Over the last 20 years, the cost of uncooled microbolometer-based imaging systems has drastically decreased while performance has increased. In the simplest terms, the figure of merit for these types of thermal detectors is given in terms of the τ-NETD product, the combination of the thermal time constant and the noise equivalent temperature difference. Considering these factors, optimal system design parameters are investigated to maximize visual information content. This dissertation focuses on improving scene information in the longwave infrared (LWIR) spectrum that has had its validity and quality degraded by noise, blur, and reflected radiance. Taken together, noise and blur degrade image quality, directly affecting system performance for object detectors trained with deep learning. Representing noise with NETD and blur in terms of equivalent angular resolution, this research provides a systematic method for relating design parameters to specific machine vision tasks that are difficult to define in a traditional imaging sense. This method provides for a system design approach based on information requirements rather than improvements to machine vision algorithms. As a machine vision function, automated target recognition (ATR) has improved with new technologies and the wide proliferation of infrared staring focal planes. Infrared search and track (IRST), which is target detection and localization at long ranges of unresolved targets, can be performed by both photon counting and microbolometer systems. The transition from broadband system design to one that involves spectral characterizations of components provides a better understanding of the performance and capabilities of new technologies. Unlike reflective bands such as visible and shortwave infrared (SWIR), reflected radiance reduces contrast in the LWIR, resulting in lost information. This research considers the sky path radiance contribution to the radiant exitance of a scene that reduces contrast, and consequently, information. Results show that reduced contrast can be overcome by utilizing multiband spectral imaging systems to remove the reflected component, thus increasing the scene information available. In addition, better scene consistency can be achieved between day and night when reflected radiance is removed. The multiband LWIR system designs presented take advantage of the low τ-NETD of modern microbolometers and demonstrate feasibility in future multiband applications.

Cubic alkaline-earth fluorides, specifically CaF2 and SrF2, have long been recognized as good laser host candidate materials for high-power amplifiers. Their low linear and nonlinear refractive indices, negative thermo-optic coefficient, high thermal conductivity, low intrinsic optical loss over a broadband spectrum and ability to incorporate rare-earth ions with low quantum defects are just a few of their attractive properties. Traditionally grown from the melt up to foot-size dimensions, these materials can also be advantageously prepared in the form of transparent ceramics to improve on gain uniformity by eliminating dopant segregation and stress-induced birefringence. Transparent ceramic gain media offer additional benefits in terms of enhanced thermal shock resistance due to their fine microstructure and process scalability. The fabrication of transparent ceramics with controlled microstructures is usually achieved by sintering fine powders below their melting point. However, due to the high surface area of these powders, oxide contamination can lead to scattering loss in the sintered parts. In an attempt to circumvent this limitation, this work investigates the fabrication of transparent ceramics by fusion casting and their ceramization via hot forging. Specifically, we have evaluated the processing conditions for fusion casting of Nd:SrF2 and shown how reducing the cooling rate from 25°C/hr to 1.5°C/hr can increase optical transmission up to the Fresnel limit in the near-IR. Similarly, we have delineated the conditions under which a randomized microstructure can be obtained by uniaxial deformation of CaF2 single crystals and Nd:SrF2 fusion cast ceramics, yielding a 30% increase in fracture toughness from 1.7 to 2.2 MPa*m1/2 for CaF2. While further optimization can still be achieved, our best samples exhibit inline scattering losses of 0.002 cm-1 at 1.5 μm compared to 0.25 cm-1 at 1.06 μm for ceramics fabricated via hot-pressing. This two-step approach has the potential to provide a cost-effective pathway to large, low-loss alkaline-earth fluoride laser gain media.

