Advances in infrared technology have made it a promising method for the food industry. The development of the Fourier Transform (FT) algorithm has shortened scan time, improved signal-to-noise ratio, and improved the accuracy of wavelength. Innovations in personal computing has made FT-infrared (FTIR) spectroscopy common place in quality control labs. Analysts routinely combine FTIR spectra with multivariate analysis to quantify components in their matrix.
Food scientists have combined FTIR with chemometrics to detect adulteration in their products. Both near- (NIR) and mid- (MIR) infrared have proven to be valuable resources for rapidly authenticating the quality of food. NIR has been applied to products such as juice, honey and milk, and for detecting acrylamide in potato chips. MIR has been successful in authenticating edible oils, juices, honeys, and for ensuring products are correctly identified as organic. As food safety continues to be a top priority for consumers, there is no doubt that improvements will continue to be made on spectrometers and in chemometric methods in order to stay one step ahead of adulterators.
One such improvement has been the development of handheld portable infrared sensors. These units are used for chemical identification in the homeland security, public safety, pharmaceutical, industrial, and medical markets. They bring the spectral resolution of benchtop instruments to field applications in rugged, battery operated units weighing less than four pounds.
A novel application of a handheld FTIR is for monitoring oxidation in edible oils. Methods currently used for oil quality testing are subjective, time consuming, and use hazardous solvents which then need to be disposed of. The aim of this research was to evaluate the capabilities of a handheld FTIR combined with multivariate analysis to characterize frying oils and to monitor chemical processes occurring during lipid oxidation as well as determining fatty acid composition.
Commercial frying oils (corn, peanut, sunflower, safflower, cottonseed and canola) were heated to 65°C in an oven for thirty days to accelerate oxidation. Reference methods included fatty acid composition (IUPAC 2301, 2302), peroxide values (PV, AOCS Cd 8-53) and free fatty acid (FFA, Shipe 1979). Aliquots were drawn at five day intervals and analyzed by benchtop and handheld mid-infrared devices and reference methods. Spectral analysis (Soft Independent Model of Class Analogy (SIMCA) and Partial Least Squares Regression (PLSR)) was carried out by pattern recognition software.
Mid-infrared spectral regions ~3000-2825 cm-1 (C-H stretching bands) and 1740 cm-1 (C=O stretching of esters) were important for classification. All six oil samples formed distinct and well-separated clusters in the SIMCA plot due to difference in their chemical composition. It was found that spectral variability could be minimized by controlling oil temperature (65°C) during data acquisition. PLSR showed good correlation coefficients (Rval) between FFA and PV on the infrared devices.
Handheld IR instruments combined with multivariate analysis showed promise for determination of oil quality parameters with similar performance as the benchtop units. Its portability and ease-of-use make the handheld IR a great alternative to traditional testing methods.