Geochimica et Cosmochimica Acta, Vol. 67, No. 23, pp.

4457– 4469, 2003
Copyright © 2003 Elsevier Ltd
Pergamon Printed in the USA. All rights reserved
0016-7037/03 $30.00 ⫹ .00
doi:10.1016/S0016-7037(03)00388-0

Kinetics and mechanism of polythionate oxidation to sulfate at low pH by O2 and Fe3ⴙ
GREGORY K. DRUSCHEL,1,* ROBERT J. HAMERS,2 and JILLIAN F. BANFIELD1,2†
1
Departments of Geology and Geophysics and
2
Chemistry, University of Wisconsin-Madison, Madison, WI 53706 USA

(Received October 16, 2002; accepted in revised form May 30, 2003)

Abstract—Polythionates (SxO62⫺ ) are important in redox transformations involving many sulfur compounds.
Here we investigate the oxidation kinetics and mechanisms of trithionate and tetrathionate oxidation between
pH 0.4 and pH 2 in the presence of Fe3⫹ and/or oxygen. In these solutions, Fe3⫹ plus oxygen oxidizes
tetrathionate and trithionate at least an order of magnitude faster than oxygen alone. Kinetic measurements,
coupled with density functional calculations, suggest that the rate-limiting step for tetrathionate oxidation
involves Fe3⫹ attachment, followed by electron density shifts that result in formation of a sulfite radical and
S3O30 derivatives. The overall reaction orders for trithionate and tetrathionate are fractional due to rearrange-
ment reactions and side reactions between reactants and intermediate products. The pseudo-first order rate
coefficients for tetrathionate range from 10⫺11 s⫺1 at 25°C to 10⫺8 s⫺1 at 70°C, compared to 2 ⫻ 10⫺7 s⫺1
at 35 °C for trithionate. The apparent activation energy (EA) for tetrathionate oxidation at pH 1.5 is 104.5 ⫾
4.13 kJ/mol. A rate law at pH 1.5 and 70°C between 0.5 and 5 millimolar [Fe3⫹] is of the form:
r⫽10⫺6.61⫾0.3[S4O62⫺]0.3⫾0.08[Fe3⫹]0.15⫾0.09
where rate, r, is in units of mol L⫺1 s⫺1. In pH ⬍1 solutions where trithionate forms by rearrangement of
tetrathionate, trithionate oxidation is the rapid pathway for conversion of tetrathionate to sulfate. The kinetics
of both trithionate and tetrathionate oxidation are several orders of magnitude slower than the formation of
polythionates from thiosulfate in acidic, Fe3⫹-rich solutions. Copyright © 2003 Elsevier Ltd

1. INTRODUCTION ates vary in chain length greatly, and contain up to 80 sulfurs
2⫺ in sulfur sols (Lyons and Nickless, 1968), up to 45 in synthetic
Polythionates (SxO ), also known as polysulfane disulfonic
6
compounds (Ancutiene and Janickis, 2000), and up to 22 sul-
acids or polysulfane disulfonates, are sulfur-based molecules in
furs in bacterial sulfur excretions (Steudel et al., 1987).
which a sulfur chain is terminated at both ends by SO3⫺ groups.
Complex mixtures of polythionates are formed by the reac-
These compounds are important intermediate species in the
tion of H2S and SO2, sometimes referred to as Wackenroeders
redox transformations of sulfur compounds in many environ-
solutions (Lyons and Nickless, 1968). In these solutions, poly-
ments and in the metabolism of sulfur-oxidizing and sulfur-
thionate chain length is generally governed by the equilibrium:
reducing microorganisms (Chambers and Trudinger, 1979;
Steudel et al., 1987; Kelly, 1999; Schippers et al., 1999; Su- SxO62⫺ ⫹ S2O32⫺ 7 Sx ⫹ 1O62⫺ ⫹ SO32⫺ (1)
zuki, 1999; Takano et al., 2000; Xu et al., 2000).
Figure 1 represents an energy-minimized geometry for the Chain lengthening reaction is favored at pH ⬍7, the pK2 of
trithionate molecule, S3O2⫺ 2⫺
6 , tetrathionate molecule, S4O6 ,
sulfurous acid (H2SO3, Moses et al., 1987). Although theoret-
and pentathionate molecule, S5O2⫺ 6 . The calculated charges on ically polythionates can form very long chains, most are ob-
each of the sulfur and oxygen atoms are also represented in served to have x ⫽ 3 to 6. The relative abundances of tetra-
Figure 1, illustrating that the central sulfurs are more reduced thionate, trithionate, and pentathionate are governed by
than the terminal sulfurs. The central S-S bond is shorter than rearrangement reactions such as:
the two symmetrical S-S bonds involving the terminal sulfurs
bonded to the oxygens. The bond lengths compare well with 2 S4O62⫺ 7 S3O62⫺ ⫹ S5O62⫺ (2)
those determined by X-ray diffraction studies of polythionate (Lyons and Nickless, 1968). Transformation of polythionate
salts (Christidis et al., 1989). Williamson and Rimstidt (1992) species can be either through redox reactions or through bimo-
predicted the pK2 of H2S4O6 to be ⫺2.3, indicating that these lecular nucleophilic displacement (SN2, Moses et al., 1987).
species are generally unprotonated at all but the most extreme Additionally, polythionates undergo a hydrolysis reaction:
pH values (Nordstrom et al., 2000).
As shown in the Eh-pH diagram (Fig. 2), polythionates have SxO62⫺ ⫹ H2O 3 Sx ⫺ 1O32⫺ ⫹ SO42⫺ ⫹ 2 H⫹ (3)
predominance fields intermediate between those of sulfide and
sulfite (diagram produced by suppressing sulfate). Polythion- SxO62⫺ ⫹ y H2O 3 (0.58x) S8 ⫹ (0.42x) SO42⫺ ⫹ (0.5 H⫹y)
(4)
* Author to whom correspondence should be addressed, at College of (Meyer and Ospina, 1982). Some of the proposed products of
Marine Studies, University of Delaware, Lewes, DE 19958, USA
(druschel@udel.edu). polythionate reactions include the polysulfane monosulfonic

Present address: Department of Earth and Planetary Sciences, Uni- acids (Steudel et al., 1987). These have the general formula
versity of California, Berkeley, CA 94720, USA. SxO32⫺ and consist of chains of sulfur atoms terminated at one
4457

