Status and prospects for GPDs studies at COMPASS

Status and prospects for GPDs studies at COMPASS Etienne Burtin, CEA/Saclay, DAPNIA/SPh. N on behalf of the COMPASS collaboration 1 - Physics impact 2 - Experimental issues 3 - Recoil detector prototype GPD 2006 June 7, 2006 - Trento, Italia

Generalized Parton Distributions * , p, r hard x+x soft Factorisation: Q 2 large, -t<1 Ge. V 2 x-x GPDs P t P’ Generalized Parton Distributions for quarks : 4 functions H(x, x, t) H(x, 0, 0) = q(x) measured in DIS F(t) measured in elastic scattering

DVCS observables Deep VCS High energy beam Cross section Bethe-Heitler Lower energy => use interference - holography Single Spin Asymmetry Polarised beam COMPASS muon beam can do all ! Beam Charge Asymmetry +/- charged beam

μ DVCS+ Bethe Heitler p The high energy muon beam at COMPASS allows to play with the relative contributions DVCS-BH which depend on 1/y = 2 mp Eℓ x. Bj /Q 2 Higher energy: DVCS>>BH DVCS Cross section Smaller energy: DVCS~BH ÞInterference term will provide the DVCS amplitude μ μ’ * φ p θ μ p BH calculable

Polarized μ+ and μ- beams Polarized beam: Ep=110 Ge. V → Em=100 Ge. V P(m+) = -0. 8 2. 108 m/spill (5 s) Same collimator settings : P(m-) = +0. 8 2. 108 m/spill (5 s) beam profile unchanged Switch in 10 mins: could be done every 8 h T 6 primary Be target Collimators 12 34 HVHV scrapers Compass target Be absorbers Protons 400 Ge. V Muon section 400 m Hadron decay section 600 m 1. 3 1013 protons/spill 2. 108 m/spill

Extraction of GPDs in the case of μ+ / μ- t, ξ~x. Bj/2 fixed dσ(μp μp ) = dσBH + dσDVCSunpol + Pμ dσDVCSpol + eμ a. BH Re ADVCS cos nφ μ μ’ * φ p Pμ+=-0. 8 Pμ-=+0. 8 θ + eμ Pμ a. BH Im ADVCS sin nφ

Kinematical domain E=190, 100 Ge. V Ix 2 Collider : H 1 & ZEUS 0. 0001<x<0. 01 Fixed target : JLAB 6 -11 Ge. V SSA, BCA? HERMES 27 Ge. V SSA, BCA COMPASS could provide data on : Cross section (190 Ge. V) BCA (100 Ge. V) Wide Q 2 and xbj ranges Limitations due to luminosity new LINAC 4 (SPS injection) in 2010 Radioprotection limits needs better shielding

Sensitivity of BCA to models BCA Model 1: H(x, ξ, t) ~ q(x) F(t) model 1 model 2 COMPASS Model 2: from Goeke, Polyakov and Vanderhaeghen t <b 2> H(x, 0, t) = q(x) e = q(x) / xα’ t sensitivity to the different spatial distribution of partons when x. Bj Good sensitivity to models in COMPASS x. Bj range

Projected errors of a possible DVCS experiment Beam Charge Asymmetry L = 1. 3 1032 cm-2 s-1 Ebeam = 100 Ge. V 6 month data taking 25 % global efficiency 6/18 (x, Q²) data samples 3 bins in x. Bj= 0. 05, 0. 1, 0. 2 6 bins in Q 2 from 2 to 7 Ge. V 2 Model 1 : H(x, ξ, t) ~ q(x) F(t) α’t H(x, 0, t) = q(x) / x Model 2 : Good constrains for models

Hard Exclusive Meson Production (ρ, ω, …, π, η… ) Scaling predictions: meson g* L hard x + ξ x-ξ soft p GPDs t =Δ 2 1/Q 6 p’ 1/Q 4 Collins et al. (PRD 56 1997): 1. factorization applies only for g*L 2. σT << σL vector mesons ρ0 largest production ρ0 π+ π- pseudo-scalar mesons present study with COMPASS

Hard Exclusive Meson Production It comes for free with the recoil detector and existing COMPASS trackers Cross section: Vector meson production (ρ, ω, …) Pseudo-scalar production (π, η… ) H & ~ E Hρ0 = 1/ 2 (2/3 Hu + 1/3 Hd + 3/8 Hg) Hω = 1/ 2 (2/3 Hu – 1/3 Hd + 1/8 Hg) H = Single spin asymmetry ed S t a AS g i st MP s g e -1/3 H - 1/8 H inv CO e nt b n ese a C pr h ~ E/H t i w for a transverse polarized target ta a d

Compass Set-up 2002 -2003 at CERN 250 physicists 26 institutes magnets muon filter Calorimeters 160 Ge. V pol. m beam RICH polarized target ~ 200 detection planes Silicon, Sci. Fi, Micromegas, Drift chambers, GEM, Straw chambers, MWPC

Incoherent exclusive r 0 production in COMPASS DATA Mpp m * N m’ r 0 pp+ f k o ew s tal i v Pre dacz’ an S A. N’ 2002 Emiss p t² Impact on GPD : s. L is dominant at high Q² (factorisation only valid for s. L)

Additionnal equipment to the COMPASS setup required for DVCS all COMPASS trackers: Sci. Fi, Si, μΩ, Gem, DC, Straw, MWPC 2. 5 m Liquid H 2 target to be designed and built μ’ ECAL 1 or 2 12° COMPASS equipment with additional calorimetry at large angle μ p’ Recoil detector to insure exclusivity to be designed and built

