STAR Highlights on Heavy Ion Physics

RHIC-STAR is a mid-rapidity collider experiment for studying high energy nuclear collisions. The main physics goals of STAR experiment are 1) studying the properties of the strongly coupled Quark Gluon Plasma, 2) explore the QCD phase diagram structure. In these proceedings, we will review the recent results of heavy ion physics at STAR.


Introduction
The experimental observations indicate the strongly coupled Quark Gluon Plasma (QGP) has been built up in the top energy heavy ion collsions at Relativstic Heavy Ion Collider (RHIC) [1,2]. With large acceptance and excellent particle identifiaction capbiltiy, the STAR experiment at RHIC is one of best mid-rapidty collider experiments to study the properties of new form of matter, QGP, the initial conditions in heavy ion collisions and the stucture of the Quantum Chromodynamics (QCD) phase diagram. STAR experiment has covered the beam energies of √ s NN = 7.7, 11.5, 14.5, 19.6, 27, 39, 62.4 and 200 GeV. In the top energy heavy ion collisions (Au + Au at √ s NN = 200 GeV), we focus on studying the nature of QGP. At the lower collision energy region, the main motivation is to explore the nuclear matter phase structure in the higher net-baryon region. The extracted baryonic chemical potental (µ B ) range based on a statistical model [3,4] from the 0 − 5% central collisions is 20 ≤ µ B ≤ 420 MeV which covers a wide region of the QCD phase diagram. In these proceedings, we are going to highlight a few results from the recent measurements of the STAR experiment.

Results on azimuthal anisotropy
The azimuthal anisotropy in the momentum space of final particles relative to the reaction plane is one of the most informative ways to study the properties of matter created in high energy heavy ion collisions. The coefficients v n from a Fourier-series expansion are used to characterize the event anisotropy quantitatively, which could be represented by the equation v n = cos[n(φ i − Ψ RP )] , where φ i means the azimuth of the i th particle in an event, Ψ RP means azimuth of the reaction plane and the angle brackets mean an average over all particles in all events. v 1 is referred to as directed flow, and v 2 as elliptic flow.

Directed flow
The hydrodynamic and nuclear transport models suggest the v 1 as a function of rapidity (y) in the midrapidity region offers sensitivity to expansion of the participant matter during the early stage of collisions. Further, a 3-fluid hydrodynamic calculation with a first-order phase transition between hadronic matter and a QGP predicts a minimum in dv 1 /dy slope of net-baryon near midrapidity at √ s NN ∼ 4 GeV, and this minimum has been termed the "softest point collapse" [5]. It motivates the energy dependence of directed flow measument. Figure 1 shows the directed flow slope in 10−40% Au + Au collisions for protons, anti-protons and net-protons. The energy dependence of proton dv 1 /dy involves an interplay between the directed flow of protons associated with baryon number transported from the initial beam rapidity to the vicinity of midrapidity, and the directed flow of protons from particle-antiparticle pairs produced near midrapidity. Thus, we propose the use of antiproton directed flow as a proxy for the directed flow of produced protons, and propose that the net-proton slope brings us a step closer to isolating the contributions from transported initial-state bayonic matter, as well as closer to the net-bayon hydrodynamic calculation. The detailed definition of net-proton slope could be found in [6]. In panel (c) of Figure 1 , it shows the dv 1 /dy slope of net-proton crosses zero between 27 and 39 GeV and remains positive up to 200 GeV, where the UrQMD model shows a monotonic trend.
The observed beam energy of the minimum in dv 1 /dy slope is higher than the energy of the minimum in the hydrodynamic prediction. The observation of net-proton dv 1 /dy slope is qualitatively consistent with the predicted signature of a first-order phase transition between hadronic and deconfined matter.    But more study on other final-state interaction effects, such as annihilation, is necessary to understand the measurement.

