Les ionomères aromatiques sont considérés comme une alternative prometteuse à Nafion pour les PEMFCs en raison de leur bonne stabilité à l'oxydation, d'excellentes propriétés thermomécaniques et de faibles coûts, etc. La plupart des ionomères aromatiques sulfonés rapportés au cours des dernières décennies présentent cependant des performances inférieures à celles de Nafion. Avec une capacité d'échange ionique (CEI) similaire, d'une part, les ionomères aromatiques sont beaucoup moins conducteurs que Nafion, notamment à faible humidité relative. Les ionomères aromatiques ayant une CEI suffisante pour donner une conductivité équivalente à celle de Nafion, d'autre part, présentent un comportement excessivement gonflant dans l'eau. Les inconvénients des ionomères aromatiques sulfonés proviennent de (i) la répartition aléatoire de groupes acides sur un squelette de polymère rigide conduisant à une séparation hydrophile-hydrophobe faible, (ii) la proximité de fractions conductrices de protons à la chaîne principale de polymère conduisant à une nanostructure basse de composés ioniques, et (iii) la faible acidité de l'acide arylsulfonique. Dans le but de surmonter ces inconvénients, mon travail de doctorat se concentre sur le développement de nouveaux ionomères aromatiques avec une morphologie et des propriétés améliorées grâce à la conception de l'architecture moléculaire, en combinaison avec une condition optimisée de traitement de la membrane. A base de cet objectif, deux séries d'ionomères aromatiques à base de copoly (arylène éther sulfone) partiellement fluoré portant des chaînes latérales pendantes d'acide perfluorosulfonique (séries InX/Y) ou perfluorosulfonimide (SiX/Y) ont été développées et caractérisées. De plus, les PEM basés sur le mélange Nafion/InX/Y ont également été ciblés. Une grande attention a été portée à l'optimisation de l'état de traitement des membranes et à l'élucidation de la relation structure-morphologie-propriété des matériaux. / Aromatic ionomers are considered as a promising alternative to Nafion for PEMFCs due to their good oxidative stability, excellent thermomechanical properties, and low cost, etc. Most sulfonated aromatic ionomers reported over the past decades, however, show lower performance than that of Nafion. With similar ion-exchange capacity (IEC), on one hand, aromatic ionomers are much less conductive than Nafion, notably at low relative humidity. Aromatic ionomers with sufficient IEC to give equivalent conduction to that of Nafion, on the other hand, exhibit excessively swelling behavior in water. The shortcomings of sulfonated aromatic ionomers derive from (i) the random distribution of acidic groups on rigid polymer backbone leading to poor hydrophilic-hydrophobic separation, (ii) the proximity of proton-conducting moieties to the polymer main chain resulting in low nanostructure of ionic clusters, and (iii) the low acidity of aryl sulfonic acid. With the aim of overcoming these drawbacks, my PhD work focuses on developing new aromatic ionomers with improved morphology and properties via molecular architecture design, in combination with optimized membrane processing condition. Based on this objective, two series of aromatic ionomers based on partially-fluorinated multi-block copoly(arylene ether sulfone)s bearing pendant perfluorosulfonic acid (InX/Y series) or perfluorosulfonimide (SiX/Y series) side chains have been developed and characterized. Moreover, PEMs based on Nafion/InX/Y blend have also been focused. Much attention has been paid to optimizing the membrane processing condition and elucidating the structure-morphology-property relation in these materials.

