Stephane Platchkov

1

• Stephane Platchkov
• DAPNIA-laboratory, CEA-Saclay, France
• For the COMPASS collaboration
• MENU-2007, Forschungszentrum Jülich, September
  10-14, 2007

RECENT RESULTS FROM COMPASS
2
OVERVIEW PHYSICS OF COMPASS

• Hadron beams
• Pion and Kaon polarizabilities
• Diffractive and Central production of exotic
  states
• Search for glueballs
• Light meson spectroscopy
• Data taking
• 2002, 2003, 2004, 2007
• 2005 CERN SPS stop

• Muon beam
• Gluon spin contribution
• Quark spin contribution
• Flavor decomposition of spin distribution
  functions
• Transverse spin distribution functions
• Transversity two hadron correlations
• Vector meson production
• Spin transfer in L-hyperon production
• Search for pentaquarks

29,
COMPASS objectives Study of QCD First priority
until now spin structure of the nucleon
3
This talk results on nucleon spin structure

• Two COMPASS results
• How large is the gluon contribution to the
  nucleon spin?
• QCD analysis of the inclusive g1(x, Q2) data
  (Q2-evolution)
• Possible now, thanks to the high accuracy of the
  data
• What is the size of the polarized valence quark
  distribution?
• Determined from Semi-Inclusive Deep Inelastic
  Scattering
• Infer also the light quark sea polarization
• Future prospects
• COMPASS program in the next years

4
The COMPASS Spectrometer
50 m
Polarized beam and target 75 50
SAT,
LAT,
PID
0.003ltxlt0.5, 10-3ltQ2lt10 GeV2
5

• How large is the gluon contribution to the
  nucleon spin?
• (using a QCD fit to world data of g1(x))
• cf also Direct determination of DG through
  Photon-Gluon Fusion
• (talk by R.Kuhn, Session 18)

6
DIS asymmetries World results on A1d
Most recent data from COMPASS 2002 2003
2004 - PLB 647(2000)8

• ? Good agreement with previous experiments
• Much improved statistics at low x
• A1 compatible with 0 at low x

7
A1 was also measured for Q2lt1
COMPASS acceptance
Asymmetry compatible with 0 down to x0.0005
8
Polarized g1(x) structure function
R
F2
9
QCD analysis of world g1(x) data

• From DGLAP evolution equations
• A common fit to the world data allows a
  separation of all flavors a3, a8, DS, Dg !
• Two different fit programs (Next to Leading
  Order)
• World data (except HERMES 2006) from p, d, 3He
  (230 points)
• Parton parameterizations at Q23 (GeV/c)2
• Fit in scheme

• All Pij and Ck are known to NLO
• Mertig, van Neerven, 1996
• Vogelsang, 1996

Polarized g1(x) structure functions
10
QCD FIT RESULTS
The fit provides two solutions one with DG gt 0
and one with DG lt 0, All other parton densities
are nearly identical
11
QCD fit results
DG lt 0
DG gt 0
Quark contribution to the nucleon spin 0.30
0.02 Total gluon polarization 0.2-0.3 DG lt 0
or DG gt 0 !
12
Strange quark distribution
DG lt 0
DG gt 0
Fit result In agreement with a determination
from integral of g1(x)
Strange sea is polarized and negative In
agreement with Hermes (talk of P. Kravchenko)
13
The two QCD fit results for DG
We also have 3 direct measurements of DG
(Photon-Gluon Fusion) - High-pt hadron pairs
with Q2gt0 - High pt hadron pairs with Q2lt0 -
Open charm events (D -- Kp)
? Talk by Roland Kuhn, Session 18
14
Gluon polarization measurements
The direct measurements of DG(x) in agreement
with the QCD fit
15
A provocative digression

• What about the proton spin?
• (in a very naïve model)
• Everything started from here
• Assume DG positive, 0.25
• Assume DG negative, -0.25

W. Pauli
N. Bohr
Pauli and Bohr puzzled by the spin physics in the
30s
16

• What is the size/shape of the polarized valence
  quark distribution?
• Is the light quarks sea flavor symmetric?

17
Access to the valence spin distribution

• Use hadron asymmetries in semi-inclusive DI
  scattering
• Combine h and h- in difference asymmetry

Q2 range
1 70 GeV

• Advantages (LO)
• Fragmentation functions cancel
• No Particle Identification
• Better statistical errors

18
Valence quark spin distribution
De Florian, Navarro, Sassot, 2005
Unpolarized distributions from MRST04 (Martin,
Stirling and Thorne, 2004 At large x, use
inclusive asymmetry (sea very small)
19
Comparison with previous experiments
Reasonable agreement, improved statistical errors
20
Is the light quark sea symmetric?

• Incluisve DIS measures valencesea contribution
• Semi-inclusive DIS measures valence contribution
• Take the difference through the first moments

Two expectations, depending on the flavor
symmetry of the sea
21
First moment (integral) of DuvDvv

• Unmeasured contribution
• (0.7 lt x lt 1.0) 0.004 (est).

HERMES PR D71(2005)012003
Gv 0.410.070.05 Our result favors a
flavor-asymmetric sea (within 2.5s)
This analysis of the HERMES data therefore
favors a symmetric polarized light flavor sea
22
Flavor separation with 2006 and 2007 data
Improved statistics, much lower x values
23
Outlook and future prospects

• OUTLOOK
• Gluon contribution to the nucleon spin was
  measured
• likely to be small or zero
• Valence spin distribution was determined
• Light quarks flavor-asymmetric polarized sea is
  favored
• (Near-) future prospects
• 2006 Data on the deuteron being analyzed more
  statistics to come
• 2007 Polarized data on the proton (L T) is
  being collected better insight into DIS, SIDIS,
  transversity
• 2008 Program with hadron beams pion
  polarizability (talk by M.-L. Collantioni,
  Thursday, session 28), search for exotics,
  light-meson spectroscopy

24
Spare slides
25
Shape of parton distributions
26
g1(x) for Q2lt1
27
Fit results and low x behavior
Leader, Sidorov, Stamenov, Phys.Rev.
D73(2006)034023. Gluck et al., Phys.Rev.
D63(2001)094005. Blümlein Böttcher, Nucl.Phys.B63
6(2002)225.
- Evolution shows little sensitivity to various
fits - Fits of LSS-05, GRSV, BB show similar
trend at low x - Sign of DG is mainly determined
by the low x behavior of g1(x)
28
Fit results on p, d, n