Experimental results on GPDs

Transcription

1 Experimental results on GPDs e e g Silvia Niccolai, IPN Orsay for the CLAS Collaboration HIX2014, 20/11/2014, Frascati (Italia) N N GPDs Interest of GPDs GPDs and DVCS Recent DVCS results from Jefferson Lab The JLab 12 GeV upgrade Future experiments on DVCS

2 Electron-proton scattering to study nucleon structure ~ ~ GPDs: H, E, H, E Fully correlated quark distributions in both coordinate and momentum space Form factors: transverse quark distribution in coordinate space Parton distributions: longitudinal quark distribution in momentum space H(ξ,t)dx = F 1 (t) ( ξ) E(ξ,t)dx = F 2 (t) ( ξ) Accessible in hard exclusive processes x H(0,0) = q(x), ~ H(0,0) = Δq(x)

3 Deeply Virtual Compton Scattering and GPDs e g* (Q 2 ) x+ξ e t x-ξ ~ ~ H, H, E, E (ξ,t) p p g Q 2 = - (e-e ) 2 x B = Q 2 /2M =E e -E e x+ξ, x-ξ longitudinal momentum fractions t = (p-p ) 2 x x B /(2-x B ) At leading order, leading twist, chiral-even, quark sector 4 GPDs for each quark flavor «Handbag» factorization valid in the Bjorken regime: high Q 2, (fixed x B ), t<<q 2 Vector: H (ξ,t) Tensor: E (ξ,t) conserve nucleon spin flip nucleon spin ~ Axial-Vector: H (ξ,t) ~ Pseudoscalar: E (ξ,t) Quark angular momentum (Ji s sum rule) = 1 1 J q - J G = 2 xdx [ H q ( x x, ) + E x x, 0 ] q, 0 (, ) X. Ji, Phy.Rev.Lett.78,610(1997) «3D» quark/gluon imaging of the nucleon

4 Transverse momentum Impact parameter Longitudinal momentum Interplay between parton distributions and reaction channels DVCS ES GTMDs GPDs FFs Thanks to C. Lorcé TMDs PDFs SIDIS DIS

5 Accessing GPDs through DVCS T DVCS GPD( t) GPD( t) ~ dx + ~ P dx i GPD( x, t) + x x + i x x -1 A = s~ T DVCS +T BH 2 s = s + - s - I(DVCS BH) s 2s I(DVCS BH) BH 2 + DVCS 2 +I s Real and imaginary parts of Compton Form Factors s, DSA Only x and t are accessible experimentally

6 Sensitivity to GPDs of DVCS spin observables ReH ImH q q = 2 e q P = 2 q q q 1 1 H ( t) - H (- t) ) + dx [ H (,, t) H (,, t) ] q q x x - x x e - x - x x + x e e g leptonic plane hadronic plane N f x= x B /(2-x B ) k=-t/4m 2 Polarized beam, unpolarized target: s LU ~ sinf Im{F 1 H + x(f 1 +F 2 )H ~ -kf 2 E}df ~ Im{H p, H ~ p } ~ Unpolarized beam, longitudinal target: s UL ~ sinfim{f 1 H+x(F 1 +F 2 )(H + x B /2E) xkf 2 E+ }df ~ Polarized beam, longitudinal target: s LL ~ (A+Bcosf)Re{F 1 H+x(F 1 +F 2 )(H + x B /2E) }df Unpolarized beam, transverse target: s UT ~ sinfim{k(f 2 H F 1 E) +.. }df Proton Neutron ~ ~ Im{H p, H p, E p } Im{H n, H n, E n } Im{H p, E p } Im{H n, E n, E n } ~ ~ Re{H p, H p } ~ Im{H n } Re{H n, E n, E n }

7 DVCS experiments worldwide DESY HERMES p-dvcs BSA,BCA, ttsa,ltsa,dsa H1/ZEUS p-dvcs X-sec,BCA CERN COMPASS p-dvcs X-sec,BSA,BCA, ttsa,ltsa,dsa JLAB Hall A p,n-dvcs (Bpol.) X-sec Hall B p-dvcs BSA,lTSA,DSA,X-sec

8 GeV Continuous Electron Beam Accelerator Facility Hall B: CLAS Large acceptance Suited for multi-particle final states L~10 34 Hall A: 2 HRS High resolution High luminosity (L~10 37 ) Limited coverage A B C E max ~ 6.0 GeV I max ~200 ma Polarization 85% 3 x 499 MHz operation Simultaneous Hall C (SOS/HMS) delivery to 3 halls Shutdown in May 2012

