Pion, kaon, proton and anti-proton transverse momentum distributions from p+p and d+Au collisions at $\sqrt{s_{NN}} = 200$ GeV

Identified mid-rapidity particle spectra of $\pi^{\pm}$, $K^{\pm}$, and $p(\bar{p})$ from 200 GeV p+p and d+Au collisions are reported. A time-of-flight detector based on multi-gap resistive plate chamber technology is used for particle identification. The particle-species dependence of the Cronin effect is observed to be significantly smaller than that at lower energies. The ratio of the nuclear modification factor ($R_{dAu}$) between protons $(p+\bar{p})$ and charged hadrons ($h$) in the transverse momentum range $1.2<{p_{T}}<3.0$ GeV/c is measured to be $1.19\pm0.05$(stat)$\pm0.03$(syst) in minimum-bias collisions and shows little centrality dependence. The yield ratio of $(p+\bar{p})/h$ in minimum-bias d+Au collisions is found to be a factor of 2 lower than that in Au+Au collisions, indicating that the Cronin effect alone is not enough to account for the relative baryon enhancement observed in heavy ion collisions at RHIC.

Identified mid-rapidity particle spectra of π ± , K ± , and p(p) from 200 GeV p+p and d+Au collisions are reported. A time-of-flight detector based on multi-gap resistive plate chamber technology is used for particle identification. The particle-species dependence of the Cronin effect is observed to be significantly smaller than that at lower energies. The ratio of the nuclear modification factor (R dAu ) between protons (p +p) and charged hadrons (h) in the transverse momentum range 1.2 < pT < 3.0 GeV/c is measured to be 1.19±0.05(stat)±0.03(syst) in minimum-bias collisions and shows little centrality dependence. The yield ratio of (p +p)/h in minimum-bias d+Au collisions is found to be a factor of 2 lower than that in Au+Au collisions, indicating that the Cronin effect alone is not enough to account for the relative baryon enhancement observed in heavy ion collisions at RHIC. Suppression of high transverse momentum (p T ) hadron production has been observed at RHIC in central Au+Au collisions relative to p+p collisions [1,2,3,4]. This suppression has been interpreted as energy loss of the energetic partons traversing the produced hot and dense medium [5]. At intermediate p T , the degree of suppression depends on particle species. The spectra of baryons (protons and lambdas) are less suppressed than those of mesons (pions, kaons) [6,7] in the p T range 2 < p T < 5 GeV/c. The baryon content in the hadrons at intermediate p T depends strongly on the impact parameter (centrality) of the Au+Au collisions with about 40% of the hadrons being baryons in the minimum-bias collisions and 20% in very peripheral collisions [6,7]. Hydrodynamics [8,9], parton coalescence at hadronization [10,11,12] and gluon junctions [13] have been suggested as explanations for the observed particle-species dependence.
On the other hand, the hadron p T spectra have been observed to depend on the target atomic weight (A) and the produced particle species in lower energy p+A collisions [14,15,16]. This is known as the "Cronin Effect", a generic term for the experimentally observed broadening of the transverse momentum distributions at intermediate p T in p+A collisions as compared to those in p+p collisions [14,15,16,17,18]. The effect can be characterized as a dependence of the yield on the target atomic weight as A α . At energies of √ s ≃ 30 GeV, α depends on p T and is greater than unity at high p T [14,15], indicating an enhancement of the production cross section. The effect has been interpreted as partonic scatterings at the initial impact [17,18]. Thus, the Cronin effect is predicted to be larger in central d+Au collisions than in d+Au peripheral collisions [19]. At higher energies, multiple parton collisions are possible even in p+p collisions [20]. This combined with the hardening of the spectra with increasing beam energy would reduce the Cronin effect [18]. At sufficiently high beam energy, gluon saturation is expected to result in a relative suppression of hadron yield at high p T in both p+A and A+A collisions and in a substantial decrease and finally in the disappearance of the Cronin effect [21].
