Automating the analysis of Langmuir probe traces: Construction of an algorithm and sensitivity analysis for probe traces from an electron cyclotron resonance plasma source

Abstract

Langmuir probe traces of plasmas are routinely analyzed to determine important characteristics, such as plasma density and electron temperatures. This article describes how a common manual technique to analyze probe data can be automated and how the sensitivity of the results to changes in the operator-specified starting values can be explored. The probe traces, which were taken in an electron cyclotron resonance plasma source, characteristically exhibit two Maxwellian electron populations, one at a bulk (Tebulk≊3–5 eV) and one at a hot (Tehot≊12–20 eV) temperature. The electron current (F) as a function of the sweep voltage (ν) of the probe traces is modeled by: F(ν) = Isate{(1 − p)exp[(Vp − ν)/Tebulk] + p exp[(Vp − ν)/Tehot]}, where Isate is the electron saturation current, p is the ratio of the hot electron current to the total, and Vp is the plasma potential. The algorithm for probe analysis partitions the sweep voltage into three intervals and uses different models and fitting techniques in each interval. The procedure’s sensitivity to different operator inputs is evaluated using factorial experiments.