CERN Accelerating science

Relativistic heavy-ion collisions offer a way to study matter and fundamental forces under the most extreme temperatures, densities, and electromagnetic fields accessible in laboratories. Analyses of these collisions provide unique insights into the physics governing the creation of matter and the Quark Gluon Plasma (QGP), an exotic state of matter expected to have been prevalent in the Universe at $\sim 10^{-6}$ seconds after the Big Bang. Dielectrons (e$^{+}$e$^{-}$) are produced at each phase of the collisions. They reach the detectors without being significantly affected by the strongly interacting medium, which is also created in the collision. Therefore, they offer an ideal opportunity to examine heavy-ion collisions. Analyses of dielectrons provide experimental input to questions regarding thermal radiation of the QGP, the partonic structure of nucleons in nuclei, chiral-symmetry restoration, and the production of matter by electromagnetic fields. The main challenge in the analysis of dielectron production is the overwhelming background originating from photon-conversion processes in the detector material close to the interaction vertex. Tracks from these conversions lead to combinatorial background in the whole dielectron mass spectrum which exceeds the signal of dielectrons produced directly in the collision by up to three orders of magnitude. Numerous track observables can be used to infer the conversion origin of a track. This motivates a multivariate classification of conversion tracks using machine learning. Applying this method enhances the statistical significance of the measured dielectron signal in comparison to conventional cut-based background rejection. In the presented analysis dielectron production is studied differentially in the pair transverse momentum. This is motivated by the recently reported dielectron excess in Au--Au collisions at the Relativistic Heavy Ion Collider (RHIC). The prevalent explanation in theoretical studies is dielectron photo-production, a process in which a dielectron is produced from the electromagnetic field in the collision. However, the scarce experimental data related to this phenomenon is not completely in line with corresponding model calculations. It is predicted that this phenomenon is observable in Pb--Pb collisions with the ALICE detector at the Large Hadron Collider. In this work the corresponding dielectron spectra are extracted. An excess with respect to the expected hadronic sources is found. It agrees to a large extent, but not entirely, with expectations for photo-production. Discrepancies between data and models for dielectron photo-production at the RHIC led to speculations about so far not considered additional effects. These are related to rescattering of leptons in the QGP, deflections due to the extremely strong magnetic field, or effects of the finite nucleus size. All of these scenarios posit a dependence of dielectron photo-production on the direction relative to the reaction plane of the collision. In this thesis first preliminary results of a corresponding analysis are presented. A further project in this thesis deals with the search for magnetic monopoles. The absence of these particles represents rather an empirical finding than a necessity of the presently known laws of physics. In addition, Dirac provided appealing theoretical motivation for their existence. He showed that they would lead to understanding of another, theoretically unanticipated, empirical finding, namely that electric charges are quantised. Searches for monopoles were unsuccessful so far. Deriving robust lower limits for the magnetic monopole mass from these searches was not possible in the past. In searches for monopole production at colliders, setting mass limit requires the calculation of the expected monopole production rates. This is currently not possible as perturbative methods cannot be applied to this problem. It was recently found that this issue can in principle be circumvented by considering monopole production in strong magnetic fields. Since heavy-ion collisions are widely believed to produce the strongest magnetic field in the present Universe, they were proposed as a promising opportunity for searches, which could be interpreted in terms of rigorous mass limits. A corresponding strategy for a monopole search in heavy-ion collisions with ALICE is developed. Due to the expected exotic properties of monopoles, a key problem is to define the search target in the detector data. It is found that nuclear fragments which stop in the detector should yield a comparable detector response. They can therefore provide data-driven guidance for the definition of search criteria. The successful identification of corresponding particle tracks in detector data demonstrates the feasibility of the foreseen monopole search.