Refining the extraction methodology of carbonate associated sulfate: Evidence from synthetic and natural carbonate samples

Chemical Geology 411 (2015) 36–48

Contents lists available at ScienceDirect

Chemical Geology journal homepage: www.elsevier.com/locate/chemgeo

Refining the extraction methodology of carbonate associated sulfate: Evidence from synthetic and natural carbonate samples Bethany P. Theiling ⁎, Max Coleman NASA Jet Propulsion Laboratory, California Institute of Technology, United States NASA Astrobiology Institute, United States

a r t i c l e

i n f o

Article history: Received 10 December 2014 Received in revised form 27 May 2015 Accepted 22 June 2015 Available online 25 June 2015 Keywords: Carbonate associated sulfate CAS Sulfur isotopes Pyrite oxidation Extraction Oxygen isotopes

a b s t r a c t Sulfur and oxygen isotope analyses of trace and whole mineral sulfate are valuable in investigating diagenetic processes and the microbial communities that produced them, seawater anoxia, and paleoclimate. Oxygen isotopes are particularly useful in that they may also record alterations to the original isotope ratio, be it from post-depositional processes or oxidation of sulfide minerals during the chemical extraction procedure. Here we rigorously test several common methodological procedures of extracting carbonate associated sulfate (CAS) for sulfur and oxygen isotope analyses in order to generate a method that will extract only the CAS, while preserving associated organic reduced sulfur and sulfides for analysis. The results of these experiments on synthetically generated carbonates demonstrate that our proposed protocol sufficiently removes all nonCAS sulfate and does not result in oxidation of included sulfides. Analytical reproducibility (standard deviation) of synthetic carbonates is 0.1‰ (1σ) for δ34S and 0.3‰ (1σ) for δ18O. Extractions of low pyrite, high organic matter geologic samples from the Monterey Formation across a range of facies types demonstrate a reproducibility (1σ) of 0.4‰–0.7‰ for δ34S and 0.8‰–1.3‰ for δ18O, resulting from sample heterogeneity. δ34S and δ18O from Monterey Formation samples do not demonstrate oxidation of organic matter, suggesting our proposed protocol will preserve organic sulfur. A high pyrite-concentration Jet Rock concretion demonstrates that additional sulfate can be produced during the non-CAS leaching processes from oxidation of pyrite. We show that pyrite from the Jet Rock concretion ceases to oxidize when the sample is leached under an anoxic environment. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Marine δ34S and δ18O are powerful parameters for characterizing seawater anoxia, perturbations in the global sulfur cycle, and productivity of microbial communities (e.g. δ34S: Thode et al., 1951; Burdett et al., 1989; Strauss, 1999; Hurtgen et al., 2002; Kamschulte and Strauss, 2004; Paris et al., 2014; Rennie and Turchyn, 2014a, δ18O: Claypool et al., 1980; Newton et al., 2004; Brunner et al., 2005; Bottrell and Newton, 2006; Rennie and Turchyn, 2014a). δ34S and δ18O of seawater sulfate are particularly useful in that the target isotope systems are coeval. Due to the long residence time of sulfate in seawater (~20 My) with respect to the oceanic mixing time (~1000 year), seawater sulfate will reflect a well-mixed system (Holland, 1984; Paytan et al., 1998), and these residence/mixing times are considered stable throughout much of the Phanerozoic. Exceptions to this stable system include the latest Permian/early Triassic (Luo et al., 2010) and the Early Jurassic (Newton et al., 2011), both demonstrating very low seawater sulfate concentrations that significantly decreased typical seawater sulfate ⁎ Corresponding author at: Department of Geosciences, University of Tulsa, 800 Tucker Drive, Tulsa, OK 74101, United States. E-mail address: [email protected] (B.P. Theiling).

http://dx.doi.org/10.1016/j.chemgeo.2015.06.018 0009-2541/© 2015 Elsevier B.V. All rights reserved.

