Realistic Modelling of Organometallic Reactivity in Solution: Computational Studies on the Mechanism of Methanolysis of Palladium-Acyl Bonds

Methoxycarbonylation is a process that converts cheap, widely available feedstocks (alkenes, carbon monoxide and methanol) into commercially-important intermediates for the chemicals industry. This process uses Pd-based catalysts and the best example is the reaction of the simplest alkene, ethene, to give the intermediate methyl methacrylate, which is used in the synthesis of plastics. More recently, alkenes such as vinyl acetate have been shown to undergo methoxycarbonylation to generate intermediates that are themselves useful as green solvents (low-volatility/biodegradable) or as monomers for the formation of biodegradable polymers. These have the potential to replace traditional materials such as polystyrene or polythene. Methoxycarbonylation of butadiene promises a new route to adipic acid, one of the co-monomers involved in the manufacture of nylon. As yet the methoxycarbonylation of vinyl acetate and butadiene have not been optimised and greater insight into these reactions is required before effective industrial processes are in place.A key issue that remains to be solved in the methoxycarbonylation reaction is the detailed mechanism by which the products are released - the so-called methanolysis step. There a several possibilities for this process, however, it is extremely difficult to obtain information on this from experiment as the reaction itself is incredibly fast. In these circumstances the use of computational modelling comes into its own, as this can readily provide information on the energies of the species involved in reactivity. The methanolysis reaction is, however, very complicated and is strongly dependent on the precise nature of the reacting species and the nature of the solvent being used. To obtain reliable modelling data these factors must be taken into account, a fact that makes the task of modelling these systems very challenging.This proposal seeks to use high level computational modelling to assess the mechanism of the methanolysis on the simplest methoxycarbonylation system - ethene/CO/MeOH - and the most effective Pd catalysts. Our approach will be to employ hybrid calculations where the catalyst and reacting molecules are dealt with at a high level of theory (density functional theory) but the solvent molecules (many 10s or hundreds) are treated at a lower level of theory based on classical force fields. Through this approach the effect of the solvent on the reactivity at the Pd catalyst will be taken into account and we aim to provide extremely reliable data to define the preferred mechanism. We will test our approach by comparing with an alternative catalyst which displays a different reactivity, thus giving a stringent test of our modelling approach.Once we have defined the correct way to treat these complicated reactions - as well as the mechanism by which methanolysis occurs - we will be in a position to tackle the new reactions of vinyl acetate and butadiene. We hope to provide sufficient insight into these processes that experimental chemists will be able to design new improved catalysts for more efficient methoxycarbonylation of these feedstocks on an industrial scale.