Adsorbate Interactions and Growth Morphology
The theory of growth has transformed substantially during the past decade. Whereas early theoretical methods were largely empirical, phenomenological, and continuum-based, today's tools include a quantitative, atomistic methodology on a first-principles basis. The ultimate length scales accessible to growth experimentalists have shrunk considerably, so that today atomic-scale resolution is routinely achieved with the scanning-tunneling microscope (STM). The time and length scales of predictive growth simulations have increased immensely, so that nowadays layer-by-layer growth of crystals can be monitored in accurate, atomistic computer experiments. Hence, experimental and theoretical methods can describe growth processes on the same level, allowing a detailed comparison between the two and providing the breeding ground for rapid scientific progress.
This thesis is a manifestation of this exciting development by efforts to extend the methodologies to a few significant and well-characterized cases. In turn, these extensions ramify to applications in other realms of science.
A kinetic Monte Carlo (KMC) study of triangular island growth on close-packed metal surfaces identifies anisotropic corner energetics as the origin of the triangular shape and orientation. It also demonstrates the value of sticking to first-principles data for activation energies when simulating island morphology. While semi-empirical parameters are sometimes the only option available as input to KMC simulations, the results of such simulations should be evaluated with caution.
It is shown how the failure to reconcile STM island-density data for weakly corrugated systems with the standard nucleation theory led to the exploration of adsorbate-interaction effects in epitaxial growth. When interaction energies calculated from first principles are properly accounted for, KMC-simulated island densities are consistent with experimental data.
In order to make mean-field-nucleation theory (MFNT) applicable also to weakly corrugated systems, a nonlocal mean-field nucleation theory is developed that takes into account the effects of adsorbate interactions on nucleation probabilities.
A generalization of the first-principles KMC methodology to more complex systems is well underway. Here it is applied to the catalytic oxidation of NO on Pt(111). The character of the reaction goes from endothermic at low O coverage to exothermic at high coverage, driven by repulsive O-O and O-NO interactions.