Arsenic mobility in the ambient sulfidic environment: Sorption of arsenic(V) and arsenic(III) onto disordered mackinawite

Arsenate, As(V), sorption onto synthetic iron(II) monosulfide, disordered mackinawite (FeS), is fast. As(V) sorption decreases above the point of zero surface charge of FeS and follows the pH-dependent concentration of positively charged surface species. No redox reaction is observed between the As(V) ions and the mineral surface over the time span of the experiments. This observation shows that As(V) dominantly forms an outer-sphere complex at the surface of mackinawite. Arsenite, As(III), sorption is not strongly pH-dependent and can be expressed by a Freundlich isotherm. Sorption is fast, although slower than that of As(V). As(III) also forms an outer-sphere complex at the surface of mackinawite. In agreement with previous spectroscopic studies, complexation at low As(V) and As(III) concentration occurs preferentially at the mono-coordinated sulfide edge sites. The K_d (L g⁻¹) values obtained from linear fits to the isotherm data are ∼9 for As(V) and ∼2 for As(III). Stronger sorption of As(V) than As(III), and thus a higher As(III) mobility, may be reflected in natural anoxic sulfidic waters when disordered mackinawite controls arsenic mobility.     

Arsenite adsorption on galena (PbS) and sphalerite (ZnS)

Arsenite, As(III), sorption on galena (PbS) and sphalerite (ZnS) was investigated as a function of solution composition and characterized using X-ray absorption spectroscopy (XAS). Adsorption conformed to a Langmuir isotherm except at the highest surface loadings, and it was not strongly affected by changes in ionic strength. Arsenite sorbed appreciably only at pH > ∼5 for PbS and pH ∼4.5 for ZnS, behavior distinct from its adsorption on other substrates. Arsenite adsorption on PbS and ZnS resulted in the conversion from As-O to As-S coordination. Arsenite does not adsorb through ligand-exchange of surface hydroxyl or sulfhydryl groups. Rather, it forms a polynuclear arsenic sulfide complex on ZnS and PbS consistent with the As₃S₃(SH)₃ trimer postulated by Helz et al. (1995) for sulfidic solutions. This complex was unstable in the presence of oxidizing agents and synchrotron light—it quickly converted to As(V), which was largely retained by the surface. These data illustrate the complexity of As(III) adsorption to even simple sulfide minerals.     

Calculation of the energetics for the oxidation of Sb(III) sulfides by elemental S and polysulfides in aqueous solution

Recent experimental studies have reported the existence of two new Sb sulfide species, Sb₂S₅²⁻ and Sb₂S₆²⁻, in alkaline sulfidic solutions in equilibrium with stibnite, Sb₂S₃, and orthorhombic S. These species contain Sb(V), which has also recently been identified in similar solutions using EXAFS by other researchers. This represents a significant change from the consensus a decade ago that sulfidic solutions of Sb contained only Sb(III) species. I have calculated from first principles of quantum mechanics the energetics for the oxidation of the Sb(III) sulfide dimer Sb₂S₄²⁻ to the mixed Sb(III,V) dimer Sb₂S₅²⁻ and then to the all Sb(V) dimer, Sb₂S₆²⁻. Gas-phase reaction energies have been evaluated using polarized valence double zeta effective core potential basis sets and Moller-Plesset second order treatments of electron correlation. All translational, rotational and vibrational contributions to the gas-phase reaction free energy have been calculated. Hydration energies have been obtained using the COSMO version of the self-consistent reaction field polarizable continuum method. Negative free energy changes are calculated for the oxidation of the dianion of the III,III dimer to the III,V dimer by both small polysulfides, like S₄H⁻, and elemental S, modeled as S₈. For the further oxidation of the III,V dimer to the V,V dimer the reaction free energies are calculated to be close to zero. The partially protonated Sb III,III dimer monoanion HSb₂S₄⁻ can also be oxidized, but the reaction is not so favorable as for the dianion. Comparison of the calculated aqueous deprotonation energies of H₂Sb₂S₄, H₂Sb₂S₅ and H₂Sb₂S₆ and their dianions with values calculated for various oxyacids indicates that the III,V and V,V dimers will have pK_a2 values <5, so that their dianions will be the dominant species in alkaline solutions. These results are thus consistent with the recent identification of Sb₂S₅²⁻ and Sb₂S₆²⁻ species. I have also calculated the Raman spectra of Sb₂S₅²⁻ and Sb₂S₆²⁻ to assist in their identification. The calculated vibrational frequencies of the III,V and V,V dimers are characteristically higher than those of the III,III dimer I previously studied. The III,V dimer may contribute shoulders to the Raman spectrum.     

