電漿輔助化學氣相沉積法(Plasma Enhanced Chemical Vapor Deposition–PECVD)已被用於直接生產晶片同品質的矽薄膜元件，並利用三氯矽甲烷(Trichlorosilane–SiHCl3)為製程氣體源將製作成本降低。然而，氫氯化非晶矽薄膜為主要具有矽氫鍵，其結構相對於結晶矽鬆散，且矽氫鍵容易在照光後結斷裂形成缺陷而發生光衰效應，為提高薄膜穩定度，故研究目標為將薄膜由非晶結構轉至微晶或奈米晶結構，藉由調整 SiHCl3 與 H2 流量比例，直接沉積奈米晶矽結構薄膜，以提升矽薄膜光電元件之效率與穩定度。
本研究採用PECVD以三氯矽甲烷氣體源在2項規劃的製程條件(射頻功率及氣體流量)和350℃的矽基板溫度下沉積氫氯化矽薄膜，並同時利用光發射光譜(Optical Emission Spectra–OES)監測紀錄沉積製程電漿的差異。接著使用傅里葉變換紅外光譜、拉曼光譜、穿透式電子顯微鏡、X–射線繞射分析和表面輪廓測量儀進行量測不同製程條件所沉積的氫氯化矽膜藉以分析薄膜的多種性質(包括有化學鍵結、結構(結晶狀態)和薄膜厚度(鍍率)等…)來確認其屬性–氫氯化非晶矽(a–Si：H/Cl)和氫氯化奈米晶矽(nc–Si：H/Cl)。
所以為節省薄膜性質分析時所耗費的龐大成本與時間，透過機器學習技術可以解決複雜的問題，因而本研究結合主成分分析(PCA)與支持向量機(SVM)之機器學習技術，處理電漿製程之OES大規模數據，並建立自動化薄膜結晶結構分類與預測系統，可判斷不同製程所沉積出來的薄膜結構(氫氯化奈米矽晶、氫氯化非晶矽薄膜)。首先，處理電漿製程之OES大規模數據，應縮減複雜度高的OES全譜數據，選擇主要前驅物的自由基(SiCl2 *，SiCl3 *，Hα和Hβ)，並通過提出的PC1–DEV算法，建立結晶相值(Value of crystalline phase – VCP)，以表徵結晶結構的趨勢變化。 本研究獲得高於0.06的VCP可以歸類為氫氯化奈米晶矽薄膜，在功率為250W流量為70sccm處有最大VCP值為0.23，其餘奈米晶矽薄膜的VCP平均值為0.11，另外低於0.06的VCP值將分類於非晶矽薄膜，然後將支持向量機法引入到對氫氯化矽薄膜處理的分類中，藉由使用三種不同的函數法線性函數核(Linear kernel)、多項式函數核(Polynomial kernel)和徑向基(高斯)函數核(Radial basis function kernel-RBF）演算法，並選出最佳訓練函數(徑向基函數核)，可建立出具有準確度為98%的高智慧判斷氫氯化矽晶薄膜結構表徵工具。

Plasma Enhanced Chemical Vapor Deposition (PECVD) has been used to improve the efficiency and stability of tantalum film optoelectronic components, including the nc–Si:H deposition film. In this paper, the nanocrystalline silicon thin films were deposited on Si substrate by PECVD from source gas of trichlorosilane (TCS, SiHCl3) at temperatures 350°C. The in–situ plasma monitoring and the resultant deposited film properties of Hydrogen chloride silicon thin film were characterized by Optical emission spectroscopy (OES), Fourier transfer infrared spectroscopy (FTIR), Raman spectroscopy (RS), X-ray Diffraction(XRD), Transmission electron microscope(TEM) and Alpha–Step profiler. In addition, principal component analysis (PCA) based on large scale OES dataset was performed and through the proposed PC1–DEV algorithm, the high–dimensional OES data of complexity should be selected and reduced to radicals of interest (SiCl2*, SiCl3*, Hα and Hβ). The value of crystalline phase (VCP) was established to differentially characterize the nanocrystalline phase as mean VCP of 0.11 and the control limits of 0.06, which can be used as the in–situ monitoring tool for crystalline phase characterization. And demonstrates use the large plasma data for PCA analysis connect with SVM algorithm method for the screening and grouping of nanocrystalline and amorphous OES spectra data, and make a decision with strong classifier performance. The support vector machine(SVM) method can classification of Hydrogen chloride silicon thin film, and using three different kernel function methods, include linear kernel, the polynomial kernel and the radial basis (Gaussian) function kernel. In this study the radial basis function kernel algorithm is the best training function used to be judgment and radial basis function kernel is selected to learn a high-smart judgment of the structure model of hydrogen chloride crystal film with an accuracy of 98%.

