![](https://sandbox4-www.anl.gov/sites/www/files/styles/article_teaser_16x9/public/16x9_Reinforcement%20learning%20figurex.jpg?h=8abcec71&itok=a3iHWcRz)
- Systems and Methods for Generating Phase Diagrams for Metastable Material States (ANL-IN-19-012)
This invention introduces a new paradigm for materials science and solid-state chemistry by moving away from an explorative experimental synthesis, followed by phenomenological modeling approach, towards a predictive process by identifying phase space where metastable states occur from machine learning and first principles.