透過您的圖書館登入
IP:3.147.66.178
  • 學位論文

鋁誘發固相磊晶技術異質磊晶矽鍺薄膜於矽基板上之研究

Hetero-epitaxial growth Si1-xGex film via a low temperature aluminum-induced solid phase epitaxy (AI-SPE) process

指導教授 : 陳福榮

摘要


減少太陽能元件製造所需的成本,將使聚光型太陽能電池在市場上更有競爭能力。其中一種方式就是利用單晶矽基板取代單晶鍺或砷化鎵(GaAs)基板,在矽基板上透過異質磊晶的方式成長不同鍺濃度之矽鍺磊晶層,最後單晶矽基板的表面具有鍺的晶格特性,因此可以將三五族太陽能電池直接生長在單晶矽基板上,大幅降低電池成本。本研究中,將利用鋁誘發固相磊晶技術於低溫環境下成長不同鍺濃度之矽鍺磊晶層。在以往的研究中,鋁誘發結晶多晶矽或多晶鍺的機制已被完善的探討;但是在鋁誘發固相磊晶的系統中仍是一知半解,在本研究中,我們將探討 (1) 非晶鍺預摻雜位置;(2) 非晶鍺預摻雜濃度; (3) 反應溫度對鋁誘發固相磊晶系統的影響。研究結果表明,矽鍺比例在鋁誘發固相磊晶反應後是可以透過預摻雜非晶鍺的濃度來控制,且反應溫度應高於400˚C。 基於上述研究結果,再透過穿透式電子顯微鏡臨場加熱系統觀察鋁誘發固相磊晶的整體反應過程及熱力學計算及分析,可以將鋁誘發固相磊晶的反應機制完整建立: (1) 非晶矽鍺受到鋁外層自由電子的影響,使共價鍵的鍵能弱化或斷鍵,形成自由原子;(2) 因擴散驅動力的驅使,自由原子會透過鋁的晶界擴散到鋁和單晶矽基板的界面,並以非晶型態穩定在此並累積(擴散);(3) 當非晶型態矽鍺累積厚度超過臨界厚度時,造成系統不穩定,為使系統趨於穩定,非晶矽鍺會在鋁和單晶矽基板界面產生一個新的成核點(結晶型態),該成核點的晶相會受到單晶矽基板表面原子排列的影響,沿著單晶矽基板有序排列;(4) 非晶矽鍺會透過鋁的晶界繼續擴散,供應成核點持續垂直成長及側向生長(長晶),由於要釋放應力的因素,鋁會往原本非晶矽鍺的位置移動;(5) 最後形成連續的矽鍺磊晶薄膜。基於上述研究之結果,最後我們利用多道鋁誘發固相磊晶製程成功的製備鍺虛擬基板,它可以提供給聚光型太陽能電池或積體電路等作為低成本基板。

並列摘要


Reducing the solar cell manufacture cost makes the concentrator photovoltaic (CPV) has more competitive in the market. One of the potential method is to use a single crystal silicon substrate (sc-Si) substituted germanium or gallium arsenide (GaAs) substrate. Growth of Si1-xGex epitaxial layer on single crystal Si substrate with different Ge concentration makes the single crystal Si substrate have Ge properties including lattice constant and so on, which maybe significant reduce the CPV manufacture cost. In this study, we will use the aluminum-induced solid phase epitaxy (AI-SPE) process to fabricate Si1-xGex epitaxial layer with different Ge concentration under low-temperature. In previous study, the mechanism of aluminum-induced crystallization (AIC), it used to grow poly-Si or poly-Ge, has well studied and build up. However, researchers have a superficial knowledge of aluminum-induced solid phase epitaxial. We believe that the mechanism of AI-SPE should be build-up if we would like to control the reaction process. In this study, we will firstly discuss the effect of (1) pre-doping Ge position, (2) pre-doping Ge concentration, and (3) reaction temperature during AI-SPE process. According to the results, the Ge concentration indeed can be well controlled via pre-doping Ge technique after AI-SPE process, and the optimal reaction temperature should higher that 400˚C. Moreover, we will use in-situ heating transmission electron microscopy to observer the AI-SPE reaction process, and the analysis of the thermodynamics exactly supports the finding from in situ TEM. Based on these results, the mechanism of AI-SPE can be concluded into five steps: (1) The covalent bond of a-Si1-xGex will be weakened and formed "free atoms" by electrons that it is surrounding the surface of aluminum (Screening effect); (2) The free atoms driven by the diffusion driving force, it will diffuse through aluminum grain boundaries to the interface between aluminum and single crystal Si substrate, and the free atoms will thermodynamically stable accumulate at interface until its thickness reaches the critical thickness (Diffusion); (3) As the accumulated thickness reaching the critical thickness, the system will become unstable. The a-Si1-xGex will generate a new crystalline phase to reduce the free energy making the system become stable, that is, Si1-xGex nuclei. the crystal orientation of Si1-xGx nuclei will affect by single crystal Si substrate and hetero-epitaxial grow on it (Nucleation); (4) The free atoms will continuously diffuse to the interface between aluminum and single crystal Si substrate, and supply to the nucleus for vertical and lateral growth. Simultaneously, the stress existing in the aluminum film generated during free atoms diffusion will release, thus, the aluminum will move to the position of a-Si1-xGex (Grain growth); (5) Finally, to form a continuous silicon-germanium epitaxial layer on the single crystal Si substrate. Based on the above study results, we finally successful prepared a germanium virtual substrate via multi-run aluminum-induced solid-phase epitaxy process, which can provide to CPV or integrated circuit as a low-cost substrate or template.

參考文獻


130. 魏松煙, 鋁誘發結晶製備結晶/磊晶薄膜應用於太陽能電池之研究, 工程與系統科學系. 2013, 國立清華大學: 新竹市. p. 140.
5. Weber, E.R., THE PRESENT STATUS AND FUTURE PERSPECTIVES OF SOLAR ENERGY TECHNOLOGIES. 2015.
6. Lewis, N.S., Basic Research Needs for Solar Energy Utilization. 2005.
7. Martí, A. and G.L. Araújo, Limiting efficiencies for photovoltaic energy conversion in multigap systems. Solar Energy Materials and Solar Cells, 1996. 43(2): p. 203-222.
8. Yu, W.J., et al., Semitransparent Polymer Solar Cells with 5% Power Conversion Efficiency Using Photonic Crystal Reflector. Acs Applied Materials & Interfaces, 2014. 6(1): p. 599-605.

延伸閱讀