Quantitative elemental analysis using Optical Emission Spectroscopy (OES) starts with a high level of confidence in spectral line assignment from reference databases. Spectral interferences caused by instrumental and line broadening decrease the resolution of OES spectra creating uncertainty in the elemental profile of a sample for the first time. An approach has been developed to quantify spectral interferences for individual line assignment in OES. The algorithm calculates a statistical interference factor (SIF) that combines a physical understanding of plasma emission with a Bayesian analysis of the OES spectrum. It can be used on a single optical spectrum and still address individual lines. Contrary to current methods, quantification of the uncertainty in elemental profiles of OES, leads to more accurate results, higher reliability and validation of the method. The SIF algorithm was evaluated for Laser-Induced Breakdown Spectroscopy (LIBS) on samples with increasing complexity: from silicon to nickel spiked alumina to NIST standards (600 glass series and nickel-chromium alloy). The influence of the user's knowledge of the sample composition was studied and showed that for the majority of spectral lines this information is not changing the line assignment for simple compositions. Nonetheless, the amount of interference could change with this information, as expected. Variance of the SIF results for NIST glass standard was evaluated by the chi-square hypothesis test of variance showing that the results of the SIF algorithm are very reproducible.
If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu
Doctor of Philosophy (Ph.D.)
College of Sciences
Length of Campus-only Access
Doctoral Dissertation (Open Access)
Chappell, Jessica, "Quantitative Line Assignment in Optical Emission Spectroscopy" (2018). Electronic Theses and Dissertations. 6387.
Restricted to the UCF community until 2-15-2020; it will then be open access.