RHIC and STAR: New Tools for Studying the Protons Spin

1
RHIC and STAR New Tools for Studying the
Protons Spin
J. Sowinski
Indiana University
Collaboration

• The STAR Spin Physics Program
• Spin related hardware improvements to STAR
• Important constraints on DG along the way
  di-jets and p0s
• Sivers Functions from jets at mid-rapidity
• Forward tracking upgrades

2
Where does the protons spin come from?
p is made of 2 u and 1d quark
S ½ S Sq
Explains magnetic moment of baryon octet
p
BUT partons have an x distribution and there are
sea quarks and gluons
Check via electron scattering and find quarks
carry only 1/3 of the protons spin!
Sz ½ ½ DS DG Lzq Lzg
3
DIS used to investigate proton structure
Virtual g momentum transfer Q,
energy loss n Measure structure functions
vs. xBjorken Q2/2MNn and Q2
Weak dependence
pparton/pproton in the momentum frame

• Spin structure functions
• Polarized beam and target
• Polarization along/opposite beam direction
• Measure difference between parallel and
  anti-parallel spin combinations
• Quark can only absorb photon of opposite helicity
  Good spin analyzer

Gluons are neutral and hence relatively
insensitive to DIS
4
Symbol abcdefghijklmnopqrstuvwxyz ABCDEFGHIJKLMNOP
QRSTUVWXYZ 1234567890- !_at_()_ \
,./ ltgt?
Univ math1 abcdefghijklmnopqrstuvwxyz ABCDEFGHIJKL
MNOPQRSTUVWXYZ 1234567890- !_at_()_ \
,./ ltgt?
abcdefghijklmnopqrstuvwxyz ABCDEFGHIJKLMNOPQRSTUVW
XYZ 1234567890- !_at_()_ \ ,./ ltgt
?
abcdefghijklmnopqrstuvwxyz ABCDEFGHIJKLMNOPQRSTUVW
XYZ 1234567890- !_at_()_ \ ,./ ltgt
?
5
Unpolarized and Polarized Structure Functions
Parton Model
2
All fixed-target data
x 3.2310?-5
x 0.008
x 0.08
x 3.2310?-3
Unpolarized DIS Structure Function(x,Q2)
g1p
Polarized DIS Structure Function(x,Q2)
x 0.75
x 0.65
100
10
1
1
100
104
103
10
105
(GeV/c)2
Q2 (GeV/c)2
Q2 (GeV)
Without e-p collider data, reduced range of x and
Q2 leaves gluon spin poorly determined
Small scaling violations with Q2 give sensitivity
to gluon distributions
2
6
Blümlein Böttcher fit to polarised data of
EMC, E142, E143, E155, SMC, Hermes
7
Parton Distribution Functions
Gluons carry 1/2 the momentum (mass)!
Maybe we shouldnt be surprised that quarks carry
only 1/3 of protons spin
DG is poorly constrained, even solutions with
zero crossing allowed
8

• Hermes and COMPASS
• Detect leading hadrons from jets
• Compass open charm
• Have announced results
• Kinematically difficult to cover broad x range
  w/o e-p collider

Semi-Inclusive Deep Inelastic Scattering

• Open charm clean
• Leading hadrons have backgrounds from QCD-Compton
  and vector meson dominance

C. Schill, Spin 2004
9
DG via partonic scattering from a gluon
Know from DIS
g-jet coinc. rare
Measure

A P 3P 3a
g
part
LL
LL
pQCD
Jets and p0s
DG
LL
Prefer

• Dominant reaction mechanism
• Experimentally clean reaction mechanism
• Large a

Heavy flavor rare

LL
10
The Relativistic Heavy Ion Collider
2.4 mile circ. Collider

• Heavy ions
• Au-Au
• Lighter ions
• Asymmetric d-Au

PHOBOS

• 4 detectors
• STAR
• PHENIX
• PHOBOS
• Brahms
• pp2pp (p-p only)

