Transcript Investigation of Hot QCD Matter with Relativistic Heavy Ions

Investigation of
Hot QCD Matter
with Relativistic Heavy Ions
Theoretical Aspects
See also:
B.M. & J. Schukraft & B. Wyslouch
First results from Pb+Pb collisions at the LHC
Ann. Rev. Nucl. Part. Sci. 62, 361 (2012)
B. V. Jacak & B.M.
The Exploration of Hot Nuclear Matter
Science 337, 310-314 (2012)
Berndt Müller
Nobel Symposium
on LHC Results
Uppsala
13 May 2013
What it’s all about
Imagine...
.
...heating a liquid (nuclear matter) until it turns into a gas
(nucleon/hadron gas) at approximately 100 billion degrees.
But when heated to 20 times higher temperature (2 trillion
degrees) you find that it suddenly turns into a liquid again,
in fact, into the most perfect liquid ever observed.
How is this possible? [We don’t really know.]
What happens at even higher temperatures? [We know.]
Is there a true phase transition? [We do not yet know.]
2
The Theory Toolkit

Perturbation theory (vacuum and thermal)


Semiclassical gauge theory


For static thermodynamic quantities
Holography


For initial state (at small x)
Lattice gauge theory


For jets and heavy quarkonia
For “universal” properties of strongly coupled gauge theory
Transport theory

