Correcting for underlying absorption interferences in Fourier transform infrared trans analysis of edible oils using two-dimensional correlation techniques

摘要

Substantive improvement in the sensitivity of the AOAC/AOCS spectral ratioing method for the determination of isolated trans isomers in edible oils was recently achieved by the application of a new spectral reconstitution (SR) technique that facilitates the FTIR analysis of edible oils in the transmission mode. However, the general applicability of the spectral ratioing method is still severely limited by the requirement to know the provenance of the oil to be analyzed and have on hand its trans-free counterpart so that the underlying triacylglycerol absorptions in the trans measurement region (990-945 cm(-1)), henceforth referred to as UA(t), may be ratioed out. To eliminate the need for a trans-free reference oil, we have employed two-dimensional (2D) correlation spectroscopy to search for other spectral features that might correlate with and serve to estimate the UA(t). The three-dimensional contour maps obtained by 2D correlation analysis of the spectra of 10 trans-free oils of different oil types, recorded using the SIR procedure, revealed such correlations in two spectral regions, 1700-1600 and 4500-4300 cm(-1), exhibiting one maximum and two maxima, respectively, with wavenumber coordinates of (968, 4407), (968, 4299), and (968, 1650). The latter two correlations, when optimized, produced excellent linear regression relationships (r > 0.95) with the UA(t). The spectra of five sets of trielaidin-spiked oils were corrected for the UA(t) using these relationships, and their trans contents were predicted from the calibration equation generated for the spectral ratioing procedure. Linear regression of predicted versus added trans over the range of 0-1.6% trans, which is below the limit of quantitation of the AOAC/AOCS spectral ratioing method, yielded r = 0.88-0.90 with an SD of similar to 0.2% trans. These results indicate that the combination of the SR technique with the UA(t) correction approach may provide a simple and accurate FTIR method for the analysis of the trans content of fats and oils that would be competitive with GC.