14 examples of how LLMs can transform materials science and chemistry: a reflection on a large language model hackathon.

Large-language models (LLMs) such as GPT-4 caught the interest of many scientists. Recent studies suggested that these models could be useful in chemistry and materials science. To explore these possibilities, we organized a hackathon. This article chronicles the projects built as part of this hackathon. Participants employed LLMs for various applications, including predicting properties of molecules and materials, designing novel interfaces for tools, extracting knowledge from unstructured data, and developing new educational applications. The diverse topics and the fact that working prototypes could be generated in less than two days highlight that LLMs will profoundly impact the future of our fields. The rich collection of ideas and projects also indicates that the applications of LLMs are not limited to materials science and chemistry but offer potential benefits to a wide range of scientific disciplines.

Solvent-Assisted Desorption of 2,5-Lutidine from Polyurethane Films.

A fundamental understanding of chemical interactions and transport mechanisms that result from introducing multiple chemical species into a polymer plays a key role in the development and optimization of membranes, coatings, and decontamination formulations. In this study, we explore the solvent-assisted desorption of a penetrant (2,5-lutidine) in polyurethane with aprotic (acetonitrile) and protic (methanol) solvents. Chemical interactions between solvent, penetrant, and polymer functional groups are characterized via time-resolved Fourier transform infrared spectroscopy (FTIR) during single and multicomponent exposures. For both solvents, an increase in the extraction rate of the penetrant is observed when the solvent is applied during desorption. Inspection of the FTIR spectra reveals two potential mechanisms that facilitate the enhanced desorption rate: (1) penetrant/solvent competition for hydrogen donor groups on the polymer backbone and (2) disruption of the self-interaction (cohesive forces) between neighboring polymer chains. Finally, the aprotic solvent is found to generate an order of magnitude greater desorption rate of the penetrant, which is attributed to a greater disruption of the self-interaction during penetrant desorption compared to the protic solvent and the inability of an aprotic solvent to form larger and potentially slower penetrant-solvent complexes.

A molecular dynamics study of the role of molecular water on the structure and mechanics of amorphous geopolymer binders.

In this paper, molecular dynamics simulations are used to study the effect of molecular water and composition (Si/Al ratio) on the structure and mechanical properties of fully polymerized amorphous sodium aluminosilicate geopolymer binders. The X-ray pair distribution function for the simulated geopolymer binder phase showed good agreement with the experimentally determined structure in terms of bond lengths of the various atomic pairs. The elastic constants and ultimate tensile strength of the geopolymer binders were calculated as a function of water content and Si/Al ratio; while increasing the Si/Al ratio from one to three led to an increase in the respective values of the elastic stiffness and tensile strength, for a given Si/Al ratio, increasing the water content decreased the stiffness and strength of the binder phase. An atomic-scale analysis showed a direct correlation between water content and diffusion of alkali ions, resulting in the weakening of the AlO4 tetrahedral structure due to the migration of charge balancing alkali ions away from the tetrahedra, ultimately leading to failure. In the presence of water molecules, the diffusion behavior of alkali cations was found to be particularly anomalous, showing dynamic heterogeneity. This paper, for the first time, proves the efficacy of atomistic simulations for understanding the effect of water in geopolymer binders and can thus serve as a useful design tool for optimizing composition of geopolymers with improved mechanical properties.

Interplay between structure and transport properties of molten salt mixtures of ZnCl2-NaCl-KCl: A molecular dynamics study.

Molten mixtures of network-forming covalently bonded ZnCl2 and network-modifying ionically bonded NaCl and KCl salts are investigated as high-temperature heat transfer fluids for concentrating solar power plants. Specifically, using molecular dynamics simulations, the interplay between the extent of the network structure, composition, and the transport properties (viscosity, thermal conductivity, and diffusion) of ZnCl2-NaCl-KCl molten salts is characterized. The Stokes-Einstein/Eyring relationship is found to break down in these network-forming liquids at high concentrations of ZnCl2 (>63 mol. %), while the Eyring relationship is seen with increasing KCl concentration. Further, the network modification due to the addition of K ions leads to formation of non-bridging terminal Cl ions, which in turn lead to a positive temperature dependence of thermal conductivity in these melts. This new understanding of transport in these ternary liquids enables the identification of appropriate concentrations of the network formers and network modifiers to design heat transfer fluids with desired transport properties for concentrating solar power plants.