Oct 2026· Small· pp.
e75978
· 0 citations· 42 references
Medicine
Abstract
Accurate prediction of ion transport properties remains a key bottleneck in computational discovery of solid-state electrolytes, particularly because experimentally measured ionic conductivities are sparse and strongly dependent on crystal structure. Here we introduce USMNet, a unified structure-aware multimodal neural network that combines crystal-structure representations, compositional features, and physically motivated descriptors to predict lithium-ion conductivity. We benchmark the model against composition-only and structure-based baselines, showing that explicit structural information improves prediction accuracy and enables data-efficient and reliable exploration of candidate materials. We then apply USMNet to 168 675 candidate materials spanning 19 lithium solid-state electrolyte frameworks, identifying 21 high-potential inorganic electrolytes. Selected candidates are synthesized and experimentally validated, with Li16ZnSiP2S16 enabling all-solid-state batteries that retain 80% capacity after 600 cycles under 0.5 C at room temperature. These results establish a generalizable structure-aware learning strategy for electrochemical property prediction and demonstrate its utility in accelerating closed-loop materials discovery under limited data regimes.
The discovery of high-performance solid-state electrolytes remains a critical challenge for all-solid-state Li-ion batteries. To accelerate the identification of novel Li-ion conductors, we developed a hierarchical screening framework combining local structure order parameters (LSOPs) and persistent homology (PH). By...
Song-Jia Kong, Zi-Heng Yu, Kota Suzuki et al.· Journal of the American Chem...· 0 citations
A hierarchical synergistic deep-learning framework that sequentially coordinates efficiency, accuracy, and reliability through four complementary modules that outperforms mainstream counterparts in its task, while retrospective validation establishes dual closed-loop verification of module-level accuracy and end-to-end...
Hong-Wei Du, Ding-Yang Lv, Bao-Le Wei et al.· 0 citations
Efficient exploration of chemical space with machine-learning interatomic potentials (MLIPs) requires a comprehensive, application-focused training database. Solid-state electrolytes for Li-ion batteries have attracted growing attention owing to their ability to improve safety and energy density. Among these, halide-ba...
Chiku Parida, Arghya Bhowmik, Juan Maria Garcia Lastra· Machine Learning: Science an...· 0 citations
Elemental substitution within existing structural frameworks is a widely applied strategy for developing advanced materials. Yet, optimizing target properties while maintaining phase purity usually demands extensive trial-and-error, which becomes substantially inefficient when navigating a complex design space. Here, w...
Bernardus Rendy, Yu-Xing Fei, T. He et al.· 0 citations
Disordered materials such as glasses are promising candidates as solid-state electrolytes for batteries, but they tend to exhibit low ionic conductivity and current methods for investigating and predicting the ionic conductivity of electrolyte materials are inefficient. Here, we propose a framework that summarizes the...
M. Pegoraro, Rasmus Christensen, S. S. Sørensen et al.· Nature Communications· 0 citations
Solid-state electrolyte (SSE) with high ionic conductivity and low cost is essential for accelerating the commercialization of all-solid-state sodium-ion batteries. Here, we provide profound atomistic insights into electronic and ionic transport properties of the SSE candidate Na2ZnS2 using the highly accurate first-pr...