The first polarized p-p collider!
11
Polarized Proton Operation at RHIC

Year 2002 2007 ?s 200 GeV Improving L and
Pol. Develop 500 GeV

• 2002 2003 2004 2005
  2006 2007
• L (s-1cm-2) 0.5x1030 2x1030
  3x1030 8x1030 17x1030
  48x1030
• Int. L (pb-1 T/L) 0.3 pb-1 0.5/0.4
  0.5/0.4 4/7 28 86
  
• Pol. 0.2 0.3
  0.40 0.45 0.65
  0.70

12
Why is RHIC the 1st Polarized Collider?
Overcoming Depolarizing Resonances

• Spin is vertical in RHIC
• In-plane components precess rapidly at spin tune
  (turns/orbit) Gg (180 at 100GeV)
• Resonances from repetitive perturbations
• Stray in-plane fields can rotate spin into plane,
  resonance when Gginteger
• Focusing fields can also do this, resonance when
  Gginteger betatron tune
• Huge number of these in RHIC
• Solution Siberian Snakes
• Idea from Novosibirsk
• First proof in tests at IUCF Cooler ring

Siberian Snake
Res. Strength

• Siberian Snake
• Helical dipole magnet
• Rotates spin 180o no net beam defl.
• 2 in RHIC
• 100 spin transmission to 100 GeV
• Same technology for spin rotators

13
AGS never designed for spin (space) Resonances
handled individually For strongest ones spin
flips!
Solution is stronger partial snake ready for
testing in coming run expect 70 pol. from AGS
Ebeam
10 ms bin width
Raw asymmetry AN ? Pbeam
36 ny
36 ny
G?
Imperfection resonances spin flip at every Gg
n Intrinsic resonances spin flip at Gg 36
ny and 36 ny
f.o.m.P4L !
14
How do we (know we) get longitudinal polarization?
Yellow
Beam-Beam Counters
BBC West
BBC East
duspinGg udefl Gg180 for 200GeV p-p
Yellow
Turn on spin rotator
eCNI online
2003
2003
eBBC 0
eBBC (x10-3)
AN BBC 1 comparable to ACNI
eBBC vert
Bottom
eCNI (x10-3)
Sideways asym. also 0
15
What I am not going to tell youeach is a seminar
in itself

• More accelerator physics Classical mechanics
• Polarized ion sources Atomic physics
• Polarimetry diffractive processes
• Polarization calibration polarized targets and
  atomic physics
• Heavy ion program
• Phobos, Brahms, pp2pp, much about Phenix

16

• Physics puts premium on a large solid angle
  detector
• Kinematic coverage
• Coincident solid angle
• Jets are big r0.7

h - ln(tan(u/2)
h 0
h 1
h -1
h 2
17

• Sacrifice full coverege for good particle ID
• m detection forward
• RICH detectors central
• Some Pb glass
• Solenoidal B field and tracking
• Fast DAQ

Upgrades planned to improve coverage for spin
18
Tools The STAR Detector at RHIC
At the heart of STAR is the worlds largest Time
Projection Chamber

• STAR Detector
• Large solid angle
• Not hermetic
• Tracking in 5kG field
• EM Calorimetry
• Slow DAQ (100Hz)
• Sophisiticated triggers

19
Detector
Lum. Monitor Local Polarim.
Triggering
Beam-Beam Counters
Special interest for spin
2lt?lt 5
h - ln(tan(u/2)
h0
h -1
h2
Triggering
Endcap EM Calorimeter
Forward Pion Detector
1lt?lt 2
-4.1lt?lt -3.3
Time Projection Chamber -2lt?lt 2
Solenoidal Magnetic Field 5kG
2003
2004
2005
Tracking
20
Endcap ElectroMagnetic Calorimeter

• Pb Scint sampling calorimeter
• 21 radiation lengths
• 720 projective towers
• Depth Segmentation
• 2 preshower layers, e/h p0/g disc.
• High position resol. SMD p0/g disc.
• Postshower layer e/h discc.
• L0 trigger- high tower, jet patches

