A Combined Molecular Simulation and Group Contribution Method to Efficiently Predict the Solubility of Nonelectrolyte Solids

Knowledge of the solubility of pharmaceutical solids in a wide range of solvents is essential for all steps of the drug discovery and development process. For drug discovery, one is interested in the design of novel compounds, for which experimental data does not exist, complicating the development process. In the present talk, our efforts to predict the solubility of nonelectrolyte solids in the absence of experimental data will be discussed. The solution phase behavior of the solute is modeled in atomistic detail via molecular simulation. Further, we will show how established solution theories may be used to supplement molecular simulation. The necessary pure solid phase properties of the solute are calculated using group contribution methods. By complementing molecular simulation with solution theories and group contribution methods, solubility predictions may be may in an extremely efficient fashion. In addition, the molecular simulations enable insight into the underlying driving forces behind the solubility process.

Results will be presented for acetaminophen in a range of pure and mixed solvents. Comparison will be made to conventional MOSCED and UNIFAC based methods. Overall, the molecular simulation based predictions are in good quantitative agreement with experimental data, and are comparable to conventional methods.