氮化鎵之緩衝層對其電性與深能階之影響

Abstract

通常在氮化鎵材料(GaN)磊晶的過程中，在磊晶層(epilayer)的成 長之前先在低溫成長一層緩衝層(buffer layer)，可以降低基板與氮化 鎵之晶格常數與熱膨脹係數差異所造成對結晶性的不良影響，進而 改善其磊晶層的電性與光性。本研究針對厚度在20nm工以下的氮化 鎵緩衝層，探討其厚度與磊晶層特性之間的關係。 本研究利用了高解析度穿透式電子顯微鏡(HRTEM)和x-ray 繞 射光譜來觀察試樣的結晶性，並以Photoluminescence(PL)和深能階 暫態響應(DLTS)來研究試樣中的缺陷。電性量測方面，主要以變溫 霍爾量測(TDH)粹取試樣中各種載子濃度與移動率等。我們並引入 了two-channel conduction model以解釋所量測到的TDH數據。參考 DLTS量測所取得的深能階缺陷縱深分佈，我們推測在接近試樣表面 處可能存在一高摻雜層，主要肇因於氮原子的逸失，且此高摻雜層 造成了TDH量測所得數據的偏差。以two-channel conduction model 修正後的TDH數據，結合上述其它量測的結果，我們發現在緩衝層 厚度不大於20nm時，緩衝層厚度的增加有助於降低氮化鎵磊晶層 中的缺陷濃度。

Buffer layer was proven effective to improve the electrical properties of GaN films. In this work, five GaN samples were studied which are grown under same conditions except for the buffer layer thickness. The buffer layer thickness of all these samples are lower than the reported optimal value: 20nm. High-resolution transmission electronic microscopy (HRTEM) and x-ray rocking curve method were carried out to study the crystalline. Photoluminescence (PL) and deep-level transient spectroscopy (DLTS) were utilized to investigate their optical properties and deep levels. Temperature- dependent Hall measurement (TDH) is also applied to extract more information of their electrical properties. A two-channel conduction model is introduced to fit the TDH data and to explain the contradictions between measurement results. Referring to the depth profile of deep trap levels obtained by DLTS, it is found that a high dopant layer is probably the reason of low measured mobility. By combining all measurement results and corrected TDH data, we found a thicker buffer layer can reduce the residual defect concentration in GaN epilayer.