Discrimination of edible oil products and quantitative determination of their iodine value by Fourier transform near-infrared spectroscopy

This work demonstrates the application of partial least squares (PLS) analysis as a discriminant as well as a quantitative tool in the analysis of edible fats and oils by Fourier transform near-infrared (FT-NIR) spectroscopy. Edible fats and oils provided by a processor were used to calibrate a FT-NIR spectrometer to discriminate between four oil formulations and to determine iodine value (IV). Samples were premelted and analyzed in gass vials maintained at 75°C to ensure that the samples remained liquid. PLS calibrations for the prediction of IV were derived for each oil type by using a subset of the samples provided as the PLS training set. For each oil formulation (type), discrimination criteria were established based on the IV range, spectral residual, and PLS factor scores output from the PLS calibration model. It was found that all four oil types could be clearly differentiated from each other, and all the validation samples, including a set of blind validation samples provided by the processor, were correctly classified. The PLS-predicted IV for the validation samples were in good agreement with the gas chromatography IV values provided by the processor. Comparable predictive accuracy was obtained from a calibration derived by combining samples of all four oil types in the training set as well as a global IV calibration supplied by the instrument manufacturer. The results of this study demonstrate that by combining the rapid and convenient analytical capabilties of FT-NIR spectroscopy with the discriminant and predictive power of PLS, one can both identify oil type, as well as predict IV with a high degree of confidence. These combined capabilities provide processors with better control over their process.     

Rapid and direct lodine value and saponification number determination of fats and oils by attenuated total reflectance/fourier transform infrared spectroscopy

A simple, rapid and reproducible method of determining the iodine value (IV) and saponification number (SN) for fats and oils was developed with an attenuated total reflectance/Fourier transform infrared spectrometer and commercially available triglycerides as calibration standards. Partial least squares was used to determine the spectral regions correlating with the known chemical IV and SN values, and the calibration set was augmented with additional standards generated by spectral co-adding techniques. The calibration model obtained was used to analyze commercially available fats and oils with a wide range of IV and SN values, and the results were compared to the values obtained by American Oil Chemists’ Society methods. With the spectrometer calibrated and programmed, IV and SN results could be obtained within 2–3 min per sample, a major improvement over conventional wet chemical methods.     

Trans fatty acid determination in fats and margarine by fourier transform infrared spectroscopy using a disposable infrared card

The use of a disposable polyethylene infrared (IR) card as a sample carrier for the quantitative determination of trans content of fats and oils and margarine by Fourier transform IR spectroscopy was investigated. Standards prepared by dissolving trielaidin in a zero-trans oil were used to develop partial least squares (PLS) calibrations for both the IR card and a 100-μm transmission flow cell. These calibrations were then used to predict a series of gas chromatographically-preanalyzed unknowns, the trans predictions obtained using the card being comparable to those obtained with the transmission flow cell. Somewhat improved performance could be obtained when the spectral data from the card were normalized to compensate for inherent variations in path length and variability in sample loading. Both IR methods tracked the gas chromatographic reference trans values very well. A series of margarine samples was also analyzed by the card method, producing results similar to those obtained using a flow cell. For the analysis of margarines, the card method has the advantage that the trans analysis can be performed directly on microwave melted emulsions because moisture is not retained on the card. Overall, the disposable IR card was shown to work well and has the benefit of allowing trans analyses to be carried out without requiring investment in a heated flow cell or attenuated total reflectance accessory.     

