Growth of Aluminum Nitride Films on Different Substrates by RF Magnetron Sputtering System

• Main Authors: WANG, JING-JIE, 王靖捷
• Other Authors: KUAN, HON
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/g3snju

碩士 === 南臺科技大學 === 光電工程系 === 107 === In this study, RF magnetron Sputter was used to sputter an aluminum nitride (AlN) film on a silicon substrate, a sapphire substrate, and a patterned sapphire substrate, this aluminum nitride film will serve as a buffer layer between the LED substrate and the n-GaN layer. The function of the buffer layer is to reduce the stress caused by the difference in lattice difference and thermal expansion coefficient between the substrate and the gallium nitride, resulting in poor epitaxial quality. The parameters of different film thicknesses were obtained by changing the process time: 165min, 82.5min, 55min, 27.5min, 13.75min, 5min, 3.75min and 3min. After this process, I will use surface profiler (α-step), Scanning Electron Microscopy (SEM), X-ray Diffraction (XRD), Transmission Electron Microscope (TEM), Energy Dispersive Spectrometer (EDS) and Atomic Force Microscope (AFM) measuring instruments were used to analyze the aluminum nitride films prepared on different substrates and different process times. In this experiment, it was found by SEM measurement that both the silicon substrate and the sapphire substrate had larger grain sizes as the process time increased (3minutes to 165minutes), and the thicker the film thickness, the easier it is to cause cracks in the film due to stress. It can be seen from the XRD measurement that the ruthenium substrate grows along the (100) and (103) directions when the film thickness is thin, and when the thickness is about 100 nm, the (002)-oriented diffraction peak appears. At about 300 nm, the (100) and (103) orientations almost disappeared, and the crystal growth direction changed to the (002) orientation. The crystal growth direction of the sapphire substrate grows along the (101) orientation and exhibits a very strong peak. The final result shows that the full width at half maximum (FWHM) obtained by the 3.75min and 27.5min of process time both showed a narrow and high peak phenomenon, so it was judged that the crystal growth effect at 3.75min and 27.5min was better.