COMPASS Results on Transverse Spin and Transverse Momentum

COMPASS Results on Transverse Spin and Transverse Momentum Effects Anna Martin Trieste University & INFN (Italy) on behalf of the COMPASS Collaboration DIS 2010

OUTLOOK • the COMPASS experiment • Collins asymmetry new results ! • Sivers asymmetry many topics not covered. . . • future measurement Anna Martin April 21, 2010

COMPASS a long story of measurements…. fixed target experiment at the CERN SPS broad physics programme SIDIS measurements: the spin structure of the nucleon 2002 2003 deuteron (6 Li. D) polarised target — L & T 2004 2006 deuteron (6 Li. D) polarised target — L only 2007 proton (NH 3) polarised target — L&T muon beam: 160 Ge. V/c longitudinal polarisation -80% intensity 2· 108 µ+/spill (4. 8 s/16. 2 s) Spectroscopy with hadron beams 2008 2009 Anna Martin April 21, 2010

COMPASS SET-UP for muon running • high energy beam • large angular acceptance • broad kinematical range two stages spectrometer Large Angle Spectrometer (SM 1) Small Angle Spectrometer (SM 2) variety of tracking detectors to cope with different particle flux from θ = 0 to θ ≈ 200 mrad Sci. Fi Silicon Micromegas GEMs Straws SDC MWPC W 45 SM 2 SM 1 ~ 50 m Polarised Target Anna Martin m beam + m F. Bradamante April 21, 2010

COMPASS SET-UP for muon running • high energy beam • large angular acceptance • broad kinematical range two stages spectrometer Large Angle Spectrometer (SM 1) Small Angle Spectrometer (SM 2) Muon. Wall SM 2 E/HCAL SM 1 Muon. Wall Polarised Target RICH Anna Martin m beam + m F. Bradamante calorimetry, PID RICH detector April 21, 2010

COMPASS SET-UP for muon running solid state target operated in frozen spin mode, 120 cm long 2002 -2004: 6 Li. D (polarised deuteron) 2 cells with opposite polarisation 2007: NH 3 (polarised protons) 3 cells with opposite polarisation Muon. Wall SM 2 E/HCAL SM 1 Muon. Wall Polarised Target RICH Anna Martin m beam + m F. Bradamante April 21, 2010

SIDIS event selection DIS cuts: Q 2>1 (Ge. V/c)2 , 0. 1<y<0. 9, Anna Martinselection: hadron ph > 1. 5 Ge. V/c, W>5 Ge. V/c 2 p. Th>0. 1 Ge. V/c, z>0. 2 April 21, 2010

• the COMPASS experiment • Collins asymmetry • Sivers asymmetry Anna Martin April 21, 2010

The Structure of the Nucleon three distribution functions are necessary to describe the quark structure of the nucleon at LO in the collinear case nucleon polarization U quark L T U polarization number density L helicity T Anna Martin transversity April 21, 2010

Transversity PDF is chiral-odd: • cannot be measured in inclusive DIS • observable effects are given only by the product of T q (x) and an other chiral-odd function can be measured in SIDIS off transversely polarised targets via “quark polarimetry” “Collins” asymmetry “Collins” Fragmentation Function “two-hadron” asymmetry “Interference” Fragmentation Function Λ polarisation Fragmentation Function of q↑ Λ D. Kang all explored in COMPASS Anna Martin April 21, 2010

Collins asymmetry if transversity PDF and Collins FF different from zero modulation in the azimuthal distribution of the final state hadrons C FC = f h + f S - p fh f. S transversity azimuthal angle of the hadron, azimuhtal angle of the spin of the nucleon “Collins FF” first measured by HERMES (proton target) and COMPASS (deuteron target) Anna Martin April 21, 2010

Collins asymmetry • clear non-zero effects seen by HERMES on proton evidence that both transversity PDF and Collins FF are non zero • ~ zero asymmetries measured by COMPASS on deuteron over the whole x-range understood as u – d cancellation Anna Martin April 21, 2010

