Printer Friendly

Quantification of Pure Refined Olive Oil Adulterant in Extra Virgin Olive Oil using Diamond Cell ATR-FTIR Spectroscopy.

Byline: _

Summay: The present study depicts spectroscopic method development to deliver a rapid simple and reproducible quantification of pure refined olive oil (PROO) adulterant in extra virgin olive oil (EVOO) using partial least square (PLS) regression (statistical parameter). Single bounce attenuated total reflectance (SB-ATR) Fourier transform infrared (FTIR) was choice in the developed method. Blended standards of PROO and EVOO were obtained by their weight by weight percentage and the values were used to construct calibration curves for quantification. The optimum regression values (i.e. greater than 0.99) were achieved using the combined frequencies of 3105-2761 1838-1687 and 1482-440 cm-1 with regression coefficients (R2) 0.99718 and achieved residual mean square error of calibration (RMSEC) 1.40% w/w. To determine the suitability of developed method principal component spectra (PCS) diagnostic was also used.

The results of the present study prove that the developed methods reported in preceding studies can be good option for more rapid and accurate determination of PROO adulteration in EVOO.

Key words: Extra virgin olive oil Adulteration FTIR Chemometrics.

Introduction

The International Olive Oil Council Olive oil can be classified into various grades [1]. Of those mainly are virgin olive oils (i.e. ordinary virgin olive oil extra virgin olive oil and virgin olive oil) olive pomace oils (i.e. crude olivepomace oil refined olivepomace oil and olivepomace oil) and refined olive oils. The significant difference among the prices of oils has motivated adulteration of costly oils with cheaper oils. Though such mixing of cheaper oils in costly oils does not cause anything that may be associated with health problem however the primary consumer is deprived of his rights which are violated by such deceiving practices [2]. Best example of the said practice is adulteration of extra virgin olive oil (EVOO) which is highly priced and mostly adulterated by mixing of low-grade olive oils olive pomace oil or refined olive oil as well as other cheaper vegetable oils such as hazelnut oil sunflower oil soybean oil and maize oil [3].

Thus the quantification of adulterants of EVOO has been desired in scenarios after highlighted above.

Detection of refined olive oil and pomace oil adulteration in EVOO often becomes difficult to accomplish especially when oils with chemical similar compositions are added [4]. As a result it was conceived that new methods should be developed for the determination of adulteration of EVOO.

Different analytical methods have been employed to detect adulterants virgin in EVOO. Most of these are based on chromatography (gas chromatography high performance liquid chromatography) spectroscopy (ultraviolet near- infrared (NIR) mid-infrared visible Raman) isotopic analysis and electronic nose systems [5-8]. Relevant applications of the chemometric techniques and electronic nose have also been reviewed [3 9- 11]. It is suggested that using chemometric analysis in the NIR adulteration of pure olive oil with soybean sunflower corn walnut and hazelnut oils could conveniently be predicted [5]. Qualitative and quantitative determination of vegetable oils (canola hazelnut pomace and high linoleic/oleic sunflower) as adulterants in commercial samples of EVOO has been reported [12].

In the study of edible fats and oils FTIR spectroscopy has been used as a powerful analytical tool especially for qualitative characterization of specific components in foods [13].

The use of diamond cell ATR-FTIR spectroscopy for the quantification of pure refined olive oil (PROO) adulterant in EVOO has been the approach in this study. The assessment of the capability of diamond cell ATR-FTIR coupled with Turbo Quant (TQ) Analyst chemometrics such as partial least square (PLS) and principal component spectra (PCS) to discriminate the EVOO mixed with PROO was the main objective of this study. This approach displays a facile and convenient means for monitoring EVOO quality. The advantages of this technique are ease of operation high sample turnover and no sample pretreatment.

Results and Discussion

A unique spectral fingerprinting of the infrared spectrum of organic molecules provides detailed information about their molecular structure. However this unique fingerprint becomes confusing when similar molecules containing structural features are mixed to each other e.g. in fats and oils complex mixture of triacylglycerols [14 15].

