recent . results for a ll and current constraints on d g

1Kieran Boyle—Gluon Polarization in the Nucleon—June 16, 2008

Recent . Results for ALL and Current Constraints

on GKieran Boyle

(RIKEN BNL Research Center) Outline:

• Quick Physics overview

• RHIC, PHENIX, and ALL

• Results from Run5 and Run6

• Constraints on G

• Current and Future Issues


Accessing G in pp scattering

• From Run5 and Run6, we consider an array of probes:– large statistics, specified trigger– large statistics, low trigger efficiency,

PID only for pT<2.8 and pT>4.7 GeV/c

– Direct photon low statistics, no G2 very clean signal– low statistics – multiparticle “cone”– J/

G2 Gq q2

Hard Scattering Process

2P2 2x P

1P

1 1x P

0

with ~25%, G not well constrained, L?

ALL~ agg * G2 + bgq * G q + cqq q2


Why ALL?

• If f = q, then we have this from pDIS• So roughly, we have

+- =

+++

+=

From ep (&pp)(HERA mostly)

pQCD NLO From e+e-


Measuring ALL

Helicity Dependent Particle Yields , , , , , etc

(Local) Polarimetry Relative Luminosity (R=L++/L+-) ALL

+ - = Opposite helicity =

++ = Same helicity

+

+=


RHIC

BRAHMS & PP2PP (p)

STAR (p)PHENIX (p)

AGS

LINAC BOOSTER

Pol. Proton Source

Spin RotatorsPartial Siberian Snake

Siberian Snakes

200 MeV Polarimeter AGS Internal PolarimeterRf Dipoles

RHIC CNI (pC) PolarimetersAbsolute Polarimeter (H jet)

Year [GeV] Luminosity [pb-1] (recorded) Polarization [%] Figure of Merit (P4L)

2003 * 200 0.35 27 0.0019

2004 * 200 0.12 40 0.0031

2005 * 200 3.4 49 0.20

2006 * 200 7.5 62 (57) 1.1 (7.9)

2006 * 62.4 0.08 48** 0.0042*** Longitudinal** Online


PHENIX Detector

BBC

ZDCZDC

EMCal detection

• Electromagnetic Calorimeter (PbSc/PbGl):• High pT photon trigger to collect trigger

to collect 0's, ’s, ’s • Acceptance: ||x • High granularity (~10*10mrad2)

• Drift Chamber (DC) for Charged Tracks• Ring Imaging Cherenkov Detector (RICH)

• High pT charged pions (pT>4.7 GeV).Relative Luminosity• Beam Beam Counter (BBC)

• Acceptance: 3.0< 3.9• Zero Degree Calorimeter (ZDC)

• Acceptance: ±2 mradLocal Polarimetry• ZDC• Shower Maximum Detector (SMD)


Local Polarimetry at PHENIX• Use ZDC and SMD to measure a

L-R and U-D asymmetry in forward neutrons (Acceptance: ±2 mrad).

• When transversely polarized, we see clear asymmetry.

• When longitudinally polarized, there should be no asymmetry.

BLUE YELLOW

Raw

as

ymm

etry

Raw

as

ymm

etry

BLUE YELLOW

Raw

as

ymm

etry

Raw

as

ymm

etry


Measured Asymmetry During Longitudinal Running (2005)

<PT/P>=10±2(%)

<PL/P> =99.48±0.12±0.02(%)

LR 2/NDF = 88.1/97p0 = -0.00323±0.00059

LR

UD 2/NDF = 82.5/97p0 = 0.00423±0.00057

XF>0 XF>0

XF<0 XF<0

2/NDF = 119.3/97p0 = 0.00056±0.00063

UD 2/NDF = 81.7/97p0 = -0.00026±0.00056

Fill NumberFill Number

<PT/P>=14±2(%)

<PL/P>=98.94±0.21±0.04(%)

• Measurement of remaining transverse component spin pattern is correct


Relative Luminosity

• R = L++/L+- ; L = luminosity.

• Luminosity is given by the number of BBC triggered events (NBBC).

• For estimate of uncertainty, measure ALL in the ratio of our two luminosity detectors

Year [GeV] ALL

2005 * 200 2.3e-4

2006 * 200 5.4e-4*** Longitudinal** Online


Strategy:1.Measure Cross Section to

confirm that pQCD is applicable to data

2.Measure ALL to extract G


pQCD works

arXiv:0704.3599 [hep-ex]

0 @ 200 GeV Direct @ 200 GeV


Calculating 0 ALL

1. Calculate ALL(+BG) and ALL(BG) separately.

2. Get background ratio (wBG) from fit of all data.

3. Subtract ALL(BG) from ALL(+BG):

ALL(+BG) = w· ALL() + wBG · ALL(BG)

+BG region :±25 MeV around

peakBG region :

two 50 MeV regionsaround peak

Two Photon Mass Spectrum


Final Run5 0 ALL

• Data is compared to GRSV model with several input values of G.

• To interpret data, we use pQCD

Question: At what pT does soft physics contribution become small enough to allow comparison to pQCD models

9.4% scale uncertainty not included

GRSV: M. Gluck, E. Reya, M. Stratmann, and W. Vogelsang, PRD 63 (2001) 094005.

