I.2)
and E σ is the value of where H σ is the fiber Hamiltonian at fixed total momentum P, P P Physically, states { σ } solving Eq. (I.1) describe a the function E zσ at the point z ≡ P. freely moving electron in the absence of asymptotic photons. It is an essential aspect of the “infrared catastrophe” in QED that Eq. (I.1) does not have any normalizable solution in the limit where the infrared cut-off σ tends to zero, and the underlying dynamical picture of a freely moving electron breaks down; see [5]. Nevertheless, the limiting behavior of the function E σ is of great interest for the P following reasons. As long as σ > 0, a natural definition of the renormalized electron mass, m r , is given by the formula ⎡ ⎤−1 ∂ 2 E σ | P| m r (σ ) := ⎣ | ⎦ . (I.3) 2 P=0 (∂| P|) (Note that E σ ≡ E σ is invariant under rotations.) Equation (I.3) is expected to remain P | P| meaningful in the limit σ → 0. In particular, the quantity on the R.H.S. of Eq. (I.3) is expected to be positive and bounded from above uniformly in the infrared cutoff σ . More importantly, one aims at mathematical control of the function ⎤−1 ⎡ ∂ 2 E σ | P| ⎦ := ⎣ m r (σ, | P|) (I.4) (∂| P|)2 in a full neighborhood, S, of P = 0, corresponding to a slowly moving electron (i.e., in the nonrelativistic regime). When combined with a number of other spectral properties of the Hamiltonian of nonrelativistic QED the condition ∂ 2 E σ
| P|
2 (∂| P|)
>0 ,
P ∈ S,
(I.5)
Renormalized Electron Mass in Nonrelativistic QED
441
uniformly in σ > 0, suffices to yield a consistent scattering picture in the limit when σ → 0 in which the electron exhibits infraparticle behavior. In fact, (I.5) is a crucial ingredient in the analysis of Compton scattering presented in [5,13]. Main results. Assuming the coupling constant, α, small enough, the following results follow. 1) The function | P| := lim
σ →0
∂ 2 E σ
| P|
2 (∂| P|)
(I.6)
< 1 }; furthermore, it is Hölder-continuous in is well defined for P ∈ S := { P | | P| 3 P. 2) The function E P := lim E σP
(I.7)
σ →0
is twice differentiable in P ∈ S and ∂ 2 E | P|
2 (∂| P|)
= | P| .
(I.8)
3) ∂ 2 E σ
| P| 2 α→0 (∂| P|)
lim
=
1 , m
P ∈ S,
(I.9)
uniformly in σ , where m is the bare electron mass. (Our results can be extended to a region S (inside the unit ball) of radius larger than 13 .) We wish to mention some related earlier results. Using operator-theoretic renormalization group methods, results (I.6) and (I.9) have been proven in [2] for the special value P = 0. The point P = 0 is exceptional, because the Hamiltonian H P is infrared regular at P = 0; it has a normalizable ground state. Thomas Chen (see [3]) has established the results in (I.8), (I.9) (using smooth infrared cut-offs) by a highly non-trivial extension of the analysis of [2] to arbitrary momenta P ∈ S. The procedure presented in our paper relies on iterative analytic perturbation theory (see Sect. II where this tool is recalled) that makes our proof substantially different and much shorter in comparison to a renormalization group approach. The main feature is a more transparent treatment of the so-called marginal terms of the interaction, where an essential role is played by explicit Bogoliubov transformations that transform the infrared representations of the CCR of photon creation- and annihilation operators = 0) back to the Fock determined by dressed one-particle states of fixed momentum P( representation. The use of these Bogoliubov transformations is a crucial device in our fight against the infrared problem. The way in which we are using them is new, at least in the context of mathematically rigorous results on the infrared problem in QED. In our paper, the regularity properties of E P come with an explicit control of the asymptotics of the fiber ground state eigenvectors σ as σ tends to zero. (This improves P earlier results in [6].) Along the lines of [1], these results are preparatory to developing an infrared finite algorithm for the asymptotic expansion of the renormalized electron mass
442
J. Fröhlich, A. Pizzo
in powers and, probably, logarithms of the finestructure constant α, up to an arbitrarily small remainder term. However, the expansion in the coupling constant α is not studied in this paper. With regard to ultraviolet corrections to the electron mass in nonrelativistic QED models, we refer the reader to [8,10,11 and 9]. In Sect. I.1 below, the model is defined rigorously. Then, for the convenience of the reader, in Sect. I.2 we outline the key ideas of the proof and present the organization of the remaining sections of the paper. I.1. Definition of the model. Hilbert space. The Hilbert space of pure state vectors of a system consisting of one non-relativistic electron interacting with the quantized electromagnetic field is given by H := Hel ⊗ F ,
(I.10)
where Hel = L 2 (R3 ) is the Hilbert space for a single Schrödinger electron; for expository convenience, we neglect the spin of the electron. The Hilbert space, F, used to describe the states of the transverse modes of the quantized electromagnetic field (the photons) in the Coulomb gauge is given by the Fock space F :=
∞
F (N ) ,
F (0) = C ,
(I.11)
N =0
where is the vacuum vector (the state of the electromagnetic field without any excited modes), and F (N ) := S N
N
h,
N ≥1,
(I.12)
j=1
where the Hilbert space h of state vectors of a single photon is h := L 2 (R3 × Z2 ).
(I.13)
Here, R3 is momentum space, and Z2 accounts for the two independent transverse polarizations (or helicities) of a photon. In (I.12), S N denotes the orthogonal projection onto the subspace of Nj=1 h of totally symmetric N -photon wave functions, which accounts for the fact that photons satisfy Bose-Einstein statistics. Thus, F (N ) is the subspace of F of state vectors corresponding to configurations of exactly N photons. Units. In this paper, we employ units such that Planck’s constant , the speed of light c, and the mass of the electron m are equal to 1. Hamiltonian. The dynamics of the system is generated by the Hamiltonian 2 x) x + α 1/2 A( −i ∇ H := + H f. 2
(I.14)
The (three-component) multiplication operator x ∈ R3 represents the position of the x . Furthermore, α > 0 electron. The electron momentum operator is given by p = −i ∇
Renormalized Electron Mass in Nonrelativistic QED
443
is the fine structure constant (which, in this paper, plays the rôle of a small parame x ) denotes the vector potential of the transverse modes of the quantized ter), and A( electromagnetic field in the Coulomb gauge, x) = 0 , x · A( ∇
(I.15)
cutoff at high photon frequencies. H f is the Hamiltonian of the quantized, free electromagnetic field. It is given by
a∗ a , H f := (I.16) d 3 k |k| k,λ k,λ λ=±
where a ∗ and ak,λ are the usual photon creation- and annihilation operators satisfying k,λ the canonical commutation relations ∗ [ak,λ , ak ,λ ] = δλλ δ(k − k ),
(I.17)
# # [ak,λ , ak ,λ ] = 0,
(I.18)
k ∈ R3 and λ, λ ∈ Z2 ≡ {±}, where a # = a or a ∗ . The vacuum vector ∈ F for k, is characterized by the condition ak,λ = 0, for all k ∈ R3 and λ ∈ Z2 ≡ {±}. The quantized electromagnetic vector potential is given by
d 3k x ∗ x −i k· ∗ i k· εk,λ , A( x ) := ak,λ k,λ ak,λ e + ε e λ=± B
|k|
(I.19)
(I.20)
3 where εk,− k,+ ,ε are photon polarization vectors, i.e., two unit vectors in R ⊗C satisfying ∗ εk,λ k,µ = δλµ , ·ε
k · εk,λ = 0,
(I.21)
for λ, µ = ±. The equation k · εk,λ = 0 expresses the Coulomb gauge condition. Moreover, B is a ball of radius centered at the origin in momentum space. Here,
represents an ultraviolet cutoff that will be kept fixed throughout our analysis. The vector potential defined in (I.20) is thus cut off in the ultraviolet. Throughout this paper, it will be assumed that ≈ 1 (the rest energy of an electron), and that α > 0 is sufficiently small. Under these assumptions, the Hamiltonian H is selfadjoint on D(H 0 ), i.e., on the domain of definition of the operator H 0 :=
x )2 (−i ∇ + Hf . 2
(I.22)
The perturbation H − H 0 is small in the sense of Kato. The operator representing the total momentum of the system consisting of the electron and the electromagnetic radiation field is given by P := p + P f ,
(I.23)
444
J. Fröhlich, A. Pizzo
x , and where with p = −i ∇ P f :=
λ=±
∗ d 3 k k ak,λ ak,λ
(I.24)
is the momentum operator associated with the photon field. The operators H and P are essentially selfadjoint on a common domain, and since = 0. The Hilbert the dynamics is invariant under translations, they commute, [H, P] space H can be decomposed into a direct integral over the joint spectrum, R3 , of the Their spectral measure is absolutely three components of the momentum operator P. continuous with respect to Lebesgue measure, and hence we have that ⊕ H P d 3 P , (I.25) H := where each fiber space H P is a copy of Fock space F. Remark. Throughout this paper, the symbol P stands for both a vector in R3 and the vector operator on H, representing the total momentum, depending on context. Simi of the total momentum larly, a double meaning is given to arbitrary functions, f ( P), operator. We recall that vectors ∈ H are given by sequences { (m) ( x ; k1 , λ1 ; . . . ; km , λm )}∞ m=0 ,
(I.26)
of functions, (m) , where (0) ( x ) ∈ L 2 (R3 ), of the electron position x and of m photon momenta k1 , . . . , km and helicities λ1 , . . . , λm , with the following properties: (i) (m) ( x ; k1 , λ1 ; . . . ; km , λm ) is totally symmetric in its m arguments (k j , λ j ) j=1,...,m , (ii) (m) is square-integrable, for all m, (iii) If and are two vectors in H, then ⎛ ∞ m
⎝ d 3 k j (m) ( x ; k1 , λ1 ; . . . ; km , λm ) d3x (, ) = m=0
λ j =±
j=1
⎞
x ; k1 , λ1 ; . . . ; km , λm )⎠ . × (m) (
(I.27)
We identify a square integrable function g( x ) with the sequence { (m) ( x ; k1 , λ1 ; . . . ; km , λm )}∞ m=0 ,
(I.28)
x ) ≡ g( x ), and (m) ( x ; k1 , λ1 ; . . . ; km , λm ) ≡ 0 for all m > 0; analowhere (0) ( gously, a square integrable function g (m) ( x ; k1 , λ1 ; . . . ; km , λm ), m ≥ 1, is identified with the sequence
{ (m ) ( x ; k1 , λ1 ; . . . ; km , λm )}∞ m =0 ,
(I.29)
Renormalized Electron Mass in Nonrelativistic QED
445
where (m) ( x ; k1 , λ1 ; . . . ; km , λm ) ≡ g (m) , and (m ) ( x ; k1 , λ1 ; . . . ; km , λm ) ≡ 0 for all m = m. From now on, a sequence describing a quantum state with a fixed number of photons is identified with its nonzero component wave function; vice versa, a wave function corresponds to a sequence according to the previous identification. The elements of the fiber space H P ∗ are obtained by linear combinations of the (improper) eigenvectors ∗ of the total momentum operator P with eigenvalue P ∗ , e.g., the plane wave ei P ·x is the eigenvector describing a state with an electron and no photon. Given any P ∈ R3 , there is an isomorphism, I P ,
I P : H P −→ F b ,
(I.30)
from the fiber space H P to the Fock space F b , acted upon by the annihilation- and
∗ i k· x a , and b∗ to e−i k· x a∗ , creation operators bk,λ , b , where bk,λ corresponds to e k,λ k,λ
k,λ
k,λ
and with vacuum f := I P (ei P·x ). To define I P more precisely, we consider a vector ψ( f (n) ; P) ∈ H P with a definite total momentum describing an electron and n photons. Its wave function in the variables ( x ; k1 , λ1 ; . . . , kn , λn ) is given by
ei( P−k1 −···−kn )·x f (n) (k1 , λ1 ; . . . ; kn , λn ),
(I.31)
where f (n) is totally symmetric in its n arguments. The isomorphism I P acts by way of I P ei( P−k1 −···−kn )·x f (n) (k1 , λ1 ; . . . ; kn , λn )
1 =√ d 3 k1 . . . d 3 kn f (n) (k1 , λ1 ; . . . ; kn , λn ) bk∗ ,λ · · · bk∗ ,λ f . (I.32) n n 1 1 n! λ1 ,...,λn Because the Hamiltonian H commutes with the total momentum, it preserves the fibers H P for all P ∈ R3 , i.e., it can be written as ⊕ H P d 3 P, (I.33) H = where H P : H P −→ H P .
(I.34)
∗ , and of the variable P, the fiber Hamiltonian Written in terms of the operators bk,λ , bk,λ H P is given by
H P :=
P − P f + α 1/2 A
where P f =
2
λ
Hf =
λ
2 + H f,
(I.35)
∗ d 3 k k bk,λ , bk,λ
(I.36)
∗ b , d 3 k |k|b k,λ k,λ
(I.37)
446
J. Fröhlich, A. Pizzo
and
A :=
B
λ
d 3k ∗ ∗ εk,λ . b + ε b k,λ k,λ k,λ |k|
(I.38)
Let < 1 }. S := { P ∈ R3 : | P| 3
(I.39)
In order to give a mathematically precise meaning to the constructions presented in the following, we introduce an infrared cut-off at a photon frequency σ > 0 in the vector potential. The calculation of the second derivative of the energy of a dressed electron – in the following called the “ground state energy” – as a function of P in the limit where σ → 0, and for P ∈ S, represents the main problem solved in this paper. Hence we will, in the sequel, study the regularized fiber Hamiltonian H Pσ :=
P − P f + α 1/2 Aσ 2
2 + H f,
(I.40)
acting on the fiber space H P , for P ∈ S, where Aσ :=
d 3k ∗ ∗ εk,λ , b + ε b k,λ k,λ k,λ B \Bσ |k|
λ
(I.41)
and where Bσ is a ball of radius σ centered at the origin. In the following, we will consider a sequence of infrared cutoffs σ j := j with 0 < <
1 2
(I.42)
and j ∈ N0 := N ∪ {0}.
Notation. 1) We use the notation AH = A|H for the norm of a bounded operator A acting on a Hilbert space H. Typically, H will be some subspace of F b . 2) Throughout the paper, we follow conventions such that 1 2πi
γ
1 dz = −1 , z
γ
1 d z¯ = ( z¯
γ
1 dz), z
where γ is an integration path in the complex space enclosing the origin.
I.2. Outline of the proof. Next, we outline the key ideas used in the proofs of our main results in Eqs. (I.6), (I.8), and (I.9). For P ∈ S, α small enough, and σ > 0, E σ is an P isolated eigenvalue of H σ |Fσ ; see Sect. II and Eq. (II.4). Because of the analyticity of P it follows that H σ in the variable P, P
Renormalized Electron Mass in Nonrelativistic QED
447
∂ 2 E σ
| P|
= ∂i2 E |σP| i iˆ | P=P 1 1 1 σ σ σ σ ∂i H P dz P , ∂i H P P = 1−2 σ σ 2πi γσ H − z H −z
2 (∂| P|)
P
, (I.43)
P=P i iˆ
P
where ∂i = ∂/∂ P i , iˆ is the unit vector in the direction i, σ is the normalized ground P state eigenvector of H σ constructed in [6]; γσ is a contour path in the complex energy P plane enclosing E σ and no other point of the spectrum of H σ |Fσ , and such that the P P distance of γσ from spec (H σ |Fσ ) is of order σ . P At first glance, the expression on the R.H. S. of (I.43) might become singular as σ → 0, because the spectral gap above E σ = inf spec (H σ |Fσ ) is of order σ . To P P prove that the limit σ → 0 is, in fact, well defined, we make use of a σ -dependent E σ ) (see Sect. II, Eq. (II.3)). This transformation has Bogoliubov transformation, Wσ (∇ P already been employed in [6] to analyze mass shell properties. In fact, conjugation of E σ ) yields an infrared regularized Hamiltonian H σ by Wσ (∇ P
P
E σ )H σ Wσ∗ (∇ Eσ ) K Pσ := Wσ (∇ P P P
(I.44)
with the property that the corresponding ground state, σ , has a non-zero limit, as P σ → 0. The Hamiltonian K σ has a “canonical form” derived in [6] (see also [12], P where a similar operator has been used in the analysis of the Nelson model): K Pσ
=
σ )2 ( P
2
+
λ
R3
σ (k)b ˆ ∗ b d 3k + E σ , |k|δ k,λ P k,λ P
(I.45)
ˆ is defined in Eq. (II.18), E σ is a c-number defined in Eq. (II.42), and σ where δ σ (k) P P P is a vector operator defined in Eq. (II.40) starting from Eqs. (II.16), (II.17), (II.34). By construction, σ σ = 0. σP , P P
(I.46)
This is a crucial property in the proof of existence of a limit of σ as σ → 0. P Equation (I.46) is also an important ingredient in the proof of (I.6), because, by apply E σ ) to each term of the scalar product on the R.H.S. of ing the unitary operator Wσ (∇ P (I.43) and using (I.46) (see Sect. III), one finds that
P
= 1−2
1 2πi
γσ
1 σ σ ∂i E j − ( j )i σj P P γj K − z j P σ σ σ ∂i E j − ( j )i j | P=P i iˆ P P P σ 1 1 σ i P ( dz , ) K σ − z P K σ − z σ P P P
1 (I.43) = 1 − 2 σ j 2 2πi 1
1 σ K j P
− zj
σ P
dz j j ,
σ σ i P ( P ) σ P
(I.47) .
P=P i iˆ
(I.48)
448
J. Fröhlich, A. Pizzo
Notice that, if one starts from the two expressions on the R.H.S. of Eq. (I.43) and on the R.H.S. of Eq. (I.47) respectively, both formally expanded in powers of α 1/2 , the Bogoliubov transformation can be seen as a tool to re-collect an infinite number of terms and to show a nontrivial identity, thanks only to Eq. (I.46) and to a vanishing contour integration (see Eq. (III.46)). Next, still using Eq. (I.46), one can show that (I.48) remains uniformly bounded in σ . To see this we use the inequality 2 1 1 ( σ )i σ , ( σP )i σP ≤ O( 1 ), (I.49) σ P P K −z α 2 σ 2δ P for an arbitrarily small δ > 0, with z ∈ γσ and α small enough depending on δ. This inequality will be proven inductively (see Theorem III.1) by introducing sequences of infrared cut-offs σ j , where σ j → 0 as j → ∞. The proof by induction is combined with an improved (as compared to the result in [6]) estimate of the rate of convergence of { σ } as σ → 0. P By telescoping, one can plug these improved estimates into (I.48) to end up with the desired uniform bound. The control of the rate of convergence of the R.H.S. in (I.48), as for arbitrary infrared cutoff σ > 0, finally σ → 0, combined with the smoothness in P, entails the Hölder-continuity in P of the limiting quantity i σ i σ 1 1 1 σ σ P P P . | P| dz , P := 1− lim 2 σ →0 2πi γσ K σ − z K σ − z σ σ i ˆ P
P
P
P
P=P i
(I.50) The Hölder-continuity in P of | P| and of lim σ →0
∂ E σ
| P|
∂| P|
, combined with the fundamental ∂2 E
theorem of calculus, imply that E P is twice differentiable and | P|2 ≡ | P| . (∂| P|) Our paper is organized as follows: In Sect. II, we recall how to construct the ground states of the Hamiltonians H σ and K σ by iterative analytic perturbation theory. This P P section contains an explicit derivation of the formula of the transformed Hamiltonians and of related algebraic identities that will be used later on. In Sect. III, we first derive inequality (I.49) and the improved convergence rate of { σ } as σ → 0, by using some key ingredients described in Sect. II. Section III.1 is P devoted to an analysis of (I.43) that culminates in the following main results. Theorem. For α small enough and P ∈ S, function | P| := lim σ →0
∂ E σ | P| 2 (∂| P|)
∂ 2 E σ
| P|
2 (∂| P|)
converges as σ → 0. The limiting
is Hölder continuous in P ∈ S. The limit lim | P| =1
α→0
(I.51)
holds true uniformly in P ∈ S. Corollary. For α small enough, the function E P := limσ →0 E σ , P ∈ S, is twice P differentiable, and ∂ 2 E | P|
2 (∂| P|)
= | P| for all P ∈ S.
(I.52)
Renormalized Electron Mass in Nonrelativistic QED
449
Remark. For the complete proof of the construction of the ground states of the Hamiltonians H σ and K σ by iterative analytic perturbation theory, the reader is advised to P P consult ref. [6]. II. Sequences of Ground State Vectors In this section, we report on results contained in [6] concerning the ground states of the σ Hamiltonians H j , where P ∈ S and j ∈ N0 , and the existence of a limiting vector for P
σ P
the sequence of ground state vectors of the transformed Hamiltonians, K j , where the σ
σ
Bogoliubov transformation used to obtain K j from H j (derived in [4]) is determined P P by ∗ bk,λ
→
E σ j )b ∗ Wσ∗ (∇ Eσj ) W σ j (∇ j k,λ P P
∗ bk,λ
=
−α
1 2
E σ j · ∗ ∇ P
k,λ
,
(II.1)
· k,λ , σj ˆ δ (k) P
(II.2)
23 δ σ j (k) ˆ |k| P
σ j Wσ∗ (∇ E σ j ) = b − α 2 bk,λ → Wσ j (∇ E )bk,λ k,λ j
1
P
Eσj ∇ P
P
|k|
3 2
for k ∈ B \Bσ j , with ⎛ σ P
E j ) := exp ⎝α W σ j (∇ where
σ ˆ δ j (k) P
1 2
λ
B \Bσ j
d 3k
Eσj ∇ P
23 δ σ j (k) ˆ |k|
⎞ ∗ ⎠ · (k,λ bk,λ − h.c.) ,
(II.3)
P
is defined in (II.18). σ P
II.1. Ground states of the Hamiltonians H j . In [6], the first step consists in constructσ P
σ P
ing the ground states of the regularized fiber Hamiltonians H j . As shown in [6], H j σ
has a unique ground state, j , that can be constructed by iterative analytic perturbaP tion theory, as developed in [12]. To recall how this method works some preliminary definitions and results are needed: • We introduce the Fock spaces Fσ j := F b (L 2 ((R3 \Bσ j ) × Z2 )) ,
σ
j Fσ j+1 := F b (L 2 ((Bσ j \Bσ j+1 ) × Z2 )). (II.4)
Note that σ
j Fσ j+1 = Fσ j ⊗ Fσ j+1 .
(II.5)
If not specified otherwise, f denotes the vacuum state in any one of these Fock spaces. Any vector φ in Fσ j can be identified with the corresponding vector, φ ⊗ f , σ in F, where f is the vacuum in F0 j .
450
J. Fröhlich, A. Pizzo
Fig. 1. The contour integral in the energy plane and the gaps
• Momentum-slice interaction Hamiltonians are defined by 1 σ σσ jj+1 )2 , σσ jj+1 + α ( A| H σ j · A| H P |σ jj+1 := α 2 ∇ P P 2
(II.6)
where σσ jj+1 := A|
λ
Bσ j \Bσ j+1
d 3k ∗ ∗ εk,λ + ε b bk,λ k,λ . k,λ |k|
(II.7)
• Four real parameters, , ρ + , ρ − , and µ, will appear in our analysis. They have the properties 0 < ρ − < µ < ρ + < 1 − Cα < 0<< where Cα , with
1 3
ρ− , ρ+
2 , 3
(II.8) (II.9)
< Cα < 1, for α small enough, is a constant such that the inequality σ E σP− k > E P − C α |k|
holds for all P ∈ S and any k = 0. Here E σ Cα →
1 3,
P−k
(II.10) := inf specH σ
as α → 0 (see Statement (I4) of Theorem 3.1 in [6]).
P−k
. We note that
By iterative analytic perturbation theory (see [6]), one derives the following results, valid for sufficiently small α, depending on our choice of , , ρ − , µ, and ρ + (see also Fig. 1): σ P
σ P
(A1) E j is an isolated simple eigenvalue of H j |Fσ j with spectral gap larger or equal σ P
σ P
to ρ − σ j . Furthermore, E j is also the ground state energy of H j |Fσ j+1 , and it σ
is an isolated simple eigenvalue of H j |Fσ j+1 with spectral gap larger or equal to P ρ + σ j+1 .
Renormalized Electron Mass in Nonrelativistic QED
451
σ
σ
σ
σ
(A2) The ground-state energies E j and E j+1 of the Hamiltonians H j and H j+1 , P P P P respectively, (acting on the same space Fσ j+1 ) satisfy the inequalities σ P
σ P
0 ≤ E j+1 ≤ E j + c α σ j2 ,
(II.11)
where c > 0 is independent of j and of α. σ σ (A3) The ground state vectors, j+1 , of H j+1 can be recursively constructed starting P P σ0 from ≡ f with the help of the spectral projection P 1 1 dz j+1 σ j+1 . 2πi γ j+1 H − z j+1 P
More precisely, σ j+1 P
1 := 2πi =
γ j+1
dz j+1
1 σ H j+1 P
− z j+1
σ P
j
∞ 1
1 dz j+1 σ j 2πi H − z j+1 n=0 γ j+1 P n 1 σj σ × −H P |σ j+1 σ j j , P H − z j+1
(II.12)
(II.13)
P
where γ j+1 := {z j+1 ∈ C | |z j+1 −
σ E j | P
σ P
= µσ j+1 }, with µ as in (II.8). j+1 is σ P
the (unnormalized) ground state vector of H j+1 |Fσ for any 0 ≤ σ ≤ σ j+1 .
II.2. Transformed Hamiltonians. In this section, we consider the (Bogoliubov-transformed) Hamiltonians σ P
σ P
σ P
σ P
E j )H j Wσ∗ (∇ E j) K j := Wσ j (∇ j
(II.14)
σ P
with ground state vectors j , j = 0, 1, 2, 3, . . .. Some algebraic manipulations to σ
express K j in a “canonical form” appear to represent a crucial step before iterative P perturbation theory can be applied to the sequence of these transformed Hamiltonians. σ In addition, some intermediate Hamiltonians, denoted Kˆ j , must be introduced to arrive P at the right kind of convergence estimates. σ The same algebraic relations that are used to obtain the “canonical form” of K j also P play an important role in the proof of our main result concerning the limiting behavior, as σ → 0, of the second derivative of the ground state energy E σ . It is therefore useful to P
σ P
derive the “canonical form” of K j and the relevant algebraic identities in some detail. σ P
The Feynman-Hellman formula (which holds because (H j ) P∈ S is an analytic famσ
σ
ily of type A, and E j is an isolated eigenvalue of H j |Fσ , 0 < σ ≤ σ j ) yields the P P identity E σ j = P − P f − α 2 Aσ j σ j , ∇ ψ 1
P
P
(II.15)
452
J. Fröhlich, A. Pizzo
where, given an operator B and a vector ψ in the domain of B, we use the notation ψ, B ψ . ψ, ψ
Bψ :=
(II.16)
As stated in [6], for α small enough, E σ j | < 1 ∀ j ∈ N0 . sup |∇ P
S P∈
We define βσ j := P f − α 2 Aσ j , 1
(II.17)
σ ˆ := 1 − kˆ · ∇ E σ j | < 1, E σ j , kˆ := k , |∇ δ j (k) P P P |k| σj E · ∗ ∇ 1 P k,λ ∗ ∗ 2 := b + α ck,λ , 3 σj k,λ ˆ |k| 2 δ (k)
(II.18)
(II.19)
P
σ j 1 ∇E · k,λ
ck,λ := bk,λ + α2
P
σ P
2 δ j (k) ˆ |k| 3
.
(II.20)
σ P
We rewrite H j as ( P − βσ j )2 + H f, 2
σ P
H j =
(II.21)
and, using (II.15) and (II.17), E σ j + βσ j σ j . P = ∇ ψ P
(II.22)
P
We then obtain σ P
σ j 2 P 2 E σ j + βσ j σ j ) · βσ j + β − (∇ +Hf ψ P 2 2 P 2 P 2 βσ j = + − βσ j ψ σ j · βσ j 2 2 P
σ j (k)b ˆ ∗ b d 3k + |k|δ k,λ k,λ
H j =
+
B \Bσ j
λ
−α
(II.25)
σ P
j (k)c ˆ ∗ c d 3k |k|δ k,λ k,λ
(II.26) σ
λ
(II.24)
P
R3 \(B \Bσ j )
λ
(II.23)
B \Bσ j
σ j (k) ˆ |k|δ P
E j · ∗ E σ j · ∇ ∇ k,λ P k,λ P
σ P
σ P
2 δ j (k) ˆ |k| 2 δ j (k) ˆ |k| 3
3
d 3 k.
(II.27)
Renormalized Electron Mass in Nonrelativistic QED
453
Adding and subtracting 1/2 βσ j 2 σ j , one finds that ψ P
σ H j P
βσ j 2 σ j (βσ j − βσ j σ j )2 ψ ψ P 2 P P = − + 2 2 2
σ j (k)b ˆ ∗ b d 3k |k|δ + k,λ k,λ +
λ
R3 \(B \Bσ j )
λ
B \Bσ j
−α
(II.29)
P
σ P
j (k)c ˆ ∗ c d 3k |k|δ k,λ k,λ
B \Bσ j
λ
(II.28)
(II.30)
E σ j · ∇ k,λ P σ j (k) ˆ |k|δ 3 σj P ˆ |k| 2 δ (k) P
E σ j · ∗ ∇ P
k,λ 3
23 δ σ j (k) ˆ |k|
d k.
(II.31)
P
Next, we implement the Bogoliubov transformation ∗ bk,λ
→
E σ j )b ∗ Wσ∗ (∇ Eσj ) W σ j (∇ j k,λ P P
=
∗ bk,λ
−α
1 2
E σ j · ∗ ∇ P
k,λ
23 δ σ j (k) ˆ |k|
,
(II.32)
,
(II.33)
P
σ j 1 ∇E · k,λ
σ j Wσ∗ (∇ E σ j ) = b − α 2 bk,λ → Wσ j (∇ E )bk,λ k,λ j P
P
P
σ P
2 δ j (k) ˆ |k| 3
E σ j ) is defined in (II.3). It is evident that Wσ j acts as the for k ∈ B \Bσ j , where Wσ j (∇ P identity on F b (L 2 (Bσ j × Z2 )) and on F b (L 2 ((R3 \B ) × Z2 )). We define the vector operators E σ j )βσ j Wσ∗ (∇ Eσj ) σ j := Wσ j (∇ j P
P
P
E σ j )βσ j Wσ∗ (∇ E σ j ) f , −Wσ j (∇ j P
(II.34)
P
noting that, by (II.15), (II.17), and (II.34), Eσj βσ j ψ σ j = P − ∇ P
=
P σj σ j σ j , P P P σ σ j , j P P
(II.35) σ P
σ P
E j )βσ j Wσ∗ (∇ E j ) f , + Wσ j (∇ j
(II.36)
σ P
where j is the ground state of the Bogoliubov-transformed Hamiltonian σ E σ j )H σ j Wσ∗ (∇ E σ j ). K j := Wσ j (∇ j P
P
P
(II.37)
P
σ
σ
Notice that in (II.36) only the ray of j enters. The sequence of vectors { j } is defined P P in Section II.3. It is easy to see that E σ j )βσ j Wσ∗ (∇ E σ j ) − βσ j σ j = σ j − σj σj . W σ j (∇ j P
P
P
P
P
P
(II.38)
454
J. Fröhlich, A. Pizzo
After inspecting straightforward operator domain questions (see [6]), the “canonical σ form” of K j is given by P
σ K j P
=
σ P
j )2 ( 2
+
σ j (k)b ˆ ∗ b d 3k + E σ j , |k|δ k,λ k,λ P
R3
λ
(II.39)
P
where σ j − σj σj , σ j := P
P
P
(II.40)
P
so that σ P
j σ j = 0,
(II.41)
P
and σ E j P
E σ j )2 ( P − ∇ P 2 P − := 2 2 E σ j · ∗ E σ j · ∇
∇ σj k,λ P k,λ P ˆ −α |k|δ (k) d 3 k. 3 σj 3 σj P ˆ ˆ 2 δ (k) 2 δ (k) B \ B | k| | k| σ
j λ P
(II.42)
P
Equation (II.39) follows by using that E σ j )c ∗ Wσ∗ (∇ Eσj ) = b ∗ , W σ j (∇ j k,λ k,λ
(II.43)
= bk,λ ,
(II.44)
P P E σ j )c Wσ∗ (∇ Eσj ) W σ j (∇ k,λ j P P
for k ∈ B \Bσ j . σ An intermediate Hamiltonian, Kˆ j+1 , is defined by P
σ Kˆ j+1 P
where
E σ j )H σ j+1 Wσ∗ (∇ E σ j ), := Wσ j+1 (∇ j+1 P
⎛ σ P
E j ) := exp ⎝α Wσ j+1 (∇
1 2
λ
B \Bσ j+1
d 3k
P
(II.45)
P
Eσj ∇ P
23 δ σ j (k) ˆ |k|
⎞ ∗ ⎠ · (k,λ bk,λ − h.c.) . (II.46)
P
σ σ We decompose Kˆ j+1 into several different terms, similarly as K j . We recall that P
P
σ P
H j+1 =
( P − βσ j+1 )2 + H f, 2
(II.47)
and, by (II.35), E σ j + βσ j σ j . P = ∇ ψ P
P
(II.48)
Renormalized Electron Mass in Nonrelativistic QED
455
It follows that (see also (II.28)–(II.31)) σ P
σ j+1 2 P 2 E σ j + βσ j σ j ) · βσ j+1 + β − (∇ +Hf ψ P 2 2 P 2 P 2 βσ j+1 = + − βσ j ψ σ j · βσ j+1 2 2 P
σ j (k)b ˆ ∗ b d 3k + |k|δ k,λ k,λ
H j+1 =
λ
+
−α
(II.51)
σ j (k)c ˆ ∗ c d 3k |k|δ k,λ k,λ
(II.52)
P
B \Bσ j+1
λ
(II.50)
P
R3 \(B \Bσ j+1 ) B \Bσ j+1
λ
(II.49)
E σ j · ∇ σj k,λ P ˆ |k|δ (k) 3 σj P ˆ 2 |k| δ (k) P
E σ j · ∗ ∇ P
k,λ 3
23 δ σ j (k) ˆ |k|
d k.
(II.53)
P
We now add and subtract 1/2 βσ j 2 σ j and conjugate the resulting operator with the uniψ P σj tary operator Wσ j+1 (∇ E ). After inspecting straightforward operator domain questions P (see [6]), we find that
j σ j + Lσσ jj+1 + Iσσj+1 ( )2
σ Kˆ j+1 =
P
P
+
λ
(II.54)
2
R3
σ j (k)b ˆ ∗ b d 3 k + Eˆ σ j+1 , |k|δ k,λ k,λ P
(II.55)
P
where 1 σ Lσ jj+1 := −α 2
Bσ j \Bσ j+1
λ
σσ jj+1 −α 2 A|
:= α
k
E σ j · ∗ b + h.c. ∇ k,λ P
k,λ σ P
2 δ j (k) ˆ |k| 3
d 3 k,
(II.56)
1
σj Iσ j+1
Bσ j \Bσ j+1
λ
+α
k
Bσ j \Bσ j+1
λ σ Eˆ j+1 P
(II.57) Eσj ∇ P
σj ∗ · k,λ ∇ E P · k,λ
23 δ σ j (k) ˆ |k| 23 δ σ j (k) ˆ |k| P
[k,λ
E σ j · ∗ ∇ P
|k|
3 2
k,λ σj ˆ δ (k) P
d 3k
P
d 3k + h.c.] , |k|
(II.58)
E σ j )2 ( P − ∇ P 2 P − := 2 2 E σ j · ∗ E σ j · ∇
∇ σj k,λ P k,λ P ˆ −α |k|δ (k) d 3 k. 3 σj 3 σj P ˆ ˆ 2 2 B \ B |k| δ (k) |k| δ (k) σ j+1
λ P
(II.59)
P
We also define the operators σj E σ j−1 )Wσ∗ (∇ E σ j ) E σ j )Wσ∗ (∇ E σ j−1 ), ˆ P := Wσ j (∇ σ j W σ j (∇ j j P
P
P
P
P
(II.60)
456
J. Fröhlich, A. Pizzo
and σj σj σj ˆ P := ˆ P − ˆ P σ j , ˆ
(II.61)
P
ˆ σj which are used in the proofs of convergence of the ground state vectors. Here, P denotes the ground state vector of the Hamiltonian σ E σ j−1 )Wσ∗ (∇ E σ j )K σ j Wσ j (∇ E σ j )Wσ∗ (∇ E σ j−1 ). Kˆ j := Wσ j (∇ j j P
P
(The sequence of vectors Notice that
P
ˆ σj } { P
P
P
P
is defined in (II.64).)
σj E σ j−1 )Wσ∗ (∇ E σ j ) E σ j )Wσ∗ (∇ E σ j−1 ). ˆ P = Wσ j (∇ σ j W σ j (∇ j j P
P
P
P
P
(II.62)
An important identity used in [6] and in the sequel of the present paper is ( j ≥ 1) σj Eσj − ∇ E σ j−1 + Lσσ j−1 ˆ P − σ j−1 = ∇ j P
P
+α
λ
+α
P
Bσ j−1 \Bσ j
λ
Bσ j−1 \Bσ j
k
σ
E j−1 · ∗ E σ j−1 · ∇ ∇ k,λ P k,λ P
σ P σ ∇ E j−1 · ∗ P k,λ [k,λ 3 σ j−1 ˆ |k| 2 δ (k) P
σ P
2 δ j−1 (k) ˆ |k| 2 δ j−1 (k) ˆ |k| 3
3
d 3k
d 3k + h.c.] . |k|
(II.63)
Equation (II.63) can be derived using (II.34), (II.36), (II.38), (II.40), (II.60), and (II.61). σ
II.3. Convergence of the sequence { j }∞ . We start from σ0 ≡ f . To pass from P j=0 P momentum scale j to j + 1, we proceed in two steps: First, we construct an intermediate ˆ σ j+1 , defined by vector, P
ˆ σ j+1 := P
∞
1 1 1 σ σ dz j+1 σ j [−K P |σ jj+1 σ j ]n j , P 2πi γ j+1 K − z j+1 K − z j+1 n=0 P
(II.64)
P
where σ σ σ σ σ K P |σ jj+1 := Kˆ j+1 − Eˆ j+1 + E j − K j P P P P 1 σ 1 σj σj 2 j · (Lσσ jj+1 + Iσσj+1 ) + h.c. + (Lσ jj+1 + Iσ j+1 ) . = P 2 2
(II.65) (II.66)
σ P
Subsequently, we construct j+1 by setting σ E σ j+1 )Wσ∗ (∇ E σ j ) ˆ σ j+1 . j+1 := Wσ j+1 (∇ j+1 P
P
P
P
(II.67)
The series in (II.64) is term-wise well-defined and converges strongly to a non-zero vector, provided α is small enough (independently of j). The proof of this claim is based on operator-norm estimates of the type used in controlling the Neumann expansion in (II.13), which requires an estimate of the spectral gap that follows from the unitarity of E σ j ) and Result (A1) described after Eq. (II.10). W σ j (∇ P
Renormalized Electron Mass in Nonrelativistic QED
457 σ
A key point in our proof of convergence of the sequence { j } is to show that the P term j σ j · (Lσσ jj+1 + Iσσj+1 ) + h.c.
(II.68)
P
appearing in (II.66), which is superficially “marginal” in the infrared, by power counting, is in fact “irrelevant” (using the terminology of renormalization group theory). This is a consequence of the orthogonality condition σ σ j σ j = 0, j , P
P
(II.69)
P
which, when combined with an inductive argument, implies that (
1 σ K j P
− z j+1
σ P
σ (+)
σ
j j · (Lσ jj+1 + Iσ j+1 ) 2 [ )] ( 1
1 σ K j P
− z j+1
1
σ P
) 2 j
(II.70)
σ (+)
(where Lσ jj+1 stands for the part which contains only photon creation operators) is of order O( η j ), for some η > 0 specified in [6]. In particular, this suffices to show that σ P
σ P
ˆ j+1 − j ≤ O(
j+1 2 (1−δ)
)
(II.71)
for any 0 < δ < 1 provided α is sufficiently small. Finally, in Theorem 3.1 of ref. [6], it is proven that there is a non-zero vector in the Hilbert space corresponding to 1 σ lim j→∞ j , and that the rate of convergence is at least O(σ 2 (1−δ) ) for any 0 < δ < 1 P provided α is sufficiently small. σ P
Remark. In Theorem 3.1 of ref. [6], for lim j→∞ j , the range of values of α such 1
that the rate of convergence, O(σ 2 (1−δ) ), holds is not claimed to be uniform in δ. The stronger result obtained in the next section (see (III.3) and (III.34)) implies that this 1 range (corresponding to the rate O(σ 2 (1−δ) )) is actually δ-independent. σ P
II.3.1. Key ingredients. To prove convergence of the sequence { j } of ground state σ K j , P
some further conditions on α, , and µ (see (II.8), vectors of the Hamiltonians (II.9)) are required, in particular an upper bound on µ and an upper bound on strictly smaller than the ones imposed by inequalities (II.8), (II.9) (for details, see Lemma A.3 in [6]). We note that the more restrictive conditions on µ and imply new bounds on 1 ρ − and ρ + . Moreover, must satisfy a lower bound > Cα 2 , with a multiplicative constant C > 0 sufficiently large. σ Some key inequalities needed in our analysis of the convergence properties of { j } P are summarized below. They will be marked by the symbol (B). In order to reach some σj important improvements in our estimates of the convergence rate of , as j → ∞ P (discussed in the next section), a refined estimate is needed that is stated in (B2), and a new inequality, see (B5), (analogous to (B3) and (B4)) is required. • Estimates on the shift of the ground state energy and its gradient. There are constants C1 , C2 such that the following inequalities hold: (B1) σ P
σ P
|E j − E j+1 | ≤ C1 α j , see [6].
(II.72)
458
J. Fröhlich, A. Pizzo
• (B2) σ 1 E σ j | ≤ C2 E σ j+1 − ∇ ˆ j+1 − σ j + α 4 j+1 . |∇ P
P
P
P
(II.73)
This is an improvement over a corresponding estimate in [6]: It can be proven after the results stated in Theorem 3.1 in [6], in particular the uniform bound from below σ σ σ σ on j , j , j , j > 23 , and following the steps in the proof of Lemma A.2 P P P P in [6]. • Bounds relating expectations of operators to expectations of their absolute values. There are constants C3 , C4 , C5 > 1 such that the following inequalities hold: (B3) For z j+1 ∈ γ j+1 , σ ( j )i P
σ j , P
1 σj i σj ( ) P σj K − z j+1 P
(II.74)
P 1 σj i σj σj i σj ( ) , ≤ C3 ( ) , σ j P P P K − z j+1 P
(II.75)
P
σ
σ
j. where ( j )i is the i th component of P P (B4) For z j+1 ∈ γ j+1 , σ (+) (Lσ jj+1 )l
σ ( j )i P
σ j , P
1 σ j (+) l σ j i σ j ) ( ) (L σj P P K − z j+1 σ j+1
P 1 σ j (+) l σj i σj σ j (+) l σj i σj (Lσ j+1 ) ( ) , ≤ C4 (Lσ j+1 ) ( ) , σ j P P P P K − z j+1
(II.76)
(II.77)
P
σ (+)
σ (+)
where (Lσ jj+1 )l is the l th component of Lσ jj+1 . (B5) For z j+1 ∈ γ j+1 , σ ( j )i P
σ j , P
2 1 σj i σj ( P ) P σj K − z j+1
P 1 σj i σj σ σ )2 ( j )i j . ≤ C5 ( ) , ( σ j P P P P K − z j+1
(II.78)
P
To prove (B3) and (B4), it suffices to exploit the fact that the spectral support (with σ σ σ σ respect to K j |Fσ j and to K j |Fσ j+1 , respectively) of the two vectors ( j )i j and
P σ j (+) l σj i (Lσ j+1 ) ( ) P
σ j P
P
is strictly above the ground state energy of σ
P P σj K , since they are both P
orthogonal to the ground state, j , of this operator. In the proof of bound (B5), it is P also required that ρ − > 3µ, as will be assumed in the following. Remarks. (1) The constants C1 , . . . , C5 are independent of α, , µ, and j ∈ N, provided that α, , and µ are sufficiently small.
Renormalized Electron Mass in Nonrelativistic QED
459
(2) For the convenience of the reader, we recapitulate the relations between the parameters entering the construction: 0 < ρ − < µ < ρ + < 1 − Cα < 0<<
ρ− , ρ+
2 , 3
(II.79) (II.80)
> Cα 1/2 , ρ − > 3µ.
(II.81) (II.82)
Moreover, we stress that the final result is a small coupling result, i.e., valid for small values of α, and that, for technical reasons, small values of the parameters , µ within the constraints listed above (that imply more restrictive bounds on ρ − , ρ + ) are required. σ P
σ P
ˆ j+1 − j obtained in [6] (see (II.71)) is The crucial estimate for the bound on R0 1 σj i σj σj i σj ( ) ≤ , ( P ) P , σ j jδ P P α K − z j+1
(II.83)
P
where R0 is independent of j, and δ, 0 < δ < 1, can be taken arbitrarily small for α and sufficiently small (depending on δ). This estimate will be improved in the next section. σ As a consequence, our estimate of the convergence rate of { j } will be improved. As P a corollary, the second derivative of E σ is proven to converge, as σ → 0. P
III. Improved Estimate of the Convergence Rate of {σ }, as σ →0, P and Uniform Bound on the Second Derivative of E σ . P
Our arguments in Sect. III rely on the results previously proven in [6] and described in Sect. II, which hold for α small enough. Therefore, in the following, we assume the constraints (II.79)–(II.82), and we make use of the estimates on the spectral gaps (see (A1) in Sect. II.1) and of the bounds (B1)–(B5) (see Sect. II.3.1). We also make use of the lower bounds ˆ σ j+1 , ˆ σ j+1 , σ j , σ j > P
P
P
P
2 3
(III.1)
uniformly in j ∈ N0 , which appear in the proof of Theorem 3.1 of ref. [6]. Assuming these bounds we can simplify the proof by induction in the theorem below. 1
Theorem III.1. For α, sufficiently small (depending on δ), > Cα 16 with C > 0, the inequality 2 R0 1 σ σ σ σ j j j j i ( )i , ( ) (III.2) σ ≤ 1 2 jδ j P P P P K − z j+1 α2 P
460
J. Fröhlich, A. Pizzo 3
holds true, where 0 < δ < 1 and R0 , O( 2 ) > R0 > O(α 8 ), is independent of j ∈ N0 := N ∪ 0, and σ P
σ P
ˆ j − j−1 ≤ α 4 j (1−δ) . 1
(III.3)
R0 can be taken arbitrarily small provided α is small enough. Proof by induction. • Inductive hypothesis. We assume that, at scale j − 1(≥ 0), the following estimate holds: 2 R0 1 σ σ ( σ j−1 )i σ j−1 , ( j−1 )i j−1 ≤ 1 . (III.4) σ j−1 P P P P K − zj 2 2( j−1)δ α P This estimate readily implies that, assuming R0 / 2 and α small enough, uniformly in j, σ
σ
ˆ j − j−1 (III.5) P ∞ P n 1 1 1 σ σ j−1 dz j σ j−1 = −K P |σ j−1 σ j P 2πi γ j K − zj K j−1 − z j n=1 P
P
1
≤ α 4 j (1−δ) .
(III.6)
An improved estimate on > i)
1 Cα 16 :
ˆ σj P
−
σ j−1 P
is based on the following bounds where
1 1 1 σ j−1 σ j−1 K P |σ j ≤ O(R02 α 4 ( j−1)(1−δ) + α 2 j−3 ), σ j−1 P K − zj
(III.7)
P
whose proof requires the use of the “pull-through formula” (see, e.g., [14]), a Neumann expansion of the resolvent, the inequality in Eq. (II.78), and Eq. (III.4); the reader can follow the similar steps used in Lemma A3 of ref. [6]; ii)
1 σ j−1 K P |σ j σ j−1 K − zj P
1
≤ O(α 2 /) ;
(III.8)
Fσ j
this estimate can be derived from standard bounds and using the “pull-through formula”. E σ j−1 )Wσ∗ (∇ E σ j ), • Induction step from scale j − 1 to scale j. By unitarity of Wσ j (∇ j P P we have that 2 1 σ σ σ σ j j j j ( )i , ( )i σj P P P P K − z j+1 P 2 1 σ σ σ σ j j j j . ˆ , ˆ ˆ )i = (ˆ )i (III.9) ( σ P P P P Kˆ j − z j+1 P
Renormalized Electron Mass in Nonrelativistic QED
461
1
For α small enough and > C α 2 , where C > 0 is large enough, we may use (B1) to re-expand the resolvent and find that 2 1 σ σ σ σ j j j j (ˆ )i ˆ , ˆ (III.10) (ˆ )i σj P P P P ˆ K − z j+1 P 2 1 σ σ σ σ j j j j . ˆ , σ ˆ ≤ 2 (ˆ )i (III.11) (ˆ P )i P P K j−1 − z P j+1 P It follows that 2 1 σ ˆ σ j i ˆ σ j ˆ σ j , σ 2 (ˆ j )i ( P ) P P P K j−1 − z j+1 P 2 1 σ j−1 i σ j−1 ˆ σj i ˆ σj ≤ 4 σ j−1 (( P ) P − ( P ) P ) K − z j+1 P 2 1 σ σ σ σ + 4 ( j−1 )i j−1 , σ j−1 ( j−1 )i j−1 . P P P P K − z j+1 P
(III.12)
(III.13)
(III.14)
Our recursion, combined with (II.78), relates (III.14) to the initial expression in (III.2), with j replaced by j − 1, while (III.13) is a remainder term. Next we note that 2 1 σ j−1 i σ j−1 ˆ σj i ˆ σj 4 σ j−1 (III.15) (( P ) P − ( P ) P ) K − z j+1 P 2 1 σ j−1 i σ j ˆ σj i ˆ σj ˆ ) (III.16) ≤ 8 σ j−1 (( P ) P − ( P ) P K − z j+1 P 2 1 σ j−1 σ j−1 i ˆ σ j (III.17) + 8 σ j−1 ( P ) ( P − P ) K − z j+1 P
R1 R2 ≤ 2 jδ + 2 jδ .
(III.18)
Here R1 ≤ O( −2 ) and R2 ≤ O( −2 ) are constants independent of α, µ, and j ∈ N, 1 provided that α is sufficiently small and > Cα 16 . In detail: – Property (B4) and the two norm-bounds
1 σ K j−1 P
− z j+1
σ P
( j−1 )i Fσ j ≤ O( −( j+1) ),
σ P
σ P
ˆ j − j−1 ≤ α 4 j (1−δ) 1
(III.19) (see (III.5)) justify the step from (III.17) to (III.18);
462
J. Fröhlich, A. Pizzo
– Concerning the step from (III.16) to (III.18), it is enough to consider Eq. (II.63) and the two bounds σ ˆ σ j ≤ O(α 2 j−1 ), ˆ σ j − σ j−1 ≤ α 4 j (1−δ) . (III.20) (Lσ j−1 )i j 1
1
P
P
P
(Hint: for the first inequality in (III.20), use the expression in (II.64).) To bound the term (III.14), we use (B5) and the key orthogonality property (II.69). For z j ∈ γ j and z j+1 ∈ γ j+1 , we find that for /ρ − sufficiently small, 2 1 σ σ σ j−1 σ j−1 4 ( j−1 )i j−1 , σ j−1 ( P )i P P P K − z j+1 P 2 1 σ σ σ σ ≤ 4C5 ( j−1 )i j−1 , ( j−1 )i j−1 σ j−1 P P P P K − z j+1 P 2 1 σ j−1 i σ j−1 σ j−1 i σ j−1 2 ≤ 8C5 ( ) , ( ) σ . P P P P K j−1 − z j
(III.21)
(III.22)
(III.23)
P
In passing from (III.22) to (III.23), we again use the constraint on the spectral support σ σ σ (with respect to K j−1 |Fσ j−1 ) of the vector ( j−1 )i j−1 . P P P Assuming that the parameters and α are so small that the previous constraints are fulfilled and that 0 < R1 + R2 ≤ (1 − 8C52 2δ )
R0 1
α2
,
we then conclude that 2 1 σ σ σ σ ( j )i j , ( j )i j σj P P P P Kˆ − z j+1 P ≤
≤
R1 R2 + 2 jδ 2 jδ 2 1 σ σ σ σ ( j−1 )i j−1 + 8C52 ( j−1 )i j−1 , σ j−1 P P P P K − zj P R0 1 2
α 2 jδ
.
(III.24)
(III.25) (III.26) (III.27) (III.28)
Notice that the bound in (III.24) induces a δ−dependent constraint on the admissible values of and, indirectly, on α. Moreover, the bound in (III.24) is fulfilled if 3
α
1
(O( 2 ) >)R0 > O(α 8 ) ≥ O( 22 ) as α → 0. • The zeroth step in the induction. Since ( σ0 )i ≡ ( P f )i , σ0 ≡ f , P
P
(III.29)
Renormalized Electron Mass in Nonrelativistic QED
463
inequality (III.2) is trivially fulfilled for j = 0; thus (III.2) holds for all j ∈ N0 and for R0 arbitrarily small, provided α is small enough. As we explain below, an improved estimate of the rate of convergence of the sequence σ follows from the bound in (III.3), but we stress that only the estimates in { j }∞ P j=0 Eqs. (III.2), (III.3) will be used for the uniform bound on the second derivative of E σ P in the next section. In fact, one can combine the bound in Eq. (III.3) with the estimate σ P
σ P
σ P
σ P
E j−1 − ∇ E j | | ln( j )|, j ≤ C α 2 |∇ j − 1
(III.30)
where C is independent of α, , µ, and j ∈ N, provided that α, , and µ are sufficiently small. The estimate in (III.30) is obtained starting from the definition in Eq. (II.67) and using the soft photon bound σ P
j 1/2 bk,λ ≤ Cα
1σ j , (k) 3/2 |k|
(III.31)
is the characteristic function of the set {k : σ j < |k| ≤ }) that follows (where 1σ j , (k) from inequality (II.10) and the identity σ P
1
j bk,λ = −α2
1σ j , (k) |k|
σ H j P−k
1 2
1 σj σj k,λ · ∇ P H P P , σj ε + |k| − E
(III.32)
P
which is derived in [4] by using a “pull-through argument”. By a standard procedure (see, e.g., [12]), one obtains similar results for the ground state vectors of the σ -dependent Hamiltonians K σ , for arbitrary σ > 0. A precise stateP ment concerning the rate of convergence is as follows: The normalized ground state vectors (that, with an abuse of notation, we call σ ) P
σP where γσ := {z ∈ C | |z − with
E σ | P
:= =
1 2πi 1 2πi
ρ− 2
γσ
dz K σ1−z f
γσ
P
1 K σ −z f
,
(III.33)
P
σ }, converge strongly to a vector P , as σ → 0, 1
σP − P ≤ O(α 4
σ 1−δ
)
(III.34)
for any 0 < δ(< 1), provided α is in an interval (0, αδ ), where αδ > 0, αδ → 0 as δ → 0. The δ-dependence of the interval αδ is an indirect consequence of the upper bound on that must be imposed through (III.24) to implement the proof by induc1
tion. The relations (II.79)-(II.82) and > Cα 16 induce a δ-dependence on the other parameters and in particular on α. Another δ-type dependence of the estimated rate of ! σ "1−δ convergence, O(
), comes from the logarithmic term in (III.30). However, this does not spoil the uniformity in δ of the interval of admissible values of α but only affects the multiplicative constant on the R.H.S. of (III.34). Moreover, with further work, the estimate in Eq. (III.30) can be improved to remove the logarithmic term.
464
J. Fröhlich, A. Pizzo
III.1. Convergence of the second derivative of the ground state energy E σ . Because of P is an analytic rotational symmetry we have that E σ ≡ E σ . Moreover, (H σ |Fσ ) P∈ S P | P| P family of type A in P ∈ S, with an isolated eigenvalue E σ . Thus, the second derivative ∂ 2 E σ
| P| 2 (∂| P|)
| P|
is well defined and ∂ 2 E σ
| P|
= ∂i2 E |σP| i = 1, 2, 3, i iˆ , | P=P
2 (∂| P|)
(III.35)
where ∂i := ∂ ∂P i . Without loss of generality, the following results are proven for the standard sequence (σ j )∞ j=0 of infrared cutoffs. By simple arguments (see [12]), limiting behavior as σ → 0 is shown to be “sequence-independent”. By analytic perturbation theory we have that ∂i2 E
σj i iˆ | P=P | P|
=1−2 ×
1 2πi
1 σ P
H j − z j
1
γj
σ P
H j − z j
σ dz j j , P
P i − (β σ j )i σj i
P − (β ) i
σ P
σ j P
,
(III.36)
i iˆ P=P
σ P
here j is the normalized ground state eigenvector of H j .
σ
E j) Next, we make use of the Bogoliubov transformation implemented by Wσ j (∇ P to show that 1 1 1 σ σ [P i − (β σ j )i ] σ j dz j j , [P i − (β σ j )i ] j P P 2πi γ j H σ j − z j H − zj P
=
1 σ j 2 P
×
P
1 2πi
1 σ P
K j − z j
(III.37)
1 γj
σ K j P
σ dz j j , P
− zj
E σ j )(β σ j )i Wσ∗ (∇ Eσj ) P i − W σ j (∇ j P
P
σ # σj σj i ∗ σj P − Wσ j (∇ E )(β ) Wσ j (∇ E ) j , (III.38) i
P
σ
P
P
σ
where j is the ground state eigenvector of K j (iteratively constructed in Sect. II). P P Recalling the definitions E σ j )βσ j Wσ∗ (∇ E σ j ) − Wσ j (∇ E σ j )βσ j Wσ∗ (∇ E σ j ) f , σ j := Wσ j (∇ j j P
P
P
P
P
σ j := σ j − σj σj , P
P
P
(III.39) (III.40)
P
and because of the identity (Feynman-Hellman, see (II.36)) Eσj βσ j ψ σ j = P − ∇ P
(III.41)
P
σ P
σ P
σ P
E j )βσ j Wσ∗ (∇ E j ) f , j σ j + Wσ j (∇ = j P
(III.42)
Renormalized Electron Mass in Nonrelativistic QED
465
we find that σ P
σ P
σ P
σ P
E j )(β σ j )i Wσ∗ (∇ E j ) = −( j )i + ∂i E j ; P i − W σ j (∇ j
(III.43)
hence, ∂i2 E
σj i iˆ | P=P | P|
1
1
(III.44)
1
σ σ j 2 2πi γ j K j − z j P P σ σ σ [∂i E j − ( j )i ] j | P=P i iˆ . P P P
=1−2
σ P
σ P
[∂i E j − ( j )i ]
1 σ K j P
− zj
σ P
dz j j , (III.45)
Using the eigenvalue equation σ P
σ P
σ P
σ P
K j j = E j j , σ P
the terms proportional to (∂i E j )2 and to the mixed terms – i.e., proportional to the σ
σ
product of ∂i E j and ( j )i – are seen to be identically 0, because the contour integral P P vanishes for each i = 1, 2, 3, e.g., 1 1 σ σ σ σ σj [∂i E j ] σ j j , [∂i E j ] j d z¯ j P P P K − z j P γj K − z j P P 2 σ ∂i E j σj σj P = , d z¯ j = 0. (III.46) σ P P E j − z¯j γj P
It follows that ∂i2 E
σj i iˆ | P=P | P|
=1+
=1+
1 πi 1 πi
(III.47)
γj
d z¯ j
γj
d z¯ j
σ
σ
P
P
j j 1 1 σj i σ j i P P ( ) ) , ( iˆ σ | P K σ − z j P K σ − z j σ j j P=P i P P
1 σ E j P
− z¯j
σ P
( j )i
1 σ K j P
−
σj σ j i P ( ) σ , P zj j P
(III.48)
σ j P i ˆ. σ | j P=P i P
(III.49) We are now ready for the key estimate. Lemma III.2. The estimate below holds true ( j ∈ N): σ j−1 σ j−1 1 1 σ j−1 i P σ j−1 i P ( ) d z¯ j−1 , σ ( ) σ σ σ P γ j−1 j−1 K j−1 − z¯ j−1 P j−1 E j−1 − z¯ j−1 P P P P σj σj 1 1 σj i σj i P P − ( ) d z¯ j ≤ j (1−2δ) , , σj ( ) σ σ σ j j j P P K − z¯ j E − z¯j γj P
P
P
P
(III.50) for any 0 < δ(< 1/2), and for α and (>
1 Cα 16 )
small enough depending on δ.
466
J. Fröhlich, A. Pizzo σ P
σ P
E j−1 )Wσ∗ (∇ E j ), Proof. By unitarity of Wσ j (∇ j γj
σ ( j )i P
σ P σ , j P
j
1 σ P
K j − z¯ j
ˆ σj 1 σ j i P = (ˆ ) σj , σj P ˆ Kˆ − γj P
σ 1 P d z¯ j σj σj E − z¯ j P P ˆ σj 1 σ j i P (ˆ ) σj σj P ˆ E − z¯ j z¯ j P P
σ ( j )i P
P
j
(III.51)
d z¯ j .
(III.52)
By assumption, α is so small that the Neumann series expansions of the resolvents below converge in Fσbj : 1 1 σ = σ j−1 + 1∞ (K j−1 , z¯ j ), σj P ˆ K − z¯ j K − z¯ j
(III.53)
P
P
1 σ E j P
− z¯ j
=
1 σ E j−1 P
− z¯j
σ P
+ (E j−1 , z¯j ),
(III.54)
where σ 1∞ (K j−1 , z¯ j ) P
:=
∞
1
σ K j−1 l=1 P
− z¯ j
σ σ σ [−(K P |σ j−1 + Eˆ j − E j−1 ) j P
P
1 σ K j−1 P
− z¯ j
]l , (III.55)
σ
is defined in Eq. (II.66), and K P |σ j−1 j σ P
(E j−1 , z¯ j ) :=
1 σ E j P
− z¯ j
σ P
σ P
(E j−1 − E j )
1 σ E j−1 P
− z¯ j
.
(III.56)
We proceed by using the obvious identity,
σ (ˆ j )i P
ˆ σj
1
σ (ˆ j )i P
ˆ σj
1
d z¯ j σ E j − z¯ j γj P P ˆ σj ˆ σj 1 1 σ j i P σ j i P (ˆ ) d z¯ j = (ˆ ) σ j−1 σ j−1 σj , σj P P ˆ ˆ K − z¯ j E − z¯ j γj P P P P ˆ σj ˆ σj 1 σ j i P σ j−1 σ j i P ∞ ˆ ˆ + d z¯ j ( ) σ j−1 σ j , 1 (K P , z¯ j )( P ) σj P ˆ ˆ E − z¯ j γj P P P ˆ σj ˆ σj 1 σ j i P σ j i P σ ˆ ˆ + ( ) (E j−1 , z¯ j )d z¯ j . ( ) σj , σj σj P P P ˆ Kˆ − z¯ j ˆ γj P P P P σj , ˆ P
σ Kˆ j − z¯ j
P σj ˆ P
(III.57)
(III.58)
(III.59)
(III.60)
Each of the expressions (III.58) and (III.59) can be rewritten by adding and subtracting σ j−1
σ ( j−1 )i σPj−1 . P P
For (III.58) we get
Renormalized Electron Mass in Nonrelativistic QED
467
(III.58) σ j−1 σ j−1 1 1 σ j−1 i P σ j−1 i P , σ j−1 = ( ) d z¯ j ( ) σ σ σ j−1 j−1 j−1 P K − z¯ j P E − z¯ j γj P P P P σ j−1 ˆ σj σ σ j−1 i P P + (ˆ j )i σ j−1 , σ j − ( P ) P ˆ γj P P σj σ j−1 ˆ 1 1 σj i σ j−1 i P P d z¯ j (ˆ ) σ j−1 σ j−1 σ j−1 σ j − ( P ) P ˆ K − z¯ j E − z¯ j P P P P σ j−1 ˆ σj σ j i P σ j−1 i P ˆ + , − ( ) ( ) σ P P ˆ σj j−1 γj P P σ j−1 1 1 σ j−1 i P ( ) d z¯ j σ j−1 σ j−1 σ j−1 P K − z¯ j E − z¯ j P P P σ j−1 σ j−1 i P + , ( ) σ P j−1 γj P σ j−1 ˆ σj 1 1 σ j i P σ j−1 i P d z¯ j . (ˆ ) σ j−1 σ j−1 σ j−1 σ j − ( P ) P ˆ K − z¯ j E − z¯ j P
P
P
(III.61)
(III.62)
(III.63)
(III.64)
P
The difference in Eq. (III.50) corresponds to the sum of the terms (III.59)-(III.60) and of the terms (III.62)-(III.64). In fact, (III.61) corresponds to the first term in (III.50) after a contour deformation from γ j−1 to γ j . The sum of the remainder terms (III.59), (III.60), and (III.62)-(III.64) can be bounded by j (1−2δ) , for , R0 and α small enough but independent of j, for any P ∈ S. (We recall 1 that R0 can be taken arbitrarily small, provided α is small enough, and that > Cα 16 .) The details are as follows. • For (III.62)-(III.64) use the following inequalities 1 R02 1 σ j−1 i σ j−1 ), ( ) ≤ O( 1 P P K σ j−1 − z¯ j α 4 ( j−1)δ P σ 1 1 σ j−1 ˆ σj i ˆ j ≤ O(α 4 j (1−δ) + α 2 j−1 ), ( P ) − ( P )i P 1 1 ≤ O( j ), σ j−1 K − z¯ j P
(III.65) (III.66) (III.67)
Fσ j
1 σ j−1 σ j−1 i ˆ σ j ( P ) ( P − P ) ≤ O(α 4 j (1−δ) ), 1 σ j−1 ˆ σj P − P ≤ α 4 j (1−δ) .
(III.68) (III.69)
In order to derive the inequality in Eq. (III.66), one uses Eqs. (II.63), (II.73), and (II.56)-(II.57).
468
J. Fröhlich, A. Pizzo σ j−1
• For (III.59), after adding and subtracting
σ ( j−1 )i σPj−1 , P
one also has to use (III.8)
P
and that
σ i σ 1 σ j−1 σj σ j−1 j−1 j−1 −(K P |σ j + EˆP − E P ) σ j−1 P P K − zj P ⎛ ⎞ 1 R04 1 j−1 ⎝ ⎠. ≤ O α2 1 α 4 ( j−1)δ
(III.70)
• To bound (III.60), note that
(III.60) =
=
ˆ σj 1 σ j i P P (ˆ ) −2πi , ˆ σ j Kˆ σ j − E σ j P ˆ σj P P P P σj ˆ ˆ σj 1 σ j i P σ j i P +2πi (ˆ ) (ˆ ) , P ˆ σ j Kˆ σ j − E σ j−1 P ˆ σj P P P P σ j−1 σj ˆ σj − E ) (E 1 σ P P P ˆ σj i 2πi (ˆ j )i σj , σj σj σj σ j−1 ( P ) P ˆ ˆ ˆ K −E K −E P P P P P σ (ˆ j )i P
ˆ σj
(III.71)
(III.72)
ˆ σj P σ P
ˆ j
,
(III.73) σ
σ
where |E j−1 − E j | ≤ O(α j−1 ). Then use the following inequality analogous to P P (III.65): 1 R02 1 σj i σj ˆ ˆ ). ( ) ≤ O( 1 P P Kˆ σ j − E σ j α 4 jδ P
(III.74)
P
Theorem III.3. For α small enough, | P| := lim σ →0 limit
∂ 2 E σ | P| 2 (∂| P|)
∂ 2 E σ | P| 2 (∂| P|)
converges, as σ → 0. The limiting function,
, is Hölder-continuous in P ∈ S (for an exponent η > 0). The lim | P| =1
(III.75)
α→0
holds true uniformly in P ∈ S. Proof. It is enough to prove the result for a fixed choice of a sequence {σ j }∞ j=0 . The estimate in Lemma III.2 implies the existence of lim j→∞ ∂i2 E
σj i iˆ . | P=P | P|
We now observe that ∂i2 E σ0 | P=P i iˆ = 1 (see Eq. (III.49)), because | P|
( σ0 )i ≡ ( P f )i , σ0 ≡ f . P
P
(III.76)
Renormalized Electron Mass in Nonrelativistic QED
469
According to the constraints in Theorem III.1, we can pick = O(α ν ) for some small exponent ν > 0 so that, for α small enough, Lemma III.2 and (III.76) yield σ σj 1 j 1 1 σj i σ j i P P d z¯ j σ j ) ( ) , ( | i ˆ σ σ σ P P πi γ j E − z¯ j j K j − z¯ j j P=P i P
P
P
P
≤ O(α ν(1−2δ) ),
(III.77)
uniformly in j ∈ N. Hence the limit (III.75) follows. The Hölder-continuity in P of | P| is a trivial consequence of the analyticity in P ∈ S of E σ , for any σ > 0, and of Lemma III.2; see [12] for similar results. P
Corollary III.4. For α small enough, the function E P := limσ →0 E σ , P ∈ S, is twice P differentiable, and ∂ 2 E | P|
2 (∂| P|)
= | P| .
(III.78)
Proof. The result follows from the Hölder-continuity of | P| , of lim σ →0
∂ E σ
| P|
∂| P|
, and from
the fundamental theorem of calculus applied to the functions E P and limσ →0 because
∂ E σ
| P|
∂| P|
,
• ∂ E σ
| P|
∂| P|
•
and
∂ 2 E σ
| P|
2 (∂| P|)
(III.79)
converge pointwise, for P ∈ S, as σ → 0, σ ∂2 Eσ ∂E | P| | P| and 2 ∂| P| (∂| P|)
(III.80)
are uniformly bounded in σ , for all P ∈ S. Acknowledgements. The authors thank Thomas Chen for very useful discussions. A.P. was supported by NSF grant DMS-0905988. Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
470
J. Fröhlich, A. Pizzo
References 1. Bach, V., Froehlich, J., Pizzo, A.: Infrared-finite algorithms in QED II. The expansion of the groundstate of an atom interacting with the quantized radiation field. Adv. Math. 220(4), 1023–1074 (2009) 2. Bach, V., Chen, T., Fröhlich, J., Sigal, I.M.: The renormalized electron mass in non-relativistic quantum electrodynamics. J. Funct. Anal. 243(2), 426–535 (2007) 3. Chen, T.: Infrared renormalization in nonrelativistic QED and scaling criticality. J. Funct. Anal. 254(10), 2555–2647 (2008) 4. Chen, T., Fröhlich, J.: Coherent Infrared Representations in Nonrelativistic QED. Proc. Symp. Pure Math. 76 (B. Simon 60-th Birthday Volume), Providence, RI: Amer. Math. Soc., 2007, pp. 25–46 5. Chen, T., Fröhlich, J., Pizzo, A.: Infraparticle scattering states in non-relativistic QED: I. the bloch-nordsieck paradigm. http://arxiv.org/abs/07092493v2[math-ph], 2009 6. Chen, T., Fröhlich, J., Pizzo, A.: Infraparticle scattering states in non-relativistic QED: II. mass hell properties. J. Math. Phys. 50, 012103 (2009) 7. Fierz, M., Pauli, W.: Nuovo cimento 15, 167 (1938) 8. Hiroshima, F., Spohn, H.: Mass renormalization in non-relativistic quantum electrodynamics. J. Math. Phys. 46(4) (2005) 9. Hainzl, C., Seiringer, R.: Mass renormalization and energy level shift in non-relativistic QED. Adv. Theor. Math. Phys. 6(5), 847–871 (2003) 10. Lieb, E.T., Loss, M.: Self Energy of Electrons in Non-perturbative QED. Conference Moshe Flato 1999, Vol. I, Dijon, Math. Phys. Stud. 21, Dordrecht: Kluwer Acad. Publ., 2000, pp. 327–344 11. Lieb, E.T., Loss, M.: A bound on binding energies and mass renormalization in model of quantum electrodynamics. J. Stat. Phys. 108, 1057–1069 (2002) 12. Pizzo, A.: One-particle (improper) states in Nelson’s massless model. Ann. H. Poincaré 4(3), 439–486 (2003) 13. Pizzo, A.: Scattering of an Infraparticle: The One-particle (improper) Sector in Nelson’s massless model. Ann. H. Poincaré 4(3), 439–486 (2003) 14. Schweber, S.S.: An introduction to Relativistic Quantum Field Theory. New York: Harper and Row, 1961 15. Reed, M., Simon, B.: Methods of Modern Mathematical Physics. Vol. I-II-III-IV, New York: Academic Press, 1972, 1975, 1977, 1978 Communicated by H. Spohn
Commun. Math. Phys. 294, 471–503 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0957-3
Communications in
Mathematical Physics
New Bounds for the Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2 Hubert Lacoin Université Paris Diderot and Laboratoire de Probabilités et Modèles Aléatoires (CNRS U.M.R. 7599), U.F.R. Mathématiques, Case 7012 (Site Chevaleret), 75205 Paris cedex 13, France. E-mail:
[email protected] Received: 18 February 2009 / Accepted: 31 August 2009 Published online: 27 November 2009 – © Springer-Verlag 2009
Abstract: We study the free energy of the directed polymer in a random environment model in dimension 1 + 1 and 1 + 2. For dimension one, we improve the statement of Comets and Vargas in [8] concerning very strong disorder by giving sharp estimates on the free energy at high temperature. In dimension two, we prove that very strong disorder holds at all temperatures, thus solving a long standing conjecture in the field. 1. Introduction 1.1. The model. We study a directed polymer model introduced by Huse and Henley (in dimension 1 + 1) [18] with the purpose of investigating impurity-induced domain-wall roughening in the 2D-Ising model. The first mathematical study of directed polymers in random environment was made by Imbrie and Spencer [19], and was followed by numerous authors [1,3–6,8,9,19,23,26] (for a review on the subject see [7]). Directed polymers in a random environment model, in particular, polymer chains in a solution with impurities. In our set–up the polymer chain is the graph {(i, Si )}1≤i≤N of a nearest–neighbor path in Zd , S starting from zero. The equilibrium behavior of this chain is described by a measure on the set of paths: the impurities enter the definition of the measure as disordered potentials, given by a typical realization of a field of i.i.d. random variables ω = {ω(i,z) ; i ∈ N, z ∈ Zd } (with associated law Q). The polymer chain will tend to be attracted by larger values of the environment and repelled by smaller ones. More precisely, we define the Hamiltonian H N (S) :=
N
ωi,Si .
(1.1)
i=1
We denote by P the law of the simple symmetric random walk on Zd starting at 0 (in the sequel P f (S), respectively Qg(ω), will denote the expectation with respect to P,
472
H. Lacoin
respectively Q). One defines the polymer measure of order N at inverse temperature β as (β)
µ N (S) = µ N (S) :=
1 exp (β H N (S)) P(S), ZN
(1.2)
where Z N is the normalization factor which makes µ N a probability measure Z N := P exp (β H N (S)) .
(1.3)
We call Z N the partition function of the system. In the sequel, we will consider the case of ω(i,z) with zero mean and unit variance and such that there exists B ∈ (0, ∞] such that λ(β) = log Q exp(βω(1,0) ) < ∞,
for 0 ≤ β ≤ B.
(1.4)
Finite exponential moments are required to guarantee that Q Z N < ∞. The model can be defined and it is of interest also with environments with heavier tails (see e.g. [26]) but we will not consider these cases here.
1.2. Weak, strong and very strong disorder. In order to understand the role of disorder in the behavior of µ N , as N becomes large, let us observe that, when β = 0, µ N is the law of the simple random walk, so that we know that, properly rescaled, the polymer chain will look like the graph of a d-dimensional Brownian motion. The main questions that arise for our model for β > 0 are whether or not the presence of disorder breaks the diffusive behavior of the chain for large N , and what the polymer measure looks like when diffusivity does not hold. Many authors have studied diffusivity in polymer models: in [3], Bolthausen remarked that the renormalized partition function W N := Z N /(Q Z N ) has a martingale property and proved the following zero-one law: (1.5) Q lim W N = 0 ∈ {0, 1}. N →∞
A series of papers [1,3,9,19,23] lead to Q lim W N = 0 = 0 ⇒ diffusivity , N →∞
(1.6)
and a consensus in saying that this implication is an equivalence. For this reason, it is natural and it has become customary to say that weak disorder holds when W N converges to some non-degenerate limit and that strong disorder holds when W N tends to zero. Carmona and Hu [4] and Comets, Shiga and Yoshida [6] proved that strong disorder holds for all β in dimension 1 and 2. The result was completed by Comets and Yoshida [9]: we summarize it here Theorem 1.1. There exists a critical value βc = βc (d) ∈ [0, ∞] (depending on the law of the environment) such that • Weak disorder holds when β < βc . • Strong disorder holds when β > βc .
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
473
Moreover: βc (d) = 0 for d = 1, 2 βc (d) ∈ (0, ∞] for d ≥ 3.
(1.7)
We mention also that the case βc (d) = ∞ can only occur when the random variable ω(0,1) is bounded. In [4 and 6] a characterization of strong disorder has been obtained in terms of localization of the polymer chain: we cite the following result [6, Theorem 2.1]: Theorem 1.2. If S (1) and S (2) are two i.i.d. polymer chains, we have ⎧ ⎫ ⎨ ⎬ (1) (2) µ⊗2 (S = S ) = ∞ . Q lim W N = 0 = Q N N −1 N ⎩ ⎭ N →∞
(1.8)
N ≥1
Moreover if Q{lim N →∞ W N = 0} = 1 there exists a constant c (depending on β and the law of the environment) such that for − c log W N ≤
N n=1
1 (1) (2) µ⊗2 n−1 (Sn = Sn ) ≤ − log W N . c
(1.9)
One can notice that (1.9) has a very strong meaning in terms of trajectory localization when W N decays exponentially: it implies that two independent polymer chains tend to share the same endpoint with positive probability. For this reason we introduce now the notion of free energy, we refer to [6, Prop. 2.5] and [9, Theorem 3.2] for the following result: Proposition 1.3. The quantity p(β) := lim
N →∞
1 log W N , N
(1.10)
exists Q-a.s., it is non-positive and non-random. We call it the free energy of the model, and we have 1 p(β) = lim Q log W N =: lim p N (β). (1.11) N →∞ N N →∞ Moreover p(β) is non-increasing in β. We stress that the inequality p(β) ≤ 0 is the standard annealing bound. In view of (1.9), it is natural to say that very strong disorder holds whenever p(β) < 0. One can c (d) the critical value of β for the free energy i.e. : moreover define β c (d). p(β) < 0 ⇔ β > β
(1.12)
c (d) is the natural critical point Let us stress that, from the physicists’ viewpoint, β c (d) > 0). In because it is a point of non-analyticity of the free energy (at least if β c (d) ≥ βc (d). It is widely believed that view of this definition, we obviously have β c (d) = βc (d), i.e. that there exists no intermediate phase where we have strong disorβ der but not very strong disorder. However, this is a challenging question: Comets and c (1) = 0. In this paper, we Vargas [8] answered it in dimension 1 + 1 by proving that β c (2) = 0. make their result more precise. Moreover we prove that β
474
H. Lacoin
1.3. Presentation of the results. The first aim of this paper is to sharpen the result of Comets and Vargas on the 1 + 1-dimensional case. In fact, we are going to give a precise statement on the behavior of p(β) for small β. Our result is the following: Theorem 1.4. When d = 1 and the environment satisfies (1.4), there exist constants c and β0 < B (depending on the distribution of the environment) such that for all 0 ≤ β ≤ β0 we have 1 − β 4 [1 + (log β)2 ] ≤ p(β) ≤ −cβ 4 . (1.13) c We believe that the logarithmic factor in the lower bound is an artifact of the method. In fact, by using replica-coupling, we have been able to get rid of it in the Gaussian case. Theorem 1.5. When d = 1 and the environment is Gaussian, there exists a constant c such that for all β ≤ 1, 1 − β 4 ≤ p(β) ≤ −cβ 4 . (1.14) c These estimates concerning the free energy give us some idea of the behavior of µ N for small β. Indeed, Carmona and Hu in [4, Sect. 7] proved a relation between p(β) and the overlap (although their notation differs from ours). This relation together with our estimates for p(β) suggests that, for low β, the asymptotic contact fraction between independent polymers 1 ⊗2 µ 1{S (1) =S (2) } , n n N N N
lim
N →∞
(1.15)
n=1
behaves like β 2 . c (2) = 0. As for the 1 + 1-dimensional case, The second result we present is that β our approach yields an explicit bound on p(β) for β close to zero. Theorem 1.6. When d = 2, there exist constants c and β0 such that for all β ≤ β0 ,
c 1 (1.16) − exp − 2 ≤ p(β) ≤ − exp − 4 , cβ β so that c (2) = 0, β
(1.17)
and 0 is a point of non-analyticity for p(β). Remark 1.7. After the appearance of this paper as a preprint, the proof of the above result has been adapted by Bertin [2] to prove the exponential decay of the partition function for Linear Stochastic Evolution in dimension 2, a model that is a slight generalisation of the directed polymer in a random environment. Remark 1.8. Unlike in the one dimensional case, the two bounds on the free energy provided by our methods do not match. We believe that the second moment method, that gives the lower bound is quite sharp and gives the right order of magnitude for log p(β). The method developed in [16] to sharpen the estimate on the critical point shift for pinning models at marginality adapted to the context of directed polymer should be able to improve the result, getting p(β) ≤ − exp(−cε β −(2+ε) ) for all β ≤ 1 for any ε.
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
475
1.4. Organization of the paper. The various techniques we use have been inspired by ideas used successfully for another polymer model, namely the polymer pinning on a defect line (see [10,14,15,24,25]). However the ideas we use to establish lower bounds differ sensibly from the ones leading to the upper bounds. For this reason, we present first the proofs of the upper bound results in Sects. 2, 3 and 4. The lower bound results are proven in Sects. 5, 6 and 7. To prove the lower bound results, we use a technique that combines the so-called fractional moment method and change of measure. This approach has been first used for the pinning model in [10] and it has been refined since in [15,25]. In Sects. 2, we prove a non-optimal upper bound for the free energy in the case of a Gaussian environment in dimension 1 + 1 to introduce the reader to this method. In Sect. 3 we prove the optimal upper bound for an arbitrary environment in dimension 1 + 1, and in Sect. 4 we prove our upper bound for the free energy in dimension 1 + 2 which implies that very strong disorder holds for all β. These sections are placed in increasing order of technical complexity, and therefore, should be read in that order. Concerning the lower–bounds proofs: Sect. 5 presents a proof of the lower bound of Theorem 1.4. The proof combines the second moment method and a directed percolation argument. In Sect. 6 the optimal bound is proven for Gaussian environment, with a specific Gaussian approach similar to what is done in [24]. In Sect. 7 we prove the lower bound for arbitrary environment in dimension 1 + 2. These three parts are completely independent of each other. 2. Some Warm Up Computations 2.1. Fractional moment. Before going into the core of the proof, we want to present here the starting step that will be used repeatedly throughout Sects. 2, 3 and 4. We want to find an upper–bound for the quantity p(β) = lim
N →∞
1 Q log W N . N
(2.1)
However, it is not easy to handle the expectation of a log, for this reason we will use the following trick. Let θ ∈ (0, 1), we have (by Jensen inequality) Q log W N =
1 1 Q log W Nθ ≤ log QW Nθ . θ θ
(2.2)
Hence p(β) ≤ lim inf N →∞
1 log QW Nθ . θN
(2.3)
We are left with showing that the fractional moment QW Nθ decays exponentially which is a problem that is easier to handle. 2.2. A non-optimal upper–bound in dimension 1 + 1. To introduce the reader to the general method used in this paper, combining fractional moment and change of measure, we start by proving a non–optimal result for the free–energy, using a finite volume criterion. As a more complete result is to be proved in the next section, we restrict to the Gaussian case here. The method used here is based on the one of [8], majorizing the free energy
476
H. Lacoin
of the directed polymer by the one of multiplicative cascades. Let us mention that is has been shown recently by Liu and Watbled [22] that this majorization is in a sense optimal, they obtained this result by improving the concentration inequality for the free energy. The idea of combining the fractional moment with a change of measure and finite volume criterion has been used with success for the pinning model in [10]. Proposition 2.1. There exists a constant c such that for all β ≤ 1, p(β) ≤ −
cβ 4 . (| log β| + 1)2
(2.4)
Proof of Proposition 2.1 in the case of Gaussian environment. For β sufficiently small, β|2 for a fixed constant C1 (here and throughout the we choose n to be equal to C1 | log β4 paper for x ∈ R, x , respectively x , will denote the upper, respectively the lower, integer part of x) and define θ := 1 − (log n)−1 . For x ∈ Z we define n 2 Wn (x) := P exp [βω(i,Si ) − β /2] 1{Sn =x} . (2.5) i=1
Note that x∈Z Wn (x) = Wn . We use a statement which can be found in the proof of Theorem 3.3. in [8]: θ log QWnm ≤ m log Q [Wn (x)]θ ∀m ∈ N. (2.6) x∈Z
This combined with (2.3) implies that p(β) ≤
1 log Q [Wn (x)]θ . θn
(2.7)
x∈Z
Hence, to prove the result, it is sufficient to show that Q [Wn (x)]θ ≤ e−1 ,
(2.8)
x∈Z
for our choice of θ and n. In order to estimate Q[Wn (x)]θ we use an auxiliary measure region where √ √Q. The the walk (Si )0≤i≤n is likely to go is Jn = [1, n] × [−C2 n, C2 n] ∩ N × Z, where C2 is a big constant. as the measure under which the ωi,x are still independent Gaussian We define Q i,x = −δn 1(i,x)∈Jn , where δn = 1/(n 3/4 variables with variance 1, but such that Qω √ 2C2 log n). This measure is absolutely continuous with respect to Q and ⎛ ⎞ 2 dQ δ δn ωi,x + n ⎠ . = exp ⎝− (2.9) dQ 2 (i,x)∈Jn
Then we have for any x ∈ Z, using the Hölder inequality we obtain, 1 1−θ
1−θ dQ dQ θ θ n (x) θ . Q Wn (x) = Q QW (Wn (x)) ≤ Q dQ dQ
(2.10)
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
477
The first term on the right-hand side can be computed explicitly and is equal to
Q
dQ dQ
θ 1−θ
1−θ
θ δn2 # Jn = exp 2(1 − θ )
≤ e,
(2.11)
where the last inequality is obtained by replacing δn and θ by their values (recall θ = 1 − (log n)−1 ). Therefore combining (2.10) and (2.11) we get that Q
(Wn (x))θ ≤ e
n (x) QW
θ
.
(2.12)
|x|≤n
x∈Z
To bound the right–hand side, we first get rid of the exponent θ in the following way:
n (x) θ ≤ n −3θ #{x ∈ Z, |x| ≤ n such that QW n (x) ≤ n −3 } n −3θ QW
|x|≤n
+
|x|≤n
n (x)n 3(1−θ) . 1{ QW n (x)>n −3 } QW
(2.13)
If n is sufficiently large ( i.e., β sufficiently small) the first term on the right-hand side is smaller than 1/n so that
n (x) QW
|x|≤n
θ
n + 1. ≤ exp(3) QW n
(2.14)
is We are left with showing that the expectation of Wn with respect to the measure Q small. It follows from the definition of Q that n = P exp (−βδn #{i | (i, Si ) ∈ Jn }) , QW
(2.15)
n ≤ P{the trajectory S goes out of Jn } + exp(−nβδn ). QW
(2.16)
and therefore
One can choose C2 such that the first term is small, and the second term is equal to √ 1/4 √ exp(−βn 1/4 / 2C2 log n) ≤ exp(−C1 /4 C2 ) that can be arbitrarily small by choosing C1 large compared to (C2 )1/2 . In that case (2.8) is satisfied and we have p(β) ≤ for small enough β.
1 β4 log e−1 ≤ − θn 2C1 | log β|2
(2.17)
478
H. Lacoin
3. Proof of the Upper Bound of Theorem 1.4 and 1.5 The upper bound we found in the previous section is not optimal, and can be improved by replacing the finite volume criterion (2.8) by a more sophisticated coarse graining method. The technical advantage of the coarse graining we use is that we will not have to choose the θ of the fractional moment close to 1 as we did in the previous section and this is the way we get rid of the extra log factor we had. The idea of using this type of coarse graining for the copolymer model appeared in [25] and this has been a substantial source of inspiration for this proof. We will prove the following result first in the case of the Gaussian environment, and then adapt the proof to the general environment. Proof in the case of Gaussian environment. Let n be the smallest squared integer bigger than C3 β −4 (if β is small we are sure that n ≤ 2C3 β −4 ). The number n will be used in the sequel of the proof as a scaling factor. Let θ < 1 be fixed (say θ = 1/2). We consider a system of size N = nm (where m is √ meant to tend √ to infinity). Let Ik denote the interval Ik = [k n, (k + 1) n). In order to estimate QW Nθ we decompose W N according to the contribution of different families path:
WN =
W (y1 ,y2 ,...,ym ) ,
(3.1)
y1 ,y2 ,...,ym ∈Z
where
W (y1 ,y2 ,...,ym )
N
β2 βωi,Si − 1 Sin ∈I y ,∀i=1,...,m . = P exp i 2
(3.2)
i=1
θ θ Then, we apply the inequality ai ≤ ai (which holds for any finite or countable collection of positive real numbers) to this decomposition and average with respect to Q to get y1 ,y2 ,...,ym ∈Z
θ QWnm ≤
θ QW(y . 1 ,y2 ,...,ym )
(3.3)
In order to estimate QW θ(y1 ,y2 ,...,ym ) , we use an auxiliary measure as in the previous section. The additional idea is to make the measure change depend on y1 , . . . , ym . For every Y = (y1 , . . . , ym ) we define the set JY as √ √ JY := ( km + i, yk n + z), k = 0, . . . , m − 1, i = 1, . . . , n, |z| ≤ C4 n , (3.4) where y0 is equal to zero. Note that for big values of n and m, # JY ∼ 2C4 mn 3/2 .
(3.5)
Y to be the measure under which the ω(i,x) are independent We define the measure Q Y ω(i,x) = −δn 1{(i,x)∈JY } , where δn = Gaussian variables with variance 1 and mean Q −1/2 Y is absolutely continuous with respect to Q and its density is n −3/4 C4 . The law Q equal to
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
479
Fig. 1. The partition of Wnm into W (y1 ,...,ym ) is to be viewed as a coarse graining. For m = 8, (y1 , . . . , y8 ) = (y1 ,...,ym )
(1, −1, 2, 3, 1, −1, −3, 1), Wn thick barriers on the figure
corresponds to the contribution to W N of the path going through the
⎛ ⎞ " ! Y dQ (ω) = exp ⎝− δn ω(i,x) + δn2 /2 ⎠ . dQ
(3.6)
(i,x)∈JY
Using Hölder inequality with this measure as we did in the previous section, we obtain θ Y dQ W θ = Q Q W(y 1 ,y2 ,...,ym ) Y (y1 ,y2 ,...,ym ) dQ
1 1−θ
θ 1−θ dQ ≤ QY Q Y W(y1 ,...,ym ) . Y dQ
(3.7)
The value of the first term can be computed explicitly,
Q
dQ Y dQ
θ 1−θ
1−θ
# JY θ δn2 = exp 2(1 − θ )
≤ exp(3m),
(3.8)
where the upper bound is obtained by using the definition of δn , (3.5) and the fact that θ = 1/2. Now we compute the second term
Y W(y1 ,...,ym ) = P exp (−βδn # {i|(i, Si ) ∈ JY }) 1{Skn ∈I y , ∀k∈[1,m]} . Q k
(3.9)
We define √ J := {(i, x), i = 1, . . . , n, |x| ≤ C4 n}, √ J := {(i, x), i = 1, . . . , n, |x| ≤ (C4 − 1) n}.
(3.10)
480
H. Lacoin
Equation (3.9) implies that (recall that Px is the law of the simple random walk starting from x, and that we set y0 = 0)
Y W(y1 ,...,ym ) ≤ Q
m # k=1
max Px exp (−βδn # {i : (i, Si ) ∈ J }) 1{Sn ∈I yk −yk−1 } . x∈I0
(3.11)
Combining this with (3.1), (3.7) and (3.8) we have ⎡ ⎤ θ
max Px exp (−βδn # {i : (i, Si ) ∈ J }) 1{Sn ∈I y } ⎦ . log QW Nθ ≤m ⎣3+log y∈Z
x∈I0
(3.12) If the quantity in the square brackets is smaller than −1, by Eq. (2.3) we have p(β) ≤ −1/n. Therefore, to complete the proof it is sufficient to show that θ
max Px exp (−βδn # {i : (i, Si ) ∈ J }) 1{Sn ∈I y } (3.13) y∈Z
x∈I0
is small. To reduce the problem to the study of a finite sum, we observe (using some well known result on the asymptotic behavior of a random walk) that given ε > 0 we can find R such that θ
θ max Px exp (−βδn # {i : (i, Si ) ∈ J }) 1{Sn ∈I y } ≤ max Px {Sn ∈ I y } |y|≥R
x∈I0
|y|≥R
x∈I0
≤ ε.
(3.14)
To estimate the remainder of the sum we use the following trivial bound: θ
max Px exp (−βδn # {i : (i, Si ) ∈ J }) 1{Sn ∈I y } |y|
x∈I0
θ ≤ 2R max Px exp (−βδn # {i : (i, Si ) ∈ J }) . x∈I0
(3.15)
Then we get rid of the max in the sum by observing that if a walk starting from x makes a step in J , the walk with the same increments starting from 0 will make the same step in J (recall (3.10)), max Px exp (−βδn # {i : (i, Si ) ∈ J }) ≤ P exp −βδn # i|(i, Si ) ∈ J . (3.16) x∈I0
Now we are left with something similar to what we encountered in the previous section P exp −βδn # i : (i, Si ) ∈ J ≤ P{ the random walk goes out of J } + exp(−nβδn ).
(3.17) If C4 is chosen large enough, the first term can be made arbitrarily small by choosing −1/4 √ C4 large, and the second is equal to exp(−C3 / C4 ) and can be made also arbitrarily
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
481
small if C3 is chosen large enough once C4 is fixed. An appropriate choice of constant and the use of (3.16) and (3.17) can lead then to
θ 2R max Px exp (−βδn # {i : (i, Si ) ∈ J }) ≤ ε. x∈I0
This combined with (3.14) completes the proof.
(3.18)
Proof of the general case. In the case of a general environment, some modifications have to be made in the proof above, but the general idea remains the same. In the change of measure one has to change the shift of the environment in JY (3.6) by an exponential tilt of the measure as follows: ⎛ ⎞ Y dQ (β) = exp ⎝− δn ω(i,z) + λ(−δn ) ⎠ . dQ
(3.19)
(i,z)∈JY
The formula estimating the cost of the change of measure (3.8) becomes
Q
dQ Y dQ
θ 1−θ
1−θ
θ δn + θ λ(−δn ) = exp # JY (1 − θ )λ 1−θ ≤ exp(2m),
(3.20)
where the last inequality is true if βn is small enough if we consider that θ = 1/2 and x→0
use the fact that λ(x) ∼ x 2 /2 (ω has 0 mean and unit variance). The next thing we have to do is to compute the effect of this change of measure in this general case, i.e.
Y W(y1 ,...,ym ) , the quantity find an equivalent for (3.9). When computing Q Y exp(βω1,0 − λ(β)) = exp [λ(β − δn ) − λ(−δn ) − λ(β)] Q
(3.21)
appears instead of exp(−βδn ). Using twice the mean value theorem, one gets that there exists h and h in (0, 1) such that λ(β − δn ) − λ(−δn ) − λ(β) = δn λ (−hδn ) − λ (β − hδn ) = −βδn λ (−hδn + h β).
(3.22)
And as ω has unit variance, lim x→0 λ (x) = 1. Therefore if β and δn are chosen small enough, the right-hand side of the above is less than −βδn /2. So (3.9) can be replaced by
Y W(y1 ,...,ym ) ≤ P exp − βδn # {i|(i, Si ) ∈ JY } 1{Skn ∈I y , ∀k∈[1,m]} . Q k 2
(3.23)
The remaining steps follow closely the argument exposed for the Gaussian case.
482
H. Lacoin
4. Proof of the Upper Bound in Theorem 1.6 In this section, we prove the main result of the paper: very strong disorder holds at all temperatures in dimension 2. The proof is technically quite involved. It combines the tools of the two previous sections with a new idea for the change of measure: changing the covariance structure of the environment. We mention that this idea was introduced recently in [15] to deal with the marginal disorder case in the pinning model. We choose to present first a proof for the Gaussian case, where the idea of the change of measure is easier to grasp. Before starting, we sketch the proof and how it should be decomposed in different steps: (a) We reduce the problem by showing that it is sufficient to show that for some real number θ < 1, QW Nθ decays exponentially with N . (b) We use a coarse graining decomposition of the partition function by splitting it into different contributions that correspond to trajectories that stay in a large corridor. This decomposition is similar to the one used in Sect. 3. (c) To estimate the fractional moment terms appearing in the decomposition, we change the law of the environment around the corridors corresponding to each contribution. More precisely, we introduce negative correlations into the Gaussian field of the environment. We do this change of measure in such a way that the new measure is not very different from the original one. (d) We use some basic properties of the random walk in Z2 to compute the expectation under the new measure. Proof for the Gaussian environment. We fix n to be the smallest squared integer bigger than exp(C5 /β 4 ) for some large constant C5 to be defined later, for small β we have n ≤ exp(2C5 /β 4 ). The number n will be used of the proof fac√ √ √ in the sequel √ as a scaling tor. For y = (a, b) ∈ Z2 we define I y = [a n, (a +1) n −1]×[b n, (b+1) n −1] so that I y are disjoint and cover Z2 . For N = nm, we decompose the normalized partition function W N into different contributions, very similarly to what is done in dimension one (i.e. decomposition (3.3)), and we refer to Fig. 2 to illustrate how the decomposition looks like: WN = W(y1 ,...,ym ) , (4.1)
y1 ,...,ym ∈Z2
where
W(y1 ,...,ym )
N " ! 2 βωi,Si − β /2 1 Sin ∈I y = P exp
i ,∀i=1,...,m
.
(4.2)
i=1
θ We fix θ < 1 and apply the inequality ( ai )θ ≤ ai (which holds for any finite or countable collection of positive real numbers) to get θ QW Nθ ≤ QW(y . (4.3) 1 ,...,ym )
y1 ,...,ym ∈Z2
In order to estimate the different terms in the sum of the right–hand side in (4.3), we Y on the the environment for every Y = (y0 , y1 , . . . , ym ) define some auxiliary measures Q ∈ Zd+1 with y0 = 0. We will choose the measures Q Y absolutely continuous with respect to Q. We use the Hölder inequality to get the following upper bound:
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
483
Fig. 2. This figure represents in a rough way the change of measure Q Y . The region where the mean of ω(i,x) is lowered (the shadow region on the figure) corresponds to the region where the simple random walk is likely to go, given that it goes through the thick barriers
≤
Q
dQ Y dQ
θ 1−θ
1−θ
θ QW(y 1 ,...,ym )
Y W(y1 ,...,ym ) Q
θ
.
(4.4)
Now, we describe the change of measure we will use. Recall that for the 1-dimensional case we used a shift of the environment along the corridor corresponding to Y . The reader can check that this method would not give the exponential decay of W N in this case. Instead we change the covariance function of the environment along the corridor on which the walk is likely to go by introducing some negative correlation. We introduce the change of measure that we use for this case. Given Y = (y0 , y1 , . . . , ym ) we define m blocks (Bk )k∈[1,m] and JY their union (here and in the sequel, |z| denotes the l ∞ norm on Z2 ): ( √ ) √ Bk := (i, z) ∈ N × Z2 : i/n = k and |z − n yk−1 | ≤ C6 n , JY :=
m *
(4.5)
Bk .
k=1
Y to be equal to We fix the covariance of the field ω under the law Q Y ωi,z ωi,z = C Y Q (i,z),( j,z ) ⎧ ⎪ ⎨1{(i,z)=( j,z )} − V(i,z),( j,z ) if ∃ k ∈ [1, m] (4.6) := such that (i, z) and ( j, z ) ∈ Bk ⎪ ⎩1 otherwise, {(i,z)=( j,z )} where
, V(i,z),( j,z ) :=
0
1{|z−z |≤C √| j−i|} √7 100C6 C7 n log n| j−i|
if (i, z) = ( j, z ) otherwise.
(4.7)
484
H. Lacoin
We define := (V(i,z),( j,z ) )(i,z),( j,z )∈B1 . V
(4.8)
One remarks that the so-defined covariance matrix C Y is block diagonal with m iden corresponding to the Bk , k ∈ [1, m], and just ones tical blocks which are copies of I − V on the diagonal elsewhere. Therefore, the change of measure we describe here exists if is definite positive. and only if I − V is associated to a positive vector and therefore is smaller The largest eigenvalue for V than
max
(i,z)∈B1
-V(i,z),( j,z ) - ≤
( j,z )∈B1
C7 . √ C6 log n
(4.9)
is less than (1 − θ )/2 so For the sequel we choose n such that the spectral radius of V is positive definite. With this setup, Q Y is well defined. that I − V Y with respect to Q is given by The density of the modified measure Q
Y 1 dQ 1t (ω) = √ exp − ω((C Y )−1 − I )ω , dQ 2 det C Y
(4.10)
where t
ωMω =
ω(i,z) M(i,z),( j,z ) ω( j,z ) ,
(4.11)
(i,z),( j,z )∈N×Z2
for any matrix M of (N × Z2 )2 with finite support. Then we can compute explicitly the value of the second term in the right-hand side of (4.4),
Q
dQ Y dQ
θ 1−θ
1−θ
. / / =/ 0
1 det
det C Y CY 1−θ
−
θI 1−θ
21−θ .
(4.12)
Note that the above computation is right if and only if C Y − θ I is a definite positive , this holds for large n matrix. Since its eigenvalues are the same of those of (1 − θ )I − V Y , thanks to (4.9). Using again the fact that C is composed of m blocks identical to I − V we get from (4.12),
Q
dQ dQ
θ 1−θ
1−θ
=
) det(I − V /(1 − θ ))1−θ det(I − V
m/2 .
(4.13)
In order to estimate the determinant in the denominator, we compute the Hilbert-Schmidt . One can check that for all n, norm of V 2 2 = V(i,z),( (4.14) V j,z ) ≤ 1. (i,z),( j,z )∈B1
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
485
We use the inequality log(1 + x) ≥ x − x 2 for all x ≥ −1/2 and the fact that the spectral /(1 − θ ) is bounded by 1/2 (cf. (4.9)) to get that radius of V
2 V V V = exp Trace log I − ≥ exp − det I − 1−θ 1−θ (1 − θ )2 (4.15)
1 . ≥ exp − (1 − θ )2 = 0 implies that det(I − V ) ≤ 1. Combining this with For the numerator, Trace V (4.13) and (4.15) we get
θ 1−θ
dQ 1−θ m Q . (4.16) ≤ exp Y 2(1 − θ ) dQ Now that we have computed the term corresponding to the change of measure, we esti
mate W(y1 ,...,ym ) under the modified measure (just by computing the variance of the Gaussian variables in the exponential, using (4.6)) : N
β2 βωi,Si − Q Y W(y1 ,...,ym ) = P Q Y exp 1 Skn ∈I y ,∀k=1,...,m k 2 i=1 ⎞ ⎛
⎜ β2 = P exp ⎜ ⎝ 2 × 1 Skn ∈I y
1
1≤i, j≤N z,z ∈Z2
2 ⎟ Y )} 1{S =z,S =z } ⎟ C(i,z),( − 1 {(i,z)=( j,z i j j,z ) ⎠
.
k ,∀k=1,...,m
(4.17)
Replacing C Y by its value we get that ⎛
2 ⎜ Y W(y1 ,...,ym ) = P exp ⎜− β Q ⎝ 2
⎞
1≤i= j≤N 1≤k≤m
1 Skn ∈I y
k ,∀k=1,...,m
.
√
1{((i,Si ),( j,S j ))∈B 2 , |Si −S j |≤C7 |i− j|} ⎟ k ⎟ √ ⎠ 100C6 C7 n log n| j − i| (4.18)
Now we do something similar to (3.11): for each “slice” of the trajectory (Si )i∈[(m−1)k,mk] , we bound the contribution of the above expectation by maximizing over the starting point (recall that Px denotes the probability distribution of a random walk starting at x). Thanks to the conditioning, the starting point has to be in I yk . Using the translation invariance of the random walk, this gives us the following (∨ stands for maximum):
Y W (y1 ,...,ym ) ≤ Q
m # i=k
max x∈I0
⎡
⎛
β2 × Px ⎣exp ⎝− 2
⎤ ⎞ 1{|Si |∨|S j |≤C6 √n, |Si −S j |≤C7 √|i− j|} ( ) ⎦. ⎠1 √ Sn ∈I yk −yk−1 100C6 C7 n log n| j − i|
1≤i= j≤n
(4.19)
486
H. Lacoin
For trajectories S of a directed random-walk of n steps, we define the quantity G(S) :=
1≤i= j≤n
1{|Si |∨|S j |≤C6 √n, |Si −S j |≤C7 √|i− j|} . √ 100C6 C7 n log n| j − i|
Combining (4.19) with (4.16), (4.4) and (4.3), we finally get ⎤ ⎡
θ m
2 m β ⎣ QW Nθ ≤ exp max Px exp − G(S) 1{Sn ∈I y } ⎦ . x∈I0 2(1 − θ ) 2
(4.20)
(4.21)
y∈Z
The exponential decay of QW Nθ (with rate n) is guaranteed if we can prove that y∈Z
θ
β2 max Px exp − G(S) 1{Sn ∈I y } x∈I0 2
(4.22)
is small. The rest of the proof is devoted to that aim. We fix some ε > 0. Asymptotic properties of the simple random walk, guarantees that we can find R = Rε such that |y|≥R
θ
θ β2 max Px exp − G(S) 1{Sn ∈I y } ≤ max Px {Sn ∈ I y } ≤ε. x∈I0 x∈I0 2
(4.23)
|y|≥R
To estimate the rest of the sum, we use the following trivial and rough bound |y|
θ θ
2
β β2 2 max Px exp − G(S) 1{Sn ∈I y } ≤R max Px exp − G(S) . (4.24) x∈I0 x∈I0 2 2
Then we use the definition of G(S) to get rid of the max by reducing the width of the zone where we have negative correlation:
β2 max Px exp − G(S) x∈I0 2 ⎛ ⎞ 2 1{|Si |∨|S j |≤(C6 −1)√n, |Si −S j |≤C7 √|i− j|} β ⎠ . (4.25) ≤ P exp ⎝− √ 2 100C6 C7 n log n| j − i| 1≤i= j≤n
√ := {(i, z) ∈ N × Z2 : i ≤ m, |z| ≤ (C6 − 1) n}. We get from the above We define B that
β2 max Px exp − G(S) ≤ P{the RW goes out of B} x∈I0 2 ⎛ ⎞ 2 1{|Si −S j |≤C7 √|i− j|} β ⎠. +P exp ⎝− √ 2 100C6 C7 n log n| j − i| 1≤i= j≤n
(4.26)
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
487
One can make the first term of the right-hand side arbitrarily small by choosing C6 large, in particular one can choose C6 such that √ 1 (4.27) P max |Sn | ≥ (C6 − 1) n ≤ (ε/R 2 ) θ . i∈[0,n]
To bound the other term, we introduce the quantity D(n) :=
1 , √ n log n| j − i|
(4.28)
1{|Si −S j |≤C7 √|i− j|} . √ n log n| j − i|
(4.29)
1≤i= j≤n
and the random variable X , X :=
1≤i= j≤n
For any δ > 0, we can find C7 such that P(X ) ≥ (1 − δ)D(n). We fix C7 such that this holds for some good δ (to be chosen soon), and by remarking that 0 ≤ X ≤ D(n) almost surely, we obtain (using the Markov inequality) P{X > D(n)/2} ≥ 1 − 2δ. Moreover we can estimate D(n), getting that for n large enough, 5 D(n) ≥ log n.
(4.30)
(4.31)
Using (4.30) and (4.31) we get ⎛ ⎞
2 2 1{|Si −S j |≤C7 √|i− j|} β ⎠ = P exp − β X P exp ⎝− √ 2 200C6 C7 100C6 C7 n log n| j − i| 1≤i= j≤n (4.32) √
β 2 log n . ≤ 2δ + exp − 200C6 C7 Due to the (recall n ≥ exp(C5 /β 4 )), the second term is less 2 1 choice of n we have made 1/2 than exp −β 2 C5 /(200C6 C7 ) . We can choose δ, C7 and C5 such that the right-hand 1
side is less than (ε/R 2 ) θ . This combined with (4.27), (4.26), (4.24) and (4.23) allow us to conclude that y∈Z
θ
β2 max Px exp − G(S) 1{Sn ∈I y } ≤ 3ε. x∈I0 2
(4.33)
So that with a right choice for ε, (4.21) implies QW Nθ ≤ exp(−m). Then (2.3) allows us to conclude that p(β) ≤ −1/n.
(4.34)
488
H. Lacoin
Proof for the general environment. The case of the general environment does not differ very much from the Gaussian case, but one has a different approach for the change of measure in (4.4). In this proof, we will largely refer to what has been done in the Gaussian case, whose proof should be read first. Let K be a large constant. One defines the function f K on R to be f K (x) = −K 1{x>exp(K 2 )} . Recall the definitions (4.5) and (4.7), and define the gY function of the environment as ⎛ ⎛ ⎞⎞ m fK ⎝ V(i,z),( j,z ) ωi,z ω j,z ⎠⎠ . gY (ω) = exp ⎝ k=1
(i,z),( j,z )∈Bk
Multiplying by gY penalizes by a factor exp(−K ) the environment for which there is too much correlation in one block. This is a way of producing negative correlation in the environment. For the rest of the proof we use the notation Uk := V(i,z),( j,z ) ωi,z ω j,z . (4.35) (i,z),( j,z )∈Bk
We do a computation similar to (4.4) to get 1 ! θ "21−θ θ − 1−θ θ Q gY (ω)W(y1 ,...,yn ) . Q W(y1 ,...,ym ) ≤ Q gY (ω)
The block structure of gY allows to express the first term as a power of m, m
" ! θ θ − 1−θ f K (U1 ) = Q exp − . Q gY (ω) 1−θ
(4.36)
(4.37)
Equation (4.14) says that
So that
and hence
Var Q (U1 ) ≤ 1.
(4.38)
) ( P U1 ≥ exp(K 2 ) ≤ exp(−2K 2 ),
(4.39)
θ f K (U1 ) Q exp − 1−θ
θ ≤ 1 + exp −2K 2 + K ≤ 2, 1−θ
(4.40)
if K is large enough. We are left with estimating the second term nm Q gY (ω)W(y1 ,...,yn ) =P QgY (ω) exp [βωi,Si − λ(β)] 1{Skn ∈I yk ,∀k=1...m} .
i=1
(4.41)
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
489
S the modified measure on For a fixed trajectory of the random walk S, we consider Q the environment with density nm S dQ [βωi,Si − λ(β)] . (4.42) := exp dQ i=1
Under this measure
,
S ωi,z = 0 Q Qωeβω0,1 −λ(β) := m(β)
if z = Si if z = Si .
(4.43)
As ω1,0 has zero-mean and unit variance under Q, (1.4) implies m(β) = β + o(β) S is around zero and that Var QS ωi,z ≤ 2 for all (i, z) if β is small enough. Moreover Q a product measure, i.e. the ωi,z are independent variables under Q S . With this notation (4.41) becomes S [gY (ω)] 1{Skn ∈I y ,∀k=1,...,m} . PQ k
(4.44)
As in the Gaussian case, one wants to bound this by a product using the block structure. Similarly to (4.19), we use translation invariance to get the following upper bound: m # k=1
S exp ( f K (U1 )) 1{Sn ∈I y −y } . max Px Q k k−1 x∈I0
Using this in (4.36) with the bound (4.40) we get the inequality ⎛ ⎞ θ m S exp ( f K (U1 )) 1{Sn ∈I y } ⎠ . max Px Q QW Nθ ≤ 2m(1−θ)⎝ y∈Z2
x∈I0
(4.45)
(4.46)
Therefore to prove exponential decay of QW Nθ , it is sufficient to show that θ max Px Q S exp ( f K (U1 )) 1{Sn ∈I y } y∈Z2
x∈I0
(4.47)
is small. As seen in the Gaussian case ( cf. (4.23),(4.24)), the contribution of y far from zero can be controlled and therefore it is sufficient for our purpose to check S exp ( f K (U1 )) ≤ δ, max Px Q x∈I0
(4.48)
for some small δ. Similarly to (4.26), we force the walk to stay in the zone where the environment is modified by writing √ S exp ( f K (U1 )) ≤ P{ max |Si | ≥ (C6 − 1) n} max Px Q i∈I0
i∈[0,n]
S exp ( f K (U1 )) 1{|S −S |≤(C −1)√n} . +max Px Q n 0 6 x∈I0
(4.49)
490
H. Lacoin
The first term is smaller than δ/6 if C6 is large enough. To control the second term, we will find an upper bound for √ , S exp ( f K (U1 )) 1{max Px Q i∈[0,n] |Si −S0 |≤(C 6 −1) n}
(4.50)
which is uniform in x ∈ I0 . What we do is the following: we show that for most trajectories S the term in f K has a large mean and a small variance with respect to Q S so that f K ( . . . ) = −K with large S probability. The rest will be easy to control as the term in the expectation is at most Q one. S is equal to The expectation of U1 under Q m(β)2
V(i,Si ),( j,S j ) .
(4.51)
1≤i, j≤n
When the walk stays in the block B1 we have (using definition (4.29))
V(i,Si ),( j,S j ) =
1≤i, j≤n
1 X. 100C6 C7
(4.52)
The distribution of X under Px is the same for all x ∈ I0 . It has been shown earlier (cf. (4.30) and (4.31)), that if C7 is chosen large enough, P
√ δ log n m(β)2 ≤ . X≤ 100C6 C7 200C6 C7 6
(4.53)
As m(β) ≥ β/2 if β is small, if C5 is large enough (recall n ≥ exp(C5 /β 4 )), this together with (4.52) gives √ δ 2 2 Px m(β) Q S (U1 ) ≤ 2 exp(K ); max |Si − S0 | ≤ (C6 − 1) n ≤ . (4.54) i∈[0,n] 6 S , we decompose the sum To bound the variance of U1 under Q
U1 =
V(i,z),( j,z ) ωi,z ω j,z = m(β)2
(i,z),( j,z )∈B1
V(i,Si ),( j,S j )
1≤i, j≤n
+2m(β)
V(i,Si ),( j,z ) (ω j,z − m(β)1{z =S j } )
1≤i≤n ( j,z )∈B1
+
V(i,z),( j,z ) (ωi,z − m(β)1{z=Si } )(ω j,z − m(β)1{z =S j } ).
(4.55)
(i,z),( j,z )∈B1
And hence (using the fact that (x + y)2 ≤ 2x 2 + 2y 2 ), Var QS U1≤16m(β)2
( j,z )∈B1
⎛ ⎝
1≤i≤n
⎞2 V(i,Si ),( j,z ) ⎠ +8
2 V(i,z),( j,z ) ,
((i,z),( j,z )∈B1
(4.56)
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
491
where we used that Var QS ωi,z ≤ 2 (which is true for β small enough). The last term is less than 8 thanks to (4.14), so that we just have to control the first one. Independently of the choice of ( j, z ) we have the bound
√ V(i,Si ),( j,z ) ≤
1≤i≤n
log n . C6 C7 n
(4.57)
Moreover it is also easy to check that
( j,z )∈B1
1≤i≤n
V(i,Si ),( j,z ) ≤
C7 n , √ C6 log n
(4.58)
(these two bounds follow from the definition of V(i,z),( j,z ) : (4.7)). Therefore ( j,z )∈B1
⎛ ⎝
⎞2
⎡
V(i,Si ),( j,z ) ⎠ ≤ ⎣
⎤ V(i,Si ),( j,z ) ⎦
( j,z )∈B1 1≤i≤n
1≤i≤n
× max
( j,z)∈B1
V(i,Si ),( j,z ) ≤ 1.
(4.59)
1≤i≤n
Injecting this into (4.56) guarantees that for β small enough, Var QS U1 ≤ 10.
(4.60)
With Chebyshev inequality, if K has been chosen large enough and S U1 ≥ 2 exp(K 2 ), Q
(4.61)
( ) S U1 ≤ exp(K 2 ) ≤ δ/6. Q
(4.62)
we have
Hence combining (4.62) with (4.54) gives √ 2 Px Q S U1 ≤ exp(K ); max |Si − S0 | ≤ (C6 − 1) n ≤ δ/3. i∈[0,n]
(4.63)
We use this in (4.49) to get S exp ( f K (U1 )) ≤ δ + e−K . max Px Q x∈I0 2 So our result is proved provided that K has been chosen large enough.
(4.64)
492
H. Lacoin
5. Proof of the Lower Bound in Theorem 1.4 In this section we prove the lower bound for the free-energy in dimension 1 in an arbitrary environment. To do so we apply the second moment method to some quantity related to the partition function, and combine it with a percolation argument. The idea of the proof was inspired by a study of a polymer model on a hierarchical lattice [21] where this type of coarse-graining appears naturally. Proposition 5.1. There exists a constant C such that for all β ≤ 1 we have p(β) ≥ −Cβ 4 ((log β)2 + 1).
(5.1)
We use two technical lemmas to prove the result. The first is just a statement about scaling of the random walk, the second is more specific to our problem. Lemma 5.2. There exists a constant c RW such that for large even squared integers n, √ √ (5.2) P{Sn = n, 0 < Si < n for 0 < i < n} = c RW n −3/2 + o(n −3/2 ). Lemma 5.3. For any ε > 0 we can find a constant cε and β0 such that for all β ≤ β0 , for every even squared integer n ≤ cε /(β 4 | log β|) we have n−1 βωi,Si − λ(β) P exp
Var Q
√ √ - Sn = n, 0 < Si < n for 0 < i < n < ε. -
i=1
(5.3) Proof of Proposition 5.1 from Lemma 5.2 and 5.3. Let n be some fixed integer and define n−1 := P exp W βωi,Si − λ(β) 1{Sn =√n,0<Si <√n for 0
(5.4)
i=1
which corresponds to the contribution to the√partition function Wn of paths with fixed √ end point n staying within a cell of width n, with the specification the environment depends only on the value of the environment on the last site is not taken into account. W ω in this cell (see Fig. 3). One also defines the following quantities for (i, y) ∈ N × Z: ! n−1 " √ √ √ (y,y+1) := P√n y e k=1 βωin+k,Sk −λ(β) 1 W {Sn = (y+1)n,0<Si −y n< n for 0
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
493
by considering only paths going from one to the other Fig. 3. We consider a resticted partition function W corner of the cell, without going out. This restriction will give us the independence of a random variable corresponding to different cells which will be crucial to make the proof work
Let N = nm be a large integer. We define = N as the set of path √ := {S : ∀i ∈ [1, m], |Sin − S(i−1)n | = n, ∀ j ∈ [1, n − 1], S(i−1)n+ j ∈ S(i−1)n , Sin }, (5.6) where the interval Si(n−1) , Sin is to be seen as Sin , Si(n−1) if Sin < Si(n−1) , and ( ) S := s = (s0 , s1 , . . . , sm ) ∈ Zm+1 : s0 = 0 and |si − si−1 | = 1, ∀i ∈ [1, m] . (5.7) We use the trivial bound
nm W N ≥ P exp (βωi,Si − λ(β)) 1{S∈} ,
(5.8)
i=1
to get that WN ≥
m−1 # s∈S i=0
1 2 (si ,si+1 ) exp βω(i+1)n,s √n − λ(β) W i i+1
(5.9)
does not take into account the site in the (the exponential term is due to the fact that W top corner of each cell). The idea of the proof is to find a value of n (depending on β) such that we are sure that (si ,si+1 ) ) for any value of m we can find a path s such that along the path the values of (W i ) and to do so, it seems natural to seek are not too low (i.e. close to the expectation of W for a percolation argument. Let pc be the critical exponent for directed percolation in dimension 1 + 1 (for an account on directed percolation see [17, Sect. 12.8] and references therein). From Lemma 5.3 and the Chebyshev inequality, one can find a constant C8 and β0 such that 8 for all n ≤ β 4 |Clog and β ≤ β0 , β| ≥ QW /2} ≥ Q{W
pc + 1 . 2
(5.10)
494
H. Lacoin
Fig. 4. This figure illustrates the percolation argument used in the proof. To each cell is naturally associated a y,y±1 ≥ 1/2Q W we open the edge y,y±1 , and these random variables are i.i.d. When W random variable W i i in the corresponding cell (thick edges on the picture). As this happens with a probability strictly superior to pc , we have a positive probability to have an infinite path linking 0 to infinity
We choose n to be the biggest squared even integer that is less than
C8 . (In particular β 4 | log β|
8 if β is small enough.) n ≥ 2β 4 |Clog β| As shown in Fig. 4, we associate to our system the following directed percolation picture. For all (i, y) ∈ N × Z such that i − y is even:
(y,y±1) ≥ (1/2)Q W , we say that the edge linking the opposite corners of the • If W i corresponding cell is open. , we say that the same edge is closed. (y,y±1) < (1/2)Q W • If W i Equation (5.10) and the fact the considered random variables are independent assures that with positive probability there exists an infinite directed path starting from zero. When there exists an infinite open path linking zero to infinity, we can define the m denote this highest path. If m is large highest open path in an obvious way. Let (si )i=1 enough, by law of large numbers we have that with a probability close to one, m ! " βωni,√nsi − λ(β) ≥ −2mλ(β).
(5.11)
i=1
Using this and the percolation argument with (5.9) we finally get that with a positive probability which does not depend on m we have ! "m . (5.12) Wnm ≥ (1/2)e−2λ(β) Q W Taking the log and making m tend to infinity this implies that p(β) ≥
1 ≥ − c log n. −2λ(β) − log 2 + log Q W n n
(5.13)
For some constant c, if n is large enough (we used Lemma 5.2 to get the last inequality. The result follows by replacing n by its value.
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
495
Proof of Lemma 5.2. Let n be square and even. Tk , k ∈ Z denote the first hitting time of k by the random walk S (when k = 0 it denotes the return time to zero). We have √ √ P{Sn = n, 0 < Si < n, for all 1 < i < n} =
n−1
P{T√n/2 = k, S j > 0 for all j < n and T√n = n}
k=1
= P{T√n/2 < n, S j <
(5.14)
√ n for all j < n and T0 = n)},
where the second equality is obtained with the strong Markov property used for T = T√n/2 , and the reflexion principle for the random walk. The last line is equal to √ √ P{ max Sk ∈ [ n/2, n)|T0 = n}P{T0 = n}. (5.15) k∈[0,n]
We use here a variant of Donsker’s Theorem, for a proof see [20, Theorem 2.6]. Lemma 5.4. The process
t →
S nt -√ - T0 = n , t ∈ [0, 1] n
(5.16)
converges in distribution to the normalized Brownian excursion in the space D([0, 1], R). know that (see for example [13, Prop. A.10]) for n even P(T0 = n) = √ We also 2/π n −3/2 + o(n −3/2 ). Therefore, from (5.15) we have √ √ P{Sn = n, 0 < Si < n, for all 1 < i < n} 5 = 2/π n −3/2 P max et ∈ (1/2, 1) + o(n −3/2 ), (5.17) t∈[0,1]
where e denotes the normalized Brownian excursion, and P its law.
Proof of Lemma 5.3. Let β be fixed and small enough, and n be some squared even integer which is less than cε /(β 4 | log β|). We will fix the value cε independently of β later in the proof, and always consider that β is sufficiently small. By a direct computation the variance of n−1 √ √ P exp [βωi,Si − λ(β)] - Sn = n, 0 < Si < n for 0 < i < n (5.18) i=1
is equal to P
⊗2
exp
n−1 i=1
where
- γ (β)1{S (1) =S (2) } -- An − 1, i i
(5.19)
( ) √ √ ( j) ( j) An = Sn = n, 0 < Si < n for 0 < i < n, j = 1, 2 ,
(5.20) ( j)
and γ (β) = λ(2β) − 2λ(β) (recall that λ(β) = log Q exp(βω(1,0) )), and Sn , j = 1, 2 denotes two independent random walks with law denoted by P ⊗2 . From this it follows
496
H. Lacoin
that if n is small the result is quite straightforward. We will therefore only be interested in the case of large n (i.e. bounded away from zero by a large fixed constant). We define τ = (τk )k≥0 = {Si(1) = Si(2) , i ≥ 0} the set where the walks meet (it can be written as an increasing sequence of integers). By the Markov property, the random variables τk+1 − τk are i.i.d. , and we say that τ is a renewal sequence. We want to bound the probability that the renewal sequence τ has too many returns before times n −1, in order to estimate (5.19). To do so, we make the usual computations with the Laplace transform. From [11, p. 577] , we know that 1 − P ⊗2 exp(−xτ1 ) =
1 (1) n∈N exp(−xn)P{Sn
(2)
= Sn }
.
(5.21)
Thanks to the local central limit theorem for the simple random walk, we know that for large n, 1 P{Sn(1) = Sn(2) } = √ + o(n −1/2 ). πn
(5.22)
So we can get from (5.21) that when x is close to zero, √ √ (5.23) log P ⊗2 exp(−xτ1 ) = − x + o( x). √ We fix x0 such that log P exp(−xτ1 ) ≤ x/2 for all x ≤ x0 . For any k ≤ n we have P ⊗2 {|τ ∩ [1, n − 1]| ≥ k} = P ⊗2 {τk ≤ n − 1} ≤ exp((n − 1)x)P ⊗2 exp(−τk x) " ! ≤ exp nx + k log P ⊗2 exp(−xτ1 ) . (5.24) 7 √ For any k ≤ 4n x0 = k0 one can choose x = (k/4n)2 ≤ x0 in the above and use the definition of x0 to get that 1 2 P ⊗2 {|τ ∩ [1, n − 1]| ≥ k} ≤ exp −k 2 /(32n) . (5.25) 6
In the case where k > k0 we simply bound the quantity by 1 2 P ⊗2 {|τ ∩ [1, n − 1]| ≥ k} ≤ exp k02 /(32n) ≤ exp (−nx0 /4) .
(5.26)
By Lemma (5.2), if n is large enough, P ⊗ An ≥ 1/2c2RW n −3 .
(5.27)
A trivial bound on the conditioning gives us 1 1 22 - 3 2 P ⊗2 |τ ∩ [1, n − 1]| ≥ k - An ≤ min 1, 2c−2 n exp −k /(32n) if k ≤ k0 , RW 3 P ⊗2 |τ ∩ [1, n − 1]| ≥ k - An ≤ 2c−2 RW n exp (−nx 0 /4) otherwise. (5.28)
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
497
8 3 We define k1 := 16π n log(2c−2 RW n ) . The above implies that for n large enough we have - P ⊗2 |τ ∩ [1, n − 1]| ≥ k - An ≤ 1 if k ≤ k1 , 1 2 - P ⊗2 |τ ∩ [1, n − 1]| ≥ k - An ≤ exp −k 2 /(64n) if k1 ≤ k ≤ k0 , (5.29) - P ⊗2 |τ ∩ [1, n − 1]| ≥ k - An ≤ exp (−nx0 /8) otherwise. Now we are ready to bound (5.19). Integration by part gives, - P ⊗2 exp (γβ|τ ∩ [1, n − 1]|) - An − 1 9 ∞ - (5.30) = γ (β) exp(γ (β)x)P ⊗2 |τ ∩ [1, n − 1]| ≥ x - An dx. 0
We split the right-hand side in three parts corresponding to the three different bounds we have in (5.28): x ∈ [0, k1 ], x ∈ [k1 , k0 ] and x ∈ [k0 , n]. It suffices to show that each part is less than ε/3 to finish the proof. The first part is 9
k1
γ (β)
- exp(γ (β)x)P ⊗2 |τ ∩ [1, n−1]| ≥ x - An dx ≤ γ (β)k1 exp(γ (β)k1 ).
0
(5.31) ε One uses that n ≤ β 4 | clog and γ (β) = β 2 + o(β 2 ) to get that for β small enough and β| n large enough if cε is well chosen we have
5 k1 γ (β) ≤ 100β 2 n log n ≤ ε/4,
(5.32)
so that γ (β)k1 exp(γ (β)k1 ) ≤ ε/3. We use our bound for the second part of the integral to get 9 γ (β)
k0
- exp(γ (β)x)P ⊗2 |τ ∩ [1, n − 1]| ≥ x - An dx
k1
9
∞
≤ γ (β)
9 1 2 2 exp γ (β)x − x /(64n) dx =
0
∞
0
x2 exp x − 64nγ (β)2
(5.33) dx.
Replacing n by its value, we see that the term that goes with x 2 in the exponential can be made arbitrarily large, provided that cε is small enough. In particular we can make the left-hand side less than ε/3. Finally, we estimate the last part 9 n - γ (β) exp(γ (β)2 x)P ⊗2 |τ ∩ [1, n − 1]| ≥ x - An dx k0 (5.34) 9 n exp(γ (β)x − nx0 /8) dx = n exp(−[γ (β) − x0 /8]n). ≤ γ (β) 0
This is clearly less than ε/3 if n is large and β is small.
498
H. Lacoin
6. Proof of the Lower Bound of Theorem 1.5 In this section we use the method of replica-coupling that is used for the disordered pinning model in [24] to derive a lower bound on the free energy. The proof here is an adaptation of the argument used there to prove disorder irrelevance. The main idea is the following: Let W N (β) denote the renormalized partition function for inverse temperature β. A simple Gaussian computation gives √ N dQ log W N ( t) -1 ⊗2 P = − 1{S (1) =S (2) } , i i dt 2 t=0
(6.1)
i=1
where S (1) and S (2) are two independent random walks under the law P ⊗2 . This implies that for small values of β (by the equality of derivative at t = 0), N ⊗2 2 Q log W N (β) ≈ − log P exp β /2 1{S (1) =S (2) } . (6.2) N
i=1
This tends to make us believe that
p(β) = − lim log P
⊗2
N →∞
exp β /2 2
N i=1
N
1{S (1) =S (2) } . N
(6.3)
N
However, things are not that simple because (6.2) is only valid for fixed N , and one needs some more work to get something valid when N tends to infinity. The proofs aim to use a convexity argument and simple inequalities to be able to get the inequality N p(β) ≥ − lim log P ⊗2 exp 2β 2 1{S (1) =S (2) } . (6.4) N →∞
i=1
N
N
The fact that convexity is used in a crucial way makes it quite hopeless to get the other inequality using this method. Proof. Let us define for β fixed and t ∈ [0, 1], N √ 1 tβ 2 N (t, β) := Q log P exp tβωi,Si − , N 2
(6.5)
i=1
and for λ ≥ 0,
N " !√ 1 Q log P ⊗2 exp tβ(ωi,S (1) +ωi,S (2) )−tβ 2 +λβ 2 1{S (1) =S (2) } . N (t, λ, β) := i i i i 2N i=1
(6.6) One can notice that N (0, β) = 0 and N (1, β) = p N (β) (recall the definition of p N (1.11)), so that N is an interpolation function. Via the Gaussian integration by par formula, Qω f (ω) = Q f (ω),
(6.7)
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
499
valid (if ω is a centered standard Gaussian variable) for every differentiable functions such that lim|x|→∞ exp(−x 2 /2) f (x) = 0, one finds 1 !√ "2 ⎛ ⎞2 2 N N 1{ S j =z } tβωi,Si − tβ2 i=1 d β 2 ⎝ P exp ⎠ "2 1 !√ N (t, β) = − Q tβ 2 N dt 2N P exp tβω − i,S j=1 z∈Z i i=1 2 N 1 2 2 √ ⊗2 β Q µ(n tβ) =− 1{S (1) =S (2) } . (6.8) i i 2N i=1
This is (up to the negative multiplicative constant −β 2 /2) the expected overlap fraction of two independent √ replicas of the random–walk under the polymer measure for the inverse temperature tβ. This result has been using the Itô formula in [4, Sect. 7]. For notational convenience, we define H N (t, λ, S (1) , S (2) ) =
N √
tβ(ωi,S (1) + ωi,S (2) ) − tβ 2 + λβ 2 1( i
i=1
i
(1) (2) Si =Si
)
.
(6.9)
We use Gaussian integration by part again, for N : H N (t, λ, S (1) , S (2) ) 1{S (1) =S (2) } j j ⊗2 exp H (t, λ, S (1) , S (2) ) P N j=1
⎛ ⎞2 ⊗2 1 (1) , S (2) ) + 1 (t, λ, S exp H P (1) (2) N N {S j =z} {S j =z} ⎟ β2 ⎜ ⎟ − Q⎜ ⎠ ⎝ ⊗2 (1) (2) 4N P exp H N (t, λ, S , S ) j=1
N P β2 d Q N (t, λ, β) = dt 2N
⊗2 exp
z∈Z
N P β2 ≤ Q 2N
H N (t, λ, S (1) , S (2) ) 1{S (1) =S (2) } j j P ⊗2 exp H N (t, λ, S (1) , S (2) )
⊗2 exp
j=1
d = N (t, λ, β). dλ
(6.10)
The above implies that for every t ∈ [0, 1] and λ ≥ 0, N (t, λ, β) ≤ N (0, λ + t, β).
(6.11)
Comparing (6.8) and (6.10), and using convexity and monotonicity of N (t, λ, β) with respect to λ, and the fact that N (t, 0, β) = N (t, β) one gets d d − φ N (t, β) = N (t, λ, β)-dt dλ λ=0 N (t, 2 − t, β) − N (t, β) ≤ N (0, 2, β) − N (t, β), ≤ (6.12) 2−t where in the last inequality we used (2 − t) ≥ 1 and (6.11). Integrating this inequality between 0 and 1 and recalling N (1, β) = p N (β) we get p N (β) ≥ (1 − e) N (0, 2, β).
(6.13)
500
H. Lacoin
On the right-hand side of the above we recognize something related to pinning models. More precisely N (0, 2, β) = where
YN = P
⊗2
exp 2β
2
1 log Y N , 2N N
(6.14)
1(
i=1
(1)
(2)
S N =S N
)
(6.15)
is the partition function of a homogeneous pinning system of size N and parameter 2β 2 with underlying renewal process the sets of zero of the random walk S (2) − S (1) . This is a well known result in the study of the pinning model (we refer to [13, Sect. 1.2] for an overview and the results we cite here) that lim
N →∞
1 log Y N = f(2β 2 ), N
(6.16)
where f denotes the free energy of the pinning model. Moreover, it is also stated h→0+
f(h) ∼ h 2 /2.
(6.17)
Then passing to the limit in (6.14) ends the proof of the result for any constant strictly bigger than 4. 7. Proof the Lower Bound in Theorem 1.6 The technique used in the two previous sections could be adapted here to prove the results but in fact it is not necessary. Because of the nature of the bound we want to prove in dimension 2 (we do not really track the best possible constant in the exponential), it will be sufficient here to control the variance of Wn up to some value, and then the concentration properties of log Wn to get the result. The reader can check than using the same method in dimension 1 does not give the right power of β. First we prove a technical result to control the variance of Wn which is the analog of (5.3) in dimension 1. Recall that γ (β) := λ(2β)−2λ(β) with λ(β) := log Q exp(βω(1,0) ). Lemma 7.1. For any ε < 0, one can find a constant cε > 0 and β0 > 0 such that for any β ≤ β0 , for any n ≤ exp cε /β 2 we have Var Q Wn ≤ ε.
(7.1)
Proof. A straightforward computation shows that the the variance of Wn is given by n ⊗2 1{S (1) =S (2) } − 1, Var Q Wn = P exp γ (β) (7.2) i=1
i
i
where S (i) , i = 1, 2 are two independent 2–dimensional random walks. As the above quantity is increasing in n, it will be enough to prove the result for n large. For technical convenience we choose to prove the result for n = exp(−cε /γ (β)) (recall γ (β) = λ(2β)−2λ(β)) which does not change the result since γ (β) = β 2 +o(β 2 ).
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
501
n The result we want to prove seems natural since we know that ( i=1 1{S (1) =S (2) } )/ log n i i converges to an exponential variable (see e.g. [12]), and γ (β) ∼ cε log n. However, convergence of the right–hand side of (7.2) requires the use of the Dominated Convergence Theorem, and the proof of the domination hypothesis is not straightforward. It could be extracted from the proof of the large deviation result in [12], however we include a full proof of convergence here for the sake of completeness. (1) (2) We define τ = (τk )k≥0 = {Si = Si , i ≥ 0} the set where the walks meet (it can be written as an increasing sequence). By the Markov property, the random variables τk+1 − τk are i.i.d. To prove the result, we compute bounds on the probability of having too many points before n in the renewal τ . As in the 1 dimensional case, we use the Laplace transform to do so. From [11, p. 577] , we know that 1 − P ⊗2 exp(−xτ1 ) =
1 (1) n∈N exp(−xn)P{Sn
(2)
= Sn }
.
(7.3)
The local central limit theorem says that for large n, 1 . πn Using this in (7.3) we get that when x is close to zero, π log P ⊗2 exp(−xτ1 ) ∼ − . | log x| P ⊗2 {Sn(1) = Sn(2) } ∼
(7.4)
(7.5)
We use the following estimate: P ⊗2 {|τ ∩ [1, n]| ≥ k} = P ⊗2 {τk ≤ n} ≤ exp(nx)P ⊗2 exp(−τk x) " ! = exp nx + k log P ⊗2 exp(−xτ1 ) .
(7.6)
Let x0 be such that for any x ≤ x0 , log P ⊗2 exp(−xτ1 ) ≥ −3/| log x|. For k such that k/(n log(n/k)) ≤ x0 , we replace x by k/(n log(n/k)) in (7.6) to get k 3k ⊗2 − P {|τ ∩ [1, n]| ≥ k} ≤ exp log(n/k) log k/(n log n/k)
k ≤ exp − , (7.7) log(n/k) where the last inequality holds if k/n is small enough. We fix k0 = δn for some small δ. We get that
k ⊗2 if k ≤ k0 , P {|τ ∩ [1, n]| ≥ k} ≤ exp − log(n/k)
(7.8) δn k0 ⊗2 = exp − if k ≥ k0 . P {|τ ∩ [1, n]| ≥ k} ≤ exp − log(n/k0 ) log(1/δ) We are ready to bound (7.2). We remark that using integration by part we obtain 9 n P exp (γ (β)|τ ∩ [1, n]|) − 1 = γ (β) exp(γ (β)x)P ⊗2 (τ ∩ [1, n]| ≥ x) dx. 0
(7.9)
502
H. Lacoin
To bound the right–hand side, we use the bounds we have concerning τ : (7.8). We have to split the integral in three parts. The integral between 0 and 1 can easily be made less than ε/3 by choosing β small. Using n ≤ exp(cε /γ (β)), we get that
9 δn 9 δn x dx γ (β) exp(γ (β)x)P ⊗2 (τ ∩ [1, n]| ≥ x) dx≤ γ (β) exp γ (β)x − log(n/x) 1 1
9 δn γ (β)βx ≤ γ (β) exp γ (β)x − cε 1 cε ≤ . (7.10) 1 − cε This is less than ε/3 if cε is chosen appropriately. The last part to bound is
9 n δn ≤ ε/3, γ (β) exp(γ (β)x)P ⊗2 (τ ∩ [1, n]| ≥ x)≤nγ (β) exp γ (β)n− log 1/δ δn (7.11) where the last inequality holds if n is large enough, and β is small enough.
Proof of the lower bound in Theorem 1.6. By a martingale method that one can find a constant c9 such that Var Q log Wn ≤ C9 n,
∀n ≥ 0, ∀β ≤ 1.
(See [6, Prop. 2.5] and its proof for more details). Therefore Chebyshev inequality gives -1 1 −1/4 Q - log Wn − Q log Wn - ≥ n ≤ C9 n −1/2 . n n
(7.12)
(7.13)
Using Lemma 7.1 and Chebyshev inequality again, we can find a constant C10 such that for small β and n = exp(C10 /β 2 ) we have Q {Wn < 1/2} ≤ 1/2.
(7.14)
This combined with (7.13) implies that − log 2 1 ≤ n −1/4 + Q log Wn ≤ n −1/4 + p(β). n n
(7.15)
Replacing n by its value we get p(β) ≥ −n −1/4 −
log 2 ≥ − exp(−C10 /5β 2 ). n
(7.16)
Acknowledgements. The author is very grateful to Giambattista Giacomin for numerous suggestions and precious help for the writing of this paper, to Francesco Caravenna for the proof of Lemma 5.2 and to Fabio Toninelli and Francis Comets for enlightening discussions. The author also acknowledges the support of ANR, grant POLINTBIO.
Free Energy of Directed Polymers in Dimension 1 + 1 and 1 + 2
503
References 1. Albeverio, S., Zhou, X.: A martingale approach to directed polymers in a random environment. J. Theor. Probab. 9, 171–189 (1996) 2. Bertin, P.: Free energy for Linear Stochastic Evolutions in dimension two. Preprint 2009 3. Bolthausen, E.: A note on diffusion of directed polymer in a random environment. Commun. Math. Phys. 123, 529–534 (1989) 4. Carmona, P., Hu, Y.: On the partition function of a directed polymer in a random Gaussian environment. Probab. Theor. Relat. Fields 124, 3, 431–457 (2002) 5. Carmona, P., Hu, Y.: Strong disorder implies strong localization for directed polymers in a random environment. ALEA 2, 217–229 (2006) 6. Comets, F., Shiga, T., Yoshida, N.: Directed Polymers in a random environment: strong disorder and path localization. Bernouilli 9(4), 705–723 (2003) 7. Comets, F., Shiga, T., Yoshida, N.: Probabilistic analysis of directed polymers in a random environment: a review. Adv. Stud. Pure Math. 39, 115–142 (2004) 8. Comets, F., Vargas, V.: Majorizing multiplicative cascades for directed polymers in random media. ALEA 2, 267–277 (2006) 9. Comets, F., Yoshida, N.: Directed polymers in a random environment are diffusive at weak disorder. Ann. Probab. 34(5), 1746–1770 (2006) 10. Derrida, B., Giacomin, G., Lacoin, H., Toninelli, F.L.: Fractional moment bounds and disorder relevance for pinning models. Commun. Math. Phys. 287, 867–887 (2009) 11. Feller W.: An Introduction to Probability Theory and Its Applications, Volume II. New York: John Wiley & Sons, Inc (1966) 12. Gantert, N., Zeitouni, O.: Large and moderate deviations for local time of a reccurent Markov chain on Z 2 . Ann. Inst. H. Poincaré Probab. Statist. 34, 687–704 (1998) 13. Giacomin, G.: Random Polymer Models. IC press, London: World Scientific, 2007 14. Giacomin, G., Lacoin, H., Toninelli, F.L.: Hierarchical pinning models, quadratic maps and quenched disorder. To appear in Probab. Theory. Rel. Fields, doi:10.1007/s00440-009-0205-y 15. Giacomin, G., Lacoin, H., Toninelli, F.L.: Marginal relevance of disorder for pinning models. to appear in Commun. Pure Appl. Math., doi:10.1002/cpa.20301 16. Giacomin, G., Lacoin, H., Toninelli, F.L.: Disorder relevance at marginality and critical point shift. http://arxiv.org/abs/0906.1942v1[math-ph], 2009 17. Grimmett, G.: Percolation. Second Edition, Grundlehren der Mathematischen Wissenschaften 321, Berlin: Springer-Verlag, 1999 18. Huse, D.A., Henley, C.L.: Pinning and roughening of domain wall in Ising systems due to random impurities. Phys. Rev. Lett. 54, 2708–2711 (1985) 19. Imbrie, J.Z., Spencer, T.: Diffusion of directed polymer in a random environment. J. Stat. Phys. 52, 3/4, 608–626 (1988) 20. Kaigh, W.D.: An invariance principle for random walk conditioned by a late return to zero. Ann. Probab. 4, 115–121 (1976) 21. Lacoin, H., Moreno, G.: Directed polymer on hierarchical lattice with site disorder. http://arxiv.org/abs/ 0906.0992v1[math.PR], 2009 22. Liu, Q., Watbled, F.: Exponential inequalities for martingales and asymptotic properties of the free energy of directed polymers in random environment. Stoch. Proc. Appl. 119, 10, 3101–3132 (2009) 23. Song, R., Zhou, X.Y.: A remark on diffusion of directed polymers in random environment. J. Stat. Phys. 85, 1/2, 277–289 (1996) 24. Toninelli, F.L.: A replica coupling-approach to disordered pining models. Commun. Math. Phys. 280, 389–401 (2008) 25. Toninelli, F.L.: Coarse graining, fractional moments and the critical slope of random copolymers. To appear in Electron. Journal Probab. 14, 531–547 (2009) 26. Vargas, V.: Strong localization and macroscopic atoms for directed polymer. Probab. Theor. Relat. Fields, 134, 3/4, 391–410 (2008) Communicated by F. L. Toninelli
Commun. Math. Phys. 294, 505–538 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0932-z
Communications in
Mathematical Physics
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1 Jens Mund Departamento de Física, Universidade Federal de Juiz de Fora, 36036-900 Juiz de Fora, MG, Brazil. E-mail:
[email protected] Received: 16 March 2009 / Accepted: 25 July 2009 Published online: 1 October 2009 – © Springer-Verlag 2009
Dedicated to the memory of Bernd Kuckert Abstract: We prove the Bisognano-Wichmann and CPT theorems for massive theories obeying braid group statistics in three-dimensional Minkowski space. We start from first principles of local relativistic quantum theory, assuming Poincaré covariance and asymptotic completeness. The particle masses must be isolated points in the mass spectra of the corresponding charged sectors, and may only be finitely degenerate. Introduction The Bisognano-Wichmann theorem states that a large class of models in relativistic quantum field theory satisfies modular covariance, namely: The modular unitary group [8] of the field algebra associated to a (Rindler) wedge region coincides with the unitary group representing the boosts which preserve the wedge. Since the boosts associated to all wedge regions generate the Poincaré group, modular covariance implies that the representation of the Poincaré group is encoded intrinsically in the field algebra. This has important consequences, most prominently the spin-statistics theorem, the particle/antiparticle symmetry and the CPT theorem [29,34]. Modular covariance also implies a maximality condition for the field algebra, namely the duality property [2], and it implies the Unruh effect [57], namely that for a uniformly accelerated observer the vacuum looks like a heat bath whose temperature is (acceleration)/2π . The original theorem of Bisognano and Wichmann [2,3] relied on the CPT theorem [45] and was valid for finite component Wightman fields. However, the physical significance of this latter hypothesis is unclear. In the framework of algebraic quantum field theory [1,31], Guido and Longo have derived modular covariance in complete generality for conformally covariant theories [9]. In the four-dimensional Poincaré covariant case the Bisognano-Wichmann theorem has been shown by the author [39] to hold under physically transparent conditions, namely for massive theories with asymptotic completeness. (Conditions of more Supported by FAPEMIG and CNPq.
506
J. Mund
technical nature have been found by several authors [5,7,30,35,54], see [7] for a review of these results.) In three-dimensional spacetime, however, there may be charged sectors with braid group statistics [19,24] containing particles whose spin is neither integer nor half-integer, which are called Plektons [23] or, if the statistics is described by an Abelian representation of the braid group, Anyons [59]. In this case, modular covariance and the CPT theorem are also expected [25, Assumption 4.1] to hold under certain conditions, but so far have not been proved in a model-independent way. The aim of the present article is to prove these theorems from first principles for massive Poincaré covariant Plektons satisfying asymptotic completeness. Let us comment in more detail on the Bisognano-Wichmann and CPT theorems, and their interrelation, in the familiar case of permutation group statistics. Let W1 be the standard wedge, W1 := { x ∈ R3 : |x 0 | < x 1 }.
(1)
The Tomita operator associated with the field algebra of W1 and the vacuum is defined as the closed anti-linear operator S satisfying S FΩ := F ∗ Ω,
F ∈ F(W1 ),
(2)
where F ∗ is the operator adjoint and F(W1 ) denotes the algebra of fields localized in W1 . Denoting its polar decomposition by S = J ∆1/2 , J and ∆it are called the modular conjugation and modular unitary group, respectively, associated with F(W1 ) and Ω. Modular covariance means that the modular unitary group coincides with the unitary group representing the boosts in 1-direction (which preserve the wedge W1 ), namely: ∆it = U (λ1 (−2π t)),
(3)
where λ1 (t) acts as cosh(t) 1 + sinh(t) σ1 on the coordinates x 0 , x 1 . Here, U is the ↑ representation of the universal covering group, P˜+ , of the Poincaré group under which the fields are covariant. (Note that then, by covariance, the modular groups associated to other wedges W = gW1 represent in the same way the corresponding boosts λW (t) = gλ1 (t)g −1 , and hence the entire representation U is fixed by the modular data.) The CPT theorem, on the other hand, asserts the existence of an anti-unitary CPT operator Θ which represents the reflexion1 j := diag(−1, −1, 1) at the edge of the standard wedge W1 in a geometrically correct way:2 ↑ Θ U (g)Θ ˜ −1 = U ( j g˜ j), g˜ ∈ P˜+ , AdΘ : F(C) → F( jC).
(4) (5)
Here, C is a spacetime region within a suitable class. Further, if a field F carries a certain charge then Θ FΘ −1 carries the conjugate charge.
