3992 papers
Condensed matter physics and materials science form a dynamic partnership, exploring how the collective behavior of atoms gives rise to the unique properties of solids and liquids. This field bridges the gap between fundamental quantum mechanics and the practical engineering of everything from flexible electronics to superconductors, turning abstract theories into tangible innovations that shape our daily lives.
At Gist.Science, we process every new preprint in this category directly from arXiv to make these complex discoveries accessible to everyone. Our team generates both plain-language overviews and detailed technical summaries for each paper, ensuring that researchers, students, and curious minds alike can grasp the latest breakthroughs without getting lost in dense jargon.
Below are the latest papers in condensed matter and materials science, organized by their most recent publication dates.
This paper introduces LLM4Mat-Bench, the largest benchmark to date comprising 1.9 million crystal structures and 45 properties across multiple input modalities, to systematically evaluate and highlight the limitations of both general-purpose and task-specific large language models in predicting materials properties.
This study systematically characterizes nitrogen-vacancy centers in amino-terminated fluorescent nanodiamonds, revealing that while zero-field splitting parameters evolve with size, wet-chemical Hofmann amino termination uniquely enables a high, laser-power-independent negative charge state in 140-nm particles by degrading surface paramagnetic defects.
This study demonstrates that uniaxial in-plane compressive strain effectively suppresses both the spontaneous anomalous Hall effect and the tiny net magnetization in the altermagnet candidate h-FeS by tuning the c-axis ferromagnetic moment, highlighting a strong correlation between these properties and offering a viable control mechanism for spintronic applications.
This study demonstrates that incorporating chiral 3-fluoropyrrolidinium cations into a two-dimensional copper chloride perovskite induces chiral magnetic order and a second-order magnetoelectric effect, whereas the racemic variant lacking structural chirality exhibits no such magnetic chirality despite sharing the same antiferromagnetic transition temperature.
This paper develops a unified thermodynamic framework for fluid-fluid and solid-fluid interfaces by extending the Larché and Cahn variational formulation to incorporate interfacial thermodynamics directly, thereby deriving both bulk and interfacial equilibrium conditions as stationarity conditions of a single functional.
The paper introduces SPEAR, a framework that enhances the physical interpretability and stability of attention-based machine learning models for materials discovery by applying learnable temperature and smoothness constraints to attention mechanisms, thereby generating coherent, causally aligned explanations without compromising predictive accuracy.
This paper reports a record-high room-temperature transverse thermoelectric figure of merit () of 0.36 in artificially tilted multilayers composed of NiFe and BiSbTe, achieved without an external magnetic field by leveraging the alloy's high power factor and the sharp transport property contrast between the constituent layers.
This paper introduces a novel simulation-based inference framework utilizing neural ratio estimation and conditional flow matching, enhanced by a physics-aware vision transformer, to efficiently infer full posterior distributions of Hamiltonian parameters from RIXS spectroscopy data, thereby enabling advanced uncertainty quantification and experimental design for transition metal compounds.
This paper reports the discovery of a robust in-plane anomalous Nernst effect in symmetry-tailored ultrathin ferromagnetic oxide films, demonstrating that the effect can arise from spontaneous in-plane magnetization and out-of-plane orbital magnetization, thereby overcoming the traditional orthogonality constraint and enabling more flexible magnetothermoelectric device designs.
This study finds that while silicon sampling using frontier models remains an early-stage method that underestimates opinion variance and distorts contextual space, its "fast" mode offers a conditional superiority over "slow" modes by delivering higher algorithmic fidelity with significantly greater efficiency in compute resources and runtime.