Elsevier

Journal of Chromatography B

Volume 910, 1 December 2012, Pages 71-77
Journal of Chromatography B

Quantification of blending of olive oils and edible vegetable oils by triacylglycerol fingerprint gas chromatography and chemometric tools

https://doi.org/10.1016/j.jchromb.2012.01.026Get rights and content

Abstract

A reliable procedure for the identification and quantification of the adulteration of olive oils in terms of blending with other vegetable oils (sunflower, corn, seeds, sesame and soya) has been developed. From the analytical viewpoint, the whole procedure relies only on the results of the determination of the triacylglycerol profile of the oils by high temperature gas chromatography–mass spectrometry. The chromatographic profiles were pre-treated (baseline correction, peak alignment using iCoshift algorithm and mean centering) before building the models. At first, a class-modeling approach, Soft Independent Modeling of Class Analogy (SIMCA) was used to identify the vegetable oil used blending. Successively, a separate calibration model for each kind of blending was built using Partial Least Square (PLS). The correlation coefficients of actual versus predicted concentrations resulting from multivariate calibration models were between 0.95 and 0.99. In addition, Genetic algorithms (GA–PLS), were used, as variable selection method, to improve the models which yielded R2 values higher than 0.90 for calibration set. This model had a better predictive ability than the PLS without feature selection. The results obtained showed the potential of this method and allowed quantification of blends of olive oil in the vegetable oils tested containing at least 10% of olive oil.

Highlights

► Reliable method for quantification of olive oils adulteration with vegetable oils. ► Determination triglycerides profile by HTGC–MS, use of all data points to get results. ► Chromatograms were pre-treated (baseline correction, peak alignment) before building models. ► Genetic algorithms were used as variable selection method to improve the models.

Introduction

Edible vegetable oils are a valuable component of a fully mature seed. They are mainly the mixtures of triacylglycerols (TAGs), with different concentration levels. The remaining nonglyceridic fraction consists of different compound classes such as hydrocarbons, tocopherols, phytosterols and sterol esters [1]. The vegetable seed or fruit from which the oil is extracted determine most of its characteristics and composition that also depends on several factors such as soil, climate, processing, harvesting and chemical process occurring during storage [2].

Among edible oils, olive oil (OO) shows important and outstanding characteristics due to its differentiated sensorial qualities (taste and flavor) and higher nutritional value which have been acknowledged internationally. Several health benefits associated with its consumption were initially observed among Mediterranean people and its dietary consumption is nowadays considered to provide many benefits to human health [3].

TAGs represent up to 95–98% (weight to weight – w/w) of vegetable oil composition and show a characteristic distribution. As a consequence, the addition of other edible vegetable oils to olive oils modifies TAG distribution and because of that, they are considered to be good fingerprints for adulteration detection purposes [4].

Companies have been taking advantage of selling OO blends at the same price as pure OO, obtaining important economic benefits. The adulterants used in blends are the ones with similar physical and chemical properties and usually they are cheaper and easy to obtain. In the case of OO this usually implies the dilution with less expensive oils or other inferior quality olive oils [5], [6]. Moreover, a lot of methods and limits were introduced into the International Olive Oil Council (IOOC) trade standard, into EC Regulation 2568/91 and into the Codex Alimentarius Standard for controlling product authenticity and quality. In addition, In the EU, requirements has being established in Regulation (EC) No. 29/2012, concerning commercialization and labeling of products which contain olive oil, blends of olive and other edible vegetable oils. The presence of olive oil higher than 50% has to be indicated on the label, but if the percentage is lower than 50% the name of olive oil cannot be used in the label [7].

However, the reasons for mixing olive oil with others are not only economical, but also nutritional. It is clear from the composition of vegetable oils, that no single oil, even olive oil, meet all the oil nutritional requirements of essential fatty acids and vitamins [8].

The interest of researchers in the authentication of vegetable oils has led to an improvement in the control of adulteration and to the development of analytical methods to establish compositional differences in olive oils blends [9]. An extensive literature, discussing the suitability of a wide assortment of analytical methods aimed at evaluating the authenticity and the presence of adulterants in OO, has been published [10], [11]. When such methods are applied in conjunction of chemometric tools, spectroscopic analytical techniques as NIR, MIR, Raman, NMR or MS, and sensor-based analytical techniques as electronic nose, have been frequently used. This techniques share as common feature that yield low-selective instrumental signals (instrumental fingerprints) which are very suitable for developing valid chemometric models for pattern recognition. However, there are not many studies about vegetable oils authentication, which use directly the raw analytical signal which come from the chromatographic instrument (chromatographic fingerprint) with multivariate statistical methods [12], [13]. In most cases, the chromatographic applications use derived information from the raw analytical signal provided by the instrument, such as peak areas or concentration profiles for the classification of edible vegetable oils and detection of their adulterations. Chemometric tools have been commonly applied for matching and discrimination, classification and prediction in assessing authenticity of vegetable oils [14], [15], [16]. Thus, GC and LC methods in combination with multivariate statistical techniques such as principal component analysis (PCA), discriminant analysis (DA), cluster analysis (CA), K-nearest neighbor, genetic algorithm (GA), partial least squares (PLS) [17] and artificial neural networks (ANN) have been applied successfully to classify and discriminate the oils [18], [19].

