Particle correlations at STAR

1
Particle correlations at STAR
Jan Pluta Heavy Ion Reactions Group
(HIRG), Faculty of Physics, Warsaw University
of Technology
Some results from the STAR HBT group, presented
recently by Z.Chajecki, A.Kisiel, M.Lisa,
M.Lopez-Noriega, S.Panitkin, F.Retiere,
P.Szarwas.
3-rd Budapest Winter School on Heavy Ion Physics,
10 XII 2003
2

• Outline
• The STAR experiment
• RHIC HBT Puzzle
• General analysis
• asHBT
• Two K0-short correlations
• p-Au, d-Au data
• Nonidentical particles - emission asymmetry
• Plans for future

3
Relativistic Heavy Ion Collider (RHIC)
?3.8 km1740 superconducting magnets

• Beam energy up to 100 GeV/A (250 GeV for p)
• Two independent rings (asymmetric beam collisions
  are possible)
• Beam species from p to Au
• Six interaction points
• Four experiments STAR, PHENIX, PHOBOS and BRAHMS

4
Solenoidal Tracker At RHIC
5
STAR Detector side view
STAR Detector side view
and STAR Collaboration face view
6
STAR Collaboration

• 500 Collaborators
• including
• 65 graduate students
• 60 postdocs
• 12 countries
• 49 institutions
• Spokesperson
• John Harris 1991 - 2002
• Tim Hallman 2002 - now

USA, Brazil, China, Croatia Czech Republich,
England, France, Germany, India, Netherlands,
Poland, Rusia
7
Particle correlations
The idea
Quantum statistics and Final-State Interaction
Space-time sizes and dynamics
Correlation function
Momenta and momentum difference
8
(No Transcript)
9
STAR Event 2
Central Event AuAu 200GeV/A
4m
Real-time track reconstruction Pictures from
Level 3 Trigger, online display.
Typically 1000 to 2000 tracks per event into the
TPC
10
Event and Particle Selection
AuAu Collisions at Sqrt(SNN)200GeV

• Centrality selection based on number of charged
  hadrons.
• three different centralities
• Midrapidity
• -0.5 lt y lt 0.5

• Particle identification via specific ionization
  (dE/dx)
• electron band removed by cuts
• Optimum performance for p HBT
• 0.150 lt pT (GeV/c) lt 0.550
• for K0s
• 0.100 lt pT (GeV/c) lt 3.500

STAR PRELIMINARY
11
Some base definitions - to be used for results
presentation
Two-particle kinematics
Main (approximative) relations Qout lt--gt
?Pt Qside lt--gt ?? Qlong lt--gt ?? KT lt--gt Pt
LCMS (P1P2)z0
12
HBT Excitation Function
Comparison with lower energies for 10 most
central events at midrapidity kT 0.17 GeV/c No
significant increase in radii with energy RO/RS
1 Gap in energy that needs to be closed
13
RHIC HBT Puzzle
Most reasonable models still do not reproduce
RHIC sqrt(SNN) 130GeV HBT radii
Hydro RQMD
PHENIX PRL 88 192302 (2002)
vSNN 130GeV

• Blast wave parameterization (Sollfrank model)
  can approximately describe data
• but emission duration must be small
• ???? 0.6 (radial flow)
• T 110 MeV
• R 13.5 ? 1fm (hard-sphere)
• ?emission 1.5 ? 1 fm/c (Gaussian)

