Quantifying inorganic sources of geochemical energy in hydrothermal ecosystems, Yellowstone National Park, USA

Combining analytical data from hot spring samples with thermodynamic calculations permits a quantitative assessment of the availability and ranking of various potential sources of inorganic chemical energy that may support microbial life in hydrothermal ecosystems. Yellowstone hot springs of diverse geochemical composition, ranging in pH from <2 to >9 were chosen for this study, and dozens of samples were collected during three field seasons. Field measurements of dissolved oxygen, nitrate, nitrite, total ammonia, total sulfide, alkalinity, and ferrous iron were combined with laboratory analyses of sulfate and other major ions from water samples, and carbon dioxide, hydrogen, methane, and carbon monoxide in gas samples to evaluate activity products for ∼300 coupled oxidation-reduction reactions. Comparison of activity products and independently calculated equilibrium constants leads to values of the chemical affinity for each of the reactions, which quantifies how far each reaction is from equilibrium. Affinities, in turn, show systematic behavior that is independent of temperature but can be correlated with pH of the hot springs as a proxy for the full spectrum of geochemical variability. Many affinities are slightly to somewhat dependent on pH, while others are dramatically influenced by changes in chemical composition. All reactions involving dissolved oxygen as the electron acceptor are potential energy sources in all hot spring samples collected, but the ranking of dominant electron donors changes from ferrous iron, and sulfur at high pH to carbon monoxide, hydrogen, and magnetite as pH decreases. There is a general trend of decreasing energy yields depending on electron acceptors that follows the sequence: O₂(aq) > NO₃⁻ ≈ NO₂⁻ ≈ S > pyrite ≈ SO₄⁻² ≈ CO(g) ≈ CO₂(g) at high pH, and O₂(aq) ≈ magnetite > hematite ≈ goethite > NO₃⁻ ≈ NO₂⁻ ≈ S ≈ pyrite ≈ SO₄⁻² at low pH. Many reactions that are favorable sources of chemical energy at one set of geochemical conditions fail to provide energy at other conditions, and vice versa. This results in energy profiles supplied by geochemical processes that provide fundamentally different foundations for chemotrophic microbial communities as composition changes.     

The oxidation state of sulfur in synthetic and natural glasses determined by X-ray absorption spectroscopy

Sulfur K-edge X-ray absorption near edge structure (XANES) spectra were recorded for experimental glasses of various compositions prepared at different oxygen fugacities (fO₂) in one-atmosphere gas-mixing experiments at 1400 °C. This sample preparation method only results in measurable S concentrations under either relatively reduced (log fO₂ < −9) or oxidised (log fO₂ > −2) conditions. The XANES spectra of the reduced samples are characterised by an absorption edge crest at 2476.4 eV, typical of S²⁻. In addition, spectra of Fe-bearing compositions exhibit a pronounced absorption edge shoulder. Spectra for all the Fe-free samples are essentially identical, as are the spectra for the Fe-bearing compositions, despite significant compositional variability within each group. The presence of a sulfide phase, such as might exsolve on cooling, can be inferred from a pre-edge feature at 2470.5 eV.The XANES spectra of the oxidised samples are characterised by an intense transition at 2482.1 eV, typical of the sulfate anion SO₄²⁻. Sulfite (SO₃²⁻) has negligible solubility in silicate melts at low pressures. The previous identification of sulfite species in natural glass samples is attributed to an artefact of the analysis (photoreduction of S⁶⁺). S⁴⁺ does, however, occur unambiguously with S⁶⁺ in Fe-free and Fe-poor compositions prepared in equilibrium with CaSO₄ at 4-16 kbar, and when buffered with Re/ReO₂ at 10 kbar. Solubility of S⁴⁺ thus requires partial pressures of SO₂ considerably in excess of 1 bar. A number of experiments were undertaken in an attempt to access intermediate fO₂s more applicable to terrestrial volcanism. Although these were largely unsuccessful, S²⁻ and S⁶⁺ were found to coexist in some samples that were not in equilibrium with the imposed fO₂.The XANES spectra of natural olivine-hosted melt inclusions and submarine glasses representative of basalts at, or close to, sulfide saturation show mainly dissolved S²⁻, but with minor sulfate, and additionally a peak at 2469.5 eV, which, although presumably due to immiscible sulfide, is 1 eV lower than that typical of FeS. These sulfate and sulfide-related peaks disappear with homogenisation of the inclusions by heating to 1200 °C followed by rapid quenching, suggesting that both these features are a result of cooling under natural conditions. The presence of small amounts of sulfate in otherwise reduced basaltic magmas may be explained by the electron exchange reaction: S²⁻ + 8Fe³⁺ = S⁶⁺ + 8Fe²⁺, which is expected to proceed strongly to the right with decreasing temperature. This reaction would explain why S²⁻ and S⁶⁺ are frequently found together despite the very limited fO₂ range over which they are thermodynamically predicted to coexist. The S XANES spectra of water-rich, highly oxidised, basaltic inclusions hosted in olivine from Etna and Stromboli confirm that nearly all S is dissolved as sulfate, explaining their relatively high S contents.     