Energia é um recurso que historicamente apresenta tendência de crescimento de demanda. Projeções indicam que, para suprir as necessidades energéticas do futuro, será necessário um uso massivo do hidrogênio como combustível. O uso de sistemas de célula a combustível baseada no uso de membrana polimérica condutora de prótons (PEMFC) tem características que permitem sua aplicação para geração de energia elétrica em aplicações estacionárias, automotivas e portáteis. O uso de hidrogênio como combustível para PEMFC apresenta a vantagem de resultar em baixa emissão de poluentes quando comparado às dos combustíveis fósseis. Para que ocorram as reações em uma PEMFC é necessária a construção de conjuntos eletrodo-membrana-eletrodo (MEA), sendo o processo de produção de MEAs e os materiais utilizados, relevantes no custo final do kW instalado para geração de energia por sistemas de célula a combustível, o que é, atualmente, uma barreira tecnológica e financeira para a aplicação em grande escala desta tecnologia. Nesse trabalho foi desenvolvido um processo de produção de MEAs por impressão a tela que apresenta alta reprodutibilidade, rapidez e baixo custo. Foram desenvolvidos o processo de impressão a tela e a composição de uma tinta precursora da camada catalisadora (TPCC), que permitem o preparo de eletrodos para confecção de MEAs com a aplicação da massa exata de eletrocatalisador adequada para cátodos 0,6 miligramas de platina por centímetro quadrados (mgPt.cm-2) e ânodos 0,4 mgPt.cm-2 em apenas uma aplicação por eletrodo. A TPCC foi desenvolvida, produzida, aplicada e caracterizada, apresentando características semelhantes a de tintas de impressão a tela para outras aplicações. Os MEAs produzidos apresentaram desempenho de até 712 mA.cm-2 a 600 mV para MEAs de 25 cm2 e o custo para produção de MEAs de 247,86 cm2 capazes de gerar 1 kW de energia foi estimado em R$ 13.939,45, considerando custo de equipamentos, materiais e mão de obra. / Energy is a resource that presents historical trend of growth in demand. Projections indicate that future energy needs will require a massive use of hydrogen as fuel. The use of systems based on the use of proton exchange membrane fuel cell (PEMFC) has features that allow its application for stationary applications, automotive and portable power generation. The use of hydrogen as fuel for PEMFC has the advantage low pollutants emission, when compared to fossil fuels. For the reactions in a PEMFC is necessary to build membrane electrode assembly (MEA). And the production of MEAs and its materials are relevant to the final cost of kW of power generated by systems of fuel cell. This represent currently a technological and financial barriers to large-scale application of this technology. In this work a process of MEAs fabrication were developed that showed high reproducibility, rapidity and low cost by sieve printing. The process of sieve printing and the ink composition as a precursor to the catalyst layer were developed, which allow the preparation of electrodes for MEAs fabrication with the implementation of the exact catalyst loading, 0.6 milligrams of platinum per square centimeters (mgPt.cm-2) suitable for cathodes and 0.4 mgPt.cm-2 for anode in only one application step per electrode. The ink was developed, produced, characterized and used with similar characteristics to ink of sieve printing build for other applications. The MEAs produced had a performance of up to 712 mA.cm-2 by 600 mV to 25 cm2 MEA area. The MEA cost production for MEAs of 247.86 cm2, that can generate 1 kilowatt of energy was estimated to US$ 7,744.14 including cost of equipment, materials and labor.

O desenvolvimento de sistemas de conversão de energia baseados na tecnologia de células a combustível tem demandado estudos de confiabilidade, uma vez que requisitos de durabilidade e custo passaram a ser fundamentais para a inserção desta tecnologia no mercado de energia. Neste trabalho foi proposta uma metodologia de análise de confiabilidade de células a combustível de membrana polimérica condutora de prótons (células a combustível do tipo PEM), tendo em vista a qualificação destes itens como protótipos de unidades comerciais. A metodologia proposta incluiu uma avaliação inicial qualitativa das possíveis falhas em células a combustível do tipo PEM, realizada pela aplicação de uma Análise de Modos de Falha e Efeitos, técnica conhecida pela sigla FMEA. Além disso, foi elaborado um plano de testes de vida para as unidades produzidas no IPEN CNEN/SP e foi efetuada a análise dos resultados usando-se métodos estatísticos para dados de confiabilidade. Assim, a parte experimental consistiu em produzir as células a combustível no laboratório e submetê-las aos testes de vida, nos quais foram simuladas condições reais de operação. Os dados amostrais foram analisados estatisticamente, gerando resultados importantes em relação às medidas de desempenho e durabilidade dos dispositivos em estudo. Por meio de uma análise não paramétrica, foi gerada uma estimativa da função de confiabilidade das células