|
Abstract
|
The aim of present study is to investigate the combination of Fourier transform infrared (FT-IR)
spectroscopy with pattern recognition to recognize the standard saffron from those which have been
adulterated with various types of food colorants. Transmittance FT-IR spectra have been obtained for
standard saffron and six mixed samples with food colorants including Tartrazine, Sunset yellow,
Azorubine, Quinoline-yellow, Allura red and Sudan II. Genetic algorithm-linear discriminant analysis (GALDA)
based on the concept of clustering of variables has been applied to transmittance FT-IR spectra for
classification of standard saffron from fraudulent samples. Analysis of the selected clusters of variables
indicates that three bands corresponding to 1800–1830, 2600–2900 and 3700–3850 cm1 are
responsible for differentiation of standard samples from fraudulent ones. Regression analysis has been
introduced in order to obtain information related to the amount of food colorant. A combination of FTIR
and the concept of clustering of variables resulted in the best performances for calibration and an
external test set with 100% sensitivity and specificity
|