Dileptons are emitted throughout the entire space-time evolution of heavy ion collisions. Being colorless, these electromagnetic probes do not participate in the final-state strong interactions during the passage through the hot medium, and retain the information on the conditions of their creation. This characteristic renders them valuable tools for studying the properties of the Quark Gluon Plasma created during ultra-relativistic heavy ion collisions. The invariant mass spectra of dileptons contain a wealth of information on every stage of the evolution of heavy ion collisions. At low mass, dilepton spectra consist mainly of light meson decays. The medium modification of the light vector mesons gives insight on the chiral symmetry restoration in heavy ion collisions. At intermediate and high mass, there are significant contributions from charm and bottom, with a minor contribution from QGP thermal radiation. The region was utilized to measure cross sections of open charm and open bottom, as well as quarkonium suppression as demonstrated by PHENIX. An earlier PHENIX measurement of dielectron spectra in heavy ion collisions, using data taken in 2004, shows significant deviations from the hadronic decay expectations. The measurement, however, su↵ered from an unfavoriii able signal to background ratio. Random combination of electron-positron pairs from unrelated sources, mostly Dalitz decay of ⇡0 and external conversion of decay photon to electrons, is the main contributor to the background. Mis-identified hadrons are another major background source. To improve the situation, the Hadron Blind Detector (HBD), a windowless proximity focusing Cerenkov detector, is designed to reduce this background by identifying electron tracks from photon conversions and ⇡0 Dalitz decays. The detector has been installed and operated in PHENIX in 2009 and 2010, where reference p+p and Au+Au data sets were successfully taken. We will present the dielectron results from the analysis of the Au+Au collisions, and compare the measured mass spectra to theoretical expectations.

Since the 1980s the spin puzzle has been at the heart of many experimental measurements. The initial discovery that only ∼30% of the spin of the proton comes from quarks and anti-quarks has been refined and cross checked by several other deep inelastic scattering (DIS) and semi inclusive DIS (SIDIS) experiments. Through measurements of polarized parton distribution functions (PDFs) the individual contributions of the u, d, ¯u, ¯d, quarks have been measured. The flavor separation done in SIDIS experiments requires knowledge of fragmentation functions (FFs). However, due to the higher uncertainty of the anti-quark FFs compared to the quark FFs, the quark polarized PDFs (∆u(x), ∆d(x)) are significantly better constrained than the anti-quark distributions (∆¯u(x), ∆ ¯d(x)). By accessing the antiquarks directly through W boson production in polarized proton-proton collisions (u ¯d → W+ → e+/µ+ and du¯ → W− → e −/µ−), the large FF uncertainties are avoided and a cleaner measurement can be done. The parity violating single spin asymmetry of the W decay leptons can be directly related to the polarized PDFs of the anti-quarks. The W± → e± measurement has been performed with the PHENIX central arm detectors at √s = 510 GeV at the Relativistic Heavy Ion Collider (RHIC) and is presented in this thesis. Approximately 40 pb−1 of data from the 2011 and 2012 was analyzed and a large parity violating single spin asymmetry for W± has been measured. The combined data for 2011 and 2012 provide a single spin asymmetry for both charges: • W+: −0.27 ± 0.10(stat) ± 0.01(syst) • W−: 0.28 ± 0.16(stat) ± 0.02(syst) These results are consistent with the different theoretical predictions at the 1σ level. The increased statistical precision enabled and required a more careful analysis of the background contamination for the this measurement. A method based on Gaussian Processes for Regression has been employed to determine this background contribution. This thesis contains a detailed description of the analysis together with the asymmetry results and future prospects.

Polarized proton-proton collisions at RHIC are being used to study the origin of proton spin, which arises from the spin and orbital angular momentum of its constituent quarks and gluons. Measurements at the PHENIX experiment at √s = 200 GeV of Aπ0LL, the double longitudinal helicity asymmetry in neutral pion production, are used in global analyses of world polarized scattering data, where they are particularly important in constraining the sector of gluon polarization. These measurements have ruled out maximal gluonic spin contributions and are consistent with a small or zero contribution. In the latest measurements, the statistical precision of the data has reached the systematic limit, prompting investigation into the largest of the systematic uncertainties, the determination of relative luminosity. Details of the 2009 measurement at PHENIX of Aπ0LL and its inclusion in the global analysis will be presented along with recent studies on systematic uncertainties, including a 2012 study that varied the angles of the beams in the PHENIX interaction region.

