超高靈敏度熱感測元件之設計與製造：基於發光電晶體於智慧科技應用

Abstract

本論文全面研究了光發射晶體管 (Light-Emitting Transistors, LETs) 的設計、製造、開發與優化,作為下一代智慧熱感測技術的先進裝置。基於III-V族化合物半導體的LETs作為創新且高速的三端口裝置,集光學與電子功能於單一元件內。透過先進的量子阱 (Quantum-Well, QW) 結構, LETs展現出在高速光通訊、光電整合電路（Optoelectronic Integrated Circuits, OEICs）以及先進熱感測應用方面的卓越潛力,特別是由於其量子阱的熱電子發射特性。通過系統性研究與創新的元件架構,本論文證明了基於LETs的裝置在超高熱敏感度與下一代熱感測解決方案中的非凡潛力,超越了傳統熱感測技術的熱敏感度限制。 本論文首先探討了使用LETs於智慧熱感測技術中的動機,介紹了針對熱感測應用所設計與製造的單量子阱異質結雙極性電晶體(Single Quantum Well-Based Heterojunction Bipolar Transistors, SQW-HBTs)。初步研究強調了在HBTs基區內整合階梯式單量子阱 (SQW) 結構,實現了在25℃到85℃溫度範圍內集電極電流提升72.23%的顯著進步。此提升歸因於增強的熱電子發射動力學,促進了電子從量子阱中的快速逃逸。與傳統HBTs的熱行為相反,該進展得到基於熱電子發射理論的修正電荷控制模型的支持,不僅解釋了觀察到的現象,還促進了QW-HBT結構在熱感測應用中的優化設計。 在這些研究基礎上,研究擴展至多量子阱（MQW）及三量子阱（TQW）HBTs,展示了熱敏感度的進一步提升。一種修正電荷控制模型被開發以考量量子阱參數（如數量與位置）對電流增益的影響。該模型經實驗結果驗證後，用於設計TQW-HBTs,其在相同溫度範圍內實現了集電極電流200%的顯著增加,以及每℃ 7 μA的電流敏感度。這些見解建立了一個優化MQW-HBT及TQW-HBT配置以提升裝置熱性能的穩健框架,用於下一代熱感測器。 儘管MQWs實現了顯著的熱敏感度增強,但由於電子捕獲多重量子阱導致的電流減少,提出了一個突破性的設計,即將光發射晶體管級聯於達靈頓電晶體中。該創新達靈頓電晶體配置利用LET的熱電子發射機制實現了卓越的熱敏感度。LET在操作溫度從25℃上升到85℃ 時集電極電流增加了153%,而達靈頓電晶體在相同偏壓和溫度條件下增加了210%。此外, LET的集電極電流對溫度信號比為8.53 μA/℃, 而在達靈頓配置中此比率提升至26.2 μA/℃, 展示了熱敏感度的顯著改善。此外,輸出電壓的電壓對溫度信號敏感度達到9.12 mV/℃, 超越了傳統熱感測器。 儘管取得了這些進步，本研究探討了解決實現線性電壓對溫度響應的挑戰，並提供了優化QW結構以平衡敏感度與線性的建議。本論文進一步詳細研究了QW寬度對熱性能的影響,突顯了敏感度與線性之間的權衡。實驗與模擬研究揭示,較窄的量子阱顯示出更高的熱敏感度,而較寬的量子阱則確保了更好的線性度。最佳的量子阱寬度為90 Å, 其在100℃ 時實現了每 ℃ 1.34 mA的熱敏感度與優異的線性度。這些發現為開發高性能熱感測器提供了關鍵見解。 最後,本論文強調了包括SQW-HBTs, MQW-HBTs和TQW-HBTs在內的基於LET裝置在智慧熱感測應用中的變革潛力。本研究中報導的新穎配置,例如與光發射晶體管級聯的達靈頓電晶體,展現了超高的熱敏感度。這些進步為下一代智慧熱感測技術奠定了堅實的基礎,提供了超高熱性能的前端元件。

This dissertation presents a comprehensive study on the design, fabrication, and optimization of Light-Emitting Transistors (LETs) as advanced devices for next-generation smart thermal sensing technologies. LETs, based on III-V compound semiconductors, emerge as innovative, high-speed three-port devices integrating optical and electronic functionalities. Leveraging state-of-the-art quantum-well (QW) structures, LETs demonstrate exceptional potential for high-speed optical communication, enhanced performance in optoelectronic integrated circuits (OEICs), and advanced thermal sensing applications, particularly due to their thermionic emission properties. These findings position LETs as strong candidates for next-generation smart thermal sensing technologies, surpassing traditional thermal sensor technologies in thermal sensitivity. The study begins with the design and fabrication of single-quantum-well heterojunction bipolar transistors (SQW-HBTs), designed to enhance thermal sensitivity. The innovative incorporation of a staircase QW into the base region of HBTs achieved a 72.23% increase in collector current across a temperature range of 25℃ to 85℃, attributed to faster electron escape dynamics from the QW. A modified charge-control model incorporating thermionic emission theory effectively explains this behavior and provides a foundation for optimizing SQW-HBT structures for thermal sensing applications. Building on these initial results, the research advances to multi-quantum-well (MQW) and triple-quantum-well (TQW) HBTs to further improve thermal sensitivity. A newly modified charge-control model for MQW-HBTs is developed to account for the influence of quantum-well parameters, such as number and position, on the current gain. This model, validated against experimental results, guides the design of TQW-HBTs, which exhibit a remarkable 200% increase in collector current over the same temperature range and achieving a current sensitivity of 7 μA/℃. These findings highlight the potential of MQW-HBTs such as TQW-HBTs for applications requiring ultra-high thermal sensitivity. To address the challenges of reduced current due to electron trapping in MQWs, the dissertation introduces a groundbreaking idea developed in the optoelectronics thermal technology by successful design and fabrication of the world’s first Darlington transistor configuration cascaded with LETs. This innovative approach combines the thermionic emission properties of LETs with the amplification benefits of the Darlington design, achieving a 153% increase in LET collector current from 25℃ to 85℃ and a further enhancement to 210% in the Darlington configuration. The collector current-to-temperature sensitivity improves from 8.53 μA/℃ in LETs to 26.2 μA/℃ in the Darlington configuration, with a voltage-to-temperature sensitivity reaching 9.12 mV/℃ surpassing conventional thermal sensor. Despite these advancements, challenges in achieving linear voltage-to-temperature responses are addressed, with recommendations for optimizing QW structures to balance sensitivity and linearity. Further investigations into the effect of QW width reveal critical trade-offs between thermal sensitivity and linearity. Narrower QWs exhibit higher sensitivity, while wider QWs improve linearity. An optimal QW width of 90 Å achieves a thermal sensitivity of 1.34 mA/℃ at 100°C while maintaining excellent linearity, providing essential insights for designing high-performance thermal sensors. Finally, this dissertation work highlights the transformative potential of LET-based devices, including SQW-HBTs, MQW-HBTs, and TQW-HBTs, for smart thermal sensing applications. The novel configurations, such as the Darlington transistor cascaded with Light-emitting transistor, reported ultra-high thermal sensitivity in this studied. These advancements lay strong ultra-high-thermal performance front-end components for next-generation for smart thermal sensing technologies.