Determination of peroxide value in thermally oxidized crude palm oil by near infrared spectroscopy

A near infrared (NIR) spectroscopic method was developed to measure peroxide value (PV) in crude palm oil (CPO). Calibration standards were prepared by oxidizing CPO in a fermentor at 90°C. A partial least squares (PLS) calibration model for predicting PV was developed based on the NIR spectral region from 1350 to 1480 nm with reference to single-point baseline at 1514 nm. The optimization of calibration factors was guided by the predicted residual error sum of squares test. The standard error of calibration obtained was 0.156 over the analytical range of 2.17–10.28 PV and the correlation coefficient (R ²) was 0.994. The method was validated with an independent set of samples which was prepared in the same manner on a different day. A linear relationship between the American Oil Chemists’ Society and the NIR methods was obtained with R ² of 0.996 and standard error of performance of 0.17. This study has demonstrated that the prediction of PV in the NIR region is possible. The method developed is rapid, with total analysis time less than 2 min, is environmentally friendly, and its accuracy is generally good for quality control of CPO.     

Industrial validation of fourier transform infrared trans and lodine value analyses of fats and oils

A Fourier transform infrared (FTIR) edible oil analysis package designed to simultaneously analyze for trans content, cis content, iodine value (IV), and saponification number (SN) of neat fats and oils by using calibrations based on pure triglycerides and derived by application of partial-least-squares (PLS) regression was assessed and validated. More than 100 hydrogenated rapeseed and soybean samples were analyzed by using the edible oil analysis package as well as the newly proposed modification of the AOCS IR trans method with trielaidin in a trans-free oil as a basis for calibration. In addition, ∼1/3 of the samples were subsequently reanalyzed by gas chromatography (GC) for IV and trans content. The PLS approach predicted somewhat higher trans values than the modified AOCS IR method, which was traced to a combination of the inclusion of trilinolelaidin in the calibration set and the effects of baseline fluctuations. Eliminating trilinolelaidin from the triglyceride standards and the use of second-derivative spectra to remove baseline fluctuations produced excellent concurrence between the PLS and modified AOCS IR methods (mean difference of 0.61% trans). Excellent internal consistency was obtained between the IV and cis and trans data provided by the edible oil analysis package, and the relationship was close to that theoretically expected [IV=0.86 (cis + trans)]. The IV data calculated for the GC-analyzed samples matched the PLS IV predictions within 1 IV unit. The trans results obtained by both IR methods were linearly related to the GC data; however, as is commonly observed, the GC values were significantly lower than the IR values, the GC and IR data being related by a slope factor of ∼0.88, with an SD of ∼0.80. The concurrence between the trans data obtained by the two FTIR methods, and between the FTIR and GC-IV data, as well as the internal consistency of the IV, cis and trans FTIR predictions, provides strong experimental evidence that the edible oil analytical package measures all three variables accurately. Co-Director, McGill IR Group.     

Quantitative determination of hydroperoxides by fourier transform infrared spectroscopy with a disposable infrared card

Disposable polyethylene infrared cards (3M IR cards) were investigated for their suitability for the quantitative determination of peroxide value (PV) in edible oils relative to a conventional transmission flow cell. The analysis is based on the stoichiometric reaction of triphenylphosphine (TPP) with hydroperoxides to produce triphenylphosphine oxide (TPPO). Preliminary work indicated that the cards, although relatively consistent in their pathlength (±1%), had an overall effective pathlength variation of ±∼5%, caused by variability in loading of the oil onto the cards. This loading variability was reduced to <0.5% by developing a normalization protocol that is based on the peak height of the ester linkage carbonyl overtone band at 3475 cm⁻¹, which allowed one to obtain consistent and reproducible spectra. The standard PV calibration approach, based on the TPPO peak height at 542 cm⁻¹, failed because of unanticipated card fringing in the region where the measurements were being made. However, the development of a partial-least-squares (PLS) calibration provided a means of eliminating the interfering effect of the fringes and allowed the TPPO band to be measured accurately. An alternate approach to the standardized addition of TPP reagent to the oil was also investigated by impregnating the 3M IR cards with TPP, thus allowing the reaction to take place in situ. The spectral analysis protocols developed (normalization/calibration) were programmed to automate the PV analysis completely. The 3M card-based Fourier transform infrared PV methods developed were validated by analyzing oxidized oils and comparing the PV predictions obtained to those obtained in a 100-μm KCI flow cell. Both card methods performed well in their ability to predict PV. The TPP-impregnated 3M card method reproduced the flow cell PV data to within ±1.12 PV, whereas the method with an unimpregnated card was accurate to ±0.92 PV over the calibrated range (0–25 PV). Our results indicate that, with spectral normalization and the use of a PLS calibration, quantitative PV data, comparable to those obtained with a flow cell, can be provided by the 3M IR card. With the analytical protocol preprogrammed, the disposable 3M card provides a simple, rapid and convenient means of carrying out PV analyses, suitable for quality control laboratories, taking about 2–3 min per analysis.     