(6)
The CPT theorem has been used as an input to the proof of modular covariance by Bisognano and Wichmann [2,3]. Conversely, the work of Guido and Longo [29], and 1 We consider j as the P T transformation. The total spacetime inversion arises in four-dimensional spacetime from j through a π -rotation about the 1-axis, and is thus also a symmetry. In the odd-dimensional case at hand, j is the proper candidate for a symmetry (in combination with charge conjugation), while the total spacetime inversion is not – in fact, the latter cannot be a symmetry in the presence of braid group statistics [25]. 2 In Eq. (4), j g˜ j denotes the unique lift [58] of the adjoint action of j from P ↑ to P˜ ↑ . + +
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
507
Kuckert [34], has shown that modular covariance implies the CPT theorem. In particular, Guido and Longo have shown [29] that modular covariance of the observable algebra A(W1 ) implies that the corresponding modular conjugation is a “PT” operator on the observable level, namely satisfies Eq. (4) on the vacuum Hilbert space and Eq. (5) with A(C) instead of F(C). Further, it intertwines a charged sector with its conjugate sector in the sense of representations, see Eqs. (63) and (65) below. Therefore, the modular conjugation can be considered a CPT operator. These results also hold in theories with braid group statistics. In the absence of braid group statistics, the CPT theorem can be made much more explicit [29,34], namely on the level of the field algebra. In fact, in this case the modular conjugation associated with F(W1 ), multiplied with the so-called twist operator, is a CPT operator in the sense of Eqs. (4), (5) and (6). In extending the Bisognano-Wichmann and CPT theorems to the case of braid group statistics, one encounters several difficulties. Since in this case there are no Wightman fields,3 the original proofs of the CPT and Bisognano-Wichmann theorems do not work. Also the proof of modular covariance in [39] and the derivation of the explicit CPT theorem on the level of the field algebra from modular covariance [29,34] do not go through, on two accounts. Firstly, the “fractional spin” representations of the universal covering group of the Poincaré group do not share certain analyticity properties of the (half-) integer spin representations which have been used in [39]. This problem has been settled in the article [40], whose results have been used to prove the spin-statistics theorem for Plektons [42]. Secondly, the derivations of modular covariance in [39] and of the CPT theorem in [29,34] rely on the existence of an algebra F of charge carrying field operators containing the observables A as the sub-algebra of invariants under a (global) gauge symmetry and such that the vacuum is cyclic and separating for the local field algebras. Such a frame, which we shall call the Wick-Wightman-Wigner (WWW) scenario, always exists in the case of permutation group statistics [17], but does not exist in the case of non-Abelian braid group statistics. Then it is not even clear what the proper candidate for the Tomita operator should be. We use here a “pseudo”-Tomita operator which has already been proposed by Fredenhagen, Rehren and Schroer [23]. A major problem then is that one cannot use the algebraic relations of the modular objects among themselves and with respect to the field algebra, and with the representers of the translations. These relations are asserted by Tomita-Takesaki’s and Borchers’ [4] theorems, respectively, and enter crucially into the derivation of modular covariance and the CPT theorem in [29,34,39]. This problem has been partially settled in [41], where the algebraic relations of our pseudo-modular objects among themselves and with the translations have been analyzed. In the present paper, we prove pseudo-modular covariance, namely that Eq. (3) holds with ∆it standing for the pseudo-modular unitary group. We also show that the pseudomodular conjugation already is a CPT operator in the sense of Eqs. (4), (5) and (6). Our line of argument parallels widely that of [39]. In the special case of Anyons, there does exist a WWW scenario and we show modular covariance in the usual sense. The article is organized as follows. In Sect. 1 we specify our framework and assumptions in some detail. As our field algebra we shall use the reduced field bundle [23]. This is a C ∗ -algebra F acting on a Hilbert space which contains, apart from the vacuum Hilbert space, subspaces corresponding to all charged sectors under consideration. It contains the observable algebra as the sub-algebra which leaves the vacuum Hilbert space invariant. However, in contrast to the field algebra in the permutation group statistics case it does not fulfill the WWW scenario. In particular the vacuum is not separating 3 However there might be, in models, string-localized Wightman type fields in the sense of [43,55].
508
J. Mund
for the local algebras, and worse: Every field operator F which carries non-trivial charge satisfies F ∗ Ω = 0. Correspondingly, there are no Tomita operators in the literal sense. This may be circumvented as proposed in [23]: There is a (non-involutive) pseudoadjoint F → F † on F which coincides with the operator adjoint only for observables. The point here is that FΩ → F † Ω, F ∈ F(W1 ), is a well-defined closable anti-linear operator. 4 We define now S ≡ J ∆1/2 as in Eq. (2), with F ∗ replaced by F † , and call S, J and ∆it the pseudo-Tomita operator, pseudo-modular conjugation and pseudo-modular unitary group, respectively, associated with F(W1 ) — We add the word “pseudo” because S is not a Tomita operator in the strict sense (it is not even an involution). In Sect. 2, we express the pseudo-Tomita operator S in terms of a family of relative Tomita operators [56] associated with the observable algebra A(W1 ) and certain suitably chosen pairs of states, and recall some algebraic properties of these objects established in [41]. Using these properties, we show that the pseudo-modular group associated with F(W1 ) leaves this algebra invariant (Proposition 1), just as in the case of a genuine modular group. In Sect. 3 we derive single particle versions of pseudo-modular covariance and the CPT theorem (Corollary 1) from our assumption that the theory be purely massive (A1). This was already partially implicit in [42]. We then prove, in Sect. 4, that this property passes over from the single particle states to scattering states. Under the assumption of asymptotic completeness (A2), this amounts to pseudo-modular covariance of the field algebra. This is our main result, stated in Theorem 1. Since the †-adjoint coincides with the operator adjoint on the observables, the restriction of S to the vacuum Hilbert space coincides with the (genuine) Tomita operator of the observables. We therefore have then modular covariance, in the usual sense, of the observables. As explained above, this also implies the CPT theorem on the level of observables. In Sect. 5 we make the CPT theorem explicit and show that the pseudo-modular conjugation of the field algebra F(W1 ) is a CPT operator in the sense of Eqs. (4), (5)5 and (6) (Theorem 2). To this end, we use the mentioned CPT theorem on the observable level [29], as well as the algebraic properties of the “pseudo”-modular objects established in [41]. (The argument used in Sect. 5 of [39] via scattering theory cannot be used since it relies on the fact that the modular conjugation maps the algebra onto its commutant, which does not hold in the present case.6 ) In Sect. 6 we finally treat the case of Anyons, where there is known to be a field algebra Fa in the WWW sense [38,48]. In particular, the vacuum is cyclic and separating for Fa (W1 ), allowing for the definition of a (genuine) Tomita operator associated with the wedge. Considering the genuine modular objects, we prove modular covariance (Theorem 3). We finally show that the modular conjugation, multiplied with an appropriate twist operator, is a CPT operator (Theorem 4). This extends the mentioned derivation of the CPT theorem in [29,34] from Bosons and Fermions to Anyons. To achieve these results, we exhibit the anyonic field algebra Fa as a sub-algebra of the reduced field bundle F, and show that the corresponding Tomita operator of Fa (W1 ) coincides with the pseudo-Tomita operator of F(W1 ) (Lemma 6).
4 To be precise, the “local” field algebras depend not only on spacetime regions such as W , but also on 1 certain paths in a sense to be specified in Sect. 1. In the definition of the pseudo-Tomita operator, W1 must therefore be replaced by a path W˜ 1 , see Eq. (22). 5 In Eq. (5), C is now understood to be a path of space-like cones as explained in Sect. 1. 6 On this occasion, I would like to rectify a minor error in the argument of [39, Sect. 5]. Namely, in Lemma 8 the modular conjugation, JW1 , must be replaced by Z ∗ JW1 , where Z is the twist operator, and twisted Haag duality for wedges [39, Eq. (1.4)] must be assumed. This does not influence the validity of its consequences, in particular of the CPT theorem (Proposition 9).
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
509
1. General Setting and Assumptions Since we are aiming at model-independent results, we shall use the general framework of algebraic quantum field theory [1,31], in which only the physical principles of locality, covariance and stability are required, and formulated in mathematical terms within a quantum theoretical setting. We now specify this setting and make our assumptions precise. Observable Algebra. The observables measurable in any given bounded spacetime region O are modelled as (the self-adjoint part of) a von Neumann algebra A0 (O) of operators, such that observables localized in causally separated regions commute. These operators act in a Hilbert space H0 which carries a continuous unitary representation U0 ↑ of the (proper orthochronuous) Poincaré group P+ which acts geometrically correctly: ↑
AdU0 (g) : A0 (O) → A0 (gO), g ∈ P+ .
(7)
(By Ad we denote the adjoint action of unitaries.) To comply with the principle of stability or positivity of the energy, the energy-momentum spectrum of U0 , namely the joint spectrum of the generators Pµ of the spacetime translations, is assumed to be contained in the forward light cone. The vacuum state corresponds to a unique (up to a factor) Poincaré invariant vector Ω0 ∈ H0 . It has the Reeh-Schlieder property, namely it is cyclic and separating for every A0 (O). Since the vacuum state should be pure, the net of observables is assumed irreducible, ∩O A0 (O) = C1. For technical reasons, we also require that the observable algebra satisfy Haag duality for space-like cones and wedges.7 Namely, denoting by K the class of space-like cones, their causal complements, and wedges, we require A0 (I ) = A0 (I ) ,
I ∈ K.
(8)
(A0 (I ) is defined as the von Neumann algebra generated by all A0 (O), O ⊂ I . The prime denotes the causal complement of a region on the left hand side, and the commutant of an algebra on the right hand side.) For the following discussion of charged sectors, it is convenient to enlarge the algebra of observables to the so-called universal algebra A generated by isomorphic images A(I ) of the A0 (I ), I ∈ K, see [20,23,28]. The family of isomorphisms A(I ) ∼ = A0 (I ) extends to a representation π0 of A, the vacuum representation. We then have A0 (I ) = π0 A(I ), and the vacuum representation is faithful and normal on the local8 algebras A(I ).9 The adjoint action (7) of the Poincaré group on the local algebras lifts to a representation by ↑ automorphisms αg of A, g ∈ P+ , which acts geometrically correctly: AdU0 (g) ◦ π0 = π0 ◦ αg , αg : A(I ) → A(g I ). 7 A space-like cone with apex a is a region in Minkowski space of the form C = a + R+ O, where O is a double cone causally separated from the origin. A wedge is a region which arises by a Poincaré transformation from W1 , see Eq. (1). 8 We call the algebras A(I ) “local” although the regions I extend to infinity in some direction, just in distinction from the “global” algebra A. 9 However, π is in general not faithful on the global algebra A due to the existence of global intertwin0 ers [23], see Footnote 10.
510
J. Mund
Charged Sectors. A superselection sector is an equivalence class of irreducible representations π of the algebra A0 of quasi-local observables, namely the C ∗ -algebra generated by all local observable algebras A0 (O). As a consequence of our Assumption A1, we shall deal only with representations which are localizable in space-like cones [12], i.e., equivalent to the vacuum representation when restricted to the causal complement of any space-like cone. Such representation uniquely lifts to a representation of the universal algebra A. If Haag duality (8) holds, it is equivalent [14,20] to a representation of the form π0 ◦ ρ acting in H0 , where ρ is an endomorphism of A localized in some specific region C0 ∈ K in the sense that ρ(A) = A
if A ∈ A(C0 ).
(9)
The endomorphism ρ is further transportable to other space-like cones, which means that for every space-like cone C1 and I ∈ K containing both C0 and C1 , there is a unitary U ∈ A(I ) such that AdU ◦ ρ is localized in C1 .10 We shall call localized and transportable endomorphisms simply localized morphisms. We further assume the representation π ∼ = π0 ρ to be covariant with positive energy. That means that there is a ↑ unitary representation Uρ of the universal covering group P˜+ of the Poincaré group with spectrum contained in the forward light cone such that ↑
AdUρ (g) ˜ ◦ π0 ρ = π0 ρ ◦ αg , g ∈ P+ ,
(10)
↑ where g˜ is any element of P˜+ mapped onto g by the covering projection. Superselection sectors, namely equivalence classes of localizable representations of A0 , are in one-toone correspondence with inner equivalence classes of localized morphisms of A. They are the objects of a category whose three crucial structural elements are products, conjugation and sub-representations. More specifically, products ρ1 ρ2 := ρ1 ◦ ρ2 of localized morphisms are again localized morphisms, leading to a composition of the corresponding sectors. A morphism ρ localized in C is said to contain another such morphism τ as a sub-representation if there is a non-zero observable T ∈ A, such that
ρ(A) T = T τ (A)
for all A ∈ A.
(If both ρ and τ are localized in a space-like cone C, then T ∈ A(C) by Haag duality.) An observable T satisfying this relation is called an intertwiner from τ to ρ. The set of all such intertwiners is denoted as Int(ρ|τ ). They are the arrows between the objects ρ and τ . Arrows can be composed if they fit together and have adjoints. Namely: If T ∈ Int(ρ|τ ) and S ∈ Int(τ |σ ) then T ◦ S ∈ Int(ρ|σ ); if T ∈ Int(ρ|τ ) then T ∗ ∈ Int(τ |ρ). It follows that if τ is irreducible (i.e., the representation π0 ◦ τ of A is irreducible), then Int(ρ|τ ) is a Hilbert space with scalar product T, S being fixed by T, S 1 := π0 (T ∗ S), T, S ∈ Int(ρ|τ ). There is also a product on the arrows, namely: if T ∈ Int(ρ|ρ) ˆ and S ∈ Int(σ |σˆ ) then T × S := T ρ(S) ˆ ≡ ρ(S) T
∈ Int(ρσ |ρˆ σˆ ).
(11)
10 In 2 + 1 dimensions, for every pair of causally separated space-like cones C , C there are two topo0 1 logically distinct ways to choose I ⊃ C0 ∪ C1 : Either one has to go clockwise from C0 to C1 within I , or anti-clockwise. This is the reason for the existence of global self-intertwiners in A which are in the kernel of the vacuum representation, and makes the enlargement from A0 to A necessary. It is also the reason for the occurrence of braid group statistics in 2 + 1 dimensions.
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
511
As a consequence of our Assumption A1, all morphisms considered here have finite statistics [18], i.e. the so-called statistics parameter λρ [14] is non-zero. This implies [15] the existence of a conjugate morphism ρ¯ characterized, up to equivalence, by the fact that the composite sector π0 ρρ ¯ contains the vacuum representation π0 precisely once. Thus there is a unique, up to a factor, intertwiner Rρ ∈ A(C0 ) satisfying ρρ(A)R ¯ ρ = Rρ A for all A ∈ A. The conjugate ρ¯ shares with ρ the properties of covariance (10), finite statistics, and localization (9) in some space-like cone which we choose to be C0 . Using the normalization convention of [15, Eq. (3.14)], namely Rρ∗ Rρ = |λρ |−1 1, the positive linear endomorphism φρ of A defined as ¯ φρ (A) = |λρ | Rρ∗ ρ(A)R ρ
(12)
is the unique left inverse [12,15] of ρ. In the present situation of three-dimensional space-time, the statistics parameter λρ may be a complex non-real number, corresponding to braid group statistics. We admit the case when its modulus is different from one (namely when ρ is not surjective), corresponding to non-Abelian braid group statistics. Field Algebra. From the observable algebra and a set of relevant sectors a field algebra can be constructed in various ways, see for example [23,25–27,37,49]. Some of these constructions have a quantum group or a more general structure playing the role of a global gauge group, however none of them fulfills the WWW scenario. Unfortunately, most constructions work only for models with a finite set of charges, whereas we wish to consider here an arbitrary (though countable) number of charges. We choose as field algebra the reduced field bundle proposed in [23], which in turn has been based on the field bundle of [15]. We start with a countable collection Σ of pairwise inequivalent localized, covariant, irreducible morphisms with finite statistics, one from each relevant sector, which is stable under conjugations and composition with subsequent reduction, and contains the identity morphism ι. We take all morphisms to be localized in the same space-like cone C0 , which we choose to be contained in W1 . The space of all relevant states is described by the Hilbert space Hρ , Hρ = H0 . (13) H := ρ∈Σ
We shall denote elements of Hρ as (ρ, ψ). For each ρ ∈ Σ there is a unitary represen↑ tation Uρ of P˜+ acting in H0 . (For ρ = ι, we take Uι ≡ U0 .) This gives rise to the direct sum representation U on H, ˜ U (g) ˜ (ρ, ψ) := (ρ, Uρ (g)ψ).
(14)
The vacuum vector Ω := (ι, Ω0 ) ∈ Hι is invariant under this representation. The observables act in H via the direct sum of all relevant representations π0 ◦ ρ =: π0 ρ, π(A) (ρ, ψ) := (ρ, π0 ρ(A)ψ). The idea of a charge carrying field is that it should add a certain charge ρc to a given state ψ ∈ Hρs . But since the product morphism ρs ρc is, in general, not contained in
512
J. Mund
the chosen set of irreducible morphisms, the new state must be projected onto an irreducible sub-representation ρr ∈ Σ of ρs ρc . (The subscripts s, c, r stand for “source”, “charge” and “range”, respectively.) This idea is realized as follows [22,23]. Given any three ρs , ρc , ρr ∈ Σ such that ρs ρc contains ρr as a sub-representation, the corresponding intertwiner space Int(ρs ρc |ρr ) has a certain finite [23] dimension N . We choose an orthonormal basis, i.e., a collection Ti ∈ A(C0 ), i = 1, . . . , N , satisfying Ti∗ T j = δi j 1,
N
Ti Ti∗ = 1 ρs ρc ,
i=1
where 1 ρs ρc is the unit in the algebra Int(ρs ρc |ρs ρc ). Following [46], we shall call the multi-index e := (ρs , ρc , ρr , i) a “superselection channel” of type (ρs , ρc , ρr ), and denote any one of the Ti from above generically as Te ∈ Int(ρs ρc |ρr ). We shall also call s(e) := ρs , c(e) := ρc and r (e) := ρr the source, charge and range of e, respectively. If s(e) or c(e) = ι, we choose Te = 1. The charge carrying fields are now defined as follows. Given e of type (ρs , ρc , ρr ) and A ∈ A, F(e, A) is the operator in H defined by F(e, A) (ρ, ψ) := δρs ,ρ ρr , π0 Te∗ ρ(A) ψ . Heuristically, this describes the action of A in the background charge ρ, addition of the charge ρc and subsequent projection onto Hρr via the intertwiner Te∗ . The norm-closed linear span of all these operators, – F := {F(e, A), A ∈ A } , e
where the sum goes over all superselection channels e, is closed under multiplication and will be called the field algebra. It is in fact a C ∗ sub-algebra of B(H) [22]. It contains the (representation π of the) observable algebra, namely π(A) = F(e, A). e: c(e)=ι
Localization. Fields are localizable to the same extent to which the charges are localizable which they carry, namely in unbounded regions in the class K. In order to have any definite space-like commutation relations, the fields need to carry some supplementary information in addition to the localization region, due to the existence of global intertwiners (see Footnote 10). The possibility we choose is to consider paths in K starting from our fixed reference cone C0 .11 By a path in K from C0 to a region I ∈ K (or 11 Two other possibilities are: To introduce a reference space-like cone from which all allowed localization cones have to keep space-like separated (this cone playing the role of a “cut” in the context of multi-valued functions) [12]; or a cohomology theory of nets of operator algebras as introduced by Roberts [50–52].
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
513
“ending at” I ) we mean a finite sequence (I0 , . . . , In ) of regions in K with I0 = C0 , In = I , such that either Ik−1 ⊂ Ik or Ik−1 ⊃ Ik for k = 1, . . . , n. Given a path (C0 = I0 , I1 , . . . , In = I ) and a morphism ρ ∈ Σ there are unitaries Uk ∈ A(Ik−1 ∪ Ik ) such that ρk := Ad(Uk · · · U1 ) ◦ ρ is localized in Ik . We shall call U := Un · · · U1 a charge transporter for ρ along the path (I0 , . . . , In ). Now a field operator F(e, A) with c(e) = ρ, is said to be localized along a path in K ending at I if there is a charge transporter U for ρ along the path such that U A ∈ A(I ). This localization concept clearly depends only on the homotopy classes (in an obvious sense [23]) of paths. We shall denote the homotopy class of a path ending at I by I˜, and ˜ The field operators localized in a given path the set of all such (classes of) paths by K. I˜ generate a sub-algebra of F which we denote by F( I˜). The vacuum Ω is cyclic for the local fields, i.e. for any path I˜ there holds – F( I˜) Ω = H. (15) Note, however, that Ω is not separating for the local12 field algebras F( I˜), since every field with non-trivial source annihilates the vacuum. We now give an alternative description of K˜ in the spirit of the “string-localized” quantum fields proposed in [43], which will be useful in the sequel. It is based on the observation that a space-like cone C is characterized by its apex a ∈ R3 and the spacelike directions contained in C. Namely, let H be the manifold of space-like directions, H := {r ∈ R3 , r · r = −1}. The set of space-like directions contained in C is C H := (C − a) ∩ H , and there holds C = a + R+ C H . C H is in fact a double cone in H .13 A similar consideration holds for causal complements of space-like cones and wedge regions (except that the apex of a wedge is fixed only modulo translations along its edge). We can therefore identify regions in K with regions of the form {a} × I H
⊂
R3 × H,
(16)
where I H is a double cone, a causal complement thereof, or a wedge, in H . Let us denote the class of such regions by K H . Regions in K H are simply connected, whereas H itself has fundamental group Z. Thus the portion of the universal covering space of H over a region I H in K H consists of a countable infinity of copies (“sheets”) of I H . We shall generically denote such a sheet over I H by I˜ H , and we denote by K˜ H the class of such sheets. We identify the universal covering space H˜ of H with homotopy classes of paths in H starting at some fixed reference direction r0 , which we assume to be contained in the reference cone C0 . (A sheet I˜ H is canonically homeomorphic to I H , but contains 12 Again, we call the algebras F ( I˜) “local” just in distinction to the “global” algebra F . 13 This is so because the boundary of C − a consists of 4 (pieces of) light-like planes through the origin. The intersection of such a plane with H is a light-like geodesic in H [44, Proof of Prop. 28]. Thus, C H is
bounded by 4 light-like geodesics emanating from two time-like separated points, and therefore is a double cone in the two-dimensional spacetime H .
514
J. Mund
Fig. 1. The (classes of the) paths γ1 and γ2 lie in the same sheet, say C˜ H , over C H . The path γ3 lies in a different sheet, which is mapped by a 2π rotation onto C˜ H
in addition the information of a winding number distinguishing it from the other sheets over I H , see Fig. 1.) We now identify paths I˜ ∈ K˜ with regions of the form {a} × I˜ H
⊂
R3 × H˜ ,
(17)
as follows. Given a path (I0 = C0 , I1 , . . . , In = I ) in K, pick a path γ = γn ∗· · ·∗γ0 in H from r0 to some r contained in I , and points a0 , . . . , an in R3 with a0 = 0 and an = a = apex of I , such that γk (t) ∈ Ik − ak for t ∈ [0, 1], k = 0, . . . , n. Then we associate with (I0 , . . . , In ) the region (17), where I˜ H is the unique sheet over I H = (I − a) ∩ H which contains the homotopy class of γ . Different paths γ lead to the same sheet, and the sheet depends only on the “homotopy class” (in the sense of [23]) of (I0 , . . . , In ). Therefore the above prescription defines a one-to-one correspondence between K˜ and R3 × K˜ H , which shall be used to identify them. In this identification, the covering space aspect of K˜ shows up in “accumulated angles”, endowing K˜ with a partial order relation. Namely, given I˜i = {ai } × I˜iH with I1 and I2 space-like separated, we shall write I˜1 < I˜2 if for any [γi ] ∈ I˜iH there holds γ1 dθ < γ2 dθ , where dθ denotes the angle one-form in a fixed Lorentz frame. (This is well-defined since the last relation is independent of the representants of [γi ] and of the Lorentz frame.) ↑ Covariance. The adjoint action of the representation U of P˜+ leaves the field algebra invariant, more specifically [15]:
U (g) ˜ F(e, A)U (g) ˜ ∗ = (e, Yρ (g) ˜ αg (A)),
(18)
where ρ = s(e) is the charge of e, and the so-called cocycle Yρ (g) ˜ ∈ A is characterized by π0 Yρ (g) ˜ = Uρ (g)U ˜ 0 (g)∗ . (19) The adjoint action on the fields is geometrically correct, i.e., AdU (g) ˜ : F( I˜) → F(g· ˜ I˜). Here, g˜ · I˜ denotes the natural action of the universal covering of the Poincaré group ˜ defined as follows. Let g˜ = (x, λ˜ ), where x is a spacetime translation and λ˜ is on K,
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
515
Fig. 2. C˜ 1 and C˜ 2 have relative winding number N (C˜ 2 , C˜ 1 ) = −1 ↑ an element of the universal covering group L˜ + of the Lorentz group, projecting onto ↑ λ ∈ L + . Then ˜ I˜ H , ˜ {a} × I˜ H := {x + λa} × λ· (x, λ)· (20)
where λ˜ · I˜ H denotes the lift of the action of the Lorentz group on H to the respective universal covering spaces. The rotations about integer multiples of 2π do not act trivially, but rather coincide with the action of the fundamental group, Z, on the universal covering space of H . Namely, they change winding numbers, see Fig. 1. Related to this, we define the relative winding number N ( I˜2 , I˜1 ) of I˜2 w.r.t. I˜1 to be the unique integer n such that r˜ (2π n)· I˜1 < I˜2 < r˜ (2π(n + 1))· I˜1 , ↑ where r˜ (·) denotes the rotation subgroup in L˜ + . See Fig. 2 for an example. (Note that this number is independent of the choice of reference direction r0 .)
Pseudo-Adjoint. Let Fι be the Banach space generated by field operators F(e, A) ∈ F which have trivial source, s(e) = ι, i.e., which have e of the form (ι, ρ, ρ). For such e, we define an adjoint channel e¯ := (ι, ρ, ¯ ρ). ¯ Following [23], we define a pseudo-adjoint F → F † on the space Fι by ¯ ρ(A ¯ ∗ )Rρ ), e = (ι, ρ, ρ). F(e, A)† := F(e, This pseudo-adjoint does not coincide with the operator adjoint (with respect to which F is a C ∗ -algebra). In fact, it is not an involution, but rather satisfies (F † )† = χρ F
(21)
if F has charge ρ. The number χρ is a root of unity in a self-conjugate sector (ρ¯ ρ), intrinsic to the sector, while in all other sectors Rρ and Rρ¯ may be chosen so that χρ = 1 [23, Eq. (3.2)]. (Those self-conjugate sectors with χρ = −1 are called pseudoreal sectors.) The adjoint preserves localization, i.e. leaves invariant the spaces of local fields with trivial source, Fι ( I˜) := Fι ∩ F( I˜): † Fι ( I˜) = Fι ( I˜).
516
J. Mund
Finally, the adjoint is preserved by Poincaré transformations: † U (g)FU ˜ (g) ˜ ∗ = U (g)F ˜ † U (g) ˜ ∗ ↑
for all g˜ ∈ P˜+ and F ∈ Fι . Due to the faithfulness of π0 ρ on the local algebras A(I ), F † Ω = 0 implies F = 0 for F ∈ F( I˜). This allows for the definition of our pseudo-Tomita operator [23] S : FΩ → F † Ω,
F ∈ Fι (W˜ 1 ).
(22)
Here, W˜ 1 is a path ending at W1 which will be specified in Eq. (28) below. Braid Group Statistics. For every pair of localized morphisms ρ, σ in Σ there is a local unitary intertwiner ε(ρ, σ ) ∈ Int(σρ|ρσ ), the so-called statistics operator. The family of statistics operators satisfies the braid relations [23, Eq. (2.6)] and determines the statistics of fields, as follows. Let C1 and C2 be causally separated, and let C˜ i be paths ending at Ci with relative winding number N (C˜ 2 , C˜ 1 ) = n, and let F(e1 , A1 ) ∈ F(C˜ 1 ) and F(e2 , A2 ) ∈ F(C˜ 2 ) be two fields with superselection channels e1 of type (α, ρ1 , β) and e2 of type (β, ρ2 , γ ), where α, β, γ , ρi ∈ Σ. Then there holds the commutation relation[23, Prop. 5.9] F(e2 , A2 ) F(e1 , A1 ) = R(δ, e1 , e2 , n) F(eˆ1 , A1 ) F(eˆ2 , A2 ). (23) δ,i 1 ,i 2
Here eˆ1 = (δ, ρ1 , γ , i 1 ) and eˆ2 = (α, ρ2 , δ, i 2 ), and the sum goes over all morphisms δ which are contained in the product representation α ◦ ρ2 . The numbers R(·) are given by
ωα ωγ n R(δ, e1 , e2 , n) = (24) π0 Te∗2 Te∗1 α(ε(ρ2 , ρ1 ))Teˆ2 Teˆ1 . ωβ ωδ The vacuum expectation values (26) of these commutation relations are already determined by the statistics phases. The statistics parameter λρ and statistics phase ωρ of a sector [ρ] are defined by the relations φρ (ε(ρ, ρ)) = λρ 1,
ωρ =
λρ , |λρ |
(25)
respectively. (They depend only on the equivalence class of ρ.) Suppose now that C1 and C2 are causally separated, and C˜ 1 and C˜ 2 have relative winding number N (C˜ 2 , C˜ 1 ) = −1, see Fig. 2 for an example. Then for F1 = F(e, A1 ) and F2 = F(e, A2 ) ∈ Fι (C˜ i ) with e = (ι, ρ, ρ), (26) ( F2 Ω, F1 Ω ) = ωρ F1† Ω, F2† Ω holds, see e.g. [15, Eq. (6.5)] and [42, Lemma A.1]. Of course ωρ = ±1 corresponds to Bosons/ Fermions, while the generic case corresponds to braid group statistics. Note that the hypothesis on the relative winding number under which Eq. (26) holds is not symmetric in C˜ 1 and C˜ 2 . Without this condition, Eq. (26) would imply ωρ ωρ¯ = 1. But ωρ and ωρ¯ are known to coincide [25], hence Eq. (26) would be be self-consistent only for ωρ = ±1, excluding braid group statistics.
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
517
Particle Structure. Our only special requirement concerns the particle structure. Namely, we shall assume that the theory is purely massive, and that asymptotic completeness holds. By purely massive, we mean that the set Σ of relevant sectors is generated by a set of elementary charges which correspond to massive single particle representations of the observable algebra. Such representation is characterized by the fact that its mass spectrum14 contains a strictly positive eigenvalue (the mass of the corresponding particle type), isolated from the rest of the mass spectrum in its sector by a mass gap (implementing the idea that there are no massless particles in the model). We also assume that there are only finitely many particle types in a given sector with a given mass, and that these have the same spin. We thus make the Assumption A1 (Massive particle spectrum). There is a finite subset Σ (1) ⊂ Σ of morphisms corresponding to massive single particle representations, which generates Σ. (In other words, Σ is exhausted by composition and subsequent reduction of morphisms in Σ (1) .) For each ρ ∈ Σ (1) , the restriction of the representation Uρ to the eigenspace of the corresponding mass value15 m ρ is a finite multiple of an irreducible representation. (Note that Σ is countable but may be infinite.) It is gratifying that this assumption, together with Haag duality (8), implies that all relevant sectors really have as representatives localized morphisms [12] with finite statistics [18] as assumed in our framework. Our second assumption is that the theory can be completely interpreted in terms of (multi-) particle states: Assumption A2 (Asymptotic Completeness). The scattering states span the entire Hilbert space H. (We shall sketch in Sect. 4 the Haag-Ruelle construction of scattering states from single particle states in the setting of the reduced field bundle for Plektons.) 2. Algebraic Properties of the Pseudo-Modular Objects As a first step, we discuss algebraic properties of the pseudo-modular objects which are independent of our special Assumptions A1 and A2, and are analogous to properties of genuine modular objects. In particular, we show that the (adjoint action of the) pseudo-modular group leaves the field algebra of the wedge invariant, and point out that Borchers’ commutation relations between the modular objects and the translations also hold in the present case. To this end, we exhibit the pseudo-Tomita operator S as a family of relative Tomita operators [56] Sρ , ρ ∈ Σ, associated with the observable algebra and certain suitably chosen pairs of states. We then use results obtained in a recent article [41] by the author on the relative modular objects. Recall that S maps each Hρ to Hρ¯ , and therefore corresponds to a family of operators Sρ acting in H0 via S(ρ, ψ) =: (ρ, ¯ Sρ ψ).
(27)
14 By mass spectrum we mean the spectrum of the mass operator P P µ . µ 15 For simplicity, we shall assume that there is only one mass eigenvalue in each sector, but our results still
hold if no restriction is imposed on the number of (isolated) mass values in each sector.
518
J. Mund
In order to calculate each operator Sρ explicitly, we first have to specify the path W˜ 1 from C0 to W1 which enters in its definition (22). For simplicity, we shall take the reference cone C0 to be properly contained in W1 (i.e., its closure is contained in W1 ), and define W˜ 1 to be the (class of the) path W˜ 1 := (C0 , W1 ).
(28)
Then, F(e, A) is in Fι (W˜ 1 ) if and only if A is in A(W1 ) and e is of the form e = (0, ρ, ρ) for some ρ ∈ Σ, in which case Te = Rρ . Thus the definition (22) reads explicitly Sρ π0 (A)Ω0 = π0 [ρ(A ¯ ∗ )Rρ ]Ω0 ,
A ∈ A(W1 ).
In the special case ρ = ι, this is just the Tomita operator of the observables, which 1/2 we shall denote by S0 ≡ J0 ∆0 . In the general case ρ = ι, Sρ is the relative Tomita operator associated with the algebra A(W1 ) and the pair of states ω0 := ( Ω0 , π0 (·)Ω0 ) (the vacuum state) and the positive functional ϕρ := |λρ |−1 ω0 ◦ φρ = Rρ Ω0 , π0 ρ(·)R ¯ ρ Ω0 . 1/2
We denote again the polar decomposition by Sρ = Jρ ∆ρ and call ∆itρ and Jρ the relative modular group and conjugation, respectively. It is known in relative TomitaTakesaki theory [56] that the operator ∆itρ ∆−it 0 is in π0 A(W1 ) for t ∈ R, giving rise to a family of unitaries ∈ A(W1 ) (29) Z ρ (t) := π0−1 ∆itρ ∆−it 0 known as the Connes cocycle (Dϕρ : Dω0 )t with respect to the pair of weights ω0 and ϕρ . Starting from the Connes cocycle, the algebraic properties of the relative modular objects have been analyzed in [41]. It has been shown there that the relative modular unitary group ∆itρ and conjugation Jρ satisfy the implementation properties Ad∆itρ ◦ π0 ρ = π0 ρ ◦ σt , AdJρ ◦ π0 ρ = π0 ρ¯ ◦ α j
(30)
on A(W1 ) ∪ A(W1 ). Here, σt is the modular group associated with A(W1 ) and the vacuum state [8], and α j is an anti-isomorphism from A(W1 ) onto A(W1 ) and vice versa. These are characterized by the fact that Ad∆it0 ◦ π0 = π0 ◦ σt , AdJ0 ◦ π0 = π0 ◦ α j ,
(31)
holds on A(W1 ) ∪ A(W1 ). Furthermore, Borchers’ commutation relations have been shown [41] to hold between the relative modular objects and the translations, namely for all t ∈ R and x ∈ R3 there holds ∆itρ Uρ (x, 1) ∆−it ρ = Uρ (λ1 (−2π t)x, 1), Jρ Uρ (x, 1)
Jρ−1
= Uρ¯ ( j x, 1).
(32) (33)
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
519
Finally, the algebraic relations among the relative modular objects, also established in [41], are completely analogous to the case of the genuine ones: Jρ ∆itρ Jρ−1 = ∆itρ¯ ,
(34)
Jρ Jρ¯ = χρ 1,
(35)
where χρ are the factors defined after Eq. (21). Note that these equations (or Eq. (21)) imply that the pseudo-Tomita operator S of the field algebra is not an involution, but rather satisfies S 2 ⊂ χ := χρ E ρ . (36) ρ∈Σ
Using relative Tomita-Takesaki theory and the fact that the pseudo-modular objects are related to the relative ones by J (ρ, ψ) = (ρ, ¯ Jρ ψ), ∆it (ρ, ψ) = (ρ, ∆itρ ψ)
(37)
due to uniqueness of the polar decomposition, we shall now calculate the adjoint action of the pseudo-modular group on the fields localized in W˜ 1 . It turns out that this action leaves F(W˜ 1 ) invariant — a non-trivial fact which enters crucially in the calculation of ∆it on scattering states in Sect. 4. For completeness’ sake we also show that the action of the pseudo-modular group commutes with the pseudo-adjoint †. Proposition 1. The adjoint action of the pseudo-modular group leaves the field algebra F(W˜ 1 ) associated to the wedge invariant, and commutes with the pseudo-adjoint † on Fι (W˜ 1 ): ∆it F(W˜ 1 ) ∆−it = F(W˜ 1 ), (∆it F∆−it )† = ∆it F † ∆−it ,
(38) (39)
F ∈ Fι (W˜ 1 ). Specifically, if e is a superselection channel with c(e) = ρ, then ∆it F(e, A) ∆−it = F(e, Z ρ (t) σt0 (A)).
(40)
It is interesting to note the resemblance of Eqs. (29) and (40) with Eqs. (19) and (18), respectively. Proof. We shall use two facts about the Connes cocycle (29) established by Longo in the present context. Namely, on A(W1 ) ∪ A(W1 ) [36, Prop. 1.1] AdZ ρ (t) ◦ σt ◦ ρ = ρ ◦ σt holds, and for any intertwiner
T∗
(41)
from ρr to ρs ρc [36, Props. 1.3, 1.4],
T ∗ ρs (Z ρc (t)) Z ρs (t) = Z ρr (t) σt0 (T ∗ ) holds.
(42)
For the proof of the proposition, let A be in A(W1 ) and e be of type (ρs , ρc , ρr ). Then ∆it F(e, A) ∆−it (ρs , ψ) = ρr , ∆itρr π0 (Te∗ ρs (A))∆−it ρs ψ = ρr , π0 (Te∗ ρs (Z ρc ))∆itρs π0 ρs (A)∆−it ψ ρs ∗ 0 = ρr , π0 (Te ρs (Z ρc ))π0 ρs (σt (A))ψ = F(e, Z ρc (t) σt0 (A)) (ρs , ψ).
520
J. Mund
In the third equation we have used Eq. (30), and in the second one we have used that ∆itρr π0 (T ∗ ) = π0 (T ∗ ρs (Z ρc (t)) ∆itρs , which is a consequence of Eq. (42). This proves the explicit formula (40). Since Z ρc (t) and σt (A) both are in A(W1 ), this also shows invariance (38) of F(W˜ 1 ). To prove Eq. (39), let e = (ι, ρ, ρ) and e¯ := (ι, ρ, ¯ ρ). ¯ Then ¯ ρ[σ ¯ t0 (A∗ )Z ρ (t)∗ ]Rρ ), (∆it F(e, A) ∆−it )† = (e, ∆it F(e, A)† ∆−it = (e, ¯ Z ρ¯ (t)σt0 [ρ(A ¯ ∗ )Rρ ]). But Eqs. (41) and (42) imply that ∗ ρ[σ ¯ t0 (A)Z ρ (t)∗ ]Rρ = Z ρ¯ (t)σt0 ρ(A)Z ¯ ¯ ρ (t)∗ )Rρ = Z ρ¯ (t)σt0 [ρ(A)R ¯ ρ¯ (t) ρ(Z ρ ].
This shows Eq. (39) and completes the proof.
3. Modular Covariance and CPT Operator on the Single Particle Space As a first step, we prove single-particle versions of the Bisognano-Wichmann and the CPT theorems. For ρ in the set Σ (1) of single particle charges, let E ρ be the projection (1) from H onto Hρ , and E ρ the projection from H onto the eigenspaces of the mass operator in Hρ (corresponding to the isolated eigenvalues in the sector ρ). We denote by E (1) the sum of all E ρ(1) , where ρ runs through Σ (1) , and call the range of E (1) the single particle space. Borchers’ commutation relations (32), (33) imply that the pseudo-modular unitary group and the pseudo-modular conjugation commute with the mass operator. Hence the pseudo-Tomita operator S commutes with E (1) . Let us denote the corresponding restriction by S (1) := S E (1) . ↑ Similarly, the representation U ( P˜+ ) leaves E (1) H invariant, giving rise to the subrepresentation
U (1) (g) := U (g) E (1) , and one may ask if modular covariance holds on E (1) H. We show in this section that this is indeed the case, the line of argument being as follows. Let K denote the generator of the unitary group of 1-boosts, U (1) (λ1 (t)) = exp(it K ). We exhibit in Eq. (46) below an anti-unitary “CPT -operator” U (1) ( j) representing the reflexion j on E (1) H, and show that S (1) coincides with the “geometric” involution (1) Sgeo := U (1) ( j) e−π K
(43)
↑ up to a unitary operator which commutes with the representation U (1) of P˜+ . By uniqueness of the polar decomposition, this will imply modular covariance on E (1) H, namely U (1) (λ1 (−2π t)) ≡ exp(−2πit K ) = ∆it E (1) . ↑ We begin by exploiting our knowledge about U (1) ( P˜+ ). By assumption, for each (1) (1) (1) ρ ∈ Σ the sub-representation Uρ := E ρ U is equivalent to a finite number,
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
521
say n ρ , of copies of the irreducible “Wigner” representation of the universal covering of the Poincaré group with mass m ρ and a certain spin sρ ∈ R. Let us denote this representation by Uρw . It acts on the Hilbert space L 2 (Hm+ ρ , dµ) ⊗ Cn ρ , which consists of momentum space “wave functions” ψ : Hm+ ρ → Cn ρ living on the mass shell Hm+ ρ := { p ∈ R3 | p · p = m 2ρ , p0 > 0} and having finite norm w.r.t. the scalar product dµ( p) (ψ( p) , φ( p))Cnρ . (ψ, φ) = Hm+ ρ
The representation Uρw acts in this space as (see e.g. [40]) ˜ Uρw (a, λ˜ )ψ ( p) = eisΩ(λ, p) eia· p ψ(λ−1 p), ˜ p) ∈ R is the where λ is the Lorentz transformation onto which λ˜ projects, and Ω(λ, so-called Wigner rotation. To the representation Uρw an anti-unitary operator Uρw ( j) can be adjoined satisfying the representation properties Uρw ( j)2 = 1
and
Uρw ( j) Uρw (g) ˜ Uρw ( j) = Uρw ( j g˜ j)
(44)
↑ for all g˜ ∈ P˜+ . Namely, it is given by [40] w Uρ ( j)ψ ( p) := ψ(− j p),
where the overline denotes component-wise complex conjugation in Cn ρ . Uρw is then a ↑ representation of the group P˜+ which we identify with the semi-direct product of P˜+ and Z2 , the latter acting in the former via the unique lift [58] of the adjoint action of j on ↑ (1) P+ . Let us denote the unitary intertwiner between the representations Uρw and Uρ of ↑ P˜+ by Wρ . In other words, Wρ is an isometric isomorphism from L 2 (Hm+ , dµ) ⊗ Cn ρ ρ
onto E ρ(1) H satisfying
Uρ(1) (g) ˜ Wρ = Wρ Uρw (g) ˜
(45)
↑ for all g˜ ∈ P˜+ . It is known that the masses[18], spins [42] and degeneracies n ρ [42] coincide for ρ and ρ. ¯ Hence the ranges of Wρ∗ and Wρ∗¯ coincide, and Cρ := Wρ¯ Wρ∗ is
a unitary operator from E ρ(1) H onto E ρ(1) ¯ H, representing “charge conjugation”, which
intertwines the representations Uρ(1) and Uρ(1) ¯ . Therefore the “CPT ”-operator U (1) ( j) := Uρ(1) ( j), Uρ(1) ( j) := Wρ¯ Uρw ( j)Wρ∗ ≡ Cρ Wρ Uρw ( j)Wρ∗ , (46) ρ∈Σ (1)
not only conjugates the charge, but also represents the reflection j, namely satisfies U (1) ( j)2 = 1, U (1) ( j) U (1) (g) ˜ U (1) ( j)∗ = U (1) ( j g˜ j).
(47)
(1) We define now a closed anti-linear operator Sgeo in terms of the representation U (1) ( P˜+ ), as anticipated, by Eq. (43). Note that the group relation j λ˜ 1 (t) j = λ˜ 1 (t) implies that (1) (1) Sgeo is an involution: it leaves its domain invariant and satisfies (Sgeo )2 ⊂ 1.
522
J. Mund
Proposition 2. There is a unitary operator D on E (1) H commuting with the represen↑ tation U (1) of P˜+ and with each E ρ , such that (1) = D S (1) . Sgeo
(48)
Proof. Let C1 denote the class of space-like cones which have apex at the origin, contain the positive x 1 -axis and are contained in the wedge W1 , and have non-zero intersection with the time-zero hyper-surface, and let C˜1 be the set of (equivalence classes of) paths in ∞ (C) ˜ be the set of field operators K of the form (C0 , W1 , C) with C ∈ C1 . Let further Fι,ρ ˜ F ∈ Fι (C) with superselection channel e = (ι, ρ, ρ) and for which g˜ → αg˜ (F) are smooth functions. ∞ (C). ˜ Then, for fixed p ∈ Hm+ , the Lemma 1 [11,42]. Let C˜ ∈ C˜1 and F ∈ Fι,ρ ρ n ρ C -valued function (49) t → ψ(t, p) := Uρw (λ˜ 1 (t))Wρ∗ E ρ(1) FΩ ( p)
extends to an analytic function in the strip t ∈ R + i(0, π ), which is continuous and bounded on its closure. At t = iπ , it has the boundary value † ψ(t, − j p)|t=iπ = Dρ¯ Wρ∗¯ E ρ(1) (50) ¯ F Ω ( p), where Dρ¯ is an isometry in Cn ρ independent of C˜ and F.16 Proof. We show how this lemma follows from [42, Prop. 2], which in turn is based on the work of Buchholz and Epstein [11]. If C˜ ∈ C˜1 , then C contains a space-like cone C1 of the special class used in [11] and [42, see Eq. (16)]. Defining C2 := −C1 and choosing a path C˜ 2 ending at C2 and satisfying N (C˜ 2 , C˜ 1 ) = −1, all conditions on C˜ 1 and C˜ 2 used in [42] are satisfied. (In particular, the “dual” [42, Eq. (17)] of the difference cone C2 − C1 contains the negative x 1 -axis, as required in Eq. (36) of [42].) For i ∈ {1, 2} we ∞ (C ˜ i ), β = 1, . . . , n ρ , with F1,1 := F of the lemma, now pick n ρ operators Fi,β ∈ Fι,ρ such that the n ρ vectors Wρ∗ E ρ(1) Fi,β Ω ( p) ∈ Cn ρ are linearly independent for all p in some open set.17 Then Proposition 2 in [42] asserts that there is an isometric matrix Dρ¯ , independent of C˜ i and Fi,β , such that the assertion of our lemma holds for all F1,β . This completes the proof of the lemma. (1) We now reformulate the lemma in terms of the operators S (1) and Sgeo . By the lemma, (1) there is an operator Aρ on E ρ H with domain (1)
∞ ˜ D0,ρ := span E (1) Fι,ρ (C) Ω ˜ C˜1 C∈
16 On the left hand side of Eq. (50), one first analytically continues into t = iπ and then conjugates component-wise in Cn ρ . 17 If E (1) FΩ is non-zero, then this is possible due to the Reeh-Schlieder property. ρ
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
defined via
523
(1) Wρ∗ Aρ φ ( p) := Wρ∗ Uρ(1) (λ˜ 1 (t))φ ( p) t=iπ , φ ∈ D0,ρ .
(51)
Denoting by Dˆ ρ the multiplication operator with the matrix Dρ , Wρ∗ Dˆ ρ φ ( p) := Dρ Wρ∗ φ ( p), Eq. (50) of the lemma reads Uρ(1) ( j) Aρ ⊂ Dˆ ρ¯ S (1) E ρ .
(52)
(1) (Note that the domain D0,ρ of Aρ is contained in the domain of S (1) .) We wish to identify (1) Aρ with e−π K ρ , where K ρ is the generator of the one-parameter group Uρ (λ˜ 1 (t)). First, (1) relation (52) shows that Aρ is closable since S is. Now for φ in the dense domain of A∗ρ , (1)
ψ ∈ D0,ρ and f ∈ C0∞ (R), one finds (by a calculation as in [39, proof of Lemma 11]): φ, e−π K ρ f (K ρ )ψ = A∗ρ φ, f (K ρ )ψ . (1)
This implies that the span of vectors of the form f (K ρ )ψ as above, Dρ , is in the domain (1) ∗∗ −π K ρ . But of the closure A∗∗ ρ of Aρ , and that on Dρ the closure Aρ coincides with e (1) (1) Dρ is invariant under the unitary one-parameter group Uρ (λ˜ 1 (t)), because for each t there is some C˜ t ∈ C˜1 such that λ1 (t)C˜ ⊂ C˜ t . By standard arguments, the closure (1)
of Aρ then coincides with the operator e−π K ρ. In particular, D0,ρ is a core for e−π K ρ , (1) (1) ρ ∈ Σ (1) , hence ρ D0,ρ is a core for Sgeo . Therefore, relation (52) implies (1) ⊂ DS (1) , Sgeo
(53)
where D := ρ Dˆ ρ . By construction, D commutes with E ρ and with the representation U (1) , as claimed in the proposition. It remains to show the opposite inclusion “⊃” in Eq. (53). To this end, we refer to the opposite wedge W1 = r (π ) W1 . Let Sˆ := U (˜r (−π )) S U (˜r (π )), ˆ (1)
S := U (1) (˜r (−π )) S (1) U (1) (˜r (π )) ≡ Sˆ E (1) , (1) (1) Sˆgeo := U (1) (˜r (−π )) Sgeo U (1) (˜r (π )). We claim that the following sequence of relations holds true: (1) ∗ (1) S (1) ⊂ ω χ ( Sˆ (1) )∗ ⊂ ω D ∗ ( Sˆgeo ) = ω U (1) (˜r (−2π ))D ∗ Sgeo . (54) Here, ω := ρ ωρ E ρ , where ωρ are the statistics phases, and χ is the operator defined in Eq. (36). To see the first inclusion, note that Sˆ is the closure of the operator FΩ → F † Ω, F ∈ Fι (˜r (−π )W˜ 1 ). Therefore Eqs. (26) and (21) imply that for F1 ∈ Fι (W˜ 1 ) and F2 ∈ Fι (˜r (−π )W˜ 1 ), there holds ω χ ∗ Sˆ F2† Ω, F1 Ω = S F1 Ω, F2† Ω .
524
J. Mund
ˆ ∗ ≡ Sˆ ∗ χ ω∗ , which coincides with ωχ Sˆ ∗ because χρ¯ = χρ This implies S ⊂ (ωχ ∗ S) and ωρ¯ = ωρ , while Sˆ ∗ is anti-linear and maps E ρ H into E ρ¯ H. Since S commutes with E (1) due to Borchers’ commutation relations, this implies the first inclusion in (54). To show the second inclusion, we first note that relations (53) and (36) imply (1) 2 (1) (1) −1 (D ∗ Sgeo ) ⊂ (S (1) )2 ⊂ χ . Since Sˆgeo = ( Sˆgeo ) , this yields (1) (1) (1) D ∗ Sgeo = χ Sgeo D ≡ Sgeo Dχ (1) invariant), and (we have equality here instead of ⊂, since D leaves the domain of Sgeo (1) ˆ the same relation holds for Sgeo . Therefore, the adjoint of Eq. (53) yields (1) ∗ (1) ∗ ( Sˆ (1) )∗ ⊂ (D ∗ Sˆgeo ) ≡ χ ∗ D ∗ ( Sˆgeo ) ,
which implies the second inclusion in (54). As to the last equality in Eq. (54), note that the group relations j r˜ (ω) j = r˜ (−ω), r˜ (π )λ˜ 1 (t)˜r (−π ) = λ˜ 1 (−t) and j λ˜ 1 (t) j = λ˜ 1 (t) (1) ∗ (1) imply that ( Sˆgeo ) coincides with U (1) (˜r (−2π ))Sgeo . This completes the proof of the sequence of relations (54). Using relation (53) then yields (1) S (1) ⊂ ω U (1) (˜r (−2π )) D ∗ Sgeo ⊂ ω U (1) (˜r (−2π )) S (1) .
This implies firstly that ω = U (1) (˜r (2π )) (which is the spin-statistics theorem) and (1) secondly that S (1) ⊂ D ∗ Sgeo , and completes the proof of the proposition. Note that the proof shows that, although we use results from [42], the spin-statistics connection needs not be assumed but rather follows. By uniqueness of the polar decomposition, Eq. (48) of the proposition implies the equations 1
∆ 2 E (1) = e−π K E (1) ,
D J E (1) = U (1) ( j) E (1) .
↑ Since the unitary D commutes with U (1) ( P˜+ ), the above equations and Eq. (47) imply the single particle versions of the Bisognano-Wichmann and CPT theorems:
Corollary 1. Let Assumption A1 of Sect. 1 hold. Then i) Modular Covariance holds on the single particle space: ∆it E (1) = U (λ1 (−2π t)) E (1) .
(55)
↑ ii) J E (1) is a “CPT operator” on E (1) H, namely, for all g˜ ∈ P˜+ , holds
J U (g)J ˜ −1 E (1) = U ( j g˜ j) E (1) . holds. Note that J is not an involution, but rather satisfies J 2 = χ , cf. Eq. (35).
(56)
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
525
4. Modular Covariance on the Space of Scattering States We shall now show that modular covariance, which we have established on the single particle space, extends to multi-particle states via (Haag-Ruelle) scattering theory. HaagRuelle scattering theory, as developed in [32,33], associates a multi-particle state to n single particle vectors which are created from the vacuum by quasi-local field operators. It has been adapted, within the field bundle formulation, to the setting of algebraic quantum field theory in [14], to theories with topological charges (i.e., charges localized in space-like cones) in [12], and to theories with braid group statistics in [21]. Since we are not aware of an exposition of scattering theory within the reduced field bundle ˜ let framework, we shall give a brief such exposition here. For ρ ∈ Σ (1) and C˜ ∈ C, ˜ F = F(e, A) be a field operator in F(C) carrying charge ρ, which produces from the vacuum a single particle vector with non-zero probability in the sense that it satisfies E (1) Fι Ω = 0. Here we have written Fι := F (eι , A) eι := (ι, ρ, ρ)
if F = F(e, A), if e = (ρs , ρ, ρr , i).
The mentioned quasi-local creation operator is constructed from F as follows. Let f ∈ S(R3 ) be a Schwartz function whose Fourier transform f˜ has compact support contained in the open forward light cone V+ and intersects the energy momentum spectrum of the sector ρ only in the mass shell Hm+ ρ . Recall that the latter is assumed to be isolated from the rest of the energy momentum spectrum in the sector E ρ H. Let now f t (x) := (2π )−2 d3 p ei( p0 −ωρ ( p))t e−i p·x f˜( p), F( f t ) := d3 x f t (x) αx (F), where ωρ ( p) := ( p2 + m 2ρ )1/2 . For large |t|, the operator F( f t ) is essentially localized in C˜ + t Vρ ( f ), where Vρ ( f ) is the velocity support of f ,
p 0 ˜ Vρ ( f ) := , p = ( p , p) ∈ supp f . 1, (57) ωρ ( p) Namely, for any ε > 0, F( f t ) can be approximated by an operator Ftε ∈ F(C˜ + t V ε ), where V ε is an ε–neighborhood of Vρ ( f ), in the sense that Ftε − F( f t ) is of fast decrease in t [6,32]. Further, Fι ( f t ) creates from the vacuum a single particle vector Fι ( f t ) Ω = f˜(P) Fι Ω
∈ E ρ(1) H ,
(58)
which is independent of t, and whose velocity support is contained in that of f. (Here the velocity support of a single particle vector is defined as in Eq. (57), with the spectral support of ψ taking the role of supp f˜.) To construct an outgoing scattering state from n single particle vectors, pick n localization regions C˜ i , i = 1, . . . , n and compact sets Vi in velocity space, such that for suitable open neighborhoods Viε ⊂ R3 the regions C˜ i + t Viε are mutually space-like separated for large t. Next, choose Fi ∈ F(C˜ i ) with respective superselection channels ei suitably chosen such that the vector Fn · · · F1 Ω does not vanish from “algebraic” reasons, i.e. satisfying s(e1 ) = ι and s(ei ) = r (ei−1 ), i = 2, . . . , n. Choose further Schwartz functions f i as above with Vρi ( f i ) ⊂ Vi . Then the standard lemma of scattering theory, in the present context, asserts the following:
526
J. Mund
Lemma 2. The limit lim Fn ( f n,t ) · · · F1 ( f 1,t ) Ω =: (ψn × · · · × ψ1 )out
t→∞
(59)
exists and depends only on the single particle vectors ψi := (Fi )ι ( f i,t ) Ω, on the localization regions C˜ i and on the superselection channels ei .18 The limit vector is approached faster than any inverse power of t, and depends continuously on the single particle vectors ψi . Further, there holds (ψn × · · · × ψ1 )out = lim Fn ( f n,t ) (ψn−1 × · · · × ψ1 )out . t→∞
(60)
Proof. The space-like commutation relations (23) imply that, for k ∈ {2, . . . , n}, Fk ( f k,t ) · · · F1 ( f 1,t ) Ω Rk−1 · · · R1 Fk−1 ( f k−1,t ) · · · F1 ( f 1,t ) Fˆk ( f k,t ) Ω ρ2 ,...,ρk−1
(61) up to terms which are of fast decrease in t. Here, Ri = R(ρi , eki , ei , n ik ), where eki has the same charge as ek , the same source as ei , and range ρi , and n ik = N (C˜ k , C˜ i ). The sum goes over all ρi contained in the product c(ek )s(ei ), i = 1, . . . , k − 1,19 plus the internal indices of the eki . Further, Fˆk is the operator arising from Fk by substituting its superselection channel ek for ek1 ≡ (ι, ρ1 , ρ1 ) with ρ1 = c(ek ). From here the proof goes through as in [14]: Differentiating Fn ( f n,t ) · · ·F1 ( f 1,t ) Ω with respect to t yields d d a sum of terms of the form Fn ( f n,t ) · · · dt Fk ( f k,t ) · · · F1 ( f 1,t ) Ω. Now dt Fk ( f k,t ) is of the same form as Fk ( f k,t ), hence can be permuted to the right by Eq. (61) up to fast decreasing terms, where it annihilates the vacuum due to Eq. (58). This shows the fast convergence in (59). Let now G k be a field operator with the same localization C˜ k and superselection channel ek as Fk and such that (G k )ι (gk ) creates the same single particle vector ψk from the vacuum as (Fk )ι ( f k ). Then Eq. (61) still holds, with the same numbers Ri , when Fk ( f k,t ) is replaced by G k (gk,t ). This implies that the scattering state (59) only depends on the single particle states, localization regions and superselection channels. The continuous dependence on the single particle vectors follows from the local tensor product structure derived in [21, Thm. 3.2]. Equation (60) follows as in [32] from the facts that Fn−1 ( f n−1,t ) · · · F1 ( f t )Ω converges rapidly to (ψn−1 × · · · × ψ1 )out , while Fn ( f n,t ) increases at most like |t|3 . Let us denote by H(n) , n ≥ 2, the closed span of outgoing n-particle scattering states as in the lemma, and by H(out) the span of all (outgoing) particle states: H(out) := H(n) . n∈N0
Here H(0) is understood to be the span of the vacuum vector Ω and H(1) := E (1) H. Asymptotic completeness (our Assumption A2) means that H(out) coincides with H. Our main result is now 18 We omit the dependence on C˜ and e in our notation. i i 19 ρ does not appear in the sum because c(e )s(e ) ≡ c(e ) contains no other representation than ρ ≡ 1 1 1 k k c(ek ). In particular there is no sum if k = 2.
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
527
Theorem 1 (Covariance of the Pseudo-Modular Groups). Let Assumptions A1 and A2 of Sect. 1 hold. Then the field algebra satisfies covariance of the pseudo-modular groups, ∆it = U (λ1 (−2π t)).
(62)
(Of course, this is equivalent to ∆itρ = Uρ (λ1 (−2π t)), ρ ∈ Σ.) Proof. On the single particle space, the claim is our Corollary 1. For the scattering states, one proves Eq. (62) by induction over the particle number. Here, we can literally take over the proof from [39, Prop. 7] due to the last lemma and the results of Sect. 2, in particular Borchers’ commutation relations (32) and invariance of the wedge field algebra under the pseudo-modular group (38). Asymptotic completeness then implies that Eq. (62) holds on the entire Hilbert space H. Since for observables the pseudo-adjoint F † coincides with the operator adjoint, this theorem asserts in particular modular covariance of the observables and implies, as mentioned in the Introduction, the CPT theorem on the observable level. (Of course for these properties to hold true, asymptotic completeness on the vacuum sector is a sufficient condition, namely H(out) ∩ Hι = Hι .) 5. The CPT Theorem We now prove that the pseudo-modular conjugation J is a CPT operator on the level of the field algebra, namely satisfies Eqs. (4), (5) and (6). As mentioned in the Introduction, Guido and Longo have shown [29] that modular covariance of the observables, which we have established in Theorem 1, implies that the corresponding modular conjugation J0 is a “PT” operator on the observable level, namely satisfies Eq. (4) on the vacuum Hilbert space and Eq. (5) with A(C) instead of F(C). As a consequence, the anti-isomorphism α j : A(W1 ) → A(W1 ) implemented by J0 , cf. Eq. (31), extends to an anti-automorphism of the entire universal algebra A (still implemented by J0 ). In particular, we have Corollary 2 ([29]). The modular conjugation J0 of the observable algebra associated with the wedge W1 implements an anti-automorphism α j of the universal algebra A, AdJ0 ◦ π0 = π0 ◦ α j ,
(63)
which has the representation properties α j αg α j = α jg j , α 2j = ι, and acts geometrically correctly: α j : A(I ) → A( j I ), I ∈ K.
(64)
Guido and Longo also show that α j intertwines any localized morphism ρ ∈ Σ with its conjugate ρ¯ up to equivalence: ¯ α j ◦ ρ ◦ α j ρ.
(65)
(In fact, we exhibit an intertwiner establishing this equivalence in Lemma 4 below.) In order to make the CPT theorem more explicit on the level of the field algebra, we begin with establishing the representation and implementation properties of the relative modular conjugations Jρ .
528
J. Mund
Proposition 3. The relative modular conjugations Jρ , ρ ∈ Σ, represent the reflection j in the direct sum representation Uρ ⊕ Uρ¯ , and implement α j in the direct sum representation π0 ρ ⊕ π0 ρ¯ of A. Namely, there holds ↑
Jρ Uρ (g)J ˜ ρ−1 = Uρ¯ ( j g˜ j), g˜ ∈ P˜+ , AdJρ ◦ π0 ρ = π0 ρ¯ ◦ α j on A.
(66) (67)
Of course, Eq. (66) implies that also the pseudo-modular conjugation J represents the reflection, J U (g)J ˜ −1 = U ( j g˜ j). Proof. Equation (66) corresponds to Prop. 2.8 of Guido and Longo’s article [29], whose proof uses Borchers’ commutation relations, commutation of the modular unitary group with the modular conjugation, and an assertion [29, Thm. 1.1] about unitary representations of the universal covering group of S L(2, R). But the commutation relations also hold in the present setting, cf. Eqs. (32) through (35), and the universal covering group of ↑ S L(2, R) is just the homogeneous part of our P˜+ , hence the proof of Guido and Longo goes through in the present case. The implementing property (67) has been shown in [41, Prop. 1] to hold on A(W1 ). By Eq. (66) it extends to all wedges W = gW1 . Considering a space-like cone C, we note that C coincides with the intersection of the wedge regions containing C. Then Haag duality for wedges and for space-like cones implies that A(C) = A(W ). W ⊃C
(This can be seen by the arguments in the proof of Cor. 3.5 in [10].) This implies that Eq. (67) also holds for space-like cones, and further, again by Haag duality, for their causal complements. Thus Eq. (67) holds for all I ∈ K, and the proof is complete. Before calculating the adjoint action of J on the fields, we discuss the item of geometrical correctness (5). In order to formulate it, one needs an action of j on the set K˜ of paths of space-like cones, such that j · I˜ is a path which starts at C0 and ends at j I . Since j cannot be continuously transformed to the identity transformation, such action must be of the form j ·(C0 , I1 , . . . , In ) = (C0 , I, jC0 , j I1 , · · · , j In ), where I ∈ K contains both C0 and jC0 . Now our requirement that C0 be contained in W1 implies that C0 and jC0 are causally separated. Hence there are two topologically distinct regions I ± containing C0 ∪ jC0 , and the action of j on K˜ depends on this choice. The action would be canonical, however, if the reference cone C0 (which enters in the ˜ were invariant under j.20 Namely, then j would act canonically as definition of K) j ·(C0 , I1 , . . . , In ) := ( jC0 , j I1 , . . . , j In ).
(68)
We therefore wish to go over to such a reference cone. To this end we first convince ourselves that all relevant structure of the field algebra is preserved under such change. More precisely, we show that two field algebras constructed as in Sect. 1 from the same observables and sectors, but with different reference cones, are isomorphic in all structural elements mentioned in Sect. 1: 20 Again, there are two topologically distinct possibilities for the choice of C : It must contain either the 0 positive or the negative x 2 -direction. The difference between the two choices shows up, in our context, only in the action of the twist operator for Anyons, see Eq. (88). So far, we leave the choice of C0 unspecified.
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
529
Lemma 3. Let Σˆ be a collection of morphisms localized in a reference cone Cˆ 0 (instead ˜ F( ˜ Hˆ and Uˆ (g) ˆ K, ˆ I˜) for I˜ ∈ K, of C0 ), one per equivalence class, and let F, ˜ be defined ˆ ˜ ˜ ˜ ˆ as in Sect. 1 with Σ replaced by Σ. Then there is a bijection I → I from K onto Kˆ˜ and an isometric isomorphism W : H → Hˆ which implements a unitary equivalence F ∼ = Fˆ preserving the respective notions of localization and pseudo-adjoints, and intertwining the representations U and Uˆ as well as the vacua. In formulas: ˆ Iˆ˜), ˜ AdW : F( I˜) → F( I˜ ∈ K, † ∗ ∗ † W F W = (W F W ) , F ∈ Fι , ↑ ˜ g˜ ∈ P˜+ , W U (g) ˜ W ∗ = Uˆ (g), ˆ W Ω = Ω.
(69) (70) (71) (72)
Proof. We choose I0 ∈ K which contains both C0 and Cˆ 0 , and pick for each ρ ∈ Σ with corresponding ρˆ ∈ Σˆ a unitary intertwiner Wρ ∈ A(I0 ) such that ρˆ = AdWρ ◦ ρ. Note that ρˆ¯ coincides with the conjugate of ρ, ˆ since Σ and Σˆ contain exactly one morphism per class. As intertwiner from ι to ρ¯ˆ ρˆ we choose ˆ¯ ρ ) Wρ¯ Rρ . ¯ ρ ) Rρ ≡ ρ(W Rρˆ := Wρ¯ ρ(W For e = (ρs , ρc , ρr , i) let eˆ := (ρˆs , ρˆc , ρˆr , i) and Teˆ := (Wρs × Wρc ) Te Wρ∗r ≡ Wρs ρs (Wρc ) Te Wρ∗r = ρˆs (Wρc )Wρs Te Wρ∗r . If i runs through 1, . . . , dim Int(ρs ρc |ρr ), then the Teˆ are an orthonormal basis of Int(ρˆs ρˆc |ρˆr ). Let finally W : H → Hˆ be defined by W (ρ, ψ) := (ρ, ˆ π0 (Wρ )ψ). ˆ e, ˆ Wρ A), where ρ is the charge of e. Defining a Then there holds W F(e, A) W ∗ = F( bijection K˜ → Kˆ˜ by (C0 , I1 , . . . , , In ) = I˜ → Iˆ˜ = (Cˆ 0 , I0 , C0 , I1 , . . . , In ), ˆ Iˆ˜). Equation (70) is ˆ e, one also checks that F(e, A) ∈ F( I˜) if and only if F( ˆ A) ∈ F( easily verified, and the intertwiner relation (71) follows from [15, Lem. 2.2]. The lemma implies of course that the pseudo-modular objects of Fˆ have the same ↑ algebraic relations among themselves and with the representation Uˆ ( P˜+ ) as those of F. Namely, it implies that the new pseudo-Tomita operator Sˆ defined via ˆ Sˆ Fˆ Ωˆ := Fˆ † Ω,
ˆ Wˆ˜ 1 ), Fˆ ∈ F(
satisfies Sˆ = W S W ∗ , and we have the following Corollary 3. All commutation relations between the relative modular objects and the Poincaré transformations, namely Eqs. (34, 35, 62, 66), as well as the implementation property (67), also hold with ρ, ∆ρ , Jρ , Uρ replaced by ρ, ˆ ∆ˆ ρˆ , Jˆρˆ , Uρˆ , respectively. Here, Jˆρˆ and ∆ˆ ρˆ are defined from Sˆ ≡ Jˆ ∆ˆ 1/2 as in Eqs. (27), (37).
530
J. Mund
Proof. Only the implementation property (67) of the new conjugation Jˆρˆ remains to be shown. But this property follows from the fact that Wρ intertwines the morphisms ρˆ and ρ by construction, cf. the proof of Lemma 3. Due to this result, we may from now on assume that the reference cone C0 in which all morphisms ρ ∈ Σ are localized is invariant under j, jC0 = C0 , and that j acts on K˜ in a canonical way as in Eq. (68). We now wish to calculate the action of J on the field algebra, and to this end introduce “cocycle” type unitaries Vρ which implement the equivalence (65) of α j ρα j and ρ: ¯ Lemma 4. The unitary operator Jρ J0 is in A0 (C0 ). Its pre-image under the (faithful) restriction of π0 to A(C0 ) is an intertwiner from α j ρα j to ρ. ¯ In other words, the unitary Vρ := π0−1 (Jρ J0 )
(73)
AdVρ ◦ α j ρ α j = ρ. ¯
(74)
is in A(C0 ) and satisfies
Proof. The implementation properties (63) and (67) imply that Ad(Jρ J0 ) ◦ π0 α j ρα j = π0 ρ. ¯ The morphisms on the left and right hand sides are localized in C0 and jC0 , respectively. Hence by Haag duality, Jρ J0 is in A0 (I ) = π0 A(I ), whenever I contains C0 ∪ jC0 . The rest of the proof is straightforward. Our explicit formula for the adjoint action of J below uses the unitaries Vρ and relies on the following observation. Recall that for each superselection channel e = (ρs , ρc , ρr , i) we have chosen an intertwiner Te in Int(ρs ρc |ρr ). One easily verifies that T¯e := (Vρs × Vρc ) α j (Te ) Vρ∗r ≡ ρ¯s (Vρc )Vρs α j (Te )Vρ∗r is an intertwiner in Int(ρ¯s ρ¯c |ρ¯r ), and can therefore be expanded in terms of the basis {Te¯ }, where e¯ is of type (ρ¯s , ρ¯c , ρ¯r ). Therefore, ∗ T¯e = ce,e with ce,e (75) ¯ Te¯ , ¯ 1 := Te¯ T¯e , e¯
where the sum goes over all superselection channels e¯ of type (ρ¯s , ρ¯c , ρ¯r ). We are now prepared for the CPT Theorem. Theorem 2 (CPT Theorem.). The pseudo-modular conjugation J is a CPT operator in the sense of Eqs. (4), (5) and (6). Namely, it represents the reflection j in a geometrically correct way: ↑ J U (g)J ˜ −1 = U ( j g˜ j), g˜ ∈ P˜+ , ˜ AdJ : F( I˜) → F( j · I˜), I˜ ∈ K,
and conjugates charges. More explicitly, AdJ is given by J F(e, A) J −1 = c¯e,e ¯ Vρc α j (A)), ¯ F(e,
(76) (77)
(78)
e¯
where the sum goes over all superselection channels e¯ of type (ρ¯s , ρ¯c , ρ¯r ) if e is of type (ρs , ρc , ρr ).
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
531
It is interesting to note the resemblance of Eqs. (73) and (78) with Eqs. (19) and (18), respectively. Proof. The representation property (76) has been shown in Proposition 3. To prove Eq. (78), let e be of type (ρs , ρc , ρr ). We have J F(e, A)J −1 (ρ¯s , ψ) = ρ¯r , π0 Vρr α j (Te∗ )Vρ∗s ρ¯s α j (A) ψ ρ¯r , π0 Te¯∗ ρ¯s (Vρc α j (A) ψ c¯e,e = ¯ e¯
=
c¯e,e ¯ Vρc α j (A)) (ρ¯s , ψ). ¯ F(e,
e¯
In the second equality we have used that Vρr α j (Te∗ )Vρ∗s = (T¯e )∗ ρ¯s (Vρc ) by definition of T¯e , and substituted T¯e as in Eq. (75). This proves Eq. (78), and shows that AdJ is an anti-automorphism of the field algebra. In order to show that it acts geometrically ˜ let U = Un · · · U1 be a charge correctly, let I˜ = (I0 = C0 , I1 , . . . , In = I ) ∈ K, transporter for ρ along I˜, and let F(e, A) be in F( I˜) with c(e) = ρ. We define U¯ k := α j (Uk ), k = 1, . . . , n. Then U¯ := U¯ n · · · U¯ 1 Vρ∗ is a charge transporter for ρ¯ along j · I˜, since Ad(U¯ k · · · U¯ 1 Vρ∗ ) ◦ ρ¯ ≡ α j ◦ Ad(Uk · · · U1 ) ◦ ρ ◦ α j is localized in j Ik , k = 1, . . . , n. Further, U¯ Vρ α j (A) ≡ α j (U A) is in A( j I ) by Eq. (64), since by hypothesis U A ∈ A(I ). Hence F(e, ¯ Vρc α j (A)) is in F( j · I˜), and the proof is complete. 6. Anyons We now consider the case of Anyons, i.e., when all sectors correspond to automorphisms of the observable algebra. In this case, one can construct a field algebra Fa [38,47,53] in the sense of the WWW scenario, in particular the vacuum vector is cyclic and separating for the local field algebras. Thus, the Tomita operator associated with Fa (W˜ 1 ) and the vacuum is well-defined and one may ask whether modular covariance holds. We show that this is indeed the case. To this end, we exhibit Fa as a sub-algebra of the reduced field bundle F, and show that the pseudo-modular operator of F(W˜ 1 ) coincides with the (genuine) Tomita operator of Fa (W˜ 1 ). Then modular covariance is implied by the results from the previous sections, and the CPT theorem follows in analogy with the permutation group statistics case. We assume for simplicity that the set Σ (1) of elementary charges consists of only one automorphism, localized in C0 = jC0 . (All results are easily transferred to the case of finitely many elementary Abelian charges, i.e., Σ (1) finite.) Σ has then the structure of an Abelian group with one generator, i.e., Z N if there is a natural number N such that the N -fold product of the generating automorphism is equivalent to the identity, and Z otherwise. In the former case there may or may not be a representant whose N -fold product coincides with the identity, as Rehren has pointed out [47]. We wish to exclude the case with obstruction for simplicity, and therefore make the following
532
J. Mund
Assumption A3. The set of relevant sectors Σ is generated by one automorphism, γ . If there is a natural number N such that γ N := γ ◦ · · · ◦ γ is equivalent to the identity ι, then γ can be chosen such that γ N = ι. 6.1. The Field Algebra for Anyons. In the assumed absence of an obstruction, there is a field algebra Fa in the sense of the WWW scenario as mentioned in the Introduction: The dual group of Σ, in our case U (1) or Z N , acts as a global gauge group on the local algebras, singling out the local observables as invariants under this action. The vacuum vector is cyclic and separating for the local field algebras. Further, the local commutation relations (governed by an Abelian representation of the braid group) can be formulated in terms of twisted locality (A.2), as in the familiar case of Fermions [13].21 The field algebra Fa for Anyons [38,48] is constructed as follows. Due to our Assumption A3, the map γ q → q, q ∈ Z N or Z establishes an isomorphism of the groups Σ and Z N or Z, respectively. We shall identify Σ with Z N or Z by this isomorphism, and write q instead of γ q , Hq and E q instead of Hγ q and E γ q , and so on. We consider the same Hilbert space H and representation U of the universal covering group of the Poincaré group as before, see Eqs. (13) and (14). The dual Σˆ of the group Σ is called the gauge group and is represented on H via V (t) := exp(2πiqt) E q . (79) q∈Σ
The anyonic field algebra Fa is now the C ∗ -algebra generated by operators Fa (c, A), c ∈ Σ, A ∈ A, acting as Fa (c, A) : (q, ψ) → (q + c, πq (A)ψ). Here we have written πq := π0 ◦ γ q to save on notation. Clearly there holds Fa (c1 , A1 ) Fa (c2 , A2 ) = Fa (c1 + c2 , πc2 (A1 )A2 ), and therefore Fa coincides with the closed linear span of the operators Fa (c, A), A ∈ A, c ∈ Σ. Furthermore, the adjoint is given by Fa (c, A)∗ = Fa −c, γ −c (A∗ ) . (80) As in the case of the reduced field bundle, we call a field operator Fa (c, A) localized in I˜ ∈ K˜ if there is a charge transporter U for γ c along I˜ such that U A ∈ A(I ). The von Neumann algebra generated by these operators is denoted by Fa ( I˜). The adjoint action of the gauge group (79) leaves each local field algebra Fa ( I˜) invariant, the fixed point algebra being the direct sum of all πq (A(I )), hence isomorphic to A(I ). The space-like commutation relations have the following explicit form. Let F1 = Fa (c1 , A1 ) ∈ Fa ( I˜1 ) and F2 = Fa (c2 , A2 ) ∈ Fa ( I˜2 ), where I˜1 , I˜2 are causally separated and have relative winding number N ( I˜2 , I˜1 ) = n. Then F2 F1 = ωc1 ·c2 (2n+1) F1 F2 21 Some results on the anyonic field algebra in d = 2 + 1 are spread out over the literature, or have not been made very explicit in the accessible literature (see however [38]), namely: Local commutation relations under consideration of the relative winding numbers, twisted duality, and the Reeh-Schlieder property. We therefore collect them, with proofs, in the Appendix for the convenience of the reader.
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
533
holds,22 where ω ≡ ωγ denotes the statistics phase of the generating automorphism γ , see Eq. (26). This may be reformulated in the form of twisted locality, which by Haag duality (8) sharpens to twisted duality, as follows. For I˜1 , I˜2 causally separated with relative winding number N ( I˜2 , I˜1 ) = n, let Z ( I˜2 , I˜1 ) be the unitary operator in H defined by 1 q 2 (2n+1) Z ( I˜2 , I˜1 ) E q := ω 2 Eq , (81) where the root of ω may be chosen at will. (This “twist operator” has been first proposed in [53].) Lemma 5 (Twisted Haag Duality.). Let I˜, I˜ be (classes of) paths in K˜ ending at I and its causal complement I , respectively. Then there holds Z ( I˜, I˜ ) Fa ( I˜ ) Z ( I˜, I˜ )∗ = Fa ( I˜) .
(82)
(We give a proof of this lemma in the Appendix.) 6.2. Modular Covariance and CPT Theorem. We now show that the anyonic field algebra satisfies covariance of the modular groups and conjugations. Since the vacuum is cyclic and separating for the local algebras, see Lemma A.3, the Tomita operator associated with Fa (W˜ 1 ) is well-defined. Let us denote this operator and its polar decomposition by Sa = Ja ∆1/2 a . Theorem 3 (Modular Covariance.). Let Assumptions A1, A2 and A3 hold. Then the modular unitary group of the anyonic field algebra satisfies modular covariance, namely coincides with the representers of the 1-boosts: ∆ita = U (λ˜ 1 (−2π t)). Proof. The theorem is a simple consequence of our Theorem 1 and the following lemma. Lemma 6. The anyonic field algebra Fa is a sub-algebra of the reduced field bundle F, and the pseudo-Tomita operator S associated with F(W˜ 1 ) and Ω coincides with the Tomita operator Sa associated with Fa (W˜ 1 ) and Ω. (The fact that then S must be an involution, S 2 ⊂ 1, is no contradiction to Eq. (36), since if γ is self-conjugate our Assumption A3 implies that γ is real and not pseudo-real [47, Remark 2 after Lem. 4.5], hence χγ = 1.) Proof. Let us first set up the reduced field bundle in the special case at hand. We identify Σ with Z or Z N as before, and denote elements by q, s, c, r . A superselection channel e = (s, c, r ) has non-zero intertwiner Te ∈ Int(s + c|r ) only if r = s + c. In this case we choose Te = 1 and write F(s, c; A) instead of F(e, A). The field algebra F (alias reduced field bundle) is thus generated by the operators F(s, c; A) : (q, ψ) → δs,c (q + c, πq (A)ψ), 22 See Appendix.
(83)
534
J. Mund
s, c ∈ Σ, A ∈ A. Clearly, Fa (c, A) = s∈Σ F(s, c; A), and hence Fa is a sub-algebra of F. Further, the map23 µ : F → Fa defined by µ (F(s, c; A)) := Fa (c, A)
preserves localization, and for F ∈ Fι satisfies µ(F) Ω = F Ω and µ F (The last equation follows from F(0, c; A)† = F 0, −c; γ −c (A∗ )
†
(84) = µ(F)∗ .
and Eq. (80).) These relations imply that for F ∈ Fι there holds S µ(F)Ω = S FΩ = F † Ω = µ(F † )Ω = µ(F)∗ Ω. But µ clearly maps F(W˜ 1 ) onto Fa (W˜ 1 ), and the proof is complete. The lemma also implies, of course, that the modular conjugation Ja associated with Fa (W˜ 1 ) represents the reflection j in the sense of Eq. (4) or (76). In order to achieve a geometrically correct action on the family of algebras Fa ( I˜) in the sense of Eq. (77), however, one has to multiply it with the twist operator Z , Z := Z (W˜ 1 , j · W˜ 1 ).
(85)
Theorem 4 (CPT Theorem for Anyons). The anti-unitary operator Θ := Z ∗ Ja is a CPT operator: It satisfies Θ 2 = 1,
ΘU (g)Θ ˜ ∗ = U ( j g˜ j),
(86)
↑ g˜ ∈ P˜+ , and acts geometrically correctly: For all I˜ ∈ K˜ there holds
AdΘ : Fa ( I˜) → Fa ( j · I˜).
(87)
Proof. The commutation relations Eq. (86) of Θ are inherited from those of Ja ≡ J , ↑ since Z commutes with U ( P˜+ ). Ja2 = 1, anti-linearity of Ja and Ja E q = E −q Ja imply 2 that Θ = 1. Tomita-Takesaki’s theorem and twisted duality (82) imply that Ja Fa (W˜ 1 ) Ja∗ = Fa (W˜ 1 ) = Z Fa ( j · W˜ 1 ) Z ∗ , which yields Eq. (87) for the case of W˜ 1 . Covariance then implies the geometric action for every I˜ ∈ K˜ which ends at a wedge region W , namely, every I˜ of the form g˜ · W˜ 1 . If I˜ ends at a space-like cone, note that twisted Haag duality (82) implies that Fa is self-dual, namely Fa (W˜ ), Fa ( I˜) = W˜ ⊃ I˜
where the intersection goes over all W˜ ∈ K˜ which contain I˜. (By I˜ ⊂ W˜ we mean that I ⊂ W and in addition I˜ H ⊂ W˜ H as subsets of H˜ .) Hence Eq. (87) also holds for such I˜. If I˜ ends at the causal complement of a space-like cone, the equation also holds, ˜ with C˜ ⊂ I˜. (This is so because the analogous since Fa ( I˜) is generated by all Fa (C) statement holds for A(I ).) This completes the proof of the theorem. 23 In fact, µ is a conditional expectation from F onto F . a
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
535
Up to here, the reference cone C0 has not been specified, and W˜ 1 may be any path from C0 to W1 . The difference in the choices only shows up in the value of the twist operator Z . Recall that there are two topologically distinct choices for C0 satisfying jC0 = C0 . With any one of these choices, the natural choice of W˜ 1 is the “shortest” path from C0 to W1 , namely W˜ 1 := (C0 , I, W1 ) for some I ∈ K. Specifying now C0 so as to contain the positive or negative x 2 -axis, respectively, the relative winding number of W˜ 1 and j · W˜ 1 is N (W˜ 1 , j · W˜ 1 ) = −1 or 0, respectively, hence 1 2
Z E q = ω∓ 2 q E q ,
(88)
respectively. A. The Anyon Field Algebra We collect some results on the anyonic field algebra which are spread out over the literature, or have not been made very explicit in the literature. Lemma A.1 (Anyonic Commutation Relations.). Let I˜1 , I˜2 be causally separated, with relative winding number N ( I˜2 , I˜1 ) = n. i) For F1 = Fa (c1 , A1 ) ∈ Fa ( I˜1 ) and F2 = Fa (c2 , A2 ) ∈ Fa ( I˜2 ) the commutation relations F2 F1 = ωc1 ·c2 (2n+1) F1 F2
(A.1)
hold, where ω denotes the statistics phase of the generating automorphism γ , see Eq. (26). ii) Equivalent with these relations is twisted locality, namely Z ( I˜2 , I˜1 ) Fa ( I˜1 ) Z ( I˜2 , I˜1 )∗ ⊂ Fa ( I˜2 )
(A.2)
if I1 and I2 are causally separated. Here, Z ( I˜2 , I˜1 ) is the “twist” operator defined in Eq. (81). Proof. Ad i) The commutation relations (23) satisfied by the reduced field operators read as follows in the present context of Anyons. Two fields F(s1 , c1 ; A1 ) ∈ F( I˜1 ) and F(s2 , c2 ; A2 ) ∈ F( I˜2 ), where s2 = s1 + c1 , which are causally separated satisfy the commutation relations F(s2 , c2 ; A2 ) F(s1 , c1 ; A1 ) = R(s1 , c1 , c2 ; n) F(ˆs1 , c1 ; A1 ) F(s1 , c2 ; A2 ),
(A.3)
where sˆ1 = s1 + c2 , and where n is the relative winding number N ( I˜2 , I˜1 ). The number R(s1 , c1 , c2 ; n) is given by, see Eq. (24),
ωα ωγ n π0 γ s1 ε(γ c2 , γ c1 ) (A.4) R(s1 , c1 , c2 ; n) = ωβ ωδ
536
J. Mund
with α = s1 , β = s2 ≡ s1 + c1 , γ = s2 + c2 ≡ s1 + c1 + c2 , δ = sˆ1 ≡ s1 + c2 . Now the statistics operator ε(γ c2 , γ c1 ) coincides, in the present situation, with ε(γ , γ )c1 ·c2 [23, Eq. (2.3)]. Further, ε(γ , γ ) coincides with a multiple ω1 of unity [14], where ω ≡ ωγ is the statistics phase of γ . Putting all this into Eq. (A.4) yields R(s1 , c1 , c2 ; n) = ωc1 c2 (2n+1) . Using this equality, the commutation relations (A.3) transfer to the fields Fa (ci , Ai ) ≡ s F(s, ci ; Ai ), proving the claim. Ad ii) For F1 , F2 as in (i) and Z := Z ( I˜2 , I˜1 ) one calculates 1 2 [F2 , Z F1 Z ∗ ] E q = ω 2 c1 +qc1 F2 F1 − ωc1 c2 F1 F2 E q . Hence the anyonic commutation relations (A.1) are equivalent with [F2 , Z F1 Z ∗ ] = 0. This implies that Z F1 Z ∗ also commutes with the von Neumann algebra generated by operators of the form F2 , and completes the proof. Lemma A.2 (Twisted Haag Duality.). Let I˜, I˜ be (classes of) paths in K˜ ending at I and its causal complement I , respectively. Then there holds Z ( I˜, I˜ ) Fa ( I˜ ) Z ( I˜, I˜ )∗ = Fa ( I˜) .
(A.5)
Proof. This follows from twisted locality and Haag duality of the observables as in the permutation group case [17, Thm. 5.4], the argument being as follows in the present setting. A standard argument [16, Remark 1 after Prop. 2.2] using the Reeh-Schlieder property and the action (79) of the gauge group implies that every operator B in Fa ( I˜) decomposes, just like a field operator, as the sum of operators Bq ∈ Fa ( I˜) carrying fixed charge, i.e. Bc E q = E q+c Bc . (Namely, Bc = Σˆ dt exp(−2πict)V (t)BV (t)∗ .) The same holds for Z ∗ Fa ( I˜) Z , where Z := Z ( I˜, I˜ ). Let now F ∈ Z ∗ Fa ( I˜) Z carry charge c, and pick a unitary Ψ ∈ Fa ( I˜ ) of the same charge. Then B := Ψ ∗ F also is in Z ∗ Fa ( I˜) Z by twisted locality, and carries charge zero. Therefore it is in Fa ( I˜) , and acts according to B (q, ψ) = (q, Bq ψ). One concludes that for every q ∈ Σ, charge transporter Uq for γ q along the path I˜ and observable A ∈ A(I ) there holds Bq π0 (Uq∗ A) = π0 (Uq∗ A) B0 . ˆ for some Bˆ ∈ A(I ). The Putting q = 0, this implies by Haag duality that B0 = π0 ( B) ∗ ˆ q ), which coincides with same equation (with A = 1) then implies that Bq = π0 (Uq BU ˆ ˆ πq ( B). Thus, B coincides with Fa (0, B) and is therefore in Fa ( I˜ ), and the same holds for F. This completes the proof. Lemma A.3 (Reeh-Schlieder Property.). The vacuum is cyclic and separating for every ˜ Fa ( I˜), I˜ ∈ K. Proof. Cyclicity of the vacuum Ω = (0, Ω0 ) for Fa ( I˜) follows from the cyclicity of Ω0 for A(I ) and the definition Fa (c, A)(0, Ω0 ) = (c, π0 (A)Ω0 ). Now by twisted locality (A.2), Fa ( I˜) Ω contains Z Fa ( I˜ )Ω, with Z unitary, which is dense. Hence the vacuum is cyclic for Fa ( I˜) , and therefore separating for Fa ( I˜). Acknowledgement. It is a pleasure for me to thank Klaus Fredenhagen, Daniele Guido, Roberto Longo and Bernd Kuckert, to whose memory this article is dedicated, for many stimulating discussions on the subject.
The CPT and Bisognano-Wichmann Theorems for Anyons and Plektons in d = 2 + 1
537
References 1. Araki, H.: Mathematical Theory of Quantum Fields. Int. Series of Monographs in Physics, no. 101, Oxford: Oxford University Press, 1999 2. Bisognano, J.J., Wichmann, E.H.: On the duality condition for a Hermitian scalar field. J. Math. Phys. 16, 985 (1975) 3. Bisognano, J.J., Wichmann, E.H.: On the duality condition for quantum fields. J. Math. Phys. 17, 303 (1976) 4. Borchers, H.J.: The CPT-theorem in two-dimensional theories of local observables. Commun. Math. Phys. 143, 315–332 (1992) 5. Borchers, H.J.: On Poincaré transformations and the modular group of the algebra associated with a wedge. Lett. Math. Phys. 46, 295–301 (1998) 6. Borchers, H.J., Buchholz, D., Schroer, B.: Polarization-free generators and the S-matrix. Commun. Math. Phys. 219, 125–140 (2001) 7. Borchers, H.J., Yngvason, J.: On the PCT-theorem in the theory of local observables. In: Mathematical Physics in Mathematics and Physics (Siena) R. Longo, ed., Fields Institute Communications, Vol. 30, Providence, RI: Amer. Math. Soc., 2001, pp. 39–64 8. Bratteli, O., Robinson, D.W.: Operator Algebras and Quantum Statistical Mechanics 1. Second ed., TMP, New York: Springer, 1987 9. Brunetti, R., Guido, D., Longo, R.: Modular structure and duality in conformal field theory. Commun. Math. Phys. 156, 201–219 (1993) 10. Brunetti, R., Guido, D., Longo, R.: Modular localization and Wigner particles. Rev. Math. Phys. 14, 759–786 (2002) 11. Buchholz, D., Epstein, H.: Spin and statistics of quantum topological charges. Fysica 17, 329–343 (1985) 12. Buchholz, D., Fredenhagen, K.: Locality and the structure of particle states. Commun. Math. Phys 84, 1–54 (1982) 13. Doplicher, S., Haag, R., Roberts, J.E.: Fields, observables and gauge transformations I. Commun. Math. Phys. 13, 1–23 (1969) 14. Doplicher, S., Haag, R., Roberts, J.E.: Local observables and particle statistics I. Commun. Math. Phys. 23, 199 (1971) 15. Doplicher, S., Haag, R., Roberts, J.E.: Local observables and particle statistics II. Commun. Math. Phys. 35, 49–85 (1974) 16. Doplicher, S., Roberts, J.E.: Fields, statistics and non-Abelian gauge groups. Commun. Math. Phys. 28, 331–348 (1972) 17. Doplicher, S., Roberts, J.E.: Why there is a field algebra with a compact gauge group describing the superselection structure in particle physics. Commun. Math. Phys. 131, 51–107 (1990) 18. Fredenhagen, K.: On the existence of antiparticles. Commun. Math. Phys. 79, 141–151 (1981) 19. Fredenhagen, K.: Structure of Superselection Sectors in Low Dimensional Quantum Field Theory. Proceedings (Lake Tahoe City), L.L. Chau, W. Nahm, eds., New York: Plenum, 1991 20. Fredenhagen, K.: Generalizations of the theory of superselection sectors. In: The Algebraic Theory of SuperSelection Sectors. Introduction and recent results. D. Kastler, ed., Singapore: World Scientific, 1990 21. Fredenhagen, K., Gaberdiel, M., Rüger, S.M.: Scattering states of plektons (particles with braid group statistics) in 2+1 dimensional field theory. Commun. Math. Phys. 175, 319–355 (1996) 22. Fredenhagen, K., Rehren, K.-H., Schroer, B.: Superselection sectors with braid group statistics and exchange algebras I: General theory. Commun. Math. Phys. 125, 201–226 (1989) 23. Fredenhagen, K., Rehren, K.-H., Schroer, B.: Superselection sectors with braid group statistics and exchange algebras II: Geometric aspects and conformal covariance. Rev. Math. Phys. SI1, 113–157 (1992) 24. Fröhlich, J., Marchetti, P.A.: Quantum field theories of vortices and anyons. Commun. Math. Phys. 121, 177–223 (1989) 25. Fröhlich, J., Marchetti, P.A.: Spin-statistics theorem and scattering in planar quantum field theories with braid statistics. Nucl. Phys. B 356, 533–573 (1991) 26. Frölich, J., Kerler, T.: Quantum Groups, Quantum Categories, and Quantum Field Theory. Lecture Notes in Mathematics, Vol. 1542, Berlin: Springer, 1993 27. Fuchs, J., Ganchev, A., Vecsernyés, P.: Rational Hopf algebras: Polynomial equations, gauge fixing, and low dimensional examples. Int. J. Mod. Phys. A 10, 3431–3476 (1995) 28. Guido, D., Longo, R.: Relativistic invariance and charge conjugation in quantum field theory. Commun. Math. Phys. 148, 521–551 (1992) 29. Guido, D., Longo, R.: An algebraic spin and statistics theorem. Commun. Math. Phys. 172, 517 (1995) 30. Guido, D., Longo, R.: Natural energy bounds in quantum thermodynamics. Commun. Math. Phys. 218, 513–536 (2001) 31. Haag, R.: Local Quantum Physics, Second ed., Texts and Monographs in Physics, Berlin-Heidelberg: Springer, 1996
538
J. Mund
32. Hepp, K.: On the connection between Wightman and LSZ quantum field theory, In: Axiomatic Field Theory, M. Chretien and S. Deser, eds., Brandeis University Summer Institute in Theoretical Physics 1965, Vol. 1, London-NewYork: Gordon and Breach, 1966, pp. 135–246 33. Jost, R.: The general theory of quantized fields. Providence, RI: Amer. Math. Soc., 1965 34. Kuckert, B.: A new approach to spin & statistics. Lett. Math. Phys. 35, 319–331 (1995) 35. Kuckert, B.: Two uniqueness results on the Unruh effect and on PCT-symmetry. Commun. Math. Phys. 221, 77–100 (2001) 36. Longo, R.: An analogue of the Kac-Wakimoto formula and black hole conditional entropy. Commun. Math. Phys. 186, 451–479 (1997) 37. Mack, G., Schomerus, V.: Conformal field algebras with quantum symmetry from the theory of superselection sectors. Commun. Math. Phys. 134, 139–196 (1990) 38. Mund, J.: Quantum Field Theory of Particles with Braid Group Statistics in 2+1 Dimensions. Ph.D. thesis, Freie Universität Berlin, 1998 39. Mund, J.: The Bisognano-Wichmann theorem for massive theories. Ann. H. Poinc. 2, 907–926 (2001) 40. Mund, J.: Modular localization of massive particles with “any” spin in d = 2 + 1. J. Math. Phys. 44, 2037– 2057 (2003) 41. Mund, J.: Borchers’ commutation relations for sectors with braid group statistics in low dimensions. Ann. H. Poinc. 10, 19–34 (2009) 42. Mund, J.: The spin statistics theorem for anyons and plektons in d = 2 + 1. Commun. Math. Phys. 286, 1159–1180 (2009) 43. Mund, J., Schroer, B., Yngvason, J.: String–localized quantum fields from Wigner representations. Phys. Lett. B 596, 156–162 (2004) 44. O’Neill, B.: Semi–Riemannian Geometry. New York: Academic Press, 1983 45. Pauli, W.: Exclusion principle, Lorentz group and reflection of space-time and charge, In: Niels Bohr and the Development of Physics, W. Pauli, ed., Oxford: Pergamon Press, 1955, p. 30 46. Rehren, K.-H.: Braid group statistics and their superselection rules. In: The Algebraic Theory of Superselection Sectors, D. Kastler, ed., Singapore: World Scientific, 1990 47. Rehren, K.-H.: Spacetime fields and exchange fields. Commun. Math. Phys. 132, 461–483 (1990) 48. Rehren, K.-H.: Field operators for anyons and plektons. Commun. Math. Phys. 145, 123 (1992) 49. Rehren, K.-H.: Weak C∗ Hopf symmetry. In: Group Theoretical Methods in Physics, A. Bohm, H.-D. Doebner, and P. Kielanowski, eds.), Lecture Notes in Physics, Vol. 504, Berlin: Heron Press, 1997, pp. 62–69 50. Roberts, J.E.: Local cohomology and superselection structure. Commun. Math. Phys. 51, 107–119 (1976) 51. Roberts, J.E.: Net cohomology and its applications to field theory. In: Quantum Fields – Algebras, Processes, L. Streit, ed., Wien-New York: Springer, 1980, pp. 239–268 52. Roberts, J.E.: Lectures on algebraic quantum field theory. In: The Algebraic Theory of Superselection Sectors. Introduction and Recent Results, D. Kastler, ed., Singapore-River Edge, NJ-London-Hong Kong: World Scientific, 1990, pp. 1–112 53. Schroer, B.: Modular theory and symmetry in QFT. In: Mathematical Physics towards the 21st Century, R.N. Sen and A. Gersten, eds., Beer-Sheva: Ben-Gurion of the Negev Press, Israel, 1994 54. Schroer, B., Wiesbrock, H.-W.: Modular theory and geometry. Rev. Math. Phys. 12, 139–158 (2000) 55. Steinmann, O.: A Jost-Schroer theorem for string fields. Commun. Math. Phys. 87, 259–264 (1982) 56. Strˇatilˇa, S.: Modular Theory in Operator Algebras, Tunbridge Wells: Abacus Press, 1981 57. Unruh, W.G.: Notes on black hole evaporation. Rev. Math. Phys. 14, 870–892 (1976) 58. Varadarajan, V.S.: Geometry of Quantum Theory, Vol. II, New York: Van Nostrand Reinhold Co., 1970 59. Wilczek, F.: Quantum mechanics of fractional-spin particles. Phys. Rev. Lett. 49, 957–1149 (1982) Communicated by Y. Kawahigashi
Commun. Math. Phys. 294, 539–579 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0961-7
Communications in
Mathematical Physics
The Dependence on the Monodromy Data of the Isomonodromic Tau Function M. Bertola1,2, 1 Department of Mathematics and Statistics, Concordia University,
1455 de Maisonneuve W., Montréal, Québec, Canada H3G 1M8
2 Centre de recherches mathématiques, Université de Montréal, Montréal,
Canada. E-mail:
[email protected];
[email protected] Received: 27 March 2009 / Accepted: 22 September 2009 Published online: 5 December 2009 – © Springer-Verlag 2009
Abstract: The isomonodromic tau function defined by Jimbo-Miwa-Ueno vanishes on the Malgrange’s divisor of generalized monodromy data for which a vector bundle is nontrivial, or, which is the same, a certain Riemann–Hilbert problem has no solution. In their original work, Jimbo, Miwa, Ueno provided an algebraic construction of its derivatives with respect to isomonodromic times. However the dependence on the (generalized) monodromy data (i.e. monodromy representation and Stokes’ parameters) was not derived. We fill the gap by providing a (simpler and more general) description in which all the parameters of the problem (monodromy-changing and monodromy-preserving) are dealt with at the same level. We thus provide variational formulæ for the isomonodromic tau function with respect to the (generalized) monodromy data. The construction applies more generally: given any (sufficiently well-behaved) family of Riemann–Hilbert problems (RHP) where the jump matrices depend arbitrarily on deformation parameters, we can construct a one-form (not necessarily closed) on the deformation space (Malgrange’s differential), defined off Malgrange’s divisor. We then introduce the notion of discrete Schlesinger transformation: it means that we allow the solution of the RHP to have poles (or zeros) at prescribed point(s). Even if is not closed, its difference evaluated along the original solution and the transformed one, is shown to be the logarithmic differential (on the deformation space) of a function. As a function of the position of the points of the Schlesinger transformation, it yields a natural generalization of the Sato formula for the Baker–Akhiezer vector even in the absence of a tau function, and it realizes the solution of the RHP as such BA vector. Some exemplifications in the setting of the Painlevé II equation and finite Töplitz/Hankel determinants are provided.
Work supported in part by the Natural Sciences and Engineering Research Council of Canada (NSERC).
540
M. Bertola
Contents 1. 2.
3. 4.
5. 6.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Riemann–Hilbert Problems . . . . . . . . . . . . . . . . . . . . . . . 2.1 Deformations and Malgrange’s form . . . . . . . . . . . . . . . . 2.2 Submanifolds of G where is closed . . . . . . . . . . . . . . . . 2.3 “Schlesinger” transformations . . . . . . . . . . . . . . . . . . . 2.3.1 Generalized Sato formula . . . . . . . . . . . . . . . . . . . . 2.3.2 Hirota bilinear relations . . . . . . . . . . . . . . . . . . . . 2.4 Right gauge equivalence . . . . . . . . . . . . . . . . . . . . . . Rational Differential Equations in Terms of Riemann–Hilbert Data . . 3.1 Monodromy map . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 Forward Birkhoff map. . . . . . . . . . . . . . . . . . . . . . Inverse Birkhoff Map: Riemann–Hilbert Problem . . . . . . . . . . . . 4.1 The set of contours . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 The jump-matrices . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Riemann–Hilbert problem . . . . . . . . . . . . . . . . . . . . . Tau Function and (Iso)monodromic Deformations . . . . . . . . . . . 5.1 Relationship between ω M and ω J MU . . . . . . . . . . . . . . . . Applications and Examples . . . . . . . . . . . . . . . . . . . . . . . 6.1 Painlevé II equation . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Variation of finite Toeplitz determinants for discontinuous symbols 6.3 Hankel and shifted Töplitz determinants . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
540 543 545 551 552 558 560 561 562 562 565 565 565 566 566 567 569 571 571 573 577
1. Introduction In the eighties Jimbo, Miwa and Ueno [9,7,8] derived a set of algebraic (in fact rational) nonlinear equations describing deformations of a rational connection on P1 which preserve the generalized monodromy data. They associated to this deformation a closed differential ω J MU on the space of deformation parameters, namely, on the space of ”isomonodromic times” which we denote collectively by t. In the simplest case of Fuchsian singularities,
(z) = A(z)(z), (∞) = 1,
A(z) =
K j=1
Aj , A j = 0, z − aj
(1.1)
the isomonodromic deformation equations were studied by Schlesinger [15] δ Ak = −
j=k
[Ak , A j ]
δ(ak − a j ) , ak − a j
δ :=
and the Jimbo-Miwa-Ueno differential reads δ(a j − ak ) 1 ω J MU = Tr(A j Ak ) . 2 a j − ak
da j ∂a j ,
(1.2)
(1.3)
j,k, j=k
a ) depend on the position of the poles It can be checked directly that if the matrices Ak ( as mandated by (1.2) then ω J MU above is a closed differential.
Dependence on Monodromy Data of the Isomonodromic Tau Function
541
This differential was generalized to an arbitrary (generic) rational connection in [9], to which we refer the reader for details. In the above situation for Schlesinger deformations, the locations of the poles constitute the “isomonodromic” parameters or times and we denote them by t. This was an important achievement because of the sweeping applications of isomonodromic deformation to integrable systems (solitons solutions to KP, solutions to Toda, etc.), Painlevé equations and, later, random matrix models. The (exponential) integral of this closed differential is the “isomonodromic tau function” τ J MU (t; m) =e
ω J MU
,
(1.4)
and the Painlevé property translates to the fact that τ J MU is a holomorphic function of the isomonodromic times that has only zeroes away from an explicit set of times where it has a branching behavior (the set of “diagonals” a j = ak , j = k in the case of (1.2)). In (1.4) we have indicated that the tau function depends necessarily on the (generalized) monodromy data, denoted generically by m; this dependence is parametric and the present paper addresses precisely the Question 1 (Naive). What is the dependence of τ J MU on the monodromy data? The question is conceptually simple but slightly ill-posed; since what JMU really defined was only a differential (in symbolic notation) ω J MU = f j (t; m) dt j , ∂t j f k = ∂tk f j , (1.5) j
the dependence of τ on m is defined only up to multiplication by an arbitrary function of the m’s only. So a better question would be Question 2 (Refined). What is the essential dependence of τ J MU on the monodromy data? Can we define an extended closed differential on the total phase space of the problem that coincides with ω J MU on the isomonodromic submanifold, namely f j (t; m) dt j + G ν (t, m) dm ν , (1.6) ωext = j
ν
∂m ν f k = ∂tk G ν , ∂m ν G µ = ∂m ν G µ ?
(1.7)
The question still admits many solutions as stated, since if we find one such extension we are still free to add any closed differential of the m’s alone. However we may and should understand the problem in a relative setting, where the answer is taken modulo closed forms of the m’s which must be holomorphic on the whole space of monodromy data. The reason for requiring holomorphicity is actually important because of the interpretation of the singularity locus of ω, as we presently explain. The meaning of the singularity locus of ω J MU . Malgrange (for Fucshian systems) [11] and later Palmer (for irregular singularities)[14] showed what the meaning of the zerolocus of τ J MU is: when τ J MU (t; m) = 0 then a vector bundle on P1 is nontrivial or –which is equivalent– a Riemann–Hilbert problem is not solvable. This is the equivalent of saying that ω J MU has only simple poles (with “residue” one) away from the non-movable singularity locus (Painlevé property). The divisor where the aforementioned bundle is nontrivial is generally termed Malgrange divisor (or simply Malgrange divisor). It is clear then that whatever extension (1.6) we are looking for, it ought to preserve the singularity locus, that is, the G ν (t, m) may have singularities only where some of the f k ’s has, and of the same type.
542
M. Bertola
So we arrive to the final formulation of a “sensible” problem, whose solution is the principal aim of the paper. Problem 1.1. Formulate a (“natural”) extended closed differential ωext (1.6) such that its tau function (locally defined up to nonzero multiplicative constants) τext (t, m) =e
ωext
(1.8)
vanishes precisely and only on the Malgrange divisor in the extended phase space of isomonodromic times t and monodromy data m. The differential we propose is in fact very natural (see Def. 2.2 and Thm. 5.1): it is the pull–back of the (partial integral of the) generator of the third cohomology of the “loop group” to the submanifold corresponding to our total phase space. In particular it expresses the same cohomology class c[ ] [11]. We will briefly indicate some interesting problems which can be addressed and that require the knowledge of the derivative of τ with respect to non-isomonodromic times (hence the knowledge of the G ν ’s). Organization of the paper. The heart of the paper is in fact Sect. 2 where we introduce (recall) the definition of the Malgrange differential ω M (Def. 2.1) associated to any (sufficiently well–behaved) Riemann–Hilbert problem. Such differential is ill–defined only when the RHP is not solvable: however its exterior differential is (admits a) smooth (extension) over the whole “loop group”, in particular also at the points where the RHP is not solvable [11]. The curvature of ω M is not zero, but it is so explicit that it is immediate to identify “simple” families of Riemann–Hilbert problems for which ω M is closed. By adding to ω M an explicit smooth differential (in particular this does not change its singularity locus) we obtain a new differential (Def. 2.2) whose curvature differs slightly from that of ω M but within the same cohomology class. In Sect. 2.3 we investigate the changes in (or ω M ) under modification of the growth conditions for the solution of the RHP (discrete “Schlesinger” transformations); this allows to interpret the matrix solution of any Riemann–Hilbert problem as a Baker– Akhiezer function via a Sato–like formula, even if a notion of tau-function is not available (see Thm. 2.1 and Sect. 2.3.1, Sect. 2.3.2). In the second part we specialize the setting from arbitrary Riemann–Hilbert problems to those that correspond to rational ODEs in the complex plane: Section 3 is a quick reminder to the reader of the classical description of the (generalized) monodromy map, i.e. how to associate to a rational ODE in the complex plane the set of “Birkhoff data” of irregular type, connection matrices, monodromy representation and Stokes’ multipliers. All the material is quite standard [17]. These Birkhoff data can be used viceversa (Sect. 4) to encode an ODE in a Riemann– Hilbert problem (which may or may not have a solution, although generically it does [16,17]). In Sect. 5 we show that is a closed differential in all the deformation parameters, which include • Monodromy representation; • Connection matrices; • Stokes’ matrices. In Sect. 5.1 we show that the restriction of to the submanifold of isomonodromic times coincides with the differential ω J MU defined in [9]. Therefore realizes the solution to our Problem 1.1. We conclude the paper with Sect. 6, where applications are provided to Painlevé II, (shifted) Toeplitz and Hankel determinants.
Dependence on Monodromy Data of the Isomonodromic Tau Function
543
2. Riemann–Hilbert Problems A Riemann–Hilbert problem consists of the data specified below: we will assume latitude in the smoothness class as this is not our primary focus. The Riemann–Hilbert data. 1. A finite collection of smooth oriented arcs γν , j = 1 . . . K , possibly meeting at a finite number of points but always in non-tangential way (Fig. 5 is a good example). We denote collectively these arcs by the symbol γ , γν . (2.1) γ = ν
If any of the contours γν extends to infinity then we shall assume that the contour has an asymptotic direction and that the corresponding jump matrix Mν (z) tends to the identity faster than any power Mν (z) = 1 + O(z −∞ ), z → ∞, with this decay valid in an open sector of nonzero angle around the direction of approach. 2. A collection of r × r matrices Mν (z), each of which is analytic at each interior point of its corresponding arc γν of a locally holomorphic function. We make the assumption that these matrices have unit determinant1 det Mν (z) ≡ 1. We will denote collectively by M(z) the matrix defined on γ that coincides with Mν (z) on γν , M : γ → S L r (C) Mν (z)χγν (z) z →
(2.2)
ν
where, for a set S, χ S denotes its indicator function. 3. At each point c where several arcs meet, denoting by γ1 , . . . , γ the arcs entering a suitably small disk at c, we impose that the jump matrices along its corresponding arc either tend to the identity matrix as O((z − c)∞ ) (faster than any power) in an open sector containing the direction of approach, or admits a local analytic extension within said disk. Remark 2.1. The last requirement allows for example the situation depicted on the right in Fig. 1; at the meeting point, c, of the circle and the “stem”, the jump matrices on each local arc admit an analytic continuation to a neighborhood of c. Note that the jump on the circle is actually a multivalued analytic function, but in a neighborhood of c each branch allows an analytic continuation. Remark 2.2. The condition for the jumps on contours that extend to infinity is perhaps overly restrictive and it is made to avoid case distinctions later on. For example, the RHP for the Airy functions (or for the linear problem associated to the Painlevé one transcendent) does not readily fall within this description. It would however be possible to reduce them to such type of situations with a sequence of transformations. Since the point is of rather technical nature, we leave it to be dealt with on a need-to basis. 1 This is not really necessary but simplifies some matters. In some situations (typically where the determinant is a rational function) it is necessary (but simple) to relax the condition and allow any matrices in G L r (C). We face the problem on a need-to basis.
544
M. Bertola
Fig. 1. The process of “cloaking” of a point of growth. The two RHPs are equivalent in the sense that one admits a solution if and only if the other does
The Riemann–Hilbert problem then consists in finding a holomorphic matrix on the complement of the contours (z) : C \ γ → S L r (C)
(2.3)
such that it admits (non-tangential) boundary values satisfying + (z) = − (z)M(z) z ∈ γ .
(2.4)
Here and throughout the paper we use the common convention that a subscript ± denotes the (nontangential) boundary value from the left (plus) or the right (minus) of the given oriented arc. The above data are insufficient to characterize uniquely the matrix (if it exists) and need to be supplemented by • growth behavior near the endpoints/intersections of the contours γν and at ∞; • an overall normalization. Therefore we will pose the following Problem 2.1 (RHP). Find a holomorphic matrix : C \ γ → G L n (C) such that • + (x) = − (x)M(x) x ∈ γ ; • (z) is uniformly bounded in C; • (z 0 ) = 1. In the following we will always choose z 0 = ∞. In case there are contours extending to infinity, this normalization needs a little of explanation: thanks to the requirement that the jump matrices at infinity are exponentially close to the identity one can see that the value of the solution (z) is actually continuous at z = ∞ and hence it can be chosen as a normalization point. The reason for allowing contours extending to infinity (under our condition for the jump matrices) is because we want to consider the case of Painlevé II
Dependence on Monodromy Data of the Isomonodromic Tau Function
545
as an example in Sec. 6.1, and also most ODE’s with polynomial coefficients exhibiting Stokes’ phenomenon can be transformed into such a RHP. In general there are obstructions to the solvability of Problem 2.1: for example let c be a point in γ where several arcs γ1 , . . . , γ meet (at nonzero tangential relative angles) and suppose they are oriented towards c. Let M j (z), j = 1, . . . , be the corresponding (locally analytic) jump matrices, then M1 (c)M2 (c) . . . M (c) = 1
(2.5)
is an obstruction. This does not mean that it is impossible to find satisfying the jumps, but it cannot be bounded at c in general and thus a relaxed growth constraint must be allowed at c. However there are situations which are of interest for our applications where a growth behavior can be “traded in” for an extra contour. Example 2.1. A typical example is depicted in Fig. 1; here on the contour that terminates at z = a there is a constant jump of the form M = C −1 e2iπ L C, where L is upper triangular (in Jordan form, for example). It is clear that the diagonal entries of L (the exponents of formal monodromy) are defined only up to addition of integers. This arbitrariness in fact corresponds to the choice of a growth behavior at the endpoint for the solution of the RHP. Once this choice has been made, we can always recast the problem into an equivalent one where the solution is bounded. This is achieved by adding a small circle at a and re-defining (z) = (z)C −1 (z − a)−L inside, with the cut of (z − a)−L along the dotted line on the right panel of Fig. 1: in the new formulation for the problem, (z) will be bounded in a neighborhood of z = a and the two problems are equivalent, in the sense that one has solution if and only if the other does, and the relation between the two solutions is also very simple. (Loop) group structure. Although we will not make any explicit use of the following fact, we mention that the matrices M : γ → S L r (C) that satisfy all the conditions above form a (infinite dimensional) Lie group, akin to the usual loop group, were it not for the fact that the contour is not a (collection of) circle(s). It is convenient to introduce a symbol for this manifold (group), G := M : γν → S L r (C), M γ (z) locally analytic . (2.6) ν
Since we will only consider finite dimensional (analytic) submanifolds of G we will not dwell on the infinite–dimensional differential-geometric issues. Most notably the tangent bundle T G will be used in a rather naive form where possible issues of topological nature (arising from the infinite dimensionality) are disposed of. In fact only the tangent space to the finite dimensional submanifolds of interest will appear, and hence the point is not relevant. 2.1. Deformations and Malgrange’s form. Suppose now that M(z) = M(z; s) depends holomorphically on additional parameters s; the reader should think of this as an explicit dependence, dictated by the problem under consideration. The parameters s could be thought of as coordinates on a manifold of deformations. When -eventually- we specialize the setting these s’s will be the isomonodromic times together with the monodromy data. On this manifold we define the one-form (differential) already used by Malgrange [11].
546
M. Bertola
Fig. 2. The local extension of + across a contour
Definition 2.1. Let ∂ denote the derivative w.r.t. one of the parameters s and assume that the Riemann–Hilbert Problem 2.1 admits a solution in an open subset of the s–parameter space.2 Then we define Malgrange’s form ω M , ω M (∂) = ω M (∂; []) := −
γ
dx −1 Tr − (x)− (x)∂ (x) 2iπ
(2.7)
∂ (z) := ∂ M(z)M −1 (z). Remark 2.3. The minus sign is –of course– conventional, but it is important because with the minus the tau-function (when it exists) has a zero and not a pole on the Malgrange divisor. Remark 2.4. The definition would hold identically also for any RHP formulated on a Riemann surface. Remark 2.5. The reader may frown upon the boundary value of the derivative of : however, since our assumptions mandate that Mν (z) is analytic in a neighborhood of the contour γν , it is simple to show that the derivative also admits (bounded) boundary values. In fact more is true: since the matrices M(z) are analytic at each interior point x of the contours, both boundary values of the solution can be analytically continued within a neighborhood of x to (germs of) analytic functions that we denote again ± (z). For example, if we want to analytically continue + (x) to an analytic function + (z) within a small disk of x ∈ γ not containing any other arc, we simply define + (z) := (z) for z on the left half disk (relative to the orientation of the arc), and + (z) := (z)M −1 (z) on the right side (see Fig. 2). Remark 2.6. In the second notation for ω M we have indicated the dependence on , which depends on the choice of jump matrices and specifications of the growth behaviors. When no ambiguity can arise, we will understand such a dependence. 2 The small–norm theorem for Riemann–Hilbert problems implies that if a RHP is solvable, then any sufficiently small deformation (in L 2 and L ∞ norms) of the jump matrices leads to a solvable RHP. With our
assumptions on the s–dependence of the jump matrices this implies that the subset of solvable RHP is an open set (if non-empty).
Dependence on Monodromy Data of the Isomonodromic Tau Function
547
Curvature of ω M . The first issue is the computation of the exterior differential: the computation can be found in [11] but in rather abstract terms and we prefer to give a direct derivation here. We have first Lemma 2.1. Let ∂ denote any vector field in the parameters of the jump matrices. Then −1 − (x)∂ (x)− (x) dx (z) (2.8) ∂(z) = x−z 2iπ γ and, consequently −1 −1 (x)(w) dx (z)(w) −1 (z)− (x) ∂ = ∂ (x) − , z−w z−x x −w 2iπ γ
−1 (z) (x) (x) −1 (x)(z) dx − ∂ − −1 ∂ (z) (z) = . 2 (z − x) 2iπ γ
(2.9) (2.10)
Proof. First of all we have ∂+ (z) = ∂− (z)M(z) + − (z)∂ M(z) , z ∈ γ .
(2.11)
This is a non-homogeneous Riemann–Hilbert problem: since the problem is normalized at infinity, (∞) = 1, it follows that ∂(z) = O(z −1 ). It is seen that the proposed expression fulfills (2.11) and this last condition. The uniqueness follows from the fact that the homogeneous part has only the trivial solution that tends to zero at ∞. The formula (2.9) follows by direct application of (2.8) and the Leibnitz rule. Finally, formula (2.10) can be obtained as a limit w → z of (2.9) as follows: −1 −1
(z)(w)−1 (z)(w) −1 ∂ (z) (z) = − lim ∂ = − lim ∂ . (2.12) w→z w→z z−w z−w
Proposition 2.1. Denoting with ∂, ∂ the derivatives w.r.t to two of the parameters s, the exterior differential of ω M is 1 d d dx ∂ (x)∂ (x) . ∂)− ∂ω M (∂) = Tr ∂ (x) ∂ (x) − η(∂, ∂) := ∂ω M ( 2 γ dx dx 2iπ (2.13) Proof. Let ∂, ∂ be two commuting vector fields so that δω(∂, ∂) = ∂ω( ∂) − ∂ω(∂).
(2.14)
We have, from Definition 2.1,
dx −1 ∂ω M ( ∂) = − Tr ∂ − (x)− (x) ∂ (x) 2iπ γ
dx
−1 Tr − (x)− (x)∂ ∂ (x) − 2iπ γ K(x, y)∂ (y)K(y, x)∂ (x) dy dx =− Tr 2 (x − y) 2iπ 2iπ γ γ x=x−
dx
−1 − , (2.15) Tr − (x)− (x)∂ ∂ (x) 2iπ γ
548
M. Bertola
where we have used (2.10) and introduced the convenience notation −1 K(x, y) = − (x)− (y).
(2.16)
Subtracting the term ∂ω M (∂) to (2.15) we have ∂ω M ( ∂) − ∂ω M (∂) =− γ
γ
+ +
γ
γ
K(x, y)∂ (y)K(y, x)∂ (x) dy dx (x − y)2 2iπ 2iπ x=x− K(x, y)∂ (y)K(y, x)∂ (x) dy dx Tr 2 (x − y) 2iπ x=x− 2iπ
Tr
γ
dx −1 Tr − (x)− (x) ∂ (x), ∂ (x) 2iπ
(2.17)
(the subscripts x=x± or y=y± indicate which boundary value is taken after the integration), where we have used the zero curvature relation for ∂ (x) = ∂ M(x)M −1 (x), namely ∂∂ (x) − ∂∂ (x) = ∂ (x), ∂ (x) .
(2.18)
In the first two terms the order of integration is important since the kernel is singular due to the denominator (x − y)2 . Let us examine the first two terms in the right-hand side of Eq. (2.17): if we denote temporarily by F(x, y) the expression
−1 −1 (z)− (w)∂ (w)− (w)− (z)∂ (z) F(z, w) := Tr −
(2.19)
the numerators are simply F(x, y) and F(y, x) respectively, due to the cyclicity of the trace; note that by Remark 2.5 the function − (z) has been extended to a local analytic function across the contours of the jumps, so that F(z, w) can be taken as locally analytic around the contours of integration in both variables. If we interchange the order 1 of integration we must evaluate the boundary value of the singular kernel (x−y) 2 on the opposite side, namely
γ
γ
F(x, y) dy dx F(x, y) dx dy = 2 2iπ (x − y)2 2iπ x=x− 2iπ (x − y) 2iπ γ γ y=y+ F(y, x) dy dx , = 2 γ γ (x − y) 2iπ x=x+ 2iπ (2.20)
Dependence on Monodromy Data of the Isomonodromic Tau Function
549
where in the last equality we simply have renamed the dummy variables. We thus have K(x, y)∂ (y)K(y, x)∂ (x) dy dx Tr 2 (x − y) 2iπ x=x− 2iπ γ γ K(x, y)∂ (y)K(y, x)∂ (x) dy dx Tr − 2 (x − y) 2iπ 2iπ γ γ x=x− F(y, x) dy dx F(x, y) dy dx − = 2 2iπ 2 2iπ (x − y) 2iπ (x − y) 2iπ γ γ γ γ x=x− x=x−
F(y, x) dy dx 2 γ γ (x − y) 2iπ x=x− 2iπ F(y, x) dy dx F(x, y) dy dx 1 1 − − 2 2 2 γ γ (x − y) 2iπ x=x+ 2iπ 2 γ γ (x − y) 2iπ x=x− 2iπ
=
1 =− 2 1 − 2
γ
γ
1 dx F(y, x) dy dx + y=x 2iπ 2 2 γ γ (x − y) 2iπ x=x− 2iπ F(x, y) dy dx (x − y)2 2iπ x=x− 2iπ
∂ y F(y, x) γ
1 dx F(x, y) dy dx + y=x 2iπ 2 2 γ γ (x − y) 2iπ x=x+ 2iπ γ F(x, y) dy dx 1 − 2 γ γ (x − y)2 2iπ x=x− 2iπ 1 dx dx 1 =− ∂ y F(y, x) y=x ∂ y F(x, y) y=x + 2 γ 2iπ 2 γ 2iπ dx 1 . ∂ y F(x, y) y=x − ∂ y F(y, x) y=x = 2 γ 2iπ 1 =− 2
∂ y F(y, x)
(2.21)
We have used, on the fourth and fifth line of (2.21) the derivative form of Cauchy integral formula. Computing explicitly the derivatives involved in this last expression we find dx 1 ∂ y F(y, x) − ∂ y F(x, y) y=x y=x 2iπ 2 γ
dx 1 Tr ∂ (x)∂ (x) − ∂ (x)∂ (x) = 2 γ 2iπ
dx −1 . (2.22) Tr − (x)− (x) ∂ (x), ∂ (x) − 2iπ γ Replacing this last expression into (2.17) yields the proof.
The Theta divisor. The reader should observe and compare the expressions (2.13) and (2.7): the crucial point is that ω M (2.7) is not defined whenever the RHP does not admit a
550
M. Bertola
solution. On the contrary, as a sort of “miracle” the expression (2.13) for its curvature is defined and holomorphic for any group–valued matrix M(z), whether or not the RHP is solvable. In other words, η is a closed two–form which is smooth also on the Malgrange divisor, whereas ω M is undefined at those points. Malgrange proved [11] (and Palmer generalized [14]) that ω M has a “pole” on such divisor: evidently, in the curvature such a pole disappears. Tau function. It is also apparent that ω M is in general not a closed one-form on the whole (infinite–dimensional) manifold G: however, it may become closed when restricted to suitable submanifolds L → G. Since the curvature η (2.13) is explicitly computable, it is easy to verify for a given explicit submanifold L whether ω M L is closed or not. When such restriction turns out to be closed we can define (locally) a function on L by τL = e ω M L . (2.23) By Malgrange’s (Palmer’s) results, this function (defined up to nonzero multiplicative constant) will vanish at the intersection with the divisor, τL (s ) = 0 ⇔ s ∈ L ∩ . (2.24) Suppose, however that ω M L is not closed; we may still seek another differential ϑ that “cures” the curvature
d ω M + ϑ L ≡ 0 (2.25) and then proceed with the construction of the tau function as before. Ideally such differential should be smooth on the whole G so that it does not change the cohomology class of ω M and also does not change the singularity structure; this way the tau function will still vanish only at L ∩ and our goal is met. Although it is not a pedagogical approach, since we know what L will be for our purposes and we have already found ϑ, we will describe it directly here. Changing the curvature of ω M . Consider the one–form ϑ that, evaluated on a vector ∂ yields
dx 1 . (2.26) Tr M (x)M −1 (x)∂ M(x)M −1 (x) ϑ(∂) := − 2 γ 2iπ Its curvature is the two–form δϑ(∂, ∂) = ∂ϑ( ∂) − ∂ϑ(∂)
dx 1 = Tr ∂ M (x)M −1 (x)∂ M(x)M −1 (x)−∂ M (x)M −1 (x) ∂ M(x)M −1 (x) 2 γ 2iπ (2.27) which the reader can verify using Leibnitz rule; here the prime denotes (as always) the derivative with respect to x. The important (but trivial) additional observation is that ϑ is a smooth differential on the whole G, since it does not require the solution of a RHP. Definition 2.2. The modified Malgrange differential is defined as := ω M + ϑ.
Dependence on Monodromy Data of the Isomonodromic Tau Function
Remark 2.7. A direct computation shows that
dx −1 Tr − (x)− (x)∂ M(x)M −1 (x) 2iπ γ
dx Tr +−1 (x)+ (x)M −1 (x)∂ M(x) − 2iπ γ
dx = 2ϑ(∂), Tr M (x)M −1 (x)∂ M(x)M −1 (x) =− 2iπ γ so that can be written in the more symmetric form 1 −1 (∂; []) = − Tr − (x)− (x)∂ M(x)M −1 (x) 2 γ
dx . + +−1 (x)+ (x)M −1 (x)∂ M(x) 2iπ
551
(2.28)
(2.29)
Combining the computation of the curvature of ω M (Prop. 2.1) with the curvature of ϑ above we have the simple Proposition 2.2. The curvature of the modified Malgrange form is δ(∂, ∂) = ∂( ∂) − ∂(∂) dx 1 = . Tr M (x)M −1 (x) ∂ M(x)M −1 (x), ∂ M(x)M −1 (x) 2 γ 2iπ (2.30) Remark 2.8. Referring to ([11] Thm. 5.5) the curvature δ clearly lies in the same cohomology class as δω M since the two differ by a smooth exact differential. If we define γ as the total Maurer–Cartan form γ (x) = M (x)M −1 (x) dx + ∂s j M(x)M −1 (x) ds j one has the expression 1 δ = Tr(γ ∧ γ ∧ γ ). (2.31) 12πi γ 2.2. Submanifolds of G where is closed. Looking at the formula of the curvature form of Prop. 2.2 we can ask ourselves what kind of jump-matrices give rise to a closed instance of the form ; we will consider cases where the integrand is identically zero. This can happen for many reasons but for our purposes and for all the applications that we could find, it is sufficient to consider those jump matrices M(z) that on each arc of γ reduce to one of the following forms: • Piecewise triangular. Matrices of the forms Mν (z) = Pν (1 + Nν (z))Pν−1 , z ∈ γν ,
(2.32)
where Nν (z) are upper–triangular analytic matrices and Pν is any constant permutation matrix (i.e. an element of the Weyl group for S L r ). • Constants. Matrices independent of z. • Torals. Matrices Mν = Dν (z) with Dν (z) diagonal matrices or any conjugation thereof by an arbitrarily chosen but fixed matrix.
552
M. Bertola
We see in Sect. 3 that any (generic) rational ODE can be encoded in a Riemann– Hilbert problem with jumps of the form indicated here above. Therefore on these submanifolds yields a closed differential. We will also show that its restriction to the isomonodromic submanifolds coincides with ω J MU . 2.3. “Schlesinger” transformations. The aim of this section is to compare the differen defined by the same jumps but allowing poles (or requiring tial ω M on two RHP for , vanishing) in the columns of the solution at specified points. We note immediately that two such solutions differ (if both exist) by a left multiplication by a rational matrix R(z); indeed –having , the same jumps by assumption– the matrix R(z) := (z) −1 (z)
(2.33)
is an analytic function in C taking away the points where has poles. Since is allowed to have poles, the matrix R(z) can have a pole of order at most equal to that of and hence it is forced to be rational by Liouville’s theorem. We now make the following observation: let R(z) be a rational matrix such that the divisor of all poles of both R(z), R −1 (z) consists of the points c1 , . . . , c K ∈ γ . Define (z) := R(z)(z).
(2.34)
Quite clearly solve a different RHP with the same jumps but possibly poles at those of R. The difference between the ω M evaluated using or is given by
dx −1 ]) − ω M (∂; []) = − Tr − (x) − (x)∂ M(x)M −1 (x) ω M (∂; [ 2iπ γ
dx −1 Tr − (x)− (x)∂ M(x)M −1 (x) + 2iπ γ
dx −1 Tr − (x)R −1 (x)R (x)− (x)∂ M(x)M −1 (x) =− 2iπ γ
dx −1 Tr R −1 (x)R (x)− (x)∂ M(x)M −1 (x)− (x) =− 2iπ γ
dx Tr R −1 (x)R (x) ∂+ (x)+−1 (x) − ∂− (x)− (x) = 2iπ γ =
K j=1
res Tr R −1 (z)R (z)∂(z) −1 (z) dz.
z=c j
(2.35)
Therefore we see that –while ω M is an object of transcendental nature, the above difference requires only the evaluation of a finite number of residues. Note also that -since (Def. 2.2) differ by ω M only in an explicit term ϑ that depends only on the jump matrices, we have ω M (∂; [ ]) − ω M (∂; []) = (∂; [ ]) − (∂; []).
(2.36)
We now proceed with the definitions in the title of the section and specialize the class of rational matrices R(z).
Dependence on Monodromy Data of the Isomonodromic Tau Function
553
Definition 2.3. Given two distinct points ξ = η and two (possibly formal) series Yξ (z) = G ξ 1 +
∞
Yξ ; z ξ
,
Yη (z) = G η 1 +
=1
z ξ := (z − ξ ), z ∞ :=
∞
Yη; z η
,
(2.37)
=1
1 , z
(2.38)
an elementary Schlesinger transformation at two distinct points ξ = η is a rational matrix R(z) such that ξ (z)z Ei , R(z)Yξ (z) = Y ξ
η (z)z η−E j , R(z)Yη (z) = Y
(2.39)
• denote formal series of the same form. In the case of ∞ we have G ∞ = 1 = where Y G ∞ . If neither ξ = ∞ = η then we impose also R(∞) = 1. [For the case ξ = η the definition should be modified in an obvious way: please see below]. The problem is purely of algebraic nature, and not a very difficult one: the computation is contained in [[7], App. A] and the derivation will not be reported (we will give below the relevant results). Suppose now we have a RHP for with jump matrices M as in Prob. 2.1: at points ξ = γ the solution to Prob. 2.1 yields a (convergent) power series which can be used as input for the above procedures. If ξ ∈ γ we cannot have a (even formal) series since –in general– not even the value of is well defined at z = ξ . In the applications to ODEs the following situation occurs; let ξ ∈ γ be a point where γ1 , . . . , γ meet ( ≥ 1). Suppose that the jump matrices are analytic in small sectors centered at ξ containing the direction of approach and that –in said sector– (see Fig. 3) M j (z) = 1 + O((z − ξ )∞ ).
Fig. 3. Illustration of the sectors of analyticity for M j where the decay (2.40) should be valid
(2.40)
554
M. Bertola
Then it is not hard to see3 that the solution (z) has the same asymptotic expansion in each of the sectors at ξ separated by the incoming arcs (z) ∼ G ξ 1 +
∞
Yξ ; j (z − ξ )
j
ξ (z). =: Y
(2.41)
1
We will allow to perform Schlesinger transformations involving either points ξ not in γ or points where the condition (2.40) is met, so that the solution defines unambiguously a (formal) analytic series centered at the point. Definition 2.4. The elementary Schlesinger transformation ξi ηj for the solution of Problem 2.1 is the solution (if it exists) of the following new RHP (where z ξ = (z − ξ ) if ξ = ∞ and z ∞ := 1z )
ξ = η
ξ = η (i = j)
+ = − M (z) = O(1)z ξ −Ei , z ∼ ξ (z) = O(1)z η E j , z ∼ η
+ = − M (z) = O(1)z ξ −Ei +E j , z ∼ ξ
(2.42)
where the normalization is fixed by requiring that (z) ∼ 1 + O(z −1 ) if ξ = ∞ = η or –if either ξ or η are infinity– that the term indicated by O(1) above is actually 1+O(z −1 ). Here and below we use the notation E i j for the elementary matrices (with a 1 on the i th row, j th column) and E i := E ii . It is immediately seen that (z) = R(z)(z)
(2.43)
with R(z) rational: indeed the ratio (z) −1 (z) for the transform ξi ηj does not have jumps and may have at most a simple pole at z = ξ . The matrix R(z) can be computed directly from the (possibly formal) series-expansion of at ξ, η (see [7]. We give below a more compact formula that the reader can check autonomously. 3 By moving slightly the jump –which can be done due to the local analyticity of the jump matrix– one sees that the solution can be “continued” analytically across the jump from the left and from the right. The ratio of these two extensions in the common sector of analyticity differs from the identity by exponentially small terms, which have no bearing on the asymptotic expansion.
Dependence on Monodromy Data of the Isomonodromic Tau Function
555
Proposition 2.3 (cf. App. A, B in [7]).Theleft-multiplier matrix R(z) that implements the elementary Schlesinger transform ξi ηj is given by
R(z) = 1 +
ξ η i j
ξ −η ( −1 (ξ )(η))i j
R −1 (z) = 1 −
(ξ = η) (η)E ji −1 (ξ ) z−ξ
R(z) = 1 +
ξ ξ i j
(i = j)
1 ( −1 (ξ ) (ξ ))i j
(ξ )E ji −1 (ξ ) z−ξ
(η)E ji −1 (ξ ) −1 (ξ )E ji −1 (ξ ) 1 R (z) = 1 − −1 z−η z−ξ ( (ξ ) (ξ ))i j z−η det R(z) = det R = 1 z−ξ ξ −η
( −1 (ξ )(η))i j
(2.44)
ξ ∞ i j
(ξ = ∞)
R(z) = 1 − E j j + R1 =
R1 z−ξ
−1 (1 − E j j ) (∞) − 1 E ji −1 (ξ ) ((ξ )−1 )i j
E ji −1 (ξ ) i−1 j (ξ )
det R(z) =
1 + (∞)E j j
(η = ∞)
R(z) = E ii (z − η) + R0 (η)E ji R0 = 1 − E ii (∞) 1 − i j (η)
R −1 (z) = E j j (z − ξ ) + R0 R0 = 1 −
∞η i j
1 R −1 (z) = 1 − E ii + z−η
R
1 (η)E (∞)(1 − E ) + 1 R1 = (η) ji ii ij
1 z−ξ
det R(z) = (z − η)
(2.45) where we have denoted by (∞) the derivative in the local parameter, that is 1 (z) = 1 + (∞) + · · · . z (The formulæ for
∞∞ i j
(2.46)
can be found in loc cit.)
If ξ (or η or both) belong to γ where condition (2.40) is in place, the computation leading to (2.35) can be still carried out with minor modifications in the process but not in the result. First of all note that the integral ω M (∂, [ ]) is still convergent since −1 ∞ ∂ = ∂ M M = O((x − c) ) (recall that we assumed M = 1 + O((x − c)∞ )) and along the contours incident at c is still hence the additional algebraic growth of −1 integrable when multiplied by ∂ . Let γ denote the contours that lie outside of disks centered at the points z = ξ, η (possibly the same) of the Schlesinger transform; we then have
dx −1 −1 ω M (∂; [ ])−ω M (∂; []) = − lim Tr − (x) − (x)−− (x)− (x) ∂ (x) →0 γ 2iπ
dx −1 Tr − (x)R −1 (x)R (x)− (x) ∂ M(x)M −1 (x) = − lim →0 γ 2iπ
556
M. Bertola
dx −1 = − lim Tr R −1 (x)R (x)− (x) ∂ M(x)M −1 (x)− (x) →0 γ 2iπ
dx Tr R −1 (x)R (x) ∂+ (x)+−1 (x) − ∂− (x)− (x) = lim →0 γ 2iπ
dx . Tr R −1 (x)R (x)∂(x) −1 (x) = lim →0 |x−ξ |=|x−η|= 2iπ
(2.47)
In the computation of the limit (2.47) we can replace ∂(x) −1 (x) by a suitable truncation of the asymptotic series at z = ξ, η (which does not depend on the direction of approach under our assumptions for M), committing an o(1) error as → 0. Then the limit equals the formal residue
ξ (z)Y ξ (z)−1 dz ω M (∂; [ ]) − ω M (∂; []) = res Tr R −1 (z)R (z)∂ Y z=ξ
η (z)Y η (z)−1 dz. (2.48) + res Tr R −1 (z)R (z)∂ Y z=η
Here the residue means simply the coefficient of the power −1 of the local parameter (note that R −1 (z)R (z) has at most a double pole and hence the formal residue involves (z)). In case only one point is involved at most the first two terms in the formal series Y in the Schlesinger transformation we have simply only one residue at the end. Remark 2.9. In the case of a Schlesinger transformation involving two distinct points the determinant of the solution det cannot remain constant, since R(z) has non-constant determinant. This does not pose any significant problem as we explain presently. To account for the different power-law of the columns at the two points z = ξ, z = η, small counterclockwise circles around those points should be added to γ imposing additional jumps of the form M1 (z) = (z − ξ )−Ei , |z − ξ | = , M2 (z) = (z − η) E j , |z − η| = ,
(2.49)
conjugating by the same matrices any jump of a contour that passes within said circles. Of course these jumps do not have unit determinant and hence the uniqueness of the solution must be argued in a different way from the one used in the unimodular case: given a solution (z) of the new RHP we see that det (z) is analytic and bounded everywhere, except for jumps on the new circles where det + (z) = (z − ξ )−1 det − (z) on |z − ξ | = and det + (z) = (z − η) det − (z) ,
on |z − η| = . (2.50)
This means that det (z) admits analytic continuation in the interior of the two disks, with a simple pole at z = η and a simple zero at z = ξ , plus the condition det (∞) = 1. z−ξ This forces det (z) ≡ z−η outside of the disks, det (z) = (z − ξ ) for |z − η| < and 1 viceversa det (z) = z−η for |z − ξ | < . Any solution will have the same determinant and hence the uniqueness is established along the same way used previously.
A direct computation based on (2.35) yields the following theorem, which is simply a rephrasing of an homologous theorem in [7], with the proper extension of understanding to the setting of RHPs.
Dependence on Monodromy Data of the Isomonodromic Tau Function
557
Theorem 2.1 (Thm. 4.1 in [7]). Given two RHPs related by the elementary Schlesinger transformation ξi ηj (Def. 2.4), the difference of the Malgrange differential on the two solutions is a closed differential on the deformation manifold given by ξη ω M (∂; [ ]) − ω M (∂; []) = ∂ ln H , (2.51) i j where
H
ξη i j
=
⎧ ⎪ ( −1 (ξ ) (ξ ))i j ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ((η))i j ⎨ ⎪ ⎪ ((ξ )−1 )i j ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ( −1 (ξ )(η)) ⎪ ij ⎪ ⎩ ξ −η
for for for for
ξ ξ i j
∞η i j ξ ∞ i j ξ η i j
(i = j) , η = ∞ , ξ = ∞
,
(2.52)
, ξ = ∞ = η
where the notation (∞) –as previously– denotes the derivative in the local parameter 1 (z) =: 1 + (∞) + . . . . z
(2.53)
For the reader’s convenience we verify Thm. 2.1 for the case ξi ηj with distinct ξ = ∞ = η and both ξ, η ∈ γ . Let ∂ be a variation of the jump-matrices M (i.e. not moving ξ, η). We have from Prop. 2.3, (η)E ji −1 (ξ ) 1 1 R −1 (z)R (z) = −1 − . (2.54) ( (ξ )(η))i j z − η z−ξ We then have to compute the (possibly formal) residue (2.35) (η)E ji −1 (ξ ) 1 1 −1 res Tr − ∂(z) (z) dz z=ξ,η ( −1 (ξ )(η))i j z − η z−ξ =
( −1 (ξ )∂(ξ ) −1 (ξ )(η))i j ( −1 (ξ )∂(η))i j − ( −1 (ξ )(η))i j ( −1 (ξ )(η))i j
( −1 (ξ )∂(η))i j (∂( −1 (ξ ))(η))i j + ( −1 (ξ )(η))i j ( −1 (ξ )(η))i j −1 ( (ξ )(η))i j . = ∂ ln ξ −η =
(2.55)
To verify the formula for ∂ξ , ∂η as well, we must add a small circle around them and a new jump M(z) = (z − ξ ) Ei , M(z) = (z − η)−E j respectively. Since ξ, η did not exist as deformation parameters in the RHP for , from the definition of ω M we need to compute (we do it only for ∂ξ , leaving the verification for ∂η to the reader) dx Ei −1 Tr − (x)− (x) ω M (∂ξ ; [ ]) = − . (2.56) x − ξ 2iπ |x−ξ |=
558
M. Bertola
Note that − (z) = R(z)(z) and + (z) = R(z)(z) (z − ξ )−E j when z belongs to the circle |z − ξ | = so that −1 R (z)R (z)(z)E i −1 (z) −1 (z) (z)E i + dz ω M (∂ξ ; [ ]) = − res Tr z=ξ z−ξ z−ξ
−1
(ξ ) (ξ ) −1 (ξ )(η) i j −1 (ξ )(η) i j 1 =− − = ∂ ln . (2.57) ξ ξ −η ( −1 (ξ )(η))i j ξ −η This proves completely the case considered. It appears quite obviously that Proposition 2.4. An elementary Schlesinger transformation exists if and only if ξ η H i j = 0. Remark 2.10. The zeroes of the matrix entries of the solution of the RHP acquire therefore the meaning of intersection of the Malgrange divisor with the space ofthe parameter η added to the problem, namely, the position of the Schlesinger transform ∞ i j . We are not going to dwell at length on the algebra of iterated elementary Schlesinger transformations and on the general transformation since the formulæ are contained in [7]; we only point out that in loc. cit. the transformations were applied to solution of either isomonodromic or isospectral deformation problems and not a general Riemann– Hilbert problem. Thus, we are mainly shifting the perspective (and compactifying some notation) of [7]. Remark 2.11. The fact that ω evaluated on two solutions of RHPs with the same jumps is a closed differential is immediate from the fact that the curvature of ω does not depend on the growth behavior of the solution but only on the jump matrices. When we specialize the setting to the case relevant to ODEs, Thm. 2.1 will hold for differentiations with respect to the monodromy data as well. 2.3.1.Generalized Sato formula Let η ∈ γ : denote by η;i j (z) the Schlesinger trans ∞η form i j of the solution of the RHP 2.1. Then the second formula (2.52) reads i j (η) ∝ exp
s
ω M (•; [η;i j ]) − ω M (•; []),
(2.58)
where the one form under integration is closed by the above remark. This is nothing but the Sato formula for the Baker Akhiezer vector; of course, at this level of generality we do not have a “tau” function because –in general– ω M will not be a closed differential of the deformation parameters of the problem. If the problem admits a τ –function, that is, if the differential ω M (or ) is closed on the submanifold of Riemann–Hilbert problems under consideration, then we have a honest version of the Sato formula, η τ ∞ i j i j (η) = , (2.59) τ η stands for the τ –function of the problem with the “insertion” of the where τ ∞ i j Schlesinger transform.
Dependence on Monodromy Data of the Isomonodromic Tau Function
559
To make the remark a bit more concrete, let us pick a point z = a ∈ γ and a small disk D. For simplicity in writing the formulæ we will simply set a = 0. Let T (z) be a diagonal matrix defined on ∂D that admits analytic continuation on P1 \ D, ∞ Tk T (z) = , Tk = diag(tk;1 , . . . , tk,r ) , zk
(2.60)
k=1
where the Laurent series is supposed to be actually convergent on P1 \ D. Let (z; [T ]) denote the solution of the RHP, + (z) = − (z)M(z) , z ∈ γ , + (z) = − (z)e
T (z)
(2.61)
, z ∈ ∂D, (∞) = 1.
(2.62) (2.63)
If the L ∞ ∩ L 2 norm of eT (z) − 1 on ∂D is sufficiently small, then the solvability of the problem is guaranteed by standard perturbation theorems. Therefore (z; [T ]) is defined at least in a ball around T (z) ≡ 0. Since the jump matrix eT (z) is diagonal (and analytic in the complement of the disk), it is immediately seen from Prop. 2.2 that ω M = 4 and that they are closed as differentials on the (infinite dimensional) manifold of T ’s. Thus there is a locally defined function such that
dx −1 . (2.64) δ ln τ (T ) := Tr − (x)− (x)δT (x) 2iπ ∂D Denote now by τ ξi ηj (T )5 the tau–function resulting after the elementary Schlesinger transformation ξi ηj . Then the content of Thm. 2.1 can be rephrased as ζ (T ) (T ) τ ∞ τ ζi ∞ i j j , −1 i j (ζ ) = . (2.65) i j (ζ ) = τ (T ) τ (T ) In fact, more is true: if ζfalls within the disk D then the RHP for the Schlesinger ∞ζ transforms i j , ζ i ∞j can be formulated as (exercise)
∞ζ i j
ζ ∞ i j
−E j + (z) = − (z)eT (z) 1 − ζz z Ei −E j , z ∈ ∂ D + (z) = − (z)eT (z) (1 − ζz ) Ei z Ei −E j , z ∈ ∂ D . + (z) = − (z)z −Ei M z Ei (∞) = 1
+ (z) = − (z)z E j M(z)z −E j (∞) = 1
z ∈ γ
z ∈ γ
(2.66) The jumps on ∂D can be written ∞ Tk + E j ζ k /k ζ −E j ζ Ei T (z) 1 − eT (z) 1 − = exp ; e z zk z k=1
∞ Tk − E i ζ k /k = exp . zk
(2.67)
k=1
4 This follows by observing that ϑ will be identically zero due to the analyticity of T outside of the disk. 5 The notation T stands for the (infinite) vector of the (matrix) coefficients of T (z), T = (T , T , . . . ). 1 2
560
M. Bertola
This leads to the following identities
i j (ζ ; [T ]) = i j 0; T − E j [ζ ] , −1 i j (ζ ; [T ]) = −1 i j (0; [T + E i [ζ ]]) , (2.68) and hence i j (ζ ; [T ]) =
τ
∞ 0 i j
(T − E j [ζ ])
τ (T )
; −1 i j (ζ ; [T ]) =
τ
0 ∞ i j
(T + E i [ζ ])
τ (T )
.
