Detection and Identification of Extra Virgin Olive Oil Adulteration by GC-MS Combined with Chemometrics

TLDR

This study proposes an analytical method to detect and identify adulteration of extra virgin olive oil with corn, peanut, rapeseed, and sunflower oils. The method evaluates 22 fatty acids and six key ratios, applying univariate and multivariate analyses, including PLS‑LDA and Monte Carlo tree classification. Univariate analysis identified specific fatty acids as markers for peanut and rapeseed adulteration, while PLS‑LDA achieved a 1 % detection limit and 90 % prediction accuracy, and a Monte Carlo tree correctly predicted adulterant type 85 % of the time.

Abstract

In this study, an analytical method for the detection and identification of extra virgin olive oil adulteration with four types of oils (corn, peanut, rapeseed, and sunflower oils) was proposed. The variables under evaluation included 22 fatty acids and 6 other significant parameters (the ratio of linoleic/linolenic acid, oleic/linoleic acid, total saturated fatty acids (SFAs), polyunsaturated fatty acids (PUFAs), monounsaturated fatty acids (MUFAs), MUFAs/PUFAs). Univariate analyses followed by multivariate analyses were applied to the adulteration investigation. As a result, the univariate analyses demonstrated that higher contents of eicosanoic acid, docosanoic acid, tetracosanoic acid, and SFAs were the peculiarities of peanut adulteration and higher levels of linolenic acid, 11-eicosenoic acid, erucic acid, and nervonic acid the characteristics of rapeseed adulteration. Then, PLS-LDA made the detection of adulteration effective with a 1% detection limit and 90% prediction ability; a Monte Carlo tree identified the type of adulteration with 85% prediction ability.