Abstract
Economically motivated adulteration of expensive coconut oil with low cost oil, like palm kernel oil and palmolein is difficult to detect and quantify by available methods primarily due to their overlapping physicochemical properties with coconut oil. In the present work, a HPLC method has been developed to detect and quantify the degree of adulteration of coconut oil with palmolein and palm kernel oil based on triglyceride structure. The normalized area percentage of trilaurin (C36) among the three major TAG molecular species dilaurin-monocaprin/myristin-caprylin-laurin (C34), trilaurin (C36) and dilaurin-monomyristin (C38) of coconut oil was chosen as detection index for quantifying degree of adulteration of coconut oil with palm kernel oil, while the area ratio of dipalmitoyl-monoolein: trilaurin was chosen as detection index for quantifying adulteration of coconut oil with palmolein. The RP-HPLC based method developed in the present work is effective with a 2–4% minimum detection limit of adulterant oils and 78–98% detection accuracy depending on the degree of adulteration and types of oil.
Keywords: Adulteration, Coconut oil, Palmolein, Palm kernel oil, HPLC-ELSD
Introduction
Edible oils and fats are the most energy rich component in our diet. Edible oil not only increases the taste and palatability of food products but also source of many fat-soluble important nutrients. However, along with the nutritional and sensory need for consumption of oil vis-a-vis health risk associated with high fat intake, there are several man-made risk factors associated with consumption of edible oil. One such risk factor is adulteration of oil at different stages of food chain. Adulteration of edible oil is a menace to the society adversely affecting the health of people. According to the database developed on Food Fraud and Economically Motivated Adulteration from 1980 to 2010, 24% of the scholarly articles are in the edible oil sector, making it the primary mode of emerging risk factors (Moore et. al. 2012). Widespread application of edible oil in Indian culinary practices concerned health professional for consumption of such adulterated oil. Adulteration is done mainly for economic reason, by mixing cheap oil or unusual oil or used oil with good oil. Cheaper edible oils are those oils which has lower acceptance level to the consumer either due to health reason or due to the taste and sensory issues. Unusual oils are those oils containing anti-nutritional components or unusual fatty acids or out of the range physico-chemical properties. A third category of adulteration is adding used/fried oil to good oil. Though, the present food regulation has some indicative tests based on either very specific fatty acids or some minor specific components for determining the presence of some specific oils, but these tests are not quantitative (FAO Manual 1997; ISI Handbook of Food Analysis Part XIII, Pages 86, 89 & 91, 1984; ICMR Manual 1990; AOCS Ce 6–25).
There are some uniqueness in each oil with respect to their fatty acid composition or presence of minor components or composition of triacylglycerol (TAG) molecular species. Each of these parameters can be studied by different chromatographic techniques such as GC, HPLC, GC–MS, LC–MS, ESI–MS, MALDI-TOFMS etc. (Zhang et. al. 2014; Yang et. al. 2013; Goodacre et. al. 2002; Park and Lee 2003; Catharino et. al. 2005; Pizzo et. al. 2018; Green et al. 2020; Park et. al. 2010; Neff et al. 2001). Different spectral techniques such as Raman spectroscopy, excitation-emission fluorescence spectroscopy, FTIR, NMR, etc., have also been used for the detection and quantification of adulteration of edible oils (Yang and Irudayaraj 2001; Bartlet 1957; Vigli et. al. 2003; Maggio et. al. 2010; Scott et al. 2003). Data obtained using all these spectral and chromatographic techniques are often modelled into different multivariate statistical methods for ease of checking adulterations of edible oil (Karthik te. al. 2012). Oil adulteration in the western world is mostly in olive oil and hence most research work was oriented towards detection of adulteration of olive oil (Zhang et. al. 2014; Yang et al. 2013; Goodacre et al. 2002; Park et. al. 2003; Green et al. 2020; Yang et. al. 2001; Bartlet 1957; Vigli et. al. 2003; Maggio et. al. 2010; Catharino et. al. 2005; Pizzo et. al. 2018). There are reports wherein authenticity of other edible oil such as sesame, rice bran, soybean, rapeseed, peanut and sunflower oils has been attempted (Zhang et. al. 2014; Azadmard-Damirchi and Torbati 2015; Mansur et. al. 2018; Karthik et. al. 2014; Lee et. al. 2001). A review article on detecting adulteration of vegetable oils by combining chemometric and chromatographic data of TAG profile is reported in the literature (Bosque-Sandra et. al. 2012). Methods for the detection and quantification of cocoa butter equivalent (CBE) in cocoa butter (CB) in dark and milk chocolate are also based on TAG structure of CBE and CB (ISO 23275 (2006) and EUR 22666N (2007)). There are very few reports on detecting as well as quantifying adulteration of coconut oil with two potential adulterant oils namely palmolein and palm kernel oil. In one report, classification of virgin coconut oil (VCNO) and a very unusual type of adulterated VCNO (mixing palm oil and mustard oil to VCNO) was made by combining FTIR spectroscopy with chemometry (Pandurangan et. al. 2017). The same technique was adopted in yet another report, for detecting adulteration of VCNO with palm kernel olein down to an adulterated level of 1% (Manaf et. al. 2007). Adulteration of coconut oil with soybean oil has also been detected by ESI–MS (Pizzo et. al. 2019), although such adulteration can be detected by many simpler techniques. A recent review article compiled the application of FTIR for determining the quality and safety of oils and fats (Li et. al. 2019). Parameters such as acid value, iodine value, peroxide value, moisture and Trans fat content (TFA) of oils and fats were accomplished with the help of FTIR technique combined with chemometry. However, only FTIR method for the estimation of TFA (1–5% w/w range) officially recognised by AOCS and AOAC.
