Generalized Parton Distributions studies @ JLab

Introduction

DVCS in JLab/Hall A

DVCS in JLab/Hall B

Comparison with models

0 electroproduction in Halls A/B

Franck SabatiéCEA Saclay

Franck SabatiéCEA Saclay

IWHSS’08 - TorinoMarch 31st, 2007

IWHSS’08 - TorinoMarch 31st, 2007

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.

Q2 and largeat xB and t fixed

but it’s not so simple…

1. Needs to be checked !!!1

1

1

1

( , , ) +

( , ,

- i + ( , )

, )

DVCS

GPD x t

GPD x tT dx

x

GPD x tdx

i

P x

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=

Experimental observables linked to GPDs

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:

42

2

4 4

2

2

m2 I

ReBH BH D

DVC

B

BH

C

S

V S

B

dT T

dx dQ dtd

d dT

dx dQ d

T

tdT

At low energy,|TDVCS|2 supposed small

42 2

2

4 42 2

2

2

Im2

Re DVCBH BH DVCS

B

BH DVCS DVCSDVCS

B

SdT T T

dx dQ dtd

d dT T T

dx dQ

T

tT

d d

Ji, Kroll, Guichon, Diehl, Pire, …

e-’

pe-*

hadronic plane

leptonic plane

0 1

42

1

2 3

1

10 1 22

2

2

1 22

4

21 2

24

2

1( , , ) cos cos 2

1 (

( ) ( )

( ) (, , ) cos cos 2 cos3

( , , )sin sin 2

)

( ) ( )

BH BH BHB

B

B

B

B

I I I I

I I

c c c c

dx Q t c c c

dx dQ dtd

x Q t

x Q td d

dxs

dQ ds

dt

Into the harmonic structure of DVCS

|TBH|2

Interference term

1 2( ) (

1

)

BH propagators dependenceBelitsky, Mueller, Kirchner

Special case of the asymmetry

4 42

4 4 0 1

1 2

0 1

sin sin 2( , , )

( ) cos ...

A B BH BH

I I

I I

d dx Q t

c cd

s s

cd c

The asymmetry can be written as:

Pro: easier experimentally, smaller Radiative corr., smaller systematics

Con: direct extraction of GPDs is model- (or hypothesis-) dependent (denominator complicated and unknown)

It was naturally the first observable extracted from non-dedicated experiments…

Potentiallymany more terms

0

20

40

60

80

100

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Year

# o

f p

aper

s

Theory GPD

Total GPD

Exp. GPD

Timeline

Invention of GPDs my Müller, who needed the theoretical object for his thesis work

Ji and Radyushkin show how the GPDs are interesting, and how to access them. Collins and Freund prove the factorization, Diehl, Pire and Gousset point out the rich harmonic structure of DVCS

First experimental paper usingH1-ZEUS non-dedicated data: Observation of DVCS at very low x

First dedicated DVCS experiment proposed and accepted by a PAC:E00-110, proton DVCS in Hall A Jefferson Lab.

HERMES and CLAS publish non-dedicated data on DVCS in the same PRL issue. They both show a large sine wave in the BSA, proving they indeed deal with the interference of DVCS and BH.

Dedicated DVCS experiment with CLAS proposed/accepted:E01-113, proton DVCS in Hall B

Intense theoretical work begings: higher twist effects, corrections, models, interpretation of various F.T., full harmonic description, …

Dedicated DVCS experiment in Hall A proposed/accepted:E03-106, neutron DVCS in Hall A

The three JLab experiments take data within 1 year (2004-2005), about 4 years after the first proposition to the JLab PAC.

