Enzymes offer a transformative solution to traditional chemical catalysts, providing both highly selective and highly pure compounds for the fine chemical, pharmaceutical, and insecticide industries. Despite significant progress, the enzymatic toolbox remains somewhat limited, spurring scientists in the biocatalysis field to strive for the expansion of chemical reactivities. The aim of this thesis was to study the Old Yellow Enzymes (OYEs) chemical versatility, in an attempt to broaden their reaction portfolio. This work is organized into five research chapters: four focusing on the chemical reactivities of OYE, and the fifth on a scaled-up reaction. We begin with an introduction outlining the state-of-the-art of OYEs and conclude with a conclusion and outlook. A detailed overview of each chapter is provided in the following paragraphs.

Chapter one provides an overview of our current understanding of OYEs, covering their history, physiological roles, classification among enzymes, structural characteristics, coenzymes, chemical reactivities, and their potential applications within industry.

In chapter two, we reveal how versatile OYEs are, by exploring an unknown reactivity, the stereoselective monoreduction of α,β-dicarbonyls towards chiral α-hydroxycarbonyls. We investigated ten aromatic, cyclic, aliphatic α,β-dicarbonyl compounds and evaluated their reduction using five OYEs and one flavin-independent double bond reductase (DBR). The most effective substrate was the aromatic α,β-dicarbonyl 1-phenyl-1,2-propanedione, which was converted to phenylacetylcarbinol with 91% conversion using OYE3 (R-selectivity >99.9% ee).

In chapter three, we continue to highlight OYEs’ versatility by yet another reactivity, the semireduction of allenes. Six activated allene substrates were screened against eighteen enzymes, including sixteen OYEs and two DBRs. The best results occurred using a class I OYE, PETNR, with 99% conversion of 10 mM ethyl-2,3-pentadienoate to ethyl-pent-3-enoate (E:Z ratio, 49:51). High selectivity was observed with class II OYE3 using methyl 2-methyl-2,3-pentadienoate as a substrate (81% conversion, E:Z ratio, 11:89), as well as with another class II OYE, EBP1 and ethyl 2-methyl-2,3-butadienoate (87% conversion, 97% ee).

Chapter four covers our re-examination of OYEs’ oxidative reaction, developing a new method for selective desaturation without requiring high temperatures. We show that by a simple pH adjustment, the active site tyrosine is deprotonated and serves as a catalytic base. Several OYEs and substrates were screened to demonstrate this desaturation method. This study expands the range of biocatalytic applications for OYEs, introducing an elegant approach to synthesizing chiral α,β-unsaturated carbonyl compounds.

In chapter five we further examined the intricacies of oxidation, by assessing whether redox potential influences desaturation. We measured the redox midpoint potential of eleven OYEs and their mutants from various classes, focusing on specific active site mutations that might shed light on desaturase activity. Our findings revealed a range of redox potentials across the different OYE classes, but no clear correlation between desaturation activity and redox potential. We examined the active site’s threonine/cysteine near the flavin N5 position and the proton-donating tyrosine with mutant enzymes to understand their role in desaturation.

In chapter six we demonstrate that OYEs are well suited for industrial use, by carrying out a 150 g/L scale-up for monoterpene asymmetric reduction. Until now, OYEs have rarely been applied in scale-up reactions, with limited turnover numbers of 102-104. We present a preparative scale using the thermostable OYE from Thermus scotoductus (TsOYE) for the asymmetric reduction of activated alkenes achieving a record turnover number of 123,000 with 1 M of (S)-carvone (98% conversion, 90% isolated yield) towards product (2R,5S)-dihydrocarvone with a diastereomeric excess of 92% (>99% ee).

In general this work advances the understanding of the biocatalytic reactivity of the OYE family, demonstrating their capacity to catalyze diverse and novel chemical reactions towards industrial applications.