The coconut tree (Cocos nucifera L.) is a tropical plant grown abundantly in coastal India. Coconut oil (CNO), obtained from fresh and matured kernel is known for its high content of lauric acid and medium chain triglyceride (MCT). This short and medium chain fatty acid rich premium oil, recognised as multipurpose nutrient supplement is being consumed in southern India as frying oil. It is important to detect any kind of adulteration of this high priced oil in order to maintain its quality and safety. However, it has been observed in recent time that the oil has been adulterated often with low cost palmolein (PML) or palm kernel oil (PKO). Though such adulteration is economically motivated and possess no health risk but the high priced CNO loses its quality and nutritive value. Beyond a certain degree of adulteration, it is possible to detect (not quantify) such adulterant in CNO based on fatty acid composition. But at lower level of adulteration, even fatty acid composition failed to detect. In fact, identification and quantification of PML and PKO in CNO is very difficult with the available methodology due to number of common compositional features in these oils. Cheaper oils are often chosen as an adulterant in a manner with overlapping physicochemical properties with expensive oil. It is needed to be seen whether differences in TAG profiles can be used for the detection as well as quantification of low-cost PML and PKO in premium quality CNO. In the present work, a RP-HPLC method is developed based on composition of TAG molecular species for detecting as well as quantifying the degree of adulteration of CNO with PML and PKO.
Experimental section
Materials
Altogether 25 coconut oil, 3 palm kernel oil and 6 palmolein collected from different manufacturers in India. HPLC-grade dichloromethane (DCM) and acetonitrile (ACN) were procured from M/s Merck India Limited, Mumbai. Standard TAG mixture (Tricaprin, Trilaurin, Trimyristin, Tripalmitin and Tristearin) (Supelco TAG Mix: Cat No. CRM17811) and standard fatty acid methyl ester mixture (Supelco FAME Mix: Cat No. CRM47885) were purchased from Sigma-Aldrich (St. Louis, USA).
Methods
Fatty acid compositional analysis using GC
Analysis of the fatty acid composition of coconut, palmolein and palm kernel oil was carried out after derivatization to fatty acid methyl esters (FAME) according to AOAC protocol (AOAC 969.33, 2000). The GC analysis was performed on a Agilent 6890 Series gas chromatograph equipped with a flame-ionization detector. A fused-silica capillary column DB225 (30 m × 0.25 mm i.d × 0.20 μm film thickness) was used with N2 as carrier gas. The column was operated at 160 °C for 2 min, raised to 230 °C at a rate of 5 °C/min and finally held at this temperature for 20 min until completion of the analysis. The injection port and detector temperatures were maintained at 230 and 260 °C, respectively and split ratio was 10:1. Peak identification was accomplished by injecting a standard fatty acid methyl ester mixture. The relative percentage of individual fatty acids was identified and quantified using ChemStation software (Rev. B.05.05).
High performance liquid chromatography (HPLC) analysis for TAG composition
The reversed phase HPLC analysis was performed on Agilent 1260 Infinity series HPLC equipped with an evaporative light scattering detector. Sample (about 10 μl of 0.5 mg/ml concentration in DCM) was injected in to the Xbridge RP column (C18-RP; Two columns of 250 mm × 2.5 mm dimension connected in series). The molecular species of TAGs were eluted using a combination of dichloromethane (DCM) and acetonitrile (ACN) at a flow rate of 1 ml/min in the following gradient mode: initial ratio of 55:45, v/v of ACN:DCM changed to 40:60 from 0 to 12 min, changed further over a period of 12–18 min to initial ratio of 55:45, v/v of ACN:DCM and continued in this ratio till 25 min in order to equilibrate the system.[31] The operating conditions for ELSD were: evaporative temperature 60 °C, flow of nitrogen 1.6 SLM, nebulizer temperature 55 °C and PMT gain 1.0. The molecular species of oils were identified by their equivalent carbon numbers (ECN), calculated as total number of fatty acids − 2x (number of double bonds in fatty acids) and the elution order was predicted as per the retention times of standard TAGs and also according to previous literature reports (Stolyhwo et. al. 1985; Shahidi 2005; Reena et. al. 2009).
