Upload
gianna
View
32
Download
1
Embed Size (px)
DESCRIPTION
High p T g -h Correlations Using the PHENIX Detector. Matthew Nguyen, Stony Brook University. Why Direct Photon Correlations?. “Leading Order” Picture Exact Momentum Balance w/ Away-Side Jet Compton Dominance p+p : Measure Gluon Distribution Function A+A: - PowerPoint PPT Presentation
Citation preview
High pT -h Correlations Using the PHENIX Detector
Matthew Nguyen, Stony Brook University
IAA
2/8/2008Quark Matter 20082
Why Direct Photon Correlations? “Leading Order” Picture
Exact Momentum Balance w/ Away-Side Jet
Compton Dominance p+p: Measure Gluon Distribution
Function A+A:
Calibrated Probe of Energy Loss More Sensitive Probe than single particle
spectra, Di-Hadron Correlations The Golden Channel for Jet Tomography?
the fine print Fragmentation photons Initial state effects (Shadowing , kT) Still sensitive to geometry / space-time
evolution Quark vs. Gluon Energy Loss
q
qg
γ q
q g
γ
Compton AnnihilationDirect Photon Processes at LO
X.N. Wang, Z. Huang PRC55 (1997) 3047-3061
3
-h Correlations in PHENIX
Direct photon cross section in p+p
PbSc
ElectromagneticCalorimeters
PbGl
2/8/2008Quark Matter 2008
Drift Chamber
h-h Correlations in Au+Au
2/8/2008Quark Matter 20084
PYTHIA 6.1
Statistical Subtraction Method Large background from meson decay Direct S/B increased by jet
suppression in central Au+Au collisions
0-h decayed via MC to get decay -h -h measured in p+p R = N(inclusive) / N(decay) Can perform a statistical
subtraction of per-trigger yields:
Method well-suited for high multiplicity environment (Au+Au)
11 1
direct inclusive decayRY Y Y
R R
Where Y = # hadrons per photon
pT [GeV]
Photon Sources in PYTHIA
R
R measurement in Au+Au 200 GeV
QM2006 Direct Photon-Hadron Results
2/8/2008Quark Matter 20085
At QM2006 PHENIX showed “proof of principle” level -h results Run 5 p+p @ 200 GeV
Subtraction method is robustNear-side disappearance consistent with +jet expectationsAway-side comparable to hadron triggered case
Away-side yields in Au+Au are suppressed compared to p+p-h measurement is statistics limited
Run 4/5 p+p/Au+Au @ 200 GeV
New Results Increased Precision
2/8/2008Quark Matter 20086
*Note: Run 6, 7 Absolute Efficiencies not finalized – In Arbitrary Units
1/N
trig
dN
/d
(A.
U.)
Systematic error dominated by uncertainty in Rand -h contribution
Subtraction technique works over wide range of pT!
Direct Photon-Hadron Correlations in p+p
2/8/2008Quark Matter 20087
[rad]
1/N
trig
dN
/d
(A.
U.)
p T, h
adro
n
pT, photon
Photon Recoil Jets
2/8/2008Quark Matter 20088
1st approximation -- recoil jets similar to hadron triggered jets in p+p <Q2> of hard scattering should be lower due to z bias in hadron-triggered correlations kT effect need not be the same – needs to be measured triggered yields sensitive to the FF
-h measurement
h-h measurementpT, photon
Per-
Trig
ger
Yiel
d (A
.U.) Run 6 p+p @ 200 GeV
– DELPHI, Eur. Phys. J. C13,543, (1996)-- OPAL Z.Phys. C 69, 543 (1996)
Run 3 -h
– Quark FF--Gluon FF
PRD 74 (2006) 072002
Vacuum Fragmentation Function
Run 5 Direct -h
Preliminary
1/Nt
rig
dN/d
x E( A
.U.)
