Hadron and Nuclear Physics with Electromagnetic Probes
Hadron and Nuclear Physics with Electromagnetic Probes
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear
Physics with Electromagnetic
Probes Proceedings of the Second KEK-Tanashi International Symposium Tanashi,Tokyo, October 25-27, 1999
edited by
K. Maruyama Center for Nuclear Study, University ofTokyo Tokyo, Japan
H. Okuno Institute for Nuclear and Particle Physics KEK Tokyo, japan
2000 ELSEVIER Amsterdam - Lausanne - NewYork - Oxford - Shannon - Singapore -Tokyo
ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 21 1, 1000 AE Amsterdam, The Netherlands
02000 Elsevier Science B.V. All rights reserved T h i s w o r k i s protected under copyright
by Elsevier Science, a n d the f o l l o w i n g terms and conditions apply t o its use:
Photocopylng Smplc photocop~rbof \ingle chapter, ,may he !made for perwnal use as allowed by natlonal copyr~ghtlaws Permlcsion o f the Puhlliher and payment of a fee 15 retlu~rcd lor all olher phillrwopy~n~, lncludnrlg r r ~ u l t ~or~ ry\~erna~lc lr cnpylng, copying for advenivng or promollonal porpo\e\. re\alr, and all forms of document dellvery Specla1 rate\ .ire d\a~labletor educa~jonal~nrl~totton* that wl\h ro make photocopies for non-profit cducatlonal classroom u5e.
OX5 IDX, UK: phone: ( 4 4 ) 1865 8.13830, fax ( 4 4 ) 1865 853333. r-nidil. perm~~~~onrOelsev~er.co.uk You may al\o contacr Global Rights dlrectly through Elsev~er'shome page (http://www.clscviernl), Perm~\\~ons nlay he \ought directly from Elwvler Science Global Rlghtr Department. PO Box 800. Oxford b) selecting 'Obtalnlng Perm~rsion\'.
I n thc IISA. urrr, ma) deal pernmr\lon* and mahe payment5 thrnugh the Copyr~gh~ Clearance Center. Inc . 222 Ro\ewuod Dnue. Danvers, MA 01923. USA; phone- (V78) 7SO84X). f i ~ r (9781 750-17-1-1. and ~n ~ h IrI K through Ihe Copyright Llcen\lng Agency Rapld Clearance Serblce (CLARCSI. 90 Tottenham Court Road. London W I P ULP. L K. phone: (4) 201 63 1 5555. fdl; (-1 207 631 5500. Other countries may have a local reprograph~crlghls agellcy for payments. Dcr~vatlreWork* Tdhle, of content, may he reproduced for Internal clrrulatlon. but permlwon of Elsevler Sclrncr 1s requued fnr external resale or dl\tr~but~on of such material. Penn~sstonof the Publisher I\ required for all other dcrlvat~veworks. mclud~ngcompllat~on~ and t n m s i a t t ~ ~ ~ s Elrclrur~lcStorage or l'\age l'crm~ss~onof the Publl'hrr I\ requlred to store or use r l c c t r u n ~ c a l.in) l ~ matecral contamed In t h work, ~ including any chapter or p;ln uf
;I
clldpter
Ehccpr ar autl~ncdaho\c, no pan of this work ma! he reproduced. sored in a re~r~e\nl \y*trm or t~ansmlttedIn any form or by any means. electronic, ntcchanlcal. photocopy~ng.record~npor othrrul\e. wlthout prlor wrltten permlsuon of the Puhllsher. Addreu permlslons reque\t\ to: El\evter Clohal Rlght, Department. at the mall. fax and e-mail ;~ddrr\\ec noted above. Yotjcr Uo r r i p o n \ ~ h l l ~~r l ya\\u~nrdhy the Puhl15hcl-for any q u r y andlor damage to per\on\ or property a\ a matter of producls I ~ i ~ b ~rlepl~gence l~ly. or otherw~he,or frum any uar oi operdllon ul sny rnelhod*. pruducli. ~n\tructlnn\or ~lledrconldlned In the matertal here~n.Bccauie of rapid advance\ In tllr rned~calsclences. In part~cular.lndependenr \er~ficatlono f diagnose\ and drug dosagei \hould he made.
First edition 2000 Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for.
ISBN: 0 444 50539 3 '@(Thepaper used in this publication meets the requirements of ANSUNIS0 239.48-1992 (Permanence of Paper). Printed in The Netherlands.
Preface The second KEK-Tanashi International Symposium on "Hadron and Nuclear Physics with Electromagnetic Probes" was held on the Tanashi campus of KEK, High-Energy Accelerator Research Organization from Oct. 25 to Oct. 27, 1999. 'The new KEK was established in April 1997 by merging three laboratories including Institute for Nuclear Study (INS): the University of Tokyo. INS has a long successful history of hosting international symposia and schools on particle and nuclear physics. The new KEK intended to continue its activity by organizing a series of international symposia. This is the second of the new series. The aim of this Symposium was to discuss recent experimental and theoretical developnients of hadron and nuclear physics, where emphases were placed on the hadron and nucleus studies wit,h electron and photon beams. At KEK-Tanashi, the 1.3-GeV Electron Synchrotron had long been operated for these purposes. In rccent years, the main research areas were photonuclear reactions and meson productions by using the first highduty tagged-photon beam and the TAGX spectrorrieter. Although this field is developing quite rapidly, the synchrotron was closed in 1999 after 37-year operation, and these activities will be continued at new facilities. n'e think it is a good time to discuss the present status and future directions of this field at this occasion. The Symposiuni was attended by 85 physicists and 35 talks were presc11tr:d. This book contains the papers presented in thc scientific program of the Symposium. We would like to thank all the speakers and the participants for their stirnulating talks and active discussions. We are also grateful t,o the members of the international advisory coninlittee and the organizing committee for polishing up the program. We appreciate all the staff members, in particular Mrs. I. Yarnamoto and Dr. K. Niki, for smooth organization. Finally, we would like to thank the Inoue Foundation for Science and the Foundation for Accelerator Science for their financial support.
KEK-Tanashi, March 2000 Editors: Koichi Maruyama and Hideki Okuno
ORGANIZATION International Advisory Committee: L.S. Cardman (Jefferson Laboratory, USA) H. Ejiri (Osaka University, Japan) R. Redwine (MIT, USA) B. Schoch (University of Bonn, Germany), T. Walcher (University of Mainz, Germany) J. Wambach (Julich, Germany) W. Weise (Miinchen, Germany) S . Yamada (KEK, Japan) K. Yazalu (University of Tokyo, Japan) President: S. Sugimoto (KEK-Tanashi) Organizing Committee: Y. Akaishi (KEK-Tanashi) M. Asakawa (Nagoya University) I. Endo (Hiroshima University) H. En'yo (Kyoto University) 0. Hashmoto (Tohoku University) J. Kasagi (Tohoku University) T. Kishimoto (Osaka University) K. Maeda (Tohoku University) K. Maruyama (CNS, University of Tokyo, Co-Chair) T. Motoba (Osaka University of Electric-Communication) H. Okuno (KEK-Tanashi, Co-Chair) T. Oshima (Nagoya University) T.-A. Shibata (Tokyo IT) Y. Sumi (Hiroshima International University) K. Tokushuku (KEK-Tanashi)
Host Institute Institute of Particle and Nuclear Studies, KEK
Sponsors Inoue Foundation Foundation for High-Energy Accelerator Science
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
CONTENTS Preface Opening address S. Sugimoto I. INTRODUCTION TO THE SYMPOSIUM Hadrons and nuclear physics with EM probes - from QCD point of view T. Hatsuda 11. MESONS IN NUCLEAR MEDIUM Vector mesons in medium and dileptons in heavy-ion collisions R. Rapp and J. Wambach p0 mesons in the nucleus
K. Maruyama Nuclear production of 4 meson at KEK H. En'yo, J. Chiba, H. Funahashi, H. Hamagaki, M. leiri, M. Ishino, S. Mihara, T. Mipashita, T. Mumkami, R. Muto, M. Naruki, M. Nomachi, K. Ozawa, O. Sasaki, M. Sekimoto, H. D. Sato, T. Tabaru, K. H. Tanaka, S. Yamuda, S. Kokkaichi and Y . Yoshimura (KEK-PSE325 Collaboration) Lepton pair spectroscopy with HADES at GSI J. Friese (HADES Collaboration) 111. NUCLEON RESONANCES IN NUCLEI AND RELATED TOPICS 61 Sll (1535) resonance in nuclei H. Yamazaki, T. Yorita, T. Kznoshita, T. Okuda, H. Matsui, T. Maruyama, J. Kasagi, T. Suda, K. Itoh, T. Miyakawa, H. Okuno, H. Shimizu, H. Y. Yoshida and T. Kinashi Delta in nuclei excited by hadronic processes J. Chiba The physics of AA exitation G. M. Huber
The neutral pion photoproduction on the proton near threshold Il-T. Cheon and M. T. Jeong Photoneutron cross section measurement on 'Be by means of inverse Compton scattering of laser photons H. Utsunomiya, Y. Yonezau~a,H. Akimune, T. Yamagata, M. Ohta, M. Fujishiro, H. Toyokawa and H. Ohgaki Nuclear disintegration induced by virtual photons at heavy-ion colliders I. A. Pshenichnov
IV. STRANGENESS PHYSICS Kaon photoproduction on nuclei K. Maeda Subthreshold and near threshold K+ meson photoproduction on nuclei E. Ya. Paryev Phenomenological aspects of kaon photoproduction on the nucleon T . Mart, S. Sumowidagdo, C. Bennhold and H. Haberzettl Hyperon polarization in Kaon photoproduction from the deuteron H. Yamamura, K. Miyagawa, T. Mart, C. Bennhold, H. Habeerzettl and W. Glockle Physics of associated strangeness production at ELSA E. Paul Retrospect and prospect of hypernuclear physics 0. Hashimoto Spin-orbit splitting of 13*C H. Kohri, S. Ajirnura, R. E. Chrien, P. M. Eugenio, G. Fkanklin, J. Franz, T. Fukuda, L. Gun, H. Hayakawa, P. Khaustov, T. Kzshzmoto, K . Matsuoka, M. May, S. Minami, Y . Miyake, T. Mori, K. Morikubo, J. Nakano, H. Noumi, H. Outa, K. Paschke, P. Pile, B. Quinn, A. Rusek, E. Saji, A. Sakayuchi, R . Sawafta, Y. Shimizu, M. Sumihama, R . Sutter, T . Tamagawa, H. Tamura, K. Tanida, L. Tang and L. Yuan
V. N-N CORRELATIONS AND FEW-BODY PHYSICS Photodisintegration reactions of 3He and *He at TAGX T . Suda
Photonuclear cross sections of three-nuclcon systcms and the role of threenucleon forces G. Orlun,dini, W. Leidernann, V. D. Efros and E. L. Tomsiak
169
Quasi-deuteron picture for %e and 4He photodisintegration S. Hirenzaki, Y . Umemoto and K . Kume Two-nucleon emission experiments at Mainz P. Grabmayr Quark substructure and isobar effects on deuteron form-factors E. Lomon
VI. NUCLEON STRUCTURE STUDIED BY HIGH-ENERGY ELECTRONS The proton and the photon, who is probing whom in electroproduction? 20 1 A . Levy High-Q2 neutral- and charged-current reactions at HERA K. Nagano Spin structure of the nucleon studied by HERMES Y. Sakemi First double polarization measurements on the way to test the GerasimovDrell-Hearn sum rule W. Meyer (GDH- and A2-Collaboration)
239
VII. NEW FACILITIES Few-body physics at Jefferson Laboratory F. W. Hersman (Hall A and C L A S Collaboration) Experimental test of the K-A relative parity - Use of polarized photon beams at high energies Y. Yamaguchz Nuclear physics experiments with 1.2-GeV STB ring at LNS-Tohoku J. Kasagi
265
Laser electron photon facility at Spring-8 T. Hotta, J. K . Ahn, H.Akimune, Y. Asano, W. C. Chang, S. Date, M. Fujiwara, K. Hicks, K. Imai, T. Iwata, T. Ishikawa, H. Kawai, 2. Y. Kim, T. Kzshimoto, N. Kumagai, S. Makino, T . Matsumura, N. Matsuoka, T. Mibe, M. Miyabe, Y. Miyachi, T . Nakano, M. Nomachi, Y. Ohashi, T. Ooba, H.Ookuma, M. Ooshima, C. Rangacharyulu, A. Sakaguchi, T. Sasaki, D. Seki, H. Shimizu, Y. Sugaya, M. Sumihama,
27 1
T. Tooyarna, H. Toyokawa, A. Wakai, C. W. Wang, S. C. Wang, K Yonehara, T. Yorita and M. Yosoi MUSES project at FUKEN RI beam factory T. Katayama, K. Maruyarna and M. Wakasugi Summary of the symposium H. OAuno Symposium program List of participants
Opening Address
Ladies and Gentlemen: On behalf of the organizers, I would like to express our hearty welcome to all of the participants of the 2nd KEK-Tanashi International Symposium, especially to those from abroad. We are very glad to have you all here, in this Tanashi Campus. Tl-le series of Tanashi Symposium was established in 1998 after the merger of three laboratories including Institute for Nuclear Study of the University of Tokyo, which was commonly known as INS-Tokyo. The INS has a long successful history of hosting a medium-sized international symposium every year on a topics concerning nuclear and particle physics. The new Research Organization, KEK-Tanashi also intends to continue such an activity in organizing a scries of intcmational symposia for the research field. This symposiunl is the 2nd of the new series and the 27th of the previous one. As you know, about 44 years ago Tanashi campus was opened for the first interuniversity research institute for nuclear and particle physics. At, the beginning our Electron Synchrotron was built here as the first high-energy accelerator in Japan. The ES played an essential role in creating high energy physics in this country and provided very important steps toward the birth of KEK-PS and Synchrotron Radiation facilities. Since we are going to move to the Tsukuba site within several months to expand our research activities, we have just closed the ES this summer after 37-year successful operation. The wide research field we have developed a.t the ES will bc taken over by new electron facilities of domestic and foreign laboratories. We think it is a good time to discuss the present status and future directions of t,his field at this occasion. So, the topics of this synlposiunl is focused on Hadron and Nuclear Physics with Electromagnetic Probes.
I hope that this symposium will contribute to further development of this field through active and fruitful discussions. In conclusion, I wish all participants a pleasant stay in Tokyo. Thank you for your attention.
Shojiro Sugimoto President of the 2nd KEK-Tanashi International Symposium
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
I. INTRODUCTION TO THE SYMPOSIUM
ThisPageIntolyLfBk
This Page Intentionally Left Blank
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Hadrons and nuclear physics with EM probes - from QCD point of view "Phys. Dept., Kyoto Univ., Kyoto 606, Japan After a brief introduction of the current problems in QCD, selected topics on the physics of in-medium hadrons are discussed. They are in-medium QCD sum rules, spectral functions in lattice QCD, and critical fluctuation of the chiral order parameter in dense medium. Special emphasis is put on the QCD spectral function which is the key observable in both theoretical and experimental studies. 1. INTRODUCTION
Quantum Chromodynamics (QCD) is the "theory of everything" in the physics of strong interaction. The QCD action can be written as [l]:
where m being the quark mass, G:, is the field strength tensor of the gluon field A:, and D p is the covariant derivative. This action has been well-established by high energy experiments such as the deepinelastic lepton-hadron scattering (DIS) and e+e- annihilation into hadrons. A key ingredient here is the asymptotic freedom of QCD 121 . . which leads to the logarithmic decrease of the QCD running coupling constant CX,(K) as K (renormalization point the typical scale of a given process) increases:
--
where N, and Nfare the number of color and flavor respectively, and AQcD is the QCD scale parameter which takes a value around 200 MeV. Due to the asymptotic freedom, the interaction among quarks and gluons become weak at high energies and the systematic perturbative analyses with appropriate factorization of the hard and the soft part can be ) high chemical potential carried out. Also, at very high temperature ( K -- T >> A Q c ~and (K p >> AQcD), the small running coupling constant as well as the Debye screening allow us to treat the system as non-interacting gas of quarks and gluons in the lowest order [3]. (Even at extremely high T a,nd p, however, the system still may show some interesting phenomena such as the spatial confinement [4]and the color superconductivity 151.1
-
At low energy scale, non-perturbative effects emerge as a result of the strong non-linear gluon self-interactions. Typical examples are the confinement of quarks and gluons and the dynamical breaking of chiral symmetry, both of which are related to the non-trivial QCD vacuum structure. Although it is still difficult to show analytically how these effects appear from the QCD action eq.(l), they are fundamental phenomena for the existence of mesons and baryons, and also of the atomic nuclei. The current physics issues in QCD may be classified into the following three categories: (1) QCD at extreme conditions such as at high temperature or at high baryon density. This is relevant to the physics of ultra-relativistic heavy-ion collisions at CERN and BNL, and also to the physics of neutron star and quark star [6]. (2) Interplay between the hard physics and soft physics in high energy QCD processes. Any physical process is always a combination of the hard scattering part and the soft part: how to factorize them and how to extract information of the non-perturbative effect are theoretical and experimental challenge.
(3) The properties of hadrons in the hotldense medium. Recent experimental developments with hadron-nucleus and nucleus-nucleus collisions give us a new insight into the properties of hadrons inside hot and/or dense matter. It is also a challenging many-body problem in QCD and could give us information on how nuclear matter undergoes a phase transition t o the quark gluon plasma. Since QCD is the theory of everything, we should always start with eq.(l) to attach the above problems. However, there are many ways to represent the same dynamics in different field choices. This is because the quantum amplitude is obtained after the integration over the field variables. For example, consider the full partition function in QCD: =
J
[dQdqdA]exp [iSQco(qlY, A)] .
This can be equivalently rewritten by integrating out the high frequency quarks and gluons above an arbitrary cutoff A, which gives the eq.(5) below. One can then make change of variables from q, Q, A to the hadronic fields (e.g., pions and nucleons), and obtain eq.(6) below.
..I
-
[ d ~ d N . exp L
A
(a,N, . . .; A)] .
Here sZf is the effective action written in terms of the low-energy quarks and gluons, while Sef is the effective action written in terms of the low-energy hadrons. Both effective actions must have explicit A dependence which should be cancelled by the A dependence from the field integration so that the full partition function is independent of A and gives a unique prediction: namely dZ/dA = 0.
-
So far, the argument is exact, but is not practical at all. In the actual application, one sometimes sets A 47r f, = 1.2 GeV (the chiral symmetry breaking scale) and construct a most general S,Hffallowed by symmetry and then determine the parameters in S,Hff by experiments. The chiral perturbation theory is a typical example of this sort [7]. If one has a big parallel computer, one may carry out the integration [dqdqdA] in eq.(5) numerically by taking A large ( E l / a with a being the lattice spacing). This is the lattice QCD numerical simulation which has considerable progress in recent years [8]. QCD sum rules which have great success in reproducing hadronic properties [9] are based on the Wilson's operator product expansion; in a way it makes use of the appropriate window of energy scale where both hadronic description and the quark-gluon description are valid (quark-hadron duality). After this introductory remarks, we will focus on in-medium hadrons in the following and review some recent theoretical developments. 2. IN-MEDIUM HADRONS
The idea of the in-medium hadrons has started in early 80's. In 1982, Pisarski discussed hadron modifications at finite temperature using phenomenological models and pointed out a possibility to observe the mass shift of the pmeson in dilepton measurement [lo]. In 1985, Kunihiro and I have studied the change of the scalar-isoscalar meson "a" near the critical point of chiral phase transition as a typical example of the dynamical critical phenomena in QCD [Ill. Brown and Rho has generalized these ideas and proposed a universal scaling hypothesis of hadron masses, which is, however, not yet proved. Heavy mesons such as J / Q have been also studied in relation to the deconfinement transition at finite T. In 1986, Hashimoto et al. has discussed the mass shift of J / Q and 7, just below the deconfinement transition [13],while Matsui and Satz proposed the disappearance of J / Q as a signature of the formation of the quark-gluon plasma [14]. In go's, it was widely recognized that the spectral functions of hadrons (instead of the ambiguous quantities such as "mass-shift" and "width-broadening") are the relevant quantity to be studied both theoretically and experimentally, although it was already implicit in ref.[11]. New theoretical tools were also developed in 90's such as the in-medium QCD sum rules [15],meson-baryon effective lagrangians [16],quark-meson coupling model [17]and so on. Since the number of references are enormous, I will refer to the Proceedings of the recent Hirschegg meeting on hadrons in dense matter [18]. In the next section, we will first show a direct connection of the spectral function in the vector channel with various experimental observables to illustrate its importance. 3. SPECTRAL FUNCTIONS FOR EM CURRENT
Among the various two point functions in QCD, the time-ordered correlation function of the electro-magnetic current is the most important one from the experimental point of view. We define the imaginary part of this correlation (the spectral function) as
where /target) is a given many-body state and J,, is the electromagnetic current in QCD:
The physical meaning of the spectral function Im(TJJ)ta,,et becomes clear by inserting the complete set between J's. Then the spectral function is a sum of contributions from all the states having non-vanishing matrix element with the state Jltarget). Namely, it probes the excited state of the many-body system which is obtained by acting the external current J onto the target. Now, let us show some example of the direct connection of Im(TJJ)tar,et with observable~. 1. For e+e- annihilation into hadrons (e+e- -+ y* 4 X ) , the cross section is directly proportional to the vacuum matrix element,
where s = q;. > 0 and the "target" is the vacuum. Therefore one is probing the structure of the vacuum through the time-like photon in this experiment. 2. For the deep-inelastic process with proton target (y* is written as
+ N -+ X), the cross section
is the Bjorken variable. Here the "target" is the where Q2 = -q:, < 0 and nucleon. Therefore, we are probing the nucleon structure through the space-like virtual photon.
+
3. For the emission of dileptons from the hot plasma (plasma -+ y* X ) , the emission rate per unit phase-space volume is
with q;, = u2 - q2 > 0. Here the "target" is the thermal distribution of hadrons or quarks and gluons. Thus we are probing the hot plasma through the emission of the time-like photon which eventually decays into dileptons. 4. If one can carry out the following hypothetical experiment ?*(time like) with A being a nucleus, the cross section is written as
which gives an information on the vector meson propagation in matter.
+A
4
X
5. In hadron-hadron collisions associated with a production of time-like photon ( A B + y* X) such as the Drell-Yan process, the cross section is written as
+
+
from which one can probe the the internal structure of the projectile and target and also the production mechanism of the time-like photon. 4. THEORETICAL TOOLS
Now, what kind of theoretical tools can be used to study the spectral functions in QCD? I will pick up two of them in the following; QCD sum rules and lattice QCD. 4.1. QCD sum rules By using the dispersion relation for the two-point function and the operator product expansion at short distance, one can derive a set of sum rules for the spectral function in the medium at finite density and at finite temperature [15]. They have the following general form of the energy weighted sum;
where Cnis the known Wilson coefficients, 0,is the gauge invariant local operators such as
Im(TJJ)pQcDis the known spectral function calculated perturbatively and thus does not depend on the target structure. The left hand side of eq.(14) can be estimated in the medium by the low energy theorem and low temperatureldensity theorem or possibly by the direct lattice QCD simulations. Therefore, one can make some constraints on the exact spectral function Im(TJJ)t,r,,t in the medium through the sum rules eq.(14). This QCD sum rule constraints have been useful for making some predictions of the spectral shift as well as for checking the validity of the effective field theory calculations ~91. 4.2. Lattice QCD The lattice QCD simulations have remarkable progress in recent years for calculating the properties of hadrons as well as the properties of QCD phase transition. In particular, the quenched QCD simulation on the masses of light mesons and baryons agree within 5-10 % with the experimental spectra [20]. However, the lattice QCD had difficulties in accessing the dynamical quantities in the Minkowski space, because measurements on the lattice can only be carried out for discrete points in imaginary time. The analytic continuation from the imaginary time to the real time using the noisy lattice data is highly non-trivial and is even classified as an ill-posed problem. Recently a first attempt to extract spectral functions (SPFs) of hadrons from lattice QCD data by using the maximum entropy method (MEM) has been reported [21]. MEM
is a method which has been successfully applied for similar problems in quantum Monte Carlo simulations in condensed matter physics, and image reconstruction in crystallography and astrophysics [22]. 4.2.1. Basic idea of MEM The Euclidean correlation function D ( r ) of an operator O(T,2)and its spectral decomposition at zero three-momentum read
D(T) =
J (ot(r,Z)O(O,
6))d3x=
1"
K ( r , w)A(w)dw,
(17) where T > 0, w is a real frequency, and A(w) is SPF (or sometimes called the image), which is positive semi-definite. The kernel K ( r , w) is proportional to the Fourier transform of a free boson propagator with mass w: At zero temperature (T = 0) in the continuum limit, K ( r , w) = exp(-TW).
(18)
Monte Carlo simulation provides D(ri) on the discrete set of temporal points 0
< ri/a <
N,. From this data with statistical noise, we need to reconstruct the spectral function A(w) with continuous variable w. This is a typical ill-posed problem, where the number of data is much smaller than the number of degrees of freedom to be reconstructed. This makes the standard likelihood analysis and its variants inapplicable 1231 unless strong assumptions on the spectral shape are made. MEM is a method to circumvent this difficulty through Bayesian statistical inference of the most probable image together with its reliability [22]. MEM is based on the Bayes' theorem: P[XIY] = P[Y IX]P[X]/P[Y], where P[XIY] is the conditional probability of X given Y. The most probable image A(w) for given lattice data D is obtained by maximizing the conditional probability P[AIDH], where H summarizes all the definitions and prior knowledge. The reliability of the obtained result can be checked by the second variation of P[AIDH] with respect to A(w). By the Bayes' theorem, (19)
P[AIDH] E P[DIAH]P[AJH],
where P[DIAH] (P[AIH]) is the likelihood function (the prior probability). For the likelihood function, the standard X 2 is adopted, namely P[D(AH]= Z;'exp(-L) with
ZLis a normalization factor. D(ri) is the lattice data averaged over gauge configurations and DA(ri) is the correlation function defined by the right hand side of (17). C is an N x N covariance matrix of the data with N being the number of temporal points to be used in the MEM analysis. The lattice data have generally strong correlations among different r's, and it is essential to take into account the off-diagonal components of C. Axiomatic construction as well as intuitive "monkey argument" [24] show that, for positive distributions such as SPF, the prior probability can be written with parameters a and m as P [ A ( H a m ]= 2;' exp(aS). Here S is the information entropy,
s=
1"[a(w)
- m(w) - ~
(l:;)]
( wlog ) - dws
Zs is a normalization factor. a is a real and positive parameter and m(w) is a real and positive function called the default model. In the state-of-art MEM [22],the output image Aout is given by a weighted average over A and a:
where A,(w) is obtained by maximizing the " - free-energy"
for a given a . Here we assumed that P[AIDHam] is sharply peaked around A,(w). a dictates the relative weight of the entropy S (which tends to fit A to the default model m) and the likelihood function L (which tends to fit A to the lattice data). Note, however, that a appears only in the intermediate step and is integrated out in the final result. Finding a global maximum of Q in the functional space of A(w), which has typically 1000 degrees of freedom in our case, can be done by utilizing the singular value decomposition (SVD) of the kernel K ( r , w) [25].
4.2.2. MEM with lattice data To apply MEM to actual lattice data, we have done quenched lattice QCD simulations with the plaquette gluon action and the Wilson quark action by the open MILC code with minor modifications [26]. The lattice size is 203 x 24 with P = 6.0, which corresponds to a = 0.0847 fm (a-I = 2.33 GeV), r;, = 0.1571 [27], and the spatial size of the lattice L,a = 1.69 fm. Hopping parameters are chosen to be r; = 0.153, 0.1545, and 0.1557 with NConf= 161 for each r;. For the quark propagator, the Dirichlet (periodic) boundary condition is employed for the temporal (spatial) direction. To calculate the two-point correlation functions, we adopt a point-source at 2 = 0 and a point-sink averaged over the spatial lattice-points. 12 for the Dirichlet boundary condition in the temporal We use data at 1 ri/a direction. We define SPFs for the PS and V channels as
<
<
so that pp,,,(w -+ large) approaches a finite constant as predicted by perturbative QCD. (the upper limit for the w We take Aw = 10 MeV for the w-integration in (17). w, integration) should be comparable to the maximum available momentum on the lattice: Wmas n / a 7.3 GeV.
-
In Fig. 1 (a) and (b), we show the reconstructed images for each K . Here we use the continuum kernel K = exp(-rw). In these figures, we have used m = mow2 with mo = 2.0(0.86) for PS (V) channel motivated by the perturbative estimate of mo We have checked that the result is not sensitive, within the statistical significance of the image, to the variation of mo by factor 5 or 115. The obtained images have a common structure: the low-energy peaks corresponding to .ir and p, and the broad structure in the high-energy region.
Figure 1. Reconstructed image pOut(w)for the PS (a) and V (b) channels. The solid, dashed, and dotted lines are for K = 0.1557, 0.1545, and 0.153, respectively. For the PS (V) channel, mo is taken to be 2.0 (0.86). w,,, is 7.5 GeV in this figure and Fig.2.
1. The mass of the pmeson in the chiral limit extracted from the peaks in Fig.l(b) reads m,a = 0.348(15). This is consistent with m,a = 0.331(22) determined by the asymptotic behavior of D ( T ) on larger lattice [27]. Although our maximum value = 12 marginally covers the asymptotic limit in r, we of the fitting range r,,,/a can extract reasonable masses for 7r and p. The width of 7r and p in Fig.1 is an artifact due to the statistical errors of the lattice data. In fact, in the quenched approximation, there is no room for the pmeson to decay into two pions. 2. As for the second peaks in the PS and V channels, the error analysis shows that their spectral "shape" does not have much statistical significance, although the existence of the non-vanishing spectral strength is significant. Under this reservation, we fit the position of the second peaks and made linear extrapolation to the chiral limit with the results,
These numbers should be compared with the experimental values:
Figure 2. Same with Fig.1 except for the use of the lattice kernel Klat.
One should remark here that, in the standard two-mass fit of D ( 7 ) , the mass of the second resonance is highly sensitive to the lower limit of the fitting range [27]. This is because the contamination from the short distance contributions from T < ?-,in is not under control in such an approach. On the other hand, MEM does not a suffer from this difficulty and can utilize the full information down to ~ , ~ ~=/ 1. Therefore, MEM opens a possibility of systematic study of higher resonances with lattice QCD data.
