Predicting cosolvency for pharmaceutical and environmental applications

School of Public Health, University of Illinois at Chicago, Chicago, IL, USA 

Cosolvency refers to the effeets of adding one or more solvents (eosolvents), whieh are different from the existing solvent in a solution, on the properties of the solution or behavior of the solute. Cosolveney has found its applications in numerous engineering and scientific disciplines. The discussion in this section will be limited to aqueous phase cosolvency (the primary solvent is water), and cosolvents will include only pure organic solvents which are miscible with water either completely (in any proportion) or partially (in only certain proportions). The extent of cosolvency will be quantitatively described by the difference in solute solubilities in pure water and in a mixture of water and cosolvent(s). 

This discussion is intended to provide an easy approach for predicting the effect of cosolvents, which are frequently involved in pharmaceutical and environmental applications, on the solubility of organic chemicals. The starting point is the widely used log-linear model. 

---

A. Li. Predicting cosolvency for pharmaceutical and environmental applications. In Handbook of... 

Many models, however, are considered to be more descriptive than predictive, because they inevitably involve one or more model parameters which are usually specific to a particular solute/solvent/cosolvent(s) system, and must be estimated from experimental data of solubility obtained for that system. On the other hand, pinely empirical models, e.g., the double-log exponential equation, aim at satisfying mathematical descriptions of measured data, and often offer little insight to the process. Comparisons among cosol-veney models have been made in several published papers. This discussion is intended to provide an easy approaeh for predicting the effect of cosolvents, which are frequently involved in pharmaceutical and environmental applications, on the solubility of organie chemicals. The starting point is the widely used log-linear model. 

---