Gluon Polarization Measurements at STAR

Transcription

1 Gluon Polarization Measurements at Matthew Walker for the Collaboration

2 Outline! Brief theoretical motivation! Inclusive measurements: Jets and pions! Correlation measurements: Di-Jets! Status and Prospects

3 Theoretical Motivation! Polarized DIS tells us that the spin contribution.4 from quark spin is only ~3%.. Without RHIC data 1 = 1 Σ + L q + G + L g With RHIC data x!g KRE (NLO) KKP (NLO) unpolarized KRE " min +1 KRE " min +% 1 - x x!g D. de Florian et al., Phys. Rev. D71, 9418 (5). D. de Florian et al., Phys. Rev. Lett. 11 (8) GRSV maxg GRSV ming x Substantial improvement for.5 < x <., but large uncertainties at low x 3

4 Theoretical Motivation! Extracting gluon polarization A LL = d σ dσ = f 1 f σ h a LL D h f f 1 f σ h D h f 1 = 1 Σ + L q + G + L g all 1.75 qg qg gg gg f 1 σ h.5 qq qq q q q q.5 f gg q q long-range short-range long-range cos!* Extract!g(x,Q ) using a global fit G(Q ) = 1 g(x, Q )dx 4

5 Detector! = -1! =! = 1 Magnet.5T! = Time Projection Chamber (TPC) Endcap Electro Magnetic Calorimeter (EEMC) West East Barrel Electro Magnetic Calorimeter (BEMC) Not shown: Trigger detectors or polarimeters 5

6 Inclusive Measurements! Inclusive measurements have:! High statistics! Simple triggers A LL = 1 (N ++ + N ) R(N + + N + ) P B P Y (N ++ + N )+R(N + + N + )! Simple reconstruction! Multiple subprocesses contribute! Wide range of x gluon in each reconstructed bin frac G Q =1GeV /c a) p = 5.6 GeV/c T p =8 GeV/c T x dn / d(log x) X gluon

7 Inclusive Jets! is well suited for Jet measurements with large acceptance (" in azimuth)! TPC provides charged tracking ( # < 1.3)! B/EEMC provide electromagnetic energy reconstruction (-1 < # < )! Jets reconstructed using a midpoint cone algorithm Jet " + " g q parton particle detector 7

8 Inclusive Jets! Shape comparison between Run 6 Data and simulation shows good agreement! Motivates use of correction based on PYTHIA MC < pt < GeV < pt < GeV Run-6 Data MC Preliminary Run p T [GeV] Jet Profile!(!r) MC: Pythia Geant < pt < GeV !r/r from Jet Axis Preliminary Run < pt < 6.19 GeV Data MC (Data - MC) / MC 1-1 Preliminary Run pt [GeV] 8

9 Inclusive Jets d! /!dp T d! [pb/gev] d!!dp T d! Inclusive Jet Cross Section GeV Cone Radius = <! <.8 Preliminary Run 6!! Ldt = 5.39 pb!1 (Data - Theory) / Theory Run-6 Data-theory Comparison of Inclusive Jet Cross Section GeV Cone Radius = <! <.8 Preliminary Run 6 Systematic Uncertainty Theoretical Uncertainty!! Ldt = 5.39 pb! ! Data agrees well with NLO pqcd calculation after hadronization and underlying event correction is applied 1 Run-6 p T [GeV].1 Systematic Uncertainty Theory NLO pqcd + CTEQ6M Pileup Preliminary Run6 Timebin Luminosity JES Had. and UE. Corrections p T [GeV] p T [GeV] 9

10 Inclusive Jets D. de Florian et al. PRL 11 (8) 71. A LL systematics (x 1-3 )! Run 6 results: GRSV-MAX/ GRSV-MIN ruled out, a gluon polarization between GRSV-std and GRSV-zero favored Reconstruction + Trigger Bias Non-longitudinal Polarization Relative Luminosity Backgrounds [-1,+3] (p T dep) ~.3 (p T dep).94 1 st bin ~.5 else ~.1 p T systematic ± 6.7% 1

11 Run 9 Projected Precision! Projected statistical precision of.1 in several pt bins! Several systematic studies underway 11

12 Neutral Pions! is able to measure neutral pions over a wide pseudorapidity range using its electromagnetic calorimeters! Forward rapidity collisions dominated by qg collisions with a low x gluon! GRSV-Max ruled out by Run 6 result 1

13 Correlation Measurements! Reconstructing multiple physics objects (di-jets, photon/jet) provides information about initial parton kinematics! Adds information about the shape of!g(x,q )! well suited for correlation measurements with its large acceptance x 1 = 1 s (p T 3 e η 3 + p T 4 e η 4 ) x = 1 s (p T 3 e η 3 + p T 4 e η 4 ) M = x 1 x s η 3 + η 4 = ln x 1 x 13

14 Charged Pions D. deflorian, Phys. Rev. D 79 (9) ! Comparison of ALL(" + ) to ALL(" - ) can give the sign of!g(x,q )! Calculating ALL as a function of z alleviates problems of trigger bias 14

15 Di-Jets Run 6! Run 6 data and simulation agreement is good! Run 6 cross section and asymmetry analyses are progressing 15

16 6 Cross Section d 3!/dMd! 3 d! 4 [pb/gev] d 3! dmd! 3 d! 4 Dijet Cross Section GeV Cone Radius =.7 max(p T ) > 1 GeV, min(p T ) > 7 GeV -.8 <! <.8,!! < 1.!! >. Preliminary!! Ldt = 5.39pb!1! Unpolarized differential cross section between 4 and 1 (GeV/c )! NLO theory predictions using CTEQ6M provided by de Florian with and without corrections for hadronization and underlying event from PYTHIA 1 1 Run-6 Systematic Uncertainty Theory NLO pqcd + CTEQ6M Had. and UE. Corrections! Statistical Uncertainties as lines, systematics as rectangles M jj [GeV] 16