Schippers and Sand.1. 1996. Williamson and Rimstidt. Kinetics Experiments Recent advances in the sampling and analysis (Druschel et al. these authors did not attempt to address either the mechanism or the kinetics of 2. 1. The oxidation of metal sulfide minerals generates acidic. Elberling et al.. MATERIALS AND METHODS oxidation of this species. The reactivity of polythionates at low pH is thus an in the lower polythionate species tri-.. and J... dithionate. has been proposed to be a leaving group in the oxidation of pyrite. 1987). as important part of understanding the governing pathways of calculated with gaussian98 employing B3LYP density functional the. 1982). 1993. in Kelly and Wood (1994) utilizing the reaction of thiosulfate with lysis.. 2001) of polythionates The kinetic investigation of polythionate oxidation was primarily have facilitated their direct identification and study in many carried out as batch reactions. utilizing ion chromatographic and spec- environments. Polythionates have been detected within sulfur globules hydrogen peroxide. Luther.. J. based on experimental results and molecular orbital theory (Goldhaber. Moses et al. SxO32⫺ ⫹ S2O32⫺ 3 Sx ⫹ 1O32⫺ ⫹ SO32⫺ (5) It is the goal of this paper to investigate the oxidation of trithionate and tetrathionate to sulfate with Fe3⫹ in acidic Similar reactions resulting in terminal sulfur oxidation form solutions. 1994. Thiosulfate. 1987. and pentathionate. 1983. 1983. 1994) and as part of the structure Solutions of sulfide. pyrite oxidation under different conditions. This difference in reaction prod- ucts must be due to either changes in the reactivity of those products in differing environments or to changes in reaction mechanism(s) and subsequent changes in pyrite oxidation path- Fig. Observations of pyrite oxida- tion in experimental and natural settings have shown more thiosulfate and polythionate at circumneutral pH than at low pH (Goldhaber.4 and 2. K. 1987. thiosulfate. Hamers.9% thiosulfate oxidation in the presence of pyrite and polythionate purity). Moses et al. 2002). 1987). 2001). Druschel. Schippers et al. known preliminary assessment of the stability of tetrathionate iron is expected to impact trithionate and tetrathionate reactiv- in acidic solutions containing Fe3⫹. Brandt and Eldik. Luther. Subsequent oxidation of thiosulfate to either polythionates or elemental sulfur and sulfite has been determined experimentally by several groups (Johnston and McAmish. Steudel et al. 2. Several researchers ory using an LANL2DZ basis set. tetra-. Polysulfane monosulfonic acids are also tetrathionate oxidation via hydrolysis reactions similar to reac- susceptible to nucleophilic displacement reactions resulting in tion 3. We focus on their reactivity with Fe3⫹ because this polythionate species and polysulfide (Sn2⫺): species is abundant and important in many low pH environ- SxO32⫺ ⫹ S2O32⫺ 3 SxO62⫺ ⫹ S22⫺ (6) ments. In a thermal spring in Yellowstone National troscopic determination of intermediate species. S2O32⫺. and sulfite were prepared of microbially formed sulfur globules (Steudel et al.. (2000) reported polythionate formation via Potassium tetrathionate salts were obtained from Sigma (⬎99. Transition metals including iron have been shown to influence the mechanism and kinetics of hydrogen sulfide. At many locations worldwide. Millero.. 1973. Aside from noting suffi. (Meyer and Ospina. 1983. Purity of the trithionate salt was determined by gravimetry and chromatographic analyses to be greater than 99%. see O’Reilly et al. mining of pyrite (FeS2)-rich deposits has increased oxidative dissolution and led to formation of environmentally damaging solutions known as acid mine drainage (AMD).. 2003. from appropriate salts under a nitrogen atmosphere immediately before Changes in polythionate concentrations through reaction with use as chromatographic standards. 1998. 1982. 1987. and sodium trithionate salts were synthesized after the method decomposition either by reaction with H2S or through hydro. cient stability in cold acidic solutions. Banfield aqueous SO2 and H2S discharged through vents may serve as an indicator of imminent volcanic eruptions (Takano. Figure 3 illustrates a general picture of some of the major pathways through which pyritic sulfur may oxidize in the environment (based on Gold- haber. F. 1998. However no attempts have yet been made towards chain lengthening/shortening: defining reactivity of polythionates with Fe3⫹. and polysulfide reactions (Williamson and The work of Takano and Watanuki (1988) includes the only Rimstidt. for review. 1999. have characterized polythionate reactions in the absence of iron (Meyer and Ospina. based on how the pyritic sulfide dissociates from the pyrite to either oxygenated or nonoxygenated forms. Thus.4458 G. 1987. 1987).. metal rich fluids. Xu and Schoonen. and sulfate chromatographic stan- . ity. and Schippers et al. 2000). Xu et al. Park. All experiments were carried out at pH between 0. Nesbitt et al. Geometry optimized structure and natural charges of S and O ways. R. in a crater lake (Takano et al. Hu. 1987). (1996) suggested pyrite oxidation pathways including end by an SO3⫺ group. 1995).. Moses et al. sulfite. thiosulfate.

The vessel. significant change of the reactant products for the tetrathionate exper- der. Two mL chromatographic techniques described below. [Fe ]. It was undersaturated with respect to ferric oxyhydroxides. with 1. was allowed to poured into the bottles. The sample was mL syringe. dards were prepared from NIST-certified stock solution (FisherBrand). Elemental sulfur found that storage of the solutions over several days did not result in HPLC standards were prepared from EM Science sublimated S8 pow. Temperature profiles were performed to show that thermal then directly injected and analyzed at different time intervals by the equilibration of the solutions occurs in a matter of minutes. Experimental sets inves- Tetrathionate formation from thiosulfate oxidation was studied by tigating [S4O2⫺ 3⫹ 6 ]initial and [Fe ] dependence were corrected to equal spiking 1 mM S2O2⫺ 3 in 50 mL of 0. The solution was pumped directly from the The bottles were stoppered. a control FeCl2 in a 125 mL glass septum bottle sealed with a rubber stopper to was run without added Fe3⫹ to assess the contribution of thermal minimize any loss of gaseous sulfur species possibly evolved in the degradation and reaction with O2 at the same pH and temperature reaction progress. Several obtained from Sigma (ACS reagent grade) and stored in 0. The vessel was left open aliquots were withdrawn in time. 90% Fe3⫹. but it was found that the trithionate was much more reactive at Batch reactions for tetrathionate oxidation experiments were carried lower temperature and therefore quenching and storage of collected out in 125 mL glass septum bottles and sealed with rubber stoppers to aliquots was not possible. capped. and reaction vessel with a single piston HPLC pump (Dionex DX10) sampled periodically through the rubber stopper with a needle and 3 directly into the sample loop of the ion chromatograph.5 mol/L HCl samples were analyzed by IC immediately to check the efficiency of the to minimize Fe2⫹ oxidation and maintain Fe3⫹ solutions at conditions storage and low-temperature quenching of reaction progress. and stored in a 4°C refrigerator for sulfur Fe3⫹ and Fe2⫹ stock solutions were prepared from chloride salts species analysis.5 mL used for immediate Fe2⫹ analysis. and 0. Polythionate oxidation kinetics and mechanism 4459 Fig. 2000) and using the LLNL v6 r8 database. placed in water baths at temperature. Duplicate experiments were conducted with 90% conditions. diluted with 18 M⍀ nanopure H2O. 3-port glass reaction vessel was minimize any loss of gaseous sulfur species possibly evolved in the utilized for the trithionate experiments. were prepared from normalized HCl (FisherBrand certified 1 N HCl ionic strength. pH. containing 500 mL reaction progress. leaving oxygenated headspace with 10 –100 equilibrate at temperature and a 1 mL spike of trithionate stock was times as much oxygen as could possibly be utilized by any reactions. All stock solutions were filter sterilized. added to begin the reaction. Fifty mL experimental solutions were prepared and of solution at the desired pH and Fe3⫹ concentration. Eppendorf tubes.5 mL to air through an open Teflon tube used to sample the solution for Fe2⫹ . 2. Fe2⫹ 10% Fe3⫹ and 10% Fe2⫹.5 mL transferred to 1. 99% minimum assay dissolved in 100% ACS grade perchloroeth. with all experimental conditions listed. activity ⌺S ⫽ 1 ⫻ 10⫺3 calculated with the Geochemist’s Workbench (Bethke. Experimental solutions Multiple sets of experiments with different [S4O2⫺ 3⫹ 6 ]initial. and T were performed. each. manner. and spiked with the imental sets reported in this work and the principle variable studied in appropriate stock solutions when in the reaction vessel. For every experiment.05 mol/L HCl solution with 1 mM ionic strength by the addition of KCl. Samples were withdrawn after several minutes and after several hours to identify the products of Trithionate experiments were originally carried out in the same this reaction. Eh-pH diagram of intermediate sulfur species at 25°C. ylene. iments. A 1-L. Table 1 lists each of the exper- solution).

