The Charm and Beauty of Heavy Quarks in Heavy Ion Collisions at RHIC

1
The Charm and Beauty of Heavy Quarks in Heavy Ion
Collisions (at RHIC)

• Thomas Ullrich, BNL/Yale
• Ohio State University Seminar
• Feb 27, 2004

2
Heavy Flavor Production in Hadronic Collisions

• Reactions that produce heavy-flavor
• involve a hard scale ? pQCD
• At LO, charm and beauty production proceeds
  through
• quark-antiquark annihilation
• or gluon fusion

• Factorize calculations
• pQCD to calculate?cc production
• ?cc propagation and hadronization

LO, NLO, NNLO calculations magnitude, shape of
spectra Charm mass ? total cross section kT ? pT
spectra
Parton Distribution Functions (PDF) rapidity
dependence , vs dependence
3
Heavy Flavor Production in pp Collisions
At high energies, gluon fusiondominates the
production cross-section ? Heavy Flavor
production directly probes gluon distributions of
colliding particles
Pythia 6.208
General Heavy-flavor and quarkonia production is
theoretically not fully understood even in pp
collisions
4
What do we know about charm production in pp?

• Data mostly from fixed target experiments at SPS
  and Fermilab, but also CDF
• energy range 200800 AGeV (?s 19 38 GeV),
  ?s1.8 TeV
• pA linear nuclear A-dependence assumed spA
  spp Aa , a 1

charm
pp
pp
beauty
pp
pp
5
Charm production in pp at RHIC energies (?s200
GeV)?

• Different PDF sets or quark masses lead to
  different energy dependences
• All the curves are normalized at low energies
• ? the predictions for higher energies have a
  certain spread
• ? changing PDF sets range 400800 mb at
  RHIC energies ? changing c quark mass by
  ?15 range 300700 mb

RHIC
CTEQ6M
compiled by H. Woeri and C. Lourenco (SQM2003)
6
Heavy Flavor Production in AA Collisons

• So far
• Production little studied (CERN/SPS low s)
• Charmonium suppression (melting of J/Y due to
  color screening) used as signature of QGP
• But
• Low x heavy quark pair is produced over long
  distances that can exceed the size of the nuclei
  (even though the matrix element of hard processes
  is dominated by short distances)
• ? sensitive to medium ? good probe for high
  density matter (QGP)
• RHIC and LHC
• gluon density in nucleii ?
• energy loss due to interaction with medium ?
• shadowing (higher twist effect ? gluon density
  steeply rises with E)
• cannot study J/Y suppression without
  understanding open charm

7
Hidden and Open Charm Both Needed
SPS ?s 17 GeV
RHIC ?s 200 GeV
At RHIC open charm production provides reference
and may be the only mean to understand charmonium
suppression (same gluon conditions in the initial
stage)
8
Heavy Flavor Production in AA Collisons

• So far
• Production little studied (CERN/SPS low s)
• Charmonium suppression (melting of J/Y due to
  color screening) used as signature of QGP
• But
• Low x heavy quark pair is produced over long
  distances that can exceed the size of the nuclei
  (even though the matrix element of hard processes
  is dominated by short distances)
• ? sensitive to medium ? good probe for high
  density matter (QGP)
• RHIC and LHC
• gluon density in nucleii ?
• energy loss due to interaction with medium ?
• shadowing (higher twist effect ? gluon density
  steeply rises with E)
• cannot study J/Y suppression without
  understanding open charm
• Heavy Flavor Production is a new frontier in
  Heavy-Ion Collisions

9
Thermalization of Heavy Quarks

• Strangeness
• 1. Lower energy threshold
• Key concept is
• TQGP TC ms 150 MeV/c2
• 2. Larger production cross-section
• 3. Pauli blocking (finite chemical potential)

• Charm
• No thermal production
• TQGP
• Charm quarks interact with evolving QGP (light
  quarks, g) makes thermalization theoretically
  possible
• Should show in
• production rate (J/? recombination)
• pT-spectra (thermal, flow)
• elliptic flow (v2)

10
Thermalization I

• Kinetic Formation Model
• Thews, Rafelski hep-ph/0305316
• Charm produced in initial nucleon-nucleon
  collision
• Competition between formation and breakup in
  region of deconfinement (QGP)
• formation c ?c g ? J/Y g
• breakup via gluons
• formation ? N2cc
• 30 effect in rate at RHIC
• negligible at SPS

11
Thermalization II

• Quark Coalescence Model
• Greco, Ko, Rapp nucl-th/0312100
• Approach successful in light-quark sector
• anomalous ?p/p 1 ratio
• constituent-quark scaling of v2
• Applied to charm production

