A detailed uncertainty analysis of EI-MBMS data from combustion experiments

• Published in: Combustion and flame 2022-09, Vol.243, p.112012, Article 112012
• Main Authors: He, Xiaoyu, Giese, Maximilian, Ruwe, Lena, Lucassen, Arnas, Moshammer, Kai
• Format: Article
• Language: English
• Subjects: Combustion experiments; Electron ionization; Molecular-beam mass spectrometry; Uncertainty analysis
• ISSN: 0010-2180; 1556-2921
• DOI: 10.1016/j.combustflame.2022.112012

Electron ionization molecular-beam mass spectrometry (EI-MBMS) is a widely used analytical method for studying the chemistry of combustion processes and for providing quantitative validation targets for kinetic model development often in form of species mole fraction profiles. The present study focuses on the determination of the uncertainties of such data, which result from the mass spectrometric measurements as well as from the data evaluation procedure, in particular through the use of different calibration methods, i.e., through a direct calibration of the instrument, a convolution procedure, and through the method of Relative Ionization Cross Sections (RICS). A series of calibration experiments were carried out using two independent time-of-flight mass spectrometers in order to check the influence of experimental and instrumental parameters on the uncertainties. The data obtained reveal several key contributors to the uncertainty, which are, among others, the signal intensity of a species in the mass spectrum, the instability of instrumental parameters, the species itself, and the calibration method. The propagation of these uncertainties from the different calibration methods as well as from the instrumental inconstancies into the final uncertainties of the mole fractions is further demonstrated using a gaussian error propagation and provision of sensitivities for different uncertainty sources is given. It was found that most of the uncertainties of the mole fractions reported in the literature are smaller than those reported here and presumably only consider the reproducibility of the signal. The use of two independent mass spectrometers for the experiments also revealed that the uncertainties determined for one instrument cannot be directly transferred to another especially for the method of direct calibration due to differences in the signal reproducibility, but that the uncertainties only change marginal for the semi-theoretical approaches.