Machine learning was utilised to swiftly identify some of the most promising fluoride-ion battery materials. The research could hasten the development of new batteries, which are expected to challenge, if not completely replace, lithium-based batteries.
Fluoride-ion batteries are great for anything from electric vehicles to consumer electronics in principle. This is due to the fact that fluoride ions are light, tiny, and extremely stable. Fluoride is also less expensive than the lithium and cobalt needed for lithium-ion batteries. Furthermore, calculations show that fluoride-ion batteries have the ability to store more energy than lithium-ion batteries.
Jack Sundberg and colleagues at the University of North Carolina at Chapel Hill’s Scott Warren lab have developed a machine learning method that uses supercomputers to quickly and correctly calculate how easily fluoride ions flow in any known fluoride-containing crystal. The researchers culled a database of 140,000 known compounds down to 10,000 fluoride-containing candidates using a typical hierarchical computation method.
“Because fluoride-ion conductors have not been studied much, we simply didn’t know what additional criteria to apply to find top candidates among the 10,000,” explains Warren. This meant that promising materials could be rejected by a single criterion before being recognised by other criteria, necessitating the development of a new technique. “The solution ended up being incredibly simple: rank structures instead of removing them,” says Sundberg. “In sports, team rankings are updated as the season progresses. Here, structure rankings are updated as we learn more about fluoride diffusion.”
Sundberg chose 300 of the 10,000 choices at random and used highly precise calculations to determine each material’s fluoride-transporting capabilities. These benchmark calculations, which took a week per material, were then used to train the system to create a far faster version of the computation, which required only one hour per material. The researchers were then able to use this method to assess the fluoride conduction properties of the remaining materials fast and precisely. The method found other fluoride conductors, giving the team confidence in their findings.
“What’s really cool is that many of the materials appear to be better conductors than the ones used in lithium-ion batteries,” says Warren. One such material is a fluoride-containing zinc-titanium compound, ZnTiF6. “This material is extremely cheap, has excellent fluoride-conduction properties, and should be especially promising as an electrolyte for a fluoride-ion battery. We’ve just submitted a patent for some of the most exciting compositions,” Warren explains.