Prediction of solubility of active pharmaceutical ingredients by semi-predictive Flory Huggins/Hansen model

In this work, solubility of four Active Pharmaceutical Ingredients (APIs) including Butyl Paraben, Fenoxycarb, Fenofibrate and Risperidone were predicted using Hansen Flory Huggins model using two different scenarios. In the first method, activity coefficient of APIs were obtained through fitting the experimental activity coefficients of solvents at particular temperature of 293 K, then components solubility in entire temperature range of study was predicted. In the second scenario, the model parameters were adjusted using experimental data of two selected solvents, then components solubility were predicted in other solvents. In order to check the physical meanings of obtained values, Molecular Dynamic (MD) simulations was utilized and the results were compared. Finally the predictive capabilities of two Hansen Flory Huggins models were compared to temperature dependent NRTL-SAC model. (C) 2017 Elsevier B.V. All rights reserved.