Model Studies of Catalytic Processes at the Solid/Liquid Interface

This thesis focuses on the fundamental understanding of different catalytic reactions at the interface between a solid catalyst and a liquid reactant phase. To this end, both in situ and operando experiments, monitored by synchrotron X-ray diffraction techniques as well as infrared spectroscopy based surface science model studies, were performed. The electrochemical surface oxidation of Pt(111) and Pt(100) electrodes under operando conditions in a potential range up to 2.25 VRHE has been investigated in a combined EC/SXRD study using a rotating disk electrode setup. The as-prepared Pt(111) surface is described best by an ideally bulk-terminated model. For Pt(100), a bulk-terminated surface with 0.17 ML of adatoms, created by lifting of a pseudohexagonal reconstruction layer, has been observed. Both surface structures are stable in O2-purged 0.1 M HClO4 for potentials between 0.0 VRHE and 1.0 VRHE, as no significant changes of the structure factors have been observed in this potential range. Starting at 1.05 VRHE, the Pt(111) surface becomes rough due to surface oxidation, which follows the well-known PE mechanism. Above 1.3 VRHE, the newly proposed PE+buckling model was used to describe the extensive surface oxidation. Increasing the potential beyond 1.55 VRHE leads to a recovery of the structure factor in the anti-Bragg positions, partially surpassing the initial SFs of the pristine surface. This SF recovery is, however, not due to a restoration of the initial surface structure, but due to a shift of the reference point of the surface layer to former deeper bulk layers. As the oxidation proceeds, the measured CTRs arise from a former deeper Pt layer, which is now the topmost, not fully oxidized platinum layer. This leads to the formation of a disordered bulk oxide at high positive potentials. The additional oxygen, present in the electrolyte, has no influence on the surface structures of the Pt(111) electrode between 0.05-2.25 VRHE. Qualitatively similar results were obtained for the surface oxidation of Pt(100) in an oxygen-purged electrolyte. Here again, an undulating trend of the structure factor in an anti-Bragg position as a function of the applied potential has been observed, but the structure factors remained below their initial value. This is associated with the formation of a rougher interface between bulk Pt electrode and disordered oxide layer, due to lateral movements of the lifted Pt atoms.A second study addressed the formation and stability of palladium carbide phases of well-defined Pd/α-Al2O3(0001) model catalysts under the conditions of catalytic liquid organic hydrogen carrier dehydrogenation. The phase composition of supported Pd nanoparticle catalysts was monitored as a function of particle size and reactant flow rate and composition in an in situ high-energy grazing incidence X-ray diffraction study. At 500 K, exposure of small Pd nanoparticles with an average particle diameter of 6 nm to methylcyclohexane leads to the immediate formation of a palladium carbide phase. The carbon content in PdxC gradually increases, until a Pd6C phase is yielded. It was shown that the stability of these carbide phases critically depends on the composition of the reactant feed and the resulting balance between carbon formation and consumption. Surface carbon species are formed by the thermal decomposition of hydrocarbons, whereas they are removed from the surface by reactions with gas phase hydrogen. At a high gas flow rate, which corresponds to a low H2 partial pressure over the catalyst, the Pd6C phase immediately decomposes after its formation. This decomposition is triggered by the nucleation and growth of graphene deposits on the surface of the Pd nanoparticles. At a low gas flow rate, which corresponds to a high H2 partial pressure over the catalyst, the Pd6C phase is stable under continuous MCH exposure. Due to the higher H2 concentration in the gas phase, the surface carbon concentration is not high enough for the nucleation of a graphene nucleus of critical size. In the absence of the carbon source from the gas phase, Pd6C decomposes due to carbon segregation to the surface of the Pd nanoparticles. Upon exposure of the larger Pd nanoparticle sample (15 nm average diameter) to MCH at the low gas flow rate, the coexistence of Pd6C and PdxC carbide phases was observed. With increasing time on stream, the PdxC phase is converted to Pd6C. Upon extended exposure to MCH, the pure Pd phase recovers at the expense of both carbide phases. Because of slow nucleation, the graphene deposits are formed only after an induction period, which triggers the carbide decomposition.Finally, the earliest stages of the ionothermal synthesis of cobalt oxide nanoparticles in an IL were scrutinized in an UHV-based surface IRAS model study. The room temperature synthesis in an IL recently attracted attention as a promising synthesis route for the production of taskspecific metal oxide nanomaterials. Deposition of Co2(CO)8 and [C4C1Pyr][NTf2] at 110 K on Au(111) leads to layered films of molecular cobalt carbonyl complexes and IL. Several bridged and non-bridged precursor species (Co2(CO)8 isomers and Co4(CO)12) are identified based on their IR spectra. These species dynamically convert between bridged and non-bridged complexes. In particular, the interaction with the pristine Au(111) surface facilitates the conversion of CO-bridged complexes to cobalt carbonyls with exclusively terminal CO ligands. Such conversions were not observed for the deposition of Co2(CO)8 on pre-adsorbed [C4C1Pyr][NTf2], as the precursor/IL interaction is weak as compared to the precursor/metal interaction. The weak precursor/IL interaction leads to molecular desorption of precursor molecules from an IL layer upon heating. In strong contrast, the formation of cobalt nanoclusters on Au(111) has been observed at elevated temperatures for precursor layers in direct contact with the gold support. However, deposition of Co2(CO)8 on Au(111) or [C4C1Pyr][NTf2]/Au(111) at 225 K resulted in the formation of Co nanoclusters already during the deposition process. Precursor-terminated layers are readily oxidized by ozone at low temperatures, forming a disordered CoxOy passivation layer. This passivation layer protects the underlying precursor layers from further oxidation. A similar protection against oxidation is achieved by the deposition of an IL layer on top of pre-adsorbed Co2(CO)8 at low temperatures. At elevated temperatures, both the cobalt oxide passivation layer and the IL protection layer get successively permeable for ozone. This results in the full oxidation of the precursor multilayer. Notably, the CoxOy passivation layer, formed at the IL/precursor interface during the exposure of [C4C1Pyr][NTf2]/Co2(CO)8/Au(111) to ozone at elevated temperatures, only becomes ozone-permeable at significantly higher temperatures as compared to the CoxOy passivation layer formed for a pristine Co2(CO)8/Au(111) film. Thus, it is concluded that the IL governs the oxide formation, which leads to a more densely packed cobalt oxide passivation layer