Oxidation and Reconstructions of Alloy Model Catalysts : Platinum-Rhodium and Palladium-Gold

Abstract: Alloys and bimetallic systems have shown many advantages in heterogeneous catalysis, but the details of their catalytic functions are not always known. For instance, the atomic details about surface reconstructions under catalytic reaction conditions, or which phases that are high-active or low-active. Furthermore, the model catalysts studied in laboratories often differ a lot from the industrial catalysts regarding working conditions and materials.Surface reconstructions depend on surface science of both physical and chemical nature. Since catalytic reactions occur on the surface of catalysts, it is desirable to understand what happens at the atomic scale on the catalyst surface under reaction conditions. There are different techniques to study surfaces at the atomic level, regarding elemental composition as well as surface structure.In this thesis, I have studied the surfaces of two alloy systems, PtRh and PdAu, under reaction conditions. CO oxidation over a Pt25Rh75(100) single crystal was studied at a commissioning beamtime as a model system using HESXRD (Paper I and Paper II). Thin films of PdAu on a sapphire substrate, with varying Pd:Au ratios, has been oxidised, and the oxide formation and alloying process has been studied with HESXRD (Paper IV). Oxidation with subsequent reduction in methane has also been studied with APXPS (Paper V). I have also analysed the results from LEED, TPD, and TSD of CO oxidation over a Pd(100) single crystal (Paper III).Two well-known oxygen structures were observed on the PtRh crystal: a (3 × 1) reconstruction with chemisorbed O and a c(8 × 2) surface oxide. Furthermore, an unexpected c(2 × 2) structure was observed under reducing conditions and elevated temperature. The quantitative analysis of these structures is presented in this thesis.The PdAu oxidation experiment was a continuation of a methane oxidation study on a Pd(100) single crystal, showing that the catalytic activity decreases if the oxide grows too thick, but increases again upon oxide decomposition. It is believed that the thick oxide exposes the low-active (100) surface orientation while the thin oxide exposes the high-active (101) surface orientation. The idea with PdAu was to limit the amount of Pd that can be oxidised in order to keep the oxide thin and high-active. The changed lattice constant of PdAu compared to Pd,however, seems to stabilsee the low-active PdO surface. We believe this can be helped by instead alloying Pd with Pt, or use PdAu samples with lower Au concentration.