ANALYSIS OF PALM OIL AS OIL ADULTERANT IN OLIVE AND PUMPKIN SEED OILS IN TERNARY MIXTURE SYSTEMS USING FTIR SPECTROSCOPY AND CHEMOMETRICS

Objective: The study was designed to develop Fourier transform infrared (FTIR) spectroscopy in conjunction with chemometrics techniques of multivariate calibration and discriminant analysis (DA) for analysis of palm oil in a ternary mixture with EVOO and PSO. Methods: FTIR spectra of pure palm oil (PO), extra virgin olive oil (EVOO), pumpkin seed oil (PSO) and its ternary mixtures randomly prepared were scanned using FTIR spectrophotometer at wavenumbers of 4000-650 cm -1 corresponding to mid-infrared region, with resolution of 8 cm -1 Results: The PLS using first derivative FTIR–ATR spectra at 3100-2750 combined with 1500-663 cm and 32 scanning using sampling technique of attenuated total reflectance (ATR). Two calibrations in multivariate models, namely principle component (PCR) and partial least square (PLS) regressions were used to facilitate quantification of PO. -1 Conclusion: This result concluded that FTIR spectra in conjunction with DA (for classification) and PLS (for quantification) is fast and accurate tools during the analysis of PO as oil adulterant in EVOO and PSO. showed the best prediction models for quantification of PO in ternary mixtures with EVOO and PSO. Using this condition, correlation coefficient (R) values for the relationship between actual values and FTIR predicted values of 0.9967 and 0.9906 were achieved in calibration and validation models, respectively. The errors in calibration and prediction models, expressed by RMSEC and RMSEP, were low, i.e. 0.0080% and 0.0152%, respectively. DA using absorbance values at the same wavenumbers also offered the optimum discrimination model for discrimination between PO and PO mixed with EVOO and PSO in ternary mixtures.


INTRODUCTION
Known as the highest olive oils, extra-virgin olive oil (EVOO) command high priced vegetable oil because EVOO is only obtained by means of the physical and mechanical processes of fresh and healthy fruits of olive (Olea europaea) in the appropriate maturation stage without any treatments which cause its original composition changed [1,2]. As a consequence, EVOO is taken into accound as high-quality edible oil due to its nutritional benefits; the appreciated features, pleasant taste, and enjoyable flavor [3,4]. The evidence from epidemiological studies (especially from dietary intervention and cohort studies) trials revealed that the consumption of EVOO is associated with increased longevity [5]. Besides, EVOO is also reported in inhibiting some disease risk factors, such as diabetes, metabolic syndrome and obesity [6] and In fats and oils industry, both EVOO and PSO revealed advantageous effects to human health, desirable characteristics and limited production; therefore, they commanded expensive price which attracted to be adulterated intentionally with other edible oils [13]. Among the adulteration practices, the addition of EVOO with low price edible vegetable oils including palm oil and seed oils like soybean, corn, sunflower, pomace, and rapeseed oils is the most common fraud [5,14]. As a consequence, numerous methods based on chemical or molecular biology properties have been developed and used for analysis of oil adulterants in EVOO, and among the most reported ones is FTIR spectroscopy [15,16]. preventing certain cancers, especially breast and digestive tract cancers [7]. EVOO is rich phenolics compound which contributed to antioxidant activities and its stability [8]. In addition, pumpkin seed oil (PSO) is also considered as one of the functional edible oils [9]. PSO contained high levels of vitamin E, carotenoids and unsaturated fatty acids, contributing to beneficial effects on human health in prevention of high cholesterol levels, prostate cancer, and molacular degeneration [10,11]. As a consequence, PSO may also be adulterated with cheaper seed oil like palm oil [12].
The combination of FTIR spectroscopy with multivariate analysis or chemometrics has been used for authentication studies of EVOO from palm oil [17], corn, soybean, sunflower, rapeseed [18], corn, sunflower, soybean and hazelnut oils [19], and analysis of palm with canola oil [20] and lard in binary mixtures [21]. From literature, there are limited studies reporting the combination of FTIR spectroscopy and chemometrics employed for analysis of palm oil as adulterant in EVOO and PSO in ternary mixtures; therefore, this study was intended to analyze palm oil mixed with EVOO and PSO in ternary mixture.

Determination of fatty acids composition
The composition of fatty acids composed PO, PSO, and EVOO was determined by gas chromatography equipped with a detector of flame ionization (GC-FID). Before being analyzed using GC-FID, the studied oils were subjected to derivatization procedure to obtain derivate methyl ester of fatty acid (FAME). The derivatization procedure and GC-FID analysis were performed according to the literature [22]. The studied oils (50 µl) were added with 1.0 ml of nhexane and 0.2 ml of 0.2 N NaOCH3. The solutions were subjected to heating at 70 °C (10 min). Every 2 min, solutions were shaken vigorously, added with 1.5 ml of BF3, and heated at 70 °C (10 min). The solutions were cooled and followed by addition of 1.5 ml of saturated NaCl solution. The supernatant containing FAME was injected into GC-FID (GC-FID Agilent 7890B GC (Germany).

