Multivariate regression models for the simultaneous quantitative analysis of calcium and magnesium carbonates and magnesium oxide through drifts data

• Autor(a) principal: Marder,Luciano
• Data de Publicação: 2006
• Outros Autores: Tomedi,Paulo, Ferrão,Marco F., Jablonski,André, Davanzo,Celso U.
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Journal of the Brazilian Chemical Society (Online)
• Texto Completo: http://old.scielo.br/scielo.php?script=sci_arttext&pid=S0103-50532006000300025
• topic: dolomite; DRIFTS; partial least squares; multivariate regression; ternary mixture design

In the present work multivariate regression models were developed for the quantitative analysis of ternary systems using Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) to determine the concentration in weight of calcium carbonate, magnesium carbonate and magnesium oxide. Nineteen spectra of standard samples previously defined in ternary diagram by mixture design were prepared and mid-infrared diffuse reflectance spectra were recorded. The partial least squares (PLS) regression method was applied to the model. The spectra set was preprocessed by either mean-centered and variance-scaled (model 2) or mean-centered only (model 1). The results based on the prediction performance of the external validation set expressed by RMSEP (root mean square error of prediction) demonstrated that it is possible to develop good models to simultaneously determine calcium carbonate, magnesium carbonate and magnesium oxide content in powdered samples that can be used in the study of the thermal decomposition of dolomite rocks.