Heavy flavours in heavy ion
collisions at the LHC
Francesco Prino
INFN – Sezione di Torino
DNP Fall Meeting, Newport Beach, October 25th 2011
Heavy Ion Collisions
Basic idea: compress large
amount of energy in a very small
volume
 produce a “fireball” of hot matter:
 temperature O(1012 K)
 ~ 105 x T at centre of Sun
 ~ T of universe 10 µs after Big Bang
Study nuclear matter at extreme
conditions of temperature and
density
 Collect evidence for a state where
3 flavours; (q-q)=0
 F. Karsch,
Nucl.Phys.A698 (2002) 199
quarks and gluons are deconfined
(Quark Gluon Plasma) and study
its properties
 Phase transition predicted by
Lattice QCD calculations
 TC ≈ 170 MeV  C ≈ 0.6 GeV/fm3
2
Heavy quarks as probes of the
medium
p
K
e,m D
D
b quark
c quark
B
n
e,m
D
Hard probes in nucleus-nucleus
collisions:
 Produced at the very early stage
of the collisions in partonic
processes with large Q2
 pQCD can be used to calculate
initial cross sections
 Traverse the hot and dense
medium
 Can be used to probe the
properties of the medium
3
Parton energy loss and nuclear
modification factor
Production of hard probes in AA expected to scale with the number
of nucleon-nucleon collisions Ncoll (binary scaling)
PbPb measurement
Parton energy loss while traversing the medium
 Medium induced gluon radiation
 Collisions with medium constituents
Observable: nuclear modification factor
1 dN AA / dpT
QCD medium
RAA ( pT ) 
~
N coll dN pp / dpT
QCD vacuum
If no nuclear effects are present -> RAA=1
Effects from the hot and deconfined medium:
-> breakup of binary scaling -> RAA1
But also cold nuclear matter effects give rise to
RAA1
 e.g. Shadowing, Cronin enhancement
 Need control experiments: pA collisions
pp reference
4
Heavy quark energy loss
Energy loss DE depends on
 Properties of the medium: density,
temperature, mean free path
 Path length in the medium (L)
 Properties of the parton:
 Casimir coupling factor (CR)
 Mass of the quark (dead cone effect)
gluonstrahlung probability ∝
1
𝜃2 +
𝑚𝑄 /𝐸𝑄
2 2
 Dokshitzer and Kharzeev, PLB 519 (2001) 199
DEquark  DE gluon
, DEmassive quark  DElight quark

RAA ( B)  RAA ( D)  RAA (p )
 Wicks, Gyulassy, Last Call for LHC predictions
5
Reaction
plane
Flow
Out-of-plane
Flow
Azimuthal anisotropy
Initial geometrical anisotropy
in non-central heavy ion
collisions
Y
In-plane
X
 The impact parameter selects a
preferred direction in the
transverse plane
Re-scatterings among produced particles convert the initial
geometrical anisotropy into an observable momentum anisotropy
 Collective motion (flow) of the “bulk” (low pT)
In addition, path-length (L) dependent energy loss in an almondshaped medium induces an asymmetry in momentum space
 Longer path length -> larger energy loss for particles exiting out-of-plane
Observable: Fourier coefficients, in particular 2nd harmonic v2,
called elliptic flow
dN N 0
1  2v2 cos2  RP   ....

d 2p


v2  cos 2  RP 
6
Heavy flavour v2
Due to their large mass, c and b quarks should take longer
time (= more re-scatterings) to be influenced by the
collective expansion of the medium
 v2(b) < v2(c)
Uniqueness of heavy quarks: cannot be destroyed and/or
created in the medium
 Transported through the full system evolution
 J. Uphoff et al., arXiv:1205.4945
7
PbPb collisions at the LHC
Pb-Pb collisions at the LHC
 √sNN=2.76 TeV (≈ 14x√sNN at RHIC)
 Delivered Integrated luminosity:
10 mb-1 in 2010
166 mb-1 in 2011
 3 experiments (ALICE, ATLAS, CMS)
8
Heavy flavour reconstruction
Semi-leptonic decays
(c,b)
Displaced J/y (from B decays)
e,m
J/y
m+
m-
B
Lxy
Primary
vertex
B,D
jet b-tagging
Full reconstruction of D meson
hadronic decays
D0  K- π+
D+  K- π+ π+
D*+  D0 π+
Ds+  K- K+ π+
9
ALICE + ATLAS + CMS
Complementary rapidity and pT coverage
DISCLAIMER: acceptance plots refer to published measurements in pp
10
How to: displaced tracks
Lower mass heavy flavour hadrons decay weakly:
 Lifetimes: ≈0.5-1 ps for D and ≈1.5 ps for B
 ct: ≈100-300 mm for D and ≈ 500 mm for B
Possibility to detect decay vertices/displaced tracks
 Tracking precision plays a crucial role
Track impact parameter:
distance of closest
approach of a track to the
interaction vertex
 ALICE, JHEP 09 (2012) 112
11
How to: particle identification
 ALICE, arXiv:1205.5423
ALICE MUON ARM
ALICE TPC
dE/dx vs. p
ALICE EMCAL
E/p for TPC e
ALICE TOF
time (ns) vs. p
 ALICE, JHEP 09 (2012) 112
12
... before going to the results
13
Is there evidence for parton
energy loss?