I
In nt te er rn na at ti io on na al l J Jo ou ur rn na al l o of f A Ap pp pl li ie ed d P Ph ha ar rm ma ac ce eu ut ti ic cs s Separation of analytes was performed using column HP-5 (30 m × 0.32 mm × 0.25 μm). The carrier gas used was ultrapure helium with the flow rate of 2.99 ml/min. The temperatures of injector and detector were set at 260 °C. The injection was carried out in split manner with ratio of 15:1. The column temperature was set in gradient manner as follows: 160 °C and hold for 2 min, 160-270 °C (10 °C/min rate), and finally hold for 5 min. In addition, flow rate of the air and hydrogen ultrapure were adjusted at 400 and 40 ml/min, respectively. The retention times of FAME standards (Sigma, Aldrich, USA) were used to identify FAMEs in studied oils. The percentage of fatty acid was computed as relative percentage (%) using internal normalization technique as in [23].

Preparation of calibration and validation samples
Preparation of calibration and validation samples was adopted from Rohman et al. [28]. For making calibration samples, 20 samples consisting of PO, EVOO, and PSO in ternary mixtures at concentration ranges of 0-100.0% v/v were prepared randomly using random number with the aid of Excel (Microsoft Inc., USA), as shown in table 1.
For preparing validation samples, a series of different samples was preparing in the concentration ranges covered by calibration samples. All samples were analyzed using FTIR spectrophotometer.

Discriminant analysis
Discriminant analysis (DA) for grouping samples was carried out, according to Putri et al. [29]. Discriminant analysis (DA) was used for discriminating between pure PO and PO in ternary mixture with PSO nd EVOO. For training sets, pure PO and ternary system of PO-PSO-EVOO were prepared in the range of 0.5-50.0% (v/v) of PO was prepared. Both groups (pure PO and PO-PSO-EVOO) were classified and discriminated using DA on the basis of Cooman's plot using FTIR spectral absorbance at specific wavenumbers as variables.

FTIR spectra acquisition
The scanning of FTIR spectra of oil samples was done using FTIR spectrophotometer (Thermo Scientific Nicolet iS10, Madison, WI), equipped with the Omnic software for spectral processing. The measurements were performed in horizontal attenuated total reflectance (HATR) at wavenumbers (1/λ) of 4000-650 cm -1 using 32 scannings with the resolution of 8 cm -1

Chemo-metrics analysis
. All FTIR spectra were corrected against FTIR spectrum of air as background. After every scan, a new reference air background spectrum was taken. These spectra were recorded as absorbance values at each data point in triplicate.
Chemometrics analysis including discriminant analysis and multivariate calibrations of principle component regression (PCR) and partial least square (PLS) was performed using TQ Analyst software version 9 (Thermo Fisher Scientific, Inc.). Statistical parameters used for evaluation of multivariate calibrations performance were coefficient of determination (R 2

RESULTS AND DISCUSSION
) and root mean square error of calibration (RMSEC), as well as root, mean square error of prediction (RMSEP). Table 2 compiled fatty acid (FA) composition of studied oils in which palmitic and oleic acids are main FAs composed PO. In addition, oleic acid was present as main FA component in EVOO, while PSO is mainly composed from linoleic and oleic acids. All FAs in PO and EVOO were in agreement with those specified in Codex Allimentarius Commission, while FAs in PSO in this study was in accordance with those reported in the literature [24]. These FAs data indicated that studied oils (PO, EVOO and PSO) were not previously mixed with other oils. The difference of fatty acids determined the different intensities of functional groups of studied oils in which oils with higher levels of oleic acids would have higher intensities in peaks related to unsaturated fatty acids. Due to its nature as fingerprint analytical technique, Fourier transforms infrared (FTIR) spectroscopy is an ideal method for analysis of complex mixtures such as edible fats and oils and other secondary metabolites extracted from specific part of plants [25]. Fig. 1 showed FTIR-ATR spectra of three edible oils, palm oil (PO), extra virgin coconut oil (EVOO), and pumpkin seed oil (PSO) scanned at mid-infrared regions corresponding to 1/λ of 4000-650 cm −1 . Each peaks and shoulders in FTIR spectra come from the IR absorption by functional groups present in these oils. The interpretation of each peaks and shoulders in relation to functional groups in FTIR spectra was compiled in table 3. All Peaks in FTIR-ATR among PO, EVOO and PSO spectra are very similar. This is not surprising because the main components composing edible fats and oils were triglyceride (95-98%), glycerol esterified with three fatty acids [27]. However, upon detailed investigation, FTIR-ATR spectra of PO, EVOO and PSO reveal slight differences in terms of band intensities (absorbance's) and the exact wavenumbers in each peak and shoulders [28].

AUTHORS CONTRIBUTIONS
Irnawati performed research activities and analyzed data. AR, SM and SR designed research, prepared manuscript and made critical thinking on manuscript.