Proceedings of the Carolina Symposium on Neutrino Physics Its Impact on Particle Physics, Astrophysics and Cosmology
Editors
J. Bahcall, W. Haxton K. Kubodera & C. Poole World Scientific
Proceedings of the
Carolina Symposium on Neutrino Physics Its Impact on Particle Physics, Astrophysics and Cosmology
This page is intentionally left blank
Proceedings of the
Carolina Symposium on Neutrino Physics Its Impact on Particle Physics, Astrophysics and Cosmology
University of South Carolina
10-12 March 2 0 0 0
Editors
J. Bahcall Institute for Advanced Study, Princeton
W. Haxton University of Washington
K. Kubodera University of South Carolina
C. Poole University of South Carolina
V f e World Scientific w l
Singapore • New Jersey • London • Hong Kong
Published by World Scientific Publishing Co. Pte. Ltd. P O Box 128, Farrer Road, Singapore 912805 USA office: Suite IB, 1060 Main Street, River Edge, NJ 07661 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
CAROLINA SYMPOSIUM ON NEUTRINO PHYSICS —ITS IMPACT ON PARTICLE PHYSICS, ASTROPHYSICS AND COSMOLOGY Copyright © 2001 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.
ISBN 981-02-4472-X
Printed in Singapore by World Scientific Printers
^:^
V„
<*j\
The Carolina Symposium is held in honor of Dr. Frank T. Avignone, III, who has made significant contributions to neutrino physics and related areas.
This page is intentionally left blank
PREFACE
Neutrinos play a key role in many areas of particle physics, nuclear physics and astrophysics. The recent discovery of neutrino oscillations gives the first hint of new physics beyond the standard model. It is clearly extremely important to further study oscillations and other fundamental properties of neutrinos. It is also important to improve our knowledge of neutrino-nucleus reactions, which are crucial for understanding a large variety of astrophysical phenomena. These and many other interesting questions can be investigated at stopped pion neutrino facilities like ORLaND proposed for the Spallation Neutron Source at the Oak Ridge National Laboratory. To place the ORLaND project in the broad context of the current status of neutrino physics in general and its relation to astrophysics and cosmology, the Carolina Symposium on Neutrino Physics - Its Impact on Particle Physics, Astrophysics and Cosmology, was organized. The purpose of the symposium is twofold: (i) To explore and exchange ideas on the latest developments in the general frontiers of neutrino physics and related fields; (ii) To address specific issues pertaining to the above-mentioned stopped-pion neutrino facility. Among the topics covered by seventeen invited talks and seventeen selected contributed papers are: cosmology and neutrinos; standard model tests with neutrinos; neutrino oscillation, experiments and theories; dark matter searches; double beta-decay; rare event detection techniques; the solar neutrino problem; supernova explosion; nucleosynsthesis; the ORLaND project. This symposium was held in honor of Dr. Prank T. Avignone, III, who has made major contributions to many of the research areas covered by this symposium. We are grateful to the U.S. National Science Foundation and the U.S. Department of Energy for their support for this symposium. We also gratefully acknowledge support from the College of Science and Mathematics and the Department of Physics and Astronomy, University of South Carolina. Thanks are also due to Dr. Vladimir Gudkov for his kind help with our editorial work.
The Editors
VII
ORGANIZATION
SCIENTIFIC P R O G R A M COMMITTEE: John N. Bahcall (IAS, Princeton) Wick C. Haxton (INT, Seattle) Kuniharu Kubodera (USC)
LOCAL ORGANIZING COMMITTEE: Horacio A. Farach, Chair (USC) Gerard M. Crawley (USC) Richard J. Creswick (USC) James M. Knight (USC) Kuniharu Kubodera (USC) Fred Myhrer (USC)
HOST I N S T I T U T E Department of Physics and Astronomy, University of South Carolina
VIII
TABLE OF C O N T E N T S
Preface
vii
COSMOLOGY Dark Energy M. S. Turner
1
Looking Back with Neutrinos L. Stodolsky
14
S U P E R N O V A E , NUCLEOSYNTHESIS Neutrino Effects in Nucleosynthesis W.C. Haxton
21
Supernova Studies at ORLaND A. Mezzacappa
36
SOLAR N E U T R I N O S Astrophysical Neutrinos: 20th Century and Beyond J. N. Bahcall Single Atom Extraction and Classification with a Hybrid Solar Neutrino Detector K. Lande, P. Wildenhain, R. Corey, M. Foygel and J. Distel
56
71
The Sudbury Neutrino Observatory J. J. Simpson
85
The Solar Core and Solar Neutrinos D. C. Kennedy
91
N E U T R I N O PROPERTIES — THEORETICAL Neutrinos and the Standard Model B. R. Holstein
97
Mass Matrix for Atmospheric, Solar, and LSND Neutrinos S. P. Rosen IX
111
X
What is Coherent in Neutrino Oscillations — the Analog with a Two-Slit Experiment H. J. Lipkin
115
N E U T R I N O PROPERTIES — E X P E R I M E N T A L Atmospheric Neutrinos in Super-Kamiokande H. Sobel
120
A Review of Neutrino Oscillation Search at Accelerators S. R. Mishra
133
The BOREXINO Project and Fundamental Achievements in the Very Low Radioactivity Techniques G. Bellini
171
Searches for Non-SM Physics with the KARMEN Experiment K. Eitel
182
Results of the Palo Verde Long Baseline Reactor Neutrino Experiment J. Wolf
187
The Observatory for Multiflavor Neutrinos from Supernovae R. N. Boyd
192
Lead Perchlorate as a Neutrino Detection Medium S. R. Elliott, P. J. Doe, R. G. H. Robertson and C. Paul
198
Direct Neutrino Mass Measurement with a Superconductive Detector M. R. Gomes, P. Valko and T. A. Girard
203
Nuclear Spin Isospin Responses and Spectroscopy of /3/3 Rays from 100 Mo for Neutrino Studies in Nuclei H. Ejiri
208
ORLaND, Oak Ridge Laboratory for Neutrino Detectors Physics Opportunities at the Proposed ORLaND Neutrino Facility F. T. Avignone III and Yu. V. Efremenko
214
Some Aspects of Neutrino Physics S. Nussinov
235
56
Fe(2/ e ,e~) 56 Co: A Technique for an Accurate Cross Section Measurement at ORLaND Yu. Efremenko, F. T. Avignone and A. Mezzacappa
248
XI
Calculating Neutrino-Nucleus Interactions D. J. Dean Search for Pseudoscalar Current Using TX° —> vv1 Decay A. R. Fazely Flavor-Degenerate Pair Production in Neutrino-Nucleus Collisions L. Chatterjee, M. R. Strayer and Jianshi Wu
258 264 268
DARK MATTER CUORE and CUORICINO E. Fiorini
274
WIMP Searches at Canfranc with Germanium Detectors A. Morales
291
Searching for Supersymmetric Dark Matter. The Modulation Effect due to Caustic Rings J. D. Vergados The ORPHEUS Dark Matter Experiment S. Casalbuoni, G. Czapek, F. Hasenbalg, M. Hauser, S. Janos, U. Moser, K. Pretzl, B. Sahli, B. van den Brandt, J. A. Konter, S. Mango, T. Ebert, K.U. Kainer and K. M. Knoop First Results from a Large Superheated Droplet Detector for Dark Matter Searches J. /. Collar, D. Limagne, J. Puibasset, G. Waysand, T. A. Girard and H. S. Miley
305 310
315
D O U B L E BETA-DECAY Background Studies for the Double Beta Decay Experiment NEMO 3 C. S. Sutton and D. Lalanne
320
Double Beta Decay and the Majorana Project L. De Braeckeleer
325
Double Beta Decay of 100 Mo V. D. Ashitkov, A. S. Barabash, S. G. Belogurov, S. I. Konovalov, R. R. Saakyan, V. N. Stekhanov, V. I. Umatov, G. Carugno, G. Puglierin and F. Massera
340
DARK ENERGY MICHAEL S. TURNER Departments of Astronomy & Astrophysics and of Physics Enrico Fermi Institute, The University of Chicago Chicago, IL 60637-1433, USA NASA/Fermilab Astrophysics Center Fermi National Accelerator Laboratory Batavia, IL 60510-0500, USA E-mail:
[email protected]
The discovery that the Universe is speeding up and not slowing down was greeted with open arms by theorists. First, because the dark energy powering the acceleration provided the "missing stuff" needed to make the Universe flat, in accord with a key prediction of inflation. Second, because theorists now have a new puzzle to solve, the nature of the mysterious dark energy. I have no doubt that the dark energy problem will be just as fundamental and just as interesting as the dark matter problem. Determining its nature will require the work of both astronomers and particle physics and will shed light on both fundamental physics and the fate of the Universe.
1
Frank Avignone: A Member of a Special Breed
A special breed of scientist is attracted to the elusive, the invisible and the difficult. Frank Avignone is definitely a member of this rare class. Frank started his career with neutrinos, the state-of-the-art in elusiveness at the time. He then graduated to dark matter. Stuff so powerful that it holds the Universe together, but so weakly interacting that it makes detecting neutrinos look easy. (For a while, it appeared that neutrinos might be the primary stuff of the Universe. But now, it appears that they are only as important as stars, fi„ ~ fi„.) Frank is deeply invovled in trying to solve the 70-year old dark matter riddle by directly detecting some of the dark matter particles that comprise our halo. Since I am confident in both his abilities and in the long, bright future that lies ahead I want to get him in on the ground floor of the next big problem: dark energy. It promises to be as fascinating and as important as dark matter. This is for you Frank.
1
2 2
The Universe Is Accelerating
The Universe is accelerating. Science magazine called this the breakthrough discovery of 1998. The evidence came from distance measurements to type la supernovae (SNe la) at redshifts of order z ~ 0.3 — O.9.1'2 The two groups involved, the Supernova Cosmology Project and the High-z Supernova Search Team, were using one of the classic tests of cosmology, the magnitude - redshift diagram or Hubble diagram, to measure the deceleration parameter q0 in hopes of determining the geometry of the Universe and the density parameter fio. Instead, both groups found that the expansion is speeding up, not slowing down. Since the gravitational effect of matter is always attractive, this acceleration must be due to something exotic! The simplest explanation is Einstein's cosmological constant, which has a repulsive gravitational effect. The discovery of the accelerating Universe was greeted with open arms by cosmological theorists because it reconciled their strong prejudice for the flat Universe predicted by inflation (and good taste) with the growing evidence that matter contributes only about 1/3 of the critical density. This cosmologist (and many others I am sure) was delighted for a second reason: We now have a new puzzle to work on, one which involves fundamental physics and will require cosmological observations to sort out. Einstein introduced the cosmological constant as a sort of fudge factor to obtain a static cosmological model that is consistent with Mach's principle. When the expansion of the Universe was discovered he tried to put the genie back in the bottle, calling the cosmological constant his greatest blunder. In context of modern physics, the cosmological constant corresponds to the quantum energy of the vacuum; so maybe Einstein was right after all. There are a couple of potential snags in this simple story. Cosmologists have been quick to invoke the cosmological to solve problems that later disappear. Could it be that the accelerating Universe will disappear too? I don't think so; it recently received independent confirmation from cosmic microwave background (CMB) anisotropy measurements which indicate that the geometry of the Universe is flat to a precision of about 7% (in terms of the density parameter, fi0 = 1-1 ± 0.07). 3 ' 4 ' 5 Since clustered matter contributes only about 1/3 of the critical density, that requires 2/3 in an unclustered form. Putting 2/3 of the critical density in vacuum energy, not only balances the books but also nicely explains the supernova results. In addition to the very checkered history of the cosmological constant, particle physicists have failed to compute it to an accuracy of better than a factor of 10 55 . The sum of the known contributions to the quantum vacuum
3
energy come to about 1055 times the critical density.6 While the cosmological term is not optional, it could prove to be zero. 3
The Dark Energy Problem
It could be that cosmologists have simply determined the weight of empty space before particle theorists have been able to compute it. There are even some hints that supersymmetry or superstring theory may hold the answer. However, for the moment, I think that we should be more open minded. Zwicky, who identified the dark matter problem, was a most creative guy (some might even say crazy), but no solution he suggested came anywhere close to the working hypothesis we have today (a new form of matter left over from the big bang). Following Zwicky's lead, I simply call this weird stuff "dark energy." By dark energy I mean a nearly smooth component of stuff with large, negative pressure, px ~ — Px- We know it is not clustered because we find no evidence of its effects in galaxies or clusters of galaxies. It is more like energy than matter, because its bulk pressure is in magnitude comparable to its energy density. (For a fluid of particles, v2/c2 ~ p/p; in general, the ratio of the pressure to the energy density quantifies how relativistic a system is.) There are actually two arguments that the bulk pressure of the dark energy must be very negative. The first involves the acceleration of the Universe directly. Acceleration implies that the deceleration parameter is negative, and in turn, that px ^ —px/2: _ (R/R)o 10 3v^0 90 = R2 = 2 ° o + 2 L n*Wi "
q0<0
=> wx < - ^ x
(1)
i 1
« ~2
(2)
Here R(t) is the cosmic scale factor and the pressure of component i,Pi = Wipi (e.g., for baryons Wi = 0, for radiation wt = 1/3, and for vacuum energy Wi — — 1). Note, the presence of a relativistic component means that go 7^ ^ o / 2 , and further, if w, < 0, the deceleration parameter can be negative! This all owes to the fact that in general relativity the source of gravity is p + 3p. The second argument involves the formation of structure in the Universe. Growing the structure seen today from the density perturbations inferred from CMB anisotropy requires a long epoch in which the Universe is matter dominated. (During a phase in which the energy density of the Universe is dominantly a smooth component such as radiation or dark energy the growth
4
of perturbations is severely impeded.) The evolution of the dark energy depends upon its "equation-of-state"; for constant wx, px oc R-3{1+Wx) PX/PM
= (ilx/nM)(l
(3) Zw
+ z) *
(4)
The more negative wx is, the lower the redshift that dark energy begins to dominate the dynamics of the Universe. A quantitative version of this argument implies that wx < —\-7,9 The first candidate for the dark energy is of course the energy density associated with the quantum vacuum (aka cosmological constant) for which px = — px (wx = — 1)- However, the inability of particle theorists to compute the energy of the quantum vacuum casts a long, dark shadow on it, and it could actually be zero.6 A host of other possibilities have been put forth: rolling scalar field (or quintessence); 9,10 a network of frustrated topological defects;11 the energy of a metastable vacuum state; 12 quantum effects of a massive scalar field;13 and "solid" or "generalized" dark matter. 14 ' 15 It could indicate a breakdown of the Friedmann equation of the standard cosmology or be the influence of extra dimensions. 16 Only for vacuum energy is the spatial distribution of dark energy absolutely uniform; however, for all the other known examples of dark energy dumpiness only manifests itself on the largest scales and can be neglected for most purposes (in the future clumping of dark energy may be of some use in determining its nature 17 ). For this reason, Martin White and I have suggested parameterizing the dark energy by its bulk equation of state: 7 wx = (px)/{px), which may be time varying (e.g., a rolling scalar field). Presently, a combination of measurements - supernovae, CMB anisotropy, large-scale structure, and determinations of the matter density - constrain Wx < —0.6 (95% cl). Interestingly, the maximum likelihood for wx is achieved for wx « — 1. 4
Getting At The Dark Energy
In determining the nature of dark energy, I believe that telescopes and not accelerators will play the leading role - you can't bottle dark energy - and even if there is a particle associated with it, it is likely to be extremely difficult to produce at an accelerator because of its gravitational or weaker interactions with ordinary matter. However, as Carroll points out, because such a particle must be extremely light (TO < H0 ~ 10 _ 3 3 eV), it couples to ordinary matter, even very weakly, it could manifest itself through a new, long range force.18
Figure 1. Comoving volume element dV/dUdz - 1 , - 0 . 8 , - 0 . 6 , - 0 . 4 (from top to bottom).
vs.
redshift
for constant
w
=
Because dark energy does not clump significantly, its presence only affects the large scale dynamics of the Universe. These effects include the growth of perturbations and the classic cosmological tests. All of the consequences of dark energy follow from its effect on the expansion rate: H2
&Gn
(PM +Px)
(5)
= H*n (l + z)3
(6)
M 3 8irGpx = i#n 3/ *[i+»(*)]«*Mi+*) jre 0
(7)
where ClM ( ^ x ) is the fraction of critical density contributed by matter (dark energy) today, a flat Universe is assumed, and the evolution of the dark energy follows from integrating its equation of motion, d(pxR3) = -pxdR3.
6
Getting at the dark energy involves the classical tests: magnitude vs. redshift (Hubble) diagram, number-count vs. redshift, and angular size vs. redshift. For a flat Universe, all of these depend upon the comoving distance to an object at redshift z, which is determined by the expansion history:
Luminosity distance, the distance inferred from measurements of the apparent luminosity of an object of known luminosity, 51og[dz,(z)] = m — M — 25, is related to r(z) dL(z) = (l + z)r(z), (9) where m is apparent luminosity, M the absolute luminosity and distances are measured in Mpc. The magnitude - redshift (Hubble) diagram is a plot of m(z) vs. log(z). The angular-diameter distance, the distance inferred from the angular size of an object of known size, <1A{Z) = D/9, is related to r(z) dA=r(z)/(l
+ z).
(10)
The angular-diameter distance comes into play in using CMB anisotropy to probe the dark energy, and the Alcock-Paczynski test which compares the angular size on the sky and the redshift extent of a spherical object (or ensemble of objects). 19 The comoving volume element is the basis of number count tests (e.g., lensed quasars, galaxies or clusters of galaxies). It is given in terms of r(z) and H(z) and depends upon the nature of the dark energy
/<*> = 5 5 } = •*<•>/»<*>•
<">
The number of objects found per solid angle dfl per redshift interval dz is given by
3 E = "<"'<'>
<12»
where n(z) is the comoving number density of objects being counted. If the evolution of n(z) is known (e.g., no evolution or evolution that can be computed from numerical simulations), then dN/dQdz, can be used to infer f(z) and wxThe gravity-driven acoustic oscillations of the baryon-photon fluid at the time of last scattering gives rises to a series of acoustic peaks in the angular power spectrum of CMB anisotropy. The peaks correspond to different Fourier
cd
a, a
• r-H
240
+J
rst
230
£
220 —
2
210
1
1
1
1
QM = 0 . 1 5 \
| l l l/l | l
2 o
1
/
CO
H-.
I l I l l l
CD
1
l l
250
0.30
M i n i 1 1 1
M
l / l 1 11 1 1 1 1 1
7
^ ^ \
0^45
-
~ ^
CO
o PU
200
i
i
i
1
i
i
i
-.8
-1
i -.6
i
-
w Figure 2. The position of the first acoustic peak as a function of w for fig/i2 = 0.02.
modes caught at maximum compression or rarefaction. The condition for this is krjsn = nir, where the odd n modes are compression maxima and 7JSH is the sound horizon: 20 r L S Vsdt _ %H
" J0
v2 =
f°°
vs{z')dz'
R(t) ~ JZLS H(z')
1/3 1 + 3/)B/4/97
(13) (14)
Modes captured at maximum compression or rarefaction provide standard rulers on the last-scattering surface. Their angular size 0t and physical size
8 d ~ n/k(l
+ 2LS) ~ »7SH/(1 + £ L s) are related by a
+ z L s) <^(LS)
%H/(1
d A (LS) = (1 + ZLS)" 1 /
(15) -%-r
(16)
where ZLS — 1100. Using the fact that the angular power at multipole / is dominated by modes around kl ~ rj^s, the positions of the peaks are given approximately by ln = nn^-
(17)
T]LS
where for a flat Universe 7JLS is J u s t the coordinate distance to the lastscattering surface r(zLs)For fixed matter and baryon density (i.e., fixed QM^2 and fifi/i2), the positions of the acoustic peaks only depend upon the distance to the last scattering surface, which depends upon fix and wx- This is the primary sensitivity that the CMB has to the equation-of-state of the dark energy. Using the approximation above for the location of the first peak, its sensitivity to cosmological parameters is found easily, AZl
h
= - 0 . 0 8 4 A W - 0 . 4 5 ^ +0.09
'
h
- 0 . 1 4 — £ - 1.25—^
Af)Bft2
nBh2 (18)
around wx = - 1 , h = 0.65, ilM = 0.35, nBh2 = 0.02 and il0 = 1. Note that the position of the first peak is least sensitive to wx- It is intriguing, but probably not significant, that the position of the first peak as determined by the BOOMERanG and MAXIMA experiments, h ~ 200, is slightly lower than expected for a vacuum energy and favors wx ~ - 2 / 3 . A recent analysis 21 that compares the power of different cosmological techniques to determine wx forecasts the following Id errors for the different techniques discussed above: aw — 0.4 for CMB anisotropy using Planck; aw — 0.1 for a sample of 2000 SNe la with 0.2 < z < 1.7 (as might be gathered by the proposed SNAP satellite mission;22) and aw ~ 0.1 for a sample of 10,000 galaxies with 0.7 < z < 1.5 assuming that their comoving number density can be determined independently to a precision of better 5% 23 (a sample of a few 1000 clusters could do similarly well provided the cluster number density is known even better 2 4 ).
-0.4 oUvW
MAP(P)
-0.6 CUr\
PiancWP) -0.8
-1.0
aM Figure 3. SNAP constraint to 0 M and ^x compared to those of MAP and Planck (with polarization) and SDSS. Also shown are the present constraints using a total of 54 SNe la. All constraints assume a flat universe and O M = 1 - O x = 0.28, wx = - 1 as iducial values of the parameters. All contours are 68% cl and are obtained using a Fisher-matrix analysis.
That the low-redshift probes do better than CMB anisotropy is not surprising - at the time of last scattering dark energy was only a tiny fraction of the total energy density. These different techniques have different systematics and in the case of CMB anisotropy and SNe la are orthogonal in the wx Ox plane. Thus, they are all likely to play a role in getting at the nature of the dark energy. Finally measurements of r(z) inferred from a large sample of SNe la can in principle be used to reconstruct the dark energy equation-of-state as a function of redshift (or scalar-field potential in the case of quintessence). 25 The "reconstruction" equation for w(z) is 1 + w(z)
=
1+ z
10
-0.4
-0.6
-0.8 -
\
\ Ji
-1.0
\
I
0.2
\
^
n*
^
_iJL 0.4
Figure 4. Constraints to f i j j and wx using galaxy counts (10,000 galaxies from z ~ 0.7 — 1.5). Inner region shows the constraint assuming Poisson errors only, while the outer two regions assume uncertainty of 10% and 20% due to uncertain knowledge of n{z). All regions are 68% cl.
3ffgn M (l + zf + 2(d2r/dz2)/(dr/dz)3 H§nM(l + zf - (dr/dz) - 2 Needless to say, reconstruction requires extraordinary data and even with such is very challenging. However, dark energy is a problem worth the effort. 5
Final Thoughts
Unless we are being terribly misled by observations of both type la supernovae and the cosmic microwave background the Universe is accelerating due to a mysterious form of smooth (or nearly smooth) dark energy. The dark energy problem at the very least rises to the stature of the dark matter problem, and raises a set of profound questions for both cosmology and fundamental
11
-0.2 -
-0.4 N -0.6
Figure 5. Reconstruction (68%, 95% cl bands) of a hypothetical equation of state (solid line) using simulated SNe data (2000 SNe la uniformly distributed out to z = 1.5 with individual distance uncertainties 7%).
physics. • How much does nothing (vacuum energy) weigh? What is the nature of the dark energy? • The ratio of matter to dark energy varies greatly with time, px /' PM ~ (1 4- z)3wx. Why is the ratio nearly one today? • In a matter-dominated Universe there is a simple connection between destiny and geometry; depending upon the nature of dark energy, a flat Universe can either expand forever or recollapse. What is our fate, and can we ever know it? 26 • The recipe for our Universe is: 5% baryons; 30% cold dark matter; and 65% dark energy. Why? Are the dark matter and dark energy related?
12
While we are far from answering any of these questions, one thing is certain, cosmological observations will pay a key role in unraveling the nature of the dark energy. Personal Note For me, the friendships one makes in pursuing the frontiers of science can be as rewarding as the science itself. Frank's intensity, humor and genuine warmth make him a very special friend. Not only was it an honor to be here to fete him, but it was a special pleasure to meet and have dinner with his wonderful children. As they say, the tree doesn't grow far from where the apples land. References 1. 2. 3. 4. 5. 6. 7. 8.
9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
S. Perlmutter et al, Astrophys. J. 517, 565 (1999). A. Riess et al, Astron. J. 116, 1009 (1998). P. de Bemardis et al, Nature 404, 955 (2000). S. Hanany et al, astro-ph/0005123. A. Jaffe et al, astro-ph/0007333. S. Weinberg, Rev. Mod. Phys. 61, 1 (1989). M.S. Turner and M. White, Phys. Rev. D 56, R4439 (1997). M.S. Turner, in Type la Supernovae: Theory and Cosmology, eds. J.C. Niemeyer and J.W. Truran (Cambridge Univ. Press, Cambridge, 2000), p. 101 (astro-ph/9904049). K. Freese et al, Nucl. Phys. B 287, 797 (1987); B. Ratra and P.J.E. Peebles, Phys. Rev. D 37, 3406 (1988); K. Coble et al, Phys. Rev. D 55, 1851 (1996); P.G. Ferreira and M. Joyce, Phys. Rev. D 58, 023503 (1998); R. Caldwell et al, Phys. Rev. Lett. 80, 1582 (1998). I. Zlatev et al, Phys. Rev. Lett. 82, 896 (1999). A. Vilenkin, Phys. Rev. Lett. 53, 1016 (1984); D.N. Spergel and U.-L. Pen, Astrophys. J. 491, L67 (1997). M.S. Turner and F. Wilczek, Nature 298, 633 (1982). L. Parker and A. Raval, Phys. Rev. D 62, 083503 (2000). M. Bucher and D. N. Spergel, Phys. Rev. D 60, 043505 (1999). W. Hu, Astrophys. J. 506, 485 (1998). N. Arkani-Hamed et al, Phys. Lett. B 480, 193 (2000). R. Caldwell et al, Phys. Rev. Lett. 80, 1582 (1998). S.M. Carroll, Phys. Rev. Lett. 81, 3067 (1998). C. Alcock and B. Paczynski, Nature 281, 358 (1979).
13
20. 21. 22. 23. 24. 25. 26.
W. Hu and N. Sugiyama, Astrophys. J. 444, 489 (1995). D. Huterer and M.S. Turner, astro-ph/OOlOxxx. http://snap.lbl.gov. M. Davis and J. Newman, Astrophys J. 513, L95 (1999). Z. Haiman, J. J. Mohr, and G. P. Holder, astro-ph/0002336. D. Huterer and M.S. Turner, Phys. Rev. D 60, 081301 (1999). L. Krauss and M.S. Turner, Gen. Rel. Grav. 31, 1453 (1999).
Looking Back w i t h Neutrinos"
Max-Planck-Institut
L. Stodolsky fur Physik,
80805
Munich
We briefly discuss the history of suggestions for time- of-flight effects due to nonzero neutrino mass and a recent proposal that such effects can be used to determine the parameters of cosmology. With neutrinos there is potentially a much deeper "look back time" than with photons. We note a new point, that if future large scale neutrino detection arrays see long-time secular variations in counting rate, this could be due to highly redshifted bursts originating in the early universe.
1
Introduction
Neutrinos are everywhere and they're getting into everything. Some have been in the newspapers and a number of them certainly penetrated the Pentagon and have gone through secrets at Los Alamos. Who knows what they won't get into next? b In view of this interesting if perhaps a bit alarming situation, it shouldn't come as a surprise that they're also getting into history. We all know of course that they play a major role in the history of the universe. Here, however we want to talk about yet another role for them: not as players, but as reporters. Back at the time of supernova 1987a and even long before, it was realized that if neutrinos had a small mass and if they traveled over astronomical distances the small deviation of their velocity from the speed of light could lead to possibly observable effects. It seems to have begun with Zatsepin l, who in 1968 suggested if a neutrino burst from a supernova would be detected it could be used to limit the neutrino mass. He used the fact that a mass would mean that the velocity would be energy dependent. Thus a burst starting out a few seconds wide in time would, due to the range of energies contained in it, spread as it traveled to us, and this spreading would reflect on the mass. He concluded one could improve the then-existing laboratory mass limit of 200 eV to about 2 eV. Later, with the discovery of different neutrino types came the realization 2 of an even more obvious effect: given different neutrino types, each one should have a different mass and so a different velocity in a burst. In SN 1987a this leads to a limit 3 of around lOeV for ve, unfortunately not very strong by present standards. We see that a neutrino mass has two potentially observable consequences: "Presented at the Carolina Symposium on Neutrino Physics, Columbia, March 2000. 6 These introductory comments are intended as an admonition to those, who despite often high ranking responsibilities, has been known to be frequently involved with neutrinos.
14
15
• A burst spreads in time due to the dispersion of velocities. • Different mass states arrive at different times. Note that one of the mass states could also correspond to photons emitted in the event, in which case m = 0 and the first point doesn't apply; or contrariwise one of the mass states could be an as yet undiscovered heavy particle (e.g. a WIMP) in which case the time-of-flight effects are magnified. Also note we should expect that the neutrino mass states are generally mixed flavor states, leading to a "flavor-echo", as we discussed in ref [2]. Then in another vein, there is the possibility of some fundamental tests of relativity 4 . The limiting velocity for all forms of radiation should be c, the speed of light. There is no reason to doubt this and many good reasons to believe it. Nevertheless, SN 1987a gave us a rare chance to test it in a novel way, and a rather strong confirmation it is, on the 10~9 level. 2
Measuring the Universe with Neutrinos
Here I would like to report on a continuation of this story, which is again nothing but kinematics, but nevertheless with amusing consequences. Once again we are concerned with relativity and neutrino mass. But now it is general relativity and instead of learning about the properties of the neutrino we want to use them to determine the large scale geometry of the universe. The main point is the following: Neutrinos with mass, when emitted from a distant source will travel somewhat slower than the speed of light. This deviation, however, will depend on the cosmological epoch, since as the universe expands the neutrinos slow down. When the neutrino finally reaches us, its total travel time represents in effect a record of the cosmological epochs it has passed through. This is a simple idea and could easily have been calculated by the founders of general relativity and cosmology, if they had known about the existence of a very light particle which travels almost, but not quite at the speed of light. But they didn't of course, and so they concentrated on "geometry" such as the measurement of angles subtended by "standard measuring rods" or the apparent brightness of "standard candles". Thus it is only very recently, stimulated by the growing evidence for neutrino mass 5 , that we realized 6 there is a "particle" as opposed to a "geometrical" way of surveying the universe. Assuming that we know or will know the neutrino masses, it turns that neutrino bursts from sources with identified red shift can give us both the Hubble constant and the acceleration parameter of cosmology7 q. No independent knowledge of the distance to the source is necessary, so difficulties involving the "cosmic distance ladder" are absent.
16
To see how this comes about, let us calculate the difference in arrival times for two mass states emitted in the same cosmic event. We take the standard FRW metric 7 ds2 = dt2 - a2(t)(dx)2, where we define a(t) to be the expansion factor of the universe normalized to its present value: a(t) = R(t)/R(now), so that a(now) = 1. We proceed by finding an equation for the coordinate velocity dx'/dt, where x' is along the particle's flight direction. First we express dxl /dt in terms of P'(t), the spatial part of the contravariant fourmomentum m dx*1 /ds. From the definition of the metric we have a(t)dx'/dt = [a{t)Pi(t)]/^m2 + {a{t)Pi(t)\2. Expanding for the relativistic case P » m we obtain a(t) dx'/dt « 1 — |m 2 /[a(t)F J (*)] 2 To find Pl(t), we now make use of the fact that the covariant or "canonical momentum" P* is constant (since nothing depends on the xcoordinate and Pi ~ <9* ). Furthermore, since the different kinds of momenta are related through the metric tensor, they all become equal at t(now), where a = 1. Hence we can identify the constant covariant momentum as P(now), the momentum at the detector. Thus from P% = g%:*Pj = l/(a 2 )Pj, we obtain P% = l/(a2)P(now) . Thus we finally have , . 1 m2 2P2{nawY
dx_ 1 dt ~ a(t)
aK)
K
'
The first term by itself will be recognized as just the equation for motion along the light cone, and then there is a small correction involving the mass. Introducing Ax for the difference in the x coordinate of two different particles of mass m-2 and mi emitted in the same event at the same time d x
^)
«a(t)i r W
L
m 2
i
_
w 2
i
1
(2)
dt ~ " 2 P (now) P (now) At the present epoch with a = 1, Ax is just the spatial separation of the two particles. Integrating, we have for this separation, or in view o f t ) « c = l for the time difference in arrival at a detector At « Ax » / a(t) dt\ [-^— ™'2 J • (3) w i 2 w J 2 P (now) Pi(now)] It is thus / a(t) dt which "records" the cosmological information. Observe a is small at early times so that most of the effect comes near the present time, as expected since this is when the neutrinos are the "slowest". The expressions in the brackets are the familiar factors giving the difference in velocity for highly relativistic particles. It now only remains to get rid of the coordinate dependent a and t and to re-express things in terms of an observable, namely the red shift parameter z
17
for the event. With the expansion of a(t) for recent epochs a(t) = 1 4- H[t — t(now)\ - \qH2[t - t(now)\2 + ..., and the redshift parameter z = l/a — 1 = -H[t - t{now)] + (1 + q/2)H'2[t - t(now)]2 + ..., we find z r
3+ q
17'-
~ 2 ~ * + -J2l-P?(ncm;) ~ / f ( W ) J
i 1r
m\
m\
-, W
giving the result in terms of the directly observable z. We thus have the measured quantities for an event, At and z, given in terms of the present Hubble constant H and the acceleration parameter q. Thus in principle two good events-assuming the neutrino masses well known by the time this all happens- fix these cosmological parameters. This was for the time delay between two distinct mass states. However it may well be that the mass differences aren't big enough to give cleanly separated pulses. In that case we can try using the pulse spreading effect. So consider the same mass but different momenta P and P'. We then get a time delay between the two of A
* * If t1 -2~V-Z+ - I \ m 2 K T T - \ ? - ("FT—?>a]" L
J
(5)
L
H 2 2 P(now) P'{now) These formulas are low z expansions, to order z1. The first term in z in the expressions is just what we would get without general relativity; it says that At is simply the velocity difference times the distance (since z = H d is just the Hubble law, where d is the distance). Even this is not entirely trivial, however, since it shows how with neutrinos one can find H without knowing the distance to the object. That traditional difficulty of observational cosmology, the "cosmological distance ladder", is gone. It is of course replaced by the difficulty of detecting neutrino bursts at cosmological distances. If the neutrino observations are ever made it will certainly be interesting to compare the "neutrino H" found this way with the "photon H" found by astronomy. 3
Reality?
The realization of these proposals, attractive as they might be, does not seem immediate or certain. First of all, we need a class of cosmic events that emits bursts of neutrinos , or neutrinos and photons simultaneously and with great intensity. This does not seem impossible, and certainly hadronic mechanisms for phenomena like the gamma ray bursts would do this, giving neutrinos from charged pions and photons from neutral pions.
18
Then there is the question of detectability. With the bursts coming from cosmological distances, they will reach us substantially weakened. Here it seems we must rely on the development of the km scale detectors in the ocean and in the ice. Assuming all this is accomplished we have the question of the good separation of the pulses. If our effects are too small compared to the duration of the original event itself, our time of flight effects will be lost or at least require an elaborate statistical analysis to be extracted. To get some feeling for this we can evaluate the kinematic factor in front of the formulas in terms of an eV (mass) 2 for the neutrinos and GeV for their energy: (m/eV) 2 2(P/GeV) 2
50/xsec/Mpc.
(6)
It appears that even at a thousand Mpc, a substantial part of the way across the visible universe, we may only expect some msec delays. While msec or even fisec speed doesn't seem very difficult for particle detectors, there is the problem of the intrinsic time scale of the burst itself. If we take supernovas or gamma-ray bursts as a guide, where the timescale is on the order of some seconds, then it seems that to have distinctly separated bursts we would need to have particles distinctly more massive than eV's. One possibility might be the third neutrino mass eigenstate, another perhaps more interesting one would be not the neutrino itself, but a heavy neutral object like the WIMP we are looking for in our dark matter searches 8 . If the objects do indeed exist they should be stable and if not overly massive, could be emitted in high energy bursts. However, there is still the possibility of obtaining information even if the bursts don't separate into clearly distinguishable pulses corresponding to the different masses. There is still the original Zatsepin effect of the spreading of the pulse, here represented by Eq [5], and involving the same cosmological information as the separation effect Eq [4]. Note our distinction of the "separation" and "spreading" effects is for the purposes of a qualitative description and the two may well overlap, necessitating a detailed analysis including a modeling of the pulse shapes. Comparing the kinematic factors in Eq [4] and Eq [5] the condition that the mass separation At be distinctly greater than that of the pulse spreading is ^ V > > ^ , where Ap is the energy spread of the burst.
19 4
Getting Into the Big Bang
Observational methods using the photon, be it the optical photon of classical astronomy, the microwave photon of the background radiation or those of radio astronomy, will never allow us to look further back than "recombination", some 100,000 years after the Big Bang. One of the most intriguing potentialities of these neutrino or neutrino - like based observations is that these particles will come to us directly from a very early epoch. The "look-back time" is much deeper. Decoupling for the neutrino is in the first few minutes, so with neutrinos we can, in principle, go back to the first minutes and study the space-time geometry at that time. Facing this exciting possibility of looking deep into the Big Bang are a few requirements and difficulties which are not exactly trivial. Again, some sort of phenomenon must exist around the epoch in question in which powerful high energy bursts of neutrinos are emitted. Perhaps collapse of overdense regions to black holes or annihilation of topological defects are a possibility. Then after being strongly redshifted and diluted by the expansion of the universe, the particles must be detected at the earth. Now we no longer have z < 0(1) but rather, if we go all the way back to neutrino decoupling z ~ (lMeV/O.lmeV) ~ 10 10 , so that these redshifts and dilutions will be very great. Furthermore, since the time scale of the event is stretched by z, a millisecond event will last a year upon reaching us. On the other hand this suggests something interesting: our grandchildren when operating very big neutrino arrays and observing long-time secular variations in the counting rate, should consider if they are perhaps seeing a burst out of the early universe and not an instability of their apparatus. Alas, in view of the detection difficulties, this way of doing cosmology will have to probably remain fantasy for quite a while, if not forever. On the other hand, it is hard to think of any other way of directly looking at the pre-recombination epoch. In a final mad fantasy we can of course push things even further back. Imagine that there is the WIMP or some other neutral, very weakly interacting particle, then we go even further back into the "Bang". Things like the formation of "Baby Universes", or other Colossal Happenings, presumably involving some activity and excitations of the particle fields on a microscopic timescale, become directly "visible" via their bursts, and allow us to study the space-time structure of the epoch. While this is indeed very far-fetched, it is amusing that at least at the level of fantasy, there is potentially an observational correspondence to such happenings.
20
References 1. G. T. Zatsepin, ZhTEF Pis. Red. 8, 333 (1968), (JETP Lett. 8 205). 2. S. Pakvasa and K. Tennakone, Phys. Rev. Lett. 28, 1415 (1972); S. Pakvasa, DUMAND Symposium 1980; N. Cabibbo, Accademia Lincei, Meeting on Astrophysics and Elementary Particles, (1980); Tsvi Piran, Phys. Lett. B 102, 299 (1981); A.K. Drukier and L. Stodolsky,F%s. Rev. D 30, 2295 (1984). P. Reinartz and L. Stodolsky, Z.f.Phys. C27, 507 (1985). 3. For the discussion with respect to SN 1987a see section 11.3.4 of G.G. Raffelt, Stars as Laboratories for Fundamental Physics, Univ. Chicago Press, (1996). 4. L. Stodolsky, Phys. Lett B 201, 353 (1988); M. J. Longo, Phys. Rev.D 36, 3276 (1987) and Phys. Rev. Lett. 60, 173 (1988); L. Krauss and S. Tremaine, Phys. Rev. Lett. 60, 176 (1988). 5. Y. Fukuda et al., (Super-Kamiokanda Collaboration), Phys. Rev. Lett. 81, 1562 (1998). 6. L. Stodolsky, Phys. Lett. B 473, 61 (2000). 7. S. Weinberg, Gravitation and Cosmology, chapt.14, John Wiley, New York, (1972). 8. S. P. Ahlen, F. T. Avignone 111, et al. Phys. Lett. B 195, 603 (1987), were the first to present laboratory limits on dark matter; for recent work see for example the CRESST report MPI-PTh/2000-04, or section VI of Proceedings, LTD-8, P. De Korte and T.Peacock editors, North-Holland (2000).
N E U T R I N O EFFECTS I N N U C L E O S Y N T H E S I S W. C. HAXTON Institute for Nuclear Theory, Box 351550, and Department of Physics, Box 351560, University of Washington, Seattle, WA 98195, USA E-mail:
[email protected] The nucleosynthesis within a Type II supernova occurs in an intense neutrino flux. I discuss some of the effects associated with neutrino interactions, including direct synthesis in the neutrino process, the role of neutrinos in controlling the r-process path and in postprocessing r-process products, and neutrino oscillation connections.
It is a great pleasure to attend this meeting in honor of a long-time friend, Frank Avignone, and dedicated to his favorite subject, neutrino physics. In contrast to the rare neutrino events that Frank has measured in the laboratory, I will talk about an environment where neutrino reactions are so frequent that they determine much of the chemistry of the matter. That environment is the progenitor star's envelope in the first seconds after a core-collapse supernova. Here the neutrinos directly synthesize new nuclei, and help to eject that matter into the interstellar medium, where it is incorporated into stars like our sun. The neutrinos also control the isospin of the nucleon soup that is the likely site of the r-process. It follows that there is an intimate connection between the properties of neutrinos, including phenomena like neutrino oscillations, and supernova nucleosynthesis. 1
Core-Collapse Supernovae
In the infall stage of a core-collapse supernova 1 neutrinos are trapped by their neutral current interactions once a density of p ~ 10 12 g/cm 3 is reached. Trapped in this sense means that the neutrino diffusion time becomes longer than the time needed to complete the collapse, thereby guaranteeing that the energy liberated by the matter falling into the gravitational potential, ~ 3 x 10 53 ergs, is contained within the protoneutron star. A small portion of this energy is later apparent in the kinetic energy of the ejected shells and in the accompanying optical display. But the vast majority, ~ 99%, is radiated in neutrinos over the ~ 3 second cooling time of the core, following core bounce. Throughout most of their outward diffusion, the various neutrino flavors remain in equilibrium Ue + Ve ++ Vp + Up 21
(1)
22
thereby ensuring that the energy is shared equally by the three flavors. However, when they reach the "neutrinosphere" at ~ 10 12 g/cm 3 , their decoupling is flavor dependent due to the reactions vx + e o vx + e ve + n f» p + e~ Pe+p++n
+ e+.
(2)
The first reaction for ves is about six times that for heavy flavors, while the second and third affect only electron neutrinos. As a result the heavy-flavor neutrinos decouple at a higher density, and thus temperature, than the electron neutrinos. The result is a characteristic temperature hierarchy Tv„Vr ~ 8MeV TPe ~ 4.5MeV T„. ~ 3.5MeV
(3)
where the ue—ue temperature difference results from the matter near the neutrinosphere being neutron rich (having experienced significant electron capture). As the energy is divided approximately equally among the flavors, it follows that the electron neutrino flux is about twice that of the heavy flavors. Supernovae are important engines driving galactic evolution, producing and ejecting the metals that enrich the galaxy. Elements produced in the hydrostatic evolution of the presupernova star (C, O, Ne, ...) are abundant in the ejecta of the explosion. The shock wave resulting from core bounce produces peak temperatures of ~ (1—3)-109K as it traverses the silicon, oxygen, and neon shells. This shock wave heating induces proton and a reactions like (7, a) <-> (a, 7) which generate a mass flow toward highly bound nuclei, resulting in the synthesis of iron peak elements as well as less abundant oddA species. Rapid neutron-capture reactions are thought to take place in the high-entropy atmosphere just above the mass cut, producing about half of the heavy elements above A ~ 80. Finally, the neutrinos themselves transmute certain nuclei within the mantel, producing rare isotopes like n B and 1 9 F in the neutrino process. 2
The Neutrino Process
The neutrino process was described independently by Domagatsky et al.2 and by Woosley, Haxton, et al. 3 Probably the simplest example occurs in the neon shell in a supernova. Because of the first-forbidden contributions, the cross section for inelastic neutrino scattering to the giant resonances in Ne is
23
~ 3 • 10~ 41 cm 2 /flavor for the more energetic heavy-flavor neutrinos. This reaction U + A-H/+A* (4) transfers an energy typical of giant resonances, ~ 20 MeV. A supernova energy release of 3 x 10 53 ergs converts to about 4 x 10 57 heavy flavor neutrinos. The Ne shell in a 20 M© star has a radius ~ 20,000 km. Thus the neutrino fluence through the Ne shell is
6V
4 . 1 fl 5 7
-i-ir
4TT(20, 000km)2
„ l0 38 /cm 2 .
(5)
K
'
'
Thus folding the fluence and cross section, one concludes that approximately l/300th of the Ne nuclei interact, often breaking up to form 1 9 F. This is quite interesting since the astrophysical origin of 1 9 F had not been understood. The only stable isotope of fluorine, 1 9 F has an abundance 19 F — 20 Ne
1 —. 3100
v(6) ;
This leads to the conclusion that the fluorine found in toothpaste was created by neutral current neutrino reactions deep inside some ancient supernova. The calculation of the 1 9 F/ 2 0 Ne ratio is somewhat more complicated than a folding of the cross section and fluence: • When Ne is excited by ~ 20 MeV through inelastic neutrino scattering, it breaks up in two ways 20
Ne(i/, i/) 20 Ne* _> " N e + n - • 20 Ne(i/,t/') 2 0 Ne*-s- 1 9 F + p
19
F + e+ + i/e + n (7)
with the first reaction occurring half as frequently as the second. The sum of these two channels is the 1/300 yield mentioned above. • The subsequent nuclear processing determines whether the 1 9 F survives. In the first 1 0 - 8 seconds the coproduced neutrons in the first reaction react via 15
0(n,p) 1 5 N
19
Ne(n,a) 1 6 0
20
Ne(n, 7 ) 2 1 Ne
19
Ne(n,p) 1 9 F
(8)
with the result that about 70% of the 1 9 F produced via spallation of neutrons is then immediate destroyed, primarily by the (n,a) reaction above. In the next 10~ 6 seconds the coproduced protons are also processed N(p,a) 1 2 C
19
F(p,a)160
23
Na(p,a) 2 0 Ne
(9)
24
with the latter two reactions competing as the primary proton poisons. This makes an important prediction: stars with high Na abundances should make more F, as the 23 Na acts as a proton poison to preserve the produced F. • A final destruction mechanism is the heating associated with the passage of the shock wave. Fluorine produced prior to shock wave passage can survive if it is in the outside half of the Ne shell. The reaction 19
F(7,a)15N
(10)
destroys F for peak explosion temperatures exceeding 1.710 9 K. Such a temperature is produced at the inner edge of the Ne shell by the shock wave heating, but not at the outer edge. If all of this physics is handled in a network code that includes the shock wave heating and F production both before and after shock wave passage, one finds3 [ 1 9 F/ 2 °Ne]/[ 1 9 F/ 2 °Ne] 0 T heavy „(MeV) 0.14 4 0.6 6 1.2 8 1.1 10 1.1 12 for a progenitor star of solar metallicity. One sees that the attribution of F to the neutrino process argues that the heavy flavor v temperature must be greater than 6 MeV, a result theory favors. One also sees that F cannot be overproduced by this mechanism: although the instantaneous production of F continues to grow rapidly with the neutrino temperature, too much F results in its destruction through the (p, a) reaction, given a solar abundance of the competing proton poison 23 Na. Indeed, this illustrates an odd quirk: although in most cases the neutrino process is a primary mechanism, one needs 2 3 Na present to produce significant F. Thus in this case the neutrino process is a secondary mechanism. While there are other significant neutrino process products (7Li, 138 La, 180 Ta, 15 N ...), the most important is 1 1 B, produced by spallation off carbon. A calculation by Timmes et al.4 found that the combination of the neutrino process, cosmic ray spallation and big-bang nucleosythesis together can explain the evolution of the light elements. The neutrino process, which produces a great deal of n B but relatively little 1 0 B, combines with the cosmic ray spallation mechanism to yield the observed isotope ratio. Again, one prediction of this picture is that early stars should be n B rich, as the neutrino process is primary and operates early in our galaxy's history; the cosmic ray production
25
of 10 B is more recent. (We return to this point below.) There is hope that abundance studies will soon be able to descriminate between 10 B and 1 1 B: as yet this has not been done. 3
The r-process
Beyond the iron peak nuclear Coulomb barriers become so high that charged particle reactions become ineffective, leaving neutron capture as the mechanism responsible for producing the heaviest nuclei. If the neutron abundance is modest, this capture occurs in such a way that each newly synthesized nucleus has the opportunity to /3 decay, if it is energetically favorable to do so. Thus weak equilibrium is maintained within the nucleus, so that synthesis is along the path of stable nuclei. This is called the s- or slow-process. However a plot of the s-process in the (N,Z) plane reveals that this path misses many stable, neutron-rich nuclei that are known to exist in nature. This suggests that another mechanism is at work, too. Furthermore, the abundance peaks found near masses A ~ 130 and A ~ 190, which mark the closed neutron shells where neutron capture rates and /3 decay rates are slower, each split into two subpeaks. One set of subpeaks corresponds to the closed-neutron-shell numbers N ~ 82 and N ~ 126, and is clearly associated with the s-process. The other set is shifted to smaller N, ~ 76 and ~ 116, respectively, and is suggestive of a much more explosive environment where neutron capture is rapid. This second process is the r- or rapid-process, characterized by: • The neutron capture is fast compared to (3 decay rates. • The equilibrium maintained within a nucleus is established by (n, 7) <+ (7, n): neutron capture fills up the available bound levels in the nucleus until this equilibrium sets in. The new Fermi level depends on the temperature and the relative n / 7 abundance. • The nucleosynthesis rate is thus controlled by the /? decay rate: each j3~ capture converting n - • p opens up a hole in the neutron Fermi sea, allowing another neutron to be captured. • The nucleosynthesis path is along exotic, neutron-rich nuclei that would be highly unstable under normal laboratory conditions. • As the nucleosynthesis rate is controlled by the /3 decay, mass will build up at nuclei where the /3 decay rates are slow. It follows, if the neutron flux is reasonably steady over time so that equilibrated mass flow is reached, that the resulting abundances should be inversely proportional to these /? decay rates. Thus large abundances are expected at the shell closures, the "waiting point" nuclei where several /3 decays must occur before the shell gap inhibiting further
26
neutron capture can be overcome. The r-process requires exceptionally explosive conditions: neutron densities in excess of ~ 10 20 /cm 3 , temperatures of (1-3) xl0 9 K, and times on the order of one to a few seconds. Evaluating the (n,7) «-» (7,n) equilibrium for typical conditions yields neutron binding energies on the order of ~ 30 kT, or about 2-3 MeV below the neutron drip line. After the r-process finishes (the neutron exposure ends) the nuclei decay back to the valley of stability by /? decay. This can involve some neutron spallation (/3-delayed neutrons) that shift the mass number A to a lower value. But it certainly involves conversion of neutrons into protons, which moves the r-process peaks at N ~ 82 and 126 to lower N, clearly. This shifted r-process peak combines with the s-process peak to produce the double-hump distributions near neutron shell closures found in nature. It is believed that the r-process can proceed to very heavy nuclei (A ~ 270) where it is finally ended by /^-delayed and n-induced fission, which feeds matter back into the process at an A ~ A m Q X /2. Thus there may be important cycling effects in the upper half of the r-process distribution. What is the site(s) of the r-process? This has been debated many years and still remains a controversial subject. Both primary (requiring no preexisting metals) and secondary (enriched in s-process elements) sites have been proposed. Some of the suggested primary sites include the neutronized atmosphere above the proto-neutron star in a Type II supernova, neutron-rich jets produced in supernova explosions or in neutron star mergers, and inhomogeneous big bangs. Secondary sites, where successful synthesis can result for lower p(n), include the He and C zones in Type II supernovae and the red giant He flash. The balance of evidence favors a primary site, so one requiring no preenrichment of heavy s-process metals. In particular, recent abundance studies 5 of very metal-poor stars ([Fe/H] ~ -1.7 to -3.12) have yielded r-process distributions very much like that of our sun (at least for Z > 56) (see Fig. 1). In these stars the iron content is variable. This suggests that the "time resolution" inherent in these old stars is short compared to galactic mixing times (otherwise Fe would be more constant). The conclusion is that the r-process material in these stars is most likely from one or a few local supernovae. The fact that the distributions match the solar r-process strongly suggests that there is some kind of unique site for the r-process: the solar r-process distribution did not come from averaging over many different kinds of r-process events. Clearly the fact that these old stars are enriched in r-process metals also strongly argues for a primary process: the r-process works quite well in an environment where there are few initial s-process metals. It may be that these and similar data make certain primary r-process
27 M i l l IT'! I | 1 1 1 ITTTTT J i l l
1 ISr
-
Pb
.5 0 -.5 -1
r
-%Zr
>4
Ru
4 «
I
: * r*l
-
>
t
- Y
M
Nd °V Ism G d *I
CB
Rh*9
-
Ni
i 1'
1.b
Tb
•
*•
Ell ! H o
Nb
; -
1 I
Ba
Hd
f m
i
•
Th :
-2 2.5
_
s OS f'?t*9? OV .ibu^dfcos
. -' i i i l n 40
u:-
i-
i • 111111111111111111111
50
60
70
. . . . i • n-
80
90
j M l I | I I I I I I I I I I I I I I I I I I • | I I I I I I 1 I I [ I I I I I I I I I | I I I I I I I I I | I l_
j? to
m-$^
0
^fk -.5 1
rTi i I
40
• 11 • • i
50
I
I
60 70 Atomic Number
I L I I
80
I • 4T
90
Figure 1: Neutron-capture abundances in the ultra-metal-poor ([Fe/H] = -3.1) halo field giant star CS 22892-052 are plotted as filled circles with error bars, along with a scaled solar system r-process abundance curve (solid line). In the bottom panel, a differential comparison between individual elements and the scaled solar system r-process abundance distribution shows excellent agreement above Z = 56, but some deviations for lighter elements. From Ref. 5.
sites, such as neutron star mergers, less probable. The reasoning rests on the expected infrequency of neutron star mergers (no more than l/100th the rate of galactic supernovae), and thus on the larger nucleosynthetic output required from such r-process sites 6 . Since the ejecta of neutron star mergers and supernovae are expected to mix over similarly sized regions, the former should produce a larger scatter of enrichments in metal-poor stars. These and other arguments have led many to suspect that core-collapse supernova may be the correct site. There is good theoretical support for this conclusion. First, galactic chemical evolution studies indicate that the growth of r-process elements in the galaxy is consistent with low-mass Type II super-
28
novae in rate and distribution. More convincing is the fact that modelers 7 have shown that the conditions needed for an r-process (very high neutron densities, temperatures of 1-3 billion degrees) might be realized in a supernova. The identified location is the last material blown off the supernova, the material just above the mass cut. When this material is initially at small r, it is a very hot, neutron-rich, radiation-dominated gas containing neutrons and protons, with neutrons dominating. As it expands off the star and cools, the material first goes through a freezeout to a particles, a step that essentially locks up all the protons in this way. Then the as interact through reactions like a + a + a ->• 12 C a + a + n-»9Be
(11)
to start forming heavier nuclei. Unlike the big bang, the density is sufficiently high to allow such three-body interactions to bridge the mass gaps at A = 5,8. The a capture continues up to A ~ 80 in network calculations. The result is a small number of "seed" nuclei, a large number of as, and excess neutrons. These neutrons preferentially capture on the heavy seeds to produce an r-process. Of course, what is necessary is to have ~ 100 excess neutrons per seed in order to successfully synthesize heavy mass nuclei. While some calculations come close to achieving this, the entropies tend to fall short of what is needed. An attractive aspect of this site is the amount of matter ejected, about 1 0 - 5 - 1 0 ~ 6 solar masses, enough to produce the present galactic r-process metallicity for a reasonable supernova rate. It is clear that neutrino physics is an intimate part of the r-process. The supernova scenario described above is usually attributed to material ejected by the protoneutron star's neutrino wind. This wind is also responsible for regulating the essential proton/neutron chemistry of this material: the reactions ve + n «-• e~ +p and ue +p o e + +n control this physics. Nonstandard neutrino physics could be critical to the r-process. An oscillation of the type v e —> ^sterile c a n alter the n/p ratio, as it turns off the ves that destroy neutrons by charged-current reactions. The nuclear physics of the r-process tells us that the synthesis occurs when the nucleon soup is in the temperature range of (3-1) -109K, which, in the hot bubble r-process described above, corresponds to a freezeout radius of (600-1000) km and a time ~ 10 seconds after core collapse. The neutrino fluence after freezeout (when the temperature has dropped below 109K and the r-process stops) is then ~ (0.045-0.015) -1051 ergs/(100km) 2 . Thus, after completion of the r-process, the newly synthesized material experiences an intense flux of neutrinos. This brings up the question of whether the neutrino flux could have any effect on the r-process.
29
4
Neutrinos and the r-process
Rather than describe the exotic effects of neutrino oscillations on the r-process, mentioned briefly above, we will examine standard-model effects that are nevertheless quite interesting. The nuclear physics of this section - neutrino-induced neutron spallation reactions - is also relevant to recently proposed supernova neutrino observatories such as OMNIS and LAND. In contrast to our first discussion of the i/-process in producing 1 9 F and 1 1 B, it is apparent that neutrino effects could be much larger in the hot bubble r-process: the synthesis occurs much closer to the star than our Ne radius of 20,000 km. The r-process is completed in about 10 seconds (when the temperature drops to about one billion degrees), but the neutrino flux is still significant as the r-process freezes out. The net result is that the "post-processing" neutrino fluence - the fluence that can alter the nuclear distribution after the r-process is completed - is about 100 times larger than that responsible for fluorine production in the Ne zone. Recalling that 1/300 of the nuclei in the Ne zone interacted with neutrinos, and noting that the relevant neutrino-nucleus cross sections scale as A (a consequence of the sum rules governing first-forbidden neutrino cross sections), one quickly sees that the probability of a r-process nucleus interacting with the neutrino flux is approximately unity. Because the hydrodynamic conditions of the r-process are highly uncertain, one way to attack this problem is to work backward in time 8 . We know the final r-process distribution (what nature gives us) and we can calculate neutrino-nucleus interactions relatively well. Thus from the observed r-process distribution (including neutrino postprocessing) we can deduce what the rprocess distribution looked like at the point of freezeout. In Figs. 2 and 3, the "real" r-process distribution - that produced at freezeout - is given by the dashed lines, while the solid lines show the effects of the neutrino postprocessing for a particular choice of fluence. One important aspect of the figures is that the mass shift is significant. This has to do with the fact that a 20 MeV excitation of a neutron-rich, weakly bound nucleus allows multiple neutrons ( ~ 5) to be emitted. The relative contribution of the neutrino process is particularly important in the "valleys" beneath the mass peaks: the reason is that the parents on the mass peak are abundant, and the valley daughters rare. In fact, it follows from this that the neutrino process effects can be dominant for precisely seven isotopes (Te, Re, etc.) lying in these valleys. Furthermore if an appropriate neutrino fluence is picked, these isotope abundances are produced perfectly (given the abundance errors). The fluences are N = 82 peak :
0.031 • 1051 ergs/ (100km)2 /flavor
30
>*
140
Figure 2: Comparison of the r-process distribution that would result from the freezeout abundances near the A ~ 130 mass peak (dashed line) to that where the effects of neutrino postprocessing have been include (solid line). The fluence has been fixed by assuming that the A = 124-126 abundances are entirely due to the i/-process.
N = 126 peak :
0.015 • 10 51 ergs/(100km) 2 /flavor
(12)
values in fine agreement with those that would be found in a hot bubble rprocess. So this is circumstantial but significant evidence that the material near the mass cut of a Type II supernova is the site of the r-process: there is a neutrino fingerprint. 5
Neutrino Oscillations and Supernova Nucleosynthesis
There are some intriguing connections between supernova nucleosynthesis, the explosion mechanism, and neutrino oscillations. Several of these have to do with the distinctive temperature hierarchy of supernova neutrinos mentioned earlier. In contrast to solar neutrinos, where detailed nuclear physics determines the neutrino spectrum, the supernova neutrino temperature dependence
31
X*
170
175
1B0
185
190
195
200
Figure 3: As in Fig. 5, but for the A ~ 195 mass peak. The A = 183-187 abundances are entirely attributed to the ^-process.
on flavor is governed by very general arguments having to do with neutrino couplings to matter, as we have noted. While modelers differ somewhat in their estimates of neutrino temperatures, there is agreement that the heavy flavor neutrino mean energy is higher than that of the electron neutrinos, and that the ue temperature is lower than that of the ue&. One consequence is that neutrino oscillation signals in terrestrial detectors could be quite obvious at the time of the next galactic supernova. For example, if ve events prove to be substantially more energetic that De events, the natural interpretation would be oscillations between heavy-flavor and ve neutrinos, leading to an anomalously hot ve spectrum. One important aspects of supernova neutrino oscillations is their potential to probe the MSW mechanism over greatly extended parameter ranges. The neutrinos have fixed spectra after they decouple at the neutrinosphere, ~ 10 12 g/cm 3 , a density ten orders of magnitude greater than that at the core of the sun. It follows that neutrinos with masses in excess of 100 eV (thus 6m2 in
32
excess of (100 eV) 2 ) will experience an MSW crossing. These crossings remain adiabatic - that is, capable of converting neutrino flavor - for mixing angles as small as 1 0 - 5 , depending on the dm2 value. It follows that oscillations unobservable by any other means could be revealed in supernovae. Relevant to the present talk is the possibility that we will not need to wait for the next supernova: oscillation effects might be deduced from their effects on nucleosynthesis. For example, we have noted that a MSW oscillation between heavy and electron flavors would lead to an unusually hot ue spectrum. The proton/neutron chemistry of the hot nucleon soup blown off the protoneutron star is governed by the competition between the reactions ve + n ->• e~ + p De + p -> e + + n.
(13)
As the oscillation leads to a hotter ve spectrum but does not affect the Pes (which, for the usual mass hierarchy, do not experience an MSW crossing), the first reaction is enhanced while the second is unchanged. The matter is thus driven proton-rich, destroying any possibility of an r-process. Thus, as Fuller has argued 9 , a demonstration that the supernova "hot bubble" is the site of the r-process would impose very stringent constraints on ve «-• v v iil r oscillations. The constraints address the entire range of cosmologically interesting uT masses. Another r-process connection arose from efforts to explain the LSND, atmospheric, and solar neutrino results in four-neutrino schemes (three active and one approximately sterile). One such scheme involves a vrlvli doublet at about 2 eV, split in order to reproduce the atmospheric 5m2, and a light ve/vsteriie doublet, split to reproduce the solar dm2. Such a scheme can have a salutary effect on the r-process because of successive MSW crossings 10 . First the v^lvj flux is removed by an oscillation with vsterUe'i then a ve —> uM/uT oscillation can take place without a corresponding back reaction. With the ue flux reduced but the ue unaffected, the matter can be driven neutron rich. This occurs at a radius where the increase in available neutrons helps the r-process to succeed in producing the A ~ 190 mass peak. Oscillations could also influence our interpretation of the abundances of the rare isotopes 1 0 , 1 1 B, 9 Be, and 6,7 Li. The neutrino process on 12 C appears to produce a great deal of 1 1 B, consistent with its abundance. It also produces significant 7 Li, but very little 1 0 B, 9 Be, and 6 Li: neutral current neutrino reactions generally do not impart sufficient energy to 12 C to populate the higher threshold channels corresponding to these products. The neutrino process produces a 1 0 B/ 1 1 B ratio of ~ 0.05, while the true abundance ratio is ~ 0.25.
33
As a primary process, it predicts a linear growth of boron with metallicity (e.g., Fe). However, the textbook explanation for the synthesis of these elements is the interaction of cosmic ray protons with 12 C and 1 6 0 in the interstellar medium u . As the reactions involve high-energy protons, 1 0 B and 6 Li are readily produced: the 1 0 B/ 1 1 B ratio is ~ 0.5, about twice that observed. The cosmic ray mechanism becomes more effective as the interstellar medium is enriched in 12 C and i e O and thus, as a secondary process, produces a quadratic growth in B with metallicity. It is very possible that nature uses a mixture of these two mechanisms. If each contributed about 50% of the n B , the correct 10 B / U B ratio would result. However recent studies of metal-poor stars show that the boron grows linearly with Fe 1 2 : the production appears to be primary. This has encouraged several efforts to reformulate the cosmic ray mechanism as a primary process, e.g., by accelerating 12 C and 1 6 0 off a supernova on to target protons in the interstellar medium. (I believe the tasking of estimating the production resulting from such a scenario is highly uncertain, however.) Surprising quantities of 9 Be have also been observed in metal-poor halo stars. Since the v process is a primary process, one could consider whether the calculated productions of 10 B and other high-threshold products might have been underestimated. It is clear that the nuclear physics uncertainties affecting these channels are much greater than in the case of 1 1 B. But another possibility that has not been explored is neutrino oscillations. In fact, if the neutrino masses have a standard seesaw pattern, the atmospheric neutrino results suggest a <$m23 somewhat below 0.01 eV 2 , producing a ve «-> vT MSW crossing near 105 g/cm 3 . This is outside the r-process region, thus leaving that synthesis unaffected, but before the carbon zone at p ~ 8 • 102 — 5 • 103 g/cm 3 . The hot ve flux leads to enhanced production of 10 B through 12C(i>e,e~), as well as increased 7 Li through the burnup reaction 10 B(p,a) 7 Be. Numerical results will be published soon 14 . 6
Summary
The connections between neutrino nucleosynthesis and the supernova mechanism are rather remarkable. We have seen that neutrinos are directly responsible for important synthesis. In turn this synthesis can be exploited as a diagnostic of the explosion, e.g., as a monitor of yet unmeasured heavy-flavor temperatures in the v process, and as a constraint on the explosion dynamics in the case of the r-process. (The neutrino fluence derived from the v postprocessing "fingerprint" on the r-process constrains a product of the freezeout
34
radius and the expansion rate.) If one adds new neutrino physics to the equation, the occurence of a supernova "hot bubble" r-process places important new constraints on the entire range of dm2 relevant cosmologically. Four-neutrino scenarios postulated to account for the LSND, atmospheric, and solar neutrino results can enhance rprocess production in the vicinity of the A ~ 190 peak, and thus could account for current underproductions in this region. Finally, the relative mix of cosmic ray and v process synthesis of Li/Be/B would be affected by mixings of the ve governed by 6m2 near the atmospheric neutrino value. The most interesting aspect of all of this is the impact new abundances observations are having. It gives one hope that new v physics might be learned from supernovae even before the next flux of supernova us hits the earth. Acknowledgements I thank Scott Buries for providing, on behalf of the authors of Ref. 5, Fig. 1. This work was supported in part by the US Department of Energy. References 1. H-Th Janka, astro-ph/008432 (submitted to Astronomy and Astrophysics); A Mezzacappa, astro-ph/0010580 (to appear in the Proc. Nuclei in the Cosmos 2000); A Burrows, astro-ph/9805170 (to appear in the Proc. 9th Workshop in Nuclear Astrophysics). 2. G V Domogatsky and D K Nadyozhin, M.N.R.A.S. 178, 33P (1977); Sov. Astr. 22, 297 (1978); Sov. Astr. Lett. 6, 127 (1980); Ap. Space Sci. 70, 33 (1980). 3. S E Woosley and W C Haxton, Nature 334, 45 (1988); S E Woosley, D H Hartmann, R D Hoffman, and W C Haxton, Ap. J. 356, 272 (1990). 4. F X Timmes, S E Woosley, and T A Weaver, Ap. J. Supp. 98, 617 (1995). 5. C Sneden, J J Cowan, 11 Ivans, G M Fuller, S Buries, T C Beers, and J E Lawler, Ap. J. 533, L139 (2000). 6. Y-Z Qian, astro-ph/0003242 (to appear in Ap. J. Lett.) 7. S E Woosley and R D Hoffman, Ap. J. 395, 202 (1992). 8. W C Haxton, K Langanke, Y-Z Qian, and P Vogel, Phys. Rev. Lett. 78, 2694 (1997) and Phys. Rev. C 55, 1532 (1997). 9. G M Fuller, Phys. Repts. 227, 149 (1993). 10. G C McLaughlin, J M Fetter, A B Balantekin, and G M Fuller, Phys. Rev. C 59, 2873 (1999); D O Caldwell, G M Fuller, and Y-Z Qian, Phys. Rev. D 61, 123005 (2000).
35
11. J Audouze, in LiBeB, Cosmic Rays, and Related X- and Gamma-Rays, ed. R Ramaty, E Vangioni-Flam, M Casse, and K Olive (ASPCS vol. 71, San Francisco, 1999). 12. L M Reball, D K Duncan, S Johansson, J Thorburn, and B Fields, Ap. J. 507, 387 (1998). 13. R Ramaty and R E Lingenfelter, in LiBeB, Cosmic Rays, and Related X- and Gamma-Rays, ed. R Ramaty, E Vangioni-Flam, M Casse, and K Olive (ASPCS vol. 71, San Francisco, 1999) p. 104. 14. G M Fuller and W Haxton, in preparation.
Supernova Studies at ORLaND A. M E Z Z A C A P P A Physics Division, Oak Ridge National Laboratory Bldg. 6010, MS 6354, P-O. Box 2008 Oak Ridge, TN 37831-6354 E-Mail:
[email protected] A new facility to measure neutrino mass differences and mixing angles and neutrino-nucleus cross sections, such as the proposed ORLaND facility at Oak Ridge, would contribute to the experimental determination of vacuum mixing parameters and would provide an experimental foundation for the many neutrinonucleus weak interaction rates needed in supernova models. This would enable more realistic supernova models and a far greater ability to cull fundamental physics from these models by comparing them with detailed observations. Chargedand neutral-current neutrino interactions on nuclei in the stellar core play a central role in supernova dynamics, nucleosynthesis, and neutrino detection. Measurements of these reactions on select, judiciously chosen targets would provide an invaluable test of the complex theoretical models used to compute the neutrinonucleus cross sections.
1
Introduction
Core collapse supernovae are among the most energetic explosions in the Universe, releasing 10 53 erg of energy in the form of neutrinos of all flavors at the staggering rate of 10 57 neutrinos per second and 1045 Watts, disrupting almost entirely stars more massive than 8-10 MQ and producing and disseminating into the interstellar medium many of the elements in the periodic table. They are a key link in our chain of origins from the Big Bang to the formation of life on Earth; a nexus of nuclear physics, particle physics, fluid dynamics, radiation transport, and general relativity; and serve as laboratories for physics beyond the Standard Model and for matter at extremes of density, temperature, and neutronization that cannot be produced in terrestrial laboratories. Current supernova theory centers around the idea that the supernova shock wave—formed when the iron core of a massive star collapses gravitationally and rebounds as the core matter exceeds nuclear densities and becomes incompressible—stalls in the iron core as a result of enervating losses to nuclear dissociation and neutrinos. The failure of this "prompt" supernova mechanism sets the stage for a "delayed" mechanism, whereby the shock is reenergized by the intense neutrino flux emerging from the neutrinospheres carrying off the binding energy of the proto-neutron star 1,2 . The heating is mediated primarily by the absorption of electron neutrinos and antineutrinos on the dissociationliberated nucleons behind the shock. This process depends critically on the 36
37
neutrino luminosities, spectra, and angular distributions, i.e., on the multigroup (multi-neutrino energy) neutrino transport between the proto-neutron star and the shock. This past decade has also seen the emergence of multidimensional supernova models, which have investigated the role convection, rotation, and magnetic fields may play in the explosion 3 ' 4 ' 5 ' 6 ' 7 ' 8 ' 9,10 ' 11 ' 12 . Realistic supernova models will require extremely accurate neutrino radiation hydrodynamics, but unless advancements in neutrino transport are matched by equally important advancements in nuclear and weak-interaction physics, the efficacy of the former must be called into question. In particular, we must move forward to the use of ensembles of nuclei in the stellar core rather than a single representative nucleus when computing electron and electron neutrino capture during the critical core collapse phase. The use of a single representative nucleus has been the standard until now in virtually all supernova models. Moreover, in computing the electron and electron neutrino capture rates, and in general all of the neutrino-nucleus cross sections, detailed shell model computations must replace the parameterized approximations that have been used in the past. For example, recent work on electron capture up to mass 65 has shown that these parameterized rates can be orders of magnitude in error13. The electron capture rate is dominated by Gamow-Teller resonance transitions, with the Gamow-Teller strength distributed over many states. Parameterized treatments place the resonance at a single energy, and this energy is often grossly over- or underestimated relative to the Gamow-Teller centroid computed in realistic shell model computations, leading in turn to rates that are often grossly too small or too large, respectively. (If the centroid is underestimated, more electrons can participate in capture. The reverse is true if the centroid is too high.) Whereas generations of nuclear structure models will afford ever greater realism in the calculation of stellar core properties and the interactions of core nuclei with the neutrinos flowing through the core, nuclear experiments must be designed and carried out that will serve as guide posts for the theoretical predictions that must be made for the countless rates that enter into any realistic supernova or supernova nucleosynthesis model. In particular, we must have neutrino-nucleus cross section measurements that will help gauge neutrino capture and scattering predictions during stellar core collapse and during the p-, r- and ^-processes after core bounce. The plan of this paper is as follows. First we give an overview of the state of the art in one- and two-dimensional supernova models, confining our discussion to issues of neutrino transport and convection. For discussions of the role of general relativity, rotation, and magnetic fields in supernova models, the reader may begin with the papers by Bruenn et o/.14, Liebendorfer et a/.15,16, Fryer
38
and Heger11, Khokhlov et alP, and MacFadyen and Woosley17. Our focus will then turn to the nuclear and neutrino science that is input to these models and important for supernova neutrino detection, with an eye toward measurements that could be made at a stopped-pion facility such as ORLaND. 2
One-Dimensional Supernova Models
Although three decades of supernova modeling have established a theoretical framework, fundamental questions about the explosion mechanism remain. Is the neutrino heating sufficient, or are multidimensional effects such as convection and rotation necessary? Can the basic supernova observable, explosion, be reproduced by detailed spherically symmetric models, or are multidimensional models required? In all of their phenomenology, core collapse supernovae are not spherically symmetric. For example, neutron star kicks18 and the polarization of supernova emitted light19 cannot arise in spherical symmetry. Nonetheless, ascertaining the explosion mechanism and understanding every explosion observable are two different goals. To achieve both, simulations in one, two, and three dimensions must be coordinated. The neutrino energy deposition behind the shock depends sensitively on the neutrino luminosities, spectra, and angular distributions in the postshock region. Ten percent variations in any of these quantities can make the difference between explosion and failure in supernova models 7,20 . Thus, exact multigroup Boltzmann neutrino transport must be considered in supernova models. Past spherically symmetric simulations have implemented increasingly sophisticated approximations to Boltzmann transport: simple leakage schemes21, two-fluid models22, and multigroup flux-limited diffusion23,24'25. A generic feature of this last, most sophisticated approximation is that it underestimates the isotropy of the neutrino angular distributions in the heating region and, thus, the heating rate 26,27 . Failure to produce explosions in the past may have resulted from the use of transport approximations. To address this question, we model the core collapse, bounce, and postbounce evolution of a 13 M© star, beginning with the precollapse model of Nomoto and Hashimoto 28 , with a new neutrino radiation hydrodynamics code for both Newtonian and general relativistic spherically symmetric flows: AGILE-BOLTZTRAN. BOLTZTRAN is a three-flavor Boltzmann neutrino transport solver29,30, now extended to fully general relativistic flows15. In the simulation we include here31, it was employed in the 0(v/c) limit. AGILE is a conservative general relativistic hydrodynamics code 15 ' 32 . Its adaptivity enables us to resolve and seamlessly follow the shock through the iron core into the outer stellar layers.
39
0
100
200 300 400 time after bounce [ms]
500
Figure 1: We trace the shock, nuclear burning, and dissociation fronts (the shock and dissociation fronts are coincident), which carve out three regions in the (r,t) plane. A: Silicon. B: Iron produced by infall compression and heating. C: Free nucleons and alpha particles.