21
Shower Maximum Detector EEMC
7,632 tot. channels
SMD profiles for a 9 GeV ?0 candidate

• Resolves closely spaced showers for p g ID
• 7000 individually read out scintillating strips
• U and V plane in each 308 sector
• Essentially no coverage gaps

22
1st p0 finder for pp 2004 production data
Preliminary
200 GeV pp (2003)
HT-low trigger ET gt 1.92 GeV
M 156 MeV s 36 MeV
M GeV
Online tower-only p0 reconstruction, 200 GeV AuAu
Charged tracks matched to fired EEMC towers
for a 62 GeV AuAu event. 2004 Data MIPs
0.3GeV g
All events
Mixed events
Difference
23
Structure Installation
I never thought anything would make the EEMC
look small. But
9/21/03 Even tornados cant stop us!
0.2 deg from vertical. John couldnt you do
better?
Rotate to vertical on strongback
Almost there. 10/1/02
24
Megatile Production Line

• There are 120 different versions of megatiles
• 1/3rd of megatiles installed. Another 1/3rd
  complete.
• Also machine SMD modules and parts of mechanical
  structure
• Will run through next summer to complete all
  megatiles at present rate. Machining two shifts.

25
SMD Construction
Fiber router layer attached and fibers installed
Strips wrapped in aluminized mylar and glued on
FR4 substrate
Fibers terminate in connector
Strips have hole down middle for wls fiber
2 SMD layers in 3 planes no gaps
26
PMT Boxes
(MA)PMT boxes and electronics on back of poletip

• 12 PMTs per box, one 68 sector of towers
• PMT housings built in Dubna
• CW bases from Dubna
• Box structure Texas AM
• MAPMT testing and LEDs at KSU
• Assembly of PMT boxes to Valpo. U.
• 4 sectors installed

MAPMT Boxes

• SMD and pre/post-shower
• 12 MAPMTs (16 ch) per box
• 192 ch. FEE internal to box
• Final testing before product.

MAPMT, CW base and 16 ch. FEE
PMT Box
Internal Fiber Harness
27

Some Assembly Required!
Upper half mounted 8/1/03
28
PMT Boxes

• 12 PMTs per box, one 68 sector of towers
• PMT housings built in Dubna
• CW bases from Dubna
• Box structure Texas AM
• MAPMT testing and LEDs at KSU
• Assembly of PMT boxes to Valpo. U.
• All 720 channels installed
• 97 fully functional

(MA)PMT boxes and electronics on back of poletip
Box lifter for installation and repair
MAPMT Boxes

• SMD and pre/post-shower
• 12 MAPMTs (16 ch) per box
• 192 ch. FEE internal to box
• 4 Sectors - 3000 ch. inst.
• 99.5 working

MAPMT, CW base and 16 ch. FEE
PMT Box
Internal Fiber Harness
29
Installed During 2003 Shutdown

• Upper half mechanical structure
• All remaining active elements 2/3rds
• All clear fiber bundles from detectors to rear of
  poletip
• All remaining PMT boxes and electronics Full
  tower coverage

• 1/3rd of MAPMT boxes (16) and associated
  electronics SMD and pre/post-shower
• Diagnostics LED and laser for all installed
  detectors
• HV system for full detector

30
Barrel ElectroMagnetic Calorimeter
One module 40 towers
24 modules FY02 60
modules FY03 90 modules
FY04 All modules (plan
all elect.)FY05

• Scinti. Pb sandwich sampling EMC
• 4800 projective towers (2p, -1lth lt1)
• Shower Max Detector-gas detector-18K strips
• Pre Shower Detector (first 2 layers)
• High tower trigger 1x1 (?, f) jet trigger

120 the last one! August 2004
pTgt3GeV
31
Quark Gluon Compton Scatteringp 1 p
Direct g 1 Jet
pT
32
Extracting DG from g Jet
Measure
Pb1, Pb2 Beam polarization 0.3-0.4 (0.7 in plan)

1 1 N - RN-
__ __ __________
N (-) Yields for same vs. opposing beam
helicity
ALL
Pb1 Pb2 N RN-
R Ratio of luminosity for different beam
helicities
From QCD
A1p
Exactly equal to A1p as measured at pol. DIS No
Fragmentation function
x
High x quark with high polarization
33
Separation of direct gs from p0s