Collins asymmetry • clear non-zero effects seen by HERMES on proton evidence that both transversity PDF and Collins FF are non zero • ~ zero asymmetries measured by COMPASS on deuteron over the whole x-range understood as u – d cancellation • independent measurement of Collins effect from BELLE data (first measurements from LEP data) all these data are well described in global fits [M. Anselmino et al. ] and allowed for first extractions of the Collins FFs and of transversity PDFs a major result ! still, a lot of interest for the higher energy COMPASS proton measurement Anna Martin April 21, 2010

Collins asymmetry – proton data COMPASS proton results from 2007 run the analysis is over and the paper almost ready to be sent new results very much the same as presented at DIS 2009 ssyst ~ 0. 5 sstat • at small x, the asymmetries are compatible with zero • large signal in the valence region of opposite sign for positive and negative hadrons same sign and ~ strength as HERMES Anna Martin April 21, 2010

Collins asymmetry comparison with predictions from fit to the HERMES proton, COMPASS deuteron, BELLE data our data support the assumption of a weak Q 2 dependence in the present energy range Anna Martin April 21, 2010

Collins asymmetry SIGN in agreement with the “recursive fragmentation model with quark spin” [X. Artru, ar. Xiv: 1001. 1061] can also describe the two-hadron asymmetries Anna Martin April 21, 2010

Two Hadron Asymmetry azimuthal asymmetry in f. RS= f. R┴ - fs’ f. R┴ is the azimuthal angle of the plane defined by the two hadrons R = (z 1 p 2 – z 2 p 1)/(z 1+z 2) A. Bacchetta, M. Radici, hep-ph/0407345 X. Artru, hep-ph/0207309 Anna Martin Interference Fragmentation Function BELLE A. Vossen April 21, 2010

Two Hadron Asymmetry deuteron: compatible with zero all x proton: DIS 2009 x. F > 0. 1 z 1, 2 > 0. 1 Z = z 1+ z 2< 0. 9 RT > 0. 07 Ge. V/c in the valence region the asymmetry is ~ larger than the Collins asymmetry larger than measured by HERMES difficult to describe both sets of results at the same time [Bacchetta et al. , Mah et al. ] it will be interesting to analyse these data in the framework of the “recursive model” Anna Martin April 21, 2010

• the COMPASS experiment • Collins asymmetry • Sivers asymmetry Anna Martin April 21, 2010

The Structure of the Nucleon three distribution functions are necessary to describe the quark structure of the nucleon at LO in the collinear case taking into account the quark intrinsic transverse momentum k. T , at leading order 8 PDFs are needed for a full description “TMDs” nucleon polarisation U quark U L T number density polarisation L helicity T Anna Martin transversity SIDIS gives access to all of them April 21, 2010

SIDIS cross-section with unpol target 3 modulations with L pol target I. Savin Collins 8 modulations (4 LO) with T pol target Anna Martin all measured by COMPASS on deuteron April 21, 2010 A Bacchetta, M Diehl, K Goeke, A Metz, P Mulders, M Schlegel (06) 3 modulations

The Structure of the Nucleon three distribution functions are necessary to describe the quark structure of the nucleon at LO in the collinear case taking into account the quark intrinsic transverse momentum k. T , at leading order 8 PDFs are needed for a full description “TMDs” nucleon polarisation U quark U L T number density Sivers function polarisation L helicity T Anna Martin transversity SIDIS gives access to all of them April 21, 2010

The Sivers PDF gives the correlation between the nucleon transverse spin and the quark kt • requires final/initial state interactions to survive time-reversal invariance • time-reversal invariance implies: … to be checked O. Denisov • in SIDIS, it is responsible of a modulation in S = fh - f. S of the hadron produced inclusively on a transversely polarized target S Anna Martin April 21, 2010

Sivers asymmetry • strong signal seen by HERMES for p+ and K+ producted on transversely polarized protons • no signal seen by COMPASS on transversely polarized deuterons, interpreted as u- and d-quark cancellation (as for the Collins asymmetry) first extractions of the Sivers function from pion SIDIS data good fits to the data Anna Martin April 21, 2010