Figure 1, taken from the simulation of Mezzacappa et al?1 shows the radius-versus-time trajectories of equal mass (O.O1M0) shells in the stellar iron core 'and silicon layer in our Newtonian simulation. Core bounce and the formation and propagation of the initial bounce shock are evident. This shock becomes an accretion shock, decelerating, the core material passing through it. At ~ 100 ms after bounce, the accretion shock stalls at a radius ~ 250 km and begins to recede, continuing to do so over the next several hundred milliseconds. No explosion has developed in this model during the first ~ 500 ms. Thus, we are beginning to answer some fundamental questions in supernova theory. We have shown results from the first ~ 500 ms of our Newtonian core collapse supernova simulation with Boltzmann neutrino transport, initiated from a 13 M© progenitor. In light of our. implementation of Boltzmann transport, if we do not obtain explosions in this model or its general relativistic counterpart when they are completed, or in subsequent models initiated from different progenitors, it would suggest that either changes in our initial conditions (precollapse models) and/or input physics or the inclusion of multidimensional effects such as convection, rotation, and magnetic fields are required ingredients in the recipe for explosion. With the implementation of Boltzmann transport, this conclusion can be made unambiguously. In the past,
40
it was not clear whether failure or success in supernova models was the result of inadequate transport approximations or the lack of inclusion of important physics. With regard to improved input physics, the use of ensembles of nuclei in the stellar core rather than a single representative nucleus, computing the neutrino-nucleus cross sections with detailed shell model computations 13 , the inclusion of nucleon correlations in the high-density neutrino opacities 33,34 , and improvements in precollapse models 35,36 all have the potential to quantitatively, if not qualitatively, change the details of our simulations. Thus, it is important to note that our conclusions are drawn in the context of the best available input physics. 3
Two-Dimensional Supernova Models: Convection
Supernova convection falls into two categories: (1) convection near or below the neutrinospheres, which we refer to as proto-neutron star convection and (2) convection between the gain radius and the shock, which we refer to as neutrino-driven convection. Proto-neutron star convection may aid the explosion mechanism by boosting the neutrinosphere luminosities, transporting by convection hot, lepton-rich rich matter to the neutrinospheres. Neutrino-driven convection may aid the explosion mechanism by boosting the shock radius and the neutrino heating efficiency, thereby facilitating shock revival. 3.1
Proto-Neutron Star Convection
This mode of convection may develop owing to instabilities caused by lepton and entropy gradients established by the deleptonization of the proto-neutron star via electron neutrino escape near the electron neutrinosphere and by the weakening supernova shock. (As the shock weakens, it causes a smaller entropy jump in the material flowing through it.) Proto- neutron star convection is arguably the most difficult to investigate numerically because the neutrinos and the matter are coupled, and, consequently, multidimensional simulations must include both multidimensional hydrodynamics and multidimensional, multigroup neutrino transport. In certain regions of the stellar core, neutrino transport can equilibrate a convecting fluid element with its surroundings in both entropy and lepton number on time scales shorter than convection time scales, rendering the fluid element nonbouyant. This will occur in intermediate regimes in which neutrino transport is efficient but in which the neutrinos are still strongly enough coupled to the matter. Figures 2 and 3, from Mezzacappa et al?, demonstrate that this equilibration can in fact occur. Figure 2 shows the onset and
41
development of proto-neutron star convection in a 25 M Q model shortly after bounce in a simulation that did not include neutrino transport, i.e., that was a hydrodynamics-only run. Figure 3 on the other hand shows the lack of any significant onset and development of convection when neutrino transport was included in what was otherwise an identical model. Transport's damping effects are obvious. (The same result occurred in our 15 MQ model.) On the other hand, in the model of Keil et al?1, vigorous proto-neutron star convection developed, which then extended deep into the core as a deleptonization wave moved inward, owing to neutrinos diffusing outward. In this model, convection occurs very deep in the core where neutrino opacities are high and transport becomes inefficient in equilibrating a fluid element with its surroundings. It is important to note in this context that Mezzacappa et al. and Keil et al. used complementary transport approximations. In the former case, spherically symmetric transport was used, which maximizes lateral neutrino transport and overestimates the neutrino-matter equilibration rate; in the latter case, rayby-ray transport was used, which minimizes (zeroes) lateral transport and underestimates the neutrino-matter equilibration rate. These different outcomes clearly demonstrate that to determine whether or not proto-neutron star convection exists and, if it exists, is vigorous will require simulations coupling three-dimensional, multigroup neutrino transport and three-dimensional hydrodynamics. Moreover, realistic high-density neutrino opacities will be needed. 3.2
Neutrino-Driven
Convection
This mode of convection occurs directly between the gain radius and the stalled shock as a result of the entropy gradient that forms as material infalls between the two while being continually heated. In Figure 5, a sequence of two-dimensional plots of entropy are shown, illustrating the development and evolution of neutrino-driven convection in our 15 M© modeP. High-entropy, rising plumes and lower-entropy, denser, finger-like downflows are seen. The shock is distorted by this convective activity. In the Herant et alf simulations, large-scale convection developed beneath the shock, leading to increased neutrino energy deposition, the accumulation of mass and energy in the gain region, and a thermodynamic engine that ensured explosion, although Herant et al. stressed the need for more sophisticated multidimensional, multigroup transport in future models. [They used twodimensional "gray" (neutrino-energy-integrated, as opposed to multigroup) flux-limited diffusion in neutrino-thick regions and a neutrino lightbulb ap-
42 ENTROPY High
Low
m
i
liii t pb :--12 m»
tpbl~. 17 HIS
t l b =.. 27 m»
Figure 2: Two-dimensional entropy plots showing the evolution of proto-neutron star convection in our hydrodynamics-only 25 M Q model at 12, 17, and 27 ms after bounce.
EMTROFY Mgh
tow
tp^ltw»
t^trifw
tpb*»i
Figure 3: Two-dimensional entropy plots showing the evolution of proto-neutron star convection in our hydrodynamics-plus-neutrino-transport 25 M Q model at 12, 17, and 27 ms after bounce.
43 ENTROPY
l^.-r 131 ma
Ipi> = 212 ms
1 ^ = 312 ma
Figure 4: Two-dimensional entropy plots showing the evolution of neutrino-driven convection in our 15 M© model at 137, 212, and 512 ms after bounce.
proximation in neutrino-thin regions. In a light bulb approximation, the neutrino luminosities and rms energies are assumed constant with radius.] In the Burrows et al simulations6, neutrino-driven convection in some models significantly boosted the shock radius and led to explosions. However, they stressed that success or failure in producing explosions was ultimately determined by the values chosen for the neutrino spectral parameters in their gray ray-by-ray (one-dimensional) neutrino diffusion scheme. (In spherical symmetry (ID), all rays are the same. In a ray-by-ray scheme in axisymmetry (2D), not all rays are the same, although the transport along each ray is a ID problem. In the latter case, lateral transport between rays is ignored.) Focusing on the neutrino luminosities, Janka and Miller 7 , using a central adjustable neutrino lightbulb, conducted a parameter survey and concluded that neutrino-driven convection aids explosion only in a narrow luminosity window (±10%), below which the luminosities are too low to power explosions and above which neutrino-driven convection is not necessary. In more recent simulations carried out by Swesty10 using two-dimensional gray iux-limited diffusion in both neutrino-thick and neutrino-thin regions, it was demonstrated that the simulation outcome varied dramatically as the matter-neutrino "decoupling point," which in turn sets the neutrino spectra in the heating region, was varied within reasonable limits. (The fundamental problem in gray transport schemes is that the neutrino spectra, which are needed for the heating rate, are not computed. The spectra
44
are specified by choosing a neutrino "temperature," normally chosen to be the matter temperature at decoupling. In a multigroup scheme, the spectra are by definition computed.) In our two-dimensional models, the angle-averaged shock radii do not differ significantly from the shock trajectories in their onedimensional counterparts, and no explosions are obtained, as seen in Figure 6. Neither the luminosities nor the neutrino spectra are free parameters. Our twodimensional simulations implemented spherically symmetric (ID) multigroup flux-limited diffusion neutrino transport, compromising transport dimensionality to implement multigroup transport and a seamless transition between neutrino-thick and neutrino-thin regions. In light of the neutrino transport approximations made, the fact that all of the simulations have either been one- or two-dimensional, and the mixed outcomes, next-generation simulations will have to reexplore neutrino-driven convection in the context of three-dimensional simulations that implement more realistic multigroup three-dimensional neutrino transport. 4
Neutrino-Nucleus Cross Sections
Neutrino-nucleus cross sections of relevance to supernova astrophysics fall into three categories: cross sections for (1) supernova dynamics, (2) supernova nucleosynthesis, and (3) terrestrial supernova neutrino detection. 4-1
Supernova Dynamics
Whether or not a supernova occurs is set at the time the shock forms and the entire post-stellar-core-bounce evolution is set in motion. Where the shock forms in the stellar core at bounce and how much energy it has initially are set by the "deleptonization" of the core during collapse. The deleptonization occurs as electrons are captured on the free protons and iron-group nuclei in the core, producing electron neutrinos that initially escape. Deleptonization would be complete if electron capture continued without competition, but at densities of order 1 0 1 1 - 1 2 g/cm 3 , the electron neutrinos become "trapped" in the core, and the inverse reactions—charged-current electron neutrino capture on neutrons and iron-group nuclei—begin to compete with electron capture until the reactions are in weak equilibrium and the net deleptonization of the core ceases on a core collapse time scale. The equilibration of electron neutrinos with the stellar core occurs at densities between 10 1 2 ~ 1 3 g/cm 3 . Additionally, as the stellar core densities increase, the characteristic nuclei in the core increase in mass, owing to a competition between Coulomb contributions to the nuclear free energy and nuclear surface tension. For densities of order 10 13
45
P (g/cm3) 10 11 10 12 10 13
Ye 0.4 0.35 0.3
T (MeV) 1 2 4
A
Z
70 88 138
30 36 52
< e„e > (MeV) 12 19 43
g/cm 3 , the nuclear mass is of order 140. Thus, cross sections for chargedcurrent electron neutrino capture on iron-group nuclei through mass 100 are needed to accurately simulate core deleptonization and to accurately determine the postbounce initial conditions. Table 1 summarizes the thermodynamic conditions in the core at the three densities discussed above and gives the representative nuclear mass and charge and mean electron neutrino energy. The data were taken from a core collapse simulation carried out by Mezzacappa and Bruenn 38,39 . Electron neutrino capture would remain important until the neutrinos equilibrate with the matter, which in our simulation would occur when the representative nucleus in the core is between mass 88 and 138. The size of the inner, unshocked core is proportional to < Ye > 2 , where < Ye > is the mean electron fraction in the inner core. Moreover, the shock loses ~ 10 51 erg of energy (an explosion energy) for every 0.1 MQ it dissociates, which is ~10-20% of the total inner core mass. Thus, an ~5-10% change in the mean electron fraction would have a significant impact on the postbounce evolution. The mean electron fraction at bounce results from many capture reactions during infall (on both protons and nuclei; we focus on nuclei here), and it is clear that accurate electron and electron neutrino capture rates are needed and that theory must be checked against experiment even if only in a few strategic cases. One goal of the proposed ORLaND facility will be to measure the cross section for electron neutrino charged-current capture on 56 Fe: •
56
Fe(z/ e ,e-) 5 6 Co
Pioneering measurements of this cross section have been performed by the KARMEN collaboration with an experimental uncertainty ~ 50%. Further measurements are required to achieve an accuracy ~ 10% to adequately test theoretical models. Moreover, the same proposed technique to measure this cross section can be used to measure the electron neutrino capture cross section on any of the following nuclei, several of which are in the critical nuclear mass range mentioned above: 7 Li, 9 Be, n B , 27 A1, 40 Ca, 51 V, 52 Cr, 55 Mn, 59 Co, 93 Nb, 115 In, 181 Ta, and 209 Bi.
46
4-2
Supernova Nucleosynthesis
There are several "processes" that define supernova nucleosynthesis: (1) Explosive nucleosynthesis, which occurs as a result of compressional heating by the supernova shock wave as it passes through the stellar layers. (2) Neutrino nucleosynthesis or a "neutrino process," which occurs due to nuclear transmutations in the stellar layers prior to shock passage. (3) A rapid neutron capture or "r" process, which occurs in the neutrino-driven wind that emanates from the proto-neutron star after the explosion is initiated. The neutrinos both drive the wind and interact with the nuclei in it. Moreover, transmutations produced in (2) are postprocessed in (1). Thus, neutrino-nucleus interactions are central to all three nucleosynthesis processes, although indirectly to process Neutrino
Nucleosynthesis
Neutrino nucleosynthesis is driven by the spallation of protons, neutrons, and alpha particles from nuclei in the stellar layers by the intense neutrino flux that is emanating from the central proto-neutron star powering the supernova40. Moreover, neutrino nucleosynthesis continues after the initial inelastic scattering reactions and the formation of their spallation products. The neutrons, protons, and alpha particles released continue the nucleosynthesis through further reactions with other abundant nuclei in the high-temperature supernova environment, generating new rare species. Neutrino nucleosynthesis occurs in two stages: (1) through the neutrino irradiation and nuclear reactions prior to shock arrival and (2) through the continuation of nuclear reactions induced by neutrinos as the stellar layers expand and cool. Neutrino nucleosynthesis is thought to be responsible for the production of, for example, 1 1 B, 1 9 F, and two of Nature's rarest isotopes: 138 La and 180 Ta. The production of the two isotopes, U B and 1 0 B, appears observationally to be linear with metalicity, i.e., primary mechanisms that operate early in the history of our galaxy produce as much of these isotopes as secondary (quadratic) mechanisms that operate after the Galaxy has been enriched with metals. On the other hand, according to current models, neutrino nucleosynthesis in supernovae, which is a primary process, is not expected to have produced much 1 0 B, unlike the secondary process, cosmic ray spallation. Thus, a laboratory calibration of the spallation channels producing these two isotopes that can be used in conjunction with future HST observations discriminating between 10 B and n B would be invaluable in resolving this controversy and in supporting the theory that neutrino nucleosynthesis in supernovae is an important source of n B in the Galaxy 41 . n B and 10 B are produced through the following spallation channels:
47 •
12
C(j/,i/'p)nB 12
C(v,v'n)nC{e+,v)nB
• •
12
C(i/,i/'d) 10 B
•
12
C(j/,z/pn) 10 B
The final abundance of 19 F produced in a supernova can serve as a "supernova thermometer." If the abundance of 1 9 F produced in the supernova is attributed to neutrino nucleosynthesis, the ratio of [ 19 F/ 2O Ne]/[ 19 F/ 2O Ne] 0 (the denominator is the measured ratio in the Sun) is a measure of the muon and tau neutrinosphere temperatures 41 . 19 F is produced through the following spallation channels41: •
20
Ne(z/, z/n) 20 Ne* —> 19 Ne + n —> 19 F + e+ + i/e + n
•
20
Ne(t/, j/n) 2 0 Ne* —> 19 F + p
No obvious site for the production of the rare isotopes, 138 La and 180 Ta, has been proposed. That they can be produced via neutrino nucleosynthesis in supernovae is compelling, and may be very important in that their existence, however rare, may be a fingerprint of the neutrino process. 138 La, and 180 Ta are produced through the following spallation channels: •
139
La(i/,^'n) 138 La
•
181
T a (i,,i/n) 1 8 0 Ta
Experiments to measure the cross sections for all of these spallation channels are being considered as part of a second wave of experiments at ORLaND. The r-Process The site for the astrophysical r-process (rapid neutron capture process) is not yet certain, but the leading candidate is the neutrino-driven wind emanating from the proto-neutron star after a core collapse supernova is initiated 42 . The r-process is thought to be responsible for roughly half of the Solar System's supply of heavy elements. As the neutrino-driven wind expands rapidly and cools, charged particle reactions "freeze out" while neutron capture reactions continue on the "seed" nuclei present at freeze-out. Neutron capture (71,7) reactions come into equilibrium with neutron disintegration (7, n) reactions as an equilibrium is established between the free neutrons and the nuclei in the wind. The (n, 7)-(7, n) equilibrium produces nuclei that are quite neutron rich. Nuclei with half lives short compared to the time scale for the r-process
48
beta decay, producing nuclei with higher Z and leading to the synthesis of heavier elements. The simultaneous operation of these three types of reactions in the wind and the accompanying nucleosynthesis constitutes the r-process 43 . Qian et al. 4 4 in both the (n,j) f+ (~f,n) equilibrium and the "postprocessing phase" after these reactions fall out of equilibrium have demonstrated that neutrino-induced reactions can significantly alter the r-process path and its yields. In the presence of a strong neutrino flux, ^-induced charged current reactions on the waiting point nuclei at the magic neutron numbers N = 50, 82,126 might compete with beta decays and speed up passage through the bottlenecks there. Also, neutrinos can inelastically scatter on r-process nuclei via i/e-induced charged-current reactions and z/-induced neutral-current reactions, leaving the nuclei in excited states that subsequently decay via the emission of one or more neutrons. This postprocessing may for example shift the abundance peak at A = 195 to smaller mass. Taking things one step further, Haxton et al. 45 pointed out that neutrino postprocessing effects would provide a fingerprint of a supernova r-process. Eight abundances are particularly sensitive to the neutrino postprocessing: 124 Sn, 125 Te, 126 Te, 183 W, 184 W, 185 Re, 186 W, and 187 Re. Observed abundances of these elements are consistent with the postprocessing of an r-process abundance pattern in a neutrino fluence consistent with current supernova models. On a more pessimistic note, Meyer, McLaughlin, and Fuller46 have investigated the impact of neutrino-nucleus interactions on the r-process yields and have discovered that electron neutrino capture on free neutrons and heavy nuclei (in the presence of a strong enough neutrino flux) can actually hinder the r-process by driving the neutrino-driven wind proton rich, posing a severe challenge to theoretical models. During the r-process and subsequent postprocessing in the supernova neutrino fluence, neutrinos interact with radioactive, neutron-rich nuclei. Thus, relevant direct neutrino-nucleus measurements cannot be made. However, indirect measurements of charged- and neutral-current neutrino-nucleus interactions on stable nuclei that serve to gauge theoretical predictions would be invaluable. 4-3
Supernova Neutrino Detection
The nineteen neutrino events detected by 1MB and Kamiokande for SN1987A confirmed the basic supernova paradigm—that core collapse supernovae are neutrino-driven events—and marked the birth of extra-Solar-System neutrino astronomy. For a Galactic supernova, thousands of events will be seen by Super-K and SNO, which, for the first time, will give us detailed neutrino
49 "lightcurves" and bring us volumes of information about the deepest regions in the explosion. In turn, these lightcurves can be used to test and improve supernova models and their offshoot predictions. Moreover, comparing these detailed neutrino lightcurves with sophisticated supernova models could provide evidence for neutrino oscillations. Among the neutrino-nucleus interactions of relevance for supernova neutrino detection are neutrino interactions on deuterium in SNO, 1 6 0 in Super-K, and 5 6 Fe and 206,207,208pb i n t h e p r o p o s e d neutrino detector, OMNIS. Deuterium: SNO T h e four main channels for supernova neutrino detection in SNO are: v + e~ —> i/+e~, i/+d — • j z + p + n , ue+d —> p + p + e ~ , and i> e +d — • n + n + e + . Measurement of the reaction • ve + d — • p + p + e at O R L a N D , which is being considered to calibrate the reaction p + p —> d+e+ + ve (part of the chain of reactions powering the Sun), would also provide a calibration of the SNO neutrino detector. Monte Carlo studies suggest t h a t two years of d a t a in approximately thirty fiducial tons of D2O would yield a cross section measurement with an accuracy of a few percent 4 7 , which, in turn, will enable a more accurate interpretation of the SNO d a t a from the next Galactic supernova. T h e deuterium measurement is among the first wave of planned experiments at O R L a N D . Oxygen: Super-K T h e charged-current reaction 160(ue, e~)16F is the principle channel for electron neutrino interactions for thermal sources in the range Tv^ > 4 — 5 MeV and its rate exceeds t h a t of neutrino-electron scattering by an order of magnitude for TUe > 7 — 9 MeV 48 . Moreover, the electron angular distribution is strongly correlated with the electron neutrino energy, providing a way to measure the incident neutrino energy and, consequently, the electron neutrino spectra. By inference, one would then be able to measure, for example, the electron neutrinosphere t e m p e r a t u r e in a core collapse supernova, i.e., we would have a supernova thermometer 4 7 . In addition, the appearance of back-angle electron emission from this reaction in, for example, Super-K would result from very energetic electron neutrinos, more energetic t h a n predicted by supernova models. This would be evidence for flavor oscillations 47 . Muon and tau neutrinos in the stellar core couple to the core material only via neutral currents, whereas electron neutrinos and antineutrinos couple via both neutral and charged currents. As a result, the former decouple at higher density and, therefore, temperature, and have harder spectra. In fact, terrestrial detection
50
of the tau and electron neutrinos from a Galactic or near-extra-Galactic core collapse supernova may be our only hope of ever observing oscillations between these two neutrino flavors if the mixing angle is small. Utilizing reactions on 1 6 0 , Langanke, Vogel, and Kolbe49 have suggested a novel way of also unambiguously identifying muon and tau neutrino signatures in Super-K. The large average energies for these neutrino flavors are sufficient to excite giant resonances via the neutral-current reactions 1 6 0 ( J ^ > T , V T)160* . These resonances are above particle threshold and subsequently decay via the emission of protons, neutrons, and gamma rays. The gamma rays would provide the muon and tau neutrino signatures. The two decay channels are: 16 0 * __j.i5 O + n + 7 and l e O* —S-15 TV + p + 7. Thus, accurate measurements of both charged- and neutral-current neutrino cross sections on 1 6 0 would be foundational to interpreting the neutrino data from the next Galactic core collapse supernova and to using that data to potentially observe, for the first time, flavor oscillations involving the tau and electron neutrinos. An experiment to measure the cross section for: 16
0(ve,e-)16F
•
is among the first proposed experiments at ORLaND. Future experiments may focus on the cross sections for: .
16
0(^,^n7)150
.
16
0(^,^P7)157V
Iron and Lead: OMNIS The use of iron and lead in OMNIS would provide yet another way of measuring neutrino oscillations in core collapse supernovae50. Iron has a sufficiently high threshold for neutron production via charged-current neutrino interactions that such production is negligible, whereas, in lead, neutrons are produced by both charged- and neutral-current interactions. Oscillations between the more energetic muon and tau neutrinos and the electron neutrinos would boost the charged-current event rate while leaving the neutral-current rate roughly unchanged. Thus, the ratio of the event rate in lead to that in iron would serve as an indicator that oscillations had occurred. To develop OMNIS, experiments to measure the neutrino-iron and neutrinolead cross sections at ORLaND have been proposed. For iron, the neutralcurrent reaction: •
56
Fe(j/, i/) 5 6 Fe* — ^ 5
Fe +
n
51 Neutrino Spectra, 100ms After Bounce — —
Bectron Neutrino Bectron Antlneutrino I* h Neutrino
./..
40 Energy [MeV]
Figure 5: Neutrino spectra from a supernova simulation with Boltzmann neutrino transport initiated from a 13 M Q progenitor. The simulation is fully general relativistic, and the spectra are computed at a radius of 500 km 1 6 .
dominates. For lead, a total cross section would be measured resulting from the following neutral- and charged-current channels: •
208
Pb(j/, z/) 208 Pb* — j - 2 0 7 Pb + n
208
Pb(i/, v1
208 Pb(i/
e ,e-)
3
Pb*
208
.206
Pb + n + n
Bi'—)- 2 0 7 Bi + n
including the channels for the isotopes 2 0 6 Pb and 2 0 7 Pb. The iron and lead cross section measurements are among the first proposed experiments at ORLaND. 5
ORLaND
A new facility to measure neutrino-nucleus cross sections, such as the proposed ORLaND facility at Oak Ridge, would provide an experimental foundation for the many neutrino-nucleus weak interaction rates needed in supernova models. Indeed, we are presented with a unique opportunity, given the intensity of the SNS as a neutrino source and given the overlap (shown in Figures 5 and 6) between the spectra of SNS and supernova neutrinos, to make such measurements. This would enable more realistic supernova models and allow us to cull
52
0
&.0S
0,2
&J5
9.2
9
&.&S
&.J
9.15
9.2
Energy, GeV
Energy, GeV
Energy, GeV
Energy, GeV
Figure 6: SNS neutrino spectra.
fundamental physics from these models with greater confidence by comparing them with detailed observations. Charged- and neutral-current neutrino interactions on nuclei in the stellar core play a central role in supernova dynamics, nucleosynthesis, and neutrino detection. Measurements of these reactions on select, judiciously chosen targets would provide an invaluable test of the complex theoretical models used to compute the neutrino-nucleus cross sections. A cknow l e d g m e n t s A.M. is supported at the Oak Eidge National Laboratory, managed by UTBattelle, LLC, for the U.S. Department of Energy under contract DE-AC05OO0E22725. AM would like to acknowledge many illuminating discussions with Frank Avignone, John Beacom, Jeff Blackmon, Dick Boyd, David Dean, Yuri Efremenko, Jon Engel, George Fuller, Wick Haxton, Raph Hix, Ken Lande, Karlheinz Langanke, Matthias Liebendorfer, Gabriel Martinez-Pinedo, Gail McLaughlin, Mike Strayer, and Friedel Thielemann, all of which contributed significantly to this manuscript. References 1. J. E. Wilson. In J. M. Centrella, J. M. LeBlanc, and E. L. Bowers, editors, Numerical Astrophysics, page 422, Boston, 1985. Jones and
53
Bartlett. 2. H. H. Bethe and J. R. Wilson. Astrophysical Journal, 295:14, 1985. 3. M. Herant, W. Benz, and S. A. Colgate. Astrophysical Journal, 395:642, 1992. 4. D. S. Miller, J. R. Wilson, and R. W. Mayle. Astrophysical Journal, 415:278, 1993. .5. M. Herant, W. Benz, W. R. Hix, C. L. Fryer, and S. A. Colgate. Astrophysical Journal, 435:339, 1994. 6. A. Burrows, J. Hayes, and B. A. Fryxell. Astrophysical Journal, 450:830, 1995. 7. H.-Th. Janka and E. Miiller. Astronomy and Astrophysics, 306:167, 1996. 8. A. Mezzacappa, A. C. Calder, S. W. Bruenn, J. M. Blondin, M. W. Guidry, M. R. Strayer, and A. S. Umar. Astrophysical Journal, 493:848, 1998. 9. A. Mezzacappa, A. C. Calder, S. W. Bruenn, J. M. Blondin, M. W. Guidry, M. R. Strayer, and A. S. Umar. Astrophysical Journal, 495:911, 1998. 10. F. D. Swesty. In A. Mezzacappa, editor, Stellar Explosions, Stellar Evolution, and Galactic Chemical Evolution, page 539, Bristol, 1998. IoP Publishing. 11. C. L. Fryer and A. Heger. astro-ph/9907433. 12. A. M. Khokhlov, P. A. Hoflich, E. S. Oran, J. C. Wheeler, and A. Yu. Chtchelkanova. Astrophysical Journal, 524:L107, 1999. 13. K.-H. Langanke and G. Martinez-Pinedo. Nuclear Physics, A673:481, 2000. 14. S. W. Bruenn, K. R. DeNisco, and A. Mezzacappa. Astrophysical Journal,, submitted, 2000. 15. M. Liebendorfer. PhD thesis, University of Basel, Basel, Switzerland, 2000. 16. M. Liebendorfer, A. Mezzacappa, F.-K. Thielemann, O. E. B. Messer, W. R. Hix, and S. W. Bruenn. Physical Review Letters, submitted, 2000. 17. A. I. MacFadyen and S. E. Woosley. Astrophysical Journal, 524:262, 1999. 18. C. L. Fryer, A. Burrows, and W. Benz. Astrophysical Journal, 496:333, 1998. 19. J. C. Wheeler. In S. S. Holt, editor, Cosmic Explosions, A Conference Proceedings, Melville, 2000. AIP. 20. A. Burrows and J. Goshy. Astrophysical Journal Letters, 416:L75, 1993.
54
21. K. A. Van Riper and J. M. Lattimer. Astrophysical Journal, 249:270, 1981. 22. J. Cooperstein, L. J. Van Den Horn, and E. A. Baron. Astrophysical Journal, 309:653, 1986. 23. W. D. Arnett. Astrophysical Journal, 218:815, 1977. 24. S. W. Bruenn. In M. W. Guidry and M. R. Strayer, editors, First Symposium on Nuclear Physics in the Universe, page 31, Bristol, 1993. IOP Publishing. 25. J. R. Wilson and R. W. Mayle. Physics Reports, 227:97, 1993. 26. H.-Th. Janka. Astronomy and Astrophysics, 256:452, 1992. 27. O. E. B. Messer, A. Mezzacappa, S. W. Bruenn, and M. W. Guidry. Astrophysical Journal, 507:353, 1998. 28. K. Nomoto and M. Hashimoto. Physics Reports, 163:13, 1988. 29. A. Mezzacappa and S. W. Bruenn. Astrophysical Journal, 405:669, 1993. 30. A. Mezzacappa and O. E. B. Messer. Journal of Computational and Applied Mathematics, 109:281, 1998. 31. A. Mezzacappa, M. Liebendorfer, O. E. B. Messer, W. R. Hix, F.-K. Thielemann, and S. W. Bruenn. Physical Review Letters, submitted, 2000. 32. M. Liebendorfer and F.-K. Thielemann. In E. Aubourg, T. Montmerle, J. Paul, and P. Peter, editors, Nineteenth Texas Symposium on Relativistic Astrophysics, Amsterdam, 2000. Elsevier Science B. V. 33. A. Burrows and R. F. Sawyer. Physical Review, C58:554, 1998. 34. S. Reddy, M. Prakash, J. M. Lattimer, and J. A. Pons. Physical Review, C59:2888, 1999. 35. G. Bazan and W. D. Arnett. Astrophysical Journal, 496:316, 1998. 36. H. Umeda, K. Nomoto, and T. Nakamura. In A. Weiss, editor, The First Stars, Berlin, 2000. Springer. 37. W. Keil, H.-Th. Janka, and E. Miiller. Astrophysical Journal Letters, 473:L111, 1996. 38. A. Mezzacappa and S. W. Bruenn. Astrophysical Journal, 405:637,1993. 39. A. Mezzacappa and S. W. Bruenn. Astrophysical Journal, 410:740,1993. 40. S. E. Woosley, D. H. Hartmann, R. D. Hoffman, and W. C. Haxton. Astrophysical Journal, 356:272, 1990. 41. W. Haxton. In Proceedings of the 1998 TASI Summer School, 1999. 42. S. E. Woosley, G. J. Mathews, J. R. Wilson, R. D. Hoffman, and B. S. Meyer. Astrophysical Journal, 433:229, 1994. 43. B. S. Meyer. Annual Reviews of Astronomy and Astrophysics, 32:153, 1994. 44. Y.-Z. Qian, W. C. Haxton, K.-H. Langanke, and P. Vogel. Physical
55
Review, C55:1532, 1997. 45. W. C. Haxton, K.-H. Langanke, Y.-Z. Qian, and P. Vogel. Physical Review Letters, 78:2694, 1997. 46. B. S. Meyer, G. C. McLaughlin, and G. M. Fuller. Physical Review, C58:3696, 1998. 47. ORLaND Collaboration. Workshop on Neutrino-Nucleus Physics Using a Stopped-Pion Facility (ORLaND), May 23-26, 2000, Oak Ridge, Tennessee 48. W. Haxton. Physical Review, D36:2283, 1987. 49. K.-H. Langanke, P. Vogel, and E. Kolbe. Physical Review Letters, 76:2629, 1996. 50. R. N. Boyd. In S. Kubono and T. Kajino, editors, Origin of Matter and Evolution of Galaxies, Singapore, 2000. World Scientific.
A S T R O P H Y S I C A L N E U T R I N O S : 20th C E N T U R Y A N D BEYOND
J. N. B A H C A L L , b a s e d u p o n I U P A P C e n t e n n i a l Lecture, N e u t r i n o 2000 Institute
for Advanced E-mail:
Study, Princeton, NJ 08540, www.sns.ias.edu/~jnb
USA
I summarize the first four decades of solar neutrino research and suggest what may be possible to learn with extragalactic neutrinos and with solar neutrinos in the next decade.
1
Introduction
I am delighted to be able to participate in this symposium in honor of Frank Avignone. There are very few really talented experimental physicists, people who can get things done and who you know will make all possible tests to make sure that the answers they report are right. There are also very few intellectuals who are also wonderful people and who light up a room wherever they happen to be. Frank is one of tiny group of people at the intersection of these two small sets. He is a very special person and I feel privileged that we are friends. I have decided to limit my remarks to two specific topics: solar neutrinos and extragalactic neutrinos. I will not say anything about the enormous achievements in the prediction and detection of supernova neutrinos and in the calculations of neutrino cooling processes for stars that are not exploding. I will also not discuss the role of neutrinos in Big Bang nucleosynthesis nor in cosmology. There are lots of grand things to say about these subjects, and many other topics in neutrino astrophysics, but I will not discuss them today. I will take a somewhat historical approach and emphasize those aspects of the development of our subject which may help guide our thinking about what we should do in the future. I will begin with solar neutrinos and then switch abruptly to extragalactic neutrinos. 56
57
Figure 1: Bruno Pontecorvo wrote In 1967: 'From the point of view of detection possibilities, an ideal object is the sun.' Figure courtesy of S. Bilenky .
2 2.1
Solar neutrinos Bruno Pontecorvo and Ray Davis
I want to begin by paying tribute to two of the great scientists and pioneers of neutrino astrophysics, Ray Davis and Bruno Pontecorvo. Bruno i r s t suggested using chlorine as a detector of neutrinos in a Chalk Eiver report written in 1946. Eay followed Bruno's suggestion and the careful unpublished feasibility study of Louie Alvarez. Using with care and skill a chlorine detector and reactor neutrinos, Eay showed in 1955-1958 that v€ and ve were different. About a
58
decade later, Ray first detected solar neutrinos, laying the foundation for the studies we shall hear about today. In 1967, one year before the first results of Ray's chlorine solar neutrino experiment were announced, Bruno published a prophetic paper entitled: 'Neutrino Experiments and the Problem of Conservation of Leptonic Charge' 1. In this paper, Bruno suggested many different experiments that could test whether leptonic charge was conserved. The grandchildren of most of these experiments are being discussed in present-day neutrino conferences. Bruno included a short section in his paper that he called 'Oscillations and Astronomy.' In this section, Bruno wrote: "Prom the point of view of detection possibilities, an ideal object is the sun." What a wonderfully contemporary statement! Bruno, like most particle physicists of the 1960's and perhaps 1970's and 1980's, did not believe astrophysical calculations could be reliable. He wrote in this same section on oscillations and astronomy: "Unfortunately, the weight of the various thermonuclear reactions in the sun, and the central temperature of the sun are insufficiently well known in order to allow a useful comparison of expected and observed solar neutrinos, from the point of view of this article." [This was 30 years before the precise confirmation of the standard solar model by helioseismology.] To support his claim, Bruno referenced only his 1946 Chalk River report, which mentioned the sun in just two sentences. Bruno did cite our calculations of the solar neutrino fluxes elsewhere in his 1967 paper, but they seem not to have affected his thinking. What can we learn from this bit of history? When Ray and I wrote our PRL papers arguing that a chlorine detector of 600 tons could observe solar neutrinos, we never discussed the possibility of using neutrinos to learn about particle physics. The only motivation we gave was "...to see into the interior of a star and thus verify directly the hypothesis of nuclear energy generation in stars." 2 Why did we not discuss using neutrinos for particle physics? Frankly, because we never thought about it. And even if we had, we would have known better than to mention it to our particle physics friends. Bruno had the insight and the vision and indeed the courage to argue that astronomical neutrinos could potentially give us unique information about neutrino characteristics. His paper is all the more remarkable because it was published a year before the first results of the chlorine experiment showed that the rate Ray observed was less than our calculated rate. We learn from these events that pioneering experiments can lead to important results in areas that are unanticipated. We will come back to this conclusion at the end of this talk.
59
Gallium
nui |ChIorm©
MI
- SuperK, SNO t~ r »
x O
c U
Z
Neutrino Energy (MeV) Solar neutrino energy spectrum Figure 2: The energy spectrum of neutrinos from the pp chain of interactions in the Sun, as predicted by the standard solar model. Neutrino l u x e s from continuum sources (such as p — p and 8 B) are given in the units of counts per cm 2 per second. The percentage errors are the calculated la uncertainties in the predicted fluxes. The p — p chain is responsible for more than 98% of the energy generation in the standard solar model. Neutrinos produced in the carbon-nitrogen-oxygen CNO chain are not important energetically and are difficult to detect experimentally. The arrows at the top of the i g u r e indicate the energy thresholds for the ongoing neutrino experiments. This spectrum is from BP98 3 .
2.2 Standard Model Predictions Figure 2 shows the calculated solar neutrino spectrum predicted by the Standard solar model. The percentage errors are the calculated la uncertainties in the predicted luxes, based upon the published errors of the measured quantities and on many calculations of standard solar models. The total intensities and the energy spectra shown in Fig. 2 are now widely used to interpret, and indeed to plan, solar neutrino experiments such as those discussed at Neutrino 2000: chlorine, Super-Kamiokande, SNO, SAGE, GALLEX, GNO, and
60
Total Hates: S t a n d a r d Model vs. E x p e r i m e n t Bahcall-Pinsonneault 98
• 7*±B **&;
74±7 0.54±0.0? i«B"
0.47±0.02 i j l2.56±0.23
SuperK
Kamioka
SAGE
GALLEX/GNO
CI
Theory •
?Be
mm
•
p
~~p*
pep
m CNO
Figure 3: Comparison of measured rates and standard-model (BP98) predictions for six solar neutrino experiments. The unit for the radiochemical experiments (chlorine and gallium) is SNU (10"" 36 interactions per target atom per sec); the unit for the water-Cerenkov experiments (Kamiokande and Super-Kamiokande) is the rate predicted by the standard solar model plus standard electroweak theory. The experimental results are described by Lande, Suzuki, Gavrin, and Belotti at Neutrino 2000.
BOEEXINO. Figure 3 compares the calculated versus the measured rates for the six solar neutrino experiments for which results have been reported. Assuming nothing happens to the neutrinos after they are created, the measured rates range from 33% ± 5 % of the calculated rate (for chlorine) to 58% ±7%. As is now well known, the observed rates cannot be fit (at a C.L. of about 99%) with any linear combination of undistorted solar neutrino energy spectra. Today we know that there are three reasons that the calculations of solar neutrino fluxes are robust: 1) the availability of precision measurements and precision calculations of input data that have been gradually refined over four
61
decades; 2) the intimate connection between neutrino fluxes and the measured solar luminosity; and 3) the measurement of the helioseismological frequencies of the solar pressure-mode (p-mode) eigenfrequencies.
- i — i — i — | — . — i — . — | — . — i — i —
0.1
Sound velocities
O.OB
Bahcall-Pinaonneault 98 L0WL1 + BISON Measurement
0.06 c 3
0.04 h 0.02
a 3 01
J_
"5 •a -0.02 o
-0.04 h -0.06 -0.08
•-Tte lowered (la off Ga)
-0.1 0.2
0.4
0.6
0.8
R/R.
Figure 4: Predicted versus Measured Sound Speeds. This figure shows the excellent agreement between the calculated (solar model BP98, Model) and the measured (Sun) sound speeds, a fractional difference of 0.001 rms for all speeds measured between 0.05H© and 0.95i?Q. The vertical scale is chosen so as to emphasize that the fractional error is much smaller than generic changes in the model, 0.09, that might significantly affect the solar neutrino predictions. The measured sound speeds are from S. Basu et al. 4 ; the figure is taken fromBP983.
Could the solar model calculations be wrong by enough to explain the discrepancies between predictions and measurements shown in Fig. 3? Helioseismology, which confirms predictions of the standard solar model to high precision, suggests that the answer is "No." Figure 4 shows the fractional differences between the most accurate available sound speeds measured by helioseismology and sound speeds calculated with our best solar model (with no free parameters). The horizontal line corresponds to the hypothetical case in which the model predictions exactly match the observed values. The rms fractional difference between the calculated
62
and the measured sound speeds is 1.1 x 1 0 - 3 for the entire region over which the sound speeds are measured, 0.05.R© < R < 0.95# Q . In the solar core, O.O5i?0 < R < O.25i?0 (in which about 95% of the solar energy and neutrino flux is produced in a standard model), the rms fractional difference between measured and calculated sound speeds is 0.7 x 1 0 - 3 . The arrow in Fig. 4 shows how different the solar model sound speeds would have to be from the observed sound speeds if one wanted to use solar physics to reduce the 7 Be neutrino flux. The position of the arrow is fixed by artificially reducing the predicted 7 Be neutrino flux that is not observed in the gallium experiments, SAGE and GALLEX plus GNO (see Fig. 3), if the p-p neutrinos are present. Remember, we believe we can calculate the p — p flux to ± 1 % . The discrepancy with the hypothetical new solar physics was estimated by using the temperature dependence of the 7 Be neutrino flux (tx T 1 0 ) and the sound speeds (ex T 1 / 2 ). The agreement with the hypothetical solar physics is more than 100 times worse than the agreement with the Standard Model physics. Figure 4 has contributed to the consensus view that the experimental results shown in Fig. 3 require new particle physics for their explanation. 2.3
Summing up and looking ahead
I want now to look back and then look ahead. I will begin by giving my view of the principal accomplishments of solar neutrino research to date. Then I will discuss two of the expected highlights of the next decade of solar neutrino research, the measurement of the neutral current to charge current ratio for 8 B neutrinos and the detection of solar neutrinos with energies less than 1 MeV. Principal achievements What are the principal achievements of the first four decades of solar neutrino research? I give below my personal list of the 'top three achievements.' • Solar neutrinos have been detected. The chlorine, Kamiokande, Super-Kamiokande, GALLEX, SAGE, GNO, and SNO experiments have all measured solar neutrino events. This is the most important achievement. The detection of solar neutrinos shows empirically that the sun shines by the fusion of light elements. • Evidence for new physics has been found. For more than thirty years, beginning with the fact that Ray's first measurements in 1968 indicated a flux lower than the standard model predictions, we have had evidence for new physics in the solar neutrino arena. This evidence has steadily deepened as
63
new solar neutrino experiments have confirmed and extended the neutrino discrepancies and helioseismology has confirmed the standard solar model. The fact that neutrino oscillations have now been observed in atmospheric neutrino phenomena further strengthens the case that oscillations occur for solar neutrinos. We are still looking for a 'smoking gun' single effect that shows up in just one solar neutrino experiment, rather than combining the results of two or more different experiments. I will discuss some possibilities below. • Neutrino fluxes and energy spectra are approximately as predicted by the standard solar model. If you had told me in 1964 that six solar neutrino experiments would give results within a factor of three of the predicted standard model results, I would have been astonished and delighted. This is especially so considering that the crucial 8 B neutrino flux depends upon the 25th power of the central temperature of the sun. This agreement exists without making any corrections for neutrino oscillations. If we correct the observed solar neutrino event rates for the effects of neutrino oscillations using the six currently allowed two-neutrino oscillation scenarios, the inferred 8 B neutrino flux at the source is rather close to the best-estimate predicted flux. At the 99% CL, one infers (see hep-ph/9911248): 0.55 <
(1)
which is a slightly tighter range than the 3er prediction of the standard solar model. SNO and the [NC]/[CC] ratio Figure 5 shows the predictions of the currently allowed neutrino oscillation solutions for the double ratio, [NC]/[CC], of neutral current to charged current event rates in the deuterium detector SNO. Art McDonald describes at Neutrino 2000 the experimental characteristics of this great observatory and outline for us the extensive program of SNO measurements. The important message of Fig. 5 is that all of the currently allowed oscillation solutions for active neutrinos predict a value for the double ratio that is different from the no oscillation value of 1.0 by at least nine times the estimated non-statistical measurement uncertainty. We all eagerly look forward to this crucial and decisive measurement. Solar neutrinos below 1 MeV More than 98% of the calculated standard model solar neutrino flux lies below 1 MeV. The rare 8 B neutrino flux is the only solar neutrino source for which
64 7
[NC]/[CC] Currently Allowed 99% C.L.
6.5 E-
5.5 5F 4.5 " i—i
O
LMA VAC, SMA
4 *
LOW
2,3.5 3 2.5
VAC, MSW
1.5 1 0.5
Sterile
Measure 3ff
No oscillation—"3SB"" Neutrino Scenario
Figure 5: The neutral current to charged current double ratio, [NC]/[CC] . The standard model value for [NC]/[CC] is 1.0. The figure shows, for a 5 MeV threshold for the CC measurement, the predicted double ratio of Neutral Current to Charged Current for different neutrino scenarios. The solid error bars represent the 99% C.L. for the allowed regions of the six currently favored neutrino oscillation solutions. The dashed error bar labeled "Measure 3
measurements of the energy have been made, but 8 B neutrinos constitute a fraction of less than 1 0 - 4 of the total solar neutrino flux. The great challenge of solar neutrino astronomy is to measure neutrino fluxes below 1 MeV. We must develop experiments that will measure the 7 Be neutrinos (energy of 0.86 MeV) and the fundamental p-p neutrinos (< 0.43 MeV). A number of promising possibilities are discussed at the LowNu workshop that precedes Neutrino 2000. The BOREXINO observatory, which can detect v — e scattering, is the only approved solar neutrino experiment which can measure energies less than 1 MeV.
65
The p-p neutrinos are overwhelmingly the most abundant source of solar neutrinos, carrying about 91% of the total flux according to the standard solar model. The 7 Be neutrinos constitute about 7% of the total standard model flux. We want to test and to understand neutrino oscillations with high precision using solar neutrino sources. To do so, we have to measure the neutrinoelectron scattering rate with 7 Be neutrinos, as will be done with the BOREXINO experiment, and also the CC (neutrino-absorption) rate with 7 Be neutrinos (no approved experiment). With a neutrino line as provided by r Be electron-capture in the sun, unique and unambiguous tests of neutrino oscillation models can be carried out if one measures both the charged-current and the neutral current reaction rates. I believe that we have calculated the flux of p-p neutrinos produced in the sun to an accuracy of ± 1 % . This belief should be tested experimentally. Unfortunately, we do not yet have a direct measurement of this flux. The gallium experiments only tell us the rate of capture of all neutrinos with energies above 0.23 MeV. The most urgent need for solar neutrino research is to develop a practical experiment to measure directly the p-p neutrino flux and the energy spectrum of electrons produced by weak interactions with p-p neutrinos. Such an experiment can be used to test the precise and fundamental standard solar model prediction of the p-p neutrino flux. Moreover, the currently favored neutrino oscillation solutions all predict a strong influence of oscillations on the lowenergy flux of ve. Figure 6 shows the calculated neutrino survival probability as a function of energy for three global best-fit MSW oscillation solutions. You can see directly from this figure why we need accurate measurements for the p-p and 7 Be neutrinos. The currently favored solutions exhibit their most characteristic and strongly energy dependent features below 1 MeV. Naturally, all of the solutions give similar predictions in the energy region, ~ 7 MeV, where the Kamiokande and Super-Kamiokande data are best. The survival probability shows a strong change with energy below 1 MeV for all the solutions, whereas in the region above 5 MeV (accessible to Super-Kamiokande and to SNO) the energy dependence of the survival probability is at best modest. The p-p neutrinos are the gold ring of solar neutrino astronomy. Their measurement will constitute a simultaneous and critical test of stellar evolution theory and of neutrino oscillation solutions.
66
Energy (MeV)
Figure 6: Survival probabilities for MSW solutions. The figure presents the yearly-averaged survival probabilities for an electron neutrino that is created in the sun to remain an electron neutrino upon arrival at the Super-Kamiokande detector.
3
Extragalactic neutrinos
Experimentalists often like to describe the power of their experiments in terms of the expected or observed number of events per year and L/E, where L is the distance between the accelerator and the detector and E is the beam energy. The quantity L/E determines, together with the square of the mass difference, the survival probability for vacuum neutrino oscillations. More generally, L/E represents the time of flight in the rest frame of the particle, the time for rare events to occur. Figure 7 shows the extraordinary sensitivity to neutrino oscillation of experiments like ANTARES, BAIKAL, ICECUBE, and NESTOR that can detect neutrinos from distant extragalactic sources. The accelerator experiments
67
lUlniBooNE
10 •
* NOMAD 10» r • CHORUS
10 »
•MINOS PH2
BOREMNO*
• SuperK •«?«*•»• (atmos.) •CHOOZ •KARMEN/LSND • KamLAND
!
•K2K 10"
ANTARESI BAIKAL ICECUBE NESTOR
10'
10-'
10«
10»
10*
10« 10» 10M L/E (km/GeV)
•
10"
10"
10»
L_
10'"
Figure 7: Very longbaseline neutrino oscillation experiments. The figure shows that experiments such as ANTARES, BAIKAL, ICECUBE, and NESTOR, which may detect high-energy neutrinos from distant gamma-ray bursts, have extraordinary sensitivity to vacuum neutrino oscillations. Neutrinos of 10s GeV from gamma-ray bursts located at cosmological distances were used to locate the positions of ANTARES, BAIKAL, ICECUBE, and NESTOR in the figure. that will be discussed at Neutrino 2000 lie in the left-hand side of Fig. 7, L/E < 104 km/GeV. Solar neutrino experiments like Super-Kamiokande, SNO, and BOREXINO can reach to 1010 km/GeV and, for the lower energy experiments, even 1011 km/GeV. Extragalactic sources such as gamma-raybursts (GRBs) have such a long baseline (~ 1010 lyrs) that the new generation of extragalactic experiments, ANTARES, BAIKAL, ICECUBE, and NESTOR will extend to the right-hand side of Fig. 7, to L/E > 1018 km/GeV. I want to say a few words about the possibilities for detecting GRB neutrinos, which is discussed in more detail at Neutrino 2000 by Eli Waxman. I believe that GRBs offer the best chance for detecting extragalactic neutrinos among all the known sources of astronomical photons. The phenomenology of the photons observed from gamma-ray bursts is
68
Figure 8: The Waxman-Bahcall upper bound on muon neutrino intensities (i/^ + 0^). This figure is from Bahcall and W a x m a n 6 . The numerical value of the bound assumes that 100% of the energy of protons is lost to n+ and n° and that the ir+ all decay to muons that also produce neutrinos. The dot-dash line gives the upper bound corrected for neutrino energy loss due to redshift and for the maximum known evolution (QSO or star-formation evolution). The lower line is obtained assuming no evolution. The solid curves show the predictions of representative AGN jet models taken from the earlier papers of Mannheim (marked M95B in the figure), Protheroe (P97), and Halzen and Zas (HZ97). The AGN models were normalized so that the calculated gamma-ray flux from ir° decay fits the observed gamma-ray background.
now relatively well understood. Many different types of observations have been carried out and the results are well summarized by the expanding fireball model. Using this model, one can work out the flux of neutrinos from shocks. Figure 8 shows the neutrino energy spectra that Waxman and I have estimated to be produced by GRBs, both from the direct burst (energies ~ 106 GeV) and from the afterglow (energies ~ 108 GeV to ~ 1019 GeV). The observed population of GRBs should give rise to ~ 10 events per km2 per year from neutrinos with characteristic energies of order 1014 eV. We shall hear at
69
Neutrino 2000 that the calculated GRB flux may be detectable in ANTARES, ICECUBE, or NESTOR. The fundamental assumption used in calculating the GRB neutrino flux is that GRBs produce the observed flux of high-energy cosmic rays, an assumption for which Eli Waxman has provided a strong plausibility argument. GRBs occur at modest to large redshifts. We know the time of the explosion to an accuracy ~ 10 sec (from the gamma rays). Therefore, GRBs can be used to test special relativity to an accuracy of 1 part in 1016 and to test the weak equivalence principle to an accuracy of 1 part in 106. If special relativity is right, the photons and the neutrinos should arrive at the same time (to an accuracy of about 10 sec, the duration of the burst). If the weak equivalence principle is valid, the arrival times of neutrinos (which traverse significant gravitational potentials) from distant sources should be independent of neutrino flavor. GRBs can also be used to probe the weak interactions to an extraordinary level of precision. Gamma-ray bursts are expected to produce only ve and v^. The large area detectors of extragalactic neutrinos are in principle sensitive to vacuum neutrino oscillations with mass differences as small as Am 2 > 10- 1 7 eV2 (from v„ - • vT). Not everything is encouraging in Fig. 8. The figure also shows the upper limit that is allowed for astrophysical neutrino production from (7,7r) interactions on high energy protons. The upper limit is established by using the observed cosmic ray flux of high energy protons. Prior to the recognition of this limit a number of authors had suggested much more optimistic models (also shown in the figure), that were normalized by fitting 7T° decay to the observed gamma-ray background. 4
Goals for Astrophysical neutrinos: 2000-2010
It seems to me that we have three principal goals for this next decade. • Determine the mixing angles and mass differences that are important for solar neutrino phenomena. • Test precisely stellar evolution by observing p — p and 7 Be neutrinos, and by determining the total flux of 8 B neutrinos. • Discover extragalactic neutrinos, perhaps from gamma-ray bursts. From time to time, friends ask me to compare the search for solar neutrinos with the search for neutrinos from GRBs. They are very different. From
70
photon studies, we know more observationally about the sun than about any other astronomical source, certainly much more than about the mysterious GRBs. Moreover the sun is in the simplest stage of stellar evolution, in quasistatic equilibrium with a characteristic time scale for evolution of 109 yr (10 16 s ). We do not even know the energy source of GRBs. We do know that GRBs are far from equilibrium, evolving explosively on a time scale of order 10- 3 s. We want to do extragalactic neutrino astronomy because it is truly an exploration of the universe. We do solar neutrino astronomy to test fundamental theories of physics and astronomy. But, perhaps solar neutrino research and extragalactic neutrino research may in the end share a fundamental characteristic: surprise. Remember, that we undertook solar neutrino research to test stellar evolution and unexpectedly (at least for everybody except Bruno Pontecorvo) we found evidence for new neutrino physics. In a sense, we are returning to our original goal in neutrino astronomy, but by a round-about path. We must first understand neutrino oscillation phenomena in order to be able to use solar neutrino observations to test precisely the theory of stellar evolution, our original goal. Perhaps with extragalactic astronomy we will participate in a similar cycle of astronomical exploration and physical clarification. T. S. Elliot in 'The Four Quartets' described the cycle succinctly and beautifully: We shall not cease from exploration And the end of all of our exploring Will be to arrive where we started And know the place for the first time. Acknowledgments I acknowledge support from NSF grant #PHY95-13835. References 1. B. Pontecorvo, Zh. Exp. Teor. Fiz. 53, 1717 (1967). 2. J. N. Bahcall, Phys. Rev. Lett. 12, 300 (1964). 3. J. N. Bahcall, S. Basu and M. H. Pinsonneault, Phys. Lett. B 433, 1 (1998). 4. S. Basu et al., Mon. Not. R. Astron. Soc. 292, 234 (1997). 5. J. N. Bahcall, P. I. Krastev and A. Yu. Smirnov, hep-ph/0002293. 6. J. N. Bahcall and E. Waxman, hep-ph/9902383.
Single Atom Extraction and Classification with a Hybrid Solar Neutrino Detector K. Lande and P. Wildenhain, University of Pennsylvania, R. Corey and M. Foygel, South Dakota School of Mines and Technology, J. Distel, Los Alamos National Laboratory, We describe an electronically triggeredradiochemicaldetector that can separate the interactions due 8B electron neutrinos from the Sun from those due to 7Be and can distinguish these events from those due to cosmicraysor localradioactivity.In this detector, a set of photomultiplier tubes detects the Cerenkov radiation generated by neutrino interaction secondaries and initiates a fast extraction of the secondary atom from the detector and directs this atom to a specific charcoal storage trap.
71
72
When, in 1965, Raymond Davis began his search for neutrinos produced by nuclear fusion reactions in the solar core, he hardly imagined where that search would lead. His goal was to demonstrate that the Sun's thermal emission was due to the fusion of hydrogen into helium by observing the neutrinos emitted in that reaction, four neutrinos per 26 MeV of released energy. Davis indeed detected these neutrinos, except that he only observed about 1/3 of the predicted flux1. Dire thoughts immediately developed; was the detector very inefficient or was there a basic flaw in the solar model? The simplest issue to attack was the behavior of the detector. The detection reaction, 37Cl(ve, e")37Ar results in a non-interacting gas atom. Davis periodically swept the produced atoms out of the detector tank. In order to measure the efficiency of that sweep, he added a known amount of isotropically labeled argon, alternately 36Ar and ^Ar, to the detector. The extraction efficiency of the 37Ar should be the same as that of the ^Ar or 38Ar "carrier" gas. The recovery yield of the carrier gas was above 96% and was measured separately for each run. That was clearly not the problem. The experimental issue was completely laid to rest when the Kamiokande, Superkamiokande and both gallium experiments, GALLEX and SAGE, all observed solar neutrino fluxes that were considerably lower than predicted. In the mid-1960's Pontecorvo2 suggested that ve from the Sun might convert to ve or v^ in the transit from the Sun to the Earth in a manner similar to the vacuum K° -> K° transition observed a decade earlier. This suggestion opened the possibility of a time dependent change in neutrino flavor. The idea was expanded in 1986 by Mikheev and Smirnov3 who extended the Pontecorvo neutrino flavor transition idea to involve transitions that might involve a matter transition (MSW) inside the Sun. Such transitions require mass differences between the two flavors of neutrinos involved and so introduce a neutrino mass scale. The combination of mass difference and coupling strength can be used to generate an energy dependence of the neutrino flavor mixing probability that can generate an energy dependent reduction of the observed neutrino solar neutrino flux. Additional evidence of unusual neutrino behavior has been observed with neutrinos produced in the upper atmosphere by cosmic rays. These observations indicate that there is a deficiency of muonic neutrinos coming upward that have traveled through most of the Earth's diameter compared to those coming downward which have a relatively short flight path from production to detection4. Electron neutrinos do not show such an effect; their flux appears independent of flight path over the available range. The interpretation of these observations is that vK convert into another neutrino flavor, but not into veand so either sterile neutrinos or vT.
73
In summary, we have two different disappearance observations that suggest neutrino flavor transitions, v(t->vt for atmospheric neutrinos and Ve-^v^ for solar neutrinos. Unfortunately, all the present observations involve the disappearance of neutrinos. There still is no observation of the appearance of the transformed flavor neutrino. Appearance of the transformed flavor neutrinos is critical to the demonstration that neutrino flavor transitions occur and so is the goal of the next generation of neutrino observations. Since the energy range of atmospheric neutrinos extends into the multi-GeV range, some fraction of these neutrinos are energetic enough to produce x mesons. Indeed, if the neutrino flavor transition hypothesis is correct, then the Superkamiokande detector already has about 100 x produced within it. The question is, can these be reconstructed and recognized? Unfortunately, the solar neutrino spectrum only extends to 14 MeV so that there is no possibility of observing neutrino interactions that can make charged muons. The only recourse is to utilize the difference in measured solar neutrino flux between a detection mode that is sensitive to all neutrino flavors and one that is sensitive to only ve. This must be done separately for 8B neutrinos and for 7Be neutrinos. The neutrino-electron elastic scattering reactions observed by Superkamiokande are the sum of charged current ve - e scattering and neutral current v^ - electron scattering. By subtracting the charged current ve flux from the total Superkamiokande measurement, we will have the flux of non-electron neutrinos from 8B in the Sun. The 8B electron neutrino flux can be obtained from inverse beta decay reactions on nuclei such as 37Cl(ve,e")37Ar as observed by the Homestake chlorine detector, or 2H(ve,e~)pp, as observed by the charged current reactions at SNO. If the Superkamiokande neutrino-electron scattering events are interpreted as all due to electron neutrinos from 8 B, men that flux is 2.44 + 0.09 x 106 Ve/cm2 sec5. To compare this result with that from the chlorine detector we must multiply by the cross section for 8B electron neutrinos in 37C1, 1.1410.04 x 10"42 cm2.6 This gives a chlorine equivalent rate assuming that the SK measured flux is all electron neutrinos of 2.78 ± 0.13 SNU. This result is quite consistent with the direct chlorine detector measurement of the solar electron neutrino flux of 2.56 ± 0.23 SNU7. The conventional interpretation of this result is that the above two measurements are of the same signal, namely all electron neutrinos from 8 B. This implies that the chlorine detector does not see any electron neutrinos from 7Be. Of course, 7Be must exist in the Sun since 8B is formed by P + 7Be -> 8B +y . Since only about
74
1/1000 7Be is involved in 8B production, the 7Be decay rate should be about 103 times the 8 B formation and decay rate. This solution, therefore, strongly suggests that essentially all 7Be electron neutrinos convert to muon neutrinos on their way from the solar core to the Earth. There is, however, another possible interpretation of the SK and chlorine data, that each of these signals contains anotiier component. For SK, that component is nonelectron neutrinos from 8B while for chlorine, that component is low energy electron neutrinos such as those from 7Be. Since there is no way to distinguish vM interactions from ve interaction in the SK detector, it is necessary to add to the information obtained for the chlorine signal, the energy distribution of the electron neutrinos that interact. 2. The Hybrid Detector - a photomultiplier triggered radiochemical detector We have proposed doing the above by converting the chlorine detector from a purely radiochemical detector to a hybrid electronic-radiochemical detector. This would be achieved by adding a set of photomultiplier tubes to the detector. These tubes would be used to detect the Cerenkov signal produced by the secondary electron from the 37Cl(ve,e")37Ar reaction. Whenever such a signal is observed, the detector would be swept with a pulse of helium and the recovered3? Ar atom would be stored in a specific charcoal trap dedicated to 8B events. There are several technical details that must be dealt with in designing such a detector. First, the extraction time must be short compared to the rate at which 37Ar atoms are produced. Second, the rate of false triggers, events in which there is a Cerenkov pulse but no production of 37Ar atoms, must also be low compared to the extraction time to avoid excessive detector dead time. Third, we must be able to distinguish between the various channels for 37Ar production, cosmic ray muons, 8B neutrinos, 7Be neutrinos and neutrons of local origin. The first two conditions can be achieved by dividing the detector into a number of separate modules, each with a separate set of triggering photomultipher tubes and a separate extraction system. Any given pm signal initiates an extraction in only one module. The remaining modules remain active. In addition, smaller modules also permit shorter extraction time constants, enhancing the effect. 3. Background rates A cosmic ray induced 37Ar production will generally consist of a very high energy through going muon, ~2 TeV, producing one or more nucleons in or near a given module followed by one of the secondary protons undergoing a 37Cl(p,n)37Ar reaction. The neutron that emerges from this reaction will moderate in the module
75
and finally capture on a 35C1 producing a 36C1 and about 7 MeV of gammas. The initial muon will result in a very large pulse. The delayed neutron capture gammas will then generate a second pulse within a millisecond of the first pulse. This signature is quite distinctive. Our estimates for cosmic ray induced 37Ar production rate at the depth of the present chlorine detector, 4200 mwe, is about 1/10 of the production rate due to neutrinos. We anticipate that at least 90% of cosmic ray muon induced 37Ar production will result in such a two pulse signal. This leaves us with a potential cosmic rav contribution to the one pulse events that is less than 1% of the neutrino induced 3 Ar rate. The local neutron induced rate is somewhat more difficult to estimate. The mechanism consists of a (n,p) process to produce a fast proton, followed by the same 37Cl(p,n)37Ar and delayed neutron capture as with the cosmic rays. Since there are no free protons in C2C14, the (n,p) reactions involve one of: 12C(n,p)12B, 35 Cl(n,p)35S or 37Cl(n,p)37S. Since 12B has a 20 ms half life and then decays by beta emission, we see a two pulse delayed pattern, only with a longer delay time. The first pulse will be due to neutron capture on 35C1 with the second pulse arising from the 12B decay. However, the high threshold, 13.4 MeV, makes this a very unlikely channel for this background reaction. Unfortunately, 35S has a 87 day half life and and beta decays with a very low energy end point so we will not see that second pulse. The low threshold for this reaction, 170 keV, suggests that this reaction could be the dominant channel for neutron induced background in C2CI4. There is a chance that some fraction of the 35S will go to an excited state and produce some deexcitation gammas that will provide a prompt, but small first pulse. Although these events will have the same time structure as the cosmic ray events, the first pulses will be very different for these two cases. The threshold for the production of 37S is higher, 4.85 MeV. Combining this threshold with the requirement that the outgoing proton have at more than 2 MeV limits this channel to neutrons above 7 MeV, a very small fraction of the neutrons that are generated by radioactivity in the local rock. The 7Be neutrino produced 37Ar will involve a 48 keV secondary electron and so will not produce any Cerenkov radiation. Since there is no Cerenkov signal for this interaction, we will periodically sweep each module, possibly every 3 hours or so, and store any extracted 37Ar in a "7Be" labeled trap. A similar analysis of neutron induced backgrounds in Nal dissolved in water leads to a much better recognition scenario. There are four target atoms of interest, 'H, O, Naand 127 I. The n,p elastic scattering reaction always results in a secondary neutron. A second such neutron is emitted from the (p,n) reaction with 127I. Thus,
76
we are guaranteed to have two neutrons and thus two neutron capture pulses that provide a clear signature of background from this reaction. This is the dominant and only significant neutron induced background reaction in the Nal solution. l6
0(n,p) 16 N has a high threshold, 10.4 MeV, so this is an unlikely reaction for neutrons from the local rock. In addition, 16N has a reasonably short life time, 7 seconds, so that a second pulse should be quite recognizable. 23Na(n,p)23Ne has a 4.4 MeV threshold. The combination of the Coulomb barrier of 127I, about 10 MeV and this reaction threshold sets the effective neutron energy threshold at 14- 15 MeV. In addition, the life time of 23Ne, 37.6 seconds, makes this a good candidate for two pulse recognition. The threshold for 127I(n,p)127Te is very low, 700 keV, so that this channel has significant possibilities for background production. However, the ratio of *H to 127I in our solution is 16/1, so that the proton producing reactions with 127I are a very small fraction of the background reactions in the iodine detector. In summary, the high Coulomb barrier greatly suppresses neutron induced background in Nal and those that do occur have a very high probability of having a clearly recognizable two pulse pattern that can be used to label these events. Since unlabeled neutron induced events might be included in the 7Be induced sample, from this point of view, Nal is a much more desirable target material than is C2C14. 4. "False" triggers The most difficult and annoying problem will be "false" triggers, namely those that are initiated by relativistic electrons that results from Compton scattering, panproduction, beta decay of radioactive contaminants or gammas from nuclear deexcitation that follows neutron capture. These occurrences do not result in 37Ar production, but do produce a Cerenkov pulse that may be difficult to distinguish from 8B neutrino induced events. Fortunately, the 8B neutrino interactions all exhibit large energies in the detector permitting a high solar neutrino event threshold. The neutrino cross section on 37C1 as a function of energy, as given by Bahcall, et. al., ref. 6, multiplied by the 8B neutrino spectrum gives an event spectrum that is strongly peaked at 10.5 MeV with only 10% of the events below 8 MeV. We obtain an estimate of the nuclear deexcitation signal following thermal neutron capture from the measurement of this signal in the Gd loaded scintillator of the Palo Verde Neutrino Oscillation Detector.8 This gives a scintillation spectrum that peaks at 6 MeV and ends at about 8 MeV. Thus, a cut at about 7 - 8 MeV would effectively separate the 8B solar neutrino interactions from the neutron induced false triggers.
77
We have used the SK total signal rate as a function of energy to get an estimate of the false trigger rate. For 25 ton modules, we extrapolated a false trigger rate of one every three hours for a 6 MeV Cerenkov threshold and a lower trigger rate for a threshold at 7 - 8 MeV. Our detector will have a lower radon content than SK since the three hour 7Be periodic sweep will also remove radon. Our shielding against external radiation will be poorer than that of SK. The above arguments suggest that most false triggers can be energetically separated from neutrino interactions. However, since it is difficult to generate reliable estimates of false trigger rates, we will set up a test module in the mine and measure this rate directly. 5. Extraction system The basic argon atom extraction technique employed in the Homestake chlorine detector is to use a flow of small helium bubbles to remove the argon atoms from the detector liquid. In that detector the bubbles were produced by use of a series of eductors, jet nozzles that employ a Bernoulli tube to pull helium gas into a stream of circulating detector liquid. The driving mechanism is a liquid circulation pump. This system works extremely well for a single large tank. Although it could also be used by a multi-module system, it would require a liquid pump per module, a fairly expensive and complex system. We have opted for a different bubble generating system, one that has been developed in recent years for the oxygenation of waste water systems. It consists of a series of flexible membranes, each of which has a large number of bubble generating openings. Pressurized helium gas is delivered to the back side of each membrane. The gas pressure forces open each of the holes emitting a stream of bubbles. We have chosen to use 1 mm diameter bubbles and set a bubble density of 10 per cm2. Thus, for a vertical cylinder module with a 1.5 meter diameter, there will be over 105 bubble emitting holes in the membrane surface on the bottom of that module. A 1 mm bubble has a rise velocity of about 20 cm/sec or takes about 40 seconds to rise to the top of an 8 m high module. Our preliminary measurements indicate that the 1/e extraction time of this system at a given horizontal slice of such a module is 15 seconds or a 99% extraction efficiency in that slice in one minute. Combining the extraction efficiency in a horizontal slice with the bubble rise time suggests a 2 minute total extraction time in such a module. These measurements are in good agreement with the results of extraction modeling by M. Foygel9. It is important to stress that our tests were done on a 2 meter high module and that we have not yet tested an 8 meter high module.
78
Each module is connected to two gas manifolds, one that brings the helium gas to the module system and a second that carries the helium gas from the modules to the charcoal traps. Two electrically controlled valves connect each module to these two gas manifolds. A number of charcoal traps are also connected to the gas manifold system. Each of these charcoal traps is also connected these two gas manifolds via two electrically operated valves. The helium gas is driven through the system by a diaphragm pump that is located in the gas return system between the exit of the charcoal traps and the input to the detector modules. A typical event sequence will begin with a neutrino interaction in a given module that results in a Cerenkov pulse detected by the photomultiplier tubes in that module. The pulse size and pulse pattern will classify the event and determine the charcoal trap that will be used. Four electrical valves are then opened by the control computer, two of these are on die gas input and output of the module in which the event occurred. The other two are on the input and output of the charcoal trap where events of this type are stored. The diaphragm pump driven helium flows through the system for a predetermined extraction time, about 2 minutes. If an event occurs in a second module during the extraction period of the first module, the control computer stores this information and then begins die second module extraction cycle as soon as thefirstone is completed. Thus, the system can tolerate a false trigger rate of one event every two minutes or 30/hour with a detector inefficiency of 1/N, where N is the number of modules in the system. For example, a 50 module system with a false trigger in each module every 1.6 hours, would be able to operate with a 98% detection efficiency. 6. System safety All module intrusions, namely the input and output gas lines and the photomultiplier cables, will be from the top. A valve or piping failure could result in air entering the module, but will not result in any loss or spill of detector liquid. The helium pump will not be able to operate until all the appropriate valves are open and all other valves are closed. Since this is a completely closed loop system, the total amount of helium gas in die system is well defined. Each module will be pressure rated so diat it can hold the entire detector gas volume without being overpressured. Each module will have an electrical pressure gauge so that overpressure can be immediately detected and stop the helium pumps. As additional pressure relief protection, each module exit valve will have a bypass pressure relief valve that connects to the return gas manifold and a second, higher set pressure valve that will connect to the outside atmosphere.
79 We do not anticipate a significant rate of valve failures. The only gas system components that are expected to require periodic maintenance are the pump diaphragms. We plan to use pumps will have double diaphragms so that the failure of one diaphragm will not allow outside air to enter the system. In addition, we plan to have a standby parallel pump in the system so that we can periodically interchange operating pumps and replace the diaphragms on the out of service pump. The indications from the manufacturer are that diaphragms should have multi-year life times in our mode of operation. Since all valves will be normally (electrical power off) closed, a power failure will not have any effect on the gas and module isolation system. The heat capacity of the charcoal traps will keep them cold as long as the failure is fairly brief. A long duration power failure may result in a warm up of these traps. Since the input and output valves of each trap are closed, there should not be any loss of trapped signal atoms. Each trap will have a pressure relief valve which will open if the gas pressure in that trap becomes excessive. The process control computer will have a standby battery power supply so that it will remain operational during a power failure. 7. Photomultiplier trigger system A set of photomultiplier tubes will be inserted into each module to detect the Cerenkov light pulses. Since our detector will not reconstruct the direction of the electron, there is no interest in the direction of the initial Cerenkov pulse nor in the precise time of light pulse arrival at each pm tube. We will add wavelength shifter to the detector liquid. This will provide isotropic emission of light and will increase the light in the blue region where the pm tubes are most sensitive by a factor of 3 4. We also plan to provide a highly reflective surface on the inside of each module so that the light that is not absorbed by a pm photocathode will be reflected and be available for a subsequent pm detection. The combined result is a total enhancement of detectable light by a factor of 20 - 25 compared to SK and SNO type detectors. This will permit a significant light collection with a fairly small number of pm tubes in each module. 8. Supernova neutrino burst detection Supernova are quite rare occurrences whose internal dynamics can be very effectively probed by investigating the flavor, energy and time structure of the emitted neutrinos. The extremely high core density is sufficient to permit MS W flavor transitions from vT-»ve. The high energy ve made in the collapsed core will find the local neutron sphere opaque. This will result in a fairly low energy at which ve can get through this obstacle. Neutrinos of other flavors will not be
80
effected by the neutron sphere. vt->-ve flavor transitions that occur outside this neutron sphere will recreate high energy ve. The measurement of the detected ve flux and energy distribution will test for this occurrence and may permit a measurement of the vx mass. The supernova signal will consist of a number of interactions within a few seconds. Some fraction of these pulses will be greater than the upper limit of the solar neutrino spectrum. The pattern is quite distinctive and will not be confused for a conventional cosmic ray interaction or solar neutrino signal. The rate of events within a few seconds will not permit extraction of individual atoms. Instead, a total detector sweep will be carried out after such an event with the argon atoms stored in a standby charcoal trap. We intend to participate in the International Supernova Watch and keep a UTC correlated clock readout on our control computer record. 9. Detector target, 37C1 or 127I? Although the above discussion is based on the use of perchloroethylene, C2C14, as the detector fill, we actually have a choice of two different targets, 37C1 in C2C14 and ,27 10 I in Nal dissolved in water. Both of these liquids are transparent to visible light, and have the same index of refraction, 1.5 and density, 1.6. Thus, they would be interchangeable in the proposed modules. There are, however, a number of differences in background sensitivities and in neutrino cross sections. The dominant background sensitivity difference between 37C1 and 127I is due to the difference in nuclear charge, Z=17 and Z=53, respectively. This difference in nuclear charge is reflected in the effective Coulomb barrier of each of these nuclei. For CI, the effective threshold for the protons in 37Cl(p,n)37 Ar is about 3 MeV while the equivalent threshold for 127I(p,n)127Xe is about 10 MeV. This difference is quite significant since neutrons from radioactivity in the local mine rock are 8 MeV or less. These neutrons then must undergo a (n,p) reaction to produce the required proton. As an example of this background effect, the rate of cosmic ray induced 127Xe production in a Nal solution is about a factor 20 smaller than the corresponding rate of 37Ar production in C2C14. In addition to a lower background sensitivity, 127I also has a larger neutrino cross section by a factor of 4 - 5 than 37C1. There is, however, one distinctive advantage to 37C1, we presently have a better knowledge of the cross section. The cross section for 7Be neutrinos is determined by time reversal from the lifetime of the final state,37 Ar. The cross section for 8B neutrinos is determined from relationships among the various constituents of the
81 A=37 system. The presently stated cross section is 1.14 ± 0.03 x 10"42 cm2. The cross section for the supernova energy region is given as 1.2 x 10"40 cm2. For 127I, the cross sections are determined from (p,n) measurements. A neutrino cross section measurement was done in the 90° neutrino room at the beam stop at LAMPF. The precision of these measurements are in the 15 - 30% range. There is in place a program to measure these cross sections for 127I to high precision. The cross section for 7Be will be done with an intense (1-2 megacurie) source of 37Ar. The emitted monoenergetic neutrinos from this source are at 814 keV, which lies just midway between the 127I-»127Xe threshold of 789 keV and the energy of the 7Be neutrinos 862 keV. The 37Ar source is being prepared in a Russian fast neutron reactor, BN-600 by 40Ca(n, 4He)37Ar. The calibration is scheduled to be carried out in the fall of 2003. The high energy neutrino, 8B and supernova, calibration will be done with neutrinos from the beam target of the SNS. It is anticipated that this calibration will be carried out in 2006. The anticipated precision is 1 - 2%. We also hope to do a similar direct neutrino cross section calibration of 37C1 at the same facility. Given the relative merits of the two detectors, it may be sensible to utilize a mix of the two nuclei and fill half of the modules with C2C14 and the other half with Nal dissolved in water. That way we can get a good check on the effect of backgrounds as well as some redundancy on the ve fluxes from 7Be and 8 B. Given the centrality of these measurements in the search for the appearance of v^ from the Sun, redundancy may be very important. 10. Detector mass There is also the question of the appropriate detector mass. For a modular detector whose mass can be gradually increased, this question is not as important as it is for a single unit detector. The initial stages of our detector mass plan are based on a combination of understanding and verifying die detector operating characteristics and the background rates. Once the various preliminary tests have been successfully completed, we plan to construct and operate one full size module. This system already involves the gas manifold system, the control system, the diaphragm pumping system and at least one charcoal trap. When we are satisfied with the single module, we will expand to a 15 - 20 module system. A 20 module C2C14 system with 25 tons of detector mass per module is reasonably close to the mass of die present chlorine detector and so provides a good basis for comparing rates between the standard radiochemical and the hybrid system. Operation of this system will also assure us that there are no complications in operating a multimodule system.
82
Once we are satisfied that the hybrid detector system is operating properly and that the data can be understood and is in agreement with that of the pure radiochemical system, we will embark on a gradual expansion of the detector to a 50 -100 module system or a total detector mass of several kilotons. Another way to approach the question of final detector mass is to focus on the statistical precision that we wish to achieve in a given running period. For example, if we wish to get a 2 - 3% statistical result in three years of operation, 900 - 2500 counted events, we need between one and three counted events per day. With C2CL| we get 0.8 37 Ar produced per day per kiloton detector mass. This should give about 0.5 counted 37Ar per day per kiloton. A 100 module C2CI4 system will have a mass of 2.5 kilotons and thus give 1.2 counted events per day or about 1300 in a three year period, a 3% statistical result. The same detector volume filled half with C2C14 and half with Nal dissolved in water will result in about 650 37Ar and 3000 127Xe. The statistical precision of such a measurement will be in the sub 2% range. Of course, we have only addressed one of the systematic errors, that associated with the cross section. It is quite conceivable that the rest of the systematic errors will be the limiting factors in the overall precision of this measurement. We thank the National Science Foundation and the U.S. Department of Energy for support of this effort.
