A key tenet of physics is that the structure, the spatial arrangement of the atoms, of any given material determines its properties. A significant impediment to understanding material properties comes from the fact that the structures themselves are intrinsically dependent on the physical properties of their constituent atoms, and the motions of the electrons about them. The best currently available measurements for the structure of a material involve the simultaneous measurement of some 1020 atoms, whereas highly accurate quantum mechanical calculations of atomic and electronic properties deal typically at most with hundreds of atoms, owing to the extreme computational cost required for this level of accuracy.
Researchers from the Centre for Science at Extreme Conditions (CSEC) have developed a new computational approach that offers a holistic, multiscale view of disordered materials such as glasses and fluids, from large-scale, bulk structural correlations down to the properties of the constituent atoms and their electronic state. Implemented in a unified software framework, AIASSE, the new method enables direct, on-the-fly feedback between the interpretation of experimental measurements (X-ray or neutron total scattering) and highly accurate quantum mechanical calculations (Born-Oppenheimer Molecular Dynamics). The broad range of applicability is demonstrated through case studies ranging from simple fluids (krypton), through atomic (silica) and molecular glasses (amorphous ice), to complex mixtures such as water-methanol.
The framework provides a novel way to study disordered systems, enabling to discover previously unexplored structure-property relations and emergent phenomena in complex materials. The new method aims to break down the longstanding divide between experiment and theory, combining the strengths of both to get the best possible understanding of the material of interest.
