G7 (IV) Environmental Biogeochemistry: Lab and Experimental Approaches

Mercury and its related compounds are widely recognized as global pollutants. Atmospheric gas, aqueous and heterogeneous chemistry are expected to occur for Hg-containing species and accurate implementation of their chemical parameters is essential for realistic modelling of mercury cycling. Although significant progress has been made, the current state of knowledge of mercury chemistry is not precise and exhibits numerous uncertainties. At McGill University, our research theme is involved in addressing a few of these uncertainties including (a) development of techniques for chemical characterization of so-called “reactive gaseous mercury”, (b) perform kinetic and product studies in the homogeneous and heterogeneous phases, including at selected environmentally relevant surfaces. In this paper, we will present some of recent data supported by the Environment Canada program (CARA), and discuss our results within the context of environmental mercury cycling while pointing to the advancement of knowledge and questions to be solved in future.

Mercury (Hg) in the environment is an important pollutant of concern worldwide. The reduction of mercury from Hg²⁺ to Hg⁰ is especially important. One pathway for this reduction to occur is through an abiotic process with humic acids (HA) in anoxic condition, which is controlled by different factors, including concentration ratio of Hg²⁺/DOC, pH, temperature and light, all of which were investigated in this study. From the scale of laboratory investigation, comparison of the reduction capacity of HA involving the initial addition status (aqueous or solid bulk) and different HA sources were also examined. Results indicated that HA were able to reduce mercury abiotically, and the ratio of Hg²⁺/HA may not be the only a parameter to decide the reduction process, but the actual absolute concentrations of Hg²⁺ and HA added were the same important. Evidently, two opposing effects including the binding affinity effect (the reduction increasing with increasing Hg²⁺/DOC ratio) and the effect of decreasing reduction with decreasing absolute concentration of HA indicating decreasing absolute redox active groups in reaction system were involved in it, which also demonstrate the existence of competition between reduction and complexation process. Thus, both of Hg²⁺/DOC ratio and absolute concentrations of HA and Hg²⁺ also need to be considered, instead of determining Hg²⁺ reduction by simply utilization of ratio relationship. Meanwhile, the results also indicated that low (3.6) or high (8.1) solution pH values decreased the HA reduction capacity. Hg⁰ production rate increased with increasing temperature, and the same trend was observed with light exposure. In addition, HA added as an aqueous solution resulted in significantly greater Hg⁰ production than addition as a bulk solid. Finally, the mercury reduction rate and capacity varied significantly (P<0.05) with HA from different sources due to the different characteristics of the HA. Although the Hg²⁺/DOC sratios used in this study were higher than levels in normal natural condition, they also demonstrated directly the existence of a possible pathway of Hg²⁺ reduction, which indicated that HA in natural environments, especially in polluted water bodies, needs to be considered not only as strong sink for Hg, but also as a source, when concentrations of Hg²⁺ and HA were in a certain range with particular conditions.

Magnetite and green rust have been shown to reduce aqueous Hg^II to Hg⁰. In this study, we have tested the ability of magnetite and green rust to reduce Hg^II sorbed to 2g/L biomass (Bacillus subtilis) at high (50 µM) and low (5 µM) Hg loadings, and at pH 6.5 and 5.0. At high Hg:biomass loading, where Hg^II binding to biomass is predominantly through carboxyl functional groups, Hg L_III edge XAS showed reduction of Hg^II to Hg⁰ by magnetite. Reduction occurred within 2 h and 2 d at pH 6.5 and 5.0 respectively. At low Hg:biomass loading, where Hg^II binds to biomass via sulfhydryl functional groups, Hg^II was not reduced by magnetite at pH 6.5 or 5.0 after 2 months of reaction. Green rust, which is generally a stronger reductant than magnetite, reduced about 20% of the total Hg^II, bound to biomass via sulfhydryl functional groups, to Hg⁰ in two days. These results suggest that Hg^II binding to carboxyl functional groups does not significantly inhibit the reduction of Hg^II by magnetite. However, the binding of Hg^II to biomass via sulfhydryl functional groups severely inhibit the ability of mixed Fe^II/III phases like magnetite and green rust to reduce Hg^II to Hg⁰. The mobility of heavy metal contaminants in aquatic and terrestrial environments is greatly influenced by their speciation, especially with respect to oxidation state. In the case of Hg, reduction of Hg^II to Hg⁰ can increase Hg mobility due to the volatility of Hg⁰. Since Hg is typically present in aquatic and terrestrial systems at low concentrations, binding of Hg^II to high affinity sites on bacteria could have important implications for the potential reduction of Hg^II to Hg⁰ and overall mobility of Hg in biostimulated subsurface environments.

