1. Deeply Virtual Compton Scattering @ JLab Franck Sabatié Saclay SPIN’06 - Kyoto October 6 th 2006 From GPDs to DVCS, to GPDs back Onto the DVCS harmonic structure E00-110 experiment in Hall A Scaling tests & GPD measurement E1-DVCS experiment at CLAS in Hall B Summary

2. Collins, Freund GPDs from Theory to Experiment Theory x+  x-  t GPDs Handbag Diagram   Physical process   Experiment   Factorization theorem states: In the suitable asymptotic limit, the handbag diagram is the leading contribution to DVCS. Q 2 and large at x B and t fixed but it’s not so simple… 1. Needs to be checked !!! 2. The GPDs enter the DVCS amplitude as an integral over x: - GPDs appear in the real part through a PP integral over x - GPDs appear in the imaginary part but at the line x= 

3. Experimental observables linked to GPDs 3. Experimentally, DVCS is undistinguishable with Bethe-Heitler However, we know FF at low t and BH is fully calculable Using a polarized beam on an unpolarized target, 2 observables can be measured: At JLab energies, |T DVCS | 2 should be small Kroll, Guichon, Diehl, Pire, …

4. The cross-section difference accesses the imaginary part of DVCS and therefore GPDs at x =  The total cross-section accesses the real part of DVCS and therefore an integral of GPDs over x Observables and their relationship to GPDs

5. e -’  p e-e- ** hadronic plane leptonic plane  Into the harmonic structure of DVCS |T BH | 2 Interference term BH propagators  dependence Belitsky, Mueller, Kirchner

6. Tests of scaling 1. Twist-2 terms should dominate  and   All coefficients have Q 2 dependence which can be tested!

7. Analysis – Extraction of observables Re-stating the problem (difference of cross-section): GPD !!! What we measure

8. Special case of the asymmetry The asymmetry can be written as: Pros: easier experimentally, smaller RC Cons: - extraction of GPDs model-dependent (denominator complicated and not well known) - Large effects of the BH propagators in the denominator Asymmetries are largely used in CLAS and HERMES measurements, where acceptance and systematics are more difficult to estimate.

9. E00-110 experimental setup and performances 75% polarized 2.5uA electron beam 15cm LH2 target Left Hall A HRS with electron package 11x12 block PbF2 electromagnetic calorimeter 5x20 block plastic scintillator array 11x12 block PbF2 electromagnetic calorimeter 15cm LH2 target Left Hall A HRS with electron package 75% polarized 2.5uA electron beam Pbeam=75.32% ± 0.07% (stat) Vertex resolution 1.2mm 5x20 block plastic scintillator array  t (ns) for 9-block around predicted « DVCS » block

10. E00-110 kinematics The calorimeter is centered on the virtual photon direction 50 days of beam time in the fall 2004, at 2.5  A intensity

11. Analysis – Looking for DVCS events HRS: Cerenkov, vertex, flat-acceptance cut with R-functions Calo: 1 cluster in coincidence in the calorimeter above 1 GeV With both: subtract accidentals, build missing mass of (e,  ) system

12. Analysis –  o subtraction effect on missing mass spectrum Using  0 → 2  events in the calorimeter, the  0 contribution is subtracted bin by bin After  0 subtraction

13. Analysis – Exclusivity check using Proton Array and MC Normalized (e,p,  ) triple coincidence events Using Proton-Array, we compare the missing mass spectrum of the triple and double-coincidence events. Monte-Carlo (e,  )X – (e,p,  ) The missing mass spectrum using the Monte-Carlo gives the same position and width. Using the cut shown on the Fig.,the contamination from inelastic channels is estimated to be under 3%.

14. Analysis – Extraction of observables

15. Difference of cross-sections Corrected for real+virtual RC Corrected for efficiency Corrected for acceptance Corrected for resolution effects Twist-2 Twist-3 Extracted Twist-3 contribution small !

16. Q 2 dependence and test of scaling =0.26 GeV 2, =0.36 No Q 2 dependence: strong indication for scaling behavior and handbag dominance Twist-2 Twist-3 Twist 4+ contributions are smaller than 10%

17. Total cross-section Corrected for real+virtual RC Corrected for efficiency Corrected for acceptance Corrected for resolution effects Extracted Twist-3 contribution small !