residence times. Perturbations in seawater sulfate δ34S specifically will reflect changes in the global sulfur cycle, and characterize local productivity of microbial communities in the depositional environment and post-burial alteration, all of which will strongly fractionate seawater sulfate δ34S. Seawater sulfate δ18O will reflect changes in riverine input and global climate regimes, as well as local bacterial or inorganic sulfide reoxidation (Fritz et al., 1989; Van Stempvoort and Krouse, 1994; Krouse and Mayer, 2000; Turchyn and Schrag, 2004). Given the power of seawater sulfate isotopes to characterize the cycle of weathering of continental material, burial, and preservation, there is a strong need to differentiate clearly between primary and diagenetic seawater sulfate δ34S and δ18O. Early studies of paleoseawater sulfate focused on evaporite deposits (gypsum, barite, etc.) (e.g. Claypool et al., 1980; Nielsen, 1989; Holser, 1992). Due to the poor spatial and temporal continuity of evaporites and their preferential deposition in shallow water, evaporitic deposits are confined to studies characterizing oxic systems or tracking turnover of stratified systems (Strauss, 1997). Carbonate associated sulfate (CAS), which is structurally bound in the calcite lattice and substitutes for carbonate (Takano, 1985; Staudt et al., 1994), is pervasive spatially and temporally in marine deposits. Burdett et al. (1989) demonstrate a remarkable similarity between seawater sulfate δ34S (20.80‰ ± 0.36‰)

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

and foraminiferal sulfate (20.64‰ ± 0.39‰), suggesting that CAS is an accurate reflection of seawater sulfate. Recent analyses of CAS using multicollector inductively coupled plasma mass spectrometry (MC– ICP–MS) demonstrate extraordinary accuracy for CAS δ34S (~0.1‰) on extremely small samples (nmol) of foraminiferal calcite (Paris et al., 2014). Biogenic carbonate shells, including foraminiferal calcite, are ideal given their high concentrations of sulfate (hundreds to thousands of ppm as opposed to tens of ppm for whole rock limestones) (Burdett et al., 1989; Kampschulte et al., 2001; Paris et al., 2014), yet have a few drawbacks. First, these recent advances using MC–ICP–MS cannot differentiate between δ18O from carbonate versus δ18O from sulfate. Therefore, carbonate shell δ18O analyzed via MC–ICP–MS will reflect some combination of seawater and seawater sulfate δ18O. Second, relying on the presence of carbonate shells imposes a limitation to the depositional environments in which these shells are generated. For example, if a foraminifera species only calcifies under oxic conditions, foraminiferal CAS from an anoxic depositional environment will only record the global seawater sulfate δ34S, and will omit the isotopic fractionations associated with anaerobic microbial ecosystems of the depositional environment. Therefore, only foraminifera that can calcify under anaerobic conditions will be useful for describing the depositional environment and early burial conditions. To date, only a few foraminifera have been shown to calcify under anoxic conditions, carefully controlled in a laboratory (Nardelli et al., 2014). As a result, whole rock CAS δ34S and δ18O are the most versatile in application, being useful from abyssal to coastal and lagoonal systems (Hurtgen et al., 2002; Kamschulte and Strauss, 2004; Loyd et al., 2012; Wotte et al., 2012a,b). One of the greatest limitations of whole-rock CAS analyses of δ34S and δ18O lies in isolating primary and diagenetic values from isotopic artifacts generated during analytical extraction of CAS. Bacterial sulfate reduction will increase δ34S and δ18O of CAS substrate during early diagenesis due to preferential uptake of 32S and 16O by sulfate reducing bacteria (e.g. δ34S: Thode et al., 1951; Strauss, 1997, δ18O: Mitzutani and Rafter, 1973; Fritz et al., 1989; Brunner et al., 2005; Antler et al., 2013). However, oxidation of sulfides (e.g. H2S, pyrite, galena, sphalerite, marcasite), whether by microbial oxidation or inorganic oxidation, generate lower δ34SCAS and may increase or decrease δ18OCAS values (Taylor et al., 1984; Böttcher and Thamdrump, 2001). Additional nonCAS sulfate can be introduced to surface outcrops and may produce a distinctive positive Δ17O (Peng et al., 2014). Oxygen isotope compositions of CAS therefore serve as a measure of both the primary and early diagenetic conditions, as well as an additional check on possible sample contamination or isotopic artifacts generated during analytical extractions. The typical CAS extraction procedure follows three phases: 1) leaching of soluble, non-CAS sulfur compounds, 2) dissolution of carbonate minerals to liberate lattice-bound SO2− 4 , and 3) precipitation of liberated SO2− 4 as BaSO4. Therefore, it is imperative first to remove all soluble, non-CAS sulfate, then remove associated sulfides without oxidizing sulfides to sulfate. Sulfides can oxidize via the following reactions (shown for pyrite): FeS2 þ 3:5O2 þ H2 O→ Fe2þ þ 2SO4 2− þ 2Hþ

