Longitudinal spin asym. and DG at COMPASS

1
Longitudinal spin asym. and DG at COMPASS
JM Le GOFF, DAPNIA/Saclay
orbital momentum
quarks
gluons
valance
sea
HEP 2007, Manchester
2
The COMPASS experiment
240 physicists from 25 institutes

• Polarized muon beam
• Long. polarized target ( Transv.)
• Quark polarization
• g1, DS, Dq, valance and sea
• Gluon polarization
• Transversity
• Production of r, F, J/?, L, ...

• Hadron beam
• Primakoff p, K polarizabilities
• see A. Guskov presentation
• Exotics q-states, glue balls
• Double charmed hadrons

project Drell-Yann for transversity

• project
• LH2 target proton recoil det.
• Generalized Parton Distributions

3
Longitudinally polarized muons 160 GeV/c
2.108 m/ spill (4.8s / 16.8s)
PB -80 Longitudinally (or transversely)
polarized deuteron target 6LiD
PT 50 Luminosity 5. 1032 cm-2s-1


LHC
COMPASS at CERN
SPS
M2 beam line

4
COMPASS Spectrometer
Trigger Hodoscopes
ECALs HCALs
MWPCs
SM2
RICH
50 m
SM1
Drift Chambers
m Filters/Walls
6LiD target
Straws
GEMs
Si
Micromegas
160 GeV m
NIM A 577(2007) 455
SciFi
veto
5
Spin asymmetry extraction
m ?
2-cell target
?
?
every 8 hours
upstream
downstream
?
?
weighted method
gain in FoM
6
(semi-)inclusive analysis
7
Inclusive asymmetries, md ? m'X
Q2gt1
PLB647(2007)8
good agreement with previous exp. factor 4 gain
in stat at low x compatible with zero at low x
Q2lt1 see PLB647(2007)330
8
g1 and contrib. to nucleon spin
G1?g1(x)dx has only 2 correction from
unmeasured low x (was a 50 negative correction
before) G1 ? DS0.350.030.05 and
Ds-0.080.010.02 (CMPS alone)
9
QCD analysis
DGLAP equations at NLO Disagreement of new data
with previous QCD fits ? new fit 2 solutions,
DGgt0 and DGlt0, DG 0.2-0.3
10
Semi-inclusive asym, m d ? m' h X
separate valance and sea
11
Difference charge asym.
LO
12
polarized valance quark distrib
for xgt0.3 sea ltlt valance g1d(x)?
Duv(x)Ddv(x)
13
Moments

• if then
  DuvDdva80.580.03
• DuvDdv0.410.070.05 2 s below
• DuvDdv G1 and a8
• ? favors asymmetric sea

DsDs-0.080.010.02 (inclusive)
14
Gluon polarization
15
Photon Gluon Fusion

• 2 signatures
• qqcc open charm (D0)
• - no physical bckg
• - s small, BR(Kp) 4
• ? low stat
• - scale mc2 (no Q2gt1 cut)
• high pT hadron pairs
• - large stat
• - phys bckg ? needs MC
• - scale pT2 (no Q2gt1 cut)

16
D0 meson reconstruction

• Thick target no decay vertex
• D0?Kp mass reconstruction
• ? combinatorial bckg
• kinematical cuts
• RICH for particule id
• D ? D0pslow
• cut on M(Kpp)-M(Kp)

17
Extraction of DG/G

• Abckg measured to be
• zero outside D0 peak
• event kin not fully
• reconstructed ?
• parametrize aLL with NN

18
High pT hadron pairs

Photon Gluon Fusion
Leading Order
QCD compton
Resolved g Q2 lt 1 (GeV/c)2
aLL calculable partonic asymmetry RPGF
Monte Carlo required for RPGF and Abckg
19
Contributions to cross-section (Q2lt1)
Resolved photons
20
Data vs MC
excellent to good agreement for all variables
21
Systematic errors

• Dominant syst error due to MC
• vary MC parameters in a range where data vs MC
  OK
• largest effect for photon kT

22
Results for DG/G
- open charm DG/G-0.570.410.17 - high pT
Q2lt1 DG/G-0.0240.0890.057 disfavors large
values of DG (2-3 needed to get DSa8)
PLB633(2006)25
23
Prospects

• all results presented 2002-04 (no data in 2005)
• improved charm analysis with S/(SB) in the
  weight
• 2006 data
• - new PT magnet ? larger acceptance
• - improved RICH faster electronics
• PMs in the inner region
• - 3-cell target kills false asym
• 2007 data on proton (NH3 target)
• (not very good for DG since FoM(fPT)2
  reduced)
• flavor separation in semi-inclusive analysis
  
• Duv and Ddv , Du and Dd

24
Improved D0 peak in 2006
S/B improved by a factor 2
25
Conclusions

• DS 0.30 0.01 0.02
• error due to low x extrapolation strongly
  reduced
• DsDs-0.08 0.01 0.02
• DuvDdv 0.41 0.07 0.05
• seems that DuDd 0
• fit to inclusive data gives 2 solutions
  DG0.2-0.3
• direct measurement
• DG not very large (anomaly contribution to
  a0 small)
• but DG0.35 still possible (in which case
  Lz0)