- 學位論文
正向入射在量子井紅外線偵測器之研究
The study of normal incidence on Quantum Well Infrared Photodetector
摘要
由於次能帶間躍遷的選擇律之限制，量子井紅外線偵測器僅能吸收電場分量垂直於量子井層的入射光，然而在研究量子井紅外線偵測器之正向入射時我們發現邊緣偶合提供了很大部分的響應，來自TM模態Brewster角入射的結果。所以我們設計了一個加大邊緣偶合(深井結構)的元件，成功地增加響應達到2~3倍相對於傳統45度角入射。 我們利用一系列的時間證明可能的光可能進入主動層的位置，理論上長寬極為不對稱的結構，對不同偏振光會有很不同的響應。首先我們設計了一個長條型的獨立平台，來探討不同偏振正向入射光的響應，偏極化的差異都小於2%，同時我們對獨立平台的邊界進行干擾，想了解漏光是否和邊緣有關，實驗結果發現邊緣塗佈光阻者響應大了1~2倍，漏光和邊緣確實有關。此外，我們做了一些實驗證明了光穿越金屬和獨立平台之cross-talk效應不是真的，全部覆蓋金屬的空片響應為0，cross-talk的響應差異不到3%。然後，我們想了解是否有偏斜方向光被獨立平台陣列所偶合，所以我們進行可變角度的量測：在不同的方向角(角度0度到90度)觀測經由結構所散射出來的訊號，我們發現有結構的比沒有結構的響應大了0.2左右，結構確實偶合了偏斜光。最後，金屬和獨立平台邊界的這個階梯也是可能的漏光來源，我們改變金屬與mesa邊界的寬度從40μm到210μm，我們發現到響應隨著寬度增加並最終達到飽和，所以我們認為這個邊界寬度和響應的來源有關。 在本篇論文中，我們成功定性地確認了可能的漏光來源—邊緣及金屬和獨立平台邊界的階梯，這對未來我們想要最佳化正向入射結構來說是重要的開始。
並列摘要
Due to the limit of intersubband transition selection rule, only the light polarized in the growth direction can cause intersubband transition. However, when doing the research of normal incidence, we found that the edge coupling provided a large part response, which came from TM incident light in Brewster’s angle. Therefore, we design a device(deep channel device) which enlarge the edge and successfully increase the responsivity two to three times of the traditional 45 facet device. By doing a series of experiment, we want to prove that the possibilities of the place light goes into the active layer. Theoretically, the structure with extreme asymmetry in length and width has different responsivity for different polarized light. First, we design a strip mesa to explore the responsivity of different polarized light, and we found that the difference of responsivity between different polarization is smaller than 2%.In the meantime, we interfere the edge of mesa by coating photoresistance to understand if there is any relation between the edge and the leakage of light, and the experimental reveals that device coated photoresistance on the edge has better responsivity than the device without coated photoresistance one to two times. The edge is important source of leakage of light. In addition, we have done some experiment, and prove that penetration of metal and cross-talk between mesas is not true. The responsivity of sample covered by metal is zero, and the difference of resposivity of cross-talk is smaller than 3%. And then, we want to understand whether there is light coupled by mesa array. So we conduct “ The measurement of changeable degree” ：observing the scattering signal through the mesa array in different direction(degree 0 to 90),and we found that the responsivity of sample with structure is larger than sample without structure 0.2.The structure indeed can couple oblique light. At least, the step between metal and the margin of mesa also can be the place of leakage of light. We changed the width of the margin of the metal and mesa from 40μm to 210μm, and found that the responsivity increased with the expansion of width and finally reached the saturation. Therefore, in our opinion, the width of the margin of the metal and mesa has relation to the responsivity. In this paper, we have successful qualitative confirmation of the possible sources of the leakage— the edge and the area of the margin of the metal and mesa. It is an important beginning for us to quantatively optimize the structure of normal incidence.