Adare et al., PRD76 (2007)051106


Soft physics• By comparing 0 data with

charged pion data, which has very good statistics at low pT, can estimate soft physics contribution

• Fitting an exponential to the low pT charged pion data (pT<1 GeV/c) gives an estimate on the soft physics contribution.

• Fit result: = 5.56±0.02 (GeV/c)−1

2/NDF = 6.2/3• From this, we see that for

pT>2 GeV, the soft physics component is down by more than a factor of 10.

exponential fit


Final Run5 0 ALL

• Measure below 1 GeV (~100% soft physics). ALL = 0.002±0.002.

• Above pT=2 GeV/c, soft physics contribution estimated to be <10%

• Contamination from soft physics asymmetry is small

• Note that DSSV used data with pT>1 GeV

Theory 2/NDF C.L. (%)

GRSV-std 10.9/8 20

GRSV G=0 12.7/8 12

GRSV G=G 256/8 0.00

GRSV G=-G 58.4/8 0.00

• Extend this approach

Adare et al., PRD76 (2007) 051106

9.4% scale uncertainty not included


Run6 0 ALL

• Run6 scaling error based on online polarization values. Final scaling error expected to be ~10%

• Grey band is systematic uncertainty due to Relative Luminosity, and is pT independent.

• Run6 Data favor “GRSV G=0” over GRSV-std

Theory 2/NDF CL(%)

GRSV-std 23.8/8* 0.25

GRSV G=0 7.9/8* 44

*Theoretical uncertainties not included


Sensitivity of 0 ALL to G

Scaling Errors not included

• Stat errors only• What about:

– exp. syst.?– theory

uncertainties?


Exp. Systematics• Polarization uncertainty:

– scales both the data and the error.

– Very small effect on interpretation in terms of G

– Final results for Run6 will be similarly small with final pol. values

• Issues:– Already below required limit of

5% per beam– Note that this is fully correlated

between all data points measured at RHIC (STAR and PHENIX) for a given year and energy

Adare et al., PRD76 (2007) 051106

P-PP+P


Exp. Systematics• Relative Luminosity uncertainty:

– Effectively a common shift of all data.– As the results for ALL are about zero, has

more significant effect than polarization uncertainty, but still small

• Relative Luminosity– In Run6, most significant experimental

uncertainty in extracting G– Larger in Run6 than Run5.– Expected to become worse in the future:

• higher luminositymore multiple collisions• squeezing longitudinal bunch profile may

increase problem

– Note that this is entirely correlated between all measurements at PHENIX, and so must be handled correctly

– To reduce effect, will REQUIRE spin flipper

ALL+ALL|RALL-ALL|R


Sensitivity of 0 ALL to G

Scaling Errors not included

present x-range


From pT to xgluon

Log10(xgluon) NLO pQCD: 0 pT=29 GeV/c xgluon=0.020.3 GRSV model: G(xgluon=0.020.3) ~ 0.6G(xgluon =01 )

Each pT bin corresponds to a wide range in xgluon, heavily overlapping with other pT bins


Sign Ambiguity

• Dominance of two gluon interaction at low pT present 0 ALL data cannot determine sign of G.

• Solution:– Higher pT higher FOM (P4L)

– Look to other probes:• Direct Photon, in which G2 term

is not allowed at lowest order.• Charged pions

G2 Gq q2


2P2 2x P

1P

1 1x P

0


Sign Ambiguity II

Fraction of pion production

• With PHENIX RICH, charged pions above 4.7 GeV can be identified.

• At higher pT, qg interactions become dominant and so qg term in ALL becomes significant allowing access to the sign of G

• Expect if G>0, then: ALL(+)>ALL(0)>ALL(-)

• Additional data will help.


Direct photon

• First step, direct photon cross section.

• ALL results from Run5 with very small statistics

• Will need significant data set for this measurement (more than Run6)

• Background ALL estimates increase errors by ~30% at low pT

– Currently studying other methods to estimate background

gq


2P 2 2x P

1P

1 1x P

qq

Gluon-photon compton dominant

RUN5


p+p + X

g2 gq q2


2P 2 2x P

1P

1 1x P

• Similar analysis technique as with 0, using two photon decay channel.

• Independent measure for gluon polarization.

• Due to different fragmentation functions, expectations are slightly different.

• Statistically limited currently


Extending the x range:0 ALL at s = 62.4 GeVfor high x measurement


0 ALL @ s=62.4 GeV

GRSV: M. Gluck, E. Reya, M. Stratmann, and W. Vogelsang, Phys. Rev. D 53 (1996) 4775.

• Short run with longitudinal polarized protonsALL

• Grey band is systematic uncertainty due to Relative Luminosity

pQCD works measure ALL


Comparison with 200 GeV

• At fixed xT, 0 cross section is 2 orders of magnitude higher at 62.4 GeV than at 200 GeV

• Converting to xT, we can get a better impression of the significance of the s=62.4 GeV data set, when compared with the Run5 final data set.

