Spin Physics Results
from RHIC
M. Grosse Perdekamp, UIUC
International Workshop on
Hadron Structure and Spectroscopy
CRNS, Paris
April 4th -6th , 2011
AnDY
pp2pp
STAR
Spin Physics Results
from RHIC
Preliminaries:
Facility Status
QCD & PDFs vs Data
Gluon spin distribution
Inclusive hadron and jet results
QCD analysis
Low x & x-dependence with di-jets and rapidity separated di-hadrons
W-production in polarized p-p
First results from pp  W  eν
Transverse Spin
Inclusive AN
Channels isolating Collins or Sivers effects
 Drell Yan measurements in Akio Owaga’s talk
RHIC Status: Running Polarized p-p
at √s= 500 GeV through 4-8/15
Luminosity in Relative Unites
4 RHIC fills, last week of March
impact from
3rd collision
point
First full length 500 GeV run
P ~ 45 % (max above 50%)
goal is P ~ 50%
L ~ 5 x 1031 cm-2s-1
goal is L ~ 10 x 1031 cm-2s-1
o AnDY commissioned
successfully
o muon-W trigger in PHENIX
in operation
Time in Store [days]
o ∫Ldt ~ 40% of planned, due
to major hardware failure
in cryo-system (repaired).
3
NLO pQCD Cross Sections vs RHIC data
for Different √s and Rapidity Intervals
√s = 200 GeV
√s = 62.4 GeV
∆ PHENIX π0, η = 0
∙ Brahms π+ ,η = 2.95
⌷ STAR
π0 ,η = 3.7
-----
o PHENIX π0, η = 0
--------
NLO pQCD
NLO pQCD
pT [GeV]
Good agreement between inclusive
hadron cross sections from RHIC
data and pQCD calculations !
pT [GeV]
See analysis in
De Florian, Vogelsang, Wagner
PRD 76,094021 (2007) and
Bourrely and Soffer
Eur.Phys.J.C36:371-374 (2004)
4
(I) Gluon Spin Distribution
Inclusive hadron and jet results
QCD analysis
Low x & x-dependence with di-jets and
rapidity separated di-hadrons
ALL for Inclusive Hadrons & Jets at mid-Rapidity η ~ 0 :
constrain ΔG(x) at 0.05 < x < 0.2
ALL for neutral pions and jets vs DSSV
de Florian, Sassot, Stratmann,
Vogelsang PRD 80:034030,2009
0
x-range for ALLπ in 3 pT bins
Courtesy Swadhin Taneja, Stony Brook
other channels in STAR and
PHENIX: eta, charged hadrons,
ALL(“charm”) at mid-rapidity
ALL(J/ψ) for η~ 2.
6
ΔG(x) from DSSV Global QCD Analysis
de Florian, Sassot, Stratmann,
Vogelsang PRD 80:034030,2009
0.2
+0.051
ΔGtrunc[0.05,0.2] = ∫ ΔG(x) dx = 0.005-0.058
0.05
x ΔG(x)
 ΔGtrunc[0.05,0.2] ≈ 0 since node at x≈0.1
Data constrains only truncated first
moment of ΔG(x) in x-interval [0.05,0.2], but
not functional form at low or high x
ΔGtrunc in large-x region constrained to be
small by requirement that ΔG(x) ≤ G(x); at
Q2=10 GeV2: ΔGtrunc[0.3,1.0] ≤ 0.03
At small x, ΔG(x) can differ from DSSV
beyond errors without violating fundamental constraints
Small-x
x≤0.05
RHIC
Range
Large-x
x≥0.2
0.05<x<0.2
7
Next steps: ALL for di-jets in STAR resolve xq, xg
projections for 500 GeV and ∫Ldt = 300 pb-1 P=70%
o Information on x1
and x2, forward jets
give access to
lower x !
o
x1 x2  M / s
De Florian,
Frixione,
Courtesy
STAR
Signer and Vogelsang
NPB 539 (1999) 455 and
PC for present calc.
Extending ΔG(x) to Lowest x with
Forward di-Jets or High pT di-Hadrons
E.g. back-to-back neutral pion pairs in the PHENIX forward EMC
back-to-back:
selects di-jets
Central Arm
forward EMC
(3.1<|η|<3.9)
(|η|<0.35)
Jet -1 (π0, pT,1)
Trigger
particle, pT,1
dφ
Associate
particle, pT,2
Back-to-back hadrons:
trigger (pT,1) and associate
(pT,2<pT,1) in separate jets:
large forward boost
Jet 2 (π0, pT,2)
x1 >> x2
 500M PYTHIA events ≈ 0.014 pb-1 , only hard
QCD processes, soft processes eliminated by
pT cuts (study by Cameron McKinney, UIUC)
enhances q-g
fraction to ~ 60%
9
Selecting x2 with pT Cuts: x2 Decreases and q-g
Fraction Increases with Magnitude of pT cut
x1
x2
log x
di-hadron pT cuts and resulting <x2>
pT,1>1.0 GeV/c,
pT,2>0.5 GeV/c
<x2> ≈ 0.014
pT,1>2.0 GeV/c,