9 CLAS: first pioneering results NON-DEDICATED experiments ep epx S. Stepanyan et al., PRL 87 (2001) S. Chen et al., PRL 97 (2006) ep epg E e =4.3 GeV E e =5.75 GeV Q 2 =1.25 GeV 2 <x B > = 0.19 <-t> = 0.19 GeV 2 <Q 2 > = 1.82 GeV 2 <x B > = 0.16 <-t> = 0.31 GeV 2 BSA ~ Im{H p, H ~ TSA ~ Im{H p,e p } p, H~ p } A(f) = asinf + bsin2f b/a << 1 twist-2 (handbag) dominance BSA, BCA and TSA (T & L) at HERMES: handbag dominance confirmed Low statistics Limited kinematics Uncertainties on determination of epg final state

10 Hall A (JLab): first dedicated DVCS experiment ep egx Polarized cross section difference C.M. Camacho et al., PRL 97 (2006) Q 2 evolution evidence for scaling (small Q 2 range) Unpolarized cross section <Q2, -t> = (2.3,-0.28) Fit BH 2 Inteference BH-DVCS Formalism of Belitsky, Mueller, Kirchner High resolution exclusivity High luminosity precision Limited phase space

11 The CLAS e1-dvcs experiment Part 1 of the e1-dvcs experiment: Data taken in the spring 2005 Beam energy ~ 5.75 GeV Beam polarization ~ 80% Target LH 2 (More data taken in 2008, under analysis) BSA ~ Im{H p, H~ p,e p } ep epγ Added to the standard CLAS Solenoid magnet + IC (inner calorimeter): 424 lead-tungstate crystals + APD readout BSA Fit = a sinf/(1+b cosf) F.X. Girod et al., PRL 100, (2008) CLAS e1-dvcs Hall A 4.3 GeV 2 VGG(*) twist-2 VGG(*) twist-2 and 3 Regge model (Laget) (*) Vanderhaegen, Guichon, Guidal

12 CLAS e1-dvcs: DVCS cross sections Work by H.S. Jo PRELIMINARY 4 d σep epγ 4 (nb/gev ) 2 dq dx Bdtd Φ BH VGG (H only) KM10 KM10a Same amount of statistics will come from e1-dvcs2 21 Q 2 -x B bins, 9 t bins, 24 φ bins

13 CLAS e1-dvcs: DVCS cross section differences Work by H.S. Jo PRELIMINARY 4 d σep epγ 4 (nb/gev ) 2 dq dx Bdtd Φ BH VGG (H only) KM10 KM10a ~ s LU ~ sinf Im{F 1 H + x(f 1 +F 2 )H -kf 2 E}df

14 Extraction of Compton Form Factors from DVCS observables 8 CFF M. Guidal: Model-independent fit, at fixed Q 2, x B and t of DVCS observables 8 unknowns (the CFFs), non-linear problem, strong correlations Bounding the domain of variation of the CFFs with model (5xVGG) M. Guidal, Eur. Phys. J. A 37 (2008) 319

15 Extraction of CFF from CLAS DVCS pol. and unpol. cross sections Im(H p ) x B CFF fits by M. Guidal using CLAS DVCS c.s. using Hall A DVCS c.s. using CLAS DVCS BSA and old TSA VGG predictions PRELIMINARY Im(H p ), flatter t slope at high x B : faster quarks (valence) at the core of the nucleon, slower quarks (sea) at its periphery -t

16 The CLAS eg1-dvcs experiment Data taken from February to September 2009 Beam energy ~5.9 GeV CLAS + IC to detect forward photons Target: longitudinally polarized NH 3 (polarization ~80%) ep epγ Analysis by: E. Seder, U Connecticut S. Pisano, INFNFR A. Biselli, Fairfield U S. Niccolai, IPNO Also: ongoing BSA and TSA for ndvcs (D. Sokhan, Glasgow U) 5 Q 2 -x B bins 4 t bins 10 f bins

17 Beam-spin asymmetry from the CLAS eg1-dvcs data Fit: a LU sinf/(1+bcosf) a LU ~ Im{H p }

18 Beam-spin asymmetry from the CLAS eg1-dvcs data Fit: a LU sinf/(1+bcosf) a LU ~ Im{H p } KMM12: Kumericki, Mueller, Murray GK: Goloskokov, Kroll GGL: Gonzalez., Goldstein, Liuti

19 Target-spin asymmetry from the CLAS eg1-dvcs data Fit: a UL sinf/(1+bcosf) ~ a UL ~ Im{H p, H p } E. Seder et al. arxiv: , submitted to PRL

20 Target-spin asymmetry from the CLAS eg1-dvcs data Fit: A UL sinf+b sin 2f CLAS: <Q 2 >= 2.4 (GeV/c) 2, <x B >= 0.31 HERMES: <Q 2 >= (GeV/c) 2, <x B >= CLAS2006: <Q 2 >= 1.82 (GeV/c) 2, <x B >= 0.28 E. Seder et al. arxiv: , submitted to PRL Agreement with world data Improved statistics x10 at low -t Extended kinematic coverage

