Modeling of plasma density, argon ion energy and ion velocity functions in a dipolar electron cyclotron resonance plasma source

Abstract

Improving the surface feature precision of semiconductor devices largely depends on the ion distribution functions, which are non-Maxwellian and varying in processing plasmas spatially. In this work, a hybrid plasma model (HPM) is developed for an Ar discharge of dipolar electron cyclotron resonance (ECR) to explore the characteristics of ion energy distributions (IEDs), ion angular distributions (IADs) and ion velocity distributions (IVDs). The hybrid model contains a fluid model that determines the electric space characteristics of plasmas, as well as a Particle in Cell/Monte Carlo (PIC/MC) model that simulates ion distributions in consideration of the collisions between ions and neutrals. The influences of the pressure, magnet current, and position on the ion distributions are also discussed. The simulation results reveal the features of IEDs which transforms from a single energy peak into a bimodal distribution with the increasing pressure. IADs have a significant peak at a small angle, and most ions strike to the substrate surface with an angle less than 10°. Increasing the magnet current contributes to broader IEDs yet does not significantly alter neither IADs nor IVDs. The sensitive position dependence of ion distributions reveals considerable anisotropy of ions on the ECR surface.