Developing safer solid-state batteries requires replacing flammable organic liquid electrolytes with solid-state electrolytes (SSEs) that can conduct ions well, remain stable, and have suitable mechanical properties. As interest in SSEs has grown, so has the volume of research in the field. However, the reported data remain scattered across the literature and vary widely in format and completeness, making systematic comparisons and the direct use of literature data for machine learning difficult.
To address this problem, the Dynamic Database of Solid-State Electrolytes (DDSE) was created to collect key data for hundreds of materials in one web-based platform.
Building on DDSE, researchers at the Advanced Institute for Materials Research (WPI-AIMR), Tohoku University, have now developed a broader platform called DigBat, which brings together more types of electrolyte data. The range of materials has been expanded to include inorganic, solid polymer, and gel polymer electrolytes. In addition, DigBat connects experimental data with atomistic simulations, machine-learning tools, and a large language model (LLM)-based assistant, bringing different types of information and analysis tools together in one platform.
“A database can be much more than a place to store information,” said Hao Li, Distinguished Professor and Principal Investigator at WPI-AIMR “We see DigBat as a map for navigating the growing world of solid-state electrolyte research. By bringing together data from different sources, including experimental results, simulations, and models, it can help researchers see the landscape more clearly and point toward promising paths for future exploration.”
As of June 2026, DigBat contains 3,816 experimental solid-state electrolytes with 27,645 conductivity data entries, as well as 852 computational materials. To improve data quality and consistency, the experimental conductivity data were manually extracted and cross-checked, duplicate entries were removed during curation, and reported values were standardized to make comparisons across publications easier. These curated experimental data are then organized into three modules in the web interface: “Inorganic solid electrolytes” (ISEs), “Solid polymer electrolytes” (SPEs), and “Gel polymer electrolytes” (GPEs).
Details of the findings were published in Nano Materials Science on July 27, 2026.
Across all modules, a common material identifier links experimental and computational records for the same material. This means that researchers can examine conductivity measurements together with structural information rather than treating each type of information separately.
For example, using the inorganic solid electrolyte data collected in DigBat, researchers can compare how well different materials conduct ions at different temperatures and explore trends in the energy barriers for ion movement across material types and publication years.
Beyond comparing existing data, DigBat can also use data to build machine-learning models. Using hydride solid-state electrolytes as a case study, the researchers developed interpretable symbolic-regression models to predict activation energy and compared the predicted values with the observed data.
“What matters is not only whether a model can make a good prediction, but whether we can understand what stands behind that prediction,” said Hao Li. “By using interpretable models, we can examine how specific material properties are related to activation energy and gain more insight from the data.”
©Qian Wang et al.
By turning scattered results into a connected and searchable resource, DigBat provides researchers with a clearer view of the solid-state electrolyte landscape. As the platform continues to grow, the team hopes it will support new connections between experiments, simulations, and data-driven methods, helping guide the search for promising materials for next-generation batteries.
- Publication Details:
Title: Digital battery platform: extending the dynamic database of solid-state electrolytes to a diversified electrolyte database with ion migration models
Authors: Qian Wang, Hanghui Liu, Seong Hoon Jang, Di Zhang, Ryuhei Sato, Kazuaki Kisu, Yusuke Hashimoto, Eric Jianfeng Cheng, Shin-ichi Orimo, Hao Li*
Journal: Nano Materials Science
