Using Higher Moments of Fluctua5ons and their Ra5os in the Search for the QCD Cri5cal Point Chris5ana Athanasiou, MIT 4 work with: Krishna Rajagopal (MIT) Misha Stephanov (University of Illinois)
Using Higher Moments of Fluctua5ons and their Ra5os in the Search for the QCD Cri5cal Point
Chris5ana Athanasiou, MIT 4
work with: Krishna Rajagopal (MIT) Misha Stephanov (University of Illinois)
Outline
• Introduc)on • Cri)cal Contribu)on to Par)cle Mul)plicity Fluctua)ons
• Ra)os of Fluctua)on Observables • Summary
QCD Phase Diagram
vacuum
Quark-Gluon Plasma
Critical point
μB / MeV
T / MeV
~ 170
~ 940
nuclear matter
0
Hadron gas Color Superconductor
crossover
Models
LaNce simula5ons
Heavy-‐Ion Collision Experiments
• Loca)ng the cri)cal point from first-‐principles – hard Heavy-‐Ion Collision Experiments
• RHIC: Au-‐Au collisions at √smax = 200 GeV
• Momentum asymmetry collec)ve flow strongly-‐coupled QGP
~ 940
vacuum
Quark-Gluon Plasma
Critical point
μB / MeV
T / MeV
~ 170
nuclear matter
0
Hadron gas Color Superconductor
crossover
QCD Phase Diagram
√s
Heavy-‐Ion Collision Experiments -‐ con5nued
• As QGP expands and cools, it follows trajectories with approx. nB/s = const.
Quark-Gluon Plasma
μB / MeV
T / MeV
~ 170
~ 940
nuclear matter
0
Color Superconductor
QCD Phase Diagram
√s
Critical point
Hadron gas
crossover
vacuum
Heavy-‐Ion Collision Experiments 2
• As QGP expands and cools, it follows trajectories with approx.
• Chemical freeze-‐out: system dilute enough that par)cle numbers freeze
nB/s = const.
• To maximize cri)cal point (CP) effects vary to get freeze-‐out point near CP
√s
Event-‐by-‐Event fluctua)ons
• Detector “sees” par)cle mul)plici)es from freeze-‐out condi)ons
• Find observables that are sensi)ve to proximity to the CP
Outline
• Introduc)on • Cri)cal Contribu)on to Par)cle Mul)plicity Fluctua)ons
• Ra)os of Fluctua)on Observables • Summary
• Cri)cal mode -‐ σ : order parameter of the chiral phase transi)on
• Correla)on length diverges at the CP ξ = m−1σ
• Develops long wavelength correla)ons at the CP
Ω(σ) =
d3x
12(∇σ)2 +
m2σ
2σ2 +
λ3
3σ3 +
λ4
4σ4 + ...
.
• Effec)ve ac)on
Cri5cal Mode Fluctua5ons Cri5cal Mode
• Near the CP: with dimensionless and known in the
Ising universality class
λ3 = λ3 T (T ξ)−3/2, λ4 = λ4 (T ξ)−1
4 λ4 200 λ3 8,
Cri5cal Mode Fluctua5ons
• at CP in the thermodynamic limit ξ →∞
• Finite system life)me compared to away from the CP (Berdnikov, Rajagopal 00)
ξmax ∼ 2 fm
∼ 0.5 fm
• Cri)cal mode fluctua)ons affect Par)cle mul)plicity fluctua)ons
Momentum distribu)ons Ra)os, etc…
of these par)cles.
• σ couples to pions and protons: Lσππ,σpp = 2 G σ π+π− + g σ p p
(t, V →∞)
3 2 1 0 1 2 3
10
20
30
40
Number
of events
Number
of protons
Measuring fluctua5ons in par5cle mul5plici5es
measure the mean, variance, skewness, etc…
• Can repeat these calcula)ons for pions, net protons, etc • Want to obtain the cri)cal contribu)on to these quan))es
• We will use cumulants, e.g.:
κ2 = N2, κ3 = N3, κ4 ≡ N4 = N4 − 3N22
mσ = ξ−1, γk =
k2 + m2/m, v2k = nk(1± nk),
gπ = G/mπ, d = 2, λ3 = λ3 T (T ξ)−3/2, λ4 = λ4 (T ξ)−1
δnk1δnk2σ = d2 1m2
σV
g2
T
v2k1
γk1
v2k2
γk2
Cri5cal contribu5on to pion/proton correlators
(Rajagopal, Shuryak, Stephanov 99, Stephanov 08)
ξ2
ξ7
ξ9/2
δnk1δnk2δnk3σ = d3 2λ3
V 2T
g
m2σ
3 v2k1
γk1
v2k2
γk2
v2k3
γk3
δnk1δnk2δnk3δnk4σ = d4 6V 3T
2
λ3
mσ
2
− λ4
×
g
m2σ
4 v2k1
γk1
v2k2
γk2
v2k3
γk3
v2k4
γk4
+ …
Net protons and mixed correlators
• Note: correlators depend on 5 parameters:
which have large uncertain)es G, g, ξ, λ3, λ4
• Net protons: Adapt previous expressions by replacing:
Np −Np
• Can also calculate mixed correlators, e.g. 2 pion – 2 proton:
δnπp1
δnπp2
δnpp3
δnpp4σ = d2
pd2π
6V 3T
2
λ3
mσ
2
− λ4
gp gπ
m4σ
2 vπ 2p1
γπp1
vπ 2p2
γπp2
vp 2p3
γpp3
vp 2p4
γpp4
vp 2k → vp 2
k − vp 2k
Calcula5ng mul5plicity cumulants
• Second cumulant – variance:
Poisson -‐ Bose-‐Einstein effects -‐ Other interac)ons -‐ Etc..
