The importance of polarisability in the modelling of solubility: quantifying the effect of charged co-solutes on the solubility of small non-polar solutes

H Docherty and PJ Dyer and PT Cummings, MOLECULAR SIMULATION, 37, 299-309 (2011).

DOI: 10.1080/08927022.2011.553225

In a recent work (Dyer et al., J. Chem. Phys. 129 (2008), 024508), we demonstrated that by accounting for the polarisability of small non- polar solutes, in an explicit manner, it is possible to approach quantitative agreement with experimental values of the excess chemical potential of the molecules in pure water. Here, we continue this line of research by considering the effects of ionic co-solutes (i.e. salts) on the solubility of an explicitly polarisable model suitable for a variety of small non-polar solutes. In doing so, we calculate the excess chemical potential that is related to the solubility of the solute in the solution, and investigate how the solubility of hydrophobic solutes varies with ion concentration, i.e. salting-out/salting-in effects, as measured by the Setchenow parameter. In particular, we consider the solubility of Ne, Ar, Kr, Xe and CH(4) in aqueous NaCl solutions for the TIP3P, TIP4P-Ew and TIP4P/2005 models of water. Using these models, we discover that the addition of explicit polarisability to the solute models decreases the excess chemical potential, but fails to obtain quantitative agreement with experimental values. In addition, we determine that explicit polarisation has minimal effect on the Setchenow parameter and, in turn, minimal effect on salting-out. Instead, we show that the over-prediction of the excess chemical potential is due to an inaccurate ion-solute interaction potential. Further to this, we demonstrate that, by accounting for polarisability explicitly, it is possible to obtain pair interaction potentials from ab initio calculations that perform as well as, or better than the models commonly used to study these systems.

Return to Publications page