Direct photons are produced during all stages of a heavy-ion collision. Due to their very small interaction cross section with the dense hadronic medium, they can escape the collision almost undisturbed and transport information about their production environment to a detector making them an excellent probe in heavy-ion physics. The observation of both a large yield and strong elliptical flow v2 of soft direct photons in heavy ion collisions at RHIC has sparked a lot of interest. While a large yield seems to point towards abundant production from the early, hot stages of the interaction, large elliptical flow can be better understood in a picture of predominately late production when the overall flow of the medium has built up. Telling different production scenarios for soft direct photons apart has been diffcult. We map out the centrality-dependence of direct photon observables and present results for dependence of the soft direct photon yield and flow as functions of centrality in the momentum range 0.4GeV/c < pT < 5.0GeV/c from a sample of externally converted photons. Here we exploit the good momentum resolution of our detector for charged particles at low momenta and reconstruct photons in electron-positron pairs from conversions in specific locations in the detector material. We find that the yield of soft direct photons has approximately a power-law dependence on the number of participants in the collision, and that their flow is en par with the flow of photons from hadron decays, indicative of relatively late production.

Relativistic heavy ion collisions have been a major research interest in the field of nuclear physics for the past few decades. Large collider facilities have been constructed to study the exotic matter produced in relativistic heavy ion collisions, one of which is the Relativistic Heavy Ion Collider (RHIC) at Brookhaven National Laboratory in Upton, NY. Essential to the study of heavy ion collisions are probes that are produced in the collision itself. Photons are a very useful probe of the collisions, since they escape the fireball virtually unmodified and carry with them information about the environment in which it was produced. Recent interest in low momentum direct photons has increased, due to the onset of the “thermal photon puzzle” and the apparent inability for typical models to explain both a large direct photon yield excess and large azimuthal production asymmetry (v2) at low momentum measured by PHENIX. The focus of this thesis will be the measurement of direct photons at low momentum with the PHENIX detector in √sNN = 200GeV Au+Au collisions. Low momentum direct photons (direct is any photon not from a hadron decay) are notoriously difficult to measure in a heavy ion environment, due to large decay photon backgrounds, neutral hadron contamination, and worsening calorimeter resolution. A novel technique for measuring direct photons via their external conversion to di-electron pairs has been developed. The method virtually eliminates the neutral hadron contamination due to the very clean photon identification based on di-electron pair invariant mass cuts. The direct photon fraction, Rγ, defined as the ratio of the yield of inclusive photons to hadron decay photons, is measured through a double ratio further reducing systematic uncertainties to manageable levels at low momentum. The direct photon fraction is converted to a direct photon invariant yield and a detailed look at the centrality dependence of the excess yield is presented. This dependence is confronted with recent theoretical calculations predicting novel production mechanisms of direct photons and possible solutions to the “thermal photon puzzle”.

The spin of the proton is known to be 1 2!. Although its angular momentum sum rule in terms of constituent quark and gluon components has been established, its detailed decomposition is poorly known. What fraction is attributed to the spin (polarization) and orbital angular momentum component is completely unknown, and how much of the spin component is from the quarks and gluons is only partially known. Dedicated experiments in the past few decades have measured the sum of quark and anti-quark spin contribution to account for only ∼25% of the proton spin, whereas separating the sea-quark polarizations or constraining the contribution of gluon polarization is still a subject of active experimental research. The Relativistic Heavy Ion Collider (RHIC) is a unique facility that provides collisions between polarized protons and thereby excellent tools to study the role of gluons in the proton intrinsic angular momentum. The double longitudinal asymmetry ALL of single inclusive production allows access to the polarized gluon distribution ∆g. It does so when the asymmetry measurements are incorporated into the so-called global analysis where polarized parton distribution functions and fragmentation functions are simultaneously fitted to best describe various measurements from different experiments. While π0 at PHENIX and jets at STAR have mainly been putting constrains on ∆G, the first moment of ∆g, other channels that provide complementary information on ∆G are critical. The high pT charged pion production is expected to be sensitive to the sign of ∆G. The isospin symmetry with other pion species will enables us to visually see the sign via the ordering of ALL of the three pion species even without performing global analysis. The interpretation can also be cross checked with the one drawn from global analysis, where the dominance of q−g scattering in π± production enhances the sensitivity. For this dissertation, high pT charged pion production at mid-rapidity in polarized p+p collisions at √s = 200 GeV has been analyzed. In this work, I developed a new analysis including the Hadron Blind Detector, a gas-based Cerenkov detector, to overcome the major challenge, a large fraction of electrons misidentified with π±, and achieved >98% purity in π± sample. Along with ALL, invariant differential cross section has been measured for different charges separately to validate the current perturbative Quantum Chromo-dynamics framework. Through these first successful measurements, we demonstrated π± is a promising channel to extract crucial information on ∆G in that complete discussions will be available with further constrained charge-separated fragmentation functions and improved statistics.

