The design of molecular water oxidation catalysts (WOCs) requires a rational approach that considers the intermediate steps of the catalytic cycle, including water binding, deprotonation, storage of oxidizing equivalents, O–O bond formation, and O2 release. We investigated several of these properties for a series of base metal complexes (M = Mn, Fe, Co, Ni) bearing two variants of a pentapyridyl ligand framework, of which some were reported previously to be active WOCs. We found that only [Fe(Py5OMe)Cl]+ (Py5OMe = pyridine-2,6-diylbis[di-(pyridin-2-yl)methoxymethane]) showed an appreciable catalytic activity with a turnover number (TON) = 130 in light-driven experiments using the [Ru(bpy)3]2+/S2O82– system at pH 8.0, but that activity is demonstrated to arise from the rapid degradation in the buffered solution leading to the formation of catalytically active amorphous iron oxide/hydroxide (FeOOH), which subsequently lost the catalytic activity by forming more extensive and structured FeOOH species. The detailed analysis of the redox and water-binding properties employing electrochemistry, X-ray absorption spectroscopy (XAS), UV–vis spectroscopy, and density-functional theory (DFT) showed that all complexes were able to undergo the MIII/MII oxidation, but none was able to yield a detectable amount of a MIV state in our potential window (up to +2 V vs SHE). This inability was traced to (i) the preference for binding Cl– or acetonitrile instead of water-derived species in the apical position, which excludes redox leveling via proton coupled electron transfer, and (ii) the lack of sigma donor ligands that would stabilize oxidation states beyond MIII. On that basis, design features for next-generation molecular WOCs are suggested.
