Recent Spin and FMS Results at STAR
STAR
Andrew Gordon
Brookhaven National Laboratory
Moriond-QCD
March 14-21, 2009
Colliding-Beam Spin Physics: Only at RHIC
2
1) Longitudinal spin program at STAR
Inclusive Jets
s=200 GeV
Strong constraint on the size of Δg from RHIC data for 0.05<x<0.2.
STAR data contributes strongly to global fits as in D. deFlorian et al., PRL 101 072001, 2008.
Run 9 currently running polarized protons at s=500 GeV for the first time. This will
allow start of measurements of W bosons.
Colliding-Beam Spin Physics: Only at RHIC
2) Transverse Spin program at STAR
Run 8 (2008) and earlier runs show large asymmetries in the forward region.
New detector will allow higher kinematic reach and extensions beyond
inclusive 0 data.
3
4
STAR Detector
New for Run 8:
Forward Meson
Spectrometer (FMS)
Stack of 1264 lead
glass cells, roughly
18 X0 in z.
FPD
(runs 3+)
Located at far West
side of Hall, at the
opening to RHIC
tunnel. Faces blue
beam.
7.5 meters from
interaction point.
FMS provides nearly 20x the coverage of previous
forward detectors at STAR
North-half, view from
the hall
FMS newly
commissioned for
Run 8
Run 85 FMS
FPD
Nearly contiguous coverage for
2.5<h<4.0.
5
East BBC
(yellow)
Expected spin up
Expected spin down
West BBC
(blue)
Polarization x Asymmetry (uncalibrated)
Bunch-by-bunch polarization from colliding beams at
STAR
Run 8 Data
~3.5 s (statistical) measurement of
polarization per bunch per hour
Statistical uncertainties only
STAR Preliminary
Bunch number at STAR
See also J. Kiryluk (STAR) ArXiv:hepex/0501072v1, 28 Jan 2005
6
7
Why high XF at a Collider?
High rapidity ’s (h~4) from asymmetric partonic collisions
p+p
0, h=3.8, √s=200GeV
<z>
Mostly high-x valence quark on low-x gluon
Fragmentation z nearly constant and high 0.7 ~ 0.8
<xq> NLO pQCD
Jaeger, Stratmann,
<xg> Vogelsang, Kretzer
Which beam is polarized (not averaged over) probes different regions
0
p
p
If Polarization here:
XF>0
Valence quark spin effects
If Polarization here:
XF<0
Low x gluons and other partons
Runs 2, 3, 5, and 6 forward
Distributions steeply
falling with PT and XF.
0
asymmetry from FPD
Data from Jan 2002
PT vs XF
Cross
sections
coverage
ofwell
data
described within pQCD
framework at √s=200
GeV
FPD set at different
distances from beam
Suggests
that asymmetry
for the different
<h>
ranges.
data
can be described
within the context of
pQCD.
PRL 92 (2004) 171801
“Forward Neutral Pion Transverse(STAR),
Single Spin
Asymmetries in p+p
Collisions at sqrt(s)=200 GeV,” PRL 101 (2008) 222001
8
Run 3, 5, and 6 asymmetry data: Theory can predict XF dependence
based on Sivers function fits to +/- asymmetries…
Run 6 Data
L=6.8 pb-1
Yellow beam
polarization =
562.6 %
AN =
ds - ds
ds + ds
dσ↑(↓)=differential cross
section when proton has
spin up (down).
Data: B.I. Abelev et al. (STAR), PRL 101 (2008) 222001
Theory (red): M. Boglione, U. D’Alesio, F. Murgia [arXiv:hep-ph/0712.4240]
Theory (blue): C. Kouvaris, J. Qiu, W. Vogelsang, F. Yuan, PRD 74 (2006) 114013
9
…but rising PT dependence is not predicted by the same fits
XF>0.4
B.I. Abelev et al. (STAR) PRL 101 (2008) 222001
Data broken
out in XF
bins
10
FMS Acceptance allows azimuthal dependence to be
measured
Run 8 data
Run 8 data
P(blue)=45.5±3.3%
L~6.2 pb-1 in plot
AN as a function of xF integrated over
the FMS acceptance.
Plots from Nikola Poljak, for STAR collaboration, “Spin-dependent
Forward Particle Correlations in p+p Collisions at s = 200 GeV,”
hep-ex/0901.2828, to be published as Spin 2008 conference
proceedings.
AN versus <cos φ> for positive (blue beam)
and negative (yellow beam) XF
Important confirmation of
previous data
11
“Jet-like” events in the FMS can potentially allow
integration over fragmentation products to measure
Sivers effect.