18. E1-DVCS @ CLAS : a dedicated DVCS experiment ~50 cm Inner Calorimeter + Moller shielding solenoid Beam Polarization:75-85% Integ. Luminosity:45 fb -1 M  (GeV 2 )

19. E1-DVCS kinematical coverage and binning W 2 > 4 GeV 2 Q 2 > 1 GeV 2

20. E1-DVCS exclusive DVCS selection Remaining  0 contamination up to 20%, subtracted bin by bin using  0 events and MC estimation of  0 (1  ) to  0  (2  ) acceptance ratio 3-particle final state

21. E1-DVCS raw asymmetries Very Preliminary Integrated over t A LU = 0.18 GeV 2 = 0.30 GeV 2 = 0.49 GeV 2 = 0.76 GeV 2

22. A LU (90°) Hall A data Old CLAS data E1-DVCS corrected A LU (90°) Very Preliminary

23. Summary Cross-section difference (Hall A):  High statistics test of scaling: Strong support for twist-2 dominance  First model-independent extraction of GPD linear combination from DVCS data in the twist-3 approximation  Upper limit set on twist-4+ effects in the cross-section difference: twist>3 contribution is smaller than 10% Total cross-section (Hall A):  Bethe-Heitler is not dominant everywhere  |DVCS| 2 terms might be sizeable but almost impossible to extract using only total cross-section: e + /e - or  + /  - beams seem necessary  Despite this, we performed a measurement of 2 different GPD integrals BSA (CLAS):  Preliminary data in large kinematic range and good statistics !

24. Outlook 2 experiments to run in Hall B in ~2008 1 experiment to run in Hall A in ~2009 Extension of all the current experiments already proposed and approved for 12 GeV running Many theoretical progress expected in the meantime: - Radiative corrections (P. Guichon) - Global analysis with adequate model (D. Mueller and others) - …

25. Backup

26. Comparison with models

27. Q 2, x t,  Designing a DVCS experiment Measuring cross-sections differential in 4 variables requires:  The high precision measurement of all 4 kinematical variables Q 2, x Scattered electron detected in the Hall A HRS: High precision determination of the  * 4-vector Emitted photon detected in a high resolution Electromagnetic Calorimeter: High precision determination of the real photon direction t, 

28. Designing a DVCS experiment Measuring cross-sections differential in 4 variables requires: Designing a DVCS experiment Measuring cross-sections differential in 4 variables requires:  A good knowledge of the acceptance Scattered electron The HRS acceptance is well known Emitted photon The calorimeter has a simple rectangular acceptance e p → e (p)  Perfect acceptance matching by design ! Virtual photon « acceptance » placed at center of calorimeter R-function cut **  Simply: t: radius  : phase

29. Measuring cross-sections differential in 4 variables requires:  Good identification of the experimental process, i.e. exclusivity Designing a DVCS experiment Without experimental resolution

30. Designing a DVCS experiment  Good identification of the experimental process, i.e. exclusivity Measuring cross-sections differential in 4 variables requires: Without experimental resolution

31. resonant or not Designing a DVCS experiment  Good identification of the experimental process, i.e. exclusivity Measuring cross-sections differential in 4 variables requires: Without experimental resolution If the Missing Mass resolution is good enough, with a tight cut, one get rids of the associated pion channels, but  o electroproduction needs to be subtracted no matter what.

32. The baby steps towards the full nucleon wave function After understanding the basic properties of the nucleon, physicists tried to understand its structure: -By Elastic Scattering, we discovered the proton is not a point-like particle and we infered its charge and current distributions by measuring the Form Factors F 1 and F 2. -By Deep Inelastic Scattering, we discovered quarks inside the nucleon and after 30 years of research, have a rather complete mapping of the Quark Momentum and Spin Distributions q(x),  q(x). Since the late 90’s, a new tool was developed, linking these representations of current/charge and momentum/spin distributions inside the nucleon, offering correlation information between different states of the nucleon in terms of partons. The study of Generalized Parton Distributions through Deep Exclusive Scattering will allow for a more complete description of the nucleon than ever before. Mueller, Radyushkin, Ji

33. E00-110 custom electronics and DAQ scheme 1. Electron trigger starts the game 2. Calorimeter trigger (350ns): - selects clusters - does a fast energy reconstruction - gives a read-out list of the modules which enter clusters over a certain threshold - gives the signal to read-out and record all the experiment electronics channels 3. Each selected electronics channel is digitized on 128ns by ARS boards t (ns) 4. Offline, a waveform analysis allows to extract reliable information from pile-up events