Fig. 1-A is representative of first derivative spectrum of EVOO and PROO standards and resultant spectrum obtained by the subtraction of PROO from EVOO with a spectral math parameter shows the clear dissimilarity between characteristic bands. The band at 3005 cm-1 shows (CH) stretching vibration of cis- double bond of unsaturatured fatty acids Fig. 1-B while symmetric and asymmetric vibration of aliphatic CH2

hydrocarbon chains represents the characteristics bands at 2922 and 2853 cm-1 [15]. Major peak at 1744 cm-1 arises from C=O stretching vibrations the peak is associated with the triglyceride ester-linkage (COOR) band and the C=O absorption of free fatty acid present in the olive oil. The band at 1461 cm-1 is attributed to asymmetric stretching in methyl and methylene groups while the peak at 1160 cm-1 is associated with the stretching of the CO bonds of aliphatic esters [16 17]. The finger print region plays a very important role in the identification of the variation among the bands.

Table-1 lists the parameters used to statistically assess the results of calibration model to determine PROO percentage in EVOO by using normal mid-ATR-FTIR spectra. The calibration model performed properly yielding good correlation coefficients and low residual mean standard error of calibration (RMSEC) values. The assessment of the errors was carried out by calculating RMSEC in the calibration model after comparing the actual concentration with those computed for each component.

Table-1: Abilities of calibration and prediction model for PROO adulterant in EVOO by ATR-FTIR.

###Spectral range

###Region type###Base line type###Factors###R2###RMSEC

###(cm-1)

###One point

###Average in range

###3105-2761

###First Derivative###One point

###1838-1687###4###0.99718###1.40

###in range###Fixed location

###1482-440

###Two points

###Fixed Location

###One point

###Average in range

###3105-2761

###Spectrum in###One point

###1838-1687###4###0.99178###2.38

###range###Fixed location

###1482-440

###Two points

###Fixed location

###Two points

###3141- 2819

###Spectrum in###Average in range

###1484- 493###4###0.95558###5.09

###range###Two points

###Average in range

###One point

###3105- 2761

###Spectrum in###Average in range

###1838-1687###2###0.8816###6.78

###range###One point

###Fixed location

###Spectrum in###1482-440###Two points

###3###0.99073###2.03

###range###Fixed location

Fig. 2-A and 2-B represent the ATR-FTIR spectra of normal and first derivative of 12 blended oils respectively which clearly show the variation in the absorption bands could be related to compositional differences among oil groups. The spectra did not show an obvious difference from visual inspection according to the varietal regions. However PLS algorithm can easily predict these minor variations in the spectrum.

Full region (4000-450 cm-1) was selected prior to three different selective regions (3105-2761) (1838-1687) (1482-440) of the mid IR spectrum which were taken to construct PLS calibration individually each region does not provided satisfactory results in term of determination coefficient (R2) and RMSEC. However combined frequencies of these regions (3105-2761 1838-1687 and 1482-440) and the baseline types for these spectral ranges were optimized as one point (average in range) one point (fixed location) and two points (fixed location) respectively with first derivative in each region also selected. The determination coefficient (R2) RMSEC for calibration set were 0.99718; 1.40 respectively therefore the values of R2 RMSEC of developed calibration proves the simplicity of method. Correspondingly for calibration model these values were within the acceptable range

Table-2 shows the mean percentage adulteration of PROO in EVOO of eight commercial extra virgin olive samples determined by ATR-FTIR spectroscopy. Amongst the analyzed EVOO samples the highest amount was determined in CS-4 (26.19%) while lowest in the CS-3 5.44% whereas sample CS-6 to CS-8 were lower from the detection range of the calibrated method.

