The Standard Model (SM) of particle physics provides the theoretical framework to describe the fundamental interactions among elementary constituents of matter. SM is supported by experiments to a high degree of accuracy, up to parts per-mil for the electroweak (EW) sector and parts-per-trillion for QED alone, but it still remains incomplete. Many observed phenomena lack explanation in the framework of the SM and its particles. They indicate the possibility of existence of particles and interactions beyond the SM (BSM). These phenomena include dark matter (DM), dark energy and baryonic asymmetry of the universe. In addition, a quantum description of gravity is still lacking. The top quark has the largest mass among the SM particles. Due to it’s heavy mass, top quark is the only colored particle which does not hadronize and hence its properties are directly accessible by studying it’s decay particles. The order one Yukawa coupling of the top quark also imbibes it with an important role in the behavior of the SM couplings at higher energy scales where possible BSM physics may contribute. As a result, precision measurements of top quark properties may provide a glimpse into BSM physics and hence making these measurements is one of the core aims of the Large Hadron Collider. In stark contrast with top quark physics is the elusive, dark matter (DM) of the universe. There exists a lot of observational evidence for it but, as of yet, with no clue with regards to its particle properties and interactions. Compelling evidence for the existence of DM comes from measurements based on cosmic microwave background radiation, astrophysical observations of distribution of visible matter in galaxy clusters, galactic cluster collisions (e.g. bullet cluster), gravitational lensing, galactic rotation curves, structure formation simulations, to name a few. It is interesting to investigate the possibility that there may be a connection between top quark and DM. In this thesis, we extend the SM with simplified models to study BSM physics at colliders and also to explain the DM puzzle. Here, we use the Top quark as a laboratory for constructing generic probes of BSM and also of the dark sector physics. In Chapter 1, we introduce some relevant background and salient aspects of the SM framework on which the following BSM theories are built. In Chapter 2 we explore an s channel and a t-channel simplified model in the context of top quark pair production using asymmetries constructed with kinematic variables of the top decay products. In Chapter 3, we then propose a simplified model which includes a colored scalar as the mediator between DM and SM particles, termed gluphillic scalar dark matter (GSDM). Monojet process is one of the primary channels to probe DM at hadron colliders. In Chapter 3, the discussion of monojet process at the Large Hadron Collider (LHC) is limited to the effective field theory (EFT) framework. In Chapter 4 we discuss collider processes in GSDM model with complete loop calculations for the diagrams involving the mediating colored scalar. We also compare the loop calculation with the EFT results to find the range of applicability of the EFT. The top quark study in Chapter 2 was initially inspired from an interesting observation made in 2008 which suggested a deviation from the SM in the forward-backward asymmetry (FBA) of a pair produced top quark. The value of FBA measured at the time was 18% ±12%. This value deviated by more than 1σ with respect to the SM leading order (LO) value of 5%. The deviation was observed by both the detectors at Tevatron, D0 and CDF, and it’s significance increased with additional data in 2012. Recent analyses of the data by D0 is now in better agreement with the latest effective-NNNLO calculations. However, the FBA measurements by CDF are still in tension with those by D0 and the value predicted by theoretical calculations. Inspired by this puzzle, which may be on its way to getting solved, we have been able to construct effective probes of BSM physics for the on-going and future searches of BSM in the top quark sector. In our analyses, we studied correlations among observables which can distinguish between different sources of BSM contributions in the top quark pair production. As a template, we use an s-channel and a t-channel mediator, both of which leave very different signatures in the kinematic asymmetry correlations. The simplified models considered by us also included parity breaking interactions which lead to polarized top quarks, providing another probe into the underlying production process. We find that all the kinematic distributions of the decay lepton get influenced by the polarization of the top quark. We show that these correlations can distinguish well between the template models of axigluon and diquark. In general, all of these observables also provide a probe into the polarization of the top quark and therefore any chiral couplings with the mediator. However, the lepton polar angle asymmetry measured in the lab frame is special in that it can not only probe the longitudinal polarization as other observables but is also sensitive to the transverse polarization of the top quark. We also show the effectiveness of the proposed top