SAMPLING AND ANALYSIS OF COMMERCIAL FATS AND OILS

APPARATUS

1. FTIR spectrometer—It should be equipped with a deuterium triglycine sulfate (DTGS) or mercury cadmium telluride (MCT) detector, capable of making measurements at 4 cm^–1 resolution in the spectral range covering 1050-900 cm^–1. Its performance must meet the following criteria. In the absence of a test sample, a 3 min data collection at 4 cm^–1 resolution must yield between 1050-900 cm^–1 a peak-to-peak noise level < 0.0005 absorbance unit (AU) for absorption spectra (Note 2).

2. Heated ATR infrared cell—It should be equipped with an internal reflection element made of zinc selenide (ZnSe), diamond, or equivalent with capacity of approximately 1–10µL. It must be capable of maintaining a constant temperature of the sample of (65 ±1)°C.

3. Steam water bath.

4. Vials with screw caps.

5. Low-lint tissue paper.

6. Disposable plastic pipets.

7. 10 mL Beakers or vials.

REAGENTS

Unless otherwise stated, use only reagents as specified in ISO 6353 (parts 2 and 3), if listed in Reference 3.

1. Reference trielaidin (TE) with purity of >99% from commercial sources (e.g., Nu-Check-Prep, Elysian, MN, Matreya, Inc., Pleasant Gap, PA, and Sigma Chemical Co., St. Louis, MO) for preparing calibration standards.

2. A high purity saturated tripalmitin (TP) trans-free reference oil with purity of >99% from same commercial sources for preparing TE calibration standards. It is noted that the impurities in TP are usually other saturated trans-free oils

PROCEDURE

1. Prepare neat (without solvent) trans reference TE calibration standards by weighing accurately to the nearest 0.0001 g, (0.3–x) g of a saturated TP trans-free reference oil, and x g of TE, into a 10 mL beaker or vial, where x equals 0.0015, 0.0030, 0.0060, 0.0150, 0.0300, 0.0450, and 0.0600 g in order to prepare 0.5, 1, 2, 5, 10, 15, and 20% (as percent of total fat) trans calibration standards, respectively.

2. Allow the ATR cell to warm up to about 65 (±1)°C for 30 min.

3. Place no sample on the clean ATR horizontal surface, collect a 256-scan single-beam FTIR spectrum at 4 cm–1 resolution, and save it. This is the open beam (air) reference single beam background spectrum.

4. Place each of the TE calibration standards on the ATR horizontal surface making sure that the surface of the ZnSe or diamond crystal is completely covered (Note 3), collect a 256-scan single-beam FTIR spectrum at 4 cm^–1 resolution, and save it.

5. After each measurement, clean the ATR crystal by thoroughly wiping it with low-lint tissue papers as many times as needed. To insure absence of cross contamination, apply the next test sample to be analyzed, and then clean the crystal again with low-lint tissue papers as many times as needed (Note 4).

6. (a) Place each of the trans fat test samples on the ATR horizontal surface making sure that the surface of the ZnSe or diamond crystal is completely covered (Note 2), collect a 256-scan single-beam FTIR spectrum at 4 cm^–1 resolution, and save it. (b) Repeat step 5. (c) Measure each test sample in duplicate (repeat steps 6a and 6b).

7. Electronically generate the absorbance spectra for all the TE reference standards. This is carried out electronically by “ratioing” each of the saved TE single-beam spectra against that of the reference single beam spectrum. Electronically convert each absorbance spectrum to its negative second derivative spectrum ( –2^nd der) (References 4–7). Save each –2^nd der spectrum.

8. Electronically generate the absorbance spectra for all the trans fat test samples. This is carried out electronically by “ratioing” each of the saved trans fat single-beam spectra against that of the reference single beam spectrum. Electronically convert each absorbance spectrum to its –2^nd der spectrum (References 4–7). Save each –2^nd der spectrum.

9. For each of the TE reference standards, display the –2^nd der spectrum (see Figs. 1–4) in the expanded wavenumber range 1030-880 cm^–1, and record the height of the –2nd der trans band at 966 cm^–1 (between zero and maximum band height, see Fig. 2). Tabulate the data.

10. For each of the trans fat test samples, display the –2^nd der spectrum (see Figs. 1–4) in the expanded wavenumber range 1030-880 cm^–1, and record the height of the –2^nd der trans band at 966 cm^–1 (between zero and maximum band height, see Fig. 2). Tabulate the data (Note 5, Figure 4).

CALCULATIONS

1. Create a calibration curve by performing a linear regression analysis of the height of the –2^nd der trans band at 966 cm^–1 versus the amount of TE (as percent of total fat) in the trans calibration standards. Calibration curves should be checked periodically to insure that they had not shifted.

2. Using the slope and intercept of the linear regression equation generated for TE calibration standards, calculate the trans level (as percent of total fat) for each fat or oil test sample by substituting the height of the –2^nd der trans band into the equation: %Trans = [ height—Intercept ] / Slope. Report results to the nearest 0.01%. This equation assumes that trans test samples consist of TE.

METHOD PRECISION

 

 

QUALITY ASSURANCE AND CONTROL

Blank test sample: The first test sample in an analysis batch is always a blank, namely air. This is the blank test sample that should give no detectable bands relative to the reference single beam spectrum (see Procedure, step 3). Repeat this test after every 10 test samples.

