edited by S. Kox, D. Lhuillier, F. Maas, and J. Van de Wiele
Serge Kox Laboratoire de Physique Subatomique et de Cosmologie 53, Avenue des Martyres 38026 Grenoble Cedex, France kox(§iii2p3. f r
David Lhuillier CEA Saclay, DSM/DAPNIA/SPhN Bat. 703 91191 GifsurYvette, France dlhuillier(§cea. f r Frank Maas Institut fiir Kernphysik A4collaboration, Parity Violation Experiment JohannesGutenbergUniversitat Mainz J.J.BecherWeg 45 55099 Mainz, Germany maasOkph.unimainz.de
Jacques van de Wiele IPNO, Universite Paris Sud Bat. 100 M 91406 Orsay Cedex, France vandewi(§ipno. irL2p3. f r The articles in this book originally appeared on the internet (springeronline.com) as open access publication of the journal The European Physical Journal A — Hadrons and Nuclei Volume 24, Supplement 2 ISSN 1434601X © SIF and SpringerVerlag Berhn Heidelberg 2005 CataloginginPublication Data applied for Bibliographic information published by Die Deutsche Bibliothek. Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at http://dnb.ddb.de
ISBN10 354025501X Springer Berlin Heidelberg New York ISBN13 9783540255017 Springer Berlin Heidelberg New York This work is subject to copyright. All rights reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from SIF and Springer. Violations are liable for prosecution under the German Copyright Law. Springer is a part of Springer Science+Business Media springer online. com © SIF and SpringerVerlag Berhn Heidelberg 2005 Printed in Italy The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Typesetting: PTPBerlin, Berlin, Germany Cover design: SIF Production Office, Bologna, Italy Printing and Binding: Tipografia Compositori, Bologna, Italy Printed on acidfree paper
l'i'uin i^L\ni'j 7jDJrinoii .MLUM/. Jutir ."w. L?<)<>:^
i M l f IJ : ( i l ( t H i h l c JUIH' W1 I . LIOO J:
Almost 50 years after the proposal of Lee and Young in 1956 to test the hypothesis of parity violation in weak interactions and the subsequent experimental verification of parity violation by C.S. Wu, parity violation has today become a useful property of weak interactions. This is due to the fact that the focus nowadays has changed: parity violation in weak interactions is no more a topic of investigation but is used as a tool in many different fields ranging from nuclear physics to the search for the hidden extra dimensions requested by string theory. For our first workshop which took place June 58, 2002, at the Institut fiir Kernphysik of the Johannes GutenbergUniversitat Mainz, we concentrated on the investigation of the strangeness contribution in the nucleon. This book contains the refereed and selected papers of the second workshop "From Parity Violation to Hadron Structure and more (Part II)", which took place June 811, in the Laboratoire de Physique Subatomique et de Cosmologie, in Grenoble. These papers appear in EPJAdirect, the electroniconly part of EPJA, and they are accessible without restrictions. They will also appear in printed form and can be ordered through Springer. The excellent presentations show the dramatic and steady progress in the accuracy of measured parity violating asymmetries over the last few years. The neutral weak current appears as a powerful new probe of matter driving experimentalists and theorists towards new frontiers in the structure of the nucleon and in the exploration of the electroweak interaction and the physics beyond. Our thanks for making the workshop such a success go to Madame J. Riffault, who as the conference secretary organized all the nonscientific aspects of the workshop. Without her tireless effort the workshop would not have become such a great success of scientific exchange. The conference also received financial support from scientific and local organizations as well as sponsors. This has allowed the event to take place and the refereed and selected papers to be pubhshed. In addition the International Advisory Committee helped us to establish the scientific program and scope of the workshop. Thanks to all of them. There are new experiments going on; plenty of data have been taken and are now being analyzed. We are very much looking forward to the exciting future where
parity violation will bring many new insights into nature. We are currently planning part III of the workshop, which will be held on Milos Island in Greece in May 2006, where we hope to celebrate with all the participants the 50th anniversary of the discovery of parity violation. S. Kox D. Lhuillier F. Maas J. Van de Wiele Guest Editors
Damon T. Spayde Updated results from the SAMPLE experiment
Richard S. Holmes The next generation HAPPEX experiments
P.A.M. Guichon Two photon effects in electron scattering
B. Pasquini, M. Vanderhaeghen Single spin asymmetries in elastic electronnucleon scattering
M. Gorchtein, P.A.M. Guichon, M. Vanderhaeghen Transverse single spin asymmetry in elastic electronproton scattering
Sebastian Baunack, for the A4collaboration Transverse spin asymmetry at the A4 experiment
F.E. Maas, for the A4Collaboration Parity violating electron scattering at the MAM! facility in Mainz
Philip G. Roos The GO experiment: Parity violation in eN elastic scattering
Luigi Capozza, for the A4 Collaboration Study of the parity violation in the Z\(1232) region
Stephen F. Pate Don't forget to measure As D lll2/Theory
P.M. King for the G^ Cohaboration Normal beam spin asymmetries during the G^ forward angle measurement
D.B. Leinweber, S. Boinepalli, A.W. Thomas, A.G. Williams, R.D. Young, J.B. Zhang, J.M. Zanotti Systematic uncertainties in the precise determination of the strangeness magnetic moment of the nucleon
123 E. Chudakov, V. Luppov M0ller polarimetry with atomic hydrogen targets
J.T. Londergan Current status of parton charge symmetry
127 Yoshio Imai, for the A4 collaboration Progress report on the A4 compton backscattering polarimeter
M.E. Sainio Pionnucleon interaction and the strangeness content of the nucleon
A. Silva, D. Urbano, H.C. Kim, K. Goeke Strange form factors of the nucleon in the chiral quarksoliton model Bastian Kubis Strange form factors and Chiral Perturbation Theory
101 Stephen R. Cotanch Timelike compton scattering and the BetheHeitler process
103 G.W. Carter, E.M. Henley Corrections to the nuclear axial vector coupling in a nuclear medium
105 HeeJung Lee, Chang Ho Hyun, ChangHwan Lee, HyunChul Kim Strangenessconserving effective weak chiral Lagrangian
109 Gordon D. Gates, Jr. Overview of laser systematics
115 Douglas H. Beck, Mark L. Pitt Beam optics for electron scattering parityviolation experiments
129 Christoph Weinrich, for the A4 Collaboration The transmission Compton polarimeter of the A4 experiment
131 Jiirgen Diefenbach Stabilization system of the laser system of the A4 Compton backscattering polarimeter
133 Jeong Han Lee Electron beam line design of A4 Compton backscattering polarimeter
D IV3/Detection 137 J. Van De Wiele, M. Morlet Background substraction in parity violation experiments 141 Boris Glaser, for the A4 Collaboration Redesign of the A4 calorimeter for the measurement at backward angles 143 Steven E. Wilhamson, for the G^ Collaboration Performance of the G^ superconducting Magnet System 145 Benoit Guillon, for the G^ collaboration Cherenkov counter for the G^ backward angle measurements 147 L. Bimbot, for the G^ CoUaboration A binperbin deadtime control technique for timeofflight measurements in the G^ experiment: the differential buddy
V Hadronic structure and more Vl/Test of the SM at low energy 151 Antonin Vacheret, David Lhuillier, representing the E158 Collaboration A precise measurement of sin?6^^ at low Q^ in M0ller scattering 155 Gregory R. Smith, for the Qweak Collaboration Qweak: A precision measurement of the proton's weak charge 159 Klaus H. Grimm, for the Qweak Collaboration The qweak tracking system
179 C.H. Hyun, C.P. Liu, B. Desplanques Parity violating asymmetry i n 7 + d ^ ^ n + p a t low energy V3/Neutrino beam
183 A. Blondel Precision physics at a neutrino factory
187 Kevin S. McFarland The MINERz/A experiment at FNAL
161 Kevin S. McFarland Neutral currents and strangeness of the nucleon from the NuTeV experiment
191 Krishna S. Kumar Frontiers of polarized electron scattering experiments
197 M.J. RamseyMusolf 167 C.J. Horowitz Parity violation in astrophysics
171 B. Desplanques Parity violation in nuclear systems
205 R.D. McKeown Workshop summary
BAUNACK Sebastian Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz baunackOkph.unimainz.de
BREUER Herbert University of Maryland Department of Physics College Park, MD 20740, USA Breuer^enp.umd.edu
CAPOZZA Luigi BECIREVIC Damir LPT, Universite Paris Sud Bat. 210) F91405 Orsay Cedex
Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz capozzaOkph.unimainz.de
CARBONELL Jaume BECK Douglas UIUC, Department of Physics 1110 West Green Str. Urbana, IL 61801, USA dhbeck^uiuc.edu
BEISE Elizabeth University of Maryland Department of Physics College Park, MD 207424111, USA [emailprotected]
GATES Gordon University of Virginia Department of Physics 382 McCormick Rd. PO Box 400714 Charlottesville, VA 229044714, USA gdc4kOTirginia.edu
Jefferson Laboratory 12000 Jefferson Avenue, Newport News, VA 23606, USA [emailprotected]
Institut fiir Kernphysk Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz [emailprotected]
COTANCH Stephen North Carolina State University Department of Physics Raleigh, NC 276958202, USA cotanchOncsu. edu
DE JAGER Kees Jefferson Laboratory 12000 Jefferson Avenue Newport News, VA 23606, USA keesOjlab.org
DELOUDI Sofia ETH Zurich Laboratory of Physical Chemistry Honggerberg HCI E243 CH8093 Ziirich deloudiOphys.chem.ethz.ch
GORSHTEYN Mikhail Genoa University Dipartimento di Fisica Via Dodecaneso 33 116146 Genova gorshtey^ge.infn.it
GRIMM Klaus College of William and Mary Department of Physics PO Box 8795 Williamsburg, VA 231878795, USA grimmOjlab.org
GUICHON Pierre CEA Saclay SPhN / DAPNIA F91191 Gif sur Yvette Cedex pguichonOcea.fr
Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz def i^kph.unimainz.de
ELLIS Jonathan R. CERN European Organization for Nuclear Research CH1211 Genve 23 John.EllisOcern.ch
FURGET Christophe LPSC, 53 avenue des Martyrs F38026 Grenoble furget^lpsc.in2p3.fr
HOLMES Richard Syracuse University 201 Physics Building Syracuse, NY 13244, USA rsholmes^phy.syr.edu
HOROWITZ Charles Indiana University Department of Physics Swain Hall West, Bloomington, IN 47405, USA [emailprotected]
Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz [emailprotected]
Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz imai^kph.unimainz.de
University Manitoba / TRIUMF 4004 Wesbrook Mah, Vancouver BC V6T2A3, Canada IryleeOtriumf.ca
Jefferson Laboratory 12000 Jefferson Avenue MS 16B Room 96 Newport News, VA 23606 USA
KOX Serge LPSC, 53 avenue des Martyrs F38026 Grenoble Cedex kox(9in2p3. f r
KUBIS Bastian ITP, University of Bern Sidlerstrasse 5 CH3012 Bern kubisOitp.unibe.ch
KUMAR Krishna University of Massachusetts Department of Physics Amherst, MA 01003, USA kkumarOphysics.umass.edu
LENOBLE Jason IPNO, Universite Paris Sud Bat. 100 M F91406 Orsay Cedex lenobleOipno.in2p3.fr
LONDERGAN John Timothy Indiana University 1331 East 10th St. Bloomington, IN 47408, USA tlonderg^indiana.edu
MAAS FYank E. Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz maasOkph.unimainz.de
MARMONIER Carole Le GOFF JeanMarc CEA Saclay, DAPNIASPhN F91191 Gif sur Yvette Cedex [emailprotected]
PHOTONIS Av. Roger Roncier, BP 520 F19106 Brive c.marmonierOphotonis.com
McFARLAND Kevin LEE HeeJung Universitat de Valencia Departamente de Fisica Teorica E46100 Burjassot (Valencia) Heej ung.Lee^uv.es
University of Rochester Department of Physics and Astronomy RC Box 270171 Rochester, NY 146270171, USA [emailprotected]
Caltech, Kellogg Radiation Laboratory 10638 Caltech Pasadena, CA 91125, USA bmck^krl.caltech.edu
UIUC, Loomis Laboratory of Physics 1110 West Green Street, Urbana, IL 61801, USA nakaharaOjlab.org
Jefferson Laboratory 12000 Jefferson Avenue 12H/C 108 Newport News, VA 23606, USA
IPNO, Universite Paris Sud Bat. 100 M F91406 Orsay Cedex
OPPER AUena Ohio University, Physics and Astronomy Edwards Accel. Lab. Athens, OH 45701, USA opperOohiou.edu
PASQUINI Barbara University of Pavia Via Bassi 6 L27100 Pavia pasquini^pv.infn.it
PATE Stephen New Mexico State University Physics Department, Box 3D Las Cruces, NM 88003, USA pateOnmsu.edu
QUEMENER Gilles LPSC, 53 avenue des Martyrs F38026 Grenoble Cedex quemenerOlpsc.in2p3.fr
RAMSEYMUSOLF Michael Caltech, Kellogg Radiation Laboratory 10638 Caltech, Pasadena, CA 91125, USA mj rm^krl. c a l t e c h . edu
SAINIO Mikko Helsinki Institute of Physics PO Box 64 FIN00014 Helsinki mikko.sainioOhelsinki.fi
SILVA Antonio CFCFCT Universidade de Coimbra CFC Faculdade de Ciencias e tecnologia Rua Larga P3004516 Coimbra aj oseOteor.fis.uc.pt
SILVESTREBRAC Bernard LPSC, 53 avenue des Martyrs F38026 Grenoble cedex silvestrebracOlpsc.in2p3.fr
SMITH Gregory Jefferson Laboratory 12000 Jefferson Avenue 12H/C 108 Newport News, VA 23606, USA smithg®jlab.org
UIUC, Loomis Laboratory of Physics 1110 West Green Street Urbana, IL 61801, USA [emailprotected]
University of Athens, I AS A University Campus GR15771 Athens StiliarisOphys.uoa.gr
Institut fiir Kernphysik Johannes Gutenberg Universitat JohannJoachimBecherWeg 45 D55099 Mainz weinrich^kph.unimainz.de
L'Institut National de Physique Nucleaire et de Physique des Particules (IN2P3), le Centre National de la Recherche Scientifique (CNRS), TUniversite Joseph Fourier de Grenoble, le Laboratoire de Physique Subatomique et de Cosmologie de Grenoble (LPSC), le Service de Physique Nucleaire du Commissariat a TEnergie Atomique (SPhNCEA), rinstitut fiir Kernphysik et la Johannes GutenbergUniversitat de Mainz, la societe CAEN, la societe Photonis, la Region RhoneAlpes, le Conseil General de Tlsere, la Ville de Grenoble, I'Office du Tourisme de Grenoble, le reseau europeen ESOP
Today's view on strangeness^ J o h n Ellis CERN, CH1211 Geneva 23, Switzerland Received: 1 December 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. There are several different experimental indications, such as the pionnucleon U term and polarized deepinelastic scattering, which suggest that the nucleon wave function contains a hidden ss component. This is expected in chiral soliton models, which also predicted the existence of new exotic baryons that may recently have been observed. Another hint of hidden strangeness in the nucleon is provided by copious (j) production in various NN annihilation channels, which may be due to evasions of the OkuboZweigIizuka rule. One way to probe the possible polarization of hidden ss pairs in the nucleon may be via A polarization in deepinelastic scattering. PACS. 12.39.Dc Skyrmions  13.75.Cs Nucleonnucleon interactions  14.20.c Baryons
2 < _p55p > Some people might argue t h a t this is a "strange" question: y why should the nucleon be strange at all  after all, is it not < p\uu\p > + < p\dd\p > just made out of three u p and down quarks? We should not j u m p to such a naive conclusion. For a start, even the Two recent determinations of the 7rnucleon U t e r m have vacuum is strange: chiral symmetry for TT, K mesons tells found large values [7] : us t h a t [1] r = 64 =b 8, (79 =b 7) MeV < 0ss0 > = (0.8 zb 0.1) < 0gg0 > . corresponding to large values ofy = 1 —cro/i^, where octet This hidden strangeness cannot be expected to disap baryon mass differences give CTQ = 36=b7 MeV [8] and hence pear when one inserts a set of three quark 'coloured test y ^ 0.5. Another example is the strange spin of the nucharges' into the vacuum. Moreover, hidden strangeness cleon: a naive interpretation of measurements of polarized will be generated in perturbative Q C D : deepinelastic structure functions would yield [9]: quark ^ gluon ^ ss pair. There are also nonperturbative mechanisms for generating ss pairs in the nucleon, such as instanton effects [2]. Another objection to this 'strange' question is the fact t h a t (at least some) experiments do not see very much strangeness in the nucleon. For example, C C F R measures a strange m o m e n t u m fraction: Pg = 4% at Q 2 = 20GeV^ [3], the H A P P E X measurement of a combination of strange electric and magnetic form factors gives a small value: GE + 0 . 3 9 G M = 0.025 =b 0.020 =b 0.014 at Q 2 = 0.48GeV^ [4], S A M P L E finds a small strange contribution to the nucleon magnetic moment: — 0.1 =b 5.1% [5], and the A4 Collaboration finds small strange contribution to another combination of form factors: GE + 0 . 2 2 5 G ' M = 0.039 4 [6]. On the other hand, a few experiments indicate quite large matrix elements for some hiddenstrangeness operators. One prominent example is the 7rnucleon U term. Send offprint requests to: [emailprotected] ^ CERNPHTH/2004231 hepph/0411369
 si{x) + s^{x)  si{x)] =  0 . 1 0 =b 0.02.
On the other hand, H E R M E S measurements of singleparticle inclusive particle production have been interpreted as indicating t h a t [10] As = dx[s^{x)  si{x)\s^{x)
= 0.03=b0.03zb0.01.
However, this estimate has been questioned on the grounds t h a t corrections to independent fragmentation may be large [11]. T h e overall picture is t h a t hiddenstrangeness matrix elements in the nucleon may be small or large, depending on the J^^ q u a n t u m numbers carried by the ss pair, which is quite compatible with theoretical ideas [12]. Even if one accepts the first estimate of As, there has been an argument about its interpretation, based on the observation t h a t in one regularization scheme As gets a large contribution from gluons Ag: As = As — {as/7T)Ag: perhaps the 'bare' As vanishes, and Ag is large and positive [13]? Since the Ag correction is schemedependent.
J. Ellis: Today's view on strangeness
.
Fig. 1. Comparison of recent determinations of AG by HERMES [16], SMC [17] and COMPASS [18] one may wonder how well defined it is [14]. However, this suggestion has at least raised the profile of the interesting question how large Ag may be. A first measurement of the gluon polarization was reported by F N A L experiment E581/704 [15], measuring TT^ production at h i g h p ^ , and Fig. 1 shows recent measurements of AG. H E R M E S find [16] < AG/G
for an average longitudinal m o m e n t u m fraction < rj = XG{'^ + s/Q'^) > = 0.07 using the asymmetry in hadronpair production at highp^^. Most recently, COMPASS has announced a new determination, again via the asymmetry in hadronpair production at high PT [18]: AG/G
at an average < XQ > = 0.13, and P H E N I X is preparing a new determination via the double helicity asymmetry in pp —^ TT^ at high PT However, all these measurements have large uncertainties, b o t h systematic and statistical. For the time being, there is no strong indication t h a t AG is large, and even its sign must still be regarded as an open question.
2 Models of the nucleon In the last millennium, the naive quark model (NQM) [19] held pride of place. It envisages the nucleon as composed of three constituent quarks Q, each with mass MQ ^ 300
MeV, in a nonrelativistic wave function. Any additional qq pairs are thought to be generated perturbatively, and few of t h e m are expected to be ss pairs. Baryons sit in the usual nonexotic SU(3) multiplets, and the combination of a UUD or U D D wave function with meagre pair creation explains the OkuboZweigIizuka (OZI) rule [20]. T h e proton spin is simply the algebraic sum of valence constituentquark spins, which add up to 1/2. Chiral soliton models [21] provide an alternative viewpoint for the new millennium. They are based on the observation t h a t the intrinsic masses of the (current) quarks defined at short distances are much smaller: niu,d ^ few MeV, rris ^ 100 MeV. Hence the quarks should be treated relativistically, and there are many intrinsic qq pairs in the nucleon wave function, which are treated as clouds of meson fields. In this picture, lowlying exotic SU(3) multiplets are predicted [22], as are evasions of the OZI rule due to the copious ss pairs in the nucleon. Moreover, the nucleon spin is obtained from orbital angular momentum, the sum of the quark spins vanishes in the limit of vanishing quark masses and a large number of colours, and the ss pairs are polarized [23]. In the chiral soliton model, baryons are constructed as clouds of TT, K , and rjs mesons, and the presence of the latter is one way to understand the copious ss pairs in the baryon wave function. Exotic baryons are expected as excitations of the meson cloud with nontrivial SU(3) transformation properties, which can also be interpreted as excitations of the qq sea in the baryon. T h e baryon spin is due to the coherent rotation of this meson cloud, motivating the interpretation of the baryon spin as orbital angular momentum, and requiring the ss pairs to be polarized. Specifically, in the limit of massless quarks and a large number of colours, the meson cloud contains no SU(3) fiavoursinglet 770 mesons, and nor do the TT, if, and rjs mesons present have any coupling to the r]o. Since axialcurrent matrix elements are related in the chiral limit to pseudoscalarmeson couplings, the absence of the 770 implies t h a t the SU(3)singlet axialcurrent matrix element between baryons also vanishes. Classically, this matrix element is in t u r n related to the sum of the quark spins in the baryon, which therefore vanishes. Since the sum of the u and d quark contributions to the proton spin is positive and does not vanish, there must be a negative, nonzero strange contribution t h a t cancels t h e m [23]. T h e presence of a nontrivial qq sea in the nucleon suggests t h a t there may exist baryons with 'exotic' quant u m numbers t h a t cannot be explained in terms of naive threequark wave functions. It is surely too naive to imagine t h a t , if one places three quarks in a vacuum containing many qq pairs, there will never be any rearrangement of the qq q u a n t u m numbers. If the qq q u a n t u m numbers do not cancel each other out exactly, the resulting baryonic state will have 'exotic' q u a n t u m numbers. In the chiralsoliton language, these can be thought of as excitations of the meson cloud. This line of argument led theorists working on chiral solitons to predict the existence of a relatively light antidecuplet of exotic baryons [22], resembling 'pentaquark'
J. Ellis: Today's view on strangeness
r.
Fig. 2. The /e/t panel shows that the /I(1520) signal can be isolated with suitable cuts. The signal for the exotic baryon O^ is the solid histogram in the right panel, and the dashed histogram is a control sample [25]
P Fig. 4. Chiralsoliton calculations of the 7rnucleon coupling depend on the model parameters Gio and p, but may be consistent with experiment (shaded) [30].
4 . 8 CT
Fig. 3 . Compilation of measurements of the O mass and decay width [281
states in t h e N Q M , of which t h e lowestlying member would have ududs q u a n t u m numbers a n d weigh about 1530 MeV [24]. O n t h e other hand, t h e Particle D a t a Group quoted in 1987 " . . . the prejudice against baryons not made of three quarks . . . " , a n d ceased t o consider their existence. This changed with t h e report by t h e L E P S Collaboration at Spring8 [25] of a candidate exotic baryon with ududs q u a n t u m numbers a n d weighing about 1540 MeV, shown in Fig. 2. This was soon followed by an avalanche of corroborating evidence from other experiments [26], which stimulated considerable theoretical enthusiasm. However, these observations are somewhat problematic. T h e masses vary outside t h e quoted statistical a n d systematic errors, as seen in Fig. 3, with peaks in nK^ final states tending to be heavier t h a n those in pK^ final states [27]. Moreover, K N partialwave analyses require t h e decay width t o be < 1 MeV [29]: this is surprisingly narrow, a n d some experiments have reported widths close t o their experimental mass resolutions, as also seen in Fig. 3. Nevertheless, t h e striking evidence in favour of the early chiralsoliton predictions motivated revisiting
t h e m [30]. It was soon realized t h a t t h e accuracy of t h e mass prediction [24] was somewhat fortuitous, as it was based on a debatable assignment of another member of the baryon antidecuplet, a n d t h e mass splittings within this multiplet were calculated using an outdated value for t h e TTnucleon U term. Using plausible ranges for t h e chiral soliton moments of inertia t h a t control t h e mean excitation energy of antidecuplet baryons, a n d t h e more modern value of t h e U t e r m discussed above, there is an uncertainty in t h e 0~^ baryon mass of at least 100 MeV. As for t h e decay width, although it vanishes in t h e limit of a large number of colours [31], leadingorder calculations with plausible baryon couplings h a d difficulty in pushing t h e decay width below about 10 MeV. We made a detailed study of SU(3)symmetry breaking effects on baryonmeson couplings in chiral soliton models, finding values of t h e 7rNucleon a n d ir — A couplings t h a t are consistent with experiment, as seen in Fig. 4. These effects tend t o reduce further t h e 6^ decay rate, as seen in Fig. 5, though a width < 1 MeV still seems unlikely [30]. One of t h e key predictions of chiral soliton models is t h e existence of other, more 'exotic' baryon multiplets, such as a 27 a n d a 35 of SU(3) t h a t are slightly heavier t h a n t h e antidecuplet. In particular, there should be a 6~^~^ state weighing < 100 MeV more t h a n t h e ^"^, as seen in Fig. 6. It is difffcult t o understand how such a state could have escaped observation in many experiments, b u t CLAS d a t a may hint at t h e existence of such a state [32]. W h a t would be t h e implications of 'exotic' baryons for our understanding of strangeness in t h e nucleon? As has already been mentioned, t h e exotic baryon spectrum is sensitive t o t h e value of the TTN U term. Using t h e chiral soliton mass formula E = 3 ( 4 M ^  3MA  MTV) + 4(Mr2  MA) m octet 4{Ms3/2
J. Ellis: Today's view on strangeness 2.00 1.75
P Fig. 5. Corrections to the 0 coupling tend to reduce its decay width, whereas the decays of other antidecuplet baryons are not strongly suppressed [30]
^ (GeV/cf Fig. 7. The ratios of (/) and uj production in association with various other particles in NN annihilation, as measured at LEAR [34], often exceed predictions based on the naive OZI rule
(f) meson production should be due only t o t h e admixture of light quarks in t h e 0 wave function, which is small, since the (f) and UJ mesons are almost ideally mixed. Generically, one would expect a production ratio
This is not very different from t h e weighted averages of experimental d a t a from TTA^ collisions:
Fig. 6. Spectroscopy of the lowestlying exotic baryons predicted in chiralsoliton models [30]
the observation of t h e S baryon reported by NA49 [33], if confirmed, would correspond t o
0.6.
This is quite consistent with t h e direct measurements of the TTN U t e r m discussed earlier [7], perhaps lending some credence t o t h e whole chiral soliton scheme.
3 OZI violation or evasion? T h e OkuboZweigIizuka (OZI) rule is based on t h e idea t h a t processes with disconnected quark lines are suppressed. As a corollary, it is not possible t o produce ss mesons in t h e interactions of nonstrange particles. Hence,
are somewhat larger, b u t not dramatically big. On t h e other hand, there are large deviations from t h e naive OZI relation in d a t a from L E A R experiments on pp annihilations, particularly in t h e following reactions: pp ^ 7 0 , pp —^ 7T(f) from t h e ^6^1 state, and pd ^ 0 n , as seen in Fig. 7 [34]. Moreover, t h e (p/uj ratio depends strongly on the initialstate spins of t h e nucleons and antinucleons, on their orbital angular momenta, on t h e m o m e n t u m transfer and on t h e isospin. For example, t h e partialwave dependence of annihilations into (f)7r is shown in Fig. 8, where we see t h a t 5wave annihilations dominate. Another example of a large (^/uu ratio is in t h e Pontecorvo reaction pd —? (pn shown in Fig. 9, where it is compared with t h e annihilation process pd ^ Tr^n. T h e OZI rule could be evaded if there are ss pairs in the nucleon wave function, since new classes of connected
J. Ellis: Today's view on strangeness 1 1 1 1 1 1 1 1 1111 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
S—wave % Fig. 8. The annihilation pp ^ (j)7v^ proceeds predominantly via the s wave [34]
Fig. 9. Signal for the Pontecorvo reaction
quark diagrams could be drawn for the production of the scattering, where pp —> ppcp is about 14 times more co(/) and other ss mesons. Motivated by the d a t a on polar pious t h a n pp ^ ppu) near threshold and the (j) and uo ized deepinelastic scattering [35], we have formulated a angular distributions are different, the 'violation' of the polarized intrinsic strange ness model [36,37], in which naive OZI rule by a factor ~ 20 in pd ^^ HeCpioo), and the ss pairs in the nucleon are assumed to have negative the negative longitudinal polarization of A baryons measured in deepinelastic neutrino scattering [38], discussed polarization, and to be in a relative 0++ state, not a 1 state as in the naive (f) wave function. T h e production of below. However, there are also some serious problems for the strangeonium states may occur via rearrangement of the s and s in different nucleons, and not via shakeout from an polarizedstrangeness model. For example, the strong OZI individual nucleon. Thus, both nucleons participate in the 'violation' in pp —^ 7 0 takes place from a ^^o initial state, production mechanism, and their relative polarization and and the spin transfer Dnn in PP ^ ^^ is small, whereas orbital angular m o m e n t u m states are important. In par Knn > 0, indicating t h a t the spin of the proton is transticular, one would expect the (j) and the /2(1525) mesons ferred to the A, not to the A [39]. Moreover, CLAS d a t a to be produced more copiously from spintriplet initial on the reaction ep ^ e'K^A indicate t h a t the spins of states t h a n from spinsinglet initial states, the (p meson the s and s are antialigned [40]. Also serious is the probto be produced preferentially from L = 0 states, and the lem t h a t pp ^ 7r^(j) is not possible from a ^ ^ i initial state /2(1525) to be produced preferentially from L = 1 states. without either flipping the spin of the 5quark or positive This model has led to several correct predictions [34]. polarization of the strange quarks in the proton [41]. In nucleonantinucleon annihilations, the pp ^ TT^CJ) rates Many of these problems would be resolved if there from the ^ ^ i and ^Pi initial states are in a ratio ^^ 15 : 1, are two components of polarized strangeness, one with in agreement with the prediction of L = 0 dominance. On Sz = —I and one with Sz = 0 [42]. This would permit the other hand, the pp —> /27r^ rates are in a ratio ~ 1 : 10, pp ^ jcj) and pp ^ TT^cp via rearrangement diagrams, and in agreement with the prediction of L = 1 dominance. the CLAS d a t a t h a t require the spins of the s and s to Moreover, there is evidence t h a t the mechanisms for 0 be antialigned could be accommodated by shakeout of and uj production are different: the ^Pi fractions in ircj)^ the Sz = 0 component. However, even this model does and UJTT^ are < 7% and ~ 37%, respectively, and (p and UJ not fit all the data, as seen in the Table. One promising production have different energy dependences in np anni possibility is to assume the dominance of a spinsinglet us hilations. Also, it has been observed t h a t the initial states diquark configuration, as indicated in the last column of in pp ^ (j)(j) are dominated by J^^ = 2++, consistent the Table [42]. Understanding the strange polarization of with Swave annihilations in a spintriplet state. Addi the proton is still a work in progress. tionally, spinsinglet initial states are strongly suppressed in pp ^ AA: the singlet fraction Fg = (0.1 =b 7.3) x 10~^. T h e polarizedstrangeness model is also consistent with 4 Probing strangeness via A polarization the available d a t a on the Pontecorvo reaction pd ^ (pn and on selection rules in pp ^ K^K*. Other success Since the polarization of the A is measurable in its decays, ful predictions include 'OZI violation' in nucleonnucleon and since the A polarization is inherited, at least in the
J. Ellis: Today's view on strangeness Table 1. Score card for various models of polarized strangeness in the nucleon wave function 0++ : ^ . =  1
from ^S\ (j)r]/urj'. small from ^^'i
fUh large from pwave (jyn/ujn: large P{A) < 0 in DIS ep ^ AKe : P{A) pp ^ AA: Dnn PP ^ AA\ Knn
naive quark model, from its constituent 5 quark, A polarization is potentially a powerful way of probing polarized strangeness. Particularly interesting from this point of view is the measurement of A polarization in leptoproduction, where two options are available: measurements in the fragmentation region of the struck quark or in t h a t of the target. T h e struck quark has net polarization, but is usually a, u, so there is no interesting spin transfer to the A baryon. However, in the target fragmentation region the 'wounded nucleon' left behind by the polarized struck quark is itself polarized in general. A priori, it is a diquark system with the possibility of a polarized ss 'sea' attached to it. Memory of this polarization may be carried by the s and s in the wounded nucleon wave function and transferred to A and A baryons produced in the target fragmentation region [43]. We have modelled this idea using the Lund string fragmentation model incorporated in L E P T O 6.5.1 and J E T S E T 7.4, and have considered various combinations of two extreme cases in which the A baryon is produced by fragmentation of either the struck quark or the remnant diquark [44]. We then fix free parameters of the model by demanding consistency with d a t a from N O M A D in deepinelastic u scattering [38]. In addition to providing a good fit to N O M A D data, as seen in Fig. 10, this procedure can be used directly to make predictions for electroproduction d a t a from H E R M E S , and agrees very well. We have then gone on to make predictions for the COMPASS muon scattering experiment. COMPASS was originally conceived to measure the polarization of the gluons in the proton, but it may also be able to cast light on the polarization of the strange quarks!
0.5 0.25 0 . . ^ 0.25 0.5 60^ .
j:^.,.
.
.  . ^ y ^ j,..r.»
Fig. 10. Predictions for longitudinal A polarization in deepinelastic scattering [44], compared with data from NOMAD [381
5 Summary As we have seen in this review, there are many pieces of experimental evidence for a significant amount of hidden strangeness in the proton wave function, notably the TTnucleon i7 term, polarized deepinelastic scattering and
J. Ellis: Today's view on strangeness large deviations from the naive OZI rule. These observations may cast light on complementary models of nucleon structure, namely the 'naive' quark model and chiral soliton models. T h e latter were recently boosted by reports of exotic baryons, whose existence was predicted years ago in the soliton model. Their spectroscopy depends, in particular, on the magnitude of the 7rnucleon U term, and the tentative indication from the difference between the masses of the 0~^ and H ~ "  b a r y o n s is t h a t this should be large, in agreement with the latest direct determinations of this quantity. T h e situation with phenomenological models of OZI 'evasion' due to polarized ss pairs in the nucleon wave function is unclear: the d a t a from L E A R and other lowenergy experiments suggests t h a t there must be many ss pairs, but their polarization states remain obscure. One thing is, however, clear: we may expect many more twists in the strange story of the nucleon! Acknowledgements. It is a pleasure to thank my collaborators on the subjects discussed here, namely S. J. Brodsky, E. Gabathuler, M. Karliner, D. E. Kharzeev, A. Kotzinian, D. V. Naumov and M. G. Sapozhnikov.
References 1. M.A. Shifman, A.I. Vainshtein, V.I. Zakharov: Nucl. Phys. B 147, 448, 519 (1979); B.L. loffe: Nucl. Phys. B 188, 317 (1981), [Erratumibid. B 191, 591] (1981); L.J. Reinders, H. Rubinstein, S. Yazaki: Phys. Rept. 127, 1 (1985) 2. G. 't Hooft: Phys. Rev. D 14, 3432 (1976), [Erratumibid. D 18, 2199] (1978) 3. A.O. Bazarko et al., [CCFR Collaboration]: Z. Phys. C 65, 189 (1995), [arXiv:hepex/9406007] 4. K.A. Aniol et al., [HAPPEX Collaboration]: Phys. Lett. B 509, 211 (2001), [arXiv:nuclex/0006002] 5. D.T. Spayde et al., [SAMPLE Collaboration]: Phys. Lett. B 583, 79 (2004), [arXiv:nuclex/0312016] 6. F.E. Maas et al., [A4 Collaboration]: Phys. Rev. Lett. 93, 022002 (2004), [arXiv:nuclex/0401019] 7. These numbers are from M.M. Pavan, I.I. Strakovsky, R.L. Workman, R.A. Arndt: PiN Newslett. 16, 110 (2002), [arXiv:hepph/0111066]; For other recent estimates, see: T. Inoue, V.E. Lyubovitskij, T. Gutsche, A. Faessler: arXiv:hepph/0311275 and references therein 8. J. Gasser, H. Leutwyler, M.E. Sainio: Phys. Lett. B 253, 252 (1991); M. Knecht: hepph/9912443 9. G.K. Mallot: Proc. of the 19th Intl. Symp. on Photon and Lepton Interactions at High Energy LP99 ed. J.A. Jaros, M.E. Peskin: Int. J. Mod. Phys. A 15S1, 521 (2000), [eConf C990809, 521 (2000)], [arXiv:hepex/9912040] 10. K. Ackerstaff et al., [HERMES Collaboration]: Phys. Lett. B 464, 123 (1999), [arXiv:hepex/9906035] 11. A. Kotzinian: Phys. Lett. B 552, 172 (2003), [arXiv:hepph/0211162] 12. B.L. loffe, M. Karliner: Phys. Lett. B 247, 387 (1990) 13. G. AltarelU, G.G. Ross: Phys. Lett. B 212, 391 (1988); R.D. Carlitz, J.C. Collins, A.H. Mueller: Phys. Lett. B 214, 229 (1988); A.V. Efremov, O.V. Teryaev: JINRE288287 (1988) 14. R.L. Jaffe, A. Manohar: Nucl. Phys. B 337, 509 (1990)
15. D.L. Adams et al., [FNAL E581/704 Collaboration]: Phys. Lett. B 336, 269 (1994) 16. A. Airapetian et al., [HERMES Collaboration]: Phys. Rev. Lett. 84, 2584 (2000), [arXiv:hepex/9907020] 17. B. Adeva et al., [Spin Muon Collaboration (SMC)]: Phys. Rev. D 70, 012002 (2004), [arXiv:hepex/0402010] 18. M. Leberig, for the COMPASS Collaboration: CERN PH Seminar, October 25th 2004, available from http://pccosrvl.cern.ch/compass/publications/talks/ 19. J.J.J. Kokkedee: The Quark Model (Benjamin, New York, 1969) 20. S. Okubo: Phys. Lett. 5, 165 (1963); G. Zweig: CERN preprint TH412 (1964); J. lizuka: Prog. Theor. Phys. Supp. 37, 21 (1966) 21. T.H.R. Skyrme: Proc. Roy. Soc. Lond. A 260, 127 (1961); Nucl. Phys. 31, 556 (1962); see also E. Witten: Nucl. Phys. B 160, 57 (1979) and 223, 422, 433 (1983); G.S. Adkins, C.R. Nappi, E. Witten: Nucl. Phys. B 228, 552 (1983); G.S. Adkins, C.R. Nappi: Nucl. Phys. B 233, 109 (1984) 22. E. Guadagnini: Nucl. Phys. B 236, 35 (1984); P.O. Mazur, M.A. Nowak, M. Praszalowicz: Phys. Lett. B 147, 137 (1984); A.V. Manohar; Nucl. Phys. B 248, 19 (1984); M. Chemtob: Nucl. Phys. B 256, 600 (1985); S. Jain, S.R. Wadia: Nucl. Phys. B 258, 713 (1985); M.P. Mattis, M. Karliner: Phys. Rev. D 31, 2833 (1985); M. Karliner, M.P. Mattis: Phys. Rev. D 34, 1991 (1986) 23. S.J. Brodsky, J.R. Ellis, M. Karliner: Phys. Lett. B 206, 309 (1988) 24. D. Diakonov, V. Petrov, M.V. Polyakov: Z. Phys. A 359, 305 (1997), [arXiv:hepph/9703373] 25. T. Nakano et al., [LEPS Collaboration]: Phys. Rev. Lett. 91, 012002 (2003), [arXiv:hepex/0301020] 26. V.V. Barmin et al., [DIANA Collaboration]: Phys. Atom. Nucl. 66, 1715 (2003), [Yad. Fiz. 66, 1763 (2003)], hepex/0304040; S. Stepanyan et al., [CLAS Collaboration]: hepex/0307018; J. Barth et al., [SAPHIR Collaboration]: hepex/0307083; V. Kubarovsky, S. Stepanyan, CLAS Collaboration: hepex/0307088; A.E. Asratyan, A.G. Dolgolenko, M.A. Kubantsev: hepex/0309042 V. Kubarovsky et al., [CLAS Collaboration]: Phys. Rev. Lett. 92, 032001 (2004), [E: ibid. 92, 049902 (2004), [arXiv:hepex/0311046]; A. Airapetian et al., [HERMES Collaboration]: Phys. Lett. B 585, 213 (2004), [arXiv:hepex/0312044]; S. Chekanov et al., [ZEUS Collaboration]: Phys. Lett. B 591, 7 (2004), [arXiv:hepex/0403051]; R. Togoo et al.: Proc. Mongolian Acad. Sci., 4, 2 (2003); A. Aleev et al., [SVD Collaboration]: arXiv:hepex/0401024 27. Q. Zhao, F.E. Close: arXiv:hepph/0404075 28. This compilation was kindly provided by M. Karliner 29. R.N. Cahn, G.H. Trilling: Phys. Rev. D 69, 011501 (2004), [arXiv:hepph/0311245] 30. J.R. Ellis, M. Karliner, M. Praszalowicz: JHEP 0405, 002 (2004), [arXiv:hepph/0401127] 31. M. Praszalowicz: Phys. Lett. B 583, 96 (2004), [arXiv:hepph/0311230] 32. M. Battaglieri, [presenting CLAS Coll. data]: talk at Pentaquark Workshop, Feb. 1012, 2004, Trento, Italy, h t t p : //www. t p 2 . ruhrunibochum. d e / t a l k s / t r e i i t o 0 4 / battaglieri.pdf 33. C. Alt et al., [NA49 Collaboration]: Phys. Rev. Lett. 92, 042003 (2004), [arXiv:hepex/0310014]
J. Ellis: Today's view on strangeness
34. For a review, see: V.P. Nomokonov, M.G. Sapozhnikov: Phys. Part. Nucl. 34, 94 (2003), [Fiz. Elem. Chast. Atom. Yadra34, 189 (2003)], [arXiv:hepph/0204259] 35. J.R. Ellis, E. Gabathuler, M. Karliner: Phys. Lett. B 217, 173 (1989) 36. J.R. Ellis, M. Karliner, D.E. Kharzeev, M.G. Sapozhnikov: Phys. Lett. B 353, 319 (1995), [arXiv:hepph/9412334] 37. J.R. Ellis, M. Karliner, D.E. Kharzeev, M.G. Sapozhnikov: Nucl. Phys. A 673, 256 (2000), [arXiv:hepph/9909235] 38. P. Astier et a l , [NOMAD Collaboration]: Nucl. Phys. B 605, 3 (2001), [arXiv:hepex/0103047]
39. B. Bassalleck et al., [PS185 collaboration]: Phys. Rev. Lett. 89, 212302 (2002), [arXiv:nuclex/0206005] 40. D.S. Carman et al., [CLAS Collaboration]: Phys. Rev. Lett. 90, 131804 (2003), [arXiv:hepex/0212014] 41. M.P. Rekalo, J. Arvieux, E. TomasiGustafsson: Z. Phys. A 357, 133 (1997) 42. J.R. Ellis, A. Kotzinian, M.G. Sapozhnikov: unpublished 43. J.R. Ellis, D. Kharzeev, A. Kotzinian: Z. Phys. C 69, 467 (1996), [arXiv:hepph/9506280] 44. J.R. Ellis, A. Kotzinian, D.V. Naumov: Eur. Phys. J. C 25, 603 (2002), [arXiv:hepph/0204206]
Strange and gluonic contributions to the nucleon spin JeanMarc Le Goff CEASaclay, DAPNIA SPhN, France Received: 15 October 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. Inclusive deep inelastic scattering (DIS) experiments lead to a small contribution of the quark spins to the nucleon spin and a negative contribution from strange quarks. Historically this triggered the interest in measuring strange form factors. However, the result of these inclusive experiments has to be reinterpreted taking into account the axial anomaly in QCD, which depends on the gluon contribution to the nucleon spin. The COMPASS experiment at CERN and experiments at RHIC are going to measure this gluonic contribution. PACS. 13.60.Hb Total and inclusive cross sections (including deepinelastic processes) production
1 Introduction In 1989 the E M C experiment [1] at C E R N concluded t h a t the contribution AH of the quark spins to the nucleon spin was compatible with zero and t h a t the strange contribution As was significantly negative, triggering the socalled spin crisis. This meant t h a t the strange axial matrix element of the nucleon was nonzero, which raised the issue whether the strange vector matrix elements could also be nonzero and motivated a large experimental program to study strange form factors through Parity Violation experiments. Here we present an overview of the spin of the nucleon, for a full review see [2].
2 Strange quark and total quark contributions T h e spin of the nucleon can be decomposed in the contributions from its constituents as 1
where Ag is the contribution from the spins of the gluons and Lq and Lg are the contributions from orbital angular momenta of quarks and gluons. W h a t do we know about AU from a theoretical point of view ? On the one hand we have the quark model which provides us with a large part of our understanding of hadrons. It gives Z\i7 ^ 0.75. On the other hand we are not able to use Q C D to compute AZ! from first principles, but using results from hyperon [3 decay experiments and assuming a strange quark contribution As = 0, we get AU ^ 0.6. We then have a qualitative agreement between Quark Model and Q C D . Send offprint requests to: [emailprotected]
In inclusive deep inelastic experiments (DIS) a lepton is scattered off a nucleon and only the scattered lepton is observed. Only two Lorentz invariants enter the problem. They can be chosen as the mass of the virtual photon, Q^ = — ^^, which gives the resolution of the probe, and Xbj = Q'^/2M{E — E') which is the fraction of the nucleon m o m e n t u m carried by the quark which absorbed the virtual photon. DIS corresponds to the limit of large Q^ at fixed Xhj. T h e cross section involves structure functions which depend only on the two Lorentz invariants. However, because the lepton scatters on a quark, which is a pointlike particle, the Q^ dependence vanishes, at least to leading order in Q C D , a property known as scaling. In the unpolarized case we have two structure functions Fi and F2. They can be expressed in terms of the p a r t o n distribution function (pdf) as Fi{x) = \ \_^u{x) \ ^u{x) \ \d{x) + ^ J ( x ) + ^5(x) + ^s(x)] and F2{x) = 2xFi{x)^ where u{x) for instance, is the probability to find inside the nucleon a quark of flavor u and a fraction x of the nucleon m o m e n t u m . In the polarized case we have in addition gi and g2] gi{x) = \ \\Au{x) + \Ad{x) + \As{x)\ with Au[x) = u^(x) — u~{x) + u^[x) — u~{x), where the polarized pdf u~^{x) is the probability to find inside the nucleon a quark of flavor u with a fraction x of the nucleon m o m e n t u m and a spin parallel to the nucleon spin. T h e integral Au = J^ Au{x)dx is the total contribution of spins of quark of flavor u to the nucleon spin. We then have AU = Au \ Ad \ As, where the contribution of heavier flavors (c, b and t) is negligible. In 1989 the E M C measured Tf = /^ g^{x)dx = ^ [^Au \ ^Ad \ ^As'j. Using SUf{3) symmetry hyperon P decays give as = Au — Ad and as = Au \ Ad — 2As. This provided 3 equations for 3 unknowns resulting in AU = 0.12 zb 0.17 and As =  0 . 1 0 zb 0.03 < 0 ! This
J.M. Le Goff: Strange and gluonic contributions to the nucleon spin
Fig. 2. AE and As as a function of the hypothesis for Ag
Phys. Rev. Lett. 92 (04) 1200^
polarized proton and d from the unpolarized one^ we get a Au d contribution. T h e opposite case gives Ad u. T h e spin asymmetry for the process then reads
Fig. 1. Results of Hermes semiinclusive analysis [4] and projections from COMPASS
came as a big surprise which was advertised as the spin crisis, so t h a t the E M C paper is one of the 3 most cited experimental papers [3]. T h e results were confirmed by SMC at CERN, SLAC, and Hermes at DESY. T h e uncertainty is now dominated by the extrapolation in the low x, unmeasured region. In order to go further and measure the x dependence of the polarized pdf one needs to perform semiinclusive DIS, I \ p ^ I' \ h \ X. T h e spin asymmetry for this process can be written as j:,el[Aqix)D^iz)^Aq{x)Dl{z)] A^
where z = E^/E^* and the fragmentation function D^{z) gives the probability t h a t the fragmentation of a quark of flavor q gives a hadron h. T h e fact t h a t D^{z) ^ D^{z) allows for the separation of sea from valence. Using A i , A^~^, A^ for proton and neutron the SMC obtained Auy, Ady, Ad^ Au without using SUf(3) flavor symmetry. However, such an analysis relies on 2, which assumes t h a t the measured hadron comes from the fragmentation of the struck quark and not from the target remnant. In order to measure also As{x) one needs to identify strange particles within the measured hadrons, either using a R I C H detector, or by reconstructing the K^ mass. Unfortunately this introduces more sensitivity to target remnants because TTIK > ^TT Figure 1 presents the results obtained by the Hermes collaboration. They exhibit no indications of Z\s < 0. D a t a down to lower x are expected from COMPASS. Polarized pdf can also be measured through parity violating 'p^ p ^ W at RHIC in Brookhaven [5]. T h e point is t h a t qq ^ VF^ selects a given quark helicity. W~^ production is dominated by ud ^ W~^. If u comes from the
T h e same formula is obtained for W~ production with u and d exchanged. T h e nice thing with such an analysis is t h a t it is independent of target fragmentation effects. It has, however, hardly any sensitivity to Z\5. We must go back to the inclusive case and note t h a t E M C does not actually measure Aq but axial matrix elements aq = (A^^7^75^1 A/"). Naively the axial matrix elements are identifled with Aq. However, due to axial anomaly we rather have ao = AU — 3 ^ Z \ ^ and as = As — ^Ag. In addition Ag oc I n Q ^ , so t h a t the gluonic contribution to aq does not go to zero at high Q^ in spite of the as factor. T h e actual value of AU and As now depends on the value of Ag as illustrated by Fig. 2. If Ag ^ 0 we are back to the spin crisis with a small AU and a negative As. If Ag is large and positive we may end u p with the expected As ^ 0 and AU ^ 0.6. Ag must be measured, b o t h for itself and in order to extract the actual value of AU and As from inclusive experiments.
3 Gluon contribution W h e n Q^ increases the resolution of the probe improves and what used to appear as a quark may start to appear as a quark and a gluon, or a gluon may appear as a quark antiquark pair. In these conditions the variation dq{x, Q'^)/d{liiQ'^) tells us something about g{x, Q^). This is formalized in the Doksh*tzerGribovLipatovAltareliParisi (DGLAP) equations. Indeed, in the unpolarized case, performing a nexttoleading order Q C D flt of F2 data, from flxed target experiments and from the H E R A collider, provides a good measurement of g{x,Q'^). This is unfortunately not the case in the polarized case because there is no polarized leptonproton collider. T h e Q^ range
J.M. Le Goff: Strange and gluonic contributions to the nucleon spin for gi data, between SLAC or Hermes and SMC, is not large enough (at most a factor 10) to allow for a precise estimate of Ag{x^ Q^) A direct measurement is needed. It is difficult to probe gluons in lepton scattering because they have no electric charge. Photons can, however, interact with gluons through the photon gluon fusion ( P G F ) process, 7*^ ^ qq. This is a higher order process which has a small cross section relative to the leading order process, j*q ^ g, so tagging is needed. T h e first possibility consists in requiring the produced qq pair to be a cc pair. Since the intrinsic charm inside the nucleon is negligible, the observation of charm is a signature of the P G F process. T h e fragmentation of charm quarks produces a D^ = cu meson in 60% of the cases. T h e easiest way to see the D^ is through its decay to Kn which has a 4% branching ratio. Due to this low branching ratio, one requires to detect either the c or the c through this channel. T h e drawback is t h a t in this case it is not possible to reconstruct the kinematics at the vertex and the m o m e n t u m fraction of the gluon cannot be evaluated. One gets the mean value of Ag{x)/g{x) averaged over the experimental acceptance. A second possibility arises from the fact t h a t in the leading order process, ^*q ^ q^ all the produced hadrons are in the direction of the virtual photon, whereas in the P G F process the qq pair can be produced at any angle and the resulting hadrons may have a transverse m o m e n t u m Pt with respect to the photon. So the idea is to search for pairs of hadrons with high pt (or two high pt jets at high energy). There is, however, a physical background, the socalled Q C D Compton process, 7*^ —7^ qg^ since in this process the final q and g b o t h can produce a high pt hadron. This background has to be evaluated by a Monte Carlo (MC) simulation, starting from the known polarized quark distribution functions. This second method was already used by Hermes [6] and SMC [7]. Hermes used a 28 GeV electron beam, they did not measure the scattered electron and they were dominated by low Q^, quasi real photons. T h e hard scale, for perturbative Q C D to be valid, is then provided by pt. T h e generator P Y T H I A was used for background estimation. T h e difficulty in this case is t h a t , in addition to Q C D Compton, there is an important background of events where the photon is resolved into its partonic structure. In this analysis the dilution due to these events was taken into account but not their contribution to the asymmetry. SMC had a 190 G e V / c muon beam. They selected only events with Q^ > 1 (GeV/c)^, in which case the resolved photon contribution could be neglected and the pure DIS generator, L E P T O , could be used. T h e results obtained by the two collaborations are presented in Fig. 4. Error bars are still pretty large. T h e COMPASS collaboration [8] at C E R N is using b o t h methods. T h e muon beam intensity was increased by a factor 5 relative to SMC to provide an average luminosity of 5 10^^ cm~^s~^. T h e target is made of ^LiD with a dilution factor 'P^ 0.5 to be compared to 0.24 for deuterated butanol, used by SMC. T h e spectrometer was commissioned and first d a t a were taken in 2002. D a t a were
taken again in 2003 and 2004 but there will be no b e a m in 2005, due to LHC installation. T h e 2002 d a t a gives an asymmetry for high pt pair =  0 . 0 6 5 zb 0.036 zb 0.010, including production, A^"^^^^' all Q^. MC studies are going on, to take into account the contribution of resolved photons to the asymmetry. P a r t of these events involve a gluon in the nucleon so t h a t their contribution to the asymmetry is proportional to Ag/g. Neglecting this fact, a statistical error 5{Ag/g) = 0.17 should be obtained. In the same conditions, all d a t a between 2002 and 2004 should provide S{Ag/g) ^ 0.05, or alternatively four bins in Xg with 8{Ag/g) ^ 0.10 in each bin. Using only d a t a with Q^ > 1 (GeV/c)^, should provide 5{Ag/g) ^ 0.16. T h e D^ ^ KTT open charm channel suffers from an important combinatorial background. Requiring t h a t the D^ comes from the disintegration D* ^ D^n ^ KTTTT, strongly reduces the background. T h e first reason is the small value of M^^ — Mjjo — M^ = 6 MeV, which leaves little space for the background. T h e second reason is the excellent experimental resolution in this mass difference, better t h a n 2 MeV, to be compared to about 25 MeV for the resolution in the D^ mass alone. Many improvements are ongoing in terms of reconstruction, in particular in the R I C H detector used to identify K. An error S{Ag/g) ^ 0.24 is then expected out of 2002 to 2004 data. T h e gluon distribution can also be probed in polarized protonproton collisions [5]. T h e golden channel is the socalled direct 7 production ^ ^ ^ 7 + jet + X. At the part on level this corresponds to qg —^ ^ 7 , where quarks from one of the protons are used to probe gluons in the other proton. So the measured asymmetry is a convolution, Ag 0 Aq. There is a physics background, qq ^ ^ 7 , which also produces 7 + jet + X. Its contribution to the asymmetry, which goes like Aq (g) Aq^ can be computed from the measured Aq{x) and Aq{x). Other possible channels at R H I C include jet production (or high Pt leading hadron) and heavy flavor production. RHIC, the relativistic heavy ion collider, is used part of the time as a polarized protonproton collider at ^/s = 5 0  5 0 0 GeV. Both P H E N I X and STAR collaborations are taking d a t a in this mode. T h e design polarization is 70%, using Siberian snakes in R H I C and partial snakes in the AGS to eliminate depolarizing resonances. T h e design luminosity is £ = 210^^ cm~^s~^. An integrated luminosity of 7 p b  ^ with P ^ 50% is expected from the 20042005 run. T h e first measured leading hadron asymmetry, ^ J ^ , is presented in Fig. 3. This asymmetry is sensitive to the convolution Ag 0 Ag. It should then be positive unless there is a node in Ag{x). T h e expected error on Ag/g in the golden channel for an integrated luminosity of 320 pb~^ with P = 70% are presented in Fig. 4. We see t h a t a wide range of Xg is covered with an excellent accuracy. Note, however, t h a t this is only the statistical accuracy, some systematic uncertainty will arise from background subtraction and deconvolution. In the long t e r m there is the EIC project [10] to build a 10 GeV polarized electron linac to collide with one of the
J.M. Le Goff: Strange and gluonic contributions to the nucleon spin
 all in all this should give an error on the integral Ag JQ Ag(x)dx on the order of 0.03 to 0.05.
GRSVstd .
PT (GeV/c) Fig. 3 . Leading hadron AJ^ asymmetry measured at RHIC by PHENIX [9], compared with prediction corresponding to the standard GRV parametrization and to a parametrization with 100% polarization at a given starting scale
 o COMPASS proposal open charm COMPASS proposal high p^  0 RHIC/STAR proposal(Y + jet, 320 pb'^ P=70%) ! HERMES published high p^ '_
we know the flavor singlet axial matrix element ao = AU— {3as/27r) Ag = 0.27 T h e strange axial matrix element is a^ ^ —0.10, but this is more sensitive to possible violations of SUf{3) symmetry t h a n GQ. T h e measurement of As{x) in semiinclusive DIS is delicate because it is sensitive to target fragmentation effects. There is hardly any sensitivity to As{x) in 'p^p collisions. A possible solution would be to get the integral As from neutrino experiments combined with parity violation experiments [11]. T h e first measurements of Aglx) by COMPASS and R H I C will appear soon. T h e experimental methods and the systematic errors are completely different and in addition each experiment has several channels, so this should provide a reliable measurement of Ag(x). In the longer t e r m the project of a polarized electronproton collider at RHIC would provide a much more accurate value of ao and Ag and then of AU. T h e contributions from orbital m o m e n t u m are very difficult to access. Generalized P a r t o n Distributions ( G P D , for a review see e.g. [12]) describe at the same time the transverse position and the longitudinal momentum, which is what is needed to compute the orbital moment u m . This is formalized in the Ji sum rule which relates an integral of G P D s to Jq = AU\Lq. G P D s can be measured in DVCS experiments, IN ^ VN'f. First DVCS measurements were performed at J L A B and H E R A and plans exist at COMPASS; due to its high luminosity an electronproton collider at RHIC would be the ideal tool for this. There is, however, a very long way before measuring the Ji sum rule.
Fig. 4. Measured values of Ag/g by Hermes and SMC, together with projections from COMPASS and RHIC RHIC polarized proton beams with a luminosity of 10 to 10^"^ c m  ^ s  i and ^ s = 100 GeV. T h e beam could start between 2012 and 2014. This would allow for:  the measurement of gi down to x = 10~^ instead of 3 10~^ now, dramatically reducing the dominant error on JQ gi{x)dx (and then on ao) which is due to the low x extrapolation.  a large range of Q^ for gi d a t a and then a precise estim a t e of Ag from Q C D NLO fit.  a direct measurement of Ag(x, Q^) through hadron pairs, jet pairs and charm production.
1. J. Ashman et al.: Phys. Lett. B 206, 364 (1988), NucL Phys. B 328, 1 (1989) 2. M. Anselmino, A. Efremov, E. Leader: Phys. Rep. 261, 1 (1995); B. Lampe and E. Reya: Phys. Rep. 332, 1 (2000) 3. h t t p : / / w w w . s l a c . S t a n f o r d . e d u / l i b r a r y / t o p c i t e s / 4. A. Airapetian et al.: Phys. Rev. Lett. 92, 012005 (2004) 5. G. Bunce et al.: Ann. Rev. Nucl. Part. Sci. 50, 525 (2000); hepph/0007218 6. A. Airapetian et al.: Phys. Rev. Lett. 84, 25842588 (2000) 7. B. Adeva et al.: Phys. Rev. D 70, 012002 (2004) 8. COMPASS, G. baum et al.: CERNSPSLC9614, http://www.compass.cern.ch/compass/proposal 9. hepex/0404027 10. www. phenix. b n l . gov/WWW/publish/abhay/Hoine_of_EIC/ 11. S. Pate: Don't Forget to Measure Z\s, these proceedings 12. M. Diehl: Phys. Rep. 388, 41 (2003) or hepph/0307382
Electromagnetic form factors of the nucleon Kees de Jager Thomas Jefferson National Accelerator Facility, Newport News, Virginia 23606, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The experimental and theoretical status of elastic electron scattering from the nucleon is reviewed. As a consequence of new experimental facilities and new theoretical insights, this subject is advancing with unprecedented precision. PACS. 13.40.Gp Electromagnetic form factors  29.27.Hj Polarized beams
1 Introduction More t h a n 40 years ago Akhiezer et al. [1] (followed 20 years later by Arnold et al [2]) showed t h a t the accuracy of nucleon charge formfactor measurements could be increased significantly by scattering polarized electrons off a polarized target (or equivalently by measuring the polarization of the recoiling proton). However, it took several decades before technology had sufficiently advanced to make the first of such measurements feasible and only in the past few years has a large number of new d a t a with a significantly improved accuracy become available. For G^ measurements the highest figure of merit at Q^values larger t h a n a few GeV^ is obtained with a focal plane polarimeter. Here, the Jacobian focusing of the recoiling proton kinematics allows one to couple a standard magnetic spectrometer for the proton detection to a largeacceptance nonmagnetic detector for the detection of the scattered electron. For studies of G^ one needs to use a magnetic spectrometer to detect the scattered electron in order to cleanly identify the reaction channel. As a consequence, the figure of merit of a polarized ^He target is comparable to t h a t of a neutron polarimeter.
2 Proton electric form factor In elastic electronproton scattering a longitudinally polarized electron will transfer its polarization to the recoil proton. In the onephoton exchange approximation the ratio of the charge and magnetic form factors is directly proportional to the ratio of the polarization components, parallel (Pi) and transverse (Pt) to the proton's momentum. T h e greatest impact of the polarizationtransfer technique was made by the two recent experiments [3,4] in Hall A at Jefferson Lab, which measured the ratio G^/G^ in a Q^range from 0.5 to 5.6 GeV^. T h e most striking feat u r e of the d a t a is the sharp, practically linear decline
as Q^ increases. This significant falloff of the formfact or ratio is in clear disagreement with the results from the Rosenbluth extraction. Segel and Arrington [5] performed a highprecision Rosenbluth extraction in Hall A at Jefferson Lab, designed specifically to significantly reduce the systematic errors compared to earlier Rosenbluth measurements. T h e main improvement came from detecting the recoiling protons instead of the scattered electrons. One of the spectrometers was used as a luminosity monitor during an e scan. Preliminary results [5] of this experiment, covering Q^values from 2.6 to 4.1 GeV^, are in excellent agreement with previous Rosenbluth results. This basically rules out the possibility t h a t the disagreement between Rosenbluth and polarizationtransfer measurements of the ratio G^/G^ is due to an underestimate of edependent uncertainties in the Rosenbluth measurements.
2.1 Twophoton exchange Two(or more)photon exchange ( T P E ) contributions to elastic electron scattering have been investigated b o t h experimentally and theoretically for the past fifty years. Almost all analyses with the Rosenbluth technique have used radiative corrections t h a t only include the infrared divergent parts of the box diagram. Thus, terms in which b o t h photons are hard (and which depend on the hadronic structure) have been ignored. T h e most stringent tests of T P E on the nucleon have been carried out by measuring the ratio of electron and positron elastic scattering off a proton. Corrections due to T P E will have a different sign in these two reactions. Unfortunately, this (e+e~) d a t a set is quite limited [6], only extending (with poor statistics) up to a Q^value of ~ 5 GeV^, whereas at Q^values larger t h a n ~ 2 GeV^ basically all d a t a have been measured at evalues larger t h a n  0.85. Blunden et al [7] carried out the first calculation of the elastic contribution from T P E effects, albeit with a
K. de Jager: Electromagnetic form factors of the nucleon
simple monopole Q^dependence of the hadronic form factors. They obtained a practically Q^independent correction factor with a linear edependence that vanishes at forward angles. However, the size of the correction only resolves about half of the discrepancy. A later calculation which used a more realistic form factor behavior, resolved up to 80% of the discrepancy. A different approach was used by Chen et al. [8], who related the elastic electronnucleon scattering to the scattering off a parton in a nucleon through generalized parton distributions. TPE effects in the leptonquark scattering process are calculated in the hardscattering amplitudes. The results for the TPE contribution fully reconcile the Rosenbluth and the polarizationtransfer data and retain agreement with positronscattering data. Hence, it is becoming more and more likely that TPE processes have to be taken into account in the analysis of Rosenbluth data and that they will affect polarizationtransfer data only at the few percent level. Experimental confirmation of TPE effects will be difficult, but certainly should be continued. The most direct test would be a measurement of the positronproton and electronproton scattering crosssection ratio at small evalues and Q^values above 2 GeV^. A measurement in the CLAS detector at Jefferson Lab has been recently approved [9]. Additional efforts should be extended to studies of TPE effects in other longitudinaltransverse separations, such as proton knockout and deepinelastic scattering (DIS) experiments.
of the beamtarget asymmetry with the target polarization perpendicular and parallel to the momentum transfer is directly proportional to the ratio of the electric and magnetic form factors. A similar result is obtained with an unpolarized deuteron target when one measures the polarization of the knockedout neutron as a function of the angle over which the neutron spin is precessed with a dipole magnet. At low Q^values corrections for nuclear medium and rescattering effects can be sizeable: 65% for ^H at 0.15 GeV^ and 50% for ^He at 0.35 GeV^ These corrections are expected to decrease significantly with increasing Q. The latest data from Hall C at Jefferson Lab, using either a polarimeter [13] or a polarized target [14], extend up to Q^ ^ 1.5 GeV^ with an overall accuracy of ^10%, in mutual agreement. From ~ 1 GeV^ onwards G^ appears to exhibit a Q^behavior similar to that of G^. Schiavilla and Sick [15] have extracted G'% from available data on the deuteron quadrupole form factor Fc2(Q^) with a much smaller sensitivity to the nucleonnucleon potential than from inclusive (quasi)elastic scattering.
The recent production of very accurate EMFF data, especially the surprising G^ data from polarization transfer, has prompted the theoretical community to intensify their investigation of nucleon structure. The first EMFF models were based on the principle of vector meson dominance (VMD), in which one assumes that the virtual 3 Neutron magnetic form factor photon couples to the nucleon as a vector meson. With this model lachello et al. [16] predicted a linear drop of A significant breakthrough was made by measuring the the proton form factor ratio, similar to that measured ratio of quasielastic neutron and proton knockout from a by polarization transfer, more than 20 years before the deuterium target. This method has little sensitivity to nu data became available. Gari and Kriimpelmann [17] exclear binding effects and to fluctuations in the luminosity tended the VMD model to conform with pQCD scaling and detector acceptance. A study of G^ at Q^values up at large Q^values. An improved description requires the to 5 GeV^ has recently been completed in Hall B by mea inclusion of the isovector TTTT channel through dispersion suring the neutron/proton quasielastic crosssection ratio relations [18,19]. By adding more parameters, such as the using the CLAS detector [10]. Preliminary results [10] in width of the pmeson and the masses of heavier vector dicate that C^j^ is within 10% of GD over the full Q^range mesons [20], the VMD models succeeded in describing new of the experiment (0.54.8 GeV^). EMFF data as they became available, but with little preInclusive quasielastic scattering of polarized electrons dictive power. Figure 1 confirms that Lomon's calculations off a polarized ^He target offers an alternative method to provide an excellent description of all EMFF data. Bijker determine G'%j^ through a measurement of the beam asym and lachello [21] have extended the original calculations metry [11]. By orienting the target polarization parallel by also including a mesoncloud contribution in F2. The to q, one measure RTI, which in quasielastic kinematics intrinsic structure of the nucleon is estimated to have an is dominantly sensitive to {G'^Y. For the extraction of rms radius of ~ 0.34 fm. These new calculations are in G'^j corrections for the nuclear medium [12] are necessary good agreement with the proton formfact or data, but do to take into account effects of finalstate interactions and rather poorly for the neutron. mesonexchange currents. Many recent theoretical studies of the EMFFs have applied various forms of a relativistic constituent quark model (RCQM). Because the momentum transfer can be 4 Neutron electric form factor several times the nucleon mass, the constituent quarks require a relativistic quantum mechanical treatment. AlIn the past decade a series of doublepolarization mea though most of these calculations correctly describe the surements of neutron knockout from a polarized ^H or EMFF behaviour at large Q^values, effective degrees of ^He target have provided accurate data on G^. The ratio freedom, such as a pion cloud and/or a finite size of the
K. de Jager: Electromagnetic form factors of the nucleon 1.5
^^^^^^^$^~f..
..
.
Fig. 1. Comparison of various calculations with available EMFF data. For G\ only polarizationtransfer data are shown. For G% the results of Schiavilla and Sick [15] have been added. The calculations shown are from [19,20,21,28,29,33]. Where applicable, the calculations have been normalized to the calculated values of /Xp,n 1.5 Gp
Fig. 2. Comparison of various RCQM calculations with available EMFF data, similarl to the comparison in Fig. 1. The calculations shown are from [22,24,26,25,27]. Miller (qonly) denotes a calculation by Miller [22] in which the pion cloud has been suppressed. Where applicable, the calculations have been normalized to the calculated values of /Xp,n
K. de Jager: Electromagnetic form factors of the nucleon
constituent quarks, are introduced to correctly describe the behaviour at lower Q^values. Miller [22] uses an extension of the cloudy bag model [23], three relativistically moving (in lightfront kinematics) constituent quarks, surrounded by a pion cloud. Cardarelli and Simula [24] also use lightfront kinematics, but they calculate the nucleon wave function by solving the threequark Hamiltonian in the IsgurCapstick onegluonexchange potential. In order to get good agreement with the E M F F d a t a they introduce a finite size of the constituent quarks in agreement with recent DIS data. T h e results of Wagenbrunn et al. [25] are calculated in a covariant manner in the pointform spectator approximation ( P F S A ) . In addition to a linear confinement, the quarkquark interaction is based on Goldstoneboson exchange dynamics. T h e P F S A current is effectively a threebody operator (in the case of the nucleon as a threequark system) because of its relativistic nature. It is still incomplete but it leads to surprisingly good results for the electric radii and magnetic moments of the other light and strange baryon ground states beyond the nucleon. Giannini et al. [26] have explicitly introduced a threequark interaction in the form of a gluongluon interaction in a hypercentral model, which successfully describes various static baryon properties. Relativistic effects are included by boosting the three quark states to the Breit frame and by introducing a relativistic quark current. All previously described R C Q M calculations used a nonrelativist ic t r e a t m e n t of the quark dynamics, supplemented by a relativistic calculation of the electromagnetic current matrix elements. Merten et al. [27] have solved the BetheSalpeter equation with instantaneous forces, inherently respecting relativistic covariance. In addition to a linear confinement potential, they used an effective flavordependent twobody interaction. T h e results of these five calculations are compared to the E M F F d a t a in Fig. 2. T h e calculations of Miller do well for all E M F F s , except for G^ at low Q^values. Those of Cardarelli and Simula, Giannini et al. and Wagenbrunn et al. are in reasonable agreement with the data, except for t h a t of Wagenbrunn et al. for G ^ , while the results of Merten et al. provide the poorest description of the data. Before the Jefferson Lab polarization transfer d a t a on G^/G^ became available Holzwarth [28] predicted a linear drop in a chiral soliton model. In such a model the quarks are bound in a nucleon by their interaction with chiral fields. Holzwarth's model introduced one vectormeson propagator for b o t h isospin channnels in the Lagrangian and a relativistic boost to the Breit frame. His later calculations used separate isovector and isoscalar vectormeson form factors. He obtained excellent agreement for the proton data, but only a reasonable description of the neutron data. Christov et al. [29] used an SU(3) NambuJonaLasinio Lagrangian, an effective theory t h a t incorporates spontaneous chiral symmetry breaking. This procedure is comparable to the inclusion of vector mesons into the Skyrme model, but it involves many fewer free parameters (which are fitted to the masses and decay constants of pions and kaons). A constituent quark mass of
Fig. 3. The ratio {Q'^F2/Fi)/ In^ ( Q V ^ ^ ) as a function of Q^ for the polarizationtransfer data and the calculations of [21, 22,28,26]. The same ratio, scaled by a factor 1/15, is shown for the neutron with open symbols. For A a value of 300 MeV has been used 420 MeV provided a reasonable description of the E M F F d a t a (Fig. 1). In the asymptotically free limit, Q C D can be solved perturbatively, providing predictions for the E M F F behavior at large Q^values. Recently, Brodsky et al. [30] and Belitsky et al. [31] have independently revisited the p Q C D domain. Belitsky et al. derive the following large Q^behavior:
where yl is a soft scale related to the size of the nucleon. Even though the Jefferson Lab d a t a follow this behavior (Fig. 3), Belitsky et al. warn t h a t this could very well be precocious, since p Q C D is not expected to be valid at such low Q^values. However, all theories described until now are at least to some extent effective (or parametrizations). They use models constructed to focus on certain selected aspects of Q C D . Only lattice gauge theory can provide a truly ab initio calculation, but accurate lattice Q C D results for the E M F F s are still several years away. One of the most advanced lattice calculations of E M F F s has been performed by the Q C D S F collaboration [32]. T h e technical state of the art limits these calculations to the quenched approximation (in which seaquark contributions are neglected), to a box size of 1.6 fm and to a pion mass of 650 MeV. Ashley et al. [33] have extrapolated the results of these calculations to the chiral limit, using chiral coefficients appropriate to full Q C D . T h e agreement with the d a t a (Fig. 1) is poorer t h a n t h a t of any of the other calculations, a clear indication of the technology developments required before lattice Q C D calculations can provide a stringent test of experimental E M F F data.
K. de Jager: Electromagnetic form factors of the nucleon
6 Experimental review and outlook In recent years highly accurate d a t a on the nucleon E M F F s have become available from various facilities around the world, made possible by the development of high luminosity and novel polarization techniques. These have established some general trends in the Q^behavior of the four E M F F s . T h e two magnetic form factors G^ and G^ are close to identical, following GD to within 10% at least u p to 5 GeV^, with a shallow minimum at  0.25 GeV^ and crossing GD at  0.7 G e V ^ Highly accurate measurements with the Rosenbluth technique have established t h a t the discrepancy between results on G^/G^ with the Rosenbluth techniques and with polarization transfer is not an instrumentation problem. Recent advances on twophoton exchange contributions make it highly likely t h a t the application of T P E corrections will resolve t h a t discrepancy. G^/G^ drops linearly with Q^ and G^ appears to drop from ^ 1 GeV^ onwards at the same rate as G^. Measurements t h a t extend to higher Q^values and offer improved accuracy at lower Q^values, will become available in the near future. In Hall C at Jefferson Lab Perdrisat et al. [34] will extend the measurements of G^^/G\^ to 9 GeV^ with a new polarimeter and largeacceptance leadglass calorimeter. Wojtsekhowski et al. [35] will measure G% in Hall A at Q^values of 2.4 and 3.4 GeV^ using the ^He(e, e'n) reaction with a 100 msr electron spectrometer. T h e Bates Large Acceptance Spectrometer Toroid facility (BLAST, h t t p : / / b l a s t . l n s . m i t . e d u / ) at M I T with a polarized hydrogen and deuteron target internal to a storage ring will provide highly accurate d a t a on G ^ and G% in a Q^range from 0.1 to 0.8 GeV^. Thus, within a couple of years G^ d a t a with an accuracy of 10% or better will be available u p to a Q^value of 3.4 GeV^. Once the upgrade to 12 GeV [36] has been implemented at Jefferson Lab, it will be possible to extend the d a t a set on G^^ and Gl^ to 14 GeV^ and on G% to 8 G e V ^ Acknowledgements. This work was supported by DOE contract DEAC0584ER40150 Modification No. M175, under which the Southeastern Universities Research Association (SURA) operates the Thomas Jefferson National Accelerator Facility.
References 1. A.I. Akhiezer, L.N. Rozentsweig, I.M. Shmushkevich: Sov. Phys. J E T P 6, 588 (1958) 2. R. Arnold, C. Carlson, F. Gross: Phys. Rev. C 23, 363 (1981) 3. V. Punjabi et al.: submitted to Phys. Rev. C; M.K. Jones et al.: Phys. Rev. Lett. 84, 1398 (2000) 4. O. Gayou et al.: Phys. Rev. Lett. 88, 092301 (2002)
5. R. Segel, J. Arrington: spokespersons, Jefferson Lab experiment EOOOOl (2000); private communication 6. J. Arrington: Phys. Rev. C 69, 032201R (2004) and references therein 7. P.G. Blunden, W. Melnitchouk, J.A. Tjon: Phys. Rev. Lett. 91, 142304 (2003) 8. Y.C. Chen et al.: Phys. Rev. Lett. 93, 122301 (2004) 9. W. Brooks et al.: spokespersons, Jefferson Lab experiment E04116 (2004) 10. W. Brooks, M.F. Vineyard: spokespersons, Jefferson Lab experiment E94017 (1994); private communication 11. T.W. Donnelly, A.S. Raskin: Ann. Phys. 169, 247 (1986) 12. J. Golak et al.: Phys. Rev. C 63, 034006 (2001) 13. R. Madey et al.: Phys. Rev. Lett. 91, 122002 (2003) 14. G. Warren et al.: Phys. Rev. Lett. 92, 042301 (2004); Zhu H. et al.: Phys. Rev. Lett. 87, 081801 (2001) 15. R. Schiavilla, I. Sick: Phys. Rev. C 64, 041002 (2001) 16. F. lachello, A. Jackson, A. Lande: Phys. Lett. B 43, 191 (1973) 17. M.F. Gari, W. Kriimpelmann: Z. Phys. A 322, 689 (1985); Phys. Lett. B 274, 159 (1992) 18. G. Hohler et al.: Nucl. Phys. B 114, 505 (1976) 19. H.W. Hammer, U.G. Meissner, D. Drechsel: Phys. Lett. B 385, 343 (1996); H.W. Hammer, U.G. Meissner: Eur. Phys. Jour. A 20, 469 (2004); P. Mergell, U.G. Meissner, D. Drechsel: Nucl. Phys. A 596, 367 (1996) 20. E.L. Lomon: Phys. Rev. C 64, 035204 (2001); Phys. Rev. C 66, 045501 (2002) 21. R. Bijker, F. lachello: Phys. Rev. C 69, 068201 (2004) 22. G.A. Miller: Phys. Rev. C 66, 032001R (2002) 23. S. Theberge, A.W. Thomas, G.A. Miller: Phys. Rev. D 24, 216 (1981) 24. F. CardarelK, S. Simula: Phys. Rev. C 62, 065201 (2000) 25. R.F. Wagenbrunn et al.: Phys. Lett. B 511, 33 (2001); S. Boffi et al.: Eur. Phys. Jour. A 14, 17 (2002) 26. M. Giannini, E. Santopinto, A. Vassallo: Prog. Part. Nucl. Phys. 50, 263 (2003); M. De Sanctis et al.: Phys. Rev. C 62, 025208 (2000); M. Ferraris et al.: Phys. Lett. B 364, 231 (1995) 27. D. Merten et al.: Eur. Phys. Jour. A 14, 477 (2002) 28. H. Holzwarth: Z. Phys. A 356, 339 (1996); hepph/0201138 29. C.V. Christov et al.: Nucl. Phys. A 592, 513 (1995); H.C. Kim et al.: Phys. Rev. D 53, 4013 (1996) 30. S.J. Brodsky et al.: Phys. Rev. D 69, 076001 (2004) 31. A.V. Belitsky X. Ji, F. Yuan: Phys. Rev. Lett. 91, 092003 (2003) 32. M. Gockeler et al.: heplat/0303019 33. J.D. Ashley et al.: Eur. Phys. Jour. A 19 (Suppl. 1), 9 (2004) 34. C.F. Perdrisat et al.: spokespersons, Jefferson Lab experiment EOl109 (2001) 35. B. Wojtsekhowski et al.: spokespersons, Jefferson Lab experiment E02013 (2002) 36. L.S. Cardman et al., eds.: PreConceptual Design Report for the Science and Experimental Equipment for the 12 GeV Upgrade of CEBAF, 2003, http://www.jlab.org /div_dept/physics_division/pCDR_public/pCDR_final
Two photon effects in electron scattering P.A.M. Guichon SPhN/DAPNIA, CEA Saclay, F91191 Gif sur Yvette, France Received: 15 November 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The apparent discrepancy between the Rosenbluth and the polarization transfer method for the ratio of the electric to magnetic proton form factors can be explained by a twophoton exchange correction which does not destroy the linearity of the Rosenbluth plot. Though intrinsically small, of the order of a few percent of the cross section, this correction is kinematically enhanced in the Rosenbluth method while it is small for the polarization transfer method, at least in the range of Q^ where it has been used until
PACS. 25.30.Bf Elastic electron scattering  13.40.Gp Electromagnetic form factors gluons, and QCD in nuclei and nuclear processes
1 Introduction T h e electromagnetic form factors are essential pieces of our knowledge of t h e nucleon structure a n d this justifies t h e efforts devoted t o their experimental determination. They are defined by t h e m a t r i x elements of t h e electromagnetic current J^{x) according to^:
where t h e electric form factor is defined by GE = Fi— TF2 with r = Q^ /4:M^ a n d CB{Q^^ S) is a known phase space factor which is irrelevant in what follows. T h e polarization parameter of t h e virtual photon has t h e expression^ y ^  M V ( l + r) y2 + M 4 r ( l + r ) '
so, at fixed Q^, giving e is equivalent t o give u. For a given Q^, (5) shows t h a t it is sufficient t o meaeu{p') \GM{Q^)r  F2{Q^)^] u{p) (1) sure t h e cross section for two values of £ t o determine t h e form factors GM a n d GE In t h e following t h e deterwhere e c^ Y^47r/137 is t h e proton charge a n d M t h e nu mination of GM a n d GE using (5) will be referred t o as cleon mass. T h e magnetic form factor GM is related t o t h e t h e Rosenbluth method [1]. T h e fact t h a t t h e combinaDirac ( F i ) a n d Pauli (F2) form factors by GM = Fi\ F2. tion da/CB{Q'^^^) is a linear function of e (Rosenbluth Here we consider only t h e proton case, so we have F i (0) = plot criterion) is generally considered as a test of t h e va1, ^2(0) = lip — 1 = 1.79. In t h e one photon exchange or lidity of t h e Born approximation. We shall see below t h a t Born approximation, elastic lepton scattering: this criterion is not strong enough. Polarized lepton beams give another way t o access t h e l{k) + N{p)^l{k')^N{p') (2) form factors [2]. In t h e Born approximation, t h e polarization of t h e recoiling proton along its motion {Pi) is progives direct access t o t h e form factors in t h e spacelike portional t o GM while t h e component perpendicular t o region (Q^ > 0) where they are real. Here we adopt t h e t h e motion {Pt ) is proportional t o GE We call this t h e usual definitions: polarization method for short. Because it is much easier to measure ratios of polarizations, it has been used mainly k + k' P + y .K kk' =p' p, (3) P through a measurement of to determine t h e ratio GE/GM Pt/Pi for which one finds t h e expression [3]: and choose K.P (4) Q' <7^ ^ GE I 2e (7) as t h e independent invariants of t h e scattering. In t h e Pi T(1 + £ ) G M ' Born approximation t h e elastic cross section is written: So, in t h e framework of t h e Born approximation, one
doB = CB{Q\e) \Gl,{Q^) +  G  ( Q 2 )
(5)
^ The spinors satisfy u{p)u{p) = 2M and the free states are normalized as < N{P')\NIP)
> = {27rf2p^S{p  p') .
has two i n d e p e n d e n t m e a s u r e m e n t s oi R = GE/GM
In
Fig. 1 we show t h e corresponding results, which we call ^ This expression assumes a negligible electron mass.
P.A.M. Guichon: Two photon effects in electron scattering
24
the polarization method result is little affected by the 2photon exchange, at least in the range of Q^ which has been studied until now.
2 Amplitude decomposition
Fig. 1. Experimental values of RRosenbluth and their polynomial fit
'^^^^
^^Polarization
Fig. 2. The box diagram. The filled blob represents the response of the nucleon to the scattering of the virtual photon
^Rosenbluth
^ n d Repolarization
^ ^ I t h e r a n g e of Q^
T h e simplest way to get the general form of the (e,p) scattering amplitude is to consider its helicity matrix elements T{h'^h'^] hehp) in the center of mass p + k = 0. Due to rotational invariance T depends on ^^cm, cos ^cmwhich do not change under the parity and time reversal operations. Since Q E D is invariant with respect to these operations, one must have T{h'X',
hehp) = T{h'^
T(h'X'.
hehp) = T{hehp; h%).
 /i^; he  hp),
(8) (9)
Note t h a t these equalities generally involve a phase factor which depends on the phase convention for the helicity states. For our purposes this factor is irrelevant. One can also check t h a t charge conjugation does not bring in additional constrains. If one applies the relations (8, 9) to the 2^ = 16 helicity amplitudes one finds t h a t only 6 of t h e m are independent. Moreover 3 of t h e m change the electron helicity which implies t h a t they are suppressed by an electron mass factor. So one ends with
which
is common to b o t h methods. T h e d a t a are taken from [4, 7,8]. T h e deviation between the two methods starts at Q^ = 2 — SGeV'^ and increases with Q^, reaching a factor 4 at about Q^ = 6 GeV^. This discrepancy is a serious problem as it generates confusion and doubt about the whole methodology of lepton scattering experiments. To unravel this problem we have to give u p the beloved one photon exchange concept and enter the not well paved p a t h of multiphoton physics. By this we do not mean the effect of soft (real or virtual) photons, t h a t is the radiative corrections. The effect of the latter is well under control because their dominant (infrared) part can be factorized in the observables and therefore does not affect the raHere we must consider genuine exchange of tio GE/GM' hard photons between the lepton and the hadron. Even if we restrict to the two photon exchange case, the evaluation of the box diagram 2 involves the full response of the nucleon to the scattering of a virtual photon and we do not know how to perform this calculation in a model independent way. Therefore we adopt a modest strategy based on the phenomenological consequences of using the full eN scattering matrix rather t h a n its Born approximation. Though it cannot lead to a full answer it produces the following interesting results [5]: — the 2photon exchange amplitude needed to explain the discrepancy is actually of the expected order of magnitude, t h a t is a few percent of the Born amplitude. — there may be a simple explanation to the fact t h a t the Rosenbluth plot looks linear even though it is strongly affected by the 2photon exchange.
1 1 1 1 2 2' 2 2
.T
1 1 1 2 2' 2
1
1
2' 2
as the only independent amplitudes. T h e next step is to write a covariant decomposition with 3 independent Lorentz structures. For obvious reasons we choose two of t h e m to be the same as in the one photon approximation. For the third structure several choices are possible and we found convenient to choose u{k')'y.Pu{k)
u{p')^.Ku{p).
so t h a t the T matrix can be written as T =
—u{k'h^u{k)
X tZ(p') [GMI^
 F2—
+ F^^J^)
u{p),
(10)
where GM , ^ 2 , ^3 are complex functions of v and Q^ and the factor e^/Q^ has been introduced for convenience. By analogy we define: GE = GM{l
+ r)F2
(11)
which is equal to GE in the Born approximation. T h e last step is to evaluate the matrix elements of T in the helicity basis and in the CM frame and to check t h a t the set of equations
2 2' 2 2
,T
1 1. 1
1
2 2' 2
2' 2
1
p.A.M. Guichon: Two photon effects in electron scattering ^
where:
F2, Fs
GM,
can be inverted, which is indeed the case. Note t h a t one can also start directly from the general amplitude for elastic scattering of two spin 1/2 particles as derived by Goldberger et al. [6]. Neglecting the amplitudes which flip the helicity of the electron this amplitude is written [6]: u{p') {S + Vf.K)
T = u{k')j.Pu{k)
+ A u{k')j^f.Pu{k)
u{p)
u{p')j^'y.Ku{p)
(12)
Using Dirac equation and elementary relations among the Dirac matrices, it is then a simple exercise to write (12) in the form (10). If one compares with the Born amplitude: 1
TB = e^u{k')f^u{k)—u{p')
25
[GMI^
F2—
~M
] u{p),
'^ME = (PM — (PE, (PSM
h(t>M.
P= j ^ 
(18)
If one substitutes the Born approximation values of the amplitudes (14) then (16, 17) give back the familiar expressions (5, 7). To simplify the general expressions (16, 17) we make the very reasonable assumption t h a t only the two photons exchange needs to be considered. This amounts to keep only the terms of order e^ with respect to the leading one in (16, 17). Using the fact t h a t (pM, (pE and F3 are of order e^ we get the approximate expressions: da
(19)
GM
Csie^Q^)
(13)
GE X <
one gets the relations:
GE
1 + '
+ 2e
1+ 
Y,27 GM
\GM
and yiBorn
(14)
{iy,Q^) = 0.
Since F3 and the phases of GM and F2 vanish in the Born approximation, they must originate from processes involving at least the exchange of 2photon. If we take care of the factor e^ introduced in the deflnition (13), we see t h a t they are at least of order e^. This, of course, assumes t h a t the phases of GM and F2 are deflned, which amounts to suppose t h a t , in the kinematical region of interest, the modulus of GM and F2 do not vanish. We take it for granted in the following but this restriction must be kept in mind if one goes in regions where some amplitude becomes very small.
(20)
Pi
2s
\GE\
1+ s
GM\
Y,27
To set a scale for the size of the two photon correction we have introduced the dimensionless ratio: lyF.
Y2^{iy,Q^)=n
M^\GM
which should be a good measure of the effect since, if we
3 Cross section and polarization transfer
neglect G,
G M with respect to r in (19) , we see t h a t 2
If we deflne:
the cross section would be of the form G M GM
=
g^^M
GE
^^i4>B
\G M (15)
GE\
Fi = e^^^ F^
our problem. In the cross section the ratio G,
1
comes with a t e r m 2 ( r
then, using s t a n d a r d techniques, we get the following expressions for the observables of interest: GM
Gsie^Q^)
+
GE
1
^2epn{
(GM^GE]F^
1
2s
+  + . T 2e r ( l + s)
Gi
ep Fs
(16)
l\e cos (j)ME + P F. cos (psM (17)
GM
+ i¥^p
cos (psM
G
GM
G M J Y2ry which at large
Q^ is essentially 2 r l«'22'7 = Q Y2^/2M^ . This produces an amplification of the two photon effect which is not present in Pt/Pi
da
(1 + 12,)^.
Therefore we expect 1^27 to be of the order of a ~ 1/137. T h e equations (19,20) already exhibit the solution to
As a rough estimate let us take Gi
GM\
4.M^ . T h e n GE{^)/GM{^) = 1/2.79 and choose Q^ the coefficient of e in (19), which is supposed to measure GE/GM in the Born approximation, is equal to 0.128 + 2.7127 So even if 1^27 is as small as 1% it produces a relative correction of 21%! By contrast if we do the same for the expression in parenthesis in (20), with e = 0.8 which is a typical value used in [7,4], we get 0.36+0.68 ^27For I27 = 1% this only produces a 2% correction. Now t h a t the origin of the discrepancy has been identified we can t r y to analyze the d a t a in a more quantitative manner.
P.A.M. Guichon: Two photon effects in electron scattering
26
4 Analysis
'
1
1
'
1
'
1
Q^^lGeV^
0.045
From (19,20) we see t h a t t h e experimental couple {da, Pt/Pi)
depends on G M
G,
and n{Fs).
Q^=2Ge\^
0.04
In first
approximation we know t h a t
—
0.035 0.03
""
Q^^3Ge\/ 2 . ,7 Q 4 Gev