quark kinematic observables, to distinguish between s and t-channel BSM physics models, in future searches for BSM particles at the run-II LHC. In a large verity of dark matter (DM) models the simplest candidate is the model of a singlet scalar particle. The scalar may couple to the standard model in a number of ways via any of the SM particles. Such models with BSM Yukawa interactions or gauge sector extensions are strongly constrained from both the direct detection and collider precision measurements. The remaining models either predict a very heavy dark matter, completely out of reach of collider searches or introduce an unnaturally weak coupling with the SM particles giving no justifications for the small numbers. An interesting corner of the space of possible DM models which has been under-explored so far includes interactions of DM particles with gluons. Although DM particles cannot themselves be charged or colored, a colored scalar mediator can allow this interaction. One such model arises when we consider the scalar DM in presence of a colored scalar particle, for example the one from t-channel model above. Such colored scalars are generically present in a number of BSM theories including SUSY and GUT. How-ever, without the need for any additional gauge symmetries, the two scalars would interact with each other via the marginal operators. In Chapter 3 we study a SM singlet scalar DM candidate which couples to SM via a colored scalar particle. In the GSDM model, DM and mediator interact via the quartic, marginal operator. DM annihilation cross-section of the order of weak interactions (∼ 0.1pb) is predicted to explain the observed dark matter relic density if arising from thermal production of a WIMP DM candidate of mass ∼ 100 GeV. On investigating the GSDM model, we find that it allows a large annihilation cross-section and is still compatible with direct detection bounds. This is so because the annihilation cross-section to a pair of colored scalars proceeds via a tree-level interaction, whereas the interaction with SM particles proceeds via loop diagrams involving the colored scalars. Our work shows that this model is compatible with the observed relic density of DM when the mediating particle is lighter than DM for a large range of the couplings. For masses of the DM and the mediator less then ∼ 50 GeV, the DM can also be lighter than the mediator where the annihilation then proceeds via loop interactions. This region of parameter space is strongly constrained from the collider physics bounds on a colored scalar particle. These bounds become much weaker in the case where the colored scalar does not couple to quarks and hence cannot decay. The bounds coming from long-lived colored scalars become relevant in those cases and also constrain the light mass window. A colored scalar interacting with quarks must do so without violating the strong flavor constraints. We consider the scalar in the framework of a class of models termed minimally flavor violating (MFV) and also assume that it couples only to the right handed up-sector quarks. Such a particle would couple to the top quark and would be observable at the LHC pair production of the top quark. We find constraints on a color triplet particle in such a case and show the coupling and mass regions allowed. Constraints from the decays to light quarks are interpreted from dijet process searches and limit the mass of a color-triplet scalar above 350 GeV. The primary process for direct search of stable particles produced at a collider is a single jet in association with missing transverse energy (MET). We find that in an effective field theory (EFT) framework, very weak bounds are obtained on the mediating scale. In Chapter 4, we perform complete loop calculations for processes involving colored scalar particles and DM at LHC in order to explore the GSDM model at LHC and FCC (Future Circular Collider). The EFT is valid only for mediator masses much heavier than the momentum transfer or the MET cuts. We show the region of applicability of the EFT by comparing it with respect to the loop induced calculation. We analyze the monojet + missing transverse energy (MET) process to find the expected bounds from LHC 13 TeV run-II. We calculate the reach of the LHC in the high luminosity run in the future and also the reach of the FCC to explore the GSDM model. We perform all our calculations for a number of representations of the colored mediator from a triplet to dimension 15. As expected, collider constraints are only significant when the dark matter is light enough (mDM ∼ 10 GeV) to be copiously produced at the LHC. We find that in the high luminosity run, LHC can probe the scalar triplet particle up-to 50 GeV mass in the monojet process though a dimension 15 particle can be probed up to 150 GeV. With an order of magnitude higher beam energy, FCC can rule out much larger parameter space or provide observational evidence for TeV scale mediating particles. In conclusion, this thesis adds to the growing body of literature which points towards BSM discoveries around the corner at high luminosity LHC in the top physics and in dark sector physics. We have also proposed avenues for precision BSM studies at the next generation colliders.