2000) with the gaussian98 output for each pyrite oxidation. tetra-.1 HCl at detected using an isocratic 30% acetonitrile:H20 eluent containing 2 25°C 500 ␮M S4O2⫺ 6 in 0..) was used to assemble polythionates. 0. Generalized diagram of some of the ideas possibly governing (Schaftenaar and Noordik. 1000. thiosulfate. Moses et al.1 (Bethke. or Cu⫹ at 70°C open to analyzed using a ferrozine method on a Hach model 2010 portable oxygenated headspace datalogging spectrophotometer. 2003).25. for reference to these conditions. All peaks except reaction HCl at 25°C 500. Table 1. Multiple linear regression of data sets for determination of an overall rate law was accomplished with the MINITAB Statistical Software package Minitab 13. and ionic strength. Nesbitt et al. log r headspace with 0. Cambridge- Soft Scientific Computing. buffered with 3 mM NaHCO3/3 at 4.7 Fig. except for a duplicate experiment that was constantly purged with argon. 1983. 1998. and were calculated with the first 3– 6 data VI Ionic strength 500 ␮M S4O2⫺ 6 in 0. III Fe2⫹.4460 G. Inc. 2002).0 I Rearrangement 500 ␮M S4O2⫺ 6 in 0. and the geom- etry was initially minimized using the molecular mechanics algorithm MM2. initial tetrathionate.10 phenanthroline. 2000. and Fe3⫹ complexes.05 M HCl and 0. molecule.1 M HCl mM tetrabutylammonium hydroxide. The Series designation is used throughout the text Chromatographic analysis of collected aliquots for tetrathionate ex. Using log r ⫽ log k⫹ n log C. 1000. electronic structure. (1993). Deviations beyond standard analytical error (which is between 2 and 5% for chromatographic runs) are likely to be the result of complexity in the reaction pathway (Brezonik. listing pertinent experimental conditions and the reaction variable 2. Local charge densities were calculated using Natural Bond Orbital 5. 25. periments was performed on a Dionex Series 500 ion chromatograph equipped with a conductivity detector and UV-Vis detector collected at Series Reaction variable Conditions 230 nm. Models were visualized using Molden3. 1993). 1987. The method was oxygenated headspace. Hamers. IV [S4O2⫺ 6 ] initial 120. no Fe3⫹ modified from Zou et al. an Alltech C18 column. Tri-. and an isocratic 95% oxygenated headspace or Ar or N2⫺ methanol:H2O eluent at 1.0 (Glen- dening et al. Sulfate. K. polysulfane monosulfonic acids. II O2 500 ␮M S4O2⫺ 6 in 0. 2000).0 mL/min after the method of McGuire and purged Hamers (2000). or 70°C open to stidt and Newcomb. Apparent activation energy and collision frequency were calculated using the Arrhenius relationship.. Schippers et al. 1. Luther. and sulfide were de- tectable using an isocratic 5mM NaHCO3/5 mM Na2CO3 eluent at a 1.3. F.1.05.5. Sample Characterization investigated in each. Multiple linear regression of several data sets at once is possible mathematically. or 2000 ␮M S4O2⫺ 6 in 0. Calculations Understanding of the geometry. R. 1.5.01. J. dithionate. 250. but were as high as 40% in some experiments. 1993).. or 10 mM Fe3⫹ at 70°C open to (Brezonik. List of tetrathionate experiments reported in this work. (after Goldhaber. Banfield tions were held constant throughout. Hu.05 M HCl and 1 Fe3⫹. oxygenated headspace .... respectively (after Rim.0.05 M HCl and 2 mM Fe3⫹ at 70°C open to oxygenated headspace 2. sulfite. 2001) on the computational cluster at the University of Wisconsin Department of Chemistry. 42. An Eh-pH diagram showing predominance fields of intermediate sulfur species was calculated using the Geochemist’s Workbench 3. All species were calculated in a spherical continuum with a dielectric constant of 80 (equivalent to water at 25°C) employing B3LYP density functional theory and the LANL2DZ basis set.5 M HCl and 1 mM points.4. 1987. Rate coefficients were calculated using a method of initial velocities 5. 2.32 (Minitab Inc. Chem3D (Version 5. 70°C all open to mM Na2CO3 and run through an IonPac NS1 column. Zn2⫹.0. Polythionates were 5000 ␮M S4O2⫺ 6 in 0.05 M HCl and 2 mM Elemental sulfur was analyzed by HPLC using a Shimadzu system Fe3⫹ at 25 and 70°C open to UV-Vis at 254 nm. log C was plotted to determine rate coefficients (k’s) and reaction KCl added orders (values of n) with respect to the changing concentrations of VII Temperature 1000 ␮M S4O2⫺ 6 in 0. Druschel. 0.1 M mL/min flow rate through an IonPac AS16 column. 2001). and J.05 M HCl and 2 mM Fe2⫹ was analyzed using 1. Errors were typically on the order of 5–15%. and electron density shifts as a result of Fe3⫹ interaction was accomplished with several computational techniques. 3. For all experiments. Zn2⫹.2. 0. 500. 25. pH and O2 concentra. mM Fe3⫹ at 4. 1993). or 2 M vs. Kinetic Analysis V [Fe3⫹] initial 500 ␮M S4O2⫺ 6 in 0. Cu⫹ 500 ␮M S4O2⫺ 6 in 0. suppressing sulfate as a species. and total iron was Fe3⫹. Reported rates are rates of tetrathionate degradation oxygenated headspace unless otherwise noted. but graphical representation with log-log plots would be in multidimensional space and is therefore not represented visually. The Lawrence Liver- analysis. sulfate were detectable with the UV-Vis detector. 60. 0. The z-matrix of the resulting molecule was output into a gaus- sian input file and run using gaussian98 (Frisch et al. and pentathionate were all present resolvable down to 5 ␮M concentrations using a 25 ␮L sample loop. Fitting errors were calculated by determining the standard error of the mean from the regression fit using SigmaPlot 4. depending on individual data set and best-fit statistics and is Fe3⫹ at 70°C open to oxygenated reported in units of mol L⫺1 s⫺1. 1996.

0 0:100:0 Fig. RESULTS as minor impurities in the tetrathionate salt.3.27 13:74:13 more National Labs V8 R6⫹ “combined” thermodynamic data set was 0.05 500 25 0.47 24:52:24 utilized the Chemical Kinetics Simulator v 1. from a starting solution containing ⬃5% trithionate and pen- tathionate and 90% tetrathionate.ibm.50 25:50:25 research at: www. 2.45 22:56:22 0.01 (available from IBM 0.1 500 4 0. Table 2 lists the results of Series I experiments.1 500 25 0. In samples analyzed imme- tion is consistent regardless of the presence of Fe2⫹ in solution diately.0 0:100:0 oxidation (or reduction) of the sulfur. 0.1 500 60 0. Tetrathionate produc. 4.45 23:54:23 0. S2O62 . and S5O6 2⫺ would be identified by the tech- 3. and pentathionate after ⬃80 min. mM Fe3⫹ 70°C pH 1. Mass balance calculations indicate that all of able elemental sulfur was produced.5 as long as excess Fe3⫹ is present (data not shown). The rearrangement reaction does not appreciably occur within several days at pH ⬎⬃ 1. The only species Reaction products of thiosulfate in acidic. verted to sulfate. and did not change significantly in the following 24 h.0 equilibrated to a 15:70:15 mix trithionate.1. 0. 1).5 (adjusted with HCl) open to O2. Tetrathionate Formation at Low pH in the Presence niques employed in the tetrathionate batch experiments if they of Excess Fe3ⴙ persist in solution for even a few minutes.almaden.01 500 25 0.1 500 25 0. (2000) noted a solution of polythionates at pH ⫽ 1.39 19:62:19 0. tions consisted of primarily tetrathionate.47 24:52:24 Kinetic simulations of reactions of thiosulfate in acidic solutions 0. Tetrathionate Oxidation by Fe3ⴙ and O2 designed to follow the production of tetrathionate from thio- sulfate and determine its fate via both rearrangement reactions A number of possible reactants and principle variables have and as the result of oxidation by Fe3⫹ and O2 to sulfate. The extent of rearrangement does change as a function of [S4O2⫺6 ]initial.2.1 2000 25 0.1 500 42 0. . S4O2⫺ 2⫺ 6 . H2SO3(aq). Fe3⫹ rich solu.1 1000 25 0. S3O2⫺ 2⫺ 6 . Based on this. Tetrathionate Rearrangement to Trithionate and Results of experiments executed under nitrogen or argon Pentathionate Investigation of the nonoxidative reactivity of tetrathionate species under the relevant pH and temperature conditions was necessary before examination of the oxidation reactions. S5O6 . Polythionate oxidation kinetics and mechanism 4461 Table 2. Results of tetrathionate rearrangement reactions under dif- ferent conditions expressed as the ratio of trithionate to tetrathionate and the ratio of trithionate to tetrathionate to pentathionate. but this is rapidly con- PCE extraction and analysis by HPLC showed that no measur. HS2O3⫺ . Intermediate species such as H2S.com/st/msim/).05 500 70 0. sulfite) were occasionally observed. where the ratio of trithionate to tetrathionate is expressed as the measure of reac- tion progress according to Eqn.0 (0. and SO4 2⫺ (Fig. An been tested for their role in the overall oxidation kinetics of outline for the experiments described below is shown in Figure tetrathionate to sulfate in the presence of Fe3⫹ at low pH (Table 4.0 0:100:0 redox reactions refer to reactions studied where there was no overall 0. The following section describes a series of experiments 3. and is minimally affected by temperature between 25 and 70 °C. samples were measured after approximately 90 min of reaction. the sulfur in these experiments is accounted for by these spe- cies.45 22:56:22 utilized for all calculations. No 0. tween polythionates and thiosulfate or sulfite. 5.1 mol/L HCl). 3. Tetrathionate oxidation experiment with 500 ␮M S4O2⫺ 6 5 rearrangement on [S4O2⫺ 6 ]initial could be due to reaction be. very small concentrations of H2SO3(aq) (detected as at pH 1. Basic structure of experiments performed in this study. S4O6 . Analysis of sulfate as a principle reaction progress vari- able is therefore appropriate under these conditions. which may occur 3. 5). The dependence of Fig. Rate laws for individual reactions were input and the starting concentrations of reactants set to approximate experimental conditions. typically identified by chromatographic analyses were S3O2⫺ 6 .1 5000 25 0. This result indicates rearrangement is not the result of a direct bimolecular reaction. S4O2⫺ 6 initial HCl concentration (M) (␮M) T (°C) S3O2⫺ 2⫺ 6 /S4O6 S3O2⫺ 2⫺ 2⫺ 6 :S4O6 :S5O6 0. Xu et al. tetrathionate.