a combustível usando-se o estimador de Kaplan- Meier. Pela modelagem paramétrica, foi possível ajustar uma distribuição exponencial para o tempo de vida destes dispositivos, gerando uma estimativa de vida média de 1.094,58 horas, com intervalo de 95% de confiança de [533,03 horas; 2.836,13 horas]. Com relação ao desempenho, foi aplicada uma regressão linear simples aos dados de potencial elétrico ao longo do tempo, gerando um valor aproximado para a taxa de queda do potencial elétrico de 80 μV h-1. Cabe ressaltar que, a metodologia proposta neste estudo deverá ser integrada ao processo de desenvolvimento das células a combustível, para permitir o acompanhamento da melhoria de sua confiabilidade. / The development of power conversion systems based on fuel cells has been demanding reliability studies since the requirements associated to cost and durability of these products have become fundamental to their acceptance in the energy market. The reliability analysis proposed in this work aimed to qualify the proton exchange membrane (PEM) fuel cells produced at IPEN CNEN/SP as prototypes of commercial units. The methodology included an initial assessment of failures of the PEM fuel cells, which was carried out by the application of a Failure Mode and Effect Analysis (FMEA). Furthermore, the methodology involved the planning of life tests and the quantitative assessment based on statistical methods for reliability data. Thus, the experimental part of this work consisted of producing the fuel cells in the laboratory, and testing them under stable and dynamic operating conditions. The main results of the quantitative analysis were the estimates of the reliability function and mean time to failure of the PEM fuel cells. These estimates were obtained by means of two different approaches, a non parametric and a parametric one. In the non parametric analysis, the Kaplan-Meier estimator was used for the reliability function. Based on the parametric analysis, an exponential distribution was fit to data and the mean time to failure of the PEM fuel cells was estimated in 1.094,58 hours, with a 95% confidence interval of [533,03 hours; 2.836,13 hours]. Concerning the PEM fuel cells performance, a linear regression analysis of the observed data (voltage vs. time) was carried out and an approximate estimate for the voltage decay rate was calculated in 80 μV h-1. Finally, the methodology proposed in this work may be integrated into the development project of the PEM fuel cells so as to allow researchers to carry out a reliability growth analysis and monitor the improvements in these products.

La pile à combustible convertit l’énergie chimique des réactants en énergie électrique continue, en chaleur et en eau. Elle est généralement utilisée autour d’un point de fonctionnement (ou zone) correspondant à un maximum de puissance électrique. Le courant continu produit par la réaction d’oxydo-réduction est proportionnel à la surface active de la pile et la tension, qui est d’environ 0,6 V au point de nominal de fonctionnement, peut être augmentée par la mise en série de plusieurs cellules (constituant un stack). En raison de son faible niveau de tension continue, son utilisation dans des systèmes électriques nécessitent de l’associer à des convertisseurs de puissance. Les travaux effectués dans le cadre de cette thèse s’intéressent au potentiel d’une source électrique continue basse tension et plus exactement à l’utilisation de la pile à combustible en fonctionnement source de courant commandée (par le débit d’un des réactants). L’expertise du laboratoire GREEN dans le domaine des supraconducteurs, nous a conduits naturellement vers une application innovante à savoir substituer les alimentations de puissance dédiées aux dispositifs supraconducteurs par une pile à combustible. Un premier essai prometteur mené sur une bobine supraconductrice de 4 mH a mis en évidence tout le potentiel d’une telle application et nous a encouragés à étendre l’étude à des bobines supraconductrices fortement inductives, des plusieurs henrys. En effet, les énergies mises en jeu sont alors plus importantes et exigent de dimensionner avec soin le banc d’essai, aussi bien du point de vue de la protection de la pile que des conditions opératoires. Pour ce faire, une modélisation et une expérimentation d’un ensemble pile à combustible/bobine supraconductrice ont également été réalisées. En parallèle du travail mené sur la partie alimentation de la bobine supraconductrice, nous avons travaillé sur le dimensionnement d’un dispositif supraconducteur innovant, communément appelé aimant vectoriel, à trois axes. Ce système peut servir comme charge pour une pile à combustible mais aussi, et surtout, comme outils de caractérisation d’échantillons supraconducteurs. Cet aimant vectoriel permet d’orienter dans les 3 directions de l’espace un champ magnétique de plusieurs teslas, uniforme à plus de 95 % dans une sphère de 100 mm de diamètre. Ce dimensionnement, nous a permis de concevoir et réaliser la structure supportant le bobinage du fil et de