Q^^SGeX"^
'%^ 0.025
(5M(^Q')
^ GM(Q'),
\GE(iyQ^)\
^
GE(Q^)
0.02 0.015
so only 7l{Fs) is really a new unknown parameter. If we look at t h e d a t a of [8] for (J/CB{S,Q'^) as a function of e we observe t h a t for each value of Q^ t h e set of points are pretty well aligned. We see on (19) t h a t this can be understood if, at least in first approximation, t h e product 1/F3 is independent of e. We do not have a first principle explanation for this b u t we feel allowed t o take it as an experimental evidence. To explain t h e linearity of t h e plots one should also suppose t h a t G M and G E are mdependent of e (that \s u) b u t since t h e dominant t e r m of these amplitudes depends only on Q^ this is a very mild assumption. We then see from (19) t h a t what is measured using t h e Rosenbluth method is: Gi (p exp \ Rosenbluth)
Gi
+ 2
T+
Yi^.
(21)
GM
GM
0.01 0.005 n
1
1
1
1
1
1
1
~
Fig. 3 . The ratio Y^^^ versus e for several values of Q^
cross section we cannot treat it as a perturbation when solving t h e system of equations. Since t h e latter is equivalent t o a quartic equation it is more efficient t o solve it numerically. For this we have fitted t h e d a t a by a polynomial in Q^. T h e result is shown on Fig. 1 a n d we shall consider this fit as t h e experimental values. In particular we do not a t t e m p t t o represent t h e effect of the error bars which can be postponed t o a more complete reanalysis of the data. Using t h e fit we solve numerically t h e system (21, 23) for t h e couple {Y^^^
with
1
jyexp
^17+27
GE
/ G M ). T h e solu
GE I GM a n d I27 essentially independent of e
tion for t h e ratio Y^^^ is shown on Fig. 3 where we can see t h a t it is actually small, of the order of a few percents. rather t h a n Also we observe t h a t it is essentially flat as a function of e which is consistent with our hypothesis. In fact it is a GE / pexp \2 (22) direct consequence of t h e smallness of 1^2^^ which multi\^Rosenbluth J GM plies t h e only factor which depends on e in (23). For t h e as implied by t h e one photon exchange approximation. same reason i^i^^27 ^^ ^^^ essentially independent of e. On t h e other hand t h e experimental results of the poT h e above result for 1^2^^^ indicates t h a t t h e correclarization method have been obtained for a narrow range tions t o t h e Born approximation are actually small in abof £, typically ^ from e = .6 t o .9. So, in practice, we can solute value. In t h e Rosenbluth method their effect is acneglect t h e e dependence of Repolarization ^ ^ ^ ^^^^ (20) cidentally amplified b u t there is no reason t o think t h a t we see t h a t this experimental ratio must be interpreted this kind of accident will also occur in GE — GE or GM — as: GM' SO it makes sense t o compare t h e value we get for ^17+27 wi^^ ^^^ starting experimental ratios R^Rosenbiuth 2e GE GE nexp (23) and Repolarization' ^ ^ 8 is showu ou Fig. 4 where we see Yi211 + 1Polarization GM GM t h a t ^ a s e x p e c t e d , i ^ i ^ ^ 2 7 ^^ close t o Repolarizationrather t h a n T^exp Polarization
GE
(24)
5 Conclusion
GM
G M and ,2 I27 t h a t we can solve for each value of Q^ . Due t o t h e
Within t h e hypothesis of our analysis, we come t o t h e conclusion t h a t t h e d a t a for GE/GM from t h e Rosenbluth a n d t h e polarization method are compatible once the exchange of 2photon is allowed in t h e analysis. T h e two photon effect is intrinsically small, as it should, b u t is strongly amplified in t h e Rosenbluth method. Assum
kinematical enhancement of t h e two photons effect in t h e
ing t h a t t h e difference between GE/GM
In order t h a t (23) be consistent with our hypothesis we should find t h a t 1^27 is small enough t h a t the factor 2 ^ / ( 1 + e) introduces no noticeable e dependence in RepolarizationWe have now a system of (21, 23) for GE
^ except at the lowest values of Q^ , where there is anyway no discrepancy between the Rosenbluth and the polarization method.
a n d GE/GM
is
of the same order as 1^2^^ ^^^ ^^^ consider t h a t t h e value ^ The calculation here has been done (arbitrarily) dX £ — 0.6 but the result is essentially independent of s.
p.A.M. Guichon: Two photon effects in electron scattering
27
1.5 ^p
Rcysenbhllh
a
{positron)
a{electron)
2BCA.
In the 2photon approximation one gets BCA = 2
sGEn(SGE) + rGMnjSGM) + SGMJGE sGl + rGl,
which in the limit GE/{TGM)
a()
1  AsYo27
+ rGM)Y2^
^ 0 implies ,n{6GM)
GM
Fig. 4. Comparison of the experimental ratios l^pR^RosenUuth According to our analysis we have 1^27 > 0, while and ^ipR^p^iarization ^ i t h the valuc of M P ^ I ^ + 2 7 = the scarce existing d a t a [11] at large Q^ indicate t h a t MP ^E I GM deduced from our analysis cr(+)/cr(—) is compatible with one or even a bit larger. negative and of the This would imply t h a t 5GM/GM^S order of I275 which is of course in line with our working hypothesis. This illustrates how dedicated beam charge is essentially the value of GE/GM we obtain for GE/GM asymmetry experiments could be used to strengthen our obtained in the Born approximation but corrected by the understanding of two photon effects in electron scattering. two photon effect due to F3. Our conclusion is then t h a t the correction which must be applied to the results of the polarization method is negligible while it is huge in the References case of the Rosenbluth method. To confirm our results one needs some model calcu1. M.N. Rosenbluth: Phys. Rev. 79, 615 (1950) lation of the central quantity of our analysis, namely F3. 2. A.I. Akhiezer, L.N. Rozentseig, I.M. Shmushkevich: Sov. An explicit calculation of the two photon effect has been Phys. J E T P 6, 588 (1958) performed in [9] and it indicates t h a t it can explain part 3. A.I. Akhiezer, M.R Rekalo: Sov. Phys. Doklady 13, 572 of the discrepancy. However this calculation limits the in(1968) termediate states to the nucleon itself, which is certainly 4. O. Gayou et al.: Phys. Rev. Lett. 88, 092301 (2002) not a realistic hypothesis. A more ambitious calculation, 5. P.A.M. Guichon, Marc Vanderhaeghen: Phys. Rev. Lett. where the intermediate states excitations are implicitly 91, 142303 (2003) taken into account through the use of generalized parton 6. M.L. Goldberger, Y. Nambu, R. Oehme: Ann. of Phys. 2, distributions, is now close to completion [10]. T h e prelim226 (1957) inary results are in agreement with our analysis. 7. M.K. Jones et al.: Phys. Rev. Lett. 84, 1398 (2000) 8. L. Andivahis et al: Phys. Rev. D 50, 5491 (1994) Another important point is the study of observables 9. P.G. Blunden, W. Melnitchouk, J.A. Tjon: Phys. Rev. which can directly test our understanding of the two phoLett. 9 1 , 142304 (2003) ton effects. As an example we consider the beam charge 10. M. Vanderhaeghen: private communication asymmetry BCA^ which is defined by: 11. J. Mar et al.: Phys. Rev. Lett. 2 1 , 482 (1968)
Eur Phys J A (2005) 24, s2, 2932 DOI: 10.1140/epjad/s2005040053
EPJ A direct electronic only
Single spin asymmetries in elastic electronnucleon scattering B. Pasquini^'^ and M. Vanderhaeghen^'^ ^ ^ ^ ^
Dipartimento di Fisica Nucleare e Teorica, Universita degli Studi di Pavia and INFN, Sezione di Pavia, Pavia, Italy ECT*, Villazzano (Trento), Italy Thomas Jefferson National Accelerator Facility, Newport News, VA 23606, USA Department of Physics, College of WiUiam and Mary, Wilhamsburg, VA 23187, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. We discuss the target and beam normal spin asymmetries in elastic electronnucleon scattering which depend on the imaginary part of twophoton exchange processes between electron and nucleon. In particular, we estimate these transverse spin asymmetries for beam energies below 2 GeV, where the twophoton exchange process is dominated by the resonance contribution to the doubly virtual Compton scattering tensor of the nucleon. PACS. 25.30.Bf Elastic electron scattering  25.30.Rw Electroproduction reactions
1 Introduction
ments of this beam normal SSA have been presented in this conference. Model calculations for such observables have been recently performed in different kinematical regimes [1,2,3, 4]. Here we report on a study of the imaginary part of the 27 exchange entering in the normal SSA's at low and intermediate b e a m energies [5]. Using unitarity, one can relate the imaginary part of the 27 amplitude to the electroabsorption amplitudes on a nucleon. Below or around twopion production threshold, one is in a regime where these electroproduction amplitudes are relatively well known using pion electroproduction experiments as input. Therefore the aim is to gain a good knowledge of the imaginary part of the twophoton exchange amplitude, and then to use such information as input for dispersion relations which will allow to quantify the contribution of the real part of the 27 exchange processes. In addition, observables such as normal SSA's are sensitive to the electroproduction amplitudes on the nucleon for a wide range of photon virtualities. This may provide information on resonance transition form factors complementary to the information obtained from pion electroproduction experiments.
Elastic electronnucleon scattering in the onephoton exchange approximation is a valuable tool to access information on the structure of hadrons. New experimental techniques exploiting polarization observables have made possible precision measurements of hadron structure quantities, such as its electroweak form factors, parity violating effects, N ^ A transition form factors, and spin dependent structure functions. However, to push the precision frontier further in electron scattering, one needs a good control of the twophoton (27) exchange mechanism and needs to understand how they may affect different observables. T h e imaginary (absorptive) part of the 27 exchange amplitude can be accessed through a single spin asymmetry (SSA) in elastic electronnucleon scattering, when either the target or beam spin are polarized normal to the scattering plane. As time reversal invariance forces this SSA to vanish for onephoton exchange, it is of order a = e^/(47r) c^ 1/137. Furthermore, to polarize an ultrarelativistic particle in the direction normal to its mom e n t u m involves a suppression factor m/E (with m the mass and E the energy of the particle), which typically is of order 10~^ — 10~^ when the electron beam energy is in the 1 GeV range. Therefore, the resulting target nor2 Single spin asymmetries in elastic mal SSA can be expected to be of order 10~^, whereas the b e a m normal SSA is of order 10~^ — 10~^. In the electronnucleon scattering case of a polarized lepton beam, asymmetries of the order p p m are currently accessible in parity violation (PV) elas T h e target spin asymmetries in elastic electronnucleon tic electronnucleon scattering experiments. While the P V scattering is defined by asymmetry measurements involve a beam spin polarized along its momentum, the SSA for an electron beam spin normal to the scattering plane can be accessed using the ^n (1) same experimental a p p a r a t u s . Results from first measure
B. Pasquini, M. Vanderhaeghen: Single spin asymmetries in elastic electronnucleon scattering
30
tual Compton scattering tensor with two spacelike photons. T h e latter is given by W^^(P',X'J,;P,\M)
=
^(2^)^(54(p + X
gapx)
X < y A'^l j t ^ ( 0 )  X > < X  J ^ ( 0 )  p A^ > ,
(5)
where t h e sum goes over all possible onshell intermediate hadronic states X. T h e number of intermediate states X Fig. 1. The 27 exchange diagram. The blob represents the which one considers in t h e calculation sets a limit on how response of the nucleon to the scattering of the virtual photon high in energy one can reliably calculate t h e hadronic tensor in (5). In addition t o t h e elastic contribution {X = N) which is exactly calculable in terms of onshell nucleon where cr^ {a\) denotes t h e cross section for a n unpolarized electromagnetic form factors, we approximate t h e remainb e a m and for a nucleon spin parallel (antiparallel) t o t h e ing inelastic part of W^^ with a sum over all TTN internormal polarization vector S^ = (k x k ' ) /  k x k '  (with k mediate states (i.e. X = TTN in t h e blob of Fig. 1). T h e and k^ t h e threemomenta of t h e initial and final electron, calculation is performed by using t h e unitarity relation respectively). Analogous expression as in (1) holds for t h e which allows t o express W^^ in terms of electroabsorption b e a m spin asymmetry {Bn) when we interpret a^ (cr^) amplitudes j*N ^ X at different photon virtualities. To as t h e cross section for a n unpolarized target a n d for a n estimate t h e pion electroproduction amplitudes we use t h e electron beam spin parallel (antiparallel) t o t h e normal phenomenological MAID analysis (version 2000) [7], which polarization vector. As h a s been shown by de Rujula et contains b o t h resonant and nonresonant pion production al. [6], t h e target a n d beam normal spin asymmetry are mechanisms. This same strategy has been used before in related t o t h e absorptive part of t h e elastic eN scattering the description of real and virtual Compton scattering in amplitude. Since t h e onephoton exchange amplitude is the resonance region, and checked against d a t a in [8]. purely real, t h e leading contribution t o SSA's is of order O(e^), and is due t o an interference between one and twophoton exchange amplitudes, i.e.
3 Results and discussion
2Im(E spins
SSA =
I7
E spins
AbsTa^) (2)
'ij\
where Ti^ is t h e onephoton exchange amplitude, a n d Abs T2j is t h e absorptive part of t h e doubly virtual Compton scattering tensor on t h e nucleon, as shown in Fig. 1. Equation (2) can be expressed in terms of a 3dimensional phasespace integral as SSA = 
2n2 e'Q
I
(v^me)^
(277)3 Z ) ( s , Q 2 ) y ^ , J^.
/
11
4^/5
^^ki ^2 na '^ {.^oniv J?"^"},
fxQl
(3)
where W^ = p^ is t h e squared invariant mass of t h e intermediate state X , and s = (p + /c)^. In (3), t h e momenta are defined as in Fig. 1, Qf = —(^ a n d Q\ = —(^ correspond with t h e virtualities of t h e two spacelike photons, D ( s , Q 2 ) = Q V e ' E . p ^ n . I^17P. a n d L « ^ . a n d B^^^ are the leptonic a n d hadronic tensors, respectively. Furthermore, (3) reduces t o t h e target or beam asymmetry once we specify t h e helicities for t h e polarized particles a n d take t h e sum over t h e helicities of t h e unpolarized particles. T h e explicit expression for t h e tensor H^^^ is given by: ^ a ^ . = W^
[«(p',A'jv)r"(p',p)t*(p,A;v)]*
(4)
where F'^^p^p) is t h e elastic photonnucleon vertex a n d W^^ corresponds t o t h e absorptive part of t h e doubly vir
In this section we show our results for b o t h b e a m and target normal spin asymmetries for elastic electronproton scattering. O u r calculation covers t h e whole resonance region, a n d addresses measurements performed or in progress at MITBates [9], MAMI [10], J L a b [11,12], a n d SLAC [13]. In Fig. 2, we show t h e beam normal spin asymmetry B^ for elastic e~^p ^ e~p scattering at a low beam energy of £^e = 0.2 GeV. At this energy, t h e elastic contribution is sizeable. T h e inelastic contribution is dominated by t h e region of threshold pion production, as is shown in Fig. 3, where we display t h e integrand of t h e 14^integration for B^. W h e n integrating t h e full curve in Fig. 3 over W, one obtains t h e total inelastic contribution to Bn (i.e. dasheddotted curve in Fig. 2). T h e present calculation (MAID) of t h e threshold pion electroproduction is consistent with chiral symmetry predictions, a n d is therefore largely model independent. One notices t h a t at backward c.m. angles (i.e. with increasing Q^) t h e 7r+n and TT^p intermediate states contribute with opposite sign. T h e peaked structure at t h e maximum possible value of the integration range in VK, i.e. Wmax = \fs — me^ is due t o the near singularity (in t h e electron mass) corresponding with quasireal Compton scattering (RCS), in which b o t h photons in t h e 27 exchange process become quasireal. This contribution at large W mainly drives t h e results for the inelastic part of t h e beam asymmetry. Furthermore, it is seen from Fig. 2 t h a t t h e inelastic a n d elastic contributions at a low energy of 0.2 GeV have opposite sign, resulting in quite a small asymmetry. It is somewhat puzzling t h a t t h e only experimental d a t a point at this energy
B. Pasquini, M. Vanderhaeghen: Single spin asymmetries in elastic electronnucleon scattering e +p ^ e +p
31
+ p ^ e +p 0 20 40 60 80 100 120 : E^ = 0.424 GeV ^ ' ^ ^ 140 50 100 150
100 120 140 160 180 Qc.m.(deg) Fig. 2. Beam normal spin asymmetry Bn for e~^p ^ e~p at a beam energy Ee = 0.2 GeV as function of the cm. scattering angle, for different hadronic intermediate states (X) in the blob of Fig. 1: N {dashed curve), nN {dasheddotted curve), sum of the N and TTN {solid curve). The data point is from the SAMPLE Cohaboration (MITBates) [9]
r^
©cmCdeg)
ec.».(deg)
Fig. 4. Beam normal spin asymmetry Bn for e~^p ^ e~p as function of the cm. scattering angle at different beam energies, as indicated on the figure. The meaning of the different lines is the same as in Fig. 2. The data points are from the A4 Cohaboration (MAMI) [10]
10 5
^cm. = ^O"*
0 5 10
6^^=60'*
/'".
50 0 50 500
Qcm. = 120"
/']
0 CAA
1.08
1.09
1.1
1.11
1.12
W (GeV)
Fig. 3 . Integrand in W of the beam normal spin asymmetry Bn for e~^p ^ e~p at a beam energy of Ee = 0.2 GeV and at different cm. scattering angles as indicated on the figure. The dashed curves are the contribution from the ir^p channel, the dasheddotted curves show the contribution from the Tr^n channel, and the solid curves are the sum of the contributions from channels. The vertical dashed line indicates the 7v~^n and the upper limit of the W integration, i.e. Wmax = ^/s — rrie
indicates a larger negative value at backward angles, although with quite large error bar. In Fig. 4, we show B^ at different beam energies below Ee = 1 GeV. It is clearly seen t h a t at energies Ee = 0.3 GeV and higher the elastic contribution yields only a very small relative contribution. Therefore Bn is a direct measure of the inelastic part which gives rise to sizeable large asymmetries, of the order of several tens of p p m
1.5
1.6 W (GeV)
Fig. 5. Integrand in W of the beam normal spin asymmetry Bn for e~^p ^ e~p at a beam energy of Ee = 0.855 GeV and at different cm. scattering angles as indicated on the figure. The meaning of the different lines is the same as in Fig. 3
in the backward angular range. At forward angles, the size of the predicted asymmetries is compatible with the first high precision measurements performed at MAMI. It will be worthwhile to investigate if the slight overprediction (in absolute value) of 5 ^ , in particular at Ee = 0.57 GeV, is also seen in a backward angle measurement, which is planned in the near future at MAMI. To gain a better understanding of how the inelastic contribution to Bn arises, we show in Fig. 5 the integrand of Bn at Ee = 0.855 GeV and at diflPerent scattering angles. T h e resonance structure is clearly reflected in the integrands for b o t h 7T~^n and n^p channels. At forward
32
B. Pasquini, M. Vanderhaeghen: Single spin asymmetries in elastic electronnucleon scattering
e +p
^e
+p 0.3 [
^
0.2
\
o ^ o ^
:^:^^^^^
^^^^^^^^^:
0.2
^cm. = ^O"*
0.4
i s 1
Qcm. = 60"
0.5 ^cm. = l^O'*
1.1
1.2
1.3
1.4
1.5
1.6
W (GeV)
Fig. 6. Target normal spin asymmetry An for e~'p^ —) e~'p as function of the cm. scattering angle at different beam energies, as indicated on the figure. The meaning of the different lines is the same as in Fig. 2
Fig. 7. Integrand in W of the target normal spin asymmetry An for e~p^ ^ e~p for a beam energy of Ee = 0.855 GeV and at different cm. scattering angles as indicated on the figure. The meaning of the different lines is the same as in Fig. 3
angles, the quasiRCS at the endpoint W = Wmax only yields a very small contribution, which grows larger when going to backward angles. This quasiRCS contribution is of opposite sign as the remainder of the integrand, and therefore determines the position of the maximum (absolute) value of Bn when going to backward angles. We next discuss the target normal spin asymmetry An. In Fig. 6, we show the results for b o t h elastic and inelastic contributions to A^ at different beam energies. At a low beam energy of £^e = 0.2 GeV, An is completely dominated by the elastic contribution. Going to higher beam energies, the inelastic contribution becomes of comparable magnitude as the elastic one. This is in contrast with the situation for Bn where the elastic contribution already becomes negligible for b e a m energies around ^ e = 0.3 GeV. T h e integrand of the inelastic contribution at a beam energy of Ee = 0.855 GeV is shown in Fig. 7. T h e total inelastic result displays a 7r+n threshold region contribution and a peak at the Z^(1232) resonance. Notice t h a t the higher resonance region is suppressed in comparison with the corresponding integrand for 5 ^ . Also the quasiRCS peak around the maximum W value is absent. As a result, the elastic contribution to An can be of comparable magnitude as the inelastic one. Due to the partial cancelation between elastic and inelastic contributions, An is significantly reduced for the proton, taking on values around or below 0.1 % for b e a m energies below 1 GeV.
Besides providing estimates for ongoing experiments, this work can be considered as a first step in the construction of a dispersion formalism for elastic electronnucleon scattering amplitudes. In such a formalism, one needs a precise knowledge of the imaginary part as input in order to construct the real part as a dispersion integral over this imaginary part. T h e real part of the twophoton exchange amplitudes may yield corrections to elastic electronnucleon scattering observables, such as the unpolarized cross sections or double polarization observables. Therefore it is of primary importance to quantify this piece of information, in order to increase the precision in the extraction of hadron structure quantities such as the nucleon form factors.
4 Conclusions
10. 11.
In this contribution, we presented calculations for beam and target normal SSAs in the kinematics where several experiments are performed or in progress.
12.
References
13.
A.V. Afanasev et al.: hepph/0208260 A.V. Afanasev, N.P. Merenkov: Phys. Lett. B 599, 48 (2004) M. Gorchtein et al.: Nucl. Phys. A 741, 234 (2004) L. Diaconescu, M.J. RamseyMusolf: nuclth/0405044 B. Pasquini, M. Vanderhaeghen: hepph/0405303 A. De Rujula et al.: Nucl. Phys. B 35, 365 (1971) D. Drechsel et al.: Nucl. Phys. A 645, 145 (1999) D. Drechsel et al.: Phys. Rep. 378, 99 (2003) S.P. Wells et al., (SAMPLE Coll.): Phys. Rev. C 63, 064001 (2001) F. Maas et al., (MAMI/A4 CoU.): nuclex/0410013 JLab HAPPEX2 experiment (E99115), spokespersons G. Gates, K. Kumar, D. Lhuillier JLab GO experiment (E00006, E01116), spokesperson D. Beck SLAG El58 experiment, contact person K. Kumar
Eur Phys J A (2005) 24, s2, 3334 DOI: 10.1140/epjad/s2005040062
EPJ A direct electronic only
Transverse single spin asymmetry in elastic electronproton scattering M. G o r c h t e i n \ P.A.M. Guichon^, and M. Vanderhaeghen^ ^ Universita di Geneva and Sezione INFN di Geneva, 16146 Geneva, Italy ^ SPhN/DAPNIA, CEA Saclay, F91191 Gif sur Yvette, France College of William and Mary & Jefferson Laboratory, Newport News, VA 23606, USA Received: 15 October 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. We discuss the twophoton exchange contribution to observables which involve lepton helicity flip in elastic leptonnucleon scattering. This contribution is accessed through the single spin asymmetry for a lepton beam polarized normal to the scattering plane. We estimate this beam normal spin asymmetry at large momentum transfer using a parton model and we express the corresponding amplitude in terms of generalized parton distributions. PACS. 25.30.Bf Elastic electron scattering netic form factors
12.38.Bx Perturbative calculations  13.40Gp Electromag
1 Introduction Elastic electronnucleon scattering in the onephoton exchange approximation gives direct access to the electromagnetic form factors of the nucleon, an essential piece of information about its structure. In recent years, the ratio GEP/GMP of the proton's electric to magnetic form factors has been measured up to large m o m e n t u m transfer Q^ in precision experiments [1,2] using the polarization transfer method. It came as a surprise t h a t these experiments for which is Q^ u p to 5.6 GeV^ extracted a ratio of GE^IGMP incompatible with unpolarized experiments [3,4,5] using the Rosenbluth separation technique. In [6], it was pointed out t h a t this discrepancy may be resolved by a precise account of the twophoton exchange effects which enter the radiative corrections to elastic form factors.
2 Elastic leptonnucleon scattering amplitude beyond the onephoton exchange In this work, we consider the elastic leptonnucleon scattering process l{k)\N{p) ^ l{k^)\N{p^). T h e general amplitude for elastic scattering of two spin1/2 particles can be parameterized in terms of six independent amplitudes. Three of t h e m describe such scattering without helicity flip on the lepton side [6], and the other three amplitudes do flip the lepton helicity [7]. We note t h a t in the onephoton exchange (Born) approximation, only two of t h e m survive, the well known electromagnetic form factors GE
Fig. 1. Handbag contribution to the 27exchange amplitude for elastic ey scattering
and GM which are real functions of the m o m e n t u m transfer Q^ only. In this work, we study the single spin asymmetry, Bn = ^^~^^ , where cr^d) denotes the cross section for an unpolarized target and a lepton b e a m spin parallel (antiparallel) to the normal polarization vector defined as Sii = (0, Sn), Si =  1 with Sn = [k X k ' ] /  [ k X k ' ]  . Its leading nonvanishing contribution is linear in the lepton mass. Furthermore, B^ vanishes in the Born approximation, and is therefore of relative order e^. Keeping only the leading t e r m in e^, Bn arises from an interference between the onephoton exchange (Born) amplitude and the imaginary part of the twophoton exchange amplitude [8]. To calculate the latter, we proceed with the model used in [7,9] where the partonic (handbag) model (cf. Fig. 1) has been adopted.
34
M. Gorchtein et al.: Transverse single spin asymmetry in elastic electronproton scattering ^^
c
pq
0.3
Keijt''Rn
0.2
E.(GeV) 11 6
0.]
i 1
1 t
1
1
1 I
; I
5
1
1
LiiL^L
0.1
_ 1
\
1
r
V
^
''
0.2 0.3 ^J
50
100
L
1
50
150
1 !
1 1
100
1
1 1 1
150
0 cm.
cm.
Fig. 2. Beam normal spin asymmetry for elastic e~N scattering as function of c m . scattering angle at different values of beam energy as indicated on the figure for the proton target (left panel), and the neutron target {right panel). The thick curves are the GPD calculations for the kinematical range where 5, —u > M^. For comparison, the nucleon pole contribution is also displayed {thin curves)
3 Results In this section we present our results for the asymmet r y Bn on the proton and neutron targets. In our calculation, we used the most recent parameterizations of the G P D s [10,11] for the evaluation of the lower blob of Fig. 1, while we calculate exactly the loop integral appearing in the upper part of Fig. 1, for details see [7,9]. As one can see from Fig. 2, the handbag mechanism predicts the effect of ^ 1.5 p p m for the proton, and ^ — 0.2 p p m for the neutron. For comparison, the contribution of the nucleon intermediate state (instead of the lower blob with G P D in Fig. 1) is shown for the same kinematics. This latter can be calculated exactly since it only contains the onshell elastic form factors of the nucleon. For the proton, the forward kinematics (30° < Ocm < 90°) looks promising for disentangling the inelastic contribution from the elastic one on the experiment. For the neutron, the effect is quite small due to partial cancellation of competing contributions. Further investigations of this observable at high m o m e n t u m transfers is necessary to obtain a valuable crosscheck for the real part of the 27exchange amplitude in order to resolve the present experimental situation with the elastic form factors. At present, there exist experimental d a t a on B^ [12,13], while the further several experiments are planned, aiming to measure this beam normal spin asymmetry in diflFerent kinematics [14,15,16].
References
10.
11. 12. 13. 14. 15. 16.
M.K. Jones et al.: Phys. Rev. Lett. 84, 1398 (2000) O. Gayou et al.: Phys. Rev. Lett. 88, 092301 (2002) L. Andivahis et al.: Phys. Rev. D 84, 5491 (1994) M.E. Christy et al.: nuclex/0401030 J. Arrington (JLab EOlOOl Collaboration): nuclex/0312017 P.A.M. Guichon, M. Vanderhaeghen: Phys. Rev. Lett. 9 1 , 142303 (2003) M. Gorchtein, P.A.M. Guichon, M. Vanderhaeghen: Nucl. Phys. A 741, 234 (2004) A. de Rujula, J.M. Kaplan, E. de Rafael: Nucl. Phys. B 35, 365 (1971) Y.C. Chen, A. Afanasev, S.J. Brodsky, C.E. Carlson, M. Vanderhaeghen: hepph/0403058 A.D. Martin, R.G. Roberts, W.J. Stirling, R.S. Thorne: Phys. Lett. B 531, 216 (2002); Feng Yuan, Phys. Rev. D 69, 051501 (2004) A.V. Radyushkin: Phys. Rev. D 58, 114008 (1998) S.P. Wells et al. (SAMPLE Collaboration): Phys. Rev. C 63, 064001 (2001) F. Maas et al. (MAMI/A4 Collaboration): in preparation SLAG El58 Experiment: contact person K. Kumar G. Gates, K. Kumar, D. Lhuillier, spokesperson(s): HAPPEX2 Experiment (JLab E99115) D. Beck, spokesperson(s): JLab/GO Experiment (JLab E0006, E01116)
Eur Phys J A (2005) 24, s2, 3538 DOI: 10.1140/epjad/s200504007l
EPJ A direct electronic only
Transverse spin asymmetry at the A4 experiment Experimental results Sebastian Baunack^, for the A4collaboration Johannes Gutenberg Universitat Mainz, Institut fiir Kernphysik, J.J. Becherweg 45, 55299 Mainz, Germany Received: 15 October 2004 / Pubhshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. The A4 cohaboration at the MAMI accelerator has measured the transverse spin asymmetry in the cross section of elastic scattering of transversely polarized electrons off unpolarized protons. An azimuthal dependence of the asymmetry has been observed, the amplitudes have been determined as =0.106 (GeV/c)^) = (8.59 0.S9stat 0.75syst) 10"^ and =0.230 (GeV/c)^) = (8.52 zb 2.31 stat 0.87 syst) 10 . arises from the imaginary part of the 27exchange amplitude. Our experimentally determined values of show that in the intermediate hadronic state not only the ground state of the proton, but also excited states contribute to the asymmetry. PACS. 13.40.Gp Electromagnetic form factors  11.30.Er Charge conjugation, parity, time reversal, and other discrete  13.40.f Electromagnetic processes and properties  14.20.Dh Properties of protons and neutrons
1 Introduction T h e A4 collaboration was founded to investigate the asymmetry in the elastic scattering of longitudinally polarized electrons off unpolarized protons. In recent years, theoretical effort has been spent to calculate the asymmetry for transversely polarized electrons [1,2]. T h e asymmetry arises from the interference between I 7 and 27 exchange (Fig. 1). T h e calculations take into account the intermediate hadronic state of the 27 exchange and might explain the discrepancy between the Rosenbluth separation technique and the polarization transfer m e t h o d for the dermination of the ratio G^/G^ of the electromagnetic form factors of the proton [3]. T h e A4 segmented lead fluoride (P6F2) calorimeter which covers the full 27r azimuthal range is an appropiate a p p a r a t u s to reveal the sinusoidal dependence of the asymmetry on the angle between the electron spin and the scattering plane. We have measured A_\_ at two b e a m energies 569.31 MeV and 855.15 MeV at scattering angles between 30° < 0 < 40° corresponding to m o m e n t u m transfers 0.106 (GeV/c)^ and 0.230 (GeV/c)^. In contrast to the parity violation measurements, the transverse electron spin causes physical asymmetries of nonnegligible order in our luminosity monitors. An extensive analysis of these asymmetries, which come from the 27 exchange in M0ller scattering, has been made in order to understand and control this effect. Our experimentally determined values of show t h a t in the intermediate hadronic state comprises part of PhD thesis
Fig. 1. 27 exchange not only the ground state of the proton, but also excited states contribute to the physical asymmetry [2].
2 Experimental setup T h e measurement principle is quite simple (Fig. 2): the polarised electron beam hits a hydrogen target and is scattered into the detector which counts the number A^^ of the elastic scattered particles for two opposite spin directions. jN'). T h e asymmetry is then A = (Ar+  Ar)/(Ar+ T h e measurement took place in the MAMI accelerator facility using the A4experiment setup, which has been described in detail in [4]. We used an electron beam with an intensity of 20 IJ,A. T h a n k s to the polarized electron source using a strained layer GaAs crystal the averaged beam polarization Pg was about 80 %. T h e spin of the electrons was reversed every 20 ms following a randomly selected pattern, (H h) or ( + +  ) . T h e polarization degree of the beam was measured by a M0ller polarimeter situated in another experimental hall. T h e spin of the
36
S. Baunack: Transverse spin asymmetry at the A4 experiment
./
N*
Fig. 4. Feynman graphs for M0ller scattering: one loop diagrams (left) and box diagrams (right). The arrow indicates the transverse polarisation of the incoming electron. The asymmetry arises from the interference term of the I7 and the 27exchange
<
20 0 20 40
60 P
Fig. 3. Schematic view of the PbF2 calorimeter from behind. The incoming electron's momentum vector ke is pointing out of the paper plane. The momentum vector ko^t of the outgoing electron can take all possible 0 values. Both together define the coordinate system. The direction of the electron polarization vector Pe for the ' + ' helicity state is indicated by the arrow. 6 is counted as indicated electrons was put into transverse direction with the help of a Wien filter located between the 100 keV electron source and the injector linac of the accelerator. T h e exact spin angle was determined by a M0llerMott polarimeter located at the b e a m d u m p . Several monitor and stabilization systems have been installed along the accelerator to minimize b e a m fluctuations t h a t introduce false asymmetries, such as differences in position, angle and energy for the two helicity states. T h e electrons were scattered on a 10 cm long liquid hydrogen target. T h e resulting luminosity is L ^ 0.5 10^^ cm~'^s~^. T h e luminosity was monitored by eight watercerenkov monitors put under small scattering angles 0 = 4.4°...8° covering the full azimuthal range (^ = 0...27r. T h e scattered electrons are detected in a total absorbing calorimeter t h a t consists of 1022 individual lead fluoride (P6F2) crystals placed in 7 rings and 146 rows. It covers scattering angles from ^ = 30°...40°, 0 = 0...2 TT and a solid angle Z\ J? = 0.64 sr. To see the azimuthal angular dependence of the asymmetry, we divide our calorimeter in the analysis into eight sectors, each sector covering A^ ?^ 45°. Figure 3 shows our convention of chosen angles and directions.
3 Physical asymmetry in M0ller scattering In order to calculate the experimental asymmetry A^xp, target density fluctuations have to be taken into account.
Fig. 5. Leading order asymmetry Aj^^ ^^ in M0ller scattering for a beam energy of E=569.31 MeV as a function of the scattering angle OCM in the center of mass system. The two dashed lines indicate the polar acceptance of the luminosity monitors
T h e elastic counts N"^ are normalised to the target density p^ = L^/I^, the ratio of luminosity L and beam current / :
^exp
N+
AT
N+
NP~
N^
N
L+  L 
/+ 
L+^L
I+ + I
I
(1) T h e approximation is valid u p to large asymmetries and certainly for the asymmetries in the A4 experiment. One sees t h a t the asymmetry measured in the luminosity monitors enters linearly into the experimental asymmetry A^xpWhile it is desired to eliminate asymmetries coming from real target density fluctuations, a physical asymmetry in the process of the luminosity measurement would enter false asymmetries into A^xp T h e eight luminosity monitors covering the same 0ranges as indicated for the calorimeter sectors in Fig. 3 detect mainly M0ller scattered electrons. It is known for many years [5] t h a t a ^depending asymmetry occurs in the M0ller scattering of transversely polarized electrons. Most recently, new calculations have been done for the E158 experiment [6]. Figure 4 shows the oneloop diagrams and the box diagrams for M0ller scattering. T h e physical asymmetry arises from the interference t e r m between these two diagrams. T h e leading order asymmetry ^M0iier .g pio^^ed in Fig. 5 for a b e a m energy of 569.31 MeV as a function of the scattering angle OCM in the center of mass system. T h e polar acceptance of the luminosity monitors are indicated by the grey box. Since the acceptance is nonsymmetric with respect to OCMI a nonvanishing physical asymmetry is observable.
S. Baunack: Transverse spin asymmetry at the A4 experiment
37
Table 1. NLO calculation for the transverse spin asymmetry A'] in M0ller scattering and comparison with the observed asymmetries A^"^^ in the luminosity monitors ^M0ller p ^ L O Calc.
Q' 0.106 (GeV/c)2 0.230 (GeV/c)2
AY""'
exp.
(26.7 (23.3
29.610"^ 16.610"^
^ ^
Table 2. Beam parameters for momentum transfer Q^=0.230 (GeV/c)^ achieved within 47 hours and Fig. 6. Extracted physical asymmetry in the luminosity mon Q^=0.106 (GeV/c)^Bachieved within 54 hours of pure data itors as a function of the azimuthal scattering angle cf) at taking Q^=0.106 (GeV/c)^ The electron spin was almost transverse Parameter 0.23 [GeV/cf 0.11 {GeV/cf with ^s = 85.1°. The fit gives A^^^^ = (24.3 2.9) 10"^ 300
^SO
Azlmulhal angle {deg)
(2.5 0.6) ppm (15.1 ) nm (126.7 ) nm (5.4 0.8) nrad (21.5 3.0) nrad (0.2 0.3) eV
Ai
Ax Ay Ax' Ay' AE
spin angle
Fig. 7. Extracted asymmetry A^^^'^ of the luminosity monitor as a function of the angle of the electron's spin angle ^s for Q^ =0.106 (GeV/c)^. The fit to the data gives ^Lumi ^ (_26.7zbl.3)10^
(0.4 ) ppm (157.0 ) nm (504.3 43.7) nm (20.5 12.9) nrad (45.4 3.9) nrad (41.4 ) eV
see M0ller electrons, but a certain 'dilution' of elastic scattered electrons. We understand qualitatively the occuring asymmetries. For a normalisation of the elastic counts, the luminosity signals have to added ^symmetrically, since the physical asymmetry in M0ller scattering then averages out.
4 Determination of transverse asymmetry An experimental verification and systematic check of the asymmetries in the luminosity monitors is possible during a spin rotation measurement. T h e Wien filter rotates the spin of the electrons from longitudinal to transverse direction in discrete steps by applying various currents. T h e corresponding spin angles are determined by the d u m p polarimeter. According to the spin angle a change in size and sign of the azimuthal dependence of the asymmetries in the luminosity monitors should be observable. Figure 6 shows the observed asymmetries for an almost transverse spin angle ^s = 85.1° and Q^ =0.106 (GeV/c)^. T h e sinusoidal behaviour is obvious. T h e amplitude is ^Lumi ^ (_24.3 =b 2.9) 1 0  ^ . Measurements at various spin angles ^s of A^^™^((^s) make a fit possible to determine A^^^. We performed such an analysis for b o t h of our so far applied m o m e n t u m transfers. Figure 7 shows the result for our low Q^ value. Due to target density fluctuations and shorter measurement times, the result for the higher Q^ value has a larger uncertainty. Table 1 gives the results of the NLO calculations for ^M0iier ^ ^ ^ ^^^ observed asymmetries ^^umi ^^ ^^^Q luminosity monitors. For a comparison of these values one has to keep in mind t h a t the luminosity monitors do not only
For the determination of the physical asymmetry we use the same technique t h a t we already used in the parity violating case [4]. T h e experimental measured asymmetry ^exp{^i) of each sector i can be written as the sum of the physical asymmetry Aphys{^i) and the helicity correlated false asymmetries Aj^ j = 1..6:
Aexpi^i)
=
P
' Aphysi^i)
+
^
ai
Xi
(2)
2=1
with P the b e a m polarisation, Xi the asymmetry in beam current, X2, X^ the diflFerences in horizontal and vertical position, X3, X4 diflFerences in horizontal and vertical angle and XQ difference in beam energy. We determine the coefficients a^ via a multiple linear regression out of the asymmetry d a t a itself. T h a n k s to the stabilisation systems, the quantities Xi could be kept small. Table 2 shows the achieved values for the two m o m e n t u m transfers. We have 47 hours b e a m d a t a for E=569.31 MeV and 54 hours for E=855.15 MeV, corresponding to a number of elastic events of Ntot = 3.9 10^^ and Ntot = 6.3 10^^ respectively. Figure 8 show the physical asymmetries as a function of the azimuthal scattering angle ^ . At the time of the
S. Baunack: Transverse spin asymmetry at the A4 experiment
38 JO
'
5
'J
<
I\
5 10
\5
nniiiiiu
{)
r.
t'
i
"uu. I.UUUI
40
Li_.
m
v\
' f ^^
t
<"d ^ ^
\
\
^=:^j^
1
^^^SN J ^N
^"1
L.^U„I.U.^. I...UL.L
120 IfjO
Li_.
20
I.UUU.I..
UI.J
J:i
\^
\ I
I.U^J.M.ULJ,
200 240 2K{) 320 360
0.2
^ \L
T
0.4
0^
0..S
I
1.2
1.4
Fig. 9. Our extracted asymmetries compared with model calculations from [2] as a function of the beam energy. For the dashed dotted line the intermediate nucleon state has been the ground state only. The dashed line represents the contribution of all possible TTN intermediate states. The solid line is the sum of both contributions
_
>;
j
i
^ 15
>1
5
f
^^V^
'
y
 ^\
< \i\
S 10 5 
y
~:r
 ^
E
t
/
^,
t»
/^
3
S 10 7^
1
Table 3 . Projected measurements for transverse single spin asymmetries with hydrogen and deuterium under forward angles ( 0 = (35 zb 5)°)
^15 ;,
40
m
1
1
1
1
1
1
1
1
J
1
1
120 160 200 240 280 320 360
Fig. 8. Extracted physical asymmetries for the eight sectors of the calorimeter for Q^=0.106 (GeV/c)^ {upper plot) and Q 2 ^ 0 . 2 3 0 (GeV/c)2
0.230 (GeV/c)^measurement, only sectors 1, 2, 5, 6 and part of sector 8 have been equipped with detectors. We fit to the d a t a an averaged function A^ which takes into account t h a t each sector covers about 45° in the azimuthal range:
Ee [MeV]
SAl hours [ppm] proton
300 420 570 854 1200 1500
0.5 0.5 0.5 0.5 1 2
20 40 90 300 260 180
5Af hours 6AI [ppm] deuteron [ppm] 0.5 0.5 0.5 0.5 1 2
20 35 70 220 180 120
20 11 7 4 6 11
/«
A1{^) =
A^ cos (t)'d(t)' = 0.765Ax cos (/)
(3)
J(P7r/8
Applying dead time corrections and corrections for aluminum dilution from the target entry and exit windows, we obtain: ^ ^ ( Q 2 = 0.11) = (  8 . 5 9 ^ ^ ( Q 2 = 0.23) = (  8 . 5 2
2.31,tat
O.Tbsyst)
10"'
0.87,^,t)
10"'
T h e first error represents the statistical uncertainty and the second error the systematic error.
5 Conclusion and outlook A comparison of our measured values of with model calculations from [2] shows t h a t the size of the observed asymmetries can not be explained if one takes into account only the elastic contribution t o the intermediate nucleon state (Fig. 9). Even a t our low Q^ nucleon resonances play an important role. For t h e future we plan a n extensive measurement program a t various energies under forward and backward angles on hydrogen as well as on deuterium (Table 3 and Table 4).
Table 4. Projected measurements for transverse single spin asymmetries with hydrogen and deuterium under backward angles ( 0 = (145 5)°) Ee
[MeV] 300 420 570 854
SAl [ppm] 3 3 4 8
hours proton 90 230 370 490
6Af [ppm] 2 2 3 7
hours deuteron 130 320 390 380
6AI [ppm] 10 10 13 28
References 1. P.A.M. Guichon, M. Vanderhaeghen: Phys. Rev. Lett. 9 1 , 142303 (2003) 2. B. Pasquini, M. Vanderhaeghen: hepph/0405303 (2003) 3. J. Arrington: nuclex/0408020 4. F.E. Maas et al.i Phys. Rev. Lett. 93, 022002 (2004) 5. A.O. Barut, C. Fronsdal: Phys. Rev. 120, 1871 (1960) 6. L. Dixon, M. Schreiber: Phys. Rev. D 69, 113001 (2004)
Eur Phys J A (2005) 24, s2, 3940 DOI: 10.1140/epjad/s2005040080
C D I A ^ ' 1CKJ A QirGCt electronic only
Normal beam spin asymmetries during the G forward angle measurement P.M. K i n g \ for the G° Collaboration^^ ^ University of Illinois at UrbanaChampaign, Urbana, IL, USA ^ Experiment E00006, Jefferson Laboratory, Newport News, VA, USA Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The vector analyzing power measured in elastic scattering of transversely polarized electrons from an unpolarized nucleon is directly proportional to the imaginary part of the two photon exchange amplitude. There has been recent interest in explorations of the two photon exchange amplitude, as the real part has been proposed as a possible resolution of the discrepancy between Rosenbluth separation and polarization observable measurements of the ratio of the electric to magnetic proton form factor. The vector analyzing power appears in the experiment as an azimuthal asymmetry. It has been measured previously in the SAMPLE and A4 experiment with different kinematics than those achievable with the G° apparatus. As part of the systematic checks for the G° forward angle measurement at TJNAF, the normal beam spin asymmetry in the G*^ detector array was measured with a 3 GeV beam incident upon a liquid hydrogen target. The experimental configuration was identical to the standard C° forward angle running except that the beam was transversely polarized in the plane of the accelerator. The data collected cover a range in center of mass angle from 19° to 37°, with an eightfold azimuthal symmetry. About 30 hours of data were taken in this configuration, resulting in an extracted vector analyzing power with a precision of a few ppm, which may already be able to provide some constraint on model predictions. PACS. 13.60.r Photon and chargedlepton interactions with hadrons  14.20.Dh Protons and neutrons 29.27.Hj Polarized beams in accelerators
1 Introduction
where Apv is the parity violating asymmetry, 0 and 0 are the electron scattering angles, and Ospin and (l)spin are the Elastic scattering of transversely polarized electrons from angles of the electron polarization vector, an unpolarized nucleon results in a n asymmetry due t o T h e G^ experiment in Hall C of Jefferson Laboratory twophoton exchange. T h e twophoton exchange produces is primarily designed to measure the parity violating eleca cross section dependence on the angle between the po tron scattering asymmetry from the proton. In the forlarization vector, P , and the scattering plane [1]: ward angle measurement, the recoil protons scattered by a 3 GeV polarized electron b e a m incident on a 20 cm long (J = (7o [1 + An'P n] (1) liquid hydrogen target are m o m e n t u m analyzed by a superconducting magnet, and are detected in 8 azimuthally where the vector analyzing power, A^, is proportional t o symmetric detector packages with 16 detectors in each the imaginary terms of the twophoton exchange ampli octant. T h e parity violating asymmetry in each detector tude. In a parity violating electron scattering measure varies from about 2 p p m t o about 20 p p m through the ment, in the general case, the normal b e a m spin asymme Q 2 ^^^^^ rp^ minimize the effect on the parity violating t r y will contribute t o the measured asymmetry, asymmetry measurement due t o t h e normal b e a m spin asymmetry, we set the polarization t o be longitudinal for A((9,(/)) = Pcos{Ospin)Apv{0) (2) ^Yie G^ measurements, and conducted the measurements + Psm{6spin) sin((/) — (f)spin)^n{0) of An discusscd in this paper. ^ The G° Experiment is supported by the U.S. National Science Foundation (NSF) under grant PHY9410768, the U.S. Department of Energy (DOE), the Natural Sciences and Engi 2 E x p e r i m e n t neering Research Council (NSERC) of Canada, and the Centre National de la Recherche Scientifique (CNRS) of France T h e transverse d a t a collection took place from 22 March through the Institut National de Physique Nucleaire et de through 26 March, 2004. At Jefferson Lab, the beam spin Physique des Particules (IN2P3). can be oriented in the horizontal plane (the plane of the ac
P.M. King: Normal beam spin asymmetries during the G° forward angle measurement
40
Table 1. Elastic proton kinematic coverage for the G° forward angle measurements. The estimated error corresponds to the statistical error on the measurements only Detectors
iQ')
(OCM)
Est. Error
14 58 912 1314 15
0.13 0.17 0.25 0.38 0.6
19.03 21.65 26.12 32.35 37.4
1.3 1.3 1.3 2.4 2.9
M
ppm ppm ppm ppm ppm 20
celerator) by adjusting the Wien filter in the injector. T h e Hall C M0ller polarimeter is only sensitive to the longitudinal component of the beam polarization. To determine the Wien filter setting which would produce transversely polarized beam in Hall C, a set of polarization measurements were made with Wien filter settings corresponding to spin angles of 95° to + 1 0 0 ° . T h e d a t a were fit to a cosine to find the setting which gave a zero of the longitudinal polarization; this was at a Wien angle of 85°. A set of three points at ° from the zero crossing were also taken at the beginning and end of the transverse running period, to check for drift of the zerocrossing; the zero crossing was found to have drifted by 3° during the four days of the measurement. T h e total spin precession through the accelerator was about 237r. T h e detector, target, and spectrometer magnet configurations were identical to those used in the s t a n d a r d G^ running. D a t a are simultaneously collected for elastically scattered protons, pions, and inelastic protons, which are separated by timeofflight from the target. T h e analysis is primarily concerned with the elastic protons, and will apply corrections for the contributions to the background under the proton peak which come from the pions and inelastic protons. T h e kinematic coverage and statistical error for the G^ detectors are shown in Table 1. T h e systematic error estimates are not shown. T h e detectors cover a range in center of mass scattering angle from 19° to 38°. T h e statistical errors should be compared to the results of a calculation [2], shown in Fig. 1, for the G^ b e a m energy which included intermediate states of the proton and TTN states with W <2 GeV.
3 Future plans T h e G^ collaboration had submitted a proposal to the Jefferson Lab PAC26 to carry out transverse asymmetry measurements in conjunction with the three beam energy running periods of the G^ backward angle measurement. T h e goal of this proposal was to make a 3 p p m measurement at a center of mass angle of about 130° at the three b e a m energies of 0.424, 0.585, and 0.799 GeV. At the highest beam energy, this would be a 10% measurement of the predictions of [2], as shown in Fig. 2.
40
60
80
100 120 140 160 180
Fig. 1. Theory prediction for G° forward angle beam energy from [2]. The dashed line is the nucleon intermediate state, the dashdot line is the TTN intermediate states, and the solid line is the total
IMI
151)
Fig. 2. Theory prediction for beam energies comparable to the G° backward angle measurements from [2]. The line types have the same meaning as in Fig. 1. The points are from the A4 measurements
4 Conclusion We have made measurements of the beam normal asymmetry for elastic ep scattering at a beam energy of 3 GeV for center of mass angles of 1938 degrees, with an estim a t e d statistical precision of a few ppm. We plan to make measurements at backward angles, with center of mass angles of about 130 degrees, at beam energies of 799, 585, and 424 MeV. T h e projected statistical error, for a dedicated experiment, is 3 ppm.
References 1. S.P. Wells et al. (SAMPLE Collaboration): Phys. Rev. C 63, 064001 (2001) 2. B. Pasquini, M. Vanderhaeghen: hepph/0405303
Ill Weak form factors of the nucleon II11 PV experiment
Eur Phys J A (2005) 24, s2, 4346 DOI: 10.1140/epjad/s200504009y
EPJ A direct electronic only
The axial form factor of the nucleon Elizabeth Beise University of Maryland, College Park, MD 20742, USA Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The parity violation programs at MITBates, Jefferson Lab and Mainz are presently focused on developing a better understanding of the seaquark contributions to the vector matrix elements of nucleon structure. The success of these programs will allow precise semileptonic tests of the Standard Model such as that planned by the QWeak collaboration. In order to determine the vector matrix elements, a good understanding of the nucleon's axial vector form factor as seen by an electron, C^, is also required. While the vector electroweak form factors provide information about the nucleon's charge and magnetism, the axial form factor is related to the nucleon's spin. Its Q^ = 0 value at leading order, QA, is well known from nucleon and nuclear beta decay, and its precise determination is of interest for tests of CKM unitarity. Most information about its Q^ dependence comes from quasielastic neutrino scattering and from pion electroproduction, and a recent reanalysis of the neutrino data have brought these two types of measurements into excellent agreement. However, these experiments are not sensitive to additional higher order corrections, such as nucleon anapole contributions, that are present in parityviolating electron scattering. In this talk I will attempt to review what is presently known about the axial form factor and its various pieces including the higher order contributions, discuss the various experimental sectors, and give an update on its determination through PV electron scattering. PACS. 12.15.Lk Neutral currents  ll.30.Er Charge conjugation, parity, time reversal, and other discrete symmetries  13.60.r Photon and chargedlepton interactions with hadrons  13.15.+g Neutrino interactions  14.20.Dh Protons and neutrons
1 Introduction
T h e neutral weak interaction between leptons and nucleons can be described by a set of three form factors t h a t contain information about nucleon structure. T h e goal of presentday experiments in parityviolating electron scattering has been to determine the two vector weak form factors, G^ and Gf^. These two form factors can be used along with the nucleon's electromagnetic form factors to disentangle the contributions of up, down and strange quarks to the nucleon's charge and magnetization distributions. However, in order to carry this out one also needs to know the third weak form factor coming from the nucleon's axial current, G^. T h e axial form factor has been determined at low m o m e n t u m transfer in b o t h quasielastic neutrino scattering and in pion electroproduction, but very little information on G^ is available at m o m e n t u m transfers greater t h a n 1 (GeV/c)^. In addition, the axial form factor as seen by an electron is substantially modified by electroweak radiative corrections t h a t cannot yet be computed with high precision, including the t e r m related to parityviolating coupling of a photon to the nucleon, known as the anapole coupling. Two experimental directions, quasielastic neutrino scattering and parityviolating quasielastic electron scattering from deuterium,
can improve our knowledge of b o t h G^ and of the anapole contribution.
2 Leptonnucleon scattering T h e nucleon electromagnetic current associated leptonnucleon scattering can be written as
{N'\j;\N)=UN
^^(^')^^ +
+
with
^^^'(^')
^Fj{q'){lh,q''l.%)l, 2MN
FE{q')
UN ,
where F^ and F2 are the wellknown Pauli and Dirac electromagnetic form factors, q^ = —Q^ is the fourm o m e n t u m transfered from the lepton to the nucleon. T h e t e r m containing the anapole form factor F']^{q^) violates parity, and F^{q^) is a form factor t h a t would arise with timereversal violation. T h e anapole form factor has been computed by several authors [1,2,3] and is expected to be small at Q^ = 0, but its computation is complicated by strong interaction effects in the nucleon, and its moment u m transfer dependence is unknown. It is negligible in
E. Beise: The axial form factor of the nucleon
44
electron scattering cross section measurements b u t it enters t h e asymmetry in parityviolating electron scattering at t h e same order as t h e weak nucleon axial form factor. T h e nucleon's neutral weak current is
{N'\J^^^J%\N)=UN
^Z(^2^^, , ^f(^')7.+
^^^^^.^^^(^2)
2Miv
+ 7M75G5(g')]^7V. T h e vector form factors F^ and F ^ are of primary interest in determining squark effects in P V electron scattering, and t h e axial form factor G\ contains information about the nucleon spin. At leading order, G\ can further be explicitly deconstructed using SU(3) symmetry into isovector and isospin singlet components t o separate out t h e contribution of 5quarks t o nucleon spin
G\m
T3GA(Q') + G ^ ( Q ' ) ,
where rs = + 1 (  1 ) for p ( n ) , G A ( 0 ) = {gA/gv) = 1.2670 5 [4] as determined in nucleon (3 decay, and G\{^) = Z\5, t h e strange quark spin content of t h e nucleon. T h e Q^ dependence of GA has generally been characterized by a dipole form, 1/(1 + Q'^/M\Y^ which can then be linked t o a determination of an axial radius in a low m o m e n t u m expansion of GA with Q^: {r\)
6 dGA,
12 MA
3 Available data Two methods have been used t o determine this lowest order Q^ behavior of t h e axial form factor. T h e most direct m e t h o d is t o use quasielastic neutrinonucleon scattering. Very little neutral current scattering d a t a is available, so cross section d a t a from t h e charged current process i'^\n ^ iJi~\p has typically been use t o extract MARecently, a new global fit t o neutrino d a t a was carried out by Budd, et al. [5], which improved over earlier fits b o t h by using t h e most recent determination of {gA/gv) along with new results for nucleon electromagnetic form factors. T h e improved fit gives MA = 1.001 =b 0.020 GeV. In pion electroproduction. MA can also be extracted from t h e transverse component of t h e nearthreshold electroproduction cross section by associating it with t h e electric dipole transition amplitude E^^^ through t h e low energy theorem of Nambu, Lurie and Shrauner [6] under t h e assumption t h a t m^r = 0. A measurement was recently carried out at t h e Mainz Microtron [7], resulting in MA = 1.068=b0.015 GeV. In a recent topical review, Bernard et al. [8] used chiral perturbation theory t o compute a finite mass correction to this extraction, which is substantial and results in a corrected MA of 1.013 5 GeV, bringing it into agreement with t h e neutrino data. Therefore, it appears t h a t MA is reasonably well determined and t h e low Q^ behavior of GA can at least be described phenomenologically. This does not, however, give a first principles theoretical description of GA{Q'^)^ nor does it provide an adequate
description at m o m e n t u m transfers above 1 (GeV/c)^. In addition, better modeling of neutrino scattering, guided by improved data, will be required for upcoming neutrino oscillation experiments. A new experiment, Minerz/a [9], has been proposed t h a t would consist of a high granularity neutrino detector located at t h e NUMI beam line at Fermilab. This experiment would be able t o provide a precise determination of G A , including possible departures from the nominal dipole behavior, at Q^ < 2 (GeV/c)^ and a first determination of GA for Q^ > 2 (GeV/c)^. Planning for another potential experiment is underway at Jefferson Lab, using t h e reaction e\p ^ jy \n^ covering t h e range Q^ '^ 1 — 3 (GeV/c)^ [10]. This very challenging experiment would require detection of t h e recoiling neutrons at very forward angles, and t h e parityviolating asymmetry in t h e detected neutrons would be measured in order t o constrain backgrounds. While t h e above measurements would be able t o better and provide improved models of cross constrain GA(Q'^) section d a t a for neutrino oscillation experiments, it is t h e axial form factor as seen by an electron t h a t is relevant t o the parity violation program t h a t is t h e topic of this workshop. T h e axial form factor seen in P V electron scattering can be written, going beyond first order, as G^iQ'
rs{l + RI=')GA{Q')
+ RI='G'A{Q')
+
G\{Q'
where Rj^~ ' are electroweak radiative corrections arising from higher order diagrams ^. T h e SU(3) octet form factor G\ is not present at treelevel, b u t appears once radiative corrections are included. Its Q^ = 0 value can be estimated from t h e ratio of axial vector t o vector couplings in hyperon (3 decay which, assuming SU(3) flavor symmetry, can be related t o t h e octet axial charge ag and to t h e hyperon F and D coefficients [4],
G^(0)
( 3 F  D) _ 1 2x/3
~ 2
as = 0.217
.
Its Q^ behavior has also not been measured, but it is usually assumed t o have t h e same dipole form as t h e isovector form factor GA{Q'^) with t h e same mass parameter MAIt should be noted t h a t while a decade of measurements related t o t h e "spin crisis" have indirectly determined GA{0) = As from polarized deepinelastic scattering, its Q^ behavior is also unknown. There has been one determination of G^j^{0) from quasielastic neutrino scattering [12], which is in reasonable agreement with t h e polarized DIS data. An improved analysis of these d a t a was carried out by Garvey et al. [13] who included possible effects of nonzero strange vector form factors. A recent further improved analysis was carried out by S. P a t e [14], who combined t h e neutrino d a t a with results from H A P P E X to perform a global fit t o t h e three strange form factors (so far with only two constraints) t o extract G\ at t h e mean of t h e two experiments, Q^ = 0.5 (GeV/c)^, rather t h a n extrapolating t o Q^ = 0. A new direct measurement of G^j^ at low m o m e n t u m transfer has been proposed using I am here following the notation in [11].
E. Beise: The axial form factor of the nucleon
45
Table 1. Electroweak radiative corrections, computed in the MS scheme, for the axial form factor measured in PV electron scattering. The values are taken from [1] Source
RA
1quark anapole total
0.18 0.06 0.24
0)
pT=0 HA
0.24 4
0.07 0.01 0.08
E E (0
4 4
<
(0
o > Q
the ratio of neutral current to charged current neutrino scattering at low m o m e n t u m transfer, FiNeSSE, using a highly segmented detector with wavelength shifting optical fibers embedded in mineral oil to identify tracks left by the recoiling protons [15]. This experiment would potentially improve the determination of As by about a factor of two over the DIS data, and with less theoretical uncertainty. Of more direct interest to the parity violation program is the radiative correction to the isovector G A ( Q ^ ) , of which the nucleon's anapole form factor F ^ is one component. T h e dominant contributions to R^^ and R^^ come from 1quark terms such as 7  Z mixing and vertex corrections, which have been computed by several authors [1,4]. Multiquark or anapole contributions were also computed [1], who modeled t h e m in terms of hadronic parityviolating NN couplings cast within a heavy baryon chiral perturbation theory framework. T h e results are shown in Table 1. While the 1quark contributions dominate the correction, the anapole contributions dominate the uncertainty. T h e axial form factor G ^ , or at least its isovector piece G^*^ \ can be determined from the P V asymmetry in quasielastic scattering from deuterium, where the strange quark effects in the neutron and proton tend to cancel. Nuclear effects, including b o t h parity conserving [17] and parityviolating [18,19] contributions, have been shown to be small. T h e first measurement of G\^ ^^^ was carried out by the S A M P L E collaboration. T h e measured asymmetry at two m o m e n t u m transfers are shown in Fig. 1. They agree fairly well with the calculation, which was carried out at Q^ = 0, indicating t h a t there is no anomalously large Q^ dependence to the anapole t e r m or to the correction at these very low m o m e n t u m transfers. However, very little else is known about its behavior away from Q^ = 0. Two model calculations [3,2] of Fj^{Q'^) have been carried out, and they indicate a much softer behavior with Q^ t h a n t h a t of ^ ^ ( Q ^ ) , even possibly an increase with Q^, as well as quite different behavior for the isoscalar and isovector pieces. These could substantially enhance the effects of radiative corrections at m o m e n t u m transfers in the range of the GO experiment. It would thus be very useful to have some experimental information on G\^ ^^^ at higher m o m e n t u m transfers. A program of backward angle measurements with a deuterium target is part of the planned running for the GO experiment, and aerogel Cerenkov detectors have been added to the detector array
ftk
d" (Gey/c) Fig. 1. Asymmetry results from the two SAMPLE deuterium experiments {solid circles, see [16]), compared to expectation from theory using the axial radiative corrections of [1] (open circles). The theory also assumes a value of GM of 0.15 nuclear magnetons, and the grey band represents a change in GM of 6 n.m. in order to identify and separate charged pions produced in the deuterium target from the desired quasielastically scattered electrons. These d a t a will not only reduce the model uncertainties in the determination of G% and G%f from the hydrogen data, but will also allow the first experimental information on G\ away from the static limit. Shown in Fig. 2 are the projected uncertainties from the GO experiment [20] in the difference between G^^^^"^^ and the treelevel G^ t h a t is seen in neutrino scattering, along with the calculation of [1] and the two S A M P L E measurements. As an aside, it should be noted t h a t , assuming t h a t a determination of the nucleon axial form factor can straightforwardly be related to electronquark interactions, the two S A M P L E measurements can be recast in terms of the two electronquark couplings C2U and C2d' Prior to the S A M P L E measurements, experimental limits on these were from the original SLAG DIS parityviolation experiment [21], and from the parityviolating quasielastic elecron scattering experiment on ^Be carried out at Mainz [22]. T h e two S A M P L E measurements are sensitive to the combination C2U — C2d' These are modified by 1quark radiative corrections, and in the case of elastic eA^ scattering the multiquark corrections as well. In order to compare directly to the SLAG DIS data, the multiquark radiative corrections must be removed, which although small, dominate the uncertainty. T h e resulting values from the 200 MeV and 125 MeV d a t a sets, respectively, are C2u  C2d =  0 . 0 4 2 C2u  C2d =  0 . 1 2
0.040 0.05
0.035 0.05
0.02
0.02
0.01,
where the first two uncertainties are statistical and experimental systematic, the third is t h a t due the radiative corrections, and, for the 125 MeV case, the last corresponds to variations in G^j^ by 6 because it is unde
E. Beise: The axial form factor of the nucleon
46
1.5
T"
T"
T"
measurements. In P V electron scattering, the axial form factor is substantially modified and very little is known about the Q^ behavior of the higher order terms. T h e GO experiment will uniquely be able to provide the higher Q^ d a t a through quasielastic scattering from a deuterium target.
Zhu calc. A SAMPLE 125 MeV SAMPLE 200 MeV
1.0 P 0.5
h
o^O.O
h References
.0.5 h ^1.0 h < 1.5 h
n GO projected uncert.
2.0 0.0
0.4
0.2 Q2
0.6
0.8
1.0
(GeV/c)2
Fig. 2. The difference between the axial form factor seen by an electron in parityviolating electron scattering and Gj^(Q'^), the treelevel form factor. Shown are the calculation of [1], SAMPLE data, and projected uncertainties in the GO experiment
termined at this m o m e n t u m transfer. These values are in good agreement with the Standard Model prediction [4], and represent a significant improvement over the earlier data.
4 Conclusion In summary, while much attention has been focused on determination of the neutral weak vector form factors, in order to extract strange quark effects in the nucleon, there are variety of experimental avenues to pursue in the near future to improve our knowledge of the nucleon's axial form factor. T h e treelevel form factor is now known reasonably well at low m o m e n t u m transfers from neutron b e t a decay, from quasielastic neutrino scattering and from pion electroproduction. Its knowledge at higher mom e n t u m transfer, including potential deviations from a generic dipole behavior, can be improved with new neutrino scattering experiments, and these experiments will help provide the required precision cross section information needed for the next generation of neutrino oscillation
1. S.L. Zhu, S.J. Puglia, B.R. Holstein, M.J. RamseyMusolf: Phys. Rev. D 62, 033008 (2000) 2. D.O. Riska: NucL Phys. A 678, 79 (2000) 3. C M . Maekawa, U. van Kolck: Phys. Lett. B 478, 73 (2000); C.M. Maekawa, J.S. Viega, U. van Kolck: Phys. Lett. B 488, 167 (2000) 4. K. Hagiwara et al.: Review of Particle Properties, Phys. Rev. D 66, 010001 (2002) 5. H. Budd, A. Bodek, J. Arrington: arXiv:hepex/0308005 6. Y. Nambu, D. Lurie: Phys. Rev. 125, 1429 (1962); Y. Nambu, E.Shrauner: Phys. Rev. 128, 862 (1962) 7. A. Liesenfeld et al.: Phys. Lett. B 468, 20 (1999) 8. V. Bernard, L. Elouadrhiri, U.G. Meissner: J. Phys. G 28, R l (2002) 9. Minerz/a proposal: K. McFarland, spokesperson 10. Jefferson Laboratory PAC25 Letter of Intent LOI04006: A. Deur, contact 11. M.J. Musolf, T.W. Donnelly, J. Dubach, S.J. Pollock, S. Kowalski, E.J. Beise: Physics Reports 239, 1 (1994) 12. L.A. Ahrens et al.: Phys. Rev. D 35, 785 (1987) 13. G.T. Garvey, W.C. Louis, D.H. White: Phys. Rev. C 48, 761 (1993) 14. S. Pate: Phys. Rev. Lett. 92, 082002 (2004) 15. FiNeSSE proposal, R. Tayloe, B. Fleming: contacts 16. T. Ito et al.: Phys. Rev. Lett. 92, 102003 (2004) 17. L. Diaconescu, R. Schiavilla, U. van Kolck: Phys. Rev. C 63, 044007 (2001) 18. R. Schiavilla, J. Carlson, M. Paris: Phys. Rev. C 67, 032501 (2003) 19. C.P. Liu, G. Prezeau, M.J. RamseyMusolf: Phys. Rev. C 67, 035501 (2003) 20. GO Backward angle proposal: D. Beck, contact. Uncertainties were projected based on 80 yit A of (normal time structure) beam on a 20 cm deuterium target, along with uncertainties achieved in the forward angle measurement that was carried out in 2004 21. C.Y. Prescott et al.: Phys. Lett. B 84, 524 (1979) 22. W. Heil et al.: NucL Phys. B 327, 1 (1989)
Eur Phys J A (2005) 24, s2, 4750 DOI: 10.1140/epjad/s2005040106
EPJ A direct electronic only
Parity violating electron scattering at the M A M I facility in Mainz The strangeness contribution to the form factors of the nucleon F.E. Maas, for the A4Collaboration Johannes Gutenberg Universitat Mainz, Institut fiir Kernphysik, J.J.Becherweg 45, 55299 Mainz, Germany email: maasQkph. unimainz. de
Received: 15 December 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. We report here on a new measurement of the parity violating (PV) asymmetry in the scattering of polarized electrons on unpolarized protons performed with the setup of the A4collaboration at the MAMI accelerator facility in Mainz. This experiment is the first to use counting techniques in a parity violation experiment. The kinematics of the experiment is complementary to the earlier measurements of the SAMPLE collaboration at the MIT Bates accelerator and the HAPPEX collaboration at Jefferson Lab. After discussing the experimental context of the experiments, the setup at MAMI and preliminary results are presented. PACS. 12.15 Electroweak interactions  ll.30.Er Charge conjugation, parity, time reversal, and other discrete  13.40.Gp Electromagnetic form factors  25.30.Bf Elastic electron scattering
1 Strangeness in the nucleon T h e understanding of q u a n t u m chromodynamics (QCD) in the nonperturbative regime of low energy scales is crucial for understanding the structure of hadronic m a t t e r like protons and neutrons (nucleons). T h e successful description of a wide variety of observables by the concept of effective, heavy ( ^ 350 MeV) constituent quarks, which are not the current quarks of Q C D , is still a puzzle. There are other equivalent descriptions of hadronic m a t t e r at low energy scales in terms of effective fields like chiral perturbation theory ( x P T ) or Skyrmetype soliton models. T h e effective fields in these models arise dynamically from a sea of virtual gluons and quarkantiquark pairs. In this context the contribution of strange quarks plays a special role since the nucleon has no net strangeness, and any contribution of strange quarks to the nucleon structure observables is a pure seaquark effect. Due to the heavier current quark mass of strangeness (m^) as compared to u p (mu) and down (m^^) with rris ~ 140 MeV ^ rriu^rad ^ 510 MeV, one expects a suppression of strangeness effects in the creation of quarkant iquark pairs. On the other hand the strange quark mass is in the range of the mass scale of Q C D {rris ~ ^QCD) SO t h a t the dynamic creation of strange sea quark pairs could still be substantial. T h e topic of the strangeness content of the nucleon has been widely discussed in other contributions to this conference from J. Ellis, A. Silva, B. Kubis and others. Parity violating (PV) electron scattering off nucleons provides experimental access to the strange quark vector current in the nucleon {N\s'^ij^s\N) which is parameterized in the
electromagnetic form factors of proton and neutron, G% and G ^ [1]. Recently the SAMPLE, H A P P E X  and A4collaboration have published experimental results. A direct separation of electric (G;) and magnetic {G%^) contribution at forward angle has been impossible so far, since the measurements have been taken at different Q^values. T h e experimental details of the S A M P L E , H A P P E X , A4 and GO collaboration is discussed in greater detail in many long and short contributions of this conference proceedings. A determination of the weak vector form factors of the proton ( G ^ and C ^ ) by measuring the P V asymmetry in the scattering of longitudinally polarized electron off unpolarized protons allows the determination of the strangeness contribution to the electromagnetic form factors G% and G%^. T h e weak vector form factors &^ ^ of the proton can be expressed in terms of the known electromagnetic nucleon form factors G^^j^ and the unknown strange form factors Gl; ^ using isospin symmetry and the universality of the quarks in weak and electromagnetic interaction ^^E,M
= i^^E,M ~ ^"E.M)
~ "^Si^
^W G^E,M ~ ^E,M'
^hc
interference between weak (Z) and electromagnetic (7) amplitudes leads to a P V asymmetry ALR{ep) = {aR — (^L)/{(^R \ CFL) in the elastic scattering cross section of right and lefthanded electrons {CFR and <7L respectively), which is given in the framework of the Standard Model [2]. ALR{'^P) is of order parts per million (ppm). T h e asymmet r y can be expressed as a sum of three terms, ALR{ep) = Ay \ As \ A A' Ay represents the vector coupling at the proton vertex where the possible strangeness contribution
48
F.E. Maas: Parity violating electron scattering at the MAMI facility in Mainz
has been taken out and has been put into Ag, a t e r m arising only from a contribution of strangeness to the electromagnetic vector form factors. T h e t e r m A A represents the contribution from the axial coupling at the proton vertex due to the neutral current weak axial form factor G^. We over the acceptance of the detector average AQ = AV\AA and the target length in order to calculate the expected asymmetry. T h e largest contribution to the uncertainty of AQ comes from the uncertainty in the axial form factor GA^ the electric form factor of the proton C ^ , and the magnetic form factor of the neutron G^.
2 The A4 experimental setup and analysis T h e A4 experiment at MAMI is complementary to other experiments since for the first time counting techniques have been used in a scattering experiment measuring a P V asymmetry. Therefore possible systematic contributions to the experimental asymmetries and the associated uncertainties are of a different n a t u r e as compared to previous experiments, which use analogue integrating techniques. More details on the A4 a p p a r a t u s can be found in the contributions of J. Diefenbach, B. Glaser, Y. Imai, J. H. Lee and C. Weinrich. T h e P V asymmetry was measured at the MAMI accelerator facility in Mainz [3] using the setup of the A4 experiment [4]. T h e polarized 570.4 and 854.3 MeV electrons were produced using a strained layer GaAs crystal t h a t is illuminated with circularly polarized laser light [5]. Average beam polarization was about 8 0 % . T h e helicity of the electron beam was selected every 20.08 ms by setting the high voltage of a fast Pockels cell according to a randomly selected p a t t e r n of four helicity states, either ( + P  P  P + P ) or (  P + P + P  P ) . A 20 ms time window enabled the histogramming in all detector channels and an integration circuit in the beam monitoring and luminosity monitoring systems. T h e exact window length was locked to the power frequency of 50 Hz in the laboratory by a phase locked loop. For normalization, the gate length was measured for each helicity. Between each 20 ms measurement gate, there was an 80 /iS time window for the high voltage at the Pockels cell to be changed. T h e intensity / = 20 fiA of the electron current was stabilized to better t h a n SI/I ~ 10~^. An additional A/2plate in the optical system was used to rotate small remaining linear polarization components and to control the helicity correlated asymmetry in the electron beam current to the level of < 10 p p m in each five minute run. From the source to the target, the electron beam develops fluctuations in beam parameters such as position, energy and intensity which are partly correlated to the reversal of the helicity from + P to P. We have used a system of microwave resonators in order to monitor beam current, energy, and position in two sets of monitors separated by a drift space of about 7.21 m in front of the hydrogen target. In addition, we have used a system of 10 feedback loops in order to stabilize current, energy [6], position, and angle of the beam. T h e polarization of the electron b e a m was measured with an accuracy of 2 % using a M0ller polarimeter
20 15
w
10 5 0 5
F
w
W
10
out out font but but but out ' out out. 15 in in in in m in m m 20 10 12 14 16 Sample No. 5 4 3 2 1 — 0 1 _ ...j...i T i 2 3 in 1 out in i out in 1 out in 4 lout
it
tj
5
1
1
2
i
1
3
1
1
4
5
i
1
6
1
i
1
7 8 Sample No.
Fig. 1. The top plot shows the data samples of 854.3 MeV data with the A/2plate in and out. The lower plots represents the data sample for the 570.4 MeV data with the A/2plate in and out as described in the text
which is located on a b e a m line in another experimental hall [7]. Due to the fact t h a t we had to interpolate between the weekly M0ller measurements, the uncertainty in the knowledge of the beam polarization increased to 4 % . T h e 10 cm high power, high flow liquid hydrogen target was optimized to guarantee a high degree of turbulence with a Reynoldsnumber of i? > 2 x 10^ in the target cell in order to increase the effective heat transfer. For the first time, a fast modulation of the b e a m position of the intense C W 20 fiA beam could be avoided. It allowed us to stabilize the beam position on the target cell without target density fiuctuations arising from boiling. T h e total thickness of the entrance and exit aluminum windows was 250 /xm. T h e luminosity L was monitored for each helicity state (R, L) during the experiment using eight waterCherenkov detectors (LuMo) t h a t detect scattered particles symmetrically around the electron b e a m for small scattering angles in the range of Oe = 4.4° — 10°, where the P V asymmetry is negligible. T h e photomultiplier t u b e currents of these luminosity detectors were integrated during the 20 ms measurement period by gated integrators and then digitized by customized 16bit analoguetodigital converters (ADC). T h e same method was used for all the beam parameter signals. A correction was applied for the nonlinearity of the luminosity monitor photomultiplier tubes. From the beam current helicity pair d a t a / ^ ' ^ and lumi
F.E. Maas: Parity violating electron scattering at the MAMI facility in Mainz
SAMPLE <
2
—
0.1
HAPPEX
1
0.05 ,A4
^A4
'f 1
G^E 0
PH
<
49
0 1
0.05
2 3
0.1
4
~
1
1
0.1
1
0.2
1
1
0.3
1
1
1
0.1
1
0.4 0.5 Q2 [(GeV/c)2]
Fig. 2. Difference between the measured parity violating asymmetry in electron proton scattering ALR(ep) and the asymmetry AQ without vector strangeness contribution from the Standard Model for the SAMPLE experiment at backward ies^ and the HAPPEX and the two A4 results at forward 'les. The new experimental result at (3^=0.108 (GeV/c)^ presented here is the most accurate measurement
nosity monitor helicity pair L ^ ' ^ d a t a we calculated t h e target density p ^ ' ^ = L^^^/JR^L ^^^ ^^^e two helicity states independently. To detect t h e scattered electrons we developed a new type of a very fast, hom*ogeneous, total absorption calorimeter consisting of individual lead fluoride (PbF2) crystals [8,9]. T h e material is a pure Cherenkov radiator and has been chosen for its fast timing characteristics and its radiation hardness. This is t h e first time this material has been used in a large scale calorimeter for a physics experiment. T h e crystals are dimensioned so t h a t an electron deposits 96 % of its total energy in an electromagnetic shower extending over a m a t r i x of 3 x 3 crystals. Together with t h e readout electronics this allows us a measurement of t h e particle energy with a resolution of 3.9%/\/E and a total dead time of 20 ns. For t h e d a t a taken at 854.3 MeV only 511 out of 1022 channels of t h e detector and t h e readout electronics were operational, for t h e 570.4 MeV d a t a all t h e 1022 channels were installed. T h e particle rate within t h e acceptance of this solid angle was ^ 50x 10^ s~^. Due t o the short dead time, t h e losses due t o double hits in t h e calorimeter were 1 % at 20 //A. This low dead time is only possible because of t h e special readout electronics employed. T h e signals from each cluster of 9 crystals were summed and integrated for 20 ns in an analogue summing and triggering circuit and digitized by a transient 8bit A D C . There was one summation, triggering, and digitization circuit per crystal. T h e energy, helicity, and impact information were stored together in a three dimensional histogram. T h e number of elastic scattered electrons is determined for each detector channel by integrating t h e number of events in an interval from 1.6 cr^ above pion production threshold to 2.0 aE above t h e elastic peak in each helicity histogram, where aE is the energy resolution for nine crystals. These cuts ensure a clean separation between elastic scattering and pion production or Z^excitation which
A4 (Q =0.230 (Ge\fc) (GeV/c)')
HAPPEX (Q'=0 0.05
0.05
0.1
Fig. 3. The solid line represents all possible combinations of G% \0.225GM as extracted from the work presented here at a Q^ of 0.230 (GeV/c)^. The densely hatched region represents the l(j uncertainty. The recalculated result from the HAPPEX published asymmetry at Q^ of 0.477 (GeV/c)^ is indicated by the dashed line, the less densely hatched area represents the associated error of the HAPPEX result
has an unknown P V cross section asymmetry. T h e linearity of t h e PbF2 detector system with respect to particle counting rates and possible effects due to dead time were investigated by varying the b e a m current. We calculate t h e raw normalized detector asymmetry as ^raw =
(N^/pR
 N^/p^)/{N^/p^
+ N^/p^).
The possible di
lution of t h e measured asymmetry by background originating from t h e production of TT^'S t h a t subsequently decays into two photons where one of t h e photons carries almost t h e full energy of an elastic scattered electron was estimated using Monte Carlo simulations to be much less t h a n 1 % and is neglected here. T h e largest background comes from quasielastic scattering at t h e thin aluminum entrance and exit windows of t h e target cell. We have measured t h e aluminum quasielastic event rate and calculated in a static approximation a correction factor for t h e aluminum of 1.030 =b 0.003 giving a smaller value for t h e corrected asymmetry. Corrections due to false asymmetries arising from helicity correlated changes of b e a m parameters were applied on a run by run basis. T h e analysis was based on t h e five minute runs for which t h e counted elastic events in t h e PbF2 detector were combined with t h e correlated b e a m parameter and luminosity measurements. In t h e analysis we applied reasonable cuts in order to exclude runs where t h e accelerator or p a r t s of t h e PbF2 detector system were malfunctioning. T h e analysis is based on a total of 7.3 x 10^ histograms corresponding to 4.8 x 10"^^ elastic scattering events for t h e 854.3 MeV d a t a and 4.8 10^ histograms corresponding to 2 10^^ elastic events for t h e 570.4 MeV data. For t h e correction of helicity correlated b e a m parameter fluctuations we used multidimensional linear regression analysis using t h e data. T h e regression parameters have been calculated in addition from the geometry of t h e precisely surveyed detector geometry. T h e two different methods agree very well within statistics. T h e experimental asymmetry is normalized to t h e electron b e a m po
50
F.E. Maas: Parity violating electron scattering at the MAMI facility in Mainz
the 570.4 MeV d a t a are displayed. A recent very accurate determination of the strangeness contribution to the magnetic moment of the proton //g = G\^{Q'^ = 0 (GeV/c)^) from lattice gauge theory [11] would yield a larger value of G% = 0.076 =b 0.036 if the Q^ dependence from 0 to 0.108 (GeV/c)^ is neglected. T h e theoretical expectations for another quenched lattice gauge theory calculation [14], for SU(3) chiral perturbation theory [15], from a chiral soliton model [16], from a quark model [17], from a Skyrmetype soliton model [18] and from an u p d a t e d vector meson dominance model fit to the form factors [19] are included into Fig. 4. Our results concerning the measurement of the transFig. 4. The solid lines represent the result on G% + O.IOGCM as extracted from our new data at Q^ = 0.108 (GeV/c)^ pre verse beam spin asymmetry have been presented in the sented here. The hatched region represents in all cases the one contribution of S. Baunack. We are also analyzing our <juncertainty with statistical and systematic and theory error d a t a in the region of the ZA(1232) resonance, which has added in quadrature. The dashed lines represent the result on been presented in the contribution of L. Capozza. We are GM from the SAMPLE experiment [10]. The dotted lines rep preparing a series of measurements of the parity violatresent the result of a recent lattice gauge theory calculation for ing asymmetry in the scattering of longitudinally polarfis [11] The boxes represent different model calculations and ized electrons off unpolarized protons and deuterons unthe numbers denote the references der backward scattering angles of 140° < Oe < 150° with the A4 a p p a r a t u s in order to separate the electric (G^^) and magnetic {G%^) strangeness contribution to the eleclarization Pg to extract the physics asymmetry, Aphys = tromagnetic form factors of the nucleon. ^ e x p / ^ e  We have taken half of our d a t a with a second A/2plate inserted between the laser system and the GaAs crystal. This reverses the polarization of the electron b e a m Acknowledgements. This work has been supported by the DFG and allows a stringent test of the understanding of sys in the framework of the SFB 201 and SPP 1034. We are intematic effects. T h e effect of the plate can be seen in debted to K.H. Kaiser and the whole MAMI crew for their Fig. 1: the observed asymmetry extracted from the dif tireless effort to provide us with good electron beam. We are ferent d a t a samples changes sign, which is a clear sign of also indebted to the AlCollaboration for the use of the Moeller parity violation if, as in our case, the target is unpolarized. polarimeter. Our measured result for the P V physics asymmetry in the scattering cross section of polarized electrons on unpolarized protons at an average Q^ value of 0.230 (GeV/c)^ References is ALR{ep) = (  5 . 4 4 zb OMstat 0.265^5^) p p m for the 1. D.B. Kaplan et al.: Nucl. Phys. B 310, 527 (1988) 854.3 MeV d a t a [12] and ALR{ep) = (  1 . 3 6 =b 0.29stat 2. M. Musolf et ah: Phys. Rep. 239, 1 (1994) 0.13syst) p p m for the 570.4 MeV d a t a [13]. T h e first error 3. H. Euteneuer et al.: Proc of the EPAC 1994 1, 506 (1994) represents the statistical accuracy, and the second rep4. F.E. Maas et al.: Eur. Phys. J. A 17, 339 (2003) resents the systematical uncertainties including b e a m po5. K. Aulenbacher et al.: Nucl. Ins. Meth. A 391, 498 (1997) larization. T h e absolute accuracy of the experiment repre6. M. Seidl et al.: Proc. of the EPAC 2000 , 1930 (2000) sents the most accurate measurement of a P V asymmetry 7. P. Bartsch: Dissertation Mainz (2001) in the elastic scattering of longitudinally polarized elec8. F.E. Maas et al.: Proc. of the ICATPP7, (World Scientific, trons on unpolarized protons. Figure 2 shows the present 2002), p. 758 situation of all the published asymmetry measurements 9. P. Achenbach et al.: Nucl. Ins. Meth. A 465, 318 (2001) in elastic electron proton scattering. From the difference 10. D.T. Spayde et al.: Phys. Lett. B 583, 79 (2004) between the measured ALii{ep) and the theoretical pre11. D.B. Leinweber et al.: heplat/0406002 (2004) diction in the framework of the Standard Model, AQ, we 12. F.E. Maas et al.: Phys. Rev. Lett. 93, 022002 (2004) extract a linear combination of the strange electric and 13. F.E. Maas et al.: nuclex/0412030 (2004) magnetic form factors for the 570.4 MeV d a t a at a Q^ 14. R. Lewis et al.: Phys. Rev. D 67, 013003 (2003) of 0.108 (GeV/c)2 of G% 0.106 G^ = 0.071 zb 0.036. 15. T.R. Hemmert et al.: Phys. Rev. C 60, 045501 (1999) For the d a t a at 854.3 MeV corresponding to a Q^ value 16. A. Silva et al.: Eur. Phys. J. A 22, 481 (2004) of 0.230 (GeV/c)2 we extract G% + 0.225 G ^ = 0.039 17. V. Lyubovitskij et al.: Phys. Rev. C 66, 055204 (2002) =b 0.034. Statistical and systematic error of the measured 18. H. Weigel et al.: Phys. Lett. B 353, 20 (1995) asymmetry and the error in the theoretical prediction of 19. H.W. Hammer et al.: Phys. Rev. C 60, 045204 (1999) AQ been added in quadrature. In Fig. 3 the results for
Eur Phys J A (2005) 24, s2, 5154 DOI: 10.1140/epjad/s2005040115
EPJ A direct electronic only
Updated results from the SAMPLE experiment Damon T. Spayde^, for the S A M P L E Collaboration Loomis Laboratory of Physics, University of Illlinois, 1110 West Green Street, Urbana, IL 61801, USA Received: 15 November 2004 / Pubhshed Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The SAMPLE collaboration has recently completed its measurements of the parity violating elastic electron scattering asymmetry from hydrogen and the quasielastic asymmetry from deuterium [1, 2,3,4,5]. Neutral weak form factors of the nucleon, both vector and axial, can be extracted from these data and used to determine the strange quark contribution to electromagnetic form factors. The results of the original measurements at a Q^ of 0.1 (GeV/c)^ [2,3] have recently been reanalyzed [4], incorporating improvements in simulation and new background data, and combined with a new measurement of the quasielastic deuteron asymmetry at a Q^ of 0.038 (GeV/c)^ [5]. Final results from this new analysis and data set will be presented. PACS. 13.60.r Photon and chargedlepton interactions with hadrons  11.30.Er Charge conjugation, parity, time reversal, and other discrete symmetries  13.40.Gp Electromagnetic form factors  14.20.Dh Protons and neutrons
1 Introduction In the simplest, nonrelativistic quark model protons and neutrons are composed of three heavy (~ 300 MeV/c^), valence u p {u) and down (d) quarks. In this model the various static properties of the nucleons such as mass and intrinsic spin are explained quite well in terms of the properties of the valence quarks. However, the full theory t h a t describes the structure of nucleons and folds in the Standard Model and q u a n t u m chromodynamics (QCD) has the valence quarks as very light ( ^ 1 0 MeV/c^) particles. Furthermore, the valence quarks are surrounded by a sea of gluons and quarkantiquark pairs. This quark sea can contain the heavier quarks such as the strange (5) as well as the lighter u and d. A full description of the properties of the nucleon must account for the presence of these quarkant iquark pairs. A variety of experiments have detected nonzero strange quark contributions to static nucleon properties such as spin, mass, and m o m e n t u m (see [6] for a recent review) . Given t h a t other low energy nucleon properties contain nonzero contributions from strange quarks it is natural to wonder how these quarks contribute to the electric and magnetic properties of the proton: the charge radius and the magnetic moment. These properties arise from the proton's vector strange matrix element (p57^5p), a quantity t h a t can be determined by measuring the proton's neutral weak magnetic form factor G'^ [7]. A method for measuring G'^ using parity violating elastic electron scattering was proposed by Beck and McKeown [8,9]. This ^ Present address: Department of Physics, Grinnell College, 1116 8th Avenue, Grinnell, lA 50112 USA
experimental technique has formed the basis for several past and present experiments trying to probe the strange quark contribution to the electric and magnetic properties of the nucleon. This paper discusses results from one of these experiments, S A M P L E , t h a t was performed at the M I T / B a t e s Linear Accelerator Center in the late 1990s/early 2000s. For a recent review of this experiment see [6]. In the scattering of electrons from nucleons via the electromagnetic interaction (exchange of a virtual photon 7) the finite extant of the nucleon is parameterized by the electric and magnetic form factors G^ and G]^. These form factors are functions of the m o m e n t u m transferred by the exchanged photon, or fourmomentum transfer squared Q^. An analogous pair of form factors, C f and Gf^, arise from the same scattering process when a neutral weak Z^ boson is exchanged instead of a 7. T h e electromagnetic and neutral weak form factors for the proton (p) and the neutron (n) can be expressed in terms of contributions from the three lightest quarks u, d, and s [10]. If charge symmetry is assumed (see [11] for a discussion of the validity of this assumption), t h a t is to say t h a t the distribution of u quarks in the proton is the same as t h a t of d quarks in the neutron, then the neutral weak form factors can be written in terms of the electromagnetic form factors and a strange quark contribution (to lowest order)
{lAsin^O
^E,M
w
G E,M'
(1)
T h e neutral weak interaction brings in a third form factor known as the axial form factor C^; to lowest order it is written as follows
G1
TSGA
+
As
(2)
52
D.T. Spayde: Updated results from the SAMPLE experiment
where TS = 1(1) for p{n), GA^Q^^ = 0) = {gA/gv) = 1.2670 =b 0.0035 as measured in /^decay experiments [12], and As is the strange quark contribution to the nucleon spin. Equations 1 and 2 both have higher order electroweak radiative corrections that are not considered here (see [10] for details). It would be extremely difficult to measure G^ ^ with a straight crosssection measurement due to the difference in coupling strengths between EM and neutral weak interactions. However, the parity violating nature of the weak interaction can be exploited to make the measurement feasible. If a beam of longitudinally polarized electrons is scattered from a target of unpolarized electrons and the helicity (product of the electron spin and momentum directions) of the electrons is changed, then the cross section asymmetry Apy is sensitive to the neutral weak interaction at leading order
2 Experiment
The SAMPLE experiment was performed at the MIT/Bates Linear Accelerator Center in Middleton, MA; the target and detector apparatus were installed in the North Hall. The proton asymmetry was measured at a beam energy of 200 MeV during the summer of 1998 [2]. The deuteron asymmetry was measured at two different beam energies: 200 MeV during the summer of 1999 [3] and 125 MeV during the fall of 2001 and spring of 2002 [5]. The original proton and deuteron results at 200 MeV were updated after the completion of the 125 MeV datataking [4,5]. Longitudinally polarized electrons were generated in the polarized source by photoemission from a bulk gallium arsenide (GaAs) crystal; a Ti:Sapphire laser provided the circularly polarized photons used to emit the electrons. The helicity Pockels cell (HFC) was used to reverse the dan  (JGL laser and electron polarizations 600 times a second. A coripy rection Pockels cell (CPC), set to function as an intensity d(jR + dGL adjuster, was also placed in the laser line to reduce the GFQ^ AA iM (3) helicity correlated charge asymmetry measured in the exAnaV2L^Glf^r[Gl,y perimental hall. A position feedback system consisting of a piece of optical glass on a piezoelectric mount was used where GF is the Fermi coupling constant and a is the fine to reduce the helicity correlated position differences meastructure constant. The other terms in 3 are defined as sured in the hall by adjusting the position of the laser spot follows on the cathode in a helicitycorrelated way [13]. The electron spins were precessed away from the longitudinal in a AE = eGUQ')Gl{Q') (4) Wien ffiter prior to injection into the main accelerator to AM = TGfj{Q^)Gl,{Q^) (5) compensate for the 36.5° bend the electrons went through as they entered the experimental hall. AA = {1  4sin2 0w) ^/T{l + T){le^) In the accelerator the electrons were accelerated to the (6) final beam energy of 125 or 200 MeV. At the end of the accelerator was an energy feedback system that reduced the fiuctuations in the electron beam energy [14]. Dipole (7) ^ 4M2, magnets bent the electrons through 36.5° onto the North Hall beam line. Prior to entering the experimental hall (8) the electrons passed through a region containing a M0ller e = l + 2(l + r ) t a n 2  polarimeter that could be used to measure the beam polarization in dedicated, lowcurrent runs. Upon entering the where M^ is the mass of the nucleon. At backward scattering angles the AE term in 3 is neg hall the electron beam passed through a series of toroids ligible relative to the other two terms so a measurement and RF cavities used to measure the beam current and of Apy on the proton Ap is sensitive to G%^ and G\. The position of each beam pulse. A set of Cerenkov detectors isovector piece of G\ is not wellconstrained by theory or (Incite coupled to photomultiplier tubes) were placed symexperiment, necessitating an additional measurement to metrically about the beamline upstream of the target to extract it. This additional measurement comes from mea monitor the halo about the electron beam and another set suring Apy for the deuteron A^. If one assumes the static were placed downstream to measure the luminosity. The experimental apparatus, shown in Fig. 1 consisted approximation is true for the deuteron, i.e. it is essentially a free proton and neutron, then one can write Ad in terms of a highpower cryogenic target encased in a leadlined scattering chamber [15]. The scattering chamber was enof Ar, and Ar, closed in a large, lighttight box that also contained the (JpAp \ (jjiAji detector package. The detector package consisted of an arA,= (9) ray of ten, large ellipsoidal mirrors symmetrically placed about the scattering chamber at backward angles. The where (Jp^n) is the crosssection for elastic e — p{n) scat mirrors focused Cerenkov light from backscattered partering. The deuteron asymmetry is sensitive to G ^ and ticles onto a corresponding array of 8 inch photomultiG\{T = 1) (the isovector piece of G^) as well, therefore, a plier tubes (PMT) encased in lead shielding. Each PMT measurement of Ap and A^ at backward scattering angles was read out by a channel of integrating electronics; the and a fixed Q^ should allow G ^ to be uniquely deter current from each PMT anode was sent to a currenttovoltage converter before being integrated and digitized. mined.
D.T. Spayde: Updated results from the SAMPLE experiment
Fig. 1. This is a schematic of the SAMPLE target and main detector apparatus. The electron beam enters from the left and strikes the target which is shown inside the scattering chamber in a cutaway view
In this mode of operation there was no way t o separate the background events from t h e signal. Dedicated runs of various types were used t o measure t h e background contribution.
exp is p or d depending on whether t h e target is full of hydrogen or deuterium, was combined with t h e theoretical asymmetry in order t o extract G%^ and G\{T = 1). A variety of improvements were made t o t h e d a t a analysis a n d theoretical calculations between t h e publication of t h e original results [2,3] a n d t h e final results [4, 5]. In b o t h t h e hydrogen a n d deuterium 200 MeV d a t a sets an improved Monte Carlo code was used, based on G E A N T [16], t h a t incorporated t h e geometry of t h e target and detectors. This code was used t o calculate t h e electromagnetic radiative corrections Re due t o internal and external bremsstrahlung processes as well as t h e theoretical asymmetry used for t h e form factor extraction [17]. T h e pion dilution factor was a new addition t o t h e analysis; it was not present in t h e original publications. New calculations by Schiavilla were incorporated into t h e theoretical deuteron asymmetry t o account for b o t h quasielastic scattering and threshold breakup. Finally, t h e t r e a t m e n t of a (/) dependent background process in t h e hydrogen measurement was altered (see [6] for details). T h e effect was t o increase t h e experimental asymmetry in each d a t a set by about 1015% a n d reduce t h e theoretical asymmetry by 35%. T h e third set of S A M P L E data, on deuterium at a beam energy of 125 MeV, were taken t o serve as a crosscheck on t h e original 200 MeV deuteron result. T h e original deuteron measurement [3] resulted in a value for G\{T = 1) t h a t deviated significantly from t h e theoretical value [18].
4 Results U p d a t e d results for t h e measured asymmetries on t h e proton a n d deuteron are listed below. T h e results are taken from [5] a n d [4] (all Q^ are in units of (GeV/c)^)
3 Analysis T h e first step in t h e d a t a analysis was t o calculate t h e raw detector asymmetry for beam pulses of positive and negative helicity. Simultaneously, t h e sensitivity of t h e detectors t o changes in t h e various beam parameters were calculated. In t h e next step of t h e analysis this information on detector sensitivity was used t o remove any b e a m parameter dependence from t h e measured asymmetry. After this correction h a d been performed, corrections for various background processes were applied as given by the following expression A^exp
53
Ad{Q^
0.091) = (7.77 =b 0.73 =b 0.72) p p m (12)
Ad{Q^
0.038) = (3.51 =b 0.57 =b 0.58) p p m (13)
where t h e first error is statistical a n d t h e second is systematic. T h e u p d a t e d theoretical asymmetries, taking into account t h e detector geometry, are Ap{Q^ = 0.1) =  5 . 5 6 + 3.37G'M
+ 1.54G^^^=^^
AdiQ'^ = 0.091) =  7 . 0 6 + 0.72Glr + 1.66(7' rj^^lMG'f=^^
Re Psflfeil
Ap{Q'^ = 0.1) = (5.61 =b 0.67 =b 0.88) p p m (11)
 U)
[Ao  (1 
fi)Ac]
(10)
where Re is an electromagnetic radiative correction, PB is t h e measured beam polarization, // is t h e fraction of t h e yield due t o light on t h e P M T s , fe is t h e fraction of t h e light signal due t o Cerenkov events, and /^^ is t h e fraction of the Cerenkov light due t o pion production. T h e quantity AQ is t h e beam parameter corrected asymmetry measured under normal running conditions; Ac is t h e same asymmetry measured with shutters drawn over t h e P M T s t o block out t h e light signal. T h e final quantity A^xp, where
(14)
(15)
AdiQ'^ = 0.038) =  2 . 1 4 + 0.276^^ + 0.76^^^^=^^ (16) from [5] a n d [4]. Figure 2 shows t h e deuteron asymmetry measured at 125 a n d 200 MeV as a function of Q^. T h e result of theoretical calculations for Ad based on [18] are shown as well. T h e good agreement between theory a n d experiment shown in this plot from [5] motivated t h e use of t h e theoretical calculation for G\{T = 1) in t h e final result for G ^ quoted in [4]. Figure 3 is taken from [4] and shows t h e region allowed by t h e measurement of Ap on hydrogen at
D.T. Spayde: Updated results from the SAMPLE experiment
54
200 MeV and the theoretical calculation of ^ ^ ^ ( r = 1). Using the theoretical value of 6 for G\{T = 1) results in G  ^ ( Q 2 = 0.1)
= 0.37 =b 0.20 =b 0.26 =b
0.07
(17)
where the errors are due to statistics, systematics, and the uncertainty on the electroweak radiative corrections [4]. This is lower t h a n , but consistent with, the earlier result quoted in [3].
J
tf (GeV/c) Fig. 2. This is a plot of the deuteron asymmetry Ad as a function of Q^. The filled circles are the data from the 125 and 200 MeV SAMPLE runs (statistical and systematic error bars are shown). The empty circles are the theoretical prediction for Ad using [18] for the multiquark radiative corrections and a value of 0.15 for G%f. The grey bands show the change in Ad as G%[ is changed by d=0.3 n.m. The figure is from [5]
1.0
T \
y