Run5 200GeV final 2.7pb-1 (49%)Run6 62.4GeV prelm. 0.04 pb-1 (48%)


A new theoretical fit• De Florian, Sassot, Stratmann

and Vogelsang (DSSV) have recently included 0 data at 200 and 62 GeV from PHENIX.– First use of any pp data.

• Only Experimental stat and syst (summed in quadrature) uncertainties – NO THEORETICAL

UNCERTAINTIES were considered.

• Results for G were driven by RHIC data in the range x=[0.05-0.2]


Conclusions• Cross sections have been measured at s=200 GeV and s=62 GeV and NLO pQCD describes our data well.

• Run5 and Run6 0 ALL are a significant constraint on the gluon spin contribution to the proton in the x range [0.02,0.3].

• Experimental systematic uncertainties are small, but are beginning to be significant.– Spin Flipper at RHIC will be needed to reduce Rel. Lumi.

uncertainty

• More luminosity at s=200 GeV will enable us to address the G sign ambiguity with charged pions as well as direct photons.

• Extending the measured x range will be needed, and has begun with measurements at s=62 GeV.


Backups


Scale


Charged pion pt spectrum

• All high quality ERT trigger tracks with an associated RICH hit with matching, dc acceptance, EM prob, E/p and momentum dependent energy cuts applied.


Polarimetry: Direction• Require longitudinally polarized beams at

PHENIX to measure ALL, and so use spin rotators.

• How can we know direction? Consider a single spin asymmetry (SSA).

SSA found in forward neutron production at RHIC in 2002 [Bazilevsky et al. PLB 650:325-330,2007]

from M. Togawa


ATT

• If large, could lead to false ALL.• Theory expects it to be small• Experimentally found to be consistent with

zero.


What about polarized?• Recall the unpolarized

data.• In the polarized case,

much less data exists, covering much smaller x and Q2 range.

• Even with this small range of data, polarized quark distribution is reasonably well constrained.

~25% of proton spin

most of proton spin in either gluon spin, or OAM.

• But from fixed target DIS, gluon is very poorly constrained.

G


Polarimetry results• HJet results for target

asymmetry are consistent between Run5 and Run6.

• Results from beam asymmetry differ, indicating change in polarization.

Run5

Run6


Extending x range is crucial!

Gehrmann-Stirling models

GSC: G(xgluon= 01) = 1 G(xgluon= 0.020.3) ~ 0

GRSV-0: G(xgluon= 01) = 0 G(xgluon= 0.020.3) ~ 0

GRSV-std: G(xgluon= 01) = 0.4 G(xgluon= 0.020.3) ~ 0.25

GSC: G(xgluon= 01) = 1

GRSV-0: G(xgluon= 01) = 0

GRSV-std: G(xgluon= 01) = 0.4

Current data is sensitive to G for xgluon= 0.020.3


Extend x Range

present x-ranges = 200 GeV

Extend to lower x at s = 500 GeV

Extend to higher x at s = 62.4 GeV

• To measure G, need as wide an x range as possible.• Planned Upgrades such as the VTX, FVTX and Nosecone will help • By measuring at different center of mass energies, we can reach

different x ranges.• We can extend our x coverage towards lower x at s =

500 GeV. Expected to start in 2009.• We can extend our x coverage towards higher x at s =

62.4 GeV. Run6


Direct Production• Gluon Compton

scattering dominates– At LO no fragmentation

function– Small contribution from

annihilation

hep-ex/0609031

Run-3

2P 2 2x P

1P

1 1x P

g

qqgq

)(ˆ)(

)(

)(

)(

2

2

1

1 qgqaxq

xq

xg

xgA LLLL


SMC Fit to DIS

Polarized gluon distribution poorly constrained


Backup: SIDIS for G

HERMES preliminaryHERMES preliminary


Backup: GFrom M. Stratmann


Proton Spin


Recent ALL Measurements

arXiv:0704.3599 [hep-ex] 0

Many independent probes to understand gluon spin


Results From DIS• Basic Idea (Cartoon):

– How many quarks have spin in direction of proton spin polarized PDF

• Look at DIS

• So measure Asymmetry of proton polarized along photon vs. against photon

lq =

= X


Nucleon Spin Structure

Naïve parton model:

1989 EMC (CERN):=0.120.090.14

Spin Crisis sdusdu

vv du 2

1

2

1

Determination of G and q-bar is the main goal of longitudinal spin program at RHIC

Gluons are polarized (G) Sea quarks are polarized:

Gqq 2

1

2

1

For complete descriptioninclude parton orbital angular momentum LZ:

ZLGqq 2

1

2

1


Polarized PDF from DISAsymmetry Analysis Collaboration

M. Hirai, S. Kumano and N. Saito, PRD (2004)

• Valence distributions well determined

• Sea Distribution poorly constrained

• Gluon can be either positive, 0, negative!


Learning about structure• Primary method is Deeply Inelastic Scattering (DIS).• Scatter off quark with momentum fraction x, and

understand probability to find a given quark at a given x, known as the Parton Distribution Function (PDF).

Q2 = -q2

recent . results for a ll and current constraints on d g

Documents