ebook img

Proton elastic form factor ratios to $Q^2$ = 3.5 GeV$^2$ by polarization transfer PDF

1.2 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Proton elastic form factor ratios to $Q^2$ = 3.5 GeV$^2$ by polarization transfer

Proton elastic form factor ratios to Q2 = 3.5 GeV2 by polarization transfer V. Punjabi,1,∗ C.F. Perdrisat,2 K.A. Aniol,7 F.T. Baker,4 J. Berthot,6 P.Y. Bertin,6 W. Bertozzi,22 A. Besson,6 L. Bimbot,26 W.U. Boeglin,10 E.J. Brash,5,30,a D. Brown,21 J.R. Calarco,23 L.S. Cardman,30 Z. Chai,22 C.-C. Chang,21 J.-P. Chen,30 E. Chudakov,30 S. Churchwell,8 E. Cisbani,14 D.S. Dale,17 R. De Leo,13 A. Deur,6,30 B. Diederich,25 J.J. Domingo,30 M.B. Epstein,7 L.A. Ewell,21 K.G. Fissum,22,b A. Fleck,5 H. Fonvieille,6 S. Frullani,14 J. Gao,22,c F. Garibaldi,14 A. Gasparian,12,17,d G. Gerstner,2 S. Gilad,22 R. Gilman,3,30 A. Glamazdin,18 C. Glashausser,3 J. Gomez,30 V. Gorbenko,18 A. Green,33 J.-O. Hansen,30 C.R. Howell,8 G.M. Huber,5 M. Iodice,14 C.W. de Jager,30 S. Jaminion,6 X. Jiang,3 M.K. Jones,2,30 W. Kahl,28 J.J. Kelly,21 M. Khayat,16 L.H. Kramer,10 G. Kumbartzki,3 M. Kuss,30 E. Lakuriki,29 G. Laveissi`ere,6 J.J. LeRose,30 M. Liang,30 R.A. Lindgren,32 N. Liyanage,22,30,32 G.J. Lolos,5 R. Macri,8 R. Madey,16,12 S. Malov,3 D.J. Margaziotis,7 P. Markowitz,10 K. McCormick,25,16,3 J.I. McIntyre,3 R.L.J. van der Meer,30,5 R. Michaels,30 B.D. Milbrath,9 J.Y. Mougey,19 S.K. Nanda,30 E.A.J.M. Offermann,30,e 5 Z. Papandreou,5 L. Pentchev,2 G.G. Petratos,16 N.M. Piskunov,15 R.I. Pomatsalyuk,18 D.L. Prout,16 0 0 G. Qu´em´ener,2,19 R.D. Ransome,3 B.A. Raue,10 Y. Roblin,6,30 R. Roche,11,25 G. Rutledge,2 P.M. Rutt,30 2 A. Saha,30 T. Saito,31 A.J. Sarty,11,f T.P. Smith,23 P. Sorokin,18 S. Strauch,2,g R. Suleiman,16,22 K. Takahashi,31 n J.A. Templon,4,h L. Todor,25,i P.E. Ulmer,25 G.M. Urciuoli,14 P. Vernin,27 B. Vlahovic,24 H. Voskanyan,34 a K. Wijesooriya,2 B.B. Wojtsekhowski,30 R.J. Woo,20 F. Xiong,22 G.D. Zainea,5 and Z.-L. Zhou22 J 0 The Jefferson Lab Hall A Collaboration 2 1Norfolk State University, Norfolk, VA 23504 2College of William and Mary, Williamsburg, VA 23187 1 3Rutgers, The State University of New Jersey, v Piscataway, NJ 08855 8 4University of Georgia, Athens, GA 30602 1 5University of Regina, Regina, SK S4S OA2, Canada 0 6Universit´e Blaise Pascal/CNRS-IN2P3, 1 F-63177 Aubi`ere, France 0 7California State University, 5 Los Angeles, Los Angeles, CA 90032 0 8Duke University and TUNL, Durham, NC 27708 / 9Eastern Kentucky University, Richmond, KY 40475 x e 10Florida International University, Miami, FL 33199 - 11Florida State University, Tallahassee, FL 32306 l 12Hampton University, Hampton, VA 23668 c u 13INFN, Sezione di Bari and University of Bari, 70126 Bari, Italy n 14INFN, Sezione Sanit`a and Istituto Superiore di Sanit`a, : 00161 Rome, Italy v 15JINR-LHE, 141980 Dubna, i Moscow Region, Russian Federation X 16Kent State University, Kent, OH 44242 r 17University of Kentucky, Lexington, KY 40506 a 18Kharkov Institute of Physics and Technology, Kharkov 310108, Ukraine 19Laboratoire de Physique Subatomique et de Cosmologie, CNRS-IN2P3, F-38026 Grenoble, France 20University of Manitoba, Winnipeg, MB R3T 2N2 21University of Maryland, College Park, MD 20742 22Massachusetts Institute of Technology, Cambridge, MA 02139 23University of New Hampshire, Durham, NH 03824 24North Carolina Central University, Durham, NC 27707 25Old Dominion University, Norfolk, VA 23508 26Institut de Physique Nucl´eaire, F-91406 Orsay, France 27CEA Saclay, F-91191 Gif-sur-Yvette, France 28Syracuse University, Syracuse, NY 13244 29Temple University, Philadelphia, PA 19122 30Thomas Jefferson National Accelerator Facility, Newport News, VA 23606 31Tohoku University, Sendai 980, Japan 32University of Virginia, Charlottesville, VA 22901 33Western Cape University, Capetown, South Africa 2 34Yerevan Physics Institute, Yerevan 375036, Armenia ∗ (Dated: February 8, 2008) Theratiooftheprotonelasticelectromagneticformfactors,GEp/GMp,wasobtainedbymeasuring Pt and Pℓ, the transverse and longitudinal recoil proton polarization components, respectively, for the elastic ~ep → ep~ reaction in the four-momentum transfer squared range of 0.5 to 3.5 GeV2. In the single-photon exchange approximation, the ratio GEp/GMp is directly proportional to the ratio Pt/Pℓ. The simultaneous measurement of Pt and Pℓ in a polarimeter reduces systematic uncertainties. The results for the ratio GEp/GMp show a systematic decrease with increasing Q2, indicating for the first time a definite difference in the distribution of charge and magnetization in theproton. Thedatahavebeenre-analyzedandsystematicuncertaintieshavebecomesignificantly smaller than previously published results. PACSnumbers: 25.30.Bf,13.40.Gp,24.85.+p I. INTRODUCTION is very large and the quarks become asymptotically free. ItisnotknownpreciselyatwhatvalueofQ2 pQCDmay start to dominate; however, expectations are that this One of the fundamental goals of nuclear physics is to willnotoccuruntilQ2 isatleastinthe tensofGeV2 [3]. understand the structure and behavior of strongly inter- Predicting nucleon form factors in the non-perturbative acting matter in terms of its basic constituents, quarks regime, where soft scattering processes are dominant, is and gluons. An important step towards this goal is the very difficult. As a consequence there are many phe- characterizationof the internal structure of the nucleon; nomenologicalmodelswhichattempttoexplainthedata the four Sachs elastic electric and magnetic form factors inthisdomain;precisemeasurementsofthenucleonform of the proton and neutron, G , G , G and G , Ep Mp En Mn factors are necessary to constrainand test these models. are key ingredients of this characterization. The elastic Only the magnetic form factor of the proton, G , is electromagnetic form factors are