34. ARS system in a high-rate environment - 5-20% of events require a 2-pulse fit - Energy resolution improved by a factor from 1.5 to 2.5 ! - Optimal timing resolution  t (ns) HRS-Calo coincidence  t =0.6 ns

35. Analysis – Calorimeter acceptance The t-acceptance of the calorimeter is complicated at high-t: 5 bins in t: -0.40-0.35-0.37 -0.35-0.30-0.33 -0.30-0.26-0.28 -0.26-0.21-0.23 -0.21-0.12-0.17 Min Max Avg X calo (cm) Y calo (cm) Calorimeter Large-t  dependence

36. Analysis –  o contamination  Symmetric decay: minimum angle in lab of 4.4° at max  o energy  Asymmetric decay: sometimes one high energy cluster… mimicks DVCS!

37. Analysis –  o subtraction using data 1.Select  o events in the calorimeter using 2 clusters in the calorimeter 2.For each  o event, randomize the decay in 2-photons and select events for which only one cluster is detected (by MC) 3.Using appropriate normalization, subtract this number to the total number of 1-cluster (e,  ) events Note: this not only suppressed  o from electroproduction but also part of the  o from associated processes Invariant Mass of 2-cluster events

38. Summary plot

39. Total cross-section and GPDs with Interesting ! Only depends on H and E

40. CLAS: Phys.Rev.Lett.87:182002, 2001 HERMES: Phys.Rev.Lett.87:182001, 2001 DVCS Results : CLAS 4.2 and 4.8 GeV and HERMES Preliminary CLAS analysis with 4.8GeV data (G. Gavalian) Preliminary

41. 1-cluster events coming from all  o 1-cluster events coming from  o electroproduction MM 2 cut M e  X (GeV 2 ) Analysis – missing mass of « subtracted »  o This method gives the number of  o for all experimental bins =-0.28 GeV 2, 100°<  <120°

42. CLAS 6 GeV DVCS Proton ep→epπ o /η Hall A 6 GeV DVCS proton neutron ep→epπ o CLAS 5.75 GeV DVCS DDVCS ΔDVCS D2VCS Polarized DVCS ep→epρ L ep→epω L ep→epπ 0 / η ep→enπ + ep→epΦ HERMES 27 GeV DVCS – BSA + BCA + nuclei d-BSA d-BCA ep→epρ σ L + DSA ep→enπ+ + …. HERA 27.5-900 GeV DVCS CLAS 4-5 GeV DVCS BSA HERMES DVCS BSA+BCA With recoil detector COMPASS DVCS  +BCA With recoil detector Published ….. Preliminary results 2004 2005 ……… ….. 2009 ? … 2010 JLab@ 12GeV Deep Exclusive experiments EVERYTHING, with more statistics than ever before

43. Deeply Exclusive Scattering program in the near future 2006-2007 HERMES 2010-… COMPASS 2010-… JLab@12GeV

44. Q 2 and t dependence with 4.8 GeV data Preliminary No clear dependence seen. Comparison with models necessary. And more accurate data clearly needed!

45. 0.15 < x B < 0.4 1.50 < Q 2 < 4.5 GeV 2 -t < 0.5 GeV 2 DVCS Beam Spin Asymmetry PRELIMINARY H. Avakian & L. Elouadrhiri GPD based predictions  0 are « suppressed » due to analysis cuts (only low t), but no subtraction or correction were done PRELIMINARY Once again, exclusivity and high statistics & precision data is the key !

46. The next generation DES experiments: a challenge We need: Resolution and exclusivity (3-particle final state) High luminosity and/or acceptance (low cross-sections) High transfers (factorization) ep  epX MAMI 850 MeV ep  epX Hall A 4 GeV ep  eγX HERMES 28 GeV N+πN Missing mass M X 2 ep  epX CLAS 4.2 GeV πoπo γ Beam energy M X 2 [MeV 2 ]

47. Dependence of  asymmetry and total cross-section as a function of x B, t, Q 2,  bins  Projected results (sample)

48. Fast Digital Trigger PbF2 block Plastic Scintillator Proton Array ARS system Fast-digitizing electronics and smart calorimeter trigger

49. Addition of a charged particle veto in front of the plastic scintillator array The veto counter consists in 2 layers of 2cm-thick scintillator paddles  Use of a deuterium target  Proton DVCS is veto-ed by new detector Neutron DVCS in Hall A at 6 GeV – E03-106 (Nov. 2004) P. Bertin, C.E. Hyde-Wright, F. Sabatié and E. Voutier Main contribution to the neutron