Perfil químico do café capixaba conilon & arábica por ESI(-)FT-ICR MS e ATR-FTIR: uma abordagem quantitativa

• Autor(a) principal: Correia, Radigya Meyrelles
• Data de Publicação: 2016
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Espírito Santo (riUfes)
• Texto Completo: http://repositorio.ufes.br/handle/10/7358
• topic: Calibração multivariada; Coffea Arábica L; Coffea canephora; Adulteração; Alimentos - Adulteração e inspeção; Espectrometria de massa; Espectroscopia de infravermelho; Quimiometria; Química; 54

Brazil is the largest coffee producer and world's second major consumer of coffee. The Espirito Santo State hasone of the largest national production centers of this product. Coffee is a very important fruit for the country'sGNP and the presence of adulterants infringe consumer rights, causing losses to the consumer and possible damage to health. Therefore, developing an analytical method to determine possible tampering would add value to the commodity and strengthening the country's economic growth. Robusta and Arabica roasted coffee samples were analyzed by negative-ion mode electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI(-)FT-ICR MS) and attenuated total reflection Fourier transform in the mid infrared region spectroscopy (ATR-FTIR) aiming to develop a quantification method of Robusta coffee in Arabica coffeeblends using uni-and multivariate data analysis models. To buildthe univariate calibration model with ESI(-)FT-ICR MS data, 10samples of Arabica coffee (Coffea Arabica) adulteratedwith different proportions of Robusta coffee (Coffea Canephora) and two pure samples (0% and 100 wt % of Robustacoffee), doped with an internal standard of 5x10-3mol L-1stearic acid (m/z283 and 567) were used. For the ATR-FTIR analysis a higher variability of adulterations were used, totaling 21blend samplesand two pure amostras(0% e 100 wt% Robusta coffee), and a partial least squares (PLS) regression model was proposed.ESI(-)FT-ICR results identified the main chemical species present in the coffee blendssuch as: quinic acid (m/z191), caffeoylquinic acid (m/z353) and feruloylquinic acid (m/z 367). The univariate calibration model obtained had a good linearity (R2of 0.9757) and limits of detection and quantification of 0.2 and 0.3 wt %, respectively. The PLS model built with ATR-FTIR data resulted in a RMSECV of 5.7 wt % and a RMSEP of 9.2 wt %, with coefficient of determination for cross-validation and prediction of 0.9635 and 0.9295, respectively.Repeatability and intermediate precision values of 4 wt% and 5wt% were found for ESI(-)FT-ICR MS respectively, and 1.7 wt% for both figures of merit in the ATR-FTIR analysis