The implementation of stable isotope tracer methods has given invaluable knowledge about Hg biogeochemical processes, including formation of CH₃Hg. It has however been acknowledged that the availability for reactions of tracers, normally added as labile dissolved complexes, likely is higher than for ambient Hg, yielding unrealistically high reaction rates.

Here we propose a novel, fundamentally different, approach utilizing solid phase isotope tracers with well-defined chemical structures mimicking as closely as possible the structure of ambient Hg in sediments. The aims are to obtain more representative tracers, and thereby be able to determine reaction rates closer to the incipient, and also be able to determine and compare methylation rates for Hg structures of different geochemical “age”.QUANTIFICATION AND SPECIATION OF SUBEstuarine sediments were incubated with isotope tracers of four different Hg^II solid phases; cinnabar: a-¹⁹⁹HgS(s), metacinnabar: ß-²⁰¹HgS(s), Hg^II bound to Mackinawite: °FeS-²⁰⁰Hg(s) and Hg^II bound to organic matter: OM-¹⁹⁶Hg(s). A traditionally used dissolved Hg^II tracer (¹⁹⁸Hg(NO₃)₂(aq)) was also included and formation of CH₃Hg was determined for all tracers as well as for ambient Hg. Using the Hg(NO₃)₂(aq) tracer overestimated the methylation rate constant, k_m, of ambient Hg up to 15 times, whereas applying solid-phase tracers corresponding to the solid-phase speciation of ambient Hg in the sediment (70% ß-HgS and 30% OM-Hg as determined independently by Hg-EXAFS) yielded k_m values within a factor of 2 of k_m for ambient Hg. Indeed the Hg(NO₃)₂(aq) tracer yielded k_m values 3-6 and 20-60 times higher than the OM-Hg(s) and ß-HgS(s) tracers.

The results show that using proper solid phase tracers will give realistic estimates of ambient Hg methylation rates and that the solid phase speciation of Hg^II is one of the principle factors controlling Hg methylation rates in sediments and that also geochemically aged Hg is available for methylation.

The usefulness of the new isotope tracer methodology to study methylation from new and aged Hg deposits was further explored in a mesocosm study utilizing 12 isolated tubes of 70 cm id and 5 m height. Intact 65 cm diameter sediment cores were spiked with OM-Hg(s), ß-HgS(s) and CH₃Hg-OM(s) isotope tracers and immersed into water-filled mesocosms followed by additions of Hg(NO₃)₂(aq) and CH₃HgCl(aq) tracers to the water phase and formation and degradation of CH₃Hg was studied during 8 weeks.

Net production of methyl mercury in eight boreal wetlands from Sweden was measured by means of; 1) input-output mass balances, 2) experimentally determined methylation (k_m) and MeHg demethylation (k_d) rate constants, and 3) measurements of %MeHg in soil. Based on the nutrient status, as characterized by type of vegetation, pH, C/N ratio and C quality, wetlands were divided into three groups; I) three northern nutrient poor fens, II) a southern nutrient gradient ranging from an ombrotrophic bog to a fen with intermediate nutrient status, and III) nutrient rich southern sites including two mesotrophic wetlands and one alder (Alnus glutinosa) forest swamp. In general, trends in k_m/k_d ratios followed soil %MeHg and was lowest at the most nutritious sites (southern mesotrophic wetlands), intermediate at the northern nutrient poor fens and the southern ombrotrophic bog, and highest at the southern fen site with intermediate nutrient status. Input-output budgets during three years (2007-2009) followed the same pattern with highest net production and export of MeHg from the fen site with intermediate nutrient status. The alder swamp, showing both high k_m and k_d, resulting in an intermediate k_m/k_d ratio and %MeHg, was a sink for MeHg all three years. We conclude that even if most boreal wetlands are sources of MeHg, the net production of MeHg varies depending on the type and nutrient status. Downstream located alder swamps may even act as sinks of MeHg on a landscape level, counteracting upstream net methylation. Results from this study have implications for the policy of wetland restoration.

