William H. Casey

Electrochemical Water Oxidation with Cobalt-Based Electrocatalysts from pH 0–14: The Thermodynamic Basis for Catalyst Structure, Stability, and Activity

Building upon recent study of cobalt-oxide electrocatalysts in fluoride-buffered electrolyte at pH 3.4, we have undertaken a mechanistic investigation of cobalt-catalyzed water oxidation in aqueous buffering electrolytes from pH 0-14. This work includes electrokinetic studies, cyclic voltammetric analysis, and electron paramagnetic resonance (EPR) spectroscopic studies. The results illuminate a set of interrelated mechanisms for electrochemical water oxidation in alkaline, neutral, and acidic media with electrodeposited Co-oxide catalyst films (CoO(x)(cf)s) as well as for a homogeneous Co-catalyzed electrochemical water oxidation reaction. Analysis of the pH dependence of quasi-reversible features in cyclic voltammograms of the CoO(x)(cf)s provides the basis for a Pourbaix diagram that closely resembles a Pourbaix diagram derived from thermodynamic free energies of formation for a family of Co-based layered materials. Below pH 3, a shift from heterogeneous catalysis producing O(2) to homogeneous catalysis yielding H(2)O(2) is observed. Collectively, the results reported here provide a foundation for understanding the structure, stability, and catalytic activity of aqueous cobalt electrocatalysts for water oxidation.

Water-oxidation catalysis by manganese in a geochemical-like cycle

Water oxidation in all oxygenic photosynthetic organisms is catalysed by the Mn₄CaO₄ cluster of Photosystem II. This cluster has inspired the development of synthetic manganese catalysts for solar energy production. A photoelectrochemical device, made by impregnating a synthetic tetranuclear-manganese cluster into a Nafion matrix, has been shown to achieve efficient water oxidation catalysis. We report here in situ X-ray absorption spectroscopy and transmission electron microscopy studies that demonstrate that this cluster dissociates into Mn(II) compounds in the Nafion, which are then reoxidized to form dispersed nanoparticles of a disordered Mn(III/IV)-oxide phase. Cycling between the photoreduced product and this mineral-like solid is responsible for the observed photochemical water-oxidation catalysis. The original manganese cluster serves only as a precursor to the catalytically active material. The behaviour of Mn in Nafion therefore parallels its broader biogeochemistry, which is also dominated by cycles of oxidation into solid Mn(III/IV) oxides followed by photoreduction to Mn²⁺.

EPR Evidence for Co(IV) Species Produced During Water Oxidation at Neutral pH

Thin-film water oxidation catalysts (Co-Pi) prepared by electrodeposition from phosphate electrolyte and Co(NO(3))(2) have been characterized by electron paramagnetic resonance (EPR) spectroscopy. Co-Pi catalyst films exhibit EPR signals corresponding to populations of both Co(II) and Co(IV). As the deposition voltage is increased into the region where water oxidation prevails, the population of Co(IV) rises and the population of Co(II) decreases. The changes in the redox speciation of the film can also be induced, in part, by prolonged water oxidation catalysis in the absence of additional catalyst deposition. These results provide spectroscopic evidence for the formation of Co(IV) species during water oxidation catalysis at neutral pH.

Dissolution of insulating oxide materials at the molecular scale.

Our understanding of mineral and glass dissolution has advanced from simple thermodynamic treatments to models that emphasize adsorbate structures. This evolution was driven by the idea that the best understanding is built at the molecular level. Now, it is clear that the molecular questions cannot be answered uniquely with dissolution experiments. At the surface it is unclear which functional groups are present, how they are arranged, and how they interact with each other and with solutes as the key bonds are activated. An alternative approach has developed whereby reactions are studied with nanometre-sized aqueous oxide ions that serve as models for the more complicated oxide interface. For these ions, establishing the structure is not a research problem in itself, and bond ruptures and dissociations can be followed with much confidence. We review the field from bulk-dissolution kinetics to the new isotope-exchange experiments in large oxide ions.

Energetic basis of catalytic activity of layered nanophase calcium manganese oxides for water oxidation

Previous measurements show that calcium manganese oxide nanoparticles are better water oxidation catalysts than binary manganese oxides (Mn3O4, Mn2O3, and MnO2). The probable reasons for such enhancement involve a combination of factors: The calcium manganese oxide materials have a layered structure with considerable thermodynamic stability and a high surface area, their low surface energy suggests relatively loose binding of H2O on the internal and external surfaces, and they possess mixed-valent manganese with internal oxidation enthalpy independent of the Mn3+/Mn4+ ratio and much smaller in magnitude than the Mn2O3-MnO2 couple. These factors enhance catalytic ability by providing easy access for solutes and water to active sites and facile electron transfer between manganese in different oxidation states.

