Overcoming the Chemical Complexity Bottleneck in on-the-Fly Machine Learned Molecular Dynamics Simulations.

We develop a framework for on-the-fly machine learned force field molecular dynamics simulations based on the multipole featurization scheme that overcomes the bottleneck with the number of chemical elements. Considering bulk systems with up to 6 elements, we demonstrate that the number of density functional theory calls remains approximately independent of the number of chemical elements, in contrast to the increase in the smooth overlap of the atomic positions scheme.

Approaches to Understanding the Solution-State Organization of Spray-Dried Dispersion Feed Solutions and Its Translation to the Solid State.

It is well established that polymers adopt a range of conformations and solution-state organization in response to varying solution environments, although very little work has been done to understand how these effects might impact the physical stability and bioavailability of spray-dried amorphous dispersions (SDDs). Potentially relevant solution-state polymer-solvent/cosolute interactions include preferential solvation, hydrodynamic size (i.e., polymer swelling or collapse), and solvent quality effects (i.e., attractive or repulsive self-interactions). Of particular interest is the investigation of preferential solvation, defined as the relative attraction or rejection of a cosolvent and/or cosolute from the local environment of a solvated macromolecule, which often occurs in multicomponent macromolecular solutions. As spray drying and other solvent-based dispersion processing necessitates the use of complex media consisting of at least three or more components (drug, polymer, solvent(s), and other possible excipients), the prevalence of this phenomenon is likely. This work characterizes largely unexplored solution-state properties in model spray-dried dispersion feed solutions using light scattering and viscometric techniques to add greater context and guidance in studying these information-rich materials. These systems are found to exhibit complex non-intuitive behavior, which serves to highlight the potential utility of preferential solvation in spray-dried dispersion processing and stability. It is hypothesized that solution-state organization of the liquid feed can be engineered and translated to the solid-state for the optimization of SDD properties.