GLUON POLARISATION

introduction

ΔG from inclusive data analysis

direct measurements of ΔG/G

1) method and previous measurements

2) recent results from COMPASS

3) measurements at RHIC

outlook

Fulvio Tessarotto

INFN – Trieste

EINN 2005

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

In inclusive polarised DIS one can measure an asymmetry in the differential cross-section:

and extract the struture function g1.

In the QPM g1 has a simple interpretation:

In QCD at NLO g1 may depend explicitely

on G:

measuring Q2 variations of g1 we can try to

get G

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

The nucleon spin

½= ½ΔΣ + ΔG + Lq + Lg

quarks gluons orb. mom.

Ellis-Jaffe: ΔΣ = a0 ≈ 0.6

experimentally: a0 ≈ 0.25

ΔΣ = a0 + 3(s/2)

ΔG

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

0.43 0.80 ΔG

0.03 0.09 - Δs

0.04 0.43- Δd

0.03 0.84 Δu

0.06 0.32 ΔΣ

0.01 0.1 1

0.0

0.1

0.2

0.3

0.4

PD(g1

NLO/F1

NLO)

x

xG(x)

0.01 0.1 1

-0.4

-0.3

-0.2

-0.1

0.0

x

-x(d+d)

0.01 0.1 1

-0.2

-0.1

0.0

x

-x(s+s)

0.01 0.1 10.0

0.1

0.2

0.3

0.4

x

PD (g1

NLO + HT)

-x(u+u)

SLAC (E142, E143, E154, E155)CERN (EMC, SMC)DESY (HERMES)

185 exp. points

E. Leader, A. Sidorov, D. Stamenov Phys. Rev. D67 (2003) 074017

G not well constrained not extracted

dΔ ,uΔ

Spin-dependent PDFs

dp,1

dp,1

dn,p,1 /Fg ,A

1/2 = ½ + + Lz = 0.96 + Lz

Lz is negative

Two years ago …

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

g and q were largely unconstrained

Direct Direct g g measurementmeasurement

Is this still true?

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

final g1 data

- Very good statistics- Smearing (resolution and radiative corr.) →

correlation between x bins

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

g1n(x) at high x from JLAB

• pQCD + no Lz →

u/u= d/d=1 at high x

• Very accurate A1n at high x

A1n > 0 at x > 0.5

• + world A1p → d/d < 0

so Lz not negligible ?

u/u

d/d

PRL 92, 012004 (2004)

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

g1d(x) at low x

• new data : =0.202 (+0.042 -0.077) → 0.237 (+0.024 -0.029)

• COMPASS systematically > SMC at low x

PLB 612 (2005) 154

g1/F1

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

NLO QCD fits with new data + new unpolarised PDFs

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

to know G we still need to measure it

possible trend toward smaller G but the results are scattered:

• Idea proposed by R.D.Carlitz, J.C.Collins and A.H.Mueller, Phys.Lett.B 214, 229 (1988)……..and many more

•revisited by A.Bravar,D.von Harrach and A.Kotzinian,Phys.Lett.B 421, 349 (1998)

• Method used in HERMES for photoproductionHERMES, A.Airapetian et al., Phys.Rev.Lett.84, 2584 (2000)

• Applied for DIS region by SMC (data with Q2 >1GeV2) SMC, B.Adeva et al.., Phys. Rev. D 70, 012002 (2004).

History:

Direct Measurements:

polarised p p collision: RHIC

semi-inclusive deep inelastic scattering: DESY, CERN (Photon Gluon Fusion)

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

• Strategies to suppress background: 1) Open charm production: q = c unique, no background charm fragmentation, 1.2 D0 per event

D0 -> K- π+ (BR 4%), D*+(~20%) -> D0 π+

2) High-PT hadron pair production: q=u,d,s kT(intr) = 450 MeV/c

pT(frag) = 350 MeV/c

• pQCD scale set by: 1) ŝ > 4mc

2 or 2) ŝ > (pt1+pt2)2

ΔG/G from Photon-Gluon Fusion

Photon-Gluon Fusion

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

HERMES & DESY

• 27.5 GeV e • Pure atomic H or D targets• Spectrometer with complete hadron identification

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Statistics decreases with pT