Tracing industrial sulfur contributions to atmospheric sulfate deposition in the Athabasca oil sands region,Alberta, Canada

Anthropogenic S emissions in the Athabasca oil sands region (AOSR) in Alberta, Canada, affect SO₄ deposition in close vicinity of industrial emitters. Between May 2008 and May 2009, SO₄-S deposition was monitored using open field bulk collectors at 15 sites and throughfall collectors at 14 sites at distances between 3 and 113 km from one of the major emission stacks in the AOSR. At forested plots >90 km from the operations, SO₄ deposition was ∼1.4 kg SO₄-S ha⁻¹ yr⁻¹ for bulk deposition and ∼3.3 kg SO₄-S ha⁻¹ yr⁻¹ for throughfall deposition. Throughfall SO₄ deposition rates in the AOSR exceeded bulk deposition rates at all sites by a factor of 2–3, indicating significant inputs of dry deposition especially under forest canopies. Both bulk and throughfall SO₄ deposition rates were elevated within 29 km distance of the industrial operations with deposition rates as high as 11.7 kg SO₄-S ha⁻¹ yr⁻¹ for bulk deposition and 39.2 kg SO₄-S ha⁻¹ yr⁻¹ for throughfall at industrial sites. Sulfur isotope ratio measurements of atmospheric SO₄ deposited in the AOSR revealed that at a few selected locations ³⁴S-depleted SO₄, likely derived from H₂S emissions from tailing ponds contributes to local atmospheric SO₄ deposition. In general, however, δ³⁴S values of SO₄ deposition at distant forested plots (>74 km) with low deposition rates were not isotopically different from δ³⁴S values at sites with high deposition rates in the AOSR and are, therefore, not suitable to determine industrial S contributions. However, O isotope ratios of atmospheric SO₄ in bulk and throughfall deposition in the AOSR showed a distinct trend of decreasing δ¹⁸O-SO₄ values with increasing SO₄ deposition rates allowing quantification of industrial contributions to atmospheric SO₄ deposition. Two-end-member mixing calculations revealed that open field bulk SO₄ deposition especially at industrial sites in close proximity (<29 km) to the operations is significantly (17–59%) affected by industrial S emissions and that throughfall generally contained 49–100% SO₄ of industrial origin. Hence, it is suggested that δ¹⁸O values of SO₄ may constitute a suitable tracer for quantifying industrial contributions to atmospheric SO₄ deposition in the AOSR.     

The role of sulfur in chemical weathering and atmospheric CO2 fluxes: Evidence from major ions, δCDIC, and δSSO4 in rivers of the Canadian Cordillera

Water samples from the Fraser, Skeena and Nass River basins of the Canadian Cordillera were analyzed for dissolved major element concentrations (HCO₃⁻, SO₄²⁻, Cl⁻, Ca²⁺, Mg²⁺, K⁺, Na⁺), δ¹³C of dissolved inorganic carbon (δ¹³C_DIC), and δ³⁴S of dissolved sulfate (δ³⁴S_SO4) to quantify chemical weathering rates and exchanges of CO₂ between the atmosphere, hydrosphere, and lithosphere. Weathering rates of silicates and carbonates were determined from major element mass balance. Combining the major element mass balance with δ³⁴S_SO4 (−8.9 to 14.1‰_CDT) indicates sulfide oxidation (sulfuric acid production) and subsequent weathering of carbonate and to a lesser degree silicate minerals are important processes in the study area. We determine that on average, 81% of the riverine sulfate can be attributed to sulfide oxidation in the Cordilleran rivers, and that 25% of the total weathering cation flux can be attributed to carbonate and silicate dissolution by sulfuric acid. This result is validated by δ¹³C_DIC values (−9.8 to −3.7‰ _VPDB) which represents a mixture of DIC produced by the following weathering pathways: (i) carbonate dissolution by carbonic acid (−8.25‰) > (ii) silicate dissolution by carbonic acid (−17‰) ≈ (iii) carbonate dissolution by sulfuric acid derived from the oxidation of sulfides (coupled sulfide-carbonate weathering) (+0.5‰).δ³⁴S_SO4 is negatively correlated with δ¹³C_DIC in the Cordilleran rivers, which further supports the hypothesis that sulfuric acid produced by sulfide oxidation is primarily neutralized by carbonates, and that sulfide-carbonate weathering impacts the δ¹³C_DIC of rivers. The negative correlation between δ³⁴S_SO4 and δ¹³C_DIC is not observed in the Ottawa and St. Lawrence River basins. This suggests other factors such as landscape age (governed by tectonic uplift) and bedrock geology are important controls on regional sulfide oxidation rates, and therefore also on the magnitude of sulfide-carbonate weathering—i.e., it is more significant in tectonically active areas.Calculated DIC fluxes due to Ca and Mg silicate weathering by carbonic acid (38.3 × 10³ mol C · km⁻² · yr⁻¹) are similar in magnitude to DIC fluxes due to sulfide-carbonate weathering (18.5 × 10³ mol C · km⁻² · yr⁻¹). While Ca and Mg silicate weathering facilitates a transfer of atmospheric CO₂ to carbonate rocks, sulfide-carbonate weathering can liberate CO₂ from carbonate rocks to the atmosphere when sulfide oxidation exceeds sulfide deposition. This implies that in the Canadian Cordillera, sulfide-carbonate weathering can offset up to 48% of the current CO₂ drawdown by silicate weathering in the region.     

