Optical Propagation in Anisotropic Metamaterials: Application to Analysis and Design of Metallo-Dielectric Filters

Date of Award

2021

Degree Name

Ph.D. in Electro-Optics and Photonics

Department

Department of Electro-Optics and Photonics

Advisor/Chair

Partha P. Banerjee

Abstract

This work involves the analysis and design of metallo-dielectric structures for potential use as transmission filters. The Berreman matrix method, effective medium theory and anisotropic transfer matrix method are used to analyze propagation of electromagnetic/optical fields through these anisotropic metamaterial structures on various substrates. The design of such structures is performed using artificial neural networks. Both transverse electric and transverse magnetic polarizations are investigated. Effective medium theory along with the Berreman matrix method is used to analyze the optical properties such as reflection and transmission spectra through anisotropic media, which can be implemented using multilayer metallo-dielectric stacks of sub-wavelength dimensions. While multilayer anisotropic stacks of arbitrary thickness can be rigorously analyzed using 4x4 transfer matrix, in this work, a simplified 2x2 anisotropic transfer matrix technique is developed to analyze optical propagation through multilayer uniaxial stacks of arbitrary thicknesses. Optical transmission of a multilayer silver-zinc oxide stack deposited on a quartz substrate is modeled with this 2x2 anisotropic transfer matrix method along with effective medium theory and reconciled with experimental observations. Results indicate that this approach can be used for in situ assessment of the complex refractive indices of constituent metal and dielectric layers. Additionally, the anisotropic 2x2 transfer matrix method enables the possibility of modeling the transmission of the same metallo-dielectric structure deposited on an uniaxial electro-optic substrate. Simulation results predict that adjusting the bias field across the substrate results in an electrically tunable transmission filter. Following the analysis of transmission filters, an artificial neural network technique is used for the design of the optimum metallo-dielectric structure to achieve a given transmission spectrum. The universal approximation theorem using ten hidden layers is used in the artificial neural network. The input parameter space has information such as layer thicknesses, operating wavelength, linewidth and peak transmittance. The activation function is used to determine optimized layer thicknesses to achieve the desired Gaussian and super-Gaussian spectra.

Keywords

Electromagnetics, Nanotechnology, Nanoscience

Rights Statement

Copyright © 2021, author

Link to Full Text

Share

COinS