Prior to analysis of TAG, the HPLC system was calibrated by injecting the CRM-TAG Mixture containing five TAG (20 μg/mL each), namely tricaprin (CCC), tricaprylin (CyCyCy), trilaurin (LaLaLa), trimyristin (MMM) and tripalmitin (PPP). Responses of all these symmetrical TAG molecules, measured as area under each peak were found to follow a linear relation versus concentration of the triglycerides (expressed as mg/mL) with regression coefficient varying in the range of 0.9866 to 0.9927. The limit of quantification (LOQ) for the five different types of TAG molecules present in CRM standard is 0.09003 mg/mL. However, the limit of detection (LOD) is found to be dependent on types of TAG molecules. Retention time, response factor, inter- and intra- day variation in retention time and response factor, and uncertainty (Type B) in the measurement of TAG molecular species was checked by injecting CRM-TAG Mixture four times in the HPLC system. The percentage of uncertainty in the measurement of area of five symmetrical standard TAG molecular species in the CRM-TAG mixture lies in the range of 2.7–3.1%. Response ratio of five peaks, namely CCC, CyCyCy, LaLaLa, MMM and PPP of CRM-TAG mixture is found to be 0.93:0.92:0.95:1.03:1.22. This indicates a little increase in response of TAGs with the increase in chain length of fatty acid. Both inter- and intra- day variation in retention time and response factor of individual TAGs are not found to be significant statistically.
Results
Fatty acid compositional analysis of coconut oil, palm kernel oil and palmolein collected from different manufacturers showed very consistent profile (Table 1). Both coconut oil (CNO) and palm kernel oil (PKO) contain similar type of fatty acids, though they differ in content at individual level. Either oil contains lauric acid as major fatty acids, but level of oleic acid is significantly high in PKO than CNO.
Table 1.
Mean fatty acid composition (in wt%) of 25 coconut oils, 3 palm kernel oils and 6 palmolein
| Fatty acida | Fatty acid composition | ||
|---|---|---|---|
| Coconut oil (Mean ± SD, n = 25) |
Palm kernel oil (n = 3) |
Palmolein (Mean ± SD, n = 6) |
|
| 6:0 (Ca) | 0.2 ± 0.06 | 2.4 ± 0.3 | – |
| 8:0 (Cy) | 5.5 ± 1.3 | 2.7 ± 0.2 | – |
| 10:0 (C) | 5.3 ± 0.4 | 3.0 ± 0.5 | – |
| 12:0 (L) | 47.7 ± 1.8 | 40.5 ± 1.2 | 0.2 ± 0.06 |
| 14:0 (M) | 20.8 ± 0.7 | 13.4 ± 0.9 | 1.0 ± 0.05 |
| 16:0 (P) | 9.2 ± 0.8 | 9.4 ± 0.8 | 39.6 ± 1.1 |
| 18:0 (S) | 3.0 ± 0.4 | 3.6 ± 0.1 | 4.5 ± 0.3 |
| 18:1 (O) | 6.7 ± 1.0 | 21.2 ± 1.2 | 42.9 ± 0.8 |
| 18:2 (L) | 1.6 ± 0.5 | 3.5 ± 0.9 | 10.9 ± 0.7 |
| 18:3 (Li) | – | – | 0.2 ± 0.0 |
| 20:0 (A) | – | 0.2 ± 0.0 | 0.4 ± 0.0 |
a Letters in parenthesis indicate abbreviation for corresponding fatty acid: Ca, Caproic acid (6:0); Cy, caprylic acid (8:0); C, capric acid (10:0); La, lauric acid (12:0); M, myristic acid (14:0); P, palmitic acid (16:0); S, stearic acid (18:0); O, oleic acid (18:1); L, linoleic acid (18:2); Li, linolenic acid (18:3); A, arachidic acid (20:0)
All these similarities and differences are reflected in the triacylglycerol (TAG) profile too (Table 2). Types of TAG molecular species present in CNO are also there in PKO due to their similarities in fatty acid profile, though there are differences in their individual content. In CNO, lauric acid containing TAG with equivalent carbon numbers (ECN) C34, C36 and C38 are the most abundant (68–69%). Among them, trilaurin (LaLaLa; C36) is the major TAG molecular species (25.6%). In PKO also, trilaurin is the predominant TAG molecular species constituting nearly 34.1% of total molecular species, followed by dilaurin-monocaprin/myristin-caprylin-laurin (LaCLa/MCyLa; C34) and dilaurin-monomyristin (LaMLa; C38). Thus adulteration of CNO with PKO at a level < 20% is difficult to detect by the change in fatty acid composition due to their similarities in type of fatty acids as well as type of TAG molecular species present in them. Both these oils contain very low level of unsaponifiable matters (< 0.3%). Though they differ in type of tocopherol present in them (15–25 ppm of α-tocopherol in CNO and 82–88 ppm γ-tocopherol in PKO), but the levels are too low to be detected by HPLC following AOCS method (2017). Hence change in composition of TAG may be an option left for detection and subsequent quantification of PKO in CNO.
Table 2.