1
R. Davis, D. Harmer and K. Hoffman, Phys Rev Lett, 20, 1205 (1968) B. Pontecorvo, ZhETF, 53, 1717 (1967) and V. Gribov and B. Pontecorvo, Phys. Lett. 28B, 493 (1969) 3 S.P. Mikheev and A.Y. Smirnov, Nuovo Cimento, 9C, 17 (1986) 4 Y. Fukuda, et. al. Phys. Lett. B467. 185, (1999) 5 M.S.Smy (for the Superkamiokande collaboration) Division of Particles and Fields 1999 Conference Proceedings 6 J.N. Bahcall, E. Lisi, D.E. Alburger, L. de Braeckeleer, S.J. Freedman and J. Napolitano, Phys Rev C54, 411( 1996) 7 B.T.Cleveland, T. Daily, R. Davis, Jr., J.R. Distel, K. Lande, C.K. Lee, P.S. Wildenhain and J. Ullman, Ap. J. 496, 505 (1998) 8 F. Boehm, et.al., hep-ex/0003022, 16Mar2000 9 M. Foygel, et. al., to be published. 10 W. C. Haxton, Phys. Rev. Lett. 60, 768 (1987) 2
83
Ji
«:
1
J
•*~1 r
1
n
tv
4=: IO
* 2 4 - dia."
V
<®
d>
§>
o
@>
oo
CM
(O "
.,
Q> 5 ft. dia.
^
FIG. 1 The photomultiplier arrangement and access ports of a single detector module are shown. The photomultiplier tubes are housed in two plexiglas tubes that enter the module through flanges in the top of the module. These plexiglas tubes serve to isolate the photomultiplier tubes from the detector fluid and permit servicing of the pm tubes without opening the detector volume. The plexiglas tubes will be filled with mineral oil. Since the detector liquid, the plexiglas and the mineral oil have the same index of refraction, there will not be any light reflections at those interfaces. The center flange of the module is used to install the gas diffuser elements and provides an exit path for the flushing gas. All detector monitors such as pressure and depth gauges are also attached to this flange. There are no holes or access points in the bottom of the module to avoid any possible sources of leaks.
84
FIG 2 This diagram shows the gas manifold system with three detector modules on the left and three charcoal traps on the right. In this example a Cerenkov trigger with a 8B pulse pattern has just occurred in the middle module. The sweeping gas flows from the diaphragm pump on the bottom through the supply gas manifold to the module Inside the module, the gas flows into the gas diffusers on the bottom of the module bubbles through the detector hquid and then exits through the top. The gas then goes through the top gas manifold to the charcoal trap array. In this case, the sweeping gas passes through the center charcoal trap, the one that collects B neutrino interaction secondaries. The charcoal trap for supernova events was omitted to save space. Additional detector modules can be attached to the gas manifold ends on the left. Modules can be added without significantly disturbing the operating system.
THE SUDBURY NEUTRINO OBSERVATORY J.J. SIMPSON Department of Physics, University of Guelph, Guelph, Ontario NIG 2W1, Canada E-mail: simpsonjcbphysics. uosuelph. ca For the SNO Collaboration*
The Sudbury Neutrino Observatory (SNO) is a large, underground heavywater Cerenkov detector which has been designed and built primarily to solve the solar neutrino problem, the shortfall in the flux of neutrinos coming from the sun relative to the best solar model predictions. As discussed in previous talks in this symposium, the neutrino flux shortfall occurs in all previous experiments which were sensitive to different energy thresholds for solar neutrinos - the gallium experiments, the chlorine experiment, and the water Cerenkov experiments. And furthermore, the shortfall seems to be a result which is independent of physicallyplausible changes in the standard solar model, changes which are consistent with helioseismological results. Because it seems to be impossible to modify the standard solar model to account for all aspects of the neutrino flux shortfall, the explanation would seem to be connected to properties of neutrinos. The favored explanation, discussed previously in this symposium by Bahcall, is neutrino-flavour oscillations; this explanation has already been successfully invoked to explain the results on atmosphere muon neutrinos obtained by Superamiokande (see H. Sobel, this symposium). The unique feature of SNO, which should enable it to determine if electron neutrinos produced in the sun are oscillating into another flavour of neutrino, is the use of deuterium in the form of heavy water to detect the solar neutrinos. There are two dominant neutrino reactions with the deuteron, the charged-current (CC) reaction, ve + d -> p + p + e" (Q = -1.44 MeV), and the neutral current (NC) reaction: vx + d -» p + n + vx (Q = -2.22 MeV). For neutrino energies of about 5-14 MeV, typical for the 8B-neutrinos from the sun, the charged-current reaction cross-section is about 2.5 times that of the neutral current reaction. In addition, there is the elastic scattering (ES) reaction vx + e" -» vx + e" which involves both charged-current and neutral current interactions for ve but only neutral-current for v^ or v t . This reaction has a much lower cross-section than the CC and NC reactions on deuterium.
85
86
One of the main physics goals of this experiment is to measure the CCreaction rate relative to the NC-reaction rate which will give a measure of the veflux to the total v-flux. If the ve-flux turns out to be less than the total v-flux, this will be strong evidence for neutrino oscillations. A secondary goal is to measure the 8B v-energy spectrum. This is determined from the CC-reaction because the electron produced in the CC-reaction carries almost all of the incident neutrino energy. Should the ve oscillate into a sterile (4th) neutrino, the measurement of a distorted v-energy spectrum might be one of the few ways to determine it. We shall also look for time-dependence of the solar v-flux, such as between day and night (due to MSW effects in the earth) or seasonal effects arising from the earth' s varying distance to the sun. The SNO detector has been running in production mode since the end of October, 1999. An outline of the SNO detector follows, but more extensive details can be found in ref. 1. The SNO detector is located in the Creighton mine of INCO Ltd. near Sudbury, Ontario, at a depth of 2070 m. The somewhat barrel-shaped cavity containing the detector (see fig. 1) is about 22 m in diameter at mid-height and 34 m high. The 1000 tonnes of heavy water, which is about 99.91% pure D 2 0, is contained in an acrylic sphere about 12 m in diameter made of ultra-violet transmitting acrylic panels about 5.5 cm thick. The acrylic sphere hangs by ten Vectran ropes in 7000-tonnes of ultra-pure water. Surrounding the acrylic sphere is a geodesic structure about 18 m in diameter which supports approximately 9700, 20 cm Hamamatsu R1408 photomultiplier tubes. These tubes were made of speciallyproduced glass (Schott 8246) which has Th and U impurities at less than 40 ppb. The geodesic PMT structure and the acrylic vessel were both assembled underground, the latter, for example, consisting of 120 thermoformed acrylic panels bonded together. The PMTs are surrounded by light concentrators which increase the effective light collection efficiency by about 75%. The distance of the PMTs from the centre of the acrylic vessel is about 8.5 metres. A variety of calibration measurements are performed on the detector at various intervals. Electronic calibration is performed with built-in pulsers on a weekly basis. Phototube calibration is performed monthly with laser light fed into a diffuser ball lowered into the heavy water. Optical calibrations are also performed with the laser-diffuser ball at various wavelengths and in various positions inside the detector on approximately half-year intervals. Until now, the energy calibration has been performed with a 16N source providing a 6.13 MeV y-ray. Several other sources are in the final phase of commissioning including a 3H(p,y)3He accelerator source for a high-energy calibration point (near 20 MeV), a continuous 8Li P-source and some radioactive sources such as 252Cf (for neutrons) and tagged 228Th and 24Na sources for low-energy studies. The energy calibration at present indicates that there are about 8 to 9 hit phototubes per MeV at low energy (essentially single photon hits). The phototube timing resolution is about 1.7 ns. The hardware trigger threshold is near 2 MeV and has a rate of about 15 Hz.
87
At the present time, the detector is running in its pure D 2 0 phase. The main purpose of this period is to obtain the ve-flux from the CC reaction as well as the ve-energy spectrum, both above a threshold of about 7.5 MeV. The reason for this threshold, which is higher than that required by backgrounds, is that the NCreaction produces neutrons which can capture on the deuteron and produce a 6.25 MeV y-ray. However, this provides an opportunity to determine the total v-flux via the NC-reaction in the pure D 2 0 phase. If the radioactive backgrounds are at their design levels, then it is expected that the realistic threshold for event detection will be at about 5 MeV. Consequently the 6.25 MeV y-ray from neutron capture on the deuteron will be detectable. It turns out that the NC-events can be distinguished statistically from the charged-current events by the fact that these n-captures on the deuteron have a non-uniform radial distribution as a function of distance from the centre of the heavy water. This is caused by the long mean distance that a thermal neutron travels between creation and absorption in very pure D 2 0. The thermal diffusion length in 99.9% D 2 0 is about 120 cm. Consequently neutrons created in the outer volume of D 2 0 have a high probability of being captured in hydrogen, which has a high n-capture cross-section, in the acrylic wall or H 2 0. Fig. 2 shows a monte carlo simulation of the radial distribution for CC and NC events; the NC distribution can also be calculated quite accurately analytically from neutrondiffusion theory. Thus it should be possible to have a preliminary estimate of reasonable accuracy from the total v-flux during the pure D 2 0 phase2. In addition the integral v-flux from the ES reaction will also be determined during the pure D 2 0 phase. Extremely important for determining the NC-reaction is a knowledge of the radioactivity levels in the light and (especially) heavy water. This is because there are post-radium isotopes in both the 232Th and 238U decay-chains which have y-rays of energy high enough (> 2.22 MeV) to photodisintegrate the deuteron producing a free neutron indistinguishable from the NC-reaction on the deuteron. Because of this we have several techniques to assay the water for 224Ra, 226Ra and 222 Rn, and there is a regular program of assaying both the light and heavy water on a regular basis (every two weeks or so). For all of these isotopes we are at or have exceeded our design goals. After about one year, we will move into a second phase which is expected to be the addition of a chlorine salt, either NaCl or MgCl2, to the heavy water. The addition of chlorine increases sensitivity to the neutral-current reaction by increasing the total energy of y-ray emission (to about 8 MeV). In addition the ncapture cross-section is large so that most neutrons capture on chlorine rather than on deuterium or hydrogen. About 2 tonnes of salt give 80% n-capture on chlorine. This phase then will be used primarily to measure the NC-reaction and the total vflux and, from the ve flux determined in the pure D 2 0 phase, to determine the ratio of ve-flux to total v-flux. After about a year, the salt would be removed from the D 2 0 and a third phase will commence in which 3He proportional counters, installed vertically in an
88
array with an approximately 1-metre horizontal spacing, will be used as a second neutral-current detector. This system has the advantage that the neutrons are detected independently of Cerenkov light. This permits simultaneous CC and NC measurements and would be especially valuable in the event of a supernova in the Galaxy. It is estimated with either the salt or the 3He detectors in combination with the pure D 2 0 run that the ratio of ve-flux to v-flux will be determined to better than 9%. Since the end of October, 1999, SNO has been running quite smoothly with approximately 85% live time for solar neutrinos over 142 days of running. References 1.
2.
J. Boger et al., be published in Nuclear Instruments and Methods in Physics Research A, accepted Feb. 2000. See also the LANL preprint server nucl-ex/9910016 v2 (Nov. 1999). J.J. Simpson and J.-X. Wang, (1988) SNO-STR-001 and SNO-STR-002.
* The SNO Collaboration includes participants from Queen's University, Carleton University, University of Guelph, Laurentian University, University of British Columbia, University of Pennsylvania, Los Alamos National Laboratory, Lawrence Berkeley National Laboratory, University of Washington, Brookhaven National Laboratory, University of Oxford.
89
17.80 M OIA.
Fig. 1 The SNO detector, showing barrel-shaped cavity, the PMT geodesic support structure, acrylic vessel and light and heavy water volumes (ref. 1).
90
0.25
0.5
0.75
(rfit/600) 3
0.25
0.5
0.75
(rfit/600) 3
Fig. 2 The volume distribution of CC (upper figure) and NC events (lower figure) for events above about 6 MeV. The radius of the acrylic vessel corresponds to 1.
T H E SOLAR CORE A N D SOLAR N E U T R I N O S DALLAS C. KENNEDY Department of Physics, University of Florida, Gainesville FL 32611, USA E-mail:
[email protected] • • WWW: http://www.phys.ufl.edu/ kennedy The long-standing deficit of measured versus predicted solar neutrino fluxes is reexamined in light of possible astrophysical solutions. In the last decade, solar neutrino flux and helioseismic measurements have greatly strengthened the case for non-astrophysical solutions. But some model-independent tests remain open.
The solar neutrino problem has nagged physicists for over 30 years and, for most of us, has a natural solution in neutrino oscillations, either in vacuo or matter-enhanced (MSW effect). This modification of neutrino properties (with mv & 1 0 - 2 eV and mixing ~ 0.001-1) requires an extension of the Standard Model not detectable in accelerators and consistent with many models of unification. x But confidence in such a solution is based on a prior exclusion of astrophysical or nuclear physics explanations; only in the past decade has such an outcome become strongly credible. Predictions of the solar neutrino flux <j) are outputs of a solar model, which in turn are special cases of hydrogen-burning (4H —> 4 He) main sequence stellar models (almost all pp chain, with small CNO contribution). These predictions are usually quoted from a particular model (here the BahcallPinsonneault 1998 model), and detailed models agree if the same inputs are used. 2 ' 3 But a more generic approach is attractive if it can free us from a specific model with fixed parameters. Greater generality is all the more important if it reveals basic properties of the Sun and <j> that depend only on simple properties. Here I outline the results of such an approach. 4 ' 5 ' 6 As far as observations now go, it confirms the results of detailed solar models but also specifies crucial solar observations that remain to be filled in. 1
Properties of Solar and Stellar Models
Simple solar models orient us with gross structure: the core (r/R& ^ 0.3, where nuclear fusion generates the luminosity LQ), the radiative zone (0.3 £> T/RQ < 0.71) and the convective zone (CZ, T/RQ > 0.71). Less massive stars (M < M Q ) have, according to stellar models, even deeper convective zones; while for M > M 0 , the outer convective zone disappears, and a convective inner core appears as the central temperature gradient surpasses a critical value. 7 The Sun might have had a convective inner core, drastically changing 91
92
the
predictions, but there is now decisive evidence against core convection (see Section 3 below). A higher central temperature would also have led to CNO-dominance and higher neutrino fluxes. 8 A solar model is standard (an SSM) if the model contains all the physics of matter, gravitation, and nuclear fusion needed to obtain a star, but nothing more. Conceptually, stellar structure and evolution divide into three levels by time scales. For the Sun, chemical evolution needs ~ 10 Gyr; thermal equilibrium, about 10 Myr (Kelvin-Helmholtz time); and hydrostatic equilibrium, about 5 minutes. The hydrodynamic time controls the helioseismic pand -modes. (Late in the Sun's evolution, the chemical time scale will be shortened and the hierarchy blurred.) Structure means only the thermal and mechanical features of a star. 7 The properties of matter needed are constitutive relations giving pressure P, opacity K, and specific luminosity generation e as functions of density p and temperature T. If the thermal structure is given, then the equation of state reduces to a barytrope P = P(p), and the mechanical structure alone becomes a closed problem characterized by the stiffness profile r = d In P/d In p. The special case of T = 7 = 1 + 1 / n = constant is a polytrope of index n. All n < 5 polytropes have finite mass and radius, n = 0 is the constant-density case. 7 ' 5 The initial conditions are fixed mass M and the element mass abundances Xi (the zero-age star assumed chemically homogeneous). The boundary conditions are zero mass and luminosity at the center and (nearly) zero pressure and density at the surface. The Xi develop gradients by evolution, as nuclear fusion and heavy element diffusion act, and additional helium accumulates in the core. 7 ' 2 Even within the SSM framework, variations are possible. The K and e functions and fusion cross sections must be calculated from atomic and nuclear physics and extrapolated into regimes not directly testable. For the dominant luminosity-producing ppl reactions (terminating through 3 He fusion), e and the reaction rates are almost fixed by LQ. But reactions not strongly connected with L© are not well-constrained: the ppl I and ppl 11 chains (terminating from 7 Be through 7 Li and 8 B, respectively) are sensitive functions of the core temperature and nuclear cross sections, as well as mildly dependent on the core density. L Q and other global properties place only weak constraints on the ppll and ppIII rates. 8 ' 7 ' 4 2
Generalizing the Standard Solar Model
Generic properties of solar structure are restricted by boundary conditions. The SSM can then be generalized in two ways. One is to calibrate with the
93 specific model. This procedure defines a generalized SSM family, although not the most general. Its most convenient implementation is through homology or power-law scaling, derivable analytically or made evident by numerical solutions. Exploration of model space by varying SSM inputs is actually a special case of homology, which amounts to "small" perturbations of the logarithms of inputs and outputs. Such "perturbative" analysis works over a surprisingly large range, so long as the power law-relations are stable. 4 ' 7 The first signs of homological behavior in SSMs were found in the 1000 SSM Monte Carlo study of Bahcall and Ulrich. 8 Subsequent work over a wider model range revealed a much broader validity for homology. The underlying analytic structure was derived by Bludman and Kennedy. 4 Starting with structure alone, one keeps only dimensional and scaling behavior of macroscopic variables, dropping the differential nature of the equations. Assuming multifactor power laws for the equation of state, opacity, and luminosity generation, we found homological relations for the mechanical and thermal structure, assuming fixed powers. This requirement restricts the homology to the radiative and core regions, below the CZ. The dominant luminosity production is by ppl, carried outwards entirely by radiative diffusion. The constitutive relations are P/p = XT/n
,
K(p,T) = Ko(Xi)pnT-s
,
e(p,T) = e0(X)pxTv
. (1)
Expanded about the SSM, the exponents are: n = 0.43, s — 2.5, A = 1.0, v = 4.2. p,, K0, and e0 are composition-dependent. The luminosity constraint reads: <j>{pp) + (O.977)0(Be) + (0.751)>(B) + (0.956)<£(CNO) = 6.55 x 1010 cm" 2 sec - 1 . 4 ' 2 The boundary conditions are imposed in a way appropriate to a single star: MQ, LQ, and RQ fixed. Homology then gives a family of possible non-convective interiors consistent with observed outer solar features and parametrized by pc and Tc: pc ~ eJ°- 34 Kj°- 40 /i°- 52 L^ 085 , Tc ~ £^013K-°034H°-22LQA1 The resulting <j> scale stably with pc and Tc over a large range, giving the two-parameter homological mechanical/thermal variations of the SSM: (j>(i) ~ p^ • Tj?', with (aj,/3j) for pp, Be, and B i/'s being (—0.1, —0.7), (0.7,9), and (0.3,21), respectively. The highest reactions in the pp chain have the famous extreme sensitivity to T c , while all sensitivities to pc are mild and arise from the small luminosity contribution made by the ppl I, ppIII, and CNO chains. It should be stressed that pc and Tc are model outputs, like the (i). These exponents reproduce the 1000-SSM Monte Carlo and clarify that the entire homological class of SSMs has the wrong pattern of fluxes to explain the observed energy dependence of : lower energies are
94 more suppressed. (Variation of nuclear cross sections also fail to explain the pattern.) This conclusion depends only on pp/-dominance in e, radiative diffusion in K, and the ideal gas law. Homology applied to the mechanical structure alone automatically leads to a polytrope. 7 But the Bludman-Kennedy homology is more general than a polytrope, as it applies to both mechanical and thermal structure. It also scales correctly in the evolved core, where the molecular weight fj, changes substantially, reflecting the chemical evolution that makes the present Sun differ from its zero-age incarnation. No polytrope fits this behavior. 4 , s A more complete version of homology is possible if the differential structure is retained and rewritten using scale-invariant homology variables. 5 For the entire mechanical, thermal, and chemical structural system, the dimensionless differential equations are not less complex than a full SSM. But if the structure is restricted to the mechanical alone and a barytrope P(p) assumed, simple dimensionless structure equations follow. The key to the mechanical structure turns out to be the T profile. Constant T gives a polytrope again; in fact, TSSM — 4/3 outside the inner core, up to the CZ, where it rises to 5/3 (the adiabatic value). But within the core, fi rises and T drops towards to the center, where Tc ~ 8/9. Described in terms of a polytrope, the effective index neff = ( r — 1) _ 1 rises in the core, diverging at F = 1. Further towards the center, neg rises from minus infinity to a finite negative value at the center. Such behavior in the inner core is not even approximately polytropic and explains why attempts to use polytropes to approximate the Sun only work (and only crudely) over the whole Sun and fail badly in the core. The other approach to generalizing the SSM is to work with a few reasonable assumptions, following these to simple, testable predictions. Some powerful results are available, although restricted to mechanical structure only, 6 to which helioseismology is the key: sound waves are mechanical perturbations with information about sound speed, equation of state, and density and pressure profiles. 9 Taking full advantage of these results requires both p- (pressure) and g- (gravity) modes. Their spectra can be inverted to yield sound speed Cad(r) and bouyancy frequency N(r) profiles, from which follows the complete mechanical structure. The thermal and chemical structures cannot be directly probed by helioseismology, as their associated time scales are so long. Simple homology implies L ~ p,4M3n~1e~0, but changes in L© large enough to explain the solar v deficit are probably ruled out by paleoclimatology. 10 Direct tests of these aspects of the SSM require comparison to other Sun-like stars. Such stars vary from the Sun somewhat in mass and chemical composition and could lie anywhere on their respective evolutionary tracks. Comparison properties
95 include luminosity, surface temperature, and photospheric radius. With accurate photometry and parallaxes, accurate luminosities and colors are achievable. n The intermediary between these observations and stellar structure is stellar atmosphere models, which have advanced considerably in the last 30 years. Although still oversimplified, the models are good enough for solartype stars to infer ranges for surface T, g, abundances, and turbulence. 12 An exciting possibility will be opened by asteroseismology of Sun-like stars, as observation of stellar seismic modes (especially g-modes) would lead to direct characterization of stellar interiors. 13
3
Observational Issues
MQ, LQ, RQ, and surface T, as well as surface and proto-solar (meteoric) abundances, are well measured. Helioseismic observations have directly or indirectly captured millions of p-modes, allowing an accurate inversion of Cad(r) down to T/RQ = 0.05. 9 A convective core (7 = 5/3) is ruled out, although circulation of heavy elements not affecting heat transport cannot be at present. 14 The inferred sound speed peaks off-center at r/R& ~ 0.07. Hydrostatic equilibrium requires dc^/dr = 0 at the center, but an off-center peak occurs where r = 1. 5 This peak and dc^/dr > 0 for r/R& < 0.07 indicate T < 1 there, a crucial confirmation of the core's chemically evolved state. A complete profile down to the center becomes possible with the lowest p-modes. But complete inversion for model-independent mechanical structure would be possible only if the higher (/-modes were observed; analogous inversion would yield N(r), and together N and Cad yield T and other mechanical profiles, the first truly independent test of the SSM. 6 Comparing the Sun with other sun-like stars has been possible for many decades, albeit at poor precision. But the recent Hipparcos-Tycho star catalogs have revolutionized astrometry, raising the accuracy of nearby stellar parallaxes by up to a factor of 10 or better (1 m-arcsec). 15 Luminosities accurate to < 5% for nearby stars (within 25 pc) can now be inferred. With good color measurements and best current stellar atmosphere models, surface T's can be limited to 1%. n ' 1 2 Even more dramatic astrometric improvements could come from the proposed SIM and GAIA orbital systems, to be launched in 2006 and 2009, respectively: 4 ju-arcsec parallax errors and luminosity errors limited only by photometry. 15
96
Acknowledgments The author thanks the CSNP for the opportunity to present our results in honor of Frank Avignone's life and work. This work was done in collaboration with Sidney Bludman (Univ. Pennsylvania and DESY) and Susana Tuzzo (Univ. Florida), who helped sample the Hipparcos and SIMBAD catalogs, and was supported by the U.S. DOE under grants DE-FG02-97ER41029 (Univ. Florida) and DE-FG06-90ER40561 (Univ. Washington), the Institute for Fundamental Theory (Univ. Florida), and the Eppley Foundation for Research. References 1. S. A. Bludman, N. Hata, D. C. Kennedy, and P. G. Langacker, Phys. Rev. D 47, 2220 (1993); N. Hata and P. Langacker, Phys. Rev. D 56, 6107 (1997). 2. J. N. Bahcall, S. Basu, and M. H. Pinsonneault, Phys. Lett. B B443, 1 (1998); see also http://www.sns.ias.edu/~jnb/. 3. A. S. Brun, S. Turck-Chieze, and P. Morel, Astrophys. J. 506, 913 (1998). 4. S. A. Bludman and D. C. Kennedy, Astrophys. J. 472, 412 (1996). 5. S. A. Bludman and D. C. Kennedy, Astrophys. J. 525, 1024 (1999). 6. D. C. Kennedy, Astrophys. J. 540 (2000), in press. 7. R. Kippenhahn and A. Weigert, Stellar Structure and Evolution (Springer-Verlag, Berlin, 1990). 8. J. N. Bahcall and R. K. Ulrich, Rev. Mod. Phys. 60, 297 (1988). 9. J. Christensen-Dalsgaard, in Proc. IAU Symp. No. 189, ed. T. R. Bedding, A. J. Booth, and J. Davis (Kluwer Academic, Dordrecht, 1997). 10. T. J. Crowley and G. R. North, Paleoclimatology (Oxford Univ. Press, New York, 1991). 11. Proc. 111th IAU Symposium, ed. D. S. Hayes, L. E. Pasinetti, and A. G. D. Philip (D. Reidel Publishing, Boston, 1985). 12. See http://kurucz.harvard.edu/. 13. T. M. Brown and R. L. Gilliland, Ann. Rev. Astron. Astrophys. 32, 37 (1994). 14. A. Cumming and W. C. Haxton, Phys. Rev. Lett. 77, 4286 (1996). 15. M. A. C. Perryman et al, ESA Hipparcos and Tycho Catalogues SP-1200, 17 vols. (ESA Publications, Noordwijk, The Netherlands, 1997); see also http://sci.esa.int/hipparcos/ and http://sim.jpl.nasa.gov/.
N E U T R I N O S A N D T H E S T A N D A R D MODEL BARRY R. HOLSTEIN Department of Physics, University of Massachusetts, Amherst, MA 01003, USA E-mail: [email protected] Since their "discovery" by Pauli in 1930, neutrinos have played a key part in confirmation of the structure of the standard model of strong and electroweak interactions. After reviewing ways in which this has been manifested in the past, we discuss areas in which neutrinos continue to play this role.
1
Introduction
The neutrino is a particle whose impact on contemporary physics far outweighs its (possible) negligible mass. Indeed even the layman has long been fascinated by a particle which is (essentially) massless, chargeless, and which can pass through the earth without interaction—cf. the poem by John Updike written nearly four decades ago (allegedly when he became bored during a Harvard physics lecture) 1 . Neutrinos they are very small They have no charge and have no mass And do not interact at all At this meeting, we will be hearing from many experts on contemporary aspects of neutrinos, especially having to do with their role in astrophysics and cosmology. I shall not attempt to compete with these experts, but rather will discuss ways in which the neutrino has impacted and continues to affect our understanding of the structure of the standard model. After a brief historical introduction, I will emphasize ways in which the neutrino has affected the evolution of standard model structure even from the beginning and then will mention areas of contemporary physics wherein the neutrino continues to play a key role. 2
Neutrino History
The "discovery" of the neutrino is quite different from that of its sibling leptons in that its existence was inferred nearly three decades before its actual experimental confirmation. Indeed it was Pauli who in 1930 postulated the existence of a light neutral particle inside the nucleus—"not larger than 0.01 proton masses"—and called by him the "neutron," in order to explain why nuclear beta decay was observed to have a continuous (three-body) rather 97
98 than discrete (two-body) electron energy spectrum 2 . This issue of the spectrum was so troublesome at the time that no less an authority than Niels Bohr had speculated that it might be necessary to abandon the idea of energy conservation, except in a statistical sense, when considering subatomic processes such as beta decay 3 . Of course, there was a serious problem with Pauli's suggestion, in that a quick uncertainty principle estimate shows that a particle this light has a position uncertainty Ax ~ l / m „ ~ 400 A and could not therefore be confined within the nuclear volume. This problem was solved by Fermi, who renamed this particle the "neutrino" and proposed his famous field theory of beta decay
nw = ^lo^^to"^
+ h.c.
(i)
wherein this particle does not exist inside the nucleus but rather is created as a byproduct of the decay itself*. The one unknown constant GF can be determined from the neutron lifetime via
r
- = ( 3 s ) 7 $?$?2"siM- -u' - E- - E-)iM-1' ri2
= ~
pMn—Mp
dEeEePe(Mn -Mp-
Eef\Mw\2
~ 4.59 x 10- 19 GeV 5 G|<|.M tt ,| 2 = 887 ± 2 sec
(2)
which yields GF — 10~ 5 Mp. This was all very convincing and Bohr soon became a believer, acknowledging "Finally, it may be remarked that the grounds for serious doubts as regards the strict validity of the conservation laws in the problem of the emission of /3-rays from atomic nuclei are now largely removed by the suggestive agreement between the rapidly increasing experimental evidence regarding /3-ray phenomena and the consequences of the neutrino hypothesis of Pauli so remarkably developed in Fermi's theory" 5 . This is all well and good but it is one thing to postulate the existence of the neutrino and quite another thing to actually detect it. The problem lies in the size of the weak coupling inferred from beta decay. One can easily calculate a resulting neutrino scattering cross section as
a =
" ( 6 § ) / T^2nSiMp + E*-M"~ E')\M*>\2 „ !fJLPeEe\Mw\2
~ 10 _44 cm 2 at Ep = 1 MeV
(3)
99 The mean free path passing through a medium of earthlike density is then expected to be Aar ~ l/(/xr„) ~ 10 21 cm
(4)
10
or 10 earth radii!! The solution to this problem, of course, is to get lots of neutrinos, many target nuclei, or better yet both\ One of the original ideas conceived by Cowan and Reines to this problem of getting many neutrinos on target was to set off a nuclear bomb near an underground detector 6 . They soon had a more realistic thought, however, and decided to place the detector near a reactor. After original work at the Hanford site, they moved their base of operations to Savannah River and in 1956 were able to announce the unambiguous discovery of the neutrino via the reaction ve + p —> n + e+7. In retrospect, this discovery took place in the middle of a tremendous amount of seminal work, which led within a decade to the picture which we now call the standard model of weak and electromagnetic interactions. This included i) Suggestion of parity violation by Lee and Yang8 and its subsequent experimental confirmation9; ii) Postulation of the V-A structure of the weak current by Feynman and Gell-Mann 10 and its confirmation; iii) Development of the quark model by Gell-Mann and Zweig11; iv) Proposal of quark mixing by Cabibbo 12 ; v) Discovery of the standard electroweak model by Weinberg and Salam 13 . By 1967 then we already had what has become one of the most successful theories in modern physics. In this picture the neutrino plays a pivotal role and has at least three fundamental aspects which have been subjected to extensive experimental tests: i) Chirality: Because of the 1 + 75 structure of the weak interaction, the neutrino (antineutrino) must be purely left-handed (right-handed). ii) Dirac Character: The neutrino is predicted of Dirac character, possessing a distinct ahtiparticle, rather then a Majorana particle which is its own antiparticle. iii) Mass: The neutrino is massless, implying that there is no lepton analog to the CKM mixing occuring in the charged weak current.
100
Each of these predictions has been studied over the years and we now have a sizable data base of experimental evidence involving each issue. I will summarize each in turn: 2.1
Chirality
The prediction of definite chirality was first studied by Goldhaber, Grodzins, and Sunyar in 1958 via electron capture on 152Eu to an excited state of 152 Sm followed by its subsequent radiative decay to the ground state 14 . By studying those photons which are emitted opposite to the direction of the outgoing neutrinos one can show that the photon and neutrino helicities must be identical. Thus one can study the neutrino helicity by measuring that of the photon. When this was done the authors announced that "our result seems compatible with ... 100% negative helicity of neutrinos emitted in orbital e~ capture," although they did not really quantify this assertion. Since that time the chirality issue has been extensively studied. The way one does this is to postulate a form for the charged current weak interaction which includes right-handed components. A typical form for the semileptonic interaction is 15 C= ^ ^ -
[ ( ^ - M M ) K - a") + (xV,, + ypAJiv"
+ a*)]
(5)
where VM, A^ {v*1, aM) are hadronic (leptonic) weak currents respectively. Here /> = (1 — x)/{\ — y) with x, y being parameters which characterize the possible existence of right-handed effects. In a minimal left-right model of spontaneous symmetry breaking, we would identify x ~ 6 - C, 2
y ^ S+ C
(6)
2
where S — M /M measures the ratio of (predominantly) left- and righthanded gauge boson masses and £ is the mixing angle defined via W\ = WL cos£ — WR sin£. The tightest limits on x, y come from precise beta decay studies. Examples include measuring the longitudinal polarization of the electron, which is given by PL ~ /?(1 — 2y'2) or of the asymmetry parameter in the decay of polarized nuclei, which for neutron decay has the form 4
_
+ gy)- ygAJygA + xgv) 92v+3g2A + (x292v+3y2g2A)
*9A{9A
(
'
Over the years a series of careful studies has produced the limits shown in Figure 1, which generally limit x,y at the several percent level16. (The reason that generally tenth of a per cent precision in beta decay measurements results in only several percent limits on x,y is due to the feature that left and right
101
0.2 Rttotivft nwoMiwiMflts. Asymmetry / Pol.
o.ia
PF/P C T
0.16 Absolute nMOMirvnwits, A. and R,
0.H
B„ and R„ 0.12
•o 0.1
0.08
0.D6
0.D4
400 500
0.02
-0.1
- 0 . 0 6 -0.06 - 0 . 0 4 - 0 . 0 2
0
0.02
0.04
0.O6 0.08 0.1
Figure 1. Present experimental limits on right-handed parameters from beta decay experiments.
handed currents do not interfere, so that any deviations from standard model predictions are quadratic in x,y as can be seen above.) 2.2
Dirac Character
The Dirac character of the neutrino has also been extensively examined, and Frank Avignone has, of course, been extensively involved in such studies. Naively one might think that the issue would already be clear from the feature that while the reaction ve + p -> n + e
(8)
ve + p -»• p + e~
(9)
does not occur while
does. Equivalently the absence of neutrinoless double beta decay (A,Z)-*(A,Z + 2) + e- +e~
(10)
102
which can occur via sequential beta decay accompanied by the exchange of a virtual neutrino (=antineutrino) would seem to argue strongly against a Majorana character. However, both of these arguments are blunted if the neutrino has definite helicity, as experimentally seems to be the case. Indeed then even if the neutrino has a Majorana character, it has the wrong helicity to bring about the scattering or double beta decay reactions above, so that their experimental absence does not bear on the Dirac vs. Majorana issue. On the other hand, if the neutrino is Majorana and has a small mass, so that helicity is not definite, then neutrinoless double beta decay is possible and it is experiment involving 76Ge which Prank pioneered and has been doing for many years. Just this week a new limit on the Majorana mass of < mff > < 0.2 eV has been announced from such measurements 17 . 2.3
Neutrino Mass
The issue of whether the neutrino has a mass is clearly a fundamental one. In the standard model the absence of mass is due to Ockham's Razor—i.e., the standard model uses only the minimal number of components. Since a right handed neutrino structure is not necessary, the standard model assumes its absence and, as it requires both a left and right handed component in order to generate a mass, the neutrino is predicted to be massless. This prediction is one which has been under experimental scrutiny for many years. (Even Fermi in his original paper suggested examination of this issue by looking at the electron energy dependence of beta decay spectra near the endpoint 4 .) Generally most such studies have utilized SH due to its low—18.6 KeV—endpoint, since that maximizes the interesting component of the electron spectrum, and such studies have become increasingly precise. An early value by Hamilton, Alford, and Gross placed the upper limit at 250 KeV 18 , which in 1972 was lowered to 60 eV by Bergkvist 19 . In 1980 Lubimov announced a nonzero value 14eV < mp < 46 eV, which set off a firestorm of new work 20 . Present experiments do not not find evidence for a nonzero neutrino mass and upper bounds have been place at 9.3 eV by Robertson et al. 21 and at 7.2 eV by Weinheimer et al. 22 , so that the Lubimov value has been superceded. Work continues on such direct mass measurements. Some interesting new ideas have been discussed at this workshop, including use of Rhenium, with an endpoint energy even lower than that of3H and the use of bolometric methods to detect the electron. In is interesting to note in this regard, that in the middle of this intense activity to measure neutrino mass, an event occurred which bears on this issue and which allows a simple limit to be set which is nearly comparable to those
103
Figure 2. Views of a region in the Large Magellanic Cloud before (right) and after (left) the morning of February 27, 1987: SN1987a is clearly visible. Copyright Anglo-Australian Observatory. Photograph by David Malin (http:// www.aao.gov.au/images/captions/aat050.html).
obtained from these careful spectral studies—-SN1987a? which blazed into the sky on February 27, 1987 and was observed not only optically, but also by neutrino detectors in the US and Japan— cf. Figure 2. If the neutrinos emitted by the supernova were massive, then the velocity would be v ~ 1 - m*/(223J) and the most energetic neutrinos would reach the earth first. It is easy to estimate the time difference as <$* <$£ ml 8EV .... T ~ T ~ El Ev and the time gap between the arrival of the high and low ener©r neutrinos could then be used to measure this mass. Experimentally, the ~ 10 MeV neutrinos arrived over a ~ 10 second time interval after travelling a distance of 165,000 light years from the supernova in the Large Magellanic Cloud but a time-energy correlation was not observed. One can then easily set a limit
104
on the mass—
•*£*(TS:)J~10~GP?;)1"'"''V
(12)
A more careful analysis sets the upper bound at about 20 eV. It is astounding to me that the relatively trivial analysis given above from an event occurring long before the dawn of civilization is able to set a limit on neutrino mass comparable to that obtained from years of precise experimental studies! Of course, I have summarized here only the direct mass measurements. Simultaneously, a series of experiments involving a search for neutrino mixing has been underway. Such mixing is prohibited in the absence of mass since neutrino identities could just be reassigned. The recent announcements of mixing signals from solar, accelerator, and atmospheric measurements then clearly, if comfirmed, indicates the existence of neutrino mass. Since this will be the subject of many talks during this workshop, I will not here summarize this data but instead will move on to discuss aspects of neutrino physics which are not as well known, but which have a bearing on contemporary physics issues. 3
Contemporary Issues
Above we have seen how the neutrino has played an essential role in development of the structure of the standard model. In this section, I argue that this is still going on and discuss ways in which neutrino interactions are involved in a number of the central issues in contemprary physics. In this discussion, I will not emphasize some of the more traditional ways in which this is manifested—e.g. i) use of neutrino scattering in order to study the Q2 evolution of deep inelastic structure functions as a test of perturbative QCD, ii) use of such deep inelastic structure functions in order to check the validity of various sum rules, such as the Adler sum rule
1 = f ^(FP(x,q*) - F-TW)),
(13)
since these are fairly well known. Instead I will discuss three lesser known applications wherein neutrino studies have a bearing on interesting standard model issues.
105
3.1
Goldberger-Treiman Discrepancy
One of the important features of QCD is its (broken) chiral symmetry, from which follows the existence of the Goldberger-Treiman (GT) relation, which connects the strong pion-nucleon coupling gnNN and the axial coupling 3A (0) measured in neutron beta decay 23 , MNgA(0)
= FngnNN(0)
(14)
where Fv — 92.3 MeV is the pion decay constant. One subtlety associated with Eq. 14 is that the pi-nucleon coupling is evaluated not at the physical point—q 2 = m 2 —but rather at the unphysical value—q2 = 0. In fact when the physical coupling is used, one expects a violation of the GT identity and this is often showcased by quoting the so-called Goldberger-Treiman discrepancy A
9A(0)MN
t
9-KNNKmi)**
Strictly speaking the value of A„ is given by a chiral counterterm, but in a reasonable model one would expect gnNN(q2) to vary with q2 in essentially the same way as its weak analog <M(<72)- In this way one finds
A
- =! - -Ei^h = r^ m - - 0034
( 16 )
9A(m2) 6 * where rA = 0.65 ± 0.03 fm is the axial charge radius measured in charged current neutrino scattering 24 ." An alternative approach is to utilize the Dashen-Weinstein relation _ y/3mlFK
*• _
o™2 IT. ZmKl'7r
fg\KN
1 9EKN
AA
I ~ ^K \g-KNN
AS
^
1= ^K VO 9nNN
/
, 1fi ,
\l°)
which predicts the pionic GT discrepancy in terms of its kaonic analogs involving A and E couplings respectively 25 . The original proof of this result argued that it was valid up to terms second order in chiral symmetry breaking. However, recently it has been shown by Goity et al. that in heavy baryon chiral perturbation theory any such difference can arise only at 0(ph) or higher 26 . Although the strong A, E couplings are not well determined, the predictions "Here the axial charge radius is defined via 9A(Q2) = 9 A(0)(l + ^q2+ D
...).
(17)
106
are only weakly dependent upon thse quantities. Thus one finds the relatively robust prediction A^Dashen - Weinstein) = 0.017
(19)
in good agreement with that expected from neutrino scattering results. Now what does experiment say? The problem here is that while the pion decay constant, the nucleon mass, and the weak axial coupling are all well known, there is still considerable debate about the value of the size of the pion nucleon coupling constant. A recent analysis of NN, NN, irN data by the Nijmegen group yields the value gVNN(m%) = 13.05 ± 0.08 27 and a VPI analysis yields similar results 28 . However, a significantly larger number— Jirivjv(mj) = 13.65 ± 0.30—has been found by Bugg and Macleidt 29 and by Loiseau30. When these values are used in order to calculate the GT discrepancy, we find An = 0.014 ± 0.006
if
gnNN
= 13.05 ± 0.08
An = 0.056 ± 0.020
if
gnNN
= 13.65 ± 0.30
(20)
The neutrino scattering number Eq. 16 then comes right in the middle, while the Dashen-Weinstein analysis strongly supports the lower value of gnNN • 3.2
Axial Charge Radius
A second interesting application of neutrino scattering results is associated with confirmation of a prediction of chiral perturbation theory and therefore of QCD. In order to understand this point, we return to the early days of current algebra and PCAC and a low energy theorem derived by Nambu and Schrauner, which argues that the axial charge radius may be obtained via measurement of the isospin odd £b+ multipole in threshold electroproduction via 31 E<;)(m,
= ftO - H
(l +klr\
+
^(
W +
I) + 0<*'))
(2.)
In this way one has determined the value TA = 0.59 ± 0.05 fm,32 differing from the number TA = 0.65 ± 0.03 fm found via direct neutrino scattering measurements. Although the discrepancy is only at the one sigma level, it is interesting that recent calculations by Bernard, Kaiser, and Meissner in heavy baryon chiral perturbation theory have shown that the old low energy theorem is incorrect and that there exists an additional contribution coming from socalled triangle diagrams, which predicts a difference between the axial charge
107
radius as measured in neutrino scattering and that from electroproduction 33 r^elec.) = r ^ n e u ) - ^ ( ^
- 1)
(22)
The 0.046 fm2 difference predicted from chiral symmetry agrees well in size and sign with that seen experimentally. 3.3
Nucleon Strangeness Content
My final example has to do with the subject of strangeness content of the nucleon, which is one of intense current interest. One of the early studies of such matters is the paper of Donoghue and Nappi 34 . The idea here is that one expects that in the limit of vanishing quark masses the nucleon mass should approach some nonzero value MQ. In the real world, with nonzero quark mass, the nucleon mass is modified to become MN = M0 + os + a
(23)
where, defining m = (mu 4- m
» = 77TT- < N\m8ss\N 2MM
>,
a = — — < N\m(uu + dd)\N > 2MN
(24)
are the contributions to the nucleon mass from strange, non-strange quarks respectively. One constraint in this regard comes from study of the hyperon masses, which yields 7YI
—
S = ——— < N\uu + dd- 2ss\N > 2-MJV
= I ™* 2 ( M s - M A ) = 25MeV (25) 2 mjt - mi and increases to about 35 MeV when higher order chiral corrections are included. A second constraint comes from analysis of TTN scattering, which says that a can be extracted directly if an isospin even combination of amplitudes could be extrapolated via dispersion relations to the (unphysical) Cheng-Dashen point F 2 D(+)( S = M 2 r ,t = m 2 r )=<7
(26)
When this is done the result comes out to be ~60 MeV, which is lowered to about 45 MeV by higher order chiral corrections. If < ./V|ss|iV > = 0, as might be expected from a naive valence quark picture, then we would expect the value coming from the hyperon mass limit and that extracted from nN
108 scattering to agree. The fact that they do not can be explained by postulating the existence of a moderate strange quark matrix element < N\ss\N
>
(27)
f = < N\uu + dd + ss\N > ~ 0 . 1
implying Mo ~ 765 Mev and as ~ 130 MeV, which seem quite reasonable. However, recent analyses have suggested a rather larger value of the sigma term, leading to / ~ 0.2, M 0 ~ 500 MeV and a8 ~ 375 MeV, which appear somewhat too large. This problem is ongoing. In any case it is of interest to study the possibility of a significant strange quark matrix element in other contexts. One quantity which has been extensively studied is the nucleon electromagnetic matrix element, which has the form < N\V™\N =
U(P')
>=< N\\ulllU
- \d^d-
b7lls\N
7M(*T(<7 2 ) + W ) ) - ^T^^^riQ2) 2M,N
> + F.!(q2)) «(PX28)
and one can look for the existence of strange quark pieces F*(q'2),F2(q2) in parity-violating electron scattering. This has been done in the forward direction by the HAPPEX experiment at JLab 3 5 and in the backward direction by the SAMPLE experiment at MIT-Bates 36 . The HAPPEX result is consistent with r{ = 0, while the Bates result seems to indicate a small positive value for /i g . Another probe comes from the realm of deep inelastic electron scattering wherein, defining the quark helicity content Aq via A<7
(29)
>
one has the constraint
/
dxgi (x) =
1
-Au + -Ad + -As
/x
a«(g2)x
(30)
JO
When combined with the Bjorken sum rule and its SU(3) generalization AM-
Au + Ad-
Ad = gA(0) =F + D 2As = 3 F - D
(31)
one finds the solution Au — 0.81, Ad = -0.42, As = - 0 . 1 1 , indicating a small negative value for the strange matrix element. So far, these results have nothing to do with our main focus, which is neutrinos. However, we note that there exists an alternative probe for such
109 a strange matrix element which is accessed via neutral current neutrino scattering. The point here is that the form of the standard model axial current is < N\A%\N > = - < iV|W7M75« - d-y^d
- s 7 ^7 5 s|iV >
(32)
which is purely isovector in the case that the strange matrix element vanishes and can be therefore be exactly predicted from the known charged current axial matrix element. This experiment was performed at BNL and yielded a result 37 As = - 0 . 1 5 ±0.09
(33)
consistent with that found from the deep inelastic sector, but a more precise value is needed. 4
Conclusion
We have argued above that the neutrino has played and continues to play an important role in the development of the standard model. In the past such studies contributed to the now accepted picture of the weak interaction. Present work looks for small deviations from this structure. However, we have also seen how neutrino experiments bear on a number of issues of great interest in contemporary physics and I suspect that neutrino measurements will continue to be exciting far into the new millenium. Acknowledgement It is a pleasure to acknowledge the hospitality of University of South Carolina and the organizers of this meeting. This work was supported in part by the National Science Foundation. References 1. J. Updike, in Telephone Poles and Other Poems, A.A. Knopf, New York (1963). 2. See, e.g. L.M. Brown, Phys. Today 31, 23 (September 1978). 3. N. Bohr, Faraday Lecture, J. Chem Soc, 349 (1932). 4. E. Fermi, Z. Phys. 88, 161 (1934); see also the translation given by A.L. Wilson, Am. J. Phys. 36, 1150 (1968). 5. N. Bohr, Nature 138, 25 (1936). 6. See, ee.g., "Celebrating the Neutrino" in Los Alamos Sci. 25, 1-191 (1997).
110
7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36. 37.
C.L. Cowan, et al., Science 124, 103 (1956). T.D. Lee and C.N. Yang, Phys. Rev. 104, 254 (1956). C.S. Wu et al., Phys. Rev. 105, 1413 (1957). R.P. Feynman and M. Gell-Mann, Phys. Rev. 109, 193 (1958). See, e.g. M. Gell-Mann and Y. Ne'eman, The Eightfold Way, Benjamin, New York (1964). N. Cabibbo, Phys. Rev. Lett. 10, 531 (1963). S. Weinberg, Phys. Rev. Lett. 19, 1264 (1967). M. Goldhaber, L. Grodzins, and A. Sunyar, Phys. Rev. 109,1015 (1958). See, e.g. B.R. Holstein and S.B. Treiman, Phys. Rev. D15, 3472 (1977). See, e.g. J. Deutsch, nucl-th/9901098, to be published in Proc. WEIN98. Cern Courier, 40, # 1 , p.8 (2000). C.S. Wu, in Alpha, Beta, and Gamma-Ray Spectroscopy, ed. K. Siegbahn, North Holland, Amsterdam (1965), Vol. II, p. 1391. K.-E. Bergkvist, Nucl. Phys. B39, 317 (1972). V.A. Lubimov et al., Phys. Lett. B94, 299 (1980). R.G.H. Robertson et al., Phys. Rev. Lett. 67, 957 (1991). Ch. Weinheimer et al., Phys. Lett. B300, 210 (1993). M.L. Goldberger and S.B. Treiman, Phys. Rev. 110, 354 (1958). T. Kitagaki et al., Phys. Rev. D28, 436 (1983); L. A. Ahrens et al., Phys. Rev. D35, 785 (1987) and Phys. Lett. B202, 284 (1988). R. Dashen and M. Weinstein, Phys. Rev. 188, 2330 (1969). J. Goity et al., Phys. Lett. B454, 115 (1999). J.J. deSwart, M.C. M. Rentmeester, and R.G.E. Timmermans, Proc. MENU97, TRIUMF Rept. 97-1, 96 (1997). R.A. Arndt, I.I. Strokovsky, and R.L. Workman, Phys. Rev. C52, 2246 (1995). D.V. Bugg and R. Machleidt, Phys. Rev. C52, 1203 (1995). B. Loiseau et al., TTN Newsletter 13, 117 (1997). Y. Nambu and E. Lurie, Phys. Rev. 125, 1429 (1962); Y. Nambu and E. Shrauner, Phys. Rev. 128, 862 (1962). A. del Guerra et al., Nucl. Phys. B107, 65 (1976); M.G. Olsson, E.T. Osypowski, and E.H. Monsay, Phys. Rev. D17, 2938 (1978); S. Choi et al., Phys. Rev. Lett. 71, 3927 (1993). V. Bernard, N. Kaiser, and U.-G. Meissner, Phys. Rev. Lett. 69, 1877 (1992). J.F. Donoghue and C. Nappi, Phys. Lett. B168, 105 (1986). K.A. Aniol et al. Phys. Rev. Lett. 82, 1096 (1999). D.T. Spayde et al, Phys. Rev. Lett. 84, 1106 (2000). L.A. Ahrens, Phys. Rev. Lett. 35, 785 (1987).
M A S S M A T R I X FOR ATMOSPHERIC, SOLAR, A N D L S N D NEUTRINOS S. P. ROSEN Office of Science, U.S. Department of Energy Presented to the Carolina Symposium on Neutrino Physics, 10-12 March 2000 In Honor of Frank T. Avignone III It is a special pleasure for me to speak at this symposium honoring Frank Avignone. I have known Frank for many years as a stalwart investigator of neutrino physics, especially the search for no-neutrino double beta decay, and I have enormous respect for his patience, persistence, and infinite care in performing these most difficult experiments. My best wishes for current and future endeavors go with this talk. Today we have sets of evidence for neutrino oscillations, and therefore neutrino mass, coming from atmospheric, solar, and LSND neutrino observations. It is very curious that the first two sets suggest large mixing angles but relatively small mass differences, whereas the third suggests small mixing with a relatively large Am 2 . It is generally accepted, in the event that all three sets of evidence are confirmed, that it will be necessary to introduce a fourth, sterile neutrino to describe the mass and mixing spectrum. Here I want to present a relatively simple scheme for the four neutrinos. In the case of atmospheric neutrinos, it has been clear for some time that the mixing of v^ and its partner in oscillations is maximal and that the corresponding Am 2 is of the order 10~ 3 eV2. Moreover, it seems most likely that the partner of the muon neutrino fM is the tau neutrino vT. In the case of solar neutrinos, there are several competing solutions, but three of them - namely, the large-angle and low MSW solutions and in vacuo oscillations - also require large mixing between the electron neutrino ve and its partner, which most likely is a sterile neutrino vs. I shall therefore assume that there is maximal mixing between ve and vs with Am 2 in the range 10~ 7 to 10~ 4 eV2. LSND provides evidence for small mixing between ve and v^, but with a "large" Am 2 , of the order 1 eV2. I thus take the point of view that v^ and vT form one maximally mixed doublet with a separation Am 2 of a few xlO~ 3 eV 2 , and that ve and vs form a second maximally mixed doublet about leV2 below i>^ and vT and with a much smaller separation Am 2 in the region of 1 0 - 5 eV2. LSND then represents a weak transition between the two doublets, as illustrated in Fig. 1. 111
112
00 sin2 29 < 1(T2
sin2 29 « 1 Atmospheric i/'s. Am 2 « K T W 2
LSND. Am 2 « lei/ 2
Solar v 's, Am2 « 10_5'6eV2
< mpfj > • "pseudo-Dirac" Figure 1.
In the case of one doublet, the most general 2 x 2 mass matrix leading to maximal mixing is * M 2 * = # ms mk * , mk ms
*6
(1)
The eigenvalues and eigenvectors are:
m±-(ms±mk),
ip± = -y=(ipa±i/jb).
(2)
We can introduce a small mixing between these two states by rotating about the y-axis through an infinitesimal angle {—289) M2 ->•
y^\
m s + mfc sin (2(5$) m* cos (2(5$) mfc cos (2(5??) m s — mk sin (2<5i9)
/
\ -ip+sm{dv)
+
ip_cos(dw)J
(3)
(4)
Note that this system issymmetric under the permutation group S-2 of two objects. To expand this matrix to the 4 x 4 case, we replace the elements
113
of M 2 by 2 x 2 matrices
m
--"sC£)-
<5>
™*^cm>
®
and thus we obtain M4 M4s
M K K M
(7)
It is not difficult to show that the eigenvalues of M4 fall into two pairs with equal splittings: A : ms + mk ± md ,
(8)
S : ms - mk ± md .
(9)
where A and 5 indicate the association with atmospheric and solar neutrinos, respectively. The differences of squared masses are then Am2A = 4 (m s + mk) md ,
(10)
A m | = 4 (m s - mfc) md •
(11)
If we define Ms = MQ> + e, mk = mo — e, we obtain Ami 's Am\
e m0
10~ 5 10- 3 '
(12)
We can also express the mass-squared difference for LSND as ^mlsND
~ ( m « +mk-
rnd)2 - (m3 - mk + md)2 = 4m s {mk - md).
(13)
As a numerical example, take AA = 3 x 10- 3 eV 2 ;
A s = 10- 5 eV 2 ;
ALSND
= leV2 .
(14)
Then m 0 Pa 1.001 eV ,
(15)
114
e w 1.5 x 1 0 - 3 eV « 2md .
(16)
If we take one of the two doublets as a pseudo-Dirac neutrino, the mass difference is consistent with the theoretical Majorana mass limit obtained from analyzing all of the current data, namely < m00 > < 6 x 1(T 3 eV .
(17)
Finally, I would like to point out that M\ has an Si x S2 permutation symmetry and that the form of K corresponds to an "equal spacing rule."
W H A T IS C O H E R E N T IN N E U T R I N O OSCILLATIONS T H E A N A L O G W I T H A TWO-SLIT E X P E R I M E N T H.J. L I P K I N Weizmann Institute, Rehovot 76100, and School of Physics and Astronomy, Tel-Aviv University, Tel-Aviv 69978, Israel and High Energy Physics Division, Argonne National Laboratory Argonne, IL 60439-4815, USA
Department
of Particle
Physics,
Israel
The standard treatment of neutrino oscillations recalls the story of the mathematics lecturer who said "It is obvious that...." then stopped and said; "Is it obvious?", went out for a half hour, and returned to say; "Yes it is obvious". The textbook neutrino-oscillation wave function, a coherent linear combination of states with different energies, is not found in any real experiments. Its application to reality is obvious. But the interval between "Is it obvious?" and "Yes it is obvious" is filled with many wrong arguments, many wrong papers, and more papers showing that the wrong arguments are wrong. We clarify this issue by describing the passage of a neutrino from source to detector as a multipath experiment where knowing the path destroys coherence, considering the beam and the detector as a correlated quantum system and applying this approach to Bragg scattering by X-rays as well as neutrinos. Amplitudes with the same energy and different masses are detected coherently and produce oscillations. Amplitudes with different energies are incoherent. Quantum mechanics alone shows the existence of a neutrino mass difference to be required to explain the observed Super-Kamiokande data.
1
Introduction - Frank Avignone and Neutrino Coherence
It is a pleasure to participate in this tribute to Frank Avignone. My first contact with Frank's physics intuition was when I was asked NOT TO REFEREE a paper by Frank and Friends1 which questioned a well-known treatment of coherence between different neutrino scattering amplitudes. The editor wanted a really impartial referee because Frank's friends were also from Israel, but he urged me to submit a companion "pedestrian" paper2 for experimenters like him who could not understand the rigorous theorems of Frank et al. Frank insists that this paper was really the work of his friends. But Frank provided the physical intuition and understanding of what coherence was all about even though he couldn't prove himself that a paper was wrong, and provided the motivation for those who could. The present talk investigates why and how neutrino states with different masses can be coherent. When a pion decays at rest -n —> [iv the energies and 115
116
momenta of the neutrino and muon are all known. This is just a "Missing Mass" experiment. The value of Mv is uniquely determined by M% = (M^ — E/j,)2 — p^. So how can there be interference between states of different mass? The answer to this paradox is found by using another paper by Frank's friends discussing which amplitudes are coherent in quantum mechanics3. Best wishes to Frank for continued fruitful activity for many years to come. 2
Interpreting the Standard Textbook Wave Function
The standard textbook neutrino-oscillation wave function has several states with the same momentum and different energies but has a constant magnitude over all space and is not going anywhere! The probability is the same to find a neutrino anywhere any time. But the phase difference between two waves having the same momentum and different masses (mi,m 2 ) and energies (.&L,£2) increases linearly with time at a rate calculated in the textbooks. M (p WM ffi-Eft 64> = (E1-E2)t= {Ei+E2)
K-mJ)* = ^
{m\-mj)x =
(1)
where we have set h = 1. E = {E\ + E2)/2 and used the velocity v = pjE = x/t to convert time into distance and obtain the OBVIOUS connection from textbook frequency to a real experiment where the v moves from a source to a detector at a distance x from the source. IS IT REALLY OBVIOUS? The two states with the same momentum and different energies also have different velocities, V\ and v2 and arrive at the point x at different times ti and t2. This suggests a different result,
5^E1t1-E2t2=(^-^)-x=Wzm^=(<^^ V vi v2 J p
P
(2)
The direct description of oscillations in space between states having same energy and different momenta4 gives a result equivalent to the OBVIOUS textbook result (1) and disagrees with (2) by a factor of two. ,, , x (Pi - P p S s = (Pl -P2)x = 2-
("*2 ~ mJ)X = Yp—
/ON (3)
Which is correct? What has happened between source and detector? Why same energy and different momenta? Many wrong papers have been written. Many more to show that these are wrong5'6.
117
3
The Analog with Two-Slit and Bragg Scattering Experiments
To understand which amplitudes are coherent and confirm that YES IT IS OBVIOUS and the textbook answer is OK, consider neutrino oscillations as two-slit experiments in momentum space with a quantum detector. In a neutrino oscillation experiment a particle passes without being observed between a source and a detector, just as in the two-slit electron diffraction experiment. In Bragg scattering a photon may be scattered by any one of the atoms in the crystal, but which atom scattered the photon is not known. In a neutrino oscillation experiment, the neutrino carrying momentum and energy from source to detector may be in any of the allowed mass eigenstates, but which carries this momentum and energy is not known. The amplitude at the detector is the coherent sum of the amplitudes from all allowed paths in energy-momentum space. Both in Bragg scattering and neutrino oscillations there would be no coherence if the energy and momenta of all relevant particles were measured precisely and momentum conservation could determine which amplitude produced a given final state at the detector. But coherence between amplitudes is not introduced by simple ignorance of which path was taken7. Coherence results only from an uncertainty required by quantum mechanics. To clarify the relation between coherence and incompleteness of knowledge consider8,9 a simple "two-slit" which-path experiment?'10 with a particle beam split into two paths and the two amplitudes, denoted by \L(x)) and \R(x)), then recombined at a point a; on a screen. A classical detector in one path determines the path taken and destroys all coherence. A quantum detector is a quantum system which undergoes a transition denoted by \Di) —> \Df), where \Di) and \Df) denote the initial and final states of the detector. If there is a quantum detector in the "R" path, the wave function for the combined system of the particle and the detector and the intensity observed at x are5: V(x,D) 2
= \L(x),Dt) 2
+ \R(x),Df)
I(x) = | \L(x)) | + | \R(x)) | + 2Re[\(L(x) \R(x)) | • e^*) . (Di \Df)\
(4) (5)
where 9{x) is relative phase of \L(x)) and \R(x)) and its variation with x gives the interference pattern on the screen. The quantum detector introduces an additional factor and phase into the interference term, given by the detector overlap (Di \Dj). Bragg scattering is a which-path experiment with many paths, one for each scattering atom, and a quantum detector in each path. The detector is the full lattice and each interference term between two paths contains two coherence factors {Di \Df) that depend on the lattice dynamics. The probability PDW that the scattering is coherent is called the "Debye-Waller" factor10'11, PDW =
118
|(-Dj \Df) | 2 which defines the relative amounts of coherent Bragg scattering and incoherently scattered photons observed outside the Bragg peak. 4
Detailed Quantum Mechanics of a v Source and Detector
We now apply the general which-path formalism developed above to a neutrino - detector system, where a neutrino vu with energy, mass and momentum E„, rrik and Pk is detected via the transition vk + n —> \x~ + p occurring on a neutron in the detector, The detector transition matrix element is (Dkf\T\Dt)
= <-Dfc/|/+e^-^)*|A)
(6)
where Di and Dkf denote the initial detector state and the final state produced in the "path A;"; i.e. after the absorption of a neutrino with mass rrik and emission of a /j,~ with energy E^ and mmentum PM, X denotes the neutron co-ordinate and i+ denotes the charge exchange isospin operator. The overlap between the final detector wave functions after the transitions absorbing neutrinos with masses mk and rrij is then (Dkf\Djf)
= (Di\ei&-p>>*\Di)
(7)
If the quantum fluctuations in the position of the active nucleon in the initial state of the detector are small in comparison with the oscillation wave length, h/(Pj — Pk), there is effectively a full overlap between final detector states after absorption of different mass neutrinos, and a full coherence between neutrino states with the same energy and different momenta.
|;3-ni2-«i (Dkf \Dif) « 1 - (1/2) • |i3- - Pk\2 • (Di\ \X2\ \D{) * 1
(8) (9)
The total energies of the final muon and detector produced after absorption of neutrinos with different energies are different. These muon-detector states are thus orthogonal to one another and there is no coherence between detector states produced by the absorption of neutrinos with different energies. 5
Simple Quantum Mechanics and Super-Kamiokande
Simple quantum mechanics alone, without the full apparatus of the standard model, shows that the Super-Kamiokande results12 require the existence of two different mass eigenstates for neutrinos. The energy spectrum of atmospheric neutrinos cannot change between their source at the top of the atmosphere
119
and their detection in a detector on earth if neutrinos are not absorbed, do not decay and interactions conserve energy. If there is only one neutrino mass value, the energy and momentum spectra will be identical for the upward and downward going neutrinos incident on the detector and no difference can be observed. The observation of the up-down difference12 therefore indicates that there are at least two different mass eigenstates, and that the difference can arise from interference between the waves of states having different masses and therefore different momenta if they have the same energy. This conclusion depends only upon quantum mechanics and previous experimental observations that ve and v^ are orthogonal. It does not depend on the details of the standard model and remains valid even if there is new physics beyond the standard model that does not violate quantum mechanics. Acknowledgements We thank Eyal Buks, Maury Goodman, Yuval Grossman, Moty Heiblum, Yosef Imry, Boris Kayser, Lev Okun, David Sprinzak, Ady Stern and Leo Stodolsky for helpful discussions. This work was supported in part by a grant from USIsrael Bi-National Science Foundation and by the U.S. Department of Energy, Division of High Energy Physics, Contract W-31-109-ENG-38. References 1. Y. Aharonov, F. T. Avignone, III, A. Casher and S. Nussinov, Phys. Rev. Lett. 58 (1987) 1173 2. Harry J. Lipkin, Phys. Rev. Lett. 58 (1987) 1176 3. Ady Stern, Yakir Aharonov and Yoseph Imry, Phys Rev. A41 (1990) 3436 4. H.J. Lipkin, Phys. Lett. B 348 (1995) 604. 5. Harry J. Lipkin, hep-ph/9907551 Physics Letters B 477 (2000) 195 and references therein 6. Harry J. Lipkin, hep-ph/9901399, in Proceedings of the Europhysics Neutrino Oscillation Workshop (NOW'98) 7-9 September 1998. Amsterdam. Published in http://www.nikhef.nl/pub/conferences/now98/ 7. Kurt Gottfried, Am. J. Phys. 68 (2000) 143 8. E. Buks et al, Nature 391 (1998) 871 9. D. Sprinzak, E. Buks, M. Heiblum and H. Shtrikman, cond-mat/9907162 10. H.J. Lipkin, Phys.Rev. A42 (1990) 49 11. H.J. Lipkin, Hyperfine Interactions, 72 (1992) 3. 12. Y. Fukuda et al Super-Kamiokande Coll. Phys. Rev. Lett. 81 (1998) 1562
ATMOSPHERIC NEUTRINOS IN
SUPER-KAMIOKANDE
H. S O B E L Department of Physics and Astronomy University of California, Irvine 92697-4575 (for the Super-Kamiokande Collaboration) In this paper I will present the updated results from the Super-Kamiokande atmospheric neutrino analysis. We have now completed the analysis of 1144 live days of d a t a corresponding to an exposure of 70 kt-years. I will discuss our measurements of contained, partially contained and entering events and the flavor ratio and angular distribution of the observed neutrino events. Using these observations, we calculate the neutrino oscillation parameters that best represent the data and our limits on a sterile neutrino interpretation of the oscillation.
1
Introduction
Atmospheric neutrinos are produced in the Earth's atmosphere as a result of hadronic cascades originating from interactions of cosmic ray primaries. A number of models have been developed to calculate these fluxes 1'2. Absolute neutrino fluxes are predicted with uncertainties of 20%, whereas the ratio of the Vy, +7^7 to ve + 7^T flux is believed to be known co better than 5%. SuperKamiokande has now collected over 70 kt-yrs of atmospheric neutrino data. This vast data set comprises ten times the data of the largest previous result. 2
The data
Super-Kamiokande is a ring-imaging water Cherenkov detector containing 50 ktons of ultra-pure water in a cylindrical stainless steel tank. The tank is 41.4 m high and 39.3 m in diameter and separated into two regions: a primary inner volume viewed by 11,146 50 cm diameter photomultiplier tubes (PMT's) and a veto region, surrounding the inner volume, and viewed by 1885 20 cm PMT's. For physics analysis, a software-defined 22.5 kton fiducial volume beginning 2m from the inner detector's PMT planes is used. Within this volume, efficiency is near 100%, with negligible contamination. Neutrino interactions occuring inside this fiducial volume are called contained events. These are reconstructed to determine their vertex and identify the number, direction, energy, and particle type of Cherenkov rings, as well as any subsequent decay electrons. Fully-contained (FC) events are those which have no exiting signature in the outer veto detector, and comprise the bulk of the contained event sample. In addition, a partially-contained (PC) sample is 120
121
Super-Kamiokande Preliminary
Super-Kamiokande Preliminary
1144.4 days
1144.4 days
700 e-like 2531
600
<
e-like 576
p-like 2486
1 >
500
400
300
200
100
•00
-15
-10
-5
0
5
10
PID likelihood, Sub-GeV, 1-ring event
15
-15
-10
-5
0
5
10
15
PID likelihood, Multi-GeV, 1-ring event
Figure 1. Particle ID distributions for sub-GeV (left) and multi-GeV (right) single-ring events. T h e histograms show the expected distributions in the absence of neutrino oscillation.
identified in which at least one particle (typically an energetic muon) exits the inner detector. The FC sample is further divided into sub-GeV (E„j s < 1330 MeV) and multi-GeV (Evis > 1330 MeV), mainly for historical reasons.
2.1
Ratio
We characterize an event as either showering (e-like) or non-showering (/i-like) based on the observed Cherenkov light pattern. The particle identification algorithm has been tuned using cosmic-ray muons, decay electrons, and with a test beam study. For both sub-GeV and multi-GeV single-ring events, the misidentification probability is roughly 2-3%. Figure 1 shows the measured and predicted particle ID distributions for subGeV and multi-GeV single-ring data. The prediction is based on a Monte Carlo sample generated by folding the calculated atmospheric neutrino fluxes with a detailed model of neutrino scattering in water, processing a sample of event vectors through a simulation of the detector and the same reconstruction programs applied to the data. As the figure illustrates, the number of singlering e-like events observed agrees well with the prediction, while there is an apparent deficit of single-ring /x-like events.
122
The extent of the fi-like deficit can be quantified by a ratio of ratios R: _
V^-Iike/
e-like/° bseri,ed
v^/i-like/
e-like^ ex P ected
Uncertainties in the absolute flux prediction (which affect both v^ and ve) cancel in this ratio. If the observed flavor composition agrees with expectation, then R = l . For the present 1144 day Super-Kamiokande data sample, RsubGeV = 0.652to^\l(stat)
± Q.Obl(syst)
RmuitiGeV = 0.661±g;g||(s*at) ± 0.079{syst) The large deviation of R from the expected value R = l , combined with the good agreement of the absolute e-like rate with prediction, is a clear suggestion of Vp disappearance. 3 3.1
Zenith angle dependence of atmospheric neutrinos Contained events
An atmospheric neutrino detector receives neutrinos that are produced everywhere in the Earth's atmosphere. Those neutrinos produced in the atmosphere above the detector travel of the order of 10's of km before interacting in the detector, while those produced on the other side of the Earth travel over 10,000 km (see figure 2). Since the production in the atmosphere is expected to be isotropic, we can expect the neutrino flux to be up-down symmetric. In other words, we expect to see as many neutrinos coming up as going down. This expectation is modified to some degree by geo-magnetic effects, especially at low energies. If we plot the observed neutrino events as a function of zenith angle, those with cosQzenith = 1-0 represent downward going neutrinos while those with cosOzenith = -1.0 represent upward-going neutrinos. This analysis will only be valid to the extent that the product lepton faithfully reflects the neutrino direction. The correlation of the reconstructed lepton direction with the neutrino direction is shown in figure 3. As can be seen, the correlation of the lepton track with the neutrino direction is poor below about 400 MeV. In figure 4 the zenith angle distribution of the data is shown for both electron-like and muon-like events for various energy regions. As can be seen, the anomalous muon deficit is a strong function of zenith angle or equivalently the neutrino flight distance.
123
Figure 2. Neutrino production in the atmosphere. The neutrino path length varies with zenith angle
180" 150" 120" 90" 60" w "rji 30"
+++4
0
500
1000
1500
2000
2500
3000
l e p t o n momentum
Figure 3. The opening angle between the neutrino direction and the direction of the reconstructed lepton track as a function of lepton momentum.
124
Multi-GeV e-like 3 a & 200 o
Multi-GeV n-iike + PC
v 100 JO c
0
-1
-0.5
0
0.5
-0.5
COS 0
0
0.5
1
COS 0
Figure 4. The observed angular distribution of the neutrino interactions. expected distributions in the absence of neutrino oscillation .
Also shown is the
1000 rev,, FCv.
750
PCv u
500
2. C0
-O250 o o
g
li ninl 111111111 i mini i mini
0
—
^ 150 c
Up-Stop v, Up-Thru v
s "J 100 50 -2
10
2
-1
10
1
10 10 E„ (GeV)
3 10
4 10
5 10
Figure 5. The spectrum of neutrino energies which contributes to each category of data.
3.2
Upward-going muons
The upward-going muon sample can be used as an independent check of the contained event results. Figure 5 shows the energy dependence of the neutrinos which contribute to each data sample.
125 Zanlth Angla D i a l of Obaarvad Upward Through Going Muon Flux Supar-Kamlokand* (ProAmlnary) ataifallcal'error only
1 -f-
o
T
-*-
i — • — i — • — i — • — i — • — ' — • — i
-1
-0.8
-0.6
-0.4
-0.2
0 cos©
Figure 6. The through-going upward muon flux vs. zenith angle. The solid histogram shows the expected distribution without oscillations.
Upward through-going muons Neutrinos which interact in the rock below the detector can produce muons which pass entirely through the detector. The parent neutrino has a typical energy of 100 GeV. Figure 6 shows the observed zenith angle dependence of this class of events. The fit to the expected distribution is poor with a probability of about 1.2%. Stopping Upward Muons Stopping upward muons are induced by neutrinos with energy comparable to the partially-contained sample. The lower energy of stopping upward muons has two implications. First, the overall suppression expected is larger, since the L/E argument of the oscillation probability is larger. Second, even neutrinos arriving from near the horizon will experience significant oscillation. Both these features are borne out in Figure 7, which shows the flux of stopping upward muons vs. zenith angle. Almost a factor two suppression of the flux is observed. While the absolute magnitude of the expected flux is uncertain, the stopping/through-going ratio can be used (in analogy to R for contained events) to remove the normalization. Since through-going upward muons do not exhibit nearly as large a depletion, the stopping muon deficit is further evidence for the expected energy dependence characteristic of neutrino oscillation.
126 Zenith Angle Dial of Obaarvad Upward Stopping Muon Flux
,--1.5 'h.
«
'at
E u o
Supvr-Kamlokand* (Preflmlnwv) •tatlatlcal«ror only Upward Stopping Muona (>7m) BG subtracted 289.6 «vanta/lt17 llvedaya (Apr.1996 - Nov.1990) . Ave.Flux 0.41±O.03±0.02<x1Q' 1 3 cmVV') Hiatogram:Exp«ctad flux (Bwto4-GRV94) -• • null oaclllatlon a=-0.167, P--0.337, wO.044 X l "K«-35.4/15 l Prob.=0.21%
b 1 V—
,x, X
0.5
4 -0.8
-0.6
f" -0.4
-0.2
0 COS0
Figure 7. Stopping upward muon flux vs. zenith angle. The solid histogram shows the expected distribution without oscillations.
4
Oscillation analysis
The contained data (e-like and /i-like) are binned in lepton momentum and angle. The angular distributions of partially-contained events, stopping upward muons and through-going upward muons are also included. The fit accounts for systematic uncertainties in the overall normalization, v^/ve ratio, flux spectral index, etc. by including additional terms which allow the uncertain quantities to vary within their estimated ranges. With the exception of the absolute normalization, which is allowed to vary freely, each systematic term contributes to x2 if it deviates from 0, but also reweights the expected distributions. The allowed regions from the global fit, for various confidence levels, are shown in Figure 8. The best fit is for Am 2 = 3.2 xlO 3 eV2 and maximal mixing, giving x2 = 135/152 degrees of freedom. The best fit without oscillation (but allowing all systematic terms to vary) gives x 2 = 315/154 degrees of freedom - a rather poor fit. Figure 9 shows the data and the best fits with and without oscillation. For both contained events and upward-going muons, the agreement of the data with the prediction of oscillation is excellent. In addition, for the best fit oscillation value, no systematic term deviates from zero by more than la.
127
• i • ' '
• i ' • • < i '
<E
68% C.L. • 9096 C.L. 99% C.L.
0.2
0.3
0.4
0.5
0.6
0.7
0.6
0.9 _
sin^e Figure 8. Allowed regions for f M «-> uT oscillation based on a global fit t o single-ring sub-GeV and multi-GeV fully-contained data, partially-contained data, and upward-going muons. The regions are la (innermost), 9 0 % confidence level (middle) and 99% confidence level (outermost).
400 .Sub-GeV e-like e !T7
T-I
i i | > i i r | » • i i | i rr-r-
. Multi-GeV e-like 2(H)
j*farT"*"^' t
Multi-GeV u-like + PC
£
4 'Upward-through going u
Upward-stopping U
£ 1
-
4^ 100
^.-+af:
_.-• i
r
i
-1
-0.5
0
cos 0
0.5
1 -1
-0.5
0
cos 0
0.5
1
-1
-0.8 -0.6 -0.4 -0.2 COS O
0
-1
-0.8 -0.6 -0.4 -0.2 COS 0
Figure 9. Zenith angle distributions for contained events (left) and upward-going muons (right). The histograms show the best-fit expected angular distribution with and without neutrino oscillation. The best fit with oscillation (sin 2 20=l.O, A m 2 = 3 . 2 x l 0 3 e V 2 ) gives x 2 = 1 3 5 / 1 5 2 dof, while the best fit without oscillation yields x 2 = 3 1 6 / 1 5 4 dof.
