Measurement of the transverse momentum of Drell-Yan lepton pairs over a wide mass range in proton-proton collisions at $\sqrt s = 13\:\mathrm{TeV}$ in CMS

The study of the Drell-Yan process, or the production of a pair of leptons in high-energy hadron collisions through the exchange of a virtual Z or $\gamma^*$ boson, is a priviledged probe to understand the production mechanisms of heavy particles such as the Higgs boson. In this thesis, data collected in 2016 by the CMS experiment in proton-proton collisions at a center-of-mass energy of $13\,\mathrm{TeV}$, corresponding to an integrated luminosity of $36.3\,\mathrm{fb}^{-1}$, are used to perform a precision measurement of the the Drell-Yan cross section diferentially in the transverse momentum of the lepton pair, for invariant massses of the pair between 50 and 1000 GeV. The measurement uses channels comprising two electrons or two muons, which have the advantage of being easily identifiable experimentally and, thus, of allowing precise measurements of cross sections. In total, about ten million two-electron events and twenty million two-muon events are analyzed. The extension of the previous analysis of the CMS collaboration, limited in invariant mass to the interval between $76$ and $106\,\mathrm{GeV}$, requires a better control of the background noises, more important below $76$ and above $106\,\mathrm{GeV}$ than in the previously considered interval. The dominant contribution, the production of a pair of top quarks decaying into leptons, is studied in detail, as well as the contamination of the data by misidentified electrons. After taking into account the detection efficiency, the results obtained in the two channels are compatible and are combined in a single measurement. The accuracy obtained, on the order of one percent, in regions hitherto little explored, allows to obtain new constraints on models widely used in particle physics. To this end, the measurement is compared to six predictions illustrating different approaches for the prediction of the transverse momentum spectrum. In spite of quite good general performances, none of them allows a complete description of the data after taking into account the theoretical and experimental uncertainties. This is particularly the case at large invariant masses or when a jet is identified in the final state. Some approaches still not widely used obtain better results in the regions they specifically target. New theoretical developments will be necessary to combine these approaches in order to obtain reliable predictions in the whole phase space.