Diffuse gas emissions at the Ukinrek Maars,Alaska: Implications for magmatic degassing and volcanic monitoring

Diffuse CO₂ efflux near the Ukinrek Maars, two small volcanic craters that formed in 1977 in a remote part of the Alaska Peninsula, was investigated using accumulation chamber measurements. High CO₂ efflux, in many places exceeding 1000 g m⁻² d⁻¹, was found in conspicuous zones of plant damage or kill that cover 30,000–50,000 m² in area. Total diffuse CO₂ emission was estimated at 21–44 t d⁻¹. Gas vents 3-km away at The Gas Rocks produce 0.5 t d⁻¹ of CO₂ that probably derives from the Ukinrek Maars basalt based on similar δ¹³C values (∼−6‰), ³He/⁴He ratios (5.9–7.2 R_A), and CO₂/³He ratios (1–2 × 10⁹) in the two areas. A lower ³He/⁴He ratio (2.7 R_A) and much higher CO₂/³He ratio (9 × 10¹⁰) in gas from the nearest arc-front volcanic center (Mount Peulik/Ugashik) provide a useful comparison. The large diffuse CO₂ emission at Ukinrek has important implications for magmatic degassing, subsurface gas transport, and local toxicity hazards. Gas–water–rock interactions play a major role in the location, magnitude and chemistry of the emissions.     

Surface chemistry of disordered mackinawite (FeS)

Disordered mackinawite, FeS, is the first formed iron sulfide in ambient sulfidic environments and has a highly reactive surface. In this study, the solubility and surface chemistry of FeS is described. Its solubility in the neutral pH range can be described by K_s^app = {Fe²⁺} · {H₂S(aq)} · {H⁺}⁻² = 10^+4.87±0.27. Acid-base titrations show that the point of zero charge (PZC) of disordered mackinawite lies at pH ∼7.5. The hydrated disordered mackinawite surface can be best described by strongly acidic mono-coordinated and weakly acidic tricoordinated sulfurs. The mono-coordinated sulfur site determines the acid-base properties at pH < PZC and has a concentration of 1.2 × 10⁻³ mol/g FeS. At higher pH, the tricoordinated sulfur, which has a concentration of 1.2 × 10⁻³ mol/g FeS, determines surface charge changes. Total site density is 4 sites nm⁻². The acid-base titration data are used to develop a surface complexation model for the surface chemistry of FeS.     

Fluorine speciation in topsoils of three active volcanoes of Sicily (Italy)

Fluorine is one of the many environmental harmful elements released by volcanic activity. The content of total oxalate-extractable and water-extractable fluorine was determined in 96 topsoils of three active volcanic systems of southern Italy (Mt Etna, Stromboli and Vulcano). Total fluorine (F) content (F _TOT) ranges from 112 to 7,430 mg kg⁻¹, F extracted with oxalate (F _OX) ranges from 16 to 2,320 mg kg⁻¹ (2–93% of F _TOT) and F extracted with distilled water ( ) ranges from 1.7 to 159 mg kg⁻¹ (0.2–40 % of F _TOT). Fluorine in the sampled topsoils derives both from the weathering of volcanic rocks and ashes and from the enhanced deposition due to volcanic gas emissions either from open-conduit passive degassing (Mt Etna and Stromboli) or from a fumarolic field (Vulcano). Fluorine accumulation in the studied soils does not generally present particular environmental issues except for a few anomalous sites at Vulcano, where measured contents could be dangerous both for vegetation and for grazing animals.     

Pathways and regulation of carbon, sulfur and energy transfer in marine sediments overlying methane gas hydrates on the Opouawe Bank (New Zealand)

This study combines sediment geochemical analysis, in situ benthic lander deployments and numerical modeling to quantify the biogeochemical cycles of carbon and sulfur and the associated rates of Gibbs energy production at a novel methane seep. The benthic ecosystem is dominated by a dense population of tube-building ampharetid polychaetes and conspicuous microbial mats were unusually absent. A 1D numerical reaction-transport model, which allows for the explicit growth of sulfide and methane oxidizing microorganisms, was tuned to the geochemical data using a fluid advection velocity of 14 cm yr⁻¹. The fluids provide a deep source of dissolved hydrogen sulfide and methane to the sediment with fluxes equal to 4.1 and 18.2 mmol m⁻² d⁻¹, respectively. Chemosynthetic biomass production in the subsurface sediment is estimated to be 2.8 mmol m⁻² d⁻¹ of C biomass. However, carbon and oxygen budgets indicate that chemosynthetic organisms living directly above or on the surface sediment have the potential to produce 12.3 mmol m⁻² d⁻¹ of C biomass. This autochthonous carbon source meets the ampharetid respiratory carbon demand of 23.2 mmol m⁻² d⁻¹ to within a factor of 2. By contrast, the contribution of photosynthetically-fixed carbon sources to ampharetid nutrition is minor (3.3 mmol m⁻² d⁻¹ of C). The data strongly suggest that mixing of labile autochthonous microbial detritus below the oxic layer sustains high measured rates of sulfate reduction in the uppermost 2 cm of the sulfidic sediment (100-200 nmol cm⁻³ d⁻¹). Similar rates have been reported in the literature for other seeps, from which we conclude that autochthonous organic matter is an important substrate for sulfate reducing bacteria in these sediment layers. A system-scale energy budget based on the chemosynthetic reaction pathways reveals that up to 8.3 kJ m⁻² d⁻¹ or 96 mW m⁻² of catabolic (Gibbs) energy is dissipated at the seep through oxidation reactions. The microorganisms mediating sulfide oxidation and anaerobic oxidation of methane (AOM) produce 95% and 2% of this energy flux, respectively. The low power output by AOM is due to strong bioenergetic constraints imposed on the reaction rate by the composition of the chemical environment. These constraints provide a high potential for dissolved methane efflux from the sediment (12.0 mmol m⁻² d⁻¹) and indicates a much lower efficiency of (dissolved) methane sequestration by AOM at seeps than considered previously. Nonetheless, AOM is able to consume a third of the ascending methane flux (5.9 mmol m⁻² d⁻¹ of CH₄) with a high efficiency of energy expenditure (35 mmol CH₄ kJ⁻¹). It is further proposed that bioenergetic limitation of AOM provides an explanation for the non-zero sulfate concentrations below the AOM zone observed here and in other active and passive margin sediments.     