D mesons
e from D decays
If complete thermalization at RHIC J/Y
dN/dy increased by factor 3
pT slope increased by factor 2 D-mesons less
pronounced effect in D?eX
but v2 2 ? larger preserved
in single-e spectra
12
Suppression of inclusive hadron yield at high pT
STAR, nucl-ex/0305015
pQCD Shadowing Cronin
energy loss
pQCD Shadowing Cronin Energy Loss

• central AuAu collisions factor 4-5
  suppression
• pT5 GeV/c suppression independent of pT
• pQCD describes data only when energy loss is
  included

13
Energy Loss of Heavy Quarks
Vacuum radiation suppressed in the dead cone for
? Dokshitzer, Kharzeev, PLB 519 (2001) 199
Medium induced radiation fills dead-cone total
E-loss comparable but smaller than for m0
Arnesto, Salgado, Wiedemann, hep-ph/0312106
? enhancement of D/? ratio at moderate
high pT (5-10 GeV/c)
14
Open Charm at SPS (pA)
Mmm-
Only measurement from NA50 via correlated m-pairs
from?DD decays in the IMR ?DD/Drell-Yan 4.2 ?
0.9 In pA at 450 AGeV with lots of PYTHIA scc
36.2 ? 9.1 mb (xF 0)
15
Open Charm at SPS (AA)
peripheral
central
Unexpected excess in the IMR in PbPb _at_ 160 AGeV

• excess grows with centrality
• mass distribution compatible with?DD
• pT, y, cos? compatible with?DD
• but checks not conclusive
• Still controversial charm, thermal,
• background subtraction method ?

16
What do we know about charm at RHIC in AA?
PHENIX AuAu_at_130GeV Phys. Rev. Lett. 88,
192303 (2002)
17
Heavy Flavor and the STAR Experiment
STAR measuring charm in leptonic channels b, c
? e X
STAR measuring charm in hadronic channels
D0 ? K ? (B.R. 3.8) and K p r (B.R. 6.2) D? ?
K ? p (B.R. 9.1) D ? D0p (B.R. 68) Lc ? p K p
(B.R. 5)
all with TPC only
18
D0 in STAR Analysis Methods
K
p

• Event-Mixing Technique
• Identify charged Kaon and Pion tracks
• through energy loss in TPC
• Produce oppositely charged K-? pair
• invariant mass spectrum in same event
• Obtain background spectrum through
• mixed event
• Subtract background and get D0 spectrum
• dAu 15.7 M events

K-? Pair Invariant Mass
19
D0 in dAu Collisions
D0D0 to increase statistics
0 Gaussian function linear Residual background

• Mass and Width consistent
• with PDG values considering
• detector effects
• mass1.8670.006 GeV/c2
• mass(PDG)1.86450.005 GeV/c2
• mass(MC)1.865 GeV/c2
• width13.76.8 MeV
• width(MC)14.5 MeV

20
D? Mesons in dAu Collisions
D ? D0p (B.R. 68) Decay Kinematics (Pythia)
Golden channel for open charm study Standard
method M(D) M(D0)145.421 MeV Width1 MeV
Difficulty the low efficiency of the soft pion
reconstruction STAR full field tracks curl up
for pT 21
D? Mesons in dAu Collisions
D0 ? K-pp0 (B.R. 13.1)
D
D0 from D decays
D0
2.42.4Masses and Widths OK m(D)-m(D0) 0.14670.00016
GeV/c2 m(D)-m(D0)(PDG)0.1454 GeV/c2 m(D)-m(D0)(
MC)0.1451 GeV/c2 width0.430.14
MeV width(MC)0.67 MeV
22
D? Mesons in dAu Collisions

• D?Kpp (B.R. 9.1)
• 3-body decay ? more background
• high-pT reach

• D mass1.8640.0052 GeV/c2
• D mass(PDG)1.869 GeV/c2
• D mass(MC)1.8680.002 GeV/c2
• width 13.833.7 MeV
• width (MC)14.91.6 MeV

23
D/D0 Ratio

• CDF hep-ex/0307080
• HERA www-h1.de/h1/www/
  publications/conf_list.html
• ee- hep-ph/0312054
• Statistical model Andronic et al.
  nucl-th/0209035

Good agreement with other experiments D/D0 ?
D/D0 0.40 ? 0.09 (stat) ? 0.13 (sys)
24
D-Meson Spectra in dAu
Assuming s(D) s(D) and scale s(D) and
s(D) to match D0 by D/D00.40
25
Charm Quark Hadronization at RHIC
NLO pQCD predictions by R. Vogt, Int. J. Mod.
Phys. E12 (2003) 211
Phenix PRL 88, 192303(2002) D.
Kharzeev,hep-ph/0310358
bare c-quark spectrum, normalized to measured
dn/dy

• Open charm spectra is hard NLO c-quark spectrum
  D spectrum
• observed in fixed target exp. at lower energies
• solved by intrinsic kT model to counter-balance
  effect of c-quark hadronization
• doesnt work at RHIC because spectrum is too
  broad
• NLO underestimates s (factor of 3 at pT10GeV/c
  for MRST)
• Harder fragmentation function ?
• Hadronize through quark recombination mechanism
  (Hwa at QM2004) ?