directly related to the Mp knownwith very good accuracyin this region. The elec- chargeandcurrentdistributionsinsidethenucleon;these tric form factor, G , was not well measuredbeyond Q2 formfactorsareamongthemostbasicobservablesofthe Ep of 1 GeV2 before this experiment. Both G and G , nucleon. En Mn theelectricandmagneticformfactorsoftheneutron,re- The first direct evidence that the proton has an inter- spectively, were also poorly knownat any Q2 value until nalstructurecamefromameasurementofitsanomalous recently. New measurements of G at Jefferson Lab Mn magnetic moment 70 yearsago by O. Stern [1]; it is 2.79 [4] up to Q2=4.8 GeV2 will bring the knowledge of this times larger than that of a Dirac particle of the same formfactorto comparablelevelsofaccuracyasfor G . Mp mass. The first measurement of the charge radius of the For the neutron electric form factor, two new Jefferson proton,byHofstadteret al. [2],yieldedavalueof0.8fm, Lab experiments [5, 6] have extended the Q2 range to quite close to the modern value. 1.5 GeV2, and two approvedexperiments [7, 8] will soon The theory that describes the strong interaction be- extendtheQ2 rangeto4.3GeV2,withanaccuracycom- tween quarks and gluons is Quantum Chromodynamics parable to that of the three other form factors. (QCD). Perturbative QCD (pQCD) makes rigorous pre- Theelectromagneticinteractionprovidesauniquetool dictions when the four-momentum transfer squared, Q2, to investigate the internal structure of the nucleon. The measurement of electromagnetic form factors in elastic, inelastic, and structure functions in deep inelastic scat- tering of electrons and muons, has been a rich source of ∗Email address:[email protected] information on the structure of the nucleon. a Present address: Christopher Newport University, Newport In the single virtual photon exchange approximation News,VA23606,USA. b Present address: Universityof Lund, Box 118, SE-221 00 Lund, forelasticscattering,thehadroncurrentoperatorcanbe Sweden. expressed in terms of two form factors: F1, the Dirac c Present address: California Institute of Technology, Pasadena, form factor, and F , the Pauli form factor. These form 2 CA91125, USA. factors and the Sachs electric and magnetic form factors dPresentaddress: NorthCarolinaAg. andTech. StateUniversity, are related according to: Greensboro,NC27411, USA. e Present address: Renaissance Technology Corp., Setauket, NY 11733, USA. GE =F1 τκF2 and GM =F1+κF2, (1) − f Present address: Saint Mary’s University, Halifax, NS, Canada B3H3C3. where τ = Q2/4M2, κ is the anomalous magnetic mo- p gPresentaddress: GeorgeWashingtonUniversity,Washington,DC mentandM themassoftheproton. InthelimitQ2 0, 20052, USA. → G =1,G =0,G =µ ,andG =µ ,whereµ h Presentaddress: NIKHEF,Amsterdam,TheNetherlands. Ep En Mp p Mn n p i Present address: Carnegie Mellon University, Pittsburgh, PA and µn are the nucleon magnetic moments. In the Breit 15217, USA. frame, GE and GM are the Fourier transforms of the 3 charge and magnetization distributions in the nucleon, respectively. A. Previous GEp Measurements Using the Rosenbluth Separation Method Both the elastic cross section and the polarization ob- servables of the elastic ep reaction can be expressed in termsofeithertheSachsortheDiracandPauliformfac- tors. These form factors are Lorentz scalars and depend only upon Q2, the four-momentum transfer squared of the reaction. A complete separation of the electric and magnetic terms is evident in the cross section formula when the Sachs form factors are used. It is then possi- ble to obtain both G2 and G2 separately, using the Ep Mp Rosenbluth method [9, 10]. In the one-photon exchange approximation, the cross section in terms of the Sachs form factors can be expressed as: dσ = α2 Ee cos2 θ2e G2 + τG2 1 , (2) dΩ 4E3 sin4 θe Ep ǫ Mp 1+τ beam 2 h i(cid:18) (cid:19) where ǫ = 1+2(1+τ)tan2(θe) −1 is the longitudinal 2 FIG. 1: World data prior to 1998 for (a) GEp/GD and (b) polarization(cid:2)of the virtual photon(cid:3), with values between GMp/µpGD versus Q2. Refs. Litt et al. [11]△, Berger et al. 0< ǫ <1, Ebeam and Ee are the energies of the incident [12]2, Price et al. [13]•, Bartel et al. [14]◦, Walker et al. andscatteredelectron,respectively,andθeistheelectron [15]⋆, Andivahiset al. [16]3 and Sill et al. [17]∗. scattering angle in the laboratory frame. Figure 1 shows previous results of G and G ob- Ep Mp tained by Rosenbluth separations, plotted as the ratios G /G and G /µ G versus Q2, up to 6 GeV2. areabletoobtaintheelectricformfactorGEp evenwhen Ep D Mp p D Here G = (1 + Q2/m2 )−2 is the dipole form fac- it is very small. D D tor, with the constant m2 empirically determined to be For one-photon exchange, in the ~ep ep~ reaction, D → 0.71 GeV2. For Q2 < 1 GeV2, the uncertainties for both thescatteringoflongitudinallypolarizedelectronsresults in a transfer of polarization to the recoil proton with G and G are only a few percent, and one finds Ep Mp that G /µ G G /G 1. For G above Q2 only two non-zero components, Pt perpendicular to, and Mp p D Ep D Ep = 1 GeV2, the lar≃ge uncertaint≃ies and the scatter in re- Pℓ parallel to the proton momentum in the scattering plane. For 100 % longitudinally polarized electrons, the sults between different experiments, as seen in Fig. 1, polarizations are [18, 19, 20, 21]: illustrate the difficulties in obtaining G by the Rosen- Ep bluth separation method. In contrast, the uncertainties for G obtained from cross section data with the as- Mp I P = 0 (3) 0 n sumption G = G /µ , remain small up to Q2 = Ep Mp p θ 31.2 GeV2[17]. In Eq. (2) the GMp part of the cross sec- I0Pt = 2 τ(1+τ)GEpGMptan e (4) tion,whichisaboutµ2 timeslargerthantheG part,is − 2 p Ep 1p θ alsomultipliedbyτ; therefore,asQ2 increases,the cross I P = (E +E ) τ(1+τ)G2 tan2 e (5) 0 ℓ M beam e Mp 2 section becomes dominated by