Abstract
The use of olive oil with cooking purposes, as final seasoning or within cooked foods is increasing worldwide due to its numerous nutritional and health benefits. These attributes are mainly determined by olive oil chemical composition, which can be altered after thermal processing, oxidation processes, or incorrect practices. For this reason, and due to the numerous factors which have influence in olive oil quality, the correct chemical characterization is highly relevant. In this study, fatty acid composition of four extra virgin olive oil (EVOO) varieties was studied. The major fatty acid (FA) determined was oleic acid (77.1% on average), followed by palmitic (11.5% on average). In addition, thermal oxidation behaviour of the four EVOO samples was studied as an indicator of their quality and stability during thermal processing. This was performed through differential scanning calorimetry (DSC) from a temperature of 40°C at six different heating rates in the range of 0.5–10°C min−1. DSC records showed the same pattern and a small shoulder in the thermo-oxidation peak was present for all samples and all heating rates. The presence of initial and final oxidation products (by monitoring K232 and K270 values, respectively) was discarded according to the International Olive Council method.
1. Introduction
Nowadays, 85% of the total fats consumed in the Mediterranean diet comes from olive oil, a vegetable oil whose consumption is associated with several health benefits such as lower incidence of cardiovascular diseases, cancer, and increased longevity [1]. Most attributes of olive oil quality are determined by its chemical composition as well as the biochemical status of the olive fruit. For the production of high-quality oil, the olives must be harvested without breaking the skins, and they must be processed within 12–24 hours of harvest [2]. Extraction must be made from healthy fruits, avoiding manipulation or treatments which could alter the chemical composition of olive oil during the extraction and storage process [3]. In addition to olive picking, storage, and processing, olive oil composition is determined by olive tree cultivation, climate, geographical area, etc. [2]. This makes every batch unique and difficult to standardize experimental conditions [4].
The group of major compounds in olive oil composition is triglycerides, which constitute between 92 and 98%. It also contains fatty acids, which contribute 94–96% of the total weight of triglycerides. In this fraction, six are major compounds: oleic (55.2–86.6%), palmitic (6.30–20.9%), linoleic (2.7–20.2%), stearic (0.32–5.33%), palmitoleic (0.32–3.52%), and linolenic (0.11–1.52%). Olive oil is also composed by minor components, fraction constituted by compounds, which derive from triglycerides and liposoluble compounds. This minority fraction can be grouped in the following: diacylglycerols (DAGs), monoacylglycerols (MAGs), free fatty acids (FFAs), oxygenated fatty acids (OFAs), cyclic fatty acids, nonlinear FAs (branched FAs), dimeric FAs, and another compounds such phenols and pigments. The total of these compounds represents between 2 and 5% of the total composition [1].
Olive oil is commonly used as final seasoning, but it is also used with cooking purposes at high temperatures. In this sense, after thermal processing, changes and degradation processes are expected in olive oil; the most usual changes consist of triglyceride polymerization and hydrolysis, fatty acid and sterol oxidation, and Maillard reactions [4]. Oxidation can also alter the flavour and nutritional quality of olive oil due to the loss of beneficial substances and the generation of new toxic compounds including oxidized fatty acids, sterols or TAG polymers, which can have a possible impact on human health and make olive oil less acceptable or unacceptable to consumers [5]. In this sense, differential scanning calorimetry (DSC) is a technique based on the measurement of the energy changes that take place when a sample is heated, cooled, or held isothermally, as well as the determination of the temperature at which these changes occur. These measurements enable the characterization of samples for several complex events such as melting processes or glass transitions [6]. Although DSC has not been established by the International Olive Council as an official method for the determination quality, variety, and geographical origin of olive oil, it has been suggested as a possible method with the advantages of being a fast and easy technique without the necessity of sample pretreatment or use of solvents [7, 8]. According to the official definition, extra virgin olive oil must be extracted by cold and mechanic conditions in an oxygen-free atmosphere in order to preserve the naturally present antioxidants. In refined olive oil, antioxidants are degraded due to refining processes and high temperatures during the olive oil production; as a consequence, the induction period is shorter in lower quality olive oils and can be used to study and compare the thermo-oxidative stability of samples [9]. In this sense, the oxidation of edible oils exhibits the induction period, and at the end of the induction period, the quality of the oil suddenly deteriorates so that the induction period is considered as a measurement of the oil stability [10].