The official method of the International Olive Council (IOC) is based on the use of the reverse phase-liquid chromatography with a refractive index detector (HPLC-RID), to establish the difference between actual and theoretical content of TAGs with Equivalent Carbon Number 42 (ECN42) [20]. In the literature, the contribution in chromatography to the authentication of vegetable oils by quantifying different major and minor compounds have been reviewed by Aparicio and Aparicio-Ruiz [21]. Detection of adulterated oils based on TAG compositions by high temperature-gas chromatography (HT)-GC was studied previously by Park and Lee [1].

This study focuses on the quantification of olive oil in blends with vegetable oils using multivariate calibration. The TAGs profiles, measured by HTGC–MS systems, have been applied for the quantification of olive oils in blends with vegetable oils (sunflower, corn, seeds, sesame and soya) and considering the different categories of olive oil (extra virgin olive oil, virgin, olive oil and pomace) and varieties (picual, hojiblanca and arbequina) at several percentages (10–90%). Multivariate statistical analyses, such as SIMCA, PLS and GA–PLS were applied to achieve this purpose.

Section snippets

Samples

The olive oil samples, to build the blends, were fourteen, including four categories [22]: extra virgin (EVOO), virgin (VOO), olive oil (OO, blend of virgin and refined) and pomace oil (POO), and three Spanish olive fruit varieties named “arbequina” (ARB), “hojiblanca” (HOJ) and “picual” (PIC).

In addition, eleven vegetable oils samples were used: two sunflowers oils (SUN), one high-oleic sunflower oil (OSUN), two corn oils (COR), one sesame oil (SES), three soya oils (SOY), and two vegetable

Results and discussion

Fig. 2 shows the baseline corrected chromatograms of the whole data set before and after iCoshift. By inspecting the figure, one can see the large difference between the two plots and how the peak shifting was perfectly corrected with this algorithm. In a first stage, iCoshift was applied separately to the different kinds of blends and successively to the whole data set with all the samples of blends together.

The first 500 data points of each chromatogram were eliminated due to the lack of

Acknowledgments

The authors would like to thank to the Andalusia Regional Government (Consejería de Innovación, Ciencia y Empresa, Project P07-FQN-02667, and Consejería de Agricultura y Pesca) for financial assistance. This work has also been partially supported by European Regional Development Funds (ERDF). The authors are grateful to the Andalusia Regional Government (Consejería de Innovación, Ciencia y Empresa) for the personal postgraduate grant awarded to CRS.

References (31)

  • R. Aparicio et al.

    J. Chromatogr. A

    (2000)
  • A. Cert et al.

    J. Chromatogr. A

    (2000)
  • E. Christopoulou et al.

    Food Chem.

    (2004)
  • P. de la Mata Espinosa et al.

    Talanta

    (2011)
  • D. Brodnjak-Voncina et al.

    Chemometr. Intell. Lab. Syst.

    (2005)
  • F. Peña et al.

    J. Chromatogr. A

    (2005)
  • R.M. Maggio et al.

    Food Control

    (2010)
  • R. Aparicio et al.

    J. Chromatogr. A

    (2000)
  • C. Ruiz-Samblás et al.

    Talanta

    (2010)
  • F. Savorani et al.

    J. Magn. Reson.

    (2010)
  • F. Marini et al.

    Chemometr. Intell. Lab. Syst.

    (2006)
  • R. Leardi et al.

    Chemometr. Intell. Lab. Syst.

    (1998)
  • Z. Xiaobo et al.

    Anal. Chim. Acta

    (2010)
  • J.R. Park et al.

    Bull. Korean Chem. Soc.

    (2003)
  • J.M. Dyer et al.

    Plant J.

    (2008)
  • Cited by (74)

    • Organic electronic nose applied to food traceability, adulteration, and authenticity

      2023, Nanotechnology-Based E-Noses: Fundamentals and Emerging Applications
    • The chromatographic similarity profile – An innovative methodology to detect fraudulent blends of virgin olive oils

      2022, Journal of Chromatography A
      Citation Excerpt :

      Both the analysis of major components such as triglycerides or the fatty acids of which the triglycerides are composed, and other minority fractions such as phytosterols, volatile organic compounds or phenolic compounds, amongst many others, are carried out. Some of these families of compounds such as triglycerides [16], fatty acids methyl esters [17,18], phytosterols [19] or volatile compounds [17,20] can be determined by gas chromatography, being the main disadvantages of these methods: (i) the use of high temperatures of analysis, (ii) the long incubation times of the samples, or (iii) the need for previous sample pre-treatment as fractionation or derivatisation. Different types of detection systems are used, being the most interesting one those based on mass spectrometry.

    • Near infrared spectroscopy combined with chemometrics for quantitative analysis of corn oil in edible blend oil

      2022, Spectrochimica Acta - Part A: Molecular and Biomolecular Spectroscopy
    View all citing articles on Scopus

    This paper belongs to the Special Issue Chemometrics in Chromatography, Edited by Pedro Araujo and Bjørn Grung.

    View full text