1

Measurement of the cross section and the single transverse spin asymmetry of forward neutrons from p-p collision

s at RHIC-PHENIX

Manabu Togawa

for the PHENIX collaboration.

Kyoto University / RBRC

2

Outline• Physics motivation

• Measurement of forward neutron at PHENIX– Experimental setup– Analysis procedure

• Result of the cross section.– Very forward case.

• Results of the single transverse spin asymmetry – xF-dependence

– Comparing tagged with/without charged particles

• Summary

3

Motivation• In the forward neutron production, there are some inte

resting behaviors.– In ISR experiment, pp √s=30~63GeV, cross section of forward ne

utron production at low pT was measured to be larger than at high pT. [left figure]

– In the RHIC IP12 experiment, pp √s=200GeV. We found large single transverse-spin asymmetry AN (-10%) [right figure]

– PHENIX data can shed new light to understand the forward neutron production mechanism.

<AN>=-0.1090.0072

Kinematics ±2.8mrad

Forward region

xFNucl. Phys. B109 (1976) 347-356

4

• Does Feynman scaling hold when going to RHIC energy?– From the ISR result, C.S. are wel

l scaling by xF at 30<sqrt(s)<63 GeV..

– Forward neutron C.S. has peak structure and it is well described by pion exchange model.

• What is the mechanism of neutron asymmetry ?– pion exchange model

• Asymmetry can be appeared by interference of spin flip.

– Twist-3 model, Siverse, Collins…

Eur.Phys.J.A7:109-119,2000

xF

xF dependence√s >=200 GeV - cross section - asymmetry

pT dependence√s >= 200 GeV - asymmetry

5

Setup

Collision point

BBC

1800 cm

Dx magnet

Schematic view from simulation. - GEANT3 (Geisha) - From the pythia simulation, Main backgrounds are gamma and proton.

ZDC

10*10cm 2.8 mrad

3 module

150X0 5.1λI

Proton shower event(In case of proton mom = 50GeV/c)

Forward counter

veto

SMD

1 ZDC

SMD

Neutron position can be decided by centroid method.

6

• Data set– In this analysis, √s=200 GeV data which was took from 2003

to 2005 is used. The data is corrected by single neutron trigger with/without MinBias trigger.

• MinBias is defined as detecting charged particles in BBC(3.03.9

• Calibration– ZDC : Looking the 1 neutron peak at heavy ion data.– SMD : Relative gain correction is done by cosmic-ray data.

• Simulation ( GEANT3 with pythia v.6220 or single particle generator )– Background and efficiency after neutron ID cut.– Unfold the neutron energy.– Smearing effect of asymmetry caused by position cut and re

solution.

Analysis outline

7

Cross section

Integrated pT area : 0<pT<0.11xF GeV/cin each point.

Cross section result is consistent with ISR data and xF scaling holds at such higher energy.

pT distribution is assuming ISR result

8

Single transverse spin asymmetry• Smearing effect is evaluated by simulation.• For asymmetry calculation, square root formula is used.

RLRL

RLRLN

NNNN

NNNN

PPA

11

xpos

ypo

s

Expected mean pT (Estimated by simulation assuming ISR pT dist.) 0.4<|xF|<0.6 0.088 GeV/c 0.6<|xF|<0.8 0.118 GeV/c 0.8<|xF|<1.0 0.144 GeV/c

neutronneutronchargedparticles

AN

Without MinBias

-6.6 ±0.6 %

With MinBias

-8.3 ±0.4 %

preliminary

* Cut is not same

9

How about charged particles at BBC ? • Asymmetry is 0 when data is selected by BBC only.• Finite asymmetries are shown tagged with forward

neutron. 4– Opposite direction as neutron (2.28±0.55±0.10)*10-2

– Same direction as neutron (-4.50±0.50±0.22)*10-2 9

Neutron analyzing power can be large with charged particles.

AN(X) < 0, AN(Y) > 0,AN(Y’)

proton

Ypolarizedproton

Y’n (neutron)

Y : charged particles

pion exchange model

pol direction

N*(*) n+Y’ ??

N* ???

Y

will be analyzed with high statistics and good polarization data.

Top view

preliminary

10

Summary• The cross section and the single transverse spin asym

metries of forward neutron production in √s=200GeV polarized p-p collision are measured at RHIC-PHENIX.

• Cross section – It is consistent with ISR data at pT ~ 0 (GeV/c). xF scaling

holds in this high energy.

• Single transverse spin asymmetry– xF-dependence looks flat, negative polarity.

– Analyzing power of forward neutron tagged with charged particles are lager than without them.

• Prefer the pion exchange model and like diffractive process(?).

