Quantitative Fourier transform infrared analysis for anisidine value and aldehydes in thermally stressed oils

Abstract

AbstractA Fourier transform infrared (FTIR) transmission‐based spectroscopic method was investigated for the simultaneous monitoring of aldehyde formation and the determination of anisidine value (AV) in thermally stressed oils. Synthetic calibration standards were prepared by adding known amounts of hexanal,t‐2‐hexenal andt,t‐2,4‐decadienal to canola oil (these compounds considered representative of aldehydic compounds formed during oxidation) plus random amounts of other compounds representative of oxidation by‐products. The standards were analyzed for their chemical AV. With the partial least squares (PLS) technique, an FTIR spectrometer was calibrated to predict both the concentrations of individual aldehyde types and AV, with the individual aldehyde contributions being related to the chemical AV by multiple linear regression to derive “apparent” AV values. The predictive capability of the PLS calibrations was assessed by analyzing canola oils that were thermally stressed at 120, 155, and 200°C. The apparent AV, predicted for these samples, matched the chemical AV values within ±1.65 AV units. A PLS calibration also was derived by using thermally stressed samples as calibration standards. This approach provided similar predictive accuracy as the use of synthetic calibration standards. As such, quantitative determination of AV by FTIR spectroscopy was shown to be feasible, and the synthetic calibration approach provided additional information on the aldehyde types present in a sample and allowed the use of a simple gravimetric approach for calibrating an FTIR spectrometer. This study provides the basis for the development of a rapid, automated FTIR method for the direct analysis for AV of thermally stressed fats and oils in their neat form without the use of chemical reagents. The implementation of such a method as a quality control tool would eliminate the use and disposal of hazardous solvents and reagents, required by the conventional chemical method, and drastically reduce analysis time (∼2 min/sample). Possible applications include monitoring of the oxidative state of frying oils or evaluation of oxidative stability of biodegradable lubricants.