New Proposal PR : Target Single Spin Asymmetry in SIDIS (e, e′π ± ) Reaction on a Transversely Polarized Proton Target and SoLID Hall-A Collaboration Proposal Kalyan Allada Jefferson Lab On Behalf of J. P. Chen (JLab), Haiyan Gao (Duke), Xiaomei Li (CIAE), Z-E. Meziani (Temple) for PR and Hall A Collaboration JLab PAC38 Meeting, Newport News, Aug 23 rd 2011 Kalyan Allada Jefferson Lab forHaiyan Gao (Duke), Z-E. Mezianne (Temple) J. P. Chen (JLab) Hall-A Collaboration Meeting 10 th June 2011

PAC35 Report

Leading Twist TMDs Sivers Transversity Pretzelosity Trans-Helicity All four distributions can be accessed in the proposed measurement

Transversity Collins Effect: Collins fragmentation function corresponds to the correlation between transverse spin polarization of the quark with the transverse momentum of outgoing hadron ● Transversely polarized quark generates ● left-right asymmetry during fragmentation ● Extraction of Transversity very crucial ● Valence like behavior ● Tensor charge – fundamental property, benchmark test for Lattice QCD + models ● Data: large and opposite signal for  + and  - on proton ● Small (near-zero) asymmetries on deuteron ● Need high precision data in high x region Transversely polarized quark generates left-right asymmetry during fragmentation E (n) ( PRL107, ) Collins Effect C. Schill, DIS2011 Workshop

Sivers Collins Effect: Collins fragmentation function corresponds to the correlation between transverse spin polarization of the quark with the transverse momentum of outgoing hadron ● Transversely polarized quark generates ● left-right asymmetry during fragmentation Correlation between transverse spin of nucleon with transverse momentum of the quark ● Clear non-zero signal for  + on proton ● Test an important QCD prediction: f 1T (SIDIS) = - f 1T (DY) ● Need high precision data in 4d ++  E (n), PRL107, C. Schill, DIS2011 Workshop

Pretzelosity and g 1T Collins Effect: Collins fragmentation function corresponds to the correlation between transverse spin polarization of the quark with the transverse momentum of outgoing hadron ● Transversely polarized quark generates ● left-right asymmetry during fragmentation ● Correlation between transverse momentum and transverse polarization of the nucleon ● Direct probe of relativistic effects ● Access to quark OAM (model dependent) ● No clear signal seen - need high precision data! Pretzelosity Trans-Helicity g 1T ● Polarized beam will allow to measure A LT ● A LT is related to g 1T ● g 1T ~ Re[(L=0) x (L=1)] -> spin-orbit correlation ● Data from JLab(n) and HERMES(p), COMPASS(p) HERMES proton C. Schill, DIS2011 Workshop M. Diefenthaler's Dissertation (HERMES) COMPASS proton

Hall-A E and World Data : (p, d, 3 He) ● Currently available data in SIDIS : ● Our collaboration has gained valuable experience with this measurement Moments integrated over other dimensions (z, P T ) Cover high x region Target SSA on transversely polarized targets

Proposed Experimental Goals Moving from exploratory measurements to precision measurements (in multi-dimensions) Provide ultimate precision 4d(x, z, P T, Q 2 ) mapping of target SSA in the valence quark region Flavor decomposition of Transversity, Sivers and Pretzelosity (when combined with neutron data) Extract tensor charge of both u and d-quark to better than 10% accuracy combined with 3 He SoLID data Extract leading-twist TMD, g 1T, using DSA Dominated by u-quark Sensitive to d-quark

Experiment Overview ● Measure SSA in SIDIS using transversely polarized proton target ● Use similar detector setup as that of two approved 3 He SoLID expts. ● Use JLab/UVa polarized NH 3 target with upgraded design of the magnet ● Target spin-flip every two hours with average in- beam polarization of 70% ● Two Beam energies: 11 GeV and 8.8 GeV ● Polarized luminosity with 100nA current: cm - 2 s -1 ● Beamline chicane to transport beam through 5T target magnetic field (already designed for g2p expt.) e + p ↑ e' +  +/- + X NH3target