12
“Collins effect”: asymmetry in fragmentation
“Sivers effect”: asymmetry from initial state KTspin correlations
STAR Preliminary
Inclusive 0 asymmetries can not distinguish
these.
Caveat for Run 8: FMS data
was acquired with a “high
tower” trigger, which creates
a bias towards “jets” that
derive from a small number
of high-z initial fragments
Simulation and Data agree well
for “jet-like” events
“Jet-shape” distribution of energy within jet-like
objects in the FMS as a function of distance
from the jet axis.
Plot from hep-ex/0901.2828.
Heavier mesons also accessible at high XF
p + p  M +X
M  + 
s  200 GeV
STAR 2006 PRELIMINARY
Di-photons in FPD with
E(pair)>40 GeV
No “center cut” (requirement that
two-photon system point at middle
of an FPD module)
With center cut and Z<0.85
Average Yellow Beam
Polarization=56%
S. Heppelmann, PANIC 2008
13
Measurement of vector mesons at large xF
14
Inclusive 0 asymmetries include feed-down from higher mass decays.
One source of 0s is the spin 1 w (782 MeV).
w decays are accessible in the FMS through their 0 (BR=9%) decay
channel.
These events are closer to the original fragmentation product.
Spin-1 production is potentially interesting, and explictly occurs in some models
of string fragmentation that provide an intuitive picture of asymmetry in
fragmentation.
w Selection
15
Look at all triplets of clusters with E>6 GeV. Apply fiducial cuts of 1/2
cell from all module boundaries
Associate two of clusters with π0 from the ω decay and one with 
Choose π0 to be pair with mass closest to 0.135 GeV, and photon from ω decay to be
the third cluster. A fast, photons-only simulation indicates that this produces the
correct identification for nearly 100% of ω→π0γ decays
Require that M(π0) be within 0.1
GeV of 0.135 GeV.
Kinematic cuts to reduce QCD
background (real π0 decays with a
third EM-rich hadronic cluster in
the FMS):
STAR Preliminary
Run 8 data
Pair mass for all
triples
All cuts applied (see below)
•PT(triplet)>2.5 GeV
•E(triplet)>30 GeV
•PT(photon cluster)>1.5 GeV
•PT(π0)>1 GeV.
M(π0 clusters) (GeV/c2)
16
Mass distribution of all triples
STAR Preliminary
For this plot simulation is
background only
Fit is gaussian + cubic polynomial
μ=0.784±0.008 GeV
σ=0.087±0.009 GeV
Scale=1339±135 Events
Run 8 data
Pythia+GEANT
(background only)
(simulation
generation
ongoing)
Comparison to Run 8 dAu data
may be interesting.
Significant (10s) w0 signal seen in the data.
Conclusions
Large-acceptance FMS is online
“jet-like” events hold the promise of a direct measurement of the Siver’s effect.
Data should allow asymmetry measurements beyond inclusive 0s
Outlook
Run 8 measurements of AN vs PT for inclusive 0s at high XF.
Run 8 pp and dAu comparisons.
Run 9 (running now) should allow STAR to extend constraints on DG and also to
begin W measurements.
Possibility of future direct- AN measurements with FMS.
Possibility of future Drell-Yan AN with FMS.
17
18
End
Kinematic cuts to reduce QCD background (real π0 decays with a third
EM-rich hadronic cluster in the FMS):
•PT(triplet)>2.5 GeV
•E(triplet)>30 GeV
•PT(photon cluster)>1.5 GeV
•PT(π0)>1 GeV.
Comparison of data (black) to simulation (red)
STAR Preliminary
Run 8 data
Pythia+GEANT
STAR Preliminary
Run 8 data
Pythia+GEANT
19
Sizeable asymmetries can be used to measure beam
polarization
STAR
BBC
Inner tiles cover
~3.5<h<5
Can use large
asymmetries to measure
(relative) bunch-bybunch polarization
See also J. Kiryluk (STAR)
ArXiv:hep-ex/0501072v1, 28 Jan
2005
20
Inner tiles of BBC accumulated every clock cycle
{
Scaler Boards
}
Discriminated
phototube outputs
East-West Coincidence
{
Bunch Crossing (7-bits)
24-bit word is
histogrammed
every clock cycle
}
21
Inclusive Jets
p + p  Jet + X

Jets at each pT are a different mix of
subprocesses
Phys. Rev. Lett. 97 (2006) 252001
Smeared x-range for jets at a few pTs
22