Elliptic flow
The most striking feature on the v 2 measurement is the observation of an energy dependent difference in v 2 between particles and their corresponding antiparticles [7]. This difference naturally breaks the number of consitutent quark scaling (NCQ) in v 2 which is regarded as an evidence of partonic collectivity in the top energy heavy ion collisions at RHIC. It indicates the hadronic degrees of freedom play a more important role at lower collision energies. Figure 2 shows the difference in v 2 between particles and their corresponding antiparticles as a function of baryonic chemical potential. Our data are compared to hydrodynamics + transport (UrQMD) hybrid model [8] and Nambu-Jona-Lasino (NJL) model [9] which considers both partonic and hadronic potential. The hydrid model could reproduce the baryon (proton) data, but fails to explain the mesons; whereas the NJL model could qualitatively reproduce the hadron splitting. By tuning the R v parameter which is related to the partonic potential, NJL model fails to reproduce the magnitude for all hadron species simultaneously. Collisions between prolate Uranium nuclei are used to study how particle production and azimuthal anisotropies depend on the initial geometry in heavy ion collisions. If body-body collisions do produce smaller multiplicities than tip-tip collisions, then we expect to see a negative slope in v 2 vs multiplicity in fully overlapping U + U collisions. A negative slope, however, can also be generated New Frontiers in Physics 2014 Colliding Energy √s NN (GeV) Au + Au Collisions at RHIC in these collisions by off-axis collisions with larger impact parameters that pass our selection criteria.
To assess the contribution from this effect, we also study Au + Au collisions as a control sample since Au nuclei are almost spherical. Figure 3 shows the v 2 of all charged particles as a function of the normalized multiplicity with | η |< 1 for two different systems. The left panel shows the results for the 1% most central events based on the smallest signal seen in the Zero Degree Calorimeter (ZDCs). Both ZDCs are used in the centrality determination. Both Au + Au and U + U show a strong negative slope, which indicates the effect of the impact parameter is still prominent. The right panel of Fig. 3 shows the same approach for the 0.1% most central events based on the ZDCs. The negative slope for Au + Au collision is smaller in magnitude, indicating the effects from non-central collisions are reduced and the variation in multiplicity is mainly driven by fluctuations. For U + U collisions, however, the slope becomes more negative as the centrality selection is tightened. This demonstrates that the variation of multiplicity in the 0.1% U + U collisions is dominated by the different geometries made possible by the prolate shape of the Uranium nucleus. We also compare the data to expectation from Glauber model and IP-Glasma calculations. It is clear that the IP-Glasma calculation based on gluon saturation describes the data better. QCD critical point. Theoretical studies indicate that higher order moments have stronger dependences on correlation length (ξ) than the variance ( (δN) 3 ∼ ξ 4.5 and (δN) 4 ∼ ξ 7 ). Thus higher moments have higher sensitivity to the QCD critical point [10,11]. This motivates the study on the kurtosis (κ) of net-proton (a proxy for net-baryon) and net-charge distributions to search for the QCD critical point at STAR. Figure 4 shows the kurtosis times variance (κσ 2 ) for net-proton (top panel) [12] and net-charge (bottom panel) [13] distributions at mid-rapidity in Au + Au collisions as a function of colliding energy for 0 − 5% and 70 − 80% collision centralities. The net-proton κσ 2 values for the 0-5% centrality selection at √ s NN = 19.6 and 27 GeV show a relatively larger deviation from Poisson and hadron resonance gas expectation values (which would correspond to uncorrelated emission and are close to unity) and the κσ 2 values from 70-80% peripheral collisions. The peripheral collisions are expected to create small systems and do not show significant bulk properties. The calculations of UrQMD which is a pure hadronic model that does not consider a phase transition show a monotonic behavior. Our data at lower energies are with large uncertainties, but a possible non-monotonic variation of the κσ 2 of the net-proton distribution is not excluded. High statistical dataset for Au + Au collisions < 20 GeV from the future second phase of the beam energy scan (BESII) will help answer the question. In the top panel, the hatched band shows the projection of the statistical errors from BESII.

Results on Chiral Magnetic Effect
The RHIC provides a good opportunity to study the parity-odd domains where the local imbalance of chirality results from the interplay of these topological configurations with the QGP. In heavy ion collisions, energetic spectator protons produce a magnetic field. This strong magnetic field coupled with the chirality asymmetry in the parity-odd domains, induces a separation of electric charge along the direction of the magnetic field. STAR has proposed a three particle correlator method to study the Chiral Magnetic Effect (CME) [14]. Figure 5 shows the collision energy dependence of the charge separation after subtraction of the background related to collective flow [15]. The results approach to zero when the beam energy is below 11 GeV. In order to draw a clear conclusion, more statistics are needed for the lower beam energies (< 20 GeV). The colored band around zero shows the estimated statistical errors for the results from RHIC BESII program. GeV and model calculations.

Results on heavy flavor production
Charm and beauty (c and b) quarks are created predominantly via initial hard scatterings in nucleon nucleon colisions and the production rate is calculable with perturbative QCD techniques. The large masses are expected to be retained during the interactions with the medium [17,18]. Therefore the heavy flavor production is a useful tool to study the medium properties of the early stages of the system when QGP is expected to exist. Figure 6 shows the nuclear modifcation factor R AA of D 0 as a function of the number of participant nucleons N part . We observe the suppression of D 0 meson production at large p T in Au + Au and U + U collisions follows a global trend as a function of N part . Compared to the light flavor hadrons (π ± ), D 0 mesons suffer similar amount of energy loss. It is consistent with physics picture of a substantial amount of charm-medium interactions. Figure 7 shows the R AA of Υ(1S+2S+3S) as a function of  disfavor Strickland model A, which is based on the heavy quark free energy scenario. The model of Strickland and Bazow incorporating lattice QCD results on screening and broadening of bottomonium with a potential based on heavy quark internal energy is consistent with the data [19]. Also, the strong binding scenario proposed by Emerick, Zhao and Rapp including additionally possible Cold Nuclear Matter (CNM) effects is consistent with our results [20].

Results on di-electron production
Di-electrons are produced throughout the whole evolution in a heavy ion collision and have negligible final state interactions with created nuclear matter. Hence, di-electrons can be considered as unique probes to study the hot and dense matter. Potentially, the di-electron could provide information of chiral dynamics in the low mass region M ll < 1 GeV/c 2 and QGP direct radiation in the intermediate mass region 1 < M ll < 3 GeV/c 2 . Figure 8 shows the efficiency-corrected di-electron invariant mass distribution from minimum-bias Au + Au collisions at √ s NN = 19.6, 17, 39, 62.4 and 200 GeV. The 200 GeV results are from [21]. The results of hadronic cocktails + in-medium model calculations are shown by the yellow bands [22]. The comparison of model calculations to the invariant mass dependence of di-electron low mass range yields supports the conclusion that, within experimental uncertainties, in-medium modifications of the ρ spectral function consistently describe the low mass region enhancement from 19.6 GeV to 200 GeV.