'y
',
5 Conclusion T h e S A M P L E collaboration has performed an important series of measurements of the parity violating electron scattering asymmetries from the proton and the deuteron. These measurements have made it possible to extract, for the first time, the strange quark contribution to the magnetic form factor of the nucleon; a quantity t h a t is directly related to the magnetic moment of the proton and the neutron. This strange quark contribution was found to be slightly positive. T h e measurements of the deuteron asymmetry can also be used to place constraints on the quantities C2u and C2d which represent electronquark axial couplings in the Standard Model [6]. T h e S A M P L E collaboration has also performed the first ever measurement of a single spin asymmetry for scattering of transversely polarized electrons from an unpolarized proton target [19]. These measurements are sensitive to twophoton exchange diagrams m. e — N scattering.
References
9. 10. 11. 12. 13. 14. Fig. 3 . This is a plot of GM versus G^j^{T = 1) and shows the 15. region allowed by the various SAMPLE measurements. The 16. open and crosshatched bands are the regions allowed at the 17. one sigma level by the 200 MeV hydrogen and deuterium mea 18. surements, respectively. The vertical, hatched band is the theoretical calculation of G\{T = 1) based on [18]. The small oval 19. is the final result reported in [4]
B. Mueller et al.: Phys. Rev. Lett. 78, 3824 (1997) D.T. Spayde et al.: Phys. Rev. Lett. 84, 1106 (2000) R. Hasty et al.: Science 290, 2021 (2000) D.T. Spayde et al.: Phys. Lett. B 583, 79 (2004) T. Ito et al.: Phys. Rev. Lett. 92, 102003 (2004) E.J. Beise, M.L. Pitt, D.T. Spayde: Prog. Part. Nucl. Phys. 54, 289 (2005) D.B. Kaplan, A. Manohar: Nucl. Phys. B 310, 527 (1988) R.D. McKeown: Phys. Lett. B 219, 140 (1989) D.H. Beck: Phys. Rev. D 39, 3248 (1989) M.J. Musolf et al.: Phys. Rep. 239, 1 (1994) G.A. Miller: Phys. Rev. C 57, 1492 (1998) K. Hagiwara et al.: Phys. Rev. D 66, 010001 (2002) T. Averett et al.: Nucl. Inst. Meth. A 438, 246 (1999) D.H. Barkhuff et al.: Nucl. Inst. Meth. A 450, 187 (2000) E. Beise et al.: Nucl. Inst. Meth. A 378, 383 (1996) GEANT: CERN Program Library D.T. Spayde: Ph.D. thesis, University of Maryland, (2001) S.L. Zhu, S.J. Puglia, B.R. Holstein, M.J. RamseyMusolf: Phys. Rev. D 62, 033008 (2000) S.P. Wells et al.: Phys. Rev. C 63, 064001 (2001)
Eur Phys J A (2005) 24, s2, 5558 DOI: 10.1140/epjad/s2005040124
EPJ A direct electronic only
The next generation HAPPEX experiments Richard S. Holmes, for the H A P P E X collaboration Syracuse University, Syracuse, NY 13244, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. A new generation of parity violating electron scattering experiments began running in Summer 2004 at the Thomas Jefferson National Accelerator Facility. HAPPEXH will measure the parity violating asymmetry in polarized electron scattering from protons at Q^ — 0.11 (GeV/c)^. This asymmetry is sensitive to a linear combination of the strange electric and magnetic form factors of the proton. HAPPEXHe will measure the parity violating asymmetry in polarized electron scattering from a "^He target at the same value of Q^, using the same spectrometers and similar detectors, accessing the strange electric form factor cleanly. PREX will use parity violating electron scattering to determine the neutron radius of the 2°^Pb nucleus. PACS. 13.60.Fz Elastic and Compton scattering  ll.30.Er Charge conjugation, parity, time reversal, and other discrete symmetries  13.40.Gp Electromagnetic form factors  14.20.Dh Protons and neutrons  24.80.+y Nuclear tests of fundamental interactions and symmetries  21.10.Gv Mass and neutron distributions  25.30.Bf Elastic electron scattering
1 Introduction
2 Experimental goals
T h e question of how sea quarks contribute to nucleon observables has been of great interest since the European Muon Collaboration's observation [1] t h a t valence quarks contribute less t h a n half of the proton spin. Kaplan and Manohar [2] suggested t h a t strange quarks (necessarily sea quarks) could contribute to the vector matrix elements of the nucleon, giving rise to nonzero strange electric and magnetic form factors. The size of these effects can be deduced from measurements of the parityviolating asymmetry in elastic scattering of longitudinally polarized electrons from protons or nuclei [3,4,5]. These asymmetries are functions of the conventional electromagnetic form factors of the nucleons, an axial form factor, and the strange form factors. Asymmetries measured in three such experiments [6,7,8,9] have been reported. Polarized electron scattering from a highZ nucleus can be used to address very different physics. The parity violating asymmetry measured in such an experiment is sensitive to the neutron radius of the target nucleus [10], an observable which has to date not been measured in a clean, model independent way. I report here on a set of secondgeneration parity violating electron scattering experiments which will be performed at the T h o m a s Jefferson National Accelerator Facility (Jlab). They will provide improved sensitivity to the strange form factors at low Q^, and a clean measurement of the neutron radius of the ^^^Pb nucleus.
T h e parity violating asymmetry in scattering of polarized electrons from protons is given by A PV
GF\Qf
X P {\\^^\^Qw) 47raV2 eG^ {G^ \ G%) \ rGj^{Gj^
e(Gj)Vr(Gl?)^
+ G^) (1)
^AA
where G^Xj^\ is the electric (magnetic) Sachs form factor for the proton, G^,j^^ is the same for the neutron, G^^ and G%^ are the electric and magnetic strange form factors, Ow is the weak mixing angle, GF is the Fermi constant, and a is the finestructure constant. A A is a t e r m proportional to G^^, the neutral weak axial form factor. T h e kinematic factors are Q^ = —q'^ > 0, the square of the fourvector m o m e n t u m transfer, r = Q^/AM^ where M is the proton mass, and e = [1 + 2(1 + r ) tan^(6>/2)]~^ where 0 is the scattering angle. Finally, p' = 0.9879 and n' = 1.0029 are parameters arising from electroweak radiative corrections [11]. At forward angles A A is small, and a linear combination of the strange form factors can be calculated from the asymmetry and measured values of the nucleon electromagnetic form factors. One can express a strangeness radius ps and a strange magnetic moment ps in terms of the strange form factors as ps = [dG^^/dT]^^^^ f^s = G ^ ( 0 ) .
R.S. Holmes: The next generation HAPPEX experiments
56
T h e original Hall A P r o t o n Parity Experiment (HAPP E X ) measured parity violating asymmetry in the elastic scattering of polarized electrons from a liquid hydrogen target at 6> = 12.3 degrees, Q^ = 0.5 ( G e V / c ) ^ T h e result [7] was A^^ =  1 4 . 9 2 8 (stat) 6 (syst) parts per million (ppm); from this we obtain G% 0 . 3 9 2 G ^ = 0.014 0.020 0 where the first error is experimental and the second is due to errors in the values of the electromagnetic form factors. T h e strange form factors may be small at this Q^, or there may be an accidental cancellation. It is of interest to measure G^^ and G ^ separately, and particularly at a smaller value of Q^ to better allow extrapolation to zero. These considerations have motivated two new experiments, H A P P E X  H (Experiment E99115) [12] and H A P P E X  H e (Experiment EOO114) [13]. H A P P E X  H , hke H A P P E X , scatters polarized electrons elastically from a hydrogen target. However, with the use of the new septum magnet upgrades to the HRS spectrometers the average laboratory scattering angle is reduced to approximately 6 degrees. T h e Q'^ for this experiment is about 0.1 GeV^ and the theoretical value for the parity violating asymmetry in the absence of strange quark effects is 1.6 ppm. T h e goal is to measure this asymmetry with statistical and systematic errors of 5% and 2.5%, respectively. Obtaining values for the two strange form factors separately requires a measurement at different kinematics or on a different target. A particularly attractive choice is to use a spinless isoscalar target in which case the parity violating asymmetry for elastic scattering of polarized electrons is simply iPV
Ao 2
2sin^6> w
GE
Gl^ + GT
CEBAF
Hall A Proton Parity 5^eriment
Fig. 1. Schematic of HAPPEXH experiment
In Hall A, elastically scattered electrons are focused onto two detectors by two identical magnetic spectrometers. Inelastic events are cleanly rejected by being focused out of the detector acceptance. T h e high rate in the detectors would result in significant dead time issues for a counting experiment. Instead, we integrate the detector signals over each helicity window using custom built integrating ADCs and a d a t a acquisition system triggered at 30 Hz. T h e beam position, energy, and intensity are measured using various detectors in the Hall A beam line and these signals are also integrated over each helicity window. T h e physics asymmetry is given by
(2)
and A^^ is sensitive only to G;. H A P P E X  H e scatters polarized electrons from a high pressure ^He target with the same kinematics as H A P P E X  H : ^ ~ 6 degrees, Q^ ^ 0 . 1 GeV^. W i t h o u t strange quark effects, the predicted asymmetry is 8.4 ppm, and the goal is to measure this asymmet r y to 2.2% statistical and 2 . 1 % systematic. Both H A P P E X  H and H A P P E X  H e began d a t a taking in Summer 2004. T h e parity violating asymmetry in elastic scattering of polarized electrons from a heavy nucleus is sensitive to the ratio Rn/Rp of the neutron and proton radii for t h a t nucleus. T h e lead radius experiment, P R E X (Experiment EOO114) [14], will measure this asymmetry with a ^^^Pb target. Kinematics will be ^ ~ 6 degrees, Q^ ~ 0.01 GeV^, with an expected asymmetry ^4^^ ~ 0.5 ppm. Our goal is to measure this asymmetry with statistical and systematic errors of 3 % and 1.1%, respectively, yielding a 1% measurement of Rn/Rp. P R E X is conditionally approved. A schematic of the H A P P E X  H experiment is shown in Fig. 1. Polarized electrons are produced at the source using circularly polarized light incident on a GaAs crystal. Polarization is set at 30 Hz, with a randomly chosen polarization followed by its complement to make a pair of opposite helicity "windows" each about 30 ms in length.
^phys
^det
AQ + aAE + 2_^
iPi^Xi
(3)
Here Adet = {S^  S^)/{S^ + S^) is the helicity correlated asymmetry in the integrated detector signals S^^^^ for right (left) polarized beam. T h e remaining terms are corrections for systematic effects. AQ and AE are the helicity correlated relative differences in the b e a m intensity and energy, respectively. Axi is the helicity correlated difference in beam parameter Xi, where the four parameters are horizontal and vertical positions and angles at the target, a and Pi are the responses of the detector to changes in beam energy and position or angle x^, respectively. This equation is correct under the assumptions t h a t the detectors and beam monitors are linear and t h a t there are no additional sensitivities to higher order parameters, such as spot size.
3 Beam and instrumentation upgrades In order to meet the challenging goals of these experiments, we must control and understand our systematics at a more precise level t h a n in the past. A number of upgrades and improvements to the polarized source, beam transport, and experimental instrumentation are required.
R.S. Holmes: The next generation HAPPEX experiments Table 1. HAPPEX beam requirements. "Jitter" is RMS width for signals integrated over 30 ms; "Difference" is size of helicity correlated difference averaged over the data set Property
Nominal
Jitter
Difference
Energy Current Position Angle Halo @ 3 mm
3.2 GeV 100 M 0 0 <100 Hz//iA
< < < <
< < < <
80 ppm 1000 ppm 12 /xm 12 /xrad
13 ppb 600 ppb 2 nm 2 nrad
including: new target cells; various improvements at the polarized source; better polarimetry; new Cerenkov detectors; new luminosity monitors; new profile scanners; new b e a m intensity and position monitors; and an improved d a t a acquisition system. T h e correction terms in 3 must be kept small and precisely measured. Potentially the largest correction is AQ, which can easily reach levels of several hundred p p m or larger. A s t a n d a r d technique for reducing this t e r m is to introduce a feedback system, by which the helicity correlated b e a m intensity difference measured in the experimental hall is used to govern modulation of the intensity of the laser light at the source cathode, driving AQ to a value consistent with zero. While it is possible to feed back on beam energy, position, and angle differences, correlations between these parameters and fluctuations in their measurement introduce complications best avoided if possible. Careful control of position and angle of the laser spot on the source cathode can keep these beam parameter differences small; then, by measuring the differences in the hall and calibrating the detector responses to parameter differences by introducing deliberate modulations of the beam during a portion of the d a t a taking, the required small corrections can be made with good precision. T h e revamped laser table setup at the polarized source features two complementary systems for modulating the laser intensity. A P Z T mirror is available for modulating position of the laser beam, if required. As in H A P P E X there is an insertable half wave plate to provide slow helicity reversal, which suppresses certain systematics; there also is a rotatable half wave plate for control of position and intensity systematics. A lens is used to image the b e a m onto the photocathode. T h e cathode provides better t h a n 80% polarization with intensities u p to 100 fiA. Some of the requirements for H A P P E X  H and H A P PEXHe quality beam are shown in Table 1. T h a n k s to hard work by the source group in collaboration with H A P P E X personnel, these requirements seem to be achievable now. P R E X requirements will be even more stringent. For further details on control of source systematics see Cates's talk at this conference. A significant portion of the final systematic error comes from polarimetry. Two polarimeters are available in Hall A, a M0ller polarimeter [15,16] and a Compton
57
polarimeter [17,15]. For the M0ller polarimeter the main uncertainty is the foil polarization; at the H A P P E X  H and H A P P E X  H e energy of ^ 3 GeV, a total relative error of 3 to 3.4% is expected. T h e Compton polarimeter, which was commissioned during the first H A P P E X experiment, has since been upgraded with the addition of an electron recoil detector. In this new configuration total relative errors of 1.4% per 1 hour measurement have been seen with a 4.5 GeV beam, and 2.0% is probably achievable at 3 GeV. Additional improvements to polarimetry will be required for P R E X , where the combination of high b e a m intensity and low energy (850 MeV) means neither of the existing polarimeters can obtain precise measurements under running conditions. To address this the Compton polarimeter will be upgraded with a green laser t h a t should enable measurements at the 1% level in under a day. T h e H A P P E X  H detectors are b r a s s / q u a r t z sandwich Cerenkov total absorption detectors. Elastically scattered electrons are contained in a ^ 1 0 0 cm x 600 cm region of the focal plane in each spectrometer arm. To cover this region with minimal transmission losses, each detector consists of two segments at right angles, oriented to put the Cerenkov peak in line with the quartz layers and aimed at the single photomultiplier t u b e coupled to each segment. For H A P P E X  H e , where the elastic peak region is smaller in size, we disassemble one of these detectors into its two segments and use one segment in each spectrometer arm. Detector design for P R E X , where the elastic peak region will be very small, has not been finalized. Interpretation of our results requires precise knowledge of our effective Q^. For H A P P E X this was done using the s t a n d a r d Hall A drift chamber package at low b e a m intensity. For the new experiments we supplement this m e t h o d with a pair of quartz Cerenkov profile detectors which can be scanned across the elastic peak at high b e a m intensity. This enables us to verify the drift chamber results and to monitor the Q^ stability during the experiment. T h e stripline position monitors used for H A P P E X have a resolution of ^ 1.8 //m. This is adequate for H A P P E X  H and H A P P E X  H e , but the stringent hmits required for P R E X demand better instrumentation. We have installed new cavity based position and intensity monitors which will be commissioned during the H A P P E X  H / H A P P E X  H e running. Target density fluctuations give rise to increased widths and correlations in the detector signals. New luminosity monitors mounted on the downstream beamline at very small scattering angles enable us to monitor such target density fluctuations. They also provide high sensitivity to helicity correlated beam parameters and enable better measurements of electronic noise. T h e monitors have been tested to ^ 2 0 0 p p m resolution per 30 ms helicity window at low b e a m intensity, and performance of ^ 1 0 0 p p m at high intensity is anticipated. A new target cell design is being commissioned for H A P P E X  H and H A P P E X  H e . Tests in early 2004 demonstrated t h a t the existing design suffered from density fluctuations in a high power b e a m which would have
58
R.S. Holmes: The next generation HAPPEX experiments quark effects. These are challenging experiments, but feasible with the improvements in equipment and techniques developed in the past several years. T h e expected impact is shown in Fig. 2. P R E X is even more challenging, but we are confident our experience in H A P P E X  H and H A P P E X  H e will establish our readiness to undertake it. H A P P E X  H and H A P P E X  H e began d a t a taking in Summer 2004, and preliminary results for b o t h have been presented [18]. We expect to complete d a t a taking for these experiments in 2005, with P R E X to be scheduled subsequently.
References
Fig. 2. Expected impact of HAPPEXH (diagonal band) and HAPPEXHe (horizontal band). Points are predictions of various theoretical models. Expected limits on ps and /is are computed assuming the full data set and an extrapolation to Q = 0, and are centered on ps = 0, /J.S 0
been unacceptable for parity experiments. T h e new "racetrack" design features fluid flow transverse to the beam direction to reduce such boiling behavior. For P R E X , a thin lead target able to withstand beam intensity of tens of //A without melting is obviously a challenge. A design in which the foil is sandwiched between thin diamond films in a liquid helium cooled copper frame has been successfully tested. Diamond has high thermal conductivity and scattering from ^^C is well understood, making this an ideal solution for this experiment.
4 Summary and prospects H A P P E X demonstrated the need for an experimental program to provide information on the electric and magnetic strange form factors separately, with high precision, at low Q^. H A P P E X and H A P P E X  H e are designed to significantly limit the space of allowed values for these strange
1. J. Ashman et al.: Phys. Lett. B 206, 364 (1988); NucL Phys. B 328, 527 (1988); Nucl. Phys. B 328, 1 (1989) 2. D.B. Kaplan, A. Manohar: NucL Phys. B 310, 527 (1988) 3. R.D. McKeown: Phys. Lett. B 219, 140 (1989) 4. E.J. Beise, R.D. McKeown: Comments Nucl. Part. Phys. 20, 105 (1991) 5. D.H. Beck: Phys. Rev. D 39, 3248 (1989) 6. K. Aniol et al. [HAPPEX Collaboration]: Phys. Rev. Lett. 82, 1096 (1999); K. Aniol et al., [HAPPEX Collaboration]: Phys. Lett. B 509, 211 (2001) 7. K. Aniol et al. [HAPPEX Collaboration]: Phys. Rev. C 69, 065501 (2004) 8. T.M. Ito et al. [SAMPLE Collaboration]: Phys. Rev. Lett. 92, 102003 (2004), and references therein 9. F.E. Maas et al. [A4 Collaboration]: Phys. Rev. Lett. 93, 022002 (2004) 10. C.J. Horowitz, S.J. Pollock, P.A. Souder, R. Michaels: Phys. Rev. C 63, 025501 (2001) 11. R.M. Barnet et al. [Particle Data Group]: Phys. Rev. D 54, 1 (1996) 12. R. De Leo et al. [HAPPEX Collaboration]. JLab Experiment E99115 Proposal (1999) 13. P.A. Souder et al. [HAPPEX Collaboration]: JLab Experiment EOO114 Proposal (2000) 14. K.A. Aniol et al. [HAPPEX Collaboration]: JLab Experiment EOO003 Proposal (1999); K.A. Aniol et al., [HAPPEX Collaboration]: JLab Experiment E03011 Proposal Update (2002) 15. J. Alcorn et al.: Nucl. lustrum. Meth. A 522, 294 (2004) 16. A.V. Glamazdin et al.: Fizika B 8, 91 (1999) 17. M. Baylac, et al.: Phys. Lett. B 539, 8 (2002) 18. K. Paschke: Baryons 2004 conference (2004)
Eur Phys J A (2005) 24, s2, 5963 DOI: 10.1140/epjad/s2005040133
EPJ A direct electronic only
The GO experiment: Parity violation in eN elastic scattering Philip G. Roos, for the GO CoUaboration^ University of Maryland, College Park, MD 20742, USA Received: 1 December 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The G*^ experiment will measure the parityviolating asymmetries in elastic electronnucleon scattering. The experiment is being performed in Hall C at the Jefferson Laboratory using a polarized electron beam and a dedicated experimental setup. Measurements of the electronproton asymmetries will be made at both forward and backward angles, and electrondeuteron asymmetries at the backward angles. These measurements will cover a momentum transfer range of 0.1  1.0 GeV^/c^. ^From these data the vector neutral weak form factors, G% and C M , and the effective axial current of the nucleon, C^, may be extracted. When combined with the known electromagnetic form factors, one will be able to extract the contributions of i^, d, and s quarks to the proton's charge and magnetization distributions. The first measurements at forward angles for the full momentum transfer range have very recently been successfully completed and preliminary results are presented here. PACS. 1 3.60.Fz Elastic and Compton Scattering  1 3.40.Gp Electromagnetic Form Factors  1 4.20.Dh Protons and Neutrons
1 Introduction A crucial aspect to our knowledge of the structure of the nucleon is the understanding of the role of the quarkantiquark sea in contributing to the various properties of the nucleon  charge, magnetism, spin, etc. To this end, over the past decade several parityviolating electron scattering experiments have been carried out in an effort to identify and quantify the strange quark contribution to the nucleon sea. Experiments for which d a t a have already been pubHshed are the S A M P L E experiment at MITBates [1, 2,3,4], the H A P P E X experiment in Hall A at Jefferson Lab [5], and the PVA4 experiment at Mainz [6,7]. Each of these experiments has been discussed at this meeting, and the contributions are included in these proceedings. In each of these experiments a measurement is made for a single Q^ value. A new parityviolating electron scattering experiment currently being carried out in Hall C at Jefferson Lab is the GO experiment [8]. T h e GO collaboration consists of approximately 100 P h . D . scientists from 20 institutions in the United States, Canada, France, and Armenia. T h e GO experiment will measure parity violation (PV) in the scattering of polarized electrons from nucleons at b o t h forward and backward angles using hydrogen and deuterium targets. T h e ultimate goal of GO is to use the measurements of the parityviolating asymmetries in elastic electron scattering from the nucleon to provide a comprehensive and ^ Jefferson Lab Experiment E00006. A list of collaborators and information about the experiment can be found at http://www.npl.uiuc.edu/exp/GO/
precise m a p of the neutral weak form factors of the proton over the range of m o m e n t u m transfers 0.11.0 (GeV/c)^, T h e interaction between electrons and nucleons involves interference between the dominant electromagnetic interaction (7 exchange) and the parityviolating weak interaction (Z^ exchange). This interference leads to a helicity dependence in the cross section for elastic scattering of polarized electrons from the nucleon, the observable measured by the GO experiment, as well as the other experiments mentioned above. T h e asymmetry has three contributing terms t h a t reflect this interference. Specifically, the parityviolating asymmetry for elastic electron scattering from the proton can be written as a sum of three terms: A 47rQ;A/2
{l4.sin^ew)e'G''^G\]
e{GlY^T{G^^) where r
=
Q2/4M^, e =
( l + 2(1 + r ) tan^ ^^
il) and
s' = ^ ( 1 — £^) r (1 + r ) . W h e n combined with known information on the nucleon electromagnetic form factors and assuming isospin conservation, it is possible to extract explicit contributions of s's pairs to the nucleon form factors. Away from the static limit, this information is characterized by the strange quark form factors, G%^ and G;, which contribute to the nucleon electric and magnetic form factors, defined by L  4 s i n 2 ^ H ^ ) ( l + i ? ^ ) G ^\M, (1 + Rv)G'^^M
 G E,M
(2)
P.G. Roos: The GO experiment: Parity violation in eN elastic scattering
60
T h e strange form factors are of particular interest because they are a direct probe of t h e quark sea contribution to ground state properties of t h e nucleon. A complication arises from t h e third t e r m in 1, t h e axial form factor G\ which has contributions b o t h from t h e firstorder exchange of a Zboson, but also from higher order processes t h a t produce a parity violating electromagnetic, or "anapole" interaction. For example, there can be a Zexchange between two quarks in t h e proton and a photon exchange between t h e electron and proton. Theoretical estimates of these axial radiative corrections have been made for t h e S A M P L E experiment [9,10], but not for higher Q^ values. In order to measure G\ and thus address t h e uncertainty in these corrections, a third measurement is needed. To obtain a third measurement t h e GO experiment will adopt t h e method used by t h e S A M P L E collaboration; specifically, we will carry out measurements of parityviolating quasielastic scattering from deuterium at back angles. T h e different sensitivities to t h e axial form factor between t h e proton and t h e deuteron permit one to extract all three terms G\^^ G%^ and G\. T h e sensitivities of t h e three measurements of t h e GO experiment t o t h e various form factors are indicated for one value of Q^ in t h e following. GO e l a s t i c s c a t t e r i n g p r o g r a m Ap: one measurement for all Q^ ^ detect recoil protons AB' three measurements for three Q^ values ^ detect electrons at 108° Ad'. Quasielastic scattering from deuterium ^ detect electrons at 108°
AF\ AB
Ad
1 ^F XF i^F = \ ^B XB \ £,d Xd ''Pd
at Q 2 = 0.44 (GeV/c)2
AF AB
Ad
7y(ppm) 13.77 25.01 34.00
^(ppm) 51.80 16.10 13.13
X(ppm) 18.63 31.41 7.07
V^(ppm) 1.01 6.96 8.41
2 GO: Forward angle scattering To carry out t h e GO measurement program on a reasonable time scale, it was necessary to meet certain specifications. Because the asymmetries were expected t o be as small as a few ppm, it is necessary to achieve counting statistics of t h e order of 10^^ t o 10^^ counts. This necessitated t h e design of a large acceptance device with high count rate detectors and electronics, and t h e capability of isolating elastic scattering from inelastic processes. In addition an electron b ea m with high intensity, high polarization and small helicitycorrelated b eam properties is highly desirable, if not absolutely essential. These requirements led to t h e design and construction of specialized equipment by the GO collaboration, as detailed next.
Focal Plane ' ^ / Q 2 = 1 GCV/C2
Recoil protons Q2=0.1 GcV/c^ Electron beam Fig. 1. A schematic diagram of the GO magnet and detection scheme for the "forward" angle mode
T h e heart of t h e GO a p p a r a t u s is a superconducting toroidal magnet t h a t focuses scattered particles through collimators onto a focal plane array of scintillators. T h e magnet contains eight superconducting coils splitting t h e magnet and detector system into octants. Each octant, as defined by collimators near t h e cryogenic target, accepted 20° in azimuthal angle. Combined with t h e polar angle acceptance, defined by t h e m o m e n t u m defining collimators, t h e device had a total solid angle of about 0.9 sr. Figure 1 shows a schematic representation of t h e principle of operation for t h e forward angle scattering mode ("forward" mode). In its "forward" mode, t h e a p p a r a t u s detects and counts recoiling protons from forward angle electron scattering (the protons are ejected at angles between approximately 60° and 75°). In this mode a simultaneous measurement for all Q^ values (0.11.0 (GeV/c)^) is made. Note t h a t for a given Q^ t h e magnet focuses protons from any point along t h e target length, at least at t h e center of each octant. After delays in delivery of t h e magnet and t h e necessity for significant modifications, in t h e end t h e magnet operated very reliably at t h e design current of 5000 A. T h e Focal Plane Detectors ( F P D ) for each octant consisted of 16 pairs of scintillation detectors (Bicron BC400) which were shaped and segmented to detect protons corresponding to a limited band of Q^ and to limit t h e ep elastic count rate to less t h a n approximately 750 kHz. These were placed at, or near, t h e focal plane of t h e magnet. F P D 114 accepted relatively narrow bands of Q^, whereas F P D 15 accepted a broad band of Q^ at t h e end of t h e range of acceptance. No elastic scattering events were recorded by F P D 16, and it was used as a monitor of background and magnet current. Pairs of scintillators were used to suppress background from neutral particles. Light from b o t h ends of each scintillator was t r a n s m i t t e d through light guides to photomultiplier tubes which were mounted in a region of very low magnetic field. A picture of one of t h e detector octants is shown in Fig. 2. T h e signals from t h e F P D photomultiplier tubes were sent t o constant fraction discriminators, mean timing was done between t h e two ends of t h e scintillators, and a coincidence was done between t h e front and back scintillator.
P.G. Roos: The GO experiment: Parity violation in eN elastic scattering
61
T0FFPD8 I
100 120 ToF{1/4ns)
Fig. 3. Typical time spectra for the GO experiment
Fig. 2. Forward proton detector octant
T h e resultant signal was sent to custom timeencoding electronics which measured the flight time of the particle using a signal associated with the beam arrival on target. This time measurement is the primary method used to separate elastic protons from inelastic protons and pious. Two different sets of timeencoding electronics were used. Four of the octants used latching time digitizers to extract the flight time with a resolution of 1 ns. This information was recorded by fast scalers, so t h a t rates of several MHz could be handled. T h e other four octants utilized electronics designed and built in France t h a t were based on flash T D C s . This system had a time resolution of 0.25 ns, and could operate u p to about 4 MHz. T h e basic d a t a transfered to the computer consisted of 128 bit timing histograms. T h e time histograms were read out every 33 msec, corresponding to the helicity reversal period. A typical time spectrum is shown in Fig. 3. T h e cryogenic target is 20 cm long and was designed to have high flow rate to minimize target density fluctuations. At the 40 Ilk. current and 3 GeV electron energy for the forward angle mode, the heat load on the target was 250 W . In tests during GO running the observed target density fluctuations were negligible at 40 ^ A , and we believe t h a t we can run at currents u p to 80 yuA for the back angle mode. T h e final requirement for the forward angle GO run is the time structure of the beam. Since we require a measurement of the timeofflight of the particles, we requested a time structure of 32 MHz. Producing a 40 /xA b e a m with this time structure, and hence very high charge density, required extensive development work on the part of the J L a b accelerator group, and I would like to ac
« J C ryoqe ni c jSiS&JPtH W'jjfi'W''^ ' i J  ^ Supply B r H ^ H H a S S R ^ J i*^
Fig. 4. GO experiment installed in Hall C at Jefferson Lab
knowledge their crucial contributions to the GO experiment. After lengthy development time, the accelerator was ultimately able to provide parity quality b e a m with the required time structure. After two commissioning/engineering runs in 2002 and 2003, followed by a production run, the forward angle measurement of the GO experiment has now been completed, and analysis is proceeding. T h e desired statistical uncertainties were achieved, and most subsystems of the apparatus performed at or above their design goals. Figure 4 shows the GO experiment as installed for the forward angle mode. T h e success of this run was due in large part to the heroic efforts of the postdocs and graduate students associated with GO. Two problems t h a t could potentially impact the error in the final results were encountered during the production running. Both were due to small background yields which had an unexpectedly large asymmetry. T h e first was due to leakage current from the other two halls at J L a b .
62
P.G. Roos: The GO experiment: Parity violation in eN elastic scattering
Increasing Q^
10
12 14 16 detector number
Fig. 5. Preliminary GO forward angle data. The data are blinded by an overall multiplicative factor These leakage beams, although small, had the s t a n d a r d 499 MHz time structure and a helicity dependent charge asymmetry not under our control. W i t h some dedicated running under various conditions, we are able to correct for this contribution to the measured asymmetry to a precision of approximately 0.1 ppm, insignificant compared to our statistical error. T h e second problem encountered is due to an unexpected background lying under the elastic peak in the time spectra. This background has a significant asymmetry which changes over the time spectrum. A great deal of effort has gone into a t t e m p t i n g to understand this background, and to devise methods of subtracting it from the elastic scattering. In addition to expected background from inelastic protons and target windows, we believe t h a t the large asymmetry in the background arises from hyperon decay, and a large simulation effort is underway. Overall we believe t h a t we have achieved the goals of the GO proposal for the forward angle measurement with the caveat t h a t our understanding, or lack thereof, of the background may increase the overall error bar somewhat, particularly for the higher Q^ values. Shown in Fig. 5 are preliminary results for the first 12 F P D s for which the background is less severe. These results are blinded by an overall multiplicative factor to reduce possible analysis biases, and should not be interpreted as anything other t h a n an indication of the overall quality and rough magnitude of the asymmetry.
3 GO: Backward angle scattering T h e second phase of the GO experiment is a measurement of the parityviolating asymmetry for backward an
gle elastic scattering. For these measurements the magnetdetector system is rotated by 180°  a process which has already been completed. In the "backward" mode (first run tentatively scheduled to begin in late 2005), there are a number of differences compared to the forward angle mode. Firstly, elastically scattered electrons are detected rather t h a n recoil protons. T h e optics of the magnet are such t h a t the scattered electrons are at an average angle of , corresponding to a single value of Q^ for each beam energy. Therefore, to cover the full range of Q^, measurements will be required at a variety of incident energies. T h e GO collaboration has proposed measurements at three values of Q^: 0.3, 0.5 and 0.8 (GeV/c)^, corresponding to incident electron energies of 0.434, 0.585 and 0.799 GeV, respectively. Another major difference is t h a t the elastically scattered electrons are not focused on specific focal plane detectors ( F P D ) , but rather on each focal plane detector there are b o t h elastic and inelastic events. To separate these, a second scintillator array (labeled the Cryostat Exit Detectors or CEDs) will be mounted near the magnet cryostat exit window for each octant. T h e C E D array consists of 9 arch shaped scintillators, similar in shape to the F P D s but smaller, since they lie closer to the target. By recording all possible combinations of coincidences between the 9 CEDs and 14 F P D s , we will be able to separate the elastic and inelastic events. In the worst case the contamination of the elastic events due to inelastic electrons will be less t h a n a few percent. While elastic and inelastic electrons can be separated this way, the F P D / C E D combination does not provide any discrimination between electrons and negative pions, b o t h of which have velocities close to t h a t of the speed of light. This is not a serious issue for the hydrogen run, since at the GO energies the 2pion production rate is very low, resulting in a contamination of less t h a n a percent even at the highest proposed energy. However, there will be a small 7r~ rate from the target windows. More importantly the TT" rate from the neutron in the deuterium target will be similar to or greater t h a n t h a t of the electrons. For this reason 8 aerogel Cerenkov detectors have been added to the design to separate electrons and pions. T h e C E D scintillators and light guides have been fabricated at T R I U M F and are currently being assembled on the support structure at J L a b . T h e Cerenkov detectors, provided by the Canadian and Grenoble collaborators, are complete and at J L a b awaiting mounting. A cartoon of the backangle setup and its operation are shown in Fig. 6. T h e diagram on the bott o m shows the combinations of CEDs and F P D s and the relative contributions from elastic and inelastic electrons. Rather t h a n use timeofflight to identify the particles of interest, we will now simply count the particles detected in every possible F P D / C E D combination t h a t did not also fire the Cerenkov. Custom electronics modules which record and sort all possible coincidences between the 9 CEDs and 14 F P D s were designed in Grenoble and Louisiana Tech. Prototypes have been fabricated and tested, and production of the final modules is nearing completion with delivery expected by Spring 2005.
P.G. Roos: The GO experiment: Parity violation in eN elastic scattering
Ceienkov
CFD
63
to simultaneously measure the parity violation asymmetry in the NZ\ transition [11]. T h e dominant contribution to the inelastic asymmetry is expected to be from the onebody, axial transition form factor, G^^. This transition form factor has been measured in charged current reactions, but the GO measurement would be the first determination of this form factor in the neutral current sector. Since we can t a g the pions, the collaboration is also considering a simultaneous measurement of the parity violating asymmetry in pion production.
V
Incident electrons
4 Conclusion CEDFPD Coincidences at Q^ = 0.3 (GeV/c)^ 
H
Ir
!
G
I
After many years and a great deal of effort on the part of many GO collaborators, we are seeing the payoff from this work. We have successfully completed the forward angle measurements, as described above, and are embarking on measurements at back angle. W i t h the completion of GO and the other measurements by the H A P P E X and PVA4 collaborations, we will have a new view into the underlying quark makeup of the nucleon. It promises to be an exciting next few years.
I
I [ D " 00 D
Q LU
.. n u cryfl_
Inelastic electrons
O
Elastic electrons I
I
I
FPD Fig. 6. GO setup for backangle eN scattering
W i t h the above differences some distinct benefits have come. Since we no longer wih be taking timeofflight information, we do not require the special 32 MHz time struct u r e needed for the forward angle measurements for which the b e a m current limit was 40 /J.A. We therefore plan to make use of the excellent performance of the cryotarget to run with 80 //A of beam current with the s t a n d a r d J L a b 2 ns beam structure. This will improve our expected statistics by a factor of two, thereby reducing the statistical error, the largest contributor to the uncertainties at the backward angle. T h e inelastically scattered electrons in the GO backward mode are primarily due to excitation of the A resonance. As stated above these events are separated from the elastic events in the space mapped out by different combinations of F P D / C E D pairs. As a result, it will be possible
References T. Ito et al.: Phys. Rev. Lett. 92, 102003 (2004) D.T. Spayde et al.: Phys. Lett. B 583, 79 (2004) D.T. Spayde et al.: Phys. Rev. Lett. 84, 1106 (2000) R. Hasty et al.: Science, 290, 2021 (2000) K. Aniol et al.: Phys. Rev. C 69, 065501 (2004), arXiV:nuclex/0402004. See also ibid, Phys. Lett B 509, 211 (2001); Phys. Rev. Lett. 82, 1096 (1999) F. Maas et al.: Phys. Rev. Lett. 93, 022002 (2004) F. Maas: private communication. Results not yet published but shown at, for example, the PAVI04 workshop in Grenoble, France, June 2004 For many details of the GO experiment see the paper of C. Furget: Proceedings of the International Workshop on Parity Violation and Hadronic Structure, Mainz, Germany, June 2002. eds. F.E. Maas, S. Kox, D. Lhuillier, J. Van de Wiele: (World Scientific 2004) in print 9. M.J. Musolf, B.R. Holstein: Phys. Lett. B 242, 461 (1990) 10. S.L. Zhu, S.J. Puglia, B.R. Holstein, M.J. RamseyMusolf: Phys. Rev. D 62, 033008 (2000) 11. Jefferson Lab Experiment E04101: T. Forest, K. Johnston, N. Simicevic, S. Wells, Spokespersons, 2004
Eur Phys J A (2005) 24, s2, 6565 DOI: 10.1140/epjad/s2005040142
EPJ A direct electronic only
Study of the parity violation in the ^ ( 1 2 3 2 ) region Response of the A4 calorimeter in the region of the ^ ( 1 2 3 2 ) resonance using MC simulations Luigi Capozza^, for the A4 Collaboration Institut fiir Kernphysik, Universitat Mainz, J. J. Becherweg 45, D55099 Mainz, Germany Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. A measurement of the parity violation (PV) asymmetry in electronproton scattering using a polarised electron beam is performed at MAML The experimental apparatus is able to detect electrons, that are scattered exciting the Z\(1232) resonance. In order to study the PV in such process, it is important to understand the energy spectrum in the corresponding region. The aim of this work is a simulation including all relevant processes in order to extract an estimate of the PV asymmetry in the Z\(1232) region. PACS. 12.15.y Electroweak interactions  ll.30.Er Charge conjugation, parity, time reversal, and other discrete  14.20.Gk Baryon resonances with S=0  13.40.f Electromagnetic processes and properties
1 Introduction Up to now, the measurements of the A4 experiments have been used for studying spin observables in the elastic scattering of polarised electrons off unpolarised protons [1]. Nevertheless, the A4 experiment offers also the possibility to study the inelastic electron scattering off the proton. In fact, b o t h elastically and inelastically scattered electrons are detected simultaneously by the lead fluoride calorimeter. In particular, the A4 detector allows the detection of the scattered electrons exciting the Z\(1232) resonance. Therefore, the measurement of spin observables in the electroproduction of the zl(1232) is possible within the A4 experiment. Among these observables the parity violation (PV) asymmetry can be measured by scattering longitudinally polarised electrons on protons. It has been shown [2, 3,4], t h a t the P V asymmetry in the cross section for the electroexcitation of the ZA(1232) resonance can yield an important insight into the proton structure. T h e scattering cross section asymmetry is obtained, in the A4 experiment, counting the number of electrons scattered with each state of helicity. T h e energy of the scattered electrons is measured by the lead fluoride calorimeter and histogrammed, giving an energy spectrum. By analysing this energy spectrum, it is possible to distinguish between electrons scattered elastically or inelastically, e.g. exciting the Z\(1232). For the analysis of the d a t a in the energy region of the Z\(1232) resonance, a detailed study of the energy spectrum is necessary for having a better knowledge of the background. This is in particular necessary in this energy range, where the ratio of the signal to the background is smaller t h a n in the case of the elastic scattering range. This results from the smaller absolute ^ Comprises part of PhD thesis
Experimental spectrum Simulated spectrum, sum of all contributions
I
Elastic ep scattering including radiative corrections
I
Scattering off Al input and output windows
Inelastic ep scattering
80
100 120 140
160 180 200
ADC channel
Fig. 1. Comparison of simulated and experimental spectra value of the signal cross section and the larger background. W i t h i n this work a detailed study of the lead fluoride detector response to the scattered electrons has been carried out by means of Monte Carlo simulations. In addition, an event generator has been implemented, which includes the elastic ep scattering considering radiative corrections, the inelastic ep scattering and the electron scattering off the aluminium input and output windows. These simulations can reproduce quite well the experimental energy spectrum in the range between the pion production threshold and the maximum of the Z\(1232) peak, as shown in Fig. 1. For this energy region, in principle, the P V asymmetry could be already extracted and compared with the theoretical predictions. More effort has to be provided in order to extend our understanding of the spectrum to include the whole Z\(1232) peak.
References 1. 2. 3. 4.
F.E. Maas et al.: Phys. Rev. Lett. 93, 022002 (2004) L.M. Nath et al.: Phys. Rev. D 25, 9 (1982) 23002305 H.W. Hammer, D. Drechsel: Z. Phys. A 353, 321331 (1995) N.C. Mukhopadhyay et al.: Nucl. Phys. A 633, 481518 (1998)
Eur Phys J A (2005) 24, s2, 6770 DOI: 10.1140/epjad/s200504015l
EPJ A direct electronic only
Don't forget to measure As Stephen F . P a t e Physics Department, New Mexico State University, Las Cruces NM 88003, USA Received: 15 October 2004 / Pubhshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. This talk explores our lack of knowledge of the strange quark contribution to the nucleon spin. As. Data on As from inclusive and semiinclusive polarized deepinelastic scattering will be reviewed, followed by a discussion of how the ongoing program of parityviolating elastic electronnucleon scattering experiments, that seek out the strange electromagnetic form factors of the nucleon, need to have an estimate for the strange axial form factor to carry out that program, and how the value of As extracted from the DIS experiments has filled that role. It is shown that elastic up, Dp, and parityviolating ep data can be combined to extract the strange electric, magnetic and axial form factors simultaneously. A proposed experiment that could address this important issue if briefly previewed. PACS. 14.20.Dh Protons and neutrons  13.40.Gp Electromagnetic form factors of elementary particles
1 Introduction In t h e current experimental program of nucleon structure studies, we find two broad areas of experimentation. O n the one hand, elastic scattering of electrons from nucleons is used t o measure t h e electromagnetic a n d axial form factors of t h e nucleon, over a range of m o m e n t u m transfer of 0.1 < Q^ < 10 (GeV/c)^. These experiments have taken place at a variety of laboratories over t h e years, with t h e current program focused at MITBates, J L a b , a n d Mainz. One of t h e highlights of t h e current program is t h e emphasis on nailing down t h e strange quark contributions t o the electromagnetic form factors, through t h e exploitation of t h e interference between photon a n d Zboson exchange processes. O n t h e other hand, deepinelastic scattering of muons a n d electrons from nucleon and nuclear targets, historically responsible for t h e discovery of t h e partonic structure of matter, continues t o play a role in t h e exploration of t h e distribution of quarks and gluons in nucleons. One of t h e highlights here is t h e focus, over t h e last 15 years, on t h e spin structure of t h e nucleon. T h e deepinelastic exploration of nucleon spin takes place now at b o t h leptonic a n d hadronic facilities, t h e spin program at RHIC being t h e most notable example of an hadronic facility taking on this physics topic. Q C D provides a simple framework in which these two experimental programs are joined together. T h e asymmetries observed in t h e polarized deepinelastic scattering experiments arise from t h e antisymmetric part of t h e virtual Compton amplitude, which contains at its heart t h e nucleon axial current, ^7^75g. In t h e quarkparton model, inclusive scattering of leptons from nucleon targets measures t h e nucleon structure function F i ,
^i(^) = jEe95(^)
where Cq a n d q(x) are respectively t h e charge and parton distribution function for quarks of flavor q. Inclusive scattering of polarized leptons from polarized nucleon targets nucleon structure function measures t h e spindependent
q
where now Aq{x) is a polarized parton distribution function. In Q C D , these distribution functions take on a scale dependence: Aq(x^Q'^). At t h e same time, t h e axial form factors G'^(Q^) measured in elastic scattering are themselves matrix elements of t h e axial current, N {P'\QII^15Q\P)N
=
u{p'hi^7^GA(Q'^Mp)
where t h e matrix element has been taken between two nucleon states of momenta p and y , a n d Q'^ = —{p^ — PY . T h e diagonal matrix elements of t h e axial current are called t h e axial charges, iv(p^7M75^b)Ar =
"^Ms^Aq
where M a n d s^ are respectively t h e mass and spin vector of t h e nucleon. T h e quantities Aq are called "axial charges" because they are t h e value of t h e axial form factors at Q^ = 0; t h a t is t o say, for example, G\{Q'^ = 0) = As. T h e connection between t h e two sets of observables lies in a wellknown Q C D sum rule for t h e axial current, namely t h a t t h e value of t h e axial form factor at Q^ = 0 is equal t o t h e integral over t h e polarized parton distribution function measured at Q^ = 00. For example,
Z\S = G ^ ( Q 2 = 0 )
/"
As{x, Q^ =
oo)dx.
S.F. Pate: Don't forget to measure As
68
In this way, the axial charges Aq provide the hnk between the lowenergy elastic scattering measurements of axial form factors and the highenergy deepinelastic measurements of polarized part on distribution functions. Of course, there are practical difficulties in the full exploration of this sum rule. No scattering experiment can reach Q^ = 0 or Q^ = oo, and no deepinelastic experiment can ever reach x = 0. However, the consequences of these difficulties are more severe in some cases t h a n in others. Our inability to reach Q'^ = oo in the deepinelastic program means t h a t Q C D corrections enter into the sum rule written above, and there is much theoretical experience in calculating these corrections. While the lowenergy elastic experiments cannot reach Q^ = 0, there is not expected any divergent behavior of the form factors near Q^ = 0 and so the idea of extrapolating to Q^ = 0 from measurements at low, nonzero Q^ is not met with alarm. On the other hand, the limitations imposed by our inability to reach a:: = 0 in the deepinelastic experiments are more problematic. T h e unpolarized part on distribution functions q{x) are all known to increase rapidly as X —^ 0 and there is no calculation of the expected behavior near x = 0 to rely upon for an extrapolation from measurements made at a: 7^ 0. Similar comments apply to the polarized parton distributions Aq{x). Unpolarized measurements of the p a r t o n distributions at H E R A have reached very low values of x, nearing x = 3x 10~^, but the corresponding measurements of the polarized distribution functions, from d a t a at SLAG, CERN, and DESY, only reach x = 3 x 10~^. Therefore, measurements of the axial charges place important constraints on the behavior of the distributions Aq{x) in the unmeasureable low2: region.
2 As from inclusive leptonic deepinelastic scattering As mentioned earlier, the doublespin asymmetries in polarized inclusive leptonic deepinelastic scattering measure the spindependent nucleon structure function gi: 9i[x
) = 5Ee'^Aq{x).
In leading order Q C D , these functions take on a scale dependence:
In nexttoleading order (NLO) Q C D , there are significant radiative corrections and the relation between gi and the Aq becomes more complex. In the discussion here, we will limit our attention to the leadingorder Q C D analysis because the NLO version of the analysis does not change the result (nor the uncertainty) for As very much, and the problems to be pointed out exist at all orders, because they are problems coming from the d a t a itself. We will use the analysis from the SMC Collaboration [2] as a model. They measured gi{x, Q^) over a wide
kinematic range, 0.003 < x < 0.70 and 1.3 < Q^ < 58.0. This coverage is not a rectangle, i. e. there is a correlation between x and Q^ in the acceptance of the experiment, and so for a reasonable analysis it is necessary to use Q C D to evolve all the d a t a to a single value of Q^, in this case Q^ = 10 GeV^. In the process of performing this evolution, a fit function for gi is produced. Then, to integrate the distribution gi over 0 < x < 1, it is necessary to extrapolate to X = 1 and X = 0. T h e extrapolation to a; = 1 makes use of the fact t h a t gi, being a difference of two quark distributions, must go to 0 as x ^ 0. This requirement is satisfied in this analysis by assuming the measured experimental asymmetry to be constant for x > 0.7. T h e extrapolation to a: = 0, on the other hand, is not straightforward, as the expected behavior of gi{x) for x —)> 0 is unknown. In this analysis, two methods were used. In one, the Q C D evolution fit was simply extrapolated to x = 0. In another, called the "Regge extrapolation," the value of gi was assumed to be constant for x < 0.003. T h e two values of the integral of gi from these two extrapolations are
A
/ gi{x)dx Jo
= 0.142 zb 0.017
"Regge"
= 0.130
Q C D fit.
7
This integral is related to the axial charges: A =
/ gi{x)dx= Jo
S^el "^ n
/ Jo
Aq{x)dx
.Au^Ad^As Now, assuming t h a t S U ( 3 ) / is a valid symmetry of the baryon octet, and using hyperon f3 decay data, then two other relations between the three axial charges are determined:
AuAd
= gA = F\D
and
Au\Ad2As
=
3FD
where gA = 1.2601 =b 0.0025 and F/D = 0.575 =b 0.016 (in 1997). Now one may solve for the axial charges, and the results are shown in Table 1. It is well to note t h a t , of course,
Table 1. Results for the axial charges from the SMC analysis [2] of their inclusive DIS data
Au Ad As
"Regge" 0.84 6 0.42 0.06 0.08 6
QCD fit 0.80 6 0.46 0.06 0.12 6
the error bars quoted here do not include any estimate of the theoretical uncertainty underlying the assumption of SU(3)/ symmetry. They do include an estimate of the uncertainty due to the extrapolations, but of course t h a t is only an estimate because the actual behavior of gi is
S.F. Pate: Don't forget to measure As
69
unknown in the x —^ 0 region. The only conclusion to be 4 Parityviolating eN elastic scattering drawn for As from this analysis is that it may be negative, One of the highlights of the current low and mediumwith a value anywhere in the range 0 to —0.2. energy electron scattering program is the measurement of the strange vector form factors of the nucleon via parityviolating eN scattering. These measurements are sensitive 3 As{x) from semiinclusive leptonic as well to the nonstrange part of the axial form factor, deepinelastic scattering but rather insensitive to the strange axial form factor due to the relative sizes of kinematic factors multiplying the In semiinclusive deepinelastic scattering experiments, various form factors that contribute to the asymmetry. a leading hadron is observed in coincidence with the To be specific, the parityviolating asymmetry observed scattered lepton, allowing a statistical identification of in these experiments, when the target is a proton, can be the struck quark, and thus a measurement of the x expressed as [6] dependence of the individual Aq{x) distributions. (Inclusive scattering only measures the total structure function A.p — gi(x).) The HERMES Experiment [3] at DESY was es47raA/2 pecially designed to make this measurement. HERMES TG^Gf^j {l'ism'^ew)e'GliG measured doublespin asymmetries in the production of charged hadrons in polarized deepinelastic scattering of e{Gir + r{Gl,r positrons from polarized targets; specifically, the asymmetry in the production of charged pions on targets of where G^^^^N are the traditional electric (magnetic) form hydrogen and deuterium, and of charged kaons in scat factors of the proton and G^ w^^ are their weak analogs, tering from deuterium. There is no assumption of SU(3)/ r = Q^/4M^, Mp is the mass of the proton, e = [1 + symmetry in their analysis. They extract the following 2(1 + r) tan^(^/2)]~^, 6 is the electron scattering angle, quark polarization distributions, over the range 0.023 < and e' = ^ r ( l + r ) ( l — e^). Lastly, G\ is the effective x < 0 . 3 0 [4]: axial form factor seen in electron scattering: Au,
^
Ad.
^
Au,
^
Ad.
,
As . ^
V^") ^(") ^^") T^") V^"^ where ^{x) is defined to be the sum of ^{x) and ^{x). Within the measured uncertainties, and within the measured xregion, the valence quarks {u and d) are polarized and the sea quarks {u^ J, and s) are unpolarized. The integral value of the measured polarized strange quark distribution is ''As' ^ /
As{x)dx = +0.03=b0.03(stat)=b0.01(syst).
[Note this would only be the true As if the integral was over the range 0 < x < 1.] Given the fact that the inclusive analysis described in the previous section produced a negative value of Z\s, it is natural to ask "where did the negative As go?" If the analyses shown of the inclusive and semiinclusive data are both correct, then all the negative contribution to the value of As must come from the unmeasured xregion, that is from x < 0.023. That would imply an average value of As{x) of approximately —5 in the range x < 0.023, which is not impossible, as s{x) is of order 20300 in the range 0.01 < x < 0.001 [5]. Some very interesting physics indeed would be revealed, if the "turn on" of the strange quark polarization in the lowx region was this dramatic. Of course, there are other explanations. The invocation of SU(3)/ symmetry in the analysis of the inclusive data is known to be problematic. And the extrapolations to a:: = 0 in those analyses do not have firm theoretical support. It is clear that a direct measurement of Z\s would serve to clarify these issues.
Here, G^^ is the nonstrange (CC) axial form factor, G\ is the strange axial form factor, and the terms Rj^~ ' represent electroweak radiative corrections [6,7,8,9]. The presence of these radiative corrections clouds the interpretation of the axial term extracted from these experiments. To solve this problem, the SAMPLE [10] experiment measured also the same asymmetry on a deuterium target, in which case the relative kinematic factors of the nonstrange (T = 1) and strange (T = 0) parts of the axial form factor are changed, allowing a separation of the two. However, one may show that this does not help in identifying the value of GJ^, because the relative size of the kinematic factors for G%^ and G\ remain the same for either target: dG'M dG'
{lAsin^Ow)
for SAMPLE.
Therefore, parityviolating eN scattering experiments can only establish a relationship between the strange magnetic and axial form factors, they cannot measure them separately.
5 Combining lyN and parityviolating eAT elastic data Recently it has been demonstrated [11] that the best (perhaps only) way to obtain all three strange quark form factors (electric, magnetic and axial) is through a combination of low energy vN and parity violating eN elastic
S.F. Pate: Don't forget to measure As
70
Table 2. Two solutions for the strange form factors at Q^ = 0.5 GeV^ produced from the E734 and HAPPEX data. (Table from [11])
G% G%
Solution 1 0.02 9 0.00 1 0.09 5
Solution 2 0.37 4 0.87 1 0.28 0
scattering. This prescription is briefly summarized here. Using the known values for the electric, magnetic and nonstrange (CC) axial form factors of the proton and neutron, one may take the difference of the vp and vp elastic cross sections and show it to be a function only of the strange magnetic and axial form factors, G\^ and G\. At the same time, the sum of the i/p and vp elastic cross sections can be shown to be a function only of the strange electric and magnetic form factors, G^^ and G%^. Measurements of forwardangle parityviolating ep elastic scattering are largely functions only of G% and G\^ as well. Therefore, combining these three kinds of d a t a can determine all three strange form factors. At the present time, there is only sufficient d a t a at Q^ = 0.5 GeV^ to make such a determination. In [11] it is shown t h a t by combining the E734 [12] results with the H A P P E X [13] d a t a from J L a b , there are two possible solutions at Q^ = 0.5 GeV^, as summarized in Table 2. In the long run, additional experimentation (already on the schedule at JLab) will select one of these solution sets, but there are already several good reasons to favor Solution 1 over Solution 2. T h e value of the strange electric form factor G% must approach zero as Q^ ^ 0, and Solution 1 is consistent with t h a t requirement. T h e value of G\ in Solution 1 is consistent with the estimated value of G\{Q'^ = 0) = As ^ —0.1 from deepinelastic data, whereas t h a t found in Solution 2 is much larger and of a different sign. T h e value of G\^ in Solution 1 is consistent with t h a t measured by S A M P L E [10] at Q^ = 0.1 (GeV/c)2, whereas the value in Solution 2 is much larger in magnitude. Finally, the value of G\^ in Solution 1 is consistent with the value of G ^ ( Q 2 = 0) =  0 . 0 5 1 T predicted recently from lattice Q C D [14]. It seems nearly certain t h a t Solution 1 is the true physical solution. Future experimentation will in any event select the correct solution set. Additional d a t a from the G^ Experiment [15], recently collected and in the process of analysis, will allow the extraction of the three strange form factors when combined with the E734 data. However, it is unlikely t h a t knowledge of the strange axial form factor over the range 0.5 < Q^ < 1.0 GeV^ will prove sufficient for the extrapolation to Q^ = 0 needed for a determination of As. New neutrino d a t a are needed at lower Q^ to permit a good determination of As.
6 A future experiment to measure the strange axial charge Even if the program I have described determines the strange axial form factor down to Q^ = 0.45 GeV^ suc
cessfully, it almost certainly will not determine the Q^dependence sufficiently for an extrapolation down to Q^ = 0. Also, questions remain about the normalization of the E734 data. Most of their target protons were inside of carbon nuclei, and there was not much known about nuclear transparency in the mid1980's. T h e E734 collaboration did make a correction for transparency effects, but this issue needs to be revisited if we continue to use the E734 data. A new experiment [16] has been proposed to measure elastic and quasielastic neutrinonucleon scattering to sufficiently low Q^ to determine the strange axial charge, As. T h e FINeSSE Collaboration proposes to measure the NC to CC neutrino scattering ratio R NC/CC
a {up —^ lyp)
and from it extract the strange axial form factor down to Q^ = 0.2 GeV^. T h e numerator in this ratio is sensitive to the full axial form factor, —Gj^ \ G ^ , while the denominator is sensitive to only Gj^p. T h e processes in the numerator and denominator have unique charged particle final state signatures. This ratio is largely insensitive to uncertainties in neutrino flux, detector efficiency and nuclear target effects. A 6% measurement of RNC/CC down to Q = 0.2 GeV provides a 4 measurement of As. Acknowledgements. This work was supported by the US Department of Energy.
References 1. L.A. Ahrens et al.: Phys. Rev. D 34, 75 (1986) 2. B. Adeva et al.: Phys. Lett. B 412, 414 (1997) 3. K. Ackerstaff et al.: Nucl. Instrum. Meth. A 417, 230 (1998) 4. A. Airapetian et al.: Phys. Rev. Lett. 92, 012005 (2004) 5. J. Pumplin et al.: JHEP 0207, 012 (2002) 6. M.J. Musolf et al.: Phys. Rep. 239, 1 (1994) 7. M.J. Musolf, B.R. Holstein: Phys. Lett. B 242, 461 (1990) 8. M.J. Musolf, T.W. Donnelly: Nucl. Phys. A 546, 509 (1992) 9. S.L. Zhu et al.: Phys. Rev. D 66, 034021 (2002) 10. D.T. Spayde et al.: Phys. Lett. B 583, 79 (2004); T.M. Ito et al.: Phys. Rev. Lett. 92, 102003 (2004) 11. S.F. Pate: Phys. Rev. Lett. 92, 082002 (2004) 12. L.A. Ahrens et al.: Phys. Rev. D 35, 785 (1987) 13. K.A. Aniol et al.: Phys. Lett. B 509, 211 (2001) 14. D.B. Leinweber et al.: submitted to Phys. Rev. Lett. [arXiv:heplat/0406002] 15. G^ Experiment, http://www.npl.uiuc.edu/exp/GO/, Doug Beck, spokesperson 16. FINeSSE Proposal, B. Fleming, R. Tayloe et al.: [arXiv:hepex/0402007]
Ill Weak form factors of the nucleon II12 Theory
Eur Phys J A (2005) 24, s2, 7378 DOI: 10.1140/epjad/s2005040160
EPJ A direct electronic only
Getting to grips with hadrons Damir Becirevic Laboratoire de Physique Theorique (Unite mixte de Recherche du CNRS  UMR8627), Universite Paris Sud, F91405 Orsay Cedex, France Received: 7 December 2004 / Pubhshed OnHne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. A short discussion concerning the lattice QCD approach to physics of hadrons is made to nonspeciahsts. A special attention is given to topics that are of particular interest to the nuclear physics community. PACS. 11.15.Ha Lattice Gauge Theory  12.39.Fe Chiral Lagrangians
1 Why lattice QCD? One of the most important questions t h a t still remains to be answered is to explain how hadrons arise from the Q C D lagrangian >CQCD
q{l^{d^+gAie)+m,]q.
(1)
T h e dynamics t h a t governs the confinement of quarks and gluons into hadrons is of notoriously nonperturbative nat u r e for which an analytic t r e a t m e n t is still missing. Although various quark models help understanding quite a number of phenomena of hadronic interactions, it should be stressed t h a t a covariant quark model t h a t solves simultaneously confinement and spontaneous chiral symmetry breaking has never been constructed. Other t h a n quark models, much effort has been put in building the effective theories of Q C D , valid for specific ranges of low energy scale. Those theories are built upon some symmetry property of the Q C D lagrangian in some specific limit. T h e most prominent example is the chiral (leftHright) symmetry, SU{Nf)L 0 SU{Nf)ji, t h a t is manifest when the quarks are massless. T h a t symmetry is spontaneously broken down to SU{Nf)y, resulting in the appearance of N'i — 1 Goldstone bosons ('pions'). Chiral perturbation theory ( C h P T ) provides us with an effective description of Q C D t h a t incorporates these features and, in addition, allows one to account for the explicit chiral symmetry breaking corrections, namely those generated by the nonzero quark mass terms in the Q C D lagrangian. T h e computation of such corrections, unfortunately, generates a bunch of low energy constants t h a t are supposed to be obtained from the matching procedure of appropriately chosen amplitudes computed b o t h in C h P T and in Q C D , at some energy scale at which C h P T can be trusted and at which direct Q C D computations can be
made. This is where lattice Q C D is expected to provide information to the Q C D piece in this matching. In the above discussion Nf stands for the number of light quark flavors. Today we are confident t h a t C h P T provides an adequate framework to describe the dynamics of strangeless hadrons {Nf = 2), whereas the situation with the strange quark {rris) is still unclear [1]. This is not only because rris is about 25 times larger t h a n rriq = {rriu + md)/2mu [2], but also because it is not much smaller t h a n the hadronic 3361^^ MeV [3]. W h a t do we Q C D scale, ^ ^ ^ ^ ^ ^ = ' ^ know about TTI^? This is one of the highlights of the lattice Q C D achievements over the past decade which is why I decided to briefly discuss it here. T h a t discussion will also allow me to introduce the methodology but also the challenges of lattice Q C D .
1.1 Lattice QCD and the strange quark mass T h e numerical solution to the problem in hands, namely to compute the hadronic spectra numerically from the Q C D lagrangian only (1), does exist. T h e crucial first step in t h a t direction is to make the analytic continuation to the euclidean metric {XQ^ ^ ^^o^), in which the Q C D generating functional reads
Z[A,q,q] = JVAVqVqexp{S[A,q,q]}
.
(2)
In euclidean space the Q C D action is real and bounded from below. Discretization of the euclidean space and time, L = Nsa and T = Nta, allows for an equivalence between the generating functional and the partition function, so t h a t the Monte Carlo methods can be employed. SU{3) gauge fields are placed on the links of the lattice whereas the quark degrees of freedom are sitting on the sites. A particularly important feature, while discretizing the Q C D action, is t h a t the gauge invariance is preserved at every stage of calculation. T h e price to
D. Becirevic: Getting to grips with hadrons
74
pay is t h a t the lattice spacing a ^ 0 breaks the Lorentz invariance, which is however recovered once we take the continuum hmit, a ^ 0 (i.e. after we send the UV cutoff to infinity). Finally, after the continuum limit has been taken appropriately, we should worry about the finite volume effects and work out the limit L^T ^ oo (i.e. send the IR regulator to zero). This is a very challenging task for numerical simulations, and it requires a lot of clever ideas and a huge computing power. W h a t is important to retain is t h a t  i n principle the Q C D simulations on the lattice offer a first principle approach to the physics of hadrons. In other words the solution to nonperturbative Q C D is provided without introducing any additional parameter except for those t h a t appear in the Q C D lagrangian, namely the quark masses and the SU{3)c gauge coupling. In practice, however, various approximations are often necessary in order to make the calculation feasible on the present day computing resources. Importantly though, all those approximations are controllable and, for the most part, we can get rid of t h e m by increasing the computing power. T h e most infamous (least controllable) is the socalled quenched approximation. It consists in neglecting the dynamical quark loops while producing the gauge field configuration. This is certainly a serious drawback of the most of currently available lattice results, but it nevertheless make a good case for developing the methodology for the computation of various physical quantities on the lattice. One way to compute the quark mass on the lattice is via the axial Ward identity ^, d^Ai^{x) = 2mqP^{x). One computes the following two correlation functions: (^9^^(x)7^75^(x) X
'
0(0)) '
and
Table 1. Strange quark mass obtained from the quenched QCD simulations on the lattice. Results by various collaborations [6] refer to the continuum limit (a —) 0) collaboration JLQCD Alpha & UKQCD QCDSF CPPACS SPQcdR
^(2 GeV) 106(7) MeV 97(4) MeV 105(4) MeV
Ultl MeV 106(2)(8) MeV
by various lattice collaborations have been obtained by means of high statistics simulations, by implementing the nonperturbative renormalization on fine grained lattices, so t h a t the continuum limit could be taken. Finite volume effects have also been examined and shown to be tiny, i.e., at the level much smaller t h a n the errors they quote. T h e results, listed in Table 1, are obtained in the quenched approximation. I m p o r t a n t qualitative outcome from the lattice studies is t h a t the quark masses are indeed small and t h a t the light hadron masses are mostly due to Q C D interaction rather t h a n to their valence quark content. Finally notice t h a t the first lattice studies in which the effects of dynamical quarks are included show t h a t the strange quark mass gets even smaller [7], but we are not yet at the stage of providing the precision unquenched computation of m..
( ^ ^ ( x ) 7 5 g ( x ) 0(0)) , X
'
'
where O is a bilinear quark operator with q u a n t u m numbers J^ = 0~, and after having properly renormalized the axial current and the pseudoscalar density, the ratio of these two correlation functions gives the quark mass. Various ways to nonperturbatively renormalize the composite quark operators on the lattice have been developed (see [4]) and they are implemented in most of the present day quark mass calculations. Besides the ratio of the above correlation functions, from the exponential dependence of the second correlator one can read off the corresponding pseudoscalar meson mass. At this point it should be stressed t h a t the lattice results are consistent with the GellMannOakesRenner (GMOR) formula, rn^^ = 2Bomq. Surprisingly, however, although the G M O R formula is expected to be valid for very small quark masses (it receives the m^corrections and higher), the lattice Q C D results (with Wilson fermions) display a rather impressive consistency with the leading G M O R formula while working with heavy pions ( m p ^ > 500 MeV). T h e strategy to reach the physical quark mass is to t u n e the quark mass in the Q C D action in such a way t h a t the corresponding pseudoscalar meson mass coincides with the physical kaon mass. T h e resulting strange quark mass ^ For alternative strategies and lattice actions to compute the strange quark mass, please see [5].
2 Hadron spectrum Lattice Q C D is particularly well suited to study the spect r a of hadrons. In the previous section we already mentioned t h a t the pseudoscalar meson masses were necessary to identify the strange quark mass. One can also study the correlation functions with the interpolating bilinear quark operators carrying other q u a n t u m numbers and thus extract the vector, axialvector, tensor and even scalar mesons (for the last the valence quarks should be nondegenerate in order to have the correlator with a discernible signal).
2.1 Glueballs In spite of the quenched approximation, some long standing problems can still be tackled. One such a problem is the existence of the mass gap in the pure YangMills theory. This problem is stated as one of Seven Millenium M a t h Prize Problems [9], to which an analytical solution is missing. On the other hand, many lattice analyses performed so far show t h a t the glueball states indeed exist. Nowadays even the spectrum of such states has been established. This is a very important prediction of lattice Q C D . T h e spectrum shown in Fig. 1, is given in multiples of ro, a quantity t h a t is defined from the condition r dV{r)/dr\^^^^ = 1.65 [10], where V{r) is the potential
D. Becirevic: Getting to grips with hadrons
12
Kinput
^ 10 [
75
^3
n
ON,=0
^ 4
3*' 1.5
8 ["
2.
'^..
O
>
io"*^^
E = 6 ^2
O
>
i 3
'^7*
o
C5
^ 2 0"
^ 1.0 ^ 1