(2.69) Here we have used the standard notation [ζ ] = (ζ, ζ2 , . . . , ζk , . . . ). 2
k
2.3.2. Hirota bilinear relations We will not go into much depth here, since all is well– known, but it may give an analytic perspective on the relations, which are usually taken only formally. As in the previous Sect. 2.3.1, let C be a counterclockwise circle around a point z = a ∈ γ (a = 0 for simplicity) and let T (z) : C → glr (C) be as (2.60). We have −1 ]) dζ = ]) dζ . + (ζ ; [T ])e−T (ζ )+T (ζ ) +−1 (ζ ; [T − (ζ ; [T ])− (ζ ; [T 2 2iπ ζ 2iπ ζ 2 C C (2.70) ]) has no jumps outside C since the other jumps Now, the matrix (x; [T ]) −1 (x; [T of the problem have been left unmodified; thus it is analytic on the complement of the disk D and goes to the identity at ∞. Thus we have the identity ]) dζ = 0 + (ζ ; [T ])e−T (ζ )+T (ζ ) +−1 (ζ ; [T (2.71) 2iπ ζ 2 C in a neighborhood of T ≡ 0. which is valid identically in T, T In view of the interpretation of as the Baker–Akhiezer vector for the τ –function (Sato-formula above), the reader may regard (2.71) as the generating function of an infinity of bilinear identities between matrix–valued τ functions; r
=1
τ
∞ 0 0 ∞ dζ (ζ )−T (ζ ) T (T − E [ζ ])e τ (T + E [ζ ]) 2 ≡ 0. ζ i
j
(2.72)
The identity (2.72) should be used as a generating function of an infinite hierarchy of 0 (T ) when expanding it PDEs for the matrix–valued tau-function [τ (T )]i j = τ ∞ i j around the diagonal in Taylor series with respect to T T = T . This generates a sort of “addition theorem” for tau-functions (see Remark 2 in [8]). The variational formula (2.9) in Lemma 2.1 takes on an added significance in view of the identity −1 1 ξη (ξ )(η) τ = . (2.73) τ i j ξ −η ij
Dependence on Monodromy Data of the Isomonodromic Tau Function
561
To explain this in a simple situation we now consider two disks D0 and D1 centered at two points –say– a = 0, 1; on the boundaries of these disks we introduce diagonal jumps exactly as in Sect. 2.3.1, T (0) (z) =
∞ ∞ (0) (1) Tk Tk (1) , T (z) = . zk (z − 1)k k=1
(2.74)
k=1
(0) (1) We denote by τ (T , T ) the tau–function as a function of the two (infinite) sets of 01 times and τ i j (T (0) , T (1) ) the Schlesinger–transformed one. Let ζ ∈ D0 and η ∈ D1 ; retracing the steps that lead to (2.69) we find −1 0 1 (0) (ξ )(η) 1 (1) τ (T + E i [ξ ], T − E j [η]) = . (2.75) i j ξ −η τ (T (0) , T (1) ) ij This formula is the content (in different notation) of (3.11) in Thm. 3.2 of [8]. The variational equation (2.9) in our Lemma 2.1 applied to derivatives with respect to some directions in T (0) , T (1) then become the generating functions for the Hirota bilinear relations that appear in Thm. 3.4 of [8]. They all boil down to the following identity, to be understood as generating functions of PDEs when evaluating its Taylor ( j) , j = 0, 1, expansion on the diagonal T ( j) = T −1 −1 )(η; T ) dx (x; T (ξ ; T )− (x; T ) − + 0= ξ −x x −η 2iπ ∂ D0 ∂ D1 =
)(η; T ) dx −1 (ξ ; T )+ (x; T ) T(0) (x)−T (0) (x) +−1 (x; T e +0 ξ −x x −η 2iπ ∂ D0
+
)(η; T ) dx −1 (ξ ; T )+ (x; T ) T(1) (x)−T (1) (x) +−1 (x; T e . ξ −x x −η 2iπ ∂ D1
(2.76)
The interested reader should compare this with Theorem 3.4 in [8]. Since it is not the primary focus of this paper (and it is certainly not a new result), we will not pursue the issue here, also because it has been dealt with at length in [8], even though in the context of isomonodromic and isospectral deformations only. 2.4. Right gauge equivalence. We will say that the two problems are (right) gauge equivalent if there exists an analytic function G : C \ γ → G L r (C) admitting boundary values (also for its derivative) at γ , with G(z) = 1 + z → ∞ and such that the jump matrices stand in the relation M(z) = G −1 − (z)M(z)G + (z) , z ∈ γ .
(2.77) O(z −1 )
as
(2.78)
It is immediate then that the two solutions are related by (z) = (z)G(z). It is then seen that the difference of ω M (or ) along the two solutions , differ only in terms that do not involve or and depend only and explicitly on M, G. Thus this equivalence
562
M. Bertola
will not modify the singularity locus of ω M and –if both deformation families admit a tau function– both tau functions will differ only by multiplication by a smooth nonzero factor. Indeed a direct computation yields ω M (∂; [ ]) − ω M (∂; [])
dx −1 −1 −1 = Tr − (x)− (x)G −1 − (x)∂G − (x) − + (x)+ (x)G + (x)∂G + (x) 2iπ γ −1 −1 Tr M −1 (x)M (x)G −1 + + (x)∂G + (x) − G − (x)G − (x)G − (x)∂G − (x) γ
−1 − G −1 − (x)G − (x)∂ M(x)M (x)
dx . (2.79) 2iπ The first term in the right hand side of (2.79) vanishes by the Cauchy theorem, since it amounts to the (boundary value of) the integral of Tr( −1 G −1 ∂G) on a collection of contours surrounding γ and contractible. Thus the difference will not involve the solution of the RHP and hence be a smooth differential in the parameters (i.e. gauge equivalence cannot modify the Malgrange divisor). Such equivalence does play a role in some cases (see [1] for examples where this happens, although not phrased in these terms). −1 −1 + G −1 − (x)G − (x)M(x)G + (x)∂G + (x)M (x)
3. Rational Differential Equations in Terms of Riemann–Hilbert Data We now describe the class of ODE’s with rational coefficients (z) = A(z)(z)
(3.1)
in an unconventional way: we will start from the formulation of a RHP and then indicate how this problem (when solvable) is equivalent to an ODE. The class of matrices A(z) that will be eventually described has poles at points a1 , . . . , ak with orders n 1 + 1, . . . , n k + 1. If n j = 0 then the pole is simple. One may take the point of view that we are providing a different (transcendental) coordinate system on the finite–dimensional vector space of rational matrices with fixed polar divisor. The forward problem, namely, the construction of the RHP from the matrix A(z) is more standard and we only sketch the main points, since it does not really play a direct rôle here. This procedure is often called the (extended) monodromy map. The standard reference for many assertions below is Wasow’s book [17] but also the paper [9] provides a concise recall. 3.1. Monodromy map. Given a rational matrix j +1 n K n 0 −1 A j, −1 A(z) = + A0, +1 z , (z − a j )
j=1 =1
(3.2)
=0
we consider the ODE (z) = A(z)(z). Without loss of generality we will assume Tr A(z) ≡ 0 so that any solution has constant determinant that we can assume to be unity det ≡ 1. We make the usual assumption:
Dependence on Monodromy Data of the Isomonodromic Tau Function
563
Assumption 3.1 (Genericity). The leading-coefficient matrices A j,n j for n j > 0 have distinct simple eigenvalues. The diagonal matrix of eigenvalues (with an arbitrarily chosen order) will be denoted by T j,n j . If n j = 0 (simple pole) then T j,0 has eigenvalues which are also distinct modulo Z (i.e. no pair of eigenvalues differ by an integer). To simplify some issues in the general description we will assume that in fact there is no pole of the connection ∂z − A(z) at z = ∞; this can always be achieved without loss of generality by a Möbius transformation that maps ∞ to a finite point (without mapping any of the other poles to infinity!). Monodromy representation. In order to fix the normalization of the initial value problem we choose a finite basepoint z 0 ∈ C (not coinciding with any of the poles) and we will denote by D the simply connected domain of P1 obtained by dissecting P1 along nonintersecting smooth arcs joining z 0 with each of the poles (Fig. 4). We consider the initial value problem (∞) = 1 (recall that ∞ is now a regular point of our ODE). By analytic continuation of the solution around a loop (based at ∞) that “goes around a j ” (i.e. has index one relative to a j and zero relative to z 0 and all other poles) we obtain (z) → (z)M −1 j , det M j = 1. These loops generate the fundamental group π1 (C \ {a1 , . . . }, ∞) and provide a representation of this fundamental group π(P1 \ {a1 , . . . }, ∞) → S L r (C).
(3.3)
Therefore, on the simply connected domain D defined here above, the matrix (z) is analytic, but on the boundary of the cuts from z 0 to the pole a j we have a jump M j . Stokes’ phenomenon in brief. Consider a higher order pole a j , with n j ≥ 1 and denote as follows the distinct eigenvalues of the leading coefficient matrix A j,n j , T j,n j = diag( j,1 , . . . , j,r ) , j, = j,s , = s ,
ζ := (z − a j ).
(3.4)
One can find at each pole a j 2n j directions (the anti-Stokes directions) angularly separated by nπj such that along each of them there exists a permutation σ (uniquely defined) yielding the definite ordering below: j,σ (2) j,σ (r ) j,σ (1) > > · · · > . (3.5) ζnj ζnj ζnj Note that if on a direction we have the above ordering, on the next (counter)clockwise we have the exact reversed (with the same permutation).
Fig. 4. The schematic arrangement of the basepoint z 0 and the dissection of P1 into a simply connected domain D by means of cuts (thick arcs). Also the generators of the homotopy group and the corresponding jumps on the arcs of the dissection
564
M. Bertola
The theorem which can be found in several places (e.g. [17]) is the following: given our IVP (z) on the simply connected domain D, there are 2n j matrices S j,ν , ν = 1 . . . 2n j (the Stokes’ matrices) which are of the form S j,ν = Pσ (1 + N j,ν )Pσ−1 ,
(3.6)
where N j,ν is upper-triangular on the directions where the ordering is as in (3.5) and lower-triangular on the directions where the ordering is reversed, and Pσ is the permutation matrix corresponding to the permutation σ appearing in (3.5). There is an invertible matrix C j ∈ S L r (C) (the connection matrix) and diagonal matrix H j (ζ ) := e
T j (ζ ) L j
ζ
, T j (ζ ) =
nj
T j, ζ − , ζ = (z − a j )
(3.7)
=1
in which T j,n j is the matrix of eigenvalues of the leading coefficient used in the definition of the anti-Stokes directions. This matrix will be called the toral element (for lack of a better name, since it belongs to the complex toral subalgebra of S L r (C)) and the diagonal matrix L j is called the exponent of formal monodromy. These matrices (Stokes’ connections and toral elements) are uniquely determined by the following set of conditions 1. In one (arbitrarily chosen and then fixed) sector separated by the two consecutive anti-Stokes directions we have the asymptotic expansion j (z)eT j (ζ ) ζ L j C j , (z) ∼ Y ∞
Y j, ζ , Y j (z) = G j 1 +
(3.8) det G j = 1,
(3.9)
=0
where the (constant in z) matrix G j depends on which pole we are considering. 2. In the next sector counterclockwise we have j (z)eT j (ζ ) ζ L j S j,1 C j , (3.10) (z) ∼ Y where we have labeled by 1 the anti-Stokes direction separating the two sectors, and in general j (z)eT j (ζ ) ζ L j S j,ν · · · S j,2 · S j,1 C j , (z) ∼ Y
(3.11)
where we are in the sector between two anti-Stokes and we have crossed ν such lines. 3. The monodromy around the pole under consideration, the connection matrix and the Stokes matrices satisfy the condition M j = C −1 j S j,2n j · · · S j,1 C j .
(3.12)
Note that at a simple pole we simply have n j = 0 and hence there are no anti-Stokes’ lines but only the exponents of formal monodromy L j and the connection matrix C j . Remark 3.1. If we had allowed a pole at infinity we would need to make only technical modifications: in that case –of course– one cannot set-up an initial value problem at ∞, but can require a specified normalized asymptotic behavior in a chosen sector. The description of the Stokes’ phenomenon is identical to the one provided above, using the coordinate ζ = 1z . We feel that there is no sufficient added value to this more general setting to justify entering into these confusing details, especially given that the properties of a rational ODE are invariant under Möbius transformations. Since these are standard facts about ODEs we will not dwell on other subtleties.
Dependence on Monodromy Data of the Isomonodromic Tau Function
565
3.1.1. Forward Birkhoff map. Given the rational matrix A(z) of our form, a chosen basepoint and dissection D, anti-Stokes lines and base-sector from where to start the counting we have associated the collection of all the data M := {a j , T j (ζ j ), L j , C j , {Sν, j : ν = 1 . . . 2n j }, } j=1,...,k .
(3.13)
The statement is that the map is (locally) injective but not surjective in general; there are some choices of data in M for which there is no corresponding ODE of the specified form. As the reader may suspect (or know), these data constitute the Malgrange divisor. There are two logically (and historically) distinct types of data in the above: the isomonodromic times T := {a j , T j (ζ j )} j=1...k
(3.14)
(the positions a j and the coefficients of the matrices T j (ζ )). On the other hand there are the genuine generalized monodromy data S := {L j , C j , {S j,ν ν = 1 . . . 2n j }, } j=1...k
(3.15)
and M is locally the product of the two. Of course this separation can be only done locally since the anti-Stokes directions depend on the leading coefficients of T j (ζ ) and the dissection D depends on the position of the poles, and hence some care is in order –if global questions are at issue– in describing the patching of these local descriptions (see for example [5]).
4. Inverse Birkhoff Map: Riemann–Hilbert Problem Given the Birkhoff data described above, the question arises as to whether one can invert the map: starting from a concrete dissection, anti-Stokes lines, etc., together with all the matrices appearing in M, can one reconstruct A(z)? To this end it is necessary to specify a Riemann–Hilbert problem in the same spirit as Sect. 2.
4.1. The set of contours. The set of contours γ consists of (see Fig. 5) 1. For each pole we draw a circle not containing any other pole: we will call this the connection circle. 2. For each pole a j a smaller circle is chosen, called the formal monodromy circle. On this circle a point β j is chosen. 3. Each β j on the formal monodromy circle is connected with a set of mutually nonintersecting paths (stems) to the basepoint. 4. At the higher poles we choose a third smaller circle called the toral circle. 5. A point (arbitrarily chosen) on the toral circle is connected finally to the pole by 2n j smooth curves that approach the singularity along the directions mentioned in Sect. 3.1; The word description is awkward but Fig. 5 should clarify all the elements.
566
M. Bertola
4.2. The jump-matrices. Rather than describing the matrices in words, we refer to the picture (Fig. 5) where we depict a situation with only two points a1 , a2 with n 1 = 0 (simple pole). The general picture is quite simply a repetition of several copies of the basic elements already manifest here. Since there is no monodromy around the basepoint z 0 chosen for the dissection, we have to impose the constraint C K−1 e2iπ L K C K · · · C1−1 e2iπ L 1 C1 = 1.
(4.1)
It will be understood that the matrices (z − a j ) L j are defined on the formal monodromy circles as continuous functions taken away from the point of insertion of the points β j (Fig. 5).
4.3. Riemann–Hilbert problem. Problem 4.1. Find a piecewise analytic matrix–valued function (z) on the complement of the indicated contours as per Fig. 5 so that • on each arc it solves + (z) = − (z)M(z) with the matrix M(z) as indicated in Fig. 5; • it is bounded on P1 ; • it satisfies the normalization (∞) = 1. Some properties follow immediately: • Any solution satisfies det (z) ≡ 1. Indeed det M(z) ≡ 1 implies that det((z)) has no jumps across the contours. Since (z) is bounded, so must det be, and hence it is an entire function, bounded everywhere, hence a constant. Since (∞) = 1 then det (z) = det (∞) = 1.
Fig. 5. The typical set of contours. The dashed lines within the formal monodromy circles are not jumps, but the domain where the determination of the powers (z − a)−L is continuously defined so as to give a precise meaning to the expressions for the various jump matrices
Dependence on Monodromy Data of the Isomonodromic Tau Function
567
Fig. 6. The constant-jump Riemann–Hilbert problem for (z)
• If a solution exists, it is unique: indeed any two solutions (z), (z) are analytically invertible by the above remark. Hence (z) −1 (z) is promptly seen to have no jumps, be uniformly bounded, and thus it is a constant. The normalization forces (z) ≡ (z). The connection with the ODE is as follows; define the piecewise analytic matrix (z) related to the solution (z) as follows: • (z) = (z) outside the formal monodromy circles; • (z) = (z)(z − a j ) L j in the annulus between toral and formal-monodromy circles; • (z) = (z)eT j (z) (z − a j ) L j inside the formal-monodromy circles. The cut of the logarithm is taken where the dotted line is traced in Fig. 5. It is promptly seen that (z) solves a new RHP where the jump matrices are piecewise constant and unimodular (see Fig. 6 for an example). Thus A(z) := (z) −1 (z)
(4.2)
is a meromorphic (traceless) matrix function with isolated singularities at the a j ’s. A local inspection shows that it has poles of finite order there, and hence it is rational. 5. Tau Function and (Iso)monodromic Deformations The goal now is to show that the modified Malgrange form is closed when evaluated along the manifold L ⊂ G consisting of jump-matrices M(z)’s described in the previous section. We emphasize that the parameters are • • • •
the toral data T j (z)’s, the connection matrices C j , the Stokes’ matrices S j,ν , the exponents of formal monodromy L j .
The reader acquainted with the literature about isomonodromic deformations should now realize that we allow many more directions of deformations. The computation of the closure of relies directly on Prop. 2.2: dx 1 . Tr M (x)M −1 (x) ∂ M(x)M −1 (x), ∂ M(x)M −1 (x) δ(∂, ∂) = 2 γ 2iπ (5.1)
568
M. Bertola
It is clear that no contribution to (5.1) can come from the arcs where M(z) is independent of z. This leaves only the contributions coming from the toral circles, the formal monodromy circles and the Stokes’ lines. Let us consider each type separately • Stokes’ lines; up to a constant conjugation by a permutation matrix the jumps are of the form 1 + N (z) with a strictly triangular N (z), and hence they do not contribute to (5.1) (the trace is identically zero). • Toral circles; on the Toral circles the jumps are diagonal and hence the commutator in (5.1) is identically zero. • Formal monodromy circles; once more, since the jumps are diagonal, no contribution to (5.1) comes from them. Note that at a simple pole the matrix L may actually be upper triangular: in this case the trace vanishes identically. Therefore we have Theorem 5.1. The differential restricted to the submanifold L of group–valued jumpmatrices described above is closed and defines a local function via the formula S, τ (T , a , L, )=e
.
(5.2)
This function is defined up to a nonzero multiplicative constant and it vanishes precisely when the Riemann–Hilbert problem is not solvable, namely, on the Malgrange divisor. The fact that τ has zeroes (and not a branching singularity) does not follow from our construction, it follows from [14 and 11]. We would like to stress that, had we chosen the Malgrange form ω M directly, the closure would have failed. Indeed the reader may check that
1 ∂a j δω M ∂, ∂ = ∂a j ∂Tr(L 2j ) K j , ∂Tr(L 2j ) − 2 j ln(x − a j ) dx 1 Kj = = = 0, 2 2iπ (x − a ) β − aj j j βj
(5.3) (5.4)
where the integrals are on the toral circles with basepoint at the necks β j . Aside from this, the correction term ϑ in (2.26) does not contribute anything for most of the contours in γ ; in fact the only contribution comes –not surprisingly in view of the above computation– from the formal monodromy circles
dx 1 ϑ(∂) = Tr M (x)M −1 (x)∂ M(x)M −1 (x) 2 γ 2iπ ln(x − a j ) dx 1 (5.5) = Tr(L j ∂ L j ) 2 x − a j 2iπ βj j
1 = Tr(L j ∂ L j )(2iπ + 2 ln(β j − a j )). 2
(5.6)
j
Note that this “problem” is invisible if we allow only isomonodromic deformations, in which case using ω M or yields the same differential. Before we turn to a list of applications of the above result to explicit examples, we make the connection with the definition of [9].
Dependence on Monodromy Data of the Isomonodromic Tau Function
569
5.1. Relationship between ω M and ω J MU . In order to establish the relationship we must “freeze” all the monodromy parts of the Birkhoff data, that is, the connection matrices, the Stokes matrices and the exponents of formal monodromy. As noted above, in this case = ω M . Then the observation that ω M and ω J MU coincide can be dug out [14] but we re-derive it here for the sake of being self-contained. If now ∂ is a derivative along the isomonodromic submanifold of the Birkhoff data, the differential reduces to an integral only on the Stokes’ lines, the toral circles and the formal-monodromy circles. Since the expression is repeated for each pole, we consider only the localization at one pole a = a j , i.e. we vary the toral data/position only of one pole. Let ∂ denote one such deformation involving only data at a = a j . Then
dx −1 (∂) = ω M (∂) = − Tr − (x)− (x)∂ M(x)M −1 (x) . (5.7) 2iπ There are two types of integrals here: the integrals along the Stokes lines and the integral around the toral circle. Along each Stokes contour the integrand can be equivalently written as follows: ⎞⎞ M(x) M −1 (x) # $% & # $% & ⎟⎟ ⎜ −1 ⎜ j (x) T ⎜ j (x) eTj (x) S −1 e−Tj (x) ⎟⎟ Tr ⎜ S j,ν e−T j (x) ∂ T j,ν ⎠⎠ ⎝− (x)− (x) ⎝∂ T j (x) − e ⎛
⎛
= Tr
: 0(traceless) + Tr(M (x)M (x)∂ T j (x)) (5.8) (5.9) T j (z) := T j (z) + L j ln(z − a j ),
−1 j (x) − +−1 (x)+ (x)∂ T j (x) − (x)− (x)∂ T
−1
where we have used the cyclicity of the trace on the second term followed by −1 (x)− (x)M(x) = +−1 (x)+ (x) − M −1 (x)M (x). M −1 (x)−
(5.10)
j (z) does not contain a logarithmic term because ∂ L j = 0 (only Recall also that ∂ T isomonodromic deformations are allowed). As a consequence of (5.8) the resulting sum of integrals along the Stokes tentacles (refer to Fig. 7) can be evaluated by excising an j (z)) on the solid black circle6 around the pole z = a j and integrating Tr( −1 (z) (z)∂ T contours indicated in Fig. 7, followed by a limit → 0. For fixed the integral along these solid “tentacles” is equal (by Cauchy’s theorem) to the counterclockwise integral along the circle and the clockwise integral along the bigger dashed circle indicated in Fig. 7, on the + side (inside) of the toral circle. −1 (x)∂ T (x)) on the right (outside) of We now have two integrals of Tr(− (x)− j the toral circle, coming from the definition of ω M minus the integral of Tr(+−1 (x)+ (x) j (x)) on the right (inside) coming from the above contour deformation: ∂T −1 j (x)) dx. Tr(− (x)− (x)∂ T j (x) − +−1 (x)+ ∂ T (5.11) − toral
Given the jump condition on the toral circle + (z) = − (z)e−T j (z) , we have −1 j (z)) = Tr(− j (z)) − T j (z)∂ T j (z). (z)− (z)∂ T Tr(+−1 (z)+ ∂(z)T
(5.12)
6 Note that (z) has a jump of the form e2iπ L j in the straight ray from a to the toral circle (Fig. 5), but the j (x)) does not have a jump there, since L j and T j are both diagonal. This is expression Tr( −1 (x) (x)∂ T the reason why we did not draw the corresponding solid line in Fig. 7.
570
M. Bertola
Fig. 7. The set of contours for the computation of ω J MU
j (z) = − Hence (5.11) reduces to (recall from (5.9 that ∂ T j (z) − ∂ T
L j ∂a j z−a j
)
0
%$# =− toral
& j (x)) + Tr(T j (x)∂ T
Tr toral
L j ∂a j −1 − (x)− (x) x − aj
dx , 2iπ
(5.13)
j (z) is a Laurent polynomial starting where the first integral is zero because T j (z)∂ T −2 from the term (z − a j ) . The remaining can be deformed within the annulus between the toral and formal-monodromy circle (and the − of the toral circle becomes the + of the local monodromy circle) and combines with the outer integral therein to give
L ∂a dx j j −1 −1 Tr + (x)+ (x) − − (x)− (x) a − x 2iπ j f or.mon. 2 2 Tr(L j )∂a j dx = 0. (5.14) =− 2 f or.mon. (x − a j ) 2iπ The remaining integrals can therefore be retracted to an integral around the circle:
j (x) dx . (5.15) Tr −1 (x) (x)∂ T 2iπ |x−a j |= In each sector we have the expansion valid for any N , ∞ N k N +1 k (z) = G j 1+ Y j,k (z − a j ) +O((z − a j ) ) ∼ G j 1+ Y j,k (z − a j ) , 0
0
(5.16)
Dependence on Monodromy Data of the Isomonodromic Tau Function
571
where the coefficient matrices G, Yk are the same irrespective of the sector. If we replace (z) in (5.15) by a suitably high truncation of the formal series, we commit an error that it is easily estimated to vanish as tends to zero. On the other hand the new integral is independent of and reduces to the formal residue,
j (z)∂ T j (z) dz , −1 (z)Y (∂; []) = ω M (∂; []) = res Tr Y j z=a j ∞ k j (z) := G j 1 + Y Y j,k (z − a j ) , (5.17) 0
to be understood simply as the coefficient of (z − a j )−1 of the above formal series with a finite Laurent tail. The reader acquainted with the definition of ω J MU will recognize that the expression is precisely the same defining ω J MU in [9] (of course one should repeat the residue computation at each pole and sum up). In conclusion we have shown Proposition 5.1. The (modified) Malgrange differential restricted to the manifold of isomonodromic deformations coincides with the Jimbo-Miwa-Ueno differential. 6. Applications and Examples 6.1. Painlevé II equation. We single out the second Painlevé for its relative simplicity in formulæ and as an illustration of potential applications to other Painlevé equations. We follow the setup in [6]. The Riemann–Hilbert description is shown in the picture, with the notations and condition * + *
+ −2i 43 z 3 +t z 1
0 L(s) := , U (s) := 1 s e , (6.1) 2i 4 z 3 +t z se 3 1 0 1 s1 − s2 + s3 + s1 s2 s3 = 0. (6.2) The matrices U (s), L(s) (where the letters stand mnemonically to remind one of the “upper” or “lower” triangularity) are the jump matrices for the RHP, and the parameter s depends on the ray under consideration as in Fig. 8. Condition 6.2 is the condition that the products of the jumps at the origin is the identity. The Riemann–Hilbert problem is then that of finding (z) uniformly bounded, with the indicated jumps and the normalization condition (z) ∼ 1 + O(z −1 ).
(6.3)
0 −1 0 1 (−z) (z) = 1 0 −1 0
(6.4)
The solution has the symmetry
which is seen by showing that if (z) solves the problem then so does the right hand side of (6.4). The matrix T (z) in the notation of Sect. 3 is given by 4i 3 1 0 T (z) = − , (6.5) z + it z σ3 , σ3 := 0 −1 3
572
M. Bertola
Fig. 8. The contours ( 1 , ..., 6 ) and jump matrices (with the notation in (6.1)) of the RHP associated to the second Painlevé transcendent
and the matrix (z) := (z)eT has constant jumps (on the anti-Stokes lines) and satisfies
(z)(z)−1 = −i 4z 2 + t + 2u 2 σ3 − 4uzσ2 − 2vσ1 −it − 2iu 2 −4i 0 2 0 4iu z + z+ = 0 4i −4iu −2v σ1 :=
0 1 1 0
, σ2 :=
0 −i i 0
, v = q (t).
−2v , it + 2iu 2 (6.6) (6.7)
Under an isomonodromic deformation the coefficients u, v in (6.6) must evolve as functions of t and –in particular– u = u(t; s) = 2 lim z 12 (z; t, s) z→∞
(6.8)
(the limit does not depend on the sector we choose) solves the second Painlevé equation d2 u = 2u 3 + tu. dt 2
(6.9)
By direct computations we have ω J MU =
du dt
2 − tu − u
∂t2 ln τ (t; s) = u(t; s)2 .
2
4
dt = ∂t ln τ (t; s) dt (6.10)
Dependence on Monodromy Data of the Isomonodromic Tau Function
573
Generically we can solve the condition (6.2) for s3 and use s1 , s2 as independent variables: more appropriately one should consider the parameter space as the algebraic manifold specified by (6.2). Introduce the matrix kernel K(x, y) :=
(x)−1 (y) . x−y
(6.11)
We then have 0 dx 0 e 2iπ
1 3 − i4 1 + s22 dx 3 x −it x 0 e −1 Tr (x) (x) +2 2 0 0 (1 + s1 s2 ) 3 2iπ i4 3 dx K12 (x, x)e 3 x +it x = −2 2iπ
1 1 + s22 i4 3 dx , (6.12) +2 K21 (x, x)e− 3 x −it x 2 (1 + s1 s2 ) 3 2iπ
∂s1 ln τ (t; s) = −2
Tr −1 (x) (x)
0
i4 3 3 x +it x
where the boundary–value indication is irrelevant since the indicated matrix element does not have a jump on the corresponding line and we have used the symmetry (6.4) which causes the factor of 2. Using formula (2.15) for the second derivatives and then (2.9), one may derive an integral formula for the derivative ∂t2 ∂s1 ln τ ; however, in view of (6.8) it is simpler to use the variational formula (2.8) directly, thus yielding ∂s1 u(t; s) = 2 lim z ∂s1 12 (z) z→∞ 0 0 −1 (x)(z) z (x) i43 x 3 +it x 0 e 12 dx = 4 lim z→∞ x−z 2iπ 1 i4 3 − 3 x −it x 0 e −1 (x)(z) z (x) 0 0 1 + s22 12 dx , −4 lim 2 z→∞ (1 + s1 s2 ) x−z 2iπ
3 (6.13) where we have used again the symmetry (6.4) to simplify the expression. Since (z) → 1 we have 1 + s22 i4 3 dx 2 − i4 x 3 −it x dx 3 − 4 . ∂s1 u(t; s) = 4 (x) e 12 (x)2 e 3 x +it x 11 (1 + s1 s2 )2 3 2iπ 2iπ
1 (6.14)
6.2. Variation of finite Toeplitz determinants for discontinuous symbols. Let S 1 = {|z| = 1}. Suppose that we have the following data: • A finite partition of S 1 in sub-arcs γ j , with β j being the separating points; • A collection of functions a j (z) : γ j → C which are analytic in a neighborhood of γ j .
574
M. Bertola
Define a(z) : S 1 → C as the piecewise analytic function that coincides with a j on each (interior of) γ j . The n th Toeplitz determinant is defined as dz µ j := z j a(z) Tn [a] := det[µ j−i ]i, j=1...n , , T0 [a] := 1. (6.15) 2iπ z |z|=1 It was shown in [1] that Tn [a] (and more general objects) are isomonodromic tau functions for the following RHP; 1 w(z) + (z) = , z ∈ S1, − (z) 0 1
(z) = 1 + O(z −1 ) z nσ3 , z → ∞, (z) = O(ln(z − β j )) , z → β j , w(z) := z n−1 a(z).
(6.16)
Note that depends on the integer n as well, and we will write n when needed to specify the value of this integer. In order to reduce to the general setting of Sect. 2 we should add a big counterclockwise circle (the “formal monodromy” circle) with jump z nσ3 and replace the asymptotics with (∞) = 1; to take care of the logarithmic growth at the endpoints β j we take –circles centered at the β j (with sufficiently small) and define (z) as (z) = (z) for z outside of the added circles, while inside we set w(x) dx 1 −C[w](z) , |z − β j | < , . (z) := (z) C[w](z) := 0 1 |x|=1 x − z 2iπ (6.17) Thus the problem (6.16) is recast to the equivalent Riemann–Hilbert problem (see Fig. 9), 1 z n a(z) , |z| = 1, min |z − β j | > , + (z) = − (z) 0 1 j 1 −C[w](z) |z − β j | = , + (z) = − (z) 0 1 + (z) = − (z)z nσ3 , |z| = R > 1,
z → ∞, (z) = 1 + O(z −1 ) ,
(6.18)
and (z) uniformly bounded in C. When necessary to specify the integer n, we will simply write n (z). Note that the additional jump on |z| = R (oriented counterclockwise) is independent of the symbol and hence undergoes no deformation. Therefore this has no impact in the definition of the differential n (•) = (•; [n ]). Note that the first column of the solution (z) to problem (6.18) consists of polynomials (of degree ≤ n) for |z| < R; indeed it has no jumps on |z| = 1 and the added small circles (due to the triangularity of the jump-matrices). Outside the circle |z| = R, said column divided by z n is bounded at infinity, which proves the assertion.
Dependence on Monodromy Data of the Isomonodromic Tau Function
575
Fig. 9. The modification to the RHP to guarantee that the solution is bounded everywhere: here C[w](z) := , w(x) dx x−z 2iπ and the new RHP has no jump within the small disk centered at β j
Similarly, the second row of −1 (z) (which solves a RHP with the jump matrix on the left) also consists of polynomials, by parallel arguments. Thus the integrand in the definition of the differential is in fact the integral of a polynomial kernel7 K21 (z, w) :=
−1 (z)(w) 21 z−w
(6.19)
of degree ≤ n − 1 in both variables (it has no jumps, it is regular on the diagonal z = w and bounded by a power at infinity separately in each variable). Now, the computations in [1] were done for an isomonodromic tau function, however the proof passes through without change to the extended -differential. Namely, the proof consisted in showing that n+1 − n = δ ln
Tn+1 [a] , Tn [a]
(6.20)
where n denotes the differential evaluated on the solution n of (6.18). Since the two RHPs differ by an elementary Schlesinger transformation of the type discussed in Sect. 2.3, this part is identical. Moreover, 0 ≡ 0 because the integrands in (2.26) and (2.7) are identically zero (all the matricesare upper triangular and ∂ is strictly upper triangular), which means that τ0 = exp 0 is a constant independent of the symbol, and can be taken to be 1. Since T0 [a] = 1 as well, this implies that δ ln Tn [a] = n ,
(6.21)
where δ is the total differential w.r.t. any parameters and may appear in the symbol and the positions of the endpoints. 7 A fact which is well known and key to matrix model computations.
576
M. Bertola
Dependence on the β j . The dependence on each of the β j is only in the jumps around the corresponding circle, and we have + * − j (β) 1 0 := w R (β j ) − w L (β j ), 2iπ z−β j ∂β j = , j (β) (6.22) 0 0 where the subscripts L ,R indicate the boundary value on the Left/Right of w along the contour, relative to the orientation of the contour, namely, is the jump of the symbol. This shows –a well–known fact– that j (β) − j (β) 1 dx = K21 (β j , β j ). ∂β j Tn [a] = − K21 (x, x) (6.23) 2iπ x − β j 2iπ 2iπ |x−β j |= Dependence on the symbol. Let ∂ denote a deformation of one of the functions a j ; then the jump matrices M(z) deforms only on the arc γ j and the small circles we have added, on a “dumb-bell” contour (Fig. 10). In particular, on the two circles the integral for reduces to dx , (6.24) K21 (x, x)C[∂w](x) 2iπ |z−β j |= x n ∂a j (x; s) dx C[∂w](z) = . (6.25) x−z 2iπ γj Since, K21 (x, x) is a analytic in the interior of the –circle (in fact, it is a polynomial!) then we can perform the integration by “contour deformation” leaving an integral of K21 (x, x) against the jump of C[∂w]. In conclusion, taking into consideration all the contours we have simply dx . (6.26) ∂ Tn [a] = K21 (x, x)x n ∂a j (x; s) 2iπ γj In general we have simply dx 0 x n ∂s a(x; s) ∂ ln Tn [a] = Tr −1 (x) (x) 0 0 2iπ |z|=1 dx = K21 (x, x)x n ∂s a(x; s) 2iπ −1 (x)(y) . K(x, y) := x−y
Fig. 10. The “dumbbell” contour
(6.27) (6.28) (6.29)
Dependence on Monodromy Data of the Isomonodromic Tau Function
577
Example 6.1. A particular case is if a j (x; s) = s j are constants, in which case the above reduces to an integral on the sub-arc γ j , dx ∂s j ln Tn [a] = K21 (x, x)x n . (6.30) 2iπ γj 6.3. Hankel and shifted Töplitz determinants. The notion of semiclassical symbols was defined in [12] and it was shown in [1] that the corresponding Hankel/Toeplitz determinant was an isomonodromic tau function, to within an explicit non-vanishing factor. The weights we are considering are all of the semiclassical type as defined in [2,4, 10,12,13]. This means that they are of the form µ(x) = e−V (x) with V (x) an arbitrary rational function. They are integrated over contours γ j which can be arbitrary contours in the complex plane as long as all integrals γ j x k µ(x) dx are convergent integrals. The range for k will be either N or Z, depending on the situation; a detailed description of the contours can be found in [2,3] and we refer thereto for a more detailed discussion. We will choose arbitrary complex constants κ j for each contour γ j and use the notation x k µ(x) dx := κj x k µ(x) dx = µk , (6.31) κ
γj
We will also use the notation κ : C → C to indicate the locally constant function that takes the (constant) value κ j on the corresponding contour γ j . The Markov function (sometimes referred to as a Weyl function) for these semiclassical weights is simply defined as the locally analytic function on C \ ∪γ j given by µ(ζ ) dζ , κ := W (x) := κ j χγ j (x). (6.32) κ ζ −x j
The function W (x) has logarithmic growth at the hard–edges, i.e. the endpoints of contours γ j , where µ is O(1). In this case one verifies that W (x) = O(ln |x − a|), where a is the hard–edge point. Consider the jump matrix supported on the contours γ j , 1 κ j e−V (z) M(z) := , z ∈ γj. (6.33) 0 1 Consider the following Riemann–Hilbert problems for a matrix = (n, ) parametrized by two integers n, + (z) = ⎧ − (z)M(z) 1 0 ⎪ ⎪ O(1) z→0 ⎨ 0 z n− −1 , (z) = n 0 ⎪ −1 )) z ⎪ z → ∞ (1 + O(z ⎩ 0 z − −1 and (z) = O(ln(z − a)) at any “hard-edge”.
(6.34)
578
M. Bertola
By definition of Schlesinger transformations (Sect. 2.3) we see that all these problems are successive Schlesinger transformations of the problem n = 0, = 1, which has the immediate solution −V (x) dx e 1 (0,1) κ x−z 2iπ (6.35) (z) = 0 1 It is also apparent that the differentials ω M (or , they coincide here because the term ϑ in Def. 2.2 vanishes identically, being the jumps 1+upper triangular) are closed differentials in the deformations of V , the endpoints and in the κ j ’s. Moreover, comparing the change ω M (∂; [ (n, ) ]) between two problems with n → n ± 1 or → ± 1 one sees exactly as in [1] that ∂ ln n = ω M (∂; [ (n, ) ]),
(6.36)
where n are the shifted Toeplitz determinants ⎞ ⎛ µ +1 · · · µ +n−1 µ µ · · · µ +n−2 ⎟ ⎜ µ −1 ⎟ ≡ 1 , ≡ 0. n := det ⎜ . .. −n ⎠ 0 ⎝ .. . µ −n+1
µ −n+2
···
µ
is –up to a Here the case of Hankel determinants corresponds to = n − 1; then n−1 n permutation of columns, hence a sign– a Hankel determinant. In particular the derivatives w.r.t. the parameters κ j are dx
∂κ j ln n = . (6.37) K (n, ) (x, x)e−V (x) 2iπ γj (n, ) −1 (n, ) (x) (y) 21 (n, ) . (6.38) K (x, y) = x−y This kernel is the usual Christoffel–Darboux kernel: for example, if = n − 1 (Hankel determinants) then K (n,n−1) (x, y) =
1 h n−1
pn (x) pn−1 (y) − pn−1 (x) pn (y) 1 = p j (x) p j (y), x−y hj n−1 j=0
(6.39) where p j are the monic orthogonal polynomials of exact degree j for the moment-functional, namely, p j (x) pk (x)e−V (x) dx = h j δ jk . (6.40) κ
Remark 6.1. Note that the RHP for (n, ) can be converted to a RHP with constant jumps V for := e− 2 σ3 and hence reduced to a rational ODE −1 = D(z). The resulting isomonodromic tau function is not exactly equal to n because the RHP for (n, ) and the RHP in the canonical form described in Sect. 4 differ by a gauge in the sense of Sect. 2.4. However the corresponding factor is easy to compute and this was accomplished in [1].
Dependence on Monodromy Data of the Isomonodromic Tau Function
579
We conclude with the remark that the variational formulæ for n are valid for very general weights, not necessarily of semiclassical type. Indeed we are really using only the results of Sect. 2, where the connection to rational ODE is not necessary. Acknowledgements. The author thanks Jacques Hurtubise and John Harnad for helpful discussion. The work was partially supported by the Natural Sciences and Engineering Research Council of Canada (NSERC).
References 1. Bertola, M.: Moment determinants as isomonodromic tau functions. Nonlinearity 22(1), 29–50 (2009) 2. Bertola, M., Eynard, B., Harnad, J.: Semiclassical orthogonal polynomials, matrix models and isomonodromic tau functions. Commun. Math. Phys. 263(2), 401–437 (2006) 3. Bertola, M., Gekhtman, M.: Biorthogonal Laurent polynomials, Töplitz determinants, minimal Toda orbits and isomonodromic tau functions. Constr. Approx. 26(3), 383–430 (2007) 4. Bertola, M.: Bilinear semiclassical moment functionals and their integral representation. J. Approx. Theory 121(1), 71–99 (2003) 5. Boalch, P.: Symplectic manifolds and isomonodromic deformations. Adv. Math. 163(2), 137–205 (2001) 6. Its, A., Kapaev, A.: The Nonlinear Steepest Descent Approach to the Asymptotics of the Second Painlevé Transcendent in the Complex Domain. Volume 23 of Progr. Math. Phys. Boston, MA: Birkhäuser Boston, 2002 7. Jimbo, M., Miwa, T.: Monodromy preserving deformation of linear ordinary differential equations with rational coefficients. II. Phys. D 2(3), 407–448 (1981) 8. Jimbo, M., Miwa, T.: Monodromy preserving deformation of linear ordinary differential equations with rational coefficients. III. Phys. D 4(1), 26–46 (1981) 9. Jimbo, M., Miwa, T., Ueno, K.: Monodromy preserving deformation of linear ordinary differential equations with rational coefficients. I. General theory and τ -function. Phys. D 2(2), 306–352 (1981) 10. Magnus, A.P.: Painlevé-type differential equations for the recurrence coefficients of semi-classical orthogonal polynomials. In: Proceedings of the Fourth International Symposium on Orthogonal Polynomials and their Applications (Evian-Les-Bains, 1992), Volume 57, London: Elsevier, 1995, pp. 215–237 11. Malgrange, B.: Sur les déformations isomonodromiques. I. Singularités régulières. In: Mathematics and Physics (Paris, 1979/1982), Volume 37 of Progr. Math., Boston, MA: Birkhäuser Boston, 1983, pp. 401–426 12. Marcellán, F., Rocha, I.A.: On semiclassical linear functionals: integral representations. In: Proceedings of the Fourth International Symposium on Orthogonal Polynomials and their Applications (Evian-Les-Bains, 1992), Volume 57, London: Elsevier, 1995, pp. 239–249 13. Marcellán, F., Rocha, I.A.: Complex path integral representation for semiclassical linear functionals. J. Approx. Theory 94(1), 107–127 (1998) 14. Palmer, J.: Zeros of the Jimbo, Miwa, Ueno tau function. J. Math. Phys. 40(12), 6638–6681 (1999) 15. Schlesinger, L.J.: Über die Lösungen gewisser Differentialgleichungen als Funktionen der singulären Punkte. J. Reine Angew. Math. 141, 96–145 (1912) 16. Sibuya, Y.: Linear Differential Equations in the Complex Domain: Problems of Analytic Continuation, Volume 82 of Translations of Mathematical Monographs. Providence, RI: Amer. Math. Soc., 1990, Translated from the Japanese by the author 17. Wasow, W.: Asymptotic Expansions for Ordinary Differential Equations. New York: Dover Publications Inc., 1987, reprint of the 1976 edition Communicated by M. Aizenman
Commun. Math. Phys. 294, 581–603 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0930-1
Communications in
Mathematical Physics
On the Relationship Between Continuousand Discrete-Time Quantum Walk Andrew M. Childs Department of Combinatorics & Optimization and Institute for Quantum Computing, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, Canada N2L 3G1. E-mail:
[email protected] Received: 13 April 2009 / Accepted: 11 August 2009 Published online: 10 October 2009 – © Springer-Verlag 2009
Abstract: Quantum walk is one of the main tools for quantum algorithms. Defined by analogy to classical random walk, a quantum walk is a time-homogeneous quantum process on a graph. Both random and quantum walks can be defined either in continuous or discrete time. But whereas a continuous-time random walk can be obtained as the limit of a sequence of discrete-time random walks, the two types of quantum walk appear fundamentally different, owing to the need for extra degrees of freedom in the discrete-time case. In this article, I describe a precise correspondence between continuous- and discretetime quantum walks on arbitrary graphs. Using this correspondence, I show that continuous-time quantum walk can be obtained as an appropriate limit of discrete-time quantum walks. The correspondence also leads to a new technique for simulating Hamiltonian dynamics, giving efficient simulations even in cases where the Hamiltonian is not sparse. The complexity of the simulation is linear in the total evolution time, an improvement over simulations based on high-order approximations of the Lie product formula. As applications, I describe a continuous-time quantum walk algorithm for element distinctness and show how to optimally simulate continuous-time query algorithms of a certain form in the conventional quantum query model. Finally, I discuss limitations of the method for simulating Hamiltonians with negative matrix elements, and present two problems that motivate attempting to circumvent these limitations. 1. Introduction Recently, quantum walk has been established as one of the dominant algorithmic techniques for quantum computers. In the black-box setting, quantum walk provides exponential speedup over classical computation [13,18]. Moreover, many quantum walk algorithms achieve polynomial speedup over classical computation for problems of practical interest. Following the development of quantum walk algorithms for search on graphs [16,48] (subsequently improved in [8,17,52]), Ambainis cemented the
582
A. M. Childs
importance of quantum walk by giving an optimal algorithm for the element distinctness problem [5]. This approach was later generalized [36,51] and applied to quantum algorithms for triangle finding [37], checking matrix multiplication [10], and testing group commutativity [35]. More recently, Farhi, Goldstone, and Gutmann used quantum walk to develop an optimal algorithm for evaluating balanced binary game trees [25], which led to optimal and near-optimal algorithms for evaluating broad classes of formulas [7,43]. Indeed, since Grover’s well-known search algorithm [30] can be interpreted as a quantum walk on the complete graph, nearly all known quantum algorithms achieving polynomial speedup over classical computation can be viewed as quantum walk algorithms. Quantum walk is defined by analogy to classical random walk, and is motivated by the ubiquity of random walk in classical randomized computation. Random walks, or Markov chains, can either evolve continuously in time or by a discrete sequence of steps. The relationship between these two notions of random walk is straightforward, and indeed it is often possible to analyze both settings simultaneously (see for example [4]). A discrete-time Markov chain with N states is specified by an N × N stochastic matrix M, a matrix with nonnegative entries whose columns sum to 1. One step of the walk transforms an initial probability distribution p ∈ R N into a new probability distribution p = M p. To approach a continuous-time Markov process, consider a lazy random walk in which we only make a transition with some small probability > 0, replacing M by M + (1 − )I . Then we have p − p = (M − I ) p.
(1)
In the limit → 0, letting one step of the discrete-time walk correspond to a time interval , we obtain the dynamics d p(t) = (M − I ) p(t), dt
(2)
a continuous-time Markov process generated by M − I . In fact, any continuous-time Markov chain with N sites can be written in the form (2) for some N × N stochastic matrix M, up to a rescaling of the time variable.1 Similarly, one can define two notions of quantum walk, the continuous-time quantum walk [15,27] and the discrete-time quantum walk [2,6,53]. But in contrast to the classical case, these two models are apparently incomparable. The continuous-time quantum walk on an undirected graph is obtained by replacing the diffusion equation (2) with the Schrödinger equation i
d q(t) = H q(t). dt
(3)
Here the Hamiltonian H is an N × N Hermitian matrix with H jk = 0 if and only if vertices j and k are connected, such as the adjacency matrix or Laplacian of the graph, and q(t) ∈ C N is a vector of complex amplitudes, one for each vertex. The state space 1 For example, given a graph with adjacency matrix A, maximum degree d, and a diagonal matrix of vertex degrees D (with D j j = deg( j)), the continuous-time random walk generated by the normalized Laplacian (A − D)/d (a negative semidefinite operator) can be viewed as the limit of the discrete-time random walk with transition matrix M = A/d + (I − D/d). The simple discrete-time random walk, in which a transition is made to a randomly selected neighbor at each step, has the transition matrix M = AD −1 , and corresponds to the continuous-time random walk generated by (A − D)D −1 . For a d-regular graph, D = d I , so these two choices are equivalent.
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
583
for the quantum walk is the same as for the corresponding classical random walk, and the dynamics are a direct quantum analog of (2). A step of a discrete-time quantum walk is a unitary operation that moves amplitude between adjacent vertices of a graph. Unlike in continuous time, there is no general definition of a discrete-time quantum walk that takes place directly on the vertices. In particular, there is no translation-invariant discrete-time unitary process on a d-dimensional lattice [38,39], and in fact, only graphs with special properties can support local unitary dynamics at all, even without homogeneity restrictions [47]. However, by enlarging the state space to include extra degrees of freedom, sometimes referred to as a quantum coin—or equivalently, by considering a walk on the directed edges of a graph—this limitation can be overcome [53]. Despite these differences, examples of continuous- and discrete-time quantum walks on graphs show many qualitative similarities. For example, the walks on the line both spread linearly in time [6,27], and the walks on the hypercube have the same instantaneous mixing time [41]. Furthermore, some quantum algorithms have been constructed that behave similarly in the two models (compare [16,17] to [8,48], and [25] to [7]). However, since the discrete-time walk takes place on a larger state space, it cannot simply reduce to the continuous-time walk as a limiting case. Indeed, no general correspondence between the dynamics of the two models has been given previously (although a particular correspondence has been noted for the infinite line and for the three-dimensional square lattice [50]). From a computational perspective, the two types of quantum walk each have their own advantages. As it does not require enlarging the state space, the continuous-time model is arguably more natural; it is simpler to define and usually easier to analyze. For example, while the exponential speedup by continuous-time quantum walk described in [13] is generally suspected to carry over to the discrete-time model, the dynamics are more involved, and to the best of my knowledge no rigorous analysis has been provided so far. On the other hand, the discrete-time model has the advantage that it is generally easy to implement using quantum circuits, whereas the main technique for implementing a continuous-time quantum walk requires the maximum degree of the graph to be small [3,9,12,13]. Thus, for example, the element distinctness algorithm [5] seems difficult to reproduce using continuous-time quantum walk. In this article, I explore the relationship between continuous- and discrete-time quantum walk, showing how to carry the desirable features of each model to the other. The foundation of the results is a formal correspondence between the two models described in Sect. 2. This correspondence employs the quantization of discrete-time Markov chains proposed by Szegedy [51], which can also be used to give a discrete-time quantum walk corresponding to any Hermitian matrix, generalizing a construction of [7,43]. Using this correspondence, Sect. 3 introduces a discrete-time quantum walk whose behavior approaches that of a related continuous-time quantum walk in a certain limit. In addition to providing a conceptual link between the two models, this limit offers a generic means of converting continuous-time quantum walk algorithms into discretetime ones. For example, it shows that there is an efficient discrete-time quantum walk algorithm for the graph traversal problem of [13]. More significantly, I apply the correspondence to the simulation of Hamiltonian dynamics by quantum circuits, giving more efficient simulations than have been known so far. For a sparse Hamiltonian, it is well-known that the evolution according to (3) for time t can be simulated in t 1+o(1) steps [9,12]. By reduction to the problem of computing parity, reference [9] established what might be called a no fast-forwarding theorem:
584
A. M. Childs
in general, a Hamiltonian cannot be simulated for time t using a number of steps that is sublinear in t. However, this left open the question of whether a truly linear-time simulation is possible. Moreover, even less has been known about the case where the Hamiltonian is not necessarily sparse, as introduced in Sect. 4. I show in Sect. 5 that by applying phase estimation to a discrete-time quantum walk, one can simulate Hamiltonian dynamics for time t using O(t) operations, not only in the sparse case, but for even more general succinctly specified Hamiltonians. As an example of this method in action, I give an optimal continuous-time quantum walk algorithm for element distinctness (Sect. 6). I also describe an application to the simulation of continuous-time query algorithms using conventional quantum queries (Sect. 7). Unfortunately, although the new simulation method can be efficient well beyond the case of sparse Hamiltonians, it can sometimes be inefficient if the Hamiltonian is not only dense, but also includes matrix elements of both signs (or, more generally, if it has complex entries). I conclude in Sect. 8 by posing some examples of Hamiltonians the method fails to address, but whose efficient simulation would lead to new quantum algorithms. 2. A Discrete-Time Quantum Walk for any Hamiltonian We begin by constructing a discrete-time quantum walk corresponding to an arbitrary Hermitian matrix, i.e., to the Hamiltonian of any finite-dimensional quantum system. This construction is based on the work of Szegedy, who defined a discrete-time quantum walk corresponding to an arbitrary discrete-time classical Markov chain [51]. It was observed in [7] that Szegedy’s framework can be used to connect continuous- and discrete-time quantum walk, assuming that the continuous-time quantum walk is generated by an entrywise positive, symmetric matrix. In [43] this was generalized to an arbitrary Hermitian matrix whose graph of nonzero entries is bipartite. Here, we describe a discrete-time quantum walk corresponding to a general N × N Hermitian matrix H . Fix an orthonormal basis {| j : j = 1, . . . , N } of C N , and let abs(H ) := N j,k=1 |H jk | | jk| denote the elementwise absolute value of H in that basis. Let N |d := j=1 d j | j be a principal eigenvector of abs(H ) (i.e., an eigenvector with eigenvalue abs(H )). If abs(H ) is irreducible, then by the Perron-Frobenius theorem, |d is unique up to a phase, and that phase can be chosen so that the entries d j are strictly positive. We focus on the irreducible case without loss of generality, as if abs(H ) is reducible, we can treat each of its irreducible components separately. Note that abs(H ) ≥ H , as can easily be proved using the triangle inequality. Define a set of N quantum states |ψ1 , . . . , |ψ N ∈ C N ⊗ C N as N 1 ∗ dk | j, k. H jk (4) |ψ j := √ dj abs(H ) k=1 It is straightforward to check that these states are orthonormal. Notice that j, k|ψ j = 0 if and only if j is adjacent to k in the graph of nonzero entries of H . The discrete-time quantum walk corresponding to H is defined as the unitary operator obtained by first reflecting about span{|ψ j } and then exchanging the two registers with the swap operation S (i.e., S| j, k = |k, j). Equivalently, two steps of the walk can be viewed as first reflecting about span{|ψ j }, and then reflecting about span{S|ψ j }. Szegedy showed that the spectrum of a product of reflections depends in a simple way
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
585
on the matrix of inner products between orthonormal bases for the two subspaces [51, Theorem 1].2 In the present case, we have ψ j |S|ψk =
H jk abs(H )
(5)
(using the fact that H = H † ), so we obtain a walk corresponding to H . More concretely, let T :=
N
|ψ j j|
(6)
j=1
be the isometry mapping | j ∈ C N to |ψ j ∈ C N ⊗ C N . Then T T † is the projector onto span{|ψ j }, so the walk operator described above is S(2T T † − 1). We have Theorem 1. Suppose that
H abs(H ) |λ
= λ|λ. Then the unitary operator
U := iS(2T T † − 1)
(7)
has two normalized eigenvectors |µ± :=
1 − e±i arccos λ S T |λ 2(1 − λ2 )
(8)
with eigenvalues µ± := ±e±i arcsin λ . This result follows directly from [51, Theorem 1], but the proof can be simplified since the walk is defined using the swap operation. We provide a proof here for the sake of completeness, along the same lines as [7, Theorem 6]. Proof. Consider the action of U on the vector T |λ. Using T † T = 1, we have U T |λ = iST |λ,
(9)
and using T † ST = H/abs(H ), we have U ST |λ = 2iλST |λ − iT |λ.
(10)
Thus, the unnormalized state |µ := T |λ + iµST |λ has U |µ = µT |λ + i(1 + 2iλµ)ST |λ, and is an eigenvector of U with eigenvalue µ provided 1 + 2iλµ = µ2 , i.e., µ = ± 1 − λ2 + iλ = ie∓i arccos(λ) = ±e±i arcsin(λ)
(11)
(12)
as claimed. The normalization follows from µ|µ = 1 + iλ(µ − µ∗ ) + |µ|2 = 2(1 − λ2 ), which completes the proof. 2 Note that this can also be viewed as a consequence of a classic result of Jordan [42].
(13)
586
A. M. Childs
3. Continuous-Time Walk as a Limit of Discrete-Time Walks Using the correspondence described in the previous section, we can construct a discrete-time quantum walk whose behavior reproduces that of the continuous-time quantum walk generated by H in an appropriate limit. First we construct a discretetime process that approximates the continuous-time one provided its eigenvalues are sufficiently small; then we construct a “lazy quantum walk” to obtain small eigenvalues. Let denote the projector onto the subspace span{T | j, ST | j : j = 1, . . . , N }, and consider the action of U restricted to this (invariant) subspace. If all eigenvalues λ of H/abs(H ) are small, then arcsin λ ≈ λ, meaning that the eigenvalues of H are approximately linearly related to the √ eigenphases of U . Furthermore, the eigenvectors of U are |µ± ≈ (1 ∓ iS + λS)T |λ/ 2, and we have 1 ∓ iS + λS 1 ± iS + λS iU ≈ ∓e±iλ T |λλ|T † . (14) √ √ 2 2 λ,± This expression appears similar to the spectral expansion of the evolution according to H/abs(H ) for unit time, e−iλ |λλ|, (15) e−iH/abs(H ) = λ
except for (i) application of the isometry T , (ii) the presence of both positive and negative phases for each eigenvalue λ, and (iii) rotations by approximately a square root of√iS, √ namely (1 ± iS)/ 2. To obtain only the ‘−’ terms, we rotate the basis by (1 + iS)/ 2. Then we find 1 − iS 1 + iS T † √ (iU )τ √ T ≈ e−iτ H/abs(H ) , (16) 2 2 where the calculation has been carried out to O(λ). More precisely, we have the following approximation: Theorem 2. Let h := H /abs(H ). Then † 1 − iS + iS −iτ H/abs(H ) T √ (iU )τ 1√ ≤ h 2 1 + ( π2 − 1)hτ . (17) T − e 2 2 Proof. By Theorem 1, we have ±i arccos λ S ∓i arccos λ S √ √ iU = λ,± ∓e±i arcsin λ 1−e T |λλ|T † 1−e . 2(1−λ2 )
2(1−λ2 )
(18)
Then a direct calculation gives 1 − iS 1 + iS T † √ (iU )τ √ T 2 2 1 ∓e±iτ arcsin λ † = T [(1 + ie±i arccos λ ) − (i + e±i arccos λ )S]T |λλ| 4 1 − λ2 λ,± T † [(1 − ie∓i arccos λ ) + (i − e∓i arccos λ )S]T 1 ∓e±iτ arcsin λ = [(1 + ie±i arccos λ ) − (i + e±i arccos λ )λ] 4 1 − λ2 λ,± [(1 − ie∓i arccos λ ) + (i − e∓i arccos λ )λ]|λλ| 1 = [−e+iτ arcsin λ (1 − 1 − λ2 ) + e−iτ arcsin λ (1 + 1 − λ2 )]|λλ|. 2 λ
(19)
(20) (21)
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
587
√ Now we can use the inequality 1 − 1√− λ2 ≤ λ2 to bound the norm of the first term by λ2 /2. By the same inequality, |(1 + 1 − λ2 ) − 2|/2 ≤ λ2 /2. Thus we have † 1 − iS + iS −iτ arcsin(H/abs(H )) T √ (iU )τ 1√ ≤ h2. (22) T − e 2 2 Using the inequalities |λ − arcsin λ| ≤ ( π2 − 1)|λ|3 and |1 − e−iθ | ≤ |θ |, we find −iτ arcsin(H/abs(H )) − e−iτ H/abs(H ) ≤ ( π2 − 1)τ h 3 , (23) e and (17) follows by the triangle inequality. To obtain an arbitrarily good approximation of the dynamics, we require a systematic means of making h = H /abs(H ) arbitrarily small. In other words, we must construct a lazy quantum walk, analogous to the lazy random walk that only takes a step with some small probability as in (1). To do this, we enlarge the Hilbert space and modify the states |ψ j from (4) to √ √ (24) |ψ j := |ψ j + 1 − |⊥ j for some ∈ (0, 1], where {|⊥ j : j = 1, . . . , N } are orthonormal states satisfying (25) ψ j |⊥k = ψ j |S|⊥k = ⊥ j |S|⊥k = 0 for j, k = 1, . . . , N . Correspondingly, we modify the isometry T to T := j |ψ j j|. Using the fact that ψ j |H |ψk = ψ j |H |ψk , it can be shown that Theorems 1 and 2 still hold, but with H replaced by H , so that λ is replaced by λ and h is replaced by h. Note that it is not necessary for the states |ψ j to arise from some Hermitian operator as in (4). Perhaps the simplest choice is to enlarge the Hilbert space from C N × C N to C N +1 × C N +1 , and let |⊥ j = | j, N + 1. Overall, we obtain a procedure for simulating a continuous-time walk by a discretetime one, as follows: 1. Given an initial state |φ0 ∈ span{| j : j = 1, . . . , N }, apply the isometry T and √ . the unitary operation 1+iS 2
2. Apply τ steps of the discrete-time quantum walk U := iS(2T T† − 1). √ T | j : j = 1, . . . , N }. 3. Project onto the basis of states { 1+iS 2
By Theorem 2 and the preceding discussion, the resulting outcomes are approximately distributed according to Pr( j) = | j|e−iτ H/abs(H ) |φ0 |2 , provided 2 h 2 and 3 h 3 τ are small. To obtain a total evolution time t, we choose = abs(H )t/τ (subject to the constraint ≤ 1), where the accuracy can be improved by increasing the number of simulation steps τ , hence decreasing . More precisely, suppose our goal is to simulate the continuous-time quantum walk according to H for a total time t, obtaining a final state |φt satisfying e−iH t |φ0 − |φt ≤ δ. Then it suffices to take ⎫ ⎧ ⎬ ⎨ 1 + ( π2 − 1)H t , abs(H )t . (26) τ ≥ max H t ⎭ ⎩ δ √ In other words, the complexity of the simulation is O (H t)3/2 / δ, abs(H )t .
588
A. M. Childs
Note that to implement the discrete-time quantum walk defined above, it must be possible to implement the isometry T (and its inverse). In particular, this requires computing ratios of nonzero components of the principal eigenvector |d, a global property of abs(H ). While this may be difficult in general, it is tractable in many cases of interest. If the graph is regular, then the principal eigenvector is simply the uniform superposition; if it is non-regular but sufficiently structured, then it may be possible to compute |d explicitly. For some applications it may be sufficient to precompute the principal eigenvector, as in [7,43]. Alternatively, another option is to replace the states in (4) by N N |H jk | ∗ |ψ j := H jk | j, k + 1 − |⊥ j (27) H 1 H 1 k=1
k=1
N for some ∈ (0, 1], where H 1 := max j k=1 |H jk |, and where the states |⊥ j N ∗ satisfy (25) with k=1 H jk | j, k playing the role of |ψ j . With this choice, no information about the principal eigenvector of abs(H ) is needed to implement the isometry T := Nj=1 |ψ j j|, and we retain the necessary properties that ψ j |ψk = δ j,k and ψ j |S|ψk ∝ H jk . However, since the quantity abs(H ) is replaced by the maximum absolute column sum norm H 1 ≥ abs(H ), the simulation may be less efficient (and in particular, still suffers from the sign problem discussed in Sect. 8). This simulation immediately shows that the exponential speedup by continuous-time quantum walk demonstrated in [13] carries over to the discrete-time model. (This is a case where the graph is slightly non-regular, but the principal eigenvector can be computed explicitly.) On the other hand, the speedup demonstrated in [18] apparently does not carry over, as abs(H ) H for the relevant Hamiltonian, and exponentially many steps of the discrete-time quantum walk would be required. We will return to this issue in Sect. 8. The estimate (26) may be overly pessimistic for some applications. To emulate the dynamics of the entire Hamiltonian, we choose so that all eigenvalues are small. But for algorithms that effectively work in a low-energy subspace (e.g., [16,17]), or that only depend on a small spectral gap (e.g., [7,43]), it may be feasible to use little or no rescaling. 4. Simulating Non-sparse Hamiltonians We now turn to the problem of simulating Hamiltonian dynamics. In this section, we review known results about the simulation of sparse Hamiltonians, introduce the problem of simulating Hamiltonians that are not necessarily sparse, and describe some preliminary results on the simulation of non-sparse Hamiltonians. Given a description of some Hermitian matrix H , Hamiltonian simulation is the problem of implementing the unitary operator e−iH t by a quantum circuit for any desired value of t. Efficient simulations of sparse Hamiltonians are well-known. In the special case that H is local—i.e., when it is a sum of terms, each acting on a constant number of qubits—efficient simulation is straightforward [33]. More generally, a Hamiltonian can be simulated efficiently provided it is sparse and efficiently row-computable. We say a Hermitian operator H acting on C N is sparse in the basis {| j : j = 1, . . . , N } if for any j ∈ {1, . . . , N }, there are at most poly(log N ) values of k ∈ {1, . . . , N } for
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
589
which j|H |k is nonzero. We say it is efficiently row-computable if there is an efficient procedure for determining those values of k together with the corresponding matrix elements. The simulability of sparse Hamiltonians was first explicitly stated in [3]; it also follows from [13] together with classical results on local coloring of graphs [32]. The main idea is as follows. Suppose we color the edges of the graph of nonzero entries of H ; then the subgraphs of any particular color consist of isolated edges, meaning that the evolution on any one of these subgraphs takes place in isolated two-dimensional subspaces, and is easily simulated. The subgraphs can be recombined using approximations to the Lie product formula, n lim e−i At/n e−iBt/n = e−i(A+B)t . (28) n→∞
As the lowest-order approximation, supposing A, B ≤ h, we have −i At/n −iBt/n n e − e−i(A+B)t = O((ht)2 /n), e
(29)
which shows that n = O((ht)2 /δ) simulation steps suffice to achieve error at most δ. Similarly, at second order, −i At/2n −iBt/n −i At/2n n e e − e−i(A+B)t = O((ht)3 /n 2 ), (30) e √ so that n = O((ht)3/2 / δ) simulation steps suffice. With approximations of increasingly high order, a simulation can be performed in nearly linear time [9,12]. Ultimately, we have Theorem 3. Suppose the graph of nonzero entries of H has N vertices and maximum degree d, and that H is efficiently row-computable, with | j|H |k| ≤ h for all j, k ∈ {1, . . . , N }. Then evolution according to H for time t with error at most δ can be simulated in ht (ht/δ)o(1) · poly(d, log N ) steps. It is natural to ask under what conditions we can simulate a non-sparse Hamiltonian. Notice that if a Hamiltonian is not sparse, then it cannot be efficiently row-computable, simply because we cannot write down the nonzero entries of a row in polynomial time. However, we can still suppose that the Hamiltonian has a succinct description of some kind and ask whether an efficient simulation is possible. One way to simulate the dynamics of a non-sparse Hamiltonian is to use information † about its spectrum. By the simple identity e−iH t = U e−iU HU t U † , we can simulate † H = U DU (with D diagonal) provided we can efficiently perform a unitary transformation U mapping the standard basis vector | j to the j th eigenvector of H , and efficiently compute the j th eigenvalue of H , under some canonical ordering. For example, this approach can be used when H is the adjacency matrix of a complete graph, a complete bipartite graph, or a star graph (and even for some more complicated cases, e.g., the Winnie Li graph in odd dimensions [18]), all of which have high maximum degree. However, it is unclear how broadly this strategy can be applied. Here, we are interested in simulations based on the graph structure of the Hamiltonian in a fixed basis. To begin, suppose the Hamiltonian is the adjacency matrix of an N -vertex graph G. We say that such a Hamiltonian is efficiently index-computable if, for any j ∈ {1, . . . , N }, it is possible to efficiently compute deg( j) and to compute the k th neighbor of j (under some canonical ordering) for any index k ∈ {1, . . . , deg( j)}.
590
A. M. Childs
The discrete-time quantum walk on an unweighted graph G is straightforward to implement with such a description. In particular, given a black box for the degree and the jth neighbors of any given vertex, a step of the discrete-time walk on G can be simulated using only four queries (to compute the degree, compute a neighbor, uncompute the neighbor, and uncompute the degree). More generally, we can consider graphs with edge weights under certain conditions mentioned below. As a preliminary observation, we can obtain efficient simulations of some non-sparse Hamiltonians by generalizing the notion of edge coloring. Suppose we have a decomposition of the graph of nonzero entries of the Hamiltonian into a union of polynomially many subgraphs, each of which is a disjoint union of simulable subgraphs; then we can efficiently simulate the dynamics using the Lie product formula. Theorem 3 can be viewed as the special case where the simulable subgraphs are single edges. As a novel example of this technique, there is an efficient simulation of Hamiltonians whose graphs are trees: Theorem 4. Suppose the graph of nonzero entries of H is a rooted, efficiently index-computable tree, with the parent of any vertex having a known index. Furthermore, suppose that the weights on the edges from any given vertex are integrable, in the sense of [29] (which holds trivially for an unweighted graph), and that for any vertex we can efficiently compute the distance to the root. Then the evolution according to H for time 1+o(1) 2 t can be simulated in O((ht) ) steps, where h := max j k |k|H | j| . Proof. Consider a decomposition of the tree into two forests of stars. The first forest contains the complete star around the root and stars centered at all vertices an even distance from the root, including all child vertices (but not the parent vertex) in each star. The second forest contains the stars around vertices an odd distance from the root, again including all child vertices but not the parent vertex. The result is a decomposition H = H1 + H2 , where the graphs of nonzero entries of H1 and H2 are both forests of stars. A single star with center vertex 0 and weight wi on the edge (0, i) for i = 1, . . . , k has exactly two nonzero eigenvalues, λ± := ± |w1 |2 + · · · + |wk |2 , with corresponding normalized eigenvectors k 1 1 wi |i . √ |0 + λ± 2 i=1
(31)
To transform from the standard √ basis to the eigenbasis√of the star, we can perform the operation |0 → (|0 + |1)/ 2, |1 → (|0 − |1)/ 2, |i → |i for i = 2, . . . , k, followed by any operation satisfying |0 → |0 and |1 → i wi |i/λ+ ; the latter can be implemented by the techniques of [29]. By applying this simulation for all disjoint stars in one of the forests, we can simulate the dynamics of either H1 or H2 . Finally, using a high-order approximation of the Lie product formula, we can simulate the full Hamiltonian H with nearly linear overhead. It is not clear how to extend this strategy to general graphs. However, we will see next that the correspondence with discrete-time quantum walk gives a broadly applicable simulation method.
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
591
5. Linear-Time Hamiltonian Simulation We now present the main result of the paper, a method of simulating Hamiltonian dynamics that exploits the connection between continuous- and discrete-time quantum walk. One such approach is to simply apply Theorem 2, which states that any Hamiltonian dynamics can be well-approximated by a corresponding discrete-time process. However, with that approach, the number of steps used to simulate evolution for time t scales as t 3/2 . Instead, we can obtain a linear-time simulation by combining Theorem 1 with phase estimation. By the no fast-forwarding theorem [9], this scaling is optimal, even for sparse Hamiltonians. However, we do not require the Hamiltonian to be sparse. The simulation of an N × N Hamiltonian H proceeds as follows. Given an input state |ψ = ψλ |λ (32) λ
in C N , where |λ denotes an eigenvector of H/abs(H ) with eigenvalue λ, apply the isometry T defined in (6) to create the state T |ψ ∈ C N ⊗ C N . This state may be written ψλ T |λ (33) T |ψ = λ
=
λ
ψλ
,
−i arccos λ i arccos λ 1−λe √ √ |µ+ + 1−λe |µ− 2(1−λ2 ) 2(1−λ2 )
(34)
where we have used (8) to write each T |λ in terms of the corresponding eigenvectors |µ± of the discrete-time quantum walk U corresponding to H . Our goal is to introduce a phase e−iλt for the λ term of this superposition. To this end, perform coherent phase estimation on the discrete-time quantum walk U corresponding to H . Recall from Theorem 1 that the eigenvalues of U corresponding to the eigenvalue λ of H/abs H are ±e±i arcsin λ . Of course, we are not directly interested in arcsin(λ), ˜ but it implicitly determines λ. Given an estimate λ˜ ≈ λ, we induce the phase e−iλt , uncompute λ˜ by performing phase estimation in reverse, and finally apply the inverse isometry T † . Overall, we claim that this procedure achieves the following: Theorem 5. A Hamiltonian H can be simulated for time t with fidelity at least 1 − δ √ using O(abs(H )t/ δ) steps of the corresponding discrete-time quantum walk (7). Proof. Let |θ denote an eigenstate of the discrete-time quantum walk with eigenvalue eiθ . Let P denote the isometry that performs phase estimation on this walk, appending a register with an estimate of the phase, as follows: aφ|θ |θ, φ. (35) P|θ = φ
Here aφ|θ is the amplitude for the estimate φ when the eigenvalue is in fact θ ; in particular, Pr(φ|θ ) = |aφ|θ |2 . Let Ft be the unitary operation that applies the desired phase, namely Ft |θ, φ = e−it sin φ |θ, φ. Then our simulation of the Hamiltonian evolution e−iH t is T † P † Ft P T .
(36)
592
A. M. Childs
Given an input state |ψ as in (32), we compute the inner product between the ideal state e−iH t |ψ and the simulated state T † P † Ft P T |ψ. Similarly to (34), we have P T e−iH t |λ = e−iλt P T |λ −i arccos λ i arccos λ √ √ = e−iλt P 1−λe |µ+ + 1−λe |µ− 2 2 2(1−λ ) 2(1−λ ) −i arccos λ −iλt 1−λe =e aφ|arcsin λ √ |µ+ , φ √ +aφ|π −arcsin λ 1−λe
i arccos λ
2(1−λ2 )
Ft P T |λ =
φ
(38)
2(1−λ2 )
φ
and
(37)
|µ− , φ ,
(39)
−i arccos λ √ e−it sin φ aφ|arcsin λ 1−λe |µ+ , φ 2 2(1−λ )
√ + aφ|π −arcsin λ 1−λe
i arccos λ
2(1−λ2 )
|µ− , φ .
Therefore, using orthonormality of the eigenstates, we have ψ|eiH t T † P † Ft P T |ψ = |ψλ |2 λ|eiH t T † P † Ft P T |λ λ
=
(41)
√ −i arccos λ 2 |ψλ |2 ei(λ−sin φ)t |aφ|arcsin λ |2 1−λe 2 2(1−λ )
λ,φ
√ i arccos λ 2 +|aφ|π −arcsin λ |2 1−λe 2 2(1−λ )
=
(40)
|ψλ |2 ei(λ−sin φ)t
λ,φ
|aφ|arcsin λ |2 + |aφ|π −arcsin λ |2 . 2
(42) (43)
Thus the fidelity of the simulation is |ψ|eiH t T † P † Ft P T |ψ| ≥ min λ
≥
φ
min
|aφ|arcsin λ |2 + |aφ|π −arcsin λ |2 cos (λ − sin φ)t 2
θ∈[0,2π )
(44) cos (sin θ − sin φ)t |aφ|θ |2
(45)
φ
2 1 (sin θ − sin φ)t |aφ|θ |2 θ∈[0,2π ) 2
≥ 1 − min
(46)
φ
≥ 1−
t2 2
min
θ∈[0,2π )
(θ − φ)2 |aφ|θ |2 ,
(47)
φ
where in the last step we have used the fact that |sin θ − sin φ| ≤ |θ − φ|. Now, to show that the fidelity is close to 1, we quantify the performance of phase estimation. To optimize the bound, we must choose the phase estimation procedure
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
593
carefully. In standard phase estimation modulo M (using M − 1 calls to the unitary operation whose phase is being estimated), we begin with the uniform superposition M−1 √1 x=0 |x in the register used to estimate the phase. However, other input states M can give better estimates depending on the application. In particular, the initial state
M−1 2 π(x + 1) |x sin M +1 M +1
(48)
x=0
minimizes the variance of the estimate [11,20,34], so it gives the best possible bound in (47). This initial state leads to the probability distribution |aθ+ |θ |2 =
2 π cos2 ( M+1 2 ) sin ( M+1 )
2M(M + 1) sin2 ( 2 +
π 2
2(M+1) ) sin ( 2
−
π 2(M+1) )
,
(49)
where the estimated phase is θ + = 2π j/M for some integer j. Of course, only the value of j mod M is significant, and we can choose the range of angles so that
= 0 + 2π j/M, where 0 ≤ 0 < 2π/M and where the integer j satisfies −M/2 + 1 ≤ j ≤ M/2. Then for sufficiently large M, |aθ+ |θ |2 ≤ ≤ ≤
π2 2M 4 sin2 ( 2
+
π 2
2(M+1) ) sin ( 2
−
π 2(M+1) )
128π 2 M 4 4 (1 −
π2
2 (M+1)2
)2
512π 2 , 9M 4 4
(50) (51)
(52)
where the last step assumes that ≥ 2π/M. Since 2π j/M ≤ ≤ 2π( j + 1)/M, φ
∞
128( j + 1)2 4π 2 + 2 M2 9M 2 j 4 j=1 256 15π 2 + π 4 + 180ζ (3) 4π 2 + = M 2 9M 2 90 186 ≤ 2. M
(θ − φ)2 |aφ|θ |2 ≤
(53)
(54) (55)
2 2 Using this bound in (47), √ we obtain a simulation fidelity of at least 1 − 93t /M . Thus we find that M = O(t/ δ) steps suffice to obtain fidelity at least 1 − δ.
Of course, to carry out the simulation described in Theorem 5, we must be able to implement the discrete-time quantum walk. As discussed at the end of Sect. 3, this may be difficult in general due to the dependence of the states (4) on the principal eigenvector of abs(H ). However, it is straightforward to implement the walk for many cases of interest, such as for an unweighted regular graph. Indeed, the discrete-time quantum walk can be carried out efficiently provided only that the Hamiltonian is efficiently indexcomputable, and that the weights appearing in the states (4) are integrable in the sense
594
A. M. Childs
of [29]. Alternatively, to avoid introducing a dependence on a principal eigenvector, one can use (27) at the expense of replacing abs(H ) by H 1 . For Hamiltonians whose graphs are trees, the simulations of Theorems 4 and 5 are incomparable. On the one hand, the simulation of Theorem 4 runs in slightly superlinear time. On the other hand, the simulation of Theorem 5 scales with abs(H ), whereas 2 Theorem 4 only scales with max j k |k|H | j| , which may be smaller.
6. Element Distinctness This section gives an application of Theorem 5 to quantum query complexity: a continuous-time quantum walk algorithm for the element distinctness problem. The intention is to formulate and analyze the algorithm entirely in terms of its Hamiltonian, but to ultimately quantify its complexity in terms of conventional quantum queries. The relationship of this algorithm to Hamiltonian-based models of query complexity is discussed at the end of Sect. 7. In the element distinctness problem, we are given a black-box function f : {1, . . . , N } → S (for some finite set S) and are asked to determine whether there are two indices x, y ∈ {1, . . . , N } such that f (x) = f (y). Ambainis found a discrete-time quantum walk algorithm that solves this problem with O(N 2/3 ) queries [5], which is optimal [1]. Unlike quantum algorithms for search on low-degree graphs [8,16,17,52], no continuous-time analog of the element distinctness algorithm has been known: the walk takes place on a high-degree Johnson graph, and sparse Hamiltonian techniques are insufficient to implement it. Since Ambainis’s algorithm appeared, it has been an open question to find a continuous-time version (see for example [14, Sect. 5]). We now describe a continuous-time quantum walk that, when simulated using Theorem 5, gives an O(N 2/3 )-query quantum algorithm for element distinctness. As in [5], the algorithm uses a walk on the Johnson graph J (N , M). The vertices of N subsets of {1, . . . , N } of size M; edges connect subsets that differ this graph are the M in exactly one element. Let M := N 2/3 , the nearest integer to N 2/3 . To simplify the analysis, suppose there is a unique pair of indices x, y for which f (x) = f (y). By a classical reduction, this assumption is without loss of generality [5]. Let A j denote the set of M-element subsets of {1, . . . , N } that contain j elements from {x, y}. We use that a set S written in a ket denotes the uniform superposi the convention √ tion |S := s∈S |s/ |S| over the elements of S. Then in the basis {|A0 , |A1 , |A2 }, the adjacency matrix of J (N , M) is ⎛ −2M √ ⎝ 2M(N − M − 1) 0
√
2M(N − M − 1) 2−N √ 2(N − M)(M − 1)
√
⎞ 0 2(N − M)(M − 1)⎠ + M(N − M). 2(M − N )
(56) We modify the graph to depend upon the black box as follows. For every subset in A2 , add an extra vertex connected by an edge to the original subset. Denote the set of all such extra vertices by B2 . Then the Hamiltonian for the algorithm is H = HU + HC , where, in the basis {|A0 , |A1 , |A2 , |B2 },
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
HU =
1 N 2/3
⎛ √ 2M(N − M − 1) √ 0 −2M √ ⎜ 2M(N − M − 1) 2 − N 2(N − M)(M − 1) ⎜ √ ⎝ 2(N − M)(M − 1) 2(M − N ) 0 0 0 0
⎛ 0 ⎜0 HC = ⎝ 0 0
0 0 0 0
0 0 0 1
⎞ 0 0⎟ . 1⎠ 0
595
⎞ 0 0⎟ ⎟, 0⎠ 0 (57)
(58)
For the initial state, take a uniform superposition over the vertices of the original Johnson graph, namely |ψ(0) := |A0 ∪ A1 ∪ A2 N −2 N −2 N −2 N = |A |A |A + 2 + 0 1 2 / M M−1 M−2 M = 1 − O(N −1/3 ) |A0 + O(N −1/6 )|A1 + O(N −1/3 )|A2 ,
(59) (60) (61)
where we have used the choice M = N 2/3 in the last line. Asymptotically, the starting state is essentially |A0 . To analyze the algorithm, we compute the spectrum of H . In principle, this could be done using techniques from [16,17]. However, since the evolution takes place within a four-dimensional subspace, we can compute the relevant eigenvalues and eigenvectors √ −1± 17 1/3 in closed form. In particular, we find two eigenvalues λ± = N + O(1) with 4 eigenvectors √ |λ+ = µ + O( N 11/3 ), O( N 11/6 ), O( N 11/2 ), 1 − µ + O( N 11/6 ) ≈ (0.6154, 0, 0, 0.7882), √ 1 − µ + O( N 11/3 ), O( N 11/6 ), O( N 11/2 ), − µ + O( N 11/6 ) |λ− = ≈ (0.7882, 0, 0, −0.6154),
(62)
(63)
where µ :=
8 √
17 +
17
≈ 0.3787.
(64)
Asymptotically, the algorithm is effectively confined to the two-dimensional subspace spanned by |A0 and |B2 . The state rotates within this subspace at a rate determined by the inverse of the gap λ+ − λ− = O(N −1/3 ). Therefore, the initial state (60) reaches a state with overlap O(1) on |B2 in time O(N 1/3 ). To quantify the query complexity of this approach to element distinctness, we invoke Theorem 5. As in [5], suppose we store the M function values along with the subset at each vertex. Then preparing (60) takes M = N 2/3 queries. Furthermore, a step of the corresponding discrete-time quantum walk can be simulated using two queries: the walk operator is local on the Johnson graph, so we simply uncompute one function value and
596
A. M. Childs
compute another. Given the function values for each subset, the extra edges for marked vertices can be included without any additional queries; indeed, by an observation in the proof of Theorem 8 of [7], they can be implemented by performing the walk on the graph with extra edges for every vertex of the Johnson graph, together with a phase factor for the edges corresponding to marked vertices. Thus, the total number of queries used by the algorithm is O(N 2/3 + abs(H )N 1/3 ). Since abs(H ) = 2N 1/3 + O(N −1/3 ), we find an algorithm with running time (and, in particular, query complexity) O(N 2/3 ). 7. The Continuous-Time Query Model Continuous-time quantum walk has proven useful in motivating new algorithmic ideas. Ultimately, though, it is most straightforward to quantify the complexity of the resulting algorithms by the number of elementary gates needed to simulate them in the quantum circuit model, as in the example of Sect. 6. However, one can also formulate a notion of query complexity directly in a Hamiltonian-based model of computation. Such a model was first introduced by Farhi and Gutmann to describe a continuous-time analog [26] of Grover’s search algorithm [30]; it was later studied in a broader context by Mochon [40] and applied to the quantum walk algorithm for evaluating balanced binary game trees [25]. In a general formulation of the continuous-time query model, an algorithm is described by a Hamiltonian of the form H D (t) + H Q , where H Q is a fixed, timeindependent Hamiltonian encoding a black-box input, and H D (t) is an arbitrary oracle-independent “driving Hamiltonian” (possibly time-dependent and with no a priori upper bound on its norm). The complexity is quantified simply by the total evolution time required to produce the result with bounded error. Equivalently, we can consider a model of fractional queries interspersed by non-query unitary operations, and take the limit in which the fractional queries can be arbitrarily close to the identity, charging only 1/k of a full query to perform the k th root of a query. It is clear that the continuous-time query model is at least as powerful as the conventional query model. Very recently, it was shown that the continuous-time query model is in fact not significantly more powerful. In particular, any algorithm using continuous queries for time t can be simulated with O(t log t/ log log t) discrete queries [19]. Here we consider the case where the driving Hamiltonian H D (t) is restricted to be time-independent. We suppose that the query Hamiltonian H Q has the form used in [25]: namely, for a binary black-box input x ∈ {0, 1} N , H Q acts on C N ⊗ C2 as H Q |i, b = |i, b ⊕ xi
(65)
for any i ∈ {1, . . . , N } and b ∈ {0, 1}. In other words, the query Hamiltonian describes a graph on 2N vertices, with an edge between vertices (i, 0) and (i, 1) if xi = 1, and no such edge if xi = 0. Consider the case where H D is time-independent. By simulating H D + H Q with highorder approximations of the Lie product formula [9,12], the evolution for time t can be approximated using (H D t)1+o(1) queries to a unitary black box for the input x. Thus, a continuous-time quantum walk algorithm with H D = O(1) gives rise to a conventional quantum query algorithm using only t 1+o(1) queries. However, this simulation incurs more overhead than that of [19]. Instead, applying Theorem 5, and again using the observation from [7, Theorem 8] to implement the discrete-time quantum walk using discrete queries (as in the simulation of the element distinctness algorithm in Sect. 6), we find
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
597
Theorem 6. Consider a continuous-time query algorithm with the time-independent Hamiltonian H D + H Q that runs in time t. Then the query complexity in the conventional discrete-time model is O(abs(H D )t). In the case where H D is time-independent and satisfies abs(H D ) = O(1), this simulation outperforms [19]. However, it may be less efficient if abs(H D ) 1, and is not even applicable to a general time-dependent H D (t). Note, however, that almost all known continuous-time quantum algorithms have H D time-independent (with the notable exception of adiabatic algorithms for search [22,44], which can nevertheless be simulated efficiently in the quantum circuit model [45]) and satisfy abs(H D ) = O(1). (It might be interesting to investigate the application of discrete-time quantum walk to simulating the dynamics of a time-dependent Hamiltonian.) Notice that the element distinctness algorithm from Sect. 6 does not naturally fit into the continuous-time query model. The simulation described in Sect. 6 uses O(N 2/3 ) initial discrete queries, followed by O(N 2/3 ) queries to simulate an evolution for time only O(N 1/3 ). However, the techniques of [19] could not be used to simulate the latter evolution using only O(N 1/3 log N ) discrete queries: the initial queries and the queries used to simulate the evolution should balance to avoid violating the (N 2/3 ) lower bound for element distinctness [1]. Despite the superficial similarity between HC and H Q , it is the term HU , rather than HC , that depends on the black-box input; overall, HU + HC does not have the form of an oracle-independent Hamiltonian plus a query Hamiltonian of the form (65).
8. A Sign Problem for Hamiltonian Simulation Although the Hamiltonian simulations described in Sects. 3–5 go considerably beyond previous techniques, they stop short of what might be possible. We conclude by considering two problems for which improved simulation methods would be valuable: approximating exponential sums and implementing quantum transforms over association schemes. Hopefully, these potential applications will motivate further work on the simulation of Hamiltonian dynamics. The essential problem with simulations based on the correspondence to discrete-time quantum walk has to do with the appearance of abs(H ), rather than H , in Theorem 5. The natural scaling parameter for Hamiltonian simulation would seem to be the basis-independent quantity H t rather than the basis-dependent quantity abs(H )t; it seems reasonable to attempt a simulation in time poly(H t). However, we do not currently know how to do this except in special cases. There are at least two potential approaches to circumventing this limitation. It might be possible to use decomposition techniques, such as the decomposition of trees into stars described in Sect. 4, in a more general context. We also might try to perform simulations in alternative bases that can be reached by efficient unitary transformations. 8.1. Approximating Kloosterman sums. An exponential sum over Fq , the finite field with q elements, is an expression of the form x∈Fq
χ ( f (x)) ψ(g(x)),
(66)
598
A. M. Childs
where χ and ψ are multiplicative and additive characters of Fq , respectively, and f, g ∈ Fq [x] are polynomials. It is well known that, under fairly mild conditions, √ χ ( f (x)) ψ(g(x)) ≤ α q,
(67)
x∈Fq
where the coefficient α depends on the degrees of f and g (see for example [46, Theorem 2.6]). However, computing the √ value of an exponential sum—or even approximating its magnitude as a fraction of q with, say, constant precision—appears to be a difficult problem in general. When f and g are both linear √ in x, the sum (66) is known as a Gauss sum. Gauss sums have magnitude precisely q provided χ , ψ are nontrivial. No efficient classical algorithm for computing the phase of a general Gauss sum is known, but this phase can be efficiently approximated using a quantum computer [23]. It is an open question whether quantum computers can efficiently approximate other exponential sums. (Note that for the special case of small characteristic, it is possible to calculate more general exponential sums using the results of [31], as observed by Shparlinski; see [18].) Another type of exponential sum of particular interest is the Kloosterman sum, which is obtained by letting χ be the quadratic character, f (x) = x 2 − c for some fixed c ∈ Fq , and g(x) = x. It is easy to see that such a sum is real-valued. A classic result of Weil says that, provided √ ψ is nontrivial and c = 0, the absolute value of the Kloosterman sum is at most 2 q [54]. However, I am not aware of an efficient algorithm to compute even a single nontrivial bit of information √ about the Kloosterman sum, such as its sign or whether its magnitude is larger than q. Kloosterman sums arise naturally in problems involving hidden nonlinear structures over finite fields [18]. In particular, the eigenvalues of a graph known as the Winnie Li graph (in even dimensions) are proportional to Kloosterman sums. An efficient quantum algorithm for approximating Kloosterman sums would provide new efficient quantum algorithms for certain hidden nonlinear structure problems. For example, this would give a way to implement a quantum walk that could be used to solve the so-called hidden flat of centers problem. Conversely, an implementation of that quantum walk could be used to estimate the sums, using phase estimation. Let us consider a simple variant of the Winnie Li graph that exemplifies the relevant problem. Let G be the Cayley graph of the additive group of Fq with the generating set X := {x ∈ Fq : χ (x 2 − c) = +1} for some c ∈ Fq× , where χ denotes the quadratic character of Fq . Let p be the characteristic of Fq , and let tr : Fq → F p be the trace map, 2 defined by tr(x) = x + x p + x p + · · · + x q/ p . Observe that δ[x ∈ X ] =
1 1 + χ (x 2 − c) − δ[x 2 = c] , 2
(68)
where δ[P] is 1 if P is true and 0 if P is false. Then for each k ∈ Fq , the Fourier vector ˜ := √1 |k ωtr(kx) |x p q x∈Fq
(69)
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
599
(where ω p := e2π i/ p ) is an eigenvector of G with eigenvalue x∈X
1 1 + χ (x 2 − c) − δ[x 2 = c] ωtr(kx) p 2 x∈Fq √ 1 2π k c = q δ[k = 0] + − cos χ (x 2 − c) ωtr(kx) p 2 p
ωtr(kx) = p
(70)
(71)
x∈Fq
√ (where the term involving c does not appear if c is not a square in Fq ). In particular, the eigenvalues of G are simply related to Kloosterman sums. Notice that k = 0 gives an eigenvalue |X | = (q + 1)/2 − δ[χ (c) = +1] = (q ± 1)/2 (the degree of G) corresponding to the uniform vector. Since we are only interested in the nontrivial Kloosterman sums, we can subtract off the projection of the adjacency matrix onto the uniform vector, defining a symmetric matrix H withx|H √|x = δ[x − x ∈ X ] − |X |/q forx, x ∈ Fq . This Hamiltonian satisfies H ≤ 2 q (by Weil’s Theorem), and a simulation of its dynamics for time t in poly(H t, log q) steps would give an efficient quantum algorithm for approximating Kloosterman sums with constant precision. √ Unfortunately, Theorem 5 is insufficient for this task. Whereas H ∼ 2 q, a simple calculation shows that abs(H ) = (q 2 − 1)/2q = (q), so abs(H )/H is exponentially large (in log q). Similar considerations hold for the actual Winnie Li graph. 8.2. Quantum transforms for association schemes. An association scheme is a combinatorial object with useful algebraic properties (see [28] for an accessible introduction). We say that a set of matrices A0 , A1 , . . . , Ad ∈ {0, 1} N ×N is a d-class association scheme on d Ai = J , the N × N N vertices provided (i) A0 = I , the N × N identity matrix; (ii) i=0 matrix with every entry equal to 1; (iii) AiT ∈ {A0 , . . . , Ad } for each i ∈ {0, . . . , d}; and (iv) Ai A j = A j Ai ∈ span{A0 , . . . , Ad } for each i, j ∈ {0, . . . , d}. An association scheme partitions the relationships between pairs of vertices into classes: we say that the relationship between vertex x and vertex y is of type i if (Ai )x,y = 1 (which implies that (A j )x,y = 0 for all j = i). Let Ni denote the number of vertices of type i relative to any given vertex (or equivalently, the number of 1s in any given row of Ai ). The matrices A0 , A1 , . . . , Ad generate a (d +1)-dimensional algebra over C called the Bose-Mesner algebra of the scheme. This algebra is also generated by a set of d+1 projection matrices E 0 , E 1 , . . . , E d called its idempotents. These projections are orthonormal d E i = I ); furthermore, (E i E j = δi, j E i for all i, j ∈ {0, . . . , d}) and complete ( i=0 the range of E i is an eigenspace of A j for every i, j ∈ {0, . . . , d}. Since they lie in the Bose-Mesner algebra, the idempotents can be expanded as E i = N1 dj=0 qi j A j , where the qi j are referred to as the dual eigenvalues of the scheme. The unitary matrices in the Bose-Mesner algebra are precisely those matrices of the form U=
d
eiφi E i
(72)
i=0
for some phases φi ∈ R. Since they are highly structured, these unitary operators can be concisely specified even when we think of N as exponentially large, and thus they
600
A. M. Childs
represent a natural class of quantum transforms for possible use in quantum algorithms. Closely related to such a transform is the POVM with elements {E 0 , E 1 , . . . , E d }: the ability to perform this measurement (coherently) can be used to implement any desired unitary transform in the Bose-Mesner algebra, and conversely, implementation of any unitary transform in the Bose-Mesner algebra with well-separated eigenvalues can be used to measure {E 0 , E 1 , . . . , E d }, by phase estimation. For a concrete application of association scheme transforms, consider the following problem. Fix a symmetric N -vertex association scheme and an (unknown) vertex t of that scheme. Suppose we are given a black box function f : {1, . . . , N } → S with the promise that for all x, y ∈ {1, . . . , N }, f (x) = f (y) if and only if x and y are of the same class with respect to t (or in other words, if and only if (Ai )t,x = (A j )t,y = 1 ⇒ i = j). The task is to learn t using as few queries to f as possible, ideally only poly(log N ). The shifted quadratic character problem is a particular instance of this problem in which the association scheme is the Paley scheme. It has d = 2 classes, meaning that it corresponds to a strongly regular graph, the Paley graph. In the Paley scheme, the vertices are elements of Fq (where q = 1 mod 4), and (A1 )x,y = 1 if and only if x − y is a square in Fq . The sequence of values of the quadratic character χ (t), χ (t + 1), χ (t + 2), . . . has been proposed as a pseudorandom generator [24], and indeed no efficient classical algorithm is known that will learn t from the function f (x) = χ (t + x) that hides t in the Paley scheme. On the other hand, van Dam, Hallgren, and Ip discovered a quantum algorithm that uses the quantum Fourier transform to solve this problem in time poly(log q) [21]. Association scheme transforms provide a general approach to this problem. Suppose we prepare a uniform superposition over all vertices of the scheme, compute the hiding function f , and then discard its value. The result is a uniform superposition over points that are of type i relative to t, 1 |ψi := √ |x, (73) Ni x:(A ) =1 i t,x
where type i occurs with probability Ni /N . Now suppose we apply the unitary operator (72), giving the state U |ψi , and measure the vertex. The probability that we obtain the vertex t is d d 2 1 iφ j Ni iφ j 2 2 e q jk ψ0 |Ak |ψi = 2 e q ji . (74) |ψ0 |U |ψi | = 2 N N j,k=0
j=0
Supposing that we know the value of i, we can choose the phases φ j to obtain a success probability Ni ( j |q ji |)2 /N 2 . Since type i occurs with probability Ni /N , by always choosing the phases to optimize the probability of obtaining vertex t given that type i occurred (and pessimistically assuming that we never obtain vertex t when another type occurs), we can succeed with probability at least Ni2 ( j |q ji |)2 /N 3 . Thus, by always choosing the phases according to the value of i that maximizes this expression, we can achieve an overall success probability of at least d 2 N2 max i3 |q ji | . (75) i N j=0
For example, in the Paley scheme, a√ straightforward calculation shows that√ this approach succeeds with probability at least ( q + 1)2 (q − 1)2 /4q 3 = 1/4 + O(1/ q).
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
601
The unitary operators (72) for the Paley scheme can be implemented efficiently because the corresponding idempotents are projections onto Fourier basis vectors. In ˜ : χ (k) = +1}, and E 2 projects onto ˜ 0|, ˜ E 1 projects onto span{|k particular, E 0 = |0 ˜ : χ (k) = −1} (recall (69)). Thus, using the Fourier transform and the ability span{|k to compute quadratic characters (which can be done efficiently using Shor’s algorithm for discrete logarithms [49]), we can efficiently carry out the above strategy to solve the shifted quadratic character problem. However, this approach is not very different from the one in [21], which employs similar tools. Instead, it would be appealing to have a purely combinatorial means of implementing (72), which could be applied even for association schemes without an underlying group structure. One way to implement (72) would be to coherently measure each of the idempotents in turn, performing a phase shift conditional on the measurement outcome and then undoing the measurement. In principle, these measurements could be performed by applying phase estimation to a simulation of Hamiltonian dynamics with the Hamiltonian given by the idempotent. Unfortunately, in many cases of interest, this approach encounters the same sign problem seen in the whereas E 1 = 1, √ previous section. For example, √ one can show that abs(E 1 ) = ( q + 1)(q − 1)/2q = ( q). Since abs(E 1 )/E 1 is exponentially large in log q, we again find that Theorem 5 does not give an efficient implementation. Acknowledgements. I would like to thank Ben Reichardt and Rolando Somma for pointing out the choice (27); Ben Reichardt, Robert Špalek, and Shengyu Zhang for discussions, in the course of writing [7], that led to the simulation described in Theorem 4; Chris Godsil and Simone Severini for discussions of association schemes; and Dominic Berry and Aram Harrow for discussions of phase estimation. This work was supported by MITACS, NSERC, QuantumWorks, and the US ARO/DTO.
References 1. Aaronson, S., Shi, Y.: Quantum lower bounds for the collision and the element distinctness problems. J. ACM 51 (4) 595–605 (2004), preliminary versions in STOC 2002 and FOCS 2002 2. Aharonov, D., Ambainis, A., Kempe, J., Vazirani, U.: Quantum walks on graphs. In: Proc. 33rd ACM Symposium on Theory of Computing, pp. 50–59, 2001, available at http://arxiv.org/abs/quant-ph/0012090, 2000 3. Aharonov, D., Ta-Shma, A.: Adiabatic quantum state generation and statistical zero knowledge. In: Proc. 35th ACM Symposium on Theory of Computing, pp. 20–29, 2003, available at http://arxiv.org/abs/quantph/0301023, 2003 4. Aldous, D., Fill, J.A.: Reversible Markov chains and random walks on graphs (in preparation), http:// www.stat.berkeley.edu/~aldous/RWG/book.html 5. Ambainis, A.: Quantum walk algorithm for element distinctness. SIAM J. Comput. 37(1), 210–239 (2007) 6. Ambainis, A., Bach, E., Nayak, A., Vishwanath, A., Watrous, J.: One-dimensional quantum walks. Proc. 33rd ACM Symposium on Theory of Computing, pp. 37–49, 2001, available at http://arxiv.org/abs/quantph/0010117, 2000 7. Ambainis, A., Childs, A.M., Reichardt, B.W., Špalek, R., Zhang S.: Any AND-OR formula of size N can be evaluated in time N 1/2+o(1) on a quantum computer. In: Proc. 48th IEEE Symposium on Foundations of Computer Science, pp. 363–372, 2007, available at http://arxiv.org/abs/quant-ph/0703015 and http:// arxiv.org/abs/0704.3628, 2007 8. Ambainis, A., Kempe, J., Rivosh, A.: Coins make quantum walks faster. In: Proc. 16th ACM-SIAM Symposium on Discrete Algorithms, pp. 1099–1108, 2005, available at http://arxiv.org/abs/quant-ph/0402107, 2004 9. Berry, D.W., Ahokas, G., Cleve, R., Sanders, B.C.: Efficient quantum algorithms for simulating sparse Hamiltonians. Commun. Math. Phys. 270(2), 359–371 (2007) 10. Buhrman, H., Špalek, R.: Quantum verification of matrix products. In: Proc. 17th ACM-SIAM Symposium on Discrete Algorithms, pp. 880–889, 2006, available at http://arxiv.org/abs/quant-ph/0409035, 2004
602
A. M. Childs
11. Bužek, V., Derka, R., Massar, S.: Optimal quantum clocks. Phys. Rev. Lett. 82(10), 2207–2210 (1999) 12. Childs, A.M.: Quantum information processing in continuous time. Ph.D. thesis, Massachusetts Institute of Technology, Cambridge, MA, 2004 13. Childs, A.M., Cleve, R., Deotto, E., Farhi, E., Gutmann, S., Spielman, D.A.: Exponential algorithmic speedup by quantum walk. In: Proc. 35th ACM Symposium on Theory of Computing, pp. 59–68, 2003, available at http://arxiv.org/abs/quant-ph/0209131, 2002 14. Childs, A.M., Eisenberg, J.M.: Quantum algorithms for subset finding. Quant. Inf. Comp. 5(7), 593– 604 (2005) 15. Childs, A.M., Farhi, E., Gutmann, S.: An example of the difference between quantum and classical random walks. Quant. Inf. Proc. 1(1-2), 35–43 (2002) 16. Childs, A.M., Goldstone, J.: Spatial search by quantum walk. Phys. Rev. A 70(2), 022314 (2004) 17. Childs, A.M., Goldstone, J.: Spatial search and the Dirac equation. Phys. Rev. A 70(4), 042312 (2004) 18. Childs, A.M., Schulman, L.J., Vazirani, U.V.: Quantum algorithms for hidden nonlinear structures. In: Proc. 48th IEEE Symposium on Foundations of Computer Science, pp. 395–404, 2007, available at http:// arxiv.org/abs/0705.2784, 2007 19. Cleve, R., Gottesman, D., Mosca, M., Somma, R.D., Yonge-Mallo, D.L.: Efficient discrete-time simulations of continuous-time quantum query algorithms. In: Proc. 41st ACM Symposium on Theory of Computing, pp. 409–416, 2009, available at http://arxiv.org/abs/0811.4428, 2008 20. van Dam, W., D’Ariano, G.M., Ekert, A., Macchiavello, C., Mosca, M.: Optimal phase estimation in quantum networks. J. Phys. A 40(28), 7971–7984 (2007) 21. van Dam, W., Hallgren, S., Ip, L.: Quantum algorithms for some hidden shift problems. In: Proc. 14th ACM-SIAM Symposium on Discrete Algorithms, pp. 489–498, 2002, available at http://arxiv.org/abs/ quant-ph/0211140, 2002 22. van Dam, W., Mosca, M., Vazirani, U.: How powerful is adiabatic quantum computation?. In: Proc. 42nd IEEE Symposium on Foundations of Computer Science, pp. 279–287, 2001, available at http://arxiv.org/ abs/quant-ph/0206003, 2002 23. van Dam, W., Seroussi, G.: Quantum algorithms for estimating Gauss sums and calculating discrete logarithms. Manuscript, 2003, available at http://www.cs.ucsb.edu/~vandam/gausssumdlog.pdf 24. Damgård, I.B.: On the randomness of Legendre and Jacobi sequences. Advances in Cryptology - CRYPTO ’88, Lecture Notes in Computer Science, vol. 403, New York: Springer, 1990, pp. 163–172 25. Farhi, E., Goldstone, J., Gutmann, S.: A quantum algorithm for the Hamiltonian NAND tree. Theory of Computing 4(1), 169–190 (2008) 26. Farhi, E., Gutmann, S.: Analog analogue of a digital quantum computation. Phys. Rev. A 57(4), 2403– 2406 (1998) 27. Farhi, E., Gutmann, S.: Quantum computation and decision trees. Phys. Rev. A 58(2), 915–928 (1998) 28. Godsil, C.: Association schemes. Lecture notes, available at http://quoll.uwaterloo.ca/pstuff/assoc.pdf, 2004 29. Grover, L., Rudolph, T.: Creating superpositions that correspond to efficiently integrable probability distributions, available at http://arxiv.org/abs/quant-ph/0208112v1, 2002 30. Grover, L.K.: Quantum mechanics helps in searching for a needle in a haystack. Phys. Rev. Lett. 79(2), 325–328 (1997), preliminary version in STOC 1996 31. Kedlaya, K.S.: Quantum computation of zeta functions of curves. Comput. Complex. 15(1), 1–19 (2006) 32. Linial, N.: Locality in distributed graph algorithms. SIAM J. Comput. 21(1), 193–201 (1992) 33. Lloyd, S.: Universal quantum simulators. Science 273(5278), 1073–1078 (1996) 34. Luis, A., Peˇrina, J.: Optimum phase-shift estimation and the quantum description of the phase difference. Phys. Rev. A 54(5), 4564–4570 (1996) 35. Magniez, F., Nayak, A.: Quantum complexity of testing group commutativity. Algorithmica 48(3), 221–232 (2007), preliminary version in ICALP 2005 36. Magniez, F., Nayak, A., Roland, J., Santha, M.: Search via quantum walk. In: Proc. 39th ACM Symposium on Theory of Computing, pp. 575–584, 2007, available at http://arvix.org/abs/quant-ph/0608026, 2006 37. Magniez, F., Santha, M., Szegedy, M.: Quantum algorithms for the triangle problem. In: Proc. 16th ACM-SIAM Symposium on Discrete Algorithms, pp. 1109–1117, 2005, available at http://arxiv.org/abs/ quant-ph/0310134, 2003 38. Meyer, D.A.: From quantum cellular automata to quantum lattice gasses. J. Stat. Phys. 85(5/6), 551– 574 (1996) 39. Meyer, D.A.: On the absence of homogeneous scalar unitary cellular automata. Phys. Lett. A 223, 337– 340 (1996) 40. Mochon, C.: Hamiltonian oracles. Phys. Rev. A 75(4), 042313 (2007) 41. Moore, C., Russell, A.: Quantum walks on the hypercube. In: Proc. 6th International Workshop on Randomization and Approximation Techniques in Computer Science, Lecture Notes in Computer Science, Vol. 2483, Berlin: Springer, 2002 pp. 164–178
On the Relationship Between Continuous- and Discrete-Time Quantum Walk
603
42. Regev, O.: Witness-preserving amplification of QMA. Lecture notes, http://www.cs.tau.ac.il/~odedr/ teaching/quantum_fall_2005/ln/qma.pdf, 2006 43. Reichardt, B.W., Špalek, R.: Span-program-based quantum algorithm for evaluating formulas. In: Proc. 40th ACM Symposium on Theory of Computing, pp. 103–112, 2008, available at http://arxiv.org/abs/ 0710.2630, 2007 44. Roland, J., Cerf, N.J.: Quantum search by local adiabatic evolution. Phys. Rev. A 65(4), 042308 (2002) 45. Roland, J., Cerf, N.J.: Quantum-circuit model of Hamiltonian search algorithms. Phys. Rev. A 68(6), 062311 (2003) 46. Schmidt, W.M.: Equations over Finite Fields: An Elementary Approach. 2nd ed., Hebercity, UT: Kendrick Press, 2004 47. Severini, S.: On the digraph of a unitary matrix. SIAM J. Matrix Anal. Appl. 25(1), 295–300 (2003) 48. Shenvi, N., Kempe, J., Whaley, K.B.: A quantum random walk search algorithm. Phys. Rev. A 67(5), 052307 (2003) 49. Shor, P.W.: Polynomial-time algorithms for prime factorization and discrete logarithms on a quantum computer. SIAM J. Comput. 26(5), 1484–1509 (1997) 50. Strauch, F.W.: Connecting the discrete- and continuous-time quantum walks. Phys. Rev. A 74(3), 030301 (2006) 51. Szegedy, M.: Quantum speed-up of Markov chain based algorithms. In: Proc. 45th IEEE Symposium on Foundations of Computer Science, pp. 32–41, 2004, available at http://arxiv.org/abs/quant-ph/0401053, 2004 52. Tulsi, A.: Faster quantum-walk algorithm for the two-dimensional spatial search. Phys. Rev. A 78(1), 012310 (2008) 53. Watrous, J.: Quantum simulations of classical random walks and undirected graph connectivity. J. Comput. System Sci. 62(2), 376–391 (2001) 54. Weil, A.: On some exponential sums. Proc. Natl. Acad. Sci. 34(5), 204–207 (1948) Communicated by M. B. Ruskai
Commun. Math. Phys. 294, 605–645 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0974-2
Communications in
Mathematical Physics
Non-Birational Twisted Derived Equivalences in Abelian GLSMs Andrei C˘ald˘araru1 , Jacques Distler2 , Simeon Hellerman3 , Tony Pantev4 , Eric Sharpe5 1 Mathematics Department, University of Wisconsin, Madison,
WI 53706-1388, USA. E-mail:
[email protected]
2 Department of Physics, University of Texas, Austin, Austin,
TX 78712-0264, USA. E-mail:
[email protected]
3 Institute for Advanced Study, School of Natural Sciences, Princeton,
NJ 08540, USA. E-mail:
[email protected]
4 Department of Mathematics, University of Pennsylvania, Philadelphia,
PA 19104-6395, USA. E-mail:
[email protected]
5 Physics Department, Virginia Tech, Blacksburg, VA 24061, USA.
E-mail:
[email protected] Received: 26 November 2007 / Accepted: 13 November 2009 Published online: 29 December 2009 – © Springer-Verlag 2009
Abstract: In this paper we discuss some examples of abelian gauged linear sigma models realizing twisted derived equivalences between non-birational spaces, and realizing geometries in novel fashions. Examples of gauged linear sigma models with nonbirational Kähler phases are a relatively new phenomenon. Most of our examples involve gauged linear sigma models for complete intersections of quadric hypersurfaces, though we also discuss some more general cases and their interpretation. We also propose a more general understanding of the relationship between Kähler phases of gauged linear sigma models, namely that they are related by (and realize) Kuznetsov’s ‘homological projective duality.’ Along the way, we shall see how ‘noncommutative spaces’ (in Kontsevich’s sense) are realized physically in gauged linear sigma models, providing examples of new types of conformal field theories. Throughout, the physical realization of stacks plays a key role in interpreting physical structures appearing in GLSMs, and we find that stacks are implicitly much more common in GLSMs than previously realized. Contents 1. 2.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . Quadrics in Projective Space and Branched Double Covers 2.1 Review of the mathematics . . . . . . . . . . . . . . . 2.2 Basic GLSM analysis . . . . . . . . . . . . . . . . . . 2.3 Berry phase computation . . . . . . . . . . . . . . . . 2.4 Monodromy around the Landau-Ginzburg point . . . . 2.5 A puzzle with a geometric interpretation of the Landau-Ginzburg point . . . . . . . . . . . . . . . . . 2.6 Resolution of this puzzle – new CFT’s . . . . . . . . . 2.6.1 Homological projective duality . . . . . . . . . . . 2.6.2 Noncommutative algebras and matrix factorization. 2.7 Summary so far . . . . . . . . . . . . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
606 608 609 610 611 615
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
618 621 622 624 625
606
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
2.8 Generalizations in other dimensions . . . . . . Example Related to Vafa-Witten Discrete Torsion . 3.1 Basic analysis . . . . . . . . . . . . . . . . . . 3.2 Some notes on the geometry . . . . . . . . . . 3.3 Relation to P7 [2, 2, 2, 2] . . . . . . . . . . . . 3.4 Discrete torsion and deformation theory . . . . 4. Non-Calabi-Yau Examples . . . . . . . . . . . . . . 4.1 Hyperelliptic curves and P2g+1 [2, 2] . . . . . . 4.2 P7 [2, 2, 2] . . . . . . . . . . . . . . . . . . . . 4.3 P5 [2, 2] . . . . . . . . . . . . . . . . . . . . . 4.4 Degree 4 del Pezzo (P4 [2, 2]) . . . . . . . . . . 4.5 P6 [2, 2, 2] . . . . . . . . . . . . . . . . . . . . 4.6 P6 [2, 2, 2, 2] . . . . . . . . . . . . . . . . . . . 5. More General Complete Intersections . . . . . . . . 5.1 P4 [3] . . . . . . . . . . . . . . . . . . . . . . . 5.2 P5 [3, 3] . . . . . . . . . . . . . . . . . . . . . 5.2.1 Basic analysis. . . . . . . . . . . . . . . . 5.2.2 Monodromy computation. . . . . . . . . . 5.2.3 Homological projective duality and fibered noncommutative K3s . . . . . . . . . . . . 6. Conclusions . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . A. Calabi-Yau Categories and Noncommutative Spaces A.1 Definitions . . . . . . . . . . . . . . . . . . . . A.2 Deformations . . . . . . . . . . . . . . . . . . A.3 Cohomology of nc spaces . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . .
3.
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
625 626 626 628 629 629 630 630 632 632 633 635 636 636 636 637 638 638
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
640 641 642 642 642 643 644 644
1. Introduction Gauged linear sigma models, first described in [1], have proven to be a crucial tool for string compactifications. They have provided insight into topics ranging from the structure of SCFT moduli spaces to curve-counting in Calabi-Yau’s. When a GLSM describes different geometries in different limits of Kähler moduli space, it has long been assumed that the different geometries are birational to one another, e.g. related by flops, blowups, blowdowns, or other such transformations. It has also been assumed that the only Calabi-Yau’s one could describe as phases of GLSM’s were built as complete intersections in toric varieties or flag manifolds (or other semiclassical moduli spaces of supersymmetric gauge theories). However, recently we have begun to learn that neither statement is always the case. In [2, Sect. 12.2] and then in [3], examples have been given of gauged linear sigma models involving (a) a Calabi-Yau not presented as a complete intersection, and (b) two non-birationally-equivalent Calabi-Yau’s. In [3], a nonabelian GLSM was analyzed, describing a complete intersection in a Grassmannian and was shown to lie on the same moduli space as the vanishing locus of a Pfaffian, and in [2, Sect. 12.2] an abelian GLSM was analyzed, describing complete intersection of four degree two hypersurfaces in P7 at one limit and a branched double cover of P3 , branched over a degree eight hypersurface (Clemens’ octic double solid) in another Kähler phase.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
607
In this paper, we shall study further examples of abelian GLSM’s describing non-birational Kähler phases. We begin by working through the example of [2, Sect. 12.2] in much greater detail, then go on to consider other examples. One natural question this work poses is: is there a mathematical relationship between the different Kähler phases, some notion that replaces ‘birational’? We propose that the different Kähler geometric phases of a given GLSM should all be understood as being related by ‘homological projective duality,’ a recent concept introduced into mathematics by Kuznetsov. Put another way, we propose that GLSM’s implicitly give a physical realization of Kuznetsov’s homological projective duality. In addition, we argue that new kinds of conformal field theories are realized as these duals. These are physical realizations of Kuznetsov’s noncommutative resolutions of singular spaces. We introduce these new conformal field theories and discuss some of their basic properties, but clearly a great deal of work should be done to properly understand them and their role in physics. The analysis of the Landau-Ginzburg points of these GLSMs revolves around subtleties in the two-dimensional abelian gauge theories with nonminimal charges, which provide one physical realization of strings on gerbes. In other words, this paper describes in detail one application of gerbes and stacks. The original application of the technology of stacks, aside from the completely obvious possibility of enlarging the number of possible string compactifications, was to understand physical properties of string orbifolds such as the fact that they give well-behaved CFT’s [4]. More recent applications outlined in [2] range from making physical predictions for certain quantum cohomology computations to reconciling different physical aspects of the geometric Langlands program. We begin in Sect. 2 with a detailed analysis of the GLSM for P7 [2, 2, 2, 2]. We find, after an analysis that involves understanding how stacks appear physically, and also after finding a crucial Berry phase, that the Landau-Ginzburg point seems, on the face of it, to be in the same universality class as a nonlinear sigma model on a branched double cover of P3 , which is another Calabi-Yau. This is already interesting in that these two geometries, the complete intersection and the branched double cover, are not birational to one another, violating the conventional wisdom that different geometric Kähler phases of the same GLSM should be birational to one another. This is also noteworthy for the novel realization of the geometry at the Landau-Ginzburg point, as something other than the simultaneous vanishing locus of a set of F-terms, realizing a complete intersection in a toric variety. Further analysis reveals further subtleties: although for analogues in lower dimensions the branched double cover at the Landau-Ginzburg point is smooth, for the particular example P7 [2, 2, 2, 2] the branched double cover is mathematically singular, whereas the GLSM does not exhibit any singularities. An additional study leads us to believe that the structure actually being realized is a ‘noncommutative resolution’ of the singular branched double cover, a conjecture which is verified by studying matrix factorizations at the Landau-Ginzburg point. (Noncommutative resolutions are defined by their sheaf theory, so, seeing that matrix factorizations match the mathematics nails down the interpretation as a noncommutative resolution.) In particular, this means that we are getting some new conformal field theories – CFT’s that look like ordinary nonlinear sigma models on smooth patches, but which are fundamentally different over singular parts of the classical geometry. We tentatively identify this duality between the large-radius and Landau-Ginzburg point geometries as an example of Kuznetsov’s ‘homological projective duality.’ Finally, at the end of Sect. 2 we also outline how this generalizes in other dimensions.
608
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
In Sect. 3 we discuss another Calabi-Yau example of this phenomenon, in which a GLSM for a complete intersection of quadrics has a (noncommutative resolution of a) branched double cover at its Landau-Ginzburg point. This particular example amounts to a fibered version of a low-dimensional example of the form from Sect. 2, and is also closely related to geometries appearing in Vafa and Witten’s work on discrete torsion [5]. We also discuss how deformation theory issues, the last remaining property of discrete torsion that has not been completely explicitly derived from B fields, can be understood from the perspective of noncommutative spaces. In Sect. 4 we extend these considerations to a series of non-Calabi-Yau examples, in which again we see GLSM’s relating complete intersections of quadrics to (noncommutative resolutions of) branched double covers. In Sect. 5 we extend these notions to more general complete intersections, not of quadrics. We find that homological projective duality continues to apply to more general cases, even cases in which the Landau-Ginzburg point does not have a geometric or nc-geometric interpretation, and we conjecture that all phases of all gauged linear sigma models are related by homological projective duality. Finally in Appendix A we review some general aspects of noncommutative resolutions and nc spaces, to make this paper more nearly self-contained, as these notions have not, to our knowledge, been previously discussed in the physics literature. There are many technical similarities between the abelian GLSMs for complete intersections described in [2, Sect. 12.2] and the nonabelian GLSMs describing complete intersections in Grassmannians in [3, Sect. 5]. In both cases, the geometry at one limit of the GLSM Kähler moduli space is realized in a novel fashion: here and in [2, Sect. 12.2] as a double cover realized by gerbes and a nonminimally-charged gauge theory, in [3, Sect. 5], through strong-coupling nonabelian gauge dynamics. In both cases, the geometries at either end of the GLSM Kähler moduli space are not birational, but instead are related by Kuznetsov’s homological projective duality. In both cases the superpotential has the form W () = i Ai j j ij
for some matrix A, giving a mass to the chiral superfields i . The primary physical difference between the gauged linear sigma model in [3, Sect. 5] and [2, Sect. 12.2] is that in the former, at least one φ always remains massless (and is removed by quantum corrections), whereas in the latter all of the φ are generically massive. Thus, in the latter case one generically has a nonminimally charged field, p, and so gerbes are relevant, whereas in the former there is never a nonminimally-charged-field story. In [6], further nonabelian examples were presented, expanding on that discussed in [3, Sect. 5], and the relevance of homological projective duality, discussed in more detail in this paper, was introduced. The physics of complete intersections of quadrics plays a central role in this paper. More information on the mathematics of complete intersections of quadrics can be found in, for example, [7]. 2. Quadrics in Projective Space and Branched Double Covers Our first example involves a gauged linear sigma model describing a complete intersection of four quadrics in P7 in the r 0 limit, and a double cover of P3 branched over a degree 8 locus in the r 0 limit. This example originally appeared in [2, Sect. 12.2]; we shall review and elaborate upon that example here.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
609
2.1. Review of the mathematics. We shall begin by reviewing pertinent mathematics. First, let us remind the reader why a double cover of P3 branched over a degree 8 hypersurface in P3 is an example of a Calabi-Yau. Let B be a complex manifold and let D ⊂ B be a smooth divisor. A double cover π : S → B branched along D is specified uniquely by a holomorphic line bundle L → B, such that L ⊗2 ∼ = O B (D). Explicitly, if s ∈ H 0 (B, L ⊗2 ) is a section with divisor D, then the double cover S is the divisor in the total space of L given by the equation z 2 = p ∗ s, where p : Tot(L) → B is the natural projection, and z ∈ H 0 (Tot(L), p ∗ L) is the tautological section. For such a cover S the adjunction formula gives K S = π ∗ (K B ⊗ L). In particular, if L = K B−1 , then the double cover S will have a trivial canonical class. In the present case, the base is P3 , with canonical bundle of degree −4, and so we see that the branched double cover is Calabi-Yau if the branch locus has degree 8. For a closely related discussion in the context of a different example, see [8, Chap. 4.4, p. 548]. Double covers of P3 branched over a degree 8 hypersurface in P3 are known as octic double solids, and are described in greater detail in e.g. [9,10]. Mathematically, the double cover can be understood as a moduli space of certain bundles on the complete intersection of quadrics. (Each quadric in P7 carries two distinct spinor bundles which restrict to bundles on the complete intersection, and when the quadric degenerates, the spinor bundles become isomorphic, hence giving the double cover of P3 .) Now, the twisted1 derived category of coherent sheaves of the branched double cover of P3 has been expected [13], and was recently proven2 [14], to be isomorphic to the derived category of a complete intersection of four quadrics in P7 . Specifically, there is a twisted derived equivalence if the double cover and the complete intersection are related as follows. Let Q a denote the four quadrics in the complete intersection, and consider the following linear combination: pa Q a (x), a
where the pa are homogeneous coordinates on P3 . Rewrite this linear combination as xi Ai j ( p)x j , ij
where Ai j is an 8 × 8 matrix with entries linear in the pa . Then the complete intersection of the four quadrics Q a is a twisted derived equivalence to a branched double cover of P3 branched over the degree eight locus det A = 0. Such derived equivalences are not unusual in gauged linear sigma models. After all, derived categories encapsulate the open string B model [15–17], and the B model is independent of Kähler moduli, hence one expects that different geometries on the same GLSM Kähler moduli space will have isomorphic derived categories. 1 Twisted in the sense described in [11]: because of a flat B field present, transition functions only close up to cocycles on triple overlaps. See [12] for a discussion of the Brauer group of P7 [2, 2, 2, 2]. 2 What was proven in [14] was a relation between the twisted derived category of a noncommutative resolution of the branched double cover, and the derived category of P7 [2, 2, 2, 2]. That noncommutative resolution will play an important role in the physics, as we shall discuss later.
610
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
On the other hand, it is also typically the case that different phases of a GLSM will be related by birational transformations, and that is not the case here: as pointed out by M. Gross [13] the complete intersection in P7 has no contractible curves, whereas the branched double cover has several ordinary double points.
2.2. Basic GLSM analysis. In this section we will work through the analysis of a gauged linear sigma model describing the complete intersection of four degree-two hypersurfaces in P7 at large radius. We will find, after careful analysis involving an understanding of how gerbes appear in physics, that the Landau-Ginzburg point of this GLSM can be interpreted geometrically as a branched double cover of P3 , the same branched double cover related to the complete intersection by a twisted derived equivalence. This gauged linear sigma model has a total of twelve chiral superfields, eight (φi , i ∈ {1, . . . , 8}) of charge 1 corresponding to homogeneous coordinates on P7 , and four ( pa , a ∈ {1, . . . , 4}) of charge −2 corresponding to the four hypersurfaces. The D-term for this gauged linear sigma model reads |φi |2 − 2 | pa |2 = r. a
i
When r 0, then we see that not all the φi can vanish, corresponding to their interpretation as homogeneous coordinates on P7 . More generally, for r 0 we recover the geometric interpretation of this gauged linear sigma model as a complete intersection of quadrics. For r 0, we find a different story. There, the D-term constraint says that not all the pa ’s can vanish; in fact, the pa ’s act as homogeneous coordinates on a P3 , except that these homogeneous coordinates have charge 2 rather than charge 1. Because of those nonminimal charges, the Landau-Ginzburg point is ultimately going to describe a (branched) double cover. The superpotential W = pi Q i (φ) i
(where the Q i are quadric polynomials) can be equivalently rewritten in the form φi Ai j ( p)φ j , W = ij
where Ai j is a symmetric matrix with entries linear in the p’s. Away from the locus where A drops rank, i.e., away from the hypersurface det A = 0, the φi are all massive, leaving only the pi massless, which all have charge −2. A GLSM with nonminimal charges describes a gerbe [18–20], and physically a string on a gerbe is equivalent via T-duality to a string on a disjoint union of spaces [2] (see [21] for a short review). For later use, let denote the locus = {det A = 0}, where the mass matrix drops rank. So far we have found that the Landau-Ginzburg point physics corresponds to a sigma model on some sort of double cover of P3 , away from the hypersurface {det A = 0} ≡ .
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
611
The Z2 gerbe on the P3 away from is a banded3 gerbe and so [2] gives rise to a disjoint union of two copies of the underlying space, i.e. a trivial double cover. However, we have claimed that we will ultimately get a branched double cover of P3 , and the branched double cover of P3 is a nontrivial4 double cover of P3 away from the branch locus . The reason for this apparent mismatch is another bit of physics; to fully understand the Landau-Ginzburg point, we must take into account a Berry phase, that exchanges the two copies as one circumnavigates the branch locus, and makes the double cover nontrivial. 2.3. Berry phase computation. We shall construct a local model for the codimensionone degenerate locus so that we can investigate the fibration structure of the Z2 gerbe over the base. We focus on a smooth point of . First let us work on the affine patch p4 = v = 0. Choose affine coordinates z a ≡ pa /v, a = 1, 2, 3. The vev of p4 breaks the U (1) spontaneously down to a Z2 subgroup under which the z are even and the φ are odd. The moduli space of the theory at r 0 is parametrized by √ z. The fields φ are massive over a generic point in moduli space. Redefine φ as y/ v, so that v drops out of the superpotential. Choose local coordinates so that the defining equation of is z 3 = 0 + o(z a2 ). Then rescale the z a by an infinite amount z a → −2 z a , yi → yi , in order to get rid of the order z a2 terms in the defining equation for . This flattens out the degenerate locus to a hyperplane z 3 = 0 in z a space. Finally, choose a basis for the φi so that the matrix A1i ( p) = 0 and Ai j = mδi j for i, j ≥ 2. In the scaling limit where we recover the local model, the superpotential is 8 1 (y i )2 . W = m z 3 (y 1 )2 + 2 i=2
The yi for i > 1 are massive everywhere in the local model and decoupled from the z degrees of freedom, so we can integrate them out trivially. Likewise the z 1,2 are decoupled, massless degrees of freedom parametrizing the two flat complex dimensions longitudinal to the degenerate locus. We shall henceforth ignore them as well. We are left with the degrees of freedom z 3 ≡ z and y 1 ≡ y, with superpotential 21 mzy 2 , and a Z2 action under which the field z is invariant but the field y → −y. Now consider a circle in the z plane surrounding the degenerate locus z = 0. Treating the theory as a fibration means doing the path integral in two steps. First hold fixed the base coordinate z and allow y to fluctuate, deriving an effective theory for z. Then quantize z, with its evolution specified by the effective Hamiltonian derived in the first step. 3 The Z gerbe on P3 is banded, hence the restriction is also banded. The restriction also should be non2 trivial, just as the original gerbe on P3 . Briefly, in light of
H 2 (P3 , Z2 ) −→ H 2 (P3 − , Z2 ) −→ H 1 (, Z2 ), if is smooth, then it is simply-connected, and so H 1 (, Z2 ) = 0, which implies that the restriction of the gerbe with characteristic class −1 mod 2 is another nontrivial gerbe on P3 − . 4 The question of triviality of the cover is local near det A = 0, and locally the cover is the subvariety in P3 × C given by z 2 = f (x), where f = det A, x indicates homogeneous coordinates on P3 , and z is a coordinate on C. So the cover is trivial if and only if we can extract locally a square root of f = det A. But if f has a square root locally, then it has a square root globally, i.e. f = g 2 for some homogeneous polynomial of degree 4. So the double cover is trivial if and only if det A is a square, which usually is not the case.
612
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
This Wilsonian treatment of the path integral breaks down only in the neighborhood of z = 0, where the y degree of freedom becomes light. However we can still ask about the boundary conditions for wavefunctions in a region defined by removing a disc D containing the origin of z-space. For values of z in Cz − D we know that because of the noneffective Z2 orbifold action, the fiber theory of y has two degenerate vacua, in one of which y is untwisted and in the other of which y is twisted. As was argued in [2], the infrared limit of the y theory over a given point in Cz − D is equivalent to a disconnected theory of two discrete points. So we have two points fibered over the complement of a disc in the z-plane. These two points are defined by the universe operators U± ≡ 21 (1 ± ϒ), where 1 is the untwisted vacuum and ϒ is the twisted vacuum. We would like to find out whether the effective theory for z defines a trivial or nontrivial fibration over Cz − D. How can we understand the monodromy of the two points over the origin? The two points must either be exchanged or remain the same as one executes a loop in the z plane around the boundary of the disc D. If the points remain the same, then P± come back to themselves, or equivalently the twisted vacuum ϒ comes back to itself. If the two fiber points are interchanged by going around ∂ D, then that is equivalent to P± being interchanged with P∓ , which in turn is the same as saying that the twisted vacuum ϒ comes back to itself up to a minus sign. Next, we need to determine whether ϒ comes back to itself with a + or a − sign when the string is moved 360 degrees around the boundary of the disc D. Take the worldsheet to be compact with radius rws , and the disc D in the z-plane to have radius R D . Then consider contributions to the worldsheet path integral in which the string moves around in a circle |z| = R > R D in a time T . We assume T rws and also T 1/(m R). Without loss of generality we shall also assume rws m R 1, so that the mass term is important only for the dynamics of zero modes on the circle and can be ignored for the nonzero modes. For any history of z the dynamics of y are exactly Gaussian. That is, the field y and its fermionic superpartners are controlled by a quadratic but time-dependent field theory. Since the field z couples only in the superpotential, the kinetic term for y is z-independent, and only its potential is z-dependent. Assume z is independent of the spatial worldsheet coordinate σ 1 . Also assume z stays exactly on the circle |z| = R and only its phase changes as a function of worldsheet time σ 0 ≡ t: z = R exp(iω(t)). Since the mass term for the scalar y in the Y multiplet is |z|2 |y|2 , it equals R 2 |y|2 for the particular z-history we consider. Thus the phase ω decouples completely from the dynamics of the boson y, which is then just a massive boson which can be integrated out trivially. y y† The fermions ψ± , ψ± do however couple to the phase of z. Their Yukawa coupling is y
y
y†
y†
LY uk. ≡ mzψ− ψ+ − mzψ− ψ+ which for our choice of history for z equals
y y y† y† LY uk. = m R exp(iω(t))ψ− ψ+ − exp(−iω(t))ψ− ψ+ .
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
613
Since z σ 1 = 0, the y-fermion theory is translationally invariant in the σ 1 direction. It is also Gaussian, so the dynamics factorizes into an infinite product of finite dimensional Hilbert spaces labelled by spatial Fourier modes. That is, if 1 y y ψ±(s) ≡ exp(−isσ 1 )ψ± (σ 1 ) 2π y†
y
y†
and similarly for ψ± , then the eight operators ψ±(±r ) and ψ±(±r ) are decoupled from all other operators with distinct absolute value of r . We are working in the limit rws m R 1 so for s = 0 the mass terms of magnitude m R make a contribution to the frequency of the oscillators which is negligible compared to the contribution s/rws from the spatial gradient. Therefore the nonzero mode oscillators can never contribute to the Berry phase. To completely specify the fermions, we have to specify their boundary conditions as we traverse the circle. Without loss of generality, we may assume the fermions are in an NS sector on the circle. In the untwisted NS sector, all values of s are half-integral, so there are no zero modes, and as a result, from the analysis above there is no contribution to the Berry phase as ω is varied from 0 to 2π . In the twisted NS sector the fermions are integrally moded, and so from the analysis above there is a contribution to the Berry y y† † phase from the zero mode oscillators b± ≡ ψ±(0) and b± ≡ ψ±(0) . If we specified that the fermions were in an R sector on the circle, the analysis would be completely symmetric, just exchanging the interpretation of twisted and untwisted sectors. The eigenvalue of the monodromy on ϒ can therefore be obtained by restricting to zero modes, and so is equivalent to the calculation of the Berry phase of the system † † H ≡ m R exp(iω)b− b+ − exp(−iω)b− b+ as ω varies from 0 to 2π . The result is that the Berry phase on ϒ is −1. We can see this as follows. Represent the fermionic oscillators as gamma matrices: 1 b+ ≡ √ ( 1 + i 2 ), 2 1 b− ≡ √ ( 3 + i 4 ). 2 It is clear that the modes then satisfy canonical anticommutation relations. Taking the representation 1 2 3 4 (5)
= = = = =
σ 1 ⊗ σ 1, σ 2 ⊗ σ 1, σ 3 ⊗ σ 1, 1 ⊗ σ 2, 1 ⊗ σ 3,
we find that 1 b− b+ = (σ 1 + iσ 2 ) ⊗ (1 + σ 3 ), 2
614
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
so the Hamiltonian is
H (t) = 2m R
0 exp(iω(t)) exp(−iω(t)) 0
1 ⊗ (1 + σ 3 ). 2
For the rest of the analysis, we will implicitly carry along the ⊗(1/2)(1 + σ 3 ) as a spectator. Omitting that factor, the Hamiltonian is 0 exp(iω(t)) H (t) = 2m R . exp(−iω(t)) 0 The Berry phase is the eigenvalue of time translation during the period [0, T ] in which ω = 2π , taking the limit ω˙ ∼ T1 2m R. For this particular system the limit is unnecessary and the Berry phase is exact even for T −1 comparable with 2m R or large compared to it. The result is a phase shift given by 21 (2π ) = π , and such a phase shift is equivalent to a sign flip: cos(x + π ) = − cos(x), sin(x + π ) = − sin(x). Now let us compute the Berry phase. Berry’s definition of parallel transport is that a state |ω always be an energy eigenstate as the Hamiltonian varies through the space of nondegenerate operators, and that δ |ψ be orthogonal to |ψ. Equivalently, for a set of energy levels |n, Berry’s parallel transport can be expressed as δ |n =
m| (δ H ) |n m =n
E 0[n] − E 0[m]
|m .
There are just two energy eigenstates |± which always have eigenvalues E 0[±] = ±K . The solutions to these equations are: ⎛ ⎞ c exp − 21 iω ⎠ |± = ⎝ 1 . ± exp + 2 iω It is clear that as ω → ω + 2π , each state gets a phase of π , or equivalently a sign of −1. Thus, since we are working in the twisted NS vacuum ϒ, we see that ϒ gets a Berry phase of −1. The Berry phase arose from fermion zero modes, and there are not any present in the untwisted NS vacuum 1, so as noted earlier the untwisted NS vacuum does not get any Berry phase. (R sectors are symmetric.) We conclude that transporting a pointlike string state5 around a loop in Cz − D which surrounds D once induces a trivial phase on the untwisted NS vacuum 1 and a phase of −1 on the twisted NS vacuum ϒ. This will also be true of all bulk NS states, since our calculation is unaffected by exciting degrees of freedom in the z 1,2 coordinates and their superpartners, in the 3+1 macroscopic Minkowski coordinates X 0,1,2,3 , ψ 0,1,2,3 , ψ˜ 0,1,2,3 of visible spacetime, or even oscillator modes of the z 3 coordinate and its superpartners. The zero modes in the angular and radial z 3 directions are what we have held fixed in order to perform the Berry phase calculation. Again, R sector states are symmetric. Since all manipulations above are entirely local, the calculation holds for any model in which the degenerate locus is a smooth hypersurface. 5 That is, a configuration which is independent of the spatial worldsheet coordinate σ 1 .
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
615
Thus, the Landau-Ginzburg point of the GLSM for P7 [2, 2, 2, 2] seems to consistently describe a branched double cover of P3 . To summarize our progress so far, away from the branch locus the GLSM at low energies reduces to an abelian gauge theory with nonminimal charges – which describes a gerbe, which physics sees as a multiple cover [2,18–21]. The gerbe in question is banded, which would imply a trivial cover, were it not for Berry phases which wrap the components nontrivially, and so gives us a nontrivial double cover. 2.4. Monodromy around the Landau-Ginzburg point. We have discussed how the Landau-Ginzburg point appears to be describing a nonlinear sigma model on a branched double of P3 . In this section we will check that interpretation indirectly by computing the monodromy about the Landau-Ginzburg point and showing that it is compatible with a nonlinear sigma model interpretation, namely, that it is maximally unipotent. As we will also discuss related monodromy computations for other models, in this section let us first set up some generalities. We consider a Calabi-Yau 3-fold, X , with a 1-dimensional Kähler moduli space. For simplicity, we will take X to be simply-connected. Let the generator of H 2 (X ) be ξ . Then, one topological invariant is the positive integer, p, such that ξ 2 = pη, where η is the generator of H 4 (X ). Let ρ = ξ η be the generator of H 6 (X ). We obtain another integer, q, by writing c2 (X ) = 2qη. As our basis for K 0 (X ), we will choose a set of generators, whose ring structure mimics that of the even-dimensional cohomology. To whit, we will choose 1. 2. 3. 4.
the class [O] of the trivial line bundle O. a = [H ] [O], where H is the line bundle with c1 (H ) = ξ . [b], where a ⊗ a = pb. [c], where c = a ⊗ b.
There is a skew-bilinear form on K 0 (X ), given by (v, w) = I nd ∂ v⊗w 1 = ch 3 (v ⊗ w) + (c1 (v) − c1 (w))c2 (X ). 12 X Expressed in our basis, this skew-form is represented by the matrix ⎡ ⎤ 0 −( p + q)/6 −1 −1 0 1 0 ⎥ ⎢ ( p + q)/6 =⎣ . 1 −1 0 0 ⎦ 1 0 0 0 On the mirror, the periods of the holomorphic 3-form obey a Picard-Fuchs equation with three regular singular points. Two of the three monodromies have very simple interpretations in terms of operations in K-theory.
616
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
The monodromy matrices one extracts from this take a nice form, when thought of in terms of natural operations in K-theory. The large-radius monodromy is M∞ : v → v ⊗ H, where H is the hyperplane bundle, corresponding to shifting the B field as one walks around the large-radius limit in the complexified Kähler moduli space. Such a largeradius monodromy is necessarily maximally unipotent, meaning, (M∞ − 1)n+1 = 0, (M∞ − 1) p = 0, 0 < p ≤ n, where n is the dimension of the space (in the present case, 3), for the simple reason that in K-theory, we can think of (M∞ − 1) as tensoring with ([H ] [O]), and tensoring with ([H ] [O]) is nilpotent – for example, ch(H O)n+1 = 0. Furthermore, if the local coordinates on the moduli space are a cover, then it might take several turns about the limit point to reproduce all of M, so that in general, the monodromy need merely be maximally unipotent in the weaker sense that N N (M∞ − 1)n+1 = 0, (M∞ − 1) p = 0, 0 < p ≤ n
for some positive integer N . In principle, by checking whether the monodromy about a given point in moduli space is maximally unipotent, we can check whether that point can be consistently described by a nonlinear sigma model on a smooth Calabi-Yau target. The monodromy about the (mirror of the) conifold is M1 : v → v − (v, O)O, where O is the trivial line bundle. This is the Witten effect, in essence. In a type II string, an electrically-charged particle becomes massless at this point, and so magnetically-charged particles pick up an electric charge proportional to the effective theta angle, which shifts when one circles the conifold point. In these one-Kähler-parameter, simply-connected, Calabi-Yau’s, only one species becomes massless: the wrapped D6-brane. Of course, the monodromy around z = 0 is the product of the other two. In our basis, these monodromies are represented by the matrices ⎡ ⎤ ⎡ ⎤ 1 0 0 0 1 −( p + q)/6 −1 −1 1 0 0 ⎥ ⎢1 1 0 0⎥ ⎢0 M∞ = ⎣ , M1 = ⎣ . 0 p 1 0⎦ 0 0 1 0 ⎦ 0 0 1 1 0 0 0 1 Now, let us restrict to P7 [2, 2, 2, 2]. This is the case p = 16, q = 32. The Picard Fuchs equation for the mirror is D (z) = 0, d where D is the differential operator (θ − z = z dz ):
D = θz4 − 16z(2θz + 1)4 . The large-radius point is z = ∞. The (mirror of the) conifold is z = 1, and our mysterious Landau-Ginzburg point is z = 0.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
617
In our chosen basis for K 0 (X ) = Z4 , the skew bilinear form, (·, ·) is represented by the matrix ⎡
⎤ 0 −8 −1 −1 1 0 ⎥ ⎢8 0 = ⎣ , 1 −1 0 0 ⎦ 1 0 0 0 and ⎡
M∞
⎤ 0 0⎥ 0⎦ 1
1 0 0 ⎢1 1 0 = ⎣ 0 16 1 0 0 1
is the large-radius monodromy, ⎤ 1 −8 −1 −1 0 0 ⎥ ⎢0 1 , M1 = ⎣ 0 0 1 0 ⎦ 0 0 0 1 ⎡
is the conifold monodromy and ⎡
⎤ 1 −8 −1 −1 ⎢ 1 −7 −1 −1 ⎥ M0 = M∞ M1 = ⎣ . 0 16 1 0 ⎦ 0 0 1 1 This last one does not look too illuminating. However: 1. By an integer change of basis (respecting the quadratic form above), M0 can be put in the form ⎡
1 ⎢1 M0 = − ⎣ 0 0
0 1 1 0
0 0 1 1
⎤ 0 0⎥ 0⎦ 1
which is minus the large-radius monodromy of the double-cover of P3 . 2. Using the basis above, we can see the monodromy is maximally unipotent. (M02 − 1)4 = 0, (M02 − 1) p = 0, 0 < p < 4. Thus, we see that the monodromy about the Landau-Ginzburg point is maximally unipotent, and hence compatible with a geometric interpretation of the Landau-Ginzburg point of this model.
618
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
2.5. A puzzle with a geometric interpretation of the Landau-Ginzburg point. So far we have described how the Landau-Ginzburg point of the GLSM for P7 [2, 2, 2, 2] describes a branched double cover of P3 , branched over a degree eight locus – the octic double solid Calabi-Yau threefold. In particular, we have argued how away from the branch locus, the Landau-Ginzburg point is a Z2 gerbe, which physics sees as a double cover, and because of a Berry phase, a nontrivial double cover. We checked this interpretation by computing the monodromy about the Landau-Ginzburg point, which we saw is consistent with a geometric interpretation. This seems to be a solid description, but there is a puzzle in the analysis of the Landau-Ginzburg point that is problematic for a strict geometric interpretation. Specifically, the geometry is singular, but the GLSM (at the Landau-Ginzburg point) behaves as if it were on a smooth manifold. In this section, we will go over this difficulty. In the next section, we will describe how this problem is resolved, and simultaneously describe how the relationship between the large-radius and Landau-Ginzburg points can be understood mathematically. Again, the problem with an interpretation of the Landau-Ginzburg point as a branched double cover is that the CFT does not degenerate at points where the branched double cover is singular – the gauged linear sigma model seems to see some sort of resolution of the branched double cover. (We will elaborate on the precise nature of this resolution later; for the moment, we merely wish to establish the physical behavior of the CFT.) Following [1], the CFT will be singular at a point in the target space if there is an extra noncompact branched over that point in the GLSM. Now, in the GLSM, the F term conditions in this model can be written Ai j ( p)x j = 0, j
ij
xi
∂ Ai j x j = 0. ∂ pk
On the branch locus, the first F term condition is trivially satisfied, but not the second, the second prevents the branch locus from having a singularity generically. Physically, the CFT will only be singular for those vectors (xi ) which are eigenvectors of zero eigenvalue of the matrix (Ai j ), and also simultaneously eigenvectors of zero eigenvalue of each matrix (∂ Ai j /∂ pk ) for each p. Furthermore, for generic quadrics, there are no such solutions – as we will see below the CFT described by the GLSM behaves as if it is describing a smooth space. Let us compare this to a mathematical analysis. If the branch locus is described as { f (x1 , . . . , xn ) = 0}, then the double cover is given by {y 2 = f (x1 , . . . , xn )}, and it is straightforward to check that the double cover {y 2 = f } will be smooth precisely where the branch locus { f = 0} is smooth. Thus, geometrically, the branched double cover will be singular only at places where the surface {det A = 0} is singular, and for generic quadrics, there will be singular points on the branched double cover. Thus, the condition that the hypersurface {det A = 0} be singular, is different from the condition for flat directions in the GLSM that we derived above, and so ultimately as a result, the GLSM behaves as if it were on a smooth space, whereas the branched double cover is singular. Global analysis. Let us now justify the statements made above regarding singularities. First, let us discuss the singularities (or rather, lack thereof) in the GLSM. For the first equation to have a non-trivial solution, p must be in the discriminant of our family
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
619
of quadrics and x must be in the kernel of the matrix A( p). Choose an affine chart on P3 which is centered at p. Let u 1 , u 2 , u 3 be the local coordinates in which p = (0, 0, 0). In these terms we have that A = C0 +C1 u 1 +C2 u 2 +C3 u 3 , where Ci are constant symmetric 8 × 8 matrices. Note that for a generic choice of quadruple of quadrics the family A has a determinant which is not identically zero as a function of the u i ’s. On the other hand the first equation says that there exists a non-zero vector x such that C0 x = 0, and the second system of equations says that B1 x = B2 x = B3 x = 0. This however implies that A(u)x = 0 for all u, i.e. det A(u) = 0 identically in u. This gives a contradiction. Next, let us turn to the singularities of the branched double cover. Start with the projectivization P35 of the 36 dimensional vector space of all 8×8 symmetric matrices. The space of singular quadrics is a divisor D ⊂ P35 - the divisor consisting of all quadrics of rank at most 7. Explicitly D = {[A] ∈ P35 | det A = 0}. Our four quadrics span a linear P3 ⊂ P35 and the branch locus is just the intersection P3 ∩ D. Now the singularities of the intersection P3 ∩ D occur at the points where P3 is not transversal to D. Note that there are two ways in which this can happen: 1. when P3 intersects D at a smooth point of D but not transversally, and 2. when P3 passes through a singular point of D. These two types of singularities behave differently: later when we discuss homological projective duality, we will see that the sheaf of Clifford algebras that we get in the h.p.d. will be locally-free at singularities of type 1 and will not be locally-free at singularities of type 2. So this sheaf will be a sheaf of Azumaya algebras on the complement of the points of type 2, i.e. on this complement we will have a gerbe over the double cover. Across these points the sheaf of Clifford algebras gives a noncommutative resolution of the singularities. Now note that the singular locus of D consists of all quadrics of rank at most 6. Every quadric of rank exactly 6 is a cone with vertex P1 over a smooth quadric on P5 . So the dimension of the locus of quadrics of rank 6 in C8 is equal to dim Gr (2, 8)+dim S 2 C6 = 12 + 21 = 33 or projectively is equal to 32. So Sing D is a non-degenerate subvariety of codimension 3 in P36 and so every P3 intersects it. So the double cover is singular and generically has finitely many singularities of type 2. To better understand this matter, we shall return to the local model of the branched cover. We consider a local model of a geometric singularity of the branched cover, where the mass matrix for the y degrees of freedom drops in rank by two. This occurs when the discriminant locus has a surface singularity of the most generic kind – an ordinary A1 singularity. This is described geometrically by a conifold singularity of the total space of the branched cover, as we shall see quite directly. However we will also establish that the CFT is nonetheless nonsingular; there is no noncompact branch, even over the point at which the degenerate locus has an A1 singularity and the total space has a conifold singularity. Setup of the local model. In this section we shall follow the same notation as in our analysis of Berry phases in Sect. 2.3. Two of the six y degrees of freedom, y1,2 , are involved in the model in a nontrivial way. These are the ones which are simultaneously massless over the singular point in the degenerate locus. Label them y α for α ∈ {1, 2}. They are coupled to the z a multiplets through a z-dependent mass matrix which vanishes at the origin. The other six y’s are massive everywhere.
