Some microorganisms have a metabolism that enables them to consume or produce molecular hydrogen by means of enzymes known as hydrogenases. Crystallographic characterization of [NiFe]-hydrogenases in 1995 and [FeFe]-hydrogenases four years later revealed the unusual presence of iron-carbonyl-cyanide centers, whereas these two types of hydrogenase have no phylogenetic relationship. Chemists seized on this finding and sought to better understand how hydrogenases work by synthesizing structural models of the catalytic sites. This approach has been facilitated in the case of [FeFe]-hydrogenases by the fact that their structure is very close to that of difer dithiolate complexes, the synthesis of which has been known for over 70 years. Recently, some of these models have begun to provide relevant information on the catalysis of hydrogen production, but many differences in reactivity remain. In particular, hydrogenases operate at a potential close to the thermodynamic limit, while using metals of the first transition series with a low affinity for hydrogen. Although there are few examples of homogeneous catalysts for hydrogen production, the coordination chemistry of hydrogen has been widely studied and its contributions are directly transposable to biological processes. Hydrogen itself has neither acid/base nor redox reactivity, but its coordination complexes possess both. This means that metal-hydrogen derivatives can be acidic and metal-hydride derivatives can be oxidized. Not surprisingly, the catalytic sites of hydrogenases contain transition metals.
In this seminar, we focus on the reactivity of chemical models of the [FeFe]-hydrogenase catalytic site. In particular, the thermodynamic and kinetic factors controlling protonation and electron transfer reactions are clarified. Through the study of mixed valence compounds, we also address the coordination of hydrogen and carbon monoxide, a common inhibitor of hydrogenases. The implications of this work for hydrogen production by coupling with a water photooxidation system will be discussed in conclusion.
