Effectuer des mesures de polarimétrie des rayons X provenant de sources astrophysiques permettrait d’obtenir de nombreuses informations sur les objets émetteurs : géométrie des disques d’accrétion de pulsars, champ magnétique au cœur des restes de supernovæ ou encore détermination du spin des trous noirs. Ces informations fondamentales sont pour l’instant inaccessibles à cause de l’absence de polarimètres X performants.L’utilisation de l’effet photoélectrique pour effectuer de la mesure spectro polarimétrique des rayons X mous dans la bande d’énergie de 1 keV à 15 keV apparaît comme une approche bien plus adaptée que l’utilisation de la diffraction de Bragg ou de la diffusion Thomson. La polarimétrie par le truchement de l’effet photoélectrique repose sur la mesure de la direction d’éjection du photoélectron, laquelle est modulée par la direction de polarisation de la lumière incidente. Il s’agit alors de construire un détecteur permettant un recul suffisant des photoélectrons afin de reconstruire leurs traces, et les détecteurs gazeux sont par nature des candidats idéaux. Cette thèse traite du développement et de la caractérisation d’un spectro-polarimètre `a rayons X-mous d’un genre entièrement nouveau : Caliste-MM. Il consiste en un détecteur gazeux, le piggyback Micromegas associé à une électronique de lecture miniature baptisée Caliste. L’une des principales innovations de ce détecteur tient au fait que son électronique de lecture est située en dehors du milieu gazeux. Les charges créées dans le piggyback diffusent dans une couche résistive répandue sur une céramique venant fermer le détecteur gazeux. Le module électronique Caliste enregistre le signal qui se répand dans la couche résistive à travers la céramique et une fine lame d’air par couplage capacitif. Le détecteur est ainsi composé de deux parties complètement indépendantes : conversion de la lumière et amplification par le piggyback, et lecture du signal par le Caliste. Les deux peuvent alors être développées indépendamment l’une de l’autre, l’électronique étant protégée des étincelles développées dans le détecteur grâce à la couche résistive du piggyback.Les caractéristiques détaillées du détecteur sont étudiées et présentées : forme des évènements, gain, résolution en énergie, ainsi que la variation de ces caractéristiques avec les différents paramètres du détecteur. Des modèles analytiques sont comparés aux résultats obtenus afin d’expliciter les phénomènes physiques responsables de la topologie des évènements enregistrés. Les différentes méthodes pour obtenir une trace reconstructible issue de photoélectrons sont aussi étudiées : utilisation d’une électronique de lecture plus finement pixélisée (utilisant ainsi pleinement le concept d’électronique découplée), test en basse pression ou utilisation de gaz légers comme l’Hélium ou le Néon.Enfin, grâce à des mesures effectuées sur le faisceau 100% polarisé de la ligne Métrologie du synchrotron SOLEIL, le facteur de modulation du détecteur est mesuré et présenté à différentes énergies de 6 à 12 keV. Une mesure du facteur de modulation de 92% à 8 keV prouve le grand potentiel de ce nouveau spectro-polarimètre et l’intérêt de son concept innovant. / Performing X-ray polarimetry of astrophysical sources could provide precious insight into the properties of the emitting objects, for example the geometry of pulsars accretion disks, magnetic field inside the core of supernovae remnants or measurement of black holes spin. These fundamental observations are today impossible due to the missing performance of X-ray polarimeters. The use of the photo-electric effect to perform spectro-polarimetry in the energy band of 1 keV to 15 keV appears to be like a much better approach than the use of Bragg diffraction or Thomson scattering. Performing polarimetry with the photo-electric effect relies on the measurement of the ejection direction of the photo-electron, which is modulated by the polarization direction of the incoming light. In order to reconstruct the photo-electron track, a detector allowing the photo electrons to recoil far enough is needed. Gaseous detectors are naturally perfect candidates. This PhD thesis focusses on the development and the characterization of a soft X-ray spectro- polarimeter of a completely new design : Caliste-MM. It consists of a gaseous detector called piggyback Micromegas associated with a miniature 3D readout electronics baptized Caliste. The main innovation of this detector comes from the fact that its readout electronics is located outside the gaseous medium. The charges created inside the piggyback diffuse in a resistive layer spread on a solid ceramic plate that closes the detector. The Caliste records the signal of the charges in the resistive layer through the ceramic and a small air layer by capacitive induction. The detector is composed of two completely independent parts : the piggyback where the X-ray conversion and amplification takes place, and the Caliste for the recording of the signal. These two parts can then be developed independently. Moreover the electronics are protected from sparks thanks to the resistive layer of the piggyback.The detailed characteristics of the detector are studied such as the shape of the events, the gain and the energy resolution. Analytical models are compared to the obtained results in order to explain the physical phenomena responsible for the topology of the recorded events. Different strategies to improve the reconstruction of the photo-electrons are explored including for example finer pitched readout electronics, low pressure and the use of lighter gases such as Neon or Helium. Finally, thanks to the measurements performed on the 100% linearly polarized beam of the Mtrologie line of the SOLEIL synchrotron facility, the modulation factor of the detector has been measured at different energies ranging from 6 keV to 12 keV. A measurement of the modulation factor of 92% at 8 keV proves the high potential of this new spectro-polarimeter and the interest into its innovative design.