In addition to DSC, spectroscopic techniques are suitable for quality control of olive oil. Fluorescence spectroscopy is a simple, rapid, economic, and nondestructive technique which is applied to determine the stage of decomposition of oils [11]. The K232 and K270 values are spectrophotometric measures for quantifying the UV absorption at 232 nm and 270 nm, respectively. It provides information about the quality of the fat, the conservation status of the oil, and any deterioration occurred during the technological processes [2]. It corresponds to the maximum absorption of the conjugated dienes and trienes, and it is expressed as specific extinctions coefficients [12].
Other technique that can be found in the literature is “Rancimat stability” which consists of exposing the olive oil to forced oxidation at 100°C until its maximum oxidation, measuring the time required for an abrupt change in conductivity from an aqueous solution where the volatile compounds carried by the oil were collected. The duration time of this period is considered as the index of resistance to rancidness of the fat being assayed [13].
In this work, the quality and stability of different varieties of olive oil were studied. The fatty acid profiles of four commercial EVOO were determined. The thermal oxidation stability and the kinetic parameters related to the oxidation process by DSC were evaluated. The specific UV extinction coefficients (K232 and K270) were determined to study the presence of oxidation products.
2. Materials and Methods
2.1. Samples
Four extra virgin olive oils samples of different brands were bought in a local store in Spain (Table 1). The samples were kept in a refrigerator at 4°C until the time of analysis.
2.2. Fatty Acid Profiles Determination
A mass between 0.10 and 0.30 g of each sample was weighted and dissolved in heptane in a reaction vessel with volume capacity equal to 1 cm3. After the sample dilution, 100 μl of sodium methoxide, the transesterification agent, was added. The time of the transesterification reaction had a duration between 15 and 20 minutes. Then, an excess of methanolic HCl (typically 100 μl) was added and the reaction was carried out at room temperature for 45 minutes. The upper heptane layer was separated and injected into the gas chromatograph [14].
Fatty acid composition was determined by the gas chromatograph GC-7890 (Agilent, USA) with a FID detector and capillary column DB-23 (60 m × 0.25 mm, with 0.25 μm stationary phase of poly(cyanopropylmethyl siloxane)). A volume of 1 μL of FAME and heptane was injected. Carrier gas flow rate was equal to 16.4 cm3 min−1 and pressure = 220 kPa. Programming chromatographic temperature was set at the initial value of 150°C (held for 6 min), followed by a heating rate of 5°C min−1 up to 170°C and heating rate of 6°C min−1 up to 220°C (held for 6 min). Next stage was a heating rate of 6°C min−1 at 220°C for 1 min and finally, heating rate of 30°C min−1 up to 240°C for 10 minutes. FID hydrogen flow and airflow rate were 40 cm3 min−1 and 450 cm3 min−1, respectively.
2.3. Differential Scanning Calorimetry
The DSC analysis was conducted on a differential scanning calorimeter, Shimadzu DSC-60 (Tokyo, Japan) equipped with an automatic gas switching unit. The temperature scale of the instrument was calibrated to the melting points of enzyl, In, Sn, and Pb. The measurement of thermo-oxidative stability was carried out in nonisothermal mode with linear heating. Samples of 3.5–4.5 mg were placed into open aluminium pans and heated in dynamic air atmosphere (50 mL min−1) from 40°C at 6 different heating rates in the range of 0.5–10°C min−1. Each measurement was terminated once an exothermic peak corresponding to thermal oxidation was observed.