F. 9) of log initial rates vs. very slow at pH⬍2 (Singer and Stumm. Simi. or elemental sulfur were detected).00 ⫻ 10⫺8 1. R.73 ⫻ 10⫺8 1. we carried out ki- netic simulations using the rates for thiosulfate disproportion- ation in acidic solutions to yield elemental sulfur and sulfite: S2O32⫺ ⫹ H⫹ 3 S0 ⫹ HSO3⫺ (7) (Johnston and McAmish.or Ar-purged solutions showed a marked lack of version of tetrathionate to sulfate. DISCUSSION analyses are within the analytical uncertainty for the methods used. Initial rates were determined between two initial points A1. 1/T yields highest Fe3⫹ concentrations (5 and 10 mM Fe3⫹) formed an apparent activation energy of 104. derived from the slope of the line. see Brezonik. Control experiments without Fe3⫹ oxidized ⬃ 10 times slower Series VII experiments are presented in Table 6 and Table under identical conditions.68 ⫻ 10⫺8 Fig.4462 G. The reaction order with respect to the initial tetrathio. The samples that contained the A4.40 ⫻ 10⫺8 6.85 ⫻ 10⫺8 1. 1993) yields a slope of nate concentration. Series II experi. This imbalance should be a measurement of species not accounted for between these end members.13 kJ/mol.52 ⫻ 10⫺8 3. Most 4. Figure 7 also plots the degree of imbalance (plotted identical experiments with atmospheric oxygen. ⫺0. Hamers. including those shown in Figures 6 and 11. Series IV experiments conducted to determine the order of reaction with respect to initial tetrathionate concentrations. log initial tetrathionate concen. the tetrathionate oxidation by their metal species over several Fe2⫹:SxO62⫺ molar ratio ranged between 1:1 and 4:1. As ionic strength increases. the reaction order calculated for SO2⫺ generation from rate of 2 ⫻ 10⫺7 mol L⫺1 s⫺1 (Fig. with a trithionate and pentathionate. Results of Series V experiments conducted to determine the order of reaction with respect to initial Fe3⫹ concentrations. An Arrhenius plot (Fig. pH 1. with 2 mM Fe 3⫹ . In numerous Series III experiments showed no evidence for catalysis of experiments. the relative roles of Fe3⫹ and oxygen as oxidants in the con- ments N2. Trithionate Oxidation Results reaction order of tetrathionate degradation was found to be fractional with respect to Fe3⫹ concentration (Fig.31 ⫻ 10 ⫺1. I1/2 tration. show an independence with respect to ionic strength.82 ⫻ 10⫺8 6. other polythionates.47 ⫻ 10⫺8 purge in the presence of Fe3⫹ show that both Fe3⫹ and oxygen The amount of Fe2⫹ produced may be used as an indicator of are required to generate appreciable sulfate. 7). K. and J. [S4O2⫺ 6 ]initial 120 ␮M 250 ␮M 500 ␮M 1000 ␮M 2000 ␮M ⫺1 ⫺1 ⫺8 ⫺8 ⫺8 ⫺8 Initial rates (mol L sc ) ⫺1.5 ⫻ 1011 J mol⫺1.72 ⫻ 10⫺8 9. Time-concentration data are in Table A2. The 3. 10. to equivalent sulfurs). Under continual argon 4 this reaction (normalized to 4 S species) was also found to be purging the oxidation rate was significantly slower than in fractional. 1973) and thiosulfate oxidation by Fe3⫹ to form tetrathionate: Table 4. normalized sulfide.5 ⫻ 10⫺8 . Series VI experiments are presented in Table 5 and Table Series IV experiments are presented in Table 3 and Table A3. (Bronsted relationship. The rate of trithionate oxidation exhibits an initial reaction larly. The plot of log initial rate vs. at 70°C (rate data available in Table 3 and Table A1). log 5000 ⫺2. determined by dividing the SO2⫺4 initial rate by four. Time-concentration data are in Table A1. decreases whereas the total rates (over 30 h of reaction time) Series V experiments are presented in Table 4 and Table A2. the initial slope proximately 0.96 ⫻ 10⫺8 variable [S4O2⫺ 6 ]initial. Figure 6 plots log rate vs. samples with lower calculated preexponential factor of 9. It is reasonable to assume reactivity over several days compared to identical solutions that Fe2⫹ is not reoxidized by oxygen because this reaction is containing oxygen. days(data not shown).89 ⫻ 10 ⫺2. Banfield Table 3. Druschel.48 ⫻ 10⫺8 7.40 ⫻ 10 ⫺1.4. 10).5 ⫾ 4. In contrast. To predict the importance of tetrathionate production in oxidizing systems that contain thiosulfate. over 45 min of reaction.29 ⫻ 10⫺9 1000 ⫺2. is ap.5. SO2⫺ 4 /4 is the sulfur-normalized formation rate.000 ⫺1. The concentra- tion of trithionate and pentathionate was not more than 10% of the total polythionate concentration for these experiments. Fe3⫹ concentrations only formed tetrathionate. [Fe3⫹] initial S4O2⫺ 6 initial rate SO2⫺ 4 initial rate SO2⫺ 4 /4 initial rate (␮M) (mol L⫺1 s⫺1) (mol L⫺1 s⫺1) (mol L⫺1 s⫺1) 500 ⫺1.Log-log plot of initial rate of Series IV experiments vs. 6. . J.3. degradation and sulfate generation for the reaction.5 (Fig. Only sulfate as the difference between calculated rates for tetrathionate and minor amounts of sulfite were produced (no thiosulfate. 1970). 8).43 ⫻ 10 ⫺2.

S4O2⫺6 ratio of ⬃0. 1998). This result reflects experiments. persist long enough to participate in reactions with tetrathion- tions of initial S4O2⫺ 6 .92 ⫻ 10⫺9 0. concentration data are in Table A4. but was sulfate produced by oxidation of pyrite would form tetrathio. served in minor quantities in several experiments. Fig.59 ⫻ 10⫺8 ⫺2. Comparison of calculated rates for Series V experiments. 11) under low pH conditions show minutes to be detected by the methods employed. Thus. (2000) and this study is probably due to differences in the concentrations of other sulfur compounds in the starting salts. our that the reaction of thiosulfate with Fe3⫹ in solution (reaction analyses are unable to detect a subset of possible intermediates 8) is much faster than disproportionation (reaction 7). Elemental sulfur was not observed in any of the low pH solutions containing excess Fe3⫹. S4O6 rates were normalized to 1 sulfur.36 ⫻ 10⫺9 45 2.24 ⫻ 10⫺8 ⫺3. acids (see Eqns. Fig.5 for 500 ␮ M S O 4 6 2⫺ starting concentrations. Fe3⫹ must be stable in these solutions for at least several Kinetic simulations (Fig. dithionates. 1990.g. .5 ⫺3. 2). Time-concentration data are in Table A3.73 ⫻ 10⫺8 2 ⫺1. 1993). After the Bronsted equation. the ratio of S3O2⫺ 2⫺ 6 : S4O6 : sulfite are so reactive in these solutions that they are unlikely to S5O2⫺6 would be approximately 25:50:25. 1. Xu et al. Results of Series VII experiments conducted to determine dependence of the rate of polythionate oxidation at pH 0. 1982). Hu.36 ⫻ 10⫺9 4.5 on ionic the temperature dependence of tetrathionate oxidation at pH 1.92 ⫻ 10⫺8 ⫺2.. and sulfoxyanion radicals). (2000) note a rearrangement ratio of ably the result of reactions involving polysulfane monosulfonic 15:70:15 for a 500 ␮M K2S4O6 salt which contained approx. and Meyer imately 5% S3O2⫺ 6 and S5O2⫺6 impurity at pH 1.48 ⫻ 10⫺8 ⫺2. At higher concentra. Sulfite was ob- acidic solutions containing excess Fe3⫹. tetrathionate rearrangement to trithionate would not be due to the rearrangement reaction. (1996). where the reaction of tetrathionate with sulfite or thiosulfate Assuming the pentathionate concentration is equal to the tri.. 2002). after Steudel et al. Round tively shields the range of electrostatic interaction. tetrathionate should be the at these conditions (Luther. yields trithionate and pentathionate. Time- strength. 8. Polythionate oxidation kinetics and mechanism 4463 Table 5. relating ionic strength to kinetics. Initial rate Total rate Temperature Initial rate M KCl (mol L⫺1 s⫺1) (mol L⫺1 s⫺1) (°C) (mol L⫺1 s⫺1) 0.5. the negative slope indicates reaction Lines indicate the difference between rates calculated for directly between oppositely charged species where greater ionic strength effec. which is consistent with prior experiments (Brandt and duced from thiosulfate instead of elemental sulfur and sulfite in Eldik. the trithionate and pentathionate products are prob- is less (Table 2).32 ⫻ 10⫺9 60 1. thionate oxidation at a pH high enough that their formation In pH 1 solution. and in (e. 7.08 ⫻ 10⫺8 1 ⫺2.1. The discrep. markers are rates calculated from initial rates while square markers are 2⫺ calculated from total rates. almost 100% of thio. This result implies that polysulfides were not the general case made by Williamson and Rimstidt (1993) that produced because polysulfides polymerize into elemental sulfur in the presence of excess Fe3⫹.1 ⫺4. and Ospina. Trithionate and predominant product of thiosulfate oxidation. The result is also pentathionate were observed as intermediate products of tetra- consistent with the results of Schippers et al. Products of Tetrathionate Oxidation by Fe3ⴙ and O2 2 S2O32⫺ ⫹ 2 Fe3⫹ 3 2 FeS2O3⫹ 3 S4O62⫺ ⫹ 2 Fe2⫹ (8) Intermediate sulfur products of tetrathionate oxidation by (Williamson and Rimstidt. Experimental results also show that tetrathionate is pro. tion. oxidized completely within several minutes after initial detec- nate. Results of Series VI experiments conducted to examine the Table 6. Dependence of tetrathionate decomposition rates on ionic strength (I) in Series VI experiments.45 ⫻ 10⫺9 25 5. These species and pentathionate (Series I experiments) yields a S3O2⫺ 6 / may be due to side reactions similar to those illustrated in Eqn. measured S4O2⫺ 6 (solid markers) and SO4 2⫺ (open markers).70 ⫻ 10⫺8 ⫺2. the degree of tetrathionate rearrangement ate.25 ⫺4. 1987. 5 and 6. However.67 ⫻ 10⫺11 0. ancy between the results of Xu et al. thiosulfate and thionate concentration (Eqn.16 ⫻ 10⫺9 70 2. Thus.

