Rapidity and kT dependence of HBT correlations in Au+Au collisions at 200 GeV with PHOBOS

Two-particle correlations of identical charged pion pairs from Au+Au collisions at √ s NN = 200 GeV were measured by the PHOBOS experiment at RHIC. Data for the most central (0–15%) events were analyzed with Bertsch-Pratt (BP) and Yano-Koonin-Podgoretskii (YKP) parameterizations using pairs with rapidities of 0.4 < y < 1.3 and transverse momenta 0.1 < kT < 1.4 GeV/c. The Bertsch-Pratt radii decrease as a function of pair transverse momentum. The pair rapidity Yππ roughly scales with the source rapidity YYKP , indicating strong dynamical correlations. Identical-particle correlation measurements (Hanbury-Brown and Twiss, HBT) yield valuable information on the size, shape, duration, and spatiotemporal evolution of the emission source in heavy ion collisions. Experimentally, the correlation function C(q) is defined as C(q) = P (p1,p2) P (p1)P (p2) (1) where P (p1,p2) is the probability of a pair being detected with relative four-momentum q = p1 − p2, and P (p1) and P (p2) are the single particle probabilities. The numerator Rapidity and kT dependence of HBT in Au+Au at 200 GeV with PHOBOS 2 is determined directly from data, while the denominator is constructed using the standard event-mixing technique. The data reported here were collected using the PHOBOS two-arm magnetic spectrometer during RHIC Run II (2001). Details of the setup have been previously described in [1]. The spectrometer arms are each equipped with 16 layers of silicon sensors, providing charged particle reconstruction both outside and inside a 2 T magnetic field. The primary event trigger was provided by two sets of 16 scintillator paddle counters, which covered a pseudorapidity range 3 < |η| < 4.5. Details of event selection and centrality determination can be found in [2, 3]. The 0–15% most central events were used in this analysis, equivalent to 〈Npart〉 = 310 as determined by a Glauber model. The details of the track reconstruction algorithm can be found in [4]. Events with a reconstructed primary vertex position between -12 cm < zvtx < 10 cm along the beam direction were selected in order to optimize vertex-finding precision, track reconstruction efficiency, and momentum resolution. Only particles which traversed the entire spectrometer were used in the analysis. A 3σ cut on the distance of closest approach with respect to the primary vertex (dcavtx < 0.35 cm) was then applied. The final track selection was based on the χ 2 probability of a full track fit, taking into account multiple scattering and energy loss. The momentum resolution is ∆p/p ∼ 1% after all cuts. To identify pions, a cut three RMS deviations away from the expected mean value of the specific ionization 〈dE/dx〉 for pions was applied. Contamination from other particle species was studied using HIJING 1.35[5] and a GEANT 3.21 simulation of the full detector. The contamination from KK, pp, and p p pairs is less than 1%; non-identical pairs contribute less than 10% throughout the entire kT range. To reject ghost pairs, only one shared hit in the weak-field region and two shared hits in the strong-field region were allowed per pair. A two-particle acceptance cut was applied to both data and background; the criterion for pair acceptance was defined by ∆φ+ 2∆θ > 0.05 rad, where ∆φ and ∆θ are the relative pair separation in azimuthal and polar angle, respectively. About 7.3 million π+π+ and 5.5 million π−π− pairs survive all cuts. Systematic errors were determined by changing two-particle acceptance cuts, cuts in azimuthal separation, random seeds used in mixed-event background generation, as well as varying the definition of “event class” to create background events from pairs within narrow and broad vertex ranges. Because the event-mixed background is the product of tracks from different events, it does not a priori include any multiparticle correlations. In order to study the HBT correlation, it is necessary to apply a weight to account for the Coulomb effect. The Coulomb correction can be expressed solely as a function of relative 4-momentum q, FR(q) = Fc(q) Fpl(q) = ∫