Characterization of uncertainties in electron-argon collision cross sections under statistical principles

Prediction for argon plasma is predicated upon the accurate characterization of electron-impact collision cross-sections, which determine the key reaction rates and transport properties of the plasma. Although these cross-sections have been the subject of experiments over decades, the resulting measurements do not always agree, and the uncertainties in the cross-sections have not been characterized. We evaluate the uncertainties in electron-argon collision cross-sections using Bayesian approach. Six collision processes (elastic momentum transfer, ionization, and 4 excitations) are characterized with semi-empirical models, whose parametric uncertainties effectively capture the essential features for the plasma chemistry/transport. These semi-empirical model are augmented by a Gaussian-process-based model to represent both systematic error in the experiments as well as the inadequacy of the semi-empirical forms. The parameters of the resulting model are calibrated via Bayesian inference using data from electron-beam experiments and ab-initio simulations. The resulting uncertainty in the cross-section model captures the scattering among the measurements, and is further validated against the swarm-parameter experiments via a 0D-Boltzmann analysis.