Avaliação da qualidade e variedade de leite em pó e leite condensado por espectroscopia Raman e análise multivariada

Loading...
Thumbnail Image

Journal Title

Journal ISSN

Volume Title

Publisher

Universidade Federal de Juiz de Fora (UFJF)

DOI

Abstract

In this work, the FT-Raman spectroscopy was explored as a rapid technique for assessment the quality of milk powder and condensed milk. The exploratory analysis employing data from Raman spectroscopy and principal component analysis (PCA) allowed the separation of samples of milk powder according to its classification. Multivariate analysis was also used to classify adulterated milk powder samples. In the model of principal components, the adulterated samples with up to 5% of whey and / or starch were very close to those of samples without addition of adulterant. Discriminant analysis by partial least squares, the commercial samples were classified correctly, since the samples with 1 and 2% w/w whey samples were classified as unadulterated. In regression models using PLS, the results were satisfactory for contaminant concentrations above 5% w/w, below this concentration and for samples without addition of adulterant the model showed high relative errors of up to 50%. The changes that occur in milk powder during storage were followed by Raman spectroscopy; the changes observed in the spectra indicate the crystallization of lactose as the main alteration of stored milk. In evaluating the quality of condensed milk of different Brazilian brands, the Raman spectra have provided information of the main constituents. The principal components analysis revealed profile of samples with different Raman spectrum from the others. In the analysis of the loadings of the first principal component, the characteristic bands of sucrose showed great contribution. These results suggest different sucrose concentrations for the samples of condensed milk. The work shows the Raman spectroscopy as an analytical tool that can be used in quality control of milk powder and condensed milk. Although the multivariate models developed are not strictly quantitative, especially for low concentrations, they can be used as screening methods for routine analysis, as demonstrated by this work.

Description

Citation

Endorsement

Review

Supplemented By

Referenced By

Rights and licensing

Acesso Aberto