For bulk QGP evolution (viscous hydro, Langevin, Boltzmann)
3
QCD EOS at μB = 0
4
QCD Phase Diagram
5
LHC vs. RHIC
Exponential fit in pT
T = 304 ±51 MeV
Exponential fit in pT
T = 221 ±23 ±18 MeV
New record “temperature” measured in
Pb+Pb at LHC:
Hydro fits
Tinit ≥ 300 MeV
TLHC = 1.37 TRHIC.
Reflects larger initial temperature Tin, but
not to be identified with Tin.
6
“LHC Bang” vs. Big Bang
“Big Bang”
Penetrating probe: photons
Chemical probes: light nuclei
Bulk probe: temperature fluctuations
lumpy initial
energy density
QGP phase
quark and gluon
degrees of freedom
Initial-state quantum fluctuations propagate
into to “macroscopic” final-state fluctuations
by hydrodynamic response.
kinetic
freeze-out
distributions and
correlations of produced
particles
QGP Phase
Boundary
“LHC Bang”
Penetrating probes: photons, jets
Chemical probes: hadrons
Bulk probe: flow fluctuations
7
Hot QCD matter properties
Which properties of hot QCD matter can we hope to determine and how ?
Easy
for
LQCD
Very
Hard for
LQCD
Hard for
LQCD
Easy
for
LQCD
Tmn
Û e , p, s
Equation of state: spectra, coll. flow, fluctuations
1 4
h = ò d x Txy (x)Txy (0)
T
Shear viscosity: anisotropic collective flow
4p 2a sC R
† a+i
a+
q̂ =
dy
U
F
(y
)UF
i (0)
2
ò
Nc - 1
ü
ï
ï
ï Momentum/energy diffusion:
4p 2a sC R
† - a+
a+
ê =
dy iU ¶ A (y )UA (0) ý
2
ò
Nc - 1
ï parton energy loss, jet fragmentation
ï
4pa s
† a0i
a
b0i
b
k=
dt U F (t )t UF (0)t
ï
ò
3N c
þ
mn
Pem
(k) = ò d 4 x eikx j m (x) jn (0)
1
mD = - lim
ln U †E a (x)UE a (0)
|x|®¥ | x |
QGP Radiance: Lepton pairs, photons
Color screening: Quarkonium states
8
What we hope to learn
Except for Πμν all medium properties are expressed as correlators of color gauge
fields. They reflect the gluonic structure of the QGP.
At high Q2 (or high T), the QGP is weakly coupled and has quasiparticle structure.
At which Q2 (T) does the QGP become strongly coupled?
Does it still contain quasiparticles?
Which observables (jets?) tell us where the transition occurs?
9
The “standard model”
CGC
“Glasma”
Hydrodynamics
Color Glass
Condensate
Hadronic gas
“Glasma”
10
Perfect liquidity
11
Viscous hydrodynamics
Hydrodynamics = effective theory of energy and momentum conservation
energy-momentum tensor
¶ mT
mn
= 0 with
T
mn
=
ideal fluid
+
= (e + P)u u - Pg
m n
mn
dissipation
+P
mn
l
é dPmn
ù
m nl
n ml du
m n
n m
mn
tP ê
+ u P +u P
=
h
¶
u
+
¶
u
trace
P
ú
dt û
ë dt
(
)
(
)
Input: Equation of state P(ε), shear viscosity, initial conditions ε(x,0), uμ(x,0)
Shear viscosity η is normalized by density: kinematic viscosity η/ρ.
Relativistically, the appropriate normalization factor is the entropy density
s = (ε+P)/T, because the particle density is not conserved: η/s.
12
Holographic argument
General argument [Kovtun, Son & Starinets, PRL 94 (2005) 111601] based on
the holographic duality (AdS/CFT) between thermal QFT and string
theory in five-dimensional curved space with a “black-hole” metric.
(3+1)-D world
(t,x)
(0,0)
r0
horizon
1
r0 =
pT
Dissipation in QFT is dual to the absorption of gravitons by the black
hole:
8p G
3
iw t
é
ù
s abs w =
dt
d
x
e
T
t,
x
,T
0,0
¾w¾¾
® a (horizon area)
xy
xy
®0
ò
ë
û
w
s abs (0)
a
s
a
h 1
Thus: h =
=
=
because s =
®
=
16p G 16p G 4p
4G
s 4p
( )
( ) ( )
13
Are αs and Nc large enough?
Coupling constant corrections [Buchel, Liu & Starinets NPB 707 (2005) 56]:
h
1
=
s 4p
-3/2
é 135
êë1+ 8 V (3)( 8pa s N c ) +
ù 1
úû » 4p [1+ 0.18 ]
for αs = 0.3
Classical gravity holography requires 4παsNc ≈ 12 ≫ 1, thus gravity dual calculations
may capture many aspects of the QGP on thermal scales, but probably not physics
of “hard” QCD probes (jets), which are controlled by pQCD.
Lattice results on thermodynamic quantities for SU(Nc)
gauge theory indicate rapid
convergence with Nc.
[Panero, PRL 103 (2009) 232001]
14
Holographic insights
