選擇性磊晶應用於介面蝕刻的氮化鎵薄膜和半極性平面的製作分析

Abstract

摘要 氮化鎵為非常熱門的光電材料，尤其近年來白光發光二極體(light emitting diode)發展迅速，而氮化鎵為其主要的材料，在做成太陽能電池和電晶體也都有很大的潛力，然而我們目前得到的氮化鎵薄膜，全部都是異質介面磊晶，基板為藍寶石或者是碳化矽。但磊晶上都有晶格不相符和熱膨脹係數不同的問題，為了改善穿晶缺陷(dislocation)的問題，在製程上，我們做基板製作，用側相磊晶的方法，降低穿晶缺陷，得到品質較好的氮化鎵。而另一方面，也有人考慮到QCSE的影響，從原本的C-plane上磊晶，改為半極性或者是非極性平面上磊晶，藉此增加內部量子效應(IQE)，但此時穿晶缺陷又會增加，所以這當中的取捨是一個很大的問題；本論文成功的做出Free standing GaN thin film也成長了半極性和非極性的平面，並對其作材料性的分析。分析的結果可以發現我們可以長出[11-22]和[10-11]的平面，這結果和德國Ulm大學光電所在2011年度的研究報告成果非常相似，我們用有機金屬化學氣相沉積(MOCVD)做了高溫和低溫的磊晶，結果和德國Ulm有一樣的結果,接著我們同時做了disk和hole的基板,，在D-180有機金屬化學氣相沉積基台(MOCVD)內，相同的參數下，同時成長；結果發現hole的基板成長出和m-plane相似的半極性平面，而disk的基板則同時成長出類似a-plane和m-plane的半極性平面，分別是[1 1 -22]和[10-11 ]的平面，即使我們採用了比以往還大尺寸的hole尺寸的基板，它還是長出了相似於m-plane的semi-polar plane[10-11]的平面，我們最大hole的尺寸為直徑20μm，在之前我們曾做過3μm的hole成長pyramid， 結果也是以m-plane 為主的半極性平面，而在德國Ulm在高溫成長時，會形成類似a-plane的半極性平面，但是在我們的實驗上，高溫成長氮化鎵時，仍然會以m-plane為主的半極性平面[10-11]，另外我們也試著做更大的disk基板，以高溫或者是低溫成長氮化鎵，此時得到的結果，仍然是12面體的兩種不同的半極性平面，另外我們也做了蝕刻後的表面分析，結果在短時間用氫氧化鉀(KOH)蝕刻後，會出現(0002)的C平面和產生些奈米線，蝕刻更久的時間後，會發現所有的半極性平面都消失，只剩下m-plane。

Abstract GaN is a very famous photonic material, especially, the white light LED for illumination for the past few year. GaN is the main material for white lighting. On the other side, GaN has a big potential for solar cell and transistor applications. However, the GaN thin film we can do are all heteroepitaxy, the substrate is most using sapphire and SiC. But both of them have the same problem in lattice mismatch and thermal expansion different. In order to improve the dislocation problem, we do substrate fabrication, for example, epitaxy lateral overgrowth. The ELOG can reduce dislocation density. Another aspect, someone considers the QCSE (quantum confined stark effects), they change epitaxy in C-plane (0001) to semi-polar plane or non-polar plane, it can enhance IQE (internal quantum efficiency), however, the dislocation density will increase. How to get two problems be the balance is a big issue. Our study successfully makes the free standing GaN thin film and we also grow semi-polar and non-polar plane and do some material property analysis. According to the result of analysis, we can grow [11-22] and [10-11] planes. The achievement is very similar to the annual report 2011, institute of Optoelectronics, Ulm University. We use MOCVD to grow in two different recipe, one is higher temperature and another is lower temperature. The result is the same with Ulm University. And then we both grow the disk and hole substrates in the same run by using D-180 MOCVD. We find that hole substrate grows analogous m-plane [10-1 1] plane and the disk substrate grow two different planes, they are similar to a-plane and m-plane semi-polar planes, respectively, [11-22] and [10-11] planes. Even we use bigger hole size substrate than before we use for growing. It still grows analogous m-plane semi-polar plane [10-11] and our hole substrate, the biggest hole diameter is 20μm. Before we grown on 3μm hole substrate for the pyramid and the result is that the pyramid is analogous to m-plane semi-polar plane. In the Ulm report, they mention about that it will form similar a-plane’s semi-polar plane. But in our experiment, we find that it is still similar m-plane’s semi-polar plane [10-11], on the other hand, we try to do bigger disk substrate, we grow them in higher and lower temperature. The result we get is still twelve facet semi-polar planes. We also do KOH etching test for our inverse pyramid. In a short period time, the inverse pyramid will show (0002) c-plane and produce some nano wire and when we etch for longer time, we find that all semi-polar planes disappear, the rest plane is m-plane only.