The PHENIX experiment at the Relativistic Heavy Ion Collider (RHIC) has measured charm and bottom quark production at midrapidity in p+p, d+Au, and Au+Au collisions at √ s = 200 GeV through their semi-leptonic decay into electrons. The large mass of the charm and bottom quarks means they are formed predominately by gluon-gluon fusion in the initial hard scatterings at RHIC and thus experience the full evolution of the medium, making them a good probe of medium effects. The yield in central Au+Au collisions is suppressed relative to p+p collisions, suggesting that the heavy quarks lose a significant portion of their initial energy in the medium. The d+Au results are enhanced relative to the p+p, pointing to cold nuclear matter effects that are masked by the hot medium in the Au+Au collisions. Studies of the intermediately sized Cu+Cu system provide a way to explore these competing effects as a function of system size and number of participating nucleons. In this dissertation, measurements of electrons from the decays of heavy quarks produced in Cu+Cu collisions are presented. We examine the interplay between hot and cold nuclear matter effects on open heavy flavor by comparing the results to those already measured in Au+Au and d+Au collisions. It has already been shown in the central Au+Au that partonic energy loss models are insufficient to describe the level of suppression. New models that include cold nuclear matter effects and the addition of meson dissociation are shown and compared to the Cu+Cu results.

A hot dense medium called the quark gluon plasma (QGP) has been created at the Relativistic Heavy Ion Collider (RHIC). Quarks and gluons are deconfined in the QGP state, but many of its properties are still under investigation. One interesting observation is that high momentum partons (quarks and gluons), which result from hard scatterings in the initial collision, lose energy as they travel through the medium. These partons fragment into the particles observed in the detector. Since fully reconstructing all the “jet” particles associated with the initial parton is complicated by the high multiplicity background produced in heavy-ion collisions, two particle correlations which trigger on a high momentum, pT , particle and measure the yield of associated particles in the event as a function of the azimuthal angle, ∆φ, are used instead. Di-hadron correlations are useful for observing suppression of the away-side (∆φ > π/2) jet yield and some features potentially due to the medium’s response to the lost energy. However, the hadron triggers, since they are fragments of a modified jet themselves, are biased to be near the surface of the medium and the jet energy is unknown. Since photons do not interact via the strong force, they are unmodified by the medium and provide an unbiased trigger. Direct photons result directly from the hard scattering. They balance the energy of the opposing parton and provide knowledge of the opposing jet momentum. Therefore, by measuring the hadron yield on the away-side, opposite the direct photon trigger, the jet fragmentation function, which describes how partons fragment into hadrons, can be measured as a function of zT = phT/pT γ. By comparing the spectra in Au+Au collisions to that in p+p collisions, the effective modifications of the fragmentation function can be quantified. Using the data collected by PHENIX during the 2007 RHIC Run, suppression of the away-side yield and the modified fragmentation function is measured via direct photon-hadron correlations. By including lower pT hadrons in the measurement, the altered shape of the modified fragmentation function is studied. Possible enhancement of the lowest zT particles suggests that the energy lost at high pT is redistributed to low pT particle production.

The Relativistic Heavy Ion Collider (RHIC) at BNL offers a unique opportunity in that it is capable of colliding protons and nuclei, including asymmetric collisions of different species. Open heavy quarks, that is charm or bottom not forming bound cc¯ or b ¯b pairs are important probes of the Quark Gluon Plasma at the Relativistic Heavy Ion Collider at BNL. They are formed at the initial collision of the nuclei and thus any effect to their transverse momentum spectra or azimuthal distribution can only come from their interaction with the matter created in the collision. One of the most powerful techniques of measuring these effects is to divide AuAu data by appropriately scaled pp data. This work focuses on providing the best possible pp reference both in scope and precision. Transverse momentum (pT ) spectra of electrons from semileptonic weak decays of heavy flavor mesons in the range of 0.3 < pT < 15.0GeV/c have been measured at mid-rapidity (|y| < 0.35) beyond the previous published range of pT < 9.0GeV/c. This is done using a new technique exploiting the observed characteristics of energy deposition in the PHENIX electromagnetic calorimeters. We present this technique as well as the final measurement compared to FONLL theory predictions of open charm and bottom cross section.

