Transcript スライド 1
Longitudinal density correlations
in Au+Au collisions at √sNN = 200 GeV
Tomoaki Nakamura
(KEK - High Energy Accelerator Research Organization)
for the PHENIX collaboration
6/25/2007
Tomoaki Nakamura - KEK
1
Pressure [Atm]
Phase diagram of He4
Tricritical point
Normal
liquid phase
Solid phase
Super fluidity
phase
Temperature [K]
[J. H. Vignos and H. A. Fairbank, Phys. Rev. Lett. 6, 265 (1961)]
6/25/2007
Tomoaki Nakamura - KEK
2
Phase transition to the Super fluidity
Phase diagram He4
Cs [J/gK]
Pressure [Atm]
critical phase
boundary
Specific heat Cs
[K]
Temperature [K]
This behavior around Tc
can be a robust signature
to indicate the phase
boundary.
6/25/2007
[mK]
|T-Tc|
[μK]
[W. M. Fairbank and M. J. Buckingam,
Int. Conf. on Low Temp. Phys. (1957)]
Tomoaki Nakamura - KEK
3
Second order phase transition
• The second order phase transition have a
close relationship on the global symmetry of
a system, rely on the universality hypothesis.
Imaginary QCD phase diagram
• If QGP phase transition corresponds to the
chiral phase transition, it would be the
second order phase transition.
• Surveying the second order derivative of
free energy i.e. Susceptibility,
2G
2 ,
h T
h T
is a standard tactic to identify the second
order phase boundary, like the specific heat
in He4,
2G
Ch T 2 .
T h
6/25/2007
K. Rajagopal,
Acta. Phys. Polon. B, 3021 (2000)
However, there is no reliable
and quantitative indication on
phase boundary in both
theoretically and experimentally.
Tomoaki Nakamura - KEK
4
Ginzburg and Landau theory
Concerning Ginzburg-Landau theory, free energy density f can be
expressed as this expansion,
1
1
1 4
2
2
f (T , , h) f 0 (T ) A(T )( ) a(T ) b h
2
2
4
spatial fluctuation
φ: scalar order parameter, h: external field
f-f0
a>0
a=0
a<0
φ
One dimensional density fluctuation from the mean density is introduced
as an order parameter.
( y) ( y)
a(T ) a0 (T Tc )
6/25/2007
Tomoaki Nakamura - KEK
5
Density correlation among
hydrodynamical sub elements
• Density fluctuation can be
measured in the final state
particle density by
introducing proper time
frame.
time
hadrons
(1)
hadron phase
(2) (3)
τc
• Differential length dz
among hydrodynamical sub
elements (1), (2), (3)…, at a
common proper time τ can
be expressed as
QGP
phase
τf
dz cosh( y )dy.
• In the case of limiting the
space region of interest only for
the mid-rapidity, cosh(y) ~
1, then,
nucleus
nucleus
dz ~ dy.
6/25/2007
Tomoaki Nakamura - KEK
6
The case in heavy-ion collision
χ, ξ
The susceptibility can be derived by
introducing correlation length ξ as
1
h
k
1
a0 | T Tc | (1 k 2 (T ) 2 )
2 f
k 2
h
hadron phase
In the long wave length limit, k = 0,
1
k 0
(T )G2 (0).
a0 | T Tc |
QGP phase
no phase transition
critical phase boundary
6/25/2007
Tomoaki Nakamura - KEK
T, ε
7
Correlation length
The correlation length is useful not only for searching the phase
transition but also for the visualization of phase order.
[H. Nishimura, 2D Ising model (2005)]
a) T < Tc
Ordered phase
→ indicate long
correlation length
6/25/2007
b) T = Tc
c) T > Tc
At critical temperature
Disordered phase
→ coexistence of various
→ indicate short
correlation length
correlation length
→ diverged measured value
Tomoaki Nakamura - KEK
8
Charged track reconstruction in PHENIX
• We use straight line tracks measured at
no magnetic field condition to optimize
low momentum charged particles.
• Minimum pT threshold.
π: 0.1 GeV/c
K : 0.25 GeV/c
p : 0.35 GeV/c
• Particle composition.
π : K : p = 94 : 4 : 2
• Acceptance: Δη < 0.7, Δφ < π/2
• Track identification: Using vector of
charged track (DC) associated with beam
vertex (BBC) and two hit points in wire
chamber (PC1, PC3). Clusters in EMC are
used for reference of association.
6/25/2007
• Mean pT for π = 0.57 GeV/c.
• For inclusive charged particle, maximum
3 % difference at η = 0.35 for the
conversion of rapidity to pseudo rapidity.
