Research Digest — 2026-10-23¶
Halide Solid Electrolytes¶
1. Cyclic Nonequilibrium Processing of Halide Solid Electrolytes for Low-Pressure Operation of All-Solid-State Batteries¶
Source: Advanced Materials (10.1002/adma.75398) · 📅 2026-10-15 · ↗ Open paper
Repeated quenching of Li3InCl6 first refines particle size and then induces local amorphization, suppressing grain-boundary resistance and introducing Li vacancies. Room-temperature ionic conductivity rises from 1.3 to 6.8 mS/cm while the Young's modulus drops from 42.3 to 17.6 GPa, enabling stable all-solid-state cell cycling at only 2 MPa stack pressure with 7.7 mAh/cm2 areal capacity.
Relevance to DENG.Group
Directly relevant to Yan Li and Mengke Li working on halide electrolyte processing and transport: it ties microstructure (amorphization, grain boundaries, vacancies) to conductivity and mechanical deformability. The local-amorphization route also resonates with Tzu-Chen Liu's amorphous electrolyte theme.
Grain Boundaries & Interfaces¶
2. Garnet-type LLZO solid electrolytes: grain-boundary engineering, electro-chemo-mechanical failure, and manufacturability¶
Source: Energy Storage Materials (10.1016/j.ensm.2026.105595) · 📅 2026-10-12 · ↗ Open paper
A review that organizes LLZO reliability around a process-state-field-failure-performance framework, showing how connected grain boundaries, pores, cracks, and surface films localize ionic current, electronic leakage, and mechanical stress. It disentangles bulk vs. grain-boundary conductivity and activation energies to explain why cubic-phase stabilization or densification alone cannot predict pellet-scale transport or short-circuit onset.
Relevance to DENG.Group
Core reading for Cheng Peng (grain boundaries in solid electrolytes): the coupled GB chemistry / mechanical softening / filament-penetration picture maps directly onto his multiscale GB modelling, and the quantitative bulk-vs-GB decomposition offers benchmarks for simulations.
ML Interatomic Potentials¶
3. Materials foundation models fail together: Correlated errors and collective blind spots¶
Source: Computational Materials Science (10.1016/j.commatsci.2026.115153) · 📅 2026-10-15 · ↗ Open paper
Quantifies error correlation across 15 public universal MLIPs on 248,623 structures: mean pairwise error correlation of 0.39 means only about 4.3 effectively independent models. Consensus screening fares poorly: the median consensus did not beat the best single model, agreement-gated screening admitted 1620 false discoveries versus 7 under a permutation null, and 27.6% of predicted-stable structures were false discoveries.
Relevance to DENG.Group
A caution for Yanhao Deng and everyone in the group using MACE/MatterSim-class universal potentials for screening: model agreement is not evidence of correctness because their errors are correlated. Reinforces the group's practice of DFT-checking uMLIP predictions before trusting them for electrolyte stability rankings.
Solid Electrolyte Interphases (SEI)¶
4. Quantifying electronic insulation of solid-electrolyte interphases through electrolyte decomposition kinetics¶
Source: npj Computational Materials (s41524-026-02362-w) · 📅 2026-10-14 · ↗ Open paper
Develops a first-principles framework combining Marcus electron-transfer theory with constrained DFT to define a 'resistance to electrolyte decomposition' (R_ED) metric that unifies electronic transmission across SEI interfaces and reorganization energy of decomposition reactions. The component-resolved, electrolyte-dependent measure ranks how effectively different SEI chemistries suppress continued decomposition.
Relevance to DENG.Group
Methodologically valuable for Umang Agarwal's electrolyte/electrode interface stability work and Yan Li's halide degradation studies: R_ED is a computable descriptor the group could apply to rank interphase compositions and connect to their own degradation thermodynamics and kinetics calculations.
5. Determining depth-resolved spatial distribution of solid-electrolyte interphase using non-destructive angle-resolved X-ray photoelectron spectroscopy for lithium-metal batteries¶
Source: Nature Communications (s41467-026-78168-7) · 📅 2026-10-22 · ↗ Open paper
Introduces angle-resolved XPS as a non-destructive depth-profiling method for the SEI in Li-metal batteries, preserving interphase integrity where sputter depth profiling damages it. It resolves compositional changes before and after Li plating and shows SEI species thickness and distribution vary strongly across electrolyte classes, controlling protective quality.
Relevance to DENG.Group
A useful experimental benchmark for Naibing Wu and Umang Agarwal: measured depth-resolved SEI composition profiles are exactly the kind of structural data that can anchor and validate the group's interfacial chemistry and ion-transport models.
kMC & Thin-Film Growth¶
6. Accelerating and scaling on-lattice Solid-On-Solid kinetic Monte Carlo simulations using an autoregressive Conditional Variational Autoencoder¶
Source: Computational Materials Science (10.1016/j.commatsci.2026.115109) · 📅 2026-10-12 · ↗ Open paper
Trains an autoregressive fully convolutional CVAE surrogate that predicts Solid-on-Solid kMC surface evolution over large time leaps, conditioned on temperature, condensing many sequential event steps into one forward pass. The surrogate preserves statistical growth and roughness dynamics while giving massive speed-ups and scaling to much larger lattices, easing coupling with continuum solvers.
Relevance to DENG.Group
Directly relevant to Jingyuan Lin's thin-film growth kMC and Tzu-Chen Liu's kMC modelling: the ML-surrogate-for-kMC idea could extend their simulations to experimentally relevant times and lengths, though the paper's statistical fidelity checks are worth studying before adopting the approach.
MOF Adsorption & Gas Kinetics¶
7. CO2 Capture in MOFs: Connecting Experiments, Multiscale Simulations, and Data-Driven Approaches¶
Source: Small (10.1002/smll.76202) · 📅 2026-10-14 · ↗ Open paper
A review integrating experimental CO2-capture studies with multiscale simulations - DFT for open-metal-site and pore-functionalization adsorption physics, GCMC and MD for capacity, selectivity, and diffusion, and CFD for process-level transport - plus emerging data-driven structure-property methods and active-learning workflows.
Relevance to DENG.Group
Right in Jason Li's territory: it is a up-to-date map of the multiscale MOF simulation stack he works in, and its data-driven design section may suggest high-throughput screening ideas for his adsorption and gas-diffusion kinetics projects.