The CO2 emissions started to peak during the start of this century and the steel manufacturing sector accounts for 25% of the global CO2 emissions. The majority of steel production is based on the route of the basic oxygen furnace (BOF) route, which is energy efficient. This production route results in the emission of BOF gas. So a new method to produce Isopropyl alcohol via BOF gas fermentation was developed for the pilot scale plant and the LCA was performed by Liew et al., (2022). The carbon-negative emissions were calculated and estimated in the pilot scale study. A few inconsistencies were identified in the calculation method, in which the components utilized for the production process contributed less than 5% to the Global Warming Potential value, and the replacements provided to the steel mill for the BOF gas redirection were cut off. The LCA value for the pilot-scale plant was incomplete and inconsistent, and only the GWP was chosen as the prominent midpoint indicator. Based on this pilot-scale plant, an industrial-scale (46 kton/ yr) base case process model was developed for the gas-fermentation process to produce IPA. The fermentation is the major step involved in the production of IPA, which involves the acetogenic bacteria Clostridium Autoethanogenum. The initial process involves the capture of the emission of BOF gas for compression and cooling performed because the temperature of BOF gas is around 1100°C. This compressed gas is fermented using of microbes and filtered out to obtain the filtered broth. Then the IPA is separated from the filtered broth using extractive distillation, using glycerol. The major product of IPA is processed out of the system. This research aims to analyze and estimate the environmental impacts of this process model of BOF gas fermentation to produce Isopropyl alcohol. The assessments were performed for the base case and the 11 process parameters of CO conversion, Volumetric mass transfer rate of CO, Product selectivity, Dilution rate, Extractive distillation glycerol mole fraction, Temperature offgas condenser, Anaerobic waste conversion, Extractive distillation molar reflux ratio, Biomass liq-liq mole fraction and the broth and glycerol purges. The impact assessment results help to identify the potential contributors in the process that affect the environment, so the process model can be optimized to yield lower impact values concerning the environmental perspectives.

The 7 midpoint indicators namely Global Warming Potential (GWP), Stratospheric Ozone Depletion (SOD), Fine Particulate Matter Formation (FPMF), Freshwater Eutrophication (FE), Marine Eutrophication (ME), Human Carcinogenic Toxicity (HCT), and Land Use (LU) were chosen to obtain a detailed view on the ecosystem, human health, and the environmental effects. The replacement calculations estimated for 1 kg of IPA production is 4.324 MJ of heat, and 0.877 kWh of electricity, has to be replaced for the steel mill. The impact assessment for the base case model was performed and compared with the conventional IPA production method (GWP: 2.026 kg CO2 eq.), the global warming potential for the BOF gas fermentation method (GWP: 27.656 kg CO2 eq.) estimated to be a 1265% increase compared with the conventional IPA method. Similarly, all seven impact categories are estimated to have a huge increase in values. An elaborate study compared and identified the most influential process parameters. The process parameter of dilution rate in which the dilution rate is lowered by 30% from the base case value appears to be the process parameter with a lower impact value of 20.464 kg CO2 eq. for the Global Warming Potential (26% lower than base case), 1.72E-05 kg CFC11 eq. for the Stratospheric Ozone Depletion (31% lower than base case), 9.618E-03 kg PM2.5 eq. for the Fine Particulate Matter Formation (38% lower than base case), for the Freshwater Eutrophication the value is 9.853E-04 kg P eq. (53% lower than base case), 5.531E-03 kg N eq. for Marine Eutrophication (33% lower than base case), 1.733E-01 kg 1,4-DCB for the Human Carcinogenic Toxicity (53% lower than base case), and the 3.265 m2a crop eq. (32% lower than base case) for the Land use impact categories. The major contributors are the high-pressure and low-pressure steam utility contributing greater than 56% for the impact category values of the GWP, FPMF, FE, and HCT particularly. The glycerol used for the extractive distillation process contributed greater than 92% for the impact categories of SOD, ME, and LU. The impact assessments across different indicators were interpreted and the major process parameters that are more influential in reducing the impact values are identified. The results indicate that the major contribution to the impact value is reduced by the emission credit ±40% for preventing the BOF gas from flaring. The carbon dioxide emission from the process, glycerol component, and steam utility collectively contribute majorly which account for more than 90% of the total impact values in the impact categories. Lowering the dilution rate, glycerol purge fraction, and glycerol mole fraction by -30%, and increasing the volumetric mass transfer rate by +30% of the process model could result in lower impact values. The CO2 emitted from the process is estimated to be 9.34 kg CO2 eq. This emission is higher than the feedstock (BOF gas) used for the gas-fermentation process. The entertainer glycerol is anaerobically digested and combusted leading to the 10% of the total CO2 emission of the process model. The process model should be updated with the process parameters listed above and a similar life cycle impact assessment has to be performed to compare the impact values. This updated process model might have comparatively better results. The sensitivity study has been performed to estimate the percentage effects of the combined glycerol and steam components on the midpoint indicator impact value. Cutting off the components of glycerol and steam from the process model still yields the GWP value of 6.16 kg CO2 eq. for industrial scale process which is a 204 % increase than the conventional IPA process. This implies that the updated process model with similar process steps cannot obtain impact values lower than the conventional IPA method. To lower the impact value of GWP, the CO2 emission from the process should be sequestrated (carbon capture) to lower the GWP value by 42% from the base case.