• Significant p0 background
• Reduced to better than 11
• Isolation cut
• SMD particle ID
• Preshower at higher energies
• Background subtraction of p0 sample from g
  sample
• Increases errors on DG(x) by factor of 1.5-2.0

34
Jet finders are currently being tested on 2004
data
And the away-side jet in central Au-Au
collisions disappears!
Phys. Rev. Lett. 91, 072304 (2003).
35
Kinematic Reconstructionevent by event

• Assume 2 body kinematics
• Neglect kT
• Measure ujet, Eg and ug
• Extract x1, x2 and u
• Assume larger of x1 and x2 xquark
• Assume lesser xgluon
• Make cut that one x gt 0.2

36
Expected DG Results based on simulations
Eventually gives best determ. of Dg(x) for
existing experiments.
500 GeV data not in plot
But L low in coming years
500 GeV important for low x
Comparison to competing expmts.
Fit integral of DG(x) determined to 60.5
Potential for constraints on DG(x)
37
Inclusive jets are sensitive to DG
But signal is mixture of multiple partonic
subprocesses
-1 lt h lt 1
ALL
Dg0 Dggmax Dg- gmax Dggstd
STAR Run-5 projections P0.4 and Lint7pb-1
Leads to small but significant ALL in 2005
-
-
-
-
-
(1/10 of these stats from 2004 currently being
processed)
-
-
-
-
-
Jet
ETjet GeV
38
1/3 of the jet energy is EM Use EM cals for
triggering jets p0s carry same physics
Jager, Stratmann, Vogelsang NLO pQCD calculations
hep-ph/0404057
-1lthlt1
Simulation
EEMC 1lthlt2
-1lthlt1 BEMC
Significant const. on DG expected in 2005 data
(1/10 stats. from 04 being analyzed) (error bar
estimates too small pTlt6 GeV)
Signal changes in EEMC due to different partonic
subprocess contrib.
39
p0
First Results
Direct g
S.S.Adler et al., Phys. Rev. Lett. 91, 072301
(2003).
Preliminary
p0
K. Okada Spin 2004
Y. Fukao Spin 2004
40
Large Analyzing Powers at RHIC

• First measurement of AN for forward ?0 production
  at ?s200GeV

Phys. Lett. B261(1991)201 Phys. Lett.
B264(1991)462
Similar to FNAL E704 result at ? s 20 GeV
E704 motivated studies of transverse spin effects
Now applied to STAR results

• Sivers spin and kT correlation in initial state
  (related to orbital angular momentum?)
• Collins Transversity distribution function
  spin-dependent fragmentation function
• Qiu and Sterman (initial-state) / Koike
  (final-state) twist-3 pQCD calculations

41
STAR Detector
TPC -1.0 lt h lt 1.0 FTPC 2.8 lt h lt 3.8 FPD
h 3.8 (pp) h 4.0 (pp, dAu)

• Forward p0 Detector (FPD)
• Pb-glass EM calorimeter
• Shower-Maximum Detector (SMD)
• Preshower

42
Do these pQCD processes apply to forward
scattering at ?s200GeV ?
Forward production at ?s ltlt 200 GeV not well
described by fixed-order pQCD calculations
(Bourelly and Soffer, hep-ph/0311110)

• Run-2 STAR data at
• ??? 3.8 (PRL 92, 171801 (2004) hep-ex/0310058)
• ??? 3.3 (hep-ex/0403012, Preliminary)
• NLO pQCD calculations (Vogelsang) at fixed ?
  with equal factorization and renormalization
  scales pT

• STAR data consistent with Next-to-Leading Order
  pQCD calculations, unlike at smaller ?s

43
Do we understand forward p0 production in p p?
Bourelly and Soffer (hep-ph/0311110)
NLO pQCD calculations underpredict the
data at low vs from ISR sdata/spQCD appears to
be function of q, vs in addition to pT
44