TAG composition (in wt%) of 25 coconut oil, 3 palm kernel oil and 6 palmolein
| ECN a | Expected TAG molecular species@ | Composition of triacylglycerol (TAG) | ||
|---|---|---|---|---|
| Coconut oil (Mean ± SD, n = 25) |
Palm kernel oil (n = 3) |
Palmolein (Mean ± SD, n + 6) |
||
| C30 | CyCLa | 1.3 ± 0.2 | 0.6 ± 0.0 | – |
| C32 | LaCyLa | 14.6 ± 2.0 | 8.0 ± 0.2 | – |
| C34 | LaCLa/MCyLa | 23.1 ± 1.9 | 10.0 ± 1.2 | – |
| C36 (1) | CyOLa | 0.5 ± 0.0 | 0.3 ± 0.1 | – |
| C36 | LaLaLa | 25.6 ± 0.7 | 34.1 ± 2.1 | – |
| C38 (1) | COLa | 0.2 ± 0.1 | 0.4 ± 0.1 | – |
| C38 | LaMLa | 20.1 ± 1.2 | 17.2 ± 1.2 | – |
| C40 (1) | LaOLa | 0.5 ± 0.2 | 4.2 ± 1.1 | – |
| C40 | MLaM/LaPLa | 9.1 ± 1.6 | 6.9 ± 1.5 | – |
| C42 (1) | LaOM | 0.4 ± 0.2 | 3.1 ± 0.8 | – |
| C42 | LaMP/LaSLa | 3.1 ± 0.7 | 6.2 ± 0.9 | – |
| C44 (4) | PLL | – | – | 1.1 ± 0.2 |
| C44 (2) | LaOO | 0.3 ± 0.1 | 3.8 ± 0.7 | |
| C44 | PLaP | 0.6 ± 0.2 | 1.0 ± 0.2 | |
| C46 (4) | LOO | – | – | 1.0 ± 0.2 |
| C46 (3) | POL | – | 9.2 ± 1.0 | |
| C46 (2) | PLP | – | – | 7.1 ± 0.6 |
| C46 (1) | MOP/SOLa | 0.4 ± 0.1 | 2.0 ± 0.9 | – |
| C48 (3) | OOO | – | 1.1 ± 0.7 | 4.0 ± 0.5 |
| C48 (2) | POO | 0.4 ± 0.1 | 0.4 ± 0.2 | 33.8 ± 2.2 |
| C48 (1) | POP | – | 0.1 ± 0.1 | 38.6 ± 3.5 |
| C48 | PPP | – | – | 1.8 ± 0.2 |
| C50 (2) | SOO | – | – | 3.0 ± 0.5 |
aECN, Equivalent Carbon Number; value in parenthesis indicates number of double bond in the TAG molecular species. @ See Table 1 footnote for abbreviation of individual fatty acid in TAG molecular species
HPLC chromatograms of pure CNO, PKO and PML is given in Fig. 1. A close look at the TAG profile of both CNO and PKO indicates that almost all the TAG molecular species are co-eluting and hence are similar in nature (Table 2) making the qualitative prediction of the presence of PKL in CNO next to impossible. However, differences in their individual levels indicate that addition of PKO to CNO will bring change in the content of three major TAGs, dilaurin-monocaprin/myristin-caprylin-laurin (LaCLa/MCyLa; C34), trilaurin (LaLaLa; C36) and dilaurin-monomyristin (LaMLa; C38) of CNO. There will be decrease in the contents of C34 and C38 and an increase in C36. But due to higher content of C36 in PKO than in CNO, the increase will be more pronounced in C36. Hence it may be possible to detect as well as quantify the presence of PKO in CNO by calculating normalized area percent of C36 (%C36) among areas of these three major TAGs (C34, C36 and C38). In the 25 well known market samples of CNO, the average %C36 was found to be 37.97 ± 0.62. Hence, a higher value of %C36 than the average value is a possible indication of adulteration of CNO with PKO.
Fig. 1.
Superimposed HPLC chromatograms of TAG mol. species of CNO (in blue), PKO (in red) and palmolein (in green). Peaks are assigned on the basis of Equivalent Carbon Number (ECN) as 1: C30, CyCLa; 2: C32, LaCyLa; 3: C34, LaCLa; 4: C36 (1), CyOLa; 5: C36, LaLaLa; 6: C38 (1), COLa; 7: C38, LaMLa; 8: C40 (1), LaOLa; 9: C40, MLaM; 10: C42 (1), LaOM; 11: C42, LaMP/C44 (4), PLL; 12: C44 (2), LaOO/C46 (4), LOO /C46 (3), POL; 13: C46 (1), MOP & SOLa /C46 (2), PLP; 14: C48 (3), OOO; 15: C48 (2), POO; 16: C48 (1), POP; 17: C48, PPP; 18: C50 (2), SOO. See Table 1 and 2 for abbreviations
For quantification, 25 samples of coconut oil were mixed equally and the combined coconut oil was injected in HPLC. The %C36 in the combined CNO was found to be 37.87% (Table 3). With this pooled CNO (n = 25), different physical blends of pooled CNO with pooled PKO (n = 3) at varying ratios (2.2849 to 22.8350% by weight of PKO) were prepared. All blends were prepared at 100 g scale and after thorough mixing for 1 h using magnetic stirrer, samples were analyzed by HPLC for TAG compositional analysis. In Table 3, the calculated normalized %C36 is given along with the varying concentration of PKO added to CNO. The increase in % C36 is plotted against increase in %PKO in coconut oil. Regression analysis indicated a linear relation [y = 37.93325 + 0.14297 x] between %C36 and the %PKO in coconut oil with a Regression coefficient (R2) of 0.9917. This linear relationship was used to determine adulteration of CNO with PKO quantitatively based on the fitted table.
Table 3.