The dielectron mass spectrum consists of light vector meson decays, correlated heavy quark contributions and decays from other hadronic and photonic sources. Thermal radiation and modifications to the light vector mesons may provide additional signals at low masses above known hadronic sources. The PHENIX √sNN = 200 GeV Au+Au and Cu+Cu analyses have measured a centrality dependent excess in the the low mass region between 0.15 GeV/c2 and 0.75 GeV/c2. Between the φ and the J/ψ, in the intermediate mass range, the correlated heavy quark decays are the primary dielectron source; direct photons may augment this region as well. The Cu+Cu system is sensitive to the onset of the dielectron excess. By studying the Cu+Cu mass spectra and yields as a function of pair pT and collision centrality we obtain further understanding of its behavior. Comparisons to the PHENIX Au+Au and p+p measurements an extrapolations from theory are presented.

A hot, dense QCD medium is created in heavy ion collisions at the Relativistic Heavy Ion Collider at Brookhaven National Laboratory. This new type of matter is opaque to energetic partons, which suffer a strong energy loss in the medium. Two particle correlations are a powerful tool to study the jet properties in the medium and provide information about the energy loss mechanism and jet-medium interactions. When triggering on high pT particles, the away-side shape depends strongly on the pT of the associated particles. In this analysis, we present the inclusive photon-hadron two particle azimuthal correlations measured in Au+Au collisions at √sNN = 200 GeV by PHENIX experiment. In order to study jet-medium interactions, we focus on intermediate pT , and subtract particle pairs from the underlying event. Jet-like correlations appear modified in central Au+Au compared to p+p, in both the trigger and opposing jet. The trigger jet is elongated in pseudorapidity (the “ridge”), while the opposing jet shows a double peak structure (”head” and “shoulder”). We decompose the structures in ∆η and ∆φ to disentangle contributions from the medium and the punch-through and trigger jets. Upon correcting the underlying event for elliptic flow, the ridge is observed for associated particle pT below 3 GeV/c; it is broad in rapidity and narrow in ∆φ. The away side correlated particle yield is enhanced in central collisions. The yield of particles in the shoulder grows with centrality while the away side punch-through jet is suppressed. Remarkably, the ridge closely resembles the shoulder in the centrality dependence of particle yield and spectra. There has been great debate about the origin of the ridge and shoulder. A favored explanation is that the structure is due to features of the collective flow of particles in the underlying event, particularly the fluctuation-driven triangular flow, quantified by the third Fourier component, v3. We measure higher order Fourier harmonics in two ways, and use the results to give the shape of particle correlations in the underlying event. We decompose the power spectrum for the medium and for jets measured in p+p collisions. When including the higher harmonics of the collective flow (v3, v4) in the shape of the underlying events in two particle correlations, the ridge and shoulder no longer exist after subtraction. The jet function in Au+Au looks like p+p in which the away side jet is suppressed and broadened. There is also a pedestal-like structure in the jet function. Since the higher harmonics only change the shape of the underlying background, the pedestal is simply the redistribution of the ridge and shoulder particle yields. In conclusion, when jets pass through the medium, the away side jet is suppressed and the shape is broadened. This also brings out extra particles with spectra slightly harder than the medium, but softer than jet fragments. These are probably from the jetmedium interaction.

