Asymmetric reductions are used to produce chiral molecules, which are important precursors for the pharmaceutical industry. Bio catalytic reductions often display high enantioselectivity without the cost and toxicity associated with metal catalysis. However, unlike metal catalysts which use H₂ directly, many useful redox-enzymes require the hydride donor NADH. NADH is expensive; therefore for a bio-catalytic process to be viable it must be recycled, usually by using a sacrificial carbon based substrate, generating super-stoichiometric amounts of waste. Two different methods for H₂-driven NADH recycling are explored in this project: using soluble hydrogenases (SH) and, carbon particles modified with a hydrogenase and an NAD⁺-reductase moiety. The conductive carbon particles allow electrons from H₂-oxidation to be channelled from the hydrogenase to the NAD⁺ reductase for reduction of NAD⁺. This project focuses on four main areas. The first looks at using the enzyme-modified particles for the production of high value chiral amines. A yield of >98% was achieved using the enzyme-modified particles with an L alanine dehydrogenase for H₂ driven conversion of pyruvate to L-alanine. Moreover, a faster rate of reaction was demonstrated with the L-alanine dehydrogenase immobilised on particles versus with the L-alanine dehydrogenase in solution. The second section focuses on elevated temperature NADH recycling: an SH and an NAD⁺-reductase from a thermophilic organism were characterised. The NAD+-reductase was subsequently used as part of a system for recycling NADH at >35 °C. When demonstrated in combination with an enoate-reductase a 62 % yield was obtained for the reduction of 2 methyl 2 cyclopentenone. In the third strand SHs and enzyme-modified particles were investigated as recycling systems for NADH analogues. In summary, this thesis expands the scope for application of H₂-driven biocatalytic reduction reactions.

The work in this Thesis addresses the challenges associated with using redox enzymes for chemical synthesis. The use of enzymes as catalysts in the synthesis of fine chemicals is becoming more wide spread, in part due their ability to catalyse reactions with incredible selectivity under relatively mild conditions. In particular, enzymes are useful for selective reduction of ketones to enantiomerically pure alcohols or amines, and partial oxidations of alkanes to alcohols. However, a key limitation to exploiting redox enzymes in these reaction pathways is the requirement for a specialised electron source, usually the expensive nicotinamide cofactors NADH or NADPH. Existing cofactor regeneration methods use a second enzyme with a sacrificial substrate which is oxidised to generate a stoichiometric waste product; this complicates isolation of the desired product and prevents the environmental benefits of biocatalysis from being fully realised. In order to provide clean and efficient biocatalytic routes, improved recycling methods for these cofactors are crucial. This Thesis develops two novel methods for in situ cofactor recycling. The first is an electro-enzymatic system; an NAD⁺-reductase enzyme is shown to use electrons directly from an electrode for supply of NADH to a co-immobilised cofactor-dependent enzyme. The second uses a hydrogenase, NAD⁺ reductase and cofactor-dependent enzyme immobilised on conducting particles for H₂-driven NADH regeneration. This relies on the thermodynamically favourable reduction of NAD⁺ by H₂ when the hydrogenase and NAD⁺-reductase are in electronic contact, provided by the conducting particle. The electro-enzymatic approach to NAD⁺ reduction is then adapted for electrochemical devices; an enzyme catalysed fuel cell and a self-powered biosensor were considered.

Thesis (Ph. D.)--University of Virginia, 2008. / Title from title page. Includes bibliographical references. Also available online through Digital Dissertations.

Catalysis is a key area of chemistry. Through catalysis it is possible to achieve better synthetic routes, exploit molecules normally considered to be inactive and also attain novel chemical transformations. The development of new catalysts is crucial to furthering chemistry as a field. Computational chemistry, arising from applying the equations of quantum and classical mechanics to solving chemical problems, offers an essential route to investigating the underlying atomistic detail of catalysis. In this thesis calculations have been applied towards studying a number of different catalytic processes. The processing of renewable chemical sources via homogeneous reactions, specifically cardanol from cashew nuts, is discussed. All routes examined for monoreduction of a diene model by [Ru(H)(iPrOH)(Cl)(C₆H₆)] and [Ru(H)(iPrOH)(C₆H₆)]⁺ are energetically costly and would allow for total reduction of the diene if they were operating. While this accounts for the need of high temperatures, further work is required to elucidate the true mechanism of this small but surprisingly complex system. Gold-mediated protodecarboxylation was examined in tandem with experiment to find the subtle steric and electronic effects that dictate CO₂ extrusion from gold N-heterocyclic carbene activated benzene-derived carboxylic acids. The origin of a switch in the rate limiting step from decarboxylation to protodeauration with less activated substrates was also clearly demonstrated. Studies of gold systems are closed with examinations of 1,2-difluorobenzene C–H activation and CO₂ insertion by [Au(IPr)(OH)]. Calculations highlight that the proposed mechanism for oxazole-derived substrates cannot be extended to 1,2-difluorobenzene and instead a digold complex offers more congruent predicted kinetics. The lens of quantum chemistry was turned upon palladium-mediated methoxycarbonylation reactions. An extensive study was undertaken to attempt to understand the bidentate diphosphine ligand dependency on forming either methylpropanoate (MePro) or copolymers. Mechanisms currently suggested in literature are shown to be incongruous with the formation of MePro by Pd(OAc)₂ and bulky diphosphines. A possible alternative route is proposed in this thesis. Four mechanisms for methoxycarbonylation with Pd(2-PyPPh₂)ₙ are detailed. The most accessible route is found to be congruent with experimental reports of selectivity, acid dependency and slight steric modifications. A modification of 2-PyPPh₂ to 2-(4-NMe₂-6-Me)PyPPh₂ is shown to improve both selectivity and turnover, the latter by four orders of magnitude (highest transition state from 22.9 kcal/mol to 16.7 kcal/mol ∆G), and this new second generation in silico designed ligand is studied for its applicability to wider substrate scope and different solvents. The final chapter of this thesis is a mixed quantum mechanics and molecular mechanics (QM/MM) examination of an enzymatic reaction, discussing the need for certain conditions and the role of particular amino acid residues in an S[sub]N2 hydrolysis reaction.