0
• e-like o |i-like "
2 1
10
10
3 10
4 10
5 10
UEV (km/GeV) Figure 10. The ratio of the number of FC data events to FC Monte Carlo events versus reconstructed L/E„. The points show the ratio of observed data to MC expectation in the absence of oscillations. The dashed lines show the expected shape for t y «-» uT at A m 2 = 2.2 x 1 0 ~ 3 e V 2 and sin226> = 1.
4.1
L/E
As a cross-check of the above analyses, we have reconstructed the best estimate of the ratio L/Eu for each event. The neutrino energy is estimated by applying a correction to the final state lepton momentum. Typically, final state leptons with p ~ 100 MeV/c carry 65% of the incoming neutrino energy increasing to ~ 85% at p = 1 GeV/c. The neutrino flight distance L is estimated following Gaisser and Stanev 3 using the estimated neutrino energy and the reconstructed lepton direction and flavor. Figure 10 shows the ratio of FC data to Monte Carlo for e-like and p-like events with p > 400 MeV/c as a function of L/E„, compared to the expectation for v^ •H- vr oscillations with our best-fit parameters. The e-like data show no significant variation in L/Eu, while the /z-like events show a significant deficit at large L/Ev. At large L/Ev, the Vy. have presumably undergone numerous oscillations and have averaged out to roughly half the initial rate. 4-2
Up <-> v'sterile Oscillation
High-energy Analysis The initial Super-K analysis of v^. disappearance 4 did not distinguish between
129 v
n «-> vT and i/M o vsteriie oscillation. Now that we have shown that oscillation is required to explain the data, we study this question with a dedicated analysis optimized for discriminating between the two possibilities. Oscillation to sterile neutrinos has two distinctive features: first, neutral current interactions will be lost if vsteriie is involved, while if vT is produced, it will still induce these reactions. Second, matter effects are applicable to vsterileThe factor sin2 26 which appears in the vacuum oscillation probability for Vp <-> vT oscillation must therefore be replaced with the effective mixing angle in matter for i/M «-> vsterUe oscillation: • 2„„ sin 20m =
sin 2 20 5 =—, ( C - c o s 2 0 ) 2 + sin 2 20
where 2VabEu Am 2 and the density-dependent factor Vab reflects the difference in interaction potential between i>M and vsterile- V06 reflects the fact that v^ undergo neutral current interactions with matter while vsterue do not. In addition, the vacuum oscillation length L must be replaced with its analog: r 4
L
L
y'(C-cos26») 2 +sin 2 26'
When E becomes large compared to (Am 2 /V a b), the C, term in the denominator of the effective mixing angle also becomes large, and drives the effective mixing angle to zero. Thus, matter effects suppress oscillation at high energy. The v^ o vsteriie analysis exploits both these differences. Three data samples are selected: First, a high-energy neutral current enriched contained sample to probe the disappearance of neutral current interactions, second, the partiallycontained event sample with E„jS>5 GeV, which is induced by neutrinos of sufficient energy to experience matter-effect suppression, and third, the stopping upward-muon sample, with neutrino energies comparable to the partially contained sample and therefore also sensitive to matter suppression. For each sample, the two hypotheses are tested by comparing the angular regions most-sensitive to matter effects. Figure 11 shows the actual angular distributions for the samples in question, along with the expected angular distributions for both scenarios. In the case of the neutral-current enriched and partially-contained samples, we test the ratio of upward to downward rates, for the stopping upward
130
J u
l—l *. «
l 4J
l—i • «1 C«i9
l M
l u
l—i H I
,t—!—l -, -«JI . «
l—l -M -M
i « tl»B
l—l B.1 I U
i O.t
l—J U.K I
F I
*•
«
«
-"
* J -W ci«6
•»'
i *J
i *l
Figure 11. Neutral Current enriched (left), partially-contained (center) and stopping upward muon (left) data samples for the usterile analysis. Solid lines show the best-fit expected distribution for Vp <-> vT oscillation, while the dotted curves show the best-fit expected distribution for f M «-• Vsterile oscillation.
muons, we test the ratio of vertical to horizontal rates. At each point in (sin 2 20,Am 2 ) space, expected ratios for both oscillation channels are determined and compared with the measured values. Figure 12 shows this comparison; clearly the u^ o vstertie hypothesis fits the data poorly, while Vy. «->• vT is consistent with the data over a large range of Am 2 , including the value derived from the global fit. Based on the individual samples, a combined x 2 is computed, reflecting the difference in goodness of fit between v^, <-» vT and v^, <-> usterue- Figure 13 shows the allowed and excluded regions. For v^ *-» vT, a large region, consistent with the best-fit value from the global analysis, is allowed. For Vft ** Vsterile, the entire (sin 2 2#,Am 2 ) plane is excluded at 99% confidence level if Am 2 >0. For Am 2 < 0, most of the parameter space is also excluded at 99% confidence level, while a small region excluded at only 90% confidence level remains. Thus Super-Kamiokande data appears to strongly disfavor, if not rule out, v^, <-> Vsterile oscillation as a viable explanation for the atmospheric neutrino anomaly.
References 1. M. Honda et al., Phys. Lett. B 248, 193 (1990); M. Honda et al., Phys. Rev. D 52, 4985 (1995). 2. G. Barr et al., Phys. Rev. D 39, 3532 (1989); V. Agraval et al., Phys. Rev. D 53, 1314 (1996).
:
•
131 upMom. mi.. rfN.C tnridwl MuM-ring . « «
upM..wBralh..rf Men l£ (K„.>5t;«V> PC rvort-
™rtl™l/Wtz..W1u rMU. «t upword U . ™ , * H » r u ««•*•
Figure 12. Comparison of vy, «-+ uT and Uy «-> usterue oscillation scenarios with the data. In each figure, the solid horizontal line shows the measured up/down (for neutral-current enriched and partially-contained) or horizontal/vertical (for stopping upward muons) ratio; the \a errors on the measurement are indicated w i t h dashed lines. The light and dark dotted curves show the expected dependence of the ratio in question on A m 2 for Uy <-> vT and Uy <-> uate.rUe oscillation. Vy •<-» vT oscillation gives a good f i t t o the data near A m 2 ~ (few) x l O - 3 eV 2 in all cases, while Vy <-> vsteTue is inconsistent with the data.
3. T.K. Gaisser and T. Stanev, Phys. Rev. D 57, 1977 (1998) 4. Y. Fukuda. et al. Phys. Rev. Lett. 8 1 , 1562 (1998)
132
excluded region from combined anaIysis(muIti+PC+upu.) 10
.
.
i
.
i
•
i
.
i
.
i
.
excluded 90% C.L. 99% C.L. allowed(FC single ring) 90% CX. 99% CX.
V —»V
-I
i
I
••!
i
!•»
•!••>•
(Am2>0) I'" '
I
-t—^
" I
(Am2<0) I"' 0.7
li,~!» MS
ll.KS 11,'J I I ' 1 ^
1
sin226 Figure 13. (Top) Allowed region for v^ <-> uT oscillation; the area above the dotted (solid) line is allowed at 9 0 % (99%) confidence level. (Center and Bottom) Excluded regions for "/* ++ ^sterile oscillation. The two bottom figures correspond to the different possible signs of A m 2 . In the center figure ( A m 2 > 0 ) , the entire shaded region shown is excluded at 9 9 % confidence level. In the bottom figure ( A m 2 < 0 ) , the area outside the solid line is excluded at 9 0 % confidence level, and the remainder of the area (shaded) is excluded at 9 9 % confidence level. Also shown in each figure is the allowed region from the analysis of FC single-ring events.
A R E V I E W OF N E U T R I N O OSCILLATION S E A R C H AT ACCELERATORS
Physics Department,
SANJIB R. MISHRA Univ. of South Carolina, Columbia, S.C., USA
A brief summary of the existing neutrino anomalies is presented below. The emphasis is laid on the accelerator related efforts to understand these anomalies. An effort is made to detail the systematic concerns in 'establishing a signal' as opposed to 'setting a limit'.
1
Neutrino Anomalies: Three Signals for Oscillations?
If neutrinos are massive, then they should exhibit flavour oscillation. This statement, although not a theorem, is most probably correct. The statement certainly finds its origin from the anticipated deep symmetry between quarks and leptons. It is probable that leptons share two basic properties with quarks: that members of different families have different masses and that the mass eigenstates do not coincide with the weak eigenstates. The existence of these properties requires that neutrino flavors oscillate into one another with a probability given by the familiar formula P("x -> vv) — sin 2 2#i 2 sin 2
/1.27Aro 2 L\
where P is the probability for flavor 'x' to oscillate into flavor 'y', Qn is the weak mixing angle between the mass eigenstates 1 and 2, L is the neutrino flight path measured in km, E is the neutrino energy measured in GeV, and Am 2 = \m\ — m\\ measured in (eV) 2 . Three neutrino anomalies exist 1. We define an anomaly as a measurement that cannot be readily explained using physics processes allowed by the standard model. All three anomalies may be interpreted as evidence for neutrino oscillations. The most compelling of these anomalies is the atmospheric neutrino anomaly. The multiple and mutually corroborative observations by the Superkamiokande experiment are sufficiently strong as to make the oscillation interpretation of these data almost certainly correct. The MINOS experiment aims to have the redundancy and control necessary to check this most important anomaly of the last decade. The other two anomalies are the observations of the solar neutrino deficit and the excess of electron-like events in the LSND experiment. 133
134
The 'standard theoretical wisdom' suggests that the three extant anomalies are one too many; possibly two too many. Fortunately, data from current and future experiments will bring the picture into focus in the next five years. Just any one of these anomalies, unequivocally verified, would constitute a breakthrough in our understanding of fundamental particles and their interactions. We know that neutrino masses are substantially smaller than the corresponding charged lepton masses. If any of the oscillation clues are correct, they are smaller by at least seven orders of magnitude. This large factor is probably related to physics on an ultra-high mass scale, possibly the scale of grand unification 2 ' 3 . Thus the confirmation and study of neutrino oscillations will give us our first window on physics at the grand-unified scale. In the sub sections below, we briefly review the present evidence for neutrino oscillations. 1.1
Atmospheric
Neutrinos
In the summer of 1998 the Superkamiokande experiment reported its neutrino oscillation 'discovery' result. It was further established with improved results in the summer of 1999. Most likely, their observations are due to fM -> vT oscillation; there could be, however, a sub-dominant v^ —>• ve oscillation, the so called evidence for a finite XJ\%. Cosmic rays hitting the earth's atmosphere produce pions, which decay into muons, which in turn decay into electrons. This chain, n -¥ (iv^ -4 eVeV^v^, produces twice as many v^ as ve . Several experiments have measured this ratio by observing the relative number of muon and electron charged current events in underground detectors. With the exception of two early, low-statistics experiments 5 , all of them have observed values of R, the double ratio of n/e observed to (j,/e predicted by simulation, to be substantially below unity 7 . The most precise value of R comes from the recent Superkamiokande experiment 10 , 0.68 ± 0.02 ± 0.05 for the sub-GeV data and 0.68 ± 0.04 ± 0.08 for the fully- and partially-contained multi-GeV data. The status of the R measurements by various groups is summarized in Figure 1. The most impressive evidence for oscillations, however, comes from the zenith angle distribution. This distribution is relevant because events from overhead have neutrino flight paths of tens of kilometers, whereas events from the far side of the earth may have neutrino flight paths as long as 10,000 kilometers. Thus for the proper combination of mass difference squared and energy, the overhead events may not have sufficient flight path to fully develop the oscillation probability, whereas multiple (rapid) oscillations occur for events coming through the earth. Figure 2 shows the "ten-bin" angular
135
A Decade-pfus of Atm. Flavor Ratios - the Latest: (Preliminary)
Water Cherenkov 52 kton * yrs.
iSuperKamiokande:
R^V,
0;68;i.0,04db 0.08
,nng
(Tc^f)
Systomatics - hmstee
iSoudan 2: R^/e W
Iron Tracking Calorimeter 4.6 kton - yrs.
« 0-68 ±'0.11 * 0.06
trfc, shwr
/ Mostly
(sMb°f||v)
A Statistics •• limited —IF-
-~-r~- -T— ft B
•
m-* i::4
fi w
•'-'•"•|v-<->-
-
THE ATMOSPHERIC NEUTSppRAT© ? f ^ ) S w , %
-
/*}MO
•
-
.
'-. „ . * • „ • T „ „ + ,y * . T ..+>: + , : j , . + -+ +.
+ **.• +
*,
bt :l f
+,«*.*
<
' ( • •
i;
j-f'-j
•
i"~""
•
:
0:iS"
»
* 1T
t
Me(*H^'
# „„„,-a™„..„......
_
Figure 1. A compilation of the double ratio, R, measured by various experiments10.
136 distributions for multi-GeV Superkamiokande data. The electron-like data agree well with the absence of oscillations, but the muon-like data deviate strongly from this hypothesis. They agree beautifully with the hypothesis of v^ - • uT oscillations with sin 2 20 = 0.99 and Am 2 = 3.1 x 10~ 3 (eV) 2 . Fits to the Superkamiokande data give 90% (99%) confidence regions for i/M ->• vT oscillations of 2 x 10~ 3 < Am 2 < 7 x 1 0 - 3 (eV) 2 ( 1 x 1 0 - 3 < Am 2 < 9 x 10" 3 (eV) 2 ), as shown in Fig 3 10 . The Superkamiokande experiment is able to rule out the possibility of v ii ~*• vs oscillations at the 99.9%C1., where v8 represents a sterile neutrino. The MINOS sensitivity, via NC/CC remeasurement in near versus far detectors, will be far superior to the Superkamiokande sensitivity to v^ ->• va Superkamiokande can also rule out that the oscillation comes entirely from Vp —• ve , but the possibility of a minor contribution from v^ —• ve is still an open question. This search for the subdominant oscillation mode is one of the outstanding challenges of the MINOS experiment. Confirmation and further investigation of these effects will be best done by accelerator experiments. The first of these will be the K2K experiment, which took its first significant data in 1999-2000. The K2K experiment could establish the Superkamiokande observations as due to oscillations, but K2K lacks the precision to 'kill' — exclude at the 99% confidence level — the effect. The much higher statistics MINOS experiment, the subject of this proposal, should be able to do the definitive check of this claim. This is detailed in a section below. 1.2
Solar Oscillations
Five experiments have measured the flux of solar neutrinos 17 , 18 . All have reported results that are a fraction of the rates calculated from models of the sun 24 as shown in Table 1. There are good arguments that the deficit of solar neutrinos cannot be due to a miscalculation of the predicted rate. The 8 B rate cannot be off by more than a factor of two because the Kamiokande and Superkamiokande experiments observe these neutrinos. The 7 Be reaction is higher in the solar chain than the 8 B reaction and has a softer temperature dependence. Thus, even if the 8 B were suppressed by a factor of two, the 7 Be reaction would be less suppressed. Finally, more than half the rate observed by the GALLEX and SAGE experiments comes from the primary pp reaction, which accounts for the vast majority of the energy output of the sun. These results taken together are inconsistent with the low rates observed in the Homestake experiment and the two gallium experiments.
137
Zenith Angle -10 bins: Vy, depletion at all angles including above-horizon 300
200
100
-1
0
1
o Sub-GeVj-i-like
1
-1 0 t Muiti-GeVp-iike+PC
Sub-G«v e-iike
i
200
-1
0
Multi-GeV e-like V e - distributions 'Well-behaved" No overt hints for sub-dominant V^ <-» V e .
With increasing £ v , V^-depletiort moves to (mostly) below-horizon,
Figure 2. Angular distribution of sub- and multi-GeV e-like and /i-like Superkamiokande atmospheric events 1 0 . For muons, the upper curves represent expectations for no oscillations; the lower curves represent the best fit for v^ -> uT oscillations with sin 2 20 = 0.99 and A m 2 3.1 x 10 3 (eV) 2 . The electron data are consistent with a null oscillation hypothesis.
138
Allowed Region: FC & PC Events ^
^
> & <X
_
™™
; i-
10
„
• .
. — —
,
S«n«r Kamiokande Prtlimimttv 848 da s
Vn *~> v T
y
r(Osc. Best Fit) = 55/67 d.o.f X~(NuN Osc.) « 177/69 d.o.f
10: I
,o1 68% C.L 9 0 % CL<
99% C.L ;
io"*!
0.2
0.4
0.5
0.S
sin228
sin 2 26 = 0.99 Am2 = 3.1 x lo^-fev 2 ] a = 0.05 Figure 3. Confidence limits set by Superkamiokande for Vy, —• vT oscillations allowed region is to the right of the curves.
10
.
The
139
Experiment
Reaction Threshold
GALLEX i/e + 7 1 Ga ->• e" + n Ge 0.233 MeV
19
SAGE
2U
Homestake 21
Kamiokande 22
Superkamiokande
23
i/e + 3 7 CI ->• e - + 3 7 Ar 0.8 MeV uee~ -> uee~ 7.3 MeV vee~ —> f e e~ 6.5 MeV
Calib. 1.04 ±0.12 0.95 ±0.14 None
N/A
Predicted Sensitivity 53% pp 28% 7 Be 12% 8 B 7% CNO 2% pep 13% 7 Be 79% 8 B 6% CNO 100% 8 B
Measured/ Predicted 0.59 ±0.06 0.52 ±0.07 0.33 ±0.028 0.54 ±0.07 0.475 ±0.015
Table 1. Solar Neutrino Results
While there is a possibility that the chemical experiments do not understand their sensitivity, both the gallium experiments have been calibrated on neutrino sources and the Homestake experiment has had a series of systematic checks. If the solar deficit is caused by neutrino oscillations, there are two possible interpretations. One is that they are simple vacuum oscillations 25 . In order to get a value as low as that reported by the Homestake experiment, a small multiple of the oscillation length has to match exactly with the earthsun distance. This leads to mass differences squared (Am 2 ) of about 10~ 10 (eV) 2 and maximum mixing angles. The second possibility is that matter oscillations are taking place in the sun 26 . Matter oscillations have the ability to convert completely ve to v^ in a triangular region of the sin2 29 versus Am 2 plot. Approximately 50% conversion will occur near the edges of the triangular region. Since the different experiments are sensitive to different energy neutrinos, points at the intersections of the appropriate triangles are candidates for accounting for the observed pattern. There are three such places, both with Am 2 equal to about 10~ 5 (eV) 2 . Two solutions have close to maximal mixing, and the other has sin220 equal to about 10~ 2 27 . Additional information that could confirm neutrino oscillations as the source of the solar neutrino deficit will come in the next few years from Super kamiokande and the Sudbury Neutrino Observatory (SNO) 2 8 . Both Super kamiokande and SNO will be able to measure the distortion in the 8 B
140
energy spectrum. This distortion is predicted to be different for each of the possible solutions, so that both sin2 26 and Am 2 will be determined. The SNO detector will have lkT of heavy water surrounded by a light water veto. It will be able to measure the distortion of the 8 B spectrum with more sensitivity than Superkamiokande by measuring the charged current reaction from deuterium, ved-t ppe~. It will also be able to measure the neutral current (NC) breakup of deuterium, which is identical for all (nonsterile) neutrino species. It will thus verify the solar model calculations independent of neutrino oscillations. 1.3
The LSND
Experiment
The Liquid Scintillator Neutrino Detector (LSND) at Los Alamos studies neutrinos from pions stopped in a water target. The chain of decay is ir+ -> / U + ^ -> e+veuIJ,i'fi. The experiment then looks for V^ ->• Ve oscillations, where the Ve is detected by observing vep -> e+n, followed by a 7 from np ->• dj. Updated results from the experiment report 22 events observed with a calculated background of 4.6 ± 0.6 events, 31 consistent with, but much more statistically significant than, their earlier results 32 . Figure 4 shows the allowed region for the LSND experiment and the excluded regions from other experiments. At high Am 2 , the LSND allowed region in sin22# is from 3 x 1 0 - 3 to 12 x 10~ 3 . For maximal mixing, the allowed region of Am 2 is between 0.05 and 0.09 (eV) 2 . Much of the allowed region for LSND has been excluded by other experiments. 33'34>35 However, this situation needs clarification, especially in the region around Am 2 of about 6 (eV) 2 , where the KARMEN experiment has a dip in its sensitivity 3 5 . We will discuss the sensitivity of the NOMAD experiment to this region below. None of the current experiments can rule out LSND. One future approved experiment, MiniBOONE at FNAL, will fully check the LSND anomaly. This is briefly discussed in the section below. 2
Cosmologically Significant Neutrinos: v^ —• vr ?
Confluence of three independent motivations, largely based upon conjectures, led us in NOMAD to search for i/M -+ vT oscillations in the large Am 2 and small mixing angle region. First, the measurement of the rotational velocity of galaxies poses a puzzle: it implies that there exists a large fraction of the mass which is dark. An attractive candidate for non-baryonic dark matter is neutrinos with a small but finite mass. Such stable massive neutrinos will contribute to the dark
141
>
10 : <
10
-1
\-
10
-3
10
10
-1 2 1
sin 26 Figure 4. Allowed region for FM to Ve oscillations from the LSND experiment and excluded regions from the BNL E-776, Bugey, and KARMEN experiments.
142
matter of the universe. For a value of the Hubble constant in the middle of its allowed range, a neutrino of mass m will contribute a fraction of about m/(30 eV) to the critical mass of the universe. Second, interpretations of the COBE results 3 e on the anisotropy of the cosmic microwave background, combined with measurements on the clustering of galaxies 37 indicate that the dark matter of the universe cannot be completely composed of cold matter - some of it had to be relativistic at the time that it decoupled from thermal equilibrium 3 8 . Of all presently known particles, only neutrinos, and most likely the vT, are candidates for this part of the dark matter. (Such a mass for vT can be "predicted" from fits to the solar neutrino deficit and the see-saw mechanism 4 .) Third, the interpretation of the solar neutrino deficit as due to matter oscillation indicate that v^ —> ue mass difference square is about 10 _ 5 (eV) 2 . The see-saw 'recipe' and the mass-hierarchy 'prejudice' (M„T » MV>1 » MVJ, indicates that the v^v,, mass difference (= \/lO _ 5 eV) implies that M„ r should be around lOeV. We emphasize that none of the above arguments was compelling. It was the confluence of the three, indicating a massive vT around lOeV, which motivated the the v^ -> uT oscillation search at the CERN SPS by the CHORUS and NOMAD experiments. The purpose of the CHORUS and NOMAD experiments at CERN was to increase the sensitivity of v^ —> vT oscillation searches in this region by an order of magnitude in sin2 26. The two experiments were in the same beam line, a double-horn focused beam created by 450 GeV SpS protons. The average neutrino energy was 24 GeV (with an average interaction energy of 43 GeV). Both experiments were located about 800m from the target. The CHORUS experiment 39 used the "traditional" method of identifying r's by searching for a kink from the r decay in photographic emulsion. The emulsion target mass was 800kg. Events were selected for scanning by kinematic criteria. Momenta of tracks were measured by fiber trackers before and after a pulsed magnet. Energies were measured by a fine-grained electromagnetic and hadronic calorimeter (lead-scintillating fibers). Muons were identified and measured in an iron toroidal spectrometer. CHORUS ran from 1994 through 1997. CHORUS has published results for about 75% of its total data; it reports no signal and an upper limit on the probability of v^ -> vT of 3.5 x 10~ 4 . The ffj, —• vT search conducted by the NOMAD experiment was also
143
negative. The 90% upper limit on the probability of fM -¥ vT oscillation was: P{vn -* vT) = 1.7 x 1(T 4 Although a negative search, this is the most sensitive of neutrino oscillation searches. We outline the salient aspects of this search in some detail below. 2.1
The u^ -> vT Search Conducted by the NOMAD
Experiment
Neutrino Oscillation MAgnetic Detector (NOMAD) is the only high-statistics neutrino detector to measure precisely the four species of neutrino: i/M V^, ve and Ve. It would have, if it could have, measured the vT interactions. It was designed for the vT detection. The NOMAD experiment used kinematics to identify events with r leptons from neutrino oscillations. For example, when a r lepton decays into a charged lepton and two neutrinos, there is missing transverse momentum in the general direction of the charged lepton. Backgrounds arise from normal charged current (CC) events in which apparent missing transverse momentum is generated from hadrons in the events that are either mismeasured or missed by the detector. Neutral current (NC) processes can also create backgrounds if a lepton or apparent lepton is produced at large transverse momentum to the hadronic shower. These processes are illustrated in Figure 5. For this kinematic technique, it was imperative that the detector was able to measure all particles as well as possible. NOMAD was designed to be an "electronic bubble chamber" enhanced with excellent electron detection and calorimetry. The outstanding features of NOMAD, in contrast to other high-statistics neutrino experiment such as CCFR, CDHS, CHARM, and FMMF — and, more recently, the MINOS experiment — are: 1. Electron Identification: NOMAD was optimized for the detection and identification of electrons. The measured electron energy resolution is about 2%. There were five independent detector systems that aided in the identification of electrons. Electrons could be identified by: • having increasing curvature in the drift chambers; • having a large pulse height in the TRD; • having a large pulse height in the preshower;
pt miss >
£
j
hadrons £ ,1
hadrons remiss
i
hadrons Figure 5. Signal and background processes for leptonic T decay. The diagrams to the left show momentum vectors in the transverse plane. In the case of neutral current background, a hadron is misidentified as a lepton.
145
• having the proper amount of energy and transverse shower shape in the electromagnetic calorimeter including a vertical tail from bremsstrahlung in the drift chambers; and • having no energy in the hadronic calorimeter. Each of these detection techniques had an independent rejection factor for pions of between 10 _ 1 and 1 0 - 3 . 2. Precise Charge Hadron Measurement: Very little multiple scattering in the magnetized volume enabled a precise measurement of the charged tracks. 3. Precise Photon and V° Measurement: Photons are measured via their conversion in the DC volume, and in the PRS-ECAL. Mesons like 7r° and r\ were readily reconstructed. These data have yielded the largest (by a factor of 20) sample of reconstructed K®, A, and A produced in ^-interactions. 4. Low Energy Threshold: NOMAD can identify particles down to about 100 MeV. A difficulty for the NOMAD concept was the leakage of low energy neutral hadrons such as neutrons and K ^ ' s . For the v^ —» vT search, however, this would induce a missing energy in the 'hadron' direction whereas the signal would produce the missing energy in the 'lepton' direction (due to the neutrinos from the tau decay). The 1995 analyses were learning experiences for the NOMAD collaboration. The process led to a consensus on how these analyses should be done, which will be discussed further below. One of the consensus items was that it was necessary to do blind analyses for oscillation searches. This means that one does not look at the signal data until the analysis is settled. This process is necessary, otherwise biases too easily creep into a search seeking a handful of events among a sample of two millions. Subsequently, the collaboration has decided upon all the subsequent searches, e.g. v^ -> i/e, searches for heavy neutrinos, etc., to be blind. The first results on i/M —> vr using blind analyses were published in the Physics Letters 14 . A third, and the final paper incorporating the entire accumulated data has been published in 12 . The results presented in this proposal are from these final analyses. The principal features of the v^ -> vT oscillation search are as follows:11 "Similar steps were also adopted for all other searches, oscillations or otherwise.
146
1. D a t a Simulator: Since these searches are sensitive to the details of the fragmentation of the hadronic jet, Monte Carlo simulations cannot be relied on to provide an accurate background estimation. Therefore, whenever possible, the analyses must use a "data simulator" (DS) to estimate the backgrounds and efficiencies. The data simulator is made by starting with a u^ charged current (CC) event and then replacing the identified muon by another lepton L If I is a neutrino, then a neutral currents (NC) is simulated; if I is an electron, then ve CC events is simulated; and if I is a r, then r CC events is simulated. To achieve the proper charge correlations, normally the NC data simulator will employ some mixture of events with positive and negative T decay candidates. In all cases, a correction is made for errors in the DS by simulating the DS in the Monte Carlo to obtain what is called a "Monte Carlo simulator" (MCS). The efficiencies and backgrounds are then calculated by the formula € =
CMC
X
£DS
.
Z-MCS
To the extent that effects factorize, this formula corrects for both the deficiencies of the Monte Carlo and the data simulator. For example, the Monte Carlo may not have the proper hadronic fragmentation, and this is corrected by the DS/MSC ratio, while the DS is biased by using only identified muons and by charge correlations, and these are corrected by the MC/MCS ratio. 2. Blind Analysis: The analysis must be blind. This means that a "box" must be defined containing possible signal events, and this box may not be opened until after the analysis-protocol is complete, i.e. the estimation of background, efficiencies, and errors are complete. It is not permitted to open the box until after the completed analysis is presented to the collaboration and the collaboration agrees that the analysis has met all of the criteria listed here. Once the box has been opened, the analysis may not be changed, except to correct clear errors, such as the inclusion of a run in which the detector was known to have been malfunctioning. 3. Consistent Control Sample: Since V^v^ event ratio is 0.025, almost no X>M —• VT is expected. Thus the positive candidates in r-search must be background. The positive sample provides a powerful control sample. Before opening the box, an analysis must show agreement between background calculations and two sets of data: data with negatively charged
147
candidates outside the box, and positively charged candidates both inside and outside the box. The probability of neutrino oscillations in these two samples is at least an order of magnitude smaller than that for the signal events. This is by the construction of the box for events outside the box, and by the small fraction of v^jv^ interactions (w 3%) for the positive events. 4. Removing Bias in Efficiency and Background Estimates: Separate samples of Monte Carlo and data simulator events must be used to set the cuts, or build the likelihood functions, and to evaluate the efficiencies and backgrounds arising from these cuts and functions. Failure to do this results in overestimates of efficiencies and underestimates of backgrounds. The search was conducted in four r decay modes: euv, ir(mr°)u, pv, and 37r±(n7r°)fc'. For each mode, two separate analyses were conducted: using the deeply inelastic scattering events (DIS) characterized by large hadronic energy and using the low multiplicity events (QEL) comprising quasi-elastic, resonance, and low-Y^j DIS events. The T -¥ e was the most sensitive channels for the v^ ->• vT search followed by r -¥ 1-prong which includes the r -+ p channel. The final v^ ->• vT analysis is summarized in Table 2. A total of 45 candidate signal events were observed, compatible with an expected background of 48.9 events. At the first glance the table gives the impression that the number of background events is quite large. We should point out, however, that in each channel, whether DIS or QEL, the search sensitivity was totally dominated by the likelihood bins where the predicted background was < 0.5 events. Thus, where as in the DIS r —> e channel the Table 2 shows a predicted 5.3 background versus 5 observed events, the best likelihood bin in this channel dominating the search sensitivity predicted 0.5 background versus 0 observed events. The total predicted background in all the decay modes (DIS+QEL), for the most sensitive likelihood bins, was estimated to be 1.62J;J;|g; the observed number of events was 1; the corresponding expected tau-signal, assuming full mixing, was 8215. The 90% upper limit on the probability of i/M —> uT oscillation was: P(v» -> vT) = 1.7 x 10~ 4 The corresponding sensitivity was 2.5 x 1 0 - 4 . The probability of getting a value of the limit smaller than the sensitivity is 39%. The excluded region in the Am 2 -sin 2 26 plane, along with published limits from other experiments, is shown in Figure 6. A 90%confidence level high-
148
Decay Mode T -»• e
DIS/QEL DIS QEL
T —> 7T,T —• p
DIS-7T
DIS-p DIS-Both T —>• 7T,T - > p
T
—t 3TT
QEL-TT
QEL-p DIS QEL
Obs. T~ 5 6 5 7 1 5 7 9 0
Bkg. 5.3r 5.4±0.9 6.8±2.1 9.5±2.5
o.o67g
6.7±2.3 5.2±1.8 9.6±2.4 0.4+ B
e(%) 3.6 3.6 0.70 1.04 0.07 0.84 2.00 1.9 2.0
Br(%) 17.8 17.8 49.8 49.8 49.8 49.8 49.8 15.2 15.2
jyMax
4110 859 2022 3307 210 357 458 1820 108
Table 2. Summary of v^ —> vT Analysis: r
Decay Mode r —> e
DIS/QEL DIS QEL
r —• 7r,r —• p
DIS-7T
T —> 7T,T —> p
QEL-TT
DIS-p
T —•
3TT
QEL-p DIS QEL
Obs. T + 9 3 19 6 19 21 6 14
Bkg. 8.0±2.4 2.2±0.5 16.0±4.0 5.6±1.5 21.9±6.4 22.2±6.6 6.9±2.5 11.0±4.0
Table 3. The Control Sample of v^ —• vT Analysis: r+.
mass limit of sin220 < 3.4 x 10~ 4 was set. At sin 2 20 = 1, values of Am 2 greater than 0.7 (eV) 2 were excluded. The v^ -> vT analysis is easily extended to ue -» vT search. The relative abundance of ve to i/M is about 1%; the limit on the ve -» vT process, thus, is about a factor of 100 worse than for the v^ -¥ vT . The resulting limit is shown in Figure 7. It should be noted that these NOMAD limits on the v^ -*• vT and ve -t vT processes are the most stringent to date.
1—r i i i i i n
1—i i i i n i |
1—i i i i i i i |
i i i n m
0 r
>
1
E531 \ i
-
~
*'
(SORUS |(
<
10'
( I rNOMAp
1 )CCFR ;/
10 r-
^ s \
•
*•
"
'''•^x>
*'•.
-
' ' 0 \ X *'». vxlx •• x X *< *xVx ;
1 r
r :
V ->V
V x 90% CL 1
10
10
••. •
X ^ > x > • ""\\
\
xx^ x X: CDHS X x ; X
1
10
i
i
I I I nil
n
10
sin 29 Figure 6. The region of the Am2-sin22i9 plane presently excluded by NOMAD for u^ oscillations at 90% confidence level and limits published by other experiments.
150 1
1
1—e—i—r—TT
T7
1
1
1—l—l
l l
i
i
i
i
M
>y[
(N
<
10'
• CCFR
10
v,->vT e
x
% C.L. i
10
i
i
CHOOZ i
i
i
i
i
10
j
i
i
i
1
sin2 26 Figure 7. The region of the Am2-sin220 plane presently excluded by NOMAD for ue —• vT oscillations at 90% confidence level and limits published by CCFR and CHOOZ.
151
3
Search for i/M —> ve Oscillation: Checking LSND
As discussed above, the LSND experiment 32 at Los Alamos has reported evidence for v^ —> ve oscillations, and has maintained this claim as it has added statistics in subsequent runs. NOMAD is sensitive to a significant portion of the parameter space corresponding to LSND's signal. An outstanding challenge for NOMAD is to be sensitive to the oscillation for values of Am 2 reaching down to the 6 (eV) 2 region at sin220 around 0.006. It should be recalled that at Am 2 = 6 (eV)2 the KARMEN experiment has a dip in its sensitivity 35 . The KARMEN experiment, or any of the current experiments, lacks the sensitivity to conclusively rule out the LSND allowed region. Since many aspects of the fM ->• ve analysis in future experiments, such as MINOS or ICARUS, will have features common to NOMAD, we discuss this analysis in some detail below. We pointed out earlier that NOMAD is the only high-statistics detector which precisely measures the ve and v^ induced events. The charge of the leading lepton tags the corresponding anti-neutrino events. Furthermore, the kinematic separation between lepton and hadron lets one identify the 'prompt' (i/-CC) from the 'non-prompt' (lepton is one of the products of the hadronic jet) interactions. The NOMAD analysis relies on the ratio of the ve and v^ fluxes as a function of energy. Because the ve and v^ fluxes peak at significantly different energies, u^ —> ue oscillations can produce a detectable distortion of the v^/ve distribution. The advantage of using the v^/ve flux ratio is that many systematic uncertainties, such as errors in simulation of the jet fragmentation, cancel in the ratio. The search is conducted as follows: • Conduct a "blind" analysis. That is, do not examine the measured e~//J,~ ratio as a function of total energy (Ev) until the analysis-protocol is complete. • Identify e^ and ^ events. Divide the sample into prompt — leading lepton kinematically isolated from the hadronic jet — and non-prompt — leading leptons 'along' the hadronic jet — subsamples. No oscillation signal is expected in the non-prompt region. This provides a calibration of the nonprompt background. • Fix all the backgrounds, having calibrated them against the control sample from NOMAD or other experiments. Since it is a blind analysis, the central values as well as errors of the backgrounds must be determined before examining the observed ratio e~ j\x~~ as a function of Ev. • The observed e+ /n+ ratio must agree with the prediction. • Form the ratio of e~/[i~ events as a function of the visible energy.
152
• Compare the data with the Monte Carlo simulation of the expected background comprising the inherent ve beam-induced events (prompt) and the hadron jet-induced events (non-prompt). The signal for an oscillation will be an energy dependent excess in the e~j[i~ ratio. Five systematic considerations hold the key to a sensitive search for v^ -> ve oscillation. • The beam ve/v^ composition: The inherent ve content of the CERN SPS neutrino beam constitutes an irreducible 'prompt' background — prompt since the leading lepton ( e - or e + ) is kinematically distinct from the particles emerging from the hadronic jet. This is the most importnat background. The beam ue content with respect to the u^ must be accurately determined. There are no theoretical guidelines or constraints on this quantity. It must be determined using the existing data. The precision of the constraining data determines the precision of the v^jv^ ratio. Determination of the relative neutrino flux entails obtaining an accurate determination of the production cross sections of 71^, K^, and K\ — the dominant sources of neutrinos — in the p-Be collision. The constraining data to do this are: u^ FM and Ve spectra measured in NOMAD, and the secondary particle production data by Atherton et al. 4 7 , and more importantly, by the SPY collaboration 46 . Using a sample of 300,000 v^ quasi-elastic-like (QEL) events — these are low hadronic energy CC — in NOMAD, the production cross-sections of 7r/K are empirically fitted with the added constraints of the Atherton and SPY measurements. The agreement between the data and the empirically parametrised (EP) flux predictions is shown in Figure 8. The agreement, over four and a half decades in number of events and Ev up to 250 GeV, shows that overall the K+ contribution relative to the 7r+ is well constrained by the i^-CC data. This comparison on a linear scale, with Ev up to 150 GeV, is also shown in Figure 9. The agreement between the radial dependence of the data and prediction is presented in Figure 10. The radial distributions distinctly show the varying K/7T contribution to the u^ flux. For the i/M -> ue search, however, the critical parameter is the K+/ir+ ratio in various energy bins, and as a function of transverse momentum. Whereas the measured v^ data determine with high precision the K/it ratio for energies above 50 to 60 GeV, at lower values, the K/-K ratio is determined to a precision of about 7% from these data. The 33,000 i/M QEL events above Ev > 50 GeV constrain the integral of the K+ /TT+ ratio to 0.6% precision. For Ev < 50 GeV, however, the kaon contribution is completely overwhelmed by the pion contribution in the i^-spectrum. This is where the direct 7r/K measurements in proton-Target (Be), at energies below 50 GeV, become im-
153 NuMu Flux Data(Symbol) -vs- Prediction(Histo)
10'
10
10
10
100 Data 000-130 Final
200
Figure 8. The energy spectrum of quasielastic-like (low v) i/M interactions (solid circles) compared with the prediction histogram. The x-axis shows the neutrino energy in GeV; the y-axis shows the number of events. The contribution of 7r+ and K+ to the i/M are also shown.
154 NuMu Flux Data(Symbol) -vs- Prediction(Histo) 40000
Entries Mean RMS
308000 25.17 23.96
30000
20000
10000
50 Data 000-130 Final
100
Figure 9. The energy spectrum of quasielastic-like v^ interactions (solid circles) compared with the prediction histogram. The contribution of 7r+ and K + to v^ are also shown.
155 NuMu Flux Radial Data -vs- Prediction
w4
1
1
Enlrla Mean
ms
80633 27.20 23.95
|
' -
•
1\ 10 3
"
102
j
\
^J \
10
u
':, \ 1
200 Data 000-030 Final
200 Data 030-065 Final 1
w4
RMS
w3 102
10
1
1
Entries
89195 23.11 23.86
|
'.
-
v*v f
'll
^ \
-
\
kL j
\ ", i
200 Data 065-090 Final
200 Data 090-120 Final
Figure 10. The energy spectrum of quasielastic-like v^. interactions (solid circles) compared with the prediction histogram in four distinct radial bins. Separate contributions of 7r+ and K+ to i/M, in each radial bin, as well as the sum, are also shown.
156 Ratio: K+/rr
o
•
M
C/3
*%l
W
• N
3 irical Par ametrizatio
'Athei•ton et al. Dati
a
p>
*& CJi CD
l_.
i? § "
+
M I*
« •N I
"
M
<:
o y M !3
o
m
3o
H
^
-s
CD
l\3
o o
1
|
M
•
n
•
1
|
R2
Figure 11. The K+/x+ production ratio in the forward direction as a function of meson energy. The solid circles represent the predicted ratio and open squares with error bars represent the SPY data 4 6 and Atherton 4 r data.
portant. The required constraint on the K+/ir+ ratio in the Ev < 50 GeV region conies from the SPY data. The NOMAD EP predictions of the K/ir ratio as a function of energy at pt = 0 are compared with the SPY/Atherton data in Figure 11. Figure 12 shows the corresponding comparison at various values of transverse momentum up to 500 MeV. The predictions agree well with the data. Figure 11 and Figure 12 show that the empirical parameter predictions agree well with the SPY measurements down to meson energies of 10 GeV.
157 Ratio: K+/TT
+
Figure 12. The K+/ir+ production ratio as a function of meson transverse momentum. The solid circles represent the fit with empirical parameters and open squares represent the SPY 4 6 and Atherton 4 7 data.
The precision, statistical and systematic, of the SPY data points is at the 1.5% level. An additional 1.5% uncertainty comes from the fit procedure. Thus, the error on the K+/ir+ ratio in the region Ev < 50 GeV is known to about 2.3% level. These errors on K+ /TT+ ratio are directly transported to the beam vtlv^ errors. The 7r~/K~ spectra are similarly determined using the NOMAD 17^-data and corresponding SPY measurements. Remarkably, the K\ contribution to ve can be determined using the
158 Ve data. The Ve flux is comprised of contributions from K^, K~, fj,~, and charmed hadrons. The dominant contribution (70%) to Ve comes from the Kl; the next contribution, K~, is already constrained by the 7^-data. The charm/hyperon contribution is parameterized from fixed target charmproduction experiments. Thus, the IVdata offers a measure of the K\ content of the beam. The K\ shape is further constrained by the measurement by Skubic et al. 15 • The low-energy events: The efficiency for the reconstruction of the lowenergy electron/muon events must be high. At lower visible energies (< 20 GeV) the ve/vn ratio is small (typically < 0.006). Furthermore, good oscillation sensitivity at low Am 2 requires reconstruction of low-energy events. • The non-prompt background: The hadronic shower, primarily via the asymmetric conversion of a photon from 7r° decay, can produce an electron (e^) signal. This background, referred to as the non-prompt background, in contrast to the prompt ve-induced background, is largely reducible via cuts on kinematic isolation variables. The amount of the background, however, must be determined from the data themselves as Monte Carlo estimation of this background is not reliable. The ability of NOMAD to distinguish e~, where the signal is expected, from e + , where no signal is expected, is critical for the determination of the background. • The electron versus muon reconstruction efficiency: The reconstruction of the electron (e*) versus muon (ji*) must be understood with high precision. The electron identification efficiency In NOMAD, using the large control sample of 7 conversions, is measured with a precision of 0.4%. • The i/-energy scale: The lepton energy is well matched in the data and Monte Carlo. The scale of the hadronic energy in the data and Monte Carlo were calibrated using the v^ CC. The NOMAD analysis, then, strongly constrains the ve flux and the nonprompt background. The ve flux comprises contributions from K+, K%, fj,+ , and charmed hadrons. Thus, all the sources of ve are constrained by the above three neutrino flux measurements, and the NOMAD experiment has a unique prediction for i/eThe expected precision on the v^jv^ flux ratio is detailed in the Table 4. The cumulative precision on the beam Ug/v^ ratio is about 1.55%, and this will permit a sensitive search for uM —• ve oscillations. Table 5 presents the preliminary estimates of the leading systematic errors in the v^ —> ve oscillation search. The NOMAD search, however, lacks the sensitivity to fully check the LSND result. The future experiment MiniBOONE at FNAL aims to do this. It is a single detector experiment which aims to determine its background — as oppose to measure it in a near detector — precisely, and compare it with
159
Sources
<EV>
K+
K+
H Others (Charm, A / S , etc.)
45.50 30.41 21.54 50.78
GeV GeV GeV GeV
Contribution
Error
0.6643 0.1465 0.1558 0.0335
2.06% 0.73% < 0.3% 0.34%
Table 4. The Sources of Relative ve Flux and Errors. The average energy and the fractional contribution to the ve flux from various sources are given. An estimate of the corresponding error in the ve / v^ ratio is presented in the last column. The number represent the central value of the empirical parameterization prediction tuned to fit the control samples.
Sources
Error
Beam Non-Prompt (Syst.) e//i Efficiency Energy Scale
2.5% 1.0% 0.4% «1%
Table 5. A preliminary estimate of the sources of systematic errors in fM —• ue search.
the observed signal. If the search is positive, the program will evolve into a definitive two detector experiment which is BOONE. 4
Future Endeavours to Check the Atmospheric Anomaly
The recent atmospheric neutrino results almost certainly suggest (imply?) that neutrino oscillations exist in the Am 2 region of 10~ 2 to 1 0 - 3 (eV) 2 and at large values of sin2 29. Efforts are underway at FNAL and CERN to check the most important of neutrino anomalies. A summary of the goals of the MINOS experiment is presented below. The goals of the MINOS experiment are to verify the atmospheric neutrino results, to measure the Am 2 value or values precisely, and to determine the extent to which muon-type neutrinos oscillate into r-type neutrinos, electrontype neutrinos, and sterile neutrinos. The experiment will be sensitive to
160 Vn -*• vT oscillations for values of Am 2 above 6 x 10~ 4 (eV) 2 at sin2 28 = 1, and for values of sin220 above 2 x 10~ 2 at high Am 2 . Given the current best value of Am 2 from the Superkamiokande experiment, the MINOS collaboration has decided to conduct its first run with the Low Energy (LE) option. Possibility of a Medium Energy (ME) or High Energy (HE) beam is built in the NuMi design. 4-1
Overview of MINOS
The detail description of the MINOS experiment is found in the Technical Design Report 5 2 . We will only briefly summarize the main aspects of the experiment here. In order to keep systematic uncertainties as small as possible, the MINOS experiment will compare neutrino induced events in two detectors, one on the Fermilab site approximately 1 km from the target, and the other on the lowest level of the Soudan mine in northern Minnesota, 736 km from the target. (See Figure 13.) For neutrino oscillations in the Am 2 range of interest, the near detector is used to predict the rate and character of neutrino interactions in the far detector. Neutrino oscillations will manifest themselves by deviations from these expectations. The neutrino beam will be derived from the 120 GeV proton beam from the Fermilab Main Injector, which is expected to provide 4 x 10 13 protons per bunch onto a segmented 160 cm-long carbon target. The pions and kaons from the target will be focused toward the detectors by two magnetic horns plus possibly a device known as a "hadronic hose," to be discussed in more detail below. By varying the currents and locations of the horns, different hadron momenta can be focused yielding beams of different neutrino energy. Three such beams have been designed, as shown in Figure 14. The optimum beam depends on the value of Am 2 . The values being reported by the Superkamiokande experiment indicate that they will want to use either the low or medium energy beams. The baseline far detector is composed of 486 layers of 8-m-wide, oneinch-thick, octagonal iron plates, giving a total mass of 5.4 kT. The iron is toroidally magnetized to approximately 1.5 T. Scintillator planes are interspersed between the iron planes. Each scintillator plane consists of 192 4-cm-wide strips of extruded scintillators with an embedded wave shifting fiber, which is read out at both ends. Figure 15 shows a drawing of the far detector. The function of the near detector is to predict the number and properties of events in the far detector, given the hypothesis of no oscillations. Thus, it
161
Figure 13. The trajectory of the MINOS neutrino beam between Fermilab and the Soudan mine. The beam is aimed down at an angle of 57 mrad at Fermilab to reach the far detector.
should be as identical to the far detector as possible and be analyzed in as similar a way as possible. However, since the rate per ton is approximately 500,000 times higher in the near detector than the far, practical considerations require some differences. The near detector has the same thickness iron plates and the same scintillator planes. However the iron has a "squashed toroid" shape, shown in Figure 16, which yields approximately the same magnetic field in the region of interest. The fiducial region is limited to 25 cm radius circle to best sample the neutrinos reaching the far detector. The match is unfortunately not perfect. The resulting difference in the near versus far v^spectra — assuming no oscillation — is due entirely to the lack of knowledge
162 , 400 50 m target pile -t- 675 m decay pipe
Perfect Focusing
30
35
40
EOJ GeV Figure 14. Neutrino interaction energy spectra predicted for the three different beam settings. "Perfect focusing" assumes that all secondary positively-charged particles are focused into a pencil beam with no divergence.
of the neutrino flux, i.e. without further empirical input (see below), one cannot predict the fM-flux in the far detector using the measured flux in the near detector. This, perhaps, is the biggest source of systematic error in the MINOS programme. Several efforts have been proposed to constrain, and possibly measure, the neutrino flux. The near detector has 6 m of iron in the longitudinal direction, divided into a 0.5 m veto region, a i m target region, a 1.5 m hadron shower region, and a 4 m muon spectrometer region. The fiducial region will only be the 1 m target region. The first three regions will have the same instrumentation as the far detector, except that the signals will be read from only one end and
163
Far Detector
FermilafcC :
25,800 m V \ 4 cm wide WLSfiber
r *
iFlm -
I ' •* ml i <•» ' Ii
Mm i > suiions 15 m Jong)
M.ignciizedFe Plates Utold\orsx2.54cmFe 5.4 kT Total Mass Magnet coil =1.5T
Figure 15. Sketch of the MINOS far detector.
will not be multiplexed. In the muon spectrometer region, only every third or fifth plane will be instrumented since this is sufficient for muon tracking. 4.2
MINOS Physics
There are three main near-far detector comparisons that MINOS can do to detect and measure neutrino oscillations: 1. The rate and spectrum of muon charged-current (CC) events. 2. The neutral-current/charged-current ratio (NC/CC). 3. The appearance of electron charged-current, or other exclusive, events.
164
4.8 m Figure 16. Sketch of the near detector looking along the beam. The shaded area is instrumented in the veto, target, and hadron shower regions (see text). The dark circle represents the 25-cm-radius fiducial target region. The diamond shaped hole represents the magnet coil.
The combination of these three measurements allows the determination of the oscillation parameters and the type or types of oscillations that are occurring. A comparison of the rate and spectrum of muon CC events, basically "long events," in the near and far detectors determines the extent to which muon-type neutrinos are oscillating into some other type of neutrino. There are only three possibilities: the muon-type neutrino oscillates into a tau-type neutrino, an electron-type neutrino, and/or a sterile neutrino. The NC/CC ratio, basically the ratio of short to long events, separates oscillations into an active neutrino (electron or tau type) from oscillations into a sterile neutrino. In the former case, the NC/CC will increase since the rate of CC events will decrease while the rate of NC events remains constant; in the latter case the NC/CC ratio will not change since the NC and CC rates will decrease by the same amount. Finally, the appearance of electron CC events will measure
165
the extent to which oscillations are occurring into electron-type neutrinos. Electron-type neutrinos can be identified by measuring the fraction of the shower which occurs in the first 5 to 10 radiation lengths. The CCFR experiment 5 4 has shown that this can be done, with a four-times coarser-grained detector albeit at much higher energies. The energy spectrum for CC events allows a relatively precise determination of the Am 2 parameter. This is demonstrated in Figure 17 for the lowenergy beam. Three different Am 2 values, all allowed by the Superkamiokande data, are easily distinguished. If the Am 2 value is in the region favored by the Superkamiokande data, about 3 x 10~ 3 (eV) 2 , then MINOS should be able to determine its value to about 7% in a two-year run. Considerably more simulation work needs to be done to understand the MINOS sensitivity with the different energy beams, including the systematic uncertainties. Preliminary simulations, however, indicate good sensitivity for MINOS for all of these tests. Figure 18 shows the sensitivity to fM —• vT and v n —• Ve oscillations for different tests for the high-energy beam. Figure 19 shows the improvement at low values of Am 2 achievable with the medium and low-energy beams for the CC test.
5
Outlook
The neutrino field is abloom with new ideas and initiative. It is the anomalies that have led to this renaissance of neutrino physics. Although in the review above, only the accelerator related research has been emphasized, there is an impressive array of investigation underway to solve these anomalies. We look forward to the next five years agog with anticipation
6
Acknowledgement
It is my previledge to have known Prof. Frank Avignone over the last several years. This symposium is in his honour. Frank is the 'Grand Old Man' of neutrino physics: he has been our mentor, our colleague, our friend. I wish him many discoveries in the years to come. I thank the organisers of this symposium, especially Prof. Kuniharu Kubodera, for their gracious hospitality in a sun-drenched Carolina, and for their patience in waiting for this article. I extend my warm thanks to Prof. Carl Rosenfeld for his critical reading of this article and essential critique.
166 i
200
'I'" i
I
i"~"i
i
r
Ph2(!ow) beam Am 2 =0.002 sin 2 (20)=1.O
150 100
50 0
0
4
6
8
10
12
14
CC reconst. Evis (GeV) T
200
Ph2(lovv) beam Am 2 =0.003 sin 2 (20)='I.O
150 100
50 0
Ife^^^^^^^^^ii 6
o
8
10
14
12
CC reconst. Evis (GeV) 1
T
' I
Ph2(!ow) beam Am 2 =0.005 sin 2 (20)=1.O
ife^^^ 4
6
8
10
12
14
CC reconst. Evis (GeV) Figure 17. Visible energy for reconstructed CC events at three different values of A m 2 for a 2-year-run of the LE beam. The open (shaded) histogram shows the expected distribution without (with sin 2 2# = 1.0) oscillations.
167
10"
I T1HMI
rrrrm—i i I I I I U
iff,-1
10" E <
i
10" 10-98 »7TA37
10"
10' ,-i
10* an
2
(28)
10° 10
i i mini i i mini i i mi -3
10'r2
1 fl-
10°
sin2 (29)
Figure 18. Calculations of 90% confidence levels that would be achieved by MINOS in a 2-year run in the absence of oscillations, using the high-energy beam. The curves represent different tests. Figure (a) is for KM -»• vT oscillations for (A) CC rate comparison, (B) N C / C C comparison, and (C) the CC energy spectrum comparison. Figure (b) is for v^ -> ve oscillations for (A) electron CC appearance comparison, (B) N C / C C comparison, and (C) the (muon) CC rate comparison.
References 1. For a review of recent data see S.R. Mishra, "Neutrino Oscillations," Plenary Talk at the COSMO '99, Trieste, September, 1999. 2. M. Gell-Mann, P. Ramond, and R. Slanksy in Supergravity (North Holland, 1978); T. Yanagida, Prog. Theor. Phys. B, 135, 66 (1978); S. Weinberg, Phys. Rev. Lett. 43, 1556 (1979). 3. S.A. Bludman, D.C. Kennedy, and P.G. Langacker, Phys. Rev. D 45 1810 (1992). 4. J. Ellis, J.L. Lopez, and D.V. Nanopoulos, Phys. Lett. B292, 189 (1992). 5. Ch. Berger et al., Phys. Lett. B227, 489 (1989); Phys. Lett. B245, 305 (1990). 6. M. Aglietta et al., Europhys. Lett. 8, 611 (1989).
168
CC Muon Neutrino Disappearance Test 10° ir-r
10-l
10 Kton-years
QQ7. conFidence level 2.07. Nu flux unct. t\l
> 10 •2 _ £
high energy beam medium"
10"
PH2 low energy 10"
_i
0.0
i—i
0.2
i
i
_l
i_
I
I
L-
0.4
_1
0.6
0.8
I
I
U
1.0
sin 2 (2G) Figure 19. Calculations of 90% confidence levels that would be achieved by MINOS for the CC comparison in a 2-year run in the absence of oscillations with the high-energy, medium-energy, and low-energy beams.
8 9
10 11. 12 13
K.S. HIrata et al., Phys. Lett. B205, 416 (1988); Phys. Lett. B280, 146 (1992); Y. Fukuda et al., Phys. Lett. B335, 237 (1994). R. Becker-Szendy et al., Phys. Rev. D 46, 3720 (1992); W.W.M. Allison et al., Phys. Lett. B449, 137 (1999). Review talk presented by A. Mann at the Lepton-Photon 1999, Stanford. "Final Results of Search for v^ -» vT and ve - • vT Oscillations in NOMAD," P. Astier, et al, paper in preparation. "Limit on ve -¥ vT Oscillations from NOMAD Experiment," P. Astier, et al, Phys. Lett. B471, 406 (2000). "Limit on ve ->• vT Oscillations from NOMAD Experiment," P. Astier, et al, Phys. Lett. B471, 406 (2000).
169
14. "A More Sensitive Search for i/M ->• uT Oscillations in NOMAD", P. Astier, et al, Phys. Lett. B453, 169 (1999). 15. Skubic et al., Phys. Rev. D18, 3115(1978). 16. "Probing Hadron Structure with Neutrino Experiments", S.R.Mishra, invited review talk presented at Workshop on Hadron Structure Functions and Parton Distributions, Fermilab, Apr.1990; World Scientific, 84-123(1990), Ed. D.Geesaman et al.; Nevis Preprint # 1426, Jun(1990). 17. Recent solar neutrino data are summarized in the review talk by Y.Suzuki, Lepton-Photon Conference, 1999, Stanford. 18. For an excellent review of solar neutrinos see W.C. Haxton, Annu. Rev. Astron. Astrophys. 33, 459 (1995). 19. P. Anselmann et al., Phys. Lett. B327, 377 (1994); Phys. Lett. B342, 440 (1995); W. Hampel, et al., Phys. Lett. B447, 127 (1999). 20. D.N. Abdurashitov et al., Phys. Lett. B328, 234 (1994); Phys. Rev. Lett. 77, 4708 (1996). 21. K. Lande in Neutrino '94 (North Holland) (1995). 22. K.S. Hirata et al., Phys. Rev. D 38, 448 (1988); Phys. Rev. D 44, 2241 (1991); Y. Fukuda et al., Phys. Rev. Lett. 77, 1683 (1996). 23. Y. Fukuda, et al., Phys. Rev. Lett. 81, 1158 (1998); 24. J.N. Bahcall and M.H. Pinsonneault, Rev. Mod. Phys. 64, 885 (1992). For an update, see J.N. Bahcall, Invited talk at 18th International Conference on Neutrino Physics and Astrophysics (NEUTRINO 98), Takayama, Japan, 4-9 Jun 1998, e-Print Archive: astro-ph/9808162. 25. S.L. Glashow and L.M. Krauss, Phys. Lett. B190, 199 (1987). 26. L. Wolfenstein, Phys. Rev. D 17, 2369 (1978); S.P. Mikheyev and A.Y. Smirnov, JVuovo Cimento 9C, 17 (1985); Prog. Part. Nucl. Phys. 23, 41 (1989). 27. J.N. Bahcall, P.I. Krastev, and A.Yu. Smirnov, Phys. Rev. D 58, 096016 (1998). 28. G. Aardsma et al., Phys. Lett. B194, 321 (1987). 29. Y. Fukuda et al., Phys. Rev. Lett. 81, 1562 (1998). 30. M. Messier, "Atmospheric Neutrinos at Superkamiokande," Talk at DPF '99, UCLA, January 7, 1999, http://www.physics.ucla.edu/dpf99/trans/2-06.pdf. 31. C. Athanassopoulos et al., Phys. Rev. Lett. 77, 3082 (1996). Also see C. Athanassopoulos et al., Phys. Rev. Lett. 81, 1774 (1998). 32. C. Athanassopoulos et al., Phys. Rev. Lett. 75, 2650 (1995); J.E. Hill, Phys. Rev. Lett. 75, 2654 (1995). 33. L. Borodovsky et al., Phys. Rev. Lett. 68, 274 (1992).
34. B. Achkar et al., Nucl. Phys. B434, 503 (1995). 35. B. Armbruster et al., Nucl. Phys. (Proc. Suppl.) B38, 235 (1995); Phys. Rev. C 57, 3414(1998). 36. G.F. Smoot et al., Astrophys. J. 396, 3,155-B5 (1992); 396, LI, 160-Bl (1992). 37. MX. Fisher et al., Astrophys. J. 388, 242, 53-Cll (1992). 38. For example see A.N. Taylor and M. Rowan-Robinson, Nature 359, 396 (1992). 39. N. Armenise et a l , CERN-SPSLC/90-42 (1990). 40. E. Eskut et al., Phys. Lett. B434, 205 (1998). 41. J. Altegoer, et al., Phys. Lett. B431, 219 (1998). 42. N. Ushida et al., Phys. Rev. Lett. 57, 2897 (1986). 43. M. Gruwe et al., Phys. Lett. B309, 463 (1993). 44. K.S. McFarland et al., Phys. Rev. Lett. 75, 3993 (1995). 45. F. Dydak et a l , Phys. Lett. B134, 281 (1984). 46. G. Ambrosini et al., Phys. Lett. B425, 208 (1998); CERN report number CERN-EP-98-018, to be published in Phys. Lett. 47. H.W. Atherton et al., CERN report number CERN 80-07 (1980). 48. "A Measurement of Absolute Electron Track Reconstruction Efficiency Using TRD Tracks," P. Hurst and F. Weber, NOMAD note 98-031. 49. B. Armbruster et al., Phys. Lett. B348, 19-28 (1995). 50. G.J. Feldman and R.D. Cousins, Phys. Rev. D 57, 3873 (1998). 51. C. Caso et al., Euro. Phys. J. C 3 , 1 (1998). 52. "The MINOS Detectors Technical Design Report," NuMI-L-337, http://www.hep.anl.gov/NDK/hypertext/minos_tdr.html. 53. The expected rate at the far location is 3000 i/^-events/year/kiloton. The near/far rate is 500,000. Thus, the number of ^-interactions in a 1.5 ton of detector at the near location is about 5 million over two years. 54. A. Romosan, et al., Phys. Rev. Lett. 78, 2912 (1997). 55. "Conceptual Design for the Technical Components of the Neutrino Beam for the Main Injector," FERMILAB-TM-2018. 56. J. Hylen, private communication. 57. "The Beam Guide, A Device for the Transport of Charged Particles," S. van der Meer, CERN 62-16. Also see "Theory of the Hylen Hadron Hose," R. H. Milbrun, NuMI-B-271. 58. "Proposal to Include Hadronic Hose in the NuMI Beam Line", J.Hylen et al., NuMI-B542, October 1999. 59. Private communication with Dr. Wes Smart. 60. http://root.cern.ch.
THE BOREXINO PROJECT A N D FUNDAMENTAL A C H I E V E M E N T S IN THE V E R Y LOW R A D I O A C T I V I T Y TECHNIQUES G. BELLINI* Dipartimento di Fisica dell'Universita' and INFN - Milan - Italy * On behalf of the Borexino Collaboration.
1
INTRODUCTION
The Borexino experiment is the first real time detector studying solar neutrinos in the sub-MeV region. In this region the expected signal ranges between 0.1-0.5 counts/day/ton. The first and hardest problem is represented by the radioactive background, especially by three different contributions: a- the cosmogenic 7 Be and the primordial 40 K, 238 U, 2 3 2 Th b- 2 2 2 Rn and 85 Kr gases c- 14 C with an end point at 156 keV. The most difficult problem is 1 4 C, which cannot be removed. A solar neutrino experiment, such as Borexino, needs a 14 C level as low as 1 4 C/ 1 2 C ~ 1 0 - 1 8 or similar. But, as I will show later, using organic liquid as a detector (as in the case of the Borexino scintillator solvent) it is possible to find a very low 14 C contamination level, if the primary product, crude oil in this case, has an underground residence of the order of 108 years. Metals like 40 K, U, Th, 7 Be show intrinsic solubility, which is vanishingly small in non-polar organic liquids. Distillation, water extraction, silica or alumina gel columns are valid tools in purifying the scintillator; they have been developed and tested by the Borexino Collaboration. Sparging with pure nitrogen (Rn-free down to the level of a few ^Bq/m 3 ) is very effective in removing Rn and Kr. Another very important tool against the contaminants is an active background tag. This is based upon the identification of alpha particles and the correlated events. About one half of the unstable nuclides of the Th and U families emit alpha particles, which are not present in the neutrino events; as a consequence, alpha identification tags a background decay. In addition, two nuclide-couples in the U chain and three in the Th chain give short time correlated decays such as 2 2 2 Rn- 2 1 8 Po, 214 Bi- 214 Po, etc. In the second case, for instance, an e~ with end-point energy 1.5 MeV is followed by an alpha with an energy of 7.688 MeV, quenched to about 750 keV in the Borexino 171
172
scintillator. These correlated decays can be easily identified using delayed coincidences. 2
T H E C O U N T I N G T E S T FACILITY (CTF)
The study of the background and the feasibility of Borexino has been carried out by means of the Counting Test Facility, a detector installed in Hall C of the Gran Sasso Laboratory. It is a simplified and scaled version of Borexino. One hundred phototubes coupled to optical concentrators detect the light emitted by 5 m 3 of scintillator, contained in a transparent nylon vessel, of lm radius. The scintillator consists of a Pseudocumene as solvent plus 1.5 g/1 of PPO as fluor; 1000 m 3 of purified water surrounds this system, providing a 2m shield in all directions. Measurements and procedures relevant to CTF and Borexino are as follows. - a calorimetric measurement of the energy with a typical resolution of 10%; the position of the events in the space obtained from the phototube timing (the choice of phototubes and a scintillator was conditioned by the need to keep the transit time jitter and the decay time as low as possible, ~ 1 ns and ~ 5 ns, respectively). - particle identification (alpha vs beta) to identify the background. - the time correlation between the events to disentangle the correlated decays of the background (A ~0.3 ns). The purification methods have been tested in CTF both for the water and the scintillator. The water purification was based upon reverse osmosis, deionization unit, nitrogen stripping, ultrafiltration at the 0.1/xm level. For the scintillator purification four different systems were installed. - water extraction and sub-micron filtration, effective for cleaning species such as metals. - vacuum distillation, a good tool against low-volatility components such as metals and dust particles. - solid column (Silica gel), in dry and wet form, with high affinity for the impurities. - nitrogen sparging to remove gases such as 2 2 2 Rn and 85 Kr; for this purpose Rn-free nitrogen, at the level of a few /xBq, is needed. All the construction materials have been carefully selected to keep their radioactive content below an acceptable limit (see table I); the selection was carried out through high-sensitivity Ge underground detectors and neutron activation. The low contamination of the materials assures not only low direct emission but also low Rn emanation.
173
Due to the high Rn level in Hall C of the Gran Sasso Laboratory (40-80 Bq/m 3 , depending on the ventilation), the auxiliary plants have to be not only air-tight but also fully impermeable to Rn; this demand has a clear impact on the choice of the materials and their thickness and on the gaskets. The surfaces have to be treated at "electronics grade", i.e. smoothed (0.60.8/jm), pickled and passivated or electropolished, to provide a clean surface and to facilitate cleaning of dust and particulate matter. The Inner Containment Vessel has been built using a copolymer nylon resin C38F. The U and Th content of the pellets was at the few ppt level, but it increases to the 10 ppt level during the extrusion and fabrication phase, even if done in a clean room. The buoyant force, due to the density difference between the internal scintillator and the external water, was 500 kg. The full installation is done in a controlled area, ranging from class 10 to class 10000. These methods and techniques, tested and applied to the CTF, are being applied presently in the construction of Borexino. 3
THE BACKGROUND PROBLEMS
The background in Borexino (and in CTF) can be classified in three categories: - External background, including emissions from the rocks of the cave (7's and neutrons) and from the construction materials, and cosmic rays - Internal background, due to the scintillator radio impurities (natural radioactivity and cosmogenic nuclides) - Surface contamination — the inner walls of the nylon containment vessel can become contaminated during the vessel fabrication (Rn daughters, etc.). The external background can be strongly supressed by proper shielding, a proper selection of the construction materials and proper software cuts to reject the muon interactions (in addition to a muon veto). The internal background depends directly on the radio purity of the scintillator. The problem we are facing is: first, the development of proper purification methods capable of reducing radioactive levels down to 10 - 1 6 g/g for the U and Th nuclide families and down to 10~ 14 g/g for K. These levels have not only never been reached, but also have never been measured. The detector CTF was developed to have a detector sensitive enough to measure such low contamination levels. The direct surface emissions are relatively innocuous, because Borexino has a buffer region defined between the vessel walls and the fiducial volume; on the other hand the release of the Rn daughter nuclei in the scintillator implanted on the nylon is very dangerous. To avoid this problem, the construction of the vessel has to be carried out in a clean room either in Rn-free
174
air or with a Rn protection. The sensitivity of the CTF was very good; the total counting rate was 0.5 counts/keV/kg/year in the range 250-2500 keV. In addition, the space reconstruction allowed a much higher sensitivity to the internal background. Let me just give an example. 85 Kr was present at the beginning of the data taking and we identified it through the decay 8 5 Kr- 8 5 m Rb- 8 5 Rb with the emission of a beta, with an end point at 175 keV, followed by a 514 keV gamma. The half-life is l^s, and the rate of this decay is only 0.43% of the total 85 Kr decays. The total rate measured at the beginning was 1.3±0.3 counts/day, corresponding to ~300±70 decays of 8 5 Kr per day. After the N 2 sparging process we measured 0.2±0.5 counts/day corresponding to ~50±35 decays/day of the total 85 Kr decay. A similar example is the detection of neutrons produced by crossing muons. They have been measured through the interactions of the neutrons with hydrogen, producing a deuteron with the emission of a 2.2 MeV gamma, with a total delay of 240 us. The neutron interactions identified were 1.2 per day, corresponding to the expected rate. The CTF internal background studies concerned in particular the U and Th families, 14 C and 222 Rn. The 14 C has been studied by looking at the low energy part of the single rate spectrum and comparing it with the predicted shape. The count rate, measured in the 50-100 keV window, gave (1.94 ±0.09) x KT 1 8 for 1 4 C/ 1 2 C. The U and Th equivalent rates have been measured via delayed coincidences: 214 Bi- 2 1 4 Po, with a lifetime of 236/JS, and 212 Bi- 2 1 2 Po with a lifetime of 433 /xs. They have been measured before any purification and after runs of water extraction and distillation. We did not observe a measurable difference between the two steps; this difference is expected to be very close to the sensitivity limit of the detector. In the case of the U family we can only determine an upper limit, because the 214 Bi and the 2 1 4 Po are daughters of 222 Rn, which permeates through the wet nylon (see later). An upper limit for the U equivalent rate is less than (3.5 ± 1.3) x 10~16 g/g. The Th equivalent rate was (4.4t};^) x 10~ 16 g/g. Further contaminants, probably introduced during the transport and the handling of the scintillator, gave 600±100 counts/day before any purification, and 0±50 counts/day after runs of water extraction and N2 sparging. The determinations of contaminants of the Th and U families have been carried out also through a very sensitive neutron activation analysis measuring the decays via a delayed coincidence technique, which detects the sequence 239TJ.239Np.239pu (239pu*) -16
& n d
233p ( j_233 U -15
T h e
mitia
J s e n s i t i v i t y WaS ~ 5 X
10 g/g for U and ~ 5 x 1 0 g/g for Th. After the installation of an active shielding the sensitivity was upgraded to ~10~ 1 7 g/g for U and ~ a
175
few x 10 - 1 6 g/g for Th. The contamination levels of the progenitors 239 U and 232 Th determined with this method were less than 2 x 10 - 1 6 g/g for U, and less than 2 x 1 0 - 1 5 g/g for Th. These results show that the production process of the Pseudocumene and the purification did not introduce breakings in the secular equilibrium of the radioactive families and that we need not worry about the build-up of nuclides in different sections of the radioactive chain during the use of the scintillator. During the CTF runs we have also tested another scintillator using PhenylorthoXylylethane (PXE) as solvent and Therphenyl as fluor. Unfortunately, the CTF runs had to be interrupted due to problems with the phototubes and, as a consequence, part of the measurements could only be done via neutron activation. In these runs with PXE a new purification system was tested: a solid purification column filled with Si gel. Before the purification the contamination level of the PXE was (1.6 ± 0.14) x 1 0 - 1 5 g/g for U and (2.5 ± 0.4) x 10~ 15 g/g for Th. After 6 runs through the wet column the measured neutron activation was less than 1.0 x 10~ 17 g/g for U, and less than 1.8 x 10~ 16 g/g for Th. The purified shielding water showed 10 - 1 4 g/g for U and Th, 1 0 - 1 0 g/g for the natural K, and 7 mBq/m 3 for radon. The Rn suppression was difficult because we must gain 6-7 orders of magnitudes with respect to the Gran Sasso water. For the radium content (determined via a charcoal filter and miniaturized proportional counters) only an upper limit of 15 mBq/m 3 was obtained. Recent upgradings of the analysis system improved the sensitivity, and the insertion of resin cartridgers lowered the radium content to ~ 1 mBq/m 3 . One of the important problems of CTF has been radon contamination. The shielding water of the CTF had a radon content of about 30 mBq/m 3 due to the radium content of the water, the emanation from the construction material, and possible leaks of the water tank. The permeability of the dry nylon to radon rapidly becomes higher when the nylon is in contact with water, reaching the level of ~ 4 x 1 0 - 1 0 cm s - 1 . Taking into account this value of the permeability and the radon content of the shielding water, we can expect a rate of ~0.5-1.0 events/day, corresponding to ~ 4 x 10 - 1 6 g/g. This is the lowest level ever measured for the U equivalent content of the scintillator. Taking into account these problems, we have decided to install a nylon balloon between the Inner Vessel and the photomultipliers and the other construction materials, both in the CTF and in Borexino. In addition, in Borexino the nylon vessel is not in contact with water (see section 5); in this case the permeability to Rn is expected to be two orders of magnitude lower.
176
As explained in section 1, a further important tool in fighting against the background is its software identification. To this end, the scintillator properties were carefully studied in CTF. A typical photoelectron yield was 300 p.e./MeV. The autoabsorption process involves 54% of the photons; 77% of them are re-emitted. This tends to lengthen the intrinsic decay time of the scintillator, which ultimately is ~4.5 ns. The alpha/beta discrimination was successful at the 90% level at 300 keV and at the 97% level at 751 keV. 4
CTF2
The Counting Test Facility, after a period of shut down, is now running; it has been refurbished and upgraded by installing a nylon balloon as a Rn barrier (see the previous section) and a simple \i veto. The present goals of this detector are: - the control of the 14 C and 7 Be content in batches of PC as received from the producer - measurements of the contaminants of the U and Th families in the Borexino scintillator and in the shielding liquid - checking the Borexino liquid handling plant (Rn leaks, cleanliness, etc.) - testing the effectiveness of the Borexino purification plants. 5
THE BOREXINO DETECTOR
The Borexino detector is shown in Fig.l. It has 300 m 3 of liquid scintillators (PC plus 1.5 g/1 of PPO) contained in a transparent nylon vessel (Inner Vessel), 8.5 m in diameter. A stainless-steel sphere (SSS), 13.7 m in diameter, supports 2200 phototubes on its inner walls; 1800 are coupled to light concentrators in order to increase the optical coverage up to ~ 30%. The region between the Inner Vessel and the SSS is filled with ~ 1000 m 3 of PC without fluor and with 5.5 m 3 of DMP as a quencher; in this region a nylon balloon is interposed as a Rn barrier between the Inner Vessel and the phototubes. Four hundred phototubes are mounted without light concentrators to have a wider acceptance, and they are part of the fi veto. The other 200 phototubes are installed on the outer walls of the SSS to detect ju's crossing the shielding water. Finally, everything is immersed in about 2500 m 3 of purified water contained in a water tank; this ensures 2 m of shielding water to SSS in all directions. The choice of PC as shielding material (buffer liquid) inside the SSS is dictated by the need to maintain the buoyant forces on the vessel as small as possible and to avoid direct contact between nylon and water. The detector is conceived as an assembly of concentric shells; the closer to the
177 Borexlno Design
- Holding Strings . Stainless Steel Water Tank 18m0
jf
2200 8" Thorn EMI PMTs In l i *• • , ' ' coflectors '." A I 1 ght cones)
Steel Shielding Plates 8 m x 8 m x 10cm and 4m x 4m x 4cm
Figure 1.
center a shell is, the lower is its level of radioactive contamination. In table I are shown radio purity levels needed for the various subsystems in comparison with what has been achieved in the CTF. Table I Material Stainless steel Shielding water Photo multipliers Buffer liquid Scintillator
Borexino Design goal
Achieved with CTF
~ i o - 9 g/g ~ i o - l u g/g
~ i o - 9 g/g
~ IO" 8 g/g 14
~ io- g/g ~ io- 1 6 g/g ~ 10- 1 8 g/g
Radio purity Th, U equiv.
~ IO" 1 4 g/g ~ IO" 8 g/g
~ io-
15
JJM
itv
»» »»
g/g
~ 4xl0-18g/g ~ 2xl0-18
i4
C
/ia
c
178
Figure 2.
The detector is complemented with several ancillary subsystems (see Fig.2): - a water purification plant (same as in CTF but with the addition of resins that filter radium (see section 2)) - scintillator purification plants, including water extraction, distillation, and purification columns with very pure Si gel (the purification rate of these plants is 1 m 3 /h) - a N 2 plant consisting of normal and high-purity lines. The former, via a boil off system, lowers the Rn contamination level down to ~0.3 mBq/m 3 , while the latter, equipped with a cryogenic charcoal filter (LTA), allows us to reduce the Rn content to ~0.5 fiBq/m3. - a storage area for the scintillator with four silos, 120 m 3 each.
