Research Digest — 2026-06-01¶
Halide Solid Electrolytes & Interfaces¶
1. Triple-Phase Boundary Instability as a Key Degradation Factor in Sulfide|(Oxy)halide Dual-Electrolyte Solid-State Batteries¶
Source: Joule (10.1016/j.joule.2026.00128) · 📅 2026-05-28 · ↗ Open paper
Systematically investigates the triple-phase boundary between sulfide separator (LPSCl) and six different (oxy)halide solid electrolytes in dual-electrolyte solid-state batteries. Using long-term cycling (100 cycles), impedance spectroscopy, ex situ and operando XPS across three cell configurations, the authors demonstrate that the triple-phase boundary is intrinsically detrimental regardless of (oxy)halide chemistry. Degradation products include metal sulfides, elemental sulfur, and sulfur gas evolution at ~4.3 V vs. Li⁺/Li. Only by introducing an (oxy)halide interlayer to avoid this boundary can high capacity retention be achieved.
Relevance to DENG.Group
Directly relevant to the Deng group's core research on halide solid electrolytes and electrode-electrolyte interfaces. The finding that triple-phase boundary instability is universal across all (oxy)halide chemistries (In-, Sc-, Zr-, Nb-, Ta-based) has major implications for the group's work on dual-electrolyte cell designs. The operando XPS evidence of sulfur gas evolution and the proposed mitigation strategies (ultrathin separator coatings, electrolyte-coated active materials, fluorinated halides) provide concrete research directions for Cheng Peng and Yanhao Deng's interface stability studies.
2. Large-Scale Screening of High-Entropy Materials for Superionic Solid Electrolytes¶
Source: npj Computational Materials (s41524-026-02116-8) · 📅 2026-05-27 · ↗ Open paper
Conducts exhaustive screening of 113,098 high-entropy solid electrolyte candidates generated by iterative elemental substitution into 93 known Li-ion conductor prototypes. Using machine learning interatomic potentials (fine-tuned M3GNet) for molecular dynamics validation, the authors identify eight halide high-entropy solid electrolytes with ionic conductivities improved by up to two orders of magnitude compared to their host materials. The conductivity enhancement is attributed to frustrated potential energy surfaces created by the high-entropy effect.
Relevance to DENG.Group
Highly relevant to the Deng group's halide electrolyte research and Yanhao Deng's MLIP work. The high-entropy design strategy combined with MLIP-accelerated screening directly parallels the group's computational approach. The identification of eight specific halide compositions with dramatically improved conductivity provides concrete targets for experimental synthesis or further computational study. The finding that high-entropy effects frustrate the potential energy surface to enhance conductivity provides a new theoretical framework for the group's understanding of ion transport in halide electrolytes.
Solid Electrolytes & Ion Transport¶
3. Disorder and Entropy Engineering in Solid-State Electrolytes for Fast Ion Conduction¶
Source: Energy Storage Materials (S2405829726003958) · 📅 2026-05-25 · ↗ Open paper
A comprehensive review covering disorder and entropy engineering strategies for enhancing ionic conductivity in solid-state electrolytes. The review systematically examines configurational disorder, site disorder, and compositional disorder across sulfide, oxide, and halide electrolyte families, analyzing how each type of disorder affects the energy landscape for ion migration and the resulting conductivity improvements.
Relevance to DENG.Group
Relevant as a reference resource for the group's solid electrolyte research. The systematic treatment of disorder engineering strategies connects directly to the group's work on doped halide electrolytes and the high-entropy screening paper also covered in this digest. The review provides a theoretical framework for understanding how the compositional complexity the group introduces into halide electrolytes affects the ion transport energetics, which can guide future computational and experimental design of new electrolyte compositions.
