Can AI help solve the lithium battery explosion problem?

Researchers from Stanford University in the United States have used artificial intelligence and machine learning to find about 21 solid electrolyte materials, hopefully to completely solve the lithium battery explosion problem.

Recently, researchers from Stanford University published their latest research results in the Journal Energy and Environmental Sciences. Unlike previous methods of randomly testing individual compounds to find new solid electrolyte materials, researchers used artificial intelligence and machine learning techniques to construct prediction models based on experimental data and found about 21 solid electrolyte materials. This is expected to completely solve the explosion problem of lithium batteries.

Usually, because the electrolyte of traditional lithium batteries is organic liquid, it is very likely that side reactions will occur at a relatively low temperature, resulting in gas, which will cause explosion after expansion of the batteries. The team's researchers trained an algorithm to simulate compounds and, after learning about existing substances in depth, screened suitable materials from large databases.

The team spent about two years designing the algorithm model. Austin Sendek, PhD candidate in Applied Physics, who led the research, said: "We screened more than 12,000 compounds containing lithium and eventually found 21 ideal materials for solid electrolytes. Screening takes only a few minutes. The vast majority of my time is actually spent collecting and managing all the data and developing more reliable measurement mechanisms for prediction models.

The team will continue to test these 21 solid materials to see if they can really solve the lithium battery explosion problem.

The page contains the contents of the machine translation.