620
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
The superpotential is α β
W = Mαβ (z)y y
+
8
(y i )2 .
i=3
We integrate out y3 , . . . , y8 trivially. The simplest choice for M which manifests an SU (2) global symmetry is Mαβ (z) ≡
1 a za , mαγ σγβ 2
where σ a are the standard Pauli matrices. The z a transform as a 3 and the yα transform as a 2. Everything else is a 1. In components we have 1 1 1 2 z 1 + i z 2 −z 3 y W = m y y 2 . −z −z + i z y 3 1 2 2 The degenerate locus is given by the equation 0 = det M = −m 2 z 12 + z 22 + z 32 = −m 2 z a2 . a
The origin is an ordinary double point singularity, or A1 surface singularity, of the variety ⊂ C3 . That is, the singularity is locally the quotient singularity C2 /Z 2 . It is easy to see that a branched cover over C3 with branch locus { a z a2 = 0} is a conifold. Introduce a fourth variable u and embed the cover into C4 by the equation z a2 . u ≡ ± a
Defining u ≡ i z 4 and squaring both sides we have the equation z 12 + z 22 + z 32 + z 42 = 0, which is the defining equation of the undeformed conifold, in standard form. Nonsingularity of the CFT. Despite the fact that the target space in this local model is geometrically a singular conifold, the CFT is nonsingular. A fortiori, this establishes that the theory is inequivalent to the CFT of the standard conifold, which is singular. More generally, as we outlined with a global analysis at the beginning of this section, the GLSM singularities are different from the geometric singularities. We will show here that the CFT can be smooth at a geometric singularity of the branched double cover, to drive home the distinction. To see this, it suffices to notice that there is no noncompact branch at the origin. The F-term equations for z a are a α β y y = 0. αγ σγβ
One can check directly, component by component, that this does indeed set both y α = 0. An easier way to see this is to note that SU (2) is transitive on spinors of fixed norm and the F-term equations are SU (2) invariant, so either all nonzero values of y α satisfy the F-term equations or else none of them does. The former possibility is obviously not true
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
621
so the y α must vanish classically, despite the fact that they both become massless at the origin. Thus, the geometric singularities of the branched double cover do not coincide with singularities of the CFT arising at the Landau-Ginzburg point, which is one problem with the proposal that the Landau-Ginzburg point flow to a nonlinear sigma model on a branched double cover. We shall resolve this discrepancy in the next section, by arguing that the technically correct interpretation of the Landau-Ginzburg point is that it flows to a nonlinear sigma model on a ‘noncommutative resolution’ of the branched double cover. In other words, the branched double cover interpretation will be correct generically, but the resulting CFT is not quite globally the same as a nonlinear sigma model on the branched double cover. 2.6. Resolution of this puzzle – new CFT’s. Although the Landau-Ginzburg point seems to be very nearly equivalent to a nonlinear sigma model on a branched double cover of P3 , the problem in the last section has made it clear that such an interpretation can not be completely correct. In addition, we also have a problem of understanding how to relate the large-radius and Landau-Ginzburg points geometrically. Ordinarily, in GLSM’s the Kähler phases are related by birational transformations, yet no birational transformation exists in this case, as pointed out earlier in Sect. 2.1. We propose that these problems are resolved and understood by virtue of Kuznetsov’s “homological projective duality” [14,22,23]. The homological projective dual of P7 [2, 2, 2, 2] is a “noncommutative resolution” of the branched double cover of P3 that we have seen. We shall describe homological projective duality in greater generality in Sect. 2.6.1, but let us take a moment to review what this means specifically in this case. The word ‘noncommutative’ in this context is somewhat misleading. Kuznetsov’s work [14,22,23] and related papers define spaces by categories of sheaves, and use the term ‘noncommutative space’ to refer to any space (or other object) whose sheaf theory yields the defining category. A noncommutative space could be an ordinary space, an ordinary space with a flat B field that twists sheaves, or even a Landau-Ginzburg model. In particular, a ‘noncommutative space’ need not be associated with a noncommutative algebra. In the present case, the noncommutative space that is homological projective dual to P7 [2, 2, 2, 2] is the pair (P3 , B), where B ∈ Coh(P3 ) is the sheaf of even parts of Clifford algebras over P3 . (The category that defines this noncommutative space is the category of coherent sheaves on P3 which are also modules over the sheaf B.) This pair (P3 , B) defines a pair (Z , A), where Z is the branched double cover of P3 and A is essentially just B but reinterpreted. In the next section, we shall elaborate on these structures and also describe how they arise physically in matrix factorization. To put this in perspective, this means that the conformal field theory obtained as the IR limit of the Landau-Ginzburg point of the GLSM, is not a nonlinear sigma model on a branched double cover, though it is close. Rather, it is a new conformal field theory, that locally on smooth patches behaves like a nonlinear sigma model on the branched double cover, but in a neighborhood of a singularity, does something different. (We will justify this interpretation in more detail later, and we will leave a more thorough examination of such new conformal field theories, associated to Kontsevich’s notion of an nc space, to future work.) This addresses the problem described in the last section. If we are describing some sort of resolution of the branched double cover, rather than the branched double cover
622
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
itself, then physical singularities will not coincide with geometric singularities of the branched double cover. 2.6.1. Homological projective duality Homological projective duality is a notion that generalizes all of the equivalences described here. It is defined in terms of both the derived categories of the spaces, and in terms of embeddings into projective spaces: varieties X and Y equipped with morphisms into the dual projective spaces f : X → PV , g : Y → PV ∗ (V a vector space) are homologically projective dual if the derived category of Y can be embedded fully and faithfully into the derived category of the universal hyperplane section of X (a subset of X × PV ∗ ) in a certain way. Homological projective duality was introduced in [22]; it is described for quadrics in [14] and for Grassmannians in [23]. The simplest versions of this correspond to classical duality6 between hyperplanes and points of projective spaces. For a vector space V , the embedding PV [1] → PV defined by the inclusion of a hyperplane (degree 1) into its ambient projective space, is homologically projective dual to the embedding pt → PV ∗ of the dual point into the dual projective space. More complicated examples can be defined by e.g. Veronese embeddings. Recall (from e.g. [24, p. 23]) that the Veronese map of degree d is a map Pn → P N of the form [x0 , . . . , xn ] −→ [. . . , x I , . . .], where the x I range over all monomials of degree d. Thus, n+d N = − 1. d In particular, a Veronese map of degree 2 defines an embedding PV → PSym2 V . Kuznetsov shows [14] that the double Veronese embedding is homologically projective dual to (PSym2 V ∗ , B0 ) −→ PSym2 V ∗ , where the pair (PSym2 V ∗ , B0 ) defines a ‘noncommutative’ or nc space. (See Appendix A for an overview of nc spaces.) This noncommutative space is defined by sheaves that are modules over B0 , where B0 is the sheaf of even parts of Clifford algebras on P(Sym2 V ∗ ): B0 = OP(S 2 V ∗ ) ⊕ 2 V ⊗ OP(S 2 V ∗ ) ⊕ 4 V ⊗ OP(S 2 V ∗ ) ⊕ · · · . As a practical matter, what arises physically is the induced action of homological projective duality on linear (hyperplane) sections, not precisely bare homological projective duality itself. Suppose we have dual maps f : X → PV and Y → PV ∗ . Now, let L ⊂ H 0 (PV, O(1)) be a set of hyperplanes, and define X L to be the complete intersection of those hyperplanes with the image of X . Since L is a set of linear forms on 6 In other words, for a projective space PV , a point in the dual projective space PV ∗ with homogeneous coordinates [a0 , . . . , an ] corresponds to a hyperplane in the original projective space defined by
a0 x0 + · · · + an xn = 0, where [x0 , . . . , xn ] are homogeneous coordinates on PV .
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
623
PV , the projectivization PL is naturally a linear projective subspace of PV ∗ . Define Y L to be the intersection of the image of Y in PV ∗ with PL. Kuznetsov proves in [22] that the derived categories of X L and Y L each decompose into several Lefschetz pieces with one essential last piece in the Lefschetz decomposition. He also shows that the essential pieces of X L and Y L are equivalent. We will see below that the nc spaces defined by these essential pieces are exactly the ones related by the change of phase in the GLSM. To be specific, let us consider complete intersections of quadrics. We have just described the induced action on hyperplanes: to describe the induced action on quadrics, we must find a way to re-embed so that the quadrics become hyperplanes,7 in effect. Now, a quadric in PV is the pullback of a linear polynomial on PSym2 V under the double Veronese embedding PW → PSym2 V . For example, if four of the homogeneous coordinates on the target are x0 x1 , x0 x2 , x20 , x1 x2 , then the hyperplane (x0 x1 ) + 3(x0 x2 ) − 2(x20 ) + 9(x1 x2 ) = 0 in the target PSym2 V is the same as a quadric hypersurface in PV . So, we consider hyperplanes on the image of PV in PSym2 V , which is equivalent to working with quadrics on PV . Let us work through a particular example, that of a complete intersection of quadrics in P7 . From the arguments above, let us begin with the double Veronese embedding P7 → P35 , which is dual to (P35 , B0 ) → P35 . Suppose we have a space L ⊂ H 0 (P35 , O(1)) of quadrics of which we wish to take the complete intersection. Let X L denote that complete intersection, i.e., X L = ∩q∈L {q = 0} ∩ P7 . (For example, if L is four-dimensional, then X L is the complete intersection of four quadrics in P7 , precisely the example we have been studying in detail so far in this paper.) Since L is a space of linear forms on P35 = PSym2 V , the projectivization PL is naturally a linear projective subspace of P35 = PSym2 V ∗ . Define the dual linear section Y L to be the intersection of PL with whatever is embedded in the P35 = PSym2 V ∗ . In particular, that means Y L = (PL , B0 |PL ). When L is one-dimensional and X L is just one quadric, then PL is a point and Y L is a point equipped with an nc structure sheaf which is an even part of a Clifford algebra. When L is four-dimensional (so that X L = P7 [2, 2, 2, 2]), then Y L = P3 , and as we shall see in the next section, the nc space defined by P3 with the restriction of B0 is a ‘noncommutative’ resolution of a branched double cover of P3 . Finally, we need to take essential pieces in the derived category, but since both sides are Calabi-Yau, the essential pieces are the entire derived category. A point to which the reader might object is that the dual spaces obtained are noncommutative spaces, at least in Kontsevich’s sense – meaning, spaces defined by their sheaf theory. In simple cases, the duals will be honest spaces, but for example when the dual variety is singular8 , then the noncommutative space will be a noncommutative resolution of singularities, matching the underlying variety at smooth points but 7 A careful reader will note that there is a potential presentation-dependence problem lurking here. If homological projective duality is defined on a choice of linear sections, then different choices, different ways of rewriting the complete intersection as a complete intersection of hyperplanes in a projective space, might give rise to different duals. This might be partly fixed by a nonobvious uniqueness theorem, and partly it might correspond to different Kähler phases in GLSM’s. We will not attempt to root out this issue here, but instead leave it for future work. 8 It is possible to also small-resolve the singularities of the branched double cover, but the result is necessarily non-Kähler. For more information on such non-Kähler small resolutions see for example [25,26]. In any event, for our purposes, this is largely irrelevant, as we can tell from the sheaf theory – matrix factorizations in the UV Landau-Ginzburg model – that physics really is seeing precisely the noncommutative resolution, a fact that will be described in detail in the next section.
624
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
doing something different at singular points. In fact, we shall see in the next section that this behavior matches physics – the Landau-Ginzburg points of these GLSM’s have sheaf theory (defined by matrix factorizations) which precisely matches the sheaf theory obtained by homological projective duality. So, physics sees noncommutative spaces; in fact, these GLSM’s give a concrete realization of what it would mean for a string to propagate on a noncommutative space, in this sense. Put another way, the CFT’s at the Landau-Ginzburg points are, in general, new types of CFT’s – they look like ordinary nonlinear sigma models close to smooth points of the branched double cover, but are different close to singular points. In the paper [6], the relevant homological projective duality began with the duality between the Plücker embedding G(2, V ) → P(Alt2 V ), which was homologically projective dual to (Pf, B0 ) → P(Alt2 V ∗ ), where Pf denotes a Pfaffian variety and B0 the sheaf of even parts of Clifford algebras that defines the structure of a noncommutative space over Pf. As here, the physically-relevant version of homological projective duality appearing there was its induced action on hyperplanes. We conjecture that Kähler phases of GLSMs are related by homological projective duality. Unfortunately, it is not possible to check this conjecture at present, as much more needs to be understood about homological projective duality. For example, the simplest flop (between small resolutions of the basic conifold) is known [22, Theorem 8.8] to work through homological projective duality, but it is not known whether more general flops are also related by homological projective duality. 2.6.2. Noncommutative algebras and matrix factorization. In this section we shall review some pertinent algebraic structures arising mathematically in homological projective duality in this example, and how they can be understood via matrix factorization. Let us begin by reviewing the mathematics [14] of homological projective duality in this case. Consider the complete intersection X of four quadrics in P7 . It is h.p.d. to a non-commutative variety (P, B), where P ∼ = P3 is the parameter space for the set of 6-dimensional quadrics that cut out X ⊂ P7 , and B ∈ Coh(P) is the sheaf of even parts of Clifford algebras associated with the universal quadric π : Q → P over P. In physics terms, the universal quadric is the GLSM superpotential i j φi Ai j ( p)φ j , and for each point on P3 we have a quadric, which defines a metric for which we can associate a Clifford algebra. The family of quadrics π : Q → P degenerates along a discriminant surface ⊂ P of degree 8. Equivalently, instead of taking the nc space (P, B), we can consider the double cover f : Z → P branched along , together with a sheaf of algebras A → Z for which f ∗ A = B. Let us take a moment to understand what happens with the data (P, B), or equivalently (Z , A) in the special situation when the octic becomes singular. The octic may become singular in two different ways. First, the plane P ⊂ P(H 0 (P7 , O(2))) can become tangent to the discriminant in P(H 0 (P7 , O(2))). In this case the double cover is singular, but the sheaf A of algebras on the double cover Z is a sheaf of Azumaya algebras. Second, it can happen that P contains a quadric of corank 2. In this case the sheaf of algebras A is not locally free at the corresponding point of the double cover. So, we get a truly non-commutative situation. The structure (P, B) arises physically via matrix factorization. Let us return briefly to the GLSM superpotential, the ‘universal quadric’ i j φi Ai j ( p)φ j . On the face of it, this describes a hybrid Landau-Ginzburg model, apparently fibered over P = P2 .
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
625
At each point on P, we have an ordinary Landau-Ginzburg model (in fact, a Z2 orbifold) with a quadric superpotential. Now, matrix factorization for quadratic superpotentials was thoroughly studied in [27]. There, it was discovered that the D0-branes in such a Landau-Ginzburg model have a Clifford algebra structure. The D0-branes in a Landau-Ginzburg model with n fields and a quadratic superpotential give rise to a Clifford algebra over those n fields, with associated metric defined by the superpotential. In the present case, where we have fibered such Landau-Ginzburg models over P, the fibered D0-branes, or more accurately D3-branes when P is three-dimensional, will have the structure of a sheaf of Clifford algebras. (After all,9 we can equivalently work in the B model, where the Born-Oppenheimer approximation for large underlying space becomes exact.) We can refine this even further. In our examples, the fibers are not just Landau-Ginzburg models with quadratic superpotential, but rather are Z2 orbifolds of Landau-Ginzburg models with quadratic superpotentials, so our fibered D0-branes will have the structure of a sheaf of even parts of Clifford algebras, as that is what survives the Z2 orbifold. This is precisely the sheaf B appearing mathematically. Furthermore, as described in [27, Sect. 7.4], all the B-branes in a Landau-Ginzburg model with a quadratic superpotential are modules over the Clifford algebra, so in particular all the B-branes in the present case should be modules over the sheaf of Clifford algebras B. Thus, we see that matrix factorization in the hybrid Landau-Ginzburg model precisely recovers the algebraic structure of homological projective duality in this example. 2.7. Summary so far. We have examined the GLSM for P7 [2, 2, 2, 2] in detail. Before proceeding, let us review what we have found. First, at the Landau-Ginzburg point of this GLSM, we have found (via an analysis that requires understanding how stacks and gerbes enter physics) that at generic points, the theory seems to describe a branched double cover of P3 , another Calabi-Yau. This geometry is realized directly as a branched double cover, rather than as a complete intersection, which is certainly novel. Furthermore, the branched double cover and the original complete intersection P7 [2, 2, 2, 2] are not birational to one another. However, the theory at the Landau-Ginzburg point is not in the same universality class as a nonlinear sigma model on the branched double cover, but rather defines a new kind of conformal field theory, one corresponding to a noncommutative resolution of the space. This noncommutative resolution is defined mathematically by its sheaf theory, which we recover physically in matrix factorizations at the Landau-Ginzburg point of the GLSM. This structure, this duality between P7 [2, 2, 2, 2], is encoded mathematically in Kuznetsov’s homological projective duality [14,22,23]. It has been discussed elsewhere [6] how homological projective duality explains analogous dualities in nonabelian gauged linear sigma models. We shall see in the rest of this paper more examples of abelian gauged linear sigma models exhibiting homological projective duality. 2.8. Generalizations in other dimensions. Examples of this form generalize to other dimensions easily. The complete intersection of n quadrics in P2n−1 is related, in the same fashion as above, to a branched double cover of Pn−1 , branched over a determinantal hypersurface of degree 2n. These are Calabi-Yau, for the same reasons as discussed in 9 The Born-Oppenheimer approximation in this context suggests a theorem regarding the behavior of matrix factorizations in families, for which we unfortunately do not yet have a rigorous proof.
626
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
[2, Sect. 12.2]. Furthermore, the complete intersections and the branched double covers are related by homological projective duality.10 In the special case n = 2, we have elliptic curves at either end of the GLSM Kähler moduli space: the branched double cover is just the well-known expression of elliptic curves as branched double covers of P1 , branched over a degree four locus. In fact, the elliptic curve obtained at the LG point is the same as the elliptic curve at large-radius (though the isomorphism between them is not canonical). Technically, this follows from the fact that the branched double cover of P1 is the moduli space of degree 2 line bundles on P3 [2, 2], and as such, it is isomorphic after one chooses a distinguished point on P3 [2, 2]. In Sect. 3, we shall see an example in which this particular example of homological projective duality is essentially fibered over P1 × P1 . In the special case n = 3, we have K3’s at either end of the GLSM Kähler moduli space: the fact that K3’s can be described as double covers branched over sextic curves, as realized here at the Landau-Ginzburg point, is described in [8, Sect. 4.5], and the relation between the branched double cover and the complete intersection of quadrics is discussed in [29, p. 145]. However, the two K3’s obtained at either end of the GLSM Kähler moduli space are not isomorphic: one has degree 8, the other has degree 2. For n = 2, 3, the branched double cover is smooth, but beginning in n = 4 and continuing for higher n, the branched double cover is singular. For 4 ≤ n ≤ 7, the branched double cover has merely ordinary double points, and for n > 7, it has worse singularities. Already for n = 4, the branched double cover cannot be globally resolved into a smooth Kähler manifold – one can perform small resolutions locally at each ordinary double point, but globally any set of small resolutions will break the Kähler property. Physically, as we have seen, for n = 4 physics does not see a non-Kähler space, but instead sees a ‘noncommutative resolution,’ an nc space. 3. Example Related to Vafa-Witten Discrete Torsion 3.1. Basic analysis. A more complicated example with analogous properties can be built as follows. Consider a complete intersection of two quadrics in the total space of the projectivization of the vector bundle O(−1, 0)⊕2 ⊕ O(0, −1)⊕2 −→ P1 × P1 . The ambient toric variety can be described by a gauged linear sigma model with fields u, v, s, t, a, b, c, d, and three C× actions, with weights u v s t a b c d λ 1 1 0 0 −1 −1 0 0 µ 0 0 1 1 0 0 −1 −1 ν 0 0 0 0 1 1 1 1 The complete intersection is formed by adding two more fields p1 , p2 , each of weights (0, 0, −2) under (λ, µ, ν). The D-terms have the form rλ = |u|2 + |v|2 − |a|2 − |b|2 , rµ = |s|2 + |t|2 − |c|2 − |d|2 , rν = |a|2 + |b|2 + |c|2 + |d|2 − 2| p1 |2 − 2| p2 |2 . 10 To check this [28], note that the space P2n+1 in the double Veronese embedding is HP-dual to the sheaf 2 of even parts of Clifford algebras on the space P2n +5n+2 of all quadrics in P2n+1 . As a corollary, the derived category of a complete intersection of n quadrics in P2n+1 contains the derived category of (a noncommutative resolution of) a double covering of Pn−1 . This is discussed in [14].
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
627
The geometry described above is reproduced when rν 0. In the phase defined by further demanding rλ 0 and rµ 0, u and v form homogeneous coordinates on one of the P1 ’s in the base, and s, t form homogeneous coordinates on the other P1 . The fields a, b, c, d form coordinates on the fibers of the P3 bundle formed by projectiving the rank four vector bundle O(−1, 0)⊕2 ⊕ O(0, −1)⊕2 . Other phases with rν 0 give birational models of the same, related by flops. For example, consider the case that rλ 0 and rµ 0, then a, b form homogeneous coordinates on one P1 , and s, t form homogeneous coordinates on a second P1 . The geometry can still be described as a P3 bundle over P1 × P1 , which is true for all phases with rν 0. We discover branched double covers when we consider phases with rν 0. Suppose that rν 0 and rλ 0, rµ 0. In this phase, u, v form homogeneous coordinates on one P1 , s, t form homogeneous coordinates on a second P1 , and p1 , p2 form homogeneous coordinates on a third P1 . To fully understand this phase we need to closely examine the superpotential, which is of the form W = p1 Q 1 + p2 Q 2 , where Q 1 , Q 2 are quadratic polynomials in the eight variables au, av, bu, bv, cs, ct, ds, dt. Let γi enumerate the four variables a, b, c, d, then the superpotential can be written W =
γi Ai j ( p)γ j ,
(1)
ij
where Ai j is a symmetric 4 × 4 matrix with entries linear in the p’s and quadratic in combinations of s, t, u, v. This superpotential is manifestly a mass term for the γi , so generically the a, b, c, d’s will be massive, except over the locus where the rank of Ai j drops. That locus is defined by det A = 0, and is a degree (4, 4, 4) hypersurface in [u, v] × [s, t] × [ p1 , p2 ]. Away from that locus, where the a, b, c, d are massive, the only fields charged under the third U (1) gauge symmetry are p1 , p2 , which both have charge −2, so we have a branched double cover, branched over the locus det A = 0, much as in the previous example. Other Kähler phases with rν 0 are very similar. Their descriptions can be obtained by switching the pair (u, v) with (a, b) and/or switching (s, t) with (c, d). If we do the former, for example, then we rewrite the superpotential in the form of Eq. (1) but with the γi running over u, v, c, d, and the matrix Ai j a symmetric 4 × 4 matrix with entries linear in the p’s and quadratic in combinations of a, b, s, t. This phase then generically is a branched double cover of P1 × P1 × P1 (with homogeneous coordinates [a, b] × [s, t] × [ p1 , p2 ] instead of [u, v] × [s, t] × [ p1 , p2 ]), branched over the degree (4, 4, 4) locus {det A = 0}. This example is believed [30] to be another example, another physical realization, of homological projective duality, or rather, for each complete intersection phase with rν 0, the corresponding phase with rν 0 is believed to be related to the rν 0 phase by homological projective duality.
628
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
3.2. Some notes on the geometry. This particular example is closely related [31] to one discussed in [5] in connection with discrete torsion.11 There, recall one started with the quotient E × E × E/(Z2 )2 , for E an elliptic curve, and deformed to a space Y , describable as a double cover of P1 × P1 × P1 branched over a singular degree (4, 4, 4) hypersurface. In more detail, let X be the quotient of a product of 3 elliptic curves by the action of G = Z2 × Z2 , where each non-trivial element of G acts by negation on two of the elliptic curves, and leaves the third one unchanged. X can be viewed as a double cover of P1 × P1 × P1 , branched over a surface S of tri-degree (4, 4, 4), highly singular. One deforms X by deforming this surface S. Following [5], let us not deform S completely, until it is smooth, but rather only until one has 64 ordinary double points in S. This does not actually give a complete description of the allowable deformed branching loci S, but we should get that from the next description. Denote by Y the typical member of the family of allowable deformations of X . It is a C-Y 3-fold with 64 ODP’s, with rk Pic(Y ) = 3, and it moves in a 51 dimensional family. By a result of Mark Gross, Br(Y ) = Z2 . Next, let us consider the other half of the story. Let E be the vector bundle O(−1, 0)2 ⊕ O(0, −1)2 on P1 × P1 . The projectivization PE of E is a P3 bundle over P1 × P1 , and as such it comes with a natural O(1). (Depending on your convention as to what projectivization means, you may need to take -1 instead of 1 in the definition of E.) Let Z be the intersection of two general sections of O(2) in PE. It is a smooth C-Y 3-fold with h 1,1 = 3, h 1,2 = 51, and the projection to P1 × P1 exhibits it as a genus one fibration with no section but with a 4-section. The analysis of the birational models of Z is straightforward, and leads to a picture with Z2 × Z2 symmetry, like a square divided into 4 triangles by the diagonals. There are two sets of 8 P1 ’s in Z which can be flopped independently, giving rise to three more birational models for Z (flop one set, flop the other, or flop both sets simultaneously). Next, let us describe how the Y ’s are related to the Z ’s. Let f denote the cohomology class of a fiber of the fibration Z → P1 × P1 (which is an elliptic curve), and consider the moduli space of stable sheaves on Z of rank 0, first Chern class 0, second Chern class f , and third Chern class 2. What is meant by this is the moduli problem whose general member is a torsion sheaf on Z , supported on a single fiber, and when this is a smooth elliptic curve, it should be a line bundle of degree 2 on that elliptic curve. Now one can show that this moduli space is precisely one of the Y ’s, and in fact there is a very explicit construction of the branch locus S of the resulting Y in terms of the two quadrics Q 1 and Q 2 whose intersection gives a given Z . (Briefly, let the first P1 have coordinates (s, t), the second (u, v), in the construction of Z . Now consider the surface S in P1 (a : b) × P1 (s : t) × P1 (u : v) defined by the property that (a : b, s : t, u : v) is in S iff the quadric a Q 1 + bQ 2 , restricted to the P3 over (s : t, u : v), is singular.) For a given Z and the corresponding Y , if we small-resolve the singularities of Y to form Y (which is unfortunately non-Kähler), then there is [25] an equivalence of derived categories D(Z ) ∼ = D(Y , α), where α is the nontrivial element of Br(Y ) and D(Y , α) denotes the twisted derived category of Y . Presumably, the physically-relevant equivalence is between D(Z ) and the (twisted) derived category of a noncommutative
11 At the time that [5] was written, discrete torsion was considered a mysterious degree of freedom, possibly intrinsic to CFT. Since that time discrete torsion has come to be completely understood [32,33] as a purely mathematical consequence of defining orbifolds of theories with B fields, neither mysterious nor intrinsic to CFT.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
629
resolution of Y , though such a noncommutative resolution has not yet been constructed mathematically. There is some additional mathematical structure which is not realized in physics. Each branched double cover of P1 × P1 × P1 can be understood as a genus one fibration in three different ways – basically, pick any one of the three P1 ’s to be the base of a branched double cover of P1 forming an elliptic curve. (Physically, one of the P1 ’s is distinguished, namely the one defined by the p’s, and moreover, a genus one fibration story does not enter the physics here at all.) More information can be found in [31]. 3.3. Relation to P7 [2, 2, 2, 2]. Not only is this example analogous to P7 [2, 2, 2, 2], as both involve complete intersections of quadrics, and at Landau-Ginzburg points describe branched double covers, but in fact in special cases there is a quantitative relationship. Given the complete intersection of two quadrics, we can embed in P1 × P1 × P7 . Specifically, given the eight variables au, av, bu, bv, cs, ct, ds, dt which have charge (0, 0, 1) under (λ, µ, ν). In the embedding above, we take these eight variables to be the homogeneous coordinates on P7 . These variables are not independent, but rather obey the two quadric relations (au)(bv) = (av)(bu), (cs)(dt) = (ct)(ds). Thus, what started as a complete intersection of two quadrics in the total space of P O(−1, 0)2 ⊕ O(0, −1)2 −→ P1 × P1 , is now a complete intersection of four quadrics (the two above, plus the two original quadrics) in P7 . This maps to P7 [2, 2, 2, 2] implicitly shrinks the 16 rational curves that are involved in the four flops between different presentations of the complete intersection of 2 quadrics in the P3 bundle on P1 × P1 , so as a result, there are no flops in P7 [2, 2, 2, 2]. 3.4. Discrete torsion and deformation theory. At this point we would like to make an observation regarding discrete torsion, that is not specific to the particular example we have discussed so far in this section. The last remaining unresolved question concerns deformation theory, namely, how can one explicitly reproduce the results of Vafa and Witten in [5]? In [32,33], the other physically observed characteristics of discrete torsion, such as its original definition in terms of phase factors in orbifolds, and its projectivization of group actions on D-branes, were explicitly derived from the idea that discrete torsion is defined by group actions on B fields. The only thing that could not be explicitly derived were the old results of [5], though for those an outline was given: just as happens for line bundles in orbifolds (and is one way of understanding the McKay correspondence), perhaps the only way to consistently deform an orbifold with discrete torsion, consistent with the orbifold Wilson surfaces, is to add nonzero H flux to exceptional submanifolds, which will play havoc
630
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
with supersymmetry, lifting previously flat directions but sometimes also creating new flat directions. If, on the other hand, we define spaces through their sheaf theory, which is the notion at the heart of the nc spaces we see appearing in e.g. the CFT at the LandauGinzburg point of the GLSM for P7 [2, 2, 2, 2], then we have another way of thinking about this issue. As has been discussed mathematically in [34], then the infinitesimal moduli should be interpreted as a suitable Hochschild cohomology, and for the example in [5] the Hochschild cohomology reproduces precisely the deformation theory seen physically. We will not comment further on this matter, but thought it important enough to warrant attention. 4. Non-Calabi-Yau Examples In this section, we will consider six non-Calabi-Yau GLSMs exhibiting behavior that can be understood in terms of Kuznetsov’s homological projective duality. Our first three examples, involving GLSMs for P2g+1 [2, 2], P7 [2, 2, 2], and P5 [2, 2], have LandauGinzburg points that can be interpreted as branched double covers. We explain, in the discussion of P2g+1 [2, 2], subtleties related to the fact that the Kähler parameter flows, and to behavior of Witten indices. In the second trio of examples, involving P4 [2, 2], P6 [2, 2, 2], and P6 [2, 2, 2, 2], there are additional complications, stemming from the fact that a branched double cover of the form one would naively expect can not exist. We discuss how, instead, to get a geometric interpretation, one must work in a different cutoff limit where the geometry is interpreted as a space with hypersurfaces of Z2 orbifolds instead of as a branched double cover. Curiously, in homological projective duality for complete intersections of quadrics, there is an even/odd distinction (reflected in the examples above) which is analogous to the distinction between duals for G(2, N ) for N even and odd in [3,23]. 4.1. Hyperelliptic curves and P2g+1 [2, 2]. A non-Calabi-Yau example of this phenomenon can be obtained as follows. Consider a gauged linear sigma model describing a complete intersection of two quadrics in P2g+1 . (We shall assume g ≥ 1.) The superpotential in this theory can be written W = p1 Q 1 (φ) + p2 Q 2 (φ) = φi Ai j ( p)φ j , ij
where the Q i are the two quadrics, and Ai j ( p) is a symmetric (2g + 2) × (2g + 2) matrix with entries linear in the pa . For r 0, the φi are mostly massive, away from the degree 2g + 2 locus det A = 0. Away from that locus, the only massless fields are the pa , and as they are nonminimally charged, they describe a gerbe, which physics sees as a double cover. So, for r 0 (and g > 1) we get a positively-curved space, namely the complete intersection of two quadrics in P2g+1 , whereas for r 0 (and g > 1) we get a negatively-curved space, namely a double cover of P1 branched over a degree 2g + 2 locus, i.e. a hyperelliptic12 curve of genus g. 12 For completeness, let us briefly repeat the analysis of Sect. 2.1 here. From that section, K S = 1 , and d = 2g + 2. Thus, π ∗ 2k+d H , where k = −2, the degree of the canonical bundle of P 2 K S = (2g − 2)π ∗ H , which is the canonical bundle of a curve of genus g.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
631
Before commenting further on the mathematics of this situation, let us review the physics of this gauged linear sigma model. For g = 1, both limits correspond to CalabiYau’s – in fact to elliptic curves. For g > 1, the story is more interesting, as neither side is Calabi-Yau. First note that for g > 1, there is an axial anomaly and so the theta angle is meaningless, so the Kähler moduli space is (at best) one real dimensional. Furthermore, the singularity near the origin semiclassically13 divides the Kähler moduli space into two disconnected halves. The fact that the Kähler moduli space splits rescues us from a problem with the Witten index. The Euler characteristic of the genus g curve at r 0 is 2 − 2g. The Euler characteristic of the r 0 complete intersection is different. The top Chern class of the tangent bundle of the complete intersection should be 4 times the coefficient of H 2g−1 in (1 + H )2g+2 (1 + 2H )2 (denominator from the two quadric equations, numerator from the Euler sequence for the tangent bundle to P2g+1 , and the factor of 4 from the fact that a general plane in P2g+1 intersects the complete intersection in 4 (= deg X ) points). One can see immediately that the Euler characteristic of the complete intersection is at least always divisible by four, whereas the Euler characteristic of the genus g curve obeys no such constraint. More, in fact: one can show the Euler characteristic of the complete intersection is not only always divisible by four, but in fact always vanishes. As a result of the Euler characteristic computations above, the only time when the Witten indices of the r 0 and r 0 theories match is when g = 1, the Calabi-Yau case where the Kähler moduli space is one complex dimensional. For g > 1, the Witten indices do not match – but since the Kähler moduli space has two distinct components for g > 1, and there is no way to smoothly move from one component to the other, the fact that the Witten indices do not match is not a concern. As another quick check of the physics, let us discuss how renormalization group flow behaves in these theories for g > 1. The gauged linear sigma model predicts that r will flow towards −∞, which is consistent with both phases. For r 0, we have a positively-curved space, so it will try to shrink under RG flow, consistent with the GLSM computation. For r 0, we have a negatively-curved space, which will try to expand – meaning, |r | should increase or, again, since r 0, r will flow towards −∞. Thus, we see both phases are consistent with the GLSM prediction that r will flow in the direction of −∞. This physics naturally latches onto some corresponding mathematics. It can be shown [36] that the moduli of (smooth) complete intersections of two quadrics in P2g+1 are naturally isomorphic to the moduli of hyperelliptic curves of genus g. The isomorphism can 13 What actually happens after we take quantum corrections into account is more interesting, and described for cases with vanishing classical superpotential in [35]. There, it was argued that the Kähler moduli space does not split apart, but rather extra Coulomb vacua emerge, and those extra Coulomb vacua fix the problem of mismatched Witten indices that we discuss momentarily. It is not completely obvious to the authors how to extend their results to cases with nonvanishing superpotential. If we simply ignore the classical superpotential (and there are arguments that this might be nearly the correct procedure), then the extra Coulomb vacua are the solutions to the quantum cohomology relation σ 2g+1−2(2) = q. That would give 2g − 3 extra Coulomb vacua, which is tantalizingly close to what we need to fix a mismatch between Witten indices of 0 and 2 − 2g that we will see shortly. Since we do not understand how to deal with cases with nonvanishing classical superpotential, which is the case throughout this paper, we will not discuss this issue further. We would like to thank I. Melnikov for a lengthy discussion of this matter.
632
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
be summarized as follows. Given a smooth quadric Q in C2g+2 , there are two families of maximal isotropic (Lagrangian) subspaces of Q. Given a pencil of quadrics (which is what one gets with a complete intersection of two), then the set of maximal isotropic subspaces of the quadrics in the pencil are a double cover of P1 minus the singular quadrics. The set of singular quadrics is the intersection of P1 with the discriminant locus, which has degree 2g + 2, and can be described in the form {det A = 0}, where A is a symmetric (2g + 2) × (2g + 2) matrix linear in the p’s, exactly as we have found physically. Thus the pencil of quadrics naturally gives rise to a hyperelliptic curve, and our physical picture of this GLSM has a natural mathematical understanding. This example can also be naturally understood in terms of homological projective duality [30]. As in the first example we studied in this paper, the homological projective duality of P2g+1 and of a sheaf of even parts of Clifford algebras implies that the derived category of P1 branched in 2g + 2 points embeds fully and faithfully into the derived category of a complete intersection of 2 quadrics. This is written up in [14]. 4.2. P7 [2, 2, 2]. A complete intersection of three quadrics in P7 is an example of a Fano manifold. Repeating the same analysis as before, one quickly finds that the Landau-Ginzburg point of the gauged linear sigma model for P7 [2, 2, 2] is a branched double cover of P2 , branched over a degree 8 locus. Let us check that this is consistent with renormalization group flow in the GLSM. As remarked in the last section, if the large-radius limit is Fano, then the Landau-Ginzburg point in a model with a one-dimensional Kähler moduli space had better be of general type, to be consistent with the behavior of RG flow in the GLSM. Applying the results of Sect. 2.1, we see that the branched double cover obeys K S = π ∗ 2k+d H , where H 2 is a hyperplane class on the base, d = 8, and k = −3 for P2 , so K S = π ∗ (H ), and hence is of general type, exactly as needed for consistency. In fact, just as for P7 [2, 2, 2, 2], there is a slight subtlety in that the branched double cover is singular, at points where the rank of the quadrics drops by two. Furthermore, these singularities do not admit crepant resolutions. However, the gauged linear sigma model is nonsingular at those points. The homological projective dual to P7 [2, 2, 2] is a noncommutative resolution of the branched double cover above. Furthermore, matrix factorizations at the LandauGinzburg point trivially reproduce the sheaf theory that defines the noncommutative resolution. Thus, as in the Calabi-Yau cases studied, we see that the gauged linear sigma model is realizing Kuznetsov’s homological projective duality. 4.3. P5 [2, 2]. Repeating the same analysis as before for the gauged linear sigma model for P5 [2, 2], one finds a branched double cover of P1 , branched over a degree 6 locus. Unlike the case above, this branched double cover is nonsingular generically, so typically no resolution (noncommutative or otherwise) need appear. As before, that branched double cover is precisely Kuznetsov’s homological projective dual to P5 [2, 2]. Let us check that this is consistent with renormalization group flow in the GLSM. As remarked earlier, if the large-radius limit is Fano, as is the case here, then the LandauGinzburg point in a model with a one-dimensional Kähler moduli space had better be of general type, to be consistent with the behavior of RG flow in the GLSM. Applying the results of Sect. 2.1, we see that the branched double cover obeys K S = π ∗ 2k+d 2 H ,
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
633
where H is a hyperplane class on the base, d = 6, and k = −2 for P1 , so K S = π ∗ (H ), and hence is of general type, exactly as needed for consistency. 4.4. Degree 4 del Pezzo (P4 [2, 2]). A complete intersection of two quadrics in P4 is a del Pezzo of degree four [37, p. 49]. This provides another Fano example to study, though in this case the physics will be more complicated. Repeating the previous analysis in this model, the superpotential in this theory can be written in the form W = p1 G 1 (φ) + p2 G 2 (φ) = φi Ai j ( p)φ j , i, j
where G i are the two quadrics, and Ai j ( p) is a symmetric 5 × 5 matrix with entries linear in the pi . As previously, this superpotential defines a mass term for the φi , except over the locus where the rank of A drops. So, away from the locus det A = 0, the only massless fields are the pi , which are of charge 2 and hence define a Z2 gerbe on P1 , which physics sees as a double cover. In other words, the resulting theory in the r 0 limit looks like a double cover of P1 branched over the locus det A = 0, which is degree five. However, there is no space fitting that description, because to be a branched double cover (here, a covering curve), the branch locus must be even. (There is a simple way to see this. Take a dimension 1 slice through the base. If a double cover exists, then we should be able to think of the intersection of each part of the branch locus with the slice as a source of a Z2 branch cut. For all example studied previously, the branch locus was even degree, and so this is consistent; but in the odd degree case, then one of the Z2 branch cuts has nowhere to end, and so the branched double cover cannot exist.) The correct interpretation of this model is slightly more subtle. It is a P1 with a Z2 orbifold structure in five points, which can be thought of locally as the quotient stack of a branched two-fold covering modulo the natural Z2 action on the cover. To see this, we have to think a bit more carefully about the appearance of branched double covers in theories with mostly-massive fields. The Z2 gerbe interpretation only holds when minimally-charged fields have been integrated out: so, we only have gerbes outside a disk surrounding the locus where minimally-charged fields become massless. The size of that disk is determined by the cutoff (or other regularization) of the theory. Outside that disk, we can talk about gerbes and/or double covers, but inside that disk, there is only a single cover. In the examples studied so far, we have been able to consistently work in a coupling limit where those disks have shrunk to zero size. We are running into problems in this example because there cannot be such a coupling limit here – there cannot be a global branched double cover interpretation. However, there is an alternate limit we can take instead. Instead of shrinking the disks to zero size, we can expand them to cover the entire space. Now, instead of getting a branched double cover, we get a single cover, with something unusual happening at the massless locus. We conjecture that at the massless locus, we have a Z2 orbifold structure. First, branched double covers project onto spaces with hypersurfaces of Z2 orbifolds along the branch locus – simply imagine orbifolding the double cover by a Z2 that exchanges the sheets. (Alternatively, think of the Z2 orbifolds as providing branch cuts. Although for the branched double cover to make sense requires that the branch locus have even
634
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
degree, it is consistent for a space to have an odd degree hypersurface of Z2 orbifolds.) In effect, we are conjecturing that changing the size of the disk is moving us from the branched double cover to its underlying space, which is extremely natural. Only in the limit that the disks have zero size or cover the space can we recover an honest geometry; varying the regularization should interpolate between the two. Notice that in order for this to be consistent, all of our examples so far should have an alternate interpretation in which we replace the branched double cover with a space with Z2 orbifolds – the description as a space with Z2 orbifolds should be universal, and the description as a branched double cover should only apply when such a double cover can exist. This also, we believe to be true. Mathematically, the branched double covers described previously have isomorphic sheaf theory to spaces with even degree hypersurfaces of Z2 orbifolds – the two descriptions are more or less isomorphic from the perspective of homological projective duality. We should be able to flip between them simply by varying the regularization of the theory. In the Calabi-Yau case, where one gets a nontrivial CFT, one should be able to vary a cutoff without fundamentally changing the theory, and so the result should follow from that. In the non-Calabi-Yau cases discussed previously, the matter is more delicate. We still have not directly addressed the reason why there should be a hypersurface of Z2 orbifolds. In principle, one ought to be able to deduce this from the Berry phase, but we have not yet worked out a consistent generalization of the computation described earlier. Part of the computation lies in the fact that we are no longer describing nontrivial CFT’s in the IR limit, so the state-operator correspondence is complicated. Another part of the complication lies in the Berry phase itself, which should now get a contribution from the curvature of the space – so for example, on a P1 with a single Z2 orbifold singularity at a point, there should be a nontrivial phase for a loop that closely wraps the point, but as one expands the loop around the space, there should be an additional curvature-dependent contribution so that ultimately, for a loop closely wrapping another ordinary point far from the Z2 , the total phase vanishes. Instead of appealing to the Berry phase, we shall appeal to homological projective duality. The matrix factorization argument we discussed in the context of P7 [2, 2, 2, 2] naively applies here (though a B twist is no longer well-defined, we can still speak about certain D-branes in the untwisted theory), and so can deduce what sheaf theory the dual theory possesses, from which we get that the defining sheaf theory is the same as that for homological projective duality. Put another way, homological projective duality tautologically matches the open string sector, as in previous examples, from which we conclude that it had better also match the closed string sector. In the present case, the homological projective dual14 [30] of P4 [2, 2] is a P1 with 5 Z2 orbifold points, so we conclude that the Landau-Ginzburg point can be described in this fashion. We can no longer claim to give a completely independent check of homological projective duality in this example, as we cannot give a completely physically rigorous derivation of the origin of the Z2 orbifold points, but at least we can see how this is consistent. We have discussed previously how the GLSM Kähler moduli space semiclassically falls apart into disjoint pieces in non-Calabi-Yau cases, and how in one-dimensional examples, if one limit is Fano, the other must be of general type, i.e. have opposite 14 Technically, the HP-dual of P4 in the double Veronese embedding is P14 with the sheaf of even parts of Clifford algebras. Therefore, the derived category of a complete intersection of 2 quadrics in P4 contains the derived category of P1 with a Z2 -orbifold structure in 5 points.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
635
curvature in order for the running of the Fayet-Iliopoulos parameter in the GLSM to be consistent. The same analysis applies here, and the P1 with 5 Z2 orbifolds is of general type, so the Fayet-Iliopoulos RG analysis discussed previously is consistent. Intuitively, the Z2 orbifolds have enough negative curvature to counterbalance the positive curvature of the rest of the projective space. Let us examine this issue in more detail, for later use. (The analysis will be closely analogous to that for double covers described in Sect. 2.1.) Given an n-dimensional projective space and D ⊂ Pn a smooth hypersurface of degree d, then consider a stack describing Pn with a Z2 orbifold along D, which shall be denoted Z ≡ Pn D/2, together ⊂Z with its natural map π : Z → Pn . On this stack there is a unique smooth divisor D ∗ n such that π OP (−2D) = O Z (− D). Now, a straightforward computation shows that15 i.e. K Z = π ∗ O(d/2 − n − 1) or c1 (T Z ) = π ∗ O(−d/2 + n + 1). K Z = π ∗ K Pn ( D), As a consistency check, recall that elliptic curves can be described as branched double covers of P1 , which project to P1 with 4 Z2 orbifold points: d = 4, n = 1, and so in this case, K Z = O, as one would expect for a Calabi-Yau. Similarly, some K3’s can be expressed as branched double covers of P2 , branched along sextics, which project to P2 with a Z2 orbifold structure along a hypersurface of degree 6: d = 6, n = 2, and so again we find K Z = O, as one would expect for a Calabi-Yau. If d/2 < n + 1, then the projective space with a hypersurface of Z2 orbifolds is Fano; if d/2 > n + 1, then it is of general type. In the present case, a P1 with 5 Z2 orbifold points, d = 5 and n = 1, so d/2 > n + 1, and so this stack is of general type, as claimed earlier. This phenomenon has appeared previously in the physics literature in discussions of nonsupersymmetric orbifolds, see for example [39–41]. There, it was discovered in (nonsupersymmetric) complex codimension one Z2 orbifolds that relevant operators in twisted sectors generically get vevs under RG flow – if we think of those Z2 orbifolds as having negative curvature, then under RG flow they should ‘expand’ (meaning, twisted sector operators get vevs). 4.5. P6 [2, 2, 2]. Following the same analysis as in the last section, the LandauGinzburg point in this model should be a P2 with a degree 7 hypersurface of Z2 orbifolds. As before, since the hypersurface is of odd degree, there cannot be a branched double cover, but there can be a space with an odd degree hypersurface of Z2 orbifolds, so we propose that that is the geometry that emerges in a relevant regularization limit, replacing the massless locus with Z2 orbifolds. As before, the homological projective dual to P6 [2, 2, 2] is exactly this, a P2 with a degree 7 hypersurface of Z2 orbifolds [42]. Let us also check that this relation is consistent with renormalization group flow in the GLSM. At large radius, P6 [2, 2, 2] is Fano, so the opposite end of the one-dimensional GLSM Kähler moduli space had better be of general type, in order to be consistent with the one-loop renormalization of the Fayet-Iliopoulos parameter in the GLSM. 15 Note that d need not be even: although O(−d/2 + n + 1) does not make sense on the projective space, the line bundle O(−d + 2n + 2) pulls back to a line bundle on Z which has a canonical square root. We can then think of this square root as the pullback of the fictional line bundle O(−d/2 + n + 1), and this agrees with the pullback map on Q-divisors. The fictional bundle O(−d/2 + n + 1) on the projective space also has an interpretation as a parabolic line bundle whose parabolic structure is specified along the divisor where the orbifold structure lies; see for example [38] and references therein for more information on parabolic vector bundles.
636
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
Indeed, using the results of the previous section, for a P2 with a degree 7 hypersurface of Z2 orbifolds, we have d = 7, n = 2, and so d/2 > n + 1, hence the P2 with hypersurface of Z2 orbifolds is of general type, exactly as needed to be consistent with physics. 4.6. P6 [2, 2, 2, 2]. Following the same analysis as previously, the Landau-Ginzburg point in this model should be a P3 with a degree 7 hypersurface of Z2 orbifolds. As before, since the hypersurface is of odd degree, there cannot be a branched double cover, but there can exist globally a space with a hypersurface of Z2 orbifolds. Also as before, this is precisely the result of applying homological projective duality [42]. In addition, the surface has several ordinary double points, which are resolved via a noncommutative resolution, of the form discussed previously. Let us also check that this relation is consistent with renormalization group flow in the GLSM. At large radius, P6 [2, 2, 2, 2] is of general type, so the opposite end of the one-dimensional GLSM Kähler moduli space had better be Fano, in order to be consistent with the one-loop renormalization of the Fayet-Iliopoulos parameter in the GLSM. Indeed, using our earlier results, for a P3 with a degree 7 hypersurface of Z2 orbifolds, we have d = 7, n = 3, and so d/2 < n + 1, hence the P3 with hypersurface of Z2 orbifolds is Fano, exactly as needed to be consistent with physics. 5. More General Complete Intersections So far in this paper we have discussed exclusively gauged linear sigma models for complete intersections of quadrics and the role that homological projective duality plays in such models. Given that plus the previous work [6], the reader might incorrectly conclude that homological projective duality can only be applied to GLSM’s with relatively exotic physics. To counter such a perception, in this section we will describe a few examples of GLSM’s for complete intersections of higher-degree hypersurfaces, and how their Kähler phase structure can also be (correctly) understood with homological projective duality. To a limited extent, this was also described in the context of nonabelian GLSM’s in [6], where homological projective duality was used to predict that the LandauGinzburg point of G(2, 6)[16 ] should be a Landau-Ginzburg model corresponding to a K3 surface. However, this was a prediction, not an observation, and moreover, LandauGinzburg models for K3’s have the same CFT as nonlinear sigma models on K3’s, so, the examples in this section should clarify that homological projective duality really can generate true Landau-Ginzburg theories. 5.1. P4 [3]. In this section we will study another example of a GLSM for a higher-degree hypersurface. The Landau-Ginzburg point of this gauged linear sigma model is well-known to be a theory defined by the Landau-Ginzburg model on [C5 /Z3 ] with a cubic superpotential. According to [23] we can recover the same result from homological projective duality. We begin with the embedding G(2, 6) → P14 = PAlt2 C6 and its homological projective dual Pf → PAlt2 C6∗ , where Pf is the Pfaffian cubic hypersurface. Now, consider the action induced on linear sections. Let L be a space of linear forms of dimension 5, so that PL ⊂ PAlt2 C6∗ has dimension 4. Intersecting both sides, we find that G(2, 6)[15 ] is dual to a cubic threefold, i.e. P4 [3].
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
637
Finally, we must consider the essential pieces in each derived category. The nc spaces defined by those essential pieces are what should be appearing at the r → ±∞ limits of the GLSM Kähler moduli space. Kuznetsov argues that there is a semi-orthogonal decomposition D b (G(2, 6)[15 ]) = O, S ∗ , (where S is the universal subbundle on the Grassmannian, restricted to the complete intersection) and D b (P4 [3]) = O, O(1), , where and are the ‘essential pieces’ of the respective categories. Kuznetsov also checks that = . Homological projective duality does not itself tell us how to interpret the nc spaces defined by the essential pieces , above. Instead, to interpret the essential piece of the derived category of P4 [3], we turn to work of Orlov [43–45], which related matrix factorizations in Landau-Ginzburg models on [C5 /Z3 ] to sheaves on P4 [3]. Specifically, he argued that the category of matrix factorizations on the Landau-Ginzburg model, which we shall denote D b ([C5 /Z3 ], W ) embeds into D b (P4 [3]), as the right orthogonal to the objects O, O(1). In other words, D b ([C5/Z3 ], W ) = = . Thus, the nc space that is homologically projective dual to P4 [3], is the Landau-Ginzburg model on [C5/Z3 ] with cubic superpotential, exactly as happens in the GLSM. Thus, the nc space which is homologically projective dual to P4 [3], is exactly the Landau-Ginzburg model appearing at the Landau-Ginzburg point in the GLSM Kähler moduli space, giving us another check that homological projective duality seems to be correctly encoding the phases of GLSM’s, and not just in physically exotic cases. In this particular example, homological projective duality acts somewhat trivially: it automatically preserves nc spaces (corresponding to the physics that the B model is independent of Kähler moduli), but typically exchanges different presentations of a fixed nc space. In the present case, homological projective duality did not itself produce a different presentation; instead, we had to invoke results of Orlov to interpret the nc space. What is important about this example is the following: 1. First, although Orlov’s work was used more crucially than homological projective duality, homological projective duality did not produce a contradiction. This is more or less guaranteed by the structure of homological projective duality – nc spaces are preserved, and as nc spaces are defined by the open string sector of the B model, which is Kähler-moduli-independent, this is a must. 2. Second, this means that these ideas encompass not only geometries appearing at limit points in GLSM Kähler moduli spaces, but also theories that one usually thinks of as being non-geometric. In the past, many researchers have spoken of the geometric phases being related by birational transformation; now we see that the replacement for ‘birational’ encompasses more than just geometric limit points. 5.2. P5 [3, 3]. In this section we will study another example of a GLSM for a higherdegree hypersurface. Again, we will see that homological projective duality encompasses not only limits in the GLSM Kähler moduli space with geometric interpretations, but non-geometric points as well.
638
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
5.2.1. Basic analysis. The Landau-Ginzburg point of this model is what is sometimes called a ‘hybrid Landau-Ginzburg model’ or ‘fibered Landau-Ginzburg model.’ It describes what might be very roughly called a family of Landau-Ginzburg models (over the vector space C6 ) fibered in some fashion over P1 . Let us first make this description more precise. A Landau-Ginzburg model is defined by a holomorphic function (the superpotential) over some noncompact space or stack. In the present case, the noncompact space is defined by a GIT quotient of six fields φi and two fields pa by a C× with weights φi −1
pa . 3
If the weights were reversed, so that the φi had weight −3 and the pa had weight 1, then this space would be the total space of a rank six vector bundle over P1 , namely O(−3)⊕6 → P1 . However, something more unusual is going on instead. A quotient C N //C× , where all the fields have weight k rather than weight 1 is seen physically not precisely as a copy of P N −1 , but rather as a Zk gerbe over P1 [2,18–21]. The quotient above is therefore not a vector bundle over P1 , but rather is a vector bundle over a Z3 gerbe on P1 . If the φi all had the same weight as the homogeneous coordinates on the underlying stack, then we would naturally think of each φ as a coordinate on the total space of the line bundle O(−1), but instead the φi ’s all have charge equal to a third of that of the underlying stack. Such a line bundle is conventionally denoted “O(−1/3),” a notation that only makes sense for gerbes and stacks, not spaces. (Bundles on gerbes will be discussed in greater detail in the upcoming work [46].) Intuitively, this description of the bundle encodes the fact that we are fibering a Z3 orbifold over a base space – the Z3 acts trivially on the base P1 (and so, we have a gerbe structure), but nontrivially on the fibers. To give a little more perspective on this language, the total space of the line bundle O(1/k) over the Zk gerbe on a point (denoted BZk ) is the same thing as the orbifold [C/Zk ]. In any event, we see that this Landau-Ginzburg model is defined by a superpotential over the noncompact stack given by the total space of the bundle O(−1/3)⊕6 → G 3 P1 , where G 3 P1 denotes the gerbe on P1 . The fibers of this hybrid Landau-Ginzburg model look like Landau-Ginzburg models with a cubic superpotential in six variables. At least at the Fermat point, such a superpotential splits into two copies of a cubic superpotential in three variables, or put another way, (T 2 × T 2 )/Z3 . This suggests that we should think of the theory at the LandauGinzburg point as some sort of K3 fibered over a P1 base, which we will indirectly check in the next section by studying monodromy computations. In the section after that, we shall see that homological projective duality will make the same prediction. 5.2.2. Monodromy computation. In this section we shall check the intuition developed in the last section, that the Landau-Ginzburg point of the GLSM for P5 [3, 3] should have the structure of a fibration over P1 , via a monodromy computation. We shall follow the conventions of Sect. 2.4. This is the case: p = 9, q = 27. The total Chern class is c(X ) = 1 + 54η − 144ρ, which yields h 2,1 = 71.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
639
The intersection form, ⎡
⎤ 0 −6 −1 −1 1 0 ⎥ ⎢6 0 ω=⎣ . 1 −1 0 0 ⎦ 1 0 0 0
(2)
The large-radius monodromy is ⎡
M∞
1 ⎢1 : v → V ⊗ H = ⎣ 0 0
0 1 9 0
0 0 1 1
⎤ 0 0⎥ . 0⎦ 1
(3)
The conifold monodromy is ⎡
⎤ 1 −6 −1 −1 0 0 ⎥ ⎢0 1 M1 : v → V − (v, O)v = ⎣ . 0 0 1 0 ⎦ 0 0 0 1
(4)
The “hybrid” monodromy is ⎡
M0−1 = (M∞ M1 )−1
−5 6 0 0 ⎢ −1 1 =⎣ 9 −9 1 −9 9 −1
⎤ 1 0⎥ . 0⎦ 1
(5)
This does not look very illuminating. But note that the standard 2-dimensional irreducible representation of Z3 over the integers (this representation is, of course, reducible over C) is generated by −1 −1 . γ = 1 0 In fact, M0−1 = AM A−1 , where ⎡
⎤ −1 −1 0 0 0 0 0 ⎥ ⎢ 1 M = ⎣ , 0 1 −1 −1 ⎦ 0 0 1 0
(6)
where A ∈ S L(4, Z). But for the extra “1” in the lower left-hand 2 × 2 block, this would simply be two copies of the two-dimensional irrep of Z3 , and we would have M 3 = 1. Instead, it is unipotent, (M 3 − 1)2 = 0. If the theory at the Landau-Ginzburg point were an ordinary Landau-Ginzburg orbifold (here presumably a Z3 orbifold due to the fact that we are intersecting cubics), then because of the Z3 quantum symmetry, we would expect the monodromy to obey M 3 = 1. Instead, we have something different. The typical interpretation of a monodromy of the form above, (M 3 − 1)2 = 0, is that we have a fibered Landau-Ginzburg
640
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
orbifold, fibered over a P1 base. Heuristically, going around three times is equivalent to shifting the B field on the P1 . Thus, we have confirmation from the monodromy computation that physically, the Landau-Ginzburg point of the GLSM for P5 [3, 3] should look like fibered LandauGinzburg models, fibered over a P1 base, which is what we argued in the last section.. In the next section, we will see that homological projective duality predicts mathematically that the Landau-Ginzburg point should be interpreted as a noncommutative K3 fibration, fibered over a P1 base. 5.2.3. Homological projective duality and fibered noncommutative K3s. Mathematically, the homological projective dual of the complete intersection P5 [3, 3] is [30] a family of noncommutative K3s, fibered over P1 , the same structure that we have seen physically. Let us review the mathematical argument below. (1) First, one can use the Grassmannian-Pfaffian homological projective duality to argue that if D ⊂ P5 is a four-dimensional smooth Pfaffian cubic hypersurface, then the dual linear section of the Grassmannian is a K3 surface S of degree 14. In this case, D b (S) embeds as a full subcategory in D b (D) and D b (D) has an induced semi-orthogonal decomposition in which D b (S) is left-orthogonal to O(−3), O(−2), O(−1). (2) Taking the observation above as a model, take any smooth four-dimensional cubic hypersurface D ⊂ P5 (not necessarily Pfaffian) and look at the left orthogonal A D ≡ ⊥ O(−3), O(−2), O(−1). Now, we can interpret the category A D to be a category of B-branes on some c = 6 SCFT, which from a mathematical point of view defines (the open string sector of) a nonlinear sigma model on a ‘noncommutative16 K3.’ The first justification is (1) where we see that A D specializes to a Pfaffian cubic. The second is that A D is a smooth proper Calabi-Yau category17 of dimension two. The third is that the Hochschild homology of A D has the same exact size as the de Rham cohomology of a K3. (3) Now let us consider families. If you have X which is the complete intersection of two cubics in P5 , then there is a pencil of cubics passing through X , i.e. we have a linear family {Dt∈P1 } of cubic fourfolds, so that X = ∩t∈P1 Dt . The object corresponding to X through taking linear sections in noncommutative HPD will be a noncommutative Calabi-Yau three-fold Y which is naturally fibered over the P1 that parametrizes the pencil. We can see this by constructing Y , as follows. Start with the universal family U = {(x, t) ∈ P(V ) × P1 |x ∈ Dt } of fourfolds in the pencil. Let p : U → P1 be the natural projection. Consider the full subcategory A X ⊂ D b (U ), defined as the left orthogonal ⊥ p ∗ D b (P1 ) ⊗ O(−3), p ∗ D b (P1 ) ⊗ O(−2), p ∗ D b (P1 ) ⊗ O(−1) . 16 Recall from our earlier discussion that the usage of ‘noncommutative’ in this context does not refer to any sort of noncommutative algebra or ring. Rather, ‘noncommutative space’ is the generic term for the base space when interpreting a category as a category of A or B model branes. The base space might be a Calabi-Yau, a Calabi-Yau with a flat B field, a nongeometric Landau-Ginzburg model, or something else. 17 See Appendix A for a review of Calabi-Yau categories.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
641
Again, it is straightforward to check that if X is smooth then A X is a smooth compact three-dimensional Calabi-Yau category. We will think of A X as the derived category of a noncommutative K3 Y which is fibered over P1 by noncommutative K3 surfaces (defined as in (2)). If we specialize X to be the complete intersection of two Pfaffian cubics, then Y will be an ordinary Calabi-Yau which is fibered by K3 surfaces of degree 14. In fact, in his paper on the Grassmannian-Pfaffian HPD [23] Kuznetsov shows that in case Y is commutative we have an equivalence of categories D b (X ) ∼ = D b (Y ). Thus, homological projective duality can be applied to more examples than just complete intersections of quadrics (or hyperplanes in Grassmannians); it makes predictions for other cases, as we have seen in this example, that correctly match physics. Unfortunately, it is very difficult to check whether analogous statements are true for GLSM’s describing hypersurfaces of degree greater than three in projective spaces. We conjecture that homological projective duality applies in general, but will have to leave a complete verification to future work.
6. Conclusions In this paper, we have done a number of different things. By studying the properties of some basic, ordinary-seeming gauged linear sigma models such as that for the complete intersection P7 [2, 2, 2, 2], we have discovered in many examples that the Landau-Ginzburg point of the GLSM realizes geometry in a novel fashion, not as a complete intersection; that the same geometry is not birational to the large-radius geometry, providing a GLSM interpolating between two non-birational geometries. We have provided a mathematical interpretation (homological projective duality) to substitute for ‘birational’ in relating GLSM phases, an interpretation which often even covers nongeometric Landau-Ginzburg models. We have discovered that many of those LandauGinzburg points do not flow under the renormalization group to nonlinear sigma models, or any other known type of conformal field theory, but rather give a physical realization of certain ‘noncommutative resolutions’ of spaces, and so give us a way of defining conformal field theories for nc spaces (in Kontsevich’s sense), a whole new kind of CFT and string compactification. In the analysis of the Landau-Ginzburg points of these otherwise ordinary-seeming GLSM’s, we have also found that understanding stacks and how stacks appear in physics plays a crucial role, as the analysis of the Landau-Ginzburg points relies crucially on facts concerning the behavior of strings propagating on gerbes. Thus, stacks play a much more important role in the physical understanding of many GLSM’s than previously realized. One of the open questions in this paper is, to what extent can all phases of gauged linear sigma models be understood via homological projective duality? To answer this question, the mathematics of homological projective duality needs to be much better understood. For example, it is known that flops of small resolutions of the most basic conifold singularity in three complex dimensions are examples of homological projective duality, but whether more general flops are also related by homological projective duality, is not yet known. In the same vein, one can ask, can the McKay correspondence or any of its variants be understood as forms of homological projective duality, especially since various forms of McKay can often be realized physically as different phases of gauged linear sigma models.
642
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
Another set of open problems arises from the fact that we have discovered more conformal field theories, not of a form previously discussed – in fact, we have even found a description of them with gauged linear sigma models. The ‘noncommutative resolutions’ arising at Landau-Ginzburg points are, close to geometric singularities, not even locally spaces. These noncommutative resolutions do not obviously have any description as any sort of conformal field theory discussed elsewhere in the literature. As such, the properties of these new conformal field theories needs to be studied. We have briefly described some properties of the B-branes in these conformal field theories, but nothing else. For example, it would be very interesting to compute Gromov-Witten invariants of these theories. (Work on other Gromov-Witten variations that might be applicable here will be discussed in [47].) Acknowledgements. We would like to thank R. Donagi, M. Gross, A. Kuznetsov, and I. Melnikov for useful conversations. In particular, we would especially like to thank A. Kuznetsov for many patient explanations, valuable discussions, and for sharing his insights. Parts of this work were done during the Aspen Center for Physics program “String Theory and Quantum Geometry,” July 2007. J.D. and T.P. would like to thank the ACP for providing excellent working conditions and a stimulating research environment. A.C. was partially supported by NSF grant DMS-0556042. J.D. was partially supported by NSF grant PHY-0455649. S.H. was partially supported by DOE grant DE-FG02-90ER40542. T.P. was partially supported by NSF Focused Research Grant DMS-0139799, NSF Research Training Group Grant DMS-0636606, and NSF grant DMS0700446. E.S. was partially supported by NSF grant DMS-0705381. S. H. is the D. E. Shaw & Co., L. P. Member at the Institute for Advanced Study.
A. Calabi-Yau Categories and Noncommutative Spaces A.1. Definitions. The notion of a smooth and proper Calabi-Yau category was introduced by Kontsevich in 1998 [48] and was subsequently discussed in detail in [49–52]. It abstracts the notion of a smooth and compact Calabi-Yau background that can be used as a target in (A or B twisted) non-linear sigma models. The basic idea is to replace the Calabi-Yau space by the category of A or B branes. Consider a small pre-triangulated18 Karoubi closed19 dg-category C. We think of the objects of this category as complexes of sheaves on a ‘noncommutative space’ or ‘nc space’ X , where X is defined by the category C. If we were trying to model an actual complex manifold X , then we would take C to be the (dg-enhancement of) the derived category of coherent sheaves on X . For that matter, X might turn out to be a Calabi-Yau with a flat B field, or even a Landau-Ginzburg model, by virtue of matrix factorizations – X need not be an actual space. ‘Noncommutative’ is a misnomer in this context, as there need not be a noncommutative algebraic structure associated with X . An nc space X is called algebraic if its sheaf theory is computed by an algebra, i.e. if we can find a dg algebra A = (A· , d) over C, such that the category C X is equivalent to the dg-category of (perfect) dg-modules over A. An nc space X is called smooth if it is algebraic and if the dg algebra A that computes its sheaf theory is perfect when viewed as a bimodule over itself. An nc space X is called compact if it is algebraic and the differential on the dg algebra A has finitely many, finite dimensional cohomology groups. That is, dimC H · (A· , d) < ∞. It is known that: 18 The derived category is then the homotopy category of this pre-triangulated category. 19 Karoubi closed means that every projector splits. Recall that a projector on an object E in a category
is an endomorphism π : E → E such that π ◦ π = π . We say a projector π splits if there is a direct sum decomposition E = E 1 ⊕ E 2 such that π is the projection map E → E 1 . Then, a category is called Karoubi closed if any projector on any object splits.
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
643
• The dg category of complexes of injective sheaves with coherent cohomology on any complex algebraic variety is computed by a dg algebra. Thus, every complex variety is algebraic when viewed as an nc space. Also, non-algebraic complex manifolds may not be algebraic nc spaces. For example, a generic non-algebraic K3 is not an algebraic nc space. One can define sheaves on non-algebraic K3’s, so there is a category, and hence an nc space; however, by a theorem of Bondal and van den Bergh [53], the category of sheaves on a generic non-algebraic K3 is not saturated, which means there is no smooth and proper algebra that computes this category. Hence non-algebraic K3’s are nc spaces, but not algebraic nc spaces. • A complex variety X is proper or smooth if and only if it is proper or smooth when we view it as a nc space. There are many other examples of proper and smooth nc spaces coming from stacks, deformation quantizations, symplectic manifolds, generalized complex manifolds, and Landau-Ginzburg models. If X is a proper and smooth nc space, then one can show that the category C X defining X has a Serre functor S : C X → C X . We say that a proper smooth nc space X is an nc Calabi-Yau, if the Serre functor is a power of the shift functor, i.e., if S = [d] for some positive integer d. (The category defining an nc Calabi-Yau is called a Calabi-Yau category.) The number d is the dimension of the nc Calabi-Yau X . Notice that one can define nc Calabi-Yau spaces of fractional dimension p/q by requiring that S q = [ p]. One example is furnished by Enriques surfaces, where the square of the Serre functor is [2], and hence the (fractional) dimension is 2/2 = 1. (In particular, note the notion of dimension can be slightly misleading, as it leads to an Enriques surface having dimension 1 as an nc space but dimension 2 as an ordinary complex manifold.) There are many more examples of such spaces and one should find their meaning eventually. A.2. Deformations. If we start with an ordinary Calabi-Yau space X , we can look at all deformations of X viewed as an nc space. In other words, we can forget about the actual X but deform the category C X as a dg category. The tangent space to such deformations can be identified with the second Hochschild cohomology of X . In the smooth proper case this tangent space has a decomposition H H 2 (X ) = H 2 (X, O X ) ⊕ H 1 (X, TX ) ⊕ H 0 (X, 2 TX ), where each piece can be interpreted geometrically: • H 2 (X, O X ) are the deformations of X as an O× gerbe; • H 1 (X, TX ) are the usual deformations of X as a variety; • H 0 (X, 2 TX ) are the deformation quantizations of the product structure on the structure sheaf of X . In the example in Sect. 5.2.3 we start with S which is a K3 surface of degree 14, and corresponds to a Pfaffian cubic fourfold D ⊂ P5 under the Grassmannian-Pfaffian duality. Next we look at the nc deformations of S that correspond to deforming the fourfold D to a non-Pfaffian cubic. In particular, if ξ is a deformation direction for D, then ξ can be interpreted as an element in H H 2 (S) and hence breaks into three components ξ = ξ20 + ξ11 + ξ02 . If ξ moves D to a non-Pfaffian cubic, then we can not have ξ = ξ11 . In the example in Sect. 5.2.3, an old geometric argument with varieties of lines shows that the gerby part ξ20 of the deformation direction ξ is non-zero, and it is natural to expect that ξ02 is also non-zero and so this nc deformation of S is of ‘mixed’ nature.
644
A. C˘ald˘araru, J. Distler, S. Hellerman, T. Pantev, E. Sharpe
A.3. Cohomology of nc spaces. Suppose that X is an algebraic nc space and let A be a dg algebra that computes the sheaf theory of X . (Recall that a nc space X is defined by a category C, but for X to be an algebraic nc space means that the category C is the category of perfect dg-modules over some dg algebra; A is that dg algebra.) Then we can attach two natural cohomology theories to X : · (X ) of X . In terms of A we have H · • The Dolbeault cohomology H Dol Dol ≡ H H· (A), · i.e. H Dol (X ) is the Hochschild homology of the dg algebra A. • The de Rham cohomology Hd· R (X ) of X . In terms of the dg algebra A we have Hd· R (X ) ≡ H P· (A), i.e. Hd· R (X ) is the periodic cyclic homology of the dg algebra A.
These definitions are justified by the observation that when X is a smooth variety we have q
• (Kostant-Hochschild-Rosenberg) H Hk (A) = ⊕ p−q=k H p (X, X ). even/odd • (Weibel) H Peven/odd (A) = Hd R (X ). · (S) If S is a nc K3 surface defined as above from a general cubic fourfold D, then H Dol · is a subspace in the ordinary Dolbeault cohomology H Dol (D) which is the orthogonal · (D). complement to {ch(O(−i))}i=1,2,3 with respect to the Mukai pairing on H Dol
References 1. Witten, E.: Phases of N = 2 theories in two dimensions. Nucl Phys. B 403, 159–222 (1993) 2. Hellerman, S., Henriques, A., Pantev, T., Sharpe, E., Ando, M.: Cluster decomposition, T-duality, and gerby CFT’s. Adv. Theo. Math. Phys. 11, 751–818 (2007) 3. Hori, K., Tong, D.: Aspects of non-Abelian gauge dynamics in two-dimensional N = (2, 2) theories. JHEP 0705, 079 (2007) 4. Sharpe, E.: String orbifolds and quotient stacks. Nucl. Phys. B627, 445–505 (2002) 5. Vafa, C., Witten, E.: On orbifolds with discrete torsion. J. Geom. Phys. 15, 189–214 (1995) 6. Donagi, R., Sharpe, E.: GLSMs for partial flag manifolds. J. Geom. Phys. 58, 1662–1692 (2008) 7. Reid, M.: The complete intersection of two or more quadrics. Ph.D. thesis, Trinity College, Cambridge, 1972, available at http://www.warwick.ac.uk/~masda/3folds/qu.pdf 8. Griffiths, P., Harris, J.: Principles of Algebraic Geometry. New York: John Wiley & Sons, 1978 9. Clemens, H.: Double solids. Adv. in Math. 47, 107–230 (1983) 10. Cynk, S., Meyer, C.: Geometry and arithmetic of certain double octic Calabi-Yau manifolds. http://arxiv. org/abs/math/0304121v1[math.AG], 2003 11. C˘ald˘araru, A., Katz, S., Sharpe, E.: D-branes, B fields, and Ext groups. Adv. Theor. Math. Phys. 7, 381–404 (2004) 12. Gross, M., Pavanelli, S.: A Calabi-Yau threefold with Brauer group (Z8 )2 . http://arxiv.org/abs/math/ 0512182v1[math.AG], 2005 13. Gross, M.: Private communication, September 27, 2006 14. Kuznetsov, A.: Derived categories of quadric fibrations and intersections of quadrics. http://arxiv.org/ abs/math/0510670v1[math.AG], 2005 15. Sharpe, E.: D-branes, derived categories, and Grothendieck groups. Nucl. Phys. B561, 433–450 (1999) 16. Douglas, M.: D-branes, categories, and N = 1 supersymmetry. J. Math. Phys. 42, 2818–2843 (2001) 17. Sharpe, E.: Lectures on D-branes and sheaves. Writeup of lectures given at the Twelfth Oporto meeting on “Geometry, Topology, and Physics,” and at the Adelaide Workshop “Strings and Mathematics 2003,” http://arxiv.org/abs/hep-th/0307245v2, 2003 18. Pantev, T., Sharpe, E.: Notes on gauging noneffective group actions. http://arxiv.org/abs/hep-th/ 0502027v2, 2005 19. Pantev, T., Sharpe, E.: String compactifications on Calabi-Yau stacks. Nucl. Phys. B733, 233–296 (2006) 20. Pantev, T., Sharpe, E.: GLSM’s for gerbes (and other toric stacks). Adv. Theor. Math. Phys. 10, 77–121 (2006) 21. Sharpe, E.: Derived categories and stacks in physics. http://arxiv.org/abs/hep-th/0608056v2, 2006 22. Kuznetsov, A.: Homological projective duality. http://arxiv.org/abs/math/0507292v1[math.AG], 2005
Non-Birational Twisted Derived Equivalences in Abelian GLSMs
645
23. Kuznetsov, A.: Homological projective duality for Grassmannians of lines. http://arxiv.org/abs/math/ 0610957v1[math.AG], 2006 24. Harris, J.: Algebraic geometry: a first course. Grad. Texts in Math. 133, New York: Springer-Verlag, 1992 25. C˘ald˘araru, A.: Derived categories of twisted sheaves on elliptic threefolds. http://arxiv.org/abs/math/ 0012083v3[math.AG], 2001 26. C˘ald˘araru, A.: N. Addington: Work in progress 27. Kapustin, A., Li, Y.: D-branes in Landau-Ginzburg models and algebraic geometry. JHEP 0312, 005 (2003) 28. Kuznetsov, A.: Private communication 29. Mukai, S.: Moduli of vector bundles on K3 surfaces, and symplectic manifolds. Sugaku Expositions 1, 139–174 (1988) 30. Kuznetsov, A.: Private communication, January 29, 2007 31. C˘ald˘araru, A.: To appear 32. Sharpe, E.: Discrete torsion. Phys. Rev. D68, 126003 (2003) 33. Sharpe, E.: Recent developments in discrete torsion. Phys. Lett. B498, 104–110 (2001) 34. C˘ald˘araru, A., Giaquinto, A., Witherspoon, S.: Algebraic deformations arising from orbifolds with discrete torsion. http://arxiv.org/abs/math/0210027v2[math.KT], 2003 35. Melnikov, I., Plesser, R.: A-model correlators from the Coulomb branch. JHEP 0602, 044 (2006) 36. Bertram, A.: Private communication, January 5, 2007 37. Beauville, A.: Complex Algebraic Surfaces. Second edition, Cambridge: Cambridge University Press, 1996 38. Iyer, J., Simpson, C.: A relation between the parabolic Chern characters of the de Rham bundles. http:// arxiv.org/abs/math/0603677v2[math.AG], 2006 39. Adams, A., Polchinski, J., Silverstein, E.: Don’t panic! Closed string tachyons in ALE spacetimes. JHEP 0110, 029 (2001) 40. Harvey, J., Kutasov, D., Martinec, E., Moore, G.: Localized tachyons and RG flows. http://arxiv.org/abs/ hep-th/0111154v2, 2001 41. Martinec, E., Moore, G.: On decay of K theory. http://arxiv.org/abs/hep-th/0212059v1, 2002 42. Kuznetsov, A.: Private communication, March 10, 2007 43. Orlov, D.: Triangulated categories of singularities and D-branes in Landau-Ginzburg models. http:// arxiv.org/abs/math/0302304v2[math.AG], 2004 44. Orlov, D.: Triangulated categories of singularities and equivalences between Landau-Ginzburg models. http://arxiv.org/abs/math/0503630v1[math.AG], 2005 45. Orlov, D.: Derived categories of coherent sheaves and triangulated categories of singularities. math.AG/0503632 46. Pantev, T., Sharpe, E.: Work in progress 47. Guffin, J., Sharpe, E.: To appear 48. Kontsevich, M.: Course on non-commutative geometry. ENS, 1998, Lecture notes at http://www.math. uchicago.edu/~mitya/langlands/html 49. Kontsevich, M.: Talk at the “Hodge centennial conference,” Edinburgh, 2003 50. Soibelman, Y.: Lectures on deformation theory and mirror symmetry. IPAM, 2003, http://www.math.ksu. edu/~soibel/ipam-final.ps, 2003 51. Costello, K.: Topological conformal field theories and Calabi-Yau categories. http://arxiv.org/abs/math/ 0412149v7[math.QA], 2006 52. Toën, B., Vaquie, M.: Moduli of objects in dg-categories. http://arxiv.org/abs/math/0503269v5[math. AG], 2007 53. Bondal, A., van den Bergh, M.: Generators and representability of functors in commutative and noncommutative geometry. Mosc. Math. J. 3, 1–36 (2003) Communicated by N.A. Nekrasov
Commun. Math. Phys. 294, 647–702 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0975-1
Communications in
Mathematical Physics
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory Richard J. Szabo1 , Alessandro Valentino2 1 Department of Mathematics and Maxwell Institute for Mathematical Sciences,
Heriot-Watt University, Colin Maclaurin Building, Riccarton, Edinburgh EH14 4AS, U.K. E-mail:
[email protected] 2 Courant Research Center, “Higher Order Structures” and Mathematisches Institut, Georg-August-Universität Göttingen, Bunsenstr. 3-5, D-37073 Göttingen, Germany. E-mail:
[email protected] Received: 28 January 2008 / Accepted: 10 November 2009 Published online: 25 December 2009 – © Springer-Verlag 2009
Abstract: We study D-branes and Ramond-Ramond fields on global orbifolds of Type II string theory with vanishing H -flux using methods of equivariant K-theory and K-homology. We illustrate how Bredon equivariant cohomology naturally realizes stringy orbifold cohomology. We emphasize its role as the correct cohomological tool which captures known features of the low-energy effective field theory, and which provides new consistency conditions for fractional D-branes and Ramond-Ramond fields on orbifolds. We use an equivariant Chern character from equivariant K-theory to Bredon cohomology to define new Ramond-Ramond couplings of D-branes which generalize previous examples. We propose a definition for groups of differential characters associated to equivariant K-theory. We derive a Dirac quantization rule for Ramond-Ramond fluxes, and study flat Ramond-Ramond potentials on orbifolds. Introduction The study of fluxes and D-branes has been of fundamental importance in understanding the nonperturbative structures of string theory and M-theory. It has also established a common ground on which a fruitful interaction between physics and mathematics takes place. For example, the seminal papers [49,62] demonstrated that D-brane charges in Type II superstring theory are classified by the K-theory of the spacetime manifold, and that ordinary cohomology alone cannot account for certain physical features induced by the dynamics of D-branes. As emphasized by refs. [2,36,55,60], and analyzed in great detail in refs. [56,57], another description of D-branes is provided by K-homology which sheds light on their geometrical nature and suggests that the standard picture of a D-brane as a submanifold of spacetime equipped with a vector bundle (and connection) should be modified. Ramond-Ramond fields are dual objects to D-branes and have also been extensively investigated, but until recently their geometric nature has remained somewhat obscure. In ref. [52] it was proposed that Ramond-Ramond fields are also classified by K-theory,
648
R. J. Szabo, A. Valentino
and that their total field strengths lie in the image of the Chern character homomorphism from K-theory to ordinary cohomology. This result led to the understanding that the Ramond-Ramond field is correctly understood as a self-dual field quantized by K-theory, and it explains various subtle issues surrounding the partition functions of these fields. In refs. [28,29] it was proposed that these properties are most naturally formulated by regarding Ramond-Ramond fields as cocycles for the differential K-theory of spacetime, an elegant description that allows one to study the gauge theory of Ramond-Ramond fields in topologically non-trivial backgrounds which naturally incorporates consistency conditions such as anomaly cancellation on branes in string theory and M-theory. These issues were among the motivations that led to the foundational paper [37], in which a detailed, elaborate construction for generalized differential cohomology theories is given. The importance of these mathematical theories has been greatly emphasized in refs. [32,33], where they are used to define and understand certain novel properties of quantum Hilbert spaces of abelian gauge field fluxes. A twisted version of differential K-theory has been proposed in refs. [11,33] and applied to the quantization of Ramond-Ramond fields in an H -flux background, while a rigorous geometrical definition of this theory has been developed recently in ref. [18]. The goal of this paper is to extend these lines of developments to study properties of Ramond-Ramond fields and D-branes in orbifolds of Type II superstring theory with vanishing H -flux. We limit our study to the cases of good (or global) orbifolds [X/G], where X is a manifold and G is a finite group acting via diffeomorphisms of X . It is possible to resolve singularities in the orbifold where it fails to be a manifold, and replace the quotient space by a non-compact manifold with appropriate asymptotic behaviour. However, orbifold singularities do not pose a problem and one can still have consistent superstrings propagating on orbifolds [24,25]. It was proposed in ref. [62] that D-branes on the orbifold spacetime [X/G] are classified by the G-equivariant K-theory of the covering space X , as defined in ref. [58]. A recent overview of related developments in the case of abelian orbifolds can be found in ref. [41]. One of the main new ingredients that we introduce into the description of D-branes and fluxes on orbifolds is the use of Bredon cohomology [16,23]. This is a powerful equivariant cohomology theory that has both advantages and pitfalls. In contrast to the more commonly used Borel equivariant cohomology, Bredon cohomology is a good “approximation” to the classification of D-brane charges. We will support this statement by showing that it correctly captures the properties of Ramond-Ramond fields on an orbifold, in particular it naturally takes into account the twisted sectors of the string theory. It thereby gives a precise, rigorous realization of stringy orbifold cohomology. We will also see that it naturally arises in the Atiyah-Hirzebruch spectral sequence for equivariant K-theory, a fact that we shall exploit to describe new consistency conditions for D-branes and fluxes on orbifolds in terms of classes in the Bredon cohomology of the covering space X . Related to this feature is the fact that this equivariant cohomology theory is the target for a Chern character homomorphism on equivariant K-theory, defined in ref. [46], which induces an isomorphism when tensored over R. By means of this technology, we present new compact and elegant expressions for the Wess-Zumino couplings of Ramond-Ramond fields to D-branes on [X/G]. This generalizes the usual Ramond-Ramond couplings [49] to orbifolds, and yields appropriate correction terms to previous flat space formulas. The major drawback of Bredon cohomology is that it is a rather difficult, abstract theory to define, and is even more difficult to explicitly calculate than other equivariant cohomology theories.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
649
Another main achievement of this paper is a proposed definition of differential K-theory suitable for orbifolds. Though extremely powerful and general, the machinery developed in ref. [37] cannot be immediately applied to an equivariant cohomology functor on the category of G-manifolds. By using Bredon cohomology and the equivariant Chern character, we define abelian groups that behave as natural generalizations of the ordinary differential K-theory groups, in the sense that they agree in the case of a trivial group and they satisfy analogous exact sequences. Although far from having the generality of the work of ref. [37], our construction gives a systematic framework in which to study Ramond-Ramond fields on orbifolds with a Dirac quantization condition, including non-trivial contributions from flat potentials, and it represents a first step in the development of generalized differential cohomology theories in the equivariant setting. It is here that the use of Bredon cohomology is particularly important, both because of the equivariant Chern character isomorphism and because the framework requires explicit use of differential forms, neither of which can be accommodated directly by the Borel construction. The outline of the remainder of this paper is as follows. In Sect. 1 we summarize some basic notions about the cohomology theories of spaces with group actions. In Sect. 2 we present a detailed definition of Bredon cohomology and the construction of the equivariant Chern character of ref. [46], as these have not made appearances before in the physics literature. These first two sections give the main mathematical background for the rest of the paper. In Sect. 3 we make a brief excursion into the description of D-branes using geometric equivariant K-homology, showing that the use of K-cycles is very well-suited to the description of fractional D-branes and their topological charges computed using equivariant Dirac operator theory. In Sect. 4 we use Bredon cohomology and the equivariant Chern character to define Ramond-Ramond couplings to D-branes on orbifolds and compare it with previous examples in the literature. Our formulas include the appropriate gravitational contributions which are derived from an equivariant version of the Riemann-Roch theorem and equivariant index theory. In Sect. 5 we give a detailed mathematical construction of the orbifold differential K-theory groups, and prove that they fit into appropriate exact sequences which are useful in applications. In Sect. 6 we use the orbifold differential K-theory to describe the flux quantization of Ramond-Ramond fields on orbifolds by writing an equivariant version of the Ramond-Ramond current in terms of the equivariant Chern character. We also study the group of flat potentials in detail, and illustrate how the spectral sequence for equivariant K-theory can be used to determine obstruction classes in Bredon cohomology which yield stability conditions for D-branes and fluxes on orbifolds. Appendix A contains some background material on functor categories used in the main text, Appendix B records the definitions of equivariant K-homology, while Appendix C demonstrates the use of geometric equivariant K-cycles in the classification of D-brane charges on orbifolds. 1. Cohomology of Spaces with Symmetries In this section we will recall some basic notions about (generalized) equivariant cohomology theories that we will need throughout this paper. In the following, X denotes a topological space and G a finite group, unless otherwise stated. Throughout a (left) action G × X → X of G on X will be denoted (g, x) → g · x. The stabilizer or isotropy group of a point x ∈ X is denoted G x = {g ∈ G | g · x = x}. Recall that a continuous map f : X → Y of G-spaces is a G-map if f (g · x) = g · f (x) for all g ∈ G and x ∈ X .
650
R. J. Szabo, A. Valentino
1.1. G-complexes. A G-equivariant CW-decomposition of a G-space X consists of a filtration X n , n ∈ N0 such that X= Xn n∈N0
and X n is obtained from X n−1 by “attaching” equivariant cells by the following procedure. Define G/K j , X0 = j∈J0
with K j a collection of subgroups of G and the standard (left) G-action on any coset space G/K j . For n ≥ 1 set ⎞ ⎛ n (1.1) B j × G/K j ⎠ ∼, X n = ⎝ X n−1 j∈Jn
where the equivalence relation ∼ is generated by G-equivariant “attaching maps” φ nj : Sn−1 × G/K j −→ X n−1 . j
(1.2)
One requires that X carries the colimit topology with respect to (X n ), i.e., B ⊂ X is closed if and only if B ∩ X n is closed in X n for all n ∈ N0 . We call the image of ˚ n × G/K j ) a closed (resp. open) n-cell of orbit type G/K j . As Bnj × G/K j (resp. B j usual, we call the subspace X n the n-skeleton of X . If X = X n and X = X n−1 , then n is called the (cellular) dimension of X and X is said to be of finite type. A G-space with a G-equivariant CW-decomposition is called a G-complex. When G = e is the trivial group, a G-complex is just an ordinary CW-complex. In general, if X is a G-complex then the orbit space X/G is an ordinary CW-complex. Conversely, there is an intimate relation between G-complexes and ordinary CWcomplexes whenever G is a discrete group. Let X be a G-space which is an ordinary CW-complex. We say that G acts cellularly on X if 1) For each g ∈ G and each open cell E of X , the left translation g · E is again an open cell of X ; and 2) If g · E = E, then the induced map E → E, x → g · x is the identity. Then we have the following Proposition 1.1. Let X be a CW-complex with a cellular action of a discrete group G. Then X is a G-complex with n-skeleton X n . In the case that X is a smooth manifold, we require the G-action on X to be smooth and there is an analogous result. Recall that the applicability of algebraic topology to manifolds relies on the fact that any manifold comes equipped with a canonical CW-decomposition. In the case in which a group acts on the manifold one has the following result due to Illman [38,39]. Theorem 1.2. If G is a compact Lie group or a finite group acting on a smooth compact manifold X , then X is triangulable as a finite G-complex. The collection of G-complexes with G-maps as morphisms form a category. We are interested in equivariant cohomology theories defined on this category (or on subcategories thereof).
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
651
1.2. Equivariant cohomology theories. We will now briefly spell out the main ingredients involved in building an equivariant cohomology theory on the category of finite G-complexes, leaving the details to the comprehensive treatments of refs. [23 and 45], and focusing instead on some explicit examples. Fix a group G and a commutative ring R. A G-cohomology theory E•G with values in R-modules is a collection of contravariant n from the category of G-CW pairs to the category of R-modules indexed by functors EG n ∈ Z together with natural transformations n n n+1 n+1 δG (X, A) : EG (X, A) −→ EG (X ) := EG (X, ∅)
for all n ∈ Z satisfying the axioms of G-homotopy invariance, long exact sequence of a pair, excision, and disjoint union. The theory is called ordinary if for any orbit G/H q one has EG (G/H ) = 0 for all q = 0. These axioms are formulated in an analogous way to that of ordinary cohomology. The new ingredient in an equivariant cohomology theory (which we have not yet defined) are the induction structures, which we shall now describe. Let α : H → G be a group homomorphism, and let X be an H -space. Define the induction of X with respect to α to be the G-space indα X given by indα X := G ×α X. This is the quotient of the product G×X by the H -action h·(g, x) := (g α(h −1 ), h·x), with the G-action on indα X given by g · [g, x] = [g g, x]. If H < G and α is the subgroup inclusion, the induced G-space is denoted G × H X . • An equivariant cohomology theory E(−) with values in R-modules consists of a • collection of G-cohomology theories EG with values in R-modules for each group G such that for any group homomorphism α : H → G and any H -CW pair (X, A) with ker(α) acting freely on X , there are for each n ∈ Z natural isomorphisms ≈
n → EnH (X, A) indα : EG (indα (X, A)) −
(1.3)
satisfying (a) Compatibility with the coboundary homomorphisms: n δ nH ◦ indα = indα ◦ δG ;
(b) Functoriality: If β : G → K is another group homomorphism such that ker(β ◦ α) acts freely on X , then for every n ∈ Z one has indβ◦α = indα ◦ indβ ◦ EnK ( f 1 ), where ≈
f 1 : indβ (indα (X, A)) − → indβ◦α (X, A) (k, g, x) −→ (k β(g) , x) is a K -homeomorphism and EnK ( f 1 ) is the morphism on K -cohomology induced by f 1 ; and
652
R. J. Szabo, A. Valentino
(c) Compatibility with conjugation: For g, g ∈ G define Adg (g ) = g g g −1 . Then the n ( f ), where homomorphism indAdg coincides with EG 2 ≈
f 2 : (X, A) − → indAdg (X, A) x −→ e , g −1 · x is a G-homeomorphism, where throughout e denotes the identity element in the group G. Thus the induction structures connect the various G-cohomologies and keep track of the equivariance. They will be very important in the construction of the equivariant Chern character for equivariant K-theory in the next section, even if we are only interested in a fixed group G. Example 1.3 (Borel cohomology). Let H• be a cohomology theory for CW-pairs (for example, singular cohomology). Define HnG (X, A) := Hn (E G ×G (X, A)), where E G is the total space of the classifying principal G-bundle E G → BG which is contractible and carries a free G-action. This is called (equivariant) Borel cohomology, and it is the most commonly used form of equivariant cohomology in the physics literature. Note that H•G is well-defined because the quotient E G ×G X is unique up to the homotopy type of X/G. The ordinary G-cohomology structures on H•G are inherited from the cohomology structures on H• . The induction structures for H•G are constructed as follows. Let α : H → G be a group homomorphism and X an H -space. Define b : E H × H X −→ E G ×G G ×α X, (ε, x) −→ (Eα(ε) , e , x), where ε ∈ E H , x ∈ X and Eα : E H → E G is the α-equivariant map induced by α. The induction map indα is then given by pullback indα := b∗ : HnG (indα X ) = Hn (E G ×G G ×α X ) −→ Hn (E H × H X ) = HnH (X ). If ker(α) acts freely on X , then the map b is a homotopy equivalence and hence the map indα is an isomorphism. Example 1.4 (Equivariant K-theory). In ref. [58], equivariant topological K-theory is defined for any G-complex X as the abelian group completion of the semigroup Vect C G (X ) of complex G-vector bundles over X , i.e., bundles E → X together with a lift of the G-action on X to the fibres. The higher groups are defined via iterated suspension. To define the induction structures, recall that if X is an H -space and α : H → G is a group homomorphism, then the map ϕ : X −→ G ×α X x −→ (e, x) is an α-equivariant map which embeds X as the subspace H ×α X of G ×α X , and which induces via pullback of vector bundles the homomorphism • ϕ ∗ : KG (G ×α X ) −→ K•H (X ).
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
653
This map defines the induction structure. It is invertible when ker(α) acts freely on X , with inverse the “extension” map E → G × H E for any H -vector bundle E over X . The induction structure can be used to prove the well-known equivariant excision theorem • • KG/N (X/N ) ∼ (X ), = KG
(1.4)
where N is a normal subgroup of G acting freely on X . Indeed, one has X/N ∼ = (G/N ) ×G X and if we define α : G → G/N to be the quotient map, then • • KG/N (X ). ((G/N ) ×α X ) ∼ = KG
since ker(α) = N acts freely on X . 2. The Equivariant Chern Character • functor and In this section we will describe the equivariant Chern character for the KG its target cohomology theory, Bredon cohomology. To this end, we will introduce some technology related to modules over functor categories, giving the necessary definitions and directing the reader to the relevant literature for further details. Some pertinent aspects of functor categories are summarized in Appendix A.
2.1. Chern character in topological K-theory. Let us begin by recalling some basic notions about the ordinary Chern character. Define π−• K to be the complex K-theory ring of the point. It is the Z-graded ring Z[[u, u −1 ]] of Laurent polynomials freely generated by an element u of degree deg(u) = 2, where u −1 ∈ K−2 (pt) is called the Bott element and is represented by the Hopf bundle over S2 . One then has a homomorphism ch : K• (X ) −→ H(X ; R ⊗ π−• K)• which induces the natural Z-graded ring isomorphism ≈
→ H(X ; R ⊗ π−• K)• K• (X ) ⊗ R − for any finite CW-complex X . This statement is true even if we tensor over Q. The use here of the K-theory of the point as the coefficient ring serves just as a re-grading of the cohomology ring H• (X ; R). For example, it is easy to check that H(X ; R ⊗ π−• K)0 ∼ = Heven (X ; R). In particular, the Chern character tells us that K-theory and cohomology are the same thing up to torsion. It is natural now to ask if there exists such a morphism for equivariant K-theory. One might naively think that the correct target theory for the equivariant Chern character would naturally be Borel cohomology. But the problem is much more subtle than it first may seem. The crucial point is that while in the ordinary cohomology of (finite) CW-complexes the building blocks are the cohomology groups of a point, in the equivariant case they are the cohomology groups of the orbits G/H for all subgroups H
654
R. J. Szabo, A. Valentino
• on the category of G, as we saw in Sect. 1.1. Any equivariant cohomology theory EG of finite G-complexes is completely specified by its value on the orbit spaces G/H . A localization theorem due to Atiyah and Segal [4] tells us that the Borel cohomology of a G-space X is isomorphic to its equivariant K-theory localized at the augmentation ideal in the representation ring R(G) consisting of all elements whose characters vanish at • (X ) as a module over R(G)). Localizing at a prime the identity e in G (regarding KG ideal of R(G) corresponds to restricting X to the set of fixed points of an associated conjugacy class of cyclic subgroups of G. In this sense, Borel cohomology does not take into account the “contributions” of the non-trivial elements in G, and hence of the fixed points of the G-action. There are several approaches to the equivariant Chern character (see refs. [1,5,15, 31,59], for example) which strongly depend on the types of groups involved (discrete, continuous, etc.) and on the ring one tensors with (R, C, etc.). As we are interested in finite groups and real coefficients for our physical applications later on, we will use the Chern character constructed in refs. [45 and 46]. Thus we proceed to the more abstract, but powerful and compact, definition of Bredon cohomology, which will turn out to be the best suited equivariant cohomology theory for all of our purposes.
2.2. Bredon cohomology. Let G be a discrete group. The orbit category Or(G) of G is defined as the category whose objects are homogeneous spaces G/H , with H < G, and whose morphisms are G-maps between them. From general considerations [23] it follows that a G-map between two homogeneous spaces G/H and G/K exists if and only if H is conjugate to a subgroup of K , and hence any such map is of the form (g H −→ g a K )
(2.1)
for some a ∈ G such that a −1 H a < K . If F is any family of subgroups of G then there is a subcategory Or(G, F) with objects G/H for H ∈ F. A simple example is provided by the cyclic groups G = Z p with p prime, for which the orbit category has just two objects, G/e = G and G/G = pt. If Ab denotes the category of abelian groups, then a coefficient system is a functor F : Or(G)op −→ Ab, where Or(G)op denotes the dual category to Or(G). With such a functor and any G-complex X ,1 one can define for each n ∈ Z the group
n CG (2.2) (X, F ) := HomOr(G) C n (X ) , F , where C n (X ) : Or(G)op → Ab is the projective functor defined by C n (X )(G/H ) := Cn X H , the cellular homology of the fixed point complex
X H := x ∈ X h · x = x ∀h ∈ H .
(2.3)
1 When G is an infinite discrete group, one should restrict to proper G-complexes, i.e., with finite stabilizer for any point of X . Some further minor assumptions are needed when G is a Lie group.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
655
In Eq. (2.2), HomOr(G) (−, −) denotes the group of natural transformations between two contravariant functors, with the group structure inherited by the images of the functors in Ab. The functoriality property of C n (X )(G/H ) is the natural one induced by the identification X H ∼ = MapG (G/H, X ). Indeed, the two maps X H −→ MapG (G/H, X ),
x −→ f x ([g H ]) = g · x,
MapG (G/H, X ) −→ X H ,
f −→ f (H )
are easily seen to be inverse to each other, and the desired homeomorphism is obtained by giving the space MapG (G/H, X ) the compact-open topology. In particular, a G-map (2.1) induces a cellular map X K → X H , x → a · x. These groups can be expressed in terms of the G-complex structure of X . If the n-skeleton X n is obtained by attaching equivariant cells as in Eq. (1.1) with K j the stabilizer of an n-cell of X , then the cellular chain complex C• (X ) consists of G-modules Cn (X ) = j∈Jn Z[G/K j ], and hence Z MorOr(G) (G/H, G/K j ) . C n (X )(G/H ) ∼ = j∈Jn n (X , F ) is the direct limit functor over all n-cells of For each n ≥ 0, the group C G orbit type G/K j in X of the groups F (G/K j ). This follows by restricting Eq. (2.2) to the full subcategory Or(G, F(X )), with F(X ) the family of subgroups of G which occur as stabilizers of the G-action on X [50]. n • (X, F ) = The Z-graded group C G n∈Z C G (X, F ) inherits a coboundary operator δ, and hence the structure of a cochain complex, from the boundary operator on cellular chains. To a natural transformation f : C n (X ) → F , one associates the natural transformation δ f defined by δ f (G/H ) : Cn X H −→ F (G/H )
σ −→ f (G/H )(∂σ ) for σ ∈ Cn−1 (X H ), with naturality induced from that of the cellular boundary operator ∂. Then the Bredon cohomology of X with coefficient system F is defined as
• H•G (X ; F ) := H C G (X, F ) , δ . This defines a G-cohomology theory. See ref. [44] for the proof that H•G (X ; F ) is an equivariant cohomology theory, i.e., for the definition of the induction structure. One can also define cohomology groups by restricting the functors in Eq. (2.2) to a subcategory Or(G, F). The definition of Bredon cohomology is independent of F as long as F contains the family F(X ) of stabilizers [50]. This fact is useful in explicit calculations. In particular, by taking F = H to consist of a single subgroup, one shows that the Bredon cohomology of G-homogeneous spaces is given by H•G (G/H ; F ) = = H0G (G/H ; F ) = = F (G/H ).
(2.4)
Example 2.1 (Trivial group). When G = e is the trivial group, i.e., in the non-equivariant case, the functors C n (X ) and F can be identified with the abelian groups Cn (X ) = C n (X )(e) and F = F (e). Then Cen (X, F) = C n (X, F)
656
R. J. Szabo, A. Valentino
and one has Hen (X ; F ) = H (C n (X, F), δ), i.e., the ordinary n th cohomology group of X with coefficients in F. Example 2.2 (Free action). If the G-action on X is free, then all stabilizers K j are trivial and X H = ∅ for every H ≤ G, H = e. In this case one may take F = e to compute the cochain complex
• CG (X, F ) ∼ = Hom G C• (X ) , F (G/e) and so the Bredon cohomology H•G (X ; F ) coincides with the equivariant cohomology
H•G X ; F (G/e) of X with coefficients in the G-module F (G/e) = F (G). In the case of the constant functor F = Z, with Z (G/H ) = Z for every H ≤ G and the value on morphisms in Or(G)op given by the identity homomorphism of Z, this group reduces to the ordinary cohomology H• (X/G; Z). Example 2.3 (Trivial action). If the G-action on X is trivial, then the collection of isotropy groups K j for the G-action is the set of all subgroups of G and X H = X for all H ≤ G. In this case the functor C n (X ) can be decomposed into a sum over n-cells of projective functors P K j with K j = G [50], and so one has
HomOr(G) C n (X ) , F ∼ = Hom Cn (X ) , lim Or(G)op F (G/H ) , ←−
where the inverse limit functor is taken over the opposite category Or(G)op . It follows that the Bredon cohomology
H•G (X ; F ) = H• X ; F (G/G) is the ordinary cohomology of X with coefficients in the abelian group F (G/G) = F (pt). 2.3. Representation ring functors. In what follows we will specialize the coefficient system for Bredon cohomology to the representation ring functor F = R(−) defined on the orbit category Or(G) by sending the left coset G/H to R(H ), the complex representation ring of the group H . A morphism (2.1) is sent to the homomorphism R(K ) → R(H ) given by first restricting the representation from K to the subgroup conjugate to H , and then conjugating by a. Since R(−) is a functor to rings, the Bredon cohomology H•G (X ; R(−)) naturally has a ring structure. Note that 0 • (G/H ) = KG (G/H ) , R(H ) ∼ = KG
(2.5)
which follows from the induction structure of Example 1.4 with X = pt and α the subgroup inclusion H → G. By Eq. (2.4) the group (2.5) also coincides with the Bredon cohomology group H•G (G/H ; R(−)), which is already an indication that Bredon cohomology is a better relative of equivariant K-theory than Borel cohomology. Indeed, using the induction structure of Example 1.3 one shows that the Borel cohomology H•G (G/H ) = H• (B H )
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
657
coincides with the cohomology of the classifying space B H = E H/H , which computes the group cohomology of H and is typically infinite-dimensional (even for finite groups H ). In this paper we will show that the Bredon cohomology H•G (X ; R(−)) gives a more precise realization of the stringy orbifold cohomology of X in the context of open string theory. In the construction of the equivariant Chern character in Sect. 2.4 below, it will be important to represent the rational Bredon cohomology H•G (X ; Q ⊗ R(−)) as a certain group of homomorphisms of functors, similarly to the cochain groups (2.2). For this, we introduce another category Sub(G). The objects of Sub(G) are the subgroups of G,2 and the morphisms are given by MorSub(H,K ) :=
Inn(K ). f : H → K ∃ g ∈ G, g H g −1 ≤ K , f = Adg
In particular, there is a functor Or(G) → Sub(G) which sends the object G/H to H and the morphism (2.1) in Or(G) to the homomorphism (g → a −1 g a) in Sub(G). If a lies in the centralizer
(2.6) Z G (H ) := g ∈ G g −1 H g = H of H in G, then the morphism (2.1) is sent to the identity map. Any functor F : Sub(G)op → Ab can be naturally regarded as a functor on Or(G)op . qt qt Define the quotient functors C • (X ) , H • (X ) : Sub(G)op → Ab by qt C • (X )(H ) := C• X H /Z G (H )
and
qt H • (X )(H ) := H• X H /Z G (H ) .
For any functor F : Sub(G)op → Ab one has Hom C• (X H /Z G (H )) , F (H ) ∼ = Hom Z G (H ) C• (X H ), F (H ) . By observing that the centralizer (2.6) is precisely the group of automorphisms of G/H in the orbit category Or(G) sent to the identity map in the subgroup category Sub(G), we finally have
qt • CG (X, F ) = HomOr(G) C • (X ) , F ∼ = HomSub(G) C • (X ) , F . (2.7) At this point one can apply Eq. (2.7) to the rational representation ring functor F = Q ⊗ R(−), which by construction can be regarded as an injective functor Sub(G)op → Ab, to prove the Lemma 2.4 ([46]). For any finite group G and any G-complex X , there exists an isomorphism of rings
≈ qt X : H•G X ; Q ⊗ R(−) − → HomSub(G) H • (X ) , Q ⊗ R(−) . 2 If G is infinite then one should restrict to finite subgroups of G.
658
R. J. Szabo, A. Valentino
2.4. Chern character in equivariant K-theory. Before spelling out the definition of the equivariant Chern character, we recall some basic properties of the equivariant K-theory of a G-complex X . Let H be a subgroup of G, and consider the fixed point subspace of X defined by Eq. (2.3). The action of G does not preserve X H , but the action of the normalizer N G (H ) of H in G does. If we denote with i : X H → X the inclusion of X H as a subspace of X , and with α : N G (H ) → G the inclusion of N G (H ) as a subgroup of G, then we naturally have the equality i(n · x) = α(n) · i(x) for all n ∈ N G (H ) and x ∈ X H . It follows that the induced homomorphism on equivariant K-theory is a map [58] • (X ) −→ K•NG (H ) X H i ∗ : KG which is called a restriction morphism. We also need a somewhat less known property [46]. Let N G be a finite normal subgroup, and let Rep(N ) be the category of (isomorphism classes of) irreducible complex representations of N . Let X be a (proper) G/N -complex, and let G act on X via the projection map G → G/N . Then for any complex G-vector bundle E → X and any representation V ∈ Rep(N ), define Hom N (V, E) as the vector bundle over X with total space Hom N (V, E) := Hom N (V, E x ), x∈X
where N acts on the fibres of E because of the action of G via the projection map. Now if H ≤ G is a subgroup which commutes with N , [H, N ] = e, then one can induce an H -vector bundle from Hom N (V, E) by defining (h · f )(v) = h · f (v), v ∈ V for any h ∈ H and any f ∈ Hom N (V, E) (remembering that G acts on E). Hence there is a homomorphism of rings • : KG (X ) −→ K•H (X ) ⊗ R(N )
defined on G-vector bundles by ([E]) :=
[Hom N (V, E)] ⊗ [V ].
(2.8)
V ∈Rep(N )
This homomorphism satisfies some naturality properties which are described in detail in ref. [46]. Note that the sum (2.8) is finite, since N is a finite subgroup. We are now ready to construct the equivariant Chern character as a homomorphism
0,1 ch X : KG X ; Q ⊗ R(−) (X ) −→ Heven,odd G for any finite proper G-complex X . The strategy used in ref. [46] is to construct Z2 -graded homomorphisms • H ch H : K (X ) −→ Hom H (X /Z (H )) , Q ⊗ R(H ) (2.9) • G X G
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
659
for any finite subgroup H , and then glue them together as H varies through the finite subgroups of G. To define the homomorphism (2.9), we first compose the ring homomorphisms i∗ • KG (X ) − → K•NG (H ) X H − → K•Z G (H ) X H ⊗ R(H ) π2∗ ⊗id −−−→ K•Z G (H ) E G × X H ⊗ R(H ), where π2 : E G × X H → X H is the projection onto the second factor. By using the induction structure of Example 1.4, one then has ≈ → K• E G × Z G (H ) X H ⊗ R(H ) K•Z G (H ) E G × X H ⊗ R(H ) − • ch⊗id −−−→ H E G × Z G (H ) X H ; Q ⊗ π−• K ⊗ R(H ), where ch is the ordinary Chern character. One finally has ≈ → H• X H /Z G (H ) ; Q ⊗ R(H ) H• E G × Z G (H ) X H ; Q ⊗ R(H ) − ∼ = Hom H• (X H /Z G (H )) , Q ⊗ R(H ) , where the first isomorphism follows from the Leray spectral sequence by observing that the fibres of the projection E G × Z G (H ) X H −→ X H Z G (H ) are all classifying spaces of finite groups, having trival reduced cohomology with Q-coefficients and are therefore Q-acyclic. The equivariant Chern character is now defined as3 ch X = ch H (2.10) X. H ≤G
By using the various naturality properties of the homomorphism (2.8) [46], one sees qt that ch X takes values in HomSub(G) H • (X ) , Q ⊗ R(−) , and by Lemma 2.4 it is thus a Z2 -graded map
qt • (X ) −→ HomSub(G) H • (X ), Q ⊗ R(−) ∼ ch X : KG = H•G X ; Q ⊗ R(−) . This map is well-defined as a ring homomorphism because all maps involved above are homomorphisms of rings. As with the definition of Bredon cohomology, the sum (2.10) may be restricted to any family of subgroups of G containing the set of stabilizers F(X ). To conclude, we have to prove that this map becomes an isomorphism upon tensoring over Q. For this, one proves that the morphism ch X in Eq. (2.10) is an isomorphism on homogeneous spaces G/H , with H a finite subgroup of G, and then uses induction on the number of orbit types of cells in X along with the Mayer-Vietoris sequences for the 3 If G is infinite then the direct sum in Eq. (2.10) is understood as the inverse limit functor over the dual subgroup category Sub(G)op .
660
R. J. Szabo, A. Valentino
pushout squares induced by the attaching G-maps (1.2). The isomorphism on G/H is a consequence of the isomorphisms (2.4) and (2.5). The details may be found in ref. [46]. • (G/H ). Then one has Let π−• K G (−) be the functor on Or(G) defined by G/H → KG the following Theorem 2.5. For any finite proper G-complex X , the Chern character ch X extends to a natural Z-graded isomorphism of rings • ≈ • (X ) ⊗ Q − → HG X ; Q ⊗ π−• K G (−) . ch X ⊗ Q : KG
3. D-Branes and Equivariant K-Cycles In this section we will make some remarks concerning the topological classification of D-branes and their charges on global orbifolds of Type II superstring theory with vanishing H -flux. Let X be a smooth manifold and G a (finite) group acting by diffeomorphisms of X . Ramond-Ramond charges on the global orbifold [X/G] are classified • (X ) as defined in Example 1.4 [34,54,62]. Dually, the by the equivariant K-theory KG G equivariant K-homology K• (X ) leads to an elegant description of fractional D-branes pinned at the orbifold singularities in terms of equivariant K-cycles. In the following we will frequently refer to Appendix B for detailed definitions and technical aspects of equivariant K-homology, focusing instead here on some of the more qualitative aspects of D-branes on orbifolds in this language. In the remainder of this paper we will assume for definiteness that G is a finite group. 3.1. Fractional D-branes. As in the non-equivariant case G = e [56,57,60], a very natural description of D-branes in the orbifold space, which captures the inherent geometrical picture of D-brane states involving wrapped cycles in spacetime, is provided by the topological realization of the equivariant K-homology groups K•G (X ). The cycles for this homology theory, called G-equivariant K-cycles, live in an additive category DG (X ) whose objects are triples (W, E, f ), where W is a G-spinc manifold without boundary, E is a G-vector bundle over W , and f : W −→ X
(3.1)
is a G-map. The group K•G (X ) is the quotient of this category by the equivalence relation generated by bordism, direct sum, and vector bundle modification, as detailed in Appendix B. Note that W need not be a submanifold of spacetime. However, since X is a manifold, we can restrict the bordism equivalence relation to differential bordism [56] and assume that the map (3.1) is a differentiable G-map in equivariant K-cycles (W, E, f ) ∈ DG (X ). In this way the category DG (X ) extends the standard K-theory classification to include branes supported on non-representable cycles in spacetime. This definition of equivariant K-homology thus gives a concrete geometric model for the topological classification of D-branes (W, E, f ) in a global orbifold [X/G] which captures the physical constructions of orbifold D-branes as G-invariant states of branes on the covering space X . In the subsequent sections we will study the pairing of Ramond-Ramond fields with these D-branes.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
661
Consider a D-brane localized at a generic point in the orbifold [X/G] with the action of the regular representation of G on the fibres of its Chan-Paton gauge bundle, i.e., the natural action of the group algebra C[G] as bounded linear operators 2 (G) → 2 (G). It corresponds to a G-orbit of such branes on the leaves X g = {x ∈ X | g · x = x}, g ∈ G of the covering space X . At a G-fixed point, this brane splits up into a set of fractional branes according to the decomposition of the representation of G on the fibres of the Chan-Paton bundle into irreducible G-modules. Stable fractional D-branes correspond to bound states of branes wrapping various collapsed cycles at the fixed points. They are thus stuck at the orbifold points and provide the open string analogs of “twisted sectors”. To formulate this physical construction in the language of equivariant K-cycles (W, E, f ), let G ∨ denote the set of conjugacy classes [g] of elements g ∈ G. It can be regarded as a set of representatives for the isomorphism classes π0 Rep(G), where Rep(G) is the additive category of irreducible complex representations of G which coincides with the category of D-brane boundary conditions at the orbifold points. There is G (X ) of DG (X ) consisting of triples (W, E, f ) for which W a natural subcategory Dfrac is a G-fixed space, i.e., for which Wg = W
(3.2)
for all g ∈ G. By G-equivariance this implies f (W )g = f (W ) for all g ∈ G, and so the image of the brane worldvolume lies in the subspace X g. f (W ) ⊂ g∈G
This is precisely the set of G-fixed points of X , and so the objects (W, E, f ) of the cateG G (X ) are naturally pinned to the orbifold points. We call Dfrac (X ) the category gory Dfrac of “maximally fractional D-branes”. In this case, an application of Schur’s lemma shows that the Chan-Paton bundle admits an isotopical decomposition and there is a canonical isomorphism of G-bundles
E ∼ E [g] ⊗ 11[g] with E [g] = = Hom G 11[g] , E , (3.3) = [g]∈G ∨
where E [g] is a complex vector bundle with trivial G-action and 11[g] is the G-bundle W × V[g] with γ : G → End(V[g] ) the irreducible representation corresponding to the conjugacy class [g] ∈ G ∨ . This isotopical decomposition defines the action of G on the Chan-Paton factors of the D-brane, and it implies the well-known isomorphism • (W ) ∼ KG = R(G) ⊗ K• (W )
(3.4)
for G-fixed spaces W [58]. This is a special case of the homomorphism defined in Eq. (2.8). From the direct sum relation in equivariant K-homology it follows that a D-brane, represented by a K-cycle (W, E, f ), splits at an orbifold point into a sum over irreducible fractional branes represented by the K-cycles (W, E [g] ⊗ 11[g] , f ), [g] ∈ G ∨ . It is important to realize that the full category DG (X ) contains much more information, and in particular the fractional D-branes will not generically form a spanning set of K-cycles for the group K•G (X ) (except in some specific examples). However, it follows from the bordism relation in equivariant K-homology that any two G-equivariant K-cycles (Wi , E i , f i ), i = 0, 1 which are bordant along the same G-orbit determine the same element in K•G (X ). This is expected since a purely topological classification
662
R. J. Szabo, A. Valentino
such as equivariant K-homology cannot capture the positional moduli associated with the regular D-branes in X/G. A related way to understand the role of fractional branes is through the connection between geometric K-homology and bordism theory [56]. Let MSpinc• (X ) be the ordinary spinc bordism group of X , which forgets about the G-action and consists of spinc bordism classes of pairs (W, f ). Then there is a map G MSpinc• (X ) ⊗MSpinc• (pt) Rep(G) −→ Dfrac (X ) , (W, f ) ⊗ V −→ (W, W × V, f ),
(3.5)
which descends to give a homomorphism MSpinc• (X ) ⊗MSpinc• (pt) K•G (pt) −→ K•G (X ). When G = e this is the isomorphism of K• (pt)-modules induced by the AtiyahBott-Shapiro orientation, and the map (3.5) determines K-cycle generators in terms of spinc bordism generators [56]. The equivariant extension of the Atiyah-Bott-Shapiro construction is given in ref. [42] in terms of finite-dimensional Z2 -graded G-Clifford modules. Since any G-Clifford module can be built as a direct sum of tensor products of G-modules and ordinary Clifford modules (see Appendix B), there is an isomor• (pt) ∼ R(G) ⊗ K • (pt) and so these representation modules phism of R(G)-modules KG = contain no new information about the orbifold group. This seems to suggest that, at least in certain cases, spanning sets of equivariant K-cycles can be taken to lie in the G subcategory Dfrac (X ). 3.2. Topological charges. The topological charge of a fractional D-brane, in a given closed string twisted sector of the orbifold string theory on a G-spinc manifold X , can be computed by using the equivariant Dirac operator theory developed in Appendix B. The equivariant index of the G-invariant spinc Dirac operator D / EX coupled to a G-vector • ∼ bundle E → X takes values in KG (pt) = R(G). We can turn this into a homomorphism on K•G (X ) with values in Z by composing with the projection R(G) → Z defined by taking the multiplicity of a given representation γ : G −→ End(Vγ )
(3.6)
of G on a finite-dimensional complex vector space Vγ . There is a corresponding class in the KK-theory group
[γ ] ∈ KK• C[G] , End(Vγ ) which is represented by the Kasparov module (Vγ , γ , 0) associated with the extension of the representation (3.6) to a complex representation of group ring C[G]. By Morita equivalence, the Kasparov product with [γ ] is the homomorphism on K-theory
K0 (C[G]) −→ K0 End(Vγ ) ∼ = K0 (C) ∼ =Z induced by γ : C[G] → End(Vγ ). We may then define a homomorphism µγ : K0G (X ) −→ Z
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
of abelian groups by µγ ([W, E, f ]) = Indexγ
f ∗ [D /W E]
:= ass f ∗ [D /W E ] ⊗C[G] [γ ]
663
(3.7)
on equivariant K-cycles (W, E, f ) ∈ DG (X ) (and extended linearly), where ass : K•G (X ) −→ K• (C[G]) is the analytic assembly map constructed in Appendix B. 3.3. Linear orbifolds. Let us now consider a simple class of examples wherein everything can be made very explicit. Let V be a complex vector space of dimension dimC (V ) = d ≥ 1, and let G be a finite subgroup of SL(V ). Our spacetime X is the G-space identified with the product X = R p,1 × V , where G acts trivially on the Minkowski space R p,1 . Fractional D-branes carrying themselves a complex linear representation of G, which is a submodule of R p,1 × V , have worldvolumes W linearly embedded in the subspace R p,1 and have transverse space given by the orthogonal complement f (W )⊥ ∼ = V with respect to a chosen inner product. Since the space of hermitian metrics is contractible, all topological quantities below are independent of this choice. As a G-space, V is equivariantly contractible to a point and hence its compactly supported equivariant K-theory is given by [4] ∨ • • KG,cpt (V ) ∼ (pt) ∼ = KG = R(G) = Z|G | .
This group coincides with the Bredon cohomology H•G,cpt (V ; R(−)), owing to the fact that the equivariant Chern character ch G/H of Sect. 2.4 is the identity map (since the non-equivariant Chern character ch = c0 : K0 (pt) → H0 (pt; Z) is the identity map). It follows that the fractional D-branes, as defined by elements of equivariant K-theory, can be identified with representations of the orbifold group4 γ =
∨ |G |
Na γa
a=1
consisting of Na ≥ 0 copies of the
a th
irreducible representation
γa : G −→ End(Va ) , a = = 1, . . . , G ∨ ,
which defines the action of G on the fibres of the Chan-Paton bundle. More precisely, each irreducible fractional brane is associated to the G-bundle V × Va over V . By Poincaré duality, it follows from Proposition 2.1 of ref. [56] that a basis for the equivariant K-homology group K•G (V ) is provided by the geometric equivariant K-cycles (V, V ×Va , id V ), a = 1, . . . , |G ∨ |. By G-homotopy invariance [56, Lemma 1.4] 4 If the transverse space V is instead a real linear G-module, then throughout one should restrict to the subring of R(G) consisting of representations associated to conjugacy classes [g] ∈ G ∨ for which the centralizer Z G (g) acts on the fixed point subspace V g by oriented automorphisms [40]. This will follow immediately from the isomorphism (4.7) below with X = V .
664
R. J. Szabo, A. Valentino
these cycles can be contracted to [pt, Va , i], where i is the inclusion of a point pt ⊂ V whose induced homomorphism i ∗ : K•G (pt) −→ K•G (V ) can be taken to be the identity map R(G) → R(G). This is simply the physical statement that the stable fractional branes in this case are D0-branes in Type IIA string theory (the pt Type IIB theory containing no such states). The G-invariant Dirac operator D / Va is just Clifford multiplication twisted by the G-module Va , and thus the topological charges (3.7) of the corresponding fractional branes in the twisted sector labelled by b are given by pt µb ([pt, Va , i]) = = Indexγb [D / Va ] = = Va ⊗ (+ − − ) ⊗C[G] [γb ] , where ± are the half-spin representations of SO(2d) on V ∼ = Cd regarded as C[G]modules. Acting on the character ring the projection gives [W ] ⊗C[G] [γb ] = χW (gb ), where χW : G → C is the character of the G-module W and [gb ] ∈ G ∨ is the conjugacy class corresponding to the irreducible representation γb . 4. Delocalization and Ramond-Ramond Couplings The purpose of this section is to describe the delocalization of Bredon cohomology and the equivariant Chern character, introduced in Sect. 2, and to apply it to the study of the coupling between Ramond-Ramond potentials and the D-branes of the previous section. In the following we will require some standard facts concerning string theory on global orbifolds, particularly its low-energy field theory content. For more details see refs. [24,25]. 4.1. Closed string spectrum. The boundary states corresponding to the fractional branes constructed in Sect. 3 have components in the twisted sectors of the closed string Hilbert space H of orbifold string theory on X . The closed string is an embedding x : S1 × R → X of the worldsheet cylinder, with local coordinates (σ, τ ) ∈ S1 × R, into the G-spinc spacetime manifold X . The Hilbert space H of physical string states decomposes into a direct sum over twisted sectors, each characterized by a conjugacy class, as H[g] (4.1) H= [g]∈G ∨
with only G-invariant states surviving in each superselection sector H[g] . Actually, the Hilbert space factorizes into one sector for each element of the group G, but the action of G mixes the sectors within a given conjugacy class. The subspaces in Eq. (4.1) are thus defined as H[g] := Hh , h∈[g]
where Hh is the subspace of states induced by the twisted string field boundary condition x(σ + 2π, τ ) = h · x(σ, τ )
(4.2)
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
665
with an analogous condition on the worldsheet fermion fields (using a lift Gˆ of the orbifold group). Then G acts on the subspace H[g] , and projecting onto G-invariant states in H[g] is equivalent to projecting onto Z G (h)-invariant states in Hh for any h in [g]. The low-energy limit of Type II orbifold superstring theory on X contains RamondRamond fields C[g] coming from the various twisted sectors. The twisted boundary conditions (4.2) on the string embedding map imposes constraints on the low-energy spectrum. For example, the untwisted sector given by g = e contains Ramond-Ramond fields defined on the entire spacetime manifold X , while the twisted sector represented by g = e gives rise to fields defined only on the fixed point submanifold X g . The GSO projection then enforces the properties that the Ramond-Ramond form potentials C[g] determine self-dual fields in each twisted sector, and that they be of odd degree in Type IIA theory and of even degree in Type IIB theory. The Ramond-Ramond fields can thus be “organised” into the differential complex
Z (g) •G (X ; R) := • X g ; R G . (4.3) [g]∈G ∨
Here we consider only fields coming from inequivalent twisted sectors and make a choice of submanifold X g , since for any conjugate element h ∈ [g] there is a diffeomorphism X g ∼ = X h . (No choice is needed in the case in which G is an abelian group.) As d ◦ g ∗ = g ∗ ◦ d for all g ∈ G, the derivation is given by dg , dG := [g]∈G ∨
where dg = d : • (X g ; R) → • (X g ; R) is the usual de Rham exterior derivative. Note that only the centralizer subgroup of g in G acts (properly) on X g . 4.2. Delocalization of Bredon cohomology. We will now show how Bredon cohomology can be used to compute the cohomology of the complex (4.3) of orbifold RamondRamond fields by giving a delocalized description of Bredon cohomology with real coefficients, following refs. [50 and 46] where further details can be found. This is the stringy orbifold cohomology of X , defined as the ordinary (real) cohomology of the orbifold resolution X = [g]∈G ∨ X g /Z G (g). Note that there is a natural surjective map π: X → X defined by (x, [g]) → x, and a natural injection X → X into the connected component of X corresponding to the untwisted sector [g] = [e]. Denote with R(−) the real representation ring functor R⊗ R(−) on the orbit category Or(G). Let G denote the set of conjugacy classes [C] of cyclic subgroups C of G. Let R C (−) be the contravariant functor on Or(G) defined by R C (G/H ) = 0 if [C] contains no representative g C g −1 < H , and otherwise R C (G/H ) is isomorphic to the cyclotomic field R(ζ|C| ) over R generated by the primitive root of unity ζ|C| of order |C|. A standard result from the representation theory of finite groups then gives a natural splitting R(−) = R C (−). [C]∈G
By definition, for any module M (−) over the orbit category one has
HomOr(G) M (−) , R C (−) ∼ = Hom NG (C) M (G/C) , R C (G/C) ∼ = M (G/C) ⊗ NG (C) R C (G/C),
666
R. J. Szabo, A. Valentino
where the normalizer subgroup N G (C) acts on R C (G/C) ∼ = R(ζ|C| ) via identification of a generator of C with ζ|C| . These facts together imply that the cochain groups (2.2) with F = R(−) admit a splitting given by
• CG X , R(−) ∼ C • X C ⊗ NG (C) R C (G/C). = [C]∈G
As the centralizer Z G (C) acts properly on X C , the natural map C • X C ⊗ NG (C) R C (G/C) −→ C • X C /Z G (C) [C]∈G
[C]∈G
⊗WG (C) R C (G/C) is a cohomology isomorphism, where WG (C) := N G (C)/Z G (C) is the Weyl group of C < G which acts by translation on X C /Z G (C). Since R C (G/C) is a projective R[WG (C)]-module, it follows that for any proper G-complex X the Bredon cohomology of X with coefficient system R ⊗ R(−) has a splitting
H•G X ; R ⊗ R(−) ∼ H• X C /Z G (C) ; R ⊗WG (C) R C (G/C). (4.4) = [C]∈G
At this point, we note that the dimension of the R-vector space R C (G/C)WG (C) ∼ = R ⊗WG (C) R C (G/C) is equal to the number of G-conjugacy classes of generators for C. We also use the fact that for a finite group G a sum over conjugacy classes of cyclic subgroups is equivalent to a sum over conjugacy classes of elements in G, and that X g = X g and Z G (g) = Z G (g). One finally obtains a splitting of real Bredon cohomology groups5
Z (g) H•G X ; R ⊗ R(−) ∼ H• X g ; R G (4.5) = [g]∈G ∨
into ordinary cohomology groups
Z (g) ∼ H• X g ; R G = H • (X g ; R) Z G (g) , d = H• X g /Z G (g) ; R ∼ with constant coefficients R. The group on the right-hand side of Eq. (4.5) corresponds to the (real) “delocalized equivariant cohomology” H• ( g∈G X g )G ⊗ R defined by Baum and Connes [6,9]. Note that this group is (non-canonically) isomorphic to R(G) ⊗ H• (X ; R) when the G-action on X is trivial. Furthermore, by using Theorem 2.5 one also has a decomposition for equivariant K-theory with real coefficients given by
Z (g) • KG K• (X g ) ⊗ R G . (X ) ⊗ R ∼ = [g]∈G ∨ • (X ). However, this decomposition captures only the torsion-free part of the group KG 5 This splitting in fact holds over Q [50].
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
667
4.3. Delocalization of the equivariant Chern character. The complex (4.3) of orbifold Ramond-Ramond fields can also be used to provide an explicit geometric description of the (complex) equivariant Chern character defined in Sect. 2.4. We will now explain this construction, referring the reader to ref. [17] for the technical details. Consider a complex G-bundle E over X equipped with a G-invariant hermitian metric and a G-invariant metric connection ∇ E . One can then define a closed G-invariant differential form ch(E) ∈ • (X ; C)G in the usual way by the Chern-Weil construction ch(E) := Tr exp(−F E /2π i ) , where F E is the curvature of the connection ∇ E . It represents a cohomology class [ch(E)] ∈ H• (X ; C)G in the fixed point subring of the action of G as automorphisms of H• (X ; C). By using the definition of the homomorphisms (2.9), with Q substituted by C and H = e, one can establish the equality [ch(E)] = cheX ([E]) . Let C < G be a cyclic subgroup, and define the cohomology class Z G (C) ∼ [ch(g, E)] ∈ H• X C ; C = H • (X C ; C) Z G (C) , d = H• X C /Z G (C) ; C ∼ represented by ch(g, E) := Tr γ (g) exp(−FCE /2π i ) , where g is a generator of C, FCE is the restriction of the invariant curvature two-form F E to the fixed point subspace X C , and γ is a representation of C on the fibres of the restriction bundle E| X C which is an N G (C)-bundle over X C . The character χC naturally identifies R(C) ⊗ C with the C-vector space of class functions C → C. By using the splitting (4.4) for complex Bredon cohomology, one can then show that chCX ([E]) (g) = [ch(g, E)] up to the restriction homomorphism R(C) ⊗ C → C C (G/C) of rings with kernel the ideal of elements whose characters vanish on all generators of C. Using Eq. (2.10) we can then define the map
Z (g) even X g ; C G chC : Vect C G (X ) −→ [g]∈G ∨
from complex G-bundles E → X given by chC (E) = Tr γ (g) exp(−FgE /2π i ) . [g]∈G ∨
(4.6)
668
R. J. Szabo, A. Valentino
At the level of equivariant K-theory, from Theorem 2.5 it follows that this map induces an isomorphism • ≈ • chC : KG (X ) ⊗ C − → HG X ; C ⊗ π−• K G (−) , (4.7) where we have used the splitting (4.5). The map (4.6) coincides with the equivariant Chern character defined in ref. [5].
4.4. Wess-Zumino pairings. We now have all the necessary ingredients to define a coupling of the Ramond-Ramond fields to a D-brane in the orbifold [X/G]. In this section we will only consider Ramond-Ramond fields which are topologically trivial, i.e., elements of the differential complex (4.3), and use the delocalized cohomology theory above by working throughout with complex coefficients. Under these conditions we can straightforwardly make contact with existing examples in the physics literature and write down their appropriate generalizations. To this aim, we introduce the bilinear product ∧G : •G (X ; R) ⊗ •G (X ; R) −→ •G (X ; R) defined on ω = [g]∈G ∨ ω[g] and η = [g]∈G ∨ η[g] by
ω ∧G η :=
ω[g] ∧g η[g] ,
(4.8)
[g]∈G ∨
where ∧g = ∧ is the usual exterior product on • (X g ; R). There is also an integration
G X
: •G (X ; R) −→ R.
If ω ∈ •G (X ; R) then we set
G X
1 ω := | G∨ |
[g]∈G ∨
X g /Z G (g)
ω[g] .
G The normalization ensures that X ω = X ω when G acts trivially on X and ω ∈ • (X ; R) is “diagonal” in R(G) ⊗ • (X ; R). Suppose now that f : W → X is the smooth immersed worldvolume cycle of a wrapped D-brane state (W, E, f ) ∈ DG (X ) in the orbifold [X/G], i.e., W is a Gspinc manifold equipped with a G-bundle E → W and an invariant connection ∇ E on E. We define the Wess-Zumino pairing WZ : DG (X ) × •G (X ; C) −→ C between such D-branes and Ramond-Ramond fields as G C˜ ∧G chC (E) ∧G R(W, f ), WZ ((W, E, f ) , C) = W
(4.9)
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
669
where C˜ = f ∗ C is the pullback along f : W → X of the total Ramond-Ramond field C= C[g] [g]∈G ∨
and the equivariant Chern character is given by Eq. (4.6) with γ giving the action of G on the Chan-Paton factors of the D-brane. The closed worldvolume form R(W, f ) ∈ even G,cl (W ; C) represents a complex Bredon cohomology class which accounts for gravitational corrections due to curvature in the spacetime X and depends only on the bordism class of (W, f ). It will be constructed explicitly in Sect. 4.7 below in terms of the geometry of the immersed cycle f : W → X and of the G-bundle ν → W given by ν = = ν(W ; f ) = = f ∗ (TX ) ⊕ TW .
(4.10)
It is easily seen that, modulo the curvature contribution R(W, f ), the very natural expression (4.9) reduces to the usual Wess-Zumino coupling of topologically trivial Ramond-Ramond fields to D-branes in the case G = e. But even if a group G = e acts trivially on the brane worldvolume W (or on the spacetime X ), there can still be additional contributions to the usual Ramond-Ramond coupling if E is a non-trivial G-bundle. This is the situation, for instance, for fractional D-branes G (W, E, f ) ∈ Dfrac (X )
placed at orbifold singularities. In this case, we may use the isotopical decomposition (3.3) of the Chan-Paton bundle along with Eq. (3.2). Then the Wess-Zumino pairing (4.9) descends to a pairing G WZfrac : Dfrac (X ) × •G (X ; C) −→ C
with the additive subcategory of fractional branes at orbifold singularities given by ⎞ ⎛
⎝ 1 C˜ [g] ∧ ch E [g] χγ (g) ⎠ ∧ R(W, f ), WZfrac ((W, E, f ) , C) = | G∨ | W ∨ [g]∈G
(4.11) where χγ : G → C is the character of the representation γ and R(W, f ) ∈ even cl (W ; C). One can immediately read off from the Wess-Zumino action (4.11) the Ramond-Ramond charges of D0-branes, and the state corresponding to the representation γ has (fractional) charge Q [g] γ =
χγ (g) | G∨ | (1)
with respect to the twisted Ramond-Ramond one-form field C[g] . These charges agree with both those of an open string disk amplitude computation and a boundary state analysis for fractional D0-branes [21]. Our general formula (4.9) includes also the natural extension to the Ramond-Ramond couplings of regular D-branes which move freely in the bulk of X under the action of G, as well as to other non-BPS D-brane states such as truncated branes.
670
R. J. Szabo, A. Valentino
4.5. Linear orbifolds. We will now “test” our definition (4.9) on the class of examples considered in Sect. 3.3. These are flat orbifolds for which there are no non-trivial curvature contributions, i.e., R(W, f ) = 1. Let us specialize to the case of cyclic orbifolds having twist group G = Zn with n ≥ d. In this case, as Zn is an abelian group, one has Z∨ n = Zn (setwise) and we can label the non-trivial twisted sectors of the orbifold string theory on X by k = 1, . . . , n − 1. The untwisted sector is labelled by k = 0. We take a generator g of Zn whose action on V ∼ = Cd is given by g · z 1 , . . . , z d := ζ a1 z 1 , . . . , ζ ad z d , where ζ = exp(2π i /n) and a1 , . . . , ad are integers satisfying a1 + · · · + ad ≡ 0 mod n.6 In this case the action of any element in Zn has only one fixed point, an orbifold singularity at the origin (0, . . . , 0). Hence for any g = e one has Xg ∼ = R p,1 and the differential complex (4.3) of orbifold Ramond-Ramond fields is given by n−1 • • • p,1 R ;R . Zn (X ; R) = (X ; R) ⊕ k=1 n Consider now a D-brane (W, E, f ) ∈ DZ frac (X ) with worldvolume cycle f (W ) ⊂ R p,1 placed at the orbifold singularity, i.e., f : W → R p,1 × (0, . . . , 0) ⊂ X . Let the generator g act on the fibres of the Chan-Paton bundle E → W in the n-dimensional regular representation γ (g)i j = ζ i δi j . The action on worldvolume fermion fields is determined by a lift Zˆ n acting on the spinor bundle S → W . Then the pairing (4.9) contains the following terms. First of all, we have the coupling of the untwisted Ramond-Ramond fields to W given by C˜ ∧ Tr exp(−F E /2π i ) ,
W
which is just the usual Wess-Zumino coupling and hence the Ramond-Ramond charge of a regular (bulk) brane is 1 as expected. Then there are the contributions from the twisted sectors, which by recalling Eq. (3.2) are given by the expression W
n−1 1 ˜ Ck ∧ Tr γ (g k ) exp(−F E /2π i ) , n k=1
where g k is an element of Zn of order k. Since γ (g k )ii = ζ ik , the coupling in this case is determined by a discrete Fourier transform of the fields C˜ k over the group Zn . The brane associated with the i th irreducible representation of Zn has charge ζ ik /n with respect to the Ramond-Ramond field in the k th twisted sector. For d = 2 and d = 3 this pairing agrees with and uniformizes the gauge field couplings computed in refs. [26 and 27], respectively. 6 Both the requirement that the representation V be complex and the form of the G-action are physical inputs ensuring that the closed string background X preserves a sufficient amount of supersymmetry after orbifolding.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
671
4.6. An equivariant Riemann-Roch formula. Let X, W be smooth compact G-manifolds, and f : W → X a smooth proper G-map. If we want to make sense of the equations of motion for the Ramond-Ramond field C, which is a quantity defined on the spacetime X , then we need to pushforward classes defined on the brane worldvolume W to classes defined on the spacetime. This will enable the construction of RamondRamond currents in Sect. 6 induced by the background and D-branes which appear as source terms in the Ramond-Ramond field equations. Some technical details of the constructions below are provided in Appendix C. Consider first the non-equivariant case G = e. Let ν → W be the Z2 -graded bundle (4.10), i.e., the KO-theory class of ν is the virtual bundle [ν] = f ∗ [TX ] − [TW ] ∈ KO0 (W ). We assume that ν is even-dimensional and endowed with a spinc structure (this is automatic if both X and W are spinc ). Then, as reviewed in Appendix C, one can define the Gysin homomorphism in ordinary K-theory f !K : K• (W ) −→ K• (X ). Using the orientations on X and W one has Poincaré duality in ordinary cohomology, inducing a Gysin homomorphism f !H : H• (W ; Q) −→ H• (X ; Q) , where here we consider the Z2 -grading given by even and odd degree. The pushforward homomorphisms in K-theory and in cohomology, under the conditions stated above, are related by the Riemann-Roch theorem which states that ch f !K (ξ ) = f !H ch(ξ ) Todd(ν)−1 (4.12) for any class ξ in K• (W ). Here Todd(E) ∈ even cl (W ; C) denotes the Todd genus characteristic class of a hermitian vector bundle E over W , whose Chern-Weil representative is F E /2π i
, Todd(E) = det tanh F E /2π i where F E is the curvature of a hermitian connection ∇ E on E. The Todd class of the Z2 -graded bundle (4.10) can be computed by using multiplicativity, naturality and invertibility to get Todd(ν) = f ∗ Todd(TX )/Todd(TW ). Thus the Chern character does not commute with the Gysin pushforward maps, and the defect in the commutation relation is precisely the Todd genus of the bundle ν. This “twisting” by the bundle ν over the D-brane contributes in a crucial way to the Ramond-Ramond current in the non-equivariant case [19,49,54]. Let us now attempt to find an equivariant version of the Riemann-Roch theorem. As the morphism f : W → X is G-equivariant, the Z2 -graded bundle ν is itself a G-bundle with even G-action. We assume that ν is KG -oriented. This requirement is just the Freed-Witten anomaly cancellation condition [30] in this case, generalized to global worldsheet anomalies for D-branes represented by generic G-equivariant K-cycles. It enables, analogously to the non-equivariant case, the construction of an equivariant Gysin homomorphism • • f !KG : KG (W ) −→ KG (X ).
(4.13)
672
R. J. Szabo, A. Valentino
We will demonstrate that, under some very special conditions, one can construct a complex Bredon cohomology class which is analogous to the Todd genus and which plays the role of the equivariant commutativity defect as above. Let us suppose that the G-action has the property that for any element g ∈ G, the N G (g)-bundle ν g = ν(W ; f )g → W g is the Z2 -graded bundle ν g = f ∗ |W g (TX g ) ⊕ TW g over the immersion f |W g : W g → X g with a Z G (g)-invariant spinc structure. Note that these are highly non-trivial conditions, because for an arbitrary G-bundle E → X one is not even guaranteed in general that E g → X g is a vector bundle, as the dimension of the fibre may jump from point to point. As a simple example of what can happen,7 let X = R and G = R+ be the group of positive reals under multiplication. Consider the G-bundle X × V → X given by projection onto the first factor, where V is a finite-dimensional real vector space and the G-action is
g · (x, v) = x , g x v for all g ∈ G. For any g = 1, (X × V )g is not a fibre bundle over X g = X , as the G-invariant fibre space over x = 0 is V while it is the null vector over any other point. When G is a finite group, one can apply a construction due to Atiyah and Segal [5]. If E is a complex G-vector bundle over X , its restriction to the fixed point subspace X g for any g ∈ G carries a representation of the normalizer subgroup NG (g) fibrewise. We can thus decompose E| X g into a Whitney sum of sub-bundles E α = Homg (11α , E| X g ) ⊗ 11α over the eigenvalues α ∈ spec(g) ⊂ C for the action of g on the fibres of E| X g , where Homg (11α , E| X g ) is a Z G (g)-bundle over X g and 11α is the N G (g)-bundle X g × Vα with Vα the corresponding eigenspace. We define the class φg (E) = α [E α ] (4.14) α∈spec(g)
in the ordinary K-theory of X g with complex coefficients. By Schur’s lemma, every element h ∈ Z G (g) commuting with g acts as a multiple of the identity on the total space of the bundle E α , and so the class obtained in this way is Z G (g)-invariant. It follows that the map (4.14) on Vect C G (X ) induces a homomorphism • (X ) ⊗ C −→ φg : K G
• g Z (g) K (X ) ⊗ C G .
By setting φ=
φg
[g]∈G ∨
we obtain a natural isomorphism leading to the splitting [5]
Z (g) • KG K• (X g ) ⊗ C G . (X ) ⊗ C ∼ =
(4.15)
[g]∈G ∨
The equivariant Chern character (4.7) provides an isomorphism componentwise between the equivariant K-theory group (4.15) and the complex Bredon cohomology of X. 7 We are grateful to J. Figueroa-O’Farrill for suggesting this example to us.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
673
• Suppose now that the equivariant Thom class Thom G (ν) ∈ KG,cpt (ν) can be decomposed according to the splitting (4.15) in such a way that the component in any subgroup Z G (g)
• (ν g ) ⊗ C Thom ν g ∈ Kcpt
coincides with the (ordinary) Thom class of the Z2 -graded bundle ν g → W g . Under these conditions, the equivariant Gysin homomorphism (4.13) constructed in Appendix C decomposes according to the splitting f !KG = f gK , [g]∈G ∨
where f gK is the K-theoretic Gysin homomorphism associated to the smooth proper map
f W g : W g −→ X g . Define the characteristic class Todd G by
Todd ν g , Todd G (ν) :=
(4.16)
[g]∈G ∨
where Todd is the ordinary Todd genus. This class defines an element of the even degree complex Bredon cohomology of the brane worldvolume W . Under the conditions spelled out above, we can now use the equivariant Chern character (4.7) and the usual RiemannRoch theorem for each pair (W g , X g ) to prove the identity (4.17) f !HG chC (ξ ) G Todd G (ν)−1 = chC f !KG (ξ ) • (W ) ⊗ C, as all quantities involved in the formula (4.17) are for any class ξ ∈ KG compatible with the G-equivariant decompositions given above. When the geometric conditions assumed above are not met, the equivariant Todd class in the formula (4.17) should be modified by multiplying it with another equivariant characteristic class G (W ) which reflects non-trivial geometry of the normal bundles N W g to the embeddings W g ⊂ W . This should come from applying a suitable fixed point theorem to the ordinary Riemann-Roch formula (4.12), but we have not found a version which is suitable to our particular equivariant Chern character in the general case on the category of G-spaces. When f is the collapsing map X → pt, this is the content of the index theorem used in Sect. 4.7 below. The formula (4.17) is, however, directly G applicable on the category Dfrac (X ) of fractional D-branes. When G is the cyclic group Zn as in Sect. 4.5 above, one can apply the Thomasson-Nori fixed point theorem [53,61] to get
Zn (W ) =
n−1
ζ k ch
!
k=0
∨ −1 N W gk
• ∈ HZ W ; C ⊗ R(−) , n
where !
−1 N W g =
g codim(W )
l=0
(−1)l
"!
l
# NW g
∈ K0 W g .
(4.18)
674
R. J. Szabo, A. Valentino
4.7. Gravitational pairings. We will now explain how to compute the curvature contributions R(W, f ) ∈ even G,cl (W ; C) to the Wess-Zumino functional (4.9) for the brane geometries described in Sect. 4.6 above and for vanishing B-field. We derive the cancelling form for the Ramond-Ramond gauge anomaly inflow due to chiral fermions on the intersection worldvolume for families of branes using the usual descent procedure [35], which is due to curvature of the spacetime manifold X itself. For this, we must explicitly use the G-spinc structure on X . The standard mathematical intuition behind this correction is to modify the equivariant Chern character to an isometry with respect to the natural bilinear pairings on equivariant K-theory and Bredon cohomology with complex coefficients [49,54]. The natural sesquilinear pairing between two classes of complementary degrees in complex Bredon cohomology, represented by closed differential forms ω, η ∈ •G,cl G (X ; C), is given by ([ω], [η])HG := X ω ∧G η. On the other hand, the natural quadratic form on fractional branes defined by classes in equivariant K-theory, represented by complex G-vector bundles E, F → X , is the topological charge
([E] , [F])KG := µ1 [X, E ∨ ⊗ F, id X ] of Eq. (3.7) in the untwisted sector corresponding to the representation γ = 1 : C[G] → C induced from the trivial representation of G. This quantity agrees with the natural intersection form on boundary states computed as the G-invariant Witten index over open string states suspended between D-branes [51], which counts the difference between the number of positive and negative chirality Ramond ground states and hence computes the required chiral fermion anomaly. The two bilinear forms are related through the local index theorem which provides a formula for Index1 [D / EX ] in terms of integrals of characteristic forms over the various singular strata of the orbifold [X/G]. It reads [17] 2 dg Index1 [D / EX ] = ch(g, E) ∧ Todd (TX g ) |g| X g /Z G (g) [g]∈G ∨ STr S ((g)) $ ∧ !