ð1Þ

Fe2þ þ Hþ þ 0:25O2 → Fe3þ þ 0:5H2 O

ð2Þ

FeS2 þ 14 Fe3þ þ 8H2 O→15 Fe2þ þ 2SO4 2− þ 16Hþ

ð3Þ

where Fe2 + production by microbially mediated sulfide oxidation (Eq. (1)) promotes further inorganic oxidation of sulfides (Eq. (3)) when Fe2+ is oxidized to Fe3+ (Eq. (2)). Sulfides can also be inorganically oxidized (Eq. (3)), which may be more common in laboratory processing of samples. Of these reactions, microbial sulfide oxidation, initiated by the use of atmospheric oxygen as the oxidant, is orders of magnitude more rapid as a process for generating sulfate (Eqs. (2) and (3)) (Nordstrom and Southam, 1997). If associated sulfides are oxidized

37

during extraction, δ34S will reflect a mixture of δ34SCAS and δ34Ssulfide, whereas δ18O will reflect mixing between δ18OCAS and δ18O derived from the extraction procedure (Eq. (3)). In their pioneering study, Burdett et al. (1989) used sodium hypochlorite (NaOCl) to remove organic and non-CAS sulfate and sulfides (e.g., organic sulfur, sulfides, sulfate salts). This method has been utilized in many CAS extraction methodologies, with several modifications. The leaching step to remove non-CAS sulfur and sulfides varies from soaking in NaOCl, NaCl, H2O2, rinsing in deionized water for 12–24 h, or a combination of all four (Burdett et al., 1989; Ohkouchi et al., 1999; Kampschulte et al., 2001; Hurtgen et al., 2002; Fike et al., 2006; Fike and Grotzinger, 2008; Marenco et al., 2008; Gill et al., 2011; Shen et al., 2011; Loyd et al., 2012; Xiao et al., 2012). NaOCl is used to remove organic sulfur and H2O2 to remove pyrite (Eqs. (1) and (3)) (Shen et al., 2011). However, in many cases, we wish to preserve the organic sulfur and associated sulfides so that they may be analyzed from the insoluble residue resulting from CAS extraction. As a result, several researchers have adopted the use of a NaCl solution, which will not remove organic sulfur or pyrite, but will leach soluble sulfate minerals such as gypsum, epsomite, and surface-adsorbed sulfate. Still others use deionized water as a leachate solution. However, Peng et al. (2014) has recently shown that deionized water will not adequately leach non-CAS phases. Therefore, the difficulties in preserving the sulfides, organic sulfur, and CAS are to a) remove all non-CAS sulfate-bearing phases without adding to the sulfate pool and b) avoid oxidizing the remaining sulfides, which will produce sulfate. As an additional measure, some researchers attempt to reduce the possibility of pyrite oxidation to sulfate (Eqs. (1) and (3)) by varying the strength of HCl used to liberate (e.g. Burdett et al., 1989; Newton et al., 2004; Fike lattice-bound SO2− 4 et al., 2006; Marenco et al, 2008). Likewise, some researchers have completed acidification under anoxic or N2 conditions to further reduce the possibility of oxidation (Fike et al., 2006; Fike and Grotzinger, 2008) or by adding HCl slowly using a dropping funnel in an unsealed vessel (Marenco et al., 2008). A few recent studies have added SnCl2 during the acidification step to reduce associated Fe3+ that may oxidize pyrite (Planavsky et al., 2012; Rennie and Turchyn, 2014a) based on the experimental monosulfide extractions of Chanton and Martens (1985). If sulfides are oxidized during CAS extraction, δ34S