Numerical Simulation Of Plasma Etching Based On CFD And Parameter Optimization

Plasma etching technology has been widely applied to semiconductor fabrication, with the increasing of integrated circuit scale and the shrinking of feature size for electronic components. Plasma etching process has been grasped for the intensive researches during in the past few decades. However, as the analysis and design of plasma etching are related to coupling multi-physics fields, such as flow field, temperature field, electromagnetic field and plasma field, and geometric multiscale features, it is difficult to explain some complex phenomena and mechanism completely now. Traditional method has high cost and long cycle, which mainly depends on the experience of trial and error to improve the quality of products and productive efficiency. So a numerical process to simulate SiO2etching with inductively coupled C4Fg/Ar plasma has been constructed using the commercial software of CFD-ACE and CFD-TOPO. The simulation model was found to be reasonable to analyze the plasma etching process, and to investigate the effects of processing parameters on the plasma character and etching profile. Based on surrogate model, the processing parameters affecting the etching rate were optimized, and the effectiveness and feasibility of proposed method were demonstrated.In chapter2, the fundamental theory and technique of plasma etching are introduced, especially the numerical simulation technique related to plasma etching.In Chapter3, effect of the processing parameters on the electron number density and temperature was researched firstly. It was founded that, the electron number density and temperature both increase with the RF power. The electron number density increases with the pressure, while the changing trend of electron temperature is opposite. The RF bias has little effect on the electron number density and temperature. Then the energy and angle distribution of ions were also investigated. The energy distribution of ions has two peaks as the alternating current on the chuck, and the average energy of ions increases with the increasing of RF bias. It was founded that the ions reached the wafer at different angles but within9degree from the surface normal. Finally the effect of processing parameters on etching profile was investigated. The result showed that the lateral etching rate increase with the RF power and bias, while the changing trend of pressure is opposite.In chapter4, the sensitivity analysis of processing parameters to etching rate has been done firstly. It was founded that the vertical etching rate monotonically increases with the RF bias and the fraction of C4F8. However, the change of vertical etching rate with RF power and pressure is not monotonous. On this basis, the RF power and pressure were optimized based on surrogate model. Compared with full factorial design, Latin hypercube design can give relatively accurate result with less sample points. And the accuracy of the Kriging model is the best, compared with the Least Square Regression and Moving Least Square modelThe simulation of SiO2etching process with inductively coupled C4F8/AR plasma provides theoretical and directional guidance for experiment study. This research is supported by the No.2national science and technology major project. And the optimization method based on surrogate model offers a feasible and effective technique for finding the optimal parameters in the process of plasma etching.