An optimum strategy for solution chemistry using semiempirical molecular orbital method: Importance of description of charge distribution

Abstract

For the purpose to execute direct dynamics calculation in solution chemistry, we propose an optimum strategy for solution chemistry using semiempirical molecular orbital (MO) method with neglect of diatomic differential overlap (NDDO) approximation with specific solution reaction parameters (SSRP), i.e., the NDDO-SSRP method. In this strategy, the empirical parameters of the semi-empirical MO method were optimized individually for target molecule or ion by reference to the ab initio MO calculation data for many configurations on the potential energy surface near the reaction path. For demonstration, the NDDO-SSRP method was applied to two molecules and two ions (OH(-), H(2)O, NH(3), NH(4)(+)) at their equilibrium states in aqueous solution, respectively. Accordingly, it was verified that both the potential energy surface and the charge distribution of these solutes in aqueous solution are dramatically improved to reproduce themselves accurately at ab initio MO calculation level. In conclusion, it is expected that the NDDO-SSRP method should become quite useful for dynamic and statistical applications to chemical reaction systems in solution.