Competition between aqueous and solid phases

Geochemical speciation codes (like 'MINTEQ') are now frequently used to show which solid phases control the concentration of a metal, e. g. in soil/sediment porewater systems (Yanful et al. 1999). Coupling aqueous speciation with surface complexation models represents the state-of-the-art in metal adsorption modelling today (see also section 5.4.7 and 5.5.2, below). But still only few sorption studies treat the adsorption of a metal on natural particles as a process of partitioning among a number of different discrete phases.

Radovanovic and Koelmans (1998) developed a conceptual kd model called 'SWAMP' (Sediment Water Algorithm for Metal Partitioning) as a function of aqueous and solid phase characteristics. The model uses the concept of additivity of discrete site binding of free and complexed metals to solid phases, accounts for electrostatic effects and competitive adsorption and includes also aqueous phase complexation (see section before). Sorption site concentrations between 10-6 M (lakes) and 10-4 M (rivers) are usually reported in the relevant literature (references given therein). Adsorption sites considered in the model are Fe oxyhydroxides (OH-sites), Mn oxyhydroxides (OH-sites), and organic matter sites (type A COOH-sites and type B OH-sites), as well as free metal ions and hydroxy complexes (MOH+ and M(OH)2) as adsorbates, and Ca2+ and Ca hydroxo complexes as main competing species. kd values obtained from suspended solids and aqueous phase measurements followed a rather constant trend at each sampling site (lake and river waters): Pb > Cd > Zn > Cu, Ni. Comparing this order with the sequence of metal hydrolysis constants (Pb > Cu > Zn > Cd > Ni) one would say that Cu seems less reactive towards particles than expected from the hydrolysis constants (opposite to Cd). However, considering the known strong affinity of Cu to dissolved organic ligands (DOM), the observed trend seems reasonable. Best fits for kd by the 'SWAMP' model were achieved for Cu and Zn. According to the modelling data, sorption of second-order hydrolysis metal species to Fe oxyhydroxide, followed by free metal sorption to Mn oxyhydroxides, in competition with Ca2+, is supposed to be a major significant process of solid phase formation occuring in these waters, consistent with the fact that mono- and divalent cations (Cu, Zn, Ni) preferably associate with Mn oxides (including Mn oxyhydroxides). The relative contribution of the selected solid binding phases for Zn were 69.6 % (Fe), 27.1 % (Mn), and 3.3 % (Corg), and for Cu 99.2 %, 0.7 % and 0.1 %, respectively. From the obtained data, the authors conclude that multiple solid

* kd = distribution coefficient phases appear to be relevant in modelling metal uptake processes. In contrast, complexes with DOM seem to be most important for Zn and Cu in aqueous phases. It was summarized that surface and solution complexation equilibria can explain the major part of the variation observed for kd in freshwaters.

Also Lofts and Tipping (2000) used a similar predictive chemical speciation code (WHAM-SCAMP) and found a good agreement between observed and model-predicted logkd-values for Zn. For Ni (and Co) the predicted kd values were higher than the observed ones, whereas the opposite was the case for Cu (and Pb). The authors resume that the differing results may depend on (1) the extent of Ca and Mg competition for available binding sites, on (2) the chemical nature of the measured particulate metal, (3) on the efficiency of the solid-solution partitioning method, and on (4) the strength of the Cu-organic matter binding. But they conclude that the used model shows a good promise to predict the solid-solution partitioning of metals in aquatic systems.

In an estuarine water, Wen et al. (1999) found a significant correlation between the colloidal metal fraction and the colloidal organic carbon fraction in the water phase suggesting colloid-bound metals as a result of metal-organic complex formation. In particular, iron was bound to high-molecular weight colloids, whereas Cu, Ni (and Pb) were more associated to colloidal low molecular weight compounds. In addition, partitioning coefficients for metals distributed between colloids and true solution (Kc) were higher than between particles (Kp) and true solution (exept for Fe), confirming a strong binding intensity of these metals to macromolecular colloidal organic matter.

Lu and Allen (2001) used batch adsorption experiments with suspended particles and river water to investigate the effect of many factors that influence the partitioning coefficient (kd) for Cu, including pH, total suspended solids (TSS), total copper concentration, dissolved organic matter (DOM), particulate organic matter (POM), hardness and ionic strength. They found organic matter binding sites both as dissolved and solid phases as most important in controlling the partitioning of Cu, beside other major factors such as pH and TSS. Measured partitioning coefficients proved to be independent of total Cu concentrations as long as TSS was high, but increased at low TSS, as expected.

Previous studies on the speciaton of dissolved nickel (Ni) in surface, waste and runoff waters by competitive ligand exchange (e. g. CSV coupled to a chelating resin) indicate complexation of this metal by an extremely strong ligand. Competitive ligand exchange techniques are increasingly used to estimate concentration and stability constants of dissolved or particulate organic carbon metal binding ligands. Significant concentrations of strongly complexed Ni have been observed in many estuarine systems. Most of these estuaries receive industrial wastewaters containing Ni ethylenediaminetetraacetate (NiEDTA2-), which is supposed to be that strongly complexed Ni species. To further identify this ligand, its source and fate, Bedsworth and Sedlak (1999) performed simultaneous measurements of NiEDTA2- by high-performance liquid chromatography (HPLC) and cathodic stripping voltametry in combination with a chelating resin column in water samples from the San Francisco Bay. Wastewater analysis indicated that NiEDTA2- accounts indeed for the strongly complexed Ni in waste effluents. Equilibrium speciation calculations suggested, what we already know from other study areas, that even other metals, like Cu(II), Zn(II), and Pb(II), are similarly discharged as EDTA complexes from wastewater treatment systems (compare for more details previous AF-MFG technical reports on this particular issue). Also seasonal variations have been observed in Ni speciation due to the discharge of stable NiEDTA2-complexes from incoming wastewater, and of weaker complexes from surface runoff, leading to higher concentrations of strongly complexed Ni in summer, when surface runoff discharges decrease.

The authors also found that dissolved Ni (i. e. complexed Ni) in the San Francisco Bay decreased with increasing salinity from ~ 160 nM down to 5 nM. Also, the dissociation rate of NiEDTA2- seemed slow relative to its residence time and sorption unlikely under the prevailing conditions. It seems most likely that part of the metal-EDTA complex is converted to chloro-complexes when mixed up with seawater. From this, the authors speculate that a substantial part of complexed metals may pass wastewater treatment plants in an unreactive form, which however may dissociate into more bioavailable and potentially toxic metal-chloro forms after released and mixing with seawater. The authors resume that the presence of stable nonbioavailable metal complexes should be more considered in geochemical modelling studies, especially when developing site-specific water quality objectives (see also chapter 8).

0 0

Post a comment