17 6 Cross Section (Data - Theory) / Theory Systematic Uncertainty Theoretical Uncertainty!! Ldt = 5.39 pb!1 Preliminary Data-theory Comparison of Dijet Cross Section GeV Cone Radius =.7 max(p T ) > 1 GeV, min(p T ) > 7 GeV -.8 <! <.8,!! < 1.,!! > M jj [GeV] Run-6 Theory: CTEQ6M NLO pqcd Had. UE. Corrections! Comparison to theory (including hadronization and underlying event correction) shows good agreement within systematic uncertainties 17

18 6 Asymmetry A LL = 1 N ++ RN + P B P Y N ++ + RN +! Run 6 Longitudinal double helicity asymmetry A LL Dijet A LL GeV Cone Radius =.7 max(p T ) > 1 GeV min(p T ) > 7 GeV -.8 <! <.8,!! < 1.!! >. Data Run-6 Sys. Uncertainty Preliminary! Systematic uncertainties show effects on trigger efficiency from different theory scenarios! Scale uncertainty (8.3%) from polarization uncertainty not shown. -. GRSV STD DSSV GRSV!g = GRSV!g =! g M jj [GeV]!! Ldt = 5.39pb!1 18

19 9 Simulation! Different detector, different trigger, updated geometry! 9 MC productions with partonic pt > GeV! PYTHIA 6.4.3, propt (PYTUNE 39)! Virtual Machine prepared with software stack and deployed to over 1 machines! Run using cloud computing resources at Clemson University in South Carolina (Ranked #85 best supercomputer)! Over 1 billion events generated by PYTHIA, filtered to allow only 36 million to undergo detector simulation (GEANT3), and 1 million through full reconstruction! Took over 4, CPU hours and generated 7 TB of files transferred to BNL N Machines Available Machines Working Machines Idle Machines! Largest physics simulation on cloud, largest simulation in CPU hours Jul17 Jul4 Jul31 Date 19

20 Data/Simulation Run 9! Run 9 data simulation agreement is good Normalized Yields (Data-Simu)/Simulation Data Simulation Run 9 Data Preliminary R cone = <! <.8 $! < 1. $% >. p+p # Jet + Jet + X s = GeV Preliminary Invariant Mass (GeV/c ) ! 34 M = x 1 x s η 34 = 1 lnx 1 x cos("*) cos θ

21 9 Projections! 8 pb -1 processed so far A LL east barrel - east barrel and west barrel - west barrel MC GRSV std GRSV m3 GRSV zero GS-C(pdf set NLO) 9 Data A LL east barrel - west barrel Scale uncertainty GRSV std DSSV! Average polarization: 59%..1..1! Figure-of-merit:.96 pb Wed Sep 15:18:55 1 M [GeV/c ] East West East West Projected Precision M [GeV/c ] October 1, 1 SPIN 1 1

22 Prospects-Inclusive Jets! First look at 5 GeV data in Run 9! Future 5 GeV runs will significantly surpass statistical precision of current constraints, including the upcoming Run 11

23 Prospects: Di-Jets 5 GeV 5 GeV Projections for 39 pb -1 at 5% polarization A LL : east barrel - endcap pb (P = 5%) NLO GRSV std DSSV A LL : west barrel - endcap ! Dijets at 5 GeV can access the gluon polarization at lower x! Expectations are for smaller asymmetries! Larger luminosities should improve statistical uncertainties A LL <! <., -1. <! 3 4 < M [GeV/c ] : east barrel - east barrel and west barrel - west barrel A LL <! <.,. <! 3 4 < M [GeV/c ] : east barrel - west barrel Scale uncertainty GRSV std DSSV <! <., -1. <! < <! < 1.,. <! < <! < 1., -1. <! 3 4 < M [GeV/c ] M [GeV/c ] 3

24 Prospects-Prompt Photons! Material removed before Run 9 significantly reduces conversion backgrounds! Forward photons measure lowest x! Correlation with mid-rapidity jet provides cleanest identification of initial parton kinematics. EEMC Photon + BEMC Jet -1 " Ldt=5 pb, Pol=.6, s= GeV.15-1 " Ldt=3 pb, Pol=.5, s=5 GeV! G/G.6-1 # Ldt=5 pb, Pol=.6, s= GeV.1 Photon + Jet, Full EEMC+BEMC.4-1 # Ldt=3 pb, Pol=.5, s=5 GeV.5 Norm. from Pythia v8 with trigger simulation. Norm. from Pythia v8 with trigger simulation Background A subtraction assumes DSSV LL LDA Efficiency " 7% and purity " 5% A LL -.5 COMPASS Background A LL subtraction assumes DSSV Analysis efficiency!.74 and purity!.4 DSSV GRSV STD s= GeV s= GeV DSSV GRSV STD s=5 GeV s=5 GeV HERMES SMC DSSV Q = 1 GeV GRSV STD Q = 1 GeV Photon x T x Gluon 4

25 Summary! 6 results improved precision at mid-rapidity and new techniques used to limit systematics! First global analysis including RHIC Spin data suggest small gluon polarization (.5 < x <.)! Correlations measurements provide constraints on parton kinematics! Run 9 provides the largest GeV data sample to date and first look at lower x with 5 GeV data! Expanded 5 GeV analyses will be possible with Run 11 data 5