L'International Linear Collider (ILC) est un projet de futur collisionneur électron positron opérant à une énergie nominale dans le centre de masse de 500 GeV. Il fera des mesures de précision, par exemple d'un boson de Higgs léger qui pourrait être bientôt découvert au Large Hadron Collider. Les détecteurs de l'ILC prévus à cet effet seront composés de calorimètres à haute granularité.Cette thèse présente l'étude de la réponse d'un calorimètre électromagnétique silicium tungstène hautement granulaire (ECAL SiW) ainsi que l'étude de la production du quark top à l'ILC.Le prototype d'un ECAL SiW développé par la collaboration CALICE a été testé sous faisceaux de particules chargées au FNAL en Mai et Juillet 2008. Après avoir sélectionné des pions chargés négativement et entrant dans le ECAL, sa haute granularité est mise à profit pour introduire une classification en quatre types d'événements, afin de décrire des interactions hadroniques.Des modèles de dimensions supplémentaires expliquent l'anomalie AFBb du LEP par une modification des couplages des quarks de troisième génération au boson Z. Ces effets motivent l'étude de la désintégration semileptonique des paires de quarks top, effectuée ici au moyen d'une simulation complète du détecteur ILD proposé pour l'ILC à une énergie dans le centre de masse de 500GeV pour une luminosité intégrée L = 500 fb-1. Les performances de ce détecteur permettent d'atteindre des efficacités de sélection de plus de 70% avec une pureté meilleure que 95%. Cela se traduit par une précision relative d'environ 1% sur l’asymétrie gauche-droite de la production du quark top ALR ainsi que sur l'asymétrie avant-arrière du quark top AFBt dans le cas où les électrons sont polarisés à 80% et pas les positrons. Les incertitudes relatives sur les couplages gauche et droit du quark top au boson Z peuvent aller jusqu'à 0.9% et 1.5%. / The International Linear Collider (ILC) is a proposed e+e- collider with a center-of-mass energy of 500 GeV or more, aimed at precision measurements, e.g. of a light Higgs boson that could be discovered soon at the Large Hadron Collider. Its detectors foresee the use of fine grained calorimeters to achieve the desired precisions.This thesis presents the study of the response to hadrons of a highly granular silicon-tungsten electromagnetic calorimeter (SiW ECAL), and the study of top quark pair production at the ILC.The SiW ECAL prototype developed by the CALICE collaboration was tested with beams of charged particles at FNAL in May and July 2008. After selecting single negatively charged pions entering the ECAL, its fine granularity is used to introduce a classification among four types of events, used to describe hadronic interactions.Motivated by extra-dimensional models which may explain the AFBb LEP anomaly by modifying the couplings of third generation quarks to the Z boson, the semileptonic decay of the top quark is studied with a full simulation of the proposed ILD detector for the ILC at center-of-mass energy of 500 GeV and integrated luminosity of 500 fb-1. Detector performances permit to reach efficiencies larger than 70% in finding those events with purity larger than 95%. This translates into a relative accuracy of about 1% on both the left-right asymmetry of top production ALR and the top forward-backward asymmetry AFBt with electrons polarized at 80% and no polarization of the positrons. The relative uncertainties in the left and right couplings of the top quark to the Z boson could be as good as 0.9% and 1.5%.