Polarized Target JLab/UVa/SLAC polarized target, used in many different experiment (SANE, RSS etc..) What is existing: – 3cm long NH 3 target – 5 Tesla superconducting magnet – Dynamic Nuclear Polarization(DNP) using microwave pumping – NMR system for polarization measurement – Target magnet design optimized for longitudinal setting Opening of +/- 45 deg in long. direction Opening of +/- 17 deg in transverse direction What is new: – New magnet design is proposed Nominal opening of +/- 28 deg in transverse direction – Spin-flip using Adiabatic Fast Passage(AFP) technique – Improved packing fraction (with target disks) Beam Axis Magnet axis Current JLab/UVA polarized target magnet (picture courtesy: C. Keith)

Target Split-Coil Design A preliminary magnet design with transverse opening of was performed for CLAS We are proposing a new design with +/ opening in transverse direction D. Crabb (UVa) is in the process of discussing the design with two companies (Oxford and Scientific Magnets) Preliminary design done for CLAS

AFP with NH 3 ● AFP was performed on different target materials ● ● Spin-flip efficiencies (  p) for different materials are shown below (  p = pol. ratio before and after spin-flip) ● Based on previous experience AFP efficiency for NH 3 was projected to be 50% ● (P.Hautle's conference proceeding, Drell-Yan workshop) ● AFP tests for NH 3 target are being planned at UVa (D. Crabb) NIM A 356 (1995) 108

Beamline Instrumentation Beam Chicane: – Two chicane magnets to steer beam through target magnetic field – Modified BZ magnets from the accelerator beamline will be used Beam Diagnostics: – BPM, BCM, slow raster will run at low current (100nA) in Hall- A – These upgrades are being done for g2p/GEp running for Fall 2011 run We will gain experience from g2p/GEp running this Fall NH 3 target

GEANT3 simulation was performed “Sheet of flame” background: High rates in localized area of acceptance – Effect is due to the target magnetic field Solution: Turn off/remove relevant areas of the detector Low momentum particles will be swept away by the target field – First GEM plane is about 1.7m away from the target Rates on GEMs < 1 kHz/mm 2 (GEMs can handle very high rates - COMPASS expt.: 30kHz/mm 2 ) Tracking is not an issue at these rates (as demonstrated in 3 He SoLID proposal) – Rates and multiplicities are much smaller outside the “sheet of flame” (compared to 3 He SoLID proposal) due to lower luminosity Radius of GEM plane 1 Azimuthal angle(lab) Backgrounds Top Bottom

Acceptance (electron) Acceptance studies done with GEANT3 model including realistic target fields Acceptance of theta extend to lower and higher angles compared to the no target field situation Large angle forward angle

Asymmetry Asymmetry formed using luminosity normalized yields Asymmetry extraction – 2D( ɸ h, ɸ s ) fitting – Maximum Likelihood Estimation (MLE) Well established analysis methods in 6GeV E experiment Systematics reduced with spin-flip

SoLID Update Electron Identification Cerenkov Two design options: – 30 spherical mirrors would foucus the Cerenkov light produced by e- in CF 4 on CsI coated GEMs GEMs + CsI used successfully by PHENIX A prototype will be tested soon – 30 spherical mirrors and Winston cones would focus the Cerenkov light produced by e - in CO 2 on a 3x3 array of 2” PMTs PMTs from H8500 series resistant in magnetic field and suitable for tiling A prototype will be tested “in beam” environment Pion Identification Cerenkov – Same design as for option 2 above; C 4 F 10 as radiator gas – Primary choice of photon detector: PMTs (will be located further away from SoLID coil --> generally smaller magnetic fields, easier to shield)

SoLID Update for COMPASS Shashlyk modules Large angle calorimeter Calorimeter Two available options explored – Shashlyk type(preferred) / SciFi calorimeter mature mass production, concentrated light readout Initial studies of Shashlyk calorimeter done – Parameter optimizations Geant4 MC for thickness, splitting, sampling ratio, etc. – Working with IHEP (Russia) Experience with Shashlyk production Moving to next level – Fine tuning and finalizing the design, prototype testing

