STARSTAR

Len K. Eun

For Collaboration

Transverse Spin Physics at STAR

26th Winter Workshop on Nuclear Dynamics Ochos Rios, Jamaica, January 2010

STARSTAR Transverse Spin Physics at STAR

• Collins effect: asymmetry comes from the transversity and the spin dependence of jet fragmentation.

• Sivers effect: asymmetry comes from spin-correlated kT in the initial parton distribution

2

PRL 92, 171801 (2004)

√s=200 GeV, <η> = 3.8

SP kT,q

p

p

SP

p

p

Sq kT,π

Large Transverse Single Spin Asymmetry(SSA) in forward meson production persists up to RHIC energy.

Photons have asymmetryJet vs. Photon sign flip predicted

No asymmetry for the jet axis

Di-jet, photon-jet not exactly back to back

STARSTAR STAR Forward Pion Detector (FPD)

• STAR forward calorimeters have gone through significant upgrades since run3. • In run6, the original FPD remained in the east, while the west FPD was expanded to FPD++.• The east FPD consists of two 7X7 Pb-glass modules, EN and ES. During run6, it was placed at the “far” position. (x-offset~30cm, <>~3.7)

3

Run6 Configuration

STARSTAR Forward 0 Single Spin Asymmetry(SSA)

1N

d d N S S NA

Pd d N S S N

At √s=200GeV, 0 cross-section measured by STAR FPD is consistent with the NLO pQCD calculation. Results at <>=3.3 and <>=3.8 have been included in the DSS global pion fragmentation function analysis. (Phys.Rev.D75(2007) 114010)

4

 Phys.Rev.Lett.101:222001,2008

4

 Black Points: Phys.Rev.Lett.101:222001,2008

From Spin2008 talk by J.Drachenberg, arXiv:0901.2763

For Fixed XF, the asymmetry AN does not fall with pT as predicted by modelsand perhaps expected on very general grounds.

STARSTAR pT Dependence of AN

NLO PQCD does describe the size and shape of this forward 0 cross section.

Model calculations (Sivers, Collins or twist-3) can explain the XF dependence of AN.

XX Flat or increasing dependence of AN on pT is very difficult to understand within any of these frameworks!

5

 Phys.Rev.Lett.101:222001,2008 

N

d dA

d d

p p M X

p p M X

1) Nominally (perhaps not significantly) larger asymmetry for Eta than Pi0.2) Large Uncertainty in Eta AN.

19.4s GeV ~1 /Tp GeV c

STARSTARPrevious Observation of Transverse SSA Forward Production of Eta Meson by FNAL Exp 704

6

proton & proton proton & anti-proton

STARSTAR Run6 FPD Acceptance for 0 and Eta

• 7x7 FPD has limited acceptance for Eta mesons. At 40GeV, a symmetrically decaying Eta needs to point to the center of the FPD to fit in. Acceptance improves greatly at higher energy.• 0 reconstruction efficiency starts to drop over 60GeV, where the separation between two photons for symmetric decay becomes on average less than 1 cell width.

Fast SimulatorThe ratio of N(reconstructed particles) to

N(generated particles with CoM within FPD) 2)(1 2

SinZEm tot

2040

60

80

100 0

0.2

0.4

0.6

0.8

02468

2040

60

80

100

Etot(GeV)

Z

Se

pa

ratio

n(F

PD

ce

ll) Eta

2040

60

80

1000

0.2

0.4

0.6

0.8

02468

2040

60

80

100

0

Etot(GeV)

Z

Se

pa

ratio

n(F

PD

ce

ll)

21

21

EE

EEz

7

Event Cuts• 2 photon events• Etotal>25GeV• Hardware threshold nominally at 25GeV• “Center Cut” for 2 CoM defined as

STARSTAR Event Selection for Run6 FPD Eta Analysis

222 )15.0()()65.3( Tan

• Etot: Detector summed energy• Z and photon separation: Fitted photon energy/locations• Reconstructs on the entire FPD • Vertex set at zero for all events

2)(1 2

SinZEm tot

691.3pion 664.3eta

0 mass region with Center Cut in black mass region with Center Cut in red

Di-Photon Center of Mass (EN&ES)

8

STARSTAR Eta Signal in Run6 FPD

0 Mass Cut

.085 .185GeV M GeV

Eta Mass Cut

.48 .62GeV M GeV

• Center Cut• 3 columns for 3 energy bins• Each column shows a single plot in log and linear scale.