Hydrogen is an interesting and promising alternative for fuels. Currently, most hydrogen is produced via a process called steam reforming, where a reaction between methane and water occurs, with carbon monoxide and carbon dioxide as waste products and released in the atmosphere. This work presents a study on the sustainable production of hydrogen. A pool of six hydrogen processes were chosen; three based on electrolysis, two based on fermentation and one based on electrolysis combined with a microbial community. On these six methods an early stage analysis was applied, screening for economics, environmental impacts and processrelated costs and impacts. Two electrolysis methods were selected for a conceptual design; alkaline electrolysis (AE) and polymer electrolyte membrane (PEM) electrolysis, both for an annual production of 1 million kg hydrogen. The fixed capital was established $86.8M for AE and $62.3M for PEM. After the design, a techno-economic analysis was performed, providing the total production cost (TPC) and the minimum fuel selling price (MFSP). For the AE, the MFSP resulted in being $21.46/kg hydrogen and an TPC of $17.16/kg hydrogen. The biggest contribution to the MFSP was the depreciation, being $10.19/kg hydrogen. The PEM electrolysis had a MFSP of $17.56/kg hydrogen and an TPC of $14.05/kg hydrogen. Also for PEM electrolysis, the biggest contribution to the MFSP was the depreciation, which was $7.32/kg. A sensitivity analysis was performed and showed the biggest uncertainty in the equipment cost and the interest rate. The MFSP for AE changes ± $3.79/kg hydrogen by a change in equipment cost of 30% and ± $3.27 - $3.72 by a 5 percentage point change of the interest rate. The MFSP for PEM changes ± $2.73/kg hydrogen by a change in equipment cost of 30% and ± $2.36 - $2.73 by a 5 percentage point change of the interest rate. Based on the results from the techno-economic analysis and sensitivity analysis, PEM electrolysis was identified as the superior method.

Our ability to tailor enzymatic properties is a critical factor for biocatalyst application in the industrial sector. Although many wild-type enzymes have been found capable of promiscuously catalysing desired anthropogenic reactions, their activity and selectivity for non-natural transformations is often poor. Consequently, it is crucial to optimise enzymes such that they can be effectively integrated into industrial processes. Notable added advantages in this context are that enzymes are considered 'green' catalysts - enhancing the perceived value of products in today's environmentally-conscious society – and that biocatalysts can carry out intricate chemistries with exceptional regio- and stereoselectivity, complementing traditional organic synthesis.