We describe an experimental setup for studying gas adsorption and chemical surface reactions by scanning tunneling microscopy (STM) at gas pressures ranging from ultrahigh vacuum (UHV) to 1 bar. The apparatus is designed for experiments to bridge the so-called pressure gap in catalysis research by obtaining atomic scale information about catalytic reaction mechanisms under steady-state conditions. It combines a UHV chamber for sample preparation and post-reaction surface analysis with a small high-pressure cell (volume 1.5 l) which contains the STM. Several concepts to improve the variable-pressure performance of existing high-pressure STM designs are described. These include access to the entire pressure range between UHV and 1 bar without triggering gas discharges, the potential for high-speed scanning and for variable temperature measurements. The design also features a fast transfer mechanism from the reactor to UHV, thus allowing for fast analysis of surface species after high-pressure experiments. First results with atomic resolution were obtained on a Ru(0001) surface at high oxygen pressures. The images show an O(1x1) adsorbate layer not observed in UHV experiments at room temperature

To test predictions about the activity of Ru catalysts the RuO 2 (110) surface was investigated in an oxygen atmosphere at ambient pressure using scanning tunneling microscopy (STM). Epitaxial RuO 2 (110) films were grown on a Ru(0001) sample following an established preparation technique from ultrahigh vacuum (UHV) investigations. The sample was then exposed to 200 mbar of O 2 at 300 K, and STM images were taken during exposure. The mesoscopic morphology of the film and the row structure of the RuO 2 (110) surface known from UHV were preserved. However, a 2-fold periodicity was observed along the [001] rows which is inconsistent with the expected surface termination by O atoms bonded to the coordinatively unsaturated sites of the RuO 2 (110) surface. In addition, a second type of features that partially form clusters within the ordered surface was observed. In a pure CO atmosphere at pressures of up to 21 mbar no atomic changes of this structure were observed, meaning that it does not contain O species that can react with CO. The new surface phase was stable after removal of the O 2 atmosphere, so that it could be further characterized in UHV. Thermodesorption spectra showed strong desorption of CO 2 with peaks at 520 and 570 K but not the expected recombinative desorption of O atoms from the coordinatively unsaturated sites. Photoelectron spectroscopy showed an O 1s state at 531.0 eV in addition to the bulk oxygen state of the RuO 2 film at 529.5 eV. The most likely interpretation of the surface species in the oxygen atmosphere is a strongly bound carbonate formed by reaction of the surface with traces of CO or CO 2 in the O 2 atmosphere. The carbonate passivates the surface, leading to complete catalytic deactivation at 300 K. It is concluded that the established model for the unusual activity of Ru catalysts, which is based on the unique chemical properties of the RuO 2 (110) surface, cannot be extrapolated to ambient conditions for temperatures below the decomposition temperature of the carbonate species.

To identify surface phases that could play a role for the epoxidation of ethylene on Ag catalysts we have studied the interaction of Ag(111) with O(2) at elevated pressures. Experiments were performed using high-pressure scanning tunneling microscopy (STM) at temperatures between 450 and 480 K and O(2) pressures in the mbar range. Below p(O(2)) approximately 1 mbar the surface largely showed the structure of bare Ag(111). At p(O(2)) above approximately 1 mbar the (4 x 4)O structure and the closely related (4 x 5 radical 3)rect structure were observed. The findings confirm theoretical predictions that the (4 x 4)O structure is thermodynamically stable at the oxygen partial pressure of the industrial ethylene oxide synthesis. However, in other experiments only a rough, disordered structure was observed. The difference is caused by the chemical state of the STM cell that depends on the pretreatment and on previous experiments. The surface was further analyzed by X-ray photoelectron spectroscopy (XPS). Although these measurements were performed after sample transfer to ultra-high vacuum (UHV), so that the surface composition was modified, the two surface states could still be identified by the presence of carbonate or a carbonaceous species, and by the absence or presence of a high-binding energy oxygen species, respectively. It turns out that the (4 x 4)O structure only forms under extremely clean conditions, indicating that the (4 x 4)O phase and similar oxygen-induced reconstructions of the Ag(111) surface are chemically unstable. Chemical reactions at the inner surfaces of the STM cell also complicate the detection of the catalytic formation of ethylene oxide.