Modeling crystal shape of polar organic materials: Applications to amino acids

Models for predicting the shape of organic crystals use geometrical rules and intermolecular interaction energies between the building blocks of the crystal but often ignore the environment in which the crystal grows. The resulting shapes predicted by these models may differ from the actual grown shapes since shape can be influenced by the solvent, impurities, and additional external factors. We present a model to predict the shape of polar organic materials crystallized from solution. The model is based on a Burton, Cabrera, and Frank growth mechanism and accounts for the solute-solvent interactions at the interface. Using this model, we have successfully predicted the shape of amino acids crystallized from solution. Amino acids were chosen as model compounds not only because they are biologically important but also because they are small organic molecules that exhibit various functional groups and have strong electrostatic interactions. Their modeling raised fundamental issues discussed in this article, including the description of charge density, determination of growth unit, selection of a force field, and estimation of crystal-solution interfacial free energy. The shape of alpha-glycine, when grown from the vapor and from aqueous solution, was predicted and compared with actual growth shapes and with predictions by others. The shape of L-alanine, when grown from aqueous solution, was predicted and compared with the experimental growth shape.