Sivers asymmetry – proton data the analysis of the 2007 data is over new results new evidence for a positive signal for h+, which extends to small x, in the region not measured before Anna Martin April 21, 2010

Sivers asymmetry – proton data the analysis of the 2007 data is over new results new evidence for a positive signal for h+, which extends to small x, in the region not measured before systematic errors Anna Martin h- ~ 0. 5 sstat h+ ~ 0. 8 sstat plus a scale (abs) uncertainty of ± 0. 01 different results from data collected in the and in Aprilfirst 21, 2010 the second half of the run

Sivers asymmetry – proton data comparison with the HERMES final results new h+ COMPASS asymmetry Anna Martin same sign smaller by a factor ~ 2 April 21, 2010

Sivers asymmetry – proton data comparison with the most recent predictions from M. Anselmino et al. based on the fit of HERMES proton and COMPASS deuteron data new Anna Martin April 21, 2010

Sivers asymmetry – proton data comparison with calculations from S. Arnold et al. [ar. Xiv: 0805. 2137] in agreement with HERMES preliminary results new Anna Martin April 21, 2010

Sivers asymmetry – proton data various investigations to understand the results possible W dependence new Anna Martin April 21, 2010

Sivers asymmetry – proton data W dependence new Anna Martin April 21, 2010

Sivers asymmetry – proton data W dependence new Q 2 W = 12 Ge. V (Ge. V 2) W = 7. 5 Ge. V W = 5 Ge. V 160 Ge. V Anna Martin x April 21, 2010

Sivers asymmetry – proton data W dependence new W = 12 Ge. V Q 2 (Ge. V 2) W = 7. 5 Ge. V W = 5 Ge. V 160 Ge. V 27. 5 Ge. V no evidence of dependence of the Collins asymmetry on W W = 3. 3 Ge. V Anna Martin x April 21, 2010

Summary the Collins and Sivers asymmetries of the proton have been measured at high energy • Q 2 and W ranges extended at large values • x range extended to considerably smaller values Collins asymmetry: • clear signal both for positive and negative hadrons in the valence region • SIDIS as appropriate tool to investigate the transversity PDF Sivers asymmetry: • signal for positive hadrons, also at small x values • indication for a possible (and unexpected) W dependence much more can be done in SIDIS off transversely polarised nucleons Anna Martin April 21, 2010

COMPASS near future 2010: one full SPS year of run with the transversely polarised NH 3 target and the 160 Ge. V muon beam the run will start in 2 weeks fron now projected statistical errors for the Sivers asymmetry ▫ 2010 run ● h+ Anna Martin (proj) these data h- April 21, 2010

HERMES p+, DIS 2007 Anna Martin April 21, 2010

Other SSAs - Deuteron data two twist-2 asymmetries can be interpreted in QCD parton model and will allow to extract unexplored DFs “pretzelosity” Collins FF on deuteron asymmetries compatible with zero : again cancellation between proton and neutron? Anna Martin April 21, 2010

Unpolarised Azimuthal Asymmetries summary • positive hadrons • negative hadrons • error bars: statistical errors only Anna Martin Transversity 2008 April 21, 2010

Recent Fits V. Barone et al. , ar. Xiv: 0912. 5194 v 1 [hep-ph] Anna Martin April 21, 2010

Compass proton data comparison with HERMES Anna Martin April 21, 2010

Collins asymmetry COMPASS proton data results very much the same as presented at DIS 2008 • Anna Martin ssyst ~ 0. 5 sstat large signal in the valence region April 21, 2010

2 H Asymmetries – Proton Calculation by Ma et al. , TPSH 09, Armenia, June 2009. Anna Martin April 21, 2010

2 H Asymmetries – Proton Prediction by Bacchetta, Radici, hep-ph/0608037 (Interference Fragmentation function scaled down to fit HERMES data) Anna Martin Recent BELLE measurement of the IFF April 21, 2010

exclusive r 0 asymmetries – proton data Anna Martin April 21, 2010

exclusive r 0 asymmetries – deuteron data Anna Martin April 21, 2010

W dependence Anna Martin April 21, 2010

W dependence Anna Martin April 21, 2010

Sivers asymmetry - 2007 proton data Transversity 2008 data collected in the second half of the run only (3 periods over 6) Anna Martin DIS 2010 data collected in the first and in the half of the run (4 periods over 6) April 21, 2010