Theoretical calculation is very welcome.

11

Future analysis

• pT dependence

– Important parameter same as xF.

• Charged particle asymmetry tagged with forward neutron.– with high statistics and good polarization data.

12

-5cm 5cm

RNN

RNNA

Horizontal scan We can see finite asymmetry in other energy RUN !

√s=62.4 GeV √s=410 GeV

13

backup

14

Cross section calculation

F

neutron

F dxLumi

N

dx

d 1

TTF

F

dppdx

xd

pd

dE

11

2

13

3

.3

312

Acc

TTFF

dpppd

dE

xdx

d

PHENIX data.

ISR data is converted to current cut. (0<pT<0.11xF GeV/c)

Input pT shape

pT=A*exp(-4.8pT)Assumed

Simulation input

comparing

• The total cross section with PHENIX neutron cut. is estimated. – pT is not evaluated from our data. As simulati

on input, ISR result is assumed.– In this analysis, radius<2cm from the collisio

n point (1800cm) is calculated. 1.1 mrad : 0<pT<0.11xF GeV/c 。

15

Asymmetry with different trigger

RLRL

RLRLN

NNNN

NNNN

PPA

11

• Smearing effect is evaluated by simulation.

• For asymmetry calculation, square root formula is used.

neutron neutronchargedparticles

Without MinBias With MinBias

16

Summary• The cross section and the single transverse spin asymmetries

of forward neutron production in √s=200GeV polarized p-p collision are measured at RHIC-PHENIX.

• Cross section

– It is consistent with ISR data at pT ~ 0 (GeV/c). xF scaling holds in this high energy.

• with 2cm radius cut in ZDC region (~1.1 mrad)

• Integrated pT range is 0~0.11xF (GeV/c)

• Single transverse spin asymmetry

– xF-dependence looks flat, negative polarity.

• Integrated pT range is 0.03xF~0.22xF (GeV/c)

– Analyzing power of forward neutron tagged with charged particles are lager than without them.

• Prefer the pion exchange model and like diffractive process(?).

Theoretical calculation is very welcome.

17

t-range

p

nF E

Ex

222 1

~ pFnF

F

F

T mxmx

x

x

pt

t

p (mp,0,0,Ep)

p (mn,pT,0,En)

xF <pT> (GeV/c) t (GeV2)

0.4~0.6 0.088 -0.458

0.6~0.8 0.118 -0.134

0.8~1.0 0.144 -0.033

ISR pT distribution is assuming.

18

Red/Black ~ 0.47 Red/Black ~ 0.71

Red/Black ~ 0.75 Red/Black ~ 0.73

Mean pT : 0.030Mean pT : 0.048

Mean pT : 0.043Mean pT : 0.070

Mean pT : 0.058Mean pT : 0.089

Mean pT : 0.072Mean pT : 0.102

19

Systematic error for C.S. • Systematic error in calibration

– Energy calibration ~5%– SMD relative calibration (it is including in SMD cut eff)– Trigger bias correction. < 10%

• Systematic error in simulation– Acceptance few % from pt distribution by SMD pos cut.– SMD cut eff. ~3%– Neutron ID

• SMD shape match. ~ 3% • Forward counter match. ~5%

– Background estimation. apply 100% error for BG ratio.• Mainly from K0 and proton • For cross section 8%

– 2 particle in 1 event ~ 5%• Systematic error in luminosity

– Cross section error 13% (22.9 mb * 0.591) for NoVtxCut• In total

– For shape after cut : ~11.6% For scaling of cross section : 17.1%

20

Systematic error for phi-asym.• Measurement : 1%

– Bunch shuffling result is ~2% of statistics, RUN5 local pol analysis note. (AN462)

– Center scan ~ 1%– Beam gas BG at ZDCN|S trigger : 0.2%

• Simulation uncertainty 4.1%– Simulation stat 2%– pT match : few%– BG estimation 3%– Center cut scan 2%

• Total : 4.2%

• Online polarization 20% scaling error.

21

Systematic error for xF-asym.• Estimation from measurement : 2%

– ~4% of statistics, by bunch shuffling. – Center scan ~ 2%– Beam gas BG at ZDCN|S trigger : 0.2%– Unfolding : calibration (bin by bin 18p)

• xF : 0.4~0.6 : 2.1% 0.6~0.8 : 0.3% 0.8~1.0 : 0.3%

• Simulation uncertainty 7.8%– BG estimation 3%– Position smearing estimated by center cut scan (Differences

between PHENIX and flat xF inputs) 7.2%– Unfolding : linearity (bin by bin 17p)

• xF : 0.4~0.6 : 12.1% 0.6~0.8 : 2.1% 0.8~1.0 : 2.1%

• Total : 8.1% of asymmetry value (not including blue term)

22

Calibration (ZDC)• Neutron cainbration is done by 1 neutron peak from heavy ion ru

n.