the G term, making p Mp p the extraction of GEp more difficult by the Rosenbluth where I0 is proportionalto the unpolarized cross section separation method. and is given by: τ B. Polarization Transfer Method I0 =G2Ep+ ǫG2Mp (6) Eqs. (4) and (5) show that I P and I P are pro- The proton form factor ratio G /G can be ob- o t o ℓ Ep Mp portional to G G and G2 , respectively. Together tained from polarization observables of the ~ep ep~ or Ep Mp Mp → these equations give: ~e~p ep reaction, the recoil proton polarization transfer → coefficients or the beam-target polarization asymmetry, respectively. Bothreactionscontainaninterferenceterm GEp Pt (Ebeam+Ee) θe = tan (7) proportionaltoGEpGMp;hencepolarizationexperiments GMp −Pℓ 2Mp 2 4 IfonlythepolarizationcomponentsP andP aremea- highest Q2 data points, a bulk GaAs photo-cathode ex- t ℓ sured,aswasthecaseinthisexperiment,thenfromEqs. cited by circularly polarized laser light produced beams (4) and (5) the form factors G and G cannot be with polarization of 0.39 and currents up to 115 µA; Ep Mp ∼ ∼ obtained separately, only their ratio can be determined. the sign of the beam helicity was changed at the rate of To obtain G and G separately, I in Eq. (6) must 30 Hz. For the lower Q2 data points, a strained GaAs Ep Mp 0 be obtained from cross section measurements. crystal was used and typical polarizations of 0.6 were ∼ The ratio G /G is obtained from a single mea- achievedwithcurrentsbetween5µAand15µA;thesign Ep Mp surement of the two recoil polarization components P of the beam helicity was changed at the rate of 1 Hz. t andP inapolarimeter,whereastheRosenbluthmethod The beam polarization was measured periodically with ℓ requiresatleasttwocrosssectionmeasurementsmadeat a Mott polarimeter in the injection line, and a Møller different energy and angle combinations at the sameQ2. polarimeter [26] in Hall A [27]. The recoil polarization method was first used in elec- The Hall A Møller polarimeter uses magnetized fer- tron scattering experiments to obtain the neutron form romagnetic supermendure foils as a polarized electron factors in the 2H(~e,e′~n)p reaction [22] and to measure target. The scattered electrons are detected in co- the form factor ratio GEp/GMp in ~ep ep~ for the free incidence in the Møller spectrometer in the range of proton [23, 24], as well as in the 2H(~e,→e′~p)n reaction for 75◦ < θ < 105◦. The Møller spectrometer con- CM the proton in the deuteron at small Q2-values [25]. sists of three quadrupoles and a dipole magnet to bend For completeness we mention here that a small nor- scattered electrons toward the detector. The detec- mal componentPind is induced by two-photonexchange tor contains two identical modules for coincidence mea- n mechanism, independent of beam polarization. The ob- surements; each module consists of a plastic scintilla- servablesPt andPℓ ofthisexperimentareentirelydueto tor and four blocks of lead glass. The Møller scat- polarizationtransfer, and the analysis method described tering cross section depends on the beam and the in this paper allows complete separation of helicity de- Møllertargetpolarizations,P andP ,respectively, e,i targ,i pendent and helicity independent polarization compo- σ (1+ (A P P )),wherei=X,Y,Z ∝ i=X,Y,Z ii targ,i e,i nents. defines the projections of polarization. The analyzing P In this paper, we present the G /G ratios, pri- power A depends on the scattering angle θ and has Ep Mp ii CM mary results from this experiment, obtained at Jefferson its maximum at θ = 90◦. Statistical uncertainty CM Lab using the recoilpolarizationmethod described here. varies between 0.2 % and 0.8 % for each measurement. Theexperimentalsetup,inparticularthe focalplanepo- Total relative uncertainty of the beam polarizationmea- larimeter(FPP),isdescribedinpartII.Thedataanalysis surement is 3 %, when systematic and statistical un- ≤ is presented in part III; this part also includes a discus- certainties are combined. In this experiment, the beam sion of the FPP calibration, the secondary results of the helicity cancels in the ratio P /P , so strictly speaking t ℓ experiment, which are independent measurements of an- measurementofthebeampolarizationisnotneeded,but alyzingpowersattenprotonenergiesbetween0.244GeV the beam polarization was measured periodically to en- and 1.795 GeV. Part IV includes the main results of the sure that the beam was polarizedand also to allow FPP experiment: the ratios GEp/GMp at 0.5 GeV2 Q2 calibration as explained in section III D. ≤ ≤ 3.5GeV2,andananalysisofsystematicuncertainties. A The beam current was monitored continuously during discussion of theoretical calculations as compared to the the experiment using resonant (RF) cavities. The beam data is in part V and conclusions are presented in part current monitor (BCM) in Hall A consists of an Unser VI. [28] monitor sandwiched between two RF cavities. The Unser monitor provides an absolute measurement of the current; the RF cavities are calibrated relative to the II. THE EXPERIMENT Unser monitor periodically. Both components are en- closedina box to shield them fromstraymagnetic fields The combinationof high energy,current,polarization, and for temperature stabilization. anddutyfactor,uniquetotheContinuousElectronBeam Beam position and direction at the target were deter- Accelerator Facility (CEBAF) of the Thomas Jefferson mined from two beam position monitors (BPM) located National Accelerator Facility (JLab), makes it possible at a distance of 7.524 m and 1.286 m upstream of the to investigate the internal structure of the nucleon with target position during this experiment. Each BPM is a higher precision than ever before. In this experiment, cavity with a four-wire antenna with wires positioned at we have measuredthe polarization transferredto the re- 45◦ from the horizontal