from spectra, v2
14
3Dimensional Pion HBT
Sqrt(SNN) 200 GeV
p p
Pratt-Bertsch Parameterization LCMS frame
(p1p2)z0 Central Events pT 0.15-0.25
GeV/c Coulomb correction ? spherical Gaussian
source of 5fm momentum resolution corrected(1
effect at 200GeV, due to higher B-field)
STAR PRELIMINARY
Statistical errors only!!
15
Centrality and mT dependence at 200 GeV
STAR PRELIMINARY
RL varies similar to RO, RS with centrality HBT
radii decrease with mT (flow) Roughly parallel mT
dependence for different centralities RO/RS 1
(short emission time)
16
Comparison 200 to 130 GeV. Longitudinal radius
(fit to STAR Y2 data only)
STAR PRELIMINARY
Longitudinal radius at 200GeV identical to 130
GeV
17
Evolution timescale from RL
Simple Mahklin/Sinyukov fit (assuming
boost-invariant longitudinal flow)
Makhlin and Sinyukov, Z. Phys. C 39 (1988) 69
STAR PRELIMINARY
Assuming TK110 MeV(from spectra at 130 GeV)
(fit to STAR Y2 data only)
18
Comparison 200 to 130 GeV. Transverse radii
STAR PRELIMINARY

• Higher B-field ? higher pT
• Transverse radii
• similar but not identical
• low-pT RO, RS larger at 200 GeV
• steeper falloff in mT
• (PHENIX 130GeV)
• Ro falls steeper with mT

19
Azimuthally sensitive HBT (asHBT)

• sensitive to interplay b/t anisotropic geometry
  dynamics/evolution
• broken symmetry for b?0 gt more detailed,
  important physics information
• another handle on dynamical timescales likely
  impt in HBT puzzle

20
HBT respect Reaction Plane
1D projections, f45
vSNN 130GeV
21
HBT(f) Results 130 GeV
T100 MeV ?r0?0.6 ?ra?0.037, R11.7 fm, t2.2
fm/c
Minbias events _at_130GeV Bolstered statistics by
summing results of p- and p analyses Blast-wave
calculation (lines) indicates out-of-plane
extended source
Star preliminary
22
A model of the freezeout - BlastWave

• BW hydro-inspired parameterization
• of freezeout
• longitudinal direction
• infinite extent geometrically
• boost-invariant longitudinal flow
• Momentum space
• temperature T
• transverse rapidity boost r

• coordinate space
• transverse extents RX, RY

• freezeout in proper time ?
• evolution duration ?0
• emission duration ??

23
A model of the freezeout- BlastWave

• BW hydro-inspired parameterization
• of freezeout
• Longitudinal direction
• infinite extent geometrically
• boost-invariant longitudinal flow
• Momentum space
• temperature T
• transverse rapidity boost r

• Coordinate space
• transverse extents RX, RY

• freezeout in proper time ?
• evolution duration ?0
• emission duration ??

7 parameters describing freezeout
24
BlastWave fits to published RHIC data
central
midcentral
peripheral

• reasonable centrality evolution
• OOP extended source in non-central collisions

25
Estimate of initial vs F.O. source shape

• estimate ?INIT from Glauber
• from asHBT

• ?FO lt ?INIT ? dynamic expansion
• ?FO gt 1 ? source always OOP-extended
• constraint on evolution time

26
asHBT at 200 GeV in STAR R(?) vs centrality

• 12 (!) ?-bins b/t 0-180? (kT-integrated)
• 72 independent CFs
• clear oscillations observed in transverse radii
  of symmetry-allowed type
• Ro2, Rs2, Rl2 cos(2?)
• Ros2 sin(2?)
• centrality dependence reasonable
• oscillation amps higher than 2nd-order 0?

() Heinz, Hummel, MAL, Wiedemann, Phys. Rev. C66
044903 (2002)
27
Pion Correlations d-Au and p-Au

• Pion correlation in d Au data selection
• p-Au selection
• 1D correlation function
• 3D correlation function
• d-Au vs p-Au
• KT dependence
• Centrality dependence

28
p-Au selection
FTPC E -Au
Au
d
All trigger events
Using information from ZDC-d STAR can separate
events with neutron spectator from deuteron
29
1D Correlation Function
Gaussian fit
STAR preliminary

• CF is very wide (rel Au-Au)
• Coulomb/merging less important
• CF looks reasonable
• 1D Gaussian fit is not good
• needed more deeply study of fit method

only statistical error included !
30
3D Correlation Function
3D Gaussian fit
Gaussian parametrization is not perfect but HBT
radii characterize the width of CF
STAR preliminary
Fit results
Rout, Rside sensitive to the number of
participants
cut on the others Q's components lt 30 MeV/c
GeV/c
31
KT dependence
STAR preliminary

