Combining Structured Data with Domain Knowledge in Battery Materials Research: The Case of Conductive Networks

Batteries contain combinations of materials that undergo electrochemical reactions to convert chemical into electrical energy. Battery research relies on experience and know-how. Important materials and processing data can get overlooked, remain undocumented, or even lost. To bridge the gap between fundamental materials research and battery process engineering, it is essential to generate, analyze, and, most importantly, link intermediate knowledge for future use. Here, it is shown how to combine domain knowledge and a data-driven approach to understanding material–property relationships in the case of conductivity networks of carbon black. The Battery Production and Characterisation Ontology (BPCO) is employed to identify hypotheses that connect battery processing to material domain knowledge. The material's interactions between carbon black, polyvinylidene flouride, and solvents in the BPCO are characterized. These materials combine to form the classical microstructure in battery electrodes for the electrical conductivity. It is demonstrated how new links to the BPCO, verified via materials-processing relationships, and the interim results are identified as intermediate data.