We do not (yet) have a faithful gravity dual of QCD → focus on universal
properties of strongly couple gauge theories:
Deviation from isotropy
Approach to hydrodynamics:
Longitudinally expanding matter
obeys minimally viscous hydrodynamics after τ ≥ 0.7/T
[Heller et al, PRL 108, 201602 (2012)]
full evolution
hydrodynamics
τT = 0.7
τT
Holographic thermalization:
Regions of space thermalize as fast as possible (at the speed of light)
⟺ information propagates with the speed of light
[Balasubramanian et al., PRL 106,191601 (2011), PRD 84, 026010 (2011)]
15
Event by event
Initial state generated in A+A
collision is grainy
event plane ≠ reaction plane
⇒ eccentricities
ε1, ε2, ε3, ε4, ... ≠ 0
⇒ flows v1, v2, v3, v4, ... ≠ 0
What controls the graininess:
Nucleon fluctuations
or
Parton fluctuations ?
æ
ö
dN
2p
= N 0 ç 1+ 2å vn ( pT ,h )cos n (f - y n ( pT ,h ))÷
è
ø
df
n
anisotropic flow coefficients
16
RHIC vs. LHC
Saturated Glasma
Gale, Jeon, Schenke, Tribedy, Venugopalan, arXiv:1209.6330
LHC
MC-Glauber
BM & A. Schäfer,
PRD 85 (2012) 114030
RHIC
17
Fluctuation spectrum
Fluctuation spectrum encodes:
Structure of quantum fluctuations in the initial state (nuclei)
Differential damping by viscous effects, propagation in matter
Can the power spectrum of vn be used to determine η/s and vsound ?
M. Luzum et al.
WMAP5
The RHIC/LHC advantage:
There are many knobs to turn, not
just a single universe to observe.
18
Color opacity
19
Parton energy loss
Elastic energy loss:
dE
= -C2 ê
dx
q
q
q
Radiative energy loss:
L
q
q
Scattering centers
⇔ color charges
g
dE
= -C2 q̂ L
dx
q
ds
+
+i
q̂ = r ò q dq
= ò dx Fi (x )F (0)
2
dq
2
2
20
Color opacity
κLHC ≈ 0.6 κRHIC
Betz & Gyulassy, arXiv:1201.0281
αs runs!
Buzzatti & Gyulassy
Is T-dependence of q^ gradual or rather a steep change for T > Tc ?
21
QGP stopping power
Nuclear suppression
RAA
(δpT)LHC ≈ 1.3 (δpT)RHIC
but:
(dN/dy)LHC ≈ 2.2 (dN/dy)RHIC
⇒ QGP at LHC is less opaque
to hard partons than at RHIC
LHC
RHIC
(T3/q^)RHIC ≈ 0.6 (T3/q^)LHC
22
Jets in the medium
Q0
Qs-1 = minimal size
of probe to which the
medium looks opaque
Momentum scale of medium
Qs = qL » mD N scatt
Transverse size of jet
r^ jet = q jet L
23
Jet collimation
Casalderrey-Solana,
Milhano & Wiedemann
JPG 38 (2011) 035006
Guangyou Qin & BM
PRL 106, 162302 (2011)
24
Di-jet asymmetry
CMS data
ATLAS data
GY Qin & BM
PRL 106 (2011)
162302
ATLAS and CMS data differ in cuts on jet energy, cone angle, etc;
results depend somewhat on precise cuts and background corrections.
Several calculations using pQCD jet quenching formalism fit the data.
General conclusion: pQCD jet quenching can explain these data.
25
Jet modification synopsis
No change at small r, high pT
Depletion at intermediate r, pT
Excess at large r, low pT
26
c-quark quenching
Vertex detector in ALICE permits prompt D-meson identification:
D-mesons appear similarly quenched as light hadrons at same momentum.
This cannot be understood by radiative energy loss alone:
Clear evidence for elastic energy loss by scattering.
27
Color screening
28
In the good old days...
... life seemed simple: It’s all color screening
Lattice
QCD
a
Q
VQQ
−
Q
mD ~ gT
mD
Only the data did not
quite fit the theory!
29
The real story...
...is more complicated (as usual).
Q-Qbar bound state interacts with
medium elastically and inelastically!
ΓQQ
lth
Q
VQQ
−
Q
mD
ép +p
ù
¶
i
i
YQQ = ê
+ VQQ - G QQ + h ú YQQ
¶t
2
êë 2M
úû
2
Q
2
Q
g
lth ~ 2π/T,
mD ~ gT
Strickland, arXiv:1106.2571, 1112.2761;
Akamatsu & Rothkopf, arXiv:1110.1203
Heavy-Q energy loss and Q-Qbar
suppression are closely related
Recombination can also contribute
when c-quark density is high enough!
c
c
J/Ψ
30
J/ψ suppression
reco!?
Less J/ψ suppression at LHC
than at RHIC:
Full range of quarkonium states
is becoming accessible.
c-cbar recombination works
31
Summary & Questions
32
Lessons Learned


