AN EXPERIMENTAL INVESTIGATION OF MULTIPLE SULFUR ISOTOPE FRACTIONATIONS DURING HETEROGENEOUS REACTIONS BETWEEN SO₂ AND ACTIVATED CARBON

Abstract

Recent theoretical and experimental investigations on multiple sulfur isotope fractionations (Lasaga et al., 2008; Watanabe et al., 2009) have suggested the possible importance of reactions between solid organic compounds and oxidized sulfur species in the creation of anomalous isotope fractionation of sulfur (AIF-S) in nature. In order to understand the details of chemical and isotopic fractionation processes involving chemisorption and redox reactions, we have conducted laboratory experiments to analyze reactions between SO_2(g) and powdered activated carbon (AC; BET surface area = ~460 m²/g) at 200° and 250 °C  in a specially constructed closed system. During each of the three series of experiments, which lasted for up to 480 hours, we monitored the changes with time in pSO_2(g) due to adsorption/desorption of SO_2(g) on/from the AC, and periodically sampled aliquots of SO_2(g) for S isotope analyses. The SO₂-reacted AC was analyzed for its chemical composition using an X-ray photoelectron spectrometer (XPS), an elemental analyzer (EA).  Different forms of S-bearing compounds were sequentially extracted by H₂O, HCl, Cr-solution, and Kiba solution as Ag₂S, and analyzed for ³²S, ³³S, and ³⁴S abundance ratios.

Results of the experiments and analyses indicate that three kinds of S-bearing species were continuously incorporated in the AC during reaction with SO_2(g) at 200250 °C: (A) weakly adsorbed SO₂ (i.e., SO_{2(w. ads)}) which was in chemical (and probably isotopic) equilibrium with SO_2(g); (B) strongly adsorbed SO₂ (i.e., SO_{2(w. ads)}), which was degassed at 300-400 °C from the AC at the end of each series of adsorption/desorption experiments; and (C) non-degassable S compounds (i.e., S_(NDG)). After reacting with a total of 10.97 mmoles of SO_2(g), the AC obtained 1.26±1.20 mmoles of SO_{2(s. ads)} and 1.12±0.05 mmoles of S_(NDG). Approximately 60% of the S_(NDG) occured as oxidized-S compounds (i.e., S-O-C compounds) that were extracted by water and recovered as BaSO₃ (and/or BaSO₄), ~20% as reduced-S compounds (i.e., S-C compounds) that were extracted by Cr-reducing solution, and the remaining ~20% as unidentified (but probably reduced S-bearing) residual S-compounds that were extracted by Kiba solution. The sequence of reactions among these S-bearing clusters was estimated to be: SO_2(g) = SO_2(w. ads) è SO_{2(s. ads)} è (S-O-C)_AC è (S-C)_AC, representing the trends of increasing S/O and S/C ratios of the AC caused by continuous oxidation of C to CO₂. The bonding energy for the SO_{2(w. ads)} was estimated to be ~30 kJ/mol.

Large kinetic isotope fractionations of sulfur isotopes occurred during the adsorption/reduction processes. Compared to the δ³⁴S of co-existing SO_2(g), the δ³⁴S values of the oxidized S-bearing species increased to 4±1‰ for SO_{2(w. ads)}, +2.4±0.4 to

+7.3±0.4‰ for SO_{2(s. ads)} and +11.6±0.7‰ for S-O-C species, but the reduced-S-C species (i.e., Cr-reducible S, and the residual S) are enriched in the lighter isotopes with the δ³⁴S values of -8.3±0.7‰. The δ³³S – δ³⁴S relationships of the S species incorporated in the AC follow the normal (i.e., mass-dependent) isotopic fractionation, but their Δ³³S values are slightly positive (Δ³³S = 0 to +0.14‰). The preferential enrichments of heavier isotopes in the adsorbed S-O bearing species, and the preferential enrichments of lighter isotopes in the reduced S species during the reduction of oxidized-S species, agree with the theoretical predictions.

Based on comparisons of our results with those of other experimental studies on

TSR and the S isotopic characteristics of H₂S (and other S-bearing compounds) in petroleum, natural gas, and some ore-forming solutions, we suggest the following: (1)

The natural TSR probably occurred by solid C-bearing compounds (e.g., dead bodies of (micro)organisms and kerogen), rather than by gaseous, aqueous, or liquid C-bearing

compounds. (2) The small isotopic fractionations between H₂S and their source SO₄^2- (i.e., Δ³⁴S = δ³⁴S_H2S – δ³⁴S_SO4 = -10 to 0‰) in petroleum and natural gas that has previously been interpreted as a result of nearly-complete thermochemical reduction of

SO₄^2- in closed systems is instead caused by TSR involving solid C-bearing compounds.

