²ÊÃñ±¦µä

Poulin, Brett A ¹ ; Aiken, George R ² ; Ryan, Joseph N ³ ; Manceau, Alain ⁴ ;Nagy, Kathryn L ⁵

¹ ²ÊÃñ±¦µä
² U.S. Geological Survey, Boulder Colorado
³ ²ÊÃñ±¦µä
⁴ ISTerre, Université Joseph Fourier
⁵ University of Illinois at Chicago

Transient redox conditions are known to strongly influence the behavior of mercury in natural environments, but the importance of soil redox status in controlling mercury fate and transport in contaminated riparian soils is not well understood. Factors such as soil mercury speciation, sulfur chemistry, organic matter quality and soil geochemistry likely control mercury transformation and mobilization within contaminated environments. Here we describe results of a study designed to identify the dominant source(s) and biogeochemical controls of mercury release from contaminated floodplain soils in Oak Ridge, TN. Mercury release behavior in two soil horizons from the East Fork Poplar Creek stream bank under saturation conditions was assessed in (1) laboratory microcosm experiments with air-dried O (0-10 cm depth) and A horizon (20-30 cm depth) soils and (2) laboratory flooding experiments with intact soil cores (0-30 cm depth) that encompass O and A horizons.

Soil characterization included total mercury concentration, mercury speciation by selective sequential extraction and X-ray absorption spectroscopy, organic matter content, elemental composition, and soil mineralogy. Individual dark microcosm experiments were performed under a nitrogen atmosphere in duplicate at a 1:1 solid-to-liquid ratio using deionized water. Porewater samples were collected at 11 time intervals over 36 days. Unfiltered porewater analyses included pH, EH, total mercury, methylmercury, dissolved gaseous mercury, and major metals (manganese, iron, aluminum). Filtered (0.20 µm) porewater analyses included anions, cations, major metals, total mercury, methylmercury, dissolved inorganic carbon, dissolved organic carbon, and dissolved organic matter (DOM) optical properties. A rapid release (1-3 days) of particulate and filter-passing total mercury was observed in O horizon porewaters following inundation, followed by a decrease between 3-11 days. The initial mercury release is likely due to the coincident release of DOM of higher molecular weight. The re-association of mercury with O horizon soils was attributed to direct (e.g., metacinnabar precipitation) and/or indirect scavenging (e.g., mackinawite precipitation and mercury sorption) of mercury by sulfide generated from microbial sulfate reduction. Conversely, slow mercury release was observed in A horizon soils for both particulate and filter-passing fractions, which suggests a release mechanism coupled to redox-sensitive processes. This hypothesis was supported by a positive correlation (R2 = 0.79, n = 22) between porewater manganese concentrations, attributed to reductive dissolution processes, and filter-passing total mercury. Flooding experiments with intact soil cores showed similar mercury release behavior, which indicates that the same processes control mercury release and transformation in intact, undisturbed soil matrices.