2.4. Determination of Specific UV Extinction Coefficients (K232 and K270)
The measurement was performed through UV/VIS spectrophotometry with a UV-1600 series spectrophotometer (VWR, Leuven, Belgium). Absorbance within a 200 to 800 nm spectral range was measured at 1 nm spectral resolution using a 1 cm path length quartz cell, in the region of 200–380 nm.
Olive oil samples were perfectly homogeneous without any suspended impurities. A mass of 0.25–0.30 g was weighted and diluted to a one percent solution in cyclohexane. Spectrophotometric analysis of olive oil was performed in accordance with the official method in the Commission Regulation (EC) [15], which involves the determination of the specific extinction in cyclohexane at wavelength of 232 and 270 nm and the determination of K232 and K270 according to the following equation:where is the extinction coefficient, is the absorbance, is the concentration of the sample in the solvent in g/100 mL, and is the path length of the cuvette in cm.
3. Results and Discussion
3.1. Fatty Acids Composition of Extra Virgin Olive Oils
The fatty acid (FA) profile of olive oil is highly relevant, and it is considered as a parameter to characterize the diverse olive varieties since the quality of the fat has a direct impact on oil quality and thus, on consumer health [16]. In addition to the clinical relevance and the nutritional value of some FA such as oleic acid, FAs are also responsible for the presence of desired and undesired volatile compounds, which have a direct influence on the positive or negative sensory perceptions in olive oil. Lipoxygenase (LOX) pathways generate most of the desired volatile aroma compounds (C5 and C6 compounds and saturated aldehydes). A series of oxidative reactions result in a large variety of metabolites from polyunsaturated FA, linoleic and linolenic acids being the main initial substrates. The importance of the FA profile is, therefore, due to the fact that high and poor quality olive oils differ by their content in these compounds derived from FA [17].
Fatty acid content of olive oils is highly variable since it is affected by numerous factors such as production and cultivation area, latitude, climate, fruit ripeness, genetic factors, etc. Environmental factors are the ones that have a greater influence on FA composition of olive oils, temperature being the one that plays an essential role in the FA profile of olive oil, since temperature regulates fatty acid desaturases. Polyunsaturated fatty acids are present in greater proportions at low temperatures [18]. In this sense, differences in the FA profile of the four studied EVOO can be explained by the different geographical areas and climate conditions in which olive fruits were grown. In addition, several agronomic, processing, and environmental variables such as degree of ripeness or storage and processing conditions have a direct influence on the olive oil chemical composition [19].
Table 2 shows the fatty acid profile (%, weight) of the different EVOO. Determined fatty acids have been grouped as total saturated (SFA), monounsaturated (MUFA), and polyunsaturated (PUFA) fatty acids. The major fatty acid percentage found was oleic acid (C18 : 1) as expected. This fatty acid content ranged from 75.2% (Ar) to 79.9% (Ko), followed by palmitic acid (C16 : 0) which ranged from 10.4% (Ko) to 12.9% (Ar), linoleic acid (C18 : 2), from 5.09% (Ko) to 8.27% (Ar), stearic acid (C18 : 0), which ranged from 1.85% in Ar to 2.08% in C + A, and linolenic acid (C18 : 3) whose content ranged from 0.59% in Ar to 2.82% in C + A. Other fatty acids such as palmitoleic acid (C16 : 1, 0.86% on average), gadoleic acid (C20 : 1; 1.24% on average), behenic acid (C22 : 0; 0.50% on average), and arachidic acid (C20 : 0; 0.27% on average) were detected in all EVOO samples and found at a concentration of less than 1%. In general, no significant variation was detected in the fatty acids composition of the different EVOO studied, showed by the standard deviation values, which varied from 0.10 (C20 : 0) to 2.23 (C18 : 1).
Saturated fatty acids comprised about 13.6% of the total fatty acids, whereas monounsaturated and polyunsaturated fatty acids represented 77.4% and 8.98%, respectively. Total unsaturated fatty acids (MUFA + PUFA) in olive oil constituted 86.4% of the total. These fractions corresponded, almost entirely, to oleic acid, while palmitic acid represented the greatest proportion of SFA.