Compact image systems bring up people's attention in the field of target recognition, surveillance, situation awareness or even photography. Conventional metrics assess image system based on image quality without considering systems' volume. More comprehensive metrics, such as General Image-Quality Equation and the Targeting Task Performance metric, incorporates all image system components from object, lenses to detector and even imaging processing algorithm. All these key factors prohibit these metrics from being applied to image system in a convenient manner. Here, we propose a simple metric, volumetric imaging efficiency, considering both image quality and volume. Only concentrate on optical lenses enables the metric being implemented onto conventional bulk optics and flat optics efficiently. Curved image sensor with monocentric lenses shows an exceptional performance based on our metric but potentially challenging in fabrication due to conventional flat substrate process. Normally, this can be done with inorganic photodetector array and perform bending as the last step or organic photodetector being directly deposited on a curved plastic substrate. Inorganic method utilizes state-of-the-art CMOS technology but the interconnects suffer great strain and stress after bending and ultimately runs the risk of device failure while organic device ensures minimum strain, but the fabrication is not compatible with CMOS technology, thus a pattern transfer method is involved for contacts deposition. Here, we introduce both techniques and addressing their challenges. For inorganic device, several interconnect deposition methods are developed and both 1D and 2D bending test are performed to test their stretchability. For organic device, without CMOS circuity, we developed a new type of photodetector, frustrated organic photodetector (F-OPD), which enables single pixel selection by biasing device in different directions. A total of 45 devices are fabricated and perform as an input of 30X30 detectors array. A variety of noise sources are discussed and applied to each pixel. The image is then restored by color leveling and 2-points or 3-points non-uniformity correction (NUC). The results are compared with original figure as a proof of concept showing the capability of the device being extended to an array.

Display market has undergone dramatic changes as the near eye displays (NED) are gaining increasing attention because they offer a deeper level of human-computer interaction with the advancement of electronic devices and computer sciences. The NEDs can be presented in two ways: virtual reality (VR) and augmented reality (AR). The former is completely immersive while the latter combines the digital information with the surrounding scenes. Although several VR headsets have been commercialized for consumers and AR products for prosumers because of their high cost, but there is still a long way to go to satisfy the strict requirements of human vision system. For example, the headset design should be ergonomic so that the users are comfortable when wearing it for a long time. It is critically important to maintain a thin form factor and lightweight while improving the viewing performance, including image quality, resolution, field of view, fatigue free, etc. In this dissertation, we focus on improving the viewing performance of AR/VR displays by developing new cholesteric liquid crystal (CLC) based reflective flat optical elements. Firstly, we introduce the basic CLC properties that are relevant to the reflective patterned optical elements. Secondly, we investigate the flat optical elements with patterned CLC structures from several aspects, including the photoalignment mechanism, polarization field generation, and device fabrication. Then we theoretically analyze the optical properties of the patterned CLC devices, providing the spectral and angular responses of the liquid crystal (LC) grating with different birefringence and device thickness. Finally, we explore new applications of these novel patterned CLC devices to address some major challenges in AR and VR displays. More specifically, a chromatic aberration correction method is applied to the pancake VR system based on our fabricated broadband CLC lens. Such a diffractive optical element exhibits an opposite dispersion behavior to the refractive lens. Thus, by combining our diffractive CLC optical element with a Fresnel lens, the chromatic aberration of the VR system is reduced significantly. In addition, a dual-depth AR system using two custom-designed CLC lenses with different optical powers is presented to mitigate the vergence-accommodation conflict (VAC) issue by generating multiple image depths. To address some existing challenges in waveguide-based AR eyeglasses, we propose and develop a switchable polarization volume grating (PVG) enabled by the patterned CLC layer. Some potential applications are demonstrated, including a significantly suppressed rainbow effect, enhanced light efficiency, and expanded field of view. The unique properties and benefits of switchable PVGs is expected to open a new door for AR and VR displays, especially the novel optical systems for waveguide-based AR displays.

This dissertation reviews the advancements made in chip-scale optical frequency combs and their applications towards optical communications and optical to RF links. We review different chip-scale comb sources and in particular, chip-scale Kerr microresonator frequency combs. Then, we establish the theoretical background in nonlinear optics which allows the formation and stabilization of Kerr solitons in nonlinear cavities. We also discuss the concept of optical injection locking and in particular, multi-tone injection locking which precedes the idea of regenerative harmonic injection locking. We then go on to show the experimental work involved in soliton generation and characterization. We show efforts towards developing an on-chip massive electronic-photonic optical communications link using Kerr soliton frequency combs as equidistant optical carriers in a DWDM based system using a PAM-4 data modulation format. Potential methods for pushing the limits of communication speeds are also highlighted involving the implementation of other degrees of multiplexing such as space division multiplexing and polarization multiplexing. The second application we explore is based on the synchronization of two pulsed sources via regenerative harmonic injection locking, one with a repetition rate in the microwave regime (10s of GHz) and the other in the mm wave domain (100s of GHz). The two sources we use here are an InP based mode locked laser PIC and the Kerr microresonator. Future goals are discussed which involve techniques for the improvement in long-term stability and chip-scale integrability. This proposal envisions future work to achieve high-capacity optical communication links and optical to RF links utilizing chip-scale Kerr microresonator frequency combs.