Fig. 2. The spectrum of baryons produced by JLQCD both in quenched (Nf = 0) and in unquenched (Nf = 2) QCD [11]. Physical (experimentally established) masses [12] are marked by the horizontal lines
Fig. 1. The spectrum of glueball states as estabhshed from the extensive quenched lattice QCD simulations in [8]. The widths of the lines reflect the error bars of lattice results
between two infinitely heavy quarks, which can be (and has been) accurately studied on the lattice. To convert to physical units, a commonly assumed value is ro = 0.5 fm (orro = 2.5GeV"^). 2.2 Baryons
As we already mentioned, from the exponential falloff of various correlation functions (made with various interpolating operators consisting of quark and gluon fields), one can extract the hadron masses with quantum numbers carried by the considered composite operator. The operators used to extract the proton mass (and its coupling to these interpolating operator) are
J{X) = £«'"= [ul{x)Cdb{x)]
A
N
PC
Jix) = e""" [ul{x)Cj5db{x)]
o
u,{x), j5Uc{x),
(3)
where C stands for the charge conjugation operator. Neutron mass is simply obtained by replacing one u quark by c/, whereas the S state arise after replacing u and d by two s quarks, and so on. The spectrum of lowest baryon states computed on the lattice is shown in Fig. 2. Strange quark mass is fixed from the physical kaon mass, as explained in the previous section. The most striking feature of that plot is that the baryon spectrum, as deduced from the quenched simulations is essentially unchanged after unquenching the QCD vacuum fluctuation by Nf = 2 dynamical quarks. This probably indicates that the most significant effect of quenching has been absorbed in the conversion of results from the lattice to physical units. Second important feature is the nucleon mass that in both cases is larger than that of the physical nucleon. To discuss the reasons for that effect we should
stress that the nucleon mass is not obtained directly on the lattice but rather after a long extrapolation. This is so because the lattice simulations are performed with the light quarks rriq > m^^^^ /2, with ruq rud, while m,, the physical limit is mq/rris = 0.04. Since the sector of light quark masses over which one has to extrapolate is expected to be highly sensitive to the effects of spontaneous chiral symmetry breaking, it is not enough to extrapolate the linear (or quadratic) quark mass dependence observed with the directly accessible quarks (i.e. in the 'heavy pion world'). Therefore, the task, that the lattice community approached very seriously, is to reduce the value of simulated quark ('pion') masses. The trouble is that reduction is very costly in computing power. Even if we manage to create clever algorithms to work closer to the chiral limit the artifacts due to finiteness of the lattice size (L) become more pronounced and the chiral extrapolations should be made by using the formulae derived by using the chiral perturbation theory in the finite volume. That issue attracted quite a bit of attention in the lattice community over the past couple of years [13]. In the case of nucleon, the leading order chiral lagrangian ^(1) "N
^{il^D^
mo)^^gA^lf,l5C^.
(4)
is consisted of the nucleon field iZ^, the covariant derivative D/^ = ^M + I K ^ ^yu*?]? in which the Goldstone boson fields enter via i7 = ^^, and ^^ = i^^d/^U^^. The standard axial coupling is used, QA = 12. Oneloop chiral corrections to the self energy of the nucleon propagator produce the shift to the bare nucleon mass, mo, as ^^9A
rriN = rno — 4cim^
+
2
_3
:m^ {^Truy
1
A
.
<
C{A)  2mo ^^^ (47r^)2 ^ + ^^ ^2
m^ + ... (5)
D. Becirevic: Getting to grips with hadrons
76
2.4
O JLQCD CPPACS O UKQCnSF
NH\/2 )
L\\]UTimc[U
>
> O
1.6 ¥UC NP imp. clover Wilson DWJ'
1.2 O.S
I
I
I
I
i
0.2
I
'^^.
0.4
I
Oil
I
i
\
N
I
O.K
nv,lGcV'l
0.8
Fig. 3 . The chiral extrapolation of the lattice QCD results with Nf = 2. Solid and dashed curves correspond to the socalled infrared regularization and to the nonrelativistic treatment of the nucleon. For more information, please see [15]
—I
M
0.2
L .
0.4
0.6
0.8
1
1.2
/H^ (GeV^) 2.8
1
1
1
'
1
*
m
2.4
where C{A) is the counterterm which cancels the yldependence t h a t otherwise arises from the renormalization of the ultraviolet divergences in the chiral loops. It t u r n s out, however, t h a t the two conventional descriptions lead to quite different results when applied to the lattice d a t a with TV/ = 2, and t h a t they coincide only for very light pions, namely TTITT <^ UIN [14,15]. This is shown in Fig. 3 where the solid curve indicates the fit to lattice d a t a and by using the expressions derived by means of the socalled infrared regularization [16]. T h e nonrelativistic treatment [17] is depicted by the two dashed lines corresponding to two specific choices of parameters ci and mo Besides providing the guidelines to extrapolate to the physically interesting limit for nucleons, C h P T is also useful in estimating the impact of the effects of the finiteness of the lattice box. It verifies the general Liischer formula [18] and provides the corrections to it. T h a t highly useful aspect of C h P T has been extended to other quantities involving baryons [19]. I m p o r t a n t to retain is t h a t the chiral logarithmic behavior gets enhanced by the finiteness of the lattice box. This makes the chiral extrapolations ever more delicate: physical chiral logarithms are mostly drowned in the finite size effects, and therefore one does not only seek the range of quark masses in which the C h P T are valid description of the lattice data, but one also has to disentangle the finite size effects from physical chiral logarithms (former being often overwhelming compared to the size of the latter). Finally, it should be noted t h a t the above discussion refers to the socalled pregime (i.e., with mjj^L ^ 1). A probably viable alternative has been recently proposed in [20] where it is claimed t h a t the eregime {TTIT^L
r
H
4
j :..
N\[/2*') «
5
J
1.6 I
*
i 14Hi"^
1 [JC Wilson J WilsonOPE 1 DWI
N
1.2 '^yy
0.8 0
0.2
0.4
0.6
0.8
1
1.2
m^ (GeV^) Fig. 4. Results for the orbital (upper plot) and radial (lower plot) excitations as a function of the pion mass directly accessible from the quenched QCD simulations on the lattice. For more details please see [21]
2.3 Fuss about the Roper resonance T h e experimental phenomenon for which the hadron physics phenomenologists do not have a viable explanation is the lightness of the radially excited state with q u a n t u m numbers of the nucleon (J^ = ^ ), also known as Roper resonance. T h e puzzle is t h a t its mass, TTIN' ~ 144 GeV, is smaller t h a n t h a t of the first orbital excitation, m ^ ^ 1.535 GeV, which contradicts most of the mass formulae derived by various forms of constituent quark model which suggest mN{^ ) < T^N{^ ) < T^N'i^ ) T h a t issue a t t r a c t e d quite a bit of attention recently since some lattice Q C D simulations claimed to have found the solution to t h a t puzzle. A special care should be devoted to choosing a good interpolating field, i.e. the one t h a t allows a good overlap with the orbital excitation J^ =  (also known as ^ n ) as discussed in [21]. T h e results of t h a t paper, along with those provided by other lattice groups [22], are shown in Fig. 4. In the region of quark masses directly accessed from the lattice studies, the mass p a t t e r n is consistent with quark models, i.e., rriNih )
D. Becirevic: Getting to grips with hadrons
Roper 2 >
o
3 Generalized parton distributions (GPD)
.i.
*
[D
S3,(I535)
m u ^ ^ l * * t '1 ^ 1.5
^
Lattice Q C D also offers the possibility to study the matrix elements of the local operators sandwiched by the hadron states. This is particularly important for the studies of the CPviolation in the Standard Model. For the recent review on t h a t topic, see [25]. A particularly interesting case to the nuclear physics community is the possibility to get some information about the G P D ' s of the nucleon. In particular, the matrix elements needed to study the 1^* moment have been studied by two lattice groups [26,27].
Nucleon
O *
1
s
77
J
T^
15 1.3
0.5
L
1
^^
'
^
Mj^/.M^.
:
:
0,1
n

0,2
E

1.7
1
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 ^P'\OU}\P)  {{P'\m,
/H/'(GeV)
D.}q\P)
Fig. 5. Results of [23]. Plot similar to those shown in Fig. 4
< Tnfj{^
) < mN'{^
), although the mass difference,
"f^N' (^ ) — '^Tv (^ )' cippears to be smaller as the quark mass is lowered. If naively extrapolated to the chiral limit, the level crossing can be envisaged too, i.e. after extrapolation rriN'i^ ) — ^ 7 v ( i ) ^ ^ y become negative. T h a t is what [23] claimed to see from their lattice study (see Fig. 5). However, as we saw with the nucleon mass, one should make sure t h a t the finite volume effects and the chiral behavior is well under control. While for the nucleon mass the help in t h a t respect comes from C h P T , there is no such effective theory t h a t provides a similar help for the nucleon excitations. For those reasons I personally think t h a t one should make a better control over various sources of systematic uncertainties in the lattice study of excited nucleons before claiming to observe the level crossing between the nucleon's orbital and radial exitations. This is the point at which one should mention t h a t the clean extraction of the mass of radially excited state has been for a long time a subject to controversies in the lattice community. A recent proposal of [24] provides a possible remedy. T h e idea is simple and it consists to consider the s t a n d a r d correlation function G(i) = ( ^ J ( x ) j t ( 0 ) ) = £ z 2 e 
ruit
(6)
i>0
We need not only to disentangle the excitations from the leading/dominant contribution (i.e. the one to the ground state), but  of all excitations we want to isolate the piece corresponding to the first radial excitation only. T h e proposal of [24] is to consider G{t) = G{t + l)G{t  1) 
+Cl{A^)—u{p')u{p)A^^A,y, rriN
where A = p — p' ^ and the operator is traceless and symmetrised over the indices in the curly brackets. T h e form factors ^ 2 , ^ 2 and C2 can be extracted for several kinematic configurations which then allows one to study their Z\^dependence. Both groups fit their d a t a {X = A2^B2^C2) to a dipole ansatz X(Z\2)
X(0) (1  M^/Z\2)2 '
OC
(9)
t h a t unfortunately does not provide us with more insight in physical mechanism t h a t governs the Z\^dependence. ^ It should be stressed, however, t h a t the calculation of the above matrix elements on the lattice is demanding for many reasons. One of the most involving problems is renormalization of the operators on the lattice containing the covariant derivative. T h e reason is t h a t at nonzero lattice spacing the Lorentz group in Euclidean space SO {A) is replaced by the group of discrete hypercubic rotations H{A)^ which additionally complicates the renormalization mixing p a t t e r n among various combinations of operators with covariant derivatives. Particularly interesting physics information from the first lattice Q C D studies of G P D ' s is the fraction of the total angular m o m e n t u m of the nucleon carried by the valence quarks. T h e total angular m o m e n t u m of a quark q in the nucleon can be expressed via [28]
j , = \[Am+Bim.
G{t)G{t)
(8)
(10)
(7)
In [26], the following values have been reported: A^{{}) = 0.40(2), ^ f (0) = 0.15(1), and 5 ^ ( 0 ) = 0.33(11), 5  ( 0 ) =  0 . 2 3 ( 8 ) , and therefore Ju = 0.37(6) and Jd =  0 . 0 4 ( 4 ) .
SO t h a t for large time separation the ground state contributes less, and the first radial excitation is then more accessible. T h e first numerical studies also seem to be encouraging in t h a t respect.
^ Resonances in the crossed tchannel are poles in the dispersion relations for these form factors. Apart from the convenient fit formula, no reasonable physical significance could be given to the resulting M^^^ ^.
2 2 ^ ^i2je(™'+™^'*, i>i=0
78
D. Becirevic: Getting to grips with hadrons
In other words, about 30% of the (quenched) proton's angular m o m e n t u m is carried by the gluons. T h e situation in the world with Nf = 2 dynamical quarks seems to show t h a t the fraction of the total angular m o m e n t u m carried by the valence quarks is even smaller. A more definite conclusion on this issue necessitates a more control over the chiral extrapolations t h a t are involved in these calculations.
7. 8. 9.
10.
4 Conclusion
11.
In conclusion I would like to point out the main benefits of the lattice Q C D approach:
12. 13.
— Lattice Q C D offers a unique possibility to study the physics of hadrons on the basis of the Q C D lagrangian only. — T h e high statistics numerical simulations of Q C D on the lattice have so far been done in the so called quenched approximation. Nowadays more and more studies are made by including the effects of dynamical quarks. — T h e methodology to extract the physically interesting information from the d a t a produced on the lattice is developed in the world with heavy pions. It is highly important to extend the range of directly accessible quark masses on the lattice to lighter ones in order to confront the quark mass dependence observed on the lattice with the expressions obtained by means of ChPT. — Many phenomenologically relevant question in particle physics have been studied by using lattice Q C D . If we are to make the precision calculation of hadronic quantities on the lattice, we first need to solve the above mentioned problems. More computing power, better algorithms, more clever physical ideas and the combination of all three aspects are essential in reaching t h a t goal.
14.
15. 16. 17.
18. 19.
20.
21.
References 1. S. DescotesGenon, N.H. Fuchs, L. Girlanda, J. Stern: Eur. Phys. J. C 34, 201 (2004) 2. H. Leutwyler: Phys. Lett. B 378, 313 (1996) 3. S. Bethke: Nucl. Phys. Proc. Suppl. 121, 74 (2003) 4. M. Luscher et al.: Nucl. Phys. B 384, 168 (1992) [heplat/9207009]; K. Jansen et al.: Phys. Lett. B 372, 275 (1996) [heplat/9512009]; G. Martinelh et al.: Nucl. Phys. B 445, 81 (1995) [heplat/9411010]; M. Gockeler et al.: Nucl. Phys. B 544, 699 (1999) [heplat/9807044] 5. R. Gupta: eConf C0304052 (2003) WG503 [hepph/0311033] 6. S. Aoki et al., [JLQCD Collaboration]: Phys. Rev. Lett. 82, 4392 (1999) [heplat/9901019]; J. Garden et al., [ALPHA Collaboration]: Nucl. Phys.B 571, 237 (2000) [heplat/9906013]; M. Gockeler et al.: Phys. Rev. D 62, 054504 (2000) [heplat/9908005]; D. Becirevic, V. Lubicz, C. Tarantino, [SPQ(CD)R Collaboration]: Phys. Lett. B
22.
23. 24. 25. 26.
27. 28.
558, 69 (2003) [heplat/0208003]; S. Aoki et al., [CPPACS Collaboration]: Phys. Rev. D 67, 034503 (2003) [heplat/0206009] PE.L. Rakow: heplat/0411036 C.J. Morningstar, M.J. Peardon: Phys. Rev. D 60, 034509 (1999) Please see the following URL: h t t p : //www.claymath.org/millerLnium/Yang —Mills_Theory/Of ficial_Problem_Description.pdf R. Sommer: Nucl. Phys.B 411, 839 (1994) [heplat/9310022] S. Aoki et al., [JLQCD Collaboration]: Phys. Rev. D 68, 054502 (2003) [heplat/0212039] S. Eidelman et al., [Particle Data Group Collaboration]: Phys. Lett. B 592, 1 (2004) G. Colangelo, S. Durr: Eur. Phys. J. C 33, 543 (2004) [heplat/0311023]; D. Becirevic, G. Villadoro: Phys. Rev. D 69, 054010 (2004) [heplat/0311028]; D. Arndt, C.J.D. Lin: Phys. Rev. D 70, 014503 (2004) [heplat/0403012]; G. Colangelo, C. Haefeli: Phys. Lett. B 590, 258 (2004) [heplat/0403025]; C.J.D. Lin et al.: Phys. Lett. B 581, 207 (2004) [heplat/0308014] D.B. Leinweber, A.W. Thomas, R.D. Young: Phys. Rev. Lett. 92, 242002 (2004) [heplat/0302020]; T.R. Hemmert, M. Procura, W. Weise: Nucl. Phys. A 721, 938 (2003) [heplat/0301021]; M. Frink, U.G. Meissner: JHEP 0407, 028 (2004) [heplat/0404018] A. AH Khan et al., [QCDSFUKQCD Collaboration]: Nucl. Phys. B 689, 175 (2004) [heplat/0312030] T. Becher, H. Leutwyler: Eur. Phys. J. C 9, 643 (1999) [hepph/9901384] N. Fettes, U.G. Meissner, M. Mojzis, S. Steininger: Annals Phys. 283, 273 (2000) [Erratumibid. 288, 249 (2001)] [hepph/0001308] M. Luscher: Commun. Math. Phys. 104, 177 (1986) S.R. Beane, M.J. Savage: Phys. Rev. D 70, 074029 (2004) [hepph/0404131]; S.R. Beane: Phys. Rev.D 70, 034507 (2004) [heplat/0403015]; R.D. Young, D.B. Leinweber, A.W. Thomas: heplat/0406001 P.F. Bedaque, H.W. Griesshammer, G. Rupak: heplat/0407009; see also W. Detmold, M.J. Savage: Phys. Lett. B 599, 32 (2004) [heplat/0407008] W. Melnitchouk et al.: Phys. Rev. D 67, 114506 (2003) [heplat/0202022] T. Burch et al., [BGR Collaboration]: Phys. Rev.D 70, 054502 (2004) [heplat/0405006]; S. Sasaki, T. Blum, S. Ohta: Phys. Rev. D 65, 074503 (2002) [heplat/0102010]; Y. Nemoto et al.: Nucl. Phys. A 721, 879 (2003) [heplat/0312033]; K. Sasaki et al.: Nucl. Phys. Proc. Suppl. 129, 212 (2004) [heplat/0309177]; M. Gockeler et al., [QCDSF and UKQCD]: Phys. Lett. B 532, 63 (2002) [heplat/0106022] S.J. Dong et al.: hepph/0306199 D. Guadagnoli, M. Papinutto, S. Simula: Phys. Lett. B 604, 74 (2004) [heplat/0409011] S. Hashimoto: hepph/0411126 M. Gockeler et al., [QCDSF Collaboration]: Phys. Rev. Lett. 92, 042002 (2004) [hepph/0304249]; see also heplat/0410023 P. Hagler et al., [LHPC and SESAM]: Phys. Rev. Lett. 93, 112001 (2004) [heplat/0312014]; see also hepph/0410017 X.D. Ji: Phys. Rev. Lett. 78, 610 (1997) [hepph/9603249]
Eur Phys J A (2005) 24, s2, 7984 DOI: 10.1140/epjad/s200504017y
EPJ A direct electronic only
Systematic uncertainties in the precise determination of the strangeness magnetic moment of the nucleon D.B. Leinweber^'2, S. B o i n e p a l l i \ A . W . Thomas^'^, A.G. W i l l i a m s \ R.D. Young^'^, J . B . Z h a n g \ and J.M. Zanotti^ ^ Special Research Center for the Subatomic Structure of Matter, and Department of Physics, University of Adelaide, Adelaide SA 5005, Austraha ^ Jefferson Lab, 12000 Jefferson Ave., Newport News, VA 23606, USA ^ John von NeumannInstitut fiir Computing NIC, Deutsches ElektronenSynchrotron DESY, D15738 Zeuthen, Germany Received: 1 December 2004 / Pubhshed Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. Systematic uncertainties in the recent precise determination of the strangeness magnetic moment of the nucleon are identified and quantified. In summary, GM — —0.046 dz 0.019 fiNPACS. 13.40.Em Electric and magnetic moments  12.38.Gc Lattice QCD calculations  12.39.Fe Chiral Lagrangians
1 Introduction Recent lowmass lattice Q C D simulation results combined with new chiral extrapolation techniques and t h e principle of charge symmetry have enabled a precise determination of t h e strangeness magnetic moment of t h e nucleon G%f [1]. In this paper, t h e systematic errors of t h e approach are explored and quantified. In particular, we examine t h e sensitivity of G%^ and t h e magnetic moments of t h e baryon octet t o t h e regulatormass scale of finiterange regularized chiral effective field theory, t h e lattice scale determination, t h e finitevolume of t h e lattice, and the quenched approximation.
2 Charge symmetry T h e approach [1] centres around two equations for t h e strangeness magnetic moment of t h e nucleon, Gl^^ ^M^ obtained from charge symmetry lQS
Gl
'm
2p + n
(r+r)
p + 2n
(")
I JDS
Gl
1  'm
(2)
Here t h e baryon labels represent t h e experimentally measured baryon magnetic moments and ^R% = G%^/^GJ^ is the ratio of s and cZquark seaquark loop contributions, depicted in t h e righthand diagram of Fig. 1. ^Rl lies m the range (0,1). T h e ratios uF ju^ and nP' ju^ are ratios of valencequark contributions t o baryon magnetic moments in full Q C D as depicted in t h e lefthand diagram of Fig. 1. T h e latter are determined by lattice Q C D calculations [1, 3,4,5,6].
u,d,s
Fig. 1. Diagrams illustrating the two topologically different insertions of the current within the framework of lattice QCD. These skeleton diagrams for the connected (left) and disconnected (right) current insertions are dressed by an arbitrary number of gluons and quark loops
Equating (1) and (2) provides a linear relationship between uP ju^ and u^/u^ which must be obeyed within Q C D under t h e assumption of charge symmetry  itself typically satisfied at t h e level of 1% or better [2]. There are no other systematic uncertainties associated with this constraint. Figure 2 displays this relationship by t h e dashed and solid line. Since this line does not pass through t h e point (1.0,1.0), corresponding t o t h e simple quark model assumption of universality, there must be an environment effect exceeding 12% in b o t h ratios or approaching 20% or more in at least one of t h e ratios. To determine t h e sign of C ^ , it is sufficient t o determine where on this constraint curve, Q C D resides.
3 vPlu^
and u^/u^
determination
Our present precise analysis has been made possible by a significant breakthrough in t h e regularization of t h e chiralloop contributions t o hadron observables [7,8,9]. Through
D.B. Leinweber et al.: Systematic uncertainties
80 1.4:
1
1
1
1 Finite Vol. QQCD QQCD Valence Sector Full QCD
1.3 1.2 
^^^^^^.
1.1
" " * .
1.0
^^
S
^'V*^^^^^
0.9 1
A Q
0.0
0.5
1
1.0
1
1
1.5
2.0
0.4
2.5
(GeV^) Fig. 2. The straight hne {dashed GM{0) < 0, solid G'M(O) > 0) indicates the constraint on the ratios u^ jvP and u^ ju^ imphed by charge symmetry and the experimentahy measured magnetic moments. The assumption of environment independent quark moments is indicated by the square. The dependence of the extrapolated ratios from lattice QCD simulations on the parameter A — 0.7, 0.8 and 0.9 GeV, (governing the size of pioncloud corrections) is illustrated by the cluster of points with A increasing from left to right
Fig. 3. The contribution of a single u quark (with unit charge) to the magnetic moment of the proton. Lattice simulation results {square symbols for m^ > 0.05 GeV^) are extrapolated to the physical point {vertical dashed line) in finitevolume QQCD as well as infinite volume QQCD. Estimates of the valence u quark contribution in full QCD and the full i^quark sector contribution in full QCD are also illustrated. Extrapolated values at the physical pion mass {vertical dashed line), are offset for clarity
the process of regulating loop integrals via a finiterange regulator (FRR) [8,10], the chiral expansion is eflPectively resummed to produce an expansion with vastly improved convergence properties. T h e chiral expansion for the 7xquark contribution to the proton magnetic moment in quenched Q C D (QQCD), has the form
runs through the lattice results corresponds to replacing the discrete m o m e n t u m sum by the infinitevolume, continuous m o m e n t u m integral. For all but the lightest quark mass, finite volume effects are negligible. Having determined the analytic coeflficients a^^A ^^^ ^ particular choice of vl, one can correct the chiral properties of the pioncloud contribution from Q Q C D to full Q C D [9, 13] by changing the coefl^icients of the loop integrals, Xr]'^ XTTB, XKB of (3), to their full Q C D counter parts [11, 12]. Valence quark contributions in full Q C D are indicated by the longdashdot curve in Fig. 3 (i.e. seaquark loop charges are zero) and the full i^quark sector including the ixseaquark loop contributions are indicated by the shortdashdot curve for yl = 0.8 GeV. Figures 4, 5 and 6 show similar results for the u quark in n, U^ ^ and E^ respectively. T h e importance of correcting for b o t h finitevolume and quenching artifacts is illustrated in Figs. 7 and 8, where the one s t a n d a r d deviation agreement between the chirally corrected lattice Q C D simulation results and the experimentally measured baryon magnetic moments is highlighted.
^XKB
lB{mK,
A) \a2ml\
a^ m ^ .
(3)
where the repeated index, B^ sums over allowed baryon octet and decuplet intermediate states. T h e dependence of the unrenormalized coefl&cients, a^, and the associated dipolevertex regulated loop integrals, I{m^,A), on the regulator parameter, A, is emphasized by the explicit appearance of A. T h e loop integrals are defined as lB{m,A)
[ 37r J Irj>{m^,A)
(4)
dk
(2Vfc2 + m2 + ABN)
k*u'^{k,A)
(fc2 + m2)3/2 (^fc2 _,_ ^ 2 _,_ z i g ^ ) ' k* =  / u{k,A), dk(fc2+m2) Jo
(5)
where ABN is the relevant baryon mass splitting and the function u{k,A) is the dipolevert ex regulator. T h e coefficients, X5 denote the known modelindependent coeflficients of the nonanalytic terms for TT and K mesons in Q Q C D [11,12]. Figure 3 illustrates a fit of F R R , quenched chiral perturbation theory ( x P T ) to our fermion lattice results (solid curve), where only the discrete momenta allowed in the finite volume of the lattice are summed in evaluating the chiral loop integrals. T h e longdashed curve t h a t also
4 Systematic errors 4.1 Regulator dependence It is important to investigate systematic errors associated with the regulatormass dependence of F R R xPT . T h e extrapolated results of finitevolume quenched chiral effective field theory should be insensitive to the choice of regulator parameter. W h e n working to sufficient order in
D.B. Leinweber et al.: Systematic uncertainties
81 1
Finite Vol. QQCD QQCD Valence Sector Full QCD
1
Finite Vol. QQCD QQCD Valence Sector  Full QCD J_
0.7
0.6
0.7
m 2 (GeV^) Fig. 4. The contribution of the u quark (with unit charge) to the magnetic moment of the neutron. Curves and symbols are as described in Fig. 3
Fig. 6. The contribution of the u quark (with unit charge) to the magnetic moment of the S^ hyperon. Curves and symbols are as described in Fig. 3
2.2 3.2
Finite Vol. QQCD QQCD Valence Sector Full QCD
3.0 2.8 2.6 2.4
^^2.2
0.8
0.0
0.1
0.2
0.3 0.4 m^2 (GeV^)
0.5
0.6
0.7
1.8 1.6 1.4 1.2 1.0
I
1
I
I
1
1
1
"I

^*}  \ \
ii
I
\ 1
1
1
1
1
1
1
1
A u. u„ V n Fig. 5. The contribution of a single u quark (with unit charge) Fig. 7. The one standard deviation agreement between the to the magnetic moment of U~^. Curves and symbols are as for chirally corrected lattice QCD simulation results {square symFig. 3 bols) and the experimentally measured baryon magnetic moments {circular symbols) having positive values. Finitevolume quenched results {crossed boxes) and infinitevolume quenched the chiral expansion, changes in the regulation of loop in results {diamonds) are also illustrated to highlight the importegrals should be absorbed by changes in the unrenormal tance of correcting for both finitevolume and quenching artiized coefficients, (IQ2A'> i^ ^ manner which preserves the facts invariant renormalized coefficients. T h e latter are reflected in Fig. 9 which illustrates the insensitivity of quenched Because the strangeness magnetic moment of (1) baryon magnetic moments in a finitevolume to the reguand (2) depends only on ratios of magnetic moments, most lator parameter A. This systematic error is small relative of this A dependence cancels in the final ratios, as illusto the statistical error. Since the finitevolume and quenching corrections are t r a t e d by the close clustering of points in Fig. 2. applied only to the loop integral contributions, the final results are A dependent. In this case, the regulator of the loop integral has become a model for the axialvector form factor of the nucleon, describing the coupling of pious to a core described by the analytic terms of the F R R expansion. This approach describes the relation between quenched and full Q C D N and A mass simulation results as a function of TTI^^ very accurately [13]. Figures 10 and 11 display results for positive and negative baryon magnetic moments in full Q C D respectively. Onestandard deviation agreement is achieved for 0.7 < ^1 < 0.9.
S
H° H"
4.2 Role of the decuplet in x P T It is often argued t h a t nexttoleadingorder nonanalytic (NLNA) contributions from the baryon decuplet are essential in describing the mass dependence of nucleon magnetic moments. While the decuplet baryon contributions are not necessarily small, we find the nonanalytic curvature induced by these contributions is sufficiently subtle t h a t it may be accurately absorbed by the analytic terms
D.B. Leinweber et al.: Systematic uncertainties
82 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
1
1
1
1
1
1
1
1


i' li

^i

_
_
1
p
1
1
1
E
n
1
H° E"
1
1
A
1
u^ u»
Fig. 8. The one standard deviation agreement between the chirally corrected lattice QCD simulation results and the experimentally measured baryon magnetic moments having negative values. Symbols are as in Fig. 7 3.5 3.0 r 2.5 2.0 1.5 la; 1.0 0.5 0.0 0.5 I 1.0 1.5 l 2.0 L 2.5
\
r
*
n
a
J
^m
p
n
Fig. 10. The dependence of positive octetbaryon magnetic moments on the parameter A = 0.6, 0.7, 0.8, 0.9 and 1.0 GeV, governing the size of pioncloud corrections associated with the finitevolume of the lattice and artifacts of the quenched approximation. Experimental measurements, illustrated at left by the filled circle for each baryon, indicate that optimal corrections are obtained for 0.6 < A < 0.9 GeV
\ J
S"" S
H"^
A
u^ u^
Fig. 9. The FRR x P T regulator mass dependence of finitevolume quenched chiral effective field theory. Experimental measurements are illustrated at left for each baryon for reference. Results for A = 0.6, 0.7, 0.8, 0.9 and 1.0 GeV are illustrated from left to right for each baryon. The small systematic dependence on A relative to the statistical error bars illustrated indicate the order of the chiral expansion is adequate for this analysis
of the chiral expansion. Figure 12 illustrates the insensitivity of finitevolume quenched chiral effective field theory to NLNA decupletbaryon contributions. Figure 13 confirms t h a t the NLNA decupletbaryon contributions are indeed large in some cases and as such are important in correcting the artifacts of the quenched approximation. However, other more highlyexcited baryon resonances have small couplings to the groundstate baryon octet relative to t h a t for the decuplet and provide negligible corrections.
4.3 Scale dependence Setting the scale in quenched Q C D simulations is somewhat problematic. Different observables lead to different
Fig. 1 1 . The dependence of negative octetbaryon magnetic moments on the parameter A = 0.6, 0.7, 0.8, 0.9 and 1.0 GeV, governing the size of pioncloud corrections associated with the finitevolume of the lattice and artifacts of the quenched approximation. Experimental measurements, illustrated at left by the filled circle for each baryon, indicate that optimal corrections are obtained for 0.7 < A < 0.9 GeV
lattice spacings, a. If one is explicitly correcting the oneloop pioncloud contributions to hadronic observables, as we are here, then clearly one must set the scale using an observable insensitive to chiral physics. This excludes observables such as the rhomeson mass, nucleon mass, or the pion decay constant commonly used in the literature to hide the artifacts of the quenched approximation. On the other hand, the heavyquark phenomenology of the staticquark potential provides an optimal case. In particular, the Sommer parameter, TQ, is ideal as it sets the scale by equating the force between two static quarks in Q Q C D and full Q C D at a precise separation of TQ = 0.49 fm.
D.B. Leinweber et al.: Systematic uncertainties 3.5 3.0 2.5 2.0 1.5
5
1.0 0.5
•
•
a
11

• ••


:l 0.0 0.5 1.0 1.5 2.0 2.5

•••
•
•••
•••

•••
•
p
n
LI
1^
H1^
A
5
1
1
1
.ii* •••^
• ••A
•••A • ••A •••A •••A
1
1
1
n
LI
HH
A
HH
u^ u^
Fig. 14. The dependence of octetbaryon magnetic moments on the matching criteria for determining the lattice spacing, a. While ttro — 0.128 fm set by ro (square symbols) is the preferred method for determining the scale, results for aa = 0.134 fm set by the string tension (triangles) and a third spacing of 0.122 fm (diamonds) are also illustrated. Experimental measurements (filled circles), are illustrated at left for each baryon
thirdorder singleelimination jackknife analysis, one finds .Mi
^
•••
= 1.092
0 and —.
1.254
.
(6)

1.0 0.5
Using the experimental magnetic moments, one observes t h a t (2)
:3. 0.0 0.5 1.0 " 1.5 2.0 2.5
p
u^ u^
Fig. 12. The insensitivity of finitevolume quenched chiral effective field theory to NLNA decupletbaryon contributions. Results including decuplet intermediate states in chiral effective field theory (squares) are compared with results excluding the decuplet (diamonds). Experimental measurements (circles), are illustrated at left for each baryon for reference
3.5 3.0 .ij 2.5 2.0 1.5 
3.5 3.0 .1** 2.5 2.0 1.5 1.0 0.5 0.0 0.5 1.0 1.5 2.0 1 2.5
83
••• •••
•••
•••
•••
"
• 1
p n E E H H A u ^ ' ^ w Fig. 13. The dependence of octetbaryon magnetic moments on the inclusion of NLNA decupletbaryon contributions in the process of correcting the pion cloud of quenched chiral effective field theory. Preferred results including decuplet intermediate states in chiral effective field theory (squares) are compared with results excluding the decuplet (diamonds). Experimental measurements (circles), are illustrated at left for each baryon for reference
T h e string tension might also be used, but screening of the potential in full Q C D makes this measure poorly defined. However, we consider it here as a measure of the systematic error encountered in setting the scale of the lattice Q C D results. Figure 14 illustrates this scale dependence on baryon magnetic moments and Fig. 15 illustrates the rather minor impact this systematic uncertainty has on the valencequark moment ratios vital to determining the sign of G%^. Collecting the variation of the valencequark moment ratios from variations in A (which maintain onestandarddeviation agreement with experiment), variations in setting the lattice scale and statistical errors determined by a
Gl
'm
1.033
(0.599)
(7)
is least sensitive to variation in the valencequark moment ratio, and hence provides the most precise determination for G ^ . Figure 16 plots G ^ as a function of ^i?^ with s t a n d a r d error limits provided by (6).
5 Estimating T h e symmetry of the threepoint correlation function [14] describing the seaquark loop contributions to the nucleon, depicted in the righthand side of Fig. 1, ensure t h a t the chiral expansion for this quantity is identical for all three quark flavours, u p to simple charge factors. For the d or 5quark loop contributions, the only difference t h a t can arise is whether one evaluates the chiral expansion at the pion or kaon mass. T h e leading nonanalytic contribution to the chiral expansion involves two pseudoscalar meson propagators, and therefore one expects contributions to ^i?^ in the ratio rn^/m^^ ~ 0.1. To be more precise, one can use the same successful (singleparameter) model, previously used to correct the quenched simulation results to full Q C D , as highlighted in Figs. 7 and 8, to provide an estimate for ^i?^. Evaluating the loop integrals with A = 0.8 0.2 GeV yields ^Rl = 0.139 with 0.096 < ^R'^ < 0.181. This uncertainty dominates the final uncertainty in G ^ .
D.B. Leinweber et al.: Systematic uncertainties
84
Table 1. Sources of uncertainty and their contribution to the strangeness magnetic moment of the nucleon, C M , in units of nuclear magnetons, /J^N Uncertainties are documented for G^ obtained from the valencequark ratio u'^/u^ in (1), from the valencequark ratio u^ ju^ in (2) and from a statistically weighted (SW) average of these two determinations
Uncertainty Source
Parameter Range
Statistical Errors Chiral corrections Scale Determination ^i^J Determination
0.7 < yl < 0.9 G e V 0.122 < a < 0.134 fm 0.096 < ^Rl < 0.181
Total Uncertainty
u^lu'', (1) G'M = 0.045
u^'lu^, (2) G'M = 0.046
SW Average G'M = 0.046
0.016 0.001 0.001 0.016
0.009 0.002 0.002 0.017
0.008 0.002 0.002 0.017
0.023
0.019
0.019
Table 1 summarizes the sources of uncertainty and their contributions to the final determination
GM
0.046
0.019 fiN ,
(8)
for the strange quark contribution to the magnetic moment of the nucleon.
1.0
1.5
Fig. 15. The charge symmetry constraint line (dashed GM{0) < 0, solid GM{0) > 0) on the ratios u^/u^ and u'^/u^. The dependence of the ratios from chirallycorrected quenched lattice QCD on the scale parameter, a = 0.134, 0.128, and 0.122 fm, is illustrated by the cluster of points with a decreasing from left to right
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
1.0
Fig. 16. The dependence of G%f on the strange to light seaquark loop ratio R^ Standard error limits have there origin in the systematic error summary of 6
Acknowledgements. We thank the Australian Partnership for Advanced Computing (APAC) for generous grants of supercomputer time which have enabled this project. Support from the South Australian Partnership for Advanced Computing (SAPAC) and the National Facility for Lattice Gauge Theory is also gratefully acknowledged. DEL thanks Jefferson Lab for their kind hospitality where the majority of this research was performed. This work is supported by the Australian Research Council and by DOE contract DEAC0584ER40150 under which SURA operates Jefferson Lab.
References 1. D.B. Leinweber et al.: arXiv:heplat/0406002 2. G.A. Miller, B.M. Nefkens, I. Slaus: Phys. Rept. 194, 1 (1990) 3. J.M. Zanotti et al.: Phys. Rev. D 65, 074507 (2002) [heplat/0110216] 4. D.B. Leinweber et al.: Eur. Phys. J. A 18, 247 (2003) [nuclth/0211014] 5. J.M. Zanotti et al.: heplat/0405015 6. J.M. Zanotti et al.: heplat/0405026 7. R.D. Young et al.: Prog. Part. Nucl. Phys. 50, 399 (2003) [heplat/0212031] 8. D.B. Leinweber, A.W. Thomas, R.D. Young: Phys. Rev. Lett. 92, 242002 (2004) [heplat/0302020] 9. R.D. Young, D.B. Leinweber, A.W. Thomas: arXiv:heplat/0406001 10. J.F. Donoghue, B.R. Holstein, B. Borasoy: Phys. Rev. D 59, 036002 (1999) 11. D.B. Leinweber: Phys. Rev. D 69, 014005 (2004) [heplat/0211017] 12. M.J. Savage: Nucl. Phys. A 700, 359 (2002) [nuclth/0107038] 13. R.D. Young et al.: Phys. Rev. D 66, 094507 (2002) [arXiv:heplat/0205017] 14. D.B. Leinweber: Phys. Rev. D 53, 5115 (1996) hepph/9512319
Eur Phys J A (2005) 24, s2, 8588 DOI: 10.1140/epjad/s200504018x
EPJ A direct electronic only
Current status of parton charge symmetry J . T . Londergan Dept. of Physics, Indiana University, Bloomington, IN, 47404, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. We review constraints on parton charge symmetry from various experiments. Recently charge symmetry violation (CSV) has been included in a global fit to high energy data. We show that CSV compatible with all high energy data would be able to remove completely the NuTeV anomaly. PACS. ll.30.Hv Flavor Symmetries  13.15.+g Neutrino Interactions crosssections, including deep inelastic structure functions
1 Experimental limits on parton CSV
Q{x)
13.60.Hb Total and inclusive
^
x[qj{x)
+ qj{x)]
(1)
= [u,d,s,c]
Charge symmetry is a restricted form of isospin invariance involving a rotation of 180° about the "2" axis in isospin space. For parton distributions, charge symmetry involves interchanging u p and down quarks while simultaneously interchanging protons and neutrons. In nuclear physics, charge symmetry is generally obeyed at the level of a fraction of a percent [1,2]. Charge symmetry violation (CSV) in p a r t o n distribution functions (PDFs) arises from two sources; from the difference Sm = rud — rriu between down and u p current quark masses, and from electromagnetic (EM) effects. Since charge symmetry is so well satisfied at lower energies, it is natural to assume t h a t it holds for p a r t o n distributions. At present, there is no direct experimental evidence of substantial violation of charge symmet r y in parton distribution functions ( P D F s ) . In a recent paper [3], we have reviewed experimental and theoretical estimates for parton CSV, and have discussed potential corrections to the extraction of the Weinberg angle in neutrino deep inelastic scattering (DIS). We summarize our arguments here. T h e most stringent upper limits on p a r t o n CSV come from comparing the structure —w _ function F2 , the average of u and u charged current reactions, and the structure function F2 for charged lepton DIS, on isoscalar targets (A^o) In leading order (LO), F2 depends on the squared charges of the quarks, while ^ 2 ^ depends on the quark weak charges. Assuming charge symmetry gives a simple relation between the structure functions, defined as the "charge ratio" Rc{x,Q'^). To lowest order in the (presumably small) CSV terms Rc{x)
F^^°{x)
+ X {s{x) + s{x)  c{x)  c{x)) / 6 5^2
1+
{x)/li
Zx (5u{x) + 5u{x) — 5d{x) — 5d{x)) lOQ(x)
Send offprint requests to: [emailprotected]
1 introduces the CSV parton distributions, 6u{x) = uP{x)d''{x);
6d{x) = (P{x)u''{x),
(2)
with analogous relations for antiquarks. Deviation of Rc{x) from unity would indicate a CSV contribution. T h e most precise neutrino measurements were obtained by the C C F R group [4], who extracted the F2 structure function for uFe and VFe reactions. Muon DIS measurements were obtained by the N M C group [5,6], who measured F2 structure functions for muon interactions on deuterium at muon energies E^ = 90 and 280 GeV. Taking into account many corrections (relative normalizations; heavy quark threshold effects; nuclear effects; corrections for excess neutrons in iron; contributions from s and c quarks), C C F R obtained results consistent with unity at about the 2 — 3 % level, in the range 0.1 < J: < 0.4. From (1), this gives an upper limit to parton CSV effects in the 6 — 9% range. At smaller x, Re appeared to deviate significantly from unity. However, upon reanalysis [7] the ratio agrees with unity at the 2 — 3 % level down to X ~ 0.03, as significant effects were found from NLO treatment of charm mass corrections, and separation of the F2 and F3 structure functions in v DIS. Other limits on p a r t o n CSV can come from measurements of W^ asymmetry in d^p—p collider. Since u quarks carry more m o m e n t u m t h a n d quarks, the direction of the W^ and p tend to be aligned, as do the W~ and p. Measurement of the W charge asymmetry is thus quite sensitive to the proton's u and d distributions. Conversely, charged lepton DIS on an isoscalar target tends to be more sensitive to u^ t h a n to d^, as it is more heavily weighted due to the squared quark charge. Comparison of, say, the C D F W charge asymmetry [8] and N M C //  D DIS can constrain some aspects of parton CSV.
J.T. Londergan: Current status of parton charge symmetry
86
d"(x)
Fig. 1. The valence quark CSV function from [9], corresponding to best fit value K = —0.2 defined in 3. Solid curve: x6dv{x); dashed curve: xSuv{x)
2 Phenomenology and theory of parton CSV Because CSV effects are typically very small at nuclear physics energy scales, all previous phenomenological P D F s have assumed the validity of p a r t o n charge symmetry. However, Martin, Roberts, Stirling and Thorne (MRST) [9] have recently studied the uncertainties in parton distributions arising from a number of factors, including isospin violation. M R S T chose a specific model for valence quark charge symmetry violating P D F s : Su^{x) = Sd^(x)
= /^(l  x)^x^^(x
 .0909).
(3)
At b o t h small and large x the M R S T CSV P D F s have the same form as the valence distributions. T h e first moment of the M R S T valence CSV function is zero; this must be the case since, e.g., the integral (Suy) is just the total number of valence up quarks in the proton minus the number of down quarks in the neutron. T h e second moment of this function represents the CSV m o m e n t u m asymmetry; SU^ = {xSu^{x)) is the difference in total m o m e n t u m carried by < and < . T h e M R S T valence CSV distributions require t h a t Su^ and Sd^ have opposite signs at large x, in agreement with theoretical predictions. This condition also insures t h a t valence quarks in the proton and neutron carry an equal amount of total m o m e n t u m (this is strictly true only at the starting scale, since the moment u m asymmetry is not constant under Q C D evolution; however M R S T find t h a t it does not change very much over a fairly wide Q^ range). The overall coefficient K was varied in a global fit to a wide range of high energy data. T h e value K = —0.2 minimised x^. Their x^ had a shallow minimum with the 90% confidence level obtained for the range —0.8 < n < +0.65. In Fig. 1 we plot the valence quark CSV P D F s corresponding to the M R S T best fit value n = —0.2. W i t h i n the 90% confidence region for the global fit, the valence quark CSV P D F s could be either four times as large as in Fig. 1, or it could be three times as big with the opposite sign. CSV distributions with this shape, and for K within this range, will not disagree seriously with any of the high energy d a t a used to extract quark and gluon P D F s . T h e M R S T group also searched for CSV in the sea quark sector. Again, they chose a specific form for sea quark CSV, dependent on a single parameter, i.e., u''{x) = dP{x)[l + 6]
= uP(x) [1  6]
(4)
Somewhat surprisingly, evidence for sea quark CSV in the global fit was substantially stronger t h a n for valence quark CSV. T h e best fit was obtained for 5 = 0.08, corresponding to an 8% violation of charge symmetry in the nucleon sea. This is considerably larger t h a n theoretical estimates of sea quark CSV [10]. T h e x^ foi" this value is substantially better t h a n with no CSV, primarily because of improvement in fits to the N M C /i — D DIS d a t a [5,6] and to the E605 DrellYan d a t a [11], when sea quark CSV is included. T h e M R S T bestfit values will necessarily give reasonable agreement with the charge ratio of (1), since b o t h the C C F R i/ Xsections and N M C muon DIS are included in the global fit. T h e M R S T group also includes the C D F W charge asymmetry measurements [8], so t h a t the M R S T global fit P D F s including CSV are compatible with all d a t a sets t h a t are most sensitive to charge symmetry violating effects. T h e M R S T phenomenological CSV distributions agree rather well with two earlier predictions using quark models. T h e Adelaide group [12] developed a m e t h o d for calculating twisttwo valence P D F s from quark model wavefunctions. Unlike earlier calculations, this model guaranteed correct support for the P D F s . Rodionov et al. [13] extended this model to calculate valence quark CSV. Sather [14] approximated the dependence of valence quark P D F s on the quark and nucleon masses, and obtained analytic approximations relating valence quark CSV to derivatives of the valence P D F s . Although there are several differences between the models of Sather and Rodionov, their predictions of valence quark CSV are quite similar. In Fig. 2, we show the theoretical valence quark CSV prediction of Rodionov. T h e solid curve is xSu^{x), while the dotdashed curve is xSd^{x), b o t h evolved to Q^ = 10 GeV^. Qualitatively, the results of Rodionov et al. are very similar to the bestfit phenomenological CSV distribution of M R S T , shown in Fig. 1. T h e sign and relative magnitude of b o t h Sd^{x) and Su^{x) are quite similar in b o t h phenomenology and theory. T h e second moments of the CSV P D F s (which give the total m o m e n t u m asymmetry between, e.g., u^ and d^) of the M R S T and Rodionov distributions are equal to within 10%.
3 Parton CSV and the NuTeV anomaly In 1973, Paschos and Wolfenstein [15] suggested t h a t the ratio of neutralcurrent (NC) and chargechanging (CC) neutrino cross sections on isoscalar targets could provide an independent measurement of the Weinberg angle (siii^Ow) T h e PaschosWolfenstein ( P W ) ratio R~ is given by R
sm
1rR^'
(5) In 5, {cr^^^) is ^^^ ^ ^ inclusive total cross section for neutrinos on an isoscalar target. T h e quantity po is one in the Standard Model.
J.T. Londergan: Current status of parton charge symmetry 0.006
<^r
0.004
\
i
/
i
0.002

K
\
]
] ]
I
0.000
^
i
]
0.002
]
0.004 0.006 ' 0.0
.
1
0.2
.
.
.
—1
i
1
1
0.4
0,6
0.8
1.0
X
Fig. 2. Valence quark CSV contributions, x5q^r{x) vs. x. Solid line: x8u^] dashdot line: xSd^. Calculated by Rodionov et al. [13] using MIT bag model wavefunctions, evolved to Q^ = 10 GeV^ T h e NuTeV group has measured NO and CC u and u cross sections on iron [16] [for more details^ see talk by K. McFarland in these proceedings]. They obtained the N C / C C ratios R"" = 0.3916 =b 0.0007 and R^ = 0.4050 =b 0.0016, from which they extracted sin^ 0^ = 0.2277 =b 0.0013 (stat) 9 (syst). This value is three s t a n d a r d deviations above the measured value for the Weinberg angle obtained from electroweak (EW) processes near the Z pole, sin^ 0^ = 0.2227=b0.00037 [17]. Such an effect can be interpreted as a 1.2% decrease in the lefthanded coupling of light valence quarks to the weak neutral current. Davidson et al. [18] have examined possible contributions from "new physics" beyond the Standard Model. It is extremely difficult to find new particles t h a t fit NuTeV without violating other experimental constraints, as several observables are constrained by very precise measurements, at or near the 0 . 1 % level, in experiments near the Z pole [17]. Even modest success in removing the NuTeV anomaly, while leaving all other measurements within Icr, can be achieved only with new particles whose masses, numbers, and couplings are very finely t u n e d  socalled "designer particles." Because of the serious difficulties in explaining the NuTeV result with particles outside the Standard Model, attention has focused on QCD effects within the Standard Model. There are many Q C D effects t h a t must be taken into account [19]. For example, there is a correction due to the excess neutrons in Fe (iron is not an isoscalar target). Although the correction is large, it has been taken into account by NuTeV and should be well under control. Radiative corrections, which affect only CC reactions, are also substantial. These effects were calculated with a standard radiative correction model [20]. Recently Diener et al. have recalculated the radiative corrections [21]. However, their calculation has yet to be incorporated into a reanalysis of the NuTeV data. T h e NuTeV group has also corrected for nuclear effects on the P D F s . There is still some uncertainty in the magnitude of such corrections; in particular, at present it is generally assumed t h a t nuclear effects are identical for v and chargedlepton DIS. Hirai
87
et al. are currently calculating nuclear effects in neutrino reactions [22]. Finally, there is a possible contribution from a strange quark m o m e n t u m asymmetry. T h e production of oppositesign dimuons in u and u reactions allows a separate extraction of s and s P D F s [23,24]. A difference in total m o m e n t u m carried by s and s would affect the NuTeV result. Currently this has been analyzed by two groups; the C T E Q group extracts the strange P D F s in a global fit to high energy d a t a [25], while the NuTeV group has analyzed the dimuon production cross Sects. [26,27]. At the moment the two results appear to disagree. T h e C T E Q analysis favors a positive m o m e n t u m asymmetry iS'v = {x{s — s)), which would remove roughly 1/3 of the NuTeV anomaly, while the NuTeV analysis is consistent with Sy either zero or slightly negative. T h e two groups are currently collaborating on the analyses, although they b o t h agree t h a t strange quark effects alone cannot remove the anomaly. Here we will review the effects of isospin violation on the NuTeV anomaly. T h e correction to the P W ratio arising from isospin violation in the P D F s has the form AR
sm
J 2(/7v + L>v)'
(6)
Only valence quarks contribute to (6), and the correction depends on the second moment of valence P D F s , where Q {x{q — q)). T h e numerator of (6) is equal to the mom e n t u m asymmetry between u p quarks and down quarks in an isoscalar nucleus, i.e., Ul^\U^ — {D^\D^). However, estimates based on the P W ratio do not accurately predict contributions to the NuTeV result. T h e NuTeV group measures the N C / C C ratios R^ and R^. Since these ratios have different cuts and acceptance corrections, one cannot simply combine t h e m as in (5). To obtain the magnitude of a given effect on the NuTeV result for the Weinberg angle, it is necessary to fold t h a t effect with f u n c t i o n a l generated by NuTeV [27]. Thus, sea quark CSV makes a correction to the NuTeV extraction of the Weinberg angle, although it is much smaller t h a n t h a t from valence quark CSV. Using the bestfit M R S T values for sea quark and valence quark CSV, would remove roughly 1/3 of the NuTeV anomaly. T h e value K = — 0.6, within the 90% confidence limit found by M R S T , would completely remove the NuTeV anomaly, while the value K = + 0 . 6 would double the discrepancy. T h e M R S T results show t h a t isospin violating P D F s are able to completely remove the NuTeV anomaly in the Weinberg angle, or to make it twice as large, without serious disagreement with any of the d a t a used to extract quark and gluon P D F s . T h e model CSV predictions t h a t we discussed earlier suggest t h a t isospin violating corrections would tend to decrease the NuTeV anomaly for the Weinberg angle. Both the Rodionov [13] and Sather [14] theoretical models would remove about 1/3 of the NuTeV anomaly. There are other models t h a t predict substantially smaller CSV effects on the NuTeV result [27,28,29], but all theoretical predictions are well within the phenomenological limits estabhshed by M R S T . The magnitude of CSV effects al
J.T. Londergan: Current status of parton charge symmetry lowed by the M R S T fit makes isospin violation one of the only viable explanations for the NuTeV anomalous value for the Weinberg angle. If CSV effects are sufficiently large to remove the Weinberg angle anomaly, such effects should be visible in various other experiments. Several possible experiments to test parton CSV were reviewed by Londergan and T h o m a s [30]. We briefly review three such possibilities. T h e first would be a comparison of DrellYan (DY) reactions from charged pions interacting with an isoscalar target. Comparison of, say, 7r+D and 7r~D DY reactions would be sensitive to the presence of p a r t o n CSV. A study of these DY reactions [31] predicted CSV effects of about 2% in magnitude. Another experiment t h a t could detect CSV effects would be semiinclusive deep inelastic scattering (SIDIS) on an isoscalar target. A study of semiinclusive 7r+ and TT" leptoproduction on deuterium [32] predicted measurable effects from CSV. However, the ability to measure CSV effects in SIDIS reactions requires accurate knowledge of "favored" and "nonfavored" fragmentation functions. In the studies of b o t h DY and SIDIS reactions, the CSV effects were three times smaller t h a n those necessary to explain the NuTeV anomaly. If isospin violating effects are really the explanation of the NuTeV effect, b o t h of these reactions should produce effects at the several percent level. This is currently under investigation. A third possible test of parton isospin violation would be the measurement of W asymmetries in highenergy p — D reactions. This could be carried out at R H I C if deuteron beams were available [33]. We are currently investigating the feasibility of this reaction, and the asymmetries t h a t would be allowed by M R S T phenomenological fits including CSV. In conclusion, despite recent progress in constraining p a r t o n isospin violation, experimental d a t a still allows p a r t o n CSV terms at the several percent level. This has been demonstrated by the M R S T global fit t h a t incorporates isospin violation, although the form of the CSV terms was fixed in their global fit. It is clearly of great interest to investigate this issue experimentally, either to decrease the allowed upper limits on isospin violating P D F s , or to measure isospin violating effects t h a t might explain the anomalous NuTeV value for the Weinberg angle. Theoretical work cited here was carried out with A.W. Thomas. T h e author t h a n k s W. Melnitchouk, K. McFarland, S. Kretzer, F . Olness, WK Tung and R. Thorne for useful discussions.
References 1. G A . Miller, B.M.K. Nefkens, I. Slaus: Phys. Rep. 194, 1 (1990)
2. E.M. Henley, G.A. Miller in Mesons in Nuclei, eds. M. Rho, D.H. Wilkinson (NorthHolland, Amsterdam 1979) 3. J.T. Londergan, A.W. Thomas, arXiv:hepph/0407247 4. CCFR Collaboration, W.G. Seligman et al.: Phys. Rev. Lett. 79, 1213 (1997) 5. NMC Collaboration, P. Amaudruz et al.: Phys. Rev. Lett. 66, 2712 (1991); Phys. Lett. B 295, 159 (1992) 6. NMC Collaboration, M. Arneodo et al.: Nucl. Phys. B 483, 3 (1997) 7. CCFR Collaboration, U.K. Yang et al.: Phys. Rev. Lett. 86, 2742 (2001) 8. CDF Collaboration, F. Abe et al.: Phys. Rev. Lett. 8 1 , 5754 (1998) 9. MRST Collaboration, A.D. Martin et al.: arXiv:hepph/0308087 10. C.J. Benesh, J.T. Londergan: Phys. Rev. C 58, 1218 (1998) 11. E605 Collaboration, G. Moreno et al.: Phys. Rev. D 43, 2815 (1991) 12. A.I. Signal and A.W. Thomas: Phys. Lett. B 191, 205 (1987); Phys. Rev. D 40, 2832 (1989) 13. E. Rodionov, A.W. Thomas, J.T. Londergan: Mod. Phys. Lett. A 9, 1799 (1994) 14. E. Sather: Phys. Lett. B 274, 433 (1992) 15. E.A. Paschos, L. Wolfenstein: Phys. Rev. D 7, 91 (1973) 16. NuTeV Collaboration, G.P. Zeller et al.: Phys. Rev. Lett. 88, 091802 (2002) 17. D. Abbaneo et al.: arXiv:hepex/0112021 18. S. Davidson, S. Forte, P. Gambino, N. Rius, A. Strumia: JHEP 202, 037 (2002) 19. J.T. Londergan: arXiv:hepph/0408243 20. D.Yu. Bardin, V.A. Dokuchaeva: report JINRE2 86260, unpublished 21. KP.O. Diener, S. Dittmaier, W. Hollik: Phys. Rev. D 65, 073005 (2004) 22. M. Hirai, S. Kumano, TH. Nagai: arXiv:hepph/0408023 23. CCFR Collaboration, A.O. Bazarko et al.: Z. Phys. C 65, 189 (1995) 24. NuTeV Collaboration, M. Goncharov et al.: Phys. Rev. D 64, 112006 (2001) 25. S. Kretzer: arXiv:hepph/0405221; S. Kretzer et al.: arXiv:hepph/0312322; F. Olness et al.: arXiv:hepph/0312323 26. K. McFarland: these procedings 27. NuTeV Collaboration, G.P. Zeller et al.: Phys. Rev. D 65, 111103 (2002) 28. C.J. Benesh, T. Goldman: Phys. Rev. C 55, 441 (1997) 29. E.G. Cao, A.I. Signal: Phys. Lett. B 559, 229 (2003) 30. J.T. Londergan, A.W. Thomas: in Progress in Particle and Nuclear Physics, Volume 41, p. 49, ed. A. Faessler: (Elsevier Science, Amsterdam, 1998) 31. J.T. Londergan, G.T. Garvey, G.Q. Liu, E.N. Rodionov, A. W. Thomas: Phys. Lett. B 340, 115 (1994) 32. J.T. Londergan, A. Pang, A.W. Thomas: Phys. Rev. D 54, 3154 (1996) 33. C. Boros, J.T. Londergan, A.W. Thomas: Phys. Rev. D 59, 074021 (1999)
Eur Phys J A (2005) 24, s2, 8992 DOI: 10.1140/epjad/s2005040199
EPJ A direct electronic only
Pionnucleon interaction and the strangeness content of the nucleon M.E. Sainio ^ Helsinki Institute of Physics, P.O. Box 64, 00014 University of Helsinki, Finland ^ Department of Physical Sciences, University of Helsinki, Finland Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. A brief review of the pionnucleon sigmaterm is given. PACS. 13.75.Gx Pionbaryon interactions  14.20.Dh Protons and neutrons
1 Introduction
where the numerator is proportional to the octet breaking piece in the hamiltonian. To first order in SU(3) breaking we have now
Sigmaterms are proportional to the matrix elements {A\mqqq\A)
;q = u,d,s
; A =
7r,K,N
m
of scalar quark currents in the framework of Q C D . These matrix elements are of interest, because they are related — to the mass spectrum, — to scattering amplitudes through Ward identities, — to the strangeness content of A, — to quark mass ratios. In the following the status of the TTN system is considered and the implications to the strangeness content in the nucleon are outlined.
rris + Trtz; — 2mN
TTis — rh
1—y
T h e pionnucleon sigmaterm is a measure of explicit chiral symmetry breaking in Q C D and it is defined as —
1
(J =—{p\uu\dd\p),
(6)
Ml
m
in terms of the kaon and pion masses. Chiral perturbation theory ( C h P T ) allows us to determine the combination
= m{p'\uu^dd\p),
t = {p'
p)'^,
'^{v\ss\p) {p\uu\ dd\p)
— 2ss\p)
1—V
36=b7 MeV, 0{ml)
[2]
33=b3MeV, ©(m^)
[3],
(2) where the difference in the last two determinations is the regularization m e t h o d used, dimensional regularization [2] or cutoff [3]. W i t h the help of the FeynmanHellmann theorem the sigmaterm can be extracted from the nucleon mass
(3) (7 =
(the OZI rule would imply 7/=0). Algebraically a can be written in the form rh {p\uu\dd
26 MeV (leading order)
(1)
i.e. cr = cr(t = 0). T h e nucleon mass is denoted by m. T h e strangeness content of the proton can be defined as
2m
(7)
35=b5 MeV, 0 ( m g / ^ ) [1]
771= {mu\md),
which is the t = 0 value of the nucleon scalar form factor a{t)
y
(5)
from the baryon spectrum. For (7 we have:
2 T h e TTN sigmaterm
u'(j{t)u
l ~ y '
where the quark mass ratio takes the value
(7 = a ( l  y)
777/
26MeV
^ dm m—:r. am
(8)
Equivalently, employing M ^ = 2mB^ (4)
M'
dm
(9)
M.E. Sainio: Pionnucleon interaction and the strangeness content of the nucleon
90
where B is the scalar vacuum condensate. T h e quark mass expansion of the nucleon mass [4] m = mo + kiM'^ + k2M^ + feM^ In
and AR is the remainder, which is formally of the order M^ [6]. To oneloop in C h P T {0{q^)) [7]
M' ^
AR
= 0.35 MeV.
Oneloop in H B C h P T (0(q^)) ^k^M^^O(M^)
(17)
gives the upper limit [8]
(10) ARC^2
MeV
(18)
yields for a and here it is notable t h a t no logarithmic contribution to order M^ appears. This allows us to write
M2
a = kiM'^ + k2M^
+ k3M^{2 In ^ + 1} mt{ + 2/c4M^ + 0 ( M ^ ) .
Uc^a(2M^).
(19)
(11)
Here the factors ki contain the lowenergy constants appearing in the respective chiral order. Numerically
W h a t remains to be fixed in order to determine the a is the form factor difference A,
=
cj(2Ml)cj{^).
cr = (75  23  7 + 0) MeV = 45 MeV,
(20)
C h P T to one loop gives [7] where the O ( M ^ ) term, /ci, has been fixed by taking a to have the value 45 MeV [5].
A^c^b
(21)
MeV.
Dispersion analysis yields [9]
3 The TTN amplitude
A^ = 15.2 =b 0.4 MeV.
To relate the sigmaterm discussion to the scattering information the s t a n d a r d representation for the TTN amplitude is adopted T^N = u'[A[iy, t) + ]p^{q + q')^B{v,
t)]u.
(12)
T h e definition of the crossing variable is s —u t (13) ^ = ~; =cj + —, 4m 4m where LJ is the pion laboratory energy. T h e amplitude D is defined as D{iy,t) = A{v,t)
+
vB{v,t)
(14)
and its imaginary part can be related to the total cross section through the optical theorem, Iml}(cc;,t = 0) = /ciab cr. T h e isoscalar ( + ) and isovector () amplitudes D^ can be written in terms of the amplitudes in the physical channels as
(22)
Becher and Leutwyler obtain [4] A^ = 14.0 MeV + 2M%,
(23)
where 62 is a renormalized coupling constant appearing in the C^^ C^ lagrangian. T h e constant 62 is expected to be small [4].
5 The Z'term Inside the Mandelstam triangle it is convenient to employ the subthreshold expansion [10], where D^ is expanded in powers of v^ and t
T h e curvature t e r m Ajj is defined as U = F ^ ( d + + 2 M ^ 4 , ) ^AD
= IJd^AD
(25)
and it is dominated by the TTTT cut giving the result [9] D^ =
(15)
Chiral symmetry allows us to write at the ChengDashen point, i.e. at (i/ = 0,t = 2 M ^ ) , = F 2 5 + ( Z / = 0, t = 2Ml)
= 11.9
6 MeV.
(26)
T h e linear part Ud is a sensitive quantity due to the cancellation of the (ioQ and (ij^ pieces in
4 A lowenergy theorem
r
AD
= cr{2Ml) + AR,
(16)
where F^^ is the pion decay constant, D^ is the isoscalar Damplitude with the pseudovector Born t e r m subtracted
rrf(A) = (  9 1 . 3 + 138.8) MeV ~ 48 MeV
(27)
^ ^ ( B ) = (  9 4 . 5 + 144.2) MeV 2^ 50 MeV
(28)
corresponding to the two solutions (A and B) discussed in [5]. These numbers lead to i7 ~ 60 MeV, which is consistent with the old result of Koch [11] T = 64 =b 8 MeV based on hyperbolic dispersion relations.
M.E. Sainio: Pionnucleon interaction and the strangeness content of the nucleon 50
6 The strangeness content of the nucleon
40
P u t t i n g all these pieces together leads to a determination of the strangeness content of the proton
30 L
(29)
SARA^
35MeV
10 0
and numerically with the solution A
O
(60  2  15) MeV,
(30)
1y
10
J
A
f ^
40 50
1 \j 0,5
\
V 1 k (GeV/c)
1
1
J
1.5
2
1
Fig. 1. The real part of the C^amplitude. The crosses refer to the tabulated values in [10]
T h e analysis discussed in Sect. 5 was based on the KH80 solution of the Karlsruhe group [12]. T h e d a t a basis used there contained mainly premesonfactoryera d a t a and, therefore, it is of great interest to perform a new analysis with the new d a t a in the spirit of the Karlsruhe group incorporating fixedt constraints. This would hopefully help in fixing the value of H more accurately. In the forward direction it is feasible to solve the dispersion relations directly, but for t < 0 it is more practical to use the expansion m e t h o d [10]. E.g., for the C^ amplitude
{C = A +
1
if
11 f
f 1
7 Partial wave analysis
'
20 30
which gives 7/ :^ 0.2 with a sizeable error. This value of y corresponds to about 130 MeV in the proton mass being due to the strange sea.
^
1 n
20 =
91
250
v/{lt/4.m^)B) N
C+{u,t)
= C + ( i / , i ) + H{Z,t)
^
C+Z,
n=0
where C^(z^, t) is the Born term, the function H is adjusted to the asymptotic behaviour of the amplitude and Z is the conformal mapping
Z{u^t)
=
a
^Ath
« + V^4  ^^
(31)
where Vth = M^^ + t/^m and a is a real parameter. T h e convergence and smoothing is taken care of by a convergence test function
1 k (GeV/c)
1.5
Fig. 2. The 7r~^p total cross section but indications are [14] t h a t the number could be 2030 % larger t h a n the numbers quoted above.
8 The relation S ^H^ threshold
N
xl
\Y,cl{n+lf, n=0
which is one component in the x^ expression to be minimized. Other contributions include 'XJ:,ATA ^ ^ ^ Xpw^ where the latter is calculated from the previous iteration of the partial wave solution. To demonstrate the working of the expansion method at t = 0 with N = 4 0 , Figs. (1) and (2) display Re C+(cj, t = 0) and crj+ . For the real part of the C+amplitude there are three d a t a points at low energy as input and they fix the subtraction constant appearing in the dispersion relation for the (7+. T h e V P I / G W U group has recently published a partial wave analysis [13], which does employ fixedt constraints. T h e publication does not, however, give a value for the U,
T h e issue of relating the U to the values of threshold parameters is an old one [15]. In general, U can be expressed in terms of the threshold parameters [16] E = F^[L{ai,T)
+ {l + ^)Tj+]+6, (32) m where L{ai,T) is a linear combination of the threshold parameters a/, r is a free parameter to single out individual scattering lengths and J + is the integral over the isoscalar combination of the total cross section. T h e remainder, S, contains contributions from the Born term, the A and the loop corrections. T h e approach of Altarelli et al. [15] corresponds to choosing r = — 1, but without loops. However, at present, one has to rely on dispersion methods to extract the threshold parameters anyway, so the value of any such formula is limited.
M.E. Sainio: Pionnucleon interaction and the strangeness content of the nucleon
92
References
9 Lattice results C h P T permits a study of the quark mass dependence of the nucleon mass. This makes it possible to have a connection to the lattice data, where, currently, only unphysically high quark masses can be dealt with. New accurate d a t a from the CPPACS, J L Q C D and Q C D S F  U K Q C D collaborations (dynamical quarks, two flavours) give [17]
1. 2. 3. 4. 5.
cr = 49 =b 3 MeV
(33) 6.
to 0{q^) in C h P T . Another approach including the leading nonanalytic and nexttoleading nonanalytic behaviour yields [18]
7. 8.
0 = 35  73 MeV.
(34) 9.
10 Conclusions
10.
T h e challenge at present seems to be in determining U. T h a t involves a number of questions — one has to deal with conflicting sets of data, — one has to rely on the Tromborg [19] formalism for the electromagnetic corrections even though there are indications [20] t h a t further improvements in this sector should be incorporated, — the extrapolation from the lowenergy region to the ChengDashen point could, to some extent, be sensitive to the dwaves, which otherwise cannot be fixed with the lowenergy scattering information [21]. T h e new direction with the lattice calculations is gradually getting very interesting as far as the sigmaterm is concerned. However, further improvements, i.e. smaller 772gvalues, will still be needed.
11. 12. 13.
Acknowledgements. I wish to thank P. Piirola for providing the figures and A.M. Green for useful comments on the manuscript. Support from the Academy of Finland grant 54038 and the EU grant HPRNCT200200311, EURIDICE, is acknowledged.
14. 15. 16.
17. 18. 19. 20. 21.
J. Gasser, H. Leutwyler: Phys. Reports 87, 77 (1982) B. Borasoy U.G. MeiBner: Ann. Phys. 254, 192 (1997) B. Borasoy: Eur. Phys. J. C 8, 121 (1999) T. Becher, H. Leutwyler: Eur. Phys. J. C 9, 643 (1999); JHEP 6, 017 (2001) J. Gasser, H. Leutwyler, M.E. Sainio: Phys. Lett. B 253, 252 (1991) L.S. Brown, W.J. Pardee, R.D. Peccei: Phys. Rev. D 4, 2801 (1971) J. Gasser, M.E. Sainio, A. Svarc: Nucl. Phys. B 307, 779 (1988) V. Bernard, N. Kaiser, U.G. Meii3ner: Phys. Lett. B 389, 144 (1996) J. Gasser, H. Leutwyler, M.E. Sainio: Phys. Lett. B 253, 260 (1991) G. Hohler: PionNucleon Scattering, LandoltBornstein Vol.I/9b2 (Springer, Berlin, 1983) R. Koch: Z. Phys. C 15, 161 (1982) R. Koch, E. Pietarinen: Nucl. Phys. A 336, 331 (1980) R.A. Arndt, W.J. Briscoe, LL Strakovsky, R.L. Workman, M.M. Pavan: Phys. Rev. C 69, 035213 (2004) M.M. Pavan, R.A. Arndt, LL Strakovsky, R.L. Workman: TTN Newsletter 16, 110 (2002) G. Altarelh, N. Cabibbo, L. Maiani: Phys. Lett. B 35, 415 (1971); Nucl. Phys. B 34, 621 (1971) J. Gasser: Proc. 2nd Int. Workshop on TTN Physics, Los Alamos 1987, ed. W.R. Gibbs and B.M.K. Nefl^ens: Los Alamos report LA11184C 266 (1987) M. Procura, T.R. Hemmert, W. Weise: Phys. Rev. D 69, 034505 (2004) D.B. Leinweber, A.W. Thomas, R.D. Young: heplat/0302020 (2003) B. Tromborg, S. Waldenstr0m, L 0verb0: Phys. Rev. D 15, 725 (1977) N. Fettes, U.G. MeiBner: Nucl. Phys. A 693, 693 (2001) J. Stahov: hepph/0206041 (2002)
Eur Phys J A (2005) 24, s2, 9396 DOI: 10.1140/epjad/s2005040204
EPJ A direct electronic only
Strange form factors of the nucleon in the chiral quarksoliton model A. Silva^ ^, D. Urbanoi'2, H.C. K i m ^ and K. Goeke^ Centre de Fisica Computacional, Departamento de Fisica da Universidade de Coimbra, P3004516 Coimbra, Portugal Faculdade de Engenharia da Universidade do Porto, R. Dr. Roberto Frias s/n, P4200465 Porto, Portugal Department of Physics, Pusan National University, 609735 Pusan, Republic of Korea Institut fiir Theoretische Physik II, RuhrUniversitat Bochum, D44780 Bochum, Germany Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The results for the strange electromagnetic and axial form factors in the chiral quarksoliton model are reviewed. The roles of a new quantization method and of the meson asymptotic behaviour are discussed. Predictions for the A4 and GO experiments are presented. PACS. 12.40.y Other models for strong interactions tromagnetic form factors
14.20.Dh Protons and neutrons  13.40.Gp Elec
1 Introduction
2 The chiral quarksoliton model (CQSM)
T h e question of the strange content of the nucleon is a very important one in the understanding of hadron structure, which eventually is to be settled by experiment. Indeed, on one hand there is no obvious theoretical reason why strange s's pairs should not contribute to the nucleon properties. On the other hand the theoretical mechanism and the actual contribution of strange quarks are not yet understood as well. Early results about the strange quark bilinears in the nucleon included the analysis of the sigma t e r m [1] and polarized DIS. These first results indicated strange quark nonvanishing contributions to the nucleon's mass and spin, respectively. T h e strange vector currents associated with electromagnetic form factors were discovered to be accessible by means of weak neutral current experiments [2,3, 4]. For reviews see [5,6]. Due to its importance, the question of the strange form factors has been addressed before in the context of the chiral quarksoliton model (CQSM) [7,8]. This work reviews the CQSM results [9,10] for the strange electromagnetic form factors and presents also the strange axial form factor. These form factors are obtained using a new quantization m e t h o d [11] and studied phenomenologically in terms of the effects of the asymptotic behavior of the meson fields.
T h e chiral quarksoliton model (CQSM) has been applied successfully to the description of many baryon observables, b o t h in flavours SU(2) and SU(3). See [12,13] for reviews. T h e Lagrangian of the CQSM reads,
^ Financial support from the organization of PAVI04 and from the Center for Computational Physics (CFC) for the participation in PAVI04 is kindly acknowledged. The work of A.S. has been partially supported by the grant PRAXIS XXI BD/15681/98.
C =
tlj{x){i^
(1)
where m = diag(m^, 772^^,777,5) is the current quark mass matrix and M the constituent quark mass. M is the only free parameter in the model. T h e main property of the model in relation with hadron physics is chiral symmetry and its spontaneous breaking. T h e auxiliary chiral fields JJJ5 = g*7 [j^ (^i^^ jr^^y ]^Q interpreted as physical meson fields and, thus, the Lagrangian (1) corresponds to a theory of constituent quarks interacting through the exchange of mesons. However, the field description embodied in (1) is not renormalizable. A regularization method is therefore necessary to completely define the model. This work uses the propertime regularization, which introduces one cutoff parameter into the model. T h e remaining parameters of the model are the current quark masses. In this work isospin breaking effects have been neglected and therefore the current mass parameters are the equal current masses (m) of the u and d quarks and the current mass (rris) of the strange quarks. In order to fit these parameters, one calculates the pion and kaon masses as well as the pion decay constant in terms of the parameters of the model. Equating the model results for these quantities with their experimental values allows to fix the parameters of the model: the current nonstrange quark masses, the strange quark mass and the cutoff come out around 8 MeV, 164 MeV and 700 MeV, respectively.
A. Silva et al.: Strange form factors of the nucleon in the chiral quarksoHton model
94
3 The model baryon state
T h e collective coordinate Hamiltonian is
In broad terms, the construction of the model baryon states necessary to compute baryon matrix elements are obtained in a projection after variation procedure, with the variation referring to the mean field solution. Taking the large Nc limit of the correlation function of two baryon currents and neglecting the strange quark mass lead to an effective action from which the mean field solution is obtained. At this level the effective action is proportional to the energy of the mean field, S^^ [U] ~ A^ [t/], which is a functional of the chiral fields U: Ai[U] = A^c^v [U] \ es [U], with ey the energy of the valence level, occupied by Nc quarks, and eg the regularized energy of the Dirac sea. Both result from the solution of the Dirac oneparticle problem e^ [U] = {n\ h (U) \n) with h (U) = ij^f'di
+ j^MU^'
+
fnj^
U = e^Pir) E L i ^" 1
(3)
^Kjr{A)
T h e quantization of the mean field is achieved in the context of collective coordinates. T h e collective coordinates are the orientations of the soliton in configuration as well as in flavour spaces (related by the hedgehog) and the position of the center of mass. These coordinates are most suited to describe the rotational zero modes, i.e. unconstrained large amplitude motion, which should be treated exactly. T h e model calculation neglects modes normal to the zero modes, which are subjected to restoring forces and hence suppressed. T h e time dependent solutions are constructed on the basis of the restriction to the zero modes by the ansatz^ A{t)Uc{x)A\t)
(4)
with the collective coordinates contained in A. In the laboratory system the Dirac operator becomes D{U)
= A \D{UC)
+ A U + j^A^mAl
= vs{r^'^'D^SiA).
(7)
with u = ( y , T , T 3 ) and ly' = {Y',J,J^) standing for the q u a n t u m numbers of the baryon state {Y' = —1). T h e symmetry breaking part of the Hamiltonian leads to a representation mixing, in first order perturbation theory, from which the baryon wave function \B^%Y^ picks terms from higher order representations [14]. 3.2 Asymptotics of the meson fields and symmetry conserving quantization T h e question of the asymptotics of the meson fields originates in the fact t h a t the embedding (3) forces all the meson fields to have a common asymptotic behavior. This is not satisfactory in the SU(3) case due to the large mass difference between pions and kaons. In order to have some information on how large these effects may be, one exploits in this work the fact t h a t there is no prescription to fix the multiples of the unity mass matrices in (2) and in the symmetry breaking piece 6m of (5): l^D{Uc)
3.1 Quantization
U{x) ^
(6)
i.e. it is composed by a flavour symmetry conserving (sc) and a symmetry breaking (sb) pieces. While the first is treated exactly, the second is treated perturbatively. T h e wave functions of the symmetry conserving part are the Wigner SU(3) matrices. For the octet
(2)
In order to solve such problem one further restricts the mesons to the hedgehog shape, TT r = P (r)h  r, with a profile function P(r) vanishing at large distances. T h e minimization of the action corresponds to a profile function Pc{r) which represents the mean field soliton. T h e formalism in flavour SU(3) is built upon the embedding of the SU(2) hedgehog into an SU(3) matrix [14]:
H = M{Uc) + H[coll r ^ c o l l '
A^^,
(5)
which clearly shows the two expansion parameters used in this work: the angular velocity A'^A (t) = i i 7 g A " / 2 and the strange mass related parameter J m , discussed below, in Sect. 3.2, in connection with the asymptotic meson behaviour effects. Not showing translational zero modes explicitly.
+ A^5mA
= i^di
+ MU^'
+ m + A^5mA
(8)
where Sm = Mi \ MgA^, with M i = (m^  m ) / 3 and Ms = (m — ms)/V^. In this case one may obtain larger mass asymptotics of the meson fields, by, in (8), increasing m at the expense of a lower Mi. T h e pion asymptotics will be denoted in the following by /x ^ 7r(~ 140 MeV) and the kaon asymptotics by // ^ i ^ ( ~ 490) MeV). T h e symmetry conserving quantization [11] used in this work avoids the problem which is encountered in this model by using the ansatz (4) to compute the relation between isospin (T) and spin (S) operators. While the result from (4) reads
T 'a
^S (^) Jp
V S D i l ' (A) ^ ,
(9)
the correct relation implies I2 = 0. T h e symmetry conserving quantization identifies the terms like I2 in the expressions for observables on the basis of the quark model limit of the CQSM.
4 Strange electromagnetic form factors T h e knowledge of the form factors of the octet vector currents is enough to provide the form factors for each flavour: V^ = uj^u
+ d/^d + sf^s ,
(10) (11)
df^d  257^5.
(12)
A. Silva et al.: Strange form factors of the nucleon in the chiral quarksoHton model Table 1. Strange magnetic moment (in n.m.) and electric and magnetic radii (in fm^). The constituent quark mass is 420 MeV and the strange quark mass 180 MeV. TT and K stand for the two asymptotic descriptions of the mesons as discussed in Sect. 3.2. TV
K
0.220 0.074 0.303
0.095 0.115 0.631
M
(r^>fe Ms {r^M
95
0.15
0.10
0.05
0.00
0.4
0.6
1.0
Q^ [GeV^] Fig. 2. Strange magnetic form factor in nuclear magnetons. Conventions and model parameters as in Table 1.
positive value, similarly to what is observed in other models by maintaining just the SU(3) structure [16]. 0.2
0.4
0.6
0.8
1.0
4.1 The SAMPLE, HAPPEX, A4, and GO experiments
Q2 [GeV2] Fig. 1. Strange electric form factor of the nucleon. Conventions and model parameters as in Table 1.
The experimental value for the magnetic form factor obtained by the SAMPLE collaboration [17] is ^1^(^2 = 0.1) = +0.37=b0.20=b0.26zb0.07 (n.m.). (14)
In this work the form factors for these currents were first obtained for the two asymptotic behaviours. The effects of the asymptotics was then studied at the level of baryon form factors by combining flavour form factors with pion asymptotics for nonstrange quarks with kaon asymptotics for strange form factors, i.e. by taking ^UB{TT) (
,dB(7r)^
ysB{K),
Gi ""E^uiQ ) — ^E,M (Q ) + ^E,M (Q ) + ^E,M
(Q ) '
(13) The most significant consequences of this ansatz for the baryon octet electromagnetic form factors is the improvement in the overall description of magnetic form factors and the neutron electric form factor. This supports the expectation that the strange form factors are better described in terms of kaon (// ^ K) rather than in pion (// ^ TT) asymptotics. The results for {ii ^ TT) are nevertheless always presented. The strange magnetic moment and the radii are presented in Table 1 and the strange electric and magnetic form factors in the CQSM are shown in Fig. 1 and Fig. 2, respectively. A particular aspect of these results is the positive value for the strange magnetic moment, which seems to be at variance with many theoretical calculations based in different approaches. The CQSM seems nevertheless to be able to accommodate higher values of this quantity, up to 0.41 n.m., as is found in the "model independent analysis" of [15]. This pinpoints the SU(3) structure of the rotational and mass corrections as the reason for such
The CQSM results of this work underestimates this experimental result, still falling however within the error bar, shown in Fig. 2 at Q^ = 1. The HAPPEX result [18] {G'E + 0 . 3 9 2 G  ^ ) ( Q 2 = 0.477) = 0.014 =b 0.030,
(15)
is, on the contrary, overestimated in the CQSM, with the nearest result (kaon asymptotics) being 0.073. Using the necessary form factors, one may study the combination G%{Q^) + (3{Q'^, 0)GI^{Q^), with /3(Q^ 0) = TG^j;i/eGl\ T = Q^/{AM%), 61 = l + 2(l + r)tan2(^/2), for the values of 0 and Q^ used in current or near future experiments. Figure 3 compares the model predictions with the recent forward data. ( G  + 0.225G1^)(Q2
(C; + 0.106G^)(O2
:0.23) = 0.039 zb 0.034 0.10) = 0.074 zb 0.036,
(16) (17)
with (16) from [19] and (17) from [20], by the A4 Collaboration and gives predictions for the backward angle. Figure 4 gives predictions for the GO experiment.
5 Strange axial form factor The axial currents of interest are expressed in term of flavour components in the same way as (12). The axial form factors for the octet axial currents have also been computed in the model with the same accuracy of the
A. Silva et al.: Strange form factors of the nucleon in the chiral quarksohton model
96
0.50
_G%m'+P{Q\e)Gl,{Q']
0.40
0 = 145
0.30
/
0.00 I
'A4

y ^
1
1
'
1
1
1
1
1
1
0.05 =r ' ^ ^."""""""^
y^ yU^^^^
1
/i ^
/
0.0ft1/^

TT
H ^
.
"
0.20 0.10
.
0.10
71

ii^K
0.15
^...<
1
iL
0.2
1
1
1
1
0.4
1
1
0.20
1
0.8
0.6
1.0
^2 rr^^AA2i Q2 [GeV^:
1
.0
1
0.2
1
1
0.4
1
1
0.6
1
1
0.
1
1.0
Q^ [GeV^]
for the A4 Fig. 3 . Predictions of G%(q^) + P{Q^,0)G'M{Q^) experiment. Conventions and model parameters as in Table 1.
Fig. 5. Strange axial form factor. Conventions and model parameters as in Table 1
0.20
G%m+m\o)GUQ^
GO
e = 108°
0.15
expected below 15 %. T h e phenomenological observed effects related to the asymptotic behaviour of the meson fields requires nevertheless a more rigorous t r e a t m e n t .
References
Q2
[GeV^
Fig. 4. Predictions for the GO experiment. Conventions and model parameters as in Table 1.
Table 2. As and strange axial r.m.s. Conventions and model parameters as in Table 1. TT
As ,2\l/2
0.086 0.554
K 0.075 0.172
electromagnetic ones. T h e axial constants and axial dipole mass reproduce reasonably well the available experimental data. T h e results for As = G\{Q^ = 0) and the strange axial r.m.s. are shown in Table 2 and the strange axial form factor is presented in Fig. 5. T h e CQSM results for As seem to fall below a recent analysis of DIS data, which yields As =  0 . 1 4 zb 0.03 [21].
6 Conclusions T h e CQSM results for the strange form factors are not ruled out by the presently available experimental data. T h e results are stable from the point of view of the model parameters, with higher order corrections in Nc and Sm
1. J. Gasser, H. Leutwyler, M.E. Sainio: Phys. Lett. B 253 (1991) 252 2. D.B. Kaplan, A. Manohar: Nucl. Phys. B 310 (1988) 527 3. D.H. Beck: Phys. Rev. D 39 (1989) 3248 4. M.J. Musolf et a l : Phys. Kept. 239 (1994) 1 5. W.M. Alberico, S.M. Bilenky, C. Maieron: Phys. Rept. 358 (2002) 227 [arXiv:hepph/0102269] 6. D.H. Beck, R.D. McKeown: Ann. Rev. Nucl. Part. Sci. 51 (2001) 189 [arXiv:hepph/0102334] 7. H.C. Kim, T. Watabe, K. Goeke: Nucl. Phys. A 616 (1997) 606 [arXiv:hepph/9606440]. 8. H. Weigel, A. Abada, R. Alkofer, H. Reinhardt: Phys. Lett. B 353 (1995) 20 [arXiv:hepph/9503241] 9. A. Silva, H.C. Kim, K. Goeke: Phys. Rev. D 65 (2002) 014016 [Erratumibid. D 66 (2002) 039902] [arXiv:hepph/0107185] 10. A. Silva, H.C. Kim, K. Goeke: arXiv:hepph/0210189, to appear in EPJA 11. M. Praszalowicz, T. Watabe, K. Goeke: Nucl. Phys. A 647 (1999) 49 [arXiv:hepph/9806431] 12. C.V. Christov et al.: Prog. Part. Nucl. Phys. 37 (1996) 91 [arXiv:hepph/9604441] 13. D. Diakonov, V.Y. Petrov: arXiv:hepph/0009006 14. A. Blotz et al.: Nucl. Phys. A 555 (1993) 765 15. H.C. Kim, M. Praszalowicz, M.V. Polyakov, K. Goeke: Phys. Rev. D 58 (1998) 114027 [arXiv:hepph/9801295] 16. S.T. Hong, B.Y. Park, D.P. Min: Phys. Lett. B 414 (1997) 229 [arXiv:nuclth/9706008] 17. D.T. Spayde et al. (SAMPLE Collaboration): Phys. Lett. B 583 (2004) 79 [arXiv:nuclex/0312016] 18. K.A. Aniol et al. (HAPPEX Collaboration): Phys. Rev. C 69 (2004) 065501 [arXiv:nuclex/0402004] 19. F.E. Maas et al. (A4 Collaboration): Phys. Rev. Lett. 93 (2004) 022002 [arXiv:nuclex/0401019] 20. F.E. Maas (for the A4 Collaboration): these proceedings 21. B.W. Filippone, X.D. Ji: Adv. Nucl. Phys. 26 (2001) 1 [arXiv:hepph/0101224]
Eur Phys J A (2005) 24, s2, 97100 DOI: 10.1140/epjad/s2005040213
EPJ A direct electronic only
Strange form factors and Chiral Perturbation Theory Bastian Kubis^ Institut fiir theoretische Physik, Universitat Bern, Sidlerstrasse 5, CH3012 Bern, Switzerland Received: 15 October 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. We review the contributions of Chiral Perturbation Theory to the theoretical understanding or notquiteyetunderstanding of the nucleon matrix elements of the strange vector current. PACS. 12.39.Fe Chiral Lagrangians  14.20.Dh Protons and neutrons
1 Introduction Chiral Perturbation Theory ( C h P T ) , as t h e lowenergy effective field theory of t h e Standard Model, ought t o be predestined t o describe t h e response of t h e nucleon to t h e strange vector current: it incorporates all symmet r y constraints of t h e fundamental theory and contains the degrees of freedom relevant at low energies, Goldstone bosons (pions and kaons) as well as m a t t e r fields (nucleons). In Sect. 2, we shall point out some specific aspects of C h P T t h a t are essential for t h e following discussion. We shall reiterate in Sect. 3 why symmetry considerations alone are not suflnicient for C h P T t o be predictive for t h e leading moments of t h e strange form factors, t h e strange magnetic moment //^ and t h e strange electric radius ( r  ; g). A lowenergy theorem for t h e strange magnetic radius {rj^ g) is presented in Sect. 4, its usefulness and limitations are discussed. Some alternative regularization schemes for loop diagrams have been suggested in order to achieve improved convergence behavior of t h e chiral series, we shall say a few words on these in Sect. 5. Finally, a conclusion is given in Sect. 6.
2 Some aspects of Chiral Perturbation Theory 2.1 Goldstone bosons and counterterms C h P T [1] systematically exploits t h e farreaching consequences of chiral symmetry: it dictates t h e appearance of (pseudo) Goldstone bosons (TT, K, rj) and tightly constrains their interaction with each other as well as with external currents and m a t t e r fields. As a consequence, t h e Goldstone boson dynamics is completely calculable, while the infiuence of heavier states can be parameterized, at low energies, by polynomials. T h e polynomial coefficients (called lowenergy constants) are not numerically known a ^ Present address: HISKP (Th), Universitat Bonn, Nussallee 1416, D53115 Bonn, Germany
priori^ b u t are still far from arbitrary, as they potentially interrelate many different observables. A particularly famous example is t h e quark mass expansion of t h e pion mass: M i = 2B7fi
^3
32^2^2
[2BmY ^0{ m^ 3 \
(1)
where m = (ruu + md)/2. T h e first t e r m on t h e right constitutes t h e wellknown GellMannOakesRenner relation. T h e correction of order m^ is proportional t o t h e lowenergy constant Is t h a t can be determined from TTTT scattering. Given such independent experimental information, t h e relative size of t h e contributions t o M^ linear and quadratic in t h e quark masses is known [2]. 2.2 Power counting In t h e Goldstone boson sector of C h P T , Lorentz invariance dictates t h a t only even powers of momenta appear in t h e effective Lagrangian. T h e typical lowenergy expansion parameters therefore are M^/A^ ?^ 0.02 (for chiral SU(2)) or M]^/Al ^ 0.2 (for chiral SU(3)), where t h e chiral symmetry breaking scale is A^ = ATTF^J; ^ 1160 MeV. In contrast, due t o t h e appearance of spin, there are also odd powers of momenta in t h e effective pionnucleon Lagrangian, such t h a t t h e convergence orderbyorder is markedly slower, typical expansion parameters are M^/rriN  0.15 in SU(2) or MK/TTIN ~ 0.5 for SU(3). Obviously, there is no way we can expect C h P T t o work as well for t h e baryon sector with strangeness as it does for, say, TTTT scattering.
It is important t o understand t h e "generic" chiral orders of t h e lowest moments in t h e nucleon vector form factors, i.e. t h e orders at which polynomial contributions occur. T h e lowenergy expansion of the strange Sachs form factors is given by
0{p)
0(p^
B. Kubis: Strange form factors and Chiral Perturbation Theory
GM,S{Q^)=
^S

