Diffractive scattering processes have two main signatures: one or both incoming protons remain intact after the interaction and one or more rapidity gaps appear as forbidden regions in the rapidity distribution of scattering products. Rapidity gaps are associated to the exchange of pomerons between the interacting protons, the pomeron being described in QCD in terms of an exchange of gluons in a colorless configuration. In the context of diffractive physics, the study of rapidity gaps is then of particular interest. In the work reported in this thesis a study of rapidity gaps has been conducted in Double Pomeron Exchange events produced in pp collisions at √s = 8 TeV, using a dataset collected in July 2012 by the TOTEM experiment at the LHC during a common data taking with the CMS experiment. In DPE events both incoming protons remain intact in the collision and a system of particles is generated in the central zone, separated from the two protons by two rapidity gaps. Thanks to the combined TOTEM-CMS apparatus, which provides an exceptionally large pseudorapidity coverage, the tagging of protons with the TOTEM Roman Pot detectors and the reconstruction of the central system with the CMS apparatus has been performed. The TOTEM T2 telescope also provided the reconstruction of charged particles in the forward region, where no information from CMS is available. The aim of this work was the development of an analysis to study the rapidity gaps in DPE events, and compare them with a sample of DPE events obtained by a Pythia8MBR Monte Carlo simulation. Since such MC sample is based on a pure 2-gluon colorless exchange during the interaction, a deviation of rapidity gap probability could represent an indication of additional exchange not related to pomerons. In this study, an important role was covered by the charged particle tracks reconstructed in the TOTEM T2 telescopes and by final-state stable particles reconstructed and identified by means of a particular algorithm, known as “particle-flow" (PF), combining the information from the CMS subdetectors. They allowed to define in a wide |η| range the two rapidity gaps in DPE event candidates. The evaluation of the size of the rapidity gaps was possible through the direct leading proton measurement by the RP detectors. As first step, an optimization in the selection of PF neutral particles (neutral hadrons and photons) at √s = 8 TeV has been performed in order to suppress most of the detector noise in data, since standard cuts were previously obtained by the CMS Collaboration for data at √s = 7 TeV. The new cuts have been found in each CMS sub-region by using a Zero-Bias sample, collected during the same data taking period of the dataset used for DPE event selection (triggered by the TOTEM RPs). Then, the typical variables for the identification of DPE processes have been introduced: the fractional longitudinal momentum loss of each scattered proton (ξ1,2) reconstructed from the proton tracks; the two values of pseudorapidity ηmin and ηmax (related to the central diffractive system) which characterize the two rapidity gaps; the mass MX of the diffractive system obtained from the RP measurements, and the central mass Mcentral measured from the PF objects in the CMS region. In the next step of this study a selection of DPE event was performed in order to remove/reduce the main sources of background. The dominant background due to elastic events overlapped with pile-up processes has been removed by vetoing on the diagonal (TB/BT) configurations for the protons in the RPs. In order to reduce the pile-up effects in the selected parallel (TT/BB) configurations, we have selected only events with one proton per arm, simultaneously tagged by the two vertical RPs in the 220-station, and with no more than one CMS vertex. Then, by requiring a value of mass of the diffractive system greater than the value of the central mass, it was possible to reject events affected by residual noise and pile-up (NP) effects. After the selection of DPE events, a data driven correction method has been applied bin per bin to the probability distributions of ηmin and ηmax as an iterative procedure in order to correct for NP effects. This procedure has been based on studies on the Zero-Bias sample. In the following step, it was necessary to also consider the contribution of another source of background: the single diffractive process. Detailed studies performed on MC simulation showed that this contribution is dominant (up to about 19%) and is due when a proton produced by the breaking of one of two incoming protons arrives to RP detectors, simulating a leading proton. This contribution has been reduced by requiring activity in the CMS region. Based on MC studies, the iterative method correction has been updated in order to account for the residual SD background. Then, a comparison with MC expectations given by the Pythia8MBR generator was made. Here, a clear enhancement in events with reduced or absent reconstructed rapidity gaps in the very forward regions is observed in DATA, leading to a discrepancy in global normalization of the probability distributions in the central region, where a substantial agreement is found for the shapes. A better agreement between DATA and MC expectations is indeed found in the subsample where forward RGs are required (T2 veto on both sides). This result cannot a priori exclude that the violation of the expected rapidity gaps is due to some other process not characterized by the exchange of colorless objects, or that the MC generator we are considering is not properly modeling the DPE processes. However, further investigation should be performed in order to be sure that the observed behaviour is not due to some subtle residual background effect, or to some detector simulation related effect as well. For instance, it should be interesting to perform a separate study of events with zero and one CMS vertex, which are expected to be characterized by a different background bias.
Scheda prodotto non validato
Scheda prodotto in fase di analisi da parte dello staff di validazione