3. As for the third bumps in Fig.1, the spectral "shape" is shown to be statistically not significant, and they should rather be considered a part of the perturbative continuum instead of a single resonance. Fig.1 also shows that SPF decreases substantially above 6 GeV; MEM automatically detects the existence of the momentum cutoff on the lattice r/a.It is expected that MEM with the data on finer lattices leads to larger ultraviolet cut-offs in the spectra. The height of the asymptotic form of the spectrum at high energy is estimated as N
where Zv is the renormalization factor of the lattice composite operator. Our estimate in eq.(25) is consistent with the high energy part of the spectrum in Fig.l(b) after averaging over w . We made a similar estimate for the PS channel: p,,(w 2: 6GeV) 2.0, which is also consistent with Fig. l ( a ) .
-
In Fig.2(a) and (b), the results using the lattice kernel Klat are shown. Klat is obtained from the free boson propagator on the lattice. It reduces to K = exp(-TW) when a -+ 0. The other parameters and boundary conditions are the same with Fig.l(a,b). The difference of Fig.1 and Fig.2 can be interpreted as a systematic error due to the finiteness of the lattice spacing a. 4.2.3. Error analysis
The statistical significance of the reconstructed image can be studied by the following procedure [22]. Assuming that P[AIDHom] has a Gaussian distribution around the most probable image A, we estimate the error by the covariance of the image,
where is a functional derivative and (.) is an average over a given energy interval. The final error for Amt is obtained by averaging the covariance over o with a weight factor P[alD H m ] . We found small error for the lowest peak, which supports our identification of the peak with p. Although the existence of the non-vanishing spectral strength of the 2nd peak and 3rd bump is statistically significant, their spectral "shape" is either marginal or insignificant. 4.2.4. Summary of SPF in lattice QCD
We have made a first serious attempt to reconstruct SPFs of hadrons from lattice QCD data. We have used MEM, which allows us to study SPFs without making a priori assumption on the spectral shape. The method works well for the mock data and actual lattice data. MEM produces resonance and continuum-like structures in addition to the ground state peaks. The statistical significance of the image can be also analyzed. Better data with finer and larger lattice will produce better images with smaller errors, and our study is a first attempt towards this goal. So far, we have calculated only the spectral functions in the vacuum. Preliminary studies on the lattice at finite temperature are under way and will be reported in the future. 5. PARTIAL CHIRAL RESTORATION IN NUCLEI 5.1. Quark condensate at finite density The direct measure of the partial restoration of chiral symmetry in medium is the chiral condensate ( q q ) . There is indeed an exact theorem in QCD for the behavior of the condensate [28]
where CnN = 45 i 10 MeV is the pion-nucleon sigma term and E(p)/A is the nuclear binding energy per particle with m being the current quark mass. If one makes low density expansion of the right hand, one obtains almost 35 % reduction of the condensate even in the nuclear matter density po = 0.17fm-3. Direct evidence of such partial chiral restoration could be a threshold enhancement of the spectral strength in the scalar-isoscalar channel as was recently pointed out in
ref.1291. A simplified explanation of the idea behind this spectral enhancement is as follows: Suppose we have a typical double-well effective potential of QCD written in terms of the order parameter a = (qq) as
where a is positive in the vacuum but changes sign at the critical point, while b remains positive. The minimum of the effective potential a0 and the curvature at the minimum read
Therefore, as a becomes small, not only the order parameter a0 but also the curvature decrease. This is nothing but the softening of the vibrational mode associated with the 2nd order phase transition [30]. In the real world, the situation is not that simple, since a has a large with decaying into two pions. Nevertheless, there is an interesting possibility that the spectral function just above the two-pion threshold could be enhanced due to the change of ao, which was originally shown in ref.1311 at finite T and later generalized to the case for finite baryon density in ref. [29]. 5.2. Model calculation Let us first describe the general features of the spectral enhancement near the twopion threshold. Consider the propagator of the scalar-isoscalar "a-meson" at rest in the medium :
where m, is the mass of a in the tree-level, and C,(w; p) is the loop corrections in the vacuum as well as in the medium. The corresponding spectral function is given by 1 p, (w) = - - ImD,(w). 'T
Near the two-pion threshold, the imaginary part in the one-loop order reads ImC, a e
( -~2m,)
.J -.43 1-
When chiral symmetry is being restored, m; ("effective mass" of a defined as a zero of the real part of the propagator ReDil(w = m:) = 0) approaches to m,. Therefore, there exists a density p, at which ReDgl(w = 2m,) vanishes even before the complete a-n degeneracy takes place; namely
At this point, the spectral function can be solely represented by the imaginary part of the self-energy;
which shows an enhancement of the spectral function at the 2m, threshold. We remark that this enhancement is generically correlated with the partial restoration of chiral symmetry. To make the argument more quantitative, let us evaluate p,(w) in the SU(2) 1'inear a-model:
where tr is for the flavor index and M = a+i?.?i. Although the model has only a limited number of parameters and is not a precise low energy representation of QCD, we emphasize that it does describe the pion dynamics qualitatively well up to 1GeV as shown by Chan and Haymaker 1331. The coupling constants p2,X and h have been determined in the vacuum to reproduce f, = 93 MeV, m, = 140 MeV as well as the s-wave T-n scattering phase shift in the one-loop order. We parametrize the chiral condensate in nuclear matter (a) as
In the linear density approximation, @(p)= 1 - C-.P (38) Po The plausible value of @(p= po) is 0.7 0.9 [30]. The spectral function together with ReDgl(w) calculated in the linear sigma model are shown in Fig.3 and 4: The characteristic enhancement of the spectral function is seen just above the 2m,. To confirm this threshold enhancement experimentally, measuring 27r0 and 27 in experiments with hadron/photon beams off the heavy nuclear targets are useful. Measuring a 27r0 -, 47 is experimentally feasible 1321, which is free from the p meson background inherent in the T+T- measurement. Measuring a --+ 27 is also interesting because of the small final state interactions, although the branching ratio is small. (One needs also to fight with large background of photons mainly coming from T's.) Nevertheless, if the enhancement is prominent, there is a chance to find the signal. Recently CHAOS collaboration [34] reported the data on the 7r+n* invariant mass distribution M$,* in the reaction A(T+,T+T*)X with the mass number A ranging from 2 to 208: They observed that the yield for M,"+,- near the 2m, threshold is close to zero for A = 2, but increases dramatically with increasing A. They identified that the T+Tpairs in this range of M,"+,- is in the I = J = 0 state. The A dependence of the the invariant mass distribution presented in 1341 near 2m, threshold has a close resemblance to our model calculation in Fig.3 and 4, which suggests that this experiment may already provide a hint about how the partial restoration of chiral symmetry manifest itself at finite density. Further theoretical works are necessary to unlabel the real nature of the threshold enhancement in the CHAOS data and some preliminary studies have been already started 135,181.
-
--+
Figure 3. Spectral function for o and the real part of the inverse propagator for severa1 values of = ( L T ) / with o~ m y k = 550 MeV. In the lower panel, decreases from bottom to top.
Figure 4. Same with Fig.1 for m y k = 750 MeV
6. SUMMARY
QCD is a theory of everything in strong interaction. Although the effective field theory descriptions are always possible at low energies in terms of the effective action S,Hff, one should keep in mind the fundamental relation such as eqs. (4,5,6) which relate the fundamental theory and phenomenology. Among many other observables, the spectral function (SPF) is the simplest but interesting quantity. I can be studied both from the QCD action (QCD sum rules, lattice QCD etc) and from the effective hadronic action. Also, SPFs are either directly or closely related to the experimental data. Namely, SPFs are the observables through which we can directly compare theory and experiment without much intermediate interface. In relation to the physics of chiral symmetry and its restoration in hot/dense matter, SPFs of light vector mesons ( p , w and 4 ) are most relevant since they can be observed by dileptons. SPF of scalar-isoscalar "oncould be also interesting: "a" does not show up clearly in the vacuum because of it huge width decaying into two pions, but it may appear as a soft and narrow collective mode when the chiral symmetry is (partially) restored. In relation to the deconfinement of quarks in hotldense matter, SPFs of heavy mesons such as J/* and Q' are most relevant. Since there are not only numerous recent developments on SPFs in QCD sum rules, in
lattice QCD and in effective field theories, it is quite likely that we may have more solid understanding of the in-medium hadrons in the near future. Experimental inputs and the cooperative works between theorists and experimentalists are, of course, essential for this purpose.
REFERENCES 1. Y. Nambu, in Preludes in Theoretical Physics, in honor of V. F. Weisskopf (North Holland, Amsteldam, 1966). T. Muta, Foundation of Quantum Chromodynamics, (World Scientific, Singapore, 1987). 2. D. J. Gross and F. Wilczek, Phys. Rev. Lett. 30 (1973) 1343. H. D. Politzer, Phys. Rev. Lett. 30 (1973) 1346. 3. J. C. Collins and M. J. Perry, Phys. Rev. Lett. 34 (1975) 1353. 4. C. Borgs, Nucl. Phys. B261 (1985) 455. 5. D. Bailin and A. Love, Phys. Rep. 107 (1984) 325. M. Iwasaki and T. Iwado, Phys. Lett. B350 (1995) 163. M. Alford, K. Rajagopal and F. Wilczek, Phys. Lett. B422 (1998) 247. R. Rapp, T . Schafer, E. V. Shuryak and M. Velkovsky, Phys. Rev. Lett. 81 (1998) 53. See the recent review; T . Schaefer, Color Superconductivity, nucl-th/9911017. 6. Quark Matter '97 Proceedings, Nucl. Phys. A638 (1998) 1. Quark Matter '99 home page: http://www.qm99.to.infn.it/ 7. See e.g. H. Leutwyler, Ann. Phys. 235 (1994) 165. A. Pich, Rept. Prog. Phys. 58 (1995) 563. 8. Lattice 98 Proceedings, Nucl. Phys. B (Proc. Suppl.) 73 (1999). 9. Vacuum Structure and QCD Sum Rules, ed. M. A. Shifman (North-Holland, Amsterdam, 1992). 10. R. D. Pisarski, Phys. Lett. BllO (1982) 155. 11. T . Hatsuda and T. Kunihiro, Phys. Rev. Lett. 55 (1985) 158. 12. G. Brown and M. Rho, Phys. Rev. Lett. 66 (1991) 2720. 13. T. Hashimoto, 0 . Miyamura, K. Hirose, and T. Kanki, Phys. Rev. Lett. 55 (1986) 2123. 14. T. Matsui and H. Satz, Phys. Lett. B178 (1986) 416. 15. T. Hatsuda and S. H. Lee, Phys. Rev. C46 (1992) R34. T . Hatsuda, Y. Koike and S. H. Lee, Nucl. Phys. B394 (1993) 221. S. H. Lee, Phys. Rev. C57 (1998) 927; erratum-ibid. C58 (1998) 3771. 16. C. Gale and J. Kapusta, Nucl. Phys. B357, 65 (1991). M. Asakawa, C.M. KO, P. Levai and X. J. Qiu, Phys. Rev. C46 (1992) 1159. M. Herrmann, B. Riman and W. Norenberg, Z. Phys. A343 (1992) 119. G. Chanfray and P. Schuck, Nucl. Phys. A545 (1992) 271c. 17. K. Saito, K. Tsushima and A. W. Thomas, Phys. Rev. C55 (1997) 2637. 18. Proceedings of Hirschegg 2000 workshop on Hadrons in Dense Matter, to appear. 19. M. Asakawa and C. M. KO, Nucl. Phys. A572 (1994) 732. B. Riman, S. H. Lee, and H. Kim, Nucl. Phys. A653 (1999) 91. F. Klingl and W. Weise, Eur. Phys. J. A4 (1999) 225.
20. 21. 22.
23.
24. 25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35.
S. Leupold and U. Mosel, Phys. Rev. C58 (1998) 2939. A. Hayashigaki, Prog. Theor. Phys. 101 (1999) 923. F. Klingl, S. Kim, S. H. Lee, P. Morath and W. Weise, Phys. Rev. Lett. 82 (1999) 3396; erratum-ibid. 83 (1999) 4224. S. Aoki et al., Phys. Rev. Lett. 84 (2000) 238. Y. Nakahara, M. Asakawa, and T. Hatsuda, Phys. Rev. D60 (Rapid Comm.) (1999) 091503; hep-lat/9909137. N. Wu, The Maximum Entropy Method, (Springer-Verlag, Berlin, 1997). M. Jarrell and J . E. Gubernatis, Phys. Rep. 269 (1996) 133. W. von der Linden, R. Preuss, and W. Hanke, J . Phys. 8 (1996) 3881. R. N.Silver et al., Phys. Rev. Lett. 65 (1990) 496. D. B. Leinweber, Phys. Rev. D 51 (1995) 6369. D. Makovoz and G. A. Miller, Nucl. Phys. B468 (1996) 293. C. Allton and S. Capitani, Nucl. Phys. B526 (1998) 463. Ph. de Forcrand et al., Nucl. Phys. B (Proc. Suppl.) 63A-C (1998) 460. J. Skilling, in Maximum Entropy and Bayesian Methods, ed. J . Skilling (Kluwer, London, 1989), pp.45-52; S. F. Gull, ibid. pp.53-71. R. K. Bryan, Eur. Biophys. J., 18 (1990) 165. The MILC code ver. 5, http://cliodhna.cop.uop.edu/-hetricklmilc . Y. Iwasaki et al., Phys. Rev. D 53 (1996) 6443. E. G. Drukarev and E. M. Levin, Prog. Paet. Nucl. Phys. A556 (1991) 467. T. Hatsuda, T. Kunihiro and H. Shimizu, Phys. Rev. Lett. 82 (1999) 2840. T . Hatsuda and T. Kunihiro, Phys. Rep. 247 (1994) 222. S. Chiku and T . Hatsuda, Phys. Rev. D57 (1998) R6, ibid. D58 (1998) 076001. H. Shimizu, in Proceedings of XV RCNP Osaka International Symposium, Nuclear Physics Frontiers with Electro-weak Probes (FRONTIER 96), (Osaka, March 7-8, 1996), p.161. L. H. Chan and R. W. Haymaker, Phys. Rev. D7 (1973) 402; ibid. D l 0 (1974) 4170. F. Bonutti et al. (CHAOS Collaboration), Phys. Rev. Lett. 77 (1996) 603; Phys. Rev. C60 (1999) 018201. R. Rapp et al., Phys. Rev. C59 (1999) R1237. Z. Aouissat et al., Phys. Rev. C61 (1999) 012202. M. J. Vicente Vacas and E. Oset, Phys. Rev. C60 (1999) 064621.
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
11. MESONS IN NUCLEAR MEDIUM
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
Vector mesons in medium and dileptons in heavy-ion collisions Ralf Rappa * and Jochen Wambachb "Department of Physics and Astronomy, SUNY Stony Brook, New York 11794-3800, USA bInstitut fur Kernphysik, TU Darmstadt, Schlofigartenstr. 9, D-64289 Darmstadt, Germany Theoretical approaches to assess modifications of vector mesons in the medium, as well as their experimental identification via electromagnetic probes, are discussed. Implications for the nature of chiral symmetry restoration in hot/dense matter are outlined and put into context with the axialvector channel. 1. INTRODUCTION
The investigation of hadron properties inside atomic nuclei constitutes one of the traditional research objectives in nuclear physics. However, in terms of the underlying theory of strong interactions QCD) even the description of the nuclear ground state remains elusive so far. Valuable insights can be expected from a careful study of transition regimes between hadronic and quark-gluon degrees of freedom. E.g., in electron-nucleus scattering experiments the corresponding control variable is the momentum transfer, whereas heavy-ion reactions, performed over a wide range of collision energies, aim at compressing and/or heating normal nuclear matter to witness potential phase transitions into a Quark-Gluon Plasma (QGP). Among the key properties of the low-energy sector of strong interactions is the (approximate) chiral symmetry of the QCD Lagrangian and its spontaneous breaking in the vacuum. This is evident from such important phenomena as the build-up of a chiral condensate and constituent quark mass (Mq N 0.4 GeV), or the large mass splitting 0.5 GeV between 'chiral partners' in the hadron spectrum (such as ~ ( 1 4 0 ) of A M a(400 - 1200))p(770)-a1(1260) or N(940)-N*(1535)). It also indicates that medium modifications of hadron properties can be viewed as precursors of chiral symmetry restoration. In this talk the focus will be on the vector (V) and axialvector (A) channels. The former is special in that it directly couples to the electromagnetic current (i.e., real and virtual photons) at which point it becomes 'immune' to (strong) final state interactions thus providing direct experimental access to in-medium properties of vector mesons, e.g., through photoabsorption/-production on nuclei, or dilepton (e+eV,p'p-) spectra in heavy-ion reactions. The key issue is then to relate the medium effects to mechanisms of chiral
-
*research supported in part by the A,-v.-Humboldt Foundation (Feodor-Lynen program) and U.S. Department of Energy under Grant No. DE-FG02-88ER40388.
restoration. This necessitates the simultaneous consideration of the axialvector channel, which, however, largely has to rely on theoretical analyses. This talk is structured as follows: Sect. 2 is devoted to vector-meson properties in nuclear matter, Sect. 3 contains applications to heavy-ion reactions and Sect. 4 finishes with conclusions. A more complete discussion of the presented topics can be found in a recent review [I]. 2. (AXIAL-) VECTOR MESONS IN COLD NUCLEAR MATTER 2.1. Correlators and duality threshold The general quantity that is common to most theoretical approaches is the currentcurrent correlation function which in the (axial-) vector channel is defined by
For simplicity we will concentrate on the (prevailing) isospin I = 1 (isovector) projections 1 j:==, = -(u rp u - d rp d) with = y p , I?$ = y5yp . (2) 2 At sufficiently high invariant mass both correlators can be described by their (identical) perturbative forms which read (up to as corrections)
( M 2 = qi - p). At low invariant masses the vector correlator is accurately saturated by the (hadronic) p spectral function within the Vector Dominance Model (VDM), i.e.,
with a similar relation involving the al meson in the axialvector channel. The spontaneous breaking of chiral symmetry (SBCS) manifests itself in both the difference of the al and p spectral functions as well as the additional pionic piece in ITA (notice that f, is another order parameter of SBCS). In vacuum the transition from the hadronic to the partonic regime ('duality threshold') is characterized by the onset of perturbative QCD around MdualN 1.5 GeV. In the medium, chiral restoration requires the degeneration of V- and A-correlators over the entire mass range. 2.2. Model-independent results: V-A mixing and sum rules In a dilute gas the prevailing medium effect can be computed via low-density expansions. Using soft pion theorems and current algebra Krippa [2] extended an earlier finitetemperature analysis [3] to the finite-density case to obtain
i.e., the leading density effect is a mere 'mixing' of the vacuum correlators IIOfiV. The 'mixing' parameter
(eN: nucleon density) is determined by the 'long-range' part of the T N sigma term, STN= ~ T ~ ~ : ( N ~ T -2 ~ 20( MeV N)
.
(8)
Chanfray et al. 141 pointed out that anNis in fact governed by the well-known nucleonand delta-hole excitations in the pion cloud of the p (or al) meson which have been thoroughly studied within hadronic models to be discussed in the following section. A naive extrapolation of eq. (7) to the chiral restoration point where E = 112, gives Q, E 2 . 5 ~ 0which , is not unreasonable. Nonetheless, as we will see below, realistic models exhibit substantial medium modifications beyond the mixing effect. Similar in spirit, i.e., combining low-density expansions with chiral constraints, is the so-called master formula approach applied in ref. [5]: chiral Ward identities including the effects of explicit breaking are used to express medium corrections to the correlators through empirically inferred TN, p N (or y N , etc.) scattering amplitudes times density. Resummations to all orders in density cannot be performed either in this framework. Model independent relations which are in principle valid to all orders in density are provided by sum rules. Although typically of little predictive power, their evaluation in model calculations can give valuable insights. One example are the well-known QCD sum rules which have been used to analyze vector-meson spectral functions in refs. [6,7]. It has been found, e.g., that the generic decrease of the quark- and gluon-condensates on the right-hand-side (r.h.s.) is compatible with the phenomenological (left-hand) side if either (i) the vector meson masses decrease (together with small resonance widths), or, (b) both width and mass increase (as found in most microscopic models). Another example of sum rules are the ones derived by Weinberg [8], being generalized to the in-medium case in ref. [9]. The first Weinberg sum rule, e.g., connects the pion decay constant to the integrated difference between the V- and A-correlators:
for arbitrary three-momentum q (here, the pionic piece has been explicitly separated out from nA).We will come back to this relation below. 2.3. Hadronic models a n d experimental constraints Among the most spectacular predictions for the behavior of vector mesons in medium is the Brown-Rho Scaling hypothesis [lo]. By imposing QCD scale invariance on a chiral effective Lagrangian at finite density and applying a mean-field approximation it was conjectured that all light hadron masses (with the exception of the symmetry-protected Goldstone bosons) drop with increasing density following an approximately universal sealing law. The scaling also encompasses the pion decay constant (as well as an appropriate power of the quark condensate) and therefore establishes a direct link to chiral symmetry restoration being realized through the vanishing of all light hadron masses.
More conservative approaches reside on many-body techniques to calculate selfenergy contributions to the vector-meson propagators Dv arising from interactions with surrounding matter particles (nucleons). They are computed from gauge-invariant (vectorcurrent conserving) as well as chirally symmetric Lagrangians. The p propagator, e.g., takes the form
for both transverse and longitudinal polarization states (which in matter, where Lorentzinvariance is lost, differ for q > 0). C,, encodes the medium modifications in the pion cloud (through NN-' and AN-' bubbles, so-called 'pisobars') [ll-13,6,14], and C p stems from direct 'rhosobar' excitations of either S-wave (N(1520)N-l, A(1700)N-1, . . .) or P-wave type (AN-', N(1720)N-', . . .) [15-181. The parameters of the interaction vertices (coupling constants and form factor cutoffs) can be estimated from free decay branching ratios of the involved resonances or more comprehensive scattering data (e.g., .irN -+ pN [19], or y N absorption) which determine the low-density properties of the spectral functions. Additional finite-density constraints can be obtained from the analysis of photoabsorption data on nuclei. Invoking the VDM, the total photoabsorption cross section can be readily related to the imaginary part of the in-medium vector-meson selfenergy in the zero mass limit (needed for the coupling to real photons). An example of such a calculation is displayed in Fig. 1 where a reasonable fit to existing data on various nuclei has been achieved. The low-density limit (represented by the long-dashed line in -10
-
0
300
600 900 q, [MeV]
1.200
1500
0.0
vacuum
0.2
0.4
0.6
0.8
1.0
1.2
M [GeV]
Figure 1. Photoabsorption spectra on nuclei Figure 2. Spin-averaged pmeson spectral with (full line) and without (long-dashed function in cold nuclear matter [14,20]. line) higher order medium effects [18]; short-dashed line: C,,, contribution.
-
Fig. 1) cannot reproduce the disappearance of especially the N(1520) as seen in the data. A selfconsistent calculation to all orders in density [I?], however, generates sufficiently large in-medium widths, on the order of 200-300 MeV (resulting in the full line). Fig. 2 shows the final result for the p spectral function [20] which has been subjected to the aforementioned constraints. The apparent strong broadening is consistent with other
~
~
calculations [6,17]. Similar features, albeit less pronounced, emerge within analogous treatments for w and 4 mesons [6]. Let us now return to the question what these findings might imply for chiral restoration. In a recent work by Kim et al. [21]an effective chiral Lagrangian including al-meson degrees of freedom has been constructed. Medium modifications of the latter are introduced by an 'al-sobar' through N(1900)N-I excitations to represent the chiral partner of the N(1520)N-I state. Pertinent (schematic) two-level models have been employed for both the p and al spectral densities which, in turn, have been inserted into the Weinberg sum rule, eq. (9) (supplemented by perturbative high energy continua). The resulting
Figure 3. Pion decay constant at finite density when evaluated through the Weinberg sum rule using schematic two-level models for in-medium p and al spectral functions [21].
-
density-dependence of the pion decay constant, displayed in Fig. 3, exhibits an appreciable decrease of 30% at QN = eO,which bears some sensitivity on the assumed branching ratio of the N(1900) + Nul decay (or N(1900) Nul coupling constant). However, the mechanism is likely to be robust: due to the low-lying pN(1520)N-I and ~ l - N ( 1 9 0 0 ) N - ~ excitations, accompanied by a broadening of the elementary resonance peaks, the p and al spectral densities increasingly overlap, thus reducing f,. 3. ELECTROMAGNETIC OBSERVABLES IN HEAVY-ION REACTIONS
In central collisions of heavy nuclei at (ultra-) relativistic energies (ranging from pl,b=l200 AGeV in current experiments to +=0.2-10 ATeV at RHIC and LHC) hot and dense hadronic matter is created over extended time periods of about 20 fm/c. Local thermal equilibrium is probably reached within the first fm/c, after which the 'fireball' expands and cools until the strong interactions cease ('thermal freezeout') and the particles stream freely to the detector. Electromagnetic radiation (real and virtual photons) is continuously emitted as it decouples from the strongly interacting matter at the point of creation. The thermal production rate of e+e- pairs per unit 4-volume can be expressed through
I0-'
-----
m -free in-medium an
----- free nn
i
in-medium m
0.0
0.2
0.4
0.6
0.8
1.0
-
1.2
M, [GeVl
Figure 4. e+e- production rates in hot hadronic matter from free 7r+7r- (dashed line), in-medium n-+7r- (full line) [20] and O(&) qq annihilation (dashed-dotted line). the electromagnetic current correlation function (summed over all isospin states I=0,1),
( f B : Bose distribution function; a similar expression holds for photons with M + 0). Fig. 4 shows that the medium effects in the p propagator (including interactions with nucleons as well as thermal pions, kaons, etc.) induce a substantial reshaping of the emission rate (full lines) as compared to free n7r annihilation (dashed line) already at rather moderate temperatures and densities (left panel). In fact, under conditions close to the expected phase boundary (right panel) the p resonance is completely 'melted' and the hadronic dilepton production rate is very reminiscent to the one from a perturbative Quark-Gluon Plasma (dashed-dotted lines in Fig. 4) down to rather low invariant masses of -- 0.5 GeV (as corrections to the partonic rate might improve the agreement at still lower masses). It has been suggested [20] to interpret this as a lowering of the in-medium quarkhadron duality threshold as a consequence of the approach towards chiral restoration. The total thermal yield in a heavy-ion reaction is obtained by a space-time integration of eq. (11) over the density-temperature profile for a given collision system, modeled, e.g., within transport [22] or hydrodynamic [23] simulations. At CERN-SpS energies (160200 AGeV) this 'thermal' component is dominant over (or at least competitive with) final state hadron decays (at low M ) and hard initial processes such as Drell-Yan annihilation (at high M) in the invariant mass range M 2 0.2-2 GeV. A consistent description of the measured data [24-271 is possible once hadronic many-body effectsare included [20,28,29], cf. Figs.5 and 6. However, at this point also the dropping mass scenario [lo] is compatible with the data [30] (cf. dashed curve in Fig. 5). Optimistically one may conclude that strongly interacting matter close to the hadronQGP phase boundary has been observed at the CERN-SpS. Other observables such as hadro-chemistry 1311 or J / Q suppression [32] also support this scenario. Nonetheless, further data are essential to substantiate the present status and resolve the open questions.
h P
<
v
central S+Au 200AGeV
1
r
9
g
lo-"
--. 22
,F 1 0 - ~
-5
Pp lo-'
.
2
"lo-"
1
I
Figure 5. Low-mass dilepton spectra at the Figure 6. Direct photon spectra at the CERN-SpS [25];Dashed-dotted line: parti- CERN-SpS: upper limits as extracted from cle decays after freezeout (hadronic 'cock- the WA80 experiment [24] compared to tail'), other lines: cocktail + .ir.ir annihila- in-medium thermal radiation contributions tion as indicated. (full line) [I].
4. CONCLUSIONS
This talk has focused on medium modifications of vector mesons in connection with chiral symmetry restoration in hotldense matter. In accordance with a variety of empirical information hadronic spectral functions are characterized by the appearance of low-lying excitations as well as a broadening of the resonance structures. A schematic treatment of the al meson on similar footings shows that these features encode an approach towards chiral restoration in nuclear matter as signaled by the decrease of the pion decay constant when evaluating the first Weinberg sum rule. The application of these model calculations to electromagnetic observables as measured in recent heavy-ion experiments at the CERN-SpS leads to a reasonable description of the data from 0 to 2 GeV in invariant mass. The structureless in-medium hadronic dilepton production rates resemble perturbative qQ annihilation in the vicinity of the expected phase boundary indicating that chiral restoration might be realized through a reduction of the quark-hadron duality threshold which in vacuum is located around 1.5 GeV. It would also corroborate the interrelation between temperatureldensity and momentum transfer in the transition from hadronic to partonic degrees of freedom. In the near future further dilepton data will be taken by the PHENIX experiment 1341 at RHIC (advancing to a new energy frontier) as well as the precision experiment HADES [33] at GSI. Thus electromagnetic observables can be expected to continue the progress in our understanding of strong interaction physics.
Acknowledgments It is a pleasure to thank G.E. Brown, E.V. Shuryak and H. Sorge for collaboration and many fruitful discussions.
REFERENCES 1. 2. 3. 4.
