Quantcast

Discriminative capacities of infrared spectroscopy and e-nose on Turkish olive oils

Research paper by Olusola Samuel Jolayemi, Figen Tokatli, Susanna Buratti, Cristina Alamprese

Indexed on: 30 Oct '17Published on: 23 May '17Published in: European Food Research and Technology



Abstract

The potentials of Fourier transform (FT) near- (NIR) and mid-infrared (IR) spectroscopy, and electronic nose (e-nose) on varietal classification of Turkish olive oils were demonstrated. A total of 63 samples were analyzed, comprising Ayvalik, Memecik, and Erkence oils. Spectra were pretreated with standard normal variate and second derivative. Classification models were built with orthogonal partial least square-discriminant analysis (OPLS-DA), considering the single data sets and also the combined FT-NIR-IR spectra. OPLS-DA models were validated both by cross validation and external prediction. All the models gave good results, being the average correct classification percentages in prediction higher than 90% for spectroscopic data and equal to 82% for e-nose data. The combined FT-NIR-IR data set gave the best results in terms of coefficients of determination (0.95 and 0.67). Different e-nose sensors discriminated Ayvalik, Memecik, and Erkence oils, explaining their distinct aromatic profiles.