Regarding FA composition, significant differences exist between olive oil and other vegetable oils. In this sense, Li et al. [20] determined the fatty acid profile of palm oil, rapeseed oil, sunflower oil, and linseed oil. Compared to these four vegetables oils, it must be highlighted the higher oleic acid content in the four EVOO studied in this work (77.1% in average) in comparison with rapeseed, palm, sunflower, and linseed oil, whose content in oleic acid was notably lower: 46.3%, 33.6%, 13.6%, and 1.2%, respectively. In addition, palmitic acid, the second most abundant FA in olive oil (11.5% on average), was found in notably lower percentages in sunflower oil (3.89%), linseed oil (3.12%), and rapeseed oil (2.69%), nevertheless, higher content of this FA was found in palm oil (29.3%) in comparison with EVOO. Content of linoleic and stearic acids in EVOO (6.44% and 1.99% on average, respectively) were lower in comparison with the other vegetable oils, whose content ranged from 8.12% (palm oil) to 51.9% (sunflower oil) for linoleic acid and between 1.51% (rapeseed oil) and 3.59% (palm oil) for stearic acid. Linolenic acid was only found in rapeseed and linseed oil, at a concentration of less than 1%. Myristic acid (C14 : 0), which was not found in olive oil, was found at 0.43% in palm oil.
Similarly, Berasategi et al. [21] studied avocado oil fatty acid composition. This oil consumption and production is significantly growing in recent years due to its beneficial health properties attributed to its high concentration of oleic acid, antioxidant vitamins, and phytosterols. This study showed that MUFA content in avocado oil was equal to 68.4% with a total content of 54.4% of oleic acid of total FA. These values are much lower in comparison with the EVOO studied in this work, which contained 78.2% on average of MUFA and oleic acid ranging from 75.2% to 79.9%. On the contrary, palmitoleic acid, whose average content in EVOO was equal to 0.86%, was found at higher concentration (7.88%) in avocado oil. The importance of MUFA content can be explained by its relation with higher concentration of minor compounds with antioxidant and hypocholesterolemic effects [21].
On the other hand, higher PUFA content was found in avocado oil (11.8%) in comparison with EVOO (7.73%). Within this group, EVOO contained 2-fold the amount of linolenic acid present in avocado oil (0.61%). Lastly, SFA content in avocado was equal to 11.8% in comparison with 7.73% in EVOO and with the main differences in palmitic and stearic acids, whose contents were equal to 18.7 and 0.51%, respectively.
3.2. Differential Scanning Calorimetry
The standard tests used for the determination of the induction period are predominantly carried out under isothermal conditions, i.e., the oxidation induction time is measured. However, under isothermal conditions, the oxidation peak measured is often flat and its onset, corresponding to the end of induction period, cannot be determined unambiguously. On the contrary, in the experiments with constant heating rate, the oxidation peak is distinct and the onset oxidation temperature can be measured accurately and unambiguously. In our previous work, a theory of the kinetic description of induction periods from nonisothermal measurements has been outlined [22] and applied for the study of thermo-oxidation of edible oils [10]. For the treatment of experimental DSC data, it was applied the procedure from the latter citation.
The DSC records of nonisothermal thermo-oxidation of olive oil C + A are depicted in Figure 1; the other EVOOs studied exhibited similar pattern. The peak corresponding to thermo-oxidation exhibits a small shoulder near its onset. The shoulder is present for all samples and for all heating rates employed; therefore, the values of oxidation onset temperatures, Ti, were evaluated as its onset extrapolated to the baseline. It can be seen from Figure 1 that higher heating rate always leads to higher oxidation onset temperature. Šimon [22] demonstrated that employing a non-Arrhenian dependence of the reaction rate on temperature, k (T) = A′exp(DT) , and assuming the same conversion for all heating rates, the dependence of oxidation onset temperature (Ti) on the heating rate can be described by the following equation:where “” is the heating rate in °C min−1 and “” and “” are kinetic parameters of thermo-oxidation. Once the values of the kinetic parameters are determined from a series of experiments carried out at different heating rates, the oxidation induction time (OIT) can be calculated as
The evaluated oxidation onset temperatures for each oil at various heating rates are listed in Table 3. These Ti vs. β dependences were further analyzed to estimate the kinetic parameters employing nonlinear least squares method applied to equation (3); the resulting parameters are listed in Table 4. Figure 2 depicts a typical result of the least squares fitting procedure.
