Ricerca di segnali di supersimmetria in eventi con 2 leptoni di carica opposta, jet ed energia trasversa mancante con l’esperimento CMS di LHC

The subject of this thesis is the study of an important channel in the search for supersymmetry through the analysis of events with two opposite-sign leptons, jets and missing transverse energy. This analysis has been performed on the data collected by the CMS detector at a center of mass system energy (c.m.s.) √s = 7 TeV in the year 2011, corresponding to an integrated luminosity of about 5 fb^-1. Supersymmetry is one of the most favored extensions of the Standard Model (SM) of particle physics, that is the theory which summarizes our current knowledge on elementary particles and their mutual interactions. Through the last decades, it has received many experimental confirmations which have confirmed its validity at the electroweak scale with a high level of precision. A fundamental element of the theory is the existence of the Higgs boson, that is necessary for all particles to acquire a mass by means of the spontaneous symmetry breaking mechanism. Despite its enormous experimental success, some theoretical issues make the Standard Model not completely satisfactory. One of these issues, best known as the herarchy problem, concerns the mass of the Higgs boson. The Standard Model is self-consistent, as well as in good agreement with all the experimental results of the electroweak physics, if the Higgs boson mass is less than 1 TeV. However, the inclusion of radiative corrections raise the Higgs mass up to the Planck scale. A possible solution of this problem could be obtained within the SM by means of a fine tuning of the fundamental parameters, but this way appears unnatural. The aforementioned problems point to a more general theory, which includes the SM as an approximation at the electroweak scale. Supersymmetry is one of the most famous models. It introduces a new kind of symmetry, where each fermion has a bosonic supersymmetric partner, and vice versa. This symmetry would be exact if the supersymmetric partners had the same mass of the currently known particles. Such a possibility has been experimentally excluded, thus a symmetry breaking mechanism must be presumed, leading to a set of new particles not yet discovered. Searches in this field aim at identifying an excess of signals with respect to those predicted by the SM, to be then related to a supersymmetric scenario. To reach this goal, a collider is necessary with a high center of mass system (c.m.s.) energy with respect to the mass of the new particles (which is supposed to be of the order of TeV), as well as a high luminosity to compensate the low cross sections predicted for supersymmetric processes. The CERN Large Hadron Collider can provide both of these requirements. It accelerates to beams of protons whose collisions, in the design configuration, will reach a c.m.s. energy √s = 14 TeV at a luminosity of 10^34 cm^-2 s^-1. Collisions occur in four experimental points, where detectors are placed. Two of these, namely ATLAS and CMS, are general purpose experiments which, among their main goals, have the search for evidence of physics beyond the Standard Model (BSM). The detection of supersymmetric particles is possible through the identification of typical final states. In the minimal supersymmetric extension of the SM, the production and subsequent decay of supersymmetric particles are characterized by the presence of high pT jets, high missing transverse energy and possibly high pT leptons. A promising research channel is represented by the analysis of events having: high missing transverse energy; high jet activity; two opposite-sign leptons. The discovery of supersymmetry is possible if the candidate signals can be clearly distinguished from the Standard Model processes, which lead to final states similar to those of the searched events. Therefore, such physical background must be well rejected. In the analysis of data performed in this work, an opportune selection of events has been made, on the basis of criteria corresponding to the aforementioned typical features of signal events, in order to reject the background as much as possible. At a first stage the use of Monte Carlo simulations was necessary to reproduce background events with analogous criteria to those adopted for the selection of events in data; thus it is possible to check if data are in agreement or if they reveal any excess of events with respect to the expectation of Standard Model. Results show a reasonable agreement between data and Monte Carlo expectations, no significant excess arises from data. After this selection is performed, a significant fraction of background events can be rejected from data. However, the residual background which stands after selection cuts must be estimated in an appropriate way. Though being fundamental in this kind of studies, Monte Carlo simulations have some intrinsic limits, due to their dependence on the theoretical model used to generate background events. Such limits lead to investigate analysis methods which are as free as possible from Monte Carlo. These techniques are instead based on the use of real data, and are therefore called data-driven methods. The method used in this work is called “ABCD method”. Distributions of pairs of variables are considered which are highly uncorrelated between them and have a good separating power between signal and background. Opportune cuts allow to identify signal-dominated regions and background-dominated regions, as well as to control the residual background in the signal region. In addition to standard kinematic variables, new variables are introduced, whose discriminating power has been evaluated. The method proceeds through two phases. First, Monte Carlo simulation help to identify the variables pairs which allow to control the background in the signal region. After they have been identified, these pairs are used for the application of the method on real data. Results show that, for each of the chosen pairs, the number of events in the signal region is compatible with the background-only hypothesis, thus no significant excess is found in data. The ABCD method has revealed itself as a valid technique to control the background in the signal region.