Understanding Scale-down of Oxygen Dependent Biocatalysis

Biocatalysis has been raising great interest in industrial chemical synthesis, particularly in the development of new biocatalytic routes for oxidation reactions. Oxidation reactions are widely used in the chemical industry, however, traditional chemocatalytic processes are poorly selective, carried out under harsh reaction conditions and generate significant quantities of waste.

In contrast, enzymes used to catalyse oxidation reactions offer exquisite selectivity and the ability to work under mild reaction conditions, which contributes to the development of sustainable industrial processes. Besides these great benefits, biocatalysis offers the possibility to modify and tune the biocatalyst, adapting it towards the process needs.

Through protein engineering, modifications can be made to an enzyme, which influence its performance. Protein engineering methods offer an extra degree of freedom to the process design task, enabling orders of magnitude improvements compared to process engineering methods. However, process limitations have to be identified at an early stage of process development to define targets for protein engineering research.

Most oxidative enzymes require molecular oxygen as an oxidant, which is conventionally supplied to the reaction by aeration in a stirred tank reactor. Although oxygen is an environmentally friendly oxidant which is inexpensive and abundant, its supply to a reaction is usually the primary constraint that prevents the implementation of biocatalytic oxidations in industry.

Moreover, it also makes these reactions difficult to study at the laboratory scale. This can be due to either poor biocatalyst stability in the presence of gas-liquid interfaces or to reaction rates being limited byeither oxygen transfer into the liquid medium or biocatalyst kinetics. In this context, this thesis proposes a scale-down approach to investigate the main limitations of biocatalytic oxidation reactions, at an early stage of process development.

This approach aims to acquire fundamental understanding of the biocatalyst kinetics and stability, under relevant process conditions dictatedby industrial targets. In the first part of the work, a systematic methodology based on reaction trajectory analyses was applied, in order to identify the major limitations of an oxidase-catalysed reaction.

Glucose oxidase was used as a model enzyme and the major process limitations identified were related to oxygen transfer rate and enzyme kinetics. The maximum biocatalyst load in order to avoid oxygen transfer limitations was found for the reactor setup used. Furthermore, the kinetic parameters for GOx were determined using a tube-in-tube reactor and the KMO (0.843 mM) was found to be higher than the solubility of oxygen in aqueous solutions (0.265 mM).

This result indicates that the enzyme efficiency is compromised when the reaction is carried out under typical operational conditions (aeration at atmospheric pressure). In the second part of the work, the stability of NAD(P)H oxidases towards gas-liquid interfaces was investigated. For the first time, an in situ experimental method that quantifies the gas-liquidinterfacial area inside an aerated stirred tank reactor was developed.

This method enabled the investigation of kinetic stability of enzymes in the presence of gas-liquid interfaces, similar to large-scale bioreactors. It was found that the half-life of NAD(P)H oxidases decreased with an increase in gas-liquid interfacial area. Results demonstrated that mechanical stirring and the presence of gas-liquid interfacial area each have an individual effect on enzyme deactivation.

Therefore, the effect of mechanical stirring was studied in the absence of gas-liquid interface and the deactivation rate of NAD(P)H oxidase increased with an increase in power input per volume of the reactor. It is proposed that the deactivation is caused by molecular collisions in the reactor and that their frequency increases with higher stirring power input.

The stability of these enzymes was also studied in the absence of stirring and in the presence of gas-liquid interfaces, in a laboratory scale bubble column. It was shown that these oxidases deactivated faster with the increase of interfacial area and that the presence of oxygen in the gas feed enhanced the deactivation rate.

With these findings, a conceptual stability map identifying the causes for enzyme deactivation as a function of mechanical stirring and gas supply was elaborated. Finally, a scale-down approach to investigate and understand biocatalytic oxidations was developed. This approach is based on an industrially driven philosophy and on the results for kinetics and stability of oxygen dependent biocatalysts obtained here.

It suggests scale-down experiments depending on the limitation that is being investigated: kinetics or stability. The outcome of this approach contributes to the establishment of practicable solutions to improve biocatalytic oxidation processes and ultimately, to accelerate their implementation in industry.