二氧化钛纳米管阵列的金属掺杂改性及人工神经网络对其形貌的预测研究-材料物理与化学专业论文.docx

二氧化钛纳米管阵列的金属掺杂改性及人工神经网络对其形貌的预测研究 THE STUDY OF DECORATING TITANIA NANOTUBE ARRAYS WITH METALS AND PREDICTING ITS MORPHOLOGY BY ARTIFICIAL NEURAL NETWORK ABSTRACT The preparation of TiO2 nanotube arrays by anodizing has been investigated in this dissertation. Ag nanoparticles were loaded on TiO2 nanotube arrays by AC electrodeposition in AgNO3 solution. The influence of concentration of AgNO3, deposition time and deposition voltage on photoelectric conversion properties of TiO2 nanotube layers were discussed. The samples were characterized by X-ray diffraction(XRD) and scanning electronic microscopy(SEM). The results show that Ag nanoparticles were loaded onto the surface of TiO2 nanotube arrays. The photocurrent density of TiO2 was increased by 5.8 times after Ag deposition under visible light and 4.0 times under UV illumination without external voltage. In addition, the photocurrent density of TiO2 was increased by 2.5 times and 2.8 times respectively when external voltage was 0.5V. It was confirmed that the presence of Ag on TiO2 catalysts could enhance the photocatalytic oxidation of methyl orange solution (15mg/L). The experimental results showed that degradation rate was different with different pH. The best degradation rate of methyl orange was obtained when the pH value equaled 13. The mechanism of enhancing photoelectric activity of decorating titania with Ag was discussed from the view of energy band theory. Based on the analysis, we confirmed that photocatalytic properties as well as photoelectric conversion properties would be enhanced by cheaper metals such as Fe, Co and Ni, which was evidenced by further experiments. The present paper tries to use the Artificial Neural Network method to make prediction about the morphology of the TiO2 nanotube grown by anodization. Firstly, the collected experimental data was simplified in an innovative way that considered the data of electrolyte content as sparse matrix which can be stated by the location and number o