The results of our study also suggest that the large AIF-S signatures (Δ³³S = -1.37 to +1.84‰) observed for SO₄^2- in some air pollutants (Ding et al., 2006) were probably created by chemisorption isotope effects between coal and the SO₂ generated by the burning of pyrite-rich coal. Because the adsorption energy for the surface S-O compounds varies depending on the physical and chemical properties of solid C-bearing compounds, as well as the S-O speciation, some kerogen in Archean sediments, at a particular maturation stage, may have had small adsorption energy to produce AIF-S signatures during TSR.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TABLE OF CONTENTS

 

List of Tables……………………………………….…………………………………. viii

List of Figures…………..…………………….……………………………………….… ix

Abbreviations…………………………………………………………………………… xi

Acknowledgements……………………………………………………………………. xii

 

1. Introduction …………………………………………………………………………………………………….. 1
   • Background ……………………………………………………………………………………….. 1
   • Objectives …………………………………………………………………………………………. 5

 

2. Experimental procedures ………………………………………………………………………………….. 6
   • Starting material …………………………………………………………………………………. 7
   • Experimental System ………………………………………………………………………….. 7
   • Experimental Procedure ………………………………………………………………………. 8

 

3. Analytical Methods
   • Recovery of SO_2(g) as BaSO₄ ……………………………………………………………… 10
   • X-ray Photoelectron Spectroscopy of Activated Carbon ………………………… 11
   • Sequential Sulfur Extraction from Activated Carbon …………………………….. 11
   • Chemical Analyses of Activated Carbon by Elemental Analyzer ……………. 11
   • Sulfur Isotope Analyses …………………………………………………………………….. 11

 

4. Results and Interpretations of the Chemical Data
   • Speciation and Amounts of S incorporated in the Activated Carbon during

Reactions with SO_2(g) ………………………………………………………………………………. 12

• Sequence and Rates of Chemical Reactions …………………………………………. 15

 

5. Results and Interpretations of the Sulfur Isotope Data
   • δ³⁴S and Δ³³S values of the initial SO_2(g) and the uncertainties ……………….. 18 5.2. δ³⁴S and Δ³³S values of the experimental products ……………………………….. 19

 

6. Discussion: Chemical and Isotopic Fractionations of Sulfur during Reactions between

Sulfate and Solid Organic Compounds ………………………………………………………………… 21

• Comparisons of experimental data on SO₂-AC reactions ………………………. 21
• Processes of Multiple Sulfur Isotope Fractionations during Sulfate – Amino

Acid Reactions ………………………………………………………………………………………. 22

• Origins of Sulfide Sulfur in Petroleum, Natural Gas and Ore-forming Fluids

……………………………………………………………………………………………………………… 23

• AIF-S in natural samples …………………………………………………………………… 25

7. Summary and Conclusions ……………………………………………………………………………… 26

References ……………………………………………………………………………………………………….. 28 Appendix A. X-ray Diffraction for the Purity Check of the Recovered BaSO₄ ………….. 34 Appendix B. Sequential Sulfur Extraction from the Activated Carbon for Chemical and

Isotopic Analyses ………………………………………………………………………………………………. 35 Appendix C. Sulfur isotope analyses: methods ……………………………………………………… 37

1. Introduction

The presence of sulfide and sulfate minerals with anomalous isotope fractionations of sulfur (AIF-S) in sedimentary rocks older than ~2.4 Ga in age, but the virtual absence of AIF-S in younger rocks, have been considered by many recent geoscientists as the definitive evidence for a dramatic change from an anoxic to oxic atmosphere ca. 2.4 Ga (e.g., Kasting et al., 2001; Pavlov and Kasting, 2002; Holland, 2006). This is because, until recently, the only known process to create AIF-S was UV photolysis of SO₂ in an O₂-poor atmosphere. However, a theoretical study by Lasaga et al. (2008) and an experimental study by Watanabe et al. (2009) have shown that chemical reactions between solid organic compounds and gaseous (or aqueous) sulfur-bearing species may cause AIF-S under certain conditions. This study is intended to investigate the details of chemical and sulfur isotopic fractionation processes during reactions between a simple solid carbon compound (activated carbon) and SO₂ gas at 200 ° and 250 °C