GEMs (Gas Electron Multiplier) and Thick-GEMs (Thick-Gas Electron Multiplier) are MPGDs (Micropattern Gas Detector) that make part of the new generation of gaseous detectors, allowing high counting rates, low cost when compared to solid state detectors, high radiation hardness and gain when using multiple structures. Besides that, the handling and maintenance of these detectors is relatively simple, being versatile to detect different types of radiation. Therefore, these detectors are an effective alternative to build imaging systems with large sensitive area. This work consists in the study and characterization of a set of gaseous detectors, more specifically the Thick-GEMs produced in the High Energy Physics and Instrumentation Center at IFUSP, which were tested showing promising results in terms of gain, energy resolution and operational stability. However, due to the low signal-to-noise ratio of the Thick-GEMs, the X-ray fluorescence imaging system was mounted using GEMs. During this work the necessary software tools for image processing and reconstruction were developed as a parallel study in computational simulations to better understand the operation of gaseous detectors. X-ray fluorescence techniques are essential in areas such as medicine and the study of historical and cultural heritage since they are non-invasive and non-destructive. Techniques to check the authenticity of masterpieces are required and museums are gradually becoming more interested in the Physics and instrumentation needed to characterize their patrimony. / Os GEMs (Gas Electron Multiplier) e Thick-GEMs (Thick-Gas Electron Multiplier) são estruturas do tipo MPGD (Micropattern Gas Detector) que fazem parte da nova geração de detectores de radiação a gás e permitem altas taxas de contagens, baixo custo quando comparados com os detectores de estado sólido, uma elevada resistência à radiação e ganhos elevados, quando utilizadas estruturas múltiplas para multiplicação. Além disso, o manuseio e manutenção desses detectores é relativamente simples, sendo versáteis em relação à montagem podendo detectar diferentes tipos de radiação. Sendo assim, a utilização desses detectores é uma alternativa eficiente para montar um sistema de imagem com grande área sensível. Este trabalho consiste no estudo e caracterização de um conjunto de detectores gasosos, mais especificamente os Thick-GEMs produzidos pelo grupo de Física de altas energias e Instrumentação do IFUSP, que foram testados para serem empregados em um sistema de imagem de fluorescência de raios-X. Os Thick-GEMs testados apresentaram resultados promissores em termos de ganho, resolução em energia e estabilidade operacional. No entanto, devido à baixa relação sinal-ruído, um sistema de imagem de fluorescência de raios-X foi montado utilizando GEMs. Durante o trabalho as ferramentas de software necessárias para processamento e reconstrução de imagens foram desenvolvidas, assim como um estudo paralelo de simulações computacionais para entender melhor o funcionamento de detectores gasosos. Técnicas como o imageamento por fluorescência de raios-X são de suma importância pois são consideradas não invasivas e não destrutivas. Sua utilização tem uma importância imprescindível nas áreas da medicina e na análise de patrimônios histórico e cultural. Atualmente, a verificação e validação de autenticidade de obras é um requisito obrigatório e alguns museus começam a se interessar cada vez mais em áreas da Física e da instrumentação necessária para caracterizar o seu patrimônio.

