Metamaterial Eigenmodes beyond Homogenization

TL;DR

This paper introduces a rigorous method for analyzing metamaterial eigenmodes that surpasses traditional homogenization theories, accurately capturing complex band structures and explaining the effectiveness of Maxwell-Garnett theory at high fill fractions.

Contribution

It presents a Greens-Galerkin based eigenproblem approach for binary periodic metamaterials, extending analysis beyond homogenization limits and explaining Maxwell-Garnett theory's success at high fill fractions.

Findings

Accurate multi-valued Floquet bandstructure computation

Method valid beyond homogenization assumptions

Explains Maxwell-Garnett theory's high fill fraction accuracy

Abstract

Metamaterial homogenization theories usually start with crude approximations that are valid in certain limits in zero order, such as small frequencies, wave vectors and material fill fractions. In some cases they remain surprisingly robust exceeding their initial assumptions, such as the well-established Maxwell-Garnett theory for elliptical inclusions that can produce reliable results for fill fractions far above its theoretical limitations. We here present a rigorous solution of Maxwell's equations in binary periodic materials employing a combined Greens-Galerkin procedure to obtain a low-dimensional eigenproblem for the evanescent Floquet eigenmodes of the material. In its general form, our method provides an accurate solution of the multi-valued complex Floquet bandstructure, which currently cannot be obtained with established solvers. It is thus shown to be valid in regimes where homogenization theories naturally break down. For small frequencies and wave numbers in lowest order, our method simplifies to the Maxwell-Garnett result for 2D cylinder and 3D sphere packings. It therefore provides the missing explanation why Maxwell-Garnett works well up to extremely high fill fractions of approximately $50\%$ depending on the base materials, provided the inclusions are arranged on an isotropic lattice.