The reactivity of CO on bimetallic Ni3M clusters (M = Sc, Ti, V, Cr, Mn, Fe, Co, Cu, Rh, Ru, Ag, Pd and Pt) by density functional theory

This work concerns the adsorption and dissociation of CO on doped Ni3M clusters, where M = Sc, Ti, V, Cr, Mn, Fe, Co, Cu, Rh, Ru, Ag, Pd and Pt, applying density functional theory to search for the best catalyst for CO dissociation. Butterfly geometry is observed for Ni3Sc, Ni3Ti and Ni3Cr, while the other clusters retain the distorted tetrahedral geometry. The presence of odd numbers of valence electrons, e.g. in Ni3Sc, Ni3V, Ni3Mn, Ni3Co, Ni3Cu, Ni3Rh and Ni3Ag clusters, allows open shell structures to show greater stability than closed shell clusters. However, the higher stability of the closed shell Ni3Ti cluster is due to the presence of a greater number of polar M-M bonds. Adsorption of CO on all the doped clusters is thermodynamically feasible under standard conditions. Dissociation is not thermodynamically feasible for any of the clusters, except on Ni3Sc. Though doping of any of the metals into the Ni-4 cluster decreases the CO dissociation barrier, none of the bimetallic clusters are good catalysts for CO dissociation. The formation of TiO, VO, CoO and MnO species reduces the catalytic activity of the doped Ni-4 cluster. Surface C binds at threefold coordination sites for most of the clusters. However, for the Ni3Fe cluster, surface C is tetra-coordinated and strongly binds to Fe, forming a sigma and a pi bond, and its greater s character favors hydrogenation. Surface O in Ni3Cu binds at threefold sites and has unpaired electrons, making it radical in nature, and more reactive than surface O in the Ni3Fe cluster. Greater positive Delta G values, the lower stability of the doped clusters, and a tendency toward adsorption all reduce the efficiency of the catalyst. Thus, Ni3Fe and Ni3Cu appear to be the best catalysts for CO dissociation.