26
Tools to Identify Electrons in STAR

• Time-of-Flight MRPC
• p/K separation up to 1.6 GeV/c
• p/K separation up to 3 GeV/c
• Thus cover wider range of (p,K,p) pT
• only 10 modules installed (p/30, 0.5

• Electromagnetic Calorimeter
• pp and dAu runs at 200 GeV/A
• 0
• Full azimuthal coverage
• 60 modules
• tower (Dh, Df)tower (0.05, 0.05)
• shower max (Dh, Df) (0.007, 0.007)
• dE/E 16/vE

27
Electron PID with MRPC TOF/TPC and EMC

• EMC
• use TPC for p and dE/dx
• use Tower E ? p/E
• use SMD shape to reject hadrons
• e/h discrimination power 105
• works for pT 1.5 GeV/c

• ToF
• use TPC and ToF PID
• works for pT

28
Inclusive Electron Spectra in pp and dAu
29
Background Subtraction
Opening Angle
? conversion p0 , ? Dalitz decays Kaon decays ? ?
F vector meson decays heavy quark semi-leptonic
decays others (Drell-Yan)
Background
Signal
Invariant Mass Square

• Select an primary electron/positron (tag it)
• Select another opposite sign track anywhere in
  TPC

Signal
Rejected
Measure background
30
Background Subtracted Electron Spectra in pp
dAu
Semantics background subtracted in STAR
non-photonic in PHENIX
31
Nuclear Effects (Cronin) ?

• Within the errors consistent with binary scaling
  ...
• NOTE RdAu for a given pT comes from heavy quarks
  from a wide pT range

32
Charm Cross-Section and PYTHIA

• Measured D combined with
• measured electron
• spectra
• better scc

• The more we learn about heavy flavor in PYTHIA
  the less
• we believe to learn something from it.
• Vary parameters (K, kT, processes, PDF, mC, etc.)
  within reasonable limits
• ? s changes up to factor 2
• ? ds/dpT at high pT up to a factor of 10

33
Consistency between electron data sets in pp at
RHIC
Single electrons in pp at ?s 200 GeV

• STAR data slightly higher than PHENIX
• STAR Generated electrons from measured D-meson
  spectra consistent with non-photonic electron
  spectra

34
Energy Loss of Heavy Flavor at RHIC ?
Central AuAu (PHENIX) vs. pp (STAR)
Central AuAu (PHENIX) vs. fit to pp (PHENIX)

• Increasing deviation towards higher pT
• Suppression of charm at high-pT likely
• magnitude still uncertain but looks big (to me)
  remember p,h factor 4-5
• need STAR AuAu/STAR pp (soon )

35
Summary

• STAR measured D0,D and D in dAu at vs200 GeV
  with a pT coverage of 0
  rapidity
• scc 1.18 0.21 0.39 mb from
  directly-measured open charm
• D/D0 0.400.09(stat)0.13(sys) consistent with
  other experiments
• D pT-spectrum is hard and coincides with the bare
  c-quark distribution from NLO calculations
• STAR measured single electrons in the range 1 pT
• D and single electron spectrum are consistent (pT
  
• scc 1.44 0.20 0.44 mb from D0 and single
  electron spectra
• Electrons with pT 3.5 GeV/c are mostly from
  beauty decays
• First RHIC measurement sensitive to beauty cross
  section!!!
• Comparison with PHENIX AuAu data indicates charm
  suppression in AuAu
• Lots of STAR AuAu data taken as we speak (? D,
  single electrons )

36
Thermalization I

• Statistical Hadronization Model
• Braun-Munzinger, Stachel, Redlich,
  nucl-th/0303036
• Charm produced in primary hard collisions and
  equilibrate
• Complete color screening (no J/?)
• At hadronization charm quarks are distributed
  into hadrons according to statistical rules
  (canonical thermodynamics)
• grand-canonical density ? hadrons
• describes J/? at SPS but needs scc increased by
  3 as compared to pQCD
• dramatic effects at LHC
• predict RHIC ratios e.g. D/D0 0.455
  
• (0.33 in
  PYTHIA)