An evaluation of copper remobilization from mine tailings in sulfidic environments

A laboratory-based assessment of copper remobilization from Cu-rich mine tailings exposed to anoxic, sulfide rich waters was performed. The results from incubation experiments, conducted over a 20 day period, were compared to thermodynamic modelling calculations of copper speciation in sulfidic waters. The tailings materials were observed to react rapidly with added sulfide, consuming 159 μmol HS⁻ g⁻¹ (dry wt) within a 24 h period. The consumption of sulfide was attributed to a two stage process involving the reduction of Fe-hydroxy phases by sulfide followed by reaction with available Fe²⁺ and Cu²⁺ resulting in the formation Fe- and Cu-sulfide phases. During incubation experiments, the dissolved copper concentrations in the absence of sulfide were approximately 0.31 μmol l⁻¹, whereas in the presence of sulfide (0.5–5 mM) concentrations were typically 0.24 μmol l⁻¹. The experiments did not indicate enhanced solubility owing to the formation of soluble copper sulfide species. The predictions (based on the most recent thermodynamic data for aqueous Cu-sulfide and Cu-polysulfide species) did not accurately explain the laboratory observations. Model predictions were greatly influenced by the assumptions made about the oxidation state of copper under anoxic conditions and the solid sulfide phase controlling copper solubility. The study emphasizes the limitations of modelling copper speciation in sulfidic waters and the need for laboratory or field verification of predictions.     

Arsenic-bearing carbonate waters: Geochemical and thermodynamic analysis of their genesis,transformation of arsenic migration species,and models for the genesis of hydrogenic as sulfide mineralization

Analysis of hydrogeochemical materials on As distribution in CO₂-bearing (carbonate) waters in various regions and the thermodynamic simulation of geochemical processes in rock-CO₂-bearing water systems made it possible to constrain the optimal conditions for As transfer from rocks into carbonate waters and the accumulation of this element in the waters. The problem was solved with regard for the various rates of As transitions from rocks to water: (a) high rates of As transitions from rocks in compliance with the ion exchange mechanism and (b) low rates of As transitions from rocks in compliance with the mechanism involving the decomposition of As-bearing minerals. Various mechanisms of As extraction from rocks result in the compositional diversity of the aqueous phase and various As migration species in CO₂-bearing waters, which, in turn, control the equilibrium concentration levels of this element. The principally important boundary conditions of As enrichment in CO₂-bearing waters are high \(P_{CO_2 } \) and R/W ratios in the geochemical systems, a preliminary increase in the Cl concentration in the CO₂-bearing waters, and the origin of these waters at high-density heat fluxes. As migration species were simulated for the model solutions and real carbonate waters of various geochemical types, and it is demonstrated that the predominant As species are oxygen-bearing HAsO ₂ ⁰ , and AsO ₂ ^? at a subordinate role of the sulfide HAs₂S ₄ ^2? , and As₂S ₄ ^2? — species even at high Σ S^2? in the system. Two models of the genesis of solid As sulfides in CO₂-bearing waters are analyzed: (1) with oxygen-bearing species (HAsO ₂ ⁰ , and AsO ₂ ^? ), which occur most widely, and (2) with sulfide species (As₂S ₄ ^2? , HAs₂S ₄ ^? , and As₄S ₇ ^2? ), which occur not as widely.     

The behavior of molybdenum and its isotopes across the chemocline and in the sediments of sulfidic Lake Cadagno, Switzerland

Molybdenum (Mo) isotope studies in black shales can provide information about the redox evolution of the Earth’s oceans, provided the isotopic consequences of Mo burial into its major sinks are well understood. Previous applications of the Mo isotope paleo-ocean redox proxy assumed quantitative scavenging of Mo when buried into sulfidic sediments. This paper contains the first complete suite of Mo isotope fractionation observations in a sulfidic water column and sediment system, the meromictic Lake Cadagno, Switzerland, a small alpine lake with a pronounced oxygen-sulfide transition reaching up to H₂S ∼ 200 μM in the bottom waters (or about 300 μM total sulfide: ΣS²⁻ = H₂S + HS⁻ + S²⁻). We find that Mo behaves conservatively in the oxic zone and non-conservatively in the sulfidic zone, where dissolved Mo concentrations decrease from 14 nM to 2-8 nM across this transition. Dissolved Mo in the upper oxic waters has a δ⁹⁸Mo_oxic = 0.9 ± 0.1‰, which matches that of the riverine input, δ⁹⁸Mo_river = 0.9 ± 0.1‰. In the deeper sulfidic waters, a subaquatic source delivers Mo at 1.55 ± 0.1‰, but the dissolved Mo is even heavier at δ⁹⁸Mo_sulfidic = 1.8‰. Sediment traps in the sulfidic zone of the lake collect particles increasingly enriched in Mo with depth, with δ⁹⁸Mo values significantly fractionated at −0.8‰ to −1.2‰ both near the chemocline and in the deepest trap. Suspended particulates in the sulfidic waters carry lighter Mo than the ambient dissolved Mo pool by ∼0.3-1.5‰. Sedimentary Mo concentrations correlate with total organic carbon and yield Mo levels which are two orders of magnitude higher than typical crustal values found in rocks from the catchment area. Solid-phase Mo in the sediment shows a slightly positive δ⁹⁸Mo trend with depth, from δ⁹⁸Mo = 1.2‰ to 1.4‰ while the pore waters show dramatic enrichments of Mo (>2000 nM) with a relatively light isotope signature of δ⁹⁸Mo = 0.9-1.0‰.These data are explained if Mo is converted to particle-reactive oxythiomolybdates in the sulfidic waters and is fractionated during removal from solution onto particles. Isotope fractionation is expressed in the water column, despite the high sulfide concentrations, because the rate of Mo removal is fast compared to the slow reaction kinetics of thiomolybdate formation. However, elemental and isotopic mass balances show that Mo is indeed quantitatively removed to the lake sediments and thus the isotopic composition of the sediments reflects sources to the sulfidic water. This efficient Mo drawdown is expected to occur in settings where H₂S is very much in excess over Mo or in a restricted setting where the water renewal rate is slow compared to the Mo burial rate. We present a model for the Mo isotope fractionation in sulfidic systems associated with the slow reaction kinetics and conclude that quantitative removal will occur in highly sulfidic and restricted marine systems.     