21 Double-spin asymmetry from the CLAS eg1-dvcs data Fit: k LL +l LL cosf)/(1+bcosf) Constant term dominated by BH

22 Double-spin asymmetry from the CLAS eg1-dvcs data Fit: k LL +l LL cosf)/(1+bcosf) ~ l LL ~ Re{H p, H p } cosf term small

23 Extraction of CFF from DVCS TSA, BSA, DSA (CLAS eg1-dvcs) ImH has steeper t-slope than ~ ImH : is axial charge more concentrated than the electromagnetic charge? Some sensitivity to ReH, ~ ~ ReE but with big uncertainties CFFs fitting code by M. Guidal [12] CLAS BSA (Girod et al.) [14] CLAS TSA (Chen et al.)

24 DVCS on nuclei: the CLAS eg6 experiment A 4 Data taken in the fall 2009 Setup: CLAS+IC+RTPC+ 4 He target Beam energy ~6.065 GeV Goals: coherent and incoherent DVCS Nuclear GPDs, EMC effect He LU ( ) = a ( ) F 1 2 A ( t) + a ( ) F Work by M. Hattawy, IPNO 2 A a ( 0 ( t) e ) FA( t) m[ H A] 2 [ H ] + a ( ) e[ H ] + a ( ) m[ H ] 2 A 3 A 3 Radial Time Projection Chamber 4 He spin 0 nucleus, at twist-2 only one CFF in DVCS BSA A e 4 He e 4 He γ e 4 He epγx <-t>= 0.11 GeV 2 <x B >= 0.18 <Q 2 >= 1.5 GeV 2 <-t>= 0.7 GeV 2 <x B >= 0.27 <Q 2 >= 2.22 GeV 2

25 Upgraded PbF 2 calorimeter DVCS on the proton in Hall A Results from E C. Munoz-Camacho et. al., PRL 97, (2006) Well described by leading-twist contribution Beam-energy separation at constant Q², x B and t: experiment E07-007

26 E07-007: Rosenbluth-like separation of DVCS DIS cross-section vs run number ±5% around expected DIS c.s. Electron acceptance checks using DIS data Working on final global normalizations (multi-tracks, dead-time, radiative corrections) M. Defurne, CEA Saclay missing mass squared Data taken in late 2010 Release of preliminary results expected in the next few months 3 kinematic points: x B =0.36, Q 2 =1.5, 1.75, 2.0 GeV 2 5 bins in t 2 beam energies per point Analysis underway (A. Marti, IPNO & Valencia U)

27 DVCS on the neutron in Hall A M. Mazouz et al., PRL 99 (2007) F. Cano, B. Pire, Eur. Phys. J. A19 (2004) 423 E08-025: Beam-energy «Rosenbluth» separation off the neutron using an LD2 target Important contribution of DVCS ² term expected in unpolarized cross section S. Ahmad et al., PR D75 (2007) VGG, PR D60 (1999) Data taken in fall 2010 concurrently with E Q²=1.75 GeV², x B =0.36 E08-025, analysis ongoing: Calibrations completed Exclusive data isolated Cross section extraction underway C. Desnault, IPNO

28 E= 2.2, 4.4, 6.6, 8.8, 11 GeV Beam polarization P e > 80% JLab upgrade to 12 GeV Add new hall 12 GeV Continuous Electron Beam Accelerator Facility CHL-2 Upgrade of the instrumentation of the 3 existing Halls

29 Kinematic coverage for the 12-GeV upgrade H1, ZEUS H1, ZEUS Sea/gluon region Valence region Study of high x B domain requires high luminosity COMPASS HERMES 11 GeV The 12 GeV Upgrade is well matched to studies in the valence-quark regime 0.7

30 New capabilities in Halls A, B & C A B High Resolution Spectrometer (HRS) pair and specialized large installation experiments C CLAS12: large acceptance, high luminosity DVCS experiments at 11 GeV have been approved for each of these three halls. Complementary programs: different kinematic coverage different precisions/resolutions focus on different observables Super High Momentum Spectrometer (SHMS) at high luminosity and forward angles

31 E : DVCS at 11 GeV in Hall A Absolute cross-section measurements Test of scaling: Q 2 dependence of ds at fixed x Bj Increased kinematical coverage JLab12 with 3, 4, 5 pass beam (6.6, 8.8, 11.0 GeV) L-HRS Scattering chamber PbF2 calorimeter 1 st experiment to run after the 12-GeV upgrade Start in 2014 for 1 year of data taking 6 GeV