, • NXg det (1 − N (g)) det 1 − N (g) exp(−F N X g /2π i ) where dg = dim(X ) − dim(X g ), |g| is the order of the element g ∈ G, and N (g) denotes the action of g ∈ G on the fibres of the normal bundle N X g to X g in X . The determinant is taken over the normal bundle N X g for each g ∈ implemented by a Berezin!G, • Grassmann integration over the exterior algebra bundle N X g . The form (g), regarded ! as an element of • N X g under the symbol map, is the action of g on the fibres of the spinor bundle S| X g , and STr S is the supertrace over the endomorphism bundle of S. This formula can be thought of as an equivariant localization of the usual AtiyahSinger index density onto the submanifolds X g ⊂ X of fixed points of the G-action on X , with the determinants reflecting the “Euler class” contributions from the non-trivial normal bundles to X g . This is the anticipated physical result arising from the closed string twisted Witten index, computed as the partition function on the cylinder with the twisted boundary conditions (4.2). Supersymmetry localizes the computation onto zero modes of the string fields which are constant maps to the submanifolds X h ⊂ X , while modular invariance requires a (weighted) sum over all twisted sectors. The index can be
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
675
rewritten using the equivariant characteristic classes defined as in Eqs. (4.6) and (4.16) to get G X / E] = chC (E) ∧G Todd G (TX ) ∧G G (X ), (4.19) Index1 [D X
where we have used (TX G (X ) :=
[g]∈G ∨
2 dg
×
!•
NXg
)g
= TX g and the element of •G,cl (X ; C) given by
|g| | G∨ | $
STr S ((g))
det (1 − N (g)) det 1 − N (g) exp(−F N X g /2π i )
(4.20)
defines a characteristic class in the complex Bredon cohomology of X . Note that the integrands of Eq. (4.20) are formally similar to the Chern characters of the virtual bundles (4.18) above. By using multiplicativity of the equivariant Chern character (4.6) to write
chC E ∨ ⊗ F = chC (E) ∧G chC (F) , it follows from Eq. (4.19) that the map √ (4.7) can be turned into an isometry by “twisting” it with the closed differential form Todd G (TX ) ∧G G (X ), which when pulled back along f : W → X gives the required anomaly cancelling form on the brane worldvolume. This should then be combined with the correction Todd G (ν)−1 contributed by the Z2 -graded bundle (4.10) to the Riemann-Roch formula (4.17). Then under the various conditions spelled out in Sect. 4.6 above, the required map R in Eq. (4.9) from G-spinc bordism classes [(W, f )] to even G,cl (W ; C) is given by % √ G (X ) f ∗ Todd G (TX ) ∧G G (X ) ∗ = Todd G (TW ) ∧G f . R(W, f ) = Todd G (ν) Todd G (TX ) (4.21) The main new ingredient in this formula is the contribution from the fixed point submanifolds X g ⊂ X , particularly their normal bundle characteristic classes (4.20). This corrects previous topologically trivial, flat space formulas, even for G-fixed worldvolumes W (see ref. [34] for example). Note that when the G-action on X is trivial, one has Todd G (TX ) = Todd(TX ) and G (X ) is constant. 5. Orbifold Differential K-Theory The main drawback of the delocalized theory of the previous section is that it cannot incorporate the interesting effects of torsion, which have been one of the driving forces behind the K-theory description of D-branes and Ramond-Ramond fluxes, and as such • (X ). In this it is desirable to have a description which utilizes the full R(G)-module KG section we will develop an extension of differential K-theory as defined in ref. [37] to incorporate the case of a G-manifold. These are the groups needed to extend the analysis of the previous section to topologically non-trivial, real-valued Ramond-Ramond fields. While we do not have a formal proof that this is a proper definition of an equivariant
676
R. J. Szabo, A. Valentino
differential cohomology theory, we will see that it matches exactly with expectations from string theory on orbifolds and also has the correct limiting properties. For this reason we dub the theory that we define ‘orbifold’ differential K-theory, deferring the terminology ‘equivariant’ to a more thorough treatment of our model (we discuss this in more detail in Sect. 5.4 below). In the following we will spell out the definition of differential K-theory groups. The crux of the extensions of these definitions to the equivariant setting will be explicit constructions of the exact sequences they are described by, which are important for physical considerations. We will determine concrete realizations of the various morphisms involved, which are given in a general but abstract framework in ref. [37].8 See refs. [28,32] for an introduction to differential cohomology theories and their applications in physics. 5.1. Differential cohomology theories. Differential K-theory of a manifold is an enrichment of its K-theory, which encodes global topological information, with local geometric information contained in the de Rham complex. Consider a (generalized) cohomology theory E• defined on the category of smooth manifolds X along with a canonical map ϕ : E• (X ) −→ H(X ; R ⊗ π−• E)•
(5.1)
which induces an isomorphism E• (X ) ⊗ R ∼ = H(X ; R ⊗ π−• E)• , i.e., the image of ϕ is a full lattice and its kernel is the torsion subgroup of E• (X ). Then one can define differential E-theory as the cohomology theory Eˇ • which lifts E• via the pullback square / cl (−; R ⊗ π−• E)• ,
Eˇ • (−) E• (−)
ϕ
(5.2)
/ H(−; R ⊗ π−• E)•
where cl (X ; R ⊗ π−• E)q denotes the real vector space of closed E• (pt; R)-valued differential forms ω on X of total degree q, and the right vertical map in the commutative diagram (5.2) is given by sending ω to its de Rham cohomology class [ω]dR . A class in Eˇ q (X ) is given by a pair (ξ, ω), with ξ ∈ Eq (X ) such that ϕ(ξ ) = [ω]dR ,
(5.3)
together with an isomorphism that realizes the equality (5.3) explicitly in H(X ; R ⊗ π−• E)q . In their foundational paper [37] Hopkins and Singer define the differential E-theory associated to any generalized cohomology theory E• , and prove its naturality and homotopy properties. This is done by generalizing the concept of function space in algebraic topology, which can be used to define the cohomology of a space, to that of differential function space, where here the term “differential” typically means something different 8 An explicit proof of these exact sequences has been given recently in ref. [18] using the geometric description of differential K-cocyles in terms of bundles with connection. Our proof is more general, but less geometric, as it exploits the realization in terms of maps to classifying spaces.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
677
from differentiable or smooth. Because of this, the differential E-groups are defined in an abstract way and are difficult to realize explicitly. An explicit construction for differential K-theory is given in ref. [37]. In the following we will go through this construction in some detail. This will be our starting point to give a definition of the differential coho• , which will reduce to ordinary mology theory associated to equivariant K-theory KG differential K-theory in the case where the group G is the trivial group. The validity of our definition will be confirmed by explicit construction of the exact sequences, that will also be important in our later physical applications.
5.2. Differential K-theory. Throughout X will denote a smooth manifold. Let Fred be the algebra of Fredholm operators on a separable Hilbert space. Recall that Fred is a classifying space for complex K-theory through the isomorphism Index(−)
[X, Fred] −−−−−→ K0 (X ) which associates to any map f : X → Fred the index bundle of f in K0 (X ). Let u ∈ Z even (Fred; R) be a cocycle of even degree which represents the Chern character of the universal bundle. Then for any map f : X → Fred representing a complex vector bundle E → X , the pullback f ∗ u is a representative of ch(E) in Heven (X ; R). ˇ 0 (X ) is defined to be the set of triples (c, h, ω), The differential K-theory group K where c : X → Fred, ω is a closed differential form in even cl (X ; R), and h is a cochain in C even−1 (X ; R) satisfying δh = ω − c∗ u.
(5.4)
The cochain h in Eq. (5.4) is precisely the isomorphism refered to in Sect. 5.1 above, which is invisible in the cohomology groups, and in this equation the closed differential form ω is regarded as a cochain under the de Rham map ω → (−) ω. Two triples (c0 , h 0 , ω0 ) and (c1 , h 1 , ω1 ) are equivalent if there exists a triple (c, h, ω) on X × [0, 1], with ω = ω(t) constant along t ∈ [0, 1], such that
and (c, h, ω) t=1 = = (c1 , h 1 , ω1 ). (5.5) (c, h, ω) t=0 = = (c0 , h 0 , ω0 ) The equivalence (5.5) can be rephrased [28] by requiring that there exists a map F : X × [0, 1] −→ Fred and a differential form σ ∈ even−2 (X ; R) such that
F t=0 = c0 ,
= c1 , F t=1
ω1 = ω0 , h 1 = h 0 + π∗ F ∗ u + dσ,
(5.6)
where π : X × [0, 1] → X is the natural projection. The relations (5.6) say that c0 and c1 are homotopic maps, hence they represent the same class in K0 (X ), and that the
678
R. J. Szabo, A. Valentino
cochains h 0 and h 1 are related by the homotopy that connects the maps c0 and c1 . We also see that the closed form ω completely characterizes the triple (c, h, ω). Borrowing terminology used in representing classes in the differential cohomology ˇ 2 (X ) as principal U(1)-bundles with connection, the class [c] ∈ K0 (X ) is called H the characteristic class, the closed differential form ω is called the curvature, while the cochain h is called the holonomy of the triple. From the defining property of the universal cocycle u and Eq. (5.4) it follows that ch ([c]) = [ω]dR . Thus the cohomology class represented by the curvature ω lies in the image of the (real) Chern character, which is a lattice of maximal rank inside the cohomology group with real coefficients. ˇ −1 (X ). Recall that the classifiying Let us now define the differential K-theory group K −1 space for K is the based loop space Fred. Thus we need a cocycle u −1 ∈ Z odd (Fred; R) which represents the universal odd Chern character. Consider the evaluation map ev : Fred × S1 −→ Fred. Then the cocycle u −1 is defined by u −1 = ∗ ev∗ u, where : Fred × S1 → Fred is the natural projection. In fact, u −1 can be defined as the slant product of ev∗ u with the fundamental class of the circle S1 , i.e., by integratˇ −1 (X ) is represented by a triple ing the cocycle ev∗ u along S1 . As above, a class in K (X ; R), and (c, h, ω), where c : X → Fred, ω is a closed differential form in even−1 cl even−2 h is a cocycle in C (X ; R) satisfying δh = ω − c∗ u −1 . Two triples (c0 , h 0 , ω0 ) and (c1 , h 1 , ω1 ) are equivalent if a relation like that in Eq. (5.5) holds. ˇ −n (X ) In an analogous way one can define the higher differential K-theory groups K for any positive integer n. One can prove that Bott periodicity in complex K-theory induces a periodicity in differential K-theory given by ˇ −n (X ) ∼ ˇ −n−2 (X ). K =K This periodicity enables one to define the higher differential K-theory groups in positive ˇ −n (X ) is given by degrees. The group composition law on K (c1 , h 1 , ω1 ) + (c2 , h 2 , ω2 ) := (c1 · c2 , h 1 + h 2 , ω1 + ω2 ), where the dot denotes pointwise multiplication. The identity element is given by the triple ( c , 0, 0), where throughout c denotes any map which is homotopic to the (constant) identity map. To allow for the presence of the characteristic class ω in the definition, the ˇ −n (X ) are generally infinite-dimensional. The definition of K ˇ −n (X ) abelian groups K
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
679
depends, up to homotopy type and cohomology class, on the choice of classifying space and of universal cocycle u [37]. A key property is the exact sequences which characterize the differential K-theory ˇ −n (X ) for any n ∈ Z as extensions of topological K-theory by certain groups groups K ˇ • (X ) is an extension of differential forms. In each case the differential K-theory group K of the setwise fibre product
A•K (X ) = (ξ, ω) ∈ K• (X ) × cl (X ; R ⊗ π−• K)• ch(ξ ) = [ω]dR by the torus of topologically trivial flat fields given by ˇ • (X ) −→ A•K (X ) −→ 0. 0 −→ K•−1 (X ) ⊗ R/Z −→ K
(5.7)
This will be useful below when we define equivariant differential K-theory. As in the case of topological K-theory, there are geometrical realizations of the groups ˇ 0 (X ) can be represented by a complex vector ˇ −n (X ) [28]. In particular, a class in K K bundle E → X equipped with a connection ∇ E . To the pair (E, ∇ E ) we can associate the triple ( f, η, ω), where f : X → BU is a map which classifies the bundle E, ω = ch(∇ E ) is a Chern-Weil representative of the Chern character of [E], and η is a Chern-Simons form such that dη = f ∗ ω BU − ω with ω BU = ch(∇ BU ) the Chern character form of the universal bundle E BU → BU with respect to the universal connection ∇ BU on E BU . In the following we will define abelian groups that can be thought of as a natural generalization of the differential K-theory of a manifold X acted upon by a (finite) group G. In this case one cannot employ the powerful machinery developed in ref. [37], • (X ) is not a cohomology theory defined on the cateas the equivariant K-theory KG gory of manifolds. Instead, we will take as our starting point the explicit definition of ˇ −n (X ) given above, and naturally generalize it to groups K ˇ −n (X ) which the groups K G accommodate the action of the group G in such a way that when G = e is trivial, one ˇ −n (X ) ∼ ˇ −n (X ). has K =K G 5.3. Orbifold differential forms. We want to generalize the commutative diagram (5.2) to the case in which our underlying cohomology theory E• (X ) is the equivariant • (X ). We first need a homomorphism ϕ from equivariant K-theory to a K-theory KG target cohomology theory which induces an isomorphism when tensored over the reals. For this, we will use the Chern character constructed in Sect. 2.4 with the target cohomology theory given by Bredon cohomology. Then we need a refinement of this cohomology which reduces to the de Rham complex when the group G is trivial. This complex may be thought of as the complex of differential forms on the orbifold X/G. For this purpose, we will use the differential complex (•G (X ; R), dG ) defined in Sect. 4.1. Using the delocalization formula (4.5) one shows that this complex is a refinement for Bredon cohomology with real coefficients, in the case when G is a finite group. It comes equipped with a natural product defined in Eq. (4.8). As a refinement for Bredon cohomology, the complex •G (X ; R) gives a well-defined map
ω −→ [ω]G−dR ∈ H •G (X ; R) , dG ∼ = H•G X ; R ⊗ R(−) and reduces to the usual de Rham complex of differential forms in the case G = e. There is an alternative complex one could construct which is “manifestly” equivariant, in the sense that its functoriality property over the category of groups is transparent.
680
R. J. Szabo, A. Valentino
It can also be generalized to the case in which G is an infinite discrete group. However, it is not evident how to define a ring structure on this complex, and its physical relation to Ramond-Ramond fields is not clear. We include its definition here for completeness.9 See Appendix A and ref. [23] for the relevant definitions concerning modules over a functor category and their tensor products. Starting from the real representation ring functor R(−) = R ⊗ R(−) over the orbit category Or(G), there is a natural map of real vector spaces
R(−) ⊗ROr(G) C • (X ; R) −→ HomROr(G) C • (X ; R) , R(−) , (5.8) where C • (X ; R) is the left ROr(G)-module obtained by dualizing the functor C • (X ; R) := R ⊗ C • (X ) defined in Sect. 2.2. Note that both C • (X ; R) and R(−), being contravariant functors, are right ROr(G)-modules. The map (5.8) is given on decomposable elements as λ ⊗ f −→ (σ → f (σ )∗ (λ)) and it is an isomorphism of real vector spaces.10 Define the differential complex
•G X ; R ⊗ R(−) := R(−) ⊗ROr(G) • (X ; R), where • (X ; R) is the functor Or(G) → Ab given by • (X ; R) : G/H → • (X H ; R), and with derivation dorb induced by the exterior derivative d. Since the de Rham map induces a chain homotopy equivalence of left ROr(G)-complexes C • (X ; R) → • (X ; R), there is a G-equivariant chain homotopy equivalence R(−) ⊗ROr(G) C • (X ; R) −→ R(−) ⊗ROr(G) • (X ; R). Combined with the isomorphism (5.8) we can thus conclude
H•G X ; R ⊗ R(−) ∼ = H •G (X ; R ⊗ R(−)) , dorb . If one chooses to work with this complex, then the construction of orbifold differential K-theory groups given in Sect. 5.4 below can be carried through in exactly the same way. But since the two complexes •G (X ; R) and •G (X ; R ⊗ R(−)) are in general not isomorphic, the two differential cohomology theories obtained will be generically distinct. 5.4. Orbifold differential K-groups. Having sorted out all the ingredients necessary to make sense of a generalization of the diagram (5.2), we will now define the differential ˇ −n (X ). First, let us recall some further basic facts about equivariant K-theory groups K G equivariant K-theory. Similarly to ordinary K-theory, a model for the classifying space 0 is given by the G-algebra of Fredholm operators Fred acting on a of the functor KG G separable Hilbert space which is a representation space for G in which each irreducible representation occurs with infinite multiplicity [3]. Then there is an isomorphism 0 KG (X ) ∼ = [X, Fred G ]G , 9 We are grateful to W. Lück for suggesting this construction to us. 10 In general, to have an isomorphism one has to require the G-manifold X to be cocompact and proper.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
681
where [−, −]G denotes the set of equivalence classes of G-homotopic maps, and the isomorphism is given by taking the index bundle. n , There is also a universal space VectnG , equipped with a universal G-bundle E G n such that X, VectG G corresponds to the set of isomorphism classes of n-dimensional G-vector bundles over X [46]. These spaces are constructed as follows. Let EG be the category whose objects are the elements of G and with exactly one morphism between each pair of objects. The geometric realization (or nerve) of the set of isomorphism classes in EG is, as a simplicial space, the total space of the classifying principal G-bundle E G. With Vect n (pt) the category of n-dimensional complex vector spaces V ∼ = Cn , the n universal space VectG is defined to be the geometric realization of the functor category n [EG, Vect n (pt)] (see Appendix A). The universal n-dimensional G-vector bundle E G is then defined as n G = Vect nG ×GL(n,C) Cn −→ Vect nG , E
(5.9)
where Vect nG is the geometric realization of the functor category defined as above but with Vect n (pt) replaced with the category consisting of objects V in Vect n (pt) together with an oriented basis of V . We assume sufficient regularity conditions on the infinite-dimensional spaces Fred G n . Since Fred G and the group completion BVect G are both classifying spaces and E G for equivariant K-theory, they are G-homotopic and we can thereby choose a cocycle even uG ∈ ZG (Fred G ; R)
representing the equivariant Chern character of the universal G-bundle (5.9). Generally, even (X ; R) is the subgroup of closed cocycles in the complex the group Z G even (X ; R) := CG
Z (g) C even X g ; R G
(5.10)
[g]∈G ∨
which, by the results of Sect. 4.2, is a cochain model for the Bredon cohomology group (X ; R ⊗ R(−)). The equivariant Chern character is understood to be composed Heven−1 G with the delocalizing isomorphism of Sect. 4.2. Since it is a natural homomorphism, for any G-bundle E → X classified by a G-map f : X → Fred G one has ch X ([E]) = f ∗ u G . ˇ 0 (X ) of the (global) orbifold Definition 5.1. The orbifold differential K-theory K G [X/G] is the group of triples (c, h, ω), where c : X → Fred G is a G-map, ω is an even−1 element in even (X ; R) satisfying G,cl (X ; R), and h is an element in C G δh = ω − c∗ u G .
(5.11)
Two triples (c0 , h 0 , ω0 ) and (c1 , h 1 , ω1 ) are said to be equivalent if there exists a triple (c, h, ω) on X × [0, 1], with the group G acting trivially on the interval [0, 1] and with ω constant along [0, 1], such that
(c, h, ω) t=0 = = (c0 , h 0 , ω0 ) and (c, h, ω) t=1 = = (c1 , h 1 , ω1 ).
682
R. J. Szabo, A. Valentino
In Eq. (5.11) the closed orbifold differential form ω is regarded as an orbifold cochain in the complex (5.10) by applying the de Rham map componentwise on the fixed point ˇ −n (X ) are submanifolds X g , g ∈ G. The higher orbifold differential K-theory groups K G defined analogously to those of Sect. 5.2 above. To confirm that this is an appropriate extension of the ordinary differential K-theory of X , we should show that the orbifold differential K-theory groups fit into exact sequences which reduce to those given by Eq. (5.7) when G is taken to be the trivial group. For this, we define the group
0
A0KG (X ) := (ξ, ω) ∈ KG (X ) × even G,cl (X ; R) ch X (ξ ) = [ω]G−dR . ˇ 0 (X ) satisfies the exact Theorem 5.2. The orbifold differential K-theory group K G sequence
Heven−1 X ; R ⊗ R(−) G 0 ˇG 0 −→ (X ) −→ A0KG (X ) −→ 0. (5.12) −→ K −1 ch X KG (X ) Proof. Consider the subgroup of Heven−1 (X ; R ⊗ R(−)) defined as the image of the G −1 equivariant K-theory group KG (X ) under the Chern character ch X . It consists of Bredon cohomology classes of the form [c˜∗ u −1 G ], where c˜ : X → Fred G . There is a surjective map 0 ˇG f : K (X ) −→ A0KG (X ), [(c, h, ω)] −→ ([c] , ω) ,
which is a well-defined homomorphism, i.e., it does not depend on the chosen representative of the orbifold differential K-theory class. By definition, the kernel of f consists of triples of the form ( c , h, 0). We also define the map
0 ˇG g : Heven−1 (X ) X ; R ⊗ R(−) −→ K G [h] −→ ( c , h, 0) , which is a well-defined homomorphism because the class ( c , h, 0) depends only on (X ; R ⊗ R(−)). Then by construction one the Bredon cohomology class [h] ∈ Heven−1 G has im(g) = ker( f ). The homomorphism g is not injective. To determine the kernel of g, we use the fact ˇ 0 (X ) can be represented as that the zero element in K G ( c , 0, 0) = ( c , π∗ F ∗ u G + dG σ, 0) (X ; R) (see Eq. (5.6)). To the map F we with F : X × S1 → Fred G and σ ∈ even−2 G can associate a map c˜ : X → Fred G such that F = ev ◦ (c˜ × idS1 ). This follows from the isomorphism −1 0 0 (X ) ∼ (X × S1 ) → KG (X ) , KG = ker i ∗ : KG where i is the inclusion i : X → X × pt ⊂ X × S1 . Now use the fact that at the level of (real) Bredon cohomology one has an equality
∗ π∗ c˜ × idS1 = c˜∗ ∗ ,
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
683
since the projection homomorphisms π∗ and ∗ both correspond to integration (slant product) along the S1 fibre. Then one has the identity # " π∗ F ∗ u G = = π∗ (c˜ × idS1 )∗ ev∗ u G = = c˜∗ ∗ ev∗ u G = = c˜∗ u −1 G . −1 It follows that ker(g) is exactly the group ch X (KG (X )), and putting everything together we arrive at Eq. (5.12).
−1 −1 X ; R ⊗ R(−) /ch X KG (X ) ∼ (X ) ⊗ R/Z is called the The torus Heven−1 = KG G group of topologically trivial flat fields (or of “orbifold Wilson lines”). We can rewrite the sequence (5.12) in various illuminating ways. Consider the characteristic class map 0 0 ˇG (X ) −→ KG (X ) f cc : K [(c, h, ω)] −→ [c]
and the map 0 ˇG gcc : even−1 (X ; R) −→ K (X ) G h −→ ( c , h, dG h) .
Let even−1 (X ; R) be the subgroup of elements in even−1 KG G,cl (X ; R) whose Bredon −1 cohomology class lies in ch X (KG (X )). Then by using arguments similar to those used in arriving at the sequence (5.12), one finds the Corollary 5.3 (Characteristic class exact sequence). The orbifold differential K-theˇ 0 (X ) satisfies the exact sequence ory group K G 0 −→
(X ; R) even−1 G even−1 (X ; R) KG
0 0 ˇG −→ K (X ) −→ KG (X ) −→ 0.
(5.13)
The quotient space of orbifold differential forms in the exact sequence (5.13) is called the group of topologically trivial fields. An element of this group is a globally defined (and hence topologically trivial) gauge potential on the orbifold X/G up to large (quantized) gauge transformations, with ω the corresponding field strength. Finally, consider the field strength map 0 ˇG f fs : K (X ) −→ even G,cl (X ; R) [(c, h, ω)] −→ ω.
(5.14)
The kernel of the homomorphism f fs is the group which classifies the flat fields (which −1 (X ; R/Z). This group will are not necessarily topologically trivial) and is denoted KG be described in more detail in the next section, where we shall also conjecture an essen−1 tially purely algebraic definition of KG (X ; R/Z) which explains the notation. In any case, we have the Corollary 5.4 (Field strength exact sequence). The orbifold differential K-theory group ˇ 0 (X ) satisfies the exact sequence K G −1 0 ˇG 0 −→ KG (X ; R/Z) −→ K (X ) −→ even KG (X ; R) −→ 0.
(5.15)
684
R. J. Szabo, A. Valentino
Higher orbifold differential K-theory groups satisfy analogous exact sequences, with the appropriate degree shifts throughout. It is clear from our definition that one recovers the ordinary differential K-theory groups in the case of the trivial group G = e, and in this sense our orbifold differential K-theory may be regarded as its equivariant generalization. At this point we hasten to add that, although our groups are well-defined and satisfy desired properties which are useful for physical applications such as functoriality and the various exact sequences above, we have not proven that our orbifold theory is a differential cohomology theory. We have also not given a definition of what a generic orbifold (or equivariant) differential cohomology theory is. For instance, it would be ˇ • (X ). In particular, the interesting to define a ring structure and an integration on K G integration requires knowledge of a relative version of orbifold differential K-theory, which we have not developed in this paper. ˇ • (X ) reduces to the ordinary We have also investigated the possibility that the group K G • ˇ • (X )⊗ R(G) ˇ differential K-theory K (X/G) in the case of a free G-action on X , and to K in the case of a trivial group action, as one might naively expect from the analogous results for equivariant topological K-theory (the equivariant excision theorem (1.4) with N = G and Eq. (3.4), respectively) and for Bredon cohomology (Examples 2.2 and 2.3, respectively). On the contrary, these decompositions do not occur, because the corresponding isomorphisms in equivariant K-theory are estabilished via the induction maps and these usually do not lift at the “cochain level” as isomorphisms. Properties such as induction structures reflect homotopy invariance of topological cohomology groups, which is not possessed by differential cohomology groups due to their “local” dependence on the complex of differential forms. We will see an explicit example of this in the next section. With this in mind, it would be interesting then to define a suitable analog of the induction structures in an equivariant cohomology theory. These and various other interesting mathematical issues surrounding the orbifold differential K-theory groups that we have defined will not be pursued in this paper.
6. Flux Quantization of Orbifold Ramond-Ramond Fields In this final section we will argue that the orbifold differential K-theory defined in the previous section can be used to describe Ramond-Ramond fields and their flux quantization condition in orbifolds of Type II superstring theory with vanishing H -flux. To formulate the self-duality property of orbifold Ramond-Ramond fields in equivariant K-theory, one needs an appropriate equivariant version of Pontrjagin duality [32]. This appears to be a very deep and complicated problem, and is beyond the scope of the present paper. Furthermore, to generalize the pairing of Sect. 4.4 to topologically non-trivial Ramond-Ramond fields, one needs to define an integration on the orbifold differential cohomology theory defined in Sect. 5.4, and regard the Ramond-Ramond fields properly as cocycles for it. In addition, one needs a graded ring structure and an appropriate groupoid representing the orbifold differential K-theory, whose objects are the Ramond-Ramond form gauge potentials C and whose isomorphism classes are the ˇ • (X ). Lacking these ingredients, most of our analysis in gauge equivalence classes in K G this section will be essentially purely “topological”. We shall study the somewhat simpler problem of the proper K-theory quantization of orbifold Ramond-Ramond fields, in particular due to their sourcing by fractional D-branes, in terms of the formulation provided by orbifold differential K-theory.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
685
6.1. Ramond-Ramond currents. We will begin by rephrasing the relation between the D-brane charge group and the group of Ramond-Ramond fluxes “measured at infinity” in the equivariant case, which is a statement about the K-theoretic classification of Ramond-Ramond fields on a global orbifold [X/G]. For this, we invoke an argument due to Moore and Witten [52] which will suggest that the equivariant Chern character ch X constructed in Sect. 2 gives the right quantization rule for orbifold Ramond-Ramond fields. Suppose that our spacetime X is a non-compact G-manifold. Suppose further that there are D-branes present in Type II superstring theory on X/G. Their Ramondi Ramond charges are classified by the equivariant K-theory KG,cpt (X ) with compact support, where i = 0 in Type IIB theory and i = −1 in Type IIA theory. We require that the brane be a source for the equation of motion for the total Ramond-Ramond field strength ω. This means that it creates a Ramond-Ramond current J . If we require that the worldvolume W be compact in equivariant K-cycles (W, E, f ) ∈ DG (X ), then J is supported in the interior X˚ of X . Let X ∞ be the “boundary of X at infinity”, which we assume is preserved by the action of G. Then • • (X, X ). Since J is trivialized by ω in X ˚ , the D-brane charge lives in KG,cpt (X ) ∼ = KG ∞ the kernel of the natural forgetful homomorphism • • f• : KG,cpt (X ) −→ KG (X )
(6.1)
induced by the inclusion (X, ∅) → (X, X ∞ ). We denote by i : X ∞ → X the canonical inclusion. The long exact sequence for the pair (X, X ∞ ) in equivariant K-theory truncates, by Bott periodicity, to the six-term exact sequence / K0 (X, X ∞ ) G
−1 KG (X ∞ ) O i∗
f0
/ K0 (X ) G i∗
−1 KG (X ) o
f−1
−1 KG (X, X ∞ ) o
0 (X ). KG ∞
It follows that the charge groups are given by ker f0 ∼ =
−1 KG (X ∞ ) −1 ∗ i KG (X )
and
ker f−1 ∼ =
0 (X ) KG ∞
0 . i ∗ KG (X )
This formula means that the group of Type IIB (resp. Type IIA) brane charges is mea−1 0 (X )) of “orbifold Ramond-Ramond fluxes (X ∞ ) (resp. KG sured by the group KG ∞ at infinity” which cannot be extended to all of spacetime X . We may then interpret, −1 0 (X )) as the group classifying (X ) (resp. KG for arbitrary spacetimes X , the group KG Ramond-Ramond fields in the orbifold X/G which are not sourced by branes in Type IIB (resp. Type IIA) string theory. The Ramond-Ramond current can be described explicitly in the delocalized theory of Sect. 4. The Wess-Zumino pairing (4.9) between a topologically trivial, complex Ramond-Ramond potential and a D-brane represented by an equivariant K-cycle (W, E, f ) ∈ DG (X ) contributes a source term to the Ramond-Ramond equations of motion, which is the class
X ; C ⊗ R(−) [Q(W, E, f )] ∈ Heven G
686
R. J. Szabo, A. Valentino
represented by the pushforward Q(W, E, f ) = f !HG chC (E) ∧G R(W, f ) . We now use the Riemann-Roch formula (4.17) and the fact that f ∗ is right adjoint to f !HG , i.e., f !HG ◦ f ∗ = idH•G (X ; C(−)) . Using the explicit expression for the curvature form in Eq. (4.21), we can then rewrite this class as & Q(W, E, f ) = chC f !KG (E) ∧G Todd G (TX ) ∧G G (X ).
(6.2)
This is the complex Bredon cohomology class of the Ramond-Ramond current J created by the D-brane (W, E, f ). In the case G = e, the expression (6.2) reduces to the standard class of the current for Ramond-Ramond fields in Type II superstring theory on X [19,49,52,54]. There is a natural extension of the current (6.2) which allows us to formally conclude, in analogy with the non-equivariant case, that the complex Bredon cohomology class • (X ) ⊗ C representing a Ramond-Ramond field is assigned associated to a class ξ ∈ KG by the equivariant Chern √ character. If the Ramond-Ramond field is determined by a differential form C/2π Todd G (TX ) ∧G G (X ) with C ∈ •G (X ; C) and dG C = ω, then this is the class [ω(ξ )] in H•G (X ; C ⊗ R(−)) represented by the closed differential form ω(ξ ) = chC (ξ ). √ 2π Todd G (TX ) ∧G G (X )
(6.3)
This is just the anticipated flux quantization condition from orbifold differential K-theory. The appearance of the additional gravitational terms in Eq. (6.3) is inconsequential to this identification. Given the canonical map (5.1) in a generalized cohomology theory E• , any other map E• (X ) → H(X ; R ⊗ π−• E)• with the same properties described in Sect. 5.1 is obtained by multiplying ϕ with √ an invertible element in H(X ; R⊗π−• E)0 . In the case at hand, the characteristic class Todd G (TX ) ∧G G (X ) is an invertible closed differential form which represents this element in Heven G (X ; C⊗ R(−)). This class √ reduces to the usual gravitational correction Todd(TX ) when G acts trivially on X . We should stress that this analysis of the delocalized theory assumes the strong conditions spelled out in Sect. 4.7, which require a deep geometrical compatibility of the equivariant K-cycle (W, E, f ) with the orbifold structure of [X/G] (or else an explicit determination of the unknown characteristic class G (W ) correcting the Riemann-Roch formula as explained in Sect. 4.6). The example of the linear orbifolds considered in Sects. 3.3 and 4.5, and in Section 6.2 below, is simple enough to satisfy these conditions. It would be very interesting to find a geometrically non-trivial explicit example to test these requirements on. In any case, the results above suggest that the orbifold differential K-theory (or more precisely a complex version of it) defined in the previous section is the natural framework in which to describe topologically non-trivial Ramond-Ramond fields on orbifolds. It would be highly desirable to determine the correct generalization ˇ • (X ) of the previous section, of Eq. (6.2) to the orbifold differential K-theory group K G and thereby extending the delocalized Ramond-Ramond currents to include effects such as torsion.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
687
6.2. Linear orbifolds. To understand certain aspects of the orbifold differential K-theory groups, it is instructive to study the K-theory classification of Ramond-Ramond fields on the linear orbifolds considered in Sects. 3.3 and 4.5. Since the C-linear G-module 0 V is equivariantly contractible, one has Hodd G (V ; R ⊗ R(−)) = 0 and KG (V ) = R(G). From Theorem 5.2 it then follows that
0
ˇG (V ) ∼ K = A0KG (X ) ∼ = (γ , ω) ∈ R(G) × even G,cl (V ; R) ch G/G (γ ) = [ω]G−dR . Since the equivariant Chern character ch G/H : R(H ) → R(H ) for H ≤ G is the identity map, the setwise fibre product truncates to the lattice of quantized orbifold differential forms and one has 0 ˇG (V ) = even K KG (V ; R).
(6.4)
This is the group of Type IIA Ramond-Ramond form potentials on V . It naturally contains those fields which trivialize the Ramond-Ramond currents sourced by the stable fractional D0-branes of the Type IIA theory, corresponding to characteristic classes [c] in the representation ring R(G) as explained in Sect. 3.3. This can be explicitly described as an extension of the group of topologically trivial Ramond-Ramond fields C of odd degree by the equivariant K-theory of V , as implied by Corollary 5.3. Since V is connected and G-contractible, one has 0G,cl (V ; R) = R ⊗ R(G) and the group (6.4) has a natural splitting 0 ˇG (V ) K
= R(G) ⊕
d
2k G,cl (V ; R)
.
(6.5)
k=1
Any closed orbifold form ω on V of positive degree is exact, ω = dG C, with the gauge invariance C → C + dG ξ . It follows that there is a natural map d
2k G,cl (V ; R) −→
k=1
odd G (V ; R) odd KG (V ; R)
which associates to the field strength ω the corresponding globally well-defined Ramond-Ramond potential C. ˇ −1 (V ) of Type IIB On the other hand, the orbifold differential K-theory group K G Ramond-Ramond fields on V can be computed by using the characteristic class exact −1 sequence (5.13) with degree shifted by −1. Using KG (V ) = 0, one finds ˇ −1 (V ) = K G
even G (V ; R) . even KG (V ; R)
(6.6)
This result reflects the fact that the Type IIB theory has no stable fractional D0-branes. Hence there is no extension and the Ramond-Ramond fields are induced solely by the closed string background. Their field strengths ω = dG C are determined entirely by the potentials C, which are globally defined differential forms of even degree. −1 (G/H ) = 0 and Note that for any G-homogeneous space G/H one has KG odd G (G/H ; R) = 0.
688
R. J. Szabo, A. Valentino
From the characteristic class exact sequence (5.13) one thus computes that the orbifold differential K-theory group 0 0 ˇG (G/H ) ∼ (G/H ) ∼ K = KG = R(H )
(6.7)
is given by the characteristic classes (of fractional D0-branes), while Theorem 5.2 (with degree shifted by −1) implies that the orbifold differential K-theory group
G/H ; R ⊗ R(−) ∼ Heven G −1 ∼ ˇ
0 KG (G/H ) = (6.8) = R(H ) ⊗ R/Z ch G/H KG (G/H ) is given by the topologically trivial flat fields. Setting H = G in Eqs. (6.7) and (6.8) shows that the differential KG -theory groups of a point generically differ from the groups (6.4) and (6.6), even though V is G-contractible. This exemplifies the G-homotopy non-invariance of the orbifold differential K-theory groups, required to capture the nonvanishing (but topologically trivial) gauge potentials on V . 6.3. Flat potentials. In Sect. 6.2 above we encountered some examples of topologically trivial Ramond-Ramond fields, corresponding to gauge equivalence classes with trivial K-theory flux [c] = 0. They are the globally defined orbifold differential forms C ∈ •G (X ; R) with the gauge symmetry C → C + ξ , where dG ξ = 0 and ξ ∈ •KG (X ; R), and field strength ω = dG C. i (X ; R/Z), The flat Ramond-Ramond fields are instead classified by the abelian group KG where i = 0 for Type IIB theory and i = −1 for Type IIA theory. In the previous section this group was defined to be the subgroup of orbifold differential K-theory with vanishing curvature. In the following we will conjecture a very natural algebraic definition of these groups which ties them somewhat more directly to equivariant K-theory groups. • (V ; R/Z) for the linear To motivate this conjecture, we first compute the groups KG orbifolds of Sect. 6.2 above, wherein the associated differential K-theory groups were determined explicitly. Using the field strength exact sequence (5.15), by definition one has −1 0 ˇG KG (V ; R/Z) ∼ (V ) → even = ker f fs : K KG (V ; R)
which from the natural isomorphism (6.4) trivially gives −1 (V ; R/Z) = 0. KG −1 Similarly, using KG (V ) = 0 one has 0 ˇ −1 (V ) → odd KG (V ; R/Z) ∼ (V ; R) . = ker f fs : K G,cl G
Using the natural isomorphism (6.6), the field strength map is f fs ([C]) = = dG C
for C ∈ even G (V ; R) ,
(6.9)
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
689
giving 0 KG (V ; R/Z) ∼ =
even G,cl (V ; R) even KG (V ; R)
.
Similarly to Eq. (6.5), there is a natural splitting of the vector space of closed orbifold differential forms given by d even 2k G,cl (V ; R) = (R(G) ⊗ R) ⊕ G,cl (V ; R) k=1
and we arrive finally at 0 KG (V ; R/Z) = R(G) ⊗ R/Z.
(6.10)
These results of course simply follow from the fact that V is G-contractible, so that every dG -closed Ramond-Ramond field is trivial, except in degree zero where the gauge equivalence classes are naturally parametrized by the twisted sectors of the string theory in Eq. (6.10). Note that both groups of flat fields (6.9) and (6.10) are unchanged by (equivariant) contraction of the G-module V to a point, as an analogous (but sim• (X ; R/Z) have at least some pler) calculation shows. This suggests that the groups KG G-homotopy invariance properties, unlike the differential KG -theory groups. This motivates the following conjectural algebraic framework for describing these groups. • (X ; R/Z) is an extension of the torus of topologWe will propose that the group KG •+1 (X ), ically trivial flat orbifold Ramond-Ramond fields by the torsion elements in KG as they have vanishing image under the equivariant Chern character ch X . The resulting group may be called the “equivariant K-theory with coefficients in R/Z”. The short exact sequence of coefficient groups 0 −→ Z −→ R −→ R/Z −→ 0 induces a long exact sequence of equivariant K-theory groups which, by Bott periodicity, truncates to the six-term exact sequence 0 (X ) KG O
/ K0 (X ; R) G
β −1 KG (X ; R/Z) o
−1 KG (X ; R) o
/ K0 (X ; R/Z) G
(6.11)
β
−1 KG (X ).
The connecting homomorphism β is a suitable variant of the usual Bockstein homomorphism. We assume that the equivariant K-theory with real coefficients is defined simply by the Z2 -graded ring
• • (X ; R) = = KG (X ) ⊗ R ∼ KG = H•G X ; R ⊗ R(−) , where we have used Theorem 2.5. The maps to real K-theory in Eq. (6.11) may then be identified with the equivariant Chern character ch X , whose image is a full lattice in the • (X ; R/Z) Bredon cohomology group H•G (X ; R ⊗ R(−)). Then the abelian group KG sits in the exact sequence β • • •+1 0 −→ KG (X ) ⊗ R/Z −→ KG (X ; R/Z) − → Tor KG (X ) −→ 0. (6.12)
690
R. J. Szabo, A. Valentino
When G = e, Eq. (6.11) is the usual Bockstein exact sequence for K-theory. In this case, an explicit geometric realization of the groups K • (X ; R/Z) in terms of bundles with connection has been given by Lott [43]. Moreover, in ref. [37] a geometric conˇ 0 (X ) in the field strength exact sequence is struction of the map K−1 (X ; R/Z) → K given. Unfortunately, no such geometrical description is immediately available for our equivariant differential K-theory, due to the lack of a Chern-Weil theory for the homotopy theoretic equivariant Chern character of Sect. 2.4. Our conjectural definition (6.12) is satisfied by the linear orbifold groups (6.9) and (6.10). • (X ; R/Z) is given, by In ref. [14] a very different definition of the groups KG defining both equivariant K-theory and cohomology using the Borel construction of Example 1.3. Then the Bockstein exact sequence (6.11) is written for the ordinary K-theory groups of the homotopy quotient X G = E G ×G X . While these groups reduce, like ours, to the usual K-theory groups of flat fields when G = e, they do not obey the exact sequence (6.12). The reason is that the equivariant Chern character used is not an isomorphism over the reals, as explained in Sect. 2.1 (see also • (X ) with respect to a cerref. [47] for a description of K• (X G ) as the completion of KG tain ideal). Moreover, an associated differential K-theory construction would directly involve differential forms on the infinite-dimensional space X G which is only homotopic to the finite-dimensional CW-complex X/G. The physical interpretation of such fields is not clear. Even in the simple case of the linear orbifolds V studied above, this description predicts an infinite set of equivariant fluxes of arbitrarily high dimension on the infinite-dimensional classifying space BG, and one must perform some non-canonical quotients in order to try to isolate the physical fluxes. The differences between the equivariant K-theory and Borel cohomology groups of V also require postulating certain effects of fractional branes on the orbifold, as in ref. [12]. In contrast, with our constructions the relation between orbifold flux groups and Bredon cohomology is much more natural, and it involves only finitely-many orbifold Ramond-Ramond fields. 6.4. Consistency conditions. As we have stressed throughout this paper, the usage of Borel cohomology as a companion to equivariant K-theory in the topological classification of D-branes and Ramond-Ramond fluxes on orbifolds has various undesirable features, most notably the fact that it involves torsion classes substantially, especially when finite group cohomology is involved. In our applications to string geometry, it is more convenient to use an equivariant cohomology theory with substantial torsion-free information. Bredon cohomology naturally accomplishes this, as instead of group cohomology the basic object is a representation ring. In fact, as we now demonstrate, the formulation of topological consistency conditions for orbifold Ramond-Ramond fields and D-branes within the framework of equivariant K-theory naturally necessitates the use of classes in Bredon cohomology. Given a Bredon cohomology class λ ∈ H•G (X ; R(−)), let us ask if there exists a Ramond-Ramond field for which ω = λ in the sense of Eq. (6.3). For this, we • (X ) of λ. As in the non-equivariant must find an equivariant K-theory lift ξ ∈ KG case [22,48], the obstructions to such a lift can be determined via a suitable spectral sequence. For equivariant K-theory the appropriate spectral sequence is described in refs. [20,50] (see also ref. [58]) using the skeletal filtration (X n ) of Sect. 1.1. We will now briefly explain the construction of this spectral sequence and its natural relationship with the obstruction theory for Ramond-Ramond fluxes in Bredon cohomology.
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
691
The E1 -term of the spectral sequence is the relative G-equivariant K-theory group p,q
= KG (X p , X p−1 )
p,q
: E1
E1
p+q
with differential d1
p,q
p+1,q
−→ E1
induced by the long exact sequence of the triple (X p+1 , X p , X p−1 ) in equivariant Kp,q theory, i.e., d1 is the composition of the map i ∗ induced by the inclusion i : (X p , ∅) → (X p , X p−1 ) with the cellular coboundary operator ofthe pair (X p+1, X p ). From Eq. (1.1) p B˚ × G/K j → X p \X p−1 , and it follows that there is a homeomorphism j∈J p
hence p,q
E1
∼ =
p+q
KG
j
q
p B˚ j × G/K j ∼ KG G/K j . =
j∈J p
j∈J p
p,q
p,q
Thus E1 = 0 for q odd, while for q even the group E1 is a direct sum of representation rings R(K j ) over all isotropy subgroups of p-cells of orbit type G/K j . It parametrizes equivariant K-theory classes defined on the p-skeleton of X which are trivial on the ( p − 1)-skeleton, and gives the supports of p-form fields and charges on the orbifold which carry no lower or higher degree fluxes. The E2 -term of the spectral sequence is the cohomology of the differential d1 . The cohomology of the cochain complex assembled from such terms is the equivariant cohomology with coefficient system R(−) on Or(G, F(X p )) for q = 0, and thus a neces• (X ) is that it define a sary condition for a p-form Ramond-Ramond field to lift to KG non-trivial cocycle in Bredon cohomology. This is consistent with Definition 5.1. The resulting Atiyah-Hirzebruch spectral sequence may then be written p,q p p+q E2 = HG X ; π−q K G (−) "⇒ KG (X ) p,q
and it lives in the first and fourth quadrants of the ( p, q)-plane. On the r th terms Er , p,q p,q the differential dr has bidegree (r, −r + 1), and Er +1 is the corresponding cohomology p,q group. Note that dr = 0 for all r even, since then either its source or its target vanishes (as Kq (C[H ]) = 0 for all q odd and H ≤ G). The E∞ -term is the inductive limit p,q
p,q
E∞ = lim Er . −→ r
p,q
p,q
For a finite-dimensional manifold X , one has Er = E∞ for all r > dim(X ) and the p+q spectral sequence converges to KG (X ). This means that the E∞ -term is the associated p+q p+q graded group of a decreasing finite filtration filt p,q KG (X ) ⊂ filt p−1,q+1 KG (X ), q q 0 ≤ p ≤ dim(X ) with KG (X ) = filt 0,q KG (X ) and p+q
filt p,q KG (X ) p+q filt p+1,q−1 KG (X )
p,q ∼ = E∞ .
(6.13)
Explicitly, if ι : X p−1 → X denotes the inclusion of the ( p − 1)-skeleton in X , then the filtration groups
p+q p+q p+q filt p,q KG (X ) := ker ι∗ : KG (X ) → KG (X p−1 )
692
R. J. Szabo, A. Valentino
consist of Ramond-Ramond fluxes where the field strength ω is a form of degree ≥ p, while the extension groups (6.13) consist of p-form fluxes with vanishing higher and lower degree fluxes. By Theorem 2.5, the equivariant Chern character ch X determines an isomorphism from the limit of the spectral sequence to its E2 -term. Thus the spectral p,q sequence collapses rationally, and so the images of all higher differentials dr , r > 2 in the spectral sequence consist of torsion classes. It follows that the next non-trivial obstruction to extending a Ramond-Ramond field is given by a “cohomology operation” p,0 p
p+3
d3 : HG X ; R(−) −→ HG X ; R(−) . (6.14) p
Thus a necessary condition for a Bredon cohomology class λ ∈ HG (X ; R(−)) to survive p,0 to E∞ is given by p,0
d3 (λ) = 0.
(6.15)
We interpret the condition (6.15) as a (partial) requirement of global worldsheet anomaly cancellation for Ramond-Ramond fluxes and, dually, the worldvolume homology cycles that they pair with. This is the orbifold generalization of the Freed-Witten condition [22,30,48] formulated in terms of obstruction classes in Bredon cohomology. It is a necessary condition for the existence of a fractional D-brane whose lowest non-vanp ishing Ramond-Ramond charge is λ ∈ HG (X ; R(−)). On the other hand, in computing the E3 -term as the cohomology of the differential (6.14), we must also take the quotient p−3,0 by the image of d3 . This means that a class λ satisfying Eq. (6.15) must be further subjected to the identifications p−3,0
λ ∼ λ + d3
(λ )
(6.16)
in E3 , for any class λ ∈ HG (X ; R(−)). We interpret the condition (6.16) as accounting for Ramond-Ramond charge violation due to D-instanton effects in the orbifold background, as explained in ref. [48] for the non-equivariant case. It means that while p−3,0 there exists a fractional brane whose lowest Ramond-Ramond charge is d3 (λ ), this D-brane is unstable. The passage from the limit (6.13) with q = 0 to the actual equivariant K-theory group p KG (X ) requires solving a typically non-trivial extension problem. Even when the spectral sequence collapses at the E2 -term, the extension can lead to important torsion corrections which distinguish the classifications of Ramond-Ramond fields based on Bredon cohomology and on equivariant K-theory. The extension problem changes the additive structure on the K-theory group of fluxes from that of the equivariant cohomology classes. This corresponds physically to non-trivial correlations between Ramond-Ramond fields of different degrees, when represented by orbifold differential forms. This torsion enhancement in equivariant K-theory compared to Bredon cohomology can shift the Dirac charge quantization condition on the Ramond-Ramond fields by fractional units and can play an important role near the orbifold points [12,13]. p,0 In the non-equivariant case G = e, the differential d3 is known to be given by the Steenrod square cohomology operation Sq3 . The vanishing condition Sq3 (λ) = 0 implies the vanishing of the third integer Stieffel-Whitney class of the Poincaré dual cycle to λ ∈ H p (X ; Z), which is just the condition guaranteeing that the corresponding brane worldvolume is a spinc submanifold of X . Unfortunately, for G = e the differential p,0
p−3
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
693
p,0
d3 is not known and the geometrical meaning of the condition (6.15) is unclear. It would be interesting to understand this requirement in terms of an obstruction theory for Bredon cohomology, analogously to the non-equivariant case, as this would open up interesting new consistency conditions for D-branes and Ramond-Ramond fields on global orbifolds [X/G]. However, we are not aware of any characteristic class theory p underlying the Bredon cohomology groups HG (X ; R(−)). Appendix A. Linear Algebra in Functor Categories In this appendix we will summarize some notions about algebra in functor categories that were used in the main text of the paper. They generalize the more commonly used concepts for modules over a ring. For further details see ref. [23]. Let R be a commutative ring, and denote the category of (left) R-modules by R−Mod. Let be a small category, i.e., its class of objects Obj( ) is a set. If C is another category, then one denotes by [ , C] the functor category of (covariant) functors → C. The objects of [ , C] are (covariant) functors φ : → C and a morphism from φ1 to φ2 is a natural transformation α : φ1 → φ2 between functors. In particular, in the main text we used the functor category R −Mod := [ , R −Mod] whose objects are called left R -modules. If one denotes with op the dual category to , then there is also the functor category Mod− R := op , R −Mod of contravariant functors → R −Mod, whose objects are called right R -modules. As an example, let G be a discrete group regarded as a category with a single object and a morphism for each element of G. A covariant functor G → R−Mod is then the same thing as a left module over the group ring R[G] of G over R. As the name itself suggests, all standard definitions from the linear algebra of modules have extensions to this more general setting. For instance, the notions of submodule, kernel, cokernel, direct sum, coproduct, etc. can be naturally defined objectwise. If M and N are R -modules, then Hom R (M, N ) is the R-module of all natural transformations M → N . This notation should not be confused with the one used for the set of all morphisms between two objects in , and usually it is clear from the context. If M is a right R -module and N is a left R -module, then one can define their categorical tensor product M ⊗ R N in the following way. It is the R-module given by first forming the direct sum M(λ) ⊗ R N (λ) F= λ∈Obj( )
and then quotienting F by the R-submodule generated by all relations of the form f ∗ (m) ⊗ n − m ⊗ f ∗ (n) = 0 ,
694
R. J. Szabo, A. Valentino
where ( f : λ → ρ) ∈ Mor( ), m ∈ M(ρ), n ∈ N (λ) and f ∗ (m) = M( f )(m), f ∗ (n) = N ( f )(n). This tensor product commutes with coproducts. If M and N are functors from to the category of vector spaces over a field K, then their tensor product is naturally equipped with the structure of a vector space over K. When is the orbit category Or(G) and R = Z, the tensor product has precise limiting cases. For an arbitrary contravariant module M and the constant covariant module N , the categorical product M ⊗ZOr(G) N is the tensor product of the right Z[G]-module M(G/e) with the constant left Z[G]-module N (G/e), M(G/e) ⊗Z[G] N (G/e). On the other hand, if the contravariant module M is constant and the covariant module N is arbitrary, then M ⊗ZOr(G) N is just N (G/G). Appendix B. Equivariant K-Homology This appendix is devoted to explaining in more detail some of the definitions and technical constructions in equivariant K-homology theories that were used in the main text to describe states of D-branes in orbifolds. B.1. Spectral definition. A natural way to define the equivariant homology theory K•G is • , which within the conby means of a spectrum for equivariant topological K-theory KG G text of Sect. 2 is a particular covariant functor Vect (−) from the orbit category Or(G) to the tensor category Spec of spectra [20]. Given any G-complex X , the corresponding pointed G-space is X + = X pt and one defines the loop spectrum X + ⊗G Vect G (−) by X +H ∧ Vect G (G/H ) ∼, (B.1) X + ⊗G Vect G (−) = G/H ∈Or(G)
where the equivalence relation ∼ is generated by the identifications f ∗ (x)∧s ∼ x∧ f ∗ (s) with ( f : G/K → G/H ) ∈ Mor(Or(G)), x ∈ X +H , and s ∈ Vect G (G/K )• . One then puts K•G (X ) := π• X + ⊗G Vect G (−) . (B.2) By using various G-homotopy equivalences of the loop spectra (B.1), one shows that this definition of equivariant K-homology comes with a natural induction structure in the sense of Sect. 1.2. For the trivial group it reduces to the ordinary K-homology K•e = K• given by the Bott spectrum BU . If G is a finite group, any finite-dimensional representation of G naturally extends to a complex representation of the group ring C[G]. Then there is an analytic assembly map ass : K•G (X ) −→ K• (C[G]) to the K-theory of the ring C[G], induced by the collapsing map X → pt and the isomorphisms K• (C[H ]) ∼ = K•G (G/H ) ∼ = R(H ) = π• Vect G (G/H ) ∼ for any subgroup H ≤ G. In the following we will give two concrete realizations of the homotopy groups (B.2).
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
695
B.2. Analytic definition. The simplest realization of the equivariant K-homology group K•G (X ) is within the framework of an equivariant version of Kasparov’s KK-theory KK •G . Let A be a G-algebra, i.e., a C ∗ -algebra A together with a group homomorphism λ : G −→ Aut(A). By a Hilbert (G, A)-module we mean a Hilbert A-module E together with a G-action given by a homomorphism : G → GL(E) such that
g (ε · a) = g ε · λg (a) (B.3) for all g ∈ G, ε ∈ E and a ∈ A. Let L(E) denote the ∗-algebra of A-linear maps T : E → E admitting an adjoint with respect to the A-valued inner product on E. The induced G-action on L(E) is given by g · T := g ◦ T ◦g−1 . Let K(E) be the subalgebra of L(E) consisting of generalized compact operators. Given a pair (A, B) of G-algebras, let DG (A, B) be the set of triples (E, φ, T ), where E is a countably generated Hilbert (G, B)-module, φ : A → L(E) is a ∗-homomorphism which commutes with the G-action,
φ λg (a) = g ◦ φ(a) ◦ g−1 (B.4) for all g ∈ G and a ∈ A, and T ∈ L(E) such that 1) [T, φ(a)] ∈ K(E) for all a ∈ A; and 2) φ(a) (T − T ∗ ), φ(a) (T 2 − 1), φ(a) (g · T − T ) ∈ K(E) for all a ∈ A and g ∈ G. The standard equivalence relations of KK-theory are now analogously defined. The set of equivalence classes in DG (A, B) defines the equivariant KK-theory groups KK •G (A, B). If X is a smooth proper G-manifold without boundary, and G acts on X by diffeomorphisms, then the algebra A = C0 (X ) of continuous functions on X vanishing at infinity is a G-algebra with automorphism λg on A given by
λg ( f )(x) := g ∗ f (x) = f g −1 · x , where g ∗ denotes the pullback of the G-action on X by left translation by g −1 ∈ G. We define K•G (X ) := KK •G (C0 (X ) , C)
(B.5)
with G acting trivially on C. The conditions (B.3) and (B.4) naturally capture the physical requirements that physical orbifold string states are G-invariant and also that the worldvolume fields on a fractional D-brane carry a “covariant representation” of the orbifold group [26]. B.3. The equivariant Dirac class. We can determine a canonical class in the abelian group (B.5) as follows. Let dim(X ) = 2n, and let G be a finite subgroup of the rotation group SO(2n).11 Let Cliff(2n) = Cliff + (2n) ⊕ Cliff − (2n) 11 Throughout the extension to K G or K −1 and dim(X ) odd can be described in the same way as in degree 1 G zero by replacing X with X × S1 .
696
R. J. Szabo, A. Valentino
denote the complex Z2 -graded euclidean Clifford algebra on n generators e1 , . . . , en with the relations ei e j + e j ei = −2 δi j . A choice of a complete G-invariant Riemannian metric on X defines a G-bundle of Clifford algebras
Cliff = Cliff TX∗ := Fr ∗ ×SO(2n) Cliff(2n) which is an associated bundle to the metric coframe bundle over X , the principal SO(2n)bundle Fr ∗ = Fr(TX∗ ) of oriented orthonormal frames on the cotangent bundle TX∗ = Fr ∗ ×SO(2n) R2n . The action of SO(2n) on the Clifford algebra is through the spin group Spin(2n) ⊂ Cliff(2n). The Lie group Spinc (2n) ⊂ Cliff(2n) is a central extension of SO(2n) by the circle group U(1), 1 −→ U(1) −→ Spinc (2n) −→ SO(2n) −→ 1 ,
(B.6)
where the quotient map in Eq. (B.6) is consistent with the double covering of SO(2n) by Spin(2n) so that Spinc (2n) = Spin(2n) ×Z2 U(1). The G-manifold X is said to have a G-spinc structure or to be KG -oriented if there is an extension of the coframe bundle to a principal Spinc (2n)-bundle Fr ∗L over X which is compatible with the G-action. The extension Fr ∗L may be regarded as a principal circle bundle over Fr ∗ ,
Gˆ G
U(1) s s ss sss s s ys / Spinc (2n) / Fr ∗ L / SO(2n)
/ X, y< y y yy yy y y / Fr ∗
where the pullback square on the bottom left defines the required covering of the orbifold group G < SO(2n) by a subgroup of the spinc group Gˆ < Spinc (2n). This lift is also necessary in order to account for the spacetime fermions present in string theory. The kernel of the homomorphism Gˆ → G is identified with the circle group U(1) < Spinc (2n) in the Clifford algebra Cliff(2n). We fix a choice of lift and hence assume that G is a discrete subgroup of the spinc group. Z2 -graded Clifford modules are likewise extended to representations of C[G] ⊗ Cliff(2n), with C[G] the group ring of G, called G-Clifford modules. The topological obstruction to the existence of a G-spinc structure on X is the equivariant third integral Stiefel-Whitney class (W3 )G (TX∗ ) ∈ H3G (X ; Z) of the cotangent bundle TX∗ in Borel cohomology. The associated bundles of half-spinors on X are defined as
± S ± = S TX∗ := Fr ∗L ×Spinc (2n) ± , (B.7) where ± are the irreducible half-spin representations of SO(2n). Since G lifts to Gˆ in the spinc group, the half-spin representations ± restrict to representations of G and
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
697
the half-spinor bundles (B.7) are G-bundles. The G-invariant Levi-Civita connection determines a connection one-form on Fr ∗ , and together with a choice of G-invariant connection one-form on the principal U(1)-bundle Fr ∗L → Fr ∗ , they determine a connection one-form on the principal Spinc (2n)-bundle Fr ∗L → X which is G-invariant. This determines an invariant connection
∇ S⊗E := ∇ S ⊗ 1 + 1 ⊗ ∇ E : C∞ X , S + ⊗ E −→ C∞ X , TX∗ ⊗ S + ⊗ E , where ∇ E is a G-invariant connection on a G-bundle E → X . The contraction given by Clifford multiplication defines a map
C : C∞ X , TX∗ ⊗ S + ⊗ E −→ C∞ X , S − ⊗ E which graded commutes with the G-action, and the G-invariant spinc Dirac operator on X with coefficients in E is defined as the composition D / EX = C ◦ ∇ S⊗E .
(B.8)
We will view the operator (B.8) as an operator on L2 -spaces
D / EX : L2 X , S + ⊗ E −→ L2 X , S − ⊗ E . X It induces a class D / E ∈ K0G (X ) as follows. The G-algebra C0 (X ) acts on the Z2 graded G-Hilbert space E := L2 (X, S ⊗ E) by multiplication. Define the bounded
X 2 −1/2 X / E) + 1 ∈ Fred G . Then D / E is represented G-invariant operator T := D / EX (D by the G-equivariant Fredholm module (E, T ).
B.4. Geometric definition. The natural geometric description of D-branes in an orbifold space is provided by the topological version of the groups K•G (X ) due to Baum, Connes and Douglas [8,7]. This can be defined for an arbitrary discrete, countable group G on the category of proper, finite G-complexes X and proven to be isomorphic to analytic • (X ) equivariant K-homology [10]. Recall that the topological equivariant K-theory KG C • is defined by applying the Grothendieck functor K to the additive category Vect G (X) • (X ) := K • Vect C (X ) . whose objects are complex G-vector bundles over X , i.e., KG G In the homological setting, the relevant category is instead the additive category of G-equivariant K-cycles DG (X ), whose objects are triples (W, E, f ), where
(a) W is a manifold without boundary with a smooth proper cocompact G-action and G-spinc structure; (b) E is an object in Vect C G (W ); and (c) f : W → X is a G-map. Two G-equivariant K-cycles (W, E, f ) and (W , E , f ) are said to be isomorphic if there is a G-equivariant diffeomorphism h : W → W preserving the G-spinc structures on W, W such that h ∗ (E ) ∼ = E and f ◦ h = f . Define an equivalence relation ∼ on the category DG (X ) generated by the operations of
698
R. J. Szabo, A. Valentino
• Bordism: (Wi , E i , f i ) ∈ DG (X ), i = 0, 1 are bordant if there is a triple (M, E, f ) where M is a manifold with boundary ∂ M, with a smooth proper cocompact G-action and G-spinc structure, E → M is a complex G-vector bundle, and f : M → X is a G-map such that (∂ M, E|∂ M , f |∂ M ) ∼ = (W0 , E 0 , f 0 ) (−W1 , E 1 , f 1 ). Here −W1 denotes W1 with the reversed G-spinc structure; • Direct sum: If (W, E, f ) ∈ DG (X ) and E = E 0 ⊕ E 1 , then (W, E, f ) ∼ = (W, E 0 , f ) (W, E 1 , f ) ; and • Vector bundle modification: Let (W, E, f ) ∈ DG (X ) and H be an even-dimensional ' = S(H ⊕11) denote the sphere bundle of H ⊕11, G-spinc vector bundle over W . Let W ' → W . Let which is canonically a G-spinc manifold, with G-bundle projection π : W S(H ) = S(H )+ ⊕ S(H )−
' = π ∗ (S(H )+ )∨ ⊗ E denote the Z2 -graded G-bundle over W of spinors on H . Set E
', E ', ' and ' f = f ◦ π . Then W f ∈ DG (X ) is the vector bundle modification of (W, E, f ) by H . We set G G (X ) = Deven,odd (X ) ∼, K0,1 where the parity refers to the dimension of the K-cycle, which is preserved by ∼. Using the equivariant Dirac class, one can construct a homomorphism from the geo- /W metric to the analytic K-homology group. On K-cycles we define (W, E, f ) → f ∗ D E and extend linearly. This map can be used to express G-index theorems within this homological framework and it extends to give an isomorphism between the two equivariant K-homology groups [10]. (See also ref. [57] for a related construction in the non-equivariant case.) Appendix C. D-Brane charges of Equivariant K-Cycles In this appendix we will review the construction of the equivariant Gysin homomorphism and how it shows that D-brane charges on the orbifold [X/G] take values in the equivari• (X ). Let X and W be smooth compact G-manifolds, and f : W → X ant K-theory KG a smooth proper G-map. We begin by dealing with the non-equivariant setting G = e. Assume that the Z2 -graded bundle ν of Eq. (4.10) is of even rank r = 2n and endowed with a spinc structure. We will generalize the construction [49,54,62], establishing that the charge of a D-brane supported on W with Chan-Paton gauge bundle E → W in Type II superstring theory without H -flux takes values in the complex K-theory of spacetime X , to D-branes represented by generic topological K-cycles (W, E, f ), i.e., including those D-branes which are not representable as wrapping embedded cycles in X . It is based on the diagram ν∼ = UI II j II π II II $ f / X W : u u u u u κ uu uu π1 X × R2q
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
699
over the brane immersion f , which we explain momentarily. The spinc condition on the bundle ν is the appropriate generalization of the Freed-Witten anomaly cancellation condition [30] to this situation. It amounts to a choice of line bundle L → W whose first Chern class c1 (L) ∈ H2 (W ; Z) obeys c1 (L) ≡ f ∗ w2 (TX ) − w2 (TW ) mod 2, where w2 (TX ) and w2 (TW ) are the second Stiefel-Whitney classes of the tangent bundles of X and W . The set of all such K-orientations is an affine space modelled on 2 H2 (W ; Z). Consider first the usual case where f : W → X is a smoothly embedded cycle. Then the virtual bundle ν can be identified (in KO-theory) with the normal bundle to W with respect to f , which is the quotient bundle π : f ∗ (TX )/TW → W . Upon choosing a Riemannian metric on X , we can identify ν with a tubular neighbourhood U of f (W ) via a diffeomorphism from the open embedding j : U → X onto a neighbourhood of the zero section embedding W → ν. Let [π ∗ S(ν)+ , π ∗ S(ν)− ; c(v)] be the AtiyahBott-Shapiro representative of the Thom class Thom(ν), in the K-theory with compact vertical support r Kcpt (ν) := Kr (ν, ν \ W ) ,
which restricts to the Bott class u −n ∈ K−r (pt) on each fibre of ν. Here S(ν)± −→ W are the half-spinor bundles associated to ν and the morphism c(v) : π ∗ S(ν)+ → π ∗ S(ν)− is given by Clifford multiplication by the tautological section v of the bundle π ∗ ν → ν which assigns to a vector in ν the same vector in π ∗ ν. Then one can define the Gysin homomorphism in ordinary K-theory f !K : K• (W ) −→ K• (X ). It is defined as the composition of the Thom isomorphism ≈
• → Kcpt (ν) K• (W ) −
ξ −→ π ∗ (ξ ) ⊗ Thom(ν) • (ν) → K • (X ) given by with the natural “extension by zero” homomorphism j : Kcpt • • • composing K (U, U \W ) → K (X, X \W ) → K (X ), where the first map is the excision isomorphism and the second map is induced by the inclusion (X, pt) → (X, W ). For a general smooth proper map f : W → X , we use the fact that every smooth compact manifold W can be smoothly embedded in R2q for q sufficiently large to define a parametrized version that yields an embedding
κ : W −→ X × R2q , whose normal bundle is spinc . The corresponding Gysin map is a homomorphism • κ!K : K• (W ) −→ Kcpt X × R2q .
700
R. J. Szabo, A. Valentino
The Gysin homomorphism f !K : K• (W ) → K• (X ) is then defined as the composition • (X × R2q ) ∼ K • (X ) for the trivial of κ!K with the inverse Thom isomorphism Kcpt = spinc bundle π1 : X × R2q −→ X. By homotopy invariance of K-theory and functoriality for pushforward maps, the map f !K is independent of the choice of identification of the normal bundle with a tubular neighbourhood and of Whitney embedding W → R2q . Let us now consider the G-actions on W and on X . In a similar way as in ordinary K-theory, if ν is KG -oriented then one has the equivariant Thom isomorphism ≈
• • (W ) − → KG,cpt (ν) KG
ξ −→ π ∗ (ξ ) ⊗ Thom G (ν) , where the equivariant Thom class Thom G (ν) ∈ KrG,cpt (ν) is defined in the same way as above using the G-spinc structure on ν and the equivariant version of the AtiyahBott-Shapiro construction [42]. The associated Gysin homomorphism, constructed as above via a choice of G-invariant Riemannian metric on X and of G-invariant Whitney embedding W → R2q with G acting trivially on R2q , is the pushforward map • (W ) → K • (X ). This establishes that the charge of a fractional D-brane in f !KG : KG G the Type II spacetime orbifold [X/G], associated to a generic G-equivariant K-cycle • (X ) in the (W, E, f ) ∈ DG (X ) on the covering space X , takes values f !KG ([E]) ∈ KG equivariant K-theory of X . Acknowlegements. We are grateful to J. Figueroa-O’Farrill, D. Freed, J. Greenlees, J. Howie, A. Konechny, W. Lück, M. Lawson, R. Reis, P. Turner and S. Willerton for helpful discussions and correspondence. This work was supported in part by the Marie Curie Research Training Network Grant ForcesUniverse (contract no. MRTN-CT-2004-005104) from the European Community’s Sixth Framework Programme. The work of A.V. was supported by the School of Mathematical and Computer Sciences at Heriot-Watt University, and in part by the German Research Foundation (Deutsche Forschungsgemeinschaft (DFG)) through the Institutional Strategy of the University of Göttingen.
References 1. Adem, A., Ruan, Y.: Twisted orbifold K-theory. Commun. Math. Phys. 237, 533–556 (2003) 2. Asakawa, T., Sugimoto, S., Terashima, S.: D-branes, matrix theory and K-homology. J. High Energy Phys. 0203, 034 (2002) 3. Atiyah, M.F.: K-Theory NewYork: Benjamin, 1967 4. Atiyah, M.F., Segal, G.B.: Equivariant K-theory and completion. J. Diff. Geom. 3, 1–18 (1969) 5. Atiyah, M.F., Segal, G.B.: On equivariant Euler characteristics. J. Geom. Phys. 6, 399–405 (1989) 6. Baum, P., Connes, A.: Chern character for discrete groups. In: A Fête of Topology, Amsterdam: North Holland, 1987, pp. 163–232 7. Baum, P., Connes, A.: Geometric K-theory for Lie groups and foliations. Enseign. Math. 46, 3–42 (2000) 8. Baum, P., Douglas, R.G.: K-homology and index theory. Proc. Symp. Pure Math. 38, 117–173 (1982) 9. Baum, P., Brylinksi, J.-L., MacPherson, R.: Cohomologie equivariante délocalisée. C. R. Acad. Sci. Paris Ser. I Math. 300, 605–608 (1985) 10. Baum, P., Higson, N., Schick, T.: On the equivalence of geometric and analytic K-homology. Pure Appl. Math. Quart. 3, 1–24 (2007) 11. Belov, D.M., Moore, G.W.: Type II actions from 11-dimensional Chern-Simons theories. Preprint. http:// arxiv.org/abs/hep-th/0611020v1, 2006 12. Bergman, O., Gimon, E.G., Kol, B.: Strings on orbifold lines. J. High Energy Phys. 0105, 019 (2001)
Ramond-Ramond Fields, Fractional Branes and Orbifold Differential K-Theory
701
13. Bergman, O., Gimon, E.G., Sugimoto, S.: Orientifolds, RR torsion and K-theory. J. High Energy Phys. 0105, 047 (2001) 14. de Boer, J., Dijkgraaf, R., Hori, K., Keurentjes, A., Morgan, J., Morrison, D.R., Sethi, S.: Triples, fluxes and strings. Adv. Theor. Math. Phys. 4, 995–1186 (2002) 15. Block, J., Getzler, E.: Equivariant cyclic cohomology and equivariant differential forms. Ann. Sci. ENS 27, 493–527 (1994) 16. Bredon, G.E.: Equivariant Cohomology Theories. Berlin-Heidelberg-NewYork: Springer, 1967 17. Bunke, U.: Orbifold index and equivariant K-homology. Math. Ann. 339, 175–194 (2007) 18. Carey, A.L., Mickelsson, J., Wang, B.-L.: Differential twisted K-theory and applications. Preprint. http:// arxiv:0708.3114v2[math.KT], 2007 19. Cheung, Y.-K.E., Yin, Z.: Anomalies, branes and currents. Nucl. Phys. B 517, 69–91 (1998) 20. Davis, J.F., Lück, W.: Spaces over a category and assembly maps in isomorphism conjectures in K- and L-theory. K-Theory 15, 201–252 (1998) 21. Diaconescu, D.-E., Gomis, J.: Fractional branes and boundary states in orbifold theories. J. High Energy Phys. 0010, 001 (2000) 22. Diaconescu, D.-E., Moore, G.W., Witten, E.: E 8 gauge theory and a derivation of K-theory from M-theory. Adv. Theor. Math. Phys. 6, 1031–1134 (2003) 23. tom Dieck, T.: Transformation Groups. Berlin: Walter de Gruyter, 1987 24. Dixon, L.J., Harvey, J.A., Vafa, C., Witten, E.: Strings on orbifolds. Nucl. Phys. B 261, 678–686 (1985) 25. Dixon, L.J., Harvey, J.A., Vafa, C., Witten, E.: Strings on orbifolds 2. Nucl. Phys. B 274, 285–314 (1986) 26. Douglas, M.R., Moore, G.W.: D-branes, quivers and ALE instantons. Preprint. http://arxiv:hep-th/ 9603167v1, 1996 27. Douglas, M.R., Greene, B.R., Morrison, D.R.: Orbifold resolution by D-branes. Nucl. Phys. B 506, 84–106 (1997) 28. Freed, D.S.: Dirac charge quantization and generalized differential cohomology. Surv. Diff. Geom. VII, 129–194 (2000) 29. Freed, D.S., Hopkins, M.J.: On Ramond-Ramond fields and K-theory. J. High Energy Phys. 0005, 044 (2000) 30. Freed, D.S., Witten, E.: Anomalies in string theory with D-branes. Asian J. Math. 3, 819–851 (1999) 31. Freed, D.S., Hopkins, M.J., Teleman, C.: Twisted equivariant K-theory with complex coefficients. Preprint. http://arxiv:math/0206257v4, 2002 32. Freed, D.S., Moore, G.W., Segal, G.B.: The uncertainty of fluxes. Commun. Math. Phys. 271, 247–274 (2007) 33. Freed, D.S., Moore, G.W., Segal, G.B.: Heisenberg groups and noncommutative fluxes. Ann. Phys. 322, 236–285 (2007) 34. García-Compeán, H.: D-branes in orbifold singularities and equivariant K-theory. Nucl. Phys. B 557, 480–504 (1999) 35. Green, M.B., Harvey, J.A., Moore, G.W.: I-brane inflow and anomalous couplings on D-branes. Class. Quant. Grav. 14, 47–52 (1997) 36. Harvey, J.A., Moore, G.W.: Noncommutative tachyons and K-theory. J. Math. Phys. 42, 2765–2780 (2001) 37. Hopkins, M.J., Singer, I.M.: Quadratic functions in geometry, topology and M-theory. J. Diff. Geom. 70, 329–452 (2005) 38. Illman, S.: Smooth equivariant triangulations of G-manifolds for G a finite group. Math. Ann. 233, 199–220 (1978) 39. Illman, S.: The equivariant triangulation theorem for actions of compact Lie groups. Math. Ann. 262, 487–501 (1983) 40. Karoubi, M.: Equivariant K-theory of real vector spaces and real projective spaces. Topol. Appl. 122, 531–546 (2002) 41. Kriz, I., Pando Zayas, L.A., Quiroz, N.: Comments on D-branes on orbifolds and K-theory. Int. J. Mod. Phys. A23, 933–974 (2008) 42. Landweber, G.D.: Representation rings of Lie superalgebras. K-Theory 36, 115–168 (2005) 43. Lott, J.: R/Z index theory. Comm. Anal. Geom. 2, 279–311 (1994) 44. Lück, W.: Chern characters for proper equivariant homology theories and applications to K- and L-theory. J. Reine Angew. Math. 543, 193–234 (2002) 45. Lück, W.: Equivariant cohomological Chern characters. Int. J. Alg. Comp. 15, 1025–1052 (2006) 46. Lück, W., Oliver, B.: Chern characters for the equivariant K-theory of proper G-CW-complexes. Progr. Math. 196, 217–248 (2001) 47. Lück, W., Oliver, B.: The completion theorem in K-theory for proper actions of a discrete group. Topology 40, 585–616 (2001) 48. Maldacena, J.M., Moore, G.W., Seiberg, N.: D-brane instantons and K-theory charges. J. High Energy Phys. 0111, 062 (2001)
702
R. J. Szabo, A. Valentino
49. Minasian, R., Moore, G.W.: K-theory and Ramond-Ramond charge. J. High Energy Phys. 9711, 002 (1997) 50. Mislin, G., Valette, A.: Proper Group Actions and the Baum-Connes Conjecture. Basel-Boston: Birkhäuser Verlag, 2003 51. Moore, G.W., Parnachev, A.: Localized tachyons and the quantum McKay correspondence. J. High Energy Phys. 0411, 086 (2004) 52. Moore, G.W., Witten, E.: Self-duality, Ramond-Ramond fields and K-theory. J. High Energy Phys. 0005, 032 (2000) 53. Nori, M.V.: The Hirzebruch-Riemann-Roch theorem. Michigan Math. J. 48, 473–482 (2000) 54. Olsen, K., Szabo, R.J.: Constructing D-branes from K-theory. Adv. Theor. Math. Phys. 4, 889–1025 (2000) 55. Periwal, V.: D-brane charges and K-homology. J. High Energy Phys. 0007, 041 (2000) 56. Reis, R.M.G., Szabo, R.J.: Geometric K-homology of flat D-branes. Commun. Math. Phys. 266, 71–122 (2006) 57. Reis, R.M.G., Szabo, R.J., Valentino, A.: KO-homology and type I string theory. Rev. Math. Phys. 21, 1091–1143 (2009) 58. Segal, G.B.: Equivariant K-theory. Publ. Math. IHES 34, 129–151 (1968) 59. Słomi´nska, J.: On the equivariant Chern character homomorphism. Bull. Acad. Pol. Sci. 24, 909–913 (1976) 60. Szabo, R.J.: D-branes, tachyons and K-homology. Mod. Phys. Lett. A 17, 2297–2316 (2002) 61. Thomasson, R.W.: Une formule de Lefschetz en K-theories equivariante algebraique. Duke Math. J. 68, 447–462 (1992) 62. Witten, E.: D-branes and K-theory. J. High Energy Phys. 9812, 019 (1998) Communicated by N.A. Nekrasov
Commun. Math. Phys. 294, 703–729 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0968-0
Communications in
Mathematical Physics
Non-Commutative Methods for the K-Theory of C ∗ -Algebras of Aperiodic Patterns from Cut-and-Project Systems Ian F. Putnam Department of Mathematics and Statistics, University of Victoria, Victoria, B.C., Canada V8W 3R4. E-mail:
[email protected] Received: 24 March 2008 / Accepted: 4 December 2008 Published online: 19 December 2009 – © Springer-Verlag 2009
Abstract: We investigate the C ∗ -algebras associated to aperiodic structures called model sets obtained by the cut-and-project method. These C ∗ -algebras are Morita equivalent to crossed product C ∗ -algebras obtained from dynamics on a disconnected version of the internal space. This construction may be made from more general data, which we call a hyperplane system. From a hyperplane system, others may be constructed by a process of reduction and we show how the C ∗ -algebras involved are related to each other. In particular, there are natural elements in the Kasparov KK-groups for the C ∗ -algebra of a hyperplane system and that of its reduction. The induced map on K-theory fits in a six-term exact sequence. This provides a new method of the computation of the K-theory of such C ∗ -algebras which is done completely in the setting of non-commutative geometry. 1. Introduction This paper is concerned with the study of aperiodic structures obtained by the so-called cut-and-project method, their associated C ∗ -algebras and the K-theory of these C ∗ -algebras. Such structures are used as models for physical materials called quasi-crystals (see [HG or J]). Around the time of the discovery of the first quasi-crystal [SBGC], original versions of the construction appeared in [DK1,DK2,E1,KKL1,KKL2,KrN,LS]. This was later made more axiomatic by using the concept of models sets introduced by Y. Meyer in 1972 [M1]. The situation is as follows: the physical space in which the atoms are actually lying is seen as a subspace of a larger Euclidean space, in which it is called the parallel direction or simply the physical space. Its orthogonal complement is usually called the internal space. In the large space, there is a lattice (called the reference lattice) which is irrationally related to the parallel direction, meaning that their intersection is just the origin. In the internal space, a window is chosen, a compact set which Supported in part by a grant from NSERC, Canada.
704
I. F. Putnam
is the closure of its interior. Most models describing real quasi-crystalline materials use the window obtained by the projection of the unit cube in the reference lattice to the internal space. In the 1980’s, it was realized that the only way to get atoms moving in a perfect quasi-crystal at very low temperatures was through the so-called flip-flops (see, in particular, Chap. 3 of Gratias and Katz in [HG]) or phasons (see [J]). The cut-and-project construction gives a very convenient representation of the phenomenon: by moving the window in the perpendicular space, every time a new point of the reference lattice enters the part of the large space which projects into the window, another point is expelled. By projecting into the physical space, such a move can be seen as a local jump of an atom from its prior position to a position nearby. The family of positions involved in the jumps is usually located on an affine hyperplane (at least whenever the window is obtained by the projection of the unit cube). It was realized quite early (see [Be1,BIT,BCL], for instance) that the window was homeomorphic to the canonical transversal (also called the atomic surface in quasi-crystalography [HG,J]) when endowed with a topology obtained by creating a gap on each affine hyperplane obtained by translating the hyperplanes which form the maximal faces of the window by the vectors in the lattice after projection to the internal space. The resulting space is totally disconnected in cases of interest. We present a specific example: the octagonal or Ammann-Beenker tiling. We follow the notation of [Be2]. Let e1 , e2 , e3 , e4 be the usual basis for R4 and our lattice is L = Z4 . Consider the orthonormal basis 1 1 1 1 1 1 v1 = (− √ , 0, − , ), v2 = (0, − √ , − , − ), 2 2 2 2 2 2 1 1 1 1 1 1 v3 = ( √ , 0, − , ), v4 = (0, √ , − , − ). 2 2 2 2 2 2 The physical space or parallel direction, G, is the span of {v1 , v2 } and the internal space, H , is the span of {v3 , v4 }. The subgroup L ⊂ H is the projection of L onto the space H and is generated by 1 1 1 1 1 1 π ⊥ (e1 ) = √ v3 , π ⊥ (e2 ) = √ v4 , π ⊥ (e3 ) = − v3 − v4 , π ⊥ (e4 ) = v3 − v4 . 2 2 2 2 2 2 Figure 1 shows the space H , the generators of L and also the window, W , which is the projection of the unit cube of the lattice L onto H . Observe that the edges of the window lie in the collection of hyperplanes, denoted P, consisting of Pk = span R π ⊥ (ek ), for k = 1, 2, 3, 4 and their translates under L. Beginning with a physical space G, an internal space H , a lattice L in G × H and a window W in H , we construct discrete point sets in G as follows. For any x in G × H , we consider the coset x + L, intersect it with G × W and project the result to G. It is convenient to require that the coset x + L does not intersect the boundary of G × W . Such a set is called a model set. This collection may then be completed in a natural way to obtain a compact set of uniformly discrete subsets of G with a natural action of G by translation. This is called the hull and is denoted (W ). An element of the space (W ) (or rather a finite part of it) for the octagonal tiling is shown below: The points of the model set are just the vertices. The edges are drawn as an aid to see the pattern; they are simply the edges given by the generators of L joining adjacent points in the coset x + L.
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
705
Fig. 1. Internal space for the octagonal tilings
Fig. 2. An octagonal tiling
We review the basic facts of this construction in Sect. 2. In particular, the space (W ) can be realized as a quotient of G × H˜ (W ) by a natural action of L, where H˜ (W ) is a totally disconnected version of the space H . We find it most convenient to follow the idea of [BIT] expressing H˜ (W ) as the spectrum of a commutative C ∗ -algebra generated by the characteristic function of the window and its translates by L. The C ∗ -algebra of interest is the crossed product C((W )) × G. Results of Rieffel immediately imply that this is Morita equivalent to C0 ( H˜ (W )) × L and we concentrate our attention on this C ∗ -algebra. These C ∗ -algebras and variants of them contain operators which are approximants of position and momentum operators for electrons moving in a quasi-crystal represented by these models sets. Moreover, if a self-adjoint operator in this C ∗ -algebra representing some observable of the system has totally disconnected spectrum, then the gaps in this spectrum may be labelled by the K 0 group of the C ∗ -algebra. For further discussion of these ideas, we refer the reader to [Be1,Be2,BHZ,KeP]. Thus, a main focus of
706
I. F. Putnam
Fig. 3. Flip-flop
research has been in computing the K-theory of these C ∗ -algebras. We mention the references [Be1,BCL,BS], but the most general scheme for computing K-theory, specifically designed for cut-and-project systems, is given in [FHK]. Returning to the construction of the space H˜ (W ), in the case where the window is the projection to H of the unit cube in the lattice L, there is an equivalent description in terms of the hyperplanes which form the boundary of W . In Sect. 3, we introduce the notion of a hyperplane system: a Euclidean space, H , a finitely generated subgroup, L and a collection of co-dimension one oriented, affine hyperplanes, P. With this data only, we construct the space H˜ (L , P) with a natural action of L and consider the crossed product. This provides a much more general construction for several reasons: we do not need all the data G and L, we do not need extra hypotheses such as L is dense and finally the topology is given without reference to model sets. As one consequence, it is not necessary that the space H˜ (L , P) be totally disconnected. For this reason, we define it as the spectrum of a commutative C ∗ -algebra only. In fact, we allow the possibility that our collection of hyperplanes P is empty, in which case H˜ (L , P) = H and the C ∗ -algebra is simply C0 (H ) × L. This is a reasonably familiar object; up to Morita equivalence, it is a non-commutative torus. Our generalization to hyperplane systems is actually a fairly obvious one; the justification for introducing it will come in the following sections. There is a reasonably simple, but somewhat imprecise, description of the space H˜ (L , P) as follows. It is obtained from H by removing each affine hyperplane in P and replacing it by two copies which are separated by a gap. Each copy is attached to one of the two half-spaces. Of course, there are some subtleties when two or more hyperplanes meet and since the collection of hyperplanes is dense. The example we provided of an octagonal tiling in Fig. 2 above actually corresponds to a new ‘doubled’ point from one of these affine hyperplanes. If one compares this model set to the one arising from its twin doubled point in H˜ (L , P), the two are identical except along a horizontal line passing through the middle of the picture. Observe that across the middle of the pattern there is a sequence of projected three dimensional cubes, each pair separated by either one or two squares. Moving the parameter point to its twin on the other doubled hyperplane, the change in the pattern is that all these cubes flip their orientation as shown in Fig. 3. This is usually refered to as a ‘flip-flop’. The points of the model set affected by this move arise from a lower dimensional cut-and-project tiling system which may be regarded as a reduction of the original one to the hyperplane where the pair of doubled points arise. We make this notion of the reduction of a hyperplane system concrete in Sect. 4. Given (H, L , P) and a hyperplane P in P, the reduction of (H, L , P) by P, which we denote by (H P , L P , P P ), is given as follows. As P is a hyperplane, it is the translate of codimension one subspace of H , which we denote H P . The group L P is just
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
707
L ∩ H P and the collection of hyperplanes P P is simply the intersections of P with the other elements of P (translated so that they lie in H P rather than P). In general, even if we begin with a cut-and-project system, its reductions may not arise from a cut-and-project system. We give a simple geometric condition in Theorem 4.5 for when this holds. It is satisfied, in particular, by the octagonal example we have given. Under these hypotheses, the flip-flops are be described explicitly in terms of the lower dimensional cut-and-project system in Theorem 4.6. The main goal of the paper is to show how this reduction to a hyperplane has a natural interpretation for the K-theory of the associated C ∗ -algebras. This is done in Sect. 5. Given (H, L , P) and its reduction to a hyperplane P in P, we let L P be a complimentary subgroup of L P in L; i.e. L = L P × L P . The set all translates of P under L can be indentified with P × L P . This is typically a dense subset of H . In passing to H˜ (L , P), each of these points is replaced by two copies. (As well, the space P is itself disconnected by the other hyperplanes, and we will write P˜ instead. For simplicity, we could imagine the case that there are no other hyperplanes so P˜ = P.) These two copies mean that we have two embeddings of P˜ × L P into H˜ (L , P), which we denote by i 0 and i 1 . If we endow L P with the discrete topology, these maps are continuous, but very far from proper. Let f be a continuous function of compact support on H˜ (L , P). The compositions f ◦ i 0 and f ◦ i 1 are continuous and bounded, but they are not compactly supported on P˜ × L P . This means that they lie in the multiplier algebra of C0 ( P˜ × L P ). More subtly and importantly, their difference does have compact support on P˜ × L P . Finally, this is all equivariant for the action of L on both spaces. This means that we have an element of the Kasparov group K K 0 (C0 ( H˜ (L , P))× L), C0 ( P˜ × L P ) × L), as interpreted by J. Cuntz [Cu]. Moreover, the C ∗ -algebra C0 ( P˜ × L P ) × L is Morita equivalent to that associated with the reduction, C0 ( H˜ P (L P , P P )) × L P . However, more is true. The results of [Pu] and show there is a six-term exact sequence relating the K -groups of these two C ∗ -algberas. The third C ∗ -algebra appearing in this sequence is that obtained from H, L and P , which is simply P after removing P and its L-orbit. In practical terms, this means that the K-theory of our C ∗ -algebra arising from (H, L , P) may be computed from that of (H P , L P , P P ) and (H, L , P ). The former is simpler because it arises from a lower dimensional hyperplane system and the latter is simpler because it involves fewer hyperplanes. It is important to note in this computation that neither of the two new simpler hyperplane systems need arise from a cut-and-project system. Indeed, continuing this way, we end with the empty collection of hyperplanes which is certainly not arising from a cut-and-project system. We carry out this computation completely for the example of the octagonal tilings in Sect. 6.
2. Cut-and-Project Systems In this section, we present the basic definitions and well-known results concerning projection method tilings. These (or variations of them) can be found in many places, e.g. [Be2,GS,Mo,Se]. They can also be found in [FHK], but the notation there seems less standard. For d ≥ 1, Rd denotes the usual Euclidean space of dimension d. The Euclidean norm of an element x in any Euclidean space is denoted |x|. We also use B(x, R) to denote the ball centred at x ∈ Rd of radius R > 0. A subset W is regular if it is non-empty and is the closure of its interior. The boundary of a set W is denoted ∂ W .
708
I. F. Putnam
Let G = Rd and H = R N be two Euclidean spaces. We let π and π ⊥ denote the two canonical projections of G × H onto G and H , respectively. There is an obvious action of G on G × H by translation in the first coordinate. With a slight abuse of notation, we denote this by u + x (or x + u), for x in G × H and u in G. In other words, we will regard G and H as subsets of G × H . ∼ Rd , H ∼ Definition 2.1. A cut-and-project system is a triple (G, H, L), where G = = N R and L ⊂ G × H is a lattice (i.e.. a co-compact discrete subgroup) satisfying the following: 1. the restriction of π : G × H → G to L is injective and 2. the map π ⊥ : G × H → H has π ⊥ (L) dense in H . We usually denote π ⊥ (L) by L. If, in addition, we have 3. the restriction of π ⊥ to L is injective, then we say the system is aperiodic. Forrest, Hunton and Kellendonk [FHK] work in somewhat greater generality. Also, they take the view that the lattice L = Zd+N . Of course, this is always the case up to isomorphism of groups, but their point of view is that the space G (which is E in their notation) is skewed, rather than the other way around. Definition 2.2. Let (G, H, L) be a cut-and-project system and let W be a compact, regular subset of H . The set of non-singular points, denoted N , is N = {x ∈ G × H | π ⊥ (x + L) ∩ ∂ W = ∅}. Furthermore, for x in N , we define x (W ) ⊂ G by x (W ) = π {y ∈ x + L | π ⊥ (y) ∈ W }. The elements of N are also sometimes called generic points and the set x (W ) is a model set. We note that usually model sets may be constructed for any x in G × H , without our hypothesis that x is in N . The following is an easy consequence of the definitions and we omit the proof. Lemma 2.3. Let (G, H, L) be a cut-and-project system and W ⊂ H be compact and regular. 1. If x is in N and s is in L, then x + s is in N and x+s (W ) = x (W ). 2. If x is in N and u is in G, then x + u is in N and x+u (W ) = x (W ) + u. The next step is to define a topology on the collection of model sets. We follow [BHZ]. We define M(G) as being in the dual space of Cc (G), the continuous compactly supported functions on G and give it the weak-* topology. The elements of M(G) are measures on G. To each set x (W ), x ∈ N , we regard it as an element of M(G) which is the sum of point measures over its elements. The fact that our sets are uniformly discrete plays an important part; we refer the reader to [BHZ] for details. The weak-* closure of x (W ), x ∈ N in M(G) will be compact. It is also important to note that a measure in the closure is again the sum of point masses over discrete point sets in G. We may suppress the distinction between point sets and measures. In fact, the space constructed above is metrizable. There are a number of possibilities for the metric and some may be more suitable than others. For our purposes here, it will be sufficient to use the following (see [FHK]), although this is not the most general.
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
709
Let , be countable subsets of G. We consider the set of all > 0 satisfying the following: there exist v, v ∈ B(0, ), ( − v) ∩ B(0, −1 ) = ( − v ) ∩ B(0, −1 ). We define d(, ) to be the infimum of all such . The closure of x (W ), x ∈ N in the topology concides with its completion in this metric. We also introduce a new topology on the set N as follows. We embed N into G × H × M(G) by sending x in N to (x, x (W )). Again the closure of the image is compact. This space is also metrizable and we define for x, y in N , d(x, y) = |x − y| + d(x (W ), y (W )). The closure of N in the larger space and the completion in this metric coincide. The following results summarize the facts we need regarding convergence of models sets in our metric. Proofs can be found in [FH]. Lemma 2.4. Suppose that xn , n ≥ 1, is a sequence in N ∩ H . 1. The sequence xn (W ), n ≥ 1, is convergent if and only if, for every R > 0, there is an N ≥ 1 such that xm (W ) ∩ B(0, R) = xn (W ) ∩ B(0, R), for all m, n ≥ N . 2. If the sequence xn (W ), n ≥ 1, converges to , then a point u in G is in if and only if there exists N ≥ 1 such that u is in xn (W ), for every n ≥ N . 3. The sequence xn (W ), n ≥ 1, is convergent if and only if, for every s in L, there exists N ≥ 1 such that either π ⊥ (xn +s) ∈ W , for all n ≥ N or else π ⊥ (xn +s) ∈ / W, for all n ≥ N . Definition 2.5. For a cut-and-project system, (G, H, L) and compact, regular set W in H , we let (W ) or simply denote the closure of {x (W ) | x ∈ N } in M(G), or ˜ ˜ denote the equivalently, its completion in the metric d. We also let (W ) or simply closure of N in G × H × M(G). ˜ We summarize the important features of our spaces (W ) and (W ) in the following theorem. We refer the reader to Chap. I of [FHK] for a proof. Theorem 2.6. Let (G, H, L) be a cut-and-project system and W be a regular subset of H. ˜ 1. There is a unique projection β˜ : (W ) → G × H which extends the identity map on N . ˜ 2. The actions of L and G on N extend to continuous actions on (W ) and β˜ is equivariant with respect to these and the obvious translation actions on the image. 3. The actions of L and G are free, wandering and commute with each other. 4. There is a unique continuous surjection β : (W ) → G × H/L which maps x (W ) to x + L, for any x in N . 5. The action of G on {x (W ) | x ∈ N } by translation extends to a continuous action on (W ) and β is equivariant with respect to it and the obvious translation action on the image. ˜ 6. (W )/L ∼ = (W ).
710
I. F. Putnam
We present an alternate approach to the definitions of the continuous hulls, (W ) ˜ and (W ) in terms of the spectrum of a commutative C ∗ -algebra. This approach was first taken in [BIT] for the Kohmoto model in the one-dimensional case. We begin with a cut-and-project system, (G, H, L), and a window, W . Since N , the set of non-singular points, is invariant under the action of G, we have N = G ×(N ∩ H ). First, consider the Hilbert space l 2 (N ∩ H ) of square summable functions on N ∩ H . Each bounded function on N ∩ H defines an operator on this space by pointwise multiplication. Let A(W ) or A denote the ∗-algebra of operators generated by C0 (H ) and all functions of the form χW −π ⊥ (s) , where s is in L. (In the aperiodic case, it is not necessary to include C0 (H ), but we do so for convenience. Having these algebras represented as operators is also for convenience in later arguments.) We let A(W ) or A denote the closure of A in the operator norm. Both A and A are commutative. Definition 2.7. The space H˜ (W ) or simply H˜ is the spectrum of the commutative C ∗ -algebra A(W ). Put another way, this means that, by the Gelfand-Naimark Theorem, there exists a locally compact Hausdorff space H˜ such that A ∼ = C0 ( H˜ ). In fact, the space H˜ is the collection of non-zero homomorphisms from A to the complex numbers. The topology is obtained by realizing each such homomorphism as an element of the dual space of A and using the relative weak topology. In our case (where A is separable), this means that a sequence φn , n ≥ 1, converges to φ if φn (a) converges to φ(a), for every a in A. For the C ∗ -algebra C0 (H ), this space is homeomorphic to H itself and the map associates to a point of H the homomorphism obtained by evaluation at that point. Since C0 (H ) ⊂ A, there is a continuous proper map q : H˜ (W ) → H such that φ( f ) = f (q(φ)), for all f in C0 (H ) and φ in H˜ (W ). We note that there is an injection of N ∩ H in H˜ which sends a point x to the functional φx ( f ) = f (x), for any f in A. The fact that any such functional extends continuously from A to A is immediate since it is given as a vector state from the Hilbert space l 2 (N ∩ H ). (As a remark, it is probably more natural to use the Hilbert space L 2 (H ), using Lebesgue measure. We avoid this route for two reasons. First, we would need to add the hypothesis that the Lebesgue measure of the boundary of W is zero. Secondly, we would need to prove at this point that the homomorphisms φx , x ∈ N ∩ H extend. This is not difficult, but we easily avoid the issue with our approach.) There is an action of L on A, denoted α, by αs ( f )(x) = f (x + π ⊥ (s)), for all x in N ∩ H . This extends to A and hence induces an action of L on H˜ . The map q is equivariant for this action. More general versions of the following appear in Sect. 3–7 of Chap. I of [FHK], but this will suffice for our purposes here. Theorem 2.8. The map sending x in N to (π(x), φπ ⊥ (x) ) in G × H˜ extends to a homeo˜ to G × H˜ which commutes with both G and L actions. morphism from The basic C ∗ -algebra of interest is the crossed product, C((W )) × G. However, ˜ from Theorem 2.6, (W ) is the quotient of (W ) by the action of L, the groups L and ˜ G are both acting on (W ) and the actions are commuting and wandering. The results of Rieffel (Situation 10 of [Ri]) apply directly and we conclude that C((W )) × G is
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
711
˜ Morita equivalent to C0 ((W )/G) × L. Moreover, it is clear from the description above ˜ that (W )/G is just H˜ (W ) and the action of L is just α as above. We summarize with the following statement, but the reader should also consult II.2.9 of [FHK]. Theorem 2.9. The C ∗ -algebras C((W )) × G and C0 ( H˜ (W )) ×α L are Morita equivalent. This implies, in particular, that C((W ))×Rd and C0 ( H˜ (W ))×α L have isomorphic K -theory groups. Henceforth, we will concentrate on the latter. We want to give some notation for crossed products. If A is a C ∗ -algebra with an action α of the discrete group L, we write elements of the crossed product as s∈L as u s , where each as is an element of A and only finitely many are non-zero. The collection of such elements is a dense ∗-subalgebra of the crossed product. The elements u s , s ∈ L, are unitary operators. In the case that A is non-unital, these lie in the multiplier algebra of A × L rather than the algebra itself. These satisfy the relation u s a = αs (a)u s , for all a in A and s in L. 3. Hyperplane Systems In the last section, we saw how C ∗ -algebras could be produced from a cut-and-project system and a window. Now, we give a different construction of C ∗ -algebras. The data in this case is called a hyperplane system. We will then establish some concrete relations between the two constructions; broadly speaking, the new construction will be more general. In the following defintion, we will consider an oriented, affine hyperplane of codimension one in a Euclidean space H . By this, we mean a set P which is the translate of a subspace H P by some vector x P . Of course, H P is just the set of all differences u − v, where u, v are in P. The point x P may be chosen arbitrarily from P, but we fix a choice. As P has codimension one, it divides the space H into two closed half-spaces whose intersection is P. By an orientation of P, we mean that we have a fixed choice of labelling these as P 0 and P 1 . ∼ R N , L ⊂ H is Definition 3.1. A hyperplane system is a triple (H, L , P), where H = a finitely generated subgroup and P is a countable collection of co-dimension 1 oriented affine hyperplanes in H which is invariant under the action of L. That is, for each P in P and s in L, P + s is also in P. A hyperplane cut-and-project system is a quadruple, (G, H, L, P), where (G, H, L) is a cut-and-project system and (H, π ⊥ (L), P) is a hyperplane system. Of course, a hyperplane system does not need G as part of its data. Moreover, the subgroup L does not need to be dense in H . Also, notice that we allow the possibility that P is empty. We note that for any two co-dimension one hyperplanes P, Q, their intersection is either P = Q, the empty set or a co-dimension 2 hyperplane. Frequently, the elements of P will be the L-orbits of subspaces of H , but this is not necessary. For any x in H , we let P(x) denote all elements of P which contain x. Let us also make clear that when we say that L acts on P, we mean that it preserves the orientation; that is, we have (P + s)i = P i + s, for all s in L and P in P.
712
I. F. Putnam
We will proceed to define C ∗ -algebras from a hyperplane system. In the case of a hyperplane cut-and-project system, we establish a link between these and the C ∗ -algebras of the last section, at least for the special class of windows which are polytopes whose boundaries are contained in the hyperplanes. Although at this point, we do not yet have any model sets, we may proceed to define a disconnected version of H , H˜ . This can be done in purely topological terms (see [Le,FHK]), but we find it most convenient to follow the ideas of Sect. 2 exploiting the Gelfand-Naimark theorem. Let N H denote the complement of the union of P in H . (This is a replacement for N ∩ H of the last section.) We consider the Hilbert space l 2 (N H ) and regard bounded functions on N H as operators. We let A(L , P) or simply A be the ∗-algebra generated by C0 (H ) and all functions of the form f χ P i , where f is in C0 (H ), P is in P and i = 0, 1, considered as operators on l 2 (N H ). Since N H is a dense G δ in H , this representation of C0 (H ) is faithful. Note that we have f χ P 0 + f χ P 1 = f , since the functions agree on N H . (We remark that we do not want our algebra to contain the function χ P i since it is not compactly supported.) We let A(L , P) or simply A be the closure of A(L , P) in the operator norm. Notice that if (H, L , P) and (H, L , P ) are hyperplane systems with P ⊂ P, then A(L , P ) ⊂ A(L , P) and A(L , P ) ⊂ A(L , P). Definition 3.2. Let (H, L , P) be a hyperplane system. We define H˜ (L , P) or simply H˜ to be the spectrum of the commutative C ∗ -algebra A(L , P). As A contains C0 (H ), there is a natural continuous surjection q : H˜ → H . Notice that H˜ need not be totally disconnected. In fact, when P is empty, it is just H . Our first result is to note that changing the data by translating P does not seriously affect the construction. Proposition 3.3. Let (H, L , P) be a hyperplane system and let x be in H . There exists a homeomorphism τx : H˜ (L , P) → H˜ (L , P − x) such that q ◦ τx (z) = q(z) + x, for all z in H˜ (L , P). Proof. The map sending f to τx ( f )(y) = f (y + x) defines an isomorphism from A(L , P) to A(L , P − x). This extends to the A algebras and therefore is induced by a homeomorphism of their spectra. That it satisfies the last statement is trivial. Our next objective is to give a better description of the elements of H˜ . Lemma 3.4. Let φ be a non-zero functional on A and let x = q(φ). 1. For any f in C0 (H ), P in P\P(x) and i = 0, 1, we have φ( f χ P i ) = f (x)χ P i (x). 2. For any P in P(x) and i = 0, 1, we have either φ( f χ P i ) = f (x), for all f in C0 (H ) or φ( f χ P i ) = 0, for all f in C0 (H ).
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
713
Proof. For the first statement, if P is not in P(x), then x is not in P and so either x is in the interior of P i or P 1−i . Let us suppose the former. We may find a function, g in C0 (H ), whose support is compact and contained in P i and satisfying g(x) = 1. Then we have φ( f χ P i ) = φ( f χ P i )g(x) = φ( f χ P i g) = φ( f g) = f (x)g(x) = f (x), and we are done. The case x in P 1−i is similar and we omit the details. For the second statement, we fix a function g in C0 (H ) such that g(x) = 1. First, we have φ(gχ P i ) + φ(gχ P 1−i ) = φ(gχ P i + gχ P 1−i ) = φ(g) = g(x) = 1. In addition, we also have φ(gχ P i )φ(gχ P 1−i ) = φ(gχ P i · gχ P 1−i ) = φ(0) = 0. We conclude that there are two possibilities, either φ(gχ P 0 ) = 1 and φ(gχ P 1 ) = 0, or vice verse. Now for any other f in C0 (H ), we have φ( f χ P i ) = g(x)φ( f χ P i ) = φ(g f χ P i ) = φ( f )φ(gχ P i ) = f (x)φ(gχ P i ). This completes the proof.
We may give a presentation of the points of H˜ as follows. Let φ be a non-zero functional on A. Let x = q(φ); this means that, for each f in C0 (H ), we have φ( f ) = f (x). Using the second part of the last result, we define δ : P(x) → {0, 1} by φ( f χ P 1 ) = δ(P) f (x), for any f in C0 (H ). In view of the last result, φ is uniquely determined by the pair (x, δ). Henceforth, we write φ = (x, δ). Notice that for a given x, not every function δ arises from some φ. In the case N = 2, if P(x) contains k lines, then q −1 {x} has 2k points, not 2k . The following provides a description of the points of H˜ and also the topology. Proposition 3.5. 1. Let x be in H and let δ : P(x) → {0, 1}. The point (x, δ) is in H˜ if and only if, for every > 0, there exists y in B(x, ) ∩ N H such that y is in P δ(P) for every P in P(x). 2. A sequence xn , n ≥ 1, in N H converges to (x, δ) in H˜ if and only if xn converges to x in H and, for all n sufficiently large, xn is in P δ(P) , for every P in P(x). Proof. First, we suppose that (x, δ) is in H˜ . Let > 0. Define W i = {P ∈ P(x) | δ(P) = i}, for i = 0, 1. Choose a function f in C0 (H ) whose support is contained in B(x, ) and so that f (x) = 1. Define the function g in A by f χP1 f χP0 = f χP1 ( f − f χ P 1 ). g= P∈W 1
P∈W 0
P∈W 1
P∈W 0
714
I. F. Putnam
We compute the value of the functional (x, δ) on g: (x, δ)(
P∈W 1
f χP1
( f − f χ P 1 )) =
P∈W 0
(x, δ)( f χ P 1 )
P∈W 1
·
(x, δ)( f − f χ P 1 )
P∈W 0
=
δ(P)
P∈W 1
= 1.
(1 − δ(P))
P∈W 0
From this, we conclude that the function g is non-zero. Hence, there is a point y in N H with g(y) = 0. It follows from Lemma 3.4 that y satisfies the desired conclusion. For the converse, it follows from the hypothesis that we may construct a sequence, yn , n ≥ 1, in N H which converges to x and such that yn is in P δ(P) , for every P in P(x). Regarding these points as vector states on A and hence in H˜ , we may find a subsequence which is convergent in the weak topology. It follows at once that (x, δ) arises from this limit. The second statement follows from Lemma 3.4. The next objective is to give a concrete link between the C ∗ -algebra of the hyperplane system arising from a cut-and-project system and that of the last section. For this, we need to consider a window W , but we must restrict to a specific class of polytopes, as follows. Definition 3.6. A subset W of H is a P-polytope if it is non-empty, compact, regular and can be written as W = ∩ Q∈W 0 Q 0 ∩ ∩ Q∈W 1 Q 1 , where W 0 and W 1 are finite subsets of P. Moreover, we say that the collections W 0 and W 1 are minimal if no proper subcollection of their union will have the same intersection. For Q in ∪i W i , we define ∂ Q W = Q ∩ W. The boundary of W can be written as the union of the sets ∂ Q W over all such Q. In some sense, this definition is going the wrong way. It is most usual to begin with a window W which is a polytope in the standard sense and then define P to be all hyperplanes which are translates of the set of hyperplanes which form the faces of W . The next result gives a specific link between our space H˜ and the topology on model sets from Sect. 2. Proposition 3.7. Let (G, H, L, P) be a hyperplane cut-and-project system. Suppose the sequence xn , n ≥ 1, in N H converges to (x, δ) in H˜ . Then for any P-polytope W , xn (W ) is Cauchy. We define (x,δ) (W ) to be its limit in the sense described in Sect. 2.
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
715
Proof. First, observe that N H is contained in N ∩ H . Since the space (W ) is compact, it suffices for us to show that any two limit points of subsequences of xn (W ) are equal. Suppose that and are limit points of the subsequences xm k (W ), k ≥ 1 and xnk (W ), k ≥ 1, respectively. Let u be in . By Lemma 2.4, for all k sufficiently large, we have u is in m k (W ) and u = π(s) for some s in L and xm k + π ⊥ (s) is in W . Write W as an intersection as in 3.6 for collections W 0 , W 1 . If xm k + π ⊥ (s) is in W , then it is in P 0 , for each P in W 0 and in P 1 , for each P in W 1 . Fix P in W i for the moment. Since xn converges to (x, δ) in H˜ , we know from Proposition 3.5 that for all n sufficiently large, xn is in ⊥ ⊥ (P − π ⊥ (s))δ(P−π (s)) = P δ(P−π (s)) − π ⊥ (s). This means that δ(P − π ⊥ (s)) = i and the subsequence xn k is also in P 0 − π ⊥ (s), hence xn k + π ⊥ (s) is in P 0 , for k large. As this holds for each P, we have xn k + π ⊥ (s) is in W , for k large. This implies that u is also in . We have shown ⊂ but the same argument shows the reverse inclusion and the conclusion follows. The precise relation between the construction of this section and the last is summarized in the following two results. Theorem 3.8. Suppose (G, H, L, P) is an aperiodic hyperplane cut-and-project system and that W is a P-polytope. Then we have C0 ( H˜ (W )) ⊂ C0 ( H˜ (L , P)) and the map sending f u s to f u π ⊥ (s) extends to an inclusion of C ∗ -algebras C0 ( H˜ (W )) ×α L ⊂ C0 ( H˜ (L , P)) ×α L . Proof. Let W 0 , W 1 be the minimal collection of elements of P as in the definition of W . Let s be in L. Choose f , a function in C0 (H ) such that f is identically one on W − π ⊥ (s). For i = 0, 1 and P in W i , the function f χ P i −π ⊥ (s) is in A(L , P). Then so is their product (over all i and P), which is just χW −π ⊥ (s) and it follows that A(W ) ⊂ A(L , P), which is the first statement. The second statement is a trivial consequence of the first and the fact of aperiodicity means that the map π ⊥ : L → L is an isomorphism. In general, the inclusion may be proper (for example, if W is a P -polytope where P is some proper L-invariant subset of P), but equality is obtained in the particular case when the window is the so-called canonical acceptance domain as follows. Let S be a set of generators for L. We define CS =
ts s | 0 ≤ ts ≤ 1, s ∈ S ,
s∈S
where C is chosen to suggest ‘cube’. We will consider the window π ⊥ (CS ). The set of hyperplanes, P, which form the boundaries of W (and all their translates under π ⊥ (L)) are described explicitly in Sect. IV.2 of [FHK]. There is also a proof of the following.
716
I. F. Putnam
Corollary 3.9. Suppose (G, H, L) is an aperiodic cut-and-project system, S is a set of generators of L, W = π ⊥ (CS ) is as above and P is the associated set of hyperplanes. Then we have C0 ( H˜ (W )) = C0 ( H˜ (L , P)) and the map sending f u s to f u π ⊥ (s) extends to an isomorphism of C ∗ -algebras C0 ( H˜ (W )) ×α L ∼ = C0 ( H˜ (L , P)) ×α L . Remark 3.10. Suppose that L is dense in H or at least that its span is all of H . We may then choose subgroups L 0 , L 1 of L so that L = L 0 ⊕ L 1 and L 0 is a lattice in H . In this case, the action of L 0 on H˜ (L , P) is free and wandering and C0 ( H˜ (L , P)) × L is strongly Morita equivalent to C0 ( H˜ (L , P)/L 0 ) × L 1 (this is just an L 1 -equivariant version of Situation 2 of [Ri]). Moreover, the space H˜ (L , P)/L 0 has a natural finite-to-one mapping onto H/L 0 , which is a torus. Remark 3.11. In the special case that P is empty, we have H˜ (L , P) = H . If we also assume L is as in the last remark, H˜ (L , P)/L 0 is just a torus and the action of L 1 is by rotation. The C ∗ -algebra C0 ( H˜ (L , P)/L 0 ) × L 1 is a non-commutative torus whose dimension is the rank of L. Remark 3.12. In the special case that H is one-dimensional and the group L has rank two, then we may choose L 0 ∼ = Z ∼ = L 1 . The action of L 1 on H˜ (L , P)/L 0 is by the restriction of a Denjoy homeomorphism to its unique minimal set, which is totally disconnected (see [PSS]). Provided that P is non-empty, the results of [PSS] show that K 0 (C0 ( H˜ (L , P) × L) ∼ = Z p+1 ,
K 1 (C0 ( H˜ (L , P) × L) ∼ = Z,
where p is the number of L-orbits in P. 4. Reduction of Hyperplane Systems In this section, we introduce the notion of the reduction of a hyperplane system to one of the hyperplanes. The result is another hyperplane system with lower dimensional data. It is important to note that if the original system is part of a hyperplane cut-and-project system, this extra feature need not pass to the reduction. Begin with a hyperplane system, (H, L , P). Choose P in P. Recall that H P = P − P is a N − 1-dimensional subspace H P = P − x P of H . Further, we define L P = L ∩ HP . The final ingredient in our reduced system is the collection of hyperplanes. We define P P to be the collection of all Q ∩ P − x P , where Q is in P, with Q ∩ P = ∅, P. (Neither H P nor L P depends on x P , and while P P does, it is only up to translation. We can then refer to Proposition 3.3 to make the relation precise.) A word of warning is appropriate: the choice of Q in P in obtaining Q ∩ P −x P is not unique. For example, in the octagonal tiling of Sect. 6, if we use P = P1 , we have P2 ∩ P1 = P3 ∩ P1 .
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
717
Definition 4.1. Let (H, L , P) be a hyperplane system. For P in P, the reduction of (H, L , P) to P is (H P , L P , P P ). We note the following easy results without proof. Proposition 4.2. Let (H, L , P) be a hyperplane system and let s be in L. We have H P+s = H P , L P+s = L P , P P+s = P P + x P − x P+s . Lemma 4.3. If P is in P, then N HP = HP \
(Q − x P ).
Q∈P ,Q= P
For any x in N H P and s in L, we have P(x + x P + s) = {P + s}. Lemma 4.4. Let W be a P-polytope, expressed minimally as W = (∩ Q∈W 1 Q 1 ) ∩ (∩ Q∈W 0 Q 0 ). Suppose that P is in W 1 or W 0 . Then ∂ P W − x P is a P P -polytope. We now consider a cut-and-project hyperplane system, (G, H, L, P), and its associated hyperplane system (H, L , P). Our aim is to show that under some (fairly strong) hypotheses, for a given P in P, the reduction of (H, L , P) to P will again arise from a cut-and-project hyperplane system. In this case, there is a precise relation between the associated model sets. Define L P = {s ∈ L | π ⊥ (s) ∈ H P }, and subsequently G P = {π(x) ∈ span R L P | π ⊥ (x) = 0} ⊂ G. Theorem 4.5. Let (G, H, L, P) be a hyperplane cut-and-project system and let P be in P. If 1. H P ⊂ span R L P and 2. L P is dense in H P , then (G P , H P , L P ) is a cut-and-project system. If (G, H, L) is aperiodic, then so is (G P , H P , L P ). Proof. The first thing to prove is that if H P ⊂ span R L P , then L P is a subset of G P ×H P . Let s be in L P , so that π ⊥ (s) is in H P ⊂ span R L P . It follows that π(s) = s −π ⊥ (s) is in G P . As L is discrete, so is L P . Since L P is dense in H P , we have π ⊥ (span R L P ) = H P and then span R L P = G P × H P . As L P is a subset of L, it follows that the restrictions of π and π ⊥ to the former are injective if their restrictions to the latter are.
718
I. F. Putnam
The last result of this section is that, under the hypothesis of the last theorem, there is a precise relation between certain model sets for the original system and those of the reduction. More precisely, suppose that (x, δ0 ) and (x, δ1 ) are two points in H˜ , where x is in P and δ0 and δ1 differ only by their values on P. Then the set theoretic difference (x,δ1 ) (W ) \ (x,δ0 ) (W ) can be expressed in terms of a model set for the reduced system (G P , H P , L P , P P ). This result is certainly known in the quasicrycstal community, but we state it here because it has an interesting contrast with a result in the next section, and also gives a proof for completeness. As L is finitely generated, the subgroup L P is a direct summand and we choose another subgroup L P such that L = L P × L P . Of course, there may be many choices; this will not effect the result. We define maps j i P : H˜ P (L P , P P ) × L P → H˜ (L , P),
for j = 0, 1. For (x, δ) in H˜ P (L P , P P ) and s in L P , we define i P ((x, δ), s ) = (x + x p + s , δ¯ j ), where δ(Q − x P − s ), for Q = P + s , δ¯ j (Q) = . j, for Q = P + s j
j
It follows from Proposition 3.5 that both i P , j = 0, 1 are well-defined and continuous; we omit the details. In particular, if we consider x which is in N HP , we have j P(x + x P + s ) = {P + s } by Proposition 4.3 and we define i P (x) = (x + x P + s , j), where the second entry is interpreted as the constant function j on the singleton {P + s }. Theorem 4.6. Let (G, H, L, P) be a hyperplane cut-and-project system. Let P be an element of P such that 1. H P ⊂ span R L P and 2. L P is dense in H P . Let W be a P-polytope associated to the minimal collections W 0 and W 1 . Suppose that (x, δ) is in H˜ P (L P , P P ) and s is in L with π ⊥ (s ) in L P . If there exists s in L such that P + π ⊥ (s) is in W 1 , then i 1 ((x,δ),π ⊥ (s )) (W ) \ i 0 ((x,δ),π ⊥ (s )) (W ) P
P
= (x,δ) (∂ P+π ⊥ (s) W − x P − π ⊥ (s)) + π(s − s ). If there is no such s, then i 1 ((x,δ),s ) (W ) ⊂ i 0 ((x,δ),s ) (W ). P
P
Proof. Case 1. We begin with the added assumptions that s = 0 and that x is in N H P . In the right-hand side of our formula, the point (x, δ) can be replaced by simply x. Of course, this simplifies things because the set on the right-hand side is a genuine model set and not a limit of them. It also implies that P(x + x P ) = {P} and we have j i P (x, 0) = (x + x P , j), for j = 0, 1. Choose a sequence xn , n ≥ 1, in P 1 − x P a sequence yn , n ≥ 1, in P 0 − x P both converging to x.
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
719
Suppose that u = π(t) is in (x+x P ,1) and not in (x+x P ,0) . From the definitions and Lemmas 2.4, this implies that xn + x P + π ⊥ (t) is in W and yn + x P + π ⊥ (t) is not in W , for n large. It follows that x + x P + π ⊥ (t) is in ∂ W and hence in ∂ Q W , for some Q in W i , where i is either 0 or 1. This implies that x + x P is in Q − π ⊥ (t) and since we know P(x + x P ) = {P}, we must have Q − π ⊥ (t) = P. Next, since xn + x P + π ⊥ (t) is in W which is contained in Q i = P i + π ⊥ (t). It follows from our choice of xn that i = 1. We have now shown that if the set difference we are considering is non-empty, then there exists an element t satisfying the condition P + π ⊥ (t) is in W 1 . Equivalently, if there is no such t, we have the containment (x,δ1 ) (W ) ⊂ (x,δ0 ) (W ). Now, we continue to prove that u = π(t) is in the right-hand side of the first expression. We have P + π ⊥ (s) and P + π ⊥ (t) are both in W 1 . If they are distinct, then by simple geometric considerations, either P + π ⊥ (s) ⊂ P + π ⊥ (t) or the reverse. But proper containment would contradict the minimality of W 1 . We conclude that these are equal and so π ⊥ (t) − π ⊥ (s) is in H P and hence t − s is in L P . Moreover, we have π ⊥ (x + t − s) = x + π ⊥ (t) − π ⊥ (s) ∈ ∂ P+π ⊥ (s) W − x P − π ⊥ (s), and it follows that π(x + t − s) is in x (∂ P+π ⊥ (s) W − x P − π ⊥ (s)). Moreover, we have u = π(t) = π(x + t − s) + π(s) ∈ x (∂ P+π ⊥ (s) W − x P − π ⊥ (s)) + π(s). We have proved ⊂ in the first expression. Now suppose that u is in x (∂ P+π ⊥ (s) W −x P −π ⊥ (s)). This means that u = π(x+t) for some t in L P , where π ⊥ (x + t) is in the interior of ∂ P+π ⊥ (s) W − x P − π ⊥ (s)), with respect to H P . This means that x + x P + π ⊥ (s + t) is in the interior of ∂ P+π ⊥ (s) W relative to P + π ⊥ (s). This, in turn, means that x + x P + π ⊥ (s + t) is in the interior of Q i , for each Q = P + π ⊥ (s) in W i and i = 0, 1. This also implies that xn + x P + π ⊥ (s + t) is in the interior of Q i , for each Q in W i (including P + π ⊥ (s)) and i = 0, 1, and n large. On the other hand, yn + x P + π ⊥ (s + t) is not in P 1 + π ⊥ (s). It follows that xn + x P + π ⊥ (s + t) is in W , while, yn + x P + π ⊥ (s + t) is not in W , for n large. Then we have π(xn + x P + s + t) = π(s) + π(t) = u + π(s) is in i 1 ((x,δ),0) (W ) and not in P i 0 ((x,δ),0) (W ). This completes the proof in Case 1. P Case 2. We remove the condition that s = 0, but still require x is in N H P . This implies that P(x + x P + π ⊥ (s )) = {P + π ⊥ (s )}. The entire argument above may be applied to the point x and the element P +π ⊥ (s ) in P. In this case, we have P +π ⊥ (s )+π ⊥ (s −s ) j is in W 1 . We choose x P+π ⊥ (s ) = x P + π ⊥ (s ). Moreover, we have i P+π ⊥ (s ) (x, 0) =
i P (x, π ⊥ (s )). The result follows at once. Case 3. We finally consider the general case. Any point (x, δ) is the limit of a sequence xn in N H P . We know the conclusion holds for each xn and the result for (x, δ) holds by virtue of Lemma 2.4. j
5. K-Theory There are already techniques available for the computation of the K-theory of the C ∗ -algebras associated with a cut-and-project system [FHK,BS]. Here, we present a variation on these results. In principle, this is more general in that they apply to hyperplane systems, but it is likely that the results of [FHK] could be generalized in this fashion.
720
I. F. Putnam
We begin with a hyperplane system, (H, L , P), and an element P in P. Let (H P , L P , P P ) denote the reduction on P. In addition, we let P = P \ {P + s | s ∈ L}. As earlier, H˜ (L , P), H˜ (L , P ) and H˜ (L P , P P ) are the spaces associated with each. For brevity, it will sometimes be useful to use the notation B(H, L , P) = C0 ( H˜ (L , P)) × L , B(H, L , P ) = C0 ( H˜ (L , P )) × L , B(H P , L P , P P ) = C0 ( H˜ P (L P , P P )) × L P . The aim of this section is to give a six-term exact sequence relating the K-groups of these three C ∗ -algebras. One map, in particular, is given in a very natural way by a class in the Kasparov group K K (B(H, L , P), B(H P , L P , P P )). The result is an immediate application of the techniques in [Pu] for the transformation groupoids H˜ (L , P) × L , H˜ (L , P ) × L and H˜ P (L P , P P ) × L P . Recall from the last section that L P ⊂ L is chosen such that L = L P × L P . We consider the space H˜ P (L P , P P ) × L P with the action of L given by (x, t ) + (s + s ) = (x + s, t + s ), for all x in H˜ P (L P , P P ), t , s in L P and s in L P . We note that L P is given the discrete topology and that this action is free. For any s in L P , we let δs denote the function on L P which is 1 at s and 0 elsewhere. If f is in H˜ P (L P , P P ) and s is in L P , we denote by f ⊗ δs the obvious function on H˜ P (L P , P P ) × L P . The linear span of such functions is dense in C0 ( H˜ P (L P , P P ) × L P ). We regard H˜ P (L P , P P ) as a subset of H˜ P (L P , P P ) × L P by identifying it with ˜ H P (L , P P ) × {0}. It is an abstract transversal in the sense of Muhly, Renault and Williams [MRW] and the groupoids ( H˜ P (L P , P P ) × L P ) × L and H˜ P (L P , P P ) × L P are equivalent. This implies that their C ∗ -algebras are Morita equivalent. In fact, one can show quite explicitly that the map which sends ( f ⊗ δt )u s+s to f u s ⊗ et ,t +s extends to an isomorphism C0 ( H˜ P (L P , P P ) × L P ) × L ∼ = (C0 ( H˜ P (L P , P P )) × L P ) ⊗ K(l 2 (L P )), where we use the notation es1 ,s2 to denote the rank one operator which maps a vector ξ in l 2 (L P ) to < ξ, δs2 > δs1 . We have already defined the two maps j i P : H˜ P (L P , P P ) × L P → H˜ (L , P),
for j = 0, 1. We extend i 0P , i 1P to maps on the groupoids j i P : ( H˜ P (L P , P P ) × L P ) × L → H˜ (L , P) × L ,
for j = 0, 1, which are the identity on L. There should be no confusion if we use the same notation. The pair (i 0P , i 1P ) defines, in a very natural way, an element of the Kasparov group K K 0 (B(H, L , P), B(H P , L P , P P )) which we describe now. First, we use the definition of K K provided by Cuntz [Cu]. Let H be a separable, infinite dimensional Hilbert space. We consider M(B(H P , L P , P P ) ⊗ K(H)), the multiplier algebra of
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
721
B(H P , L P , P P ) ⊗ K(H). It contains B(H P , L P , P P ) ⊗ K(H) as an ideal. A quasihomomorphism from B(H, L , P) to B(H P , L P , P P ) is a pair, (ρ0 , ρ1 ) of ∗-homomorphisms from B(H, L , P) to M(B(H P , L P , P P ) ⊗ K(H)), such that, for every a in B(H, L , P), ρ0 (a) − ρ1 (a) is in B(H P , L P , P P ) ⊗ K(H). In our case, we use the Hilbert space l 2 (L P ) and we have already noted that (C0 ( H˜ P (L P , P P )) × L P ) ⊗ K(l 2 (L P )) ∼ = C0 ( H˜ P (L P , P P ) × L P ) × L . j j Notice that if f is in C0 ( H˜ (L , P)), then, for j = 0, 1, the function i P ( f ) = f ◦ i P is a continuous bounded function on H˜ P (L P , P P ) × L P . Of course, it fails to vanish at infinity, but it does lie in the multiplier algebra of C0 ( H˜ P (L P , P P × L P )). The action j of L extends to the multiplier algebra and the map i P is equivariant. That is, we define
i P : B(H, L , P) → M(C0 ( H˜ P (L P , P P × L P )) × L), j
by j
iP(
s∈L
for j = 0, 1 and
s∈L
fs u s ) =
j ( f s ◦ i P )u s s∈L
f s u s in B(H, L , P) as before.
Lemma 5.1. Let f be in C0 (H ) ⊂ C0 ( H˜ (L , P)), Q be in P and i = 0, 1. If there exists s in L P such that Q = P + s , then we have i 0P ( f χ Q i ) − i 1P ( f χ Q i ) = (−1)i f ⊗ δs , where f in C0 (H P ) ⊂ C0 ( H˜ P (L P , P P )) is defined by f (x) = f (x + x P + s ), x ∈ H P . If there is no such s , then i 0P ( f χ Q i ) − i 1P ( f χ Q i ) = 0. In particular, i 0P ( f χ Q i ) − i 1P ( f χ Q i ) is in C0 ( H˜ P (L P , P P ) × L P ). Proof. For j = 0, 1, we have j i P ( f χ Q i )((x, δ), s ) = ( f χ Q i )(x + x P + s , δ¯ j ).
This first case is when x + x P + s is not in Q. In this case, the result is f (x + x P + s )χ Q i (x + x P + s ) and is independent of j. The second case is when x + x P + s is in Q and Q = P + s . In this case, the result is f (x + x P + s )δ(Q) and again is independent of j. Finally, if x + x P + s is in Q and Q = P + s , the result is f (x + x P + s ) if j = i and zero otherwise. The conclusion follows.
722
I. F. Putnam
Since functions of the form f χ Q i as in the lemma generate the C ∗ -algebra C0 ( H˜ (L , P)), which is an ideal in its multiplier algebra, it follows that i 0P (a) − i 1P (a) is in C0 ( H˜ P (L P , P P ) × L P ), for every a in C0 ( H˜ (L , P)). It follows immediately that i 0P (a) − i 1P (a) is in B(H P , L P , P P ), for every a in B(H, L , P). Following [Pu], we now want to take the quotient of H˜ (L , P))× L by the equivalence relation i 0P ((x, δ), s ) ∼ i 1P ((x, δ), s ), for all ((x, δ), s ). In fact, it is easy to see that for any ((y0 , δ0 ), s0 ) and ((y1 , δ1 ), s1 ) in H˜ (L , P)) × L, they are related by ∼ if and only if y0 = y1 , s0 = s1 and δ0 |P (y0 ) = δ1 |P (y1 ). This is the essential idea of the proof of the following; the remainder is a matter of checking topological details which are fairly simple by using Proposition 3.5 and we omit them. Theorem 5.2. The map α sending (y, δ) in H˜ (L , P) to (y, δ|P (y)) in H˜ (L , P ) is continuous, proper and L-invariant. If we also denote by α the map between the groupoids H˜ (L , P) × L and H˜ (L , P ) × L which is α in the first coordinate and the identity in the second, then it induces an isomorphism of topological groupoids between the quotient of H˜ (L , P) × L by i 0P ((x, δ), s ) ∼ i 1P ((x, δ), s ), for all ((x, δ), s ), and ( H˜ (L , P )) × L. Notice that our map α induces a *-homomorphism, also denoted α, α : C0 ( H˜ (L , P )) × L → C0 ( H˜ (L , P)) × L by α( s f s u s ) = s ( f s ◦ α) u s . The hypotheses of Theorem 2.1 of [Pu] are satisfied and we conclude the following holds. Theorem 5.3. Let (H, L , P) be a hyperplane system, P be an element of P and P = P \ {P + s | s ∈ L}. There is a six-term exact sequence α∗
K 0 (B(H,O L , P )) K 1 (B(H P , L P , P P )) 0o
[i P ,i 1P ]∗
/ K 0 (B(H, L , P)) K 1 (B(H, L , P)) o
[i 0P ,i 1P ]∗
/ K 0 (B(H P , L P , P P )) ,
α∗
K 1 (B(H, L , P )),
where [i 0P , i 1P ] is the element of K K 0 (B(H, L , P), B(H P , L P , P P )) described above. Finally, we note that this computation of the map can be carried out quite explicitly, at least for projections which are functions on H˜ (L , P), as follows. Theorem 5.4. Suppose that W = (∩ Q∈W 0 Q 0 ) ∩ (∩ Q∈W 1 Q 1 ) is a P-polytope, expressed minimally. For j = 0, 1, if there exists (a necessarily unique) s j ∈ L P such that P + s j ∈ W j , then set W j = ∂ P+s j W − x P − s j and set W j to be the empty set otherwise. Then we have [i 0P , i 1P ]∗ ([χW ]) = [χW 0 ] − [χW1 ].
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
723
j
Proof. Let p j = i P (χW ), for j = 0, 1. These are projections in the multiplier algebra of C0 ( H˜ P (L P , P P ) × L P ) × L and their difference is in C0 ( H˜ P (L P , P P ) × L P ) × L. Cuntz describes K 0 (C0 ( H˜ P (L P , P P ) × L P ) × L) in terms of such pairs. If a pair [ p, p ] have both p, p in the algebra itself, then this pair is the same as [ p] − [ p ] in the usual description of K-zero. In our case, we exploit the extra feature that p0 and p1 commute (since they are both in the commutative algebra C0 ( H˜ P (L P , P P ) × L P )). In Cuntz’ approach, the pair ( p0 p1 , p0 p1 ) is the trivial element. Let p j = χW j ⊗ δs j if s j exists and is zero otherwise. Each of these is in C0 ( H˜ P (L P , P P ) × L P ). Under the explicit isomorphism between C0 ( H˜ P (L P , P P ) × L P ) × L and B(H P , L P , P P ) ⊗ K(l 2 (L P )), the projection p j is mapped to χW j ⊗ δs j and its class in K 0 (B(H P , L P , P P )) is [χW j ].
Since s0 and s1 are clearly distinct if they both exist, we have p0 p1 = 0. We will show that p0 − p1 = p0 − p1 . Assuming this for the moment, if we subtract p0 from both sides and multiply by p0 , we get − p0 p1 = p0 (1− p0 ). The left-hand side is clearly negative, while the right is positive. We conclude that they are both 0. Analogous results hold replacing p0 with p1 . From this it follows that p0 and p1 are both orthogonal to p0 p1 and we have p0 p1 + p j = p0 p1 + p j p j = p j ( p1− j + p j ) = p j ( p1− j + pj) = 0 + pj = pj.
This means that the pair ( p0 p1 , p0 p1 ) may be added to the orthogonal pair ( p0 , p1 ) and the result is ( p0 , p1 ). We conclude that in K -zero, we have [i 0P , i 1P ]∗ ([χW ]) = [( p0 , p1 )] = [( p0 , p1 )] = [ p0 ] − [ p1 ] = [χW 0 ] − [χW1 ]. It remains to prove that p0 − p1 = p0 − p1 . Find a function f in C0 (H ) which is identically equal to one on the compact set W . We can express χW = ( Q∈W 0 f χ Q 0 )( Q∈W 1 f χ Q 1 ). Let us assume for the moment that both s0 and s1 exist. It follows from Lemma 5.1 that i 0P ( f χ Q i ) = i 1P ( f χ Q i ), for all Q = P +s0 , P +s1 . Also, let f i (x) = f (x+x P +si ), i = 0, 1 be as in Lemma 5.1, so that i 0P ( f χ P i +s ) − i 1P ( f χ P i +s ) = (−1)i f i ⊗ δsi i
i
for i = 0, 1. Next, for Q = P + s0 , P + s1 and j, k = 0,, we claim that i P ( f χ Q k ) f i ⊗ δsi = ( f i )2 χ Q k ∩P−x P −s ⊗ δsi . j
i
We evaluate both sides at (x, s ) ∈ H(L P , P P ) × L P . It is clear that both sides are zero unless s = si . Moreover, since Q is not in the L-orbit of P, we have i P ( f χ Q k )(x, si ) = f (x + x P + si )χ Q k (x + x P + si ) = f i (x)χ Q k ∩P−x P −s (x). j
i
724
I. F. Putnam
This establishes the claim. Let a denote the product of all factors f χ Q k over k = 0, 1 and Q in W k with Q = P + si . It follows from the claim above that i P (a) f i ⊗ δsi = ( f i )n χ∂ P+s W −x P −si ⊗ δsi , j
i
where n is the number of factors involved. Since f is identically 1 on W , f i is identically 1 on ∂ P+si W − x P − si and so χ∂ P+s W −x P −si f i = χ∂ P+s W −x P −si . i
i
Next, we consider the product f χ P 0 +s f χ P 1 +s = f 2 χ(P 0 +s )∩(P 1 +s ) . 0
1
0
1
As this is a factor in χW , it must be non-zero. The hyperplanes P + s0 and P + s1 are parallel and for the two opposite half-spaces to have non-trivial intersection, P + s1 must be contained in P 0 + s0 and P + s0 must be contained in P 1 + s1 . It follows that i P ( f χ P 0 +s ) f 1 ⊗ δs1 = i P ( f ) f 1 ⊗ δs1 , j
j
0
i P ( f χ P 0 +s ) f 0 ⊗ δs0 = i P ( f ) f 0 ⊗ δs1 . j
j
0
We compute p0 − p1 = i 0P ( f χW ) − i 1P ( f χW ) = i 0P ( f χ P 0 +s f χ P 1 +s a) − i 1P ( f χ P 0 +s f χ P 1 +s a) 0
=
1
0
1
(i 0P ( f χ P 0 +s )i 0P ( f χ P 1 +s a) − i 1P ( f χ P 0 +s )i 0P ( f χ P 1 +s a)) 0 1 0 1 +(i 1P ( f χ P 0 +s a)i 0P ( f χ P 1 +s ) − i 1P ( f χ P 0 +s a)i 1P ( f χ P 1 +s )) 0
1
0
1
= ((i 0P ( f χ P 0 +s ) − i 1P ( f χ P 0 +s )))i 0P ( f χ P 1 +s a) 0
0
1
+i 1P ( f )i 1P ( f χ P 0 +s a)(i 0P ( f χ P 1 +s ) − i 1P ( f χ P 1 +s )) 0 1 1 = ( f 0 ⊗ δs0 )i 0P ( f χ P 1 +s a) − i 1P ( f χ P 0 +s a)( f 1 ⊗ δs1 ) 1
=
( f 0 )n+2 χ∂ 0 W −x P −s0 P +s 0
0
⊗ δs0 − ( f 1 )n+2 χ∂ 1 W −x P −s1 P +s 1
⊗ δs1
= χW0 ⊗ δs0 − χW1 ⊗ δs1 = p0 − p1 . This completes the proof in the case that both s0 and s1 exist. The proofs in the three remaining cases are similar or even somewhat easier since the argument two paragraphs above is no longer needed. We omit the details. Remark 5.5. It is interesting to compare this result with Theorem 4.6. This is a dynamical/operator theory analogue of that result for model sets.
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
725
Remark 5.6. The complexity of a hyperplane system (H, L , P) can be roughly measured by three numbers: the dimension of H , the rank of L, and the number of distinct L-orbits in P. (The last may be infinite.) We compare the three systems (H, L , P), (H, L , P ) and (H P , L P , P P ). The second is simpler than the first; the first two numbers remain the same while the third is reduced by one. The third is also simpler than the first; the first number is reduced by one, the second by at least one (except in rather trivial cases) and the third is reduced by at least one. We may regard Theorem 5.3 as computing the K-theory of the C ∗ -algebra associated with (H, L , P) in terms of the other two (simpler) systems. In principle, the same techniques could be applied to these simpler systems. In the end, there would presumably be a spectral sequence. We do not pursue this here. It is important to note that even in the case that (H, L , P) arises from a cut-and-project system as earlier, the intermediate systems may not, but are merely hyperplane systems. For example, it is quite possible that L P = 0. 6. Example: The Octagonal Tiling We now use the methods of the last section to compute the K-theory of the C ∗ -algebra of the octagonal tiling, as described in the introduction. Along the way, we establish a number of results in the general setting. We do not give complete proofs, which are somewhat lengthy. One aspect which is nice for the octagonal case is that as we proceed along, all the groups are free abelian, so it most convenient for us to simply list their generators. (The reader should note this is not always the case.) Let us just start with some notation which is particular to the octagonal case. Here H is dimension two, while L is generated by π ⊥ (e1 ), π ⊥ (e2 ), π ⊥ (e1 ), π ⊥ (e4 ), which we now denote by l1 , l2 , l3 , l4 , for convenience. The four hyperplanes Pk , k = 1, 2, 3, 4 are in fact subspaces so that H Pk = Pk . For simplicity, we denote this by Hk . We also denote its stabilzer under L as L Pk = L k . It is fairly easy to see that L 1 is generated by l1 and l3 − l4 . There are similar descriptions of the others. If one considers all non-trivial pairwise intersections of hyperplanes in the collection P, this set of points is, of course, invariant under L. It consists of exactly three L-orbits, that of H1 ∩ H2 (which is the origin), (H1 + l4 ) ∩ H2 and (H3 + l1 ) ∩ H4 . We denote these three points by V0 , V1 and V3 respectively. We let L i denote the stabilzer of Hi , for each 1 ≤ i ≤ 4. First of all, we note that in the case that the collection of hyperplanes P is empty, then H˜ (L , ∅) = H . It follows that ˆ × Hˆ , C0 ( H˜ (L , ∅)) × L ∼ = C0 (H ) × L ∼ = C ∗ (L) × Hˆ ∼ = C( L) where the second isomorphism is via Fourier transform. As L is isomorphic to Z4 , we have Lˆ ∼ = H , since it is a Euclidean space. Applying Connes’ = T4 , the 4-torus, while Hˆ ∼ analogue of the Thom isomorphism, the K-theory of this C ∗ -algebra is the same as that of the torus (since dim H is even). We conclude from this that K 0 (B(H, L , ∅)) ∼ = ∧even L , K 1 (B(H, L , ∅))) ∼ = ∧odd L . Since our group L and various subgroups of it will be acting, not just on H , but also on various subspaces determined by the hyperplanes, we will denote elements of the K-theory above by l H , for l in L and exterior powers of such symbols. We may list the
726
I. F. Putnam
generators of K 0 (B(H, L , ∅))) as 1H ,
l1H ∧ l2H , l1H ∧ l3H ,
l1H ∧ l4H ,
l2H ∧ l3H , l2H ∧ l4H , l3H ∧ l4H , l1H ∧ l2H ∧ l3H ∧ l4H , and similarly for K 1 . The next step to consider is the exact sequence of Theorem 5.3 in the special case that P is the L-orbit of a single hyperplane P, and hence P is empty. It can be computed explicitly from the exact sequence and in notation above, it simply sends l H P in K i (B(H P , L P , ∅)) to l H in K i+1 (B(H, L , ∅)) and similarly for exterior powers. (Notice that the dimension shift from the exact sequence is taken care of by the isomorphism above and the fact that dim H = dim H P + 1. For the octagonal tiling, we apply this for P = H1 . Notice that L 1 = H1 ∩ L is generated by l1 and l3 − l4 . Let P1 be the L-orbit of H1 . The generators of K 0 (B(H1 , L 1 , ∅)) are l1H1 , (l3 − l4 ) H1 . The generators of K 1 (B(H1 , L 1 , ∅)) are 1 H1 , l1H1 ∧ (l3 − l4 ) H1 . These groups are mapped injectively into the groups K i (B(H, L , ∅)) and the quotients are K i (B(H, L , P1 )). The generators for K 0 (B(H, L , P1 )) are l1H ∧ l2H , l1H ∧ l3H = l1H ∧ l4H , l2H ∧ l4H ,
l3H ∧ l4H ,
l2H ∧ l3H , l1H ∧ l2H ∧ l3H ∧ l4H .
The generators for K 1 (B(H, L , P1 )) are l2H ,
l3H = l4H ,
l1H ∧ l2H ∧ l3H ,
l1H ∧ l2H ∧ l4H , l1H ∧ l3H ∧ l4H , l2H ∧ l3H ∧ l4H . We next turn to disconnecting along the hyperplane H3 and its L-orbit. Notice that L 3 is generated by l3 and l1 + l2 . Let P2 denote the L-orbits of H1 and H3 . We will apply Theorem 5.3 to P2 , with P = H3 and so P = P1 . The first task is to compute K ∗ (B(H3 , L 3 , (P2 ) H3 )). In this case, we have (P2 ) H3 is just the L 3 -orbit of V0 . As a result, the generators of K 0 (B(H3 , L 3 , (P2 ) H3 )) are l3H3 , (l1 + l2 ) H3 and the generator of K 1 (B(H3 , L 3 , (P2 ) H3 )) is (l1 + l2 ) H3 ∧ l3H3 . Again, the maps from these groups into the K-theory of B(H, L , P1 ) are injective and the K-theory of B(H, L , P2 ) is just the quotient. We are able to list the generators of K 0 (B(H, L , P2 )) as l1H ∧ l2H , l1H ∧ l3H = l1H ∧ l4H = −l2H ∧ l3H , l2H ∧ l4H , l3H ∧ l4H ,
l1H ∧ l2H ∧ l3H ∧ l4H .
The generators for K 1 (B(H, L , P2 )) are l1H ∧ l2H ∧ l3H , l1H ∧ l2H ∧ l4H , l1H ∧ l3H ∧ l4H , l2H ∧ l3H ∧ l4H . We next turn to disconnecting along the hyperplane H2 and its L-orbit. Let P3 denote the L-orbits of H1 , H3 and H2 . We will apply Theorem 5.3 to P3 , with P = H2 and so P = P2 . The first task is to compute K ∗ (B(H2 , L 2 , (P3 ) H2 )). It is easy to see that (P3 ) H2 is simply the L 2 -orbits of the points V0 , V1 . We would like to apply the same theorem to compute this, for the hyperplane system (H2 , L 2 , {V0 , V1 } + L 2 ). Ordinarily, this would require two applications of the theorem, since there are two distinct
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
727
hyperplane orbits which must be removed. However, since these hyperplanes are parallel, we may make a single application, letting P = ∅ be the result of removing them both and the third term is simply the direct sum of the K-theory groups of B({V0 }, 0, ∅) and B({V1 }, 0, ∅). The K 0 group is just the free abelian group on 10V , 1V1 and K 1 is trivial. From this, we obtain generators for K 0 (B(H2 , L 2 , (P2 ) H2 )): 1V0 − 1V1 , l2H2 , (l3 + l4 ) H2 , where the first is a slightly rough notation for an element whose image under the KK-map is 1V0 − 1V1 , which is in the kernel of the map into K 1 (B(H2 , L 2 , ∅)). We also obtain a single generator for K 1 (B(H2 , L 2 , (P3 ) H2 )), which is l2H2 ∧ (l3 + l4 ) H2 . In this case, the map from K 0 (B(H2 , L 2 , (P3 ) H2 )) to K 1 (B(H, L , P2 )) is zero while the map from K 1 (B(H2 , L 2 , (P3 ) H2 )) to K 0 (B(H, L , P2 )) is injective. We conclude that a set of generators for K 0 (B(H, L , P3 )) is 1V0 − 1V1 ,
l2H2
(l3 + l4 ) H2
l1H ∧ l3H = l1H ∧ l4H = l2H ∧ l3H = −l2H ∧ l4H , l1H ∧ l2H ,
l3H ∧ l4H ,
l1H ∧ l2H ∧ l3H ∧ l4H .
while a set of generators for K 1 (B(H, L , P3 )) is the same as for K 1 (B(H, L , P2 )) above. Finally, we consider P which is the translation of all four hyperplanes and reduce of P = H4 , so that P = P3 . The computation of K ∗ (B(H4 , L 4 , (P) H4 )) is analogous to that above for K ∗ (B(H2 , L 2 , (P3 ) H2 )) and the generators of its K 0 -group are 1V0 − 1V2 , l4H4 , (l1 − l2 ) H4 , while the generator of its K 1 -group is (l1 − l2 ) H4 ∧ l4H4 . Here, the map from both of these groups is zero and so we may write generators for K 0 (B(H, L , P)) as 1V0 − 1V1 ,
l2H2
(l3 + l4 ) H2 ,
1V0 − 1V2 ,
l4H4
(l1 − l2 ) H4 ,
l1H ∧ l3H = l1H ∧ l4H = l2H ∧ l3H = −l2H ∧ l4H , l1H ∧ l2H ,
l3H ∧ l4H ,
l1H ∧ l2H ∧ l3H ∧ l4H .
and the generators for K 1 (B(H, L , P2 )) as l1H ∧ l2H ∧ l3H , l1H ∧ l2H ∧ l4H , l1H ∧ l3H ∧ l4H , l2H ∧ l3H ∧ l4H , (l1 − l2 ) H4 ∧ l4H4 . Remark 6.1. As will be clear to the experts at this point, the proper way to organize calculations like the one above is by means of a spectral sequence. We do not pursue this for two reasons. First, it would take considerable effort and, secondly, it seems that the result would be very similar to the spectral sequence obtained by Forrest, Hunton and Kellendonk in [FHK]. There are some differences; the computation of [FHK] is actually computing the cohomology of the hull, which is a slightly different thing. Our version for K-theory would have the effect of ‘bundling together’ all the even and all the odd cohomology groups into the two K-theory groups.
728
I. F. Putnam
Acknowledgements. The author would like to thank the referee for a thorough reading of the manuscript and many helpful suggestions.
