X-Ray Energy Dispersive Spectrometer Development in a Triaxial Geometry

• Autor(a) principal: Inocêncio, Rúben Miguel Gonçalves
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/160734
• topic: Triaxial X-ray fluorescence spectrometry; Exponential modified gaussian; Quantification software; Fundamental parameters quantification; Domínio/Área Científica::Engenharia e Tecnologia::Outras Engenharias e Tecnologias

Energy dispersive X-ray fluorescence (EDXRF) spectrometry is a non-destructive method to perform qualitative and quantitative analysis of elements present in solid and liquid samples. This technique is widely used in industry and academia despite facing some difficulties on offering high signal-to-noise, due to Bremsstrahlung background noise, and because it deeply relies on a software capable of analysing the data assertively. This dissertation presents the development of an EDXRF spectrometer in a triaxial geometry (3EDXRF) and a software for spectra analysis. The adaptation of the classical XRF setup into a triaxial geometry raises the signal-to-noise ratio by progressively eliminating the orthogonal components of the eletromagnetic field that makes Bremsstrahlung photons. Furthermore, this 3EDXRF spectrometer is coupled with a double crystal wavelength dispersive spectrometer (DCS) that enables analyses with greater energy resolution which makes this setup exclusive with high scientific importance. It was further developed a software for spectra analysis in Python that improves spectra reading by introducing exponential modified gaussians (EMG) to better fit the spectra components, achieving more reliable results. By modelling the observed phenomena using complex physical models and solving them with non-linear optimization algorithms, the elements in a sample are quantified. While developing the cluster of spectrometers, the fluorescence spectra of a set of metallic alloys containing Ag, Cu and Zn were taken as a case study. The results from fitting the spectra are presented along with the quantification results. By implementing the EMG, the spectra fitting improved and the errors associated with the fit were reduced. Concerning the quantification, the methodology produced consistent results with the ones considered as a control, showing a potential success on modelling the physical phenomena considered in the quantification. However, quantification results show relative errors on the order of tens percent, indicating our model needs further developments in order to be more complete and accurate. Such improvements may start from considering multiple fluorescence events.