GEMs (Gas Electron Multiplier) and Thick-GEMs (Thick-Gas Electron Multiplier) are MPGDs (Micropattern Gas Detector) that make part of the new generation of gaseous detectors, allowing high counting rates, low cost when compared to solid state detectors, high radiation hardness and gain when using multiple structures. Besides that, the handling and maintenance of these detectors is relatively simple, being versatile to detect different types of radiation. Therefore, these detectors are an effective alternative to build imaging systems with large sensitive area. This work consists in the study and characterization of a set of gaseous detectors, more specifically the Thick-GEMs produced in the High Energy Physics and Instrumentation Center at IFUSP, which were tested showing promising results in terms of gain, energy resolution and operational stability. However, due to the low signal-to-noise ratio of the Thick-GEMs, the X-ray fluorescence imaging system was mounted using GEMs. During this work the necessary software tools for image processing and reconstruction were developed as a parallel study in computational simulations to better understand the operation of gaseous detectors. X-ray fluorescence techniques are essential in areas such as medicine and the study of historical and cultural heritage since they are non-invasive and non-destructive. Techniques to check the authenticity of masterpieces are required and museums are gradually becoming more interested in the Physics and instrumentation needed to characterize their patrimony. / Os GEMs (Gas Electron Multiplier) e Thick-GEMs (Thick-Gas Electron Multiplier) são estruturas do tipo MPGD (Micropattern Gas Detector) que fazem parte da nova geração de detectores de radiação a gás e permitem altas taxas de contagens, baixo custo quando comparados com os detectores de estado sólido, uma elevada resistência à radiação e ganhos elevados, quando utilizadas estruturas múltiplas para multiplicação. Além disso, o manuseio e manutenção desses detectores é relativamente simples, sendo versáteis em relação à montagem podendo detectar diferentes tipos de radiação. Sendo assim, a utilização desses detectores é uma alternativa eficiente para montar um sistema de imagem com grande área sensível. Este trabalho consiste no estudo e caracterização de um conjunto de detectores gasosos, mais especificamente os Thick-GEMs produzidos pelo grupo de Física de altas energias e Instrumentação do IFUSP, que foram testados para serem empregados em um sistema de imagem de fluorescência de raios-X. Os Thick-GEMs testados apresentaram resultados promissores em termos de ganho, resolução em energia e estabilidade operacional. No entanto, devido à baixa relação sinal-ruído, um sistema de imagem de fluorescência de raios-X foi montado utilizando GEMs. Durante o trabalho as ferramentas de software necessárias para processamento e reconstrução de imagens foram desenvolvidas, assim como um estudo paralelo de simulações computacionais para entender melhor o funcionamento de detectores gasosos. Técnicas como o imageamento por fluorescência de raios-X são de suma importância pois são consideradas não invasivas e não destrutivas. Sua utilização tem uma importância imprescindível nas áreas da medicina e na análise de patrimônios histórico e cultural. Atualmente, a verificação e validação de autenticidade de obras é um requisito obrigatório e alguns museus começam a se interessar cada vez mais em áreas da Física e da instrumentação necessária para caracterizar o seu patrimônio.

