Effect of radio-frequency electric power applied to a boron nitride unbalanced magnetron sputter target on the deposition of cubic boron nitride thin film

• Published in: Metals and materials international Vol. 19; no. 6; pp. 1317 - 1321
• Main Authors: Ko, Ji-Sun, Park, Jong-Keuk, Lee, Wook-Seong, Huh, Joo-Youl, Baik, Young-Joon
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.11.2013; Springer Nature B.V; 대한금속·재료학회
• Subjects: Bias; Boron nitride; Characterization and Evaluation of Materials; Chemistry and Materials Science; Cubic boron nitride; Deposition; Electric potential; Electric power generation; Engineering Thermodynamics; Heat and Mass Transfer; Machines; Magnetic Materials; Magnetism; Manufacturing; Materials Science; Metallic Materials; Processes; Silicon substrates; Solid Mechanics; Thin films; Voltage; 재료공학; crystal structure; residual stress; cubic boron nitride; Fourier transformed infrared spectroscopy; sputtering
• ISSN: 1598-9623; 2005-4149
• DOI: 10.1007/s12540-013-6028-5

Cubic boron nitride (c-BN) films were deposited by an unbalanced magnetron sputtering method. A (100) Si wafer with a nanocrystalline diamond thin film as a surface coating layer or that without it was used as a substrate. The target power was varied from 100 to 400 W. A boron nitride target was used, which was connected to a radio frequency power supply. High frequency power connected to a substrate holder was used for self-biasing. The deposition pressure was 0.27 MPa with a flow of Ar (18 sccm) — N 2 (2 sccm) mixed gas. The existence of threshold bias voltages for c-BN formation and resputtering were observed irrespective of target power. The bias voltage window for c-BN formation broadened with increased target power. The deposition rate decreased with enhanced bias voltage and decreased target power. Residual stresses of the films did not vary noticeably with target power within the target power range of c-BN formation. A parameter space for c-BN formation according to the target power and the bias voltage, as two variables, was suggested.