Desorption from Contaminated Sediment and the Organic-Carbon Normalized Sediment-Water Partition Coefficient, Koc, for Dioxin

Abstract

We use the “solids concentration effect” in an attempt to measure the organic-carbon normalized sediment-water partition coefficient, Koc, for 2,3,7,8-tetrachlorodibenzo-p-dioxin (dioxin) using desorption from contaminated sediment from Lake Ontario. The sediment was collected from Station 208 (Onuska et al., J. Great Lakes Res. 9(2), 169-182, 1983); it contains 470 pg/g dry weight of dioxin and 2.37% of organic carbon. We prepared a series of sediment suspensions of decreasing solids concentration; after various periods of equilibration followed by separation of sediment solids and water by sedimentation, we analyzed the aqueous phase for dioxin and organic carbon. The ranges of solids concentration (Cs,w ), of total aqueous phase dioxin (Clc,w ), and of aqueous phase organic carbon (Coc,w) are 0.06-60 g/L, 2-1000 pg/L, and 0.23-65 mg/L respectively. For the measured partition coefficient, Kmoc , our data fit the functions, log Kmoc = a constant - log Cs,w, and log Kmoc = a constant - log Csorb,w; values of R2 range from 0.61 to 0.92. Csorb,w is related to the organic carbon content of the water by Csorb,w = Coc,w - 0.2 mg / L; the prime indicates that the aqueous organic carbon levels are corrected for blank contributions. The total aqueous phase dioxin concentration only depends strongly on the concentration of organic carbon when Coc,w > 0.2 mg / L. We think this is an example of the phenomenon described by the term “critical micelle concentration”, CMC, which is usually reserved for the description of surfactant solutions, but here we have evidence that it applies to natural organic carbon also. For the system here the CMC is about 0.2 mg/L. By spiking the apparatus with C13-labeled dioxin, we are able to judge the degree to which the system has reached equilibrium or steady state. We establish clearly thereby that the dioxin in the water is at steady state in four experiments. Another feature is the recovery of the settled sediment solids at the end of experiments, and we show that the dioxin concentration, on a dry weight basis, decreases; levels are in the range 230-380 pg/g. This begs the question: What is the organic-carbon normalized dioxin concentration in the sediment at the end of the experiment? With assumptions, mass balance shows that significant quantities of organic carbon are adsorbed on the walls of the apparatus, and it appears that equilibrium is reached when concentrations of dioxin on organic material in the sediment, in the water and on the walls of the apparatus are the same. When doing partitioning experiments at low solids in large apparatus, it is important to recover the settled solids, and determine the level of adsorbed chemical and the organic carbon content. Then, we can complete the mass balance and determine uniquely the quantities of material adsorbed to the walls of the apparatus. Work with less hydrophobic compounds than dioxin shows that Kmoc becomes independent of solids concentration as the solids concentration decreases, whence Koc = Kmoc. We did not reach a sufficiently small concentration of solids; however, we are able to conclude that for dioxin Koc > 7.1; this observation agrees with other partitioning work done with dioxin and its isomers.