HERMES ALL from high pT hadron pairs

ehhXeNTeLL fAPP

NN

NNA

Proton target

Promising channel, but …

theoretical uncertainties

may be large

<xg> = 0.17

<Q2> = 0.06 (GeV/c)2

photoproduction

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ΔG/G from SMC

µ+ 190 GeV Pµ= -0.78±0.03

DIS eventsQ2 > 1 GeV2, 0.4<y<0.9

xF > 0.1, pT2 > 2.5

(GeV/c)2

1% of inclusive sample

∆G/G = -0.200.290.11

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

COMPASS SPS

LHC

COMPASS EXPERIMENT

Beam: 2 . 108 µ+/ spill (4.8s / 16.2s)Luminosity: ~ 4 . 1032 cm-2 s-1 Beam polarisation: 76%Beam momentum: 160 GeV/c

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

SM1 dipoleSM1 dipole

SM2 dipole

Polarised Target

HCAL1

Muon-filter1,MW1SPS 160 GeV 2.108/spill

Micromegas,DC,SciFi

Gems,SciFi,DCs,straws

MWPC Gems Scifi

trigger-hodoscopes

Silicon

Spectrometer 2002 -> 2004

RICH_1 Gem_11

ECAL2,HCAL2

straws,MWPC,Gems,SciFi

straws Muon-

filter2,MW2

DW45

SciFi

Veto

New detector technologies

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ΔG/G from open charm

G

GaA LL

)s,x(G )s( sd

)s,x(G )s( sdccN

GPGF

GPGF

PGF at NLO:Bojak, Stratmann NPB 540 (1999) 345; A.P. Contogouris and G. Grispos, Phys. Rev. D 62 (2000). ŝ = m2

cc

D0 K- π+ (BR 4%)

D*+ D0 π+ + c.c.

Special trigger: quasi realphotoproduction

c

c

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

D* tagging: D* D0 π (Kπ)π

KD0MeV/c2)

sK

M

eV/c

2 )

Cuts:

zD > 0.2

|cos *| < 0.85 (Background)

10 < pK < 35 GeV/c(RICH PID)

D0 K π

D*

D

0 π

no vertex detector(solid target!)

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ΔG/G from open charm

MC: Aroma + PolDis

ΔG/G = -1.08 ± 0.73

D0 (untagged, 2003) 4200 eventi

D*+→D0π+ (2003) 1100 eventi

<xg> ~0.15

0.73 0.43 with 2004 data

D0 → K π π0

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Pairs of hadrons with high pt and Q2 > 1 GeV/c

Photon Gluon Fusion

h1

h2

• current fragmentation:

• event cuts:

xF > 0.1z > 0.1

pt > 0.7 GeV/cpt1

2+pt22 > 2.5 (GeV/c)2

m(h1h2) > 1.5 GeV/c2

• 2 high pt hadrons:

N

Q2 > 1 (GeV/c)2

xBj < 0.050.1 < y < 0.9

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Result for high pt asymmetry

Aγ*d = -0.015 ± 0.080 (stat.) ± 0.013 (syst.)2002/03 data

Asymmetry for high-pt hadron pairs at Q2 > 1 (GeV/c)2:

Systematic uncertainty contains contributions from:• target and beam polarisation measurement (5% resp. 3%)• upper limit on false asymmetries (0.013)• determination of D and f (5% each)

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

determination of ΔG/G

Photon Gluon Fusion QCD-Compton Leading Order

DISLOq

q

D

a

G

G

D

a

D

AA

tot

ComComLL

tot

PGFPGFLL

hhNLLd

ˆˆ

Relative Fractions determined by Monte Carlo (LEPTO)

05.074.0DaPGFLL

0qq

DaComLL

cut on xBj < 0.05:

0Ad1

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Data Monte Carlo comparision• Use LEPTO Monte Carlo for Q2>1 (GeV/c)2 including RADGEN

σPGF/σtot = 0.34 ± 0.07 (syst.)

Determination of thefraction of PGF eventsin LEPTO:

• Variation of the cut on pt2

• Standard and modified* set of fragmentation parameters

* SMC, B. Adeva et al., Phys. Rev. D70:012002, 2004

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

10% ofstatistics

COMPASS result for ΔG/G

ΔG/G = 0.06 ± 0.31stat. ± 0.06syst.2002/03 dataQ2 > 1 (GeV/c)2

at <xg> = 0.13 ± 0.08

GSA,B,C: T. Gehrmann and W. Sterling, Z. Phys. C65 (1994) 461.