Solvation processes in steam: Ab initio calculations of ion-solvent structures and clustering equilibria

Reports of the high ion content of steam and low-density supercritical fluids date back to the work of Carlon [Carlon H. R. (1980) Ion content of air humidified by boiling water.J. Appl.Phys.51, 171-173], who invoked ion and neutral-water clustering as mechanism to explain why ions partition into the low-density aqueous phase. Mass spectrometric, vibrational spectroscopic measurements and quantum chemical calculations have refined this concept by proposing strongly bound ion-solvent aggregates and water clusters such as Eigen- and Zundel-type proton clusters H₃O⁺·(H₂O)_m and the more weakly bound water oligomers (H₂O)_m. The extent to which these clusters affect fluid chemistry is determined by their abundance, however, little is known regarding the stability of such moieties in natural low-density high-temperature fluids. Here we report results from quantum chemical calculations using chemical-accuracy multi-level G3 (Curtiss-Pople) and CBS-Q theory (Peterson) to address this question. In particular, we have investigated the cluster structures and clustering equilibria for the ions and H₃S⁺·(H₂O)_m(H₂S)_n, where m ? 6 and n ? 4, at 300-1000 K and 1 bar as well as under vapor-liquid equilibrium conditions between 300 and 646 K. We find that incremental hydration enthalpies and entropies derived from van’t Hoff analyses for the attachment of H₂O and H₂S onto H₃O⁺, and H₃S⁺ are in excellent agreement with experimental values and that the addition of water to all three ions is energetically more favorable than solvation by H₂S. As clusters grow in size, the energetic trends of cluster hydration begin to reflect those for bulk H₂O liquids, i.e. calculated hydration enthalpies and entropies approach values characteristic of the condensation of bulk water (ΔH^o = −44.0 kJ mol⁻¹, ΔS^o = −118.8 J K mol⁻¹). Water and hydrogen sulfide cluster calculations at higher temperatures indicate that a significant fraction of H₃O⁺, and H₃S⁺ ions exists as solvated moieties.     

Internally consistent thermodynamic data for magnesium sulfate hydrates

Experimental studies on the stability of several Mg-sulfate hydrates including epsomite (MgSO₄·7H₂O), hexahydrite (MgSO₄·6H₂O), starkeyite (MgSO₄·4H₂O), and kieserite (MgSO₄·H₂O) as a function of temperature and relative humidity are in poor agreement with calculations based on thermodynamic properties of these substances taken from the literature. Therefore, we synthesized four different MgSO₄ hydrates and measured their enthalpies of formation by solution calorimetry at T = 298.15 K. The resulting enthalpies of formation from the elements are:

•

Δ_fH⁰₂₉₈ (epsomite) = −3387.7 ± 1.3 kJmol⁻¹

•

Δ_fH⁰₂₉₈ (hexahydrite) = −3088.1 ± 1.1 kJmol⁻¹

•

Δ_fH⁰₂₉₈ (sanderite, MgSO₄·2H₂O) = −1894.9 ± 1.3 kJmol⁻¹

•

Δ_fH⁰₂₉₈ (kieserite) = −1612.4 ± 1.3 kJmol⁻¹

Using mathematical programming (MAP) techniques, standard thermodynamic values consistent both with our calorimetric data and previously published humidity brackets could be derived:

•

Epsomite: Δ_fH⁰₂₉₈ = −3388.7 kJmol⁻¹, S⁰₂₉₈ = 371.3 Jmol⁻¹ K⁻¹, Δ_fG⁰₂₉₈ = −2871.0 kJmol⁻¹

•

Hexahydrite: Δ_fH⁰₂₉₈ = −3087.3 kJmol⁻¹, S⁰₂₉₈ = 348.5 Jmol⁻¹ K⁻¹, Δ_fG⁰₂₉₈ = −2632.3 kJmol⁻¹

•

Starkeyite: Δ_fH⁰₂₉₈ = −2496.1 kJmol⁻¹, S⁰₂₉₈ = 259.9 Jmol⁻¹ K⁻¹, Δ_fG⁰₂₉₈ = −2153.8 kJmol⁻¹

•

Kieserite: Δ_fH⁰₂₉₈ = −1611.5 kJmol⁻¹, S⁰₂₉₈ = 126.0 Jmol⁻¹ K⁻¹, Δ_fG⁰₂₉₈ = −1437.9 kJmol⁻¹

Additionally, heat capacity measurements and standard entropy determinations of several magnesium sulfate hydrate minerals from the literature are analyzed and judged against estimates obtained from a linear combination of the heat capacities of MgSO₄ and hexagonal ice. The results of the MAP analysis are compared to these estimates to conclude that heat capacity and entropy correlate well with the number of waters of hydration. However, even the good correlation is not good enough to capture the fine variations in these properties. Consequently, their experimental measurement is inevitable if reliable thermodynamic data are sought. Our MAP thermodynamic data show that epsomite, hexahydrite, and kieserite have stability fields in the T-%RH space. Starkeyite is metastable. Although no MAP data could have been derived for pentahydrite (MgSO₄·5H₂O) and sanderite, their transient existence suggest that both of them are metastable as well.     