4. Lithium Metal Batteries with a Bone-Inspired Solid-Sol Electrolyte Based on Natural CaF2¶
Source: ACS Nano (10.1021/acsnano.6c03537) · 📅 2026-05-27 · ↗ Open paper
Develops a bone-inspired solid-sol electrolyte using naturally abundant CaF2 as the inorganic matrix with only 12.8 wt% organic solvent. The multilevel nonbonding interactions between CaF2 and the organic solvent create a continuous ion transport network. The CaF2 matrix provides mechanical support and stability, while in situ reaction with Li forms an SEI enriched with LiF and Li-Ca alloy, enhancing interfacial kinetics and suppressing dendrite growth. The electrolyte exhibits a 5.26 V electrochemical window, Li⁺ transference number of 0.77, and stable cycling at temperatures up to 100°C.
Relevance to DENG.Group
Relevant to the group's interest in solid-state electrolytes and interface engineering. The CaF2-based solid-sol concept represents an unconventional approach to quasi-solid electrolytes that could inspire new composite designs. The in situ formation of LiF-rich SEI through CaF2-Li reaction is notable for the group's dendrite suppression research — the LiF and Li-Ca alloy SEI composition provides a model system for understanding how mechanically robust interphases form and function. The bone-inspired architecture of combining rigid inorganic scaffolds with organic ion-conducting pathways could inform the group's thinking about composite electrolyte design.
ML Interatomic Potentials & Workflows¶
5. Machine-Learning-Assisted Discovery of a Stable Li3As2 Intermediate Phase in the Li-As Binary System and Its Electrochemical Implications¶
Source: Physical Chemistry Chemical Physics (10.1039/D6CP01519K) · 📅 2026-05-28 · ↗ Open paper
Proposes a hierarchical computational workflow combining global structure search with a fine-tuned machine learning interatomic potential (25.3 meV/atom energy RMSE on ~3500 DFT-labeled configurations) to explore the Li-As binary system. The approach reveals a thermodynamically stable intermediate phase C2/c-Li₃As₂ on the convex hull between LiAs and Li₃As. This phase is dynamically stable, metallic, and shows a low Li⁺ migration barrier, with an equilibrium potential of ~0.95 V vs. Li/Li⁺ and a theoretical capacity of 536.6 mAh/g with ~68.6% volume expansion — making it promising for shallow-lithiation fast-charging anodes.
Relevance to DENG.Group
Directly relevant to Yanhao Deng's MLIP development work. The hierarchical workflow of fine-tuning pretrained potentials on relatively small DFT datasets (~3500 configurations) and using them for global structure search mirrors the approach the group could use for exploring new solid electrolyte or electrode compositions. The 25.3 meV/atom accuracy achieved with fine-tuning demonstrates the practical feasibility of transfer learning for materials discovery. The Li₃As₂ discovery also has implications for the group's understanding of alloy anode–electrolyte interfaces, where such intermediate phases may form during cycling.
6. Benchmarking Universal Machine Learning Interatomic Potentials for Solid-State Electrolyte Applications¶
Source: ACS Materials Letters (10.1021/acsmaterialslett.6c00134) · 📅 2026-05-27 · ↗ Open paper
Provides a systematic benchmark of universal MLIPs (including MACE, CHGNet, and M3GNet) specifically for solid-state electrolyte applications, evaluating their accuracy on key properties including Li⁺ migration barriers, phonon spectra, and thermodynamic stability across sulfide, oxide, and halide electrolyte families. The study identifies systematic failure modes and provides practical recommendations for when universal potentials can be trusted versus when system-specific fine-tuning is necessary.
Relevance to DENG.Group
Directly relevant to Yanhao Deng's MLIP work. This benchmark provides exactly the type of systematic evaluation the group needs to decide which universal potential to use as a starting point for their halide electrolyte simulations. The identification of specific failure modes for SSE applications will help the group avoid common pitfalls when using off-the-shelf potentials. The practical recommendations for fine-tuning thresholds directly inform the group's workflow decisions and could save significant computational resources by identifying which properties can be reliably predicted with universal models versus requiring custom training.