Detection of adulteration in honey by discriminant analysis using FTIR spectroscopy

The application of mid-infrared (MIR) spectroscopy as a screening tool for the determination of sugar adulteration in honey is discussed Honey samples with different adulterant sugars were scanned using the attenuated total r eflectance accessory of the Bio-Rad FTS-6000 Fourier transform spectrometer Adulterants considered were glucose, fructose, sucrose, and corn syrup. Pr edictive models were developed to classify the adulterated honey samples us ing discriminant analysis. Spectral data was compressed using principal com ponent analysis, and linear discriminant and canonical variate techniques w ere used to discriminate adulterated honey samples. The correct classificat ion of 100% was achieved using canonical variate analysis for honey samples adulterated with corn syrup and sugar mixtures. The best predictive models achieved using linear discriminant analyses correctly classified 90% of va lidation test for sucrose and corn syrup adulterated samples. Results demon strated that discriminant analysis of the spectra of adulterated honey samp les could be used for classification.