Regulation of bacterial sulfate reduction and hydrogen sulfide fluxes in the central namibian coastal upwelling zone

The coastal upwelling system off central Namibia is one of the most productive regions of the oceans and is characterized by frequently occurring shelf anoxia with severe effects for the benthic life and fisheries. We present data on water column dissolved oxygen, sulfide, nitrate and nitrite, pore water profiles for dissolved sulfide and sulfate,³⁵S-sulfate reduction rates, as well as bacterial counts of large sulfur bacteria from 20 stations across the continental shelf and slope. The stations covered two transects and included the inner shelf with its anoxic and extremely oxygen-depleted bottom waters, the oxygen minimum zone on the continental slope, and the lower continental slope below the oxygen minimum zone. High concentrations of dissolved sulfide, up to 22 mM, in the near-surface sediments of the inner shelf result from extremely high rates of bacterial sulfate reduction and the low capacity to oxidize and trap sulfide. The inner shelf break marks the seaward border of sulfidic bottom waters, and separates two different regimes of bacterial sulfate reduction. In the sulfidic bottom waters on the shelf, up to 55% of sulfide oxidation is mediated by the large nitrate-storing sulfur bacteria, Thiomargarita spp. The filamentous relatives Beggiatoa spp. occupy low-O₂ bottom waters on the outer shelf. Sulfide oxidation on the slope is apparently not mediated by the large sulfur bacteria. The data demonstrate the importance of large sulfur bacteria, which live close to the sediment-water interface and reduce the hydrogen sulfide flux to the water column. Modeling of pore water sulfide concentration profiles indicates that sulfide produced by bacterial sulfate reduction in the uppermost 16 cm of sediment is sufficient to account for the total flux of hydrogen sulfide to the water column. However, the total pool of hydrogen sulfide in the water column is too large to be explained by steady state diffusion across the sediment-water interface. Episodic advection of hydrogen sulfide, possibly triggered by methane eruptions, may contribute to hydrogen sulfide in the water column.     

