Electrical Engineering and Computer Science electronic theses and dissertations (MU)
Permanent URI for this collection
The items in this collection are the theses and dissertations written by students of the Department of Electrical and Computer Engineering. Some items may be viewed only by members of the University of Missouri System and/or University of Missouri-Columbia. Click on one of the browse buttons above for a complete listing of the works.
Browse
Recent Submissions
Item The transverse thermoelectric power in tellurium films(University of Missouri--Columbia, 1971) Sandfort, Robert Melvin; Charlson, Earl J."Thermoelectric phenomena were first observed more than a century ago.1.1,1.2 Since that time a vast amount of progress has been made in understanding the thermodynamic and electronic principles which govern these phenomena. Most of the early thermoelectric research work in semiconducting materials was oriented towards optimising the parameters for direct conversion of heat to electrical power or for using electrical power to pump heat. Often overlooked is the insight into the nature of semiconduction processes which has been gained as a result of using the thermoelectric effects, particularly the Seebeck effect, as an experimental tool. One of the most frequently used experimental techniques to determine carrier concentration type in semiconductors is the thermal probe. This technique consists of placing a heated metal probe on the surface of a semiconductor to which another large area ohmic contact has been made. Charge carriers will diffuse from the region of the hot probe under the influence of a carrier concentration and kinetic energy gradient. Equilibrium is restored by the gradient in electrostatic potential which results from the average increase in majority carriers in the cooler regions of the semiconductor. The potential of the hot probe relative to the cooler contact will have a sign depending on the majority carrier type--negative for p-type and positive for n-type. While conceptually very simple, this technique is quite complicated in terms of the physical processes taking place in the semiconductor, and frequently erroneous results are obtained with its use."--Introduction.Item Vapor-chamber-based modular cold plate for TO-247 power semiconductor CFD setup, desgin & verification(University of Missouri--Columbia, 2025) Auduong, Jason; Huang, QingyunTO-247 packaged power semiconductors offer a low-cost alternative to power modules for high-volume, high-power applications, but their highly localized heat fluxes create significant thermal management challenges. This work proposes a compact 13.5 × 18.95 mm² two-phase cold plate that integrates a vapor chamber (VC) with a liquid-cooled heat sink, designed specifically for fabrication via laser powder bed fusion (L-PBF). The additively manufactured design incorporates enclosed channel networks and micropillar structures to enhance coolant distribution and heat spreading. A key contribution of this study focuses on the use of computational fluid dynamics (CFD) simulations in ANSYS Icepak to evaluate the thermal performance of the proposed dual-phase architecture. The CFD analysis directly compares a single-phase baseline with the integrated VC–liquid-cooling configuration under a 400 W/cm² heat load. CFD results demonstrate that the proposed dual-phase architecture substantially reduces peak temperature and improves spatial temperature uniformity compared with a single-phase cold-plate baseline. Under identical operating conditions, the integrated VC–liquid-cooling design achieves approximately 25% lower maximum temperature, demonstrating the effectiveness of combining L-PBF design freedom with two-phase thermal management for TO-247 devices.Item A study of panorama monocular depth estimation based scene understanding using deep learning methods(University of Missouri--Columbia, 2025) Mohadikar, Payal Vinayak; Duan, YePanorama images carry comprehensive scene representation and are widely useful for applications like AR/VR, robotics, and autonomous driving that require holistic scene understanding. Depth estimation, being the core component of scene understanding, has been widely researched, achieving significant improvement for perspective inputs. However, for 360/panorama input, the methods still produce low-quality, globally inconsistent depths, indicating poor generalization ability due to challenges such as inherent spherical distortion and relatively few training data. Recent state-of-theart (SOTA) methods utilize multiple projected distortionless tangent patches to mitigate spherical distortion, but they lose the learning of holistic contextual information, leading to global discrepancies and merging artifacts in the final merged-back 360 depths. In this study, we propose to mitigate the existing global inconsistency, merging artifact and poor-generalization issues via novel panorama depth estimation networks that take a single panorama image to estimate 360 depth maps. Extensive experiments on benchmark datasets demonstrate that our networks achieve benchmark quantitative and qualitative performance compared to the existing methods, with zero-shot performance (of the SN360 method) attaining SOTA results, improving Abs Rel by 22.8% and RMSE by 11.0%. Panorama images also offer a compelling input modality for immersive 3D scene reconstruction. However, existing SOTA scene reconstruction 3D Gaussian Splatting (3DGS) methods for panoramas require multiple views and/or per-scene optimization, limiting their usability in practical settings. We present a novel feed-forward 3DGS framework that reconstructs a full indoor 3D scene and enables high-quality panoramic novel view synthesis from only a single input panorama image. For this, we utilize a depth foundational model prior to initialize Gaussian parameters. The proposed 3DGS network then estimates per-pixel Gaussian parameters to form an explicit 3D Gaussian scene representation. To improve high-frequency appearance and reduce splat-induced artifacts, we introduce a latent diffusion–based enhancement module that conditions on the input panorama to denoise and sharpen the rendered novel views while maintaining consistency with the input image. Experiments on multiple datasets show that our method achieves strong performance compared to multi-view panorama methods, despite requiring only a single image. Our results demonstrate the feasibility of efficient, feed-forward, single-panorama image 3D reconstruction for indoor environments.Item Lead-free perovskites for solid-state lighting : integrating advanced materials