Multi-parametric analysis of green coffee. Geographical Origin Profiling

SERRA, F.; ALONSO-SALCES, R. M.; RENIERO, F.; GUILLOU, C.

Praga (República Checa)

Simposio; 2nd International Symposium on Recent Advances in Food Analysis; 2005

Institute of Chemical Technology Prague, IAEA and RIKILT of the University of Wageningen

Due to its broad diffusion and high price, it is not unusual that coffee is subject to adulteration. In the strict sense of consumer safety, a coffee blended with cheaper products (chicory, cereals and leguminous plants) represents a fraud affecting the basic quality of the product when the ingredients listed on the label offer an untrue description of the content on the package. More often frauds are related to the intrinsic quality, concerning the sensorial properties associated with the geographic origin of the coffee. Several attempts have been made to develop analytical tools that are able to confirm appellation systems already in use. These approaches deal with the analysis of coffee tocopherols and triglycerides by HPLC (Gonzalez A.G., 2000), the determination of more than 50 elements in coffee by semi–quantitative scan ICP–MS (Prodolliet J. et al., 2001), the analysis of d13C and d15N coupled with the site–specific (D/H) isotope ratio by 2H–NMR on caffeine extracted from green coffee (Prodolliet J. et al., 1997), or the characterization of green coffee by d13C, d15N and d11B analysis (Serra F. et al., 2005). Total polyphenols together with other chemical descriptors (metals, amino-acids, furfurals, caffeine) has been used to perform varietal classification of coffee or discriminate among different kinds of roasted coffees (Martin M. J. et al., 1998). In this study, a huge sample set covering all the regions producing coffee in America, Africa and Asia are studied. The polyphenolic profiles analysed by HPLC-DAD-MS and the isotopic content by IRMS of coffee beans, are evaluated by pattern recognition techniques in order to classify coffees according to their geographical origin.