Baseline for modified FF in Au+Au
0-h xE distributions were shown to poorly discriminate between quark (– ) and gluon
(--) fragmentation due to 0 trigger bias (z<1)
2/8/2008Quark Matter 20089
p+p @ 200 GeV
2triggerT
triggerT
partnerT
Ep
ppx
)(zDdxdNppE
jetT
triggerT
In “leading order” picture:
Direct photon xE distributions should be sensitive to underlying(quark) fragmentation functionUse Run 6 data to reduce errors
Au+Au Run 7 Direct Photon-Hadron Correlations (7 < pT, < 9)
2/8/200810 Quark Matter 2008
Little or no near-side production associated with direct photon triggers Away-side yields indicate large jet suppression in +jet channelWorking to reduce systematics and more data to analyze (~2X)
0
Au+Au analysis is challenging: Additional sources of uncertainty from underlying event subtraction (ZYAM), elliptic flow subtraction and 0 combinatoric background
The pT Landscape
2/8/2008Quark Matter 200811
Direct Photon S/B (R)0 Combinatoric S/BpQCD Accuracy
pT, hadron
pT, photon
Run7 Au+Au 200 GeV
Jet S/B
Rate
Near/Away-Side > 1
Run7 Au+Au 200 GeV
Near/Away-Side ≈ 1
12-15 X 3-5 GeV7-9 X 3-5 GeV
Near-Side Fragmentation
2/8/2008Quark Matter 200812
Near-Side yield small but non-zero for z 0Expect near-side contribution from photon fragmentationSystematic errors for -h correlations large on the near-side
-h (h-)measurement
h-h measurement Per-
Trig
ger
Yiel
d (A
.U.)
z0.10.3 0.25 0.2
Run 6 p+p @ 200 GeV
pT, photon
Fragmentation Photons
2/8/2008Quark Matter 200813
q
qg
γ
Bremsstrahlung diagrams
q
qgγ
gFragmentation diagrams must be included at NLO to cancel divergences |z| < 1 for fragmentation photons, but how much less?Dparton/photon(z) not well constrainedMay oppose either quark or gluon jetsImportant for LHCSuppressed by E-loss
I. Vitev, 2007 RHIC-AGS Meeting
Rate of fragmentation photons may be increased by presence of medium partons
InducedBremstrahlung
--RHIC– LHC
h- Correlations
2/8/2008Quark Matter 200814
10-
6
X10-6
X10-3
X10-3
h-inclusive
h-decay
h-direct
Idea: By triggering on a hadron and looking for near-side direct photon partners one can measure the fragmentation photon yield directly
Measure hadron - inclusive and hadron - decay correlations
Decay corr’s are made by tagging0 and by invariant mass
Must know tagging efficiency and false tagging rate precisely dominant source of systematic uncertainty
2.5 < pT, < 3.5
h- Correlations II
2/8/2008Quark Matter 200815
Outlook
2/8/2008Quark Matter 200816
Direct photon-hadron correlations are moving from the proof of principle stage towards the precision measurement stage Finalizing Run 6 p+p for increased precision and broader
kinematic range compared to previous results Analyzing Run 7 Au+Au data set and working to reduce
systematic uncertainties Baseline measurements of -h and h- correlations are
being made to ensure that theory is well under control Fragmentation component measured – Nfrag ≈ 0.1 Ninc Away-side jet suppression in the +jet channel appear
strong PHENIX is currently working on Run 6 and Run 7 p+p and
Au+Au -h correlations at high pT
Backup Slides
2/8/200817 Quark Matter 2008
2/8/2008Quark Matter 200818
Constructing the Decay Background To start: Construct (0) - h
from 0-h Problem is that tagging the
0’s introduces a bias <pT> tagged = <pT>
true Developed a pair-by-pair
correction procedure
Per-Trigger Yield from 0):
07
5
( , )T T TP p p dp
Correction Procedure:
Decay kinematics reproduced by Monte Carlo include energy resolution effects and acceptance (P)This is just for 0’s, What about contributions from other mesons? This approach implicitly assumes they have the same associated yield
00 0
00
0
( )d, , ,T
hhTT T TT TT p YPY pp p p ppp
000( ) (( )TrueTagged )
TT T
d dNdNdp
Nd dpp
– Analytic Form – w/ Detector Response
Sample MC weight function:
2/8/2008Quark Matter 200819
-h CorrelationsRun 5 p+p @ 200 GeV
Parent-h
PRC 75, 024909 (2007)
First measurement of -h correlations at RHIC!