SoLID Update GEM Update SoLID GEM modules: 100 cm x 40cm CERN can now produce as large as 200cm x 50cm foils UVa GEM development Update: – Beam test at Mainz in Sept. with 10 cm x 10cm prototype + APV25 electronics – Construct 40 cm x 50 cm prototype this Fall and beam testing at JLab (Nov- May) – Funding received for large prototype: ~100 cm x 40 cm – Plan to fabricate in the Spring Collaboration Update – Major progress from our chinese collaboration: MoUs signed between various institution with JLab, two large NSFC grants approved. – Already have factory production ability for MRPC (RHIC-STAR ToF upgrade) and plan for large GEM detector production

Kinematics Coverage Coverage with 11 GeV beam (both forward and large angle) x B = 0.05 – 0.68 Q 2 = 1.0 – 9.0 (GeV/c) 2 P T = 0 – 1.8 GeV/c z = 0.3 – 0.7 W > 2.3 GeV Q 2 vs x W vs x W' vs x P T vs z z vs x P T vs x Transversity

Systematic Uncertainties ● Other systematics: ● Detector efficiency/acceptance/luminosity : < 2 % in each spin-pair ● Target polarization direction, random background: negligible ● For contributions from A UL, we will take dedicated data ● Systematics < statistical precision

A UT Projections Projections for one out of 48 panels in Q 2 and z Partial loss of azimuthal coverage due to “sheet of flame” background taken into account Dilution and packing fraction included  + Sivers  + Collins

A UT Projections (  + Sivers) Multi-dimensional binning in x, Q 2, p T, z (674 bins in total) Q 2 = 8 (GeV/c) 2 Q 2 = 1.0 (GeV/c) 2 z = 0.3z = 0.7

Impact u-quark Siversdistribution function A. Prokudin Region covered Impact of this measurement on u-quark Sivers distribution function

Beam Time Request We request a total of 120 days of beam time

Precision 4-d mapping of target SSA using SoLID and polarized NH 3 (p) target – Flavor separation of TMDs when combined with 3 He(n) data Extract tensor charge of u and d quark to better than 10% SoLID is an ideal device to carry out such precision measurements: – High luminosity, large acceptance and full azimuthal coverage – Will provide most precise SSA/DSA data in the kinematic region 0.05<x<0.68, 0.3<z<0.7, P T up to 1.8 GeV/c and Q 2 up to 9 (GeV/c) 2, which is crucial input to global analysis Will provide ultimate precision data on proton in the global efforts to study TMDs: – SIDIS: JLAB, HERMES, COMPASS, Future EIC – Drell-Yan: RHIC spin, Fermilab DY program, PAX Summary

Backup Slides

SoLID vs CLAS12 (Figure of Merit) FOM = N*P 2 *f 2 – N = no. of events, P = polarization, f=dilution – N SoLID (N CLAS ) = 1.56e9 (0.9e8) – P SoLID (P CLAS ) = 70% (75%) – f SoLID (f CLAS ) = 0.13 (0.33) FOM SOLID : FOM CLAS = 2.4 : 1

SoLID and CLAS12 Phase Space SoLID CLAS W > 2.3 GeVW > 2.0 GeV

Interaction between SoLID and target magnet Force on target magnet due to SoLID – SoLID designed to have low field at target ( for 3 He) – Induced field found to be negligible for g2p/GEp expt. at 1.1m away from the target (TOSCA calculations) Force on SoLID due to target field – Initial calculation using POISSON show the about 0.5 ton at 85cm from the target – Will take into consideration in supporting structure design We plan to do a detailed TOSCA calculation to know the exact forces

Tensor Charge Unmeasured region <3-4% – Will constrain by our data Higher twist <5% – Can be improved with comparison of EIC and our data Q2 evolution of transversity: <3% – EIC data can definitely help. PT dependence: <2-3% – Our data will put strong constrains Collins fragmentation 10% now (5% with future data and our data in terms of shape) NLO formalism + evolution on Collins fragmentation function (<3%)? Final determination better than 10%, current best educated guess ~7-15% level (can only determined by data) – Also will improved with additional data from EIC, e+e-, and DY data. – Currently 30-50% level, also violation of Soffer’s bound?

Event Display