Multivariate classification of medicinal herbs from the Amazon Region and their infusions according to their mineral composition [Classifica??o multivariada de ervas medicinais da Regi?o Zmaz?nica e suas infus?es de acordo com sua composi??o mineral]

摘要

In this study, the mineral composition of leaves and teas of medicinal plants was evaluated. Ca, Cu, Fe, Mg, Mn e Zn were determined in the samples using flame atomic absorption spectrometry. Principal component analysis was applied to discriminate the samples studied. The samples were divided within the 2 groups according to their mineral composition. Copper and iron were the variables that contributed most to the separation of the samples followed by Ca, Mg, Mn and Zn. The information in the principal component analysis was confirmed by the dendrogram obtained by hierarchical cluster analysis.