The role of sulfate groups in controlling CaCO3 polymorphism

The nucleation and growth of CaCO₃ phases from aqueous solutions with SO₄²⁻:CO₃²⁻ ratios from 0 to 1.62 and a pH of ∼10.9 were studied experimentally in batch reactors at 25 °C. The mineralogy, morphology and composition of the precipitates were characterized by X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscopy and microanalyses. The solids recovered after short reaction times (5 min to 1 h) consisted of a mixture of calcite and vaterite, with a S content that linearly correlates with the SO₄²⁻:CO₃²⁻ ratio in the aqueous solution. The solvent-mediated transformation of vaterite to calcite subsequently occurred. After 24 h of equilibration, calcite was the only phase present in the precipitate for aqueous solutions with SO₄²⁻:CO₃²⁻ ? 1. For SO₄²⁻:CO₃²⁻ > 1, vaterite persisted as a major phase for a longer time (>250 h for SO₄²⁻:CO₃²⁻ = 1.62). To study the role of sulfate in stabilizing vaterite, we performed a molecular simulation of the substitution of sulfate for carbonate groups into the crystal structure of vaterite, aragonite and calcite. The results obtained show that the incorporation of small amounts (<3 mole%) of sulfate is energetically favorable in the vaterite structure, unfavorable in calcite and very unfavorable in aragonite. The computer modeling provided thermodynamic information, which, combined with kinetic arguments, allowed us to put forward a plausible explanation for the observed crystallization behavior.     

Sulfur speciation in natural hydrothermal waters, Iceland

The speciation of aqueous dissolved sulfur was determined in hydrothermal waters in Iceland. The waters sampled included hot springs, acid-sulfate pools and mud pots, sub-boiling well discharges and two-phase wells. The water temperatures ranged from 4 to 210 °C, the pH_T was between 2.20 and 9.30 at the discharge temperature and the SO₄ and Cl concentrations were 0.020-52.7 and <0.01-10.0 mmol kg⁻¹, respectively. The analyses were carried out on-site within ∼10 min of sampling using ion chromatography (IC) for sulfate (SO₄²⁻), thiosulfate (S₂O₃²⁻) and polythionates (S_xO₆²⁻) and titration and/or colorimetry for total dissolved sulfide (S²⁻). Sulfite (SO₃²⁻) could also be determined in a few cases using IC. Alternatively, for few samples in remote locations the sulfur oxyanions were stabilized on a resin on site following elution and analysis by IC in the laboratory. Dissolved sulfate and with few exceptions also S²⁻ were detected in all samples with concentrations of 0.02-52.7 mmol kg⁻¹ and <1-4100 μmol kg⁻¹, respectively. Thiosulfate was detected in 49 samples of the 73 analyzed with concentrations in the range of <1-394 μmol kg⁻¹ (S-equivalents). Sulfite was detected in few samples with concentrations in the range of <1-3 μmol kg⁻¹. Thiosulfate and SO₃²⁻ were not detected in <100 °C well waters and S₂O₃²⁻ was observed only at low concentrations (<1-8 μmol kg⁻¹) in ∼200 °C well waters. In alkaline and neutral pH hot springs, S₂O₃²⁻ was present in significant concentrations sometimes corresponding to up to 23% of total dissolved sulfur (S_TOT). In steam-heated acid-sulfate waters, S₂O₃²⁻ was not a significant sulfur species. The results demonstrate that S₂O₃²⁻ and SO₃²⁻ do not occur in the deeper parts of <150 °C hydrothermal systems and only in trace concentrations in ∼200-300 °C systems. Upon ascent to the surface and mixing with oxygenated ground and surface waters and/or dissolution of atmospheric O₂, S²⁻ is degassed and oxidized to SO₃²⁻ and S₂O₃²⁻ and eventually to SO₄²⁻ at pH >8. In near-neutral hydrothermal waters the oxidation of S²⁻ and the interaction of S²⁻ and S⁰ resulting in the formation of S_x²⁻ are considered important. At lower pH values the reactions seemed to proceed relatively rapidly to SO₄²⁻ and the sulfur chemistry of acid-sulfate pools was dominated by SO₄²⁻, which corresponded to >99% of S_TOT. The results suggest that the aqueous speciation of sulfur in natural hydrothermal waters is dynamic and both kinetically and source-controlled and cannot be estimated from thermodynamic speciation calculations.     

