Organic matters, but inorganic matters too: column examination of elevated mercury sorption on low organic matter aquifer material using concentrations and stable isotope ratios

Abstract. Sorption of mercury (Hg) in soils is suggested to be predominantly associated with organic matter (OM). However, there is a growing collection of research that suggests that clay minerals and Fe/Mn oxides are also important solid phases for the sorption of soluble Hg in soil–groundwater systems. We use a series of (60 mL syringe based) column experiments to examine sorption and subsequent desorption of HgCl2 solutions (experiment 1 (EXP1): 46.1 ± 1.1 mg L−1; experiment 2 (EXP2): 144 ± 6 mg L−1) in low-OM (0.16 ± 0.02 %) solid-phase aquifer materials. Analyses of total Hg concentrations, Hg speciation (i.e. pyrolytic thermal desorption (PTD)), and Hg stable isotopes are performed on both solid- and liquid-phase samples across sorption and desorption phases of the experiments. The sorption breakthrough curve best fitted a Freundlich model. Despite the very low-OM content, the Hg equilibrium sorptive capacity in these columns is very high: 1510 ± 100 and 2320 ± 60 mg kg−1 for EXP1 and EXP2, respectively, and it is similar to those determined for high-OM soils. Data from the experiments on mass-dependent Hg stable isotope fractionation data from these experiments (described by δ202Hg) support preferential sorption of lighter isotopes to the solid-phase materials with results indicating an isotopically heavier liquid phase and an isotopically lighter solid phase. Desorption fits exponential decay models, and 46 ± 6 % and 58 ± 10 % of the sorbed Hg is removed from the solid-phase materials at the termination of desorption in EXP1 and EXP2, respectively. The divergence of δ202Hg values between liquid and solid phases also continues into desorption. This desorption profile is linked to the initial release of easily exchangeable Hg(II) species physically sorbed to Fe/Mn oxides and clay mineral surfaces (liquid phase enriched in heavy isotopes) and then slower release of Hg(II) species that have undergone secondary reaction to more stable/less-soluble Hg(II) species and/or diffusion/transport into the mineral matrices (processes favouring lighter isotopes; solid phase enriched in lighter isotopes). The secondary production of Hg(0) within the columns is confirmed by PTD analyses that indicate distinct Hg(0) release peaks in solid-phase samples at <175 ∘C, which again agrees with field observations. Retardation (RD) and distribution (KD) coefficients are 77.9 ± 5.5 and 26.1 ± 3.0 mL g−1 in EXP1, respectively, and 38.4 ± 2.7 and 12.4 ± 0.6 mL g−1 in EXP2, respectively. These values are similar to values derived from column experiments on high-OM soil and provide the basis for future Hg fate and transport modelling in soil–groundwater systems.