Authentication of the geographical origin of virgin olive oils from the main worldwide producing countries: A new combination of HS-SPME-GC-MS analysis of volatile compounds and chemometrics applied to 1217 samples. (June 2020)

• Record Type: Journal Article
• Authors: Cecchi, Lorenzo; Migliorini, Marzia; Giambanelli, Elisa; Rossetti, Adolfo; Cane, Anna; Mulinacci, Nadia; Melani, Fabrizio
• Is Part Of: Food control. Volume 112(2020)
• Publication Date: 2020-06
• Subjects: Volatile compounds -- LDA -- Authentication -- Geographical origin -- Olea europaea L. -- Frauds; Food -- Quality -- Periodicals; Food -- Analysis -- Periodicals; Food handling -- Periodicals; Food industry and trade -- Quality control -- Periodicals; Aliments -- Industrie et commerce -- Qualité -- Contrôle -- Périodiques; Aliments -- Qualité -- Périodiques; Aliments -- Analyse -- Périodiques; Hygiène alimentaire -- Périodiques; Food -- Analysis; Food handling; Food -- Quality; Periodicals; Electronic journals; 664.07
• DOI: 10.1016/j.foodcont.2020.107156
• Languages: English
• ISSNs: 0956-7135

Abstract

Abstract: Authentication of geographical origin of virgin olive oils is necessary to protect consumer and producers from frauds. A method able to classify virgin olive oils from the main worldwide producing countries is still missing. In this work, we developed 3 chemometric approaches for classification of virgin olive oils from Italy, Spain, Greece, Portugal, Tunisia and other countries all over the world. The approaches were developed starting from a data-set containing fatty acid composition and the amount of 72 volatile compounds, evaluated by a never applied HS-SPME-GC-MS quantitation method, of 1217 oil samples from three different olive oil campaign. The approach that gave the best predictive results is based on Linear Discriminant Analysis run on quantitative data from only 25 volatile compounds selected by one-way ANOVA as the most capable in discriminating between the diverse origins. The method was built and internally validated using a training-set of 1000 samples and externally validated with a test-set of 217 independent samples. The method was able to classify the geographical origin of 94.5% samples, with a percentage of correct classification even higher than 97% for some origins. Preliminary studies also suggested the proposed approach is able to correctly classify the geographical origin of binary mixtures of oils from different origins. The approach proposed in this manuscript is easily applicable in testing laboratories and represents a very useful tool for the olive oil field, helping in protecting consumers and producers from frauds. Highlights: 1217 virgin olive oils from the main producing countries analyzed by HS-SPME-GC-MS. Three chemometric approaches for authentication of geographic origin were proposed. The ANOVA-LDA model gave 87.3% of correct prediction only using 25 VOCs. For some origins, the model correctly predicted more than 97% of samples. The model correctly classifies the geographic origin of binary mixtures of VOOs.