How to find an optimum cluster size through topological site properties: MoSxmodel clusters

Computational investigations in catalysis frequently use model clusters to represent realistically the catalyst and its reaction sites. Detailed knowledge of the molecular charge, thus electronic density, of a cluster would then allow physical and chemical insights of properties and can provide a procedure to establish their optimum size for catalyst studies. For this purpose, an approach is suggested to study model clusters based on the distributed multipole analysis (DMA) of molecular charge properties. After full density functional theory (DFT) geometry optimization of each cluster, DMA computed from the converged DFT one‐electron density matrix allowed the partition of the corresponding cluster charge distribution into monopole, dipole, and quadrupole moments on the atomic sites. The procedure was applied to MoS₂ model clusters Mo₁₀S₁₈, Mo₁₂S₂₆, Mo₁₆S₃₂, Mo₂₃S₄₈, and Mo₂₇S₅₄. This analysis provided detailed features of the charge distribution of each cluster, focused on the 101 0 (Mo or metallic edge) and 1 010 (sulfur edge) active planes. Properties of the Mo₂₇S₅₄ cluster, including the formation of HDS active surfaces, were extensively discussed. The effect of cluster size on the site charge distribution properties of both planes was evaluated. The results showed that the Mo₁₆S₃₂ cluster can adequately model both active planes of real size Mo₂₇S₅₄. These results can guide future computational studies of MoS₂ catalytic processes. Furthermore, this approach is of general applicability. © 2011 Wiley Periodicals, Inc. J Comput Chem, 2011