NON-DESTRUCTIVE ASSESSMENT OF LIPID OXIDATION IN TREE NUTS USING NEAR-INFRARED SPECTROSCOPY

Abstract

This study aimed to evaluate the performance of prediction models for assessing lipid oxidation in tree nuts using near-infrared spectroscopy (NIRs). A total of 86 samples of tree nuts, including walnuts, macadamias, and pecans, were randomly selected and scanned using a Fourier transform near-infrared (FT-NIR) spectrometer over the wavenumber range of 4,000-12,500 cm^-1. The acquired spectra were preprocessed using chemometric methods and the samples were subsequently analyzed for moisture content (MC), peroxide value (PV) and free fatty acid (FFA) as reference quality parameters. Partial least squares regression (PLSR) was then applied to establish the relationship between the spectral data and the reference values. The results showed that spectral preprocessing using first derivative combined with standard normal variate (SNV) or multiplicative scatter correction (MSC), combined with wavelength selection, improved the prediction performance of the quality models. The best PLSR models for MC, PV and FFA yielded coefficient of determination (R²) values of 0.918, 0.973 and 0.764, and ratio of standard error of performance to standard deviation (RPD) values of 3.49, 6.07 and 2.03 for MC, PV and FFA, respectively. These findings demonstrate that NIRs, combined with appropriate spectral preprocessing, has the potential to serve as a rapid, non-destructive screening tool for quality assessment of tree nuts.

Keywords: Tree Nuts, Lipid Oxidation, Rancidity, Near-Infrared Spectroscopy (NIRs)