^(rL,.>0' + ...
o{p^)
,
Oip^
iT, i^^ ^ ^ ^
where one has Qs = 0 due t o gauge invariance for t h e strange charge. We note in particular t h a t polynomial contributions t o t h e radius terms (r^.^), {rj^ g) appear at leading and subleading oneloop order, respectively.
A,
E", S ^
Fig. 1. Diagram generating the leading contribution to {r]^^s)
3 Why ChPT cannot predict /j^s and {r^ E,s/ T h e reason why it is impossible t o predict }is a n d ( r  ; ^) from first principles in C h P T was identified several years ago [3]. There are three independent diagonal vector currents in SU(3),
o (3
i = 3, 8, 0,
4'^ = «Y7M9
(2)
where A^/^ are t h e usual GellMann matrices a n d A^ = ^ 2 / 3 d i a g ( l , 1,1). T h e three are proportional t o t h e isovector a n d isoscalar electromagnetic currents a n d t h e baryon number current, respectively. T h e electromagnetic and t h e strangeness currents are linear combinations of these.
Q ' [GeV']
Fig. 2. Extrapolation of the SAMPLE value GM,S(Q'^)
^O M^
4.1 A lowenergy theorem for {r\j g) J.
M 
/(3) , _]_ 7(8)
Jt = f^j'^'^/^'
^ (3) After
i.e. t h e response t o one component of t h e strangeness current, t h e baryon number, is going t o be completely independent of what we know from electromagnetic probes. In the effective Lagrangian language, this means t h a t wherever matrix elements of t h e electromagnetic current depend on lowenergy constants, there will be a new, independent constant for t h e strangeness current. As an example, consider t h e leading terms contributing t o t h e magnetic moments, uD/F £(2)
SrriN
(B
T h e constants 6g can be fitted alternatively t o t h e magnetic moments of proton and neutron or t o all octet moments, b u t 6g only appears in t h e strange magnetic moment. T h e same p a t t e r n emerges for all lowenergy constants, therefore instead of predicting t h e strange vector form factors, C h P T can only adjust its constants in order to reproduce experimental findings.
t h e pessimistic finding of t h e last section, how can there possibly be any lowenergy theorem for any strange vector form factor? T h e answer is, through leading nonanalytic loop effects. T h e diagram of order 0{p^) displayed in Fig. 1 generates a contribution t o t h e strange magnetic radius according t o [4]
{rl,J''
5L>2  QDF + 9 F 2 niN 487rF.
As a lowenergy constant can only contribute t o {rj^ g) at t h e next order, generating a t e r m of ( 9 ( M ^ ) , t h e t e r m in (5) is (at least formally) dominant. All t h e masses a n d coupling constants in (5) are known, hence we have a parameterfree prediction for ( r  ^ g ) , (r'M,s
Besides being a quantity of interest in its own right, t h e strange magnetic radius (r^ g) is also of high importance for t h e experimental determination of t h e strange magnetic moment fig for t h e following reason: Experimental measurements of GM,S{Q'^) always have t o be performed at finite, nonvanishing Q^, therefore one needs t o extrapolate t o GM,S{0)
= /is [4].
(6)
0.115 fm^
We can use t h e leading chiral prediction for {rj^ ^) t o extrapolate from t h e S A M P L E result [5] G'M,S(0.1 GeV^) = 0.37 0 6 7 t o t h e strange magnetic moment. /Is = 0.32
4 The strange magnetic radius
+ 0{M%) . (5)
0.20
0.26
0.07
(7)
This extrapolation is visualized in Fig. 2. Furthermore, in [7] this was combined with a second experimental result, t h e H A P P E X measurement [6] for GE,S\0.39GM,S, in order t o fix b o t h unknowns in t h e 0{p^) representation of t h e strange form factors (for fig and (r; ^)) a n d predict also t h e Q^dependence of t h e strange electric form factor, see Fig. 3. We note from t h a t figure t h a t even over a large Q^ range, t h e form factor looks essentially linear
B. Kubis: Strange form factors and Chiral Perturbation Theory STRANGE ELECTRIC FORM FACTOR
^
99
^
"—»<^<—*
Fig. 5. Resummation of relativistic corrections in the infrared regularization scheme. The crosses denote l/rriN insertions in the nucleon propagator
Q' [GeVI
Fig. 3 . Q^dependence of GE,S at 0{p^). The fi sure is taken from [7]
nexttoleading order corrections are sizeable. We therefore want to discuss here two alternative schemes to evaluate loops t h a t may resum these corrections more efficiently, or may give a fairer estimate of these contributions altogether.
5.1 Infrared regularization
Fig. 4. Extrapolation of the SAMPLE value GM,S{Q'^) to fis, using the 0{p^) value for (rj^^s) The grey band indicates a dispersiontheoretical prediction. The figure is taken from [8] and displays only very little curvature. T h e reason for this is t h a t the closest infrared singularity in this representation is the KK cut, which has a threshold far away from the physical region, and t h a t counterterms providing curvature in the polynomial part are suppressed to twoloop order.
4.2 Stability of the lowenergy theorem Even though the lowenergy theorem (5) is strictly a theorem, it is an important problem to study its stability when subject to higherorder corrections. A calculation to 0{p^) has been performed [8] (see also [9]), from which we only quote the numerical result.
{rh,s)'''
(0.04 + 0 . 3 &^)
W
(8)
where the lowenergy constant b^ is expected to be of order 1. We note t h a t the corrections to the central value as compared to (6) are sizeable, and t h a t the unknown constant induces a theoretical uncertainty in the extraction of /j,s from experiment. T h e corresponding plot is shown in Fig. 4 (taken from [8]).
5 Variants of loop regularization We have seen t h a t the lowenergy theorem for (rj^s) ^^ given by a leadingorder kaonloop effect, but t h a t the
All loop results discussed so far have been calculated in what is called Heavy Baryon Chiral Perturbation Theory. In this formalism, the nucleons are treated as heavy, nonrelativist ic fields, for which the relativistic corrections suppressed by powers of I/TUN can be calculated systematically. This method has the advantage t h a t loop diagrams always follow naive power counting rules and t h a t leading loop calculations are technically simple, but the downside t h a t the analytic structure is sometimes distorted even in the lowenergy region, and t h a t in practice I/TUN corrections are awkward to calculate. One alternative scheme t h a t has been suggested to overcome these shortcomings is "infrared regularization" [10], which is a variant of dimensional regularization t h a t preserves power counting also in relativistic baryon C h P T . In this scheme, all the l/rriN corrections are automatically resummed as depicted schematically in Fig. 5. It has been shown in [11] t h a t in some cases particularly sensitive to relativistic "recoil effects" (in [11]: the neutron electric form factor), this relativistic resummation leads to a much improved convergence behavior. It is therefore interesting to see what happens to the strange magnetic radius in this scheme, even beyond 0{p^). We find the following partial corrections of 0{p^) [9]:
MrlJ''
3 5 ( 5 Z ) ^  6 D F + 9 F ^ ) MK 3847rF. TOjV
+0.053 W
,.
+0
f^niA/E 
MK
(9)
where the additional terms proportional to AniA/jj = mA/E — TTiN have been included in the numerical value. Although these expressions are not complete (there will also be twoloop contributions), they are nonanalytic in the quark masses and therefore not modified by counterterms. Numerically, they t u r n out to be large, about half the size of the leading 0{p^) value. This supports once more the conclusion t h a t , unfortunately, the lowenergy theorem for {rj^s) ^^ 0{p^) is numerically not very reliable.
B. Kubis: Strange form factors and Chiral Perturbation Theory
100
scheme. T h e problem of large l/rriN corrections in chiral SU(3) is by no means solved here.
6 Summary and conclusions
Fig. 6. Cutoff dependence of (rj^^s)
5.2 Cutoff regularization Various a t t e m p t s at cutoff regularization have been undertaken in the context of C h P T . We do not t r y to give an overview of these, but instead concentrate on one specific approach [12] in which (rj^g) has exphcitly been considered [13]. T h e idea is t h a t the nucleon has an intrinsic size, and t h a t C h P T can only reliably predict the TT or K fluctuations t h a t are longranged on this scale. Dimensional regularization does not separate these different ranges of momenta in the loop integration, such t h a t it might be more efficient to employ a cutoff in order to reduce unrealistically strong shortdistance effects. T h e method in [12] amounts to the use of a KNN form factor instead of a constant vertex. In [13], a dipole form has been chosen, which leads to the analytic result (^M,s) cutoff
X{x)
\''M,s)dim.reg. X
^\MK>
T h e matrix elements of the strange vector current seem to remain elusive quantities for a description in Chiral Perturbation Theory. They are problematic as Goldstone boson dynamics is not overly dominant here, and in most cases, unknown lowenergy constants appear at leading order. We have pointed out t h a t there is an exception, the strange magnetic moment for which a parameterfree lowenergy theorem exists. However, higher order corrections are found to be sizeable, such t h a t the lowenergy theorem is very unstable once these are taken into account. Also cutoff calculations indicate a smaller absolute size of ( r  f g). T h e role of C h P T remains however to aid to interrelate more data, as they become available [15]. Acknowledgements. I would like to thank the organizers of PAVI 04 for the great opportunity to participate in this workshop, and for the wonderful organization of the whole event. I am grateful to T.R. Hemmert and U.G. Meii3ner for the fruitful collaboration that originally introduced me to this subject, and to various colleagues for interesting discussions that deepened my understanding of the field, in particular H.W. Hammer, M.J. RamseyMusolf, J.F. Donoghue, B.R. Holstein, T. Huber, and A. RoB. This work was supported in part by RTN, BBWContract No. 01.0357, and ECContract HPRNCT200200311 (EURIDICE).
(10)
References {x + lf
This cutoff dependence is displayed in Fig. 6. It is obvious t h a t for cutoffs of the order of yl = 4 0 0 . . . 600 MeV, (^Afs) cutoff is sizeably reduced compared to the dimensional regularization result. Despite the host of plausible arguments in favor of cutoff schemes, there are also some drawbacks: 1. The cutoff is an additional parameter t h a t can only, at best, be estimated. 2. It presents a deviation from the strict effective Lagrangian approach in C h P T , as a consequence of which gauge invariance usually has to be cured "by hand", and chiral symmetry is by no means guaranteed.^ 3. The cutoff upsets the analytic structure of the form factors: it produces additional unphysical poles a n d / o r cuts. This means in particular t h a t is is unclear how a marriage with dispersion relations (see e.g. [8] and references therein) might be achieved. 4. Finally, the reduction of the numerical value of {r\j g) must not be confused with the inclusion of higher order corrections as done in the infrared regularization
This problem has recently been addressed in [14].
1. J. Gasser, H. Leutwyler: Annals Phys. 158, 142 (1984); J. Gasser, H. Leutwyler: Nucl. Phys. B 250, 465 (1985) 2. G. Colangelo, J. Gasser, H. Leutwyler: Phys. Rev. Lett. 86, 5008 (2001) 3. M.J. Musolf, H. Ito: Phys. Rev. C 55, 3066 (1997) 4. T.R. Hemmert, U.G. MeiiSner, S. Steininger: Phys. Lett. B 437, 184 (1998) 5. D.T. Spayde et a l , [SAMPLE Collaboration]: Phys. Lett. B 583, 79 (2004) 6. K.A. Aniol et al., [HAPPEX Collaboration]: Phys. Rev. Lett. 82, 1096 (1999); Phys. Lett. B 509, 211 (2001) 7. T.R. Hemmert, B. Kubis, U.G. MeiBner: Phys. Rev. C 60, 045501 (1999) H.W. Hammer, S.J. PugUa, M.J. RamseyMusolf, S.L. Zhu: Phys. Lett. B 562, 208 (2003) 9 B. Kubis: Thesis, Berichte des FZ Julich, jm4007 (2002) 10 T. Becher, H. Leutwyler: Eur. Phys. J. C 9, 643 (1999) 11 B. Kubis, U.G. MeiBner: Nucl. Phys. A 679, 698 (2001) 12. J.F. Donoghue, B.R. Holstein, B. Borasoy: Phys. Rev. D 59, 036002 (1999) 13 T. Huber: Master thesis (Univ. of Massachusetts, 2002); A. RoB: Master thesis (Univ. of Massachusetts, 2003) 14 D. Djukanovic, M.R. Schindler, J. Gegelia, S. Scherer: (2004), arXiv:hepph/0407170 15 F.E. Maas et al. [A4 Collaboration]: Phys. Rev. Lett. 93, 022002 (2004)
Eur Phys J A (2005) 24, s2, 101102 DOI: 10.1140/epjad/s2005040222
EPJ A direct electronic only
Timelike compton scattering and the BetheHeitler process Stephen R. Cotanch^ Department of Physics, North Carohna State University, Raleigh NC 276958202, USA Received: 15 October 2004 / Pubhshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. Arguments are presented for ^(7, e^e~)p measurements to obtain new information about the offshell timelike nucleon form factors, especially in the (f) meson region governing the (t)N couplings 9(hNN Theoretical predictions based upon a Quantum Hadrodynamic model and vector meson dominance are highlighted for both the proton form factor and the timelike Compton scattering cross section. The BetheHeitler process is also calculated but is only important at low momentum transfer t permitting a novel high \t\ (f) enhancement in the Compton cross section related to nucleon strangeness to emerge. PACS. 12.40.Nn Regge theory  12.40.Vv Vectormeson dominance factors  13.40.Hq Electromagnetic decays
1 Introduction
13.40.Gp Electromagnetic form
covariants. This study addresses the vector and tensor elements, 7^ and cr^jy, corresponding to the coupling constants gXj^jsi 9.nd g^jsiN^ respectively, since the (p is predominantly ss. These (pN coupling constants govern the Q u a n t u m Hadrodynamical [QHD] Lagrangian
While Compton scattering using real, p{j,'y)p, and even virtual (spacelike), p(e, e^7)p, photons has been measured, the timelike virtual Compton scattering [TVCS] process, p ( 7 , e + e ~ ) p , has received little attention. However such measurements can provide new information about the timelike proton form factor in the unobserved region 0 < g^ < 4 M ^ (all measurements utilized e+e~ ^ NN where g^ > 4M^) since the threebody final state has an essentially unrestricted virtual photon mass q^ > 4Mg ~ 0. Previous theoretical studies using vector meson dominance [VMD] have predicted t h a t the p ( 7 r ~ , e + e ~ ) n [1,2] and ^ ( 7 , e + e ~ ) p [3] cross sections b o t h exhibit a dramatic, dual peaked resonant signature for timelike virtual photon fourmomentum spanning the vector meson masses {q^ ~ My for V = p^uo^cj)). This paper extends the work of [3] by calculating the competing BetheHeitler [BH] process, 7p ^ 77vP ^ e+e~p, and documenting t h a t it is only important for small t, significantly below the interesting high t region where the s and u channel processes embodying the V M D resonant signature dominate.
used for b o t h the V M D proton form factor and T V C S cross section predictions discussed below. If there is no or insignificant nucleon strangeness then (j)N coupling should be suppressed due to the dominant ss structure of the (j) and the OZI rule. However, significant OZI violations have been observed in inelastic fip and elastic vp scattering, pp annihilation experiments and measurements of the TiN sigma term, which collectively suggest appreciable strangeness in the proton. Evidence for nucleon strangeness is further discussed and reviewed in [4]. Related, a previous analysis [3] of spacelike neutron electric form factor d a t a and high \t\ (f) photoproduction d a t a yielded gX^^^i — 13, 9^NN — ' Accurate T V C S measurements at high t will permit extraction of these couplings which quantify the degree of nucleon strangeness.
2 Hidden strangeness
3 Predictions for TVCS
In addition to mesons, other eigenstates of the Q C D Hamiltonian also contain hidden strangeness. One clear example is the ground state vacuum with nonzero quark condensates {0\ss\0)  {0\uu\0)  A^^^. Of current intense interest and debate is nucleon hidden strangeness which is completely specified by the n = 1, 2 ... 16 matrix elements, {N\srns\N), involving the Lorentz bilinear ^ Email address: [emailprotected]
As detailed in [3] the nucleon form factors were calculated using a generalized V M D model. A good description of the baryon octet form factors was obtained, especially the sensitive spacelike neutron electric form factor and the proton E M form factors in b o t h the spacelike and timelike regions as depicted in Fig. 1. Note the resonant peaks in the unmeasured timelike vector meson region, in particular the 0 peak which scales with the (pN coupling.
S.R. Cotanch: Timelike compton scattering and the BetheHeitler process
102 i(r
_.
1
.
1
1
>
1
10 p^UP
n..
p(77v)P ^
.1
= 4,0 GcV
10
ll
!0' 10"
10"
:
^ ]()
l./i\
10
11
lio^i
\
^ ] ( )  ^ l /
K^
10
10"
1
;i
]{)'
w .6
..^
4
?
^2
I
I
2
3
4
.^
6
10""
q(GcV)
.i
10
1.^
^^.._.x.
20
25
30
33
6^^,,^"^ [deg]
Fig. 1. Data and VMD (absolute value) for the proton electric ^^ ^ ^ ^ ^
T , ^ rr^Tr^n / N rx^n ^. ^ TT^TT^ Fig. 2. VMD prediction for TVCS, ^(7,7^)^. The smaller angle (j) peak quantifies the proton's strangeness
T h e form factor peaks are reflected in the T V C S cross section displayed in Fig. 2. A novel, dual peak proflle arises from the quadratic relation between q^ and the recoil proton lab angle. T h e smaller angle (j) peak corresponds to high t and is dominated by the u channel proton propagator with QcfyNN coupling (sparse dotted curve). T h e magnitude of this peak represents the degree of the proton's strangeness. T h e other two peaks near 30^ entail lower t and involve (j) and uo coupling, respectively, to TT (dense dotted curve) and rj (short dashed curve) t channel exchange. These two peaks represent the expected (j) production background since they are based upon established results. This dual peak (j) signature follows from only V M D and should occur in other dynamic models. Therefore V M D predicts t h a t a measurement of the high t T V C S cross section ratio R = a[q^ = M^)/a{q'^ = M^) is proportional to g'^NN/d^NN "^^^^ ^^^ been numerically conflrmed in this model giving R = 0.14 / (where f is a kinematic quantity of order unity) which is an order of magnitude larger t h a n the OZI prediction [4], R = tan'^ S f = 0.0042 / , where S = 3.7^ is the deviation from the ideal quark flavor mixing angle in the 0. T V C S measurements would therefore appear to be an excellent probe of the proton's strangeness content.
Cparity 1 (single photon production) while for the theoretically known BH process they have C = 1 (two photon production). For the kinematics listed in Fig. 2, the calculated BH cross section dominates the T V C S cross section for \t\ < 0.01 G e V ^ is comparable for t up to 0.04 GeV^ and is an order of magnitude smaller for t > 0.06 GeV^. Hence the charge asymmetry measurement will only be necessary for small t where meson and pomeron (long dashed curve in Fig. 2) exchange dominate the T V C S process. To extract the (pN couplings at high t, the T V C S cross section is sufl&ciently large for direct measurement without competition from the BH process.
4 The BetheHeitler process Because the BH and T V C S processes compete it is necessary to assess their relative magnitudes. Even if the amplitudes are comparable it is still possible to extract the TVCS amplitude by measuring the charge asymmetry (cr(e+e~) — cr(e~e+)) since in TVCS the e+e~ pair has
5 Conclusion W i t h GeV electron facilities, such as Jlab, T V C S experiments appear quite feasible, providing an opportunity to obtain the unknown nucleon on and offshell timelike form factors. If V M D is valid the ^A/" couplings can then be extracted which in t u r n permits a direct assessment of nucleon hidden strangeness. R. A. Williams and C. W. Kao are acknowledged. This work was supported by D O E grant DEFG0297ER41048.
References 1. R.A. WilUams, S.R. Cotanch: Phys. Rev. Lett. 77, 1008 (1996) 2. S.R. Cotanch, R.A. WilUams: Nucl. Phys. A 631, 478 (1998) 3. S.R. Cotanch, R.A. Wilhams: Phys. Lett. B 549, 85 (2002) 4. J. Ellis, M. Karliner, D.E. Kharzeev, M.G. Sapozhnikov: Phys. Lett. B 353, 319 (1995)
Eur Phys J A (2005) 24, s2, 103103 DOI: 10.1140/epjad/s200504023l
EPJ A direct electronic only
Corrections to the nuclear axial vector coupling in a nuclear medium G.W. Carter and E.M. Henley Department of Physics,Box 351560, University of Washington, Seattle, WA 981951560, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The temporal component of the weak axial vector current in nuclei is increased due to meson exchange currents. We consider further corrections from pions and the sigma meanfield. PACS. 23.40.s Beta decay; double beta decay; electron and muon capture  23.40.Bw Weak interaction and lepton (including neutrino) aspects
T h e weak axial current j ^ is not conserved in a nuclear medium or nucleus; t h e space component is reduced whereas t h e time component, g\ increases by 50 — 100% [1]. Already in t h e 1970's it was shown [2] t h a t nuclear exchange currents are responsible for t h e major part of the increase of g^^. A simplistic explanation is t h a t t h e pion adds an extra 7^ which makes t h e space component of 0{p/M), whereas t h e time component becomes of 0 ( 1 ) . Pion exchange currents have been examined by several authors [3]. Some aspects of t h e exclusion principle have been included in pion exchange calculations, e.g., via use of t h e shell model. However, t h e theoretical enhancement falls about 10% short of t h e measurements [4]. There is, to the same order, t h e one nucleon pion loop contribution. It differs from t h e free nucleon counterpart by t h e effect of t h e Pauli exclusion principle. It is this effect we have examined in order to see whether it might contribute t h e missing ^ 10% of g^. We use t h e chiral Lagrangian of Carter, et al. [5], treating t h e a field in mean field theory. Ng 1 D L2
2CJI
(cr + i r
7r75) +  ^ ' 7 ^ ^
a^75
TV, (1)
N
[7^r . [TT X Z\^7r + 75(crZ\^ TT  irA^a)]] N , (2) (3) ^^^
d^TT
(4)
where g' is t h e coupling to t h e chiral vector mesons p and ^ 1 , and bold characters indicate isospin vectors. To start with, t h e nucleon has a m o m e n t u m < kpSince we need t h e propagators in t h e presence of t h e nucleus, they have to be above t h e Fermi sea (e.g.  p  > kp)
T h e effect of t h e inmedium sigma field is to reduce the nucleon mass, M ^ M* ~ 0.8M. We assume t h a t t h e pion coupling to t h e nucleon has a dipole form factor with a cutoff A = 0.9 or 1.1 GeV. T h e constant D is fixed t o fit gA = 1.26 for t h e free nucleon. gA
(5)
where CTQ is t h e vacuum expectation value of t h e cr field, (To = 102 MeV, and a is t h e mean field result at finite density. We fnd t h a t g\ = ^ A ( 1 + S), with 6 =  . 1 3 (  . l l ) for A = 1.1(0.9) GeV. Thus, t h e omitted eflFect is indeed of the order of magnitude required (1015%), b u t with t h e wrong sign. These corrections only serve to increase t h e discrepancy between theory and experiment. T h e authors t h a n k Mary Alberg for helpful comments. This work was supported by t h e U. S. Department of Energy grant DEFG0297ER4014.
References 1. See, e.g., D.H. Wildenthal, M. S. Curtin, B. A. Brown: Phys. Rev. C 28, 1343 (1983) 2. L. Kubodera, J. Delorme, M. Rho: Phys. Rev. Lett. 40, 755 (1978) 3. See e.g., TS. Park, H. Jung, DP. Min: Phys. Lett. B 409, 26 (1997) 4. E.K. Warburton: Phys. Rev. Lett. 66, 1823 (1991); Phys. Rev. C 44, 233 (1991) 5. G.W. Carter, P.J. Ellis, S. Rudaz: Nucl. Phys. A 603, 367 (1996), [Erratumibid., 608, 514] (1996)
Eur Phys J A (2005) 24, s2, 105105 DOI: 10.1140/epjad/s2005040240
EPJ A direct electronic only
Strangenessconserving effective weak chiral Lagrangian HeeJung Lee^, Chang Ho Hyun^, ChangHwan Lee^, and HyunChul Kim^ ^ Departament de Fisica Teorica, Universitat de Valencia E46100 Burjassot (Valencia), Spain ^ Institute of Basic Science, Sungkyunkwan University, Suwon 440746, Republic of Korea ^ Department of Physics and NuRI, Pusan National University, Busan 609735, Republic of Korea Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. We consider the strangenessconserving effective weak chiral Lagrangian based on the nonlocal chiral quark model from the instanton vacuum. We incorporate the effect of the strong interaction by the gluon into the effective Lagrangian. The effect of the Wilson coefficients on the weak pionnucleon coupling constant is discussed briefly. PACS. 12.40.y, 14.20.Dh Effective chiral Lagrangian, Wilson coefficients, weak pionnucleon coupling constant
In this work, we briefly illustrate the AS = 0 effective weak chiral Lagrangian from the instanton vacuum, which is essential to study the weak interaction of hadrons at lowenergy regions. Since parity violation (PV) can provide highprecision tests of the electroweak s t a n d a r d model (SM) [1], a great deal of attention has been paid to P V in the SM. It is known t h a t there is discrepancy in the weak charge between measurements in atomic physics and the prediction from the SM [2]. Moreover, it is well known t h a t there is still disagreement theoretically as well as experimentally [3] in determining the weak pionnucleon coupling constant h^.. Recently, Meissner et al. [4] studied /i^ within the SU(3) Skyrme model, based on the effective c u r r e n t current interaction which is equivalent to a factorization scheme at the leading order in Nc. However, processes such as nonleptonic weak processes defy any explanation from the factorization. Therefore, we first derive the AS = 0 effective weak chiral Lagrangian, incorporating the effective Hamiltonian [5] within the nonlocal chiral quark model from the instanton vacuum. Using the derivative expansion, we obtain the effective weak Lagrangian to order 0{p^) and to the n e x t  t o  l e a d i n g order (NLO) in Nc. We will use this derived effective weak chiral Lagrangian as a starting point for investigating h\. We include two effects from the strong interaction in the effective Lagrangian : T h e first one is the Q C D vacu u m effect which is implemented in the chiral quark model from the instanton vacuum, and the other is the perturbative gluon effect (the strong enhancement effect) which is encoded in the Wilson coefficients [5]. T h e terms at the leading order of Nc in the effective Lagrangian are expressed in terms of the currentcurrent interactions which correspond to the factorization scheme. It can be easily
shown t h a t they are identical to those in [4] when the perturbative gluon effect is turned off. On the other hand, the NLO terms from the nonfactorization scheme have a more complicated form. T h e explicit form of the AS = ^ effective weak Lagrangian with the Wilson coefl&cients to the NLO can be found in [6]. T h e role of the Wilson coefl[icients can be investigated by calculating h\ from the effective weak chiral Lagrangian. As done in [4], we employ the chiral soliton with the zeromode quantization. T h e pion which couples to the nucleon can be introduced from the meson fluctuation around the soliton fleld. W h e n the perturbative gluon effect is turned off and NLO terms are not considered, h\ t u r n s out to be the same as t h a t in [4]. On the other hand, if the strong enhancement effect is taken into account at the leading order in A^c, a rough estimation of the effect shows t h a t h\ is enhanced by 20 %, compared to t h a t without the Wilson coefficients. However, it should be noted t h a t if we restrict ourselves to SU(2), h\ vanishes anyway. T h e investigation in the SU(3) flavor space is under progress.
References 1. M.A. Bouchiat, C.C. Bouchiat: Phys. Lett. B 48, 111 (1974) 2. C.S. Wood et al.: Science 275, 1759 (1997); S.C. Bennett, C.E. Wieman: Phys. Rev. Lett. 82, 2484 (1999) 3. W.S. Wilburn, J.D. Bowman: Phys. Rev. C 57, 3425 (1998) 4. U.G. Meissner, H. Weigel: Phys. Lett.B 447, 1 (1999) 5. B. Desplanques, J.F. Donoghue, B.R. Holstein: Annals Phys. 124, 449 (1980) 6. H.J. Lee et al.: hepph/0405217
IV Experimental techniques in PV electron scattering IV1 Beam asymmetry
Eur Phys J A (2005) 24, s2, 109114 DOI: 10.1140/epjad/s200504025y
EPJ A direct electronic only
Overview of laser systematics Gordon D. Gates, Jr. Department of Physics, University of Virginia, Charlottesville, Virginia, USA Received: 15 January 2005 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. This paper discusses systematic effects in parity experiments that originate from the laser and optics system that are used in a polarized electron source. Covered are both the sources of systematics, as well as strategies for their minimization. PACS. 29.25.Bx Electron sources  29.27.Hj Polarized beams  42.25.Ja Polarization  42.25.Lc Birefringence  ll.30.Er Charge conjugation, parity, time reversal, and other discrete symmetries
1 Introduction
2 The various types of systematics
In general, experiments t h a t study parity violation in electron scattering utilize a polarized electron source t h a t is based on photoemission from various types of gallium arsenide (GaAs) crystals. T h e photoemission is induced using circularly polarized light from a laser. Because the helicity of the electron beam is determined by the polarization state of the laser light, the electron polarization can be reversed or "flipped" quickly and in a quasirandom manner by using an electrooptical device such as a Pockels cell to circularly polarize the light. Parity experiments typically measure tiny helicitydependent asymmetries in the scattering of polarized electrons off unpolarized targets. T h e asymmetries themselves might range from 0.1 parts per million (ppm) to 100 ppm, and it is sometimes necessary to control helicitycorrelated systematic effects at the level of parts per billion. In experiments where electronic crosstalk is sufficiently under control, the helicitycorrelated changes in the parameters of the electron b e a m generally originate in helicitycorrelated changes in the light used to induce photoemission. Laser systematics are thus critical to achieving increasingly accurate measurements of parity violation in electron scattering. Ever since the first pioneering experiment t h a t observed parity violation in electron scattering [1], the understanding of laserbased systematics has grown. It is now possible to catalog some of the dominant effects, including how they can be diagnosed, and in some cases corrected. Indeed, the improvement of our understanding of laser systematics has played a critical role in making increasingly sensitive parity experiments possible. This paper will examine some of the laser systematics t h a t were dominant during several experiments with which the author was involved [25], as well as discussing strategies for their minimization.
In the absence of any effort to control them, the largest systematic in a parity experiment will generally be helicitycorrelated asymmetries in the charge delivered to the target. While it is certainly important to measure and correct for "charge asymmetries" this can only be done up to a point. Beam current measuring devices will always have nonlinearities at some level. Even if one had a perfect device for measuring b e a m current, there is still the possibility t h a t through interaction with the accelerator, such as beam loading, charge asymmetries could be translated into other helicitycorrelated effects. Fortunately, it is reasonably straightforward to "balance" the charge delivered to the target over the course of an experiment at the level of a few hundred parts per billion (ppb) resulting in systematic uncertainties in the parity violating asymmetry of a few ppb. If charge asymmetries are the "zerothorder" effects, the firstorder effects are then helicitycorrelated differences in the b e a m position. Since charge asymmetries are reasonably straightforward to control, these "position differences" end u p being a more troublesome problem. Many recent parity experiments have had significant contributions to their systematic errors from helicitycorrelated position differences [4,5]. Even if charge and position differences are reduced to negligible levels it is still possible to be troubled by higherorder effects. For instance the spot size of the laser can systematically change while position and charge are held fairly constant. One issue t h a t needs to be considered is whether the elimination of lowerorder effects simply results in the increase of higherorder effects. This makes it desirable to have diagnostics for higherorder effects even if there is no obvious way to control them.
no
G.D. Gates, Jr.: Overview of laser systematics
3 The sources of systematics 3.1 Charge asymmetries Charge asymmetries result when the average current associated with one helicity state is different from the average current associated with the other helicity state. T h e dominant mechanisms associated with this effect are well understood, and have been described in some detail for b o t h simple [6] and more complex [7] optics setups. T h e asymmetries stem from the fact t h a t when making circularly polarized light, there are always small admixtures of linear polarization which cause a small degree of ellipticity. W h e n the helicity of the light is flipped, it is often the case t h a t the major axis of the polarization ellipse will rotate by 90°. Since most optics systems have many elements (for instance mirrors) t h a t transport one linear polarization better t h a n another (a property we will refer to as a transport asymmetry), flipping the helicity can cause a change in the efficiency with which the light delivered to the cathode. Historically, this type of effect has sometimes been referred to as the "PITA" effect, where PITA is an acronym standing for "polarization induced transport asymmetry" [6]. T h e P I T A effect thus results from the fact t h a t the optics system has an "analyzing power" with an accompanying analyzingpower axis. In polarized electron sources, the optics transport syst e m is not the only component with an analyzing power. Bulk GaAs has a theoretical maximum polarization of 50%, and values of 3545% are typical. Recently it has become common to use modified GaAs crystals such as strained GaAs or superlattice GaAs because in such crystals a degeneracy associated with the valence band is broken raising the theoretical maximum polarization to nearly 100%, with values of 7082% being typical. T h e improved polarization of these photocathodes comes at a price. W h e n irradiated with linearly polarized light, these photocathodes have a q u a n t u m efficiency (QE) t h a t depends on the orientation of the light's polarization axis with respect to an axis t h a t lies in the plane of the crystal's surface. T h e crystal itself thus has an analyzing power. Figure l a illustrates the direction of the analyzingpower axis. T h e Q E anisotropy associated with the analyzingpower axis can be as much as 15%. T h e analyzing power of the crystal has essentially the same effect on beam current as does a t r a n s p o r t asymmetry. W h e n the crystal is illuminated with elliptically polarized light, the photoemitted current depends critically on the position of the major axis with respect to the analyzingpower axis. If the polarization ellipses associated with the two helicity states are as indicated in Fig. l b , a maximal charge asymmetry will result. If the polarization ellipses are oriented as indicated in Fig. Ic, a minimal charge asymmetry will result. W h e t h e r a charge asymmetry results from a laserb e a m transport asymmetry, or from an anisotropy in the photocathode's QE, it is straightforward, and useful, to characterize the effect quantitatively. For definiteness, we will assume t h a t the device used to produce circular polarization is a Pockels cell, oriented so t h a t its fast axis
a) A GaAs crystal with an "analyzingpower" axis as indicated.
b) Most ser\s\Y\we orientation for polarization ellipses.
c) Least ser\s\Y\ve orientation for polarizaton ellipses.
Fig. 1. a Illustrated is a GaAs crystal with a quantum efficiency that is sensitive to the orientation of linear polarization with respect to the indicated analyzingpower axis, b Polarization ellipses for nominally positive and negative helicity light resulting in maximum charge asymmetry, c Polarization ellipses for nominally positive and negative helicity light resulting in minimum charge asymmetry
is at ° with respect to horizontal. We will further assume t h a t prior to traveling through the Pockels cell, the light is linearly polarized in the horizontal direction. It is convenient to parameterize the phases introduced by the Pockels cell as (1)
<5^ = + ( I + ai)  / \ i
(2)
If Ai = ai = 0, the phases introduced by the Pockels cell are =b, and in principle, the light will have perfect circular polarization. If either Ai or ai are nonzero, however, elliptical polarization will result. I note in passing t h a t we are using the subscript 1 for a and A to be consistent with the notation of [7]. Let us now assume t h a t the light, after passing through the Pockels cell, passes through an asymmetric t r a n s p o r t system, characterized by two orthogonal axes x' and ?/', where the x^ axis makes an angle 0 with respect to the horizontal. We will assume t h a t light linearly polarized along the x\y^) axis will be transported with a transmission coefficient Tx'{Tyf), and define the quantities T = (T^/ + r ^ O / 2 , and s = T^^  Ty>. W i t h these definitions, the charge or equivalently current asymmetry Aj can be written to first order as AI
JRJL
Ai cos 20
(3)
where I^{I^) are the electron b e a m intensities associated with the Pockels cell phases 5^{5^). T h e reason we chose to write the phases as we did in 1 and 2 is immediately apparent. T h e equation for Ai depends linearly on Z\i, but not at all on a i . T h e reason is as follows. If Z\i 7^ 0, polarization ellipses result whose major axes will rotate by 90° when the helicity is flipped. T h a t is, if Z\i 7^ 0 we have a situation such as is illustrated in Fig. l b . If ai ^ 0, the polarization ellipses for the two helicity states will be
G.D. Gates, Jr.: Overview of laser systematics
111
Large Ai and charge asymmetry Medium Ai and charge asymmetry Small Ai and charge asymmetry
Fig. 2. Illustrated is a GaAs crystal being irradiated by light in which the residual linear polarization is varying from a maximum toward the top of the crystal to a minimum toward the bottom of the crystal
coincident with one another. Only the direction that the electric vector travels around the ellipse will change. We will discuss later how 3 can give us guidance in suppressing charge asymmetries. One final comment regarding the phase Z\i. It should be noticed that the sign of Z\i does not change when the phase on the Pockels cell is flipped. Thus, any object in the optics system, a vacuum window, a mirror, or even residual birefringence in the Pockels cell, can cause a nonzero value of Z\. 3.2 Position difFerences from phase gradients
In the previous section we discussed how a charge asymmetry can result from a nonzero value of the phase Z\i. If we consider a laser beam spot illuminating a GaAs crystal, it can also be the case that the phase Z\i, and hence the associated charge asymmetry, varies in some manner across the laser spot. Such a situation is illustrated in Fig. 2. If the charge asymmetry for the emitted electrons changes as we move from the top of the crystal to the bottom, the beam profiles for the two helicity states will have centroids that are shifted vertically with respect to one another. From the perspective of our beam position monitors, these shifts will be seen as helicitycorrelated position differences. To first order, the position differences will be proportional to the gradient of the phase Z\i, and independent of the average value of Z\i.
Fig. 3. Illustrated is steering due to a Pockels cell having lenslike properties when it is pulsed at high voltages
3.4 Position differences from gradients in t h e analyzing power of t h e cathode
The last class of effects we will mention arises from changes in the QE anisotropy as we move across the cathode. For instance, assume that the direction of the analyzingpower axis is constant, but that the magnitude of the anisotropy changes from 5% at the top of the cathode to 10% at the bottom. If the incident light is perfectly circularly polarized, there is no component of linear polarization and hence no charge asymmetry. If there is residual linear polarization, however, there could be a charge asymmetry whose size varies across the crystal, resulting in changes in the centroid of the electron beam's position much as was discussed in Sect. 3.1. The reader should note that this type of position difference should be proportional to the degree of ellipticity of the light, and hence Z\i. This is in contrast to the position differences from phase gradients (Sect. 3.1) where to first order we would expect no dependence on Z\i. This difference in response to Ai is a useful diagnostic tool for distinguishing between phase gradients in the optics system and analyzingpower gradients on the cathode.
4 Controlling systematics 4 . 1 Charge asymmetries
There are at least three strategies for controlling charge asymmetries and 3 is useful for understanding two of them.
3.3 Position differences from steering effects
Another source of helicitycorrelated position differences is steering caused by the Pockels cell. The Pockels cell is alternately pulsed to positive and negative high voltage in order to introduce the phases given by 1 and 2. Empirically, it appears that this results in the Pockels cell behaving alternately as a diverging and converging lens. If a laser beam is sufficiently small in diameter and goes through the very center of the Pockels cell, the steering effects can be kept quite small. As one goes off center, however, some steering occurs as would also be the case with any lens. This effect, which is illustrated in Fig. 3 can be quite significant.
4.1.1 Phase adjustments
The phase Z\i can be controlled using the Pockels cell. The nominal voltage at which the Pockels cell is pulsed is 7 kV. If a fixed voltage is added to the voltage associated with each polarity, one can introduce an arbitrary Z\i. For instance one could run at +2, 900V and —2, 500V. Notice that it is not the magnitude of each voltage that is changed, but rather its actual value. If instead we changed the magnitude of each voltage by a few hundred volts, we would be adjusting ai and not Z\i, and no change to the charge asymmetry would occur.
112
G.D. Gates, Jr.: Overview of laser systematics
Fig. 4. Illustrated is the basic setup for using a rotating halfwave plate (RHWP) to rotate the polarization ellipses associated with the laser light
40
60
80
100
120
140
asym_bpm1l02ws, Asymmetry vs. 6
Aq = 7.98 + 1211.75 sin (2x+ 75.52) + 3151.04 sin (4x+ 158.47)
4.1.2 Rotating halfwave plate Another way of controlling charge asymmetries is through 0 (the angle t h a t appears in 3). W h e n t h e light emerges from t h e Pockels cell, t h e major axes of t h e ellipses are typically either vertical or horizontal. This is determined by t h e orientation of t h e fast axis of t h e Pockels cell, which . By introducing a halfwave plate, howis usually at ever, we can rotate t h e orientation of t h e ellipses. T h e basic setup is illustrated in Fig. 4. T h e goal is t o rotate the ellipses toward t h e orientation illustrated by Fig. Ic, where t h e major axes are at ° with respect to t h e analyzingpower axis. This is equivalent t o adjusting 0 to 45°. In a practical situation it is actually desirable to retain a small amount of sensitivity to t h e analyzing power of t h e cathode. In this way t h e charge asymmetry induced by adjustments of Z\i can be used as a diagnostic for maximizing t h e circular polarization of t h e light t h a t actually strikes t h e cathode. This helps with, among other things, the position differences due to analyzingpower gradients on t h e cathode. The effect of t h e rotating halfwave plate is illustrated dramatically by t h e d a t a shown in Fig. 5. T h e charge asymmetry is shown as a function of t h e rotation angle ^ of a halfwave plate using a setup similar to t h a t shown in Fig. 4 (this 0 is distinct from t h e 0 appearing in 3). From t h e discussion presented here, we should expect t h e charge asymmetry t o vary sinusoidally with 4 0. This is because a 90° rotation of t h e halfwave plate will rotate t h e polarization ellipses by 180°, at which point t h e p a t t e r n should repeat. This is close to what we see in Fig. 5, b u t not exactly. An analysis of t h e d a t a (shown on t h e figure) reveals b o t h A6 and 2 6 components. This is because t h e halfwave plate itself will have imperfections and introduce a "Z\like" phase, and t h e fast and slow axes associated with t h a t phase rotates with t h e halfwave plate. For parity experiments at J L a b , a rotating halfwave plate is an essential part of controlling charge asymmetries. I note in passing t h a t at SLAG, we accomplished essentially t h e same thing (rotating t h e ellipses to an arbit r a r y angle) using two Pockels cell. For t h e sake of brevity, however, I will not discuss t h a t approach here.
4.1.3 lA cell T h e final technique for controlling charge asymmetries is to use what has come t o be called an "Intensity Asymme
Fig. 5. Charge asymmetry (in ppm) is plotted as a function of the angle of the RHWP. Also shown are the results of a fit including a constant, a 2^, and a 4 ^ component
try" or lA cell. T h e lA cell is a Pockels cell t h a t is set u p between two polarizers so t h a t it can be used as part of an electrooptical shutter. T h e charge asymmetry can be carefully measured, and t h e IA cell pulsed to a slightly different voltage for each helicity to achieve balance. This is sort of a bruteforce technique, in t h a t it does not address the underlying problems t h a t are causing t h e asymmetry. It is well suited for use in a feedback mechanism, however, and it can be done very quickly. For this reason, t h e use of an lA cell has become s t a n d a r d at SLAC, J L a b , and other labs. During H A P P E X II (an IA cell was not used during H A P P E X I) and during E l 5 8 , however, great care was taken to reduce charge asymmetries to something on the order of 100 p p m independent of t h e function of t h e IA cell. At SLAC this was accomplished using a "doublefeedback" method. T h e lA cell was used in a relatively fast feedback loop t o minimize charge asymmetries, and t h e size of t h e correction being applied by t h e lA cell became the error signal for a second feedback loop which corrected the value of Ai. For H A P P E X II Z\i was adjusted manually before turning on t h e lAcell feedback loop, and t h e size of t h e lA cell correction was monitored "by hand" to determine when further adjustments to Z\i were necessary. Keeping charge asymmetries small before turning on the lA cell is a good way t o help ensure t h a t higherorder effects do not become a problem.
4.2 Controlling position asymmetries As it t u r n s out, it is relatively easy t o control charge asymmetries. Position asymmetries, however, can present a real challenge to parity experiments. There are several techniques, however, t h a t are useful.
4.2.1 Minimizing steering To t h e extent t h a t a Pockels cell behaves like a lens, t h e minimization of steering can be accomplished by centering t h e laser beam on t h e Pockels cell. T h e Pockels cell is translated in two dimensions while monitoring t h e position differences. T h e one difficulty comes from t h e fact
G.D. Gates, Jr.: Overview of laser systematics that there is no obvious way to separate position differences due to steering from position differences due to other sources. Thus, the experimenter can fool themselves that they are centering the Pockels cell when actually they are compensating for several problems at once. In addition to precise centering of the Pockels cell, another technique for reducing the effects of steering is imaging. By using lenses to create an image of the Pockels cell at the location of the photocathode, steering effects can hypothetically be eliminated. In this configuration, rays of light emanating from the same point on the Pockels cell will all be imaged to the same spot on the cathode, regardless of their exit angle leaving the Pockels cell. Between centering and imaging, steering effects can be greatly suppressed.
4.2.2 Minimizing the effects of phase gradients
One important source of phase gradients is the Pockels cell itself. A Pockels cell will often have a residual birefringence that varies across the aperture of the cell. It is straightforward, however, to construct a small setup on an optics table to characterize such gradients. We have found that if we communicate to the vendor that this is a specification about which we are concerned, they are capable of controlling it at some level. If we subsequently characterize the phase gradients of each Pockels cell and select the best one, the effects from the Pockels cell can be substantially reduced. Another source of phase gradients is the vacuum window through which the laser beam enters the polarized electron source. This window is generally under stress, which causes induced birefringence. It is useful to explore ways of minimizing these effects, even if this means nothing more than carefully selecting the window that is used. Once the best Pockels cell and vacuum window have been chosen, it may still be possible to reduce the effects of phase gradients further. The position differences due to phase gradients are independent of the average value of Z\i, but they do depend on the orientation of the fast axis associated with the gradients to the analyzingpower axis of the system. If the phase gradients are dominated by the Pockels cell, the rotating halfwave plate can in principle be used to "dial away" the effect. If the phase gradients are dominated by the vacuum window, the rotating halfwave plate will have little or no effect.
4.2.3 Minimizing the effects of QE anisotropy gradients
One obvious way of minimizing the effects of QE anisotropy gradients is to choose a photocathode in which the anisotropics are small. Assuming this has been done, it is still possible to reduce the effects to a negligible size by ensuring that the light falling on the cathode is perfectly circularly polarized. It must be remembered, however, that even if light of arbitrary polarization state can be produced outside of the vacuum system, the vacuum window will have an effect, the exact nature of which is
113
generally unknown. In general it is not trivial to completely eliminate the effect. If a single Pockels cell and a rotating halfwave plate are being used, there is a way to eliminate the effect of QE anisotropy gradients in the limit where the vacuum window has no birefringence. As mentioned earlier, QE anisotropy gradients are proportional to Z\i. With a large Pockels cell voltage offset (large Z\i), the effect will be large, and the rotating halfwave plate can be used for zeroing. In a practical situation where the vacuum window does matter, this orientation may or may not be optimal to minimize all the sources of position differences. 4.2.4 Active feedback
It is always possible to employ a piezoelectric driven mirror to actively steer the laser beam in a helicitycorrelated fashion to suppress position differences. Such techniques can be quite effective, but it should be remembered that higher order moments of the laser spot will still be varying in a fully helicitycorrelated manner. It is thus prudent to minimize the sources of position differences as much as possible before using bruteforce feedback for suppression. 4.2.5 Adiabatic damping
If the accelerator is appropriately tuned and free of XY coupling, helicitycorrelated position differences are damped as ^JAjP^ where ^4 is a constant and P is the momentum. This is due to the adiabatic damping of phase space that takes place as the beam is accelerated. In practice, adiabatic damping has resulted in factors of 310 in suppression of position differences. Factors of 50 do not appear unrealistic, but have not yet been achieved.
5 Reversals Once everything has been done to minimize laser and optics related systematics, the experimenter can still reduce the size of systematic errors by employing a range of "reversals" in which some action is taken that changes the sign of the physics asymmetry without changing the sign of certain helicitycorrelated systematics. In this manner the effect of the systematic cancels out when computing the final physics asymmetry. In certain cases reversals can play a large role in taking a very troublesome systematic and suppressing it to a negligible level. One such reversal utilizes an "insertable halfwave plate", not to be confused with the rotating halfwave plate discussed earlier. Inserting such a halfwave plate will fiip the helicity of the light striking the cathode, ultimately resulting in a flip of the sign of the physics asymmetry. Certain systematics, however, will not flip sign, including steering effects and electronic cross talk. Helicitycorrelated steering is one of the important sources of beamposition differences, making the the insertable halfwave plate an important part of the full setup.
114
G.D. Gates, Jr.: Overview of laser systematics
Another reversal t h a t was used during SLAG E l 5 8 was an "asymmetry inverter" [7], which is essentially a pair of two beam expanders which, together with the other optical elements, provides either a positive or negative magnification of equal magnitude between the the object point near the Pockels cell and the image point at the cathode. Both position and angle differences should in theory be eliminated by using an asymmetry inverter. T h e final reversal we will mention involves the use of g — 2 precession, the fact t h a t when an electron b e a m at high energies traverses an angle 0 the spins will precess by an amount {g — 2 ) 7 ^ / 2 , where 7 is the Lorentz factor. If it is possible to run at two different energies such t h a t the electron will arrive at the target with two different helicity states while other energydependent quantities are sufficiently well understood, one has a way of fiipping the physics asymmetry t h a t truly has nothing to do with the polarized electron source, and would thus leave sourcerelated helicitycorrelated systematics unchanged. Perhaps the largest drawback of this approach is t h a t changing energy is time consuming, and hence cannot be done as frequently as one might want from the perspective of controlling systematics. It is very reassuring, however, to see a physics asymmetry fiip sign during such an energy change.
6 Summary There has been significant progress in the understanding of laser systematics since the early studies of parity nonconservation in electron scattering in the late 1970's. T h e control of such effects has made it possible to study progressively smaller asymmetries with increasing accuracy. Certainly an area of notable progress has been an improved understanding of the origins of position differences. Whereas steering effects have been identified for some time [2], the identification of phase gradients and Q E anisotropy gradients as a source of position differences has brought an important new perspective to the field. I t r y to summarize some of what we have discussed in Table 1. In the first row of this 2 x 2 matrix are a few of the things to which the effects of phase gradients are sensitive. T h e first column represents the sensitivities of the effects of phase gradients in the Pockels cell and the second column represents the sensitivities of the effects of phase gradients in the vacuum window. It is noted t h a t in b o t h cases, the position differences are independent of Z\i. In the case of phase gradients from the Pockels cell, however, the rot a t i n g halfwave plate ( R H W P ) can be used to zero out the effect. In the second row of this 2 x 2 matrix are a few of the things to which the effects of Q E anisotropy gradients are sensitive. T h e first column represents the interaction of Q E anisotropy gradients with the residual birefringence of the Pockels cell. T h e second column repre
Table 1. The sensitivities of position differences Pockels Gell
Vacuum window
Phase gradients
ind. of Z\i sens, to RHWP
ind. of Z\i insens. to RHWP
QE anisotropy gradients
prop, to Z\i sens, to RHWP
sens, to degree of circ. pol.
sents the the interaction of Q E anisotropy gradients with the birefringence of the vacuum window. T h e table indicates t h a t b o t h Z\i and the R H W P can be used to control the effects of the Pockels cell residual birefringence with the Q E anisotropy gradients. T h e table entry under vacu u m window in the second column is indicating t h a t these effects are zeroed in the limit of perfect circular polarization, but no specifics are given regarding how to achieve t h a t limit. Table 1 is useful for formulating strategies to minimize position differences. One possibility is to magnify the effects of Q E anisotropy gradients due to the Pockels cell by using a large offset voltage ( ^ 1 ) . T h e R H W P can then be scanned, and should show four zerocrossings for the position differences. Z\i can then be set to a nominal zero, and the R H W P scan repeated. T h e experimenter can then set the R H W P to a position close to the zero from the first scan for which the position differences are minimal during the second scan. The effects of column one are then minimal, and the effects of "column two row two" are probably also fairly small. W h a t is left is presumably mostly due to phase gradients in the vacuum window. T h e discussion given here certainly does not represent a definitive analysis of all forms of laser systematics. I hope, however, t h a t this paper provides a useful perspective as we advance forward in our understanding of these often subtle effects. Future parity experiments will make increasingly stringent demands on the control of laser systematics. There is every reason to remain optimistic t h a t these challenges can be met.
References 1. 2. 3. 4. 5. 6.
G.Y. Prescott et al.: Phys. Lett. 77B, 347 (1978) P A . Souder et al.: Phys. Rev. Lett. 65, 694 (1990) K. Aniol et al.: Phys. Rev. Lett. 82, 1096 (1999) K. Aniol et a l : Phys. Lett. B 509, 211 (2001) P L . Anthony et al.: Phys. Rev. Lett. 92, 1816021 (2004) G.D. Gates et al.: Nucl. Instr. and Meth. in Phys. Res. A 278, 293 (1989) 7. W.B. Humensky et al.: Nucl. Instr. and Meth. in Phys. Res. A 521, 261 (2004)
Eur Phys J A (2005) 24, s2, 115118 DOI: 10.1140/epjad/s200504026x
EPJ A direct electronic only
Beam optics for electron scattering parityviolation experiments Douglas H. Beck^ and Mark L. Pitt^ ^ Nuclear Physics Laboratory and Loomis Laboratory of Physics, University of Illinois at UrbanaChampaign, 1110 West Green Street, Urbana, IL 61801, USA ^ Department of Physics Virginia Polytechnic Institute and State University Blacksburg, VA 240610435, USA Received: 15 November 2004 / Pubhshed Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. Parityviolating electron scattering experiments at intermediate energies measure asymmetries in the 10~^ — 10~^ range and therefore require stringent control of false asymmetries. One of the primary sources of such asymmetries is the combined effect of helicitycorrelated changes in a certain beam property, accompanied by a change in the detector response. Careful control of the beam, including the optical properties of the acceleration and transport system, is required in order to reduce these false asymmetries to a manageable level. Developments in beam optics associated with the HAPPEX and GO experiments at Jefferson Lab are presented. PACS. 29.27.Hj Polarized beams  13.60.r Photon and chargedlepton interactions with hadrons  25.30.c Elastic electron scattering
1 Introduction Parityviolating electron scattering experiments measure the interference between the electromagnetic and neutral weak interactions with the target. As such, the experimental asymmetries are of order 10~^ with nucleon/nuclear targets at intermediate energies. T h e counting rates required to accumulate 10^^ or more events are achieved with the combination of high intensity beams, thick targets and detectors with large solid angle acceptance as discussed extensively at this conference. Control of false asymmetries is achieved by careful control of the beam, starting on the laser Table [1] and extending throughout the accelerator to the target. T h e most important aspect of b e a m control is maintaining a constant current independent of the beam helicity. This is achieved by measuring the electron b e a m current for each helicity state and feeding the resulting error signal back to control the laser intensity. Such systems have become fairly standard. T h e emphasis here is instead on the control of helicitycorrelated position and angle differences to which the present and future experiments have a significant sensitivity. To set the scale, suppose the response of the detector yield (Y) to a change in beam position x is {l/Y){dY/dx) = 0.1%/mm. In order for the asymmetry induced by such a sensitivity to be Af < 10~^, the helicitycorrelated b e a m motion must be Ax < 10~^ mm, or 100 nm. Control of b e a m position and angle at these scales is nontrivial. Helicitycorrelated motion of the polarized laser beam in the electron source (used to produce electrons by the photoelectric effect) can easily be 1000 n m without careful attention.
Two different approaches were used by the GO and HAPPEX experiments to reduce the helicitycorrelated beam position and angle differences. In the GO experiment, the position of the laser b e a m on the cathode was actively controlled to reduce b e a m position differences at the target with an automatic feedback system. To date, the HAPPEX experiment has essentially relied on careful setup of the optical polarization and transport system for the laser beam, together with "damping" of the position differences inherent in the acceleration of the beam. They have also recently introduced careful control of the optics near the target ("phase trombone"). T h e status of b o t h techniques is discussed below.
2 General beam optics Beams which are launched off the axis ("central orbit") of an optical system will oscillate about the axis with a motion known as b e t a t r o n oscillation. T h e formalism of linear beam optics can be used to determine the transverse motion of beam particles in the vicinity of the nominal beam trajectory. T h e particle orbit is described by the quantities [x, x^ y, y'] which are a function of the distance s along the central orbit. Here, x and y are the transverse displacements from the central orbit, while x' = dx/ds and y' = dy/ds are the inclination angles of the particle orbit relative to the central orbit. Under the assumptions of linear beam optics (small inclination angles and only constant or linearly increasing magnetic restoring forces) the solution of the equations of motion for the transverse
D.H. Beck, M.L. Pitt: Beam optics for electron scattering parityviolation experiments
116
Position Difference Suppression Factors from 100 keV to Hall C
30
5 MeV
100 keV
'
'
'
'
'
'
3 GeV 1
1
SH25
0 0
.a 15
1

yio ^

5 n
#
^ 4
Fig. 1. A schematic illustration of the phasespace ellipses at the end of a wellmatched and poorlymatched optical transport system. In the case of the badly matched system, the phasespace ellipse area is preserved, but the ellipse is distorted leading to a larger betatron amplitude than expected from the adiabatic damping factor
displacement in the case of no acceleration is given by [3]: x{s) = V^\/^p(s)
cos {^(s) + 0 }
(1)
5 6 BPM n u m b e r
Fig. 2. Measured damping factors in the GO experiment. The arrows below the graph show the kinetic energy of the electron beam at the various beam position monitors where the measurement are made. The damping factor is defined to be the helicitycorrelated position difference observed at the first BPM (BPM number 1) divided by the helicitycorrelated position difference observed at the BPM of interest. The expected damping in the injector (from 100 keV to 5 MeV kinetic energy) is not observed because of optics mismatches. Damping is observed in the accelerator (from 5 MeV to 3 GeV kinetic energy)
spacing, the expression for the transverse position is modified in a simple way
where
i>{s)
f W) ds'
(2)
Jo
is the accumulated b e t a t r o n phase, (j) is the initial phase, e is the emittance, and (3{s) is referred to as the b e t a function or amplitude function. So, the transverse position of the b e a m at a given point along the central orbit (e.g. at the target) depends on the initial position and angle of the b e a m as well as on the intervening optics (the integrated b e t a function along the orbit). T h e solutions for x and x' can be combined in the form 7(5) x'^{s) + 2 a{s) x{s) x'{s) + (5{s) x''^{s)
(3)
which defines a phasespace ellipse in the x — x' plane. T h e three parameters t h a t characterize this ellipse (a, (3 and 7) are referred to as the "Twiss parameters". Note t h a t these parameters are functions of the p a t h length s along the central trajectory, so the phasespace ellipse can change shape as the particle moves along its orbit. T h e area of this phasespace ellipse is simply Tre, where e is the emittance. For the case of no acceleration, Liouville's theorem states t h a t the area of the phasespace ellipse (and therefore the emittance) remains constant. W h e n there is acceleration, and the relative moment u m changes are small over the scale of the optical element
x{s) = ^/e^J(3[s)J—
cos {'0(s) + 0 } ,
(4)
where po and p are the initial and final b e a m particle momenta, respectively. This reduction in the amplitude of the b e t a t r o n motion as the beam m o m e n t u m is adiabatically increased is referred to as adiabatic damping. Therefore the acceleration of the machine gradually reduces the amplitude of the b e t a t r o n oscillation, and the transverse displacement from the central orbit can be reduced further at a given point by controlling the overall accumulated b e t a t r o n phase. In practice, one works to achieve the minimum possible helicitycorrelated position and angle differences in the beam in the injector by proper alignment and configuration of elements in the polarized injector laser p a t h as described in reference [1]. One then expects additional reduction in the position and angle differences observed at the experimental target by the adiabatic damping factor  A/PO/P T h e full expected adiabatic damping factor is often not achieved due to an optically mismatched beam transport system. In a perfectly matched system, the Twiss parameters after passing through each beamline element match the design parameters. As discussed in the next section, various types of imperfections in the transport system lead to a deviation from this ideal case.
D.H. Beck, M.L. Pitt: Beam optics for electron scattering parityviolation experiments I Betatron phases 3 REFRESH
DETACH
ll\1 '
/ y^
1J
/

/^ j
. .
Q_X J. i J .
^
1
F
D a.
/
iJ. .1.
I.
1
/
1
.
J\ /
, /
117 iPixj
r
1
1
1
1
1
1
^ _ _ , . , J  ^
' j f
1
^
J.L . i i 3 l e D i s a J B l , e « l f a l . n l a n J . Jl.. .IJ.fl . i. A .
« JJIID U..
inlxj REFRESH
DETACH
Fig. 3 . Calculated effects of the phase trombone in the Hall A beam line at Jefferson Lab. The upper panel shows a difference in the phase of the Px function of 60° (upper curves at right: ^x and ^x) for slightly different tunings. The small difference in the tuning can be seen in the lower panel where the nearly identical pair of curves show /3aj, /^^ for the two cases. The y functions are unaffected T h e result of a mismatched t r a n s p o r t setup are shown schematically in Fig. 1. At a given location, t h e phasespace ellipse preserves its area, b u t in a badly matched system t h e ellipse becomes distorted leading t o a larger b e t a t r o n amplitude ("orbit blowup") t h a n ideal adiabatic damping would predict.
3 Active feedback of beam position difference In t h e GO experiment, t h e relatively large bunch charge (running at 31 MHz pulse rate, necessitated by timeoffiight measurements, see [2]) required a nonstandard tuning of t h e injector which made reduction of t h e helicitycorrelated beam position differences with t h e standard damping difficult t o achieve. Nominally, with 3 GeV incident energy, t h e damping from t h e injector t o t h e target would be 3 GeV 95 (5) 355 keV However t h e damping factors measured in t h e x a n d y directions were typically 25 a n d 10, respectively as shown in Fig. 2.
T h e precise cause of the reduced damping is not clear; however, there are several likely contributors. In order for the damping t o be realized, t h e beam, characterized by its Twiss parameters, must have t h e envelope t o which the subsequent optical elements are matched. In practice, correction elements are used t o restore t h e envelope after sections of the beamline (e.g. linac, arc, etc.). Matching is particularly important in t h e injector where t h e relative acceleration is large, b u t is particularly difficult because of t h e focusing effects of t h e accelerating sections. T h e impact of this focusing is larger in t h e injector because of t h e lower energy of t h e beam; it is further complicated by t h e focusing components t h a t mix t h e x a n d y phase spaces. Work continues t o make improvements in this difficult tuning problem [4]. During t h e GO experiment, t h e lack of damping was overcome with active feedback on t h e beam position at the target [5]. Helicitycorrelated beam position difference measurements were made near t h e target a n d used to move t h e polarized source laser beam (using a piezoelectric actuator t o move a reflecting mirror) in a helicitycorrelated manner t o null t h e error signal. In practice, this feedback was somewhat more difficult as t h e b e a m current and motions in t h e x a n d y directions were fully coupled.
118
D.H. Beck, M.L. Pitt: Beam optics for electron scattering parityviolation experiments
e.g. changing the y position of the laser b e a m at the source changed the x and y positions at the target, as well as the b e a m current. Successful operation of the feedback system required periodic calibration of this "response" matrix, as well as some optical adjustments to insure t h a t the matrix was nonsingular.
4 Phase trombone In the most recent run of the H A P P E X experiments [6], beam position differences were mitigated using different techniques. As H A P P E X ran with the s t a n d a r d bunch charge (499 MHz pulse rate), the tuning of the injector was more s t a n d a r d and larger damping factors (though not the theoretical values) were routinely obtained. T h e position differences of the laser beam on the polarized source crystal were also reduced through a combination of a new, larger diameter Pockels cell (the electrooptic A/4 plate t h a t sets the laser beam polarization) and more careful optical alignment. In the current run, a newly developed technique was also initiated. Starting with reduced position differences at the target, the H A P P E X group, in collaboration with the J L a b Accelerator Division used a group of eight quadrupole magnets in the arc to adjust the b e t a function phase advance at the target (hence the name phase trombone) to t r a d e off helicitycorrelated position and angle differences (see Fig. 3). T h e basic idea is to change the phase advance periodically during the experiment to trade, e.g., a large position difference for a large angle difference or even to reverse the sign of a position difference to cancel the effect in an earlier part of the run. Technically, this amounts to rotating the phase space ellipse t h a t describes the beam envelope. Development of this tool is also continuing.
5 Conclusion Because the asymmetries in parityviolating experiments are small, careful control of the b e a m is required to reduce false asymmetries to acceptable levels. Therefore, in a real sense, the a p p a r a t u s for these experiments involves the entire accelerator as an integral part. Among other accelerator challenges, controlling helicitycorrelated beam positions and angles at the target at a level of a few n m and a few nrad, respectively, requires sophisticated control mechanisms. Achieving the natural (due to acceleration) damping of motion in the transverse planes is becoming easier as new sources of optical mismatches are identified. In the GO experiment, active feedback was used with some success to reduce the position and angle differences, although changing b e a m conditions required periodic adjustments. T h e present H A P P E X run has seen the beginning of development of a new optical tool to allow position and angle differences to be traded off, reducing the overall effect on the experiment. Continued development of these techniques will be important to the success of future measurements of even smaller asymmetries. We gratefully acknowledge contributions to this presentation from A. Bogacz, Y. Chao, K. Nakahara, and K. Paschke.
References 1. G. Gates: this workshop 2. P. Roos: this workshop 3. K. Wille: The Physics of Particle Accelerators: An Introduction (Oxford University Press, New York 2000) 80ff 4. Y. Chao et al.: proceedings of the 9th European Particle Accelerator Conference, 2004 5. K. Nakahara: this workshop 6. R. Holmes: this workshop
Eur Phys J A (2005) 24, s2, 119120 DOI: 10.1140/epjad/s2005040279
EPJ A direct electronic only
GO beam quality and multiple linear regression corrections Kazutaka Nakahara University of Illinois at UrbanaChampaign 1110 West Green St. Urbana, IL 61801, USA email: nakaharaOjlab. org Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The GO experiment measures parityviolating elastic ep scattering asymmetries to probe the strange quark content of the nucleon. The goal is to measure the asymmetries with an overall uncertainty of 5% of the measured asymmetries which will be of order 10~^ to 10~^. In order to achieve the above precision, systematic effects which can induce false asymmetries must be controlled. One such systematic error is helicitycorrelated changes in beam parameters, which, coupled with the sensitivity of the GO spectrometer to such beam variations, can induce false asymmetries. Beam parameters monitored for helicitycorrelation are beam current, beam position, beam angle, and beam energy. A feedback loop was successfully used to reduce the helicitycorrelation in beam current and beam position. The sensitivity of the GO spectrometer to fluctuations in beam parameters has also been measured, and the false asymmetries have been determined to be of order 10~^. This contribution will address the sensitivity of the GO spectrometer to changes in beam conditions, the performance of the feedback loops as well as the resulting parity quality of the GO beam, and the resulting false asymmetry. PACS. 29.27.HJ Polarized beams 13.60.r Photon and chargedlepton interactions with hadrons  25.30.c Elastic electron scattering
1 Introduction
contribution will detail the methods specifically used by the GO collaboration to reduce these false asymmetries.
T h e physics process responsible for the asymmetry measured in the GO experiment is the weak parityviolating amplitude in elastic scattering [1]. A = ( o  +  a _ ) / ( a + + o_)
2 Parity quality beam
(1)
where cr+/ are the elastic ep cross sections in the positive/negative helicity states of the electron beam. Thus, the measured asymmetries are susceptible to helicity correlated fluctuations in b e a m conditions. These fluctuations can alter the measured cross section in a manner unrelated to the physics process we wish to probe, and consequently appear as unwanted false asymmetries. As parityviolation experiments strive to achieve greater precision in the measurements of their asymmetries, the tolerance on helicitycorrelated fluctuations in b e a m conditions have become more stringent. However, recent developments in improving beam quality for such experiments have yielded significant results in reducing false asymmetries coming from helicitycorrelated beam fiuctnations. These developments, together with the s t a n d a r d multiple linear regression analysis techniques, have rendered such false asymmetries to be a wellcontrollable systematic effect [2,3]. Although there have been many improvements made b o t h in controlling the helicitycorrelation at the source as well as through improved b e a m transport, this
Processes which can induce helicitycorrelation in the beam may include effects such as an imperfect circular polarization of the laser, helicitycorrelated laser motion at the cathode, and beam loading effects in the accelerating cavities [4]. In the GO experiment the cumulative effect from all such processes was measured, and an active feedback device was used to null out the resulting helicitycorrelated differences. An IA (intensity attenuator) cell and a P Z T (piezoelectric transducer) mirror were used to modulate the intensity and position of the laser at the cathode in order to null out any charge asymmetry and position differences t h a t exist within the beam. Existing correlations between charge, position, angle and energy made it possible to reduce all 6 helicitycorrelated differences simply by feeding back on the above 3 parameters, although it was not guaranteed t h a t the energy difference would be reduced. A halfwave plate was inserted/retracted at the source to reverse the circular polarization of the laser every few days to monitor any further systematics effects coming from polarization reversal.
K. Nakahara: GO beam quality and multiple linear regression corrections
120
Slopes vs. Oclantt Detector 1
\
i f
^
. NSU^T
^
< NE3M IN Call OUT Co.l IM
J
Table 1. The false asymmetries from the helicitycorrelated differences as well as the sensitivity of the spectrometer
1
Beam
Helicity
Slopes (octant
False
Parameter
correlated
dependent)
asymmetry
10 05
difference
1 0 1 **0i5
E
St
'
1
[
_
I
(octantsum)
(INOUT)
4
1 I
X
6
4 nm
1 to 1 % / m m
109
Y
8
4 nm
1 to 0.8 % / m m
109
Ox 6y
2
0.3 m a d
6 to 8 % / m r a d
io«
3
0.5 nrad
4 to 5 % / m r a d
E
58
4 eV
0.01 to 0.02 % / M e V
10"^ 109
I
0.28
~103 %/nC
10"^
0.28 ppm
DDSF
>
I 0 02 I
t
n 1 1
1 11
'
t t
! 1
i L
Fig. 1. Linear regression slopes determined from natural beam motion (NBM) and coil modulation show consistent results. The beam current slopes shows a large odd vs even octant dependence due to differing deadtimes from the 2 distinctly separate electronics that were used in those octants
3 Multiple linear regression In order to determine how much false asymmetry results from the helicitycorrelation in the beam, the sensitivity of the GO spectrometer to changes in beam parameters was measured. T h e false asymmetry resulting from the helicitycorrelated parameter differences is thus, [2] Afalse = ^d/Y
* {dY/2dPi)
* SPi
(2)
where Y is the yield seen on the detectors, SPi denotes the helicitycorrelated differences in the six beam parameters, and dV/dPi is the slope which characterizes the sensitivity of the spectrometer yield to fluctuations in the beam. T h e spectrometer is composed of 8 azimuthally symmetric octants, each with an array of 16 detectors which corresponds to different Q^, so the false asymmetry contribution due to position and angular motion of the beam tends to cancel out when summed over all octants. In addition to determining the slopes with the n a t u r a l motion of the beam, a set of steering coils were used upstream of the target to modulate the b e a m by large amounts in order to gain more dynamic range in determining the sensitiv
ity of the spectrometer to beam position and angle. T h e two methods have shown consistent results throughout the run and the expected octantdependence is seen. Figure 1 shows the determined sensitivity to changes in the 6 beam parameters for one particular detector element. T h e resulting false asymmetries coming from the slopes and the helicitycorrelation in the 6 beam parameters are summarized on Table 1.
4 Conclusions T h e parity b e a m feedback system for the GO forward angle production was successfully implemented, and parity quality b e a m was achieved to the level where the resulting false asymmetry of order 0.01 p p m is considered a negligible contribution to the 5% overall uncertainty t h a t the GO experiment aims for in the determination of its parityviolating asymmetries. Acknowledgements. Thanks to Doug Beck, Mark Pitt, Matt Poelker, Joe Grames, YuChu Chao, Reza Kazimi, and the Jefferson Lab HallC and Accelerator groups for their tireless effort and support.
References 1. Douglas H. Beck, Barry R. Holstein: International Journal of Modern Physics E (2001), p. 510 2. Damon Spayde: Thesis at the U. of Maryland, Measurement of the Strange Magnetic Form Factor of the Proton using Elastic Electron Scattering, (2001) p. 55 and p. 101117 3. K.A. Aniol, et al.: Phys. Rev. C 69, 065501 (2004), p. 46 4. P.A. Souder et al.: Phys. Rev. Lett. 65, 694 (1990), p. 695697
IV Experimental techniques in PV electron scattering IV2
Polarimetry
Eur Phys J A (2005) 24, s2, 123126 DOI: 10.1140/epjad/s2005040288
EPJ A direct electronic only
M0ller polarimetry with atomic hydrogen targets E. Chudakov^^ and V. Luppov^^ ^ Thomas Jefferson National Accelerator Facility, Newport News, VA23606, USA ^ University of Michigan Spin Physics Center, Ann Arbor, MI 481092036, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. A novel proposal of using polarized atomic hydrogen gas, stored in an ultracold magnetic trap, as the target for electron beam polarimetry based on M0ller scattering is discussed. Such a target of practically 100% polarized electrons could provide a superb systematic accuracy of about 0.5% for beam polarization measurements. Feasibility studies for the CEBAF electron beam have been performed. PACS. 07.60.Fs Polarimeters  29.25.Pj Polarized targets  67.65.+Z Spinpolarized hydrogen and helium
1 Motivation Precise electron beam polarimetry will become increasingly important for the next generation of parity violation experiments. T h e systematic errors (polarimetry excluded) and statistical errors of some of these experiments will become better t h a t 0.5%. For example, the measurement of the neutron skin of the ^^^Pb nucleus, proposed at Jefferson Lab [1], requires a 1% polarimetry accuracy for the 850 MeV, 50 /xA polarized electron beam, and would benefit from a polarimetry accuracy of 0.5%. Compton polarimetry, while accurate enough at the energies > 4 GeV [2, 3] has difficulties at low energies ^ 8 0 0 MeV and a 1% accuracy has not been achieved so far. M0ller polarimetry does not depend considerably on the beam energy, but the accuracy is limited by the choice of the polarized electron target. Ferromagnetic foils, used so far, provide electron polarization of about 8%, known either with an accuracy of about 23% (see, for example [4,5,6]) if the foil is magnetized along its surface in a a field of 1030 m T , or with an accuracy ^ 0 . 3 % , if it is magnetized in a very strong field of ^ 4 T [6]. There are other systematic errors, associated with ferromagnetic targets. A kinematics difference in scattering on the external and internal a t o m shells lead to a systematic error (the socalled Levchuk effect [7]). T h e target heating limits the b e a m current to 23^A, a factor of 1030 below the typical currents needed for the experiments. Also, the dead time gives a systematic error. ^ This work was supported by the Southeastern Universities Research Association (SURA), which operates the Thomas Jefferson National Accelerator Facility for the United States Department of Energy under contract DEAC0584ER40150. Now with Janis Research Company, Wilmington, MA 018870696
W i t h all this in mind it seems very attractive to use atomic hydrogen gas, held in an ultracold magnetic t r a p [8], as the source of 100% polarized electrons. M0ller polarimetry with such a target would be free of the accuracy limitations discussed above. T h e target polarization would be close enough to 100% and there will be no need to measure it. There will be no Levchuk effect or noticeable dead time. Here, a feasibility study of such an option is presented.
2 Polarized atomic hydrogen target 2.1 Hydrogen atom in magnetic field T h e magnetic field Bs and the hyperfine interaction split the ground state of hydrogen into four states with different energies. T h e low energy states are \a) =  ^ ^ )  cos^— f^) s i n ^ and \b) = \ 4^), where the first and second (crossed) arrows in the brackets indicate the electron and proton spin projections on the magnetic field direction. As far as the electron spin is concerned, state \b) is pure, while state I a) is a superposition. T h e mixing angle 0 depends on the magnetic field Bs and t e m p e r a t u r e T: tan 20 ^ 0.05 T/Bs. At Bs = 8 T and T = 0.3 K the mixing factor is small: s i n ^ « 0.003. State \b) is 100% polarized. State I a) is polarized in the same direction as \b) and its polarization differs from unity by ^ 10~^.
2.2 Storage cell In a magnetic field gradient, a force — V ( / / H B ) , where I^H is the atom's magnetic moment, separates the lower and the higher energy states. T h e lower energy states are pulled into the stronger field, while the higher energy states are repelled from the stronger field. T h e 0.3 K cylindrical storage cell, made usually of pure copper, is located
124
E. Chudakov, V. Luppov: M0ller polarimetry with atomic hydrogen targets
beam
Fig. 1. A sketch of the storage ceU
in the bore of a superconducting ~ 8 T solenoid. T h e polarized hydrogen, consisting of the low energy states, is confined along the cell axis by the magnetic field gradient, and laterally by the wall of the cell (Fig. 1). At the point of statistical equilibrium, the state population, p follows the Boltzmann distribution: p oc exp
(fieB/kT),
(1)
where jj^e is the electron's magnetic moment (IHH ~ Me) and k = ks is the Boltzmann constant. T h e cell is mainly populated with states \a) and \b), with an admixture of states \c) and \d) of exp (2//e5/A;T) ^ 3 10"^^. In the absence of other processes, states \a) and \b) are populated nearly equally. T h e gas is practically 100% polarized, a small ( ^ 10~^) oppositely polarized contribution comes from the  t ^ ) component of state \a). T h e atomic hydrogen density is limited mainly by the process of recombination into H2 molecules (releasing ~ 4 . 5 eV). T h e recombination rate is higher at lower temperatures. In gas, recombination by collisions of two atoms is kinematically forbidden but it is allowed in collisions of three atoms. On the walls, which play the role of a third body, there is no kinematic limitation for two a t o m recombination. At moderate gas densities only the surface recombination matters. In case of polarized atoms, the cross section for recombination is strongly suppressed, because two hydrogen atoms in the triplet electron spin state have no bound states. This fact leads to the possibility of reaching relatively high gas densities for polarized atoms in the traps. A way to reduce the surface recombination on the walls of the storage cell is coating t h e m with a thin film (~50 nm) of superfiuid ^He. T h e helium film has a very small sticking coefficient^ for hydrogen atoms. In contrast. ^ The sticking coefficient defines the atom's adsorption probability per a collision with a surface.
hydrogen molecules in thermal equilibrium with the film are absorbed after a few collisions and are frozen in clusters on the metal surface of the t r a p [9]. T h e higher energy states are repelled from the storage cell by the magnetic field gradient and leave the cell. Outside of the heliumcovered cell, the atoms promptly recombine on surfaces into hydrogen molecules which are either p u m p e d away or are frozen on the walls. Some of the higher energy states recombine within the cell and the molecules eventually are either frozen on the heliumcoated wall, or leave the cell by diffusion. T h e cell is filled with atomic hydrogen from an R F dissociator. Hydrogen, at 80 K, passes through a Teflon^ pipe to a nozzle, which is kept at ~ 3 0 K. From the nozzle hydrogen enters into a system of heliumcoated baffles, where it is cooled down to ^ 0 . 3 K. At 30 K and above, the recombination is suppressed because of the high temperature, while at 0.3 K it is suppressed by helium coating. In the input flow, the atoms and molecules are mixed in comparable amounts, but most of the molecules are frozen out in the baffles and do not enter the cell. T h e gas arrives at the region of a strong field gradient, which separates very efficiently the lower and higher atomic energy states, therefore a constant feeding of the cell does not affect the average electron polarization. This technique was first successfully applied in 1980 [10], and later a density^ as high as 310^^ atoms/cm^ was achieved [8] in a small volume. So far, the storage cell itself has not been put in a highintensity particle beam. For the project being discussed a normal storage cell design can be used, with the beam passing along the solenoid axis (Fig. 1). T h e double walls of the cylindrical copper cell form a dilution refrigerator mixing chamber. T h e cell is connected to the b e a m pipe with no separating windows. T h e tentative cell parameters are (similar to a working cell [11]): solenoid m a x i m u m field of Bs = 8 T , solenoid length of Ls = 30 cm, cell internal radius of To = 2 cm, cell length of Lc = 35 cm and t e m p e r a t u r e of T = 0.3 K. T h e effective length of such a target is about 20 cm. For the guideline, we will consider a gas density of 3 10^^ cm~^, obtained experimentally [12], for a similar design.
2.3 Gas properties I m p o r t a n t parameters of the target gas are the diffusion speed. At 300 m K the RMS speed of the atoms is ^ 8 0 m / s . For these studies we used a calculated value [13] of the hydrogen atoms cross section a = 42.3 10~^^ cm^, ignoring the difference between the spin triplet and singlet cross sections. This provided the mean free p a t h i = 0.57 m m at density of 3 10^^ cm~^. ^ Teflon has a relatively small sticking coefficient for hydrogen atoms. ^ This parameter is called concentration, but we will use the word density in the text, since the mass of the gas is not important here.
E. Chudakov, V. Luppov: M0ller polarimetry with atomic hydrogen targets T h e average time, Td for a "low field seeking" a t o m to travel to the edge of the cell, assuming its starting point is distributed according to the gas density, is^: Td ~ 0.7 s. This is the cleaning time for an a t o m with opposite electron spin, should it emerge in the cell and if it does not recombine before. T h e escape time depends on the initial position of the atom, going from ^^ 1 s at z = 0 to 0.1 s at z = 8 cm. T h e average wall collision time is about 0.5 ms.
2.4 Gas lifetime in the cell For the moment we consider the gas behavior with no b e a m passing through it. Several processes lead to losses of hydrogen atoms from the cell: thermal escape through the magnetic field gradient, recombination in the volume of gas and recombination on the surface of the cell. T h e volume recombination can be neglected u p to densities o f  10^^ c m  2 [8]. T h e dominant process, limiting the gas density, is the surface recombination. In order to keep the gas density constant the losses have to be compensated by constantly feeding the cell with atomic hydrogen. Our calculations, based on the theory of such cells [8], show, t h a t a very moderate feed rate of (^ — 1 10^^ a t o m s / s would provide a gas density of 7 10^^ cm~^. This can be compared with the measurement [12] of 310^^ cm~^. T h e average lifetime of a "high field seeking" a t o m in the cell is — 1 h.
2.5 Unpolarized contamination T h e most important sources of unpolarized contamination in the target gas in absence of b e a m have been identified: 1) hydrogen molecules: — 10~^; 2) high energy atomic states \c) and \d): r^ 10~^; 3) excited atomic states < 10~^^; 4) other gasses, like helium and the residual gas in the cell:  10^ T h e contributions l)3) are present when the cell is filled with hydrogen. They are difficult to measure directly and we have to rely on calculations. Nevertheless, the behavior of such storage cells has been extensively studied and is well understood [8]. T h e general parameters, like the gas lifetime, or the gas density are predicted with an accuracy better t h a n a factor of 3. T h e estimates l)3) are about 100 times below the level of contamination of about 0 . 1 % which may become important for polarimetry. In contrast, the contribution 4) can be easily measured with beam by taking an empty target measurement. Atomic hydrogen can be completely removed from the cell by heating a small bolometer inside the cell, which would remove the helium coating on this element, and catalyze a fast recombination of hydrogen on its surface. However, it is important ^ This time was estimated using simulation, taking into account the gas density distribution along z and the repelling force in the magnetic field gradient.
125
to keep this contamination below several percent in order to reduce the systematic error associated with the background subtraction.
3 Beam impact on storage cell We have considered various impacts the I^ = 100 /iA C E B A F b e a m can inflict on the storage cell. T h e beam consists of short bunches with r = CTT ~ 0.5 ps at a JT = 499 MHz repetition rate. T h e beam spot has a size of about ax ~ cry ^ 0 . 1 m m . T h e most important depolarization effects we found are: A) gas depolarization by the R F electromagnetic radiation of the beam: ~ 3 10~^; B) contamination from free electrons and ions: ^ 10~^; C) gas excitation and depolarization by the ionization losses: ~ 10~^; D) gas heating by ionization losses: ~ 10~^^ depolarization and a ^ 3 0 % density reduction. The effects A) and B) are described below.
3.1 Beam RF generated depolarization T h e electromagnetic field of the b e a m has a circular magnetic field component, which couples to the \a)^\d) and \b)^\c) transitions. T h e transition frequency depends on the value of the local magnetic field in the solenoid and for the bulk of the gas ranges from 215 to 225 GHz. T h e spectral density function of the magnetic field can be presented in the form of Fourier series with the characteristic frequency of ujo = 27rjr. The Fourier coeflficients are basically the Fourier transforms of the magnetic field created by a single bunch. T h e bunch length is short in comparison with the typical transition frequency {ujtrans^ ~ 0.1). T h e resonance lines of the spectrum (a reflection of the 499 MHz repetition rate) populate densely the transition range (see Fig. 2). T h e induced transition rate depends on the gas density at a given transition frequency. This rate was calculated taking into account the b e a m parameters and the field m a p of a realistic solenoid. Provided t h a t the field of the solenoid is fine t u n e d to avoid the transition resonances for the bulk of the gas in the cell (see Fig. 2), the depolarization described has the following features:  the transition rate is proportional to I^;  the average rate of each of the two transitions is about 0.5 10~^ of the target density per second;  at the center around the b e a m the full transition rate is about 6% of the density per second. In order to estimate the average contamination we take into account t h a t each resonance line presented in Fig. 2 corresponds to a certain value of the solenoid field and, therefore, affects the gas at a certain z. Using a realistic field m a p of the solenoid we obtained t h a t the average depolarization in the beam area will be reduced to about ^ 0.3 10~^ by the lateral gas diffusion and by the escape of the "low field seeking" atoms from the storage cell.
126
E. Chudakov, V. Luppov: M0ller polarimetry with atomic hydrogen targets
221
222
223
224
225
T r a n s i t i o n f r e q u e n c y v (GHz)
Fig. 2. Simulated spectra of the transitions on the axis of the hydrogen trap with the maximum field of 8.0 T. The density of atoms depends on the field as exp{—jieB/kT). The two curves show j^dN/dh'ad and ^dN/dubc  the relative number of atoms which can undergo \a) ^ \d) and \b) —^ \c) transitions at the given frequency, per one GHz. The resonant structure of the spectral function of the beaminduced electromagnetic field is shown as a set of vertical bars, 499 MHz apart
tion of < 0 . 0 1 % , coming from several contributions. T h e impact of the most important of these contributions can be studied, at least their upper limits, by deliberately increasing the effect. For example, the b e a m R F induced transitions can be increased by a factor of ~ 7 0 , by fine tuning of the solenoid magnetic field. T h e contribution from the charged particles in the b e a m area can be varied by a factor u p to ~ 10^, by changing the cleaning electric field. T h e systematic errors, associated with the present Hall A polarimeter, when added in q u a d r a t u r e give a total systematic error of about 3 % [5]. Scaling these errors to the hydrogen target option reduces the total error to about 0.3%.
5 Conclusion In order to study experimentally the depolarization effect discussed, one can t u n e the solenoid magnetic field to overlap a resonance line with the transition frequency of the gas at the cell center. This would increase the transition rate by a factor of ^ 7 0 .
3.2 Contamination by free electrons and ions T h e beam would ionize per second about 20% of the atoms in the cylinder around the beam spot . T h e charged particles would not escape the b e a m area due to diffusion, as the neutral atoms would do, but will follow the magnetic field lines, parallel to the beam. An elegant way to remove t h e m is to apply a relatively weak ~ 1 V / c m electric field perpendicular to the beam. T h e charged particles will drift at a speed ofv = 'Ex B / 5 ^ ~ 12 m / s perpendicular to the b e a m and leave the beam area in about 20 /is. This will reduce the average contamination to a 10~^ level.
4 Application of the atomic target to M0ller polarimetry This feasibility study was done for the possible application of the target discussed to the existing M0ller polarimeter in Hall A at J L a b [5].The results are, however, more generic and are largely applicable to other facilities with "continuous" electron beams. T h e b e a m polarization at J L a b is normally about 80%, at beam currents below 100 //A. Scaling the results of the existing polarimeter to to the hydrogen target discussed we estimated t h a t at 30 /iA a 1% statistical accuracy will be achieved in about 30 min. This is an acceptable time, in particular if the measurements are done in parallel with the main experiment. There is no obvious way to measure directly the polarization of the hydrogen atoms in the b e a m area. T h e contamination from the residual gas is measurable. T h e rest relies on calculations. All calculations show t h a t the polarization is nearly 100%, with a possible contamina
T h e considerations above show t h a t a stored, longitudinally electronspinpolarized atomic hydrogen can be used as a pure, 100% electron polarized gas target. A thickness of at least 6 10^^ electrons/cm^ can be reached with a target diameter of 4 cm and a length of 20 cm along the beam. T h e polarized hydrogen gas should be stable in the presence of a 100 ^ A C E B A F beam. A M0ller polarimeter, equipped with such a target would provide a superb systematic accuracy of about 0.5%, while providing a 1% statistical accuracy in about 30 min of running at a b e a m current of 30 fiA .
References 1. C.J. Horowitz, S.J. Pollock, P.A. Souder, R. Michaels: Phys. Rev. C 63, 025501118 (2001), [arXivinuclth/9912038] 2. M. Baylac et al.: Phys. Lett. B 539, 812 (2002), [arXiv:hepex/0203012] 3. M. Woods [SLD Collaboration]: arXiv:hepex/9611005 4. RS. Cooper et al.: Phys. Rev. Lett. 34, 15891592 (1975) 5. A.V. Glamazdin et al.: Fizika B 8, 9195 (1999), [arXiv:hepex/9912063] 6. M. Hauger et al.: Nucl. Instrum. Methods A 462, 382392 (2001), [arXiv:nuclex/9910013] 7. L.G. Levchuk: Nucl. Instrum. Methods A 345, 496499 (1994) 8. LF. Silvera, J.T.M. Walraven: "Spin Polarized Atomic Hydrogen," in Progress in Low Temperature Physics, vol. X (Amsterdam: Elsevier Science Publisher B.V., 1986) 139370 9. LF. Silvera: Phys. Rev. B 29, 38993904 (1984) 10. LF. Silvera, J.T.M. Wahaven: Phys. Rev. Lett. 44, 164168 (1980) 11. T. Roser et al.: Nucl. Instrum. Methods A 301, 4246 (1991) 12. M. Mertig, V.G. Luppov, T. Roser, B. Vuaridel: Rev. Sci. Instrum. 62, 251252 (1991) 13. M.D. Miller, L.H. Nosanow: Phys. Rev. B 15, 43764385 (1977)
Eur Phys J A (2005) 24, s2, 127128 DOI: 10.1140/epjad/s2005040297
EPJ A direct electronic only
Progress report on the A4 Compton backscattering polarimeter Yoshio Imai^, for the A4 Collaboration Inst, fiir Kernphysik, Universitat Mainz, J.J.BecherWeg 45, D55128 Mainz, Germany Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The A4 collaboration at the Dept. of Nuclear Physics, University of Mainz, is conducting experiments on parity violation in the elastic electronnucleonscattering which require the use of polarized beams. In order to measure the absolute beam polarization, we have installed a Compton backscattering polarimeter in front of the target, using for the first time the internal cavity concept. A maximum intracavity intensity of 90 W has been measured, and in August 2003, first backscattered photons have been detected. This article describes the new design concept and the current status and results. PACS. 29.27.Hj Polarized beams  41.85.Qg Beam analyzers, beam monitors and Faraday cups  42.60.By Design of specific laser systems  42.60.Da Resonators, cavities, amplifiers, arrays and rings
1 Introduction
Table 1. Luminosity requirements for green light, Pe=0.8
T h e A4 experiment at the Mainz Microtron (MAMI) is designed to determine the strange quark contribution to the nucleon properties by measuring the parityviolating crosssection asymmetry in the elastic electronnucleon scattering with polarized beams. T h e measured asymmet r y is related to the physics asymmetry via
^.
= P. A.phys
2 Compton polarimetry T h e Compton crosssection for polarized light on polarized electrons can be written as follows [1]: dap
dai
dQ
^ dQ
da,long
vp;'""*' dn
(2)
Piong^ Ptr dcuotc the lougitudiual and transverse electron polarizations, Q, V the Stokes parameters describing the light polarization and (f the azimuthal scattering angle. W h e n using purely circular light {Q = 0), switching the light helicity {V = ) leads to an asymmetry in the spatial and energetic distributions of the backscattered photons, from which Piong ctnd Ptr can be extracted. T h e measurement time depends on the crosssection, the asymmetry, and the luminosity C via [2] t oc
1
C{a){A^) ^ comprises part of PhD thesis
10% 5% 3% 1%
C{Ee = SbbMeV) 1.15 4.59 12.76 114.86
kHz/barn kHz/barn kHz/barn kHz/barn
C{Ee = ^rOMeV) 2.51 10.05 27.91 251.16
kHz/barn kHz/barn kHz/barn kHz/barn
(1)
where Pg is the (longitudinal) b e a m polarization. For an absolute measurement of Pe? ^ Compton backscattering polarimeter using a new design concept has been installed in the A4 beamline.
da ~dQ
APe/Pe
(3)
where (a) is the detector eflftciencyweighted average of the unpolarized crosssection over the energy range, and ( ^ ^ ) the crosssection and efficiencyweighted meansquared asymmetry. Asymmetry and crosssection are mostly fixed by kinematics and available devices, so only the luminosity can be optimized. Table 1 shows the luminosity required to achieve various accuracies within 15 min in absence of background. W h e n calculating the expected luminosity for reasonable setup parameters (green laser light, 10 W of o u t p u t power), the maximum value is about 4 k H z / b a r n even for optimum light focusing. It strongly depends on the crossing angle and decreases by a factor of 20 within 20 mrad. It is therefore desireable to use a collinear geometry and necessary to increase the laser intensity.
3 Polarimeter layout One possibility to increase the available intensity is to feed the laser beam into a FabryPerot resonant cavity. This concept has been reported to work successfully [3] but is difficult to build because the small bandwidth makes a frequency stabilization of the laser necessary. T h e A4 polarimeter implements for the first time the internal cavity concept [4]: lasers already consist of a F  P cavity with
Y. Imai: Progress report on the A4 Compton backscattering polarimeter
128
fibre detector
magnetic chicane
plasma tube I quadrant diode
Fig. 1. Schematic view of the laser resonator. It is installed in a magnetic chicane in front of the A4 target
Fig. 2. ADC spectra in the Nal calorimeter with and without laser beam. The intracavity power was 49W
4 Status and results a lasing medium, and the o u t p u t light is only a fraction of the internally circulating light. Our m e t h o d is to extend the cavity length, use highreflecting mirrors on b o t h ends and guide the electron beam through the laser resonator where it interacts with the high intracavity power. Since the laser medium will adapt to cavity length fluctuations, no frequency stabilization is necessary; however, the achievable maximum power is lower t h a n with an external cavity. Figure 1 shows an overview drawing. T h e laser is an Arion laser delivering 10 W at 514.5 n m in factory conflguration. T h e lens is used to preserve the original beam proflle in the medium while optimizing it in the interaction region. T h e cavity is now 7.8 m long and therefore vulnerable to vibrations of the optical elements. Since the influence of vibrations to the b e a m axis depends on the optics spacing and the vulnerability of the luminosity to b e a m axis fluctuations depends on the focusing, MonteCarlo simulations of the effective luminosity as a function of vibration amplitude have been performed for various focusings. T h e flnal value is a focusing oi ZR = 2.5 m. with a maximum luminosity of 2.1 k H z / b a r n per 10 W of power. Also, a stabilization system for the laser beam position has been designed. T h e position is measured with quadrant diodes and stabilized using piezoactuated mirrors [5]. In the interaction region, three wire scanners measure the positions of b o t h beams simultaneously to establish overlap. T h e backscattered photons are detected in a Nal calorimeter. T h e electrons involved in the scattering lose energy and are displaced with respect to the main beam. A scintillating flbre array behind the chicane is used to detect t h e m in coincidence with the photons to improve the d a t a quality. T h e circular polarization of the light is created using two quarter waveplates, one being rot at able to select the helicity. Two waveplates are necessary because the polarization optics is installed inside a resonator. One of the vacuum windows is used as a beam splitter to measure the polarization state. T h e extracted light (0.6%) is transmitted through a rotating quarter waveplate and a GlanLaser prism; the intensity is thereby modulated with modulation amplitudes proportional to the beam's Stokes parameters.
T h e magnetic chicane has been set up in December 2002. It does not affect the beam quality on the target. The laser system has also been installed and works reproducibly. After installation of the polarization optics, intracavity intensities of up to 90 W have been measured in singleline (514.5 nm) conflguration. Procedures to bring electron and laser beem to overlap have been established, and backscattered photon spectra have been recorded with the Nal calorimeter. A calibration procedure for the detector has been established which uses muons from cosmic radiation with triggerdeflned track lengths inside the detector. T h e flbre array detector has been commissioned, and the d a t a quality was improved by imposing a coincidence condition between electrons and photons. T h e signaltonoise ratio was increased from 1:7.1 to 1:2.1. T h e laser beam stabilization system has been installed, and tests have shown a signiflcant reduction of beam axis fluctuations [5]. T h e measurement device for the laser Stokes parameters has been installed and tested. T h e next steps will comprise the improvement of the vacuum and an analysis of stress birefringence in the vacuum windows. T h e N a l calorimeter is (/^averaging and the polarimeter therefore only sensitive to longitudinal polarization. It is planned to install a positionsensitive detector to measure also the transverse polarization. W i t h this setup, a statistical accuracy of 2.5% without and about 5% with background seems to be achievable in 15 min. Systematic uncertainties can arise from detector response and the Stokes parameter measurement. We are currently working to control and minimize them. This work has been supported by the Deutsche Forschungsgemeinschaft and the U.S. DoE.
References 1. F. Lipps, H. Tolhoek: Physica X X , (1954) 8599, 395405 2. G. Bardin et ah: Conceptual Design Report of a Compton Polarimeter for Cebaf Hall A (JLab, 1996) 3. N. Falletto et al.: N.I.M. A 459, 412425 (2001) 4. M. Diiren: HERMES internal report 00005, (2000) 5. J. Diefenbach: this volume
Eur Phys J A (2005) 24, s2, 129130 DOI: 10.1140/epjad/s2005040302
EPJ A direct electronic only
The transmission Compton polarimeter of the A4 experiment A simple simultaneous monitor for the longitudinal electron beam polarisation Christoph Weinrich^, for the A4 Collaboration JohannesGutenbergUniverstitat, Institut fiir Kernphysik, J.J.BecherWeg 45, 55099 Mainz, Germany Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. A transmission Compton polarimeter as a simultaneous, relative monitor for the longitudinal polarisation of a stabilised, polarised electron beam (polarisation degree ~ 80 %) has been constructed in a new design. It is located in the vacuum between the target and the beam dump of the A4 parity violation experiment at MAML The analysing power is ~ 80 ppm at 854 MeV and ~ 115 ppm at 570 MeV. The measurement precision for the polarimeter asymmetry (which is proportional to the longitudinal beam polarisation degree) is ~ 3 ppm within 5 min. at 854 MeV and ~ 7 ppm within 5 min. at 570 MeV and the systematical error is about 0.5 ppm + 1  2 %. This simple polarimeter consists mainly of two graphite scatterers, a watercooled SamariumCobalt permanent magnet, an aluminum secondary electron emission monitor, an understructure and electronics. PACS. 29.27.Hj Polarized beams  29.27.Fh Beam characteristics
1 Introduction T h e A4 collaboration at MAMI measures (parity violating) beam spin asymmetries in elastic electronnucleon scattering. T h e transmission Compton polarimeter serves as a relative monitor for the longitudinal polarisation between the absolute M0ller polarimeter measurements and for spin angle measurements.
2 Functionality and setup In a transmission Compton polarimeter the polarised electron b e a m (with flipping polarisation direction) produces proportionally polarised bremsstrahlung in some target. T h e polarisation of the bremsstrahlung beam is then analysed by measuring the asymmetry in its transmission through a magnet, arising from the polarisation dependence of Compton scattering. T h e beam electrons superposing the bremsstrahlung b e a m have to be diminished in front of the magnet because they dilute the measured asymmetry. T h e quantitatively unknown background from electromagnetic shower production by the photons (and electrons) in the magnet make the polarisation measurement only relative. T h e A4 transmission Compton polarimeter is located in the 60 cm wide exit beam pipe between the target and the beam d u m p . Its compact design allows the b e a m to pass by and reach the beam d u m p without beeing dissipated too much. Electron background diminution is ^ comprises part of PhD thesis
achieved by scattering, i.e. spreading the electron beam to a limited extent. For space saving and background minimisation, a permanent magnet is used as analysing magnet and secondary electron emission for the transmission measurement. Dimensions and positions of the main polarimeter parts were optimised on the basis of calculations of electron scattering, bremsstrahlung and pair production and Compton scattering. Figure 1 shows the design of the polarimeter. Polarised bremsstrahlung is produced in the liquid hydrogen target and in two graphite scatterers. These also spread the electron beam and strongly reduce the number of electrons impinging the magnet. T h e analysing magnet is an axially magnetised Sm2Coi7 (Vacomax 225 H R ^ ^ ) permanent magnet. It is highly remanent and coercive and heatproof. T h e estimated electron polarisation in the magnet is about 3.2 % [1]. T h e length of the magnet has been chosen in order to nearly minimise the relative error for the measured asymmetry. T h e magnet is cooled by water of the beam d u m p cooling water circuit. A c o p p e r / C u F e P body is shrinkfitted to the magnet and conducts the heat to the cooling pipes. T h e aluminum converter is used to measure the photons t r a n s m i t t e d through the magnet. Therein the photons produce electronpositron pairs, which generate a secondary electron emission signal. T h e signal of the 1^* scatterer is used as the reference (normalising) signal for the calculation of the transmission (which is the ratio of the photon flux in front of and behind the magnet). To allow the measurement of secondary electron emission currents, the scatterers, the magnet and the converter are electrically isolated by an aluminum shielded ceramic isolator. T h e
C. Weinrich: The transmission Compton polarimeter of the A4 experiment
130
\ 6000
40
Target, 0,012 radJength liquid hydrogen
2965
"3145
f * Scatterer. 0,16radJength graphite
Cboffrig device \
362
100
50
156
2"'' Scatterer, 0,63 rad. length graphite
/
90
Converter. 1 rad. length aluminum y
Magnet, 3,6 rad. length Snfi2Co.,y
Beam dump, 0 50 cm
End flange Fig. 1. The design of the A4 transmission Compton polarimeter (all parts are symmetric around the 8
2"^ scatterer, the magnet and the converter are mounted on one adjustable ( M a y T e k ^ ^ ) aluminum carriage which can be rolled into the exit beam pipe and is kept in position by pins. T h e 1^* scatterer is mounted on a proper frame. T h e rods holding the scatterers are made of titanium for heat resistance. T h e signal cables t h a t are connected to the polarimeter pieces to conduct the secondary electron emission currents are radiation resistant in the rear ("hot") region. T h e polarimeter electronics consists of MAMI and A4 s t a n d a r d parts: amplifying currentvolt age converters close to the signal source with differential output and integrationADC and histogrammingmodules as designed for the A4 beam monitors (histogramming modules are used in timing mode). T h e signals are integrated over the 20 ms measurement gates between which the polarisation is flipped in the p a t t e r n ( + P ,  P ,  P , + P ) , the sign of P beeing chosen pseudorandomly before repetition. T h e transmission asymmetry A^^^^ = (T+  T  ) / ( r + + T ~ ) with T = SclSsi {Scisi converter/1, scatterer signal; + /  : polarisation state) is calculated offline using the timings of the (pedestal corrected) signals. A challenge was the pedestal correction of the converter signal. This pedestal is strongly drifting probably caused by activation of the material. We solved this problem by measuring the signal ratio of the converter and 1^* scatterer signals c = AScI^Ssi in switching off the beam. The converter pedestal is then calculated ^^ S% = Sc — c Ssi {Sc/sisignal averages over polarisation 4tuple). Since the signal ratio is drifting slowly as well it has to be measured regularly in order to keep the systematical error small.
3 Results T h e measured analysing power of the polarimeter is about 80 p p m at 854 MeV and  115 p p m at 570 MeV beam energy. T h e variance (precision) of the measured polarimeter asymmetry (for 5 min. runs) is ^ 3 p p m at 854 MeV and ^ 7 p p m at 570 MeV. Sensitivity to helicity correlated b e a m parameters was found to be < 30 p p m / / i m for position differences (at target), < 0.015 p p m / p p m for current
'
I 100
1 '
1
1
1
1
1
1
1
'
1 '
A" 50
y'