179
The Borexino detector is under construction in Hall C of the Gran Sasso Laboratory. It is expected to start data taking in the spring 2002. 6
THE BOREXINO PROGRAM.
The main interaction studied in Borexino is neutrino elastic scattering on an electron. A software threshold of E e = 0.25 MeV is defined by the software, and it corresponds to a neutrino energy threshold of 0.4 MeV. In Fig. 3, neutrino signals expected in Borexino are shown and compared with the expected background. The energy resolution has been folded in. The Borexino program consists of the following items. - Rates The v event rates expected in Borexino are calculated for the values of <5m2 and sin2 (26) corresponding to the currently allowed regions for the MSW and Vacuum oscillations hypotheses. The allowed regions are defined by the Standard Solar Model (Bahcall and Pinsonneault, 1998) or by the results of the CI and Ga experiments and the SuperKamiokande experiment (total rate). The calculated event rates are summarized in Table II. Table II: Expected rates (events/day) S.S.M. 55
MSW-SMA 12
MSW-LMA 32
MSW-LOW 29
VACUUM 26 (20-39)
The prediction for the vacuum oscillation (V.O.) hypothesis has a large range due to the width of the still allowed region. These rates should be compared with an expected background of ~15 events/day. A small rate, less than the MSW-SMA prediction, could favour a sterile neutrino. On the other hand, a high rate should disfavour the SMA solution. - Seasonal modulation. The seasonal modulations of the Borexino rates can be due to two origins: the eccentricity of the Earth's orbit with respect to the Sun (geometric modulation) and vacuum oscillations. The geometric modulation produces a ~ 3 % effect and gives a direct demonstration of the solar origin of the detected neutrinos. Borexino is able to reproduce its typical 1/R2 behavior with a C.L. at 5cr in 3 years of data taking.
180
The modulations due to vacuum oscillations are definitely higher and can reach a factor of ~ 4 between different annual seasons. Borexino is particularly favoured due to the monochromaticity of the 7 Be line. According to a Fourier analysis applied to the Borexino data that have been generated by assuming the V.O. hypothesis, the Borexino sensitivity extends to large Am 2 values. - Zenith angle distribution and day/night difference. These two observables can be affected by the existence of the MSW effect. The solar neutrinos from the 7 Be source show an important day/night effect (a factor of 2 difference) in the LOW region. Borexino is unique in this domain. Similar considerations are also valid for the zenith angle distribution. - 8 B solar neutrinos. The event rate in Borexino for the 8 B solar neutrinos is ~300/year. Borexino can measure the neutrino energy spectrum with a threshold at ~3.5 MeV. - Borexino program beyond the solar neutrinos. Borexino can provide a very powerful tag for V events via the following reactions: v+ p = e + + n (threshold = 1.8 MeV) n + p = 2 H + 7(2.2 MeV). The latter reaction has a delay of ~200^us with respect to the V reaction. Therefore the following studies are possible. - Study of F's emitted by the mantle and the crust of the Earth The v production in the Earth is due to the /? decay of four nuclides of the Th and U chains. The measurement of their rates allows us to determine the U and Th abundances separately, because the positron-deposited energy is different for the two nuclide-chains. The expected rate in Borexino is ~10 events/year with a negligible background. - Long base line measurement of reactor z/'s Borexino detects ~30 events/year of reactor 77's. The reactors are located in Europe, mainly in France, and the closest one is 600 km distant. These events can be easily disentangled if we introduce a cut at ~2.5 MeV on the positron deposited energy. It is possible to explore with Borexino the large mixing angle and the large Am 2 region, gaining an order of magnitude in the mass difference with respect to the Chooz results. - Supernova explosion detection Calculations done for a supernova explosion at the center of our Galaxy
181
Figure 3.
(10 kpc) foresee more than 100 neutrino interactions in Borexino. The most important contributions are from T>e + p interactions with 79 events (threshold at 1.8 MeV), and from V2C{vx, vx)v2G* reactions with 21 events (threshold at 15.1 MeV ). This last reaction can be easily identified because 1 2 C* decays with the emission of a 15.1 MeV 7-ray. The expected energy spectrum of the neutrinos from a supernova event is dominated by ve and ve at low energies and by vx and Vx at higher energies. As a consequence we can obtain an indication on the mass difference between neutrinos of different flavours from the time difference between the two classes of events mentioned above. - Appearance experiment. We can search for the presence of TVs in the solar flux. This study is connected to a possible spin-flavour conversion of the solar neutrinos in the solar magnetic field range (0.0-3.0xl0~ lo eV 2 ).
SEARCHES FOR N O N - S M PHYSICS W I T H T H E K A R M E N EXPERIMENT K. E I T E L a Institut fur Kernphysik, Forschungszentrum Karlsruhe Postfach 3640, 76021 Karlsruhe, Germany The neutrino experiment KARMEN is taking data in its upgraded configuration since February 1997. A search for v^—> 9e oscillations through p ( ve ,e+ ) n shows no oscillation signal. A maximum likelihood analysis of the data leads to an upper limit for the mixing angle of s i n 2 ( 2 e ) < 2.1 • 1 0 - 3 (90% CL) at large A m 2 . An anomaly in the time distribution of v-induced events is persistent through the entire KARMEN1+2 data set. A new analysis including recent data is presented showing consistency with a hypothetical non-SM decay 7r+—> fi++ X of pions.
1
Neutrino Production and Experiment Setup
The Karlsruhe Rutherford Medium Energy Neutrino experiment KARMEN is performed at the neutron spallation facility ISIS. In addition to neutrons, pions are produced by stopping 800 MeV protons in a beam stop target of T&-D2O. Muon neutrinos emerge from the 7r+—>• / i + + ^M decay at rest (DAR). The produced fi+ are stopped within the target and decay via JX+ —> e + + ve + v^. There is only a very small intrinsic contamination of vejve < 6.2 • 10~ 4 from 7r~ decay in flight followed by \i~ DAR which is further reduced by evaluation cuts. The energy spectra of the j/'s are well defined due to the DAR of both the 7r+ and fi+ (Fig. la). The v^ are monoenergetic with E(i^)=29.8MeV, the continuous energy distributions of ue, PM up to 52.8 MeV can be calculated using the V-A theory. Two proton pulses of 100 ns base width and a gap of 225 ns are produced with a repetition frequency of 50 Hz leading to a unique time structure of the neutrinos (Fig. lb,c). The neutrinos are detected in a rectangular segmented tank with 561 of liquid scintillator at a mean distance of 17.7 m which is well shielded against beam correlated spallation neutrons and cosmic induced background. The detector and shielding are described in detail in 1 and 2 . 2
The search for 9^De
Oscillations
The signature for the detection of ve is a spatially correlated delayed coincidence of positrons from p ( v e , e + ) n with energies up to Ee+ = Epe - Q — 52.8 - 1.8 = 51.0 MeV and 7 emission of either of the two neutron capture a
for the KARMEN collaboration; electronic address: [email protected]
182
183
10
20
30
40
50
energy Ev [MeV]
time [[is]
Figure 1: Neutrino energy spectra (a) and production times of u^ (b) and v e ,i^, (c) at ISIS.
processes p (n/y) d with one 7 of E(j) — 2.2 MeV or Gd (11,7) Gd with 3 7 quanta in average and a sum energy of J3 E(l) — 8 MeV. The positrons are expected in a time window of several fxs after beam-on-target with a 2.2 (is exponential decrease due to the /u+ decay. The time difference between the e + and the capture 7 is given by the thermalization, diffusion and capture of neutrons. The raw data investigated for this oscillation search were recorded in the measuring period of Feb. 1997 through Feb. 1999 which corresponds to 4670 C protons on target. A positron candidate is accepted only if there is no previous activity in the central detector, the inner anti nor outer shield up to 24jus before. The required cuts in energy and time are: 0.6 < tp < 10.6 /JS, 16.0 < Ep < 50.0 MeV. The cuts on the delayed expected neutron event are as follows: 5.0 < td-tp < 300/is, Ed < 8 MeV and a volume of 1.3 m 3 for the spatial coincidence. Applying all cuts, the total background expectation amounts to 7.8 ± 0.5 sequences where the individual background sources are: 1.9 ± 0.1 events induced by cosmic fi, 2.6 ± 0.3 12 C (v e ,e~ ) 12 N g . s . sequences, 2.3 ± 0.3 !/-induced accidental coincidences and 1.1 ± 0.1 sequences from the intrinsic ISIS ve contamination. Analysing the data results in 8 sequential events which satisfy all conditions. To extract a possible PM-» ve signal a maximum likelihood analysis of these 8 sequences is applied making use of the precise knowledge of all background sources and a detailed MC description of the oscillation signature in the detector. As this analysis shows no hint for an oscillation signal, an upper
184
io "3
10 "2
io 'l
i sin 2 0 2
Figure 2: KARMEN 90% CL exclusion curve for the oscillation parameters A?7i 2 ,sin 2 (20). Also shown are limits from other experiments and the confidence region for P^ -> Pe from a preliminary maximum likelihood analysis of the LSND 1993-98 data.
limit of 4.0(3.4) oscillation events for Am 2 < l e V 2 / c 4 ( A m 2 > 20eV 2 /c 4 ) can be extracted at 90% CL. Assuming maximal mixing (sin 2 (20) = 1), 1605±176 (e + ,n) sequences from oscillations with large Am 2 were expected leading to an upper limit on the mixing amplitude of sin 2 (20) < 2.1 x IO" 3 . The complete exclusion curve in (Am 2 ,sin 2 (20)) can be seen in Fig. 2. Also shown are limits from other experiments 3 ' 4 as well as the 90% and 95% confidence regions from LSND based on a preliminary analysis of the entire 1993-98 data set 5 . Regions of a fixed confidence level such as in Fig. 2 only represent a somewhat discrete picture of 2-dimensional and rather complex likelihood functions. Both the KARMEN and LSND experimental data were analysed with such likelihood functions. In addition, the statistics to deduce confidence regions were built in the same way as suggested by Feldman and Cousins 6 (F-C). Therefore, it is possible to combine the likelihood functions and extract combined confidence regions based on a combination of the individual F-C statistics created by Monte Carlo procedures. The results of such an analysis 7 can be summarized as follows: Though the experiments have different central statements (KARMEN: no signal, LSND: excess interpreted as Vp-+ve), there is a
185
region in (Am 2 ,sin 2 (26)) statistically compatible at 90% CL with both outcomes. Regions with Am 2 < 2 eV2 have much higher combined likelihood than a couple of 'islands' at higher Am 2 . These islands disappear when considering a more stringent region of 80% combined confidence. Finally, at a confidence level of « 70% for the individual experiments the outcomes are statistically incompatible. 3
The time anomaly
In the window of 0.6-10.6/is after beam-on-target the contribution of reactions on 1 2 C, 13 C and 56 Fe induced by ve and PM should reflect the 2.2 fj,s life time of the muons decaying in the target superimposed on a flat cosmic induced background. This fact is well underlined in the total KARMEN1+2 data set 2.2 IJS + flat CR bg. (ML-fit)
100
2
4
6
8
10
time after beam-on-target [|js]
5
6 7 8 velocity v x [m/ps]
Figure 3: left: Time distribution of events after beam-on-target with a fit (solid line) assuming cosmic background and i/-induced events with fixed lifetime of muons in the ISIS target. right: Likelihood as a function of hypothetical particle velocities. Shown is the experimental result as well as a high statistics MC data set scaled down to the experimental values.
shown in Fig. 3. However, there is a visible distortion between 3.1-4.1/iS corresponding to an excess of about 100 events first reported in 1995 on a smaller data set 8 . Since then, any attempt to explain this peak-type structure by systematic effects failed. Besides a statistical fluctuation, only an explanation in terms of new physics seems to be left to explain the anomaly. Each such new physics hypothesis explaining the time anomaly has to fulfill certain very distinct conditions imposed by the observations. These
186
boundary conditions are the observed base width of approximately 1 /is of the distortion about 3.6 /Lis after beam-on-target. One hypothesis which does fulfill the above conditions without any contradictions throughout the analysis is a very rare branch of the decay of positive pions at rest 7r+-> H++ X. The new X particle slowly penetrates the massive iron shielding, therefore it must be massive, neutral and only weakly interacting. The rest mass of the X particle is close to the mass difference between ir+ and fi+ corresponding to a very small velocity of fix ~ 1/60. The signal in the scintillator would then be induced by the decay of the X particle into e~e + pairs or photons and neutrinos. With a velocity of 5 m//us the travelling time of an X particle in the central detector along the 3.5m long module axes is approximately Ait r . o „ e (=700ns. With a time resolution of less than At<2ns and a spatial resolution of Aa; <15cm there must be a unique time of flight correlation (ToF) of each X particle decay between t and x. Figure 3 shows the result of a maximum likelihood (ML) analysis of the data under such a ToF hypothesis as function of the particle velocity. A MC simulation based on a X particle contribution is also analysed demonstrating good agreement with the experimental result. More details of this new statistical analysis can be found in 9 . The strength of the distortion corresponds to very small branching ratios r of n+—> fi++ X down to 10~ 17 and a broad range of lifetimes r up to several hundreds of seconds for the X particle. Up to now, there are experimental limits from P S I 1 1 and Fermilab 10 which cover only minor parts of the allowed region. KARMEN will continue taking data until April 2001 to enhance the sensitivity for the v^-^De oscillation search, but also to further investigate the time anomaly while detector design studies for a dedicated X particle search are under way. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
G. Drexlin et al, Nucl. Instrum. Methods A 289, 490 (1990). G. Drexlin et al, Prog. Part. Nucl. Physics 40, 193 (1998). M. Mezzetto et al, Nucl. Phys. B 70, 214 (1999). B. Achkar et al, Nucl. Phys. B 434, 503 (1995). R. Tayloe et al, Proc. of the Lake Louise Winter Institute (1999) G. Feldman and R. Cousins, Phys. Rev. D 57, 3873 (1998) K. Eitel, New Journal of Physics 2, 1.1 (2000). B. Armbruster et al, Phys. Lett. B 348, 19 (1995). C. Oehler et al, http://www-ikl.fzk.de/karmen/publist_e.html#1999 J. Formaggio et al, hep-ex/9912062 M. Daum et al, http://wwwl.psi.ch/~jschott/
RESULTS OF THE PALO V E R D E LONG BASELINE REACTOR NEUTRINO EXPERIMENT
J. Wolf (Palo Verde Collaboration) Department of Physics and Astronomy, University of Alabama, Tuscaloosa AL 35487 We report on the initial results from a measurement of the anti-neutrino flux and spectrum from the three reactors of the Palo Verde Nuclear Generating Station using a segmented gadolinium-loaded scintillation detector at a distance of about 800 m. We find that the anti-neutrino flux agrees with that predicted in the absence of oscillations, excluding at 90% CL P e — #x oscillations with Am2 > 1.12 x 10~ 3 eV 2 for maximal mixing and sin 2 20 > 0.21 for large A m 2 . Our results support the conclusion that the atmospheric neutrino oscillations observed by Super-Kamiokande do not involve ue.
1
Introduction
Nuclear reactors have been used as intense sources of ue in experiments searching for neutrino oscillations 1 . These experiments usually detect ve by the process Pe + p —>• n + e + , where the cross-section-weighted energy spectrum of ve, peaking at about 4 MeV, can be deduced from the measured e + spectrum. Any Pe flux deficit or distortions of the ve energy spectrum would indicate oscillations. The low energy of reactor Pe allows these experiments to reach very small mass parameters, albeit with modest mixing-angle sensitivity. Past experiments 2 with detectors at 50-100 m from a reactor have explored the mass-parameter range down to 10~2 eV 2 . The work described here, and a similar experiment at the Chooz reactor in France 3 , are the first long baseline (~1 km) searches, designed to explore the parameter range down to 10~ 3 eV2 as suggested by the early Kamiokande atmospheric neutrino anomaly 4 . Although later results from Super-Kamiokande 5 , which appeared while this work was in progress, seem to disfavor the v^ — ue channel, a direct experimental exploration amply motivated this work. The Palo Verde neutrino oscillation experiment 6 is located at the Palo Verde Nuclear Generating Station near Phoenix, Arizona. The total thermal power from three identical pressurized water reactors is 11.6 GW. Two of the reactors are 890 m from the detector, while the third is at 750 m. Our detector is placed in a shallow underground site (32 meter-water-equivalent overburden), thus eliminating the hadronic component of cosmic radiation and reducing the muon flux by a factor of ~ 5. The fiducial mass, segmented to reject the remaining background, consists of 11.3 tons of 0.1% Gd-loaded liquid 187
188
scintillator contained in a 6 x 11 array of 9 m-long acrylic cells, as shown in Fig. 1. Each cell is viewed by two 5-inch photomultiplier tubes, one at each end. A ve is identified by space- and time-correlated e + and n signals. Positrons deposit their energies in the scintillator and annihilate, yielding two 511 keV 7's, giving a triple coincidence. Neutrons thermalize and are captured in Gd, giving a 7-ray shower of 8 MeV total energy. The Gd loading of the scintillator has two advantages: it reduces the neutron capture time from 170 /is (on protons) to 30 [is and provides a high energy 7 shower to tag the neutron capture, resulting in a substantial background reduction. Both the positron and the neutron are triggered by a triple coincidence requiring at least one cell above a "high" threshold set at about 600 keV (positron ionization or neutron shower core), and two cells above a "low" threshold set at about 40 keV (Compton scattering from annihilation photons or neutron shower tails). The triple coincidences are required to be within a 3 x 5 matrix anywhere in the detector. The central detector is surrounded by a 1 m water shield to moderate background neutrons produced by muons outside the detector and to absorb 7's from the laboratory walls. Outside the water tanks are 32 large liquid scintillator counters and two end-caps to veto cosmic muons. The rate of cosmic muons is approximately 2 kHz. The pattern of muons traveling through veto chambers and their timing relative to the central detector hits are recorded for subsequent off-line analysis. In order to reduce natural radioactivity, all building materials for the detector were carefully selected, including the aggregate (marble) used in the concrete of the underground laboratory. 2
Efficiency, Calibration and Neutrino Flux
Since the ultimate sensitivity of the experiment relies on a disappearance measurement, precise knowledge of the detector efficiency and of the expected ve flux from the reactors is essential. The efficiency calculation is based upon a primary measurement performed a few times per year with a calibrated 22 Na e + source and an Am-Be neutron source, placed at various positions inside the detector. The 2 2 Na source is inserted trough calibration pipes inside the detector and mimics the effects of the positron from the Pe interaction by providing annihilation radiation and a 1.275 MeV photon which simulates the e + ionization in the scintillator. The neutron detection efficiency is measured by scanning the detector with the Am-Be source where the 4.4 MeV 7 associated with the neutron emission is tagged with a miniaturized Nal(Tl) counter. Other radioactive sources are used to measure the energy response of the
189
detector. A Th source placed at 7 positions along each cell is used more frequently to track the scintillator transparency. Weekly runs of fiber-optic and LED flasher systems are used to monitor the gain and linearity of photomultipliers and the timing/position relationship along the cells. The i>e flux and spectrum from a fission reactor and the Pe + p —> n + e + cross section are well known 1 ' 2 ' 7 and are calculated by tracking the 235 U, 238 U, 239 Pu, and 2 4 1 Pu fission rates in the three plant reactors, taking into account both power level and fuel age. The uncertainty in the Pe reaction rate is less than 3%. 3
Neutrino Selection
The data presented here were collected in periods of 67.3 days in 1998 and 134.4 days in 1999. During the 98 (99) data taking one of the far (near) reactors was off for 31.3 (23.4) days. Here we outline the principles of the analysis and the results. A detailed description can be found elsewhere6. Neutrino candidates were selected by requiring an appropriate pattern of energy to be present in the detector for the positron- and the neutron-like parts of the events. In addition the two sub-events are required to occur within 1 m from each other. At our depth the background to Pe events consists of two types of events: uncorrelated hits from cosmic-rays and natural radioactivity and correlated ones from cosmic-muon-induced neutrons. The first type can be measured by studying the time difference between positron-like and neutron-like parts of an event. By requiring that the time lapse between the two sub-events ten be 5 /is < ten < 200 [is, the uncorrelated background is reduced to 3.4 ± 0.2 events d-"1 (4.8 ± 0.2 events d _ 1 ) for 1998 (1999), as measured from a fit to an appropriate combination of exponential functions. From the distribution of time intervals between a cosmic-ray /x and a Pelike event we infer that the majority of correlated background is produced by pairs of neutrons, where the capture of the first neutron in each pair mimics the positron signature. The requirement that no cosmic-ray hits be present in a window of 150 /is preceding the ve candidate completes the event selection. The resulting rates N of ve candidates per day in different periods are given in Tab. 1. Along with the ue events this final data set contains both random and correlated background, as mentioned above. Two independent techniques were used to estimate and subtract the background. The most straightforward method ("Method 1") relies on the changes of the De signal when different reactors are turned off. A x 2 -analysis is performed, comparing the expected rate to the efficiency-corrected measured rate, assuming a constant background.
190 Table 1: Summary of results from the Palo Verde experiment. The values in the second part of the table are derived from Method 2. B = B u n c + Bna + Bnp. Robs a n d ^Calc a r e the observed and calculated Ve rates corrected by the efficiencies r) for the case of no-oscillations. Uncertainties are statistical only. (Reactor at 890 m* (750 m T ) distance off) Period Duration (d) Efficiency r\ Event rate N ( d - 1 ) Pe rate SV ( d - 1 ) Background B ( d _ 1 ) Robs ( d " 1 ) ttcalc ( d - 1 )
98 "on" 36.0 0.0746 38.2 ± 1 . 0 16.5 ± 1 . 4 21.7 ± 1 . 0 221 ± 18 218
98 "off"* 31.3 0.0772 32.2 ± 1.0 13.4 ± 1 . 4 18.8 ± 1.0 174 ± 17 155
99 "on"
99 "off't
111.0 0.112 52.9 ± 0 . 7 25.2 ± 0 . 9 27.7 ± 0 . 6 225 ± 8 218
23.4 0.111 43.9 ± 1.4 15.1 ± 1.9 28.8 ± 1 . 3 136 ± 17 130
The systematic error for "Method 1" is 10%. We find good agreement with the hypothesis of no oscillation, as shown in Fig 1 (curve a). In "Method 2", that is described in detail elsewhere 8 , we make use of the intrinsic symmetry of the dominant background components under interchange of the energy cuts for positron and neutron sub-events. The rate of ve candidates after all cuts can be written as N = Bunc+Bnn+Bpn + Sv, where the contribution of the uncorrelated Bunc, two-neutron Bnn and other correlated backgrounds Bpn are explicitly represented, along with the Pe signal 5„. The event selection with swapped cuts results in a rate N' = Bunc + Bnn + e\Bpn + e 2 S„ where ei and e2 account for the different efficiency for selecting asymmetric events after the swap. The dominant background Bnn and Bunc is symmetric under exchange of sub-events, so that it cancels when we subtract N — TV' = (1 — e i ) 5 p n + (1 — e 2 )5^. The efficiency correction for the neutrino signal is e2 — 0.2. Monte Carlo simulations and data above a 10 MeV energy cut are used to estimate (1 - ei)Bpn. We find that the processes of //-spallation in the laboratory walls and capture of fi's passing the veto counter untagged contribute to (1 - €i)Bpn, while other backgrounds are negligible. We obtain (1 - e i ) B p n = - 0 . 3 ± 0 . 7 d"""1 (-0.4 ± 0.8 d _ 1 ) in 1998 (1999). This represents only a small correction to N - N'. This technique makes the best possible use of the statistical power of all data collected. The systematic error for "Method 2" is 8%. The results are shown in the second part of Tab. 1 for different running periods. Clearly Method 2 is also in agreement with the no-oscillation hypothesis (Fig. 1 curve b). In conclusion, the data from the first period of running from the Palo Verde detector show no evidence for v^-v^ oscillations. This result, together with the data already reported by Super-Kamiokande 5 and a more stringent limit by
191 .1 . 1 1 1 1 1 1 1> 1• • • ( • • •
(b)
Y
Kamiokande atmospheric v —> v
±h?~ ^
10
I*
e
1
S
LED ESJl
oil
\\
-M optical fiber
LED
scintillator 1
oil
Jpa
1 . . . .
0.1
1 . . . .
0.2
1 . . . .
0.3
1 . . . .
0.4
1 . . .1
0.5
0.6
1
0.7
1 . . . .
0.8
1 . . .
0.9
.
1
Figure 1: Left: Schematic view of the Palo Verde neutrino detector Right: Limits on A m 2 and sin22d from the present work (90% CL). Curves (a) and (b) are based on Method 1 and 2 for background subtraction, respectively as described in the text.
Chooz 3 , exclude the channel v^ — ve as being responsible for the atmospheric neutrino anomaly reported by Kamiokande 4 . Data-taking at Palo Verde is scheduled to continue until the summer of 2000. Acknowledgments We would like to thank the Arizona Public Service Company for the generous hospitality at the Palo Verde plant. This project was supported in part by the US DoE. It also received support from the Hungarian OTKA fund and from the ARCS foundation. 1. See e.g. F. Boehm and P. Vogel, "Physics of Massive Neutrinos" Cambridge University Press 1992 (2 nd ed.). G. Zacek et al, Phys. Rev. D34 (1986) 2621 ; B. Achkar et al., Nucl. Phys. B434 (1995) 503. 3. M. Apollonio et al, Phys. Lett. B466 (1999) 415. 4. Y. Fukuda et al, Phys. Lett. B335 (1994) 237. 5. Y. Fukuda et al, Phys. Rev. Lett. 81 (1998) 1562. 6. F. Boehm et al, hep-ex/0003022, subm Phys. Rev. D. 7. P. Vogel and J.F. Beacom, Phys. Rev. D60 (1999) 053003. Y.F. Wang et al, hep-ex/0002050, subm. Phys. Rev. D.
The Observatory for Multiflavor Neutrinos from Supernovae R . N . Boyd Departments of Physics and Astronomy Ohio State University, Columbus, Oh 43210 USA E-mail: [email protected] OMNIS, the Observatory for Multiflavor Neutrinos from Supernovae, will consist of 14 kT of lead and iron which, when irradiated by neutrinos from a supernova, will produce secondary neutrons which will then be detected. A supernova at the center of the Galaxy will produce about 2000 events in OMNIS, mostly from neutral current interactions, creating an unprecedented d a t a base from which to understand the processes that affect and govern stellar collapse. OMNIS' lead modules give it particular sensitivity to neutrino oscillations of the type v^ —¥ ve or i>r —• ve. Its intrinsic timing capability, better than 0.1 ms, might allow it to measure neutrino mass from the time-of-flight shifts in the luminosity curves of the neutrinos of different flavors to 20-30 e V / c 2 . In the event of collapse to a black hole, OMNIS may be able to determine the mechanism that caused the collapse, measure the stellar geography, and determine neutrino masses to a few e V / c 2 from the differences in the luminosities of the neutrinos of different flavors.
1
Introduction
The evolution of a massive star ultimately produces a core collapse supernova (SN), which emits about 10 53 ergs of neutrinos 1 . The standard model 2 of this process suggests that i/e's have a mean energy of 11 MeV, j / e ' s of 16 MeV, and ^ ' s , i^'s, i'r's, and t/ r 's of 25 MeV. The ve's and ve's interact with matter through both the charged-current (CC) and neutral-current (NC) interactions, whereas all others interact only through NC interactions. This traps the i/ e 's and i/e's farther from the star's core, so they emerge (from a cooler region) with a lower energy than do the other neutrinos. The i/e luminosity should exhibit a "neutronization spike" that would signal the beginning of the stellar collapse, from p + e~ ->• n + ve.
(1)
Subsequent broad distributions lasting at least several seconds would then be expected from neutrinos being produced in the hot core of the star by e + + e~ -> Vi + in
(2)
where i = e, //, or r. This process occurs only about once in 1019 e + e annihilations, but it can cool the core of the star in a few seconds. 192
193 Detection of the neutrinos from a SN can provide diagnostics of the environment from which they are produced and trapped. Indeed, the observation of the ve signal 3, 4 from SN 1987a produced a qualitative confirmation of the theoretical description of the trapping, as the neutrinos, thought to be produced in tens of milliseconds l , were observed over several seconds. However, much more can be learned both by observation of a larger statistical sample of SN neutrinos and by observation of neutrinos other than i>e's. 2 2.1
Technical Characteristics of OMNIS Mechanical Details
Thus we are designing the Observatory for Multiflavor Neutrinos from Supernovae. OMNIS involves collaborators from a dozen universities and national laboratories. It was originally designed 5, 6 to utilize nuclei in rock to convert neutrinos into neutrons, which would then be detected. However, the conversion efficiency of possible types of rock was found to be much less than that of iron or lead 7 . Thus OMNIS will consist of 4 kT of lead and 10 kT of iron. Lead is an efficient converter of the neutrinos to neutrons, as its threshold for neutron emission via NC interactions is only 7.37 MeV. In addition, ve's can interact with lead through the CC interaction to produce 208 Bi, a process with a higher threshold for neutron emission of 9.77 MeV, but a larger cross section. A SN at a distance of 8 kpc should produce 8 about 880 neutrons per kT of lead for all flavors for the standard model SN neutrino spectra, but primarily from NC interactions induced by J^'s, v^s, z/T's and i/ r 's. Sufficiently highenergy neutrinos can produce events in which two neutrons are emitted in lead (threshold = 14.98 MeV). The yield of events from the CC process is expected to be only ~55 events per kT in the lead, but that yield is extremely dependent on the energy of the ^ e 's and i>e's, a point discussed further below. Iron, by contrast, has a high threshold for neutron emission via NC interactions, 11.20 MeV, and negligible production for CC processes. This results in a lower efficiency, but a simpler calculation of its yield than for lead. Iron should produce 9,10 around 140 neutrons per kT for the standard model neutrino spectrum for a SN at 8 kpc, with virtually all events coming from t/^'s, O^s, i/T's and i>T's. The two types of OMNIS converters will thus provide three time dependent spectra: two from the single- and two-neutron events from lead and the third from iron. The yields anticipated from a SN at the Galactic center, i.e., at 8 kpc from earth, are indicated in Table 1. They assume a neutron detection efficiency determined n from detailed Monte-Carlo simulations. The yields anticipated for some other potential detectors of SN neutrinos are also
194 Table 1: Yields of Supernova Observatories from an 8 kpc Distant Supernova
Detector SuperK SNO SNO OMNIS OMNIS no osc.
Target Material H20 H20 D20 Fe Pb
Mass (Ton) 32000 1600 1000 10000 4000
Target Element p, e, O p, e, O d, e, O Fe
ZVT-^e
^e
Ve
Yield 180 16 190 20
Yield 8300 520 180 40
vti-\-vii-\-vTJr vT Yield 50 6 300 500
160 <4420
70 40
1400 640
indicated. As can be seen, SuperKamiokande will produce the most events, but they will almost all be ve induced. OMNIS will provide, in addition, a statistically large sample of v^, v^, i/T, and i>T induced events. OMNIS will be sited in the Center for Applied Repository and Underground Science, CARUS, 700 m underground, in New Mexico. The detectors will consist of alternating slabs of lead or iron and racks of scintillators with PM tubes at both ends. The ends of the modules will be covered by doors of lead or iron respectively, which will slide on rails to allow access to PM tubes or removal of scintillators for maintenance. The scintillator will contain 0.1% of Gd to produce the signatures of neutron induced events. 2.2
Fast Timing Capability of OMNIS
An important feature of OMNIS is its fast timing capability. Once neutrons are produced they lose little energy until they reach the scintillator, where they lose most of their energy in a few scatterings from the protons therein. Monte-Carlo simulations 11 have shown that these occur in less than 200 ns. The subsequent thermalization of the neutrons requires ~30 //s, when they are captured by a Gd nucleus, producing four 7-rays with a total energy of almost 8 MeV. The combination of these fast-slow signals produces both the signature for the neutron-induced events, and OMNIS' fast timing capability. 3 3.1
Astrophysics and Physics from OMNIS Astrophysics of Stellar Collapse
The mechanisms by which neutrinos are produced in SNe are inevitable, but the energies of the neutrinos are not. A low entropy collapse is essential to
195 produce the long trapping time observed in the neutrinos from SN 1987a, but those were observed only for a small sample of i>e's. Neutrino b r e m s s t r a h l u n g 1 2 can affect both the number and energies of the neutrinos emitted, while inelastic s c a t t e r i n g 1 2 ' 1 3 and convection can affect their energy distributions. In addition, the energy distributions may be "pinched" at both the high- and lowenergy ends in dense stellar cores 1 4 . Information about these effects could be obtained from a high-statistics observation of SN neutrinos. This information is i m p o r t a n t to using the subsequent neutrinos for, e.g., neutrino mass measurements and black hole collapse diagnostics, in addition to understanding the processes of stellar collapse and cooling. 3.2
Measuring
Neutrino
Mass
Because of O M N I S ' intrinsic timing capability, the timing of the onset of any neutrino luminosity curve will probably be limited by statistics. This would be ~ a few ms for a SN at the Galactic center 1 5 even so. This level of timing could produce a measurement of neutrino mass from the effect of the time-of-flight of the neutrinos on their distributions. If one neutrino is very light, the arrival time difference between those neutrinos and those of larger mass m and energy E, for a SN at distance D, is At = 0M5[(m/eV)/(E/MeV)]2{D/l0kpc))
(3)
A SN at the Galactic center would determine the onset of the distributions to a few ms, which in turn would determine 1 5 the mass of the heavier neutrino to 20-30 eV. This level of accuracy could not be achieved in any other way. 3.3
Sensitivity
to
Oscillations
T h e lead yield is especially sensitive to v^ —y i/e or vT —>• ve oscillations; either would transfer the energy of the i/^'s or z/T's to f e ' s , so would produce much more energetic ve's t h a n would be expected from the SN. If maximal mixing occurred, the yield would be enhanced in l e a d 8 by about a factor of four, and the two-neutron event yield by a factor of 40, as is indicated in Table 1. T h u s the ratio of one- to two-neutron events from the lead would produce a clear signature of such oscillations. 3.\
Diagnosing
Collapse to a Black
Hole
Perhaps the most dramatic manifestation of O M N I S ' fast timing would be the abrupt termination of the neutrino luminosities if the stellar collapse went fairly promptly to a black hole. Because the i/e's and ^ e ' s would be trapped
196 farther from the center of the star than would the v^s v^s, t/ T 's, and i>T's, the ve and ve luminosities would terminate after those of the u^s, i>^s, vT's and vT's, since the black hole would grow outward from the center of the star. This difference has been estimated 16 ' 17 to be of the order of one ms. It might depend on stellar geography, i.e., the locations of the neutrinospheres 18 . It might also allow for diagnosis 18 of the details of collapse to a black hole, including whether the collapse was triggered by a softening of the nuclear equation of state resulting from deleptonization 17 or from enough infall for the core to exceed the Chandrasekar mass 19 , as the luminosities would be expected to vary in different ways just prior to the collapse. 4
Frequency of Galactic Supernovae
The historical record identifies one or two Galactic SNe per century. However, a more careful examination of those events shows that they all occurred in the 10% of the Galaxy nearest our solar system 2 0 ' 2 1 . This suggests that most SNe were obscured from observation in optical photons by intervening dust, so that the actual rate may be one every 10 to 20 years. 5
Results from a Close Supernova
A SN could occur much closer to earth than the Galactic center; Betelgeuse, a 30 M0 red giant, is some 30 times closer to earth. It would produce several million events in OMNIS if it became a SN during OMNIS' lifetime. This would not necessarily improve the neutrino mass measurement, as the statistics would improve, but the time-of-flight difference would decrease. But, if the SN went fairly promptly to a black hole, this would again allow for an unprecedented time determination, hence improved neutrino mass measurement, and an extraordinary diagnostic of the event itself. However, OMNIS (and other potential SN neutrino detectors as well!) must be designed to accommodate the very high count rate from a close SN if its benefits are to be realized. Thus OMNIS will be segmented electronically to insure that the signal rate from a close SN would not saturate it. 6
Conclusions
The unique astrophysics and physics that could be obtained from a statistically large sample of neutrinos from the next Galactic SN, including luminosities of all neutrino flavors, argue strongly for building an observatory to provide those data. We are planning OMNIS to fulfill that need. Furthermore, the fast timing characteristics of OMNIS give it the capability to measure neutrino masses at
197 levels that would be difficult to achieve with any other technique, as well as possibly to diagnose the process of collapse to a black hole. Acknowledgements NSF support through grant PHY9901241 is gratefully acknowledged. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
A. Burrows, Ann. Rev. Nucl. Part. Sci. 40, 181-212 (1990) Y.-Z. Qian et al, Phys. Rev. Letters 71, 1965 (1993) K.W. Hirata et al, Phys. Rev. Letters 58, 1490 (1987) R.M. Bionta et al, Phys. Rev. Letters 58, 1494 (1987) D.B. Cline, et al, Astrophys. Lett. Commun. 27, 403-409 (1990) D.B. Cline, et al, Phys. Rev. D50, 720-729 (1994) P.F. Smith, Astropart. Phys. 8, 27-42 (1997) G.M. Fuller, W.C. Haxton, and G.C. McLaughlin, Phys. Rev. D59, 085005-1 (1999) S.E. Woosley et al, Astrophys. J. 356, 272-301 (1990) E. Kolbe and K. Langanke, submitted to Phys.Rev. (2000) J.J. Zach et al, private communication (1999) S. Hannestad and G. Raffelt, Astrophys. J. 507, 339 (1998) H.-T. Janka et al, Phys. Rev. Letters 76, 2621 (1996) T. Totani et al, Astrophys. J. 496, 216 (1998) J. Beacom, private communication, 2000 A. Mezzacappa, private communication, 2000 T.W. Baumgarte et al, Astrophys. J. 468, 823 (1996) J. Beacom, R.N. Boyd, and A. Mezzacappa, private communication, 2000 C.L. Fryer, W. Benz, and M. Herant, Astrophys. J. 460, 801 (1996) P.M. Dragicevich, D.G. Blair, and R.R. Burman, Mon. Not. R. Astron. Soc. 302, 693 (1999) K. Hatano, A. Fisher, and D. Branch, Mon. Not. R. Astron. Soc. 290, 360 (1997)
LEAD PERCHLORATE AS A N E U T R I N O D E T E C T I O N MEDIUM S. R. ELLIOTT, P. J. DOE, R. G. H. ROBERTSON University of Washington, Seattle, WA 98195, USA E-mail: sreQu.washington.edu C. PAUL Raytheon, 2200 East Imperial Highway, El Segundo, CA 90245, USA E-mail: [email protected] We introduce lead perchlorate as a Cerenkov medium for neutrino detection, describe the principal neutrino reaction signatures and discuss applications of such a detector.
1
Introduction
Because of the high cross section, lead is an attractive neutrino detection medium and has been proposed for use in studies of supernovae l'2 and neutrino oscillations3. This interest has stimulated new calculations of the neutrino cross sections 1,4 ' 5 . The detectors proposed for these studies are typically of a segmented construction employing proportional counters interleaved with liquid or plastic scintillators. These techniques do not generally lend themselves to realizing massive, high resolution, cost effective detectors. Lead perchlorate (Pb(C104)2) dissolves readily in water yielding a high density, transparent liquid. This raises the possibility of building massive detectors using the well demonstrated water Cerenkov technique. The detector would be capable of measuring not only charged particles and gamma rays but also neutrons which are produced in the v-Vb interactons. These neutrons are detected via the gamma rays emmitted as a result of neutron capture on the chlorine contained in Pb(C104)-2- The ability to simultaneously measure the energy, spatial and temporal information of charged particles, gamma rays and neutrons make this a potentially powerful detector technique. 2
Lead as a Neutrino Target
The neutrino interactions of interest in Pb may occur by either the charged current (CC) or the neutral current (NC) reactions given below 198
199 ve + 2 0 8 Pb ->
208
Bi* + e~
(CC)
I
207
Bi + X-y + Yn
vx + 2 0 8 Pb -> 208P&* + */
(7VC)
'^Pb + X-y + Yn The number of neutrons produced, 0,1 or 2, depends on the neutrino energy and the particular reaction. The CC reaction proceeds mainly through the Gammow-Teller resonance, which occurs at approximately 16 MeV. At 30 MeV, the energy regime of interest to supernova and accelerator beam stop neutrino studies, the CC cross section is approximately 600 times that of carbon, while the NC cross section is approximately 100 times that of carbon. The ve cross section on Pb is quite small, however solutions of Pb(ClC"4)2 contain large quatities of hydrogen and the reaction: ve + P ->• n + e+ has a significant rate. The NC and CC events are distinguished from each other by the amount of prompt j + e~ energy. NC events only have significant prompt energy when no neutrons are produced. Thus events that have significant prompt energy and are accompanied by one or more neutrons are candidates for ve or i>e interactions. The ve and ve events would be separated off-line. Since only the ve can produce two neutron events, these events provide information on the ve energy spectrum. The energy spectrum of the one neutron events is a mixture of ve and ve and one would extract the information on the ve by comparison of the one and two neutron prompt energy spectra. The presence of hydrogen in the Pb(C104)2 solution results in the neutrons quickly thermalizing and capturing on 35 C1, which has a high neutron capture cross section of 44.0 b, in approximately 10 /usee. The neutron capture results in multiple gamma rays totaling 8.6 MeV which are detected by way of the Compton scattered electrons. 3
Relevant Properties of Lead Perchlorate
An 80% solution can be made by dissolving approximately 500g of Pb(C104)2 in lOOg of water 6 . The resulting transparent liquid has a density of 2.7 gm/cc and a refractive index of 1.5. The refractive index as a function of the strength of the Pb(C10 4 ) 2 solution is given in Fig. 1 An energetic electron will produce about 185 Cerenkov photons per centimeter in the wavelength region to which photomultipliers are generally sensitive. The stopping power of the 80% solution is approximately 0.2 cm/MeV,
200 1.55 T
c t>
.
15
y /
.»
1.45 -
°
14
/
/
/
^
JK^
_C
^ ^ - ^
1.35
^_____—-•*-"""^
1.3 H
1
1
1
1
0
20
40
60
80
'
Concentration (% by mass) Figure 1. The variation of the refractive index as a function of the strength of the Pb(C10 4 )2-
resulting in a photon yield of around 37 photons/MeV. To realize a large C'erenkov detector with the photomultipliers distributed around the periphery requires that the liquid has excellent transparency in the wavelegths of interest. We have made measurements of the transmission and attenuation of light in an 80% solution of Pb(ClC>4)2. No attempt was made to either purify or filter particulates from this liquid prior to making the measurements. Fig. 2 shows the spectral transmission of light through a solution of Pb(C10 4 )2 in the region of 200 to 650 nm. The transmission is referenced to a sample of deionized water. It is encouraging to note that in this unpurified sample there are no significant absorption lines. To measure the attenuation length we constructed a 1 m tall cell through which monochromatic, collimated light from a light emmitting diode was passed. The transmitted light was recorded by measuring the base current from a photomultiplier tube as the height of the liquid was varied in the cell. Fig. 3 shows the results obtained by passing 460 nm light through an unpurified solution of Pb(C104)2The attenuation length of 43 cm was unexpectedly low and is insufficient for a large C'erenkov detector. Reducing the concentraton of the Pb(C104)2 solution to 40% using deionized water with an attenuation length of greater than 20 m, resulted in a further drastic reduction in attenuation length of the solution. This suggests that the problem is due to the formation of insoluble
201
200
300
400
500
600
700
Wavelength (nm) Figure 2. Spectral transmission through a 1 cm cell filled with an 80% solution of Pb(C104)2 referenced to deionized water.
15
25
35
45
55
65
Path Length (cm) Figure 3. The attenuation of 460nm light in an 80% solution of Pb(C104)2.
lead hydroxide in suspension in the liquid, which scatters light from the liquid cell.
202
4
Discussion
We are currently working to understand the processes that influence the light transmission in Pb(C104)2. A modest improvement in the attentuation length would be appropriate for an experiment at a stopped pion neutrino source such as ISIS at the Rutherford-Appleton Laboratory or the proposed OrLAND neutrino facility at the Oak Ridge Laboratory Spallation Neutron Source (SNS) 7 . Such a detector would simultaneously measure the i/-Pb cross sections and search for v^ -» ue- oscillations as reported by the LSND experiment. The signal for this oscillation would be a distortion in both the time and energy spectrum of the ue. Since Ve are not appreciably present at a beam stop v source one can consider a ve appearance search, selecting for ve in the manner described earlier. With this demonstration of the detector technology plus a measurement of the i/-Pb cross sections, one would be in a position to consider a massive Pb(C104)2 based supernova detector. Acknowledgments We would like to think Edwin Kolbe, Gail McLaughlin, George Fuller, Wick Haxton, Richard Hahn and Geoff Miller for useful conversations. This research was supported by grant from the Department of Energy. References 1. C. K. Hargrove, et al., Astroparticle Physics 5, 183 (1996). 2. D. B. Cline et al, Phys. Rev. D50, 720 (1994); P. F. Smith, Astroparticle Physics 8, 27 (1997). 3. C. K. Hargrove and R. T. Siegel, "A Proposal to Measure the i/-Pb Charged and Neutral Current Cross Sections for Neutrinos from Stopped Pion Decay", 15 January 1998, (unpublished). 4. George M. Fuller, Wick C. Haxton, and Gail C. McLaughlin, Phys. Rev. D59, 085005 (1999). 5. E. Kolbe, K. Langanke, arXiv:nucl-th/0003060, 27 March 2000. 6. CRC handbook of Chemistry and Physics, 66th edition, CRC Press, Inc. 7. B. D. Anderson, F. Avignone et al., A Large Neutrino Detector Facility at the Spallation Neutron Source in Oak Ridge Natonal Laboratory. Proposal to the US Department on Energy (in preparation).
DIRECT NEUTRINO MASS MEASUREMENT WITH A SUPERCONDUCTIVE DETECTOR M.R. GOMES, P. VALKO AND TA GIRARD Centro de Fisica Nuclear, Universidade de Lisboa, 1649-003 Lisboa, Portugal Email: [email protected] We briefly describe the motivations for, and possible advantages of, a superconductive measurement of the 187 Re decay in rhenium, and provide preliminary experimental results with a 25 micron thick geometrically-metastable strip.
1
Introduction
The best direct limits on electron neutrino mass currently derive from precision measurements of the near-endpoint tritium decay spectrum 1 , which also continue to yield an endpoint anomaly. These are generally mass-limited in order to avoid the effects of multiple event summing (pileup) into the interest region of the spectrum, and have considerable difficulty to achieve resolutions of 10 eV near the decay spectrum endpoint. The Genova group 2 , and more recently Milan 3 , have pursued the alternative measurement of 187 Re decay using cryogenic ^-calorimeters towards addressing the rate limitations: the decay is characterized by an endpoint energy of ~ 2.6 keV, and the fraction of decays into the last 10 eV of the spectrum is three orders of magnitude larger than the tritium experiments. The /z-calorimeter essentially consists of a thermistor mounted on a natural rhenium substrate, operated at ~ 100 mK. Results to date impressively demonstrate a full undistorted decay spectrum down to 100 eV, with a resolution of ~ 13 eV at 5.9 keV. The timing resolution is however of order 10 ms, so that these measurements are pileup-limited to a rate of 3 Hz. Although rhenium is a type I superconductor with T c = 1.7 K and H c (0) = 210 G, both the Genova and Milan measurements make use of only the reduced specific heat of the material. We here describe an alternative measurement technique which relies on the nucleation of normal (N) state flux tubes in a geometrically-metastable Type I superconducting (S) strip by decay-induced heating of the material 4 ; pulses induced in a surrounding loop by the flux changes associated with the loss of the Meissner effect are detected to yield the interaction signal. The timing resolution of the acquisition electronics is however of order 10 fis, implying a data acquisition rate already some 102 larger than bolometers. Timing resolutions of order 10 ns have been previously demonstrated 5 , suggesting the ability to eventually accomodate event rates 203
204 c
NORMAL
]
- - V . . AT AH fp
I I 1 1
SUPERCONDUCTING
temperature
"~"~---\
c
Figure 1. Phase diagram of a geometrically-metastable superconducting foil in a perpendicular magnetic field.
105 larger than the bolometers. 2
The Geometrically-metastable Superconducting Detector
The S—»N phase transition of a type I superconducting foil in a perpendicular magnetic field is initiated when the field at the foil edge reaches the thermodynamic critical field (H c ); the applied field is however well below this, owing to the geometry-dependent demagnetization of the foil6. Moreover, the first flux to enter at H/ p is driven to the foil center by the Gibb's free energy potential, lowering the edge field and requiring increase of the applied field to generate further flux entry. In consequence, the transition is delayed over a large AH, and the foil is generally in an intermediate (or mixed) state. During the phase transition, local heating generated by energy-deposition of incident radiation may also nucleate the normal state, if the AT is sufficient to raise the temperature above the phase line as shown in Fig. 1. The signal is then generally composed of two distinct contributions; these can be resolved by the insertion of a pause during the field increase, which yields the signal shown in Fig. 2. This can be deconstructed by S(t) = S M + ST, with Sj(t) = Sj(l-e-^).
(1)
where the first term corresponds to the magnetic contribution, and the second to the thermal. Fitting of the data typically yields TM=0-29±0-06 S, corresponding approximately to the damping constant of the field magnet. The saturation of the curve implies a time-dependent measurement efficiency (e), which can be expressed as
205
H P
2 103 1 10* 0
time (s) Figure 2. Timing curve of transitions during a pause inserted in the field ramp.
e(t) = f[l~rT(A.t)-l(l-e-^)].
(2)
where At is the pause time and So is the source activity, yielding e ~ 70% with typical experimental values. In general, pulse-height = d^/dt ~ j B . d S = /xoHcS,,>, where S^ is the surface area of the nucleated flux tube and flux variation are assumed to occur on superconducting time scales. Measurements with 109 Cd conversion electron irradiations of a tin device have shown the pulse-height to be linear with energy, such that AE/LyS^ = constant 7 . Energy resolution is thus predicated on the ability to resolve flux bundle size. 3
Preliminary Results
In Fig. 3, we show a typical thermal-only spectrum, displayed as the variation of [events]1^2 with energy. This was obtained with a 99.99+% pure, 25 /jm thick natural polycrystalline ribbon, with width and length 0.900±0.030 and 15±0.01 mm, respectively. The foil was zero field cooled to T = 330±10 mK; the field ramped up from zero to 88 G at a rate of 16 G/s, stopped for 20 s, and then continued to 250 G, well above H c (210 G) before being returned to zero. The spectrum exhibits a linearity down to 1.2 keV, below which it becomes dominated by noise. The integrated thermal rate is ~ 8.5 Hz over the pause period, comparable with the 12 Hz anticipated from the device volume. Also shown, in comparison, is a recent 2 bolometer spectrum of the decay, normalized to the events of the present work via a multiplication factor of 15.
206
0
20
40
60
80
100
threshold (mV)
Figure 3. Preliminary decay spectrum of
187
Re.
Figure 4. Preliminary 1 8 7 R e spectrum obtained from repeated threshold variation measurements under otherwise identical conditions at a pause field below H^ p .
The 17.5 mV noise limit in these measurements corresponds to flux tubes of 600 (/>(,, or ~ 660 eV. The threshold step-size in these measurements was 5 mV, equivalent to 135 eV, insufficient to resolve the Ka, K@ lines of the 55 Fe decay; the minimum step-size permissable with present equipment is 0.1 mV, implying an ultimate step-size of ~ 2<^o ~ 2.7 eV. While these results are encouraging, we stress that the basic physics underlying the foil response to irradiation is not yet understood. Experiments in which the pause is inserted below H/ p in fact yield similar pulse height spectra, as shown in Fig. 4, with linear timing spectra. Although this is consistent with a recent description of explosive nucleation of the normal state with
207
laser irradiation by Ghinovker et. al. 8 , the description neglects the geometric barrier and its influence on the signal response remains to be elaborated. 4
Summary
Preliminary measurements of the decay of 187 Re in thin geometricallymetastable foils appears to yield an electron spectrum with linear response down to ~ 1 keV. The current device is two orders of magnitude faster than bolometers, and can be improved by another three orders. The noise level can be reduced by use of cooled, low noise preamplification electronics. Energy resolution in this device is predicated on the ability to resolve flux bundle size, which has been shown to be linear in the deposited energy 7 . The minimum step-size permissable with present equipment suggests an ultimate sensitivity of ~ 2.7 eV. Implementation of a SQUID readout, sensitive to variations of 10~3(j>o, would yield an ultimate single o resolution of ~ 1 eV, but at the expense of fast data acquisition. Acknowledgments We thank J. Collar, D. Limagne and G. Waysand of the Groupe de Physique des Solides, Universites Paris 7/6 for technical discussions, and C. Oliveira of ITN-Sacavm for MCNP simulations of the calibrations. This work was supported in part by grant CERN/FAE/1211/98, under the Ministry of Science & Technology of Portugal. References 1. Review of Particle Properties, Eur. Phys. J. C 3 , 1 (1998). 2. F. Fontanelli, F. Gatti, A. Swift and S. Vitale, Nucl. Instr. & Meth. A370, 247 (1996). 3. A. Alessandrello et. al., Phys. Lett. B457, 253 (1999). 4. V. Jeudy et. al., Journ. Low Temp. Phys. 93, 515 (1993). 5. M. Furlan et. al., in Low Temperature Detectors for Neutrinos and Dark Matter-Ill (Ed. Frontiers, Gif-sur-Yvette, 1991) 21. 6. H. Castro, B. Dutoit, A. Jacquier, M. Baharami and L. Rinderer, Phys. Rev. B59, 596 (1999-1). 7. V. Jeudy et. al., Nucl. Instr. & Meth. A373, 65 (1996); Nucl. Instr. & Meth. A370, 104 (1996). 8. M. Ghinovker, I. Shapiro and B. Ya. Shapiro, Phys. Rev. B59, 9514 (1999-11).
208
Nuclear Spin Isospin Responses and Spectroscopy of /?/? Rays from 100 Mo for Neutrino Studies in Nuclei H. Ejiri" NPL, Dept. Physics, University of Washington, Seattle, WA. 98195, USA Nuclear spin isospin responses for neutrino(i/)'s in nuclei can be well studied by means of the intense v beam from SNS and the highly efficient detector of ORLaND. They provide a unique opportunity for exclusive studies of the nuclear responses for J/'S, which are crucial for v mass studies by double beta decays(/3/3) and for solar and supernova v studies by inverse beta decays. Spectroscopic studies of two 0 rays from 1 0 0 M o are of potential interest for u mass studies by 00 decays and realtime studies of low energy solar u 's by inverse 0 decays.
1
Nuclear Spin Isospin Structures and Responses for Neutrinos
Nuclei, which are quantum systems of nucleons, are used as excellent microlaboratory for studying elementary particles and fundamental interactions of astroparticle physics interests. Nuclear isospin and spin-isospin responses for the vector and axial-vector weak interactions are crucial for studying neutrinos(zz) and nuclear weak processes of astroparticle physics interests 1 . Nuclear spin isospin interactions give rise to spin isospin giant resonances at the high excitation region of Eex = 10 ~ 30 MeV, and spin isospin core polarizations at the low excitation region of Eex = 0 ~ 5 MeV. Consequently nuclear spin isospin responses for weak, electromagnetic and strong processes are modified much in nuclei 2 . It is of great interest to study the nuclear spin isospin structures in a wide excitation region of Eex = 0 ~ 50 MeV * 2 . Nuclear weak processes are used for studying neutrino physics beyond the standard model(SM) of SU(2)L X U(l) and new aspects of astronuclear processes in the sun and stars. Actually, Majorana v masses, right-handed weak interactions, Majoron v couplings and other neutrino interactions, which are beyond SM, are effectively studied by nuclear double beta decays(/?/3). Here the nuclear spin isospin responses(matrix elements) are crucial for extracting the v properties and the weak interactions. Solar and supernova v's provide important information on v oscillations, v masses and stellar evolution mechanisms. They are studied using nuclear inverse /? decays by the charged-current interaction and nuclear inelastic scatterings by the neutral-current interaction. Nuclear weak processes are important also for astronuclear reactions and nu"On leave from RCNP, Osaka University,Ibaraki, Osaka 567, Japan.
209 clear synthesis. T h e n the nuclear spin isospin responses for the solar and supernova z/s and for the astronuclear weak processes are essential for the studies of the v properties and the weak processes. T h e weak, electromagnetic and strong interactions involve spin and isospin dependent terms, and accordingly the weak, electromagnetic and strong processes depend largely on the nuclear spin isospin structures. T h e y are modified in nuclei by nuclear spin isospin medium effects which reflect the nuclear spin isospin structures. T h e nuclear spin isospin structures are important for studies of the nuclear weak, electromagnetic and strong processes. T h e intense SNS v beam, together with the highly efficient detector of O R L a N D 3 , provides excellent opportunities for selective studies of the nuclear spin isospin structures and the spin isospin medium effects, and thus for studies of the nuclear spin isospin responses for the weak, electromagnetic and strong processes. T h e nuclear responses concerned with the astroparticle physics are mainly in the low excitation region of Eex — 0 ~ 50 MeV. This is just the region covered by the SNS v beam and O R L a N D . T h e spin(cr) isospin(r) operator is in general expressed as TTSLJ
= 9%SLjTT[rLYL
x
(1)
where T,S,L, and J are isospin, spin, orbital angular m o m e n t u m and total angular m o m e n t u m , respectively, and YL is the spherical harmonics. T h e coupling constant is given by gj-sLJ w i t h a standing for the weak, electromagnetic or strong interaction. T h e vector and axial-vector weak operators are given, respectively, by the isospin and spin-isospin operators of T I O L J and T\n,j with T = T± for the charged-current interaction and T = T3 for the neutral-current one. T h e electric and magnetic operators are written similarly by the spin isospin operators with the isospin T = T 3 . T h e nuclear weak responses investigated by nuclear /? decays are limited to the charged-current ground state transitions. Hadronic nuclear reactions have been used extensively to study nuclear spin isospin responses in a wide excitation region, charge-exchange spin-flip and non spin-flip reactions for the charged-current spin-isospin and isospin responses and inelastic scatterings for the neutral-current responses. T h e hadronic processes, however, involve necessarily strong interactions mediated by various kinds of mesons, and accordingly the spin isospin interaction is not just the simple one of the central spin isospin interaction. T h e interaction strength itself is strong and thus the reaction proceeds by multistep processes as well as the direct single step process. Projectile particles and emitted particles in nuclear reactions are distorted by the strong nuclear potential. Therefore it is not straightforward to deduce the spin isospin responses from hadronic reactions. In fact, they are used only for limited cases
210
of strong spin-isospin excitations with TSL J=1101 modes in the T± channel. Electromagnetic probes are used to study nuclear spin isospin structures and nuclear spin isospin responses. These electromagnetic probes, however, include isoscalar currents and orbital currents as well. In general it is not easy to separate experimentally these contributions from isovector spin currents relevant to the weak interaction. The isospin spin responses studied by the electromagnetic probes are only for those in T3 channel. The v probe is an excellent and ideal probe for studying nuclear spin isospin structures and nuclear spin isospin responses of the nuclear and astroparticle physics interests. It can probe directly for the spin isospin structures and the spin isospin weak responses through the charged- and neutral-current interactions. The v projectile is free from the nuclear distortion and multistep processes since the interaction itself is very weak. The reaction rate with the v beam is extremely small. The intense SNC v beam and the highly efficient ORLaNd make it possible to study directly the spin isospin structures and the spin isospin responses. 2
Nuclear Spin Isospin Responses Studied by S N S / O R L a N D
Nuclear spin isospin structures and nuclear weak responses to be studied by means of the SNS v beam and ORLaND are as follows. 1. Spin isospin responses for the charged- and neutral-current weak interactions of the isospin (TSLJ = 10LL) and spin-isospin(T5LJ = IILJ) modes with L = 0,1, 2 in a wide excitation region. They elucidate nuclear spin isospin structures of the nuclear and astroparticle physics interests, including spin isospin core polarizations and spin isospin giant resonances with 5 = 1,0 and L=0,l,2 in both T = T± and T 3 channels. 2. Spin-isospin responses for 71 Ga, 100 Mo and other nuclei used for solar v studies. The v probe gives straightforwardly the spin isospin responses for the ground and excited states of these nuclei. They are important for quantitative studies of solar-^ oscillations. 3. Spin isospin responses with 1=0,1 for 1 6 0 , 100 Mo, 2 0 8 Pb and other nuclei used for supernova v studies. The spin isospin responses in a wide excitation region of the 7± = 1 and T3 channels are essential for quantitative studies of v oscillations, v masses, stellar evolution mechanisms, nuclear synthesis processes and others of astroparticle physics interests. 4. Nuclear spin isospin strength distributions with L = 0 ~ 5 in the T_ channel for 100 Mo, 116 Cd, 136 Xe, 150 Nd and other nuclei used for studies of neutrinoless double beta decays(0^/?/3) and those in the T + channel for daughter nuclei. The responses are related to the nuclear matrix elements associated
211
with Of/?/?, which are used for studying v masses and the weak interactions. 5. Comparison of neutrino weak processes with corresponding hadronic and electromagnetic processes to investigate reaction mechanisms involved in the hadronic and electromagnetic processes. 3
Spectroscopy of Two j3 Rays from
100
Mo for Electron Neutrinos
It is of great interest to study directly u mass with sensitivity down to ~0.03 eV 4 . Double beta decay may be the only probe presently able to access such small v masses. Actually, observation of Of/?/? would identify a Majorana-type electron v with a non-zero effective mass < mv > x 5 6 7 . The Of/?/? process is sensitive not only to the v mass (< mv >), but also to the right-handed weak current, the Majoron neutrino coupling and other terms beyond SM 5 6 7 8 . Spectroscopic studies of the energy and angular correlations for two /?-rays are useful to identify the terms responsible for Of/?/?. Low energy solar-f studies, so far, have been carried out with 7 1 Ga and 37 C1 detectors 9 10 u . They are non-realtime and inclusive measurements that do not identify the f sources in the sun. Realtime spectroscopic studies of low energy solar v are important for studies of solar-f oscillation. Recently spectroscopic studies of two /? rays from 100 Mo are shown to be of potential interest for investigating both the Majorana v mass by Of/?/? and low energy solar f's by inverse /? decay 4 . Coincidence studies with a multiton 100 Mo detector for correlated /?/? from Of/?/?, together with the large Q value(<5/3/3), permit identification of the f-mass term with a sensitivity of ~ 0.03 eV. Correlation studies of the inverse 3 and the successive /?-decay of 100 Tc, together with the large capture rates for low energy solar f's, make it possible to detect in realtime individual low energy solar f's in the same detector. The isotope 100 Mo is just the one that satisfies the conditions for the /?/? — v and solar-f studies. The large Q value of (5/3/3=3.034 MeV gives a large phase-space factor G°" to enhance the Of/?/? rate and a large energy sum of E1 + E2 = Qpp to place the Of/?/? energy signal well above most BG except 208 T1 and 214 Bi. The energy and angular correlations for the two /?-rays can be used to identify the f-mass term 8 . The low threshold energy of 0.168 MeV for the solar-f absorption allows observation of low energy sources such as pp and 7 Be. The GT strength to the 1+ ground state of 100 Tc is measured to be (gA/gv)2B(GT)=0.52±0M 1 2 13 by both charge-exchange reaction and electron capture . Capture rates are large even for low energy solar f's. The rates are 639SNU and 206SNU for pp-f and 7 Be v, respectively 4 . The solar-v sources are identified by measuring
212
the inverse-/? energies. Only the 100 Tc ground state can absorb Be v and pp v. Therefore the ratio of pp-i^ to 7Be-i/, which is sensitive to the j/-oscillation parameters, can be determined without ambiguity of the B(GT) value for the ground state. The measurement of two /3-rays (charged particles) enables one to localize in space and in time the decay-vertex points for both the Ov/3/3 and solar-i^ studies. The tightly localized /?-/? event in space and time windows, together with relevant /? , 7 and X ray measurements, are key points for selecting Oi/fi/3 and solar-f signals and for reducing correlated and accidental BG by factors 1 0 - 5 ~ 1(T 6 as shown in ELEGANT detectors 8 . One possible detector for this experiment is an ensemble of plastic scintillator modules. It may consist of 2000 modules, each with x=6m, y=6m, z=0.25cm. Thin 100 Mo foils with thickness of 0.05 g/cm 2 are interleaved between the modules. Scintillation signals of two /? rays from 100 Mo are read by WLS(wave length shifter)fibers stretched at both x and y directions in each module so as to get adequate energy and position resolutions. WLS read outs have been well developed 14 15 , and Mo foils with realistic impurity level of 1 ~0.1 ppt can be used 16 . The detector can be used also for supernova v studies, other rare nuclear processes, and for other isotopes. Another option is a liquid scintillator 3 in place of the solid one, keeping similar configurations of the WLS readout. The energy and spatial resolution are nearly the same. Then 150 Nd with the large Qpp may be used either in solid or solution in the liquid scintillator for 0uf3/3. Of particular interest is 136 Xe because liquid Xe is a scintillator. The present spectroscopic method for 0^/?/?, which measures selectively the v mass term, is complementary to the calorimetric /?/? measurements with high energy resolutions for 7 6 G e 1 7 and 1 3 0 Te 1 8 . The present method provides important data for solar u, which are supplementary to existing geochemical and planned realtime experiments 1 9 2 0 . The author would like to express his hearty thanks to Prof. Frank T Avignone III for valuable discussions, strong encouragements, and for the strong leadership in fundamental nuclear and astroparticle physics. The author thanks NPL and INT, University of Washington for support and discussions. 1. 2. 3. 4.
H .Ejiri, Int. J. Mod. Phys. E6 No 1 (1997) 1; Phys. Rep. C (2000). H. Ejiri and J.I. Fujita, Phys. Rep. 38 C (1978) 85. F. Avignone III et al., ORLaND Coll, ORLaND proposal (1999). H. Ejiri, J. Engel, R. Hazama, P. Krastev, N. Kudomi, and R.G.H. Robertson, nuclexp/9911008 v2, 23 Nov 1999.
213 5. W. C. Haxton and G. J. Stephenson Jr, Prog. Part. Nucl. Phys. 12 (1984) 409; M. Doi et a l . Prog. Theor. Phys. 83 (Suppl.)(1985) 1; M. Moe and P. Vogel, Ann. Review Nucl. Science 44 (1994) 247. 6. F.T. Avignone III and R.L. Brozinski, in Neutrino, ed H. Klapdor (Springer-Verlag 1988) p.147. 7. A. Faessler and F. Simcovic, J. Phys. G 24 (1998) 2139; J.Suhonen and O.Civitarese, Phys. Rep. 300 (1998) 123. 8. H. Ejiri et a l , Nucl. Phys. A611 (1996) 85; Nucl. Instr. Method, A302 (1992) 304; H. Ejiri et a l , J. Phys. Soc. Japan Lett. 65 (1996) 7. 9. J. N. Bahcall and M. Pinsonneault, Rev. Mod. Phys. 64 (1992) 885, and 67 (1995) 781; J. N. Bahcall et a l , Phys. Lett. B433 (1998) 1. 10. GALLEX Coll, W. Hampel, et a l , Phys. Lett. B388 (1996) 384; SAGE Coll, J. N. Abdurashitov, et a l , Phys. Rev. Lett. 77 (1996) 4708. 11. B. T. Cleveland et a l . Astrophysics J,496 (1998) 505. 12. A. Akimune, et a l , Phys. Lett. B394 (1997) 23. 13. A. Garcia et a l , Phys. Rev. C47 (1993) 2910. 14. The MINOS Collaboration, NuMI-L-337. 15. A.Konaka, Proc.NNN99 workshop, Sept.1999, Stony Brook, ed.C.K.Jung and M.Diwan (1999). 16. R.G.H.Robertson, Prog. Part. Nucl. Phys. 40 (1998) 113. 17. L. Baudis et al. Phys. Rev. Lett. 83 (1999) 41; H.V. KlapdorKleingrothaus et a l , J. Phys. G 24 (1998) 483. 18. A. Alessandrello et a l , Phys. Lett. B433 (1998) 156; O.Cremonesi, Nucl. Phys. B (Proc.Suppl.) 77 (1999) 369. 19. L. Oberauer,Nucl. Phys. B (Proc.Suppl.) 77 (1999) 48. 20. R. S. Raghavan, Phys. Rev. Lett. 78 (1997) 3618.
PHYSICS OPPORTUNITIES AT THE PROPOSED ORLaND NEUTRINO FACILITY FRANK T. AVIGNONE III Department of Physics and Astronomy, University of South Carolina, Columbia, SC 29208 USA E-mail: [email protected] YURI V. EFREMENKO Department of Physics and Astronomy, University of Tennessee, Knoxville, TN27996-1200 USA E-mail: efremenk@utkux. utcc. utk. edu ORLaND, the Oak Ridge Laboratory for Neutrino Detectors, is a proposed underground laboratory adjacent to the target station of the Spallation Neutron Source (SNS) under construction at the Oak Ridge National Laboratory (ORNL). The neutrons, copious nmesons, and hence neutrinos, will be generated by a 1.3 GeV proton beam (2 MW) impinging on a mercury target with 600 ns wide pulses at 60 Hz. The neutrino production rates are predicted to be: R(ye)= R[yMJ= R\pMJ =9.4x 1014 sec1. Fluxes from this source will be adequate to support a broad program of neutrino-nucleus cross-sections important in nuclear physics, nuclear astrophysics, and elementary particle physics. A broad selection of possible experiments is discussed.
1
Introduction
Neutrinos, postulated by Pauli in 1932 and discovered by Cowan and Reines in 1957, are still very much a mystery. Hundreds of experiments to study their properties have been done at low, medium, and high energies. With all this activity, their most fundamental properties are not yet known. What are their masses? Are they Dirac or Majorana particles? Do they have magnetic moments? While there now is evidence that those associated with different charged leptons mix and oscillate between flavors, the parameters of these processes are not known. In fact, their allowed ranges are very inconsistent from experiment to experiment. Their interaction cross-sections on nuclei, so important in astrophysics, are measured in only a very few cases and, in some important ones, only very crudely. Many more accurate measurements are needed. Neutrinos impact the dynamics of supernova collapse and explosion as well as the formation of heavy elements. For example, they carry off a large fraction of the gravitational energy of collapse, cooling the core, while possibly reheating stalled shock fronts that create the ejecta observed in supernova remenants. Accurate crosssection measurements of neutrino-nucleus reactions are needed to explain this.
214
215 In this article, we describe a proposed stopped-pion neutrino facility in an underground laboratory near the Spallation Neutron Source (SNS) at Oak Ridge National Laboratory (ORNL), designed to address these and many other questions. It would be built underground to accommodate a large Cherenkov detector (2000 ton) to measure, for example, differential and total cross-sections of the reaction 16 le 0(v e ,e") 1 6 F, the neutral current processes O(vx,v'xnr)l50, and 16 15 0( vx, v'xpy ) N, when filled with water containing a chloride salt. It could later be filled with a dilute mineral oil scintillator to create a scintillating Cherenkov detector to make very precise measurements of the cross-sections: 12C( » / e ,e - ) 12 N gs , l2 C(v e ,e~) 12 N*, ,2C(vx,v'xy)uC, to search for v^ -> ~ve oscillations and, if found, measure 8m2 and Sin220 accurately. This detector would also be ideal to measure Sin20w at low energies to an accuracy of about 1%, and could place much improved bounds on the muon-neutrino magnetic moment. Neutrino oscillations and Sin26w could also be measured in the water detector. The present design has three floor levels, outside of the large detector area, that could accommodate six to eight smaller detectors ranging from 20 to 200 tons. For example, a 200 ton scintillating Cherenkov detector could be constructed as a research and development tool for the large scintillating Cherenkov detector. It could be probing v^ —» ve oscillations with a sensitivity better than that of MiniBoone during the time the large detector is measuring reactions involving ls O in direct support of the Super Kamiokande and Solar Neutrino Observatory (SNO) experiments and astrophysical theory, including supernova modeling. Similar Cherenkov detectors, for example, could be built there to study the crosssections of 127I( v g ,e" )127Xe and 37C1( v e ,e~) 37 Ar used as detection mechanisms in the Homestake solar neutrino detectors, and 28Si( ve,e~ )28P of importance in astrophysics. A -30 ton heavy water Cherenkov detector could be accommodated to measure o{d( ve,e~ )pp} with statistical precision better than 1%. This result would support the interpretation of SNO data as well as theoretical studies of the weak interaction in nuclei. This reaction is particularly interesting because the nuclear structure of the deuteron is theoretically well understood. This reaction cross-section also determines that of the reaction pp -> v^e+d, which is the most important reaction in the sun. There is no other way to accurately measure this cross-section. The authors have designed, and simulated by Monte Carlo computations, a highly segmented detector based on the general principle of the Soudan-II detector, consisting of a large array of iron tubes, each surrounding a gas detector. This device could be used to make accurate measurements of (v e ,e~) cross-sections on 56 Fe, or any mechanically stable material that can be made into tubes. Finally, the proposed neutrino facility could support a program of research and development of new and innovative technologies to detect neutrinos from the sun
216 and from supernovae, and to measure neutrino-nucleus cross-sections with the required accuracy to be useful in nuclear physics and nuclear astrophysics. The proposed Oak Ridge Laboratory for Neutrino Detectors (ORLaND) could be the focus of a long-term study (20-25 years) in neutrino physics in direct support of many programs in elementary particle physics, nuclear physics, astrophysics, and cosmology.
2
The SNS as a Source of Neutrinos
The SNS at ORNL will consist of a 70 mA, H" ion source, a -0.4 km long superconducting 1.3 GeV proton accelerator, and a 220 m circumference accumulator ring that can deliver 2 MW (9.6 x 1015 protons sec"1) on a mercury target. The protons will impinge on the target in pulses of full-time duration of 600 ns (FWHM 380 ns) at a frequency of 60 Hz. This design is optimized to produce intense, short-time-duration pulsed neutron beams required for studying the structure of materials. In additional to the intense neutron beams, 1.3 GeV protons on mercury is a copious source of pions. About 17% of the incident protons will produce pions that decay into muons, electrons, and neutrinos. The n~ are slowed and captured on nuclei very rapidly in the high-Z target and shielding. Those that decay prior to capture produce a u." that also captures on a nucleus with high probability. Only about 2.3 x 10"5 of the negative muons decay before capture and produce neutrinos via: n~ —> n~ + vM (T1/2 =26 ns) and //" —> e~ + vM + ve (T1/2 = 2.2 msec). The high probability of nuclear capture of TT and uT result in a very small flux of ve. The 7t+, on the other hand, are produced with an average energy of -200 MeV and are rapidly stopped (<0.3 ns) but are not captured on nuclei. They decay via: n+ -> /J+ + Vp and jU+ -> e + + ve + v^. At 1.3 GeV, each proton produces 0.098 7t+ and 0.061 iC. This results in the production of 0.94 x 1015 neutrinos of each flavor ( v vM, and ve) per second with 9.6 x 10'5 protons sec"1 on target. The high probability of rapid nuclear capture of n~ and u." results in the ratio ve I vM < 3 x 10"4. This small value enhances the sensitivity of a v^ -> ve appearance oscillation experiment. All of the neutrinos are emitted isotropically so that the flux of each of the v^, vM, andv e neutrinos would be 3 x 106 sec"1 cm"2 at 50 m from the target. The calculated energy and time distributions of all four flavor neutrinos produced are shown in Figs. 1 and 2.
217 3
The ORLaND Facility
The ORLaND collaboration proposes to build a large underground bunker close the SNS target station. There are several arguments in favor of an underground location. First, it will completely decouple all activity at ORLaND from the operation and any future upgrade of the SNS. It will not require any modification or re-verification of the present SNS design. An underground location has the extra advantage of protection from neutrons escaping through the SNS target shielding. Also, the overburden will be enough to eliminate the hadron component of cosmic rays and to reduce the cosmic ray muon flux by a factor of four.
16 s 4
10
t
v»
" " " • - • • ^
.....!T^s<
^v^ *V
10'
ie2
0
2000
4000
6000
£000 J0000
0
2000
4860 6000 SOO0 10000
Time, nsec i0V
n
Time, nsec
j*
r
10* -.
•tm io •
ioJ
16
10
"--. \
1 J
1
^
""i\
10 10
10
W]
^i,
II ilk 6000 8000 J0000 IE
0
2000
4000
Time, nsec
: 0
2000
4000
6000 8000 J0000
Time, nsec
Figure 1. Time distributions for the four neutrino types produced at the SNS.
218 The optimal position of the bunker will be as close as possible to the first target station of the SNS in order to subtend the maximum solid angle for the maximum neutrino flux. To insure smooth integration of the ORLaND and SNS construction schedules, the only reasonable bunker position will be just outside the target building at 90° relative to the direction of the proton beam. To insure the maximum flux, the bunker position should be as close as possible to the target station. These constraints leave only one option for the location — next to the wall of the first target station but on the side, near the second target station that is being considered by the National Science Foundation.
16
16
t
i
+
f.
V.
t
| = """"Htt-H-
J
16
i
_ :::::::::::::::+:::::::::::::::
•f "ft:r"T'T"f 6.1 6.6S
16
,..r.±~r..,..T..t. 6.1S
1, , , , 6.2
, ,, , I , , , ,
6.6S
6.1
Energy, GeV
:::r:r :
16's-
z 16'
h
16'te-r le 16 16
I
V
1
i
16
16
16
16
, 6.6S
6.2
1/ •J-
«
-+
•—
16
, _
^
6.1S
Energy, GeV
TT H , , , , i . , . ,'6.2 6.1 6.1S Energy, GeV
L
L^S.
