FT-IR extra virgin olive oil classification based on ethyl ester content
Introduction
Extra-virgin olive oil (EVOO) is a premium vegetable oil obtained from fresh fruits only by means of physical and mechanical processes (Council Regulation (EC) No 1513/2001). It has a great market value due to its appreciated features. Over the years, several European Regulations have modified the quality and purity characteristics of virgin olive oils (VOO) for their commercial classification and labelling. In 2011, the European Commission introduced a limit to the content of fatty acid alkyl esters (FAAE) in extra virgin olive oils (Commission Regulation (EU) No 61/2011). According to the mentioned Regulation, a virgin olive oil labelled as EVOO must contain a maximum of 75 mg kg−1 for the sum of ethyl esters (FAEE) and fatty acid methyl esters (FAME) or their sum could be between 75 and 150 mg kg−1 in case their ratio (FAEE/FAME) is guaranteed to be ≤ 1.5.
FAAE are neutral lipids originating from the esterification of free low-weight alcohols with free fatty acids. Among involved alcohols, the most important are methanol and ethanol, yielding to FAME and FAEE, respectively. Among fatty acids, palmitic and oleic acids are the most common. Methanol and ethanol rise respectively from the progressive degradation of drupe cell walls and from fermentation processes mainly occurring during olive and/or oil storage in improper conditions. Fatty acids, instead, are commonly found in VOO to some extent, depending on the maturation stage of olives but, mainly, on their quality and integrity. FAAE content could be also affected by the extraction process (Alcala et al., 2017, Caponio et al., 2018, Squeo et al., 2017). The Commission Regulation (EU) No 61/2011 has been finally modified by the Commission Delegated Regulation (EU) 2016/2095 focusing only on the FAEE content and setting the maximum value for EVOO at 35 mg kg−1 of oil.
According to the official method (Commission Regulation (EU) No 61/2011), FAAE determination requires their separation from triacylglycerols and other oil constituents by chromatography on a hydrated silica gel column using Sudan 1 (1-phenylazo-2-naphthol) as indicator for the elution. Then, the FAAE fraction is collected, dried and re-suspended in n-heptane or iso-octane. Alkyl esters are finally separated by capillary gas-chromatography. Quantification is achieved by the addition of a proper internal standard. Overall, around 5 h are needed to complete the analysis, without considering the preparation steps such as silica conditioning. Besides, a large amount of organic solvents is used for the determination, mainly n-hexane. Indeed, approximately 250 mL are required for one determination. Considering that the analysis should be performed at least in duplicate in order to obtain reliable results, around 1 L of solvent is required for the analysis of only two oil samples.
Hexane is toxic for humans as well as for the environment as extensively reported in hexane safety data sheet, according to the Regulation (EC) No 1272/2008. Thus, the possibility to significantly reduce the use of this solvent, together with the other organic solvents required for the analysis (diethyl ether, isoctane), matches the sustainability goals desired from Institutions all over the world (United Nations, 2016). The possibility of overcoming these issues (time-consuming analysis, health and environment hazards) lies in green approaches like the use of spectroscopic techniques that is one of the most promising.
Spectroscopic techniques are non-destructive, green, fast and easy to use. Among them, mid-infrared (MIR) spectroscopy is one of the most used, having an illustrious history in lipid chemistry, and it has experienced growing interest and applications thanks to the introduction of the Fourier transform instruments (FT-IR) (Dobson, 2001). The MIR range goes from around 2.5 to 25 μm or, as most commonly reported, from 4000 to 400 cm−1. Absorption of a MIR photon typically excites one of the fundamental vibrations, associated with a change of the dipole moment of an oscillating molecule (Sikorska, Khmelinskii, & Sikorski, 2014). Despite the complexity of spectra collected along the food systems, the association of MIR spectroscopy with chemometrics allows the extraction of the significant and valuable information (Gómez-Caravaca, Maggio, & Cerretani, 2016). Indeed, when spectra are recorded from real food samples, they contain information about different components of the sample matrix together with their interactions, and multivariate methods are successfully used in interpreting the spectra signals for analytical purposes (Bro, 2003, Kjeldahl and Bro, 2010, Sikorska et al., 2014). Several chemometric approaches might be used, falling in two main classes: qualitative and quantitative methods. As regard to alkyl esters, in a previous study by Valli et al. (2013), Partial Least Square (PLS) regression models were tentatively developed for the quantification of FAAE based on VOO FT-IR spectra. However, they were aimed at the quantification of the sum of ethyl and methyl esters as well as their ratio, parameters that are no longer considered for the EVOO classification.