 

1.1. Background

Sulfur has four isotopes ³²S, ³³S, ³⁴S and ³⁶S, with the average abundances of 95.04%, 0.75%, 4.20%, and 0.015% respectively (Coplen et al., 2002). Since the theoretical works by Bigeleisen and Mayer (1947), researchers have recognized that most natural samples show the following sulfur isotopic relationships:

δ³³S_i (‰) = 0.515 x δ³⁴S_i                                                                            (1)

and                   δ³⁶S_i (‰) = 1.89 x δ³⁴S_i                                                                              (2)

where subscript i refers to a sulfur-bearing compound, and δ³³S_i, δ³⁴S_i and δ³⁶S_i are defined as permil deviations of the ³³S/³²S, ³⁴S/³²S, and ³⁶S/³²S ratios of compound i from that of the international standard (V-CDT: Vienna Canyon Diablo Troilite), such as

δ³³S_i (‰) = [(³³S/³²S)_i / (³³S/³²S)_V-CDT -1] x 1000                                      (3) The ³²S/³⁴S ratio of V-CDT is determined as 22.649 (Ding et al., 2001).

Relationships (1) and (2) occur because the magnitude of isotopic fractionation during most (bio)chemical reactions depends primarily on differences in isotope mass. For this reason, (1) and (2) are often called as equations for the “mass-dependent fractionation (MDF)”, “terrestrial fractionation (TF)”, or the “normal fractionation (NF)” of sulfur isotopes.

Deviations of isotopic compositions of i from the MDF relationships are typically expressed by:

Δ³³S_i = δ³³S_i  – 0.515 x  δ³⁴S_i                                                                      (4)

Δ³⁶S_i = δ³⁶S_i – 1.89 x  δ³⁴S_i                                                                        (5)                                     ³³θ’ = δ³³S_i / δ³⁴S_i                                                                                       (6) and                  ³⁶θ’ = δ³⁶S_i / δ³⁴S_i                                                                                (7)

 

When Δ³³S_i and/or Δ³⁶S_i values fall outside the ranges of 0±0.2‰ and 0±0.4‰ respectively, and ³³θ’ and/or ³⁶θ’ values fall outside the ranges of 0.51±0.01 and 1.9±0.1 respectively, the sample is considered to contain mass independently fractionated sulfur isotopes (MIF-S) (Farquhar and Wing, 2003) or anomalously-fractionated sulfur isotopes (AIF-S) (Lasaga et al., 2008).

Hoering (1988) reported that some Archean-aged barites displayed distinct AIF-S signatures: Δ³³S = -0.46‰ and Δ³⁶S = +2.15‰. However, it was a report by Farquhar et al. (2000a) that generated great interests among geoscientists studying multiple sulfur isotope geochemistry. Farquhar et al. discovered that sulfide and sulfate minerals from some sedimentary rocks of Archean age possessed AIF-S signatures, mostly positive Δ³³S values (-0.8 to +2.0‰) for sulfides and mostly negative Δ³³S values (-1.3 to +1.9‰) for sulfates.  Farquhar et al. (2001) reported that SO₂ photolysis in an O₂-free atmosphere by a UV laser with a 193 nm wavelength produced S^o and SO₄^2- with large positive Δ³³S values (+59.9 to +69.3‰) and negative Δ³³S values (-23.3 to -12.8 ‰), respectively. Subsequent researchers recognized that many (but not all) sedimentary rocks >2.45 Ga in age possessed AIF-S signatures (e.g., Mojzsis et al., 2003; Ono et al., 2003; Papineau et al., 2006; Philippot et al., 2007; Goldman et al., 2008; Ono et al. 2009a,b; Shen et al., 2009). Through photochemical modeling, Pavlov and Kasting (2002) suggested that the S⁰ produced by photo-dissociation of SO₂ would not be preserved if the atmospheric pO₂ was higher than ~1 ppm. Based on these studies, many geoscientists have accepted the theory that the AIF-S signatures in sedimentary samples were created by UV photolysis of volcanic SO₂ in an oxygen poor atmosphere, and that the AIF-S record is the smoking gun for the “Great Oxidation Event (G.O.E.)” at 2.45 Ga. However, many serious problems exist with this theory:

