太陽能為極具發展潛力的替代性能源，目前使用太陽能發電的成本仍高於大多數國家市電與大部分的其他能源。因此提高太陽能電池轉換效率與降低生產成本，為此產業的競爭要素。

本研究針對提高轉換效率當中的“減少表面光反射所造成的能量損失”作為研究主題，提出一最佳化製程模式的方法論、減少實驗的次數與成本、縮短製程參數最佳化的時間。

經由矽晶太陽能電池的鹼蝕刻，與抗反射製程中的氟化鎂(MgF2)、氮化矽(SiNx)鍍膜的實驗結果，發現與驗證類神經網路具備預測與最佳化太陽能電池製程模式的功能，可幫助找到更佳的參數組合，顯著提升製程的品質指標達12 % 以上。使得鹼蝕刻在400~1000 nm波長的平均反射率，使用積分球量測可達15.50 %，最低反射率為13.69 %；而氮化矽在400~1000 nm平均反射率可達8.99 %。

Solar energy is a kind of alternative energy and possesses the potential of development. The generation cost of electricity using solar energy is still higher than most of the other energy sources. Therefore, improving the conversion efficiency of solar cells and reducing the production cost have become the key elements for the competition of this industry.

This thesis aimed to reduce the surface reflection of light energy loss to improve the conversion efficiency of solar cells. A methodology of the process optimization is proposed in this thesis to reduce the number and cost of tests and shorten the time of process parameter optimization. The experimental results of the alkaline etching and the anti-reflection coating process using magnesium fluoride (MgF2), silicon nitride (SiNx) on silicon solar cells demonstrate that of, the artificial neural networks possesses the ability of predicting and optimizing the process model features of solar cells. The results will help engineers to find a better combination of parameters and significantly improve process quality indicators for more than 12% to enhance the competitiveness of solar cells manufacturers. The experimental results include the average reflectance between 400~1000 nm of alkaline etch is 15.50 %, and the 8.99 % average reflectance between 400~1000 nm of silicon nitride can be achieved.

[1] W. Shockley, H. J. Queisser, “Detailed balance limit on efficiency of p-n junction solar cells,” Journal of Applied Physics, vol. 32, no. 3, pp. 510–519, Mar. 1961.

[2] P. Campbell, and M. A. Green, “The limiting efficiency of the silicon solar cells under concentrated sunlight,” IEEE Transactions on Electron Devices, vol. ED-33, no. 2, pp. 234-239, Feb 1986.

[3] C. T. Sah, “Reduction of Solar Cell Efficiency by Edge Defects Across the Back Surface Field Junction,” Solid State Electronics, vol. 25, no. 9, pp. 851-858, Sep. 1982.

[4] M. A. Green, “Limits on the open-circuit voltage and efficiency of silicon solar cells imposed by intrinsic Auger process,” IEEE Transactions on Electron Devices, vol. ED-31, no. 5, pp. 671–678, May 1984.

[5] 羅華強，「類神經網路 :MATLAB的應用」，高立出版社，2005。

[6] 葉怡成，「類神經網路模式應用與實作」，儒林圖書有限公司，2006。

[7] 張文山，「利用類神經網路進行電漿化學氣相沈積製程特性預測之研
究」，國立成功大學碩士在職專班論文，2007。

[8] S.-S. Han and G. S. May,“Using Neural Network Process Models to Perform PECVD Silicon Dioxide Recipe Synthesis Via Genetic Algorithms, ” IEEE Transactions on Semiconductor Manufacturing, vol. 10, issue 2, pp.279-287, May 1997.

[9] L.C.-K. Liau, C.-J. Huang, C.-C. Chen, C.-S. Huang, C.-T. Chen, S.-C. Lin, and L.-C. Kuo,“Process modeling and optimization of PECVD silicon nitride coated on silicon solar cell using neural networks,” Solar Energy Materials & Solar Cells, vol. 71, pp. 169-179, 2002.

[10] W.-C., Chen, H.-I., Lee, W.-J., Deng, and K.-Y., Liu, “The implementation of neural network for semiconductor PECVD process,” Expert Systems with Applications, vol.32, issue 4, pp. 1148-1153, May
2007.

[11] E-Ton Solar Technology.
http://www.e-tonsolar.com/Technology.asp?le=tchinese&fid=64]

[12] Gintech Energy Corporation.
http://www.gintechenergy.com/about_process.php]

[13] T. A. Kovacs, N. I. Maluf, and K. E. Petersen, “Bulk Micromachining of Silicon,” Proceedings of the IEEE, vol. 86, no. 8, pp.1536-1551, August 1998.

[14] M. Elwenspoek, “The form of etch rate minima in wet chemical anisotropic etching of silicon,” Journal of Micromechanical and Microengineering, vol. 6, pp. 405-409, 1996.

[15] 王宗新，「金字塔抗反射結構之製作及其單晶矽太陽能電池之應
用」，國立中山大學光電工程研究所碩士論文，2007。

[16] Y.-D. Ko, P. Moon, C.- E. Kim, M.-H. Ham, J.-M. Myoung, and I. Yun, “Modeling and optimization of the growth rate for ZnO thin films using neural networks and genetic algorithms,” Expert Systems with Applications, vol. 36, pp. 4061-4066, 2009.

[17] J. Yoo, S.K. Dhungel, and J. Yi, “Properties of plasma enhanced chemical vapor deposited silicon nitride for the application in multicrystalline silicon solar cells,” Thin Solid Films, vol. 515, pp. 5000–5003, 2007.

[18] A. J. M. van Erven, R. C. M. Bosch and M. D. Bijker, “Textured Silicon Surface Passivation by High-Rate Expanding Thermal Plasma Deposited SiN and Thermal SiO2/SiN Stacks for Crystalline Silicon Solar Cells,” Progress in Photovoltaics: Research and Applications, vol. 16, issue 7, pp. 615–627, Aug. 2008.

[19] U. Gangopadhyaya, K. Kimb, S.K. Dhungelb, P.K. Basub, and J. Yib, “Low-cost texturization of large-area crystalline silicon solar cells using hydrazine mono-hydrate for industrial use,” Renewable Energy, vol. 31, pp. 1906–1915, 2006.

[20] M. M. Hilali, K. Nakayashiki, A. Ebong, and A. Rohatgi, “High Efficiency (19%) Screen-printed Textured Cells on Low-resistivity Float-zone Silicon with high Sheet-resistance,” Progress in Photovoltaics: Research and Applications.vol. 14, pp. 135–144, 2006.