[1]. 郭文傑、李祺菁、董鍾明，”兩岸電子材料市場及技術競合分析”，工研院IEK化材組，2015
[2]. Liu, H., Du, Y., Deng, Y. & Ye, P. D. Semiconducting black phosphorus: synthesis, transport properties and electronic applications. Chem. Soc. Rev. 44, 2732–2743 (2015).
[3]. Lin, I.-K., Bai, H. & Wu, B.-J. Analysis of Relationship between Inorganic Gases and Fine Particles in Cleanroom Environment. Aerosol Air Qual. Res. 10, 245–254 (2010).
[4]. Leland Chang et al. Extremely scaled silicon nano-CMOS devices. Proc. IEEE 9, 1860–1873 (2003).
[5]. Richter, A. et al. Versatility of doped nanocrystalline silicon oxide for applications in silicon thin-film and heterojunction solar cells. Solar Energy Materials and Solar Cells 174, 196–201 (2018).
[6]. Han, C.-W. et al. Hydrogenated Microcrystalline Silicon Film Growth by Inductively Coupled Plasma–Chemical Vapor Deposition on ZrO 2 Gate Dielectric for Thin Film Transistors. Jpn. J. Appl. Phys. 45, 4365–4369 (2006).
[7]. Kim, W. et al. Paper-Based Surface-Enhanced Raman Spectroscopy for Diagnosing Prenatal Diseases in Women. ACS Nano 12, 7100–7108 (2018).
[8]. Chen, F. F. Physical mechanism of current-free double layers. Physics of Plasmas 13, 034502 (2006).
[9]. R.O. Dendy (ed.), Plasma Physics: An Introductory Course (Cambridge University Press, Cambridge, 1993)
[10]. Plasma Electronic.URL: https://www.plasma-electronics.com/chemical-vapor-deposition.html
[11]. Lebib, S. & Roca i Cabarrocas, P. Effects of ion energy on the crystal size and hydrogen bonding in plasma-deposited nanocrystalline silicon thin films. Journal of Applied Physics 97, 104334 (2005).
[12]. Venables, J. A. & Spiller, G. D. T. Nucleation and Growth of Thin Films. in Surface Mobilities on Solid Materials (ed. Binh, V. T.) 341–404 (Springer US, 1983). doi:10.1007/978-1-4684-4343-1_16
[13]. Chaudhary, D., Sharma, M., Sudhakar, S. & Kumar, S. Plasma Impedance Analysis: A Novel Approach for Investigating a Phase Transition from a-Si:H to nc-Si:H. Plasma Chem Plasma Process 37, 189–205 (2017).
[14]. Habuka, H. et al. Model on transport phenomena and epitaxial growth of silicon thin film in SiHCl3H2 system under atmospheric pressure. Journal of Crystal Growth 169, 61–72 (1996).
[15]. Sakurai, A., Saito, A. & Habuka, H. Surface and gas phase reactions induced in a trichlorosilane–SiH x system for silicon film deposition. Surface and Coatings Technology 272, 273–277 (2015).
