Prediction of the optimized solvent composition for solubilization of drugs in water-cosolvent mixtures

The capability of the Jouyban-Acree model for predicting the optimized solvent composition of binary solvents for solubilization of drugs is shown employing solubility of drugs in aqueous mixtures of dioxane, ethanol and polyethylene glycol 400. The established model constants of the Jouyban-Acree model and solubility of drugs in water and cosolvent are used to predict the maximum solubility of in the binary solvent mixture (log X-m(max)) and the corresponding solvent composition (f(1,max)). The accuracy of the predicted log X-m(max) and f(1,max) is studied using average absolute error (AAE) of predicted and observed values. The AAEs were 0.10 +/- 0.12 and 0.08 +/- 0.10, respectively for log X-m(max) and f(1,max). The method provided acceptable predictions and is recommended for practical applications. The main advantage of the proposed method is its extension to temperatures higher/lower than room temperature.