The experimental collaborations at the Relativistic Heavy Ion Collider (RHIC) have established that dense nuclear matter with partonic degrees of freedom is formed in collisions of heavy nuclei at √sNN = 200 GeV. Information from heavy quarks has given significant insight into the dynamics of this matter. Charm and bottom quarks are dominantly produced by gluon fusion in the early stages of the collision, and thus experience the complete evolution of the medium. The production baseline measured in p + p collisions can be described by fixed order plus next to leading log perturbative QCD calculations within uncertainties. In central Au+Au collisions, suppression has been measured relative to the yield in p + p scaled by the number of nucleon-nucleon collisions, indicating a significant energy loss by heavy quarks in the medium. The large elliptic flow amplitude v2 provides evidence that the heavy quarks flow along with the lighter partons. The suppression and elliptic flow of these quarks are in qualitative agreement with calculations based on Langevin transport models that imply a viscosity to entropy density ratio close to the conjectured quantum lower bound of 1/4π. However, a full understanding of these phenomena requires measurements of cold nuclear matter (CNM) effects, which should be present in Au+Au collisions but are difficult to distinguish experimentally from effects due to interactions with the medium. This thesis presents measurements of electrons at midrapidity from the decays of heavy quarks produced in d+Au collisions at RHIC. A significant enhancement of these electrons is seen at a transverse momentum below 5 GeV/c, indicating strong CNM effects on charm quarks that are not present for lighter quarks. A simple model of CNM effects in Au+Au collisions suggests that the level of suppression in the hot nuclear medium is comparable for all quark flavors.

The Relativistic Heavy Ion Collider (RHIC) was built to produce and study Quark Gluon Plasma (QGP), the phase of matter thought to exist under conditions sufficiently hot and dense to create a medium in which the degrees of freedom are quarks and gluons rather than color neutral hadrons. Already in its early years of running, the data from RHIC provided tantalizing evidence of QGP signatures in Au+Au collisions at √sNN = 200 GeV. A crucial part of understanding the putative QGP in Au+Au collisions is to have both a well understood reference as well as a robust control experiment. Proton-proton collisions at the same √sNN serve as the baseline for heavy ion collisions at RHIC, and play an invaluable role in setting our frame of reference in interactions that do not create any nuclear medium. For the control experiment, RHIC’s ability to collide asymmetric beams is utilized and d+Au collisions are used. Unlike p+p collisions, in the d+Au system there is a nuclear medium present - the heavy Au nucleus - and so we may study this system to distinguish initial state cold nuclear matter effects from final state effects that occur in the hot dense medium of Au+Au collisions. Beyond its use as a control experiment, the d+Au collision system presents the opportunity for important study of nuclear and nucleonic structure, it is after all necessary for our colored parton theory to operate in the nucleus as well as in a QGP. Deuteron - gold collisions at RHIC are a powerful tool for shedding light on cold nuclear matter effects. This thesis describes two analyses of d+Au collisions measured by the PHENIX experiment at RHIC. The first is a measurement of the midrapidity yield of unidentified charged hadrons in the 2003 RHIC run. This is used a key baseline for understanding particle production in Au+Au collisions as well as a detailed look at the Cronin effect. The second analysis measures rapidity separated two-particle production where one of the particles is at either forward or backward rapidity and the other at midrapidity. These measurements probe different x regions of the Au nucleus and there investigate shadowing, anti-shadowing and other cold nuclear matter effects.

Deconfinement of color charge in nuclear matter at high energy density is a topic of considerable theoretical interest and experimental effort. Predicted in QCD, a new phase of deconfined matter, the quark gluon plasma, is thought to describe a transitional period of the early universe following the Big Bang. The extremely high energy density medium created in relativistic collisions of large nuclei at RHIC afford an opportunity to study the properties of quark gluon plasma in a laboratory setting. Fast partons (quarks and gluons) transiting the produced medium have been observed to experience a large energy loss. Correlations between pairs of final state particles at high transverse momenta (pT ! 4 GeV/c) map the hadron jets resulting from these partons and show that partons crossing the medium are nearly fully absorbed. The mechanism of energy loss on length scales comparable to the nucleus is not fully understood, so more differential measurements are needed to constrain theoretical models. Quenching as a function of the path length through the medium adds a new dimension of experimental discrimination on energy loss and initial state geometry. The resulting away-side suppression patterns indicate that surviving fast partons cross the nuclear overlap region with little energy loss. The transiting partons deposit energy locally in the medium. The resulting medium excitations may lead to measurable signals related to the medium properties. Pair correlations at low pT (" 4 GeV/c) can reflect the medium response. Comparison of correlations in heavy ion collisions with baseline measurements in protonproton collisions show modifications to the correlation shape and yields. Two new structures are found, both extended in rapidity, one centered at small azimuthal opening angle ∆φ (known as the “ridge”) and the other occurring at ∆φ = π ± 1.1 rad (“shoulder”). Comparisons between the two raise the possibility that both phenomena may result from the same mechanism. The medium response correlations are consistent with collective excitation theory, but pose challenges to Cherenkov gluon radiation and deflected jet models.