Tailoring the redox enzyme-material interface for enhanced electro- and photocatalytic fuel synthesis

The utilisation of sunlight to convert carbon dioxide (CO2) to glucose in natural photosynthesis serves as inspiration for the design of (semi)artificial systems which convert electrical or solar energy into chemical fuels. These artificial (photo)synthetic systems are generally comprised of either an electrode or a cheap and robust light harvester interfaced with a highly efficient catalyst. Redox enzymes serve as model natural catalysts that can convert substrates such as protons (H+) and CO2 to useful fuels such as hydrogen (H2) and formate (HCOO-) with higher efficiencies than artificial systems. Although their oriented non-covalent attachment to electrodes and photosensitisers enables substrate conversion at high efficiencies, in both cases, stability is limited to hourly timescales, the reasons for which remain convoluted. This ambiguity stems from the need for a more thorough characterisation of the redox enzyme-material interface. In addition, the resulting non-covalent biohybrids are expected to be fragile to external stimuli such as pH and salt concentration. Therefore, in this thesis, an in-depth characterisation of the enzyme-electrode interface is combined with insights into the effect of material surface chemistry, external components, and covalent binding at the enzyme-photosensitiser interface with the aim of significantly advancing light-driven semi-artificial photosynthetic systems. First, [NiFeSe]-H2ase and [W]-FDH from Desulfovibrio vulgaris Hildenborough (DvH) is immobilised on a range of charged and neutral self-assembled monolayer (SAM)-modified gold electrodes with varying hydrogen bond (H-bond) donor capabilities. The key factors dominating the activity and stability of the immobilised enzymes are determined using protein film voltammetry (PFV), chronoamperometry (CA), and electrochemical quartz crystal microbalance (E-QCM) analysis. Electrostatic and H-bonding interactions are resolved, with electrostatic interactions responsible for enzyme orientation while enzyme desorption is strongly limited when strong H-bonding is present at the enzyme-electrode interface. Conversely, enzyme stability is drastically reduced in the absence of strong H-bonding, and desorptive enzyme loss is confirmed as the main reason for activity decay by E-QCM during CA. This study provides insights into the possible reasons for the reduced activity of immobilised redox enzymes and the role of film loss, particularly H-bonding, in stabilising bioelectrode performance, promoting avenues for future improvements in bioelectrocatalysis. Semi-artificial approaches to renewable fuel synthesis exploit the integration of enzymes with synthetic materials for kinetically efficient fuel production. The CO2 reductase enzyme FDH is interfaced with carbon nanotubes (CNTs) and amorphous carbon dots (a-CDs). Each carbon substrate, tailored for electro- and photocatalysis, is functionalised with positive (-NHMe2+) and negative (-COO−) chemical surface groups to understand and optimise the electrostatic effect of protein association and orientation on CO2 reduction. Immobilisation of FDH on positively charged CNTs results in efficient and reversible electrochemical CO2 reduction via direct electron transfer (DET) with >90% Faradaic efficiency and −250 µA cm−2 at -0.6 V vs SHE (pH 6.7 and 25 °C) for formate production. In contrast, negatively charged CNTs only result in marginal currents with immobilised FDH. QCM analysis and attenuated total reflection infrared (ATR-IR) spectroscopy confirm the high binding affinity of active FDH to CNTs. FDH has subsequently been coupled to a-CDs, where the benefits of the positive charge (-NHMe2+-terminated a-CDs) were translated to a functional CD-FDH hybrid photocatalyst. High rates of photocatalytic CO2 reduction (turnover frequency: 3.5 × 103 h−1; AM 1.5G) with DL-dithiothreitol as the neutral sacrificial electron donor were obtained when compared to charged sacrificial electron donors, providing benchmark rates for homogeneous photocatalytic CO2 reduction with metal-free light absorbers. This work provides a rational basis to understand interfacial surface/enzyme interactions at electrodes and photosensitisers to guide improvements with catalytic biohybrid materials. Consequently, non-covalent supramolecular biohybrids are susceptible to pH and external salt and buffer components. To mitigate for this, the CD surface chemistry can be further tuned to incorporate unsaturated carbonyls such as maleimide functional groups which can selectively bind to thiols under neutral conditions. In this case, a site-directed mutant FDH baring a cysteine close to the distal FeS cluster was attempted at covalent conjugation with maleimide-CDs to enhance DET photocatalytic CO2 reduction and stability. However, preorientation from the underlying surface charge of the CDs was found to be pivotal in enabling the conjugation of FDH in an electroactive orientation, while additional stability issues were uncovered with the maleimide group under reducing conditions. This work expands the understanding of the non-covalent interactions that govern the enzyme-material interface, with the aim of improving the electro- and photocatalytic activities of redox enzyme biohybrid systems. However, by solely harnessing non-covalent interactions, electro- and photocatalytic activities of the biohybrids fall short of the intrinsic solution assay activities of the isolated enzymes. The use of alternative interactions, such as covalent binding, may enable continued development of biohybrid semi-artificial photosynthetic systems to reach, and possibly exceed, intrinsic enzyme activities.