Deep neural network enabled active metasurface embedded design


Loading...
Thumbnail Image
Type
Article
Change log
Authors
Zheng, B 
Julian, M 
Tang, H 
Abstract

jats:titleAbstract</jats:title> jats:pIn this paper, we propose a deep learning approach for forward modeling and inverse design of photonic devices containing embedded active metasurface structures. In particular, we demonstrate that combining neural network design of metasurfaces with scattering matrix-based optimization significantly simplifies the computational overhead while facilitating accurate objective-driven design. As an example, we apply our approach to the design of a continuously tunable bandpass filter in the mid-wave infrared, featuring narrow passband (∼10 nm), high quality factors (jats:italicQ</jats:italic>-factors ∼ 10jats:sup2</jats:sup>), and large out-of-band rejection (optical density ≥ 3). The design consists of an optical phase-change material Gejats:sub2</jats:sub>Sbjats:sub2</jats:sub>Sejats:sub4</jats:sub>Te (GSST) metasurface atop a silicon heater sandwiched between two distributed Bragg reflectors (DBRs). The proposed design approach can be generalized to the modeling and inverse design of arbitrary response photonic devices incorporating active metasurfaces.</jats:p>

Description
Keywords
active metasurface, deep neural network, embedded design, phase change material, tunable metasurface
Journal Title
Nanophotonics
Conference Name
Journal ISSN
2192-8606
2192-8614
Volume Title
0
Publisher
Walter de Gruyter GmbH