16 - , , , , [ , , , ,
6.6S
6.1
[
m
6.1S
6.2
Energy, GeV
Figure 2. Energy distributions of the four neutrino types produced at the SNS. The monoenergetic line of the muon neutrino near 100 MeV does not include broadening due to nuclear effects.
219 An artist's conception of the facility is shown in Fig. 3. The large cylindrical bunker with 30 m of water equivalent of overburden will create a protective environment for a large (2000 ton) scintillator-Cherenkov detector, with enough room on the side for a number of smaller detectors. To provide space close to the target position for the smaller detectors, the tank of the large detector is shifted to one side. The facility would have an access tunnel to move detector components in and out, a staging area, and an overhead bridge-crane. All electronics and computers, and the control room, would be located inside the bunker.
4
Scientific Program at ORLaND
There are three major goals of the proposed ORLaND facility: the search for new phenomena, the measurement of parameters and constants that are of importance in various areas of physics, and providing an environment for the R&D of neutrino detectors. The search for new phenomena includes neutrino oscillations, the measurement of Sin29w at stopped pion neutrino energies, and, among others, searching for the magnetic moment of muon-neutrinos.
Figure 3. Overview of the ORLaND facility with the SNS target building on the right hand side.
220 4.1
Search for New Phenomena
So far, the titanic efforts made by the physics community to look for deviations from the standard model of electroweak interactions have not provided strong, confirming evidence. However, there are several hints that deserve serious consideration and vigorous investigation. ORLaND will have the capability to shed light on a number of them. 4.1.1
Neutrino Oscillations
Because of the unique parameters of the SNS neutrino beam, ORLaND would be in an exceptional position. With a large, 2000 ton detector, it would be able to significantly extend the region of study of Sin229 and 5m2 parameters for five different neutrino oscillation channels — three with appearance experiments and two with disappearance: vM -> ve , ve -> vM , ve -> ve , ve —> vx , and vM —> vx. The first two transitions can be tested down to Sin220 less than 10'4 due to the strong suppression of ve at the SNS target. A sensitive search for the second transition is possible with the monoenergetic line of v^ from the fast decay of n* at rest (see Fig. 2). The last two disappearance transitions can be tested with an accuracy of 1% by measuring the following ratio of elastic scattering on electrons N(v^,e)/[N( v e ,e) + N(v^,e)]. Because all three neutrinos from this ratio are produced in equal quantities, the absolute measurement of neutrino flux is not necessary. An example of the mixing parameter region that can be studied at ORLaND for vM —» ve oscillations is shown in Fig. 4. On the same plot, the sensitivities of several other proposals are shown, together with the region favored by the LSND experiment [1]. 4.1.2
Other Searches
There are several other opportunities to look for new phenomena for which ORLaND would be able to significantly improve present limits, among them a search for the muon-neutrino magnetic moment effect in the electromagnetic part of neutrino elastic scattering on electrons in addition to the standard weak interaction. The electromagnetic part has a characteristic energy behavior in which the differential cross-section increases with decreasing energy of the recoiling target. Another interesting possibility is the precise measurement of Sin29w with neutrinos at low energy. ORLaND would be able to do this with an accuracy of about 1% by measuring the following ratio:
R= , y L
v
(1)
221
10
_^L_L_ a> 1 0
E I^X/^N^-
^
10
\ •
lisjfcj""^
10
1
^£?
^4
S65
10
10
J
10
^
^i
10
»2.-
10
0
SirT20 Figure 4. Predicted limits for ORLaND for vM —> ve oscillations compared with those of MiniBoone, MINOS, the region favored by the LSND experiment. which depends only on Sin20w. Also u-e universality can be tested by comparing the following ratios with theoretical predictions:
R,
=•
"we (V'+~M)
(2)
<*.) and
&=
&NC (Ve
+
V
M)
2
^c4 C(ve,e^Ngs)
(3)
In these ratios, the systematic uncertainty of the neutrino flux cancels out. ORLaND would be able to improve previous results from the KARMEN experiment [3] by one order of magnitude, reaching a limit governed by systematic errors. Measurements quoted by Czarmecki and Marciano [4] made at three values of Q2 differ significantly. A measurement at low energy would be very valuable. 4.2
Measuring Constants and Parameters
There are several fields in which accurate measurements of neutrino cross-sections will have major impact on future development. Two of the most obvious are astrophysics and nuclear physics.
222 4.2.1
Astrophysics
Core collapse supernovae occupy a special place in the cosmic hierarchy for many reasons. They are responsible for producing and disseminating most of the heavy elements in the periodic table, without which life would not exist. They are among the most energetic explosions in the cosmos, releasing energy in a span of tens of seconds at the staggering rate of 1045 Watts. They are the precursors to neutron stars and black holes. They play a major role in galactic chemistry and dynamics, supplying newly formed elements to the interstellar medium, and compressing and heating this medium out to distances of tens of parsecs from the point of explosion. Current supernova theory revolves around the "delayed shock mechanism." This shock revival mechanism is extremely sensitive to the details of neutrino transport — in particular, neutrino luminosities, spectra, and cross-sections in the shock heating region. During the shock reheating phase, neutrino energies are time dependent, falling in the range 10-30 MeV. The similarity of that range and the spectra of neutrinos produced during core collapse supernovae to the SNS neutrino energy spectrum is serendipitous. It is this overlap that opens up the exciting possibility of connecting ground-based neutrino-nuclear experiments with the ultimate fate of massive stars and of this universe. Measurements at ORLaND that would be of great importance to supernova theory include: 1.
measurements that could provide evidence for neutrino oscillations and, given oscillations, the mixing parameters and mass square differences; and
2.
measurements of neutrino-nucleus cross-sections of relevance to the rprocess, and neutrino nucleosynthesis. Of particular interest are those that convert 12C into U B or U C, or even to 7Be or 7Li, and 20Ne to 19F or 19Ne, or 4 He to 3He or 3H, etc. Some technology development will be necessary for some of these examples.
3.
Very specific oscillations may be important to making the r-process happen in the neutrino wind from a core collapse supernova, as pointed out by G. Fuller et al. [2].
4.2.2
Nuclear Physics
The ORLaND laboratory and detectors would be unique tools to measure crosssections of inelastic neutrino-nucleus scattering for both neutral current and charged current interactions with various nuclei. For nuclear theory, there are vast points of interest; for example, study of Giant-resonances induced by the weak-interactions, or precise comparisons of neutrino and antineutrino cross-sections on nuclei. Several techniques are under consideration to measure neutrino-nucleus interaction cross-sections: scintillating-Cherenkov detectors, which are particularly good for measuring v-carbon cross-sections, as demonstrated by the LSND collaboration [1], and as proposed above for neutrino oscillations; other liquid
223 Cherenkov detectors (water, for example, or other clear liquid compounds); segmented scintillation detectors like KARMEN [2]; and other types of segmented detectors. Here we discuss a few specific examples under investigation by the ORLaND collaboration. Specific Reactions Under Investigation 5.1
Spectra and Flux of Neutrinos from the SNS
The decay-at-rest (DAR) energy spectra shown in Fig. 2 are shown in more detail on a different scale on a linear plot in Fig. 5. The monoenergetic spectrum of vM is fixed by the two-body kinematics of the decay n+ —» // + + vM. The spectra of ve and vM have been derived in detail [6]. Normalized expressions are given in equations (4) and (5), 12
P £
( - « ) =F74 T £ U £ ° - ^
(4)
and
p
6£.
K)=-#^-T^
(5)
Intensity
where E0 is the maximum neutrino energy, 52.8 MeV.
,.r
7
L,
/
1
I,
I
,J
7
r
L
...I
'
O
O.Ol
0.02
O.03
0.0,4
0.05'
Energy,
•
O.Q6
GeV
Figure 5. Spectra of neutrinos from stopped 7t* and \i* decay.
224 The flux of ve is severely suppressed, as discussed above. To determine this important ratio, ve I v;, < 3 x 10"4, detailed Monte-Carlo simulations were performed in which the details of the structure of the mercury target shielding and moderator were accounted for. For vM —» ve appearance oscillation experiments, this is a crucially important quantity. Nevertheless, in all of these experiments done at stopped pion facilities to date, the (v e ) background expected from \JT decay is calculated. No experimental verification of these calculations has been possible because of the low statistical accuracy. At ORLaND, this issue can be confronted experimentally. One can see in Fig. 1 that there is a significant fraction of the ves emitted during the 600 ns pulse. Accordingly, by observing ve candidate events and binning them according to time, one can deduce important experimental information about the irreducible background of ve from \C decay — a background that tends to decrease the sensitivity of vM —> ve experiments and can actually produce a false positive signal. In our current feasibility studies, we start by considering all of the target materials that nature provides in pure or almost pure isotopic form (Table 1). Other criteria for material selection are availability and affordability in ton or multi-ton quantities. There are other materials not in nearly pure isotopic abundance that are also useful targets. Lead is one such example. It is a candidate for the bulk of a large supernova search detector (OMNIS) that uses neutrino spallation of neutrons from 206,207,2O8pL r c l
Table 1. Available Target Materials for Detector Construction (numbers in parentheses are natural isotopic abundances) 7
Li
i»p 35
K
56 l33
,2
Fe Cs
(92)
,J
(100)
23
(93) (92) (100)
C
(98.9)
- 8 Si
(92)
5,
V
93
Nb
:.« Bi
4
Be
(100)
"B
Na
(100)
27
(97)
45
"Ca
w
Co
13!
(100)
la
(100)
14
N
(99.6)
3I
P
(100)
(99.8)
53
(100)
115,
(100)
'"Ta
Cr
A1
(80) (100)
Sc
(100)
K9y
(I'OO)
15
*Tb
(100)
"•O
(99.8)
32
s
(95)
(84)
55
(96)
127,
(100)
"Tm
(100)
(100)
Mn
(100)
225 5.2
Neutrino-Nucleus Reactions Thus Far Considered
The following is a list of reactions, in no special order, for which we have made preliminary investigations of feasibility at ORLaND: d(v e ,e")pp, 1 6 0(v ( ,,e") 1 F, l6 12 56 0 ( m / « ) l s O , * ,2C(vc,e-)12Ngs> C(v e ,e-) 12 N*, Fe( ve,e-)56Co, and 28 2S Si( ve,e') ?. Other groups cooperating with the ORLaND collaboration are studying the feasibility of observing 127I( ve,e~ )127Xe [7], 208Pb( v, xnv'?°s"?b [5], 208 Pb( v, Vny )207Pb [8], and 208Pb( v c ,
The i/-Induced, Charge Current Disintegration of the Deuteron
The disintegration of the deuteron by reactor v e s was first observed by Reines, Sobel and Passierb and more recently by Riley et al. [9], and by stopped n+ neutrinos from the decay of u+ by Willis et al. [10] at the LAMPF facility at LANL. The LANL result was: a[d(v e ,e~)pp] = (5.2 ± 1.8) x 10"41 cm2. The calculated value is 5.4 x 10"41 cm2, and it is generally believed that this calculation is good to an accuracy of a few percent [11]. This must be demonstrated experimentally. This reaction is very important because it is the inverse of the primary reaction in the sun and is one of the main mechanisms utilized in the Solar Neutrino Observatory (SNO) detector [12,13]. Accordingly, it is necessary to measure this cross-section to a few percent accuracy if possible. While the majority of the ve from the SNS will be higher than ves from the sun, an accurate measurement at ORLaND can test the nuclear structure calculations and can calibrate SNO for the observation of the neutronization- ve pulse from a supernova. If measured to a few percent, we would know all about the deuteron that we need to know. A 20-ton fiducial volume of D 2 0 would be a cylinder 2.84 m diameter by 2.84 m high. It would contain 1.2 x 1030 deuterons. At 50 m from the SNS target, the flux will be 3.0 x 106 ve/cm2/sec. Accordingly, the rate in a D 2 0 Cherenkov detector would be 5900 ± 77 per year for a 1.3% statistical accuracy in one year, or 11,812 ± 110 in two years for a statistical accuracy of <1%. This can be expanded to a 30-ton D 2 0 target in an acrylic sphere immersed inside the 2000-ton H 2 0 Cherenkov detector to observe reactions on D 2 0 and H 2 0 simultaneously. The ve flux can be accurately monitored by electron-neutrino elastic scattering. 5.2.2
Measurement of the Cross-Section of 16 0( ve,e~ )16F
The charge-current interaction ve + 16 0 -» 16F + e" is interesting for several reasons. As pointed out by Haxton [14], for energetic neutrinos from a supernova, the reaction le O( v e ,e") 16 F exceeds (ve,e~) elastic scattering in rate by more than a factor of 10. In a large water Cherenkov detector, Super Kamiokande for example, this could be a more sensitive mechanism for identifying the burst of ve from the
226 neutronization pulse of a supernova collapse. For this application, an accurate crosssection measurement would be necessary and could be done only at ORLaND. In addition, the cross-sections for charge-current and neutral-current scattering off oxygen can be calculated accurately using data obtained from P-decay rates, (p,n) reactions used to calibrate Gamow-Teller strengths, and electron scattering form factors [14]. An approximate computation of the reaction rate per oxygen nucleus has been performed (Fig. 1 in ref. [14]) and the spectrum given (Fig. 5 in ref. [14]); the result is shown in equation (6): (CTO) = fofay^VeyiE^
I j®(EvyEVe
= 2.65 x 10-35/sec/16O
(6)
The feasibility of that experiment, and very probably of differential crosssection measurements, depends on the 2000-ton ORLaND Cherenkov detector filled with high purity water. The fiducial volume and mass are 1660 m3 and 1472 tons of 1 6 0, respectively (9.22 x 1027 moles or 4.93 x 1031 atoms of 16 0). The reaction rate at 50 m from the target is: R = N( 1 6 0)(CT(D) = 1.3 x 10"3 sec" 1 ,
(7)
corresponding to 4.7 hr"1, 112 d"1, and 40,880 y"1. These rates will be reduced by an efficiency of 0.30. Over a period of several years, these rates imply the feasibility of measuring da/dQdE with at least some angular and energy resolution. Haxton [14] has pointed out the importance of this angular distribution. For \TV) = 8 MeV, the maximum cross-section at 165° is 13 times larger than at 15°. For \TV) = 8 MeV, the maximum cross-section is 28 times larger than that for \TV) = 4 MeV. The enhanced back-angle emission is due to forbidden transitions which are stimulated by higher energy ve. Therefore, the accurate measurement of these cross-sections will be very valuable for the analysis of future supernovae data collected by Super Kamiokande. A large rate, compared to that of the neutronization pulse of ve or compared to ve + p —> e+ + n, with a large backangle fractions, would imply that these latter ve reactions come from neutrinos that decoupled from the core at a higher temperature. This could imply vv -> ve oscillations, on which there are no constraints. This effect could be large enough to observe if the collapsing star is close enough to allow significant statistical accuracy, because most models imply (E v ) ~ 11 MeV, whereas lEVr) Both vr,vr
~ 25 MeV.
and v^,v^ interact only via neutral currents and therefore decouple
from the core more readily and earlier when the core temperature is higher. ORLaND is the only proposed facility at which these measurements could be done.
227 5.2.3
Neutral-Current Reactions in
16
0
Recently, Langanke, Vogel and Kolbe [15] and Kolbe, Langanke, and Thielemann [16] have calculated rates for the reactions 16 0( vx, v'xpy )15N and 16 0( vx, v'jn ) 15 0, among others, for a Fermi-Dirac neutrino spectrum with parameters (J. = 0, T = 8 MeV and u = 3T, T = 6.26 MeV where the FD distribution is: m
=
l
+ e*j%E-N)/T
•
(8)
The resulting de-excitation gamma rays in 15N and 15 0 are above 6 MeV; therefore, using a water Cherenkov detector with a threshold of a few MeV, these can be observed in coincidence with the proton pulse from the SNS accelerator. If we include ve, v^, v^, the cross-section is 3.7 x 10"42 cm2 for 16 0( vx, v'xpy), and the proton is not observable. For the 2000-ton water detector at 50 m, this results in 5.5 x 10"4 sec"1 or -48 per day and 17,345 per year. Kolbe et al. [16] suggest this as a search for strangeness in nuclei because the cross-section above would increase by ~21% if the strangeness axial form factor, Gs, changed from 0 to -0.3. The cross-section for 16 0( vx, v'xny)xiO, including vM,vM,ve, is 9.3 x 10"43 cm2, 4 times smaller, which would result in 4,336 events per year. The ratio of the rates of (py) to (ny) final states is ~4 with Gs = 0 and ~ 6 for Gs = -0.3. By adding a NaCl solution to the water, it might be possible to distinguish the (ny) reaction. In any case, 16 0 is also interesting because its charge current cross-section is one that might contribute to the reheating of a supernova shock front, would be useful in analyzing water Cherenkov solar neutrino and supernova data, and can test nuclear structure calculations like those of Haxton [14], Langanke et al. [15], and Kolbe et al. [16]. Finally, the possibility of analyzing Super Kamiokande data from a future supernova to determine the ratio of back-angle e"-emission to front-angle emission would depend on accurate measurements of this reaction with a known spectrum and flux. 5.2.4
Measurement of the Reaction 56Fe( ve,e~ )56Co
There are two main motivations for measuring this cross-section as accurately as possible. 56Fe comprises the majority of the inner core of a pre-collapse large star. Accordingly, much of the infalling material in a supernova collapse is 56Fe. As the neutrinos emitted from the inner core meet the infalling material and a stalled shock front, this reaction could reheat the material in the shock front and restart it. The effectiveness of this mechanism depends very much on the cross-section of 56 Fe( v e~ )56Co, which is a direct input to supernova collapse model simulations.
228 The second motivation stems from the fact that these simulations have many theoretical cross-sections for various v-nuclear processes. Many can never be measured. Those that can be measured accurately can be used to test the nuclear structure models. An accurate measurement of this cross-section would be very valuable in evaluating the importance of 56Fe( v e ,e - ) 5 6 Co in stellar evolution and collapse [17]. A third motivation stems from the fact that iron is one of the materials being considered as a converter in large supernova detectors in the planning stage — OMNIS for example [18]. This cross-section would be valuable in testing nuclear models as well as assisting in the design of experiments. The first estimate of the cross-section was made by Bugeev {cited in ref. [19]), using a shell model. Recently, it has been calculated and averaged over the stoppedpion neutrino spectrum by Kolbe, Langanke and Martinez-Pinedo [19], and by Kolbe and Langanke [20]. In the latter [20], the neutral current cross-sections for the production of neutrons and their spectra are given for Fe and Pb. The basic theoretical expressions for charge current v-nucleus reactions were derived by Walecka [21] and by Donelly and Peccei [22]; they involve both Gamow-Teller (GT) and forbidden contributions. The GT contributions are reliably calculated within the shell model, whereas RPA models are more reliable for the forbidden transitions [19]. In the calculations used in those studies [19,20], the GT contributions were computed with the large-space Strasbourg-Madrid codes developed by Caurier [23,24] while the forbidden transitions were computed within the RPA. The resulting cross-section, weighted by the stopped-pion neutrino spectrum is 2.64 x 10"40 cm2. The only measurement was reported by the KARMEN collaboration [25]; their result was [2.56 ± 1.08 (stat) ± 0.43 (syst)] x 10"40 cm2. According to the feasibility study reported here, the cross-section can be measured at ORLaND with a statistical accuracy of about 1-2% using a high granularity detector called SOUDANINO that is described below (see also Y. Efremenko, these proceedings). The geometry and principle of operation of SOUDANINO is similar to the SOUDAN-II detector from which it derives its name [26]. It is a cubic matrix of parallel iron tubes with thin walls, each containing a position-sensitive gas detector. Monte-Carlo simulations indicate that the optimum thickness for 30 MeV electrons is 0.5 mm, with diameters OD = 10 mm and ID = 9 mm. The total tube length is 3.5 m, only 2.5 m being in the fiducial volume. At 50 m from the SNS target, o ( v e ) = 3 x 106 ve cm"2 sec"1, while the number of 56Fe atoms in iron is 9.9 x 1027 ton"1. The rate is approximately (er)N( 56 Fe)0(i/,) = 250 y'1 ton"1. This type of detector has an estimated efficiency of 0.30; therefore, to collect 1000 events per year for analysis, the fiducial mass must be 13.3 tons, the length of each tube in the fiducial volume is 2.5 m, with an iron volume of 37.3 cm3,
229 corresponding to 0.29 kg per tube. The fiducial volume must contain 45,862 tubes, or an array of 214 by 214 tubes. To render this array fiducial, 50 tubes must be added to each row and to each column. The total detector is an array of 314 by 314 tubes, or 98,596 tubes, so that the fiducial volume is surrounded by 0.5 m of live detector on all sides. The total active detector is a 3.5 x 3.5 x 3.5 m array of parallel tubes. This concept can be utilized to measure charge current ( ve,e~ ) reaction crosssections in any of the available target materials given in Table 1 that can be formed into rigid tubes. For some of the heavier elements — for example 181Ta, 209Bi, and 206 207 2O8 ' ' Pb(natural), the cross-sections are theoretically large, so that the detector could be much smaller in mass. The tubes might well be thinner walled, but in general the detectors could be significantly smaller. 5.2.5
Neutrino Interactions on 12C
The 12C isotope forms one of the "onion skin" shells of a large star just prior to collapse, because stellar evolution creates 12C in the 3a burning cycle. Naturally, it will constitute much of the ejecta that later will interact with the neutrinos to possibly reheat the shock front. Also, l2C forms an isospin triplet with 12B, 12C, and 12 N(J",T = 1+,1) in its 15.11 MeV excited state and can be excited from its (J",T = + 0 ,0) ground state by the weak neutral current. This reaction allows investigation of the details of the isovector and axial vector part of the weak hadronic current. The reaction is well understood theoretically [27-30]; however, this is a reaction that can be used to tune small nuclear details. 12 C( v e ,e-) 12 N*: This excites 12N to 5 excited states that decay by proton emission. Several states in U C decay by gamma emission, but the single-pronged proton and gamma events will suffer more background; however, it may be feasible. 12 C( vx,v'x)nC (15.1 MeV): This reaction can be done at short time intervals following the proton pulse, thereby being dominated by (vM,v'M) reactions, but including some ve and vM. Then cuts favoring longer times following the proton pulse will involve only ( vM, v'^) and ( ve, Ve) reactions, which cannot be separated from one another. A summary of previous cross-section measurements and theoretical calculations are given in Tables 2 and 3. In addition, the neutral current to charge current cross-section ratio, averaged over the stopped-pion neutrino spectrum, was measured only by the KARMEN group. Their result was 1.17 ± 0.11 ±0.012 [3]. The three calculated values are 1.08 [27], 1.13 [28], and 1.27 [29].
230 Table 2. Charge Current Cross-sections on 12C (in units of 10~42 cm2)
"C(v e ,e-)"N P
12
Model
[ref]
9.2 9.4
—
EPM
3.7 6.3
DWM CRPA
[27] [29]
5.4 ±1.9
E225 LSND KARMEN
C( ve,e~-
9.3 9.3 ±0.4 ±0.8 9.1 ±0.4 ±0.9 9.3 ± 0.4 ± 0.8
)i2N*
5.7 ±0.6 ±0.6 5.1 ±0.6 ±0.5
[28] [30] [1] [3]
EPM = elementary particle model; DWM = Donnelly-Walecka model; CRPA = continuum random phase approximation. Table 3. Neutral Current Cross-sections on l2C( Ve) 12C* (in units of 10'42 cm2)
<<x(ve,v4
(*fo.v;)
(°( v « + ^)
4.6 4.5 4.9 4.7 4.6
5.7 5.4
—
—
10.3 9.9 10.6 11.9 10.5 10.9 ±0.7 ±0.8
5.8 7.2 5.9
Mv^| — 2.74 2.70 2.64 2.80 3.1 ±0.8 ±0.5
Model
[ref]
EPM EPM EPM DWM CRPA KARMEN
[31] [27] [32] [29] [28] [3]
The larger volume of a detector experiment at ORLaND should allow these cross-sections and ratios to be measured extremely accurately in either a 200-ton or a 2000-ton detector. Moreover, 12C reactions should be measured at all times as a monitor of the neutrino flux for the other experiments. 6
Detectors Being Developed at Other Laboratories
There are several other detector concepts being developed at other sites that are reported at this conference. These groups propose to develop, test, and calibrate their prototypes at ORLaND. The first is the Nal in aqueous solution as a Cherenkov detector, with an extra technique to identify the 127Xe from the reaction 127I( ve,e' ),27Xe. The details are covered in the paper by Lande et al. [7] in this proceedings. The main motivation is to calibrate the 127I solar neutrino experiment of the University of Pennsylvania in the Homestake gold mine. An accurate measurement will also be valuable to test nuclear models in this mass range. Also, ve + 37C1 can be investigated with a perchloroethylene Cherenkov detector at ORLaND.
231 The second is the array of lead converters and neutron detectors being designed for the OMNIS detector discussed by Boyd [5] in this conference. Detecting only neutrons, the detector would be sensitive to two reactions: 208
Pb( vx, Vxxn )208xPb and 208Pb( ve,e~ )108Bi* -> 207Bi + n + ys.
When ORLaND is operating, the OMNIS collaboration intends to place a prototype in the laboratory that is large enough to measure the rate accurately enough to determine the sensitivity of a large array to a supernova collapse at a given distance. The third is the lead perchlorate, Pb(C104)2, that is being developed at the University of Washington and is described in these proceedings by Doe [8]. This clear liquid contains 80% lead by weight. It can be used as a supernova detector or a neutrino oscillation detector using the liquid as a Cherenkov medium. The research and development continues and, if successful, this could represent an efficient new technique in neutrino detection.
7
Summary
A large number of examples of interesting physics that can be done at the ORLaND laboratory at the SNS at Oak Ridge have been discussed. These range from neutrino oscillations with very low backgrounds, a precision measurement of Sin29w, new bounds on the magnetic moment of the muon neutrino, and a sensitive probe of the KARMEN anomaly. Neutrino nucleus interactions on deuterium and 16 0 have been examined from the point of view supporting the interpretation of the data from SNO and Super Kamiokande from solar neutrinos and supernova neutrinos. A number of neutrino nucleus experiments have been discussed in detail to support theoretical astrophysics and, in particular, the theoretical efforts in supernova simulations by the group at ORNL [33]. Finally, several examples of the proposed experiments from three groups, not in the original ORLaND collaboration, demonstrate the growing interest in the nuclear and astrophysics community in the success of the ORLaND proposal to build a multi-experiment stopped-pion neutrino facility.
8
Acknowledgments
One of us (FTA) was supported by the U.S. Department of Energy's Office of High Energy Physics and the University of South Carolina. Much of the work done by YuE was supported by the Oak Ridge National Laboratory. The authors thank H. K. Carter, F. Plasil, and G. Young for their significant efforts in the development of the project, and the entire ORLaND collaboration for their many contributions over
232 several years. We thank Richard Boyd for his helpful comments and his critical reading of the manuscript. The ORLaND Collaboration F. T. Avignone llla, B. D. Anderson 6 , T. C. Awes c , S. Berridge^, W. Bilpuch e , C. Britton c , W. Bryan c , W. M. Bugg^, R. L. Burmai/, J. Busenitz£, H. K. Carter^, L. Chatterjee'', V. Cianciolo c , H. O. Cohn^, M. Danilov', D. Dean c , L. DeBraeckeleei*, P. Degtiarendo/, Yu. Efremenko^, M. A. Elaasar^, A. R. Fazely', T. A. Gabriel c , C. R. Gould™, V. Gudkov a , R. Gunasingha', Z. D. Greenwood", T. Handled, E. L. Hart c , R. L. Imlay 0 , Yu. A. Kamyshkov c , D. D. KoetkeP, K. Kubodera*2, C. Lane?, R. W. MainweilerP, W. J. Metcalf0, A. Mezzacappa c , S. Mishra a, L. W. Mo r , V. Nosik', T. A. Nunamaker r , S. Nussinov5, A. Piepke8, F. PlasiF, J. Reidy', C. Rosenfeld0, D. Smith", I. Stancuqv, T. D. S. Stanislaus?, R. Steinberg?, M. Strayed, R. Svoboda 0 , J. W. Watson 6 , A. Wittenberg c , J. Wolf8, O. Zeldovich', W.-M. Zhang 6 . a) b) c) d) e) f) g) h) i) j) k) 1) m) n) o) p) q) r) s) t) u) v)
University of South Carolina, Columbia SC Kent State University, Kent OH Oak Ridge National Laboratory, Oak Ridge TN University of Tennessee, Knoxville TN Duke University and TUNL, Durham NC Los Alamos National Laboratory, Los Alamos NM University of Alabama, Tuscaloosa AL Oak Ridge Associated Universities, Oak Ridge TN Institute for Theoretical and Experimental Physics, Moscow, Russia Jefferson Laboratory, Newport News VA Southern University, New Orleans LA Southern University, Baton Rouge LA North Carolina State University, Raleigh NC Louisiana Technical University, Ruston LA Louisiana State University, Baton Rouge LA Valparaiso Universisty, Valparaiso IN Drexel University, Philadelphia PA Virginia Polytechnic and State University, Blacksburg VA Tel Aviv University, Tel Aviv, Israel University of Mississippi, Oxford MS Embry Riddle Aeronautical University, Prescott AZ University of California, Riverside CA
233 References 1. 2. 3. 4.
5. 6. 7. 8. 9. 10. 11.
12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
C. Athanassopoulos et al., Phys. Rev. Lett. 81, 1774 (1998). G. Fuller, private communication. R. Maschew et al., Prog. Part. Nucl. Phys. 40, 183 (1998). A. Czarnecki and W. Marciano, Talk given at the "International Workshop on Electron-Electron Interactions at TeV Energies," December 1999, Santa Cruz, CA; hep-ph/0003049, March 2000. R. Boyd, these proceedings G. Kallen, Nuclear Theory. Springer-Verlag, New York (1962). K. Lande, these proceedings. P. Doe, these proceedings. F. Reines, H. Sobel, and E. Paisierb, Phys. Rev. Lett. 45, 1307 (1980). See also Riley et al., UCI Preprint #98-37; hep-ex/9904001, 1 April 1999. S. E. Willis et al., Phys. Rev. Lett. 44, 522 (1980). K. Kubodera and F. Myhrer, in "Proceedings of the Accelerator Production of Tritium Symposium," 14-15 May 1996, Columbia SC, eds. F. T. Avignone and T. A. Gabriel. World Scientific, 148 (1997). J. Bahcall, these proceedings. J. Simpson, these proceedings. W. C. Haxton, Phys. Rev. D36, 2283 (1987). K. Langanke, P. Vogel, and E. Kolbe, Phys. Rev. Lett. 76, 2629 (1996). E. Kolbe, K. Langanke, and F.-K. Thielemann, Eur. Phys. J. A3, 389 (1998). S. E. Woosley, D. Hartmann, R. D. Hoffman, and W. C. Haxton, Ap. J. 356, 272 (1990). R. Boyd, these proceedings; P. Smith and D. Cline, private communication. E. Kolbe, K. Langanke, and G. Martinez-Pinedo, Preprint: arXiv:nuclth/9905001 V2, 3 August 1999. E. Kolbe and K. Langanke, Preprint: arXiv:nucl-th/0003060, 27 March 2000. J. D. Walecka, in "Muon Physics," eds. V. W. Hughes and C. S. Wu. Academic Press, 113(1975). T. W. Donnelly and R. D. Peccei, Phys. Repts. 50, 1 (1979). E. Caurier, Computer Code ANTOINE. CRN, Strasbourg (1989) [quoted in 18]. E. Caurier et al., Phys. Rev. C59, 2033 (1999). K. Eitel [quoted in 18]; G. Ruf, Diploma Thesis, University of Bonn (1998). W. W. M. Allison, Nucl. Instr. Meth. A376, 36 (1996). M. Fukugita et al., Phys. Lett. B212, 139 (1988). E. Kolbe et al., Phys. Rev. C49, 1122 (1994). T. W. Donnelly, Program NUEE to calculate cross-sections for ISIS: private comm. to KARMEN collaboration (1991). D. A. Krakauer et al., Phys. Rev. D45, 2250 (1992).
31. J. Bernabeu and P. Pascual, Nucl. Phys. A324, 365 (1979). 32. M. Pourkaviani and S. L. Minz, J. Phys. G: Nucl, Part. Phys. 16, 569 (1990). 33. A. Mezzacappa, these proceedings.
SOME ASPECTS OF NEUTRINO PHYSICS SHMUELNUSSINOV Department ofHigh Energy Physics, Tel Aviv University, School of Physics and Astronomy, Tel Aviv, 69978 ISRAEL
1
Introduction
I first came across the name "Avignone" (not in the context of bridges or dames !) about twenty years ago via the "Avignone spectrum" of reactor neutrinos — a key to the analysis of reactor neutrino oscillations then and now. While the standard electro-weak model for gauge interactions is well established, questions pertaining to neutrino mixings and masses, the existence of axions/Wimps or decays involving quantum number violations (and neutrinoless double-beta decays, in particular) still remains largely open. The physics of these phenomena has been an ongoing saga in which wrong theories/experiments keep inspiring further experiments/theories as the true pattern slowly unfolds — a saga in which Frank Avignone has been a dominant figure. Over the decades, Avignone and Brodzinsky and their collaborators pioneered the developement of underground detectors of exceedingly low background. These detectors opened up whole new fields of experimental research and, one may say with some poetic license, that Avignone et al. lived up to Archimedes: "Give me a sufficiently powerful instrument and I will move the world." Such instruments detected double-beta decays with lifetimes of =1021 years [1] and established subeV bounds on the majorana mass of the electron neutrino [2]. Also, interactions of relic heavy (LSP?) halo particles of =50 GeV mass and subweak cross-sections are now, in principle, detectable. Of all the many varied applications that Frank himself used these instruments for, I will mention only one [3,4]. The need to have ultra-pure germanium detectors (so as to avoid radioactive background) forced making the detectors out of single crystals. At Frank's instigation, this led to a more sensitive search for solar axions. The point is that solar axions of specific energy are more effectively (coherently) converted to photons via Primakoff scattering on the germanium atoms when the planes of the lattice are oriented relative to the sun's direction so as to satisfy the Bragg condition. Even with the limited O (KeV) energy resolution, this implies a time-energy correlation for the purely calorimatric signal in the detector, drastically enhancing a solar axion signal relative to the unstructured background. In this talk, I will focus on two issues. First I will discuss the future of neutrino oscillation physics in light of the recent SuperKamiokande discovery of atmospheric 235
236
neutrino oscillations. In this connection I will briefly mention ORLaND, which has become Frank's new passion and mission. Next I will discuss some exotics connected with neutrino physics and how a careful combination of astrophysics and particle physics considerations can limit them. Indeed, it is only by taking the existing few positive indications for where the new physics is, and combining them with the vast body of knowledge and arguments for where it is not, that we can start delineating the outlines of the future reality that the present ongoing saga will eventually reach. 2
The Physics of Neutrino Physics — the High and Low AM2 Options
Neutrino oscillations were first suggested by the relative paucity of neutrinos observed in various experiments designed to detect solar neutrinos as compared with theoretical expectations [5]. Neutrino oscillations have definitely been found in the atmospheric neutrino data of the SuperK [6] and may soon be confirmed by the first long-baseline exeriment with a controlled neutrino beam between KEK and SuperK. The last discovery entailed two surprises: • the mixing between v^ and vT is almost maximal, and •
the observed AAfj*r =3.10"3 (eV)2 is small.
Higher AM2 values and smaller mixing were expected from the orthodox hierarchical "see-saw" inspired extrapolations utilizing the then favored small-angle MSW solution to the solar neutrino problem. There are two known approaches to explaining the small neutrino mass. In the conventional see-saw approach [7], the small neutrino mass simply reflects the relatively high-mass scale of the right-handed SU(2)L singlet partners of the ordinary left-handed neutrinos. In the absence of a primary left left majorana neutrino mass, the diagonalization of the 2x2 mass matrix generates M(v L ) = Mj=>,Vac / M(vR). The other alternative is to introduce a new triplet Higgs [8] with a light VeV. This led to a very rich and intriguing phenomenology, but it is now practically ruled out since the light triplet majorons would have almost doubled the invisible part of the Z width. Thus, unless some truly novel and convincing mechanism can generate light neutrino masses, most theoreticians are reluctant to give up at least some form of the see-saw mechanism. Maximal mixing has been encountered before in particle physics, in the celebrated KK system. However, in this case a rather profound principle underlies the maximal mixing — the fact that, by the CPT theorem, the masses of the particle and anti-particle are identical and hence the smallest off diagonal (AS = 2) matrix element will cause maximal mixing. There is no similar flavor degeneracy principle here; in fact, a striking NONdegeneracy is suggested by the analogous lepton and/or quark masses. Furthermore, we do not have a completely symmetrical maximal mixing 3x3 matrix analogue as
237
SuperK data [8], and reactor data limit vT -» ve mixing. This suggests a bimaximal mixing pattern [10] corresponding to the matrix: f
0 V2 -A/2
V2 1 1
-V2 A 1 1
It is amusing that all non-vanishing entries in this matrix are ±
V2
and ±
1
i.e., the entries of the completely symmetrical real 2x2 and 4x4 mixing matrices! It is not clear whether a natural explanation of this pattern [11] exists, requiring a drastic paradigm change as compared with the standard hierarchical see-saw and the Fritch-Minkowsky-Stech relations between mass ratios and mixings, or whether the latter can still be somehow resuscitated [12]. Certain modified versions of the see-saw mechanism [13] can yield "small" AM2 values, along with substantial generation independent M2. This then avoids hierarchical neutrino masses — i.e., M(vT) » A/lv^l » M{ye) , allowing for Af(v,) = few eV with some hot dark matter component due to neutrinos. Also, such electron-neutrino majorana masses are detectable in the germanium underground experiments alluded to above. The modified see-saw variants often maintain a weaker form of hierarchical neutrino mass differences: AM^, > AMJ^ . (In passing we note that all the solutions, small and large mixing MSW and vacuum, for the solar neutrino problem dictate AM2, much smaller than AM^.) The small AM2 values severely limit the prospects of finding any neutrino oscillations in terrestrial accelerator experiments, which is not a (very!) long baseline type. We generally have 1 > (AM2)L/ E, amd the oscillation probability of Vfi -> vr or vM -> ve is: AM2 3.10-3(eV)z
L /750 km^ sin 2 20 UT E/lOGeV
(1)
(Since a > sin a, the above constitutes an upper bound for P.) In Eq (1) we used the proposed path lengths in the long baseline Minos and Opera experiments to normalize L; 10 GeV, the typical neutrino energies in these experiments, normalizes E. It is above the threshold for tau production, as required, for tau appearance experiments.
238
An analogous expression for neutrinos from decays of stopped pions is: Pfa
-> v r ) - 3.10-
AM2 L/50m 3.10-3 (eV)2 £730MeV
-\2
sin2 20. en
(2)
where we used the flight path length of LSND/KARMEN and ORLaND set-ups, and 30 MeV is the fixed (accelerator independent) typical decay at rest energy of the Michel spectrum neutrinos. The total number of interactions of neutrinos in a detector of mass M(D), at a M{D)ov{E) distance L, scales with Hence the number of interactions of oscillated
n
neutrinos — i.e., the number of tau appearances, or the reduction of the number of charge current v^ interactions, is independent of L! If we further use ov(E) = G\EMN for the high energy neutrinos, we find that the number of interactions of oscillated neutrinos in HE beams is: 105
W megawatt
M(D) IS
sin2 2# year
AM2 3.10-1 {eVy
fv
(3)
where fv ~ 10 -4 - 10~3 is the fraction of the total energy input in neutrinos that emerge from pions decaying in a given decay length: L(Decay) « L in the appropriate small solid angle AQ. so as to arrive at the far detector. Likewise for decays at rest with - and isotropic angular distribution, we find that the number of interactions of oscillated neutrinos for DAR is: 10
W megawatt
( > sin2 20
M(D) year
AM2 3.10-3 (eV)2
(4)
with nv ~ 0(1) or O(few) the average number of neutrinos produced per GeV of incoming energy in the beam of accelerator particles or at the reactor. The above arguments are rather sketchy and many simplifications were made in an effort to obtain nice, universal, expressions involving only the power of the reactor/accelerator, the detector's mass M(D), and overall time T. Still, it is quite clear that terrestrial neutrino oscillations would be largely doomed if indeed 3.10_3(eV) = AM2 > AM2^. Should we, then, abandon altogether experiments looking for v^ -» ve oscillations with "high" AM2 and small mixings? The tantalizing LSND result mentioned in these proceedings by Rosen suggests that we should not yet do this. The proposed ORLaND experiment at the neutron spallation source at Oak Ridge will, among many other things, enable us to perform sensitive searches for oscillations precisely in this area. This is the case because of the intense, pulsed
239 beam and an unusually low ve I v^ ratio. The latter is suppressed by the strong absorbtion of the negative pions and muons from which the anti-electron neutrinos originate in the heavy target and shielding in the spallation source. Also, a set-up with a segmented detector with good angular resolution and time separation of the electron and muon neutrino pulses, will permit measuring the scattering of the different species off electrons, considerably improving the value of sin2 9W measured this way. These and the many neutrino nuclear cross-sections that can be addressed at ORLaND have been discussed at length in the detailed talk by Avignone in these proceeding. In vigorously pursuing the ORLaND project, Frank is clearly suggesting that experimental physics is not only the "art of the impossible" — that is, verifying most current theoretical predictions, rather it is also the "art of the possible" — seizing this unique opportunity to install a new research program in experimental neutrino physics in conjunction with the world's largest neutron facility. 3
Supernova 1987a and Neutrino Mixing
The neutrino pulses from SN 1987a observed at Kamiokande and 1MB lasted for few seconds and only =18 interactions, presumably ve + P —> e+ + n, were recorded. Still, these observations significantly restrict neutrino properties [15]. Interestingly, there may be some implications for neutrino mixing as well. (-) (-) (-) Simulations [16] suggest that each of the six neutrino species ve v ^ v T carries 1/6 of the total gravitational collapse energy. However, due to the smaller cross-sections, v^ and vT are emitted from deeper, hotter regions of the core than ve , and hence have higher energies: ( £ , ) = (£ T ) = (l.3-1.6)(£ e )
(5)
If one quarter of the emitted v^ and vT oscillate enroute from the LMC into ve , as the bi-maximal mixing suggests, then (recalling that the cross-section for neutrino reactions at these energies scale like the square of the energy) about half of the interactions observed originated from the higher energy converted neutrinos. This would tend to make the energy distribution almost bimodal and considerably broader. Due to the paucity of SN1987a neurino events, it is not clear if this is indeed the case [17,18]. Also, a more recent analysis [19], which more carefully accounts for the neutrino energy losses due to recoils in elastic nuclear scatterings, suggests much smaller disparities between the energy distributions of the electron neutrino and the other species.
240
A future galactic supernova will manifest in SuperK via thousands of events, and the above test for electron neutrino mixing could be carried out more reliably. During the short (0 (millisecond)) time of the gravitational collapse, most of the lepton numbers of the original stellar core of mass M(core) = two solar masses and Ne = O(1057) electrons is converted, via electron capture on protons, into electron neutrinos. Simulations indicate that roughly half of these neutrinos escape as a millisecond pulse of "neutronization-neutrinos." The smaller number of these neutrinos as compared to the "thermal" electron anti-neutrinos.and the lower crosssections for the (neutrino-electron scattering) detection interaction of the former explain why the neutronization pulse has not been seen in the SN1987a data. Such a signal should, however, clearly manifest in future galactic supernovae, although its expected magnitude would be reduced by about a factor of two for maximally mixed electron neutrinos. In certain exotic cases, such as an electron neutrino with nonvanishing tiny electric charge or for massive tau neutrinos which decay to an electron, a positron, and a neutrino, the asymmetric neutronization pulse (comprising neutrinos but not anti-neutrinos) could have dramatic and amusing manifestations that we briefly discuss next. 4
Supernovae and Nonvanishing Neutrino Charge
It has been noted [20] that, unless |#(v e )/gj = x{ve) - 10~17, the curving of the neutrino trajectories in the galactic (and LMC) magnetic fields, which is stronger for lower energy neutrinos, would broaden the arrival time of neutrinos from SN 1987a beyond the observed few seconds duration. The following considerations may suggest an even more stringent upper bound on the neutrino charge. The neutronization pulse of neutrino electrons will charge up the core to a total charge: Q = ^Ne\qe\x(ve). The resulting electrostatic energy, Q2/Rcore, would then exceed the total gravitational collapse energy ^ R, - ^ = ( M22 R,- 3 ) > 0 PL
B
^ .
once * ( v ^ 2 ^ ^ - = 4 . 1 0 ^ N. MPL
<6>
with MN,MPL = 1019 GeV, the nucleon and Planck masses N„ and the number of nucleons in the collapsing core of mass and radius MC,RC. Clearly the RHS of Eq (6) would then furnish a new upper bound on x{ve)- A fraction, Rc/Rprogenitor ~ 5.10"7, of the electrostatic energy resides outside the progenitor star. The large E fields at
241
the surface of the star would then generate discharges in the ambient stellar atmosphere, transforming some of the above electrostatic energy into an optical flash. Such a flash, which should have been coincident in time with the neutrino pulse, has not been seen. This may suggest an even more stringent bound on the neutrinos charge x(v ) ^ 10"21 - 10"20. We note, however, that the large fields inside and oustside the core/progenitor may generate counter-flowing currents, preventing excessive charging. 5
Lepton Asymmetry in Supernovae and Its Implications for the Case of Decaying Tau Neutrinos
Ignoring for the moment the suggestions from SuperK (and possible difficulties with Big Bang nucleosynthesis), let us assume that the tau neutrino mass is in the 515 MeV range. If the mixing (Ver) > 0.01, a value consistent with known bounds, the decay vT —» e+e~ve proceeds with a lifetime of less than 100 seconds. These decays will then occur mainly outside the progenitor star of SN 1987a and might lead to the dramatic effect discussed next. In the present scenario, we have, in roughly 1% of the e~p —> vn collisions, tau neutrinos rather than electron neutrinos. We therefore expect the lepton number excess, due to the trapped electrons and electron neutrinos, to equilibrate among all the lepton species so that N(yx) - N(yT) = —- = 1056. In analogy with the Michel spectrum for muon decay, the electron in the decay of the tau neutrino should have, on average, almost twice the energy of the positron, and the reverse — namely, (Ee+) = l(Ee-), should hold for the decay of v T . Therefore, on average, the electrons from tau neutrino decay move faster than the positrons by A)3 = Mj HE} ~ 10"3, and a gradual separation between the location of the electrons and positrons would ensue. Due to the tau neutrino excess, the resulting charge separation will not be cancelled by the reverse effect in the decays on the anti-neutrinos. The total asymmetry energy,
( V )Vt ~ (Ee+ )Vi ) K " "vr ) " ^ K ) ^ - 1051 CT& > could eventually be stored as electrostatic energy in the giant "stellar capacitor" thus formed. Just as in the case discussed in the previous section, this could then be manifest via an early optical flash. For a more comprehensive discussion of this and the more standard effects of tau decay outside the supernova, see [21,22].
242 6
Neutrino Anomalies
Experimental neutrino physics has had.over the years, more anomalies than almost any other branch of experimental physics. Most anomalies fade away; a few (very few!), like the "atmospheric neutrino anomaly," become real effects. Here we would like to focus on the relatively recent vintage, the KARMEN time anomaly described here in [23], which suggests the existence of a new particle (denoted here by n°) with a precisely tuned mass Mn- = M ^ - M ^ - 5 KeV = 33.906 MeV. The n" particle is created in the target in a small fraction of the stopped pion decays, travels with a velocity of pV = 1/60 the distance to the detector in =4 microseconds, then decays therein via n° —> e+e~vd, depositing electromagnetic energy of more than 17 MeV (the lack of a sharp line at 17 MeV disfavors the n° -» yvd decay). To fit the magnitude of the anomaly, we require that Br(7T+ -» n+n"). T(n° -> e+e~y) = 2.6 x 10"17 (sec)"1.
(8)
Due to putative invisible n° decays, its actual lifetime is shorter than r(n° -» e+e-y) = T = r _ 1 (n° -» e+e~y) .
(9)
The latter partial lifetime should therefore exceed a microsecond in order to prevent the n° particles from decaying before arriving at the detector. The recent stringent PSI accelerator bound on the decay of n+ into a monochromatic muon: Br(n+ -> n+n°) < 2.6 x 10"8, implies via the basic relation, Eq (8), that t is shorter than 1000 seconds. This still allows T and the corresponding Br(^ + —»fi+n°) to vary over nine decades. However, by combining minimal particle theory and data with supernova constraints, most of this range can be excluded. 6.1
Particle Physics
i) The decays of interest, n+ -> fi+n° and n° -» e+e~vd, can be described by the following four effective Fermi lagrangians: Gll%Y5{lo[Ya}^%{YSorYa75}^n-
•
0°)
and
Gn.vn.rvVt,ve.-rvt-.
(ii)
243
where f is some Dirac matrices. Note that the interactions here may differ from those in the standard V-A form. Disregarding the 0(1) corrections stemming from this difference, we can relate the new and standard pion and muon decays: +
r(n° -> e~e vd) = T(fi -» evv)
2.10J
f
G *
GF
(r ^
2
•
GF
Mu
; 2 . 1 0 % (sec)"1
(12)
and Br(7r+ -> /J + n°) Br(w+ -> n+v^
«1
f ^
0.016
'6'2
s 0.016 e^.
(13)
KGFJ
In the last equation, pn- (the velocity of the n° particle for the charged pion DAR), represents the relevant phase space ratio and a common F„ factor has been cancelled. The basic relation, Eq (8), then implies £,, £„• = 10"12, and from Eq (9), E^ <1.5xl0" 6 ore„- > 0 . 7 x l 0 - 6 . ii) The n" particle should be an SU(2)xU(l) singlet; otherwise, it would have increased the Z width like an extra neutrino. If the left-handed u,d quarks participate in the (n+ -» /J.+n°)decay, then a global SU(2)L transformation takes the charged pion and muon to the neutral pion and muon neutrino, respectively — leaving the n° invariant. Hence, we have a n° —> n + vM vertex of equal strength, as n+ —> n + JX+. iii) If we maintain lepton number conservation, the n° particle carries a (muon) lepton number and and so should vd in n° -» e~e+vd . The simplest alternative, avoiding the introduction of yet another new light fermion, is to identify the "decay" neutrino with the ordinary muon neutrino. In this case, we can apply an SU(2)L rotation to the n° decay process and, after crossing, achieve an additional channel for muon decay: fi~ —> n° + e~~ + v„ , controlled by the same Gn. and therefore of a rate roughly £„• times smaller than that of n~ —> e~ + ve + //^ . The electron spectrum in muon decay has been studied extensively and it agrees very well with the expectations for the standard decay into two massless neutrinos. Therefore, one can exclude the new channel at the promile level. Hence, if the assumptions made above are indeed correct, we have en- < 10"3, or equivalently x > 1 sec. This, then, would leave only the range 103 sec > x > 1 sec as acceptable.
244
iv) Even without the specific assumptions as in (iii) above, we can still motivate a weaker bound e„- < 1 and the corresponding lifetime bound % > 1 sec. In any renormalizeable theory, the four-Fermi coupling G, just like the ordinary GF, should emerge from tree diagrams with dimensionless couplings. Hence, in analogy with
GFF = & L , we would have Gn, = g™
or
G
„ "
g
^Se-v M\.
X° X~~ some new bosons. The mass of the new putative charged boson should exceed that of the W boson, or else it would have been produced in LEP. Likewise, we should have the neutral X° particle heavier than the Z boson or else it would manifest e+e~ colliders. Recalling that gw = 1/2, we cannot then have £„• » 1 unless we have new dimensionless couplings — i.e., the g's, which considerably exceed unity. 6.2
The Effect of Emission ofn' Particles from and Decays out of Supernova Cores
A key observation is that n° particles are copiously produced in the initial hot core (temperature T = 30 MeV). The reaction e+e~ -» vdn°, which is a crossed version of the original n° decay vertex, and v^N —> n°N (N = nucleon), which is generated via neutral pion exchange and its muon neutrino — n° vertex (see a(ii) above), have cross-sections G?„ E} - E„,„ and G}.El ~ £,,_ .respectively. Since the neutrinos and positrons are confined in the core for (1 sec), they suffer ( R ) many collisions, ncMisions = ——
2
(4ifP = (ncr)"1 is the corresponding mean free
paths) — roughly 108 and 106 for the neutrinos and positrons, respectively (the reduced number of the latter reflects the smaller density of electrons as compared with nucleons in the core). In each weak interaction mediated collision with sufficient energy in the center mass system of the colliding particles, there is a probability £„• or eM of producing an n° particle. The total number of n° produced is therefore: Nm(n')
- W ( l 0 % + 10«£„.)/|, > N(2.Wje^fB)
= 20A^
(14)
with N the total number of neutrinos and f the "Boltzman Factor," reflecting the mild suppression of neutrinos (or of positrons) with the requisite energy E > M„- = 33 MeV. Finally we used a + b> l4ab and the basic condition £„• £n = 10"12. Eq (14) does not imply that we produce more n° particles than the total number of neutrinos, but rather that we produce as many n° particles as neutrinos over a very short time
245
interval (0(0.1 sec) or less). We next follow the fate of these many n° particles and show that, in almost every scenario, they cause drastic, unacceptable changes in observed features of supernovae. i) Since the n° particle has no electroweak charges and is not hadronic, we expect it to interact more weakly then neutrinos. If £„• < 1, then the reverse n° scattering into neutrino will not suffice to recapture the n° particles in the core. Also the corresponding n° lifetime is then longer than 0.01 sec, allowing the n° particle to escape from the core before decaying. This will then result in a catastrophically rapid cooling of the core, and the SN 1987a neutrino pulse should have been drastically shortened and weakened compared with observations. ii) Even if the n° particles possess new interactions that confine them to the core for the duration of the neutrino confinement, the n° particles will equilibrate with the neutrinos and will eventually be emitted along with the neutrinos. Rather then carry most of the collapse energy, the n° particles will, in this case carry only a fraction, f/6 = 0.05, with f = 1/3 the Boltzman factor for a final temperature of T = 10 MeV, and we account for the fact that we have six neutrino species. If the lifetime of the n° particles is longer than 100 seconds, most will decay outside the progenitor SN1987a. The annihilation of the decay positrons near the progenitor would lead to a very strong gamma flash almost coincident with the neutrino pulses — which has been strongly ruled out [24]. Also, had the n° particle lifetime been considerably longer, then the escaping positrons from all prior galactic supernovae accumulate in the galaxy and build up a strong gamma annihilation line [25]. Note, however, that lifetimes longer than 1000 seconds are anyway excluded by the PSI accelerator bound. If the lifetime lies in the range of 0.01 to 100 seconds, the n° particles will escape the core and decay within the progenitor. It is well known that the actual supernova explosion carries only few promiles of the collapse energy. The extra energy deposition would have strongly enhanced the shock and shortened the observed three hours lapse between the neutrino pulse and the optical flash; see ref. [26] for further details. 6.3
The KARMEN Anomaly
Our discussion has not completely excluded the n° particle as an explanation for the KARMEN anomaly. Thus, if we do not assume, as we did in section 6.2(iii) above, that the decay neutrino in the 3-body final state is the muon neutrino, the n° lifetime can be shorter than 0.01 seconds — corresponding to £„• > 0.1. In this case, the n° particles will not escape from the core and the supernova bounds will be void. The same conclusion holds if we postulate a new n°-nucleon interaction which is stronger rather than weak. In this case, again the n° particles will be confined to the core and the collapse energy will predominantly escape via the conventional
246
neutrino emission. In either case, new four-Fermi constants of the order of the standard Fermi constant or larger, is required. E„- > 1 is particularly puzzling as eneu = 10"12 implies eM < 10"12 — a rather striking disparity between the two new couplings which are associated with the same new physics. We note that the large couplings of n° particles and - to electron positron or of the n° particle to nucleons will prevent their decoupling in the early universe evolution at tempratures higher than GeV, the decoupling temperature of neutrinos. Since, however, the particles decay at t = 0.001 seconds when the temperature is of the order of 30 MeV — i.e., the mass of the n° particle itself, the effect on nucleosynthesis may be ameliorated. 7
Summary and Conclusions
This talk and write-up has been rather heterogenous, containing crude estimates, fancy (and unlikely) scenarios, and a lot of arguments — which by now may be redundant — why the KARMEN anomaly is unlikely to be explained by a new particle. Yet it is perhaps proper in this meeting in honor of Frank Avignone, who has through honest, solid, relentless efforts helped forge the present field of neutrino physics and who is always open to the mysterious and the unexpected. References 1. S. R. Elliot, A. A. Hahn, and M. K. Moe, Phys. Rev. Lett. 59 2020 (1987); H. S. Miley, F. T. Avignone III, R. L. Brodzinski, H. J. Collar, and J. H. Reeves, Phys. Rev. Lett. 65 3092 (1990) 2. C. E. Aalseth et al., Phys. Rev. C59 2108 (1999) 3. R. J. Creswick et al., Phys. Lett. B427 235 (1998) 4. F. T. Avignone III et al., Phys. Rev. Lett. 81 5068 (1998) 5. J. N. Bahcall, these proceedings 6. H. W. Sobel, these proceedings 7. M. Gell-Mann, P. Rammond, and R. Slansky, in "Supergravity," eds. P. van Niuweuizen and D. Z. Freedman, North Holland Pub. (1979); T. Yanagida, lectures at the meeting "Grand Unification and Baryon Number of the Universe," KEK, Tokyo (1979) 8. G. Gelmini and M. Roncadelli, Phys.Lett. B99 441 (1981); H. Georgi, S. L. Glashow, and S. Nussinov, Nucl. Phys. B193 297 (1981) 9. S. Nussinov, Phys. Lett. B63 201 (1976); L. Wolfenstein, Phys. Rev. D18 958 (1978) 10. Baltz, A. S. Goldhaber, and M. Goldhaber, Phys. Rev. Lett. 81 5730 (1998); H.Georgi and S. L. Glashow, hep-ph/9808293 11. R. N. Mohapatra and S. Nussinov, Phys. Rev. D60 013002 (1999) 12. S. Babu, J. Pati, and F. Wilzeck, Nucl. Phys. B566 33 (2000)
247
13. R. N. Mohapatra and G. Senjanovic, Phys. Rev. Lett. 44 912 (1980) 14. F. T. Avignone III, these proceedings 15. Georg Raffelt, "Stars as Laboratories for Fundamental Physics," Univ. of Chicago Press (1996) 16. J. R. Wilson and R. W. Mayle, Phys. Rev. 163 63 (1998) 17. Y. Smirnov, D. N. Spergel, and J. N. Bahacall, Phys. Rev. D49 1389 (1994) 18. P. J. Kernan and L. M. Krauss, Nucl. Phys. B437 243 (1995) 19. H.-J. Janka, W. Keil, G. Raffelt, and D. Seckel, Phys. Rev. Lett. 767 2621 (1997); S. Hannestad and G. Raffelt, APJ 507 339 (1998) 20. G. Cocconi, Phys. Lett. B206 705 (1988) 21. Dar, J. Goodman, and S. Nussinov, Phys. Rev. Lett. 58 2146 (1987) 22. R. N. Mohapatra, S. Nussinov, and X. Zhang, Phys. Rev. D51 3843 (1994) 23. K. Eitel, these proceedings 24. Dar and S. Dado, Phys. Rev. Lett. 59 2368 (1987) 25. Goldman, R. Mohapatra, and S. Nussinov, Phys. Lett. B281 151 (2000) 26. A.D. Dolgov, S. H. Hansen, G. Raffelt, and D.V. Semikoz, hep-ph/0002223
5ti
Fe(i/ e ,e ) 56 Co: A Technique for an Accurate Cross Section Measurement at ORLaND Yu.Efremenko Department of Physics and Astronomy, University of Tennessee, Knoxville, TN 37996, E-mail: efremenkQunix.utk.edu F.T.Avignone Department of Physics and Astronomy, University of South Carolina, Columbia, SC 29208, E-mail: [email protected] A.Mezzacappa Physics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, E-mail [email protected]
A detector design is presented that would allow the measurement of the cross section of S6 Fe(i/ e ,e - ) 56 Co with a statistical accuracy of a few % at the ORLaND facility at the SNS. Monte Carlo simulations are presented to document the detector response.
1
Motivation
Core collapse supernovae are stellar explosions disrupting almost entirely stars more massive than 8—10 M 0 and extremely important for Galactic dynamical and chemical evolution because of their energetics and nucleosynthesis. Neutrino-nucleus weak interactions play a central role in both supernova dynamics and supernova nucleosynthesis. These interactions are also central to existing and proposed terrestrial facilities to detect neutrinos from the next Galactic or near-extra-Galactic supernova, which in turn will provide detailed neutrino "lightcurves" from which supernova models and supernova nucleosynthesis models can be diagnosed and improved. Although many nuclei participate in these weak interactions and it is impossible to measure all of the relevant cross sections, measurement of a few key cross sections would provide invaluable checks on the elaborate theoretical models used to compute them 1 . At the end of a massive star's life, its core is composed of iron-group nuclei. This core is surrounded by layers of successively lighter elements, such as silicon, oxygen, carbon, helium, and hydrogen. The core becomes gravitationally unstable, collapses on itself, rebounds at supernuclear densities, and generates a shock wave that propagates out of the core, through the outer 248
249
layers, disrupting the star in a supernova. The stage is set, i.e., where the shock forms in the core and how much energy is imparted to it when it does form, by the "deleptonization" of the core during collapse. Deleptonization occurs via electron capture on iron-group nuclei and on free protons. Initially, the electron neutrinos produced escape the core, carrying away lepton number. However, as the core densities increase beyond 1011 g/cm 3 , the electron neutrinos become trapped, and electron neutrino capture begins to compete with electron capture until, eventually, these two inverse weak interactions are equilibrated. At this point, the deleptonization "ceases" on the timescale of core collapse. Thus, one or two measurements of the cross sections for electron neutrino capture on iron-group nuclei (e.g., <j[56Fe(i/e, e~) 56 Co]) would provide important input to models of stellar core collapse and the post core-bounce supernova evolution. After the shock forms, the stellar core is stratified. Neutrinos of all flavors are radiating from a proto-neutron star at the center of the explosion at the staggering rate of 1057 neutrinos per second and 10 45 Watts. The electron neutrinos and antineutrinos power the supernova via neutrino absorption on the dissociation-liberated nucleons behind the shock. All three flavors of neutrinos interact with nuclei in the outer stellar layers, via neutral- and charged-current reactions, producing nuclear transmutations that play an important role in supernova nucleosynthesis. In some cases, these neutrino-nucleus interactions provide the channels for the production of some of Nature's rarest isotopes 2 . The same technique proposed to measure the neutrino-iron cross section can be used to measure the neutrino-nucleus cross section on any of the following nuclei: 7 Li, 9 Be, n B , 27 A1, 4 0 Ca, 51 V, 52 Cr, 55 Mn, 59 Co, 9 3 Nb, 115 In, 181 Ta, and 209 Bi. Finally, iron is one of the materials proposed for the next-generation supernova neutrino detector, OMNIS, now in its planning stages 3 . Accurate measurements of both charged- and neutral-current neutrino-iron cross sections would be invaluable in its design and, ultimately, in the interpretation of its supernova neutrino data. 2
Theoretical Background
The first estimate of the neutrino-iron cross section was made by Bugaev 4 using the shell model. More recently, the inclusive cross section of 56 Fe(i/ e , e _ ) 5 6 Co, averaged over the stopped muon decay neutrino spectrum, was calculated by Kolbe, Langanke, and Martinez-Pinedo 5 , and by Kolbe and Langanke 6 . In ref. 6 , the neutral current cross sections for the production of neutrinos and their spectra are presented for both Fe and Pb.
250
The charged-current neutrino-nucleus cross section between nuclear states with specific angular momenta and isospin was derived by Walecka 7 , and by Donnelly and Peccei 8 . This cross section involves a Gamow-Teller (GT) contribution and forbidden transitions. Random phase approximations (RPA) are not as reliable as shell model calculations for transitions mediated by the GT operator; however, they do describe forbidden transitions adequately 5 . The shell model calculation of the GT contributions to the cross section 56 Fe(i/ e ,e _ ) 5 6 Co of refs. 5 and 6 were performed within the complete pf-shell using the large-space Strasbourg-Madrid codes developed by Caurier 9 , 10 . The forbidden transitions were calculated with the RPA. The resulting cross section, weighted by the neutrino spectrum from decay of a stopped muon, is 2.64 • 10~ 40 cm 2 . The only measurement of this cross section was reported by the KARMEN collaboration [2.56 ± 1.08(stat) ± 0A3(sys)] • 10 _4O cm 2 n . A much more accurate measurement is needed. According to a computational feasibility study reported here, it can be measured with a statistical accuracy of a few percent with a high-granularity detector described below called SOUDANINO. 3
Feasibility Study of the Proposed Detector
The geometry and principle of operation of the SOUDANINO detector are similar in principle to those of the Soudan-II detector, from which it gets its name 12 . It is composed of iron tubes with thin walls, each containing a position-sensitive gas tube; see Fig. 1. The signal would be read out separately from each individual tube. Particle energy is reconstructed by the range of the particle track or by the total number of fired cells. Directional information can be extracted from the reconstruction of the track. In principle, this detector can be constructed in such way that iron tubes will be replaceable. After making measurements with one set of "target" tubes, they might be replaced by the tubes made from a new material (new target). As a result, neutrino interactions with different nuclei can be studied with the same detector. Detectors with gas tubes as the sensitive elements have a number of advantages compared, for example, to those with scintillator rods. In general, they are less expensive, and do not require a complex light-readout system. In addition, the low mass of this part of the detector eliminates the necessity to subtract interactions in the target from the interactions in the detector itself. The energy resolution for electrons in the energy range of a few tens of MeV is better by measurement of the track length than by measurement of total energy deposition. The design of such a detector should be a reasonable compromise between
251
Figure 1: Schematic view of proposed SOUDANINO detector.
cost and performance. The thinner the tube walls, the better the energy and angular resolution. On the other hand, the thinner tube walls give the detector less average density and, as a result, increase the size and the number of channels. We obtained an initial optimization of the detector using a GEANT Monte Carlo code. As a result of this study, we conclude that the optimum iron wall thickness is 0.5 mm. The dimensions are OD = 10 mm and ID = 9 mm. In the rate calculations, the detector was assumed to be at a distance of 50 m from the 2-megawatt target of the Spallation Neutron Source (SNS). For the theoretical cross section of 2.64 • 10~ 40 cm 2 , the neutrino flux at the detector of 3 • 10 6 i/ e cm _ 2 sec _ 1 , and the number of iron atoms in one ton (9.9 • 10 2 7 fon - 1 ), the expected reaction rate is 250 events y~1ton~1. We estimate a detection efficiency of 30%. For a desired reconstructed event rate of 1000 y~l, the active detector mass should be 13.3 tons, corresponding to a fiducial volume of 2.5 x 2.5 x 2.5 m 3 . This fiducial part of the detector should be surrounded by at least 50 cm of an active detector. As a result, the total detector size would be 3.5 x 3.5 x 3.5 m3 and would consist of a 314 x 314 array of parallel tubes. During the Monte Carlo study of the SOUDANINO performance we compared two types of active elements for the detector, scintillator rods and gaseous drift tubes. The results of the simulation of SOUDANINO, for 30-MeV electrons for both gas tubes and scintillator rods, are shown in Fig. 2. For this simulation, an ideal light readout (no photo-electron statistical fluctuations)
252
was assumed. The energy resolution by measurement of the total energy is 21.4%; the energy resolution measured by track range is 23.4%. For the real detector, the first value will be worse because of the fluctuations in the light collection system. On the lower part of the figure, the simulation of the detector with gas tubes is shown. The average number of fired cells is less and the resolution is slightly worse for gas tube detectors compared with scintillators, because of the lower efficiency to low-energy photons.
400
;
^*"&-»
'
\
)
Enei gy resolution 21 %
200 0
*r , , . J
0.095
. . . . iK-_. . .... I ... . 0.01 0.015 0.02 0.025 0.03
Energy, GeV •2*00 400 200 0
600 400
prift tubei
:
200 0
"S.— . i . . 15
20
. 25
30
Nhits Figure 2: Detector energy resolution, with scintillator and gas drift tube sensitive volume. Energy reconstructed by total energy deposition (upper figure), by the number of fired cells for scintillator (middle figure), and by number of gas drift tubes (lower figure). The electron energy is 30 MeV.
For the energies of interest, the probability to start an electromagnetic shower is small; therefore, the measurement of the energy using the number of fired cells for electrons has excellent linearity as can be seen in Fig. 3. For the iron detector with tubes 1 cm in diameter and walls 0.5 mm thick, the average density will be 1.35 g cm~3. The 13.5-ton cubic detector will have a fiducial dimension of 2.5 meters on a side. With a fiducial cut of 0.5 meters around the entire detector, the total dimension should be 3.5 x 3.5 x 3.5 m 3 . The number of tubes to be read out will be about 105. The measurement of the energy by range does not require amplitude information from the tubes, and as a result, a relatively simple, inexpensive electronic system can be used. The results of simulations for this type of detector are presented in Fig. 4 where
253
/* -/»
u
Close -...lfj mm..j ipea. uth_C ,£mni_wfllla
14 12 10 8
i
/
6
(l^"^
4 2
o,
—l
Open =J20 nm..pipea..withJl.tamjsiiills....
l*^
10
15
20
25
30
35
40
45
SO
55
E,MeV Figure 3: Linearity of the SOUDANINO detector for two different tube thickness for gas drift tubes.
the energy and angular resolutions are presented. For electrons with an energy of 30 MeV, the predicted energy resolution is 25% and the angular resolution is about 15°. The major source of beam-correlated background will be interactions of high -energy neutrons produced in the SNS target and penetrating into the ORLaND facility. Their interaction might knock out a high-energy proton which could imitate the track of an electron. However, to fire the same number of cells as an electron, the proton should have a much higher energy, see Fig. 5. For example, a 140-MeV proton can produce the same number of hits as a 35MeV electron. Already this energy-cut significantly reduces the contribution from this source of background. In addition, it is possible to separate electrons from protons with the same number of hits by the shape of the track. The proton tracks are more linear, and the electron tracks are irregular in shape because of the emission of photons. In Fig. 6, examples of the coordinates of firing cells, for 40-MeV electrons and 125-MeV protons are shown. By the introduction of the "track linearity parameter," which is the \2 of how well one can fit the track with a straight line, one can separate electrons from protons for a signal with the same number of hits. See Fig 7, and Fig 8. Of course, more sophisticated algorithms can produce better separation.
254 ^
45
sf & •3 35
•pp©fTt«»-^QtTrm"|^p®s-iTth"thm-mte"i
% 30 K 15
J „ ..
4 MthJtBmrA. wMi
35
40
„1C1CUUBL
!
10
15
20
25
30
45r
50
55
E,Mef p 3
tt °
! !
c
= <> If— -: - . ,'„ —% 6 / 5 6; g / 8 t-;
"
"
'
•
i
-j
•"" *
"
..„.„4ji
.„_
>
|J
JUJ-J^.
10
15
20
25
30
,
; !
„
1
:
-
.... J
35
40
I*1"
M
i
"T
1
;•
|
.-Lj-w-LxjL^i
45
S
*E, hef
Figure 4; Energy and angular resolution for the the SOUDANINO detector for two different tube thicknesses.
38 35 * ^ 30 25 20
L JE r oJ. L ji . _ i ipnai \r
1
:
|
A ;
4
\
j
{
i_ _ j _ J 0
OM
|
|
i
;
|
j
J l l
.J
- } 4iiil^^^^^ -jiiiiliiiiiiSSp^^^^^B
• - ' • ' • . ' • \ 10
i
\ |gff^^
\,,m ffHM
0.04 OM
0.08
0.1
0.12 0.14 0.16 0.18
02
Energy, GeV
Energy, GeV
Figure 5: Detector response for protons and electrons.
255 i
i
i
I
1
f «,-;•-
§ /#5 125.MaSf.ptatoD
f- 5-
-f-
!
!
I.....
1
il5a.Mg^protor|
.,.,.,..r.i^.,..r.
i
j
/
1
—•
fcECCT^-r-rr
i
1 v t
!
jj..,.. r
w
90
JS9
JJ0
«
Xcorr
§ 1/5 I
r
:
;
j4Q M«Mkkcirfl4
j4Q May>l«ciro4
1
_
p B-r-j ., 100
110
„4,..,-,- .r.| -,..,-,-.»-.[.,...,-,.,.. Iffl
129
110
129
Xcorr
Xcorr
Figure 6: Sample track for 125-MeV proton and 40-MeV electron in the SOUDANINO detector.
*"p22
l
1
!
1
:
L
Emtdas.„.„..]
!
S17.5 i-
••;: "*;::
~ ,«._,,..,„„„, ••Ili^SSS
7,5
tzS
llilB Nhtts
Nhiis
Figure 7: Track Linearity parameter for the protons and electrons in the detector.
256
6
7
8
9
10
Track nonlinearity i
i Iprotc ns
\
1
j > * * " ^
; ;
i
u
!
.—-|§iectfons
|
J j j n
I,*?
|
— i — i,,.. 2
J
4
5
S
7
S
P
/
0
Running summ of track nonlinearity
Figure 8: Rejection of the protons from the electrons by Track Linearity parameter.
4
Conclusions
We have investigated in detail the response of a highly segmented detector with a fiducial mass of 13.3 tons of iron-clad gas tube detectors. An accurate cross section of the reaction 5 6 Fe(i/ e ,e - ) 5 6 Co can be measured with such a detector placed at 50 meters from the target station of the SNS (under construction at ORNL). Detailed Monte Carlo computations show that simple gas tube detectors with position sensitivity give an energy resolution comparable to scintillation detectors. This technique can be applied to any targets that can be fabricated into tubes of 1 cm OD, with thin walls and with lengths of 3.5 m. These measurements would yield ve -nucleus cross section of great importance to neutrino astrophysics. 5
Acknowledgments
YuE is supported by the Oak Ridge National Laboratory and the University of Tennessee, FTA is supported by the U.S. DoE Office of High Energy Physics, and the University of South Carolina, AM is supported at the Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U.S. Dept. of Energy under contract DE-AC05-00OR22725.
257
References 1. E. Kolbe, and K.-H. Langanke, and G. Martinez-Pinedo, Phys. Rev. C 60, 52801 (1999). 2. S. E. Woosley, and D. Hartmann, and R. D. Hoffman, and W. C. Haxton, The Astrophysical Journal 356, 272 (1990). 3. D. B. Cline, et al, Phys. Rev. D D50, 720 (1994). 4. E. V. Bugaev et al., Nucl. Instrum. Methods A324, 350 (1979) 5. E. Kolbe, K. Langanke and G. Martinez-Pinedo, Preprint: arXiv:nuclth/99050013 Aug, 1999. 6. E. Kolbe and K. Langanke, Preprint: arXiv:nucl-th/0003060 27 March 2000. 7. J. D. Walecka, in Muon Physics, eds. V. W. Hughes and C. S. Wu (Academic Press, New York, 1975) p. 113. 8. T. W. Donelly and R. D. Peccei, Phys. Rep. 50, 1 (1979). 9. E.Caurier, computer code ANTOINE, CRN, Strasbourg, 1989 10. E. Courier, G. Mctrt'nez-Pirwjdo, F. Nowacki, A. PoveSj J-Retamosci and A. P. Zukes Phys. Rev. C C59, 2033 (1999) 11. K. Eitel, Private Communication, G. Ruf, diploma the sis, University of Bonn. 12. W.W.M. Allison, Nucl. Instrum. Methods A376, 36 (1996)
CALCULATING N E U T R I N O - N U C L E U S I N T E R A C T I O N S D.J. D E A N Physics
Division,
Oak Ridge National Laboratory, Oak Ridge, E-mail: [email protected]
TN 37831,
USA
I briefly review the role of weak interactions on nuclei. I then review the status of calculations necessary to provide high-quality results for a description of relevant experimental data that may come from the proposed ORLaND facility.
