Desorption of Adsorbate Layers on rystal Surfaces

Research project P 14628 Desorption of Adsorbate Layers on Crystal Surfaces Michael HOHAGE 09.10.2000 This project aims at a better understanding of the behaviour of adsorbate layers on crystal surfaces. The analysis will be based on the promising combination of thermal desorption spectroscopy (TDS) with a kinetic Monte-Carlo (KMC) simulation, specially designed for this purpose. First results of this combination have already proven the high efficiency of this procedure. It is planned to identify the basic properties of the adsorbate layers as well as to analyze the underlying atomic processes in a quantitative manner, by the investigation of the desorption behaviour. In particular, the extraction of detailed information about finite size effects and kinetic limitations from the desorption spectra is subject of our studies. We intend to investigate a broad spectrum of adsorbate systems. We will chose adsorbates which differ in the type and strength of their interaction with the substrate, namely physisorbates like Xe or CF4 in contrast to chemisorbates such as atomic oxygen or hydrogen. In adition, substrates with fundamentally different properties will be used. We may expect a substantially different adsorption and desorption behaviour, if instead of flat substrates, vicinal or nanostructured template surfaces (as, for instance, the Cu-CuO stripe phase) are used. This project is placed into a larger context through ist relation with FWF-projects (P12317-NAW and P 13 841 - CHE). The combination of the main experimental methods of each project, Scanning Tunneling Microscopy (STM), Thermal Energy He Atom Scattering (TEAS), and the TDS (in the present project), should allow us to reach a deeper understanding of the atomic processes active during adsorption and desorption of adsorbates on crystal surfases. n of adsorbates on crystal surfases.

The understanding of the behavior of adsorbates on surfaces is a crucial prerequisite to gain insights in more complex processes like catalytic reactions. Accordingly, this project aimed at the analysis of adsorbate layers on crystal surfaces. We used a combination of temperature programmed desorption (TPD) and kinetic Monte-Carlo (KMC) simulation to study such systems. The application of this combination was, indeed, successful in linking the observed desorption behavior directly to processes on the atomic scale. After developing a novel KMC scheme optimized to deal with desorption phenomena, we have proven its ability to reproduce well understood desorption phenomena. Thereafter, we applied this simulation to more challenging adsorbate systems. Continuously, the KMC code has been further improved to deal with problems like long-range and multi-body interactions. The standard analysis of desorption spectra usually fails to determine atomic interaction energies for adsorbates with small lateral interactions. This difficulty originates from a phase transition from condensed adsorbate islands to a 2D adsorbate gas during desorption. On the contrary, the KMC simulation is perfectly suited to analyze such systems and to link their behavior to the underlying atomic processes. For Xe on Pt(111) and N 2 on Cu(110)(21)O we quantified the atomic interaction parameters which are responsible for their interesting behavior. Recently it turned out that also metallic layers show similar effects during desorption underlining the general importance of these studies. Adsorbates may act considerably differently on stepped surfaces like Pt(997) than on their barely stepped counterparts. Surprisingly, the Xe adsorption on Pt(997) takes place row-by-row. To explain this adsorption scenario various models have been proposed in literature. The KMC analysis allowed us to identify the model which is most suitable to explain the experimental observations. Furthermore, by assuming a small lateral Xe-Xe interaction energy the KMC simulation has reproduced the observed desorption spectra quantitatively. Another important phenomenon is the influence of the finite size of substrate structures on the adsorbate. By studying the desorption of CF4 on the nano-structured Cu-CuO stripe phase, we identified a considerable finite size effect: The desorption energy is strongly dependent on the width of the CuO stripe. By applying the KMC simulation we identified the origin of this energy shift as the change in the shape of the CF4 islands located on the CuO stripes. The width of the islands is obviously limited to the width of the CuO stripes. The success to study such desorption phenomena has shown that the combination of TPD and KMC simulation is the method of choice to gain deep insights into the interplay of atomic processes which determines the macroscopic behavior of the adsorbates.