• The discrepancy between the AIF-S signatures produced by photochemical reactions and those of sedimentary rocks (Fig. 1): UV photolysis of SO₂ with 193 nm created elemental sulfur with large positive Δ³³S (+59.9 to +69.3‰) and large negative δ³⁴S (-85.0 to -39.1‰) values (Farquhar et al., 2001). In contrast, natural pyrites with AIF-S signatures typically possess positive Δ³³S (0 to +11‰) and positive δ³⁴S (0 to +30‰) values (e.g., Farquhar et al., 2000a; Hu et al., 2003; Mojzsis et al., 2003; Ono et al., 2003; Bekker et al., 2004; Papineau et al., 2005; Whitehouse et al., 2005; Cates and Mojzsis, 2006; Johnston et al., 2006; Ono et al., 2006; Bao et al., 2007; Farquhar et al.,

2007; Hou et al., 2007; Kamber and Whitehouse, 2007; Kaufman et al., 2007; Ono et al.,

2007; Papineau et al., 2007; Philippot et al., 2007; Domagal-Goldman et al., 2008; Johnston et al., 2008; Partridge et al., 2008; Ueno et al., 2008; Guo et al., 2009; Ono et al., 2009a,b; Shen et al., 2009; Thomazo et al., 2009). There is no obvious mechanism to significantly increase the δ³⁴S values from the photochemically produced S⁰ values (-85.0 to -39.1‰) to those of natural FeS₂ (0 to +30‰). Naraoka and Poulson (2008) and Lyons (2008) have recognized that the absorption spectra of SO₂ isotopologues are extremely complex in a wavelength region ~180 to ~230 nm, which causes the AIF-S signature of the products of UV photolysis to be extremely sensitive to the wavelength of a UV source. When a broad-band UV lamp, which produces the UV spectra similar to the natural sun light, was used in the SO₂ photolysis, the S⁰ had positive Δ³³S (+15±5‰) and positive δ³⁴S (+180±10‰) values with δ³³S/ δ³⁴S values of 0.54-0.6 (Matsterson et al., 2011); these δ³⁴S-Δ³³S trends are not consist with the majority of the Archean data (Fig.

1).

• The existence of AIF-S signatures in post-2.45 Ga samples: AIF-S signatures were found in post-2.45 Ga materials. Young (2009; personal communication 2010) found AIF-S signatures (-0.55 to +1.25‰) in 1.9 Ga black shales near the Outokumpu massive sulfide deposits in Finland. AIF-S signatures have also been found in present-day sulfate aerosols which were probably produced by burning of pyritic coal: Δ³³S values of -1.37 to +1.84‰ for Beijing aerosols (Ding et al., 2006) and 0 to 0.47‰ for California-Los Angeles aerosols (Romero and Thiemens, 2003). Sulfate associated with some volcanic ashes in ice cores are found to show AIF-S signatures: Δ³³S = -0.50 to +0.67‰ (Savarino et al., 2003) and -1.03 to +1.41‰ (Baroni et al., 2007).  
• The existence of general correlations between AIF-S signatures and lithology of sedimentary rocks: Pyrite crystals in many Archean-age shales have been found to show no AIF-S signatures (Farquhar et al., 2000; Whitehouse et al., 2005; Ohmoto et al., 2006; Ono et al., 2006; Farquhar et al., 2007); this is difficult to explain if the S came from the atmosphere. Pyrites with no AIF-S are generally hosted in C_org-poor shales (Watanabe et al., 1997; Ohmoto et al., 2006). Most pyrite crystals with AIF-S signatures occur in organic C-rich black shales (C_org = 1-15 wt%) that have been hydrothermally altered (Ohmoto et al., 2006; Kaufman et al., 2007; Watanabe et al., 2008). The existence of positive correlations between contents of pyrite and C_org, suggests that pyrite was generated by H₂S derived from reduction of sulfate by organic matter, either through bacterial sulfate reduction (BSR), or thermochemical sulfate reduction (TSR).
• SO₂-poor Archean volcanic gas: If the Archean continental mass was much smaller than today, most of the mantle degassing would have occurred through submarine hydrothermal systems where the dominant S-bearing specie was H₂S, rather than SO₂ (Kump and Barley, 2007). Also H₂S would have been more abundant than SO₂ in subaerial volcanic gas, if the Archean mantle beneath the continental crust was not as oxidized as today (Ohmoto, 2009). Because UV photolysis of H₂S does not produce AIFS (Farquhar et al., 2000b) volcanic outputs during the Archean likely did not acquire AIFs signatures in the atmosphere.