1
Introduction
A variety of experiments may be performed using the neutrinos originating from the decay at rest pions that will be produced at the Oak Ridge Spallation Neutron Source. For example, one fundamentally important quantity is the strangeness content of the nucleon which could be measured provided that a useful experiment could be identified.1 Neutrino interactions with the deuteron (e.g. ve + d -> ue + n + p) are very important reactions used by the Sudbury Neutino Observatory (SNO) to study the solar neutrino flux. This reaction has recently been extensively studied in effective field theory,2 and was shown to depend on one coefficient in the next-to-leading order expansion. SNO is unable to accurately measure this cross section, and a calibration experiment would be extremely valuable. Weak interactions on nuclei occur naturally during the various phases of supernova evolution. During the final stages of the stellar evolution of a ten to twenty solar mass star, an iron core develops as the nuclear fuel is exhausted. Outward pressure from liberated electrons keeps this core from collapsing until it reaches the Chandrasekhar mass. At this point, electrons are captured by nuclei through the Gamow-Teller resonance, thus depleting the lepton number within the core region. The interaction e~ + A(N, Z) ->• A*(N +1, Z - 1) + ve becomes a source of electron neutrinos within the supernovae. As the collapse proceeds, a bounce develops when the central density in the neutron rich core exceeds nuclear matter density; however, numerical studies indicate that the energy obtained from this nuclear bounce is not enough to overcome the gravitational attraction of the matter. Neutrinos again play an important role as they may deposit their energy to the matter in the stalling shock front, imparting sufficient energy to overcome gravitation. Once a supernovae explodes, neutrinos again become important in r-process nucleosynthesis, particularly of heavy elements through a mechanism of neutrino spallation. 258
259
2
Neutrino-nucleus interaction
The formalism necessary to calculate total and differential neutrino cross sections on nuclei has been known for many years. 3 The charged-current cross section at a given incident neutrino energy Ev is °cc{Ev)
=^ j
dcosOdiEi - Ef + Ev - EfaE,
\ M |2 ,
(1)
where Et and Ef are initial and final energies of the nuclear states involved, Ei and pi are the energy and momentum of the emitted lepton, p„ is the neutrino momentum, and cos# = ip'i'fc1' i • The reduced matrix elements of the transition matrix element M, which connects states i and / (jf,Tf,Tzf J2 (jfTfTzf
|| QJ'T || JhThTZi) || [a]a x ajb]J'T
= \\ J^Tz^
{jata \\ QJT(q) \\ jbtb)
(2)
3a jb
where J ; j , Ttj, and Tztj are the initial and final spin, isospin, and isospin projection of the initial and final nuclei, and j a and ta are single-particle spin and isospin labels of the individual nucleons in the nucleus. We represent the expansion of the weak interaction in terms of multipoles ilJT of a given spin and isospin (JT) multipolarity. We note that fi carries a momentum transfer dependence q. The one-body matrix elements (/ || flJT \\ i) were tabulated, 4 and are readily calculable. The more difficult computation involves the onebody density matrix elements which are calculated between the final and initial nuclear many-body states. 3
M e t h o d s of solution to t h e nuclear m a n y - b o d y problem
Various approaches to solutions of the nuclear many-body problem exist. For light nuclei (A < 8) Greens Function Monte Carlo methods have been employed to solve for ground-state properties using realistic two- and threenucleon interactions. 5 ' 6 These methods are, however, limited to light systems and have not been developed for calculating scattering cross sections (although their antecedent Variational Monte Carlo methods have been used to study proton-nucleus reactions. 7 ) The shell model provides a tractable solution to the nuclear many-body problem in a truncated model space. Thus one divides the complete Hilbert space into an active and inactive space. The active space is usually referred to as the valence space. One natural form of truncation is to work in a major
260
oscillator shell. For example, if we choose 4 0 Ca as a core of inactive nucleons, we can work in the region of 21 < N, Z < 39. Thus, by solving the resulting eigenvalue problem for a given nucleus, we obtain detailed information concerning its spectroscopy and one-body density matrix. One must necessarily use many-body perturbation theory to project the realistic nucleon-nucleon interaction into the model space.8 This approach, while theoretically correct, encounters problems for nuclei towards the midshell for a variety of reasons. Perhaps the least well investigated at the moment involves inclusion of a realistic three-nucleon interaction into the many-body perturbation theory. GFMC results seem to indicate that one cannot properly describe the mass dependence of nuclei as a function of nucleon number without the three-body interactions, and thus it is not surprising that the resultant shell-model interaction derived from many-body perturbation theory still requires some modifications by slightly tuning to experimental data. In light nuclear systems such as 1 2 C, it is possible to obtain directly from experimental data enough information to fit nuclear matrix elements. 9 This is also possible in the scf-shell (nuclei from 1 6 0 to 4 0 Ca). 1 0 Over the years, good effective interactions have also been developed for the /p-shell (iron group nuclei). 11 Even in cases where reasonable effective interactions are available, one always faces the explosion in the dimension of the many-body basis in which one must diagonalize the effective Hamiltonian. Increasing computer power allows one to incrementally solve for larger problems by diagonalization. Alternate methods of solution based on quantum Monte Carlo techniques have been developed and used in regions where standard diagonalization could not at the time be successfully applied. 12 Another complication comes when one extends calculations beyond one major oscillator shell. In such cases, contamination from spurious center-of-mass effects must be corrected. Furthermore, overlapping methods to account for the giant-resonance regions are probably necessary. Thus, calculations of the neutrino-nucleus cross sections within a shell model approach will always have a systematic uncertainty arising from a) the effective interaction used; b) the size of the active model space; and c) center-of-mass contaminations. RPA-based approaches 13 typically begin with Skyrme Hartree-Fock generated basis states, thus determining the mean-field and single-particle occupied levels of the nucleus. The unoccupied levels are obtained by diagonalizing the Hartree-Fock mean field using a harmonic oscillator basis. Thus, the continuum part of the single-particle spectrum is discretized, and discrete particlehole configurations coupled to the appropriate spin and parity are used as a
261
basis in order to cast the RPA equations into a matrix form. Hartree-Fock is sometimes augmented by BCS pairing correlations. In this case, uncertainties arise from the effective interactions used and the allowed correlations available for the ground state. The lack of correlations beyond RPA typically causes an increased total cross section and a narrower distribution of strength within the various interaction channels. Recent applications of continuum RPA and the shell model to 12 C scattering give 13.4 - 14.5 x l O - 4 0 cm2 (shell model 15 ) and (18 - 20) x KT 4 0 cm'2 (CRPA 14 ) for the total cross sections. The experimental value is 12.3 x 10~ 40 cm 2 . 16 At present theoretical calculations have an associated 25-35% systematic uncertainty. 4
A guide from electron capture
Based on the above assessment, one could argue that we should simply avoid complicated calcuations with which to calibrate and interpret experimental data. This conclusion would be mistaken, as can be shown from investigations of the Gamow-Teller transitions in nuclei. A number of studies have shown that inclusion of the full space is necessary if one wishes to describe the Gamow-Teller strength distributions. Such Ohui calculations have been carried out in the p-, sd-, fp-, and giis-shells. For example, total GamowTeller strengths are well reproduced for a variety of /p-shell nuclei using a modified version of the Kuo-Brown interaction for this region. 17 The strength distributions for several nuclei are shown in Fig. 1. These calculations were carried out using shell-model Monte Carlo techniques. 12 We renormalize the total strength by 1/1.262, which is a standard factor for spin operators in Qhu) model spaces. These shell-model results should be compared with the more naive non-interacting shell-model results of Fuller, Fowler, and Newmann. 18 The total strengths of FFN are usually a factor of 2-10 larger than the full calculation. FFN places that strength into one transition peak (solid verticle line in Fig. 1). Clearly, the strength is fragmented across many states, and the centroid is usually displaced from the FFN value. These results may have serious consequences for the evolution of the collapse phase of type II supernovae. 19 ' 20 The inverse reaction, neutrino scattering or capture by iron, should also be studied both experimentally and theoretically. 5
Conclusions
Several important reactions that have a significant scientific impact may be studied at the proposed ORLaND facility. Each reaction requires sufficient
262
o.o
2.0
4.0
6.0 8.0 E (MeV)
10.0
12.0
Figure 1. Gamow-Teller strength distributions for selected nuclei. The dashed line represents experimental data, while the solid curve are the shell-model Monte Carlo calculations. Solid vertical lines represent the FFN parameterization.
nuclear structure computation in order to understand the resulting experimental signal. In the Introduction, I mentioned two such experiments: 1) ve + d ->• ve + p + n, and 2) a well-designed neutral current experiment to detect strangeness content of the nucleon. Additionally, several interesting experiments could be done with iron in order to study neutrino heating. For light systems such as the deuteron, effective field theory may be a useful theoretical framework in which to understand experimental findings. For heavier nuclei (carbon, oxygen, iron) the shell model and RPA will be useful tools.
263
Acknowledgments This research was sponsored by the Division of Nuclear Physics of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725 managed by UT-Battelle, LLC. The author acknowledges useful discussions with W. Haxton and B. Holstein. References 1. W. Haxton, private comm. 2. M. Butler and J.-W. Chen, arXiv:nucl-the/9905059, (June, 1999). 3. There are many review articles on this subject. See for example T.W. Donnelly, Prog. Part. Nucl. Phys. 13, 183 (1985). 4. T.W. Donnelly and W.C. Haxton, At. Dat. Nucl. Dat. Tab. 23, 103 (1979). 5. B.S. Pudliner, V.R. Pandharipande, J. Carlson, S.C. Pieper, and R.B. Wiringa, Phys. Rev. C56, 1720 (1997). 6. R.B. Wiringa, J. Carlson, S.C. Pieper, and V.R. Pandaripande, Phys. Rev. C, (in press, 2000). 7. R. Schiavilla et al, Phys. Rev. C58, 1263 (1998). 8. M. Hjorth-Jensen, T. T. S. Kuo, and E. Osnes, Phys. Rep. 261, 125 (1995). 9. S. Cohen and D. Kurath, Nucl. Phys. 73, 1 (1965). 10. B.A. Brown and B.H. Wildenthal, Ann. Rev. Nucl. Part. Sci. 38, 29 (1988). 11. A. Poves and A. P. Zuker, Phys. Rep. 70, 235 (1981). 12. S.E. Koonin, D.J. Dean, and K. Langanke, Phys. Repts. 278, 1 (1997). 13. C. Volpe, N. Auerbach, G. Colo, T. Suzuki, and N. Van Giai, arXiv:nuclth/0001050, January 2000. 14. E. Kolbe, K. Langanke, S. Krewald, Phys. Rev. C49, 1122 (1994). 15. A.C. Hayes, Phys. Repts. 315, 257 (1999). 16. As quoted in Ref. 15. 17. K. Langanke, D.J. Dean, P.B. Radha, Y. Alhassid, and S.E. Koonin, Phys. Rev. C52, 718 (1995). 18. G.M. Fuller, W.A. Fowler and M.J. Newman, ApJS 42, 447 (1980); 48, 279 (1982); ApJ 252, 715 (1982); 293, 1 (1985). 19. D.J. Dean, K. Langanke, L. Chatterjee, P.B. Radha, and M.R. Strayer Phys. Rev. C58, 536 (1998). 20. G. Martinez-Pinedo, K. Langanke, and D.J. Dean, ApJ. Supp. 126, 493 (2000).
S E A R C H F O R P S E U D O S C A L A R C U R R E N T U S I N G TT° -> vv' DECAY
Department
A. R. FAZELY of Physics, Southern University, Baton Rouge, E-mail: [email protected]
LA 70813
USA
The observation of the decay n° —^ vu' would imply new physics. The pion with a hadronic state of Jp = 0~ can decay to vacuum, Jp = 0 + only via weak axial vector (A) or pseudoscalar (P) operators. Momentum and angular momentum conservation require that the decay neutrinos possess the same helicity. A- and P-interaction favor leptons with opposite and the same helicity, respectively. Therefore, in A-interaction the unfavored helicity —/? is proportional to (1 — /?) and in the P-interaction the decay rate is proportional to (1 + f)). This implies that A-interaction forbids 7r° —• up decay for massless neutrinos with 0—1 and that P coupling maximizes decay rate for neutrinos with zero rest mass. Possible search for this decay in high intensity accelerators such as Spallation Neutron Source with suppressed cosmic ray background is discussed.
264
265
The observation of the decay 7r° -> vv would imply physics beyond the standard model. The pion decay with a hadronic state of Jp = 0~ to vacuum, Jp = 0 + can only be achieved by axial vector (A) or pseudoscalar (P) operators. In this purely neutral current process, momentum and angular momentum conservation require that the decay v and v possess the same helicity. Aand P-interaction favor leptons with opposite and the same helicity, respectively. Therefore, in A-interaction the unfavored helicity —/? is proportional to (1 — (3) and in the P-interaction the decay rate is proportional to (1 -f f3). This implies that A-interaction forbids IT0 -> vv decay for massless neutrinos and that P coupling maximizes decay rate for neutrino with zero rest mass. Equally interesting are the limits on decays n° -» vev^ which violates individual lepton number conservation in the neutral current sector. An experimental upper limit, r(7r° -» vev€)/r(n° -> all) < 2.4 x 10~5 (90% l CL), was set by Herczeg and Hoffman who used the K+ -> TT+VV data and focused on the K+ -> 7r + (108MeV)f v region of the decay spectrum. Hoffman 2 has set more stringent limits for the branching ratio for 7r° —> vv by using the data from several beam-dump experiments. Similar limits were obtained by Dorenbosch et al? For the LSND detector and the neutrino beam facility at the Los Alamos Meson Physics Facility (LAMPF) the TT° -> vv has been simulated. A 780-MeV beam of protons incident on a predominantly water-copper target was used as the source for the 7r°'s. The n" production is peaked forward, i/'s in the energy range 0-700 MeV, resulting from possible n° —• vv decaying in flight provided the neutrino beam. The LSND detector, located 30 m from the beam stop, contained 167 tons of diluted liquid scintillator which served as the active target. The liquid scintillator was viewed by 1220, 8"-diameter Hamamatsu photomultiplier tube (PMT) mounted inside the tank. An active shield with 292, 12.7-cm diameter PMT's vetoed cosmic rays4. The read-out and data acquisition system are described in more detail elsewhere5. The 7T° production rate was calculated using the "High Energy Transport Code"6. Our calculation shows that for every 780-MeV incident proton, 0.11 ± 0.02 7T° are produced in the beam stop. The ir° spectra is expected to be stiffer than the well-known TT+ spectra because of absence of energy loss due to ionization in the case of neutral pions. The pion spectra were extracted by Gaussian fits8 to the measured 7r+ spectra of Cochran et al.7. The energy spectra down to 0° and out to 180° were obtained by a linear extrapolation
266
-14 » 10
[D
-
0.7
Entries Mean RWS UOFLW OVFLW ALLCHAN
0.6
2011 1159809 2Z9.8 129.0 0.000
o.ooo 0.2249E-12
0.5
0.4
0.5
0.2
-
0.1
0
I
, . 1 , , , , 1 , , . , 1 , , , , 1 , , , , 1 . i , , l -<|< > nl i i i 100 200 300 +00 500 600 700
i 800
E, f r o m TT DIF (MeV)
Figure 1: Energy distributions for the v^'s as seen by the LSND detector
from the above data. A pion production Monte Carlo program based on these fits agrees well with measurements of stopped 7T+ 8 . The spectrum of neutrinos as seen by the LSND detector 9 is shown in figure 1. Because chiralities of both neutrinos are the same in this decay (both are either left handed or right handed) the actual number of neutrinos that are of the proper chirality to interact on nuclei is assumed to be (y + v)/2. The cross sections for the above reactions obtained from a calculation by Geisser and O'Connelf 0 . This cross section is calculated to be 0.5 x 10 - 3 8 cm 2 /neutron. The dominant reaction is Ve,»+12C -+e,n + p + X (1) This is the so-called quasi-free reaction and it has the largest cross section of all competing reactions. The search is done by identifying e's and yu's with energies above 60 and 120 MeV, respectively. These energies are chosen to eliminate electrons from stopping muons and muons from i/M 's interactions on detector nuclei due to 7r+'s decaying in flight. In the /J, case, a muon signature
267
is established by identifying a michel electron with spatial and time correlation with the muon. Given a decay-in-flight source at SNS for pions in the Oak Ridge Laboratory for Neutrino Detectors (ORLaND), one can search for 7r° —• vv decay. For a two kilo-ton detector, we expect limits of 5.0 x 1 0 - 9 for electrons and 8.0 x 1 0 - 9 for muons both at 90% C.L.. Similar limits are possible for the flavor violating decay. In summary, we believe that a dedicated decay-in-flight source at SNS for ORLaND would provide the necessary tools to search for -n° —> vv decay for both electron, and muon channels, For a two kilo-ton detector in ORLaND, we expect limits of 5.0 x 10~9 for electrons and 8.0 x 1 0 - 9 for muons as well as the flavor violating decay all at 90% C.L..
Acknowledgments This work was supported in part by the U.S. Department of Energy under the contract number DE-FG02-95ER40911. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
P. Herczeg and C M . Hoffman, Phys. Lett. B 100, 347 (1981). C M . Hoffman, Phys. Lett. B 208, 149 (1988) J. Dorenbosch, et al, Z. Phys. C404971988. J. Napolitano, et al, Nucl. Instrum. Methods A 274, 152 (1989). C. Athanassopoulos et al Nucl. Instrum. Methods A 388, 149 (1997). T.W. Armstrong, et al, Nucl. Sci. Eng. 49821972 and LAMPF report No. LA-UR 82-1589 1982. D.R.F. Cochran, et al, Phys. Rev. D 16, 3085 (1972). R.L. Burman, M.E. Potter, and E.S. Smith, Nucl. Instrum. Methods A 291, 621 (1990). Myuongkee Sung, private communication. T.K. Gaisser and J.S. O'Connell, Phys. Rev. D 34, 822 (1986).
FLAVOR-DEGENERATE PAIR P R O D U C T I O N I N N E U T R I N O - N U C L E U S COLLISIONS LALI CHATTERJEE Physics Department, Cumberland University, Lebanon, TN 37087, USA E- mail: Ichatterjee ©Cumberland, edu M. R. STRAYER Physics Division, Oak Ridge National Laboratory, P. O. Box 2008 Oak Ridge, TN 37831-6373 E-mail: [email protected] JIANSHI WU Department of Natural Sciences, Fayetteville State Fayetteville, NC 28301-4298 E-mail: [email protected]
University
Standard Model cross sections for flavor-degenerate pair production in neutrinonucleus collisions are presented for the first time. In the V-A approximation, results are seen to agree with published work from the pre-neutral current era. Cross sections are of the order of 1 0 - 4 1 cm2 for vc-gold collisions for 100 MeV incident neutrinos. Signatures of such events and similarities with leptonic electroweak channels are discussed.
Pair production in neutrino-nucleus collisions is representative of a class of processes which produce lepton-antilepton pairs in the electroweak sector and couple these electromagnetically to nuclei. As the nucleus participates via virtual photon exchange, the resultant cross sections are of order (GZa) 2 , where G is the Fermi constant and (Za) represents the electromagnetic coupling to the nucleus. These may be compared to processes of order (G 2 a), including bremstrahlung of neutrino pairs by electrons, 1 neutrino-photon collisions,2 and alpha order corrections to the lowest order electroweak channels. 3 The contributions from pair production could therefore surpass the (G 2 a) ones for values of Z exceeding 12, depending on the other effects such as phase space. Neutrino-electron scattering channels 4 are also of comparable order at sufficiently high Z values. The leading order neutrino-nucleus cross sections, behaving as (G 2 A), 5 remain higher than the pair-production channel. In this conference report, we focus on the production of first-generation flavor pairs only and consider incident neutrinos of electron and muon flavor. The channels can be represented as v± + A —> Vi + A + e + + e~, where i includes all flavors. The electron neutrinos can initiate the reaction 268
269
*e+
Figure 1. Feynman diagrams for both charged current and neutral current pair production in the field of a nucleus.
by both charged and neutral current channels, while the muon neutrino can only do this via neutral currents. The processes can also be described as bremstrahlung of charged pairs by neutrinos in the field of nuclei and represented by the Feynman diagrams of Fig. 1. We restrict our study to spin-zero nuclei that respond coherently and elastically. At the energies of interest, the effective Lagrangian approximation of the Standard Model, where the gauge boson propagators are collapsed to local couplings, may be used. Since the electroweak sectors in Fig. 1 represent purely leptonic currents, the form of the effective Lagrangian is well known. 4,6 The second-order S-matrix elements corresponding to electron-positron pair production by neutrinos can be written in a generalized form in terms of the appropriate effective Lagrangian as M=
SL(Ze2)F(q2)^[U(k')r(l-l5)U(k)} 1 1 -7aV(r + )] (1) [U(r-ha:ln(a ~ Hs) + ln(av ~ bj5); '((BL-m)"" " " "* '°'(®+-m)
where Q_ = r_-q and Q + = q-r + , a = -1/2 + 2 s i n 2 ^ and b = -1/2 for i/M, a = +1/2 + 2 sin20„, and b = +1/2 for ve and a = 1, b = 1 for the V-A case. 6,4 Equation (1) includes the direct and cross terms. U and V refer to the spinors for the respective particles or antiparticles. Q_ and Q+ are the fourmomenta of the fermion propagators, 'e' is the usual electromagnetic coupling constant, and q is the four-momentum transfer to the nucleus. P and q are connected to the four-momenta of the initial and final nuclei through q =
270
p-p' and P = p+p'. F(q 2 ) represents the nuclear form factor, and the other variables represent the four-momenta of the external legs. The Fierz rotation has been applied to the charged current channel as is customary. Introducing the phase space factors, the cross section can be written in terms of the matrix element as
° = lE^-J
-^^2El2E; r
£ y
I
M
I
fajsW
+
r++r-k-q).
spins
(2) The square of the matrix element is summed over final spins. There is no averaging over initial spins, as the neutrino is helical. The delta function ensures four-momentum conservation for the process. Conventions for phase space factors and gamma matrices are those of Bjorken.7 The phase space is evaluated numerically using Monte Carlo techniques. The charged lepton sector of the matrix element of Eq. (1) can be reduced to a compact form by using the Dirac equation and expressed as the sum of two terms. Mch = Mo+M', where Mo = [?(r_) 7 „(« - m5)V(r+)]
[ j g ^ L
- ^ ± L ]
(3)
and M'-Q(r
) r _ A M ° ~ frft) • „ (Q ~ ^ X ^
(4)
such that M = ~^[U{k')^{\
- -y5)U(k)]Mch.
(5)
This decomposition allows us to isolate the leptonic part of (M0) that coincides with the standard electroweak leptonic processes like e-v scattering and muon decay, for which the spin summations and angular correlations are well known. We have seen numerically that the M' term contribution is small over a wide range of neutrino energies. This is not surprising since for real photon channels this term does not contribute, and the two cross sections should approach each other as q2 —• 0, in accordance with the equivalent photon descriptions. 89 We have compared our results calculated for 56 Ni in the V-A approximation with similar results from Czyz10 and observe excellent agreement. We have also tested the analytic approximation Mo for Mc/j against exact numerical calculations for the full neutrino cross sections, and we find excellent agreement for gold. This agreement indicates the term M' does not contribute
271
appreciably, and thus the simplified form, Mo, may be used for various applications. From the laboratory detection angle, the event signature would be the produced pair. Such pairs should be distinct from those arising from deexcitation of nuclei, as the energy available to the latter pair is fixed by the de-excitation energy of the nucleus and the correlation signature of such a deexcitation pair. In contrast, the energy distributions of the neutrino-initiated direct pairs display continuum characteristics, and the angular distributions respect helicity constraints of the electroweak vertex and the physics of the photon vertex. The compact form of the matrix element squared in its Mo form and the similarity of the weak leptonic sector with the similar quantity from e-v scattering suggested the possibility of processes like this serving as a tool in low-energy explorations of sin29w. In fact, we could express the ratio of ue to Up cross sections as
ZZL = {(s+l)2B1+s2B2+m2s(s-l)B3} avll {(s-l)2B1+s2B2+m2s(s-l)B3}'
[)
Bi, B2, and B3 can be found numerically at given energies, and s = sin 2 0 w . We present in Fig. 2 the above ratio as a function of energy for a range of s in gold. Other similar ratios that may be of interest to experimentalists can be easily computed. For instance, the ORLaND proposal 11 proposes to use a combined electron-neutrino, anti-muon neutrino cross section in the ratio to determine s from electron scattering. This choice is influenced by the time distributions of the SNS neutrinos. Similar incident particle combinations could be used in the ratios for the pair production cross sections. The connection between the processes studied in this paper and photonneutrino pair production can be easily seen by cutting off the nuclear legs in the Feynman diagrams of Fig. 1. We would like to thank Professor Leo Stodosky for pointing out the possibility of using the neutrino-nucleus pair production as an experimental tool to study the photon-neutrino cross sections. 12 The question of possible roles in astrophysics also arises. What might be the astrophysical environments to spawn such neutrino-bremstrahlung pairs? High Z environments such as pre-collapse supernova cores, scenes of the stalled shock, or the neutrino-driven winds hosting the r-process and neutrino nucleosynthesis suggest themselves as potential sites. The density of the Fermi sea, however, would inhibit low-energy electron production in pre-collapse cores. Including flavor variety, the formalism can be readily extended to degenerate flavor production beyond the first generation and to nondegenerate flavor production using appropriate choices for the constants in Eq. (1).
272 10.0
9.0
-&
-
• — • Sin 2 (i\,) = 0.2
* - - - » SinVw) = 0-3 8.0
7.0
-Jill
-
6.0
5.0
4.0 -
3.0
2.0
-
-
1.0 0.0
10
100
1000
E v (MeV)
Figure 2. The ratio of electron neutrino cross section to the muon neutrino cross section ; a function of neutrino energy for a range of values of the Weinberg angle.
Acknowledgments
This research was sponsored by the Division of Nuclear Physics of the U. S. Department of Energy under Contract No. DE-AC05-00OR22725 managed by UT-Battelle, LLC. Lali Chatterjee would like to acknowledge research support from a National Science Foundation Grant No. PHY-0074759 and travel support from Cumberland University. Jianshi Wu would like to acknowledge support from a DOE Grant No. DE-FG02-97ER410144.
273
References 1. P. J. Schinder et al, Astrophysical Journal 313, 531 (1987); D. A. Dicus, Phys. Rev. D 6, 941 (1972). 2. D. A. Dicus and W. W. Repko, Phys. Rev. Lett. 79, 569 (1997). 3. J. N. Bahcall, M. Kamionkowski, and A. Sirlin, Phys. Rev. D 51, 6146 (1995). 4. F. Scheck in Leptons, Hadrons and Nuclei (North Holland Publishing, 301, 1983). 5. S. E. Wooseley et al, Astrophysical Journal 356, 272 (1990); J. S. O'Connell, T. W. Donnelly, and J. D. Walecka, Phys. Rev. C 6, 719 (1972). 6. C. Caso et al, European Physical Journal C 3, 1 (1998). 7. J. D. Bjorken and S. D. Drell in Relativistic Quantum Mechanics (McGraw-Hill Book Co., 1964). 8. C. Bottcher and M. R. Strayer, Phys. Rev. D 29, 1330 (1989); M. Dress, J. Ellis, and D. Zeppenfeld, Phys. Lett. B 223, 454 (1989). 9. L. B. Oken in Leptons and Quarks (North-Holland Publishing Co., Amsterdam, 1982); M. A. Kozhushner and E. P. Shabalin, Sov. Phys. JETP 14, 676 (1962); E. P. Shabalin, Sov. Phys. JETP 16, 125 (1963). 10. W. Czyz, G. C. Sheppey, and J. D. Walecka, Nuovo Cimento 34, 404 (1964). 11. ORLaND proposal 12. L. Stodosky (Munich), private communication, March 12, 2000.
CUORE AND CUORICINO ETTORE FIORINI Dipartimento
di Fisica G.Occhialini dell'lNFN,
dell' Universita'di
Milano-Bicocca
e Sezione
diMilano
via Celoria 16, 20133 Milan (Italy)
E-mail:
[email protected]
After a brief introduction on the principle of cryogenic detectors and on the role they are going to play in fundamental physics, their specific possibilities in experiments on the search for double beta decay and for other rare events will be discussed. The large planned experiment CUORE and its already approved first step CUORICINO will then be described.
1
Introduction
The technique of cryogenic thermal detection has been initiated 16 years ago in Europe [1] and in the United States [2]. It consists in the development of new type of detectors, where the incoming particle loses almost entirely its energy in form of heat [3] . The main approach is the so called bolometric one, and is based on the use of a diamagnetic and dielectric crystal kept at very low temperature (tens of millikelvin). In this conditions its heat capacity is due to lattice vibrations only and, as such, is proportional to the cube of the ratio between the operating and Debye temperatures. As a consequence its heat capacity can become so low that even the tiny energy delivered in the crystals by a charged particle produces a substantial increase of temperature that can be revealed and measured by means of a suitable thermometer. These bolometers can in principle achieve energy resolutions larger by orders of magnitude than any other existing detector. In practice one has to overcome however severe problems like thermal instabilities and consequent variations of gain , electronic and mechanical noises etc. The collaboration and common activity among physicists involved in solid state, material science, nuclear and subnuclear physics, has however allowed [4] to achieve impressive resolutions both in the low and in the high energy regions. Microbolometer of masses of the order or lower than a milligram have reached resolutions around 5 eV at 6 keV, about twenty time better than in any other detector [5] . In the case of "macrobolometers" detectors with masses near to one kilogram have been constructed [6]. Despite their obviously larger heat capacity, their energy resolutions in the spectroscopy of high energy y rays is similar to those of germanium diodes. For a particle spectroscopy their energy resolution has been proved to be twice better than for any other type of detector , including solid state diodes. 274
275
Hybrid detectors were also developed. In view of experiments on dark matter with cryogenic detectors, like CDMS and EDELWEISS, American and European groups have implemented detectors where heat pulses are measured together with ionisation pulses. This allows discriminating particles with low ionisation, like nuclear recoils, from electrons or y rays. Simultaneous measurement of heat and scintillation was first performed in 1992 by the Milano group with the construction of a "scintillating bolometer" [7]. It is presently being actively studied by the Munich and Oxford groups, which were able to achieve discrimination between electrons and nuclear recoils with a rejection factor of 99.7 % for the former ones. One of the main advantages of cryogenic detectors , especially in the study of rare events which will be considered here , is their flexibility in the choice of the material as target of the interaction or source of the decay. 2. Possible "thermal" experiments on double beta decay The above mentioned flexibility of cryogenic detectors could be particularly useful in searches of double beta decay where a nucleus (A,Z) decays in the isobar (A,Z+2) with the emission of two neutrinos or , if lepton number is violated, without the emission of neutrinos [8-11]. As it is well known the evaluation of the rate of this decay depends strongly of the calculated values of the nuclear matrix elements , which are obviously subject to considerable uncertainties It is therefore essential , especially in experiments for neutrinoless double beta decay, to search for the decay of different type of nuclei. This is particularly true in the experiments based on the detector = source , or "calorimetric", approach where the detector itself contains the double beta decay candidate nucleus. When planning for a double beta decay experiment one should find a compromise between the "thermal" properties of a compound and its content of the double beta decay candidate nucleus. Some interesting "thermal " candidates for double beta decay searches are reported in Table I. Table I. Compound 4
*CaF2
7
*Ge
1,t,
Isotopic abundance of the candidate nucleus .0187 %
Transition energy
7.44 "
2038.7 "
4272 keV
CdW04
7.49 "
2804
"
1J0
TeO2
34
"
2528
"
lsu
NdF2
4.64
"
3368
"
276
I would like to add some comment on each of these "thermal' candidates:
This would be in principle an excellent candidate due to the large transition energy of 48Ca and to the good thermal properties of a calcium fluoride crystal. One problem could come from the fact that in this high energy region, while the background from y rays is negligible, the background from a particles can jeopardise the sensitivity of the experiment. On this point of view the Milano group has developed the above mentioned "scintillating" calcium fluoride bolometer where the heat pulse is measured simultaneously with the scintillation one [7]. Operating this detector with the two pulses in coincidence should strongly suppress the background of the poorly scintillating a particles. The difficulty in searches with this nucleus stays in its very low natural abundance, which makes the massive production of strongly enriched material very expensive. One could add that no compound of calcium has been so far obtained in gas form, which prevents enrichment by centrifugation.
I76Ge
I
Large crystals of germanium have already been operated in bolometers with a resolution approaching that of Ge diodes. It is not impossible that in the future the energy resolution of a thermal detector made with this material could become definitely superior than the resolution of a semiconductor. The cost of enrichment of large masses, even if feasible with centrifugation of Germanium fluoride, is however considerable.
CdW0 4 Large crystals of this material, an excellent scintillator, have been already operated in bolometers [12]. The heat vs. scintillation is therefore definitely possible. Cadmium can be enriched with centrifugues, even if this procedure is still rather expensive. 130
f^
277
The great advantage of this compound lies in the large isotopic aboundance of 130 Te and in the reasonably good thermal and mechanical properties of tellurium oxide. The transition energy is located in the valley of the 2615 peak of 208T1, one of the main y ray contaminants: reduction of background it therefore very promising.
™NHF; An excellent candidate for thermal detection of double beta decay would be 150Nd in a 150NddF2 crystal. Test measurements carried out by the Milano and Santa Barbara groups have however encountered considerable difficulty to cool to low temperatures these crystals, probably due to the paramagnetic properties of Neodimium. The only "thermal" experiments on double beta decay are those on Te carried out by the Milano group in the Gran Sasso Underground Laboratory [13] (fig.l). Various types of bolometers have been implemented with single crystals or arrays of crystals of TeC^- These crystals are of 3x3x6 cm3 with a mass of about 340 grams. The present setup is made by an array of 20 of these crystals and represents therefore the largest cryogenic detecting apparatus operating in the world.
Faraday cage
Helium re-liquefier
Pump's Fig. 1: The cryogenic set-up operating in Hall A of the Gran Sasso Laboratory
278
The temperature sensors are Neutron Transmutation Doped (NTD) Germanium thermistors thermally coupled to the crystals by means of 6 epoxy glue spots of ~ 0.6 mm diameter. In order to calibrate and stabilize the gain of the bolometer, which is very sensitive to the temperature, a heater consisting by a resistor of 100 to 200 k£2 was attached to each crystal. It has been specifically realized by depositing a heavily doped meander on a 1 mm3 silicon chip. The detector is shielded with the following materials: - Copper 2 cm - Roman lead (with less than 4 mBq/kg 210Pb activity) 1 cm - Low activity lead ( 16 ± 2 Bq/kg 210Pb activity) 10 cm ^ n r >Bq/kg „ / i , „ 210 210r>u „„*:.,;».A - vr„,™„i Normal 1leadi i( ~ i160 Pb activity) 10 cm The perfect reproducibility of the 20 detectors is shown by the overposition of the calibration spectra reported in Fig.2. The "background" spectrum collected so far corresponds to t 65,000 hours per crystal of effective running time, namely to about one kg x year of 130Te. The background in the region of neutrinoless DBD is of about 0.5 counts keV'1 kg'1 y"1.
%Wy 600
1000
1400
1800
2200
2600
Energy [keV] Fig.2: Sura of the twenty calibration spectra in the array operating in Hall A
279
An analysis based on the coincidence and anticoincidence of pulses from nearby detectors is being carried out. Preliminary results show that a considerable contribution to this background comes from degraded a particles from radioactive contamination on the surface of the detectors and on the materials immediately surrounding them. We are going to cure this contribution in the future runs by chemically and mechanically treating these surfaces. No evidence appears in the region of neutrinoless DBD. By applying the usual maximum likelihood procedure and taking into account our Monte Carlo evaluated 83.14% detection efficiency, we can set at 90% confidence level a lower limit of 1.3 x 1023 years on the half life for neutrinoless DBD of 130Te. The corresponding limit on the average antineutrino mass <mv> and on the presence of right right (r|) and left right (A) current parameters depends strongly on the evaluation of the relevant nuclear matrix elements. According to the theories reported in a recent review [10] and by [14] our result allows to extract upper limits on <mv> , (T|) and (A) in the ranges of 1-2.7 eV, 2-4 10 6 and 1-3 x 108 , respectively. On the basis of most of these calculations our result is the most restrictive one in direct experiments after those achieved with Germanium diodes [15,16]. No peak appears at the energy of 867 corresponding to neutrinoless DBD of 128 Te. By the usual likelihood method we can set an upper limit on the lifetime of 3.5 x 1022 years at 90% confidence level. The corresponding much less significant upper limit for the average neutrino mass range from to 10 to 22 eV. 3. "Thermal " experiments on direct interactions of WIMPS Thermal detectors, unlike semiconductors and scintillators, are expected to present a Quenching Factor, namely an efficiency to record the energy of a nuclear recoil as those produced by direct interaction WIMPS, near to 100 %. In the case of Te0 2 detectors this has been proved, and also shown to be practically independent on the recoil energy from 20 to 200 keV [17]. Thermal detectors provide in addition a large choice of target nuclei . We note that in natural tellurium 93% of isotopes are 0+. Three thermal experiments [4] specifically devoted to searches on direct interactions of WIMPS are presently running : CRESST in Gran Sasso, EDELWEISS in the Frejus Tunnel , and CDMS at Standford. None of them has a mass sufficient to test directly the annual variation of a possible signal due to WIMPS indicated by the DAM A group. The exclusion plot of the last of these experiments has already reached a sensitivity near to the one of the DAMA experiment. Our experiment MIBETA being run in Gran Sasso with tellurite crystals was not specifically designed to search for WIMPS, but already provides a resonably low counting rate in the low energy region (a few counts/keV/day at about 15 keV). In view of CUORICINO and especially of CUORE different thermal detectors can be considered. As an example the Milano group was able to operate in the Gran
280
Sasso Laboratory a bolometer where the absorber was a crystal of PbW0 4 made with Roman lead of 3x3x6 cm3, with a mass of ~ 440 grams. The resolution was similar as the one of Te0 2 detectors. 4. "Thermal " experiments on solar axions The possibility to search for solar axions by coherent Primakoff conversion on crystal planes has recently been suggested [18-20]. This conversion would lead to a subdiurnal modulation of the signal in a detector sensitive to low energy events. This modulation has already been searched for with a Ge diode. No evidence was found with upper limits on the coupling constant of ~ 3 x 10"9 in 1/M in GeV"1 . Thermal detectors , where the absorber provides well defined crystal planes, allow a wide choice of detector materials . The already tested crystals of PbW0 4 made with Roman lead look very promising for these searches [24}. 5. CUORE The potentialities of CUORE (for Cryogenic Underground Observatory for Rare Events) and some plans for CUORICINO have been already presented in many Conferences [22]. In the present design study CUORE would consist (Fig.3) in an array made by 17 towers. Each of them would contain 15 "floors" made by four bolometers, which in the present design study are cubic crystals of 5 cm side (Fig.4).
Fig.3: Scheme of CUORE. CUORICINO pratically represent a column of CUORE
281
This apparatus would represent a powerful tool to search for rare events like single and double beta decay, rare decays in alpha particles or in other complex nuclei and interactions of axions and Weakly Interacting Massive Particles (WIMPs). Excellent choice for double beta decay would be crystals made by Te0 2 , of 760 g mass, taking advantage of the large isotopic abundance of 130Te (34%) and its large (2528.8 keV) transition energy to i30Xe. This is however by no means the only one: for instance an appropriate choice could be to use the external layers of CUORE, made by Te0 2 crystals, in anticoincidence with the central ones, made by different materials particularly apt to WIMPs and Solar Axions searches. The construction of CUORE would however imply large funding ( around ten Millions dollars) and the special installation of a new large cryogenic facility in the Gran Sasso Laboratory. We have therefore presented to the Gran Sasso Scientific Committee and to the funding authorities a more limited set-up named CUORICMO (for "little CUORE"), which has been approved and funded
Fig. 4: A module of CUOMCINO (the sametobe used for CUORE)
282 6. CUORICINO CUORICINO is a large extension of the previously mentioned Mibeta experiment presently running in the Gran Sasso Underground laboratory. The CUORICINO setup (Fig.4) has been studied in order to assure a filling as complete as possible of the experimental space available in the dilution refrigerator presently operating for the Mibeta experiment in Hall A. CUORICINO is therefore the most massive cryogenic detector for searches on rare events compatible with the sizes of the existing cryogenic facilities operating in the Gran Sasso laboratory. In addition CUORICINO is not only a powerful self-consistent rare event experiment, but also a general test for the feasibility of CUORE. The CUORICINO structure corresponds in fact substantially to one of the 17 towers of which CUORE would consist. The only difference is that the CUORICINO tower will contain 14 planes instead of 15, as foreseen for CUORE, due to the limitation in height of the present dilution refrigerator. The CUORICINO array is a set of 56 cubic tellurite crystals of 5 cm side and 760 g mass, placed in a tower, which hosts four crystals per plane. Each plane consists of two copper frames (upper and lower) held by 4 cylindrical bars in copper. Small supports of PTFE hold tightly the crystals: their contraction at low temperature increases mechanical crystal stability , according to a principle well experimented in the presently operating 20 detector array. The gap between crystals in CUORICINO is of 0.6 cm only, assuring a very compact structure, which helps in study and reduction of the background by means of operation in coincidence and anticoincidence. The relevant dimensions of CUORICINO are reported in Table I. Table I. Dimensions in cm of the main CUORICINO components
STRUCTURE Copper Frame Crystals Small copper bars Plane Tower Long Copper bars
Side length " Diameter Side Length " Cross section
DIMENSIONS 12.2 Height • 5 0.4 Axes distance 12.2 Height 12.2 0.5 x0.2 it
0.8 5 8.0
Thick
0.4
6.6 92.4 92.4
The entire CUORICINO tower is placed within a closed cylindrical vessel in copper (the 50 mK radiation shield), surrounded by various coaxial cylinders in the dilution refrigerator. The CUORICINO detector tower will be topped with a Roman lead cylinder of 10 cm height and 15 cm diameter, which shields the intrinsic,
283 unavoidable radioactivity of some components of the dilution unit, such as the sintered silver contained in the discrete heat exchangers. The dilution refrigerator will be shielded externally by three layers of different types of Lead. The inside one, of 2 cm minimum thickness, will be made by Roman lead whose 210Pb activity has been measured to be below 4 mBq kg"![23]. The intermediate and external shields will be provided by layers of 10 cm minimum thickness made by modern lead of 16 ± 4 and 150 ± 20 Bq kg'1, respectively. The electrolytic copper of the thermal shields will provide an additional shield of 2 cm minimum thickness. A Faraday cage to eliminate electromagnetic interference will surround the refrigerator.
Fig.5 : The presently operating array for R&D of CUORICINO
For CUORICINO and CUORE we are planning the use of crystals of a mass 760 g, the largest than in any underground cryogenic experiments and more than twice
284
more massive than our presently operating absorbers. The temperature sensors will be Neutron Transmutation Doped (NTD) Ge thermistors. The optimization of the signal to noise ratio in bolometers with a so large absorber is a difficult task, and has required specific studies on the NTD Ge thermistors properties and on the thermal couplings among the various elements which constitute the detector. In order to perform this optimization, we have developed a thermal model able to predict with reasonable accuracy the pulse structure of the bolometers. On the basis of these considerations we have recently tested in hall C arrays of one, two and four crystals of 5x5x5 cm3 with a mass of 760 g, identical to those to be run in CUORICINO and CUORE. We have followed only partially the prescriptions of the model because a full realization of them would have implied the production of a new thermistor batch and substantial differences in the mounting procedure. We have used thermistors with the same doping as those of the 20 crystals experiment, but with dimensions 3x1x3 mm3 , namely 5 times larger. The most recent results have been obtained with an array of 4 bolometers with an original system of suspension with a spring (Fig.5). The excellent energy resolution of these crystals is shown in Fig.6. We would like to note that one of these crystals achieves an energy resolution [6] to high energy y r a y s similar to those of the best Germanium diodes. For 5407 keV a particles a FWHM resolution of ~ 3.2 keV was obtained, definitely superior to for any particle detector. As mentioned before our first choice, at least for CUORICINO, are crystals of Te0 2 . We believe however that options should be left open for different detecting absorbers. Other high Z materials should be conceived for coherent detection of WIMPs or solar axions as the already tested large crystals of PbW0 4 made with Roman lead pre-measured in order to ensure against the low energy contamination due to 2,0 Pb. FWHM @ 5407 keV : 3.2 ±.3 keV
3 30O O 20
Counts
, 1 .^np-n.^J 46
50
54
Energy
58
5420
[keV]
Fig.6: The peaks of 2,0Pb at 46 keV (FWHM resolution of 1 keV) and of resolution of 4.2 keV).
Energy 210
[keV]
Po at 5407 keV (FWHM
285
The South Carolina and Zaragoza groups are simulating the background counting rate for CUORICINO, with a modified EGS and by the GEANT program from CERN, respectively. A peculiar property of thermal detectors is that they are sensitive over their entire volume. As a consequence the problems connected with the study and the reduction of the background are entirely different than for detectors like the Ge diodes which are insensitive in a layer on the surface of about 0.5 mm thickness. As a consequence the a background from the surface radioactivity comes only from the internal contact, of much reduced surface area. This is clearly shown by a recent comparison with the background spectra obtained by the Heidelberg-Moscow experiment in a run of a 2.758 kg Ge diode totaling 660.22 days. The integral counting rate from 3 to 8.2 MeV was of 0.38 counts d' only. The only available spectra in this high energy region with thermal detectors are those obtained by the Milano group in the Mibeta experiment. The counting rate in the same region is about 15 times larger, per detector, even after subtraction of the peak due to internal radioactivity of 21° Po, a typical contaminant in Tellurium. One has in addition to take into account that the surface of the thermal detector (90 cm2) is about four times less that the entire surface of the above mentioned Ge diode. Since however the active internal surface of the Ge diode is of a few cm2 only, the effective surface activity on the two detectors is similar. The spectrum obtained in the Mibeta experiment shows no clear peak, indicative of an internal contamination, apart the above mentioned peak due to 210Po. There is therefore no indication of an internal contamination of Uranium and Thorium at a level of more than 0.1 pg/g. On the other side the spectrum becomes difficult to be understood. There are various bumps with a structure, which is definitely not gaussian with a large spread towards low energies. As a consequence the alpha energies are very difficult to be evaluated, even if we take the clear 5407 keV 210Po peak as a reference. A very tentative explanation could be the following: There is a surface contamination which "implants" the nuclei of the chains at a very shallow depth inside the crystal or in the material immediately surrounding it. If the decay occurs in the materials surrounding the detector the alpha particles (and the recoils!) reach the bolometers with a degraded energy. If the decays occur in the external layer of the bolometer an energy between the full transition energy and the recoil is recorded. Due to the very low range of the recoil the contributions from contamination in the external layer of the crystals and in the materials immediately surrounding them are difficult to be distinguished. This insidious background could be very dangerous both for neutrinoless double beta decay (degraded alpha particles) and for Dark Matters searches (degraded recoils). As a consequence special care should be devoted to the preparation of the crystals and of the frame of CUORICINO and obviously also of other thermal detectors. In particular the presence of radon should be absolutely avoided, in any stage of preparation of the bolometers.
286 7. The role on CUORICINO in underground physics The field and subjects of underground physics are evolving very rapidly, and searches on new phenomena especially in astroparticle physics are continuously proposed in addition to the "classical' ones (e.g. double beta decay, WIMPs interactions etc.). A peculiar advantage of a cryogenic detector like CUORICINO is the extensive choice of possible detecting nuclei 7.1. Double Beta Decay Due to the uncertainty in nuclear matrix calculations, searches for neutrinoless double beta decay should be extended to as many favorable nuclear candidates as possible. Particularly effective to search for this channel are the experiments where the detecting material contains the candidate nucleus for double beta decay. As pointed out before, a peculiar advantage of thermal detectors especially in the 4=detector approach [24], is their ample choice of detecting materials. We refer for the sake of simplicity only to the tested approach of a natural Te0 2 source. The expected sensitivity of CUORICINO under different hypothesis on resolution and background are shown in Table II together with the corresponding expectations for CUORE.
Table II: Expectations on the limits on half lifetime for neutrinoless double beta decay and effective neutrino mass (t is in year)
CUORICINO 56 crystals of 760g of natural Te0 2 FWHM= 10 keV
FWHM= 5 keV B(c/keV kg y)
<mv> (eV)
T,/2(y)
<mv> (eV)
T,„(y)
0.2
5.7 xlO x t"
0.30 x t"
4.06 xlO x t"
0.36 x f"4
0.1
8.1xl0 24 xt ,/2
0.25 x f"4
5.74xl0 24 xt" 2
0.30 x t4'4
0.01
2.6 xlO25 x t"2
0.14 xt" 4
1.81 xlO24 xt" 2
0.17 xt" 4
24
2
4
24
2
We would like to point out that the conservative expectations in the first row are based on an extrapolation of the present background based only on the lower surface/volume ratio of CUORICINO and the natural decrease of cosmogenic activity. We note that the excellent result of the Heidelberg-Moscow experiment has been obtained by not including in the final analysis the first 200 days of run and that a similar approach has been adopted by IGEX. We obviously expect
287
considerable improvements in the future, as those indicated in the two last rows of Table II. The possibility to fabricate CUORICINO (if not CUORE) with enriched crystals can obviously be considered and has been recently discussed with the Kurchatov Institute in Moscow, where a considerable amount of 128 Te0 2 and 130 TeO2 has been previously produced for us with clean centrifugues. Due to the large natural abundance of 130Te, enrichment in this isotope is much less expensive than e.g. for 76Ge (an order of magnitude less per unit isotope mass). The recent measurements on 130TeO2 by the Milano group, reported previously, indicate that some impurity is introduced in tellurite by the enrichment procedure. As a consequence this procedure should be further refined if the sensitivity on the lifetime for neutrinoless double beta decay has to be improved by a factor of three with an "enriched CUORICINO". 7.2 Searches on direct interactions of WIMPs in CUORICINO The CUORICINO detector does not exhibit the ionization vs. heat background rejection property of CDMS and EDELWEISS, but we believe that it is indeed complementary to these experiments due to the different target nucleus and much larger mass. We believe however that CUORICINO will be particularly sensitive to the annual modulation analysis due to its large mass and to its large quenching factor, which has been proved to be constantly near one at various recoiling energies [17]. For a correct evaluation of the sensitivities of CUORICINO in the seasonal modulation analysis, a better knowledge of long time stability and of systematic effects in the 20 detector array is essential. The unique versatility of cryogenic detectors could allow in the next years to search for WIMPs interactions in various nuclei, particularly in those where a possible effect would be found. Many other materials are of obvious interest for searches of Dark Matter with CUORICINO. Examples of good "thermal" candidates are A1 2 0 3 , CaF2, LiF etc., which present, especially in the first case, a large Debye temperature. We have operated crystals of these materials, but we prefer to consider for this preliminary version of CUORE only materials with high Z nuclei. They are complementary to those already adopted in the presently running and planned experiments and the large N2 enhancement factor is very promising to search for vector interactions of WIMPs in the high mass region. Further work has therefore to be done with the above mentioned crystals of PbW0 4 made with Roman lead. 7.3. Searches for Axions from the Sun CUORICINO seems to us an ideal detector to search for coherent interactions of axions coming from the Sun and from other cosmic sources. The Te0 2 crystals, which we are presently using, have a tetragonal structure and (100)(010)(001) orientation. In the crystal cell the (100) side is equal to the
288
(010) one and different from (001). The crystal is normally grown along the (100) side. As a consequence different coherent interactions would take place in correlation with the position of the (100)(010) and (100)(001) planes relative to the Sun and therefore the signal modulation would be repeated twice. After discussion with the Shanghai Qinghua Nonmetal Co. we plan for CUORICINO the <110> orientation since this option, equivalent for the sensitivity in the axion search, is easier and less expensive. We have calculated the expected gaY|. bound achievable by the CUORICINO detector due to the Primakoff coherent conversion of solar axions into photons via Bragg scattering. As shown in [20], the values used for input parameters are: a
< alim
k(
ckeV'kg'day"
1
kg years l/8 ) xlO-'GeV 1 M T
Where k depends on the crystal structure and material, as well as on the experimental threshold and resolution. For CUORICINO k has been calculated to be 3.0 for an energy resolution of 2 keV. The computation of this expression with the above values yields the limits reported in the fourth column of Table III for CUORICINO. As it can be seen there is only a very weak dependence of the limit on gayy from the energy resolution. Table III:Limits on g ^ for CUORICINO
Mass (kg)
Resolution
Background
Limit on gayy 1
(keV)
(counts keV'kg'd" )
(2 years) GeV"1
42
2
0.5
1.6x10"*
42
2
0.1
1.3x10-"
42
2
0.05
1.2xl0" y
42
1
o.5
1.5x10"
42
1
0.1
1.2 x 10"
42
1
0.05
1.1x10-"
289 6.4 Searches in Low Energy Nuclear Physics with CUORICINO We believe that too little effort has been devoted so far to the application of cryogenic detectors to the field, which is rather improperly defined as "low energy nuclear physics". A list of subjects of great interest, even if limited to underground experiments, would be out of the aims of this paper (and also outside our specific competencies). As an example concerning rare processes we would like to remind the interest in naturally occurring decays involving electron capture and emission of electrons, positrons , a particles and complex nuclei [24] . Some results have been already obtained with thermal detectors: the first correct evidence of electron capture of m Te [25] and the first measurement of the spectrum of "3Cd [12]. With CdWO„ crystals one could search for the existence of the a decaying nucleus 180W etc.
&. Acknowledgment I would like to express my gratitude to the Milan cryogenic detectors group for the exciting work carried out together since more than 15 years. This work has been funded in part by the TMR CEE program, under Contract FMRX-CT98-0167
References 1. 2 3 4 5 6 7 8 9 10 11
Fiorini E. and Niinikoski T., Nucl.lnstrum. and Meth. 224 (1984) 83 McCammon D. et al, J.Appl.Phys. 58:1263 (1984) Twerenbold D., Rep.Prog. Phys.59 (1996) 349 and Booth N., Cabrera B. and Fiorini E., Ann. Rev. of Nucl.Sci. 46 (1996) 471 For recent results on cryogenic detectors see, Proc. of the VIII Int. Workshop on Low Temperature Detectors Dalfsen (Netherlands) August 15-21 Nuclear Instrum. and Meth (in the press) Alessandrello A. et al, Phys.Rev. Lett. 82 (1999) 513 Alessandrello A. et al, NIM A 440 (2000) 397 Alessandrello A. et al, Phys.Lett.B 420 (1998) 109 and previous references Morales A., Nucl. Phys.B (Proc.Suppl) 77 (1999) 235 Fiorini E.: Double Beta Decay, Invited paper to Neutrino Telescope (Venice, February 23-27, 1999) Ed. By M. Baldo Ceolin (in the press) Suhonen J. and Civitarese O. , Physics Reports 300 (1998) 123 Tretyak V.I. and .Zdesenko V, Atomic Data and Nuclear Data Tables 61 (1995) 43 and references therein
290
12 13 14 15 16 17 18 19 20 21 22
23 24 25 26
Alessandrello A. et al, Nucl.Phys. B (Proc.Suppl.) 35 (1994) 394 Alessandrello Yu et al, Phys.Lett. B 420 (1998) 109 and " New Experimental Results on Double Beta Decay of 130Te" submitted to Phys.Lett.B Barbero C , Krmpotic J.M., and Tadic D., Nucl.Phys. A 650 (1999) 485 Aalseth C.E. et al; Phys.Rev.C 59 (1999) 1 Baudis L. et al, Phys. Rev. Lett. 83 ( 1999) 41 Alessandrello A. et al, Phys.Lett.B 408 (1997) 465 Creswick R.J. et al, Phys.Lett.B 427 (1998) 235 Avignone F.T. et al, Phys. Rev. Lett. 81 (1998) 5068 Cembrian S. et al: Prospects of Solar Axions Seaches with Crystal Detectors, to be pubblished in Astroparticle Physics Private communication by F. Avignone Fiorini E., Phys.Rep. 307 (1998) 309; Report by F. Avignone, Proc. of the Sixth Intern. Symposium on Particles, Strings and Cosmology (PASCOS), Boston MA, USA Wordl Sci. (in Press) ; Report by A. Giuliani, Int.Symp. on Lepton and Baryon Number Violation, April 1998, Trento (Italy); O. Cremonesi, Nucl. Phys.B (Proc.Suppl)77 (1999) 369; Report by E. Previtali to the 2nd Int. Conf. On Dark Matter in Astro and Particle Physics (DARK 98), Heidelberg (Germany), July 1998 ; Report by F. Avignone, International workshop on Dark Matter (DM-98), Boxton , U.K. Sept. 1998 , Wordl Sci. (in press) Alessandrello A. et al, Nucl. Instrum. and Meth. B 142 (1998) 163 Dell' Antonio G.F. and Fiorini E., Suppl.Nuovo Cim. 17 (1960) 132 See for instance the recent reviews in "Nuclear Decay Modes", ed. by D.N. Poenaru, Inst.of Physics Pu. (Bristol-Philadelphia), 1996 Alessandrello A. et al, Phys. Rev.Lett. 77 (1996) 3319
W I M P SEARCHES AT C A N F R A N C W I T H GERMANIUM DETECTORS ANGEL MORALES Laboratory of Nuclear and High Energy Physics University of Zaragoza. 5009 Zaragoza. Spain E-mail: amorales @posta. unizar. es An overview of the searches for Weak Interacting Massive Particles (WIMPs) through their scattering off Germanium nuclei carried out in the Canfranc Tunnel Astroparticle Laboratory (at 2450 metres of water equivalent (m.w.e.)) in a collaboration between the Universities of South Carolina and Zaragoza is given. The main experimental results are sketched both for natural abundance (COSME) and 7 6 G e enriched (IGEX) Germanium detector experiments are summarized and a briefing on the GEDEON project is also presented.
1
Historical N o t e
I first met Frank Avignone ("Paco" for his Spanish friends) in 1986 during a conference in Heidelberg. We had known each other's names for a few years through our publications, which happened to be in the same field of science and interest. At that time my group of the Zaragoza University was preparing at the Canfranc tunnel, in the Spanish Pyrenees, a double beta decay experiment of 76 Ge to excited states of 76 Se, looking for coincidences 2/3/7, trying to settle a disturbing result obtained in a previous similar experiment we had carried out in Frejus (France) in a collaboration with CENBG Bordeaux. We needed a good germanium detector and Frank put, very kindly, at our disposal his expertise in the design of such spectrometer, which was made by Princeton Gamma-Tech (New Jersey, USA) under his supervision in 1987. The detector—which was nicknamed PACO for obvious reasons—is still working perfectly in Canfranc. The first joint scientific research that Frank and I carried out together was, however, in the field of particle dark matter searches in the old Canfranc facility, with a small germanium detector—COSME—which Frank brought from USC-PNNL in 1988-1990. He had pioneered in the field of WIMP direct searches with germanium detectors making the first published contribution in this field (1987), as a by-product of a double beta decay investigation, but this time we were to carry out the first dedicated WIMP experiment with a Ge detector. I picked up Frank at the airport of Barcelona where he landed with the detector COSME and we drove directly to Canfranc to put the detector underground. The detector COSME was specifically designed for very low energy thresholds and good energy resolution and, al291
292
though the background was not its best performance, it played a leading role in the exclusion of low mass WIMPs as dark matter candidates (COSME-1). This joint venture was only the first of a long series of fruitful collaborations that we have had—and still have—Frank and I along more than ten years, and the starting of an enjoying scientific relationship which has turned into a true friendship extended also to our families. After the COSME-1 project, the IGEX (International Germanium Experiment) Collaboration enters into the stage in 1990, also in Canfranc. IGEX has been one of our longer lasting joint ventures in double beta decay, and is taking now a new look as a WIMP search. Other common undertakings have followed, in summary, making Frank a frequent visitor to Zaragoza-Canfranc, where he is at home and as in his own laboratory. This note explains why I have chosen the COSME and IGEX dark matter topic to dedicate it to Frank in his Jubilee Conference. 2
Introduction
There is a substantial evidence1 that most of the matter of the universe is dark and a compelling motivation to believe that it consists mainly of nonbaryonic objects. From the cosmological point of view, two big categories of non- baryonic dark matter have been proposed: cold (WIMPs, axions) and hot (light neutrinos) dark matter according to whether they were slow or fast moving at the time of galaxy formation. A general consensus coming out from a variety of observations and well founded theoretical arguments is that a large portion of non-baryonic cold dark matter is needed in the universe. This form of dark matter could be filling the galactic halo in enough amount to attempt its detection 2 . The indirect detection of WIMPs proceeds currently through two main experimental lines: either by looking in the space for positrons, antiprotons, antideuterons or other antinuclei produced by the WIMPs annihilation in the halo, or by searching in large underground detectors or underwater neutrino telescopes for upward-going muons produced by the energetic neutrinos emerging as final products of the WIMPs annihilation in celestial bodies (Sun, Earth...) The direct detection of WIMPs relies in the measurement of their elastic scattering off the target nuclei in a suitable detector 3,4 . Pervading the galactic halos, slow moving (~ 300 km/s), and heavy (10 ~ 103 GeV) WIMPs could, for intance, make a Ge nucleus recoil with a few keV (only about 1/4 of this energy is visible in the detector), at a rate which depends of the type of WIMP and interaction. Because of the low interaction rate and the small
293
energy deposition, the direct search for particle dark matter through their scattering by nuclear targets requires ultralow background detectors of a very low energy threshold. Moreover, the (almost) exponentially decreasing shape of the predicted nuclear recoil spectrum mimics that of the low energy background registered by the detector. All these features together make the WIMP detection a formidable experimental challenge. Customarily, one compares the predicted event rate with the observed spectrum. If the former turns out to be larger than the measured one, the particle under consideration can be ruled out as a dark matter component. Such non-appearance of the WIMP is expressed as a contour line
Strategies for W I M P detection
The rarity and smallness of the WIMP signal dictate the experimental strategies for its detection: Reduce first the background, controlling the radiopurity of the detector, components, shielding and environment. The best radiopurity has been obtained in the Ge experiments. In the case of the Nal scintillators, the backgrounds are still one or two orders of magnitude worse than in Ge. The next step is to use discrimination mechanisms to distinguish electron recoils (tracers of the background) from nuclear recoils (originated by WIMPs or neutrons). Various techniques have been applied for such purpose from which two of them will be mentioned: a statistical pulse shape analysis (PSD) based on the different timing behaviour of both types of pulses and an identification (on an event-by-event basis) of the nuclear recoils by measuring at the
294 same time two different mechanisms of energy deposition, like the ionization and heat, capitalizing the fact that for a given deposited energy (measured as phonons) the recoiling nucleus ionizes less than the electrons. The other obvious strategy is to make detectors of very low energy threshold and high efficiency to see most of the signal spectrum, not just the tail. That is the case of bolometer experiments. Finally, one should search for distinctive signatures of the WIMP, to prove that you are seeing a WIMP. Suggested identifying labels are: an annual modulation of the signal rate and energy spectrum (of a few percent) due to the seasonal June-December variation in the relative velocity Earth-halo and a forward-backward asymmetry in the direction of the nuclear recoil due to the Earth motion through the halo. The annual modulation signature has been already explored. Pioneering searches for WIMP annual modulation signals were carried out in Canfranc, Kamioka and Gran Sasso. The DAMA experiment at Gran Sasso, using a set of Nal scintillators has reported an annual modulation effect interpreted as due to a WIMP of about 60 GeV of mass and scalar cross-section on protons of <rp = 7 x 10~ 6 picobarns. The universities of Zaragoza and South Carolina have been collaborating for a decade in a search for WIMP dark matter in the underground facility of the University of Zaragoza in the Canfranc Tunnel (Spain). The techniques which both groups were using in double beta decay searches with Ge detectors lead in a natural way to attempting to detect directly this non-baryonic dark matter through the nuclear recoil in WIMP-Ge nucleus scattering. In the following, we will review the status, results and prospects of these experiments. 4
Germanium Experiments 3 , 4
The high radiopurity and low background achieved in Germanium detectors, their fair low energy threshold, their reasonable Quenching Factor (about 25%) (nuclear recoil ionization efficiency relative to that of electrons of the same kinetic energy, or ionization yield) and other nuclear merits make Germanium a good option to search for WIMPs with detectors and techniques fully mastered. The first detectors applied to WIMP direct searches (as early as in 1987) were, in fact, Ge diodes, as by-products of 2/3-decay dedicated experiments. The exclusion plots c(m) for spin-independent couplings obtained by former Ge experiments [PNNL/USC/Zaragoza (TWIN and COSME-1), UCSB, CALT/NEU/PSI, H/M] are still remarkable and have not been surpassed till recently by Nal experiments (DAMA) using statistical PSD. A Ge detector of natural abundance (COSME) of the U. Zaragoza / U. S.Carolina / PNNL Collaboration and another one (RG-II) made of enriched
295 76
Ge of the IGEX Collaboration are being used in Canfranc to search for WIMPs interacting coherently with the Ge nuclei of the detectors. The COSME detector was fabricated at Princeton Gamma-Tech, Inc. in Princeton, New Jersey, using naturally abundant germanium. The refinement of newly-mined germanium ore to finished metal for this detector was expedited to minimize production of cosmogenic 68 Ge. The COSME detector 6,7 is a p-type coaxial hyperpure natural germanium crystal with a mass of 254 g and an active mass of 234 g which has a long term resolution of 0.43 keV full width at half maximum (FWHM) at the 10.37 keV gallium X-ray. In its first installation (at 675 m.w.e) the detector was placed within a shielding of 10 cm of 2000 yr. old (Roman) lead (inner layer) plus 20 cm of low activity lead (about 70 yr old). A 3 mm thick PVC box sealed with silicone closes the lead shielding to purge the radon gas. The PVC box is covered by 1 mm of cadmium and 20 cm of paraffin and borated polyethylene. All the shielding and mounting is supported by 10 cm of vibrational and acoustic insulator sandwiched within two layers of 10 cm of wood mounted on a floor of concrete (20 cm). It was first operated 6,7 in the former Canfranc underground facility at 675 m.w.e. in a small gallery of the Canfranc (shutdown) railway tunnel in the Spanish Pyrenees. In that set-up, the energy threshold was Ethr = 1 . 6 keV and the background at threshold was about 10 counts k e V - 1 k g - 1 d a y - 1 . After about 500 days of data taking (referred as COSME-T experiment) no signal originated by WIMP appeared and, as stated before, an exclusion plot in the cross-section versus WIMP masses plane was derived, assuming WIMP-matter spin-independent couplings. To derive such exclusion plots cr(m), the predicted signal was required to be not larger than the 90% C.L. upper limit of the poissonian background counts recorded in a chosen energy bin. In the derivation of the interaction rate signal, the WIMPs were supposed to form an isotropic, isothermal non-rotating halo of density p = 0.3 GeV c m - 3 , and have a maxwellian velocity distribution with v r m s = 270 km s _ 1 , (with an upper cut corresponding to an escape velocity of 650 km s _ 1 ), and a relative earth-halo velicity of v r = 230 km s _ 1 . In the COSME-1 data, the energy range chosen to derive the exclusion plot was from 1.6 to 8 keV where the background was, approximately, 5 counts/(keV kg day). In spite of such modest figure, the results after 130 kg day of exposure improved the exclusion plots at low masses (from 9 to 20 GeV) obtained with other Ge experiments because of the low threshold energy of COSME-1. The COSME detector has been reinstalled, in better background conditions (at 2450 m.w.e.) inside a Marinelli beaker in Roman lead (COSME-2) in the same shielding (described later) as the three 2.1 kg enriched germanium detectors of IGEX (the International Germanium Experiment on Dou-
296
Figure 1. Low energy background spectrum of COSME-2.
ble Beta Decay) 8 ' 9 . In its new installation, COSME-2 has an energy threshold of Ethr = 2.5 keV, and an energy resolution of T(FWHM) = 0.4 keV at 10 keV. The average background rate, in 311 days of exposure (Mt=72.8 kg day) is 0.6 c/(keV kg day) from 2 to 15 keV and 0.3 c/(keV kg day) from 15 to 30 keV which is significantly better than in COSME-1. The COSME-2 spectrum is shown in Fig. 1 and the numerical data given in Table 1. The IGEX experiment 8 ' 9 , optimized for detecting 76 Ge double-beta decay, has been described in detail elsewhere. A recent upgrade of the IGEX detectors allow them to also be used in the search for WIMPs interacting coherently with germanium nuclei. The IGEX detectors were fabricated at Oxford Instruments, Inc., in Oak Ridge, Tennessee. Russian Ge0 2 powder, isotopically enriched to 86% 76 Ge, was purified, reduced to metal, and zone refined to ~ 10 13 p-type donor impurities per cubic centimeter by Eagle Picher, Inc., in Quapaw, Oklahoma. The metal was then transported to Oxford Instruments by surface in order to minimize activation by cosmic ray neutrons, where it was further zone refined, grown into crystals, and fabricated into detectors. All of the cryostat parts were electroformed using a high purity OFHC copper/CuS0 4 /H2S04 plating system. The solution was continuously filtered to eliminate copper oxide, which causes porosity in the copper. A Ba(OH)2 solution was added to precipitate BaS0 4 , which is also collected on the filter. Radium in the bath exchanges with the barium on the filter, thus minimizing radium contamination in the cryostat parts. The CUSO4 crystals were purified
297 E (keV) 2.5
3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5 16.5 17.5
counts 1098
76 62 57 56 37 51 45 56 46 34 30 34 38 30 19
E (keV) 18.5 19.5 20.5 21.5 22.5 23.5 24.5 25.5 26.5 27.5 28.5 29.5 30.5 31.5 32.5 33.5
counts
25 19 28 20 18 17 22 20 23 28 19 19 19 15 27 22
E (keV) 34.5 35.5 36.5 37.5 38.5 39.5 40.5 41.5 42.5 43.5 44.5 45.5 46.5 47.5 48.5 49.5
counts
21 13 20 12 14 20 13 17 17 17 20 18 48 20 17 19
Table 1. Low-energy data from the COSME-2 detector (Mt = 73 kg-d).
of thorium by multiple recrystallization. The IGEX detector used for dark matter searches, designated RG-II, has a mass of ~ 2.2 kg. The active mass of this detector, ~ 2.0 kg, was measured with a collimated source of 152 Eu in the Canfranc Laboratory and is in agreement with the Oxford Instruments efficiency measurements. The full-width at half-maximum (FWHM) energy resolution of RG-II was 2.37 keV at the 1333-keV line of 60 Co. Energy calibration and resolution measurements were made every 7-10 days using the lines of 22 Na and 60 Co. Calibration for the low energy region was extrapolated using the X-ray lines of Pb. For each detector, the first-stage field-effect transistor (FET) is mounted on a Teflon block a few centimeters from the center contact of the germanium crystal. The protective cover of the FET and the glass shell of the feedback resistor have been removed to reduce radioactive background. This first-stage assembly is mounted behind a 2.5-cm-thick cylinder of archaeological lead to further reduce background. Further stages of preamplification are located at the back of the cryostat cross arm, approximately 70 cm from the crystal. The IGEX detectors have preamplifiers modified for the pulse-shape analysis used in the double-beta decay searches. The detectors shielding is as follows, from inside to outside. The inner-
298 most shield consists of 2.5 tons of 2000-year-old archaeological lead forming a 60-cm cube and having < 9 m B q / k g of 2 1 0 P b ( 2 1 0 B i ) , < 0.2 m B q / k g of 2 3 8 U , and < 0.3 m B q / k g of 2 3 2 T h . T h e detectors fit into precision-machined holes in this central core, which minimizes the empty space around the detectors available t o radon. Nitrogen gas, a t a rate of 140 1/hour, evaporating from liquid nitrogen, is forced into the detector chambers to create a positive pressure and further minimize radon intrusion. T h e archaeological lead block is centered in a 1-m cube of 70-year-old low-activity l e a d ( ~ 10 tons) having ~ 30 B q / k g of 2 1 0 P b . A minimum of 15 cm of archaeological lead separates the detectors from the outer lead shield. A 2-mm-thick cadmium sheet surrounds the main lead shield, and two layers of plastic seal this central assembly against radon intrusion. A cosmic muon veto covers the t o p and sides of the central core, except where the detector Dewars are located. T h e veto consists of B I C R O N BC-408 plastic scintillators 5.08 cm x 50.8 cm x 101.6 cm with surfaces finished by diamond mill to optimize internal reflection. BC-800 (UVT) light guides on the ends taper to 5.08 cm in diameter over a length of 50.8 cm and are coupled to H a m a m a t s u R329 photomultiplier tubes. T h e anticoincidence veto signal is obtained from the logical O R of all photomultiplier t u b e discriminator o u t p u t s . An external polyethylene neutron moderator 20 cm thick (1.5 tons) completes the shield. T h e entire shield is supported by an iron structure resting on noise-isolation blocks. T h e experiment is located in a room isolated from the rest of the laboratory and has an overburden of 2450 m.w.e., which reduces the measured muon flux to 2 x 1 0 _ 7 c m ^ 2 s _ 1 . T h e d a t a acquisition system for the low-energy region used in dark m a t t e r searches (referred t o as IGEX-DM) is based on s t a n d a r d NIM electronics and is independent from t h a t used for double-beta decay searches (IGEX-2/3). It has been implemented by splitting the normal preamplifier o u t p u t pulses of each detector and routing t h e m through two Canberra 2020 amplifiers having different shaping times enabling noise rejection 6 . These amplifier o u t p u t s are converted using 200 MHz Wilkinson-type Canberra analog-to-digital converters, controlled by a P C through parallel interfaces. For each event, the arrival time (with an accuracy of 100 /JS), the elapsed time since the last veto event (with an accuracy of 20 fia), and the energy from each ADC are recorded. T h e detector IGEX-RG-II features an energy threshold of 4 keV and an energy resolution of 0.8 keV at the 75 keV P b x-ray line. T h e background rate recorded was ~ 0.3 c/(keV-kg-day) between 4-10 keV, ~ 0.07 c/(keV-kg-day) between 10-20 keV, and ~ 0.05 c/(keV-kg-day) between 20-40 keV. Fig. 2 shows the RG-II 30-day spectrum; the numerical d a t a are given in Table 2.
299 OJi 0.7
nn M
u
n l-
n
i^ n ^ ^nI ^ V ^ J ^ M Un i 20
30
Energy (keV) Figure 2. Low-energy spectrum of the IGEX RG-II detector (Mt = 60 kg-d).
5
Results
The IGEX-DM results obtained correspond to 30 days of analyzed data (Mt=60 kg-days) from IGEX detector RG-II (Table 2 and Fig. 2). The corresponding
300
E (keV) 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5 16.5 17.5 18.5
counts 44 23 29 17 8 14 12 14 10 5 3 3 10 3 5
E (keV) 19.5 20.5 21.5 22.5 23.5 24.5 25.5 26.5 27.5 28.5 29.5 30.5 31.5 32.5 33.5
counts 9 5 3 4 4 0 2 1 2 3 1 2 2 3 2
E (keV) 34.5 35.5 36.5 37.5 38.5 39.5 40.5 41.5 42.5 43.5 44.5 45.5 46.5 47.5 48.5
counts 3 2 1 4 2 3 1 5 0 3 4 3 10 1 2
Table 2. Low-energy data from the IGEX RG-II detector (Mt = 60 kg-d).
sections above 1.3xl0~ 8 nb for masses corresponding to the 50 GeV DAMA region 12 . Also shown is the combined germanium contour, including the last Heidelberg-Moscow data 1 3 (recalculated from the original energy spectra with the same set of hypotheses and parameters), the DAMA experiment contour plot derived from Pulse Shape Discriminated spectra 14 , and the DAMA region corresponding to their reported annual modulation effect12. The IGEX-DM exclusion contour improves significantly on that of other germanium experiments for masses corresponding to that of the neutralino tentatively assigned to the DAMA modulation effect12 and results from using only data without background subtraction. The COSME-2 exclusion contour also slightly improves the Ge-combined plot for masses between 20 and 40 GeV. Data collection of IGEX-RG-II is currently in progress with improved background below 20 keV. Based on present IGEX-DM performance and reduction of the background to ~ 0.1 c/(keV-kg-day) between 4-10 keV, the complete DAMA region ( m = 5 2 ^ ° GeV, CTP=(7.2i°;4)xl0-9 nb) could be tested after an exposure of 1 kg-year. There exist some projects of USC and UZ with germanium detectors to pursue the search for WIMPs with improved sensitivity and larger detector masses to look for annual modulation of dark matter signals. They are essentially based in the use of segmented Ge detectors in various options and
301 io- 5
Itt I
i
\
£10"7
IO"8
10
1
5
10
50 100 M w (GeV)
5001000
Figure 3. IGEX-DM exclusion plot for spin-independent interaction obtained in this work (thick solid line). Results obtained in other Germanium experiments are also shown: Canfranc COSME-1 data 7 (dot-dashed line), recent COSME-2 d a t a 3 ' 1 0 (thick dashed line), and the previous Ge-combined bound (thin dashed line) —including the last Heidelberg-Moscow d a t a 1 3 . The result of the DAMA NaI-0 experiment 1 4 (thin solid line) is also shown. The "triangle" area corresponds to the (3
geometries. One is, for instance GEDEON (Germanium Diodes in One Cryostat, MOZA Collaboration, Canfranc) 4 ' 15 . GEDEON is a detector project which uses the conventional, well-known, Ge-diode technology to look for cold dark matter. The GEDEON single cell is a cylindrical cryostat in electroformed copper (dimensions 20 cm diameter x 32 cm long) made with IGEX technology, hosting 28 (natural abundance) germanium crystals which share the same common copper cryostat (0.5 mm thick). A schematic view of the GEDEON single cell is shown in Fig. 4. The Ge crystals, of 860 g each, are arranged in four plates suspended from copper rods. The cell is embedded in a precision-machined hole in a Roman lead block providing a shield of 20 cm, surrounded by another lead shielding 20 cm thick. A cosmic veto and a neutron shield complete the shielding. The preliminary MC estimated background in the 1 ~ 50 keV region ranges from 2 x 10~ 2 to 2 x 10~ 3 c/keV kg day, according to the level of radioimpurities included as input. The radiopurity assays have been carried out in the Canfranc Laboratory for the lead and copper components of the shielding. The background goal of GEDEON, below 100 keV, is < 10~ 3 c/keV kg day and this value has been used to calculate anticipated o~(m) exclusion
302
Figure 4. GEANT-generated picture of the GEDEON single cell (four planes of six 860 g Ge diodes). The single cryostat hosting the diodes is not shwown for clarity.
plots. The expected threshold is Ethr = 2 keV and the energy resolution in the low energy region is T ~ 1 keV. The MC estimated background of the GEDEON unit cell (28 crystals) is given in Fig. 5, whereas the predicted o-(m) plots are given in Fig. 6 which also shows the region of the neutralino parameter space which GEDEON attempts to explore. The study design of this project is being performed, including also a second phase with four cryostats (112 detectors and a total mass of 92 kg of Ge).
Figure 5. Montecarlo estimaged background of the GEDEON detector project (energy interval 0—50 keV).
