Biocatalytic methods of synthesis are becoming increasingly important in industry. Using enzymes as catalysts allows highly selective reactions to be performed under milder physical conditions and in a more environmentally benign fashion than most corresponding chemical catalysts.

Enzymes have in general evolved to perform one type of reaction on a limited set of molecules, and hence there is often a need to alter the specificity of an enzyme to suit a desired process. Understanding the details of enzymatic catalysis at a quantum mechanical level enables the intelligent redesign of these macromolecules. For this purpose, density functional theory (DFT) has been shown to epitomise a suitable balance of accuracy and computational cost. Thus, this thesis describes the quantum chemical rationalisation of the reaction mechanism and sources of selectivity of the bacterial alcohol dehydrogenase TbSADH – an enzyme highly suited to modification for industrial processes.

ADHs catalyse reversibly the interconversion of alcohols and ketones or aldehydes. Herein, the general ADH reaction mechanism was shown to be viable for this enzyme. In addition, the experimental enantiopreference of the enzyme was reproduced, and thus the reversal of selectivity seen with the slight increase in substrate size was captured. The main determinant of selectivity was found to be a fine balance of repulsive steric interactions and attractive dispersion effects between the substrate and the hydrophobic binding pockets. The ability of the modelling methodology to capture effects such as these represents further evidence of its usefulness as a complement to experimental work in designing the biocatalysts of the future.

The development of protocols to allow quantum mechanical investigation of the production of large and industrially interesting axially chiral alcohols is also presented. The work described has showed that quantum chemical models of many hundreds of atoms are now within our grasp, and although they were unable to correctly describe the selectivity for the large 4-(bromomethylene)cyclohexan-1-one in TbSADH, the protocols devised can be very useful for future investigations of enzymatic catalysis.