Novos métodos para autenticação e determinação de umidade e dos teores totais de proteínas e fósforo em ovos de galinha em pó usando imagens digitais, espectrometria NIR e quimiometria

Abstract

Powdered chicken eggs are used in various food formulations in the bakery and confectionery industries, mainly due to their long shelf life, ease of storage and transportation, and their gelling, foaming and emulsifying properties. Assessing their quality is important to ensure food safety for consumers. Therefore, this work aimed to develop fast, labor-free and accessible methods to assess the quality of powdered chicken eggs using near-infrared (NIR) spectra, digital images, data fusion and calibration and authentication techniques. Digital images were captured using a scanner, while NIR spectra were recorded using benchtop and portable spectrometers. In the first study, new strategies were developed to determine moisture and total protein and phosphorus contents in powdered chicken eggs. Multivariate calibration methods to predict total protein content, in terms of correlation coefficient (r) and ratio performance deviation (RPD), were better using digital images (r = 0.995 and RPD = 9.55) than those using NIR spectra from bench and portable equipment. On the other hand, in predicting humidity (r = 0.974 and RPD = 4.35) and total phosphorus content (r = 0.98 and RPD = 4.3), the benchtop NIR achieved better analytical performance. When applying data fusion techniques, the results obtained were slightly inferior to the best methods developed using pre-processed NIR spectra or digital images separately. In the second study, new strategies were devised to authenticate powdered chicken eggs for adulteration with soy protein and corn starch. To this end, classification techniques were applied using data-driven Independent and Flexible Modeling of Class Analogy (DD-SIMCA). The best method was the one using portable NIR spectra when compared with digital images, in which all adulterated samples were classified correctly and only one pure sample was classified as adulterated in the test set, achieving sensitivity of 95.7%, specificity of 100%, and accuracy 99.4%. Therefore, the methods proposed in this thesis work proved to be a good alternative tool for destructive chemical analysis since samples of powdered chicken eggs could be analyzed without using chemicals and without generating waste harmful to health and the environment, following the basic principles of Green Chemistry.