Dans la technologie à l'hydrogène, l'oxydation préférentielle du CO en excès d'hydrogène (réaction PrOx) est un processus important pour l'obtention d'hydrogène sans CO pour les piles à combustible à membrane échangeuse de protons (PEMFC). Les catalyseurs à base de PtCu sont l’un des systèmes les plus étudiés pour les dispositifs mobiles en raison de leur bilan d’activité / sélectivité élevé et de leurs propriétés chimiques et mécaniques appropriées pour les procédures de démarrage / arrêt dans les conditions de fonctionnement des processeurs de combustible.Récemment, l'utilisation de catalyseurs bimétalliques Pt-Cu avec une activité et une sélectivité excellentes vis-à-vis de l'oxydation du CO a été rapportée pour la réaction PrOx. Cependant, la nature des phases actives et le rôle des deux métaux au cours de la réaction ne sont pas clairement démontrés.Pour comprendre ce système, il est nécessaire de créer un catalyseur modèle qui facilite l’étude. Ainsi, des nanoparticules d'alliage bimétallique Pt-Cu bien définies ont été synthétisées et étudiées par des techniques d'Operando permettant de comprendre les modifications électroniques de surface de l'interface solide-gaz du catalyseur modèle mentionné ci-dessus dans des conditions de réaction PrOx.Dans ce travail, la composition et la nature des espèces présentes à la surface du catalyseur dans des conditions de réaction bien contrôlées ont été étudiées, en particulier du point de vue de la dynamique de surface, des transitions structurelles et des effets possibles de l’atmosphère de réaction et des couches superposées adsorbées sur la surface. composition superficielle et structure de l'alliage. / In the Hydrogen technology, the preferential oxidation of CO in excess of hydrogen (PrOx reaction) is an important process for obtaining CO-free hydrogen for proton exchange membrane fuel cells (PEMFCs). PtCu based catalysts are one of the most studied systems for mobile devices because of their high activity/selectivity balance and their appropriate chemical and mechanical properties for the start-up/shut-down procedures during fuel processors operation conditions.Recently, the use of Pt-Cu bimetallic catalysts with excellent activity and selectivity towards CO oxidation was reported for PrOx reaction. However, there are not clear evidences off the nature of the active phases and the role of both metals during the reaction.To understand this system it is necessary to create a model catalyst which facilitates the study. Thus, well-defined Pt-Cu bimetallic alloy nanoparticles were synthetized and studied by Operando techniques allowing the comprehension of the surface electronic modifications in the solid-gas interface of the above mentioned model catalyst under PrOx reaction conditions.In this work, the composition and nature of the species present on the catalyst surface upon well-controlled reaction conditions were studied, in particular from the point of view of surface dynamics, structural transitions and the possible effects of reaction atmosphere and adsorbed overlayers on the surface composition and structure of the alloy.

Fuel cells are more efficient and cleaner than combustion engines. Ethanol as a fuel has a high energy density and is safer and easier to handle than hydrogen which is normally used in fuel cells. If efficient fuel cells on alcohol were available, they could be used for engines and power sources for electronics. Platinum-tin surfaces have proven to be good catalysts for ethanol and an improvement over pure platinum. The mechanism and the structure during catalysis are not well known. An experiment was performed at the Hippie beam line at Max IV to improve the knowledge in this area. The (111) surface of Pt and Pt3Sn alloy and the (223) surface of Pt, was exposed to ethanol and oxygen. Pt and Pt3Sn both have face-centered-cubic (FCC) crystal structures. The (111) surface is the most close-packed in an FCC crystal. A (223) surface is a (111) surface cut at a low angle. So it has the appearance of a stepped (111) surface. The edges on the (223) surface should increase the activity compared to the (111) surface. The surfaces and the gas phases were measured in situ with ambient pressure x-ray photoelectron spectroscopy and a quadrupole mass spectrometer was used to analyze the gas composition. The hypothesis that increasing the number of edges as with the Pt(223) surface should increase the activity is accurate. Pt(223) was more active than Pt(111). Pt(223) and Pt3Sn(111) have similar ethanol conversion rate. Increasing the oxygen-to-ethanol ratio increased the activity both with Pt(111) and Pt(223), Pt3Sn(111) was not tested with increased oxygen-to-ethanol ratio. The gas phases were analyzed, and the existing compounds were identified. Acetaldehyde shows up in the C1s gas spectrum in all of the sequences. When ethanol decreases acetaldehyde increase. The difference between these two compounds is only two hydrogen atoms. This reaction is the start of the catalytic process and it is the same for all tested crystals. Ethylene (CH2CH2) shows up as a vague peak in the gas phase. It is only present at higher temperatures and with a low oxygen rate. Compared to the other crystals the Pt3Sn(111) sample doesn't produce CO2, at least not to a detectable degree. In the gas phases of the other crystals, the CO2 peak was visible. Pt(223) creates CO2 but to a lesser degree than Pt(111). The goal of the experiment was to investigate which Sn phases are present during ethanol oxidation. This turned out to be difficult. The Pt3Sn crystal was carbon poisoned during the first test sequence and the graphite layer was not possible to remove during the beam time. Curve fitting of the Sn3d peak resulted in two components. The components were Pt3Sn alloy and Sn with adsorbed molecules. The expected SnO2 and SnO peaks notably absent. The oxygen probably bonds with carbon instead of tin. Carbon was present on the surface due to insufficient cleaning. In the oxygen spectrum, chemically bonded oxygen seems to be present from 100 °C, as SnO2 or SnO. This peak is most likely from some other component containing oxygen. If oxygen is bonded to Sn, it should be visible in the Sn3d peak, unless it is hiding underneath one of the present peaks. According to Batzill et al. a quasimetalic state consisting of oxidized Sn alloyed with Pt has a similar binding energy as Pt3Sn alloy. So it could be that the oxygen is hiding underneath the Pt3Sn alloy component. The experiment has improved the knowledge of ethanol oxidation on platinum and platinum-tin surfaces. The knowledge gained here is a good start for further experiments and simulations.