Arsenic and Antimony in Groundwater Flow Systems: A Comparative Study

Arsenic (As) and antimony (Sb) concentrations and speciation were determined along flow paths in three groundwater flow systems, the Carrizo Sand aquifer in southeastern Texas, the Upper Floridan aquifer in south-central Florida, and the Aquia aquifer of coastal Maryland, and subsequently compared and contrasted. Previously reported hydrogeochemical parameters for all three aquifer were used to demonstrate how changes in oxidation–reduction conditions and solution chemistry along the flow paths in each of the aquifers affected the concentrations of As and Sb. Total Sb concentrations (Sb_T) of groundwaters from the Carrizo Sand aquifer range from 16 to 198 pmol kg⁻¹; in the Upper Floridan aquifer, Sb_T concentrations range from 8.1 to 1,462 pmol kg⁻¹; and for the Aquia aquifer, Sb_T concentrations range between 23 and 512 pmol kg⁻¹. In each aquifer, As and Sb (except for the Carrizo Sand aquifer) concentrations are highest in the regions where Fe(III) reduction predominates and lower where SO₄ reduction buffers redox conditions. Groundwater data and sequential analysis of the aquifer sediments indicate that reductive dissolution of Fe(III) oxides/oxyhydroxides and subsequent release of sorbed As and Sb are the principal mechanism by which these metalloids are mobilized. Increases in pH along the flow path in the Carrizo Sand and Aquia aquifer also likely promote desorption of As and Sb from mineral surfaces, whereas pyrite oxidation mobilizes As and Sb within oxic groundwaters from the recharge zone of the Upper Floridan aquifer. Both metalloids are subsequently removed from solution by readsorption and/or coprecipitation onto Fe(III) oxides/oxyhydroxides and mixed Fe(II)/Fe(III) oxides, clay minerals, and pyrite. Speciation modeling using measured and computed Eh values predicts that Sb(III) predominate in Carrizo Sand and Upper Floridan aquifer groundwaters, occurring as the Sb(OH)₃⁰ species in solution. In oxic groundwaters from the recharge zones of these aquifers, the speciation model suggests that Sb(V) occurs as the negatively charged Sb(OH)₆⁻ species, whereas in sufidic groundwaters from both aquifers, the thioantimonite species, HSb₂S₄ ⁻ and Sb₂S₄ ²⁻, are predicted to be important dissolved forms of Sb. The measured As and Sb speciation in the Aquia aquifer indicates that As(III) and Sb(III) predominate. Comparison of the speciation model results based on measured Eh values, and those computed with the Fe(II)/Fe(III), S(-II)/SO₄, As(III)/As(V), and Sb(III)/Sb(V) couples, to the analytically determined As and Sb speciation suggests that the Fe(II)/Fe(III), S(-II)/SO₄ couples exert more control on the in situ redox condition of these groundwaters than either metalloid redox couple.     

Evidence for aqueous clusters as intermediates during zinc sulfide formation

Using zinc sulfide as an example, we demonstrate a plausible stepwise process for the formation of minerals from low temperature aqueous solutions. The process occurs with the formation of soluble complexes that aggregate into soluble rings and clusters. The final moiety in solution has a structure similar to the moiety in the first formed solid, which is a restatement of the Ostwald step rule. Titrations of aqueous Zn(II) with bisulfide indicate that sulfide clusters form at concentrations of 20 μM (or less) of metal and bisulfide. Precipitation does not occur according to voltammetric measurements using a mercury electrode and UV-VIS (ultra-violet to visible) spectroscopic data. UV-VIS data and filtration experiments indicate that the material passes through 0.1 μm Nuclepore and 1000 dalton filters. The complexes form rapidly (k_f > 10⁸ Ms⁻¹), are kinetically inert to dissociation and thermodynamically strong. Although a neutral complex of 1:1 (ZnS) empirical stoichiometry initially forms, an anionic complex with an empirical 2 Zn:3 S stoichiometry results with continued addition of sulfide. Gel electrophoresis confirms the existence of a cluster that is negatively charged with a molecular mass between 350 and 750 daltons. On the basis of known mineral and thiol complex structures for these systems, a tetrameric cluster unit of Zn₄S₆(H₂O)₄⁴⁻ is likely. Molecular mechanic calculations show that this cluster is structurally analogous to ZnS minerals (particularly sphalerite) and is a viable precursor to mineral formation and a product of mineral dissolution.The formation of Zn₄S₆(H₂O)₄⁴⁻ can occur from condensation of Zn₃S₃(H₂O)₆ rings, which are neutral molecular clusters. The Zn atoms on one Zn₃S₃(H₂O)₆ ring combine with the S atoms on another Zn₃S₃(H₂O)₆, to lead to higher order clusters with loss of water. The Zn₄S₆⁴⁻ species form by the cross-linking of two neutral Zn₃S₃ rings by added sulfide; thus a Zn–S–Zn bridge forms across the rings with subsequent rearrangement and condensation to Zn₄S₆⁴⁻; this combination results in a sphalerite-like cluster. If the rings condense without additional sulfide, a wurtzite-like structure could form. All condensations result in sulfide displacement of water from Zn to form Zn–S bonds. Water loss is an example of an entropy-driven process, which leads to a more favorable thermodynamic process. These clusters would be resistant to oxidation by O₂. Voltammetric experiments indicate neutral and anionic clusters for Zn and agree with ion chromatographic data from the sulfidic waters of the Black Sea.     

Arsenic speciation in Mono Lake, California: Response to seasonal stratification and anoxia

Mono Lake is a closed-basin, alkaline, hypersaline lake located at the western edge of the Great Basin in eastern California. We studied the distribution of arsenic (As) species in the water column of Mono Lake between February and November, 2002. This period captured the seasonal progression from winter mixing, through summer thermal stratification, to autumn overturn. Arsenic speciation was determined by ion chromatography-inductively coupled-plasma-mass spectrometry of samples preserved in the field by flash-freezing in liquid nitrogen. We found that arsenic speciation was dominated (>90%) by arsenate when oxygen was detectable. Once levels fell below 6 μmol/L O₂, arsenic speciation shifted to dominance by reduced species. Arsenate and arsenite co-occurred in a transition zone immediately below the base of the oxycline and low but significant concentrations of arsenate were occasionally detected in sulfidic hypolimnion samples. Thio-arsenic species were the dominant form of As found in sulfidic waters. Maxima of thio-arsenic species with stoichiometries consistent with mono-, di- and trithio-arsenic occurred in succession as sulfide concentration increased. A compound with a stoichiometry consistent with trithio-arsenic was the dominant As species (∼50% of total As) in high sulfide (2 mmol/L) bottom water. Lower concentrations of total As in bottom water relative to surface water suggest precipitation of As/S mineral phases in response to sulfide accumulation during prolonged anoxia.     

