First Principle Theory for pKa Prediction at Molecular Level: pH Effects Based on Explicit Solvent Model

Abstract

pKa is one of the fundamental properties of molecules and its accurate prediction by theoretical method is indispensable in the current era. At present, the most common approach is based on the free energy difference evaluated in dielectric continuum model for pure water, epsilon=80, which completely ignores ionic influence, i.e., ionic strength. In the present paper, a molecular level theory to predict pKa is proposed based on the reference interaction site model, which is a statistical mechanics for molecular liquids. By regarding an acidic aqueous solution as a three component system including water, proton species (cation), and anion, aqueous solutions with desired pH can be theoretically realized by controlling the number density of the proton species. Using computed free energy changes on the deprotonation of glycine at various pH conditions, a titration curve and pKa can be obtained from the first principle, showing excellent agreement with experimental data. To our best knowledge, this is the first theoretical attempt to directly evaluate pKa under the condition where ionic influence is explicitly taken into account by using statistical molecular theory.