Glass transition temperature (Tg) is an important material property, which predetermines the kinetic stability of amorphous solids. In the context of active pharmaceutical ingredients (API), there is motivation to maximize their Tg by forming amorphous mixtures with other chemicals, labeled excipients. Molecular dynamics simulations are a natural computational tool to investigate the relationships between structure, dynamics, and cohesion of amorphous materials with an all-atom resolution. This work presents a computational study, addressing primarily the predictions of the glass transition temperatures of four selected API (carbamazepine, racemic ibuprofen, indomethacin, and naproxen) with two nucleobases (adenine and cytosine). Since the classical non-polarizable simulations fail to reach the quantitative accuracy of the predicted Tg, analyses of internal dynamics, hydrogen bonding, and cohesive forces in bulk phases of pure API and their mixtures with the nucleobases are performed to interpret the predicted trends. This manuscript reveals the method for a systematic search of beneficial pairs of API and excipients (with maximum Tg when mixed). Monitoring of transport and cohesive properties of API-excipients systems via molecular simulation will enable the design of such API formulations more efficiently in the future.

• Keywords
• active pharmaceutical ingredients, amorphous dispersion, glass transition, molecular dynamics,
• Publication type
• Journal Article MeSH