Fe3ⴙ and O2 as Oxidants in Tetrathionate and Rate laws from Johnston and McAmish (1973) and Williamson and Trithionate Oxidation Rimstidt (1993) were used in this calculation. thiosulfate in a pH 1. Thus. reoxidation of Fe2⫹ by 6 represent SO2⫺ 4 and the diamonds represent Fe2⫹ measured in the O2 is extremely slow. Observation of the slower rate at which sulfoxyanion prod- ucts form in the absence of O2 and the molar ratio of Fe3⫹ as hexaaquo complex Fe(H2O)63⫹. Druschel. course of the experiment. Results of kinetic calculations for the fate of 1000 ␮M Series VII experiments. Note that negligible S8 was formed as a result of disproportionation in this solution and that ⬃100% of the thiosulfate is converted to tetrathionate.2. oxygen is a more efficient probably rate-limiting. Based on our experimental results. Hamers. Given the lack of amount of Fe2⫹ generated is much less than that expected if significant concentrations of observed intermediates. Thus. The squares which sulfate is generated.g. even though it is not the only oxidant (see above).5 solution containing 2 mM Fe3⫹. Fe3⫹ is a catalyst. sulfite are more efficiently oxidized by Fe3⫹ compared to O2. Over the pH The oxidation of polythionates in acidic solutions containing range investigated. the overall reaction may be approximated: S4O62⫺ ⫹ 3 Fe3⫹ ⫹ 2.5 In the average case (reaction 9). Diagram of a trithionate oxidation experiment with 500 ␮M S3O2⫺ at 35°C in a pH 1. J. 11. It remains unclear from experimental results whether or not this result suggests that polysulfane monosulfonic acids and the transition state yields Fe2⫹..4464 G. Because species such as thiosulfate and sulfonic acids (SxO32⫺) and sulfite. R. In contrast. Regardless of whether or not the Fe3⫹ degradation of S4O2⫺ 6 and S3O6 2⫺ (Figs. the data clearly suggest that Fe3⫹ has an important is a direct participant in the oxidation of these polythionates. The slow rate of polythionate an electron donor in these experiments illustrate the role of O2 oxidation in the absence of Fe3⫹ suggests that Fe3⫹ is involved in the overall reaction pathway (Series II. V experiments). and J. Singer and Stumm 1970). The in the initial reaction step of the oxidation. K. 55% of the excess oxygen in sulfate is derived from O2. 4. this step is Fe3⫹ is the sole oxidant.3. It not. 9. Fe-oxidizing microorganisms present . the Fe3⫹ in solution exists primarily as the excess Fe3⫹ occurs by complex pathways containing many steps. This suggests that a FeS4O6⫹ complex is oxidant of some intermediate products between tetrathionate a transition state towards the formation of polysulfane mono- and sulfate than Fe3⫹. 4. Banfield Fig. 70% of the excess oxygen in sulfate would be O2-derived in a pathway that only utilized O2 as the oxidant. role in the overall oxidation kinetics of tetrathionate.5 solution with 2 mM Fe3⫹. This difference could be measured in future work. At low pH. the dependence of polythionate oxidation on Fe3⫹ means that the rate of Fe2⫹ reoxidation by oxygen could determine the rate at Fig. Arrhenius plot showing the apparent activation energy for the Fig. as a catalyst (none of the other metal ions tested catalyzed The kinetics of Fe2⫹ generation correlates well with the tetrathionate oxidation).5 H2O 3 4 SO42⫺ ⫹ 3 Fe2⫹ ⫹ 9 H⫹ (9) 3⫹ The ratio of Fe to oxygen consumed in experiments varied between 2:3 and 4:2. Implications of the Dependence on [Fe3ⴙ] In the absence of an unlimited external supply of Fe3⫹. 10. As has been pointed out previously (e.75 O2 ⫹ 4. F. 5 and 10). Fe3⫹ could be described sulfur radicals are in integral part of the reaction pathway.

