Deep-learning-enabled inverse engineering of multi-wavelength invisibility-to-superscattering switching with phase-change materials

Abstract

Inverse design of nanoparticles for desired scattering spectra and dynamic switching between the two opposite scattering anomalies, i.e. superscattering and invisibility, is an interesting topic in nanophotonics. However, traditionally the design process is quite complicated, which involves complex structures with many choices of synthetic constituents and dispersions. Here, we demonstrate that a well-trained deep-learning neural network can handle these issues efficiently, which can not only forwardly predict scattering spectra of multilayer nanoparticles with high precision, but also inversely design the required structural and material parameters quite efficiently. Moreover, we show that the neural network is capable of finding out multiwavelength invisibility-to-superscattering switching points at the desired wavelengths in multilayer nanoparticles composed of metals and phase-change materials. Our work paves a way of deep learning for the inverse design of nanoparticles with desired scattering spectra.

Hermitian and Non-Hermitian Dirac-Like Cones in Photonic and Phononic Structures

“…Deep learning, as a data-driven methodology, could overcome complicated design problems with growth of structural complexity and higher degree of freedom through training the artificial neural network [36,181]. A trained neural network can be used as a fast, general purpose predictor of optical and electromagnetic responses of complicated 3D structures, and is particularly efficient in solving notoriously difficult inverse problems in nanophotonics [182][183][184][185]. Another way is the utilization of the systematic band engineering method that has realized Dirac cones at arbitrary k point in the Brillouin zone of PhCs without symmetry [35].…”

Intelligent metaphotonics empowered by machine learning

“…Namely, the scattering efficiency was below 10 −2 in the range from 400 nm to 700 nm, dropping below 10 −4 between 510 nm and 550 nm. Another demonstration utilizes phasechange materials for realizing invisibility-to-superscattering switching [43]. Developed DL approach allowed to predict required materials and structural parameters to realize simultaneously satisfied conditions for super-and near-zero scattering for two phase states, see Fig.…”

Inverse Design of Nanophotonic Devices Using Generative Adversarial Networks with the Sim-NN Model and Self-Attention Mechanism

“…However, this method is not heuristic such that if the target response changes, the optimization process must be restarted from the very beginning. The emergence of data-driven methods [ 8 ] based on neural networks (NNs) is now showing promising signs for solving these problems and making major breakthroughs in these tasks [ 9 , 10 , 11 ]. However, the non-uniqueness of inverse scattering is the root cause that hinders the convergence of inverse NN models [ 12 , 13 ].…”