Quantitative determination of fat and total protein in milk based on visible light scatter

A new optical spectroscopic method for milk fat and total protein analysis has been developed. In contrast to the conventional approach that generally relies on the components' absorption, the suggested method is based on the phenomenon of light scatter by fat and protein particles. This fundamental distinction enables shifting the measurement to the cost-effective visible and adjacent near infrared region (below 1000 nm), where the scatter strongly dominates. Partial Least-Squares regression modelling on a designed set of training and validation milk samples resulted in root mean-square prediction errors of 0.05% and 0.03% for fat and protein content, respectively, which is close to the accuracy of reference analysis. It has been shown that multivariate data analysis is capable of distinguishing individual scatter spectra of fat and protein. This conclusion has been supported by Mie-theory calculations. The method is suitable for routine laboratory analysis or in-line quality monitoring in the dairy production. (C) 2012 Elsevier Ltd. All rights reserved.