Reservoirs are regarded as sensitive ecosystems for Hg methylation and subsequent MeHg bioaccumulation in food chain. Here, we present data from six reservoirs with different ages (from 2 years old to more than 30 years old) in Wujiang River Basin to investigate mercury biogeochemical cycling. We investigated the input and output fluxes of total and methyl mercury for all reservoirs to establish a mass balance of mercury for all reservoirs. We found that all reservoirs are the sink for total Hg. We found that newly built reservoirs are the sink of MeHg as well, but old reservoirs are the source of MeHg. We identified that low organic matter content in flooded soil and high pH of water are the main reasons resulting in low MeHg production in reservoir. Eutrophication of reservoirs could result in elevated MeHg production in reservoirs. Aquaculture activity in reservoir is one of the major driving forces of eutrophication of reservoirs in Southwestern China. We investigated the distribution of Hg species in water column and sediment profiles in these reservoirs. We found that organic matter content in sediments which is governed by the primary production of the reservoirs controls the net mercury methylation rates. For newly built reservoirs, organic matter content is low, which is not in favor of Hg methylation. With increase of ages of reservoirs, organic matter content in sediment increased, and mercury methylation process started to increase. Hg methylation processes occurred in the surface of the sediment. This observation is different from those observed in North America and Europe, where elevated Hg methylation processed were observed in newly built reservoirs.

The biogeochemistry of mercury in the Guamium River, basin, a 7,051 ha catchment, that drains an agricultural urban area (upstream), farm land (middle basin) and an urban industrial area (downstream) was assessed. We sampled bottom, and suspended river sediments and the characid fish lambari (Astyanax altiparanae) at nine locations, that were distributed among the three distinct regions of land usage. Fish were caught by electro fishing in the dry season (September, 2005) and in the rainy season (February, 2006), while bottom sediments were collected in July, 2008 (dry season) and March, 2009 (rainy season). Suspended sediment were sampled in March, 2009 at only one location of the three areas. Sediment samples were analyzed for both, bioavailable and total Hg in the <0.063mm fraction. Concentrations of total Hg were determined from liophylized fish samples. Hg concentrations in bottom sediments ranged from 0.040-0.105 mg kg^-1 and 0.018-0.064 mg kg^-1 for total and available Hg respectively for the dry season. During the wet season bottom sediments contained 0.055-0.133 mg kg^-1 total Hg and 0.010-0.052 mg kg^-1 available Hg. Concentrations in the suspended sediments were higher upstream, intermediated in the middle basin and increased downstream, with absolute concentrations being in the same order of magnitude as the bottom sediments. Sediment isotopic composition of ¹³C, ¹⁵N, indicated that up to 70% of the carbon originated from C₄ plants like sugarcane and from river plankton. A C/N ratio of 10-15 suggested sewage inputs to the river. Lambari, which was the most abundantly caught fish species, measured between 44 and 135 mm total length and ranged in Hg concentrations from 0.026 to 0.381 mg Kg^-1 (wet weight). Mean Hg concentrations were higher in fish captured upstream (0.139 mg Kg^-1), than in fish caught from the middle basin (0.074 mg Kg^-1) or downstream area (0.060 mg Kg^-1). No significant relationship between fish length and Hg concentration existed. Hg concentrations in bottom sediments increased in a downstream direction, suggesting that Hg inputs were highest in the urban industrial area.

Porewaters are considered an important biogeochemical compartment for assessing metal contamination and are used to predict the bioavailable and mobile fractions of contaminants in sediment. Yet, concentrations of mercury species in porewaters can be difficult to relate to MeHg concentrations in other environmental compartments, in part because porewaters are generally considered as a single “dissolved” fraction, which ignores the role of colloidal and organic ligand binding on geochemical behavior. In this study of Hg contamination downstream from a point source of contamination, size fractionation of Hg in porewaters was examined by three different techniques: flow-field flow fractionation, size exclusion chromatography, and ultracentrifugation, all coupled with isotope dilution ICP-MS, to interpret the behavior of Hg from a contaminated site.

Sediment, porewater, surface water, and biotic samples were collected from the Berlin Superfund site, in northeastern New Hampshire, USA. This site is a former chlor-alkali facility on the banks of the Androscoggin River, where liquid Hg has been observed seeping from rock fractures along the river bank. Samples from five dam reservoirs, within 12 miles downstream of the site, as well as an upstream site were collected in the summer of 2010. Reservoirs were chosen as they act as both deposition zones for contaminated sediment, as well as good environments for Hg methylation.

Concentrations of Hg in sediment, porewater, surface water, and pelagic fish were all elevated downstream of the former chlor-alkali facility. Partitioning of MeHg and THg from porewater to sediment was correlated to the organic matter content of the sediment. A weak correlation was observed between MeHg in porewater and in suspended particulates in surface water, whereas no relationship was evident with filtered surface water. Conversely, a strong correlation was observed between porewater THg and overlying water. Size fractionation of the porewaters by three techniques showed inorganic Hg to be predominantly colloid-bound, whereas MeHg was distributed between the colloidal and small (<3kDa) organic particles. Separation of “dissolved” MeHg into a smaller size compartment is suggested to give a more useful parameter for assessing mobility and availability.