Di-Photon Invariant Mass Spectra in 3 Energy Bins

9

AN(xF) will be reported for di-photon events in these two shaded mass regions. We will not separate contributions from backgrounds under the Eta and 0 peaks.

STARSTAR Mass Dependence of AN

• Forward asymmetry clearly reveals the shape of two mass resonances.

• There is an “asymmetry valley” in between 0 and Eta mass regions.

1. Nphoton = 22. Etotal > 40GeV3. No Center Cut4. Average Beam

Polarization = 56%

p p M X M 200s GeV

STAR 2006 PRELIMINARY

10

STARSTAR AN(xF) in 0 and Eta Mass Regions

1. Nphoton = 22. Center Cut ( and )3. Pi0 or Eta mass cuts4. Average Beam

Polarization = 56%

0.361 0.064NA

0.078 0.018NA

.55 .75FX

For , theasymmetry in the Eta mass region is greater than 5 sigma above zero, and about 4 sigma above the asymmetry in the 0 mass region.

.55 .75FX

p p M X M 200s GeV

11

0I

1I *Assume mixing angle: ~ 19.5P , '

12

𝜂= 1ξ3൫𝑢𝑢ത+𝑑𝑑ҧ−𝑠𝑠ҧ൯

𝜂′ = 1ξ6൫𝑢𝑢ത+𝑑𝑑ҧ+2𝑠𝑠ҧ൯

𝜋0 = 1ξ2൫𝑢𝑢ത−𝑑𝑑ҧ൯

STARSTAR Should AN be larger for Eta than π0 ?

• Gluons or η has Isospin I=0.• u quark has Isospin I=1/2• π0 has Isospin I=1.

• But we expect both mesons to come from fragmentation of quark jets.

• For Sivers Effect: Asymmetry is in the jet and should not depend on the details of fragmentation.

• For Collins Effect: Asymmetry reflects fragmentation of the quark jet into a leading η or π0 meson. Differences in fragmentation could relate to:

• Mass differences?• Isospin differences?• Role of Strangeness?• But Collins Effect Should be suppressed when Z~1

New for Run 8New for Run 8: FMS

• Stack of 1264 lead glass cells, roughly 18 X0 in z.

• Located at far West side of Hall, at the opening to RHIC tunnel. Faces blue beam.

• 7.5 meters from interaction point

East FPD (since runs 3)

STARSTAR STAR Forward Meson Spectrometer (FMS)13

With installation of FMS, STAR With installation of FMS, STAR EM calorimeter coverage spans EM calorimeter coverage spans most of the pseudo-rapidity most of the pseudo-rapidity region from region from -1<-1<<4.<4.

STARSTAR

• EMC CalorimetersbEMC -1<η<1eEMC 1<η<2FMS 2.5<η<4FPD movable

•Tracking TPDFTPC

FMS Greatly Enhances STAR EM Coverage14

15

Readout of 1264 Readout of 1264 channels of FMS channels of FMS provided by QT provided by QT boards. Each board boards. Each board hashas

• 32 analog inputs32 analog inputs

• 5-bit TDC / channel5-bit TDC / channel

• Five FPGA for data Five FPGA for data and triggerand trigger

• Operates at 9.38 Operates at 9.38 MHz and higher MHz and higher harmonicsharmonics

• Produces 32 bits for Produces 32 bits for each RHIC crossing for each RHIC crossing for triggertrigger

• 12-bit ADC / channel12-bit ADC / channelDesigned and built at UC Berkeley/SSL

QT boardboard

Cockcroft-Walton Cockcroft-Walton HV bases with HV bases with computer control computer control through USB. through USB. Designed/built in Designed/built in house for FEU-84.house for FEU-84.Designed and built at Penn State University

Small Cell PSU TypeSmall Cell PSU Type224 of 476224 of 476

Forward Meson Spectrometer Forward Meson Spectrometer Lead Glass From FNAL E831804 cells of 5.8cm5.8cm60cmSchott F2 lead glass

15

From SPIN08 talk by N. Poljack, arXiv:0901.2828

Estimate tot. ≤ 1.2 stat.

stat.errors only

STAR Preliminary

16

STARSTAR Preliminary Run8 FMS π0 AN

Azimuthal Angle Dependence of ANAN vs. xF

Consistent with previous measurements

Run3+5+6

17

STARSTAR pT-dependence of π0 AN

Run8

• Large solid angle of FMS allows simultaneous mapping of xF vs pT with greater statistics

Blue: From Spin2008 talk by J.Drachenberg, arXiv:0901.2763

 Black: Phys.Rev.Lett.101:222001,2008

F.o.M. was smaller in run8 than in run6 More statistics needed

• The Sivers effect predicts spin dependent left/right bias in kT due to parton orbital angular momentum.

