Adsorbate Induced Reconstructions of Metal Surfaces

In this dissertation I present work on the modeling of adsorbate-metal interactions with a specific focus on carbon monoxide (CO) induced restructuring of platinum stepped surfaces. Additionally, I will present work on the development of a multiple-minima fluctuating charge (mm-FlucQ) potential which has been used to simulate charge responsive platinum surfaces.

New Pt-CO and Au-CO forcefields were developed to study coverage-dependent restructuring of high-index platinum & gold (557) surfaces. It was observed that the weak Au-CO binding led to minimal disruption of the surface, whereas the strong Pt-CO interactions resulted in significant disruption of the step-edges which led to increased step-wandering. Specifically, the strong CO-CO quadrupolar repulsion caused an increase in adatom mobility. This increased mobility eventually led to large-scale surface reconstructions including the formation of a metastable double layer.

A platinum/palladium (Pt/Pd) (557) surface was simulated to explore CO-induced reconstructions on a more complicated bimetallic surface. The Pt-CO forcefield was retuned while a new Pd-CO forcefield was developed. The difference in binding strengths was found to play an important role in the disruption of the Pt/Pd surface while the preferred binding sites on each system (Pt: atop, Pd: bridge/hollow) led to significantly different behaviors with regards to surface diffusion and mobility.

The importance of step-edge energetics for the formation of adatoms encouraged us to directly examine straight edged (557) & (112) and kinked edged (765) & (321) platinum surfaces. As expected, the systems with rougher edges experienced a greater amount of surface diffusion and a concomitantly increased amount of step-wandering. The length of the (111) plateaus between step edges also played an important role with regard to the extent of surface reconstruction observed.

Accurate treatment of the electrostatic interactions between adsorbates and surfaces is often neglected in molecular dynamics simulations because of the increased computational cost and difficulty of implementation. Our new mm-FlucQ potential allows us to more properly describe a metal surface’s response to impinging charged species in a dynamic fashion by allowing the charges on each atomic site to fluctuate in response to the local electronic environment.