NiFe hydrogenases are promising candidates for sustainable energy approaches using H2 as a clean fuel. Comparison with the rare earth metal containing H2-conversion catalysts that are used industrially highlights a number of distinct advantages in the use of NiFe hydrogenases or hydrogenase-inspired catalysts. They are able to catalyse H2-conversion with high turnover frequencies while containing only base metals, they tolerate a greater variety of conditions and don’t require the use of fossil fuel derived feedstocks. However, implementation of hydrogenase-based catalysts is hampered by the incomplete understanding of some of the aspects of its catalysis. While a general outline for its catalytic mechanism has been established, the specific determinants for efficient H2-cleavage and H+-reduction remain vague. NiFe hydrogenases with different activity profiles, physiological functions and parent organisms all display a highly conserved active site environment yet the way in which the active site chemistry is modulated by the surrounding protein architecture is poorly understood. The importance of understanding this modulation is demonstrated by the small molecule hydrogenase-inspired catalysts that do exhibit efficient H2-conversion invariably featuring either a significant second coordination sphere, or are cofactors implanted within a protein matrix.
The carbonyl and cyanide ligands present in the NiFe hydrogenase active site are ideal probes for IR experiments, indeed vibrational spectroscopy has been used extensively to identify and monitor the interconversion of redox structural states formed by NiFe hydrogenases. Conventional IR approaches use the fundamental frequencies of the carbonyl and the two cyanide stretching modes (vCO, vCN1 and vCN2, respectively) as fingerprints to identify specific redox states. When evaluated in isolation these frequencies can provide only limited catalytically relevant insight as they are subject to a variety of structural and electronic factors including the ligands present in the particular active site state, H-bonding and long range electrostatic interactions with the protein matrix. The interpretation of vCO and vCN mode frequencies is hampered by uncertainty in the anharmonicities along the CO and CN stretching coordinates, in addition to the precise composition of their associated normal modes. While the vCN1 mode is generally acknowledged to have greater symmetric stretching character and the vCN2 mode to have greater asymmetric stretching character, a thorough understanding of their potential energy surfaces (contributions from each cyanide ligand) has yet to be established. This situation is further complicated by the highly localised nature of the vCO and vCN modes to their respective bond coordinates, which results in their stretching frequencies being insensitive to isotopic substitution of adjacent atoms, which would otherwise provide useful information relating to the molecular composition proximal to the [NiFe] site. Here, ultrafast pump-probe and 2D-IR spectroscopy are used to characterise the vCO and vCN modes of the regulatory hydrogenase from Ralstonia eutropha, the membrane bound hydrogenase (Hyd-1) from Escherichia coli and the Fe-site mimic K[CpFe(CO)(CN)2]. A number of previously unexplored spectroscopic parameters are evaluated and a high degree of similarity is observed between different redox states and enzymes, implying an evolutionarily selected active site environment. Via comparison of these observables with those from the Fe-site mimic dissolved in a variety of solvents, some parameters can be ascribed to intrinsic properties of the active site Fe(CO)(CN)2 moiety, whilst others are indicative of the protein scaffold forming a highly specialized environment that tightly controls the structure of the active site. This situation is similar to that observed in both native [FeFe]-hydrogenase and in active site mimics, for which the spectroscopy of the isolated active site is distinctly different to that of the active site implanted within the protein matrix.