This thesis describes searches for new heavy resonances that decay into dielectron final state and searches for new physics in the top quark sector. The standard model of elementary particle is introduced in the first chapter. After that, a selection of theories beyond the standard model that predict the existence of new massive resonances are described together with an introduction to the effective field theory that is used to search for new physics in top quark sector. Then, the Large Hadron Collider (LHC) and the Compact Muon Solenoid (CMS) detector are introduced, and the techniques used in order to reconstruct the particles produced in the collisions are discussed afterwards. Finally, two separate analyses are presented.The first analysis is searching for new heavy resonances using dielectron final state. As some beyond Standard Model theories predict the existence of new heavy resonances that can decay into dielectron pair, such as the grand unified theories and theories that introduce extra space-like dimensions. An observation of a local “bump” in the dielectron invariant mass spectrum will be an evidence for the existence of a new heavy resonance. The data used is from CMS experiment collected in 2016 with 35.9 fb-1 and in 2017 with 41.4 fb-1. The event selection is optimized in order to be highly efficiency for high energy electron and avoid loosing potential signal events. The leading background is the Drell-Yan process and it is estimated from simulation. The sub-leading background is from ttbar and ttbar-like processes and it is estimated from simulation also. A data-driven method is used to validate the simulation of sub-leading background. The last background from quantum chromodynamics processes is determined by data-driven approach. After having inspected the final dielectron invariant mass spectrum, no significant excess over the standard model background is observed, and upper limit at 95% confidence level is set on the ratio of production cross-section times branching ratio of a new resonance to the one at the Z boson peak.The second analysis is the search for new physics in the top quark sector with dielectron and dimuon final states using data collected by the CMS experiment in 2016 with 35.9 fb-1. Because of its high mass and close to electroweak symmetry breaking scale, the top quark is expected to play an important role in several new physics scenarios. The new physics in top quark pair production and in single top quark production in association with a W boson are investigated and a dedicated multivariate analysis is used to separate these two processes. No significant deviation from the standard model expectation is observed. Results are interpreted in the framework of an effective field theory and constraints on the relevant effective couplings are set at 95% confidence level. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

<p>This thesis presents two measurements of the top quark pair production cross section at sqrt{s} = 1.96 TeV using data from the DØ experiment. Both measurements are performed in the dilepton final state and make use of secondary vertex b-tagging. With 158 pb^-1 of data in the electron-muon final state, the measured cross section is:</p><p>σ(top-antitop) = 11.1 +5.8 -4.3 (stat) +- 1.4 (syst) +- 0.7 (lumi) pb.</p><p>With 425 pb^-1 of data in the electron+track and muon+track final states, the measured cross section is:</p><p>sigma(top-antitop) = 6.3 +2.1 -1.8 (stat) +- 1.1 (syst) +- 0.4 (lumi) pb.</p><p>Both measurements are in agreement with the prediction from perturbative QCD calculations. In addition, an estimate of the DØ silicon detector lifetime is presented. The radiation damage is determined by studying the depletion voltage of the silicon sensors as a function of time. Based on this data the silicon detector is estimated to remain operational up to delivered luminosities of 6-8 fb^-1.</p>

This thesis presents two measurements of the top quark pair production cross section at sqrt{s} = 1.96 TeV using data from the DØ experiment. Both measurements are performed in the dilepton final state and make use of secondary vertex b-tagging. With 158 pb-1 of data in the electron-muon final state, the measured cross section is: σ(top-antitop) = 11.1 +5.8 -4.3 (stat) +- 1.4 (syst) +- 0.7 (lumi) pb. With 425 pb-1 of data in the electron+track and muon+track final states, the measured cross section is: sigma(top-antitop) = 6.3 +2.1 -1.8 (stat) +- 1.1 (syst) +- 0.4 (lumi) pb. Both measurements are in agreement with the prediction from perturbative QCD calculations. In addition, an estimate of the DØ silicon detector lifetime is presented. The radiation damage is determined by studying the depletion voltage of the silicon sensors as a function of time. Based on this data the silicon detector is estimated to remain operational up to delivered luminosities of 6-8 fb-1.

