For the etching of silicon or silicon dioxide, CF 4 and NF 3 plasmas have been extensively explored, and many studies have been devoted to the understanding of the discharges in terms of plasma physics, chemistry, and surface chemistry. Among the studies, some works are reporting on modeling to grasp the main idea of the etch process.In an earlier work on modeling, 1 CF 4 plasma was approached as a "pseudo-black-box," in which expressions based on the conservation of fluorine and carbon atoms were derived which relate the concentration of the various species in the effluent gas to the etch rate. Chapman and Minkiewicz 2 proposed a simple model for flow rate effects in plasma etching, in which the etch rate is determined by the generation rate of active species and limited by lack of reactant gas at low flows and by pumping of active species at high flows. In a work by Economou and Alkire, 3 a mathematical model was formulated for a parallel plate plasma etching reactor for the case of the oxygen discharge. The model took account of the variation of electron density and energy with operating conditions, the potential distribution and ion transport in the sheath, the ion-assisted reaction kinetics on the surface, and the transport and reactions of neutral and charged species. Though the above mentioned models provided a semiquantitative understanding of several aspects of plasma etching and captured the salient features observed experimentally, quantitative comparisons were not possible owing to the lack of accurate rate reaction kinetics data.Meanwhile, several works on modeling have tried to make quantitative comparisons by using computer simulation with known kinetics data for a number of chemical reactions. 4-7 A model by Edelson and Flamm 4 simulated the CF 4 plasma etching of silicon using 40 reactions including gas-phase and surface reactions, where processes occurring at the gas-surface interface play a very important role even in the absence of silicon and gas-phase reactions of CF 2 are neglected. Disagreement in explaining the experimental results of Smolinsky and Flamm 8 was raised by Ryan and Plumb, 5,6 who developed a similar simulation model with different emphasis on the importance of gas-phase reactions. In a recent work by Meeks et al. 7 about 160 gas-phase reactions with accompanying rate coefficients are considered for modeling of chemical downstream etch systems for NF 3 /O 2 mixtures. It was noted, however, that the dominant gas-phase reaction paths can be limited to a much smaller reaction set through detailed sensitivity analysis of the fluorine atom concentration for the conditions of interest to the study.In a work on deposition kinetics of silicon carbide from plasmas of hexamethyldisiloxane (HMDSO), 9 we introduced a simple plasma chemistry model to simulate the deposition process. The plasma model is based on the postulation of direct proportionality of the electron density in the discharge to the discharge power. The postulation is supported by a recent work of Ahlrichs et al., 10 wher...