• clear KT dependence
• Rout and Rside - sensitive to the number of
  participants
• Rlong the same KT dependence for dAu and pAu

p Au d Au
32
KT dependence d-Au Au-Au divided by p-p
STAR preliminary

• the same trend of KT dependence for d-Au
• and Au-Au as for p-p
• HBT radii are scaled by constant factors

33
MT dependence of Rlong
STAR preliminary
Sinyukov fit
Rlong const (mT)-a
mT ? kT2 massp2
a for different collisions
STAR preliminary
p-p d-Au Au-Au Au-Au

peripheral midcentral
34
Centrality definition in d-Au
FTPC-Au charged primary particle multiplicity in
-3.8lt?lt-2.8
most peripheral
most central
3
2
1
FTPC E -Au
Au
d
35
centrality
Centrality dependence

• clear centrality dependence
• similar to AuAu
• connection to geometry

STAR preliminary
p Au d Au
minbias
- Glauber calculations (Mike Miller)
36
K0sK0s Correlations
37
mt scaling violation?
inv
Next RHIC HBT puzzle ?
38
The asymmetry analysis
Non-identical particle correlations

• Catching up
• Interaction time larger
• Stronger correlation

• Moving away
• Interaction time smaller
• Weaker correlation

• Double ratio
• Sensitive to the space-time asymmetry in the
  emission

Kinematics selection on any variable e.g. kOut,
kSide, cos(v,k)
R.Lednicky, V. L.Lyuboshitz, B.Erazmus,
D.Nouais, Phys.Lett. B373 (1996) 30.
39
Double ratio definitions
2k p1 p2
P p1 p2
simulation
kside lt 0
kside sign selection arbitrary
Correlation functions
p1
kout gt 0
p2
kout sign selection determined by the
direction of the pair momentum P
Double ratios
kside gt 0
kout gt 0
p2
2k GeV/c
klong is the z component of the momentum of first
particle in LCMS
p1
40
What to expect from double ratios

• We are directly sensitive to time shift, the
  space shift arises from radial flow possibility
  of a new radial flow measurement

• Initial separation in Pair Rest Frame (measured)
  can come from time shift and/or space shift in
  Source Frame (what we want to obtain)

41
What do we probe?
Separation between source 1 and 2 in pair rest
frame
Source of particle 1
Source of particle 2
sr
Boost to pair rest frame
Dr
Dt
ltDrgt
Dr (fm)
r (fm)
Dr gpair (Dr bpair Dt)
Separation due to space and/or time shift

• Mean shift (ltDrgt) ? seen in double ratio
• Sigma (sr) ? seen in height of CF

42
Correlation functions and ratios
Good agreement for like- sign and unlike-sign
pairs points to similar emission process for K
and K-
CF
Out
Clear sign of emission asymmetry
Side
Two other ratios done as a double check
expected to be flat
Long
Preliminary
43
Results for Pion-Proton 130 AGeV
? peaks

• Similar preliminary analysis done for pion-proton
• We observe Lambda peaks at kminv of ?
• Good agreement for identical and non-identical
  charge combinations

STAR preliminary
44
Preliminary results for Kaon-Proton

• Using data from Year2 (200 AGeV) sufficient
  statistics
• No corrections for momentum resolution done
• No error estimation yet fit indicates
  theoretical expectations

45
Modeling the emission asymmetry

• Need models producing strong transverse radial
  flow
• Blast-wave as a baseline
• RQMD
• UrQMD
• T. Humanic's rescattering model

• What do we measure and how to compare it to the
  models?
• Is our fitting method working? And if yes, what
  does it tell us?
• Need to disentangle flow and time shift