[16]. Habuka, H., Sakurai, A. & Saito, A. By-Product Formation in a Trichlorosilane-Hydrogen System for Silicon Film Deposition. ECS J. Solid State Sci. Technol. 4, P16–P19 (2015).
[17]. Luo, P. et al. Effects of deposition pressure on the microstructural and optoelectrical properties of B-doped hydrogenated nanocrystalline silicon (nc-Si:H) thin films grown by hot-wire chemical vapor deposition. Microelectronics Journal 39, 12–19 (2008).
[18]. Matsuda, A. Microcrystalline silicon. Journal of Non-Crystalline Solids 338–340, 1–12 (2004).
[19]. Kushner, M. J. On the balance between silylene and silyl radicals in rf glow discharges in silane: The effect on deposition rates of a ‐Si:H. Journal of Applied Physics 62, 2803–2811 (1987).
[20]. Matsuda, A. Plasma and surface reactions for obtaining low defect density amorphous silicon at high growth rates. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 16, 365–368 (1998).
[21]. 張濟忠，”現代薄膜技術”，冶金工業出版社，2009
[22]. 王增福，”實用鍍膜技術”，電子工業出版社，2008
[23]. Moriaki Wakaki、周海憲、程云芳譯，”光學材料手冊”，化學工業出版社，2010
[24]. Physics and technology of amorphous-crystalline heterostructure silicon solar cells. (Springer, 2012).
[25]. Pearson, K. LIII. On lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 2, 559–572 (1901).
[26]. Hotelling, H. Analysis of a complex of statistical variables into principal components. Journal of Educational Psychology 24, 417–441 (1933).
[27]. J. Hogenboom and L. Barina, "Principal component analysis and sidechannel attacks-master thesis;' Master's thesis, 2010.
[28]. Cristianini, N. & Shawe-Taylor, J. An Introduction to Support Vector Machines and Other Kernel-based Learning Methods. (Cambridge University Press, 2000). doi:10.1017/CBO9780511801389
[29]. Vapnik, V. N. 統計學習理論的本質. (淸華大學出版社, 2000).
[30]. The 5th Annual Conference of Taiwan's Economic Empirics
[31]. Burges, C. J. C. [No title found]. Data Mining and Knowledge Discovery 2, 121–167 (1998).
[32]. 陳建勳，”非晶矽繞射光學元件的製作與分析”，國立中央大學物理研究所碩士論文，2005
[33]. Martins, R. et al. Role of ion bombardment and plasma impedance on the performances presented by undoped a-Si:H films. Thin Solid Films 383, 165–168 (2001).
[34]. Jagodzinski, H. H. P. Klug und L. E. Alexander: X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials, 2. Auflage. John Wiley & Sons, New York-Sydney-Toronto 1974, 966 Seiten, Preis: £ 18.55. Berichte der Bunsengesellschaft für physikalische Chemie 79, 553–553 (1975).
[35]. Amor, S. B., Bousbih, R., Ouertani, R., Dimassi, W. & Ezzaouia, H. Correlation between microstructure and properties of hydrogenated Si thin films grown by plasma enhanced chemical vapor deposition under different hydrogen flow rates. Solar Energy 103, 12–18 (2014).
[36]. Bustarret, E., Hachicha, M. A. & Brunel, M. Experimental determination of the nanocrystalline volume fraction in silicon thin films from Raman spectroscopy. Appl. Phys. Lett. 52, 1675–1677 (1988).
[37]. Zhou, H. P. et al. Dilution effect of Ar/H 2 on the microstructures and photovoltaic properties of nc-Si:H deposited in low frequency inductively coupled plasma. Journal of Applied Physics 110, 023517 (2011).
[38]. Sriraman, S., Agarwal, S., Aydil, E. S. & Maroudas, D. Mechanism of hydrogen-induced crystallization of amorphous silicon. Nature 418, 62–65 (2002).
