Hydrotropy: binding models vs. statistical thermodynamics

Hydrophobic drugs can often be solubilized by the addition of hydrotropes. We have previously shown that preferential drug-hydrotrope association is one of the major factors of increased solubility (but not "hydrotrope clustering'' or changes in "water structure''). How, then, can we understand this drug-hydrotrope interaction at a molecular level? Thermodynamic models based upon stoichiometric solute-water and solute-hydrotrope binding have long been used to understand solubilization microscopically. Such binding models have shown that the solvation numbers or coordination numbers of the water and hydrotrope molecules around the drug solute is the key quantity for solute-water and solute-hydrotrope interaction. However, we show that a rigorous statistical thermodynamic theory (the fluctuation solution theory originated by Kirkwood and Buff) requires the total reconsideration of such a paradigm. Here we show that (i) the excess solvation number (the net increase or decrease, relative to the bulk, of the solvent molecules around the solute), not the coordination number, is the key quantity for describing the solute-hydrotrope interaction; (ii) solute-hydrotrope binding is beyond the reach of the stoichiometric models because long-range solvation structure plays an important role.