and vertical. The relative po- ± coilproton,withalongitudinallypolarizedelectronbeam sition of the beam on the target can be determined to scattered by an unpolarized hydrogen target. about100µmforcurrentsabove1µAbyusingthe tech- TheexperimentwasperformedinHallAatJLab. The niqueofdifference-over-sumbetweenthesignalsfromthe longitudinalandtransversepolarizationsofthe outgoing four antenna wires. To obtain the absolute position of proton were measured for the ~ep ep~ reaction, in a the beam, the BPMs are calibrated with respect to wire range of Q2 from 0.5 GeV2 to 3.5 G→eV2. The beam en- scanners which are located close to each of the BPMs at ergy ranged from 0.934 GeV to 4.091 GeV. For the five 7.353mand 1.122mfrom the target. The wire scanners 5 are surveyed with respect to the hall coordinates. The dispersive, and mixed focusing in the non-dispersive di- beam position is recorded for every event. rection. The front quadrupole Q1 has a magnetic length The cryogenictargetcontainedthree loops. Eachloop of0.941mandaninner radiusof0.15m, andit isfocus- included one 15 cm and one 4 cm aluminum cell; both ing in the dispersive direction. The quadrupoles Q2 and cells have a diameter of 6.35 cm. The sidewall thickness Q3 provide focusing in the non-dispersive direction and of each cell was 178 µm, and entrance and exit window theybothhaveaninnerradiusof0.30mandamagnetic thicknesseswere71µmand102µm,respectively. Asthis length of 1.82 m. All three are superconducting, cos(2φ) experiment required only a liquid hydrogen target, only quadrupoles with an outside cylindrical magnetic field loop 3 was used; the other two loops were filled with he- return iron yoke. The quadrupole fields are monitored liumgasat0.12MPatosavecoolingpower. Thenominal using Hall probes and Gauss-meters. temperature and pressure for the liquid hydrogen target The dipole in the HRS has a superconducting coiland during the experiment were 19 K and 0.17 MPa, respec- warmironconfigurationwithshapedpole-facesandfield tively. Thetargetdensitydecreasedbyabout5%[29]at gradienttohelpfocusinthedispersivedirection;itsmag- anincidentbeamcurrentof120µAcomparedtoitsden- neticfielddeflectstheparticlesinaverticalplaneby45◦. sity at 10 µA (measured in an earlier experiment). The Thebendingradiusofthe dipoleis8.40mwithacentral targetassembly was housed inside a scattering chamber. gapof0.25mandaneffectivelengthof6.6m. Thenom- The scatteringchamberinHallA is dividedinto three inalmaximumcentraltrajectorymomentumis4GeV/c, sections. The vacuum in all three sections is maintained themomentumresolutionisoftheorderof10−4,andthe atalevelof0.13mPa. Thebottomsectionisfixedtothe momentum acceptance is 5 %. The field in the dipole ± Hall A pivot and the top part of the scattering chamber ismonitoredandregulatedbyanNMRprobe;itisstable contains the target’s cryogenic plumbing. The middle at the 10−5 level. section of the chamber has an inner diameter of 103.7 The focal plane detector assembly for each spectrom- cm,awallthicknessof5cmofaluminumandaheightof eter is enclosed in a metal and concrete shield house to 91 cm. The entrance and exit beam pipes are connected reduce the backgroundradiation. Bothdetector systems to this section. The scattered particles go through alu- contain two vertical drift chambers (VDC), and scintil- minumexitwindows18cmhighand406µmthicktothe lator arrays called S1 and S2. In addition, the electron entrances of two high resolution spectrometers (HRSs). detector system contains a gas Cˇerenkov detector, and In order to reduce the heat deposition in a very small lead-glass arrays used as pre-shower and shower detec- area of the target from an intense electron beam and tors; the hadron detector package contains an aerogel to minimize corresponding target density changes, the Cˇerenkov, a gas Cˇerenkov and the focal plane polarime- beam was rastered before it strikes the target. The fast ter (FPP). The assembly of the hadron arm detectors is rastering system is located 23 m upstream of the target. shown in Fig. 2. In this experiment only the VDCs and The rastering system contains two sets of magnets, one S1 and S2 detectors were used on the electron side, and to deflect the beam vertically, and the other to deflect the VDCs, S1, S2 and the FPP on the hadron side. it horizontally. The magnetic field varies sinusoidally at ThetwoVDCs,installedclosetothefocalplaneofeach 17.7 kHz in the vertical direction and 25.3 kHz in the HRS, giveprecisereconstructionofpositions andangles. horizontaldirection. Thetypicalrasteredbeamspotsize The central ray of the spectrometer passes through the at the target was 3.5 3.5 mm2. centerofeachVDCat45◦tothevertical. ThetwoVDCs ≈ × A box located at the entrance window of each spec- are separated by 33.5 cm. The active area of each one trometercancontainthreemovablecollimators. Theup- is 211.8 28.8 cm2 with two wire planes at 45◦ to the × per collimator is a stainless steel 5 mm thick sieve slit; dispersivedirectionandperpendicularto eachother. All it is used to study the optics of the spectrometers. The VDCs are operated at a high voltage of 4 kV, and the middlecollimatorismadeoftungsten,andis8cmthick, gas mixture used in these chambers is 62 % argon and 6.29cmwide,and12.18cmhigh,andislocatedatadis- 38 % ethane. The position resolution of each plane is ≈ tance of 110.9 cm from the target. The bottom position 100 µm. isemptyandperformsnocollimationandistheoneused InbothHRSdetectorsystemstherearesixscintillator inthisexperiment;the collimationisthendefinedbythe paddles in plane S1 and six in S2. Each paddle is seen apertureofthemagneticelementsoftheHRS.Thespace byonephotomultiplierateachend. Thepaddlesinboth between the exit window of the scattering chamber and planes are oriented such that they are perpendicular to the entrance window of each HRS consists of 