Biogeochemical redox cycling of arsenic in mine-impacted lake sediments and co-existing pore waters near Giant Mine,Yellowknife Bay,Canada

Lacustrine sediments, submerged tailings, and their pore waters have been collected at several sites in Yellowknife Bay, Great Slave Lake, Canada, in order to investigate the biogeochemical controls on the remobilization of As from mining-impacted materials under different depositional conditions. Radiometric dating confirms that a mid-core enrichment of Pb, Zn, Cu and Sb corresponds to the opening of a large Au mine 60 a ago. This was evident even in a relatively remote site. Arsenic was enriched at mid-core, coincident with mining activity, but clearly exhibited post-depositional mobility, migrating upwards towards the sediment water interface (SWI) as well as down-core. Deep-water (15 m) Yellowknife Bay sediments that contain buried mine waste are suboxic, relatively organic-rich and abundant in microbes with As in pore waters and sediments reaching 585 μg/L and 1310 mg/kg, respectively. Late summer pore waters show equal proportions of As(III) and As(V) (16–415 μg/L) whereas late winter pore waters are dominated by As(III) (284–947 μg/L). This can be explained by As(III) desorption mechanisms associated with the conversion of FeS to FeS₂ and the reduction of As(V) to As(III) through the oxidation of dissolved sulfide, both microbially-mediated processes. Processes affecting As cycling involve the attenuating efficiency of the oxic zone at the SWI, sediment redox heterogeneity and the reductive dissolution of Fe(hydr)oxides by labile organic matter, temporarily and spatially variable.     

Geochemical evolution of Sb migration species in carbonate and thermal waters in relation to Sb hydrogenic mineralization

Based on the synthesis of hydrogeochemical materials on Sb occurrence in carbonate and thermal waters and thermodynamic simulations, genetic analysis was conducted of the transformations of probable Sb migration species (particularly oxygen-bearing and sulfide ones), and their transformations were calculated depending on the main parameters of hydrogeochemical systems (\(P_{CO_2 } \), T, R/W, Eh, and pH). The oxygen 2HSbO ₂ ⁰ + 3H₂S = Sb₂S₃ + 4H₂O (2SbO ₂ ^? + 3HS^? + 5H⁺ = Sb₂S₃ + 4H₂O) and sulfide HSb₂S ₄ ^? + H⁺ = Sb₂S₃ cr + H₂S (Sb₂S ₄ ^2? + 2H⁺ = Sb₂S₃cr + H₂S) models for the genesis of hydrogenic Sb₂S₃(cr) were simulated. Information on occurrences of carbonate and thermal waters in various regions worldwide was generalized, and the reasons were identified for the geochemical separation of As and Sb in carbonate and thermal waters. The causes and conditions of an increase in Sb concentrations in thermal waters were revealed, and Sb migration species in carbonate and thermal waters were identified for various parameters of hydrogeochemical systems. Variations in Sb speciation were demonstrated for hydrogeochemical systems depending on their boundary conditions (\(P_{CO_2 } \), T, and R/W). Models were outlined responsible for the precipitation of Sb₂S₃(cr) from carbonate and thermal waters.     

Arsenic Geochemistry of the Great Dismal Swamp,Virginia, USA: Possible Organic Matter Controls

