Scholarlink

<p>Topological states, which often manifest as localized modes at interfaces between distinct topological phases, and their manipulation have attracted considerable interest in research. Here, we demonstrate a versatile approach to sculpting topological modes into desired shapes by incorporating various artificial gauge fields —including scalar, vector, and imaginary gauge potentials—and leveraging the power of artificial neural networks. The chose gauge fields enable precise tuning of the dissipation of the topological modes across that of bulk modes, facilitating a transition from localized states to fully delocalized ones. These eigen modes can be precisely engineered by neural networks, achieving tailored profiles of topological states, which remain spectrally isolated from bulk bands and exhibit minimal loss compared to other modes. Our theoretical results are experimentally validated on silicon photonic platforms, demonstrating flexible manipulation of mode profiles. This approach enables the design of topological states with customized properties, offering potential for diverse applications in photonics and beyond.</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

Artificial gauge fields for sculpting topological modes on photonic chips

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月4日

查看全文
  • Zhiyuan Lin,
  • Jian Li,
  • Wange Song,
  • Xueyun Li,
  • Haoran Xin,
  • Xian Long,
  • Chen Chen,
  • Shining Zhu,
  • Tao Li,
  • Shuang Zhang

摘要

Topological states, which often manifest as localized modes at interfaces between distinct topological phases, and their manipulation have attracted considerable interest in research. Here, we demonstrate a versatile approach to sculpting topological modes into desired shapes by incorporating various artificial gauge fields —including scalar, vector, and imaginary gauge potentials—and leveraging the power of artificial neural networks. The chose gauge fields enable precise tuning of the dissipation of the topological modes across that of bulk modes, facilitating a transition from localized states to fully delocalized ones. These eigen modes can be precisely engineered by neural networks, achieving tailored profiles of topological states, which remain spectrally isolated from bulk bands and exhibit minimal loss compared to other modes. Our theoretical results are experimentally validated on silicon photonic platforms, demonstrating flexible manipulation of mode profiles. This approach enables the design of topological states with customized properties, offering potential for diverse applications in photonics and beyond.