HONEY AUTHENTICATION BY MEANS OF A POTENTIOMETRIC ELECTRONIC TONGUE

Be able to classify honeys according to their floral or geographical origin has become a priority issue as a results of the frauds observed. The European Community legislation on honey packaging recommends the use of labels indicating floral and geographical origin, as specific quality criteria. The identification of botanical source of honey responds to consumer demands and guarantees the quality of the products, avoiding frauds which have a significant economic impact and undeniable nutritional and organoleptic consequences. The work is part of a line of research very timely, aimed at the development of non-destructive techniques for botanical classification of honeys introducing, this one, the goal of authentication. For the purpose, it seems necessary to develop appropriate database. Given the lack of detailed studies on the honey of one of the most important regions of Italy for the production, this study was focused on honey of Calabria. A biomimetic system, α Astree e-tongue based on the different response of seven potentiometric sensors (ZZ, JE, BB, CA, GA, HA, JB) and chemometric data processing was used. The sensor array was applied to four honey groups of different botanical origin (Citrus, Citrus spp, Chestnut, Castanea Sativa Miller, Acacia, Robinia Pseudoacacia, and sulla, Hedysarum coronarium) being a representative sampling of an italian region (Calabria). The honey sensor signal patterns were treated by two different pattern recognition techniques: Principal Component Analysis (PCA) and Discriminant Function Analysis (DFA). The e-tongue was able to classify the honeys of different floral origins by PCA and DFA with a classification rates of 90 % and 93.75 %, respectively. The results showed the high potential of e-tongue as a tool for honey authentication.