Profiling of phenolic compounds using UPLC–MS for determining the geographical origin of green coffee beans from Ethiopia

Abstract

A total of 100 samples of green coffee (Coffea arabica L.) beans from the major producing regions of Ethiopia were studied using ultra performance liquid chromatography–mass spectroscopy to determine if the phenolic content could be linked to their geographical origins for authentication purposes. Principal component analysis allowed the most discriminating compounds to be identified. Based on their concentrations, 3-O-caffeoylquinic and 4,5-O-dicaffeoylquinic acids were found to be characteristic markers for Northwest and East (Harar) region coffees, respectively. Sub-regional coffee types from West, except Jimma B, could be distinguished by their 3,5-O-dicaffeoylquinic to 4,5-O-dicaffeoylquinic acid concentration ratios, while Yirgachefe coffees from South could be distinguished by their 4,5-O-dicaffeoylquinic to 3,4-O-dicaffeoylquinic acid concentration ratios. Linear discriminant analysis provided a classification model with recognition and prediction abilities of 91% and 90%, respectively, at regional level, and 89% and 86%, respectively, at sub-regional level. This is important for the detection of fraud, including the selling of inferior Ethiopian coffees under the label of the more expensive Harar coffees.