Nonlinear optical effects occur when strong electromagnetic waves induce changes in a medium that affect its own propagation or that of another wave. Third order optical nonlinearities scale linearly with irradiance and lead to effects like two-photon absorption and nonlinear refraction. This work focuses on the experimental and theoretical study of two-photon absorption in crystalline solids. We begin by detailing the quantum mechanical states of electrons in solids along with the computational approaches to calculate their band structure. Next, a theoretical model for the linear and nonlinear optical interaction of light with matter is presented in a many-body formalism. This first-principles approach derives first and third order nonlinear optical coefficients directly from the many-body Schrödinger equation coupled to the electromagnetic wave equation through the current densities excited by incident electromagnetic fields. The following work examines nondegenerate two-photon absorption in semiconductor quantum well waveguides to determine their suitability as a two-photon lasing medium under population inversion. Experimental pump-probe measurements are presented for a structure comprising GaAs/32% AlGaAs quantum wells. The data is first analyzed by devising a theoretical model for the co-propagation of a strong pump and weak probe pulse within the wave guide sample. After, we present a quantum mechanical model for the electronic states and corresponding optical response of our system to compare to the two-photon absorption coefficients determined from the experimental investigation. The model's excellent agreement with the measured results allows us to extrapolate to the extremely nondegenerate regime, predicting large enhancements in the nondegenerate two-photon absorption coefficients when one pulse has a mid-infrared wavelength. Next, we detail phonon-assisted nondegenerate two-photon absorption in silicon, with the goal of determining which transition pathway best explains the dispersion of ND-2PA coefficients near the band gap energy. After discovering many cases where simplified models break down, we introduce a tool to calculate linear and nonlinear optical properties of materials using density functional theory. This work focuses on the efficient calculation of one- and two-photon absorption, as well as the nonlinear effects arising from excited carrier populations. Finally, a theoretical model is proposed to determine the exact many-body quantum states of materials including electron-electron Coulomb interactions.

Today, most of the photovoltaic cells in the market are made of silicon. Great achievements are being attained every year in terms of reducing the price of this kind of cells and improving their efficiency, reliability and durability. However, further improving the cell performance is a challenging task because of the presence of optical, recombination and resistive loss mechanisms in the cell. This work is focused on the measurement and mitigation of these losses. Mitigation of the optical, recombination and resistive losses at first require quantifying those losses and their impacts on the cell performance metrics accurately. Traditionally, solar cells have been measured using characterization techniques like current-voltage, and Suns-Voc, which express the performance metrics in terms of the global cell parameters for the entire cell. However, solar cell is a large area device and different parts on a cell produce different amount of electricity because of the nonuniform distribution of the crystalline defects over the cell and the process variations. Spatial distributions of the cell parameters are valuable because they provide the in depth information about the root causes behind the performance drop of a cell, and points to its remedy. Camera based luminescence imaging and point by point measurement of quantum efficiency and reflectance on the cell are used in this work to find the spatial distribution of the parameters. A new method of parameter imaging is implemented by incorporating the quantum efficiency scanning with the luminescence measurement. A comprehensive methodology to evaluate losses and process variations in silicon solar cell manufacturing is also presented here. The nature of the distributions and correlations in this study provide important insights about loss mechanisms in industrial solar cells, helping to prioritize efforts for optimizing the performance of the production line. As an effort to mitigate the optical, recombination and resistive losses in the silicon solar cells, self-assembled multifunctional nanostructures are developed. These nanostructures can reduce the optical losses in the near band edge, thus contribute in increasing the photogenerated current density. They also contribute in reducing the surface recombination loss by passivating the silicon surface. Additionally, they shows promising results in reducing spreading resistance, which eventually helps the charge transport mechanism in the cell. An overview of the recent trends and endeavors in silicon photovoltaics is first given, followed by a chapter on the important concepts in silicon photovoltaics. The next chapter describes the solar cell manufacturing process and different performance issues related to it. Chapter 4 introduces different measurement techniques used for quantifying the optical, recombination and resistive losses. The following chapters present the crux of this work: method developed for measurement and mitigation of optical, recombination and resistive losses in silicon photovoltaics.