pT,2>1.0 GeV/c
<x2> ≈ 7.2*10-3
pT,1>3.0 GeV/c,
pT,2>1.5 GeV/c
<x2> ≈ 4.8*10-3
Projected ALL(π0) in MPC for different ΔG(x) at low x
for ∫Ldt = 300 pb-1, √s = 500 GeV (RHIC W program, 2011 to 2015)
Increasing Δgtrunc [10-4,0.05]
ΔGtrunc= -.1
ΔGtrunc= -.2
ΔGtrunc= -.5
pT,1>3 GeV, pT,2 > 1.5 GeV error bars are statistical only
ΔGtrunc [10-4,0.05] = -0.1 will be observable
11
First Forward EMC ALL , Run 2009 at √s=200 GeV
and Projections for Run 2011, √s=500 GeV
ALL forward EMC clusters run 2009
ALL forward clusters, projected 2011
DSSV-MAX PYTHIA
DSSV NLO
First step towards acquisition of
large integrated luminosity for
ΔG(x) at small x !
12
(II) W-production in polarized p-p
pp  W  eν
first results
pp  W  μν
look at ongoing run
Quark and Anti-Quark Helicity Distributions
from Inclusive ALe,μ in W-Production
De Florian at Berkely RSC meeting Nov, 2009
• Large Q2, knowledge of
FFs not needed
• pQCD analysis of
inclusive lepton AL
• DSSV analyzed MC data
of 200 pb-1 and 800 pb-1
from STAR and PHENIX
• Significant improvement of
knowledge with 200 pb-1
First exploratory run at
√s=500 GeV in 2009
P ~ 35%
∫Ldt ~ 9 pb-1
STAR, run 2009, √s= 500 GeV:
Parity Violating AL in p+p  high pT e
Phys.Rev.Lett. 106 (2011) 062002
Jacobian Peak for e- and e+
AL for e- and e+
Not yet (!) sensitive to quark and
anti-quark helicity distributions
15
W-Cross Sections for p-p:
PHENIX
&
ATLAS
Phys.Rev.Lett. 106 (2011) 062001
arXiv:1012.5382 [hep-ex]
Consistent with NNLO QCD
16
News from Present p-p Run at √s = 500 GeV
First full length 500 GeV run
PHENIX Muon Trigger
Installed & Operating
muTr trigger
electronics(JSPS)
P ~ 45 % (max above 50%)
goal is P ~ 50%
L ~ 5 x 1031 cm-2s-1
goal is L ~ 10 x 1031 cm-2s-1
o AnDY commissioned
succesfully
o muon/W trigger in PHENIX
in operation
o ∫Ldt ~ 40% of planned, due
to major hardware failure
in cryo system (repaired).
RPCs in Urbana (NSF)
RPCs in PHENIX (NSF)
FPGA based
level-1 trigger
processors
17
PHENIX Muon Trigger Performance
muTracker trigger efficiencies
RPC-Inner Ring Efficiency
problems to be solved:
RPC-gas -> mixture & pressure differentials
timing -> RPC south is 1 beam clock late
Taking data with muTr part of trigger in run 2011, use RPC
offline for background rejection
First Look at Data from Fast Production (Ralf Seidl)
Arbitrary normalization
(III) Transverse Spin
Inclusive AN
Collins or Sivers effects
AN in Very Forward Neutron Production
using the Zero Degree Calorimeter
 Large negative SSA observed for xF>0
 Diffractive physics
 Highly useful as local polarimeter for PHENIX
neutron
At Hard Scale: AN  0 , QCD Test !?
AN 
mq
s
example, mq  3MeV , s  20 GeV , AN  104
22
Experiment: Sizeable SSA Observed
over Large Range of Scales !
Experiment: AN >> 10-4 for 4 GeV < √s < 200 GeV for charged pions !
ZGS √s=4.7 GeV
AGS √s=6.5 GeV
FNAL √s = 20 GeV
RHIC √s = 200 GeV
π+
π-
Soft effects due to QCD dynamics in hadrons
remain relevant up to scales where pQCD can
be used to describe the scattering process!
from Christine Aidala, Spin 2008 and
Don Crabb & Alan Krisch in then Spin
2008 Summary, CERN Courier, 6-2009
23
AN vs xF almost unchanged for
√s=19.4, 62.4 and 200 GeV
24
Origin of Large SSA for Hard Scattering -Two Solutions: Final State vs Initial State
(I) “Transversity” quark-distributions
and Collins fragmentation
STAR, PRL-92:171801, 2004
Correlation between proton- und quark-spin
and spin dependent fragmentation
 q( x)  H1 ( z2 , k 2 )
Quark transverse
spin distribution
Collins FF
(II) Sivers quark-distribution+
First measurement at
RHIC √s = 200 GeV
Correlation between proton-spin and
transverse quark momentum
q
1T
f
( x, k )  D ( z)
2