References [Be1] [Be2] [BCL] [BHZ] [BIT] [BS] [Cu] [DK1] [DK2] [E1] [E2] [FH] [FHK] [GS] [HG] [J] [KKL1] [KKL2] [KeP] [KrN] [Le] [LS] [M1] [M2] [Mo] [MRW] [Pe] [Pu]
Bellissard, J.: K-theory of C ∗ -algebras in solid state physics. In: Statistical Mechanics and Field Theory, Mathematical Aspects, Dorlas, T.C., Hugenholz, M.N., Winnink, M. (eds.), Lecture Notes in Physics, 257, Berlin-New York: Springer, 1986, pp. 99–156 Bellissard, J.: Noncommutative geometry for aperiodic solids. In: Geometric and Topological Methods for Quantum Field Theory (Villa de Leyva, 2001), River Edge, NJ: World Sci. Publishing, 2003, pp. 86–156 Bellissard, J., Contensou, E., Legrand, A.: K-theorie des quasi-crystaux, image par la trace: le cas du reseau octoganol, C.R. Acad. Sci. (Paris), t. 327, Serie I, 197–200 (1998) Bellissard, J., Herrmann, D., Zarrouati, M.: Hull of aperiodic solids and gap labelling theorems. In: Directions in Mathematical Quasicrystals, Baake, M.B., Moody, R.V. (eds.), CRM Monograph Series, Volume 13, Providence, RI: Amer. Math. Soc., 2000, pp. 207–259 Bellissard, J., Iochum, B., Testard, D.: Continuity properties of electronic spectrum of 1D quasicrystals. Commun. Math. Phys. 141, 353–380 (1991) Bellissard, J., Savinien, J.: A spectral sequence for the K-theory of tiling spaces. Eng. Th. Dyn. Syst. 29, 997–1031 (2009) Cuntz, J.: Generalized homomorphisms between C ∗ -algebras and KK-theory. In: Dynamics and Processes (Bielefeld, 1981), Blanchard, P., Streit, L. (eds.), Lecture Notes in Math. 1031, Berlin: Springer-Verlag, 1983, pp. 31–45 Duneau, M., Katz, A.: Quasiperiodic patterns. Phys. Rev. Lett. 54, 2688–2691 (1985) Duneau, M., Katz, A.: Quasiperiodic patterns and icosehedral symmetry. J. Phys (France) 47, 181–196 (1986) Elser, V.: Indexing problems in quasicrystal diffraction. Phys. Rev. B 32, 4892–4898 (1985) Elser, V.: Comment on “quasicrystals: a new class of ordered structures”. Phys. Rev. Lett. 54, 1730 (1985) Forrest, A., Hunton, J.: The cohomology and K-theory of commuting homeomorphisms of the cantor set. Erg. Th. Dyn. Syst. 19, 611–625 (1999) Forrest, A., Hunton, J., Kellendonk, J.: Topological invariants for projection method patterns. Mem. A.M.S. 159 (758) (2002) Grünbaum, B., Shephard, G.C.: Tilings and Patterns. New York: Freeman, 1987 Hibbert, F., Gratias, D. (eds.): Lectures on Quasicrystals. Les Ulis: Editions de Physique, 1994 Janot, C.: Quasicrystals. A Primer. Oxford: Clarendon, 1992 Kalugin, P.A., Kitaev, A.Y., Levitov, L.S.: AL 0.86 M N0.14 : a six-dimensional crystal. JETP Lett. 41, 145–149 (1985) Kalugin, P.A., Kitaev, A.Y., Levitov, L.S.: 6-dimensional properties of AL 0.86 M N 0.14 . J. Phys. Lett. (France) 46, L601–L607 (1985) Kellendonk, J., Putnam, I.F.: Tilings, C ∗ -algebras and K-theory. In: Directions in Mathematical Quasicrystals, Baake, M., Moody, R.V. (eds.), CRM Monograph Series, Volume 13, Providence, RI: Amer. Math. Soc., 2000, pp. 177–206 Kramer, P., Neri, R.: On periodic and non-periodic space fillings of E m obtained by projections. Acta. Cryst. A 40, 580–587 (1984), and (Erratum) Acta. Cryst. A 41, 619 (1985) Le, T.T.Q.: Local rules for aperiodic tilings. In: The Mathematics of Long Range Aperiodic Order, Moody, R.V. ed., Dordrecht: Kluwer, 1996, pp. 331–366 Levine, D., Steinhradt, P.J.: Quasicrystals: a new class of ordered structures. Phys. Rev. Lett. 53, 2477–2480 (1984) Meyer, Y.: Algebraic Numbers and Harmonic Analysis. Amsterdam: North-Holland, 1972 Meyer, Y.: Quasicrystals, Diophantine approximation and algebraic numbers. In: Beyond Quasicrystals, eds. Axel, F., Gratias, D., Berlin: Springer-Verlag, 1995, pp. 3–16 Moody, R.V.: Model sets: a survey. In: From Quasicrystals to More Complex Systems, eds. F. Axel, F. Dnoyer, J.P. Gazeau, Centre de physique Les Houches, Berlin-Heidelberg-New York: Springer-Verlag, 2000 Muhly, P.S., Renault, J.N., Williams, D.P.: Equivalence and isomorphism for groupoid C ∗ -algebras. J. Op. Th. 17, 3–22 (1987) Pedersen, G.K.: Analysis NOW. Graduate Texts in Mathematics, Vol. 118, Berlin-Heidelberg-New York: Springer-Verlag, 1988 Putnam, I.F.: On the K-theory of C ∗ -algebras of principal groupoids. Rocky Mountain J. Math. 28, 1483–1518 (1998)
C ∗ -Algebras from Model Sets Obtained by the Cut-and-Project Method
[PSS] [Ri] [SBGC] [Se]
729
Putnam, I.F., Schmidt, K., Skau, C.F.: C ∗ -algebras associated with denjoy homeomorphisms of the circle. J. Op. Th. 16, 99–126 (1986) Rieffel, M.A.: Applications of strong morita equivalence to transformation group C ∗ -algebras. Proc. Symp. Pure Math. 38, 299–310 (1982) Schechtman, D., Blech, I., Gratias, D., Cahn, J.W.: Metallic phase with long-range orientational order and no translational symmetry. Phys. Rev. Lett 51, 1951–1953 (1984) Senechal, M.: Quasicrystals and Discrete Geometry. Cambridge: Cambridge University Press, 1995
Communicated by A. Connes
Commun. Math. Phys. 294, 731–744 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0973-3
Communications in
Mathematical Physics
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit Christian Bär, Frank Pfäffle Universität Potsdam, Institut für Mathematik, Am Neuen Palais 10, 14469 Potsdam, Germany. E-mail:
[email protected];
[email protected] Received: 7 May 2008 / Accepted: 23 October 2009 Published online: 10 January 2010 – © Springer-Verlag 2010
Abstract: Let H = 2 L + V , where L is a self-adjoint Laplace type operator acting on sections of a vector bundle over a compact Riemannian manifold and V is a symmetric endomorphism field. We derive an asymptotic expansion for the heat kernel of H as 0. As a consequence we get an asymptotic expansion for the quantum partition function and we see that it is asymptotic to the classical partition function. Moreover, we show how to bound the quantum partition function for positive by the classical partition function. 1. Introduction In this paper we study semi-classical approximations for the heat kernel of a general self-adjoint Laplace type operator in a geometric context. More precisely, let M be an n-dimensional compact Riemannian manifold without boundary and let E → M be a Riemannian or Hermitian vector bundle. Let L be a self-adjoint Laplace-type operator with smooth coefficients acting on sections of E. Important examples of such operators are the Laplace-Beltrami operator acting on functions, more generally, the HodgeLaplacian acting on differential forms, and the square of the Dirac operator acting on spinors. We fix a symmetric endomorphism field V (the potential) which need not be a scalar multiple of the identity. For any > 0 we consider the self-adjoint operator H := 2 L + V . One is now interested in the behavior of H as 0. The solution operator e−t H for the heat equation ∂u ∂t + Hu = 0 has a smooth integral kernel k(x, y, t, ) which we briefly call the heat kernel. Our main result, Theorem 3.3, states that there is an asymptotic expansion k(x, y, t, )
0
∼
χ (d(x, y)) · q(x, y, t2 ) ·
∞ (t2 ) j · ϕ j (x, y, t). j=0
Here d(x, y) denotes the Riemannian distance of x and y, χ is a suitable cut-off function, and q is explicitly given by the Euclidean heat kernel, q(x, y, τ ) = (4π τ )−n/2 ·
732
C. Bär, F. Pfäffle
exp −d(x, y)2 /4τ . The ϕ j are smooth sections which can be determined recursively by solving appropriate transport equations. Theorem 3.3 is optimal in the sense that the asymptotic expansion holds for all derivatives with respect to x, y, and t and is uniform in x, y ∈ M and t ∈ (0, T ] for any T > 0. Now fix t > 0 and define the quantum partition function Z Q () := Tr(e−t H ). As a direct consequence of Theorem 3.3 we get Corollary 4.1, Z Q ()
0
∼
∞ √ −n (2 π t) · a j (t) · (t2 ) j .
(1)
j=0
This result has been proved in [10, Sec. 2] by technically rather involved methods from pseudo-differential calculus. The corresponding classical partition function is given by Z C () = (2π )−n T ∗ M tr exp(−t (| p|2 · id + V (x))) dpdx. We directly obtain Corollary 4.2: lim
0
Z Q () = 1. Z C ()
(2)
This asymptotic statement corresponds to the following bound valid for small positive (Corollary 4.7): √ v0,n t2 Z Q () 2 √ . ≤ c3 · ec2 ·t · Z C () v t2 K ,n
Here v K ,n (r ) denotes the volume of the geodesic ball of radius r in the n-dimensional model space of constant curvature K . The constant c3 depends only on the dimension n of M, the constant c2 depends on n, on a lower bound for the Ricci curvature, and on a lower bound for the potential of L. In the case that the underlying manifold is Euclidean space and the operator L is the classical Laplace operator acting on functions, the optimal inequality Z Q () ≤ 1 Z C () was independently obtained by Golden [5], Symanzik [14], and Thompson [15]. In this classical situation (2) can be found in Simon’s book [13, Thm. 10.1]. In the case of a general compact manifold the asymptotic expansion of Z Q () is contained in [10, Prop. 2.4]. Our methods are rather direct and explicit. We use standard facts from geometric analysis. The estimates on the quantum partition function are based on the Golden-Thompson inequality, the Hess-Schrader-Uhlenbrock estimate and an estimate on the heat kernel of the Laplace-Beltrami operator due to Schoen and Yau. No pseudo-differential calculus or microlocal analysis are needed. 2. The Heat Kernel Let M be an n-dimensional compact Riemannian manifold without boundary. In local coordinates the Riemannian metric is denoted by gi j , its inverse matrix by g i j .
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit
733
Let E → M be a Riemannian or Hermitian vector bundle. We denote the metric on E by ·, · . Let L be a self-adjoint Laplace-type operator with smooth coefficients acting on sections of E. In local coordinates x = (x1 , . . . , xn ) and with respect to a local trivialization of E we have L = −g i j (x)∂i ∂ j + bk (x)∂k + c(x). Here the Einstein summation convention is understood, bk (x) and c(x) are matrices depending smoothly on x, and ∂i = ∂x∂ i . It is well-known [1, Prop. 2.5] that L can globally be written in the form L = ∇ ∗ ∇ + W, where ∇ is a metric connection on E, ∇ ∗ is its formal L 2 -adjoint, and W is a smooth section of Sym(E), the bundle of fiberwise symmetric endomorphisms of E. Example 2.1. Let E be the trivial line bundle, i. e. sections of E are nothing but functions. Let the connection be the usual derivative, ∇ = d, and let W = 0. Then L = = d ∗ d = −div ◦ grad, the Laplace-Beltrami operator. Example 2.2. More generally, let E = p T ∗ M be the bundle of p-forms. Then the Hodge-Laplacian L = d ∗ d + dd ∗ is a self-adjoint Laplace-type operator. Here ∇ is the connection induced on E by the Levi-Civita connection. For example, for p = 1, the Bochner formula says L = ∇ ∗ ∇ + Ric, see [2, p. 74, formula (2.51)] or [7, Chap. 2, Cor. 8.3]. Example 2.3. Let M carry a spin structure and let E be the spinor bundle. If D is the Dirac operator, then by the Lichnerowicz-Schrödinger formula L := D 2 = ∇ ∗ ∇ + 14 scal, see [9] or [7, Chap. 2, Thm. 8.8]. Now fix another section V ∈ C ∞ (M, Sym(E)). For > 0 we define the self-adjoint operator H := 2 L + V.
(3)
For t > 0 we use functional calculus to define the operator e−t H as a bounded selfadjoint operator on the Hilbert space of square integrable sections of E, L 2 (M, E). For any u 0 ∈ L 2 (M, E) we can put u(x, t) := (e−t H u 0 )(x) and we get the unique solution to the heat equation ∂u + Hu = 0 ∂t subject to the initial condition u(x, 0) = u 0 (x). By elliptic theory e−t H is smoothing and its Schwartz kernel k(x, y, t, ) depends smoothly on all variables x, y ∈ M, t, > 0, see [1, Sec. 2.7]. By E E ∗ → M × M we denote the exterior tensor product bundle of E with its dual bundle E ∗ . Its fiber over (x, y) ∈ M × M is given by (E E ∗ )(x,y) = E x ⊗ E y∗ =
734
C. Bär, F. Pfäffle
Hom(E y , E x ). Note that for fixed t and the heat kernel k(·, ·, t, ) is a section of E E ∗ . We define d(x, y)2 −n/2 , · exp − q : M × M × (0, ∞) → R, q(x, y, t) := (4π t) 4t where d(x, y) denotes the Riemannian distance of x and y. For technical reasons we define M M := M × M \ {(x, y) | x and y are cut-points}. Recall that x and y are cut-points if either there are several geodesics of minimal length joining x and y or there is a Jacobi field along the unique shortest geodesic which vanishes at x and y. For example, on the standard sphere cut points are exactly antipodal points. One always has that M M contains the diagonal and is an open dense subset of M × M. The function q is smooth on M M × (0, ∞). We will often abbreviate r := d(x, y). Then r 2 is a smooth function on M M and r itself is smooth on M M away from the diagonal. By gradx we denote the gradient with respect to the variable x, similarly for x , H,x , and ∇x . Straightforward computation yields q gradx (r 2 ), 4t 2 r x (r 2 ) , + x q = −q · 4t 2 4t 2 r ∂q n =q· , − ∂t 4t 2 2t ∂ 2n + x (r 2 ) + x q = −q · . ∂t 4t gradx q = −
(4) (5) (6) (7)
3. The Formal Heat Kernel Now we make the following ansatz for a formal heat kernel of H over M M: ˆ y, t, ) = q(x, y, t2 ) · k(x,
∞
(t2 ) j · ϕ j (x, y, t).
(8)
j=0
Lemma 3.1. There are unique continuous sections ϕ j over M M × [0, ∞), smooth over M M × (0, ∞), such that (i) ϕ0 (y, y, 0) = id E y for all y ∈ M, (ii) ∂t∂ + H,x kˆ = 0. Assertion (i) means that for fixed y and any partial sum of the formal heat kernel kˆ converges to the delta function at y since this is clearly the case for q. Assertion (ii) is to be understood in the sense that the formal series in the definiton (8) of kˆ is differentiated termwise and then regrouped by powers of .
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit
735
Proof of Lemma 3.1. Using (4), (7), and ∇ ∗ ∇( f ϕ) = ( f )ϕ − 2∇grad f ϕ + f ∇ ∗ ∇ϕ we compute ∂ + H,x kˆ ∂t ∞ ∂ ∂ = + 2 x (q(x, y, t2 )) · (t2 ) j · ϕ j + q · ((t2 ) j · ϕ j ) ∂t ∂t j=0
− 22 · (t2 ) j · ∇gradx q ϕ j + q · (t2 ) j · H,x ϕ j ∞ 2n + x (r 2 ) j ∂ 2 j =q· (t ) −2 ϕj + ϕj + ϕj 2 4t t ∂t j=0 − 22 ∇ −1 grad (r 2 ) ϕ j + 2 L x ϕ j + V ϕ j 4t 2
=q·
∞ j=0
x
2n + x (r 2 ) ∂ j−1 2 j − ϕ j + jϕ j + t ϕ j t 4 ∂t
+ ∇ 1 grad (r 2 ) ϕ j + 2 t L x ϕ j + t V ϕ j x 2 ∞ 2n + x (r 2 ) ∂ j− + tV ϕj + t ϕj =q· t j−1 2 j 4 ∂t j=0 + r ∇gradx (r ) ϕ j + L x ϕ j−1 . Thus assertion (ii) is equivalent to the recursive transport equations ∂ 2n + x (r 2 ) + t V ϕ j = −L x ϕ j−1 t ϕ j + r ∇gradx (r ) ϕ j + j − ∂t 4
(9)
for j = 0, 1, . . . where we use the convention ϕ−1 ≡ 0. The function G := (2n + x (r 2 ))/4 appearing in (9) is smooth on M M and vanishes on the diagonal {r = 0}. We observe that the transport equation (9) is an ordinary differential equation along ∂ the integral curves of the vector field t ∂t∂ + r ∂r in the r -t-surface, see Fig. 1. More precisely, if we fix an angle θ and put t = cos(θ )·s and r = sin(θ )·s, then (9) translates into ∂ ϕ j + ( j − G + cos(θ )sV ) ϕ j = −L x ϕ j−1 . (10) ∂s Here we identify the fibers of E by parallel transport along the radial geodesics γ ema∂ . nating from y so that r ∇gradx (r ) becomes identified with r ∂r Since this differential equation is singular at s = 0 we need to introduce integrating factors. In order to write them down we solve the linear ODE, d A(s) = A(s) · cos(θ ) · V (γ (sin(θ )s)), A(0) = id, (11) ds in the space of endomorphisms of E y . Of course, A depends smoothly on all data such as s, θ , y = γ (0), and γ˙ (0). As long as A(s) is regular the determinant of A(s) satisfies the linear ODE, d det(A(s)) = det(A(s)) · cos(θ ) · tr(V (γ (sin(θ )s))), ds s
736
C. Bär, F. Pfäffle
Fig. 1. Various parameters and their relation
hence det(A(s)) cannot vanish anywhere since it would otherwise have to be identically 0 contradicting the initial condition det(A(0)) = 1. Thus A(s) remains invertible for all s. We now define s G dσ R j (s) := s j · exp − · A(s). σ 0 Since G is smooth and vanishes at s = 0 the integrand G/σ is smooth along the segment parametrized by [0, s]. Direct computation shows that (10) is equivalent to ∂ (R j ϕ j ) = −s −1 R j L x ϕ j−1 . ∂s For j = 0 this means that R0 ϕ0 = C0 is constant, i. e. ϕ0 = C0 ·R0−1 = C0 · exp A(s)−1 . The initial condition (i) is now equivalent to C0 = 1. Thus we have
s
ϕ0 = exp 0
G dσ σ
(12) s 0
· A(s)−1 .
For j ≥ 1 we note that s −1 R j is smooth also at s = 0 and we get R j ϕ j = − R j L x ϕ j−1 dσ + C j . Evaluation at s = 0 shows C j = 0. We therefore have ϕ j = −R −1 j
s
G dσ σ
·
(13) s 0
σ −1
σ −1 R j L x ϕ j−1 dσ.
0
We have established uniqueness of the ϕ j . As to existence we only need to ensure that (13) and (3) define smooth sections over M M × (0, ∞). For (x, y) ∈ M M and r ∈ [0, 1] we let xy : [0, 1] → M be the unique shortest geodesic with xy(0) = y and xy(1) = x. In other words, in terms of the Riemannian exponential map xy(r ) = expy (r exp−1 y (x)). The map M M × [0, 1] → M, (x, y, r ) → xy(r ), is smooth. Substituting σ = us, Eq. (13) can be rewritten as ϕ0 (x, y, t) = exp 0
1
u −1 G(xy(u), y) du · A( d(x, y)2 + t 2 )−1
(14)
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit
737
and (3) becomes ϕ j (x, y, t) −1
1
= −R j (x, y, t)
u −1 R j (xy(u), y, ut) · (L x ϕ j−1 )(xy(u), y, ut)du,
0
where R j (x, y, t)
= (d(x, y)2 + t 2 ) j/2 · exp −
1 0
G(xy(u), y) du u
· A( d(x, y)2 + t 2 ).
This shows smoothness of the ϕ j on M M × (0, ∞) and continuity on M M × [0, ∞). Remark 3.2. If θ = 0, i.e., if s = t and x = y, then (11) becomes a linear ODE with constant coefficients, d A(s) = A(s) · V (y), ds
A(0) = id,
and can be solved explicitly, A(s) = exp(sV (y)). In particular, since G(yy(u), y) = G(y, y) = 0, (14) becomes ϕ0 (y, y, t) = exp(−t V (y)).
(15)
Construction of the approximate kernel. Now we fix η > 0 smaller than the injectivity radius of M. This means that {r ≤ η} ⊂ M M. We choose a smooth cutoff function χ : R → R such that χ ≡ 1 on (−∞, η/2] and χ ≡ 0 on [η, ∞). For N ∈ N we set kˆ (N ) (x, y, t, ) := χ (d(x, y)) · q(x, y, t2 ) ·
N
(t2 ) j · ϕ j (x, y, t).
(16)
j=0
Note that kˆ (N ) is a smooth section of E E ∗ over M × M × (0, ∞) × (0, ∞) and not just over M M × (0, ∞) × (0, ∞). Theorem 3.3. Let M be an n-dimensional compact Riemannian manifold without boundary. Let E → M be a Riemannian or Hermitian vector bundle and let H be as in (3). Let k(x, y, t, ) be the heat kernel of H. Let T > 0 and , j ∈ N0 . For N > n + + j let kˆ (N ) (x, y, t, ) be the approximate heat kernel as defined in (16). Then j ∂ ˆ (N ) k − k sup ∂t j
t∈(0,T ]
C (M×M)
= O(2N +1−2n−2 −2 j ) ( 0).
738
C. Bär, F. Pfäffle
Proof. By construction of kˆ (N ) we have (using that q = O((t2 )−n/2 ) and q = O((t2 )∞ ) in the region where χ (d(x, y)) is not locally constant): ∂ + H,x (kˆ (N ) − k) r N := ∂t ∂ + H,x (kˆ (N ) ) = ∂t ⎞ ⎛ N ∂ + H,x ⎝q · = χ (d(x, y)) · (t2 ) j · ϕ j ⎠ + O((t2 )∞ ) ∂t j=0
= χ (d(x, y)) · q · t = O(t N −n/2 2N +2−n ).
N 2N +2
L x ϕ N + O((t2 )∞ )
Similarly, for any m ∈ N we get (H,x )m (r N ) = O(t N −n/2−2m 2N +2−n−2m ) and (L y )m (r N ) = O(t N −n/2−2m 2N +2−n−4m ). By the choice of the initial condition kˆ (N ) − k vanishes at t = 0. Thus uniqueness of the solution to the heat equation (Duhamel’s principle) implies
t (N ) ˆ (k − k)(x, y, t, ) = e−(t−s)H,x r N (·, y, s, )ds. 0
The spectrum of L is bounded from below, hence H = 2 L + V ≥ −c1 for all ≤ 1, where c1 is a suitable positive constant. Thus we have for the L 2 -L 2 -operator norm, e−(t−s)H,x L 2 ,L 2
≤
e c1 t .
Therefore we get for all t ∈ (0, T ], (kˆ (N ) − k)(·, y, t, ) L 2 (M) ≤
t
0
≤
t
e−(t−s)H,x L 2 ,L 2 · r N (·, y, s, ) L 2 (M) ds ec1 t · c2 · s N −n/2 2N +2−n ds
0
≤ c3 (T ) · 2N +2−n . Furthermore, L x (kˆ (N ) − k)(·, y, t, ) L 2 (M) = −2 (H,x − V ))(kˆ (N ) − k)(·, y, t, ) L 2 (M)
t −2 −(t−s)H,x ≤ H,x e r N (·, y, s, )ds 2
L (M)
0
+ c4
−2
ˆ (N )
(k
− k)(·, y, t, ) L 2 (M)
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit
739
t −(t−s)H,x ≤ −2 e H r (·, y, s, )ds ,x N 0
L 2 (M)
+ OT (2N −n )
≤ −2 · c5 (T ) · H,x r N (·, y, s, ) L 2 (M) + OT (2N −n ) = OT (2N −n−2 )
and t −(t−s)H,x e L r (·, y, s, )ds L y (kˆ (N ) − k)(·, y, t, ) L 2 (M) = y N 0
= OT (
2N −2−n
L 2 (M)
).
Here the lower index T in OT (· · · ) indicates that the constant bounding the OT (2N −n−2 )-term by 2N −n−2 depends on T . Integration with respect to y yields (L x + L y )(kˆ (N ) − k)(·, ·, t, ) L 2 (M×M) = OT (2N −2−n ). Inductively, we get (L x + L y )m (kˆ (N ) − k)(·, ·, t, ) L 2 (M×M) = OT (2N +2−n−4m ). By the elliptic estimates we have for the Sobolev norms (kˆ (N ) − k)(·, ·, t, ) H 2m (M×M) = OT (2N +2−n−4m ) and by the Sobolev embedding theorem (kˆ (N ) − k)(·, ·, t, )C (M×M) = OT (2N +1−2n−2 ). Similarly, we have r N (·, ·, t, )C (M×M) = OT (2N +1−2n−2 ). It remains to control the t-derivatives. We compute ∂ ˆ (N ) (k − k) = −H,x (kˆ (N ) − k) + r N ∂t = −2 L x (kˆ (N ) − k) − V · (kˆ (N ) − k) + r N , thus ∂ (N ) (kˆ − k) ∂t
C (M×M)
≤ · c6 · kˆ (N ) − kC +2 (M×M) + c7 · kˆ (N ) − kC (M×M) + r N C (M×M) 2
= OT (2N +1−2n−2 −2 ). An induction finally proves the theorem.
740
C. Bär, F. Pfäffle
Remark 3.4. We were somewhat generous in the application of the Sobolev embedding theorem. With a little more care we can improve the statement of Theorem 3.3 as follows: If in addition to T, , j and N we are given ε > 0, then j ∂ ˆ (N ) k − k sup = O(2N +2−ε−2n−2 −2 j ) ( 0). ∂t j t∈(0,T ] C (M×M) Remark 3.5. It would be nice to extend Theorem 3.3 to operator families of the form H = 2 L + D + V, where D is a formally self-adjoint differential operator of first order. However, this seems not to work. The resulting transport equations can no longer be solved. There is a different approximation in [10] working also for D = 0. It is less explicit and makes heavy use of asymptotic expansions of total symbols of pseudodifferential operators. For the trace of the heat operator it seems to give the same result. We discuss this in the next section. 4. The Classical and Quantum Partition Functions Applying Theorem 3.3 with j = = 0 we have in particular |k(x, y, t, ) − kˆ (N ) (x, y, t, )|
≤
C · 2N +1−2n
for all t ∈ (0, T ], x, y ∈ M, and for 0. We put x = y, take the pointwise trace and integrate over M to obtain Corollary 4.1. Let M be an n-dimensional compact Riemannian manifold without boundary. Let E → M be a Riemannian or Hermitian vector bundle and let H be as in (3). Let T > 0. Then we have an asymptotic expansion Tr(e−t H )
∼
∞ √ (2 π t)−n · a j (t)t j · 2 j
( 0)
j=0
uniform in t ∈ (0, T ]. Proof. This follows directly from Theorem 3.3 together with q(y, y, t2 ) = (4π t2 )−n/2 and a j (t) = M tr(ϕ j (y, y, t)) dy. We fix t > 0 and we define the quantum partition function Z Q () := Tr(e−t H ) and the corresponding classical quantity
−n Z C () := (2π ) tr exp(−t (| p|2 · id + V (x))) dpdx. T∗M
(17)
(18)
If E is the trivial line bundle and V is the potential energy this is the partition function in statistical mechanics with t = 1/k B T , where T is the temperature and k B is Boltzmann’s constant. For convenience, we call it the classical partition function also in our more general situation. Then we have
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit
741
Corollary 4.2. Let M be an n-dimensional compact Riemannian manifold without boundary. Let Z Q () and Z C () be defined as in (17) and (18). Then lim
0
Z Q () = 1. Z C ()
Proof. Using Rn exp(−t| p|2 ) dp = (π/t)n/2 and (15) we get
−n Z C () = (2π ) · tr exp(−t| p|2 ) exp(−t · V (x)) dpdx T∗M
−n = (2π ) · (π/t)n/2 · tr(exp(−t · V (x))) dx M
√ = (2 π t)−n · tr(ϕ0 (x, x, t)) dx M √ = (2 π t)−n · a0 (t). The assertion follows from Corollary 4.1.
In the remainder of this section we will contrast this asymptotic comparison of Z Q () and Z C () with an inequality of the two partition functions which works for positive . For this we need the following version of the Golden-Thompson inequality (see [4,8] or [12]). Lemma 4.3. Let B and C be self-adjoint operators on a Hilbert space H , both bounded from below and such that B + C is essentially self-adjoint on the intersection dom(B) ∩ dom(C) of the domains of B and C. Then Tr (exp(−(B + C))) ≤ Tr (exp(−B) exp(−C)) . We will also need the following elementary assertion. Lemma 4.4. Let A1 and A2 be complex N × N -matrices. Let A2 be Hermitian and nonnegative. Then |tr (A1 A2 ) | ≤ |A1 | · tr (A2 ) ,
(19)
where |A1 | denotes the operator norm of A1 . Proof. Since A2 is Hermitian and nonnegative there is a nonnegative Hermitian matrix B such that A2 = B 2 . For the standard basis e1 , . . . , e N of C N and the standard Hermitian scalar product (·, ·) we get N N |tr (A1 A2 ) | = |tr (B A1 B) | = (B A1 Be j , e j ) = (A1 Be j , Be j ) j=1 j=1 ≤
N j=1
|A1 | · |Be j |
2
N = |A1 | · (B 2 e j , e j ) = |A1 | · tr (A2 ) . j=1
742
C. Bär, F. Pfäffle
Proposition 4.5. Let M be an n-dimensional compact Riemannian manifold without boundary. For any x ∈ M and r > 0, let ωx (r ) denote the volume of the geodesic ball about x with radius r . Then there are constants c1 > 0 and c2 ∈ R such that for any t > 0 and any > 0 one has
tr e−t V (x) c2 ·t 2 dx. (20) Z Q () ≤ c1 · e √ M ωx ( t2 ) Proof. We apply Lemma 4.3 with B = t2 ∇ ∗ ∇ and C = t (V + 2 W ) and use the ∗ ∗ notations kt∇ ∇ (x, y) and kt (x, y) for the kernels of the heat operators e−t∇ ∇ and e−t respectively, where is the Laplace-Beltrami operator acting on functions: 2 ∗ 2 Z Q () ≤ Tr e−t ∇ ∇ e−t (V + W )
∗ 2 = tr kt∇2∇ (x, x)e−t (V (x)+ W (x)) dx
M (19) 2 ∇∗∇ ≤ kt 2 (x, x) · tr e−t (V (x)+ W (x)) dx
M 2 ≤ kt2 (x, x) tr e−t (V (x)+ W (x)) dx, M
where for the last inequality we have the Hess-Schrader-Uhlenbrock inequality, i.e. ∇∗∇ kt (x, y) ≤ kt (x, y) for any x, y ∈ M and t > 0 (see [6]). Applying Lemma 4.3 once more (with B = t V (x) and C = t2 W (x)), using Lemma 4.4 and choosing a w0 ∈ R with W ≥ w0 leads to
2 Z Q () ≤ kt2 (x, x) tr e−t V (x) e−t W (x) dx
M 2 ≤ (21) kt2 (x, x) e−t w0 tr e−t V (x) dx. M
From [11, Thm. 4.6] we get a pointwise estimate for the heat kernel of the Laplace-Beltrami operator: There are constants c1 , c > 0 such that for any τ > 0 and any x ∈ M one has kτ (x, x) ≤ c1 ·
1 √ · ec·τ . ωx ( τ )
Inserting this into (21) and taking c2 = c − w0 yields the claim.
(22)
Remark 4.6. The constants c1 , c2 can be given explicitly: Let α > 1, δ > 0 and n = dim(M), let κ > 0 with Ric ≥ −κ and let w0 ∈ R with W ≥ w0 . In (22) one can αn and c = α−1 take c1 = (1 + δ)nα exp 1+α · κ · δ (compare [11, Thm. 4.6]) and then δ c2 = c − w0 . Given K ∈ R and r ∈ (0, √π ) (where we use the convention √π = ∞ for K ≤ 0) K K let v K ,n (r ) denote the volume of a geodesic ball of radius r in the n-dimensional model space of constant curvature K . Recall that this model space is hyperbolic space, Euclidean space, or the sphere with their appropriately scaled standard metrics.
Asymptotic Heat Kernel Expansion in the Semi-Classical Limit
743
Corollary 4.7. Let M be an n-dimensional compact Riemannian manifold without boundary, and let K ∈ R be an upper bound for the sectional curvature of M and let ι denote the injectivity radius of M. Then there are constants c2 , c3 ∈ R such that for any t, > 0 2 with t2 < πK and t2 < ι2 we have √ v t2 0,n Z Q () 2 √ . ≤ c3 · ec2 ·t · Z C () v t2 K ,n
Proof. For r < min{ι, √π } the Bishop-Günther Theorem [3, Thm. 3.7] states ωx (r ) ≥ K v K ,n (r ) for all x ∈ M. Inserting this into (20) yields
1 2 √ · tr e−t V (x) dx Z Q () ≤ c1 · ec2 ·t · M v K ,n t2 = c1 · ec2 ·t · 2
for any t, > 0 with t2 < proof.
π2 K
v K ,n
1 √
t2
√ · (2 π t)n · Z C ()
and t2 < ι2 . Putting c3 := c1 ·
√ (2 π )n v0,n (1)
concludes the
Remark 4.8. Even optimal choices of α and δ yield a constant c3 which is much larger than 1. Therefore Corollary 4.7 does not even imply half of Corollary 4.2, namely lim
0
Z Q () ≤ 1. Z C ()
We do not know whether Corollary 4.7 holds with c3 = 1. Acknowledgements. We would like to thank Markus Klein for very helpful discussions on the topics of this paper and SFB 647 funded by Deutsche Forschungsgemeinschaft for financial support.
References 1. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Berlin: Springer, 1992 2. Besse, A.L.: Einstein Manifolds. Berlin: Springer, 1987 3. Chavel, I.: Riemannian Geometry: A Modern Introduction. Cambridge: Cambridge University Press, 1993 4. Elworthy, K.D., Ndumu, M.N., Trumam, A.: An elementary inequality for the heat kernel on a Riemannian manifold and the classical limit of the quantum partition function. Pitman Res. Notes Math. Ser., 150, Harlow: Longman Sci. Tech., 1986 5. Golden, S.: Lower bounds for the Helmholtz function. Phys. Rev. B 137, 1127–1128 (1965) 6. Hess, H., Schrader, R., Uhlenbrock, D.A.: Kato’s inequality and the spectral distribution of Laplacians on compact Riemannian manifolds. J. Diff. Geom. 15, 27–37 (1980) 7. Lawson, B.H., Michelsohn, M.-L.: Spin Geometry. Princeton, NJ: Princeton University Press, 1989 8. Lenard, A.: Generalization of the Golden-Thompson inequality Tr(e A e B ) ≥ Tr e A+B . Indiana Univ. Math. J. 21, 457–467 (1971/1972) 9. Lichnerowicz, A.: Spineurs harmoniques. C. R. Acad. Sci. Paris 257, 7–9 (1963) 10. Schrader, R., Taylor, M.E.: Small asymptotics for quantum partition functions associated to particles in external Yang-Mills potentials. Commun. Math. Phys. 92(4), 555–594 (1994) 11. Schoen, R., Yau, S.-T.: Lectures on Differential Geometry. Cambridge, USA: International Press, 1994
744
C. Bär, F. Pfäffle
12. Simon, B.: Trace Ideals and their Applications. Second edition. Providence, RI: Amer. Math. Soc. 2005 13. Simon, B.: Functional Integration and Quantum Physics. Second edition. AMS Chelsea Publishing, Providence, RI: Amer. Math. Soc., 2005 14. Symanzik, K.: Proof and refinements of an inequality of Feynman. J. Math. Phys. 6, 1155–1156 (1965) 15. Thompson, C.J.: Inequality with applications in statistical mechanics. J. Math. Phys. 6, 1812–1813 (1965) Communicated by A. Connes
Commun. Math. Phys. 294, 745–760 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0970-6
Communications in
Mathematical Physics
Existence, Regularity, and Properties of Generalized Apparent Horizons Michael Eichmair Department of Mathematics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA. E-mail:
[email protected] Received: 25 June 2008 / Accepted: 17 October 2009 Published online: 25 December 2009 – © Springer-Verlag 2009
Abstract: We prove a conjecture of Tom Ilmanen and Hubert Bray on the existence of the outermost generalized apparent horizon in an initial data set and that it is outer area minimizing. We use the features of the construction in [E07] to prove the following conjecture in [BK09]: Conjecture 1. (Hubert Bray, Tom Ilmanen, 2006) Let (M n , g, p) be a complete asymptotically flat initial data set which contains a generalized trapped surface, and let 2 ≤ n ≤ 7. Then there exists a unique outermost generalized trapped surface n−1 ⊂ M n . Moreover, this n−1 is a generalized apparent horizon and it is outer area minimizing (every hypersurface which encloses it has at least its area). 1. Introduction, Overview, Notation In this paper, we consider initial data sets of general relativity, by which we will mean triples (M n , g, p) consisting of a complete oriented Riemannian manifold (M n , g), whose regularity is at least C 3 , together with a symmetric (0, 2)-tensor p that is required to be C 1,γ for some γ ∈ (0, 1). We will always assume that n ≥ 2. An initial data set is said to be asymptotically flat in the literature if the complement of a compact set in M n consists of a disjoint union of Euclidean ends {N1 , . . . , N p }, each diffeomorphic to Rn \B(0, 1) with appropriate decay of the metric tensor gi j to the Euclidean metric δi j and the second fundamental form pi j to 0 in these coordinate charts. For the purpose of Conjecture 1, the following weak decay conditions are sufficient: |gi j − δi j | + |x||∂k gi j | = O(|x|− p ) and | pi j | = O(|x| for some powers p > 0 and q > 1.
−q
)
(1)
746
M. Eichmair
Definition 1.1 ([BK09]). The compact embedded C 2 -hypersurface n−1 in an asymptotically flat initial data set (M n , g, p) is a generalized trapped surface with respect to a chosen end Nk , if it is the boundary of an open set U ⊂ M n which contains the ‘points at infinity’ of all asymptotically flat ends but the chosen one, and if H ≤ | tr ( p)|. If n−1 satisfies H = | tr ( p)| then n−1 is called a generalized apparent horizon. Here, the mean curvature is computed with respect to the unit normal ‘pointing to infinity’ so that when U = {x : |x| > R} ⊂ Rn one has H = (n − 1)/R. The decay conditions in (1) are chosen so that if U is the complement of a large coordinate ball in an asymptotically flat manifold, then its boundary is generalized untrapped in the sense that H∂U > | tr ∂U ( p)|. Note also that by virtue of their definition, there is a partial ordering on the collection of all generalized trapped surfaces with respect to a chosen end Nk : if = ∂U and = ∂U then lies outside if U includes U . The term “outermost” in Conjecture 1 is understood in this sense. Conjecture 1 is well-known to be true when p ≡ 0, see [HI01, Sect. 4] and the references therein. In this case, generalized apparent horizons are minimal surfaces which put variational methods at one’s disposal. However, unless p is a constant multiple of the metric, (generalized) apparent horizons are not known to arise as critical points of an elliptic variational problem, which complicates their existence and regularity theory. In their celebrated proof of the spacetime positive mass theorem [SY81], R. Schoen and S.-T. Yau observed that apparent horizons can appear as an obstruction to proving the existence of entire solutions of Jang’s equation on asymptotically flat initial data sets. Schoen proposed that this ‘defect’ of Jang’s equation can be turned into a device to prove existence of apparent horizons. Using this observation, L. Andersson and J. Metzger [AM07] proved that closed apparent horizons exist between inner and outer trapped surfaces. The Plateau problem for apparent horizons was settled in [E07]. The method in [E07] shows that the apparent horizons appearing in the construction have a certain almost minimizing property (in the sense of Almgren) which likens them with minimal and constant mean curvature surfaces. Explicit bounds for their hypersurface measure in all dimensions and strong geometric estimates are a direct consequence of this property. Using very different techniques, a lower bound on the “outer injectivity radius” for a certain class of closed apparent horizons was derived in [AM07] by a delicate surgery procedure based on a priori curvature estimates. Such curvature estimates were obtained by the authors from stability and the Gauss-Bonnet theorem in [AM05], and then used for the surgery procedure to derive area bounds for certain 2-dimensional horizons. These area bounds and the estimate on the outer injectivity radius were then applied in [AM07] to show that the boundary of the trapped region of a 3-dimensional initial data set is smooth and embedded. This important result of L. Andersson and J. Metzger is the analogue of Theorem 1.2 for marginally outer trapped surfaces in dimension n = 3. In this paper we adapt the features of the construction in [E07] to generalized apparent horizons to prove Conjecture 1. Our methods here work in all dimensions n ≤ 7 and can be used to extend the result on the trapped region in [AM07] to these dimensions, and without recourse to their surgery procedure. A variant of the standard calibration argument used in [E07, Example A.1] to establish the almost minimizing property shows
Generalized Apparent Horizons
747
that the outermost generalized apparent horizons is indeed outer area minimizing, as was conjectured by Bray and Ilmanen. Our basic existence result for generalized apparent horizons is as follows: Theorem 1.1. Assume that 2 ≤ n ≤ 7 and let (M n , g, p) be a complete initial data set. ˙ 2 Let ⊂ M n be a bounded open subset with embedded C 2 -boundary ∂ = ∂1 ∪∂ such that H∂1 > | tr ∂1 ( p)| (mean curvature with respect to the normal pointing out of ) and H∂2 ≤ | tr ∂2 ( p)| (with respect to the normal pointing into ). Then there exists ˙ n−1 , where n−1 is a closed embedded an open set ⊂ such that ∂ = ∂1 ∪ 2,α C -hypersurface satisfying H = | tr ( p)| (normal pointing into ) with uniform n estimates depending only on Hn−1 (∂1 ), | p|C 1 () ¯ , and the local geometry of (M , g). n−1 intersects with a component of ∂2 , then these comIf a connected component of ¯ and ponents coincide. Moreover, n−1 minimizes area with respect to variations in it is C-almost minimizing in all of for a constant C = C(| p|C () ¯ ). The existence statement in Theorem 1.1 is the analogue for generalized apparent horizons of [AM07, Theorem 3.1]. Our proof of Theorem 1.1 in Sect. 4 is based on the Perron method used in [E07] together with an approximation argument. The C-almost miniziming property with the constant C depending only on | p|C () ¯ and the outer area minimizing property in the preceding theorem are important features of this approach and at the heart of our proof of Conjecture 1. A summary of results related to what we call the C-almost minimizing property here and the class FC of boundaries with this property is given in [E07, App. A] with concise references to the geometric measure theory literature. We derive the C 2,α -estimates from results in [SS81], which we appropriate to our context in Appendix A. The robust ‘low order approach’ to regularity used here via geometric measure theory and the stability based analysis of [SS81] is available also when n ≥ 8 if we accept thin singular sets. It provides a satisfactory theory for limits of regular embedded horizons in arbitrary dimensions that is friendly towards analysis, see Remark A.3 for details, and compare with the curvature estimates that were obtained in [AM05] by generalizing the iteration method of [SSY75]. The modification of Jang’s equation used in this paper can be applied to find solutions to the Plateau problem for generalized apparent horizons as in [E07]. The outer area minimizing property of generalized apparent horizons is appealing from the point of view of a spacetime Penrose inequality, and we refer the reader to the paper [BK09] for their original motivation in this context. See also the recent construction by A. Carrasco and M. Mars in [CM09] of a counterexample to a conjectured spacetime Penrose inequality in [BK09]. It has been shown by M. Khuri in [K09] that 3 + 1-dimensional Minkowski space does not contain generalized apparent horizons. G. Galloway has an argument to construct generalized trapped curves in 2 + 1-dimensional Minkowski space (private communication). In Sect. 5 we observe that the limit of an increasing sequence of generalized apparent horizons {in−1 }, where the in−1 satisfy uniform C 2,α -estimates and are all outer area minimizing, is embedded. Embeddedness is a critical issue here because (generalized) apparent horizons, unlike minimal surfaces, do not satisfy a two-sided maximum principle, cf. [AM05, Remark 8.3]. This observation leads directly to the proof of Conjecture 1: Theorem 1.2. Conditions as in the preceding theorem. Then there exists a unique outermost generalized apparent horizon n−1 in . n−1 is a closed embedded C 2,α -hypersurface. Moreover, n−1 is outer area minimizing. A refined statement and the proof of this theorem are given in Sect. 5. The following notation will be used throughout this paper:
748
M. Eichmair
Definition 1.2. Given an initial data set (M n , g, p), an open set U ⊂ M n , and a C 2 function u : U → R we write Di u H(u) := Di , 1 + | D u|2 Di u D j u pi j , tr( p)(u) := g i j − 1 + | D u|2 and for every ε > 0, | tr( p)|ε (u) :=
tr( p)(u)2 + ε2 .
Note that H(u) is the mean curvature of graph(u, U ) ⊂ M n × R computed with respect to the product metric g + dx2n+1 and its downward pointing unit normal √(D u,−1) 2 . If 1+| D u|
we think of the tensor p as being extended to the product manifold M n × R by zero in the vertical direction, then tr( p)(u) represents its trace over the tangent space of graph(u, U ). We introduce the auxiliary expressions | tr( p)(u)|ε , which regularize the absolute value of tr( p)(u), to facilitate exposition and analysis in the paper. 2. Solving (H − |tr( p)|ε − t)(u) = 0 for t > 0 In this section we use the Perron method as in [E07] to prove the following existence result: Lemma 2.1. Let (M n , g, p) be a complete initial data set and let ⊂ M n be a bounded ˙ 2 , where H∂1 > open subset with embedded C 2 -boundary such that ∂ = ∂1 ∪∂ | tr ∂1 ( p)|ε (with respect to the outward unit normal) and H∂2 < | tr ∂2 ( p)|ε (with respect to the normal into ) for some ε ∈ (0, 1). There exists θ > 0 such that for every 2,α sufficiently small t > 0 there is a non-positive function u εt ∈ Cloc () ∩ C 0,1 ( ∪ ∂1 ) ε ε satisfying (H −| tr( p)|ε − t) u t = 0 in with u t = 0 along ∂1 , u εt ≤ − θt on {x ∈ ¯ : dist(x, ∂1 ) ≤ θ }, and : dist(x, ∂2 ) ≤ θ }, u εt ≥ ln 1 − dist(∂1 ,·) on {x ∈ θ
0 ≥ u εt ≥ − Ct on all of , where C := 1 + n| p|C () ¯ . We discuss the required modifications of the method used in [E07, Sect. 2, Sect. 3] for the proof of Lemma 2.1. We refer the reader to that paper for details and concise references for the techniques involved. Recall first that if u : U ⊂ → R is a C 3 -function and if G n := graph(u, U ) = {(x, u(x)) : x ∈ U } ⊂ M n × R (with the product metric g + dx2n+1 ) denotes its graph, then 1 1 = 0, G + | h |2 + Ric M×R ( ν , ν ) + ν H(u) v v where v := 1 + | D u|2 is the ‘area stretch factor’ of the graph, cf. [SY81, Eq. (2.28)]. This is the second variation formula of the area element of G n applied to the variation coming from vertical translation in M n ×R. Here, G denotes the non-positive hypersurface Laplacian with respect to the induced metric g¯ on G n , h denotes the second funda1 mental form of G n , and the downward pointing unit normal ν = (1+| D u|2 )− 2 (D u, −1)
Generalized Apparent Horizons
749
differentiates the mean curvature H(u) of G n as a function on U × R (invariant in the vertical direction). If t > 0 and if u solves (H −| tr( p)|ε − t) u = 0 on U , then this identity implies a differential inequality of the form
1 1− 3(n − 1)
| h |2 1 κ2 + G ≤ v v v
(2)
for v −1 , where κ is a constant depending only on | Ric M |, | p|C 1 , and the dimension n, but not on ε > 0 or t > 0. The scale of κ is one over length. The method of Korevaar-Simon (see [E07, Lemma 2.1] for a precise statement) provides interior gradient estimates from oscillation bounds for such functions u. We now construct the particular solutions u εt described in Lemma 2.1. First note that the constant function 0 is a super solution of (H −| tr( p)|ε − t) u = 0. Similarly, the constant function at height − Ct , where C = 1+n| p|C () ¯ (the dimension times the largest ¯ plus one) is a sub solution. Let dist(∂2 , ·): M n → R measure eigenvalue of p on geodesic distance in (M n , g) to ∂2 and choose δ > 0 so small that dist(∂2 , ·) is a C 2 -function on {x ∈ : dist(∂2 , x) ≤ 2δ} (cf. [GT98, Lemma 14.16]), such that the latter set is disjoint from ∂1 , and such that the mean curvature of the distance surfaces (∂2 )γ := {x ∈ : dist(∂2 , x) = γ } satisfies 2δ + H(∂2 )γ < | tr (∂2 )γ ( p)|ε for all ¯ defined by γ ∈ [0, 2δ). It follows that the function u¯ t ∈ C() u¯ εt (x)
:=
dist(∂
2 ,x)−δ
t
0
if d∂2 (x) ≤ δ if d∂2 (x) > δ
is a Perron super solution for the operator L εt = H −| tr( p)|ε − t on for t > 0 sufficiently small. Similarly, it follows that for all t > 0 sufficiently small the function defined by
u t (x) :=
ln 1 − − Ct
dist(∂1 ,x) δ
if 0 ≤ d∂1 (x) ≤ δ 1 − exp(− Ct ) if d∂1 (x) > δ 1 − exp(− Ct )
is a Perron sub solution (possibly for some smaller δ > 0 depending on the geometry of ∂1 ). We emphasize that this sub solution is independent of ε > 0 and also of t > 0, when sufficiently small, in a fixed neighborhood of ∂1 . Also observe that ¯ : 0 ≤ dist(∂2 , x) ≤ δ }, and that u εt ≥ ln 1 − dist(∂1 ,·) u¯ εt ≤ − 2tδ on {x ∈ 2 δ ¯ : 0 ≤ dist(∂1 , x) ≤ δ }. The point in the construction of both sub and on {x ∈ 2 super solution is that the non-constant portions of their graphs converge to the cylinders (∂i )γ × R uniformly in C 2 on compact sets, as the parameters δ, t tend to zero. This classical construction of boundary barriers for the prescribed mean curvature equations is due to J. Serrin [Se69], see [GT98, Sect. 14], cf. [AM07, Prop. 3.5], and also [E07, Lemmas 2.2 and 3.3] for concise references and a ‘geometric’ discussion of these barriers in this context. The Perron method employed in [E07] now carries over verbatim to the present con2,α text and shows that there exists a Cloc -function u εt : → R with u t ≤ u εt ≤ u¯ εt and such that (H −| tr( p)|ε − t) u εt = 0 on . It follows easily that u t extends to a C 0,1 function near ∂1 . Taking θ = δ/2 this concludes the proof of Lemma 2.1.
750
M. Eichmair
The pointwise inequality (2) implies a stability-type inequality for G n in a standard way: given a test function φ ∈ Cc1 (U × R), multiply (2) by vφ 2 and integrate by parts on G n to obtain
1 ¯ 2 + κ2 1− | h |2 φ 2 ≤ |Dφ| φ2 (3) 3(n − 1) G G G adjusting κ depending on the dimension. A related computation is implicit in the proof 1 of [SY81, Prop. 1], see also [AM05, Lemma 5.6]. Instead of 3(n−1) we could take any 1 other positive constant less than (n−1) to make the arguments below work; this cut-off however is critical for the use of [SS81], as we discuss in the appendix. As in [SY81] we will eventually pass the graphs of solutions of the equation (H − | tr( p)|ε − t) u εt = 0 to a geometric subsequential limit as t 0. In order to avoid a dimensional restriction, we take this limit in a certain class of almost minimizing boundaries (in the sense of Almgren) and use compactness and regularity results from geometric measure theory to analyze the limit. More precisely, we use the classes FC of C-almost minimizing boundaries discussed in [E07, App. A]. As we will see below, we are able to bound |tu εt | and hence the mean curvature of all graphs in the construction uniformly by a constant that only depends on | p|C () ¯ so that their graphs are C-almost minimizing, cf. [E07, Ex. A.1]. Exactly as in the analysis of [SY81, Prop. 4], the limiting surface will contain a graphical component that is asymptotic to a vertical cylinder, and the cross section εn−1 ⊂ M n of this cylinder satisfies the geometric equation Hε = | tr ε ( p)| and inherits the stability-type inequality (3) as well as the C-almost minimizing property in from the original graphs. 3. Existence of Surfaces εn−1 with Hε = |trε ( p)|ε Following [SY81], we now pass the graphs of the solutions u εt from Lemma 2.1 to a geometric limit as t 0. Just as with the regular Jang’s equation, the vertically unbounded limiting surface will be asymptotic to a cylinder whose cross section is a closed surface εn−1 such that Hε = | tr ε ( p)|ε , see also [AM07, Theorem 1.1], [E07, Theorem 1.1]. Theorem 3.1. Assume that 2 ≤ n ≤ 7, let (M n , g, p) be a complete initial data set, and ˙ 2 , let ⊂ M n be a bounded open subset with embedded C 2 -boundary ∂ = ∂1 ∪∂ where H∂1 > | tr ∂1 ( p)|ε (with respect to the outward unit normal) and H∂2 < | tr ∂2 ( p)|ε (with respect to the normal pointing into ). There exists an open set ⊂ ˙ εn−1 such that εn−1 is an embedded hypersurface disjoint from ∂ with ∂ = ∂1 ∪ that satisfies Hε = | tr ε ( p)|ε (unit normal pointing into ). There are C 2,α -estimates for the surface εn−1 arising in the construction which only depend on Hn−1 (∂1 ), n n−1 is stable in the sense of | p|C 1 () ¯ , and the local geometry of (M , g). In fact, ε ¯ , and it is C-almost miniAppendix A, it minimizes area with respect to variations in mizing in all of for a constant C = C(| p|C () ¯ ). 2,α () ∩ C 0,1 ( ∪ ∂1 ), C > 0 be as in Lemma 2.1, and let Proof. Let θ > 0, u εt ∈ Cloc G εt := {(x, u εt (x)) : x ∈ ∪ ∂1 } be the corresponding graphs in (M n × R, g + dx2n+1 ). Since 0 ≥ u εt ≥ − Ct the mean curvatures of the graphs G εt are bounded uniformly by 2C and hence are 2C-almost minimizing in the language of [E07, App. A]. Using the compactness and regularity theory for these almost minimizing boundaries (see also [E07, Remark 4.1]) we can pass G εt to a smooth subsequential limit G ε along a sequence
Generalized Apparent Horizons
751
t 0. Moreover, the connected components of G ε are either cylindrical or themselves entirely graphical by the Harnack principle, cf. [E07, Lemma 2.3] for reference. In fact ˙ εn−1 so that below we see that there exists an open subset ⊂ with ∂ = ∂1 ∪ ε × {0} the hypersurface G is given as the graph of a function u ε : ∪ ∂1 → R satisfying H(u ε ) = | tr( p)|ε (u ε ). This is because the sub solution u εt is independent of t in a fixed neighborhood of ∂1 preventing the limit from diverging downwards there. We have that u ε = 0 on ∂1 and that u ε tends to negative infinity with graph(u ε , ) asymptoting εn−1 × R on approach to εn−1 through , cf. [SY81, Prop. 4]. It follows also that εn−1 satisfies Hε = | tr ε ( p)|, where the mean curvature is computed with respect to the unit normal pointing into . Using that the unit normal vector field − 1 1 + | D u ε |2 2 (D u ε , −1) of G ε has non-negative divergence, a standard calibration argument (cf. [E07, Ex. A.1]) shows that εn−1 minimizes area with respect to variations ¯ . The remarks in Sect. 2 imply that εn−1 satisfies a stability-type inequality (3) in with a constant κ that only depends on Ric M and | p|C 1 () ¯ making the results discussed in Appendix A available. It follows that the locally defining functions of εn−1 satisfy uniform C 1,α -estimates; since they also satisfy the (geometric) divergence form equation Hε = | tr ε ( p)|, C 2,α -estimates follow from standard elliptic theory [GT98]. 4. Existence of Generalized Apparent Horizons In this section we prove an existence theorem for generalized apparent horizons by passing the hypersurfaces εn−1 constructed in the previous section to a subsequential limit as ε 0. The lower order geometric properties (C-almost minimizing, outer area minimizing) of the surfaces εn−1 descend to this limit, as does the stability-type inequality (3) with uniform constant κ = κ(Ric M , | p|C 1 () ¯ , n), so that we stay in the class of surfaces to which the regularity and compactness theory of [SS81] discussed in Appendix A is applicable. In Proposition 4.1 we use the Perron method to prove that given two generalized trapped surfaces, there always exists a stable outer minimizing generalized apparent horizon enclosing both of them. The purpose of this proposition in the proof of Conjecture 1 corresponds to that of Lemma 8 in [KH97] and more specifically to that of Lemma 7.7 in [AM07]. Theorem 4.1. Assume that 2 ≤ n ≤ 7 and let (M n , g, p) be a complete initial data set. ˙ 2 Let ⊂ M n be a bounded open subset with embedded C 2 -boundary ∂ = ∂1 ∪∂ such that H∂1 > | tr ∂1 ( p)| (mean curvature with respect to the normal pointing out of ) and H∂2 ≤ | tr ∂2 ( p)| (with respect to the normal pointing into ). Then there exists ˙ n−1 , where n−1 is a closed embedded an open set ⊂ such that ∂ = ∂1 ∪ hypersurface satisfying H = | tr ( p)| (normal pointing into ) with uniform C 2,α n estimates depending only on Hn−1 (∂1 ), | p|C 1 () ¯ , and the local geometry of (M , g). n−1 is disjoint from ∂1 , and if a connected component of n−1 The hypersurface intersects with a component of ∂2 , then these components must coincide. Moreover, ¯ and it is C-almost minimizing in n−1 minimizes area with respect to variations in all of for a constant C = C(| p|C () ). ¯ Proof. By the result of Sect. 3 we can find for every sufficiently small ε > 0 an open set ˙ εn−1 with Hε = | tr ε ( p)| (normal ε ⊂ with embedded boundary ∂ε = ∂1 ∪ n−1 pointing into ε ) so that (i) ε ∈ FC () with a constant C independent of ε > 0, ¯ ε , and (iii) εn−1 is C 2,α with (ii) εn−1 minimizes area with respect to variations in estimates independent of ε > 0.
752
M. Eichmair
By property (iii) (alternatively using Theorem A.2) we can pass εn−1 to a C 2,β -subsequential limit n−1 , where we fix some β ∈ (0, α). The C-almost minimizing property (i) of εn−1 gives that the only place where n−1 could fail to be embedded is along ∂2 . However, we can use the convergence in C 2,β and property (i) to rule out sheeting near the boundary exactly as in the proof of Corollary A.1. It follows that n−1 is properly embedded, that n−1 = ∂ for an open set ⊂ , and that H = | tr ( p)|. The strong maximum principle shows in a standard way that if a component of n−1 intersects with a component of ∂2 , then these components must coincide. (See [AM07, Prop 3.1, AG05, Prop. 2.4] for marginally outer trapped surfaces; in the notation of the latter reference, 0 ≤ u 2 − u 1 (the difference of the two locally defining functions) satisfies a linear elliptic equation to which the strong maximum principle can be applied.) That ¯ and is C-almost minimizing in all n−1 minimizes area with respect to variations in of follow easily now. Lemma 4.1. Let (M n , g, p) be a complete initial data set, 2 ≤ n ≤ 7, and let , ⊂ ˙ 2 with H∂1 > | tr ∂1 ( p)| ⊂ M n be open bounded subsets so that ∂ = ∂1 ∪∂ ˙ and ∂ = ∂1 ∪ ˙ , where H = | tr ( p)| (norand such that ∂ = ∂1 ∪ mal pointing into ) and H = | tr ( p)| (normal pointing into ) for embedded C 2 -hypersurfaces ∂1 , , . Then for every sufficiently small ε > 0 there exists an ˙ open set ε ⊂ ∩ such that ∂ε = ∂1 ∪ε , where ε is embedded and satisfies Hε = | tr ε ( p)|ε (normal pointing into ε ). The surface ε arising in our construction satisfies C 2,α -estimates depending only on Hn−1 (∂1 ), | p|C 1 () ¯ , and n the local geometry of (M , g), but not on ε > 0. Moreover, ε minimizes area with ¯ respect to variations in ε , and it is C-almost minimizing in all of ∩ for some C = C(| p|C () ¯ ). Proof. As in Sect. 2 we can use the Perron method to construct for every sufficiently small 2,α 0,1 t > 0 a non-positive function u ε t ∈ Cloc ( ) ∩ C ( ∪ ∂1 ) such that (H −| tr( p)|ε − C ε ε t)u t = 0 with u t = 0 on ∂1 . Moreover, we can arrange that u ε t ≥ − t on all of θ and u ε ¯ ) and where t ≤ − t on {x ∈ : dist(x, ) < θ }, where C = C(| p|C () θ > 0 depends on ε > 0 but not on t. We construct u ε t with identical properties, but with respect to . ε ε ε We now consider the function u¯ ε t = min(u t , u t ) ∈ C( ∩ ). Then u¯ t is a Perron super solution with respect to the operator H −| tr( p)|ε −t on the open set ∩ . We θ also have that u¯ ε t ≤ − t on the set {x ∈ ∩ : dist(x, ) < θ or dist(x, ) < θ }, ε so that in particular u¯ t tends to negative infinity on approach to ∂( ∩ )\∂1 uniformly as t 0. Let u εt be the Perron solution constructed from this super solution and the Perron sub solution described in Sect. 2 (which is constant away from ∂1 and independent of t, ε > 0 near ∂1 ). The graphs G εt := graph(u εt , ∩ ) ⊂ × R are C-almost minimizing in the cylinder ( ∩ ) × R. As in the proof of Theorem 3.1 ˙ we conclude that there exists an open set ε ⊂ ∩ with ∂ε = ∂1 ∪ε such ¯ ε , C-almost minimizing that ε is embedded, outer area minimizing with respect to in all of ∩ , and so that Hε = | tr ε ( p)|ε . The C 2,α -estimates follow from the comments succeeding the statement of Lemma 2.1 and the results in Appendix A. Proposition 4.1. Assumptions as in Lemma 4.1. Then there exists an open set ⊂ ˙ such that is disjoint from the ∩ with embedded boundary ∂ = ∂1 ∪ intersecting (but not coinciding) components of , and so that H = | tr ( p)| (with respect to the normal pointing into ). The surface arising in our construction satisfies C 2,α -estimates depending only on Hn−1 (∂1 ), | p|C 1 () ¯ , and the local
Generalized Apparent Horizons
753
geometry of (M n , g). It is stable in the sense of Appendix A, it minimizes area with ¯ , and it is C-almost minimizing in ∩ . respect to variations in Proof. Let {ε } be the surfaces constructed in Lemma 4.1. Using the uniform volumeand C 2,α -estimates (alternatively using Theorem A.2), we can pass these surfaces to an immersed subsequential limit n−1 . By the C-almost minimizing property of εn−1 in ∩ which descends to n−1 , it follows that n−1 could only fail to be embedded on the boundary of ∩ . As above, the argument in Corollary A.1 rules out sheeting near this boundary. Hence n−1 is an embedded generalized apparent horizon. Finally, the strong maximum principle applied as in the proof of Theorem 4.1 shows that if a component of n−1 touches a component of or then these components must coincide. We emphasize that in this proposition is stable and outer minimizing. Hence the conclusion is not trivial when ∩ = ∅ or even when = . This will be relevant in the proof of Theorem 5.1. Remark 4.1. The particular ε-regularization we chose helped us with the construction of appropriate super solutions for the problems (H −| tr( p)|ε − t)u εt = 0 from the conditions H = | tr ( p)| and H = | tr ( p)| in Lemma 4.1. For marginally outer trapped surfaces, where the natural conditions for the inner boundaries are H + tr ( p) ≤ 0 and H + tr ( p) ≤ 0, one can proceed similarly by first finding auxiliary surfaces εn−1 ⊂ ∩ for which Hε + tr ε pε = 0. Here, pε := p − εφg for some fixed smooth function φ : → R with φ ≡ 1 near ∪ and φ ≡ 0 near the outer boundary ∂1 . The apparent horizon enclosing ∪ is then found by letting ε 0 as above (note that the C-almost minimizing property is independent of ε > 0 here as well, so we have geometric estimates that allow us to pass to a limit). The trick of modifying the second fundamental form tensor in the direction of g to get strict barriers was used in [AM07, Theorem 5.1].
5. The Outermost Generalized Trapped Surface In this section we give the proof of Conjecture 1. The outermost generalized trapped surface is constructed as the boundary of the union of all generalized trapped domains. This is analogous to the construction of the apparent horizon as the boundary of the trapped region in [HE73,KH97,HI01,AM07]. The idea of replacing the total union by one increasing union is contained in [KH97] and has been used in [AM07] to prove existence of a smooth outermost apparent horizon in 3-dimensional data sets. A major technical challenge in [AM07] was to show that the boundary of this increasing union is smooth and embedded. We survey the steps in their argument for comparison: the authors first reduce to the case where the boundaries of the increasing sets in this union are stable marginally outer trapped surfaces and hence have bounded curvature [AM05, Theorem 1.2] (see also Sect. 4 in [HI01]), and by a further reduction to the case where these surfaces have a lower bound on their “outer injectivity radius,” cf. [AM07, Sect. 6]. The latter step depends on the a priori curvature bound coming from stability and a very delicate surgery procedure. The lower bound on the outer injectivity radius and the curvature bound give an area estimate [AM07, Theorem 6.1] for the boundary surfaces in this union. It follows that the limit of these surfaces exists as a smooth immersed marginally trapped surface which is the boundary of an open set, and hence cannot touch
754
M. Eichmair
itself on the inside. The lower bound on the outer injectivity radius of these surfaces implies that the limit is embedded [AM07, Sect. 7]. Our proof of Conjecture 1, which also applies to marginally outer trapped surfaces and works (at least) in dimensions 2 ≤ n ≤ 7, is based on the lower order properties (outward minimizing and outward almost minimizing) of these surfaces: Theorem 5.1. Let (M n , g, p) be a complete initial data set and assume that 2 ≤ n ≤ 7. ˙ 2 Let ⊂ M n be a bounded open subset with embedded C 2 -boundary ∂ = ∂1 ∪∂ such that H∂1 > | tr ∂1 ( p)| (mean curvature with respect to the normal pointing out of ) and H∂2 ≤ | tr ∂2 ( p)| (with respect to the normal pointing into ). Then there exists ˙ , where is an embedded C 2,α -hypersurface ⊂ open such that ∂ = ∂1 ∪ with H = | tr ( p)| (with respect to the normal pointing into ), and such that is the outermost generalized trapped surface in the following sense: if ⊂ has ˙ , where is an embedded C 2 -hypersurface with H ≤ | tr ( p)|, boundary ∂1 ∪ then ⊂ (so that encloses ). Moreover, this minimizes area with respect ¯ . to variations in ¯ , where the intersection is taken over all open Proof. Consider the closed set F = ∩ ˙ such that is an embedded, C 2 -hypersursubsets ⊂ for which ∂ = ∂1 ∪ face with H ≤ | tr ( p)|. As in [KH97], F is already the intersection of a countable ¯ of such sets. (This is because is second countable.) We can use Proposifamily i tion 4.1 to arrange for these sets to be decreasing 1 ⊃ 2 ⊃ . . . and such that i (= ∂i \ ∂1 ) is a stable (in the sense of inequality (3) and Appendix A) generalized apparent horizon (cf. [KH97,HI01,AM07]) which minimizes area with respect to vari¯ . Then Corollary A.1 implies that := int F has the required properties. ations in i
Appendix A. Remark on [SS81] In this appendix we explain how the regularity theory of [SS81] for stable critical points of elliptic variational problems can be applied to the context of generalized apparent horizons, even though these latter surfaces are not associated with a particular functional. The proofs in [SS81] generalize to our situation with only a few very minor modifications, which we discuss here. The results stated in this appendix provide a general C 1,α -regularity and compactness theory for limits of smooth embedded hypersurfaces of bounded area and mean curvature that also satisfy a stability-type inequality such as (3), with the usual estimate of the singular set of such a limit. In particular, these results are available in all dimensions and in other situations where a priori curvature estimates may not be readily available. We follow the notation and conventions of [SS81] closely in this appendix to facilitate reference. Given a Riemannian manifold (M n , g), a point p ∈ M n , and 0 < ρ0 < inj p (M n , g), one can use geodesic normal coordinates centered at p to identify the geodesic ball B n ( p, ρ0 ) ⊂ M n with the Euclidean ball {|X | < ρ0 } ⊂ Tan p (M n ). Given a hypersurface G n−1 ⊂ M n ∩ B n ( p, ρ0 ) one can use this identification to compute geometric quantities of G n−1 either with respect to gi j or with respect to the Euclidean metric δi j
Generalized Apparent Horizons
on {|X | < ρ0 }. One has that
755
gi j 0) = δi j , ∂k gi j 0) = 0, and sup ∂kl2 gi j ≤ µ21
{|X |<ρ0 }
for some constant µ1 ≥ 0 depending only on the geometry of (B n ( p, ρ0 ), g) (specifically on the C 0 size of the curvature tensor). Denoting all quantities computed with respect to the Euclidean metric δi j with a hat, one has that 2 ˆ 2 ≤ c1 µ1 |X ||h| ˆ 2 + µ21 , (4) | h |g − |h| ˆ ≤ c1 µ1 |X ||h| ˆ + µ1 , H −H provided that µ1 |X | < 1. Here, X is the position vector in {|X | < ρ0 }, c1 is a dimensional constant, and the norms of hatted quantities are taken with respect to the Euclidean metric δi j , cf. [SS81, Sect. 1, Sect. 6]. Suppose now that the embedded C 2 -hypersurface G n−1 ⊂ M n satisfies a stabilitytype inequality of the form
η 2 2 2 2 ¯ 1− | h |g φ ≤ |Dg φ| + κ φ 2 ∀φ ∈ Cc1 (M n ), (5) 2 G G G where η ∈ (0, 1) and that in addition its mean curvature is bounded | H | ≤ κ. Using the estimates in (4) we see that on B n ( p, ρ0 ) this inequality carries over to the Euclidean geometry in the form
ˆ 2φ2 ≤ |h| (1 − η) G
ˆ |Dφ|2 + κˆ 2 φ 2 for all φ ∈ Cc1 ({|X | < ρ0 }), (6) G
G
provided that µ1 ρ0 is sufficiently small (depending only on the dimension), and where the new constant κˆ depends on κ, µ1 , η and the dimension n. All integrals here are comˆ denotes the tangential puted with respect to the Euclidean metric induced on G, and Dφ ˆ of G on the geodesic ball gradient along G. From (4) we see that the mean curvature H {|X | < ρ0 } can be estimated by ˆ ≤ c1 µ1 |X ||h| ˆ + µ1 + κ. (7) |H| Absorbing κ into the constant µ1 , we see that the structural assumptions (1.16) and (1.17) of [SS81] are satisfied with a marginally worse constant 1 − η multiplying the left-hand side of the stability-type inequality (6). We note that the fundamental integral curvature estimate in [SS81, Sect. 2] still follows from these inequalities provided that 1 0 < η < n−1 and that the proof of the basic regularity estimate [SS81, Theorem 1] carries over verbatim to our setting if we assume a priori that G n−1 ⊂ B n ( p, ρ0 ) is an embedded, relatively closed C 2 -hypersurface. In the statement of the result below, C(X, ρ) denotes the cylinder {y ∈ Rn−1 : |y − x| < ρ} × R for X = (x, xn+1 ) ∈ {|X | < ρ0 }:
756
M. Eichmair
Theorem A.2 ([SS81]). Let G n−1 ⊂ {|X | < ρ0 } be a relatively closed, embedded C 2 -hypersurface such that for some constants µ, µ1 > 0 one has Hn−1 (G n−1 ) ≤ µωn−1 ρ0n−1 , ˆ ≤ c1 µ1 |X ||h| ˆ + µ1 , |H| and
ˆ 2φ2 ≤ |h|
(1 − η) G
ˆ 2 + µ21 |Dφ|
G
G
φ 2 for all φ ∈ Cc1 ({|X | < ρ0 })
1 for some η ∈ (0, n−1 ). Then there exists a number δ0 ∈ (0, 1) depending only on n, µ, η, and µ1 ρ0 such that if X ∈ G n−1 ∩ {|X | < ρ0 /4}, ρ ∈ (0, ρ0 /4), and G is the connected component of G n−1 ∩ C(X, ρ) containing X , and if for some δ ∈ (0, δ0 ),
sup |yn+1 − xn+1 | ≤ δρ,
Y ∈G
µ1 ρ ≤ δ,
where X = (x, xn+1 ) and Y = (y, yn+1 ), then G ∩ C(X, ρ/2) consists of a disjoint union of graphs of functions u 1 < u 2 < · · · < u k defined on B n−1 (x, ρ/2) ⊂ Rn−1 such that 2 ρ
sup y∈B n−1 (x,ρ/2)
+
ρ α 2
|u i (y)| +
sup y∈B n−1 (x,ρ/2)
sup y,y ∈B n−1 (x,ρ/2),y=y
ˆ i (y)| |Du
ˆ i (y) − Du ˆ i (y )| 1 |Du =: |u i |1,α,B n−1 (x,ρ/2) ≤ δ 3 α |y − y |
(8)
for all i = 1, 2, . . . k, where α ∈ (0, 1), k, and c2 depend only on n, µ, η, and µ1 ρ0 . 1
Remark A.1. The appearance of δ 3 in estimate (8) rather than plain δ as in [SS81] is due to the fact that we don’t want to appeal to Schauder theory at this point (to avoid mention of a defining equation for the functions u i ). Careful screening of the proof in [SS81] (note in particular (1.21), (3.32), (4.33)-(4.37) in that paper) shows that this power is sufficient. Theorem A.2 says that a regular closed embedded submanifold with controlled mean curvature and area that is stable in the sense of (6) decomposes into a union of graphs with C 1,α -estimates whenever its support is sufficiently close to a hyperplane. In the case where the submanifolds are stationary with respect to an elliptic functional (i.e. satisfy an appropriate equation), Theorem A.2 is combined in [SS81] with the compactness theorem for rectifiable varifolds, a version of Federer’s dimension reduction argument, and the fact that there exist no stable minimal hypercones in Rn other than planes when 3 ≤ n ≤ 7 to obtain curvature estimates in these dimensions. (Note that there are stable minimal tangent cones in dimension n = 2 which are singular at the origin. Such tangent cones for the limiting surfaces are ruled out in [SS81, p. 786 and 787]. We point out that in this paper, one-dimensional generalized apparent horizons arise as cross-sections of stable almost minimizing cylinders and that we can carry out the regularity argument on that level. The almost minimizing property and the fact that all our surfaces are boundaries would also rule out such singular cones.) If we do bring in the defining equation HG = | tr G ( p)|ε of the surfaces in this paper, uniform C 2,α -estimates in terms of µ, µ1 ρ0 , n and | p|C 1 follow from the same argument.
Generalized Apparent Horizons
757
The following regularity property was formulated in [SS81, p. 780]: Definition A.1 ([SS81]). Let δ0 , µ, µ1 be positive constants. We say that a countably (n − 1)-rectifiable varifold V ∈ Vn−1 (Rn ) has the property Pδ0 µµ1 with respect to an open subset U of {|X | < ρ0 } if µ−1 σ n−1 ωn−1 ≤ ||V ||(B n (X, σ )) ≤ µσ n−1 ωn−1 for all X ∈ spt ||V || and σ > 0 such that B n (X, σ ) ⊂ U,
(9)
and provided that whenever the hypotheses Y ∈ spt ||V ||, B n (Y, ρ) ⊂ U, µ1 ρ < δ and spt ||V || ∩ B n (Y, ρ) ⊂ {X ∈ Rn : dist(X, ) < δρ} hold for some hyperplane ⊂ Rn containing Y and some δ ∈ (0, δ0 ), then there exists an isometry O of Rn with O(Y ) = 0, O() = Rn−1 × {0}, and O(spt ||V || ∩ B n (Y, ρ)) ∩ C(0, ρ/2) =
l
graph(u i )
i=1
for some integer l, 1 ≤ l ≤ µ, where u i ∈ C 1,α (B n−1 (0, ρ/2)) are such that u 1 ≤ u 2 ≤ · · · ≤ u l and 1
|u i |1,α,B n−1 (0,ρ/2) ≤ δ 3 . Here we are using scale-invariant Schauder norms on the left. Remark A.2. The upper bound in condition (9) is implied by a total mass bound ||V ||(U ) and an estimate on the first variation δV of V of the form
δV (ϕ)| ≤ µ1 (|ϕ| + |X || D ϕ|) d||V || for all ϕ ∈ C 1 (U, Rn ) (10) c through the monotonicity formula, see [SS81, pp. 778, 779]. Here, D ϕ denotes the ambient covariant derivative of the vector field ϕ. Such an estimate is implied by (7). The lower bound in (9) follows from the same principle if we assume that V satisfies a positive lower bound on its density (for example if it is integer multiplicity). Note that we do not require that the defining graphs u 1 ≤ · · · ≤ u l be disjoint. It is also evident that the class Pδ0 µµ1 is preserved under varifold limits (by Arzela-Ascoli). Note also that any relatively closed, embedded C 2 -hypersurface G n−1 ⊂ {|X | < ρ0 } satisfying the hypotheses of Theorem A.2 belongs to the class Pδ0 µµ1 provided one chooses δ0 sufficiently small, cf. [SS81, Remark 11]. The following definitions of the singular and regular sets differ marginally from the definition in [SS81, p. 777]. Definition A.2. Let U ⊂ {|X | < ρ0 } be an open set and let V ∈ Vn−1 (Rn ) have the property Pδ0 µµ1 with respect to this set. The regular set reg(V ) of the varifold V in U is defined as the set of all points Y ∈ spt ||V || ∩ U such that for some small radius ρ > 0 one can write spt ||V || ∩ B(Y, ρ) as the union of weakly ordered C 1,α -graphs u 1 ≤ u 2 ≤ · · · ≤ u l defined on a common hyperplane. The singular set sing(V ) is defined as the complement of reg(V ) in spt(V ) ∩ U . Note that it follows from the property Pδ0 µµ1 that Y ∈ reg(T ) if and only if there exists a varifold tangent T ∈ Tan Var(V, Y ) so that ||T || = m|| for some hyperplane of Rn and some positive integer 0 < m ≤ µ. It also follows trivially that varifold tangents at regular points are unique. From the definition one sees that reg(V ) is
758
M. Eichmair
relatively open. There is some minor subtlety in allowing the graphs of the locally defining functions u 1 ≤ u 2 ≤ · · · ≤ u k to touch. This is related to the fact that, unlike minimal surfaces, (generalized) apparent horizons don’t satisfy a two-sided maximum principle. If the first variation of V also satisfies (10), then its tangent varifolds are stationary cones [SS81, p. 780]. The Hopf maximum principle and the constancy theorem then show that the tangent cones of such varifolds are smooth hypersurfaces with constant (integer) multiplicity near their regular points. It is easy to see that this notion of singular set is consistent with the basic assumptions A.1, A.2, and A.3 of the abstract dimension reduction procedure given in Appendix A of [Si83]. The compactness theory of [SS81] for stable minimal hypersurfaces with area bounds takes the following form in the present context, with virtually the same proof: Theorem A.2 ([SS81]). Let G in−1 ⊂ {|X | < ρ0 } be a sequence of embedded relatively 1 closed C 2 -hypersurfaces such that there exist constants µ1 , µ > 0 and η ∈ (0, n−1 ) so that Hn−1 (G in−1 ) ≤ µρ0n−1 ωn−1 ,
2 2 2 2 ˆ ˆ |h i | φ ≤ |Dφ| + µ1 φ 2 ∀φ ∈ Cc1 ({|X | < ρ0 }), (1 − η) Gi Gi Gi ˆ i | ≤ c1 µ1 |X ||hˆ i | + µ1 . |H Let G n−1 ∈ Vn−1 (Rn ) be a subsequential varifold limit of {G in−1 }. Then G n−1 is integer rectifiable and the Hausdorff dimension of its singular set sing(G n−1 ) is ≤ n − 8. In particular, if 2 ≤ n ≤ 7, the limit G n−1 is an immersed C 1,α -hypersurface. In the following corollary we note that increasing limits of stable outer minimizing hypersurfaces with uniformly bounded mean curvature remain regular and embedded. Corollary A.1. Assume that 2 ≤ n ≤ 7, let (M n , g) be a complete Riemannian mani˙ 2 . fold, and let ⊂ M n be a bounded open set with smooth boundary ∂ = ∂1 ∪∂ Consider a decreasing sequence ⊃ 1 ⊃ 2 ⊃ . . . of open subsets of with ˙ in−1 such that the in−1 are embedded C 2 -hypersurfaces which are outer ∂i = ∂1 ∪ ¯ i . Assume that there are minimizing, i.e., minimize area with respect to variations in constants µ1 , µ so that the assumptions of Theorem A.2 hold uniformly for in−1 and that the surfaces in−1 stay away in Hausdorff distance from the outer boundary ∂1 . ˙ n−1 such that n−1 is an Then there exists an open set ⊂ with ∂ = ∂1 ∪ n−1 1,α n−1 1,β embedded C -hypersurface, i → in C for any 0 < β < α, and so that n−1 ¯ minimizes area with respect to variations in . The corollary holds true if ‘outer minimizing’ is replaced by ‘outer C-almost minimizing’ with essentially the same proof. The uniform area bound (expressed in the constant µ) required for the use of Theorem A.2 is given by Hn−1 (∂1 ), respectively Hn−1 (∂1 ) + CLn (). The Hölder exponent α ∈ (0, 1) is as in Theorem A.2 and depends only on n, η, µ, and µ1 ρ0 . Proof. Note first that by Allard’s integral compactness theorem, the sequence in−1 converges to a countably (n − 1)-rectifiable integer multiplicity varifold n−1 with bounded mass and first variation. From Theorem A.2 we know that sing( n−1 ) = ∅.
Generalized Apparent Horizons
759
Let T = m|| be the (unique) varifold tangent at X ∈ spt n−1 , where is a hyperplane and m is a positive integer. The bounds on the mean curvature imply that T ∩ B n (0, 1) is approached in Hausdorff distance by appropriate rescalings of the embedded hypersurfaces in−1 (cf. [SS81, p. 780]), which by Theorem A.2 decompose after an appropriate rotation into graphs u i1 < u i2 < · · · < u im over B n−1 (0, 1/2) ⊂ Rn−1 with |u|1,α,B n−1 (0,1/2) → 0. Since the in−1 are boundaries, the set of points {(x, xn ) : x ∈ B n−1 (0, 1/2) and u ij (x) < xn < u ij+1 (x)} either belongs to i or its complement for ¯ i we immediately obtain that every j = 1, . . . , m − 1. Since in−1 minimizes area in m ≤ 2. Finally note that since the sets i are decreasing, m = 1 (cf. [AM07, p. 971]). Since X ∈ spt n−1 was arbitrary it follows that n−1 is embedded. That n−1 = ∂ ¯ now follow easily. and that n−1 minimizes area with respect to variations in Remark A.3. Theorem A.2 describes the regularity of varifold limits of smooth embedded hypersurfaces which have bounded area and mean curvature, and which satisfy a uniform stability-type inequality. An immediate consequence are curvature estimates for stable embedded generalized apparent horizons n−1 in dimensions 2 ≤ n ≤ 7 in terms of the injectivity radius of (M n , g), | Rm |∞ (which enters through µ1 ), | p|C 1 () ¯ , and a bound on the hypersurface measure (coming from the outer minimizing property of the surfaces in this paper). Such estimates have been obtained for immersed stable marginally outer trapped surfaces in dimensions 2 ≤ n ≤ 6 in [AM05], by generalizing the iteration method in [SY81], in particular by deriving an appropriate analogue of the Simons identity for χ (which is the second fundamental form plus the restriction of p to the surface). The authors derive an inequality like (5) with |h| replaced by |χ | using the first eigenfunction of the stability operator associated to a marginally trapped surface in an initial data set in [AM05, Lemma 5.6]. Their lemma shows that the theory in this appendix also applies to stable marginally trapped surfaces. The ‘lower order approach’ of [SS81] that we are taking here is quite flexible and applies nicely to the class of generalized apparent horizons for which one expects C 2,α -regularity at best. These results are available in all dimensions if we accept singular sets of Hausdorff co-dimension 7. Note that the important estimate on the outer injectivity radius in [AM07, §6] can be recovered for stable outer C-almost minimizing surfaces n−1 in all dimensions 2 ≤ n ≤ 7 by combining Theorem A.2 with the argument of Corollary A.1. The point is that the increasing property of the surfaces in−1 in Corollary A.1 only enters when we concluded that m = 1. That there can be at most two sheets merging from the inside follows from the one-sided (almost) minimizing property and the fact that the in−1 ’s are boundaries. Acknowledgements. This work forms part of my thesis, and I am very much indebted to my adviser Richard Schoen for his constant support and encouragement. Thanks very much to Hubert Bray for great discussions, and for drawing my attention to this problem. I would like to sincerely thank Simon Brendle, Leon Simon, and Brian White for the great example they set and for everything they have taught me over the years. I am grateful to Greg Galloway, Tom Ilmanen, Marcus Khuri, and Jan Metzger for their interest in this work.
References [AM05] [AM07]
Andersson, L., Metzger, J.: Curvature estimates for stable marginally trapped surfaces. Preprint (2005), http://arxiv.org/abs/gr-qc/0512106v2, 2006 Andersson, L., Metzger, J.: The area of horizons and the trapped regions. Commun. Math. Phys. 290(3), 941–972 (2009)
760
[AG05] [CM09] [BK09] [BL07] [E07] [GT98] [HE73] [HI01] [KH97] [K09] [SS81] [SSY75] [SY81] [Se69] [Si83]
M. Eichmair
Ashtekar, A., Galloway, G.: Some uniqueness results for dynamical horizons. Adv. Theor. Math. Phys. 9, 1–30 (2005) Carrasco, A., Mars, M.: A counter-example to a recent version of the Penrose conjecture. http:// arxiv.org/abs/0911.0883v1[gr-qc], 2009 Bray, H., Khuri, M.: P.D.E.’s which imply the Penrose Conjecture. http://arxiv.org/abs/0910. 4785v1[math.DG], 2009 Bray, H., Lee, D.: On the riemannian penrose inequality in dimensions less than 8. Duke Math. J. 148(1), 81–106 (2009) Eichmair, M.: The Plateau problem for marginally trapped surfaces. http://arxiv.org/abs/0711. 4139v2[math.DG], 2008 preprint (2007) Gilbarg, D., Trudinger, N.: Elliptic Partial Differential Equations of Second Order. BerlinHeidelberg-New York: Springer, 1998 Hawking, S., Ellis, G.: On the Large Scale Structure of Space-Time. Cambridge: Cambridge University Press, 1973 Huisken, G., Ilmanen, T.: The inverse mean curvature flow and the riemannian penrose inequality. J. Differential Geometry 59(3), 353–437 (2001) Kriele, M., Hayward, S.: Outer trapped surfaces and their apparent horizons. J. Math. Phys. 38, 1593 (1997) Khuri, M.: A note on the nonexistence of generalized apparent horizons in minkowski space. Class. Quant. Grav. 26(7), 078001 (2009) Schoen, R., Simon, L.: Regularity of stable minimal hypersurfaces. Comm. Pure and Appl. Math. 34(6), 741–797 (1981) Schoen, R., Simon, L., Yau, S.-T.: Curvature estimates for minimal hypersurfaces. Acta Math. 134(3-4), 275–288 (1975) Schoen, R., Yau, S.-T.: Proof of the positive mass theorem. Commun. Math. Phys. 79(2), 231–260 (1981) Serrin, J.: The problem of dirichlet for quasilinear elliptic differential equations with many independent variables. Philos. Trans. Roy. Soc. London. Ser. A 264, 413–496 (1969) Simon, L.: Lectures on Geometric Measure Theory. Centre for Mathematical Analysis, Australian National University, Volume 3, 1983
Communicated by P.T. Chru´sciel
Commun. Math. Phys. 294, 761–825 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0950-x
Communications in
Mathematical Physics
Infraparticle Scattering States in Non-Relativistic QED: I. The Bloch-Nordsieck Paradigm Thomas Chen1 , Jürg Fröhlich2,3 , Alessandro Pizzo2 1 Department of Mathematics, University of Texas at Austin,
Austin, TX 78712, USA. E-mail:
[email protected]
2 Department of Mathematics, University of California Davis,
Davis, CA 95616, USA. E-mail:
[email protected]
3 IHÉS, Bures sur Yvette, France.
E-mail:
[email protected] Received: 30 August 2008 / Accepted: 25 August 2009 Published online: 20 December 2009 – © The Author(s) 2009. This article is published with open access at Springerlink.com
Abstract: We construct infraparticle scattering states for Compton scattering in the standard model of non-relativistic QED. In our construction, an infrared cutoff initially introduced to regularize the model is removed completely. We rigorously establish the properties of infraparticle scattering theory predicted in the classic work of Bloch and Nordsieck from the 1930’s, Faddeev and Kulish, and others. Our results represent a basic step towards solving the infrared problem in (non-relativistic) QED.
I. Introduction The construction of scattering states in Quantum Electrodynamics (QED) is an old open problem. The main difficulties in solving this problem are linked to the infamous infrared catastrophe in QED: It became clear very early in the development of QED that, at the level of perturbation theory (e.g., for Compton scattering), the transition amplitudes between formal scattering states with charges and a finite number of photons are ill-defined, because, typically, Feynman amplitudes containing vertex or electron selfenergy corrections exhibit logarithmic infrared divergences; [14,22]. A pragmatic approach proposed by Jauch and Rohrlich, [21,27], and by Yennie, Frautschi, and Suura, [31], is to circumvent this difficulty by considering inclusive cross sections: One sums over all possible final states that include photons whose total energy lies below an arbitrary threshold energy > 0. Then the infrared divergences due to soft virtual photons are formally canceled by those corresponding to the emission of soft photons of total energy below , order by order in perturbation theory in powers of the finestructure constant α. A drawback of this approach becomes apparent when one tries to formulate a scattering theory that is -independent: Because the transition probability P for an inclusive process is estimated to be O( const.α ), the threshold energy cannot be allowed to approach zero, unless “Bremsstrahlungs” processes (emission of photons) are properly incorporated in the calculation.
762
T. Chen, J. Fröhlich, A. Pizzo
An alternative approach to solving the infrared problem is to go beyond inclusive cross sections and to define α-dependent scattering states containing an infinite number of photons (so-called soft-photon clouds), which are expected to yield finite transition amplitudes, order by order in perturbation theory. The works of Chung [12], Kibble [23], and Faddeev and Kulish [13], between 1965 and 1970, represent promising, albeit incomplete progress in this direction. Their approaches are guided by an ansatz identified in the analysis of certain solvable models introduced in early work by Bloch and Nordsieck, [2], and extended by Pauli and Fierz, [14], in the late 1930’s. In a seminal paper [2] by Bloch and Nordsieck, it was shown (under certain approximations that render their model solvable) that, in the presence of asymptotic charged particles, the scattering representations of the asymptotic photon field are a coherent non-Fock representation, and that formal scattering states with a finite number of soft photons do not belong to the physical Hilbert space of a system of asymptotically freely moving electrons interacting with the quantized radiation field. These authors also showed that the coherent states describing the soft-photon cloud are parameterized by the asymptotic velocities of the electrons. The perturbative recipes for the construction of scattering states did not remove some of the major conceptual problems. New puzzles appeared, some of which are related to the problem that Lorentz boosts cannot be unitarily implemented on charged scattering states; see [19]. This host of problems was addressed in a fundamental analysis of the structural properties of QED, and of the infrared problem in the framework of general quantum field theory; see [30]. Subsequent developments in axiomatic quantum field theory have led to results that are of great importance for the topics treated in the present paper: i) Absence of dressed one-electron states with a sharp mass; see [4,28]. ii) Corrections to the asymptotic dynamics, as compared to the one in a theory with a positive mass gap; see [3]. iii) Superselection rules pertaining to the space-like asymptotics of the quantized electromagnetic field, and connections to Gauss’ law; see [4]. In the early 1970’s, significant advances on the infrared problem were made for Nelson’s model, which describes non-relativistic matter linearly coupled to a scalar field of massless bosons. In [15,16], the disappearance of a sharp mass shell for the charged particles was established for Nelson’s model, in the limit where an infrared cut-off is removed. (An infrared cutoff is introduced, initially, with the purpose to eliminate the interactions between charged particles and soft boson modes). Techniques developed in [15,16] have become standard tools in more recent work on non-relativistic QED, and attempts made in [15,16] have stimulated a deeper understanding of the asymptotic dynamics of charged particles and photons. The analysis of spectral and dynamical aspects of non-relativistic QED and of Nelson’s model constitutes an active branch of contemporary mathematical physics. In questions relating to the infrared problem, mathematical control of the removal of the infrared cutoff is a critical issue still unsolved in many situations. The construction of an infraparticle scattering theory for Nelson’s model, after removal of the infrared cutoff, has recently been achieved in [26] by introducing a suitable scattering scheme. This analysis involves spectral results substantially improving those in [16]. It is based on a new multiscale technique developed in [25]. While the interaction in Nelson’s model is linear in the creation- and annihilation operators of the boson field, it is non-linear and of vector type in non-relativistic QED. For this reason, the methods developed in [25,26] do not directly apply to the latter.
Infraparticle States in QED I
763
The main goal of the present work is to construct an infraparticle scattering theory for non-relativistic QED inspired by the methods of [25,26]. In a companion paper, [11], we derive those spectral properties of QED that are crucial for our analysis of scattering theory and determine the mass shell structure in the infrared limit. We will follow ideas developed in [25]. Bogoliubov transformations, proven in [10] to characterize the soft photon clouds in non-relativistic QED, represent an important element in our construction. The proof in [10] uses the uniform bounds on the renormalized electron mass previously established in [9]. We present a detailed definition of the model of non-relativistic QED in Sect. II. Aspects of infraparticle scattering theory, developed in this paper, are described in Sect. III. To understand why free radiation parametrized by the asymptotic velocities of the charged particles must be expected to be present in all the scattering states, we recall a useful point of view based on classical electrodynamics that was brought to our attention by Morchio and Strocchi. We consider a single, classical charged point-particle, e.g., an electron, moving along We suppose that, for t ≤ 0, it a world line (t, x (t)) in Minkowski space, with x (0) = 0. moves at a constant velocity vin , and, for t > t¯ > 0, at a constant velocity vout = vin , | vout |, | vin | < c, where c is the speed of light that we set equal to 1. Thus, x (t) = vin · t, for t ≤ 0,
(I.1)
x (t) = x ∗ + vout · t, for t ≥ t¯,
(I.2)
and
for some x ∗ . For times t ∈ [0, t¯], the particle performs an accelerated motion. We propose to analyze the behavior of the electromagnetic field in the vicinity of the particle and the properties of the free electromagnetic radiation at very early times (t → −∞,“in”) and very late times (t → +∞, “out”). For this purpose, we must solve Maxwell’s equations for the electromagnetic field tensor, F µν (t, y), given the 4-current density corresponding to the trajectory of the particle; (back reaction of the electromagnetic field on the motion of the charged particle is neglected): ∂µ F µν (t, y) = J ν (t, y)
(I.3)
y − x (t)), x˙ (t) δ (3) ( y − x (t)) ), J ν (t, y) := −q ( δ (3) (
(I.4)
with
2(2π )3 α 1/2
(α is the finestructure constant). where, in the units used in our paper, q = µν We solve Eq. (I.3) with prescribed spatial asymptotics (| y | → ∞): Let F[v L .W. ] (t, y) be y | → ∞), approaches the Liénarda solution of (I.3) that, to leading order in | y |−1 (| Wiechert field tensor for a point-particle with charge −q and a constant velocity vL .W. at all times. Let us denote the Liénard-Wiechert field tensor of a point-particle with charge −q moving along a trajectory (t, x (t)) in Minkowski space with x (0) =: x and x˙ (t) ≡ v, µν µν for all t, by Fx ,v (t, y). Apparently, we are looking for solutions, F[v L .W. ] (t, y), of (I.3) with the property that, for all times t, µν
µν
y |−2 ), |F[v L .W. ] (t, y) − Fx ,v L .W. (t, y)| = o(|
(I.5)
764
T. Chen, J. Fröhlich, A. Pizzo
as | y | → ∞, for any x . This class of solutions of (I.3) is denoted by CvL .W. . It is important to observe that, by causality, the class CvL .W. is non-empty, for any vL .W. , with | v L .W. | < 1(= c). This can be seen by choosing Cauchy data for the solution of (I.3) satisfying (I.5) at some time t0 , e.g., t0 = 0. µν Let us now consider a specific solution, F[v L .W. ] (t, y), of Eq. (I.3) in the class CvL .W. . We are interested in the behavior of this solution at very early times (t 0). We expect that, for | y − x (t)| = o(|t|), µν
µν (t, y) 0, vin
F[v L .W. ] (t, y) F
(I.6)
(here the symbol means: up to a solution of the homogeneous Maxwell equation decaying at least like t12 ). However, for | y − x (t)| → ∞, µν
µν (t, y), 0, v L .W.
F[v L .W. ] (t, y) F
(I.7)
as quantified in (I.5). µν We note that, by (I.1), F (t, y) solves Eq. (I.3), for times t < 0. Thus, 0, vin
µν
µν
µν (t, y) 0, vin
φin (t, y) := F[v L .W. ] (t, y) − F
t <0
(I.8)
solves the homogenous Maxwell equation, i.e., Eq. (I.3) with J ν ≡ 0. For t t¯, we expect that, for | y − x (t)| = o(t), µν
µν
F[v L .W. ] (t, y) Fx ∗ ,vout (t, y)
(I.9)
(here the symbol means: up to a solution of the homogeneous Maxwell equation decaying at least like t12 ). But, for | y − x (t)| → ∞, µν
µν (t, y), 0, v L .W.
F[v L .W. ] (t, y) F
(I.10)
µν
as quantified in (I.5). We note that, by (I.2), Fx ∗ ,vout (t, y) solves Eq. (I.3), for times t > t¯. Thus, µν
µν
µν
φout (t, y) := F[v L .W. ] (t, y) − Fx ∗ ,vout (t, y)
t > t¯
(I.11)
solves the homogenous Maxwell equation. µν Next, we recall that φas (t, y), with as = in/out, can be derived from an electroµ magnetic vector potential, Aas , by µν ν µ φas (t, y) = ∂ µ Aas (t, y) − ∂ ν Aas (t, y).
(I.12)
µ µ · We can impose the Coulomb gauge condition on Aas : Aas = (0, Aas (t, y)), with ∇ Aas (t, y) ≡ 0. It turns out (and this can be derived from formulae one finds, e.g., in y | → ∞), Aas (t, y) is given by [20]), that, to leading order in | y |−1 (| 3k vas · ε∗ 1 d k,λ −i k· y +i| k|t Aas (t, y) := α 2 + c.c. εk,λ e 3 ˆ 2 (1 − v | k| · k) as λ |k| vL .W. · ε∗ 1 d 3k y +i|k|t k,λ −i k· −α2 + c.c. , (I.13) ε e 23 (1 − vL .W. · k) ˆ k,λ |k| λ |k|
Infraparticle States in QED I
765
and ε , ε are transverse polarization where kˆ is the unit vector in the direction of k, k,+ k,− ∗ ˆ vectors with k · ε = 0, λ = +, −, and ε · ε = δλ,λ . k,λ
k,λ
k,λ
The free field µν
µν
φ µν (t, y) = φout (t, y) − φin (t, y)
(I.14)
is the radiation emitted by the particle due to its accelerated motion, as t → ∞. It is well known that Eqs. (I.6)-( I.13) can be made precise within classical electrodynamics under some standard assumptions on the Cauchy data for the solutions in addition to the condition in Eq. (I.5). We will see that analogous statements also hold in our model of quantum electrodynamics with non-relativistic matter. In this paper, we treat the quantum theory of a system consisting of a nonrelativistic charged particle only interacting with the quantized e.m. field. The motion of the quantum particle depends on the back-reaction of the field, and the asymptotic in- and out-velocities of this particle are not attained at finite times. However, the infrared features of the asymptotic radiation in the classical model, described above for a given current, are reproduced in this interacting quantum model. In fact, the set of classes CvL .W. , associated with different currents but at fixed vL .W. , corresponds to one of the superselection sectors of the quantized theory; see e.g. [3]. In particular, the Fock representation, which is the usual (but not the only possible) choice for the representation of the algebra of photon creation- and annihilation operators, corresponds to vL .W. = 0. This implies that, in the Fock representation of the interpolating photon creation- and annihilation operators, an infrared-singular asymptotic electromagnetic-field configuration must be present for all values of the asymptotic velocity of the electron different from zero. In particular, after replacing the classical velocities with the spectral values of the quantum operators vout/in , the background field with vL .W. = 0 (given by (I.13)) corresponds to the background radiation described by the coherent non-Fock representations of the algebra of asymptotic photon creation- and annihilation operators labeled by vout/in ; see also Sect. III.5. II. Definition of the Model The Hilbert space of pure state vectors of the system consisting of one non-relativistic electron interacting with the quantized electromagnetic field is given by H := Hel ⊗ F,
(II.1)
where Hel = L 2 (R3 ) is the Hilbert space for a single electron; (for expository convenience, we neglect the spin of the electron). The Fock space used to describe the states of the transverse modes of the quantized electromagnetic field (the photons) in the Coulomb gauge is given by F :=
∞
F (N ) ,
F (0) = C ,
(II.2)
N =0
where is the vacuum vector (the state of the electromagnetic field without any excited modes), and F (N ) := S N
N j=1
h,
N ≥ 1,
(II.3)
766
T. Chen, J. Fröhlich, A. Pizzo
where the Hilbert space h of a single photon is h := L 2 (R3 × Z2 ).
(II.4)
Here, R3 is momentum space, and Z2 accounts for the two independent transverse polarizations (or helicities) of a photon. In (II.3), S N denotes the orthogonal projection onto the subspace of Nj=1 h of totally symmetric N -photon wave functions, to account for the fact that photons satisfy Bose-Einstein statistics. Thus, F (N ) is the subspace of F of state vectors for configurations of exactly N photons. In this paper, we use units such that Planck’s constant , the speed of light c, and the mass of the electron are equal to unity. The dynamics of the system is generated by the Hamiltonian
2 x) x + α 1/2 A( −i ∇ H := (II.5) + H f. 2 The multiplication operator x ∈ R3 corresponds to the position of the electron. The x ; α ∼ electron momentum operator is given by p = −i ∇ = 1/137 is the finestructure con x) denotes the (ultraviolet stant (which, in this paper, is treated as a small parameter), A( regularized) vector potential of the transverse modes of the quantized electromagnetic field at the point x (the electron position) in the Coulomb gauge, x) = 0. x · A( ∇
(II.6)
H f is the Hamiltonian of the quantized, free electromagnetic field, given by a∗ a , H f := (II.7) d 3 k |k| k,λ k,λ λ=±
where a ∗ and ak,λ are the usual photon creation- and annihilation operators, which k,λ satisfy the canonical commutation relations ∗ [ak,λ , ak ,λ ] = δλλ δ(k − k ) ,
(II.8)
# [ak,λ ,
(II.9)
ak# ,λ ]
= 0,
with a # = a or a ∗ . The vacuum vector obeys the condition ak,λ = 0, for all k ∈ R3 and λ ∈ Z2 ≡ {+, −}. The quantized electromagnetic vector potential is given by d 3k y ∗ y −i k· ∗ i k· y ) := εk,λ A( ak,λ + ε e a e k,λ , k,λ λ=± B |k|
(II.10)
(II.11)
3 where εk,− k,+ ,ε are photon polarization vectors, i.e., two unit vectors in R ⊗ C satisfying ∗ εk,λ k,µ = δλµ , ·ε
k · εk,λ = 0,
(II.12)
Infraparticle States in QED I
767
for λ, µ = ±. The equation k · εk,λ = 0 expresses the Coulomb gauge condition. Moreover, B is a ball of radius centered at the origin in momentum space; represents an ultraviolet cutoff that will be kept fixed throughout our analysis. The vector potential defined in (II.11) is thus regularized in the ultraviolet. Throughout this paper, it will be assumed that ≈ 1 (the rest energy of an electron), and that α is sufficiently small. Under these assumptions, the Hamitonian H is selfadjoint on D(H0 ), the domain of definition of the operator H0 :=
x )2 (−i ∇ + Hf . 2
(II.13)
The perturbation H − H0 is small in the sense of Kato. The operator measuring the total momentum of a state of the system consisting of the electron and the electromagnetic field is given by P := p + P f ,
(II.14)
x is the momentum operator for the electron, and where p = −i ∇ f ∗ d 3 k k ak,λ P := ak,λ
(II.15)
λ=±
is the momentum operator for the radiation field. The operators H and P are essentially selfadjoint on the domain D(H0 ), and since = 0. The Hilbert the dynamics is invariant under translations, they commute: [H, P] space H can be decomposed on the joint spectrum, R3 , of the component-operators of Their spectral measure is absolutely continuous with respect to Lebesgue measure, P. ⊕ H := H P d 3 P, (II.16) where each fiber space H P is a copy of Fock space F. Remark. Throughout this paper, the symbol P stands for both a variable in R3 and a vector operator in H, depending on the context. Similarly, a double meaning is also associated with functions of the total momentum operator. (E.g.: In Eq. (III.1) E σ is an P operator on the Hilbert space H, while in Eq. (III.3) it is a function of P ∈ R3 .) To each fiber space H P there corresponds an isomorphism I P : H P −→ F b ,
(II.17)
where F b is the Fock space corresponding to the annihilation- and creation operators x x ∗ ∗ ∗ i k· −i k· bk,λ ak,λ a , with vacuum f = , b , where bk,λ is given by e , and b by e k,λ
k,λ
k,λ
I P (ei P·x ), where x is the electron position. To define I P more precisely, we consider an (improper) vector ψ( f (n) ; P) ∈ H P with a definite total momentum, which describes an electron and n photons. Its wave function, in the variables ( x; k1 , . . . , kn ; λ1 , . . . , λn ), is given by
ei( P−k1 −···−kn )·x f (n) (k1 , λ1 ; · · · · · · ; kn , λn ),
(II.18)
768
T. Chen, J. Fröhlich, A. Pizzo
where f (n) is totally symmetric in its n arguments. The isomorphism I P acts by way of I P ei( P−k1 −···−kn )·x f (n) (k1 , λ1 ; · · · · · · ; kn , λn ) (II.19) 1 = √ d 3 k1 . . . d 3 kn f (n) (k1 , λ1 ; · · · · · · ; kn , λn ) bk∗ ,λ · · · bk∗ ,λ f . n n 1 1 n! λ1 ,...,λn (II.20) The Hamiltonian H maps each fiber space H P into itself, i.e., it can be written as ⊗ H P d 3 P, (II.21) H = where H P : H P −→ H P .
(II.22)
∗ , and of the variable P, the fiber Hamiltonian Written in terms of the operators bk,λ , bk,λ H P has the form
2 P − P f + α 1/2 A + H f, (II.23) H P := 2 where ∗ P f = (II.24) d 3 k k bk,λ , bk,λ λ
Hf =
b∗ b , d 3 k |k| k,λ k,λ
(II.25)
d 3k ∗ ∗ bk,λ ε + ε b k,λ k,λ . k,λ |k|
(II.26)
λ
and A :=
λ
B
Let 1 }. (II.27) 3 In order to give precise meaning to the constructions used in this work, we restrict the total momentum P to the set S, and we introduce an infrared cut-off at an energy σ > 0 in the vector potential. The removal of the infrared cutoff in the construction of scattering states is the main problem solved in this paper. The restriction of P to S guarantees that the propagation speed of a dressed electron is strictly smaller than the speed of light. However, our results can be extended to a region S (inside the unit ball) of radius larger than 13 . We start by studying a regularized fiber Hamiltonian given by
2 P − P f + α 1/2 Aσ + Hf H Pσ := (II.28) 2 < S := { P ∈ R3 : | P|
Infraparticle States in QED I
769
acting on the fiber space H P , for P ∈ S, where Aσ :=
λ
d 3k ∗ ∗ bk,λ ε + ε b k,λ k,λ , k,λ B \Bσ |k|
(II.29)
and where Bσ is a ball of radius σ . Remark. In a companion paper [11], we construct dressed one-electron states of fixed σ σ momentum given by the ground state vectors j of the Hamiltonians H j , and we σ
σ
P
P
compare ground state vectors j , j corresponding to different fiber Hamiltonians P P σ σ H j , H j with P = P . We compare these ground state vectors as vectors in the Fock P
P
space F b . In the sequel, we use the expression σ P
σ
j − j F P
(II.30)
as an abbreviation for σ P
σ
I P ( j ) − I P ( j )F ; P
(II.31)
· F stands for the Fock norm. Hölder continuity properties of σ in σ and in P are P proven in [11]. These properties play a crucial role in the present paper. II.1. Summary of contents. In Sect. III, time-dependent vectors ψh,κ (t) approximating scattering states are constructed, and the main results of this paper are described, along with an outline of infraparticle scattering theory. In Sects. IV and V, ψh,κ (t) is shown to out/in in the Hilbert space H, as time t tends to infinity. converge to a scattering state ψh,κ out/in
This result is based on mathematical techniques introduced in [26]. The vector ψh,κ represents a dressed electron with a wave function h on momentum space whose support is contained in the set S (see details in Sect. III.1), accompanied by a cloud of soft photons described by a Bloch-Nordsieck operator, and with an upper cutoff κ imposed on photon frequencies. This cutoff can be chosen arbitrarily. In Sect. VI, we construct the scattering subspaces Hout/in . Vectors in these subspaces are obtained from certain subspaces, H˚ out/in , by applying “hard” asymptotic out/in photon creation operators. These spaces carry representations of the algebras A ph out/in
and Ael of asymptotic photon creation- and annihilation operators and asymptotic electron observables, respectively, which commute with each other. The latter property proves asymptotic decoupling of the electron and photon dynamics. We rigorously out/in establish the coherent nature and the infrared properties of the representation of A ph identified by Bloch and Nordsieck in their classic paper, [2]. In a companion paper [11], we establish the main spectral ingredients for the construction and convergence of the vectors { σ }, as σ tends to 0. These results are obtained P with the help of a new multiscale method introduced in [25], to which we refer the reader for some details of the proofs. In the Appendix, we prove some technical results used in the proofs.
770
T. Chen, J. Fröhlich, A. Pizzo
III. Infraparticle Scattering States Infraparticle scattering theory is concerned with the asymptotic dynamics in QFT models of infraparticles and massless fields. Contrary to theories with a non-vanishing mass gap, the picture of asymptotically freely moving particles in the Fock representation is not valid, due to the inseparability of the dynamics of charged massive particles and the soft modes of the massless asymptotic fields. Our starting point is to study the (dressed one-particle) states of a (non-relativistic) electron when the interactions with the soft modes of the photon field are turned off. We then analyze their limiting behavior when this infrared cut-off is removed. This amounts to studying vectors ψ σ , σ > 0, in the Hilbert space H that are solutions to the equation H σ ψ σ = E σP ψ σ ,
(III.1)
⊕ σ 3 E σ is where H σ = H d P, and E σ is a function of the vector operator P; P P P the electron energy function defined more precisely in Sect. III.1. Since in our model non-relativistic matter is coupled to a relativistic field, the form of E σ is not fixed by P symmetry, except for rotation invariance. Furthermore, the solutions of (III.1) give rise to vectors in the physical Hilbert space describing wave packets of dressed electrons of the form σ d 3 P, ψ σ (h) = h( P) (III.2) P where the support of h is contained in a ball centered at P = 0, chosen such that i.e., we must impose the condition that the maximal E σ | < 1, as a function of P, |∇ P group velocity of the electron which, a priori, is not bounded from above in our nonrelativistic model, is bounded by the speed of light. (For group velocities larger than the velocity of light, the one-electron states decay by emission of Cerenkov radiation.) The guiding principle motivating our analysis of limiting or improper one-particle states, ψ σ (h) for σ → 0, is that refined control of the infrared singularities, which push these vectors out of the space H, as σ → 0, should enable one to characterize the soft photon cloud encountered in the scattering states. The analysis of Bloch and Nordsieck, [2], suggests that the infrared behavior of the state describing the soft photons accompanying an electron should be singular (i.e., not square-integrable at the origin in photon momentum space), and that it should be determined by the momentum of the asymptotic electron. In mathematical terms, this means that the asymptotic electron velocity is expected to determine an asymptotic Weyl operator (creating a cloud of asymptotic photons), which when applied to a dressed one-electron state ψ σ =0 (h) yields a well defined vector in the Hilbert space H. This vector is expected to describe an asymptotic electron with wave function h surrounded by a cloud of infinitely many asymptotic free photons, in accordance with the observations sketched in (I.6)–(I.13). Our goal in this paper is to translate this physical picture into rigorous mathematics, following suggestions made in [15] and methods developed in [9,10,25,26]. III.1. Key spectral properties. In our construction of scattering states, we make extensive use of a number of spectral properties of our model proven in [11], and summarized in Theorem III.1 below; (they are analogous to those used in the analysis of Nelson’s model in [26]).
Infraparticle States in QED I
771
We define the energy of a dressed one-electron state of momentum P by E σP = inf specH Pσ ,
E P = inf specH P = E σP=0 .
(III.3)
We refer to E σ as the ground state energy of the fiber Hamiltonian H σ . We assume that P P the finestructure constant α is so small that E σ | < νmax < 1 |∇ P
(III.4)
< 1 }, for some constant νmax < 1, uniformly in σ . for all P ∈ S := { P ∈ R3 : | P| 3 σ E , we introduce a Weyl operator Corresponding to ∇ P
⎛ E σ ) := exp ⎝ α 2 W σ (∇ P
1
λ
B \Bσ
d 3k
⎞
Eσ ∇ P
23 δ ( |k| P,σ k)
∗ ⎠ · (εk,λ bk,λ − h.c.) ,
(III.5)
where σ k , δ P,σ (k) := 1 − ∇ E P · |k|
(III.6)
acting on H P , which is unitary for σ > 0. We consider the transformed fiber Hamiltonian E σ )H σ Wσ∗ (∇ E σ ). K Pσ := Wσ (∇ P P P
(III.7)
E σ ) acts on the creation- and annihilation operators We note that conjugation by Wσ (∇ P as a linear Bogoliubov transformation (translation) E σ ) b# Wσ∗ (∇ E σ ) = b# − α 1/2 W σ (∇ P k,λ P k,λ
1σ, (k) E σ · ε # , ∇ 3 P k,λ 2 δ ( |k| k)
(III.8)
P,σ
stands for the characteristic function of the set B \Bσ . Our methods where 1σ, (k) rely on proving regularity properties in σ and P of the ground state vector, σ , and of P the ground state energy, E σ , of K σ . These regularity properties are summarized in the P P following theorem, which is the main result of the companion paper [11]. Theorem III.1. For P ∈ S and for α > 0 sufficiently small, the following statements hold. (I 1) The energy E σ is a simple eigenvalue of the operator K σ on F b . Let Bσ := P P ≤ σ }, and let Fσ denote the Fock space over L 2 ((R3 \Bσ ) × Z2 ). {k ∈ R3 | |k| Likewise, we define F0σ to be the Fock space over L 2 (Bσ × Z2 ); hence F b = Fσ ⊗ F0σ . On Fσ , the operator K σ has a spectral gap of size ρ − σ or larger, sepP arating E σ from the rest of its spectrum, for some constant ρ − , with 0 < ρ − < 1. P The contour γ := {z ∈ C ||z − E σP | =
ρ−σ } , σ >0 2
(III.9)
772
T. Chen, J. Fröhlich, A. Pizzo
bounds a disc which intersects the spectrum of K σ |Fσ in only one point, {E σ }. P P The ground state vectors of the operators K σ are given by P
σP
:=
1 2π i
2π1 i
1 γ K σ −z
dz f
P
1 γ K σ −z P
(III.10)
dz f F
and converge strongly to a non-zero vector P ∈ F b , in the limit σ → 0. The 1
rate of convergence is at least of order σ 2 (1−δ) , for any 0 < δ < 1. (Although it is not relevant for the purposes of this paper, we note that the results in [17] imply 1 the uniformity in δ of the range of values of α, where the rate estimate σ 2 (1−δ) holds; analogous conclusions follow for the rate estimates below.) The dependence of the ground state energies E σ of the fiber Hamiltonians K σ P P on the infrared cutoff σ is characterized by the following estimates:
|E σP − E σP | ≤ O(σ ),
(III.11)
and
Eσ − ∇ E σ | ≤ O(σ 2 (1−δ) ), |∇ P P 1
(III.12)
for any 0 < δ < 1, with σ > σ > 0. (I 2) The following Hölder regularity properties in P ∈ S hold uniformly in σ ≥ 0: 4 −δ σP − σP+ ≤ Cδ | P| P F
(III.13)
4 −δ , Eσ − ∇ Eσ |∇ | ≤ Cδ | P| P P+ P
(III.14)
1
and 1
P + P ∈ S, where Cδ and Cδ are finite for any 0 < δ < δ < 41 , with P, constants depending on δ and δ , respectively. (I 3) Given a positive number νmin , there are numbers rα = νmin + O(α) > 0 and νmax < 1 such that, for P ∈ S\Brα and for α sufficiently small, E σ | > νmin > 0, 1 > νmax > |∇ P
(III.15)
uniformly in σ . (We also notice that the control on the second derivative of E σ P in P uniformly in the sharp infrared cut-off σ ≥ 0 (see [17]) would allow us to E σ arbitrarily close take νmin ≡ 0, rα ≡ 0, and to include electron velocities ∇ P to 0, but we prefer to work with an assumption self-contained in the paper). (I 4) For P ∈ S and for any k = 0, the following inequality holds uniformly in σ , for α small enough: σ E σP− k > E P − C α |k|,
where E σ
P−k
:= inf spec H σ
P−k
and
1 3
< Cα < 1, with Cα →
(III.16) 1 3
as α → 0.
Infraparticle States in QED I
773
(I 5) Let σ ∈ F denote the ground state vector of the fiber Hamiltonian H σ , so that P
P
Eσ ) σP = ζ Wσ (∇ P
σ P
σ F
ζ ∈ C, |ζ | = 1.
(III.17)
1σ, (k) , 3/2 |k|
(III.18)
P
For P ∈ S, one has that bk,λ
σ
P σ F P
F ≤ C α 1/2
where σ is the ground state of H σ and C is a positive constant; see Lemma 6.1 P P of [10] which can be extended to k ∈ R3 using (I 4). Detailed proofs of Theorem III.1 based on results in [25,10] are given in [11]. III.2. Definition of the approximating vector h,κ (t). We construct infraparticle scattering states by using a time-dependent approach to scattering theory. We define a time-dependent approximating vector ψh,κ (t) that converges to an asymptotic vector, as t → ∞. It describes an electron with wave function h (whose momentum space support is contained in S), and a cloud of asymptotic free photons with an upper photon frequency cutoff 0 < κ ≤ . This interpretation will be justified a posteriori. We closely follow an approach to infraparticle scattering theory developed for Nelson’s model in [26], (see also [15]). In the context of the present paper, our task is to give a mathematically rigorous meaning to the formal expression σ
i Ht out Wκ,σ ( v (t), t) e−i H t ψ σ (h), κ (h) := lim lim e t→∞ σ →0
where
⎛
v (t), t) := exp ⎝α 2 Wκ,σ (
1
λ
d 3k
Bκ \Bσ
(III.19)
∗ −i|k|t − ε i|k|t } ∗ ak,λ v(t) · {εk,λ a e e k,λ
k,λ
− |k|(1 k · v(t))
|k|
⎞ ⎠.
The operator v(t) is not known a priori; but, in the limit t → ∞, it must converge to the asymptotic velocity operator of the electron. The latter is determined by the operator E , applied to the (non-Fock) vectors . This can be seen by first considering the ∇ P P infrared regularized model, with σ > 0, which has dressed one-electron states ψ σ (h) in H, and by subsequently passing to the limit σ → 0. Formally, for σ → 0, the Weyl operator ei H t Wκ,σ ( v (t), t) e−i H
σt
(III.20)
is an interpolating operator used in the L.S.Z. (Lehmann-Symanzik-Zimmermann) approach to scattering theory for the electromagnetic field, where the photon test functions (in the operator Wκ,σ ( v (t), t)) are evolved backwards in time with the free evolution, and the photon creation- and annihilation operators are evolved forward in time with the interacting time evolution. Moreover, the photon test functions in (III.20) coincide E σ ) defined in (III.5), after replacing with the test functions in the Weyl operator Wσ (∇ P
774
T. Chen, J. Fröhlich, A. Pizzo
E σ by the operator v(t). We stress that, while the Weyl operator Wσ (∇ Eσ ) the operator ∇ P P leaves the fiber spaces H P invariant, the Weyl operator Wκ,σ ( v (t), t) is expressed in terms of the operators {a, a ∗ }, as it must be when describing real photons in a scattering process, and hence does not preserve the fiber spaces. E σ , as t → ∞ and σ → 0, two Guided by the expected relation between v(t) and ∇ P key ideas used to make (III.19) precise are to render the infrared cut-off time-dependent, < 1 }, with a grid with σt → 0, as t → ∞, and to discretize the ball S = { P ∈ R3 | | P| 3 size decreasing in time t. This discretization also applies to the velocity operator v(t) in expression (III.19). The existence of infraparticle scattering states in H is established by proving that the corresponding sequence of time-dependent approximating scattering states, which depend on the cutoff σt and on the discretization, defines a strongly convergent sequence of vectors in H. This is accomplished by appropriately tuning the convergence rates of σt and of the discretization of S. Our sequence of approximate infraparticle scattering states is defined as follows (for t 1): i) We consider a wave function h with support in a region R which is a union of cubes contained in S\Brα ; (see condition (I 3) in Theorem III.1). We introduce a time-dependent cell partition G (t) of R. This partition is constructed as follows: At time t, the linear dimension of each cell is 2Ln , where L is the diameter of R, and n ∈ N is such that 1
1
(2n ) ≤ t < (2n+1 ) ,
(III.21)
for some > 0 to be fixed later. Thus, the total number of cells in G (t) is N (t) = (t) 23n , where n = log2 t ; (x extracts the integer part of x). By G j , we denote the jth cell of the partition G (t) . ii) For each cell, we consider a one-particle state of the Hamiltonian H σt , (t) σt d 3 P, ψ j,σt := h( P) (III.22) G j(t)
P
where ∈ C 1 (S\Brα ), with supp h ⊆ R; • h( P) 0 • σt := t −β , for some exponent β (> 1) to be fixed later; • in (III.22), the ground state vector, σt , of H σt is defined by P
P
σt := Wσ∗t (∇ E σt ) σt , P
P
P
(III.23)
where σt is the ground state of K σt ; (see Theorem III.1). P
P
(t)
iii) With each cell G j we associate a soft-photon cloud described by the following “LSZ (Lehmann-Symanzik-Zimmermann) Weyl operator”: ei H t Wκ,σt ( v j , t) e−i H where
⎛
v j , t) := exp ⎝α Wκ,σt (
1 2
σt t
,
(III.24)
⎞ ∗ −i|k|t − ε∗ ak,λ ei|k|t } ak,λ d 3 k v j · {εk,λ e k,λ ⎠. − |k|(1 k · v j ) Bκ \Bσt |k|
λ
(III.25)
Infraparticle States in QED I
775
• Here κ, with 0 < κ ≤ , is an arbitrary (but fixed) photon energy threshold or counter threshold. E σt ( P ∗ ) is the c-number vector corresponding to the value of the • v j ≡ ∇ j in the center, P ∗ , of the cell G (t) . E σt ( P) “velocity” ∇ j
j
iv) For each cell, we consider a time-dependent phase factor e
σ P
E t ,t) iγσt ( v j ,∇
,
(III.26)
with E σt , t) := −α γσt ( vj, ∇
P
E σt · ∇
t 1
P
cos(k · ∇ ) d 3 k dτ, E σt τ − |k|τ v (k) j P
Bσ S \Bσt τ
(III.27) and := 2 vl j (k)
l
(δl,l −
1 kl kl l . ) vj 2 2 |k| |k| (1 − k · v j )
(III.28)
Here, στS := τ −θ , and the exponent 0 < θ < 1 will be chosen later. Note that, in E σt is interpreted as an operator. (III.26), (III.27), ∇ P v) The approximate scattering state at time t is given by the expression ψh,κ (t) := ei H t
N (t)
Wκ,σt ( v j , t) e
E σt ,t) −i E σt t iγσt ( v j ,∇ P
e
P
(t)
ψ j,σt ,
(III.29)
j=1
where N (t) is the number of cells in G (t) . E σt ,t) iγ ( v ,∇
P is similar to that of the Coulomb The role played by the phase factor e σt j phase in Coulomb scattering. However, in the present case, the phase has a limit, as t → ∞, and is introduced to control an oscillatory term in the Cook argument which is not absolutely convergent (see Sect. III.3).
III.3. Statement of the main result. The main result of this paper is Theorem III.2, below, from which the asymptotic picture described in Sect. III.5, below, emerges. It relies on the assumptions summarized in the following hypothesis. Main Assumption III.1. The following assumptions hold throughout this paper: (1) The conserved momentum P takes values in S; see (II.27). (2) The finestructure constant α satisfies α < αc , for some small constant αc 1 independent of the infrared cutoff. (3) The wave function h is supported in a set R and is of class C 1 , where R is contained in S\Brα , as indicated in Fig. 1, and rα is introduced in (I 3).
776
T. Chen, J. Fröhlich, A. Pizzo
Fig. 1. R can be described as a union of cubes with sides of length 2Ln 0 , for some n 0 < ∞
Theorem III.2. Given the Main Assumption III.1, the following holds: There exist positive real numbers β > 1, θ < 1 and > 0 such that the limit (out) s − lim ψh,κ (t) =: ψh,κ t→+∞
(III.30)
(where ψh,κ (t) is defined in Eq. (III.29) and κ, see (III.25), is the threshold frequency)
(out) 2 d 3 P. Furthermore, the rate of conexists as a vector in H, and ψh,κ 2 = |h( P)| −ρ vergence is at least of order t , for some ρ > 0. We note that this result corresponds to Theorem 3.1 of [26] for Nelson’s model. The limiting state is the desired infraparticle scattering state without infrared cut-offs. We shall verify that {ψh,κ (t)} is a Cauchy sequence in H, as t → +∞; (or t → −∞). In Sect. III.4, we outline the key mechanisms responsible for the convergence of the approximating vectors ψh,κ (t), as t → ∞. We note that, in (III.30), three different convergence rates are involved: • The rate t −β related to the fast infrared cut-off σt ; • the rate t −θ , related to the slow infrared cut-off σtS (see (III.27)); • the rate t − of the grid size of the cell partition. We anticipate that, in order to control the interaction, • β has to be larger than 1, due to the time-energy uncertainty principle. • The exponent θ has to be smaller than 1, in order to ensure the cancelation of some “infrared tails” discussed in Sect. IV. • The exponent , which controls the rate of refinement of the cell decomposition, will have to be chosen small enough to be able to prove certain decay estimates.
Infraparticle States in QED I
777
III.4. Strategy of convergence proof. Here we outline the key mechanisms used to prove that the approximating vectors ψh,κ (t) converge to a nonzero vector in H, as t → ±∞. Among other things, we will prove that 1 2 d3 P ) 2 . (III.31) lim ψh,κ (t) = h2 := ( |h( P)| t→∞
From its definition, see (III.29), one sees that the square of the norm of the vector ψh,κ (t) involves a double sum over cells of the partitions G (t) , i.e., ψh,κ (t)2 =
N (t)
e
E σt ,t) −i E σt t iγσt ( vl ,∇ P
e
P
(t)
∗ ψl,σt , Wκ,σ ( vl , t) t
l, j=1
× Wκ,σt ( v j , t) e
σ P
σ P
E t ,t) −i E t t iγσt ( v j ,∇
e
(t) ψ j,σt ,
(III.32)
where the individual terms, labeled by (l, j), are inner products between vectors labeled by cells Gl(t) and G j(t) of G (t) . A heuristic argument to see where (III.31) comes from is as follows. Assuming that • the vectors ψh,κ (t) converge to an asymptotic vector of the form σ
v (t), t) e−i H t ψ σ (h) = Wκ,σ =0 ( v (±∞)) ψ σ =0 (h), lim lim ei H t Wκ,σ ( out/in
t→±∞ σ →0
(III.33) where
⎛
out/in Wκ,σ =0 ( v (±∞))
out/in ∗
:= exp ⎝ α
1 2
λ
Bκ
out/in ∗
− ε∗ a ak,λ d 3 k v(±∞) · {εk,λ k,λ k,λ | k|(1 − k · v (±∞)) |k|
out/in
}
⎞ ⎠,
out/in
and a , a are the creation- and annihilation operators of the asymptotic k,λ k,λ photons; • the operators v(±∞) commute with the algebra of asymptotic creation- and anniout/in ∗ out/in hilation operators {a , a }; (this can be expected to be a consequence of k,λ k,λ asymptotic decoupling of the photon dynamics from the dynamics of the electron); • the restriction of the asymptotic velocity operators, v(±∞), to the improper dressed E , i.e., one-electron state is given by the operator ∇ P E ; v(±∞) P ≡ ∇ P P
(III.34)
then, the two vectors Wκ,σt ( v j , t) e
σ P
σ P
iγσt ( v j ,∇ E t ,t) −i E t t
e
(t)
ψ j,σt and Wκ,σt ( vl , t) e
σ P
σ P
iγσt ( vl ,∇ E t ,t) −i E t t
e
(t)
ψl,σt
(III.35) corresponding to two different cells of G (t) (i.e., j = l) turn out to be orthogonal in the limit t → ±∞. One can then show that the diagonal terms in the sum (III.32) are the only ones that survive in the limit t → ∞. The fact that their sum converges to h22 is comparatively easy to prove.
778
T. Chen, J. Fröhlich, A. Pizzo
A mathematically precise formulation of this mechanism is presented in Sect. IV. In Sect. IV.1, part A., the analysis of the scalar products between the cell vectors in (III.35) is reduced to the study of an ODE. To prove (III.31), we invoke the following properties (t) (t) of the one-particle states ψ j,σt and ψl,σt located in the jth and lth cell, respectively: • Their spectral supports with respect to the momentum operator P are disjoint up to sets of measure zero. • They are vacua for asymptotic annihilation operators, as long as an infrared cut-off σt for a fixed time t is imposed: For Schwartz test functions g λ , we define σt i H σt s λ i|k|s aσout/in (g) := lim e e−i H s d 3 k, (III.36) ak,λ g (k) e t s→±∞
on the domain of
λ
H σt .
An important step in the proof of (III.31) is to control the decay in time of the offdiagonal terms. After completion of this step, one can choose the rate, t − , by which the diameter of the cells of the partition G (t) tends to 0 in such a way that the sum of the off-diagonal terms vanishes, as t → ∞. Precise control is achieved in Sect. IV.1, part B., where we invoke Cook’s argument and analyze the decay in time s of
d i H σt s E σt ,s) −i E σt s (t) iγ ( v ,∇ P e (III.37) Wκ,σt ( v j , s) e σt j e P ψ j,σt ds = i ei H
σt s
+ i ei H
[H Iσt , Wκ,σt ( v j , s)] e
σt s
Wκ,σt ( v j , s)
σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
ψ (t) j,σt
(III.38)
E σt , s) dγσt ( vj, ∇ E σt ,s) −i E σt s iγσt ( v j ,∇ P P P ds
e
e
(t)
ψ j,σt , (III.39)
for a fixed infrared cut-off σt , and a fixed partition. As we will show, the term in (III.38) can be written (up to a unitary operator) as 1 E σt ,s) (t) iγ ( v ,∇ 3 3 P d y Jσt (s, y) · q ) cos( q · y − | q |s) d q e σt j ψ j,σt α2 v j ( Bκ \Bσt
(III.40) plus subleading terms, where Jσt (s, y) is essentially the electron current at time s, which is proportional to the velocity operator i [H σt , x ] = p + α 2 Aσ t ( x). 1
(III.41)
In (III.40), the electron current is smeared out with the vector function v ( q ) cos( q · y − | q |s) d 3 q, gt (s, y) := j Bκ \Bσt
(III.42)
which solves the wave equation s,y gt (s, y) = 0,
(III.43) σ
E t ,s) iγ ( v ,∇
(t)
P ψ j,σt . Because of the dispersive and is then applied to the one-particle state e σt j properties of the dynamics of the system, the resulting vector is expected to converge to 0 in norm at an integrable rate, as s → ∞. An intuitive explanation proceeds as follows:
Infraparticle States in QED I
779
i) A vector function gt (s, y) that solves (III.43) propagates along the light cone, and supy∈R3 | gt (s, y)| decays in time like s −1 , while a much faster decay is observed
when y is restricted to the interior of the light cone (i.e., | |ys | | < 1). ii) Because of the support in P of the vector ψ (t) j,σt , the propagation of the electron current in (III.38) is limited to the interior of the light cone, up to subleading tails. Combination of i) and ii) is expected to suffice to exhibit decay of the vector norm of (III.38) and to complete our argument. An important refinement of this reasoning process, involving the term (III.39), is, however, necessary: A mathematically precise version of statement ii) is as follows: Let χh be a smooth, approximate characteristic function of the support of h. We will prove a propagation estimate σt σt χh ( x ) eiγσt (v j ,∇ E P ,s) e−i E P s ψ (t) j,σt s σt σt E σt ) eiγσt (v j ,∇ E P ,s) e−i E P s ψ (t) − χ h (∇ j,σt P ≤
1 1 | ln(σt )|, s ν t 32
(III.44)
as s → ∞, where ν > 0 is independent of . Using result (I 3) of Theorem III.1, and our assumption on the support of h formulated in point ii) of Sect. III.2, this estimate provides sufficient control of the asymptotic dynamics of the electron. An important modification of the argument above is necessary because of the dependence of v ( gt (s, y) := q ) cos( q · y − | q |s) d 3 q, (III.45) j Bκ \Bσt
on t, which cannot be neglected even if y is in the interior of the light cone. In order to exhibit the desired decay, it is necessary to split gt (s, y) into two pieces, v ( q ) cos( q · y − | q |s)d 3 q (III.46) j Bκ \Bσ S s
and
Bσ S \Bσt
v ( q ) cos( q · y − | q |s) d 3 q j
(III.47)
s
for s such that σsS > σt , where σsS = s −θ , with 0 < θ < 1. (The same procedure will also be used in (III.50), below.) The function (III.46) has good decay properties inside the light cone. Expression (III.40), with gt (s, y) replaced by (III.46), can be controlled by standard dispersive estimates. The other contribution, proportional to (III.47), is in principle singular in the infrared region, but is canceled by (III.39). This can be seen by using a propagation estimate similar to (III.44). This strategy has been designed in [26]. However, because of the vector nature of the interaction in non-relativistic QED, the cancelation in our proof is technically more subtle than the one in [26]. After having proven the uniform boundedness of the norms of the approximating vectors ψh,κ (t), one must prove that they define a Cauchy sequence in H. To this end,
780
T. Chen, J. Fröhlich, A. Pizzo
we compare these vectors at two different times, t2 > t1 (for the limit t → +∞), and split their difference into ψh,κ (t1 ) − ψh,κ (t2 ) = ψ(t2 , σt2 , G (t2 ) → G (t1 ) ) + ψ(t2 → t1 , σt2 , G (t1 ) ) + ψ(t1 , σt2 → σt1 , G (t1 ) ),
(III.48)
where the three terms on the r.h.s. correspond to I) changing the partition G (t2 ) → G (t1 ) in ψh,κ (t2 ): ψ(t2 , σt2 , G (t2 ) → G (t1 ) ) := ei H t2
N (t1 )
Wκ,σt2 ( v j , t2 ) e
σt
σt
E 2 ,t2 ) −i E 2 t2 iγσt ( v j ,∇ 2
e
P
(t )
ψ j,σ1 t
P
j=1
− e i H t2
N (t1 )
Wκ,σt2 ( vl( j) , t2 )e
σ
2
σ
E t2 ,t2 ) −i E t2 t2 iγσt ( vl( j) ,∇ 2
e
P
(t )
ψl( 2j),σt ,
P
2
j=1 l( j)
(III.49) where l( j) labels all cells of G (t2 ) contained in the jth cell G j(t1 ) of G (t1 ) . Moreover, σt
E ∗2 vl( j) ≡ ∇
and
Pl( j)
σt
E ∗1 ; v j ≡ ∇ Pj
II) subsequently changing the time, t2 → t1 , for the fixed partition G (t1 ) , and the fixed infrared cut-off σt2 : ψ(t2 → t1 , σt2 , G (t1 ) ) := e
i H t1
N (t1 )
Wκ,σt2 ( v j , t1 ) e
σt
σt
E 2 ,t1 ) −i E 2 t1 iγσt ( v j ,∇ 2
e
P
P
(t )
ψ j,σ1 t
j=1 N (t1 )
− e i H t2
Wκ,σt2 ( v j , t2 ) e
σ
σ
E t2 ,t2 ) −i E t2 t2 iγσt ( v j ,∇ 2
e
P
P
2
(t )
ψ j,σ1 t ; 2
j=1
(III.50)
and, finally, III) shifting the infrared cut-off, σt2 → σt1 : ψ(t1 , σt2 → σt1 , G (t1 ) ) := ei H t1
N (t1 )
Wκ,σt1 ( v j , t1 )e
σ
σ
E t1 ,t1 ) −i E t1 t1 iγσt ( v j ,∇ 1
e
P
P
j=1
−e
i H t1
N (t1 )
Wκ,σt2 ( v j , t1 )e
σt
σt
E 2 ,t1 ) −i E 2 t1 iγσt ( v j ,∇
j=1
2
P
e
P
(t )
ψ j,σ1 t
1
(t )
ψ j,σ1 t . 2
(III.51)
It is important to take these three steps in the order indicated above. In Step I), the size of ψ(t2 , σt2 , G (t2 ) → G (t1 ) )2 in (III.49) is controlled as follows: The sum of off-diagonal terms yields a subleading contribution. The diagonal terms are shown to tend to 0 by controlling the differences vl( j) − v j .
Infraparticle States in QED I
781
In Step II), Cook’s argument, combined with the cancelation of an infrared tail (as in the mechanism described above), yields the desired decay in t1 . Step III) is more involved. But the basic idea is quite simple to grasp: It consists in rewriting e i H t1
N (t1 )
Wκ,σt2 ( v j , t1 ) e
σ
σ
E t2 ,t1 ) −i E t2 t1 iγσt ( v j ,∇ 2
e
P
P
(t )
ψ j,σ1 t
j=1
(III.52)
2
as e
i H t1
N (t1 )
σt2
E σt2 ) Wσt (∇ E σt2 ) eiγσt2 (v j ,∇ E P Wκ,σt2 ( v j , t1 ) Wσ∗t (∇ 2 P
2
σt
,t1 ) −i E 2 t1
e
P
P
j=1
(t )
ψ j,σ1 t . 2
(III.53) (t1 )
The term in (III.53) corresponding to the cell G j acting with the “dressing operator”
of G (t1 ) can then be obtained by
σt
E 2)e ei H t1 Wκ,σt2 ( v j , t1 ) Wσ∗t (∇ on the “infrared-regular” vector σ
E t2 ,t1 ) iγσt ( v j ,∇ 2
P
(t )
1 j,σ := e t
P
P
2
e
σt
−i E 2 t1
σ
E t2 ,t1 ) iγσt ( v j ,∇ 2
P
(III.54)
σt
(t1 )
Gj
2
,
2 3 h( P) d P
P
(III.55)
E σ ) σ (see (III.23)), for all j. The advancorresponding to the vectors σ = Wσ (∇ P
P
P
1) inherits the regularity properties of tage of (III.53) over (III.52) is that the vector (tj,σ t 2
(t )
1 σ described in Theorem III.1. In particular, the vectors j,σ converge strongly, as t2 P σt2 → 0, and the vector
(t )
1 e−i q·x j,σ t
(III.56)
2
depends on q in a Hölder continuous manner, uniformly in σt2 . This last property entails enough decay to offset various logarithmic divergences appearing in the removal of the infrared cut-off in the dressing operator (III.54). Our analysis of the strong convergence of the sequence of approximating vectors culminates in the estimate
ρ ψh,κ (t2 ) − ψh,κ (t1 ) ≤ O (ln(t2 ))2 /t1 , (III.57) for some ρ > 0. By telescoping, this bound suffices to prove Theorem III.2. Indeed, to estimate the difference between the two vectors at times t2 and t1 , respectively, where t2 > t1 > 1, we may consider a sequence of times {t12 , ..., t1n }, such that t1n ≤ t2 < t1n+1 , and use Estimate (III.57) for each difference ψh,κ (t2 ) − ψh,κ (t1n ), ψh,κ (t1m ) − ψh,κ (t1m−1 ),
(III.58) 2 ≤ m ≤ n.
(III.59)
Then, one can show that there exists a constant ρ > 0 such that the rate of convergence of the time-dependent vector is at least of order t −ρ , as stated in Theorem III.2.
782
T. Chen, J. Fröhlich, A. Pizzo (out/in)
III.5. A space of scattering states. We use the asymptotic states ψh,κ to construct a out/in subspace, H˚ κ , of scattering states invariant under space-time translations, and with a photon energy threshold κ, H˚ κout/in :=
out/in
ψh,κ
∈ C 1 (S\Brα ), τ ∈ R, a ∈ R3 , (τ, a ) : h( P) 0
(III.60)
where out/in
ψh,κ
out (τ, a ) ≡ e−i a· P e−i H τ ψh,κ .
(III.61)
This space contains states describing an asymptotically freely moving electron, accompanied by asymptotic free photons with energy smaller than κ. out/in Spaces of scattering states are obtained from the space H˚ κ by adding (and subtracting) “hard” photons, i.e., Hout/in :=
ψ
out/in h, F
∈ C 1 (S\Brα ), Fˆ ∈ C ∞ (R3 \0 ; C3 ) , : h( P) 0 0
(III.62)
where out/in ψ h, F
:= s −
lim
t→+/−∞
e
Ft ,t]− A[ Ft ,t] i A[
ψh,κ (t),
(III.63)
and Ft , t] := i A[
y) y ) ∂ A(t, ∂ Ft ( − · Ft ( A(t, y) · y) d 3 y ∂t ∂t
(III.64)
is the L.S.Z. photon field smeared out with the vector test function y ) := Ft (
λ=±
d 3k y e−i|k|t+i k· √ ε∗ Fˆ λ (k) (2π )3 2 |k| k,λ
(III.65)
with ˆ k) ∗ ˆλ ∞ 3 3 := F( εk,λ F (k) ∈ C 0 (R \{0} ; C ).
(III.66)
λ
An a posteriori physical interpretation of the scattering states constructed here emerges by studying how certain algebras of asymptotic operators are represented on the spaces of scattering states: out/in
• The Weyl algebra, A ph •
, associated with the asymptotic electromagnetic field.
out/in The algebra Ael
generated by smooth functions of compact support of the asymptotic velocity of the electron.
These algebras will be defined in terms of the limits (III.67) and (III.69), below, whose existence is established in Sect. VI.2.
Infraparticle States in QED I
783
Theorem III.3. Functions f ∈ C0∞ (R3 ), of the variable ei H t its, as t → ±∞, as operators acting on Hout/in , s−
x −i H t , have strong limt e
x out/in out/in ei H t f ( ) e−i H t ψ =: ψ , h, F f ∇ E h, F t→+/−∞ t lim
(III.67)
:= limσ →0 f (∇ E σ ). where f ∇ E ( P) P
Theorem III.4. The LSZ Weyl operators
G t ,t]− A[ G t ,t] i A[ ∈ L 2 (R3 , (1 + |k| −1 )d 3 k), λ = ± , e : Gˆ λ (k)
(III.68)
have strong limits in Hout/in ; i.e., W
out/in
:= s − (G)
lim
t→+/−∞
e
G t ,t]− A[ G t ,t] i A[
(III.69)
exists. The limiting operators are unitary and satisfy the following properties: i) W out/in (G ) = W out/in (G +G ) e− W out/in (G) where
G ) = 2i I m ρ(G,
λ
G ) ρ(G, 2
(III.70)
Gˆ (k) d k G (k) ˆλ
,
3
.
(III.71)
λ
defines a strongly continuous one-parameter ii) The mapping R s −→ W out/in (s G) group of unitary operators. iii) e−i H τ = W out/in (G −τ ), ei H τ W out/in (G)
(III.72)
−τ is the freely time-evolved (vector) test function at time −τ . where G out/in
out/in
The two algebras, A ph and Ael , commute. This is the precise mathematical expression of the asymptotic decoupling of the dynamics of photons from the one of the electron. The proof is non-trivial, because non-Fock representations of the asymptotic out/in photon creation- and annihilation operators appear. (For the representation of A ph , which is non-Fock but locally Fock see Sect. VI.2.) We will show that 2 G out/in − 2 2 ∇ E P (G) out/in = |2 d 3 P, e e | h( P) (III.73) ψh,κ , W out (G)ψ h,κ where 2 = G
k)| 2d 3k |G(
1/2 ,
(III.74)
784
and
T. Chen, J. Fröhlich, A. Pizzo
1 u (G) := 2i Re α 2 λ
Bκ
Gˆ λ (k)
u · ε∗
k,λ
2 (1 − u · k) ˆ |k| 3
d k . 3
(III.75)
out/in
on the space of scattering states can be More precisely, the representation of A ph decomposed in a direct integral of inequivalent irreducible representations labelled by the asymptotic velocity of the electron. For different values of the asymptotic velocity, these representations turn out to be inequivalent. Only for a vanishing electron velocity, the representation is Fock; for non-zero velocity, it is a coherent non-Fock representation. The coherent photon cloud, labeled by the asymptotic velocity, is the one anticipated by Bloch and Nordsieck in the non-relativistic approximation. These results can be interpreted as follows: In every scattering state, an asymptotically freely moving electron is observed (with an asymptotic velocity whose size is strictly smaller than the speed of light, by construction) accompanied by a cloud of asymptotic photons propagating along the light cone. Remark. We point out that, in our definition of scattering states, we can directly accommodate an arbitrarily large number of “hard” photons without energy restriction, i.e., we can construct the limiting vector out Ft(1) , t]ψh,κ (t) Ft(m) , t] . . . A[ Aout [ F (m) ] . . . Aout [ F (1) ]ψh,κ := s − lim A[ t→+∞
(III.76) out plus m asymptotic photons with wave functions which represents the state ψh,κ F (m) , . . . , F (1) , respectively. Analogously, we define
) (m ) ] . . . Ain [G (1) ]ψ in := s − lim A[ G (1) G (m Ain [G , t] . . . A[ t t , t]ψh,κ (t). h,κ t→−∞
(III.77) This is possible because, apart from some higher order estimates to control the commutator i[H, x ], and the photon creation operators in (III.76)–(III.77) (see for example [18]), we use the propagation estimate (III.44), which only limits the asymptotic velocity of the electron. This fact is very important for estimating scattering amplitudes involving an arbitrary number of “hard” photons. In particular, for any m, m ∈ N, we can define the S-matrix element j } ) = Aout [ F (m) ] . . . Aout [ F (1) ]ψ out Sαm,m ( { Fi }, {G h out ,κ out ,
(m ) ] . . . Ain [G (1) ]ψ inin in Ain [G (III.78) h ,κ which corresponds to the transition amplitude between two states describing an incoming electron with wave function h in , accompanied by a soft photon cloud of free photons (1) , . . . , G (m ) ), of energy smaller than κ in , plus m hard photons (with wave functions G and an outgoing electron with wave function h out and soft photon energy threshold κ out , plus m hard photons, respectively. j } ) in the finestructure constant α can be carried The expansion of Sαm,m ( { Fi }, {G out, at least to leading order, along the lines of [1]. This yields a rigorous proof of the
Infraparticle States in QED I
785
transition amplitudes for Compton scattering in leading order, and in the non-relativistic approximation, that one can find in textbooks. Moreover, as expected from classical electromagnetism, “close” to the electron a Liénard-Wiechert electromagnetic field is observed. The precise mathematical statement is out/in i Ht 2 e−i H t ψ out/in d 3 y Fµν (0, y ) δ˜ (y − x ψ lim|d|→∞ lim | d| , e − d) t→±∞ h, F h, F E ∇ ψ σt , ψ σt |h( P)| 2 d 3 P = 0, − Fµν P (0, d) (III.79) P P where δ˜ is a smooth, -dependent delta function which has the property that its Fourier transform is supported in B , x is the electron position, Fµν = ∂µ Aν − ∂ν Aµ
(III.80)
with (2π )2 α 1/2 (III.81) | y − x |
d 3k y ∗ y i −i k· ∗ i i k· (k,λ , (III.82) ak,λ k,λ ak,λ Ai (0, y) := − ) e + ( ) e λ |k|
A0 (0, y) := −
E ∇
and Fµν P is the electromagnetic field tensor corresponding to a Liénard-Wiechert solution for the current
1 1 E t), 2(2π )3 α 2 ∇ E t) , |∇ E | < 1; E δ (3) ( Jµ (t, y) := −2(2π )3 α 2 δ (3) ( y −∇ y −∇ P P P P (III.83) see the discussion in Sect. I. IV. Uniform Boundedness of the Limiting Norm Our first aim is to prove the uniform boundedness of ψh,κ (t), as t → ∞; more precisely, that 2 d 3 P. lim ψh,κ (t), ψh,κ (t) = |h( P)| (IV.1) t→∞
The sum of the diagonal – with respect to the partition G (t) introduced above – is
terms 2 easily seen to yield |h( P)| d 3 P in the limit t → ∞, as one expects. Thus, our main task is to show that the sum of the off-diagonal terms vanishes in this limit. In Sect. V, we prove that the norm-bounded sequence {ψh,κ (t)} is, in fact, Cauchy. We recall that the definition of the vector ψh,κ (t) involves three different rates: • The rate t −β related to the fast infrared cut-off σt ; • the rate t −θ of the slow infrared cut-off σtS (see (III.27)); • the rate t − of refinement of the cell partitions G (t) .