20 cm of the spectrometer central ray. The distance between the air. two planes is 1.933 m on the electron side and 1.854 m Elastic ep events were selected by detecting scattered on the hadron side. The active area of the S1 plane is electrons and the recoiling protons in coincidence, us- 170 36 cm2, and of the S2 plane 212 60 cm2. The × × ing the twoidenticalHRSs ofHallA. Eachspectrometer thickness of each paddle in both planes is 0.5 cm. The consists of three quadrupoles and one dipole. The con- scintillator paddles of both planes overlap by 0.5 cm to figuration is QQD Q, two quadrupoles followed by an achieve complete coverageof the focal plane area. n indexed dipole (n=-1.25), and a quadrupole. Both spec- The trigger for both spectrometers is similar and is trometers are designed for point-to-point focusing in the formed from a coincidence between the signal from two 6 temporarily for on-line analysis to monitor the exper- iment and then transfered to tapes in the JLab mass storagesystem to be used later for final off-line analysis. A. Focal Plane Polarimeter Polarizationexperimentshavebecomeincreasinglyim- portant in the study of nuclear reactions. Focal plane polarimeters were standard equipment at intermedi- ate energy proton accelerators, such as LAMPF [31], TRIUMF[32], SATURNE [33], and PSI [34]. Experi- mentsatthesefacilitieshavedemonstratedthesensitivity of spin observables to small amplitudes. Similar consid- erations have more recently led to the development of protonpolarimetersforuseatelectronaccelerators,such as the MIT-Bates laboratory [35] and at the Mainz Mi- FIG. 2: Schematic of the hadron arm detector package in- crotron [36]. cluding thepolarimeter. The FPP in Hall A at JLab was designed, built, in- stalledandcalibratedby acollaborationofRutgersUni- versity, the College of William and Mary, Norfolk State scintillator planes, S1 and S2. The first requirement is University,the UniversityofGeorgia,andthe University to form a coincidence between the left and right signals of Regina [37]. from eachpaddle in the S1 and S2 planes. The time res- olutionperplaneisabout0.3ns(1σ). Thesecoincidence signals are fed into a memory lookup unit (MLU) which 1. Physical Description of Focal Plane Polarimeter is programmed to form a second trigger, called “S-Ray” (trigger T1 for electron HRS and trigger T3 for hadron The FPP is a part of the hadron detector package of HRS), by requiringthat the paddlesthat firedin the S1- the Hall A high resolution spectrometer. As shown in and S2 planes belong to an allowed hit pattern. The al- Fig. 2, the polarimeter is installed downstreamfrom the lowed hit pattern requires that if paddle N is fired in focal plane VDCs; it is oriented along the mean particle the S1 plane, then in the scintillator plan|e |S2 a signal directioninthefocalplanearea,at45◦ tothevertical. It mustcomefrompaddle N-1 or N or N+1 ortheover- consists of two front detectors to track incident protons, | | | | | | lap between two of those. Each spectrometer also has followed by a carbon analyzer and two rear detectors to a “loose” trigger: for the electron arm it is formed by track scattered particles. requiring that signals from two out of three detectors be ThefourtrackingdetectorsoftheFPParedriftcham- present, S1-plane, S2- plane and the Cˇerenkov detector; bers made ofstrawtubes; the strawtube designis based the hadron arm loose trigger requires signals from just on the one used for the EVA detector at Brookhaven S1-plane and S2-plane. These loose triggers are used to National Laboratory [38]. The four drift chambers con- obtaindetectorefficiencies. Finally,theS-raytriggersig- tain a total of 24 planes of straw tubes. Twelve of these nalsT1andT3formthe“coincidence”triggerT5forthe are in the 2 front chambers CH1 and CH2, where they experiment. These five different triggers, T1 to T5, are are oriented along the u and v directions at +45◦ and sent to the trigger supervisor (TS) and are also counted 45◦ relative to the spectrometer dispersive direction − by scalers. (x); each chamber has the configuration vvvuuu. In the The TS was designed and built by the CEBAF online backchambersCH3andCH4theconfigurationisuuvvxx dataacquisition(CODA)group. ThefunctionsoftheTS and uuuvvv, respectively, where x indicates that the x- include interface between the hardware trigger electron- coordinateismeasured,andthestrawsareorientedalong ics andthe computer dataacquisitionsystem,producing the y direction. acomputerbusysignalthatisusedtocalculatethecom- Theindividualstrawsarethinwalledmylartubescon- puterdead-time,andpre-scalingofthetriggerinputs T1 taining the anode wire at their center, and the gas mix- to T5. ture. They are made by wrapping an inner 10 µm thick The data acquisitionsystem was entirely developed at aluminum foil, and two 50 µm thick mylar foils around JLabbytheCODAgroup[30]. InthisexperimentCODA a mandrel, together with a heat setting glue. The inner wasrunningonasingleHewlett-Packard(HP9000)com- diameter of the straw is 10.44 mm. The ground con- puter. The two main tasks of CODA are to transmit nection to outside is made with silver epoxy to a brass information from the detectors to the computer via read ferrule inserted at both ends. As shown in Fig. 3, a del- outcontrollers(ROC)andbuildeventsbycollectingdata rin feed-through inserted in the ferrule provides gas feed from all the ROCs. The data are stored on a hard disk and exhaust, and a positioning hole for a brass slit pin, 7 Theanalyzerconsistsoffivesetsofgraphiteplateswith differentthicknesses. Eachsetismadeoftwohalvesthat can be moved separately on left and right. The plates are beveled at an angle of 45◦ so that the two halves overlap when closed. The plates have thicknesses of 1.9 cm, 3.8 cm, 7.6 cm, 15.2 cm and 22.9 cm and they are separated by 1.6 cm. The ability to vary