\
\ \ h 1
5
1
2.5
1
1
1
1
1
1
1
1
1
1
1
0 2.5 5 (Spin angle) Wien filter current / Ampere
Fig. 2. Result of the spin rotation measurement at 570 MeV with a cosinefit. The ordinate is the current of the Wien filter dipole which is proportional to the spin angle. The abscissa is the polarimeter asymmetry
asymmetries and negligible for energy (preliminary, probably overestimated values at 570 MeV). Averaged over a d a t a sample (~ 500 runs) typical position differences are r^ 50 nm, typical current asymmetries are less t h a n 1 ppm. T h e signal ratio measurement error (used for pedestal correction) contributes as a systematical error of 12 % of the measured asymmetry. Runwise decorrelation of the measured polarimeter asymmetry of the beam parameters (based on the polarisation 4tuples) had negligible effect under normal b e a m conditions. Figure 2 shows the result of a spin rotation measurement using a Wien filter. Spin rotation measurements were also used to determine the spin angle at the experiment.
References 1. M. Huth: personal communication
Eur Phys J A (2005) 24, s2, 131131 DOI: 10.1140/epjad/s200504031l
EPJ A direct electronic only
Stabilization system of the laser system of the A4 Compton backscattering polarimeter Jiirgen Diefenbach^ Institut fiir Kernphysik, Universitat Mainz, J. J. Becherweg 45, D55099 Mainz, Germany Received: 1 November 2004 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The A4 Compton backscattering polarimeter is used to measure the degree of spin polarization of the 855 MeV electron beam provided by the MAMI accelerator facility. Therefore special care must be taken to optimize and stabilize the overlap of electron and laser beam in order to get the highest luminosity for shortest measurement times. For this purpose an active stabilization system for the laser beam position of our intracavity polarimeter's optical resonator has been developed and commissioned. PACS. 07.50.Qx Signal processing electronics  29.27.Hj Polarized beams  42.60.By Design of specific laser systems  42.60.Lh Efl^iciency, stability, gain, and other operational parameters
1 Introduction T h e laser resonator of the A4 backscattering polarimeter has a total length of 7.8 m. Therefore the proper alignment of the laser mirrors is sensitive to any mechanical instabilities caused for example by seismic vibrations, the flow of cooling water through laser t u b e and dipole magnets, air turbulences etc. Furthermore there are slow drifts of the laser pointing stability due to small changes in the t e m p e r a t u r e of the laser's plasma tube, e.g. from changes in the t u b e current. All this affects the laser beam position and leads to variations of the overlap between electron and laser beam and of the laser power, b o t h causing changes in the luminosity of the polarimeter.
and the attached laser mirror, and has been measured as a transfer function. T h e n an active polezero cancellation circuit was t u n e d carefully to extinct electronically the resonance structure of the piezo platform measured before. T h e —6 dB bandwidth of the disturbance rejection of the stabilization system in a test setup has been increased from 210 Hz to 800 Hz by this method. T h e system is fully remotely programmable using MAMI s t a n d a r d electronics. T h e elements of the decoupling matrices and the control loop amplifications are respectively given by eighteen and six 12bit multiplying DACs.
3 First results and outlook 2 Hardware These changes in the laser beam position and intracavity power have been measured using quadrant photodiodes installed along the laser cavity. To suppress the changes in the beam position, three of the four laser mirrors are mounted on piezo driven platforms. These platforms can be used to optimize the laser resonator alignment so t h a t the position of the laser beam and consequently its power can be stabilized. From the measured b e a m positions, tilt angles for the piezo mounted mirrors are calculated in the decoupling circuits. An automatic shutdown circuit protects the syst e m from resonant oscillations t h a t could cause damage to the piezo platforms. A passive singlepole lowpass filter at 35 Hz provides the control loops with stability. T h e mechanical resonance of each piezo platform depends on the geometry and the m o m e n t u m of inertia of the platform ^ comprises part of PhD thesis
T h e stabilization system has been operated so far for test purposes with four out of its six control loops running. W h e n reconceived in a naive picture, where all beam position noise is projected onto tilting vibrations of the (so far) two stabilizable piezomounted laser mirrors, one can compute the mean tilt amplitudes of the mirrors from the obtained laser beam position d a t a and decoupling matrices. Assuming this, the eflFective tilt amplitude was found to be reduced from about 90 ytxrad without stabilization to about 10 /irad with the stabilization system running and the pointing stability on the diodes from 8 to 12 /irad without to 0.8 //rad with stabilization. Longterm drifts of the beam went down from u p to 21 / / m / h to 0.3 /xm/h. Furthermore it should be possible to optimize the performance of the stabilization system by retuning the polezero cancellation circuits and then increasing the loop amplifications. Also a DAC module will be designed to add offsets to the position signals and thereby shifting the laser beam to provide finetuning of the overlap of b o t h beams.
Eur Phys J A (2005) 24, s2, 133133 DOI: 10.1140/epjad/s2005040320
EPJ A direct electronic only
Electron beam line design of A4 Compton backscattering polarimeter Jeong Han Lee^ Johannes Gutenberg Universitat Mainz, Institut fiir Kernphysik, J.J. Becherweg 45, 55299 Mainz, Germany Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. The new beam hue has been built to measure the polarization of the electron. The transverse position and emittance of the MAMI beam are determined from measurements of the beam size on three wire scanners. The position measurement has been done and the emittance measurement is still progressing. PACS. 29.27.Eg Beam handling; beam transport  29.27.Fh Beam characteristics
1 Introduction T h e A4 collaboration measures the parityviolating asymmetry for elastic electron scattering off an unpolarized proton. To extract the physics asymmetry, it will be crucial to determine the polarization of the electron beam to high accuracy. For this reason, Compton polarimeter has been installed. T h e one of advantages of Compton polarimeter is t h a t it is not a destructive method. T h a t means the continuous online monitoring of the electron polarization would be possible with the same electron beam conditions.
2 Beam conditions in the chicane A b e a m line, which is called a chicane, was designed to measure the polarization of the electron. T h e chicane has the same four dipole magnets, which separate the scattered photons from the electrons, and contributes to the dispersion functions due to the magnets. T h e dispersion must be eliminated to preserve the same beam properties by introducing two quadrupoles. T R A N S P O R T [1], which uses a matrix formalism to design a beam transport system, was used to design the chicane. T h e chicane has been built and is fully operational. T h e motion of particles b e a m can be represented by the ellipse (e.g., Twiss) parameters, /3, a, and 7. For transport lines, there is no periodic boundary condition about the parameters, which depend on the initial b e a m profiles. However, although the chicane is the t r a n s p o r t line, two requirements must be fulfilled to achieve the good luminosity and to keep the same beam properties at the beginning and at the end of the chicane; the first one is t h a t the electron beam has a waist at the middle of the chicane(a^t; = 0 and P^ = 1/^w) and the second one is the Twiss parameters are the same at the beginning ^ comprises part of PhD thesis
and end of the chicane(/3/ = /?^, a / = a^, and 7 / = 7^). Moreover, it is necessary for the good luminosity to determine the Rayleigh length of the electron beam. T h e Rayleigh length can be determined by using the equation X = \/e{P — 2as \ 75^), where e stands for the emittance, s denotes the location of the beam, and x is the beam size at s. T h e design value of the Rayleigh length is 9.617 meter.
3 Position and emittance measurement To make the scattering between the electron and the photon possible, it might be necessary to determine a transverse position of b o t h of the beams. Moreover, to get the actual values of the Twiss parameters and the emittance of the electron beam [2], it would be needed to measure the size of the beam. T h e achievement of the two purposes will be done by three wire scanners t h a t are the most commonly tool used to diagnose a b e a m profile in the middle of the chicane. However, the wire scanner, which A4 uses, is not the common type but the fourbar linkage type. T h e equation of motion of the wire scanner is therefore needed. T h e analysis of the wire scanner and the beam position measurement were done. T h e horizontal and vertical emittance of the MAMI electron beam are Ch = 7.7607r x 10~^mm m r a d and Cy = 1.0177r X 10~^mm mrad. T h e actual values of Twiss parameters and the emittance can be deduced from the measured electron beam profiles. It is a challenge to make t h e m coincide with their design values.
References 1. D.C. Carey, K.L. Brown, F. Rothacker: SLACR0530 2. M.G. Minty, F. Zimmermann: Measurement and Control of Charged Particle Beams (Springer, 2003), 99
IV Experimental techniques in PV electron scattering IV3
Detection
Eur Phys J A (2005) 24, s2, 137140 DOI: 10.1140/epjad/s200504033y
EPJ A direct electronic only
Background substraction in parity violation experiments J. Van De Wiele and M. Morlet Institut de Physique Nucleaire/IN2P3/CNRS, F91406 Orsay Cedex, France Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. The importance of the knowledge of the background in parity violating (PV) experiments is shown. Some improvements in Monte Carlo simulations are presented and discussed. PACS. 01.30.Cc Conference proceedings  25.30.Bf Elastic electron scattering  13.60.Le Meson production  13.60.r Photon and charged lepton interactions with hadrons
It is well known t h a t the asymmetry in P.V. experiments due to the exchange of the ZQ boson is small ( ^ 10~^ — 10~^). Much care has to be taken in the measurement of such a small quantity. Since a few years, impressive improvements in technical aspects have been achieved and some of t h e m have been presented in this workshop. W i t h o u t such improvements, the extraction of the physical quantity would be obtained with too large a systematic error and so would be meaningless. In any experiment, simulation of all the processes which populate the "good" events as well as some "background" events, is a good tool to be sure t h a t the experiment and in particular the experimental setup is under control. Simulation of very small effects is not an easy task for many reasons:  T h e accuracy of the simulation depends strongly on the statistics and s t a n d a r d methods, which are time consuming, therefore may become inefficient.  Some of the physical effects, which are usually considered as small and therefore are neglected, may contribute.  It is necessary to improve the description of some processes which are usually treated only in an approxim a t e way.  Accurate models and d a t a needed do not exist. T h e basic formula is the following: iPhys
AMeas
A  — /
V.
Fig. 1. Typical measured spectrum. X is a measured quantity energy or time of flightTable 1. Ratio Aphys/^Meas
as a function oi AI/AQ
for cfi/cfQ
=10% Ai/Ao
Aphys/AM eas
3. 2. 1. 0. 1. 2. 3.
0.63 0.73 0.82 0.91 1.00 1.09 1.18
1 + Ei^i/^O ^i/^o) 1 + E^i/^O
In this expression the index " 0 " stands for the elastic events and the index "i^^O" stands for any background event. If the background asymmetry vanishes, the denominator acts as a dilution factor to the physical asymmetry. As can be seen in Fig. 1, any measured spectrum of a physical quantity (energy or time of flight) shows u p a the physical signal above some background events. In Table 1 Aphys/AMeas ^rc givcu as a function of AI/AQ if
the ratio of bad events to good events is equal to 10 %. In Fig. 2, this effect is plotted for different contamination rates. T h e change in the sign of Ai/Ao could be dramatic. In principle, it is possible to experimentally study the background but the statistical precision could be poor as compared to the elastic peak. In some cases, it is possible only to extrapolate the the background and the experimental study becomes less efficient (see Fig. 3)
J. Van De Wiele, M. Morlet: Background substraction in parity violation experiments
138
" 1 r^ T ^ M l
Ml
1 '
" i
1.5
y^—
0 ri
U.J
^
1 ^.
^
'

0.5
^

U
a ^
WA \
1 1 . 1 5
4
3
Fig. 2. Ratio AphysjAu^as values of di/ao
111
1
I
II
Ml 1
1 3
A
?
as a function of AijA^
for several
Monte Carlo (M.C.) simulations are supposed to be a powerful tool to understand both the elastic observables and the background features. They give complementary information to the measurement. This method is powerful if the physical laws under consideration are well known. Monte Carlo results are accurate if we are able to generate a large number of events. M.C. is based on a onetoone correspondence between uniform random number rj G ]0,1[ and a physical law. For example, in one dimension
Fig. 3. Different extrapolations of the background where £, A^^ are respectively the luminosity and the number of random events. [AOi] = Oimax — ^imin is the angular range for the polar angle and [^0i] = (t>imax{^i) — 01mm (^i) is the angular range for the azimuthal angle. Example 2:
a \ b
1 +
2+3
^Imin
'^Imin
[Adi
[Ml El — Eimin{Ol)
V
v
N
f{x') dx\
V:
f{x')
dx\
[AE,
and in two dimensions:
M
n. = ^ .
w
pXlmc
M = /
h{A)dx'^,
Di=/
dh ^ [AO,] [A^,] [AE,] sm Ui. df2i dEi N^
h{x[)dx\
[AEi] = EimaxiOi)  EiminiOi) is the energy range. The two methods are equivalent when these weights are corrX2 rectly introduced. with fi{xi) = / / ( ^ i , 4 ) dx2, and r]^ Another method to improve the efficiency of the simuV2' ^X2min{xi) lation is to calculate the crosssections of the needed processes. In the following part, we will concentrate on the A4 and the G^ experiment and will give some specific with PX2 rX2max{xi) examples. In the A4 experiment [4], the rate of inelastic elecAf2= f{xi,X2) dx'2, V2= f{xi,X2) dX2. ^X2min(xi) ^ X2min(xi) tron produced by one pion electroproduction e + p —> Using the definition of M.C. method is time consuming e' + y + TT^ and e \ p —> e' \ n' \ 7r+ reactions has because we need to invert the expression in the numerator been measured and calculations with effective lagrangians to get the physical quantity. To increase the efficiency of for Eg lower than 1 GeV are accurate enough. It is more the M.C. method, it is possible to introduce some weights difficult to calculate the number of photons coming from the TT^ decay. With an electromagnetic shower calorimeW to take into account of the crosssections. ter, it is impossible to disentangle electrons and photons. Example 1: a + 6 —> 1 + 2 A part of these photons have the same energy as the elastic electrons and thus contribute to the background under ^1 ^Imin y^lmin the elastic peak. The standard method  calculation of the V, [Z10, [A0i TT^ electroproduction followed by TT^ decay after Lorentz boost  is not appropriate to calculate the rate of photons if we want to take into account energy loss and ex«. = A[^,.,(^*.l^s.„«,. ternal radiative corrections. Furthermore, it is impossible ^1
"
Xlmin
J. Van De Wiele, M. Morlet: Background substraction in parity violation experiments e
= 0.6b0 GeV 1
'
'
1
'
1
1
1
:
1
1
1
1 p
—>
e
p
139
f X
1
^ 
Gp=60 deg.

0p=6O (deg)
LO
yx
Solid line
.
.
1
......

X
. . . . . . ... 7^^"^"^^,..
/C =
I
IC S{E^E^)
dQ^rQ dE^
1 4^
E^
Ery =^
PT^OU^O
U.
7
dQ^o dE^
E^
p^oU^o
U
T h e TT^ differential crosssection, which depends on the helicity of the incident electron beam, is calculated through the models described above. From such accurate calculations, we conclude t h a t the photon contamination does not depend on the helicity and acts only as a dilution factor. We note here t h a t the inclusive pion or proton crosssections has to be calculated by replacing the usual flux factor r which is divergent when 6e' ^ 0 and mg = 0 by
T: a
Orsoy
Dashed line
Oconnel
E'^E^
1
r
a 87r2
Q.:'H
g.:?/h
g.j
i.i/b
r.a^
T, (GeV)
to estimate the false asymmetry carried by the photons. Nevertheless, it is possible to include all these effects if we calculate directly the photon production crosssection (the lagrangian for the TT^ —> 27 is known):
dEj
DMled lirte rr" p
y.??b
Fig. 4. Inclusive proton differential crosssection at 650 MeV
d^a
Orsoy

^lOtv)
dQj
Totol
Tolo" ccnLritutron (ScliJ line)
Rflol pl^Qton contribution {Dotled lire]
.
I,
t^^^ virt
T h e expression of Fyirt is given in [1]. In G^ experiment phase I, with an electron b e a m at 3 GeV, we are interested to know the inclusive pion and proton crosssection. Unfortunately, at such an energy, the crosssection is not well known and the calculations performed with effective lagrangians on one hand, accurate at energies below 1 GeV, and with Regge models on the other hand, well suited at energies greater t h a n 56 GeV, are not very reliable. LightBody and O'Connel have devel
Fig. 5. Comparison between our calculation and the EPC code developed by Light Body and O'Connel oped a code (EPC) to compute such crosssections but it is based on high energy d a t a and the validity of the extrapolation to our energy is questionable. We have developed a code with an other approach. T h e inclusive spectrum is obtained by integrating the five times differential crosssection over the electron angles. Because of the exchange of a virtual photon in the electroproduction reaction, the main contribution in the integration is expected to come from the terms with small values of q^. We then assume t h a t the square of the matrix element may be extracted from photoproduction measurements which exist between 200 MeV and 3 GeV [2]. We have checked this approximation at an energy of 650 MeV where an exact calculation with an effective lagrangian can be used. T h e results are displayed in Fig. 4. T h e agreement between electroproduction and photoproduction is better t h a n 5%. An event generator based on this model has been written for the G^ collaboration. T h e angular distribution of one pion photoproduction d a t a have been included. For two pion (or more) photoproduction reactions, there are only few angular distribution d a t a but several measurements of the total crosssection exist. All the channels u p to 3 pions have been included. A comparison with the E P C code is shown in Fig. 5. T h e Time of Fhght (TOE) proton spectrum includes, in addition to the electroproduction, some contribution from photoproduction reactions. This photoproduction is due to the competition, in any material, between electroproduction of the incoming electron and the real bremsstrahlung photons. T h e rate of inelastic protons is proportional to the number of bremsstrahlung photons, more precisely to I^{Eo, E^^t) which is the number of photons in the energy bin E^, E^\dE^ after an electron, initially with an an energy E'o? has passed through a target of thickness t measured in unit of the radiation
J. Van De Wiele, M. Morlet: Background substraction in parity violation experiments
140 —
0.0'^
0.0'1
—
•
—
.
—
1
—
1
—
1
—
—
1
—
•
—
•
—
1
—
1
—
1
—
1
—
1
—
1
—
—
—

1
'
proximation, with g{y) = 1. Within this approximation, Tsai and Van Whitis have derived an analytical expres
> "

[I^{Eo,E^,t)] approx
0.01

0.009
o.ooa
0.007
^

^
^
^
\r
^^ 
\
0.006
O.OOD
:
:
1
.
.
:
1
:
.
1
.
.
.
1
:
._JJ
.
T,(GeV)
Fig. 6. Value of I^{Eo, E^, t) for a LH2 target (t=20 cm, Eo= 3 GeV) as a function of E^\ Approximate formula {dotted line), complete screening approximation {dashed line) and exact calculation {solid line) length of the material [3]. This number of photons is also proportional to the bremsstrahlung crosssection:
^iE,E,)
=
1 / 2
4
4\
, ,
E^
where XQ is the radiation length of the material. In the s t a n d a r d calculations, we assume a complete screening ap
1 E^
(1
EQ ^
ln(l
:)
As was stated by Y. Tsai, this approximate expression may be a poor if the detailed shape at the highenergy tip of the Bremsstrahlung is needed. This is our case because highenergy inelastic protons, produced by highenergy photons, will have the same T O F compared to elastic protons. We have performed some exact calculations of I^{Eo,Ej,t) and new expressions have been derived for hydrogen and aluminium. For this material, we have used the ThomasFermi Moliere Model [3]. Comparison between the approximate expression, the complete screening case and the exact calculation for liquid hydrogen is shown in Fig. 6. T h e approximate formula or the complete screening calculations overestimate the highenergy number of photons.
References 1. S. Ong, J. Van de Wiele: Phys. Rev. C 63, 024614 (2001) 2. P. Corvisiero et al.: Nucl. Instr. Meth. A 346, 433 (1994) 3. Y.S. Tsai: Rev. Mod. Phys. 46, 815 (1974); Y.S. Tsai, Van Whitis, Phys. Rev. 149, 1248 (1966) 4. F.E. Maas et al.: Phys. Rev. Lett 93, 022002 (2004)
Eur Phys J A (2005) 24, s2, 141141 DOI: 10.1140/epjad/s200504034x
EPJ A direct electronic only
Redesign of the A4 calorimeter for the measurement at backward angles Boris Glaser^, for the A4 Collaboration Institut fiir Kernphysik, Johannes GutenbergUniversitat Mainz, J.J.BecherWeg 45, 55099 Mainz, Germany Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. Backward angle measurements of the parity violating asymmetry in elastic electron proton scattering are planned with the A4 calorimeter. At present the experiment measures at forward angles. For the measurement at backward angles the support of the PbF2detector has to be redesigned. For this purpose it will be positioned on a rotatable platform, so that measurements at forward and backward angles will be possible with little effort. We will highlight the new experimental setup and its special features. PACS. 07.05.Fb Design of experiments  13.40.Gp Electromagnetic form factors
1 Introduction Backward kinematics measurements in parity violating electron proton scattering on hydrogen and deuterium are planned with the A4 calorimeter. For this purpose the support for the PbF2calorimeter will be positioned on a rotatable platform, so t h a t measurements at forward and backward angles will be possible.
Angular Acceptance Calorimeter Scattering Chamber
v4
2 The calorimeter
3 The scattering chamber and the luminosity monitors Figure 1 shows the solid angle of the crystals of the calorimeter and of the luminosity monitors. An elongation chamber for the main scattering chamber will be installed, so t h a t the luminosity monitors have in forward and backward scattering position the same solid angle with respect to the target centre (Fig. 1). Comprises part of diploma thesis
w
Luminosity Monitors
Target Calorimeter
T h e A4 Calorimeter is at present mounted on a linear moveable platform. T h e support is used to adjust the focus of the crystal calorimeter on the target. A special requirement for a possible backward angle setup comes through the fact, t h a t not only the detector but also the cryogenic hydrogen target with its scattering chamber and the luminosity monitors have to be adjusted. A new support for the calorimeter and the scattering chamber is needed.
Angular Acceptance Luminosity Monitors
Scattering Chamber Elongation
Fig. 1. New Scattering Chamber. The electron beam comes from the left
4 The new experimental setup We have designed a new, common support structure so t h a t the calorimeter and the scattering chamber are supported on a rotatable platform. To ensure a vibration free, shock free and easy positioning of the platform, it is supported on three hydraulic oil sliding feet. By its small stroke (1/10 mm) contrary to air cushions (stroke some m m ) , an easy adjustment without impacts during set off is ensured. Vibrations during rotating, t h a t can damage the fragile crystals, are minimal too. T h e support platform, the scattering chamber elongation and the hydraulic oil sliding feet are already delivered and will be installed in December 2004. T h e new setup will be in operation by end of February 2005.
Eur Phys J A (2005) 24, s2, 143143 DOI: 10.1140/epjad/s2005040359
EPJ A direct electronic only
Performance of the G superconducting magnet system Steven E. Williamson, for the G^ Collaboration University of Illinois, with support from the U.S. National Science Foundation under grant PHY9410768, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. At the heart of the G° Spectrometer is the toroidal superconducting magnet system (SMS). The SMS has been in use at Jefferson Lab since the fall of 2002. Experience with the operation and reliability of the magnet over that period is reported. Some measured performance parameters are compared with the magnet specification. PACS. 29.30.Aj Chargedparticle spectrometers: electric and magnetic
T h e G^ superconducting magnet system (SMS) is an ironfree toroid with zero magnification optics. Its field, peaking at 3.5 T (3 T in conductor) is generated by eight coils, each with 144 t u r n s (4X36 windings), in a common cryostat. T h e stored energy is 6.6 M J at the normal operating current of 5 kA. Coil locations were measured at room t e m p e r a t u r e after installation at Jefferson Lab using photogrammetry to locate 128 targets (16 on each coil). Design and measured target locations were compared, while adjusting the overall position and orientation of the magnet for a best fit. T h e average deviation of measurements from the ideal was found to be 1.6 m m , less t h a n the 2.0 m m specification. T h e locations of the coils, when cooled, were deduced from known coefficients of thermal expansion. A measurement of the Q^ associated with each focal plane detector was extracted [1] from the difference between the timeoffiight of elastic protons and of TT"^ particles. This difference is sensitive to the particle trajectory through the magnet and thus to the magnetic field configuration. Measurements were compared to a simulation based on the design magnetic field and found to agree to a precision of 100 ps, which implies an uncertainty on Q^ within the 1% requirement of the experiment. T h e SMS cooldown, specified to take 7 days, actually required about 21 days. This rate was limited by the requirement t h a t Z\T between inlet and coil average be < 75 K. Heat load to LHe was specified to be < 40 W, but boiloff studies indicate a load of about 107 W. T h e steadystate LHe requirement of the magnet at full power was found to be about 8 g/s, consistent with the measured heat load with some additional load from the supply lines. During a fast d u m p of magnet stored energy, the current decays with a 10.4 s time constant into the 0.05 i? d u m p resistor. This implies an inductance of 0.52 H which matches the design inductance of 0.53 H. Redundant quench protection systems, a "digital" system ( D Q P ) , which relied on the operation of the control system pro
grammable logic controller (PLC), and an independent "analog" system (AQP), were used to trigger a fast d u m p when a quench was detected. The D Q P initially suffered from the failure of series "safety" resistors on voltage t a p s due to thermal cycling. Circuitry was added to detect broken resistors. For each coil, a b a t t e r y provided an isolated current, which circulated through the coil and adjacent voltage t a p safety resistors. Diodes were used to ensure t h a t the isolated current was only seen by the corresponding input stage to the D Q P . Offsets voltages produced by the b a t t e r y current were measured and subtracted by the P L C software. The absence of the offset voltage was the signature for a broken resistor. After the first commissioning run (October 2002 to J a n u a r y 2003), the safety resistors were relocated outside of the cryostat. About 160 of the 3270 hours of available d a t a collection time during commissioning and production running were lost due to magnet problems. This is about 48% of all lost d a t a collection time. Most (70.3%) of the magnet problems were caused by radiation damage to control syst e m components. A typical failure began with a halt of P L C program execution due to a radiationrelated memory error, which caused the "heartbeat" interlock to open. This shut down the power supply. A transient at the start of the shutdown caused the A Q P to erroneously detect a "quench" and initiate a fastdump. Eddycurrent heating then evaporated LHe in the coils and reservoir requiring a minimum 2.5hour recovery time. LHe supply and return problems were the second largest cause (18.7%) of magnet related lost time.
References 1. G.Batigne: Proceedings of the Fourth International Conference on Perspectives in Hadronic Physics, Trieste, May 1216, 2003
Eur Phys J A (2005) 24, s2, 145145 DOI: 10.1140/epjad/s2005040368
EPJ A direct electronic only
Cherenkov counter for the G^ backward angle measurements Benoit Guillon, for the G^ collaboration Laboratoire de Physique Subatomique et de Cosmologie, 53 avenue des Martyrs, 38000 Grenoble, France Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The G^ program consists of a set of parity violation experiments performed at Jefferson Lab (Va, USA) dedicated to the determination of the strange quark contribution to the charge and magnetization distributions of the nucleon [1]. This paper describes the final design of the Cherenkov counter used to reject the significant charged background of the G^ backward angle measurements. PACS. 29.40.Ka Cherenkov detectors
1 Introduction T h e main focus of the G^ experiment is to measure the neutral weak form factors G^ and Gf^ of the nucleon, over a large m o m e n t u m transfer range. This will allow us to determine the strange quark contributions to the charge and magnetization densities of the nucleon. For this purpose, parityviolating asymmetries in elastic ep scattering have already been measured at forward angles over the Q^ range 0.11 (GeV/c)'^, and will be measured at backward angles for Q^ values of 0.3, 0.5 and 0.8 (GeV/c)'^. At backward angles we also measure quasielastic scattering from a LD2 target to extract precisely the axial form factor [1]. However, negatively charged pions, as well as their decay products (//~), will produce a significant background to the elastic and quasielastic rates detected by the G^ spectrometer. Aerogel Cherenkov counters aflFord the best ir/e discrimination at these energies and allow implementation in the current G^ setup. L P S C Grenoble has designed and constructed half of the 8 Cherenkov counters needed.
2 The aerogel Cerenkov counter T h e aerogel refractive index (n = 1.03) was fixed by the m o m e n t u m distributions of the background. It should be less t h a n 1//?, but as large as possible to maximize light yield. T h e aerogel was supplied by M a t s u h i t a Electric Works in the format of tiles (113x113x10 mm^). T h e geometry of the counter and the necessary number of photomultipliers ( P M T ) was studied by using simulations [2] (Geant, Litrani) and validated with measurements made on a prototype. To allow the mounting of the aerogel in France and a safe transportation to Jlab, the Cerenkov box (see Fig. 1) is composed of two parts. T h e first part is the aerogel radiator which is 5 cm thick, and the second one is the lightbox itself. To ensure the best photon collection
Affrocjffl
i
*
I*n ] 'j
nillitjort." I'ui>c^iLight Box
nun«iBi
Fig. 1. Cherenkov design and magnetic shielding
using 4 P M T s , the inner parts of these two boxes are covered by three layers of diflFuse reflecting paper (Millipore). T h e XP4572b 5 inch tubes (Photonis) are very sensitive to magnetic fields. As they will be located in the fringe field of the G^ magnet (4.4 m T in the axial direction and 11 m T in the transverse one), eflicient magnetic shielding is needed. T h e design finally retained after tests at the LCMI Laboratory in Grenoble is made of three layers of 2 m m soft iron separated by 2 m m of air, and one layer of 0.8 m m of //metal. During these tests, one had to find the best compromise (light yield and field) for the positioning of the tubes inside the shielding. It was found t h a t a backoff of 15 cm was suflftcient. The 4 Cerenkov counters have been also successfully tested with cosmic rays and typical total numbers of 8 photoelectrons were measured with a collection time lower t h a n 25 ns.
References 1. D.H Beck, the G^ collaboration: Jeff"erson Lab proposal E04115 (originally E91017) 2. G. Quemener, S. Kox: G^ Report GO0052 (2000)
Eur Phys J A (2005) 24, s2, 147147 DOI: 10.1140/epjad/s2005040377
EPJ A direct electronic only
A binperbin deadtime control technique for timeofflight measurements in the G experiment: The differential buddy L. Bimbot, for the G^ Collaboration Institut de Physique Nucleaire d'Orsay, BP n° 1, F91406 OrsayCedex, France Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The general principle is presented. The application to the G° experiment [1] was enabled by the specificity of the time encoding ASIC component. On the poster, some encountered difl^iculties are exposed, together with a possible software remedy. An internal report is in preparation [2]. PACS. 01.30.Cc Conference proceedings  25.30.Bf Elastic electron scattering  07.05.Dz Control system  07.05.Hd Data acquisition  07.05.Kf Data analysis
Since PAVI02 the G^ experiment successfully passed its commissioning and the first phase of d a t a taking at forward angles [3]. This poster contribution aims to depict an experimental approach to control the deadtime in the timeofflight histogramming of events. Every part of the electronics  Constant Fraction Discriminator (CFD), Mean Timer (MT) and Time to Digital Converter (TDC)  has its own deadtime contributing. Different types of events must be considered: single C F D (only one of the left or right C F D s sees an event), single M T (only one of the Front or Back M T s sees an event) and good events (4 C F D signals leading to T D C encoding). Only the last type is subject to the differential buddy t r e a t m e n t . T h e correction for deadtime from other types of events is done with the help of a slow, event by event, acquisition made in parallel with a fastbus setup, and giving the probability of each kind of event [4] . T h e s t a n d a r d procedure of linear regression [5] for helicity correlated b e a m properties will help to remove residual deadtime effects. T h e idea, here, is to measure the loss of good events resulting from acquisition blocking. T h e direct counting of these events is impossible because they occur when the electronics is busy, analyzing a previous event, but an image can be obtained by checking, for each event if an associated detector (buddy), supposed to count identically, is busy or not. This is experimentally possible because it involves two different channels. However, as one needs time to process the signals, the comparison is made after a delay corresponding exactly to one pulse of beam duration. There are two basic assumptions: t h a t counting is identical despite a spatial rotation (180°) and a time translation (32 nsec). After solving tuning and analyzing problems, it is possible to use the d a t a from the buddy histograms to estim a t e the deadtime of each detector and its stability (see figures below). T h e deadtime is the ratio, within proton
Fig. 1. Measured deadtime according to detector number
Fig. 2. Evolution of measured deadtime for specific channels peak limits, of the corrected number of events having encountered the b u d d y busy, to the total number of events. Special t h a n k s to J.C. Cuzon, A. Gauvin, H. Guler, G. Quemener, J. Lenoble and R. Sellem for discussions and technical assistance.
References 1. see http://www.npl.uiuc.edu/exp/GO/GOMain.html 2. http://www.npl.uiuc.edu/exp/GO/docs/docs.html, Internal report GO04009 3. P.G. Roos: talk at this workshop 4. G. Batigne: Thesis, LPSC 0341, 2003, p. 147172 5. http://gOweb.jlab.org/manual/Analysis_manuals.html, A. Biselli et al.: G° Analysis replay engine, p. 29
V
Hadronic structure... and more
V1 Test of the SM at low energy
Eur Phys J A (2005) 24, s2, 151154 DOI: 10.1140/epjad/s2005040386
EPJ A direct electronic only
A precise measurement of siri^Ow at low Q^ in M0ller scattering Antonin Vacheret and David Lhuillier, representing the E158 Collaboration CEASaclay, Dapnia/SPhN, 91191 GifsurYvette, France Received: 15 December 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The E158 experiment has performed the first measurement of the parityviolating asymmetry in electronelectron (M0ller) scattering using a 50 GeV polarized electron beam and a fixed unpolarized liquid hydrogen target in End Station A at SLAG. Our preliminary results is: Apv = —128^ 14{stat)^ 12(syst) x 10"^. From this quantity we extract sin^ OW{Q'^ = 0.026GeV/c^)j^ = , consistent with the Standard Model prediction. PACS. ll.30.Er Gharge conjugation, parity, time reversal, and other discrete symmetries  12.15.Lk Electroweak radiative corrections  12.15.Mm Neutral currents  13.66.Lm Processes in other leptonlepton interactions  13.88.+e Polarization in interactions and scattering  14.60.Gd Electrons
1 Introduction and physics motivation In the scattering of longitudinally polarized electrons from unpolarized targets, the reversal of the helicity of incoming electrons is equivalent to applying the parity symmetry. Hence the quantity Apy = {o'R—(7L){aR\aL)j where a^^L is the cross section for incident right (left)handed electrons, is a pseudoscalar arising from the parity violating part of the interaction in the scattering process. To first order this corresponds to the interference of the neutral weak and electromagnetic amplitudes [1]. W h e n considering the M0ller process, Apy is proportional to the electron's weak charge, written at tree level as Q^^ = 1 — 4 sin^ Ow where Ow is the weak mixing angle. T h e motivation of the E l 5 8 experiment is to measure sin^ Ow at low energy [2], far away from the Zpole (Q^
V27ra
iy l^y^^{ly)
nQ'wiQ^)
(1)
where Gp and a are the Fermi and fine structure constants, Q^ is the square of the fourvector m o m e n t u m transfer and y = Q^ js^ where s is the square of the centerofmass energy. T h e T factor accounts for photon radiation eflFects in initial and final states. Most of the loop and vertex electroweak corrections are absorbed into the definition of an effective weak mixing angle sin^^vF(Q^) which acquires a Q^ dependence. Taking advantage of the large cross section and little theoretical uncertainty of the purely leptonic M0ller reaction, a precise measurement of these electroweak radiative corrections probes physics beyond the Standard Model (SM) at the TeV scale.
T h e figure of merit of the measurement peaks at high energy and to less extent at centerofmass scattering angles around 90°. W i t h a 50 GeV b e a m and Q^ = 0.026 GeV/c^, Apv is predicted to be ^ 3 2 0 parts per billion (ppb) at tree level. Electroweak radiative corrections [4] reduce sin^ Ow by 3%. However this quantity remains numerically very close to 1/4 resulting in an experimental asymmetry reduced by more t h a n 40%. Hence the difficulty of measuring an extremely small asymmetry is compensated by a great sensitivity to sin^ Ow
2 Beam T h e development of the intense and highly polarized 50 GeV b e a m in SLAG E n d Station A (ESA) is a key element in the measurement of ApvLongitudinally polarized electrons are produced by optical pumping of a strained GaAs photocathode [5] by circularly polarized laser light. T h e sign of the laser circular polarization state determines the electron beam helicity. T h e beam is pulsed at 120 Hz with an intensity of 5.10^^ electrons in a 300 ns pulse. T h e time structure consists of quadruplets of two consecutive pulses with pseudorandomly chosen helicities, followed by their complements, yielding two independent leftright pulse pairs every 33 ms. This sequence is phaselocked to the 60 Hz of the power line in order to reduce electronic noise.One of the main experimental challenges when measuring a physics asymmetry at the few 100 p p b level is to keep the beam parameters (intensity, energy, position, angle) with negligible leftright asymmetries. T h e strategy used to reduce these asymmetries is threefold [6]: a passive minimization of helicity correlated intensity and position diflFerences t h a t result from imperfections in the laser light and the
152
A. Vacheret, D. Lhuillier: A precise measurement of sin^Ow at low Q^ in M0ller scattering
photocathode response is first performed by a careful optimization of all the optical elements in the laser p a t h . T h e n helicitydependent corrections are applied to the laser beam in a feedback loop using periodic average measurement of beam asymmetries at upstream and downstream ends of the accelerator. These measurements are performed by cavity monitors implemented by pairs on the beam line [7]. This redundancy is critical to access the intrinsic resolution of the monitors and be able to read the b e a m parameters with a precision far better t h a n their pulse to pulse jitter. Finally a set of slow helicity reversals, performed on a 1 or 2 days basis, further suppress effects of the remaining asymmetries. A halfwave plate can be inserted across the laser b e a m of the source to flip its polarization. Also the main linac can be operated at 45.6 or 48.7 GeV resulting in a 180° extra spin precession in the arc leading to ESA. In b o t h cases the electron beam helicity is flipped, hence the sign of Apv, while certain classes of beam asymmetries remain unaffected yielding to partial cancellations. T h a n k s to these active minimizations, cumulative beam asymmetries are reduced to < 200 p p b in intensity, < 10 p p b in energy, and < 5 n m in position, orders of magnitude below the " n a t u r a l " beam asymmetries one could expect from the accelerator.
3 Instrumentation in ESA T h e a p p a r a t u s in ESA is illustrated in Fig. 1. T h e high luminosity and good background rejection required for the measurement of the tiny M0ller asymmetry are achieved by the use of an extended target, a spectrometer/collimator system and integrating detectors. T h e target cell [8] is a 1.57 m long cylinder filled with liquid hydrogen circulating at :^5 m / s . Turbulences are enhanced by Aluminum meshes in the fluid p a t h allowing the absorption of ~ 5 0 0 W deposited by the beam while keeping density fluctuations below 40 p p m per pulse pair. These fluctuations are monitored by "luminosity monitors" consisting of eight ionization chambers arranged around the beam pipe downstream the detectors and intercepting particles scattered at ~ 1 mr. T h e huge rate provides great sensitivity to target boiling as well as a powerful check of the null asymmetry expected at such forward angles. T h e choice of hydrogen as an electron target results from the best Z / A ratio. T h e low Q^ and high energy kinematics lead to very forward laboratory scattering angles and b o t h the primary b e a m and the scattered particles propagate through the spectrometer. A horizontal 3dipole chicane defines the energy acceptance and shields the detectors from direct lineofsight to the target. At the exit of the chicane the primary beam is put back on its axis and the main acceptance collimator, 3QC1B, selects particles between 4.4 and 7.5 mr. T h e n a series of 4 quadrupoles separates spatially the M0llerscattered electrons from the Mott {ep scattering) background at the detector plane 60 meters downstream the target. Selected M0ller electrons form a ring
target
li
iMoller
Fig. 1. Overview of the experimental setup in ESA. The horizontal axis is shrunk by a factor ^50 in order to enhance the horizontal deviation of the dipole chicane and the focusing of the M0ller events toward the inner ring of detectors
approximately symmetric about the beam axis with an energy range of 1324 GeV and containing ~2.10^ electrons per pulse. For the asymmetry measurement, the particles are intercepted by the M0ller and ep detectors, consisting of 25 cm thick concentric cylinders with a 1525 cm and 2535 cm radius coverage respectively. They are assembled by layering planes of flexible fusedsilica fibers between elliptical copper plates so as to withstand a 100 Mrad radiation dose. T h e fibers are oriented at the Cherenkov angle and gathered in bunches at the back of the detector. They direct the Cherenkov light into shielded P M T s via air lightguides. A total of 60 bunches provides radial and azimuthal segmentation (Fig. 2). T h e asymmetry is simply measured by extracting the fractional difference in the integrated calorimeter response for incident right and lefthanded beam pulses. This technique allows reaching very high counting rate with no dead time. T h e drawback is a "blind" detector integrating all events in the acceptance, including background. Therefore a complete set of auxiliary detectors is implemented to monitor the flux and the asymmetry of the different backgrounds. A profile detector is located just upstream of the main calorimeter providing a complete radial and azimuthal m a p of the charged particles flux. Numerous profile measurements allow a very accurate calibration of the simulations (Fig. 2). T h e dominant background is the ep flux with 8% contamination of the M0ller fiux. T h e small contribution of photons and neutrons to the calorimeter response is measured in calibration runs. T h e pion flux and asymmetry are measured behind the M0ller detector and a heavy lead shielding using a set of 10 Cherenkov detectors. Finally the b e a m polarization is measured once every 2 days using a polarized supermendur foil target just upstream the LH2 target. T h a n k s to a dedicated retractable collimator, doublypolarized M0ller events are isolated inside a small element of the acceptance of the same spec
A. Vacheret, D. Lhuillier: A precise measurement of sin^Ow at low Q^ in M0ller scattering
153
Table 1. Corrections A A and dilutions / to Araw for run III AA (ppb)
Source
25
30 r (cm)
Fig. 2. Radial profile of charged particles at the calorimeter. The points are data, the histograms are simulations of the M0ller (shaded) and ep (hatched) flux. For the detection of M0ller electrons, regions I+II are segmented in three rings of 10, 20 and 20 PMTs respectively
trometer as for the main experiment. They are detected by a small t u n g s t e n / q u a r t z stack fixed at the end of an air lightguide with P M T readout. This setup is mounted on the profile detector frame.
4 Analysis and preliminary results Every 33 ms the d a t a from all beam monitors and detector channels are collected for the 4 beam pulses of a quadruplet. T h e experimental detector asymmetries (Aexp) and beam parameters differences (Axi, i = E^x^y^Ox^Oy^Q) for the corresponding two pairs are extracted. T h e analysis is performed with a blinding offset added to the detector asymmetries. This offset is pickedup by a random algorithm in a range comparable to the expected physics asymmetry and kept secret until the very last stage of the analysis, preventing any psychological influence on the final results. T h e cuts applied to the d a t a (16% of rejected events) select periods of stable beam and operational equipement in an helicityindependent way. In total ^ 4 X 10^ pulse pairs satisfy all criteria for the three runs taken in 20023. Final results have already been published for the first run [9], we present here preliminary results for the complete statistics. T h e detector signals are corrected for differences in the rightleft beam properties as measured by the beam monitors according to /
OLi^AXi
(2)
where the coefficients oti are the sensitivities of the detector to each b e a m parameter. They are determined by two independent methods. T h e first one applies an unbinned least squares linear regression to the pulses used for physics, taking advantage of the beam parameters jitter. T h e second method, socalled "dithering", uses calibration subset of the pulses (4% duty cycle), where each parameter is modulated periodically around its average value
Beam (first order) Beam (higher order) Transverse polarization e~ p ^ e~ p(\^) e~ ( 7 ) P ^ e  p ( + 7 ) High energy photons Synchrotron photons Neutrons Pions
/ 
libl
4 8 22
4 2 2 6 3 2
libl libl
0.058 0.007 0.009 0.003 0.004 0.002 0.0015 zb 0.0005 0.0006 zb 0.0002 0.001 1
by an amount large compared to b e a m jitter. Because of its very forward kinematics E158 is quite sensitive to beam fluctuations therefore the extra lever a r m given by the large amplitudes of the dithering method is not critical to extract accurate a coefficients. Final analysis relies on the first method, statistically more powerful, and uses the dithering as a crosscheck of systematic errors. These corrections remove b o t h the beaminduced random and systematic effects. T h e RMS of the A^aw distribution is reduced to ^ 2 0 0 ppm, very close to the expected pure statistical width and a factor 2 smaller t h a n the RMS of the Aexp distribution. T h e cumulative asymmetry correction is listed in Table 1 for run III. T h e impressively small systematic error of 1 p p b follows from the excellent agreement between the two correction methods when applied on the same d a t a sample. After correcting the effects of all the beam parameters nonstatistical fluctuations of Araw around its mean value are observed in the, most sensitive, outer ring of the M0ller detector, pointing to extra systematic effects at the subpulse scale. Such effects couldn't be treated with the abovementioned methods which only deal with mean values per pulse. To study and correct the induced false asymmetries, the signals of the beam monitors are oversampled with 4 timeslices inside the duration of a beam pulse. T h e final correction removes all the nonstatistical fluctuations with no significant effect on the mean value and a total systematic error of 4 p p b . After a complete regression a remaining azimuthal modulation of Araw is observed due to a small nonzero component of the transverse beam polarization leading to a correction of —4 zb 2 p p b . A separate study limited the bias due to beam spotsize fluctuations on Araw to 1 ppb, using d a t a from a retractable wire array. T h e physics asymmetry is finally formed from Araw by correcting for background contributions, detector linearity € and b e a m polarization Pg1 ^phys
—
Pee
Ar
E.^^. lE^f^
(3)
where AAi and fi are the asymmetry corrections and dilutions for various background sources listed in Table 1.
A. Vacheret, D. Lhuillier: A precise measurement of sin^Ow at low Q^ in M0ller scattering
154
^
^
w3^^
0.2292 ± 0.0019
Qw(Cs)
—•—I
0.2361 ± 0.0017
0.2330 ± 0.0015 60 70 DATA SAMPLE
Fig. 3 . M0ller physics asymmetry. The solid line indicates the expectation for the asymmetry for all combinations of beam energy and halfwave plate state T h e largest correction is due t o electrons from inelastic electron a n d photonproton interactions. T h e associated asymmetry was measured during t h e first r u n in t h e ep detector a n d reasonable assumptions were used for t h e kinematic extrapolation t o t h e M0ller region. T h e flux contamination is determined from a simulation validated with t h e numerous profile detector data. % with dominant T h e averaged beam polarization is 88 errors arising from t h e knowledge of the magnetization of the foil target a n d t h e background. T h e linearity of t h e calorimeter response is determined t o be e = 0.99 =b 0.01. As a final check t h e mean Aphys is plotted for each of the energy a n d halfwave plate configurations. Figure 3 show t h a t d a t a are compatible with a perfect sign reversal, giving confidence in t h e removal of all spurious asymmetries. After removing the blinding offset the grand average of the sign corrected asymmetries gives t h e final physics result:
0.2311 ± 0.0002 ± 0.0006
0.24
0.245
0.25
sin^e^CM^)
Fig. 4. Summary of siii^Ow{MS) measurements evolved to the Zpole. QwiCs): parity violation in Cesium atoms, NuTeV: ^/nuclei scattering, PDG2002: world average, see [14] to measurements in t h e Cesium atoms [11,12] a n d in unucleon scattering [13]. T h e significant deviation of t h e NuTeV result can't be a t t r i b u t e d t o new physics yet as detailed analysis of nuclear effects is still going on. T h e sensitivity of the new E158 result t o new physics is complementary t o t h e existing d a t a a n d t h e measurement of an asymmetry with a 15 p p b error is a benchmark for future high accuracy experiments [15].
References Apv
=  1 2 8 =b 14 (stat.)
12 (syst.) p p b
(4)
1. Y.B. Zel'dovich: Sov. Phys. J E T P 94, 262 (1959) 2. K.S. Kumar, R.S. Holmes, E.W. Hughes, P.A. Souder: Mod. Phys. Lett. A 10, 2979 (1995) 3. LEP and SLD collaborations: hepex/0312023 (2003) 4. A. Czarnecki, W.J. Marciano: Int. J. Mod. Phys. A 15, 2365 (2000); J. Erler, A. Kurylov, M.J. RamseyMusolf: Phys. Rev. D 68, 016006 (2003); F.J. Petriello: Phys. Rev. D 68, 033006 (2003); A. Ferroglia, G. Ossola, A. Sirlin: eprint hepph/0307200 (2003) 5. T. Maruyama et al.: Nucl. Inst. Meth. A 492, 199 (2002) sin^ OwiQ'^ = 0.026 G e V / c ^ ) ^ ^ = (5) 6. T.B. Humenski et al.: Nucl. Inst. Meth. A 521, 261 (2004) 0.2403 zb 0.0010 (stat) 0.0009 (syst) 7. H. Whittum, Y. Kolomensky: Rev. Sci. lustrum. 70, 23002313 (1999) consistent at t h e 1.2 a level with t h e SM expectation [4] 8. J. Gao et al.: Nucl. lustrum. Meth. A 498, 90 (2003) sin^ Ow{0)jis = 0.2385zb0.0006 (theory). This result pro9. P.L. Anthony et al.: Phys. Rev. Lett. 92, 181602 (2004) vides significant new limits on physics beyond the SM. For 10. M.J. RamseyMusolf: Phys. Rev. C 60, 015501 (1999) example t h e scale ALL of a new lefthanded contact inter 11 C.S.Wood et al.: Science 275, 1759 (1997) action [10] is set t o yl+^ > 6.4 TeV a n d yl~^ > 13.9 TeV 12 V.A. Dzuba, V.V. Flambaum, J.S.M. Ginges: Phys. Rev. D 66, 076013 (2002); A. Derevianko, B. Ravaine, W.R. Johnfor potential positive and negative deviations respectively, son: physics/0401043, (2004) and t h e lower limit for t h e mass of a new Z^ boson [4] is 13. K.S. McFarland: these proceedings and references therein set t o Mz^ > 860 TeV at 95% C.L. 14. Review of Particle Properties, K. Hagiwara et al.: For a coherent comparison between all precise measurePhys. Rev. D 66, 010001 (2002) ments of sin^ Ow Fig. 4 shows all results evolved t o t h e 15. K.S. Kumar: these proceedings and references therein Zpole. At low energy t h e E158 accuracy is comparable establishing parity violation in M0ller scattering with t h e most precise measurement of any asymmetry in electron scattering. Using (1) we can extract the weak charge of the electron a n d then sin^OwiQ'^)^ T h e average kinematics as well as photon radiation effect are determined from a Monte Carlo simulation: Q^ = 0.026 (GeV/c)^, y ^ 0.6 and Tb = 1.01 zb 0.01. We find
Eur Phys J A (2005) 24, s2, 155158 DOI: 10.1140/epjad/s2005040395
EPJ A direct electronic only
Qweak: A precision measurement of the proton's weak charge Gregory R. Smith, for the Qweak Cohaboration Jefferson Lab, Newport News, VA 23606, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa ItaUana di Fisica / SpringerVerlag 2005 Abstract. The Qweak experiment at Jefferson Lab will measure the parityviolating asymmetry in ep elastic scattering at very low Q^ using a longitudinally polarized electron beam and a liquid hydrogen target. The experiment will provide the first measure of the weak charge of the proton, Qlw, to an accuracy of 4%. Qtt; is simply related to the weak mixing angle 0^^ providing a precision test of the Standard Model. Since the value of ^\Y? Qyj is approximately 1/4, the weak charge of the proton Q^ = 1 — 4sin^^u; is suppressed in the Standard Model, making it especially sensitive to the value of the mixing angle and also to possible new physics. The experiment employs an 85% polarized, 180 /xA, 1.2 GeV electron beam, a 35 cm liquid hydrogen target; and a toroidal magnet to focus electrons scattered at 8° ib2°, corresponding to Q^ ~ 0.03 (GeV/c)^. With these kinematics the systematic uncertainties from hadronic processes are strongly suppressed. To obtain the necessary statistics this 2200 hours experiment must run at an event rate of over 6 GHz. This requires current (integrating) mode detection of the scattered electrons, which will be achieved using synthetic quartz Cherenkov detectors. A tracking system will be used in a lowrate counting mode to determine the average Q^ and the dilution factor of background events. The theoretical context of the experiment and the status of its design are discussed. PACS. 24.80.+y Nuclear tests of fundamental interactions and symmetries scattering
25.30.Bf Elastic electron
1 Introduction ^ The Qweak Collaboration: D. Armstrong^, T. Averett^, J Birchall^, T. Botto^, J.D. Bowman"^, A. Bruell^, R. Carlini^, S. Chattopadhyay^, C. Davis^, J. Doornbos^, K. Dow^, J. Dunne'^, R. Ent^, J. Erler^ W. Falk^ M. Farkhondeh^ J.M. F i n n \ T. Forest^ W. Franklin^ D. Gaskell^ K. G r i m m \ F.W. Hersman^°, M. Holtrop^°, K. Johnston^, R. Jones^\ K. J o o ^ \ S. Kowalski^, C. Keppel^^ M. Kohl^ E. Korkmaz^^ L. Lee^ Y. Liang^^ A. Lung^, D. Mack^, S. Majewski^, J. Mammei^^, R. Mammei^^, D. Meekins^ G. Mitchell^, H. Mkrtchyan^^ N. Morgan^^ A. Opper^^, S. Page^ S. Pentillo^, M. Pitt^^ M. Poelker^ T. Porcelh^^, W.D. Ramsay^, M. RamseyMusolf^^, J. Roche^, N. Simicevic^ G.R. Smith^ T. Smith^^ R. Suleiman^^ S. Taylor^ E. Tsentalovich^ W.T.H. vanOers^ S. Wells^ W.S. Wilburn^ S. Wood^ H. Zhu^°, and C. Zorn^ ^College of Wilham & Mary ^University of Manitoba ^Massachusetts Institute of Technology ^Los Alamos National Laboratory ^Thomas Jefferson National Accelerator Facility ^TRIUMF Mississippi State University ^Universidad Nacional Autonoma de Mexico ^Louisiana Tech University ^°University of New Hampshire ^^ University of Connecticut ^^Hampton University ^^University of Northern British Columbia
We describe a new experiment at Jefferson Lab to make the world's first measurement of the weak charge of the proton. Qweak is a well defined experimental observable with a definite prediction in the Standard Model (SM). T h e experiment would constitute the first SM test at Jefferson Lab. To lowest order, the weak charge can be expressed as Q ^ = 1 — 4sin^ ^w, where sin^ ^^i; ?^ 0.23 is the weak mixing angle. T h e goal of the Qweak experiment [1] is a 4% measurement of Q ^ (combined statistical and systematic errors), which corresponds to a 0.3% measurement of sin^ dyj. A measurement of this precision is possible because hadronic corrections are small at low Q^ [2], and measured by many different experiments aimed at electromagnetic and weak hadronic form factors.
2 Physics motivation T h e SM makes a firm prediction for Qweak based on the running [3,4] of '^\Y? dy^ from the Z^ pole. As shown in ^^Ohio University ^^Virginia Polytechnic Institute ^^Yerevan Physics Institute ""^California Institute of Technology ""^^Dartmouth College
G.R. Smith: Qweak: A precision measurement of the proton's weak charge
156 0.25
SMy
0.24
h
INUTCV
IAPV
A^*^""**^
1
0.23
J^ Zpole
Qweak
E158 1
1
1
1
1
1
1
0.001
0.01
0.1
1 Q [ GeV]
10
100
1000
Fig. 1. Running of the weak mixing angle [3] in the Standard Model, calculated in the MS scheme. Shown are results from atomic parity violation (APV) [810], NuTeV [7], and the Z pole [5]. The error bar shown for the E158 [6] M0ller experiment is that anticipated for their final result Fig. 1 t h e value of t h e weak mixing angle is predicted t o change (in t h e MS renormalization scheme) by ^ 3 % from t h e energy scale of t h e Z pole (where a decade of precision measurements have been made at high energy colliders like SLAC and L E P [5]) to t h e low energy scale of Qweak W i t h t h e proposed 0.3% measurement of t h e weak mixing angle, t h e Qweak experiment will make a ~ lOcr verification of this effect. A deviation from t h e SM prediction would imply new physics beyond t h e SM. In t h e case of t h e Qweak experiment, there is even sensitivity t o which SM extension is responsible for t h e deviation, especially when taken in conjunction with t h e results of other experiments. Should t h e result of this experiment agree with t h e prediction of the SM, t h e n t h e result would dramatically constrain possible extensions to t h e SM. T h e Qweak experiment [1] at Jefferson Lab (JLab) will make such a precision measurement of t h e asymmetry between crosssections for positive and negative helicity electrons in polarized elastic electronproton scattering. T h e asymmetry violates parity, and arises from t h e interference of electromagnetic and weak amplitudes (photon and Z^ boson exchange). At this low energy scale, t h e asymmetry is a measure of t h e weak charge of t h e proton, Q ^ , which is t h e strength of t h e weak vector coupling of t h e Z^ boson t o t h e proton. At tree level t h e value of Qweak is expected t o be ~ 0.072. In t h e limit of small scattering angle and small m o m e n t u m transfer (Q^ ^ 0), the asymmetry is given by [2]: a^ — G
GF
47ra\/2
][Q^Ql^Q^B{Q^)]
 0 . 3 p p m at
Q2
(1)
= 0.03 GeV^
where B{Q^) is a contribution from electromagnetic and weak form factors. It is sensitive to G ^ ^ and G^ ^ , t h e target of many recent and ongoing experiments at Jefferson Lab, Bates, and MAMI. As 1 makes clear, t h e B t e r m is suppressed as Q^ goes to zero by an extra factor of Q^ relative to Qweak On t h e other hand, t h e experimental asymmetry grows with Q^. At Jefferson Lab a Q^
of ^ 0.03 GeV^ has been chosen as t h e most ideal compromise between these two competing considerations. T h e expected measured asymmetry A^eas of —0.29 p p m consists of three main terms. T h e biggest is t h e t e r m sought in this experiment, namely A.QW^ t h e first t e r m in 1. It has an expected value of —0.19 ppm, or about 2 / 3 of t h e expected Ameas T h e second t e r m comes from t h e hadronic form factors discussed above, and has a magnitude expected to be about 3 1 % of Ameas, or about —0.09 ppm. It will be well constrained by t h e precise measurements provided by H A P P E X , PVA4, and GO. T h e third t e r m is almost negligible, and amounts to only about 3 % of t h e expected Ameas It accounts for t h e axial contribution, contains G ^ , and includes substantial electroweak radiative corrections. It will be constrained by S A M P L E and GO. W h e n added in quadrature, t h e two background terms Ahadronic and Aaxiai contribute only about 2% t o t h e expected final error on Qweak We reemphasize t h a t t h e bulk of this socalled background contribution can be accounted for solely with the results of precision experiments aimed at hadronic form factors, and does not rely on theoretical calculations or models. In addition, t h e Qweak collaboration is considering augmenting t h e existing experimental information on B{Q'^) by performing a short independent measurement at a Q^ slightly higher t h a n 0.03 GeV^. This should not be necessary if t h e ongoing form factor experiments achieve their proposed goals. T h e Qweak experiment will provide tight constraints on SM parameters associated with potential new physics. For example, t h e up and down quark couplings in t h e SM electronquark Lagrangian: GF.
^SM =  ^ e 7 ^ 7 5 e ^ C i g ^ 7 ^ g
T2'
are constrained by Q^
according to
Q ^ ( ^ M ) =  2 ( 2 C i , + Cid)  0.0721 . This constraint is complementary t o other existing constraints, and when combined with those other measurements, dramatically reduces t h e allowed phase space for new physics. T h e physics reach of this experiment can be expressed in terms of its precision as: A ^
1
4.6 TeV.
JV2GF\AQP I w\
T h e discovery potential of weak charge measurements will be unmatched until t h e LHC t u r n s on. If new physics is uncovered at t h e LHC, such as an extra Z boson, then precision experiments like Qweak undertaken at low Q^ will be essential to determine t h e charges, coupling constants, etc. There are many reasons to expect t h a t t h e SM is a lowenergy effective theory of some more fundamental description of nature. Neutral current experiments like Qweak provide important consistency checks of t h e SM, are complementary to direct searches for new physics, and can
G.R. Smith: Qweak: A precision measurement of the proton's weak charge help distinguish and characterize the new physics once it is found [3,4]. One of the most plausible extensions to the SM which can be tested for by the Qweak experiment is the presence of extra Z's. T h e present lower bound for Z's provided by Tevatron experiments is only 0.6 TeV, which leaves plenty of discovery potential for the Qweak experiment. TJ'S would resolve the ^^ 2(j discrepancy from the SM in the generation number (N^ = 2.986=b0.008) derived from the measured Z^ lineshape. Furthermore, Z's would improve the agreement of b o t h the A P V and NuTeV experiments with the SM. Another SM extension the Qweak experiment is particularly sensitive to is Rparity violation, one of the most interesting varieties of SUSY. Finally, the Qweak experiment will provide a sensitive test for leptoquarks. Weak charge and mixing angle results can be obtained in other types of experiments. T h e SLAG E l 5 8 [6] experiment is a purely leptonic M0ller scattering experiment recently completed at a similar Q^. Their preliminary results are consistent with the SM. T h e Fermilab NuTeV experiment [7] explored z/A scattering at ~ 10 GeV/c^, and found an intriguing Scr discrepancy with the SM. T h e interpretation of this discrepancy remains somewhat controversial for now. Finally, there have been atomic parity violation (APV) measurements [810] made of the Gs a t o m ( Q ^ = 0 ) . Those results have been plagued by extraordinarily difficult and changing theoretical corrections (see, for example [11,12]), however, at present their result is consistent with the SM. To provide some perspective, a 13% measurement of Q^eak ^^ ^^^ ^~P experiment would provide the same sensitivity to new physics as a 1% measure of Qweak(N, Z) usiug A P V , but without the theoretical uncertainties. On the other hand, our ability to test for extensions to the SM will be greatest when the results of the Qweak experiment are combined with those from the complementary triad of experiments referred to above, especially E158.
3 The qweak experiment T h e conceptual design of the Qweak experiment is illust r a t e d in Fig. 2. T h e experiment consists of scattering a longitudinally polarized 1.2 GeV electron beam by a 35 cm liquid hydrogen target. Elastically scattered electrons at 8° =b 2° are selected by a collimation system, and then focused by a large toroidal resistive magnet onto a set of eight synthetic quartz Cerenkov detectors. At the average experimental m o m e n t u m transfer of Q^ = 0.03 GeV^ the expected Qweak asymmetry is small, 0.3 parts per million (ppm). T h e expected event rate for scattered electrons of ^ 6 GHz precludes counting the individual events. Instead, the experiment will use current mode detection and low noise front end electronics. In brief, the various experimental systems are as follows. T h e liquid hydrogen target will be patterned after the successful design of the GO and S A M P L E targets except t h a t a centrifugal p u m p will deliver several times the flow velocity achieved in those targets. To meet the more challenging requirements of the Qweak experiment.
157
the b e a m will be rastered into a uniform 4mm x 4mm p a t t e r n . T h e target cell will be 35cm in length. W i t h the proposed b e a m current of 180/iA, this target will require almost 2.5kW of cooling power. T h e ability of the target to provide the necessary small boiling contribution to the asymmetry width will be backed u p by an array of high rate, small angle quartz luminosity monitors near the end of the Jefferson Lab Hall G b e a m line. T h e toroidal magnet consists of eight resistive coils each approximately 3.7m long and 1.5m tall composed of a double pancake of Gu conductor 3.8cm x 5.8cm in cross section, with a 2cm diameter central cooling channel. T h e magnet will provide an J B dl ~ 0.7 Tm with a 9500 A, 1.2 M W DG power supply. T h e coils will be supported in an aluminum stand, and will permit a 1 foot diameter, Pbshielded b e a m pipe to pass through the central axis. Simulations show t h a t this magnet will cleanly separate photons and inelastic events from the elastic events of interest at the focal plane. T h e detectors themselves will be relatively insensitive to 7's, n's, and TT'S, be capable of withstanding > 300 kRad, and operate at counting statistics. Eight fused silica Cerenkov detectors will be used, each approximately 2m X 16cm x 2.54cm. These bars are characterized by n=1.47, ^Cerenkov = 47°, and a total internal reflection angle of 43°. T h e bars will be read out by S20 photocathodes at each end, and should provide about 100 photoelectrons per event. A tracking system will be used with the b e a m current reduced by four orders of magnitude, allowing individual events to be observed. This will enable b o t h a measurement of the dilution of the Cerenkov detector signal by background and a precise determination of the average Q^. T h e tracking system components will rotate to cover all octants, and will include the following sets of detectors. A gas electron multiplier closest to the target will serve as a vertex detector. Horizontal drift chambers near the magnet entrance will measure the scattering angle. These will be augmented by a minitorus to sweep away the otherwise dominant M0ller electrons. At the focal plane, vertical drift chambers will m a p the analog response of the Cerenkov system. Finally, large scintillation counters will provide a charged particle trigger. Precision beam polarimetry is required in order to have a polarization contribution to the systematic error of less t h a n 1.5%. To achieve this a twopronged approach is being pursued. First, efforts are already underway to increase the operating current of the existing < 10/iA, < 1% M0ller polarimeter to 100 //A b e a m currents. Methods being investigated include kicking the beam across wire M0ller targets, and use of rotating foil targets. In addition, a Compton polarimeter is being developed for J L a b Hall C which would provide continuous relative measurement of the b e a m polarization, normalized by the M0ller polarimeter. Other requirements on b e a m properties have been determined by extensive simulations. T h e Qweak b e a m property specifications have for the most part already been achieved at Jefferson Lab during the GO and H A P P E X
158
G.R. Smith: Qweak: A precision measurement of the proton's weak charge Election Beam
L2 GeV 180 ^A, P=80% 30 Hz leveisal
Middle Tiacking Cham be is Re a I T L ac ki ng C h am be is & Scintillatois Current Mode Sj'iithetic Quart £ Cerenkov Detectore  6 G H i total rate
35 cm LH^ Taiget & Sc arte ling Chambei
Fiist Collimatoi (Tungsten) & G E M detectois
Second Collimatoi (Conciete)
Magnet
Detectoi Shielding
Fig. 2. Conceptual design for the Qweak experimental setup, in Hall C at Jefferson Lab. The eight quartz detectors, each 2 m x l 6 cmx2.5 cm, are shown inside their reentrant shielding enclosure. The spectrometer provides clean separation of elastic and inelastic electrons at its focal plane
experiments. T h e Qweak requirements are very similar to those for H A P P E X  P b , which runs with s t a n d a r d equipment and therefore should be scheduled well before QweakT h a t experiment, however, does not enjoy the common mode rejection inherent in a toroidal spectrometer like Qweak In addition, intensity and position feedback will be separated in the Qweak experiment by making use of position information just upstream of the target. In spite of the fact t h a t the b e a m properties at Jefferson Lab seem to already be very close to what is required for the Qweak experiment, the collaboration is working with the J L a b source group to explore the possibility of implementing helicity reversal faster t h a n the canonical 30 Hz, which shall considerably ease the demands this experiment places on many b e a m properties as well as target density fluctuations, among others. T h e Qweak physics proposal was approved in January, 2002 by the J L a b PAC, and has since become an import a n t new thrust of the J L a b scientific program. T h e collaboration presented a successful technical design review in J a n u a r y 2003. A management plan is in place for the project, and all the required funding has been secured. T h e project is being supported by the U.S. Department of Energy, the National Science Foundation, and the Natural Sciences and Engineering Research Council of Canada. T h e Qweak experiment will proceed in two stages. First, a statistics limited run with a low power target will aim for a < 8 % result on the asymmetry in 2007. This will be followed by runs totalling about 2200 hours at 180 fiA in
order to achieve a 4% result. T h e collaboration is investigating whether the absolute limits of the technique can be pushed any further, recognizing t h a t the physics impact of an even more precise experiment would be enormous.
References 1. Qweak Collaboration, R. Carlini et al.: Jefferson Lab Proposal E02020. Proposal and collaboration member list available at http://www.jlab.org/qweak/. 2. M.J. Musolf et a l : Phys. Rep. 239, 1 (1994) 3. J. Erler, A. Kurylov, M.J. RamseyMusolf: Phys. Rev D 68, 016006 (2003) 4. A. Kurylov, M.J. RamseyMusolf, S. Su: hepph/0303026, to appear in Phys. Rev. D 5. Particle Data Group: K. Hagiwara et al.: Phys. Rev. D 66, 010001 (2002); In particular, the section Electroweak Model and Constraints on New Physics 6. P L . Anthony et al.: Phys. Rev. Lett. 92, 181602 (2004) 7. NuTeV CoUaboration, G.P. Zeller et al.: Phys. Rev. Lett. 88, 091802 (2002) 8. N.H Edwards et al.: Phys. Rev. Lett. 74, 2654 (1995) 9. P A . Vetter et al.: Phys. Rev. Lett. 74, 2658 (1995) 10. C.S. Wood et al.: Science 275, 1759 (1997) 11. M.Y. Kuchiev, V.V. Flambaum: Phys. Rev. Lett. 89, 283002 (2002) 12. A.L Milstein, G.P Sushkov, LS. Terekhov: Phys. Rev. Lett. 89, 283003 (2002)
Eur Phys J A (2005) 24, s2, 159160 DOI: 10.1140/epjad/s2005040400
EPJ A direct electronic only
The Qweak tracking system Measurement of the average Q^ and the dilution by background events Klaus H. Grimm, for the Qweak Collaboration The College of Wilham & Mary, Department of Physics, Wilhamsburg, VA 231878795, USA Received: 01 November 2004 / Pubhshed Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. The Q^^^k experiment will measure the parityviolating elastic ep scattering asymmetry to extract the weak charge of the proton. The experiment employs a toroidal magnet to focus electrons scattered at 8° , corresponding to Q^ ~ 0.03 (GeV/c)^, on eight Cerenkov detectors located in the focal plane of the spectrometer. Since the asymmetry is proportional to Q^, it is crucial to obtain an accurate measure of the acceptanceaveraged value of Q^. A tracking system will be used in a lowrate counting mode, allowing individual events to be observed. This will enable a determination of the average Q^ by measuring the scattering angle and interaction vertex, for mapping the response across the surface of the Cerenkov detectors, and for the dilution of the Cerenkov detector signal by background. PACS. 25.30.Bf Elastic electron scattering  29.40.Gx Tracking and positionsensitive detectors
1 Introduction ^ The Qweak Collaboration: D. Armstrong^ T. Averett\ J. Birchall^, T. Botto^ J.D. Bowman^, A. Bruell^, R. Carlini^, S. Chattopadhyay^, C. Davis^, J. Doornbos^, K. Dow^, J. Dunne'^, R. Ent^, J. Erler^ W. Falk^ M. Farkhondeh^ J.M. F i n n \ T. Forest^ W. Franklin^ D. Gaskell^ K. G r i m m \ F.W. Hersman^°, M. Holtrop^°, K. Johnston^, R. Jones^\ K. J o o ^ \ S. Kowalski^, C. Keppel^^ M. Kohl^ E. Korkmaz^^ L. Lee^ Y. Liang^^ A. Lung^, D. Mack^, S. Majewski^, J. Mammei^^, D. Meekins^, G. Mitchell^, H. Mkrtchyan^^ N. Morgan^^ A. Opper^^, S. Page^ S. Pentillo^, M. Pitt^^ M. Poelker^ T. Porce^i^^ W.D. Ramsay^, M. RamseyMusolf^^, N. Simicevic^, G.R. Smith^, T. Smith^^, R. Suleiman^^ S. Taylor^, E. Tsentalovich^, W.T.H. vanOers^ S. Wells^ W.S. Wilburn^, S. Wood^ H. Zhu^°, and C. Zorn^ ^College of Wilham & Mary ^University of Manitoba ^Massachusetts Institute of Technology ^Los Alamos National Laboratory ^Thomas Jefferson National Accelerator Facility ^TRIUMF ^Mississippi State University ^Universidad Nacional Autonoma de Mexico ^Louisiana Tech University ^°University of New Hampshire ^"^ University of Connecticut ^^Hampton University ^^University of Northern British Columbia ^^Ohio University ^^Virginia Polytechnic Institute ^^Yerevan Physics Institute ^California Institute of Technology ^^Dartmouth College
In elastic electronproton scattering the parityviolating asymmetry is proportional to Q^. In the Q^^^j^ experiment [1,2] a collimator system, optimized by G E A N T simulations, defines the average value of the accepted Q^ by setting kinematical cuts on the electron scattering angles. For Q^g^y^ it is critical to determine the acceptance averaged value of Q^ for the electrons from the ep elastic scattering events of interest with o^ 1% accuracy. This translates into a O.Gmrad precision in the measure of the scattering angle. Since Q^g^^j^ is an integrating experiment, the Q^ dependence of the ep crosssection will bias the average detected Q^. This and any positiondependent detector bias must be taken into account since Q'^ may be correlated with position at the focal plane. Although in principle these m a t t e r s can be simulated, it is essential to check the Monte Carlo using an ancillary calibration measurement: to measure the scattering angle, interaction vertex, shape of the focal plane distributions, and the positiondependent detector bias. This information will be extracted from ancillary measurements at low b e a m current in which the Cerenkov detectors will be read out in pulse mode and individual particles tracked through the toroidal spectrometer with a tracking system.
2 Tracking system design concept T h e tracking system consists of a set of drift chambers at three locations along the track. It will be capable of mapping two opposing octants simultaneously where
K.H. Grimm: The Qweak tracking system
160
Torofdal AAarn Magnet
collfmator 2 collrmaTor 1
Cerenkov detector electron beam shfefdihg wall liquid hydrogen target $EM
Horizontal Drift chambers
Vertical Drift chambers
Fig. 1. Conceptual design of the Q^^^k tracking system. The toroidal magnetic field spatially separates inelastic events from elastic events which will be focus on a fused silica (quartz) Cerenkov detector per octant located in the focal plane of the spectrometer. The tracking for single particle will be performed at low beam current using a set of drift chambers at three locations in two opposite octants at a time the tracking chambers will be mounted on a rotating wheel assembly. T h e Cerenkov detectors are operated in a low beam current counting mode (1^10 nA,) equivalent to an ep elastic rate of ~ 5 0 k H z / o c t a n t ) for background/acceptance studies using the tracking system, and in a classic integrating mode (1^180 ytxA, equivalent to an ep elastic rate of ^ 8 0 0 M H z / o c t a n t ) during high beam current production running where the tracking system will be retracted. A Triple G E M (Gas Electron Multiplication) chamber is situated between the primary and secondary collimator. A G E M was chosen because of its high rate capability, fast time response and good position resolution, but it does not provide any track angle information. Each chamber has an active area of 1 6 c m x l 6 c m , consists of 3200 channels supporting a rate of 10 MHz and an expected position resolution of c:^200 /xm. A pair of horizontal drift chambers (HDC) separated by 50 cm is situated between the secondary collimator and the main magnet. Each chamber has an active area of 53 cm X 49 cm, and consists of 374 signal wires. T h e expected position resolution will be ~200/xm. T h e HDCs in combination with the G E M will measure the scattering angle, interaction vertex, and establish the trajectory upstream of the main magnet. T h e expected vertex resolution along the target will be ~ 1 m m ; the reconstruction of the scattering angle will be archived with an angular resolution of c^O.Gmrad. T h e HDCs will see b o t h elastic and inelastic events and are in direct line of sight to the target. A low field toroidal sweeping magnet will be installed between the GEMs and the HDCs to prevent M0ller events (3070 MeV,10 MHz) from reaching the HDCs. T h e sweeping magnet will remain on during production run to maintain the same conditions between production running and Q^ calibration runs. A pair of vertical drift chambers (VDC) separated by 30 cm will be situated in front of the Cerenkov detectors,
Fig. 2. GEANT simulation of the spatial separation of the elastic and inelastic events in the focal plane. The /lshaped Cerenkov detector made from joint quartz bars will cover/detect the elastic events with a ~0.02% inelastic rate contribution.
after the shielding wall. Each chamber has an active area of 200 cm X 50 cm, and consists of 800 signal wires. T h e expected position resolution will be ^100/xm. T h e VDCs will m a p out the positiondependent Cerenkov detector response across the surface of each quartz bar. W i t h a known field m a p of the main and sweep magnets and with the local track information of GEMs, HDCs, and VDCs it is possible to reconstruct the particle momentum. This allows the separation between elastic and inelastic track events needed for the mapping and background measurements. T h e global tracking in a toroidal magnetic fields is not a trivial issue since with a nonhom*ogeneous field there is no simple analytic expression to describe the track. T h e Qlueak ^ ^ i i i magnet has a moderate focusing for electrons regarding the polar scattering angle but it has a strong defocusing in the azimuthal angle, so each track will have an unique trajectory parametrization. Therefore the final track fitting method will depend upon a iterative simulation and comparison method similar to existing algorithms used by B L A S T and CLAS. In addition to the tracking technique for the background determination long calibration timeofflight spectra out to 500 ns will be taken in order to verify t h a t there are no longlived backgrounds or delayed lights coming out of the quartz detectors t h a t would require a correction to the central value of the measured asymmetry.
References 1. Qweak Collaboration, R. Carlini et al.: Jefferson Lab Proposal E02020, available at: http://www.jlab.org/qweak/
2. See G.R. Smith, in these proceedings, Qweak: A Precision Measurement of the Proton's Weak Charge
Eur Phys J A (2005) 24, s2, 161164 DOI: 10.1140/epjad/s200504041y
EPJ A direct electronic only
Neutral currents and strangeness of the nucleon from the NuTeV experiment Kevin S. McFarland^ University of Rochester, Rochester, NY 14610, USA Received: 15 December 2004 / PubHshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. The NuTeV neutrino experiment ran in 19971998 at Fermilab and accumulated the world's highest statistics samples of high energy (20300 GeV) separated neutrino and antineutrino interactions. The NuTeV collaboration has used this data to extract the electroweak parameter, sin.^ Ow, from the measurement of the ratios of neutral current to charged current neutrino and antineutrino deep inelastic scattering cross sections. This result, though consistent with previous neutrino electroweak measurements, is not consistent with predictions. One interpretation involves the possibility that the strange quark sea carries significantly more momentum in the nucleon than the antistrange sea. We report on the direct study of this possibility from measurements of chargedcurrent interactions on strange quarks in our neutrino and antineutrino beams. PACS. 12.15.Mn Neutral Currents  13.60.Hb Crosssections in inelastic leptonhadron scattering
1 Introduction and motivation Neutrino scattering played a key role in establishing t h e structure of t h e Standard Model of electroweak unification, and it continues to be one of t h e most precise probes of t h e weak neutral current available experimentally today. W i t h t h e availability of copious d a t a from t h e production and decay of onshell Z and W bosons for comparison, contemporary neutrino scattering measurements serve to validate t h e theory over many orders of magnitude in m o m e n t u m transfer and provide one of t h e most precise tests of t h e weak couplings of neutrinos. In addition, precise measurements of weak interactions far from t h e boson poles are inherently sensitive to processes beyond our current knowledge, including possible contributions from leptoquark and Z' exchange [1] and new properties of neutrinos themselves [2]. T h e ratio of neutral current t o charged current crosssections for either ly or V scattering from isoscalar targets of u and d quarks can be written as [3]
jl^H
a[ jy N ^
v
X)
(gi^r^'^gf^).
(1)
where
a{VN ^ i^X) a{iyN ^ iX)
1 ^ 2'
(2)
and ^£ ^ = (^L i?)^ + (^i i?)^7 isoscalar combinations of quark chiral couplings to t h e Z. Many corrections to 1 ^ presenting on behalf of the NuTeV Collaboration
are required in a real target [4], b u t those most uncertain result from t h e suppression of t h e production of charm quarks in t h e target, which is t h e CKMfavored final state for chargedcurrent scattering from t h e strange sea. This uncertainty has limited t h e precision of previous measurements of electroweak parameters in neutrinonucleon scattering [5,6,7]. One way to reduce t h e uncertainty on electroweak parameters is to measure t h e observable R
1r
(gl
(3)
first suggested by Paschos and Wolfenstein [8] and valid under t h e assumption of equal m o m e n t u m carried by t h e u and d valence quarks in t h e target. Since a^^ = a^^ and a^^ = a^^, t h e effect of scattering from sea quarks, which are symmetric under t h e exchange q ^q^ cancels in the difference of neutrino and antineutrino crosssections. Therefore, t h e suppressed scattering from t h e strange sea does not cause large uncertainties in R~. R~ is more difficult to measure t h a n R^^ primarily because t h e neutral current scatterings of u and V yield identical observed final states which can only be distinguished through a priori knowledge of t h e initial state neutrino. T h e experimental details and theoretical t r e a t m e n t of crosssections in t h e NuTeV electroweak measurement are described in detail elsewhere [4]. In brief, we measure t h e experimental ratio of neutral current to charged current candidates in b o t h a neutrino and antineutrino beam. A Monte Carlo simulation is used to express these exper
K.S. McFarland: Neutral currents and strangeness of the nucleon from the NuTeV experiment
162
imental ratios in terms of fundamental electroweak parameters. This procedure implicitly corrects for details of the neutrino crosssections and experimental backgrounds. For the measurement of sin^ Ow, the sensitivity arises in the V beam, and the measurement in the V b e a m is the control sample for systematic uncertainties, as suggested in the PaschosWolfenstein R~ of 3.
1 0.75

