High-precision modeling of dynamic etching in high-power magnetron sputtering

Abstract

Etching of the cathodes in magnetron sputtering determines the plasma discharge properties and deposition efficiency. In high-power and high-ionization discharges, etching becomes more complicated, resulting in inaccurate results if the conventional models are still used. This work aims at establishing an accurate dynamic model for high-power and high-ionization discharges by combining the cellular automata (CA) method and particle-in-cell/Monte Carlo collision (PIC/MCC) method, in which all the interactions pertaining to the etching morphology, plasma density, electric field, and magnetic field are considered. In high-power discharges such as continuous high-power magnetron sputtering (C-HPMS), strong self-sputtering and intense gas rarefaction stemming from the high temperature in the vicinity of the target influence the etching behavior. Compared to the experimental results, the morphology simulated by the dynamic etching model shows an error of only 0.8% in C-HPMS, which is much less than that obtained by the traditional test-electron Monte Carlo (MC) method (10.1%) and static PIC/MCC method (4.0%). The dynamic etching model provides more accurate results to aid the development and industrial application of high-power magnetron sputtering.