Dual AI boost for solid-state battery material development
Cette publication existe aussi en Français
Two major projects are using generative AI to find and produce new materials for solid state batteries.
Researchers at Google have used the DeepMind generative AI to find and simulate thousands of new materials, particularly electrolytes for solid state batteries.
At the same time researchers at Carnegie Mellon University in the US have used generative AI to for a system called Coscientist that develops autonomous experiments to speed up the verification of materials.
At Google the graph networks for materials exploration (GNoME) were trained at scale can reach unprecedented levels of generalization, improving the efficiency of materials discovery by an order of magnitude. Building on 48,000 stable crystals identified in continuing studies, improved efficiency enables the discovery of 2.2 million structures, many of which escaped previous human chemical intuition.
The work is an order-of-magnitude expansion in stable materials known to humanity and used the DeepMind AI to find 321,000 potential materials. These are then simulated at a molecular level to determine the functions to an accuracy of 11meV/atom.
Of the stable structures, 736 materials have already been independently experimentally produced to verify the AI output.
At Carnegie Mellon the Coscientist tool is driven by GPT-4 and autonomously designs, plans and performs complex experiments by incorporating large language models empowered by tools such as internet and documentation search, code execution and experimental automation.
This is used to accelerate research across six diverse tasks, including the successful reaction optimization of palladium-catalysed cross-couplings, while exhibiting advanced capabilities for autonomous experimental design and execution.
www.nature.com/articles/s41586-023-06792-0; www.nature.com/articles/s41586-023-06735-9
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