Table-2: ATR-FTIR determination of PROO adulteration in EVOO of eight commercial samples

Samples

###Percentage by FTIR

CS-1###11.23 0.25

CS-2###16.12 0.35

CS-3###5.44 0.16

CS-4###26.19 0.42

CS-5###23.26 0.64

CS-6###2.3/ND

CS-7###1.5/ND

CS-8###3.2/ND

The spectral information can be used as a measurable property which could make possibility of establishing a calibration thus advantage of using advanced chemometric techniques such as PLS is there [14]. The information provided by the calibration results diagnostic can help in identifying standards that may be outliers. A typical percentage difference plot will show data points distributed randomly above and below the zero line within a narrow concentration range.Fig. 3 presents the calibration plot and percentage difference plot between actual and predicted values. In developing the PLS model the percent values for standard oils obtained from pre- constituted mixtures of EVOO with PROO (w/w) were put along with the spectra into the Turbo Quant (TQ) Analyst program. At the time of the optimization process the combined frequency regions of 3105-2761 1838-1687 and 1482-440 cm-1 were selected. The developed calibration model offers highest values of R2 and lowest value of RMSEC.

Principal component spectra (PCS) diagnostic has also been employed. Ten PCS of blending samples/standards were obtained using the advanced diagnostic option in Turbo Quant Analyst software (Fig. 4). These noisy or featureless PCS indicate that the corresponding (and any subsequent) principal component contributes little useful information to the calibration model. The PCS show how the spectral information in a calibration set is represented by the principal components. PCS is the orthogonal spectrum that represents the amount of variability described by a principal component measured across the entire spectral range of the standards. The data obtained from these spectra were put into the Microsoft excel software to obtain a calibration plot between percentage variance against cumulative percentage variance with R2 value at 0.973 which further confirmed the reliability and accuracy of data (Fig. 5).

Experimental

Samples and Reagents

EVOO and PROO were bought from the local industry in Karachi Pakistan. The oil samples were stored in glass bottles in dark before being used for analysis. The dates of manufacturing and expiry of samples were also mentioned. All chemicals (e.g. reagents and solvents) to be used in the study were purchased from E. Merck (Darmstadt Germany).

Blending of Oils for ATR-FTIR Analysis

PROO was added into EVOO in the range of 560% w/w at interval of 5 units (Table-3). The blended samples were kept in controlled room temperature 25 0C during authentication studies.

Table-3: Pre-constructed blending (w/w) of PROO

in EVOO.###

###%###EVOO PROO

Samples###x=A+B###(A/x)A-100###(B/x)A-100

###Blending (A)###(B)

###1###5%###90.50 05.51###100.01###94.90###5.10

###2###10%###90.02 10.05###100.07###89.56###10.44

###3###15%###80.51 10.52###100.03###84.87###15.13

###4###20%###80.03 20.02###100.05###79.88###20.12

###5###25%###70.52 20.52###100.04###74.89###25.11

###6###30%###70.05 30.01###100.06###70.04###29.96

###7###35%###60.54 30.51###100.05###65.05###34.95

###8###40%###60.02 40.01###100.03###60.01###39.99

###9###45%###50.52 40.50###100.03###55.08###44.92

###10###50%###50.02 50.01###100.03###50.04###49.96

###11###55%###40.55 50.51###100.07###45.24###54.76

###12###60%###40.03 60.02###100.04###40.11###59.89

FTIR Spectral Measurements

Infrared spectra of the blended samples were recorded on a Thermo Nicolet Avatar 320 FTIR spectrometer. It was equipped with a Diamond Cell Smart Accessory (ID: 060-5013) which was removable. The detector was deuterated triglycine sulfate (DTGS) and KBr optics. For data acquisition and instrument control OMNIC software version 7.0 (Thermo Nicolet Analytical Instruments Madison WI) from Thermo was employed. All spectra were collected by co-addition of 32 scans at a resolution of 4 cm-1 in the range of 4000400cm-1 at 1.93 data spacing. The spectrum of each standard or sample was ratioed against a fresh background spectrum recorded from the uncovered removable diamond crystal. All analyses were carried out at room temperature and three spectra were recorded for each sample. ATR crystal was carefully cleaned with a cellulose tissue soaked in n-hexane and then rinsed

with acetone to remove any lipo- or hydrophilic residues of previous sample. The main benefits of using a diamond cell ATR smart accessory is its simplicity in handling It only requires a sample to be placed on the crystal and the spectrum is taken against the fresh background of the clean crystal.