and two symmetrical terminal S-S bonds may be ruptured.09 for [Fe3⫹]. Fig. [S4O2ⴚ 6 ]0. Fe3⫹-rich solutions is generally slow compared to reac. Thus. It is within these experiments. Formation of the Fe3⫹ complex 0.1 and 2 mol/L (Series VI experiments. [Fe3ⴙ]0. The degree of [H⫹] on the oxidation rate.5 mM and 1 mM [Fe3⫹]initial and in the instantaneous rates causes electrons to shift to Fe3⫹. The molecular geometry was [Fe3⫹] at very high initial [Fe3⫹] (Fig. (1996). Thus. I1/2 slope of approximately zero Removal of the highest value (10 mM Fe3⫹ experiment) in the (Table 5). ated. but (contained in Chem3D) and imported to gaussian98. Ab initio acidity investigated. optimization of the FeS4O6⫹ molecule using optimization func- nate by Fe3⫹ is that the rate constant decreases with increasing tions in the gaussian98 program. is in units of mol L⫺1 s⫺1. Oxidation of tetrathionate below a pH of 1 calculations of polythionate structures (this study) and X-ray is therefore controlled by the rearrangement reaction of tetra. the rate depends directly. Fe3⫹ concentrations.06 ⫾ 0. 1989) thionate to form trithionate. separation in the original S4O2⫺ 6 molecule and the calculated on [Fe3⫹]. tion of the FeS4O6⫹ complex probably leads directly to rupture Between pH 1 and 2. the consequence of the nate oxidation rates measured over a range of concentrations dependence of polythionate oxidation rates on iron oxidation (Tables 3 and 4) was used to investigate the dependence of rate rates may be a dependence of the rate on microbial activity.5. there is no measured dependence on of a terminal S-S bond in tetrathionate (Fig. In almost all experiments (those conducted at S3O3⫺ and SO3⫺ (Fig. Although the dependence of charged species (Brezonik 1993). 4. but caution must be exercised in applying the rate law with neutral species may also be a significant part of the overall to solutions containing iron outside the range used in these reaction pathway. the oxidation of polythionates at low pH in the presence of 1995). course of pyrite oxidation has been proposed by Schippers et al. on [Fe3⫹]initial and [S4O2⫺ 6 ]initial at pH 1 to 2 and 70°C. The anal- between the reactants. oxygen. The slope of log k vs. Fe3⫹-rich solu. but not linearly. 7). The results the observed increase in rate in the presence of Fe3⫹. 13).08[Fe3⫹]0. and that reaction reaction. but tions is many orders of magnitude faster than tetrathionate one that should be considered as a potential parallel pathway in oxidation (Williamson and Rimstidt. and [H ] Dependence of Tetrathionate Oxidation Calculations of the electronic structure of the S4O2⫺6 molecule (Fig. Results of multiple linear regression of the log values yields a rate law 4. reaction of thiosulfate to Shifts in electron density towards the Fe3⫹ following forma- tetrathionate is very fast in the presence of Fe3⫹). 8). the pathway is likely ⴙ to favor terminal S-S cleavage reactions. S4O2⫺6 may be expected to yield a slope of ⫺6. based in part on hydrolysis reactions of polythionates to tions involving sulfide. 4. and sulfite. Proposed Mechanism of Polythionate Oxidation Oxidation A mechanism for polythionate oxidation formed in the Tetrathionate and trithionate oxidation in low pH.07 initial rates of oxidation of tetrathionate at between 0. This can be seen by comparison of initial rates for electron density for FeS4O6⫹ . The total rates of reaction exhibit a log k vs. and therefore reaction between Fe3⫹ and ysis shows a possible zero-order overall dependence on initial Fe3⫹.7. where the rate of sulfate production from tetrathionate oxida- I1/2 is negative. calculations. excess Fe3⫹.. Here we propose a mechanism independent of this. Multiple linear regression of tetrathio- by up to 5 orders of magnitude. 7).g. important to note that we were unable to obtain geometry An intriguing feature of the results for oxidation of tetrathio.15⫾0. Cleavage of the S-S bonds is critical in the overall oxidation tude faster than that of tetrathionate under identical conditions of tetrathionate to sulfate. reaction rate on [Fe3⫹]initial is not linear (Fig. r. Thus. Xu and Schoonen. illustrating like-charged interactions as the reaction progresses.61 ⫾ 0. Polythionate oxidation kinetics and mechanism 4465 in natural acidic environments accelerate Fe2⫹ oxidation rates precipitation of FeOOH. indicating a greater influence on neutral species and regression yields an order of 0.6. 1) show that breaking one of the terminal The dependence of sulfate production on Fe3⫹ concentration S-S bonds without electron density redistribution would yield is very complex. oxygen. Role of Ionic Strength of: Ionic strength was found to have an inverse effect on the r ⫽ 10⫺6. but that is a function of a poor fit to nonlinear data. where may be due to rapid back reaction of S2Ox species to yield only the energy was minimized. The Bronsted equation pre.4. Figure 12 illustrates the partial charge ⬍ ⬃2 mM [Fe3⫹]).3 ⫾ 0. reducing it to Fe2⫹. consistent with reaction between oppositely tion. 12). 1993. This form qualitatively indicate that an electrostatic interaction between of the rate law more accurately reflects the nature of the reactants is a part of the reaction pathway. suggest that the terminal S-S bonds are longer and weaker in comparison to the inner S-S bond. The explanation for optimized using the molecular mechanics algorithm MM2 the inverse dependence of the rate on [Fe3⫹]initial is unclear. Relative Rates of Tetrathionate and Trithionate 4. The oxidation rate of trithionate is several orders of magni. tetrathionate should accumulate in solution. while solution pH ⬎ ⬃2 limits [Fe3⫹]aq due to following four reactions: .. Solution pH ⬍1 can influence the electron shift across the S-S bond affects the charge distribution rate of sulfate production through conversion of tetrathionate to in the resulting species and can be represented by one of the trithionate. tetrathionate at high [Fe3⫹] (e. the rate of dicts this slope to be equivalent to the product of the charges reaction varies more significantly with [S4O2⫺6 ]initial.3[S4O62⫺]0. diffraction studies of tetrathionate salts (Christidis et al. thiosulfate. The central S-S bond or one of the throughout the range of temperatures. The rate of polysulfane monosulfonic acids and oxidation by molecular thiosulfate oxidation to tetrathionate in acidic.

as this would involve stripping of all valence electrons from S. withdrawal of an electron from a filled 2p energy level). gaussian98 calculations of tetrathionate and proposed tran. on Fe3⫹ concentrations. illustrating the degree of electron density shift as the pathway. 12). 6.g.g. 12.. F. 12 and 16) is an accepted part of the chain mechanism commonly invoked for the oxidation of sulfite to sulfate (Er- makov. 14. 1998). 1997). However.e. Banfield S3O42⫺ ⫹ O2 3 S3O62⫺ (15) S3O30 may also undergo chain-lengthening/shortening reactions (e. Hamers.4466 G. The ab initio results suggest that electron density shifts of the order of magnitude that would result in SO3*⫺ (where * denotes a radical species) via reaction 10 are most plausible (Fig.8. It is conceivable that S3O30 formed by reaction 10 could oxidize to S3O2⫺6 via series of hydrolysis and oxidation reac- tions: Fig. SO3*⫺ will react very quickly to sulfate (Brandt and Eldik. trithionate oxida- S4O62⫺ ⫹ Fe3⫹ 3 S3O3⫺ ⫹ SO30 ⫹ Fe2⫹ (11) tion should control overall tetrathionate oxidation rates at so- lution pH’s ⬍1 where rearrangement of tetrathionate to or trithionate occurs. 5) before oxidation of the terminal sulfur (sulfone) group (Eqn. Kinetic Parameters and Implications for Pyrite Oxidation The order of the tetrathionate and trithionate oxidation reac- tions is fractional and the rate exhibits a nonlinear dependence Fig. and 15). 14 and 15). .. This generalized pathway is proposed based S3O ⫹ H2O 3 S3O 0 3 2⫺ 4 ⫹2H (14) on empirical observations and theoretical computational evidence. based on a combination of theoretical calculations and experimental results. Oxidation rates of tetrathionate range between 10⫺11 s⫺1 at 25°C and 10⫺8 s⫺1 at 70°C (pseudo first order rate coeffi- S4O62⫺ ⫹ Fe3⫹ 3 S3O32⫺ ⫹ SO3⫹ ⫹ Fe2⫹ (10) cients). The SO3*⫺ tetrathionate and trithionate oxidation product (Eqns. The mechanism of trithionate oxidation may be very similar to that for tetrathionate. S3O62⫺ ⫹ Fe3⫹ 3 S2O30 ⫹ SO3*⫺ ⫹ Fe2⫹ (16) The S2O30 product may then undergo a series of hydrolysis and oxidation reactions to form SO3*⫺ radicals (as postulated for S3O30 in Eqns. A proposed overall reaction pathway. 13. is shown in Figure 13. R. Eqn. e. Formation of sulfite via reaction 13 would be possible if the degree of electron density shift is significantly greater than that predicted by the DFT model.. K. The oxidation of trithionate exhibits a pseudo first order rate constant of 2 ⫻ 10⫺7 s⫺1 at 35 °C. and J. tions and parallel reactions. Because trithionate or oxidation is several orders of magnitude faster than tetrathio- nate oxidation at comparable temperatures. The slow oxidation of polythionates in low pH solutions S4O62⫺ ⫹ Fe3⫹ 3 S3O30 ⫹ SO3*⫺ ⫹ Fe2⫹ (12) containing excess Fe3⫹ (which is many orders of magnitude slower than the formation of polythionates from thiosulfate) or indicates that a pathway progressing through thiosulfate in S4O62⫺ ⫹ Fe3⫹ 3 S3O3⫹ ⫹ SO32⫺ ⫹ Fe2⫹ (13) these conditions should yield significant quantities of polythio- nates. resulting in SVIO30 (it is also unclear how this molecule might form SO2⫺ 4 with no electrons left to covalently bond oxygen with). and specifically to the importance of side reac- a result of Fe3⫹-complexation and cleavage of a terminal S-S bond. this is not supported by experimental or field Formation of SO3⫹ via reaction 10 is highly unlikely because it would require S is in a ⫹7 state (i. Druschel. J. This is due in part to the complexity of sition state FeS4O⫹6 . 4. Proposed pathway of tetrathionate oxidation at low pH in the ⫹ presence of excess Fe3⫹. Formation of SO30 is unlikely (reaction 11).

