Electronic reaction field cavity optimization: Extension to solvation of molecules

We have recently proposed a new quantum reaction field model that does not require the empirical specification of cavity size for spherical ions and molecules. When the solute wavefunction is optimized under the boundary condition that it is fully contained within the cavity, reaction field stabilization and repulsive kinetic energy compression are the competing forces that determine the optimal cavity size. When this model was investigated with the simplest reaction field method, the Born theory, we found that Born theory with wavefunction containment gives good agreement with experimental enthalpies of solvation for cations, but is highly imperfect for anions. Furthermore, that model was not extensible to molecules that are not spherically symmetric. Our new model relaxes the strict separation of solute and solvent electron densities to allow leakage of the solute wavefunction into the surrounding dielectric with a density functional penalty that is a function of solvent and solute electron density. in the limit of infinite solvent density we recover the original electronic reaction field cavity optimization model, thereby allowing us to predict energies and cavity sizes based solely on theoretical grounds for molecules of arbitrary symmetry. By interpreting the solvent density parameter in molecular terms, we show significant improvement for anion energies and radii for low solvent densities, while correct free energies for cations require high local densities like that of full containment. Future prospects for our parameter free model are discussed.