.
0.5 0.25
F[sIn^0vv,ljp(x)c "n(x);X]
:
1.1 QCD corrections Equations 1 and 3 assume targets symmetric under the exchange of u and d quarks, and t h a t quark seas consist of quarks and antiquarks with identical m o m e n t u m distributions. T h e NuTeV analysis corrects for the significant asymmetry of d and u quarks t h a t arises because the NuTeV target, which is primarily composed of iron, has a 5.74 zb 0.02% fractional excess of neutrons over protons. However, this correction is made under the assumption of isospin symmetry. () () () d n{x), d p{x) = u n{x)
u p{x)
This assumption, if significantly incorrect, could produce a sizable effect in the NuTeV extraction of sin^ Ow [9,10,11, 12]. Similarly, the cancellation of charm production from the strange quarks (3) assumes t h a t the m o m e n t u m distributions of the strange and antistrange seas are identical, i.e., s{x) = s(x). NuTeV's analysis is done to leading order in p Q C D , but perhaps surprisingly, the NLO corrections to R~ are very small [13,14,15]. Dropping the assumptions of symmetric heavy quark seas and isospin symmetry, but assuming small deviations in all cases, the effect of these deviations on R~ is [18]: 6R
{Up UpDn^
Dn)  {Dp DpUn^ 2{Up Up^Dp
br)
UpUp^
Dp
D,
Dp)
f
^ + Z^^K),(4)
F[sm^ Ow, ^; x] S{x) dx,
1
=0.75  \ 1 = 1.25
.s^^''"'
 \
FEsin' 0 w , s ( x )  ~s{x)x\
, , , , i T ^ ,
0.1
, , , 1 ,1
0.2
1 ,
1 1 , 1
0,1> 0.4
, 1 , , , , 1 , , , , 1 , , , , 1 , , , , 1 , , , ,
0.5
0.6
0.7
0.8
0.9
1
Fig. 1. The functionals describing the shift in the NuTeV sin^ Ow caused by not correcting the NuTeV analysis for isospin violating u and d valence and sea distributions or for (s(x)) / (six)). The shift in sin^ Ow is determined by convolving the asymmetric momentum distribution with the plotted functional it can be seen t h a t the level of isospin violation required to shift the sin^ Ow measured by NuTeV to its s t a n d a r d model expectation would be, e.g.. Dp — Un ^ 0.01 (about 5% of Dp \ Un), and t h a t the level of asymmetry in the strange sea required would be 5 — 6' ^ +0.007 (about 30% of 5 + 6 ) .
2 Electroweak results
where Al^^ = {e'^^Y  {e'jfY, QN is the total m o m e n t u m carried by quark type Q in nucleon N^ and ec denotes the ratio of the scattering cross section from the strange sea including kinematic suppression of heavy charm production to t h a t without kinematic suppression. NuTeV does not exactly measure R~, in part because it is not possible experimentally to measure neutral current reactions down to zero recoil energy. To parameterize the exact effect of the symmetry violations above, we define the functional F[sin^ Ow, ^] x] such t h a t Z\sin Ow
=0.5
^^<^^^
 1 F[sin^0w,d; ( x )  i 'n(x).X]
Un)
xi^^l^^l) \
0.25
(5)
Jo
for any symmetry violation 5{x) in P D F s . All of the details of the NuTeV analysis are included in the numerical evaluation of the functionals shown in Fig. 1. For this analysis.
As a test of the electroweak predictions for neutrino nucleon scattering, NuTeV performs a singleparameter fit to sin^ Ow with all other parameters assumed to have their s t a n d a r d values, e.g., s t a n d a r d electroweak radiative corrections with po = 1. This fit determines sm
2 ^(on—shell)
0.22773
0.00135(stat.)
0.00093(syst.)
(175 GeV)2 M? 0.00022 X ( ^ ^ ^ ) (50 GeV)2 + 0.00032 X ln(
MHig
150 GeV
).
(6)
A fit to the precision electroweak data, excluding neutrino measurements, predicts a value of 0.2227 0.00037 [16, 17], approximately 3a from the NuTeV measurement. In where Mw the onshell scheme, sin^ Ow = 1 — M^/M^, and Mz are the physical gauge boson masses; therefore, this result implies Mw = 80.14 + 0.08 GeV Although this deviation is statistically significant, it is not immediately apparent what the cause of this discrepancy might be.
K.S. McFarland: Neutral currents and strangeness of the nucleon from the NuTeV experiment We discuss, in turn, possibilities of new physics, nuclear effects, large isospin violation and an asymmetric strange
2.1 New physics
The primary motivation for embarking on the NuTeV measurement was the possibility of observing hints of new physics in a precise measurement of neutrinonucleon scattering. NuTeV is well suited as a probe of nonstandard physics for two reasons: first, the precision of the measurement is a significant improvement, most noticeably in systematic uncertainties, over previous measurements [5,6,7], and second NuTeV's measurement has unique sensitivity to new processes when compared to other precision data. In particular, NuTeV probes weak processes far offshell, and thus is sensitive to other tree level processes involving exchanges of heavy particles. Also, the initial state particle is a neutrino, and neutrino couplings are the most poorly constrained by the Z^ pole data, since they are primarily accessed via the measurement of the Z invisible width. An often useful lowenergy parameterization of new physics is to consider a unitcoupling "contact interaction" in analogy with the Fermi effectively theory of lowenergy weak interactions. Assuming a contact interaction described by a Lagrangian of the form =b47r
{hj^hqU^l^qH,
+ hl^hW^l^QH,
+C.C.) (7)
the NuTeV result can be explained by an interaction with mass scale ylj^ ^ 4 zb 0.8 TeV. However, post hoc it appears that well motivated and complete models for such an interaction seem to be difficult to find [12]. An extra U{1) gauge group giving rise to a heavy Z' boson is a possibility [12,19], but the U{1) gauge group suggested by NuTeV would not necessarily be one motivated by models of unification of known forces [1]. There are few other precision measurements of neutrino neutral current interactions. Measurements of neutrinoelectron scattering from the CHARM II experiment [20] and the direct measurement of r{Z ^ vV) from the observation of Z ^ vV^ at the Z^ pole [16] provide measurements of a few percent precision. The two most precise measurements come from the inferred Z invisible width [16] and the NuTeV result. Both of the precise rate measurements are significantly below the expectation. Again, models how such a deviation of neutrino couplings might fit against other constraints from the data are difficult to find, and the most complete attempts come from models which mix heavy and light neutrinos to form the eigenstates of the neutral weak interaction [21]. 2.1.1 Nuclear Effects
It has been suggested by several authors, correctly, that differences in nuclear effects between charged and neutral current neutrino scattering could affect the NuTeV
163
sin Ow analysis. There are constraints on process dependent nuclear effects, notably the agreement between F2 from chargedlepton and neutrino chargedcurrent scattering [22]. However, effects such Vector Meson Dominance in shadowing [23] or models of antishadowing [24] could still affect the NuTeV result. The former model predicts large increases in R^ and R^ as measured by NuTeV of 0.6% and 1.2%, respectively. This effect, however, largely cancels in R~, and furthermore the NuTeV sin^ Ow data itself disfavors this model through its separate measurements of R^ and R^^ which are both below predictions, while this model increases those very predictions. The latter model of antishadowing would primarily effect R^ and increase it modestly. This effect would be consistent with, though not favored by, the NuTeV data, and although it could reduce slightly the measured sin^ Ow from R~, it would not improve the overall agreement of the NuTeV R^ and R^ with the data.
2.2 Isospin violation
Several recent classes of models predict isospin violation in the nucleon [9,10,11]. The earliest estimation in the literature, a bag model calculation [9], predicts large valence asymmetries of opposite sign in Up — dn and dp — Un at all X, which would produce a shift in the NuTeV sin^ Ow of —0.0020. A more complete calculation done by Thomas et al. [10] concludes that asymmetries at very high x are larger, but the asymmetries at moderate x are smaller and of opposite sign at low x. This calculation is sensitive to the amount of smearing allowed in the energy of the remaining diquark at the bag scale after scattering, agreeing qualitatively with the Sather result with no smearing, but reducing the effect to a negligible —0.0001 when assuming smearing of order TIQCD The effect is also evaluated in the Meson Cloud model [11], and there the asymmetries are much smaller at all x^ resulting in a modest shift in the NuTeV sin^ Ow of+0.0002. Finally, Thorne et al [25] have proposed that isospin violation may arise from QED corrections to PDFs, and have estimated the possible size of the effect in R~ to be ~ —0.002. However, this calculation is very sensitive to assumptions about the quark mass, and the value above assumes quark masses of a few MeV are appropriate. The assumption of constituent quark masses would drastically reduce the size but would retain the sign of the effect. Models for isospin violation aside, the NuTeV data itself cannot provide a significant independent constraint on this form of isospin violation. A recent global analysis has also attempted to constrain this possibility, but found no sufficiently significant constraint. It allows isospin violation large enough to move NuTeV into agreement with prediction or large enough to double the discrepancy [26]. We conclude that NuTeV may have indeed found strong evidence for large (compared to even generous predictions) isospin violation in PDFs in a direction favored by most models, but that there is no independent evidence to support this hypothesis.
164
K.S. McFarland: Neutral currents and strangeness of the nucleon from the NuTeV experiment a sample positive asymmetry (black)
(0 X
NuTeV experiment benefited greatly from significant contributions from the Fermilab Particle Physics, Computing, Technical and Beams Divisions.
References 1. P. Langacker et al.: Rev. Mod. Phys. 64, 87 (1991) 2. K.S. McFarland, D. Naples et al.: Phys. Rev. Lett. 75, 3993 (1995) 3. C.H. Llewellyn Smith: Nucl. Phys. B 228, 205 (1983) 4. G.P. Zeller et al.: Phys. Rev. Lett. 88, 091802 (2002) 5. K.S. McFarland et al. Eur. Phys. Jour. C I , 509 (1998) Red = s*s" param 6. A. Blondel et al.: Zeit. Phys. C 45, 361 (1990) 7. J. Allaby et al. Zeit. Phys. C 36, 611 (1985) Blue = Ka param 8. E.A. Paschos, L. Wolfenstein: Phys. Rev. D 7, 91 (1973) Black = CTEQlike xs' 9. E. Sather: Phys. Lett. B 274, 433 (1992) 10. E.N. Rodionov, A.W. Thomas, J.T. Londergan: Mod. Phys. Lett. A 9, 1799 (1994) 11. F. Cao, A.I. Signal: Phys. Rev. C 62, 015203 (2000) 12. S. Davidson, S. Forte, P. Gambino, N. Rius, A. Strumia: hepph/0112302 13. K. McFarland, S.O. Moch: Proceedings of MiniWorkshop on Electroweak Precision Data and the Higgs Mass, hepFig. 2. NuTeV's results for the strange quark momentum ph/0306052 asymmetry, s{x) — s(x), using different parameterizations. The 14. S. Kretzer: Procedings of the Rencontres de Moriond "A€ — a" parameterization, allowing differences in magnitude on QCD and HighEnergy Hadronic Interactions, hepand power in (1x) is shown in blue; a parameterization sugph/0405221 gested by CTEQ for s~^,s~ is shown in red; and a sample pos15. B. Dobrescu, R.K. Ellis: Phys. Rev. D 69, 114014 (2004) itive asymmetry using that parameterization with the central 16. "A Combination of Preliminary Electroweak Measurevalue of [31] is shown in black ments and Constraints on the Standard Model", CERNEP/200198, hepex/0112021 17. M. Griinewald: private communication, for the fit of [16] 2.2.1 Strange Sea Asymmetry without neutrinonucleon scattering data included 18. G.P. Zeller et al.: "On the effect of asymmetric strange If the strange sea is generated by purely perturbative seas and isospinviolating parton distribution functions Q C D processes, then neglecting electromagnetic effects, on sin^ Ow measured in the NuTeV experiment," hepone expects {s{x)) = (s(x)). However, it has been noted ex/0203004 t h a t nonperturbative Q C D effects can generate a sig 19. E. Ma, D.P. Roy: Phys. Rev. D 65, 075021 (2002) nificant m o m e n t u m asymmetry between the strange and 20. P. Vilain et al: Phys. Lett. B 335, 248 (1994) antistrange seas [27,28,29,30]. Outside the context of the 21. W. Loinaz, N. Okamura, T. Takeuchi, L.C.R. WijewardNuTeV electroweak data, measurements of this momenhana: Phys. Rev. D 67, 073012 (2003) t u m asymmetry constrain the properties of the intrinsic 22. U.K. Yang et al., [CCFR/NuTeV Collaboration]: Phys. Rev. Lett. 86, 2742 (2001) strange sea of the nucleon and helps to discriminate among 23. G.A. Miller, A.W. Thomas: hepex/0204007 the models suggested above. ^ ii^ii~X^ the 24. S. Brodsky, I. Schmidt, J.Y. Yang: SLACPUB9677, By measuring the processes I'N.^N USMTH136, hepph/0409279 C C F R and NuTeV experiments constrain the difference 25. A. Martin, R. Roberts, W. Stirling, R. Thorne: IPPPbetween the m o m e n t u m distributions of the strange and 0462, DCPT04124, CAVENDISHHEP200428, hepantistrange seas. A recent analysis from the C T E Q colph/0411040 laboration [31] has claimed t h a t this d a t a favors a positive 26. R.S. Thorne (Cambridge U.): Int. J. Mod. Phys. A 19, S — S, perhaps large enough to explain one sigma of the 1074 (2004) NuTeV discrepancy. However, the NuTeV fully NLO Q C D 27. A.I. Signal, A.W. Thomas: Phys. Lett. B 191, 205 (1987) analysis [32], does not confirm these results and instead 28. M. Burkardt, B.J. Warr: Phys. Rev. D 45, 958 (1992) weakly prefers negative S — S, as illustrated in Fig. 2. 29. S. Brodsky, B. Ma: Phys. Lett. B 381, 317 (1996) T h e constraint represented by this analysis, with an un 30. W. Melnitchouk, M. Malheiro: Phys. Lett. B 451, 224 (1999) certainty t h a t translates to less t h a n one NuTeV s t a n d a r d deviation in sin^ Ow, makes it unlikely t h a t a positive S—S 31. S. Kretzer et al.: Phys. Rev. Lett. 93, 041802 (2004); F. 01ness et al.: MSUHEP030701, BNLNT0317, RBRC329, is responsible for what we observe. hepph/0312323 32. D. Mason [NuTeV CoUaboration]: Proceedings of 39th Acknowledgements. We gratefully acknowledge support for Rencontres de Moriond on QCD and HighEnergy this work from the U.S. Department of Energy, the National Hadronic Interactions, hepex/0405037 Science Foundation and the Alfred P. Sloan Foundation. The
V \l2
Hadronic structure... and more P V in nuclear systems
Eur Phys J A (2005) 24, s2, 167170 DOI: 10.1140/epjad/s200504042x
EPJ A direct electronic only
Parity violation in astrophysics C.J. Horowitz Nuclear Theory Center and Department of Physics, Indiana University, Bloomington, IN 47405, USA Received: 15 October 2004 / Pubhshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. Core collapse supernovae are gigantic explosions of massive stars that radiate 99% of their energy in neutrinos. This provides a unique opportunity for large scale parity or charge conjugation violation. Parity violation in a strong magnetic field could lead to an asymmetry in the neutrino radiation and recoil of the newly formed neutron star. Charge conjugation violation in the neutrinonucleon interaction reduces the ratio of neutrons to protons in the neutrino driven wind above the neutron star. This is a problem for rprocess nucleosynthesis in this wind. On earth, parity violation is an excellent probe of neutrons because the weak charge of a neutron is much larger than that of a proton. The Parity Radius Experiment (PREX) at Jefferson Laboratory aims to precisely measure the neutron radius of ^°^Pb with parity violating elastic electron scattering. This has many implications for astrophysics, including the structure of neutron stars, and for atomic parity nonconservation experiments. PACS. 25.30.Bf Elastic electron scattering  ll.30.Er Charge conjugation, parity, time reversal, and other discrete symmetries  26.60+c Nuclear matter aspects of neutron stars  97.60.Bw Supernovae
1 Introduction Can there be large scale parity violation in astrophysics? For example, the 1960s science fiction movie Journey to the Far Side of the Sun visits a parity double of earth. In the real world, core collapse supernovae provide a unique opportunity for macroscopic parity violation. These gigantic explosions of massive stars are dominated by neutrinos t h a t transport 99% of the energy. This is because no other known particle can diffuse through the very dense m a t t e r of the collapsed star. In Sect. 2 we discuss how the weak interactions of these neutrinos may lead to large scale parity or charge conjugation violation. On earth, parity violation can be used to probe neutrons because the weak charge of a neutron is much larger t h a n t h a t of a proton. In Sect. 3 we describe the Parity Radius Experiment ( P R E X ) at JeflFerson Laboratory t h a t aims to use parity violating electron scattering to measure the neutron radius in ^^^Pb. This measurement has many implications for astrophysics including properties of neutron stars, nuclear structure, and atomic parity nonconservation.
2 Parity violation in supernovae In this section we explore large scale parity or charge conjugation violation in core collapse supernovae. W h e n the center of a massive star has burned to ^^Fe, thermonuclear reactions cease and the core of the star collapses, Send offprint requests to: email: [emailprotected]
in a fraction of a second, all the way to nuclear densities. This forms a protoneutron star with half the mass of the sun and a radius of order 10 kilometers. T h e protoneutron star is initially hot and lepton rich. Over a period of seconds, its very large gravitational binding energy of over 100 MeV/nucleon is radiated away in neutrinos. Note, t h a t any electromagnetic or strongly interacting particles diffuse very slowly at these high densities. This immense neutrino burst involves 10^^ neutrinos and carries about 3 X 10^^ ergs. In an historic first, about 20 neutrinos were detected from SN1987A. T h e protoneutron star is so dense t h a t even weakly interacting neutrinos diffuse. They interact thousands of times before leaving the star. Could there be a signature of all of these weak interactions? One possibility is parity violation in a strong magnetic field. This may lead to an asymmetry in the number of neutrinos emitted along the field compared to those emitted against the field, see Fig. 1. T h e neutrinos carry away so much m o m e n t u m t h a t a few % asymmetry in the neutrino radiation can lead to a recoil of the neutron star at hundreds of kilometers per second [1]. Indeed, many neutron stars are observed to have large galactic velocities t h a t presumably came from kicks at birth of 500 k m / s or larger. Unfortunately, explicit calculations of the neutrino asymmetry from parity violation have proved complicated and uncertain. A very strong magnetic field, of order 10^^ gauss or more, may be required to produce the few percent asymmetry t h a t can explain the observed velocities of neutron stars. Charge conjugation (C) violation is closely related to parity (P) violation. In the s t a n d a r d model, the product C P is approximately conserved so the large P violation
C.J. Horowitz: Parity violation in astrophysics
168
larger than that of a proton. Determining neutron densities of heavy nuclei has important implications for the neutron rich matter that is present in many astrophysical objects. In this section we describe the Parity Radius Experiment (PREX) at Jefferson Laboratory to measure parity violation in elastic electron scattering from ^^^Pb, and mention some of its implications. The parity violating asymmetry A for elastic electron scattering provides a purely electroweak probe of neutron densities. In Born approximation, A is
nu
ProtoNeutron Star
GFQ^
47r2V2Q,
0 nu Fig. 1. Parity violation in a strong magnetic field could lead to an u p / down asymmetry in the neutrino flux
implies large C violation. As a result of C violation, neutrino nucleon cross sections are systematically larger than antineutrinonucleon cross sections. This changes the composition of the neutrino driven wind during a supernova and may be important for nucleosynthesis. The intense neutrino flux blows some baryons oflF of the neutron star and into a high entropy wind. As this wind cools, the nucleons can condense to form heavy nuclei. This wind is the most promising site for rprocess nucleosynthesis. In the rprocess, seed nuclei rapidly capture several neutrons to form about half of the elements heavier than Fe. The ratio of neutrons to protons in the wind is set by the rates of neutrino and antineutrino capture. i^e \ n ^ p \ e
(1)
Ue \ p ^ n \ e'^
(2)
Because of C violation, the cross section for 2 is significantly smaller than the cross section for 1. This decreases the ratio of neutrons to protons in the wind by 20%. As a result, the wind is unlikely to be significantly neutron rich [2] and present simulations do not have enough neutrons to synthesize all of the rprocess elements. Furthermore, it appears very difficult to change conditions to make the wind significantly neutron rich. This is an important problem for rprocess nucleonsynthesis in the neutrino driven wind. Possible alternative sites, although these also have problems, include neutron star mergers and accretion disks around black holes. The site of the rprocess remains an important open problem in nuclear astrophysics.
3 The parity radius experiment (PREX) Parity violation is a uniquely clean probe of neutron densities. This is because the weak charge of a neutron is much
Fw{Q^)/Fch{Q^),
(3)
and involves the ratio of the weak form factor Fw{Q^) to the charge form factor FchiQ'^) The charge form factor is just the Fourier transform of the charge density at the momentum transfer Q of the experiment and is known from electron scattering. Likewise the weak form factor is the Fourier transform of the weak charge density. This directly gives the neutron density since most of the weak charge is carried by neutrons. At low momentum transfers A is very sensitive to the difference in neutron and proton root mean square radii. A heavy nucleus such as ^^^Pb is expected to have a neutron rich skin because of the neutron excess and because of the Coulomb barrier that prevents protons from being at large distances. The thickness of this skin has many implications for the properties of neutron rich matter as we discuss below. However the skin thickness has never been cleanly measured because all hadronic probes of neutron densities have significant uncertainties from the strong interaction. Parity violation provides a very clean way to measure the neutron skin thickness. The Parity Radius Experiment (PREX) is proposed at Jefferson Laboratory to measure A for elastic scattering of 850 MeV electrons off ^^^Pb at a scattering angle of six . This allows degrees [3]. The goal is to determine A to one to determine the neutron rms radius to 1% 5 fm) [4]. Because it is a purely electroweak reaction, the theoretical interpretation of the results is very clean. 3.1 Coulomb distortions and transverse analyzing power
A heavy nucleus has a large charge Z that will significantly distort the electron waves. This invalidates the simple Born approximation formula in 3. However, Coulomb distortions can be included exactly by numerically solving the Dirac equation for an electron moving in the Coulomb potential plus the very small (axialvector) weak potential [5]. This involves summing over a very large number of partial waves, because of the long range of the Coulomb potential. However, there are standard techniques for improving the convergence of this sum. We find that Coulomb distortions reduce A by about 30% [5]. Since the experimental goal is a 3% measurement it is very important that distortions are accurately included. One way to check the distortions is to look at the parity allowed transverse analyzing power Ay. If the initial
C.J. Horowitz: Parity violation in astrophysics
1006
2006 h
3c06h
4006.
25 !0 15 SciiUcring Angle (degrees) Fig. 2. Analyzing power Ay for elastic scattering of 850 MeV electrons from ^°^Pb versus laboratory scattering angle electron spin is transverse to the scattering plane, the analyzing power Ay gives the asymmetry for scattering to the left compared to scattering to the right. Time reversal invariance requires Ay to vanish in Born approximation. However, Ay is nonzero when Coulomb distortions are included. Therefore, Ay provides a direct test of the same Coulomb distortions. Furthermore, a nonzero Ay could be an important source of systematic errors. This is because the beam may have a small residual transverse polarization, and this coupled with Ay could lead to a false asymmetry t h a t changes sign with the beam helicity. There is considerable interest in Ay for electronproton scattering [6]. In Fig. 2 we show the transverse asymmetry Ay for the scattering of 850 MeV electrons off ^^^Pb. This is based on an exact numerical solution of the Dirac equation t h a t sums u p any number of photon exchanges. T h e nucleus is assumed to remain in its ground state at all times. Intermediate states involving excited states of P b are not included. However, the elastic scattering is coherent and leads to a cross section proportional to Z^. Each intermediate state, involving the excitation of a given proton, is not coherent and contributes a cross section of order unity. Even when one sums over the Z protons this only gives a total contribution of order Z and this is small compared to the Z^ coherent contribution. Therefore our elastic approximation should be good for a heavy nucleus like ^^^Pb. We find t h a t Ay is relatively large, comparable in magnitude to A and has significant structure in diffraction minima. It would be useful to measure Ay using a modest amount of running with transversely polarized beam. This can be done during P R E X . Note, t h a t Ay grows with Z because it is sensitive to Coulomb distortions. Therefore Ay is smaller for lighter nuclei. We will present more Ay results in a future publication [7].
3.2 Atomic parity nonconservation and PREX Atomic parity nonconservation experiments provide imp o r t a n t low energy tests of the s t a n d a r d model. Parity
169
violation involves the overlap of atomic electrons with the weak charge density of the nucleus and this is primarily carried by the neutrons. Therefore, precision atomic experiments are sensitive to the neutron radius. In the future, the most accurate low energy parity test of the standard model may involve the combination of an atomic experiment and P R E X to constrain the neutron density [4]. Note, t h a t the Colorado Cs experiment is presently limited by atomic theory uncertainties in the electron density at the nucleus [8]. One way to reduce these electron wave function uncertainties is to measure the ratio of parity violation in two different isotopes. Such a ratio measurement can be very sensitive to the neutron radii of the isotopes. These radii can be constrained by P R E X or future parity violating electron scattering experiments.
3.3 The equation of state of neutron rich matter, PREX, and neutron stars Measuring the neutron radius in P b constrains the Equation of State (EOS), pressure versus density, of neutron rich matter. T h e pressure forces neutrons in the skin of a heavy nucleus out against the surface tension. Therefore, the higher the pressure, the larger will be the neutron radius. Thus a measurement of the neutron radius in ^^^Pb determines the pressure of neutron m a t t e r at just below normal nuclear density [9]. A neutron star is a gigantic nucleus, see for example [10] for an introduction. Its structure depends only on the equations of General Relativity and the E O S of neutron rich matter. T h e central density of a neutron star can be a few or more times the nuclear density. Therefore, measuring the radius of a neutron star, which is expected to be of order 10 km, will determine the E O S at high densities [11]. Astronomers are working hard to measure neutron star radii. One approach is to determine b o t h the Xray luminosity and surface t e m p e r a t u r e . Thermodynamics allows one to determine the surface area assuming black body radiation plus model corrections for nonblack body effects in the atmosphere [12]. It is very interesting to compare the low density information on the EOS from the P b radius with the high density information from a neutron star radius. There are many possible exotic phases of Q C D at high densities. These include quark matter, strange matter, pion or kaon condensates, and color superconductivity. An exotic phase could show u p as an abrupt softening of the E O S with increasing density. Note, if an exotic phase has a stiff EOS (high pressure) then it will not be thermodynamically favored. T h e following scenario would provide an exciting indication of an exotic high density phase. P R E X could measure a large P b radius. This would show t h a t the EOS is stiff at low density. If the radius of a neutron star then t u r n s out to be small, this shows t h a t the high density E O S is relatively soft. Although these measurements would not, by themselves, determine the n a t u r e of the high density phase, they would strongly suggest interesting behavior. It has proved very difficult to find other signatures
C.J. Horowitz: Parity violation in astrophysics
170
of an exotic phase in the center of neutron stars. Finally, we note t h a t P R E X has many other implications for the solid crust of neutron stars [13] and for how neutron stars cool [14].
3.4 Neutron rich nuclei and PREX P R E X also has important implications for the structure of neutron rich nuclei t h a t can be studied in radioactive beams. T h e Rare Isotope Accelerator (RIA) would produce such nuclei, and their properties are interesting for several reasons. For example as discussed in Sect. 2, rprocess nucleosynthesis involves the capture of many neutrons to produce very neutron rich nuclei. T h e radius of P b determines the density dependence of the symmetry energy S{p). T h e symmetry energy describes how the energy of nuclear m a t t e r rises as one moves away from equal numbers of neutrons and protons. It is very important for the structure of neutron rich nuclei. T h e energy of pure neutron m a t t e r at density p, Eneutronip), IS approximately, E,neutron (p) ^ Enuclearip)
+
S(p),
(4)
where Enudear is the energy of symmetric nuclear m a t t e r (N = Z) and is largely known. Determining the pressure of neutron m a t t e r by measuring the neutron radius gives dEneutron/dp ^ud this givcs dS/dp. Thus P R E X will determine the density dependence of the symmetry energy which is poorly constrained at present.
4 Conclusions: Parity violation and astrophysics Core collapse supernovae are dominated by neutrinos. This provides a unique opportunity for large scale parity or charge conjugation violation. Parity violation in a strong magnetic field could lead to an asymmetry in the neutrino radiation and recoil of the newly formed neutron star. Charge conjugation violation in the neutrino driven wind above a neutron star reduces the ratio of neutrons to
protons. This is a large problem for rprocess nucleosynthesis in the wind. On earth, parity violation probes neutrons because the weak charge of a neutron is much larger t h a n t h a t of a proton. T h e Parity Radius Experiment ( P R E X ) at Jefferson Laboratory aims to measure the neutron radius of ^^^Pb to 1% by using parity violating elastic electron scattering. Because it is an electroweak reaction, the theoretical interpretation of this measurement is very clean. T h e neutron radius of P b has many implications for atomic parity nonconservation experiments, neutron stars, and the structure of neutron rich nuclei. Acknowledgements. Some of this work was done in collaboration with Jorge Piekarewicz. We thank E. D. Cooper for help on the Ay calculations. This work was supported in part by DOE grant DEFG0287ER40365.
References 1. C.J. Horowitz, J. Piekarewicz: Nucl. Phys. A 640, 281 (1998) 2. C.J. Horowitz: Phys. Rev. D 65, 083005 (2002) 3. P.A. Souder, R. Michaels, G. Urciuoli, spokespersons: Jefferson Laboratory Experiment E03011, See also http://hallaweb.jlab.org/parity/prex 4. C.J. Horowitz, S.J. Pollock, P.A. Souder, R. Michaels: Phys. Rev. C 63, 025501 (2001); See also http://cecelia.physics.indiana.edu/prex 5. C.J. Horowitz: Phys. Rev. C 57, 3430 (1998) 6. F.E. Maas et al.: nuclex/0410013 7. E.D. Cooper, C.J. Horowitz: to be published 8. Wood CS, Bennett SC, Cho D, Masterson BP, Roberts JL, Tanner CE, Wieman CE, Science 5307, 1759 (1997) 9. B.A. Brown: Phys. Rev. Lett. 85, 5296 (2000) 10. J.M. Lattimer, M. Prakash: Science 304, 536 (2004) 11. J. Carriere, C.J. Horowitz, J. Piekarewicz: Astrophys. J. 593, 463 (2003) 12. Jose Pons et al.: Astrophys. J. 564, 981 (2002) 13. C.J. Horowitz, J. Piekarewicz: Phys. Rev. Lett. 86, 5647 (2001) 14. C.J. Horowitz, J. Piekarewicz: Phys. Rev. C 66, 055803 (2002)
Eur Phys J A (2005) 24, s2, 171174 DOI: 10.1140/epjad/s2005040439
EPJ A direct electronic only
Parity violation in nuclear systems B. Desplanques Laboratoire de Physique Subatomique et de Cosmologie (UMR CNRS/IN2P3UJFINPG), F38026 Grenoble Cedex, France Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. Parity violation in nuclear systems is reviewed. A few ingredients relevant to the description of the parityviolating nucleonnucleon force in terms of meson exchanges are reminded. Effects in nuclear systems are then considered. They involve pp scattering, some complex nuclei and the deuteron system. PACS. 24.80.+y Nuclear tests of fundamental interactions and symmetries
1 Introduction A large number of paritynonconserving (pnc) effects has been observed in various nuclear systems. While their expected size at low energy is of the order of 10"'' (for the amplitude), they can be strongly enhanced in some cases, due to the closeness of states with opposite parities or the suppression of the regular transition. Thus, effects of the order of 10~^ have been measured in neutronnucleus scattering in the vicinity of low energy pwave resonances (see [1] for some review). Qualitatively, such effects are understood. However, little quantitative information could be obtained on the pnc component of nucleonnucleon (NN) forces expected to account for them. From the knowledge of this interaction, one can expect to learn about the pnc mesonnucleon coupling constants which they depend on and, thus, get information on the underlying hadronic physics. This one is complementary to the information t h a t can be obtained from nonleptonic hyperon decays. It concerns in first place the TTNN coupling t h a t has been at the center of many theoretical and experimental works. This one can be most easily compared to nonleptonic hyperon decay amplitudes. Another less fundamental but important motivation for the study of pnc nuclear effects is the necessity to determine the effective strength of the various pnc NN amplitudes. These ones can indirectly contribute to other pnc effects, especially in electron scattering mainly discussed at this meeting. Though the effect is not large, its knowledge is required to determine the reliability of the information t h a t is looked for in such high accuracy measurements. Some recent developments in the field are reviewed here. T h e plan of the paper is as follows. In the second part, we briefly remind ingredients entering the pnc NN force, while emphasizing a few points of interest for the following part devoted to pnc effects. T h e third section is concerned with pnc pp scattering. This process is the only one t h a t provides a calibration of the strength of pnc NN forces at the present time. A few nuclear pnc effects in complex
^ 71, p , CO
a— Q / ' ^ .0
O N.
'—'
9^'^ Q  ^^
N
^^^^^^
o
&
Fig. 1. Diagram representation of the pnc NN interaction
nuclei, especially in ^^F and in ^^^Cs, are discussed in the fourth section. The fifth section is devoted to pnc effects in the np system, including the deuteron. This particular field has been particularly active these last years. A conclusion and an outlook are presented in the sixth section.
2 PNC NN
potential: Ingredients
T h e pnc NN force is generally described as resulting from meson exchanges, TT, p and LJ. A diagrammatic representation is given in Fig. 1. One of the vertex, represented by a circle, corresponds to the strong interaction. T h e other one, represented by a box, corresponds to the weak, pnc, interaction. As isospin is not conserved, there are many couplings in some cases. They are:  h^, which governs the long range part of the force and necessarily involves the AT = 1 component of the weak interaction  / i ^ ' i ' 2 . AT = 0, 1, 2 hZ^^:AT = 0, 1  h]^: AT = 1, (different type pNN coupling). T h e vectormeson couplings determine the shortrange part of the pnc NN force. As is well known, the contribu
172
B. Desplanques: Parity violation in nuclear systems
tion of this part is sensitive to shortrange correlations in the strong NN interaction as well as to other correlations. Many contributions going beyond the above ones have been considered in the literature. They involve for instance twopion exchanges displayed in Fig. 1, either with the same coupling as for the onepion exchange or with the same coupling as for the p exchange. In the last case, the pexchange force acquires a longer range t h a t could show u p in the analysis of pnc effects in pp scattering. Some discussion and references could be found in [1]. At low energy, only gross features may be relevant. T h e NN interaction can then be parametrized by five S ^ P NN transition amplitudes [2]:  ^iS^o ^ ^Po , ^T = 0 , 1 , 2, (pp, pn and nn forces)  ^^'i ^ i P i , ZAT = 0 {pn force)  ^Si ^^Pi , AT =1 {pn force). It was shown t h a t this description could be extended to higher energy by singularizing the pionexchange contribution which, due to its long range, contributes sizeable P ^ D transition amplitudes [3]. Apart from the name, these works largely anticipated recent effective fieldtheory approaches [4], which also consider P ^ D transitions. Many works have been devoted to the pnc mesonnucleon couplings, which enter NN interaction models. A large part of them, prior to the DDH work [5] or later, fit in this framework. Due to the lack of space, we again refer to [1] for references and detailed discussion. We only present here some estimates and make a few pertinent remarks. T h e sample of results given in Table 1 corresponds to the predictions of two significantly different models for the most relevant couplings, h\, h^ and h^. They are based on a quark model (DDH), partly updated, and a chiral soliton model by Kaiser and Meissner (KM) [6] (see also [7]). Despite appearances, results t u r n out to be qualitatively similar. Discrepancies can be ascribed to the weight a t t r i b u t e d to individual contributions in DDH. It is noticed t h a t the dominant contribution to h\ is produced by strange quarks, of particular interest at this meeting while the consistency of this estimate with the Q C D sum rules ones remains an open problem. It was proposed to use a chiral quark model to make a new estimate (Lee et al. [8], this conference). It is also noticed t h a t D D H estimates, relying for a part on experimental data, should be less sensitive to "rescattering effects" evoked in the literature [9] whereas h^^ is likely to be negative.
3 Longitudinal asymmetry in pp scattering T h e lowenergy longitudinal asymmetry in pp scatering is the most important benchmark in the field at present. A complete theoretical analysis can be done. It shows t h a t measurements at 13.6 and 45 MeV are in complete agreement with each other, thus fixing the strength of the ^^0 ^ ^Po pnc pp transition amplitude. For a given description of the strong interaction model, the strength of a combination of h^^ and h^ couplings or, in first approximation, the h^ and h^ couplings, can be obtained. At
Table 1. Mesonnucleon pnc coupling constants: a few estimates from different works. The question mark at the last line indicates that the original value could be actually close to 0 DDH (range)
DDH("best" )
KM
W hi
0 ^ 11
4.6
0.2
W hi
0 ^ 2.5 (update, K = 3)
10^/1°
31 ^ 6
11
4
10^/1°
10 ^ 6?
2
6
0.8 1.3 (with ss)
A(107) 1
) ^  ^ p 100
1^6
CO
1
y^^^^ 1
\ I
\ 1S03P0
/
A
T(MeV)
3P21D2
y
2
Schematic
Fig. 2. Schematic representation of p and cjexchange contributions to the longitudinal asymmetry in pp scattering higher energy, around 221 MeV, it was noticed t h a t the contribution of the ^6*0 ^ ^Po transition amplitude was vanishing, providing a window to determine the ^P2 ^ ^D2 transition amplitude. By combining this measurement with the lowenergy one, contributions due to p and {jj exchanges can thus be disentangled. A schematic representation of the two contributions with couplings close to the D D H "best guess" ones, is given in Fig. 2. It is seen t h a t the cjexchange one has a negligible contribution around 221 MeV while a yoexchange contribution alone is not doing badly. A better agreement is obtained by increasing the strength of this one and compensating for the overestimate at low energy by adding a cjexchange contribution with a sign opposite to the D D H "best guess" or K M one. This provides a simple explanation for the couplings obtained by Carlson et al. [10]:
hf = 22.3 hf = 15.5
1 0  ^ hFJ 1 0  ^ hFJ
+5.2 10"^ (fit) 3.0 1 0  ^
(DDH).
(1)
T h e fit evidences a striking feature as the value for the uj coupling has a sign opposite to expectations and, thus, can point to missing ingredients in predictions. We however notice t h a t the significance of the result is not strong (the p alone is already giving a good account). A refined theoretical analysis and a more accurate measurement could be quite useful. It would be interesting to investigate for instance the role of a longer range pexchange contribution mentioned in the second section.
B. Desplanques: Parity violation in nuclear systems
173
4 Paritynonconservation in complex nuclei
5 Paritynonconservation in the deuteron
Many pnc effects in complex nuclei have been measured and analyzed. Considered individually, it is however difficult to draw some conclusion from them. Moreover, when they are considered together, it is not rare that the information obtained from one process is at the limit to contradict that one from another process. Two effects nevertheless deserve some attention: the circular polarization of photons emitted in the transition 0~(1.08 MeV) ^ l+(g.s.) in ^^F and the ^^^Cs anapole moment. They are successively discussed in the following (references for both theory and experiment may be found in [1]). The interest of the pnc effect in '^^F is that the calculation of the relevant pnc nuclear matrix element (0~ ^ 0"^) can be checked by looking at the first forbiden P decay of the neighboring nucleus, ^^Ne. It implies the AT = 1 part of the weak interaction and the measurement can thus provide information on the pnc TTNN coupling, /i^. From the experimental limit of P^^ one gets the following upper limit: \hi\< 1.310'. (2) This result is supported by the absence of effect in two other processes in ^^Ne and in ^^Tc. In these cases, the contribution of the pion exchange is a priori large. To agree with the upper experimental limit, one has first to assume that the coupling hi. is not too large and, moreover, that the corresponding contribution be cancelled for a part by some isoscalar contribution (for ^^Ne). Contrary to ^^F, there is no available check on the relevant pnc nuclear matrix element. Accepting that this one be uncertain by up to a factor 3 would however give a limit on /i^ similar to (2). The ^^^Cs anapole moment has been analyzed by different authors. To a large extent, this quantity involves a combination of the pnc NN force close to that one governing pnc effects in several oddproton systems as different as pa scattering, ^^F, "^^K, ^^^Lu, ^^^Ta. At first sight, it appears that the above combination should be two times larger for the anapole moment than for the other processes. The discrepancy has the order of a typical uncertainty in the field but there was some belief that the estimate in the first case could be less uncertain than for other effects (for a part, it involves a longrange operator). On the other hand, the effects in the other oddproton systems overdetermine the above combination of parameters. If these last processes are ignored, it appears that a large value of /i^, of the order of 110~^, at the upper limit of the original DDH range, is needed. This is inconsistent both with the upper limit, (2), and the DDH updated range. We notice that the last calculation of the anapole moment [11] relies on an approximation that allows for an improved calculation in one respect but implies some contribution from orbitals below the Fermi level with a wrong sign in another respect. A correct account of these ones could enhance the theoretical estimate but will not reach a factor 2. The validity of a similar approximation, which omits 3body terms, was discussed in [12].
Most recent pnc studies in nuclear systems have concentrated on the np system (deuteron and scattering state). This emphasis is largely motivated by both the feasibility of the corresponding experiments in a near future (see Stiliaris's talk at this conference) and a safer interpretability of possible effects. These ones include the photonemission asymmetry in the thermalenergy radiative capture of polarized neutrons by protons, n\p ^ d\j [13,14,15,16], presently performed at LANSCE, the asymmetry in the deuteron photodisintegration depending on the photon helicity [17,18,19,20], which could be performed at JLab, lASA, SPring8, , the deuteron anapole moment [21,22, 23,24], the pnc deuteron electrodisintegration in relation with the SAMPLE experiment [25,16], and the longitudinal asymmetry and the neutronspin rotation in np scattering [20]. Some earlier works could be quoted. The recent ones involve new methods (effectivefield theories [13, 15]), improved NN interaction models (AV18+ [15, 24,16,18,20]), more complete calculations (twobody currents [25,16]), and increased attention to gauge invariance [23,24,20]. A few remarks are made below about these different works. The earlier pionexchange contribution to the asymmetry in the thermalneutron radiative capture on protons, n + p ^ c/ + 7. 0.11 hi (3) is confirmed by recent estimates, indicating that the coro np scattering length, was approrection for a wrong priately made. It is also found that the above estimate results from a strong cancellation when a calculation is performed without relying on the Siegert theorem [15,16]. Amazingly, this weak interaction problem provides information on the accuracy of effectivefield theory methods employed for the strong interaction. The approach used in [13], for instance, overestimes (3) by 60% at leading order (almost a factor 2 for comparable ingredients). On the basis of an estimate by Oka [26], it was thought that the study of the photonhelicity dependence of the deuteron photodisintegration cross section could provide an alternative way to determine the coupling /i^. This motivated several works that disproved the above estimate and its main conclusion [17,18,19,20]. An account of the new results can be found in the Hyun's talk at this conference. For the inverse process near threshold ("Lobashov experiment"), it should be noticed that a circular polarization of photons as large as 1 10~^ is not excluded for some reasonable models of both the strong and the weak NN interaction [20]. The deuteron anapole moment is largely academic as there is not much hope it could be measured in a near future. It however provides a nice laboratory for studying the implications of gauge invariance, which is essential for getting a consistent estimate of this quantity. A contribution required by chiral symmetry [21], absent in [22], has thus been recovered in potential based approaches [23].
174
B. Desplanques: Parity violation in nuclear systems
On the other hand, this last work confirms the conclusion obtained from the study of A^ about the reliability of some effectivefield theories. It is likely t h a t an alternative approach [27], which is nothing but the one initiated by Danilov's work [2], extended later on to higher energy [3], should do better. In pncelectron experiments performed on the deuteron, aiming at determining the contribution of strange quarks to nucleon form factors, there was some concern about the role of a nuclear pnc effect. This one was studied in two different works [25,16] which showed t h a t the effect, a few percents, would be negligible. Actually, the main role of pnc nuclear effects in this process (together with t h a t one involving the proton) is an indirect one. They allow one to put limits on coupling constants t h a t enter radiative corrections [28]. Paritynonconservation in np scattering has been recently revisited [20]. T h e main feature evidenced by the new results is the dominance of the pionexchange contribution, as far as the DDH "best guess" is used for the corresponding pnc coupling.
To determine this sector of the pnc NN interaction, appropriate experiments are heavily needed, preferentially with light systems where theoretical uncertainties are reduced. T h e np amplitude with "polarized" neutrons is better studied in the neutronspin rotation. T h e nn amplitude could be obtained from the neutronspin rotation in n e u t r o n  a scattering, after removing the previous contribution of the np amplitude. T h e best process for determining the last np amplitude, which involves b o t h "polarized" neutrons and "polarized" protons, is the circular polarization of photons in the thermal neutronproton radiative capture ("Lobashov experiment"). Evidently, the asymmetry A^, already mentioned, is part of the needed experiments. While it involves the difference in the two np amplitudes with a "polarized" neutron and an "unpolarized" proton on the one hand, the inverse configuration on the other hand, it also allows one to get information on the most debated pnc coupling, /i^.
6 Discussion and conclusion
References
Many lowenergy pnc nuclear effects, involving mainly protons, are within expectations. However, one has often to be satisfied with discrepancies up to a factor 2. This is not enough to constrain the different pnc mesonnucleon couplings if one refers to a potential approach or the lowenergy A^A^ scattering amplitudes if one rather relies on the less ambitious approach represented by effective field theories. Most probably, the pnc TTNN coupling, /i^, is small and within the D D H u p d a t e d range. Some processes could require a significantly larger value but, in our opinion, they have not the weight of the other ones t h a t point to a small value. Concerning the vector mesonnucleon couplings, there is a slight hint t h a t the isoscalar cu one, /i^, could have a sign opposite to expectations. This should motivate further studies to confirm the hint on the one hand, to see whether this opposite sign is conceivable. An analysis in terms of couplings has some interest but, apart from the fact it assumes t h a t multimeson exchanges can be ignored, it does not necessarily provide a pertinent clue at which part of the pnc interaction is rather unconstrained. Looking at the various NN scattering amplitudes can thus represent a complementary view. Among the five amplitudes required for the description of pnc effects at low energy, only one (pp) is determined with a good accuracy. From the study of oddproton systems, and after removing the contribution of the pp amplitude, a pn amplitude involving "unpolarized" neutrons can be obtained. Being derived indirectly, from complex systems moreover, the accuracy of this amplitude is not as good as for the pp one. For the three other amplitudes, which involve "polarized" neutrons (with "unpolarized" protons, with "unpolarized" neutrons and with "polarized" protons), only upper limits are known.
1. B. Desplanques: Phys. Rep. 297, 1 (1998) 2. G.S. Danilov: Sov. J. Nucl. Phys. 14, 443 (1972); Phys. Lett. 18, 40 (1965); Phys. Lett. B 35, 579 (1971) 3. B. Desplanques, J. Missimer: Nucl. Phys. A 300, 286 (1978) 4. B. Holstein et al.: nuclth/0407087 5. B. Desplanques, J. Donoghue, B. Holstein: Ann. of Phys. 124, 449 (1980) 6. N. Kaiser, U.G. Meissner: Nucl. Phys. A 489, 671 (1988); A 499, 69 (1989); A 510, 759 (1990) 7. U.G. Meissner, H. Weigel: Phys. Lett. B 447, 1 (1999) 8. H.J. Lee et al.: hepph/0405217 9. S.L. Zhu et al.: Phys. Rev. D 63, 033006 (2001) 10. J. Carlson et al.: Phys. Rev. C 65, 035502 (2002) 11. W.C. Haxton et al.: Phys. Rev. C 65, 045502 (2002) 12. B. Desplanques: Phys. Lett. B 47, 212 (1973) 13. D.B. Kaplan et a l : Phys. Lett. B 449, 1 (1999) 14. B. Desplanques: Phys. Lett. B 512, 305 (2001) 15. C.H. Hyun et al.: Phys. Lett. B 516, 321 (2001) 16. R. Schiavilla et al.: Phys. Rev. C 67, 032501R (2003) 17. LB. Khriplovich et al.: Nucl. Phys. A 690, 610 (2001) 18. C.P. Liu, C.H. Hyun, B. Desplanques: Phys. Rev. C 69, 065502 (2004) 19. M. Fujiwara et al.: Phys. Rev. C 69, 065503 (2004) 20. R. Schiavilla, J. Carlson, M. Paris: nuclth/0404082 21. M.J. Savage, R.P. Springer: Nucl. Phys. A 644, 235 (1998); A 657, 457(E) (1999) 22. LB. Khriplovich et al.: Nucl. Phys. A 665, 365 (2000) 23. C.H. Hyun, B. Desplanques: Phys. Lett. B 552, 41 (2003) 24. C.P. Liu, C.H. Hyun, B. Desplanques: Phys. Rev. C 68, 045501 (2003) 25. C.P. Liu et al.: Phys. Rev. C 67, 035501 (2003) 26. T. Oka: Phys. Rev. D 27, 523 (1983) 27. M.J. Savage: Nucl. Phys. A 695, 365 (2001) 28. S.L. Zhu et al.: Phys. Rev. D 62, 033008 (2000)
Acknowledgements. We are very grateful to C. H. Hyun and C.P. Liu for a stimulating and fruitful collaboration.
Eur Phys J A (2005) 24, s2, 175178 DOI: 10.1140/epjad/s2005040448
EPJ A direct electronic only
Parity violation in nuclear systems Experimental considerations in the deuteron photodisintegration with polarized photons Efstathios Stiliaris^ Physics Department, University of Athens, GR15771 Athens, Greece and Institute of Accelerating Systems & Apphcations (lASA), P.O. Box 17214, GR10024 Athens, Greece Received: 15 December 2004 / Pubhshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. Experimental measurements of Parity NonConserving (PNC) asymmetries in simple nuclear systems represent always a keytool for the study of the weak nucleonnucleon interaction and consequently an accurate experimental method for the determination of the mesonnucleon weak coupling constants of the underlying theory. Recent theoretical analysis on the deuteron photodisintegration with polarized photons, a few MeV above threshold, have drastically improved previous theoretical estimates. Based on that, the feasibility of measuring the photon asymmetry A^ in the reaction j\d ^ n\p with the 10MeV CW Linac at the Institute of Accelerating Systems and Applications (lASA) is considered here. A brief review on previous experimental results obtained in the deuteron photodisintegration and in the thermalneutron radiative capture on protons (inverse reaction) is given. The most important parameters in the design of a nuclear parity experiment are presented and the crucial factors, such as beam intensity, beam polarization and neutron detection techniques with the required high accuracy are outlined. PACS. 24.80.+y Nuclear tests of fundamental interactions and symmetries  25.20.x Photonuclear reactions
1 Introduction T h e deuteron has played an important role in the study of the weak nucleonnucleon interaction. Together with scattering experiments the studies of Parity NonConserving (PNC) transitions in the nucleonnucleon system are very attractive because of the simplicity and the well understood structure of the system [1,2]. T h e first experiment which received a lot of attention during the 1970's was the "Lobashov" experiment. In this experimental study the Leningrad group has investigated the net polarization of the emitted photons in the radiative thermal neutron capture by proton, n \ p ^ d \ ^. T h e experimental work was characterized by the novel techniques applied for measuring the integrated current. T h e nonzero polarization obtained in this reaction [3], P^ =  ( 1 . 3 zb 0.45) X IQ^, which is 30 times larger in magnitude t h a n the theoretical prediction and, moreover, of opposite sign, motivated many theoretical calculations in the frame of strong and weak interaction models known at t h a t time. Later experimental work for the circular polarization P^ of the emitted 2.23 MeV photons reported values more consistent with theoretical estimates but with too poor accuracy to allow any definite conclusion about the strength of the P N C forces [4]. At present, a new P N C asymmetry measurement for the radiative neutronproton email address: s t i l i a r i s O p h y s . u o a . g r
capture with polarized neutrons, n \ p —> <^ + 7, is in preparation at LANSCE in order to reduce the experimental error of the spatial asymmetry A^ of the emitted photons [5,6]. Another tool to study P N C forces is the inverse reaction 7 + d ^ n + p, where a deuteron is disintegrated by absorbing a circularly polarized photon. T h e asymmetry Aj in this reaction near threshold is expected to be sensitive to the same components of the weak nucleonnucleon interaction as the thermal neutron capture. A key measurement of this asymmetry has been also performed in the past at the Chalk River Nuclear Laboratories, established as a "Chalk River" experiment [7]. Unfortunately, the results of this measurement were characterized by poor control on the systematic errors, mainly due to b e a m instabilities. W i t h the most recent advances in the beam instrumentation and in the detection techniques, taking also in account the new theoretical considerations on this subject, a measurement of A^ becomes nowadays more realistic. T h e feasibility of measuring the photon asymmetry A^ in the deuteron photodisintegration at low energies with the 10MeV C W Linac at the Institute of Accelerating Systems and Applications (I AS A), Athens, is considered here. First, a brief review on the recent theoretical calculations and predictions of the expected asymmetry will be given. T h e most important parameters in the design of a nuclear parity experiment will be presented and the
E. Stiliaris: Parity violation in nuclear systems
176
crucial factors, such as beam intensity, b e a m polarization and neutron detection techniques with the required high accuracy will be outlined. At the end, the lASA accelerator and the possibilities of using it as a dedicated machine for the nuclear parity study will also be discussed.
2 Theoretical calculations and predictions T h e asymmetry in the deuteron photodisintegration 7 d ^ n\p oi diii unpolarized target is defined as a^
— CF
yi^ —
<^+
(j
where cr+ and a_ denotes the total cross section using right and lefthanded polarized photons respectively. This asymmetry A^ becomes asymptotically equal to the circular polarization P^ of the emitted 2.23MeV photons in the thermal neutron capture by protons (inverse reaction n\p ^ d\^) at threshold energy. First calculations of A^ done by Lee [8], u p to photon energy 1 MeV above the disintegration threshold, show a reasonable result within the theoretical range of P^ in this energy domain, where the regular transition M l dominates. However, a later work by Oka [9], extended to higher photon energies uj^ ^ 35 MeV, suggests t h a t P N C effects are different compared to low photon energy, since the major contribution to the cross section comes now from the E l transition (compare also Fig. 1). Due to the initiated parityviolating 7rexchange contribution, it was predicted t h a t A^ shows a great enhancement at energies cjj > b MeV. T h e experimental observation of this enhancement would provide an important and unambiguous determination of the weak TTNN coupling constant /i^. Unfortunately, a recent calculation of A^ by Khriplovich and Korkin [10] showed critical contradictions to Oka's result, with a huge suppression of A^ at energies uj^ >3 MeV. It seems t h a t in Oka's work, a nonvanishing value for the pionexchange contribution to the asymmet r y was obtained as a result of the incomplete account for Podd mixing, since only the ^Pi admixture to the deuteron ground state was included there. In order to examine the situation in a most accurate way, Liu, Hyun and Desplanques [11] did a careful calculation of this process by completing the missing parityadmixed components in the final state, in particular in the ^Pi channel, and by including tensor and spinorbit forces in the nucleonnucleon interaction using the realistic Argonne AV18potential. T h e result of this improved work confirmed the strong suppression of the 7rexchange contribution to the asymmetry and is illustrated in Fig. 2. At the same time, another independent theoretical work by Fujiwara and Titov [12] analyzed the energy dependence of the Aj asymmetry in the deuteron photodisintegration. Although this second paper was based on the Paris potential with soft repulsion at short distances and the H a m a d a  J o h n s t o n potential, b o t h studies come to consistent to each other results. T h e predicted theoretical values for the asymmetry A^ of all mentioned models are summarized in Table 1.
0 2 4 6 8 10 E, E„, (MeV) Fig. 1. Total cross section for the deuteron photodisintegration as a function of the energy excess above the disintegration threshold, calculated with the Paris potential (From [12]). The contributions of the Ml (dotdashed) and E l (dashed) transitions are shown together with experimental data Table 1. Predicted values for the Aj 7 + (i ^ n + p reaction
asymmetry in the
7Energy
A,
Dependence
Ref.
10 MeV 30 MeV threshold 3 MeV threshold
^ 2.4 X 10"^  5 . 7 x 10^ ^ 1.0 X 10"^  1.0 X 10"^  2.5 X 10"^
hi hi
[9] [9]
Ml dominance Ml dominance h^ h^ h^
[10] [10]
Itpj
Itpj
It^j
[11],[12]
T h e experimental consideration following in the next section is based on the most realistic estimates by Liu, Hyun and Desplanques [11], which are in agreement with the results by Fujiwara and Titov [12]. T h e predicted value A^ ~ 2.5 X 10~^ near threshold will be the starting point for the discussion of possible future experiments measuring this asymmetry in the deuteron photodisintegration with polarized photons. From Fig. 2 it is clear t h a t , when the photon energy gets larger, the asymmetry gets smaller; as the the photon energy reaches 1 MeV above the threshold, the asymmetry drops by an order of magnitude.
3 The future experiment In the "Chalk River" experiment the obtained results for the A^ asymmetry were mainly affected by the big systematic errors. T h e final values reported there A^ = (2.7 =b 2.8) X10^ at E^ = 4.1 MeV and A^ = (7.7=b5.3) x IQ^ at E^ = 3.2 MeV [7] put an experimental upper limit on the value of the P N C effect. But, as this experimental study
E. Stiliaris: Parity violation in nuclear systems 0.251
1
1
r
1
r
1
1
1
1
1
1
1
1
p
1
1
Pobrised & Beom E = 3  8 McV
'I y \ d —^ n \ p 
OJI c 1^ o.l^i^ ^'
1
1
' ' 
1 \
"[
= 200MA
P = 80%.
^ ,
u 1 = oilo,D^l'
I
1
1
i
\
i
1
1
1
1
1
1
1
1
1
1 — J — 1  ^
Fig. 2. Theoretical prediction for the asymmetry Aj in the deuteron photodisintegration with polarized photons. The calculation is based on the DDH best values, taking in account tensor as well as spinorbit forces in the nucleonnucleon interaction (From [11])
concluded, "... it is possible to consider future measurements with techniques similar to those used in the present measurement but having improved b e a m intensity, beam polarization and control of systematics" [7]. T h e experimental goal is therefore to reduce the systematic errors to a level better t h a n 10~^ in measuring the neutron asymmetry of the reaction in the deuteron photodisintegration with polarized photons. T h e most imp o r t a n t factors in the design of such an experiment are: — Reasonable flow of polarized photons — An improved neutron detection system — Beam quality and stability with fast feedback systems Starting with some conservative numbers all experimental requirements are presented in the following section.
3.1 Experimental considerations T h e polarized photons are produced via Bremsstrahlung with a high density polarized electron beam and a gold radiator. Development of a highly polarized electron beam with energy stability on the level of a few 10~^ can be routinely achieved. Having in mind a recent LetterOfIntent presented at J L a b for a similar project [13], following parameters are outlined here for the proposed apparatus: — Beam energy in the range 2.3  8 MeV (200/iA) — Beam polarization P^ ^ 80% — P h o t o n production target (radiator) made of 1mm Au plate — 30 cm long Liquid Deuterium {LD2) target — Main detector consists of two components (slow neutrons & photons) — Lead shielding to reduce the intensity of scattered photons on neutron detectors — Heavy water moderator to slow down neutrons — Additional Compton detectors for monitoring
177
Au Radiator f = 60%
Photon Flux $ = 6 X 10'^ ph/sec Efficiency = 4S7,
I
Poiorized y's $ = 3 X 10^^ ph/sec Reaction Absorption Length 1.4 X lO^^cm REACTION VtELD 6 X 10^ n/sec
LD2 Target {30cm) d{v.n)p
i ^^B DETECTOR Efficiency ^ 50% 3x10=^ Hz
~i~
Events nsedcd 2 X 10'^
I Aver. Physics Asymmetry
0.5x10^ Experimer>tQl AccurQcy
t>AQ T I M E  77 days J!!
Fig. 3. Experimental statistics expected in a future experiment and the required acquisition time for data taking
Statistics obtained with the above mentioned parameters are presented in the diagram of Fig. 3. It is clear from this simple analysis, t h a t the D a t a Acquisition (DAQ) time needed to successfully perform such an experiment and to reach the required accuracy comes to the order of several months! T h e experimental a p p a r a t u s is shown in Fig. 4. T h e main detector has segmentation on the forward and the back parts to get sensitivity to the directional asymmetry in addition to P N C asymmetry in the total cross section. T h e expected high counting rate of the neutron detectors (~ 1^^'^Hz) will help to study tiny systematic effects. The choice of the spin flip frequency has to be optimized for minimum fluctuation of the beam energy and position.
3.2 The lASA electron accelerator A 10MeV C W linear electron accelerator has been already installed at the Institute of Accelerating Systems & Applications (IASA), which is currently under commissioning. This machine comprises a thermionic electron gun, a 100 keVLine with a buncherchopping system followed by a 5 MeV Linac with R F structures of the sidecoupled type and a 4mbooster section of the same type [14]. Both are powered with a 500 k W multicavity C W klystron amplifier at 2380 MHz. T h e machine is hosted in the basem*nt of the lASA building, which is extended with a new experimental hall (Fig. 5). T h e lASA accelerator meets exactly the needs for a future parityviolation experiment as previously discussed. T h e energy range is optimally covered by this machine and
E. Stiliaris: Parity violation in nuclear systems
178
Lead Shield Heovy Wcicr
1 LCD,
I
PNC
V
N e u t r o n A Photon Detcctoi*5
D "H D
o
Forbvard Photon D e t e c t o r
RREPS
Detector Photon Beam D j m p
Fig. 4. Experimental setup for the study of the 7 + ( reaction in a future experiment
n\p Fig. 6. Experimental area in the accelerator vault. Indicated are the area reserved for the future parity experiments (PNC) together with the area devoted to novel radiation sources (RREPS)
a precise measurement of parityviolating forces seems to be feasible. T h e 10MeV lASA electron accelerator could serve as a machine dedicated to this kind of research. Acknowledgements. I am indebted and would like to thank B. Desplanques from LPSC Grenoble, and B. Wojtsekhowski from JLab for the valuable suggestions and discussions on the deuteron parityviolation subject. I am also grateful to Prof. C.N. Papanicolas from the University of Athens for his continuous interest to this project. Fig. 5. The 10MeV CW electron hnear accelerator at lASA
References the beam current and beam characteristics, already measured at 100 keV, could guarantee a high quality beam. At present, the beam is not polarized, but there are fut u r e plans for the installation of a high intensity polarized electron gun. Space has been already reserved in the accelerator vault for future parity experiments as indicated in Fig. 6. Taking into account the long acquisition time required by the parityviolation experiments, the I AS A 10MeV electron accelerator could ideally be devoted to the research of the P N C studies in the deuteron photodisintegration.
4 Conclusion Parity NonConserving (PNC) experiments at low energy got large attention in the last years. W i t h the recent theoretical improvements in the calculation of the asymmetry in the deuteron photodisintegration induced by polarized photons a few MeV above threshold and with the technical progress in the accelerator domain with polarized beams.
1. B. Desplanques, J.F. Donoghue, B.R. Holstein: Ann. Phys. (N.Y.) 124, 449 (1980) 2. B. Desplanques: Phys. Rep. 297, 1 (1998) 3. V.M. Lobashov et al.: Nucl. Phys. A 197, 241 (1972) 4. V.A. Knyazkov et al.: Nucl. Phys. A 417, 209 (1984) 5. V.M. Snow et al.: Nucl. lustrum. Methods A 440, 729 (2000) 6. V.M. Snow et al.: Nucl. lustrum. Methods A 515, 563 (2003) 7. E.D. Earle et al.: Can. J. Phys. 66, 534 (1988) 8. H.C. Lee: Phys. Rev. Lett. 41, 843 (1978) 9. T. Oka: Phys. Rev. D 27, 523 (1983) 10. LB. Khriplovich, R.V. Korkin: Nucl. Phys A 690, 610 (2001) 11. C.P Liu, C.H. Hyun, B. Desplanques: Phys. Rev. C 69, 065502 (2004) 12. M. Fujiwara, A.L Titov: Phys. Rev. C 69, 065503 (2004) 13. B. Wojtsekhowski, W.T.H. van Oers: JLab LetterOfIntent 00002 for PAC 17 (2000) 14. E. Stiliaris et al.: Proceedings of EPAC 2000, loP, pp. 866868
Eur Phys J A (2005) 24, s2, 179180 DOI: 10.1140/epjad/s2005040457
EPJ A direct electronic only
Parity violating asymmetry i n 7 + d ^ ^ n + p a t low energy C.H. H y u n \ C.P. Liu^, and B. Desplanques^ ^ Institute of Basic Science, Sungkyunkwan University, Suwon 440746, Korea ^ Kernfysisch Versneller Instituut, Zernikelaan 25, Groningen 9747 AA, The Netherlands 3 Laboratoire de Physique Subatomique et de Cosmologie (UMR CNRS/IN2P3UJFINPG), F38026 Grenoble Cedex, France Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. We calculate the parityviolating asymmetry in the 7 + d ^^ n + p process where the deuteron is disintegrated by circular photons. The photon energy is considered up to 10 MeV above threshold, where the lowest electromagnetic transition modes M l and El dominate. We employ the Argonne vl8 potential for the strong interaction and the DDH potential for the parityviolating weak interaction of the twonucleon systems. The asymmetry is about 2.5 x 10~^ at threshold, and decreases rapidly to have magnitude of order 10~^ or less for photon energies larger than 3 MeV. The exchange of vector mesons dominates the asymmetry, while the pion contribution is negligible. PACS. 24.80.+y Nuclear tests of fundamental interactions and symmetries
1 Introduction T h e interest to test the s t a n d a r d model in the realm of nuclear and atomic physics becomes more and more growing and popular nowadays. T h e activeness of the subject has been driven by recent experiments: measurement of the anapole moment of ^^^Cs [1], parityviolating (PV) longitudinal asymmetry in p p scattering [2] and strange form factor of the nucleon from ep scattering [3]. In addition to these experiments, a P V asymmetry in up ^ 6/7 is being measured at LANSCE [4], and the possibility to measure a P V asymmetry in jd ^ np is also being discussed. These experiments provide important information about the weak interaction of hadrons at low energy. From the theoretical viewpoint, weak as well as strong interactions can be described by means of the oneboson exchange, which introduces P V mesonnucleon coupling constants. Knowing precise values of the P V coupling constants is an important ingredient to understand the nuclear weak interaction. Unfortunately, for some of the P V coupling constants, recent experiments mentioned above give incompatible values with those from other experiments or theory calculations. We expect t h a t future experiments will provide a way to resolve the present uncertainties, and shed light on the nuclear weak interaction problem. In this work, we calculate the P V asymmetry A^ in 7(i ^ np. Confronting the possibility of its experimentation, it is necessary to calculate A^ with u p d a t e d modern potentials, and compare it with old ones t h a t show significant difference. In our calculation, the strong interaction is accounted for by the Argonne v l 8 (Avl8) potential [5], and the P V potential given by Desplanques, Donoghue
and Holstein (DDH) [6] is employed for the weak interaction. In the next section, we briefly describe the basic formalisms. Numerical results are shown and discussions follow.
2 Result and discussion We consider the photon energies u p to 10 MeV above threshold, which leaves only a few loworbital states in the final state sufficient for the result being without significant error. Oppositeparity states are admixed in the wave function by the P V mesonexchange potential. T h e D D H P V potential [6], adopted in our work, includes TT, p and cjexchanges with the vertices specified by isospin transfer AI = 0 , 1 , 2. At the considered energies. Ml and El modes dominate the electromagnetic transition, and therefore we use the Ml and the El operators given as
2mN
(1) (2)
OEI
where e^ is the charge of the nucleon, //^ the magnetic moment, 1^ the angular momentum, cr^ the spin and x^ the position. Given the wave functions and the transition operators, the asymmetry A^ can be calculated from the definition ^
' ^ " ' cr+ +cr_
(3)
C.H. Hyun, C.P. Liu, B. Desplanques: Parity violating asymmetry inj\d^n\p8it
180 U.VU3
^"^ T
'
1
T
T
,
T
r
.
^
T
T
'
low energy
' T ^1
1
_.^.^—^"^
0 J
1 .0.005
/
—
1
^
U.(MHJ2