• Confirms previous results
• Rapid h dependence
• Small AN at negative xF

Released at Spin 2004 A. Ogawa
45
Analyzing Powers at Mid-Rapidity
Do processes invoked in forward scattering show
up at large angles?
STAR Collab. Phys. Rev. Lett. 92 (2004) 171801
Measure
Jet
D. Boer and W. Vogelsang, Phys.Rev. D 69 (2004)
094025
Sivers Function correlation between kT and spin
Jet
46
Partonic kT from Dijet Analysis
Sivers Effect Prediction
T. Henry
4.1 x 10 -4
0.03 0.05
sf 0.23 0.02
D. Boer and W. Vogelsang, Phys.Rev. D 69 (2004)
094025
8 lt pT1,2 lt 12 GeV ?1,2 lt 1
AN
kT ?ltkT2gt ET sin (sf) ET 13.0 6 0.7sys ?
Trigger Jet
STAR agrees well with World Data on Partonic kT
df
kT distribution

• Curves are for various gluonic Sivers functions
• Connection to partonic orbital angular momentum
• Suppressed by Sudakov effect

47
Transversity
48
(No Transcript)
49
Examples of Future Transverse Spin Measurements

• Attempt to determine transversity by measuring
  transverse poln xfer from incident proton to
  final-state quark. For latter, rely on
  fragmentation analyzing powers (Collins or
  interference fragmentn fcns.) calibrated at ee?
  collider.
• Probe transverse quark and gluon motion in
  incident protons by measuring (leading-twist) AN
  with respect to kT direction, inferred from
  misalignment of not quite back-to-back dijets.

• Measure quark transvty2 via (pQCD-allowed) ANN
  for qq-dominated high-pT dijets.
• AN for W prodn sensitive only to transverse
  quark motion preference in pold p, can be used
  in principle to map out flavor-dependence.

Measurements will progress among channels and
observables as L,P improve and theory provides
better guidance!
50
How is the Sea Generated?
Middle Ground chiral symmetry large Nc
models ps Goldstone bosons nucleons solitons
p
p
n
51
What pp Spin Observables Does QCD Allow?

• L- and R-handed q sectors are separate (chiral
  symm.) but equal (parity).

• ? need 2-spin observables probe g poln with
  highly polarized quarks!

LO QCD spin observ-ables
52
Until the 1st Measurements
NMC and

• Non-perturbative processes seem to be needed in
  generating the sea
• What about flavor asymmetry in the spin of u and
  d for different models?

_
_
53
_
E866 Results
Polarized q Flavor Asymmetry

• d(x)?u(x) and Du(x)?Dd(x) better for Q2
  evolution
• E866 Results are qualitatively consistent with
  pion cloud models, instanton models, chiral quark
  soliton models, etc.

• Most quark-based models predict
  ?01?u(x)??d(x)dx ? ?01d(x)?u(x)dx
• Most meson-based models disagree

B. Dressler et al., Chiral Quark Soliton Model
Predictions
?2 (5 GeV)2


x(?u??d)
?2 (600 MeV)2


x(d?u)
54
Measurements by Hermes have been used to extract
Du(x)?Dd(x)
_
_
PRL 92 (2004) 012005
_
_
x(Du(x)?Dd(x))
e
h p,p-,k,k-

• Combined p and d data fit for all identified
  hadron A1s
• Input unpolarized q(x) pdfs
• JETSET fragmentation fctns.
• Simplifying assumptions on symmetries in sea
  quark dist.
• Results Controversial

Dressler et al., Eur. Ph. J. C14 (2000) 147
55
W(-) Production in p-p at ? s 500 GeV/c2
Dressler et al. predict large sensitivity

• V-A coupling
• only LH u and RH d couple to W
• Likewise LH d and RH u to W-
• Only LH Ws produced
• Neutrino decay gives preferential directionality
  in decay

Parity violating single spin asymmetry
AL (Helicity flip in one beam while averaging
over other)
_
_
_
_