R. Rapp and J . Wambach, to appear in Adv. Nucl. Phys. (2000))and hep-ph/9909229. B. Krippa, Phys. Lett. B427 (1998) 13. M. Dey, V.L. Eletsky and B. Ioffe, Phys. Lett. B252 (1990) 620. G. Chanfray, J . Delorme, M. Ericson and M. Rosa-Clot, nucl-th/9809007. 5. J.V. Steele, H. Yamagishi and I. Zahed, Phys. Rev. D56 (1997) 5605. 6. F. Klingl, N. Kaiser and W. Weise, Nucl. Phys. A624 (1997) 527. 7. S. Leupold, W. Peters and U. Mosel, Nucl. Phys. A628 (1998) 311. 8. S. Weinberg, Phys. Rev. Lett. 18 (1967) 507. 9. J.I. Kapusta and E.V. Shuryak, Phys. Rev. D49 (1994) 4694. 10. G.E. Brown and M. Rho, Phys. Rev. Lett. 66 (1991) 2720. 11. M. Herrmann, B. Friman and W. Norenberg, Nucl. Phys. A560 (1993) 411. 12. G. Chanfray and P. Schuck, Nucl. Phys. A555 (1993) 329. 13. M. Asakawa, C.M. KO, P. Lkvai and X.J. Qiu, Phys. Rev. C46 (1992) R1159. 14. M. Urban, M. Buballa, R. Rapp and J . Wambach, Nucl. Phys. A641 (1998) 433. 15. B. F'riman and H.J. Pirner, Nucl. Phys. A617 (1997) 496. 16. R. Rapp, G. Chanfray and J. Wambach, Nucl. Phys. A617 (1997) 472. 17. W. Peters et al., Nucl. Phys. A632 (1998) 109. 18. R. Rapp, M. Urban, M. Buballa and J . Wambach, Phys. Lett. B417 (1998) 1. 19. B. Friman, in Proc. of ACTP Workshop on 'Hadron Properties in Medium' (Seoul, Korea, 27.-31.10.97); and nucl-th/9801053. 20. R. Rapp and J . Wambach, Eur. Phys. J. A6 (1999) 415; R. Rapp, Nucl. Phys. A661 (1999) 33c. 21. Y. Kim, R. Rapp, G.E. Brown and M. Rho, nucl-th/9912061. 22. W. Cassing and E.L. Bratkovskaya, Phys. Rep. 308 (1999) 65. 23. J . Sollfrank et al., Phys. Rev. C55 (1997) 392; C.M. Hung and E.V. Shuryak, Phys. Rev. C56 (1997) 453. 24. R. Albrecht et al., WA80 collaboration, Phys. Rev. Lett. 76 (1996) 3506. 25. G. Agakichiev et al., CERES collaboration, Phys. Lett. B422 (1998) 405; B. Lenkeit, Doctoral Thesis, University of Heidelberg, 1998. 26. A.L.S. Angelis et al. (HELIOS-3 collaboration), Eur. Phys. J. C5 (1998) 63. 27. E. Scomparin et al. (NA50 collaboration), J. Phys. G25 (1999) 235; P. Bordalo et al. (NA50 collaboration), Nucl. Phys. A661 (1999) 538c. 28. R. Rapp and E.V. Shuryak, Phys. Lett. B473 (2000) 13. 29. K. Gallmeister, B. Kampfer and O.P. Pavlenko, Phys. Lett. B473 (2000) 20. 30. G.Q. Li, C.M. KO, G.E. Brown and H. Sorge, Nucl. Phys. A611 (1996) 539; G.Q. Li and C. Gale, Phys. Rev. C58 (1998) 2914. 31. P. Braun-Munzinger and J. Stachel, Nucl. Phys. A638 (1998) 3c. 32. H. Satz, Nucl. Phys. A661 (1999) 104c. 33. J . F'riese, these proceedings. 34. see, e.g., the PHENIX homepage a t http://www.phenix,bnI.gov.
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
meson in the nucleus K. Maruyama
"
"Center for Nuclear Study, University of Tokyo Midori-cho 3-2-1, Tanashi, Tokyo, Japan, 188-0002
meson in nuclear Theoretically predicted modification in the mass spectrum of the medium (3He) has been tested by a tagged-photon experiment at Tokyo. The p0 mesons are identified in the coherent photoproduction process on the nucleus. The mass and width of determined by using the Soding model parameterization are consistent with the vacuum values. While, the T + T - invariant mass spectrum observed in the incoherent process requires a modified-mass contribution in addition to the vacuum-mass pO.
1. INTRODUCTION
For the mass of the hadrons in vacuum, there are several types of theoretical calculations such as the constituent-quark models. the bag models. and the Lattice models. The masses and widths predicted by the theories have been examined by observations, and one believes that the experimental values are the true hadron masses and widths. While, a variety of theories predict the modification in the mass spectrum of the vector meson in nuclear medium. These predictions are due to be tested by experimental observations. Although. many theoretical papers of these topics are available at present [1,13], no direct mass measurement is in hand. A direct way is to determine the mass of the hadron which decays in nuclear medium by constructing the invariant mass from the measurements of the decayed particle momenta. A mass-shifted peak structure is the signal of the modification. This is a direct experimental determination of the p0 meson mass in nuclear medium (3He). The p0 meson (the lightest vector meson) is selected from the hadrons as the easiest candidate to examine the prediction. Two talks in these Proceedings [14,15] are excellent experimental and theoretical reviews to the physics topics. I will make a very brief summary of the theoretical status. Then, I will concentrate on a experimental test of the mass shift of the p0 meson in 3He. The measurement was carried out at the Institute for Nuclear Study. University of Tokyo (INS), using the 1-GeV photon beam from the 1.3GeV Electron Synchrotron (ES) by the TAGX (ES134) collaboration, which is consisted of 25 members from 9 institutions of four countries: Canada, Italy, Korea, and Japan.
2. PHYSICS MOTIVATION 2.1. Theoretical predictions The Chiral symmetry between quarks, building blocks of matter, and gluons interactions in Quantum Chromodynamics (QCD) is spontaneously broken in vacuum. The symmetry is restored in the high-energy hadron collisions, where the strong coupling constant (a,) decreases to reach an asymptotic small value. This decrease is predicted by QCD, and is experimentally proved to be correct. Many theories predict the partial restoration of the symmetry also in high-temperature as well as in high-matter-density environments [1,13]. Chiral effective Lagrangian theories and QCD-Sum-Rule calculations predict the change in the vacuum expectation values of the antiquark-quark condensates (Gq) depending on the environmental baryon density. The change is due to partial restoration of the Chiral symmetry breaking in such environment. This dictates to mass shift and/or width shift of the vector mesons such as p, w , and 4 even at the normal nuclear-matter density from the values in vacuum. Mass-shift formulae for vector mesons in nuclear medium are given by some model calculations such as: scaling hypothesis in Chiral effective Lagrangian by Brown-Rho [3], QCD sum rules calculation by Hatsuda-Lee [12], VDM calculation by Asakawa et al. [16], hadron rescattering calculation by Rapp et al. [17]. and the one that includes momentum dependence by Lee [la]. All predictions are summarized in the following expression,
mass in the nucleus, mv the p0 mass in vacuum, p* the matter where mv*is the density of environment, p the normal nuclear-matter density, and a a parameter which describes the magnitude of mass shift. In order to determine a , where it is predicted to be cr = 0.0-0.25, a feasible task for experimentalists at present is to determine it at the normal matter density. 2.2. How t o observe Mass shift of the vector meson in nuclear medium might be observed as effects in spacelike region by finding the modification of the meson propagator. The other, which takes place in time-like region, is to observe the decay of the vector meson in nuclear medium. We take this more direct way to determine the magnitude of the mass shift. One has to assume that it is possible to place a vector meson in the nucleus without losing its entity as a hadronic particle. The vector meson decays into particle pairs (e+e-, .rrfn-, K+Kd etc.). The decay takes place inside the nuclear sphere at a certain probability, and the decayed particles carry out the information of the parent mass in nuclear medium. The probability depends on vector-meson type, the energy, and the size of the nucleus as shown in Figure 1. By observing these particles, one can determine the mass, and compare it with the mass in vacuum. Table 1 lists the properties of the three lightest vector mesons p, w , and 4. In order to get a high decay probability inside the nucleus, the p0 meson with the shortest life time is the best solution. The photon is superior to the hadron as the method to produce the
Table 1 The lightest vector mesons. I)
Mass (MeV/c2) Width (MeV/c2) CT (fm) ree/r rhadrons/r
768 151.5 1.3 4.5~ 1.0
W
05
782 8.4 23.4 7~ 0.90
1019 4.4 44.4 3x 0.84
vector meson because of it weaker interaction to produce the vector meson even in the interior of the nucleus.
EP (GeV)
Figure 1. Decay probabilities of the p0 meson inside nuclei are calculated as a function of the energy. P is the velocity. At around 1 GeV, the difference between 3He and Pb, where the numbers are radii of the nuclei, is smaller than those a t higher energies.
M,,
(MeV/c2)
Figure 2. A schematical T + T - invariant mass spectrum, which is a superposition of the pn mass shape in vaccum and that with mass shift of AM. Broadening of the pn mass shape or a two-peak structure depending A M is the signal of the mass modification.
2.3. E x p e r i m e n t a l design (TAGX options) Bertin and Guichon [19] proposed a method of measurement of the vector-meson (VM) mass inside nuclear medium by using the vector meson photoproduction on nuclear targets. The procedure is: (1) produce a stable VM inside the nucleus by the photon, (2) the VM interacts with nuclear medium, (3) the VM decays inside the nucleus into two particles, (4) detect the particles to reconstruct the effective VM mass. No experiment using this method has been carried out up to now.
We have examined the conditions that interfere a prompt experimental test. They are: (1) to reduce escaping probability of the VM from the nucleus, and to give enough time interval interacting with the nuclear medium, the momentum of the VM should be low, (2) to reduce the distortion due to final state interaction (FSI), detect decayed lepton pairs. and (3) to enhance the decayed signals, the nuclear radius should be large. Decay This fact in addition to the small fractions of the VMs to lepton pairs are small photoproduction cross section requires a high-current, and high-duty factor electron beam producing the photons, which had not been available before the operation of the JLAB electron accelerator. This is the reason why the test has been delayed. Our strategy is to change the conditions at the minimum level (TAGX options): (1) use VM+r+.ir- decays which has a high decay fraction instead of e+e- pairs, and (2) to compensate this, use light nuclei instead of heavy nuclei to reduce FSI of the pions suffering before escaping from the nucleus. The decay probability in Figure 1 shows that the use of light nuclei is not bad option at lower energies. TAGX options are to measure the 3He(y, n+.ir-)reaction just at the p0 meson production threshold. 2.4. What is the signal ?
The VM photoproduced inside the nucleus decays both inside and outside. When the mass shift takes place inside the nucleus with a significant magnitude, the resulting mass spectrum is a superposition of the two mass spectra. Figure 2 shows a schematic such .irf n- invariant mass spectrum. In reality. the mass spectrum is integration over nuclear matter density, momentum space, and production points. 3. EXPERIMENT 3.1. Photon beam
ES supplied a tagged-photon beam of energies up to 1.2 GeV with a duty factor of less than 1 %. This small number has not been appropriate for the coincidence measurement such as the present ,oO-mesonproduction. An idea was proposed to extend the duty factor with minor investment by Yoshida et al. [20], and it was successfully realized in 1987. The tagged-photon beam with more than 10%-duty factor had been available, which was the best in the world in the energy region above 1 GeV. Figure 3 shows the accelerators in the world before the operation of MAMI-I1 and CEBAF. 3.2. TAGX spectrometer
A magnetic spectrometer with a large solid angle of .ir sr has been constructed as a general-purpose detector of photoreaction measurements by using the higher-duty factor beam in 1987. This TAGX spectrometer was used for an experiment that realized the design in the preceding section. It was for the measurements of the r+.ir-photoproduction on the 3He nucleus in the E, region close to the ,oO production threshold. No measurement of p0 photoproduction on any nuclei has been carried out so far in the E, region. The experiment used the tagged-photon beam and the large-acceptance TAGX spectrometer [21] at the 1.3-GeV ES. The photons ( 5 x 1 0 ~l / s in intensity) with energies of 0.80-1.12 GeV ( 5 MeV in the energy resolution) were incident on a cryogenic 3He tar-
CEBAFJ
zc.
4 -
0
3 -
W
-
x
2 -
B
0 ELSA :Bonn)
i
Frascati
.,
'87
A
1
-.~ ---
('
0 Tokyo
>O
--
-
MIT-Bates
o
~1
SPrinp-8
10
I
Duty Factor (%)
MAMl Mainz) ~
ai 100
Figure 3. The ES extended its duty factor from 1% up to 20% in 1987 in prior to other electron accelerators in the world. Those accelerators with 100% duty factor with the energies greater than 1 GeV were in operational several years after this ES operation.
get (0.35 g/cm2 in thickness). Four-momenta of photoproduced charged particles were measured simultaneously with TAGX. The methods were time-of-flight measurements for velocity determination and trajectory measurements in the magnetic field for momentumvector determination. The results of event-by-event analysis of the recorded two-chargedpion events such as the reaction point distribution which shows non-target background events and the mass of the spectrometer particles which show a clear peak of pions were used for identification of the 3He (y,.ir+nP) events [22] . Detection efficiencies of TAGX for the observed events were estimated by a detector simula! ! tion calculation, and they were used for correcting for detector biases and acceptances to obtain the .ir+n- invariant mass (MT,) distribution. 3.3. Calibration
The capability of the TAGX spectrometer as charged particle detector has been examined in several experiments, and the performance was well known on this stage [21]. The mass of the meson is reconstructed to be the invariant mass of two pions. Calibration of TAGX was made by looking for the narrow-width K, which decays into two pions. Taking into account the longer life of K, of cr=2.54 cm than that of pO, a cut is applied in the distance between the n+.ir- vertex from the beam (23cm) to enhance the K, contribution. MT, distribution (Fig.5) shows a clear peak. A gaussian fit t o the mass spectrum gives the peak position to be 500.61.4.6 MeV/c2 and the width 19.9f 5.4 MeV/c2, which are consistent with the K, assumption. We obtain absolute mass accuracy of less than 4 MeV/c2, and Albfn,=20MeV/c2, which are much smaller than the p0 width. 3.4. Data analysis
Data accumulation with TAGX was made at Tokyo in 1994 by the TAGX(ES-134) collaboration. The data were analyzed in Tokyo, and then delivered to collaborators in Regina, where another physics analysis has been performed by Kagarlis et al. [23]. In this paper, the results obtained mainly at Tokyo by Yamashita et al. [22] are presented.
Figure 4. A schematical layout of the TAGX spectrometer.
4. RESULTS AND DISCUSSION 4.1. P r o d u c t i o n mechanisms There are two types of p0 production mechanisms on the nucleus, i.e. the coherent production mechanism that leaves the target nucleus untouched, and the incoherent one that breaks the nucleus. In the former mechanism, the vector dominance model predicts that the photon fluctuates into the vector mesons in the nuclear field. The meson and the nucleus make an elastic diffractive scattering. All the nucleons in the nucleus contribute coherently to produce the meson. While the photon produces the meson also by the quasi-free production on the nucleon. The nucleons contribute incoherently. Neither data is available in the threshold region. p0 photoproduction threshold on the free nucleon is 1080 MeV (at the mass peak of 768 MeV/c2), and it is 870 MeV on the 3He nucleus for coherent production. The results obtained in our highest photon energy bin of 1040-1120 MeV, which is just above the threshold of both production mechanisms, are discussed in this section [22,28]. The invariant mass distribution, which has a broad peak structure, can be a superposition of several n+n- production mechanisms. They are (1) coherent (diffractive) and
300 400 500 600 700 800
M,,
(MeV/c2)
Figure 5. A .ir+n- invariant mass for the events with the vertex away from the beam (see text). A peak at around 500 MeV/c2 is the K,.
Table 2 Mass and width for the T + T - mass spectrum in the coherent process are reproduced by the Soding model. Mass (MeV/c2) Width (MeV/c2) x2 /dof No. of free parameters 747517 168+24 0.98 4
(2) incoherent (quasi-free) p0 meson production, and (3) the .ir'.ir- production in addition to (4) 357 production which is excluded from the discussion in the following by rejecting M, where m, is the pion the events whose missing mass (Mx) is higher than the Am, mass and M is the 3He mass. The rest of the events are categorized into two mechanisms according to their missing mass values: one is for the events which are consistent with the coherent (Mx- M within resolution), and the other is consistent with the incoherent.
+
4.2. Coherent production Figure 6 shows the invariant mass distribution of the T + T - system in the coherent kinematical region. The data may be reproduced by the p-wave Breit-Wigner mass formula,
r the full width of the resonance, where M,, is the invariant mass of the ~ + n system, and Mo the resonance mass. A Breit-Wigner fitting was made. The width obtained of 151121 MeV/c2 is consistent with the table value, but the mass peak of 706%9 MeV/c2 is significantly low (cf. Table 1). This cannot be interpreted to be the mass shift, because the shift should associate with the non-mass-shift peak which comes from the p0 decays outside the nucleus. Possible causes of the apparent mass shift are examined. There are several known massskewing effects such as the pw interference, the Ross-Stodolsky mechanism, rescattering, and the Soding model [25]. We used the Soding model which takes into account the
7
lo ytd->pOd
8
M,,
'
1
y 3 ~ e - > p 03 ~ (TAGX) e -----Ai,z
(MeVlc2)
Figure 6. A T + T - invariant mass distribution obtained in the photon energy range 1040-1120 MeV. Thick dashed curve is a fit with the p-wave Breit-Wigner mass formula to the invariant mass distribution. Thick solid curve is a fit with the Soding model, where the solid is for the vacuum-mass p0 and the dotted for the interference.
Figure 7. Cross sections for coherent production. Present data (solid circle) and the data on the deuteron [24] (solid depensquares). Dashed curve shows dence.
interference between the p0 production with the Drell type
T+T-
production.
ADrellis an interference between the Breit-Wigner and the Drell type background. The result of a fit using this model is shown in Fig. 6. It reproduces the present data with parametrization listed in Table 2. The results are consistent with the values of the p0 meson in vacuum, i.e. cr--0. Therefore, we observed the photoproduced p0 meson in the 3He nucleus at around the threshold for the first time. As for the production cross section, the value integrated over the measured mass range is shown in Fig. 7 , which is slightly larger than the total cross section for p0 production on the deuteron measured by Benz et al. [24]. When one assumes a mass number dependence of A2I3, two sets of the data are consistent each other as seen in Fig. 7. 4.3. Incoherent production Figure 8 shows the invariant mass distribution of the T + T - system determined in the incoherent region. The peak position and the width are different from the coherent distribution. An excess in the mass region 600-700 MeV/c2 is not reproduced reasonably well by neither the p-wave Breit-Wigner mass formula nor the Soding model. The difference might be filled with other mechanisms such as .ir+.ir- production through T A channel or the phase space production. Fittings including one or two of these mechanisms do not
improve the reproducibility. This is because no mechanism enhance the 600-700 MeV/c2 region. Another mechanism which might deform the invariant mass spectrum is proposed by Yamazaki-Akaishi [26]. They estimate the effects of the average potential energy of the target nucleon and the correlation with surrounding nucleons. A substructure in the mass region 500-600 MeV/c2 appears when the matter density is much higher than the normal density, that is not the case in this experiment. Saito-Tsushima-Thomas [27] predicts p0 Inass reduction in the 3He nucleus by using the quark-meson coupling model under the mean field approximation. According to the a meson mass reduction they get 7 2 5 ~ 7 3 2MeV/c2 as the pO meson mass, which is not large enough to explain the present low mass enhancement. A fit with the sum of the free ,oO mass shape and a mass-shifted. whose magnitude is a free parameter to be determined by the fitting, Breit-Wigner formula is tried. The results are shown in Fig. 9 and listed in Table 3. The inclusion of the mass-shifted p0 reduces the fitting X2/dof frorn 8.9 to 2.0. Obtained value is a = 0.15h0.01. The value and its error are subject to change when assumed conditions are not satisfied. (1) The width is also modified, (2) the collisional broadening is large, (3) momentum dependence of mass modification is taken into account. and (4) the Breit-Wigner formula does not work. 40 o Coherent Incoherent -Soedlng
30
0
TAGX ........
20
10 .............................
0 400
600
M,,
800
1000
(MeV/c2)
Figure 8. Invariant mass distributions obtained in the incoherent region (solid) and the coherent region (open). The drop higher above 850 MeV/c2 is due to the acceptance cut. A fitting with the Soding model does not work for incoherent data.
M,,
(MeV/c2)
Figure 9. Incoherent process. A fitting with a sum of the ,oO mass shape, a massshifted shape (first peak), and the phase space contribution is successful in reproducing the present data.
5. CONCLUSIONS Theoretically predicted mass modification of the p0 meson in the 3He nucleus has been tested by a tagged-photon experiment at Tokyo. Coherently produced pO mesons observed have the vaccum mass and the vacuum width, and these facts confirm no mass
Table 3 The n+.ir- invariant mass spectrum in the incoherent process is reproduced by a sum of three contributions: (1) pO,(2) mass-shifted pO,and phase space. Numbers without errors are fixed in the fitting. Width (MeV/c2) Relative pOfraction Mass (hleV/c2) (1) 655&8 151 82&14
modification in the coherent production mechanism. Incoherently produced nfn- mesons have the mass spectrum that is consistent with an assumption that the p0 meson mass reduction by 15 %.
REFERENCES 1. 2. 3. 4. 5. 6.
G.E. Brown, Nucl. Phys. A488 (1988) 689. H. Kurasawa and T . Suzuki, Prog. Theor. Phys. 84 (1990) 1030. G.E. Brown and M. Rho, Phys. Rev. Lett. 66 (1991) 2720. G . Chanfray, Z. Aouissat. P. Schuck, and W. Norenberg, Phys. Lett. B256 (1991) 325. M. Lutz, S. Klimt. and W. Weise, Nucl. Phys. A542 (1992) 521. M. Herrmann, B.L. Riman, and W. Norenberg. Z. Phys. A-Hadrons and Nuclei 343 (1992) 119. 7. M. Herrmann et al., Nucl. Phys. A560 (1993) 411. 8. G.E. Brown, M.Rho, and M. Soyeur. Nucl. Phys. A553 (1993) 7 0 5 ~ . 9. M. Asakawa and C.M. KO, Phys. Rev. C48 (1993) R526. 10. H-C. Jean, J . Piekarewicz, and A.G. Williams. Phys. Rev. C49 (1994) 1981. 11. Y. Koike, Phys. Rev. C51 (1995) 1488. 12. T. Hatsuda and S.H. Lee, Phys. Rev. C46 (1992) R34. 13. T. Hatsuda and T . Kunihiro. Phys. Rep. 247 (1994) 221. 14. T. Hatsuda, in these Proceedings. 15. R. Rapp, in these Proceedings. 16. M. Asakawa, C.M. KO, P. Lkvai, and J. Qju. Phys. Rev. C46 (1992) R1159. 17. R. Rapp et al., Xucl. Phys. A617 (1997) 472. 18. S.H. Lee, Phys. Rev. C57 (1988) 927. 19. P.Y. Bertin and P.A.M. Guichon, Phys. Rev. C42 (1990) 1133. 20. K. Yoshida et al., IEEE Trans. Nucl. Sci. NS-32 (1985) 2688. 21. K. Maruyama et al. (TAGX Collaboration), Nucl. Instrum. Methods A376 (1996) 335. 22. H. Yamashita. Ph. D. Thesis (in Japanese), Tokyo University of Agriculture and Technology (1997). 23. M.A. Kagarlis et al., Phys. Rev. C60 (1999) 025203-1. 24. P. Benz et al., Nucl. Phys. B79 (1974) 10. 25. P. Soding, Phys. Lett. B19 (1966) 702. 26. T . Yamazaki and Y. Akaishi, Phys. Lett. B453 (1999) 1. 27. K. Saito, Phys. Rev. C56 (1997) 566. 28. K. Maruyama, Nucl. Phys. A629 (1998) 351c.
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Nuclear production of
4 meson at KEK
H. En'yo ", J . Chiba b , H. F'unahashia, H. Hamagaki, ', M. Ieirib, M. Ishinoa, S. Miharaa * , T . Miyashitaa, T. Murakamia R. Mutoa, M. Narukia, M. Nomachi d , K. Ozawaa, 0 . Sasakib, M. Sekimotob, H. D. Satoa, T. Tabarua, K. H. Tanakab, S. Yamadaa, S. Yokkaichia and Y. Yoshimuraa t (KEK-PS-E325 collaboration) "Department of Physics, Kyoto University, Kyoto 606-8502, Japan bKEK, 1-1 Oho, Tsukuba, Ibaraki 305-0801 Japan "CNS, University of Tokyo, 3-2-1,Midori-cho, Tanashi-shi, Tokyo, 188-8501, Japan dRCNP, Osaka University,lO-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan
+
We have observed q5 -+ K+K- decays in 12-GeV p A interaction for three different targets, polyethylene, carbon and lead. The experiment E325 is being performed at the primary beam-line E P l B at the KEK Proton Synchrotron. The data from the first physics run in 1997 are reported. The kinematical region of the observed q5 mesons covers 1 to 3 in /371ab where modification of q5 meson in nuclear matter is expected to be visible. The shape of the invariant mass spectra of the observed q5 mesons are consistent with that observed in free space within the statistics which are presently available. The observed mass-number dependence of the production cross section follows A" (a=0.98f 0.10). No significant p, dependence has been observed on the a parameter. 1. INTRODUCTION
In QCD, hadrons are described as a bound state of confined valence quarks which obtained the effective mass due to the spontaneous breaking of the chiral symmetry. This mechnism determines the properties of hadrons in free space, such as the mass and the decay width. If the vacuum structure was changed from that in free space, those properties should also change. The quark condensate (qq) is the order parameter of the chiral symmetry which is subject to change in hot (T# 0) or dense ( p # 0) matter providing a vacuum structure different from that in free space. Although (qq) is not an observable, the mass of vector mesons can provide a good measure of the mass of valence quarks[l]. An experimental observation of in-medium modifications of vector mesons is thus desirable to study the nature of QCD in the non-perturbative region. *Present Address: ICEPP, University of Tokyo, 7-3-1 Hongo, Tokyo 113-0033, Japan +Present Address: Xaxon Cooperation. 1-3-19, Tanimachi, Chu-ou, Osaka, Japan
In hot matter, (qq) as a function of temperature is expected to decrease rapidly to zero around the critical temperature T, where the QCD phase transition will occur and the chiral symmetry will be restored. The observed excess in lepton pairs by the CERN heavy-ion experiments may already hint a precursor of the symmetry restoration[2]. On the other hand, in dense matter, linear decrease of (qq) is expected theoretically [I]. Several experimental efforts, including this work, have been started to investigate effects of the partially-restored chiral symmetry at normal nuclear-matter density.
Table 1 Experiments to study nuclear density effect on vector mesons lace channel status Published KEK(Tanashi)-ES y + A -, p + X (p -+ T + V ) KEK-PS p A -+ 4 X ( 4 -t K+K-/e+e-) Running (this work) y + A - + $ + X ($-+K+K-) Ready to run Spring-8 GSIIHADES Preparation T A -+ w X (w + e+e-) GSI/FRS Ready to run d A d 3He + A*(bound)
+ + +
+ +
Ref. 131 [4] [51 161 171
Table 1 lists those experiments. The present work, KEK-PS E325, is characterized by the simultaneous measurements of the hadronic and the leptonic decay modes of the $. Based on the QCD sum rule, Hatsuda and Lee predicted[l] that the $-meson mass decreases as shown in Figure 1. According to this model, the mass decrease of 4 meson is in the range of 20 to 40 MeV at normal nuclear density po. Figure 2 shows the expected invariant mass spectra of the q5 for beryllium, copper and lead nuclei, according to the predictions of the mass shift in Fig.1. The Woods-Saxon type nuclear density distribution is employed in the Monte Carlo calculation which treats the $ with P Y ~1. The ~ ~dispersion of mass in media, which is experimentally measurable, is neglected in the calculation. In copper case, about 10% of 4 mesons are expected to decay inside the nucleus when any in-medium modification is absent. And in the invariant mass spectra, they make the second peak on the left side of the main peak which consists of the 4 mesons decayed outside the nucleus. It should be noted that in the K + K - channel, the invariant mass spectra could be deformed due to rescattering and modification of kaons in nucleus. Measurements in the ete- channel are thus more desirable. The decay branching ratios are sensitive to the mass shift of 4 meson and kaon because of the small Q-value, 32MeV, in the K + K - channel. Even if the kaon mass shifted and the q5 mass was unchanged, the effect would result in a change of the branching ratios. If the kaon mass were not modified and the $-meson mass decreased more than 32 MeV , the decay to K + K - would be forbidden in nucleus. Recently, Klingl et a1.[8] calculated the in-medium modification of 4 meson being at rest in nucleus. They obtained the mass shift of $ meson being consistent with the results of Hatsuda and Lee, with significant broadening of the width as large as 45 MeV/c2, 10 times larger than in free space. Their calculations was done for the 4 at rest, and the
-
Invariant Mass (GeV)
Figure 1. Hatsuda and Lee's predictions for the density dependence of the mass of $ meson. Dashed lines are the threshold of the K K channels. The two solid lines represent the range of the strange-quark condensate expected from the experimental observation (Y=~(SS)/((UG)(dl)).
Invariant Mass (GeV)
Figure 2. Expected signals for 3 different targets and for the different assumptions of the strangeness condensate as of Fig.1.
+
dispersion, i.e. the width change as a function of the momentum of 4 meson is yet to be worked out. This calculation, however, implies a significant increase of the decay rate in nucleus for $ mesons in flight and a modification of the invariant mass shape in the $ + K+K- channel. Experimentally a detection of 4 mesons at rest is nearly impossible since the daughter kaons decay immediately. Nevert,heless with the slower 4 mesons, the larger decay probability inside a nucleus and the lager nuclear matter effect are expected. It should be noted that the production of 4 meson with a nuclear target is not well studied so far. The cross section of particle productions with a nuclear target normally follows the relation, a(A) = a(1) x Aa, where A is the nuclear mass number. In the production of the J/+ and the Drell-Yan lepton-pairs, a is close to unity in high energy, as measured in the p+p- pairs by Binkley et al. in 300-GeV/c n A interaction[9]. To produce those particles it requires partonic interactions like g g -+ J / + g or q q --+ y*,i.e. the production is supposed to be perturbative and all the nucleons in a nucleus contribute equally to the production. In the production of n,p and w mesons a is almost 2/3, which corresponds to the projective surface area of nucleus. Those particles, consisting of light u(G)and d(d) quarks, are produced in the fragmentation, and the yield is supposed to be governed by the first collision of an incident proton whose mean free path in nucleus is much shorter than a typical nuclear radius. On the production of 4 meson Binkley et al. showed a to be 0.66 f0.03, close to 213, although their identification of the 4 meson was not perfect due to the poor mass resolution. Bailey et al. measured the mass number dependence of the $ production in the K+K- channel in 120-GeV/c p + A
+
+ +
+
interaction and obtained cr = 0.86 f 0.02 [lo]. Aleev et al. measured 4 -+ K + K - decays in 70-GeV n A interaction and obtained cr = 0.81 f 0.06 [ l l ] . All the 4 productions mentioned above were measured in the kinematical region of 0 5 XF I 0.3 and 0 5 p~ 5 1 GeV/c. No measurement of the target mass dependence of the 4 meson production had been performed in the region of XF 5 0 , close to the target rapidity. The simultaneous measurements of 4 -+ K + K - and 4 -+ e f e- channels being planned in the present experiment can cancel the effect of unknown production mechanism by taking the ratio of the A-dependence between the both channels, and the media effect on the produced 4 mesons can be extracted. It is, however, very important and interesting to study of the nuclear dependence of the production itself, to understand how the nature of the $ reveals in nuclear interaction. The experiment E325 has been designed to explore the physics discussed above. The spectrometer newly built for the experiment is located at the primary beam line of KEK 12-GeV Proton Synchrotron. The special emphases are put on the detection of slowly moving $ mesons (Pyla65 2 ) which have a larger probability to decay inside nucleus, and the simultaneous measurements of 4 -+ K + K - and 4 -, e+e- decays in the same apparatus. It should be noted that high quality primary beam on thin targets is needed to suppress the background from y-conversion in the e+e- channel. We started the construction of the spectrometer on June 1996. The engineering run was performed on November 1996, and the first physics run was executed in June 1997, focusing on the 4 -+ K + K - decay channel. The results from the run in 1997 are reported in this manuscript.
