| 摘要: | 在此研究中，採用獨特的脈衝直流濺射多步沉積技術來獲得氮化鋁 (AlN) 薄膜中的殘餘應力及薄膜特徵趨勢。實驗以一步沉積及多步（兩步和四步）沉積製程分別濺鍍60分鐘和30分鐘進行對比。其中透過多步沉積進行比對的實驗總濺鍍時間相同，並且在每次中斷期間關閉電源10分鐘。通過一步沉積和多步沉積薄膜特性的比較，研究薄膜晶體取向、殘餘應力和薄膜特徵之間的相關性。
通過 X光繞射儀 (XRD)、傅立葉變換紅外光譜 (FT-IR)、掃描電子顯微鏡 (SEM) 和原子力顯微鏡 (AFM) 對氮化鋁薄膜的厚度、微觀結構、殘餘應力和晶態進行了檢測。結果表明，氮化鋁薄膜在相同的總濺鍍反應時間，不同的沉積層數下，具有不同的應力特性和微觀結構。此部分可以透過離子轟擊程度的增加和基板感應溫度升高導致對薄膜中不同微觀結構變化進行解釋。
在這項研究中，用於一步沉積的厚度約為 1300奈米的氮化鋁薄膜的均方根 (RMS) 表面粗糙度為 2.67奈米，晶粒尺寸為 65.1奈米，殘餘應力為 1575 MPa。但通過使用四步沉積，這些值降低到 2.46奈米的表面粗糙度、51.4奈米晶粒大小和841 MPa 殘餘應力。在研究中更詳細地描述了多步沉積製程下的更多對應量測值和趨勢。
光學發射光譜 (OES) 數據收集了180至850奈米波長的電漿沉積過程中的光波，並透過預處理篩選特定的主要波長（例如 N2 （315 和 336奈米）和 Al（394 和 396 奈米）），用於大規模數據分析。透過分析大規模 OES 數據，以查看 OES 數據、結晶狀態和薄膜質量之間是否存在聯繫。基於 OES 數據預處理執行的交叉驗證測試，採用主成分分析 (PCA) 和微觀結構特徵值 (VMC) 算法的方法表明，這種多步沉積技術可以從微觀結構分析中提供良好的應力梯度控制並且可以有效地對氮化鋁沉積的薄膜特性進行分類。;In this work, a unique multi-step deposition technique in a pulsed DC sputtering system was employed to obtain trends of residual stress and film characters in aluminum nitride (AlN) thin films. The experiments were carried out by the 60-minute and the 30-minute process in one-step deposition. A multi-step (two and four-step process) deposition was carried out with equally separate processing time of the overall processing time being the same and turn off the power for 10 minutes during each interruption. The comparison of film characteristics between one-step deposition and multi-step deposition was used to study the correlation among film crystal orientation (state), residual stress and film features. The thickness, microstructure, residual stress, and crystalline state of AlN thin films were examined by scanning electron microscopy (SEM), X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FTIR), and atomic force microscope (AFM). The results show that the AlN films have different stress characteristics in different microstructures when processed at the same amount of time in different deposition intervals. The increase in the degree of ion bombardment and induced substrate temperature raised leads to the interpretation for different microstructural changes in films. In this study, AlN films with thickness of about 1300 nm for the one-step deposition have a root mean square (RMS) surface roughness of 2.67 nm, a grain size of 65.1 nm, and residual stress of 1575 MPa. But by utilizing four-step deposition, these values were reduced to 2.46 nm RMS surface roughness, 51.4 nm grain size, and 841 MPa residual stress. More corresponding values and trends about the effect of multi-step deposition are described more details in the research. In-situ optical emission spectroscopy (OES) data were collected during plasma deposition of selected dominant wavelengths such as N2 (315 & 336 nm), and Al (394 & 396 nm) for large-scale data analysis. Residual stress data were analyzed to see if there is a link among OES data, crystalline state, and thin-film quality. Based on the cross-validation test executed from OES data preprocessing, the methodology with the principal component analysis (PCA) and value of microstructure characteristics (VMC) algorithm demonstrated that this multi-step deposition technique can provide a good stress gradient control from microstructural analysis and can be effectively classified for film characteristics of AlN deposition. |