NUMBERED NOTES

1. The present negative second derivative method was developed to overcome the high variability in precision measurements obtained for trans fats and oils levels below 5% (as percent of total fat) with infrared Official Methods AOCS Cd14d-99 (Reference 1) and AOAC 2000.10 (Reference 2).

2. To achieve a minimum signal-to-noise ratio (SNR) of 10:1, a high sensitivity linearized MCT detector operating at liquid nitrogen temperature would be recommended when a single-reflection ATR crystal is used. For room temperature DTGS detectors, a 3-reflection (or higher) ATR crystal would be recommended.

3. ATR sampling—It is essential to ensure that the test portion of the fat being analyzed completely covers the horizontal surface of the ATR crystal for the quantitative determination to succeed. In rare cases, after placing a fat test portion on the ATR crystal, the melted or liquid fat beads up and partially rolls off the surface of the ATR crystal. The only recourse the analyst would have is to try again.

4. Cleaning the ATR crystal—Thoroughly wiping the horizontal surface of the ATR crystal with low-lint paper has been the method of choice for cleaning it without the use of any solvent. However, to insure the complete removal of a test sample the analyst should also apply the subsequent test portion to be analyzed, and clean the crystal once again. Additionally, the analyst could make an infrared measurement after cleaning the crystal; the absence of a fat spectrum would confirm that the crystal is clean; this last step is particularly useful when a test sample having a relatively low level of trans fat is measured after another unknown test sample that is high in trans fat. However, for accurate background measurement, more precautions are required: Use a lint- and glue-free cotton swab with a very small amount of ethanol, then a dry cotton swab. A blank sample should be run to confirm the removal of all residues.

5. If a spectral feature is observed at slightly lower wavenumbers, near or below 960 cm^–1 (Figure 4), it should not be attributed to a band for isolated trans double bonds (reference 5).

REFERENCES

1. AOCS Official Method Cd14d-99, Rapid determination of isolated trans geometric isomers in fats and oils by attenuated total reflection infrared spectroscopy.

2. AOAC Official Method 2000.10, Determination of total isolated trans unsaturated fatty acids in fats and oils. ATR-FTIR spectroscopy.

3. ISO 6353, Reagents for Chemical Analysis, Part 2 (1983) and 3 (1987); Specifications.

4. M. Milosevic, V. Milosevic, J.K.G. Kramer, H. Azizian, and M.M. Mossoba, Lipid Technology, 16:252–264 (2004).

5. M.M. Mossoba, J.K.G. Kramer, M. Milosevic, V. Milosevic, and H. Azizian, (2007), J Am Oil Chem Soc, 84:339-342.

6. M.M. Mossoba, V. Milosevic, M. Milosevic, J.K.G. Kramer, H. Azizian, (2007), Analytical Bioanalytical Chemistry (special issue on Food and Dietary supplements), 389, 87-92.

7. M.M. Mossoba, A. Seiler, J.K.G. Kramer, V. Milosevic, M. Milosevic, H. Azazian, H. Steinhart, (2009), J Am Oil Chem Soc, 86:1037-1045.

8. AOCS Official Method Ce1h-05, Determination of cis-, trans-, saturated, monounsaturated and polyunsaturated fatty acids in vegetable or non-ruminant animal oils and fats by capillary GLC.

 

Figure 1.
Negative second derivative (top) and Absorption (bottom) spectra for a test sample consisting of trielaidin in tripalmitin with a trans level of 12.58% (as percent of total fat). A vertical line indicates the position of the unique band due to isolated trans double bonds at 966 cm^–1. The second derivative spectrum was multiplied by –1 only to have the bands point upwards for convenience.

 

Figure 2.
The region of the spectra that exhibit the deformation band for isolated trans double bonds at 966 cm^–1 is expanded for the Negative second derivative (solid line) and Absorption (dotted line) spectra for a test sample consisting of trielaidin in tripalmitin with a trans level of 12.58% (as percent of total fat). The height of the negative second derivative band, as indicated by the vertical arrow, can be accurately measured from a horizontal baseline. It is noted that several weak bands observed in the same spectral region are also more pronounced in the –2nd derivative spectrum (narrower bandwidths) than in the absorption spectrum.

 

Figure 3.
Expanded spectral region that exhibits the –2nd derivative of the deformation band for isolated trans double bonds at 966 cm^–1 for representative test samples covering approximately the 1–12% range and containing trans fat levels determined by IR to be for Lard 1.21%, and Canola oil mixtures 2.21%; 4.20%; 4.72%; 7.35%; 9.11%; and 12.62%, as percent of total fat. The height of the –2nd derivative band can be easily measured from the horizontal baseline (dotted line).

 

Figure 4.
Expanded spectral region that exhibits the –2nd derivative of the deformation band for isolated trans double bonds at 966 cm^–1 for several test samples containing trans fat (solid lines), as well as for coconut oil (dotted line) that is high in saturated fat and contains only a trace (approximately 0.1%, reference 8) of trans fat. Coconut oil exhibited a spectral feature at slightly lower wavenumbers, near 960 cm^–1, which is easy to misidentify as a band for isolated trans double bonds (reference 5).