Grant: Discovery Grants Program - Individual

The liquid state is both dense and disordered, making its properties more challenging to predict than a gas (disordered but not dense) or a crystalline solid (dense but not disordered). Despite their complexity, liquids are central in most of science, industry and the natural world, especially in the form of water and solutions based on water. Our research uses computer simulations to study liquids, and especially water, by modelling them at the molecular scale, in order to predict their properties and understand the processes by which they transform to the solid state. We study the properties of metastable supercooled liquids, that is, liquids that have been cooled below their freezing temperature but have not yet solidified. Supercooled water is a major focus of our work. Supercooled water has a large number of unusual properties and may undergo a novel liquid-liquid phase transition between two distinct forms. We will conduct simulations to support experimental efforts to find this phase transition. Supercooled water, and crystalline and amorphous ice, play a central role in many fields, including atmospheric and climate science research on droplet and ice formation in clouds, cryogenic preservation of food and tissue, control of ice formation on surfaces such as aircraft wings, and in the study of ice-rich planetary bodies in the outer solar system. Our research supports these fields by advancing our fundamental understanding of the properties of water. The processes that transform a metastable liquid to a solid are complex and can lead to an ordered crystal or a disordered glass. In many systems, instead of a crystal appearing directly from the liquid, an intermediate solid may form first, which then later converts to the final solid form. Such two-step nucleation processes have been identified in a wide variety of contexts, including the formation of clathrates from natural gas, the biomineralization process by which ocean organisms create shells, and the formation of protein aggregates associated with Alzheimer's and Parkinson's disease. We will use simulations to test molecular-scale theories of two-step nucleation that will advance our ability to predict and control these processes. We will also study new approaches for tracking the solidification process at the molecular scale using novel order parameters to detect solid-like local regions in a supercooled liquid. These solidity order parameters quantify the localization of a molecule by its neighbors without assuming that the local environment is crystalline or glassy. If successful, such solidity order parameters will provide simulations with a powerful tool to model solidification without biasing the outcome to a particular structure. This work will contribute to theoretical efforts to develop a unified understanding of solidification that encompasses the formation of both crystals and glasses from the supercooled liquid state.