CLEANER QUANTIFICATION OF DIETARY FIBRE IN CRACKER BISCUITS THROUGH FTIR AND CHEMOMETRICS

Abstract

In this work we present the optimization and implementation of a new cleaner method based on Fourier transform Infrared Spectroscopy (FTIR) and chemometrics for the quantification of fibers in cracker biscuits. For this purpose, 60 samples of cracker biscuits from the local market in Santa Cruz do Sul (RS, Brazil) were crushed and subjected to fiber analysis by acid and alkaline digestion according to the reference method. The samples were also analyzed by the technique of Attenuated Total Reflectance (ATR) in the range of 4000-650 cm^-1. The spectral data set was modeled via orthogonal partial least squares regression (OPLS) with data from the reference method in different spectral bands and pre-processing. An optimized ATR/OPLS model was obtained in the range of 4000-650 cm^-1 with 2 latent variables and data without pre-processing, transformed by the 1st derivative and by 1 component of orthogonal signal correction. The cross-validation returned a coefficient of determination of 0.9999 and a mean squared error of cross-validation (RMSECV) of 0.0115 %, thus proving the validity of the FTIR method with chemometrics for this purpose.