Sivers distribution
h
q
(III) Initial or final state twist-3+
Qiu/Sterman and Koike
+ unified picture: Ji, Qiu, Vogelsang
and Yuan in PRL-97:082002, 2006
25
BRAHMS: AN for Charged Pions vs pT
and xF at √s=62.4 GeV and √s=200 GeV
√s=62.4 GeV
AN increases with
xF (valence quarks)
0.4<pT<0.6 GeV/c
AN increases with
pT ? Limited pT range!
0.5 pT<0.6 GeV/c
200 GeV
0.6 pT<0.8 GeV/c
0.8 pT<1.0 GeV/c
1.0 pT<1.2 GeV/c
1.5<pT<2.0 GeV/c
2.0<pT<2.5 GeV/c
√s=200 GeV
0.5<pT<0.75 GeV/c
1.0<pT<1.25 GeV/c
1.25<pT<1.5 GeV/c
26
STAR Run 2008: pT Dependence
of AN at √s=200 GeV Ogawa at CIPANP 2009
Consistent with zero for
all pT
Positive xF
Negative xF
Decrease as ~1/pT expected is not observed.
AN constant from pT>2.5 GeV. Need more statistics to extend
measurement to pT > 4 GeV !
Expectations for AN with PHENIX MPC and
Transverse Spin Running in 2012 or 2013
Red: Zhong-Bo Kang possible pT dependence
if all even orders of twist expansion contribute
Blue: pT dependence if sub-leading twist dominates
28
BRAHMS: AN for Charged Pions ,
Kaons and Protons at √s=200 GeV
p
K
p
Large AN for K-  significant Sivers asymmetries for
sea quarks ?!
Large AN for anti-Proton unexplained.
Another Surprise:
AN for Eta Mesons larger than for Pions !
p  p   X