786
T. Chen, J. Fröhlich, A. Pizzo
IV.1. Control of the off-diagonal terms. We denote the off-diagonal term labeled by the pair (l, j) of cell indices l = j contributing to the l.h.s. of (IV.1) by σ σ σ σ iγ ( v ,∇ E t ,t) −i E t t (t) iγ ( v ,∇ E t ,t) −i E t t (t) P P e P ψl,σt , Wκ,σt ,l, j (t)e σt j e P ψ j,σt , Ml, j (t) := e σt l (IV.2) where we use the notation Wκ,σt ,l, j (t) := exp λ
Bκ \Bσt
· ε a ∗ e−i|k|t − ε ∗ a ei|k|t d 3 k , ηl, j (k) k,λ k,λ k,λ k,λ (IV.3)
E σt ( P ∗ ), P ∗ being the center of the cell G (t) , and where v j ≡ ∇ j j j := α 2 ηl, j (k)
v j
1
vl
1
(1 − |k| k · v j ) 3 2
− α2
(1 − |k| k · vl ) 3 2
.
(IV.4)
We study the limit t → +∞; the case t → −∞ is analogous. A. Asymptotic orthogonality. In order to prove that the off-diagonal terms (IV.2) vanish in the limit t → +∞, we separate the role played by the time variable t as the parameter determining the dynamical cell decomposition and infrared cutoffs, from its usual role as the conjugate variable to the energy. For the latter, we introduce an auxiliary variable s ≥ t. Then, for fixed t (such that the cell decomposition and the cutoffs are constant), µ we interpret the terms (IV.2) as special values Ml, j (t) = Ml,1 j (t, t) of families Ml, j (t, s) introduced below, which depend on t, s, and an additional auxiliary parameter µ ∈ R. µ Our strategy will be based on proving that the dispersive properties of Ml, j (t, s) as a function of s ≥ t alone, for fixed t and µ, imply that Ml, j (t) has a sufficiently fast decay in t such that our desired result of asymptotic orthogonality follows. More precisely, we introduce a family of operators µ ∗ −i|k|s ∗ i|k|s 3 W d k (s) := exp µ ηl, j (k) · ε a e − ε a e κ,σt ,l, j
k,λ k,λ
Bκ \Bσt
λ
k,λ k,λ
(IV.5) depending on a parameter µ ∈ R, and define σt σt σ σ iγσt ( v j ,∇ E t ,s) −i E t s (t) µ (t, s) := eiγσt (vl ,∇ E P ,s) e−i E P s ψ (t) , W µ P P ψ M (s)e e l,σt j,σt l, j κ,σt ,l, j (IV.6) µ=1 (t, t) = Ml, j (t). for s ≥ t ( 1). Obviously, M l, j The phase factor γσt ( v j , ∇ E σt , s) is chosen as follows: P
γσt ( v j , ∇ E σt , s) := −α P
1
s
E σt · ∇ P
Bσ S \Bσt
cos(k · ∇ )d 3 kdτ E σt τ − |k|τ v (k) j P
τ
(IV.7)
Infraparticle States in QED I
787
for s −θ ≥ σt , and γσt ( v j , ∇ E σt , s) P
− θ1
σt
:= −α
E σt · ∇ P
1
Bσ S \Bσt
cos(k · ∇ )d 3 kdτ, E σt τ − |k|τ v (k) j P
τ
(IV.8) s −θ
for < σt . As a function of µ, the scalar product in (IV.6) satisfies the ordinary differential equation µ
(t, s) dM l, j dµ
µ (t, s) + rσµ (t, s), = − µ Cl, j,σt M l, j t
where
Cl, j,σt :=
and
Bκ \Bσt
(IV.9)
2 d 3 k, | ηl, j (k)|
(IV.10)
E σt ,s) −i E σt s (t) iγ ( v ,∇ P rσµt (t, s) := − e σt l e P ψl,σt , µ
ηl, j )(s) e Wκ,σt ,l, j (s) aσt (
σ P
E σt ,s) −i E σt s (t) iγ ( v ,∇ P + aσt ( ηl, j )(s)e σt l e P ψl,σt , µ
Wκ,σt ,l, j (s)e with aσt ( ηl, j )(s) :=
E σt ,s) −i E σt s iγσt ( v j ,∇ P
λ
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
P
e
The solution of the ODE (IV.9) is given by C µ − l,2j,σt µ2 0 Ml, j (t, s) + Ml, j (t, s) = e
0
µ
rσµt (t, s) e−
Cl, j,σt 2
(t) ψ j,σt ,
· ε ∗ a ei|k|s d 3 k. ηl, j (k) k,λ k,λ
Bκ \Bσt
(t)
ψ j,σt
(µ2 −µ2 )
(IV.11)
dµ ,
(IV.12)
where the initial condition at µ = 0 is given by l,0 j (t, s) = 0, M
(IV.13)
(t) (t) since the supports in P of the two vectors ψl,σt , ψ j,σt are disjoint (up to sets of measure 0), for arbitrary t and s. (t) (t) Furthermore, condition (I 4) in Theorem III.1 implies that the vectors ψl,σt , ψ j,σt are vacua for the annihilation part of the asymptotic photon field1 under the dynamics generated by the Hamiltonian H σt . As a consequence, we find that
lim r µ (t, s) s→+∞ σt
= 0,
(IV.14)
1 The existence of the asymptotic field operator for a fixed cut-off dynamics is derived as explained in part B. of this section.
788
T. Chen, J. Fröhlich, A. Pizzo
for fixed µ and t. To arrive at this conclusion, the following is used: The one-particle E σt ,s) iγ ( v ,∇
P state, multiplied by the phase e σt j continues to be a one-particle state for the σ t Hamiltonian H ; for large s (see (IV.8)) the phase is s-independent; the operator E σt P coincides with the operator H σt when applied to one-particle states of the Hamiltonian H σt . Therefore, by dominated convergence, it follows that
l,1 j (t, s) = 0. lim M
s→+∞
(IV.15)
1 (t, t) ≡ Ml, j (t), we have Since M l, j |Ml, j (t)| l,1 j (t, +∞)| = |Ml, j (t) − M +∞
σ σ d i H σt s iγσt ( v j ,∇ E t ,s) −i E t s (t) (t) P P e Wκ,σt ( v j , s)e e ψ j,σt ds . ≤ 2 sup(ψl,σt ) sup ds t l j (IV.16) To estimate the r.h.s. of (IV.16), we proceed as follows. Since we are interested in the limit t → +∞, and the integration domain on the r.h.s. of (IV.16) is [t, +∞), our aim is to show that
σ σ d i H σt s iγ ( v ,∇ E t ,s) −i E t s (t) P e Wκ,σt ( v j , s)e σt j e P ψ j,σt ds
(IV.17)
is integrable in s, and that the rate at which the r.h.s. of (VI.16) converges to zero offsets the growth of the number of cells in the partition. This allows us to conclude that N (t)
Ml, j (t) −→ 0
(IV.18)
l, j (l= j)
in the limit t → +∞, and, as a corollary, lim
t→+∞
N (t)
Ml, j (t) =
2 d 3 P, |h( P)|
(IV.19)
l, j
as asserted in Theorem III.2. The convergence (IV.18) follows from the following theorem. Theorem IV.1. The off-diagonal terms Ml, j (t), l = j, satisfy | Ml, j (t) | ≤ C
1 | ln σt |2 t −3 , tη
(IV.20)
for some constants C < ∞ and η > 0, both independent of l, j, and > 0. In particular, Ml, j (t) → 0, as t → +∞.
Infraparticle States in QED I
789
As a corollary, we find
| Ml, j (t) | ≤ C N 2 (t)
1≤l= j≤N (t)
1 1 | ln σt |2 t −3 ≤ C η | ln σt |2 t 3 , tη t
(IV.21)
since N (t) ≈ t 3 . (Throughout the paper, C, C , c, and c denote positive constants.) We conclude that, for < η4 , (IV.18) follows. B. Time derivative and infrared tail. We now proceed to prove Theorem IV.1. The arguments developed here will also be relevant for the proof of the (strong) convergence of the vectors ψh,κ (t), as t → +∞, which we discuss in Sect. V. To control (IV.16), we focus on the derivative
d i H σt s E σt ,s) −i E σt s (t) iγ ( v ,∇ P e (IV.22) Wκ,σt ( v j , s) e σt j e P ψ j,σt ds = i ei H
σt s
+ i ei H
[H Iσt , Wκ,σt ( v j , s)] e
σt s
Wκ,σt ( v j , s)
E σt ,s) −i E σt s iγσt ( v j ,∇
e
P
ψ (t) j,σt
P
(IV.23)
E σt , s) vj, ∇ dγσt ( E σt ,s) −i E σt s iγσt ( v j ,∇ P e
ds
P
e
P
(t)
ψ j,σt ,
(IV.24)
where 1 x) · Aσt ( x) Aσt ( . x) + α H Iσt := α 2 p · Aσt ( 2
(IV.25)
We have used that Wκ,σt ( v j , s) = e−is H0 Wκ,σt ( v j , 0)eis H0 , where H0 := H σt − H Iσt is the free Hamiltonian, to obtain the commutator in (IV.23). We rewrite the latter in the form " ! [H Iσt , Wκ,σt ( v j , s)] = Wκ,σt ( v j , s) Wκ,σt ( v j , s)∗ H Iσt Wκ,σt ( v j , s) − H Iσt , (IV.26) and use that Wκ,σt ( v j , s) Aσt ( x)Wκ,σt ( v j , s) = Aσt ( x) + α 2
1
∗
Bκ \Bσt
3 cos(k · x − |k|s)d v (k) k j
(IV.27) is defined in (III.28). We can then write the term (IV.23) as (see (III.8)), where vl j (k) (IV.23) = i e
i H σt s
×e
v (k) d 3k j
Wκ,σt ( v j , s)α i[H , x ] ·
e × cos(k · x − |k|s) +ie
σt
i H σt s
Bκ \Bσt σ E σt ,s) −i E t s iγσt ( v j ,∇ P
α2 Wκ,σt ( v j , s) 2
E σt ,s) −i E σt s iγσt ( v j ,∇ P
e
P
e
Bκ \Bσt
P
(t)
ψ j,σt
3 cos(k · x − |k|s)d v (k) k j
(t)
ψ j,σt ,
where we recall that i[H σt , x ] = p + α 2 Aσt ( x); see (III.41). 1
(IV.28)
2 (IV.29)
790
T. Chen, J. Fröhlich, A. Pizzo
From the decay estimates provided by Lemma A.2 in the Appendix one concludes that the norm of (IV.29) is integrable in s, and that +∞ 3 1 ds (IV.29) ≤ η | ln σt |2 t − 2 (t 1) (IV.30) t t for some η > 0 independent of . The analysis of (IV.28) is more involved. Our argument will eventually involve the derivative of the phase factor in (IV.24). To begin with, we write (IV.28) as (IV.28) = i ei H ×(H σt + i)
σt s
Wκ,σt ( v j , s)α i[H σt , x ] ·
1 H σt
Bκ \Bσt
+i
cos(k · x − |k|s) e v (k) d k j 3
σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
(t)
ψ j,σt . (IV.31)
Pulling the operator (H σt + i) through to the right, the vector (IV.31) splits into the sum of a term involving the commutator [H σt , x ], 1 σt s v (k) i ei H Wκ,σt ( v j , s)α i[H σt , x ] · σ d 3k j H t + i Bκ \Bσt ]e × [ H σt , cos(k · x − |k|s) and i ei H
σt s
×
Wκ,σt ( v j , s)α i[H σt , x ] · 1
H σt + i
cos(k · x − |k|s)e
E σt ,s) −i E σt s iγσt ( v j ,∇
Bκ \Bσt
e
P
P
(t)
ψ j,σt ,
(IV.32)
v (k) d 3k j σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
(t)
(E σt + i)ψ j,σt . P
(IV.33)
Note that [H σt , x ] H σ1t +i is bounded in the operator norm, uniformly in σt . To control (IV.32) and (IV.33), we invoke a propagation estimate for the electron position operator as follows. Due to condition (I 3) in Theorem III.1, we can introduce a C ∞ −function χh ( y ), y ∈ R3 , such that • χh ( y ) = 1 for νmin ≤ | y | ≤ νmax . • χh ( y ) = 0 for | y | ≤ 21 νmin and | y| ≥
1+νmax 2
.
It is shown in Theorem A.3 of the Appendix that, for θ < 1 sufficiently close to 1 and s large, the propagation estimate σt σt σt σt χh x eiγσt (v j ,∇ E P ,s) e−i E P s ψ (t) − χh (∇ E σt )eiγσt (v j ,∇ E P ,s) e−i E P s ψ (t) j,σt j,σt P s 1 1 ≤ ν 3 | ln(σt )|, (IV.34) s t2 Eσ holds, where ν > 0 is independent of . The argument uses the Hölder regularity of ∇ P and and of σ listed under properties (I 2) in Theorem III.1, differentiability of h( P), P (III.18).
Infraparticle States in QED I
791
We continue with the discussion of the expressions (IV.32) and (IV.33). We split (IV.33) into two parts: i ei H + i ei H
σt s σt s
Wκ,σt ( v j , s) J |κσ S (s) e s σsS σt
Wκ,σt ( v j , s) J | (s) e
σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
E σt ,s) iγσt ( v j ,∇ P
e
σ −i E t s P
(E σt + i) ψ (t) j,σt P
(t)
(E σt + i) ψ j,σt , P
(IV.35) (IV.36)
using the definitions 1
3 cos(k · x − |k|s)d v (k) k if s −θ ≥ σt j H σt + i Bκ \B S σs 1 v (k) := αi[H σt , x ] · d 3k cos(k · x − |k|s) if s −θ < σt , j σt + i H Bκ \Bσt (IV.37)
J |κσ S (s) := αi[H σt , x ] · s
and σS J |σts (s)
1
σt
:= α i[H , x ] ·
H σt + i B S \Bσt σ
3 cos(k · x − |k|s)d v (k) k if s −θ ≥ σt j
s
:= 0 if s
−θ
< σt ,
(IV.38)
where we refer to σsS := s −θ as the slow infrared cut-off. (We consider s, t large enough such that κ > σsS , σt .) σS
To control J |σts (s) in (IV.38), we define the “infrared tail” σ
σ
t t v j , xs , s) d γσt ( d(ei H s x h (s)e−i H s ) i H σt s 1 σt := α e−i H s σ e · ds H t +i ds cos(k · ∇ d 3 k if s −θ ≥ σt , E σt s − |k|s) v (k) · j
P
Bσ S \Bσt s
:= 0 if s −θ < σt ,
(IV.39)
where x h (s) := x χh ( xs ). Summarizing, we can write (IV.22) as (IV.22) = (IV.29) + i ei H + i ei H + i ei H + i ei H ×
σt s σt s
σt s
Wκ,σt ( v j , s) J |κσ S (s) e s σsS σt
Wκ,σt ( v j , s) J | (s) e Wκ,σt ( v j , s)
σt s
Bκ \Bσt
σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇ E σt ,s) iγσt ( v j ,∇ P
E σt , s) vj, ∇ dγσt ( P
ds
Wκ,σt ( v j , s) α i [H σt , x ] ·
e
e e
σ −i E t s P σ P
(t)
(E σt + i) ψ j,σt P
(t)
(E σt + i) ψ j,σt P
σ P
iγσt ( v j ,∇ E t ,s) −i E t s
e
ψ (t) j,σt
1 H σt + i
(IV.40) (IV.41) (IV.42) (IV.43)
σt
σt
E ,s) −i E s (t) iγ ( v ,∇ [H σt , cos(k · x − |k|s)] v (k) P d 3k e σt j e P ψ j,σt , j
792
T. Chen, J. Fröhlich, A. Pizzo
where we recall that (IV.29) satisfies (IV.30). We claim that +∞ ≤ 1 | ln σt |2 t −3/2 , (t 1) [(IV.40) + (IV.41) + (IV.42)] ds tη t for some η > 0 depending on θ , but independent of . This is obtained from +∞ [(IV.40) + (IV.41) + (IV.42)] ds t +∞ 1 cos(k · x − |k|s) v (k) ≤ ds α i[H σt , x ] · σ d 3k j H t + i Bκ \Bσt t $ # σt σt x iγ ( v , ∇ E ,s) −i E s (t) σ σ σ j E t ) − χh ( ) e t P × χ h (∇ e P (E t + i) ψ j,σt P P s +∞ x iγ (v ,∇ E σt ,s) −i E σt s σt P + ds J |κσ S (s)χh ( )e σt j e P (E + i) ψ (t) j,σt P s s t % & +∞ d γσt ( v j , xs , s) x σsS i H σt s + ds e Wκ,σt ( v j , s) J |σt (s)χh ( ) − t s ds E σt ,s) −i E σt s iγ ( v ,∇ (t) P e P (E σt + i) ψ j,σt × e σt j P % & +∞ E σt , s) d vj, ∇ dγσt ( γσt ( v j , xs , s) i H σt s P + − ds e Wκ,σt ( v j , s) t ds ds E σt ,s) −i E σt s iγ ( v ,∇ (t) P e P (E σt + i) ψ j,σt × e σt j P
(IV.44)
(IV.45)
(IV.46) (IV.47) (IV.48)
(IV.49)
using the following arguments: • The term (IV.46) can be bounded from above by t1η | ln σt |2 t −3/2 , for some η > 0 independent of , due to the propagation estimate for (III.44) and Lemma A.2, which show that the integrand has a sufficiently strong decay in s. • In (IV.47), the slow cut-off σsS and the function χh ( xs ) make the norm integrable in s with the desired rate (i.e., to get a bound as in (IV.30)), for a suitable choice of θ < 1. In particular, we can exploit that ' ' θ ' v j ) cos(k · x − |k|s) χh ( x ) '' ≤ O( s ), (IV.50) k, sup ' d 3 k ( s s2 Bκ \B S x ∈R3 σs
see Lemma A.2 in the Appendix. • In (IV.48), only terms integrable in s and decaying fast enough to satisfy the bound (IV.30) are left after subtracting d γσt ( v j , xs , s) ds σS
(IV.51)
from J |σts (s). This is explained in detail in the proof of Theorem A.4 in the Appendix.
Infraparticle States in QED I
793
• To bound (IV.49), we use the electron propagation estimate, combined with an integration by parts, to show that the derivative of the phase factor tends to the “infrared tail” for large s, at an integrable rate that provides a bound as in (IV.30). We note that due to the vector interaction in non-relativistic QED, this argument is more complicated here than in the Nelson model treated in [26] where the interaction term is scalar. Here, we have to show (see Theorem A.4) that, in the integral with respect to σt σt s, the pointwise velocity ei H s i[ x, H σt ]e−i H s can be replaced by the (asymptotic) E σt at asymptotic times. mean velocity ∇ P
Finally, to control (IV.43), we observe that the commutator introduces additional decay in s into the integrand when xs is restricted to the support of χh . It then follows that the propagation estimate suffices (without infrared tail) to control the norm, by the same arguments that were used to estimate (IV.46), (IV.47). Combining the above arguments, the proof of Theorem IV.1 is completed. V. Proof of Convergence of ψh,κ (t) In this section, we prove that ψh,κ (t) defines a bounded Cauchy sequence in H, as t → +∞. To this end, it is necessary to control the norm difference between vectors ψh,κ (ti ), i = 1, 2, at times t2 > t1 1.
V.1. Three key steps. As anticipated in Sect. III.4, we decompose the difference of ψh,κ (t1 ) and ψh,κ (t2 ) into three terms ψh,κ (t1 ) − ψh,κ (t2 ) = ψ(t2 , σt2 , G (t2 ) → G (t1 ) ) + ψ(t2 → t1 , σt2 , G (t1 ) ) + ψ(t1 , σt2 → σt1 , G (t1 ) ),
(V.1)
where we recall from (III.49) – (III.51): I) The term ψ(t2 , σt2 , G (t2 ) → G (t1 ) ) = e i H t2
N (t1 )
Wκ,σt2 ( v j , t2 ) e
σt
σt
E 2 ,t2 ) −i E 2 t2 iγσt ( v j ,∇ 2
P
e
P
(t )
ψ j,σ1 t
j=1
− e i H t2
N (t1 )
Wκ,σt2 ( vl( j) , t2 )e
σ
2 σ
E t2 ,t2 ) −i E t2 t2 iγσt ( vl( j) ,∇ 2
P
e
P
(t )
ψl( 2j),σt , (V.2) 2
j=1 l( j)
accounts for the change of the partition G (t2 ) → G (t1 ) in ψh,κ (t2 ), where l( j) (t ) (t ) labels the sub-cells belonging to the sub-partition G (t2 ) ∩ G j 1 of G j 1 , and where we define E σt∗2 vl( j) ≡ ∇
Pl( j)
and
E σt∗1 ; v j ≡ ∇ Pj
794
T. Chen, J. Fröhlich, A. Pizzo
II) the term ψ(t2 → t1 , σt2 , G (t1 ) ) = e i H t1
N (t1 )
Wκ,σt2 ( v j , t1 ) e
σt
σt
E 2 ,t1 ) −i E 2 t1 iγσt ( v j ,∇ 2
e
P
P
(t )
ψ j,σ1 t
j=1
− e i H t2
N (t1 )
Wκ,σt2 ( v j , t2 ) e
σ
σ
E t2 ,t2 ) −i E t2 t2 iγσt ( v j ,∇ 2
P
e
P
2
(t )
ψ j,σ1 t , 2
j=1
(V.3)
accounts for the subsequent change of the time variable, t2 → t1 , for the fixed partition G (t1 ) , and the fixed infrared cut-off σt2 ; and finally, III) the term ψ(t1 , σt2 → σt1 , G (t1 ) ) = e i H t1
N (t1 )
Wκ,σt1 ( v j , t1 )e
σt
σt
E 1 ,t1 ) −i E 1 t1 iγσt ( v j ,∇ 1
P
e
P
(t )
ψ j,σ1 t
j=1
− e i H t1
N (t1 )
Wκ,σt2 ( v j , t1 )e
σt
σt
E 2 ,t1 ) −i E 2 t1 iγσt ( v j ,∇ 2
P
e
P
1
(t )
ψ j,σ1 t
j=1
2
(V.4)
accounts for the change of the infrared cut-off, σt2 → σt1 . Our goal is to prove
ρ ψh,κ (t2 ) − ψh,κ (t1 ) ≤ O (ln(t2 ))2 /t1 ,
(V.5)
for some ρ > 0. To this end, it is necessary to perform the three steps in the order displayed above. As a corollary of the bound (V.5), we obtain Theorem III.2 by telescoping (see the comment after Eq. (III.57)). The arguments in our proof are very similar to those in [26], but a number of modifications are necessary because of the vector nature of the QED interaction. For these modifications, we provide detailed explanations. V.2. Refining the cell partition. In this section, we discuss step (V.2) in which the momentum space cell partition is modified. It is possible to apply the methods developed in [26], up to some minor modifications. We will prove that
ψ(t2 , σt , G (t2 ) → G (t1 ) ) ≤ O (ln(t2 ))2 /t ρ (V.6) 2 1 for some ρ > 0. The contributions from the off-diagonal terms with respect to the subpartition G (t2 ) of G (t1 ) can be estimated by the same arguments that have culminated in (t ) the proof of Theorem IV.1. That is, we first express ψ j,σ1 t as (t )
ψ j,σ1 t = 2
(t Gj 1
σt
2 3 h( P) d P = )
P
l( j)
2
σt
2 3 h( P) d P = )
(t2 Gl( j)
P
l( j)
(t )
ψl( 2j),σt . 2
(V.7)
Infraparticle States in QED I
795
Then, ψ(t2 , σt 2 , G (t2 ) → G (t1 ) )2 N (t1 )
=
(
σt ) −i E 2 t2 (t2 ) κ,σt ( P W v , t ) − W ( v , t ) e ψl( j),σt , κ,σt2 j 2 l( j) 2 2 2
j, j =1 l( j),l ( j )
( ) σt −i E 2 t2 (t2 ) κ,σt ( κ,σt ( P ( j ) , t2 ) − W , t2 ) e W , v v ψ ( j ),σ l j l 2 2 t
(V.8)
2
where we define σt2
E iγ ( v ,∇ κ,σt ( P W v j , t2 ) := Wκ,σt2 ( v j , t2 ) e σ t 2 j 2
κ,σt ( W vl( j) , t2 ) := Wκ,σt2 ( vl( j) , t2 ) e 2
,t2 )
,
σ E t2 ,t2 ) iγσt ( v ,∇ 2 l( j) P
(V.9) .
(V.10)
Following the analysis in Sect. IV.1, one finds that the sum over pairs (l ( j ), l( j)) with either l = l or j = j can be bounded by O(t2− ), provided that < η4 , as in (IV.21). * . Then, we are Let * stand for the expectation value with respect to the vector
left with the diagonal terms N (t1 )
(
∗ ∗ κ,σ κ,σt ( κ,σ κ,σt ( 2−W ( vl( j) , t2 ) W v j , t2 ) − W ( v j , t 2 )W vl( j) , t2 ) t t 2 2 2
2
j=1 l( j)
) *
,
(V.11) σt −i E 2 t2 P
(t )
*≡e labeled by pairs (l( j), l( j)), where ψl( 2j),σt in the case considered here. 2 For each term ∗ κ,σ κ,σt ( W ( vl( j) , t2 )W v j , t2 ) , (V.12) t 2 *
2
we can again invoke the arguments developed for off-diagonal elements indexed by (l, j) (where l = j) from Sect. IV.1. In particular, we define for s > t2 , σ −i E t s (t2 ) ∗ µ v j , s) µ P ψ M (t , s) := e 2 l( j),σt2 , Wκ,σt2 ( vl( j) ] [l( j), v j ],[l( j), σ −i E t s (t ) µ κ,σ (V.13) W ( vl( j) , s) e P ψl( 2j),σt , t 2
2
where σt2
E −iγ ( v ,∇ µ∗ µ κ,σ κ,σ P W ( v j , s) W ( vl( j) , s) = e σt2 j t t 2
,s)
2
µ∗ Wκ,σ ( v j , s) t 2
µ ×Wκ,σ ( vl( j) , s) e t 2
and µ∗ µ ( v j , s) Wκ,σ ( vl( j) , s) Wκ,σ t2 t2
= exp
µ
λ
Bκ \Bσt
∗ i|k|s −εk,λ e ak,λ
with µ a real parameter.
σt
E 2 ,s) iγσt ( vl( j) ,∇ 2
P
,
(V.14)
· ε a ∗ e−i|k|s η j,l( j) (k) k,λ k,λ 2
3
d k
,
(V.15)
796
T. Chen, J. Fröhlich, A. Pizzo
Proceeding similarly as in (IV.12), the solution of the ODE analogous to (IV.9) for µ M vl( j) ] (t2 , s) at µ = 1 consists of a contribution at µ = 0, which remains [l( j), v j ],[l( j), non-zero as s → ∞, and a remainder term that vanishes in the limit s → ∞. In fact, 1 (t2 , s) lim M s→+∞
=e
−
vl( j) ] [l( j), v j ],[l( j),
C j,l( j) σt2 2
σt
e
−1
E 2 ,σt θ ) iγσt ( v j ,∇ 2 2
P
where
) ψl((t2j),σ , t2
C j,l( j),σt2 :=
Bκ \Bσt
σt
e
−1
E 2 ,σt θ ) iγσt ( vl( j) ,∇ 2 2
P
) , (V.16) ψl((t2j),σ t 2
2 d 3 k, | η j,l( j) (k)|
(V.17)
2
defined in (IV.4). Hence, (V.11) is given by the sum of as in (IV.10), with η j,l( j) (k) −
N (t2 ) +∞ j=1 ( j) t2
−
N (t2 ) +∞ j=1 ( j) t2
d 1 (t2 , s) ds M ds [ ( j),v j ],[ ( j),v ( j) ] d 1 (t2 , s) ds M ds [ ( j),v ( j) ],[ ( j),v j ]
(V.18)
and #
Cl( j), j,σt2 $ σt2 ) (t2 ) − 2 ψ ψl((t2j),σ , 2 − 2 cos γ ( v − v , ∇ E , t ) e σ j 2 l( j) t2 l( j),σt , t
N (t2 ) j=1 ( j)
P
2
2
(V.19) where E σt2 , t2 ) := γσt ( E σt2 , σt− θ ) − γσt ( E σt2 , σt− θ ). γσt2 ( v j − vl( j) , ∇ vl( j) , ∇ vj, ∇ 2 2 2 2 1
P
1
P
P
(V.20) The arguments that have culminated in Theorem IV.1 also imply that the sum (V.18) can be bounded by O(t2−4 ), for η > 4. The leading contribution in (V.6) is represented by the sum (V.19) of diagonal terms (with respect to G (t2 ) ), which can now be bounded from above. It suffices to show that '
Cl( j), j,σt '' ' E σt2 , t2 ) e− 2 2 ' ≤ 1 ln t2 (V.21) sup '' 2 − 2 cos γσt2 ( v j − vl( j) , ∇ ' η P S t1 P∈ for some η > 0 that depends on . To see this, we note that the lower integration bound in the integral (V.17) contributes a factor to (V.21) proportional to ln t2 . In Lemma A.1, it is proven that ' ' 1 1 ' E σt2 , (σt2 )− θ ) − γσt ( E σt2 , (σt2 )− θ ) '' ≤ O(| v j − vl( j) |). vj, ∇ v , ∇ ' γσt2 ( l( j) 2 P
P
(V.22) We can estimate the difference v j − vl( j) =
E σt∗1 ∇ P j
−
E σt∗2 ∇ P
l( j)
, which also appears in
using condition (I 2) of Theorem III.1. This yields the -dependent negative ηl( j), j (k), power of t1 in (V.21).
Infraparticle States in QED I
797
V.3. Shifting the time variable for a fixed cell partition and infrared cut-off. In this subsection, we prove that
ρ (V.23) ψ(t2 → t1 , σt2 , G (t1 ) ) ≤ O (ln(t2 ))2 /t1 for some ρ > 0; see (V.3). This accounts for the change of the time variable, while both the cell partition and the infrared cutoff are kept fixed. It can be controlled by a standard Cook argument, and methods similar to those used in the discussion of (IV.22). For t1 ≤ s ≤ t2 , we define s σt2 σt2 cos(k · ∇ ) d 3 k dτ. E σt2 τ − |k|τ v (k) v j , ∇ E , s) := − ∇E · γσt2 ( j P
P
1
Bσ S \Bσt τ
P
2
(V.24) Then, we estimate t2 σ σ σt2 d E t2 ,s) −i E t2 s (t1 ) iγ ( v ,∇ P ei H s Wκ,σt2 ( ds v j , s)e σt2 j e P ψ j,σt 2 ds t1
(V.25)
cell by cell. To this end, we can essentially apply the same arguments that entered the treatment of the time derivative in (IV.22), see also the remark after Theorem A.3, by defining an infrared tail in a similar fashion. The only modification to be added is that, apart from two terms analogous to (IV.23), (IV.24), we now also have to consider i ei H s (H − H σt2 ) Wκ,σt2 ( v j , s)e
σ
σ
E t2 ,s) −i E t2 s iγσt ( v j ,∇ 2
e
P
P
(t )
ψ j,σ1 t , 2
(V.26)
which enters from the derivative in s of the operator underlined in ei H s e−i H
σt2
s i H σt2 s
e
Wκ,σt2 ( v j , s)e
σt
σt
E 2 ,s) −i E 2 s iγσt ( v j ,∇ 2
P
e
P
(t )
ψ j,σ1 t . 2
(V.27)
To control the norm of (V.26), we observe that H − H σt2 = α 2 i[H, x ] · A<σt2 − α 1
where A<σt2 :=
λ=± Bσt2
A<σt2 · A<σt2 2
,
d 3k ∗ ∗ εk,λ , k,λ bk,λ + ε bk,λ |k|
(V.28)
(V.29)
and we note that [H, x ] · A<σt2 = A<σt2 · [H, x ], because of the Coulomb gauge condition. Moreover, ∗ Wκ,σ ( v j , s) i[H, x ] Wσt2 ( v j , s) = i[H, x ] + hs ( x) t 2
x) < O(1) and with hs ( x)] = [hs ( [bk,λ x), A<σt2 ] = 0. , h s (
(V.30)
798
T. Chen, J. Fröhlich, A. Pizzo
Furthermore, we have (t )
1 bk,λ ψ j,σt = 0 2
for k ∈ Bσt2 , and A<σt2 ψ (tj,σ1 )t , A<σt2 · A<σt2 ψ (tj,σ1 )t ≤ O(σt2 ψ (tj,σ1 )t2 ).
(V.31)
(t ) (t ) (t ) A<σt2 · [H, x ] ψ j,σ1 t ≤ O σt2 [H, x ] ψ j,σ1 t + ψ j,σ1 t ,
(V.32)
2
2
The estimate
2
2
2
holds, where − 3 2
(t )
[H, x ]ψ j,σ1 t ≤ O(t1 2
),
(V.33)
because (t )
(t )
[H, x ]ψ j,σ1 t ≤ c1 (H σt2 + i)ψ j,σ1 t 2
2
for some constant c1 , and − 3 2
ψ (tj,σ1 )t = O(t1 2
).
Consequently, we obtain that σ σ E t2 ,s) −i E t2 s (t1 ) iγ ( v ,∇ P v j , s)e σt2 j e P ψ j,σt (H − H σt2 ) Wκ,σt2 ( 2 (t1 ) (t1 ) −3/2 ≤ O( σt2 [H, x ] ψ j,σt + ψ j,σt ) ≤ O( σt2 t ). 2
2
(V.34)
Following the procedure in Sect. II.2.1, B., one can also check that t2 σ σ E t2 ,s) −i E t2 s (t1 ) iγσt ( v j ,∇ i H s −i H σt2 s d i H σt2 s 2 P P e e e Wκ,σt2 ( v j , s)e e ψ j,σt ds 2 ds t1 3 1 2 −2 ≤ O , (V.35) η | ln σt2 | t1 t1 for some η > 0 independent of . Similarly as in (IV.21), we choose small enough such that η4 > . The number of cells in the partition G (t1 ) is N (t1 ) ≈ t13 . Therefore, summing over all cells, we get 1 − 3 − 3 (V.36) O N (t1 ) t1 2 σt2 t2 + O N (t1 ) η | ln σt2 |2 t1 2 , t1 as an upper bound on the norm of the term in (V.3). −β The parameter β in the definition of σt2 = t2 can be chosen arbitrarily large, independently of . Hereby, we arrive at the upper bound claimed in (V.5).
Infraparticle States in QED I
799
V.4. Shifting the infrared cut-off. In this section, we prove that
ρ ψ(t1 , σt2 → σt1 , G (t1 ) ) ≤ O (ln(t2 ))2 /t1
(V.37)
for some ρ > 0; see (V.4). The analysis of this last step is the most involved one, and will require extensive use of our previous results. The starting idea is to rewrite the last term in (V.4), e
i H t1
N (t1 )
Wκ,σt2 ( v j , t1 ) e
σt
σt
E 2 ,t1 ) −i E 2 t1 iγσt ( v j ,∇ 2
e
P
P
(t )
ψ j,σ1 t ,
(V.38)
2
j=1
as e i H t1
N (t1 )
σt2
E σt2 ) Wσt (∇ E σt2 ) eiγσt2 (v j ,∇ E P Wκ,σt2 ( v j , t1 ) Wσ∗t (∇ 2 P
2
σt
,t1 ) −i E 2 t1
e
P
P
(t )
ψ j,σ1 t , 2
j=1
(V.39) and to group the terms appearing in (V.39) in such a way that, cell by cell, we consider the new dressing operator σt
E 2)e ei H t1 Wκ,σt2 ( v j , t1 ) Wσ∗t (∇
(t )
1 j,σ := t 2
P
P
2
which acts on
σt
−i E 2 t1
,
(V.40)
σt
(t1 )
Gj
2 d 3 P, h( P)
(V.41)
P
(t )
E σ ) σ , see (III.23). The key advantage is that the vector 1 where σ = Wσ (∇ j,σt2 P P P inherits the Hölder regularity of σ ; see (III.13) in condition (I 2) of Theorem III.1. P We will refer to (V.41) as an infrared-regular vector. Accordingly, (V.39) now reads e
i H t1
N (t1 )
σt2
E σt2 ) eiγσt2 (v j ,∇ E P Wκ,σt2 ( v j , t1 ) Wσ∗t (∇ P
2
σt
,t1 ) −i E 2 t1
e
P
(t )
1 j,σ , (V.42) t 2
j=1
and we proceed as follows. A. Shifting the IR cutoff in the infrared-regular vector. First, we substitute e
i H t1
N (t1 )
σt2
E σt2 ) eiγσt2 (v j ,∇ E P Wκ,σt2 ( v j , t1 )Wσ∗t (∇ 2
P
σt
,t1 ) −i E 2 t1
e
P
j=1
−→ ei H t1
N (t1 )
σt
E 2)e Wκ,σt2 ( v j , t1 )Wσ∗t (∇ 2
j=1
P
(t )
1 j,σ t
σ
2
σ
E t1 ,t1 ) −i E t1 t1 iγσt ( v j ,∇ 1
P
e
P
1) (tj,σ , t 1
(V.43) where σt2 is replaced by σt1 in the underlined terms. We prove that the norm difference of these two vectors is bounded by the r.h.s. of (V.37). The necessary ingredients are:
800
T. Chen, J. Fröhlich, A. Pizzo
1) Condition (I 1) in Theorem III.1. 2) The estimate E σt2 , t1 ) − γσt ( E σt1 , t1 )| ≤ O vj, ∇ vj, ∇ |γσt2 ( 1 P
P
1 2 (1−δ) t1
σ
t11−θ
! " + O t1 σt1 ,
for t2 > t1 1, proven in Lemma A.1. The parameter 0 < θ < 1 is the same as the one in (III.26). 3) The cell partition G (t1 ) depends on t1 < t2 . 4) The parameter β can be chosen arbitrarily large, independently of , so that the −β infrared cutoff σt1 = t1 can be made as small as one wishes. First of all, it is clear that the norm difference of the two vectors in (V.43) is bounded by the norm difference of the two underlined vectors, summed over all N (t1 ) cells. Using 1) and 2), one straightforwardly derives that the norm difference between the two underlined vectors in (V.43) is bounded from above by 1
O( t1 σt12
(1−δ) − 23 t1
),
(V.44)
−3
where the last factor, t1 2 , accounts for the volume of an individual cell in G (t1 ) , by 3). The sum over all cells in G (t1 ) yields a bound 1
O( N (t1 ) σt12
(1−δ) 1− 23 t1
),
(V.45)
where N (t1 ) ≈ t13 , by 3). Picking β sufficiently large, by 4), we find that the norm −η difference of the two vectors in (V.43) is bounded by t1 , for some η > 0. This agrees with the bound stated in (V.37). B. Shifting the IR cutoff in the dressing operator. Subsequently to (V.43), we substitute e
i H t1
N (t1 )
σt1
E σt2 )eiγσt1 (v j ,∇ E P Wκ,σt2 ( v j , t1 )Wσ∗t (∇ 2
P
σt
,t1 ) −i E 1 t1
e
P
1) (tj,σ t
j=1
−→ ei H t1
N (t1 )
σt
E 1 )e Wκ,σt1 ( v j , t1 )Wσ∗t (∇ 1
P
σ
1
σ
E t1 ,t1 ) −i E t1 t1 iγσt ( v j ,∇ 1
P
e
P
1) (tj,σ t
j=1
1
(V.46)
where σt2 → σt1 in the underlined operators. A crucial point in our argument is that σt when σt1 (> σt2 ) tends to 0, the Hölder continuity of 1 in P offsets the (logarithmic) P divergence in t2 which arises from the dressing operator. We subdivide the shift σt2 → σt1 in σt
σt
E 2 ) −→ Wκ,σt ( E 1) Wκ,σt2 ( v j , t1 )Wσ∗t (∇ v j , t1 )Wσ∗t (∇ 1 2
P
1
P
(V.47)
into the following three intermediate steps, where the operators modified in each step are underlined: Step a) E σt2 ) v j , t1 )Wσ∗t ( v j )Wσt2 ( v j )Wσ∗t (∇ Wκ,σt2 (
P 2 ∗ ∗ σt2 v j , t1 )Wσt ( v j )Wσt2 ( v j )Wσt (∇ E ), Wκ,σt1 ( P 1 2 2
−→
(V.48)
Infraparticle States in QED I
801
Step b) σt
E 2) Wκ,σt1 ( v j , t1 )Wσ∗t ( v j )Wσt2 ( v j )Wσ∗t (∇
P ∗ ∗ σt1 Wκ,σt1 ( v j , t1 )Wσt ( v j )Wσt2 ( v j )Wσt (∇ E ), P 1 2 1
−→
2
(V.49)
Step c) E σt1 ) Wκ,σt1 ( v j , t1 )Wσ∗t ( v j )Wσt2 ( v j )Wσ∗t (∇
P 2 ∗ ∗ σt1 Wκ,σt1 ( v j , t1 )Wσt ( v j )Wσt1 ( v j )Wσt (∇ E ). P 1 1 1
−→
(V.50)
Analysis of Step a). In Step a), we analyze the difference between the vectors σt1
E σt2 )eiγσt1 (v j ,∇ E P ei H t1 Wκ,σt2 ( v j , t1 )Wσ∗t ( v j )Wσt2 ( v j )Wσ∗t (∇ 2
P
2
σt
,t1 ) −i E 1 t1
e
P
(t )
1 j,σ t
1
(V.51) and σt1
E σt2 )eiγσt1 (v j ,∇ E P v j , t1 )Wσ∗t ( v j )Wσt2 ( v j )Wσ∗t (∇ ei H t1 Wκ,σt1 ( 1
2
P
σt
,t1 ) −i E 1 t1
e
P
(t )
1 j,σ , t 1
(V.52) for each cell in G (t1 ) . Our goal is to prove that (V.51) − (V.52) ≤ const ln t2 P(t1 , t2 ),
(V.53)
where E σt2 )) P(t1 , t2 ) := sup (e−i(|k|t1 −k·x) − 1)Wσt2 ( v j )Wσ∗t (∇ P Bσ k∈ t
×e
2
1 σt
σt
E 2 ,t1 ) −i E 1 t1 iγσt ( v j ,∇ 2
P
e
P
−η (t1 ) j,σ ≤ O( t1 ln t2 ) t 1
(V.54)
as t1 → +∞, for some η > 0 independent of and for β large enough. Using the identity v j , t1 )Wσ∗t ( v j ) = Wκ,σt1 ( v j , t1 )Wσ∗t ( vj) Wκ,σt2 ( 2 1 sin(k · x − |k|t 1) v (k) · v iα j j × exp d 3k − 2 Bσt \Bσt |k|(1 k · v j ) 1 2 ⎛ ⎞ 1 −k· x) ∗ (e−i(|k|t 3k v − 1) − h.c. } j · {εk,λ bk,λ 1 d ⎠, × exp ⎝α 2 − |k|(1 k · v j ) Bσt \Bσt λ |k| 1 2 (V.55)
802
T. Chen, J. Fröhlich, A. Pizzo
the difference between (V.51) and (V.52) is given by
sin(k · x 1) v (k) v j · − |k|t j − |k|(1 k · v j ) Bσt \Bσt 1 2 ⎤ ⎞ 1 −k· x) ∗ −i(| k|t − 1) − h.c. } bk,λ d 3 k v j · {εk,λ (e ⎠ − I⎦ − |k|(1 k · v j ) |k|
ei H t1 Wκ,σt1 ( v j , t1 )Wσ∗t ( v j ) exp 1
⎡
⎛
× ⎣exp ⎝α 2
1
λ
Bσt \Bσt 1 2
σt
E 2)e × Wσt2 ( v j )Wσ∗t (∇ 2
+e
i H t1
σt P
1
%
Wκ,σt1 ( v j , t1 )Wσ∗t ( vj) 1 σt
2
d3k
σt P
E 1 ,t1 ) −i E 1 t1 iγσt ( v j ,∇
P
E 2)e v j )Wσ∗t (∇ × Wσt2 (
iα 2
iα exp 2
1
P
(t )
1 j,σ t
1
sin(k · x 1) v (k) − |k|t v j · j d k |k|(1 − k · v j ) 3
Bσt \Bσt
σ
E t1 ,t1 ) iγσt ( v j ,∇
P
e
e
1 2 σt −i E 1 t1 (t1 ) P j,σt1
,
(V.56) &
−I
(V.57)
where I is the identity operator in H. The norm of the vector (V.56) equals ⎡ ⎛ ⎞ ⎤ 1 −k· x) ∗ (e−i(|k|t 3 · { ε b − I) − h.c. } v j k,λ k,λ d k ⎣exp ⎝α 21 ⎠ − I⎦ − | k|(1 k · v ) Bσt \Bσt j λ |k| 1 2 σt1 σt1 σ t2 iγσt ( v j ,∇ E ,t1 ) −i E t1 (t1 ) ∗ P . (V.58) v j )Wσt (∇ E )e 1 e P j,σt × Wσt2 ( P 2 1 We now observe that • for k ∈ Bσt1 , σt
E 2) v j )Wσ∗t (∇ bk,λ Wσt2 (
(V.59)
=
(V.60)
P ∗ σt2 Wσt2 ( v j )Wσt (∇ E ) bk,λ P 2 σ t E 2 ) f ( v j )Wσ∗t (∇ + Wσt2 ( k,λ v j , P 2 2
where
Bσt \Bσt 1
P),
2 ≤ O(| ln σt2 |) d 3 k | f k,λ v j , P)| (
(V.61)
(V.62)
2
uniformly in v j , and in P ∈ S, and where j enumerates the cells. • for k ∈ Bσt1 , σ
bk,λ e
σ
E t1 ,t1 ) −i E t1 t1 iγσt ( v j ,∇ 2
P
e
P
(t )
1 j,σ = 0, t 1
(V.63)
(t )
1 because of the infrared properties of j,σ . t 1
From the Schwarz inequality, we therefore get (V.58) ≤ c | ln σt2 | P(t1 , t2 ),
(V.64)
Infraparticle States in QED I
803
for some finite constant c as claimed in (V.53), where E σt2 ) P(t1 , t2 ) = sup (e−i(|k|t1 −k·x) − 1) Wσt2 ( v j )Wσ∗t (∇ Bσ k∈ t
P
2
1
×e
σ
σ
E t2 ,t1 ) −i E t2 t1 iγσt ( v j ,∇ 2
e
P
P
(t1 ) j,σ , t
(V.65)
1
as defined in (V.54). To estimate P(t1 , t2 ), we regroup the terms inside the norm into σt2
E σt2 ) eiγσt2 (v j ,∇ E P v j ) Wσ∗t (∇ (e−i(|k|t1 −k·x) − 1)Wσt2 (
e
P
2
=
σt
,t1 ) −i E 2 t1
1 −k· x) E σt2 ) (e−i(|k|t Wσt2 ( v j ) Wσ∗t (∇ k P− 2
− I) e
P
σt
1) (tj,σ t σt
1
E 2 ,t1 ) −i E 2 t1 iγσt ( v j ,∇ 2
e
P
P
(t )
1 j,σ t
1
(V.66) σt2
E σt2 ) eiγσt2 (v j ,∇ E P v j ) Wσ∗t (∇ + Wσt2 ( 2
e
P−k
E σt2 ) e − Wσt2 ( v j ) Wσ∗t (∇ 2
σt
,t1 ) −i E 2 t1
P
σ E t2 ,t1 ) iγσt ( v j ,∇ 2 P
e
P
σt −i E 2 t1 P
(t )
1 j,σ t
1
(t )
1 j,σ . t 1
(V.67) (V.68)
We next prove that (V.66) , (V.67) − (V.68) ≤ O( (σt1 )ρ t1 ln t2 )
(V.69)
for some ρ > 0 independent of . To this end, we use: E σ described under condition (I 2) in Theorem i) The Hölder regularity of σ and ∇ P P III.1. ii) The regularity of the phase function σt
E 2 , t1 ) γσt2 ( vj, ∇
(V.70)
P
with respect to P ∈ supph ⊂ S expressed in the following estimate, which is similar to (A.3) in Lemma A.1: For k ∈ Bσt1 and t1 large enough, ' ' 1 ' 4 (1−δ ) t (1−θ) ), (V.71) E σt2 , t1 ) − γσt ( E σt2 , t1 ) '' ≤ O( |k| vj, ∇ v , ∇ ' γσt2 ( j 1 2 P P−k where 0 < θ (< 1) can be chosen arbitrarily close to 1. iii) The estimate σ bk,λ P ≤ C
1σ, (k) 3/2 |k|
(V.72)
from (I 5) in Theorem III.1 for P ∈ S, which implies
1/2 N f Pσ
=
λ
1/2 d
3
σ 2 k bk,λ P
≤ C | ln σ |1/2 .
(V.73)
804
T. Chen, J. Fröhlich, A. Pizzo
Likewise,
1/2 N f σP
=
2
−3/2 ) σ d 3 k bk,λ + O(|k| P
B \Bσ
λ
1/2
≤ C | ln σ |1/2 ,
(V.74) σt
which controls the expected photon number in the states { 1 }. As a side remark, P we note that the true size is in fact O(1), uniformly in σ , but the logarithmically divergent bound here is sufficient for our purposes. iv) The cell decomposition G (t1 ) is determined by t1 < t2 . Moreover, since β(> 1) −β can be chosen arbitrarily large and independent of , σt1 = t1 can be made as small as desired. We first prove the bound on (V.66) stated in (V.69). To this end, we use
e−i(|k|t1 −k·x) − I =e
σt
e
σ E t1 ,t1 ) iγσt ( v j ,∇ k 2 P−
+e
e
σ E t1 ,t1 ) iγσt ( v j ,∇ k 2 P−
−e
σt
E 1 ,t1 ) −i E 1 t1 iγσt ( v j ,∇
σ E t1 ,t1 ) iγσt ( v j ,∇ 2 P
2
σt −i E 1 t1 P−k
e e
e
P
P
1) (tj,σ t
(V.75)
1
(t )
1 (e−i(|k|t1 −k·x) − I) j,σ t
σt −i E 1 t1 P−k
σt −i E 1 t1 P
1) (tj,σ t
1
(V.76) (V.77)
1
(t )
1 j,σ . t
(V.78)
1
σt
The Hölder regularity of 1 from i) yields P
( 1 −δ ) − 3 t1 2
(V.76) ≤ O( t1 σt14
),
(V.79)
where δ can be chosen arbitrarily small, and independently of . The derivation of a similar estimate is given in the proof of Theorem A.3 in the Appendix, starting from σt E σt1 , again (A.27), to which we refer for details. The Hölder continuity of E 1 and ∇ P P from i), combined with ii), with θ sufficiently close to 1, implies that, with k ∈ Bσt , 1
1
− 3 2
(V.77) − (V.78) ≤ O( t1 σt15 t1
),
(V.80)
as desired. To prove the bound on (V.67) − (V.68) asserted in (V.69), we write E σt2 ) − Wσ∗ (∇ E σt2 ) = Wσ∗ (∇ E σt2 )(Wσ∗ (∇ E σt2 ; ∇ E σt2 ) − I), Wσ∗t (∇ t t t 2
P−k
2
P
2
P
P−k
2
P
(V.81) where E σt2 ; ∇ E σt2 ) := Wσt (∇ E σt2 ) Wσ∗ (∇ E σt2 ), Wσ∗t (∇ t 2 2
P−k
P
P
2
P−k
Infraparticle States in QED I
805
and apply the Schwarz inequality in the form E σt2 ; ∇ E σt2 ) − I) * (Wσ∗t (∇ k P+ P 2 1/2 ' ' d 3q ' σt2 E σt2 '' N 1/2 * , ∇ E ≤C sup − ∇ ' f k P− P q |3 Bσt | P∈supp h, k∈B
(V.82)
σt1
2
*≡e where in our case,
σ
σ
E t1 ,t1 ) −i E t1 t1 iγσt ( v j ,∇ 2
P
1/2 σt N f 1 P
e
P
(t )
1 j,σ . We have t 1
≤ c | ln σt1 |1/2 ≤ c ( ln t1 )1/2 ,
(V.83)
as a consequence of iii). Due to i), ' ' 1 ' σt2 E σt2 '' ≤ O( σt 4 −δ ), − ∇ ' ∇ E P− 1 k P
sup Bσ P∈supp h, k∈ t
(V.84)
1
where δ > 0 is arbitrarily small, and independent of (see (III.14)). Therefore,
sup (V.67) − (V.68) ≤ O((ln t2 ) (σt1 )ρ )
Bσ k∈ t
1
for some ρ > 0 which does not depend on (recalling that t1 < t2 ). We may now return to (V.53). From iv), and the fact that the number of cells is N (t1 ) ≈ t13 , summation over all cells yields N (t1 )
(V.51) − (V.52) ≤ O(
j=1
ln(t2 ) ρ ) t1
(V.85)
for some ρ > 0, provided that β is sufficiently large. This agrees with (V.37). / N (t ) The sum j=11 (V.57) can be treated in a similar way. Analysis of Step b). To show that the norm difference of the two vectors corresponding to the change (V.49) in (V.46) is bounded by the r.h.s. of (V.5), we argue similarly as for Step a), and we shall not reiterate the details. One again uses properties i) – iv) as in Step a). Analysis of Step c). Finally, we prove that the difference of the vectors corresponding to (V.50) satisfies N (t1 ) 2 σt1 σt1 ( σ ) σt i H t1 t E σt1 )−I eiγσt1 (v j ,∇ E P ,t1 ) e−i E P t1 ψ (t1 ) Wκ,σt1 ( v j , t1 ) W |σt12 ( v j )W ∗ |σt12 (∇ e j,σt P j=1
1
ρ ≤ O (ln(t2 ))2 /t1
(V.86)
for some ρ > 0, where we define σt
v j ) := Wσ∗t ( v j )Wσt2 ( v j ), W |σt12 ( 1
(V.87)
806
T. Chen, J. Fröhlich, A. Pizzo
and likewise, σt E σt1 ) := Wσ∗ (∇ E σt1 ) Wσt (∇ E σt1 ). W ∗ |σt12 (∇ t 1 P
P
2
P
(V.88)
We separately discuss the diagonal and off-diagonal contributions to (V.86) from the sum over cells in G (t1 ) . • The diagonal terms in (V.86). To bound the diagonal terms in (V.86), we use that, with E σt1 | ∗ , v j ≡ ∇ P= P j
σt σt E σt1 ) := W |σσtt1 ( E σt1 ) vj ; ∇ v j )W ∗ |σt12 (∇ W |σt12 ( 2 P
P
(V.89)
allows for an estimate similar to (V.82), where we now use that −( 41 −δ )
σt
E 1 − v j | ≤ O( t sup |∇ 1 P
G (t1 ) P∈ j
).
(V.90)
E σ , due to condition (I 2) in Theorem The latter follows from the Hölder regularity of ∇ P III.1; see (III.14). Moreover, we use (V.83) to bound the expected photon number in the σt1 states { }. P Hereby we find that the sum of diagonal terms can be bounded by −( 41 −δ )
(t )
O( N (t1 ) ψ j,σ1 t 2 (ln t2 ) t1 1
−ρ
) ≤ O( t1
ln t2 )
(V.91)
(t )
for some ρ > 0, using N (t1 ) = O(t13 ), and ψ j,σ1 t 2 = O(t1−3 ). 1 • The off-diagonal terms in (V.86). Next, we bound the off-diagonal terms in (V.86), corresponding to the inner product of vectors supported on cells j = l of the partition 1 (t, s) in (IV.6) that were discussed G (t1 ) . Those are similar to the off-diagonal terms M l, j in detail previously. Correspondingly, we can apply the methods developed in Sect. IV.1, up to some modifications which we explain now. Our goal is to prove the asymptotic orthogonality of the off-diagonal terms in (V.86). We first of all prove the auxiliary result σt
σt
E 1 )e−i H lim aσt1 ( ηl, j )(s) W ∗ |σt12 (∇
σt1
s
P
s→+∞
(t )
ψ j,σ1 t = 0. 1
(V.92)
To this end, we compare σt E σt1 )1 W ∗ |σt12 (∇ P
(where 1
(t1 )
Gj
(t1 )
Gj
( P)
(V.93)
is the characteristic function of the cell G j(t1 ) ) to its discretization:
1. We pick t¯ large enough such that G (t¯) is a sub-partition of G (t) ; in particular, /M (t¯) N (t¯) G j(t) = m( j)=1 Gm( j) , where M = N (t) . E σt∗1 , where P ∗ is the center of the cell G (t¯) , 2. Furthermore, defining um( j) := ∇ Pm( j)
(t¯) we have, for P ∈ Gm( j) ,
σt
E 1| ≤ C | u m( j) − ∇ P
where C is uniform in t1 .
m( j)
( 1 −δ ) 4 1 , t¯
m( j)
(V.94)
Infraparticle States in QED I
807
3. We define M
σt
Wσt12 (M) :=
σt
W ∗ |σt12 ( u m( j) ) 1G (t¯) ( P)
(V.95)
m( j)
m( j)=1
and rewrite the vector σt E σt1 )e−i H aσt1 ( ηl, j )(s) W ∗ |σt12 (∇
σt1
s
P
(t )
ψ j,σ1 t
(V.96)
1
in (V.92) as λ
Bκ \Bσt
( ) σt E σt1 ) − Wσσtt1 (M) d 3 k e−i k·x W ∗ |σt12 (∇ 2 P
1 σt1
−i E s · ε ∗ b ei|k|s × ηl, j (k) e P ψ (tj,σ1 )t k,λ k,λ M
+
σt
W ∗ |σt12 ( u m( j) )
λ
m( j)=1
×e
σt −i E 1 s P
(V.97)
1
Bκ \Bσt
· ε ∗ a ei|k|s d 3 k ηl, j (k) k,λ k,λ 1
(t¯)
ψm( j),σt .
(V.98)
1
We now observe that, at fixed t1 , t2 , (III.18) and the bound (V.94) imply that the vector in (V.97) converges to the zero vector as t¯ → +∞, uniformly in s. Moreover, the norm of the vector in (V.98) tends to zero, as s → +∞, at fixed t¯. This proves (V.92). The main difference between the vector corresponding to the jth cell in (V.86) and the similar expression in (IV.22) that is differentiated in s is the term proportional to the operator σt σt E σt1 ), W |σt12 ( v j )W ∗ |σt12 (∇
(V.99)
P
which is absent in (IV.22). To control it, we first note that the Hamiltonian ⊕ σt1 := σt1 d 3 P, H H P
(V.100)
where σt1 := H
f P − P>σt1 + α 1/2 Aσt1
P
with f P>σt1 :=
and f
H>σt1 :=
2 R3 \Bσt
f
+ H>σt1
(V.101)
∗ 3 k bk,λ d k, bk,λ
(V.102)
b∗ b d 3 k, |k| k,λ k,λ
(V.103)
1
R3 \Bσt
2
1
808
T. Chen, J. Fröhlich, A. Pizzo
satisfies σt1 σt1 = E σt1 σt1 , H
(V.104)
σt σt σt1 ] = 0. E σt1 ), H v j )W ∗ |σt12 (∇ [W |σt12 (
(V.105)
P
P
P
P
and P
Using (V.105), the vector in (V.86) corresponding to the jth cell can be replaced by σt1 s
ei H
Wκ,σt1 ( v j , s)e
σt1 s ! iγσt ( v j ,∇ E 1 ,s) −i H 1 P σt
e
'σ " E σt1 )−I ψ (t) , for s = t1 , W 'σt1 ( vj; ∇ j,σ t1 P
t2
(V.106) where we recall that σt σt E σt1 ) = W |σσtt1 ( E σt1 ). vj ; ∇ v j )W ∗ |σt12 (∇ W |σt12 ( 2 P
P
1 (t, s); Similarly to our strategy in Sect. IV.1, we control M l, j 'σt E σt1 ) in the vj ; ∇ ative in s of the term proportional to W 'σ 1 ( t2 P form
d ds
e
σt1 s iH σt1
= i ei H ×e
s
Wκ,σt1 ( v j , s) e
σt1 s iH
·
Bκ \Bσt
×e
e
σ iγσt ( v j ,∇ E t ,s) 1 P
e
Bκ \Bσt
P
Bκ \Bσt
jth cell vector has the
σt W |σt12 ( vj
E σt1 ) ψ (t1 ) ;∇ j,σt P
1
cos(k · x − |k|s) d 3k v (k) j 1
E σt1 ) ψ (t) ;∇ j,σt P
cos(k · x − |k|s) d 3k v (k) j 1
v ( q ) cos( q · x − | q |s)d 3 q j 1
1
σt1 s
σ P
σt
e
P
Wκ,σt1 ( v j , s) σt
σt E σt1 ) ψ (t) W |σt12 ( vj ; ∇ j,σt
P σt1 v j , ∇ E , s) dγσt1 ( P
σt
E 1 ,s) −i E 1 s iγσt ( v j ,∇ 1
σt
P
σt W |σt12 ( vj
Wκ,σt1 ( v j , s) α 2
iγσt ( v j ,∇ E t ,s) −i E 1 s
+ i ei H ×e
1
σt1 , x ] · Wκ,σt1 ( v j , s) α i[ H
σt −i E 1 s P
+i e
σt
E 1 ,s) −i E 1 s iγσt ( v j ,∇
see (IV.16). The deriv-
P
e
P
1
ds σt W |σt12 ( vj
E σt1 ) ψ (t1 ) . ;∇ j,σt P
1
(V.107)
Due to the similarity of this expression with (IV.22) – (IV.29), we can essentially adopt σt1 instead the analysis presented in Sect. IV.1. The only difference here is the operator H of H σt1 , and the additional term involving the commutator #
$ x ∗ σt1 σt1 χh , W |σt2 ∇ E (V.108) P s applied to the one-particle state σ
e
σ
E t1 ,s) −i E t1 s iγσt ( v j ,∇ 1
P
e
P
(t )
ψ j,σ1 t . 1
(V.109)
Infraparticle States in QED I
809
However, the latter tends to zero as s → +∞, at a rate O( s1η ), for some -independent E σ (condition (I 2) in Theorem η > 0. This follows from the Hölder regularity of ∇ P III.1), and (III.18). Similarly, we treat the commutator (V.108) with the infrared tail σt1 replacing H σt1 ). It is then straightforward to (IV.39) in place of χh ( xs ) (and with H see that we arrive at (V.5). " VI. Scattering Subspaces and Asymptotic Observables This section is dedicated to the following key constructions in the scattering theory for an infraparticle with the quantized electromagnetic field: i) We define scattering subspaces Hout/in which are invariant under space-time transout/in lations, built from vectors { h,κ }. out/in To this end, we first define a subspace, H˚ κ , depending on the choice of a threshold frequency κ with the following purpose: Apart from photons with energy smaller than κ, this subspace contains states describing only a freely moving (asymptotic) electron. out/in Adding asymptotic photons to the states in H˚ κ , we define spaces of scattering states of the system, where the asymptotic electron velocity is restricted to the < 1 }. E : | P| region {∇ P 3 out/in We note that the choice of H˚ κ is not unique, except for the behavior of the dressing photon cloud in the infrared limit. It is useful because – in the construction of the spaces of scattering states, we can separate “hard out/in photons” from the photon cloud present in the states in H˚ κ , which is not completely removable – each state in the scattering spaces contains an infinite number of asymptotic photons. – from the physical point of view, every experimental setup is limited by a threshold energy κ below which photons cannot be measured. out/in out/in ii) The construction of asymptotic algebras of observables, A ph and Ael , related to the electromagnetic field and to the electron, respectively. The asymptotic algebras are out/in – the Weyl algebra, A ph , associated to the asymptotic electromagnetic field; out/in
– the algebra Ael generated by smooth functions of compact support of the asymptotic velocity of the electron. out/in out/in The two algebras A ph and Ael commute. This is the mathematical counterpart of the asymptotic decoupling between the photons and the electron. This decoupling is, however, far from trivial: In fact, in contrast to a theory with a mass gap or a theory where the interaction with the soft modes of the field is turned off, the system is characterized by the emission of soft photons for arbitrarily long times. In this respect, the asymptotic convergence of the electron velocity is a new conceptual result, obtained from the solution of the infraparticle problem in a concrete model, here non-relativistic QED. Furthermore, the emission of soft photons for arbitrarily long times is reflected in the representation of the asymptotic electromagnetic algebra, which is non-Fock but only locally Fock (see Sect. VI.2). More precisely, the representation can be decomposed on the spectrum of the asymptotic
810
T. Chen, J. Fröhlich, A. Pizzo
velocity of the electron; for different values of the asymptotic velocity, the repre E = 0, the representation is sentations turn out to be inequivalent. Only for ∇ P Fock, otherwise they are coherent non-Fock. The coherent photon cloud, labeled by the asymptotic velocity, is the well known Bloch-Nordsieck cloud. All the results and definition clearly hold for both the out and the in-states. We shall restrict ourselves to the discussion of out-states. VI.1. Scattering subspaces and “One-particle” subspaces with counter threshold κ. In Sect. III, we have constructed a scattering state with electron wave function h, and a dressing cloud exhibiting the correct behavior in the limit k → 0, with maximal photon frequency κ. To construct a space which is invariant under space-time translations, we may either focus on the vectors
out e−i a· P e−i H τ ψh,κ ,
(VI.1)
or on the vectors obtained from s − lim ei H t t→+∞
where
N (t)
τ, a Wκ,σ ( v j , t)e t
σ P
σ P
E t ,t) −i E t t iγσt ( v j ,∇
e
(t)
ψ j,σt (τ, a ),
(VI.2)
j=1
⎛
τ, a Wκ,σ ( v j , t) := exp ⎝ α 2 t
1
λ
d 3k
Bκ \Bσt
∗ −i|k|(t+τ ) e−i k· a − h.c.} v j · {εk,λ a e k,λ
− |k|(1 k · v j )
|k|
⎞ ⎠,
(VI.3) and (t) ψ j,σt (τ, a )
:=
G j(t)
−i E σt τ
e−i a· P e
P
σt d 3 P. h( P)ψ
(VI.4)
P
Using Theorem III.2, one straightforwardly finds that
out e−i a· P e−i H τ ψh,κ
(VI.5)
= s − lim e−i a· P e−i H τ ei H t t→+∞
N (t)
Wκ,σt ( v j , t)e
σ P
σ P
E t ,t) −i E t t iγσt ( v j ,∇
e
(t)
ψ j,σt
(VI.6)
j=1
= s − lim ei H t t→+∞
N (t+τ )
τ, a Wκ,σ ( v j , t)e t+τ
σ P
σ P
E t+τ ,t+τ ) −i E t+τ t iγσt+τ ( v j ,∇
e
(t+τ )
ψ j,σt+τ (τ, a ).
j=1
(VI.7) The two limits (VI.2) and (VI.7) coincide; this follows straightforwardly from the line of analysis presented in the previous section. Therefore, we can define the “one-particle” space corresponding to the frequency threshold κ as out/in ∈ C 1 (S\Brα ), τ ∈ R, a ∈ R3 . (VI.8) ψh,κ (τ, a ) : h( P) H˚ κout/in := 0 out/in
By construction, H˚ κ
is invariant under space-time translations.
Infraparticle States in QED I
811
General scattering states of the system can contain an arbitrarily large number of “hard” photons, i.e., photons with an energy above a frequency threshold, say for instance out/in κ. One can construct such states based on H˚ κ according to the following procedure. We consider positive energy solutions of the form d 3k y e−i|k|t+i k· k) y ) := (VI.9) Ft ( √ F( 3 2 (2π ) |k| of the free wave equation y) ∂ 2 Ft ( y · ∇ y Ft ( ∇ y) − = 0, ∂t 2
(VI.10)
∈ C ∞ (R3 \{0}). k) which exhibit fast decay in | y | for arbitrary fixed t, and where F( 0 We then construct vector-valued test functions d 3k y ∗ λ −i|k|t+i k· y ) := (VI.11) Ft ( √ εk,λ F (k) e 3 2 (2π ) |k| λ=± satisfying the wave equation (VI.10), with ∗ λ ∞ 3 3 k) := εk,λ F( F (k) ∈ C 0 (R \{0} ; C ).
(VI.12)
λ
We set y) := ei H t A( y )e−i H t ; A(t,
(VI.13)
y ) is the expression in (II.11) with = ∞. An asymptotic vector potential is here A( constructed starting from LSZ (t → ±∞) limits of interpolating field operators, y) t ( ∂ A(t, y ) ∂ F y) · Ft , t] := i − · Ft ( A(t, y ) d 3 y, (VI.14) A[ ∂t ∂t with Ft as in (VI.11) for the negative-energy component, and with − Ft for the positiveenergy component. We define ψ
out/in h, F
:= s −
where ψh, F (t) := e
lim
t→+/−∞
ψh, F (t),
Ft ,t]− A[ Ft ,t] i A[
ψh,κ (t).
(VI.15)
(VI.16)
out in H ˚ κout/in (we temporarily drop the depenHere, ψh,κ (t) approximates a vector ψh,κ dence on (τ, a ) in our notation). The existence of the limit in (VI.15) is a straightforward consequence of standard decay estimates for oscillating integrals under the assumption in (VI.12), combined with the propagation estimate (III.44). Finally, we can define the scattering subspaces as out/in ∈ C 1 (S\Brα ), Hout/in := : h( P) F ∈ C0∞ (R3 \0 ; C3 ) . (VI.17) ψ 0 h, F
812
T. Chen, J. Fröhlich, A. Pizzo
VI.2. Asymptotic algebras and Bloch-Nordsieck coherent factor. We now state some theorems concerning the construction of the asymptotic algebras. The proofs can be easily derived using the arguments developed in Sects. IV and V; for further details we refer to [26]. Theorem VI.1. The functions f ∈ C0∞ (R3 ), of the variable ei H t xt e−i H t , have strong limits for t → ∞ in Hout/in , namely: x out/in out/in i Ht e−i H t ψ =: ψ s − lim e f , (VI.18) h, F h f ∇ E , F t→+/−∞ t := limσ →0 f (∇ E σ ). where f ∇ E ( P) P
The proof is obtained from an adaptation of the proof of Theorem A.3 in the Appendix. For the radiation field, we have the following result. Theorem VI.2. The LSZ Weyl operators
G t ,t]− A[ G t ,t] i A[ λ 2 3 −1 3 e : G (k) ∈ L (R , (1 + |k| )d k), λ = ±
(VI.19)
have strong limits in Hout/in : W
out/in
:= s − (G)
lim
t→+/−∞
e
G t ,t]− A[ G t ,t] i A[
.
(VI.20)
The limiting operators are unitary, and have the following properties: i) out/in +G )e− W out/in (G)W (G ) = W out/in (G
where
G ) = 2i I m ρ(G,
G ) ρ(G, 2
,
(VI.21)
λ
G (k)d 3k . λ (k) G
(VI.22)
λ
defines a strongly continuous, one paramii) The mapping R s −→ W out/in (s G) eter group of unitary operators. iii) −i H τ = W out/in (G −τ ), ei H τ W out/in (G)e
(VI.23)
−τ is a freely evolved, vector-valued test function in the time −τ . where G Next, we define out/in
as the norm closure of the (abelian) *algebra generated by the limits in • Ael (VI.18). out/in as the norm closure of the *algebra generated by the unitary operators in • A ph (VI.20).
Infraparticle States in QED I
813 out/in
From (VI.21) and (VI.23), we conclude that A ph is the Weyl algebra associated to a free radiation field. Moreover, from straightforward approximation arguments applied out/in out/in to the generators, we can prove that the two algebras, Ael and A ph , commute. Moreover, we can next establish key properties of the representation of the algeout/in bras A ph for the concrete model at hand that confirm structural features derived in [19] under general assumptions. out/in To study the infrared features of the representation of A ph , it suffices to analyze } of the algebra with respect to arbitrary the expectation of the generators { W out (G) out , states of the form ψh,κ out ψ out ψh,κ , W out (G) (VI.24) h,κ = lim
t→+∞
N (t)
e
σ P
σ P
E t ,t) −i E t t iγσt ( vl ,∇
j=1, l=1
G t ,0]− A[ G t ,0] i A[
∗ Wκ,σ ( vl , t)e t
e
(t)
ψl,σt ,
Wκ,σt ( v j , t)e
(VI.25) σ P
σ P
E t ,t) −i E t t iγσt ( v j ,∇
e
(t) ψ j,σt .
In the step passing from (VI.24) to (VI.25), we use Theorem III.2. One infers from the arguments developed in Sect. IV.1 that the sum of the off-diagonal terms, l = j, vanishes in the limit. Therefore, (VI.25) = lim
t→+∞
N (t)
e
j=1
e where
σ P
σ P
E t ,t) −i E t t iγσt ( v j ,∇
e
(t)
ψ j,σt ,
σt t ,0] v (G) G t ,0]− A[ G iγσ ( σt i A[ t v j ,∇ E ,t) −i E t j
e
:= 2i Re α u (G)
1 2
Bκ
λ
e
λ (k) G
P
e
ψ (t) j,σt , (VI.26)
u · ε∗
k,λ
2 (1 − u · |k| k) 3
P
d 3k .
(VI.27)
After solving an ODE analogous to (IV.9), we find that the diagonal terms yield C out (G) G out out 2 d 3 P, |h( P)| (VI.28) ψh,κ , W (G)ψh,κ = e− 2 e ∇ E P where
C G :=
k)| 2 d 3 k. |G(
(VI.29)
E σt∗ , combined with the convergence ∇ E σt → ∇ E (as Here, we also use that v j ≡ ∇ P Pj
P
t → ∞ and P ∈ S). out/in ) Now, we can reproduce the following results in [19]: The representation (A ph out/in
is given by a direct integral on the spectrum of the operator vas in Hout/in , defined by x −i H t out/in f ( vas e ) := s − lim ei H t f (VI.30) t→ +/−∞ t
814
T. Chen, J. Fröhlich, A. Pizzo
for any f ∈ C0∞ (R3 ), of mutually inequivalent, irreducible representations. These out/in representations are coherent non-Fock for values vas = 0. The coherent factors, out/in labeled by vas , are out/in
α
1 2
vas
2 (1 − vas |k| 3
out/in
· εk,λ out/in
α
and
· k)
out/in k,λ
for the annihilation and the creation part, a
1 2
vas
· ε∗
k,λ out/in
2 (1 − vas |k| 3
out/in ∗ , k,λ
and a
· k)
,
(VI.31)
respectively.
out/in (A ph )
The representation is locally Fock in momentum space. This property is equivalent to the following one: 0 κ ∈ C ∞ (R3 \Bκ ; C3 ), the operator For any κ > 0, and G 0 κt , t] G A[−
(VI.32)
out in the limit t → +∞, i.e., annihilates vectors of the type ψh,κ
G κt , t]ψ out = 0. lim A[− h,κ
(VI.33)
t→+∞
To prove this, we first consider Theorem III.2, then G κt , t]ψ out = lim A[− G κt , t]ψh,κ (t). lim A[− h,κ
t→+∞
(VI.34)
t→+∞
Next, we rewrite the vector κt , t]ψh,κ (t) = ei H t A[− G κt , 0] G A[−
N (t)
Wκ,σt ( v j , t)e
σ P
σ P
E t ,t) −i E t t iγσt ( v j ,∇
e
(t)
ψ j,σt
j=1
(VI.35) as − t
+∞
N (t)
d i Hs κ E σt ,s) −i E σt s (t) iγ ( v ,∇ P { e A[−G s , 0] Wκ,σt ( v j , s)e σt j e P ψ j,σt } ds ds j=1
(VI.36) + lim ei H s s→+∞
N (t)
G κs , 0]e Wκ,σt ( v j , s) A[−
E σt ,s) −i E σt s iγσt ( v j ,∇ P
e
P
(t)
ψ j,σt . (VI.37)
j=1
The integral in (VI.36), and the limit in (VI.37) exist. To see this, it is enough to follow the procedure in Sect. IV.1, taking into account that the operator G κs , 0] A[−
1 [H σt , x ] (H σt + i)
(VI.38)
is bounded, uniformly in t and s. The limit (VI.37) vanishes at fixed t because of condition (I 4) in Theorem III.1. Therefore we finally conclude that the limit (VI.34) vanishes. Liénard-Wiechert fields generated by the charge. Now we briefly explain how to obtain the result stated in (III.79). The assertion is obvious for the longitudinal degrees of
Infraparticle States in QED I
815
freedom; see the definition of Fµν in (III.80). For the transverse degrees of freedom, we argue as follows. Similarly to the treatment of (VI.24), we arrive at a sum over the diagonal terms, out/in i Ht tr e−i H t ψ out/in ψ ,e (0, y) δ˜ ( y − x − d) lim d 3 y Fµν t→±∞
h, F
N (t)
= lim
t→±∞
h, F
t ψ σj,σ , t
tr ψ σt . d 3 y Fµν (0, y) δ˜ ( y − x − d) j,σt
j=1
Then, one uses the pull-through formula as in Lemma 6.1 in [10], and Proposition 5.1 in [10] which identifies the infrared coherent factor by showing that ' ' ' ' 1 1σ, (k) 1 ' σ ' σ σ −1 2 + α ε · ∇ E ' P , bk,λ ' ≤ α 1/2 C |k| 1 σ k,λ P P ' ' | k| − k · ∇ E 2 |k| P
(VI.39) for k → 0. These ingredients imply that ' ⎧ ' N (t) ⎨ ' σt 2' tr ψ σt ψ j,σt , d 3 y Fµν (0, y) δ˜ ( y − x − d) | d | ' lim j,σt 't→±∞ ⎩ j=1 ⎫ N (t) ⎬ ∇ E σt tr P 2 ψ σt , ψ σt d3 P F − |h( P)| (0, d) µν P P ⎭ G (t) j=1
j
' ' ' ' ≤ O(|d| −1/2 ), ' ' (VI.40)
→ ∞, as asserted in (III.79). which vanishes in the limit |d| Appendix A. In the Appendix, we present detailed proofs of auxiliary results used in Sect. III. Lemma A.1. The following estimates hold for P ∈ S: (i) For t2 > t1 1, E σt2 , (σt2 )− θ ) − γσt ( E σt2 , (σt2 )− θ )| ≤ O(| vj, ∇ vl( j) , ∇ v j − vl( j) |), |γσt2 ( 2 1
1
P
P
(A.1) σt
σt
Pj
P ( j)
E ∗1 and v ( j) ≡ ∇ E ∗2 . where v j ≡ ∇ (ii) For t2 > t1 1,
1 E σt2 , t1 ) − γσt ( E σt1 , t1 ) | ≤ O [(σt1 ) 2 (1−δ) t 1−θ + t1 σt1 ] . | γσt2 ( vj, ∇ v , ∇ j 1 1 P
P
(A.2)
816
T. Chen, J. Fröhlich, A. Pizzo
(iii) For s, t 1 and q ∈ { q : | q | < s (1−θ) }, E σt , s) − γσt ( E σt | γσt ( vj, ∇ vj, ∇
P+ qs
P
θ
, s) | ≤ O(s − 4 (1−δ ) s (1−θ) ),
(A.3)
E σt , s) is defined. whenever γσt ( vj, ∇ P
Proof. The proofs only require the definition of the phase factor, and some elementary integral estimates, using conditions (I 1) and (I 2) in Theorem III.1. " Lemma A.2. For s ≥ t 1, the estimates ' ' ' ' | ln σt | ' ' l 3 sup ' ), v j (k) cos(k · x − |k|s)d k ' ≤ O( ' ' s 3 Bκ \Bσt x ∈R ' ' ' ' ' ' tθ x l 3 ' sup ' v j (k) cos(k · x − |k|s)d k χh ( ) '' ≤ O( 2 ), s ' s x ∈R3 ' Bκ \Bσ S
(A.4)
(A.5)
t
hold, where vl j (k)
1 kl kl l , := 2 (δl,l − )v j 2 2 |k| |k| (1 − k · v j ) l
(A.6)
and where σt := t −β , στS := τ −θ , with β > 1, 0 < θ < 1. Moreover, χh ( y ) = 0 for 1+νmax 1 | y | ≤ 2 νmin and | y | ≥ 2 with 0 < νmin < νmax < 1 (see (III.15)). Proof. To prove the estimate (A.4), we consider the variable x first in the set { x ∈ R3 : | x| < (1 − ρ)s, 0 < ρ < 1}. Integration with respect to |k| yields We denote by θk the angle between x and k. ' ' ' ' ' ' l 3 (A.7) v j (k) cos(k · x − |k|s)d k' ' ' Bκ \Bσt ' ' ' ' ' sin(κ k · x − κs) − sin(t −β k · x − t −β s) ' l ( = '' (A.8) k) d
' v j k k · x − s 2 v ( ≤ k)|d k , (A.9) | j ρs 2 l l ( where v j k) := |k| v j (k). For x in the set { x ∈ R3 : | x| > (1 − ρ)s, 0 < ρ < 1}, we integrate by parts with respect to cos θk , and observe that the two functions l ( v j k) and
l ( d( v j k)) d cos θk
(A.10)
Infraparticle States in QED I
817
belong to L 1 (S 2 ; d k ). This yields 3 cos(k · x − |k|s)d vl j (k) k Bκ \Bσt
= −
κ
σt κ
l ( v j k)|θk =π
(A.11)
| sin(|k| x| + |k|s) d|k|dφ k | |k| x|
(A.12)
| sin(|k| x| − |k|s) l ( d|k|dφ k)| θ =0 v j k k | |k| x| σt l ( d( · x − |k|s) d 3k v j k)) sin(k − . | 2 |k| x| |k| Bκ \Bσt d cos θk
−
The absolute values of (A.12), (A.13), and (A.14), are all bounded above by O
(A.13) (A.14)
| ln σt | s−ρs
,
as one easily verifies. This establishes (A.4), uniformly in x ∈ R3 . To prove (A.5), we consider x in a set of the form { x ∈ R3 : (1 − ρ )s > | x| > (1 − ρ)s, 0 < ρ < ρ < 1}.
(A.15)
in (A.12), (A.13), and (A.14) in the We apply integration by parts with respect to |k| case σtS . As an example, we get for (A.12), cos(κ (| x| + s)) l ( (A.12) = (A.16) dφk v j k)|θk =π κ (| x| + s) | x| cos(t −θ (| x| + s)) l ( dφk − (A.17) v j k)|θk =π t −θ (| x| + s) | x| κ | cos(|k| x| + |k|s) l ( dφk d|k|. k)| (A.18) + θ =π v j k 2 | |k| x|(| x| + s) σtS Since x is assumed to be an element of (A.15), it follows that the bound (A.5) holds for (A.12). In the same manner, one obtains a similar bound for (A.13) and (A.14). " Theorem A.3. For θ < 1 sufficiently close to 1, and s ≥ t, the propagation estimate x E σt ,s) −i E σt s (t) iγ ( v ,∇ P e σt j e P ψ j,σt (A.19) χh s σt σt E σt )eiγσt (v j ,∇ E P ,s) e−i E P s ψ (t) − χ h (∇ j,σ t P ≤ c
1 1 | ln(σt )| s ν t 32
(A.20)
holds, where ν > 0 is independent of . Proof. Since the detailed proof of a closely related result is given in Theorem A2 of [26], we only sketch the argument. Expressing χh (which we assume to be real) in terms of its Fourier transform χ h , start from the bound
818
T. Chen, J. Fröhlich, A. Pizzo
d 3q χ h ( q )(e
E σt −i q·∇
≤
d 3q χ h ( q )(e
E σt −i q ·∇
−e
P
d 3q χ h ( q) e
+
x
− e−i q· s ) e
P
σ P
i(E t −E
σt q P+ s
E σt ,s) −i E σt s iγσt ( v j ,∇
σ P
i(E t −E
)s
e
P
σt q P+ s
)s
)e
x
P
ψ (t) j,σt
E σt ,s) iγσt ( v j ,∇
(e−i q· s − 1) e
P
σ P
E t ,s) iγσt ( v j ,∇
(A.21)
ψ (t) j,σt
(A.22)
ψ (t) j,σt .
(A.23)
We split the integration domains of (A.22) and (A.23) into the two regions q : | q | > s 1−θ } I+ := {
I− := { q : | q | ≤ s 1−θ }.
and
(A.24)
In both (A.22) and (A.23), the contribution to the integral from I+ is controlled by the decay properties of χ ( q ), and one easily derives the bound in (A.20). For the contributions to (A.22) from the integral over I− , the existence of the gradient of the energy, the Hölder property in P of the gradient, and the decay properties of χ ( q ) are enough to produce the bound in (A.20). To control (A.23), we note that the two vectors σt
x
σt
and
P− qs
P− qs
:= e−i q· s σt
(A.25)
P
σt belong to the same fiber space H P− q , and that, as vectors in Fock space,
P− qs
s
coincide, i.e., x
−i q · s I P− σt ) ≡ I P ( σt ). q (e P
s
and σt P
(A.26)
P
We split and estimate (A.23)| I− , i.e., (A.23) where the integration domain is restricted to I− , by σ σ i(E t q −E t )s E σt ,s) σt iγ ( v ,∇ P q d 3 Pd 3 q s P (A.23)| I− = χ h ( q) e P− h P e σt j (A.27) (t)
− ≤
χ h ( q) I−
χ h ( q)
I−
I−
I−
χ h ( q) χ h ( q) I−
χ h ( q) I−
i(E
e
(t) j
(t)
σ P
−E t )s
σt q P− s
σt q P− s
i(E
j
σ P
h P e h P e
σ P
σ P
σt d 3 Pd 3 q P
E σt ,s) σt iγσt ( v j ,∇ q P P−
e
e
σt q P− s
E σt ,s) iγσt ( v j ,∇ P
h P e
σt
σ P
P− qs
E t ,s) iγσt ( v j ,∇
h P− q e
σ P
−E t )s
E iγσt ( v j ,∇
σt
σ P
σt q P+ s
)s
h P− q e
h P e
(A.29)
σt q P− s
E iγσt ( v j ,∇
σ P
E t ,s) iγσt ( v j ,∇
d 3 Pd 3 q
P− qs
,s)
σt q P− s
s
i(E t −E
d 3 Pd 3 q
(A.28)
s
h P e
−E t )s
−E t )s
σ P
E t ,s) iγσt ( v j ,∇
s
−
(t) j
e
)s
−E t )s
σt q P− s
i(E
(t) j
I−
σt q P− s
e
χ h ( q)
−
e
i(E
(t) j
+
e
σ σ i(E t −E t q P P+ s
i(E
(t) j
χ h ( q)
(t) j
− +
P− s
j
I−
d 3 Pd 3 q
σt
P− qs
,s)
σt
d 3 Pd 3 q
P− qs
d 3 Pd 3 q
σt d 3 Pd 3 q . P
(A.30)
(A.31)
Infraparticle States in QED I
819
The terms (A.29), (A.30), and (A.31) can be bounded by % | χh ( q )|
(A.29) ≤
(t)
j
I−
&1
σt |h P |2 I P ( σt ) − I P− )2 q ( P− q F P s
2
d3 P
d 3 q,
s
(A.32) %
(A.30) ≤
| χh ( q )|
(t)
j
I−
| q (h P e
E σt ,s) 2 iγσt ( v j ,∇ P
)|
s
&1 σt I P− )2 d 3 P q ( P− q F s s
2
d 3 q, (A.33)
and
⎡
⎤1
2
| χh ( q )| ⎣
(A.31) ≤ I−
j
O q
|h P |
2
I P ( σt )2F P
d P ⎦ d 3 q, 3
(A.34)
s
where q (h P e
E σt ,s) iγσt ( v j ,∇ P
s
(t)
j
(t),
and O q := ( j ∪ j s
q s
) := h P e (t)
E σt ,s) iγσt ( v j ,∇ P
− h P− q e
E σt iγσt ( v j ,∇
q P− s
,s)
s
(t),
)\( j ∩ j
q s
(t),
), where j
q s
is the translate by
, (A.35) q s
of the
(t)
cell j . Using (A.3), the C 1 −regularity of h P , and the definition of I− , one readily shows that the terms (A.33), (A.34) satisfy the bound (A.20), as desired. To estimate (A.32), we use the inequality
σt I P σt − I P− F (A.36) q q P P− s s
E σt ) σt − I q Wσt (∇ E σt q ) σt q F (A.37) ≤ I P Wσt (∇ P− s P P P− P− s s σt σt ∗ σt ∗ σt (W F , (A.38) ( ∇ E ) − W ( ∇ E )) W ( ∇ E ) +I P− q σ t σt σt q q q s
P
P− s
P− s
P− s
where it is clear that E σt ) σt = σt . Wσt (∇ P
P
(A.39)
P
Moreover, we use properties (I 2), (I 5) in Theorem III.1, where we recall that (I 2) Hölder regularity in P ∈ S uniformly in σ ≥ 0 holds in the sense of 4 −δ σP − σP+ ≤ Cδ | P| P F 1
(A.40)
and 4 −δ , Eσ − ∇ Eσ |∇ | ≤ Cδ | P| P P+ P 1
(A.41)
P + P ∈ S, and where Cδ and Cδ are for any 0 < δ < δ < 41 , with P, finite constants depending on δ and δ , respectively. We can bound (A.37) by use of (A.40).
820
T. Chen, J. Fröhlich, A. Pizzo
In order to bound (A.38), we recall the definition of the Weyl operator ⎛ ⎞ Eσ ∇ 1 ∗ P E σ ) := exp ⎝ α 2 ⎠ d 3k W σ (∇ · (εk,λ bk,λ 3 − h.c.) , P 2 δ ( B \ B | k| k) σ λ P,σ and we note that ' E σt E σt − ∇ (A.38) ≤ c ' ∇
P− qs
P
⎛ ' ' Rt ⎝ Rt + λ
B \Bσt
σ d 3 k bk,λ
P− qs
(A.42)
1 ⎞ 2 ⎠ 2F (A.43)
from a simple application of the Schwarz inequality, where 1 1 d 3k 2 Rt := = O(| ln σt | 2 ). 3 B \Bσt |k| Moreover, we have
λ
B \Bσt
σ d 3 k bk,λ
P− qs
2F ≤ c | ln σt |,
(A.44)
(A.45)
which is derived similarly as (V.74). (A.41), we obtain a contribution to the upper E σ in P, From Hölder continuity of ∇ P bound on (A.38) which exhibits a power law decay in s. We conclude that (A.32) is bounded by (A.20), as claimed. " Remark. By a similar procedure, one finds that for t2 ≥ s ≥ t1 , x iγ (v ,∇ E σt2 ,s) −i E σt2 s (t1 ) P e P ψ j,σt χ h ( ) e σt2 j 2 s σt2 σt2 1 | ln(σt2 )| E σt2 ) eiγσt2 (v j ,∇ E P ,s) e−i E P s ψ (t1 ) ≤ c ν − χ h (∇ . j,σ 3/2 t P 2 s t1
(A.46)
Analogous extensions hold for the estimates in the next theorem. Theorem A.4. Both % & +∞ d γσt ( v j , xs , s) x σsS i H σt s e Wκ,σt ( v j , s) J |σt (s)χh ( ) − t s ds E σt ,s) −i E σt s iγ ( v ,∇ (t) P e P (E σt + i) ψ j,σt ds × e σt j P and
(A.47)
+∞ d v j , xs , s) iγσt (v j ,∇ E σt ,s) −i E σt s σt γσt ( σt (t) P e ei H s Wκ,σt ( v j , s) e P (E + i) ψ j,σt P t ds E σt , s) vj, ∇ dγσt ( E σt ,s) −i E σt s iγσt ( v j ,∇ (t) σ P t P (A.48) e − e P (E + i) ψ j,σt ds P ds
Infraparticle States in QED I
821
are bounded by c
3 1 | ln(σt )|2 t − 2 , tη
d γ ( v , xs ,s)
σS
where η > 0 is -independent. J |σts (s), σt dsj (IV.38), (IV.39), and (IV.7) – (IV.8), respectively.
Proof. We recall from (IV.38) that for σsS ≥ σt , σS v (k) J |σts (s) = α i[H σt , x ] · j Bσ S \Bσt
(A.49) σ
E t ,s) dγσt ( v j ,∇ P ds
, and
1 H σt
are defined in
3 k, cos(k · x − |k|s)d
+i
s
where
σsS
:=
1 sθ
is the slow cut-off, and from (IV.39), σ
σ
t t v j , xs , s) d γˆσt ( d{ei H s x h (s)e−i H s } i H σt s 1 σt := α e−i H s σ e · ds H t +i ds 3 cos(k · ∇ E σt s − |k|s)d v (k) k. · j
P
Bσ S \Bσt
(A.50)
s
For s such that σsS ≤ σt unitarity of ei H s and σS
σt the expressions (A.47) and (A.48) are identically zero. By Wσt ( v j , s), we can replace the part in the integrand of (A.47)
proportional to J |σts (s) by e
i H σt s
σt
Wκ,σt ( v j , s)αi[H , x ]
1 H σt
x χh ( ) · +i s
B σ S \B σ t
cos(k · ∇ E σt s − |k|s) v (k) d 3k j P
s
×e
σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
(t)
(E σt + i)ψ j,σt ,
(A.51)
P
up to a term which yields an integral bounded in norm by 3 1 | ln(σt )|2 t − 2 , tη
(A.52)
for some η > 0 and independent of . To justify this step, we exploit the fact that the operator i[H σt , x ]
1 H σt + i
(A.53)
is bounded. Moreover, we are applying the propagation estimate 3 3 cos(k · x cos(k · ∇ E σt s − |k|s)d v (k) v (k) − |k|s)d k− k j j P Bσ S \Bσt
B σ S \ B σt
s
·e
E σt ,s) iγσt ( v j ,∇ P
s
e
σ −i E t s P
(E σt P
(t) + i) ψ j,σt
1 1 ≤ c 1+ν 3 | ln(σt )|, s t2
(A.54)
for some ν > 0, which is similar to (A.19). To obtain the upper bound, we exploit σS
the fact that due to the slow cut-off σsS = s −θ , θ > 0, in J |σts (s), the upper integration bound in the radial part of the momentum variables vanishes in the limit s → ∞.
822
T. Chen, J. Fröhlich, A. Pizzo
We note that we have to assume θ < 1 as required in (IV.37), in order to use the result in Lemma A.2. Next, we approximate (A.51) by ei H
σt s
Wκ,σt ( v j , s)α e−i H
σt s
1 H σt
x (s) dx (s) cos(k · ∇ E σt s − |k|s) v (k) d 3k χh ( )· j P + i ds s B σ S \B σ t s
· (E σt P
+ i)e
σ P
E t ,s) iγσt ( v j ,∇
where x (s) := ei H
σt s
(t) ψ j,σt ,
x e−i H
(A.55)
σt s
. To pass from (A.51) to (A.55), we have used
1 1 d x (s) 1 d[ x(s), H σt ] 1 d x (s) = σ − σ , σ σ ds H t + i H t + i ds H t +i ds H t +i
(A.56)
and we have noticed that the term containing 1 H σt
d[ x(s), H σt ] 1 +i ds H σt + i
(A.57)
can be neglected because an integration by parts shows that the corresponding integral 3 is bounded in norm by t1ν | ln(σt )|2 t − 2 for some ν > 0 and -independent. This uses ' ' ' ' ' ' σ 3 t ' E s − |k|s) ' ≤ O | ln (σt ) | (A.58) sup ' d k v j (k) cos(k · ∇ ' P s S ' B S \Bσt ' P∈ σs
and
' ' ' ' ' ' d 1 cos(k · ∇ ' ≤ O E σt s − |k|s) v (k) sup '' , (A.59) d 3k j ' P s 1+θ S ' ds Bσ S \Bσt ' P∈ s
which can be derived as in Lemma A.2. To bound the integral corresponding to (A.55), we note that up to a term whose integral is bounded in norm by (A.49), one can replace d xds(s) χh ( x (s) s ) by
d i H σt s σt e x h (s)e−i H s , ds
(A.60)
y ) defined as in Sect. IV.1. This is possible because where x h (s) := x χh ( xs ), with χh (
d i H σt s σt e x h (s)e−i H s ds x x −i H σt s σt = − ei H s x [ 2 · ∇χ h ( )]e s s x σ σt t + ei H s i[H σt , x ]χh ( )e−i H s s # $ x i[H σt , x ] −i H σt s σt x e + ei H s ∇χh ( ) · s s 2 # $ i[H σt , x ] x σt x σt · ∇χh ( ) e−i H s , + ei H s s 2 s
(A.61) (A.62) (A.63) (A.64) (A.65)
Infraparticle States in QED I
823
where (A.63) corresponds to d xds(s) χh ( x (s) s ). Moreover, we use the fact that the vec1 σ t tor operator H σt +i i[H , x ] is bounded, and apply the propagation estimate (A.19) to xi x j j ∇ χh ( xs ) s2
and to xs ∇ j χh ( xs ) with appropriate modifications (see (A.72) and recall h (∇ E σt ) = 0 for P ∈ supp h). that ∇χ P We observe that σt σt d( ei H s x h (s)e−i H s ) 1 σt σt v (k) ei H s Wκ,σt ( v j , s) α e−i H s σ d 3k · j H t +i ds B S \Bσt i
σs
(E σt + i) e E σt s − |k|s) · cos(k · ∇ P
P
P
corresponds to e
E σt ,s) iγσt ( v j ,∇
%
i H σt s
Wκ,σt ( v j , s)
v j , xs , s) d γˆσt ( ds
(t)
ψ j,σt
& e
(A.66)
E σt ,s) iγσt ( v j ,∇ P
(t)
ψ j,σt .
(A.67)
This immediately implies (A.47). To prove (A.48), we need to control the integral s¯ σt σt α d( ei H s x h (s)e−i H s ) σt σt v (k) · ei H s Wκ,σt ( v j , s)e−i H s σ d 3k j H t +i ds Bσ S \Bσt t s
σt
eiγσt (v j ,∇ E P ,s) (E σt + i)ψ (t) ds E σt s − |k|s) · cos(k · ∇ j,σt P
(A.68)
P
for s¯ → +∞. An integration by parts with respect to s yields α i H σt s v (k) x h (s) · e Wκ,σt ( v j , s) σ d 3k j H t +i B S \Bσt σs
' t t e−i E P s eiγσt (v j ,∇ E P ,s) (E σt + i)ψ (t) 's¯ E σt s − |k|s) × cos(k · ∇ j,σ t t P P s¯ d i H σt s α σt σt (e − Wκ,σt ( v j , s)e−i H s ) ei H s σ t +i ds H t σ
× x h (s) ·
Bσ S \Bσt
σ
(A.69)
cos(k · ∇ E σt s − |k|s) v (k) d 3k j P
s
σ E σt ,s) −i E t s iγσt ( v j ,∇
×e −
P
e
P
(t)
(E σt + i)ψ j,σt ds
(A.70)
P
s¯
α σt ei H s Wκ,σt ( v j , s) σ H t +i t ⎧ ⎤⎫ ⎡ ⎨ d ⎬ σ t cos(k · ∇ eiγσt (v j ,∇ E P ,s) ⎦ E σt s − |k|s) ⎣ v (k) × xh (s) · d 3k j P ⎩ ds ⎭ B S \Bσt σs
×e
σ P
−i E t s
(E σt P
(t)
+ i)ψ j,σt ds
Here, we notice that x x h (s) = x χh ( ) = −is s
(A.71)
x
χh ( ∇ q )e−i q· s d 3 q.
(A.72)
824
T. Chen, J. Fröhlich, A. Pizzo
Furthermore, the operator
−i
tends to
−i
x
χh ( ∇ q )e−i q· s d 3 q
(A.73)
σt
E −i q·∇ χh ( P d 3 q ∇ q )e
(A.74)
for s → ∞, if it is applied to the vectors e or
Bσ S \Bσt
σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
e
(t)
ψ j,σt ,
cos(k · ∇ E σt s − |k|s)e v (k) d 3k j
(A.75) σ P
σ P
E t ,s) −i E t s iγσt ( v j ,∇
P
e
ψ (t) j,σt , (A.76)
s
or ⎧ ⎡ ⎨d ⎣ ⎩ ds B
\Bσt σsS
v j ) cos(k · ∇ e E σt s − |k|s) k, d 3 k ( P
⎤⎫ ⎬
E σt ,s) iγσt ( v j ,∇ ⎦ P
⎭
e
σ P
−i E t s
(t)
ψ j,σt . (A.77)
The rate of convergence of the corresponding expression in (A.48) is bounded by (A.49). Therefore, we can replace expressions (A.69),(A.70), and (A.71) by σt
E σt v j , s)α s ∇ (A.78) ei H s Wκ,σt ( P ' E σt ,s) −i E σt s (t) 's¯ iγσt ( v j ,∇ cos(k · ∇ E σt s − |k|s)e v (k) P d 3k e P ψ j,σt ' · j Bσ S \Bσt s
− ·
s¯
ds t
P
t
d i H σt s σt E σt (e Wκ,σt ( v j , s)e−i H s ) α s ∇ P ds
(A.79) σt
Bσ S \Bσt
E ,s) (t) v j ,∇ 3 iγσt ( cos(k · ∇ E σt s − |k|s)d v (k) P ke ψ j,σt j P
s
s¯
σt E σt − ei H s Wκ,σt ( v j , s)α s ∇ (A.80) P ⎧t ⎡ ⎤⎫ ⎬ ⎨d E σt ,s) iγσt ( v j ,∇ 3 cos(k · ∇ E σt s − |k|s)d ⎦ ⎣ v (k) P k e · j P ⎭ ⎩ ds B S \Bσt σs
×e
σ −i E t s P
(t)
ψ j,σt ds.
Recalling the definition of the phase factor in (IV.7)-(IV.8), the sum of the expressions (A.78), (A.79), and (A.80) can be written compactly as s¯ v j , ∇ E P , s) iγσt (v j ,∇ E σt ,s) −i E σt s (t) dγσt ( σt P e ds ei H s Wκ,σt ( v j , s) e P ψ j,σt , (A.81) ds t after an integration by parts. This implies the asserted bound for (A.47).
"
Infraparticle States in QED I
825
Acknowledgements. The authors gratefully acknowledge the support and hospitality of the Erwin Schrödinger Institute (ESI) in Vienna in June 2006, where this collaboration was initiated. T.C. was supported by NSF grants DMS-0524909 and DMS-0704031. A. P. was supported by NSF grant DMS-0905988. Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
References 1. Bach, V., Fröhlich, J., Pizzo, A.: An infrared-finite algorithm for Rayleigh scattering amplitudes, and Bohr’s Frequency Condition. Commun. Math. Phys. 274(2), 457–486 (2007) 2. Bloch, F., Nordsieck, A.: Phys. Rev., 52, 54 (1937); Bloch, F., Nordsieck, A.: Phys. Rev. 52, 59 (1937) 3. Buchholz, D.: Commun. Math. Phys. 52, 147 (1977) 4. Buchholz, D.: Phys. Lett. B 174, 331 (1986) 5. Bach, V., Fröhlich, J., Pizzo, A.: Infrared-finite algorithms in QED I. The groundstate of an atom interacting with the quantized radiation field. Commun. Math. Phys. 264(1), 145–165 (2006) 6. Bach, V., Fröhlich, J., Pizzo, A.: An infrared-finite algorithm for Rayleigh scattering amplitudes, and Bohr’s frequency condition. Commun. Math. Phys. 274(2), 457–486 (2007) 7. Bach, V., Fröhlich, J., Sigal, I.M.: Renormalization group analysis of spectral problems in quantum field theory. Adv. in Math. 137, 205–298 (1998) 8. Bach, V., Fröhlich, J., Sigal, I.M.: Spectral analysis for systems of atoms and molecules coupled to the quantized radiation field. Commun. Math. Phys. 207(2), 249–290 (1999) 9. Chen, T.: Infrared renormalization in nonrelativistic QED and scaling criticality. J. Funct. Anal. 254(10), 2555–2647 (2008) 10. Chen, T., Fröhlich, J.: Coherent infrared representations in nonrelativistic QED. In: Spectral Theory and Mathematical Physics: A Festschrift in Honor of Barry Simon’s 60th Birthday. Proc. Symp. Pure Math., Providence, RI: Amer. Math. Soc., 2007 11. Chen, T., Fröhlich, J., Pizzo, A.: Infraparticle scattering states in non-relativistic QED: II. Mass shell properties. J. Math. Phys. 50(1), 012103 (2009) 12. Chung, V.: Phys. Rev. 140 B, 1110 (1965) 13. Faddeev, L., Kulish, P.: Theor. Math. Phys. 5, 153 (1970) 14. Fierz, M., Pauli, W.: Nuovo. Cim. 15, 167 (1938) 15. Fröhlich, J.: On the infrared problem in a model of scalar electrons and massless, scalar bosons. Ann. Inst. Henri Poincaré, Section Physique Théorique 19(1), 1–103 (1973) 16. Fröhlich, J.: Existence of dressed one electron states in a class of persistent models. Forts. der Phys. 22, 159– 198 (1974) 17. Fröhlich, J., Pizzo, A.: Renormalized electron mass in non-relativistic QED, Preprint http://arxiv.org/abs/ 0908.1858v1[math-ph], 2009 18. Fröhlich, J., Griesemer, M., Schlein, B.: Asymptotic electromagnetic fields in models of quantum-mechanical matter interacting with the quantized radiation field. Adv. Math. 164(2), 349–398 (2001) 19. Fröhlich, J., Morchio, G., Strocchi, F.: Charged sectors and scattering state in quantum electrodynamics. Ann. Phys. 119(2), 241–284 (1979) 20. Jackson, J.-D.: Classical Electrodynamics. New York: Wiley, 1998 21. Jauch, J.M., Rohrlich, F.: Helv. Phys. Acta 27, 613 (1954) 22. Jauch, J.M., Rohrlich, F.: Theory of Photons and Electrons. Reading, MA: Addison-Wesley, 1980 23. Kibble, T.: J. Math. Phys. 9, 315 (1968) 24. Nelson, E.: Interaction of nonrelativistic particles with a quantized scalar field. J. Math. Phys. 5, 1190–1197 (1964) 25. Pizzo, A.: One-particle (improper) states in Nelson’s massless model. Ann. H. Poincaré 4(3), 439–486 (2003) 26. Pizzo, A.: Scattering of an infraparticle: The one-particle (improper) sector in Nelson’s massless model. Ann. H. Poincaré 6(3), 553–606 (2005) 27. Rohrlich, F.: Phys. Rev. 98, 181 (1955) 28. Schroer, B.: Infrateilchen in der Quantenfeldtheorie. (German) Fortschr. Physik 11, 1–31 (1963) 29. Reed, M., Simon, B.: Methods of Modern Mathematical Physics. Vol. I – IV New York: Academic Press, 1972, 1975, 1977, 1978 30. Strocchi, F., Wightman, A.S.: Proof of the charge superselection rule in local relativistic quantum field theory. J. Math. Phys. 15, 2198–2224 (1974) 31. Yennie, D., Frautschi, S., Suura, H.: Ann. Phys. 13, 375 (1961) Communicated by H. Spohn
Commun. Math. Phys. 294, 827–862 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0972-4
Communications in
Mathematical Physics
Gravity, Twistors and the MHV Formalism Lionel Mason, David Skinner The Mathematical Institute, 24-29 St. Giles’, Oxford OX1 3LP, United Kingdom. E-mail:
[email protected] Received: 3 September 2008 / Accepted: 28 April 2009 Published online: 24 December 2009 – © Springer-Verlag 2009
Abstract: We give a self-contained proof of the formula for the MHV amplitudes for gravity conjectured by Berends, Giele & Kuijf and use the associated twistor generating function to define a twistor action for the MHV diagram approach to gravity. Starting from a background field calculation on a spacetime with anti-self-dual curvature, we obtain a simple spacetime formula for the scattering of a single, positive helicity linearized graviton into one of negative helicity. Re-expressing our integral in terms of twistor data allows us to consider a spacetime that is asymptotic to a superposition of plane waves. Expanding these out perturbatively yields the gravitational MHV amplitudes of Berends, Giele & Kuijf. We go on to take the twistor generating function off-shell at the perturbative level. Combining this with a twistor action for the anti-self-dual background, the generating function provides the MHV vertices for the MHV diagram approach to perturbative gravity. We finish by extending these results to supergravity, in particular N = 4 and N = 8. 1. Introduction Recent advances in understanding the perturbative structure of gravity (see e.g. [1–16]) have uncovered structures that are not visible in the standard spacetime formulation of general relativity. A particularly striking development has been the chiral MHV (Maximal Helicity Violating) diagram formulation [2,13–15]. In this approach, the full perturbation theory for gravity, at least at tree level, is built up out of standard massless scalar propagators and MHV vertices. These vertices are off-shell continuations of amplitudes describing interactions of n linearized gravitons in momentum eigenstates, two of which have positive1 helicity while n −2 have negative helicity. Such amplitudes were first conjectured for Yang-Mills by Parke & Taylor [17] (and proved by Berends and Giele [18]) 1 We will use Penrose conventions for twistor space, in which the amplitudes supported on a twistor line are ‘mostly minus’; these amplitudes are usually thought of as MHV, but will be called MHV here. Our conventions are detailed at the end of the introduction.
828
L. Mason, D. Skinner
Fig. 1. Reversing the momentum of one of the positive helicity particles leads to the interpretation of the MHV amplitude as measuring the helicity-flip of a single particle which traverses a region of ASD background curvature
and later a more complicated formula (95) for gravity was conjectured by Berends, Giele & Kuijf [19]. Both in gravity and Yang-Mills, MHV amplitudes are considerably simpler than a generic tree-level helicity amplitude. In particular, they may involve an arbitrary number of negative helicity gravitons (gluons) at little or no cost in complexity. Why should this be? Bearing in mind that a negative helicity graviton that has positive frequency is anti-self-dual [20,21], the picture in Fig. 1 interprets MHV amplitudes as measuring the helicity-flip of a single particle as it traverses a region of anti-self-dual (ASD) background curvature. The asd Einstein equations, like the ASD Yang-Mills equations, have long been known to be completely integrable [22,23] and lead to trivial scattering at tree-level. From this perspective, the key simplification of the MHV formalism arises because the ASD background, despite its non-linearities, can effectively be treated as a free theory. The MHV amplitudes themselves represent the first departure from antiself-duality. The MHV formulation is essentially chiral. For gravity, this chirality suggests deep links to Plebanski’s chiral action [24–26], to Ashtekar variables [26,27] and to twistor theory [22,28]. It is the purpose of this article to elucidate these connections further and to go some way towards a non-linear formulation that helps illuminate the underlying nonperturbative structure. Thus we begin in Sect. 2 with a brief review of the Plebanski action, explaining how it can be used to expand gravity about its anti-self-dual sector. Similar discussions have been given in [24,29] and more recently [30] whose treatment we follow most closely. On an ASD background, a linearized graviton has a canonically defined self-dual part, but its anti-self-dual part shifts as it moves through the spacetime. We show in Sect. 2.2 that the tree-level amplitude for this shift to occur is precisely measured by a simple spacetime integral formula. This integral is a generating function for all the MHV amplitudes. To obtain them in their usual form, one must expand out the background field in terms of fluctuations around flat spacetime. Understanding how a non-linear anti-self-dual field is composed of linearized gravitons is feasible precisely because the asd equations are integrable, but nonetheless the inherent non-linearity makes this a rather complicated task on spacetime [31]. However, by going to twistor space and using Penrose’s non-linear graviton construction [22], the ASD background can be reformulated in an essentially linear way. Hence in Sect. 3, after reviewing the relevant twistor
Gravity, Twistors and the MHV Formalism
829
theory of both linear gravity and non-linear ASD gravity, we obtain a twistor representation of the generating function using twistor integral formulæ for the spacetime fields. We will see that it is straightforward to construct a twistor space for a non-linear ASD spacetime that asymptotically is a linear superposition of momentum eigenstates. This uses a representation for the twistor space as an asymptotic twistor space constructed from the asymptotic data and is closely related to Newman’s H -space construction [32]. Thus, we can use the twistor description to expand our generating function around Minkowski spacetime. A completely analogous story is true in Yang-Mills [33,34], with the corresponding twistor expression yielding all the Parke-Taylor amplitudes. This is reviewed in Appendix B; some readers may find it helpful to refer to the (somewhat simpler) Yang-Mills case for orientation. Performing the expansion, one finds that the n-point amplitude comes from an integral over the space of holomorphic twistor lines with n marked points. The marked points support operators representing the external gravitons; the 2 positive helicity gravitons are represented by 1-form insertions while the n − 2 negative helicity gravitons give insertions of vector fields. These vectors differentiate the external wavefunctions, leading to what is sometimes called ‘derivative of a δ-function support’. The 1-forms and vector fields really represent elements of certain cohomology classes on twistor space. It is interesting to note that these are the same cohomology groups that arise as (part of) the BRST cohomology in twistor-string theory [35,36], but here there are extra constraints which ensure that they represent Einstein, rather than conformal, gravitons. A string theory whose vertex operators satisfy these extra constraints was constructed in [37], although these models do not appear to reproduce the MHV amplitudes [38]. Integrating out the twistor variables finally yields the formula n κ n−2 (4) (n) M = δ pi i=1 n−1 [1n]8 1 k| pk+1 + · · ·+ pn−1 |n] × +P{2,...,n−2} , (1) [1 n−1][n−1 n][n 1] C(n) [kn] k=2
√ for the n-particle amplitude M(n) , where κ = 16π GN and we have used the spinor-helicity formalism: the i th external graviton is taken to have null momentum pα(i)α˙ = |i[i|, where |i and [i| respectively denote the anti-self-dual and self-dual spinor constituents of pi , and C(n) is the cyclic product [12,23] · · · [n − 1 n][n1]. The symbol P{2,...,n−2} denotes a sum over permutations of gravitons 2 to n − 2; the amplitude is completely symmetric in the external states (up to the overall factor [1n]8 from the two positive helicity gravitons) once these permutations are accounted for. Equation (1) is not the original expression of BGK [19] and an analytic proof that the two forms coincide for arbitrary n ≥ 4 is given in Appendix A. The twistor formula also yields the correct 3-point amplitude, which is non-zero in complexified momentum space (although yields zero on a Lorentzian real slice). Our generating function may be simply extended to the case of MHV amplitudes in supergravity, and this is discussed in Sect. 6 for N = 4 and N = 8 supergravity. In the MHV diagram formalism, the full perturbation theory is reproduced from MHV amplitudes that are continued off-shell to provide vertices. These vertices are then connected together with propagators joining positive and negative helicity lines. With p such propagators, one obtains a N p MHV amplitude, usually thought of in terms of the scattering of 2 + p positive helicity gravitons and an arbitrary number of negative
830
L. Mason, D. Skinner
helicity gravitons. In Sect. 5 we continue our twistorial generating function off-shell and couple it to the twistor action for anti-self-dual gravity constructed in [39]. The Feynman diagrams of the resulting action reproduce (in a certain gauge) the MHV diagram formalism for gravity. At present, we understand this action only in perturbation theory, and its validity as an action for gravity rests on the validity of the MHV diagram formalism. It would be very interesting to learn how the off-shell twistor action generates off-shell curved spacetime metrics, or to see if the existence of the twistor action implies that the MHV expansion is indeed valid. The gravitational MHV amplitudes were originally calculated [19] using the Kawai, Llewelyn & Tye relations [42], and subsequently recalculated in a different form using the Britto, Cachazo, Feng & Witten recursion relations [43], suitably modified for gravity [5–7]. Although the BGK expression is strongly constrained by having the correct soft and collinear limits, strictly speaking, BGK were only able to prove that their formula followed from the KLT relations for n ≤ 11 external particles. The formulæ obtained from BCFW recursion relations have also only been verified to be equivalent to the BGK expression up to this level. Our derivation is a complete constructive proof of the BGK formula (the formulæ of [5–7] are also independently proved). Evidence for a MHV diagram formulation of perturbative gravity has been discussed in [13–15], based on recursion relations. It has been established [15] that the MHV diagrams yield the correct n-graviton amplitudes, again for n ≤ 11. Reference [15] also gives a generating function for MHV amplitudes in N = 8 supergravity, taking the BGK amplitudes as an input. Some steps towards an MHV action for gravity have been taken in [44], starting from lightcone gauge in spacetime and inspired by the work of Mansfield in YangMills [45,46]. A twistorial generating function which reproduces the gravity MHV amplitudes was constructed by Nair in [47]. Nair’s paper has influenced this one; the main difference is that we give an independent derivation of the amplitudes, starting from a spacetime formula for scattering off an ASD background. We also take a more geometrical perspective than [47]. A treatment of the MHV amplitudes that emphasizes their close connection to the integrability of asd backgrounds has been given in [31,48] using ‘perturbiners’. 1.1. Conventions and notation. Flat Minkowski spacetime M is taken to be R4 with metric of Lorentz signature (+ − −−) and with vector indices a = 0, 1, 2, 3. Let S+ and S− be the self-dual and anti-self-dual spin spaces. Elements of S± will be taken to ˙ 1; ˙ α, . . . = 0, 1. We have dotted and undotted Greek indices respectively, i.e. α, ˙ . . . = 0, denote the Levi-Civita alternating spinor by εαβ = ε[αβ] , with ε01 = −1, etc. We often ˙ use the notation r α ↔ |r and s α˙ ↔ |s] and then [ p r ] = p α r β εαβ and s t = s α˙ t β εα˙ β˙ denote the S L(2, C)-invariant inner products. In complexified spacetime the two spin bundles will also be denoted S+ and S− . On a Lorentzian real slice they are related by complex conjugation S+ = S− , which therefore exchanges dotted and undotted spinor indices. Vector indices a = 0, 1, 2, 3 can be replaced by spinor indices, so that the position vector of a point can be given as 0 1 x + x3 x1 + i x2 (2) x α α˙ = √ 1 2 0 3 . 2 x − ix x − x ˙ , so that the rhs of (2) is a Hermitian matrix. The Lorentz reality condition is x α α˙ = x¯ αα We will often work on complexified spacetime, where x a and x α α˙ are complex and the reality condition is dropped.
Gravity, Twistors and the MHV Formalism
831
Projective twistor space PT is the space of totally null self-dual two planes (α-planes) in complexified spacetime. We describe PT using homogeneous coordinates (ωα , πα˙ ), with the incidence relation being ωα = ix α α˙ πα˙ ; the solutions for x α α˙ holding (ωα , πα˙ ) constant defines the α-plane. In these conventions, an element of H 1 (PT , O(−2s − 2)) corresponds to an on-shell massless field of helicity s in spacetime by the Penrose transform. Thus a negative helicity gluon has homogeneity zero in twistor space, and the amplitudes supported on degree 1 holomorphic curve are ‘mostly minus’. We call such + + − − · · · − − amplitudes MHV, although they are the complex conjugate of what is called an MHV amplitude in much of the scattering theory literature. With our conventions, Witten’s twistor-string theory [40] is really in dual twistor space. In Lorentzian signature, twistor space and its dual are related via complex conjugation, i.e. (ωα , πα˙ ) ∈ PT → (π¯ α , ω¯ α˙ ) ∈ PT ∗ , reflecting the Lorentzian conjugation of Weyl spinors. For complexified spacetime, one often gives dual twistor space independent coordinates (λα , µα˙ ) which are the coordinates used in [40]. 2. MHV Amplitudes on ASD Background Fields 2.1. The Plebanski action. The (complexified) spin group of a Lorentzian four manifold M is S L(2, C) × S L(2, C). Correspondingly, the tangent bundle T M decomposes into the self-dual and anti-self-dual spin bundles S± as T M S+ ⊗ S− . Each S L(2, C) factor acts non-trivially on only either S+ or S− and so any connection on T M may be decomposed into connections on the two spin bundles as ⊕ . Splitting the curvature two-form into its self-dual and anti-self-dual parts R ± , one finds that R + = R + ( ) and R − = R − ( ) so that the self-dual (ASD) part of the curvature depends only on the connection on S+ (S− ). (On a Lorentzian four-manifold ∗2 = −1, so the SD/ASD curvatures are complex and = , R + = R − . In Euclidean or split signature the spin connections and R ± are real and independent. We will mostly work on complexified spacetime, imposing reality conditions only at the end.) Plebanski [24] gave a chiral action for Einstein’s general relativity that brings out this structure (see also [25,26]). In his approach, the basic variables are the self-dual spin connection , together with a tetrad of 1-forms eα α˙ which define the metric by ˙
ds 2 = εαβ εα˙ β˙ eα α˙ eβ β ,
(3)
where εαβ = ε[αβ] , ε01 = 1 and similarly for εα˙ β˙ . The components of the tetrad are defined by eα α˙ = eaα α˙ d x a and form a vierbein. Plebanski’s action is a first-order theory in which and the tetrad are treated as independent a priori. In the absence of a cosmological constant, the action is
1 ˙ S[ , ] = 2
α˙ β ∧ (d + ∧ )α˙ β˙ , (4) κ M ˙
where κ 2 = 16π GN and α˙ β are three self-dual two-forms, given in terms of the tetrad ˙ β) ˙ α(α˙ ∧ eα . It is a striking fact that plays no role in this action2 . It neverby α˙ β = e theless describes full (non-chiral) Einstein gravity, as follows from the field equations: ˙
d α˙ β + 2
γ˙ d α˙ β˙ + (α˙
(α˙ γ˙
˙ γ˙ β)
∧
= 0, α α˙ ∧ β) = 0. ˙ γ˙ ∧ e
2 Of course, one can still construct an ASD spin connection from the tetrad.
(5) (6)
832
L. Mason, D. Skinner
The first of these is the condition that is torsion-free, which fixes it in terms of the tetrad. Since (after an integration by parts) appears in the action only algebraically, this equation may be viewed as a constraint. Imposing it in (6) implies that the Ricci curvature of the metric (3) vanishes, so that M satisfies the vacuum Einstein equations. Thus Plebanski’s action is equivalent to the Einstein-Hilbert action (up to a topological term). ˙ It is also possible to take α˙ β to be an arbitrary set of self-dual 2-forms and view ˙ them as the basic variables, as was done in [25]. The condition that α˙ β comes from a ˙ β) ˙ α(α˙ tetrad (i.e. α˙ β = e ∧ eα ) is ensured by including a Lagrange multiplier to enforce ˙ ˙
(α˙ β ∧ γ˙ δ) = 0. In the present paper, this constraint will naturally be solved as part of ˙ ˙ the construction of α˙ β from twistor space. We also remark that α˙ β and α˙β˙ may be thought of as a 4-covariant form of Ashtekar variables [27]: if C is a spacelike Cauchy α˙ β˙ i ˙ surface in M, then the restriction of α˙ β to C gives Ashtekar’s densitized triads3 σ[ jkl] via
α˙ β˙ α˙ β˙ k α˙ β˙ α˙ β˙ i
[ jk δ il] = σ[ jkl] , (7)
[i j] = 3σ[i jk] C
C
whereas the restriction of to C is the Ashtekar-Sen-Witten connection (see [26] for details). 2.2. Linearizing around an anti-self-dual background. We will be particularly interested in anti-self-dual solutions to (5)–(6). On an ASD solution, the self-dual spin bundle S+ → M is flat, so vanishes up to a gauge transform. The torsion-free constraint (5) becomes ˙
d α˙ β = 0 ,
(8)
so that the self-dual part of the spin connection constructed from the tetrad eα α˙ must also be pure gauge. There are no constraints on the anti-self-dual part of this connection, so the associated Riemann tensor R abcd (e) need not vanish, but is purely asd. Decomposing a general Riemann tensor into irreducibles gives [49] α˙ β˙ γ˙ δ˙ εαβ εγ δ + αβ γ˙ δ˙ εα˙ β˙ εγ δ + γ δ α˙ β˙ εαβ εγ˙ δ˙ Rabcd = αβγ δ εα˙ β˙ εγ˙ δ˙ + R εαγ εβδ εα˙ β˙ εγ˙ δ˙ + εαβ εγ δ εα˙ γ˙ εβ˙ δ˙ , + 12
(9)
α˙ β˙ γ˙ δ˙ = (α˙ β˙ γ˙ δ) where ˙ and α˙ βγ ˙ δ = (α˙ β)(γ ˙ δ) are the spinor forms of the self-dual part of the Weyl tensor and the trace-free part of the Ricci tensor, respectively, and R is the scalar curvature. With vanishing cosmological constant, R abcd (e) is anti-self-dual if α˙ β˙ γ˙ δ˙ , α˙ βγ and only if ˙ δ and R vanish. The ASD part αβγ δ of the Weyl tensor need not vanish (at least in complexified or Euclidean spacetime), but it obeys ∇ α α˙ αβγ δ = 0 as a consequence of the Bianchi identities on the ASD background. Anti-self-dual spacetimes are sometimes known as ‘half-flat’ or ‘left-flat’. As discussed in [30], such left-flat spacetimes are all that survive in a chiral limit of the Plebanski action, obtained by rescaling → κ 2 and then taking the limit κ 2 → 0. In this chiral theory, is independent of the tetrad even after the field equations are imposed. 3 i, j, k, . . . are indices for the tangent space to C.
Gravity, Twistors and the MHV Formalism
833
In the full theory (4), set = 0 + σ and = 0 + γ to consider a small fluctuation on a background ( 0 , 0 ). We will eventually take all fluctuations to be proportional to the coupling κ. When the background is anti-self-dual (so 0 is closed and 0 vanishes), the fluctuations are subject to the linearized field equations ˙
˙ γ˙ β)
dσ α˙ β = −2γ (γα˙˙ 0
and
β β˙
dγα˙ β˙ ∧ e0 = 0.
(10)
Note that the exterior derivatives d here can be thought of as acting covariantly on the dotted spinor indices, since S+ → M is flat in the background. After some algebra, the second of these equations implies that α˙ β˙ γ˙ δ˙ γ˙ δ˙ , dγα˙ β˙ = ψ 0
(11)
(α˙ β˙ γ˙ δ) α˙ β˙ γ˙ δ˙ = ψ where ψ ˙ . Taking the exterior derivative of this equation and using (8) α α ˙ α˙ β˙ γ˙ δ˙ = 0, so ψ α˙ β˙ γ˙ δ˙ may be intepreted as a linearized self-dual Weyl tensor yields ∇ ψ propagating on the asd background. Since (10) are linearized, their space of solutions is a vector space V . If S is the infinite dimensional space of solutions to the nonlinear field equations (5)-(6), then V may be thought of as the fibre of T S over the ASD background ( 0 , 0 ) ∈ S. An on-shell linearized fluctuation (σ, γ ) preserves the anti-self-duality of the Riemann tensor if and only if it lies in a subspace V − ⊂ V defined by γ αβ˙˙ = 0, modulo gauge. However, we cannot invariantly define an analogous subspace V + of self-dual solutions modulo gauge, because (e.g.) the condition that the variation of the ASD Weyl tensor should vanish is not true for infinitesimal diffeomorphisms and so such a definition is not gauge invariant. (In fact, it would be over-determined.) We can nevertheless define V + as the quotient V + = V /V − so that V + = {(σ, γ ) ∈ V }/{(σ, γ )|γβα˙˙ = dµαβ˙˙ } = {γ |dγα˙ β˙ ∧ eα α˙ = 0}/{γβα˙˙ = dµαβ˙˙ }. (12) An element [σ, γ ] ∈ V + determines a unique non-zero linearized self-dual Weyl tensor by (11). The definitions of V ± are summarized in the exact sequence 0 → V − → V → V + → 0,
(13)
where the second arrow is inclusion, and the third arrow is the map sending (σ, γ ) → γ modulo linearized gauge transformations. Exactness means that if a linearized solution projects to zero in V + , then it necessarily comes from one in V − . On a flat background, V decomposes as V = V + ⊕ V − , but on an ASD background such a global splitting is obstructed because elements of V + cannot globally be required to have non-vanishing anti-self-dual parts. We will see that the MHV amplitudes precisely measure this obstruction. 2.3. Scattering of linearized fields. Figure 1 in the introduction realises the MHV amplitudes as the plane wave expansion of the amplitude for the scattering of a single, linearized graviton off an ASD background. The linearized graviton is taken to have positive helicity in the asymptotic past. To fix ideas, we consider a scattering process to take initial (characteristic) data from I − to data on I + . Here, I ± are future/past null infinity [49] and form the future/past boundaries of the conformal compactification of an asymptotically flat spacetime. They have the structure of lightcones (whose vertices are usually
834
L. Mason, D. Skinner
taken to be at infinity), so they have topology S 2 × R. In the conformal compactification of Minkowski space, the lightcone of a point on I − refocuses on a corresponding point of I + and thus I ± are canonically identified. The inversion x a → x a /x 2 sends the lightcone of the origin to I ± in the conformal compactification. For our scattering process, the linearized graviton is prepared to have positive helicity on I − and scatters off the ASD background to emerge with negative helicity in the asymptotic future I + . For positive frequency fields, states of positive or negative helicity are self-dual or anti-self-dual, respectively [20,21]. On a curved spacetime, one can sometimes (perhaps with some gauge choices) define the positive/negative frequency splitting on an arbitrary Cauchy surface, but in general the results on different Cauchy surfaces will not agree, as is familiar e.g. from Hawking radiation. However, for an asymptotically flat spacetime, I ± are lightcones at infinity’ and have the same S 2 × R topology as in Minkowski space. For these spacetimes, we can use Fourier analysis in the R factors to perform the positive/negative frequency splitting at4 I + or I − . Equivalently, one can split a field into parts that analytically continue into the upper and lower half planes respectively of the complexification C of the R generators. On an asymptotically flat spacetime that is anti-self-dual, one can say more: as in Minkowski space, the lightcone emitted from an arbitrary point of I − refocusses at a point of I + , so I ± may again be canonically identifed. (The reason for this will become transparent in the twistor formulation of the next section; essentially, identified points of I ± correspond to the same Riemann sphere in twistor space.) Thus, on an ASD background, the positive/negative frequency splittings at I − and I + agree, and it is easy to check they reproduce the standard splitting when the spacetime is flat. Thus we wish to find an expression for the scattering of a self-dual linearized graviton by an arbitrary asymptotically flat, asd spacetime M. In the path integral approach, to compute the scattering amplitude, we formally consider the integral [D D ] eiS/, taken over all fields that approach the prescribed behaviour at I ± . In the tree-level approximation, the path integral is given simply by evaluating eiS/ on fields that extend this boundary configuration throughout the spacetime in accordance with the equations of motion, i.e. on ( 0 + σ, γ ). To leading order in the fluctuations, this is eiS/ ≈ 1 +
i κ 2
α˙ β˙
M
γ˙
0 ∧ γ α˙ ∧ γγ˙ β˙ .
(14)
The first term on the right-hand side is the diagonal part of the S-matrix. The remaining part is the classical approximation to the transition amplitude we seek. This term is simply i/ times the part of the Plebanski action that is lost in the chiral limit mentioned ˙ above. Indeed, because satisfies d α˙ β˙ ∧ eβ β = 0 in the chiral theory, is indistinguishable from the linearized fluctuation in in the full theory. This field equation for γ also implies that the formula is gauge invariant since if we change γ → γ + dχ with χ of compact support, the change in the integrand is clearly exact with compact support γ˙ α˙ β˙ ˙ since d(γα˙ β˙ ∧ eβ β ∧ eβ ) = 0 and d 0 = 0. In the MHV diagram formulation, the full classical theory can be built up from the complete set of MHV vertices, together with a propagator derived from the chiral 4 Strictly, to split into positive/negative frequency at I ± , we must first perform a conformal rescaling so as to make sense of the limits of the fields at infinity. Such conformal rescalings can be canonically restricted to be constant along the generators [49] so there is no ambiguity in the splitting.
Gravity, Twistors and the MHV Formalism
835
theory5 . Thus it is perhaps not surprising that all of the infinite number of MHV amplitudes should somehow be contained in this term. We will see later how to use this expression as a generating function for all the gravitational MHV amplitudes. 2.3.1. An alternative derivation We will now rederive the expression for the scattering amplitude in more detail. Although this derivation is instructive, the impatient reader may prefer to skip ahead to the next section. Consider canonical quantization of the Plebanski action around an anti-self-dual (rather than flat) background. The amplitude we seek might then be written as out |in asd , where · | · asd is the inner product on the Hilbert space of the theory describing fluctuations around the asd background, and in , out are in and out states of the appropriate helicity. We can construct this inner product from the symplectic form on the phase space of the classical theory as follows (see e.g. [52,53]). The space of solutions S to (5)-(6) possesses a naturally defined closed two-form defined using the boundary term in the variation of the action S. Letting δ denote the exterior derivative on the space of fields, ˙ so that δ α˙ β and δ α˙β˙ are one-forms on S, is given by 1 = 2 κ
C
˙
δ α˙ β ∧ δ α˙ β˙ ,
(15)
where C is a Cauchy surface in M. is independent of the choice of Cauchy surface, because if C1 and C2 are two such surfaces bounding a region D ⊂ M (i.e. ∂ D = C1 − C2 ) then
˙ ˙ ˙ (16) δ α˙ β ∧ δ α˙ β˙ = δ
α˙ β ∧ δ α˙ β˙ = δ d α˙ β ∧ δ α˙ β˙ . C1 −C2
∂D
D
Provided the field equations hold throughout D, this last term is δ 2 S and so vanishes because δ is nilpotent. Therefore, is invariant under diffeomorphisms of M (whether or not these preserve C) and under rotations of the spin frame (it has no free dotted spinor indices). Moreover, vanishes when evaluated on any changes in and that come from such a diffeomorphism or spin frame rotation, so it descends to a symplectic form on S/Diff0+ (M). This symplectic form is real for real fields in Lorentzian signature. The quantum mechanical inner-product · | · is then defined as i · · = ( · , P+ · ),
(17)
where P+ projects states onto their positive frequency components6 , defined at I ± as above. We can use the symplectic form to define a duality between V + and V − . The symplectic form vanishes on restriction to the anti-self-dual linearized solutions V − (which 5 As mentioned in the Introduction, the status of the MHV formalism in gravity - justified using recursion relations - requires a more complete understanding of the possible contribution from the ‘pole at infinity’ [15]. However, tree-level MHV diagrams in (super) Yang-Mills are known to be equivalent to Feynman diagrams [50,51]. 6 P is a choice of ‘polarization’ of the phase space in which positive/negative frequency states are taken + to be holomorphic/antiholomorphic. We make this choice by defining it at null infinity, and no ambiguity arises as to whether future or past infinity is chosen in an asymptotically flat, ASD spacetime. One can check that (17) is positive definite, and linear/anti-linear in its left/right entries, with respect to the complex structure of the polarization.
836
L. Mason, D. Skinner
have γ = 0, mod gauge). So, if h a,b = (σa,b , γa,b ) are two elements of V T S| Masd and h a ∈ V − ⊂ V , then ( · , h a ) annihilates any part of h b that is in V − and we have (h b , h a ) = −
1 κ2
C
˙
σaα˙ β ∧ γb α˙ β˙
(18)
for any (σb , γb ) ∈ V . We see from this formula that the pairing only depends on γb , i.e.
∗
the projection of (σb , γb ) into V + . Therefore, we have an isomorphism V + V − . We need to prepare our incoming field so that it is purely self-dual, so we need to construct a splitting of the sequence (13). This is easily done on I ± using the standard expression of characteristic data for the gravitational field in terms of the asymptotic shear σ [49]. Since this expression may not be familiar to many readers, we give a somewhat formal, but equivalent definition: motivated by (18) we will say that a linearized field (σb , γb ) is self-dual at I ± if, given a one-parameter family Ct of Cauchy hypersurfaces, with Ct → I ± as t → ±∞, then
lim
t→±∞ C t
α˙ β˙
σb ∧ γc α˙ β˙ = 0 ,
∀γc ∈ V + .
(19)
We wish to consider the amplitude for a positive frequency, linearized solution h 1 that has positive helicity at I − to evolve into a positive frequency, negative helicity linearized solution at I + by scattering off the ASD background. That is, h 1 is purely self-dual at I − so it satisfies (19), and we wish to know its anti-self-dual part after evolving it to I + . From the discussion above, we can extract this by computing the ∗ inner product with a linearized field h 2 that is purely self-dual (in V − ) at I + . Taking + this inner-product at I , for positive frequency states the amplitude is
i i α˙ β˙
h 2 h 1 = (h 2 , P+ h 1 ) = − 2 σ ∧ γ2 α˙ β˙ κ I+ 1
(20)
because (σ2 , γ2 ) is purely self-dual at I + . Now, ∂ M = I + − I − , so Stokes’ theorem gives
i i α˙ β˙ α˙ β˙ α˙ β˙ h 2 h 1 = − 2 dσ1 ∧ γ2 α˙ β˙ + σ1 ∧ dγ2 α˙ β˙ − 2 σ ∧ γ2 α˙ β˙ κ M κ I− 1
i γ˙ α˙ β˙ α˙ β˙ 2 α˙ β˙ γ˙ δ˙ γ˙ δ˙ = 2
0 ∧ γ1 α˙ ∧ γ2 β˙ γ˙ − σ1 ∧ ψ (21) 0 κ M
i γ˙ α˙ β˙ = 2
0 ∧ γ1 α˙ ∧ γ2 β˙ γ˙ . κ M In going to the second line, we used the linearized field equations (10) together with the α˙ β˙ fact that I − σ2 ∧ γ1 α˙ β˙ = 0 because h 1 is purely self-dual at I − . The third line fol˙
˙ γ˙ δ)
˙
˙
lows because σ (α˙ β ∧ 0 = 0 from the linearization of the constraint (α˙ β ∧ γ˙ δ) = 0 that ensures = 0 + σ comes from a tetrad. Equation (21) agrees with the form of the tree amplitude computed before, as it should.
Gravity, Twistors and the MHV Formalism
837
3. Twistor Theory for Gravity Although we have argued that they are related, the expression (14) (or (21)) is still a far cry from the usual form of the MHV amplitudes, which live on a flat background spacetime. To connect the two pictures, we must expand out the ASD background in (21) in terms of plane wave perturbations away from Minkowski space. This background is α˙ β˙ explicitly present in (21) through 0 and also implicit through the equations satisfied by the γ s. In order to perform the expansion we will have to use the integrability of the ASD interactions. Even so, constructing a fully nonlinear asd background that is asymptotically a superposition of negative helicity momentum eigenstates, and then using this background to evaluate (21) is a very complicated task. What enables us to proceed is the use of twistor theory, which brings out the integrability of the ASD sector and is therefore well-adapted to the problem at hand. We now briefly review the twistor theory of linearized gravity on flat spacetime, before moving on to discuss Penrose’s non-linear graviton construction [22] which gives the twistor description of an asd spacetime (see e.g. [49,54,55] for textbook treatments). 3.1. Linearized gravity. We first review the basic twistor correspondence. The twistor space PT of flat spacetime is CP3 with a CP1 removed. We can describe CP3 using homogeneous coordinates Z I = (ωα , πα˙ ), where I = 0, . . . , 3, while α = 0, 1 and ˙ 1˙ are spinor indices as before. In these coordinates, the line that is removed is α˙ = 0, given by πα˙ = 0, so that πα˙ = 0 on PT . Hence PT fibres over the CP1 whose homogeneous coordinates are πα˙ . Points x ∈ C4 of (complexified) spacetime with coordinates x α α˙ correspond to lines (CP1 s) in PT by the incidence relation ωα = ix α α˙ πα˙ .
(22)
We will denote this line by L x . The removed line πα˙ = 0 corresponds to a point at infinity in spacetime (the vertex of the lightcone at infinity). We use the standard notation O(n) to denote the line bundle on CPm of Chern class n. Sections of O(n) can be identified with functions on the non-projective space of homogeneity degree n, so that Z I ∂ f /∂ Z I = n f . We will use the same notation for line bundles over a projective line (m = 1) and over twistor space (m = 3). The normal bundle to L x in PT is N L x |PT O(1) ⊕ O(1). In particular, for x = 0, α ω are coordinates along the fibres of the normal bundle to L 0 . Thus, in this flat case, PT is the total space of the normal bundle to a line. The incidence relation (22) identifies a point x with a holomorphic section CP1 → PT and the space of such sections H 0 (L x , N L x |PT ) C4 is (complexified) flat spacetime. The correspondence with flat spacetime can also be expressed in terms of the double fibration P(S+ ) p PT
@ q , R @ M
(23)
where P(S+ ) is the projectivization of the bundle of dotted spinors, coordinatized by (x α α˙ , πβ˙ ) up to scaling of the π s, and M C4 is complexified Minkowski space. The bundle P(S+ ) → M is necessarily trivial, and the fibres q −1 (x) are CP1 s coordinatized by πα˙ . Conversely, the fibres p −1 (ωα , π α˙ ) are the set of points (x α α˙ , πα˙ ) such that
838
L. Mason, D. Skinner
ωα = ix α α˙ πα˙ ; given one such point (x0 , π ), this is the totally null, complex two-plane x0α α˙ + λα π α˙ . The Penrose transform represents linearized gravitons of helicities −2 and +2 on spacetime as elements of the twistor space cohomology groups H 1 (PT , O(2)) and H 1 (PT , O(−6)), respectively. In a Dolbeault framework, these are described locally by (0, 1)-forms h(Z ) and h(Z ), homogeneous of degrees 2 and −6. h and h thus obey ¯ = 0 = ∂¯ ¯ , ¯ We will ∂h h and are defined up to the gauge freedom h ∼ h + ∂χ h ∼ h + ∂λ. suppress (0, p)-form indices in what follows (and some readers may prefer to think in ˇ terms of a Cech picture of cohomology). The Penrose transforms of h and h are
ψαβγ δ (x) = α˙ β˙ γ˙ δ˙ (x) = ψ
[π dπ ] ∧ p
∗
Lx
∂ 4h , ∂ωα ∂ωβ ∂ωγ ∂ωδ
(24)
[π dπ ] ∧ πα˙ πβ˙ πγ˙ πδ˙ p ( h), ∗
Lx
where the pullback p ∗ simply imposes the incidence relation (22). Differentiating under obey the usual spin-2 (i.e. linearized Einstein) the integral sign shows that ψ and ψ α α ˙ α α ˙ ¯ equations ∂ ψαβγ δ = 0, ∂ ψα˙ β˙ γ˙ δ˙ = 0 provided only that h and h are ∂-closed. The cohomology class h plays an active role through its associated Hamiltonian vector field, V := I (dh, · ) = I J K
∂h ∂ . ∂ZJ ∂ZK
(25)
Here I is a holomorphic Poisson bivector of homogeneity −2. It is determined by the line that was removed from CP3 to reach PT and has components αβ ∂ ∂ ε 0 so that I = εαβ α ∧ I JK = . (26) 0 0 ∂ω ∂ωβ It follows that V in (25) represents an element of H 1 (PT , TPT ) and so describes a linearized complex structure deformation. We will study these deformations further in the next subsection. A positive helicity graviton may also be represented by an element B ∈ H 1 (PT , 1,0 ⊗ O(−4))
(27)
¯ of a cohomology class, if, as well as having the standard gauge freedom B → B + ∂χ B is also subject to the additional gauge freedom B → B + ∂m + n[π dπ ].
(28)
Here, m and n are (0, 1)-forms of homogeneity −4 and −6 respectively, while χ is a (1, 0)-form of weight −4. The freedom to add on arbitrary multiples of [π dπ ] means that only the part Bα dωα of B along the fibres of PT → CP1 contains physical information; the remaining freedom B → B + ∂m means that this physical information is captured by I (d B) = I I J ∂ I B J = εαβ
∂ Bβ . ∂ωα
(29)
Gravity, Twistors and the MHV Formalism
839
I (d B) is again in H 1 (PT , O(−6)) and so can be identified with h. The Penrose transform of B is
α˙ [π dπ ] ∧ π α˙ πβ˙ p ∗ (B) (30) γ β˙ = 2 Lx
which, as our notation suggests, may be interpreted as a linearized self-dual spin connection. (The factor of 2 is for later convenience.) To see this, note first that (30) respects the gauge freedom (28) because any piece of p ∗ B proportional to [π dπ ] wedges to zero in (30), while adding on a total derivative B → B + dm corresponds to the linearized gauge freedom γ αβ˙˙ → γ αβ˙˙ + dµαβ˙˙ of a spacetime connection. (µαβ˙˙ is the Penrose transform of m and satisfies the asd Maxwell equation ∂α α˙ µα˙β˙ = 0.) The linearized spin connection generates a linearized curvature fluctuation as it ought, since
˙ ∂ ∗ dγα˙ β˙ = 2d x δ δ [π dπ ] ∧ π π p (B) ˙ α ˙ β ∂ x δ δ˙ L
x ∂ Bγ ˙ [π dπ ] ∧ πα˙ πβ˙ πγ˙ πδ˙ p ∗ = 2d x δ δ ∧ d x γ γ˙ ∂ωδ L
x ˙ γ˙ = d x δδ ∧ d xδ [π dπ ] ∧ πα˙ πβ˙ πγ˙ πδ˙ p ∗ h Lx
α˙ β˙ γ˙ δ˙ d x =ψ
δ δ˙
γ˙
∧ d xδ ,
(31)
where in the second line we used the fact that p ∗ B = iBγ (ix · π, π )d x γ γ˙ πγ˙ (mod [π dπ ]), which depends on x only through ωγ = ix γ γ˙ πγ˙ . kα kα˙ may be Plane wave gravitons (linearized spin-2 fields) of momentum pα α˙ = described by twistor functions h(Z ) = κ δ¯(2) ([π k]) exp ω k , h(Z ) = κ δ¯(−6) ([π k]) exp ω k , (32) where, √ for later use, we have taken all fluctuations to be proportional to the coupling κ = 16π GN and we follow [58,41] in defining [π α] r +1 ¯ 1 δ¯(r ) ([π k]) := ∂ . (33) [k α] [π k] In this definition, |α] is a fixed dotted spinor introduced so that the δ-function (0,1)-forms ¯ δ¯(r ) ([π k]) have homogeneity r in |π ]. On the support of the δ-function, πα˙ ∝ kα˙ so the momentum eigenstates (32) are in fact independent of the choice of |α]. Note that, ¯ because of the weight of the δ-function, h has weight −4 in the momentum spinor |k] (counting | k as weight −1), while h has weight +4. This is as expected for states of helicity −2 and +2, respectively. Likewise, the one-forms B may be taken to be B(Z ) = κ
dω β δ¯(−5) ([π k]) exp ω k , β k
(34)
| arises from the gauge freedom (28) in the definiwhere the constant undotted spinor β | is arbitrary provided β tion of B. The choice of β k = 0 reflecting the gauge freedom (28). It is easy to check that I (d B) = h, with h as above in (32).
840
L. Mason, D. Skinner
We remark in passing that V represents an element of H 1 (PT , TPT ) together with the extra requirement (25) that it be Hamiltonian with respect to I , while (incorporating ¯ + ∂m) B represents an element of H 1 (PT , 2 ⊗ O(−4)), the redundancy B → B + ∂χ cl 2 where cl is the sheaf of closed (2,0)-forms, together with the extra requirement that n[π dπ ] be taken equivalent to zero. Without the Hamiltonian and n[π dπ ] ∼ 0 conditions, these cohomology groups represent states in conformal gravity. The extra conditions eliminate half the conformal gravity spectrum, reducing it to Einstein gravity as above. The cohomology groups H 1 (PT , TPT ) and H 1 (PT , 2cl ⊗ O(−4)) (together with their N = 4 completions) define vertex operators in the Witten, Berkovits or heterotic twistor-string theories [35,36]. String theories that impose the extra conditions were constructed in [37], but these theories only seem to describe the asd interactions of Einstein (super)gravity [38].