303
io- 4 io- 6 S3 t IO"8 ~b IO" 1 0
10
f
—g--Yp50 ibo M w (GeV)
5661000
Figure 6. Projections for GEDEON (thick dashed line) assuming MT = 24 kg y, E t h = 2 keV, F W H M = 1 keV and b = 2 x 10~ 3 c/keV/kg/day for GEDEON. The regions enclosed by thin solid lines are the Ge-combined and DAMA excluded regions. The DAMA modulation experiment allowed region (thin dots line) and MSSM region (thin dashed line) are also shown.
Acknowledgments The unpublished data presented here resulted from collaborative research with the COSME (F.T. Avignone, R.L. Brodzinski and H.S. Miley) and IGEX (C.E. Aalseth, F.T. Avignone, R.L. Brodzinski, S. Cebrian, E. Garcia, D. Gonzalez, W.K. Hensley, I.G. Irastorza, I.V. Kirpichnikov, A.A. Klimenko, H.S. Miley, J. Morales, A. Ortiz de Solorzano, S.B. Osetrov, V.S. Pogosov, J. Puimedon, J.H. Reeves, M.L. Sarsa, S. Scopel, A.A. Smolnikov, A.G. Tamanyan, A.A. Vasenko, S.I. Vasiliev, J.A. Villar) members to whom I am deeply indebted. Results will be published in due course. I wish to thank especially S. Cebrian, I.G. Irastorza and S. Scopel for their invaluable collaboration in the making of the exclusion plots. The present work was partially supported by the CICYT under grant number AEN99-1033. References 1. G. Jungman, M. Kamionkowski and K. Griest, Phys. Rep. 267 (1996) 195 2. A. Morales, "Dark Matter and its Detection", Summary Talk given at the NUPECC Workshop on Present and Future of Neutrino Physics, Frascati, NUPECC Report in Highlights and Opportunities in Nuclear Physics, Ed.
304 by J. Vervier et al., December 1997. ( a s t r o - p h / 9 8 1 0 3 4 l ) . 3. For a recent survey of the s t a t u s of W I M P searches, see for instance, A. Morales, "Direct Detection of W I M P Dark M a t t e r " ( a s t r o - p h / 9 9 1 2 5 5 4 ) . Review Talk at the T A U P 99 Workshop, College de France, Paris. Nucl. Phys. B (Proc. Suppl.) 87 (2000) 477. T h e experiments quoted in the text are fully referred to in our review papers of Refs. 3 ' 4 4. A. Morales "Selected Projects in Direct Detection of Dark M a t t e r " , P r o c . Neutrino Telescopes Workshop, Venice, February 1999. Ed. M. BaldoCeolin, p. 249 and F . T . Avignone and A. Morales, P r o c . Int. Conference on Neutrino Physics and Astrophysics. Helsinki, June 2000, ed. K. Enkvist et al. in World Scientific P u b . (1997) p. 413. 5. A.K. Drukier et a l , Phys. Rev. D 3 3 (1986) 3495. 6. J. Morales et a l , Nucl. lustrum. & Meth. A 3 2 1 (1992) 410 7. E. Garcia et a l , Nucl. Phys. B (Proc. Suppl.) 2 8 A (1992) 286 and Phys. Rev. D 5 1 (1995) 1458. 8. C. Aalseth et al., Phys. Rev. C 5 9 (1999) 2108 9. D. Gonzalez et al., P r o c . T A U P 99 Workshop, College de France, Paris, Nucl. Phys. B (Proc. Suppl.) 8 7 (2000) 278. 10. S. Cebrian et al., New Journal of Physics 2 (2000) (http://www.nj p.org). 11. J. Engel, Phys. Lett. B 2 6 4 (1991) 114. 12. R. Bernabei et al., Phys. Lett. B 4 5 0 448 (Preprint R0M2F/2000/01 J a n u a r y 2000). 13. L. Baudis et al., Phys. Rev. D 5 9 (1999) 022001. 14. R. Bernabei et al., Phys. Lett. B 3 7 9 (1996) 299. 15. A. Morales et al., G E D E O N , a project for W I M P searches with a set of n a t u r a l abundance Ge diodes in a single cryostat. Preliminary Study for Submission to C I C Y T (Spain), J a n u a r y 1999.
S E A R C H I N G FOR S U P E R S Y M M E T R I C D A R K MATTER. T H E M O D U L A T I O N EFFECT D U E TO CAUSTIC R I N G S .
Theoretical
Physics
J. D . V E R G A D O S Section, University of loannina, E-mail: [email protected]
GR-45110,
Greece
The detection of the theoretically expected dark matter is central to particle physics and cosmology. Current fashionable supersymmetric models provide a natural dark matter candidate which is the lightest supersymmetric particle (LSP). The theoretically obtained event rates are usually very low or even undetectable. So the experimentalists would like to exploit the modulation effect. In the present paper we study a specific class of non-isothermal models involving flows of caustic rings. We find that the modulation effect arising from such models is smaller than that predicted by the isothermal models.
1
Introduction
In recent years the consideration of exotic dark matter has become necessary in order to close the Universe*. Recent data from the Supernova Cosmology Project 2 ' 3 suggest that the situation can be adequately described by a barionic component HB = 0 . 1 along with the exotic components QCDM = 0.3 and Q\ — 0.6 (see also Turner, these proceedings). Since this particle is expected to be very massive, mx > 30GeV, and extremely non relativistic with average kinetic energy T < lOOKeV, it can be directly detected 4 _ 5 mainly via the recoiling nucleus. Using an effective supersymmetric Lagrangian at the quark level, see e.g. Jungman et al 1 and references therein , a quark model for the nucleon 6 ' 9 and nuclear wave functions 5 one can obtain the needed detection rates. They are typically very low. So experimentally one would like to exploit the modulation of the event rates due to the earth's revolution around the sun. In our previous work 7 ' 8 we found anhanced modulation, if one uses appropriate asymmetric velocity distribution. The isolated galaxies are, however, surrounded by cold dark matter , which, due to gravity, keeps falling continuously on them from all directions 10 . It is the purpose of our present paper to exploit the results of such a scenario. 2
The Basic Ingredients for LSP Nucleus Scattering
The differential cross section can be cast in the form 8 : du — — i) — ^ da(u,v) = [(^s + S v -^) F\u) + £W»*ii(«)] rbvy2 305
(1)
306
with
£s
="»0 2 «"*K/S - ' ^ > a i = - i * *'<£>' <2>
where cr?xo is the relevant nucleon cross section. The functions T,spin, associated with the spin, and the small coherent term, Ey, associated with the vector contribution, are not going to be discussed further (see oue earlier work 8 ). In the above expression mjv is the proton mass, fir is the reduced mass, F(u) and Fn(u) are the usual and isovector spin nuclear form factors and u = q2b2/2, with b the harmonic oscillator size parameter and q the momentum transfer to the nucleus. The scale is set by
P(0)
m
da(u,v)\v\
A
mx
where dcr(u,v) is given by Eq. ( 1) One normally assumes p(0) = as the LSP density in our vicinity. 3
(3)
ATTIN
0.3GeV/cm3
Convolution of the Event Rate
In this section we will examine the consequences of the earth's revolution around the sun (the effect of its rotation around its axis is expected to be negligible) i.e. the modulation effect. Following Sikivie we will consider 2 x JV caustic rings, (i,n) , i=(+.-) and n=l,2,...N (N=20 in the model of Sikivie et al), each of which contributes to the local density a fraction pn of the the summed density p of each type i = +,-. and has velocity y„ = (ynx,yny,ynz) , in units of v0 = 220 Km/s, with respect to the galactic center. We find it convenient to choose the z-axis in the direction of the motion of the the sun, the y-axis is normal to the plane of the galaxy and the x-axis is in the radial direction. The needed quantities are takan from the work of Sikivie (table 1 of last Ref. 10 ) by the definitions yn = vn/v0,ynz = vn
/(*/) = £
S(v -v0yn)
(4)
307
The velocity of the earth around the sun is given by 5 . vE = v0 + vi = wo + fi (sina x - cosa cosj y + cosa sin-y z)
(5)
where a is the phase of the earth's orbital motion, Q = 0 around second of June. In the laboratory frame we have 8 v = v — VE 4
The N o n Directional Event Rates
Integrating Eq. (3) we obtain for the total undirectional rate R = Rt —7^r[l - h(a,Qmin)cosa]
(6)
where Qmin is the energy transfer cutoff imposed by the detector, and a — [y/2fj,rbv0]~l• Also pn = dn/p,p = J2n=i ^n (for each flow +,-). In the Sikivie model 1 0 2p/p(0) = 1.25. In the above expressions R is the rate obtained 4 by neglecting the folding with the LSP velocity and the momentum transfer dependence, i.e. by R=
p(0)
m
/737rf,
V{V)[I:S
-^A^;
, (v2) ^ ,
,f, +
pin +
*°
—*
v]
(7.
(7)
and it contains all SUSY parameters except mx The modulation is described in terms of the parameter h. The effect of folding with LSP velocity on the total rate is taken into account via the quantity t. We like to stress that it is common to extract the LSP nucleon cross section from the the meusured event rates in order to compare with the SUSY predictions . It is clear that in going from the data to the cross section, one should divide by t. The undirectional differential rate takes the form = R^tT(u)[l
- cosa H(u)}
(8)
The factor T(u) takes care of the u-dependence of the unmodulated differential rate. It is defined so that duT{u) = 1.
(9)
•>u„
i.e. it is the relative differential rate. umin is determined by the energy cutoff due to the performance of the detector. umax is determined by the escape 2
velocity ve8C via the relations: umax = max{^ n = 1,2, ...,N. On the other hand H(u) gives the energy transfer dependent modulation amplitude (relative to the unmodualated one).
308
-0.005
Figure 1: The quantities T(u) and H(u) entering the differential amplitude. Thick solid line corresponds to m^i = 30 GeV the intermediate thickness line to mch,i = 80 GeV, the fine line to m^i = 100 GeV. The rest correspond to larger LSP masses and fall on top of each other.f
5
Discussion of the Results and Conclusions
We have calculated the the total event rates for elastic LSP-nucleus scattering including realistic nuclear form factors. We focused our attention on those aspects of the problem, which do not depend on the parameters of supersymmetry other than the LSP mass. The parameter R, normally calculated in SUSY theories, was not considered in this work. The interested reader is referred to the literature , for a review 1 and references therin and, in our notation, to our previous work 4 ' 5 . We specialized our results for the target 1271. We considered the effects of the detector energy cutoff, by considering two typical cases Qmin = 10 and Qmin = 20 KeV. We assumed that the LSP density in our vicinity and the velocity spectrum is that of caustic rings of Sikivie et al 1 0 . The total rates are described in terms of t and h. In TABLE I we show how they vary the detector energy cutoff and the LSP mass. The parameters T(u) and H(u) entering the differential amplitude are shown in Fig. 1. The shape of T{u) is analogous to that of the isothermal models except that the maximum occurs at u — 0.0, rather than at u = 0.1. The function H(u) shows oscillations, which result in a smaller total modulation. Another way of understanding how the cancellations arize is to note that for some rings ynz > 1, while for others ynz < 1. The maximum occurs around the 2nd of December, something already noticed by Sikivie et a l 1 0 . Furthermore the modulation is small, h = 0.025, i.e. a 5% difference betwen the maximum and the minimum (see TABLE I). It is a bit smaller than that of the symmetric models, but a lot smaller than that predicted by the asymmetric ones 7 ' 8 , i.e h — 0.46
309 Table 1: The quantities t and h in the case of the target 5 3 / 1 2 7 for various LSP masses and Qmin i n KeV (for definitions see text). LSP Quantity
Wmin
mass
in GeV
10
30
50
80
100
125
250
t h
0.0 0.0
1.451 0.022
1.072 0.023
0.751 0.024
0.477 0.025
0.379 0.026
0.303 0.026
0.173 0.026
t h
10.0 10.0
0.000 0.000
0.226 0.013
0.356 0.023
0.265 0.025
0.224 0.025
0.172 0.026
0.098 0.026
t h
20.0 20.0
0.000 0.000
0.013 0.005
0.126 0.017
0.139 0.024
0.116 0.025
0.095 0.026
0.054 0.026
Acknowledgments The author would like to acknowledge partial support of this work by TMR No ERB FMAX-CT96-0090. References 1. For a recent review see e.g. G. Jungman et al, Phys. Rep. 267, 195 (1996). 2. R.S. Somerville, J.R. Primack and S.M. Faber, astro-ph/9806228; Mon. Not. R. Astron. Soc. (in press). 3. S. Perlmuter et al, Astrophys.J., in press (astro-ph/9812133); S. Perlmuter, M.S. Turner and M. White, Phys. Rev. Let. 83,670 (1999). 4. J.D. Vergados, J. of Phys. G 22, 253 (1996). 5. T.S. Kosmas and J.D. Vergados, Phys. Rev. D 55, 1752 (1997). 6. M. Drees and M.M. Nojiri, Phys. Rev. D 48, 3843 (1993); Phys. Rev. D 47, 4226 (1993). 7. J.D. Vergados, Phys. Rev. Let. 83, 3597 (1999) 8. J.D. Vergados, Phys. Rev. D62 (2000) 0235XX-1 ;astro-ph/0001190 9. T.P. Cheng, Phys. Rev. D 38 2869 (1988); H -Y. Cheng, Phys. Lett. B 219 347 (1989). 10. P. Sikivie, I. Tkachev and Y. Wang Phys. Rev. Let. 75, 2911 (1995; Phys. Rev. D 56, 1863 (1997); P. Sikivie, Phys. Let. b 432, 139 (1998); astro-ph/9810286
THE ORPHEUS DARK MATTER EXPERIMENT S. CASALBUONI, G. CZAPEK, F. HASENBALG, M. HAUSER, S. JANOS, U. MOSER, K. PRETZL, B. SAHLI Laboratory for High Energy Physics, University of Bern, Sidlerstrasse 5, CH 3012 Bern, Switzerland. B. VAN DEN BRANDT, J. A. KONTER, S. MANGO Paul Scherrer Institute, CH 5232 Villigen, Switzerland. T. EBERT, K. U. KAINER, K. M. KNOOP Institut fur Werkstoffkunde und Werkstofflechnik, Technical University of Clausthal, Agricolastrasse 6, D 38678 Clausthal-Zellerfeld, Germany. The ORPHEUS dark matter experiment is being built and most of it is already installed at our underground facility in Bern (70 m.w.e). The detector relies in an initial phase on 0.45 kg (1.6 kg maximum capacity) superheated superconducting tin granules (SSG) to measure recoils from weakly interacting massive particles (WIMPs). The current status of the installation as well as parallel ongoing studies are presented.
1
Introduction
The ORPHEUS detector is made of a homogeneous mixture of SSG in a dielectric filling material immersed in an external magnetic field. The type I superconductor granules are kept slightly below the boundary of their superconducting-to-normal phase transition in a metastable state. The recoil energy released by a particle interacting with a granule causes a temperature increase inversely proportional to the specific heat of the granule. Due to the rather small value of the specific heat below 500 mK, the energy deposited is enough to make a granule normal inducing a flux change because of the disappearence of the Meissner effect. The flux change, or "flip", is measured by a sensitive magnetometer, e.g. an LRC circuit. A general description of SSG detectors is given in Refs. 1 2 and of their readout electronics in Ref. 3 . 2
Experimental s e t u p
ORPHEUS is being built in the underground facility of the University of Bern (70 m.w.e.). A scheme of the experimental setup is shown in Fig. 1. Active shielding is provided by 2 cm thick plastic scintillators, followed by a passive shield: 15 cm normal lead, 4 cm OFHC-copper, and 18 cm of 310
311
boron-doped (5%) polyethylene. The whole shielding is mounted on rails to open it in two halves and to access the detector. The cold box consists of concentric copper thermal shields held at temperatures of 2.5, 4.5, and 80 K. The detector chamber is installed inside and is made of electroformed copper. For an initial trial, the detector will be filled with Sn grains between 30-34 pm in diameter mixed with PTFE (Teflon) powder as a filling dielectric material at 10% filling factor. For reading out the flipping granules, 56 pickup coils, 1.8 cm in diameter, 6.8 cm long and roughly 1500 windings will be used. The detector chamber will be maintained at a base temperature of «300 mK using an Oxford dilution refrigerator (300 [iW cooling power).
1n»«r
Figure 1. ORPHEUS experimental setup.
Last year the dilution refrigerator and the side access were successfully tested with a small prototype cold box at a base temperature of 200 mK 4 . At present, the dilution refrigerator, side access, cold box, detector chamber, and most of the shielding are already underground. The pickup coil bodies and solenoid are being built and from the shielding only the active veto is
312
missing. Before starting the first trial runs, however, a measurement of the actual background inside the shield with a high purity germanium detector is foreseen. 3
G r a n u l e studies
Improvement of the phase transition homogeneity in specially treated tin granules produced by gas atomization has been observed. In a recent study 5 , granules melted with a laser beam and fast cooled in a liquid nitrogen bath, exhibited better properties over untreated granules. Interesting results were obtained when the superheating critical field of single grains are measured at several angles with respect to the external magnetic field. While previous measurements of this kind 6 presented large variations (~20%) of the superheating fields, the latest measurements show that the treated granules possess rather flat (~2%) curves of superheating fields as a function of the rotation angle. A second sample of regularly spaced tin cylinders produced by tin evaporation onto a glass substrate also shows an improved phase transition homogeneity. While ramping up the external magnetic field «75% of the phase transitions occurr in a narrow magnetic field range, of the order of 3%, around the superheating critical field of «28 mT at 1.4 K 5 . 4
Conventional r e a d o u t
The conventional readout consists of a LRC circuit given by a pickup coil L, a cooled shunt resistor R, and a capacitance C determined by the combination of cable capacitances and the input capacitance of the low noise amplifier used to measure the signal. The noise level is determined measuring the Vtms after passing the amplified signal through a low-pass filter. A previuos study found an optimized configuration, in terms of signal-to-noise ratio (SNR), for the following set of parameters: L=9.8 mH, fl=10 kft and a cutoff frequency of of 15.9 kHz (-3 dB) for the low-pass filter. The pick-up coil used was 1.8 cm in diameter and 6.8 cm in length and had 1500 windings. The typical SNR obtained was 10 for granules 30-34 /an in diameter. Using the definitive ORPHEUS chamber, these SNR measurements were repeated at 1.4 K in a testing cryostat with several granule diameters. Phase transitions of Sn granules of 30-34 /xm in diameter produced this time a typical SNR of 18, improving our previous measurements. This result encourages us to use smaller granules (27-30 /im, SNR=13) toghether with the previously foreseen probes of 30-34 /zm for the firts trials of the ORPHEUS detector.
313
This experiment also tested the performance of a new test pulse design. 5
SQUID readout
To gain sensitivity to flux variations of small size granules (~20 /mi in diameter) and decrease the number of signal channels, a SQUID readout system will be implemented at a later stage. Tests of noise levels have started in a chamber with dimensions similar to the ORPHEUS detector. This chamber contains a superconducting shield made of a 1.5 /on Nb film coated on a copper cylinder 15 cm in diameter and 52 cm long. Inside, a superconducting solenoid of similar dimensions provides the external magnetic field. In order to operate a SQUID system inside this solenoid, the induced magnetic field has to be rather homogenous within the detection volume. Thanks to the superconducting shield, the inhomogeneity achieved has been measured to be < 2% in a region comprising w75% of the detection volume. At present, several runs with a flux-gate magnetometer have been performed to determine the flux attenuation of the superconducting shield as well as the solenoid homogeneity. The initial values of —57 dB attenuation have now been improved to —68 dB after the addition of a mu-metal cylindrical shield with open ends outside the chamber. Runs at 4.2 K have been performed in the absence of an applied field using two SQUID channels, one hooked up to a magnetometer sensing coil (3.5 cm in diameter, N = 5) and the other to a gradiometer (3.5 cm diameter, 2 cm separation between opposite windings). A two-turn loop between a magnetometer and a gradiometer is used as a calibration test pulse. Preliminary results suggest that either the shielding is insufficient or there is a leak through the shield at some point. According to measurements based on the calibration pulse, the minimum level of the magnetic fluctuations registered is « 15$ 0 - This value is too high in comparison with the expected flux variation from a single 20 /mi flip of ~ 10$ o - Further tests are required to clarify the issue. 6
Radiopurity measurements
A series of radiopurity measurements are performed continuosly at the Gotthard tunnel (3000 m.w.e, muon flux attenuation = 1.7xl0 - 6 ) with a small high purity germanium detector inside a passive shield. A background level of 10 (5) cts/keV.kg.d, at 15 (45) keV, has been achieved. The main contribution to the background comes from 2 1 0 Pb contamination in the inner low activity Pb of the shield. With this Ge detector, typical sensitivity levels
314
of ~ 5 , 10, 5, and 100 mBq k g - 1 have been obtained for the parent isotopes of 2 3 2 Th, 238 U, 137 Cs, and 40 K, respectively. Samples of OFHC Cu, Teflon powder (polytetrafluorethylene), Delrin (polyacetal polyoxymethylene), high purity Sn, and Nb(Ti) wire were measured with activity levels comparable to that of the background. A major concern is the activity of the normal lead to be used in the ORPHEUS shield. Several probes were measured for their 2 1 0 Pb content by a and 7 spectroscopy. It was found that normal ORPHEUS lead has an activity of ss200 Bq k g - 1 . This implies that an ultra-low activity inner Pb cylinder would be required to reduce the 7 ray flux due to the normal lead. 7
Outlook
The current status of the ORPHEUS project has been presented. We hope to finish the active shield and start data acquisition in the near future. Acknowledgments This work was supported by Schweizerischer Nationalfonds zur Forderung der wissenschaftlichen Forschung and by the Bernische Stiftung zur Forderung der wissenschaftlichen Forschung an der Universitat Bern. References 1. 2. 3. 4.
A. Drukier and L. Stodolosky, Phys. Rev. D 30, 2295 (1984). K. Pretzl, J. Low Temp. Phys. 93, 439 (1993)). K. Borer and M. Furlan, Nucl. Instrum. Methods A 365, 491 (1995). S. Casalbuoni et al. in The Identification of Dark Matter, Proceedings of the Second International Workshop, Buxton, England, 1998, edited by N. J. C. Spooner and V. Kudryavtsev (World Scientific, Singapore), 1999 p. .377 5. S. Calatroni et al. in Low Temperature Detectors, Proceedings of the 8 t h International Workshop, Dalfsen, The Netherlands, 1999. 6. M. Frank et a/., Nucl. Instrum. Methods A 287, 583 (1990).
FIRST RESULTS F R O M A LARGE SUPERHEATED DROPLET DETECTOR FOR D A R K MATTER SEARCHES J.I. C O L L A R , D. L I M A G N E , J. P U I B A S S E T , G. W A Y S A N D Groupe de Physique
des Solides (UMR CNRS 75-88), Universites 75251 Paris Cedex 05, France
Paris
7 & 6,
T.A. GIRARD Centro de Fisica Nuclear,
Universidade
de Lisboa,
1649-003
Lisbon,Portugal
H.S. M I L E Y Pacific
Northwest
National
Laboratory,
Richland,
WA 99352,
USA
SIMPLE (Superheated Instrument for Massive ParticLE searches) employs superheated droplet detectors (SDDs) to search for Weakly Interacting Massive Particle (WIMP) dark matter. As a result of the intrinsic SDD insensitivity to minimum ionizing particles and high fluorine content of target liquids, a competitive WIMP sensitivity has been obtained already at the prototype stage. We comment here on the expected immediate increase in sensitivity of the program, which aims at an exposure of >25 kg-day during the year 2000.
N.B: Three out of six authors in this paper got their Ph.D. 's under Frank Avignone 's direction... old crackers stick together! 1
Introduction
Superheated Droplet Detectors (a.k.a. SDDs or Bubble Detectors) consist of a dispersion of small drops (radius ~ 10 fim) of superheated liquid in a gel or viscoelastic matrix. This recent development in radiation detection technology 1 combines the well-known characteristics of bubble chambers with several advantageous features: the SDD matrix isolates the fragile metastable system from vibrations and specially from convection currents (absent in gels), while the smooth liquid-liquid interfaces impede the continuous triggering on surface impurities that occurs in the walls and gaskets of even the cleanest bubble chambers. In this way, the lifetime of the superheated state is extended to the point that new practical applications such as personnel and area neutron dosimetry become possible. In the moderately superheated industrial refrigerants used in SDDs, bubbles can be produced only by particles having elevated stopping powers, as is the case for low-energy nuclear recoils. This behavior is described by the "thermal spike" model 2 : for the transition to occur, a vapor nucleus or "protobubble" of radius > rc must be created, 315
316
while only the energy deposited along a distance comparable to this critical radius rc is available for its formation. Hence, a double threshold is imposed: the requirement that the deposited energy E be larger than the thermodynamical work of formation of the critical nucleus, Ec, and that this energy be lost by the particle over a distance 0 ( r c ) , i.e., a minimum stopping power. Protobubbles formed by energy depositions not meeting both demands simply shrink back to zero; otherwise, the transition is irreversible and the whole droplet vaporizes, producing a characteristic violent sound that can be used for detection purposes. Both thresholds can be varied by tuning the operating temperature and pressure 4 : keV nuclear recoils are detectable at room temperature and atmospheric pressure, allowing for a low-cost WIMP search free of the complications associated to cryogenic equipment. Most importantly, the minimum stopping power requirement provides a total insensitivity to most backgrounds interfering with WIMP searches, alpha emitters being the only internal contaminants of concern.
2
Detector Fabrication
SDDs of active mass O(l) kg can considerably extend the present experimental sensitivity to WIMPs well into the region where new supersymmetric particles are expected 3 . The modest active mass of commercially available SDDs (~ 0.03 g refrigerant/dosimeter), together with a desire to control and understand the fabrication process, lead us to develop a large-volume pressure reactor dedicated to SDD production. Three years of R&D have been necessary to produce SDD modules simultaneously meeting all requirements of alpha-emitter radiopurity, stability and large-mass (a factor 1000 more than commercial dosimeters) necessary for this new application. The particulars on fabrication process, condensed matter issues, purification and testing in a shallow underground gallery can be found in 4 . The emphasis of the program has been on keeping the number of detector components down to a minimum for reasons of radiopurity, safety and cost, imposing the need to solve multiple condensed-matter issues with as few elements as possible. The target mass of present SIMPLE modules is limited only by the size of the pressure reactor, and they can be operated continuously for up to ~ 40 d (most commercial bubble dosimeters are intended for just a few hours of continuous exposure). The design of recompression chambers that would allow for an indefinite exposure is under study. The cost of the SDD matrix has been kept down, allowing for a much larger future design.
317
3
Calibrations
The response of smaller SDDs to various neutron fields has been extensively studied 5 ' 6 , 7 and found to match theoretical expectations. However, largesize, opaque SDDs require independent calibration: acoustic detection of the explosion of the smallest droplets or of those most distant from the piezolelectric sensors is not a priori guaranteed. Two separate types of calibration have been performed to determine the target mass effectively monitored in SIMPLE modules and to check the calculation of the temperature- and pressuredependent threshold energy Ethr above which WIMP recoils can induce nucleations 3 , s . SIMPLE SDD modules have been exposed to a well-characterized 252 Cf neutron source at the TIS/RP calibration facility (CERN) 4 . A satisfactory agreement between the observed nucleation rate and Monte Carlo simulations was found at different pressures and temperatures 4 . As a result of this calibration, the sound detection efficiency with present sensors was found to be 34 ± 2% (74 ± 4%) at P =2 atm (P =1 atm), a decisive piece of information to obtain dark matter limits. A second type of calibration can be performed by homogeneously diluting a liquid 241 Am source (an alpha-emitter) into the matrix prior to gel setting. The 5.5 MeV alphas and 91 keV recoiling 237 Np daughters cannot induce nucleations at temperatures below Ta and Tar, respectively 3 | 4 . These alpha calibrations allowed us to corroborate the calculation of Ec for each refrigerant (a good knowledge of Ec is crucial to determine the expected SDD sensitivity to WIMP recoils in different operating conditions).
4
First Results
The installation 500 m underground of SIMPLE modules started in July 1999. A decommissioned nuclear missile launching control center has been converted into an underground laboratory 9 ' 10 , facilitating this and other initiatives. The characteristics of this site (microphonic silence, unique electromagnetic shielding of the halls 9,1 °) make it specially adequate for rare-event searches. Modules are placed inside a temperature-regulated water bath, surrounded by three layers of sound and thermal insulation. A 700 1 water pool acting as neutron moderator, resting on a dual vibration absorber, completes the shielding. Events in the modules and in external acoustic and seismic monitors are time-tagged, allowing to filter- out the small percentage (~ 15%) of signals correlated to human activity in the immediate vicinity of the experiment. The signal waveforms are digitally stored, but no event rejection based on pulse-shape considerations is performed at this stage (the sharply-resonant
318
piezoelectric sensors presently employed will eventually be substituted by others with a flatter spectral response, allowing for univocal identification of the nucleation sounds). Accounting for sound detection efficiency and a 62% fluorine mass fraction in R-115 (C2CIF5), limits on the spin-dependent WIMPproton cross section awp (Fig. 1) have been extracted from the first 0.19 kg-day of exposure of the prototype module 4 . Present SIMPLE limits are impaired by the large statistical uncertainty associated to the short exposure accumulated so far, and not yet by background rate. A considerable improvement is expected after the ongoing expansion of the bath to accommodate multiple modules (Fig. 1). SIMPLE 2000 aims at an exposure of ~ 25 kgday in the next few months, by replacing the detectors (in batches of seven) every four to six weeks, repeating this cycle several times. A weak Am/Be neutron source will be used at the end of each run to assess in situ the sound detection efficiency of each module. In parallel to this, plastic module caps have been replaced by a sturdier metallic design: runs using refrigerant-free modules showed that a majority of the prototype events arose from pressure microleaks in the old caps, correlated to the sense of T ramping and able to stimulate the piezoelectric sensor 4 . In principle, if this source of background is controlled, the maximum sensitivity of SIMPLE 2000 can already start to probe the spin-dependent neutralino parameter space (Fig. 1). 5
Prospects
The smallness of spin-dependent WIMP scattering rates vis-a-vis their coherent (spin-independent) counterparts has regrettably produced a diminished interest in developing detectors able to exploit the first. However, the importance of the spin-dependent channel, for which fluorine-rich detectors are by far the optimal target u , has been recently underlined by its relative insensitivity to CP-violation parameter values, which may severely reduce coherent interaction rates 12 . To further stress the significance of this channel, we have emphasized 4 a not-so-obvious complementarity of spin-dependent and spinindependent searches in exploring the neutralino phase space, arising from a lack of correlation between both crossections in theoretical models: If an exhaustive test of the neutralino-as-cold-dark-matter hypothesis is ever to be completed, the development of fluorine-rich detectors cannot be neglected: in this respect SDDs represent an ideal opportunity. At the time of this writing, SIMPLE and PICASSO (a similar Canadian search) are considering merging into a large international effort (U. Montreal, U. Paris, U. Lisbon, PNNL, SNO, Yale U., Bose Institute) to maximize the returns from this promising approach to dark matter detection.
319
Figure 1. 95% C.L. limits on uvp p extracted from 0.19 kg-day of SDD exposure, compared with other experiments (DAMA limits stem from a ~ 1.5 • 10 4 kg-day exposure). The lower lines indicate the expected sensitivity of SIMPLE 2000 after an exposure of 25 kg-day, if no improvement in the background is obtained (dashed line) or at the maximum reachable level for this exposure (zero background, solid line off the scale). "MSSM" marks the tip of the region where a lightest supersymmetric partner is expected.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
R.E. Apfel, Nucl. Instr. Meth. 162, 603 (1979). F. Seitz, Phys. Fluids 1, 1 (1958). J.I. Collar, Phys. Rev. D 54, R1247 (1996). J.I. Collar, J. Puibasset, T.A. Girard, D. Limagne, H.S. Miley and G. Waysand, Phys. Rev. Lett. 85, 3083 (2000); New J. Phys. 2, 14.1 (2000). M. Harper, Nucl. Sci. Eng. 114, 118 (1993); Nucl. Instr. Meth. A 336, 220 (1993). Y.-Ch. Lo and R. Apfel, Phys. Rev. A 38, 5260 (1988). F. d'Errico, Radiat. Prot. Dosim. 84, 55 (1999). M. El-Nagdy et al., J. Br. Nucl. Eng. Soc. 10, 131 (1971). http://home.cern.ch/collar/RUSTREL/rustrel.html h t t p : / / i a p e t u s . unice. f r / ~ r u s t r e o u / G. Waysand et al., astro-ph/9910192. J. Ellis, R.A. Flores, Phys. Lett. B 263, 259 (1991). T. Falk et al., hep-ph/9806413; P. Gondolo, priv. comm.
BACKGROUND STUDIES FOR THE DOUBLE BETA DECAY EXPERIMENT NEMO 3
C. SEAN SUTTON Mount Holyoke College, South Hadley, MA 01075, USA E-mail: [email protected] DOMINIQUE LALANNE LAL, IN2P3-CNRSet Unversite Paris-Sud, 91405 Orsay, France E-mail: lalanne@lal. in2p3.fr Neutrinoless double beta decay will be sought with the NEMO 3 detector to investigate physics beyond the standard model. Combinations of Monte-Carlo calculations and background measurements indicate that the detector will ultimately be limited by the background from two neutrino double beta decay events. Source purity is discussed in the context of HPGe measurements, and backgrounds from gamma rays and neutrons are presented. Given the current configuration, NEMO 3 is expected to be sensitive to half-life limits of the order of 1025y which for 100Mo will reach an effective neutrino mass as low as 0.2 eV.
1
Introduction
The NEMO 3 detector [1] has been designed to study neutrinoless double beta (BBOu) decay to a half-life limit of 102S years. Renewed interest in GBOu decay has been generated by the compelling neutrino mass observations of such experiments as SuperKamiokanda [2]. These observations and their consequences will be discussed elsewhere in the CSNP Proceedings. In the NEMO 3 detector, the source is not the detector as is the case with a number of other double beta decay experiments like the 76Ge experiments [3,4]. Rather, the detector looks at thin foils of double beta decay isotopes (100Mo, 82Se, 130 Te, ...). In one scenario, the detector will house 10 kg of 100Mo foils that collectively form a 20 m2 by 50 um/cm2 thick source. These foils are fixed vertically in a volume designed to track pairs of electrons. The tracking volume is a cylindrical annulus, and particle trajectories are resolved with 6,180 open drift cells with octagonal cross sections that also run vertically and operate in Geiger mode. This tracking volume is surrounded by 1,920 very low activity scintillators and photomultiplier tubes that function as a calorimeter. The signature of BBOu decay is 320
321
the emission of two electrons, the sum of whose energy is the Q-value of the decay. This process has not been seen to date, whereas the two neutrino double beta decay process has been successfully studied. Though the Q-values of the isotopes to be studied are rather high, so that the naturally occurring backgrounds are small, there are a variety of possible background events that mimic double beta decay. One class of events comes from electrons that are created in the scintillators or structural members of the detector, and then cross through the detector. These events are easily recognized by time-offlight measurements and are excluded by cuts in the data [5,6]. There are two other troublesome events that cannot be excluded by this cut. The first of these comes from contamination in the source foil and is identified here as internal background. The second, external background, comes from neutral particles entering the detector and interacting with the source foil.
2
Internal background
The Q-value for 100Mo is 3.03 MeV, thus the beta decay of 214Bi (3.3 MeV) and 208T1 (5.0 MeV) are capable of producing two electron events by various scattering processes which will mimic double beta decay. To minimize the contribution of these two decays, the decay chains of 238U and 232Th must be eliminated to the level where the activities of 214Bi and 208T1 are less than 0.3 mBb/kg and 0.02 mBq/kg respectively. These activity limits were calculated to allow for two 214Bi events and two 208T1 events that will mimic BBOu decay events when 10 kg of 100Mo are studied with a live time of five years. To meet these stringent requirements, a chemical purification process was developed at the Idaho National Engineering and Environmental Laboratory. The process involves the extraction of uranium and thorium through precipitation of Mo0 3 . This process is coupled with the introduction of a few barium atoms to block the sites otherwise occupied by Ra. Thus, uranium, thorium and radium are removed from the molybdenum powder. The process was developed with samples that were measured with the HPGe detectors in the Frejus Underground Laboratory. These tests were initially carried out on natural molybdenum, whose starting radiopurity was roughly 100 times less than the 100Mo currently being purified. In Table 1, the extraction factors for two samples are given. The sample INEEL-3 was a "atMo sample and Batch 7 is a 100Mo sample. On inspection of Table 1, one notices that for the natMo sample there is a reduction factor on the order of 100 for U and Th, while other limits suggest a reduction factor of at least 39 for Ra. It is not clear that the same reduction factors will hold on the initially cleaner 100Mo sample. However, the 100Mo sample was
322 enriched with 235U, which still shows a large reduction factor and that owing to the similar chemical properties of 238U and 232Th, the extraction factors should not differ profoundly. The extraction limits on Ra are greater than 12 and may be greater than 39. The HPGe measurements go on directly to confirm that the 214Bi activity is less than the experimental design requirement of 0.3mBq/kg. The task of measuring the 208 T1 activity to less than 0.02mBq/kg requires larger samples and longer integration times than have been practical. However, if one assumes an extraction factor of about 25, one reaches the limit. Thus, it is possible that the actual activity is much less than the limit sought.
Table 1: Extraction Factors from Chemical Processing Isotope
238 235
u
U
228Jh 226 228
Ra Ra
Nat
Mo INEEL-3
Mo Batch 1
100 ± 3 0 140 ± 20
>16
>95 >39 >34
>8 >12 >2
100
80 ± 3 0
The other internal background for studies of 13130u decay is the tail [7] of the two neutrino double beta decay curve. In an experiment with NEMO 2 [8], the half-life of this decay mode was determined to be T1/2 = 0.95±0.04(stat)±0.09(syst)xl0 19 y. Given the resolution of the detector, it will be shown that this background will be the dominate one and ultimately limit the performance of NEMO 3.
3
External backgrounds
The NEMO 3 detector is being assembled in the Frejus Underground Laboratory. The depth is 4,800 m.w.e. and the cosmic ray muon flux is reduced by more than one million to 5xl0"9s"'cm"2. This, combined with the orientation of the detector, liberates the detector from a background component introduced by cosmic ray muons. To study the effect of gamma rays and neutrons it is useful to study events identified as "one crossing electron events" (OCE events), where time-of-flight measurements eliminate them from the data sample but provide useful checks on
323
the external background. The events are typically produced by gamma rays giving Compton electrons in the scintillators which then cross the detector The energy spectrum in the laboratory is known for gamma rays. Simulations of OCE events with a 20 cm iron shield were performed. These simulations indicated that in the energy region of interest, which is 2.8 to 3.2 MeV (NEMO 3 has a resolution of 400 keV FWHM at 3 MeV), the gamma ray contribution to OCE events is negligible. Thus, it will be negligible as an external background for flfiOu decay. Neutrons are more complicated. The measured thermal (» 0.025 eV) and fast (>1 MeV) neutron fluxes were found to be (1.6±0.1)xlOVcm- 2 and (4.0±1.0)xl0" Vcm" 2 , respectively. For the intermediate energies (epithermal), values which were measured in other underground laboratories, such as Gran Sasso, were used and found to be of the same order of magnitude as the thermal and fast neutron fluxes. This information coupled with data taken with an Am-Be neutron source on the earlier prototype detector, NEMO 2, gives an indication of the neutron contribution to the OCE event channel. The energies of the neutron-induced OCE events can exceed 8 MeV. The thermalized neutrons are captured on the iron and copper frame of the detector and emit high-energy gamma rays which then Compton scatter. Additionally, a peak in this spectrum appears at 1.8 MeV due to 2.2 MeV gamma rays coming from neutron captures in the hydrogen component of the scintillators. Simulations of neutron interactions with the GEANT/MICAP code were make. The use of an extended gamma ray library was applied to various shielding configurations for NEMO 3. The shield selected is 20 cm of borated polyethylene inside of which is 20 cm of iron and finally 1cm more of borated polyethylene. This, coupled with a magnetic field (30 Gauss) for charge identification will optimally limit the number of "two electron" events. For neutron and gamma rayinduced two electron events this gives 0.02+004_o02 and 0.10±0.03 events per five years respectively. Consequently, the external background has been effectively eliminated for the NEMO 3 detector.
4
Current Status of the NEMO 3 Detector and Expected Performance
As indicated above the NEMO 3 detector is currently being assembled. Three of the 20 sectors, which will form the cylindrical detector, are operational and presently collecting preliminary data. The remaining mechanical components have all been built and the assembling of the detector continues as of May, 2000. In summarizing the backgrounds and running a possible positive detection of an effective neutrino mass, <mu>, one will have the following backgrounds: the external background from gamma rays and neutrons will not contribute
324
significantly over five years; the internal background will have a total of four events from 2I4Bi and 208T1, and perhaps less depending on the chemistry. The twoneutrino background will introduce 5.5 events in five years. The expected background will be 9.5 events/5y. If a signal of 24 events is seen then the 5 cr effect would give T,/2 = 3.5xl0 24 y, and <m„> = 0.3 to 0.7 eV depending on the nuclear matrix element. References 1. Dassie, D. et al., NEMO 3 Proposal, LAL Preprint 94-29 (1994). 2. Fukuda, Y. et al., Evidence for Oscillation of Atmospheric Neutrinos, Phys. Rev. Lett. 81 (1998) pp. 1562-1567. 3. Gunther, M. et al., Heidelberg-Moscow (3(3 experiment with 76Ge: Full set up with five detectors, Phys. Rev. D 55 (1997) pp. 54-67. 4. Brodzinski, R. L. et al., Status Report on the International Germanium Experiment, Nucl. Phys. B Proc. Suppl. 31, (1993) pp.76-79. 5. Arnold, R. et al., Double Beta Decay of 82Se, Nucl. Phys. A 636 (1998) pp. 209-223. 6. Longuemare, C. et al., Measurement and Control of the 214Bi Contamination in the PP NEMO-2 Experiment, Nucl. Instr. Meth. A 401 (1997) pp.144-155. 7. Boehm, F. and Vogel, P., Physics of Massive Neutrinos (Cambridge University Press, 1987). 8. Dassie, D. et al., Two-neutrino double-P decay measurement of 100Mo, Phys. Rev. D 51 (1995) pp. 2090-2100.
D O U B L E BETA D E C A Y A N D T H E M A J O R A N A P R O J E C T LUDWIG DE BRAECKELEER* TUNL, Duke University, Durham, North Carolina E-mail: [email protected] A novel design for a large-mass double-beta decay experiment is presented. Based on the expertise gathered over the last decade, we argue that the project is feasible, and could reach a Majorana mass sensitivity of a few hundredths of an eV.
1
Introduction
Neutrinos are enigmatic particles. They have been at the forefront of fundamental physics for the better part of the last century, yet very little is known about them. Do they have mass? Do they oscillate? In fact, it is still unknown if they are their own anti-particles. From a theoretical view, it is natural to expect them to have mass, to oscillate and to be their own anti-particles. The search for neutrinoless double-beta decay is the only known practical method to investigate the Majorana vs. Dirac nature of the neutrino. Indeed, the black-box theorem 1 ' 2 shows that in the context of most gauge theories, the observation of the neutrinoless mode of double-beta decay would imply the existence of one massive neutrino eigenstate. Various authors 3 have recently argued in favor of a neutrino mass between 0.01 and 1 eV. Consequently, the search for this process is timely and may prove very fruitful. We are proposing an experiment, the Majorana Project, which could probe such a small neutrino mass. The project relies entirely on proven technology and takes advantage of two major opportunities of our time. The first is the opening of a US underground facility in Carlsbad, New Mexico for scientific activities. The second is the end of the cold war which makes it possible to collaborate with former Russian weapons facilities in order to produce large amounts of isotopically-enriched materials needed. It is worth noting that besides the neutrinoless mode of double-beta (0i//3/?) decay, there is a second mode with the emission of two neutrinos (2i//?/?). Although this mode is allowed, it is very suppressed since it is a second order weak interaction process. In the last decade, several 2i//3/3 transitions have been observed. These observations provide nuclear theorists with a test of their calculations of the nuclear matrix elements (NME). Similar calculations are needed not only to deduce a lower limit on a Majorana neutrino *FOR THE MAJORANA COLLABORATION 325
326
mass but also to compare the relative merits of different experiments. Further, the experimental study of the 2v(3f3 decay generates ideas which can be used to improve previous searches of (V/?/? decay. We describe in this paper the evolution of such an approach. A few years ago, we suggested a novel method to look for 2z//?/3 decays to excited 0 + states. The technique is simple and solid. The double-beta decay is unambiguously observed by the measurement in coincidence of the two 7 rays emitted in the decay of the excited daughter state. This technique was applied to the decay of 100 Mo which is described in section 2. However, this method is quite general and can be applied to many isotopes which have a similar nuclear structure. During the summer of 1999, we formed a collaboration to build an improved apparatus. We decided to investigate the /?/? transition of 76 Ge to the first excited 0 + state of 76 Se. This apparatus is described in section 3. It is straightforward to build an active sample for this measurement and the signature of the decay is a triple coincidence between two 7 rays and the deposited energy of the two electrons. Because the electrons have a short-range, it is natural to design a segmented Germanium detector. It was soon realized that the segmentation strongly suppresses the background and this insight gave rise to a new concept for a Qvfifi decay apparatus which is described in section 4. Finally, in section 5, we comment on the unique features of 100 Mo for the detection of low-energy solar neutrinos. 2
The TUNL-ITEP
100
Mo Experiment
Previous experiments have focused on very low-background detection systems 4,5 ' 6 . This has been obtained by building the detectors with lowradioactivity materials and by operating the experiments in underground laboratories which offer an efficient shielding against the cosmic-ray induced background. An alternative approach to background reduction is to employ a coincidence technique, in which two separate detectors simultaneously detect the two emitted 7 rays from the 2i//?/3(0+ -» Of) decay of 100 Mo with E 7 i = 590.76 keV and E 7 2 = 539.53 keV (Fig. 1). This approach was accomplished for the first time in the present work by using two HPGe detectors (8.5-cm diameter, 5-cm width, 1.8-keV FWHM energy resolution at 1.33 MeV and 0.7-keV resolution at 0.122 MeV) in coincidence. A disk sample of molybdenum is sandwiched between these two 7-ray detectors which are inserted into a Nal annulus that acts as an active veto. Plastic plates on either side of the apparatus act as a veto for the regions not covered by the Nal annulus. The entire apparatus is surrounded by a passive shielding made of lead bricks (Fig. 2). The experiment is conducted inside the Low Background Counting Facil-
327
0,
1130.3
Figure 1. Decay scheme involving 00 decay of 1 0 0 Mo and EC and 0 decay of plified 7 decay scheme of 1 0 0 R u is also indicated.
100
T c . Sim-
Figure 2. A cut-away view of the TUNL-ITEP 1 0 0 Mo 0 0 decay apparatus, including the 100 Mo disk, two HPGe detectors, active veto shielding (consisting of a Nal annulus and two plastic end-plates), and passive lead shielding.
328 •
•
•
•
i
•
•
•
1
•
•
•
•
1
'
•
•
•
8 -
«,[«•«
1lull
IR.UIJIII
«,«•»]
Figure 3. The left plot shows the 7-ray spectrum in coincidence with 540 ± 2.5 keV. The plot on the right shows the 7-ray spectrum in coincidence with 591 ± 2.5 keV. Note the 22 coincidence events in 440 d of measuring time.
ity of the Triangle Universities Nuclear Laboratory (TUNL), a well-shielded room located in the basement of the Physics Department of Duke University. A 98.4% enriched 100 Mo sample (1.05-kgmass, 10.6-cm diameter, 1.1-cm thickness) was studied for 440 days. Twenty-two 540 keV-591 keV coincidence events were detected, and Figure 3 shows the 7-ray spectra in coincidence with 540 ± 2.5 keV (left panel) and 591 ± 2.5 keV (right panel). A measurement of the background rate yields 2.5 events per 5 keV. From these data, the half-life of the P/3 decay of 100 Mo to the first excited 0+ state of 100 Ru can be determined. After subtraction of the background, one obtains an half-life time of (4.9taJ) x 10 20 yr. Before our result obtained above ground can be compared to the previous results for 10o Mo measured underground, one must investigate the possibility whether or not the signal observed in our 100 Mo double-beta decay experiment, or at least part of it, is due to 100 Mo(p,n) and/or 100 Mo(n, 7) reactions. The 100 Mo(p, n) reactions to bound states in 100 Tc will produce, after prompt gamma-ray de-excitation, the ground state of 1 0 0 Tc. The latter has a half-life time of 15.8 s and a 5.7% branching ratio to the 0+ state of 100 Ru which decays to the ground state by emission of two 7 rays of 540 and 591 keV. It is worth noting that our Nal veto is not acting against this process due to the 15.8-s half-life of the 100 Tc ground state. One way to estimate the magnitude of the 1M Mo(p,n) process is by searching for the very similar 95Mo(p,n) reactions. Natural molybdenum has a 16% abundance of 95 Mo. A 1-kg disk of natu-
329 ral molybdenum (10-cm diameter, 0.965-cm thickness, 9.6% 100 Mo), has been stored for one year in the apparatus and subsequently investigated for 180 days. The production of the isomer (9/2+, half-life time of 20 h) of 9 5 Tc can proceed via the (p, n) reaction on 95 Mo and is very practical for our purpose. This element decays with large branching ratios to two excited states: 41% to the 786.2-keV state and 32% to the 1039.2-keV state. In both cases, these states decay to the first excited state at 204 keV with a branching ratio of 78 and 88% , respectively. Moreover, the first excited state has a short lifetime of 0.8 ns. Data from 180 days of observation with the natural molybdenum sample were analyzed, and no such coincidence events were detected. Thus, it can be concluded that proton background does not contribute significantly to the 0(3 signal. Another potential source of background are neutrons which could contribute via neutron capture in the 100 Mo(n, 7) reaction. This is a possible background because the decay of 101 Mo yields some 7 rays with energies very close to the ones of interest in the /3/3 decay of 100 Mo. However, this neutron capture process should also produce other 7 rays in coincidence with a 105 higher probability. Since these 7 rays have not been observed in our experiment, theoretically this possible background is completely ruled out. 3
The Guernica Project
Continuing on the work completed with the TUNL-ITEP experiment discussed in Section 2, the Guernica Project (Germanium Underground Experimental Research In CArlsbad) plans to investigate additional 2i//?/3 decays to excited states. The experimental setup will consist of 14 HPGe detectors: 2 central detectors surrounded by 12 additional ones, all of which are inserted into a Nal annulus acting as a veto (see Figure 4). One of the central HPGe detectors will be constructed from enriched 7 6 Ge. This detector will then be segmented both axially and azimuthally to create 18 segments - 3 internal axial segments and 6 external azimuthal segments (see Figure 5). Such segmentation will permit detailed study of 2i//?/3 decay of 76 Ge to the 0+ state of 76 Se, whose signal consists of two /3 particles ( E e i + e 2 = 0 - 917 keV), detected in one segment due to the electrons' short range, and two 7 rays (E 7 i = 563.2 keV and E 7 2 = 559.1 keV) detected elsewhere - see Figure 6. As Table 1 shows, the theoretical predictions for this 2i//3/?(0+ - • 0+) half-life vary widely based on the model used to make the prediction even through all of the listed models are able to easily reproduce the the half-life of the transition to the ground-state. Indeed, in the framework of QRPA models 9 ' 12 ' 13 , the NME are functions of
330
Figure 4. An overview of the Guernica Project apparatus. One of the central HPGe detectors is made from 76 Ge and consists of 18 segments. #8cm Interna] segmentation lines - 2
External segmentation lines 6 - each at 60 degrees apart
Figure 5. A schematic showing how the 76 Ge crystal wil be segmented - side and end views are shown. It will be made into an n-type detector.
the particle-particle strength parameter gpp. The NME for transitions to the ground state depend strongly on gpp. This provides a given model with the freedom to choose an appropriate value for gpp such that predictions of the half-life match that measured experimentally. On the other hand, transitions
331
0
As
2039.3
0,
1122.3
2,
559.1
0
(keV)
Figure 6. Double-beta decay scheme of 76 Ge.
Table 1. QRPA-calculated transition half-lives for the 2vf}f} decay of 76 Ge to the ground (g.s.) and Of excited states of 76 Se. Also experimental half-lives are given. r 2 ; 2 ( g .s.) (yr)
TfhOt) (y)
Model
1.4 X 10 21 9 X 10 2 0 7.7 x 10 2 0 1.1 x 10 2 1 1.4 x 10 2 1 1.4 x 10 2 1
> 1.7 x 10 2 1 4 x 10 2 2 7.5 x 10 2 1 1.7 x 10 2 4 1.0 x 10 2 3 3.1 x 10 2 3
Experiment Woods-Saxon Adjusted Woods-Saxon Woods-Saxon Woods-Saxon Adjusted Woods-Saxon
Reference 7 8 9 10 11 11
to the Of state have little dependence on the choice of gpp. Therefore, the measurement of the half-life of the 2i//3/3 transition to the Of state will constitute a more severe test of these models. At the same time, the Guernica Project apparatus can be used simultaneously to study passive samples placed between the two central HPGe detectors. Monte-Carlo simulations have shown that this apparatus will provide an efficiency for gamma-gamma coincidence that is at least 10 times larger than that obtained with the TUNL-ITEP experiment discussed in Section 2. This increase will enable study of the 2i//3/3 decays to excited 0 + states in other isotopes such as 82 Se, 96 Zr, 130 Te and 150 Nd.
332
Figure 7. An overview of the Majorana Project apparatus, with the lead shielding removed to show detail. Each "pull-out" section will consist of two super-cryostats.'
4 4-1
The Majorana Project Overview
The Majorana Project (see Figure 7) is a 500-kg double-beta decay experiment to be located deep underground. The apparatus will consist of 210 76 Ge crystals placed in 10 super-cryostats (see Figure 8). Each crystal will be similar to the segmented 76 Ge detector described in section 3. The 76 Ge (86% enrichment) will be separated in Krasnoyarsk, Russia. After zone-refinement of the germanium at Eagle Picher, Oklahoma, the crystals will be grown at PerkinElmer in Tennessee. The apparatus will be located in the US underground facility in Carlsbad, New Mexico. 4.2
The WIPP Facility
The Department of Energy, Carlsbad Area Office, has committed to provide an underground setting for our double-beta decay experiments at the Waste Isolation Pilot Plant (WIPP) in the southern half of the Room Q alcove. As part of this commitment, they will provide two enclosures in which the experiments will be located. These enclosures will be outfitted with power supplies,
333
Figure 8. A Yiew of one of the super-cryostats. Each super-cryostat will consist of 21 individual segmented 76 Ge crystals encased in copper and to be cooled to LN2 temperature.
air conditioning and insulation, and telephone and internet connections. The WIPP facility is located in a salt bed at a depth of 2000 mwe underground. The background radiation at WIPP has been measured. Potassium levels of 209 ± 200 ppb, thorium levels of 96 ± 59 ppb and uranium levels of 49 ± 2 4 ppb have been found. The neutron flux has been estimated as 340 m~ 2 d % and the cosmic muon flux as 1-2 x ! 0 " 7 cnr^s"" 1 .
4.3
The Production of1QGe in Krasnoyarsk
The end of the cold war makes it possible to consider the production of enriched 7 6 Ge in Russia. This element can be separated by the centrifuges that were developed and built for the weapons program in the USSE. With small (10-20 kg) quantities of 7 6 Ge already produced at Krasnoyarsk using equipment never employed for weapons enrichment, the technology and ability to create large quantities of this isotope are assured. As a potentially significant element of DOE's Nuclear Cities Initiative, the enrichment of 500 kg of 76 Ge as a way of supporting Russian weapons scientists could become a reality in only a few yeaxs. An initial cost-estimate for 500 kg of enriched 7 6 Ge, based on discussions with Majorana Project collaborators in Russia, is about 20 million US dollars. .
334
0
Z
4
6 B 10 12 14 16 IB segment hit*
Figure 9. Monte-Carlo simulation of 6 0 C o background. The plot on the left shows the spectrum before and after segmentation. The plot on the right displays the number of segments in which energy was deposited from the internal 6 0 C o background.
4-4
Monte-Carlo
Simulations
We have modeled the geometry of a ~2-kg HPGe detector, including mounting materials similar to those used in existing ultra-low background detectors. Simulations using this geometry have been completed for internal cosmogenic 60 Co (see Figure 9) and 68 Ga, the most important backgrounds expected in an ultra-low background 76 Ge /?/3-decay experiment. Simulations of internal 76 Ge Qvfiji decay have also been completed. The result of these simulations has been analyzed to predict the background suppression resulting from detector segmentation. Detector designs with six outer azimuthal segments and from one to four internal axial segments have been studied. The results have been summarized and will guide the final detector design. Three-fold axial and six-fold azimuthal segmentation give projected limits on the half-life of 76 Ge Oi//?/? decay 2.84 times more stringent than those resulting from a similar experiment without segmentation. Our projections have not yet taken into account the benefits expected from the self-shielding properties of an array of Ge detectors where anti-coincidence requirements are enforced.
335 1 1 11 1 11 1
§3000
' DEP e«lclency= 0.8018 gamma ettlclency = 0.2645
2000-
Figure Of Merit = 1.96
\ 1500-
Jl 1000-
900_ J"W^, ~ ~ T ^ 4 < 1
I960
1580
1600
1620
1640
1660
1680
1700
keV Figure 10. Plot which shows the effectiveness of Pulse-Shape Discrimination.
4-5
Pulse-Shape
Discrimination
Drawing on earlier work in Pulse-Shape Discrimination (PSD), we have developed at PNNL a PSD hardware and software architecture based on commercially-available preamplifiers and transient recorders. In contrast to earlier techniques, this new architecture does not rely on specially-modified preamplifiers, or hardware differentiators. Experimental trials of the new system have been conducted with 2.4-kg HPGe detectors using the proposed preamplifiers and instrumentation. PSD efficacy is based on measurements of a 1593-keV double-escape peak and adjacent 7 rays. These results give projected limits on the half-life of 76 Ge 0i//3/J decay 1.56 times more stringent than those resulting from an experiment without PSD (see Figure 10). Development of the PSD analysis software continues, and we anticipate some improvement over our current results. 4-6
The Super-cryostat
The super-cryostat encloses 21 segmented 76 Ge detectors. This compact configuration insures a small ratio of copper to germanium mass and benefits from self-shielding. Indeed, /?/3-decay events induce a signal in a single detec-
336 Table 2. Theoretical nuclear structure factors F^ and Majorana neutrino mass parameters corresponding to T°J2 = 10 2S yr for 7 6 Ge. FN ( y r ) " 1 1.56 9.67 1.21 1.12 1.41
x x x x x
13
1010~ 1 5 10~13 10~ 1 3 10~ 1 4
Model
(ro«) eV
Weak coupling shell model QRPA QRPA QRPA Shell model
0.41 1.69 0.47 0.48 1.36
Reference 15
16,17 18 19 20
tor while 7 rays from the background tend to scatter among several detectors. All the parts of the cryostat will be electroformed onto stainless steel mandrels from a CUSO4 solution. To reduce the thorium contamination, the CUSO4 will be purified by crystallization. The radium will be removed from the plating bath by introducing a barium scavenge. The entire procedure 14 was successfully developed by the IGEX collaboration. 4-7
Majorana Mass Sensitivity
The total improvement that detector segmentation and PSD will bring to the Majorana Project has been projected. The limits on the half-life of 76 Ge 0i//3(3 decay are projected to be 4.49 times more stringent than an experiment without segmentation or PSD. Table 2 shows the sensitivity of the 76 Ge experiment to the Majorana neutrino mass. The theoretical nuclear structure factor F/v is defined such that: K > = me
I
(1)
n
These theoretical predictions, combined with conservative background estimates, imply that the sensitivity of the proposed 500-kg Majorana Project is about 0.02 eV. 5
Observation of 7 B e Neutrinos with a
100
Mo Detector
The attractive features of 100 Mo for the study of /3/3 decay have been appreciated for a long time. However, the unique properties of this isotope for the study of low-energy solar neutrinos have been widely ignored to this day." In °We wish to thank P. Vogel for a private communication. After this symposium, we learned from J. Bahcall that H. Ejiri had drafted a preprint on the subject.
337
his masterpiece 21 on the topic, John Bahcall does not even list it as one of the possible isotopes for a new generation of detectors. We wish to stress the importance of a 100 Mo detector for 7 Be solar neutrino detection. The Q-value of the reaction ve + 100 Mo -> 1 0 0 Tc + e~ is 168 keV, well below the energy of the 7 Be v (E„ = 861 keV). For background considerations, it is extremely important that the electron is monoenergetic. The matrix element of this transition can be measured (electron capture branch of the 1 0 0 Tc decay) and has been done so although not accurately: log ft — 4.45loio- Certainly, this branching ratio should be remeasured with better accuracy, but the strength of this Gamow-Teller transition is experimentally guaranteed by the observation of similar transitions in this mass area. The nuclear spin sequence, 0 + ->• 1 + , gives the largest possible cross-section for a given NME: a =
a0(PeEeF(Z,Ee)) With (To =
2.6 x 10- 4 1 [21' + 1 cm 2 . ^ 21 + 1
(2)
Using the solar flux prediction of Bahcall 21 , one finds a rate of 60 7 Be v reactions per year per ton of 100 Mo, assuming no oscillation. The final nucleus of this reaction, 1 0 0 Tc, is unstable and beta-decays (see Figure 1) to the ground state of 100 Ru with an half-life time of 15.8 s. Therefore, one can set up a coincidence, in space and time, between the monoenergetic electron and the delayed /? particle. This possibility reminds us of the measurement described in section 2 which demonstrated a near-absence of background even though that particular experiment was located above ground. We believe that the background of the above-ground experiment involving two 7 rays in coincidence should be similar to the one observed in this experiment where one would measure a coincidence between a monoenergetic signal and a continuous P signal ( 0 - 3 MeV) with an apparatus located in an underground facility. The Guernica Project should establish the validity of this assumption. The detection of 7 Be neutrinos without any significant amount of background would make it possible to confirm the MSW Effect and the SMA solution of the long-standing solar neutrino puzzle. Indeed, this solution, which is widely preferred by theory, predicts that all 7 Be neutrino oscillate into v,j,. Dramatically enough, a one ton 100 Mo detector would observe zero events per year in place of about 60 in the case of no oscillation. Since the 7 Be nucleus is a necessary precursor of the 8 B nucleus whose neutrinos have already been observed, no other theory could explain the absence of 7 Be neutrinos. This project shares many similarities with the Majorana Project. The enriched 100 Mo needed for this experiment could also be separated by the
338
Krasnoyarsk centrifuges. In addition, the DOE, Carlsbad Area Office, would strongly support this effort as a flagship for the new US underground facility. 6
Conclusion
At the beginning of his presentation at this symposium, Dr. Rosen stated that a large 0/3 decay experiment was the most important experiment needed today because it is the only kind of measurement that can determine whether the neutrino is a Majorana or a Dirac particle. We certainly agree with this statement and gladly suggest a concrete proposal for this exciting endeavor. Acknowledgements I would like to thank all of the Majorana Project collaborators: W. Tornow and M. Hornish (Duke University and TUNL), A. Young and S. Hoedl (North Carolina State University and TUNL), H. Karwowski, A. Champagne and R. Fitzgerald (University of North Carolina and TUNL), R. Janssens and M. Carpenter (Argonne National Laboratory), V. Brudanin, S. Egorov, S. Sandukovski and O. Kotchetov (JINR, Dubna), A. Barabash, V. Yumatov, V. Ashitkov and V. Stechanov (ITEP, Moscow), J. Webb (New Mexico State University), H. Miley, W. Wilcox, B. Thompson, J. Reeves and R. Brodzinski (PNNL), P. Sangsingkeow (PerkinElmer), C. Aalseth, F. Avignone III and H.A. Farach (University of South Carolina). In addition, I want to thank the Department of Energy, Carlsbad Area Office, and in particular Ines Triay, Beth Bennington and Roger Nelson for allowing us to locate our experiment in their underground facility. Also, I would like to extend our gratitude to Dennis Hofer and Howard Vasquez of the Westinghouse Waste Isolation Division for their help in dealing with the technical issues involved with moving our experiment underground. I thank Jeffrey Hughes from the Office of the DOE Undersecretary for his assistance with the issues concerning the use of the Krasnoyarsk separators for scientific purposes. Finally, I wish to thank Dr. P. Rosen and Timothy Harms (DOE EM) for their continued interest in the Majorana Project. This work was supported in part by the U.S. Department of Energy, Office of High Energy and Nuclear Physics, under grant No. DE-FG02-97ER41033. References 1. J. Schechter and J.W.F. Valle, Phys. Rev. D 25, 2951 (1982).
339
2. S.P. Rosen in Symmetries and Fundamental Interactions in Nuclei, ed. W.C. Haxton and E.M. Henley (World Scientific, Singapore, 1995), p. 251. 3. S.M. Bilenky et al, Phys. Lett. B 465, 193 (1999). 4. D. Blum et al, Phys. Lett. B 275, 506 (1992). 5. A.S. Barabash et al, Phys. Lett. B 345, 408 (1995). 6. A.S. Barabash et al, J. Phys. At. Nucl. 62, 2039 (1999). 7. A. Balysh et al, Phys. Lett. B 322, 176 (1994); A.S. Barabash et al, Z. Phys. A 352, 231 (1995). 8. O. Civitarese and J. Suhonen, Nucl. Phys A 575, 251 (1994). 9. M. Aunola and J. Suhonen, Nucl. Phys A 602, 133 (1996). 10. S. Stoica and I. Mihut, Nucl. Phys A 602, 197 (1996). 11. T. Toivanen and J. Suhonen, Phys. Rev. C 55, 2314 (1997). 12. J. Suhonen and O. Civitarese, Phys. Rep. 300, 123 (1998). 13. A. Faessler and F. Simkovic, J. Phys. G 24, 2139 (1998). 14. C.E. Aalseth et al, Phys. Rev. C 59, 2108 (1999). 15. W.C. Haxton and G.J. Stephenson, Jr., Prog. Part. Nucl. Phys. 12, 409 (1984); W.C. Haxton, Nucl. Phys. B (Proc. Suppl.) 31, 82 (1993). 16. P. Vogel and M.R. Zirnbauer, Phys. Rev. Lett. 57, 3148 (1986); J. Engel, P. Vogel, and M.R. Zirnbauer, Phys. Rev. C 37, 731 (1988). 17. M.K. Moe and P. Vogel, Annu. Rev. Nude. Part. Sci. 44, 247 (1994). 18. O. Civitarese, A. Faessler, and T. Tomoda, Phys. Lett. B 194, 11 (1987);T. Tomoda, Rep. Prog. Phys. 54, 53 (1991) and references therein. 19. K. Muto and H.V. Klapdor, Phys. Lett. B 201, 420 (1988); A. Staudt, K. Muto, and H.V. Klapdor-Kleingrothaus, Europhys. Lett. 13, 31 (1990). 20. E. Caurier et al, Phys. Rev. Lett. 77, 1954 (1996); P.B. Radha et al, Phys. Rev. Lett. 76, 2642 (1996). 21. J. Bahcall, Neutrino Astrophysics, (University Press, Cambridge, 1989).
D O U B L E BETA DECAY OF
100
MO
V . D . A S H I T K O V , A.S. B A R A B A S H , S.G. B E L O G U R O V , S.I. K O N O V A L O V , R . R . S A A K Y A N r V . N . S T E K H A N O V , V.I. U M A T O V Institute
of Theoretical
and Experimental Physics, B. Cheremushkinskaya 117259 Moscow, Russia E-mail:[email protected]
25,
G. C A R U G N O A N D G. P U G L I E R I N Dipartimento
di Fisica e INFN,
Universita di Padova, Padova, Italy
via Marzolo
8,
1-35131
F.MASSERA INFN, Sezione
di Bologna,
40126,
via Berti Pichat,
6/2 (Bologna),
Italy
Using a liquid argon ionization chamber the 2vf3/3 decay of 1 0 0 Mo was detected with its half-life of [6.8 ± l.l(stat) ± lA(syst)] • 10 18 y. The limits on half-lives for Oi/- and Oi/x°-decays of 1 0 0 Mo were estimated as 6.4(3.4) • 10 2 1 y and 2.8(2.2) • 10 2 0 y respectively at 68(90)% CL. Available world d a t a for the 2v(30 decay of 1 0 0 Mo lead to the mean "world" value of the half-life, T1/2 = (7.9 ± 0.7) • 10 18 y, that corresponds to the nuclear matrix element, MQT = 0.119 ± 0.005.
1
Introduction
Intensive research for the neutrinoless double beta decay is due to its connection with fundamental aspects of particle physics (see, for example, reviews 1 2 3 ' ' ). The main interest in this process is certainly concerned with neutrino mass, because if the Qvj3f3 decay were detected then according to the theory the mass of at least one neutrino must be nonzero and this mass is of the Majorana type. At the moment only lower limits on half-lives (T1/2) have been obtained experimentally. These limits are used to deduce upper limits on the Majorana neutrino mass, the right-handed current admixture parameter, the MajoronMajorana neutrino coupling constant etc. However, uncertainties in nuclear matrix elements (NME) calculations do not allow reliable limits to be placed on these fundamental values. In this context the detection of 2v/3/3 decay becomes of particular importance because information on experimental values of NME(2i/) for different nuclei enables a more accurate calculation of both NME(2^) and NME(Of). Besides that, more precise study of the 1v decay •PRESENTLY AT UNIVERSITY COLLEGE LONDON, UK
340
341
mode is interesting from the point of view of a search for a possible time variation of the weak interaction coupling constant 4 ' 5 . The nuclide 100 Mo is one of the most attractive for investigations of /?/? decay. It has a large /?/? transition energy, 3034 keV. In addition the transition 100 Mo(0+J - 100 Ru(0+ s .) is characterized by the highest value of NME for both the 1v decay mode (extracted from experiments, see ref.6, for example) and the OJ' decay mode (as predicted by recent calculations 2 , T ). As it was pointed out in 4 | 5 , 100 Mo is a good candidate for the geochemical 2f/?/? decay experiments. Follow-up comparison of the geochemical results with the results of the direct (counter) experiments will allow a conclusion on the variability of the weak interaction constant to be deduced (it is presented in detail in 4 ' 5 ) . In this connection the problem of measuring the half-life of the 2v(3(3 decay of 100 Mo with a high accuracy assumes great importance. Up to date there are a few positive results for the 2v(i(3 decay of 100 Mo [8-12]. The experimental values of the half-life are inside the interval from [6.75_0 4 J ^ J ± 0 . 6 8 ( s y s i ) ] • 10 1 8 y 1 2 to 11.5±f:g -10 1 8 y 8 . This work presents the results of a new independent detection of the 2i//3/3 decay of 100 Mo using a liquid ionization chamber. Besides that, the mean "world" value of 100 Mo half-life and the corresponding NME for the 2v(3f3 transition are given.
2
Experimental Procedure
The experiment is being carried out in the Gran Sasso Underground Laboratory (3500 m w.e. deep). The experimental setup consists of a liquid argon ionization chamber placed in passive shielding (15cm of lead), a gas system, electronics and a data acquisition system. The active detection portion of the chamber is composed of alternating circular planes of anodes and cathodes with screen grids placed between them. The cathodes are made of molybdenum foil approximately 50 mg/cm 2 thick. The chamber contains 14 cathodes, 15 anodes and 28 screening grids. The chamber has been assembled with eight cathodes containing enriched molybdenum (98.4% 100 Mo) and six cathodes with natural molybdenum (9.6% 100 Mo). Radioactive impurities of the Mo samples are less than 0.015 Bq/kg for 214 Bi, 0.0015 Bq/kg for 208 T1 and 0.04 Bq/kg for 2 3 4 m Pa. Each anode is connected to a charge-sensitive preamplifier, followed by an amplifier and a flash ADC with the 50 ns sampling time. The energy resolution (FWHM) is 6% at an energy of 3 MeV. The trigger for data collection requires that at least one anode signal exceeds the threshold ( 0.8 MeV).
342
Each trigger causes digitized signals from all anodes to be written to a data tape. Data processing is performed off-line. Two-electron events (events with two neighboring anode signals with a time difference of < 0.6//s) are selected. The detection efficiency is calculated by Monte Carlo. The more detailed description of the experimental setup can be found in 13>14>15. 3
Results
The results presented here were obtained with 137.8 g (313 hours of data taking) and 306 g (1355 hours of data taking) of enriched Mo. Fig. 1 shows the energy spectra of two-electron events for enriched (458 kg-h) and natural (303 kg-h) molybdenum. The threshold for the first electron is equal to 0.8 MeV, for second one is 0.5 MeV. Events from natural Mo cathodes are used for background estimation. O^-decay. To reduce the background the energy threshold for each electron of a pair has been selected to be 1 MeV. The energy range (2.8-3.1) MeV has been studied with an additional selection on signal shape. As a result 3 events in the enriched molybdenum and 3 events in the natural Mo (i.e. 4 events if recalculated for 458 kg-h) have been detected. Using the detection efficiency (6.9%) one can obtain the limit on the Of/?/? decay of 100 Mo, Ti/ 2 > 6.4(3.4) -10 21 y at 68%(90%) C.L. 0ux°-decay. The energy interval of 2.3 — 3.0 MeV has been investigated. 785 events for the enriched Mo foils and 558 events for the natural foils (or 843 events if recalculated for 458 kg-h) have been recorded. For the efficiency of 5.7% we have obtained the limit, T 1 / 2 > 2.8(2.2) • 1020 y at 68%(90%) C.L. 2^-decay. Events have been analysed in the energy interval of 1.4 — 2.3 MeV where the signal-to-background ratio is maximal. Background subtraction have led to the final value of the effect, 636 ± 102 events. Using the calculated detection efficiency of the 2*//?/? decay of 100 Mo (2.2%) one gets the half-life: T 1 / 2 = [6.8 ± l.l(stat)
± lA{syst)] • 10 18 y.
The systematic error is mainly due to the possible contributions of radioactive impurities in the foils. 4
Discussion
Table 1 presents all the available positive experimental results on the half-life of 100 Mo. Only the preliminary result of M. Moe et al. 9 is not given because
343
700 600
h
-
500 400
a)
r
300
Counts/0.1 MeV
200 100 0
1111
<)
l
400
rk^j-x,,
2
3
i, 4
i
.. i
J
i
.. I
5
300
i
i...
7
b)
M
350
i
6
250 200 150 100 50 i
0 0
i
1
.I i
2
A.
i . . ' — i
_i_ 3
1 1 1 1 1
4
Energy (MeV)
Figure 1. Energy spectra of two-electron events for (a) enriched and (b) natural molybdenum.
we use their more precise final result 12 . The last line shows the mean of the half-lives of all five experiments. The mean estimate has been performed according to the usual technique of calculation of the mean with different variances summing up the statistical and systematic errors in quadrature. Using the phase factor G = 8.9 • 1 0 - 1 8 y _ 1 (for gA = 1.25) 6 and our mean "world" half-life one can get NME (2v) for the transition of 100 Mo(0+ ) - l o o Ru(0+ s ), M G T = (0.23 ±0.01) MeV" 1 or MGT = (0.119 ±0.005) (scaled by the electron rest mass). The passive antineutron shielding which is going to be installed for the next set of measurements will decrease the background in the energy range of
344 Table 1. Experimental half-lives for the 2t//?/3 decay of
Year, reference
100
Mo.
T&.xlO^y
1991, / 8 /
n-5±l:S
1995, / 1 0 / 1997, / l l /
9.5 • OA(stat) ± 0.9{syst) 7 g+2-2
1997, / 1 2 /
6-75+S:3±0.68(^)
2000, present work
6.8±l.l(stat)±lA(syst)
mean
7.9 ± 0 . 7
the Oi/p/3 decay of 100 Mo and increase sensitivity to Of- and Qvx°-processes up to the levels exceeding the best current results 16 . We would like to thank Dr. J.A. Thomas for an accurate reading of the manuscript and for valuable remarks. References 1. H.V. Klapdor-Kleingrothaus, A. Staudt, Non-accelerator Particle Physics, Bristol: Institute of Physics Publishing, 1995. 2. A. Faessler, F. Simkovic, J. Phys. G 24, 2139 (1998). 3. H.V. Klapdor-Kleingrothaus, in: Proc. 5-th Int. WEIN Symp. Physics Beyond the Standard Model, Eds. P. Herczeg, C M . Hoffman and H.V. Klapdor-Kleingrothaus, World Scientific, 1999, p. 275. 4. A.S. Barabash, JETP Lett, 68, 1 (1998). 5. A.S. Barabash, preprint ITEP 39-99, Moscow, 1999. 6. J. Suhonen and O. Civitarese, Phys. Rep. 300, 123 (1998). 7. F. Simkovic et al., Phys. Rev. C 60, 055502 (1999). 8. H. Ejiri et al., Phys. Lett. B 258, 17 (1991). 9. S.R. Elliot, A.A. Hahn, M.K. Moe, J. Phys. G 17, S145 (1991). 10. D. Dassie et al., Phys. Rev. D 51, 2090 (1995). 11. M. Alston-Garnjost et al., Phys. Rev. C 55, 474 (1997). 12. A. De Silva et al, Phys. Rev. C 56, 2451 (1997). 13. V.D. Ashitkov et al, Phys. At. Nucl.Gl, 910 (1998). 14. V.D. Ashitkov et al, Nucl.Phys. B (Proc. Suppl.) 70, 233 (1999). 15. V.D. Ashitkov et al, Phys. At. Nucl. 62, 2044 (1999). 16. H. Ejiri et al, Nucl. Phys. A 611, 85 (1996).