Research Digest — 2026-03-31¶
Papers¶
1. Importance of Electronic Entropy for Machine Learning Interatomic Potentials¶
Source: arXiv (cond-mat.mtrl-sci), Mar 27, 2026 · 📅 2026-03-27 · ↗ Open paper
Conventional MLIPs fail to reproduce correct stability of intermediate Na concentrations in battery cathode \(\ce{NaFePO4}\) because structural optimization leads to incorrect \(\ce{Fe^{2+}}\)/\(\ce{Fe^{3+}}\) charge assignments. The root cause is missing electronic entropy associated with charge ordering. Solution: embed charge-state information directly into MLIP representation by distinguishing \(\ce{Fe^{2+}}\) and \(\ce{Fe^{3+}}\) environments during training. Retrained CHGNet, cPaiNN, and MACE with charge-state-aware representations. Validated against DFT convex hulls for \(\ce{NaFePO4}\) across Na concentration range. Analyzed magnetic moments during structural optimization to reveal charge-ordering failures. First explicit demonstration that electronic entropy missing from MLIPs causes qualitative failures in battery cathode thermodynamics. Provides a practical fix (charge-state embeddings) that works across multiple MLIP architectures. Shows that this is a general problem for mixed-valence transition-metal systems.
Relevance to DENG.Group
Directly relevant to Yanhao Deng's MLIP work and the group's battery materials research. Addresses a critical failure mode of MLIPs for mixed-valence materials like battery cathodes — incorrect charge ordering predictions that break thermodynamic stability calculations. Yanhao should test charge-state embeddings for \(\ce{Li3YCl6}\) and related halide electrolytes where transition metals (if present) exhibit mixed valence. The approach is architecture-agnostic and could be combined with existing MACE workflows. For Naibing's polymer electrolyte work, if redox-active additives are considered, this methodology becomes critical.
2. Kinetic Monte Carlo simulations for suppressing dendrite growth in lithium-ion batteries¶
Source: Journal of Power Sources · 📅 2026-03-30 · ↗ Open paper
KMC simulations can identify operating conditions (current density, temperature, electrolyte properties) that suppress dendrite nucleation and growth. The method captures rare-event kinetics that deterministic continuum methods may miss. Kinetic Monte Carlo simulations of Li deposition/dissolution with explicit tracking of dendrite nucleation events. Parameterized with experimental kinetic data and validated against literature dendrite observations. Application of KMC methodology to dendrite suppression in Li-ion batteries. Most prior dendrite modeling uses phase field or molecular dynamics — KMC offers a different timescale regime (seconds to hours) that bridges atomistic and continuum approaches.
Relevance to DENG.Group
Complements Shoutong Jin's phase field dendrite work with a stochastic KMC approach. Provides an alternative computational framework that may capture kinetic effects missing from continuum phase field models. Shoutong Jin should compare KMC predictions against his phase field simulations for the same dendrite systems. If KMC and phase field agree on critical current densities, it strengthens confidence in both methods. If they disagree, the discrepancy reveals physics missing from one or both approaches. The KMC framework could also be adapted to study dendrite suppression strategies (pulsed current, electrolyte additives).
3. Influence of sintering temperature on the performance of garnet-based all-solid-state Li-metal batteries¶
Source: Journal of Power Sources · 📅 2026-03-30 · ↗ Open paper
Sintering temperature controls grain size, porosity, and grain boundary chemistry in garnet electrolytes, which in turn govern ionic conductivity and cycling stability. Optimal sintering balances densification against Li loss and secondary phase formation. Systematic variation of sintering temperature for garnet electrolyte pellets. Characterization by XRD, SEM, EIS. Electrochemical testing in Li|garnet|Li symmetric cells and full solid-state batteries. Provides quantitative relationships between sintering conditions, microstructure, and electrochemical performance for garnet electrolytes. Identifies the temperature window that maximizes ionic conductivity while minimizing interfacial resistance.
Relevance to DENG.Group
Directly relevant to solid electrolyte research — garnet-type oxides (\(\ce{LLZO}\)) are leading candidates for solid-state batteries. Processing-microstructure-property relationships are critical for optimizing ionic conductivity and interfacial stability. The processing-structure-property maps are directly applicable to Jerry's group if they synthesize garnet electrolytes or collaborate with experimental groups. The optimal sintering conditions can be used as a baseline for Li loss mitigation strategies. The grain boundary resistance analysis methodology is transferable to halide electrolyte processing.