with 3D-printed optics for efficient and stable white light emission(University of Missouri--Columbia, 2025) Yue, Yang; Zhu, Peifen[EMBARGOED UNTIL 08/01/2026] Perovskite materials have emerged as a promising class of optoelectronic materials with significant commercial potential in lighting and display applications. This study explores three key areas: the synthesis and characterization of germanium-based lead-free metal halide perovskites, the fabrication and evaluation of high-performance lead-free white light-emitting diodes (WLEDs), and the theoretical and experimental investigation of light extraction efficiency (LEE) in 3D-printed transparent resin lenses. To develop lead-reduced perovskite materials, germanium-based CsGexPb(1-x)Br3 perovskite nanocrystals (NCs) were synthesized using a modified ligand-assisted reprecipitation (LARP) method. These NCs were structurally characterized through X-ray diffraction (XRD) and high-resolution transmission electron microscopy (HR-TEM) and optically analyzed using absorption and photoluminescence spectroscopy. The introduction of germanium into halide perovskites presents a viable pathway toward sustainable optoelectronic devices, offering several advantages, including high efficiency (>25%) for solar cells and light-emitting diodes, cost-effective raw materials and production, tunable bandgap through compositional adjustments, a broad color gamut, flexible shape and dimensionality, and lightweight properties. Lead-free perovskites-based white light-emitting diodes (WLEDs) were fabricated using 3D-printed lead-free perovskite color conversion layers (CCLs). The fabricated WLEDs consist of an ultraviolet (UV) LED chip array and three CCLs: yellow-emitting double perovskites, green-emitting Cu-based perovskites, and a blue-emitting photon downconverter. Liquid crystal display (LCD)-based 3D-printing technique was employed to encapsulate lead-free photon downconverters into three thin resin-based layers, respectively. The WLED devices show excellent optical performance with luminous efficacy of radiation (LER) up to 260 lm/W, color rendering index (CRI) up to 97, tunable correlated color temperature (CCT) from warm white light, neutral white light (x=0.32, y=0.33) to cold white light, tunable circadian action factor (CAF), and extremely small delta u v (Duv). On the other hand, the daily measurement of photoluminescence quantum yield (PLQY) shows good stability of the printed CCLs, which proves encapsulating perovskites in transparent resin by 3D-printing technique is effective in protecting perovskite from degradation. These experimental results confirm the validity of fabricating WLEDs using the 3D-printing method to form anti-degradation perovskite-resin CCLs. A significant portion of LED-emitted light is trapped within LEDs due to total internal reflection, caused by refractive index mismatch and suboptimal lens geometry. To address this issue and improve light extraction efficiency (LEE), lens shape and dimensions are critical. This study fabricated dome, cone, and array resin lenses with varying dimensions using LCD-based 3D printing. LEE was then evaluated both theoretically, using the Monte Carlo method, and experimentally, by measuring external quantum efficiency (EQE) from spectral power distribution (SPD) data. The LEE trends obtained from both methods were consistent. The results demonstrate that, within each lens type, a dome lens with a height/radius ratio of 1, a cone lens with a 20° semi-angle, and a 20Ã--20 semi-sphere array lens exhibited the highest LEE. This work validates two complementary theoretical and experimental approaches for assessing LEE across diverse lens geometries.Item Engineering sustainable and cost-effective perovskite alternatives for advanced optoelectronic devices(University of Missouri--Columbia, 2025) Dzorkpata, Christopher Selom; Zhu, Peifen[EMBARGOED UNTIL 08/01/2026] Over the past decade, the surge in global energy consumption driven by lighting technologies has fueled the need for innovative electroluminescent materials that are both energy-efficient and exhibit exceptional emission properties. In response, metal halide perovskites (MHPs) have emerged as promising semiconductors due to their remarkable optical and electronic characteristics, offering the potential to revolutionize modern optoelectronic devices, such as light-emitting diodes (LEDs), by providing more cost-effective and efficient solutions. These low-cost, solution-processed materials exhibit pure-color electroluminescence, making them invaluable for next-generation display and lighting technologies. Despite recent advancements in the efficiency of perovskite-based LEDs, significant challenges remain in addressing the environmental and health concerns associated with the lead content in these materials, as well as their long-term stability under environmental conditions. Leveraging the diverse structures and elemental compositions within the perovskite family, researchers have developed various perovskite forms that aim to mitigate these limitations. This research focuses on the experimental development of lead- free and highly stable MHPs for white LEDs (WLEDs) through the use of low-cost, solution-based methods, including the hot-injection and precipitation techniques. Our work demonstrates that doping CsPbBr₃ with up to 50% Zn(SCN)₂ via the hot-injection method significantly enhances stability and photoluminescence quantum yield (PLQY), although the presence of lead remains a concern due to its toxicity. To address this, we synthesized CsZn(SCN)3 by completely replacing PbBr2 with Zn(SCN)₂, resulting in improved stability, though the PLQY was reduced. In an effort to overcome the limitations of lead-free alternatives, we synthesized double perovskite (DP) microcrystals, including CsAgBiCl₆ doped with Na, Bi, and Y, using precipitation and hydrothermal methods. This DP structure achieved an impressive PLQY of approximately 95%, marking a significant improvement in luminescence. Further exploration led to the synthesis of Cs₂ZrCl₆ using precipitation and hydrothermal techniques, a material with unique properties, including tunable emission wavelengths. The synthesis of these lead-free and environmentally friendly perovskites not only addresses the toxicity issues but also offers stability and tunability, providing a pathway towards more sustainable and efficient lighting technologies, such as LEDs and solar cells. This research opens new avenues for the development of cost-effective, environmentally responsible, and stable perovskite materials for next-generation optoelectronic devices.