()-h Yields generated by same procedure as (0)-h
Combinatorial background under h peak subtracted using a sideband procedure
•-h•(h)-h
Zharkarov –JETP Lett. 80 (2004)
Turbide et al., Phys. Rev. C72 (2005)
2/8/2008Quark Matter 200820
QGP Induced Photon Production
InducedBremstrahlung
Jet Conversion– Hard partons re-scatter off medium partons via Compton process
Induced Bremsstrahlung – Rate of fragmentation photons increased by presence of medium partons
Both effects still speculative Are there experimental signatures? pT, v2 dependence (see other talks in session) h- correlations
Jet Conversion
Prompt Photon Spectrum
2/8/2008Quark Matter 200821
p+p Integrated Yields
2/8/2008Quark Matter 200822
kT Effect
2/8/2008Quark Matter 200823
kT measurement for 0-h
• kT = 2.5 GeV• kT = 0
kT Effect: Net momentum of di-jets(and -jets) due to initial state radiation
xE slopes for -h from PYTHIA
kT measurements the di-jet channelshow a large effect -- 2.5 – 3.0 GeV
PYTHIA shows that kT changes the xE distribution significantlykT measurement in -h is needed
PRD 74 (2006) 072002
Gluon Fragmentation Functions
2/8/200824 Quark Matter 2008
Compilation by Hirai et al.Phys Rev D75 (2007) 94009
Gluons contribute significantly to single inclusive measurements at RHIC However different parameterizations give divergent estimations of gluon contribution e.g., AKK FF’s indicate much higher gluon contribution than previous fits
Inclusive production from AKK FF
Rg from Run 4 Au+Au 200 GeV
2/8/2008Quark Matter 200825
2/8/2008Quark Matter 200826
Using p+p data in FF‘sSTAR, Phys Lett B, 637 (2006) 161
200 GeVs
STAR identified hadron spectra constrain the gluon FF at large zUse twoparticle correlations to constrain g p at all z?
New set of fragmentation functions include p+p (and DIS) data and obtain very different parameterizations of the gluon FF’s
Phys. Rev. D75, 114010 (2007).
De Florian, Sassot, Stratmannhep-ph:0707.1506
recoil jets more likely to be a quark jet than 0 recoil jets Quark and gluon jets may have different particle content p/ on the away-side of vs. 0 recoil sensitive the FF’s
0
0
1/
/ ( recoil)/ ( re
2 + 1/ 2
coi
/ ( recoil)/ ( recoi
l)
l)
2 1
p pq q g g
pq
pq q
pg
q
g
D D D Dpp D D
D D DD pp
2/8/2008 Quark Matter 200827
Identified Hadron Correlations
For definiteness assume:
Quar
k/Gl
uon
Cont
ribut
ion
q
qg
γ
Double ratio gives partial cancellation of systematics due to particle identificationand direct photon subtraction
Kretzer Acta Phys. Polon. B36, 179 (2005)
qDD )recoil(
gq DDD 2/12/1)recoil ( 0 &
In this simplified picture particle ratios can be written in terms of FF’s:
pT()
Well constrained from e+e-Quark
Gluon
2/8/2008Quark Matter 200828
-Identified h Results (Under Construction)
PHENIX5 < pT
< 7 X 1 < pTh < 2
Hadron efficiencies not finalized Systematic errors are large but will most cancel in ratio So far run 5 only Run 5 + 6 should enable a significant measurement (integrated over away-side)
Arbi
trar
y N
orm
aliza
tion
- identified hadron correlations can be performed in the range 5 < pT
< 12 X 1 < pTh <2
This corresponds to about 0.1 < z < 0.2
2/8/2008Quark Matter 200829
R from p+p CollisionsIn p+p there is no published Rbut deriveable from 0 and spectraFor pure power law: N(0)/N() = (n-1)/2; n = 0 slope parameter
0
0total decay( ) /
1( /
1)2
T
T
dN dpRdN dp
n
Quoted from x-sec analysis
2/8/2008Quark Matter 200830
kT effect affects:–away-side jet width–Inclusive high pT x-sections–Recoil spectra (dN/dxE)
Trigger Biases and the kT effect
kT Effect: Acoplanarity of di-jets(and -jets) due to initial state radiation
Insensitivity of dN/dxE to true FF is the interplay of two trigger biases:
Trigger Bias #1: Trigger requirement
biases the near side FF to high z Does not affect away-side jet Absent in g-h
kT measurement for di-jets
Trigger Bias #2: Longitudinalcomponent of kT biases jet CM indirection of the trigger for pT,t >
pT,a Does affect the away-side jet Present in -h
kT described by NLO, Why do we need to measure it?
2/8/2008Quark Matter 200831
Gluon Fragmentation
The gluon fragmentation function has large uncertainties! Is there an independent test of gp?
2/8/2008Quark Matter 200832
pQCD Limitations
at 200 GeVp p s+ =
Data Fit
Auranche 2006
Agreement w/ data is remarkable!Better than jet cross-sections
Closer: Agreement only 20% Not good enough to constrain kT
2/8/2008Quark Matter 200833
kT in the -jet channel
M. Begel, Nucl. Phys. B 79, 244 1999;Ph.d thesis (1998)
Tk s Measurement for -0 at lower sIndependent of process
2/8/2008Quark Matter 200834
2/8/2008Quark Matter 200835