A study of nuclear effects using forward-rapidity hadron production and di-hadron angular correlations in sqrt(sNN) = 200 GeV d+Au and p+p collisions with the PHENIX detector at RHIC

Measurements using the PHENIX forward detectors at the Relativistic Heavy Ion Collider (RHIC) in high-energy deuteron-gold (d+Au) collisions enable us to study cold nuclear matter effects in nucleon structure at small parton-momentum fraction, or Bjorken-x. The large gluon densities in Lorentz-contracted gold nuclei enable us to search for the yet-unobserved saturation of the gluon distribution at small x, which is caused by a balance between gluon fusion and splitting. Gluon saturation is described by the Color Glass Condensate (CGC) theory [1], which predicts a suppression of inclusive particle production in heavy-ion collisions, in particular at forward rapidity, because of a decreased gluon density. In addition, it has been suggested that forward rapidity di-hadron correlations may elucidate CGC effects with two signatures that are specific predictions from CGC: awayside-yield suppression and angular broadening [2]. This thesis describes the first experimental measurements of these forward di-hadron correlations in PHENIX. Previously, RHIC experiments have shown a suppression in the single-particle nuclear modification factors (RdA, Rcp) for sqrt(sNN) = 200 GeV d+Au collisions in the forward (deuteron) direction [3, 4]. Multiple theories can explain the observed suppression (including CGC), but a conclusive measurement discriminating amongst the models has yet to be carried out. Two new forward-rapidity electromagnetic calorimeters (Muon Piston Calorimeters or MPCs, −3.7 < eta < −3.1, 3.1 < eta < 3.9) enable the PHENIX experiment to measure the single-particle nuclear modification factors in addition to further understanding forward particle production with the forward di-hadron correlation measurements. Azimuthal correlations of di-hadron pairs at different pseudorapidities allow us to scan the x-dependence of correlated di-hadron production, which can then be used to discriminate amongst the models that compete to explain the observed levels of forward particle production. More specifically, the x-dependence of the yields and widths of the correlated peaks can be measured, rigorously testing the theoretical models that attempt to explain the forward particle production. The forward-rapidity correlations are especially interesting because it is expected that they provide a test of gluon saturation down to x ~ 5 × 10−4 in the Au-nucleus. In this thesis, we present results based on the high integrated-luminosity data sample of proton-proton (p+p) and d+Au collisions at sqrt(sNN) = 200 GeV taken at RHIC in 2008. In order to produce the results in the relatively new MPC, a significant effort was concurrently invested into tasks related to the detector efficacy, including: improving the electromagnetic-shower reconstruction algorithm, devising and implementing the detector calibration, and creating the simulated particle-reconstruction efficiency and simulated particle-identification schemes. The relevant details of this work are shown followed by the physics analyses, which we summarize in what follows. We first present the PHENIX sqrt(sNN) = 200 GeV inclusive neutral pion RdA results as well as the p+p neutral pion differential cross section and the d+Au invariant yields for the pseudorapidity ranges of 3.0 < eta < 3.4, 3.4 < eta < 3.8, and 3.0 < eta < 3.8. We observe a similar trend in the suppression of RdA as first observed by the BRAHMS experiment in the forward direction wherein the suppression increases with decreasing collision impact parameter [3]. We also observe a larger suppression in the higher-rapidity bin (3.4 < eta < 3.8) as compared to the lower (3.0 < eta < 3.4). These results are compared with the nuclear-shadowing model of Qiu and Vitev [5], and the comparison shows that nuclear shadowing alone is unable to explain the observed level of suppression. We then proceed to show three sets of di-hadron correlation functions: two sets wherein a particle is at midrapidity (|eta| < 0.35) and the other is at forward rapidity (3.0 < eta < 3.8) in the MPC (termed mid-forward correlations), and another wherein both particles are detected at forward rapidity in the MPC (forward-forward correlations). For both the mid-forward and forward-forward correlations, we quantify the yields of the correlated awayside signal and form the di-hadron nuclear modification factor JdA, which is the correlated two-particle analogue of RdA. We again observe an increasing suppression with decreasing impact parameter, but in addition we see a very large suppression in JdA that reaches JdA ~ 0.1 for the forward-forward correlations; it is these correlations that are expected to be most sensitive to gluon-saturation effects [6]. To summarize the JdA data, the measured values are plotted versus our estimate of the parton momentum-fraction in the Au-nucleus, xAufrag. We observe an increasing level suppression with decreasing xAufrag, which would seem to support the predictions of CGC. However, predictions from other models, notably initial-state energy loss with nuclear shadowing, are necessary to eliminate other possible explanations. We also present simulation studies in the appendix that raise questions about the nature of particle production for the forward-forward correlations. In particular, the forward-forward correlated di-hadron signal in the PYTHIA p+p monte-carlo simulations does not seem to originate from di-jet production, but from some other momentum-conserving process. While PYTHIA admittedly does not correctly simulate the partonic interactions, this study still raises questions about the nature of di-jet production in this region