Conclusion

The present approach indicated that the diamond cell ATR-FTIR spectroscopy can be a suitable tool for the determination of PROO adulterant in EVOO oil samples. No costly standard reagent and chemicals are required in applying the developed method for analysis of samples. Thus it is concluded that the method is simple sensitive and reproducible after the stabilization of the instrument under optimized environmental conditions especially temperature and humidity. For the determination of adulteration in oils the proposed method could be easily applied.

References

1. S. Mildner-Szkudlarz and H. H. Jelen Journal of Food Quality 33 21 (2010).

2. F. Ulberth and M. Buchgraber European Journal of Lipid Science and Technology 102 687 (2000).

3. M. C. C. Oliveros J. L. P. Pavon C. G. Pinto M. E. F. Laespada B. M. Cordero and M. Forina Analytica Chimica Acta 459 219 (2002).

4. D. L. Garcia-Gonzalez and R. Aparicio Lipid Technology 18 81 (2006). A. A. Christy S. Kasemsumran Y. Du and Y. Ozaki Analytical Sciences 20 935 (2004).

5. F. PeAa S. Cardenas M. Gallego and M. Valcarcel Journal of Chromatography A 1074 215 (2005).

6. G. Flores M. L. Ruiz del Castillo G. P. Blanch and M. Herraiz Food Chemistry 97 336 (2006).

7. G. Flores M. L. Ruiz del Castillo M. Herraiz and G. P. Blanch Food Chemistry 97 742 (2006).

8. M. Lorenzo J. L. P. Pavon M. E. Laespada C. G. Pinto and B. M. Cordero Journal of Chromatography A 945 221 (2002).

9. S. Biswas K. Heindselmen H. Wohltjen and C. Staff Food Control 15 19 (2004).

10.S. Mildner-Szkudlarz and H. H. Jelen Food Chemistry 110 751 (2008).

11.R. M. Maggio L. Cerretani E. Chiavaro T. S. Kaufman and A. Bendini Food Control 21 890 (2010).

12.R. R. Stewart In Flora of West Pakistan An Annotated Catalogue of the Vascular Plants of West Pakistan and Kashmir Fakhri Press Karachi p. 223

13.O. J. Ress In Fourier Transform Infrared Spectroscopy (2010): development techniques and application in the analyses of fats and oils Nova Science Publishers New York pp. 1-36.

14.S. T. H. Sherazi M. Y. Talpur S. A. Mahesar A. A. Kandhro and S. Arain Talanta 80 600 (2009).

15.S. A. Mahesar S. T. H. Sherazia A. A. Kandhro M. I. Bhanger A. R. Khaskheli and M. Y. Talpur Vibrational Spectroscopy 57 177 (2011).

16.G. Downey P. McIntyre and A. N. Davies Applied Spectroscopy 57 158 (2003). A. Bendini L. Cerretani F. Di Virgilio P. Belloni M. Bonoli-Carbognin and G. Lercker Journal of Food Quality 30 424 (2007).

17.N. Sinelli M. Casale V. Di Egidio P. Oliveri D. Bassi D. Tura and E. Casiraghi Food Research International 43 2126 (2010).
COPYRIGHT 2014 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Journal of the Chemical Society of Pakistan
Article Type:Report
Date:Aug 31, 2014
Words:2541
Previous Article:Ammonia Nitrogen Removal by Single-stage Process: A Review.
Next Article:Optimization of Dissolution of Ulexite in Phosphate Acid Solutions.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters