Heavy Quark Probes of Hadronization of Bulk Matter at RHIC Huan Zhong Huang Department of Physics and Astronomy University of California at Los Angeles Department of Engineering Physics Tsinghua University

Collisions at high p T (pQCD) At sufficiently large transverse momentum, let us consider the process: p + p  hadron + x 1) f(x,  2 ) – parton structure function 2)  ab->cd – pQCD calculable at large  2 3) D(z h,  2 ) – Fragmentation function To produce heavy quark pairs, the CM energy must>2m

Heavy Quark Production Mechanism Sensitive to initial gluon density and gluon distribution D0D0 J/  K+K+ l l K-K- e-/-e-/- e+/+e+/+ e-/-e-/- e+/+e+/+ Energy loss when propagating through dense medium Different scaling properties in central and forward region indicate shadowing, which can be due to CGC. Suppression or enhancement of charmonium in the medium is a critical signal for QGP. Sensitive to initial gluon density and gluon distribution

Parton Distribution Function Important

Uncertainties in gluon structure function of the proton x CTEQ5M1 CTEQ5HJ MRST2001 Band – experimental constraints J. Pumplin et al, JHEP07(2002)012

Fragmentation Functions

Fragmentation Functions from e+e Collisions Belle Data

Charm Mesons from Hadronic Collisions Charm meson p T ~ follow the NLO charm quark p T -- add k T kick -- harder fragmentation (  func or recombination scheme)

k T Kick? What about k L ? The x F distribution matches the NLO charm quark x F !

Belle Puzzle ! PRL 89, (2002)  (e + e -  J/  cc)  (e + e -  J/  X) = An order of magnitude higher than theoretical predictions -- B.L. Ioffe and D.E. Kharzeev, PRD 69, (2004) These results challenge our current understanding of how charm quarks/mesons are produced. We may question our view for the underlying charm production process, e.g., the universality of fragmentation process and the fragmentation schemes !

K ~ 1.5 Neutral D mesons LO QCD does not reproduce the cross sections ! K Factor !!

K ~ 4.5 Charged D mesons K factor energy, particle dependent ! Charm-Beauty different ! We don’t know the production mechanism at all !

Detecting D-Mesons via Hadronic Decays Hadronic Channels: –D 0  K  (B.R.: 3.8%) –D 0  K  (B.R.: 6.2%  100% (     ) = 6.2%) –D   K  p(B.R.: 9.1%) –D *±  D 0 π(B.R.: 68%  3.8% (D 0  K  ) = 2.6%  ) –  c  p K  (B.R.: 5%)

General Techniques for D Reconstruction 1.Identify charged daughter tracks through energy loss in TPC 2.Alternatively at high p T use h  and assign referring mass (depends on analysis) 3.Produce invariant mass spectrum in same event 4.Obtain background spectrum via mixed event 5.Subtract background and get D spectrum 6.Often residual background to be eliminated by fit in region around the resonance Exception D*: search for peak around m(D*)-m(D 0 ) = GeV/c 2 D0D0 D0D0 D*D*

Detecting Charm/Beauty via Semileptonic D/B Decays Semileptonic Channels: –D 0  e + + anything (B.R.: 6.87%) –D   e  + anything (B.R.: 17.2%) –B  e  + anything (B.R.: 10.2%)  single “ non-photonic ” electron continuum “ Photonic ” Single Electron Background: –  conversions (  0   ) –  0,  Dalitz decays – , , … decays (small) –Ke3 decays (small)

~7.6M AuAu 200GeV Run IV P05ia production 0~80% Min. Bias. |Vz| < 30cm Electrons can be separated from pions. But the dEdx resolution is worse than d+Au Log 10 (dEdx/dEdx Bichsel ) distribution is Gaussian.  2 Gauss can not describe the shoulder shape well. Exponential + Gaussian fit is used at lower p T region. 3 Gaussian fit is used at higher p T region.  2 /ndf = 65/46 0.3<p T <4.0 GeV/c TOF electron measurements |1/  -1|<0.03  2 /ndf = 67/70

Mass(e + e - )<0.15 GeV/c 2  Combinatorial background reconstructed by track rotating technique.  Invariant mass < 0.15 for photonic background. γ conversion π 0 Dalitz decay η Dalitz decay Kaon decay vector meson decays Dominant source at low p T Electron Spectrum

Charm p T Spectra Power-law function with parameters dN/dy, and n to describe the D 0 spectrum D 0 and e  combined fit Generate D 0  e  decay kinematics according to the above parameters Vary (dN/dy,, n) to get the min.  2 by comparing power-law to D 0 data and the decayed e  shape to e  data =1.20  0.05(stat.) GeV/c in minbias Au+Au =1.32  0.08(stat.) GeV/c in d+Au

Charm Total Cross Section 1.13  0.09(stat.)  0.42(sys.) mb in 200GeV minbias Au+Au collsions 1.4  0.2(stat.)  0.4(sys.) mb in 200GeV minbias d+Au collisions Charm total cross section per NN interaction Charm total cross section follows roughly Nbin scaling from d+Au to Au+Au considering errors Indication of charm production in initial collisions Systematic error too large !

Experimental Statistical and Systematic Errors c-cbar production CS PHENIX mb STAR mb Errors taken seriously High pT region does not contribute to total CS much. STAR data need to be compared with PHENIX data!