Calculated normalized area% of C36 among the three major TAGs (C34, C36 and C38) in CNO on addition of different amount of PKO
| % PKO in CNO | %C36a | |
|---|---|---|
| 0 | 37.87 ± 0.62 |  |
| 2.2849 | 38.43 ± 0.39 | |
| 4.1028 | 38.73 ± 0.30 | |
| 6.1672 | 38.69 ± 0.63 | |
| 8.0469 | 39.09 ± 0.04 | |
| 9.9949 | 39.37 ± 0.29 | |
| 12.2309 | 39.68 ± 0.11 | |
| 14.0804 | 39.97 ± 0.08 | |
| 16.3778 | 40.37 ± 0.17 | |
| 18.0364 | 40.58 ± 0.45 | |
| 20.2789 | 40.83 ± 0.07 | |
| 22.8350 | 41.27 ± 0.31 |
a based on mean areas of three independent injections; %C36 = (C36 × 100)/(C34 + C36 + C38)
An intra-laboratory assay was conducted by randomly blending three PKO with different CNO at concentration within the calibrated range (2–22%) to validate the concept. Altogether 10 such blends were prepared and injected in HPLC. From the chromatogram, the normalized abundance of %C36 among the three major TAGs was calculated and compared with the fitted table to know quantitative content of PKL in CNO. Based on the above regression analysis and data obtained in 10 randomly blended sample, it has been observed that the accuracy in predicting the adulteration of PKO in CNO is in the range of 78–94% based on degree of adulteration and type of coconut oil.
Palmolein (PML), on the other hand contains higher content of palmitic and oleic acid compared to CNO (Table 1). Hence, detection of adulteration of CNO with PML is possible up to a certain concentration based on fatty acid composition. As PML is rich in tocotrienols (both α and γ) and CNO contains only 15–25 ppm of α-tocopherol, it is also possible to detect the presence of PML in CNO by the analysis of tocotrienols following AOCS method.[35] However, due to oil-to-oil variation of this parameter in both these oils, quantification of PML in CNO is quite difficult based on the composition of tocols. However, PML contains two major TAG molecular species, dipalmitin-monoolein (POP) and diolein-monopalmitin (POO) (Fig. 1) constituting together 72.4% of total molecular species and which are at a negligible concentration in CNO (Table 2). As both these peaks of PML are far away from all prominent peaks of CNO, it may be possible to quantify PML in CNO based on ratio of dipalmitin-monoolein (POP) of PML to trilaurin (LaLaLa) of CNO.
In a similar fashion, 6 samples of PML were mixed equally and the pooled oil was used to adulterate pooled CNO (n = 25) at different proportions (1.05131 to 24.26403% by weight). All blends were prepared at 100 g scale and after thorough mixing for 1 h using magnetic stirrer, samples were analyzed by HPLC for TAG compositional analysis. In Table 4, mean ratios of POP:LaLaLa of all prepared blends are given. Ratios of POP:LaLaLa was then plotted against %PML added to CNO. The curve showed an exponential increase in the ratio with the increase in concentration of PML in CNO.Regression analysis indicated an exponential relation [y = 0.09087 × exp {− x/(− 17.218)} + (− 0.09182)] between the POP:LaLaLa ratio and the %PML in CNO with a regression coefficient (R2) of 0.99694. This relationship was used to determine adulteration of CNO with PML quantitatively based on the fitted table.
Table 4.
Mean area ratios of POP:LaLaLa of all prepared blends at different degree of adulteration of CNO with PML
| %PML in CNO | Mean of POP:LaLaLaa | |
|---|---|---|
| 1.05131 | 0.0115 ± 0.00014 |  |
| 2.07199 | 0.01295 ± 0.0007 | |
| 4.11504 | 0.02185 ± 0.0009 | |
| 6.07266 | 0.03386 ± 0.0006 | |
| 8.01862 | 0.04775 ± 0.0007 | |
| 10.07316 | 0.06658 ± 0.0014 | |
| 11.99368 | 0.08962 ± 0.0010 | |
| 14.10327 | 0.11524 ± 0.0027 | |
| 16.17778 | 0.14324 ± 0.0012 | |
| 18.19791 | 0.16933 ± 0.0012 | |
| 19.98828 | 0.20415 ± 0.0045 | |
| 21.98890 | 0.23971 ± 0.0043 | |
| 24.26403 | 0.28274 ± 0.0043 |
abased on mean areas of three independent injections
An intra-laboratory assay was conducted by randomly blending 6 PML with different CNO at concentration within the calibrated range (1–24%) to validate the concept. Altogether 10 such blends were prepared and injected in HPLC. The mean ratios of POP:LaLaLa of three injections of each blend were calculated. Based on the above regression analysis and concurrent fitted table, the percentage of PML in CNO was calculated. The range of accuracy in the measurement of %PML in CNO lies in the range of 78–98% based on degree of adulteration and type of CNO and PML.
Discussion
An ideal method is one which require least sample preparation step without any pre-treatment, easy to understand by an average analyst and low capital cost. The present manuscript is one such attempt to quantify adulteration of coconut oil with two low cost oils having overlapping physicochemical properties using HPLC-ELSD, which has not been reported earlier. Ease of sample preparation, easy to understand by an average analyst, short run time, not so high capital cost (compared to NMR, LC–MS/MS) and without any chemometry need to treat final data makes this HPLC method better than reported methods, most of which are reported to check adulteration of a less complicated oil (olive oil). IR operation, on the other hand is easier, faster and cheaper, but data needs to be modelled into different multivariate statistical methods for ease of checking adulterations. IR method fails to give conclusive result with a narrow difference in iodine value between the major and minor oil.