STAR arXiv:0905.2840 (Heppelmann, DIS08)
s  200 GeV
AN(η) > AN(π0) for 0.55 < xF < 0.75
AN

AN
p
 0.361  0.064
 0.078  0.018
Possibly large effects
in the fragmentation for
eta-mesons?
STAR
Understanding of AN in terms of Collins
and Sivers Effect: Work in Progress!
Future goal:
Extract Sivers and transversity quark distributions from
global anlaysis to all SIDIS, pp and e+e- data!
Present work: Extract Sivers + transversity from SIDIS and e+eand predict AN in pp
BRAHMS π+,-
STAR π0
Presently: Poor agreement with many problems to solve!
Universality, evolution, pdf and fragmentation
functions not sufficiently known.
For example, note the impact of un-polarized FFs
thick line DSS
thin line Kretzer
Global analysis of SIDIS & e+eAnselmino, Boglione, D’Alesio,
Kotzinian, Murgia, Prokudin, Turk
Phys. Rev. D75:05032,2007
AN calculation from D’Alesio, 2008
31
Measurements to Isolate
Different Mechanisms
o Transversity & Collins
o Sivers
Ideas for Measurements of
Transversity Observables at RHIC
p p 
 l  l  X
Drell Yan:
Required luminosity not available at RHIC.
pp  
   X
Spin dependent Lambda-FF unknown.
Measure Λ-FF in e+e- ?
p p 
[ jet  h]hemisphere  X
Collins effect in jets; possible in STAR ?
 hadron ID at high p, z- and ϕ-resolution ?
pp  
[p   p  ]hemisphere  X
Di-hadron intereference fragmentation
function. IFF data available from e+e- Belle !
33
Interference Fragmentation –IFFfor Di-Hadrons at Mid-Rapidity in PHENIX
AUT compatible ~0 with
present statistics
Dilution from gg processes!
Future:
 Update with more statistics
from runs 2012 and 2013
With 2012 & 2013 statistics
extend measurements in the
forward direction for smaller
g-g process fraction and
large x !
34
IFF Measurement in e+e- at BELLE
PHENIX & STAR collaborators have joined Belle: BNL-Illinois-Indiana-RBRC-RIKEN
electron
e e  (p p  ) jet1 (p p  ) jet 2 X
Artru and Collins,
Z. Phys. C69, 277 (1996)
Boer, Jakob, and Radici,
ϕpair-2
PRD67, 094003 (2003)
θ
p 2
p  2 q2
zpair-1
quark-2
spin
z1,2 relative pion-pair
momenta
p
q1
quark-1
spin
ϕpair-1

1
p 1
zpair-1
positron
sin 2 
2
2
a12 ~
H
(
z
,
M
)
H
(
z
,
M
1
pair

1
pair

1
1
pair

2
pair
2 )
2
1  cos 
35
Belle IFF- Asymmetries vs Hadron Pair
Momentum Fraction zi BNL-Illinois-Indiana-RBRC-RIKEN
a12
to be published
this month …
9x9
z1 z2 binning
z1
Ideas for Measurements of Sivers
Observables at RHIC
pp  
 h  0   X
Precision measurement of AN at mid-rapidity.
pp  
 jet  jet  X
Back-to-back correlations for jets.
pp  
 heavy flavor  X
AN for heavy flavor.
pp  
 jet  X
AN for inclusive jets.
pp  
   X
AN for direct photons.
pp  
 jet    X
AN in jet-photon production.


pp 
 l  l

AN in Drell Yan.
significant improvements from upgrades:
forward calorimeters
+ silicon vertex
detectors
New Experiment:
AnDY
37
AN from p0 and h+/- at Central Rapidity
Anselmino et al, Phys. Rev. D 74 094011
Process Contribution to
p0, η=0, s=200 GeV
PRL 95, 202001 (2005)
Constrain gluon Sivers effect
using PHENIX 2002 p0 data !
38
AN from p0 Update
• pT range extended from 5 to 12 GeV/c
• Results consistent with previous PHENIX analysis
• Statistical uncertainties reduced by more than factor of 30
39
Impact of 2006 + 2008 Data Sets
0.02 < xSampled < 0.08
Sea quark Sivers maximized +
Gluon Sivers function
Gluon Sivers parameterized to
1 sigma of data
u + d quarks Sivers w. no gluon or sea
quark contribution.
Low pT pi0 at mid-rapidity is not sensitive
to valence quark Sivers function
Naïve expected impact of new
data.
Maximized Gluon Sivers function
Violates <kT> of partons = 0
Theoretical analysis
to be carried out.
AN in Di-Jet Production in STAR
y
x
z
STAR: PRL-99:142003,2007
protonspin
180º
S1
Di-jet pT
Di-jet pT
  
