Early Transition Metal Oxides as Catalysts: Crossing Scales from Clusters to Single Crystals to Functioning Materials

The overall goal of this program is to investigate the electronic structure and chemical bonding of early transition metal oxide clusters and use them as well-defined molecular models to obtain insight into properties and mechanisms of oxide catalysts, as well as to provide accurate spectroscopic and molecular information to verify theoretical methods used to predict materials properties. A laser vaporization cluster source is used to produce metal oxide clusters with different sizes, structures, and compositions. Well-defined inorganic polyoxometalate clusters in solution are transported in the gas phase using electrospray. Two state-of-the-art photoelectron spectroscopy apparatuses are used to interrogate the oxide clusters and polyoxometalate anions in the gas phase to obtain spectroscopic and electronic structure information. The experimental effort is assisted by theoretical calculations to understanding the structures, chemical bonding, and catalytical properties of the transition metal oxide clusters. The research approach combines novel and flexible experimental techniques and advanced theoretical/computational methodologies and seeks molecular-level information to aiding the design of new catalysts, as well as mechanistic understanding. We have focused on the investigation of tungsten oxide clusters containing three W atoms: W{sub 3}O{sub x}{sup -} (x = 7-11). A number of interesting findings have been made. We observed that the oxygen-poor W{sub 3}O8 cluster contains a localized W{sup 4+} center, which can be used as a molecular model for O-deficient defect sites. A chemisorption energy was obtained through density functional calculations for W{sub 3}O8 + O{sub 2} {yields} W{sub 3}O{sub 10} as -78 kcal/mol. We further found that the neutral stoichiometric W{sub 2}O{sub 6} and W{sub 3}O{sub 9} clusters do not react with O{sub 2} and they only form physi-sorbed complexes, W{sub 2}O{sub 6}(O{sub 2}) and W{sub 3}O{sub 9}(O{sub 2}). However, the negatively charged W{sub 2}O{sub 6}{sup -} and W{sub 3}O{sub 9}{sup -} clusters are found to form chemisorbed complexes due to the presence of the extra electron. Thus, the W{sub 2}O{sub 6}{sup -} and W{sub 3}O{sub 9}{sup -} negative clusters can be viewed as models for O{sub 2} interaction with a reduced W site (W{sup 5+}) on the oxide surface. These studies also led to the surprising observation of the first d-orbital aromatic clusters in W{sub 3}O{sub 9}{sup 2-} and Mo{sub 3}O{sub 9}{sup 2-}, which each contains a completely delocalized three-center two-electron bond made entirely made of the metal d orbitals. This last result was highlighted in both Chem & Eng. News and Nature. We further studied a series of small metalate anions using electrospray, including the hydroxo and methoxo oxometalate MO{sub 3}(OH){sup -} and MO{sub 3}(OCH{sub 3}){sup -}, and the dimetalates: M{sub 2}O{sub 7}{sup 2-}, MM{prime}O{sub 7}{sup 2-}, and M{sub 2}O{sub 7}{sup -} (M, M{prime} = Cr, Mo, and W)