(67c) Adding Complexity, and Perhaps Cooperativity, to Dispersed Metal Oxide Catalysts

Highly dispersed metal oxide domains either on supports or in oxide frameworks, e.g. TS-1 or Sn-beta, are useful Lewis acid catalysts. Beginning with work in the Iglesia and Katz groups and continuing to the present day, Notestein developed methods for using bulky, multidentate oligophenols (calix[n]arenes) as ligands to graft dispersed Ti, Ta, Fe, and other metals on oxide supports. These materials are proficient at Lewis-acid catalyzed alkane and alkene oxidation. In the case of Ta-silicasynthesized by this method, the epoxidation of cyclohexene by hydrogen peroxide was independent of surface density at ~100 TON/h and occurred with less than 5% allylic oxidation, in contrast both with more typical Ta precursors and with Ti synthesized by a similar method.

Recently, the Notestein group has sought to expand this method to study these well-behaved isolated sites working alongside, and perhaps cooperatively, with other functional groups. The majority of this talk will briefly discuss our recent advances in three areas. First, added Lewis base sites to this surface via aminopropyltriethoxy silane (APTES) or a carbamate-protected form of the same silane, and we have probed emergent cooperative interactions and the avoidance of poisoning by base pairing using UV-visible, XANES, carbon dioxide uptake, C-C bond formation, and other reactions as probes. Secondly, we have added low loadings of Pd to create new catalysts for quinoline hydrodenitrogenation. Having known, controllable amounts of Lewis acid sites, in contrast to being constrained by those found naturally on the support, is allowing detailed mechanistic investigations changes product selectivity from primarily fully saturated compounds (propylcyclohexane), to 50% or greater propylbenzene. Finally, we are using these synthesis methods to understand photocatalytic oxidation of benzyl alcohols and the photocatalytic reduction of carbon dioxide. Here, a tradeoff is seen between needing larger crystalline domains for light harvesting (manifested as absolute rates increasing with crystallinity), and large numbers of exposed sites for reactivity (manifested as moderately increasing turnover frequencies at increasing dispersion). Overall, these studies show that controllable, systematic syntheses of supported metal oxide domains are essential to understanding the relative severities of various needs of multifunctional materials and controlling their interplay.