An attenuated total reflectance IR study of silicic acid adsorbed onto a ferric oxyhydroxide surface

Silicic acid (H₄SiO₄) can have significant effects on the properties of iron oxide surfaces in both natural and engineered aquatic systems. Understanding the reactions of H₄SiO₄ on these surfaces is therefore necessary to describe the aquatic chemistry of iron oxides and the elements that associate with them. This investigation uses attenuated total reflectance infrared spectroscopy (ATR-IR) to study silicic acid in aqueous solution and the products formed when silicic acid adsorbs onto the surface of a ferrihydrite film in 0.01 M NaCl at pH 4. A spectrum of 1.66 mM H₄SiO₄ at pH 4 (0.01 M NaCl) has an asymmetric Si-O stretch at 939 cm⁻¹ and a weak Si-O-H deformation at 1090 cm⁻¹. ATR-IR spectra were measured over time (for up to 7 days) for a ferrihydrite film (≈1 mg) approaching equilibrium with H₄SiO₄ at concentrations between 0.044 and 0.91 mM. Adsorbed H₄SiO₄ had a broad spectral feature between 750 and 1200 cm⁻¹ but the shape of the spectra changed as the amount of H₄SiO₄ adsorbed on the ferrihydrite increased. When the solid phase Si/Fe mole ratio was less than ≈0.01 the ATR-IR spectra had a maximum intensity at 943 cm⁻¹ and the spectral shape suggests that a monomeric silicate species was formed via a bidentate linkage. As the solid phase Si/Fe mole ratio increased to higher values a discrete oligomeric silicate species was formed which had maximum intensity in the ATR-IR spectra at 1001 cm⁻¹. The spectrum of this species suggests that it is larger than a dimer and it was tentatively identified as a cyclic tetramer. A small amount of a polymeric silica phase with a broad spectral feature centered at ≈1110 cm⁻¹ was also observed at high surface coverage. The surface composition was estimated from the relative contribution of each species to the area of the ATR-IR spectra using multivariate curve resolution with alternating least squares. For a ferrihydrite film approaching equilibrium with 0.044, 0.14, 0.40 and 0.91 mM H₄SiO₄ the area of the spectra accounted for by monomeric species were 92%, 49%, 23% and 6%, respectively. The remainder was oligomer apart from a small amount (<5%) of polymerized silica at the two higher H₄SiO₄ concentrations. The solid phase Si/Fe mole ratios for these samples were 0.020, 0.037, 0.071 and 0.138, respectively.     

Anaerobic methane oxidation and a deep H2S sink generate isotopically heavy sulfides in Black Sea sediments

The main terminal processes of organic matter mineralization in anoxic Black Sea sediments underlying the sulfidic water column are sulfate reduction in the upper 2-4 m and methanogenesis below the sulfate zone. The modern marine deposits comprise a ca. 1-m-deep layer of coccolith ooze and underlying sapropel, below which sea water ions penetrate deep down into the limnic Pleistocene deposits from >9000 years BP. Sulfate reduction rates have a subsurface maximum at the SO₄²⁻-CH₄ transition where H₂S reaches maximum concentration. Because of an excess of reactive iron in the deep limnic deposits, most of the methane-derived H₂S is drawn downward to a sulfidization front where it reacts with Fe(III) and with Fe²⁺ diffusing up from below. The H₂S-Fe²⁺ transition is marked by a black band of amorphous iron sulfide above which distinct horizons of greigite and pyrite formation occur. The pore water gradients respond dynamically to environmental changes in the Black Sea with relatively short time constants of ca. 500 yr for SO₄²⁻ and 10 yr for H₂S, whereas the FeS in the black band has taken ca. 3000 yr to accumulate. The dual diffusion interfaces of SO₄²⁻-CH₄ and H₂S-Fe²⁺ cause the trapping of isotopically heavy iron sulfide with δ³⁴S = +15 to +33‰ at the sulfidization front. A diffusion model for sulfur isotopes shows that the SO₄²⁻ diffusing downward into the SO₄²⁻-CH₄ transition has an isotopic composition of +19‰, close to the +23‰ of H₂S diffusing upward. These isotopic compositions are, however, very different from the porewater SO₄²⁻ (+43‰) and H₂S (−15‰) at the same depth. The model explains how methane-driven sulfate reduction combined with a deep H₂S sink leads to isotopically heavy pyrite in a sediment open to diffusion. These results have general implications for the marine sulfur cycle and for the interpretation of sulfur isotopic data in modern sediments and in sedimentary rocks throughout earth’s history.     

Atmospheric sources and sinks of volcanogenic elements in a basaltic volcano (Etna, Italy)

This study reports on the first quantitative assessment of the geochemical cycling of volcanogenic elements, from their atmospheric release to their deposition back to the ground. Etna’s emissions and atmospheric depositions were characterised for more than 2 years, providing data on major and trace element abundance in both volcanic aerosols and bulk depositions. Volcanic aerosols were collected from 2004 to 2007, at the summit vents by conventional filtration techniques. Precipitation was collected, from 2006 to 2007, in five rain gauges, at various altitudes around the summit craters. Analytical results for volcanic aerosols showed that the dominant anions were S, Cl, and F, and that the most abundant metals were K, Ca, Mg, Al, Fe, and Ti (1.5-50 μg m⁻³). Minor and trace element concentrations ranged from about 0.001 to 1 μg m⁻³. From such analysis, we derived an aerosol mass flux ranging from 3000 to 8000 t a⁻¹. Most analysed elements had higher concentrations close to the emission vent, confirming the prevailing volcanic contribution to bulk deposition. Calculated deposition rates were integrated over the whole Etna area, to provide a first estimate of the total deposition fluxes for several major and trace elements. These calculated deposition fluxes ranged from 20 to 80 t a⁻¹ (Al, Fe, Si) to 0.01-0.1 t a⁻¹ (Bi, Cs, Sc, Th, Tl, and U). Comparison between volcanic emissions and atmospheric deposition showed that the amount of trace elements scavenged from the plume in the surrounding of the volcano ranged from 0.1% to 1% for volatile elements such as As, Bi, Cd, Cs, Cu, Tl, and from 1% to 5% for refractory elements such as Al, Ba, Co, Fe, Ti, Th, U, and V. Consequently, more than 90% of volcanogenic trace elements were dispersed further away, and may cause a regional scale impact. Such a large difference between deposition and emission fluxes at Mt. Etna pointed to relatively high stability and long residence time of aerosols in the plume.     

