Machine Learning‐Based Optimization of Chiral Photonic Nanostructures: Evolution‐ and Neural Network‐Based Designs

Abstract

Chiral photonics opens new pathways to manipulate light–matter interactions and tailor the optical response of metasurfaces and ‐materials by nanostructuring nontrivial patterns. Chirality of matter, such as that of molecules, and light, which in the simplest case is given by the handedness of circular polarization, have attracted much attention for applications in chemistry, nanophotonics and optical information processing. The design of chiral photonic structures using two machine learning methods, the evolutionary algorithm, and neural network approach, for rapid and efficient optimization of optical properties for dielectric metasurfaces, is reported. The design recipes obtained for visible light in the range of transition‐metal dichalcogenide exciton resonances show a frequency‐dependent modification in the reflected light's degree of circular polarization, that is represented by the difference between left‐ and right‐circularly polarized intensity. Our results suggest the facile fabrication and characterization of optical nanopatterned reflectors for chirality‐sensitive light–matter coupling scenarios employing tungsten disulfide as possible active material with features such as valley Hall effect and optical valley coherence.

AI-Based Metamaterial Design

“…A single-layer focusing lens in the near-infrared region with eight controllable responses subjected to different combinations of working frequencies and linear polarization states was developed. Similar capability and automatization via the data-driven scheme for optical lens design were presented with the transition to NN. ,− …”

Recent Progress on Plasmonic and Dielectric Chiral Metasurfaces: Fundamentals, Design Strategies, and Implementation

“…Reproduced with permission. [52] Copyright 2022, Mey et al published by Wiley-VCH GmbH. B) The deep-learning-based optimization system used the reverse design strategy based on a target-driven conditional generative network (TCGN) to predict the optimal parameters of an all-silicon chiral structure working in the terahertz regime for the required CD response.…”

“…For rapid and ef-ficient optimization, Mey et al demonstrated the utilization of two machine-learning approaches named evolutionary algorithm and neural network (Figure 9A). [52] The dielectric resonator's spectral properties were adjusted to the spectrum of transition metal dichalcogenide exciton resonances which are popular materials [123,124] for nanophotonic and optoelectronic applications. Moreover, the proposed framework provides a frequencydependent modification of CD responses.…”

“…Moreover, machine-learning-based design techniques are reported in the literature using plasmonic and dielectric materials to optimize chiral metasurfaces. [52][53][54][55][56][57][58][59] Furthermore, there has been a development to generate nonlinear optical responses using chiral metasurfaces. [60][61][62][63] Considering chiral metasurfaces as the hot research area in optics and photonics, this paper reviewed the technology shift from plasmonic to dielectric chiral metasurfaces with fundamental concepts, design strategies, and their implementation for various applications.…”

Deep learning for circular dichroism of nanohole arrays

“…DL is being increasingly utilized in diverse fields, such as fiber optics [24], semiconductors [25], and design of electromagnets [26,27]. Researchers have already used deep learning to implement the study of chiral nanostructures [28][29][30]. In nanophotonics, DL has been adopted for resonant mode analysis [31,32] and spectra calculation [33,34].…”