Update on Prospects for Drell Yan at RHIC

1
Update on Prospects for Drell Yan at RHIC
L.C. Bland Brookhaven National Laboratory Transver
se Partonic Structure Workshop Yerevan, 26 June
2009
2
Outline

• Motivation for transverse spin Drell Yan
• Requirements
• Progress towards large-y DY

3
Transverse Spin Drell Yan at RHIC vs p-Sivers
Asymmetry in Semi-Inclusive Deep Inelastic
Scattering

• Important test at RHIC of recent fundamental QCD
  predictions for the Sivers effect, demonstrating
  
• attractive vs repulsive color charge forces
• Possible access to quark orbital angular momentum
  
• requires very high luminosity (RHIC II)
• both STAR and PHENIX can make important,
  exciting, measurements
• o

Transverse-Spin Drell-Yan Physics at RHIC L.
Bland, S.J. Brodsky, G. Bunce, M. Liu, M.
Grosse-Perdekamp, A. Ogawa, W. Vogelsang, F.
Yuan http//spin.riken.bnl.gov/rsc/write-up/dy_fin
al.pdf
4
Attractive vs Repulsive Sivers Effects Unique
Prediction of Gauge Theory !
Simple QED example
Drell-Yan repulsive
DIS attractive
Same in QCD
As a result
5
Experiment SIDIS vs Drell Yan SiversDIS -
SiversDY Probes QCD attraction and QCD
repulsion
HERMES Sivers Results
RHIC II Drell Yan Projections
0
Sivers Amplitude
M. Diefenthaler arXiv0706.2242 (final results
from HERMES just released arXiv0906.3918)
0
0.1 0.2 0.3 x
6
Rapidity and Collision Energy Transverse Spin
Asymmetries for the DY Processhttp//spin.riken.b
nl.gov/rsc/write-up/dy-final.pdf
Light mass DY, Mggt 4 GeV/c2 Rapidity
distributions for different ?s
Large rapidity acceptance required to probe
valence quark Sivers function, also where
p?p?pX transverse spin asymmetries are found to
be large at RHIC.
7
Benchmarking Simulations
pp ? J/?X ? ll-X, ?s200 GeV
PHENIX, PRL 98(2007) 232002
mm- 1.2lthlt2.2
ee- hlt0.35
J/? is a critical benchmark that must be
understood before Drell-Yan
8
Dilepton Backgrounds
Drell-Yan
J/?
?
?
Isolation needed to discriminate open heavy
flavor from DY, and is possible at PHENIX with
planned upgrades.
9
Summary of Transverse SSA Drell-Yan Requirements

• 250 pb-1 transverse polarization Drell-Yan data
  sample probes Sivers function sign relative to
  SIDIS.
• Isolation required to discriminate low-mass DY
  from open-heavy flavor.
• Large rapidity will require tracking for
  charge-sign discrimination.
• Likely optimal at ?s 500 GeV, given
  demonstrated luminosity
• Useful to pursue DY measurement where ANgt0 is
  found for p?p?pX
• Large rapidity will require benchmarked
  simulations to establish tracking requirements.
  Benchmarking starts with J/? (see below).

10
LuminosityRun-9 performance
Source RHIC Collider Projections, W. Fischer et
al. (2009)

• Challenges remain to be overcome to realize the
  best-case scenarios
• Luminosity increases at ?s500 GeV relative to
  ?s200 GeV were realized
• Depolarizing resonances in RHIC will require new
  tunes to reduce their impact

11
LuminosityFuture Projections
Source RHIC Collider Projections, W. Fischer et
al. (2009)

• Luminosity projections for ?s 500 GeV are
  sufficient for transverse-spin DY
• Improved polarization would be useful to achieve
  sufficient accuracy

12
STAR Detector
Forward Meson Spectrometer commissioned/operated
in RHIC run 8.
Cluster-pair triggered readout of Forward Time
Projection Chamber in ongoing RHIC run 9.
13
From FPD To FMS

• 1264 Lead Glass Detectors
• Full azimuthal coverage for 2.5 lt ? lt 4

FPD Prior to Run 8
FMS Run 8 and beyond
14
FMS Detector
C.Perkins, QM09

• 20x more acceptance than previous forward
  electromagnetic calorimeters at STAR

Geometric Efficiency J/Psi xF
FPD
FMS

• Increased acceptance not only increases pion
  yields and kinematic range but also give much
  higher geometric efficiency for high-xF J/Psi

15
Field Effects

• Does the signal survive the field?

