Discrimination of geographical origin of oranges (Citrus sinensis L. Osbeck) by Headspace Solid Phase Microextraction Mass spectrometry-based Electronic Nose (HS-SPME MS-eNose) and Chemometrics

Introduction: In the last years the Italian imports of oranges, mainly from Spain and South-Africa, have significantly increased, therefore there is a real possibility for the Italian consumer to buy foreign products sold fraudulently as Italian. For this reason, in this work an HS-SPME/MS-eNose method for the discrimination of the geographical origins of oranges was developed and validated. Methods: Oranges samples coming from Italy, Spain and South Africa were analyzed by an HS-SPME/MS-eNose method. Subsequently, three multivariate statistical models, i.e. PCA/LDA, SELECT/LDA and PLS-DA, were built for the geographical origin discrimination and the relevant performances were compared. Moreover, an HS-SPME/GC-MS method combined with ANOVA was used to identify discriminating compounds. Results: Although all tested statistical models gave acceptable performance, the SELECT/LDA model showed the highest percentages in terms of prediction ability in cross-validation and external validation, with average values of 97.8% and 95.7%, respectively. In particular, the external prediction ability of 95.7% was obtained with all South African and Spanish samples correctly recognized while only 2 samples out of 19 Italian samples were not correctly assigned, with a specific prediction rates of 89.5%. Moreover HS-SPME/GC-MS analysis showed that, although 28 out of 65 identified VOCs had a different content in samples belonging to different origin classes, no compound was able to discriminate at the same time the three geographical origins. Conclusions: In this study, a rapid and inexpensive method based on MS-eNose analysis in combination with chemometrics was successfully used to discriminate oranges coming from Italy, Spain and South Africa. Although HS-SPME/GC-MS analysis showed the absence of specific markers, differences in the pattern and content of VOCs of orange samples of the three different geographical origins were observed.