^.1
<:.
1s L I
O.OOW I  !
Cj
—  c^ C.
^
0.0005
DDIJ 0.00061 1
2
1
J
1
1
4
.
1
1
1
1
6 W
I
.
J
7 S iMtV)
1
1
y
L
_i
1
2
10
.1
4
5
I
I
A
»
J
7
I
i_
K
y
tn
l]
12
Fig. 2. Suppressed coefficients
Fig. 1. Dominant coefficients
where cr+ (cr_) denotes the total cross section for right(left) handed circular photons. To first order in the weak coupling constants, A^ can be written as A^ = cihl + C2/i^ + cshl + c^hl + c^hl + c^/ii.
(4)
Numerical results for the coefficients Q ' S are shown in Fig. 1 and 2. T h e magnitude of Q ' S for heavymeson isoscalar and isotensor vertices (c2, C4, C5) are larger t h a n those of isovector ones (ci, C3, CQ) by an order or more. Since the weak coupling constants calculated from a quark model [6] or t h e Skyrme model [7,8] give a similar magnitude, A^ will be dominated by isoscalar and isotensor P V vertices. In Fig. 3, we show A^ calculated with t h e D D H best values for the weak coupling constants. At threshold, A^ ~ 2.5 X 10~^, and it decreases to the order of 10~^ for uj^ > 3 MeV. T h e small magnitude of Aj can be understood from the strong cancellation of C4 with C2 and C5, and similar values of /z^(=  1 1 . 4 x 10"'^) with hl{=  9 . 5 x 10"'^). T h e strong suppression of A^ confirms the result of a recent schematic calculation [9], but does not support t h e earlier significant enhancement from the contribution of h^ [10]. It is shown in [11] t h a t the enhancement in [10] is due to t h e omission of t h e contribution from ^Pi ^ ^Si — ^Di transition, whose inclusion gives suppressed A^ values. T h e asymmetry A^ was recently calculated with A v l 8 and CDBonn [12] potentials by Schiavilla et al. [13]. For A v l 8 , they obtain a result similar to ours, but the CDBonn potential gives a larger result t h a n A v l 8 by a factor of about 2. A^ at threshold with rspace Bonn [14] and BonnA, B [15] is calculated, and, with t h e D D H best values, the results t u r n out to be similar with t h a t of CDBonn [16]. T h e enhancement for the Bonn potentials is due t o strong attraction at short distance in t h e ^Pi channel, which leads to a bound state in ^Pi state [17]. Therefore, this artifact of the ^Pi Bonn potential should be treated properly, e.g. by introducing a cutoff, to obtain a sound result.
0,05 h
Ci) ( M u V )
Fig. 3. A^ with DDH best values
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
C.S. Wood et al.: Science 275, 1759 (1997) A.R. Berdoz et al.: Phys. Rev. Lett. 87, 272301 (2001) F.E. Maas et a l : Phys. Rev. Lett. 93, 022002 (2004) W.M. Snow et al.: Nucl. lustrum. Meth. A 440, 729 (2000) R.B. Wiringa, V.G.J. Stoks, R. Schiavilla: Phys. Rev. C 51, 38 (1995) B. Desplanques, J.F. Donoghue, B.R. Holstein: Ann. Phys. (N.Y.) 124, 449 (1980) N. Kaiser, U.G. Meissner: Nucl. Phys. A 489, 671 (1988); A 499, 69 (1989) U.G. Meissner, H. Weigel: Phys. Lett. B 447, 1 (1999) LB. Khriplovich, R.V. Korkin: Nucl. Phys. A 690, 610 (2001) T. Oka: Phys. Rev. D 27, 523 (1983) C.P. Liu, C.H. Hyun, B. Desplanques: Phys. Rev. C 69, 065502 (2004) R. Machleidt: Phys. Rev. C 63, 024001 (2001) R. Schiavilla, J. Carlson, M. Paris: nuclth/0404082 R. Machleidt, K. Holinde, Ch. Elster: Phys. Rep. 149, 1 (1987) R. Machleidt: Adv. Nucl. Phys. 19, 189 (1989) C.H. Hyun el al.: in preparation J. Haidenbauer: private communication
V
Hadronic structure... and more
\/3 Neutrino beam
Eur Phys J A (2005) 24, s2, 183186 DOI: 10.1140/epjad/s2005040466
EPJ A direct electronic only
Precision physics at a neutrino factory A. Blondel Departement de Physique Nucleaire et corpusculaire, Faculte des Sciences, Universite de Geneve, Quai Ansermet 24, Geneve 4, CH1211 Switzerland Received: 1 November 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. Neutrino beams of unprecedented flux could be produced in a Neutrino Factory from muon decays. In the vicinity of the storage ring, short baseline experiments would perform a new class of precise tests of the theory and original deepinelasticscattering (DIS) studies. Thanks to the availability of high energy Ve and Ve , the long baseline experiments will be capable of very precise measurements of neutrino oscillations, including ability to solve parameter ambiguities and study of leptonic CP violation, for any value of the mixing angle 0\z above a fraction of a degree. Finally, the Neutrino Factory is the first step towards muon colliders. PACS. 1 3.88.+e Polarization in interactions and scattering  1 2.15.Ff Quark and lepton masses and mixing  1 4.60.Pq Neutrino mass and mixing  2 9.25.t Particle sources and targets  2 9.27.a Beams in particle accelerators  2 9.20.C Cyclic accelerators and storage rings
1 Introduction Neutrinos have historically played an essential role in particle physics, with t h e discovery of Neutral Currents (NC), the first observations of open charm, precautious scaling and its violations and t h e early description of t h e struct u r e of t h e nucleon. More recently L E P established t h a t there are three species of light neutrinos, thus probably only three families of fermions. Finally, neutrinos have recently be in t h e limelight with t h e final demonstration t h a t neutrinos have mass and mix. One of t h e main goals of t h e upcoming years will be t h e observation of leptonic C P violation, which is one of t h e leading explanations for the m a t t e r  a n t i m a t t e r asymmetry of t h e Universe. T h e importance of this search will justify substantial investments for t h e future. One of t h e leading ideas is t h a t of a neutrino factory, in which neutrinos are produced in a controlled way by means of a stored muon beam. This presentation and article summarize t h e work of several hundred of members of t h e C E R N and ECFA studies of a Neutrino Factory Complex, which is contained in the very complete [1]. T h e Neutrino Factory has been proposed in 1998 [2] and studied extensively thanks in particular to t h e pioneering work of t h e Neutrino Factory and Muon Collider collaboration [3]. As we will see in t h e following sections, such a machine is very polyvalent, and offers opportunities in short baseline physics, neutrino oscillations and is also t h e first step towards muon colliders.
2 The neutrino factory T h e basic layout of a Neutrino Factory is shown in Fig. 1. T h e principle [4] is to produce t h e largest possible inten
w^ j*/\^
A pttHsiblt: layout of ii
^[ M m . li..r\
iM^
1.2 1 0 '  ' M  » = 1 . 2
. r
/
\
lO'jivr
i
Fig. 1. Possible layout of a neutrino factory sity of low energy muons, accelerate t h e m to an energy of 20 to 50 GeV and store t h e m in a decay ring. A short (13 ns) high power (several M W ) proton beam of energy in excess of a few GeV hits a renewable target (liquid mercury is t h e baseline design). T h e secondary pions are captured by a magnetic device (tapered solenoid or magnetic horn) and fed into a solenoidal magnetic channel where they decay into muons. T h e muons are then subject to phase rotation and ionization cooling after which t h e energy spread and transverse emittance are considerably reduced. T h e b e a m is by then small enough t o be accelerated by means of recirculating linac or Ffa*g's, and stored in a decay ring equipped with long straight sections. Typically 10^^ muons are injected per second, producing at t h e end of t h e straight sections a very intense, bunched, beam of neutrinos from t h e decay of muons, (i^ ^ e+z/g i^i^ or /i~ ^ e~i>e Ufj^ , t h e sign being determined by t h e polarity of a few magnetic elements in t h e system or by timing.
A. Blondel: Precision physics at a neutrino factory
184
10
20 30 E^ (GeV)
10
40
20 30 E^ (GeV)
40
50
Fig. 2. Event rates at the near detector station of a neutrino factory. Note the vertical scale
x10 V events from i decay
5000 +
4500
Ll
R=1 Om radius, L=732 Km, 31.4 KT 3.10^ 50 GeV 1^^
Fig. 4. Expected achievable precision on polarized structure functions gi and g^ on proton {lower part) and deuteron (upper part) and for neutrino beam (left part) and antineutrino beam (right part) at the near detector of a neutrino factory
4000 3500 CCv^,P^,=+1
3000
CCv.,P
=1
/
\
2500 2000 1500 1000 500 0
10 15 20 25 30 35 40 45 50 V energy (GeV)
Fig. 3 . Event rate at a far detector station of a neutrino factory for positive muon decay. The effect of muon beam polarization is shown. In the racetrack or triangle geometry for the storage ring, the polarization can be preserved but averages out for each muon fill T h e resulting event rates are shown in Fig. 2 for t h e near detector station and in Fig. 3 for t h e far detectors. Originating from a stored and monitored beam, t h e flux of neutrinos should be known t o a fraction of a percent [5].
3 Short baseline physics In t h e near detector station, a kilogram of material placed on t h e b e a m axis would see typically 100 millions of interactions per year, more t h a n in t h e old 1200 t o n CDHS detector, b u t still only one event every few accelerator pulses. This, being obtained with a well defined flux of neutrinos (which are polarized by nature), opens a new realm of experimentation since t h e target material can be varied ad infinitum and t h e final state products can bemeasured in detail. T h e physics potential of high intensity
short baseline physics has been discussed quantitatively in [6], although no experimental set u p has been simulated so far. T h e range of physics accessible t o t h e near detector station is quite large. First, definitive measurements of unpolarized structure functions measurements and their flavour composition, in particular t h e strange sea, will be possible for t h e whole accessible kinematic range. This is, in particular, t h a n k s t o t h e assumed capability of t h e detectors t o t a g charm production. These ingredients should lead t o an improved determination of t h e strong coupling constant from either global structure function fits or from the GLS sum rule. Then, t h e use of small targets being possible, using polarized hydrogen or deuterium targets should allow a detailed decomposition of t h e spin structure functions, as shown in Fig. 4. Again, t h e use of charm t a g should be determinant in t h e study of t h e polarized structure function of t h e strange quarks. Maybe one of t h e most interesting topics shold be t h e possibility t o use a variety of nuclear targets and m a p systematically t h e nuclear effects in structure functions. T h e study of final states in neutrino scattering offers new possibilities. For instance t h e study of final state A and Ac polarization b o t h in charged current and neutral current processes has been given as an example, allowing extraction of t h e newly introduced polarized fragmentation functions. T h e high statistics available should allow precision measurements of s t a n d a r d model processes on electrons. This has a twofold interest. First, t h e charged current inverse muon decay process, z/^ e~ ^ /i~z/e , although unfortunately applicable only t o t h e muon neutrino component of t h e beam from /x+ decay, should allow a very precise flux normalization (provided an adequate target and detector can be built), in a way similar t o t h a t provided by B h a b h a scattering in e"^e~experiments. Secondly, t h e electron final state, which is very rich due t o t h e presence
A. Blondel: Precision physics at a neutrino factory
185
Table 1. Event rates in a 50 kton magnetized iron detector for one year running at a neutrino factory Baseline
730 km 3500 km
CCv^
3.5 10^ 1.2 10^
CC Ve
5.9 10^ 2.4 10^
Golden signal sin^ (9i3 = 0.01 1.1 10^ 1.0 10^
"
10
15 (b)
20
25
30
35
E^ ,W} [GeV]
Fig. 5. Expected achievable precision on the weak mixing angle sin^ ^w at small Q^ from neutrino scattering off electrons. The precision is shown as a function of the cut on the final state electron energy. It is clear that the /x^ exposure is more interesting for the mixing angle measurement, on the other hand the fi~ exposure is more sensitive to the interference between the NC and CC processes of the electron neutrino component in the beam, is sensitive to sin^ O^at small Q^ with a precision of the order of 2 X 10~^, similar to t h a t available at the Z pole, as shown in Fig. 5. Finally, the high statistics allied with the improved knowledge of charm production should allow a complete revision of the measurement of the hadronic neutral current processes ( N C / C C ratio) which have recently [7] exhibited a discrepancy with model expectations at the level of three s t a n d a r d deviations.
Fig. 6. Comparative sensitivity of various future measurements of ^13
asymmetry between the neutrino and antineutrino oscillations. In addition this channel involving electron neutrinos is sensitive to m a t t e r effects and should allow a determination of the sign of the mass difference Am^^ , which is presently unknown. T h e simulataneous presence of z^e and P^ in the beam has for consequence t h a t the detector has to be magnetic to separate the CC neutrino interactions generated by the P^ contained in the beam from those generated by 4 Neutrino oscillations zy^ originating from z/g ^ z/^ oscillations. A large 50 kton magnetic detector has been suggested, in extrapolation A large part of the present excitement for Neutrino Fac from the well known CDHS or MINOS experiments. T h e tories is, understandably, the long baseline oscillation rates are astounding, as shown in Table 1, many times physics. This has been extensively reviewed in the littera higher t h a n in the case of more conventional neutrino t u r e [5],[8] and only the main results are summarized here. beams. For this process the backgrounds are very small. Neutrino oscillations are now well established and T h e other very strong point of the Neutrino Factory demonstrate t h a t neutrinos have mass. T h e most striking is the capability to study the z/g —> v^ oscillation. This result of recent measurements are contained in our knowl channel is particularly valuable since it allows to lift the edge of the neutrino mixing matrix. For three flavours of unavoidable parameter ambiguities. T h e sensitivity or preneutrinos there are, naturally, three mixing angles, O12 , cision of the Neutrino Factory to the angle ^13 and to the ^13 5 ^23 , and two mass diflFerences / \ m f 2 , ^ ^ 2 3 ? ^^^^ P^^y C P violating phase 5 are shown in Figs. 6 and 7, and are a role in the oscillation process. T h e typical oscillation clearly superior to any other device imagined so far. length is 500 k m / G e V for the 'atmospheric oscillation' driven by Am^'^ , and 18000 k m / G e V for the 'solar oscillation' driven by / \ m f 2 In addition one expects the presence of a phase, yielding perhaps observable leptonic C P 5 Muon collider violation. T h e most interesting channel (socalled 'golden channel') to be studied is the v^ ^ i^/^ (and i/e ^ ^^L) Finally, it is worth keeping in mind t h a t the Neutrino oscillation which is suppressed at 'atmospheric' distances Factory is the first step towards muon colliders. As shown by the small value of the so far unknown ^13 . This sup in [9], the relevant characteristics of muons are t h a t , compression makes this channel particularly interesting since pared to electrons, i) they have a much better defined enit makes it possibly sensitive to the interference between ergy, since they hardly undergo synchrotron radiation or the solar and atmospheric oscillations, and thereby to the beamstrahlung, ii) their coupling to the Higgs bosons is resulting C P violation, which would manifest itself by an multiplied by the ratio (r72^/?Tie)^, thus allowing schannel
A. Blondel: Precision physics at a neutrino factory
186 fi
>
4
6 Conclusions
2 , Nufficl
SPLSB SPL+BflEa . SB+BB. lMion
We acknowledge the support of the European Community  Research Infrastructure Activity under the F P 6 "Structuring the European Research Area" programme (CARE, contract number RII3CT2003506395)
Best LMA aftar SNO Salt
O.S
References
1. ECFA/CERN Studies of an European Neutrino Factory Complex, A. Blondel (ed.) et a l : CERN2004002  ECFAJ Pa re sons ill VI ty 04230, h t t p : / / p r e p r i n t s . cern. ch/cerrLrep/2004/ _L 2004002/2004002.html 10 10 10 2. S. Geer: Phys. Rev. D 57, 6989 (1998) sin^e 13 3. http://www.cap.bnl.gov/inuinu Fig. 7. Sensitivity of various future neutrino options to the 4. P. Gruber et al.: The Study of a European Neutrino Factory Complex, in [1] p. 7; Feasibility Study on a Neutrino Source CPviolating phase S Based on a Muon Storage Ring, D. Finley, N. Holtkamp, eds.: (2000), Ex: m^ ^ 400 6eV/c^, m^ = 115 (&eV/c^ m^usv = 1 TcV/c^^. http://www.fnal.gov/proj ects/muon_collider/ 6E/E = 3 10^^ one week of running r e p o r t s . h t m l ; Feasibility StudyII of a MuonBased X^ 150 Neutrino Source, S. Ozaki, R. Palmer, M. Zisman, rGri:ii" 10 J. Gallardo, eds.: BNL52623, June 2001, available at I 120 h t t p : //www. cap.bnl. gov/miimu/study!i/FS2report.html; M.M. Alsharoa et al.: Phys. Rev. ST Accel. Beams 6, Ofher decay imcdes X ton^ 8 081001 (2003); Neutrino Factory and Beta Beam Experiments and Developments, (S. Geer, M. Zisman, eds.): i^ay be suff kJenfly copi 3. depiending on tian[l, Report of the Neutrino Factory and Beta Beam Working Group, APS MultiDivisional Study of the Physics of 6 CO Neutrinos, July 2004 5. M. Campanelli et al.: Oscillation Physics with a Neutrino 3a Factory, arXiv:hepph/0210192 in [1] p. 138 6. M.L. Mangano et al.: Physics at the frontend of a Neutrino BocldrotFNi level i , . ^ Factory: a quantitative appraisal, arXiv:hepph/0105155, r""r 396 3^>6 in [1] p. 187 7. K. McFarland: these proceedings; G.P. Zeller et al.: CCFR coll., Phys. Rev. Lett. 88, 091802 (2002); hepex/0110059 Fig. 8. Study of the supersymmetric H,A system at a muon 8. A. De Riijula, M.B. Gavela, P. Hernandez: Nucl. Phys. B colhder 547, 21 (1999); B. Autin, A. Blondel, J. Ellis (eds.): CERN yellow report CERN 9902, ECFA 99197; C M . Ankenbrandt et al.: Phys. Rev. ST Accel. Beams 2, 081001 (1999); production with a useful rate. These remarkable properA. Blondel et al.: Nucl. Instr. Meth. Phys. Res. A 451, ties make muon colliders superb tools for the study of 102 (2000); C. Albright et al.: FERMILABFN692, hepHiggs resonances, especially if, as predicted in supersymex/0008064; D. Harris et al.: Snowmass 2001 Summary, metry, there exist a pair H, A of opposite C P q u a n t u m hepph/0111030; A. Cervera et al.: Nucl. Phys. B 579, 17 numbers which are nearly degenerate in mass, as evi(2000), Erratumibid. B 593, 731732 2001; M. Koike, J. denced in Fig. 8. T h e study of this system is extremely Sato: Phys. Rev. D 62, 073006 (2000) difficult with any other machine and a unique investiga 9. See for instance S. Kraml et al: Physics opportunities at tion of the possible C P violation in the Higgs system would /j,^/j,~ Higgs factories, in [1], p. 337 become possible.
Eur Phys J A (2005) 24, s2, 187187 DOI: 10.1140/epjad/s2005040475
EPJ A direct electronic only
The MINERz/A experiment at FNAL Kevin S. McFarland^ University of Rochester, Rochester, NY 14610, USA Received: 15 December 2004 / Published Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. I present detector and physics capabihties of the MINERi/A experiment. PACS. 13.15.+g Neutrino interactions  25.30.Pt Neutrino scattering from nuclei
MINERi/A is a dedicated neutrino crosssection experiment planned for the NuMI beamline at Fermilab [1]. T h e detector (Fig. 1) consists of a lowmass active scintillator target surrounded by calorimetric detectors and upstream heavy nuclear targets. T h e lowmass target allows for separation of final state particles and therefore the identification of exclusive final states. T h e surrounding calorimeters ensure complete energy collection in the events, except for final state muons, which may be measured in the MINOS experiment's near detector located immediately downstream of MINERz/A. T h e physics goals of the experiment include measurements of the ^dependence of quasielastic {vn ^ li~p) scattering, measurement of the axial form factor of the nucleon at high Q^ (shown in Fig. 2), tests of quarkhadron duality in the axial current and measurements of coherent singlepion prodution in the Coulomb field of a target nucleus (Fig. 3). T h e physics of neutrino crosssections is an exciting subject in its own right and explores physics in the axial current similar to t h a t being probed at high precision in the vector current at high intensity electron scattering machines. These measurements are also important for fut u r e neutrino oscillation experiments planned with beams of energies 1a few GeV, where neutrino crosssections are difficult to predict theoretically and are poorly measured [2]. Results from the MINERz/A experiment will significantly reduce errors from unknown neutrino crosssections in the MINOS, T 2 K and NOz^A experiments.
Outer Detect01 (OD)
Vi;to I
I
Fig. 1. A schematic side view of the MINERz/A detector. Neutrinos enter from the right
J 00
>^ > BNL a i . [>a. B * k e r i?X ffi.
0 10. 0 00
Fig. 2. Estimation of FA from a sample of Monte Carlo neutrino quasielastic events recorded in the MINERi/A active carbon target, assuming a dipole form with MA = 1.014 GeV/c^. Also shown is FA from bubble chamber experiments
References 1. D. Drakoulakos et al. [Minerva Collaboration]: "Proposal to perform a highstatistics neutrino scattering experiment using a finegrained detector in the NuMI beam," arXiv:hepex/0405002 2. D.A. Harris et al. [MINERvA Collaboration]: "Neutrino scattering uncertainties and their role in long baseline oscillation experiments," arXiv:hepex/0410005 presenting on behalf of the MINERz/A Collaboration
.J/
Fig. 3. Coherent crosssections for 5 GeV neutrinos vs. atomic number. The solid curve and circles are two different models; crosses show expected MINERz/A measurements. The shaded band indicates the A region of previous experiments
VI
Concluding talks
Eur Phys J A (2005) 24, s2, 191195 DOI: 10.1140/epjad/s2005040484
EPJ A direct electronic only
Frontiers of polarized electron scattering experiments Krishna S. Kumar^ Department of Physics, University of Massachusetts, Amherst, MA 01003, USA Received: 15 December 2004 / Pubhshed Onhne: 8 February 2005 © Societa Itahana di Fisica / SpringerVerlag 2005 Abstract. Parityviolating electron scattering has developed into a precise and sensitive tool to probe the structure of weak neutral current interactions at Q^
1 Introduction
to the fourmomentum transfer Q^:
T h e past couple of decades have seen the emergence of parityviolating electron scattering (PVES) experiments to the forefront of probing electroweak and strong interactions at low energy. Major experimental innovations have been made [1], such as the development of highly polarized electron beams, dense cryogenic targets, radiationhard integrating detectors t h a t effectively count at GHz rates, counting detectors t h a t can count at MHz rates, precision beam monitoring and precision measurements of b e a m polarization with fractional accuracy approaching 1%. W h e n the accuracy of P V E S measurements is combined with new accelerator capabilities, the investigation of novel aspects of weak neutral current (WNC) interactions becomes feasible. Further, the high rate capabilities and the ability to measure small asymmetries opens the window into measurements of beamnormal asymmetries, where the incident b e a m polarization is transverse to the scattering plane. In the following, we introduce the observables of interest and discuss possible new measurements.
1.1 Parityviolating asymmetries
Apv
In P V E S , one measures the fractional difference Apy of the scattered flux from an unpolarized target for incident longitudinally polarized electrons whose spins are aligned along or against the momentum. For Q^
^PV
CTR
 CFL
crR\(TL
\A2 \A^
4x/2'ira
(1)
For typical Q^ values of interest (between 0.01 and 1 GeV^), Apy ranges from 0.1 to 100 parts per million (ppm). T h e W N C amplitude is typically dominated by the piece of the low energy W N C interaction t h a t arises from the product of the axialvector electron coupling and the vector coupling of the target particle. T h e discovery potential of P V E S measurements arises in several diflFerent kinematic situations. In some cases, the weak vector coupling of the target provides access to a new linear combination of the underlying quark vector currents, leading to novel insights of nuclear and nucleon structure [2]. Experiments H A P P E X , S A M P L E , A4 and GO (probing the strange structure of the nucleon) and the planned P R E X experiment (probing the neutron skin of ^^^Pb), fall under this category. In special circ*mstances, if the vector coupling of the target is very wellknown or if there are cancellations t h a t make the ratio of weak and electromagnetic amplitudes insensitive to hadron structure, one can probe new particle physics at TeV scales [3]. T h e SLAG E122 (Deep Inelastic Scattering), SLAG E158 (M0ller scattering) and Qweak (elastic electronproton scattering) experiments fall under this category. A third class of measurements access the small amplitudes t h a t arise from the product of the electron vector coupling and the target axialvector coupling. T h e A4, S A M P L E and GO experiments probe the axial nucleon current in this manner.
K.S. Kumar: Frontiers of polarized electron scattering experiments
192
2 Beamnormal asymmetries AT
Hx) =  [{C,^   C 2 , ) ^ ^ ^ y ^ ^ + corrections
If the incident beam is polarized transverse to the beam direction, one can construct an asymmetry: 27r d(cr'^  a^) d(j) a^ \ a^
(2)
which is proportional to Se (ke x k^), where Se is the incident electron spin vector and the incident and final electron 3vectors ke and k^ define the scattering plane. The primary contribution to AT comes from two photonexchange. The transverse asymmetry is suppressed by the electron boost and thus an order of magnitude estimate is AT ^ arrie/y/s. However, even for elastic electronproton scattering, the intermediate states between the two virtual photons have to be summed over the full virtuality kinematically available to each photon, resulting in enhancements that are quite sensitive to nucleon excited state structure [4,5,6]. Recently, analysis of data on proton elastic form factors have shown a disagreements between crosssection and asymmetry measurements. The real part of twophoton exchange amplitude is a good candidate to explain a signficant part of the discrepancy [7]. While AT measurements are sensitive to the imaginary part of the same amplitude, detailed studies might be possible since the amplitude constitutes the leading contribution. Thus, AT measurements have emerged as an important component of precision studies of nucleon structure.
3 PV deep inelastic scattering at high x The upgrade of Jefferson Laboratory (Jlab) to 11 GeV incident energy will allow precision measurements in parityviolating deep inelastic scattering (PV DIS). For the first time, high statistics can be accumulated at high x ~ 0.7, where x is the fraction of the nucleon momentum carried by the struck quark. PV DIS provides access to novel aspects of nucleon structure, complementing and enhancing precision electromagnetic DIS studies. Apy in DIS can be written as
(7)
where q{x) = u{x) \ d{x). For scattering off the proton a{x)
\u{x)\0.91d{x) u{x) + 0.25d{x)
(8)
3.1 Charge symmetry violation
As can be seen from 6, a{x) ;^ 1.15 for an isoscalar target, independent of x. This results from the assumption of charge symmetry, where the ^xquark distribution in the proton is the same as the dquark distribution in the neutron, with a similar assumption for the proton c/quark distribution: u^ = d^ and d^ = u^. If a(x) can be measured with high precision over a range of x values, one is thus quite sensitive to charge symmetry violation (CSV). If one defines CSV parameters: Su{x) = uP{x)  d^(x);
Sd{x) = dP{x)  ix^(x),
(9)
then the dependence on the parityviolating asymmetry for an isoscalar target is [8]: SAPV ipy
^Su — Sd u\ d
0.28
0.2SRcsv
(10)
While Rcsv is known to be less than 0.01 for x < 0.4 from neutrino DIS measurements [9], a bag model calculation suggests that Rcsv ^ 0.01 for x ^ 0.4 and rising to 0.02 for X ~ 0.6. At high x, knowledge of u\ d is limited. As X ^ 1, ii u \ d falls off more rapidly than 6u — Sd, then Rcsv might rise to 0.1 at x ^ 0.7, which would be observable with a 1% Apy measurement. Further, Rcsv is quite unconstrained at large x. There is the tantalizing possiblity that Rcsv in the moderate and high x region is a factor of 3 bigger than abovementioned values, which would be large enough to explain the 3cr discrepancy in the neutrinonucleon DIS measurement (NuTeV anomaly) [9] and would be a very important discovery. 3.2 d/u at high x
G, ipy
2A/2'ira
a{x) +
b{x)\, l + (ly)2'
a{x) = Sifi{x)Cuqi/
Sifi{x)qf,
b{x) = Sifi{x)C2iqi/
I^ifi{x)qf.
(3) (4)
(5) Here, Cii{C2i) are the weak vector (axialvector) weak charges for the ith. quark flavor, fi{x) are parton distribution functions and qi are the electromagnetic charges. The a(x) term arises from the product of the electron axialvector coupling and the quark vector coupling and is typically the dominant term. For an isoscalar target such as deuterium, the dependence on structure largely cancels out in the Apy ratio of the weak and electromagnetic amplitudes: a{x)
{Ciu — 7:Cid) \ corrections
(6)
As can be seen from 8 for PV DIS off the proton, a(x) is quite sensitive to the ratio d{x)/u{x). The value of d/u as J: ^ 1 is a very important parameter to pin down in DIS physics. It is required in order to properly constrain fits of parton distribution functions and impacts predictions for QCD processes at high energy colliders. More importantly, d/u at high x provides new information on important pieces of the nucleon wave function. There is empirical evidence that the minority quark in the nucleon is suppressed at high x, an intuitive notion in terms of a hyperfine interaction. While the SU(6) wave function would predict d/u ^ 0 . 5 , simple SU(6)breaking arguments would predict d/u ~ 0. However, a perturbative QCD calculation predicts d/u = 0.2 as J: ^ 1 [10]. Currently, the best estimates of d/u comes from ^H DIS structure function data, but uncertainties in the ^H
K.S. Kumar: Frontiers of polarized electron scattering experiments wavefunction limits the ability to discriminate between various predictions for d/u [11]. There are plans to measure d/u via the ratio of ^H and ^He structure functions and also via measurements of deuteron structure functions with tagged slow recoiling protons. A precise enough measurement oi a{x) for the proton at X ^ 0.7 would be able to distinguish between competing predictions for d/u as x ^ 1. T h e advantage of Apy measurements over other methods is t h a t there are no nuclear corrections since the P V DIS measurement can be made on a proton target.
193
A conventional magnetic spectrometer with the requirements specified above would be prohibitively expensive. However, it might be possible to employ a calorimeter with the ability to identify and count clusters at rates of 10 to 20 MHz, such as the one employed by the Mainz A4 experiment. T h e scattered electron energies of interest would range from 2 to 4 GeV. A large acceptance toroid could serve as a sweeping magnet, to remove low energy pions, M0ller electrons and to shield the calorimeter from lineofsight photons. T h e conceptual design for such an a p p a r a t u s is now under way, and might be potentially useful for other applications besides P V DIS.
3.3 Higher twist effects T h e topics discussed above can be accessed only if the physics of leading twist dominates the weak and electromagnetic amplitudes. However, it is wellknown from unpolarized structure function studies t h a t hadronic corrections from higher twist (HT) effects might add Q^ dependence to the asymmetry: ^PV
(a:,Q2)
ipy
(a;)(l + C ( x ) / Q 2 ) .
(11)
It is not possible to calculate C{x) from first principles, so ideally it must be extracted from data. P V DIS has the potential to probe for interesting H T effects, since the Apy weakelectromagnetic amplitude ratio would be sensitive to the amount of contributions from coherent quark combinations such as diquarks. For example, if spin0 diquarks dominate, then the H T effects are expected to be small, with a 1/Q^ dependence [12]. Additionally, novel H T effects arising from the interference of H T effects involving 7 exchange on one quark and Z exchange on a different quark might contribute [13]. Recently, an NNLO Q C D analysis has shown t h a t H T effects in the unpolarized structure functions are small for X < 0.6 but does not rule out 10 to 20% effects for X ^ 0.7 [14]. It would be interesting to search for H T effects in P V DIS. If they are large, they would point to dynamics specific to DIS processes involving 7 — Z interference. Determining whether H T effects are small or large is critical to ensure t h a t the physics of CSV and the d/u ratio can be extracted cleanly. The size of H T effects can already be constrained or discovered with a 6 GeV beam; there is a proposal under consideration at Jlab to pursue this measurement [15].
3.4 Experimental equipment for PV DIS at high x To comprehensively address the physics topics discussed above experimentally, a series of Apy measurements in the range of 1 to 2% accuracy are required for the x range from 0.3 to 0.7, with a lever a r m of a factor of 2 in Q^ while keeping W^^^ > 4 and Q^in > 1 W i t h the upgrade of Jlab, high luminosity with a beam energy of 11 GeV becomes possible for the first time. However, to achieve sufficient statistics at the highest possible Q^, a spectrometer with at least 50% acceptance in the azimuth is required.
3.5 Transverse asymmetries As a bonus, the use of a device such as the one described in the previous section makes possible a precision study of an entirely new process. Beamnormal asymmetries in the DIS region would become measureable with high precision. While leading twist contributions to transverse asymmetries might be of the order of ppm, H T effects might be enhanced by one to two orders of magnitude [16] due to large logarithms in the sum over intermediate states in the 2photon amplitude. Thus, H T effects could potentially be studied in detail.
4 Asymmetries at low Q^ and forward angle T h e unprecedented high luminosity available at Jlab as well as the upgrade of the energy provides new opportunities to measure asymmetries with sufficient precision at very forward angles, in a Q^ range between 0.05 and 1 GeV^, b o t h in elastic scattering as well as in highly inelastic scattering (W'^ > 4 GeV^).
4.1 Longitudinal asymmetries T h e E l 5 8 experiment has carried out an auxiliary measurement of the P V asymmetry in electronproton scattering at Q^ ~ 0.05 GeV^. T h e measurement is consistent with roughly Apy ^ —10~^Q'^ for the inelastic scattering component (mostly real and virtual photoproduction), which constituted about 30% of the flux. T h e remaining signal comes from elastic electronproton scattering, with an asymmetry t h a t is a factor of 5 smaller. It would be interesting to measure Apy in inelastic electronproton scattering at forward angles, mapping out the asymmetry variation as function of the target recoiling mass W. This would provide new information on parityviolating real and virtual photoproduction, which is presumably related to electromagnetic photoproduction via an isospin rotation. Thus, prevailing parametrizations on leptoproduction, primarily from the H E R A collider [17], would be tested in a new way.
194
K.S. Kumar: Frontiers of polarized electron scattering experiments
4.2 Transverse asymmetries
As described in Sect. 2, AT measurements have emerged as an important probe of nucleon structure. These asymmetries are enhanced as the incident beam energy is increased, since inelastic intermediate states with one quasireal photon makes a significant contribution. At very low Q^ (< 0.1 GeV^) and forward angle, AT can be predicted using the optical theorem [4,5]. At intermediate Q^ (0.1 < Q^ < 1 GeV^), the inelastic amplitudes from single and multiple pion production are expected to dominate [4]. As Q^ is increased. AT receives increasing contributions from offforward structure functions. At Q^ ^ 1 GeV^, AT can be calculated in a perturbative QCD framework where it is related to Generalized Parton Distributions [18]. There are currently plans for AT measurements at the GO and A4 experiments. The kinematics are wellsuited to test the regime of single pion production. On the other hand, it would interesting to measure AT at forward angle at very low Q^ (< Q.I GeV^) and at high Q^ _ i QeV^ in order to connect to the optical theorem at one extreme and parton distributions at the other extreme. 4.3 Experimental program
The longitudinal and transverse asymmetry measurements discussed above require the ability to measure very high flux rates (~ 100 MHz) and small asymmetries {^ 1 ppm). A spectrometer/detector package that provides this along with azimuthal coverage to measure AT with high efficiency would have to be similar in concept to the El58 design, where the entire primary and scattered beam were enclosed in a set of quadrupole doublets. This allows the separation of elastic and inelastic electronproton scattering events from background while providing acceptance in the full range if the azimuth. A compact radiationhard calorimeter would be placed downstream of the quadrupole spectrometer to integrate the scattered fiux. Alternatively, a more conventional forward angle spectrometer setup perhaps enhanced with septum magnets could be contemplated, although the solid angle would be significantly smaller. Indeed, these studies can be launched already with a 6 GeV beam and a proposal is under consideration at Jlab to pursue this measurement [19].
5 Electroweak physics As mentioned in the introductory paragraphs, with judicious choice of target and kinematics, it is possible to probe the structure of the WNC interaction itself, with little uncertainty from hadron structure. High precision measurements of such amplitudes at Q^
less sensitive to PV contact interaction scales by an order of magnitude. There are only a select few reactions that can probe the 10 TeV scale in fixed target WNC interactions [3]. Successful measurements have been reported in atomic parity violation [21], neutrino DIS (NuTeV) [22] and fixed target M0ller scattering (E158) [23]. One legacy of such precision measurements is that they are sensitive enough to observe electroweak radiative corrections (higherdiagrams that involve W and Z bosons in quantum loops). A convenient manifestation of this sensitivity is the "running", as a function of momentum transfer Q^, of the weakmixing angle sin2^w(Q')MS [24]. It is important to measure sin^ Ow with sufficient precision to observe the running in as many different reactions as possible in order to comprehensively probe for physics beyond the standard model at the TeV scale. The El58 measurement has established the running of sin^ Ow at the 7a level. The NuTeV measurement has a 3a descrepency with the standard model prediction, the origin of which is a subject of active theoretical and experimental debate. In the following, we describe plans and new ideas to make more precision measurements of sin^ Ow at low Q^.
5.1 Elastic electronproton scattering
At sufficiently forward angles and low Q^, the hadronic structure undertainty in the WNC elastic electronproton amplitude becomes small enough such that one can measure the underlying coherent 2u\d eq amplitude combination to high precision. This combination is proportional to 1 — 4 sin^ Ow so that a 4% measurement of Apy would achieve a precision of (5(sin^ Ow) = 0.0007 The experiment would use a 1 GeV 180//A electron beam incident on a 35 cm liquid hydrogen target. Scattered electrons would be focused by a toroidal field in the full range of the azimuth on to Cerencov detectors. The predicted asymmetry is 0.3 ppm. The experiment, named Qweak [25], has been approved for initial construction at Jlab and the first data collection is projected for 2008. It turns out to be quite complementary to compare Apy measurements from elastic electronproton scattering and M0ller scattering. Various models for new physics at the TeV scale can induce constructive or destructive interferences in one or both reactions. In particular, one might be able to distinguish between conventional and RParity breaking SUSY models, which is relevant to the viability of a SUSY dark matter candidate [26].
5.2 Deep inelastic scattering
If the PV DIS asymmetry can be measured to an accuracy of 1% at Q^ ^ 5 GeV^ and x ^ 0.35 for an isoscalar target such as deuterium, then the parameter a{x) can be measured with high precision, free from hadron structure uncertainties [27]. This measurement would be robust only if the program described in Sect. 3 is carried out. Indeed,
K.S. Kumar: Frontiers of polarized electron scattering experiments as can be seen from 6, a{x) is independent of x and simply a function of sin^ Ow under the assumption of charge symmetry and assuming t h a t H T effects are either directly measured or constrained. T h e measurement is interesting for several reasons. Firstly, it would test the W N C amplitude in the leptonquark sector, where there is currently a 3 a discrepancy in the NuTeV result. Secondly, combined with other measurements in elastic electronproton scattering, precise constraints would be possible on the lesser known axialvector quark couplings C2i This would, among other things, provide complementary constraints on various models with new heavy Z' bosons [3].
5.3 M0ller scattering T h e P V asymmetry in fixed target M0ller scattering is unique among W N C processes at Q^
6 Outlook Parityviolating electron scattering will continue to remain in the forefront of studies of low energy electroweak interactions in the future. Over the next few years, there are exciting possibilities for novel experimental programs as new facilities are developed, addressing physics topics from nucleon structure to physics beyond the s t a n d a r d model at the TeV scale.
195
Acknowledgements. It is a pleasure to thank the organizers for a stimulating workshop. Input and discussions with A. Afanasev, B. Holstein, T. Londergan, K. McFarland, W. Melnitchouk, B. Pasquini, M.J. RamseyMusolf, P.A. Souder, A.W. Thomas and M. Vanderhaeghen are gratefully acknowledged.
References 1. K.S. Kumar, P A . Souder: Prog. Part. Nucl. Phys. 45, S333S395 (2000), and references therein; D.H. Beck, B.R. Holstein: Int. J. Mod. Phys. E 10, 141 (2001) 2. M.J. Musolf et al: Phys. Kept. 239, 1 (1994) 3. M.J. RamseyMusolf: Phys. Rev. C 60, 015501 (1999) 4. B. Pasquini, M. Vanderhaeghen: Phys. Rev. C 70, 045206 (2004) 5. A. Afanasev, N.P. Merenkov: Phys. Rev. D 70, 073002 (2004) 6. F. Maas et al: nuclex/0410013 (2004) 7. P.A.M. Guichon, M. Vanderhaeghen: Phys. Rev. Lett. 91, 142303 (2003); P G . Blunden et al: Phys. Rev. Lett. 91, 142304 (2003) 8. P.A. Souder: to be published in the proceedings of mX2004, Marseilles, France, (2004) 9. J.T. Londergan, A.W. Thomas: hepph/0407247 (2004), and references therein 10. G.R. Farrar, D.R. Jackson: Phys. Rev. Lett. 43, 246 (1979) 11. W. Melnitchouk et al: Phys. Rev. Lett. 84, 5455 (2000); W. Melnitchouk, A.W. Thomas: Phys. Lett. B 377, 11 (1996) 12. R.L. Jaffe: private communication 13. S. Brodsky: In Proceedings from Jlab/Temple University HiX2000 Workshop, Philadelphia, USA, (2000) 14. A.D. Martin et al: Eur. Phys. J. C 35, 325 (2004) 15. Jlab Proposal P05007, X. Zheng: contactperson 16. A. Afanasev: private communication 17. A.C. Caldwell: Acta Phys. Polon. B 33, 35993608 (2002) 18. M. Gorshteyn et al: Nucl. Phys. A 741, 234248 (2004) 19. Jlab Proposal P05005, P. Bosted: contactperson 20. K.S. Kumar et al: Mod. Phys. Lett. A 10, 2979 (1995) 21. S.C. Bennett, C.E. Wieman: Phys. Rev. Lett. 82, 24842487 (1999) 22. G.P Zeller et al: Phys. Rev. Lett. 88, 091802 (2002) 23. P L . Anthony et al: Phys. Rev. Lett. 92, 181602 (2004) 24. A. Czarnecki, W.J. Marciano: Phys. Rev. D 53, 1066 (1996); A. Czarnecki, W.J. Marciano: Int. J. Mod. Phys. A 15, 2365 (2000); J. Erler, M.J. RamseyMusolf: hepph/0409169; A. Ferrogha, G. Ossola, A. Sirlin: Eur. Phys. J. C 34, 165 (2004) 25. D.S. Armstrong et al: Qweak Collaboration, Published in AIP Conf. Proc. 698, 172175 (2004) 26. A. Kurylov, M.J. RamseyMusolf, S. Su: Phys. Rev. D 68, 035008 (2003) 27. D. Mack et al: in PreConceptual Design Report for The Science and Experimental Equipment for the 12 GeV Upgrade of CEBAF, (2004) unpublished 28. K.S. Kumar: in DPF/DPB Summer Study on New Directions in HighEnergy Physics, econf C960625, (1996) NEW168
Eur Phys J A (2005) 24, s2, 197204 DOI: 10.1140/epjad/s2005040493
EPJ A direct electronic only
Parityviolation with electrons: Theoretical perspectives M.J. RamseyMusolf Kellogg Radiation Laboratory, California Institute of Technology, Pasadena, CA 91125, USA, email: mjrmOcaltech.edu Received: 1 January 2005 / Published Online: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. I review recent progress and developments in parityviolating electron scattering as it bears on three topics: strange quarks and hadron structure, electroweak radiative corrections, and physics beyond the Standard Model. I also discuss related developments in parityconserving scattering with transversely polarized electrons as a probe of twophoton processes. I conclude with a perspective on the future of the field. PACS. ll.30.Er Parity symmetry  25.30.c Leptoninduced reactions  12.15.Mm Neutral currents
1 Introduction Parityviolating electron scattering (PVES) was once considered something of a specialized  almost exotic  subfield of nuclear physics. In the past decade, however, the field has made substantial advances and has become something of a mainstream area of research. This m a t u r a t i o n of the field has been nicely summarized in the two PAVI meetings: Mainz (2002) and Grenoble (2004). My own involvement has now spanned a decade and a half, so I feel somewhat justified in providing a theoretical perspective on the status of the field. A rather extensive review can be found in my contribution to the Mainz PAVI proceedings [1], and this one will be somewhat abbreviated. Here, I will t r y to highlight what I think we have learned since the early 1990's and what I think may be the important directions for the future. In reviewing the evolution of P V E S , one can identify three rough eras: (1) the 1970's and 1980's, which Wiem a n and Masterson have called "ancient history" [2]; (2) the 1990's and first half of this decade, the "modern era"; and (3) the future, lasting perhaps into the 2020's. T h e focus in the early era was on testing the neutral current structure of the Standard Model (SM). Pioneering experiments were carried out in b o t h atomic P V (see [2] for a review) as well as in the SLAG deep inelastic scattering (DIS) experiment [3,4] t h a t were followed by the Mainz ^Be quasielastic [5] and MITBates ^^G elastic [6] P V E S measurements. In the 1990's, the structure of the SM neutral currents had been well established, so the emphasis shifted to using the SM neutral current as a probe of nucleon and nuclear structure. Here, the basic idea was t h a t the weak neutral current depends on a different linear combination of the light quark currents t h a n enters the electromagnetic current. Thus, by making a judicious choice of targets and kinematics, one can use P V E S  in combination with ordinary electron scattering  to perform a fiavor decomposition of the nucleon's vector cur
rent response. More recently, there has been a resurgence of interest in exploiting P V E S to study the electroweak interaction itself. In this regard, the pioneering experiment has been the SLAG M0ller experiment [7] t h a t will be followed by the equally demanding J L a b QWeak experiment [8]. Looking down the road, one may see followups to these measurements, possibly including more precise versions of the original SLAG deep inelastic experiment at J L a b or the M0ller experiment at either J L a b or the linear collider. W i t h this context in mind, let me t r y to summarize the recent developments in three fields, focusing primarily on the "modern era": strange quarks and nucleon structure; radiative corrections, including b o t h electroweak and QED; and the search for new physics.
2 Strange quarks T h e heyday of P V E S has been dominated by strange quarks in the nucleon. Goming amidst the "spin crisis" of the early 1990's, wherein there were indications t h a t strange quarks carried a more substantial fraction of the nucleon's spin (and mass) t h a n one might naively think (for a review, see [9]), the question arose as to whether strange quarks might also play a substantial role in the nucleon's electromagnetic structure. Three developments  two theoretical and one experimental  catalyzed the use of P V E S to address this question. Theoretically, Kaplan and Manohar observed how the weak neutral current (WNG) of the SM  in conjunction with the electromagnetic (EM) current  provided the tool needed to carry out the fiavor decomposition of the nucleon's vector current structure [10]. Subsequently, JaflPe noted t h a t dispersion theory analyses of the nucleon's isoscalar form factors implied large couplings of the nucleon to the 0(1020) and t h a t this OZIviolation would imply sizeable strange quark vector current form factors as a consequence [11]. On the
198
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives
experimental side, McKeown showed t h a t by measuring the backward angle P V asymmetry under conditions t h a t were experimentally feasible, one would be sensitive to strange magnetic form factors of the magnitude implied by Jaffe's analysis [12]. (For other important, early work on this topic, see [13,14,15,16]). The field was shortly off and running, resulting in the S A M P L E program at MITBates [17,18,19,20,21,22], the H A P P E X [23,24] and GO [25] experiments at J L a b , and the A4 program at Mainz [26,27,28]. A related development t h a t I will not focus on here is the development of the J L a b experiment designed to study the neutron distribution of ^^^Pb (see [29] and references therein). T h e idea  for which Donnelly, Dubach, and Sick deserve the credit [30]  is again to exploit the complementarity of the W N C and E M charge operators: the E M charge operator is sensitive primarily to protons, while the W N C primarily sees neutrons. T h e details of each of these experiments has been discussed elsewhere in this meeting, so I will not comment on any further on the specifics. While the GO, H A P P E X , and A4 experiments are not yet completed, the results obtained to date have taught us t h a t the role of strange quarks in the nucleon vector current form factors is likely to be relatively less important t h a n in the case of the nucleon mass and spin. In particular, the benchmark Jaffe predictions appear to be too large compared to experiment. At least, it is not consistent to have b o t h sizable strange magnetism and strange electricity in the nucleon. T h e completion of the experimental program will definitively tell us whether there is room for either one to be relatively important. Theoretically, we have learned t h a t strange quark dynamics are more subtle t h a n the simplest pictures might suggest and t h a t  because the effects are not so large obtaining a quantitatively reliable description and realistic physical picture remains a challenging task. In terms of the simplest physical pictures t h a t one might use to guide intuition, we have learned t h a t neither vector meson dominance nor kaon cloud dominance is right. Vector meson dominance works well in describing the dynamics of pseudoscalar mesons; it has long been known, for example, t h a t the a priori unknown low energy constants (LECs) in the 0{p^) chiral Lagrangian for the octet of pseudoscalar mesons are welldescribed by vector meson dominance [32]. Although a similar ansatz allows one to obtain a reasonable fit to the nucleon isoscalar vector form factor in dispersion theory, Jaffe's generalization of it to the strange quark sector is now ruled out by experiment (for an update of the analysis in [11]). T h e other picture t h a t used to guide one's intuition is t h a t of a kaon cloud around the nucleon. Its field theoretic description involves fiuctuations of the nucleon into kaonhyperon intermediate states whose effects are computed using loop graphs in perturbation theory. Now, one might rightfully object t h a t given the large [0(1)] meson nucleon coupling constants {e.g., QA ~ 1.26), one has no basis for believing perturbation theory. Nevertheless, many theorists (myself included) proceeded to carry out such calculations anyway, using various models for the hadronic
vertex form factors (for a list of references, see [18]). T h e motivation behind this miniindustry was an old computation of the nucleon's isovector form factors by Bethe and DeHoffman using pion loops [33]. Despite the presence of the 0 ( 1 ) couplings, it worked, and so, based purely on this phenomenological success, people proceeded with the same approach for the strange form factors. W h a t most people did not realize, however, is t h a t in the late 1950's, Federbush, Goldberger, and Treiman (FGT) showed  using dispersion theory  t h a t the success of the Bethe and DeHoffman pion loop calculation was a big accident [34]. In fact, the pion loop calculation as naively performed does not even respect unitarity. W h e n the latter requirement is imposed, the pion loop calculation falls short of the nucleon's isovector magnetic moment by half. F G T showed how to compute the pion cloud contribution correctly in dispersion theory by using measured strong interaction TTN scattering amplitudes and TT form factor data, effectively summing the loop graphs to all orders in the strong couplings. Remarkably, they obtained stunning agreement with the measured nucleon form factors. In the 1990's, my collaborators and I applied the F G T dispersion theory approach to the strange quark form factors, focusing on the kaon cloud contribution t h a t undergirded so many model computations [35,36,37,38]. Using experimentally obtained KN scattering amplitudes and e+e~ data, we performed the allorders summation of the kaon loop graphs analogous to the one carried out by F G T . We found t h a t the resulting spectral functions for the strangeness vector form factors have a peak at invariant mass t ~ mn?,. In short, the kaon cloud contribution is dominated by the 0(1020) resonance  consistent with the earlier poledominance ansatz. This result has two implications. First, the plethora of kaon cloud model calculations are all wrong; only an allorders summation is credible. Second, the kaon cloud cannot be the whole story, because the large effect implied by the 0resonance is ruled out by experiment. In principle, this situation is not incompatible with dispersion theory, which implies contributions from a tower of intermediate states. In practice, however, we have reached the end of the road, because there simply is not enough highquality e^e~ and strong interaction scattering d a t a involving these other states^ . In the case of the isovector form factors, F G T got lucky because the pion cloud contribution  when computed correctly  saturated the experimental form factor result. There was no need to analyze other, higher mass intermediate states. T h e strange quark, in contrast, is more elusive. T h e P V E S experiments suggest t h a t contributions from other intermediate states cancel against the kaon cloud, but theoretically, we simply cannot compute these canceling effects. Parenthetically, my experience with this dispersion theory analysis leads me to view simple meson cloud model computations of other nucleon structure effects with a degree of skepti^ For a quark model treatment of part of this tower of states, see [39]
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives It is reasonable to ask, however, why one has to rely on dispersion theory in the "modern era". Indeed, the effects of pseudoscalar mesons in various aspects of nucleon struct u r e  such as the isovector form factors or polarizablities  have been treated with considerable success in chiral perturbation theory ( C h P T ) . In C h P T , any observable is given by the sum of a loop contribution and an LEG, or counterterm^. Even though one never carries out an allorders loop calculation as dispersion theory suggests one should, this approach is consistent with dispersion theory because of the presence of the LEG contribution and because for lowenergy processes, higherorder effects are suppressed by p/vl, where A is either the nucleon mass or chiral symmetrybreaking scale. T h e actual value of the LEG, however, cannot be predicted theoretically; it has to be taken from experiment. To the extent t h a t one has enough independent experiments to determine all the LEGs at a given order in p/A^ one can then make predictions for other observables. T h e problem for strange quarks is t h a t we don't have the required set of independent experiments. As Ito and I pointed out some time ago, the leadingorder LEG's contain an SU(3)singlet component t h a t cannot be taken from any existent experiments [40]. In order to determine these constants, one has to measure the very strange quark form factors t h a t one would like to predict  a situation of circular logic. Formally, an exception occurs in the case of the strange magnetic radius, (r^)M As shown by Meissner, Hemmert, and Steininger, the leadingorder contribution is entirely nonanalytic, so t h a t a countertermfree prediction can be obtained from a oneloop computation [41]. Unfortunately, the chiral expansion for SU(3) is slowlyconverging, since the relevant expansion parameter is rrik/A ~ 1/2. Thus, one might worry t h a t higherorder effects are not negligible. In fact, my collaborators and I showed t h a t  in the case of the strange magnetic radius the nexttoleading order (NLO) loop contribution cancels most of the LO loop effect, exposing one to a dependence on the NLO LEG, b'^ [42]. T h e resulting expression for {rl)M is ( r 2 ) ^ =  [ 0 . 0 4 + 0.3 6^] fm'
(1)
where the first t e r m on the RHS gives the loop contribution evaluated at a renormalization scale /i = 1 GeV and the second t e r m is the NLO counterterm contribution. Naturalness considerations suggest t h a t b^ should have magnitude of order unity, implying t h a t the second t e r m on the RHS of (1) dominates over the first. Since the precise value of b^ cannot be determined except by measuring the strange magnetic radius itself, one is back to the original problem. It seems there is no free lunch with strange quarks. T h e situation is not entirely bleak, as one can get some indications of the size of the LEG from either dispersion theory or lattice QGD, and a reasonable (though not rigorous) range of predictions can be ^ Usually, one takes for the loop contribution only the part of a loop result that is nonanalytic in external momenta or masses that cannot be written down in a Lagrangian.
199
obtained in this case. Indeed, the final S A M P L E result for the strange magnetic moment  which requires extrapolation to Q^ = 0 by using the magnetic radius  includes this range in its quoted error [18]. It goes without saying t h a t one would like to carry out reliable, microscopic calculations of the strangeness form factors in QGD, and the lattice is our only tool for doing so. Some time ago, the KentuckyAdelaide collaboration obtained a nonzero signal for the form factors in a quenched calculation, the results of which suggested a negative strangeness magnetic moment [43]. More recently, the authors of [44] performed an quenched computation and found no evidence for nonzero strangeness form factors. The two computations differed in the number of gauge configurations and Z2 noise vectors employed. To my knowledge, the two groups have not yet sorted out the reasons for the difference in their two results. As we also heard in this meeting, D. Leinweber and collaborators have exploited quenched lattice results and the assumption of charge symmetry to predict the strangeness magnetic moment. T h e details may be found in his talk, but the approach has successfully reproduced the measured octet baryon magnetic moments. T h e predicted strangeness magnetic moment is negative: jj^s = —0.051 =b 0.021 [45]. T h e conundrum for this prediction as well as for the previously reported lattice results is t h a t the P V E S experiments suggest a positive value for fis (see, e.g., [26]). If these indications are solidified with the final results from the experiments, more work will be needed to understand the sign of /j^s from first principles in QGD (for an observations regarding the possible role of C transformation properties, see [46]) . Finally, let me note t h a t hadron models may provide some insights into the strangeness form factors. At present, one model  the chiral quark soliton model gives the only prediction for a positive JJ^S t h a t I am aware of [47]. In light of all other theoretical work t h a t seems to suggest a negative sign, this prediction must be taken seriously. Since I am not an expert on this model, I cannot comment in detail on what it may mean for the QGD dynamics of strange quarks. It is suggestive, however, t h a t the topology of the QGD vacuum has a nontrivial impact on sea quark dynamics.
3 Radiative corrections A significant consideration in the theoretical interpretation of the P V asymmetries and their implications for strange quarks has been contributions from electroweak radiative corrections. In high energy processes, one can compute these corrections with a high degree of reliability. At the low energies relevant to the P V E S experiments, the situation is quite different due to the interplay of the strong interaction and higherorder electroweak effects. T h e electroweak Ward Identities protect the weak neutral vector current from strong interaction renormalization, but the the axial vector current can experience substantial effects. Since the axial vector response contributes to a generic P V asymmetry, one has to take into
200
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives
account the uncertainties associated with these axial vector radiative correction effects. Uncertainties also arise in the vector channel via box diagrams in which two electroweak gauge bosons are exchanged; no symmetry protects one from QCD effects in this case. Fortunately, however, the potentially largest uncertainties, which would appear in the Z — 7 box graphs, are fortuitously suppressed by 1—4sin^ Ow ^ 0.1 [48,76]. Consequently, one's primary concern involves the axial vector channel. As part of my P h . D . research  carried out in collaboration with Barry Holstein  I showed t h a t one class of axial vector corrections  the socalled "anapole moment" terms  can be b o t h surprisingly large and theoretically uncertain [49]. These corrections involve the exchange of a virtual 7 between the lepton and hadron target, so their effect in neutrinohadron scattering is suppressed by an additional power of a/Air compared to the effect in P V E S . Consequently, it is useful to distinguish between the axial vector form factor as measured in electron scattering, G\, from the corresponding form factor probed by neutrinos, G^. As Holstein and I pointed out in [49], the uncertainties associated with the anapole contributions to G\ are at least as large as the contribution expected from strange quarks (as extrapolated from polarized deep inelastic scattering) , so P V E S provides a rather poor probe of axial vector strangeness. Neutrino scattering is theoretically much cleaner, since the radiative correction uncertainties are insignificant. Subsequent to this work, Donnelly and I observed t h a t the radiative correction uncertainties in G\ could mimic the effect of strange magnetism on the elastic, ep P V asymmetry [50]. Consequently, without a better handle on G ^ , one faced an intrinsic, theoretical uncertainty in the value of G%^ t h a t could be extracted from P V E S experiments. More generally, this kind of delicate interplay between various theoretical inputs  including radiative corrections  and the extracted values of the strange quark form factors was laid out in [51]. Fortunately, Hadjimichael, Poulis, and Donnelly observed t h a t by measuring the P V asymmetry in quasieslastic (QE) eD scattering, one could obtain a different linear combination of G\^ and G\ t h a n enters elastic ep scattering, thereby facilit a t i n g an experimental separation of b o t h [52]. W i t h this goal in mind, the S A M P L E Collaboration measured b o t h asymmetries. As initially reported, the results implied a value for G\ consistent with zero and a value of G%j t h a t was somewhat positive [20]. T h e result for G\ was particularly surprising. Originally, Holstein and I predicted t h a t radiative corrections would reduce G\ by roughly 30%, but the S A M P L E results suggested a 100% reduction. This surprise set off a flurry of theoretical efforts to explain the large effect. My collaborators and I updated the earlier work with Holstein using heavy baryon chiral perturbation theory and found little room for a larger effect t h a n originally predicted [53]. Others explored the Q^dependence of the anapole form factor [54,55], quark model estimates of the anapole contribution [56], and nuclear P V effects in the eD reaction [57,58]. Despite this hard work, no one could explain the S A M P L E result. One
remaining suspect remained to be analyzed: the Z — 7 box contribution. In contrast to its effect in the vector channel, its contribution to the axial response is not 1 — 4 sin^ Ow suppressed, so it was thought t h a t a large effect might be present here. Subsequently, however, the S A M P L E collaboration reanalyzed the deuterium d a t a and found three previously underestimated corrections [19]. T h e largest involved neutral pion backgrounds. T h e resulting shifts in the P V Q E asymmetry was just enough to change the extracted value of G\, bringing it into agreement with the original Musolf and Holstein prediction. W h a t we have learned from this experience is t h a t nonperturbative Q C D effects on electroweak radiative corrections in the axial vector channel can be significant, and one must take t h e m into account somehow. Doing so has become important not only for the strange quark program, but also for interpretation of the GO measurement of the P V N ^ A asymmetry. Here, the same kind of anapole effects t h a t enter the elastic asymmetry can be as significant [59]. Consequently, the normalization of the axial vector transition form factor probed by P V E S , G^'^ at Q^ = 0, will contain a theoretical radiative correction uncertainty of order 1020%. As Shilin Zhu and I recently showed, a determination of these corrections could be achieved by combining an experimental value for G^'^((5^ = 0) with the offdiagonal GoldbergerTreiman relation ( O D G T R ) , as the latter predicts the value of G^{Q'^ = 0) in the absence of these corrections [60]: 2 QKNAF,
GiiO) = \h
rriN
(1^.)
(2)
where QKNA is the strong TTNA coupling constant t h a t is known with 5% accuracy from TIN scattering in the resonance region, F^^ is the piondecay constant, and ZA^^ is a chiral correction t h a t is of order a few percent. Since G^'^(O) differs from G'^(O) by the large electroweak radiative corrections, a measurement of the former with P V E S will primarily probe these radiative corrections. In addition, a new effect arises in the inelastic asymmet r y t h a t does not occur for elastic scattering. Specifically, the asymmetry no longer vanishes at Q^ = 0. This result is a consequence of Siegert's theorem [61], which implies t h a t elastic matrix elements of the El operator vanish (even in the presence of paritymixing in the initial a n d / o r final states), but t h a t they need not do so for inelastic reactions. T h e resulting, lowQ^ N ^ A asymmetry has the form
^L^^(Q^ = 0) =  2 ^ ^ + 
(3)
where C^ is the transition magnetic form factor, d^ parameterizes the N ^ A El amplitude, and the " + indicate higher order chiral corrections. This effect for the N ^ A asymmetry was first pointed out by in [62], where we argued t h a t measuring d^ could be important for two reasons. First, without knowing its value, any a t t e m p t to determine the Q^dependence of G^'^ with P V E S could be confused by this term. Second, d/^ is the neutral current analog of the El amplitudes for
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives P V , electromagnetic hyperon decays, such as S + ^ pj. T h e asymmetries associated with the latter are of order four times larger in magnitude t h a n one would expect based on simple symmetry considerations  a puzzle t h a t has largely eluded explanation. Studying the neutral current analog {i.e., d^) could, in principle, shed new light on this old problem. New efforts are underway to develop experiments t h a t would determine d^More recently, another aspect of radiative corrections those involving the exchange of two photons  has grabbed the attention of the P V E S community. While this effect is pure Q E D and conserves parity, one manifestation of it can be measured using similar methods to those used in the P V experiments. Specifically, by scattering transversely polarized electrons from a given target and measuring the transversespin asymmetry (or vector analyzing power). An, one probes the imaginary part of the twophoton exchange amplitude, Ad^^. T h e vector analyzing power (YAP) has been measured by the S A M P L E Collaboration at relatively low energy [65] and by the Mainz A4 Collaboration at higher energy [66]. T h e S A M P L E result differs substantially from an old, potential scattering prediction carried out by Mott [67], leading to considerable theoretical interest in this observable. Although the VAP has been discussed by several others in this meeting, I would like to comment on why I think it is an important topic to pursue experimentally and theoretically. By now it is wellknown t h a t a proper t r e a t m e n t of the real part of Ai^^ may help resolve the apparent differences in the Q^dependence of the proton electric form factor as determined polarization transfer experiments vs. Rosenbluth separation (for a discussion, see, e.g., [63,64] and references therein). Clearly, having an experimental test of theoretical calculations of Ai^^ is important from this standpoint, and the VAP provides such a test. It is also important for our understanding of electroweak radiative corrections. Indeed, the kinds of twoboson exchange box graphs mentioned above have to be computed theoretically in order to arrive at Standard Model predictions for various electroweak observables. In particular, the dominant theoretical uncertainty associated with the interpretation of neutron and nuclear /5decay involves such graphs, where one of the exchanged bosons is a W^ and the other is a 7. In order to extract a value for the C K M matrix element Vud from these measurements, one must compute the Wj box contribution. T h e theoretical machinery needed for this computation is the same as required for Ai^^, but there is no hope of ever measuring A4wf directly. Hence, the VAP may provide our only experimental test of the theoretical framework used in computing this important electroweak radiative correction. T h e /3decay correction applies to a very lowenergy process, for which the use of effective field theory ( E F T ) might be particularly suitable. Similarly, the recent SAMP L E measurement of the VAP was performed in the lowe n e r g y / E F T domain. W i t h this context in mind, my student L. Diaconescu and I recently computed the VAP for lowenergy, elastic ep scattering using an E F T approach. Our primary objective was to determine if one could use
201
E F T to resolve the discrepancy between the S A M P L E result and the old potential scattering calculation performed by Mott. If successful, we would have some confidence in applying the same framework to the electroweak box correction. T h e details of our calculation can be found in [68]. Let me comment, however, on a few aspects of the calculation. First, given the technical challenge in carrying out analytic computations, we decided to work initially with an E F T involving only electrons, photons, and nucleons. In the heavy baryon formalism, the nucleon is essentially a "static" source, and there is no large m o m e n t u m associated with this degree of freedom. In this E F T , the pion is integrated out, being treated as "heavy". While this assumption would certainly hold for the /^decay correction, it is admittedly questionable for the VAP at the SAMP L E kinematics. Nonetheless, we explored how much of the experimentally observed VAP could be accounted for by this simplest E F T . In this case, A^ has a power series expansion in p/M, where M is the nucleon mass and p is either the incident electron energy or electron mass. T h e Mott computation corresponds to the 0(p/M)^ contribution  the one t h a t survives for an infinitely heavy target. We showed t h a t , working to 0{p/M)'^, one can make a parameterfree prediction for the VAP. To this order, there are no unknown LEC's, and the VAP arises entirely from a oneloop effect. The first unknown constants appear at 0{p/M)^. To obtain a consistent computation to 0{P/MY, one must include b o t h the nucleon magnetic moment and charge radius at the jNN vertices in the loop, as well as "kinetic" terms in the nucleon propagator. To our surprise, we found t h a t the 0{p/M)'^ VAP agrees beautifully with the S A M P L E result. Given t h a t the m o m e n t u m transfer is of order 7717^, one would have expected t h a t inclusion of pions would be necessary to produce agreement. W h a t the result suggestions, however, is t h a t for this particular observable, pions do not appear to play a significant role at these kinematics. As a corollary, it seems reasonable t h a t using the same E F T (without dynamical pions) at the lowerenergies relevant for /3decay will produce a reliable result for t h a t process. Carrying out the latter computation is clearly a task for the future. On the other hand, taking the E F T expression for A^ to the higher energies relevant to the Mainz VAP measurement leads to significant disagreement  not a surprising outcome given t h a t the Mainz energies are well beyond the limit of this E F T . For this domain, one must likely include dynamical pions and the Z\resonance. I don't know how to do t h a t in a modelindependent way t h a t retains the systematic power counting of E F T , so some other approach may be necessary. B. Pasquini has reported on one such a t t e m p t using the MAID program [69]. As I understand the results, the MAID calculation comes closer to the Mainz results t h a n the E F T calculation, but fails to reproduce the lower energy S A M P L E result t h a t can be explained using E F T (for other model computations, see, e.g., [70,71]). It is clearly of interest to understand the reasons behind these differing theoretical results, and having
202
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives
more experimental d a t a in the low to mediumenergy domain would be quite helpful.
4 New physics A subject t h a t is close to my heart  but one t h a t I will spend less time on here  is the use of P V E S to search for physics beyond the Standard Model. Two experiments are currently on the books: the E l 5 8 M0ller experiment at SLAC [7] t h a t has finished its d a t a taking and the QWeak experiment t h a t will take place in the future at J L a b [8]. T h e beauty of these two experiments is t h a t they are highly precise, t h a t they are being carried out at essentially the same Q^, and t h a t they are theoretically "clean" probes of the Standard Model and its extensions. Indeed, these two experiments, when taken in t a n d e m with the results of the Cesium atomic P V experiment, provide a unique diagnostic of new physics. My collaborators and I have recently illustrated this point in several papers, focusing particularly intently on supersymmetric extensions of the Standard Model [72,73,74,75,76]. I refer the reader to those papers for details. Even in the absence of new physics effects, the two experiments will provide the most precise determinations of the running sin^ Ow at a scale below the Z^po\e (see, e.g., [77,72,78,79]). Looking to the future, the diagnostic power of the P V E S experiments could be amplified with additional measurements. A more precise version of the M0ller experiment is under consideration for J L a b in its post12 GeV upgrade phase. As I understand it, this experiment would be quite challenging, but several talented people are working on ways to meet this challenge. A second possibility would be to carry out a more precise version of the SLAC deep inelastic experiment. Recently, a proposal to do so was given high marks by the SLAC EPAC, but the funds to carry out this experiment do not seem to be available. There is also considerable interest in performing a lowerenergy version of the P V deep inelastic, or "DISParity" experiment at the upgraded J L a b . In this case, sorting out the theoretical implications of a precision measurement would be more involved t h a n for the SLAC kinematics, primarily because of potential highertwist (HT) contributions. T h e SLAC Q^ would be sufficiently large t h a t one could neglect these H T effects and perform a precision electroweak test. On the other hand, "twist pollution" will creep in at the J L a b kinematics, and we do not have enough information on H T contributions to make definitive statements about their size at these kinematics. It may be t h a t carrying out multiple measurements at different m o m e n t u m transfers would allow one to simultaneously constrain the degree of twist pollution and deviations from the Standard Model electroweak contribution. Given t h a t the figure of merit for the deep inelastic measurements is relatively large, such a program may be realistic. Theoretically, several questions pertaining to the deep inelastic asymmetry remain to be studied extensively. In the case of H T , for example, one does not know what Q C D predicts for the evolution of the corresponding structure
functions. Similarly, the degree to which isospin violation in the leading twist p a r t o n distribution functions would affect the asymmetry is an interesting question. As we heard in K. McFarland's talk, the currently favored explanation of the NuTeV anomaly [80] is just such isospin violation. To be a viable one, this violation would have to be large, so t h a t one might expect similarly large effects on the deep inelastic asymmetry. Finally, there is the question as to what combination of measurements would yield the most useful information on the electroweak sector of the Standard Model as well as on the aforementioned aspects of nucleon structure. Clearly, there is considerable room for future theoretical work on this topic.
5 Conclusions It seems to me t h a t the "modern era" of P V E S experiments will be coming to an end in the next few years. W i t h the conclusion of the S A M P L E , H A P P E X , GO, and A4 programs, the experimental questions about the size of the strange quark form factors and axial vector radiative corrections will have been settled; the ^^^Pb measurement will have provided new information about the distribution of neutrons in heavy nuclei; and the first generation of precision electroweak tests ( E l 5 8 and QWeak) will have produced results. I believe t h a t those who have been involved in this field will be able to look back with considerable satisfaction at the unique and varied physics t h a t this subfield of nuclear physics has illuminated. At the same time, there will be ample grist for the theoretical mill. If the present indications of a positive strangeness magnetic moment persist, it will be a challenge to provide a Q C D explanation for this result and to understand the dynamics behind it. Similarly, if GO and possible new experiments are able to cleanly separate the G ^ ' ^ and d^ contributions to the inelastic asymmetry, then new confrontations with hadron structure theory will be available. Finally, comparisons of first results from the LHC with information obtained from precision measurements  such as Cesium atomic P V , E158, QWeak, and possible second generation progeny  may help us determine which of the currently popular extensions of the Standard Model  if any  are most viable. In short, it appears t h a t we have hardly heard the last word from parity violation. Acknowledgements. I am indebted to my miany theoretical and experimental collaborators and colleagues who are too numerous to list and who, over the years, have helped me understand various aspects of this field with greater clarity and insight. This work was supported in part by U.S. Department of Energy DEFG0302ER41215 and by a National Science Foundation Grant PHYOO71856.
References 1. M.J. RamseyMusolf: arXiv:nuclth/0302049
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives 2. B.P. Masterson, C.E. Wieman: in Precision Tests of the Standard Model, P. Langacker ed. (World Scientific, Singapore, 1995), Ch. IV. D 3. C.Y. Prescott et al.: Phys. Lett. B 77, 347 (1978) 4. C.Y. Prescott et al.: Phys. Lett.B 84, 524 (1979) 5. W. Heil et al.: Nucl. Phys. B 327, 1 (1989) 6. P.A. Souder et al.: Phys. Rev. Lett. 65, 694 (1990) 7. P.L. Anthony et al. [SLAC E158 CoUaboration]: Phys. Rev. Lett. 92, 181602 (2004), [arXiv:hepex/0312035] 8. D.S. Armstrong et al. [Qweak Collaboration]: AIP Conf. Proc. 698, 172 (2004), [arXiv:hepex/0308049] 9. B.W. Filippone, X.D. Ji: Adv. Nucl. Phys. 26, 1 (2001), [arXiv:hepph/0101224] 10. D.B. Kaplan, A. Manohar: Nucl. Phys. B 310, 527 (1988) 11. R.L. Jaffe: Phys. Lett. B 229, 275 (1989) 12. R.D. Mckeown: Phys. Lett. B 219, 140 (1989) 13. G. Feinberg: Phys. Rev. D 12, 3575 (1975), [Erratumibid. D 13, 2164 (1976)] 14. J.D. Walecka: Nucl. Phys. A 285, 349 (1977) 15. D.H. Beck: Phys. Rev. D 39, (1989) 3248 16. J. Napolitano: Phys. Rev. C 43, 1473 (1991) 17. E.J. Beise, M.L. Pitt, D.T. Spayde: Prog. Part. Nucl. Phys. 54, 289 (2005) 18. D.T. Spayde et al. [SAMPLE Collaboration]: Phys. Lett. B 583, 79 (2004), [arXiv:nuclex/0312016] 19. T.M. Ito et al. [SAMPLE Collaboration]: Phys. Rev. Lett. 92, 102003 (2004), [arXiv:nuclex/0310001] 20. R. Hasty et al. [SAMPLE Collaboration]: Science 290, 2117 (2000), [arXiv:nuclex/0102001] 21. D.T. Spayde et al. [SAMPLE Collaboration]: Phys. Rev. Lett. 84, 1106 (2000), [arXiv:nuclex/9909010] 22. B. Mueller et al. [SAMPLE Collaboration]: Phys. Rev. Lett. 78, 3824 (1997), [arXiv:nuclex/9702004] 23. K.A. Aniol et al. [HAPPEX Collaboration]: Phys. Rev. C 69, 065501 (2004), [arXiv:nuclex/0402004] 24. K.A. Aniol et al. [HAPPEX CoUaboration] Phys. Lett. B 509, 211 (2001), [arXiv:nuclex/0006002] 25. G. Batigne [GO Collaboration]: Eur. Phys. J. A 19, SUPPL1206 (2004) 26. F.E. Maas et al.: arXiv:nuclex/0412030 27. F.E. Maas et al. [A4 Collaboration]: Phys. Rev. Lett. 93, 022002 (2004), [arXiv:nuclex/0401019] 28. S. Baunack [A4 Collaboration]: Eur. Phys. J. A 18, 159 (2003) 29. C.J. Horowitz, S.J. Pollock, P.A. Souder, R. Michaels: Phys. Rev. C 63, 025501 (2001), [arXiv:nuclth/9912038] 30. T.W. Donnelly, J. Dubach, L Sick: Nucl. Phys. A 503, 589 (1989) 31. H.W. Hammer, U.G. Meissner, D. Drechsel: Phys. Lett. B 385, 343 (1996), [arXiv:hepph/9604294] 32. G. Ecker, J. Gasser, A. Pich, E. de Rafael: Nucl. Phys. B 321, 311 (1989) 33. H. Bethe, F. deHoffman: Mesons and Fields (Row and Peterson, Evanston, 1955), Vol. II 34. P. Federbush, M.L. Goldberger, S.B. Treiman: Phys. Rev. 112, 642 (1958) 35. H.W. Hammer, M.J. RamseyMusolf: Phys. Rev. C 60, 045204 (1999), [Erratumibid. C 62, 049902 (2000)], [arXiv:hepph/9903367] 36. H.W. Hammer, M.J. RamseyMusolf: Phys. Rev. C 60, 045205 (1999), [Erratumibid. C 62, 049903 (2000)], [arXiv:hepph/9812261]
203
37. M.J. RamseyMusolf, H.W. Hammer: Phys. Rev. Lett. 80, 2539 (1998), [arXiv:hepph/9705409] 38. M.J. Musolf, H.W. Hammer, D. Drechsel: Phys. Rev. D 55, 2741 (1997), [Erratumibid. D 62, 079901 (2000)], [arXiv:hepph/9610402] 39. P. Geiger, N. Isgur: Phys. Rev. D 55, 299 (1997), [arXiv:hepph/9610445] 40. M.J. Musolf, H. Ito: Phys. Rev. C 55, 3066 (1997), [arXiv:nuclth/9607021] 41. T.R. Hemmert, U.G. Meissner, S. Steininger: Phys. Lett. B 437, 184 (1998), [arXiv:hepph/9806226] 42. H.W. Hammer, S.J. Puglia, M.J. RamseyMusolf, S.L. Zhu: Phys. Lett. B 562, 208 (2003), [arXiv:hepph/0206301] 43. S.J. Dong, K.F. Liu, A.G. Williams: Phys. Rev. D 58, 074504 (1998), [arXiv:hepph/9712483] 44. R. Lewis, W. Wilcox, R.M. Woloshyn: Phys. Rev. D 67, 013003 (2003), [arXiv:hepph/0210064] 45. D.B. Leinweber et al.: arXiv:heplat/0406002 46. X.D. Ji, J. Tang: Phys. Lett. B 362, 182 (1995), [arXiv:hepph/9507465] 47. A. Silva, H.C. Kim, K. Goeke: Eur. Phys. J. A 22, 481 (2004), [arXiv:hepph/0210189] 48. W.J. Marciano, A. Sirlin: Phys. Rev. D 29, 75 (1984), [Erratumibid. D 31, 213 (1985)] 49. M.J. Musolf, B.R. Holstein: Phys. Lett. B 242, 461 (1990) 50. M.J. Musolf, T.W. Donnelly: Nucl. Phys. A 546, 509 (1992), [Erratumibid. A 550, 564 (1992)] 51. M.J. Musolf, T.W. Donnelly, J. Dubach, S.J. Pollock, S. Kowalski, E.J. Beise: Phys. Rept. 239, 1 (1994) 52. E. Hadjmichael, G.I. Poulis, T.W. Donnelly: Phys. Rev. C 45, 2666 (1992) 53. S.L. Zhu, S.J. Puglia, B.R. Holstein, M.J. RamseyMusolf: Phys. Rev. D 62, 033008 (2000), [arXiv:hepph/0002252] 54. C M . Maekawa, U. van Kolck: Phys. Lett. B 478, 73 (2000), [arXiv:hepph/0006161] 55. C M . Maekawa, J.S. Veiga, U. van Kolck: Phys. Lett. B 488, 167 (2000), [arXiv:hepph/0006181] 56. D.O. Riska: Nucl. Phys. A 678, 79 (2000), [arXiv:hepph/0003132] 57. R. Schiavilla, J. Carlson, M. Paris: Phys. Rev. C 67, 032501 (2003), [arXiv:nuclth/0212038] 58. C P . Liu, G. Prezeau, M.J. RamseyMusolf: Phys. Rev. C 67, 035501 (2003), [arXiv:nuclth/0212041] 59. S.L. Zhu, C M . Maekawa, G. Sacco, B.R. Holstein, M.J. RamseyMusolf: Phys. Rev. D 65, 033001 (2002), [arXiv:hepph/0107076] 60. S.L. Zhu, M.J. RamseyMusolf: Phys. Rev. D 66, 076008 (2002), [arXiv:hepph/0207304] 61. A.J.F. Siegert: Phys. Rev. 52, 787 (1937) 62. S.L. Zhu, C M . Maekawa, B.R. Holstein, M.J. RamseyMusolf: Phys. Rev. Lett. 87, 201802 (2001), [arXiv:hepph/0106216] 63. J. Arrington: Phys. Rev. C 69, 022201 (2004), [arXiv:nuclex/0309011] 64. P.A.M. Guichon, M. Vanderhaeghen: Phys. Rev. Lett. 91, 142303 (2003), [arXiv:hepph/0306007] 65. S.P. Wells et al. [SAMPLE collaboration]: Phys. Rev.C 63, 064001 (2001), [arXiv:nuclex/0002010] 66. F.E. Maas et al.: arXiv:nuclex/0410013 67. N.F. Mott: Proc. R. Soc. London, Ser. A 135, 429 (1932) 68. L. Diaconescu, M.J. RamseyMusolf: Phys. Rev. C 70, 054003 (2004), [arXiv:nuclth/0405044]
204
M.J. RamseyMusolf: Parityviolation with electrons: Theoretical perspectives
69. B. Pasquini, M. Vanderhaeghen: Phys. Rev. C 70, 045206 (2004), [arXiv:hepph/0405303] 70. A.V. Afanasev, N.P. Merenkov: Phys. Lett. B 599, 48 (2004), [arXiv:hepph/0407167] 71. A.V. Afanasev, N.P. Merenkov: Phys. Rev. D 70, 073002 (2004), [arXiv:hepph/0406127] 72. J. Erler, A. Kurylov, M.J. RamseyMusolf: Phys. Rev. D 68, 016006 (2003), [arXiv:hepph/0302149] 73. A. Kurylov, M.J. RamseyMusolf, S. Su: Phys. Rev. D 68, 035008 (2003), [arXiv:hepph/0303026] 74. A. Kurylov, M.J. RamseyMusolf, S. Su: Phys. Lett. B 582, 222 (2004), [arXiv:hepph/0307270]
75. M.J. RamseyMusolf: Phys. Rev. D 62, 056009 (2000), [arXiv:hepph/0004062] 76. M.J. RamseyMusolf: Phys. Rev. C 60, 015501 (1999), [arXiv:hepph/9903264] [77,78,79] 77. A. Czarnecki, W.J. Marciano: Phys. Rev. D 53, 1066 (1996), [arXiv:hepph/9507420] 78. A. Ferrogha, G. Ossola, A. Sirlin: Eur. Phys. J. C 34, 165 (2004), [arXiv:hepph/0307200] 79. J. Erler, M.J. RamseyMusolf: arXiv:hepph/0409169 80. G.P. Zeller et al. [NuTeV Collaboration]: Phys. Rev. Lett. 88, 091802 (2002), [Erratumibid. 90, 239902 (2003)], [arXiv:hepex/0110059]
Eur Phys J A (2005) 24, s2, 205208 DOI: 10.1140/epjad/s200504050x
EPJ A direct electronic only
Workshop summary R.D. McKeown California Institute of Technology, Pasadena, CA 91125, USA Received: 15 October 2004 / Published Onhne: 8 February 2005 © Societa Italiana di Fisica / SpringerVerlag 2005 Abstract. This latest in a series of workshops on parityviolating electron scattering comes at a momentous time in the history of this subject. The first experiments to determine strange form factors of the nucleon have produced intriguing final results, and several powerful new experiments are now producing data. In addition, the precision of the technique has been improving and new experiments testing the electroweak theory have reported remarkably precise data. There has also been a great deal of progress on both the theory of strange form factors and interpretation of electroweak symmetry tests. PACS. 01.30.Cc Conference proceedings  25.30.Bf Elastic electron scattering gation, parity, time reversal, and other discrete symmetries
1 Introduction It has been 15 years since the first papers [1,2,3] proposing to study strange quark contributions to nucleon electroweak form factors gave new impetus to the field of parityviolating electron scattering. Indeed shortly after those papers in February of 1990, the first of these workshops [4] was held in Pasadena, California. T h e warm and sunny winter weather seemed to portend a bright future (we didn't know the workshop should have been called PAVI90!) and provided additional motivation for participants from colder climes in the US, Canada, and Germany to attend. T h a t workshop certainly reaffirmed the motivation for planning a program of parityviolating electron scattering experiments to explore the strangeness in the nucleon. But it also served to highlight many concerns on b o t h the theoretical and experimental sides, which have fortunately all been mitigated through diligent and clever efforts in the subsequent years. Since t h a t time, we have seen truly remarkable progress in this field with great strides made by theorists as well as experimentalists. This workshop was an excellent opportunity to take stock and chart a new course for the future. T h e organizers should be commended for providing a stimulating program and delightful environment, and I certainly hope t h a t their plans to continue this tradition are fruitful.
2 Strange form factors T h e first proposed experiment to study the strangeness in the nucleon using parityviolating electron scattering was the S A M P L E [5] experiment at M I T / B a t e s . As reported at this workshop by D. Spayde, the S A M P L E experiment
ll.30.Er Charge conju
has measured the strange magnetic form factor for the first time, with the result [6] G ^ ( Q 2 = 0.1(GeV/c)^) = 0.37 =b0.20 =b0.26 =b0.07. (1) Although this result is consistent with zero strangeness, it is not consistent with the prevailing theoretical view as presented in Fig. 1. Most theoretical predictions favor a substantially negative strange magnetic moment, whereas the S A M P L E result indicates /i^ > 0. Figure 1 also indicates the level of interest in the quantity fig among theorists in this field. There have been a great number of studies, and this has made for very pleasant and productive interactions between theorists and experimentalists. W h a t is difficult to tell from Fig. 1 is t h a t we have actually learned something from this discourse. Early models with simplistic t r e a t m e n t s of vector meson dominance or lowest order K — A loops are known to be inadequate to reproduce the the experimental result, have wellstudied theoretical shortcomings, and so are no longer used in modern calculations. Unfortunately, as reported in this conference by Kubis and RamseyMusolf, effective field theory is not very effective for these observables as there are too many unknown counterterms and the convergence of the series is, at best, very slow. It now appears t h a t the remaining theoretical t r e a t m e n t s t h a t are consistent with the d a t a are the chiral soliton model [9] (as discussed by Silva at this workshop) and the latticebased t r e a t m e n t [10] as discussed by Leinweber. T h e prediction of the chiral soliton model for the S A M P L E result is shown as the black diamond at entry 25 in Fig. 1, one of the few results with G%j > 0. T h e Leinweber et. al. result is shown as the last magenta star (entry 27), and represents a bold and precise prediction Ms
0.051
1
(2)
R.D. McKeown: Workshop summary
206
AQaTlheTa=a5.0deg
0.5
..
*/J ri r\
i 'A 0.51.0
10
15
20
25
Fig. 1. Theoretical predictions for fig are shown as symbols, with the SAMPLE experimental result indicated by the hatched region, statistical uncertainty (inner) and total uncertainty (outer). The theoretical results are compiled and discussed in [7,8]
t h a t lies just outside the la error bar of the experiment. There was much discussion about the justification for the very small quoted theoretical error, and I am sure t h a t this discussion will continue at many additional conferences and workshops. Of course during the last few years we have also seen the remarkably precise d a t a [11] from H A P P E X . This experiment has the notable distinction of being the first parityviolation experiment to use a high polarization b e a m from a strained GaAs crystal. They obtain a result for the combination of form factors at Q^ = 0.477 (GeV/c)2:
Gl
0.392G^ ^M/
l^p
0.091 =b 0.054 =b 0.039
(3)
which also can be used to rule out many models. Future planned running of H A P P E X involves lower Q'^ = 0.1 (GeV/c)^ measurements of elastic asymmetries for Helium as well as Hydrogen targets. And the major news at this conference are the new results from the Mainz A4 collaboration. They report asymmetries at ^ = 35° at two Q^ values, as shown in Fig. 2. These are very interesting results t h a t may indicate a nonvanishing strange quark contribution, and the A4 collaboration has plans for many more measurements including backward angles. In addition, we heard t h a t the G^ experiment has completed its production run at forward angles. T h e d a t a are under analysis and will hopefully be available in the very near future. Clearly, many years of planning, systematic studies, and equipment construction are now yielding an impressive d a t a set t h a t will surely provide us with a clearer picture of strangeness in the nucleon. T h e broader context of strangeness in the nucleon was also discussed at the workshop. M. Sainio u p d a t e d the situation regarding the sigma term. And J. Ellis presented an overview discussing the relation to the spin of the nucleon, the possible connection with exotic baryon states, and ideas for evading the OZIrule. It is intriguing to note t h a t the chiral soliton model, which is uniquely successful in producing a result for C ^ t h a t quantitatively agrees with
Fig. 2. Results reported at this workshop by the A4 collaboration. The solid line is the prediction for the case where both strange form factors vanish the S A M P L E experimental data, is also achieving significant notoriety for it's recent predictions of pentaquark states (for which there is increasingly substantial, but also controversial, experimental evidence).
3 Transverse spin asymmetries T h e subject of transverse spin asymmetries is receiving increased attention in the last year or so. Due to the fact t h a t it is higher order in the fine structure constant (i.e., 2 photon exchange) and it is further suppressed by the lorentz factor I / 7 , there was not much previous interest in this subject. T h e existence of recent d a t a from the S A M P L E experiment [12] demonstrated the feasibility of the method for measuring the small asymmetries (although larger t h a n parityviolation asymmetries). In addition, there has been increased recent interest due to the apparently significant 2 photon exchange effects in the interpretation of proton form factor measurements at high Q^ (see Sect. 4). At this workshop, we heard presentations of the S A M P L E d a t a by D. Spayde as well as new d a t a from A4 by S. Baunack. In addition, high energy d a t a from SLAC E158 will be available soon. T h e A4 collaboration measures transverse asymmetries with their luminosity monitors, or "Lumis" (mostly M0ller scattering), as well as with their detectors for elastic scattering off protons. The Lumi d a t a are quite precise (5%), but disagree with the NLO calculations for M0ller scattering by 25 a. This is not understood. T h e theoretical interpretation of the transverse asymmetries in elastic ep scattering was discussed by Pasquini and by RamseyMusolf. Pasquini et al. use the MAID description of photon couplings to the nucleon and the continuum. This prescription results in qualitatively correct behavior, but misses all the d a t a by about 2 a (they underpredict the magnitude of the S A M P L E asymmet r y but overpredict the magnitude of the A4 asymmetries). RamseyMusolf presented calculations [13] in the framework of effective field theory. They obtain reasonable agreement with the S A M P L E data, but it seems t h a t
R.D. McKeown: Workshop summary the A4 data are at too high a Q^ for this treatment to be vaUd.
207
have been achieved in the pursuit of strange quark form factors during the last decade have also motivated new higher precision experiments to test the standard electroweak model. At this workshop we heard a report by A. Vacheret on the SLAG El58 experiment, which mea4 Other form factors sures parityviolating M0ller scattering and a presentation K. DeJager presented a thorough review of the situation of the plans for Qweak at JLab by G. Smith. with the elastic electromagnetic form factors of the nuIn a fundamental sense, the parityviolating effect in cleon. There are recent G% data that are beginning to map electron scattering relates to the weak neutral couplings out the Q^ dependence of this elusive form factor with im to the electrons and quarks. At treelevel, these couplings pressive precision. This form factor is an important con depend on one parameter, the weak mixing angle Ow straint on nucleon structure as well as an important input Thus the various precision parityviolating electron scatin the interpretation of parityviolation experiments. An tering experiments aim to constrain sin^ Ow and compare other major topic of current interest relates to the contro with the very precise value measured at the Zpole in e+versy regarding the G\jG\^ data from recoil polarization e~ scattering. Due to radiative corrections, the value of measurements [14] at Q^ > 1 GeV^. The form factor ra sin^ Ow "runs" with Q^ in a predictable fashion accordtio extracted from these measurements disagrees with the ing to the standard model. Thus these experiments test results determined previously by Rosenbluth separation. for new particles in loops and exchanges associated with More recently, new Rosenbluth separation data [15] taken these radiative corrections to search for evidence of new by detection of the recoil proton (to reduce systematic er physics beyond the standard model. rors and radiative corrections) give strong support for the Another method to precisely determine sin^ Ow at low earlier Rosenbluth separation data. Q^ is via neutral current neutrino scattering. Recently, This conundrum appears to be resolvable by consid the NuTeV experiment at Fermilab has reported a meaeration of 2 photon exchange effects [16]. Guichon pre surement [20] of ratios of neutral current to charged cursented their analysis at this workshop, and it appears that rent cross sections for both neutrinos and antineutrinos. quantitative agreement with the two data sets is possible They observe a substantial effect in the neutrino ratio, although additional free parameters must be employed. which they interpret as an anomalous value of ("onshell" Their analysis indicates that 2 — 3% two photon exchange scheme) contributions with reasonable magnitudes can distort the sin^ Ow = 0.2277 =b 0.0013 =b 0.0009 (4) Rosenbluth plots by about the amount necessary to resolve the discrepancy. Clearly further measurements are which is 3cr from the standard model prediction of 0.2227=b desirable, including comparison of positron and electron 0.00037. K. McFarland presented these results and disscattering cross sections to test the model quantitatively. cussed many alternate explanations for the discrepancy. If this explanation is correct, then it would seem prudent It appears that relatively large isospin violation in the to revisit all cases where delicate Rosenbluth separations parton distribution functions could generate the observed have been performed to extract small amplitudes (e.g. the effect and would be consistent with all other experiments. famous R = (TL/GT in deep inelastic scattering). The cause of the NuTeV anomaly remains a subject of E. Beise discussed a recent reanalysis [17] of the Q^ much active study. dependence of the axial form factor of the nucleon as deThe recently published El58 results [21] are based on termined in quasielastic neutrino charged current interacthe data from the first of three data runs. The results tions. That work produced a slightly smaller value of the reported at this workshop included the first and second axial mass as compared to previous studies: MA = 1.001 =b data sets: 0.020 GeV. This quantity is relevant to the interpretation of backward angle parityviolation experiments such as fusin^ Ow = 0.2379 =b 0.0016 =b 0.0013 (5) ture planned measurements by G^ and A4. These collaborations also plan to run with deuterium, which allows sep which agrees well with the standard model prediction of aration of the axial term (as was done for SAMPLE [18]). 0.2386zb0.0006. (Note that E158 uses a different renormalization scheme to quote sin^ Ow) The situation is shown graphically in Fig. 3. The Qweak proposal [23] to JLab is to measure parity5 Electroweak tests violating elastic electronproton scattering at very forward The birth of parityviolating electron scattering was mar angles {0  8°) to achieve a low Q^ = 0.028 GeV^ This ked by the famous Prescott experiment at SLAG [19], will ensure that the strange quark effects in the electric which had the distinction to provide the first quantita form factor are small enough to be manageable. The high tive evidence for violation of parity in the neutral cur statistics will be achieved by using 180//A of electron beam rent as predicted by the standard model. Other early ex from strained GaAs with high polarization. The 35 cm periments in parityviolating electron scattering also were long liquid hydrogen target must absorb 2.5kW of beam attempts at testing the standard electroweak theory, no power without boiling effects. (Previous high power tartably the Mainz ^Be experiment and the Bates ^^C ex gets at SAMPLE and E158 achieved 500W and mainperiment. The advances in experimental techniques that tained excellent thermal stability.) The goal is to measure
208
R.D. McKeown: Workshop summary
sin^e;;?(Q')
In the shorter term, it seems t h a t the methods developed for parityviolation experiments can be used to obtain more d a t a on transverse asymmetries to explore the 2 photon exchange process in more detail. At high Q^ there could be a useful connection with the Generalized P a r t o n Distributions (see the talk by M. Gorshteyn) and at low Q^ there could be a connection with the nucleon polarizabilities. Clearly there is a great deal of experimental and theoretical work to do in this area, and we are just seeing the beginning of t h a t endeavor. So it seems t h a t at least one more workshop on this topic is welljustified, and if the organizers can find a location t h a t is of comparable quality to the Grenoble workshop I am sure we (and many new younger people) will be there to hear of the exciting developments t h a t are sure to come. Q (GeV/c)
Fig. 3. Plot of effective sin^ 6w vs Q^ showing the E158 result, NuTeV result, Zpole value, and Cs atomic parityviolation. The theoretical curve is from [22]
References
sin^ Ow to a precision of 7 (compared to the ultim a t e precision of E158 of r^ . It is hoped t h a t this ambitious experiment could be mounted in 2007 and begin commissioning studies shortly thereafter. One should keep in mind t h a t this is the time scale for LHC to start, and the future of such experiments in the postLHC era is far from clear.
6 Outlook T h e field of parityviolating electron scattering has entered an extremely productive phase. Over the next 35 years we should have in hand a definitive dataset mapping out the role of strange quarks in the electroweak form factors of the nucleon. It is difficult to judge how t h a t will t u r n out and how the theoretical interpretation will develop in response to the data. Perhaps a round of higher precision experiments will be indicated. Already we see t h a t if one takes the Leinweber, et al. prediction seriously, one could justify building a superSAMPLE experiment with high polarization C W beam to achieve precision on lis comparable to this theoretical prediction. Perhaps we will even have new d a t a from Qweak in this time frame to further test the s t a n d a r d electroweak theory. And it might even be worth revisiting the m e t h o d of elastic scattering from a spinless nucleus like ^^C to perform other high precision tests.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
D. Kaplan, A. Manohar: Nucl. Phys. B 310, 527 (1988) R.D. McKeown: Phys. Lett. B 219, 140 (1989) D.H Beck: Phys. Rev. D 39, 3248 (1989) E.J. Beise, R.D. McKeown: Parity Violating Electron Scattering^ (World Scientific, Singapore, 1990) Bates proposal 8906, R.D. McKeown, D.H. Beck: spokespersons D. Spayde et al.: Phys. Lett. B 583, 79 (2004) D.H. Beck, B.R. Holstein: Int. Journ. Mod. Phys. E 10, 1 (2001) D.H. Beck, R.D. McKeown: Ann. Rev. Nucl. 51, 189 (2001) A. Silva, H.C. Kim, K. Goeke: Nucl. Phys. A 721, 417 (2003) D.B. Leinweber et al.: heplat/0406002 K.A. Aniol et al.: Phys. Lett. B 509, 211 (2001) S.R Wells et al.: Phys. Rev. C 63, 064001 (2001) L. Diaconescu, M.J. RamseyMusolf: nuclth/0405044 M.K. Jones et al.: Phys. Rev. Lett. 84, 1398 (2000); O. Gayou et al.: Phys. Rev. Lett. 88, 092301 (2002) JLab experiment EOlOOl, J. Arrington, R. Segel: spokespersons P.A.M. Guichon, M. Vanderhaeghen: Phys. Rev. Lett. 91, 142303 (2003) H. Budd, A. Bodek, J. Arrington: hepex/0308005 T. Ito el al.: Phys. Rev. Lett. 92, 102003 (2004) C.Y. Prescott et al.: Phys. Lett. B 77, 347 (1978) G.P. Zeller et al.: Phys. Rev. Lett. 88, 0918021 (2002) P.L. Anthony et al.: Phys. Rev. Lett. 92, 181602 (2004) A. Czarnecki, W.J. Marciano: Int. J. Mod. Phy. A 15, 2365 (2000) JLab experiment E02020: J. Bowman, R. Carlini, J. Finn, S. Kowalski, S. Page, spokespersons