A high throughput computational framework applied to advanced battery materials
This project will use the massive, multilevel parallelism of crystal structure prediction and crystal structure analysis to vet candidate electrode materials for advanced battery applications. With potential materials systems numbering in the millions, an experimental approach that can exhaustively - or even significantly - consider all candidate electrode materials is obviously impractical. Many of these systems can be considered and culled using high throughput algorithms that, individually, impose a minimal computational expense, but collectively require large-scale computational resources. These algorithms can be easily parallelized, with nearly 100% scaling efficiency, due to the inherent independence of each calculation. For those materials that survive the initial screening, a more thorough examination at a higher level of theory is necessary. This involves identifying first the crystal structures then the properties of novel crystal chemistries.
Atomistic Simulation Environment (python)