+
2. EXPERIMENT E325 AT KEK-PS
Figure 3 and 4 show a schematic layout of the experiment. Beam protons were delivered to three targets ( carbon, lead, and polyethylene ) placed in-line at the center of the spectrometer magnet. The spectrometer has two arms for kaon and electron detection, sharing the magnet and the tracking devices as common. The kaon arms cover from 112" to f54" horizontally and f6" vertically, where the horizontal angle was measured from the beam line and the vertical angle was measured from the horizontal plane. The electron arms have a much larger acceptance covering from f12" to 690" horizontally and f22" vertically. The trajectories and momenta of K+K- pairs and e+e- pairs from $-meson decays were determined with the Cylindrical Drift Chamber (CDC) and the Barrel Drift Chambers (BDC). Kaons were identified with the Time-Of-Flight (TOF) method using the Start Timing Counters (STC) and the Forward TOF Counters (FTOF) in the off-line analysis. The Aerogel cerenkov counters (AC) and the Hodoscope Counters (HC) were used with FTOF for the on-line trigger to select kaons. Electron identification was performed with the segmented Gas ~ e r e n k o vcounters (FrontGC / RearGC) and the lead glass EM Calorimeters (SideEMC / RearEMC). Details of the experiment can be found elsewhere[4]. To evaluate the achieved tracking performance, A -+ pn- decays were studied. In Figure 5, the invariant mass spectrum of pn- pairs are plotted. We required the decay points are more than 20 mm apart from the target. The clear resonance peak of A shows
.
Rear GC
magnet return yoke
Aerogel Cerenkov
T target chamber
I Figure 3. Top view of the E325 setup
near EMC
T"pl-
1
KDC Hodoscope
1 .
Figure 4. Side view of the E325 setup
the mass resolution of 2.2 MeV/c2. By rescaling this mass resolution by taking account of the differences in the Coulomb scattering effects and in the momentum distributions of protons, pions, and kaons, we evaluated the mass resolution of the q!~ + K+K- invariant mass to be 1.2 MeV/c2, which is good enough for the present analysis. The peak value of the Gaussian fit is 1115.11 MeV/c2, which is consistent with the mass of A in the literature[l2], 1115.68 MeV/c2.
Figure 5. The p r - invariant mass spectra. The line is the best fit result of a Gaussian with a linear background. A is seen at 1115.11 MeV/c2 with a resolution of 2.2 MeV/c2.
3. RESULTS AND DISCUSSION
The invariant mass spectra of the K + K - pairs are shown in Figure 6 for three different targets, polyethylene, carbon and lead. Clear resonance peaks corresponding to 4 + K + K - decays are observed. To evaluate the shape of the background, the event-mixing method was applied. We have assumed that the background is due to an uncorrelated pair of particles so that the two tracks from different events should have the same distribution as the background shape. The hatched areas in the Figure 6 are the shape obtained by the event mixing.
pigure 6 . ~h~ invariant K + K - spectra. In (A) we include all the three targets. The spectra (B);(C)and (D) are torresponding to polyethylene, carbon and lead targets respectively. The hatched areas are the backgrounds evaluated by the event-mixing method.
Figure 7. The resonance position (A) and width (B) of the observed 4 mesons. The line is drawn at the values in free space.
The spectra were fit with the sum of the Breit-Wigner function and the background shape. The resonance positions and the widths are plotted in Figure 7. The resonance positions obtained by the present work are consistent with the known value of 1019.4 MeV/c2. The widths of 4 meson from the polyethylene and the carbon target are also consistent with the natural width of 4 meson, 4.43f0.05 MeV/c2[12]. For the lead target case, the width looks larger but the present statistical significance is not enough to conclude the difference. We analyzed the mass number dependence of the production of 4 meson using the standard parametrization, a(A) = a ( A = 1) x A" , where A is the atomic mass number
45
of the target nucleus. The results are plotted in Figure 8 together with the best fit line.
r - u(Pb)/o(C) rctio n-0.98 0.10
*
a=0.97
* 0.13
1
Figure 8. Mass number dependence of the cross section. The solid line is obtained by the ratio of the lead to the carbon data. The dashed line is the best fit over the three points. The data point for hydrogen is obtained by the CH2-C subtraction. The obtained a parameter is plotted in Figure 9 together with other available data[lO, 111 as a function of XF of observed $ mesons. In spite of the large difference in incident energies and in kinematical coverages, the obtained a is consistent with those of other experiments. Table 2 summarized those a parameters. The region of the transverse momentum p~ is higher in the present experiment than in others. One has to consider that the Cronin effect[l3],characterized by the increase of a as a function of p ~ can , be the origin of the large a value obtained in the present experiment. In the upper part of the Figure 10 we plot, the p~ dependence of a parameter of the present measurement, together with the a measurements in the 4.3-GeV/cp+A -+ T++X interactions[l4]. In the pion measurements, the increase of a above p~ = 0.6 is attributed to the Cronin effect. We did not observe any increase of a in the $-meson production in the present data. The origin of the Cronin effect is not well understood so far. One of possible explanations is the change of the production mechanism from non-perturbative interaction to perturbative one. If we assume the cy value is determined with the Q2 involved in the interaction, the plot of p~ dependence of a mentioned above should be re-plotted as a function of transverse mass, m r . as is shown in the lower part of the Figure 10. The a value of the present data, then, is reasonably understood together with the cr increase seen in the pion data.
Table 2 Comparison of the a parameters and the kinematic regions.
+
4 (GeV)
12GeV p Pb/C(present work) 120GeV/c p Ta/Be[lO] 100GeV/c p Be/p[lO]
+ +
5.1 15.1 14.2
a 0.9850.10 0.86k0.02 0.9650.04
PT (GeV/c) XF 0.4< p~ <1.5 -0.8< XF <0.3 O< p~ <1 O< XF <0.3 O< p~ <1 0.11< XF <0.24
"' E
+ +
1 2 GeV 2 + A s x (our dcto) 0 43GeV/cp+A+n+X
02
F . . '02
, , 0 4
. 0 6
I 08
1
~ ' 2
1.4
rn, (GeV/cZ)
Figure 9. The XF dependence of the a: parameter
~ of Figure 10. The p~ and r n dependence the a: parameter.
The fact that the a parameter is close to unity at this energy region fi = 5.1 GeV) is surprising. It is generally believed that incident energy of 12GeV is dissipated rapidly and the nucleons in the back half of the nucleus can not contributed to the particle production. The present results suggests that the production mechanism of 4 mesons is similar to that of J/$ at higher energy, which is visible because the production from the fragmentation is largely suppressed due to the significant heavy mass of the sS and the OZI suppression. The result that the a: is close to unity suggests that the q5 mesons are generated at anywhere in the lead nuclei. And the average pylabof the observed C#J mesons from the lead target is 1.8. The in-medium decay probability can then be deduced with an assumption of in-medium width of d meson. If we take the width 45 MeV from the theoretical
I
calculation of Klingl et a1.[8], The in-medium decay probability becomes as large as 40%, and the resonance peak shape should deformed consequently. The present observation is not consistent with such a picture. As was explained in the Section 1, their calculation was made for the q5 at rest, and the decay width is subject to change as the momentum of the q5 increases. Their calculation should be considered as the maximum broadening of the q5 meson in nuclear media, because the major mechanisms of the broadening is attributed to the K--nucleon interactions in nucleus and the interactions gets weaker as the kaon momentum increases. Further data accumulation and data analysis of the present experiment will enable us to determine the in-medium decay rate and in-medium decay width simultaniousely. 4. CONCLUSION AND OUTLOOK
The present experiment is the first attempt to measure the production and the decay of $ mesons in the kinematical region where the effect of nuclear media should play an important role. The experiment is on going at KEK with protons of 12-GeV incident energy, which is the lowest& amongst the $-meson measurements in the hadron induced nuclear reactions. The data from the first physics run in 1997 were reported. The K + K - invariant mass spectra for three different targets, carbon, polyethylene and lead, have been obtained. The widths of the resonance are 4.9h2.8 MeV for carbon and 11.5f 5.0 for lead, for the 4 mesons with /3yLabin the region from 1 to 3. The widths as well as the resonance positions are consistent with those in free space, and any significant resonance-shape modification, which is expected from the in-medium modification of hadrons, was not seen in the statistical accuracy which is presently available. The mass number dependence of the production cross sections in the $ + K + K channel has been also obtained using carbon and lead targets. The parameter a! is found to be 0.98 f 0.10, very close to unity. This value is consistent with the ones measured at higher incident energies from 70 to 120 GeV using hadron induced nuclear reaction. The present measurement has firstly covered the region of XF < 0 closer to the target rapidity, while other measurements covered only XF > 0. No significant p~ dependence has been observed on the a! parameter. In other words, the steepness of the A-dependence of the $ production at this energy is not likely from the Cronin effect. We have the data being analyzed from the runs in 1998 and 1999. Further data accumulation is scheduled in 2000. The better statistics in the q5 + K + K - decay mode as well as the new information from the $ + e+e- decay mode will soon be available.
REFERENCES
1. T. Hatsuda and S. H. Lee, Phys. Rev. C46 (1992) R24. 2. G. Agakichiev et al., Phys. Lett. B422 (1998) 405. 3. K. Maruyama, proceedings of this symposium. G.J. Lolos et al. (TAGX Collaboration),Phys. Rev. Lett. 80 (1998) 241-244. 4. E325 proposal, http://www.pn.scphys.kyoto-u.ac.jp/- enyo/e325/phi-root.ps.Z). S. Yokkaichi, Memoir of The Faculty of Science, Kyoto University, (doctor thesis), in print.
5. T . Kinashi, K. Takanashi, M. Fujiwara, T. Hotta, T. Nakano, In Hyogo 1997, Exciting physics with new accelerator facilities 55-62. 6. W. Schoen et al.( HADES Collaboration ), Acta Phys. Polon. B27:2959-2963,1996. 7. GSI/SIS proposal S214, Search for bound q- and w- nuclear states using the recoilless (d:3H e ) reaction. 8. F . Klingl, T. Wass: W. Weise, Phys. Lett. B 431 (1998) 254. 9. M.Binkley et al., Phys. Rev. Lett. 37 (1976) 571. 10. C. Daum et al., Z. Phys. C 18 (1983) 1. R. Bailey et al., Z. Phys. C 22 (1984) 125. 11. A. N. Aleev et al., JINR, Dl-90-168(1990), Dubna. H 12. Europ. Phys. J. C3 (1998) 376. 13. J . W. Cronin et al., Phys. Rev. D l 1 (1975) 3105. 14. M.Ono et al., Phy. Lett. B84 (1979) 514.
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
"
Lepton pair spectroscopy with HADES at GSI J. Friesea, for the HADES collaboration "Physik Department E12, Technische Universitat Miinchen, D-85748 Garching, Germany The High Acceptance DiElectron Spectrometer HADES is presently set up at GSI, Darmstadt, for the systematic study of medium modifications of hadron properties. Short lived hadrons will be produced in nuclear matter from normal up to medium baryon densities utilizing pion, proton and heavy ion induced reactions. Electron pair spectroscopy will be used to measure resonance widths and effective masses of vector mesons as well as transition formfactors of neutral mesons and baryon resonances in the low mass region M 5 1 GeV/c2. The spectrometer is designed for an invariant mass resolution AMtn,/Min, 21 1% (g),a large signal to background ratio, and a large acceptance Rpi, 2: 40%. 1. INTRODUCTION
Over the past two decades the properties of strongly interacting particles embedded in hot and dense matter has increasingly attracted both, experimental and theoretical efforts (for recent reviews see [I-31). The availability of relativistic and ultrarelativistic heavy ion beams allows in a unique way to prepare and explore such states of matter in earthbound laboratories. The long range goal of such studies is to verify experimentally a state of matter, in which chiral symmetry, one of the fundamental symmetries in nature, is restored. Under normal conditions this symmetry is broken and the quark condensate < q, q >, the order parameter of the chiral phase transition, has a non-zero vacuum value. Various approaches within the theory of strong interaction (QCD) predict a reduction of the quark condensate with increasing temperatures and particle densities finally leading to at least a partial restoration of that symmetry [4-121. One expected consequence of this phenomenon is the modification of hadron properties (masses, decay widths, etc.) which can already become significant even for hadrons embedded in groundstate nuclear matter. However, the propagation of hadrons in the medium leads to additional modifications in the particle spectral functions and screens the fundamental QCD effects. Experimental verification of these modifications is therefore one of the main guidelines in nowadays heavy ion physics. The light and short lived vector mesons p, w , and q5 are particularly attractive for such investigations, as they may be abundantly produced in relativistic nuclear collisions and decay at least partly within the hot and dense reaction volume. The w meson is of particular interest, since a modification of its narrow decay width [13,14] might be observed already in cold nuclei when the w is produced in T - induced reactions under adjusted
kinematical conditions [15]. Experimentally, the invariant mass reconstruction and the particle spectral functions can be obtained preferably from lepton pair decay spectroscopy, as electrons and myons experience only very little rescattering in the hadronic medium. Up to now, lepton pair spectroscopy has been performed for light and heavy collision systems at incident energies of 1 - 4AGeV [16,17] and at 160 - 200AGeV [18-231. All experiments observe in the low invariant mass range 200 MeV/c2 < M,,, < 600 MeV/c2 a significant excess of yield as compared to current theoretical calculations [24]. However, the limited invariant mass resolution and background separation has prevented so far from unambiguous conclusions with respect to medium modifications. To improve and extend the existing experimental evidence the new High Acceptance Di-Electron Spectrometer HADES [25,26] is presently being set up at GSI, Darmstadt. This second generation experiment aims at in-medium meson spectroscopy in the medium temperature and density domain with unprecedented invariant mass resolution and signal to background discrimination. After a short motivation for our choice of the GSI-SIS energy regime we give a short description of the spectrometer and report the present status of the experiment. Finally we present simulated dilepton spectra for two examples of reactions and summarize the physics programm for the first runs. 2. L E P T O N P A I R S P E C T R O S C O P Y I N T H E SIS E N E R G Y R E G I M E
In central heavy ion collisions of 1 - 2 AGeV incident energy a hadronic fireball is formed, with temperatures up to about T = 50 - 90 MeV and densities ranging up to 3 times normal nuclear matter densities. At this energy density the system is still in a "pretransition" phase, i.e. far beyond the one needed for the deconfinement and chiral phase transition. Nevertheless, these conditions are advantageous for studies of medium modifications, since in contrast to much higher energies, where the fireball expands violently and experiences rapid changes in temperature and density, the formation of the reaction volume lasts for as long as 10 fm/c. This transient state is already accompanied by a significant reduction of the quark condensate down to as much as 21 30% of its vacuum value [27]. Furthermore, only low lying resonances ( A , ..) and light mesons (n, v, p, w ) are produced and propagate in a baryonic environment with small pion content. In this way all background due to pion induced secondary reactions is minimized. Electron pairs are not only produced in purely binary vector meson decays but come also from 3-body no, A, 7, ...- Dalitz decays and from the nucleon-nucleon Bremsstrahlung. To disentangle these continuum type of contributions from the measured spectra and to allow for a sound interpretation of the anticipated experimental findings in heavy ion collisions, additional knowledge on the fundamental electromagnetic properties of all involved hadrons is needed. These properties can be be gained from exclusive measurements of elementary reactions in n- and nucleon induced reactions on cold nuclei. The importance of this aspect may be seen already from the fact that the interpretation of the DLS yield [16] as arising from an enhanced 17- Dalitz decay would be inconsistent with results obtained by the TAPS collaboration for r]- production in C C and Ca Ca collisions [28]. The GSI accelerator facility offers a unique variety of projectile beams ranging from pions to Uranium ions and allows for a systematic study of lepton pair production with various projectile target combinations in a broad rapidity and transverse momenta regime.
+
+
The available projectile energies can be chosen such that certain reactions channels (multi pion, higher resonances, etc.) are either below threshold or are populated less abundantly as compared to higher energies. Consequently, TO -+ y~ and 7 --t yy decays with subsequent pair conversion occur less frequently and hence limit the combinatorial e+ebackground such that a large signal / noise ratio > 1 can be obtained, in particular in the mass region of the p- and w meson. Therefore it is expected that systematic precision measurements with a mass resolution comparable to the natural resonance width of the wmeson (F = 8MeV/c2) might allow to disentangle the underlying QCD phenomena from the background processes and perhaps verify a partial restoration of chiral symmetry. 3. THE HADES SPECTROMETER 3.1. General properties The design of the spectrometer is governed by the high-multiplicity environment of heavy ion collisions, the production cross sections close to threshold, and the small branching ratio for the dielectron decay channel due to the electromagnetic coupling constant. Au collisions will produce an e+eFor example, only one out of lo5 - lo6 central Au pair from i.e. a p meson decay. For this reason, the key features of the new instrument are:
+
A large acceptance
E
40% combined with a rate capability
A resolution for invariant mass reconstruction AMi,,/Mi,,
Y
A signal to background ratio > 1 for invariant masses up to M
E
lo6 s-'
1% ( a ) .
-- 1GeV/c2
High granularity for multi hit capability and redundant track reconstruction. The spectrometer (see left panel in Fig. 1) is azimuthally segmented into 6 sectors and 0 85'. Four planes of Multi-wire Drift Chambers (MDC) tocovers polar angles 18' gether with a 6-coil superconducting magnet form the magnetic spectrometer for charged particle momentum measurement. A RICH (Ring Imaging Cherenkov) counter surrounding the target is used for electron identification [29]. A set of electromagnetic PreShower detectors (at polar angles below 45O) and a Time-of-Flight (TOF) wall provide additional lepton identification and a multiplicity signal to trigger on central collisions. The areal mass density of all detector materials is optimized as to limit the effective radiation length pd/Xo E and hence to keep multiple scattering and combinatorial background from external conversion low.
< <
3.2. The magnet and the "low mass" drift chambers The toroidal magnet is formed by 6 superconducting coils producing a maximum field of 3.5 Tesla with a current density of 120 A/mm2 per coil. For a magnetic field region of 0.5 m length the transverse momentum kick for leptons is less than 100MeV/c. Under these conditions the required momentum resolution can only be obtained by a position resolution of 11 100 p m and minimum multiple scattering in the detector materials. The tracking system consists of 24 trapezoidal multi cell drift chamber modules arranged in a frustum like geometry to form 4 tracking planes of increasing size. Each module
MDC
-
Beam
-.
Figure 1. The new dielectron spectrometer HADES at GSI. Leftpanel: Schematic cross section showing the main components and detectors. The system is designed with a sixfold symmetry around the beam axis. The magnet coils are projected into the plane. Right panel: Upstream view of the present spectrometer setup with the superconducting toroid in the center. Two sectors are equipped with detectors. Four driftchambers and the RICH are mounted inside the magnet. The outer detectors (PreShower and TOF) are pulled back to the service position.
consists of six sense-field wire layers separated by cathode wire layers and oriented at different angles with drift cell sizes increasing from 5 x 5 mm2 (plane 1) to 14 x 10 mm2 (plane 4). The use of 20 pm sense wires, bare aluminum cathode and field wires ( 80 resp. 100 p m ), and a He-iC4Hlo (60140) gas mixture keeps the radiation length and multiple scattering on a tolerable level. The intrinsic position resolution of a full size chamber (plane 2) was measured with 2.1 GeV/c protons and determined to < 70 pm(a) [30] utilizing both, a self tracking method in subsequent cells and external silicon p-strip trackers. This spatial resolution ensures the required precision in momentum measurement of about Aplp 5 1.5% ( a ) . 3.3. Detectors for electron identification The RICH consists of a gaseous decafluorobutan (C4FIo)radiator around a segmented target, a spherical VUV mirror, and a fast gaseous photon detector separated by a CaF2 window from the radiator gas. The Cherenkov photons are reflected and focused by the mirror t o rings of roughly constant diameter (E 5 cm for all polar angles) onto a multiwire proportional chamber (MWPC) operated with pure methane and a solid CsI photon converter covering the cathode pad plane [31]. For the given geometry about 12 to 17
Cherenkov photons will be detected for each incident lepton, corresponding to a figure of merit No = 108. The Time of Flight (TOF) wall is designed to discriminate electrons from pions up to 0.5 GeV/c and from protons up to 2 GeV/c. Simultaneously the device will be used as a multiplicity trigger array for central heavy ion collisions. The whole system consists of altogether 1148 scintillators arranged in an outer (# > 45O, 648 sc.) and an inner (0 < 45'; 500 sc.) wall. The detector is calibrated with a UV nitrogen laser coupled to the scintillators by means of quartz fibres. The start detector will be placed in-beam close to the target and will consist either of a diamond strip detector (for heavy ions, a 2i 50ps) or a scintillating fiber hodoscope (for hadrons). A time resolution loops 5 at,f 5 160ps has been measured with a full size detector [32],corresponding to an electron identification efficiency of 96% in the low momentum region at large polar angles. With ~ has been obtained. Including the modified prototype modules a resolution 0 ~ =072ps contribution from the start detector this would lead to OTOF N 90ps for the whole system. For particles of high momenta emitted at polar angles 19 5 45' the electron - pion separation is improved by an additional electromagnetic PreShower detector. The detector consists of two lead converters enclosed by three layers of wire chambers with pad readout operated in a limited self-quenching streamer mode. A comparison of signals from the preand the two post-converter layers allows for separation of showering electrons and nonshowering hadrons. The granularity (6192 pads per layer) limits the double hit probability to < 2% for central Au Au collisions at 1 AGeV. With a full size detector module a detection efficiency of 82% for single electron showers has been obtained from 800 MeV electron beam measurements. Investigations with proton beams in the momentum range of interest ( 5 2 GeVlc) lead to a hadron / electron misidentification probability of < 10%. Only for lower momenta (200 MeV/c) the hadron / electron misidentification probability reaches li 25%, while the pure electron identification efficiency is still 40%.
+
3.4. Trigger and data aquisition system The trigger system [33]is organised in three levels. The first level trigger is obtained by a fast multiplicity signal from the TOF wall and selects central collisions. From these the second level selects di-electron candidates with a minimum opening angle by matching the positions of recognised rings in the RICH with those in the TOF and PreShower detectors. Finally the third level suppresses background by making use of an online tracking analysis from driftchamber signals. With a beam intensity of lo8 s-l and a 1% interaction target about lo5 s-' events with lepton content are expected. Therefore the signal processing and data acquisition is laid out to accept the signals from all 73,000 detector channels every 10 ,us. The primary data rate of 4 Gbytels is reduced to a maximum of about 4 Mbytels for Au Au collisions by online hit pattern and data analyis performed in hardware image processing units. The data acquisition system [34] is based on a switched ATM network connecting the VME crates for detector readout and the 2nd level trigger processor with the eventbuilder workstation and an optional 3rd level trigger computing stage. Subevent data of each detector are collected for a large number of 2nd level triggers before they are sent asynchronously to the eventbuilder. The system allows for a sustained transfer rate of 14 Mbytels.
+
3.5. Status of installation and first results Although the spectrometer is still in the set-up phase, many components have been already installed (see right panel in Fig. 1) and tested in commissioning runs. The superconducting magnet has reached full field (B,,, = 3.8 T ) at a current of 3500 A and the field mapping measurements showed good agreement (< 1%) with the calculations. The cave, beamline, and mainframe assembly is finished and two of the six sectors are meanwhile equipped with detectors including the new readout electronics, part of the trigger system, and the ATM network for data acquisition. The integration of the remaining detectors and components is presently under way and, after further commissioning, first physics runs are planned for autumn 2000.
+
Ti at E = 1.75 Figure 2. Response of the RICH in central heavy ion collisions Ar AGeV. Left panel: Ring image measured for one electron or positron from external pair conversion in the target or radiator gas. A coincidence track signal is found in the drift chamber. Note the low background. Right panel: Distribution of ring centers found in one of the six sectors of the RICH for about 2 . lo5 central collisions.
The two sector setup has been operated successfully in two commissioning runs with heavy ion reactions C+C at E = 1.2 AGeV and Ar Ti at E = 1.75 AGeV. All detectors showed stable operation at beam intensities up to lo7 s-' and a performance close to or even beyond the design values. The available trigger modules and the ATM data acquisition system worked reliably and allowed for first data taking. The segmented (8 strips) CVD diamond detector used as in-beam T O F start detector withstood beam intensities up to lo8 s-l with a time resolution a = 50 ps. For the T O F detectors typical time resolutions of a = 100 - 150 ps were obtained. The drift chambers showed routinely a noise level of < 40mV (peak-to-peak) while signals from minimum ionizing particles saturate the amplifiers at 11 400 mV. The PreShower detector was read out successfully utilizing the new image processing unit. The electron identification capability
+
+
of the RICH is illustrated in Fig. 2. In a preliminary analysis of 2 . lo5 central Ar Ti collisions at E = 1.75 AGeV about 2500 candidates for Cherenkov rings were found, in reasonable agreement with simulation results. The corresponding electrons and positrons come mostly from external pair conversion of y-s in the target and the radiator gas. In almost all events the ring candidates (left panel of Fig. 2 ) were not obscured by electronic noise or signals from charged particle background. The extracted ring centers (right panel of Fig. 2 ) show a rather homogeneous distribution declining with increasing polar angle as expected from simulations. A more quantitative analysis of the available data is presently under way. 4. HIGH QUALITY INVARIANT MASS SPECTRA
The expected quality of dielectron spectra has been studied by detailed Monte Carlo simulations [35,36] anticipating the performance of detectors and trigger as discussed in the preceding chapters. The following two examples may give an impression of the capabilities of HADES and the perspectives for meson spectroscopy. 4.1. Exclusive q- Dalitz decay studies in p p collisions The low mass (100 M e v / c 2 < Mt,, < 500 M e V / c 2 ) range of dielectron spectra in nuclear collisions is characterised by comparatively large production cross sections and is dominated by contributions from nucleon - nucleon Bremsstrahlung, the no- and 7-Dalitz decays and - less pronounced - the A- Dalitz decay. In p p reactions the electron background from secondary sources can be kept at minimum. Since the excellent time resolution of the TOF wall allows also for n - p discrimination. both, the leptons and the protons can be identified and detected simultaneously. For these events the reconstruction of a Dalitz decay is kinematically complete without requiring an additional photon detector. Applying a cut on the corresponding particle mass in the p p missing mass distribution suppresses the contribution from background sources considerably. Measurements at incident energies below and slightly above the qproduction threshold will allow to study the continuum distributions above the no- mass. The production of q- mesons in p p reactions has been studied extensively [37-391 and the inclusive cross sections are well known. The two published measurements of the qDalitz decay are based on a sample of 80 counts [40]and on a sample of 2366 counts [41], the latter corresponding to a hypothetical background free yield of 106 signal pairs. For the A- Dalitz decay experimental cross sections may be deduced from the DLS data at energies below the q- production threshold Epp < 1.27GeV [17]. In the simulation the q production is assumed to be mediated by the N*(1535) resonance : pp -+ pN*(1535) -+ ppq -+ ppe+e-(7). For the Dalitz decays the y goes undetected. Considerable background of electron pairs will come mostly from pp --+ ppno -+ p p y y with subsequent photon conversion and the no- Dalitz decay pp -+ ppno -+ pp,l)ye+e-. In the q mass region also pp -+ pp2n0 -+ pp7e+eV ye+e- reactions will contribute. The expected electron pair invariant mass distribution for the 7 Dalitz decay is shown in Fig. 3. The combinatorial background from external two photon conversion (no-+ y y ) is not included but is estimated to be small. Once measured, the observed electron pair spectrum for the 7-Dalitz decay branch can be used to calibrate the spectrometer acceptance, to check the contributions from other sources (A- Dalitz, combinatorial background, etc.), and to
-.
___-_L-
Au+Au 1 AGeV
I
10
1
2n0 M, cut 3
0
,.
-
0 '
~-
,
O.?
.I
,, 0.3
. . ! . .~,-.-A Me+e-[G~VC-'1 0 4
e'e- Mass
0.5
0.6
0.7
(cev)
-
Figure 3. Simulated invariant mass spectra of e+e- pairs from light and heavy collision systems. Left panel: Yield of e+e- pairs from lo6 7- Dalitz decays (line) expected for cx 100h of pp collisions at E, = 2 GeV. The small background from pp27ro (dotted) reactions is further reduced to the dashed curve by requiring the missing mass of the Au collisions ppq- reaction. Right panel: Relative yield of e+e- pairs per central Au at E, = 1AGeV. The contributions from the various sources are indicated. Medium modifications are not taken into account. Note the low combinatorial background at high invariant masses.
