Correlation analyses on binding behavior of heavy metals with sediment matrices

This article presents the amounts of heavy metals bound to the sediment matrices (carbonates, Fe-oxides, Mn-oxides, and organic matter), the correlations between any two heavy-metal binding fractions, and the correlations between sediment matrices and their heavy-metal binding fractions. Data consisted of 313 sets obtained from five main rivers (located in southern Taiwan) were analyzed by statistical methods. Among six heavy metals analyzed (Zn, Cu, Pb, Ni, Cr, and Co), the statistical results show that Zn is primarily bound to organic matter, and Cr is primarily bound to Fe-oxides. Principal component analysis (PCA) and correlation analysis (CA) result in significant correlations between carbonates bound Ni and carbonates bound Cr, Fe-oxides bound Ni and Fe-oxides bound Cr, and Mn-oxides bound Cu and Mn-oxides bound Cr. From linear regression results, the levels of the six heavy metals bound to either organic matter or Fe-oxides is moderately dependent on the contents of organic matter or Fe-oxides, especially true for Cr and Pb. According to slope values of linear regression, Cu and Cr have the highest specific binding amounts (SBA) to organic matter and Fe-oxides, respectively. A significant correlation between organic matter and organically bound heavy metals implied that organic matter contained in the sediments of the Potzu river and the Yenshui river can be adequately used as a normalizing agent. However, the six heavy metals bound to either carbonates or Mn-oxides do not correlate with carbonates or Mn-oxides. The obtained results also imply that competitions of various sediment phases in association with heavy metals occur, and organic matter and Fe-oxides are more accessible to heavy metals than other sediment phases. (C) 2001 Elsevier Science Ltd. All rights reserved.