Le détecteur ATLAS est un expérience généraliste placé auprès du collisionneur proton-proton circulaire de 27 Km de circonférence, LHC. Le LHC est conçu pour produire des collisions avec une énergie nominale au centre de masse de l'ordre de 14 TeV et une luminosité instantané de 1034cm-2s-1.Il donne accès à des processus à l'échelle du TeV. En 2010 et 2011 le LHC a fournit des collisions proton-proton avec une énergie dans le centre de masse de 7 TeV donnant la possibilité d'enregistrer plus de 5 fb-1 de données. En 2012 le LHC a ensuite fournit des collisions avec une énergie dans le centre de masse de 8 TeV. La présente thèse traite de la mesure de la section efficace de production des paires de quarks top anti-top dans le mode de désintégration complètement hadronique utilisant les données enregistrées par l'expérience ATLAS en 2011 avec une énergie dans le centre de masse de 7 TeV correspondant à une luminosité intégrée de 4.7 fb-1. Le canal complètement hadronique a l'avantage d'être caractérisé par un rapport d'embranchement de 46%. Il souffre par contre d'un bruit de fond multi-jet QCD élevé. Le principal bruit de fond pour la production des paires des quark top-antitop est dû aux processus QCD avec au mois six de quarks/gluons dans l'état final. Ce bruit de fond n'est malheureusement pas bien connu et donc difficile à reproduire par simulation Monté-Carlo. Il est estimé à partir des données. La section efficace ainsi mesuré est 168±12 (stat.) +60/-57(syst.) ± 7 (lum.) pb. La deuxième analyse présentée dans la cette thèse se focalise sur la recherche du boson du Higgs produit en association avec une paire de quark top, où le boson de Higgs se désintégré en paires de quarks b, et le système top-anti-top se désintégrant dans le canal complètement hadronique. L'analyse est effectuée sur les mémés données utilisées par la mesure de la section efficace top-anti top. Deux bruits de fond principaux peuvent être distingués: la production des paires de quarks top avec la présence des jets supplémentaires et la production QCD. Le premier bruit de fond est estimé en utilisant le modelé Monté-Carlo différenciant la production de jets légers des jets lourds. Le deuxième bruit de fond est estimé avec les données. Le but de cette étude est de montrer le potentiel du canal complètement hadronique en regardant le signal sur la racine du bruit du fond dans les régions de signal et en faisant une comparaison avec les résultats publics présentés par l'ATLAS dans le canal semileptonic.

La masse du quark top est un paramètre fondamental du Modèle Standard de la physique des particules. L'un des défis du programme de physique du LHC, où le quark top est produit en abondance, est la détermination précise de ce paramètre. Actuellement, il est mesuré via la reconstruction directe de tous les produits de désintégration du quark top. L'incertitude systématique sur la calibration en énergie des jets et sur l'étiquetage de quarks b limite grandement la précision de cette mesure et il est donc indispensable de développer des mesures alternatives. Une possibilité très prometteuse – jusque-là jamais exploitée – est l'étude de désintégrations de paires de quarks top, où l'un des deux quarks se désintègre dans le canal leptonique avec un J/psi créé lors de la fragmentation du quark b. La masse du quark top est extraite à travers sa corrélation à la masse invariante de la combinaison du J/psi et du lepton chargé. Cette méthode présente l'avantage de n'utiliser que des leptons chargés, dont l'énergie-impulsion est mesurée avec une grande précision dans le détecteur CMS / The top quark mass is a fundamental parameter of the Standard Model of Particle Physics. One of the challenges in the LHC physics case, where the top quark is produced in abundance, is the precise determination of this parameter. Currently, it is measured through the direct reconstruction of all the top quark decay products. The systematic uncertainty on the jet energy calibration and b tagging sensibly limits the precision on this measurement and it is thus vital to develop alternative measurements. A very promising possibility – never exploited until now – is the study of top pair decays, where one of the two quarks decays in the leptonic channel with a J/psi created in the b quark fragmentation. The top quark mass is extracted through its correlation to the invariant mass of the J/psi-charged lepton combination. This method presents the advantages of using charged leptons only, whose energy momentum is measured with great accuracy in the CMS detector

