MULTIVARIATE CORRELATIONS OF GEOCHEMICAL BINDING PHASES OF HEAVY METALS IN CONTAMINATED RIVER SEDIMENT

Abstract

Distributions of geochemical binding phases of seven heavy metals (Cr, Cu, Co, Zn, Ni, Pb, and Cd) in sediment cores taken from six heavily polluted sites of the Ell-Ren River in Southern Taiwan were studied. Sequential extraction procedures (SEP) were used to determine the variations of heavy metal binding phases (exchangeable, bound to carbonates, bound to manganese-oxides, bound to iron-oxides, and bound to organic matter) in different sediment depths. Multivariate analyses were used to explore the correlations among these geochemical binding phases of heavy metals.

Results showed that the total amounts of various binding phases of heavy metals significantly varied with sediment depth, but their binding behaviors in various phases did not significantly change with depth. The organic matter content in the sediments increased with increasing Fe-oxide content. In addition, the binding affinities of carbonates with Zn, Pb, and Ni were higher than the affinities of carbonates with the other heavy metals. The binding affinity of Fe-oxides with Cr was higher than the affinities of Fe-oxides with the other heavy metals. Both correlation matrixes and principal component analyses demonstrated that distributions of Cu, Zn, Ni, and Cd had significant correlations with each other in both different depth horizons and various geochemical binding phases. The results indicate that these heavy metals might be discharged from the same pollution sources in the past, and also showed stable geochemical binding behaviors with the high silt sediment. However, Co had a poor correlation with the other six heavy metals in various binding phases, except with organic matter. Binding behaviors of Pb in the phases of bound to carbonates and exchangeable were different from the other six heavy metals. Cu was inversly correlated with the other six heavy metals in its binding behavior with reducible phases (Fe-/Mn-oxides).

Keywords:

• Correlation
• Heavy metal
• Multivariate analysis
• Sediment
• Speciation

ACKNOWLEDGMENT

The authors are grateful to National Science Council for financial support (NSC 86-2211-E-041-005).