hide
Free keywords:
-
Abstract:
The adsorption of a particle or molecule from the gas-phase is what is typically referred to as the first step in heterogeneously catalyzed processes. The adsorbed particles are activated for the subsequent chemical reaction by the chemical bond being formed between the particle and the specific adsorption site on the surface. How reactive an adsorbed particle will be is strongly influenced by the strength of this bond to the surface. A strong bonding to the surface can lead to a low reactivity. In case of a weak bond, it is possible that the particle desorbs again before a reaction could take place. This bond strength is not only influenced by
the formation energy of the chemical bond to the surface but also by the interactions between neighboring particles.
These interactions can have either a stabilizing or destabilizing effect. Their importance increases the more densely packed the adsorbates are on the surface. The higher the coverage, the more important the impact of these interactions. Ordering behavior or even the promotion of certain cluster or island structures are two examples for properties which are
steered by these interactions on a microscopic level. Quantifying these effects on a microscopic scale is not always an easy task.
It is, therefore, of great importance for model descriptions which are used in computer simulations for these kind of systems to properly describe the interactions of the adsorbates in the plane parallel to the surface — the lateral interactions. A major motivation for the
development of such model descriptions is presented by the fact that even nowadays supercomputers are not powerful enough to simulate, on the level of electronic structure theory, system sizes large enough for a realistic description.
An ideal model would be presented by one whose parameters are derived from a set of
calculations on the level of electronic structure theory. In addition, the derived parameters of this model should be chosen such that even in the mesoscopic model all parameters would retain their microscopic meaning. This is the major topic of the thesis presented here. This work presents an important contribution because it takes a systematic look at this important interplay between different kinds of interaction and derives appropriate model descriptions for the simulation of large scale systems. Starting from a simple qualitative model for the description of the interplay between lateral interactions and chemisorption, two limiting cases will be discussed.
In the first case a strong bond to the specific adsorption site is established and neighboring particles interact only weakly. This scenario will be investigated throughout this thesis by means of a cluster expansion (CE) approach. Based on calculations on the level of density-functional theory a parameterization of interaction parameters will be presented.
In this model description lateral interaction parameters are identified with a certain geometrical configuration of neighboring particles adsorbed in surface adsorption sites. Within the parameterization process a certain energy contribution to the total binding energy of a particle will be assigned to this specific adsorption configuration. By means of this parameterization procedure it is ensured that all interaction parameters in fact resemble a microscopic arrangement of particles. This approach will be applied to a classical example system of surface science. It will be used to describe the adsorption of carbon monoxide (CO) on a palladium (100) surface. As a result of this application a closed set of interaction parameters
will be derived which is in very good agreement with previously published values of certain interaction parameters. These previously published values, in contrast, had been derived by a fitting process to experimental data. Limitations of this description will be discussed for the specific example system and their origins will be explained.
The second case presents an example of strongly interacting adsorbates. The binding to the surface is either rather weak or at most of the same order of magnitude as the particle interactions. In such a scenario usually a non site specific adsorption is observed. A typical example is presented by a system where a metal is deposited on a surface of another metal. In the work presented here, the example system is presented by small copper islands on a
silver surface. Motivated by a set of experimental observations for which an explanation was desired, this system was investigated by a theoretical approach. Based on simulations on the level of electronic structure theory a semi-empirical potential description of the example
system was validated. This potential was then used for large scale simulations accessing the experimental system sizes. The insight gained by our simulations allowed us to develop a model which is able to explain the driving force behind the experimentally observed reconstructions.
We showed that the driving force is originated in the lateral interactions of island
atoms and the strain induced by the large lattice mismatch of silver and copper. Our results were able to initiate new experiments. In addition, based on our simulations, the required experimental parameters could be suggested.
