Towards an accurate theoretical description of surface processes

Molecular modelling methods are indispensable for both helping to understand experimental results and to explore new materials. In this thesis we focus on theoretical methods that are used to study activated processes at surfaces as well as those which can account for van der Waals dispersion forces. To begin, we examine existing methods and develop a few new ones that are suitable for identifying transition states in chemical reactions. We discuss in detail how the various methods compare in efficiency for some simple chemical processes on an NaCl surface (water diffusion and HCl dissociation). The interaction of water with salt is then extended, focussing on the interaction of water clusters (up to 20 water molecules) with clean and defected salt surfaces. The aim of this part of the thesis is to understand in detail the initial stages of NaCl dissolution. In the remainder of the thesis we focus on a problem suffered by many current density functional theory methods, namely their inability to accurately account for dispersion forces. We test a recently proposed non-local functional and show how its accuracy can be dramatically improved with some simple modifications. The new functional(s) are tested on a wide variety of materials and highly encouraging results have been obtained.