Theoretical framework becomes mature

LHC provides lever arm to probe log / low power dependence
of observables on kinematic variables

Theoretical framework developed at RHIC is confirmed
QGP at LHC is less strongly coupled than at RHIC

Average η/s at LHC larger than at RHIC


QGP at LHC is less color opaque than at RHIC


ηRHIC ≈ 0.6 ηLHC
(T3/q^)RHIC ≈ 0.6 (T3/q^)LHC
Growing richness of versatile matter probes

Flow, E-by-E fluctuations, jets, heavy quarks (quarkonia)
33
Experimental limitations

The combination RHIC+LHC has proven to be very
powerful, allowing for a meaningful variation of QGP initial
conditions

3 months HI beam time in 5 years severely limit the ability to
vary system size and beam energy at LHC for

making baseline p+p and p+A measurements

separating initial state fluctuations from geometric effects

exploring effects of strong transient magnetic fields

filling the gap between 100 GeV/A and 1.37 TeV/A

RHIC runs 4−5 months/year, but beam time still limits its
ability to perform all desirable measurements

Is it time for a reconsideration in the light of other priorities of
34
the LHC physics program?
Achievements and Opportunities

The QGP is a substance like no other:






relativistic, yet strongly coupled
a liquid that cools into a gas
non-superfluid liquid near the quantum limit of viscosity
matter that admits only diffusive transport of particles, yet
transports information at the speed of light (how?)
matter that requires a new vacuum state to exist
The QGP poses a challenge to theorists:

What is the microscopic structure of a relativistic liquid?
 How to solve a strongly coupled quantum system for which we do
not (yet) have a gravity dual?

The QGP continues to bring experimental surprises:

E.g., how small can a thermal QGP be and behave as “matter”?
35
Four Questions


What is the structure of the initial state, how does it thermalize?

Is the initial state (at small x) a color glass condensate?

Is thermalization governed by strong coupling?
How precisely can we measure QGP transport coefficients?


At what scale does the QGP become strongly coupled?


What does <25% uncertainty require for η and q^ ?
Can jet modification be used to determine the kinematic scale
separating quasiparticle and liquid domains in the QGP ?
Can we measure color screening in the QGP?

Can sequential quarkonium melting determine λDebye ?
36

Can (heavy) quark recombination demonstrate deconfinement?
Additional slides
37
Jets: core questions

What is the mechanism of energy loss ?
“radiative” = into non-thermal gluon modes
 “collisional” = directly into thermal plasma modes


How are radiative and collisional energy loss affected by
the structure of the medium (quasiparticles or not)?

Quasiparticle masses in weak coupling
 AdS/CFT inspired models with weak-strong coupling transition?

What happens to the lost energy and momentum ?
If “radiative”, how quickly does it thermalize = what is its
longitudinal momentum (z) distribution ?
 What is its angular distribution (the jet “shape”) = how much is
found in a cone of angular size R ?


How do the answers depend on the parton flavor ?
38
Anomalous viscosity?
Dusling, Epelbaum, Gelis &
Venugopalan, arXiv:1206.3336
4
Φ
hydro-like
domain
Anomalous viscosity
due to dynamically
local field domains:
Momentum change
per domain:
hA
9T 3
» 2
s 5g N c B 2 rm
Perturbative η/s
Effective η/s
η/s Bound
Dp » gQ a Ba rm
Asakawa, Bass & BM, PRL 96 (2006)
252301
PT - PL
e
2 éh ù
= 1/4 ê ú
t e ë s ûeff
39
Ridge in p+Pb
Initial-state 2-gluon correlations...
...or QGP hydrodynamics?
40
CGC or QGP ?
origin of the Ridge.”
41
“Fat” protons
Like all quantum systems, protons fluctuate in size.
Small protons - when all quarks are nearby - cause color transparency.
Large protons have large cross sections when colliding with a nucleus.
Assume a Gaussian model: P(σ) ~ e-σ/σ0 with σ0 ≈ 65 mb. Then ask:
What is the average σ when the proton collides with 30 nucleons in the
Pb nucleus? The answer is ⟨σ⟩ ≈ 150 mb !
What does a “fat” proton look like?
Is it three valence quarks far apart, connected by long gluon flux tubes?
Is it a proton surrounded by N virtual pions?
These configurations will have very different parton distributions.
42
Pion swarms
We can estimate P(Nπ) using data from Fermilab E866, which measured
DY muon pairs in p+1H and p+2H and deduced the isospin asymmetry of
the light quark sea distribution in the proton.
Most plausible explanation: p → n+π+
Integral = 0.118 ± 0.012
Implies: p → N+π has probability ~0.18.
⇒ probability of N+3π is ~ 10-3
Additive quark model:
⟨σ⟩N+3π ≈ 3⟨σ⟩N ≈ 175 mb
43