Grand canonical generative diffusion model for crystalline phases and grain boundaries
This addresses a critical bottleneck in materials science for generating complex atomic structures, though it is an incremental improvement over existing diffusion models.
The authors tackled the inability of existing particle-based diffusion models to generate ordered crystalline structures by developing a grand canonical diffusion model using a voxel-based representation with a variable number of particles, successfully generating common crystalline phases and grain boundary structures.
The diffusion model has emerged as a powerful tool for generating atomic structures for materials science. This work calls attention to the deficiency of current particle-based diffusion models, which represent atoms as a point cloud, in generating even the simplest ordered crystalline structures. The problem is attributed to particles being trapped in local minima during the score-driven simulated annealing of the diffusion process, similar to the physical process of force-driven simulated annealing. We develop a solution, the grand canonical diffusion model, which adopts an alternative voxel-based representation with continuous rather than fixed number of particles. The method is applied towards generation of several common crystalline phases as well as the technologically important and challenging problem of grain boundary structures.