Development of Computational Methodologies for the Study of Metal Complex Structures.

This thesis aims to extend the range and sophistication of computational methods available for the study of the structures of metal coordination complexes. To this end, a number of different approaches have been adopted. Molecular mechanics force field calculations have been successfully employed to investigate chemical and structural problems in a range of coordination complexes. New force field parameters are presented for Uranium(IV), Palladium(II) and Tungsten complexes, along with modified Cobalt(II) parameters for Schiff-base complexes. Widespread application of molecular mechanics is, however, shown to be inhibited as a result of the complex and time consuming parameterisation process, and the lack of available structural data. The Cambridge Structural Database is the major repository of structural data for inorganic complexes, but the way in which it is stored makes it difficult to extract the desired information. A new data format is proposed specifically for metal coordination complexes, whereby the metal forms a focal point with information separated into ligand sub-sets. The application of the new data format is demonstrated using a number of examples with particular relevance to addressing problems in force field parameterisation. With the technological applications of coordination compounds often manifested in their solid-state structures, there is a need to understand and predict crystal structure based only on a knowledge of the molecular unit. Several methods of investigating crystal structure, namely shape and structure similarity and packing efficiency, have been investigated with a view to using these parameters in a predictive manner