TUNABLE PERFECT LIGHT ABSORBER FOR PHOTONIC APPLICATIONS

Abstract

Recent progress in nanophotonics has shown remarkable control over the electromagnetic eld at the nanoscale using multilayer thin lm and metamaterial structures. The ability to control the electromagnetic eld at the nanoscale has led to the development of a new class of optical devices with applications in imaging, sensing, and communication. One key application of these devices is the realization of perfect light absorption, which is essential for a wide range of applications, including solar cells, thermal emitters, and sensors. Perfect light absorbers are typically designed using planar multilayer thin lm structures and metal/dielectric nanostructures. However, these devices are limited by their static optical response and cannot be dynamically tuned once they are fabricated. To overcome this limitation, the use of active materials such as liquid crystals, transparent conductive oxides, and phase change materials (PCM) in these devices has been proposed. In particular, PCMs can change their optical properties by undergoing a phase transition, allowing for the dynamic tuning of the absorption spectrum. Chalcogenide alloys such as Ge2Sb2Te5 (GST) are a class of PCMs that have been widely used in optical data storage and recon gurable photonics. The main advantage of using PCMs in the design of perfect absorbers is that the optical properties of PCMs can be dynamically tuned by applying external stimuli such as electrical, optical, or thermal signals. Furthermore, they are non-volatile and can maintain their optical properties even after the stimuli are removed. The design of PCM-based tunable perfect absorbers is challenging due to the complex design space and large number of design parameters. To overcome this challenge, inverse design approaches based on deep learning and optimization algorithms have been proposed. This thesis investigates the use of PCMs in the design of planar multilayer and metamaterial structures for the realization of tunable perfect light absorber for applications including structural color lters, spectrally switchable perfect absorption operating in the current and emerging optical communication wavelengths, and ii molecular identi cation using surface-enhanced infrared absorption spectroscopy. The dynamic tuning of the absorption spectrum is achieved by changing the phase of the PCM using external electrical signals. The thesis also presents various inverse design approaches based on deep learning to design PCM-based tunable perfect absorbers.