Quantitative determination of hydroperoxides by fourier transform infrared spectroscopy with a disposable infrared card

Disposable polyethylene infrared cards (3M IR cards) were investigated for their suitability for the quantitative determination of peroxide value (PV) in edible oils relative to a conventional transmission flow cell. The analysis is based on the stoichiometric reaction of triphenylphosphine (TPP) with hydroperoxides to produce triphenylphosphine oxide (TPPO). Preliminary work indicated that the cards, although relatively consistent in their pathlength (±1%), had an overall effective pathlength variation of ±∼5%, caused by variability in loading of the oil onto the cards. This loading variability was reduced to <0.5% by developing a normalization protocol that is based on the peak height of the ester linkage carbonyl overtone band at 3475 cm⁻¹, which allowed one to obtain consistent and reproducible spectra. The standard PV calibration approach, based on the TPPO peak height at 542 cm⁻¹, failed because of unanticipated card fringing in the region where the measurements were being made. However, the development of a partial-least-squares (PLS) calibration provided a means of eliminating the interfering effect of the fringes and allowed the TPPO band to be measured accurately. An alternate approach to the standardized addition of TPP reagent to the oil was also investigated by impregnating the 3M IR cards with TPP, thus allowing the reaction to take place in situ. The spectral analysis protocols developed (normalization/calibration) were programmed to automate the PV analysis completely. The 3M card-based Fourier transform infrared PV methods developed were validated by analyzing oxidized oils and comparing the PV predictions obtained to those obtained in a 100-μm KCI flow cell. Both card methods performed well in their ability to predict PV. The TPP-impregnated 3M card method reproduced the flow cell PV data to within ±1.12 PV, whereas the method with an unimpregnated card was accurate to ±0.92 PV over the calibrated range (0–25 PV). Our results indicate that, with spectral normalization and the use of a PLS calibration, quantitative PV data, comparable to those obtained with a flow cell, can be provided by the 3M IR card. With the analytical protocol preprogrammed, the disposable 3M card provides a simple, rapid and convenient means of carrying out PV analyses, suitable for quality control laboratories, taking about 2–3 min per analysis.     

Rapid determination of cis and trans content, iodine value, and saponification number of edible oils by fourier transform near-infrared spectroscopy

Fourier transform near-infrared (FT-NIR) spectroscopy was evaluated as a means of simultaneously determining the cis and trans content, iodine value (IV), and saponification number of neat fats and oils. Reference values for these parameters were obtained from oils using a previously developed mid-FTIR Edible Oil Analysis Package. Two partial least squares calibrations were developed for a 5-mm heated flow cell, the first a process calibration based on hydrogenated soybean samples and the second a more generalized calibration based on an oil samplematrix containing many oil types and designed to remove any correlations among the parameters measured. Each calibration performed well with its own validation samples; however, only the noncorrelated calibration was able to analyze oil samples accurately from a variety of sources. It was found that NIR analysis maintained the internal consistency between cis/trans and IV, and the accuracy and reproducibility of the predictions were on the order of ±1.5 and ±1.0 units, respectively, for all parameters evaluated. FT-NIR is shown to be a very workable means of determining cis/trans/IV values and saponification number for edible fats and oils, and it provides a rapid alternative to the commonly used chemical and physical methods presently employed in the industry.