choisir un certain fil supraconducteur. Le système complet devant coûter moins de 50 k€, cryostat inclus, nous nous sommes orientés vers du fil supraconducteur à basse température critique, refroidi à l’hélium liquide / The fuel cell (FC) converts the chemical energy of the reactants into direct electrical energy, heat and water. The FC is generally used around an operating point (or area) corresponding to a maximum of electric power. The direct current produced by the redox reaction is proportional to the active surface of the single cell and its voltage, which is approximately 0.6 V at the nominal operating point, can be increase by connecting several cells in series (constituting a stack). Due to its low DC voltage amplitude, its use in electrical systems requires the use of power converters. In this work, we have been interested taking benefit of such DC low voltage power source and more precisely the use of the FC as a current source controllable by the one of the reactant flow rates. The expertise of GREEN laboratory in the field of superconductors has naturally led us to an innovative application, namely to substitute the power supplies dedicated to the superconducting devices by a FC. A first promising test conducted on a 4 mH superconducting coil highlighted the full potential of such an application and encouraged us to extend the study to highly inductive superconducting coils where the energies involved are more important. This requires to carefully design the test bench with a protection system for the FC as well as operating conditions. To this end, a FC model supplying a superconducting coil has been developed and tested experimentally. At the same time, we have focused on the supply part of the superconducting coil by designing an innovative superconducting device, commonly called a three-axis vector magnet. This system can be used as a load for a fuel cell, but also, and above all, as a tool for the characterization of superconducting samples. This vector magnet allows to orient a magnetic field of several tesla in the three space directions, with a uniformity of more than 95 % in a 100 mm sphere of diameter. This design allowed us to realize the windings supporting structure and to choose a superconducting wire. The complete system has to cost less than 50 k€, including the cryostat, we have finally choose a superconducting wire with low critical temperature, cooled by liquid helium

De nouveaux électrocatalyseurs avec nanofilm de platine pour la réaction de réduction de l'oxygène avec application dans des piles à combustible à membrane échangeuse de protons ont été développés. Ces catalyseurs comprennent des films minces de platine déposés sur des réseaux de nanofibres de carbone. Des supports de nanofibres de carbone et de nanobrosse ont été préparés par électrofilage suivi de traitements thermiques pour la stabilisation et la graphitisation. Une méthode innovante d’électrodéposition pulsée à surpotentiel élevé a été développée pour le dépôt de nanofilm de platine sur des supports de nanofibres de carbone et de nanobrosse, ainsi que sur du graphite pyrolytique hautement orienté dont la planéité permet de caractériser le dépôt avec microscopie à force et électronique. Ces approches ont conduit à des électrodes en nanofibres autosupportées avec une porosité qui a été accordée à un matériau de plus en plus dense d'un côté à l'autre, où le côté présentant la plus grande surface était utilisé pour déposer du platine. Les électrodes ont été caractérisées ex situ en utilisant voltampérométrie cyclique, en démontrant une activité plus élevée pour la réaction de réduction de l'oxygène et une durabilité contre des cycles de tension plus élevée que les catalyseurs classiques au platine sur carbone. Ces électrodes ont été assemblés directement avec une membrane et une anode et caractérisés in situ dans une pile à combustible. Des films minces de platine ont également été préparés à la surface des nanofibres de nickel en utilisant le nouvelle approche de l'échange galvanique assisté par micro-ondes ; divers paramètres expérimentaux ont été étudiés pour déterminer leur effet sur l'échange et la morphologie du platine. Les fibres de nickel@platine résultantes ont présenté une électroactivité élevée pour la réaction de réduction d'oxygène et ont été caractérisées comme des électrocatalyseurs non supportés à la cathode d'un assemblage d'électrodes à membrane; des travaux supplémentaires sont nécessaires pour les stabiliser contre la perte de nickel de l’électrocatalyseur vers l’électrolyte. / Novel platinum thin film electrocatalysts for the oxygen reduction reaction of proton exchange membrane fuel cells were developed. These catalysts comprise platinum thin films deposited on carbon nanofibrous webs. Carbon nanofibres and nanobrush supports were prepared by electrospinning followed by thermal treatments for stabilisation and graphitisation. An innovative pulsed high overpotential electrodeposition method was developed to deposit platinum thin films both on carbon nanofibre and nanobrush