+
deduce the electromagnetic formfactor of the 7 meson to a better precision
+
4.2. Electron positron pairs in Au Au collisions Central collisions of the heaviest ions are expected to produce the largest space-time volume of compressed nuclear matter. Simultaneously they are the most challenging systems to be investigated. The invariant mass spectrum of lepton pairs produced in central Au Au collisions at E = 1AGeV was obtained from a GEANT simulation and is shown in the right panel of Fig. 3. The events have been generated in a BUU transport code [42]without taking any medium modifications into account. The lepton pairs originate from decaying sources with Maxwellian momentum distribution in a fireball of about 65 MeV temperature. The different contributions from continuum sources (pnBremsstrahlung and Dalitz decays) and from purely 2 particle resonance decays form a cocktail which shows a significant structure in the region around the masses of free w and 4 mesons. It becomes obvious that the good suppression of combinatorial background and the high mass resolution should allow to detect any mass shift or resonance broadening, in particular of the w. A detailed and quantitative analysis will be carried out by systematic investigations as function of transverse momentum and rapidity and by further exclusive
+
meas~irementsof the Dalitz decay channels. The latter cxperiments are planned for the future utilizing coir~cidcricemeasurements with the photon spectrometer TAPS and/or the GSI large area neutron detector LAND. 5. CONCLUSIONS AND OUTLOOK
HADES is a high rate, high acceptance apparatus for electron pair spectroscopy with < 1% invariant mass resolution and a fast and selective multi level trigger scheme. Presently, two out of six sectors of the spectrometer are completed and first commissior~ingruns have been performed. Results obtained with this setup as well a,s with full size detector prototypes bcfore dcmonstratc the feasibility of the cxpcriment and in particular the efficient and redundant lepton identification. The spectrorrleter will be completed in 2001, with first physics runs beginning a,lrcatly in 2000. Thc physics program is broad and includes the study of lepton pair emission in relativistic hcavy ion collisions, dielect,ron productiori in elemcntary reactions and experiments aimed at studying the structure of hadrons. Particular attention will be paid to the decay properties of vector mesons with respect to rnass shifts and resonance broadening. The shape of thc invariant mass spectrum will give access to electromagnetic transition form fact,ors of neutral mesons. The long-range aim is to better describe the dynamical propcrties of hot and comprcssed hadronic matter, an environmerlt in which partial restoration of chiral symmetry is expected to show up and information about the origin of hadron masses may be obtained. The experimental program of HADES will focus first on pp collisions and light heavy ion systems to verify the yield enhancements found in the DLS measurements. When finally the designed invariant resolution becon~esavailable, dedica,ted investigations of thc w meson decay are foreseen utilizing both, pion and heavy ion induced reactions. The series of experinients planned will exploit the full range of primary and secondary bearns available at GSI, including pions and heavy ions up to uranium. Future combinations with an clectromagnetic calorimeter and/or a neutron detector for missing mass measurements will allow more exclusive measuremerits arid - for a long tirne - guarant,ec substantial scientific output.
ACKNOWLEDGEMENTS The collaboration gratefully acknowledges the support by EC (HTM, TMR program), ASCR, BMBF, DAAD, DFG, DLR (WTZ), GSI, IN2P3, INFN, INTAS, NATO, as well as the local support in all collaborating institutions. REFERENCES C.M. KO, V. Koch and G.Q. Li, Annu. Rev. Nucl. Part. Sci. 47 (1997) 505 W. Cassing and E.L. Bratkovskaya, Phys. Rep. 308 (1999) 65 R. Rapp and J. Wambach, hep-ph/9909229 G.E. Brown and M. Rho, Phys. Rev. Lett. 66 (1991) 2720 T. Hatsuda and S.H. Lee, Phys. Rev. C 46 (1992) R34 6. M. Herrmann, B. Friman and W. Norenberg, Nucl. Phys. A560 (1993) 411 7 . C.M. Shakin and W.-D. Sun, Phys. Rev. C49 (1994) 1185 8. J. Engels, F. Karsch and K. Redlich, Nucl. Phys. B435 (1995) 295 9. K . Saito. k. Tsushima and A.W. Thomas, Nucl. Phys. A609 (1996) 339 10. F. Klingl and W. Weise, Nucl. Pllys. A606 (1996) 329
1. 2. 3. 4. 5.
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. 38. 39. 40. 41. 42.
Y. Koike and A. Hayashigaki, Prog. Theor. Phys. 98 (1997) 631 R . Rapp, G. Chanfray and 3. Wambach, Nucl. Phys. A617 (1997) 472 F . Klingl and W. Weise, hep-ph/9802211 M. Effenberger et al., Phys.Rev. C60 (1999) W. Schon et al., Acta Phys. Pol. B27 Vol. 11 (1996) R . J . Porter et al., Phys. Rev. Lett. 79 (1997) 1229 W.K. Wilson et al., Phys. Rev. C57(4) (1998) 1865 G. Agakichiev et al., Phys. Rev. Lett. 75 (1995)1272 T h . Ullrich e t al.. Nucl. Phys. A610 (1996) 317c, and A. Drees, Nucl. Phys. A610 (1996) 5 3 6 ~ G. Agakichiev et al., Phys. Lett. B422 (1998) 405 T . Akesson et al., Z. Phys. C68 (1995) 47 M.Masera: Nucl. Phys. A590 (1995) 93c M.C. Abreu e t al.. Nucl. Phys. A590 (1995) 117c E.L. Bratkovskaya, W . Cassing, R. Rapp, J. Wambach, Nucl. Phys. A634 (1998) 168 HADES, Technical Proposal: GSI 1994 J . Friese et al., Nucl. Phys. A 654, (1999) 1017c B. Friman, W. Norenberg and V.D. Toneev, Eur. Phys. J . A3 (1998) 165 R . Holzmann et al., Phys. Rev. C56 (1997) R2920 K. Zeitelhack et al., Nucl. Instr. Meth. A 433 (1999) 201 C. Garabatos et al., Nucl. Instr. Meth., A 412 (1998) 38 J . Friese et al., Nucl. Instr. Meth. A 433 (1999) 86 Agodi et al., IEEE Trans. Nucl. Sci., 45, No.3 (1998) 665 M. Petri et al., Proc. ofXXXVIInt. Wint. Meet. Nucl.Phys., Bormio, Italy (1998), 622 M. Miinch et al., Proc. ofXXXVIIInt. Wint. Meet. Nucl.Phys., Bormio, Italy (1999), 509 R. Schicker et al., Nucl. Instr. Meth., A 380 (1996) 586 H. Neumann et al., Acta Phys. Slov. 44 (1994) 195 H. Calkn et al., Phys. Lett. B366 (1996) 39 A.M. Bergdolt et al., Phys. Rev. D48 (1993) R2969 E . Chiavassa et al., Phys. Lett. B322 (1994) 270 M.R. Jane et al., Phys. Lett. 59B (1975) 103 and erratum Phys. Lett. 73B (1978) 503 G. Agakichiev et al., Eur. Phys. J. C4 (1998) 231 G. Wolf et al., Nucl. Phys. A517 (1990) 614
111. NUCLEON RESONANCES IN NUCLEI AND RELATED TOPICS
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Sll (1535) resonance in nuclei H. Yamazakia * , T . Yoritaa t : T . Kinoshitaa, T . Okudaa, H. Matsuia, T . Maruyamaa J. Kasagia, T . sudab 5 , K. Itohc, T. Miyakawac, H. Okunod, H. Shimizue t H. Y. Yoshidae and T . Kinashie
,
"Laboratory of Nuclear Science, Tohoku University, Sendai 982-0826, Japan bDepartment of Physics, Tohoku University, Sendai 980-8577, Japan 'Department of Applied Physics, Tokyo University of Agriculture and Technology, Koganei, Tokyo 184-8588, Japan dKEK-Tanashi, Midori-cho, Tanashi, Tokyo 188-0002, Japan eDepartment of Physics, Yamagata University, Yamagata 990-8560, Japan In order to study the S11(1535) resonance in the nuclear medium, total cross sections of the (7,Q ) reactions on C, A1 and Cu have been measured for photon energies between 680 and 1000 MeV. A broad resonance structure due to the excitation and decay of the ,311 resonance in the nucleus has been clearly observed for the first time. The apparent energy and width of the resonance are 900 and 300 MeV, respectively. Model calculations based on the quantum molecular dynamics (QMD) have been performed. The Q photoproduction on nuclei can be basically understood as the quasi-free reaction with the reasonable medium modifications. 1. INTRODUCTION
The property . of nucleon resonance in nuclear medium is one of the most important issues in the nuclear physics. Recent measurements of the total photoabsorption cross section have shown clear evidence of nuclear medium effects on nucleon resonances /1,2]. . . The DI3(1520) and FI5(l68O) resonances are not observed in the photoabsorption cross section on nuclei, while the A resonance is clearly seen. The behavior of the A resonance which corresponds up to 500 MeV of incident photon energy can be described well by the A-hole model. For higher energy region from 600 MeV to 1000 MeV, proposed was an interpretation based on the collision broadening of the resonance. However, this attempt needs very large broadening, which are hardly justified in terms of NN* -+ NN cross section. Hirata et al. [3] proposed that disappearance of nucleon resonance is caused by the cooperative effect of the collision broadening of A and N*, the pion distortion *Corresponding author e-mail
[email protected],Tel +81 22 743 3433, Fax +81 22 743 3401. +Present address Research Center for Nuclear Study, Osaka University, Ibaraki 567-0047, Japan %Presentaddress College of Bioresource Sciences, Nihon University, Fujisawa, 252-8510, Japan §Present address RIKEN, Wako 351-0198, Japan
and interference in the two-pion photoproduction processes in the nuclear medium. Their calculation can reproduce the experimental disappearance of D13 and F15 by using modest collision broadening of about 80 MeV in width. Thus. measurements other than the total absorption are highly desirable. In the (7,T ) reactions, however, it is hard to determine the resonance properties in nuclei, because the D13 and FI5 resonances mainly decay to T-N channel and are overlapping with the tail of the A resonance strongly. We have aimed at the Sll(1535) resonance, which strongly couples with IKq and decays to the Nq channel with the branching ratio of 45 55 %. There is no nucleon resonance which has such a large branching ratio to 7-N decay around this energy region. Because of this unique property, one can excite only the Sll resonance by using ( y , q ) reactions below 1.0 GeV of incident photon energy. Thus we have measured the total (y, 7) cross sections on C, A1 and Cu.
Figure 1. Schematic drawing of the experimental setup. Tagged photon beam bombarded targets and produced the q mesons. The charged particles accompanied with the photons were removed by the sweep magnet which was placed at 1 m upstream of the target. Two decaying photons from the 77 mesons were detected by the two pure CsI calorimeters.
2. EXPERIMENT
The experiment was performed using the 72 beam line in the 1.3 GeV electron synchrotron of KEK-Tanashi (former INS). The schematic drawing of the experimental setup is shown in Figure. 1. The 72 beam line supplied a tagged photon beam with its intensity of 9 x 105/sec for 680 5 E, 5 1000 MeV. The tagging efficiency was measured once a day. and its typical value was about 70 80 %. Tagged photons bombarded the C, A1 and Cu targets and produced 7 mesons. which were identified by the traditional invariant mass analysis of two decaying photons. Photons were detected by the two sets of pure CsI calorimeters with plastic scintillators for charged particle veto. which were placed in front of the calorimeters. Each calorimeter consists of 29 pure CsI crystals with 13.5 radiation length in thickness. The energy of the incident photon was deduced from the sum of the energies deposited in the 7 crystals. The energy weighted average of the center position of the 7 crystals was regarded as the incident position of the photon. The energy resolution of the pure CsI scintillator was 3 % for photons of 1 GeV and the angular resolution of the photon emission direction was about 5". The invariant mass resolution of two photons is about 6 % at 7 mass region, which is consistent with the energy and angular resolution ) were measured for of the calorimeters. Differential cross sections of the ( 7 , ~ reaction the polar angles from 0" to 90" in the laboratory system with several opening angles of two CsI calorimeters. Detection efficiency was deduced from the GEANT3 Monte Carlo simulation [4]. The total cross section of q photoproduciton on nuclei was deduced by integrating the differential cross section.
-
3. RESULTS
In Figure. 2, the total q photoproduction cross sections are shown as a function of incident photon energy. Open circles, closed circles and squares correspond to the total cross section of the (7.7) reaction on C, A1 and Cu, respectively. It should be noticed that almost whole structure of the S11(1535) photoexcitation in nuclei has been shown for the first time. The reaction cross sections on C, A1 and Cu have similar dependence on incident photon energy. The cross section shows the maximum value at around 900 MeV of photon energy; this value is much higher than expected in the simple quasi free picture. The width of the resonance structure is about 300 MeV, which is also wider than expected. In order to clarify the cause of the phenomena, we have performed the calculation based on the semi-classical model, Quantum Molecular Dynamics(QMD). QMD was originally developed to describe the high energy heavy ion reactions [6]. We have modified the QMD code to describe the photon induced 7 production. At the first step, one nucleon absorbs a photon and becomes an SI1 resonance. Two set of the resonance parameter of the Sll(1535) resonance were used in the QMD calculations. One parameter set is 1540 MeV for the resonance energy, 150 MeV for the width at the resonance pole and 0.55 for the branching ratio of the 7-N channel. These values were determined so as to reproduce the H(y, 7) cross section in reference [7-91. The other is 1544 MeV for the resonance energy, 212 MeV for the width at the resonance pole, which were obtained by B. Krusche et al. by using recent measurement of the H(y, q) cross section at Maintz and Bonn [lo]. The results of the two parameterizations are plotted in Figure 3 with the experimental data. A solid line corresponds to the one with r = 212 and a dotted line with r = 150.
Figure 2. Total cross sections of ( y , q ) reactions on C, A1 and Cu with the QMD calculations (solid line) for the resonance width of 212 MeV, and calculation (dotted line) on C for the width of 150 MeV. Open circles, closed circles and squares shows the results on C, A1 and Cu, respectively.
The results of the QMD calculation are plotted in Figure. 2. In the calculations, the Fermi motion, the Pauli blocking and the 7 absorption in a nucleus have been considered; the 7 absorption cross section was deduced from the detailed balance analysis of the i7-p -+ qn cross section. The collision broadening of SI1(1535) in a nucleus have been also considered; the cross section given in ref. [ll]is used. A dotted line is the result on C with the resonance width of 150 MeV. Solid lines are the results on C, A1 and Cu with 212 MeV. As can be seen in Figure 2, the former parameterization underestimates the measured cross section on C target by about 20 p b above 800 MeV of incident photon energy, but the latter can reproduce the experimental results well on all targets. 4. SUMMARY AND OUTLOOK
The present measurements show that the 7 photoproduciton on nuclei can be basically understood as the quasi-free reaction with the reasonable nuclear medium modifications. They include the Fermi motion, the Pauli blocking, the 7 absorption in nuclei and the N-N* collision broadening. It was found that the resonance parameter with F = 212 MeV
V
B. Krusche et al. M. Wilhelm S. Homma et al.
Figure 3. Total cross section of H(y, 7). Two parameterizations are shown with a solid, dotted line for the resonance width of 212 and 150 MeV, respectively.
can reproduce the experimental data well. Recently Armstrong et al. [12] reported precise measurements of S1l resonance parameters based on the p(e,e'p)q experiment at TJNAF; b, = 0.55, M R = 1532 and FR = 154 f 20 MeV. As shown in Figure. 2, this parameterization cannot reproduce present data with the quasi-free picture. More precise measurements of the elementary processes are highly desirable to reduce the uncertainty of the resonance parameters of s11.
REFERENCES 1. N. Bianchi et al., Phys. Lett. B299 (1993) 219; ibid., B309 (1993) 5 2. M.Anghinolfi et al., Phys. Rev. C47 (1993) R922 3. M. Hirata et al., Phys. Rev. Lett. 80(1998)5068 4. CERN Program Library Long Writeup W5013: Geant Detector Description and Simulation Tool, 1993. 5. M. Robig-Landau et al., Phys. Lett. B373 (1996) 45 6. K. Niita et al., Phys. Rev. C52 (1995) 2620 7. B. Krusche et al., Phys. Rev. Lett.74 (1995) 3736 8. M. Wilhelm, Ph.D. Thesis, Bonn, BN-IR-93-43; B.H. Schoch, Prog. Part. Nucl. Phys. Vol. 34 (1995) 43
9. S.Homma et al., J. Phys. Soc. Jpn. 57 (1988) 828 10. B. Krushe et al., Phys. Lett. B397 (1997) 171 11. M. Effenberger et al., Nucl. Phys. A613 (1997) 353 12. C.S. Armstrong et al., Phys. Rev. D60 (1999) 052004
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
"
Delta in nuclei excited by hadronic processes Junsei Chibaa "IPNS, The High Energy Accelerator Research Organization (KEK) 1-1 Oho, Tsukuba, Ibaraki, Japan 305-0801 The A's in nuclei excited by proton induced reaction is presented. The experiment was carried out at the National Laboratory for High Energy Physics (KEK) about 10 years ago, but the data are still highest quality in this field. Charged particles emitted in the central region were measured with a coincidence of a forward going neutron in (p, n) reactions at 1.5 GeV/c in the A excitation region. The A peak in the neutron spectrum in coincidence with p.ir+shows no shift, as in the (p,p'AO) reactions at 4 GeV/c. On the other hand, the peak in the nuetron spectra in coincidence with other particles, especially with a T + , shifts toward higher momentum. The coincidence data set stringent constraints on theoretical models introduced to interpret the peak shift in inclusive spectra. 1. INTRODUCTION
It is one of the most interesting questions in nuclear physics whether basic properties of particles are modified in nuclear matter compare with those in free space. The A is the first excited state of a nucleon and plays an important role in the dynamics of the excitation of nuclear matter. In 1980's, a shift of the peak position in the momentum spectra of outgoing particles (or in the excitation energy of the A produced in nuclei) in the ( 3 H e ,t ) reactions were reported [I]. It is very interesting and exciting if the shift is caused by a change of the resonance mass of the A, because it would be the first evidence that particles are modified in nuclear medium. However, magnitude of the peak shift depends on incident energy [2] which implies that the peak shift could not be attributed to a simple modification of mass of A in nulcear matter. Such peak shift, though its magnutude is very small, had been seen also in (p,n) reactions at 800 MeV [3,4]. In those days, we carried out coincidence measurements in (p,pl) reactions at the incident momenta of 3 and 4 GeV/c. In this experiment, we found that the A peak in the momentum distribution of forward protons in p.ir-events (a proton and a .ir-are emitted in the target region in coincidence with a forward proton) is consistent with the mass and width of A in free space [5]. It means the A mass does not change when we see the decay products of A, while the mass seems to be modified in inclusive momentum distribution. This was a puzzle. In order to investigate A production in nuclei for further details, we have carried out an experiment to measure both inclusive and coincidence spectra for (p, n) reactions at 1.5 GeV/c. The reaction channel was chosen because (1)the cross section for A productions
is maximum at around 1.5 GeV/c, (2) the contribution of background processes is smaller in (p, n) reactions than in (p,pl) due to the dominance of the pp -+ nA'+ channel, and (3) inclusive data exist at 1.46 GeV/c [6]. 2. EXPERIMENT
Experiments were performed at the 7r2 beam line of the 12-GeV Proton Synchrotron (PS) at the National Laboratory for High Energy Physics (KEK, now named the High Energy Accelerator Research Organization) using a large-acceptance spectrometer called FANCY for coincidence particle detection. This is a cylindrical spectrometer [7] composed of a solenoid (3kG magnetic field), a cylindrical drift chamber (jet-chamber type) and a cylindrical hodoscope. By measuring pulse heights from both ends of resistive sense wires, we are able to reconstruct tracks in three dimensional space. A momentum resolution o p / p 0.1 . p (in GeV/c) was achieved. Particles with momenta up to 800 MeV/c were well identified by the energy loss (dE/dx) measurements. The spectromenter covered a wide range of polar angles between 25 and 105 degrees (with the requirement of a hodoscope hit) or between 12 and 141 degrees (without this requirement). The forward neutron detector was made of a stack of plastic scintillators whose dimensions are 3 cm (thickness) x 20 cm (width) x 100-150 cm (height). Fifty scintillators were assembled into a 10 x 5 array to form a 200cm-wide by 15cm-thick detector. The detector was surrounded by veto scintillators with a thichness of 1 cm to reject charged particles. The flight path of the neutrons was about 10 meters. The neutron detector covered polar angles between 0 and 10 degrees, but only the data between 0 and 6 degrees were analyzed. A schematic drawing of the detector system is shown in Fig. 1.
Hodoscope Solenoid
Neutron Counter
Dipole
beam -C
tArget
Veto
Figure 1. A plan view of the detector system for the El73 experiment. An unseparated beam with an intensity of (1 - 2) x lo5 particles per 0.5-sec beam spill were delivered from the internal target of the PS to the experimental target. Target
materials used were carbon and polyethylene (CH2) with thicknesses of 0.87 and 0.94 g/cm2 respectively. H-target data were deduced by subtraction. The beam momentum measured by time of flight over a 12m - path was 1.49 GeV/c with a momentum spead of 0.02 GeV/c. Event triggers were generated by a coincidence of a beam proton and a hit in the neutron detector. A beam-veto scintillator with a dimension of 15 x 15 cm was placed at 1.3 m downstream of the target in order to reduce the trigger rates for accidental coincidences. Since the cylindrical hodoscope was not used in the trigger logic, the acceptance of the cylindrical spectrometer was very large (between 12 and 141 degrees, covering a solid angle of 88 % of 4n). Tracking efficiecy was more than 95 % in the above angular range. Tracking is possible for particles with momenta as low as 50 MeV/c, but low momentum particles could not be detected due to energy loss in the target material. Cutoff momenta for pions and protons were effectively 75 and 220 MeV/c, respectively. The detection efficiency of the neutron detector was calculated by a Monte Carlo program in which the detector geometry including veto counters was realistically treated. The efficiency was (5.5 f 1) % in the momentum range between 0.8 and 1.4 GeV/c. It slightly decreases monotonically as neutron momentum inreases because events in which a high momentum particle was emitted were rejected by the veto counter. Intrinsic T O F resolution of each counter was 200 psec (rms) for charged particles. Although long-term time drift was calibrated and corrected using laser pulses, overall T O F resolution deteriorated to 350 psec in summed data obtained for a data-taking period of 5 months. The corresponding momentum resolution was 15 MeV/c at 1.0 GeV/c and 50 MeV/c at 1.5 GeV/c. 3. RESULTS
All events were classified into 6 types according to the combination of particles detected in the cylindrical spectrometer(CDC); none in CDC (hereafter called none-event), only a n+(n-event), only a proton (p-event), a proton and a n+ (pn-event), 2 protons, and others. All types except none-events were almost free from background because of the excellent detection capability of the CDC. However, the cross sections for none-events must be deduced by subtracting data for target-out runs. The target inlout ratio was about 2:l for none-events. Inclusive cross sections were thus obtained as a sum of all 6 event types. Using the calculated efficiency mentioned above, inclusive cross sections agreed well with previous measurements at LAMPF [6]. Fig. 2 shows doubly differential cross sections for C(p, n) reactions, together with fractions of each event type in the A excitation region. We'd like to emphasize that our detector acceptance is so large that at least one charged particle is detected in 80 % of all (p,n) events. The fraction of 71.- and pn-events changes rather rapidly as a function of neutron momentum, while the fraction of other event types stays almost constant. The fractions of pn-, T-, none- and p-events in the A region for H target were about 65 %, 20 %, 10 % and 5 %, respectively, which agree well with a Monte Carlo calculation in which the cutoff for low momentum particles was taken into account. Fig. 3(a) shows the differential cross sections for (p, npnf) reactions on C and H targets. The peak in the spectrum for the C target shifts slightly (-50 MeV/c) toward
C
800
Carbon (inclusive)
1000
1200
i/
1400
Neutron Momentum (MeV/c)
-
Figure 2. Doubly differencial cross sections of (p,n) reactions at 1.5 GeV/c on C at 6' = 0 6 degrees. Solid line showed cross sections of the same reactions at 1.46 GeV/c at 4 degrees with a shift of -35 MeV/c to account for the difference of incident momentum. Calculation of quasifree A(with mass of 1200 MeV and width of 160 MeV) production is shown in dashed line. Above picture shows the fraction of each event type.
the low momentum side with significant broadening compared to the H target. But the spectra were well reproduced as shown in the Figure (solid line) if we account for the binding energy and the Fermi motion of nucleons in nuclei. The model used in the calculation was exactly the same as that in Ref. [5]. The width (aF)of the Fermi motion expressed in a gaussian form was 120 MeV/c. The best match to the experimental data was achieved with a binding energy of 40 MeV, while it was 25 MeV for (p,p'AO) reactions at 4 GeV/c. This model is, in some sense, self inconsistent because the binding energy and the Fermi motion are treated independently. In order to overcome this weakness, we also calculate the momentum spectra based on the model introduced by Gugelot [8] in which the two quantities are treated in a consistent manner. If we used a binding energy of 25 MeV in our model, the two model agree with each other in terms of quasifree two-body kinematics. Therefore the difference in the binding energy in (p, nA++) and (p,p'AO) reactions seems not to be caused by a trivial kinematic effect but by some as yet unknown dynamical effect. The ratio of cross sections O ~ ~ , , ~ + + / O ~ ~ , , ~ +is+ 1.2 f 0.1 which agrees well with the (P,~'AO)data at 4 GeV/c [5] where the target mass dependence of the cross section was discussed based on an intranuclear cascade model. While the forward momentum spectrum is well understood with a simple quasifree picture, an anomaly was found in the distribution of the invariant mass of a proton and a r+(MP,+). Fig. 3 (b) shows Mp,+ distributions
0 800
1000 1200 1400 Pn (Neutron Momentum, MeVIc)
. ..,
1100 1150 1200 1250 1300 M +(px+Invariant Mass, MeV)
1350
pn
Figure 3. (a) Momentum spectra of forward neutrons for PT-eventson C and H targets. Lines show our model calculation with the mass and width of Abe the same as those in free space (M=1232 MeV and r=115 MeV). Binding energy (Q) and Fermi motion (OF) are Q=40 MeV, crF=120 MeV/c (solid line), Q=O. ffF=O (dashed line; for H target), and Q=O, aF=120 MeV/c (dotted line), respectively. (b) Distributions of pn+ invariant mass for C and H targets. Breit-Wigner functions with M=1225 MeV and r=115 MeV (solid). M=1207 MeV and r=190 MeV (dashed) are also shown.
for C and H targets. The distribution for the H target was well expressed by a BreitWigner function with standard values for the mass (1232 MeV) and width (115 MeV) of the A. The distribution for the C target significantly differs from that for the H target. Fitting to a Breit-Wigner function yields a mass of 1207 i 7MeV and width of 190k25MeV. Using these values for A's in nuclei, the momentum distribution of forward neutrons from quasifree A production was calculated and is shown in Fig. . Agreement with the experimental data is fair. Those observations suggest two possible scenarios to understand quasifree A production: (a) A's in nuclei differ from A's in free space as shown by the Mp, distribution, and some additional momentum is transferred to a target nucleus for pr-events, or (b) A's in nuclei don't differ from free A's but rescatterings and final state interactions of a proton and a T+ decaying from a A result in a distorted M,, distribution. More studies are necessary to make clear conclusions. Fig. 4 shows momentum spectra of forward neutrons for T-, 2p- and pn-events along with results of a simple model calculation. In this model. a A with M = 1207 MeV and
r = 190 MeV is produced in a quasifree process.
When the momentum of a proton decaying from the A is less than 400 MeV in the target rest frame, the proton is captured by the target nucleus and only a pion comes out from the collision (coherent pion process). The probability for such events depends strongly on the mass of A, and therefore momentum spectrum for T-events differs significantly from that for pr-events as shown in the figure. Angular distributions of pions are also well reproduced by this model. Considering all experimental data obtained so far in hadronic processes, the following scenario seems to be most reasonable, although more sophisticated theoretical calculation is necessary to make a final conclusion: The mass of A in nuclei is effectively smaller than that in free space as had been already seen in r A total cross sections [lo]. Fate of a A depends strongly on its mass; probability of so-called coherent pion process increases when its mass is smaller, while probability of quasifree (decay) process increases when its mass is larger. Probability of absorption process is almost independent of the A mass because the cross section of N A -t N N shouldn't depend on the mass.
l
'
l
'
l
'
l
'
-
pPf
c Pq f n*r%f
@t*qGsa 3,grg @I@@
@
9icSB m i
99
800
9
n+
-
!ti
' ~@@,PI ' ~@
a I
'
.
HH
-
l
a pn+PP-
.p+C+n+(pn+, PP, x+)
.
I
1000
,
I
4,
.
1200
1400
Neutron Momentum (MeVIc)
Figure 4. Momentum spectra of forward neutrons for T-, 2p- and PT-events on C target. Now we qualitatively understand the behaviour of A in nuclei produced in hadronic processes. There are two big questions left; (1) what is the origin of the A mass shift? and (2) why A's produced in hadronic processes differ from those produced in electromagnetic processes? In addition to these questions, we should carefully examine the coherent pion process by experiments with high energy resolution. An experiment to reveal the coherent pion process was carried out at LAMPF [ll]but still clear answer was not yet obtained. 4. SUMMARY
In summary, we have carried out first coincidence and inclusive measurements of (p, n) reactions at 1.5 GeV/c on H and C targets. All (p,n) events were classified into 6 types according to particles emitted in the target region and fractions of each type were determined as a function of neutron momentum. Two types, PT-eventand T-event,
showed larger changes of fractions. For pn-events, monentum spectra of forward neutrons were well understood by a quasifree picture, but the p~ invariant mass distribution differs significantly from that in free A decays. The fate of a A in a nucleus seems to depend strongly on its mass. Probabilities of quasifree (decay), coherent pion, and absorption processes are qualitatively understood.
I would like to thank all collaborators of the KEK PS experiment El33 and E173; K. Nakai, T . Kobayashi, T . Nagae, K. Tokushuku, H. Sano, M. Sekimoto, I. Arai, H. Sakamoto, A. Manabe, K. Aoki, H. Nunokawa, M. Tanaka, M. Nimomiya, M. Tomizawa, H. Kitayama, iV.Kato, T. Baker, C. Glashausser, R. Ransome, G. Edwards and G. Kumbartzki. REFERENCES 1. D. Contardo et al., Phys. Lett. 168B(1986) 331. 2. V.G. Ableev et al., J E T P Lett. 40(1984) 763. 3. B.E. Bonner et al., Phys. Rev. C 18(1978) 1418.
4. C.W. Bjork et al., Phys. Lett. 63B(1976) 31. 5. T . Nagae et al., Phys. Lett. lglB(1987) 31. 6. R.G. Jeppesen, Ph.D. thesis, Univ. of Colorado, 1986 (unpublished). 7. K. Ichimaru et al., Nucl. Instr. Meth. A237(1985) 559. 8. P.C. Gugelot, Phys. Rev. C 30(1984) 654. 9. J. Chiba et al., Phys. Rev. Lett. 67(1991) 1982. 10. A.S. Carrol et al., Phys. Rev. C 14(1976) 635. 11. R. Gilman et al., Nucl. Phys. A577(1994) 227c.
