A Digital Workflow Supporting the Selection of Solvents for Optimizing the Crystallizability of p-Aminobenzoic Acid

We present a grid-based molecular modeling approach and software application for screening the solute–solvent and solute–solute interactions of organic molecules. This tool can provide a deeper understanding of solubilization of organic molecules, intended to guide scientists to intuitive conclusions about whether a solute/solvent pair may provide desired physical properties, such as crystallizability, solubility, and crystal polymorphism. This study focused on solutions of p-aminobenzoic acid in acetonitrile, ethanol, and water. Acetonitrile molecules are found to form the weakest interactions with the solute molecule, although they also form weak interactions with themselves. In contrast, water forms strong interactions with the solute molecule, with a strong preference to interact with the carboxylic acid group, although they also form strong self-interactions. Ethanol forms strong interactions with all of the solute molecules, along with reasonably strong interactions with itself. The looser solvation of the carboxylic acid group by acetonitrile is thought to drive the crystallization of the α polymorph, by lowering the crystallization kinetic energy barrier. In ethanol, the strong interactions of the solvent are thought to contribute to significant undercooling of ethanolic solutions observed in previous studies. Water’s strong interactions with the carboxylic acid of the solute may drive the self-assembly of the α-form by interactions of the phenyl groups and also contribute to the nucleation of the β-form from this solvent. This workflow can provide valuable guidance on the solvation properties of organic molecules and clusters, producing low-energy solvation shells of molecules and clusters to be utilized as starting points for more sophisticated simulations, such as molecular dynamics.