Le spectromètre à muons de CMS doit permettre l'identification rapide et efficace des muons produits lors des collisions proton-proton au LHC. Cependant, à cause d'un environnement de détection extrême, seules les chambres à pistes cathodiques équipent actuellement les bouchons de CMS. Cette faiblesse dans le système de détection pourrait devenir problématique après l'amélioration du LHC. L'augmentation du taux de particules dans les bouchons va dégrader les performances du système de déclenchement L1 ainsi que l'efficacité de sélection des phénomènes physiques intéressants. Le but du programme d'amélioration de CMS est de maintenir le taux de déclenchement L1 tout en gardant une efficacité de sélection maximale. La collaboration CMS GEM propose d'équiper les régions vacantes des bouchons avec des détecteurs basés sur la technologie d'amplification des électrons dans un gaz (GEM), appelés GE1/1. Le sujet de thèse proposé par la collaboration CMS GEM a pour but de justifier le choix de la technologie GEM pour l'amélioration de CMS. Trois projets ont été suivis pendant la thèse. La première partie du projet de thèse consistait à mesurer précisément les caractéristiques fondamentales et les performances de détection des détecteurs triple-GEM produit par la technique simple-masque. Ces éléments sont essentiels pour s'assurer que les détecteurs GE1/1 pourront fonctionner en toute sécurité dans l'environnement des bouchons de CMS. Le deuxième projet consistait à prouver que les excellentes performances des détecteurs GE1/1 ne vont pas se dégrader pendant leur utilisation dans CMS. Cette étape comprend l'étude du fonctionnement à long terme des détecteurs GE1/1, en particulier le phénomène de vieillissement, qui inclue tous les processus physiques et chimiques qui provoquent la dégradation graduelle et permanente des performances de détection. Plusieurs tests de vieillissement ont été menés dans des zones d'irradiations spécifiques au CERN pour reproduire un minimum de 10 ans de fonctionnement réel dans l'environnement de CMS après la montée en puissance du LHC. Enfin, les excellentes propriétés mesurées lors de la phase de R&D ont permis de valider la technologie triple-GEM, qui fut ensuite approuvée par la collaboration CMS et le comité du LHC. La production de 144 grands détecteurs GE1/1 sera partagée entre différents sites de production à travers le monde. Le troisième projet de thèse fut donc le développement des principales étapes du contrôle qualité nécessaire pour assurer une production uniforme et une même qualité pour tous les détecteurs GE1/1. / The muon system of CMS aims to provide an efficient and fast identification of the muons produced in the proton-proton collisions. However, the forward region of the end-caps is only instrumented with Cathode Strip Chambers. This lack of redundancy will be problematic after the high-luminosity upgrade of the LHC (HL-LHC), for which the increase of the background rate would degrade the Level-1 trigger performance and thus the selection of interesting physics channels. The goal of the CMS muon upgrade is to maintain the L1 trigger rate with maximum selection efficiency in order to fully exploit the HL-LHC. The CMS GEM Collaboration has proposed to instrument the vacant highetaregion of the muon end-caps with Gas Electron Multiplier (GEM) detectors, called GE1/1chambers.The Ph.D. subject proposed by the CMS GEM Collaboration aims to demonstrate that the GE1/1technology is the most suitable choice for the upgrade of the muon end-caps. Three main researchprojects were conducted in this context. The first project included the precise measurement of the fundamental characteristics and the detection performance of the triple-GEM detectors produced with the single-mask technique. Those characteristics are essential to ensure that the detectors can operate in the forward region of CMS. The second project was focused on the long-term operation of GE1/1 detectors, in particular the study of the aging phenomenon, which includes all the processes that lead to a significant and permanent degradation of the performance of the detectors. Several aging tests were performed at the CERN irradiation facilities to prove that the GE1/1 chambers can operate during at least 10 yearsat HL-LHC without suffering from performance losses. The excellent properties measured during the R&D phase led to the approval of the GE1/1 project by the CMS Collaboration. The third project, conducted in the framework of the mass production,consisted of developing of the main steps of the Quality Control of the GE1/1 chambers.

Lab measurements and Monte Carlo simulations have been carried out for the evaluation of the Straw-type detectors used in the NA62 experiment at CERN. In addition, analyses of experiment data was used in corrections to improve the reconstruction of particle tracks, ultimately leading to improved resolution of the detector system as a whole. 97.7 percent of the Straws were aligned to within 30 microns, quantified as the deviation from zero of the mean of the inherent residual distribution of each Straw. A drift time dependence on where along the Straw the particle ionized have been corrected for; before the correction the dependence was as big as 6 ns. A radius-drift time relation based on the leading edge timing distribution has been deduced and implemented. Upon implementation artifacts from the piecewise fits used became evident. An alternative approach using residuals has been put forward.