the thick- ∼ nessoftheanalyzerisnecessarytooptimizetheefficiency whilemaintainingtheCoulombmultiplescatteringangle within acceptablelimits. The carbonthicknesses usedin FIG. 3: Schematicshowing theend assembly of straw tubes. this experiment at different proton energies are given in Table I. The main contribution to small angle multiple Coulomb scattering originates in the analyzer, with ad- into which the high voltage anode wire is solderedunder ditional contributions from the scattering in the S2 pad- prescribedstress to insure that the gravitationalsagging dles, the straw tubes in the two front chambers and the andelectrostaticdeflectionofthewirearesmall. Thean- air between them. Table I gives typical multiple scatter- odewireisgoldplated25µmdiametertungsten-rhenium inganglesforthe relevantkinematicsofthisexperiment. wire. As Coulomb scattering is largely spin independent, in The two front chambers are identical to one another firstorderitdoesnotaffectthepolarimeterperformance; and contain 1008 straw tubes each. In these two cham- however,multiplescatteringsmearsoutthenuclearscat- bers the straw tubes are precisely spaced by inserting teringdistribution,andwillthereforeresultinasmallde- their ends into aluminum blocks in which holes of diam- pendenceofanalyzingpowerupontheanalyzerthickness. eter 10.75 mm have been drilled with 10.95 mm spacing Coulomb scattering results in a strong forward peak of center-to-center. Each block has 3 layers of such holes protons which did not undergo nuclear scattering; these separatedverticallyby9.5mmandshiftedbyhalfahole events are suppressed by requiring a minimum scatter- separation, providing a very tight packing. Each block ingangle2 to3 times largerthanthe multiple scattering accommodates 16 straws in each of the 3 layers, for a rmsangle;herethisminimumanglewasfixedat47mrad total of 48. The spacing between the straws of each (2.7◦). plane are maintained with mylar shims glued every 30 cmalongthe length. The activeareaforthe frontcham- TABLE I: Multiple scattering for the ten proton kinetic en- bers is 60×209 cm2. The nominal distance between the ergies. Tpinc. is the incident proton kinetic energy, Cthick. two frontchambers is 120 cm center to center; the inter- the analyzer thickness, ϑrms the root mean square Coulomb veningspacewasoccupiedduringthisexperimentbythe scattering angle, in FPP chambers (fpp),in theanalyzer (C) 100cmgasCˇerenkovdetector: althoughnotusedinthis and added quadratically (total). experiment, it contributed an additional 3 mrad to the multiple scattering at the lowest proton energy. Tpinc. Cthick. ϑrfmpps ϑrCms ϑrtomtasl (GeV) (cm) (mrad) (mrad) (mrad) Thetwobackchamberscontainatotalof3102straws. Eachchambercontainssixplanesofstraws,withthesuc- 0.267 7.62 7.9 16. 17.8 cessive layers of straws glued together using precision 0.426 22.86 5.9 20.8 21.6 guiding plates and pins to insure accurate positioning, 0.639 41.91 3.8 18.1 18.5 andis protectedby0.36mmthickcarbonfiberpanels at 0.799 41.91 3.1 16.1 16.4 the top and the bottom. Both chambers are positioned 0.959 49.53 2.8 14.8 15.1 on a 1.9 cm thick and 31.5 cm wide plastic honeycomb, 1.014 49.53 2.7 13.1 13.4 aluminum faced composite, that also provides a mount- 1.156 49.53 2.2 11.6 11.8 ingsurfaceforgas,highvoltagedistribution,andreadout 1.333 49.53 2.1 10.4 10.6 boards. ChambersCH3andCH4haveactiveareasof124 272cm2 and142 295cm2, respectively. The distance 1.599 49.53 1.8 9.3 9.5 b×etween these two×chambers is fixed and equal to 38.0 1.865 49.53 1.6 8.2 8.4 cm, center to center. ThegasmixtureusedintheFPPchambersis62%/38 To reduce the cost of electronics, the signal output % argon/ethaneby volume. The straw chambersareop- of the individual straws is multiplexed in sets of eight. erated at a high voltage of 1875 V. The drift velocity With multiplexing the maximum rate each tube can ac- of electrons for this gas mixture and for this voltage is cept safely is 100 kHz. One end of each straw is con- about 50 µm/ns over almost the entire volume of the nected to the high-voltage distribution board, and the tube. The efficiency of an individual straw tube for sin- other to a readout board. The Rutgers University elec- gle track events is greater than 97 % after correction for tronics shop designed and built readout boards with 16 the small gap between them. parallelchannels. Theinputsignalfromeachstraw(typ- 8 section D describes the FPP calibration. TABLEII:Theoutputpulsewidthsfortheeightchannelsof themultiplexing circuitry on the chamberread out cards. Channel 1 2 3 4 5 6 7 8 A. Kinematic Settings and Selection of Good Width (ns) 25 45 35 55 85-90 65 100-105 75 Events This experiment was performed at ten different values ofQ2;theseQ2 valuesaswellasotherusefulkinematical ically10mVinsize)iscoupledtogroundthrougha1500 quantities are given in Table III. pF capacitor and fed into the input of an NEC1663 am- The Hall A analysis program ESPACE calculates the plifier. The amplifier output, a 100 mV positive signal, position and angle at the focal plane, xfp,yfp,θfp,φfp, is fed into a LeCroy MVL407 quad comparator. This foreacheventfromtherawVDCdata. Theposition,an- is a leading edge discriminator that gives a logical true glesandmomentumforprotonandelectronatthetarget, whentheinputsignalexceedsasuppliedpositivethresh- y,θ,φ,δ, are then calculated using the HRS optics ma- old voltage. The output of the comparator is then fed trix; δ is the relative momentum δ=p−pc, with p and p into pulse shaping circuitry. The readout board is di- pc c beingthe scatteredparticle’smomentumandthe central vided into two halves, each of eight channels. The shap- momentum of the spectrometer, respectively. ingcircuitryfortheeightchannelsgivesadifferentwidth logic pulse for each of the channels. This allows all eight to be multiplexed together with OR chips