Recent results on inclusive hadron production from d+Au collisions indicate that hadron suppression at intermediate p T in Au+Au collisions is due to final-state effects [4,22,23]. The rapidity dependence of the particle yield at intermediate p T shows suppression in forward rapidity (deuteron side) and enhancement in the backward rapidity (Au side) in d+Au collisions at RHIC [24,25]. A study of particle composition will help understand the origin of the rapidity asymmetry [10]. In order to further understand the mechanisms responsible for the particle dependence of p T spectra in heavy ion collisions, and to separate the effects of initial and final partonic rescatter- * URL: www.star.bnl.gov ings, we measured the p T distributions of π ± , K ± , p and p from 200 GeV d+Au and p+p collisions. In this letter, we discuss the dependence of particle production on p T , collision energy, and target atomic weight.
The detector used for these studies was the Solenoidal Tracker at RHIC (STAR). The main tracking device is the Time Projection Chamber (TPC) which provides momentum information and particle identification for charged particles up to p T ∼ 1.1 GeV/c by measuring their ionization energy loss (dE/dx) [26]. Detailed descriptions of the TPC and d+Au run conditions have been presented in Ref. [22,26]. A prototype time-offlight detector (TOFr) based on multi-gap resistive plate chambers (MRPC) [27] was installed in STAR for the d+Au and p+p runs. It extends particle identification up to p T ∼ 3 GeV/c for p andp. In p+p and d+Au collisions, the dE/dx resolution from TPC was found to be better than 8% and there is 2 ∼ 3σ separation between the dE/dx of pions at relativistic rise and the dE/dx of kaons and protons at p T > ∼ 2 GeV/c [26]. By combining the particle identification capability of dE/dx from TPC and Time-of-Flight from TOFr, we are able to extend pion identification to ∼3 GeV/c [26,28]. MRPC technology was first developed by the CERN ALICE group [29] to provide a cost-effective solution for large-area time-offlight coverage.
TOFr covers π/30 in azimuth and −1 < η < 0 in pseudorapidity at a radius of ∼ 220 cm. It contains 28 MRPC modules which were partially instrumented during the 2003 run. Only particles from −0.5 < η < 0 are selected where most of the MRPC modules were instrumented. Each module [27] is a stack of resistive glass plates with six uniform gas gaps. High voltage is applied to electrodes on the outer surfaces of the outer plates. A charged particle traversing a module generates avalanches in the gas gaps which are read out by 6 copper pickup pads with pad dimensions of 31.5 × 63 mm 2 . The MRPC modules were operated at 14 kV with a mixture of 95% C 2 H 2 F 4 and 5% iso-butane at 1 atmosphere. In d+Au collisions, TOFr is situated in the outgoing Au beam direction which is assigned negative η. The average MRPC TOFr timing resolution alone for the ten modules used in this analysis was measured to be 85 ps for both d+Au and p+p collisions. The "start" timing was provided by two identical pseudo-vertex position detectors (pVPD), each 5.4 m away from the TPC center along the beamline [30]. Each pVPD consists of 3 detector elements and covers ∼ 19% of the total solid angle in 4.4 < |η| < 4.9 [30]. Due to the low multiplicity in d+Au and p+p collisions, the effective timing resolution of the pVPDs was 85 ps and 140 ps, respectively.