K k
Gluon
radiation
parton
Di-jet
with 0
Additional kT kick to jet
axisAfrom
Sivers effect
N consistent
 Boer & Vogelsang, PRD 69, 094025 (2004)
41
Summary
o Gluon Spin contribution constraint for 0.05 < x < 0.2
use di-jet and di-hadron measurements to probe
x-dependence ΔG(x) and forward jet production to
reach low x, x~0.001.
o W-program has started with electrons (STAR & PHENIX)
and muons (PHENIX). Luminosity accumulation will take
3-4 runs.
o Precision data on AN are available. Exciting new Drell Yan
experiment at IP2: AnDY (see Akio Ogawa’s talk).
Initial measurements to isolate Collins -and Siversasymmetries. Much improvement from detector upgrades
and increased statistics.
42
Backup
43
STAR Run 2006: pT Dependence
of AN at √s=200 GeV PRL 101,222001
For given η strong
correlation between
xF and pT:
AN(pT)
integrated
over xF
AN(pT, xF)
AN increases with pT up to
pT ~ 3 GeV/c -- Models: AN ~ 1/pT
44
PHENIX: AN vs XF
for p0’s at √s=62.4 GeV
AN = 0 for xF < 0 no sizeable
asymmetries at small x!
Larger forward asymmetries
at higher pseudo-rapidity, η ?
Limited by statistics and
correlations between xF,
pT and η !
Sivers Effect in Heavy Flavor Production
 Heavy flavor production gives sensitivity to gluon Sivers effect .
 Significant improvement with vertex detector upgrades.

Work needed to connect theory and experimental observable.
Gluon Sivers=Max
Measurement for m-
Gluon Sivers=0
Calculations for D mesons
Anselmino et al, PRD 70, 074025 (2004)
A RHIC and US-Japan Contribution to Transverse Spin
Analysis: Measurement of the Collins Effect in e +eAnnihilation into Quarks at Belle
BNL-Illinois-RBRC-RIKEN
Measurement of the Collins
effect in e+e- at Belle:
Belle Collins
asymmetries
& global fit
e++e-  π+ + π- + X
~ Collins(z1) x Collins (z2)
electron
p
q1
q2
quark-2
spin
quark-1
spin
p
Collins FF
extracted from
Belle data.
positron
47
Collins Effect in Quark Fragmentation
J.C. Collins, Nucl. Phys. B396, 161(1993)

sq
q

k

sq

ph

ph 

h, p h

k
: quark momentum
: quark spin
: hadron momentum
: transvers e hadron momentum
z h  Eh Eq
 2 Eh
s : relative hadron momentum


ph 
Collins Effect:
Fragmentation of a
transversely polarized
quark q into spin-less
hadron h carries an
azimuthal dependence:


 

 k  ph   s q
 sin 
48
General Form of Fragmentation Functions
Number density for finding
hadron h from a transversely
polarized quark, q:

Dq ( z, ph  )  D1q ,h ( z )  H1q ,h ( z, ph2 )
h
unpolarized FF




ˆ
k  ph   s q
zM h
Collins FF
49
IFF- a12 vs Invariant Mass
a12
8x8 m1 m2 binning
Systematic errors shown.
a12 increases with m1 and m2
reaches |a12 | ~ 0.1 at large mi.
m1 [GeV/c2]
50
The Sivers Effect
Sivers:
Correlation between the transverse spin of the
proton and the transverse momentum kT of
quarks and gluons in the proton (link to orbital
angular momentum?)
Sivers function:
proton
D. Sivers 1990
Sp
Observed asymmetry:
Sp
proton


( Pˆ  kT )  S P
q
2
AN  q( x1 )  f1T ( x2 , k  ) 
M
51
ANphoton+Jet : An Alternative Test of the Process
Dependence of the Sivers Effect at RHIC
Weighted moment of ANphoton+Jet
Bachhetta, Bomhof, D’Alesio, Mulders, Murgia
Phys.Rev.Lett.99:212002,2007.
no process dependence
Measurement:
AN in jet-photon production
forward photon η > 2
jet -1 < η < 0
Correct process dependence
Much less luminosity hungry …
NSAC milestone for transverse spin
(HP-13, 2015) ! Reachable at STAR.
Requires FOCAL upgrade in PHENIX.
52