A cosolvency model to predict solubility of drugs at several temperatures from a limited number of solubility measurements

A cosolvency model to predict the solubility of drugs at several temperatures was derived from the excess free energy model of Williams and Amidon. The solubility of oxolinic acid, an antibacterial drug, was measured in aqueous (water+ethanol) and non-aqueous (ethanol+ethyl acetate) mixtures at several temperatures (20, 30, 35, 40 degrees C). Oxolinic acid displays a solubility maximum in each solvent mixture at solubility parameter values of 32 and 22 MPa(1/2). The temperature and heat of fusion were determined from differential scanning calorimetry. The solvent mixtures do not produce any solid phase change during the solubility experiments. The experimental results and those from the literature were employed to examine the accuracy and prediction capability of the proposed model. An equation was obtained to represent the drug solubility changes with cosolvent concentration and temperature. The model was also tested using a small number of experimental solubilities at 20 and 40 degrees C showing reasonably accurate predictions. This is important in pharmaceutics because it save experiments that are often expensive and time consuming.