Plasma processes are used industrially for etching or deposition of materials for example for the microelectronics industry. The plasma etching/deposition process is very complicated due to the interplay and link among discharge physics, chemistry, and surface chemistry. Predictive models do not exist, and modelling efforts concentrate only on part of the whole problem, e.g. on the plasma physics. The modelling approach described here first analyses the plasma in modules, which are studied separately: They include the plasma physics, the plasma chemistry, and the surface chemistry modules. After the analysis and testing of the modules, their synthesis into a complete plasma simulator with predictive capabilities is attempted. The interactions among the physics and chemistry modules, and the surface and plasma chemistry modules are taken into account: an iterative solution procedure shows that the complete simulator converges to a consistent solution. Substantial differences between the converged solution and that without any interactions are demonstrated.
In this paper the one-dimensional in space, and complete plasma simulator is applied for a radio frequency CH4 plasma used for the deposition of Diamond-Like Carbon at 100 mTorr. The physics module calculations show that the negative ion density in the bulk plasma is less than one tenth of the electron density. A generalised gas phase chemistry model is also written and the concentrations of CH3 (1012/cm3), CH2 (1010/cm3) , and H are calculated. A detailed surface deposition model is then formulated using a multi-adsorption model. Carbon deposition from direct ion incorporation and ion induced stitching of neutrals is considered. It is found that the synergy between plasma and surface chemistries is very important for correct calculations of species densities and deposition rates, which are in good agreement with experiment.