The kinetic parameters obtained from the treatment of nonisothermal data were used to predict the values of OIT. The prediction of the values of oxidation induction time, OITs, based on equation (3) for each olive oil are presented in Figure 3. Two representative temperatures were chosen (25°C and 150°C). The lower temperature represents the usual storage conditions. However, care should be taken since both representative temperatures chosen (25°C and 150°C) are outside the experimental range of DSC measurements. The higher representative temperature chosen (150°C) is much closer to the experimentally investigated temperature range and the corresponding OIT values are expected to be both more precise and accurate.
(a)
(b)
Figure 3 shows that all the OITs values predicted at 150°C lie in a relatively narrow range of 30 to 50 min with oil Arbequina being least stable. Considering the OITs uncertainty, all the olive oils exhibit approximately the same high-temperature thermo-oxidative stability.
Results for 25°C also suggest that Arbequina is the least stable oil, and the Coupage Changlot Real and Arbosana has about four times longer shelf life—the differences between the oils are now much more pronounced. However, it should be kept in mind that the temperature (25°C) lies far away from the experimental range, and nonlinear extrapolation affects both accuracy and precision of the results (as demonstrated by much longer error bars compared to high-temperature prediction).
Similarly, Li et al. [20] studied thermal oxidation stability of four different vegetable oils (palm, rapeseed, sunflower, and linseed oil) through DCS at different heating rates (1, 5, 7.5, 10, 15, and 20°C/min). According to Ti obtained for the different oils, the following order for oxidation stability was obtained: palm oil > rapeseed oil > sunflower oil > linseed oil. When comparing Li et al.’s [20] results with the present study, it can be concluded that for all heating rates, the four vegetable oils showed higher Ti in comparison with the EVOO studied in the present work. Ti at a heating rate of 10°C/min was equal to 250.2, 233.3, 221.1, and 202.9°C for palm, rapeseed, sunflower, and linseed oil, respectively. In contrast, Ti values between 190 and 196.9°C were obtained for the EVOO samples at the same conditions. Similar pattern was observed for all heating rates. In addition, similar behaviour was registered in both studies when comparing thermal decomposition profiles at different heating rates: higher heating rate resulted in higher degradation rate and increased Ti.
Differences in oxidation stability of these vegetable oils are directly related to FA composition: vegetable oils with higher UFA content are usually less stable than those with higher SFA proportion. This can be explained by FA chemical structure, determined by chain length, unsaturation degree, and ramifications. Oxidation mostly occurs in double bonds; for this reason, FAs with higher unsaturation degree are more prone to oxidation and less stable, as a consequence, than SFA [23, 24].
3.3. Ultraviolet Spectrophotometry
The four EVOO varieties studied showed similar UV spectra in the UV and visible range (Figure 4). Evaluation of the spectra of the four samples according to equation (1) yields the values summarized in Table 5. As shown, all olive oils fulfill the criteria for extra virgin olive oil laid down by the International Olive Oil Council and the Commission Regulation [15] since K232 and K270 values were lower than the limits established (2.50 and 0.22, respectively).
K232 is related to the presence of hydroperoxides, conjugated dienes, carboxylic compounds, and conjugated trienes. On the other hand, K270 is dependent on the secondary products formed from the oxidation products detected at 232 nm [11, 26]. Therefore, results indicated the absence of oxidation products in the olive oils studied as well as the absence of refining oil in the commercial EVOO samples.
