No abstract

The abundances, molecular weights, molar absorptivities, and polydispersities of sediment pore water dissolved organic matter (DOM) and ferrous iron, Fe(II), were measured from three sites within a freshwater wetland (Old Woman Creek) adjacent to Lake Erie. DOM concentrations (expressed as organic carbon) increased with depth. Number-average molecular weight, weight-average molecular weight, and molar absorptivities (at 280 nm) also increased with depth up to 5 cm and remained relatively constant below this depth. These properties are indicative of the respective changes in the size of the DOM constituents and the number of aromatic moieties present. Moreover, a strong correlation was observed between Fe(II) and DOM, which suggests that both constituents coaccumulate in these pore fluids, It is hypothesized that reductionof iron oxides coated with organic matter releases both DOM and Fe(II). Increases in the molecular weights and molar absorptivities of the DOM constituents suggest that the "sorbed" component released by iron oxide dissolution is composed of larger and more aromatic molecules. The molecular-weight data also suggest that much of the refractory organic carbon in the interstitial waters comprises smaller-than-expected polymeric molecules. Thus, the release of the sorbed organic materials during the reductive dissolution of the oxide phases in anoxic sediments coupled with the selective preservation of refractory organic components may play an important role in both the DOM-mediated fate of pollutants and the cycling of carbon in wetlands.Natural dissolved organic matter (DOM) is ubiquitous in wetland surface waters and sedimentary pore fluids. These poorly characterized materials are capable of participating in a number of chemical reactions that are geochemically important. These include (but are not limited to) the complexation of metals (Cabaniss and Shuman 1987; Bartschat et al.

Sorption experiments were conducted with naphthalene, phenanthrene, and pyrene on low organic carbon sediments at 4 and 26 °C using batch and column techniques. Experimental controls ensured the absence of biologic and photolytic activity and colloidfree solution supernatants. Equilibrium distribution coefficients (K d ) increased 1.1-1.6 times with a decrease in temperature of 22 °C. Fraction instantaneous sorption (F) values did not change significantly with a decrease in temperature of 22 °C. Desorption rate constants (k 2 ) decreased 1.2-2.6 times with a decrease in temperature of 22 °C. Times to equilibrium were at least 40 h. The magnitude of observed K d and k 2 values and the effect of temperature on K d (e.g., low enthalpy of sorption) are consistent with sorbate partitioning between the aqueous phase and small amounts of organic matter (f oc ) 0.02%) on the sediments. The temperature dependence of K d and k 2 may be small as compared to the effects of heterogeneities in field-scale aquifer systems. Thus, thermal gradients may not be of major importance in most saturated subsurface regimes when predicting solute transport. However, aquifer remediation pump-and-treat times could be decreased because increased temperature decreases both retardation and tailing.

It has recently been recognized that mobile colloids may affect the transport of contaminants in ground water. To determine the significance of this process, knowledge of both the total mobile load (dissolved + colloid‐associated) and the dissolved concentration of a ground‐water contaminant must be obtained. Additional information regarding mobile colloid characteristics and concentrations are required to predict accurately the fate and effects of contaminants at sites where significant quantities of colloids are found. To obtain this information, a sampling scheme has been designed and refined to collect mobile colloids while avoiding the inclusion of normally immobile subsurface and well‐derived solids. The effectiveness of this sampling protocol was evaluated at a number of contaminated and pristine sites. The sampling results indicated that slow, prolonged pumping of ground water is much more effective at obtaining ground‐water samples that represent in situ colloid populations than bailing. Bailed samples from a coal tar‐contaminated site contained 10–100 times greater colloid concentrations and up to 750 times greater polycyclic aromatic hydrocarbon concentrations as were detected in slowly pumped samples. The sampling results also indicated that ground‐water colloid concentrations should be monitored in the field to determine the adequacy of purging if colloid and colloid‐associated contaminants are of interest. To avoid changes in the natural ground‐water colloid population through precipitation or coagulation, in situ ground‐water chemistry conditions must be preserved during sampling and storage. Samples collected for determination of the total mobile load of colloids and low‐solubility contaminants must not be filtered because some mobile colloids are removed by this process. Finally, suggestions that mobile colloids are present in ground water at any particular site should be corroborated with auxiliary data, such as colloid levels in “background” wells, colloid‐size distributions, ground‐water geochemistry, and colloid surface characteristics.