An all-inorganic, oxidatively and thermally stable, homogeneous water oxidation catalyst based on redox-active (vanadate(V)-centered) polyoxometalate ligands, Na10[Co4(H2O)2(VW9O34)2]·35H2O (Na101-V2, sodium salt of the polyanion 1-V2), was synthesized, thoroughly characterized and shown to catalyze water oxidation in dark and visible-light-driven conditions. This synthetic catalyst is exceptionally fast under mild conditions (TOF > 1 × 10(3) s(-1)). Under light-driven conditions using [Ru(bpy)3](2+) as a photosensitizer and persulfate as a sacrificial electron acceptor, 1-V2 exhibits higher selectivity for water oxidation versus bpy ligand oxidation, the final O2 yield by 1-V2 is twice as high as that of using [Co4(H2O)2(PW9O34)2](10-) (1-P2), and the quantum efficiency of O2 formation at 6.0 μM 1-V2 reaches ∼68%. Multiple experimental results (e.g., UV-vis absorption, FT-IR, (51)V NMR, dynamic light scattering, tetra-n-heptylammonium nitrate-toluene extraction, effect of pH, buffer, and buffer concentration, etc.) confirm that the polyanion unit (1-V2) itself is the dominant active catalyst and not Co(2+)(aq) or cobalt oxide.

Distinguishing between homogeneous and heterogeneous catalysis is not straightforward. In the case of the water oxidation catalyst (WOC) [Co4(H2O)2(PW9O34)2](10-) (Co4POM), initial reports of an efficient, molecular catalyst have been challenged by studies suggesting that formation of cobalt oxide (CoOx) or other byproducts are responsible for the catalytic activity. Thus, we describe a series of experiments for thorough examination of active species under catalytic conditions and apply them to Co4POM. These provide strong evidence that under the conditions initially reported for water oxidation using Co4POM (Yin et al. Science, 2010, 328, 342), this POM anion functions as a molecular catalyst, not a precursor for CoOx. Specifically, we quantify the amount of Co(2+)(aq) released from Co4POM by two methods (cathodic adsorptive stripping voltammetry and inductively coupled plasma mass spectrometry) and show that this amount of cobalt, whatever speciation state it may exist in, cannot account for the observed water oxidation. We document that catalytic O2 evolution by Co4POM, Co(2+)(aq), and CoOx have different dependences on buffers, pH, and WOC concentration. Extraction of Co4POM, but not Co(2+)(aq) or CoOx into toluene from water, and other experiments further confirm that Co4POM is the dominant WOC. Recent studies showing that Co4POM decomposes to a CoOx WOC under electrochemical bias (Stracke and Finke, J. Am. Chem. Soc., 2011, 133, 14872), or displays an increased ability to reduce [Ru(bpy)3](3+) upon aging (Scandola, et al., Chem. Commun., 2012, 48,8808) help complete the picture of Co4POM behavior under various conditions but do not affect our central conclusions.

No abstract