Investigation of Lamotrigine Crystal Polymorphs Using Crystal Structure Prediction and Terahertz Spectroscopy

Abstract Crystal structure prediction (CSP) plays a significant role in exploring the crystal polymorphism of pharmaceutical solids. Determination of the atomic-level structures of crystalline materials is a fundamental step in understanding their bulk physical and chemical properties and behavior in drug products. CSP can be performed using numerous computational approaches, but all ultimately require comparison with experimental measurements to confirm the CSP findings. In this work, terahertz (THz) vibrational spectroscopy is combined with powder X-ray diffraction (PXRD) to demonstrate a new metrics framework for the validation of CSP investigations. Terahertz spectroscopy probes the low-frequency (≤150 cm−1) lattice vibrations of molecular crystals and provides a polymorph specific spectral profile that complements diffraction methods. This methodology is applied to the anti-epileptic drug lamotrigine, the subject of multiple crystal engineering studies to improve its formulated performance. A CSP search was performed on lamotrigine starting with a fast molecular mechanics survey for initial crystal structure candidates (∼800,000 structures), followed by high-accuracy solid-state density functional theory to refine the most promising candidates. The accuracy of the CSP output was evaluated by numerical comparison of simulated and experimental PXRD and THz datasets, with both benchmarks unambiguously indicating the lowest energy CSP candidate to be a structural match to the published lamotrigine polymorph (Form I, C 2/c space group). The success of the overall approach highlights THz spectroscopy as an effective orthogonal experimental benchmark for CSP and the combined experimental validation process represents a compelling procedure for advancing predictive solid-state modeling of pharmaceutical materials.