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ALL/D = Rpgf G/G aLLpgf

+ Rqcdc q/q aLLqcdc

all Q²

Pairs of hadrons with high pt and Q2 < 1 GeV/c

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

+ Rqq q/q aLLqq q/q

Physical background

ALL/D = Rpgf G/G aLLpgf

+ Rqcdc q/q aLLqcdc

Q² < 1

all Q²

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

+ Rqg G/G aLLqg q/q

Physical background

ALL/D = Rpgf G/G aLLpgf

+ Rqcdc q/q aLLqcdc

+ Rqq q/q aLLqq q/q

all Q²

Q² < 1

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

+ Rgq q/q aLLgq (G/G)

Physical background

ALL/D = Rpgf G/G aLLpgf

+ Rqcdc q/q aLLqcdc

+ Rqq q/q aLLqq q/q

+ Rqg G/G aLLqg q/q

all Q²

Q² < 1

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

+ Rgg G/G aLLgg (G/G)

Physical background

ALL/D = Rpgf G/G aLLpgf

+ Rqcdc q/q aLLqcdc

+ Rqq q/q aLLqq q/q

+ Rqg G/G aLLqg q/q

+ Rgq q/q aLLgq (G/G)

all Q²

Q² < 1

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Inputs:

-generation of Monte Carlo evts-parametrization (world data)-pQCD calculations

Physical background

ALL/D = Rpgf G/G aLLpgf

+ Rqcdc q/q aLLqcdc

+ Rqq q/q aLLqq q/q

+ Rqg G/G aLLqg q/q

+ Rgq q/q aLLgq (G/G)

+ Rgg G/G aLLgg (G/G)

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Resolved photon processes

polarised parton distribution functions of (virtual…) photon areunknown!

BUT:

-q(x,²) < q(x,²) < q(x,²)

leads to 2 extreme (min&max) scenarios

additional (theoretical) uncertainty

measured

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

MC/data comparison

Q²

PYTHIA generator is used for Q²<1GeV²GEANT is used for spectrometer simulation

.data

- MC

data/MC

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

MC/data comparison

pT

.data

- MC

.data

- MC

data/MC data/MC

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Results from MC

The Monte Carlo describes our data quite well!

Rpgf ≈ 31%

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Q2>1GeV2 :

Q2<1GeV2:

Gluon polarisation, high pT pairs 2002-2003

data

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

G = 2.48

G = 0.62

G = 0.16

Gluon polarisation results:

ΔG/G (xg ≈ 0.1) is small G>1 disfavoured

GRSV2000NLO fits to g1

2 = 3 Gev2

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ΔG/G from two jet events

LLAˆ

Hard Scattering Process

2P2 2x P

j 2f x

i 1f x1P

1 1x P

zhqDs

1ps

2ps

gggg gqgq qqqq

qg+gq

qq

gg

0

Fraction

's produced

Pt [GeV/c]

Leading hadron used as jet tag

G

G

G

G

G

G

q

q

q

q

q

q

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Prompt photon production asymmetry• Gluon Compton Dominates

• At LOLO no fragmentation function• Small contamination from annihilation

•

2i i 2

i u,d,s

2i i 2

i u,d,s

LL1

1LL

e f (xg(

)

e f (A

x )a (gq

x )

g(x )q )

A1

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

NSRL

2 superconducting rings3.8 km circumference

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

RHIC – first polarized hadron collider

BRAHMS

STAR

PHENIX

AGS

LINACBOOSTER

Pol. H- Source

Spin Rotators(longitudinal polarization)

Solenoid Partial Siberian Snake

Siberian Snakes

200 MeV Polarimeter Internal Polarimeter

Rf Dipole

pC PolarimetersAbsolute Polarimeter (H jet)

pC Polarimeter

25% Helical Partial Siberian Snake

5% Helical Partial Siberian Snake

PHOBOS

Spin Rotators(longitudinal polarization)

Spin flipper

Siberian Snakes

Achieved ~ 50% beam polarization at 100 GeVAchieved ~ 30% beam polarization at 205 GeV on first try!

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

New AGS helical snakes

2.6 m

2.6 m

25% s.c. helical snake build at BNL

Warm snake avoids polarization mismatch at AGS injection and extraction.

Cold strong snake eliminates all depolarizing resonances in AGS.

6 % helical snake build at Tokana Industries funded by RIKEN.

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

STAR spin physics program: ∆G, ∆q/∆q, δq

Large acceptance TPC and EMC -1<η<2

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ΔG – Star: ALL for jets (2005)

STARSTAR

0 5 10 15 20 25 30

0.2

0.4

0.6

0.8

1.0

0.0

gg qq

qg

Inclusive Jets :LOW. Vogelsang

pT (GeV)

2005

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

PHENIX Detector

• Central Arms: ||<0.35, =2900

Charged particle ID and tracking; photon ID

• Muon Arm:1.2<||<2.4Muon ID and tracking

• Global Detectors Collision triggerCollision vertex characterizationRelative luminosityLocal Polarimetry

Philosophy: High rate capability & granularity Good mass resolution and particle ID Sacrifice acceptance

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ppπ0/prompt + X

prompt γπ0

Excellent agreement with NLO & NLL

calculation with collider data !