ALW- u(x1)Dd(x2)d(x1)Du(x2)
Allows kinematic separation especially for W- in
EEMC
56
STAR Simulations of W Prodn
By L.C. Bland

• Separate ud from du by detecting the e and e?
  from ud ? W? e du ? W?? e-
• Sensitivity to u vs. d comes from which beam
  spin is flipped and h distributions
• W momentum in direction of higher-x parton
  (usually q as opposed to q)
• PV decay of L-handed W ? CP ? in W rest frame
  e (e?) emitted prefly along (opposite) W (W?)
  spin
• ? e? focused in q direction while e is more
  spread out
• W? prod., e? in endcap strongly emphasizes
  dtoward uaway collisions
• Less clean separation for W ? e
• Separation of antiquark and quark polarizations
  is kinematically cleanest in endcap region

57
x Sensitivity and Possible x Dependence
Extraction
One may approximately reconstruct x1,2 from ?(e),
pT(e) event-by-event from 2-body fusion 2-body
decay, i.e., neglecting W width transverse mom.
Works best in Endcap and for W-
Ws in Barrel Endcap
Ws in Endcap
58
AL for leptons from W production
Leptonic kinematics, no detector simulation other
than acceptance
Nadolsky and Yuan, Nucl. Phys. B666 (2003) 31.

_
_
Large Dd-Du
? s 500 GeV 800pb-1
59
Good Tracking is Essential
Existing tracking degrades with h for hgt1

• Fewer pad rows hit as h increases
• Pad row density decreases for hgt1.5
• Miss SVT

Nadolsky and Yuan

• Need to determine sign of pT 20-40 GeV/c
  electrons (minimum requirement)
• pT/ETcal would help e/h discrimination
• Has to be done in high density of tracks
• Need TPC design resolution

These are all crucial for tracking in the region
-1 lt h lt 1 as well, where tracking upgr.
also of use
60
Future Upgrades Inner and Forward Tracking

• Sensitivity in forward region
• Requires tracking for up to pT40 GeV e/e- sign
  determination
• Tracking upgrade

Parity violating long. asymmetry in W production
allows extraction of
Nadolsky and Yuan, Nucl. Phys. B666 (2003) 31.
61
Upgrade STAR Forward Tracking for W ? vs. W ?

• Add inner tracking for forward region Si with
  50 mm resolution
• Add tracking in front of endcap GEM with 100
  mm resolution
• Simulations by N. Smirnoff for uniform 30 GeV p-
  illumination of endcap region show added
  detectors can eliminate sign misidentification
  (sagitta 2.5mm)
• GEMS as fast detectors can help with pile up

62

• Timescale
• Integrated Cent. Tracking Upgrade planned
  installation for FY09 run
• Forward Tracking planned installation for FY10
  run
• Some plans have first 500 GeV p-p in FY08 some in
  FY09.
• Any earlier is unrealistic CAD upgrades
  complete for FY07
• 500 GeV also for DG Dont want to delay waiting
  for Fwd Track. Upg.
• Note high luminosity and polarization at 500 GeV
  needs to be developed
• Need 800pb-1
• 200pb-1 is useful

From 20-Year Planning Study for RHIC 12/31/03
63
What can be done before tracking upgrade?

• Restrict to -1 lt h lt 1
• Not so many counts lost
• But lose the range with best separation between
  flavors
• Dd and Du reasonable
• Du and Dd look hard to separate
• First run likely to have low luminosity
• Extending range to 1.? needs detailed simulation
  expect 1.5
• Add EEMC SMD point
• Add vertex from other tracks
• Displace vertex away from EEMC?
• First 500 GeV run w/o full tracking upgrade
  would still be a good start

_
_






64
STAR Spin Physics Program Near and Long Term

• Proton Spin Structure
• Gluon contributions to the protons spin
• ltDGgt jets and p0s
• q g g jet, DG(x)
• Heavy flavors
• Spin/momentum correlations
• Sivers Functions dijets
• Collins Functions Leading particle correl. in
  jets
• Transversity
• Flavor separated q, q Origin of the sea
• Standard Model tests
• Parity violation in jet production

_
65
Background slides follow
66
Flavor Dependence of Sea Antiquark Spin
Preferences

?u(x) vs. ?d(x) will provide critical clues to
the non-perturbative nature of the nucleon sea

• Many models of nucleon spin structure (like
  chiral quark sol-iton model at right) predict
  larger polarized than known un-pold sea flavor
  asymmetry.

