深紫外光發光二極體之結構設計與磊晶之研究

Abstract

近年來，深紫外光發光二極體越來越受到人們的重視，因為其小體積、節能、 瞬間開關、對環境友善與客製化空間大等優點，深紫外光發光二極體被視為取代 汞燈的下一個深紫外光燈源。深紫外光發光二極體可以改善生活中的許多產品， 它可以應用在水的消毒、消毒隱形眼鏡盒與消毒冰箱等等。許多需要消毒功能的 產品都有深紫外光發光二極體的需求。 於第一部分的研究中，我們展示如何用濺鍍與退火的方式，做出高穿透且高 晶格品質的氮化鋁模板， XRD 的(002)與(102)半高寬為 100/290 角秒，而傳透率 與吸收率分別為 77.97%與 5.38%的良好表現。在這個部分中會討論線差排密度下 降的機制和比較此模板與傳統兩段式 MOCVD 模板的吸收。 另外， 我們藉由成長超 晶格過渡層於模板上改善 n-AlGaN 之表面形貌，超晶格過渡層可以藉由先釋放應力並且 降低模板中的線差排密度使 n-AlGaN 上的 pits 密度降低 95%。 第二部分的研究中，我們藉由模擬不同的電子阻擋層結構，得到電洞注入佳、溢流 問題小且內部量子效應高的鋁含量漸變升高電子阻擋層。在這個部分中也會展示深紫外 光發光二極體的電致發光結果。

Nowadays, deep-ultraviolet light-emitting diodes have attracted much attention. DUV LEDs have due to the small size, instant on/off, and environmental-free. Therefore, DUV LEDs have been expected for replacement of low-pressure mercury lamp. DUV LEDs could improve life by the application in sterilization, biochemistry, water purification and air purification. We developed a high quality and highly-transparent annealing sputter AlN template. The FWHM of (002) plane and (102) plane x-ray rocking curve was 100 arcsec and 290 arcsec, respectively. The transmittance and absorption were 77.97% and 5.38%, respectively. The comparison between template using annealed sputter AlN buffer and conventional MOCVD template was introduced. We also improved the morphology of n-AlGaN by inserting multi-stack superlattice transition layers. The transition layers would pre-relax the template and reduce the TDD of the template. As a result, the pits density of n-AlGaN was reduced 95%. In order to improve the hole injection, we used the one-dimensional drift-diffusion charge control solver (DDCC) to design the electron-blocking layer (EBL) structure. We concluded that a compositionally grade-up EBL LED possessed a better hole injection and less loss overflow than those of other designs. As a result, the IQE of LED with a grade-up EBL is also the highest among all. Base on the simulation results, we fabricated a DUV LED epilayer structure with an electroluminescence peak at 291nm. The defect band was alleviated by growth conditions optimization.