Krzysztof Kurek

1
?G from COMPASS

• Krzysztof Kurek
• Andrzej Soltan Institute for Nuclear Studies,
  Warsaw

on behalf of the COMPASS Collaboration
DIS2006 XIV International Workshop on Deep
Inelastic Scattering 20/April/2006 -
24/April/2006 Tsukuba
2

• longitudinally polarised muon beam
• longitudinally or transversely polarised
  deuteron (6LiD) target
• momentum and calorimetry measurement
• particle identification

LHC
SPS
N
3
THE COMPASS SPECTROMETER
Hodoscopes
E/HCAL2
E/HCAL1
SM2
RICH1
Muon Wall 2, MWPC
SM1
Polarised Target
MWPC, Gems, Scifi, W45 (not shown)
SPS 160 GeV m beam
Muon Wall 1
Straws, Gems
Micromegas, SDC, Scifi
Scifi, Silicon
4
Content

• Motivation and Nucleon spin decomposition.
• Three methods of accessing in Compass.
• Open charm channel method and results.
• High pT hadron pairs method
• - Results for events with low Q2.
• - Results for events with Q2 gt 1 GeV2.
• Conclusions.

5
Nucleon Spin decomposition
½ ½?S ?G L

• A very small fraction of the proton spin is
  carried
• by the spin of the quarks - put the naive but
  well-
• accepted quark model into serious
  questioning!
• (EMC (1988) a0 ?? 12 9 14 while ?
  60 expected,
• confirmed by SMC, SLAC and Hermes ?? 20 -
  30)
• The possible role of axial anomaly
• measured quantity a0 ?? - (3 ?S/2?) ?G

6
Nucleon Spin decomposition
½ ½?S ?G L

• A very small fraction of the proton spin is
  carried
• by the spin of the quarks - put the naive but
  well-
• accepted quark model into serious
  questioning!
• (EMC (1988) a0 ?? 12 9 14 while ?
  60 expected,
• confirmed by SMC, SLAC and Hermes ?? 20 -
  30)
• The possible role of axial anomaly
• measured quantity a0 ?? - (3 ?S/2?) ?G
• warning orbital angular momentum anomaly
  cancelled (X.Ji)

7
Direct measurements of ?G/G
Photon Gluon Fusion (PGF) probes gluons
LL

• Open charm golden channel
• ? no background asymmetry, less MC
  dependent.
• ? small statistics, NLO corrections can be
  important
• 2 high pT hadrons (pT gt 0.7, then selection on
  ?pT2)
• ? Large statistics
• ? physical background model (MC)
  dependent,
• requires very good description of data
  by MC.

8
Direct measurements of ?G/G

• 2 high pT hadrons
• Low Q2 analysis (Q2lt1) perturbative scale fixed
  by pT,
• complicated physical background e.g. resolved
  ?,low pT
• High Q2 analysis (Q2gt1) scale Q2 ,physical
  background
• better controlled in the frame of pQCD.

Physical background
resolved ? (Q2lt1)
QCD-Compton
resolved ? (Q2lt1)
9
Direct measurements of ?G/G
Due to limitations of every methods - high pT
hadron pairs methods are complementary to one
another and complementary with the method using
selection of open charm production!
resolved ? (Q2lt1)
10
?G/G from open charm channel
11
?G/G from open charm channel
D0 ? K ? (untagged) D ? D0 ? ? K
? ?

ALL/D
Scale ltQ2gt ? 13 GeV2 4 mc2
12
?G/G from open charm channel

ALL/D
?LL
?LL calculated with help of MC and parametrized
by measured quantities (Neural Network used)


13
?G/G from open charm channel
D0 D result 2002 2004
-0.57 ? 0.41(stat.) _at_ xg ? 0.15,
scale ?13 GeV2
The studies on the systematical uncertainty are
ongoing
14
?G/G from 2 high pT hadrons
15
?G/G from 2 high pT hadrons (low Q2 )
Low Q2 Q2 lt 1 GeV2
30
50
Resolved ?
MC event generator PYTHIA is used for low Q²
analysis
16
?G/G from 2 high pT hadrons (low Q2 )
Data MC
Data/MC
Q²
Q²
pT
pT
Example of good description of the data by the
simulation
17
Resolved photon asymmetry
?G/G from 2 high pT hadrons (low Q2 )
Problem polarized PDFs of virtual photons not
measured!
?q?,pert(xp,?²) calculable
-q?,npert(xp,?0²) lt ?q?,npert(xp,?0²) lt
q?,npert(xp,?0²)
measured

• Allows us to obtain a range
• for (?q/q)? and (?G/G)?

? Adds a limited uncertainty to the estimation of
(?G/G)(xg)
18
?G/G from 2 high pT hadrons (low Q2 )
Values used for extraction of ?G/G
2002-2003 publishedPLB 633 (2006) 25-32
19
?G/G from 2 high pT hadrons (Q2 gt1)

• Statistics smaller than in low Q2 analysis (10)
• Background better controlled pQCD (QCD-C, LP)
• LEPTO MC generator has been used for data
• description (tunning similar to SMC)
• ?pT2 gt 2.5 GeV2 used.

2002-2003 data result
0.06 0.31(stat.) 0.06 _at_ xg 0.13 0.08
20
?G/G from 2 high pT hadrons (Q2 gt1)

• Analysis is ongoing 2002-2004 results expected
  soon
• Scale is determined by Q2 and in contrast to
  low Q2
• analysis the cut ?pT2 gt 2.5 GeV2 can be
  released to
• smaller value to optimize working point
• (question higher fraction RPGF and small
  statistics or
• lower fraction and higher statistics?)

• Neural Network is tested to improve selection of
  PGF
• subprocess and optimize working point.
• The significant improvement is expected.

21
Results for ?G/G - summary
Only statistical errors plotted
22
Results for ?G/G - summary
Open charm
23
Outlook

• More results soon available with 2004 data for
• Q2 gt1 GeV2 high pT events
• For the future
• Optimization of event selection with a neural
  network
• Bins in xg (requires improvement of xg
  reconstruction)
• NLO in open charm analysis

? No resolved photons!
? No physical background!

• 2006 data with new COMPASS magnet (larger xg)

24
Summary

• New measurements of ?G/G have been presented.
• Small ?G is preferred or ?G(xg) has a node
• around 0.1.
• Ellis-Jaffe sum rule seems to be violated if
  large ?G
• is excluded (axial anomaly).

• ?G ? 0.4 not excluded and scenario when L is
  small
• still possible.
• ?G ? 0 indicates the important role of angular
  orbital
• momentum in nucleon spin decomposition
  described
• in the frame of parton model and pQCD.