TRANSVERSE SPIN MEASUREMENTS INAnna Martin University of Trieste and INFN Triesteon behalf of the COMPASS CollaborationVilla Mondragone, June 14, 2006

CONTENTTHE EXPERIMENTCollins and Sivers asymmetriespositive and negative hadrons p, KTwo hadron asymmetriesall pairsleading hadrons polarimetryPHYSICS RESULTS three different ways to access transversitySUMMARY AND OUTLOOK

the COMPASS experimentfixed target experiment at the CERN SPS 240 physicists from 28 institutes, 11 Countries

approved in 1998 with a broad physics programmehadron beams p,K,p (2004) 2007, muon beam 2002-2004, 2006, . transversitygluon contribution to nucleon sping1, DS, Dq flavour decomposition.r 0 production A. SandaczSPSmuon beam:momentum 160 GeV/c longitudinal polarisation -76%intensity 2108 +/spill (4.8s/16.2s)

SM1SM2 m beamMuonWallMuonWallE/HCALE/HCALRICHPolarised TargetSciFiSiliconMicromegasGEMsStrawsSDCMWPCW45the spectrometertwo stages spectrometer Large Angle Spectrometer (SM1)Small Angle Spectrometer (SM2)tracking, calorimetry, PIDhigh energy beambroad kinematical rangelarge angular acceptance

the target system

6LiD: polarization PT = 50% dilution factor f = 0.38 relaxation time > 2000 hrs

during data taking with transverse polarizationdipole field always polarization reversal in the 2 cells after ~5 days (systematics)

two 60 cm long cellswith opposite polarization (systematics)dN/dz2000zvtx (mm)00-100010004000solid target operated in frozen spin mode

the RICH detectorfully efficient for transverse data in 2003 and 2004 threshold momenta:pp = 2 GeV/cpK = 9 GeV/cpP =17 GeV/cphoton detectors

data taking 2002-2004transversely polarised deuteron target ~ 20% of the running time

Collins and Sivers asymmetriespublished results 2002prel results 2002-2004 with RICH PID

Two hadron asymmetriesprel results 2002-2004

polarimetryprel results 2002-2003

200211 days of data taking (19),2 periods2003 9 days of data taking (14),1 period200414 days of data taking (24),2 periods

trigger (large x, Q2)DAQ, on line filter

Collins and Sivers asymmetriespositive and negative hadrons p, KTwo hadron asymmetriesall pairsleading hadrons polarimetryPHYSICS RESULTS

single hadron asymmetriesSIDIS cross-sectionh azimuthal angle of the hadron s azimuthal angle of the transverse spin of the struck quark s azimuthal angle of the transverse spin of the initial quark long pol B and/or Ttransvely polarisedtargetunpol B and TC = h - s

Collins angle

S = h - s Sivers angle

s = p - s

single hadron asymmetriesSIDIS cross-sectionh azimuthal angle of the hadron s azimuthal angle of the transverse spin of the struck quark s azimuthal angle of the transverse spin of the initial quark long pol B and/or Ttransvely polarisedtargetunpol B and Tindependent angles independent extraction of the asymmetriesCollinsSiversbeam polarizations = p - sat leading order

Collins asymmetryAColl evaluated as function of x, z and pt forLeading Hadrons and All Hadronsminimizes acceptance effectsspin independent terms cancel at 1st order

Sivers asymmetry Sivers distribution functionf target polarisation dilution factorPT target polarisationASivl evaluated as function of x, z and pt forLeading Hadrons and All Hadrons

event selection

DIS eventsQ2 > 1 GeV2W2 > 25 GeV20.1 < y < 0.9

hadronspht > 0.1 GeV/c z > 0.2 (all h)z > 0.25 (leading h)Q2 = 2.4 (GeV/c)2xBj 0.035W = 9.4 GeV/c22002 data

event selection cont2002 datapt0.51 GeV/cz 0.45

Asymmetries 2003-2004 vs 2002increase in statistics by a factor of 7!results from 2003-2004 data released in April (DIS2006)comparison with the published 2002 dataPRL 94 (2005) 202002

systematicsextensive studies, in view of the small statistical errors:false asymmetriessplitting target cells and scrambling the datastability in time of acceptance / efficiencyR-teststability of the measured asymmetriesdividing periods in time, splitting target cells and spectrometer,making bins in z and pt ,changing F binning, use of different estimators2-dimensional fitquality of the fit.

systematics (cont).consistency of the results from the different data taking periods

expected: normal standard distributionCONCLUSION:all the effects are compatible with the statistical fluctuations our systematic errors are considerably smaller than the statistical errors

Collins Asymmetries 2002-2004small errors (~1%)small asymmetriescancellation between p and n (lh and ah)

Sivers Asymmetries 2002-2004small errors (~1%)small asymmetriescancellation between p and n (lh and ah)

Collins - comparison with theoryW. Vogelsang and F. YuanPRD72 (2005) 054028

DTq from saturation of Soffer inequality 2 parameter fit to HERMES p data to extract the Collins FF unfav FF ~ - fav FF (c2/n ~ 0.8)

very small asymmetries for d target comparison with COMPASS 2002 data okCOMPASS 2002 data some of the recent phenomenological studies

