arXivDaily arXiv每日学术速递 周一至周五更新
重置
2412.13012 2026-06-13 cs.LG cond-mat.mtrl-sci cond-mat.str-el

Deep Learning Based Superconductivity: Prediction and Experimental Tests

基于深度学习的超导性:预测与实验测试

Daniel Kaplan, Adam Zhang, Joanna Blawat, Rongying Jin, Robert J. Cava, Viktor Oudovenko, Gabriel Kotliar, Anirvan M. Sengupta, Weiwei Xie

发表机构 * Department of Physics and Astronomy(物理与天文学系) Rutgers University(罗格斯大学) Department of Chemistry(化学系) Michigan State University(密歇根州立大学) University of South Carolina(南卡罗来纳大学) Princeton University(普林斯顿大学) Center for Computational Quantum Physics(计算量子物理中心) Flatiron Institute(Flatiron研究所) Center for Computational Mathematics(计算数学中心)

AI总结 本文提出基于深度学习的超导材料预测方法,并通过实验验证,发现新型三元化合物Mo₂₀Re₆Si₄在5.4K以下超导,同时讨论AI预测的局限性与未来研究方向。

Comments 14 pages + 2 appendices + references. EPJ submission

详情
Journal ref
Eur. Phys. J. Plus (2025) 140:58
AI中文摘要

新型超导材料的发现一直是材料科学中的长期挑战,具有在能源、交通和计算领域的广泛应用潜力。近年来,人工智能(AI)的进步使通过高效利用庞大的材料数据库来加速新材料的搜索成为可能。本文提出了一种基于深度学习(DL)的方法来预测新超导材料。我们从DL网络中合成了一种化合物,并确认其超导性质与预测一致。我们的方法还与基于随机森林(RFs)的先前工作进行了比较。特别是,RFs需要了解化合物的化学性质,而我们的神经网络输入仅依赖于化学组成。借助网络的提示,我们发现了一种新的三元化合物Mo₂₀Re₆Si₄,在5.4K以下表现出超导性。我们进一步讨论了使用AI进行预测所存在的现有限制和挑战,并提出了潜在的未来研究方向。

英文摘要

The discovery of novel superconducting materials is a longstanding challenge in materials science, with a wealth of potential for applications in energy, transportation, and computing. Recent advances in artificial intelligence (AI) have enabled expediting the search for new materials by efficiently utilizing vast materials databases. In this study, we developed an approach based on deep learning (DL) to predict new superconducting materials. We have synthesized a compound derived from our DL network and confirmed its superconducting properties in agreement with our prediction. Our approach is also compared to previous work based on random forests (RFs). In particular, RFs require knowledge of the chemical properties of the compound, while our neural net inputs depend solely on the chemical composition. With the help of hints from our network, we discover a new ternary compound $\textrm{Mo}_{20} \textrm{Re}_{6} \textrm{Si}_{4}$, which becomes superconducting below 5.4 K. We further discuss the existing limitations and challenges associated with using AI to predict and, along with potential future research directions.