發展太陽能應用可以減少石化能源的使用及二氧化碳排放，對永續發展與保護生態環境有顯著的貢獻。建立穩定且快速製造品質優良太陽能板的技術將對此領域的發展有所助益，其中，有機金屬氣相沉積(MOCVD)製程技術可有效用於生長大面積與高品質的薄膜太陽能電池。本研究目標為利用電腦輔助軟體將太陽能電池薄膜之MOCVD磊晶過程，透過有限元素法計算低壓MOCVD反應腔體內部零件之溫度分佈，並藉由相關零件量測點的溫度量測數值與模擬結果進行比對，以驗證有限元素模型的有效性。 本研究利用有限元素法模擬一款商用低壓MOCVD反應腔體，從加熱絲功率設定，以及考慮加熱板、托盤、玻璃板熱傳條件，計算各元件的溫度分佈，並利用實機測試量測之溫度結果來驗證模型的有效性，在模擬過程中逐一檢查與修正邊界條件與零件中熱傳導條件。經比對計算與實驗結果，發現在達到相似的加熱板熱電耦溫度設定時，其玻璃板熱電偶溫度也十分接近，驗證了本研究所建立計算模型的有效性。在加熱板各量測點的溫度比較方面，模擬結果與實機測試量測值的差異皆小於3.3%。而在預設玻璃板目標溫度為190°C的操作條件下，實機測試各量測點的溫差值為3.7°C (1.9%)，而模擬結果的溫差值為8°C (4.2%)。此比較結果顯示模擬與實機測試量測的溫度分佈結果具有可接受的一致性，確認了本研究所建立有限元素模型的有效性。因此，借助電腦輔助計算，可以花費較少時間及較低的成本獲得不同MOCVD反應腔體操作條件的溫度分佈結果，有效率地求取最佳操作條件。本研究還探討加改良熱板模組化設計的可能性，模擬結果顯示，改良設計配合均勻的功率輸入可降低玻璃板的溫差，提升製程工作溫度的均溫性。 ;Solar energy not only reduces the use of petrochemical energy and emissions of carbon dioxide, but also contributes to the ecological environment and sustainable development. Stability and mass production are the most important consideration for producing high-quality thin-film solar panels. Metal organic chemical vapor deposition (MOCVD) is widely used in depositing large-area and high-quality thin-film solar cells. The aim of this study is using finite element method (FEM) to develop a computer aided engineering (CAE) technique for simulating and analyzing temperature distributions in a low-pressure MOCVD reactor for epitaxial growth of thin films used in solar cells. Temperature at selected positions in certain components is experimentally measured to validate the FEM modeling. An FEM model is constructed based on the design of a commercial low-pressure MOCVD reactor. Simplified components and proper assumptions are applied in the FEM modeling for saving computational time. For practical purpose, the heat transfer settings and thermal boundary conditions in the FEM model are assumed in a way similar to those employed in operation of the given MOCVD reactor. Temperature distributions are calculated for all components and those in heating plates and glass panels are analyzed and compared with the experimental measurements. The difference in the temperature of each measured point on heating plates between experiment and simulation is less than 3.3%. With regard to the temperature uniformity on the top surface of glass panels, the temperature difference measured in experiment is 3.7°C (1.9%) while it is 8°C (4.2%) in the FEM simulation with consideration of a gap between tray and glass panels, for a target temperature of 190°C. The difference between simulation and experiment is attributed to the simplification of transportation components and relevant boundary conditions. Effectiveness of the constructed FEM model is validated by the agreement of overall temperature distributions between simulations and experimental measurements. Therefore, the CAE developed in this study is applicable for effectively designing an MOCVD reactor or optimizing the operational settings for a given MOCVD process. A modified design of heating plates is also considered in this study. Simulation results show that the modified design of heating plates can reduce the temperature difference on glass panels and improve the temperature uniformity.
