Aplicação de imagens digitais e técnicas espectroanalíticas combinadas com quimiometria para detecção e quantificação de adulteração em leite bovino

Abstract

Milk provides essential nutrients (carbohydrate, fat, protein, minerals and vitamins) of great nutritional relevance for humans, particularly during childhood. Nowadays, the occurrence of milk adulteration is a major issue in the dairy industry, and has been causing concerns among costumers and food manufacturers. In this context, the objective of this study was developed a simple, fast and non-expensive method to identify and quantify milk adulteration. In this study, three different methods were proposed based on digital image, infrared spectroscopy (IR) and time domain nuclear magnetic resonance (TD-NMR). Milk samples were adulterated by addition of tap water, whey, synthetic milk, synthetic urine, urea, hydrogen peroxide and caustic soda, in different concentrations, except for caustic soda. In this case, fresh milk samples were put at ambient temperature and waited to turn sour. Later, a 10- mol/L commercial NaOH solution was added into the sourish milk to establish its original pH. Classification models obtained with digital images exhibited tight and well-separated clusters allowing the discrimination of control from adulterated milk samples. Results obtained with TD-NMR and IR allowed the discrimination of milk samples according to the level of adulteration and the type of adulterant used in the adulteration process. In the case of adulteration with NaOH, the method developed with digital image was able to discriminate fresh from adulterated milk samples. Regression models obtained with digital image showed that the method was able to detect adulteration level ≥ 6% v/v. In the other hand, the models obtained with IR and TD-NMR allowed the quantification of adulteration at levels ≤ 3% v/v.