After the introduction of the Commission Delegated Regulation (EU) 2016/2095, few authors have taken interest in FAAE determination by green methods. Indeed, near infrared (NIR) and Vis-NIR spectroscopy has been used to develop regression models for measuring total FAAE content, as well as FAEE and FAME content separately (Cayuela, 2017, Garrido-Varo et al., 2017). However, even though chemometric approaches can overcome the overlapping NIR signals resulting from first and second overtones and combinations of the fundamental vibrations, more accurate assignments of absorption bands can be reached by MIR spectroscopy. This is particularly relevant when assessing differences among molecules having similar bonds that scatter in a complex matrix such as oil. As far as we know, despite the importance of FAEE, no other attempts have been carried out by IR spectroscopy to develop a rapid procedure for their analysis. Starting from these considerations, the aim of this work was the application of IR spectroscopy to the development of classification models (based on Linear Discriminant Analysis and Soft Independent Modelling of Class Analogy) able to discriminate between EVOO and non-EVOO based on FAEE content. Though FAEE is a continuous variable, a classification approach was chosen instead of quantification since, by a practical point of view, the proposed method should address a discrimination issue regarding the authentication of EVOO, where authentication is intended as the compliance of a food with its label description (Danezis, Tsagkaris, Camin, Brusic, & Georgiou, 2016). A similar approach based on discriminant classification techniques has already been applied in the literature in order to develop fast sorting tests for olive oils, based on the content of α-tocopherol or squalene (Cayuela and García, 2017, Cayuela and García, 2018). Supervised classification techniques use the information about the known class membership of training samples in order to create classification rules able to assign new unknown samples to one of the defined classes, based on their fingerprint measurement (Berrueta, Alonso-Salces, & Héberger, 2007). Thus, these chemometric techniques perfectly fit in authentication issues where the goal is to verify if a sample belongs to a predefined class, such in the case of EVOO and non-EVOO differentiated by the FAEE content.
Section snippets
Sampling
A set of 159 VOO (113 extra virgin and 46 virgin) from Apulia region (southeast Italy) were collected during 2016/17 and 2017/18 production seasons directly from olive mills located in different provinces (i.e., 100 samples from Bari province; 15 from Brindisi province; 13 from Barletta-Andria-Trani province; 11 from Foggia province; 11 from Lecce province; 9 from Taranto province). All samples were bulk oils, blends of the principal Apulian olive cultivars, all extracted by continuous plants
Fatty acid ethyl ester content of olive oil samples
Table 1 reports the descriptive statistics of the olive oil samples, divided by classes: EVOO (class 1), non-EVOO (class 2). The number of objects was quite different between the two classes since no previous information about the amount of FAEE in the collected samples were provided by producers. In any case, this situation is representative of a real scenario, because during the authentication of EVOO a low number of non-EVOO should be expected.
Class 2 (non-EVOO), although made up of a lower
Conclusions
The potential of FT-IR spectroscopy coupled with chemometrics as a tool for a rapid and fast discrimination of extra virgin olive oils and virgin olive oils based on fatty acid ethyl ester content has been assessed. The legal limit for FAEE content enables to highlight two well-defined classes of products, thus a discriminant classification approach can be considered as the most suitable in this context, to be preferred over the SIMCA class-modelling algorithm. LDA models based on selected
Acknowledgements
This work was supported by AGER 2 Project, grant no. 2016-0105. The authors declare that there are no conflicts of interest regarding the publication of this paper. All the authors have contributed to the same extent to the realization of this paper.
References (31)
- et al.
Supervised pattern recognition in food analysis
Journal of Chromatography A
(2007) Multivariate calibration: What is in chemometrics for the analytical chemistry?
Analytica Chimica Acta
(2003)- et al.
Sorting olive oil based on alpha-tocopherol and total tocopherol content using near-infra-red spectroscopy (NIRS) analysis
Journal of Food Engineering
(2017) - et al.
Nondestructive measurement of squalene in olive oil by near infrared spectroscopy
Lebensmittel-Wissenschaft und -Technologie- Food Science and Technology
(2018) - et al.