Tomoaki Nakamura - KEK
9
Spectator neutrons are counted
as the energy sun in ZDC.
Centrality and Npart and εBj
• Centrality class is determined
by forward detectors.
[PHENIX, arXiv:0704.2894 (nucl-ex)]
• Transverse total energy is measured
by central arm EM calorimeter.
[PHENIX, Phys. Rev. C71, 34908 (2005)]
Relative number of fragment particles
from participant nucleons are measured by
charge sum in BBC.
Number of participants
• Number of participant nucleons • Bjorken energy density is
calculated as
are obtained by the Glauber
1 dET
Bj
.
model calculation for each
cAT dy
centrality class.
6/25/2007
Tomoaki Nakamura - KEK
10
Multiplicity fluctuation and correlation
[PHENIX, arXiv:0704.2894 (nucl-ex)]
Negative Binomial distribution.
n
Pn( k )
(n k ) / k
1
(n 1)(k ) 1 / k 1 / k k
δη=0.09
2 1 1
2
2
,
n
n
2 k
Second order factorial moment as a
representation of two particle correlation.
n(n 1) n 2 n
F2
2
n
n 2
2 n 2 n
2 1
1 2
2
n
1
1
k
6/25/2007
δη=0.7
• Uncorrected charged particle
multiplicity distribution in various
pseudo rapidity gap and NBD fits
for most central (10%) events in
Au+Au collisions at √sNN=200GeV.
• Accuracy of fits : 80% C.L.
Tomoaki Nakamura - KEK
11
Extraction of the correlation
Using Ornstein-Zernike formula, 1D two
particle correlation function,
[PHENIX, arXiv:0704.2894 (nucl-ex)]
C2 (1 , 2 ) 2 (1 , 2 ) 1 (1 ) 1 ( 2 ),
can be parameterized as
C2 (1 , 2 )
/
e
.
2
1
α: correlation strength
ξ: correlation length
β: constant.
Using relation with NBD k:
k 1 ( ) F2 1
0
0
C2 (1 , 2 )d1d 2
2 12
2 ( / 1 e
6/25/2007
2
2
/
)
k ( )
Tomoaki Nakamura - KEK
1
2 /
( )
99% C. L.
12
αξ, β vs. Npart
[PHENIX, arXiv:0704.2894 (nucl-ex)]
• β absorb effects on the finite
resolution of centrality binning i.e.
the fluctuation of Npart.
β
●5%
○10%
• αξ product, which is
monotonically related with χk=0
indicates the non-monotonic
behavior around Npart ~ 90.
1
2
1
| 1 Tc / T |
αξ
1
k 0
| T Tc |
●5%
○10%
• The difference of data points
between 5% and 10% binning can
be understood the smearing
5% binning
effects around the peak.
10% binning
6/25/2007
Tomoaki Nakamura - KEK
13
Other correlation sources
Pseudorapidity independent correlations are all absorbed by the
constant term β. e.g. elliptic flow etc.
Npart fluctuations (residual effect) are also absorbed owing to the β.
Trivial particle correlations originating from charged track
reconstructions in tracking detectors have been suppressed a priory.
Effects from weak decay particles (Λ, Ks) were estimated for the
NBD k by the MC calculation. It is less than 1% for each.
Effects from photon conversion electrons is about 10-3%, which was
obtained by GEANT MC simulation.
Effects from knock on electron in detector material is about 10-5%.
Above contribution is negligible as compared to total error on k.
6/25/2007
Tomoaki Nakamura - KEK
14
Evaluation of the non-monotonicity
χ2/NDF = 2.76 : 0.60
[PHENIX, arXiv:0704.2894 (nucl-ex)]
5%
χ2/NDF = 2.10 : 1.23
5%
Power law
Linear
Power law
+ Gaussian
Linear
+ Gaussian
10%
10%
χ2/NDF = 1.23 : 0.79
χ2/NDF = 1.23 : 0.79
Power law + Gaussian:
Linear + Gaussian:
6/25/2007
3.98 σ (5%), 3.21 σ (10%)
1.24 σ (5%), 1.69 σ (10%)
Tomoaki Nakamura - KEK
15
On the relation with HBT effect
• If all correlations are originated
in HBT effect,
α corresponds to the
chaoticity parameter λ
ξ corresponds to the radius
parameter R
used in HBT analysis.
Au+Au √sNN=200GeV
• However, λ is constant as a
function of Npart, and R
monotonically increases with
increasing Npart.
• Therefore, known HBT effects
cannot explain the non-monotonic
behavior of αξ.
6/25/2007
One dimensional radius parameters.