3.2. The non-linear graviton. Penrose’s non-linear graviton construction [22] associates a deformed twistor space PT to a spacetime M with anti-self-dual (ASD) Weyl tensor. In this correspondence, the structure of M is encoded into the deformed complex structure of the twistor space. For ASD spacetimes that also obey the vacuum Einstein equations, the twistor space fibres over CP1 and admits an analogue of the Poisson structure I along the fibres. We can still describe such a PT using homogeneous coordinates (ωα , πα˙ ), where πα˙ are holomorphic coordinates that are homogenous coordinates for the CP1 base. As in PT , ωα parametrize the fibres of PT → CP1 , but in general they will no longer be holomorphic coordinates throughout the deformed twistor space. As in flat space, M is reconstructed as the space of degree-1 holomorphically embedded CP1 s inside PT . For some fixed x ∈ M, we will again denote the corresponding CP1 by L x . Although it will no longer have all the properties of a ‘straight line’, the normal bundle N L x |PT will still be O(1) ⊕ O(1) (as it was in the flat case) so that H 0 (L x , N L x |PT ) C4 , which is identified as the tangent space T M|x . Just as spacetime is no longer an affine vector space, PT is no longer isomorphic to the total space of N L x |PT . The correspondence may again be interpreted in terms of a double fibration P(S+ ) p PT
@ q R @ M
(35)
as in (23). For a half-flat spacetime M that is sufficiently close to flat spacetime M, the spin bundle is the product CP1 × M. The complex structure on PT may be described in terms of a finite deformation of ¯ the flat background ∂-operator: ∂¯ → ∂¯ + V = ∂¯ + I (dh, · )
(36)
¯ with I (dh, · ) as in Eq. (25). Only allowing Hamiltonian deformations of the ∂-operator ensures that PT also fibres over CP1 and has a holomorphic Poisson structure I on the fibres. This will be essential in the construction of the spacetime metric below. The ¯ deformed ∂-operator defines an integrable almost complex structure if and only if the Nijenhuis tensor N = (∂¯ + V )2 ∈ 0,2 (PT , TPT )
(37)
Gravity, Twistors and the MHV Formalism
841
vanishes. For Hamiltonian deformations (36), one finds [39] N = 0 if ¯ + 1 {h, h} = 0. ∂h 2
(38)
There is a ‘Poisson diffeomorphism’ freedom generated by Hamiltonians χ which are smooth functions of weight two, because changing ¯ + {h, χ } h → h + ∂χ
(39)
does not alter the complex structure. The diffeomorphism freedom can be fixed by requiring h to be holomorphic in ωα and proportional to π¯ dπ¯ , so that its (0,1)-form is purely along the base of the fibration PT → CP1 . Any such h automatically leads to a vanishing Nijenhuis tensor. This gauge condition is natural in a scattering theory context, being essentially the same condition as is utilised in Newman’s formulation of the nonlinear graviton [23,32,56]. In Newman’s formulation (which will not be emphasized here), the holomorphic lines L x are obtained from lightcone cuts of (complexified) null infinity CI and can thus be reconstructed simply from the asymptotic data of the spacetime M, while h is interpreted as an integral of the asymptotic shear (the asymptotic characteristic data of M). Requiring that h be holomorphic in ωα and proportional to π¯ dπ¯ does not completely fix the gauge freedom (39). In Newman’s picture, the remaining freedom is fixed by additionally requiring that h depends on ωα only through ω π¯ . We will implicitly use ‘Newman gauge’ in what follows: in particular, the twistor representatives of momentum eigenstates introduced in Eq. (32) are adapted to Newman gauge. As mentioned above, each point x ∈ M corresponds to a holomorphically embedded CP1 denoted by L x . The flat space incidence relation ωα = ix α α˙ πα˙ must be generalized, because ωα is no longer a globally holomorphic coordinate on PT . We thus represent L x ⊂ PT by the deformed incidence relation ωα = F α (x, π ),
(40)
where F α has homogeneity one in πα˙ . The condition that L x be holomorphic with respect to the deformed complex structure (36) is
0 = (∂¯ + V )(ωα − F α (x, π )) L = V α L − ∂¯ F α (x, π ), (41) x
x
so that we obtain the condition ∂¯ F α (x, π ) = V α (F α (x, π ), π ),
(42)
see Fig. 2. The restriction of V α to L x means that we set ωα = F α (x, π ) in V , so that (42) is a nonlinear differential equation for F α . This generally makes it very difficult to find explicit expressions for the holomorphic curves. As in PT , for fixed x the curve L x ⊂ PT defined by (42) is a section of the fibration PT → CP1 , holomorphic with respect to the deformed complex structure, and has normal bundle N L x |PT O(1) ⊕ O(1). The deformation theory of Kodaira & Spencer implies that the family of lines in PT survive small deformations of the complex structure and form a four parameter family. Thus there will be a four parameter space of solutions to the nonlinear equation (42) and it is this parameter space that we identify with M.
842
L. Mason, D. Skinner
Fig. 2. Deformations of the complex structure induce deformations of the holomorphic curves. Identifying the four parameters x on which F α (x, π ) depends with spacetime coordinates, the normal vector α α α ˙ α F − ix πα˙ ∂/∂ω on L x connects the original twistor line ωα = ix α α˙ πα˙ to the deformed curve
3.2.1. Constructing the spacetime metric The space of degree one curves is naturally endowed with a conformal structure by requiring two points x, y ∈ M to be connected by a null geodesic if their corresponding ‘lines’ L x , L y ⊂ PT intersect. Let us now show explicitly how to use the twistor data to construct a spacetime metric [22]. Consider the (weighted) 1-forms [π dπ ] and dωα − V α . These forms are annihilated by contraction with the antiholomorphic vector fields of the deformed complex structure, and so define a basis of holomorphic forms on PT . Note that the holomorphic form [π dπ ] is unaltered compared to PT ; this is a consequence of restricting to Hamiltonian complex structure deformations in (36). The holomorphic 3-form of weight +4 is therefore PT = [π dπ ]∧(dωα − V α )∧(dωα − Vα ). Pulling back PT to P(S+ ) (i.e. imposing the incidence relation ωα = F α (x, π )) gives p ∗ PT = [π dπ ] ∧ p ∗ (dωα − V α ) ∧ p ∗ (dωα − Vα ) = [π dπ ] ∧ dx F α ∧ dx Fα ,
(43)
where dx denotes the exterior derivative on P(S+ ) holding πα˙ constant, i.e. dx = dx a ∂/∂ x a . (Possible terms in dπ¯ vanish by virtue of the holomorphy of these sections, while terms in dπ vanish by virtue of the fact that the expressions are wedged against [π dπ ].) The requirement (42) that L x ⊂ PT is a holomorphic line ensures ∂¯ F α (x, π ) = V α (F(x, π ), π ) so
∂¯ dx F α ∧ dx Fα = 2 dx (∂¯ F α ) ∧ dx Fα = 2 ∂β V α L dx F β ∧ dx Fα , (44) x
where in the second term we used the fact that V α depends on x only through F β (x, π ). The wedge product implies this expression is antisymmetric in α, β and so in fact it vanishes because V is Hamiltonian. Therefore dx F α ∧ dx Fα is a two-form of homogeneity +2 in πα˙ that is holomorphic along each CP1 . Consequently, by Liouville’s theorem, ˙
p ∗ PT = −[π dπ ] ∧ q ∗ α˙ β (x)πα˙ πβ˙ , ˙
(45)
where α˙ β ∈ 2 (M, Sym2 S+ ) are three spacetime two-forms, pulled back to P(S+ ) by q ∗ . (The minus sign is for convenience.) We drop the pullback symbol q ∗ in what follows.
Gravity, Twistors and the MHV Formalism
843
˙
˙
α˙ β is automatically closed on spacetime, because α˙ β πα˙ πβ˙ = dx F α ∧ dx Fα . The discussion around Eq. (8) then shows that the spacetime M is necessarily anti-self-dual. ˙ Moreover, α˙ β is simple by construction, so ˙
˙
πα˙ πβ˙ α˙ β = πα˙ eα α˙ ∧ πβ˙ eαβ
(46)
for some tetrad eα α˙ . This decomposition does not uniquely fix the tetrad: we can freely replace eα α˙ by αβ eβ α˙ for αβ (x, π ) an arbitrary element of S L(2, C), as any such αβ drops out of Eq. (46). Comparing definitions shows that p ∗ (dωα − V α ) = dx F α = iαβ e
mod [π dπ ], β β˙
πβ˙
mod [π dπ ].
(47)
Equations (46) & (47) generalize the flat spacetime formulæ7 ˙
p ∗ (dωα ∧ dωα ) = −dx α α˙ ∧ dxαβ πα˙ πβ˙ p ∗ dωα = i dx α α˙ πα˙
mod [π dπ ] mod [π dπ ]
(48)
arising from the incidence relation ωα = ix α α˙ πα˙ in PT . In (47), a choice of αsβ fixes a choice of spin frame (for the undotted spinors) and hence a choice of tetrad eα α˙ . However, although αβ (x, π ) has weight zero in πα˙ , generically it is not π -independent. Because of this, it is not simply a local Lorentz transform on spacetime, but is best thought of as a holomorphic frame8 trivializing N L x |PT ⊗ O(−1) over L x (see also [25,26]). Note that since the normal bundle N L x |PT O(1) ⊕ O(1), the bundle N L x |PT ⊗ O(−1) is indeed trivial on L x . Its
space of global holomorphic sections H 0 (L x , O ⊕ O) C2 is precisely the fibre S− x of the bundle of anti-self-dual spinors on M. 4. Gravitational MHV Amplitudes from Twistor Space We now provide a twistorial description of h 2 |h 1 by translating the right-hand side of (21) using the Penrose integral transform. Finally, we will use the twistor description to expand around Minkowski space in plane waves, thus recovering the standard form of the MHV amplitudes. Underlying much of what follows is a presentation for the twistor data, going back to Newman [32], that relates directly to the asymptotic data at I for the fields involved. By using sums of momentum eigenstates for the data at I we guarantee that the fields and backgrounds that we work with are asymptotically superpositions of plane waves. Technically, ASD spacetimes constructed in this way are not asymptotically flat along the directions of the plane waves. It is nevertheless possible to incorporate them into an asymptotically flat formalism at the expense of having to consider δ-function singularities in the asymptotic data (the asymptotic shear) as already apparent in (32) for the twistor representatives. 7 Strictly, Eq. (48) also include a α in the definition of p ∗ dωα . Such a relates the twistor coordinate β index on ωα to the undotted spacetime spinor index on d x α α˙ . On a flat background these indices can be
identified directly. 8 α (x, π ) is thus somewhat analogous to the choice of holomorphic frame H (x, π ) that arises in a similar β context for Yang-Mills, see Eq. (111)–(114).
844
L. Mason, D. Skinner
In Sect. 2.2, the classical amplitude for a positive helicity graviton to cross an asymptotically flat ASD spacetime and emerge with negative helicity was shown to be
i γ˙ α˙ β˙ h n h 1 = 2
0 ∧ γn α˙ ∧ γ1 β˙ γ˙ , (49) κ M α˙ β˙
where 0 is formed from the tetrad of the half-flat background and γ1,n are two linearized self-dual connections that are on-shell with respect to the linearized field equations (10). (The labelling 1, n is for later convenience.) We seek a twistorial interpretation of this term. Firstly, the Penrose transform (30) of the linearized self-dual spin connection 1-form
γ αβ˙˙ = 2 [π dπ ] ∧ π α˙ πβ˙ p ∗ (B) (50) Lx
also makes sense on an ASD background. To see this, first recall from Sect. 2.2 that the background self-dual spin connection is flat on an ASD spacetime. It is therefore at most pure gauge and can be taken to vanish. The space of dotted spinors is then globally trivialized both on spacetime and on twistor space, so there is no difficulty in adding π α˙ πβ˙ at different points of L x in (50). As in Eq. (30) for flat space, the Penrose transform (50) is the pullback of the (2, 1)-form [π dπ ] ∧ π α˙ πβ˙ B to P(S+ ), pushed down to the ASD q
spacetime (i.e. integrated over the CP1 fibres of P(S+ ) → M). To see that this pushdown is well-defined, note that for any vector field X on M, there is a unique vector field X ∈ T P(S+ ) that obeys X [π dπ ] = 0 and whose projection q∗ ( X ) to T M is again α ˙ ∗ X . So for any such X , the integral 2 L x X [π dπ ] ∧ π πβ˙ p B is well-defined and
equal to X γ αβ˙˙ . Hence the integral (50) is also unambiguous. In particular, if {∇γ γ˙ } is a basis of T M dual to the tetrad, the components of the spin connection in this basis are given by contracting (50) with ∇γ γ˙ :
α˙ α˙ [π dπ ] ∧ π α˙ πβ˙ πγ˙ Bα (F, π )αγ (x, π ), (51) (γγ γ˙ ) β˙ := ∇γ γ˙ γ β˙ = 2 Lx
γ γ˙ p ∗ B. The holomorphic frame α trivializes where we have used (47) to evaluate ∇ β the anti-self-dual spin bundle over L x , thus allowing us to make sense of the integral of the indexed quantity9 Bα . To construct the Penrose transform of the expression for h n |h 1 , we extract the components of each γ to obtain
i i γ˙ γ γ˙ α˙ β˙ α˙ β˙
∧ γ ∧ γ = dµ γn γ1 γ γ˙ α˙ β˙ , (52) ˙ 1 β γ˙ 0 n α˙ 2 2 κ M 2κ M α˙ β˙
where dµ := 0 ∧ 0 α˙ β˙ is the volume form on M. Using the Penrose transform (51) in Eq. (49) gives
2i h n h 1 = 2 dµ [πn dπn ][π1 dπ1 ] Bn α (F, πn )αγ (x, πn ) κ M×CP1 ×CP1 B1 β (F, π1 )βγ (x, π1 )[πn π1 ]3 ,
(53)
9 A similar rôle is played by the holomorphic frame H in the Penrose transform of a background coupled self-dual Yang-Mills field, see (116).
Gravity, Twistors and the MHV Formalism
845
where M × CP1 × CP1 is the fibrewise product of P(S+ ) with itself. The spinors |π1 ] and |πn ] label to two copies of the CP1 fibres. This formula currently describes the scattering of two positive helicity gravitons off a (fully non-linear) ASD background spacetime. In order to obtain the BGK amplitudes, we must expand the background spacetime M around Minkowski space M. In principle, this can be done by iterating deformations of the twistor space caused by adding in negative helicity momentum eigenstates, and keeping track of the holomorphic degree-1 curves to construct the function F α (x, π ) explicitly (see [31,57] for a discussion along these lines). In practice, constructing F α in this way is complicated, and the difficulties are compounded by having to expand all the terms in (53). Instead, motivated by an analogous step at the same point in the Yang-Mills calculation (equation (118)), we seek a coordinate transformation of the spin bundle P(S+ ) → M that simplifies our task. The desired coordinate transformation takes the form (x α α˙ , πβ˙ ) → (y α α˙ (x, π ), πβ˙ )
such that
iy α α˙ πα˙ = F α (x, π ),
(54)
and may be viewed as a π -dependent coordinate transformation of M. Equation (54) replaces F α by iy α α˙ πα˙ , so that from the point of view of the (y, π ) coordinates, we never need face the complicated problem of constructing F α (x, π ) explicitly! The price to be paid for this seemingly magical simplification is that generically, the CP1 fibres of P(S+ ) → M do not coincide with those of P(S+ ) → M where here the ys are taken to be coordinates on M; in other words, the CP1 s of constant x (the twistor lines in PT ) are not the same as the CP1 s of constant y (the twistor lines in PT ). There is some freedom in the definition of y α α˙ in (54). One natural choice that fits the bill is y α α˙ (x, π ) = i
F α (x, ξ )π α˙ − F α (x, π )ξ α˙ , [ξ π ]
(55)
where ξ α˙ is an arbitrary constant spinor. Note that if the background is actually flat, then F α = ix α α˙ πα˙ and we have simply y α α˙ = x α α˙ . Also note that the numerator vanishes at |π ] = |ξ ], so the apparent singularity when [ξ π ] = 0 is removable. Hence y α α˙ is smoothly (but not holomorphically) defined, and (y, π ) are good coordinates on P(S+ ), at least when the departure from flat spacetime is not too severe. Equation (55) explicitly shows that the CP1 s of constant x do not coincide with those of constant y, because y varies as we move along a CP1 fibre L x . β β We now pick a spacetime spin frame by requiring α (x, ξ ) = εα . Then, using Eq. (47), the Jacobian of the coordinate transformation (55) with the spacetime tetrad (x) ∇α α˙ is found to be 1 ˙ ˙ ˙ (x) ∇α α˙ y β β = −iαβ (x, π )πα˙ ξ β − εαβ ξα˙ π β . (56) [ξ π ] β
This Jacobian has unit determinant because α ∈ S L(2, C), so dµ = d4 y (mod [π dπ ]). Furthermore, we see that ∂ ∂ ˙ (x) (x) = π α˙ α α˙ , (57) π α˙ ∇α α˙ = π α˙ ∇α α˙ y β β ˙ β β ∂y ∂y which will be used in what follows. We are not quite ready to put this coordinate transformation to use, because our expression (53) is written as an integral over the fibrewise product of the spin bundle with itself, rather than just as an integral over P(S+ ). Since (54) does not map the
846
L. Mason, D. Skinner
fibres of P(S+ ) → M to the fibres of P(S+ ) → M, if the coordinate transformation is given by say y(x, π1 ), the πn integral in Eq. (53) will not hold y constant for fixed (x, π1 ). We will deal with this by reformulating this second fibre integral as an inverse ¯ of the ∂-operator up the CP1 fibres of P(S+ ) over M and perturbing about the fibres of constant y. We can understand ∂¯ −1 as follows. First recall that on a single CP1 , any (0,1)-form ¯ is automatically ∂-closed for dimensional reasons. The cohomology groups H 0,1 (CP1 , O(k)) vanish for k ≥ −1 and so any (0,1)-form of homogeneity k ≥ −1 is necessarily ¯ ∂-exact. Thus, if ν ∈ 0,1 (P1 , O(k)) with k ≥ −1 then ∂¯ −1 ν makes sense and is an element of 0 (P1 , O(k)). When k ≥ 0, ∂¯ −1 ν is not uniquely defined because we can add to ∂¯ −1 ν a globally holomorphic function ρ of weight k, since ∂¯ ∂¯ −1 ν + ρ = ν. However, there are no global holomorphic functions of weight −1, so when k = −1, ∂¯ −1 ν is unique. Explicitly, in terms of homogeneous coordinates πα˙ on the CP1 , one takes10
[π1 dπ1 ] 1 −1 ∧ ν(π1 ) , ν := (58) ∂¯21 2π i CP1 [π2 π1 ] which is indeed a 0-form of weight −1 in π2α˙ . Taking ∂¯ (with respect to |π2 ]) of both sides shows that ∂¯ ∂¯ −1 = 1, because the only π2 -dependence on the right is from the homogeneous form 1/2π i[π2 π1 ] of the standard Cauchy kernel for ∂¯ −1 . (In affine coordinates z on the Riemann sphere, π = (1, z) and [π2 π1 ] = z 1 − z 2 .) To exploit this in our situation, first use (47) & (51) to rewrite (53) as
2i dµ [πn dπn ][π1 dπ1 ] Bn α (F, πn ) h n h 1 = 2 κ M×CP1 ×CP1 γ˙ 2 . (59) ×αγ (x, πn )πn ∇ γ γ˙ B1 (F, π1 )[πn π1 ] β˙
γ˙
Next, note that p ∗ B1 π1α˙ π1 π1 has weight −1 in |π1 ] and is a (0, 1)-form on (the second) CP1 (valued also in T ∗ M ⊗ Sym3 S+ ). We then define
[π1 dπ1 ] 1 β˙ γ˙ −1 α˙ β˙ γ˙ ¯ ∧ B(F, π1 ) π1α˙ π1 π1 , := (60) x ∂n1 B π π π 2π i CP1 [πn π1 ] where the presubscript x emphasizes the fact that in this formula, ∂¯ involves the (0, 1)˙ vector tangent to the CP1 fibres q −1 (x). As above, this defines x ∂¯ −1 (B π α˙ π β π γ˙ ) uniquely. Using this in Eq. (59) allows us to rewrite that equation as
4π α α˙ x ∂¯ −1 B1 [πn π1 ]3 , (61) dµ [πn dπn ] αβ Bn β (F, πn )πnα˙ ∇ h n h 1 = − 2 n1 κ P(S+ ) now interpreted as a (two-point) integral over the projective primed spin bundle. We can now use the coordinate transformation to simplify the integral (61). Transforming to the (y, π ) coordinates using Eqs. (54) & (57) gives
∂ 4π −1 h n h 1 = − 2 B1 (y, π1 )[πn π1 ]3 , d4 y[πn dπn ] Bnα (y, πn )πnα˙ α α˙ x ∂¯n1 κ P(S+ ) ∂y (62) 10 The numerical prefactor 1/2π i of course involves the ratio of a circle’s area to its radius, rather than the coordinate πα˙ .
Gravity, Twistors and the MHV Formalism
847
now written as an integral on the spin bundle over flat spacetime. It remains to reformulate the operator x ∂¯ −1 , the inverse of the ∂¯ −1 operator on the CP1 s of constant x, in ¯ ¯ terms of y ∂¯ −1 the inverse of the ∂-operator on the CP1 s of constant y. These ∂-operators 1 are essentially just the antiholomorphic tangent vector to the CP s of constant x or y and the relationship between them follows by the chain rule. Using Eqs. (42) & (55) we find x ∂¯
¯ α α˙ ) = y ∂¯ + (∂y
∂ p ∗ (V α )ξ α˙ ∂ = y ∂¯ − i , α α ˙ ∂y [ξ π ] ∂y α α˙
(63)
where the extra term is the difference between an anti-holomorphic vector field tangent to the fibres of P(S+ ) → M and the anti-holomorphic vector field tangent to the fibres of P(S+ ) → M. Consequently, we see that11 (somewhat formally) 1 1 , = ¯ ¯ ∂ ∂ + LV x y
(64)
¯ where the right-hand side of this equation involves the ∂-operator along the CP1 fibres in the (y, π ) coordinates, together with the Lie derivative LV along the vector field ∗ α α˙ := −i p (V )ξ ∂ . V [ξ π ] ∂y α α˙
(65)
= V to from its pushdown p∗ V (We will often abuse notation by not distinguishing V twistor space.) The Lie derivative takes account of the fact that this operator acts on the form Bα dy α α˙ πα˙ ; both the components Bα and basis forms dy α α˙ depend on y. The operator (∂¯ + LV )−1 may be computed through its expansion 1 1 1 1 1 1 1 = − LV + LV LV − · · · , ¯∂ + LV ¯∂ ¯∂ ¯∂ ∂¯ ¯∂ ∂¯
(66)
¯ where all the inverse ∂-operators now imply an integral over the CP1 s at constant y α α˙ (the holomorphic lines in PT ). We have
∞ ∂ n+1 4π
hn h1 = (−) d4 y [πn dπn ] Bnα πnα˙ α α˙ κ 2 ∂y n=2 1 1 1 1 3 (67) LVn−1 · · · LV2 B1 [πn , π1 ] . ∂¯ ∂¯ ∂¯ ∂¯ ¯ The inverse ∂-operators always act on sections of 0,1 (CP1 , O(−1) ⊗ T ∗ M) and so are canonically defined as in Eq. (60), although here it is y rather than x that is being held point in the y-direction, the Lie derivatives may be constant. Because the vector fields V ¯ brought inside all the CP1 integrals, effectively commuting with the inverse ∂-operators. So the n th -order term in the expansion is
n 4π [πi dπi ] α α˙ ∂ in (n) 4 B π y Mtwistor := d n−2 2 (2π ) κ [π π ] n n ∂y α α˙ i=1 i+1 i LVn−1 · · · LV2 B1 [πn , π1 ]4 , (68) 11 Equation (63) is analogous to the Yang-Mills equation A = −∂¯ H H −1 , while Eq. (64) is analogous to H ∂¯ −1 H −1 = 1/(∂¯ + A) used in Appendix B.2.
848
L. Mason, D. Skinner
Fig. 3. Yang-Mills (l) and gravitational (r) MHV amplitudes are supported on holomorphic lines in twistor space. For gravity, the negative helicity gravitons arise from insertions of normal vector fields, giving a perturbative description of the deformation of the line
n where we have compensated for the fact that the integration measure i=1 [πi dπi ]/ [πi+1 πi ] includes an extra factor of 1/[πn π1 ] by increasing the power of [πn π1 ] in the numerator. In this expression the n-point amplitude comes from an integral over the space of lines twistor space, with n insertions on the line, each of whose insertion point is integrated over. This is exactly the same picture as described in the appendix for Yang-Mills. For gravity, the n − 2 vector fields differentiate the wavefunctions (as we will see explicitly later), leading to what is sometimes called ‘derivative of a δ-function’ support (see Fig. 3). Our final task is to evaluate this expression when the external states are each the plane waves of (32), (34). From (25), the associated twistor space vector fields are α V (Z ) = κ δ¯(1) ([π k]) eω k k ∂/∂ωα , so the vector fields on P(S+ ) become V (y, π ) = −iκ δ¯(1) ([π k]) exp (i p · y)
k α ξ α˙ ∂ , [ξ π ] ∂y α α˙
(69)
using Eq. (65). Pulling the plane wave formula (34) for B back to P(S+ ) gives B = iκ
|dy|π ] β δ¯(−5) ([π k]) exp (i p · y) β k
(70)
in the (y, π ) coordinates. To evaluate (68), use the Cartan formula LV = V d + dV to replace LVn−1 . The second term in Cartan’s formula leads to a contribution Bnα πnα˙
∂ ∂ d Vn−1 LVn−2 · · · LV2 B1 = Bnα πnα˙ α α˙ Vn−1 LVn−2 · · · LV2 B1 (71) α α ˙ ∂y ∂y
to the integrand of (68). On the right-hand side, Bnα πnα˙ ∂/∂y α α˙ simply differentiates the scalar Vn−1 LVn−2 · · · LV2 B1 . Because Bn is pulled back to P(S+ ) from twistor space, it depends on y only through y α α˙ πα˙ , so Bnα may be brought inside the πnα˙ ∂/∂y α α˙ derivative. Hence (71) is a total derivative and may be discarded. Now, using the fact that [d, LV ] = 0 for any vector field V , the remaining terms involve ∂ Vn−1 dLVn−2 · · · LV2 B1 α α ˙ ∂y ∂ h 1 γ γ˙ ˙ , (72) = Bnα πnα˙ α α˙ Vn−1 LVn−2 · · · LV2 dy ∧ dyγδ π1γ˙ π1δ˙ ∂y 2
Bnα πnα˙
Gravity, Twistors and the MHV Formalism
849
β˙ where we have used dB = h/2 dy α α˙ ∧ dyα πα˙ πβ˙ , which again follows because B is pulled back from a field on twistor space. The key simplification that allows us to evaluate (72) comes from making the gauge choice |ξ ] = |n], where |n] is the dotted momentum spinor of the positive helicity graviton represented by Bn . With this choice, the two-form LVn−2 · · · LV2 h 1 /2 d β ˙ ˙ y γ γ˙ ∧ dyγδ π1γ˙ π1δ˙ is contracted into the bi-vector Bnα πnα˙ Vn−1 n β ∂/∂y α α˙ ∧ ∂/ ˙ ∂y β β . But the momentum eigenstate Bn has support only when |πn ] = |n], so this bi-vector is purely self-dual: B
α
πnα˙
β
˙
Vn−1 n β
∂ ∂ 1 Bn Vn−1 α˙ β˙ ∂ ∂ πn n ∧ = ∧ . ˙ ˙ β β β β [n πn−1 ] ∂y α α˙ 2 [n π ] ∂yβα˙ ∂y ∂y n−1
(73)
It is straightforward to check that because the vectors Vi are Hamiltonian, with our gauge β˙ ˙ choice, the bi-vector πnα˙ n β ∂/∂y α α˙ ∧ ∂/∂yα commutes with all the remaining Lie deriv˙ atives. Therefore, we may immediately contract this bivector with dy γ γ˙ ∧ dyγδ π1γ˙ π1δ˙ to obtain (n)
Mtwistor=
n 2π [πi dπi ] Bn Vn−1 in 4y h 1 )[πn π1 ]5 [n π1 ]5 , d Vn−2 · · · V2 ( n−2 2 [π (2π ) κ i+1 πi ] [n πn−1 ] i=1
(74) where the remaining vector fields V2 to Vn−2 act simply by differentiating everything to their right. To take account of the possible orderings of the external states, we insert Vm = κ
n−1
i δ¯(1) ([π i]) ei pi ·y
i=2
i α n α˙ ∂ [n πm ] ∂y α α˙
(75)
β˙
for each vector field Vm at (y α α˙ , πm ), where the i are expansion parameters labelling the physical external states. (We use the shorthand piα α˙ = i α i α˙ .) Extracting the coefficient n−1 ¯ of i=2 i and using the δ-functions to integrate over the n insertion points gives the n-particle MHV amplitude as n κ n−2 (4) (n) δ pi [1 n]8 Mtwistor = i=1 n−2 β n − 1 1 k| pk−1+pk−2 +· · ·+ p2 + p1 |n] + P(2,...,n−1) , × n[n−1 n][n 1]2 C(n) [n k] β k=2
(76) where P(2,...,n−1) is a sum over permutations of the vector fields. Consider the first (displayed) permutation. This is the same as the first term in M(n) in Eq. (1), except for a factor n − 1 | pn−1 |1] β β × [1 n − 1] = − . | pn |1] β n[n 1] β
(77)
850
L. Mason, D. Skinner
This factor is independent of 2, . . . , n − 2, permuting the first term over gravitons 2 to n − 2 will yield the same factor times the corresponding permutation of (1). Therefore we have M(n) twistor = −
| pn−1 |1] (n) β M + other perms. | pn |1] β
(78)
The remaining permutations in (78) involve exchanging graviton n − 1 with each of gravitons 2 to n − 2. But since M(n) is equal to the standard BGK amplitude (as proved in Appendix A), we know (e.g. from Ward identities [60]) that it is in fact symmetric under exchange of any two like-helicity gravitons. Hence each term is proportional to M(n) and we are left with an overall factor −
n−1 | p2 + p3 + · · · + pn−1 |1] β β | pi |1] =− = 1. | pn |1] | pn |1] β β i=2
(79)
(n)
(n) , and that M Thus we have shown that (76) is really independent of β twistor = M . It is remarkable that the infinite series of n-particle MHV amplitudes may be constructed by expanding the square of the self-dual spin connection on an anti-self-dual spacetime (49).
5. A Twistor Action for MHV Diagrams in Gravity According to the MHV diagram formalism, initiated in [58] for Yang-Mills and [2,13,14] for gravity, one can recover the full perturbation theory by continuing the MHV amplitudes off-shell and connecting them together using propagators connecting positive and negative helicity lines12 . The MHV diagram formalism was first developed in the context of the ‘disconnected prescription’ of twistor-string theory [58], but soon after it was realized that one could also construct actions whose Feynman diagrams generate the Yang-Mills MHV diagram formalism [33,34,41,45,46,62,63]. We now give a twistor action whose perturbation theory generates the MHV diagram formalism for gravity. In Section 3.2, ASD spacetimes were reformulated in terms of deformed twistor spaces by the nonlinear graviton construction [22]. The field equation on twistor space is the vanishing of the Nijenhuis tensor 1 IJ ¯ N = I ∂ J ∂h + {h, h} (80) 2 so that the almost complex structure ∂¯ + I (dh, · ) is integrable and PT is a complex threefold, obtained as a deformation of PT (see the discussion around Eq. (36)). In [39], a local twistor action whose field equations include the condition N = 0 was constructed. The action is written in terms of a field13 h ∈ 0,1 (PT , O(2)) and h ∈ 0,1 (PT , O(−6)), 1,1 although we also here use B ∈ (PT , O(−4)). It takes a ‘BF’-like form
1 1 IJ ¯ ¯ S= ∧ I B I ∂ J ∂h + {h, h} = − ∧ h ∂h + {h, h} , (81) 2 2 PT PT 12 At the quantum level, this program works as stated only for supersymmetric theories [61]. 13 We abuse notation by not distinguishing the (0,1)-forms h, h from their cohomology classes.
Gravity, Twistors and the MHV Formalism
851
where = I J K L Z I dZ J ∧ dZ K ∧ dZ L /4! is the canonical holomorphic 3-form of weight +4, I I J is the Poisson structure introduced in Eq. (26) and { · , · } its associated Poisson bracket. Note that this Poisson bracket has weight −2 so that the action is well-defined on the projective space. In the first version, B plays the rôle of a Lagrange multiplier ensuring the vanishing of the Nijenhuis tensor. The second form follows upon integration by parts. In the second form, the field equations of this action are ¯ + 1 {h, h} = 0 ∂h 2
and
∂¯h h = 0,
where
∂¯h f := ∂¯ f + {h, f }.
(82)
We also have the gauge freedom h → h + ∂¯h χ , h → h + ∂¯h χ . In the linearized theory, these imply that on-shell, h and h are representatives of the cohomology classes used to describe linearized gravitons of helicities ±2 in the Penrose transform, as reviewed in Sect. 3.1. The gauge freedom may be fixed by using14 ‘CSW gauge’ [34,58]: choose an antiholomorphic vector field η¯ tangent to the fibres of PT → CP1 and impose the axial gauge condition that η¯ h = 0 and η¯ h = 0. Imposing this gauge in (81), the cubic vertex vanishes and one is left with an off-diagonal kinetic term and a linear theory. The other main ingredient in the MHV diagram formulation is the infinite set of MHV vertices: off-shell continuations of the MHV amplitudes. Using coordinates (y, π ) for the spin bundle P(S+ ) → M over Minkowski space, it follows from the previous section that in the twistor formulation these vertices arise from the expansion of
1 4 α α˙ ∂ 3 d y ∧ [πn dπn ] ∧ B (y, πn )πn α α˙ B(y, π1 )[πn π1 ] , (83) ∂y ∂¯ + LV P(S+ ) where we interpret B as the pullback to P(S+ ) of an arbitrary element of 1,1 (PT , is here interpreted as in O(−4)) (i.e., not necessarily obeying ∂¯ B = 0). Likewise, V (65): ∗ α α˙ ∂ = −i p (V )ξ V [ξ π ] ∂y α α˙
where
Vα =
∂h . ∂ωα
(84)
The inverse operator 1/(∂¯ + LV ) is again understood through its infinite series expansion (66) leading to an infinite sequence of MHV vertices. These only involve the components of the (0,1)-forms B and h that are tangent to the CP1 base of PT → CP1 . The choice of the vector field η¯ corresponds to the choice of spinor used by [58]. As described for Yang-Mills in [41], it enters into the definition of the propagator which gives the CSW rule for extending the MHV amplitudes off-shell. The discussion in [41] applies here directly with just a shift in homogeneities. Therefore, treating h and B as the fundamental fields, in CSW gauge, the Feynman diagrams of the action
1 IJ ¯ S[B, h] = ∧ I B I ∂ J ∂h + {h, h} 2 PT
1 ∂ + d4 y ∧ [πn dπn ] ∧ B α (y, πn )πnα˙ α α˙ B(y, π1 )[πn π1 ]3 ∂y ∂¯ + LV P(S+ ) (85) reproduces the MHV diagram formulation of gravity. 14 The Newman gauge of Sect. 3.2 implies CSW gauge, but also enforces other conditions appropriate only when the fields are on-shell.
852
L. Mason, D. Skinner
6. Supergravity Supertwistor space PT[N ] is the projectivisation of C4|N , where we have adjoined N anticommuting homogeneity degree 1 coordinates ψ A , A = 1, . . . , N . In Penrose conventions, the space of holomorphic lines in PT[N ] is anti-chiral superspace M[N ] with coordinates (x α α˙ , θ Aα˙ ), where θ Aα˙ are anti-commuting. The flat space incidence relation (40) is augmented to ωα = ix α α˙ πα˙
ψA = θ Aα˙ πα˙ .
(86)
The linear Penrose transform of Sect. 3.1 extends [64] to one between cohomology classes on PT[N ] and superfields on M[N ] . In particular, h naturally extends to an (onshell) superfield15 H ∈ H 1 (PT[N ] , O(2)) that is holomorphic in ψ i . That is, H has component expansion H(Z , ψ) = h(Z ) + ψ A λ A (Z ) + · · · + (ψ 1 ψ 2 · · · ψ N )φ(Z ),
(87)
where the coefficient of (ψ)k may represented by a (0,1)-form on the standard twistor space PT and has homogeneity 2 − k. H generates Poisson deformations of the complex structure of the twistor superspace through its associated Hamiltonian vector superfield I (dH, · ) [39]. As a superfield it represents on-shell spacetime fields of helicities −2, − 23 , . . . , −2 + N2 . When N < 8, the conjugate graviton supermultiplet is repre ∈ H 1 (PT , O(N −6)). As in the non-supersymmetric sented by a twistor superfield H [N ] may equivalently be represented by a superfield B ∈ H 1 (PT[N ] , 1 (N − 4)), case, H modulo the gauge equivalence B → B + dm(Z , ψ) + n(Z , ψ)[π dπ ]. A particularly interesting case is N = 4, for which twistor space is a CalabiYau supermanifold, i.e., it admits a global holomorphic volume (integral) form. The Calabi-Yau property singles out N = 4 twistor space as a natural target for a string theory [40]. When N = 4, B(Z , ψ) has homogeneity zero and there is a natural extension of the action (85):
1 4 IJ ¯ SN =4 = d ψ ∧ I B I ∂ J ∂H + {H, H} 2 PT[4]
Bnα ξ α˙ ∂ 1 4|8 + d x ∧ [πn dπn ] ∧ B , (88) [ξ πn ] ∂ x α α˙ ∂¯ + LV P(S+[4] ) where in the second term, B is pulled back to the superspace spin bundle P(S+[4] ) and V is the vector field on P(S+[4] ) defined by (65) (or (84)) with V replaced by V. In this N = 4 ¯ formula, the inverse ∂-operators act on (0,1)-forms of vanishing weight. Although ∂¯ −1 is not obstructed on such forms, it is ambiguous. The freedom can be fixed by adding a −1 B vanishes when |π2 ] = |ξ ]. With this choice, constant so that ∂¯21
[π1 dπ1 ] [ξ π2 ] 1 −1 B(x, θ , π1 ) B(x, θ , π2 ) := (89) ∂¯21 2π i CP1 [π2 π1 ] [ξ π1 ] which has homogeneity zero in |π2 ] and satisfies ∂¯ ∂¯ −1 B = B; the integrand in the second term of (88) then has vanishing weight in each CP1 and is thus well-defined. It is 15 PT [N ] is a split supermanifold, whose cohomology is generated by that of the base.
Gravity, Twistors and the MHV Formalism
853
easy to check that the truncation of (88) to N = 0 reproduces (85). As in [65], when B and V are on-shell with respect to the local N = 4 twistor action and are taken to be the twistor momentum eigenstates ∂ V(Z , ψ) = κ δ¯(1) ([π k]) exp ω kα α , k + ψ A ζ A (90) ∂ω β dω B(Z , ψ) = κ δ¯(−5) ([π k]) exp ω k + ψ A ζ A , (91) β k then the coefficients of the external Grassmann parameters ζ A in an expansion of the non-local term give the MHV amplitudes for arbitrary external members of the N = 4 supermultiplet. Although N = 4 twistor supersymmetry seems natural in twistor-string theory, N = 8 supergravity is usually thought of as more fundamental. The N = 8 graviton supermultiplet is CPT self-conjugate, and this fact has recently been argued to underlie many surprising simplifications in the S-matrix [9]. Thus, on twistor space, the complete multiplet is represented by a single superfield H(Z , ψ) = h(Z ) + ψ A λ A (Z ) + · · · + (ψ)8 h(Z ),
(92)
In the case that the external states are on-shell momentum eigenstates, represented on twistor space by the Newman gauge expression (90), the MHV scattering of arbitrary members of the N = 8 multiplet is described by the formula (n)
MN =8 =
P(S+[8] )
d4|16 x
n H1 Hn−1 [πi dπi ] Hn Vn−2 · · · V2 . [πi+1 πi ] [π1 πn ] [πn πn−1 ] [πn−1 π1 ] i=1
(93) Unlike the previous formulæ (68) & (88), this expression singles out three of the external fields, representing them in terms of the Hamiltonian function H rather than the vector field V. This is closely related to the formula obtained by Nair in [47]. It is easy to check that (93) reproduces the BGK amplitudes for external gravitons, and satisfies the supersymmetric recursion relations of [15]. 7. Conclusions and Future Directions A perspective of this paper has been that the MHV vertices provide a bridge between perturbative treatments of gravity and the fully nonlinear, non-perturbative structure that is such a key part of General Relativity. When we are on-shell with respect to the chiral action (81) (or the chiral limit of the Plebanski action), we may take advantage of the integrability of the anti-self-dual Einstein equations to interpret the infinite sum of MHV amplitudes as simply the square of a linearized fluctuation γ of the selfdual spin connection on the ASD background. The techniques of this paper, both for Yang-Mills and gravity indicate that it is possible to develop a background field formalism on fully nonlinear asd backgrounds within which explicit computations are tractable and generate amplitudes for processes with an arbitrary number of negative helicity legs. This programme would allow one to incorporate the integrability of the anti-self-dual Einstein equations into the study of perturbation theory in such a way as to bridge the gap between perturbative and non-perturbative treatments of gravity.
854
L. Mason, D. Skinner
The status of the MHV diagram formulation for gravity is currently less clear than that for Yang-Mills, although it has now been verified for up to 11 external particles [15]. At this stage there is no reason to doubt that the MHV picture for gravity should be successful, at least classically. The validity of our twistor action (85) for gravity currently depends on that of the MHV formalism whereas, in the case of Yang-Mills, the twistor action of [33,34] and reviewed in Appendix B provides an independent non-perturbative derivation of the MHV formalism [41]. A future goal is to construct a twistor action for gravity that works in the same way—for this it will be necessary to build a formalism in which the background is off-shell and PT possesses only an almost complex structure. A search for a spacetime MHV Lagrangian for gravity has been initiated in [44], following the path of [45] in Yang-Mills. N = 4 supergravity is not unique, and (88) is not the unique N = 4 completion of the non-supersymmetric action. Firstly, the Poisson structure I may also point along the fermionic directions, and in [39,66] this was shown to be responsible for gauged supergravities in the self-dual sector. Secondly, unlike the N = 4 completion of the MHV amplitudes in Yang-Mills, there seems to be no compelling reason that the nonlocal term in (88) should be only quadratic in B. It would be interesting to know if additional terms are required in the case of gauged supergravity. A key motivation for much of the work here is to reverse engineer a twistor-string ¯ theory for gravity. The Lie derivatives and inverse ∂-operators in the second term in the action (88) are suggestive of a worldsheet OPE interpretation, and it would be fascinating to see if this term (taken on-shell) can arise as an instanton contribution in some form of twistor-string theory. In particular, h and B enter just as they do in the vertex operators of [37]. Acknowledgements. We would like to thank Mohab Abou-Zeid, Paolo Benincasa, Rutger Boels, Freddy Cachazo, Henriette Elvang, Dan Freedman and Chris Hull for useful discussions. LM is partially supported by the EU through the FP6 Marie Curie RTN ENIGMA (contract number MRTN–CT–2004–5652) and through the ESF MISGAM network. This work was financed by EPSRC grant number EP/F016654, http://gow.epsrc.ac.uk/ViewGrant.aspx?GrantRef=EP/F016654/1.
A. Simplifying the BGK Amplitudes In this appendix we will show analytically that the Berends, Giele & Kuijf [19] form of the graviton MHV amplitude agrees with the simplified expression (1) used in the text. Similar manipulations have been performed in [5,16,47]; our version of the amplitude is nearest to one given implicitly in [47], although we believe the detailed form is new. Berends, Giele & Kuijf give the MHV amplitude MBGK =
κ n−2 (4) δ p M,
(94)
where for n ≥ 5, M(1+ , 2− , 3− , . . . , n − 1− , n + ) ⎧ ⎫ ⎨ 12n − 2 n − 1 F n−3 ⎬ n−1 = [1n]8 [i j] + P(2,...,n−2) ⎩ ⎭ [1 n − 1] N (n) i=1 j=i+2
(95)
Gravity, Twistors and the MHV Formalism
with N (n) :=
i< j
855
[i j], and where
F :=
n−3
k| pk+1 + pk+2 + · · · + pn−1 |n]
(96)
k=3
when n ≥ 6 and F = 1 when n = 5. In (95), the symbol P2,...,n−2 denotes a sum over all permutations of gravitons 2 to n − 2. We begin by writing 12n − 2 n − 1 21[1n] n − 2 n − 1[n − 1 n] = [1 n − 1] [1 n − 1][n − 1 n][n1] 2| p3 + p4 + · · · + pn−1 |n]n − 2| pn−1 |n] , =− [1 n − 1][n − 1 n][n1]
(97)
using momentum conservation in the second step. Combining this with F in Eq. (96) gives a factor n−2 −
k=2 k| pk+1
+ pk+2 + · · · + pn−1 |n] . [1 n − 1][n − 1 n][n1]
(98)
Next, by carefully altering the limits of the products, we may re-express N (n) as N (n) =
n−1
n
[i j] = −C(n)
i=1 j=i+1
⎧ ⎨n−3 n−1 ⎩
i=1 j=i+2
[i j]
⎫ ⎬ n−2 ⎭
[kn],
(99)
k=2
where C(n) is the cyclic product [12][23] · · · [n − 1 n][n1]. The term in braces now cancels an identical term in the numerator of (95). Hence we obtain κ 2 δ p MBGK (1+ , 2− , 3− , . . . , n − 1− , n + ) = n−2 1 k| pk+1 +· · ·+ pn−1 |n] [1n]8 +P(2,...,n−2) , × [1 n−1][n−1 n][n 1] C(n) [kn]
(100)
k=2
which is the form of the amplitudes in Eq. (1).
B. Yang-Mills In this appendix, we will review the twistor construction of the Parke-Taylor amplitudes in Yang-Mills theory (see [33,34] for further details). Although this section is not strictly necessary for an understanding of the gravitational case, there are nonetheless many analogies between the two and some readers may find it useful to refer here for comparison.
856
L. Mason, D. Skinner
B.1. Scattering off an anti-self-dual Yang-Mills background. On spacetime, Yang-Mills theory may be described by the Chalmers & Siegel [67] action
1 S[A, G + ] = 2 (101) tr G + ∧ F − G + ∧ G + , g M where F = d A + A2 and G + is a Lie algebra-valued self-dual 2-form. We will frequently drop the superscript from G + , but it is always self-dual. The field equations are G+ =
1 + F 2
and
DAG+ = 0 ,
(102)
where D A is the covariant derivative. The first of these equations may be viewed as a constraint; enforcing it in (101) one recovers the standard Yang-Mills action, up to a topological term. Using the Bianchi identity, the second equation is the standard YangMills equations D A ∗ F = 0. Anti-self-dual solutions to (102) have F + = 0. Replacing A → A+a and G → G +g and expanding the full field equations to linear order, one finds 2g = (D A a)+
and
DAg = 0
(103)
when the background is anti-self-dual. The solution space of these linear equations is an (infinite dimensional) vector space U (to be considered modulo gauge transformations). If R denotes the space of solutions to the full Equations (102), then U may be interpreted as the fibre of T R over a particular ASD solution. As for gravity, we identify U − ⊂ U as the subspace with g = 0. Since F → F + D A a = F + (D A a)+ + (D A a)− , Eq. (103) shows that linearized solutions in U − preserve the anti-self-duality of the Yang-Mills curvature. U + is defined asymmetrically to be U + := {g ∈ 2+ (M, EndE) D A g = 0}, modulo gauge transformations. From Eq. (103), such g fields generate linear fluctuations in the self-dual part of the curvature. However, on an ASD background it does not make sense to ask for U + to be the solutions that are purely self-dual, because under a background gauge transformation with parameter χ , the variation a → a + D A χ implies D A a → D A a + D A (D A χ ) = D A a + [F − , χ ]
(104)
so that requiring (D A a)− = 0 would not be invariant under background gauge transformations. Again, this is summarized by the exact sequence 0 → U − → U → U + → 0,
(105)
where U − → U is an inclusion and the map U → U + is (a, g) → g. The fact that a selfdual fluctuation may or may not have an anti-self-dual component obstructs the global splitting U = U − ⊕ U + . Once again, this obstruction may be attributed to the MHV amplitudes, interpreted as scattering a linearized self-dual field off the ASD background. Evaluating the action (101) on (A, G + ) = (A0 + a, g), where (A0 , 0) are an ASD background and (a, g) obey the linearized equations (103), we find
i i S[A0 + a, g] = 2 tr (g ∧ g) (106) g M which, according to the path-integral argument in Sect. 2, is the tree-level amplitude for a positive helicity gluon to scatter off the background and emerge with negative helicity. We can again confirm this with a separate calculation.
Gravity, Twistors and the MHV Formalism
857
The space of solutions R of (102) again possesses a naturally defined closed twoform
1 := 2 tr (δG ∧ δ A) . (107) g C As a consequence of the field equations, is independent of the Cauchy surface C and descends to a symplectic form on Y/gauge. If A1,2 = (a1,2 , g1,2 ) are two sets of linearized solutions, then g 2 (A1 , A2 ) = C tr (g1 ∧ a2 − g2 ∧ a1 ). As for gravity on an anti-self-dual spacetime, the symplectic form (107) can be used to define a splitting of U that depends on a choice of Cauchy surface C. Clearly, U − forms a Lagrangian subspace with respect to (107) and we can ensure U + is likewise Lagrangian by defining a fluctuation A2 to be purely self-dual if
1 (A2 , A1 ) = − 2 tr (g2 ∧ a1 ) (108) g C for an arbitrary fluctuation A1 . The quantum mechanical inner-product is defined in the same way as in the text (on Minkowski space M the positive/negative frequency splitting can be performed straightforwardly) and agrees with the symplectic form (107) on positive frequency states. The amplitude for a linearized fluctuation A1 that has positive helicity and positive frequency at I − to emerge at I + with negative helicity (and positive energy) after traversing the region of anti-self-dual Yang-Mills curvature (in M) is A2 |A1 asd , where A2 is purely self-dual at I + . In exact analogy to Eq. (21), we find
i A2 A1 asd = 2 tr (g2 ∧ a1 ) g I+
i i = 2 tr (D A g2 ∧ a1 + g2 ∧ D A a1 ) + 2 tr (g2 ∧ a1 ) (109) g M g I−
i = 2 tr (g2 ∧ g1 ) g M after using the linearized field equations (103) and the fact that A1 is purely self-dual at I −. Equation (109) is a generating function for the Parke-Taylor amplitudes. To obtain them in their usual form, one must construct a background ASD field A that is a (nonlinear) superposition of n − 2 plane waves and solve the equation D A g = 0 with such an A. Finally, one must expand the above integral to the appropriate order. As for gravity, these problems are considerably simplified by the use of twistor theory, which brings out the integrability of the ASD Yang-Mills equations. B.2. The twistor theory of Yang-Mills. For the basic notation of twistor space, we refer to the beginning of Sect. 3. Anti-self-dual connections on spacetime correspond to holomorphic bundles E on twistor space, by the Ward construction [68]. In the Dolbeault framework used in this paper, such a bundle is determined by an operator ∂¯ +A satisfying ¯ F (0,2) := (∂¯ + A)2 = 0, where ∂¯ is the standard ∂-operator on twistor space and A is the (0, 1)-form part of a connection on E (and has homogeneity degree 0). Note that ¯ ∂¯ + A may be regarded as a deformation of the ∂-operator on a flat gauge bundle, while the integrability condition F (0,2) = 0 arises as the field equations of the action
858
L. Mason, D. Skinner
PT
∧ tr(G ∧ F),
(110)
where G is a (0, 1)-form of homogeneity −4 with values in End(E) and is the canonical holomorphic (3,0)-form of weight +4 on PT . Thus, (110) is the twistor equivalent of the g 2 → 0 limit of (101) on spacetime. Following Sparling [69], the spacetime Yang-Mills connection can be reconstructed by first solving (111) ∂¯ + A L H = 0 , x
where, for a Yang-Mills field on spacetime with gauge group G, A takes values in the complexified Lie algebra of G whereas H is valued in the complexification of G itself. The notation (∂¯ + A) L means the restriction of the twistor space operator ∂¯ + A to x L x . A solution H of (111) is a global holomorphic frame of E| L x , related to the twistor connection one-form by A| L x = −∂¯ H H −1 .
(112)
The generic existence of such frames for each x is guaranteed by standard properties of holomorphic vector bundles16 . To reconstruct the spacetime connection A, first note that H −1 π α˙ ∂ H/∂ x α α˙ has homogeneity one in πα˙ . Moreover, H −1 π α˙ ∂ H/∂ x α α˙ is holomorphic on L x , since ¯∂ H −1 π α˙ ∂ H = H −1 A π α˙ ∂ H − H −1 π α˙ ∂ (AH ) = 0 , (113) ∂ x α α˙ ∂ x α α˙ ∂ x α α˙ where π α˙ ∂A/∂ xα α˙ = 0 because A has been pulled back from PT and so depends on x only through the combination x α α˙ πα˙ . Thus H −1 π α˙ ∂α α˙ H must in fact be linear in πα˙ and so may be written as H −1 π α˙ ∂α α˙ H = π α˙ Aα α˙ (x)
(114)
for some Lie-algebra valued functions Aα α˙ that depend only on spacetime. This provides the spacetime connection A = Aα α˙ d x α α˙ . To construct a twistor expression for A2 |A1 , recall that for a flat Yang-Mills bundle, the Penrose transform of a linearized fluctuation g is related to G by
gα˙ β˙ (x) = [π dπ ] ∧ πα˙ πβ˙ p ∗ (G) , (115) Lx
β˙ ¯ = 0 then g ˙ automatically obeys where g = gα˙ β˙ d x α α˙ ∧ d xα . Moreover, if ∂G α˙ β α α ˙ ∗ ∂ gα˙ β˙ = 0, again because the pullback p G depends on x only through x α α˙ πα˙ . The ¯ = 0 and ∂ α α˙ G ˙ = 0 are the linearized field equations of (110) and (101) equations ∂G α˙ β in the case that the background bundles are flat17 so that we can find a gauge where A = 0 and A = 0. However, on a ASD Yang-Mills background, (115) does not quite 16 The Penrose-Ward transform requires E| L x to be trivial. This will generically be the case and arises because the fibre of the Yang-Mills bundle over a spacetime point x is by definition the space of global holomorphic sections of E| L x ; these ‘jump’ if E| L x becomes non-trivial, so any twistor bundle that comes from a spacetime bundle will necessarily be trivial over L x . 17 They can also be thought of as Abelianized versions of the full theory.
Gravity, Twistors and the MHV Formalism
859
make sense. In order to add up an End(E)-valued form over L x , in the presence of a non-flat Yang-Mills bundle we need first to pick a holomorphic trivialization of E| L x that is global over L x : this is just the solution H of Eq. (111). The background-coupled twistor integral formula for g+ is then
gα˙ β˙ (x) = [π dπ ] ∧ πα˙ πβ˙ H −1 (x, π ) p ∗ (G) H (x, π ). (116) From Eq. (116) we now find (dropping the pullback symbol p ∗ )
∂ −1 α α˙ H GH ∂ gα˙ β˙ = [π dπ ] ∧ πα˙ πβ˙ ∂ xα α˙
= [π dπ ] ∧ πα˙ πβ˙ −πα˙ Aα α˙ H −1 G H + H −1 G H Aα α˙ = − Aα α˙ , gα˙ β˙ , (117) or in other words D A g = 0, which is the linearized field equation (103) for g on an ASD background. Therefore, the scattering amplitude we seek is given by
i
A2 A1 asd = 2 d4 x [π1 dπ1 ][π2 dπ2 ] [π1 π2 ]2 tr H2−1 G2 H2 H1−1 G1 H1 , g (118) where the integral on the right is then taken over R4 × CP1 × CP1 . To obtain the Parke-Taylor amplitudes we must expand the frames H as a perturbation series around a flat background by inverting the relation A| L x = −∂¯ H H −1 . Rather than do this directly (see [48]), it is simpler to note that the Green’s function K 12 for the ¯ ∂-operator on L x , acting on sections of End(E)| L x , is related to H by K 12 (x, π1 , π2 ) =
1 2π i
H (x, π1 )H −1 (x, π2 ) [π1 π2 ]
(119)
and may formally be thought of as (∂¯ + A| L x )−1 . This is analogous to Eq. (64) in Sect. 4. The Green’s function thus depends non-polynomially on A; expanding the right-hand
side of (118) as a series in A using K i j A=0 = 1/2π i[πi π j ], one obtains i g2
n n [πi dπi ] [π1 π p ]4 tr An · · · A p+1 G p A p−1 · · · A2 G1 d x [πi πi+1 ] 4
(120)
p=2
i=1
for the vertex involving n fields. To obtain the Parke-Taylor amplitudes, take A and G to be linear combinations of momentum eigenstates pα α˙ = kα kα˙ , with helicities −1 and +1, respectively. As in [58], these can be represented by the twistor functions, A=g G=g
n i=3 2 i=1
i Ti δ¯(0) ([π i]) exp
ω i[i σ ] , [π σ ]
ω i[i σ ] , i Ti δ¯(−4) ([π i]) exp [π σ ]
(121)
860
L. Mason, D. Skinner
where Ti are arbitrary elements of the Lie and the i are nalgebra of the gauge group, expansion parameters. The coefficient of i=1 i in (120) is the n th -order Parke-Taylor amplitude (complete with the appropriate colour-trace). Treating the fields A and G as End E-valued (0,1)-forms, rather than representatives of cohomology classes, we can combine (120) with the action (110) to obtain a twistor action for the MHV diagram formulation of Yang-Mills. It is straightforward to extend this to an action for N = 4 SYM. See [33,34,41] for details.
References 1. Bern, Z., Dixon, L.J., Perelstein, M., Rozowsky, J.S.: Multi-leg one-loop gravity amplitudes from Gauge theory. Nucl. Phys. B 546, 423 (1999) 2. Bern, Z., Bjerrum-Bohr, N.E.J., Dunbar, D.C.: Inherited twistor-space structure of gravity loop amplitudes. JHEP 0505, 056 (2005) 3. Bern, Z., Carrasco, J.J., Dixon, L.J., Johansson, H., Kosower, D.A., Roiban, R.: Three-loop superfiniteness of N = 8 supergravity. Phys. Rev. Lett. 98, 161303 (2007) 4. Bern, Z., Carrasco, J.J., Forde, D., Ita, H., Johansson, H.: Unexpected cancellations in gravity theories. Phys. Rev. D 77, 025010 (2008) 5. Bedford, J., Brandhuber, A., Spence, B., Travaglini, G.: A recursion relation for gravity amplitudes. Nucl. Phys. B 721, 98 (2005) 6. Cachazo, F., Svrcek, P.: Tree level recursion relations in general relativity. http://arXiv.org/abs/hep-th/ 0502160v3, 2005 7. Benincasa, P., Boucher-Veronneau, C., Cachazo, F.: Taming tree amplitudes in general relativity. JHEP 0711, 057 (2007) 8. Cachazo, F., Skinner, D.: On the structure of scattering amplitudes in N = 4 super Yang-Mills and N = 8 supergravity. http://arXiv.org/abs/0801.4574v2[hep-th], 2008 9. Arkani-Hamed, N., Cachazo, F., Kaplan, J.: What is the simplest quantum field theory?. http://arXiv.org/ abs/0808.1446v2[hep-th], 2008 10. Bjerrum-Bohr, N.E.J., Vanhove, P.: Explicit cancellation of triangles in one-loop gravity amplitudes. JHEP 0804, 065 (2008) 11. Bjerrum-Bohr, N.E.J., Vanhove, P.: Absence of triangles in maximal supergravity amplitudes. http:// arXiv.org/abs/0805.3682v2[hep-th], 2008 12. Bjerrum-Bohr, N.E.J., Dunbar, D.C., Ita, H., Perkins, W.B., Risager, K.: The no-triangle hypothesis for N = 8 supergravity. JHEP 0612, 072 (2006) 13. Bjerrum-Bohr, N.E.J., Dunbar, D.C., Ita, H., Perkins, W.B., Risager, K.: MHV-vertices for gravity amplitudes. JHEP 0601, 009 (2006) 14. Nasti, A., Travaglini, G.: One-loop N = 8 supergravity amplitudes from MHV diagrams. Class. Quant. Grav. 24, 6071 (2007) 15. Bianchi, M., Elvang, H., Freedman, D.Z.: Generating tree amplitudes in N = 4 SYM and N = 8 SG. JHEP 0809, 063 (2008) 16. Elvang, H., Freedman, D.Z.: Note on graviton MHV amplitudes. JHEP 0809, 063 (2008) 17. Parke, S.J., Taylor, T.R.: An amplitude for n gluon scattering. Phys. Rev. Lett. 56, 2459 (1986) 18. Berends, F.A., Giele, W.T.: Recursive calculations for processes with n gluons. Nucl. Phys. B 306, 759 (1988) 19. Berends, F.A., Giele, W.T., Kuijf, H.: On relations between Multi-Gluon and Multi-Graviton scattering. Phys. Lett. B 211, 91 (1988) 20. Bialynicki-Birula, I., Newman, E.T., Porter, J., Winicour, J., Lukacs, B., Perjes, Z., Sebestyen, A.: A note on helicity. J. Math. Phys. 22, 2530 (1981) 21. Ashtekar, A.: A note on helicity and selfduality. J. Math. Phys. 27, 824 (1986) 22. Penrose, R.: Non-linear gravitons and curved twistor theory. Gen. Rel. Grav. 7, 31 (1976) 23. Hansen, R.O., Newman, E.T., Penrose, R., Tod, K.P.: The metric and curvature properties of H space. Proc. Roy. Soc. Lond. A 363, 445 (1978) 24. Plebanski, J.F.: On the separation of Einsteinian substructures. J. Math. Phys. 18, 2511 (1977) 25. Capovilla, R., Jacobson, T., Dell, J., Mason, L.: Selfdual two forms and gravity. Class. Quant. Grav. 8, 41 (1991) 26. Mason, L., Frauendiener, J.: The Sparling 3-Form, Ashtekar Variables and Quasi-Local Mass. Lond. Math. Soc. Lect. Notes 156, Cambridge: Cambridge University Press, 1990, p. 189
Gravity, Twistors and the MHV Formalism
861
27. Ashtekar, A.: New variables for classical and quantum gravity. Phys. Rev. Lett. 57, 2244 (1986) 28. Penrose, R., MacCallum, M.A.H.: Twistor theory: An approach to the quantization of fields and spacetime. Phys. Rept. 6, 241 (1972) 29. Ashtekar, A., Jacobson, T., Smolin, L.: A new characterization of half flat solutions to einstein’s equation. Commun. Math. Phys. 115, 631 (1988) 30. Abou-Zeid, M., Hull, C.M.: A chiral perturbation expansion for gravity. JHEP 0602, 057 (2006) 31. Rosly, A.A., Selivanov, K.G.: Gravitational SD perturbiner. http://arXiv.org/abs/hep-th/9710196v1, 1997 32. Newman, E.T.: Heaven and its Properties. Gen. Rel. Grav. 7, 107 (1976) 33. Mason, L.J.: Twistor actions for non-self-dual fields: a derivation of twistor-string theory. JHEP 0510, 009 (2005) 34. Boels, R., Mason, L., Skinner, D.: Supersymmetric Gauge theories in twistor space. JHEP 0702, 014 (2007) 35. Berkovits, N., Witten, E.: Conformal supergravity in twistor-string theory. JHEP 0408, 009 (2004) 36. Mason, L., Skinner, D.: Heterotic twistor-string theory. Nucl. Phys. B 795, 105 (2008) 37. Abou-Zeid, M., Hull, C., Mason, L.: Einstein supergravity and new twistor string theories. Commun. Math. Phys. 282, 519–573 (2008) 38. Nair, V.P.: A note on graviton amplitudes for new twistor string theories. Phys. Rev. D 78, 041501 (2008) 39. Mason, L.J., Wolf, M.: A twistor action for N = 8 self-dual supergravity. Commun. Math. Phys. 288, 97–123 (2009) 40. Witten, E.: Perturbative Gauge theory as a string theory in twistor space. Commun. Math. Phys. 252, 189 (2004) 41. Boels, R., Mason, L., Skinner, D.: From twistor actions to MHV diagrams. Phys. Lett. B 648, 90 (2007) 42. Kawai, H., Lewellen, D.C., Tye, S.H.H.: A relation between tree amplitudes of closed and open strings. Nucl. Phys. B 269, 1 (1986) 43. Britto, R., Cachazo, F., Feng, B., Witten, E.: Direct proof of tree-level recursion relation in Yang-Mills theory. Phys. Rev. Lett. 94, 181602 (2005) 44. Ananth, S., Theisen, S.: KLT relations from the Einstein-Hilbert Lagrangian. Phys. Lett. B 652, 128 (2007) 45. Mansfield, P.: The Lagrangian origin of MHV rules. JHEP 0603, 037 (2006) 46. Ettle, J.H., Morris, T.R.: Structure of the MHV-rules Lagrangian. JHEP 0608, 003 (2006) 47. Nair, V.P.: A Note on MHV amplitudes for gravitons. Phys. Rev. D 71, 121701 (2005) 48. Rosly, A.A., Selivanov, K.G.: On amplitudes in self-dual sector of Yang-Mills theory. Phys. Lett. B 399, 135 (1997) 49. Penrose, R., Rindler, W.: Spinors and Spacetime 1 & 2. Cambridge Monographs on Math. Phys., Cambridge: CUP, 1984 & 1986 50. Risager, K.: A direct proof of the CSW rules. JHEP 0512, 003 (2005) 51. Elvang, H., Freedman, D.Z., Kiermaier, M.: Recursion relations, generating functions, and unitarity sums in N = 4 SYM theory. JHEP 0904, 009 (2009) 52. Woodhouse, N.M.J.: Geometric Quantization. Second edition, Oxford Mathematical Monographs. OUP, Oxford, 1992 53. Ashtekar, A., Engle, J., Sloan, D.: Asymptotics and hamiltonians in a first order formalism. Class. Quant. Grav 25, 095020 (2008) 54. Wardm, R.S., Wells, R.O.: Twistor Geometry and Field Theory. Cambridge Monographs on Math. Phys. CUP, Campridge, 1990 55. Huggett, S.A., Tod, K.P.: An Introduction To Twistor Theory. London Mathematical Society Student Texts 4, Campridge: CUP, 1985 56. Eastwood, M., Tod, P.: Edth - a differential operator on the sphere. Math. Proc. Camb. Phil. Soc. 92, 317 (1982) 57. Porter, J.R.: The nonlinear graviton: superposition of plane waves. Gen. Rel. Grav. 14, 1023 (1982) 58. Cachazo, F., Svrcek, P., Witten, E.: MHV vertices and tree amplitudes in gauge theory. JHEP 0409, 006 (2004) 59. Bena, I., Bern, Z., Kosower, D.A.: Twistor-space recursive formulation of Gauge theory amplitudes. Phys. Rev. D 71, 045008 (2005) 60. Bern, Z., Dixon, L.J., Dunbar, D.C., Perelstein, M., Rozowsky, J.S.: On the relationship between YangMills theory and gravity and its implication for ultraviolet divergences. Nucl. Phys. B 530, 401 (1998) 61. Brandhuber, A., Spence, B., Travaglini, G.: From trees to loops and back. JHEP 0601, 142 (2006) 62. Boels, R.: A quantization of twistor Yang-Mills theory through the background field method. Phys. Rev. D 76, 105027 (2007) 63. Gorsky, A., Rosly, A.: From Yang-Mills Lagrangian to MHV diagrams. JHEP 0601, 101 (2006) 64. Ferber, A.: Supertwistors and conformal supersymmetry. Nucl. Phys. B 132, 55 (1978) 65. Nair, V.P.: A Current algebra for some Gauge theory amplitudes. Phys. Lett. B 214, 215 (1998) 66. Wolf, M.: Self-dual supergravity and twistor theory. Class. Quant. Grav. 24, 6287 (2007) 67. Chalmers, G., Siegel, W.: The self-dual Sector of QCD Amplitudes. Phys. Rev. D 54, 7628 (1996)
862
L. Mason, D. Skinner
68. Ward, R.S.: On self-dual Gauge fields. Phys. Lett. A 61, 81 (1977) 69. Sparling, G.: Dynamically broken symmetry and global yang-Mills in Minkowski space. Sect. 1.4.2 In: Further Advances in Twistor Theory, Mason, L., Hughston L. (eds), Pitman Research Notes in Maths 231, Essex: Longman, Harlow, 1995 Communicated by G.W. Gibbons
Commun. Math. Phys. 294, 863–889 (2010) Digital Object Identifier (DOI) 10.1007/s00220-009-0971-5
Communications in
Mathematical Physics
Twisted K -Theory and Finite-Dimensional Approximation Kiyonori Gomi The Department of Mathematics, The University of Texas at Austin, 1 University Station C1200, Austin, TX 78712-0257, USA. E-mail:
[email protected];
[email protected] Received: 4 September 2008 / Accepted: 20 October 2009 Published online: 24 December 2009 – © The Author(s) 2009. This article is published with open access at Springerlink.com
Abstract: We provide a finite-dimensional model of the twisted K -group twisted by any degree three integral cohomology class of a CW complex. One key to the model is Furuta’s generalized vector bundle, and the other is a finite-dimensional approximation of Fredholm operators. 1. Introduction Since the work of Atiyah and Hirzebruch [2], K -theory has been recognized as a fundamental notion in topology and geometry. Twisted K -theory is a variant of K -theory originating from the works of Donovan-Karoubi [10] and Rosenberg [19]. Much focus is on twisted K -theory recently, due to applications, for example, to Dbrane charges ([17,20]), Verlinde algebras [12] and quantum Hall effects [8]. As is well-known, the K -group K (X ) of a compact space X admits various formulations. The standard formulation of K (X ) uses finite dimensional vector bundles on X . One can also formulate K (X ) by using a C ∗ -algebra as well as the space of Fredholm operators. To define twisted K -theory, we usually appeal to the latter two formulations above, involving some infinite dimensions. K -theory enjoys numerous applications to topology and geometry because of its realization by means of vector bundles. To give a similar realization of twisted K -theory seems to be an interesting problem to be studied not only for better understanding but also for further applications. So far, as a partial answer to the problem, twisted vector bundles or bundle gerbe K -modules [6] are utilized to realize the twisted K -group whose “twisting” satisfies a condition. The condition is that the degree three integral cohomology class corresponding to the twisting is of finite order. A complete answer to this realization problem, valid for twistings corresponding to any degree three integral cohomology classes, was known Current address: Department of Mathematics, Kyoto University, Kyoto 606-8502, Japan
864
K. Gomi
by André Henriques. The aim of the present paper is to give another complete answer by generalizing the following result announced in [15]: Theorem 1. Let X be a CW complex, P a principal bundle over X whose structure group is the projective unitary group of a separable Hilbert space of infinite dimension, and K P (X ) the twisted K -group. We write KFP (X ) for the homotopy classes of P-twisted (Z2 -graded) vectorial bundles over X . Then there exists a natural isomorphism α : K P (X ) −→ KFP (X ). The notion of vectorial bundle is a generalization of the notion of vector bundles due to Mikio Furuta [13]. Vectorial bundles realize the ordinary K -group K (X ), and arise as finite-dimensional geometric objects approximating families of Fredholm operators. We can think of the approximation as a linear version of the finite-dimensional approximation of the Seiberg-Witten equations [14]. Since K P (X ) consists of certain families of Fredholm operators, a twisted version of vectorial bundles provides a suitable way to realize twisted K -theory. As a simple application of Theorem 1, we can generalize some notions of 2-vector bundles [5,7]. The notion of 2-vector bundles in the sense of Brylinski [7] uses the category of vector bundles, and a 2-vector bundle of rank 1 reproduces the category of twisted vector bundles, so that the twisted K -group whose twisting corresponds to a degree three integral cohomology class of finite order. By using the category of vectorial bundles instead, we get a proper generalization of Brylinski’s 2-vector bundles. This generalization reproduces the category of twisted vectorial bundles, and hence the twisted K -group with any twisting. A similar replacement may generalize 2-vector bundles of Baas, Dundas and Rognes [5], which they studied in seeking for a geometric model of elliptic cohomology. Also, Theorem 1 allows us to construct Chern characters of twisted K -classes in a purely finite-dimensional manner [16]. In a word, the proof of Theorem 1 is a comparison of cohomology theories: as is wellknown, the twisted K -group K P (X ) fits into a certain generalized cohomology theory K P∗ (X, Y ). The group KFP (X ) also fits into a similar cohomology theory KFP∗ (X, Y ), and the homomorphism α : K P (X ) → KFP (X ) extends to a natural transformation between these two cohomology theories. Then, appealing to a standard method in algebraic topology, we compare these cohomology theories to show their equivalence. According to the outline of the proof above, this paper is organized as follows. In Sect. 2, we review a definition of twisted K -theory and a construction of twisted K -cohomology K P∗ (X, Y ). In Sect. 3, we introduce the notion of (Z2 -graded) vectorial bundles and its twisted version. In Sect. 4, we construct the cohomology theory KFP∗ (X, Y ). In Sect. 5, we construct the natural transformation between K P∗ (X, Y ) and KFP∗ (X, Y ). A key to the construction is a finite-dimensional approximation of a family of Fredholm operators. After a study of the natural transformation, we compare the cohomology theories to derive Theorem 1. Finally, in the Appendix, proof of Furuta’s results, crucial to the present paper, are culled from the Japanese textbook [13] for convenience. 2. Twisted K -Theory We here review twisted K -cohomology theory, following [3,9] mainly. 2.1. Review of twisted K -theory. Let PU (H) = U (H)/U (1) be the projective unitary group of a separable Hilbert space H of infinite-dimension. We topologize PU (H) by
Twisted K -Theory and Finite-Dimensional Approximation
865
using the compact-open topology in the sense of [3]. Let F(H) be the set of bounded linear operators A : H → H such that A∗ A − 1 and A A∗ − 1 are compact operators: F(H) = {A : H → H| A∗ A − 1, A A∗ − 1 ∈ K(H)}. Note that F(H) is a subset of the space of Fredholm operators on H. We induce a topology on F(H) by using the map F(H) −→ B(H)co × B(H)co × K(H)norm × K(H)norm , A → (A, A∗ , A∗ A − 1, A A∗ − 1), where B(H)co is the space of bounded linear operators B(H) topologized by the compact-open topology, and K(H)norm is the space of compact operators K(H) topologized by the usual operator norm. Then F(H) is a representing space for K -theory, and PU (H) acts continuously on F(H) by conjugation ([3]). In this paper, the “twist” in twisted K -theory is given by a principal PU (H)-bundle. For a principal PU (H)-bundle P → X given, the conjugate action gives the associated bundle P × Ad F(H) → X whose fiber is F(H). Definition 2.1. Let X be a compact Hausdorff space, and P → X a principal PU (H)bundle. We define the twisted K -group K P (X ) to be the group consisting of fiberwise homotopy classes of the sections of P × Ad F(H) → X : K P (X ) = (X, P × Ad F(H))/homotopy, where the addition in K P (X ) is given by fixing an isomorphism H ⊕ H ∼ = H. Remark 1. As in [3], we can consider a more refined “twist” by introducing a Z2 -grading to the Hilbert space and using unitary transformations of degree 1. However, the present paper does not cover the case. Remark 2. In [3], a projective space bundle plays the role of a “twisting”. Since the structure group of the projective space bundle is PU (H), Definition 2.1 gives the same twisted K -group as that in [3]. Remark 3. Instead of the compact-open topology, we can also work with the topology on PU (H) given by the operator norm. In this case, the formulation of twisted K -theory uses the space of (bounded) Fredholm operators on H equipped with the operator norm topology, instead of F(H). An advantage of the compact-open topology, other than that pointed out in [3], is that it simplifies some argument in Subsect. 5.1. 2.2. Review of twisted K -cohomology. We formulate twisted K -cohomology as a certain generalized cohomology. We write C for the category of CW pairs: an object in C is a pair (X, Y ) consisting of a CW complex X and its subcomplex Y . A morphism f : (X , Y ) → (X, Y ) is a continuous map f : X → X such that f (Y ) ⊂ Y . We also write C for the category of CW pairs equipped with PU (H)-bundles: an object (X, Y ; P) in C consists of a CW pair (X, Y ) ∈ C and a principal PU (H)-bundle P → X . A morphism ( f, F) : (X , Y ; P ) → (X, Y ; P) consists of a morphism f : (X , Y ) → (X, Y ) in C and a bundle map F : P → P covering f . A CW complex X equipped with a principal PU (H)-bundle P → X will be identified with (X, ∅; P) ∈ C.
866
K. Gomi
Let (X, Y ) ∈ C be a CW pair and P → X a principal PU (H)-bundle. The support of a section A ∈ (X, P × Ad F(H)) is defined to be the closure of the set consisting of the points at which A is not invertible: SuppA = {x ∈ X | Ax : H → H is not invertible}. We define K P (X, Y ) by using sections A ∈ (X, P × Ad F(H)) such that SuppA ∩ Y = ∅. In K P (X, Y ), two sections A0 and A1 are identified if they are connected by a sec˜ ∈ (X × I, (P × I ) × Ad F(H)) such that SuppA ˜ ∩ (Y × I ) = ∅, where I tion A is the interval [0, 1]. We then define the twisted K -cohomology groups K P−n (X, Y ) as follows: K P×I n (X × I n , Y × I n ∪ X × ∂ I n ), (n ≥ 0), K P−n (X, Y ) = K Pn (X, Y ), (n < 0). Clearly, a morphism ( f, F) : (X , Y ; P ) → (X, Y ; P) induces a homomorphism ( f, F)∗ : K Pn (X, Y ) → K Pn (X , Y ) for all n ∈ Z. In the case of P = f ∗ P, we simply write f ∗ : K Pn (X, Y ) → K nf ∗ P (X , Y ) for the homomorphism induced from ( f, f ), ∗ where f : f P → P is the canonical bundle map covering f . Now, we summarize basic properties of twisted K -cohomology theory ([3,9,11]): Proposition 2.2. The assignment of {K Pn (X, Y )}n∈Z to (X, Y ; P) ∈ Chas the following properties: (1) Homotopy axiom. If ( f i , Fi ) : (X , Y ; P ) → (X, Y ; P), (i = 0, 1) are homotopic, then the induced homomorphisms coincide: ( f 0 , F0 )∗ = ( f 1 , F1 )∗ . (2) Excision axiom. For subcomplexes A, B ⊂ X , the inclusion map induces the isomorphism: n n K P| (A ∪ B, B) ∼ (A, A ∩ B). (n ∈ Z). = K P| A∪B A
(3) Exactness axiom. There is the natural long exact sequence: δn−1
δn
n−1 n · · · → K P| (Y ) → K Pn (X, Y ) → K Pn (X ) → K P| (Y ) → · · · . Y Y
the inclusion maps X λ → (4) Additivity axiom. For a family {(X λ , Yλ ; Pλ )}λ∈ in C, λ X λ induce the natural isomorphism: −n ( λ X λ , λ Yλ ) ∼ K P−n (X λ , Yλ ), (n ∈ Z). K = Pλ λ λ
λ
(5) Bott periodicity. There is the natural isomorphism: βn : K Pn (X, Y ) −→ K Pn−2 (X, Y ), (n ∈ Z). The homotopy axiom and the additivity axiom are clear. The excision axiom is due to the fact that the set of invertible operators in F(H) is contractible [3]. The periodicity is a consequence of the homotopy equivalence F(H) 2 F(H), ([3,4]). The exactness axiom follows essentially from the cofibration sequence: · · · ←− 2 Y ←− (X/Y ) ←− X ←− Y ←− X/Y ←− X ←− Y,
Twisted K -Theory and Finite-Dimensional Approximation
867
where the reduced suspension. Noting the homotopy equivalence n stands for n (X pt) X × I /(Y × I n ∪ X × ∂ I n ), we obtain the non-positive part of the long exact sequence for a pair (X, Y ) in a way similar to that used in [1]. Then we get the positive part by using the Bott periodicity. A similar construction of the exact sequence will be performed in Subsect. 4.3 in our finite-dimensional model. Remark 4. The definition of K P (X, Y ) in [9] is equivalent to that in this paper, because of the definition of the support of A ∈ (X, P × Ad F(H)). The definition of K P−n (X ) in [3,9], which utilizes sections of the bundle P × Ad n F(H) over X , is also equivalent to our definition. 3. Vectorial Bundle We here introduce Furuta’s generalized vector bundles [13] as vectorial bundles. Our formulation differs slightly from the original formulation in [13]. Twisted vectorial bundles are also introduced in this section. 3.1. Vectorial bundle. Definition 3.1. Let X be a topological space. For a subset U ⊂ X , we define the category HF(U ) as follows. An object in HF(U ) is a pair (E, h) consisting of a Z2 -graded Hermitian vector bundle E → U of finite rank and a Hermitian map h : E → E of degree 1. The homomorphisms in HF(U ) are defined by HomHF (U ) ((E, h), (E , h )) = {φ : E → E | degree 0, φh = h φ}/ , where stands for an equivalence relation. That φ φ means: For each point x ∈ U , there are a positive number µ > 0 and an open subset V ⊂ U containing x such that: for all y ∈ V and ξ ∈ (E, h) y,<µ , we have φ(ξ ) = φ (ξ ). In the above, we put (E, h) y,<µ =
Ker(h 2y − λ) =
λ<µ
{ξ ∈ E y | h 2y ξ = λξ }.
λ<µ
By abuse of notation, we just write φ for the equivalence class [φ] of a map φ : (E, h) → (E , h ) in HomHF (U ) ((E, h), (E , h )). For a subset V ⊂ U , the restriction (E, h) → (E, h)|V defines a functor HF(U ) → HF(V ), which composes properly for a smaller subset in V . Definition 3.2. For a space X , we define the category KF(X ) as follows. (1) An object (U, (E α , h α ), φαβ ) in KF(X ) consists of an open cover U = {Uα }α∈A of X , objects (E α , h α ) in HF(Uα ), and homomorphisms φαβ : (E β , h β ) → (E α , h α ) in HF(Uαβ ) such that: φαβ φβα = 1
in HF(Uαβ );
φαβ φβγ = φαγ
in HF(Uαβγ ),
where Uαβ = Uα ∩ Uβ and Uαβγ = Uα ∩ Uβ ∩ Uγ as usual. We call an object in KF(X ) a vectorial bundle over X .
868
K. Gomi
(2) A homomorphism ({Uα }, (E α , h α ), φα β ) → ({Uα }, (E α , h α ), φαβ ) consists of homomorphisms ψαα : (E α , h α ) → (E α , h α ) in HF(Uα ∩ Uα ) such that the following diagrams commute in HF(Uα ∩ Uα ∩ Uβ ) and HF(Uα ∩ Uβ ∩ Uα ), respectively: E α O φα β
E β
ψαα
/ Eα ~> ~ ~~ ~~ψαβ ~ ~
Eα }> O } }} φαβ }} }} / Eβ
ψαβ
E β
ψββ
An isomorphism of vectorial bundles is a homomorphism in KF(X ) admitting an inverse. Vectorial bundles E0 and E1 over X are said to be isomorphic if there exists an isomorphism E0 → E1 . To indicate the relationship, we will write E0 ∼ = E1 . Vectorial bundles E0 and E1 are said to be homotopic if there exists E˜ ∈ KF(X × I ) such that ˜ X ×{i} ∼ E| = Ei for i = 0, 1. We will write [E] for the homotopy class of a vectorial bundle E ∈ KF(X ). Lemma 3.3 ([13]). Let KF(X ) be the homotopy classes of vectorial bundles on X . Then KF(X ) is an abelian group. Proof. The addition in KF(X ) is given by the direct sum of vector bundles, and the inverse by reversing the Z2 -grading in vector bundles. Then the present lemma will be clear, except for the consistency of the definition of the inverse. To see it, we define (F, η) ) by taking F = F 0 ⊕ F 1 to be F i = I × C and η : F → F to be ∈ HF(I 0t η= . We multiply (E, h) ∈ HF(X ) by (F, η) to get (E ⊗ F, h ⊗ id F + ⊗ η) ∈ t 0 HF(X × I ), where : E → E acts on the even part E 0 of E = E 0 ⊕ E 1 by 1 and the odd part E 1 by −1. Then, as a homotopy, the object above connects the trivial object in HF(X ) with (E, h) ⊕ (E ∨ , h ∨ ), where (E ∨ , h ∨ ) stands for (E, h) with its Z2 -grading reversed. We can readily globalize this construction, so that the inverse is well-defined. For a Z2 -graded vector bundle E over X , we can construct a vectorial bundle over X by taking an open cover U of X to be X itself and a Hermitian map h : E → E of degree 1 to be h = 0. This construction of vectorial bundles induces a well-defined homomorphism K (X ) → KF(X ). The following result of Furuta will be used in Subsect. 5.3, and its proof is included in the Appendix. Theorem 3.4 ([13]). If X is compact, then K (X ) → KF(X ) is bijective. Remark 5. As Definition 3.1 works without Z2 -grading, the vectorial bundles in Definition 3.2 should be called Z2 -graded vectorial bundles. However, we drop the adjective “Z2 -graded”, since ungraded ones will not appear in this paper. 3.2. Twisted vectorial bundle. Definition 3.5. Let X be a topological space, P → X a principal PU (H)-bundle, and U ⊂ X a subset.
Twisted K -Theory and Finite-Dimensional Approximation
869
(a) We define the category P(U ) as follows. The objects in P(U ) consist of sections s : U → P|U . The morphisms in P(U ) are defined by HomP (U ) (s, s ) = {g : U → U (H)| s π(g) = s}, where π : PU (H) → U (H) is the projection. The composition of morphisms is defined by the pointwise multiplication. (b) We define the category HF P (U ) as follows. The objects in HF P (U ) are the same as those in P(U ) × HF(U ): Obj(HF P (U )) = Obj(P(U )) × Obj(HF(U )). The homomorphisms in HF P (U ) are defined by: H om HF P (U ) ((s, (E, h)), (s , (E , h )))
= HomP (U ) (s, s ) × HomHF (U ) ((E, h), (E , h ))/ ∼,
where the equivalence relation ∼ identifies (g, φ) with (gζ, φζ ) for any U (1)-valued map ζ : U → U (1). Definition 3.6. Let X be a paracompact space, and P → X a principal PU (H)-bundle. We define the category KF P (X ) as follows: (1) An object (U, Eα , αβ ) in KF(X ) consists of an open cover U = {Uα }α∈A of X , objects Eα in HF P (Uα ), and homomorphisms αβ : Eβ → Eα in HF P (Uαβ ) such that: αβ βα = 1
in HF P (Uαβ );
αβ βγ = αγ
in HF P (Uαβγ ).
We call an object in the category KF P (X ) a twisted vectorial bundle over X twisted by P, or a P-twisted vectorial bundle over X . (2) A homomorphism ({Uα }, Eα , α β ) → ({Uα }, Eα , αβ ) consists of homomorphisms αα : Eα → Eα in HF P (Uα ∩ Uα ) such that the following diagrams commute in HF P (Uα ∩ Uα ∩ Uβ ) and HF P (Uα ∩ Uβ ∩ Uα ), respectively: Eα O α β
Eβ
αα
/ Eα ? αβ
? EOα ~~ ~ αβ ~~ ~~ / Eβ
αβ
Eβ
ββ
It may be helpful to give a more explicit description than that in the definition above. We can describe a twisted vectorial bundle as the data (U, sα , gαβ , (E α , h α ), φαβ ) consisting of: • an open cover U = {Uα } of X ; • local sections sα : Uα → P|Uα ; • lifts gαβ : Uαβ → U (H) of the transition functions g¯ αβ ;
870
K. Gomi
• Z2 -graded Hermitian vector bundles E α → Uα of finite rank; • Hermitian maps h α : E α → E α of degree 1; • maps φαβ : E β |Uαβ → E α |Uαβ such that h α φαβ = φαβ h β and: φαβ φβα 1 φαβ φβγ z αβγ φαγ
on Uαβ ; on Uαβγ .
In the above, the transition function g¯ αβ : Uαβ → PU (H) is defined by sα g¯ αβ = sβ . A lift gαβ of g¯ αβ means a function gαβ : Uαβ → U (H) such that π ◦ gαβ = g¯ αβ . The function z αβγ : Uαβγ → U (1) is defined by gαβ gβγ = z αβγ gαγ . Note that the data sα , gαβ of P are crucial in considering isomorphisms classes of twisted vectorial bundles. Definition 3.7. We denote by KFP (X ) the homotopy classes of twisted vectorial bundles over X twisted by P. The notion of homotopies of P-twisted vectorial bundles over X is formulated by using (P × I )-twisted vectorial bundles over X × I . As in the case of KF(X ), the set KFP (X ) gives rise to an abelian group. Clearly, if P is trivial, then a trivialization P∼ = X × PU (H) induces an isomorphism K FP (X ) ∼ = K F(X ). Remark 6. Consider the following property of a topological space X : (L) For any principal PU (H)-bundle P → X and an open cover of X , there is a refinement U = {Uα } of the cover such that we can find local trivializations sα : Uα → P|Uα and lifts gαβ of the transition functions g¯ αβ . As P is locally trivial, the existence of lifts gαβ matters only. In general, paracompact spaces have the property (L). Thus, through this property, the paracompactness assumption in Definition 3.6 ensures that KFP (X ) is non-empty. Remark 7. The assignment of P(U ) to each open set U ⊂ X gives a U (1)-gerbe over X , where U (1) is the sheaf of germs of U (1)-valued functions. In general, for a U (1)-gerbe G, we can construct a category KF G (X ) similar to KF P (X ). On a manifold X , the assignment U → KF G (U ) becomes a stack and gives the generalization of Brylinski’s 2-vector bundle mentioned in Sect. 1. 4. Cohomology Theory KF By means of KFP (X ), we construct in this section a certain generalized cohomology theory similar to twisted K -cohomology theory. Then we describe and prove some basic properties. 4.1. Construction. Let X be a paracompact space, and P → X a principal PU (H)bundle. We define the support of a twisted vectorial bundle E = (U, sα , gαβ , (E α , h α ), φαβ ) ∈ KF P (X ) to be: SuppE = {x ∈ X | (h α )x is not invertible for some α }.
Twisted K -Theory and Finite-Dimensional Approximation
871
For a (closed) subspace Y ⊂ X , we denote by KF P (X, Y ) the full subcategory in KF P (X ) consisting of objects E such that SuppE ∩ Y = ∅. Then we define KFP (X, Y ) to be the homotopy classes of objects in KF P (X, Y ), where homotopies are given by objects in KF P×I (X × I, Y × I ). For n ≥ 0, we put: KFP−n (X, Y ) = KFP×I n (X × I n , Y × I n ∪ X × ∂ I n ). We also put KFP1 (X, Y ) = KFP−1 (X, Y ). By means of the pull-back, a morphism ( f, F) : (X , Y ; P ) → (X, Y ; P) in C clearly induces a homomorphism ( f, F)∗ : KFPn (X, Y ) → KFPn (X , Y ). In the case of P = f ∗ P and F = f , we will write f ∗ for the induced homomorphism. Proposition 4.1. The assignment of {KFPn (X, Y )}n≤1 to (X, Y ; P) ∈ C has the following properties: (1) Homotopy axiom. If ( f i , Fi ) : (X , Y ; P ) → (X, Y ; P), (i = 0, 1) are homotopic, then the induced homomorphisms coincide: ( f 0 , F0 )∗ = ( f 1 , F1 )∗ . (2) Excision axiom. For subcomplexes A, B ⊂ X , the inclusion map induces the isomorphism: n n KFP| (A ∪ B, B) ∼ (A, A ∩ B). = KFP| A∪B A
(3) “Exactness” axiom. There is the natural complex of groups: δ−1
δ0
−1 0 · · · → KFP| (Y ) → KFP0 (X, Y ) → KFP0 (X ) → KFP| (Y ) → KFP1 (X, Y ). Y Y 0 (Y ). This complex is exact except at the term KFP| Y the inclusion maps X λ → (4) Additivity axiom. For a family {(X λ , Yλ ; Pλ )}λ∈ in C, X induce the natural isomorphism: λ λ −n ( KF P λ
λ
λ X λ,
λ Yλ )
∼ =
λ
KFP−n (X λ , Yλ ). λ
The homotopy axiom and the additivity axiom follow directly from the definition of KFP (X, Y ). The excision axiom and the “exactness” axiom will be shown in the following subsections. Remark 8. The Bott periodicity for KFP−n (X, Y ) is not yet established at this stage. This is the reason that the “exactness” axiom in Proposition 4.1 is formulated partially. At the end, the periodicity will turn out to hold, and we will obtain the complete exactness axiom. Remark 9. A generalization of KFP (X ) is given by incorporating actions of Cliffordalgebra bundles into vectorial bundles. Another generalization is to use real vector bundles with inner product instead of Hermitian vector bundles. These generalizations also satisfy properties similar to those in Proposition 4.1.
872
K. Gomi
4.2. Excision axiom. We here prove the excision axiom in Proposition 4.1. For untwisted KF(X, Y ), the excision theorem is shown in [13]. The following argument is essentially the same as that used in the untwisted case. Lemma 4.2. (Meyer-Vietoris construction) Let X be a paracompact space, P → X a principal PU (H)-bundle, and U, V ⊂ X open subsets such that U ∩ V = ∅. If E ∈ KF P|U (U ) and F ∈ KF P|V (V ) are isomorphic on U ∩ V , then there is G ∈ KF P|U ∪V (U ∪ V ) such that G|U ∼ = E and G|V ∼ = F. Proof. Suppose that E = (U, Eα , αα ) and F = (V, Fβ , ββ ). We can construct the object G = (W, Gγ , ϒγ γ ) as follows. We let W be the open cover of U ∪ V consisting of the open sets belonging to U or V. The object Gγ is Eα or Fβ . Then αα , ββ and the data of the isomorphism E|U ∩V ∼ = F|U ∩V together give the morphisms ϒγ γ . Proposition 4.3. (Excision axiom) Let X be a paracompact space, and P → X a principal PU (H)-bundle. For an open set U and a closed set Y such that U ⊂ Y ⊂ X , the inclusion i : X − U → X induces the isomorphism: ∼ =
i ∗ : KFP (X, Y ) −→ KFP| X −U (X − U, Y − U ). Proof. It suffices to construct the inverse of i ∗ . Suppose that we are given E ∈ KF P| X −U (X − U, Y − U ). We put V = X − Y and W = X − SuppE. We let O = (U, sα , gαβ , (E α , h α ), φαβ ) ∈ KF P|W (W ) be an object such that E α , h α and φαβ are trivial. Note that O represents 0 ∈ KFP|W (W ). Clearly, the support of E does not intersect V ∩ W . Thus, there is a natural isomorphism E|V ∩W ∼ = O|V ∩W , so that Lemma ˜ ˜ ˜ X −U ∼ 4.2 provides us an object E ∈ KF P (X ) such that SuppE ∩ Y = ∅ and E| = E. ˜ above Note that the construction in Lemma 4.2 is natural. Hence the construction of E behaves naturally with respect to the pull-back. Consequently, the assignment E → E˜ induces a well-defined map KFP| X −U (X − U, Y − Y ) → KFP (X, Y ), giving the inverse to i ∗ . Now, the excision axiom in Proposition 4.1 follows from the proposition above: Setting X = A∪ B, Y = B and U = B− A∩ B, we get KFP (A∪ B, B) ∼ = KFP| A (A, A∩ B), n (A, A ∩ B), (n < 0). which leads to K FPn (A ∪ B, B) ∼ = K FP| A 4.3. Exactness axiom. We show the “exactness” axiom in Proposition 4.1 in a way similar to that used in [1]. To define the connecting homomorphism δ−n , we begin with: Lemma 4.4. Let X be a paracompact space, P → X a principal PU (H)-bundle, and Z ⊂ Y ⊂ X subspaces. If Y → X is a cofibration, then we have the exact sequence: i∗
j∗
KFP (X, Y ) −→ KFP (X, Z ) −→ KFP|Y (Y, Z ), where i : (X, Z ) → (X, Y ) and j : (Y, Z ) → (X, Y ) are the inclusion maps. Proof. Clearly, j ∗ i ∗ = 0. Suppose that [E] ∈ KFP (X, Z ) is such that j ∗ ([E]) = 0. This means that there is a homotopy F˜ ∈ KF P×I (Y × I, Z × I ) connecting E|Y with a trivial object O on Y . Then we have an object G ∈ KF P×[0,1] (X ∪ {0} ∪ Y × [0, 1], Z × [0, 1])
Twisted K -Theory and Finite-Dimensional Approximation
873
such that G| X ×{0} ∼ = E and G|Y ×{1} ∼ = O. We can construct such an object G by ˜ Y ×(0,1] and the pull-back of E under the projection X × {0} ∪ applying Lemma 4.2 to F| Y × [0, 1) → X × {0}. Now, because Y → X is a cofibration, we have a map η making the following diagram commutative: X × {0} ∪ Y × [0, 1]
/ X × [0, 1] l l l l id l η vl l X × {0} ∪ Y × [0, 1] ˜ = η∗ G ∈ Then H = η(·, 1)∗ G defines [H] ∈ KFP (X, Y ). Since the homotopy H ∗ ∼ KF P×I (X × I, Z × I ) connects G| X ×{0} = E with H, we have i ([H]) = [E]. the group KF −n+1 (X, Y ) is isomorphic to Lemma 4.5. For (X, Y ; P) ∈ C, P KFP×I n (X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, Y × I n−1 × {1} ∪ X × ∂ I n−1 × I ). Proof. Recall KFP−n+1 (X, Y ) = KFP×I n−1 (X × I n−1 , Y × I n−1 ∪ X × ∂ I n−1 ) by definition. Consider the following maps: (X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, n−1 × I ) Y × I n−1 _ × {1} ∪ X × ∂ I i
(X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, (Y × I n−1 ∪ X × ∂ I n−1 ) × I ) O p
j
? (X × I n−1 × {0}, (Y × I n−1 ∪ X × ∂ I n−1 ) × {0}), where i and j are induced from the inclusions, and p from the projection. The map p = p ◦ i is also induced from the projection. We will prove below that p ∗ provides us the isomorphism in the present lemma. For the aim, we show: (a) p ∗ is injective; (b) p ∗ is surjective; and (c) i ∗ is surjective. For (a), we use the map H given by the homotopy extension property: / X × I n−1 × I X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I g g g g g id g gH g sg g X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I.
874
K. Gomi
If we put h(·) = H (·, 1), then p ◦ h is homotopic to the identity of X × I n−1 relative to Y × I n−1 ∪ X × ∂ I n−1 , so that p ∗ is injective. For (b), it is enough to apply the excision axiom. For (c), we define X ⊃ Y ⊃ Z as follows: X = X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, Y = (Y × I n−1 ∪ X × ∂ I n−1 ) × I, Z = Y × I n−1 × {1} ∪ X × ∂ I n−1 × I. Lemma 4.4 gives the exact sequence: KFP×I n (X , Y) −−−−→ KFP×I n (X , Z) −−−−→ KFP×I n (Y, Z), in which the first map coincides with i ∗ . Hence the surjectivity of i ∗ will follow from KFP×I n (Y, Z) = 0. To see this vanishing, we let (Z, Z) → (Y, Z) be the inclusion, and (Y, Z) → (Z, Z) the map given by composing the following projection and inclusion: Y −−−−→ (Y × I n−1 ∪ X × ∂ I n−1 ) × {1} −−−−→ Z. These maps give homotopy equivalences between (Y, Z) and (Z, Z), so that we have KFP×I n (Y, Z) ∼ = KFP×I n (Z, Z) = 0. the group KF −n (Y ) is isomorphic to Lemma 4.6. For (X, Y ; P) ∈ C, P|Y KFP×I n (X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, X × I n−1 × {0} ∪ Y × I n−1 × {1} ∪ X × ∂ I n−1 × I ). Proof. The present lemma straightly follows from the excision axiom.
we define the natural homomorphism Now, for (X, Y ; P) ∈ C, −n (Y ) −→ KFP−n+1 (X, Y ), (n ≥ 1) δ−n : KFP| Y
to be the composition of the isomorphism in Lemma 4.6, the following homomorphism induced from the inclusion map: KFP×I n (X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, X × I n−1 × {0} ∪ Y × I n−1 × {1} ∪ X × ∂ I n−1 × I ) ⏐ ⏐ KFP×I n (X × I n−1 × {0} ∪ (Y × I n−1 ∪ X × ∂ I n−1 ) × I, Y × I n−1 × {1} ∪ X × ∂ I n−1 × I ), and the isomorphism in Lemma 4.5. Proposition 4.7. (Exactness axiom) For (X, Y ; P) ∈ Cand n ≥ 0, we have the following exact sequences: −n (Y ). (a) KFP−n (X, Y ) −→ KFP−n (X ) −→ KFP| Y δ−n−1
−n−1 (b) KFP| (Y ) −→ KFP−n (X, Y ) −→ KFP−n (X ). Y δ−n−1
−n−1 (c) KFP−n−1 (X ) −→ KFP| (Y ) −→ KFP−n (X, Y ). Y
Twisted K -Theory and Finite-Dimensional Approximation
875
−n In the above, the maps KFP−n (X, Y ) → KFP−n (X ) and KFP−n (X ) → KFP| (Y ) are Y induced from the inclusions (X, ∅) → (X, Y ) and Y → X , respectively.
Proof. We define X ⊃ Y ⊃ Z to be X = X × I n , Y = Y × I n ∪ X × ∂ I n and Z = X × ∂ I n . Then we consider the diagram: KFP×I n (X , Y) −−−−→ KFP×I n (X , Z) −−−−→ KFP×I n (Y, Z) ⏐
⏐
KFP−n (X, Y ) −−−−→
KFP−n (X )
−−−−→
−n KFP| (Y ), Y
where the upper row is the exact sequence in Lemma 4.4, the lower row is the sequence in (a), and the third vertical map is the isomorphism in the excision axiom. The diagram above commutes, so that (a) is proved. We can prove (b) and (c) in the same way. For (b), we use X ⊃ Y ⊃ Z given by: X = X × I n × {0} ∪ (Y × I n ∪ X × I n ) × I, Y = X × I n × {0} ∪ Z, Z = Y × I n × {1} ∪ X × ∂ I n × I. For (c), we use X ⊃ Y ⊃ Z given by X = X × I n × [−1, 0] ∪ X × I n × {0} ∪ Y × I n × [0, 1], Y = X × I n × {−1} ∪ X × I n × {0} ∪ Y × I n × [0, 1] ∪ X × ∂ I n × [−1, 0], Z = X × I n × {−1} ∪ Y × I n × {1} ∪ Y × ∂ I n × [0, 1] ∪ X × ∂ I n × [−1, 0]. Then we obtain (b) and (c) by identifying the groups in the exact sequence in Lemma 4.4, and checking the compatibility of the identifications. The details of the check are left to the reader (cf. [1]). Finally, to complete the proof of the “exactness” axiom in Proposition 4.1, we extend the exact sequence obtained so far in Proposition 4.7 as a complex. For the purpose, we let F = F 0 ⊕ F 1 be the Z2 -graded Hermitian vector bundle over the unit disk D 2 ⊂ C defined by F i = D2 × C. We also let T : F → F be the Hermitian map of degree 1 0 z¯ . As is known [1], the “Thom class” (F, T ) represents a generator given by Tz = z 0 of K (D 2 , S 1 ) ∼ = Z. For a moment, let X be a paracompact space and P → X a principal PU (H)-bundle. Multiplying E = ({Uα }α∈A, sα , gαβ , (E α , h α ), φαβ ) ∈ KF P (X ) and (F, T ), we get the following object β(E) in KF P×D 2 (X × D 2 , X × S 1 ): ∗ ∗ ˆ ∗ 2 T ), π X∗ φαβ ⊗ 1), ({Uα × D 2 }α∈A, π X∗ sα , π X∗ gαβ , (π X∗ E α ⊗π D 2 F, π X h α ⊗π D
where π X and π D 2 are the projections from X × D 2 to X and D 2 respectively. The ˆ ∗ 2 T of degree 1, acting on the Z2 -graded tensor product Hermitian map π X∗ h α ⊗π D ∗ ∗ π X E α ⊗π D 2 F, is given by: ˆ ∗ 2 T = π X∗ h α ⊗ 1 + ⊗ π ∗ 2 T, π X∗ h α ⊗π D D
876
K. Gomi
where is 1 on the even part of π X∗ E α , and −1 on the odd part. The assignment E → β(E) gives rise to a functor, and induces a natural homomorphism: β : KFP (X ) −→ KFP−2 (X ) = KFP×D 2 (X × D 2 , Y × S 1 ). Now, we complete the proof of the “exactness” axiom in Proposition 4.1: we define δ0 : KF 0 (Y ) → KF 1 (X, Y ) to be Proposition 4.8. For (X, Y ; P) ∈ C, P|Y P δ0 = δ−2 ◦ β. Then the following maps compose to give the trivial map, i.e. δ0 ◦ i ∗ = 0. i∗
δ0
0 (Y ) − KFP0 (X ) −−−−→ KFP| −−−→ KFP1 (X, Y ). Y
Proof. Notice the commutative diagram: i∗
δ0
i∗
δ−2
0 (Y ) − KFP0 (X ) −−−−→ KFP| −−−→ KFP1 (X, Y ) ⏐Y ⏐
⏐β ⏐
β
−2 KFP−2 (X ) −−−−→ KFP| (Y ) −−−−→ KFP−1 (X, Y ). Y
Now, Proposition 4.7 (c) completes the proof.
5. Finite-Dimensional Approximation In this section, we construct a natural transformation between K P∗ (X, Y ) and KFP∗ (X, Y ). The key to the construction is a notion of a finite-dimensional approximation of a family of Fredholm operators. We then study some properties of the natural transformation to prove our main theorem (Theorem 5.13).
5.1. Approximation of family of Fredholm operators. First of all, we introduce some ˆ the space of notations: let Hˆ be the Z2 -graded Hilbert space Hˆ = H ⊕ H, and F(H) ˆ self-adjoint bounded operators on H of degree 1 whose square differ from the identity by compact operators: ˆ = { Aˆ : Hˆ → H| ˆ bounded, self-adjoint, degree 1, Aˆ 2 − 1 ∈ K(H)}. ˆ F(H) ∗ ˆ through the assignment A → Aˆ = 0 A . We identify F(H) with F(H) A 0 2 For A ∈ F(H), we write ρ( Aˆ ) for the resolvent set of the operator Aˆ 2 , and σ ( Aˆ 2 ) = C−ρ( Aˆ 2 ) for the spectrum set. If µ is such that 0 < µ < 1, then σ ( Aˆ 2 )∩[0, µ) consists of a finite number of eigenvalues, since Aˆ 2 − 1 is compact. In particular, corresponding eigenspaces are finite-dimensional, and so is the following direct sum: ˆ A) ˆ Aˆ 2 ξ = λξ }. ˆ <µ = (H, Ker( Aˆ 2 − λ) = {ξ ∈ H| λ<µ
λ<µ
The purpose of this subsection is to establish:
Twisted K -Theory and Finite-Dimensional Approximation
877
Proposition 5.1. Let X be a topological space, and A : X → F(H) a continuous map. For an open set U ⊂ X and a number µ ∈ (0, 1) ∩ x∈U ρ( Aˆ 2x ) given, the family ˆ Aˆ x )<µ ⊂ U × Hˆ gives rise to a (finite rank, Z2 -graded, of vector spaces x∈U (H, Hermitian) vector bundle over U . ˆ Aˆ x )<µ approximates the original family { Aˆ x }. The restriction of Aˆ x to (H, Notice that the next lemma ensures the hypothesis in the proposition: Lemma 5.2. Let A : X → F(H) be a continuous map. For each point x0 ∈ X and a number µ ∈ ρ( Aˆ 2x0 ), there is an open neighborhood U of x0 such that: µ∈
ρ( Aˆ 2x ).
x∈U
ˆ norm , Proof. This lemma follows from the following facts: (i) the map X → B(H) (x → Aˆ 2x − µ) is continuous; (ii) the operator Aˆ 2x0 − µ is invertible; and (iii) invertible ˆ norm . bounded operators on Hˆ form an open subset in B(H) For the proof of Proposition 5.1, we show some lemmas. Lemma 5.3. Let U and µ be as in Proposition 5.1. For each point x0 ∈ U , there exists an open neighborhood V ⊂ U of x0 such that: ˆ Aˆ x )<µ = dim(H, ˆ Aˆ x0 )<µ dim(H,
< +∞
for all x ∈ V . ˆ Aˆ x0 )<µ is finite-dimensional. We put Proof. Because Aˆ 2x0 − 1 is compact, (H, ˆ Aˆ x0 )<µ . Let λ1 (x) ≤ λ2 (x) ≤ · · · denote eigenvalues of Aˆ 2x , where each r = dim(H, eigenvalue is included as many times as the dimension of its eigenspace. As is known, λk (x) is a continuous function in x, because of the expression: λk (x) =
sup
E⊂H dimE=k−1
u, Aˆ 2x u . |u|2 u∈E ⊥ −{0} inf
We choose ε so as to be 0 < 2ε < min{µ − λr (x0 ), λr +1 (x0 ) − µ}, and define the open set V such that x0 ∈ V ⊂ U to be: V =
r +1
{x ∈ U | |λi (x) − λi (x0 )| < ε}.
i=1
ˆ Aˆ x )<µ = r . For all x ∈ V , we have λr (x) < µ < λr +1 (x), so that dim(H,
Lemma 5.4. Let U and µ be as in Proposition 5.1. The orthogonal projections πx : ˆ Aˆ 2x )<µ constitute the continuous map Hˆ → Hˆ onto (H, π = {πx }x∈U : U −→ B(H)norm .
878
K. Gomi
Proof. It suffices to prove that, for a point x0 ∈ U , we have πx − πx0 → 0 as x → x0 . For this aim, we choose ε and V as in the proof of Lemma 5.3. If x ∈ V , then πx has the expression: 1 πx = R(z; Aˆ 2x )dz, 2πi C where R(z; Aˆ 2x ) = (z − Aˆ 2x )−1 is the resolvent, and C is a counterclockwisely oriented circle in C such that: its center lies on the real axis; and the open disks B(λi (x0 ); 2ε), (i = 1, . . . , r ) are inside C, but B(λr +1 (x0 ); 2ε) is outside. Notice ε < |z − λ| for (z, λ) ∈ C × x∈V σ ( Aˆ 2x ). Thus, for (z, x) ∈ C × V , we have: 1 R(z; Aˆ 2x )u v 1 = < . = sup ˆ 2x )v 2 (z− A ˆ u ε (z − A )v u=0 v=0 x inf v=0 v
R(z; Aˆ 2x ) = sup
Now, thanks to the integral expression of πx , we get a constant M such that πx −πx0 ≤ M Aˆ 2x − Aˆ 2x0 for x ∈ V . Hence πx − πx0 → 0 as x → x0 . Lemma 5.5. Let U and µ be as in Proposition 5.1. For x0 ∈ U , there is an open neighˆ Aˆ x0 )<µ induces an borhood W ⊂ U of x0 such that: the projection p : Hˆ → (H, ∼ ˆ ˆ ˆ ˆ isomorphism (H, A x )<µ = (H, A x0 )<µ for all x ∈ W . ˆ Aˆ x0 )<µ . We write p ⊥ : Hˆ → Proof. Let F ⊥ be the orthogonal complement of F = (H, F ⊥ for the projection, and i ⊥ : F ⊥ → Hˆ for the inclusion. The operator πx⊥ = 1 − πx is apparently Fredholm, and x → πx⊥ is norm continuous by Lemma 5.4. Thus, in the same way as that used in the appendix of [1], we can find an open neighborhood V of x0 such that: the map p ⊥ πx⊥ i ⊥ : F ⊥ → F ⊥ is bijective for all x ∈ V . Now, by the map of exact sequences: i⊥
0 −−−−→ F ⊥ −−−−→ ⏐ ⏐ p ⊥ πx⊥ i ⊥
p Hˆ −−−−→ ⏐ ⏐ ⊥ ⊥ p πx
0 −−−−→ F ⊥
F ⊥ −−−−→ 0 −−−−→ 0,
F −−−−→ 0 ⏐ ⏐
we see that p induces an isomorphism Ker p ⊥ πx⊥ ∼ = F for x ∈ V . Note that ⊥ ⊥ ⊥ ˆ Aˆ x )<µ . By Lemma 5.3, the dimension of (H, ˆ Aˆ x )<µ is Ker p πx ⊃ Kerπx = (H, ˆ Aˆ x0 )<µ , provided that x ∈ V . Thus, p induces an isomorequal to that of F = (H, ˆ Aˆ x )<µ ∼ ˆ Aˆ x0 )<µ for all x ∈ W = V ∩ V . phism (H, = (H, ˆ Aˆ x )<µ is locally Proof of Proposition 5.1. It suffices to see that the family x∈U (H, trivial. We consider the open neighborhood W of a point x0 ∈ U in Lemma 5.5. Then, ˆ Aˆ x0 )<µ induces a local trivialization on W , the map id × p : W × H → W × (H, ˆ ˆ ˆ ˆ x∈W (H, A x )<µ → W × (H, A x0 )<µ . Remark 10. Instead of F(H), we can use the space of Fredholm operators with the norm topology to obtain the same claim as Proposition 5.1. A key to this case is that 0 is a discrete spectrum of a non-invertible Fredholm operator.
Twisted K -Theory and Finite-Dimensional Approximation
879
5.2. Natural transformation. Let X be a paracompact space, and P → X a principal PU (H)-bundle. We construct a natural homomorphism: α : K P (X ) −→ KFP (X ) as follows: suppose that a section A ∈ (X, P × Ad F(H)) is given. We choose an open cover U = {Uα }x∈A such that there are local sections sα : Uα → P|Uα and lifts of transition functions gαβ : Uαβ → U (H). The local sections of P allow us to identify −1 on Uαβ . A with a collection of maps Aα : Uα → F(H) such that Aα = gαβ Aβ gαβ Because of Lemma 5.2, taking a refinement of U if necessary, we can find a positive number µα such that µα ∈ x∈Uα ρ(( Aˆ 2α )x ). By Proposition 5.1, we get a finite rank ˆ ( Aˆ α )x )<µα over Uα . The restricZ2 -graded Hermitian vector bundle E α = x∈Uα (H, tion of Aˆ α to E α defines a Hermitian map h α : E α → E α of degree 1. On Uαβ , we define φαβ : E β → E α to be the composition of the maps: id×g ˆ ( Aˆ β )x )<µβ → Uαβ × Hˆ −→αβ Uαβ × Hˆ → ˆ ( Aˆ α )x )<µα , (H, (H, x∈Uαβ
x∈Uαβ
where the first and third maps are the inclusion and projection, respectively. The data E = (U, sα , gαβ , (E α , h α ), φαβ ) is a P-twisted vectorial bundle over X . The isomorphism class of E is independent of the choice of µα , sα , gαβ and U. Now, the homomorphism α : K P (X ) → KFP (X ) is given by α([A]) = [E]. The same construction yields a natural map α : K P (X, Y ) → KFP (X, Y ) for Y ⊂ X , and hence αn : K Pn (X, Y ) → KFPn (X, Y ). Lemma 5.6. For a paracompact space X and a principal PU (H)-bundle P → X , the following diagram commutes: β
K P (X ) −−−−→ K P×D 2 (X ⏐ ⏐ α
× D2, X × S1) ⏐ ⏐α
β
KFP (X ) −−−−→ KFP×D 2 (X × D 2 , X × S 1 ), where the upper map β induces the Bott periodicity K P0 (X ) ∼ = K P−2 (X ). Before the proof of this lemma, we explain the map β inducing the Bott periodicity for twisted K -cohomology. Roughly, the map is a “multiplication of a Thom class”. ˆ This identification To be more precise, recall the identification of F(H) with F(H). is compatible with the conjugate actions of PU (H), through the diagonal embedding ˆ of the assoˆ So we can represent an element in K P (X ) by a section A U (H) → U (H). ˆ ˆ ˆ ciated bundle P × Ad F(H) over X . We identify the section with a map A : P → F(H). For ( p, z) ∈ P × D 2 , we define a degree 1 self-adjoint Fredholm operator β(A)( p,z) on the Z2 -graded Hilbert space Hˆ ⊗ (C ⊕ C) by ˆ p ⊗ 1 + ⊗ Tz ˆ ( p,z) = A β(A) 0 A∗p 0 z¯ 1 0 1 0 . ⊗ + ⊗ = z 0 0 −1 0 1 Ap 0 ˆ of (P × D 2 ) × Ad F(Hˆ ⊗ (C ⊕ C)) over X × D 2 , This operator defines a section β(A) and induces the Bott periodicity map β : K P (X ) → K P×D 2 (X × D 2 , X × S 1 ).
880
K. Gomi
ˆ ∈ (X, P × Ad F(H)) ˆ be a section given. First, we describe Proof of Lemma 5.6. Let A ˆ ∈ KF −2 (X ) as follows: let an object βα([A]) P
E = ({Uα }α∈A, sα , gαβ , (E α , h α ), φαβ ) ∈ KF P (X ) ˆ represent the element α([A]) ∈ KFP (X ). We suppose that E α is given by ˆ ( Aˆ α )x )<µα under a choice of a positive number µα . Taking a finer E α = x∈Uα (H, open cover if necessary, we can assume that the rank of E α is rα , and that there exists a positive number α such that α < min{µα − λrα (x), λrα +1 (x) − µα } for all x ∈ Uα . Here λ j (x) is the j th eigenvalue of ( Aˆ 2α )x , which varies continuously in x (cf. Lemma 5.3). We define an open cover {V (s; α )}s∈[0,1] of D 2 by setting V (s; α ) = {z ∈ D 2 | s − α < |z|2 < s + α }, s ∈ [0, 1]. ˆ ∈ KF −2 (X ) by Then we can represent βα([A]) P ˆ ∗ 2 T ), π X∗ φαβ ). ({Uα × V (s; α )}, π X∗ sα , π X∗ gαβ , (π X∗ E α ⊗ (C ⊕ C), π X∗ h α ⊗π D ˆ ∈ KF −2 (X ). In applying Proposition 5.1 to the Next, we consider the element αβ([A]) P 2 ˆ of (P × D )× Ad F(H⊗(C⊕C)), ˆ section β(A) we use the open set Uα ×V (s; α ) and the positive number µα +s. The j th eigenvalue of the square of ( Aˆ α )x ⊗1+⊗Tz is λ j (x)+|z|2 . Since λrα (x) + |z|2 < µα + s < λrα +1 (x) + |z|2 holds for (x, z) ∈ Uα × V (s; α ) by construction, we obtain: ˆ ( Aˆ α )x )<µ ⊗ (C ⊕ C). (Hˆ ⊗ (C ⊕ C), ( Aˆ α )x ⊗ 1 + ⊗ Tz )<µα +s ∼ = (H, ˆ also represents αβ([A]). ˆ Hence the representative of βα([A])
Proposition 5.7. The homomorphisms αn : K Pn (X, Y ) → KFPn (X, Y ), (n ≤ 1) constitute a natural transformation of cohomology theories. Proof. It suffices to see that the natural homomorphisms αn are compatible with the axioms in Proposition 2.2 and 4.1. The homotopy axioms, the excision axioms and the additivity axioms are clearly compatible with αn . For n < 0, inclusion maps define δn , so that δn αn = αn+1 δn . This formula also holds for n = 0, because of Lemma 5.6. As a result, the “exactness” axioms are compatible with αn . 5.3. Finite-dimensional approximation in untwisted case. In untwisted case, the map α has the following property: Proposition 5.8. If (X, Y ; P) ∈ Cis such that P → X is trivial, then the homomorphism α : K P (X, Y ) → KFP (X, Y ) is bijective. If P is trivial, then we can identify K P (X, Y ) with the set of the homotopy classes of maps A : X → F(H) such that A y , (y ∈ Y ) is the identify. Accordingly, we identify the map in Proposition 5.8 with: α : [(X, Y ), (F(H), 1)] −→ KF(X, Y ), where the group KF(X, Y ) consists of homotopy classes of vectorial bundles whose supports do not intersect Y . For the proof of Proposition 5.8, we notice Furuta’s result:
Twisted K -Theory and Finite-Dimensional Approximation
881
Theorem 5.9 ([13]). For a compact space X and its closed subspace Y ⊂ X , there is an isomorphism K (X, Y ) → KF(X, Y ). The K -group K (X, Y ) above is formulated by means of vector bundles, rather than Fredholm operators. In the case of Y = ∅, Theorem 5.9 gives Theorem 3.4. The proof of Theorem 5.9 is also included in the Appendix. Thanks to the above result of Furuta, we have: Lemma 5.10. For a compact space X , the map α : [X, F(H)] → KF(X ) is bijective. Proof. For (a), we consider the following diagram: [X, F(H)] MMM rr MMMα MMM rr r M& r yr / KF(X ), K (X ) indrrr
where ind : [X, F(H)] → K (X ) is the isomorphism constructed in [1,3]. The homomorphism K (X ) → KF(X ) is introduced in Subsect. 3.1. The method showing the sur jectivity of ind in [1] allows us to realize any vector bundle E → X as E = x∈X Ker Aˆ x by means of a map A : X → F(H) such that σ ( Aˆ 2x ) = {0, 1} for all x ∈ X . Thus, the above diagram is commutative, so that Theorem 3.4 implies the present lemma. Lemma 5.11. For (X, pt) ∈ C, the map α : [(X, pt), (F(H), 1)] → KF(X, pt) is bijective. Proof. We use the exact sequences for (X, pt). By Proposition 5.7, the diagram [(I, ∂ I ), (F(H), 1)] −−−−→ [(X, pt), (F(H), 1)] −−−−→ [X, F(H)] ⏐ ⏐ ⏐ ⏐α ⏐α ⏐α KF(I, ∂ I )
−−−−→
KF(X, pt)
−−−−→
KF(X )
is commutative. Because F(H) is a representing space for K -theory [3], we have π1 (F(H), 1) = 0. By Theorem 5.9, we also have KF(I, ∂ I ) ∼ = K (I, ∂ I ) = 0. Hence Lemma 5.10 leads to the present lemma. The following is also a result of Furuta: Lemma 5.12 ([13]). If X is compact and Y ⊂ X is closed, then the quotient map q : X → X/Y induces an isomorphism KF(X, Y ) ∼ = KF(X/Y, pt). Proof. Under the assumption, the topology of (X/Y ) − pt induced from X/Y coincides with the topology of X − Y induced from X . Hence the isomorphism classes in KF(X, Y ) correspond bijectively to those in KF(X/Y, pt) via q. Since this correspon dence respects homotopies, the lemma is proved. Proof of Proposition 5.8. The quotient q : X → X/Y gives the diagram: [(X, Y ), (F(H), 1)] ⏐ q∗⏐
α
−−−−→
KF(X, Y ) ⏐ ∗ ⏐q
[(X/Y, pt), (F(H), 1)] −−−−→ KF(X/Y, pt). α
882
K. Gomi
By Lemma 5.12, the right q ∗ is bijective. Since Y → X is a cofibration, the left q ∗ is also bijective, and the diagram above is commutative. Now Lemma 5.11 establishes Proposition 5.8. 5.4. Main theorem. Theorem 5.13. For a CW complex X and a principal PU (H)-bundle P → X , the homomorphism α−n : K P−n (X ) → KFP−n (X ), (n ≥ 0) is bijective. To prove this theorem, we begin with the case that X is finite. Lemma 5.14. Let X = eq ∪ Y be a finite CW complex given by attaching a q-cell to another finite CW complex Y . Then, for any principal PU (H)-bundle P → X , we have natural isomorphisms compatible with α−n : K P−n (X, Y ) ∼ = K −n (D q , S q−1 ), KFP−n (X, Y ) ∼ = KF −n (D q , S q−1 ). Proof. Let f : D q → X denote the map attaching the q-cell. By definition, f induces a homeomorphism from D q − S q−1 to its image eq = f (D q − S q−1 ). We write e¯q for the closure of eq in X , and ∂ e¯q for its boundary. Regarding D q as the unit disk in Rq , we decompose it as D q = D ∪ A, where D is the disk of radius 1/2, and A the annulus q q whose radius r ranges from 1/2 ≤ r ≤ 1. Then, setting D+ = f (D ) and e¯− = f (A), q q q q q we can describe e¯ as the union e¯ = D+ ∪ e¯− . Since ∂ e¯ is homotopy equivalent to q e¯− , the excision axiom gives q q q −n −n −n K P−n (X, Y ) ∼ (e¯q , ∂ e¯q ) ∼ (e¯q , e¯− ) ∼ (D+ , ∂ D+ ). = K P| = K P| = K P| q q q e¯
e¯
D+
Because all principal PU (H)-bundles over the disk
q D+
−n are trivial, we have K P| q
D+
(D+ , ∂ D+ ) ∼ = K −n (D+ , ∂ D+ ), and hence the first isomorphism. The same argument is valid for the second isomorphism. These isomorphisms come from the axioms in Proposition 2.2 and 4.1, so that the compatibility with α−n follows. q
q
q
q
Lemma 5.15. Let X be a finite CW complex. Then, for a principal PU (H)-bundle P → X , the homomorphism α−n : K P−n (X ) → KFP−n (X ), (n ≥ 0) is bijective. Proof. We prove this lemma by an induction on the number r of cells in X . If r = 1, then X consists of a point, so that α−n , (n ≥ 0) is bijective by Proposition 5.8. If r > 1, then we can express X as X = eq ∪ Y , where eq is a q-dimensional cell, and Y is a subcomplex with (r −1) cells. Proposition 5.7 gives the following commutative diagram for n ≥ 0: −n−1 K P| (Y ) Y
/ K −n (X, Y ) P
/ K −n (X ) P
/ KF −n (X, Y ) P
−n−1 KFP| (Y ) Y
α−n
/ KF −n (X ) P
/ K −n (Y ) P|Y
/ K −n+1 (X, Y ) P
/ KF −n (Y ) P|Y
/ KF −n+1 (X, Y ) P
As the hypothesis of the induction, we assume the first and fourth vertical maps are bijective. Lemma 5.14 and Proposition 5.8 imply that the second and fifth maps are bijective. Therefore the third map is also bijective. Notice that the five lemma works 0 (Y ). even if the lower row is not exact at KFP| Y
Twisted K -Theory and Finite-Dimensional Approximation
883
Proof of Theorem 5.13. Let X q ⊂ X be the subcomplex consisting of cells whose dimensions are less than or equal to q. Identifying X q × {q + 1} ⊂ X q × [q, q + 1] with X q × {q + 1} ⊂ X q+1 × [q + 1, q + 2], we get the “telescope” X˜ of X : X˜ = X 0 × [0, 1] ∪ X 1 × [1, 2] ∪ X 2 × [2, 3] ∪ · · · . In a similar way, we get a principal PU (H)-bundle P˜ → X˜ from P → X . As is known [18], the projections X q ×[q, q +1] → X q induce a homotopy equivalence : X˜ → X . −n ˜ we have K −n ( X˜ ) ∼ ˜ ∼ Since ∗ P ∼ = P, = K P−n (X ) as well as KFP−n ˜ ( X ) = KFP (X ). P˜ Thus, to show Theorem 5.13, it suffices to prove that α−n : K −n ( X˜ ) → KF −n ( X˜ ) P˜
P˜
is bijective for n ≥ 0. For this purpose, we let Y˜ be the subcomplex in X˜ given by Y˜ = q X q = q X q × {q}. For n ≥ 0, Proposition 5.7 gives: (Y˜ ) K −n−1 ˜ P|Y˜
KF −n−1 (Y˜ ) ˜ P|Y˜
/ K −n ( X˜ , Y˜ ) ˜
˜ / K −n ˜ (Y )
/ K −n+1 ( X˜ , Y˜ ) ˜
˜ / KF −n ˜ (Y )
/ KF −n+1 ( X˜ , Y˜ ). ˜
/ K −n ( X˜ ) ˜
P
P|Y˜
P
α−n
/ KF −n ( X˜ , Y˜ ) ˜
/ KF −n ( X˜ ) ˜
P
P
P
P|Y˜
P
The first and fourth columns in the commutative diagram above are bijective: the additivity axiom in Proposition 2.2 implies −n ˜ ∼ K −n ( q Xq) ∼ = ˜ (Y ) = K P| q P|Y˜
q
X
−n K P| (X q ). q
q
X
−n q q (Y˜ ) ∼ Similarly, we have KF −n = q KFP| X q (X ). Because X is a finite CW com˜ P| Y˜
−n −n plex, the map α−n : K P| (X q ) → KFP| (X q ) is bijective by Lemma 5.15, and so is X q Xq −n −n α−n : q K P| q (X q ) → q KFP| q (X q ). The second and fifth columns can be shown X X to be bijective in the same way, since we have
−n K −n ( X˜ , Y˜ ) ∼ ( q X q × I, q X q × ∂ I ) = K q P| ×I P˜ X q −n −n−1 ∼ K P| q ×I (X q × I, X q × ∂ I ) = K P| (X q ). = q q
X
Now, the five lemma leads to the bijectivity of the third.
q
X
We have Theorem 1 by setting n = 0 in Theorem 5.13. We also have: Corollary 5.16. (Bott periodicity) Under the assumption in Theorem 5.13, there is a natural isomorphism KFP−n (X ) ∼ = KFP−n−2 (X ) for n ≥ 0. Thus, on CW complexes, we can extend Proposition 4.1 to get a cohomology theory {KFPn (X, Y )}n∈Z equivalent to twisted K -cohomology.
884
K. Gomi
Appendix A. Proof of Furuta’s Theorem We provide proof of Furuta’s theorems (Theorem 3.4 and 5.9) culling from [13]. Most parts of this appendix are devoted to the proof of Theorem 3.4, since the proof of Theorem 5.9 is almost the same. To prove Theorem 3.4, we begin with preliminaries in Subsect. A.1, and then construct a vector bundle from a vectorial bundle in Subsect. A.2. We use the vector bundle to prove that the map K (X ) → KF(X ) is surjective in Subsect. A.3. Finally, the injectivity of K (X ) → KF(X ) is shown in Subsect. A.4. A.1. Preliminary. Let E = ({Uα }α∈A, (E α , h α ), φαβ ) ∈ KF(X ) be a vectorial bundle on a compact space X . Taking a finer open cover if necessary, we can assume that E α is a trivial bundle E α = Uα × Vα , where Vα = Vα0 ⊕ Vα1 is a Z2 -graded Hermitian vector space of finite rank. Since X is compact, we can also assume that {Uα }α∈A is a finite cover of X . For x ∈ X , we put A(x) = {α ∈ A| x ∈ Uα }. Lemma A.1. There is a positive number λ such that: for x ∈ X and α, β ∈ A(x), the map (φαβ )x : Vβ → Vα induces an isomorphism (Vβ , (h β )x )<λ ∼ = (Vα , (h α )x )<λ . Proof. By the definition of vectorial bundles, we can find, for each point x ∈ X , an open neighborhood Ux of x and a positive number λx such that: for y ∈ Ux and α, β ∈ A(x), the map (φαβ ) y induces an isomorphism (Vβ , (h β ) y )<λx ∼ = (Vα , (h α ) y )<λx . Because X is compact, we can choose a finite number of points x1 , . . . , xn ∈ X so that Ux1 , . . . , Uxn cover X . The minimum among λx1 , . . . , λxn gives the λ. We choose and fix a positive number λ in the lemma above. Then, for x ∈ X , we define a Z2 -graded Hermitian vector space (E)x to be (E)x = (Vα , (h α )x )<λ / ∼, α∈A(x)
where ∼ is an equivalence relation: for vα ∈ Vα and vβ ∈ Vβ , we have vα ∼ vβ if and only if (φαβ )x vβ = vα . We also define a Hermitian map of degree 1, (h E )x : (E)x → (E)x , [vα ] → [(h α )x vα ], where [vα ] stands for the element in (E)x represented by vα ∈ (Vα , (h α )x )<λ . Lemma A.2. For a point x0 ∈ X and a number µ such that µ ∈ (0, λ) ∩ ρ((h E )2x0 ), there is an open neighborhood U of x0 on which the family of vector spaces ((E)x , (h E )x )<µ x∈U
gives rise to a vector bundle.
We remark that the family of vector spaces x∈X (E)x is not generally a vector bundle over X since the dimension of (E)x may jump as x varies.
Proof. We take and fix α ∈ A(x0 ). The same argument as in Subsect. 5.1 implies that there is an open neighborhood U of x0 on which x∈U (Vα , (h α )x )<µ gives rise to a vector bundle over U . Hence the natural bijection between x∈U (Vα , (h α )x )<µ and x∈U ((E)x , (h E )x )<µ establishes the lemma.
Twisted K -Theory and Finite-Dimensional Approximation
885
2 } be a partition of unity subordinate to A.2. Construction of vector bundle. Let {ρ02 , ρ∞ the open cover {[0, λ), (0, ∞)} of [0, ∞). For x ∈ X and α ∈ A(x), the functional calculus induces the Hermitian map ρ02 ((h α )2x ) : Vα → Vα . We can think of ρ02 ((h α )2x ) as an “approximation” of the orthogonal projection onto (Vα , (h α )x )<λ . In fact, the image of Vα under ρ02 ((h α )2x ) is in (Vα , (h α )x )<λ . In particular, the image of the odd part Vα1 is in (Vα , (h α )x )1<λ = Vα1 ∩ (Vα , (h α )x )<λ , since ρ02 ((h α )2x ) is of degree 0. Now, we use a partition of unity {α }α∈A subordinate to the open cover {Uα }α∈A of X to define the linear map (g)x , (x ∈ X ) as follows: 1 2 2 1 Vα1 −→ (E 1 )x , (g)x : α vα → α α (x)[ρ0 ((h α )x )vα ], α∈A
where (E)x = (E 0 )x ⊕ (E 1 )x and (E i )x = α∈A(x) (Vα , (h α )x )i<λ / ∼. By means of (g)x , we also define the linear maps Vα1 −→ (E 1 )x (g i )x : (E i )x ⊕ α∈A
to be
= ρ∞ ((h E )2x )(h E )x v 0 + (g)x ( α vα1 ), (g 1 )x (v 1 ⊕ ( α vα1 )) = ρ∞ ((h E )2x )v 1 + (g)x ( α vα1 ).
(g 0 )x (v 0 ⊕ (
1 α vα ))
Lemma A.3. The maps (g 0 )x and (g 1 )x are surjective at each x ∈ X . Proof. Consider the eigenspace decomposition (E 1 )x = κ Ker((h E )2x − κ). For each 2 (κ) = 0. In the case of ρ 2 (κ) = 0, we can see Ker((h )2 − κ, we have ρ02 (κ) = 0 or ρ∞ E x 0 1 2 (κ) = 0, we κ) ⊂ (g)x (Vα ) for an α ∈ A(x) such that α (x) = 0. In the case of ρ∞ clearly have Ker((h E )2x − κ) ⊂ (g 1 )x ((E 1 )x ). Since κ = 0 in this case, we also have Ker((h E )2x − κ) ⊂ (g 0 )x ((E 0 )x ). Accordingly, we have the following exact sequence at each x ∈ X : 0 −−−−→ (F i )x −−−−→ (E i )x ⊕
(g i )x
Vα1 −−−−→ (E 1 )x −−−−→ 0,
α∈A
where (F i )x = Ker(g i )x . The exact sequence implies that dim(F i )x is locally constant in x, because dim(E 0 )x − dim(E 1 )x is. While the family of vector spaces x∈X (E i )x is not generally a vector bundle, we have: Proposition A.4. For i = 0, 1, the family of vector spaces F i = x∈X (F i )x gives rise to a vector bundle over X . Proof. We take and fix x0 ∈ X and α0 ∈ A(x0 ). For x ∈ Uα0 , we introduce linear maps as follows: (ταi 0 )x : (E i )x −→ Vαi 0 , 1 1 (gα0 )x : α∈A Vα −→ Vα0 ,
[vαi ]
1 α vα
→ (φα0 α )x vαi , → α α (x)(φα0 α )x ρ02 ((h α )2x )vα1 .
Note that (gα0 )x = (τα10 )x ◦ (g)x . We also introduce (gαi 0 )x : Vαi 0 ⊕ Vα1 −→ Vα10 α∈A
886
K. Gomi
by setting (gα00 )x (v 0 ⊕ (
1 α vα )) (gα10 )x (v 1 ⊕ ( α vα1 ))
= ρ∞ ((h α0 )2x )(h α0 )x v 0 + (g)x ( α vα1 ), = ρ∞ ((h α0 )2x )v 1 + (g)x ( α vα1 ).
In the same way as in the proof of Lemma A.3, we see that (gαi 0 )x are also surjective for x ∈ Uα0 . Now, we have the commutative diagram: 0 −−−−→ Ker(gαi 0 )x −−−−→ Vαi 0 ⊕ ⏐ ⏐ 0 −−−−→
(F i )x
(gαi )x
Vα1 −−−−→ Vα10 −−−−→ 0 0
α∈A ⏐(τ 1 ) ⏐ ⏐ α0 x ⏐ (g i )x −−−−→ (E i )x ⊕ Vα1 −−−−→ (E 1 )x −−−−→ 0. (ταi )x ⊕id 0
α∈A
Since (ταi 0 )x is injective, so is the map (F i )x → Ker(gαi 0 )x . Because dim(E 0 )x − dim(E 1 )x = dim Vα00 − dim Vα10 , we have dim(F i )x = dim Ker(gαi 0 )x . This implies that the map (F i )x → Ker(gαi 0 )x is bijective. Consequently, we can identify F i |Uα0 with the family of vector spaces x∈Uα Ker(gαi 0 )x . Because (gαi 0 )x is continuous in 0 x ∈ Uα0 , the family x∈Uα Ker(gαi 0 )x becomes a vector bundle. Hence the identifica0
tion makes F i into a vector bundle.
A.3. Surjectivity. In this subsection, we prove: Proposition A.5. If X is compact, then K (X ) → KF(X ) is surjective. So far, a Z2 -graded vector bundle F = F 0 ⊕ F 1 over X is constructed from a given vectorial bundle E ∈ KF(X ). To prove the proposition above, it suffices to introduce a Hermitian metric and a Hermitian map h on F so that (F, h) is isomorphic to E as a vectorial bundle. For x ∈ X , we define a Hermitian metric on (E i )x ⊕ ( α∈A Vα1 ) by: 1 vα 2 , λ α vα 2 . v 1 ⊕ ( α vα1 ) 2 = v 1 2 +
v 0 ⊕ (
1 2 α vα )
= v 0 2 +
α
We induce the Hermitian metric on (F i )x by restriction. We then define (h 10 )x : (F 0 )x −→ (F 1 )x , v 0 ⊕ (
1 α vα )
→ (h E )x v 0 ⊕ (
1 α vα ).
We put h x = (h 10 )x + (h 01 )x , where (h 01 )x is the adjoint of (h 10 )x : (h 01 )x : (F 1 )x −→ (F 0 )x , v 1 ⊕ (
1 α vα )
→ (h E )x v 1 ⊕ (
α
λvα1 ).
Twisted K -Theory and Finite-Dimensional Approximation
887
Lemma A.6. Let µ0 > 0 be such that 0 (r ) = 1 for r ∈ [0, µ0 ]. Then for x ∈ X the composition of the inclusion and the projection (F i )x −→ (E i )x ⊕ Vα1 −→ (E i )x α∈A
induces an isomorphism (F, h)x,<µ0 ∼ = ((E)x , (h E )2x )<µ0 . Proof. The point x ∈ X will be fixed in this proof. So we omit subscripts x from (F i )x , (E i )x (h E )2x , and so on. The map h 2E ⊕ ( α h 2α ) is Hermitian with respect to the Hermitian metric on E i ⊕ ( α∈A Vα1 ) introduced just before this lemma. Hence we get the orthogonal decomposition Vα1 = Fˆµ = Ker(h 2E ⊕ ( α h 2α ) − µ). Ei ⊕ Fˆµ , α∈A
Since g i ◦ h 2E ⊕ (
α
µ≥0
h 2α ) = h 2E ◦ g i , we also get the orthogonal decomposition Fµ , Fµ = F ∩ Fˆµ . F=
µ≥0
Because h ◦ h 2E ⊕ ( α h 2α ) = h 2E ⊕ ( α h 2α ) ◦ h, the map h preserves the orthogonal decomposition of F. Thus, in the following, we will verify the present lemma on each space Fµ . First, we suppose µ ≥ µ0 . Then we have (F, h)<µ0 ∩ Fµ = {0}. To see this, notice λ ≥ µ0 . For a vector v 0 ⊕ ( α vα1 ) in the even part Fµ0 of Fµ , we have h 10 (v 0 ⊕ ( α vα1 )) 2 µ v 0 2 + α vα1 2 ≥ ≥ min{µ, λ} ≥ µ0 . v 0 ⊕ ( α vα1 ) 2 v 0 2 + λ1 α |vα1 2
Hence the eigenvalue of h 2 is greater than or equal to µ0 on the even part Fµ0 , and so is on the odd part Fµ1 . Next, we consider the case of µ < µ0 . Because 0 (µ) = 1 and ∞ (µ) = 0, a vector v i ⊕ ( α vα1 ) in Fˆµi belongs to Fµi if and only if α α vα1 = 0. Hence Fµi has the orthogonal decomposition Fµi = E µi ⊕ Vµ1 , where E µi = E i ∩ Ker(h 2E − µ), Vµ1 = { α vα1 ∈ α∈A Vα1 | α α vα1 = 0} ∩ Ker( α h 2α − µ). The Hermitian map h 2 preserves the decomposition. In particular, h 2 = λ on Vµ1 , so that (F, h)<µ0 ∩ Fµ = E µ . Thus, (F, h)<µ0 = (E, h E )<µ0 . The isomorphism in Lemma A.6 is compatible with the Hermitian maps (h)x and (h E )x . In addition, the isomorphism in Lemma A.6 induces an isomorphism of vector bundles locally, provided that µ0 is chosen suitably. Now, the construction of (E)x implies that (F, h) is isomorphic to E as a vectorial bundle, which completes the proof of Proposition A.5. If Y ⊂ X is a closed subspace and SuppE ∩ Y = ∅, then we also have Supp(F, h) ∩ Y = ∅. This means that we have the pair of vector bundles (F 0 , F 1 ) and h 10 : F 0 → F 1 is invertible on Y . As is known [1], such data (F 0 , F 1 , h 10 ) constitute the K -group K (X, Y ). Thus we get:
888
K. Gomi
Proposition A.7. For a compact space X and its closed subspace Y ⊂ X , there is a surjection K (X, Y ) → KF(X, Y ). A.4. Injectivity. Proposition A.8. If X is compact, then K (X ) → KF(X ) is injective. Proof. Suppose that two Z2 -graded Hermitian vector bundles F0 and F1 over X give the same element in KF(X ). Then there is a vectorial bundle F˜ on X × [0, 1] such that ˜ X ×{i} is isomorphic to (Fi , h i ) as a vectorial bundle, where h i is the trivial Hermitian F| map h i = 0. Thanks to the construction proving Proposition A.5, we can replace F˜ ˜ h), ˜ where F˜ is a Z2 -graded Hermitian vector bundle on X × [0, 1] and by a pair ( F, ˜ h)| ˜ X ×{i} as vectorial ˜h : F˜ → F˜ is a Hermitian map of degree 1. That (Fi , h i ) ∼ = ( F, 2 bundles means that Fi ∼ = Ker h˜ | X ×{i} as vector bundles. This implies that the pairs (Fi0 , Fi1 ) and ( F˜ 0 | X ×{i} , F˜ 1 | X ×{i} ) are in the same class in K (X ). Therefore the pairs (F00 , F01 ) and (F10 , F11 ) represent the same class in K (X ). Proposition A.9. For a compact space X and its closed subspace Y ⊂ X , the map K (X, Y ) → KF(X, Y ) is injective. Lemma A.10. For i = 0, 1, let Fi = Fi0 ⊕ Fi1 be a Z2 -graded Hermitian vector bundle over a compact space X , and h i : Fi → Fi a Hermitian map of degree 1. If (F0 , h 0 ) and (F1 , h 1 ) are isomorphic in KF(X ), then F00 ⊕ F11 and F01 ⊕ F10 are isomorphic as vector bundles. Proof. By the definition of the equivalences in KF(X ), we have a map g : F0 → F1 of degree 0 compatible with h 0 and h 1 such that: for each x ∈ X , there are a positive integer λx and an open neighborhood Ux of x such that: (g) y induces an isomorphism (F0 , h 0 ) y,<λx ∼ = (F1 , h 1 ) y,<λx for all y ∈ X . Since X is compact, we can find a positive number λ such that: (g)x induces an isomorphism (F0 , h 0 )x,<λ ∼ = (F1 , h 1 )x,<λ for each x ∈ X. 2 } subordinate to the We choose such λ as above, and take a partition of unity {ρ02 , ρ∞ open cover {[0, λ), (0, ∞)} of [0, ∞). The Hermitian map ρ0 ((h 1 )2x ) : (F1 )x → (F1 )x plays a role of a continuous “approximation” of (a square root of) the orthogonal projection onto (F1 , h 1 )x,<λ . In fact, the image of ρ0 ((h 1 )2x ) is in (F1 , h 1 )x,<λ . Hence the map 2 (g)−1 x ρ0 ((h 1 )x ) : (F1 )x → (F0 )x makes sense. In particular, this map is continuous in x, so that we obtain a vector bundle map g −1 ρ0 (h 21 ) : F1 → F0 of degree 0. Similarly, the image of ρ∞ ((h 1 )2x ) is in (F1 , h 1 )x,≥λ = κ≥λ Ker((h 1 )2x − κ). Thus, we obtain a vector bundle map sgn(h 1 )ρ∞ (h 21 ) : F1 → F1 of degree 1, where sgn(t) = t/|t|. Now, we define h˜ : F0 ⊕ F1 → F0 ⊕ F1 to be h˜ = h˜ 0 + h˜ ∞ , where h˜ 0 = gρ0 (h 20 ) + g −1 ρ0 (h 21 ), h˜ ∞ = sgn(h 0 )ρ∞ (h 20 ) − sgn(h 1 )ρ∞ (h 21 ). Then h˜ 20 = ρ0 (h 21 )2 + ρ0 (h 20 )2 , h˜ 2∞ = ρ∞ (h 21 )2 + ρ∞ (h 20 )2 and h˜ 0 h˜ ∞ + h˜ ∞ h˜ 0 = 0. Therefore h˜ 2 = 1 and h˜ is an isomorphism. By construction, h˜ carries the component F00 ⊕ F11 to F01 ⊕ F10 , and vice verse.
Twisted K -Theory and Finite-Dimensional Approximation
889
Proof of Proposition A.9. For i = 0, 1, we let (Fi , h i ) represent an element in K (X, Y ), where Fi = Fi0 ⊕ Fi1 is a Z2 -graded Hermitian vector bundle on X and h i : Fi → Fi is a Hermitian map of degree 1 such that h i is invertible on Y . Suppose that (Fi , h i ) define the same element in KF(X, Y ). Because X × [0, 1] is compact, the construction showing Proposition A.5 gives a Z2 -graded Hermitian vector bundle F˜ on X × [0, 1] and a Hermitian map h˜ of degree 1 such that: h˜ is invertible on Y × [0, 1] and we ˜ h)| ˜ X ×{i} in KF(X, Y ). Now, Lemma A.10 implies that (Fi , h i ) and have (Fi , h i ) ∼ = ( F, ˜ h)| ˜ X ×{i} represent the same class in K (X, Y ). Thus (F0 , h 0 ) and (F1 , h 1 ) are in the ( F, same class in K (X, Y ). Acknowledgements. I benefited a lot from discussion with M. Furuta at various stages of the work, and I am grateful to him. I thank T. Moriyama for suggestions regarding the proof of Theorem 1. I also thank A. Henriques for suggestions about the proof and for comments on a draft. I am indebted to D. Freed and B-L. Wang for useful discussions. The author’s research is supported by a JSPS Postdoctoral Fellowship for Research Abroad. Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
References 1. Atiyah, M.F.: K -theory. Lecture notes by D. W. Anderson, New York-Amsterdam: W. A. Benjamin, Inc., 1967 2. Atiyah, M.F., Hirzebruch, F.: Vector bundles and homogeneous spaces. Proc. Sympos. Pure Math., Vol. III Providence, R.I.: Amer. Math. Soc., 1961, pp. 7–38 3. Atiyah, M.F., Segal, G.: Twisted K -theory. Ukr. Mat. Visn. 1(3), 287–330 (2004); translation in Ukr. Math. Bull. 1(3), 291–334 (2004) 4. Atiyah, M.F., Singer, I.M.: Index theory for skew-adjoint Fredholm operators. Inst. Hautes Études Sci. Publ. Math. No. 37, 5–26 (1969) 5. Baas, N.A., Dundas, B.I., Rognes, J.: Two-vector bundles and forms of elliptic cohomology. In: Topology, Geometry and Quantum Field Theory, London Math. Soc. Lecture Note Ser., 308, Cambridge: Cambridge Univ. Press, 2004, pp. 18–45 6. Bouwknegt, P., Carey, A.L., Mathai, V., Murray, M.K., Stevenson, D.: Twisted K -theory and K -theory of bundle gerbes. Commun. Math. Phys. 228(1), 17–45 (2002) 7. Brylinski, J.-L.: Categories of vector bundles and Yang-Mills equations. Contemp. Math. 230, Providence, RI: Amer. Math. Soc., 1998 8. Carey, A.L., Hannabuss, K.C., Mathai, V., McCann, P.: Quantum Hall effect on the hyperbolic plane. Commun. Math. Phys. 190(3), 629–673 (1998) 9. Carey, A.L., Wang, B.-L.: Thom isomorphism and Push-forward map in twisted K-theory. J. K-Theory 1(2), 357–393 (2008) 10. Donovan, P., Karoubi, M.: Graded Brauer groups and K -theory with local coefficients. Inst. Hautes Études Sci. Publ. Math. No. 38, 5–25 (1970) 11. Freed, D.S., Hopkins, M.J., Teleman, C.: Loop groups and twisted K -theory I. http://arXiv.org/abs/0711. 1906v1[math.AT], 2007 12. Freed, D.S., Hopkins, M.J., Teleman, C.: Twisted K -theory and loop group representations. http://arXiv. org/abs/math/0312155v3[math.AT], 2005 13. Furuta, M.: Index theorem, II. (Japanese) Iwanami Series in Modern Mathematics. Tokyo: Iwanami Shoten, 2002 14. Furuta, M.: Monopole equation and the 11 8 -conjecture. Math. Res. Lett. 8(3), 279–291 (2001) 15. Gomi, K.: An approach toward a finite-dimensional definition of twisted K -theory. Travaux mathématiques. Vol. XVII, Fac. Sci. Technol. Commun., Luxembourg: Univ. Luxemb., 2007, pp. 75–85 16. Gomi, K., Terashima, Y.: Chern-Weil construction for twisted K -theory. preprint 17. Kapustin, A.: D-branes in a topologically nontrivial B-field. Adv. Theor. Math. Phys. 4(1), 127–154 (2000) 18. Milnor, J.: On axiomatic homology theory. Pacific J. Math. 12, 337–341 (1962) 19. Rosenberg, J.: Continuous-trace algebras from the bundle theoretic point of view. J. Austral. Math. Soc. Ser. A 47(3), 368–381 (1989) 20. Witten, E.: D-branes and K -theory. J. High Energy Phys. 1998(12), Paper 19, 41 pp. (1998) (electronic) Communicated by A. Kapustin