Le travail présenté dans cette thèse porte d’une part sur le développement et la caractérisation d’une nouvelle génération de détecteurs Micromegas (MICROMEsh GAseous Structure) pour l’expérience COMPASS au Cern, et d’autre part sur l’estimation de la contribution de canaux diffractifs à la production de pions et de kaons, dans le cadre de l’étude des fonctions de fragmentation de quarks en hadrons à COMPASS. De nouveaux détecteurs Micromegas d’une surface active de 40 x 40 cm² sont en cours de développement pour le futur de l’expérience COMPASS, à partir de 2015. Ces détecteurs devront fonctionner dans des flux de muons et hadrons approchant 8 MHz/cm². Pour cela, une zone centrale d’environ 5 cm de diamètre, traversée par le faisceau, sera constituée de 1280 pixels, et des technologies permettant de fortement réduire l’impact des décharges seront adoptées. Quatre prototypes de géométrie finale utilisant deux types de technologies de réduction des décharges différentes ont été produits au Cern et testés en conditions réelles à COMPASS entre 2010 et 2012. Trois d’entre eux sont des détecteurs hybrides utilisant une feuille GEM (Gas Electron Multiplier) en tant qu’étage de préamplification pour réduire la probabilité de décharge. Le dernier détecteur est équipé d’une structure résistive à « résistances enterrées », permettant la réduction de l’amplitude des décharges. Leurs performances sont présentées dans cette thèse. L’impact de ces résultats sur la production et la mise en œuvre de la série finale de détecteurs est également discuté. Les fonctions de fragmentation de quarks en hadrons l’hadronisation d’un quark de saveur q en un hadron h. Ces fonctions universelles interviennent dans l’expression de la section efficace de nombreux processus. Elles sont accessibles à COMPASS via la réaction de diffusion profondément inélastique semi-inclusive de muons sur des nucléons. Les observables permettant leur extraction dans ce cadre sont les multiplicités de hadrons, soit le nombre moyen de hadrons produits par événement de diffusion profondément inélastique. Les mésons vecteurs issus de processus diffractifs produisent également des pions et kaons par leur désintégration. Ces processus n’impliquant pas l’hadronisation d’un quark, il semble justifié de ne pas les comptabiliser dans la mesure des multiplicités. Ce travail propose une étude Monte-Carlo des contributions de la production diffractive de mésons vecteurs rho et phi à la production de hadrons léger et aux événements inclusifs. Des facteurs de correction des multiplicités sont finalement établis. Enfin, l’effet de cette correction sur l’extraction des fonctions de fragmentation en pions est présenté. / This thesis is dedicated to the development and characterisation of a new generation of Micromegas (MICROMEsh GAseous Structure) detectors for the COMPASS experiment at Cern, and the estimation of the diffractive processes’ contribution to the production of pions and kaons, concerning the study of the quark fragmentation functions into hadrons at COMPASS. New Micromegas detectors with a 40 x 40 cm² active area are being developed for the future physics program of the COMPASS experiment starting in 2015. These detectors will have to work in high muon and hadron flux (close to 8 MHz/cm²). In this context, a central area of about 5 cm diameter, crossed by the beam, will be composed of 1280 pixels, and discharge-reduction technologies will be used. Four prototypes with a final read-out geometry, using two different discharge-reduction technologies have been produced at Cern and studied in nominal conditions at COMPASS between 2010 and 2012. Three are hybrid detectors using a GEM (Gas Electron Multiplier) foil as a preamplification stage to reduce the discharge probability. The other is equipped with a so called “buried resistors” resistive structure allowing to strongly reduce the discharge amplitude. Their performances are presented in this thesis. The impact of these results on the production and implementation of the final series of detectors is also discussed. Quark fragmentation functions into hadrons describe the hadronisation of a quark of flavor q into a hadron h. These universal functions take part in the cross-section expression of several processes. They can be accessed at COMPASS via semi-inclusive deep inelastic scattering of muons off nucleons. The relevant observables for fragmentation function extraction are hadron multiplicities, corresponding to the mean number of hadrons produced per deep inelastic scattering event. Vector mesons produced by a diffractive process can decay into pions and kaons. As such processes do not involve quark hadronisation, they should a priori be excluded from multiplicity measurements. This work presents a Monte-Carlo study of the impact of diffractive rho and phi production on light hadrons and inclusive events yields. Multiplicity correction factors are finally established. The effect of this correction on the extraction of pion fragmentation functions is also discussed.