Food authentication: Techniques, trends & emerging approaches
TRAC Trends in Analytical Chemistry
(2016) - et al.
Confirmation of brand identity in foods by near infrared transflectance spectroscopy using classification and class-modelling chemometric techniques - the example of a Belgian beer
Food Research International
(2011) - et al.
Stepwise orthogonalization of predictors in classification and regression techniques: An “old” technique revisited
Chemometrics and Intelligent Laboratory Systems
(2007) - et al.
Chemometric applications to assess quality and critical parameters of virgin and extra-virgin olive oil. A review
Analytica Chimica Acta
(2016) - et al.
Handheld NIR device: A non-targeted approach to assess authenticity of fish fillets and patties
Food Chemistry
(2018) - et al.
An orthogonal feature selection method
Pattern Recognition
(1976)
Application of class-modelling techniques to near infrared data for food authentication purposes
Food Chemistry
Applications of Fourier transform-infrared spectroscopy to edible oils
Analitica Chimica Acta
Alkyl esters content and other quality parameters in oil mill: A response surface methodology study
European Journal of Lipid Science and Technology
Calcium carbonate effect on alkyl esters and enzymatic activities during olive processing
Italian Journal of Food Science
Rapid NIR determination of alkyl esters in virgin olive oil
Grasas Y Aceites
Cited by (22)
Application of NIR spectroscopy coupled with DD-SIMCA class modelling for the authentication of pork meat
2023, Journal of Food Composition and AnalysisA green and direct method for authentication of rosewood essential oil by handheld near infrared spectrometer and one-class classification modeling
2022, Microchemical JournalCitation Excerpt :Therefore, NIRS technology (NIRS associated with chemometrics) has been recognized as a fast, effective, non-destructive, and cost-effective technique for the authentication and determination of the purity of various products. Among chemometric exploratory analysis techniques applied in solving problems of adulteration or oils identification, those that stand out are Principal Component Analysis (PCA) [33,34], Linear Discriminant Analysis (LDA) [35], Hierarchical Cluster Analysis (HCA) [36], as well as supervised chemometric methods SIMCA [33,35], Discriminant Analysis by Partial Least Square (PLS-DA) [34] and PLSR [37,38]. Taking into account the growing concern with the preservation of this native Brazilian forest species, the requirements of industry for a high purity oil, and all restrictions of Brazilian regulatory agencies and CITES, this work proposes a simple, fast and direct method for the identification and authentication of the purity of commercial essential oil reported as A. rosiodora, originating from different sources, based on NIRS technology.
Assessment of macronutrients and alpha-galactosides of texturized vegetable proteins by near infrared hyperspectral imaging
2022, Journal of Food Composition and AnalysisCitation Excerpt :On the opposite, rapid, environmentally friendly, and non-destructive methods of analysis are available, which are also suitable for on-line or at-line applications. These are based on spectroscopic techniques, among which infrared spectroscopy has been proven to be an efficient technique for food quality control (Cen and He, 2007; Sinelli et al., 2008; Squeo et al., 2019; Woodcock et al., 2008). The potential of infrared spectroscopy, specifically near infrared (NIR), has been enhanced by coupling it with imaging techniques.
Metabolic fingerprinting strategy: Investigation of markers for the detection of extra virgin olive oil adulteration with soft-deodorized olive oils
2022, Food ControlCitation Excerpt :High price together with quantitative expansion of high-quality olive oil consumption makes olive oil vulnerable for economically motivated adulteration practices (Bajoub et al., 2018; Casadei et al., 2021). Fraudulent practices involve misdescription of botanical origin (Carranco, Farrés-Cebrián, Saurina, & Núñez, 2018; Filoda et al., 2019) and/or geographical origin (Bontempo et al., 2019; Gertz, Gertz, Matthäus, & Willenberg, 2019); addition of cheaper olive oil (e.g. refined olive oil, ROO) into higher grade olive oil (Squeo, Grassi, Paradiso, Alamprese, & Caponio, 2019; Yan, Oey, Van Leeuwen, & Van Ruth, 2018) is another common way of adulteration. In 2016, the FoodIntegrity project (www.foodintegrity.eu) launched a survey addressing all olive oil actors (producers, wholesalers, retailers, researchers, analysts, etc.) to find out their priorities in olive oil authentication issues.