• For di-jets at mid-rapidity, this initial state kT asymmetry leads to a spin dependent shift in the azimuthal opening angle for the jets.

• In contrast, di-jet measurements are insensitive to the Collins effect, which produces an asymmetry within a jet.

STARSTAR Sivers Effect with Mid-Rapidity Di-Jets

180open

1

2

spin

di-jetbisector

kTx

18

STARSTAR Sivers Effect with Mid-Rapidity Di-Jets

PRL 99 (2007) 142003

• The observed AN is an order of magnitude smaller than what was seen in SIDIS by HERMES• Might be due to cancellations between initial state and final state, u and d quark, and/or very small gluon Sivers effect.

19

STARSTAR Summary

1. The STAR Forward Pion Detectors (FPD) at RHIC measured cross-section for 0 meson in <>=3.3~4.0 region during √s=200GeV p+p collision. It was found to be consistent with pQCD calcuations.

2. From RHIC run3 to run8, the FPD measured large forward single spin asymmetry, AN, for 0. The xF dependence of AN was qualitatively consistent with theoretical predictions. pT dependence, however, differed significantly from predictions based on all currently existing models

3. In addition to 0, Eta mesons were observed in the east FPD during RHIC run6. We measured the single spin asymmetry in the 0 and the Eta mass regions, at <>~3.65 and xF above 0.4. We found the AN in Eta mass region to be ~4 standard deviation greater than the AN in 0 mass region from 55GeV to 75GeV. (xF=0.55~0.75)

4. Forward Meson Spectrometer (FMS), commissioned in RHIC run8, greatly enhances EM coverage of STAR. The preliminary results from run8 show that 0 AN(xF) is consistent with previous measurement, while the azimuthal angle dependence of AN is as expected. With additional transverse running, we can also significantly improve our measurement of AN vs. pT utilizing much improved acceptance of the FMS

Len Eun

20

STARSTAR

Back Up

1. Majority valence quark in polarized proton.• for proton • for antiproton.

2. Minority valence quark• for proton • for antiproton.

1. Pion containing only majority quarks: large positive AN.

2. Pion containing only minority quarks: large negative AN.

3. Pion containing both majority and minority quarks: intermediate positive AN.

uu

dd

Pattern from Previous Transverse SSA Measurements of Forward Pion / Eta Production with High Energy Polarized Proton / Antiproton Beams

STARSTAR

6

STARSTAREta and 0 Energy Sharing (Z) Distribution

21

21

EE

EEz

Z Z

Nev

ents

0 mass region with Center Cut Eta mass region with Center Cut

9

STARSTAR Mass Dependence of AN

• Yellow beam asymmetry clearly reveals the shape of two mass resonances.

• There is an “asymmetry valley” in between 0 and Eta mass regions.

1. Nphoton = 22. No energy cut3. With Center Cut4. Z< 0.855. Average Yellow Beam

Polarization = 56%

p p M X M 200s GeV

STAR 2006 PRELIMINARY

7

STARSTAR AN(xF) in 0 and Eta Mass Regions

1. Nphoton = 22. Center Cut ( and )3. Pi0 or Eta mass cuts4. Z < 0.855. Average Yellow Beam

Polarization = 56%

.55 .75FX

For , theasymmetry in the Eta mass region is greater than 5 sigma above zero, and about 4 sigma above the asymmetry in the 0 mass region.

.55 .75FX

p p M X M 200s GeV

8

018.0078.00 NA

065.036.0 NA

Run8 FPD east result : arXiv:0901.2763 (James Drachenberg– SPIN08)

Negative xF consistent with zero

20

STARSTAR pT-dependence of π0 AN for Negative xF