Since the acceptance of TOFr is small, a special trigger selected events with a valid pVPD coincidence and at least one TOFr hit. A total of 1.89 million and 1.08 million events were used for the analysis from TOFr triggered d+Au and non-singly diffractive (NSD) p+p collisions, representing an integrated luminosity of about 40 µb −1 and 30 nb −1 , respectively. The d+Au minimum- bias trigger required an equivalent energy deposition of about 15 GeV in the Zero Degree Calorimeter in the Au beam direction [22]. Minimum-bias p+p events were triggered by the coincidence of two beam-beam counters (BBC) covering 3.3 < |η| < 5.0 [1]. The NSD cross section was measured to be 30.0 ± 3.5 mb by a van der Meer scan and PYTHIA [31] simulation of the BBC acceptance [1]. A small multiplicity bias ( < ∼ 10% in d+Au and 18% in p+p) at mid-rapidity was observed in TOFr triggered events due to the further pVPD trigger requirement and was corrected for using minimum-bias data sets and PYTHIA [31] and HIJING [32] simulations. The effect of the trigger bias on the mid-rapidity particle spectra was found to be independent of particle p T at p T >0.3 GeV/c [33]. Centrality tagging of d+Au collisions was based on the charged particle multiplicity in −3.8 < η < −2.8, measured by the Forward Time Projection Chamber in the Au beam direction [22,34]. The TOFr triggered d+Au events were divided into three centralities: most central 20%, 20 − 40% and 40− ∼ 100% of the hadronic cross section. The average number of binary collisions N bin for each centrality class and for the combined minimum-bias event sample is derived from Glauber model calculations and listed in Table I.
The TPC and TOFr are two independent systems. In the analysis, hits from particles traversing the TPC were reconstructed as tracks with well defined geometry, momentum, and dE/dx [26]. The particle trajectory was then extended outward to the TOFr detector plane. Fig. 1 shows inversed velocity (1/β) from TOFr measurement as a function of momentum (p) calculated from TPC tracking in TOFr triggered d+Au collisions. The raw yields of π ± , K ± , p andp are obtained from Gaussian fits to the distributions in m 2 = p 2 (1/β 2 − 1) in each p T bin. For π ± at p T >1.8 GeV/c, an additional cut on dE/dx was applied at 50% efficiency [28]. The dE/dx distribution was measured by selecting on pure pion and proton samples from TOFr. The uncertainty of this cut was evaluated by systematically studying the yield as a function of the cut. Acceptance and efficiency were studied by Monte Carlo simulations and by matching TPC track and TOFr hits in real data. TPC tracking and MRPC hit matching efficiencies were both about 90%. Weak-decay feeddown (e.g. K 0 s → π + π − ) to pions is ∼ 12% at p T <1 GeV/c and ∼ 5% at higher p T , and was corrected for using PYTHIA [31] and HI-JING [32] simulations. Inclusive p andp production is presented without hyperon feeddown correction. p andp from hyperon decays have the same detection efficiency as primary p andp [35] and contribute about 20% to the inclusive p andp yield, as estimated from the simulation.
The invariant yields d 2 N/(2πp T dp T dy) of π ± , K ± , p andp from both NSD p+p and minimum-bias d+Au events are shown in Fig. 2. The average bin-to-bin systematic uncertainty was estimated to be of the order of 8%. The systematic uncertainty is dominated by the uncertainty in the detector response in Monte Carlo simulations (±7%). The normalization uncertainties in d+Au minimum-bias and p+p NSD collisions are 10% and 14%, respectively [1,22]. The charged pion yields are consistent with π 0 yields measured by the PHENIX collaboration in the overlapping p T range [2,23].