Preliminary results fromJLab experiments

0

20

40

60

80

100

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Year

# o

f p

aper

s Theory GPD

Total GPD

Pentaquark

Exp. GPD

Published non-dedicated results on ALU and AUL

PRL 97, 072002 (2006)

JLab/Hall B - E1 & HERMES

JLab/Hall B - Eg1

Both results show, with a limited statistics, a sin behavior(necessary condition for handbag dominance)

In the ALU result, DD models (VGG) tend to over-estimate the data

ALU

AUL

CLAS: PRL 87, 182002 (2001) HERMES: PRL 87, 182001 (2001)

E00-110 experimental setup

• 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

50 days of beam time inthe fall 2004, at 2.5A

113294 fbLu dt

Designing the Hall A DVCS experiment

Measuring cross-sections differential in 4 variables requires:Measuring cross-sections differential in 4 variables requires:

A good knowledge of the acceptance

Scattered electronThe HRS acceptance

is well known

Emitted photonThe calorimeter has a simple

rectangular acceptance

e p → e (p)

Perfect acceptancematching by design !Virtual photon « acceptance »placed at center of calorimeter

R-functioncut

*

Simply:t: radius: phase

Analysis by missing mass (e)X

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

Analysis – o subtraction effect on missing mass spectrum

Using 0→2 events in the calorimeter,the 0 contribution is subtracted bin by bin

After0 subtraction

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,)

2 cutXM

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%.

Difference of cross-sections

2 22.3 GeV

0.36B

Q

x

Corrected for real+virtual RCCorrected for efficiencyCorrected for acceptanceCorrected for resolution effectsChecked elastic cross-section @ ~1%

Twist-2Twist-3

Extracted Twist-3contribution small !

PRL97, 262002 (2006)

Q2 dependence and test of scaling

<-t>=0.26 GeV2, <xB>=0.36

No Q2 dependence using BMK separation:strong indication for scaling behavior and handbag dominance

Twist-2Twist-3

Twist 4+ contributions are smaller than 10%

PRL97, 262002 (2006)

1 1 2 22()

2 4( )

B

I BCx t

F F Fx

F FM

Supposedly mostly H (~80%)

Total cross-section

2 22.3 GeV

0.36B

Q

x

Corrected for real+virtual RCCorrected for efficiencyCorrected for acceptanceCorrected for resolution effects

Extracted Twist-3contribution small !

PRL97, 262002 (2006)

but impossible to disentangle DVCS2

from the interference term !

DVCS on the neutron in JLab/Hall A: E03-106

LD2 target24000 fb-1

xB=0.36, Q2=1.9 GeV2

MODEL-DEPENDENTJu-Jd extraction

E1-DVCS with CLAS : a dedicated DVCS experiment in Hall B

~50 cm

Inner Calorimeter+ Moller shielding solenoid

Beam energy: ~5.8 GeVBeam Polarization: 75-85%Integ. Luminosity: 45 fb-1

Next half of data in 2008

M(GeV2)

E1-DVCS kinematical coverage and binning

W2 > 4 GeV2

Q2 > 1 GeV2

Integrated over t

E1-DVCS : Asymmetry as a function of xB and Q2

<-t> = 0.18 GeV2 <-t> = 0.30 GeV2 <-t> = 0.49 GeV2 <-t> = 0.76 GeV2

Accurate data in a large kinematical domain

E1-DVCS : ALU(90°) as a function of -t + models

JM Laget

VGG twist-2

VGG twist-2+3

Accepted by PRL

0.09<-t<0.2

0.2<-t<0.4

0.4<-t<0.6

0.6<-t<1

1<-t<1.5

1.5<-t<2

PRELIMIN

ARY,

Analysis in

Progre

ss

PhD Thesis H.S. Jo

E1-DVCS : Cross-sections over a wide kinematical range

New developments in GPD modeling : minimal dual model

Central quantities are t-dependent parton distributions q and e(x,t), obtained as →0 limits of the nucleon GPDs:

q and e are estimated using a dual representation (parton distributions as an infinite series of t-channel exchanges).

Only the leading function is kept for simplicity. Only the GPD H is kept, supposedly dominant in DVCS.

It is a simple one-parameter model with strong predictive content.If part of the data cannot be fit, it will mean higher twists/genuine non-forward effects need to be included: first step to understand DVCS data.

