Thomson scattering versus modeling of the microwave plasma torch: a long standing discrepancy almost solved

Abstract

The characterization of plasmas gives much more insight if the results of modeling are compared with those of experiments. In the past, several modeling-experiment studies were performed on the microwave plasma torch (MPT). For the modeling a Global Plasma Model (GPM) was employed and for the experiment Thomson scattering (TS) was used. The same GPM-TS comparison was successfully applied to several other, more robust plasmas. However for the MPT it was found that the comparison led to severe discrepancies. The most salient feature was that the value of the electron temperature, T_e, found by TS, was substantially higher than what GPMs predicted. Several attempts were undertaken to solve this problem by refining the model and its most important input: the gradient of the electron density. The attention for the model implied that the TS technique was not scrutinized sufficiently. Recently it was discovered that the outcome of TS is very sensitive to the spectral fitting procedure. In the past this was carried out by employing the calibrated intensity using Gaussian fitting. However, a procedure based on the fitting of the spectral shape by using modified Gaussian fitting gives better results and leads to a lower T_e value, more in the range of the GPM results. This paper discusses the background of the GPM and gives the basis of TS fitting procedures and an analysis of the systematic error that can be made when TS is performed on filamentary MPT-like plasmas.