46
Understanding modelsBlast wave Flow baseline

• Blast wave
• Parameterizes source size (source radius) radial
  flow (average flow rapidity) and momentum
  distribution (temperature)
• No time shift
• Only spatial shift due to flow
• Infinitely long cyllinder (neglects long
  contribution)

Parameterization of the final state
47
Blast wave how does the flow work
Average emission points
Pion pt 0.15 GeV/c ?t 0.73
Kaon pt 0.5 GeV/c ?t 0.71
Proton pt 1. GeV/c ?t 0.73
Particle momentum
Spatial shifts (Dr)
48
Fitting and quantitative comparisons

• Fits assume gaussian source in PRF
• rout distributions have non-gaussian tails
• Use the same fitting procedure for models and
  data - correlation functions constructed with
  Lednicky's weights

Example of rout distribution from RQMD
49
Comparing models to data

• Rescattering models and blast-wave are consistent
  with data
• Blast wave parameters constrained by STAR
  measurements
• In models flow is required to reproduce the data

• More points in ßt needed to map and discriminate
  the flow profile needs STAR upgrades in PID
  capability (TOF barrel)

50
STAR HBT Matrix (circa 2003)
Analysis in progress
published
submitted
Not shown
3p Correlations (accepted PRL) asHBT Phase space
density Correlations with Cascades dAu,
pp Cascades
51
What have we learned so far?

• RHIC HBT puzzle
• Break down of theoretical description of
  correlations at RHIC
• Indication of short source lifetime and
  freeze-out duration at RHIC
• Short lived hadronic phase?
• Out of plane extended pion source in non-central
  collisions
• Also points to short emission times
• Weak energy dependence of the HBT radii
• Where is the phase transition?
• Large pion phase space densities (non-universal)
• Small entropy per pion?
• Chaotic pion source from 3p correlations
• No multiparticle effects above Pt200 MeV/c
• Source asymmetries from non-identical
  correlations
• Consistent with collective flow and short time
  scales
• Only systematics measurements may provide
  answers!

52
What will affect STAR HBT analysis?

• RHIC upgrades progress
• STAR upgrades
• Various other measurements (e.g. spectra, high
  Pt,
• strangeness, flow, etc)
• New theoretical ideas

53
Consequences for STAR HBT

• Large statistics AuAu datasets
• Plans for 2004 14 weeks of AuAu physics
  running
• 30M central, 50M peripheral events
• What can be done? Many analysis which were
  statistics limited!
• Rare particle correlations W, X,L, etc
  (identical, non-identical)
• Early freeze-out, sequence of emission, flow,
  FSI, etc
• Correlations relative to reaction plane
• Kaons
• Non-identical
• Baryon correlations ppbar, LLbar, pL, etc
• Coalescence, light nuclei and anti-nuclei
• Large statistics pp (100M events) datasets _at_200,
  500 GeV
• STAR HBT matrix (e.g. non identical correlations)
• HBT in Jets?
• spin dependent HBT? (with polarized beams)
• Different energies
• Different beams

Add dependencies on centrality, Kt, reaction
plane Event by Event HBT New analyses ideas
(S.Pratt, imaging, etc)
54
Consequences for STAR HBT

• Better particle identification
• Extension of HBT systematics to higher Kt 1-3
  GeV/c
• Region of transition from Hydro to pQCD
• What is space-time picture in this region?
• Correlations of identical particles
• Scan in Pt for Non-identical correlations
• Sensitivity to flow profile, model details
• asHBT
• Higher efficiency of hyperons reconstruction
• x10 for W compare to TPC alone
• High statistics correlations with hyperons

55
Summary

• RHIC and STAR future seems to be certain for next
  5-10 years
• Upgrade path is visible
• The number of available datasets and possible
  analysis topics will be rapidly increasing
• Data volumes will be unprecedented (at least for
  us)
• Can we do analysis in a reasonable time?
• Analyses will be moving to rare particles
• Shall we continue with systematic approach?
• Probably yes
• If new results or theoretical predictions will
  suggest promising measurement - we will
  concentrate on it