Research Digest — 2026-07-09¶
Dendrite Growth & Mechanics in Solid Electrolytes¶
1. Multiphysics Modeling and Analysis for Dendrite Problems in Solid-State Lithium/Sodium Metal Batteries¶
Source: Nano-Micro Letters (s40820-026-02200-0) · 📅 2026-06-25 · ↗ Open paper
A comprehensive review of multiphysics modeling approaches for dendrite problems across liquid and solid-state battery systems. Covers phase-field, continuum mechanics, and electrochemical coupling frameworks for both Li and Na metal anodes with solid electrolytes, systematically summarulating experimental observations of dendrite morphologies.
Relevance to DENG.Group
Excellent reference for Shoutong Jin's dendrite modeling. The Na metal coverage may also be relevant for future expansion of group research directions.
Grain Boundaries & Interfaces in Solid Electrolytes¶
2. Charged grain boundaries limit short-circuit endurance in garnet solid-state battery electrolytes¶
Source: Nature Materials (s41565-026-02206-0) · 📅 2026-06-30 · ↗ Open paper
This study shows that grain boundaries in LLZO garnet electrolytes feature elevated electronic conduction and act as preferential pathways for lithium deposition, limiting short-circuit endurance. The charged nature of GBs creates localized electronic leakage that promotes dendrite nucleation along GBs before mechanical fracture occurs.
Relevance to DENG.Group
Core relevance to Cheng Peng's grain boundary research. This directly connects GB electronic structure to dendrite vulnerability — a computational target for ML potential-based GB modeling.
3. Machine-learning interatomic potentials for interfaces in all-solid-state batteries¶
Source: OSTI / DOE (3024472) · 📅 2026-06-18 · ↗ Open paper
Reviews the emerging use of MLIPs for large-scale, high-accuracy simulations of interfaces in all-solid-state batteries. Covers training data strategies, active learning schemes, and applications to electrode/electrolyte interface stability and ion transport across interfaces.
Relevance to DENG.Group
Core relevance to Yanhao Deng and Umang Agarwal's work. The active learning schemes described could improve the group's ML potential training efficiency.
Halide Solid Electrolytes¶
4. Machine learning pipelines for the design of solid-state electrolytes¶
Source: Materials Horizons (d5mh01525a) · 📅 2026-06-18 · ↗ Open paper
Comprehensive survey of ML pipelines for solid-state electrolyte design, from data resources and feature engineering through classical models, deep learning architectures, and cutting-edge generative models. Covers the full workflow from data curation to property prediction and materials discovery.
Relevance to DENG.Group
Important methodological reference for the group's ML potential and materials discovery efforts. Could inform new research directions combining MLIPs with generative design.