Collins - comparison with theoryA.V. Efremov, K. Goeke and P. Schweitzer, PRD73 (2006) 094025

including transverse momenta DTq from chiral quark soliton model fit to HERMES data to get Collins FF unfav FF ~ - fav FF

agreement (absolute values) with analysis of BELLE DTd ~ unconstrained

comparison with COMPASS 2002 data ok

Sivers - comparison with theoryW. Vogelsang and F. YuanPRD72 (2005) 054028

2 parameter fit to HERMES p data to extract Sivers function (u and d) d ~ - 2 u due to the large p- asymmetry (c2/n ~ 1.2)

comparison with COMPASS 2002 data ok

COMPASS 2002 data

Sivers - comparison with theoryM. Anselmino et al., PRD73 (2006) 094025

including transverse momenta FF from literature 13 parameter fit to HERMES and COMPASS 2002 data to extract Sivers functions (c2/n ~ 1.06)

Collins and Sivers asymmetriespositive and negative hadrons, p, KTwo hadron asymmetriesall pairsleading hadrons polarimetryPHYSICS RESULTS

p, K Asymmetries 2003-2004same DIS event selection and hadron definition as beforeplus PID based on RICH response

additional systematic tests: RICH stability systematic errors smaller than statistical onesfinal sample

negativeallleadingpions 4.2 M2.8 Mkaons0.6 M0.4 M

positiveallleadingpions 5.3 M3.4 Mkaons0.9 M0.7 M

p, K Sivers Asymmetries 2003-2004GPD2006last weekcancellation between p and n ?

p, K Collins Asymmetries 2003-2004again, difficult to see an effect GPD2006last weekjust released

comparison with HERMES

comparison with HERMES

Collins and Sivers asymmetriespositive and negative hadrons, p, KTwo hadron asymmetriesall pairsleading hadrons polarimetryPHYSICS RESULTS

Two Hadron Asymmetrieslooking at two hadron production, a different asymmetry can be measuredFRS = fR - fsA. Bacchetta, M. Radici, hep-ph/0407345X. Artru, hep-ph/0207309

Two Hadron Asymmetries 1selection of DIS eventsQ2 >1 (GeV/c)20.9 > y > 0.1W > 5 GeVand hadrons z1, 2 > 0.1 and xf1, 2 > 0.1 z1+ z2 < 0.9 all combinations of +ve (h1) and ve (h2) hadrons (~1.3/ev)presently no RICH PIDtotal statistics 2002-2004: 6.1 M combinations

Two Hadron Asymmetries 1final sampleK0 r0

Two Hadron Asymmetries 1all combinations of +ve (h1) and ve (h2) hadronssmall errors (order of ~%)small asymmetries! cancellation?

Collins and Sivers asymmetriespositive and negative hadrons, p, KTwo hadron asymmetriesall pairsleading hadrons polarimetryPHYSICS RESULTS

different hadron pairs selections have been tried still based on the string fragmentation model

pT - leading hadron pairs h1 = +ve hadron with largest pT, h2 = -ve hadron with second largest pT and h1 = -ve hadron with largest pT, h1 = +ve hadron with second largest pTTwo Hadron Asymmetries 2 SPIN 20051 entry/event

Two Hadron Asymmetries / 2z-leading hadron pairsCIPANP06as for p and K, just released

Collins and Sivers asymmetriespositive and negative hadrons, p, KTwo hadron asymmetriesall pairsleading hadrons polarimetryPHYSICS RESULTS

L polarimetryevent selection

L polarimetry2002-2003 data TRANSVERSITY 2005

L polarimetry2002-2003 data 2004 data: analysis ongoing TRANSVERSITY 2005

SUMMARY AND OUTLOOK

summaryNew precise DIS deuteron data from COMPASS are now availableIn all the channels investigated in so far

the measured deuteron asymmetries are very small, compatible with zeroSURPRISING RESULTS FROM KAON ASYMMETRIESCollins and Sivers asymmetries +ve and -ve hadrons, p, KTwo hadron asymmetries all pairs, leading hadrons polarimetryPhenomenological studies can describe at the same time the BELLE (FF), the HERMES (proton) and 2002 COMPASS (deuteron) Collins and Sivers asymmetries Is the emerging picture satisfactory ?!!

outlook 1in the near future COMPASS will produce some new resultsfrom the present deuteron data:plustransverse effects from longitudinal dataother two asymmetries from single hadron data g2Cahn effectexclusive r production on transversely polarised target .

Two hadron asymmetriesRICH PID 2003-2004 polarimetry 2004

outlook 2in 2006 COMPASS plans to take data with a transversely polarised proton target (NH3)with 30 days running, precision ~ deuteron in 2006, new PT magnet (larger acceptance at large x)The COMPASS Collaborationis considering a new proposal (*) based on spectroscopy,GPDs, and TRANSVERSITY

SIDIS essential for flavour separation ! H. Fischer

(*) Input to CERN Council Strategy Group (Jan. 15, 2006) SPSC-EOI-005 (Jan. 18, 2005)