The binding of silver(I) by organic matter was investigated in batch equilibrations with Suwannee River Fulvic Acid (SRFA), mor (Oe) and peat materials. For the SRFA systems, binding isotherms were determined at pH 4 and 8 using potentiometric titrations, with 0.05 M NaNO3 as background electrolyte. The binding showed a strong pH dependency and was about one order of magnitude stronger at pH 8 compared to pH 4. The data set was consistent with published data sets for isolated fulvic and humic acids. The binding of silver(I) by mor and peat materials was studied as a function of pH in the range 2.5-5.0 in dilute NaNO3 solutions (ca. 0.01 M), at a wide range of silver-to-soil ratios (10-2 – 10-4.5 mol kg-1). The silver(I) binding properties of the two materials were similar, with increasing binding strength with pH. The slope of sorption isotherms determined at pH 2.5 and 4.0 was significantly less than one, indicating heterogeneous binding sites. Ion competition experiments with added 1 mM iron(III) or 1 mM aluminium(III) at pH 2.5 and 4.0 showed no interaction between these ions and the silver(I) ion, indicating highly specific silver(I) binding sites.
Silver K-edge XANES in the region 25280-26080 eV was used to characterize the binding mode of silver(I) in the mor and peat materials. A silver-to-soil ratio >10-2.5 mol kg-1 was needed to obtain a significant absorption edge. Comparison of spectra with oxygen, sulfur and nitrogen containing model compounds, indicated that the silver(I) ions, on average, was bound to one oxygen and one nitrogen donor ligand in an approximately linear fashion in both samples. Speciation of sulfur in the mor and peat materials using sulfur K-edge XANES, suggested that thiol groups could contribute to maximum about 20% of the silver binding at the conditions used in the Ag K-edge XANES measurements.
The Stockholm Humic Model (SHM) was used as a tool to make a unified interpretation of SRFA and soil data. First, the model was calibrated using SFRA data together with published data on silver binding to isolated fulvic and humic substances, in order to obtain a consistent generic silver binding parameter set. Second, the calibrated SHM was used to predict the experimental silver(I) binding data obtained with the mor and peat materials. Taking this approach the model largely underestimated the silver(I) binding by the mor and peat materials; roughly by a factor of 10-100, dependent on experimental conditions. Thus, our data suggest that silver(I) binding properties of isolated fulvic and humic acids are very different from those of intact soil and peat materials. This will have great implications for how to calibrate and apply geochemical models describing the behavior of silver(I) in soils and natural waters.
Athens, Georgia, 2013.
12th International Conference on the Biogeochemistry of Trace Elements (ICOBTE), June 16-20, Athens, USA