0 0( , ) lim ( , , ), ( , ) lim ( , , )q x t H x t e x t E x t

arXiv:0803.1271v2Polyakov, Vanderhaeghen

Minimal dual model, confrontation with data

DD model (VGG)Minimal dual model

Data: Hall A polarized cross section data at Q2=2.3 GeV2

Difference of cross sections (imaginary part of interference term)

Data are perfectly described by minimal dual model

Minimal dual model, confrontation with data (2)

Beam Spin Asymmetries(Hall B data)

Total cross sections(Hall A data)

DD model (VGG)Minimal dual model

e’

e

0

leptonic plane

hadronicplane

p’

• Q2= - (k-k’)2

• xB = Q2/2M=Ee-Ee’

• t = (p-p’)2

• (angle between leptonic and hadronic plane)

= (Q2,xB)d4

dQdxBddtd2

ddtreduced cross section for *p→p0

dTT

dt+

d2ddt

=1

2(

dT

dt

dL

dt+

dLT

dt+√ 2+1) cos cos2

+h√ 2-1)dL’T

dtsin

0 electroproduction cross section

VGGLaget

-Low-t cross-section largelyovershoots GPD & JML models !!!

-TT term is large, suggesting a large transverse component.

-But 0 cross-section ratio for proton and deuteron targets is found ~ 1

More data needed !

Submitted soon

0 cross sections in Hall A – a puzzle !

0 cross sections in Hall B

Arbitrary unitsStatistical errors only

T+L vs. –t (GeV2/c2)

S. Niccolai

PRELIMIN

ARY,

Analysis in

Progre

ss

Summary

Scaling test of DVCS cross section difference

→ DD models over-estimate it→ Minimal dual model provide a good description

Asymmetry measured in a large kinematical range,badly described by models.

Cross section measured both accurately and(soon) in a large kinematical range, also badly described by models: suggest genuine off-forward contribution and/or higher twist at work.

0 cross sections (unseparated) measured: not well described by either Regge model or GPD model: a challenge for theory !

Like Stefano said: mode data are needed, but modelling is only in its first phase and shows interesting progress recently.

JLab 6 & 12 GeV,HERMES recoil,COMPASS + upcomingTheory/Phenomenology work

Use of base CLAS12 equipment, including Inner Calorimeter (IC)

Detection of the full (e,p,) final state

Perform 2 experiments for the extraction of the BSA and the TSA

Experimental Setup and proposed experiments at 11 GeV

IC tungsten shielding

Distance, target-IC: ~1.75m currently(note: this is NOT optimized for DVCS!)

Beam Spin Asymmetry

From 1% statistical error on extracted Twist-2 coefficient

to 10% statistical errorat high xB

IC in standard position – 80 days – 10^35 Lum – VGG model

The cross-section difference accesses the imaginary part of DVCS and therefore GPDs at x =

1

1

1

1

( , , ) +

( , ,

- i + ( , )

, )

DVCS

GPD x t

GPD x tT dx

x

GPD x tdx

i

P x

The total cross-section accessesthe real part of DVCS and therefore an integral of GPDs over x

Observables and their relationship to GPDs

CLAS 6 GeV

DVCSBSAProton

ep→epπo/η

Hall A 6 GeV

DVCSprotonneutron

ep→epπo

CLAS 5.75 GeV

DVCS

DDVCS

ΔDVCS

D2VCS

PolarizedDVCS

ep→epρL

ep→epωL

ep→epπ0/η

ep→enπ+

ep→epΦ

HERMES 27 GeV

DVCS – BSA + BCA

+ nuclei BSA d-BCA

+ Polarized target DVCS

ep→epρσL + DSA

CLAS 4-5 GeV

DVCSBSA

HERMES

DVCS

BSA+BCA

With recoil detector

COMPASS

DVCS

+BCA

With recoil detector

Published or Finalized Published or Analysis in 2009 –

2001 close to be progress 2010 2010+ 2014?

JLab@

12GeV

(low/medium energy) Deep Exclusive experimentsE

VE

RY

TH

ING

, with

more statistics than ever before

JLab

Hall A

DVCS²

CLAS

BSA+TSA

In red: selected topics for JLab