• Radial and Azimuthal fields impart impulses in
  the F direction
• These impulses are small and in opposite
  directions (they nearly cancel each other)
• Field effects on our signal are small

C.Perkins, DNP08
16
FMS Minbias Simulations and Association Analysis

• Simulation
• PYTHIA 6.222 full GEANT simulations
• 9.2 nb-1 Integrated Luminosity

• Fast J/? generator full GEANT simulations

• Data
• Plot includes lt 1 of full data set

• Reconstructed quantities match generated
  quantities quite well

• Full simulation models Mpair data very well

C.Perkins, QM09
17
Forward pp J/? 2-Cluster Analysis
Reconstructed 2-cluster invariant mass

• Fit with Gaussian Polynomial
• Gaussian Fit Parameters
• µ 3.083 0.017 GeV/c2
• s 0.028 0.011 GeV/c2
• ?2/d.o.f. 24.6/25
• Significance from fit
• 2.1 s
• Background Simulation
• Needs more statistics
• Normalized to integral of data

• Cuts Applied
• Epair gt 60.0 GeV
• energy sharing, Zpair lt 0.7
• Isolation Radius

• 0.5 Dh-Df

C.Perkins, QM09

• pair mass background modeled well by simulation
• significance of J/? not large here, but including
  requirements on pT,cluster further reduces
  background by suppressing low-pT p0 production.

18
Forward pp J/? 2-Cluster Analysis
Reconstructed 2-cluster invariant mass / ( 6
pb-1 Sampled Luminosity)
C.Perkins, QM09

• Fit with Gaussian Offset
• Gaussian Fit Parameters
• µ 3.080 0.020 GeV/c2
• s 0.082 0.026 GeV/c2
• ?2/d.o.f. 20.83/26
• Significance from fit
• 4.5 s

• Cuts Applied
• E_pair gt 60.0 GeV
• Z?? lt 0.7
• Isolation Radius

• 0.4 Dh-Df
• pT_cluster gt 1.0 GeV/c

• high-xF J/? may have implications for intrinsic
  charm at large Bjorken-x in proton
• use to benchmark simulations for future
  transverse-spin Drell-Yan experiment

19
Forward pp J/? 3-Cluster Analysis

• Reconstructed invariant mass of candidate ?C ?
  J/? ? events

• Peak Counts 8.40 2.88
• 2.9 s Significance
• µ 2.97 0.025 GeV
• s 0.070 0.025 GeV
• ?2/d.o.f. 0.7 with 14 points fit.

• Significance depends on background model
• 2.9 s significance with currently estimated
  background.

C.Perkins, QM09
20
Why does high-xF intrinsic heavy flavor matter?

• Diffractive Higgs production at the LHC via QQ in
  proton
• May provide a clear signal for Higgs production
  due to small background
• How can high-xF intrinsic heavy flavor happen?
• Not from Gluon Splitting (extrinsic heavy flavor)
• Heavy quarks are expected to be multi-connected
  to the valence quarks within a proton and appear
  at large x via

Phys.Rev. D73 (2006) 113005
QED
QCD

• Can intrinsic heavy flavor expectations be tested
  experimentally?

21
Status/Plan of Large-xF DY

• Large-xF J/? production has been observed from
  bare large-y calorimeter response in RHIC run 8.
• Cluster-pair trigger is operational for acquiring
  large-y tracking data in RHIC run-9. Pending
  analysis, requirements for future DY can be
  established (e.g., fast-tracking inside solenoid,
  space points in front of FMS).
• Sufficient luminosity for pp ?s500 GeV
  collisions has been established further
  development of polarization is required, as is
  measurement of AN(xF) for p?p?p0X at ?s500 GeV
  and measurement of large-xF J/? and U production
  at ?s500 GeV, to bracket light-mass DY region.
• Technical solutions exist for fast tracking
  inside solenoid (GEM trackers) and space points
  in front of FMS (forward meson preshower).
  Construction to span 2.5lthlt4 region is required,
  and could be completed in 2 years, pending
  approval.
• RHIC schedule is oversubscribed ? DY would be
  after RHIC run 11 (gt2011).
• Run-10 will be AuAu energy scan for
  deconfinement critical point search, and AuAu at
  ?sNN200 GeV.
• Run-11 is expected to be polarized pp, with
  unknown mix of ?s200,500 GeV and
  longitudinal/transverse polarization.

22
Backup
23
PHENIX Detector

• p0/g/h detection
• Electromagnetic Calorimeter (PbSc/PbGl)
• High pT photon trigger to collect p0's, hs, gs
• Acceptance hlt0.35, f 2 x p/2
• High granularity (1010mrad2)
• p/ p-
• Drift Chamber (DC) for Charged Tracks
• Ring Imaging Cherenkov Detector (RICH)
• High pT charged pions (pTgt4.7 GeV).
• Relative Luminosity
• Beam Beam Counter (BBC)
• Acceptance 3.0lt hlt3.9
• Zero Degree Calorimeter (ZDC)
• Acceptance 2 mrad
• Local Polarimetry
• ZDC
• Shower Maximum Detector (SMD)

EMCal
24
Energy Dependent Corrections
C.Perkins, QM09

• Reconstructed photon energies were individually
  corrected as dictated by detector simulations.

• Electromagnetic shower shape profiles and
  analysis code have been successfully adapted from
  FPD for use with FMS
• When the p0(h) mass is properly reconstructed, so
  is its energy