A taste sensor device for unmasking admixing of rancid or winey-vinegary olive oil to extra virgin olive oil

• Autor(a) principal: Harzalli, Ussama
• Data de Publicação: 2018
• Outros Autores: Rodrigues, Nuno, Veloso, Ana C. A., Dias, Luís G., Pereira, José A., Oueslati, Souheib, Peres, António M.
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/1822/48443
• topic: Electronic tongue; Chemometrics; Olive oil; Physicochemical analysis; Sensory analysis; Adulteration detection; Science & Technology

Electrochemical sensor devices have gathered great attention in food analysis namely for olive oil evaluation. The adulteration of extra-virgin olive oil with lower-grade olive oil is a common worldwide fraudulent practice, which detection is a challenging task. The potentiometric fingerprints recorded by lipid polymeric sensor membranes of an electronic tongue, together with linear discriminant analysis and simulated annealing meta-heuristic algorithm, enabled the detection of extra-virgin olive oil adulterated with olive oil for which an intense sensory defect could be perceived, specifically rancid or winey-vinegary negative sensations. The homemade designed taste device allowed the identification of admixing of extra-virgin olive oil with more than 2.5% or 5% of rancid or winey-vinegary olive oil, respectively. Predictive mean sensitivities of 84±4% or 92±4% and specificities of 79±6% or 93±3% were obtained for rancid or winey-vinegary adulterations, respectively, regarding an internal-validation procedure based on a repeated K-fold cross-validation variant (4 folds×10 repeats, ensuring that the dataset was forty times randomly split into 4 folds, leaving 25% of the data for validation purposes). This performance was satisfactory since, according to the legal physicochemical and sensory analysis, the intentionally adulterated olive oil with percentages of 2.510%, could still be commercialized as virgin olive oil. It could also be concluded that at a 5% significance level, the trained panelists could not distinguish extra-virgin olive oil samples from those adulterated with 2.5% of rancid olive oil or up to 5% of winey-vinegary olive oil. Thus, the electronic tongue proposed in this study can be foreseen as a practical and powerful tool to detect this kind of worldwide common fraudulent practice of high quality olive oil.