Mapping the coupled degradation pathways of solid-state batteries through granular text mining

The commercial deployment of solid-state batteries (SSBs) remains constrained by intricate, interdependent degradation pathways occurring across solid–solid interfaces. Despite a vast body of empirical studies, a systematic, literature-scale quantitative synthesis of how distinct failure modes are reported and associated across diverse material systems remains limited. Here, we establish an automated text-mining and bibliometric framework encompassing 1314 peer-reviewed publications retrieved through February 2026, with temporal analyses restricted to the complete 2012–2025 interval. The methodology combines a physics-informed hierarchical taxonomy with a validated natural language processing (NLP) pipeline that achieves 75.0% strict single-label accuracy and 78.8% weighted precision against double-blind expert annotations, complemented by transformer-based semantic modeling (BERTopic). Non-linear temporal regression characterizes a sustained exponential increase in degradation-focused publications over 2012–2025 ( R 2 = 0.888 ), consistent with increasing research emphasis on interfacial and chemo-mechanical degradation. Tripartite flow mapping and normalized material–mechanism association analyses reveal distinct literature-level emphasis differentials. Sulfide-electrolyte studies show comparatively stronger emphasis on chemical/electrochemical decomposition, whereas oxide/ceramic studies exhibit substantial morphological and bulk/grain-boundary degradation discussion alongside the broadly prevalent interfacial-impedance literature. These patterns reflect differences in reporting and research emphasis rather than mutually exclusive physical degradation regimes, consistent with the coupled electro-chemo-mechanical nature of SSB failure. Such coupled trade-offs are also consistent with growing literature interest in composite solid electrolytes (CSEs) as candidate architectures for combining interfacial stability with mechanical compliance. Finally, we identify important literature gaps, particularly the limited availability of standardized, machine-extractable in situ mechanical, thermal, and contact-pressure descriptors required for future predictive multiphysics models.