Discovery of new high-pressure phases – integrating high-throughput DFT simulations, graph neural networks, and active learning

Researchers have combined high-throughput density functional theory (DFT) simulations with graph neural networks and active learning to accelerate the discovery of new material phases under high pressure. This integrated approach aims to efficiently explore the vast landscape of potential material structures and identify novel configurations that emerge in extreme pressure environments. By leveraging machine learning, the process can predict and prioritize promising candidates, significantly reducing the experimental and computational effort typically required for such discoveries.

This advancement could lead to the development of new materials with unique properties for various technological applications.