 CMS, EPJC 72 (2012) 1945
 ALICE, arXiv:1210.4520
Charged particle spectra suppressed in AA w.r.t. pp (RAA<1)
 Larger suppression at LHC than at RHIC
 Maximum suppression for charged particles at pT≈6-7 GeV/c
First results from pilot pPb run confirm that it comes from a
final state effect
14
Are heavy flavours well
calibrated probes?
Do we understand
their production
in pp?
 ALICE, arXiv:1205.5423
 CMS, EPJC 71 (2011) 1575
 ALICE, JHEP 1201 (2012)
 CMS, PRL 106 (2011) 112001
YES! pQCD
predictions agree
with data within
uncertainties
15
Nuclear modification factor
E
E-DE
16
Heavy flavour decay electrons
e
Inclusive electron spectrum with
two different PID analyses:
TPC+TOF+TRD and TPC+EMCAL
Subtract background electrons
 Electron pair invariant mass method
 Cocktail method
Inclusive-background = c+b
pp reference:
 7 TeV pp data sacled to 2.76 TeV
for pT<8 GeV/c
 FONLL for pT>8 GeV/c
17
Heavy flavour decay electrons
e
Inclusive electrons – cocktail
 = c+b
pp reference:
 7 TeV pp data sacled to 2.76
TeV for pT<8 GeV/c
 FONLL(pQCD) for pT>8 GeV/c
1 dN AA / dpT
RAA ( pT ) 
N coll dN pp / dpT
Clear suppression in the pT range 318 GeV/c
-> amounts to a factor of 1.5-3 in
3<pT<10 GeV/c
18
Heavy flavour decay muons
at forward rapidity
Single muons at forward rapidity
(-4<h<-2.5)
 Punch-through hadrons rejected by
m
requiring match with trigger chambers
 Subtract background m from p/K decay
 Extrapolated from mid-rapidity measurement with an
hypothesis on the rapidity dependence of RAA
pp reference measured at 2.76 TeV
Suppression by a
factor 2-4 in 010% centrality
Less suppression
in peripheral
collisions
 ALICE, PRL 109 (2012) 112301
19
Heavy flavour decay muons
at midrapidity
Single muons in |h|<1.05, 4<pT<14 GeV/c
 Match tracks from Inner Detector and
Muon Spectrometer
 Use discriminant variables with different
distribution for signal and background
 Background: p/K decays in flight, muons from
hadronic showers, fakes
RCP ( pT ) 
N coll
Per
dN / dpT
Cent
N coll
Cent
dN / dpT
Per
Approximately flat vs. pT
 Trend difficult to evaluate
due to fluctuations in
peripheral bin
20
Electrons vs. muons
Similar RAA for heavy flavour
decay electrons (|h|<0.6) and
muons (2.5<y<4) in 0-10%
centrality
Direct comparison between RAA and RCP not possible
 Assuming ~no suppression for 60-80% centrality -> same size
of suppression also for muons in |h|<1.05
21
Can we separate charm and
beauty?