Rates of water exchange for two cobalt(II) heteropolyoxotungstate compounds in aqueous solution.

Polyoxometalate ions are used as ligands in water-oxidation processes related to solar energy production. An important step in these reactions is the association and dissociation of water from the catalytic sites, the rates of which are unknown. Here we report the exchange rates of water ligated to Co(II) atoms in two polyoxotungstate sandwich molecules using the (17)O-NMR-based Swift-Connick method. The compounds were the [Co(4)(H(2)O)(2)(B-α-PW(9)O(34))(2)](10-) and the larger αββα-[Co(4)(H(2)O)(2)(P(2)W(15)O(56))(2)](16-) ions, each with two water molecules bound trans to one another in a Co(II) sandwich between the tungstate ligands. The clusters, in both solid and solution state, were characterized by a range of methods, including NMR, EPR, FT-IR, UV-Vis, and EXAFS spectroscopy, ESI-MS, single-crystal X-ray crystallography, and potentiometry. For [Co(4)(H(2)O)(2)(B-α-PW(9)O(34))(2)](10-) at pH 5.4, we estimate: k(298)=1.5(5)±0.3×10(6) s(-1), ΔH(≠)=39.8±0.4 kJ mol(-1), ΔS(≠)=+7.1±1.2 J mol(-1) K(-1) and ΔV(≠)=5.6 ±1.6 cm(3) mol(-1). For the Wells-Dawson sandwich cluster (αββα-[Co(4)(H(2)O)(2)(P(2)W(15)O(56))(2)](16-)) at pH 5.54, we find: k(298)=1.6(2)±0.3×10(6) s(-1), ΔH(≠)=27.6±0.4 kJ mol(-1) ΔS(≠)=-33±1.3 J mol(-1) K(-1) and ΔV(≠)=2.2±1.4 cm(3) mol(-1) at pH 5.2. The molecules are clearly stable and monospecific in slightly acidic solutions, but dissociate in strongly acidic solutions. This dissociation is detectable by EPR spectroscopy as S=3/2 Co(II) species (such as the [Co(H(2)O)(6)](2+) monomer ion) and by the significant reduction of the Co-Co vector in the XAS spectra.

Metastable structures and isotope exchange reactions in polyoxometalate ions provide a molecular view of oxide dissolution

Reactions involving minerals and glasses in water are slow and difficult to probe spectroscopically but are fundamental to the performance of oxide materials in green technologies such as automotive thermoelectric power generation, CO2 capture and storage and water-oxidation catalysis; these must be made from geochemically common elements and operate in hydrous environments. Polyoxometalate ions (POMs) have structures similar to condensed oxide phases and can be used as molecular models of the oxide/water interface. Oxygen atoms in POM exchange isotopes at different rates, but, at present, there is no basis for predicting how the coordination environment and metal substitution influences rates and mechanisms. Here we identify low-energy metastable configurations that form from the breaking of weak bonds between metals and underlying highly coordinated oxygen atoms, followed by facile hydroxide, hydronium or water addition. The mediation of oxygen exchange by these stuffed structures suggests a new view of the relationship between structure and reactivity at the oxide/solution interface.

Oxygen-Isotope Exchange Rates for Three Isostructural Polyoxometalate Ions

We compare oxygen-isotope exchange rates for all structural oxygens in three polyoxoniobate ions that differ by systematic metal substitutions of Ti(IV) --> Nb(V). The [H(x)Nb(10)O(28)]((6-x)-), [H(x)TiNb(9)O(28)]((7-x)-), and [H(x)Ti(2)Nb(8)O(28)]((8-x)-) ions are all isostructural yet have different Brønsted properties. Rates for sites within a particular molecule in the series differ by at least approximately 10(4), but the relative reactivities of the oxygen sites rank in nearly the same relative order for all ions in the series. Within a single ion, most structural oxygens exhibit rates of isotopic exchange that vary similarly with pH, indicating that each structure responds as a whole to changes in pH. Across the series of molecules, however, the pH dependencies for isotope exchanges and dissociation are distinctly different, reflecting different contributions from proton- or base-enhanced pathways. The proton-enhanced pathway for isotope exchange dominates at most pH conditions for the [H(x)Ti(2)Nb(8)O(28)]((8-x)-) ion, but the base-enhanced pathways are increasingly important for the [H(x)TiNb(9)O(28)]((7-x)-) and [H(x)Nb(10)O(28)]((6-x)-) structures at higher pH. The local effect of Ti(IV) substitution could be assessed by comparing rates for structurally similar oxygens on each side of the [H(x)TiNb(9)O(28)]((7-x)-) ion and is surprisingly small. Interestingly, these nanometer-size structures seem to manifest the same general averaged amphoteric chemistry that is familiar for other reactions affecting oxides in water, including interface dissolution by proton- and hydroxyl-enhanced pathways.