[A. Enokizono, Ph. D. thesis,
Hiroshima Univ.]
Tomoaki Nakamura - KEK
16
Conclusion I
• The charged particle multiplicity distributions for the
various pseudorapidity gap, δη < 0.35, in Au+Au
collisions at √sNN = 200 GeV are found to be well
described by NBD as well as the other collision system.
• We found the constant β parameter is necessary to
avoid the residual effects in the measurement for the
extraction of correlations from the integrated
correlation function.
• Upper limit of correlation length over all centrality
bins is less than 0.035, which is obtained by the free
parameter fits.
6/25/2007
Tomoaki Nakamura - KEK
17
Conclusion II
• αξ product, which monotonically related to
susceptibility in the long wavelength limit, χk=0, show a
non-monotonic behavior as a function of the number of
participant nucleons, Npart.
• A possible indication of a local maximum or critical
behavior is seen at Npart ~ 90 and the corresponding
energy density is εBjτ ~ 2.4GeV/(fm2c).
6/25/2007
Tomoaki Nakamura - KEK
18
Backup
6/25/2007
Tomoaki Nakamura - KEK
19
Systematic check on αξ
Fit range:
0.066 < δη < 0.306
6/25/2007
Tomoaki Nakamura - KEK
20
Correlation length vs. temperature
Cube of the equilibrium
correlation length (thin
lines) and nonequilibrium correlation
length (thick lines) as a
function of temperature.
[C. Nonaka and M. Asakawa, Phys. Rev. C71, 44904 (2005)]
6/25/2007
Tomoaki Nakamura - KEK
21
Two point correlation function
G2 ( y1 , y2 ) ( y1 ) ( y2 )
dy dy
G ( y , y )e
( y ) ( y ) e
ik ( y 2 y1 )
2
1
2
1
1
2
ik ( y 2 y1 )
2
dy1dy2
y y2 y1
Y G2 ( y )e
k
6/25/2007
iky
dy
( y)
e
iky
2
dy
2
Tomoaki Nakamura - KEK
22
Charged particle multiplicity distributions
and negative binomial distribution (NBD)
DELPHI: Z0 hadronic Decay at LEP
2,3,4-jets events
E802: 16O+Cu 16.4AGeV/c at AGS
most central events
[DELPHI collaboration] Z. Phys. C56 (1992) 63
[E802 collaboration] Phys. Rev. C52 (1995) 2663
Universally, hadron multiplicity
distributions agree with NBD in
high energy collisions.
6/25/2007
Tomoaki Nakamura - KEK
23
NBD and NFM
Bose-Einstein distribution
Pn /(1 )
n
n 1
Negative binomial distribution
(k )
n
P
n
(n k ) / k
1
k
(n 1)(k ) 1 / k 1 / k
2 1 1
2
2
,
n
n
2
k
σ: standard deviation
NBD (k→∞) = Poisson distribution
NBD (k<0) = Binomial distribution
μ: average multiplicity
n(n 1) n 2 n
F2
2
2
n
n
2 n 2 n
2 1 1
1 2 1
2
n
k
6/25/2007
Tomoaki Nakamura - KEK
24
Integration of correlation function
R2 ( y1 , y2 ) ae| y1 y2 |/ b
F2
0
0
1
(ae| y1 y2 |/ b)dy1dy2
2
0 y1
0 y2
6/25/2007
Tomoaki Nakamura - KEK
25
Centrality determination
•
•
•
Event characterization in terms of impact
participants
go into BBC
parameter (b) in Au+Au collisions.
– Large : peripheral collision
– Small : central collision
Coincidence between BBC and ZDC.
go into ZDC
spectator
– Determine collision centrality.
– 93 % of inelastic cross section can be seen.
Extract variables using Glauber Model
– Number of participants (N_part).
• Number of nucleons participate in a
collision.
• Represents centrality.
• Related with soft physics.
15-20%
– Number of binary collisions (N_binary).
1015%5-10%
• Number of Nucleon-Nucleon collisions.
• Related with hard physics.
0-5%
• Incoherent sum of N-N collisions
becomes a baseline for A-A collisions.
0-5%
6/25/2007
Tomoaki Nakamura - KEK
26
Glauber model, BBC and ZDC
Dencity profile
1
(r ) 0
rR
1 exp
a
R 6.38 fm, a 0.54 fm, NN 42 mb
Glauber R. J., Phys. Rev. 100 242 (1955); in: Lectures
in the theoretical physics, ed. W. E. Brittin, L. G.
Dunham, Interscience, N. Y., 1959, v. 1, p. 315.
BBC
6/25/2007
Tomoaki Nakamura - KEK
27