BACKGROUND: Phenolic compound profiles of 20 honeys of different botanical origin (eucalyptus, citrus, chestnut and linden) were obtained by high-performance liquid chromatography with ultraviolet detection after solid phase extraction, in order to evaluate the effectiveness of the fingerprint method for monofloral honey discrimination. RESULTS: A total of 58 peaks were detected at λ = 280 nm. Distinctive phenolic compound profiles were obtained in which both the nature and the relative amount of the detected compounds were characteristic for different botanical source honeys. In order to detect sample groupings, chromatographic peak areas were submitted to principal component analysis. Then linear discriminant analysis was carried out on the first three principal components. In addition, linear discriminant analysis was carried out on the 58 variables, allowing the selection of five variables able to discriminate honeys of different botanical origin. CONCLUSION: The chemometric evaluation of the phenolic compound profiles yielded classification models able to group honey samples according to their floral source with an excellent degree of agreement. The main advantage of the fingerprint approach with respect to traditional methods is that it does not require time-consuming identification and quantification of the analytes. The method proved to be effective for the assessment of honey authenticity.
BACKGROUND: Phenolic compound profiles of 20 honeys of different botanical origin (eucalyptus, citrus, chestnut and linden) were obtained by high-performance liquid chromatography with ultraviolet detection after solid phase extraction, in order to evaluate the effectiveness of the fingerprint method for monofloral honey discrimination. RESULTS: A total of 58 peaks were detected at λ = 280 nm. Distinctive phenolic compound profiles were obtained in which both the nature and the relative amount of the detected compounds were characteristic for different botanical source honeys. In order to detect sample groupings, chromatographic peak areas were submitted to principal component analysis. Then linear discriminant analysis was carried out on the first three principal components. In addition, linear discriminant analysis was carried out on the 58 variables, allowing the selection of five variables able to discriminate honeys of different botanical origin. CONCLUSION: The chemometric evaluation of the phenolic compound profiles yielded classification models able to group honey samples according to their floral source with an excellent degree of agreement. The main advantage of the fingerprint approach with respect to traditional methods is that it does not require time-consuming identification and quantification of the analytes. The method proved to be effective for the assessment of honey authenticity.