Solvation of Li+ by argon: how important are three-body forces?

A new analytical potential for Li+Ar2 including three-body interactions has been modeled by employing ab initio energies that were calculated within the CCSD(T) framework and a quadruple-zeta basis-set (i.e., cc-pVQZ for lithium and aug-cc-pVQZ for argon) and, then, corrected for the basis-set superposition error (BSSE) with the counterpoise method. Departing from this function, we have constructed the potential energy surface for Li+Arn clusters by summing over all two-body and three-body terms. We have employed our evolutionary algorithm (EA) to perform a global geometry optimization that allows for the study of a Li+ ion microsolvated with argon atoms. For the smaller clusters, the putative global minimum geometry obtained for the analytical potential has been used as a starting point for an ab initio optimization at the MP2 level. For clusters up to n = 10, the energetics and structures from the analytical potential energy surface (PES) that includes three-body interactions show good agreement with the corresponding ones optimized at the ab initio level. Removing the three-body terms from the analytical PES leads to global minima that fail to represent the main energetic features and the structures become wrong in the case of the Li+Ar2, Li+Ar3 and Li+Ar10 clusters. For n > 10, the comparison between potentials with and without three-body forces shows significant structural and energetic differences for most of the cluster sizes.