| Summary: | 碩士 === 國立中山大學 === 海洋科學系研究所 === 103 === In this study, experiments were carried out using a two-column (Chelex-100 and AG MP-1 resin) ion exchange technique, operated in different sequences (Chelex-AG MP1, normal phase, or AG MP1-Chelex, reversed phase). The model solutions, containing various concentrations of metal complexing agents (humic acid, HA and/or ethylenediaminetetraacetic acid, EDTA), of different pH were prepared in Milli-Q water and seawater, and then passed through the column sets, in order to investigate the variations in trace metal (Cd, Co, Cu, Ni and Zn) partitioning. Chelex-100 is a cation exchange resin, and AG MP-1 is an anion exchange resin. The fractions obtained were operationally defined as Chelex-labile, AG MP-labile, non-labile and zwitterionic (retained by both Chelex-100 and AG MP-1 resins, obtained by calculation).
Experimental results showed that: (1) Cd, Co, Ni and Zn were mostly present as Chelex-labile fraction, while Cu was equivalently present as Chelex-labile and AG MP-labile fractions. When the column sequence was reversed, Cd was not retained by Chelex-100 resin, and some can be retained by AG MP-1 resin in seawater medium, suggesting the predominant fraction of Cd being non-labile fraction (probably CdCl2). Solutions with higher pH (8.0) showed enhanced AG MP-labile metal fractions, indicating organic complexation, than those at pH 5.5 in freshwater medium; (2) Organic complexes in model solutions were retained mostly by AG MP-1 resin, with a small part being retained by Chelex-100 resin. In addition, part of non-labile fraction exhibited organic complex characteristics; (3) AG MP-labile metal concentrations and the concentration of humic acid had a positive correlation, indicating the complexation with humic acid for different metals is in the following order: Cu &;gt; Ni &;gt; Co &;gt; Zn &;gt; Cd; (4) When model solutions contain only trace metals and EDTA, and the total metal concentration was lower than the complexation capacity of EDTA, Cd (89~92%), Co (77~83%), Cu (92~96%), Ni (88~93%) and Zn (80~83%) were predominantly present as AG MP-labile fraction (negatively charged M-EDTA complexes, MHEDTA-) in freshwater model solutions, and a minor were present as non-labile fraction (neutrally charged M-EDTA complexes, MH2EDTA). Fe was equivalently present as AG MP-labile and non-labile fractions in model solutions, comparing with other metals, and accounted for most of the non-labile fraction (FeHEDTA) among the elements determined. In seawater model solutions, Cd, Co, Cu, Ni and Zn were present mostly as Chelex-labile and non-labile fractions. Fe was equivalently present as Chelex-labile, AG MP-labile and non-labile fractions in model solutions. It was found that metal-organic complexes could present in different terms of metal speciation in model solutions (MHEDTA-, MH2EDTA); (5) Metal-EDTA complexes showed stronger complexation than those of metal-HA.
This analytical scheme was applied to estuarine and coastal waters. Results showed that, column yields had comparable trends of differences between operation modes (normal and reversed), and between samples in natural pH (~8.0) and buffered (5.5) when compared with lab experiments. Metal concentrations in the coastal and estuarine waters were in general lower than those in model solutions prepared in the lab, thus the distributions of metal species derived from the dual-column preconcentration technique represent results closer to natural waters.
Results obtained in this study suggest that column yields vary when the dual-column preconcentration technique is operated under different column sequences. Modification of sample/solution, especially pH, results in shifts of metal species. Therefore, the use of technique with multiple preconcentration resins requires more careful definition of operating conditions and result interpretation.
|
|---|
