Lawrence Berkeley National Laboratory

Sol-gel synthesis is a versatile method for producing inorganic materials that offers advantages such as atomic-level mixing and lower processing temperatures over traditional solid-state methods. Despite these benefits, achieving phase purity remains a challenge due to the process’s sensitivity to synthesis parameters such as precursor choice, solvent selection, and heating conditions. Phase-impure outcomes are essential for understanding sol-gel reaction pathways, but are largely absent from existing curated datasets. This work leverages Gemini 3.0 Flash to construct a text-mined inorganic sol-gel database with phase purity information. The dataset contains ~148,000 synthesis records, including ~28,000 phase-impure entries. All extracted entities achieved F1 scores greater than 0.89. At the record level, 68% of extracted records match human annotation on all evaluated fields. Furthermore, to validate the scientific utility of our dataset, we demonstrate that the aggregated data reproduces previously reported phase-purity trends for BiFeO3 thin films.