Recoil detector design Goals: Detect protons of 250 -750 Me. V/c t resolution => s. TOF = 200 ps exclusivity => Hermetic detector Design : 2 concentric barrels of 24 scintillators counters read at both sides European funding (127 k€) through a JRA for studies and construction of a prototype ( Bonn, Mainz, Saclay, Warsaw)

Physical Background to DVCS Source : Pythia 6. 1 generated DIS events Apply DVCS-like cuts one m’, , p in DVCS range no other charged & neutral in active volumes detector requirements: 24° coverage for neutral 50 Me. V calorimeter threshold 40° for charged particles in this case DVCS is dominant

Timing and triggering issues PMT Slow protons TOF~60 ns PMT Fast protons TOF=3 ns B: 0. 7 MHz/counter PMT A: 2 MHz/counter Time window - kinematics (60 ns) - light propagation (up to 40 ns) - safety margin (30 ns) => all signal in a 128 ns window Light attenuation - ADC(PMT) = Edep e-x/l Analog trigger - hard to implement - background… (next slides ) Digital trigger - could work at 50 MHz - t>trigger latency… One solution : use the inclusive trigger and sample the signals DAQ rate : 100 k. Hz

Part I : Simulation of the Recoil Detector Goals : - evaluate front-end requirements - tune reconstruction algorithms Dedicated Géant 3. 21 program - LH 2 target with envelope - inner scint. 4 mm thick - outer scint. 5 cm thick - light attenuation - all processes turned on - custom beam halo generation - PMT waveform generation

Effect of d-rays With simulation of d-rays

Digitization : Waveform generation • For each Geant step in active volume : – record DE, z, t – propagate step to photocathode exit by applying: • • decay constant of the scintillator speed of light in the medium attenuation length light yield + quantum efficiency – fill the histogram of time of production of gelectrons (0. 1 ns bins) • Add contribution from additional muons according to intensity (2 108 m in a 5 s long spill) • Smear with a gaussian of s=3 ns to mimic the time response of the PMT now 1 ns bins (1 Gs/s) up -30 ns down +120 ns

PMT signals : only 1 m in the set-up Blue is background 1 2 3 4 7 8 9 13 14 19 20 1 2 upstream 7 PMT downstream PMT 5 6 10 11 12 15 16 17 18 21 22 23 24 3 4 5 6 INNER OUTER Red is DVCS proton 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

PMT signals : 2 8 10 m/spill (5 s) recording the waveform of all signals and segmentation are mandatory Hints for analysis: - facing counters - m m’ vertex (z, t) - DEINNER vs DEOUTER - DEOUTER vs b

Simple waveform analysis • For each PMT signal (up to 5 “hits”): – perform leading edge discrimination – correct for time-walk effect – extract pulse height and integrate charge (35 ns window) • For each up/down pair (up to 5*5=25 ”points”): – calculate time & position of crossing point using time information – unfold attenuation length to extract energy loss – request that position is within scintillator volume +/- 10 cm

Criteria for proton candidates • Have points in corresponding A and B counters Ai +1 • For each pair of “points” • Energy loss correlation • Energy loss vs bmeas correlation A Bi i DEB ( no additional beam muons in this plot – just for pedagogy ) DEA b

Coincidence with the scattered muon Use reconstructed muon vertex time to constraint proton candidates Use vertex position to evaluate the effective signal

Performances : Efficiency = number of events with proton identified number of “triggers” Seff for 1000 events trigger = one event with at least one good combination of A and B with hits identified proton = of proton candidate matches of generated proton m/5 s spill

Recoil Detector Prototype 4 m Mechanical design for the prototype 30° sector , Scale 1 of what could be the DVCS recoil detector

tests on the m beam in fall-2006 Outer Layer Inner Layer CH Target Ai+1 Bi Ai Goals for this 2 -month data taking period: - Validate detection with s. TOF ~ 250 ps Scintillators + Light guides + PMT + electronics - Measure background with equivalent CH 2 target ( ~1 MHz/detector) use 1 GHz sampler of the signals with a stand-alone DAQ and standard multihit TDCs and ADCs

Conclusions This initiative is now an “Express of Interest” : SPSC-EOI-005 http: //doc. cern. ch//archive/electronic/cern/preprints/spsc/public/spsc-2005 -007. pdf Towards a GPD experiment using COMPASS… - COMPASS is complementary to other experiments - has good sensitivity to GPD models through BCA - has good Q² range for 0. 03<xbj<0. 2 Experimental challenges - measure TOF with good resolution in high background conditions - 1 GHz sampling and recording of the PMT signals - increase beam intensity … - recoil detector prototype test this fall next step : write the proposal !

Spare

Roadmap for GPDs at COMPASS • • • 2005: Expression of interest SPSC-EOI-005 2006: Test of recoil detector prototype Proposal 2007 -2009: construction of – recoil detector – LH 2 target – ECAL 0 ≥ 2010: Study of GPDs at COMPASS In – – – parallel present COMPASS studies with polarised target Complete analysis of ρ production Other channels: , 2π … GPD E/H investigation with the transverse polarized target

Proton luminosity upgrade at CERN compass Gatignon

Front-End Electronics COMPASS DAQ : 100 k. Hz – no dead time Data flow from recoil detector : 2. 5 GByte/s • 100 k. Hz, 100 channels, 128 samples, 10 bits => 25 % increase in event size • discussion about the DC 282 module – 10 bits resolution (16 bits used for coding) – 350 ns dead time – memory depth 256 -1024 MBytes – 400 MBytes/s PCI transfer