Self-Consistent Numerical Framework for Modeling Planar DC Magnetron Sputtering

In this study, we propose a self-consistent numerical framework for modeling planar Direct Current (DC) magnetron sputtering for a single alloy target. Unlike other reported numerical methods in literature which need experimental data of racetrack and discharge current to calculate the argon ion flux on target, the present self-consistent model obtains that necessary information from discharge simulation without doing experiment. Furthermore, the model also shows the advantage of more accurate calculation of total knocked out target atoms due to take into consideration of spatial argon ion flux and energy distribution on the target. In the model, we account for electromagnetics, fluid dynamics, discharge chemistry and thermal transport across multiple length scales. Specifically, we employ Lattice Boltzmann Method (LBM) for transport phenomena, Finite Elements Method (FEM) for magnetron discharge transport, and binary collision approximation Monte-Carlo method for sputtering and atomic deposition. Our modeling framework could predict the composition, uniformity, and deposition flux of deposited thin film alloys. Furthermore, the simulation also provides the information of plasma in the chamber such as electron, ion densities, argon ion energy and argon ion flux bombarding on the target. Through parametric analyses, we found that a reduction in substrate radius and substrate-target distance can lead to improved deposition efficiency and film uniformity, and that magnetic fields can significantly affect the substrate deposition flux and film uniformity. Substrate composition is found to scale with both chamber pressure and inverse chamber temperature, although the effect of chamber temperature is considerably marginal under low pressure conditions. Our work provides useful insights to operational heuristics involved in magnetron sputtering and other physical chemical processes.