arXiv subjects

Yongheng Li

Publications and source records attributed to Yongheng Li.

5 recordsLinked to original sources

arXiv2026

A Hierarchical Synergistic Deep Learning Framework Integrating Composition, Structure, and Ionic Transport for Solid-State Electrolyte Discovery

Inorganic solid-state electrolytes must combine high room-temperature ionic conductivity, a wide electrochemical window, excellent electronic insulation, and favorable mechanical compliance. Single models struggle to support reliable multi-objective screening across vast chemical spaces because of training-data distribution mismatch, cross-property dataset heterogeneity, and scarce kinetic transport data. To overcome these limitations, we develop a hierarchical synergistic deep-learning framework that sequentially coordinates efficiency, accuracy, and reliability through four complementary modules. The in-house-developed L-G-DCNN and a multi-fidelity implementation built on DenseGNN serve as compositional and structural experts for thermodynamic coarse screening and multi-property evaluation, respectively; MatterSim and system-specific DeePMD models provide transport pre-assessment and kinetic validation. Systematic benchmarks show that each module outperforms mainstream counterparts in its task, while retrospective validation establishes dual closed-loop verification of module-level accuracy and end-to-end workflow reliability. Applied to 30,364,908 Alex/ICSD-derived candidates, the framework identifies 97 high-performance candidates with room-temperature ionic conductivities of 0.109--59.0 mS/cm, including 94 halides, one borohydride, and two oxides. Consistency with independent experimental data confirms that 76 of the 94 halides fall within reported high-conductivity structural regions. Analysis reveals that Li$^{+}$ jump-network connectivity, rather than the number of geometric Li sites, is the core determinant of room-temperature ionic conductivity. Li-defect engineering effectively enhances oxide transport, whereas the inherent rigidity of the O$^{2-}$ framework suggests a potential upper limit on oxide electrolyte performance.

cond-mat.mtrl-sci↗

arXiv2026

Adapting a Pre-trained Single-Cell Foundation Model to Spatial Gene Expression Generation from Histology Images

Spatial transcriptomics (ST) enables spot-level in situ expression profiling, but its high cost and limited throughput motivate predicting expression directly from HE-stained histology. Recent advances explore using score- or flow-based generative models to estimate the conditional distribution of gene expression from histology, offering a flexible alternative to deterministic regression approaches. However, most existing generative approaches omit explicit modeling of gene-gene dependencies, undermining biological coherence. Single-cell foundation models (sc-FMs), pre-trained across diverse cell populations, capture these critical gene relationships that histology alone cannot reveal. Yet, applying expression-only sc-FMs to histology-conditioned expression modeling is nontrivial due to the absence of a visual pathway, a mismatch between their pre-training and conditional ST objectives, and the scarcity of mixed-cell ST supervision. To address these challenges, we propose HINGE (HIstology-coNditioned GEneration), which retrofits a pre-trained sc-FM into a conditional expression generator while mostly preserving its learned gene relationships. We achieve this by introducing SoftAdaLN, a lightweight, identity-initialized modulation that injects layer-wise visual context into the backbone, coupled with an expression-space masked diffusion objective and a warm-start curriculum to ensure objective alignment and training stability. Evaluated on three ST datasets, ours outperforms state-of-the-art baselines on mean Pearson correlation and yields more accurate spatial marker expression patterns and higher pairwise co-expression consistency, establishing a practical route to adapt pre-trained sc-FMs for histology-conditioned spatial expression generation.

cs.CV↗

arXiv2025

Pressure-Induced Reversal of Thermal Anisotropy in Bi2O2Se

Bi2O2Se is an emerging semiconductor with intrinsically low thermal conductivity, making it a promising material for thermoelectric applications. Hydrostatic pressure can effectively tunes the thermal conductivity, with various pressure-dependent trends reported. However, its impact on thermal anisotropy, particularly in the highly anisotropic Bi2O2Se, remains poorly understood. Here, we report a pressure-driven reversal of thermal anisotropy: k_z < k_x at 0 GPa transforms into k_z > k_x at 60 GPa without phase transition. This stems from distinct phonon dispersions along the x- and z-directions under pressure, leading to a reshaped group velocity landscape. Below 10 meV, vz > vx at both pressures, with a much greater advantage at 60 GPa. Above 10 meV, vx > vz at 0 GPa; however, the difference nearly vanishes at 60 GPa. These changes result from anisotropic lattice compression, with the z-axis shrinking more significantly than the x-axis and suppressing the lone pair activity of Bi atoms. This study calls for revisiting the pressure dependence of thermal conductivity anisotropy and provides insights for pressure-driven thermal switching applications.

cond-mat.mtrl-sci↗

arXiv2025

Incipient ionic conductors: Ion-constrained lattices achieving superionic-like thermal conductivity by extreme anharmonicity

Phonon liquid-like thermal conduction in the solid state enables superionic conductors to serve as efficient thermoelectric device candidates. While liquid-like motion of ions effectively suppresses thermal conductivity (κ), their high mobility concurrently triggers material degradation due to undesirable ion migration and consequent metal deposition, making it still a challenge to balancing low κand high stability. Here, we report a superionic-like thermal transport alongside restricted long-range ion migration in CsCu_2I_3 with incipient ionic conduction, using synchrotron X-ray diffraction, inelastic X-ray scattering, and machine-learning potential-based simulations. We reveal that the Cu ions exhibit confined migration between CuI_4 tetrahedra at high temperatures, displaying extreme anharmonicity of dominated phonons beyond conventional rattling and comparable to that in superionic conductorsl. Consequently, a glass-like κ(~0.3 W m^{-1} K^{-1} at 300 K) following the relationship of κ~ T^{0.17}, was achieved along the x-direction, where Cu ion migration is three oders of magnitude lower than in superionic conductors. These results highlight the advantage of incipient ionic conductors in simultaneously maintaining both low κand high stability, elucidate the thermal transport mechanism via ion migration constraints, and pave an effective pathway toward ultralow thermal conductivity in ionic conductors.

cond-mat.mtrl-sci↗

arXiv2025

Low and Anisotropic Thermal Conductivity in Mixed-Valent Sn$_2$S$_3$

Compounds of Sn, such as SnSe and SnS, exhibit novel phonon characteristics and low thermal conductivity, making them emerging star materials in the thermoelectric family. In this work, through the Boltzmann transport equation scheme and the Wigner thermal transport model, quasi-1D mixed-valent Sn$_2$S$_3$ were found to exhibit a low thermal conductivity along c-axis with a weak temperature dependence. The low thermal conductivity is attributed to the anharmonic rattling vibrations of weakly bonded Sn(II) atoms, which are influenced by the coulomb interaction of lone pairs at adjacent Sn(II) atoms. The rattling of Sn(II) induces low-frequency flat optical phonons and avoids crossing behavior. The atomic displacements and mean square displacement (MSD) analysis reveal that Sn(II) atoms exhibit significantly greater and anisotropic displacements compared to Sn(IV) and S, confirming that Sn(II) behaves as a rattler. The results obtained from this work suggest an opportunity to discover low thermal conductivity in mixed-valent compounds.

cond-mat.mtrl-sci↗