An Improved Thermodynamic Model to Experimentally Determine the Solubility and Interaction Parameter of Drug-Polymer Binary Systems Using Differential Scanning Calorimetry

Indomethacin-PVP K29/32 systems with drug weight fractions fd from 0.5-90% w/w were prepared by annealing physical mixtures in a conventional oven for 48 hours. Differential scanning calorimetry (DSC) was performed to determine observed heats of melting h * m of undissolved crystalline drug in the polymer. Plots of h * m vs. fd were made, and the drug-in-polymer solubility was taken as the horizontal intercept (corresponding to the highest drug load with no observed melting). The drugpolymer interaction parameter was obtained from the solubility using a recently published model (Bellantone et al. 2012), which specifically accounts for the effects of heat capacity differences between solid solutions and physical mixtures. Results Plots of h * m vs. fd were linear, and the experimental solubilities of indomethacin in PVP at 100C, 115C and 130C were 29%, 36% and 40% w/w, respectively, which were in agreement with solubilities predicted when heat capacity effects are included. Using these data in the model, corresponding interaction parameter values of were 0.86, 0.69 and 0.56. The positive values of these interaction parameters are in disagreement with those obtained by neglecting the heat capacity effects, which were negative. Conclusion A new experimental method was successfully developed to determine drug-in-polymer solubilities, which is in agreement with predictions of a recently published method that accounts for heat capacity differences between solid solutions and physical mixtures. This method leads to positive values for drug-polymer interaction parameters, which may be more reasonable than negative values when considering hydrophobic drugs and hydrophilic polymers.