Multivariate Statistical Techniques to Determine Essential and Toxic Elements in Biological Samples by X-Ray Fluorescence

Abstract

X-ray fluorescence spectrometry was used to determine essential and toxic elements in tea leaves. Quantitative measurements were obtained for K, Mg, Cu, Zn, As, Sr, Fe, Co, Ca, P, S, Pb, Rb, and Mn. Pearson's product moment correlation coefficient with principal component analysis and hierarchical clustering analysis were performed to investigating the relationships between essential and nonessential elements in the dry herbal mass. From multivariate statistical analysis, Ca, Fe, P, Zn, and Mn were found to have strong positive correlations. While principal component analysis demonstrates that tea leaves were dominated by Fe and Cl in the first principal component, Pb and As were in the second principal component with cumulative variability of 84.4%. However, hierarchical clustering analysis indicated the presence of a cluster of four essential elements, one cluster containing toxic and essential elements, and one cluster composed of toxic elements. The results of this study improve understanding of the correlation between essential and toxic elements in biological samples such as tea leaves.

Keywords:

• hierarchical clustering analysis
• Pearson correlation
• principal component analysis
• X-ray fluorescence spectrometry
• XRF

ACKNOWLEDGMENTS

The authors acknowledge Dr. Mouhamadou Diallo Malaria Section, Molecular Biology Unit, Bacteriology and Virology Laboratory at Aristide Le Dantec Hospital for providing technical support and gratefully acknowledge Dr. Lutgen Pierre Luxembourg IFBV NGO for providing tea samples.