Heavy Quarks Unique Heavy Quark Flavors (Charm or Beauty) Heavy Flavors once produced – do not change to light flavor easily heavy quark production can be calculated from pQCD approach more reliably than light quarks Trace heavy quark flavors in nuclear collisions -- collision dynamics and hadronization mechanism Fragmentation versus Recombination/Coalescence Fragmentation p(heavy quark meson)/p(heavy quark) < 1 Recombination/Coalescence p(heavy quark meson)/p(heavy quark) >= 1

Nuclear Modification Factors Use number of binary nucleon-nucleon collisions to gauge the colliding parton flux: N-binary Scaling  R AA or R CP = 1 simple superposition of independent nucleon-nucleon collisions !

Charm and Non-photonic Electron Spectra 1.13  0.09(stat.)  0.42(sys.) mb in 200GeV minbias Au+Au collsions Total charm  Binary Scaling suppression at high pT

Charm Nuclear Modification Factor STAR: Phys. Rev. Lett. 91 (2003) Suppressions!! R AA suppression for single electron in central Au+Au similar to charged hadrons at 1.5<p T <3.5 GeV/c Heavy flavor production IS also modified by the hot and dense medium in central Au+Au collisions at RHIC

electrons  Kp d hadrons High p T Electron ID dE/dx from TPC SMD from EMC

hadrons electrons High p T Electron ID p/E from EMC After all the cuts

The shape and yield at high pT Note: FONLL – effective fragmentation function harder than commonly used Peterson function! STAR – difference ~ 5.5 PHENIX -- ~1.7 (?)

Non-photonic electrons R AA -- similar magnitude as light hadrons -- STAR-PHENIX data consistent in the overlapping region The high pT region n-p electron R AA surprising ! Non-photonic electron R AA

Heavy quark energy loss: Early Expectations Y. Dokshitzer & D. Kharzeev PLB 519(2001)199 Radiative energy loss of heavy quarks and light quarks --- Probe the medium property ! Heavy quark has less dE/dx due to suppression of small angle gluon radiation “Dead Cone” effect M. Djordjevic, et. al. PRL 94(2005) J. Adams et. al, PRL 91(2003) What went wrong?

Radiative Energy Loss not Enough Moore & Teaney, PRC 71, (2005) Large collisional (not radiative) interactions also produce large suppression and v 2

Charm Quark in Dynamical Model (AMPT) Large scattering cross sections needed !

Does Charm Quark Flow Too ? Reduce Experimental Uncertainties !! Suppression in R AA  Non-zero azimuthal anisotropy v 2 !

B and D contributions to electrons Experimental measurement of B and D contributions to non-photonic electrons ! Direct measurement of D and B mesons

Poor (Wo)Man’s Approach to Measure B/D Contributions to Electrons – e-h correlations B D PYTHIA Simulations of e-h correlations from p+p X. Lin hep-ph/

B does not seem dominant at pT 4.5 GeV/c Preliminary STAR Data Xiaoyan Lin – STAR presentation at Hard Probe 2006

Open Issues Phenix and STAR results  Converge?! Systematic errors on non-photonic electrons under control ! Quantitative description for energy loss and p T spectra for light/heavy quarks Collectivity for heavy quarks?

Recombination  D S /D 0 PYTHIA Prediction Charm quark recombines with a light (u,d,s) quark from a strangeness equilibrated partonic matter  D S /D 0 ~ at intermediate p T !!!

J/  –Small: r ~ 0.2 fm –Tightly bound: E b ~ 640 MeV HG QGP  Observed in dileptons invariant mass spectrum  Other charmonia  ’ ~ 8%  ~ 32% Color Screening

J/psi Suppression and Color Screening QCD Color Screening: (T. Matsui and H. Satz, Phys. Lett. B178, 416 (1986)) A color charge in a color medium is screened similar to Debye screening in QED  the melting of J/ . cc Charm quarks c-c may not bind Into J/  in high T QCD medium The J/  yield may be increased due to charm quark coalescence at the final stage of hadronization (e.g., R.L. Thews, hep-ph/ ) Recent LQCD Calculation:

J/  Quark Potential Model

Lattice QCD Calculations

J/  from di-lepton Measurements J/    - PHENIX Data Branching ratios: e + e %;     5.88%

J/psi is suppressed in central Au+Au Collisions ! Factor ~ 3 the same as that at SPS Satz: Only  states are screened both at RHIC and SPS. Alternative: Larger suppression in J/psi at RHIC due to higher gluon density, but recombination boosts the yield up !

V 2 of J/psi V 2 of J/psi can differentiate scenarios ! pQCD direct J/psi should have no v 2 ! Recombination J/psi can lead to non-zero v 2 !

The case for partonic DOF/Deconfinement can be made with strange vector meson  cannot be made from KK coalescence !

J/  Suppression or Not Nuclear Absorption of J/  important at low energy important (SPS) ! Both QCD color screening and charm quark coalescence are interesting, which one is more important at RHIC? At RHIC the J/  measurement requires high luminosity running! Centrality and p T dependence important !

p T Scales and Physical Processes R CP Three P T Regions: -- Fragmentation -- multi-parton dynamics (recombination or coalescence or …) -- Hydrodynamics (constituent quarks ? parton dynamics from gluons to constituent quarks? ) Where does heavy quark fit?

The End