金/鉍/矽 金屬-半金屬-半導體接面的研究

Abstract

本論文研究金/鉍/矽 金屬-半金屬-半導體接面的電壓電流特性。在實驗上，我們以分子束磊晶法在n型與p型的(111)Si基板成長不同厚度的鉍薄膜，再以電子束蒸鍍法鍍上金薄膜，完成接面並進行電壓電流量測以獲取接面能障。在理論計算上，我們以一維Poisson方程式求解接面的位能分布，獲取能帶結構圖，據以獲得鉍矽接面的能障。並與實驗獲得的能障比較；兩者的結果大致相符。 我們從理論計算發現，鉍矽接面能障會隨著偏壓與鉍薄膜的厚度而改變。此點與金屬半導體蕭基接面具有固定的能障值不同。此可變的能障源自於鉍的低本質載子濃度與量子侷限效應，加上鉍的雙極性使的鉍矽接面類似於np或nn半導體之異質接面特性。此點也獲得了實驗的支持。 我們也在金/鉍/n-矽的電性逆偏量測時，發現了不尋常的指數型上升電流。我們利用前述的能帶結構圖計算獲取不同逆偏壓下的矽鉍接面處電場，以熱游子場穿隧放射模型(TFE model)計算由鉍注入電子進入矽的電壓電流特性。此計算結果與實驗值相當接近，證實了模型的正確性。此項發現顯示了鉍的電子電洞雙極特性，及其在新穎元件應用的潛力。

This thesis studied on Au/Bi/Si, metal/semimetal/semiconductor, contact IV characteristics. In experimental work, Bi thin films with different thickness were deposited on the (111) p- and n-Si substrate by MBE. After the MBE growth, Au dots were deposited on the Bi film with an e-beam evaporator to complete contact structure. In theoretical calculation, we solved 1D Poisson’s equation to find contact potential profile, get the band diagram and determine SBH at the Bi/Si interface. The comparison between theoretical calculation and experimental results showed rough similarity. From theoretical calculation, we found out that Bi/Si contact SBH would change with different bias voltage and bismuth film thickness. This result is far different from well-known fixed SBH value. The variable SBH is due to the low intrinsic carrier density of Bi, the quantum confinement effects and Bismuth bipolar characteristic, making Bi/Si contact behave like np or nn semiconductor heterojunction. And above points meet with experimental support. While measuring Au/Bi/n-Si reverse current, we found an unusual exponential increasing current. Applying aforementioned band diagram calculation, we got contact field values under varying reverse bias and fit the current with TFE model as electrons injecting from Bi to Si. The calculations fit well with experimental results proving the accuracy of applied model. The results of our study indicate bismuth possessing electron-hole bipolar characteristic and show the promising potential in novel device applications.