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
The physics of AA excitation G.M. Hubera "Department of Physics, University of Regina, Regina, SK S4S OA2 Canada The physics of double delta excitation has been of interest since the early 19801s,when Brown. Toki, Weise and Wirzba [I] pointed out that in the simplest SU(3) consituent quark coupling model A N -+ AA is expected to be much stronger than A N + N N . Since it is generally believed that the A N + NN process plays a major (if not the dominant) role in pion and photoabsorption processes in the resonance region, this claim could have a major implication on our understanding of both processes. In the simplest case of a free deuteron target, A A excitation will require an incident energy of about 600 MeV to be placed completely on shell. On heavier nuclear targets, the process is helped by the high Fermi momentum tail, and so could be proportionately more important at lower energies than it would be on deuterium. The task, clearly, is to find a unique signature which would indicate the presence of this exotic, and otherwise unexpectedly important process. The first detailed calculations involving the A N -+ AA process were done for n absorption. motivated by the greater activity at the meson factories at that time. Arguments were made [I] that A A excitation following n absorption should have a rather unique kinematic signature, in which both A's internally de-excite, leading to the emission of four energetic nucleons via the mechanism shown in figure l a ) . For the sake of clarity, it should be pointed out that diagrams involving the excitation of two sequential A's, on two different nucleons. are not considered to be double delta excitation; the latter process requires both A's to be present simultaneously in the nucleus. In order for two nucleons to reach A excitation, it is clear that they must be at close range, and so the reaction is potentially sensitive to the importance of short range correlations (SRC) in the nuclear environment. The smallest nucleus for which this unique four nucleon emission signature can be studied is 4He, and so a detailed calculation was performed by Schwesinger, Wirzba and Brown [2]. This calculation not only involved A N + AA diagrams of the type shown in figure l a ) . but also virtual n + np decay diagrams such as in figure l b ) . In this case, the n + pn coupling is calculated from p + 27r decay. It is found that both processes play major roles in the 4He(n,4N) reaction, with the two diagrams having significant interference at lower energies. Unfortunately, a nonrelativistic approximation was used for the nucleon degrees of freedom, so the validity of the calculation may be increasingly doubtful above 600 MeV. Nonetheless, a AA cross section of 12 mb at 400 MeV would account for 30% of the measured pion absorption total cross section on 4 H e at that energy 131. . This work result,ed in a number of interesting findings. The calculation is especially ~
Figure 1. Pion absorption through formation of a A A state which subsequently decays, leading to emission of four energetic nucleons. a) the A N -+ A A mechanism. b) absorption through formation of a virtual p~ pair.
sensitive to the treatment of SRC, since it involves four nucleons close together. Their mutual short range repulsion was treated with a range of llm,, where m, = 730 MeV, and led to a reduction in the A A --+ 4 N cross section. In addition, because the two nucleons associated with the A's are initially in a relative S-state, the intermediate A A is dominantly in a I = 2 or a S = 2 state. The I = 2 states can only be reached by a tensor force acting on A N + A A , while the S = 2 states have three open channels reached by the tensor force, and only one by spin-spin forces. A further enhancement of the tensor contribution comes from the fact that the momenta transferred in A N + A A can be arbitrarily small, and so cancellations between T and p tensor contributions are less severe than in A N + N N transitions. Thus, it appears that the major contribution to the A A mechanism comes from the tensor force, while this force can be usually safely disregarded in the A N -, N N transition. To date, relatively little progress has been made in identifying the role of the A A -+ 4 N process in pion absorption. Experimentally, this is because of the high pion energy required, and the tremendous increase in the complexity of the pion absorption reaction at these energies, which mask any possible signature. Further theoretical investigation has also lowered the expectations somewhat. For example, a 1991 paper by Abbas [4] showed that inclusion of proper nucleon deformation in the N -+ A transition leads to a suppression of ( k ) 2 by 27% compared to the earlier SU(4) calculations by Brown, et al. These complications contrast with the early optimistic words of Schiffer: " T h e A A m e c h a n i s m almost has t o be present, its predominance is n o t a m a t t e r of ad hoc assumptions but of Clebsch-Gordan coeficients in a simple SU(4) quark model." [5]. The only significant experimental evidence for the AA mechanism using pion beam has come from studies of (T, 27r) at PSI. Rahav et al. [6] compared 12C(r-, 2 ~ and ~ ) 12C(r-, T-T+) yields at T, = 292 MeV using the CALLIOPE spectrometer, and found that there was a proportionally higher number of 2 r - events than anticipated from quasifree 27r production. If we compare the fundamental reactions, we expect a ratio
- Full calculation -- AN->AA (fig la)
Figure 2. Total cross section of the 4He(7r,4N) process versus total energy of the incoming pion, calculated by Schwesinger, Wirzba, and Brown [2].
while the observed ratio is
so there are nearly twice as many 27r- events as expected. However, according to a spinisospin Clebsch-Gordan coefficient calculation, the AA mechanism is expected to yield a ratio of 0.59, and so provide a more balanced ratio between the two charge states. Assuming that only quasifree and AA mechanisms contribute to the observed yield ratio, A (0.30 f 0.14)aQF. they concluded that ~ A = The clearest evidence of AA excitation has come from photoabsorption studies using the TAGX spectrometer at INS (Tokyo). Asai et al. [7] measured the d(y, r + r - p ) n reaction using a tagged photon beam with energy between 570 and 850 MeV, and employed Monte Carlo simulations of a number of competing r+7r-production mechanisms: Quasifree yp
-+
A++7rP + 7r+7r-p1 with the neutron as a spectator
Double delta ypn
-+
A++A-
-t
7r+r-pn
Four body phase space ypn -+ n+7r-pn, which could mimic a number of overlapping, higher mass resonances not individually resolved by the experiment. This analysis yielded acceptable fits to the experiment kinematic observables. Events with missing momentum < 250 MeV/c were entirely consistent with the quasifree mechanism
Asai et al. (TAGX) Wada et al. (Bonn) - Figs. 4 a)+b)+c) Figs. 4 a)+b)
Figure 3. A++A- cross sections obtained from the d(y, n+n-)n reaction at TAGX [7] and SAPHIR [B], and compared with the calculations of Gomez-Tejedor, Oset and Toki [9]both with and without the contribution of intermediate N*.
(i.e. the neutron was a spectator), while events with p,,,, > 250 MeV/c were best fit with a combination of AA and phase space contributions. The same analysis technique was also applied to d(y, 7r+7r-p)n data obtained by the SAPHIR collaboration at Bonn using tagged photon beam 750 5 E, 5 1280 MeV [B]. Although the greater contribution of N* resonances at this higher energy requires further analysis before final results can be published. the preliminary results are quite consistent with the TAGX experiment, and are clearly of higher quality. The conclusion of both experiments is that A++A- excitation is observed, but that A'AO is too weak to be identified. This is expected on the basis of isospin. It is also seen that the AA mechanism increases in prominence quite rapidly with increasing photon beam energy. attaining nearly 40% of the quasifree cross section by E, = 710 MeV, which is similar to the ratio quoted earlier by the (T, 27r) experiment at lower energy. The results from TAGX prompted a detailed calculation by Gomez-Tejedor, Oset and Toki [9]. Like the earlier pion absorption calculations, this model's foundation is A N -+ AA transitions of types shown in figures 4a) and 4b). In addition, the contribution of the N&,, resonance was taken into account via diagrams similar to that shown in figure 4c). The result of the calculation is shown in figure 3. There is a constructive interference between the A and N* diagrams below E, z 820 MeV and a destructive interference above this energy, which is due to the change of sign of the real part of the N* propagator. Although the importance of the N* diagram is small, the low energy experimental cross sections are better reproduced with its addition. The model acceptably describes the photoproduction data up to E, = 800 MeV, but deviates from the data in a steadily worse manner at the higher energies. This discrepancy was anticipated, because diagrams involving the virtual yp meson exchange current, such
Figure 4. a).b),c) sample diagrams included in the model of the yd + A++A- reaction by Gomez-Tejedor, Oset and Toki [9].d) diagrams including intermediate vector mesons, such as this one, were not included.
as in figure 4d), were not included. If we compare with the result of Schwesinger et al., we see that the inclusion of p diagrams indeed cannot be neglected at these energies. There is also the question of the deuteron D-state, which was neglected in these calculations, but also expected to play a more important contribution as the energy is increased. More recently, these studies were extended to lower energy on 3He by the TAGX collaboration [lo]. Inclusive 3He(:i, n+n-) data were obtained from 380 to 700 MeV incident photon energy. In the case of the lowest energy, this is only 50 MeV above the absolute n+n- detection threshold. The detection of additional protons, which would have made the experiment more exclusive in nature, was precluded by proton rangeout in the target, and the 300 MeV/c proton momentum detection threshold. The results are shown in figure 5, and are generally consistent with the deuteron data, after multiplication by 1.5. However, we also see that the data are in excess of the Gomez-Tejedor et al. prediction (after multiplication by the same factor), a difference which cannot be explained by the x 25 MeV difference in reaction threshold between 2H and 3He. Analysis of higher energy 3He(y. nf T-p) data from TAGX is still underway. In conclusion, we have seen that AA mechanisms do indeed play an identifiable role in n and 7 induced reactions above x 500 MeV, although the importance of their role has been tempered somewhat from the first optimistic expectations. Nonetheless, it would appear that the study of this mechanism can play a role in our better understanding of nuclear dynamics, as it touches on a large number of topics which have sustained interest in the nuclear community. These issues include the role that repulsive SRC play in setting the strength of the A N + A A process, the interference of N * and p r contributions, and the A N tensor force.
1
a Ref. 7 (d*1.5) o Ref. 8 (d*1.5) w Ref. 10 ( 3 ~ e ) - d*1.5 Ref. 9 with N'
1 T
Figure 5. 3He(y, AAp) cross sections obtained at TAGX [lo], and compared with the data and calculations on deuterium target (after multiplication by 1.5).
The author wishes to thank INS and the TAGX collaboration for their role as hosts and collaborators over the last ten years. It was a very enjoyable and valuable experience. This work was supported in part by the Natural Sciences and Engineering Research Council of Canada (NSERC).
REFERENCES G.E. Brown, H. Toki, W. Weise, A. Wirzba, Phys. Lett. 118B (1982) 39. B. Schwesinger, A. Wirzba, G.E. Brown, Phys. Lett. 132B (1983) 269. D. Ashery, J.P. Schiffer, Ann. Rev. Nucl. Part. Sci. 36 (1986) 207. A. Abbas, J . Phys. G 17 (1991) L181. J.P. Schiffer, Comrn. Nucl. Part. Phys. 14 (1985) 15. A. Rahav, et al., Phys. Rev. Lett. 66 (1991) 1279. M . Asai, et al., Z. Phys. 344 A (1993) 335. Y. Wada, representing the SAPHIR collaboration, Particles and Nuclei International Conference, Williamsburg, Virginia, 1996 (unpublished). 9. J.A. Gornez-Tejedor, E. Oset, H. Toki, Phys. Lett. 346B (1995) 240. 10. D.G. Watts, et al., Phys. Rev. C 55 (1997) 1832.
1. 2. 3. 4. 5. 6. 7. 8.
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
"
The neutral pion photoproduction on the proton near threshold 11-T. Cheona and M. T . ~ e o n g ~ "Department of Physics, Yonsei University, Seoul 120-749, Korea bDepartment of Physics, Dongshin University, Naju 520-714, Korea , have been calculated by taking the vector The Eo+and MI+ amplitudes for ~ ( yirO)p mesons into account in the pole model of t-channel. A conclusion is led that the rescattering process below the n + threshold plays a key role in this process. Our results are Eo+= -1.23 x 10-3mG, MI+ = 7.56 x 10-3m;$kq and MI- = -2.15 x 10-~m;$kq at the n+ threshold. The PCAC hypothesis is well restored in our frame of calculation.
1. INTRODUCTION The neutral pion photoproduction on the proton near threshold drew tremendous attention of physicists in the context of a possibility that the low energy theorem(LET) might be broken down. Since the LET is based on the partially conserved axial current (PCAC) hypothesis [I], a break of the LET implies a break of PCAC. For such a reason, the results caused a sensation [2,3]. In 1994, we pointed out [4] that the pion p-wave contribution should be properly substracted from the total cross section to extract the electric dipole amplitude near threshold and that the experimental results could well be reproduced consistently with the LET if vector meson contributions were taken into account. Fuchs et al. [5] performed again the experiment to measure the total and diffenrential cross sections near threshold with improved accuracy. Their results are in good agreement with those calculated by the chiral perturbation theory (ChPT) [6]. However, the ChPT cannot make an exact numerical prediction for Eo+amplitude so far, because the calculation of the loop diagrams converge too slowly. Even in this situation, a key question is left whether the charge exchange process, yp -+ n + n -+ irop, can contribute to Eo+even below the ir+ threshold because of mass differences between nn+ and nop. Such a process is expressed as [7]
where a(n.ir+ + p r o ) = 0.172f m is the pion-nucleon scattering length, Eo+(v nn+)= 26.6 x 10-3m;1 is larger than Eo+(yp -+ pnO) by one order of magnetitude and q = 0.190 i fm-' is the ir+ momentum below the ir+ threshold. This imaginary momentum makes Eo+real, i.e. -+
Hence, a question arises whether rescattering of the pion carrying an imaginary momentum below the threshold can contribute to the process, i.e. the real part of amplitude. Notice that the propagation particle is carrying a real momentum in usual rescattering processes. It is very interesting to analyse carefully the experimental results to find a definite answer to the question. 2. CALCULATION
The LET prediction of the electric dipole amplitudes for yp is given as
-+
reaction a t threshold
where p = m,o/M, f2/47r = 0.08, e2/47r = 11137, K, = 1.79, mro = 135MeV and M = 938.3MeV. The amplitude E& given in eq.(2) is the result of on-shell calculation. On the basis of quark model, we estimated the off-shell contribution to be +0.15 x 10-~m$. Therefore, the rescattering amplitude is E+ : = -0.72 x 10-~m;:. Magnetic dipole amplitudes were calculated in the framework of the BL theory [8]. The -: = -6.31 x l~-~rn;:k~ Born term has been estimated as M E = 3.13 x l~-~rn;:kg and M . The A contribtions calculated in both s- and u-channels are M& = 4.17 x l~-~rn;;kg and Mf_= 1.84 x 10-~m;:k~. Let us evaluate the anomaly term [9] due to the vector mesons using the pole model in the t- channel of the dispersion theory [lo]. Although the vector coupling term has been considered by several authors [8,10,11],tensor coupling term was usually neglected. Since it is expected to play an important role to reproduce the experimental results, we take it into account in our calculation. Vector meson exchange is described by
where E ~ , , ~is antisymmetric pseudotensor of rank 4, i p is the photon polarization vector and t = (p2 - p1)2. The pion photoproduction amplitude in the 7rN center-of-mass frame is given in the general form
F = i ( o . e ) F l + ( 0 . 6 ) ~[El. x k ] ~ ~ + i ( a . & ) ( E l . ~ ) F ~ + i ( u ~ ~ ) ( ~ l ~ $ ) (5) F~ for the energy up to the (33)- resonance. Here, k and are the unit vectors of the photon and pion momenta. Each amplitude, F,, can be connected to the invariant amplitudes Ai, by a set of linear equations as
+
+
where ($ = m,+/(8nWs), Xv = eg,,,/m,o(m:o mb), C = m: (mi/2W,) and D, = (m E,)lI2 with El = (m2 + k2)lI2 and E2 = (m2 q2)1/2. The coupling constants : = 0.1 f 0.5, in the N N V vertex are enpirically known as [12] gLT/4n = 8.1 f 1.5, gT gPv/4n = 0.55 & 0.06 and g i T / 4 ~= 20.5 2.1, which induce g,v = 10.09, g , ~= 1.12, = 0.1 and gpT/gpV = 6.1. The Qpv - 2.63 and gPT = 16.05. Here, we find g,,/g,v coupling constants, g,, and g,,,, are evaluated with the decay widths [13],I'(w -+ irOy)= (0.717 f 0.042)MeV and r ( p 0 -+ noy) = (0.118 f 0.030)MeV. The results are g,, = 10.313 and g,, = f 0.130. Their ratio is lg,,,/g,,,j = 2.4, which is smaller than the S U ( 3 ) result of 3. Notice that the data for I'(po -t r O y )has rather a large error. The data ir-y) = (0.067f 0.007)MeV for p -+ ir-y decay shows a better precision [13],i.e. r ( p which implies g,, = f0.0983. Their value gives the ratio Ig,,,/g,,,l = 3.18 being very close to the SU(3) result. The signs of these coupling constants are undetermined at this stage. To determine their signs, let us examine the deuteron magnetic moment. The experimental value is pd = 0.843nm which is in deficit of 0.014nm. This deficit might be caused by neglect of contributions from the ynp and ynw currents [13,14]. Our estimates of those contributions are Ap; = f0.019nm and Ap: = 10.006nm. To compensate the above deficit in pd, we must have a positive sign for w contribution at least, i.e. g,, = f0.313. On the other hand, if Ap: is positive we have Apd = +0.025nm, which is too large compared to the deficit Apd = +0.014nml while the negatve value of Apd gives a reasonable result Apd = +0.013nm. Thus, we conclude g,, to be negative, i.e. g,,, = -0.0983. With these values, we find the A E L amplitude at threshold as
+
+
+
-+
3. RESULTS
The s-wave amplitude for the neutral pion photoproduction at threshold yields
Table 1 Values of the electric and magnetic dipole amplitudes, Eo+and MI+, at threshold. The experimental result(I1) for the magnetic amplitudes is available for MI = MI+ - MI-. The values of the Eo+ and MI+ amplitudes are in units of 10-3m;i and 10-3m;tkq, respectively. Experimental results I and I1 are taken from refs. 3 and 5, respectively. Exp. (11) Born A(1232) (p,w ) Rescatt. Sum Exp. (I) Eo+ -2.27 0 1.76 0.72 -1.23 -1.32 f 0.11 -1.31 f 0.08 MI+ +3.13 +4.17 0.26 0 +7.56 +7.41 MI =
The amplitude Eo+ near threshold without taking account of the rescattering process, yp -+ ~ + -+n TOP, mentioned in the beginning is shown by a dot-dashed curve in Fig.1. It is obvious that the dot-dashed curve does not agree with the experimental result. The amplitude Eo+at threshold without taking account of the rescattering procces is is shown by the dotted curve in Fig.1, and it yields
Figure 1. Electromagnetic dipole amplitudes, EO+ and MI. The dot-dashed and dotted curves are the Eo+amplitude obtained without and with the rescattering below the .ir+ threshold. The dashed curve represents the MI amplitude. Experimental results are taken from ref. 5.
Figure 2. Total cross sections. The solid and dotted curves are obtained without and with the cusp effect due to the rescattering, respectively. The dashed curve represents the s-wave contribution with the rescattering. Experimental data are taken from refs. 2, 3, 5.
Contributions of the vector mesons to the MI+ amplitudes can be evaluated with Eqs. ( 7 ) and (8). The results are = 0.26 x 10-~rn;$kq and MK = 2.32 x 10-~rn;$kq. Thus, the total values for the magnetic dipole amplitudes are found as
ML
+
We define Ml = 3E1+ MI+ - MI-. This result is in Fig. 1 with a dashed curve. Figure 2 shows the total cross section calculated with these electromagnetic amplitude near threshold. The solid and dotted curves are obtained without and with the rescattering process, respectively. j R o m these figures, one may confirm the existence of the rescattering process below the T + threshold. The dashed curve in Fig. 2 is the s-wave
I
20
-
,
1
1
I
= 146.1 MeV
Er = 145.3 MeV
E,.
ET=147.8 MeV
E,z148.9 MeV
I
,
~r=1~6.~'?4,4,~./
E, =149.8J&aiJ ,- \
Q
$40-
PION ANGLE
,B,
(7
Figure 3. Differential cross sections. The solid and dotted cures are obtained without and with the rescattering. The dashed curves are calculated with the amplitude given in eq. (14). Experimental data are taken from ref.[5].
contribution with the cusp effect due to the rsscattering process. Our results are consistent with the recent experimental data obtained by Fuchs et al. 151. Figure 3 shows the differential cross sections with recent experimental data [5], which improved accuracy and shifted globally downward compared with the previous results [2]. The solid curve shows the results without the rescattering contribution, while the dotted curve represents the results with the rescattering contribut'ion below the n+ threshold. We can see that the dotted curves agree with the experimental results [5] much better than the solid curves. Since the rescattering discussed here appears only below the .ir+ threshold, we do not have any dotted curves for the cases of E, = 151.7MeV and 152.5MeV. Although we used g , ~= 16.05 evaluated from the empirical value gPT/g,v = 6.1 in the present calculation, our previous analysis used rather small value for the tensor coupling of rho meson, i.e. g p =~ 9.73 evaluated from g p ~ l g P V= 3.7 based on the SU(3) vector dominance. Since AE& = AE,", AE{+ = 0.933 x 10-~rn%in our previous calculation, the amplitude ~ h h was ,
+
This value is somewhat close to that given in eq.(12). Therefore, it may be supposed that
if g p = ~ 9.73 is used, the experimental results can be explained without the rescattering process below the .irf threshold. Indeed, the differential cross sections up to the .ir+ threshold energy can reasonably be reproduced. The results are shown by dashed curves. However, the improved data at E, = 151.7MeV and 152.5MeV can not be reproduced, as is seen in Fig. 3. Agreement with the data of Eo+shown in Fig.2 could not be satisfied either. 4. C O N C L U S I O N
z , = 7.56 x 10-~rn;ikq and MI- = -2.15 x Our results are Eo+= -1.23 x ~ O - ~ r n MI+ 10-3rn;i kg at the .ir+ threshold. We are led to a conclusion that the rescattering of the pion carrying an imaginary momentum below the .ir+ threshold can play important roles as a physically meaningful process. Our present investigation leads to a conclusion that the recent experimental results are consistent with the LET and the PCAC hypothesis remains valid. REFERENCES 1. P. de Baenst, Nucl. Phys. B24, 633 (1970). 2. R. Beck et al., Phys. Rev. Lett. 65, 1841 (1990). 3. J. C. Bergstrom et al., Phys. Rev. C53, R1052(1996). 4. B. G. Yu, 11-T. Cheon and M. T . Jeong, J . Phys. Soc. Japan 63, 78 (1994). 5. M. Fuchs et al., Phys. Lett. B368, 20 (1996). 6. V. Bernard, N. Kaiser and U. Meissner, Z. Phys. C70, 483 (1996). 7. A. N. Kamal, Phys. Rev. Lett. 63, 2346(1989). 8. I. Blomqvist and J . Laget, Nucl. Phys. A280, 405 (1977). 9. J. Wess and B. Zumino, Phys. Lett. 3 7 B , 95 (1971). 10. F. A. Berends, A. Donnachie and D. L. Weaver, Nucl. Phys. B 4 , 103 (1967). 11. T. Schafer and W . Weise, Nucl. Phys. A531, 520 (1991). 12. T. Ericson and W. Weise, "Pion and Nuclei", Clarendon Press, Oxford 1988. 13. R. J . Adler, Phys. Rev. 141, 1499(1966). 14. M.Chemtob, E. J. Moniz and M. Rho, Phys. Rev. C10, 344 (1974).
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Photoneutron cross section measurement on 'Be by means of inverse Compton scattering of laser photons H. Utsunomiya, Y. Yonezawa, H. Akimune, T. Yamagata and M. Ohtaa M. Fujishiro H. Toyokawa, and H. Ohgakic
" Department
of Physics, Konan University, Okamoto 8-9-1, Higashinada, Kobe 658-8501, Japan
bOsaka Prefecture University, Gakuencho 1-2, Sakai, Osaka 599-8570, Japan 'Quantum Radiation Div., Electrotechnical Laboratory, Umezono 1-1-4, Tsukuba, Ibaraki 305-8568, Japan Photoneutron cross sections were measured for 'Be in the energy range from 1.8 to 4.0 MeV using a laser-electron MeV photon source. These cross sections are relevant to the reaction rate of the first step of a-process, a ( a n , y)gBe, in a neutrino-driven wind of core-collapse supernovae. Results for the 1/2+, 512-, 112-, and 5/2+ states are presented in comparison with the data taken with other photon sources. 1. INTRODUCTION
The neutrino-driven wind formed in core-collapse supernovae may be an ideal site for the r-process [I]. In this rapidly expanding environment, n- and a-rich material is first processed into 9Be through a(a n, y)gBe by bridging mass gaps A = 5 and 8, which is followed by the 'Be(a, n)12C. This sequence of 12C production proceeds more efficiently than the triple a process. The subsequent reaction network is considered to produce seed nuclei beyond j6Fe for r-process. 'Be is a loosely-bound nuclear system consisting of two a ' s and a neutron. Non of any two constituents alone can form a bound system. The lifetimes of 5He and 'Be are 1.1 x sec and 0.97 x 10-l6 sec, respectively. Three adjacent thresholds exist at 1.573 MeV for a a n, at 1.665 MeV for 8Be n and at 2.467 MeV for 5He a . In view of the lifetimes, the synthesis of 9Be is considered to proceed more dominantly through 8Be than through jHe in two successive binary processes. The ternary process is likely to play a minor role. Resonances near these thresholds determine the rate of the 'Be synthesis. Resonance parameters of these states can be obtained by photodisintegrating 'Be. Previously, photoneutron cross sections for 'Be were measured near thresholds using two kinds of real photon sources, Bremsstrahlung and radioactive isotopes. Bremsstrahlung measurements with a photon difference method lacked great accuracy due to poor energy resolution [2]. Measurements with radioactive isotopes (24Na,"A1, 38C1,j8C0, 65Ni,88Y, lo5Ru, lZ4Sb,142,144Prand 206Bi)were limited below 2.75 MeV [3-81. Low-lying states in
+ +
+
+
'Be were also investigated in electron scattering [9,10]. However, virtual photons populated the first excited state (1/2+) only very weakly. No single measurements provided data for all resonance states near thresholds (1/2+, 512-, 112-, 512'). 2. EXPERIMENTAL METHOD 2.1. y-ray production and neutron detection Inverse Compton scattering of laser photons incident on relativistic electrons in the storage ring TERAS of the Electrotechnical Laboratory (ETL) produces quasi-monochromatic y rays in the energy range from 1 to 40 MeV [ l l ] . This new MeV photon source, LEMPS (laser-electron MeV photon source), is available as a pencil-like beam typically with a flux of lo4 photons/sec/mm2 and energy resolution of a few %. The beam properties were previously investigated with an anti-Compton spectrometer [12]. A Nd:YLF laser (A = 1053 nm) was used. The laser system was operated at 1 kHz and produced 100 % linearly polarized light. Unpolarized light was also produced by passing polarized light through an optical element called depolarizer. Electron energy was varied from 314 MeV to 475 MeV to produce y rays in the energy range of 1.8 MeV - 4.0 MeV. y rays were collimated into 2 mm in diameter with a 20 cm thick P b block at 5.5 m from a head-on collision point between electrons and laser photons. y-ray energy was measured with a 155 cm3 pure-Ge detector of coaxial type. A full energy peak was energy-calibrated with natural sources of 40K and 208T1as well as a standard source of 60Co. A 4 cm thick 9Be rod of 99.5 % enrichment (2.5 cm in diameter) was irradiated. Neutrons were measured with four 60 cmHg BF3 counters (MITSUBISHI ND8534-60: 2.5 cm in diameter x 24.1 cm in effective length) embedded in a 30 cm polyethylene cube. The BF3 counters were located, two each vertically and horizontally, in a concentric ring at 7.5 cm from the beam axis. Neutron moderation time in the polyethylene was measured in l ms time range with an ORTEC 566 TAC module. Fig. 1 shows a moderation time distribution. Background neutrons, a time-independent component, made a minor contribution to the neutron measurement. Incident photons were detected directly with a BGO detector (2" in diameter x 6" in length) placed at the end of the beam line. Pile-up spectra were obtained for the photon beam pulsed at 1 kHz. The average number of photons per beam pulse was determined from the pile-up spectra with use of a single photon spectrum that was separately taken with a DC beam. The total number of photons was obtained from the frequency (IkHz) and the data acquisition time. Independent measurements were carried out with a DC beam, in which foreground neutrons and background neutrons were measured before and after each measurement. The results of the pulsed- and DC-beam measurements agreed to each other within the accuracy of 5 %. 2.2. Neutron detection efficiency A 2.16 mCi 24Na(half-life 15 hours) was made by irradiating a 50 mg sample of NaOH for 15 minutes with thermal neutrons at the Research Reactor Institute of Kyoto University. It was mixed with 2 cm3 D 2 0 in an aluminum container (2 cm in diameter x 4 cm in length). The 24Na beta-decay to the 4+ state in 24Mg with a 99.94 % branching ratio, producing 2754 keV y rays in transition to the 2+ transition. These y rays disintegrated deuterons and produced 265 keV neutrons. Contributions of other states in 24Mg
0
200
400
600
800
1000
Channel Number
Figure 1. Moderation time distribution for neutrons in a polyethylene cube. Full scale is 1 ms. The time constant for moderation is about 140 p sec. Background neutrons that arrived at BF3 counters time-independently are clearly distinguished from photoneutrons of interest.
to neutron production were neglected because of small branching ratios. The absolute intensity of the 265 keV neutron source was determined with the water bath method [13]. The source was placed at the center of a 400 liter water bath. Neutron flux was measured from 2.5 cm to 17.5 cm in 2.5 cm steps with a 1.5 cm3 BF3 counter (MITSUBISHI ND8522-60). The efficiency calibration of the small counter for thermal neutrons had separately been done with the standard graphite pile at ETL [14]. Background neutrons were measured with the same counter covered by a 5 mm thick cadmium. Time elapsed during the measurement was corrected for. The source intensity at the time of production was determined to be 2.83 x lo3 counts/sec/4~. The neutron source was placed at the center of the neutron detector. A neutron counting was performed five times, each for 300 seconds, with two bare BF3 counters and two Cdcovered counters. The contribution of background neutrons was very small. The detection efficiency thus determined for 265 keV neutrons was 1.67 & 0.09 % per BF3 counter. The uncertainty stemmed from 1 % systematic error in neutron counting, 5 % in calibration of the small BF3 counter and 1.7 % in determination of the source intensity. Fig. 2 shows the total efficiency of the neutron detector as a function of neutron energy. The efficiency was also measured with a 252Cfsource whose average neutron energy is 2.35 MeV. The energy dependence of the efficiency was calculated with a Monte Carlo code MCNP [15]. 3. RESULTS
Cross sections were obtained with a formula
8
/
'.N~OH
+ D20
Monte Carlo Cal. MCNP
~
Figure 2. Neutron detection efficiency as a function of neutron energy. The efficiency was experimentally determined with sources of 24NaOH D 2 0 and 252Cf. The neutron energy dependence of the efficiency was calculated with the MCNP Monte Carlo code.
