Semiempirical quantum chemical calculation of the energy of interaction of solute molecules with solvent

Computed stabilities for a few small molecules solvated in an aqueous medium have been reported. The methods adopted are CNDO/2 and ab-initio Hartree-Fock treatment. The CNDO/2 stabilities arising from the medium polarization effects compare favorably with the corresponding ab-initio stabilities for the same species. The heat of solution is determined theoretically for each of the gaseous species Cl- ion, H+ ion and hydrogen chloride. The CNDO/2 hydration numbers, 1 for Cl- ion and < 1 for H+, ion, are in good agreement with experiment. The CNDO/2 parameters are not adequate enough to yield the stability of negative ions with closed-shell electronic configurations. Hence the CNDO/2 calculation of the heat of solution of gaseous hydrogen chloride requires the solvent molecules to be extensively associated with the ions. The total hydration number required for the computed enthalpy change to validate the experimental heat of solution is greater than 5. The ab-initio calculations ape carried out with different basis sets [6-31g, 6-31g(d), mixed basis and 6-311+g(2d,p)], and with or without electron correlation effects. The results are only mildly encouraging. They all point out the need for considering a more extensive basis set and a rather large hydration number (greater than or equal to 3 for Cl-, and 2 for H+). The hydration number 2 is indicative of the mechanism by which protons achieve an abnormally high mobility in water. The total hydration number required by the 6-31g: calculations to explain the experimental heat of solution of HCl is only 3. The 6-31g(d) calculations need a total hydration number greater than 5 that is in agreement with the hydration numbers predicted ab-initio for individual ions. A good value for the enthalpy change is easily obtained from a mixed basis calculation on the solvation of HCl. Nevertheless the Hartree-Fock treatment requires a total hydration number of about 3 for both chloride and hydrogen ions whereas the fourth-order Moller-Plesset treatment needs a total hydration number around 2. This clearly indicates that the mixed basis set is less reliable for the computation of a thermochemical property as the agreement between the theoretical and the experimental results is basically fortuitous.