Edible oils and fats are primarily consists of a complex mixture of triglycerides (TAG). The analysis of TAG is a challenging task due to the presence of a large number of individual TAGs arising out of positioning a group of fatty acids over the three position of glycerol backbone. Stolyhwo et al. first published composition of TAG molecular species of some edible oils by coupling reversed phase HPLC with ELSD detector (1985). Identification of TAG molecular species is based on Equivalent Carbon Number (ECN = Total carbon–2 × No. of double bond), a concept used way back in 1979 to understand the polarity of individual TAGs (Herslof et. al. 1979). Even GC analysis of TAG molecular species was also reported in the literature using very specific capillary column like CP-TAP, RTX-65TG, DB 17 etc. (Buchgraber et. al. 2004).
Fatty acid composition of 25 CNO and 6 PML samples showed very consistent composition with variation in the content of major fatty acids in the range of 0.3 to 1.8% (Table 1). During the profiling of TAGs of 25 CNO samples, very consistence composition was also observed with variation in the content of major TAGs in the range of 0.7–2.0% (Table 2). Very consistent composition of TAG was also observed in 6 samples of PML. However, profiling of only three PKO samples was carried out due to its non-availability in open market in India. In CNO, the major TAGs were LaCLa/MCyLa (23.1 ± 1.9), LaLaLa (25.6 ± 0.7), LaMLa (20.1 ± 1.2). PKO, on the other hand though depicted similar TAG profile but differ in their individual content. Hence, it is practically impossible to detect the presence of PKO in CNO. However, a close look at their contents indicates that on addition of PKO in CNO, there will be decrease in content of LaCLa/MCyLa (C34) and LaMLa (C38) and increase in content of LaLaLa (C36). Therefore, a calculation of normalized variation in the area of C36 (LaLaLa) among the areas of three major affected TAGs (LaCLa/MCyLa, LaLaLa and LaMLa) is used as detection index for detecting and quantifying the degree of adulteration of CNO with PKO. It was also postulated based on the average value of %C36 (37.97 ± 0.62) in 25 studied CNO that a higher value of %C36 than the average value is a possible indication of adulteration of CNO with PKO. A linear relationship was observed between %C36 and %PKO added to CNO. Based on regression analysis of data of prepared blends, an intra-laboratory assay was conducted. Results indicated that the percentage of accuracy in the measurement of this detection index for the detection of PKO in CNO is in the range of 78–94% depending on the degree of adulteration and type of oils used in the assay. Lower percentage of accuracy was observed when %PKO in CNO is below 4%.
Two most abundant TAGs in PML are POO and POP, fortunately eluting at higher retention time than most TAGs of CNO. Thus it is possible to detect presence of PML in CNO at a concentration as low as 1% in the mixing ratio. Very concept of this method for quantification of minor oil to major oil is to identify a major TAG from each oil and finding a relation between them. Remarkable difference in retention time of trilaurin (LaLaLa) of CNO and dipalmitin-monoolein (POP) of PML and complete absence of trilaurin (LaLaLa) in PML and negligible presence of dipalmitin-monoolein (POP) in CNO, the ratio of POP:LaLaLa is used as detection index for quantifying the level of PML in CNO. Regression analysis showed an exponential increase in the ratio of POP: LaLaLa with the increase in concentration of PML in CNO. Based on regression analysis of data of prepared blends, an intra-laboratory assay was also conducted. The range of accuracy in the measurement of %PML in CNO lies in the range of 78–98%.
In conclusion, a reversed-phase HPLC-based method developed in the present work to detect and quantify the degree of adulteration of CNO with two most common and potential adulterant oils, namely PML and PKO. The method does not require any derivatization prior to its injection in HPLC. The normalized variation in %area of LaLaLa (C36) among the three major TAGs (C34, C36 and C38) has been considered as detection index for checking adulteration of CNO with PKO. The mean %C36 in pure CNO is 37.97 ± 0.62 and a higher value will indicate its adulteration with PKO. The method is effective with a minimum detection limit of 2–4% adulterant oils and 78–94% measurement accuracy depending on the degree of adulteration and types of oils to quantify the presence of PKO in CNO. In case of adulteration of CNO with PML, the ratio of POP: LaLaLa has been considered as detection index. Effectiveness of this method is as low as 1% and accuracy in the quantification of PML in CNO lies in the range of 78–98%. Hence, it is possible to detect and quantify adulteration of CNO with PMO and PKO based on differences in their TAG profile. The present research work was carried out to prove the concept and needs cross validation study for its final acceptance by the regulatory authority.
Acknowledgements
The research work was carried out with the financial grant received from Food Safety and Standard Authority of India (FSSAI), Government of India (No. 75/R&D/Adulteration in Edible Oil/RARD/2016-FSSAI). We thank Director, CSIR-Indian Institute of Chemical Technology for the support (IICT/Pubs./2019/325).