+
where Pin is the number of neutrons detected, N, is the number of y rays incident on the BGO detector, Nt is the number of target nuclei per cm2, and E,(E,) is the neutron detection efficiency. A correction factor, f,was needed for the measurement with a thick target.
where p is the total y-attenuation coefficient in 9Be, and t is the length of the 'Be target material. The attenuation coefficient was taken from [16]. Fig. 3 shows measured photoneutron cross sections as a function of y energy. Note that results for unpolarized y rays (open circles) agree with those for polarized ones (solid circles) within statistical uncertainties. The present measurement fully covered the excitation-energy range of three resonances (512-, 112-. 5/2+), while it covered only a high-energy tail of the first excited state (1/2+). This is because as approaching to the 1.665-MeV threshold, it became difficult to accurately evaluate what fraction of the y-ray flux lay above the neutron threshold. A Monte Carlo calculation of the asymmetric energy spread of LEMPS may help deduce the necessary y flux. This will be done in the future. Currently, a least-square fit to the data is in progress to deduce resonance parameters (7-decay widths etc.) towards an evaluation of the relevant reaction rate.
Present results (polarized) Present results(unoolarized) Fujishiro et al. (i982) RI BHarnermesh et ai.(1953) B.Russell et al. (1948) RI J.H.Gibbons et a1.(1959) RI W.John et a1.(1962) RI ..... Jakobson et al. (1961) Bremss. o
1.5
2.0
2.5
1
3.0
3.5
4.0
1 4.5
E (MeV) 7
Figure 3. Photoneutron cross sections for gBe. For comparison, results of the preceding measurements with Bremsstrahlung and radioactive isotopes are also shown.
4. ACKNOWLEDGMENT
We are indebted to Mr. K. Okamoto (Research Reactor Institute of Kyoto University) for his technical assistance in the irradiation of NaOH with thermal neutrons, to Mr. Y. Ogawa (Kinki University) for his help in running the MCNP code, and to Dr. K. Kudo (Electrotechnical Laboratory) for his cooperation in the detector calibration with the standard graphite pile. This work was supported by the Japan Private School Promotion Foundation.
REFERENCES 1. S.E. Woosley and R.D. Hoffman, Astrophys. J . 395 (1992) 202. 2. M.K. Jakobson, Phys. Rev. 123 (1961) 229. 3. B. Russel e t al., Phys. Rev. 73 (1948) 545. 4. B. Hamermesh and C. Kimball, Phys. Rev. 90 (1953) 1063. 5. J.H. Gibbons et al., Phys. Rev. 114 (1959) 1319. 6. W. John and J.M. Prosser, Phys. Rev. 163 (1967) 958. 7. M. Fujishiro et al., Can. J . Phys. 60 (1982) 1672. 8. M. Fujishiro et al., Can. J . Phys. 61 (1983) 1579. 9. H.-G. Clerc et al., Nucl. Phys. A120 (1968) 441. 10. G. Kuechler et al., Z. Phys. A326 (1987) 447.
11. H. Toyokawa et al., Nucl. Instr. and Methods in Phys. Res. A422 (1999) 95. 12. H. Ohgaki et al., IEEE Tran. Nucl. Sci. 38 (1991) 386. 13. Experimental nuclear physics, Vol. 2, edited by E. Segre (John, Wiley & Son, Inc., New York, NY 1978) p. 455. 14. Y. Inoue et al., Bull. Electrotech. Lab. 31 (1967) 60. 15. J.F. Briesmeister, MCNP, A General Monte Carlo N-Particle Transport Code, Version 4B, Las Alamos National Laboratory, 1997. 16. Engineering Compendium on Radiation Schielding, Vol. 1, edited by R.G. Jaeger, E.P. Blizard, A.B. Chilton, M. Grotenhuis, A. Honig, Th. A. Jaeger, and H.H. Eisenlohr (Springer-Verlag New York Inc. 1968) p. 173.
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Nuclear disintegration induced by virtual photons at heavy-ion colliders I.A. Pshenichnova * t "Institute for Nuclear Research, Russian Academy of Science, 117312 Moscow, Russia A model of electromagnetic interaction of relativistic heavy ions which makes it possible to calculate the inclusive and exclusive characteristics of such interactions is presented. The multistep model includes (1) the single and double virtual photon absorption by a nucleus, (2) intranuclear cascades of produced hadrons and (3) statistical decay of excited residual nuclei described by evaporation, fission and multifragmentation mechanisms. Components of the model were verified by comparison of calculation results with experimental data obtained with monoenergetic photons and heavy ions at KEK-Tanashi Synchrotron and CERN SPS accelerator, respectively. 1. DESCRIPTION OF THE MODEL
The relativistic Coulomb excitation of nuclei is a well-known phenomenon [I]. It is expected to play an important role in collisions of ultrarelativistic heavy nuclei at new colliders, RHIC and LHC [2,3]. The process may be described in the framework of the Weizsacker-Williams method of equivalent photons. According to this method the impact of Lorentz-boosted Coulomb field of a nucleus on the collision partner may be treated as the absorption of equivalent photons. Due to the coherent nature of photon emission by the nucleus, such photons are almost real, Q2 5 1/R2, R is the nuclear radius. Therefore photonuclear reaction data obtained in experiments with real monoenergetic photons and appropriate theoretical models describing photonuclear reactions may be used to predict the properties of electromagnetic dissociation process. Since an ultrarelativistic nucleus moving at the impact parameter b with the velocity p = V / C % 1, (y >> 1) spends a short time At near the collision partner, the virtual photon spectrum contains all the frequencies up to the maximum energy E,ma" l / A t y / R :
where a is the fine structure constant, KO and K1 are the modified Bessel functions and x = E,b/(yp). The mean number of photons absorbed by the collision partner of mass A, is defined by:
*Supported by INTAS fellowship YSF-98-86 te-mail:
[email protected]
-
where OA,(E-() is the appropriate total photoabsorption cross section obtained starting from the photoneutron threshold at Eyin 7 MeV. Assuming the Poison distribution for the multiphoton absorption with the mean multiplicity m(b),Eq.(I), the integral cross sections are calculated for the first- and second-order processes [4] for a particular dissociation channel i :
with the following single and double photon spectral functions:
/ b d b e - m ( b ) ~ (b),~ l , N ( ~ ) ( EE2)~ , DC,
~ ( " (=4I n)
=n
bmm
/m
b d b e - m c b ) ~ (b)~N(E2, l, b).
bmtn
The values f:')(El) and f , ( 2 ) ( ~ 1 ,E2)defined as the branching ratios for the considered disintegration channel i in the single and double photon absorption, respectively, have to be calculated by our multistep model. Similar expressions may be written for the differential distributions of produced particles on the rapidity, transverse momentum and other variables. In order to obtain f:') and f,(2)several mechanisms are included in a new Relativistic ELectromagnetic DISsociation (RELDIS) code for the Monte Carlo simulation, namely, the intranuclear cascade of fast particles produced after the photon absorption on a nucleon or nuclear pair [5], and the evaporation of nucleons and lightest fragments, binary fission or multifragmentation [6] at a later stage of interaction. The multifragmentation process dominates when the excitation energy of residual nucleus, E*, exceeds 3 - 4 MeV/nucleon and is described by the statistical multifragmentation model (SMM) [6]. For fissile nuclei the evaporation may take place before or after the fission. The competition of evaporation and fission is also described with the SMM. The lightest fragments may be also created via coalescence of fast nucleons into dl t , 3He or 4He [7]. The decay of highly excited light residual nuclei with A 16 is treated by the Fermi break-up mechanism 161. Other details of the calculation scheme may be found in Refs. [4,5,8].
<
2. EMISSION OF NUCLEONS AND LIGHTEST FRAGMENTS
Depending on the photon energy, E,, and mass number, A,, different processes take place in the nuclear photoabsorption. Due to the contributions of several mechanisms: y N -+ 7rN, y N -+ 27rN and y(np) + np the calculated double differential spectra d2a/dRdP of pions and protons have complex shapes above the pion production threshold [5]. The spectra of fast particles n+, n-, 7 and p predicted by the model were compared 1000 MeV including in Ref. [5] with available sets of experimental data at 140 E,
<
<
the data obtained with KEK-Tanashi, the 1.3-GeV Electron Synchrotron. A satisfactory description of the spectra was obtained. Fast hadrons produced after the photon absorption initiate a cascade of subsequent collisions with the intranuclear nucleons leading to the heating of a residual nucleus. Later the nucleus undergoes de-excitation by means of the emission of nucleons and fragments. Because of a low Coulomb barrier in light nuclei the rates of proton and neutron emission are comparable. Fragment spectra in the photoabsorption on a carbon nucleus are given in Fig. 1. The low energy part of the deuteron spectra is explained by the explosive Fermi break-up while the high energy part is attributed to the coalescence mechanism. Since the main part of fast nucleons is emitted in the forward direction, the coalescence contribution dominates at small angles. On the contrary, the distribution of deuterons from Fermi break-up is nearly isotropic. Because of the option to accelerate also light oxygen ions at RHIC, the
y(380 MeV)
+C
+d+X
y(580 MeV)
+C
+d+X
1 1 1 1 1 1 1 ' ' ~ 1 1 1
Figure 1. Deuteron emission in photoabsorption on carbon. Points: KEK-Tanashi data [9]. Solid histograms: results of the intranuclear cascade calculations with coalescence and Fermi break-up. Contributions from coalescence mechanism are shown by the dashed, dotted and dash-dotted histograms for different values of coalescence parameter [7]:po = 90,129 and 200 MeV/c, respectively.
calculations of the electromagnetic dissociation contribution for such ions should take into account both of the considered mechanisms.
The photoabsorption scenario is very different for heavy nuclei, gold and lead. In this case neutrons are the most abundant particles produced in the electromagnetic collisions of ultrarelativistic nuclei [4]. When a heavy nucleus absorbs photons in the Giant Resonance region, 6 5 E, 5 30 MeV, the evaporation model may be used with the assumption E* = E, which leads to the emission of one or two neutrons. On the other side of the equivalent photon spectrum, when E, reaches the value of several GeV, the multiple pion photoproduction on intranuclear nucleons becomes the main absorption mechanism. In this case up to 95% of the photon energy is released in the fast particles on average. Nevertheless, the remaining energy deposited in the residual nucleus is sufficient for evaporating many neutrons or even multifragmentation [8]. As it was found in Ref. [4]for PbPb collisions at SPS (158A GeV beam) and LHC (2.75A+2.75A TeV beams) the mean neutron multiplicities are 4.2 and 8.8, respectively. The same value for AuAu collisions at RHIC (100A + lOOA GeV) is equal to 7.2. The neutron multiplicity distributions are shown in Fig. 2. They are strongly peaked at the single neutron emission channel due to GDR decay, while there is a long tail of the multiple neutron emission originating from high excitations. These results provide important information for designing large-rapidity detectors and zero-degree calorimeters at RHIC and LHC.
Ll"
" I "
" I ' " ' I " " I U "
W
Figure 2. RELDIS predictions for multiplicity distributions of neutrons in the electromagnetic dissociation of P b nuclei at LHC and SPS energies (solid and dotted lines, respectively) and Au nuclei at RHIC energies (dashed lines).
3. CHARGE CHANGING REACTIONS AT SPS
CERN SPS accelerator currently supplies the highest energy available for heavy ion studies. A detailed understanding of fragmentation mechanisms at SPS energies provides an important point for the extrapolation of the theoretical and experimental results to RHIC and LHC energies.
<
<
Partial break-up charge changing cross sections (-7 AZ 1) for 158A GeV 208Pb ions interacting with various targets, from H to Pb, were measured recently [lo]. Although the contributions to fragmentation due to the electromagnetic and nuclear forces were not separated in the experiment [lo], one can use RELDIS code to evaluate the role of the electromagnetic dissociation for each particular dissociation mode. As it is shown in Fig. 3, the main part of charge changing cross sections in PbPb collisions may be explained by the electromagnetic interaction.
158A GeV '-Pb on '07pb
Figure 3. RELDIS results for fragmentation charge changing cross sections of 208Pbions at SPS. Points: experimental data [lo].
Figure 4. RELDIS predictions for mass distributions of isotopes with charges Z= 75,76, ...,82 produced in the electromagnetic dissociation of 208Pbions at SPS.
Calculations show that multiple neutron emission also plays a role iri each of the partial charge changing channels with AZ = -6, ..., -1 for PbPb collisions, see Fig. 4. The lost of each proton is accompanied with a high probability by the multiple neutron emission. For example, up to 20 neutrons may be emitted in AZ = -2 fragmentation channel, and the emission of 8-10 neutrons is the most probable process.
This is evident from the fact that due to a high Coulomb barrier of P b nucleus, photoemission of protons takes place well above the GDR region, particularly in the region of quasideuteron absorption, where the excitation energy, E* 20 - 50 MeV, is sufficient to evaporate many neutrons. On the contrary, the mass distribution for AZ = 0 channel is completely different since in this case In and 2 n emission in GDR region strongly dominates. However, a long tail of the multiple neutron emission also exists in this channel. Our expectation of the intense neutron emission in the charge changing electromagnetic , collisions of heavy nuclei is very different from a simple picture assumed in Ref. [ l l ]where the rates of neutron and proton emission above the photoproton threshold were determined by Ar/Z ratio of the nucleus which undergoes fragmentation. A detailed understanding of the fragmentation mechanism may be obtained by measurements of fragment masses and detection of neutrons produced in collisions of ultrarelativistic heavy ions.
-
4. CONCLUSIONS
A variety of fragmentation mechanisms takes place in the electromagnetic dissociation of ultrarelativistic heavy ions, namely the coalescence, Fermi break-up, evaporation, fission and multifragmentation. A partial or complete disintegration of the colliding nuclei is possible despite the absence of geometrical overlap of the nuclear densities. Comparison with CERN SPS data on PbPb collisions demonstrates the dominance of electromagnetic dissociation in the partial charge-changing fragmentation channels with AZ = -6, ..., -1. RELDIS model predicts also a very intense neutron emission in such channels. The author gratefully acknowledges the fruitful collaboration with A.S. Botvina, J.P. Bondorf, A.S. Iljinov and 1.N. Mishustin on the subjects of this talk. Discussions with A.B. Kurepin, G. Giacomelli, M. Giorgini, L. Patrizii and P. Serra are greatly appreciated. The author is very indebted to the Organizing Committee of the KEK-Tanashi Symposium for the kind hospitality and financial support.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
C.A.Bertulani and G.Baur, Phys. Rep. 163, 299 (1988). G.Baur, K.Hencken, D.Trautmann, J . Phys. G24, 1657 (1998). F.Krauss, M.Greiner and G.Soff, Prog. Part. Nucl. Phys. 39, 503 (1997). I.A.Pshenichnov, I.N.Mishustin, J.P.Bondorf et al., Phys. Rev. C 60, 044901 (1999). A.S.Iljinov, I.A.Pshenichnov, N.Bianchi et al., Nucl. Phys. A616, 575 (1997). J.P.Bondorf, A.S.Botvina, A.S.Iljinov et al., Phys. Rep. 257, 133 (1995). A.S.Sudov, A.S.Botvina, A.S.Iljinov, Nucl. Phys. A554, 223 (1993). I.A.Pshenichnov, I.N.Mishustin, J.P.Bondorf et al., Phys. Rev. C 57, 1920 (1998). K.Baba, I.Endo, H.Fukuma et al., Nucl.Phys. A444, 578 (1986). H.Dekhissi, G.Giacomelli, M.Giorgini et al., Nucl.Phys. A, 1999, in print. S.E.Hirzebruch, E.Becker, G.Hiintrup et al., Phys. Rev. C 51, 2085 (1995).
IV. STRANGENESS PHYSICS
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
"
Kaon photoproduction on nuclei
" Physics Department, Graduate School of Science, Tohoku University, Sendai 9808587, Japan The photoproductions of a A through the 12C(y,K + ) and 3He(y,KO) reactions were studied near the threshold region. The experimental data were compared with a numerical calculation in terms of the quasifree A photoproduction. The cross sections were also discussed by comparing with the different model predictions. We discuss the use of polarized photon beams for the (y, K) reactions. It enables us to study the elementary kaon photoproduction amplitudes in detail. 1. INTRODUCTION
Photonuclear reactions associated with a strangeness degree of freedom, such as (y, KO) and (7,K + ) reactions, leave S = -1 in nuclei converting a nucleon into a hyperon. A hyperon production in a nucleus provides a unique opportunity to study fundamental information on the nucleon-hyperon and the nucleus-hyperon interactions.[l] They can be studied by use of the quasifree (QF) process and the hyper-nucleus spectroscopy. Until now, these investigations have been done using hadoronic probes. In particular, the ( T + , K + ) reactions have been used extensively to study the production and decay of A-hypernuclei. [2] The A-producing (7,K + ) and (y, KO) reactions are believed to be complementary to the hadoronic reaction.[3] The dominance of the spin-flip amplitudes in (y, K ) near threshold region enables us to study the spin dependent behavior of a A in nuclear potential. Since the mean free path of y in nuclear medium is much longer than any hadrons, we can probe A in nuclei with less distortion. In spite of the importance of (y, K ) measurements on nuclear targets, suitable y beams have not been available. High duty and high intensity y beams in GeV-region are required to study the A photoproductions because of the small (7,K ) cross sections and the requirements of the photon tagging. In this article, we describe the 12C(y,K + ) and 3He(y,KO) data near the threshold region. These measurements are the first observation of the K + and KO photoproduction on nuclei using the tagged photon beam. The experimental data will be introduced and compared with theoretical interpretations. Finally, we notice the importance of the detailed study of the elementary kaon production amplitudes, which can be investigated by using the polarized photon beams. They can be done at the Laser-Electron-Photon facility at SPring8.141 They will give us a complete data set for the elementary (7,K ) reaction amplitudes.
2. TAGX EXPERIMENTS 2.1. K+ photoproduction at TAGX 2.5
I
I
I
,
1
Inclus~vcI2C(y.K+)Cross Section
.
8
I
. I GcV
+-1.0-1
50
0 0.7
0.8
0.9
1.0
I 1
1.2
E., (GcV)
Figure 1. Differential cross section for 12C(y,K + ) reaction. TAGX data were compared with the elementary cross section and the numerical estimates in terms of the quasifree process.
11.2
1
1
114
115
11.6
11.7
Missing Mass ( G C V I C ~ )
Figure 2. Missing mass distribution for 12C(y,K t ) reaction in the photon energy range of 1.0 - 1.1 GeV. Lines represent the results of the Monte Carlo calculation. The arrow indicates the mass of the i2B ground states.
The 12C(-y,K + ) reaction was carried out using the TAGX spectrometer[5] at the 1.3GeV Tokyo Electron Synchrotron Laboratory.[6] The incident photon energy range was E, = 0.78 - 1.1 GeV. The charged particle events were momentum analyzed from the hits of Si-Strip Counters and Cylindrical Drift Chambers in the magnetic field. The timing and triggering counters were inner- and outer-scintillation-counter hodoscopes, and a time-offlight wall. They covered the angular ranges of 10 to 40 degrees on the left-hand side and 0 to 30 degrees on the right-hand side. A typical momentum resolution was 6% for protons of 0.4 GeV/c. Figure 1 shows the obtained differential cross section. Solid squares show the TAGX data compared with an average of the elementary cross sections multiplied by a factor of four. Thick curves show the Monte Carlo calculations, in which the QF process is assumed. The proton momentum distribution were taken from the 12C(e,e'p) data[7]. The dot-dashed curves represent the contributions of two protons in the p shell and four protons in the s shell, respectively. The thick solid line shows the sum of these two contributions. We can understand that the nuclear Fermi motion causes the K+ yield to rise more slowly near the threshold than the elementary process. Figure 2 is the missing mass distribution for the kaon events in the energy interval of 1.0 - 1.1 GeV. The curves show the QF calculation. We can see a broad peak structure centered a t 11.55 GeV/c2. It is reproduced well by the sum of the kaons from p and s
shells through the QF process. We find an excess of the data above the QF calculation in the missing mass region below 11.4 GeV/c2. The bound A states are expected in this region. The integrated differential cross section over this region is 0.21 f O.OSpb/sr. The effective proton number can be interpreted as a number of the participating nucleons in the reaction process. Since a proton changes to a A, the effective nucleon number is replaced by an effective proton number in (y, K + ) process. Our result of effective proton number is 4.2 f 0.6. It suggest the (y, K + ) is a useful tool to probe the nuclear interior with less distortion. Theoretical investigations of 12C(?.K + ) have been done in terms of the quasi-free process leading to an unbound A with a three body model,[8,9] and an intranuclear cascade model[lO]. Lee et al. [8] predicted both the inclusive and exclusive 12C(y,K + ) cross sections by using the recently developed amplitudes[ll] of yp -t K+A and wave functions from the relativistic mean-field theory[12]. They obtained O.19pb/sr for the sum of the cross sections for all bound A states. which is close to the experimental value 0.21 f0.05pb/sr. The inclusive cross section in the energy region up to E, = 1.1GeV was calculated based on the three body model. Their numerical estimates are shown in Figure 1. The thin-dotted line is the calculated cross section of the exclusive 12C(y,K+)i2B. The thin-dashed curve is the cross section of the QF 12C(y,K + ) X process. The thin-solid curve is the sum of these two contributions. The QF A production dominates the (y, K') process in the threshold region. A reduction factor 2.2 was used to fit the data. It is mainly due to the medium effects on y, K + and A. When we use the experimental value of the total y N and K + N cross section. the reduction factor become R 1.6. It suggest that the nuclear medium effects on the A propagation in nuclei must be significant. Another numerical approach is an intranuclear cascade model. [lo] It describes the y A reactions in the framework of a two-step cascade process. The first step is a rapid process. The intranuclear cascade develops through the binary collisions. The second step is the decay process of the excited nucleus. They include both the meson production channel and multicollisional intranuclear cascade process. The general behavior of the cascade calculation is in good agreement with TAGX data. The subthreshold cross section can be explained by Fermi motion and effective mass of bound nucleons.
-
-
+
2.2. KO photoproduction at TAGX The KO photoproduction on3He were measured using the TAGX spectrometer. The invariant mass spectrum constructed from two pion events is shown in Figure 3, where the photon energy range is above the reaction threshold for the elementary KO photoproduction and the vertex cut was employed outside of the target cell. We can identify the KO signature at -- 0.5 GeV/c2 in the invariant mass spectrum. In order to estimate the experimental cross section, the TAGX Monte Carlo was used to determine the detection efficiency.[5] We obtained the cross section 0 = 0.69pb for Ko-short photoproduction on 3He. Therefore the g , , ~ o becomes to about 1.4pb, which is similar to a,,K+ at E, 1 GeV. The differential cross section is shown in Figure 4. Although the K+ cross sections are strongly peaked in forward direction, the TAGX data does not show the forward peaking. It is understood by a lack of the t-channel diagram in the elementary KO photoproduction. The curves in Figure 4 are the numerical calculation using a computer code provided
-
Target (Full - Empty)
3
: 3 ~ e ( y , ~ 0TAGX ) DATA : E,= I . I GeV M=499.38 MeVlc2
h
L
25 tn
V
500
1000
MASS (MeV)
Figure 3. Invariant mass spectrum constructed from two pion events. The shadow indicates the KO events a t M -0.5GeV/c2.
21 -
:
-
-
AS1 C4 ....... AW4
.
0 0
60
1 20
180
ec.~.
Figure 4. Differential cross section for 3He(y,KO). Lines show the numerical estimates with different models.
by Sotona.[l3] Input parameters for the production amplitudes are obtained from K + photoproduction models.[l4-161 It is reported that the predicted (y, K f ) cross section using these different models show rather similar behavior in the 1 GeV energy region. In Figure 4, different model predictions denoted by AS1[14], C4[15] and AW4[16] are compared with experimental data. The numerical values of the (y, KO) cross sections are strongly model dependent in contrast to the K + photoproduction. 3. KAON PHOTOPROCUCTION WITH POLARIZED PHOTON BEAMS
In the A producing kaon photoproduction, a spin 0 kaon is photoproduced on a spin 112 nucleon leading to a spin 112 A. There are eight possible spin states in the system. Therefore the elementary kaon photoproduction process can be expressed with four independent amplitudes. The defined experimental observables in (y, K ) are differential cross sections, three single polarization asymmetries of P(recoil), C(beam) and T(target), and twelve double polarization asymmetries. It is sufficient to measure eight observables among them t o complete the data sets for the analysis of elementary N(y, K)A processes. The Laser-Electron-Photon facility a t Spring8 provides almost 100% polarized photon beams.[4] The photon beams are tagged in an enegy range from 1.5 to 2.4 GeV with the energy spread of 15 MeV. The average intensity is 107/sec which is limited by the light source operation. A large acceptance charged particle spectrometer will be completed in 2000. In this condition, we have a chance to measure the elementary A producing kaon photoproduction observables, doldS2, P, C and T in an energy range of E, = 1.5 - 2.4 GeV. It must be the most important information to study the kaon photoproduction on nuclei.
4. S U M M A R Y We measured the differential cross sections of the 12C(y,K + ) and 3He(y,KO) reaction in the threshold region at 1.3-GeV Tokyo electron synchrotron laboratory. The 12C(y,K + ) cross sections are compared with numerical estimates and theoretical interpretations. We found that the accurate study of the elementary (y, K ) process and nuclear medium effects of y, kaons, A propagation must be important for the future investigation of kaon photoproduction on nuclei. We noted that the measurements of the spin observables in (y, K ) is very effective to understand the kaon photoproduction. They can be done at new Laser-Backward-Compton facility in the near future. They will give us important information to study the A-nucleus interaction and the structure of the A-hypernuclei. REFERENCES 1. S.S. Hsiao and S. R. Cotanch, Phys. Rev. C28, (1983) 1668, M. Sotona, et al. A547 (1992) 63c, and C. Bennhold, Nucl. Phys. 547 (1992) 79c. 2. O.Hashimoto, S.Ajimura, K.Aoki, H.Bhang, T.Endo, Y.Fujii, 0 .Hotchi, E.Hungerford, J.H.Kim, Y.D.Kim, T.Kishimoto, K.Koshino, K.Kubota, K.Maeda, T.Nagae, H.Noumi, Y.Ohta, K.Omata, H.Outa, H.Park, Y.Saito, T.Saito, Y.Sato, M.Sekimoto, T.Shibata, T.Takahashi, T.Tamagawa, H.Tamura, L.Tang, H.Tanita, M.Youn, Nucl.Phys. A629, (1998) 405c and references therein. 3. C. B. Dover, Nucle. Phys. A547 (1992) 27c. 4. T.Nakano, H.Ejiri, M.Fujiwara, T.Hotta, K.Takanashi, H.Toki, S.Hasegawa, T.Iwata, K.Okamoto, T.Murakami, J.Tamii, K.Imai, K.Maeda, K.Maruyama, S.Date, M.M.Obuti, Y.Ohashi, H.Ohkuma, N.Kumagai, Nucl.Phys. A629, (1998) 559c 5. TAGX Collaboration: Nucl. Instrm and Methods A376 (1996) 335. 6. K.Maeda, et all Nuclear Physics A577 (1994) 227c, and H.Yamazaki,, et al, Physical Review C 52, R1157 (1995). 7. J . Mougey, M. Bernheim, A. Bussikre, A. Gillebert, Phan Xuan Ha, M Priou, D. Royer, I. Sick and G. J. Wagner, Nucl. Phys. A262, 461 (1976). 8. T.S. Lee, Z.Y. Ma, B. Saghai and H. Toki, Phys. Rev. C58 (1998) 1558. 9. E. Ya. Paryev, private communication. 10. S. de Pina, E. C. de Oliveira, E. L. Medeiros, S. B. Duarte and M. Gonalves, Physics Letters B434 (1998) 1. 11. J . C. David, C Fayad, G. H. Lamot and B. Saghai, Phys. Rev. C53 (1996) 2613. 12. Z. Ma, J. Speth, S. Krewald, B. Chen, and A. Ruber, Nulc. Phys. A608 (1996) 305. 13. M. Sotona, K. Itonaga, T. Motoba, 0 . Richter and J . iofka, Nucl. Phys. A547, 63c (1992), and M. Sotona and S. Frullani, Prog. Theor. Phys. 1 1 7 (1994) 151. 14. R. A. Adelseck and B. Saghi, Phys. Rev. C42 (1990) 108. 15. R. Williams, Ch. R. Ji and S. R. Cotanch, Phys. Rev. C46 (1992) 1617. 16. R. A. Adelseck and L.E. Wright, Phys. Rev. C38 (1988) 1965.
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
"
Subthreshold and near threshold K f meson photoproduction on nuclei E.Ya. Paryev
"
"Institute for Nuclear Research, Russian Academy of Sciences, Moscow 117312, Russia The inclusive K + meson production in photon-induced reactions in the near threshold and subthreshold energy regimes is analyzed with respect to the one-step ( y N 4 K + Y , Y = A, C) incoherent production processes on the basis of an appropriate new folding model, which takes properly into account the struck target nucleon removal energy and internal momentum distribution (nucleon spectral function). Comparison of the model calculations of the K + differential cross sections for the reaction y C12 in the threshold region with the existing experimental data is given. The predictions for the K + total and differential cross sections from yC12 and yPb208 reactions at subthreshold and near threshold energies are provided.
+
An extensive investigations of the production of K + mesons in proton-nucleus reactions [I-91 at incident energies lower than the free nucleon-nucleon threshold have been carried out over the last years. There are essentially less extensive investigations [lo, 111 of the inclusive subthreshold kaon production in pion-nucleus reactions. Finally, the electromagnetic production of K + mesons on nuclei in the threshold region has up to now received very little consideration [12], probably, because of a lack of suitable facilities and associated detectors. The inclusive (y, K+) measurements on nuclear targets in the threshold region are planned to be conducted in the near future at the Continuous Electron Beam Accelerator Facility (CEBAF) [13, 141 using the tagged photon beam in the CLAS (the CEBAF Large Acceptance Spectrometer) detector system. It is clear that, in order to analyze the results of such measurements, a relevant formalism has to be developed. The main goal of the present work is to extend the spectral function approach [6, 91 that has been employed for the description of the measured total [l]and differential [4, 71 kaon production cross sections from pA collisions in the near threshold and subthreshold energy regimes to K+-producing electromagnetic processes as well as to study the role played by nucleon-nucleon correlations in these processes. In this paper we present the predictions for the K + total and differential cross sections from y C12 and y P b 2 0 8 reactions in the threshold energy region obtained in the framework of the first collision model [6, 151 based on nucleon spectral function and compare part of them with the available data. An incident photon can produce a K + directly in the first inelastic y N collision due to nucleon Fermi motion. Since we are interested in the bombarding energy region up to approximately 1.4 GeV, we have taken into account the following elementary processes which have the lowest free production thresholds (respectively, 0.911, 1.046 and
+
+
1.052 GeV):
y+p+ K++A,
In our method the K+ production cross section for yA reactions from primary reaction channels (1)-(3) can be expressed [6, 151 as the respective integral of the free inclusive elementary K+ production cross sections [15] and nucleon spectral function P(pt,E ) over the struck target nucleon momentum pt and removal energy E . The nucleon spectral function P(pt,E) represents the probability to find in the nucleus a nucleon with momentum pt and removal (binding) energy E. The specific expressions for the high momentum and high removal energy part (correlated part) of the nucleon spectral function as well as for the single-particle (uncorrelated) part of the one given in [6, 151 were used in our calculations of K+ production in yC and yPb collisions.