into a single TABLE III: Beam energies and spectrometer settings of the outputchannel,reducingthenumberofcablesandchan- experiment. E is the beam energy, Q2 is the four mo- beam nels of TDC needed. To limit potential noise problems, mentum transferred square, pec, θec, ppc and θpc are central ground planes are inserted within chamber readout card values of momentum and angle for the spectrometers detect- stacks, and differential output signals of amplitude 0.1 ing electrons and protons, respectively. V are generated. Level shifter boards located away from the chambers, near the TDCs, convert these signals to Ebeam Q2 pec θec ppc θpc usual ECL levels for input to the TDCs. (GeV) (GeV2) (GeV/c) (deg) (GeV/c) (deg) The output pulse widths for eight channels, generally 0.934 0.50 0.675 52.59 0.756 45.28 adjustedto1-2%ofthewidth,aregiveninTableII.The 0.934 0.80 0.516 79.81 0.991 30.84 identification of a wire within a group of 8 is obtained 1.821 1.20 1.193 43.45 1.268 40.36 by decoding the information from pipeline TDC’s which 3.592 1.50 2.815 22.11 1.463 46.52 digitize both the leading and trailing edge times of the 3.592 1.80 2.656 25.01 1.649 42.92 signal. Thewiregroupisgivenbythemultiplexedoutput 4.087 1.90 3.093 22.28 1.712 43.35 carryingasignal;theactualpositionofthetrackrequires 4.087 2.17 2.950 24.40 1.872 40.68 decoding of the timing information to first identify the wire in the groupand then calculate the drift time using 4.091 2.50 2.774 27.08 2.068 37.68 driftvelocitycalibrationdata. Multipletrackswithinthe 4.091 3.00 2.507 31.29 2.357 33.59 samesubsetof8wirescannotbedecoded,unlessthehits 4.087 3.50 2.241 35.90 2.642 29.87 are separated by at least 250 ns. The final optics matrix for each spectrometer was de- termined subsequentto this experiment using the proce- III. DATA ANALYSIS dureasdescribedinRef. [40]. Figs. 4and5showacom- parison between the distributions of the target variables ThedatawereanalyzedwiththestandardHallAanal- y,φ,δ,θ obtainedfromthe dataandcalculatedusing the ysis program called ESPACE [39]. The output from ES- Monte Carlo programMCEEP [41], for proton and elec- PACE includes histograms, two-dimensional plots and tron at Q2 of 3.5 GeV2, respectively. The MCEEP re- multi-dimensional arrays (ntuples), which are used in sultsarenormalizedbyafactorofabout0.85. Thisvalue further analysis to obtain quantities of interest, the seems quite reasonable,as in this experiment the trigger GEp/GMp ratio and the analyzing power Ay. efficiency and the effect ofboiling ontargetdensity were The kinematic settings of this experiment, the cuts neither measured nor considered explicitly in the simu- applied in ESPACE, the selection of elastic events, the lations. In addition, the BCM was not calibrated, as it reconstruction of tracks in the FPP chambers, and the would have been in the case of an absolute cross section cuts on FPP variables are described in subsection A. A measurement, for example. The agreement between the descriptionoftheazimuthaleventdistributionandasym- data and MCEEP results is good. Cuts were applied to metry after scattering in the carbon analyzer, and spin all events for θtar ( 65 mrad), φtar ( 32 mrad), ytar ± ± transportthroughmagneticelementsofthehadronHRS, ( 6.5 cm) and δ ( 5 %), to eliminate the events seen in ± ± is given in subsection B. The methods to calculate the the tails of Figs. 4 and 5. G /G ratio are described in subsection C, and sub- The experimental event rates for each Q2 and the one Ep Mp 9 calculated with the Monte Carlo MCEEP are given in the Table IV. The actual event rates are always lower 30000 than the MCEEP rates by about 15 %, indicating that 20000 there is no significant background. The large difference s15000 s seen between MCEEP and experimental event rate at t t20000 un un Q2 of 0.8 GeV2 is due to a physical aperture cutting o10000 o C C acceptance for the open collimator in the electron arm. 10000 5000 Table IV also includes the total number of good events, averagecurrent, computer dead time, and averagebeam 0 -5 0 5 0 -40 -20 0 20 40 polarization value for each Q2 value. Y (cm) f (mrad) 20000 TABLE IV:Experimental conditions. 15000 s s15000 Q2 Expt. MCEEP total no. average dead average beam t t n n u10000 u rate rate of events current time polarization o o10000 C C (GeV2) Hz Hz (µA) (%) 5000 5000 0.50 1050 1120 2.0 × 106 4 21 0.560±0.030 0 0 0.80 250 420 4.6 × 106 10 1 0.544±0.006 -0.05-0.025 0 0.0250.05 -100 -50 0 50 100 d q (mrad) 1.20 1100 1310 1.6 × 107 24 14 0.497±0.012 1.50 420 480 9.2 × 106 10 4 0.483±0.021 1.80 330 360 1.7 × 107 13 3 0.611±0.020 FIG.4: Comparisonbetweenthetargetvariablesy,φ,δ,θob- tained from the data (dots) and calculated using the Monte 1.90 1110 1160 5.6 × 107 63 24 0.391±0.002 CarloprogramMCEEP(solidline)forprotonsdetectedinthe 2.17 830 970 3.8 × 107 78 20 0.385±0.002 right HRS. The MCEEP results are normalized by a factor 2.50 540 650 5.6 × 107 74 4 0.390±0.008 of about 0.85, agreement between the data and the MCEEP 3.00 300 370 3.2 × 107 88 1 0.395±0.002 results is good. 3.50 140 160 2.0 × 107 94 1 0.384±0.007 Selection of elastic ep events was accomplished by im- plementing a correlated cut on the missing energy E m and missing momentum p . Due to the kinematic con- m straints of ep elastic scattering, no further cuts were 15000 20000 needed to remove background events. The missing en- s s nt nt15000 ergy Em is defined as: u10000 u Co Co10000 Em =Ebeam+Mp (Ee+Ep) (8) 5000 − 5000 where E is the scattered proton energy, and M is the p p 0 0 mass of the proton. From conservation of momentum, -5 0 5 -40 -20 0 20 40 Y (cm) f (mrad) the missing momentum, P , is defined as: m 15000 P = P2 +P2 +P2 (9) 15000 m mx my mz unts10000 unts10000 Pmx = pqe·sinθe−pp·sinθp (10) Co Co Pmy = pe·sinφe−pp·sinφp (11) 5000 5000 Pmz = Pbeam−(pe·cosθe+pp·cosθp) (12) 0 