Nuclear effects on hadron production in d+Au collisions are measured through comparison to the p+p spectrum, scaled by the number of underlying nucleonnucleon inelastic collisons using the ratio where T dAu = N bin /σ pp inel describes the nuclear geometry, and d 2 σ pp inel /(2πp T dp T dy) for p+p inelastic collisions is derived from the measured p+p NSD cross section. The difference between NSD and inelastic differ- ential cross sections at mid-rapidity, as estimated from PYTHIA [31], is 5% at low p T and negligible at p T > 1.0 GeV/c. Fig. 3 shows R dAu of π + + π − , K + + K − and p +p for minimum-bias and central d+Au collisions. The systematic uncertainties on R dAu are of the order of 16%, dominated by the uncertainty in normalization. The R dAu of the same particle species are similar between minimum-bias and top 20% d+Au collisions. In both cases, the R dAu of protons rise faster than R dAu of pions and kaons. We observe that the spectra of π ± , K ± , p andp are considerably harder in d+Au than those in p+p collisions. The R dAu of the identified particles has characterstics of the Cronin effect [14,15,16,18] in particle production with R dAu less than unity at low p T and above unity at p T > ∼ 1.0 GeV/c. On the contrary, the R CP (nuclear modification factor between central and peripheral collisions) of identified particles in Au+Au collisions at √ s N N = 200 GeV as measured by PHENIX and STAR collaborations [6,7] do not have the above features. The R CP of p +p follows binary scaling and that of π 0 shows large suppression of meson production in central Au+Au collisions [7] as depicted in the bottom panel of Fig. 3. It is notable that the R dAu of proton and anti-proton are greater than unity in both central and minimum-bias d+Au collisions while the proton and antiproton production follows binary scaling in all centralities in Au+Au collisions [7].  [7]. The systematic uncertainties on these ratios were estimated to be of the order of 10% for p T < ∼ 1.0 from a Glauber model calculation, (p +p)/h averaged over the bins within 1.2 < pT < 2.0 GeV/c (left column) and within 2.0 < pT < 3.0 GeV/c (right column) and the R dAu ratios between p +p and h averaged over 1.2 < pT < 3.0 GeV/c for minimum-bias, centrality selected d+Au collisions and minimum-bias p+p collisions. A p+p inelastic cross section of σ inel = 42 mb was used in the calculation. For R dAu ratios, only statistical errors are shown and the systematic uncertainties are 0.03 for all centrality bins. GeV/c, decreasing to 3% at higher p T . At RHIC energies, the anti-particle to particle ratios approach unity (p/p = 0.81 ± 0.02 ± 0.04 in d+Au minimum-bias collisions) and their nuclear modification factors are similar. The (p +p)/h ratio from minimum-bias Au+Au collisions [7] at a similar energy is about a factor of 2 higher than that in d+Au and p+p collisions for p T > ∼ 2.0 GeV/c. This enhancement is most likely due to final-state effects in Au+Au collisions [5,8,9,11,12,13]. The ratios show little centrality dependence in d+Au collisions, as shown in Table I. The identified particle yields can also provide important information and constraints for other studies even when our measurements are in a limited rapidity range (-0.5< y <0.0). Our measurement of (p +p)/h ratio shows that baryons account for only about 20% of the total inclusive charged hadrons with little centrality dependence. Therefore, the measurement of rapidity asymmetry of inclusive charged hadrons around midrapidity by the STAR collaboration [24] is unlikely due to a change in particle composition or baryon stopping. For p T < 2.0 GeV/c, the (p +p)/h ratio in p +p collisions at √ s N N = 1.8 TeV [36] is very similar to those in d+Au and p+p collisions at √ s N N = 200 GeV. Also shown are p/h + ratios in p+p and p+W minimum-bias collisions at √ s N N = 23.8 GeV [14,15]. Although the relative yields of particles and anti-particles are very different, the Cronin effects are similar. At √ s < 40 GeV, there is a general trend of decreasing Cronin effect of all particles with beam energies at high p T [15,16], however, the Cronin effects ofp data are less conclusive [16].
The difference between R dAu at √ s N N = 200 GeV for p +p and h can be obtained from the (p +p)/h ratios in d+Au and p+p collisions. Table I shows R p+p dAu /R h dAu determined by averaging over the bins within 1.2 < p T < 3.0 GeV/c. Alternatively, we can study Cronin effect of the identified particles by comparing the α parameters of protons and pions. At lower energy, the α parameter in the power law dependence on target atomic weight A α of identified particle production falls with √ s [15,16] at high p T (p T ≃ 4.6 GeV/c). From the ratios of R dAu between p +p and π + + π − , we may fur-