HYDRODESULFURIZATION OVER MODEL SULFIDE CLUSTER-DERIVED CATALYSTS

There is growing interest in the use of organometallic clusters as precursors to model heterogeneous catalysts. This interest is partially derived from the expectation that "solid-state metal complex catalysts" will be uniform in composition and structure thus more amenable to characterization than catalysts prepared using conventional techniques. This research used a series of sulfide clusters as precursors to model heterogeneous catalysts in an effort to enhance our understanding of the active sites in hydrodesulfurization catalysts. Alumina supported catalysts derived from the sulfide clusters Cp2Mo2(mu-SH)2(mu-S)2, Cp2Mo2Co2(mu-3-S)2(mu-4-S) (CO)4, and Cp2Mo2Fe2(mu-3-S)2(CO)8 (Cp = cyclopentadienyl) shared important catalytic and spectroscopic features with a commercial catalyst and unsupported MoS2. These strong similarities implied that the active sites in the sulfide cluster-derived catalysts could be used as models of the sites in conventionally prepared hydrodesulfurization catalysts. The active sites in the bimetallic sulfide cluster-derived catalysts appeared to be highly dispersed ensembles containing Mo-delta+ (delta < 4), cobalt or iron, and sulfur. The structure of these ensembles remains to be determined but our results provide evidence that the promotional effect in hydrodesulfurization catalysis is at least partly due to strong, direct interactions between molybdenum and the promoter. There were differences between the thiophene hydrodesulfurization product distributions. The sulfide cluster-derived catalysts produced significant amounts of C2 and C3 hydrocarbons while the commercial catalyst produced mostly C4 hydrocarbons. These results suggested that C-C bond hydrogenolysis preceded C-S bond cleavage over the sulfide cluster-derived catalysts. More subtle differences between the product distributions of the cluster-derived catalysts have been attributed to differences in the nature of sulfide ligands in their precursors.