Redox-active silica-supported catalysts can be used to create alternatives to stoichiometric (i.e., non-catalytic) processes that are widely used in oxidation reactions and often require toxic and/or environmentally-unfriendly reagants. Catalytically active metal sites can be introduced to a solid support either as part of the framework of porous materials during their synthesis (co-condensation), or through post-synthetic attachment. The strategy of catalyst immobilization emphasized in this work involves linking a metal complex to a silica material via a grafted spacer or tether. This approach requires modifying one of the ligands of the catalyst precursor with a functional group or linker capable of creating the anchor, via a covalent bond, to the silica support. Much of the synthetic effort is devoted to preserving structure and reactivity of the molecular complex upon anchoring.

In this work, three silica-supported complexes were prepared. Each was characterized in detail at each step of the attachment procedure, to verify the reaction and its completeness. The corresponding homogeneous catalysts were also prepared in order to make structural and reactivity comparisons.

First, we report a bis(8-hydroxyquinolinato)(2-propanato)oxovanadium(V) catalyst, V(O)(HQ)2(OiPr), that selectively oxidizes alcohols under mild reaction conditions (60 °, air) and operates via an unusual base-assisted mechanism. A silica-supported analog, V(O)(HQ)2(OiPr) x-A380, was prepared via a multi-step grafting procedure and tested for catalytic aerobic oxidation under similar conditions as the homogeneous system. The activity of V(O)(HQ)2(OiPr)x-A380 was much lower than that of V(O)(HQ)2(OiPr), suggesting that the two are not precisely structurally equivalent. Further analysis reveals that during the grafting procedure, unexpected oxidation of the ligand tether occurred, giving rise to a V(IV) product.

In a separate study, we prepared two silylpropyl bis(pentafluorophenyl)boranes as molecular models for the Lewis acid component of supported frustrated Lewis pairs. The borane precursors were thoroughly characterized using NMR to confirm their structures, and to provide reference signals for silica-supported analogs. Three types of supported propyl bis(pentafluorophenyl)boranes, were prepared using a two-step methodology. First, allyl propylsilanes were installed, followed by hydroboration using Piers' borane to give tethered Lewis acid sites. As for molecular FLPs, combination with a Lewis base could lead to applications as heterogeneous metal-free catalysts for organic synthesis.

Lastly, two periodic mesoporous organosilicas (PMO) were synthesized by co-condensing either phenylpyridyl (ppy) silane or bipyridyl (bpy) silane, into the silica framework. The two materials were metalated using Cp*Ir(OH2) 3(OTf)2 to give Cp*Ir-ppy-PMO and Cp*Ir-bpy-PMO, respectively, as supported analogs for the molecular catalysts Cp*Ir(ppy)(OTf) and Cp*Ir(bpy)(OTf)2. When tested in the catalytic oxidation of propionaldehye using water as the oxygen source (aldehyde-water shift catalysis, AWS), the activities of the supported catalysts were comparable to those of the molecular systems, although all had the activities that are rather low. In spite of the low activity seen for AWS catalysis, Cp*Ir-ppy-PMO proved competent in the hydrogenation of levulinc acid.

We describe approaches for the design and synthesis of silica-supported molecular catalysts, using ligand anchoring techniques to create site-isolated, discrete catalyst.