Allouche et al. [27] studied the evolution of K232 and K270 values of two extra virgin olive oils from Arbequina and Picual cultivars during heating at 180°C. Results showed that both indexes increased notably during the heating process, obtaining the higher values for Arbequina oil. Similarly, it was experimentally proved in [11] that during oil oxidation, high levels of peroxides are generated from primary oxidation compounds, resulting in higher K232 and K270 values and fluorescence spectra with peaks in the 415–600 nm region. In addition, it was demonstrated in this study that the combination of fluorescence techniques with multivariate analysis is a suitable method to characterize olive oil on the basis of the main quality parameters of olive oil: peroxide value, K232, K270, and acidity.
The suitability of K232 and K270 to determine the quality and conservation status of vegetable oils was also proved by Rodrigues et al. [28]. In this work, oil from Jatropha curcas L seeds was stored for 42 days, at 35°C and 75% or 92% relative humidity (RH). Results showed that higher RH resulted in a higher increment in K232 and K270 values. Regarding K232, an increase of 0.029 absorbance units/day was observed at 75% RH; nevertheless, a faster increase was observed at 92% RH (0.059 absorbance units/day). Similar results were obtained for K270, showing an increase from 0.07 to 0.22 after storage in higher humidity conditions.
4. Conclusions
Authentication and traceability of extra virgin olive oils are highly in demand in the market. The International Olive Oil Council and the Commission Regulation [15] has defined the quality of olive oil according to a series of parameters such as free fatty acids content and UV-specific extinction coefficients (K232 and K270). These parameters were determined in this work; results showed that oleic acid is the most abundant in the four EVOO (77.1% on average), followed by palmitic (11.5% on average). The importance of FA profile is due to its high contribution to olive oil oxidative stability. K232 and K270 values confirmed the absence of oxidation primary and secondary products.
In addition, the results showed that oil analysis can be performed with differential scanning calorimetry, an alternative technique for the evaluation of olive oil quality and stability as well as the determination of the heating effect on olive oil. DSC is an efficient, fast, accurate, and environmentally friendly method for the identification of peaks related to olive oil chemical composition. Nevertheless, in terms of authenticity, the information provided by the DSC analysis is not enough to detect adulterated olive oils due to the large number of possible adulterants [1].
In the four different EVOO varieties studied, DSC provided thermal fingerprints of the samples. For all heating rates, the peak corresponding to thermo-oxidation exhibits a small shoulder near its onset and all samples shown similar DSC record. It also can be concluded from the analysis of the Ti vs. β dependences that, for all samples, higher heating rate always leads to higher oxidation onset temperature. When comparing results obtained at two representative temperatures (25°C and 150°C), higher temperature is much closer to the experimentally investigated temperature range, as a consequence, OIT values obtained are more precise and accurate, exhibiting all the oils approximately the same thermo-oxidative stability. Much longer error bars as a consequence of less accuracy and precision of the results are obtained at 25°C.
It can, therefore, be concluded that the control of storage conditions of olive oil (temperature, humidity, etc.) is extremely relevant in order to preserve its quality. Evaluation of FA profile, K232 and K270 values, and Ti through DSC is a suitable, simple, and accurate technique to predict the quality, conservation status, and oxidation stability of different vegetable oils.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The Slovak Research and Development Agency supported this work under the contract nos. APVV-0850-11, APVV-14-0393, and APVV-15-0052. The Slovak Scientific Grant Agency Vega also supported this work by the contract nos. 1/0353/16 and 1/0592/15. In addition, the authors would like to thank the Operation Research and Development Program for the projects: “National Centre for Research and Application of renewable energy sources” (ITMS 26240120016 and ITMS 26240120028), “Competence centre for new materials, advanced technologies and energy” (ITMS 26240220073), and “University Science Park STU Bratislava” (ITMS 26240220084), cofinanced by the European Regional Development Fund.