We have developed an analytical method using UHPLC-UV/ESI-TOF MS for the comprehensive profiling of the metabolites found in the methanolic extracts of 13 different varieties of avocado at two different ripening degrees. Both chromatographic and detection parameters were optimized in order to maximize the number of compounds detected and the sensitivity. After achieving the optimum conditions, we performed a complete analytical validation of the method with respect to its linearity, sensitivity, precision, accuracy and possible matrix effects. The LODs ranged from 1.64 to 730.54 ppb (in negative polarity) for benzoic acid and chrysin, respectively, whilst they were found within the range from 0.51 to 310.23 ppb in positive polarity. The RSDs for repeatability test did not exceed 7.01% and the accuracy ranged from 97.2% to 102.0%. Our method was then applied to the analysis of real avocado samples and advanced data processing and multivariate statistical analysis (PCA, PLS-DA) were carried out to discriminate/classify the examined avocado varieties. About 200 compounds belonging to various structural classes were tentatively identified; we are certain about the identity of around 60 compounds, 20 of which have been quantified in terms of their own commercially available standard.
Copyright © 2011 Elsevier B.V. All rights reserved.