supports, and also on highly oriented pyrolytic graphite, the planarity of which allowed detailed characterisation of the conformity, contiguity and thickness of the platinum films using atomic force and electron microscopy. These approaches led to self-standing nanofibre electrodes with porosity that was tuned to increasingly dense material from one side to the other, where the side presenting highest surface area was used to deposit platinum. The electrodes were characterised ex situ using cycling voltammetry where they demonstrated higher activity for the oxygen reduction reaction and greater durability on voltage cycling than conventional platinum on carbon catalysts. They were also assembled directly with a membrane and anode and characterised in situ in a single fuel cell. Thin platinum films were also prepared at the surface of nickel nanofibres using a novel approach to galvanic exchange assisted by microwaves, and a range of experimental parameters was investigated to determine their effect on the extent of exchange and the resulting platinum morphology. While the resulting nickel@platinum core@shell fibres demonstrated high electroactivity for the oxygen reduction reaction and were characterised as unsupported electrocatalysts at the cathode of a membrane electrode assembly, further work is required to stabilise them against nickel leaching from the catalyst to the electrolyte.

Energia é um recurso que historicamente apresenta tendência de crescimento de demanda. Projeções indicam que, para suprir as necessidades energéticas do futuro, será necessário um uso massivo do hidrogênio como combustível. O uso de sistemas de célula a combustível baseada no uso de membrana polimérica condutora de prótons (PEMFC) tem características que permitem sua aplicação para geração de energia elétrica em aplicações estacionárias, automotivas e portáteis. O uso de hidrogênio como combustível para PEMFC apresenta a vantagem de resultar em baixa emissão de poluentes quando comparado às dos combustíveis fósseis. Para que ocorram as reações em uma PEMFC é necessária a construção de conjuntos eletrodo-membrana-eletrodo (MEA), sendo o processo de produção de MEAs e os materiais utilizados, relevantes no custo final do kW instalado para geração de energia por sistemas de célula a combustível, o que é, atualmente, uma barreira tecnológica e financeira para a aplicação em grande escala desta tecnologia. Nesse trabalho foi desenvolvido um processo de produção de MEAs por impressão a tela que apresenta alta reprodutibilidade, rapidez e baixo custo. Foram desenvolvidos o processo de impressão a tela e a composição de uma tinta precursora da camada catalisadora (TPCC), que permitem o preparo de eletrodos para confecção de MEAs com a aplicação da massa exata de eletrocatalisador adequada para cátodos 0,6 miligramas de platina por centímetro quadrados (mgPt.cm-2) e ânodos 0,4 mgPt.cm-2 em apenas uma aplicação por eletrodo. A TPCC foi desenvolvida, produzida, aplicada e caracterizada, apresentando características semelhantes a de tintas de impressão a tela para outras aplicações. Os MEAs produzidos apresentaram desempenho de até 712 mA.cm-2 a 600 mV para MEAs de 25 cm2 e o custo para produção de MEAs de 247,86 cm2 capazes de gerar 1 kW de energia foi estimado em R$ 13.939,45, considerando custo de equipamentos, materiais e mão de obra. / Energy is a resource that presents historical trend of growth in demand. Projections indicate that future energy needs will require a massive use of hydrogen as fuel. The use of systems based on the use of proton exchange membrane fuel cell (PEMFC) has features that allow its application for stationary applications, automotive and portable power generation. The use of hydrogen as fuel for PEMFC has the advantage low pollutants emission, when compared to fossil fuels. For the reactions in a PEMFC is necessary to build membrane electrode assembly (MEA). And the production of MEAs and its materials are relevant to the final cost of kW of power generated by systems of fuel cell. This represent currently a technological and financial barriers to large-scale application of this technology. In this work a process of MEAs fabrication were developed that showed high reproducibility, rapidity and low cost by sieve printing. The process of sieve printing and the ink composition as a precursor to the catalyst layer were developed, which allow the preparation of electrodes for MEAs fabrication with the implementation of the exact catalyst loading, 0.6 milligrams of platinum per square centimeters (mgPt.cm-2) suitable for cathodes and 0.4 mgPt.cm-2 for anode in only one application step per electrode. The ink was developed, produced, characterized and used with similar characteristics to ink of sieve printing build for other applications. The MEAs produced had a performance of up to 712 mA.cm-2 by 600 mV to 25 cm2 MEA area. The MEA cost production for MEAs of 247.86 cm2, that can generate 1 kilowatt of energy was estimated to US$ 7,744.14 including cost of equipment, materials and labor.

