DigBat maps 3,816 solid-state electrolytes

A researcher comparing solid-state electrolytes no longer has to treat every paper as an isolated island. DigBat, a web-based platform developed by the Advanced Institute for Materials Research (WPI-AIMR) at Tohoku University, brings 3,816 experimental electrolytes and 27,645 conductivity data entries into one searchable research environment, alongside 852 computational materials as of June 2026.

The problem it tackles sits beneath the promise of safer solid-state batteries. These designs replace flammable organic liquid electrolytes with solid materials that must conduct ions well, remain stable and have suitable mechanical properties. Yet published results are scattered and formatted inconsistently, making comparisons—and the direct use of literature data for machine learning—awkward. DigBat expands an earlier Dynamic Database of Solid-State Electrolytes to cover inorganic, solid polymer and gel polymer electrolytes.

The platform’s useful detail is in the connections. A common material identifier ties experimental records to atomistic simulations, so conductivity measurements can be examined alongside structural information. Researchers can compare conductivity at different temperatures and study energy barriers for ion movement across material types and publication years. The team says it manually extracted and cross-checked experimental conductivity data, removed duplicates and standardized reported values.

DigBat also turns the collection into a modeling workspace. In a case study on hydride solid-state electrolytes, the researchers used symbolic regression—machine learning that produces interpretable equations—to predict activation energy. Hao Li, a distinguished professor and principal investigator at WPI-AIMR, said the point is not only accurate prediction but understanding which material properties sit behind it.

For researchers, the platform offers a way to compare reported materials, connect measurements to simulations and inspect model-based leads. It does not, on the evidence reported by the team, deliver a finished battery or guarantee a promising material will work outside its published conditions. DigBat is a growing research platform, not a commercial battery deployment.