Scholarlink

<p>In the realm of semiconductor technology, the precise tuning of band gaps is crucial for the development of sophisticated devices. Monolayer transition metal dichalcogenides (TMDMs) are of particular interest due to their transition from indirect to direct band gap semiconductors, characterized by effective light absorption and enhanced charge transfer capabilities. This study utilizes machine learning to efficiently identify TMDMs with direct band gaps, employing an Artificial Neural Network that boasts an 88% accuracy rate. Complementary methods like XGBoost and Random Forest also demonstrate promising potential. Shapley Additive Explanations analysis uncovers key factors influencing the likelihood of a compound being a TMDM, including oxygen and sulfur content, the range of atomic numbers, the average number of s-valence electrons, and the presence of elements like Co, Zr, and Ti. This investigation not only accelerates the identification of single-layer TMDMs but shows potential for predicting novel materials, significantly advancing research in the field.</p> Graphical abstract

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

Accelerated exploration of direct band gap transition metal dichalcogenide monolayers through machine learning-driven discovery

  • 期刊论文

  • 发表日期：

    2025年7月31日

查看全文
  • Wanli Jia,
  • Pengcheng Ma,
  • Qi Guan,
  • Lin Zhang,
  • Xinmei Wang,
  • Ruofan Shen,
  • Wei Wang

摘要

In the realm of semiconductor technology, the precise tuning of band gaps is crucial for the development of sophisticated devices. Monolayer transition metal dichalcogenides (TMDMs) are of particular interest due to their transition from indirect to direct band gap semiconductors, characterized by effective light absorption and enhanced charge transfer capabilities. This study utilizes machine learning to efficiently identify TMDMs with direct band gaps, employing an Artificial Neural Network that boasts an 88% accuracy rate. Complementary methods like XGBoost and Random Forest also demonstrate promising potential. Shapley Additive Explanations analysis uncovers key factors influencing the likelihood of a compound being a TMDM, including oxygen and sulfur content, the range of atomic numbers, the average number of s-valence electrons, and the presence of elements like Co, Zr, and Ti. This investigation not only accelerates the identification of single-layer TMDMs but shows potential for predicting novel materials, significantly advancing research in the field.

Graphical abstract