DISSOLVE: Database of ionic solutes? solvation free energies

Predicting values of the solvation free energy of ionic solutes is essential in chemical and biological processes as this property is required to model kinetics, electrochemical potentials, and acidity constants; however, developing computational prediction methods is challenging, partly because of the lack of accurate reference data. The state-of-the-art Minnesota solvation database reports an uncertainty of 3.0 kcal mol-1 in the absolute solvation free energy of ionic solutes, which is large for further method development and validation. To reduce this uncertainty, we have created an accurate reference database, the database of ionic solutes' solvation free energies (DISSOLVE). It consists of 330 ionic solvation free energy data entries of various chemical classes in water, dimethyl sulfoxide, methanol, and acetonitrile. The solvation free energy of these ionic solutes is determined by combining measured and computed values of the physical properties. We have extracted aqueous solvation energies of neutral solutes and pKa values from experimental studies in the literature. Using ab initio methods, we compute the gas phase acidity and non-aqueous solvation energy values of neutral solutes. On the basis of our analysis, we find that DISSOLVE entries are more accurate than the previously reported reference data in the literature, achieving a lower uncertainty of 1.5 kcal mol-1 for aqueous and 2.6 kcal mol-1 for non-aqueous absolute solvation free energies of ionic solutes. Additionally, we find an uncertainty of 1.0 kcal mol-1 and 1.1 kcal mol-1 for the conventional solvation free energies of ionic solutes in aqueous and non-aqueous solutions, respectively. With this improvement, our database enables the prediction of equilibrium and reaction rate constants within an order of magnitude, as well as the development and validation of new computational chemistry approaches for predicting the solvation free energy of ionic solutes.