D.. 3).D. Z. W. A. Pople J. Daniels A.. and Weinhold F.4).. Poskrebyshev G. W.15 ⫾ 0. University of Illinois. this brings into Wisconsin). Fox D. 1987). Interf. C... 116. Komaromi I. Act2. M.. sulfate. 13) are based on K2S4O6.. Stratmann 4.. J.Theoretical Chemistry Institute. Malick initial step involving oppositely charged species.0.. (1989) between them indicates some intermediates in the reaction Experimental charge density in polythionate anions: II.. (1973) A study of the rates of sulfur kinetic data provide new insights into sulfur transformations production in acid thiosulfate solutions using S-35. This reaction of tetrathionate Frisch M. X-ray study pathway are more efficiently oxidized by one or the other. Head-Gordon M. r ⫽ 10⫺6. Peng C. L. implicating a fast reaction pathway for tetrathionate lacombe M. 188. (2000) Bacterial and chemical oxidation of pyritic mine tailings at low temperatures. Wyoming. W. Coll. D. ation of stable sulfur isotopes: A review and critique. Gonzalez C. Am. will persist in these environments where they form from thio.. a wide range of pH conditions (Luther. Ermakov A. Cammi R. J. Adamo C. University of Wisconsin. Kinetics of intermedi- ate sulfur species reactions in the environment.. E. 41.. Chen W. Piskorz P. 42.. Uni- consistent with oxygenation and charge balances in the trans.. (1979) Microbiological fraction- With both Fe3⫹ and molecular oxygen present in these reac. Pomelli C. Burant J.. with a calculated preexponential factor of 9. Liashenko A. and Sand W. Scuseria J. Schippers sulfate (Fig. J.. Keith T. E. L... a combination of experimental and calculation results. and Will G. and are Druschel G. Carpenter J. Chal- ditions.. 38. Chambers L. K. (2002) Sulfur biogeochemistry.... A. Foresman J... 31– 42.. Al-Laham M. low Guide to Rxn.and polysulfane mono.. K. (2000) Chemical deposition of copper molecular oxygen to sulfate through an initial step in which sulfide films in the surface of polyethylene by the use of higher formation of Fe3⫹-polythionate complexes result in the break polythionic acids. and van Eldik R.. 249 –293. 112–119.D. and to J. Cossi M.3[S4O62⫺]0. S4O62⫺ ⫹ 3 Fe3⫹ ⫹ 2.. Replogle E. and McAmish L. 129 –133. CO) question the pathway for pyrite oxidation involving release of for very helpful suggestions and comments on this manuscript. 225–238. A. (2000) The Geochemist’s Workbench 3. Wong M. under low pH conditions in the presence of excess Fe3⫹. Montgomery J. Strain M. 23. J. where the rate of sulfate production from tetrathionate oxida.. L. The overall reaction of tetrathionate in oxic. Dr.. The overall rates of tetrathionate and trithionate oxidation Glendening E. Zakrzewski V.13 kJ/mol.. 12–19. Acknowledgments—We would like to thank Dr. thesis. Trans. Associate editor: D. K. Gill P... Geomicrobiol. Prog.. C. (forming tetrathionate. is in units of mol L⫺1 s⫺1. (1997) Sulfite oxidation: The state-of-the-art of the problem. thesis. A.. A. Ph.. D.. Transit. A. Kristallogr. 667– 675. J. Petersson G. (1998) Kinetics and mechanism of the pH conditions with excess Fe3⫹ may be represented by: iron(III)-catalyzed autooxidation of sulfur(IV) oxides in aqueous solution. These Johnston F..5 ⫻ R. C. Nanayakkara A. 193–217. 2002). A. formation of polythionates to sulfate. Martin Schoonen and ments containing significant concentrations of Fe3⫹ (Schippers Michael Borda (State University of New York at Stony Brook).75 O2 ⫹ 4. B. A. Druschel G. B. Cui Q. Druschel. Farkas O. A. Ph. Raghavachari K.. magnitude faster than tetrathionate oxidation under these con. The of the electron density distribution in potassium tetrathionate. and Cohn C. A... N.. II: An anion-ex- change resin technique for sampling and preservation of sulfoxyan- both Fe3⫹ and initial tetrathionate concentration In oxic. Brezonik P. tions as potential oxidants. (2001) NBO 5. This thiosulfate and the oxidation of dissolved thiosulfate by Fe3⫹ to research was supported by grant CHE 9521731.. 283... This indicates that polythionates nion formation during pyrite oxidation at pH 6 –9 and 30 C. S. Schlegel H. Polythionate oxidation kinetics and mechanism 4467 observations of pyrite oxidation products in low pH environ. N. proposed details of the reaction pathway (Fig. Contam. W. G.5 ⫾ Cheeseman J. and Gtplot. M. and Dr. 1. 1993) in Goldhaber M. Bethke C. Xu The rate law for the overall oxidation of tetrathionate to Y.. and thus may be metabolized by organisms that utilize Hu B. Mark Williamson (Geochemical Solutions..1: A User’s sulfonic acids. D.. Bo Hu (University of et al. K. Reed A. Ball J. Nordstrom D.. P. Clifford S.5 and 5 mM Fe3⫹ the rate law is: Elberling B. 4(3). CHE 9807598 from the form tetrathionate. Morales C. Geochem. Kinet. are markedly slower than rates of oxidation of thiosulfate Bohmann J. Metal.. Trucks G. Morokuma K. Although the mechanism of formation of and two anonymous reviewers for their comments towards improving this manuscript.. Gonzalez C. Sci. Cioslowski The reaction of trithionate is shown to be several orders of J... React. and Purmal A.. Robb M.... J. K. CONCLUSIONS REFERENCES 3⫹ Trithionate and tetrathionate oxidizes with both Fe and Ancutiene I. Ochter- strength that supports a reaction pathway rate-limited by an ski J. E.. Catal. Fe3⫹-rich solutions. r. A.. Dr. .. low ions in natural waters. to sulfate is characterized by a high activation energy (104. B. Y. Schoonen M.. Ortiz J.. and Janickis V.. Kudin 1011 J mol⫺1).. Brandt C. 295–308. Stefanov B. M. (1983) Experimental study of metastable sulfur oxya- acidic... Sci. tion. Pol. Liu G. (1993) Chemical Kinetics and Process Dynamics in ⫹ 3 Fe2⫹ ⫹ 9 H⫹ (9) Aquatic Systems. D. and coupled with a rearrangement reaction. A. Millam J. Martin R. Ayala P. USA. the competition for reduced species Christidis P. Tact. oxidation below pH 1 that may be through trithionate oxidation Andres J.. Barone V. 1996. Rimstidt for editorial handling and a very thiosulfate-like moieties on the pyrite surface may be valid over valuable review. R. Williamson and Rimstidt. S.. Schippers A.. and Trudinger P.5 H2O 3 4 SO42⫺ Chem. Kirfel A.. We would also like to thank A. Tomasi J. (2001) Gaussian98 (Revision A. pH conditions between 0.09 Hydrol. followed by conversion of tetrathionate to National Science Foundation. B.3 ⫾ 0. Dapprich S. Gomperts R. Sci. At low pH. CRC Press. V. Rabuck A. (2003) Sulfur geochemistry of hydrothermal waters sulfate has been defined as fractional in order with respect to in Yellowstone National Park. intermediate sulfur compounds for energy generation.. Mennucci B. Y.61 ⫾ 0. Rentzeperis P. Badenhoop J... George Luther (University of Delaware). of a terminal S-S bond to form SO3n. the thiosulfate detachment step may not be favored at lower pH.. Colloid. versity of Wisconsin. and a relationship of reaction rate with ionic K.08[Fe3⫹]0. The influence of pH. M. Gaussian Inc.. Johnson B. medium and aging. (2002) Chemical and structural study at the interface between metal sulfides and acids. Rimstidt 5. J.