32 DVCS BSA and TSA with CLAS12 & 11 GeV beam 85 days of beam time P beam = 85% L = cm 2 s 1 1< Q 2 < 10 GeV <x B < 0.65 t min < t < 2.5 GeV 2 Statistical error: 1% to 10% on sinf moments Systematic uncertainties: ~6-8% 120 days of beam time P beam = 85%, P target = 80% L = cm 2 s 1 1< Q 2 < 10 GeV <x B < 0.65 t min < t < 2.5 GeV 2 Statistical error: 2% to 15% on sinf moments Systematic uncertainties: ~6-8%

33 DVCS BSA and TSA with CLAS12 & 11 GeV beam 85 days of beam time P beam = 85% L = cm 2 s 1 1< Q 2 < 10 GeV <x B < 0.65 t min < t < 2.5 GeV 2 Statistical error: 1% to 10% on sinf moments Systematic uncertainties: ~6-8% Impact of CLAS12 DVCS-BSA data on modelindependent fit to extract Im(H)

34 CLAS12: p-dvcs transverse target-spin asymmetry 100 days of beam time Beam pol. = 80% ; target pol. (HDIce) = 60% ; Luminosity = 5x10 33 cm -2 s -1 1< Q 2 < 10 GeV 2, 0.06 <x B < 0.66, t min < t < 1.5 GeV 2 Transverse-target spin asymmetry for p-dvcs is highly sensitive to the u-quark contributions to proton spin. Proposal conditionally approved by PAC39

35 Work by M. Guidal Projections for CLAS12 for H Im

36 From CFFs to spatial densities How to go from momentum coordinates (t) to space-time coordinates (b)? (M. Guidal, H. Moutarde, M. Vanderhagen, Rept.Prog.Phys. 76 (2013) ) Applying a model-dependent deskewing factor: Fit By Burkardt (2000) Projections for CLAS12

37 -0.2 BSA 0.2 ( H, E) ( H, E) u d ( t) = 9 15 ( t) = 9 15 ed e(p)ng CLAS12 + Forward Calorimeter + Neutron Detector 80 days of data taking L = cm 2 s 1 /nucleon BSA for DVCS on the neutron with CLAS12 ~ [ 4 H, E) p( t) - H, E) n( t) ] [ 4 H, E) ( t) - H, E) ( t) ] n p -t s LU ~ sinf Im{F 1 H + x(f 1 +F 2 )H -kf 2 E}df The most sensitive observable to the GPD E First-time measurement Final phases of construction at IPN Orsay

38 -0.2 BSA 0.2 ( H, E) ( H, E) u d ( t) = 9 15 ( t) = 9 15 ed e(p)ng CLAS12 + Forward Calorimeter + Neutron Detector 80 days of data taking L = cm 2 s 1 /nucleon J u =.3, J d =.1 J u =.1, J d =.1 BSA for DVCS on the neutron with CLAS12 ~ Model predictions (VGG) for different values of quarks angular momentum [ 4 H, E) p( t) - H, E) n( t) ] [ 4 H, E) ( t) - H, E) ( t) ] n p -t s LU ~ sinf Im{F 1 H + x(f 1 +F 2 )H -kf 2 E}df The most sensitive observable to the GPD E First-time measurement J u =.3, J d =.3 J u =.3, J d =-.1 f

39 Timelike Compton Scattering with CLAS12 TCS: sensitivity to the real part of CFFs Exploratory measurement with GeV (R. Paremuzyan, IPNO & Yerevan) Projections for 120 days of running with CLAS12

40 E : DVCS at 11 GeV in Hall C Energy separation of the DVCS cross section Higher Q²: measurement of higher twist contributions Low-x B extension (thanks to sweeping magnet) Sweeping magnet 1116-block PbWO 4 calorimeter Hall C HMS Tentative running: ~

41 Summary GPDs are a unique tool to explore the internal landscape of the nucleon: 3D quark/gluon imaging of the nucleon orbital angular momentum carried by quarks Their extraction from experimental data is very difficult: there are 4 GPDs for each quark flavor they depend on 3 variables, only two ( t) experimentally accessible they appear as integrals in cross sections Recently-developed fitting methods allow to extract CFFs from DVCS observables We need to measure several p-dvcs and n-dvcs observables over a wide phase space A wealth of new results on various DVCS observables is coming from recent CLAS and Hall-A experiments (on the proton, deuterium and 4 He targets) The 12-GeV-upgraded JLab will be the only facility to perform DVCS experiments in the valence region, for Q 2 up to 11 GeV DVCS experiments on both proton and deuterium targets (polarized and unpolarized) are planned for 3 of the 4 Halls at JLab@12 GeV