Cette thèse s’inscrit dans une démarche d’optimisation des performances de la PEMFC, à travers le développement d’outils de diagnostic dédiés à la gestion de l’eau. Celle-ci est un des éléments clé de l’augmentation des performances et de la durée de vie des piles à combustible. La technique du bruit électrochimique a été utilisée pour détecter des phénomènes liés au comportement de l’eau lors du fonctionnement de la pile dans différentes conditions opératoires (humidité relative des gaz, température et courant). Le bruit électrochimique enregistré pendant ces expériences a été associé à des mécanismes source grâce à une démarche expérimentale appropriée et à un traitement de signal basé sur l’analyse fréquentielle et temporelle. Les résultats de l’analyse spectrale et statistique ont permis de décrire l’état de santé de la pile à combustible (assèchement et noyage), par la génération des descripteurs dans le domaine fréquentiel et temporel, signatures de chaque état de fonctionnement de la pile à combustible. Enfin, de manière à compléter l’approche effectuée sur la détection de l’assèchement de la membrane et le noyage de la cellule par l’étude des bruits électrochimiques, des capteurs de mesure de pression en amont de la cellule de PEMFC ont été utilisées et comparer au signaux de tension pour détecter le phénomène de noyage d’une mono-cellule de PEMFC. / This thesis deals to optimize the performance of the PEMFC, through the development of diagnostic tools dedicated to water management. This water management is one of the key elements to increase the performance and life of fuel cells. The electrochemical noise technique was used to detect the phenomena related to the water behavior during the operation of the cell, under different operating conditions (relative humidity of the gases, temperature and current). The electrochemical noise recorded during these experiments was associated with mechanisms through an appropriate experimental approach and signal processing based on frequency and time analysis. The results of spectral and statistical analysis made it possible to describe the state of health of fuel cell (drying and flooding), by the generation of the descriptors in the frequency and time domain, signatures of each operating state of the fuel cell. Finally, in order to complete, the approach carried out on detecting the drying of the membrane and flooding of the cell by electrochemical noise analysis, the pressure measurement was combined with voltage signals, to detect the flooding phenomena of fuel cell.