Surface water samples for arsenic (As) concentration and speciation analysis were collected from organic matter-rich blackwaters of the Lake Drummond portion of the Great Dismal Swamp in southeastern Virginia, USA. Arsenic concentrations and speciation were determined by selective hydride generation, gas chromatography with photoionization detection. Surface waters from the Great Dismal Swamp are high in dissolved organic carbon (DOC) concentrations (445–9,600 μmol/kg) and of low pH (4.2–6.4). Total dissolved As concentrations [i.e., As(III) + As(V)], hereafter As_T, range from 2.2 nmol/kg to 21.4 nmol/kg. Arsenite, As(III), concentrations range from ∼1 nmol/kg to 17.7 nmol/kg, and As(V) ranges from ∼1 nmol/kg to 14.1 nmol/kg. Arsenate, As(V), is the predominant form of dissolved As in the inflow waters to the Great Dismal Swamp, whereas within the swamp proper arsenite, As(III), dominates. Arsenite accounts for 8–37% of As_T in inflow waters west of the Suffolk Scarp, and between 54% and 81% of As_T in Lake Drummond and Great Dismal Swamp waters east of the scarp. Arsenite is strongly correlated to DOC (r = 0.94) and inversely related to pH (r = −0.9), both at greater than the 99% confidence level. Arsenate is weakly related to pH and DOC (r = 0.4 and −0.37, respectively), and neither relationship is statistically significant. No statistical relationships exist between As(V) or As(III) and PO₄ concentrations. The predominance of As(III) and its strong correlation with DOC in Great Dismal Swamp waters suggest that DOC may inhibit As(III) adsorption or form stable aqueous complexes with As(III) in these waters. Alternatively, phytoplankton and/or bacterially mediated reduction of As(V) may be important processes in the organic-rich blackwaters and/or sediment porewaters of the swamp, leading to the prevalence of As(III) in the water column.     

The influence of redox processes on trace element mobility in a sandy aquifer—an experimental approach

This study is concerned with the geochemical behavior of Cu, Zn, As, Mo, Ba, La and Ce under the reducing conditions of a bank infiltration system. To identify and interpret individual processes laboratory experiments were performed on columns containing sandy sediments of an anoxic pleistocene aquifer from the Ruhr valley (western Germany). The flow rate being the key variable was varied from 0.21 to 0.46 and 0.82 m/d (meters per day), whereas the concentration of the reducing agent (acetate) remained constant. A second experiment lasting 80 weeks was carried out at a flow rate of 0.80 m/d, in order to characterize interactions between the pore water and solid phase. The results of these experiments show that the geochemistry of the trace elements involved can be explained to a large extent by the major redox processes of manganese, iron and sulfate turnover. The trace elements As, Mo, Ce and La were released into the pore water as a result of the reduction of Mn- and Fe-(hydr)oxides within the zone of major redox processes. Cu and Zn were removed from the infiltrating water within the first centimeters of the column by sulfide (co)precipitation, whereas the mobility of Mo was controlled by sulfidic fixation further down the flow path. As a result of the decreasing S²⁻-concentrations along the flow path, dissolved As(III) was re-oxidized (E_H>− 250 mV) and precipitated as As(V) in a barium–arsenate-phase.     

Arsenite sorption on troilite (FeS) and pyrite (FeS2)

Arsenic is a toxic metalloid whose mobility and availability are largely controlled by sorption on sulfide minerals in anoxic environments. Accordingly, we investigated reactions of As(III) with iron sulfide (FeS) and pyrite (FeS₂) as a function of total arsenic concentration, suspension density, sulfide concentration, pH, and ionic strength. Arsenite partitioned strongly on both FeS and FeS₂ under a range of conditions and conformed to a Langmuir isotherm at low surface coverages; a calculated site density of near 2.6 and 3.7 sites/nm² for FeS and FeS₂, respectively, was obtained. Arsenite sorbed most strongly at elevated pH (>5 to 6). Although solution data suggested the formation of surface precipitates only at elevated solution concentrations, surface precipitates were identified using X-ray absorption spectroscopy (XAS) at all coverages. Sorbed As was coordinated to both sulfur [d(As-S) = 2.35 Å] and iron [d(As-Fe) = 2.40 Å], characteristic of As coordination in arsenopyrite (FeAsS). The absorption edge of sorbed As was also shifted relative to arsenite and orpiment (As₂S₃), revealing As(III) reduction and a complete change in As local structure. Arsenic reduction was accompanied by oxidation of both surface S and Fe(II); the FeAsS-like surface precipitate was also susceptible to oxidation, possibly influencing the stability of As sorbed to sulfide minerals in the environment. Sulfide additions inhibit sorption despite the formation of a sulfide phase, suggesting that precipitation of arsenic sulfide is not occurring. Surface precipitation of As on FeS and FeS₂ supports the observed correlation of arsenic and pyrite and other iron sulfides in anoxic sediments.     

Natural arsenic attenuation via metal arsenate precipitation in soils contaminated with metallurgical wastes: III. Adsorption versus precipitation in clean As(V)/goethite/Pb(II)/carbonate systems

