Scanning tunneling microscopy at high pressure – an approach to bridge the pressure gap in catalysis research

Most experiments on the surface processes in heterogeneous catalysis are conducted in ultra high vacuum (UHV), while catalytic processes in an industrial environment take place at pressures which are typically ten orders of magnitude higher. The use of UHV in the experiments helps to achieve sample cleanliness and is a prerequisite for the application of most surface science methods. This approach has been very successful in clarifying the basic steps of surface reactions, but the discrepancy in the experimental conditions raises the question in how far UHV-results can be extrapolated to high pressures. This problem is referred to as the pressure gap. Scanning tunneling microscopy (STM) has been successfully used in UHV to obtain information about surface reactions on an atomic scale. Since the tunneling effect is not limited to vacuum conditions, this technique is, in principle, a tool for bridging the pressure gap. This thesis describes the setup of a new experiment in which a scanning tunneling microscope (STM) is integrated in a small chamber for variable pressure from UHV to 1 bar. This STM-chamber serves as a reactor cell and is coupled to a conventional UHV-chamber equipped with standard surface science techniques for sample preparation and characterization. Thus, well-defined single crystal surfaces can be studied over a wide pressure range in situ and with atomic resolution. The new experimental setup opens the possibility to investigate high pressure phenomena relevant to catalytic processes such as modifications of the surface morphology, or changes in the reaction mechanism. The first system studied was the interaction of oxygen with ruthenium. The experiments were motivated by the fact that ruthenium exhibits a strong pressure gap effect as a catalyst for the CO-oxidation, with negligible activity in UHV but high activity at high pressures. Based on recent UHV-experiments, it was suggested that this behavior may be linked to the formation of a RuO2(110)-film which acts as the actual catalyst. With the new STM, the adsorbate structures were imaged both on the clean and oxidized Ru(0001) surface at ambient oxygen pressure and room temperature. On the pristine Ru(0001) surface, a O(1x1) structure was observed which corresponds to a coverage twice as high as the saturation coverage that can be achieved in UHV at room temperature. The RuO2(110)-surface was prepared in UHV and then also studied at high oxygen pressures. For the first time, theoretical predictions for the surface termination of the oxide film could be tested in situ. Surprisingly, the predicted high coverage oxygen phase was not observed. Instead, an ordered c(2x2)/(2x1)-phase forms which is stable in a CO atmosphere and thus inactive for the oxidation of CO. Thermodesorption spectra suggest that this new structure is formed by a carbon containing species (possibly carbonate) that accumulates on the surface. Obviously, side reactions play a major role on the RuO2(110) surface. This aspect may help to explain anomalies found in kinetic measurements at high pressures, i.e. the steady decline of the catalytic activity of ruthenium from an initially high value which is reported in literature