LPWS法を用いたAr RF 平行平板プラズマのモンテカルロシミュレーション

The geometry, pressure and power coupling conditions of most plasma sources for semiconductor manufacturing lend themselves to particle simulations such as Monte Carlo simulations(here after MCS). Usually the kinetics solvers are coupled to solvers for Poisson’s Equation. MCS usually employ averaging over discrete regions of parameter space. When the number of discrete regions is increased, two problems result: one is the instability of the solution because of a statistical change and the other is the increase of the calculation time. Ventzek and Kitamori (J. Appl. Phys., vol. 75, pp. 3785-3788, 1994) proposed Legendre Polynomial Weighted Sampling (here after LPWS) which aimed to optimize sampling statistics with an economy of particles. In this paper, we characterize an Ar RF parallel plate discharge using a MCS employing LPWS based on the Date’s model (T. IEE Japan, 111-A, 11, pp. 962-972, 1991). The method is shown to replicate the behavior of RF discharges with high fidelity.