Scholarlink

<p>The discovery of high-performance thermoelectric (TE) materials remains a grand challenge in energy science, typically constrained by slow, high-throughput screening and empirical intuition. Here, we report on the emergent intermetallic compound Fe–Cu–Sn, a previously uninvestigated TE system computationally predicted using an Artificial Neural Network (ANN) model that took optimized TE properties as input. We transition from the predicted solution to a rigorous experimental validation, with the primary focus being a comprehensive metallurgical investigation. The results demonstrated that the optimal performance was achieved via the Mechanical Alloying and Spark Plasma Sintering (SPS) route. The resultant nanostructured Fe–Cu–Sn sample exhibited synergistic transport properties, driven by the energy filtering effect at nanoscale Fe<sub>2</sub>Sn and Cu<sub>6</sub>Sn<sub>5</sub> phase boundaries. The low <i>κ</i><sub><i>L</i></sub> was attributed to dual-mechanism phonon scattering, characteristic of a crystalline-electron–phonon-glass structure. Critically, the experimental results validated the ANN model’s core prediction, confirming its high accuracy in identifying the optimal electronic composition.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Fe–Cu–Sn: metallurgical investigation and thermoelectric properties of a novel ANN-discovered compound

  • 期刊论文

  • 发表日期：

    2026年4月29日

查看全文
  • Seyed Ali Hosseini Khorasani,
  • Ehsan Borhani,
  • Mohammad Yousefieh,
  • Amin Janghorbani

摘要

The discovery of high-performance thermoelectric (TE) materials remains a grand challenge in energy science, typically constrained by slow, high-throughput screening and empirical intuition. Here, we report on the emergent intermetallic compound Fe–Cu–Sn, a previously uninvestigated TE system computationally predicted using an Artificial Neural Network (ANN) model that took optimized TE properties as input. We transition from the predicted solution to a rigorous experimental validation, with the primary focus being a comprehensive metallurgical investigation. The results demonstrated that the optimal performance was achieved via the Mechanical Alloying and Spark Plasma Sintering (SPS) route. The resultant nanostructured Fe–Cu–Sn sample exhibited synergistic transport properties, driven by the energy filtering effect at nanoscale Fe2Sn and Cu6Sn5 phase boundaries. The low κL was attributed to dual-mechanism phonon scattering, characteristic of a crystalline-electron–phonon-glass structure. Critically, the experimental results validated the ANN model’s core prediction, confirming its high accuracy in identifying the optimal electronic composition.