Gradient-based optimization of core-shell particles with discrete materials for directional scattering.

Designing nanophotonic structures traditionally grapples with the complexities of discrete parameters, such as real materials, often resorting to costly global optimization methods. This paper introduces an approach that leverages generative deep learning to map discrete parameter sets into a continuous latent space, enabling direct gradient-based optimization. For scenarios with non-differentiable physics evaluation functions, a neural network is employed as a differentiable surrogate model.

Self-Assembled Silicon@Silica Metasurfaces with High-Quality Resonances in the Infrared.

2D assemblies of resonant dielectric particles constitute promising materials for the next generation of photonic devices, thanks to their low optical losses and intense electromagnetic response. However, bottom-up synthesis methods present many difficulties when targeting metasurface applications, particularly due to the high degree of positional disorder and the size dispersion of the resonant particles. This work presents the fabrication of core-shell silicon@silica particles with multipolar resonances in the visible and near-infrared.