Abbreviations
- TAG
Triacylglycerol
- LaLaLa
Trilaurin
- CCC
Tricaprin
- CyCyCy
Tricaprylin
- MMM
Trimyristin
- PPP
Tripalmitin
- LaMLa
Dilaurin-monomyristin
- LaCLa
Dilaurin-monocaprin
- MCyLa
Myristin-caprylin-laurin
- POP
Dipalmitin-monoolein
- POO
Diolein-monopalmitin
- ECN
Equivalent carbon number
- CNO
Coconut oil
- VCNO
Virgin coconut oil
- PKL
Palm kernel oil
- PML
Palmolein
- FAME
Fatty acid methyl esters
- CRM
Certified reference material
Compliance with ethical standards
Conflict of interest
The Authors declare that they have no conflict of interest.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- AOAC . Official methods of analysis of AOAC International. 17. Gaithersburg, MD: AOAC international; 2000. [Google Scholar]
- AOCS (1989) Test for the presence of palmolein in other oil. Official Method of AOCS Ce 6–25
- AOCS (2017) Tocopherols and tocotrienols in vegetable oils and fats. Official method of AOCS Ce 8–89
- Azadmard-Damirchi S, Torbati M. Adulterations in some edible oils and fats and their detection methods. J Food Qual Hazard Contr. 2015;2:38–44. [Google Scholar]
- Bartlet JC. Identification of edible oils and the detection of oil adulteration by differential infra-red spectroscopy. Nature. 1957;180:1071–1072. doi: 10.1038/1801071a0. [DOI] [PubMed] [Google Scholar]
- Bosque-Sandra JM, Cuadros-Rodrigue L, Ruiz-Samblas C, Paulina de la Mata A. Combining chromatography and chemometrics for the characterization and authentication of fats and oils from triacylglycerol composition data—A review. Anal ChimActa. 2012;724:1–11. doi: 10.1016/j.aca.2012.02.041. [DOI] [PubMed] [Google Scholar]
- Buchgraber M, Ulberth F, Emons H, Anklam E. Triacylglycerol profiling by chromatographic techniques. Eur J Lipid SciTechnol. 2004;106:621–648. doi: 10.1002/ejlt.200400986. [DOI] [Google Scholar]
- Catharino RR, Haddad R, Cabrini LG, Cunha IBS, Sawaya ACHF, Eberlin MN. Characterization of vegetable oils by electrospray ionization mass spectrometry fingerprinting: classification, quality, adulteration, and aging. Anal Chem. 2005;77:7429–7433. doi: 10.1021/ac0512507. [DOI] [PubMed] [Google Scholar]
- EUR 22666N (2007) Detection and quantification of cocoa butter equivalent (CBE) in milk Chocolate
- FAO (1997) Manuals of Food Quality Control (14/8) Test for the presence of cottonseed oil, Page 271
- Goodacre R, Vaidyanathan S, Bianchib G, Kella DB. Metabolic profiling using direct infusion electrospray ionisation mass spectrometry for the characterization of olive oils. Analyst. 2002;127:1457–1462. doi: 10.1039/b206037j. [DOI] [PubMed] [Google Scholar]
- Green HS, Li X, De Pra M, Lovejoy K, Steiner F, Acworth IN, Wang SC. A rapid method for the detection of extra virgin olive oil adulteration using UHPLC-CAD profiling of triacylglycerols and PCA. Food Control. 2020;107:106773–106780. doi: 10.1016/j.foodcont.2019.106773. [DOI] [Google Scholar]
- Herslof B, Podlaha O, Toregard B. HPLC of triglycerides. J Am Oil Chemist's Soc. 1979;56:864–866. doi: 10.1007/BF02909537. [DOI] [Google Scholar]
- ICMR (1990) Manual methods of analysis for adulterants and contaminates in foods: Test for presence of rice bran oil, Page 5
- ISI (1984) Handbook of Food Analysis (Part XIII) Test for the presence of sesame oil, page 86
- ISI (1984) Handbook of Food Analysis (Part XIII) Modified test for presence of Castor oil, page 91
- ISI (1984) Handbook of Food Analysis (Part XIII) Test for the presence of mineral oil in edible oil, page 89
- ISO 23275 (2006) Method for the Detection and Quantification of cocoa butter equivalent (CBE) in Dark Chocolate
- Karthik D, Vijayarekha K, Manjula K. Multivariate analysis for detecting adulteration in edible oil: a review. IEEE-IntConfAdvEngSciManag. 2012;31:272–277. [Google Scholar]
- Karthik D, Vijayarekha K, Manickkam V. A simple method for detecting adulteration in sunflower oil using chemometry. Rasayan J Chem. 2014;7:99–103. [Google Scholar]
- Lee DS, Lee ES, Kim HJ, Kim SO, Kim K. Reversed phase liquid chromatographic determination of triacylglycerol composition in sesame oils and the chemometric detection of adulteration. Anal ChimActa. 2001;429:321–330. doi: 10.1016/S0003-2670(00)01289-7. [DOI] [Google Scholar]
- Li Q, Chen J, Huyan Z, Kou Y, Xu L, Yu X, Gao J-M. Application of fourier transform infrared spectroscopy for the quality and safety analysis of fats and oils: a review. Crit Rev Food SciNutr. 2019;59:3597–3611. doi: 10.1080/10408398.2018.1500441. [DOI] [PubMed] [Google Scholar]
