Enhancing Process Stability of AIN Thin Films through Pre-Deposition in Chamber and Optical Emission Spectroscopy Assisted by Big Data Analysis

In order to improve the quality of aluminum nitride (AlN) films, this study used 13.56 MHz radio frequency reactive sputtering technology through pre-sputtering chamber and used optical emission spectroscopy (OES) to measure elemental composition of gases to ensure the stability of the gas environment in the chamber [1]. Our focus is on evaluating differences in film properties such as crystallinity, surface morphology, and thickness uniformity. Our results show that the film quality of AlN is stable after the predetermined time exceeds 40 minutes, showing better crystalline quality, as evidenced by sharper and more intense XRD peaks (especially the <002> peak). Furthermore, AFM measurements show that the surface roughness of films deposited by RF sputtering decreases with increasing nitrogen concentration, with root mean square (RMS) values ranging from 1.54 nm to 3.24 nm. OES has been used to study the production of active species in plasmas excited by pulsed DC and RF power sources in nitrogen atmospheres. The higher spectral intensity observed in RF sputtering, especially the emission characteristics of ionized nitrogen (N<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>) near 390.93 nm [2], indicates that RF plasma provides energy conditions favorable for the formation of high-quality AlN films. Furthermore, we apply principal component analysis (PCA) algorithm for big data analysis to reduce the dimensionality and visualize the clustering results of OES data recorded during different pre-deposition times.