An exploratory study for the technological classification of egg white powders based on infrared spectroscopy. (October 2018)

• Record Type: Journal Article
• Authors: Grassi, Silvia; Vitale, Raffaele; Alamprese, Cristina
• Is Part Of: Lebensmittel-Wissenschaft + Technologie =. Volume 96(2018)
• Page Start: 469
• Page End: 475
• Publication Date: 2018-10
• Subjects: Gelling properties -- Foaming properties -- Duplex algorithm -- PLS-DA -- Variable selection; EWP egg white powder -- FT Fourier transform -- IR infrared -- MIR mid infrared -- MSC multiplicative scatter correction -- NIR near infrared -- PLS partial least squares regression -- PLS-DA partial least squares-discriminant analysis -- VIP variable importance in projection; Food industry and trade -- Periodicals; Food -- Composition -- Periodicals; Microbiology -- Periodicals; Nutrition -- Periodicals; 664.005
• DOI: 10.1016/j.lwt.2018.05.065
• Languages: English
• ISSNs: 0023-6438

Abstract

Abstract: This work aims at the evaluation of FT-NIR and FT-IR spectroscopy as rapid, easy, and cost-effective tools for the classification of egg white powder (EWP) based on its technological properties. Up to 100 commercial spray-dried EWP samples with known gelling and foaming properties were used to acquire FT-NIR and FT-IR spectra. An appropriate data-splitting algorithm (Duplex) was applied in order to create, for each dataset, a calibration set and a representative validation test set for prediction. Different spectral pre-treatments and their combinations were investigated for the calculation of Partial Least Squares–Discriminant Analysis models in order to classify samples according to gel strength, foam height, and foam instability. A variable selection strategy based on the so-called Variable Importance in Projection scores was also evaluated. Both FT-NIR and FT-IR spectroscopy showed good potential in discriminating EWP samples with different technological properties. Correct classification percentages in prediction ranging from 59% to 89% were obtained with the best models calculated with selected wavenumbers. These results show a promising industrial perspective, demonstrating the possibility of developing cheap and fast instruments spanning a limited spectral range, which can be implemented on the production lines for EWP sorting and quality control. Highlights: IR spectroscopy correctly predicts albumen gel strength, foam height and stability. FT-IR predicts dried albumen gelling and foaming properties better than FT-NIR. Variable selection improves IR models' classification ability. IR spectroscopy implementation in industrial production lines is promising.