(600f) Morphology Prediction of Crystal Grown from Mixture Solvents

Crystal morphology refers to those crystallographic planes and their Miller indices that appear on the crystal surface, which can significantly affect the bioavailability for drugs, efficiency for catalysts, as well as sensitivity for high energetic materials, etc. For example, needle-shaped crystals would increase downstream processing costs and sensitivity to explosives, so they are typically undesirable for pharmaceuticals and high energy materials. Crystal morphology is determined by the external growth environment and internal structure, which can be controlled by changing temperature, supersaturation, solvent, additives, etc. The solvent is perhaps the most important external influence on the crystal morphology. Correctly interpreting solvent effects is essential to engineer the desired crystal morphology with specific properties.

Our group at UCSB developed ADDICT software to predict crystal morphology [1]. Later we upgraded its framework [2] and completely rewrote the code based on Matlab using object-oriented programming in ADDICT3 (the third version of ADDICT) [3]. Recently, we added different force fields into ADDICT so that it can be used for predicting the morphology of various organic crystals containing a wide variety of atom types [4]. Most recently, we also developed a method in ADDICT to predict the crystal morphology grown from a single solvent based on quantum chemistry solvation models (e.g., COSMO [5], SMD models [6], etc.). Considering that solvent mixtures can provide more flexible options to obtain the desired crystal morphology, e.g., antisolvent operation, it is desirable to extend crystal morphology prediction to mixed solvents.

In this work, we propose a method to predict the morphology of crystals grown from binary solvent mixtures using ADDICT based on quantum chemistry solvation models. We systematically tested the morphology of several crystals grown from binary mixture solvents (e.g., hexanitrohexaazaisowurtzitane (CL-20) grown from ethyl acetate and benzene, doravirine (MK-1439, form I) grown from ethanol and water, etc.). The predicted crystal morphologies are in good agreement with the experimental results indicating our proposed method is reliable. Furthermore, the approach proposed in this study can be extended to ternary solvent mixtures.

References

[1] Li, J., Tilbury, C. J., Kim, S. H., & Doherty, M. F. (2016). Prog. Mater. Sci., 82, 1.

[2] Zhao, Y., Tilbury, C. J., Landis, S., Sun, Y., Li, J., Zhu, P., & Doherty, M. F. (2020). Cryst. Growth Des., 20, 2885.

[3] Landis, S., Zhao, Y., & Doherty, M. F. (2020). Comput. Chem. Eng., 133, 106637.

[4] Zhao, Y., Gee R., & Doherty, M. F. (2023). AIChE J. ( accepted).

[5] Zhao, Y., Guo, Y., Gee R., & Doherty, M. F. (2023). AIChE J. (in preparation).

[6] Zhao, Y., Gee R., & Doherty, M. F. (2023). AIChE J. (in preparation).