- Maggio RM, Cerretani L, Chivaro E, Kaufman T, Bendini A. A novel chemometric strategy for the estimation of extra virgin olive oil adulteration with edible oils. Food Control. 2010;21:890–895. doi: 10.1016/j.foodcont.2009.12.006. [DOI] [Google Scholar]
- Manaf MA, Man YBC, Hamid NSA, Ismail A, Abidin SZ. Analysis of adulteration of virgin coconut oil with palm kernel olein using Fourier transform infrared spectroscopy. J Food Lipid. 2007;14:111–121. doi: 10.1111/j.1745-4522.2007.00066.x. [DOI] [Google Scholar]
- Mansur AR, Jeong H-R, Lee BH, Koo M, Seo D-H, Hwang SH, Kim D-O, Nam TG. Comparative evaluation of triacylglycerol, fatty acids, and volatile organic compounds as markers for authenticating sesame oil. Int J Food Prop. 2018;21:2509–2516. doi: 10.1080/10942912.2018.1534123. [DOI] [Google Scholar]
- Moore JC, Spink J, Lipp M. Development and application of a database of food ingredient fraud and economically motivated adulteration from 1980 to 2010. J Food Sci. 2012;77:R118–126. doi: 10.1111/j.1750-3841.2012.02657.x. [DOI] [PubMed] [Google Scholar]
- Neff WE, WmC B, List GR. Triacylglycerol structure of food fats high in saturated acids by HPLC and mass spectrometry. J LiqChromatogrRelatTechnol. 2001;24:837–854. [Google Scholar]
- Pandurangan MK, Murugesan S, Shettu N, Gajivaradhan, Detection of adulteration of coconut oil using Fourier transform infrared spectroscopy and chemometrics. Int J Stat Appl Math. 2017;2:46–51. [Google Scholar]
- Park YW, Chang P-S, Lee J. Application of triacylglycerol and fatty acid analyses to discriminate blended sesame oil with soybean oil. Food Chem. 2010;123:377–383. doi: 10.1016/j.foodchem.2010.04.049. [DOI] [Google Scholar]
- Park JR, Lee D-S. Detection of adulteration in olive oils using triacylglycerol compositions by high temperature gas chromatography. Bull Korean ChemSoc. 2003;24:4527–4530. [Google Scholar]
- Pizzo JS, Galuch MB, Santos PDS, Santos OO, Visentainer L, Eberlin MN, Visentainer JV. Assessment of adulteration of cosmetics based on vegetable oils by GC-FID and lipid profile using direct infusion electrospray ionization mass spectrometry (ESI-MS) J BrazChemSoc. 2018;29:2457–2465. [Google Scholar]
- Pizzo JS, Galuch MB, Santos PDS, Manin LP, Zappielo CD, Filho OJS, Santos OO, Visentainer JV. Determination of coconut oil adulteration with soybean oil by direct infusion electrospray ionization mass spectrometry. J BrazChemSoc. 2019;30:1468–1474. [Google Scholar]
- Reena MB, Reddy SRY, Lokesh BR. Changes in triacylglycerol molecular species and thermal properties of blended and interesterified mixtures of coconut oil or palmolein with rice bran oil or sesame oil. Eur J Lipid SciTechnol. 2009;111:346–357. doi: 10.1002/ejlt.200800065. [DOI] [Google Scholar]
- Scott SM, James D, Ali Z, Ohare WT, Rowell FJ. Total luminescence spectroscopy with pattern recognition for classification of edible oils. Analyst. 2003;128:966–973. doi: 10.1039/b303009a. [DOI] [PubMed] [Google Scholar]
- Stolyhwo A, Colin H, Guiochon G. Analysis of triglycerides in oils and fats by liquid chromatography with the laser light scattering detector. AnalytChem. 1985;57:1342–1354. doi: 10.1021/ac00284a038. [DOI] [PubMed] [Google Scholar]
- Shahidi F. Baileys industrial oil and fat product. New York: Willey; 2005. pp. 61–719. [Google Scholar]
- Vigli G, Philippidis A, Spyros A, Dais P. Classification of edible oils by employing 31P and 1H NMR spectroscopy in combination with multivariate statistical analysis. A proposal for the detection of seed oil adulteration in virgin olive oils. J Agri Food Chem. 2003;51:5715–5722. doi: 10.1021/jf030100z. [DOI] [PubMed] [Google Scholar]
- Yang Y, Ferro MD, Cavaco I, Liang Y. Detection and identification of extra virgin olive oil adulteration by GC-MS combined with chemometrics. J Agri Food Chem. 2013;61:3693–3702. doi: 10.1021/jf4000538. [DOI] [PubMed] [Google Scholar]
- Yang H, Irudayaraj J. Comparison of near-infrared, fourier transform-infrared and fourier transform-raman methods for determining olive pomace oil adulteration in extra virgin olive oil. J Am Oil Chemist's Soc. 2001;78:889–895. doi: 10.1007/s11746-001-0360-6. [DOI] [Google Scholar]
- Zhang L, Li P, Sun X, Wang X, Xu B, Wang X, Ma F, Zhang Q, Ding X. Classification and adulteration detection of vegetable oils based on fatty acid composition. J Agri Food Chem. 2014;62:8745–8751. doi: 10.1021/jf501097c. [DOI] [PubMed] [Google Scholar]