Synthetic access to a framework-stabilized and fully sulfided analogue of an Anderson polyoxometalate that is catalytically competent for reduction reactions

Polyoxometalates (POMs) featuring 7, 12, 18, or more, redox-accessible transition-metal ions are ubiquitous as selective catalysts, electrocatalysts, and sensitized photocatalysts, especially for oxidation reactions. The corresponding synthetic and catalytic chemistry of stable, discrete, and capping-ligand-free polythiometalates (PTMs), which could be especially attractive for reduction reactions, is much less well developed. Among the challenges is the propensity of PTMs to agglomerate and form larger clusters of indeterminate size, as well as the tendency for agglomeration to block access of candidate reactants to potential catalyst active-sites. Nevertheless, the pervasive presence of transition-metal sulfur clusters metalloenzymes or cofactors that catalyze reduction reactions, and the justifiable proliferation of studies of 2D metalchalcogenides, and especially their edge sites, as reduction catalysts, point to the promise of well-defined and controllable PTMs as catalysts for reduction reactions, including complex, bond-forming, many-electron reactions. Here we report the fabrication of agglomeration-immune, reactant-accessible, capping-ligand-free CoIIMoIV6S24n- clusters as periodic arrays in a water-stable, hierarchically porous Zr-metal-organic-framework (MOF; NU1K) by first preparing and installing a disk-like Anderson polyoxometalate, CoIIMoVI6O24m(-), in size-matched (<1 nm) micropores termed c-pores, where the siting is established via DED (difference electron density) X-ray diffraction experiments. Prolonged treatment with flowing H2S while heating, uniformly reduces the six molybdenum(VI) ions to Mo(IV) and quantitatively replaces oxygen anions with similarly ligating sulfur anions in the form S(2-), HS(-), and S2(2-). Further DED measurements show that the templated POM-to-PTM conversion leaves the clusters individually isolated in open-channel-connected c-pores. The structure of the immobilized cluster as determined, in part, by XPS, XAFS, and PDF (pair-distribution function) analysis of total X-ray scattering agrees very well with the theoretically simulated structure. Preliminary, proof-of-concept experiments show that electrode-supported thin-films of CoMo6S24@NU1K are electrocatalytically competent for hydrogen evolution in aqueous acid (e.g. 10 mA·cm-2 of current density at an overpotential of 100 mV). Suspensions of CoMo6S24@NU1K in acetonitrile + triethanolamine, are photocatalytically competent for hydrogen evolution via sensitization with chromophoric MOF linkers. Nevertheless, the initially installed PTM appears to be a pre-catalyst, as hydrogen evolution is observed only after four hours of photolysis. Reduction-assisted loss of ~3-to-6 sulfurs, as H2S, likely is responsible for pre-catalyst-to-catalyst conversion, as the loss opens coordination sites on multiple cluster-sited metal ions, perhaps enabling hydrogen evolution via a Mo-hydride intermediate. Given the great variety of sizes and compositions available for both POMs and Zr-MOFs, we suggest that the approach described here can be adapted for the synthesis and stabilization of periodic arrays of other non-agglomerating, capping-ligand-free PTMs of well-defined metal-nuclearity, presumably including catalytically functional PTMs.