Continuous Oxidative Biocatalysis

Enzymes have a number of characteristics that make them attractive as biocatalytic alternatives to conventional transition-metal based catalysts for the production of both high- and low-value products. These include, but are not limited to, mild operating conditions, high enantioselectivity and the possibility of modifying their properties to better suit a process. In particular, the use of oxidases to catalyze selective oxidations using molecular oxygen as an environmentally friendly oxidant, is gaining attention. However, as most industries are trending towards continuous processing to improve efficiency, biocatalysis must do the same. Since continuous biocatalysis is an emerging field, the goal of this thesis was to gain a deeper understanding of the limitations of oxidase-catalyzed reactions and then propose a practical reactor configuration to facilitate their industrial implementation.
Plug-flow is the preferred mode of continuous operation to allow good control of residence times and enable full conversion. However, the lack of mixing makes plug-flow reactors unsuitable for multiphase systems like gas-liquid biocatalytic oxidations. Therefore, operation in stirred tanks is proposed as a scalable alternative that affords better mass transfer. Even so, the biocatalytic oxidation of glucose was found to be predominantly oxygen limited in a continuous stirred tank reactor. This is due to the low water-solubility of molecular oxygen, which severely limits the driving force for gas-liquid mass transfer and results in ineffective use of the enzyme due to its comparatively low affinity towards oxygen. To raise its solubility, the partial pressure of oxygen in the reactor must be increased, either by raising the total pressure, which increases costs, or the oxygen content of the feed gas, which was found to potentially deactivate the enzyme above concentrations of 60-80%. A better alternative would thus be to employ protein engineering to improve the affinity of the enzyme towards oxygen. In the meantime, a model of the system was used to demonstrate that just one additional reactor in series with the first enables more effective enzyme use and the possibility of near-complete conversion.
The gas-liquid mass transfer coefficient, kLa, which is largely thought of as a reactor property, was found to be highly influenced by the composition of the reaction media. This is likely based on the complex interactions of the individual media components at the gas-liquid interface. Unfortunately, these influences could not be modelled with sufficient accuracy to allow reliable prediction of kLa. Since the kLa is a critical process parameter for the design, scale-up and operation of an oxygen-limited biocatalytic reaction, an alternative means of estimating its value that accounts for the reactor operating conditions, as well as the reaction itself, is required. This was done by fitting the model of the system to experimental data. The estimated kLa values were significantly higher than those measured in pure water. Additionally, the kLa in the second reactor was higher than the first. This indicates that, in the case of gas-liquid biocatalytic reactions, the enzyme may be used more effectively with each successive stirred tank in a series, at least until the primary substrate becomes rate-limiting. The digital model of the system was further used to explore alternative strategies to improve reaction performance. If provided with economic constraints on each operating parameter, it would allow independent optimization of the reactors in a series to achieve near-complete conversion, as well as targets for improvements to the enzyme, through protein engineering, to be set.