22
D mesons
K p
Analysis strategy
 Invariant mass analysis of fully
reconstructed decay topologies
displaced from the primary vertex
Feed down from B (10-15 %
after cuts) subtracted using
pQCD (FONLL) predictions
 Plus in PbPb hypothesis on RAA of D
from B
D0  K- π+
D+  K- π+ π+
D*+  D0 π+
23
D meson RAA
pp reference from measured D0, D+ and D* pT -differential
cross sections at 7 TeV scaled to 2.76 TeV with FONLL
 Extrapolated assuming FONLL pT shape to highest pT bins not
measured in pp
D0, D+ and D*+ RAA agree
within uncertainties
Strong suppression of prompt
D mesons in central collisions
-> up to a factor of 5 for pT≈10
GeV/c
24
Charm + strange: Ds+
First measurement of Ds+ in AA collisions
Expectation: enhancement of the
strange/non-strange D meson yield at
intermediate pT if charm hadronizes via
recombination in the medium
Strong Ds+ suppression
(similar as D0, D+ and D*+) for
8< pT <12 GeV/C
RAA seems to increase (=less
suppression) at low pT
 Current data do not allow a
conclusive comparison to other
D mesons within uncertainties
 Kuznetsova, Rafelski, EPJ C 51 (2007) 113
25
 He, Fries, Rapp, arXiv:1204.4442
D vs. heavy flavour leptons and
light flavours
To properly compare D and
leptons the decay kinematics
should be considered
 pTe ≈0.5·pTB at high pTe
Similar trend vs. pT for D,
charged particles and p±
 Maybe a hint of RAAD > RAAπ
at low pT
26
Data vs. models
Little shadowing at high pT
 suppression is a hot matter effect
 need pPb data to quantify initial state effect
D mesons
HF muons
 ALICE, PRL 109 (2012) 112301
Models of in-medium parton energy loss can describe
reasonably well heavy flavour decay muons at forward
rapidity and D mesons at midrapidity
27
J/y from B feed-down
J/y from B decays to access beauty inmedium energy loss
 Long B-meson lifetime -> secondary J/y’s
from B feed-down feature decay vertices
displaced from the primary collision vertex
 Fraction of non-prompt J/y from
simultaneous fit to m+m- invariant mass
spectrum and pseudo-proper decay length
distributions
m+
m-
J/y
B
Lxy
 J /y  Lxy ( J /y ) 
M J /y
pT ( J /y )
28
RAA of non-prompt J/y
Slow decrease of RAA with increasing centrality
Hint for increasing suppression (-> smaller RAA)
with increasing pT
 CMS, PAS HIN-12-014
29
Beauty vs. charm
Caveat: different
y and pT range
In central collisions, the expected RAA hierarchy is observed:
RAAcharm < RAAbeauty
30
b-jet tagging
Jets from b quark
fragmentation identified
(tagged) for the first time in
heavy ion collisions by CMS
jets are tagged by cutting on
discriminating variables
based on the flight distance
of the secondary vertex
 Enrich the sample in b-jets
 An alternative tagger based
only the impact parameter of
the tracks in the jet is used as
cross check
b-quark contribution
extracted using template fits
to secondary vertex invariant
mass distributions
 CMS, PAS HIN-12-003
31
Beauty vs. light flavours
Low pT: different suppression
for beauty and light flavours
 BEWARE:
1) not the same centrality
2) B->J/y decay kinematics
High pT: similar suppression
for light flavour and btagged jets
32
Reaction
plane
Flow
Out-of-plane
Flow
Azimuthal anisotropy
Y
In-plane
X
33
D meson v2
First direct measurement of D anisotropy in heavy-ion collisions
Yield extracted from invariant mass spectra of Kp candidates
in 2 bins of azimuthal angle relative to the event plane
1 p N IN  N OUT
v2 
R2 4 N IN  N OUT
-> indication of non-zero D meson v2 (3s effect) in 2<pT<6 GeV/c
34
Challenge the models
The simultaneous description of D meson RAA and v2 is a
challenge for theoretical models
35
Challenge the models
The simultaneous description of heavy flavour decay
electrons RAA and v2 is a challenge for theoretical models
36
Heavy flavours:
what have we learned so far?
Abundant heavy flavour production at the LHC
Allow for precision measurements
Can separate charm and beauty (vertex detectors!)
Indication for RAAbeauty>RAAcharm and RAAbeauty>RAAlight
More statistics needed to conclude on RAAcharm vs. RAAlight
Indication (3s) for non-zero charm elliptic flow at low pT
Hadrochemistry of D meson species
First intriguing result on Ds+ RAA, not enough statistics to conclude
37
Heavy flavours:
what next?
So far, an appetizer
What will/can come in next years (2013-2017):
pPb run -> establish initial state effects
Separate charm and beauty also for semi-leptonic channels
Improved precision on the comparison between charm and light
hadron RAA
More differential studies on beauty
And even more with the upgrades (2018):
High precision measurements of D meson v2 and comparison to light
flavours -> charm thermalization in the medium?
Charm baryons (Lc) -> study baryon/meson ratio in the charm
sector
High precision measurement of Ds+ RAA and v2
...
38
Backup
39
D meson dN/dpT
40
D and charged particle RAA
 ALICE, JHEP 09 (2012) 112
41
D meson RAA: LHC vs RHIC
42
Heavy Flavour electrons: LHC vs
RHIC
43
Ds/D0 and Ds/D+
44
RAA of non-prompt J/y
Hint of slow decrease of RAA with increasing
rapidity
Non-prompt J/y at midrapidity slightly less
suppressed than at forward rapidity
45
b-jet tagging
Jets from b quark
fragmentation identified
(tagged) for the first time in
heavy ion collisions by CMS
jets are tagged by cutting on
discriminating variables
based on the flight distance
of the secondary vertex
 Enrich the sample in b-jets
 An alternative tagger based
only the impact parameter of
the tracks in the jet is used as
cross check
b-quark contribution
extracted using template fits
to secondary vertex invariant
mass distributions
46
b-jet fraction vs. centrality
Fraction of b-jets over inclusive jet
Does not show a strong centrality dependence
47