Scholarlink

Abstract <p>The targeted modulation of catalytic performance in sub-nanometer Pt–Sn alloy clusters is hampered by the unclear atomic-level structure–property relationship. This knowledge gap stems from their intrinsically diverse and irregular configurations lacking long-range order at sub-nanometer scale. Herien, the machine-learning atomic simulations and DFT calculations were integrated to accurately identify the structural evolution of sub-nanometer Pt–Sn clusters, establishing descriptors of structural and electronic properties across the full compositional range. Global exploration of the potential energy surface establishes a comprehensive structural stability phase diagram for the Pt–Sn system, unveiling a size-dependent “promotion then inhibition” effect on cluster stability and confirming the thermodynamic preference of Pt<sub>1</sub>Sn<sub>1</sub> stoichiometry. Quantitative analysis of cluster structure uncovers a counter-intuitive phenomenon: Sn surface segregation drives the counter-intuitive decrease of surface Pt site density with increasing cluster size. Moreover, when the Pt:Sn ratio reaches 1:2 or lower, surface Pt sites achieve mono-dispersion independently of the overall cluster size. The coordination number of surface Pt is identified as the key quantitative descriptor governing the d-band center, with Pt–Pt bonding playing a significantly stronger role in modulating the d-band center than Pt–Sn interactions. These findings provide a “handbook style” fundamental basis for the rational design of improved sub-nanometer Pt–Sn cluster catalysts.</p> Graphical abstract <p>The machine-learning atomic simulations and DFT calculations were integrated to accurately identify the structural evolution of subnanometer Pt-Sn clusters, establishing descriptors of structural and electronic properties across the full compositional range.</p> <p></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

Deciphering the fine structure of Pt–Sn sub-nanometer clusters and d-band center tuning mechanism via machine learning

  • 期刊论文

  • 发表日期：

    2026年5月14日

查看全文
  • Zhuang Qu,
  • Qiang Li,
  • Kai He,
  • Ye Yang,
  • Hanbing Song,
  • Xu Wang,
  • Chen Yang,
  • Yucai Qin,
  • Lijuan Song

摘要

Abstract

The targeted modulation of catalytic performance in sub-nanometer Pt–Sn alloy clusters is hampered by the unclear atomic-level structure–property relationship. This knowledge gap stems from their intrinsically diverse and irregular configurations lacking long-range order at sub-nanometer scale. Herien, the machine-learning atomic simulations and DFT calculations were integrated to accurately identify the structural evolution of sub-nanometer Pt–Sn clusters, establishing descriptors of structural and electronic properties across the full compositional range. Global exploration of the potential energy surface establishes a comprehensive structural stability phase diagram for the Pt–Sn system, unveiling a size-dependent “promotion then inhibition” effect on cluster stability and confirming the thermodynamic preference of Pt1Sn1 stoichiometry. Quantitative analysis of cluster structure uncovers a counter-intuitive phenomenon: Sn surface segregation drives the counter-intuitive decrease of surface Pt site density with increasing cluster size. Moreover, when the Pt:Sn ratio reaches 1:2 or lower, surface Pt sites achieve mono-dispersion independently of the overall cluster size. The coordination number of surface Pt is identified as the key quantitative descriptor governing the d-band center, with Pt–Pt bonding playing a significantly stronger role in modulating the d-band center than Pt–Sn interactions. These findings provide a “handbook style” fundamental basis for the rational design of improved sub-nanometer Pt–Sn cluster catalysts.

Graphical abstract

The machine-learning atomic simulations and DFT calculations were integrated to accurately identify the structural evolution of subnanometer Pt-Sn clusters, establishing descriptors of structural and electronic properties across the full compositional range.