MULTIVARIATE ANALYSIS OF THE MINERAL CONTENT OF GLUTEN-FREE SNACKS

HIDALGO, M.; FECHNER, DIANA C.; SGROPPO, SONIA CECILIA; MARCHEVSKY, E

San Luis

Congreso; SOCIEDAD DE BIOLOGIA DE CUYO; 2015

SOCIEDAD BIOLOGIA DE CUYO

In this work, the mineral concentrations of seventeen elements (Al, Ca, Co, Cr, Cu, Fe, K, Li, Mg, Mn, Mo, Na, Ni, P, Pb, Sr and Zn) in commercial gluten-free snacks were determined by using inductively coupled plasma optical emission spectrometry (ICP-OES). Microwave-assisted acid digestion of samples was used to eliminate the organic matrix of samples. The analytical method was validated by linearity, detection limits, precision, and recovery experiments, obtaining satisfactory values in all cases. The multielemental composition results were evaluated using multivariate analysis. Multivariate pattern recognition tools applied to data sets included principal component analysis (PCA) as a visualization method, and hierarchical cluster analysis (HCA) as an unsupervised learning method. In PCA the data matrix is decomposed into scores and loadings matrices. The scores vectors describe the relationship between the samples in the model subspace and the loadings vectors describe the importance of each descriptor within the model. It can represent graphically intersample and intervariable relationships and provides a way to reduce the dimensionality of the data. Similarly to PCA, clustering of samples reveals similarities among the samples while clustering of variables pinpoints intervariable relationships. As a result, PCA demonstrated that the elements that contributed most for the variability inter-samples were: Ca, K, Ni and Na. A trend was observed towards the classification of samples according to the recipe ingredients. In addition, the results obtained by HCA were in a good agreement with PCA results.