Particulate sulfur species in the water column of the Cariaco Basin

The biogeochemistry of iron sulfide minerals in the water column of the Cariaco Basin was investigated in November 2007 (non-upwelling season) and May 2008 (upwelling season) as part of the on-going CARIACO (CArbon Retention In A Colored Ocean) time series project. The concentrations of particulate sulfur species, specifically acid volatile sulfur (AVS), greigite, pyrite, and particulate elemental sulfur, were determined at high resolution near the O₂/H₂S interface. In November 2007, AVS was low throughout the water column, with the highest concentration at the depth where sulfide was first detected (260 m) and with a second peak at 500 m. Greigite, pyrite, and particulate elemental sulfur showed distinct concentration maxima near the interface. In May 2008, AVS was not detected in the water column. Maxima for greigite, pyrite, and particulate elemental sulfur were again observed near the interface. We also studied the iron sulfide flux using sediment trap materials collected at the Cariaco station. Pyrite comprised 0.2-0.4% of the total particulate flux in the anoxic water column, with a flux of 0.5-1.6 mg S m⁻² d⁻¹.Consistent with the water column concentration profiles for iron sulfide minerals, the sulfur isotope composition of particulate sulfur found in deep anoxic traps was similar to that of dissolved sulfide near the O₂/H₂S interface. We conclude that pyrite is formed mainly within the redoxcline where sulfur cycling imparts a distinct isotopic signature compared to dissolved sulfide in the deep anoxic water. This conclusion is consistent with our previous study of sulfur species and chemoautotrophic production, which suggests that reaction of sulfide with reactive iron is an important pathway for sulfide oxidation and sulfur intermediate formation near the interface. Pyrite and elemental sulfur distributions favor a pathway of pyrite formation via the reaction of FeS with polysulfides or particulate elemental sulfur near the interface. A comparison of thermodynamic predictions with actual concentration profiles for iron sulfides leads us to argue that microbes may mediate this precipitation.     

Stream geochemistry, chemical weathering and CO2 consumption potential of andesitic terrains, Dominica, Lesser Antilles

Recent studies of chemical weathering of andesitic-dacitic material on high-standing islands (HSIs) have shown these terrains have some of the highest observed rates of chemical weathering and associated CO₂ consumption yet reported. However, the paucity of stream gauge data in many of these terrains has limited determination of chemical weathering product fluxes. In July 2006 and March 2008, stream water samples were collected and manual stream gauging was performed in watersheds throughout the volcanic island of Dominica in the Lesser Antilles. Distinct wet and dry season solute concentrations reveal the importance of seasonal variations on the weathering signal. A cluster analysis of the stream geochemical data shows the importance of parent material age on the overall delivery of solutes. Observed Ca:Na, HCO₃:Na and Mg:Na ratios suggest crystallinity of the parent material may also play an important role in determining weathering fluxes. From total dissolved solids concentrations and mean annual discharge calculations we calculate chemical weathering yields of (6-106 t km⁻² a⁻¹), which are similar to those previously determined for basalt terrains. Silicate fluxes (3.1-55.4 t km⁻² a⁻¹) and associated CO₂ consumption (190-1575 × 10³ mol km⁻² a⁻¹) determined from our study are among the highest determined to date. The calculated chemical fluxes from our study confirm the weathering potential of andesitic-dacitic terrains and that additional studies of these terrains are warranted.     

Geochemical and mineralogical impacts of H2SO4 on clays between pH 5.0 and −3.0

Natural and constructed clay liners are routinely used to contain waste and wastewater. The impact of acidic solutions on the geochemistry and mineralogy of clays has been widely investigated in relation to acid mine drainage systems at pH > 1.0. The impact of H₂SO₄ leachate characterized by pH < 1.0 and potentially negative pH values on the geochemistry and mineralogy of clays is, however, not clear. Thus, laboratory batch experiments were conducted on three natural clay samples with different mass ratios of smectite, illite and kaolinite to investigate the impact of H₂SO₄ on the geochemistry and mineralogy of aluminosilicates from pH 5.0 to −3.0. Batch testing was conducted at seven pH treatments (5.0, 3.0, 1.0, 0.0, −1.0, −2.0 and −3.0) using standardized H₂SO₄ solutions for four exposure periods (14, 90, 180, and 365 d). Aqueous geochemical and XRD analyses showed: increased dissolution of aluminosilicates with decreasing pH and increasing exposure period, that smectite was more susceptible to dissolution than illite and kaolinite, precipitation of an amorphous silica phase occurred at pH ? 0.0, and anhydrite precipitated in Ca-rich clays at pH ? −1.0. In addition, global dissolution rates were calculated for the clays and showed good agreement to literature smectite, illite and kaolinite dissolution rates, which suggests global dissolution rates for complex clays could be determined from monomineralic studies. A stepwise conceptual model of the impact of H₂SO₄ on aluminosilicate geochemistry and mineralogy between pH 5.0 and −3.0 is proposed.     