Cette thèse présente des études pour l’International Linear Collider (ILC), un collisionneur électron-positron linéaire avec une énergie nominale dans le centre de masse de 250 GeV à 500 GeV. Les données analysées ont été enregistrées avec le prototype physique CALICE d’un calorimètre électromagnétique silice-tungstène (Si-W ECAL) à FermiLab en 2008. Au cours de cette thèse, un algorithme de recherche de traces a été développé, qui trouve des traces secondaires dans les événements hadroniques enregistrés par le prototype Si-W ECAL. Cet algorithme révèle des détails sur les interactions hadroniques dans le volume du détecteur et les résultats sont comparés avec des simulations basées sur le GEANT4 toolkit. Les recherches indirectes de nouvelle physique nécessitent une haute précision sur les mesures des paramètres de Modèle Standard. Théories de la physique au-delà de Modèle Standard, comme théories de dimensions supplémentaires ou modèles composite, impliquent des modifications des couplages électrofaibles des quarks lourds, top et bottom. La deuxième partie de la thèse est une étude de simulation complète des algorithmes de vertexing dans l’environnement ILD et la reconstruction de la charge de quark b. La reconstruction de la charge du quark bottom est essentielle pour de nombreux canaux de physique à l’ILC, particulièrement, pour les réactions e⁺e⁻ → bb̄ et e⁺e⁻ → tt̄ . L’algorithme développé améliore la performance de reconstruction de la charge du quark bottom. Les méthodes de reconstruction de la charge du quark bottom sont appliquées à l’analyse du mécanisme de production tt̄ . Cela permet d’augmenter la statistique pour l’estimation du facteur de forme électrofaible du quark top par rapport à une étude antérieure et donc de diminuer les incertitudes statistiques correspondantes. Les résultats de l’étude du détecteur permettent d’estimer la précision de l’ILC sur les couplages et les facteurs de forme électrofaibles du quark bottom. L’ILC sera capable de résoudre l’anomalie du LEP dans le processus de production bb̄. La précision de l’ILC sur le couplage droite Z⁰bb̄, un candidat majeur pour les effets de la nouvelle physique, est calculée et est au moins 5 fois mieux que celle des expériences de LEP. / This thesis presents studies for the International Linear Collider (ILC),a linear electron-positron collider with a nominal center-of-mass energy of 500 GeV. Data are analysed that were recorded with the physics prototype of the CALICE silicon-tungsten electromagnetic calorimeter (Si-W ECAL) prototype at FermiLab in 2008. During this thesis, a track-finding algorithm was developed, which finds secondary tracks in hadronic events recorded by the Si-W ECAL physics prototype. This algorithm reveals details of hadronic interactions in the detector volume and the results are compared with simulations based on the geant4 toolkit.Indirect searches of New Physics require a high precision on the measurements of the Standard Model parameters. Many Beyond Standard Model theories, like extradimentional or composite models, imply modifications of electroweak couplings of the heavy quarks, top and bottom. The second part of the thesis is a full simulation study of vertexing algorithms in the ILD environment and the reconstruction of the b-quark charge. The b-quark charge reconstruction is essential for many physics channels at the ILC, particularly, for the e+ e− → bb̄ and the e+ e− → tt̄ channels. The developed algorithm improves the b-quark charge reconstruction performance.The b-quark charge reconstruction methods are applied to the tt̄ production process. This allows to increase statistics for the top quark electroweak form factor estimation w.r.t an earlier study and thus to decrease corresponding statistical uncertainties.The results of the detector study allow for an estimation of the ILC precision on the b-quark electroweak couplings and form factors. The ILC will be able to resolve the LEP anomaly in the bb̄ production process. The ILC precision on the right-handed Z⁰bb̄ coupling, a prime candidate for effects of new physics, is calculated to be at least 5 times better than theLEP experiments.

Ce travail de thèse a été effectué au sein de la collaboration ATLAS au LHC. Il s'intéresse dans un premier temps au développement de nouveaux détecteurs Micromegas pour le projet NSW, et dans un second temps à la recherche d'un processus physique très rare : la production de quatre quarks top au LHC, à une énergie dans le centre de masse de 13 TeV. Le programme de physique du LHC prévoit une augmentation significative de la luminosité de l'accélérateur dans les années à venir, en particulier pour la phase haute-luminosité dont le démarrage est prévu en 2026. Certains composants du détecteur ATLAS doivent être mis à jour, notamment les petites roues du spectromètre à muons qui vont devoir supporter une irradiation bien supérieure au niveau actuel. De nouvelles petites roues (NSW) sont en cours de développement et une partie des détecteurs Micromegas pour ces petites roues est construite par l'IRFU. Les performances de ces détecteurs sont validées sur un banc de test utilisant des muons cosmiques. Dans cette thèse, l'objectif a été de développer un code d'analyse des données de ce nouveau banc de test afin de caractériser les détecteurs Micromegas pour le projet NSW. Le code fournit notamment des informations de gain et d'efficacité de détection, et sera utilisé jusqu'à la fin du projet. Les efficacités mesurées sont globalement en accord avec les attentes sauf sur certaines parties des détecteurs qui doivent être investiguées. Le quark top est un élément central de plusieurs théories au-delà du modèle standard, comme par exemple les modèles avec des