Find in Library

The false declaration of honey authenticity requires the use of rapid and efficient analytical tools in order to be detected. In this study, the use of a green, rapid and emerging approach for food authentication, FT-Raman spectroscopy, proved to obtain reliable and efficient honey botanical and harvesting year differentiation models, when the spectroscopic data was processed by employing a supervised statistical method, namely Partial Least Squares Discriminant Analysis (PLS-DA). The peaks and bands present in the Raman spectra were discussed based on honey composition. In order to increase the efficiency of the models, different preprocessing methods were used and a variable reduction step was employed. The new authentication approach is capable of distinguishing among four botanical sources and two harvesting periods of honey with a correct prediction rate higher than 97 %. The Raman markers that proved to contribute the most to the discrimination were correlated with the honey composition.