Recognition of the above problems in the SO₂-UV-photolysis model led Watanabe et al. (2009) to hypothesize that AIF-S signatures in sedimentary rocks were generated during reactions involving organic matter and SO₄^2- under hydrothermal conditions. To test this hypothesis Watanabe et al. (2009; in prep.) conducted a series of experiments of TSR at 150-200 °C using powdered amino acids, solid sulfate, and water. The experiments produced reduced compounds (H₂S and Cr-reducible S compounds) with distinct AIF-S signatures, ranging in δ³⁴S values from -16.2 to +5.1‰ relative to the initial sulfate S, and Δ³³S values from +0.1 to +2.3‰ (Fig. 2). They have recognized that the AIF-S generation occurred in TSR experiments using simple amino acids (glycine and alanine), but not with complex amino acids (e.g., arginine) or with kerogen extracted from sediments (Watanabe et al., in prep.). Based on these experimental data and the correlations between AIF-S and the lithology of host rocks, Watanabe et al. (2009) suggested that the creation of AIF-S was favored prior to ~2.4 Ga when simple, reactive organic matter accumulated in large rift-related basins under large-scale submarine hydrothermal conditions. Therefore, according to their model, the AIF-S record of sedimentary rocks may be linked to the thermal and biological evolution of early Earth, not to the atmospheric evolution.

Concurrently with the experimental investigations by Watanabe et al. (2009; in prep), Lasaga et al. (2008) carried out a theoretical study on multiple sulfur isotope fractionations during heterogeneous reactions (i.e., chemisorption) between solid surfaces and gaseous (or aqueous) sulfur-bearing species. The results of their computations, which were based on quantum chemistry and ab-initio models, predicted that AIF-S signatures could be generated if the chemisorption energy was less than about ~30 kJ/mol and a discontinuity occurs in the number of bound energy levels for the adsorbed species (e.g., C_surface-SO₂ species on a graphite surface) of different sulfur isotopes (Fig. 3). The magnitude of AIF-S during a heterogeneous reaction increases with increasing temperature. Balan et al. (2009) computed the wavefunction probabilities of the bound and unbound species in a finite-size box where the boundary wall was set at 30 Å from a solid surface and found that significant overlaps between the two wavefunction probabilities existed, and thereby concluded that chemisorption processes would not produce AIF. However, in a laboratory or natural system, the boundary wall typically occurs more than 1 µm (>10⁴ Å) away from a solid surface. Lasaga et al. (in prep) have shown that in such systems, the unbound states (i.e., free molecules, such as SO₂) are distributed equally throughout such a system and the overlap of bound and unbound wavefunctions is zero. Thus, they conclude that the suggestion by Lasaga et al. (2008) of the importance of chemisorption for the creation of AIF-S is valid.

 

1.2. Objectives

Considering the theory of chemisorption isotope effects, we hypothesized that the TSR in the Watanabe et al. experiments proceeded through the following three steps: (1) adsorption of aqueous SO₄^2- on the surfaces of C-bearing solid compounds; (2) reduction of the adsorbed SO₄^2- by C (and H and N) atoms to form reduced S-bearing clusters in the solid phase; (3) reactions between the reduced S-bearing clusters and H₂O to form H₂S_(aq.

1. g) and (4) desorption of H₂S_{(aq, g)}. The theory by Lasaga et al. (2008) predicted that: (a) the AIF-S signatures may be generated during the first step; (b) the adsorbed SO₄^2- species may have positive δ³⁴S and Δ³³S values with respect to the initial SO₄^2-; (c) the second step may cause mass-dependent isotopic fractionation; and (d) the third step may cause AIF-S with negative δ³⁴S and negative Δ³³S values for the desorbed H₂ This study was conceived to test the above hypothesis by investigating the details of chemical and isotopic fractionations during reactions between a simple solid C-bearing compound (activated carbon) and a simple sulfur-bearing gaseous compound (SO₂) at elevated temperatures in a dry system. The results from this study were also expected to help us understand the conditions and isotopic characteristics of H₂S (and other S-bearing compounds) in petroleum, natural gas, and some ore-forming fluids.

AN EXPERIMENTAL INVESTIGATION OF MULTIPLE SULFUR ISOTOPE FRACTIONATIONS DURING HETEROGENEOUS REACTIONS BETWEEN SO₂ AND ACTIVATED CARBON

Sharing is caring!