ignore
ω2p, σ =κ2p,σ
Np= d2
pg2ξ2
T
k
v2k
γk
2
knk
−1
• Normalizing:
ωipjπ =κipjπ
Ni
i+jp N
ji+jπ
• For mixed cumulants with i protons and j pions:
• Non-‐cri)cal contribu)on to ωipjπ = δi,i+j + δj,i+j + (few %)
κ2p,σ = (δNp)2σ =
k1
k2
δnk1δnk2σ ∝ V 1
Mul5plicity cumulants – cri5cal point signature
• Higher cumulants depend stronger on ξ: ω2 ∝ ξ2,
ω3 ∝ ξ9/2,
ω4 ∝ ξ7
• As we approach the CP ξ increases and then decreases as we move away from it
• CP signature: Non-‐monotonic behavior, as a func5on of collision energy, of mul5plicity cumulants
ξ(µB) =2 fm
(1 + (µB − 400)2/W 2)1/3
0.1 0.2 0.3 0.4 0.5 0.6ΜBGeV
0.5
1.0
1.5
2.0
Ξfm0.1 GeV
0.2 GeV
!
• E.g. toy example
0.0 0.1 0.2 0.3 0.4 0.5ΜBGeV
2
4
6
8
10
Ω
Ω1 p2Π
Ω3 Π
Ω2 p1Π
Mul5plicity cumulants – example plots
0.0 0.1 0.2 0.3 0.4 0.5ΜBGeV0
100
200
300
400
Ω4 p
g7, 0.2 GeVg7, 0.1 GeVg5, 0.1 GeV
0 0.1 0.2 0.30
1
2
0.0 0.1 0.2 0.3 0.4 0.5ΜBGeV
10
20
30
40
50
60
Ω
Ω4 Π
Ω2 p2Π
Ω3 p
Parametriza)on (Cleymans et al 05):
T (µB) = a− bµ2B − cµ4
B
a = 0.166 GeV, b = 0.139 GeV−1, c = 0.053 GeV−3
and using λ3 = 4, λ4 = 12, G = 300 MeV, g = 7
Data on net proton cumulants
κ σ2 ≡ κ4
κ2where
(STAR Collabora)on 2010)
Cri5cal contribu5on to proton ω4
0.0 0.1 0.2 0.3 0.4 0.5ΜBGeV0
100
200
300
400
Ω4 p
g7, 0.2 GeVg7, 0.1 GeVg5, 0.1 GeV
0 0.1 0.2 0.30
1
2
Mul5plicity cumulants – movie
Changing the cri)cal μB – the loca)on of the CP:
Outline
• Introduc)on • Cri)cal Contribu)on to Par)cle Mul)plicity Fluctua)ons
• Ra)os of Fluctua)on Observables • Summary
Uncertain5es of parameters
• Cumulants depend on 5 non-‐universal parameters: κ2π ∼ V T−1G2ξ2N2
π ,
κ3π ∼ V T−3/2G3λ3ξ9/2N3
π ,
κ4π ∼ V T−2G4(2λ23 − λ4)ξ7N4
π
• have large uncertain)es hard to predict the cri)cal contribu)on to cumulants
G, g, ξ, λ3, λ4
• By taking ra)os of cumulants can cancel some parameter dependence
minimize observable uncertain)es
Ra5os of mul5plicity cumulants
No parameter dependence
Ra)os taken aher subtrac)ng Poisson and defined λ4 ≡ 2λ2
3 − λ4
r = i + j
r1 =skewnesspskewnessπ
, r2 =kurtosispkurtosisπ
,
Parameter independent ra5os
• Parameter and energy independent ra)os:
where skewness =κ3
κ3/22
, kurtosis =κ4
κ22
• All equal to 1 if CP contribu)on dominates
• How to use these ra)os: • If one sees peaks in the measured cumulants at some μB • Calculate these ra)os around the peak • If equal to 1 Parameter independent way of verifying
that the fluctua)ons you see are due to the CP
r4 =κ2
2p2π
κ4πκ4p
r1 = (Nπ/Np)1/2 , r2 = Nπ/Np,• Poisson contribu)on: r3 = r4 = 0
r3 =κ2p1π
κ2/33p κ1/3
3π
,
Constraining parameters
• If CP found, can constrain parameters by measuring cumulant ra)os near the CP
• Parameters appear in certain combina)ons in the cumulants can only constraint 4 independent (but not unique) combina)ons
κ4pN2π/κ4πκ2
2p,
• For example, some choices are:
1. using or 2. using or 3. using 4. using
G ξ
κ3p1πNp/κ4pNπ,G/g
λ4/λ2
3 κ2pκ4p/κ23p,
κ2p2πNπ/κ4πκ2p
κ2p2πN2p /κ4pN
2π
κ3pN3/2p /κ9/4
2p N1/4π .λ2
3/g3
Outline
• Introduc)on • Cri)cal Contribu)on to Par)cle Mul)plicity Fluctua)ons
• Ra)os of Fluctua)on Observables • Summary
Summary
• We used par)cle mul)plicity fluctua)ons as a probe to the loca)on of the CP
• Higher cumulants of event-‐by-‐event distribu)ons are more sensi)ve to cri)cal fluctua)ons
• Constructed cumulant ra)os to iden)fy the CP loca)on with reduced parameter uncertain)es
• CP signature: Non-‐monotonic behavior, as a func)on of collision energy, of mul)plicity cumulants
• If CP is found, showed how to use cumulant ra)os to constraint the values of the non-‐universal parameters
Thank you!