3424 –3431. Takano B. 34. (1968) The lower oxy-acids of sulphur. 199 –216.4468 G. Saitoh H. 45. Japan. J. . monitoring aqueous polythionates in an active crater lake. Edit. and Wood A. decomposition of thiosulfate and polythionates in Cinder Pool. (1982) Raman spectrometric study of the graphic determination of polythionates in natural water. 3555–3561. 97. Wiley. Dicinoski G. (2001) presence of pyrite in low-temperature aqueous solutions.. Technol. D.. In Aquatic Chemical Kinetics (ed. 407– 423. Min. (2001) The Physical Chemistry of Natural Waters. 171. Jozsa P.. Steudel R. J. Acta 57. K.. W. 52. H. and Rimstidt J. M. A. and Newcomb W. Geochim.. 5443–5454. Pratt A. (2000) Molden: a pre. Shaw M. Mol. Am. (1970) Acidic mine drainage: The rate- Luther G. G. (1995) The stability of thiosulfate in the O’Reilly J. and J. J. Res. Microb. and Nickless G. (1999) Intermediary sulfur chemolithotrophic sulfur bacteria in relation to the sulfur oxidation compounds in pyrite oxidation: Implications for bioleaching and pathways. 475–501. Acta 51. Lyons D. 165–192. Gobel T.22) and Luther G. (1987) Correlation of volcanic activity with sulfur oxyanion McGuire M. D. 97–105. (1994) Synthesis and determination of Schippers A. Nordstrom D. 1121–1123. J. N. Shirane Volcano.. W. and Ospina M. Cosmochim. Environ. Biot. (1999) Bacterial leaching of metal sulfides thiosulfate and polythionates. A. Anal. (2000) Extraction and quantitative speciation in a crater lake. Takano B. and Hazeu W. Herman J. and Mills A. and Sand W. and Hamers R. Quality Tools. 1561–1571. W.. J.. In Methods in Enzymology. A. 34. Nordstrom D.. (1998) of the ferric-thiosulfate complex in acidic aqueous solutions. A. 1919 –1934. D. (1992) Correlation between Moses C. Cosmochim. J. F. and Rimstidt J. (2000) A telemetering system for Millero F. Int. Release 13 for Windows. analysis of elemental sulfur from sulfide mineral surfaces by high. M. Appl. 14. and J. Kusatsu- 4651– 4655... J. Schoonen M. Rohwerder T. 65. Alpers C. (1993) Separation of aqueous put. Chem. (1994) The kinetics and electro- Nordstrom D. Des. O. J. . Hamers. Sci. (1993) The rate of decomposition Nesbitt H. (2000) Sulfur geochemistry of hydrothermal waters in of rate data: The rate of reaction of ferric iron with pyrite. Williamson M. K.and post. and enzymatic reactions. Nickless). Zhang Y. Microb... (1993) Measurement and analysis Ball J. Volcanol.Vol. Technol. K. Geochim. (1994) Geochemical implications performance liquid chromatography.. W. 83. Res. Xu Y. and Sand W. Acta 284. and Noordik J. Cosmochim. Takano B. Angew. In Suzuki I. Appl. III. 1633–1635. polythionates by reversed-phase ion-pair liquid chromatography Schippers A. Xu Y. P. Takano B. Aqueous pyrite oxidation by dissolved oxygen and by ferric iron. composition of bacterial sulfur secretions. and Rimstidt J. W. decomposition of aqueous tri. 97. (1987) structure and thermodynamic properties of aqueous sulfur species. Stumm). L. R. 3867–3880. Elsevier. (2000) chemical rate-determining step of aqueous pyrite oxidation. 173–198. D. orbital theory considerations. Acta 432. Peck Jr. Appl. Jia Z. H. W. Druschel. (1988) Quenching and liquid chromato- Meyer B. Sulfur and iron surface states on fractured pyrite surfaces. Arch. Williamson M. 59 – bacterial leaching of pyrite. 123–134. K.. ic-separation of higher polythionates SnO6(2⫺) (n ⫽ 3. M.. Microbiol. Banfield Kelly D.. 23–38. W. and Lu P. (1999) Oxidation of inorganic sulfur compounds: Chemical Inorganic Sulphur Chemistry (ed.. Wiley. Environ. 65. Environ. Chim. USA. Geochim.. Chromatographic and electrophoretic separation of inorganic sulfur Cosmochim. 509 –534. W. Sci. Zheng Q. Can. Yellowstone National Park.. Geochim. 1067–1076. and Haddard P.. fides and elemental sulfur. Geochim. 219 –229. and Scaint M. III (1990) The frontier-molecular-orbital theory approach their detection in cultures of thiobacillus-ferrooxidans—Molecular in geochemical processes. Geochim. (1987) Chromatograph- 3193–3. Science 167. R. and Watanuki K. Chim. Geoth. Volca- Minitab Inc. and Blowes D. Kelly D. and sulfur-oxygen species. Anal. Cosmochim. Geochem. Aid. (2003) MINITAB User’s Guide 2: Data Analysis and nol. California. 151–153.. Singer P. Geochim. (1999) Thermodynamic aspects of energy conservation by Schippers A.. Zou H. Microbiol. 104 –110. and Takano E. R. pp. Ptacek C. D. and Schoonen M. Wyoming. Acta 57.. . 62. and Stumm W. proceeds by two indirect mechanisms via thiosulfate or via polysul- Inorganic Microbial Sulfur Metabolism (eds. 254 –258. processing program for molecular and electronic structures. Acta 51. Academic Press.. Schaftenaar G. 4605– 4622. C. 14. 397– 406. pp. Acta 56. of subaqueous molten sulfur at Yugama Crater Lake. and Ohsawa S. (1996) Sulfur chemistry in with suppressor-conductivity detection. Acta 59. LeGall). Bancroft G. Cosmochim. 26. (1987) Pyrite oxidation and reduction: Molecular determining step. D. Science 235. P. 28. biodepyritization of coal.. Phosphorus Sulfur 847– 854.. J. Com.. and Sand W. A. Environ. and Rimstidt J. 243. Talanta 35. pp. Acta 58. A. Geoth. 319 –321. P. S. Williamson M. Cunningham K. A. Holdt G. Negative pH and extremely acidic mine waters from Iron Mountain. J... Microbiol. J. II: Formation and Cosmochim.and tetrathionate.

Data for experiments 2⫺ investigating tetrathionate degrada. Total elapsed time SO4 conc. 70°C. (min) (␮M) (min) (␮M) (min) (␮M) (min) (␮M) 0 0 0 0 0 0 1 61 380 10 380 368 380 847 40 219 1830 40 635 516 635 1664 100 814 3450 60 1460 978 1460 2243 160 1064 6545 90 2290 1683 1820 2426 220 1468 3440 2030 2290 2529 280 1592 4355 2224 2935 2622 340 1726 5820 2442 3440 2695 475 2171 7260 2492 4355 2756 595 2427 5820 2890 715 2522 7260 2948 845 2681 1295 3034 . Time elapsed SO4 conc.5 1 24 34 47 61 107 (min) M M M 1M 2M 40 143 146 138 219 377 100 336 543 577 814 812 1 0 0 0 0 0 160 368 640 750 1064 1224 65 721 704 602 344 227 220 377 727 1009 1468 1457 120 891 881 765 489 355 280 454 867 1155 1592 2000 185 1023 1010 911 651 468 340 464 896 1342 1726 2279 250 1160 1094 1015 794 565 475 474 1005 1484 2171 2854 330 1237 1156 1154 957 639 595 454 naa 1605 2427 3015 440 1331 1289 1240 961 na 715 485 na 1655 2522 3831 570 1398 1383 1335 1082 862 845 478 na 1694 2681 3663 690 1454 1414 1374 1111 912 1295 na 1045 1801 3034 4188 825 1488 1463 1427 1294 969 935 naa na 1452 na na a na ⫽ not analyzed. 70°C. 2⫺ 500 ␮M S4O6 . 2⫺ Table A2. 2⫺ Table A4. tion rate dependence 2⫺ with ionic strength dependence (pH 0.5. 2000 ␮M Fe3⫹. 25°C 45°C 60°C 70°C 2⫺ 2⫺ 2⫺ 2⫺ Total elapsed time SO4 conc.000 ␮M Fe3⫹ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ 2⫺ Time elapsed S3O6 S4O6 S5O6 SO4 S3O6 2⫺ S5O6 SO4 S3O6 2⫺ S5O6 SO4 2⫺ S5O6 S3O6 SO4 (min) (␮M) (␮M) (␮M) (␮M) (␮M) S4O6 (␮M) (␮M) (␮M) (␮M)S4O6 (␮M) (␮M) (␮M)S4O6 (␮M) (␮M) (␮M) (␮M) 0 nda 442 nd 16 nd 465 nd 33 nd 457 nd 47 446 nd 4 105 45 nd 370 nd 237 nd 330 nd 404 nd 286 17 589 310 6 16 497 110 nd 314 nd 341 nd 254 nd 627 nd 221 20 772 276 5 6 653 160 nd 294 nd 392 nd 227 nd 700 nd 203 20 858 254 nd 3 718 430 nd 216 nd 534 nd 203 nd 922 nd 170 19 1063 233 nd nd 919 310 nd na nd 616 nd 131 nd 1190 nd 138 17 1125 176 2 nd 1070 1060 nd na nd 754 nd 89 nd 1451 nd 115 16 1305 114 nd nd 1350 a nd ⫽ not detected.5. 500 ␮M Fe3⫹ 1000 ␮M Fe3⫹ 5000 ␮M Fe3⫹ 10. 1410 1580 1517 1608 1412 1180 1770 na na 1653 na 1253 a na ⫽ not analyzed. Initial [S4O6 ] Time elapsed Ionic strength (M KCl) (min) 120 ␮M 250 ␮M 500 ␮M 1000 ␮M 2000 ␮M Time elapsed 0.5.25 0.5. Data for experiments investigating tetrathionate degradation rate dependence with temperature at pH 1.1 0. 1000 ␮M Fe3⫹). 70°C with 2 mM Fe3⫹. Table A3. Polythionate oxidation kinetics and mechanism 4469 APPENDIX Table A1. Data for experiments investigating tetrathionate degradation rate dependence with [Fe3⫹] at pH 1. 1000 ␮M S4O6 . Data for experiments investigating tetrathionate degrada- tion rate dependence with [S4O6 ] at pH 1. 500 ␮M S4O6 . Total elapsed time SO4 conc.