A Soluble Phosphorus‐Centered Keggin Polyoxoniobate with Bicapping Vanadyl Groups

A water-soluble tetramethylammonium (TMA) salt of a novel Keggin-type polyoxoniobate has been isolated as TMA9[PV2Nb12O42]·19H2O (1). This species contains a central phosphorus site and two capping vanadyl sites. Previously only a single example of a phosphorus-containing polyoxoniobate, [(PO2)3PNb9O34](15-), was known, which is a lacunary Keggin ion decorated with three PO2 units. However, that cluster was isolated as an insoluble structure consisting of chains linked by sodium counterions. In contrast, the [PV2Nb12O42](9-) cluster in 1 is stable over a wide pH range, as evident by (31)P and (51)V NMR, UV/Vis spectroscopy, and ESI-MS spectrometry. The ease of substitution of phosphate into the central tetrahedral position suggests that other oxoanions can be similarly substituted, promising a richer set of structures in this class.

Sarcoplasmic reticulum calcium ATPase interactions with decaniobate, decavanadate, vanadate, tungstate and molybdate

Over the last few decades there has been increasing interest in oxometalate and polyoxometalate applications to medicine and pharmacology. This interest arose, at least in part, due to the properties of these classes of compounds as anti-cancer, anti-diabetic agents, and also for treatment of neurodegenerative diseases, among others. However, our understanding of the mechanism of action would be improved if biological models could be used to clarify potential toxicological effects in main cellular processes. Sarcoplasmic reticulum (SR) vesicles, containing a large amount of Ca(2+)-ATPase, an enzyme that accumulates calcium by active transport using ATP, have been suggested as a useful model to study the effects of oxometalates on calcium homeostasis. In the present article, it is shown that decavanadate, decaniobate, vanadate, tungstate and molybdate, all inhibited SR Ca(2+)-ATPase, with the following IC(50) values: 15, 35, 50, 400 μM and 45 mM, respectively. Decaniobate (Nb(10)), is the strongest P-type enzyme inhibitor, after decavanadate (V(10)). Atomic-absorption spectroscopy (AAS) analysis, indicates that decavanadate binds to the protein with a 1:1 decavanadate:Ca(2+)-ATPase stoichiometry. Furthermore, V(10) binds with similar extension to all the protein conformations, which occur during calcium translocation by active transport, namely E1, E1P, E2 and E2P, as analysed by AAS. In contrast, it was confirmed that the binding of monomeric vanadate (H(2)VO(4)(2-); V(1)) to the calcium pump is favoured only for the E2 and E2P conformations of the ATPase, whereas no significant amount of vanadate is bound to the E1 and E1P conformations. Scatchard plot analysis, confirmed a 1:1 ratio for decavanadate-Ca(2+)-ATPase, with a dissociation constant, k(d) of 1 μM(-1). The interaction of decavanadate V(10)O(28)(6-) (V(10)) with Ca(2+)-ATPase is prevented by the isostructural and isoelectronic decaniobate Nb(10)O(28)(6-) (Nb(10)), whereas no significant effects were detected with ATP or with heparin, a known competitive ATP binding molecule, suggesting that V(10) binds non-competitively, with respect to ATP, to the protein. Finally, it was shown that decaniobate inhibits SR Ca(2+)-ATPase activity in a non competitive type of inhibition, with respect to ATP. Taken together, these data demonstrate that decameric niobate and vanadate species are stronger inhibitors of the SR calcium ATPase than simple monomeric vanadate, tungstate and molybdate oxometalates, thus affecting calcium homeostasis, cell signalling and cell bioenergetics, as well many other cellular processes. The ability of these oxometalates to act either as phosphate analogues, as a transition-state analogue in enzyme-catalysed phosphoryl group transfer processes and as potentially nucleotide-dependent enzymes modulators or inhibitors, suggests that different oxometalates may reveal different mechanistic preferences in these classes of enzymes.