Investigating the factors leading to crystallisation in drug melts and glasses

Objectives: 1) we will elucidate the structural connections between the melt and the resulting crystal form to explain why some crystal forms can only be obtained from the glass or melt, and 2) we will generate fundamental data on the glass or melt as a model for the solvent-free dense phase in solution crystallisation.

Crystallisation from the neat compound in its liquid or solid amorphous phase, the melt or glass, respectively, can yield crystal forms that are not available through solution crystallisation. This long-known phenomenon has gained renewed interest in recent years, where it could be shown that well-examined compounds can be crystallised in novel crystal forms even after decades of study.

The reasons for this ability are not yet clear and many studies concentrate on the nucleation and growth step of the melt crystallisation with few considering molecular interactions, density, and local ordering. In addition, the amorphous phase has important applications, such as in drug delivery by increasing bioavailability of poorly soluble drugs.

Here, crystallisation has to be avoided at any cost, but despite recent exciting studies, why and when a material crystallises is poorly understood. In this project we will fill the following knowledge gap: What are the molecular interactions in the neat liquid that drive specific crystal forms to crystallise?

We will follow the intermolecular interactions in melt or glass phases using variable temperature Fourier Transform Infrared Spectroscopy (FTIR) and variable temperature Magic Angle Spinning (MAS) ssNMR to detect molecular interactions (including hydrogen bonding and [PIE]-stacking), mobility and flexibility in the melt. From the shifts observed by both techniques, we can obtain an aggregation constant and calculate the energy of aggregating.

Through the generation of van't Hoff plots, we can follow if the aggregation mode is temperature depending and changes at specific points. Using ssNMR, we will additionally follow the spin-lattice relaxation times (T1, T1p) to obtain information about molecular mobility and specific rotations present in the amorphous phase.

In parallel to the spectroscopic studies, we will develop a structural model of the melt that allows us to probe intermolecular interactions and clustering to support our interpretation of the spectroscopic data. For this we will use neutron total scattering experiments in combination with Monte Carlo simulations tethered to the experimental data. Total scattering methods yield structure factors, the Fourier transform of the related pair distribution functions (PDFs), which are the distribution of all intra- and intermolecular atomic distances in a sample. They are a powerful tool in the investigation of local order.