Select “coulomb” event by BBC

ZDC detector

10cm

10cm

We can see 100 GeV peakin CuCu event

Black : all eventRed : select center region.

pp eventafter calibrated.

23

Calibration (SMD)Relative gain of each channels are matched by cosmic ray data.

Cosmic peakTook by Alexei.

Compared with simulation data.See simulation term.

24

nID study using pythia input.

pid pid

1 Gamma 10 K0 long

2 Positron 11 K+

3 Electron 12 K-

5 Mu+ 13 Neutron

6 Mu- 14 Proton

8 Pi+ 15 Antiproton

9 Pi- 25 Antineutron

Black : sumGreen : gammaBlue : neutronRed : proton

n

p

ZDC threshold is ~5GeV

25

Correction of BG (estimation)• With SMD cut and center cut (r<0.5), BG are

mainly K0 and proton– From ISR paper,

E (GeV) K0 (%) Proton (%)

10-20 16.67 11.11

20-30 5.97 2.99

30-40 4.69 2.34

40-50 1.75 3.59

50-60 1.78 4.14

60-70 1.50 5.00

70-80 1.18 1.76

80-90 0 2.14

90-100 0 9.71

Pythia out.

Pythia out is agree K0 ration with ISR data. I used BG distribution after cut from the pythia. Error will be added 100% of this value

26

Neutron measurement stability

1BBCLLofN

neutronIDofN

Stability = 7.65762e-06/4.76148e-03*sqrt(315.5/95) = 0.3 %.

27

RUN5 pp MB looking 22.9 ± 9.7mb For neutron C.S. correct BBCLL1 cut efficiency.

cm

Red : BBCLL1(>0 tube)Blue : BBCLL1(noVtxCut) BBCLL1/NoVtxCut

0.591 ± 0.05 (for maximum error) 8.46%

BBCLL1 bias study.

28

Stability of neutron peak after cut

This peak

29

NORSH vs. SOUTH (nID num)

nID number are agree with SOUTH and NORTH after SMD cut efficiency Amplitude of C.S. will be same

30

BG estimation from pythia

• With BBC trigger– K0 : 1.16% proton : 1.60% sum : 2.76%

• Without BBC trigger– K0 : 0.97% proton : 2.27% sum : 3.24%

correction assuming A of K0 and p = 0.if 3%, asymmetry is smeared as 1.0/1.03 =

0.97

31

Radius distribution for calculating asymmetry

Center cut by r>0.5cm

Cut radius 0.5 1.0 2.0

Real data 0.0308 0.0320 (1.039) 0.0341 (1.107)

sim (center) 0.0829 0.0853 (1.029) 0.0922 (1.112)

sim (flat) 0.0814 0.0829 (1.018) 0.0884 (1.086)

* () means ration to cut radius = 0.5.

sim : C.S. is flat sim : C.S. is pT ~ 0real data

Differences of ratio will be systematic error : 2%

32

前方中性子測定

33

ISR 実験

xFNucl. Phys. B109 (1976) 347-356

○ 重心系エネルギー 30~63 GeV○STAC 検出器 40*40*80 cm3 (iron) 5.5I

res. 26%@20GeV 19%@30GeV○They were placed in 0o,20o,66o,119o,(±1)mrad 　 (56m for 0o)

xF スケーリング前方ピーク

34

One pion exchange modelEur.Phys.J.A7:109-119,2000

xF

メソンクラウドモデル

レッジェポール

B. Kopeliovich et al. Z.Phys.C73:125-131,1996

35

ZEUS 測定

Leading neutron spectrometer

Beam pipe

Neutron exit window

106m

Lead plates sampling cal.(7I) EE /65.0/

3 枚の charge veto counter(70*50*2 cm3)

GeV300~s

30 GeV 電子　 920 GeV 陽子

36

ZEUS results

Forward neutronC.S. of neutron-tagged D*±

ep での重クォーク生成は g 融合

もし交換されたパイオンと仮想光子との反応ならパイオン内部のグルーオン構造関数が見えるはず

S. Chekanov et al. physics Letters B 590 (2004) 143–160

S. Chekanov et al. nuclear Physics B 637 (2002) 3–56

Scaling is unreliable

37

• 生成断面積に関して– 今のところパイオン交

換モデルでよく記述できている。

• 横偏極非対称度– パイオンはスピンフリ

ップなので、記述できるだろう。

　 PHENIX での新しいデータが重要だ。

Eur.Phys.J.A7:109-119,2000

xF