The effects of xylose, ribose, glucose, fructose, and galactose on the structural and flavor properties of Maillard reaction products (MRPs) derived from enzymatically hydrolyzed soybean meal were investigated. The extent of browning and ultraviolet (UV) spectroscopy analysis indicated that pentose exhibited higher reactivity than hexose, and aldose demonstrated greater reactivity than ketoses. Fluorescence spectroscopy analysis revealed an increase in the fluorescence intensity of MRPs, while infrared spectroscopy indicated alterations in absorption peak intensities of MRPs, compared with soybean meal hydrolysate. The molecular masses of MRPs were mainly distributed in the ranges of 128-500 and 500-1000 Da. In addition, the contents of bitter amino acids and umami amino acids in ribose-derived MRPs (R-MRPs) were relatively high, at 9.29 and 3.60 mg/g, respectively. Pentose-derived MRPs exhibited superior umami value, richness value and lower bitterness compared with hexose-derived MRPs. Furans, pyrazines, ketones, and aldehydes were the most abundant volatile substance in R-MRPs, presenting stronger caramel, nutty and grilled meat-like aromas. Furthermore, strong-flavored soybean oil was successfully prepared using the ribose-soybean meal hydrolysate (SBMH) system, indicating that while the system imparted a rich flavor to the oil, the fatty acid composition and physicochemical indices of the oil remained stable and safe.

To meet the different functional needs for soybean protein isolate (SPI) in different food applications, this study utilized infrared spectroscopy to rapidly analyze 70 SPI samples subjected to different pH treatments, and explored the effect of pH changes on the secondary structure content of SPI. Mean centralization (MC), multivariate scattering correction (MSC), standard normal variate transformation, and normalization were used for infrared data preprocessing. Feature wavebands were identified based on two-dimensional correlation infrared spectra, and predictive modeling of the secondary structure content of SPI against pH was performed using partial least squares (PLS) and arithmetic optimization algorithm-random forests (AOA-RF). The results showed that the relative standard deviations of the α-helix, β-sheet, β-turn, and random coil prediction models developed by the combined use of MC and MSC were 1.29%, 1.60%, 1.37%, and 7.28%, respectively, indicating their combination to be the best spectral pre-processing method. The optimal model for predicting α-helix and β-sheet contents was AOA-RF (characteristic wavebands), with calibration determination coefficients of 0.935 0 and 0.926 6 and prediction determination coefficients of 0.856 8 and 0.870 1, respectively. The optimal model for predicting β-turn and random coil contents was PLS (characteristic wavebands), with calibration determination coefficients of 0.915 4 and 0.881 7 and prediction determination coefficients of 0.891 3 and 0.784 3, respectively. The results of this study provide a theoretical basis for product quality detection and processing condition control in industrial settings.

The existing methods for the determination of phosphorus content are unable to regulate the addition of acid and base in the refining process of crude soybean oil through real-time monitoring. Therefore, a novel rapid method for determining the phosphorus content of crude soybean oil based on near-infrared spectroscopy was proposed in this study. It was found that standard normal variate transformation was more effective than two other spectral preprocessing methods evaluated for denoising the spectral data indicative of the phosphorus content in soybean crude oil. The characteristic absorption band of phosphorus was optimized by synergy interval partial least squares (SiPLS). A back propagation (BP) neural network prediction model of the phosphorus content in crude soybean oil was established with learning efficiency of 0.005 and 108 training cycles. The determination coefficient (R²), root mean square error (RMSE) and relative standard deviation (RSD) for the correction set were 0.9797, 0.8593 and 1.89%, respectively. The R², RMSE and RSD for the validation set were 0.9785, 0.9638 and 2.15%, respectively. The above results showed that NIR spectroscopy can achieve rapid, accurate and non-destructive detection of the phosphorus content in, and provide a feasible method for the refining of crude soybean oil.