Multi-diagnostic experimental validation of PIC/MCC simulations of symmetric capacitively coupled plasmas in CF4

In plasma science, particle-in-cell/Monte Carlo collision (PIC/MCC) simulations are an important tool. Using them as digital twins for knowledge-based plasma process development requires quantitative agreement with experiments. Based on multiple experimental diagnostics applied to a geometrically symmetric capacitively coupled radio frequency plasma, a 1d3v PIC/MCC code is validated for CF4 gas. Measurements of the ion flux energy distribution function, electron density, and spatio-temporally resolved F atom excitation rate at a driving frequency of 13.56 MHz, a peak-to-peak voltage ranging from 150 V to 900 V, and pressures of 5 Pa and 30 Pa, are compared to simulation results. Ion-neutral cross section data, commonly used in previous work, are found to provide a limited accuracy only for the ion dynamics and conversion in the sheaths, resulting in ion flux energy distribution functions at the electrodes that differ significantly from measured data. In agreement with more recent datasets, the ion-neutral cross sections are, therefore, scaled to reproduce the experimental results. The coefficients for ion-ion recombination, secondary electron emission, and electron reflection at the electrodes are systematically varied to achieve precise alignment with the measured plasma properties including the accurate description of electron power absorption mode transitions as a function of the driving voltage amplitude. The choice of the ion-ion recombination coefficient is found to have a strong effect on this mode transition in the simulation. With these key input parameters adjusted, good quantitative agreement between simulation and experimental results is achieved over a wide range of external control parameters providing the basis for quantitatively accurate digital-twin plasma simulations.