first results on two-particle correlations determined by phenix

Stephen C. Johnson forthe PHENIX Collaboration

First Results on Two-Particle Correlations Determined by PHENIX

Stephen C. Johnson

Lawrence Livermore National Lab

for the PHENIX Collaboration

Stephen C. Johnson forthe PHENIX Collaboration

Why study particle correlations?(Preaching to the choir)

• Measures the relative spatial-temporal separation at the point of last interaction – Sensitive to relative

separation in phase space of pairs at freeze-out

)(r

)(~1)(

)(2 q

qB

qAC

Assuming a chaotic source with no dynamical correlations and no final state interactions the resulting correlation function is rather straight-forward:

C2

1/R

q

p1

p2

p2

p1

q

2

Stephen C. Johnson forthe PHENIX Collaboration

But Beware!!

• “C2 depends on more variables than you can think of.”– Anonymous HBT expert

• In particular– Dynamical correlations

– Lorentz kinematics

**This first figure unabashedly stolen

from M. Lisa

Be careful with assumptions of symmetric and static sources!!

flow

RO

RS

**

RS2+22=RO

2

e.g.

Stephen C. Johnson forthe PHENIX Collaboration

Agenda

• First PHENIX HBT results from Run 1.

• Proof of principle – capability of PHENIX to perform these measurements– PID capabilities and analysis– 1D correlations– Bertsch-Pratt 3D fit to ROSL

– Confirm dependencies of R on kt

• Comparisons and Conclusions

• A glance at future capabilities.

Stephen C. Johnson forthe PHENIX Collaboration

Particle identification

Tracking with Drift Chamber/Pad Chambers

Time of flight with TOF & Beam-Beam Counter

“Particle identification capability of the PHENIX experiment”

H. Hamagaki

• For this analysis: = • 2 within peak

• 3 away from K peak

• 180<p< 2000 MeV/c

Stephen C. Johnson forthe PHENIX Collaboration

Pair Acceptance• Acceptance about time-of-

flight wall restricts correlation measurement to ypair<.3

• PHENIX excellent PID allows pair reconstruction to very high kt– => larger statistics of year-2

will allow high kt 3D measurements

All pairsLow q pairs

Raw

Cou

nts

(A.U

.)

MeV

ppppk yyxxt

350

)()(2

1 221

221

Stephen C. Johnson forthe PHENIX Collaboration

q Acceptance

~qTside

~qout qlong

• Finite acceptance in y and => finite qTside and qlong reach.

• Excellent particle identification to large p => large qTout bite.

kt

Stephen C. Johnson forthe PHENIX Collaboration

MeVq

MeVq

Tout

Tside

100

100

MeVq

MeVq

Tside

long

100

100

MeVq

MeVq

Tout

long

100

100

q acceptance

Stephen C. Johnson forthe PHENIX Collaboration

Centrality• All data presented today

are for ‘minimum bias collisions’ (i.e. only with trigger on zvtx)

• However: finite detector + minimum bias sample + pairs requirement– => bias towards central

collisions in min bias sample

• Truly a central event sample

%17

%15

events

pairs

centrality

centrality

PHENIX Preliminary

Stephen C. Johnson forthe PHENIX Collaboration

Qinv distributions

Qinv (MeV/c)

C2

• Full coulomb correction…– Analytic coulomb

correction assumes R(Qinv) source

– Resulting radii independent of RQinv input within ~10%

• Qinv => measurement of relative separation of the pair in the pair rest frame– Acceptance dependent!!

PHENIX Preliminary

Stephen C. Johnson forthe PHENIX Collaboration

Rl = 4.0 1.2RTside = 7.9 1.1RTout = 6.23 .54 = .493 .050

Rl = 6.7 .9RTside = 5.8 1.5RTout = 5.49 .54 = .493 .061

Bertsch-Pratt Fit Results

Stephen C. Johnson forthe PHENIX Collaboration

1D correlations

– Full coulomb correction

– q =>Measure of relative separation of pair in the rest frame of the collision.

q (MeV/c)

C2

R=6.5 .9=.38 .08

R=4.5 .6=.28 .05

R=4.5 .5=.34 .06

R=6.0 .8=.38 .08

R=4.8 .5=.48 .08

R=3.7 .6=.37 .09

PHENIX Preliminary

MeVkt 195 MeVkt 524MeVkt 307

Stephen C. Johnson forthe PHENIX Collaboration

As a function of kt

• kt dependence is well behaved

• Measured relative separation of pions from source expected to decrease versus kt possibly due to collective behavior of source or resonance contributions.

PHENIX Preliminary

Stephen C. Johnson forthe PHENIX Collaboration

Comparisons• No indication of large

volumes

• Comparable results to SPS and preliminary STAR with similar <kt>/centrality

– Not exactly the same, so be careful…

• Perhaps indicative of same effective freeze-out due to convolution of– larger source and

– stronger dynamical correlations / high flow velocities / resonances(?)

Star Preliminary (M. Lisa)NA44NA49E866E895PHENIX Preliminaryroot-s

Stephen C. Johnson forthe PHENIX Collaboration

Conclusions*• First measured correlation function from PHENIX at

RHIC including most Run 1 dataset:

– 1D correlation looks well behaved => used to determine coulomb correction.

– 3D correlation comparable to SPS published and STAR preliminary results when comparing similar kt bins

– Dependencies of Rl and Rs on beam energy are gradual and well behaved.

– No indication of long lived source in correlation function• But be careful what you conclude from that…

• First proof-of-principle methodology -- looking forward to increased statistics and systematic studies…

* Don’t leave yet, I still have 3 more slides…

Stephen C. Johnson forthe PHENIX Collaboration

For the future...• This year

– Complete systematic studies of correlations

– h-h- -p, -K (1D)

– with EMCAL:• , KK, pp (1D)

• Next year– 100x statistics =>

• vs centrality and kt

• pp, KK

– with EMCAL• anti-neutron, (?)

• In general– Are there other methods we can use to extract info from

correlations? ==>

Coherent interpretation of source size

Stephen C. Johnson forthe PHENIX Collaboration

Other methods

– Explicit inversion of correlation function to deduce relative separation of pairs at freeze-out.

– How sensitive is this method to the source?

• Can we learn more than just the RMS?

• Extract multi-D source parameters?

• Systematic studies underway.

Brown and Danielewicz nucl-th/0010108

Figure and studies by Mike Heffner

See poster by D. Brown

Stephen C. Johnson forthe PHENIX Collaboration

Special Thanks

R. SoltzM. Heffner