A Computational Study of Transition Metal Clusters for Dehydrogenation Catalysis

Developing more reactive and selective catalysts for petrochemical refining and synthesis, specifically the dehydrogenation of propane to form propylene, is extremely important for the US and global economy. Studies have suggested that sub-nanometer transition metal (TM) clusters can be synthesized with superior properties to those of the bulk metal. The use of alloy clusters can potentially tune their catalytic properties for specific reactions. Using computational methods, the dehydrogenation reaction pathways of possible catalysts with propane can be tested; however, this can be very time consuming. By looking at possible simple descriptors of promising catalytic behavior, the time spent testing specific alloys can be greatly reduced. We have found a possible correlation between the hydrogen binding energy of a four-atom TM cluster and its activation energy for the rate-limiting step of the propane dehydrogenation reaction. If such a correlation exists, the far less time consuming calculation of hydrogen binding energy could provide a clear indicator of catalytic activity. We have calculated the hydrogen binding energies for M4 (M = TM atom) as well as Pt4-x Mx alloy clusters, (x = 0-3) and have tried to understand how differences in these energies are related to properties of the clusters such as ionization energies, electron affinities, and cohesive energies. We report our results and their implications for the design of new dehydrogenation catalysts