Additional Contributions to Interionic Forces

The solvent-averaged pair potential as given by Eqs. (6) and (4) can be expected to be quite realistic for large r or for small r. In the range 

If we represent the solvent as a structureless fluid (in this paragraph only) characterized by a dielectric constant e, then CaCblre is the exact potential function for two point charges or for two conducting spheres (for r sufficiently large). If we now think of each ion as a sphere of dielectric with a charge at its center, then classical electrostatics predicts a term due to the polarization of each ion by the charge and induced moments of the other. The first approximation to this term, called CAVo (r), has been included in many ionic solution calculations but seems to be quite unimportant.  

London dispersion contributions to Uab should be especially important when the ions are relatively large. Taking them into account in a consistent manner, however, would also require calculation of the solute-solvent dispersion forces and this has not been done. When the dispersion interactions are significant and the model potential in Eq. (7) is used, then their effect may be expected to show up in the species dependence of the Gurney parameters, as is indeed the case.  

---