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Outlook 2005/7

2003/4

PHENIX ALL for Inclusive 0 productionat √s=200 GeV, 2003 + 2004 data with about ∫Ldt~0.3pb-1 and <Pb>~0.3

Run 5 (April –June 2005):

absolute polarimeter: ∆Pb/Pb~7%

∫Ldt ~ 3pb-1, <Pb>~0.45

reduces statistical error by factor 7

pp -> 0 + X

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

B. Jaeger et al. Phys. Rev. D67, 054005 (2003)

Y. Fukao, spin2004

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

ΔG at RHIC – summary & outlook

• Various channels, large coverage in xg

- Jet/leading hadron production : First results of ALL pp 0 +X from PHENIX Very good perspective for +- at PHENIX and jets at STAR - Direct photon production: hard scale 100 GeV 2, xg reconstructed in STAR

- Heavy flavours, ...

• RHIC Spin program on track, major effort to reach performance

Run s1/2(GeV) Pbeam L(pb-1)

3 200 0.27 0.35

4 + 5 200 >0.40 10

>5 200 0.70 320

500 0.70 800

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Physics vs Luminosity and Polarization at RHIC

2005 2006 2007 2008 2009 …. 2012 (RHIC II)

10 pb-1 …………………………………… 275pb-1 …….. 950pb-1

√s= ……………………….. 200 GeV …………………......... 500 GeV|

P= 0.5 0.6 0.7 ……………………………………

Inclusive hadrons + Jets ~ 25% Transverse Physics Charm Physics direct photons bottom physics W-physics

ALL(hadrons, Jets) ALL(charm)

ALL(γ) AL(W)

L= 1x1031cm-2s-1 6x1031cm-2s-1 1.6x1032cm-2s-1

see Spin report to DOE http://spin.riken.bnl.gov/rsc/

@ 200GeV @ 500GeV

eRHIC

ELIC

Outlook

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

PHENIX Forward Upgrade: W-physics and jet+photon

Muon Trigger Upgrade

RPC + FPGA based

trigger logic

Nose cone calorimeter

W-Si calorimeter

Trigger RPC Locations

RPC2

June 2005: $1.99M NSFMRI grant approved + $280k institutional support

UC Riverside, Abilene Christian, Iowa State, U. Illinois, U. Colorado, Columbia, Georgia Tech., RBRC, Kyoto, Peking U.,

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Final results on ∆G will come from combined NLO analysis of RHIC and DIS

RHIC measurements will span broad range in x with good precision. Multiple channels with independent theo. and exp. uncertainties.

Uncertainty through extrapolation to small x

s=200 GeV incl. 0 prod’n s=500 GeV incl. jet prod’n

∆G Measurements by 2012 see Spin report to DOE http://spin.riken.bnl.gov/rsc/

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

COMPASS

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

COMPASS upgrade in 20065

m5

m

6 m6 m3 m3 m

Photon Photon Detectors:Detectors:CsI MWPCCsI MWPC

mirrormirrorwallwall

vesselvessel

radiatorradiator::

CC44FF1010

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

• 2002 – 2004 -> 2006• • FoM(G/G)

G/G)=1/FoM1/2

hard work to increase FoM

0

2

4

6

8

10

12

14

16

2002 2003 2004 2006

per year accumulated FoM

Figure of Merit for open charm

2006 will be a very important year

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

μ’

p’μ

DVCS μp μ’p’

Additional equipment to the COMPASS setup

At these energies (for μ, μ’, and ) the missing mass technique is not adapted

M2required = (mp+mπ)2-mp

2 = 0.25 GeV2

M2observed > 1 GeV2

ECal 1 or 2

12°

Nμ=2.108/SPS cycle(duration 5.2s, each 16.8s)

2.5m liquid H2 targetto be designed and built

L = 1.3 1032 cm-2 s-1

Recoil detector to insure exclusivityto be designed and built

+ additional calorimeter at larger angle

Fulvio Tessarotto EINN 2005

Milos, 23 09 2005

COMPASS far future:

GPD’s, transversity, G/G and hadron programwill benefit from a possible upgrade of the SPS:

a Superconducting SPS (1 TeV) is better suited to LHC needs;

A bright future for G both at short and long term