Interaction between sulphide and H2O in silicate melts

Reaction between dissolved water and sulphide was experimentally investigated in soda-lime-silicate (NCS) and sodium trisilicate (NS3) melts at temperatures from 1000 to 1200 °C and pressures of 100 or 200 MPa in internally heated gas pressure vessels. Diffusion couple experiments were conducted at water-undersaturated conditions with one half of the couple being doped with sulphide (added as FeS or Na₂S; 1500-2000 ppm S by weight) and the other with H₂O (∼3.0 wt.%). Additionally, two experiments were performed using a dry NCS glass cylinder and a free H₂O fluid. Here, the melt was water-saturated at least at the melt/fluid interface. Profiling by electron microprobe (sulphur) and infrared microscopy (H₂O) demonstrate that H₂O diffusion in the melts is faster by 1.5-2.3 orders of magnitude than sulphur diffusion and, hence, H₂O can be considered as a rapidly diffusing oxidant while sulphur is quasi immobile in these experiments.In Raman spectra a band at 2576 cm⁻¹ appears in the sulphide - H₂O transition zone which is attributed to fundamental S-H stretching vibrations. Formation of new IR absorption bands at 5025 cm⁻¹ (on expense of the combination band of molecular H₂O at 5225 cm⁻¹) and at 3400 cm⁻¹ was observed at the front of the in-diffusing water in the sulphide bearing melt. The appearance and intensity of these two IR bands is correlated with systematic changes in S K-edge XANES spectra. A pre-edge excitation at 2466.5 eV grows with increasing H₂O concentration while the sulphide peak at 2474.0 eV decreases in intensity relative to the peak at 2477.0 eV and the feature at 2472.3 eV becomes more pronounced (all energies are relative to the sulphate excitation, calibrated to 2482.5 eV). The observations by Raman, IR and XANES spectroscopy indicate a well coordinated S²⁻ - H₂O complex which was probably formed in the glasses during cooling at the glass transition. No oxidation of sulphide was observed in any of the diffusion couple experiments. On the contrary, XANES spectra from experiments conducted with a free H₂O fluid show complete transformation of sulphide to sulphate near the melt surface and coexistence of sulphate and sulphide in the center of the melt. This can be explained by a lower H₂O activity in the diffusion couple experiments or by the need of a sink for hydrogen (e.g., a fluid which can dissolve high concentration of hydrogen) to promote oxidation of sulphide by H₂O via the reaction S²⁻ + 4H₂O = SO₄²⁻ + 4H₂. Sulphite could not be detected in any of the XANES spectra implying that this species, if it exists in the melt, it is a subordinate or transient species only.     

Landscape variations in stream water SO4 and δSSO4 in a boreal stream network

Despite reduced anthropogenic deposition during the last decades, deposition sulphate may still play an important role in the biogeochemical cycles of S and many catchments may act as net sources of S that may remain for several decades. The aim of this study is to elucidate the temporal and spatial dynamics of both SO₄²⁻ and δ³⁴S_SO4 in stream water from catchments with varying percentage of wetland and forest coverage and to determine their relative importance for catchment losses of S. Stream water samples were collected from 15 subcatchments ranging in size from 3 to 6780 ha, in a boreal stream network, northern Sweden. In forested catchments (<2% wetland cover) S-SO₄²⁻ concentrations in stream water averaged 1.7 mg L⁻¹ whereas in wetland dominated catchments (>30% wetland cover) the concentrations averaged 0.3 mg L⁻¹. A significant negative relationship was observed between S-SO₄²⁻ and percentage wetland coverage (r² = 0.77, p < 0.001) and the annual export of stream water SO₄²⁻ and wetland coverage (r² = 0.76, p < 0.001). The percentage forest coverage was on the other hand positively related to stream water SO₄²⁻ concentrations and the annual export of stream water SO₄²⁻ (r² = 0.77 and r² = 0.79, respectively). The annual average δ³⁴S_SO4 value in wetland dominated streams was +7.6‰ and in streams of forested catchments +6.7‰. At spring flood the δ³⁴S_SO4 values decreased in all streams by 1‰ to 5‰. The δ³⁴S_SO4 values in all streams were higher than the δ³⁴S_SO4 value of +4.7‰ in precipitation (snow). The export of S ranged from 0.5 kg S ha⁻¹ yr⁻¹ (wetland headwater stream) to 3.8 kg S ha⁻¹ yr⁻¹ (forested headwater stream). With an average S deposition in open field of 1.3 kg S ha⁻¹ yr⁻¹ (2002-2006) the mass balance results in a net export of S from all catchments, except in catchments with >30% wetland. The high temporal and spatial resolution of this study demonstrates that the reducing environments of wetlands play a key role for the biogeochemistry of S in boreal landscapes and are net sinks of S. Forested areas, on the other hand were net sources of S.