Overview

Materials science studies how a material’s composition, atomic and microstructure, and processing history shape properties such as strength, conductivity, stability, and catalytic activity. Researchers use machine-learning interatomic potentials and equivariant graph networks to approximate first-principles calculations, physics-informed surrogates to model mechanics and microstructure evolution, and generative models to propose crystals, polymers, and metamaterials. Bayesian active learning and autonomous laboratory systems help select experiments and refine candidates, while X-ray diffraction, microscopy, and synthesis benchmarks test predictions against measurements. A central challenge is ensuring that strong results on curated or simulated data generalize to new materials and real experiments while remaining physically valid, synthesizable, and reproducible.

Latest papers

CardsList

333 papers36 papers in the last four weeks, level with the four weeks before. 0.3% of all new papers.

  1. Interpretable Nanoporous Materials Design with Symmetry-Aware Networks

    Sep 19, 2025Zhenhao Zhou, Salman Bin Kashif, Jin-Hu Dou +4Interpretable MLMaterials Property Prediction

  2. GenShin: Guiding Rational Liposome Design by Ranking Liposomal Protein Corona through a Docking-Pose-Free GNN

    Mar 16, 2025Pingfei Zhu, Hongyi Liu, Xueyan Liu +2Graph Neural NetworksMaterials Science

  3. Learning Chern Numbers of Topological Insulators with Gauge Equivariant Neural Networks

    Feb 21, 2025Longde Huang, Oleksandr Balabanov, Hampus Linander +3Equivariant Neural NetworksMaterials Science