Highlights
Redefining the state of the art in crystal free energy calculations
The ability to accurately predict crystal form stability is crucial in the pharmaceutical industry as it directly impacts the efficacy, stability, and bioavailability of drugs. Different crystal forms of the same molecule can exhibit vastly different solubility and stability profiles depending on environmental factors, such as temperature and humidity, which can affect the drug's effectiveness and shelf life.
The study compares the relative stability of hydrates and anhydrates, which are compounds with and without water molecules incorporated into their crystal structure. Predicting the conditions under which different solid forms of these materials transition from one to the other is a complex but critical task for the pharmaceutical industry. Any meaningful prediction of these transitions relies on accurate free energy calculations and their validation by experimental measurements. This data, however, is very difficult to find and is rarely hosted publicly. To address this bottleneck, Dr. Marcus Neumann and his team at AMS, together with several partners from industry and academia, have compiled the first reliable experimental benchmark for solid-solid free energy differences of industrially relevant compounds. They demonstrate that their computational methodology, leveraging high performance computers to perform quantum chemistry calculations with FHI-aims, can accurately predict these free energy differences without any empirical input.
However, the authors acknowledge that there is still room for improvement. The current method allows for the prediction of probable phase transition temperatures between two crystal forms with an accuracy on the order of about 100 K. To predict phase-transition temperatures to within 10 K, a further improvement in the accuracy of the free energy differences by a factor of 10 is required. This presents a challenging yet vital goal for future research in this field. As computational tools continue to evolve, their role in materials science research is becoming increasingly critical, promising new innovations and applications in various fields.
“FHIaims has proven to be a game changer for accurate lattice free energy calculations because of its scalability, numerical accuracy and small basis set superposition error. Of particular value for the TRHu(ST) method is the efficient handling of the PBE0 functional.“
- Marcus Neumann, Founder and CEO of Avant-Garde Materials Simulation
The full reference for the published article is:
Firaha, D., Liu, Y.M., van de Streek, J. et al., Nature 623, 324–328 (2023), DOI: 10.1038/s41586-023-06587-3