Near-eye display systems, which project digital information directly into the human visual system, are expected to revolutionize the interface between digital information and physical world. However, the image quality of most near-eye displays is still far inferior to that of direct-view displays. Both light engine and imaging optics of near-eye display systems play important roles to the degraded image quality. In addition, near-eye displays also suffer from a relatively low optical efficiency, which severely limits the device operation time. Such an efficiency loss originates from both light engines and projection processes. This dissertation is devoted to addressing these two critical issues from the entire system perspective. In Chapter 2, we propose useful design guidelines for the miniature light-emitting diode (mLED) backlit liquid crystal displays (LCDs) to mitigate halo artifacts. After developing a high dynamic range (HDR) light engine in Chapter 3, we establish a systematic image quality evaluation model for virtual reality (VR) devices and analyze the requirements for light engines. Our guidelines for mLED backlit LCDs have been widely practiced in direct-view displays. Similarly, the newly established criteria for light engines will shed new light to guide future VR display development. To improve the optical efficiency of near eye displays, we must optimize each component. For the light engine, we focus on color-converted micro-LED microdisplays. We fabricate a pixelated cholesteric liquid crystal film on top of a pixelated QD array to recycle the leaked blue light, which in turn doubles the optical efficiency and widens the color gamut. In Chapter 5, we tailor the radiation pattern of the light engine to match the etendue of the imaging systems, as a result, the power loss in the projection process is greatly reduced. The system efficiency is enhanced by over one-third for both organic light-emitting diode (OLED) displays and LCDs while maintaining indistinguishable image nonuniformity. In Chapter 6, we briefly summarize our major accomplishments.

Physical-layer key distribution is an area of active research due to vulnerabilities in current digital encryption methods, which rely on computational security. We propose and demonstrate a high-speed physical-layer key generation and distribution method based on random mode mixing in multimode optical fibers. We refer to this method as random-channel cryptography (RCC). In RCC, a key is extracted from the channel state of a shared multidimensional reciprocal channel. Projection operators reduce the signal to a single degree of freedom (DOF) for the legitimate users, while the signal is spread over many DOFs anywhere accessible to eavesdroppers. This produces a large asymmetry between the eavesdroppers' and legitimate users' measurement complexities. Furthermore, signal-to-noise ratio analysis reveals that RCC is information-theoretically secure under certain attacks. However, initial demonstrations of RCC offered very low key rate-distance products. A nine orders-of-magnitude increase in the key rate-distance product was demonstrated using techniques from traditional telecommunications, such as high-speed modulation, wavelength-division multiplexing, and advanced modulation formats. Compared to other physical-layer key distribution methods, RCC is easy-to-implement, robust, and offers high security.

Since the concepts of augmented reality (AR) and virtual reality (VR) were introduced, they have attracted people's attention worldwide, both in the industry and research areas. As the most promising hardware architecture that can bring AR/VR into daily life, near eye displays (NEDs) have been studied and investigated heavily over the past half-century, especially the concept of "Metaverse" introduced by some top companies in recent years. However, the form factor and optical efficiency are two major bottlenecks for the current NEDs before they can become the major platform. Liquid crystal (LC) flat optics have several advantages, including compact, high diffraction efficiency, easy to pattern, highly transparent and low cost. Therefore, they are idea candidates for NEDs applications. In this dissertation, we focus on the novel LC flat optics applications in the NEDs, aiming to reduce the system form factor and enhance the system optical efficiency. The first half surrounds VR applications and systems, which adopt transmission-type LC flat optics. The second half covers AR system design and demonstration, which takes the advantages of reflection-type LC flat optics. In VR part, we demonstrate an approach to double the optical efficiency of VR systems based on a directional backlight and a diffractive deflection film (DDF), which is a specially designed LC flat optics. Our approach works well in both Fresnel and "pancake" VR systems. We also have the simulation model, which exhibits results highly consistent with the experiment. What's more, a new ultra-compact VR system is also proposed and demonstrated in this dissertation. In this ultra-compact VR system, an LC deflector is inserted into the imaging optics and it can achieve a process called polarization interpolation. This process helps reduce the distance from the display panel to the imaging optics by 50% in theory. In AR part, we design and demonstrate a gaze matched Maxwellian-view AR system pupil steering system. This system applies the LC flat optics as the optical combiner. In the demo, this system achieves many good properties, including compact form factor, high optical efficiency, gaze matching, extended eyebox, aberration free, good ambient light transmittance and relatively large field of view. The proposed applications and systems with LC flat optics are attractive for next-generation NEDs.