[39]. Das, C. & Ray, S. Power density in RF PECVD: a factor for deposition of amorphous silicon thin films and successive solid phase crystallization. J. Phys. D: Appl. Phys. 35, 2211–2216 (2002).
[40]. Smets, A. H. M. & van de Sanden, M. C. M. Relation of the Si  H stretching frequency to the nanostructural Si-H bulk environment. Phys. Rev. B 76, 073202 (2007).
[41]. Finger, F. et al. Stability of microcrystalline silicon for thin film solar cell applications. IEE Proc., Circuits Devices Syst. 150, 300 (2003).
[42]. Hussain, S. Q. et al. Efficient light trapping for maskless large area randomly textured glass structures with various haze ratios in silicon thin film solar cells. Solar Energy 173, 1173–1180 (2018).
[43]. Habuka, H., Katayama, M., Shimada, M. & Okuyama, K. Nonlinear increase in silicon epitaxial growth rate in a SiHCl3H2 system under atmospheric pressure. Journal of Crystal Growth 182, 352–362 (1997).
[44]. Habuka, H. et al. Chemical process of silicon epitaxial growth in a SiHCl3–H2 system. Journal of Crystal Growth 207, 77–86 (1999).
[45]. Li, X. et al. Effect of deposition rate on the growth mechanism of microcrystalline silicon thin films using very high frequency PECVD. Optik 180, 104–112 (2019).
[46]. Goh, B. T., Wah, C. K., Aspanut, Z. & Rahman, S. A. Structural and optical properties of nc-Si:H thin films deposited by layer-by-layer technique. J Mater Sci: Mater Electron 25, 286–296 (2014).
[47]. Tabuchi, T., Toyoshima, Y., Fujimoto, S. & Takashiri, M. Remotely induced high-density hollow-anode plasma and its application to fast deposition of photosensitive microcrystalline silicon thin film with preferential orientation. AIP Advances 9, 055125 (2019).
[48]. Sharma, M., Juneja, S., Sudhakar, S., Chaudhary, D. & Kumar, S. Optimization of a-Si:H absorber layer grown under a low pressure regime by plasma-enhanced chemical vapor deposition: Revisiting the significance of the p/i interface for solar cells. Materials Science in Semiconductor Processing 43, 41–46 (2016).
[49]. Zhou, H. P. et al. Rapid and controllable a-Si:H-to-nc-Si:H transition induced by a high-density plasma route. J. Phys. D: Appl. Phys. 50, 385103 (2017).
[50]. Sriraman, S., Agarwal, S., Aydil, E. S. & Maroudas, D. Mechanism of hydrogen-induced crystallization of amorphous silicon. Nature 418, 62–65 (2002).
[51]. Shanks, H. et al. Infrared Spectrum and Structure of Hydrogenated Amorphous Silicon. phys. stat. sol. (b) 100, 43–56 (1980).
[52]. Losurdo, M. et al. Enhanced absorption in Au nanoparticles/a-Si:H/c-Si heterojunction solar cells exploiting Au surface plasmon resonance. Solar Energy Materials and Solar Cells 93, 1749–1754 (2009).
[53]. Naikal, N., Yang, A. Y. & Shankar Sastry, S. Informative feature selection for object recognition via Sparse PCA. in 2011 International Conference on Computer Vision 818–825 (IEEE, 2011). doi:10.1109/ICCV.2011.6126321
[54]. Vapnik, V. N. The nature of statistical learning theory. (Springer, 2000).
[55]. Dingari, N. C., Barman, I., Myakalwar, A. K., Tewari, S. P. & Kumar Gundawar, M. Incorporation of Support Vector Machines in the LIBS Toolbox for Sensitive and Robust Classification Amidst Unexpected Sample and System Variability. Anal. Chem. 84, 2686–2694 (2012).
[56]. Huang, Z., Chen, H., Hsu, C.-J., Chen, W.-H. & Wu, S. Credit rating analysis with support vector machines and neural networks: a market comparative study. Decision Support Systems 37, 543–558 (2004).