Sivers asymmetry comparison with HERMES new asymmetry smaller at COMPASS by a about a factor of 2 Anna Martin April 21, 2010

COMPASS SET-UP radiator C 4 F 10 threshold: p ~2 Ge. V/c K ~ 10 Ge. V/c Polarised Target RICH Anna Martin m beam RICH detector April 21, 2010

The Target System solid state target operated in frozen spin mode 2002 -2004: 6 Li. D (polarised deuteron) during data taking with transverse polarization, polarization reversal after ~ 4 -5 days • dilution factor f = 0. 38 • polarization PT = 50% • two 60 cm long cells with opposite polarization 2006: • PTM replaced with the large acceptance COMPASS magnet (180 mrad) • 2 target cells 3 target cells 2007: NH 3 (polarised protons) • dilution factor f = 0. 14 • polarization PT = 90% Anna Martin April 21, 2010

DIS Event Selection DIS cuts: Q 2>1 (Ge. V/c)2 0. 1<y<0. 9 W>5 Ge. V/c 2 hadron selection: p > 1. 5 Ge. V/c p. T>0. 1 Ge. V/c, z>0. 2 Anna Martin April 21, 2010

Structure of the Nucleon three distribution functions are necessary to describe the quark structure of the nucleon at LO: q(x) : number density or unpolarised distribution probability density of finding a quark with a fraction x of the longitudinal momentum of the parent nucleon very well known Dq(x) = q - q : longitudinal polarization or helicity distribution unpolarised DIS (CERN, HERA, SLAC) a longitudinally polarised nucleon, probability and hard nucleon in scattering measurements density of finding a quark with a momentum fraction x and spin parallel to that of the parent nucleon well known measurements at or transversity DTpolarised q(x) = q -DIS/SIDIS q : transverse polarization SLAC, EMC, SMC, HERMES, COMPASS distribution in a transversely polarised nucleon, probability density of finding a quark with a momentum fraction x and spin parallel to that of the parent new nucleon poorly known SPIN-Praha-2009 A. Martin

Transversity distribution • proposed in ’ 79 (Ralston & Soper), reappraised in ’ 90 • properties • DTq(x) Dq(x) • probes the relativistic nature of quark dynamics • no contribution from the gluons simple Q 2 evolution • positivity (Soffer) bound • first moment: tensor charge • sum rule for transverse spin in Parton Model framework • it is related to GPD’s Bakker, Leader, Trueman, PRD 70 (04) • is chiral-odd • more difficult to measure SPIN-Praha-2009 A. Martin

How to measure DTq(x) is chiral-odd cannot be measured in inclusive DIS it can be measured in SIDIS on transversely polarized targets: the observable is the convolution of DTq(x) with another chiral-odd quantity in l. N↑ l’h. X the chiral-odd partner is the “Collins” fragmentation function, which describes a possible left-right asymmetry of the hadrons in the hadronization process of a transversely polarized quark Collins effect: a quark moving “horizontally” and polarized “upwards” would emit the leading meson preferentially on the “left” side of the jet SPIN-Praha-2009 A. Martin

Collins asymmetry • clear non-zero effects seen by HERMES on proton evidence that both transversity PDF and Collins FF are non zero • ~ zero asymmetries measured by COMPASS on deuteron over the whole x-range understood as u – d cancellation Anna Martin April 21, 2010

SIDIS cross-section leading order Collins asymmetry unpol L pol. target - T pol. target pol. beam & target Anna Martin April 21, 2010

V W =5 W = 7. 5 Ge 12 = W y= 0. 1 . 9 y= 0 Anna Martin April 21, 2010