• Latest (contro-versial) HERMES results do not
  support this picture!

HERMES SIDIS results analysis
67
Bjorken x Sensitivity of STAR W Prodn

• 1? ?e ? 2 probes either asymmetric qtoward qaway
  or ? symmetric dtoward uaway collisions.
• Sensitivity generally good for xq ? 0.1, where
  chiral soliton model predicts largest flavor dep.
  of ?q/q

• One may approximately reconstruct x1,2 from
  ?(e), pT(e) event-by-event from 2-body fusion
  2-body decay, i.e., neglecting W width
  transverse mom.
• x1x2 MW / ? s 0.16 _at_ ?s 500 GeV
• (x1?x2)/( x1x2) tanh ?emeas ? ?erest tanh
  ?emeas ? cosh-1(MW / 2pTmeas)

68
Does pQCD Apply to RHIC pp Collisions?
Absolute differential cross sections for
inclu-sive ?0 prodn in ?s 200 GeV pp collisions

• Next-leading-order pQCD reproduces measured
  cross sections well, at both mid and forward
  rapidity, with common parton densities and
  fragmentation functions, down to pT 2 GeV/c !
  This is not the case in lower-energy fixed-target
  experiments!

69
EMC Shower-Maximum Detectors ? ?/?0 Discrimination
20-60 GeV ?0 3 lt ? lt 4
Fine-grained scint. strips (endcap, FPD) 5 R.L.
deep ? trans-verse shower profile, different for
? vs. ?0.
Simulation at right ? 80 ? retention 80 ?0
rejection over pT 10-20 GeV/c range in endcap,
using residuals to profile fit with single-photon
peak shape.
70
EMCs Facilitate Triggering on Jet Events
Or pp at full luminosity with TPC pileup?
Algorithms developed to filter out pileup tracks
-- by demanding consistency with vertex from
tracks leading to prompt EMC hits, and/or with
EMC hits themselves work well in simulations.
Unfortunately, pp luminosities obtained to date
not high enough to test in real data!
71
Transverse Spin Measurements Have Stimulated
Rapid Development of Theory

• Large AN seen by STAR for forward ?0 (cross
  section consistent with NLO pQCD), and sizable
  azimuthal asyms. in HERMES SIDIS, can arise from
  various contributions.
• Unraveling these requires complementary
  measurements for a number of channels, involving
  transverse analyzing powers, spin correlations
  and polarization transfer.
• Potential payoff mapping out transverse spin
  vs. transverse motion preferences of partons in
  transversely polarized proton.

pQCD
Non-pert.
Non-pert.
Factorization
Hard hadronic ampl. PDFs ? hard partonic ampl.
? fragmentn fcn.
AN?0 can arise, in principle, from (e.g.)
?
?
? ?q in p ? aNpartonic (mq/pT)?s ? 0 for light
quarks
?
?
? ?q in p ? dNNpartonic poln xfer ? Collins
fragmentn
?
?
?
?sq?(pq?kTfragment)? ? 0
?
?
?
?sp?(pp?kTquark)? ? 0
?
?
? Sivers correln in p ? kT / pT effect on ?
partonic
?
?
?
?
?
?
? ?q in p ?sq?(pp?kTquark)? in unpold p?
aNNpartonic kT / pT effect on ? partonic
72
DG via partonic scattering from a gluon
Know from DIS
Measure
g-jet coinc. rare

A P 3P 3a
g
part
LL
LL
pQCD
Jets and p0s
DG
Prefer

• Dominant reaction mechanism
• Experimentally clean reaction mechanism
• Large a
• But jet and p0 rates are sufficient to give
  significant DG const. in 2005 data

Heavy flavor rare

LL