La durabilité des piles à combustible à membrane échangeuse de protons, jugée à l’heure actuelle insuffisante, est principalement liée à la dégradation des supports de catalyseur cathodiques carbonés. Afin d’augmenter cette durabilité, un fin revêtement de dioxyde d’étain est effectué sur différents types de carbone (nanotubes, noirs et aérogel de carbone). L’objectif est d’obtenir un matériau alliant la morphologie et la conductivité électrique des carbones avec la stabilité thermodynamique du dioxyde d’étain. Plusieurs types de carbone avec des propriétés intrinsèques différentes ont été choisis afin d’avoir un large un panel de textures et de structures. Dans ces travaux de thèse, des études expérimentales ont été effectuées dans le but d’obtenir des revêtements d’oxyde d’étain fins, homogènes et couvrants. Ces études ont mis en l’avant l’influence primordiale de la texture et de la structure du carbone, mais aussi de la valeur du pH du milieu réactionnel sur la qualité et la quantité de revêtement. Il a aussi été montré que les mécanismes de transformation du précurseur en dioxyde d’étain dépendent de la valeur de ce pH. En améliorant les interactions entre la surface des carbones et des espèces réactives, il a été possible dans certains cas de diminuer grandement la quantité de précurseur tout en améliorant la qualité du revêtement. Les dépôts de nanoparticules de platine effectués sur des aérogels de carbone bruts et revêtus ont mis en avant un comportement différent du platine qui a tendance à s’agglomérer lorsqu’il est en contact avec le dioxyde d’étain. Les performances initiales ainsi que la durabilité des électrocatalyseurs en fonction de deux tests de vieillissement accéléré (classique ou démarrage/arrêt) ont ensuite été discutées, mettant en avant des résultats mitigés. / The proton exchange membrane fuel cell’s lifespan is insufficient because of the degradation of carbon used as cathodic catalyst supports. In order to reduce this degradation, a thin tin dioxide coating is synthesized on the surface of different carbonaceous materials (nanotubes, carbon blacks and aerogel). The aim is to combine the morphology and the electric conductivity of the carbon with the thermodynamic stability of the tin dioxide. Carbonaceous materials with different intrinsic properties are chosen for this study to test a wide range of textures and structures. Experimental studies were carried out in order to synthesize a thin, homogeneous and covering tin dioxide coating. The major influences of the texture and structure of carbonaceous materials but also the influence of the pH value on the quantity and quality of the coating are highlighted. It turns out that the mechanism of formation of tin dioxide depends on this pH value. Thanks to the improvement of the interactions between the carbon surface and the reactive species, it was possible, in some cases, to reduce drastically the quantity of precursor. Platinum nanoparticles deposition performed on various materials (raw or coated carbon aerogel) highlights a different platinum behavior. In fact, on the tin dioxide surface, nanoparticles tend to agglomerate together instead of making a homogeneous dispersion. Then, the initial performances and the durability of electrocatalysts tested with two accelerated stress tests (load protocol or start/stop protocol) are evaluated, spotlighting mitigate results.

Orientador: Luis Alberto Martínez Riascos. / Dissertação (mestrado) - Universidade Federal do ABC. Programa de Pós-graduação em Energia.

Le développement des Piles à Combustible (PàC) est en plein essor dans le contexte de transition énergétique mondial. La production d'énergie électrique par les PàCs possède l'atout majeur de ne rejeter que de l'eau et de la chaleur, sans émission de gaz à effet de serre. Pour un développement et une commercialisation plus large des PàCs comme générateurs d'énergie, leur fiabilité et leur durée de vie. Cette thèse est dédiée au Contrôle Tolérant aux Fautes appliquée à la gestion de l'eau dans les Piles à Combustibles de type PEM. Une gestion appropriée de l'état d'hydratation de la PEMFC contribue à éviter les dégradations irréversibles de ses composants et ses performances, et par conséquent à un allongement de sa durée de vie et une augmentation de sa fiabilité. La stratégie de Contrôle Tolérant aux Fautes Actif proposée est constituée d'une méthode de diagnostic basée modèle pour la détection et l'isolation des fautes et d'un contrôleur PID auto-adaptatif régulant la surstoechiométrie en oxygène. Il est à noter que cette stratégie a été validée en ligne sur un système PàC réel. / Fuel Cell (FC) development is expending due to global energy transition. Power generation using FC results in water and heat as by-products, without emission of greenhouse gases. To continue developing and expanding its use as power generators, FC lifetime and reliability have to be enhanced. This thesis work is dedicated to Fault Tolerant Control System (FTCS) applied to water management in PEM Fuel Cells. An appropriate water management of FC allow to avoid irreversible degradations of FC components and performance that lead to an improvement of FC reliability and lifetime.The proposed FTCS is composed of a model-based diagnosis method applied to fault detection and isolation, and a self-tuning PID strategy for oxygen excess ratio control. This strategy is tested and validated on-line on a real FC system.