0 where pe (pp) is the scattered electron (proton) momen- -0.05-0.025 0 0.0250.05 -100 -50 0 50 100 d q (mrad) tum, θe (θp) is the scattered electron (proton) Cartesian angleinthehorizontalplane,andφ (φ )isthescattered e p electron(proton)Cartesianangleintheverticalplane. In FIG.5: Comparisonbetweenthetargetvariablesy,φ,δ,θob- Fig.6,ahistogramofthemissingenergyE andmissing m tained from the data (dots) and calculated using the Monte momentumP isshown. ThereisapeakatE =0and Carlo program MCEEP (solid line) for electrons detected in m m a peak at about P = 10MeV/c. The peak in the P the left HRS. The MCEEP results are normalized by a fac- m m histogram is not centered at zero because P is defined tor of about 0.85, the agreement between the data and the m positive, and it has a larger width due to finite angular MCEEP results is good. resolution in both the transverse ( 2.0 mrad) and dis- ± persive ( 6.0 mrad) directions. A radiative tail is seen ± 10 Todetermineatrackforthefront(back)chambers,there mustbeatleastonehitinCH1andCH2(CH3andCH4) 30000 andatleastthreehitstotalinthefront(back)chambers. 25000 The efficiency for an individual straw to detect a proton ts20000 isabout97%;thetotalnumberofhitsinthefrontcham- n u15000 bers is usually about 5 or 6, and it is 4 or 5 for the rear o C chambers(astheX-planeisnotused). Thetotalnumber 10000 ofpossiblefront(back)tracksisthenumberofclustersin 5000 CH1 (CH3) times the number of clusters in CH2 (CH4). 0 0 25 50 75 100 125 150 175 The straws have cylindrical symmetry so from a single E (MeV) m hit it is impossible to distinguish if the proton passed to the left orrightofthe centerwire (left/rightambiguity). 12000 Foreachpossibletrack,aleast-squaresstraightline fitis 10000 s done for all possible combinations of left/right for each t8000 n hit. Then, out of all possible tracks, the one with the u o6000 smallest χ2 is selected as the good track. The only ex- C 4000 ception is when the track with the smallest χ2 has only 2000 three hits; then if there is another possible track with 0 more hits, it is selected as the good track if its χ2 is 0 25 50 75 100 125 150 175 P (MeV/c) reasonable. m The FPP chambers are aligned by using a software FIG.6: Histogram ofthemissingenergyandmomentumEm procedure. Thecarboncanbemovedouttomakeaclear andPm,respectively. ThepeaksatEm=0andataboutPm path between the front and back FPP chambers. Then =10MeV/ccontaintheelasticepevents. ThepeakinPmhas the tracks in the front and back chambers are aligned to a larger width than Em because of finiteangular resolution. the tracks in the VDC so that the FPP has the same coordinate system as the VDC. This is done for each FPP chamber by adjusting the three positions: z which is distance from the VDC and the zero of the u and v intheE histogramuptoabout125MeV.Thetailseen m axes,andadjustingthe threeanglesoftheFPPchamber inthe P histogramincludes the radiativetailaswellas m (angleoftheuzplane,angleofthevzplaneandtheangle events that are multiple scattered in windows. of the uv plane). The actual alignment of the chambers InHallA, twomethods ofmeasuringthe beam energy was good, since in software the angles of the chambers are now available, but at the time of this experiment are adjusted by less than a degree. neither was operational. The beam energy can be de- termined using the ep elastic kinematics, either fromthe A good track in the front and back FPP chambers is measuredscatteringangles of the electronandprotonor required to determine the polar angle (ϑ) and azimuthal from Eq. (8) by forcing Em = 0. Subsequent to this angle(ϕ) ofthe scatteredprotoninthe carbonanalyzer. experiment, the beam energy was measuredto a relative Next, we calculate the distance of closest approach be- precisionofabout1 10−3 andthecentralmomentumof tween the tracks from the front chambers and the back × the spectrometer was determined to the same precision; chambers,andatwhatdistance,Z ,fromtheVDCthe fpp usingthiscentralmomentumofthespectrometerandEq. closest approach occurred. In Fig. 7, Z is plotted for fpp (8), the beam energies were determined and are listed in Q2 =3.5GeV2. Thetotalthicknessofcarbonis49.5cm; Table III. Comparison to the beam energies determined it consists of 4 successive blocks which are separated by from the scattering angles lead to the conservative con- 1.6 cm, so one can see a peak for each block in Fig. 7. clusion that the beam energy was known with a relative ∼Chamber3islocatedatZ ofabout395cm,asseenby fpp precision of 2 10−3. theslightbumpinFig.7fromscatteringinthischamber. × Once an event was identified as an elastic ep scatter- For a selection of good events, a cut is placed on Z fpp ing, the next step was to search for a good track in the depending on the thickness of carbon used for the given front and back FPP drift tube chambers. The analysis proton momentum. partfor the FPPwasincorporatedinthe mainESPACE program by the FPP group [37]; this part of the pro- To reduce the false asymmetries a cone-test is applied gram reconstructs position and angles, in the front and for each event. An event passes the cone-test when the back FPP chambers, then calculates the polar and az- track that hit the back chambers at the measured polar imuthal scattering angles,ϑ and ϕ, respectively, and the angle would have hit the chambers for any possible az- positionof the interactionpoint, Z , in the carbonan- imuthal angle. In Fig. 8, the polar angle (ϑ) is plotted fpp alyzer. Tracking in the chambers is done in the u and v versusZ with the cone-testapplied. As expected, the fpp coordinates separately. Tracking starts with identifying rangeofacceptedpolaranglesincreasesastheinteraction clusters of hits in chambers CH1, CH2, CH3 and CH4. point in the carbon is closer to the back chambers.

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.