dimensions supplémentaires, ou encore qui incluent une modification du secteur du boson de Higgs. Certains de ces modèles prédisent une augmentation de la section efficace de production de quatre quarks top qui est un phénomène très rare dans le modèle standard et dont la section efficace ne vaut qu'environ 12 fb à 13 TeV. L'étude du processus à quatre tops dans le canal à deux leptons de même charge électrique est particulièrement intéressante car le niveau de bruit de fond y est très faible. Ce travail de thèse s'intéresse d'abord à l'analyse des événements à quatre tops avec le détecteur ATLAS, portant sur les données prises en 2015 et 2016. L'étude de ce processus est un défi de par le faible nombre d'événements disponibles, mais aussi à cause du bruit de fond dû aux faux leptons qui est difficile à estimer. Plusieurs limites sont posées sur des modèles de nouvelle physique, et sur la production conforme au modèle standard. Cette dernière limite est combinée avec un résultat provenant du canal à un lepton pour fournir la meilleure limite au moment de sa publication. Le processus de production de quatre quarks top conforme au modèle standard est à nouveau étudié avec l'entièreté des données prises entre 2015 et 2018, dans le but d'en effectuer une première observation. L'analyse de données est en cours, et cette thèse se concentre sur le développement d'une nouvelle méthode d'estimation du bruit de fond dû aux faux leptons. / This doctoral work was carried out as a member of the ATLAS collaboration at the LHC, an accelerator with a center-of-mass energy of 13 TeV. The development of new Micromegas detectors for the NSW project has been studied, along with the search for a very rare physics process : the production of four top quarks at the LHC, at a center-of-mass energy of 13 TeV. The physics programme of the LHC plans a significant increase of the luminosity in the coming years, especially for the high-luminosity phase beginning in 2026. Some parts of the ATLAS detectors must be upgraded to cope with the increased radiation levels. The small wheels of the muon spectrometer will be replaced by new small wheels (NSW) using Micromegas detectors. IRFU is responsible for producing some of the detectors, which performances are being validated on a test bench using cosmic muons. During this doctoral work, the aim was to develop an analysis code for the test bench data, in order to characterize the Micromegas detectors of the NSW project. The code provides gain and detection efficiency informations, and will be used until the end of the project. The measured efficiencies are in agreement with the expectations except in some parts of the detectors that are being investigated. The top quark plays a central role in many theories beyond the standard model, for instance in models with extra dimensions, or with a modification of the Higgs boson sector. Many of these models predict an enhanced cross section for the very rare four top quark process, which is predicted to be approximately 12 fb at 13 TeV. The study of the four top quark process is particularly interesting in the channel with leptons having the same electric charge because of the very low background level. This doctoral work is firstly carried out on the data taken in 2015 and 2016 by the ATLAS detector, in order to study the four top quark process and some new physics models. This process is very challenging given the very low number of events, and the difficult estimation of the fake lepton background. Several limits are set on new physics models and on the standard model four top quark production. The results on the standard model process are combined with the results from the single lepton channel to set the tightest limit at the time of its publication. The four top quark production is being studied again, using the data taken from 2015 to 2018 this time, with the aim of making a first observation of this process. The doctoral work is focused on a new method for estimating the fake lepton background.

This thesis presents the measurement of the differential cross section of the top-antitop pair production in proton-proton collisions at center-of-mass energy of 8 TeV at the ATLAS experiment. The measurement is performed for top-antitop events in the single lepton decay channel in the boosted topol- ogy. The measured differential cross section is expressed as a function of the top quark transverse momentum at particle level and at parton level. The mea- sured distributions are compatible with the theoretical predictions of the Standard Model. The experimental techniques used in this measurement are discussed with emphasis on the jet reconstruction and on the identification of jets originating from b-quark fragmentation. The jet reconstruction is influenced by simultane- ous proton-proton collisions (pileup), and the mitigation of these pileup effects is studied. A novel pileup subtraction technique for jets is proposed using an ex- tension of the methods currently being employed by the LHC experiments. The new method has a good performance in removing the pileup contributions at the level of jet constituents. 1