| 摘要註: |
由於c平面氧化鋅會有自發性極化的特性，造成量子侷限史塔克效應，與發光效率下降，使用非極性(11-20) a平面氧化鋅來避免此問題。我們利用低壓化學汽相沉積法(LPCVD)，成長在不同成長時間(厚度)下的非極性(11-20) a平面氧化鋅薄膜，我們進行掃描電子顯微鏡(SEM)、陰極發光(CL)、原子力顯微鏡(AFM)、光激螢光(PL)以及拉曼(Raman)等實驗。成長時間最長的樣品，其陰極發光(CL)強度是所有樣品當中最強的，反之成長時間最短的樣品為最弱。進一步由陰極發光(CL)波峰位置顯示，成長時間最短的樣品其激發波長最短。隨著成長時間越長，沿著x軸的晶格拉伸變小與沿著y軸的壓縮應力變小，其極化偏振度(DOP)越大，這樣的趨勢歸因於磊晶層內部應力的分佈。此外由光激螢光(PL)光譜量測到兩個主要的波峰分別分佈在371奈米和450奈米，371奈米和450奈米的激發波峰分別來自自由電子至束縛電洞之間的躍遷和晶格缺陷，然而隨著成長時間越長光激螢光(PL)強度積分面積越大。再者，由理論模擬結果顯示，除E2和E3 躍遷以外，對於E1和所有躍遷的極化偏振度與我們實驗結果相符合。透明導電薄膜(TCO)使得薄膜太陽能電池能夠截取電流，並且使光線能夠到達吸收層。高透明度且高導電性的氧化鋅使其成為極佳的透明導電薄膜(TCO)材料。除此之外，產量豐富且便宜的氧化鋅加上無毒的特性是其最大的優點。我們利用低壓化學汽相沉積法(LPCVD)，成長在不同成長時間(厚度)下的極性(0001) a平面氧化鋅薄膜，我們進行掃描電子顯微鏡(SEM)、陰極發光(CL)、光吸收度、原子力顯微鏡(AFM)、光激螢光(PL)以及拉曼(Raman)等實驗。隨著薄膜厚度增加得到更長的陰極發光(CL)激發波長、更大的光吸收度及更大的表面粗糙度，此趨勢表示成長時間最長樣品是製作透明導電薄膜(TCO)的最佳選擇。隨著成長時間從20分鐘到60分鐘，發現光激螢光(PL)激發波長峰值變得更長，然而成長時間直到70分鐘激發波長峰值開始變短，這是因為晶格應力在成長時間70分鐘時開始鬆弛。隨著薄膜厚度增加到1022奈米，晶格拉伸應力變大，然而當厚度增加到1388奈米則開始變小，這是因為累積在晶格內的熱能使得晶格應力產生鬆弛現象。 Because spontaneous polarization of polar c-plane ZnO causes Quantum Confined StarkEffect (QCSE) and decreases the light efficiency, nonpolar a-plane ZnO could solve theproblem. We have shown the experimental results of SEM, CL, AFM, PL, and Ramanmeasurements of the (11-20) a-plane ZnO sample with different growth time (thickness),grown by low pressure chemical vapor deposition (LPCVD). The CL intensity of the longestgrowth time sample is the strongest among three samples, while that of the shortest growthtime sample is the weakest. The CL peak position of emission wavelength for the shortestgrowth time sample is the shortest. As the growth time increases, the degree of polarization(DOP) increases. As the in-plane tensile strain Δε xx and compressive strain Δε zzdecreases, the DOP increases. This trend should be related to the strain distribution inside theepilayer. In addition, there are one main emission peak around ~371 nm (3.342 eV) and oneemission around ~450 nm (2.756 eV) in the PL spectrum. The emission peak around ~371nm (3.342 eV) is due to free-electron-to-bound-hole (e-A0) transition of ZnO at 20K, whilethe emission peak around ~450 nm (2.756 eV) is related to a deep level or trap state in theZnO, such as interstitial zinc and oxygen vacancies that act as donors at energy levels locatedbelow the conduction band edge. As the growth time increases, the integral PL intensitybecomes larger. Furthermore, our simulation results show that except for the E2 and E3transition, as the anisotropic in-plane strains are relaxed for ε xx ≥ 0 and ε zz ≤ 0 , thedegree of polarization for E1 and all transitions increases. The simulation result is consistentwith our experiment results.Thin film solar cells require a transparent conductive oxide (TCO) to extract theelectrical current and allow the light to reach the absorber layers. ZnO is a good TCOcandidate as it can be both highly transparent and highly conductive; ZnO is an abundant, lowcost, and non-toxic material. We have shown the experimental results of SEM, CL,UV-Visible absorption spectra, AFM, PL, and Raman measurements of the (0001) c-planeZnO sample with different growth time (thickness), grown by low pressure chemical vapordeposition (LPCVD). As the thickness increases, a longer CL emission wavelength, largerabsorbance, and larger surface roughness are observed. This implies the longest growthtime sample is the best TCO candidate. As the growth time increases from 20 to 60 minutes,the PL emission wavelength becomes longer. As the growth time increases to 70 minutes, theemission wavelength becomes shorter. It is related to lattice strain relaxation. As the thicknessincreases to 1022 nm, the in-plane tensile strain becomes larger. As the thickness furtherincreases up to 1388 nm, the in-plane tensile strain becomes smaller, it is related to therelaxation of thermal strain. |