Figure 1. Differential cross sections for K+ production in y+C12 reactions in the angular domain lo0 O K + 2 40° in the lab system as functions of the laboratory energy of the photon.
<
Figure 1 shows a comparison of the calculated differential cross sections for the production of K+ mesons at the laboratory angles of 10' 5 OK+ 5 40' from primary yN + K+Y channels with the experimental data [12] for y C12 + K+ X reaction at the various bombarding energies. The solid and dot-dashed lines represent our calculations with the total nucleon spectral function for primary production processes (1)-(3) and (2), (3), respectively. The dashed line denotes the same as solid line, but it is supposed that the total nucleon spectral function is replaced by its correlated part. The arrow indicates the threshold for the reaction yp -t K+A occuring on a free proton. One can see that:
+
+
1) the contributions to the Kf production from the primary reaction channels (1) and (2), (3) with A and C particles in the final states are comparable at bombarding energies E, 2 1.2 GeV, whereas at lower incident energies the primary production process (1) is essentially more important than (2) and (3); 2) the kaon yield from the one-step Kt production mechanism is entirely governed by the single-particle part of the nucleon spectral function at all considered beam energies (0.8 GeV 4 E, 4 1.3 GeV), what makes difficult to extract the information on the high momentum and high removal energy components within the C12 target nucleus from the first kaon photoproduction experiment [12];
3) while our calculations for the one-step reaction channels (1)-(3) reproduce reasonably well the experimental data [12]in the energy region of interest, they nevertheless provide different energy dependence of the excitation function as compared to the data, what might be due to possible in-medium modifications of the elementary y N -+ K+Y reactions discarded in the present work.
Figure 2. Double differential cross sections for the production of K+ mesons at an angle of 10' in the interaction of 0.8, 0.9 and 1.3 GeV photons with the C12 nuclei as functions of kaon momentum.
Figure 2 presents the results of our calculations for the double differential cross sections for the production of K+ mesons at an angle of 10' in the interaction of photons with energies of 0.8, 0.9 and 1.3 GeV with C12 nuclei. The solid and dashed lines are our calculations for kaon production processes (1)-(3) with the use of total nucleon spectral function and its correlated part, respectively, and correspond from the top to the bottom 1.3, 0.9 and 0.8 GeV incident energy. It is seen that the main contribution to
the K+ production both at subthreshold and above the free yN threshold beam energies considered here comes from the use of the uncorrelated part of the nucleon spectral function in the calculation of the corresponding momentum-energy-averaged differential cross sections for kaon production, what makes rather difficult to extract the information on the correlated part of the n~icleonspectral function even through analysis of the experimental double differential cross sections for K+ production at adopted photon energies (E, L 0.8 GeV).
tL-li,
K'0 1
G(CeV1
Figure 3. Total cross section for K+ production in y C12 reactions as a function of the laboratory energy of the photon.
+
Figure 4. Total cross section for K+ production in y Pb208reactions as a function of the laboratory energy of the photon.
+
Finally, Figures 3 and 4 present the results of our calculations for the total cross sections for K+ production in y C12 and y + Pb208reactions, respectively. The arrows indicate the thresholds for the reactions yp -+ K+Co,yp --, K+A occuring on a free proton and the absolute production threshold. The rest of the notation is identical to that in Figure 1. It can be seen that in these cases the kaon yield from the one-step K+ production mechanism is almost completely determined by the correlated part of the nucleon spectral function only in the vicinities of the absolute reaction thresholds (at bombarding energies of E, 5 0.75 GeV).This conclusion is in line with our findings inferred above (cf. Figures 1 and 2) from the analysis of differential and double differential kaon production cross sections. The values of the total kaon production cross sections in the far subthreshold region (E, 5 0.75 GeV) are very small (in the range of 0.1-10 nb), but one should expect to measure these values at the CEBAF [13, 141. As in the preceding cases, the direct K+ production processes (2), (3) play a minor role in kaon production in y A interactions at beam energies of E, < 1.2 GeV. Thus, our results demonstrate that it is difficult to distinguish the contribution to K+ production in yA reactions from the high momentum and high removal energy part
+
of the nucleon spectral function which is generated by ground-state two-nucleon shortrange correlations inside the target nucleus with the measurements of the K+ total and differential cross sections at subthreshold photon energies. REFERENCES 1. V.P.Koptev, et al., ZhETF. 94 (1988) 1. 2. ASibirtsev and M.Buscher, 2.Phys. A347 (1994) 191. 3. W.Cassing, et al., 2.Phys. A349 (1994) 77. 4. M.Debowski, et al., 2.Phys. A356 (1996) 313. 5. ASibirtsev, et al., 2. Phys. A358 (1997) 357. 6. S.V.Efremov and E.Ya.Paryev, Eur. Phys. J. A1 (1998) 99. 7. A.V.Akindinov, et al., APH N.S., Heavy Ion Physics 4 (1996) 325. 8. A.Badala, et al., Phys. Rev. Lett. 80 (1998) 4863. 9. E.Ya.Paryev, Eur. Phys. J. A5 (1999) 307. 10. S.V.Efremov and E.Ya.Paryev, 2.Phys. A351 (1995) 447. 11. S.V.Efremov and E.Ya.Paryev, Z. Phys. A354 (1996) 219. 12. H.Yamazaki, et al., Phys. Rev. C52 (1995) R1157. 13. CEBAF experiment 91-014 ( C . E . Hyde-Wright spokesman). 14. B.A.Mecking, Nucl. Phys. A639 (1998) 559c. 15. E.Ya.Paryev, Preprint INR-lOO4/99, Moscow (1999); EPJA, in the press.
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with Electromagnetic Probes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
"
Phenomenological aspects of kaon photoproduction on the nucleon T. Marta* , S. Sumowidagdoa*, C. Bennholdbt and H. Haberzettlbt "Jurusan Fisika, FMIPA, Universitas Indonesia, Depok 16424, Indonesia bCenter for Nuclear Studies, Department of Physics, The George Washington University, Washington, D.C. 20052, USA Using an isobar model which can reproduce the existing experimental data of kaon photoproduction on the nucleon we investigate some related phenomenological aspects, i.e. the hadronic form factors, contributions of kaon-hyperon final states to the anomalous magnetic moment of the nucleon, and missing nucleon resonances. By reggeizing the appropriate propagators we extend the model to the higher photon energy regime. 1. INTRODUCTION
A wealth of new high-statistics data on elementary kaon photo- and electroproduction has recently become available in three isospin channels. Along with some new progress in the theoretical side this has made the field of kaon electromagnetic production to be of considerable interest. New models, span from chiral perturbation theory to the relatively simple isobar approach. have been proposed in the recent years as the SAPHIR collaboration made their precise data publicly available. Because the error-bars are sufficiently small, an interesting structure can be resolved in K'A total cross section. This leads to a critical question, as to whether the structure comes from less known resonances or other reaction channels start to open at the corresponding energy. In this paper we discuss some phenomenological aspects, which can be investigated by means of the isobar model. The model has been constructed by including three states that have been found to have significant decay widths into KA and K C channels, the 5 1 1 (1650), Pll (1710), and P13(1720) resonances, to fit all elementary data by adjusting some free parameters. which are known as coupling constants.
2. THE MODEL AND SOME PHENOMENOLOGICAL ASPECTS 2.1. Hadronic form factors Previous analyses of kaon photoproduction have never included a form factor at the hadronic vertex. However, since most of the present isobaric models diverge at higher energies, the need for such hadronic form factors has been known for a long time. Furthermore, it has been demonstrated that models which give a good description of the (7,K + ) data can give unrealistically large predictions for the (y, KO) channels [I]. It 'Supported in part by the University Research for Graduate Education (URGE) grant. +Supported in part by US DOE with grant no. DE-FG02-95ER-40907
is well known that incorporating a hadronic form factor helps alleviate this divergence and, simultaneously, leads to a problem with gauge invariance, since not every diagram in the Born terms retains gauge invariance by i t ~ e l f .The ~ question of gauge invariance is actually one of the central issues in dynamical descriptions of how photons interact with hadronic systems. While there is usually no problem at the tree-level with bare, pointlike particles, the problem becomes very complicated once the electromagnetic interaction is consistently incorporated within the full complexity of a strongly-interacting hadronic system. In the previous work [2] we have studied the influence of hadronic form factors on kaon production, by multiplying the whole amplitude with an overall, monopole, form factor F ( t ) , i.e.
where the subscripts B and R refer to the Born and resonance terms, to simulate the average effect of the fact that nucleons are not point-like. In spite of the success to suppress the divergence and to avoid the problem of gauge invariance, this ad hoc fashion does not have any microscopic foundation. In order to restore gauge invariance properly, one needs to construct additional current contributions beyond the usual Feynman diagrams to cancel the gauge-violating terms. One of the most widely used methods is due to Ohta [3]. For kaon photoproduction off the nucleon, Ohta's prescription amounts to dropping all strong-interaction form factors for all gauge-violating electric current contributions in Born terms. Symbolically, this may be written as
The recipe, however, does not completely solve the problem of divergence, since the electric terms do not have suppression and, therefore, could violently increase as a function of the coupling constants. As shown in Ref. [4], even at the coupling constants values accepted by the SU(3) symmetry, Ohta's recipe already yields very large x2. On the other hand, Haberzettl [5] has put forward a comprehensive treatment of gauge invariance in meson photoproduction. This includes a prescription for restoring gauge invariance in situations when one cannot handle the full complexity of the problem and therefore must resort to some approximations. In our language, this method can be translated as
+
+ +
with P(s,t, u)= a l F ( s ) a 2 F ( t ) + a3F(u) and al a2 a3 = 1. Clearly, Haberzettl's method removes the Ohta's problem by an additional form factor in the electric terms. By fitting to the kaon photoproduction data we found that the method proposed by Haberzettl t o be superior rather than the Ohta's, since the former can provide a reasonable description of the data using values for the leading couplings constants close to the SU(3) prediction. Such couplings cannot be accommodated in Ohta's method due to the absence of a hadronic form factor in the electric current contribution. 3 ~ i n c ethe resonance terms are individually gauge invariant, the discussion will be limited t o the Born terms.
Table 1 Numerical values for the contribution of kaon-hyperon final states to the square of anomalous magnetic moments of proton and neutron. Column ( 1 ) is obtained from Eq. ( 5 ) ,while column ( 2 ) is evaluated by using Eq. ( 6 ) . Experimentally, K: = 3.214 and K: = 3.660.
n;(K) Channel yp-+K+A yp K+CO y p --+ K°C+ Tot a1 -+
(1) -0.026 -0.024 -0.013 -0.063
&i(K) (2) 0.044 0.030 0.031 0.105
Channel y n -+ K O A y n -+ K+Cy n -+ K°CO Tot a1
(1) 0.075 -0.025 -0.019 0.031
(2) 0.110 0.050 0.031 0.191
2.2. The anomalous magnetic moment of the nucleon One of the important ground state properties of the nucleon is the anomalous magnetic moment, which exists as a direct consequence of its internal structure. More than 30 years ago Gerasimov, and independently Drell and Hearn, proposed that this ground state property is related to the nucleon's resonance spectra by a sum rule which was then called the Gerasimov-Drell-Hearn (GDH) sum rule [6]. In the limit of photon point, the sum rule may be written as
where g3/2 and 0112 denote the cross sections for possible combinations of the nucleon and photon spins. Experiment with polarized beam and target has been performed at MAMI with photon energy up to 850 MeV and data are being analyzed [7].Using higher photon energies, experiments have been planned at ELSA and JLab. For practical purpose, instead of Eq. ( 4 ) we use
where OTT, denotes the cross section with polarized real photon and target. In terms of polarization observables this cross section corresponds to the double polarization E [8]. Since there are no data available for OTTI, previous work [9]approximated Eq. ( 5 ) with
to estimate the upper bound of contributions, where OT represents the total cross section. To calculate Eqs. ( 5 ) and ( 6 ) we use our elementary operator with urn, = 2.2 GeV. The result is shown in Table 1. Our calculation yields values of K ; ( K ) = -0.063 and n : ( K ) = 0.031, or I n P ( K ) j / n p5 0.14 and n , ( K ) / n , 5 0.094. This shows that the kaonhyperon final states contributions to the proton's and neutron's magnetic moment are very small. An interesting feature is that our calculation yields a negative contribution , is obviously consistent for the K ; ( K ) and a positive contribution for the K ~ ( K )which with the result of Karliner's work [ l o ] .
Figure 1. Total cross section for K f A photoproduction on the proton. The dashed line shows the model
3.0
without the D13(1960) resonance, while the solid line is obtained by including the D13(1960) state. The new
52
5.
V
,2.0
-
-
0
0
SAPHIR data [12] are denoted by the solid squares, old data are shown by the open circles.
1.0 -
1.6
1.7
1.8
1.9
2.0
2.1
2.2
W (GeV)
2.3. Investigation of missing resonances A brief inspection to the particle data book reveals that less than 40% of the predicted nucleon resonances are observed in .irN + .irN scattering experiments. Quark model studies have suggested that those "missing" resonances may couple strongly to other channels, such as the KA and K C channels [ l l ] . Interestingly, the new SAPHIR total cross section data [12] for the p(y, K+)A channel, shown in Fig. 1, indicate for the first time a structure around W = 1900 MeV. Using the current isobar model we investigate this structure. As shown in Fig. 1, our previous model cannot reproduce the total cross section. Although a structure in total cross section data does not immediately imply a new resonance, the energy region around 1900 MeV represents a challenge not only because of possible broad, overlapping resonances, but also because there are additional production thresholds nearby, such as photoproduction of q', K *A, and KA* final states, which can all lead to structure in the K+A cross section through final-state interaction. Here, we limit ourselves only to the possibility that this structure is in fact due to one of the missing or poorly known resonances. The constituent quark model of Capstick and Roberts [ll]predicts many new states around 1900 MeV. However, only a few of them have been calculated to have a significant KA decay width [ l l ] . These are the S11(1945),P11(1975),P13(1950),and D13(1960) states. We have performed fits for each of these possible states, allowing the fit to determine the mass, width and coupling constants of the resonance. We found that all four states can reproduce the structure at W around 1900 MeV, while reducing the x2, but only for the D13(1960) state we found a remarkable agreement, up to the sign, between the quark model prediction and our extracted result [13]. The result is shown in Fig. 1, where without this resonance the model shows only one peak near threshold, while inclusion of the new resonance leads to a second peak at W slightly below 1900 MeV, in accordance with the new SAPHIR data. The difference between the two calculations is much smaller for the differential cross sections. The largest effects are found in the photon asymmetry. Therefore, we would suggest that measuring this observable is well suited to shed more light on the contribution of this state in kaon photoproduction.
Figure 2. Total cross section for kaon photoproduction on the proton. The dashed line shows the isobar model with hadronic form factors, but without reggeization. The solid line is obtained by reggeizing the K and K* propagators in the model. For K+ photoproduction, notation for the data is as in Fig. 1, for KO production old (open circles) and preliminary data (solid circles) are shown [14]. 2.1
2.2
W (GeV)
3. EXTENSION TO HIGHER ENERGIES
Extending the model t o the higher energy regime requires a non-trivial task, since the Born terms increase rapidly as a function of energy. As shown in Fig. 2, even the hadronic form factors are unable to suppress the cross sections for the energy region above 2 GeV as demanded by the data. Especially in the case of K°C+ production, where the predicted cross section starts to monotonically increase at this point. However, in order to explore the higher-lying nucleon resonances or to account for higher energies contributions to the GDH integral, an isobar model which also properly work at higher photon energies would be demanded. In Ref. [14] it has been shown that the contributions from the t-channel resonances are responsible for the divergence of the cross section, thus indicating that the Regge propagatlor should be used instead of the usual Feynman propagator. While a proper reggeization of the model is considerably complicated and the study is still underway, we investigate here only the qualitative effects of using Regge propagators in the model. Following Ref. [15], we multiply the Feynman propagators l / ( t - m i . ) of the K*(892)
and K1(1270) resonances in the operator with a factor of P R ~(t ~ m$.), ~ ~ where . PRegge indicates the Regge propagator given in Ref. 1151. For the K * intermediate state it has the form
where a ( t ) = a0 +a' t denotes the corresponding trajectory. Equation (7) clearly reduces to the Feynman propagator in the limit of t -+ mg., thus approximating the low energy behavior of the amplitude. The model is then refitted to kaon photoproduction data and the result is shown in Fig. 2, where we compare the isobar model with and without reggeization. Obviously, Regge propagators strongly suppress the cross section at high energies and, therefore, yield a better explanation of data at this energy regime. For the K°Cf process, the use of Regge propagators seems to give more flexibility in reproducing the cross section data. This cannot be achieved without reggeization, since the high energy behavior of both t-channel resonances is less controllable by the hadronic form factors. However, since the data for the K°C+ channel shown in Fig. 2 are still preliminary [16],we have to wait before any further conclusion can be drawn. In future we will include the high energy data in the fit and investigate the model in the transition between medium and high energy regions.
REFERENCES 1. T. Mart, C. Bennhold, and C.E. Hyde-Wright, Phys. Rev. C 51 (1995) R1074. 2. C. Bennhold, T. Mart, and D. Kusno in Proceedings of the CEBAF/INT Workshop on N* Physics, Seattle, USA, 1996 (World Scientific, Singapore, 1997), p.166. 3. K. Ohta, Phys. Rev. C 40 (1989) 1335. 4. H. Haberzettl, C. Bennhold, T. Mart, and T. Feuster, Phys. Rev. C 58 (1998) R40. 5. H. Haberzettl, Phys. Rev. C 56 (1997) 2041. 6. S.B. Gerasimov, Sov. J. Nucl. Phys. 2 (1966) 430; S.D. Drell and A.C. Hearn, Phys. Rev. Lett. 16 (1966) 908. 7. See W. Meyer, these proceedings. 8. R.L. Workman and R.A. Arndt, Phys. Rev. D 45 (1992) 1789. 9. D. Drechsel, Prog. Part. Nucl. Phys. 34 (1995) 181; H.-W. Hammer, D. Drechsel, and T. Mart, nucl-th/9701008; S. Sumowidagdo and T. Mart, Phys. Rev. C 60 (1999) 028201. 10. I. Karliner, Phys. Rev. D 7 (1973) 2717. 11. S. Capstick and W. Roberts, Phys. Rev. D 58 (1998) 074011. Tran et al., Phys. Lett. B 445 (1998) 20. 12. SAPHIR Collaboration: M.Q. 13. T . Mart and C. Bennhold, Phys. Rev. C 61 (2000) 012201(R). 14. F.X. Lee, T. Mart, C. Bennhold, H. Haberzettl, and L.E. Wright, nucl-th/9907119 and references therein. 15. M. Guidal, J.M. Laget, and M.Vanderhaeghen, Nucl. Phys. A 627 (1997) 645. 16. By the time of finishing this paper we realized that the final version of SAPHIR data for the y p + K°C+ channel has been published in: S. Goers et al., Phys. Lett. B 464 (1999) 331.
+
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Hyperon polarization in Kaon photoproduction from the deuteron H. Yamamura", K. Miyagawa ", T. Mart W . Glockle
b,
C. Bennhold ', H. Haberzettl 'and
"Department of Applied Physics, Okayama University of Science, 1-1 Ridai-cho, Okayama 700, Japan bJurusan, Fisika, FMIPA, Universitas Indonesia. Depok 16424, Indonesia 'Center for Nuclear Studies, Department of Physcs, The George Washington University, Washington, D.C. 20052 dInstitut fur Theoretische Physik 11, Ruhr-Universitat Bochum, D-44780 Bochum, Germany
+
+
+
We analyze the reaction y d -+ K + A(C) N using the Nijmegen soft core hyperonnucleon interactions NSC97f and NSC89 and a recently updated production operator for the y N 4 K + A(E) processes. Significant effects of the Y N final state interaction are found near both the K + A N and K + E N thresholds.
+
+
1. INTRODUCTION
Since hyperon-nucleon scattering experiments are difficult to perform, hyperon producY N appear as tion processes such as y d -t K + Y N and e + d -+ e' K t natural candidates that allow exploring the Y N interaction. One can obtain the information of the Y N interaction by analyzing the correlated Y N final states. An inclusive d(e, e l K + ) Y N experiment has already been performed in Hall C at TJNAF, while the data for d(y. K + Y ) N are being analyzed in Hall B. Recently, we found that various meson-theoretical Y N interactions generate S-matrix poles around the AN and C N thresholds[l]. The pole near the E N threshold is related to the strength and the property of the AN - C N coupling and causes enhancements in the AN elastic total cross sections. The hope is that the pole structure of the Y N t matrix will have visible effects in such production processes mention above. In this paper, we study the inclusive d(y, K + ) Y N and exclusive d(y, K + Y ) N processes for OK = 0' and predict various observables including polarization observables.
+
+ +
+
+ +
2. FORMALISM
The reaction processes y
TiIQd>=
1Utj . t y i ' ~ >, j
+d
7-
K+
+ A(C) + N are expressed by the operator Tias
i, j = Ah[, EN,
= Figure 1. Inclusive d ( y ,K + ) cross section as a function of lab momentum p ~ for + 0" and the photon lab energy E, = 1.3 GeV. The two thresholds K f A N and K + C N are indicated by the arrows. The results around the K + C N threshold are enlarged in (b).
where the operator t y i describes the elementary processes y+ N + K++A(C), and j9,j > represents the deuteron state which is generated by the Nijmegen 93 N N interaction [2]. The operator Uijcorresponds to the Y N final state interaction processes, and is represented as
where V,, is Y N interaction including AN - C N coupling. We ignore the K + meson interaction with the nucleon and hyperon in the final states. From Eqs.(l) and (2), one can deduce the coupled set of integral equations for Ti,
We solve this set (3) after partial-wave decomposition in momentum space. The three elementary process y + p -+ K+ + A(Co) and y + n --+ K + + C- are properly incorporated in the driving term in Eq.(3). Equation (3) is solved on isospin bases A N and E N , but the resulting amplitudes are transformed into those on the particle bases An, Con and C - p by which the inclusive d(y, K + ) , exclusive d(y, K + Y ) cross sections and hyperon polarizations are calculated. For details, we refer the reader to ref.[3]. 3. RESULTS
At present, we calculate the observables only for the K+ meson scattered to 0 degree = 0"). The Nijmegen soft-core Y N interactions NSC97f[4] and NSC89[5] and a
(OK+
+
+
+
recently updated production operator[6] for the y N -+ K+ A(C) N processes are used. Figure l ( a ) shows the inclusive cross sections which sum up the contributions of the K'An, K + C O nand K+C-p final states. The solid and dashed lines are the predictions of the NSC97f and NSC89 Y N interactions, respectively. The dotted line shows the results of the plane wave impulse approximation (PWIA). The arrows indicate the two thresholds K + A N (pK = 977.30 MeV/c) and K + C N (pK = 869.14 MeV/c). The two pronounced peaks around p~ =945 and 809 MeV/c are due to the quasi-free processes of A and C, where one of the nucleon in the deuteron is spectator and has zero momentum in the laboratory system. Significant FSI effects are found around the K + A N and K f C N thresholds. The cross section is increased up to 86% by FSI near the K + A N threshold. Around the K f C N threshold, as shown Fig.l(b), the strengths and shapes of the enhancements by the NSC97f and NSC89 are quite different.
Figure 2. (a) Exclusive d ( y , K+A) cross section and (b) A recoil polarization with incoming polarized photon as a function of the A scattering angle in the An c.m. system. The photon lab energy is E, = 1.3 GeV. The outgoing kaon lab momentum and angle are p ~ =+ 870 MeV/c and 8 ~ =+ 0" respectively.
Figure 2(a) illustrates the exclusive d(y, K f A ) cross sections just below the K f C N threshold ( p K = 870 MeV/c). The FSI effects are seen both at very forward and at large angles. The PWIA cross sections are basically zero at backward angles, while the FSI calculations still show some strength. Figure 2(b) demonstrates the A recoil polarizations with incoming polarized photon. The A recoil polarizations in PWIA are almost one. This is because the incoming photon is polarized along the z axis, but the target deuteron is unpolarized and the outgoing K + meson carries no spin and angular momentum in this case (OK = 0"). However, the final state interactions cause the large deviations from one, and the prediction by NSC97f is quite different from that of NSC89.
- --
0.08 -
1
,
,
,
's
hN .-.,
-- NSC89
8
-.----------
-
0 0
I
\
u
0.02
,
- - - PWIA - NSC97f '
-
(b)
(a)
0.1
, 90
[degl
- , - -- , - . . 180
o
0
-
l
'
'
90
0,-,
'
'
[degl
Figure 3. (a) Exclusive d(y. K+C-) cross section and (b) C- recoil polarization with incoming polarized photon as a function of the C- scattering angle in the C-p c.m. system. The photon lab energy is E, = 1.3 GeV. The outgoing kaon lab momentum and angle are PK+= 865 MeV/c and Q K + = 0' respectively.
Figure 4. Inclusive d(y, K + ) cross section in the PWIA as a function of lab momentum p ~ + for Q K + = 0' and photon lab energy E, = 1.3 GeV. The solid curve shows the prediction by the present version[6] of the production operator for y N -t K+ + A @ ) , while the dashed line corresponds to the prediction by an old version[7].
+
The exclusive results to the K+C-p final states just above this threshold (pK =865 MeV/c) are shown in Fig.3. The prominent FSI effects are seen both in the cross sections and in the double polarization observable. Finally, we briefly discuss the production operator for 7 N -+ K+ A(C) processes. In Fig.4, the inclusive cross sections in PWIA in which an old version[7] of the operator is used are compared to those with the present version[6]. The latter has been improved in the fitting to the data of y + N -+ K+ + A(C) including the new SAPHIR data. The difference between the predictions by the two versions are quite large as in Fig.4, which suggests this reaction d(y, K + ) Y N is another promising candidate for investigating the operator.
+
+
REFERENCES 1. K. Miyagawa, H. Yamamura, Phys. Rev. C60 (1999) 024003; nucl-th/9904002. 2. V. G. J. Stoks, R. A. M. Klomp, C. P. F . Terheggen, and J. J. de Swart, Phys. Rev. C49 (1994) 2950. 3. H. Yamamura, K. Miyagawa, T . Mart, C. Bennhold, W. Glockle, Phys. Rev. C61 (1999) 014001; nucl-th/9907029. 4. Th. A. Rijken, V. G. J. Stoks, and Y. Yamamoto, Phys. Rev. C59 (1999) 21. 5. P. M. M. Maessen, T h . A. Rijken, and J . J. de Swart: Phys. Rev. C40 (1989) 2226. 6. C. Bennhold, T. Mart, A. Waluyo, H. Haberzettl, G. Penner, T . Feuster, and U. Mosel, in Proceedings of the Workshop on Electron-Nucleus Scattering, Elba, Italy, 1998, edited by 0.Benhar, A. Fabrocini, and R. Schiavilla (Edizioni ETS, Pisa, 1999), p. 149; nucl-th/9901066. 7. C. Bennhold, T . Mart, and D. Kusno, in Proceedings of the CEBAF/INT Workshop on N* Physics, Seattle, USA, 1996, edited by T.-S. H. Lee and W. Roberts (World Scientific, Singapore,1997), p.166
ThisPageIntolyLfBk
ThisPageIntolyLfBk This Page Intentionally Left Blank
Hadron and Nuclear Physics with ElectromagneticProbes K. Maruyama and H. Okuno (Editors) 2000 Elsevier Science B.V. All rights reserved.
'
Physics of associated strangeness production at ELSA Ewald Paul Physikalisches Institut, Universitat Bonn, Germany 1. INTRODUCTION
Associated strangeness production has been measured by the SAPHIR Collaboration [I] with the SAPHIR detector [2] at the Electron Stretcher and Accelerator ELSA in Bonn. Results will be reported from about 6% of the data on tape [3,4]. 2. THE STRANGE PARTICLE PROGRAMME AT SAPHIR
The SAPHIR experiment is devoted to measuring photoproduction with a beam of real photons at energies following a bremsstrahlung spectrum from electrons up to 3 GeV from ELSA. The strange particle reactions measured are essentially those with two-body final states:
Results from these reactions [3,4] are presented and discussed in this talk. Data analyses are in progress for the reactions:
The measurements with SAPHIR allow events to be reconstructed and identified over the whole angular space of produced charged particles. The results consist of total and differential cross sections, and transverse polarization of the hyperons. They are compared to models in three distinct parts. In the first part we consider isobar models. They describe the data at the level of hadronic degrees of freedom by fitting a superposition of Born terms and resonant contributions to the data. Parameters which are common with other reactions are determined in a coupled-channel approach. In the second part, the question whether hyperon polarizations also have a more general origin than stemming from interference of specific amplitudes is discussed. This is suggested since A and C states are polarized in similar manner for various beams and over
orders of magnitude in energy [5]. Moreover hyperons indicate the polarization of the s quark within the framework of the static quark model. If the s-quark carries polarization this implies that the polarizations of A and C states are related to each other. This is what is observed in general. The third part is concerned with chiral model calculations based on an effective chiral Lagrangian and on chiral perturbation theory. Such perturbative calculations describe pion photoproduction processes near threshold. When going to kaon-hyperon final states the mass of the kaon, which is considerably larger than the mass of the pion, implies new questions concerning the perturbative calculations directly as well as the handling of chiral symmetry breaking in general.
,
\;:,
drift chambers ,
:, ,
P
,
Figure 1. Topological and kinematical reconstruction of an event of the type yp
+
K°C'
3. THE EXPERIMENT
The ELSA accelerator provides electron beams in the energy range 0.5 to 3.5 GeV with nearly 100% duty cycle. Such a beam was used to produce bremsstrahlung photons which were tagged one-by-one in a tagging system for SAPHIR. The SAPHIR detector comprises a large magnet with a target inside filled with liquid hydrogen or deuterium and surrounded by a cylindrical drift chamber. Downstream and to the sides are planar drift chambers and large scintillator hodoscopes. Further downstream follows an electromagnetic calorimeter, which was not used for the strange particle data analyses. Details of the experimental setup are given elsewhere [2]. The SAPHIR drift chamber system is well suited to measure tracks of charged particles