Cultivar discrimination and prediction of mixtures of Tunisian extra virgin olive oils by FTIR

Abstract

Fourier‐transform infrared spectroscopy, followed by multivariate treatment of the spectral data, was used to classify Tunisian extra virgin olive oils (EVOOs) according to their cultivar. Moreover, these data were also employed to establish the composition of binary mixtures of EVOOs from different cultivars. For this purpose, the spectra were divided in 20 regions, being the normalized peak areas within these regions used as predictors. Using linear discriminant analysis, an excellent resolution between EVOOs from different cultivar was obtained. Moreover, multiple linear regression models were used to predict the composition of binary mixtures of EVOOs, being in all cases capable of predicting the percentage of one of the oils with average validation errors below 6%. Thus, the FTIR method proposed in this work, followed by chemometric analysis, could be used in an industrial setting since no complicated laboratory facilities are required, which could save great deal of time and money. Practical applications: EVOO quality can be affected by different parameters, such as cultivar and maturity index, among others, consequently, it is necessary to develop analytical methods able to verify EVOO quality, and also to control its origin since olive oil producers have to include in their manufactured products the cultivar (monovarietal oils) from which the oil was obtained. Currently, almost all the research related to Tunisian EVOOs has been mainly focused on the improvement and characterization of the two main cultivars (Chétoui and Chemlali). Thus, to diversify Tunisian olive oil resources and improve the quality of olive oil produced in Tunisia, research on additional cultivars needs to be conducted. Thus, the present methodology (including the application of chemometric techniques) could be used as an authentication tool for olive oil industry to assess cultivar of EVOO. ATR‐FTIR data of Tunisian EVOOs are used to predict EVOO cultivar using LDA.

Application of an electronic tongue for Tunisian olive oils’ classification according to olive cultivar or physicochemical parameters

“…The results showed that olive oils could be correctly classified (100%) according to olive cultivar based on linear discriminant analysis (LDA) models. These same authors [13] also demonstrated that attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectroscopy could be used to discriminate Tunisian EVOO according to olive cultivar as well as to predict olive oil mixtures, using LDA and multiple linear regression (MLR) models. The results showed that the FTIR spectra with a stepwise LDA allowed establishing a multivariate classification model that could correctly classify all EVOO according to the olive cultivar (sensitivity of 100% for both training evaluation sets).…”

A multivariate data approach for FTIR-ATR monitoring of virgin olive oils: Lipid structure during autoxidation

“…Evaluation of spectral data requires some of software application with prediction or validation models. FTIR spectroscopy was used by several olive oil researchers for classification (Jimenez-Carvelo et al, 2017), monitoring of adulteration (Rohman et al, 2017), characterisation of cultivars (Abdallah et al, 2016) and geographic origin (Hennessy et al, 2009). Different data processing techniques were applied in these previous studies (Sinelli et al, 2007;Maggio et al, 2009;Lerma-García et al, 2011).…”

Mid-Infrared Spectroscopy as a Valuable Tool to Tackle Food Analysis: A Literature Review on Coffee, Dairies, Honey, Olive Oil and Wine

“…FT-MIR has also been broadly used to evaluate the quality and authenticity of olive oils [ 125 , 126 ], as well as to determine its geographic and varietal origin [ 127 , 128 , 129 , 130 ]. In this context, several chemometric models have been developed taking in account the fingerprint of different chemical compounds such as triglycerides, fatty acids or steroids [ 106 ].…”

“…Correct classification of samples were obtained for NIR and combined MIR and NIR techniques (higher than 90%), in comparison with e-nose technique (82%) that is suggested to be used as a complementary method to human sensory analysis [ 127 ]. A total of fifty-five EVOOs from seven Tunisian cultivars were successfully discriminated using MIR and linear discriminant analysis [ 128 ]. Using 20 spectral regions, mainly corresponding to vibrations associated with C–H, C–O, C=C (aromatic), and =C–H groups all the samples were correctly classified.…”

“…Using 20 spectral regions, mainly corresponding to vibrations associated with C–H, C–O, C=C (aromatic), and =C–H groups all the samples were correctly classified. Moreover, multiple linear regression was used to detect mixtures of EVOOs from different cultivars with an average validation error below 6% [ 128 ].…”