Soil contamination with As and potentially harmful metals is a widespread problem around the world especially from mining and metallurgical wastes, which release substantial amounts of these elements to the environment in potentially mobile species. Recently, it has been found that in various Mexican soils contaminated with these types of wastes, arsenate is not in the form of sorbed species on Fe oxides present in the soils, as generally reported in the literature, but in the form of very insoluble compounds such as Pb, Cu and Ca arsenates. Here a thermodynamic model is applied and validated with the results from wet chemical experiments to determine the fundamental geochemical conditions governing the mobility of As in the presence of Pb. For this purpose, a relatively simple but fundamental system of goethite (α-FeOOH)/As(V)/Pb(II)/carbonate was defined as a function of the As(V)/Fe(III) ratio, in a pH range of 5–10. The speciation model included the simultaneous inclusion of triple layer surface complexation and arsenate precipitation equilibria. The model predicts that from very low total As(V)/Fe(III) molar ratios (0.012 at pH 7) the precipitation mechanism significantly influences the attenuation of As(V), and rapidly becomes the dominant process over the adsorption mechanism. Model results identify the quantitative conditions of predominance for each mechanism and describe the transition conditions in which relatively large fractions of adsorbed, precipitated and dissolved As(V) species prevail. Experimental measurements at selected As(V)/Fe(III) ratios and pH confirmed the predictions and validated the coupled thermodynamic model utilized.     

Element discharge from pyritic mine tailings at limited oxygen availability in column experiments

Sulfide mineral oxidation in mine tailings deposits poses a long term threat to surrounding ground water and surface waters. Soil or water cover remediation aims at reducing the rate of sulfide mineral oxidation by decreasing the O₂ ingress rate. In this study, the authors addressed the rate of sulfide oxidation and pH buffering in ∼33 months long, well-controlled laboratory studies of water saturated columns of sulfidic mine tailings from the Kristineberg site in Sweden at reduced O₂ availability. The element discharge rates slowly declined towards a quasi-steady state over hundreds of days. Non-reactive tracer tests showed an anomalously large dispersion, indicating strong flow heterogeneity, possibly including preferential flow and/or stagnant water zones. Congruent dissolution of pyrite and sphalerite by injected oxidants (dissolved O₂ and Fe(III)) adequately explained the discharge rate of Fe, S and Zn at quasi-steady state. Arsenic, Pb and Cu were partly retained in the tailings. Base cation discharge rates, and thus pH buffering, were apparently controlled by the rate of acidity production, with actual pH levels, available mineral surface area, and water residence times being of less importance.     

Geochemical modeling of arsenic sulfide oxidation kinetics in a mining environment

Arsenic sulfide (AsS (am), As₂S₃ (am), orpiment, and realgar) oxidation rates increase with increasing pH values. The rates of arsenic sulfide oxidation at higher pH values relative to those at pH∼2 are in the range of 26-4478, 3-17, 8-182, and 4-10 times for As₂S₃ (am), orpiment, AsS (am), and realgar, respectively.Numerical simulations of orpiment and realgar oxidation kinetics were conducted using the geochemical reaction path code EQ3/6 to evaluate the effects of variable DO concentrations and mineral reactivity factors on water chemistry evolution during orpiment and realgar oxidation. The results show that total As concentrations increase by ∼1.14 to 13 times and that pH values decrease by ∼0.6 to 4.2 U over a range of mineral reactivity factors from 1% to 50% after 2000 days (5.5 yr). The As release from orpiment and realgar oxidation exceeds the current U.S. National Drinking Water Standard (0.05 ppm) approximately in 200-300 days at the lowest initial dissolved oxygen concentration (3 ppm) and a reactivity factor of 1%. The results of simulations of orpiment oxidation in the presence of albite and calcite show that calcite can act as an effective buffer to the acid water produced from orpiment oxidation within relatively short periods (days/months), but the release of As continues to increase.Pyrite oxidation rates are faster than orpiment and realgar from pH 2.3 to 8; however, pyrite oxidation rates are slower than As₂S₃ (am) and AsS (am) at pH 8. The activation energies of arsenic sulfide oxidation range from 16 to 124 kJ/mol at pH∼8 and temperature 25 to 40°C, and pyrite activation energies are ∼52 to 88 kJ/mol, depending on pH and temperature range. The magnitude of activation energies for both pyrite and arsenic sulfide solids indicates that the oxidation of these minerals is dominated by surface reactions, except for As₂S₃ (am). Low activation energies of As₂S₃ (am) indicate that diffusion may be rate controlling.Limestone is commonly mixed with sulfide minerals in a mining environment to prevent acid water formation. However, the oxidation rates of arsenic sulfides increase as solution pH rises and result in a greater release of As. Furthermore, the lifetimes of carbonate minerals (i.e., calcite, aragonite, and dolomite) are much shorter than those of arsenic sulfide and silicate minerals. Thus, within a geologic frame time, carbonate minerals may not be present to act as a pH buffer for acid mine waters. Additionally, the presence of silicate minerals such as pyroxenes (wollastonite, jadeite, and spodumene) and Ca-feldspars (labradorite, anorthite, and nepheline) may not be important for buffering acid solutions because these minerals dissolve faster than and have shorter lifetimes than sulfide minerals. However, other silicate minerals such as Na and K-feldspars (albite, sanidine, and microcline), quartz, pyroxenes (augite, enstatite, diopsite, and MnSiO₃) that have much longer lifetimes than arsenic sulfide minerals may be present in a system. The results of our modeling of arsenic sulfide mineral oxidation show that these minerals potentially can release significant concentrations of dissolved As to natural waters, and the factors and mechanisms involved in arsenic sulfide oxidation warrant further study.     

