Date of Award

2019

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy in Electrical Engineering (PhD)

Administrative Home Department

Department of Electrical and Computer Engineering

Advisor 1

Durdu Guney

Committee Member 1

Paul Bergstrom

Committee Member 2

Christopher Middlebrook

Committee Member 3

Miguel Levy

Abstract

Much research effort has been spent in the 21st century on superresolution imaging techniques, methods which can beat the diffraction limit. Subwavelength composite structures called ``metamaterials" had initially shown great promise in superresolution imaging applications in the early 2000s, owing to their potential for nearly arbitrary capabilities in controlling light. However, for optical frequencies they are often plagued by absorption and scattering losses which can decay or destroy their interesting properties. Similar issues limit the application of other superresolution devices operating as effective media, or metal films that can transfer waves with large momentum by supporting surface plasmon polaritons. In this dissertation, new methods of mitigating the loss of object information in lossy and noisy optical imaging systems are presented. The result is an improvement in the upper bound on lateral spatial resolution. A concentration is placed on metamaterial and plasmonic imaging systems, and the same methods are subsequently adapted to more conventional far-field imaging systems. First, through numerical simulation it is shown that a lossy metamaterial lens has degraded imaging performance which can be partially compensated by deconvolution post-processing of the resultant image. This post-processing procedure is then shown to emulate a physical process called plasmon injection, which has been previously implemented to effectively remove the losses in a plasmonic metamaterial. Next, a more realistic scenario is considered; a thin film of silver acting as a near-field plasmonic ``superlens." In this case, methods are implemented to model incoherent light propagation so that the image can be reconstructed using only intensity data, removing the need for phase measurement. The same procedure from above is followed, and the resolution is enhanced. To push the resolution further, a spatial filtering method called active convolved illumination is developed to overcome the resolution limit set by the noise floor of the system. Finally, the spatial filtering methods are applied to more a more conventional far-field imaging system. Supported by experiment, the lateral resolution of a low numerical aperture imaging system is improved by blocking photons at the Fourier plane. For coherent light, a diffractive superlens is designed which uses the same principles from the above theory, except it encodes the high spatial frequency waves into propagating waves via a diffraction grating. The result is lateral resolution performance that surpasses similar previously published devices by 10 nm at a wavelength more than 80 nm longer.

Available for download on Wednesday, August 12, 2020

Share

COinS