A carregar...
Publicação
QUANTIFICAÇÃO DE ELEMENTOS TRAÇO E MENORES COM FLUORESCÊNCIA DE RAIOS-X: ABORDAGEM OTIMIZADA PARA DETERMINAÇÃO DA COMPOSIÇÃO DA MATRIZ
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 32.68 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A análise da distribuição dos elementos traço e menores em diferentes tipos de amostras
é fundamental para melhor compreender a composição e os processos envolvidos em
materiais, em áreas como a biomedicina e a geologia. A técnica de Fluorescência de Raios-X
Dispersiva em Energia (EDXRF) é uma excelente escolha para essa análise por ser rápida,
não destrutiva, e devido aos seus bons limites de deteção, o que a torna eficaz na deteção
de elementos traço (~10 ppm) e menores (~100 ppm). No entanto, os efeitos de matriz
influenciam a precisão e exatidão na quantificação.
Neste trabalho, desenvolveu-se uma metodologia otimizada para determinar automa-
ticamente a matriz de amostras desconhecidas, com base na análise da região de dispersão
de um espetro EDXRF e no método dos Parâmetros Fundamentais (FP) que permite
uma quantificação melhorada ao simular os efeitos de matriz. Primeiro, foi criado um
modelo para ajustar os picos de dispersão (Compton e Rayleigh) das riscas características
do ânodo do tubo de raios-X, permitindo a construção de uma curva de calibração que
relaciona o rácio Compton/Rayleigh com o número atómico médio ( ¯𝑍) da amostra. Com
esse fim, foram utilizadas amostras modelo. Posteriormente, foi desenvolvido um método
para determinar a composição da matriz a partir do ¯𝑍, aplicando diferentes abordagens
consoante o tipo de amostra, e desenvolvendo um script de otimização. Para isso, foram
utilizados Materiais Certificados de Referência (CRMs).
O modelo de ajuste dos picos de dispersão apresentou boa qualidade quando com-
parado a um software comercial. As amostras biológicas (tecidos moles, sangue e cabelo)
apresentaram os melhores desempenhos quantitativos, em comparação com as amostras
geológicas (solos, sedimentos e rocha fosfática) que, embora com alguns desvios, também
foram adequadamente quantificadas. O script de otimização sugeriu boa capacidade de
previsão das concentrações dos elementos da matriz. Esta metodologia reduz o tempo de
análise quantitativa e, ainda assim, obtém resultados promissores. É viável para diferentes
tipos de amostra, embora sejam necessárias melhorias para maximizar a sua eficácia.
The analysis of trace and minor element distribution in different types of samples is essential for a better understanding of the compostion and processes involved in materials, in fields such as biomedicine and geology. The Energy Despersive X-Ray Fluorescence (EDXRF) technique is an excellent choice for this analysis because it is fast, non-destructive, and has good detection limits, making it effective for detecting trace () and minor () elements. However, matrix effects influence the accuracy and precision of quantification. In this work, an optimized methodology was developed to automatically determine the matrix of unknown samples based on the analysis of the scattering region of an EDXRF spectrum and the Fundamental Parameters (FP) method, which allows improved quantification by simulating matrix effects. First, a model was created to fit the scattering peaks (Compton and Rayleigh) of the characteristic lines of the X-ray tube anode, allowing for construction of a calibration curve that correlates the Compton/Rayleigh ratio with the average atomic number ( ¯𝑍) of the sample. To achieve this, model samples were used. Subsequently, a method was developed to determine the matrix composition from ¯𝑍, applying different approaches depending on the sample type, and developing an optimization script. Certified Reference Materials (CRMs) were used for this purpose. The scattering peak fitting model showed good quality when compared to a commercial software. Biological samples (soft tissues, blood and hair) demonstrated the best quan- titative performances compared to geological samples (soils, sediments and phosphate rock), which, despite some deviations, were also adequately quantified. The optimization script suggested good predictive capability for the matrix element concentrations. This methodology reduces quantitative analysis time while still achieving promising results. It is viable fo different types of samples, although further improvements are necessary to maximize its efficacy.
The analysis of trace and minor element distribution in different types of samples is essential for a better understanding of the compostion and processes involved in materials, in fields such as biomedicine and geology. The Energy Despersive X-Ray Fluorescence (EDXRF) technique is an excellent choice for this analysis because it is fast, non-destructive, and has good detection limits, making it effective for detecting trace () and minor () elements. However, matrix effects influence the accuracy and precision of quantification. In this work, an optimized methodology was developed to automatically determine the matrix of unknown samples based on the analysis of the scattering region of an EDXRF spectrum and the Fundamental Parameters (FP) method, which allows improved quantification by simulating matrix effects. First, a model was created to fit the scattering peaks (Compton and Rayleigh) of the characteristic lines of the X-ray tube anode, allowing for construction of a calibration curve that correlates the Compton/Rayleigh ratio with the average atomic number ( ¯𝑍) of the sample. To achieve this, model samples were used. Subsequently, a method was developed to determine the matrix composition from ¯𝑍, applying different approaches depending on the sample type, and developing an optimization script. Certified Reference Materials (CRMs) were used for this purpose. The scattering peak fitting model showed good quality when compared to a commercial software. Biological samples (soft tissues, blood and hair) demonstrated the best quan- titative performances compared to geological samples (soils, sediments and phosphate rock), which, despite some deviations, were also adequately quantified. The optimization script suggested good predictive capability for the matrix element concentrations. This methodology reduces quantitative analysis time while still achieving promising results. It is viable fo different types of samples, although further improvements are necessary to maximize its efficacy.
Descrição
Palavras-chave
EDXRF Elementos Traço e Menores Matriz Dispersão de Compton e de Rayleigh Parâmetros Fundamentais Número Atómico Médio