Arsenic repartitioning during biogenic sulfidization and transformation of ferrihydrite

Iron (hydr)oxides are strong sorbents of arsenic (As) that undergo reductive dissolution and transformation upon reaction with dissolved sulfide. Here we examine the transformation and dissolution of As-bearing ferrihydrite and subsequent As repartitioning amongst secondary phases during biotic sulfate reduction. Columns initially containing As(V)-ferrihydrite coated sand, inoculated with the sulfate reducing bacteria Desulfovibrio vulgaris (Hildenborough), were eluted with artificial groundwater containing sulfate and lactate. Rapid and consistent sulfate reduction coupled with lactate oxidation is observed at low As(V) loading (10% of the adsorption maximum). The dominant Fe solid phase transformation products at low As loading include amorphous FeS within the zone of sulfate reduction (near the inlet of the column) and magnetite downstream where Fe(II)_(aq) concentrations increase; As is displaced from the zone of sulfidogenesis and Fe(III)_(s) depletion. At high As(V) loading (50% of the adsorption maximum), sulfate reduction and lactate oxidation are initially slow but gradually increase over time, and all As(V) is reduced to As(III) by the end of experimentation. With the higher As loading, green rust(s), as opposed to magnetite, is a dominant Fe solid phase product. Independent of loading, As is strongly associated with magnetite and residual ferrihydrite, while being excluded from green rust and iron sulfide. Our observations illustrate that sulfidogenesis occurring in proximity with Fe (hydr)oxides induce Fe solid phase transformation and changes in As partitioning; formation of As sulfide minerals, in particular, is inhibited by reactive Fe(III) or Fe(II) either through sulfide oxidation or complexation.     

Speciation of rare earth elements in natural terrestrial waters: assessing the role of dissolved organic matter from the modeling approach

Humic Ion-Binding Model V, which focuses on metal complexation with humic and fulvic acids, was modified to assess the role of dissolved natural organic matter in the speciation of rare earth elements (REEs) in natural terrestrial waters. Intrinsic equilibrium constants for cation-proton exchange with humic substances (i.e., pK_MHA for type A sites, consisting mainly of carboxylic acids), required by the model for each REE, were initially estimated using linear free-energy relationships between the first hydrolysis constants and stability constants for REE metal complexation with lactic and acetic acid. pK_MHA values were further refined by comparison of calculated Model V “fits” to published data sets describing complexation of Eu, Tb, and Dy with humic substances. A subroutine that allows for the simultaneous evaluation of REE complexation with inorganic ligands (e.g., Cl⁻, F⁻, OH⁻, SO₄²⁻, CO₃²⁻, PO₄³⁻), incorporating recently determined stability constants for REE complexes with these ligands, was also linked to Model V. Humic Ion-Binding Model V’s ability to predict REE speciation with natural organic matter in natural waters was evaluated by comparing model results to “speciation” data determined previously with ultrafiltration techniques (i.e., organic acid-rich waters of the Nsimi-Zoetele catchment, Cameroon; dilute, circumneutral-pH waters of the Tamagawa River, Japan, and the Kalix River, northern Sweden). The model predictions compare well with the ultrafiltration studies, especially for the heavy REEs in circumneutral-pH river waters. Subsequent application of the model to world average river water predicts that organic matter complexes are the dominant form of dissolved REEs in bulk river waters draining the continents. Holding major solute, minor solute, and REE concentrations of world average river water constant while varying pH, the model suggests that organic matter complexes would dominate La, Eu, and Lu speciation within the pH ranges of 5.4 to 7.9, 4.8 to 7.3, and 4.9 to 6.9, respectively. For acidic waters, the model predicts that the free metal ion (Ln³⁺) and sulfate complexes (LnSO₄⁺) dominate, whereas in alkaline waters, carbonate complexes (LnCO₃⁺ + Ln[CO₃]₂⁻) are predicted to out-compete humic substances for dissolved REEs. Application of the modified Model V to a “model” groundwater suggests that natural organic matter complexes of REEs are insignificant. However, groundwaters with higher dissolved organic carbon concentrations than the “model” groundwater (i.e., >0.7 mg/L) would exhibit greater fractions of each REE complexed with organic matter. Sensitively analysis indicates that increasing ionic strength can weaken humate-REE interactions, and increasing the concentration of competitive cations such as Fe(III) and Al can lead to a decrease in the amount of REEs bound to dissolved organic matter.