Search for: All records

A deep learning aided optimization algorithm for the design of flat thin-film multilayer optical systems is developed. The authors introduce a deep generative neural network, based on a variational autoencoder, to perform the optimization of photonic devices. This algorithm allows one to find a near-optimal solution to the inverse design problem of creating an anti-reflective grating, a fundamental problem in material science. As a proof of concept, the authors demonstrate the method’s capabilities for designing an anti-reflective flat thin-film stack consisting of multiple material types. We designed and constructed a dielectric stack on silicon that exhibits an average reflection of 1.52 %, which is lower than other recently published experiments in the engineering and physics literature. In addition to its superior performance, the computational cost of our algorithm based on the deep generative model is much lower than traditional nonlinear optimization algorithms. These results demonstrate that advanced concepts in deep learning can drive the capabilities of inverse design algorithms for photonics. In addition, the authors develop an accurate regression model using deep active learning to predict the total reflectivity for a given optical system. The surrogate model of the governing partial differential equations can then be broadly used in the design of optical systems and to rapidly evaluate their behavior.

Graphene has transformed the fields of plasmonics and photonics, and become an indispensable component for devices operating in the terahertz to mid-infrared range. Here, for instance, graphene surface plasmons can be excited, and their extreme interfacial confinement makes them vastly effective for sensing and detection. The rapid, robust, and accurate numerical simulation of optical devices featuring graphene is of paramount importance and many groups appeal to Black-Box Finite Element solvers. While accurate, these are quite computationally expensive for problems with simplifying geometrical features such as multiple homogeneous layers, which can be recast in terms of interfacial (rather than volumetric) unknowns. In either case, an important modeling consideration is whether to treat the graphene as a material of small (but non-zero) thickness with an effective permittivity, or as a vanishingly thin sheet of current with an effective conductivity. In this contribution we ponder the correct relationship between the effective conductivity and permittivity of graphene, and propose a new relation which is based upon a concrete mathematical calculation that appears to be missing in the literature. We then test our new model both in the case in which the interface deformation is non-trivial, and when there are two layers of graphene with non-flat interfacial deformation.