Support Vector Regression Analysis Based on the Process Parameter Optimization of SnO2∶Sb Thin Film Deposition
CHEN Yuanhao1,2, XIAO Li1,3, LIANG Changxing1,2, LUO Yueting1,2, GONG Hengxiang1,2
1 Institute of Photovoltaic New Energy Application Technology and Equipment, Chongqing University of Technology, Chongqing 400054, China 2 Physics Experiment Center, Chongqing University of Technology, Chongqing 400054, China 3 Chongqing Key Laboratory of Green Energy Material Technology and Systems, Chongqing University of Technology, Chongqing 400054, China
Abstract: Developing the low-cost transparent conductive oxide film material with high quality is of vital important for the development of modern optoelectronic devices. Upon facing the multi-dimensional thin film growth parameter space, how to reduce the time and material cost effectively in the process of seeking the optimal thin film is an urgent concern for researchers. Based on this, a process parameter optimization approach is proposed focusing on the process of growing SnO2∶Sb thin films on quartz substrates using the mist-CVD method. The mist-CVD system is adopted because its environmentally friendly process to prepare metal oxide of high quality with simple equipment configuration, and shows potential for industrial development. Firstly, A experimental design methods is adopted to design experiments for preparing the transparent conductive thin films considering process parameters of doping concentration, deposition temperature, HCl content and deposition time. Furthermore, the value of figures of merits is introduced to evaluate the transparent conducting property of SnO2∶Sb film. Moreover, to establish the effective prediction model of the transparent conductivity of SnO2∶Sb film under different process parameters, the support vector regression method based on Bayesian optimization is used. The two-dimensional cloud image distribution results show the effective prediction results of the influence of preparation parameters on properties of SnO2∶Sb films, which including 4-dimensional parameter space with a limited combination of 27 sets of experimental results. Finally, the prominent process parameters preparing SnO2∶Sb films with high figures of merits is predicted from the two-dimensional cloud image. The SnO2∶Sb films with high transparency and conductivity can be obtained based on the mist-CVD method adopting the optimized process parameter. The average thickness of the SnO2∶Sb films is about 380 nm, the average visible light transmittance of the sample can reach 86.61%, and the sheet resistance is 21.1 Ω·□-1. This work proposes an effective method for optimizing process parameters during exploring materials or device, which can speed up the development of materials or devices, and save time and materials cost meanwhile.
陈远豪, 肖黎, 梁昌兴, 罗月婷, 龚恒翔. 基于SnO2∶Sb薄膜沉积工艺参数优化的支持向量回归分析[J]. 材料导报, 2023, 37(11): 21120097-6.
CHEN Yuanhao, XIAO Li, LIANG Changxing, LUO Yueting, GONG Hengxiang. Support Vector Regression Analysis Based on the Process Parameter Optimization of SnO2∶Sb Thin Film Deposition. Materials Reports, 2023, 37(11): 21120097-6.
1 Hosono H. Thin Solid Films, 2007, 515, 6000. 2 Park J S, Maeng W J, Kim H S, et al. Thin Solid Films, 2012, 520, 1679. 3 Park J S, Kim H, Kim I D. Journal of Electroceramics, 2013, 32, 117. 4 Yu X, Marks T, Facchetti A. Nature Materials, 2016, 15, 383. 5 Liu X Z, Zhang X, Tao H, et al. Acta Physica Sinica, 2020(69), 248. (in Chinese). 刘贤哲,张旭,陶洪,等. 物理学报, 2020(69), 248. 6 Kawaharamura T. Japanese Journal of Applied Physics, 2014, 53, 05FF08. 7 Ikenoue T, Sakamoto S I, Inui Y. Japanese Journal of Applied Physics, 2014, 53, 05FF06. 8 Sun Z, Oka D, Fukumura T. Chemical Communications, 2020, 56, 9481. 9 Onyema C C. Mist-CVD deposited ZnO thin films for metal-semiconductor field effect transistors. Ph. D. Thesis, University of Canterbury, New Zealand, 2020. 10 Tanaka K, Mao K, Yoshihisa F, et al. Thin Solid Films, 2020, 697, 137820. 11 Cao B, Adutwum L A, Oliynyk A O, et al. ACS Nano, 2018, 12, 7434. 12 Wen Y F, Cai C Z, Pei J F, et al. Acta Physica Sinica, 2009, 58, 15 (in Chinese). 温玉锋, 蔡从中, 裴军芳, 等. 物理学报, 2009, 58, 15. 13 Zhang Y, Xu X. Optik, 2020, 217, 164808. 14 Attia P M, Grover A, Jin N, et al. Nature, 2020, 578, 397. 15 Zheng Y J, Liang X B, Zhang Q, et al. Materials Reports, 2021, 35, 6 (in Chinese). 郑玉杰, 梁鑫斌, 张起, 等. 材料导报, 2021, 35, 6. 16 Balachandran P V, Kowalski B, Sehirlioglu A, et al. Nature Communications, 2018, 9, 1668. 17 Wei L, Xu X, Gurudayal, et al. Chemistry of Materials, 2019, 31, 7340. 18 Haacke G J. Journal of Applied Physics, 1976, 47, 4086. 19 Pan L J, Chen J Q. Experimental design & data processing, Southeast University Press, China, 2008 (in Chinese). 潘丽军, 陈锦权. 试验设计与数据处理, 东南大学出版社, 2008. 20 Scholkopf B, Sung K K, Burges C J C, et al. IEEE Transactions on Signal Processing, 2002, 45, 2758. 21 Buhmann M. Scholarpedia, 2010, 5, 9837. 22 Cortes C, Vapnik V. Machine Learning, 1995, 20, 273. 23 Cui J X, Yang B. Journal of Software, 2018, 29, 3068 (in Chinese). 崔佳旭,杨博. 软件学报, 2018, 29, 3068. 24 Yao P. Desalination, 2011, 267, 170. 25 Gupta S, Yadav B C, Dwivedi P K, et al. Materials Research Bulletin, 2013, 48, 3315. 26 Ramarajan R, Kovendhan M, Thangaraju K, et al. Applied Surface Science, 2019, 487, 1385. 27 Zhang B. Research on prepareation and optical and electrcal properties of SnO2 low emission films. Ph. D. Thesis, Harbin Institute of Technology, China, 2011 (in Chinese). 张波. SnO2低辐射薄膜的制备与光电性能研究. 博士学位论文, 哈尔滨工业大学, 2011. 28 Yasuoka T, Liu L, Ozaki T, et al. AIP Advances, 2021, 11, 45123. 29 Bayram A B, Ko? M, Akyürekli· S, et al. Journal of the Institute of Science and Technology, 2021, 11, 1102. 30 Leidich P, Mondeshki M, Barton B, et al. Zeitschrift Für Naturforschung B, 2018, 73, 865. 31 Ponja S D, Williamson B A D, Sathasivam S, et al. Journal of Materials Chemistry: C, 2018, 6, 7257. 32 Liu L, Kawaharamura T, Dang G T, et al. Japanese Journal of Applied Physics, 2019, 58, 25502. 33 Scott J I, Martinez-Gazoni R F, Allen M W, et al. Journal of Applied Physics, 2019, 126, 135702.
渝公网安备50019002502923号 版权所有：《材料导报》编辑部
2023, Vol. 37 