will reflect a mixture of sulfide and CAS δ34S. Similarly, oxidation of sulfides during extraction will produce δ18O values that reflect a mixture of CAS δ18O and atmospheric or water δ18O (with associated isotopic fractionations) used in the leaching and/or dissolution steps (Eq. (3)). Our goal is to identify the extraction techniques that will preserve organic sulfur and sulfides in the resulting residue, while maintaining the original δ34S and δ18O of CAS. Therefore, here we test the most common methods of CAS extraction on synthetically generated CAScarbonates. We compare synthetic CAS that are leached and unleached using a NaCl solution, with and without the presence of pyrite, and under atmospheric and N2 headspace conditions. By testing various methods on synthetic CAS-carbonates in which pyrite concentrations can be controlled, we establish a refined protocol for CAS extraction. In addition, we compare δ34S and δ18O analyses from the proposed procedure to samples in which we anaerobically and abiotically forced associated pyrite to oxidize. We evaluate the purity of several BaSO4 precipitates by comparing chromatography and energy dispersive x-ray spectroscopy (EDX). Our proposed protocol is then applied to several natural, geologic samples from different facies types of the midMiocene Monterey Formation to test the reproducibility of extraction and analysis of natural samples. 2. Development of analytical methods 2.1. Synthetic CAS precipitation Synthetic CAS-carbonates were generated for these experiments so that leaching and CAS extraction techniques could be compared based

38

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

on the presence and absence of pyrite. Synthetic carbonates were precipitated from anhydrous compounds via the following reaction: NaHCO3 þ CaCl2 →NaCl þ CaCO3 þ Hþ : Na2SO4 was added to the NaHCO3 solution before addition of CaCl2 to promote incorporation of SO24 − into the resulting CaCO3 lattice. Batches of CAS CaCO3 were generated for each experiment using the same stock chemicals to minimize fluctuations in synthetic CAS isotopic ratios. The molarities of Na2SO4, NaHCO3, and CaCl2 for each experiment are shown in Table 1 and produced an excess of Na2+ in solution. 2.2. Non-CAS leaching and CAS extraction Experiment variables are compiled in Table 2. Each batch of experiments contained at least one sample in which no pyrite was added to the CAS precipitate, to serve as a reference for % SO24 −, δ34S and δ18O in the precipitate and a frame of reference by which pyrite-added samples are compared. Powdered NIST 8455 (bioleaching pyrite ore) was used as a pyrite additive for these experiments, whose grain sizes are b0.5 mm, common in marine systems. Pyrite concentrations for these experiments ranged from ~25 ppm to ~1000 ppm and were run in replicate to determine the reproducibility of the experimental results (Table 2). Due to the higher possibility of oxidation by NaOCl and H2O2, removal of non-CAS sulfates in these experiments was completed by constantly stirring the sample in a 10% NaCl solution in 18 MΩ de-ionized water for 24 h. Non-leached samples were extracted for comparison. The leaching step used 10 mL of 10% NaCl/g of sample. Samples were then either filtered through Whatman 45 μm cellulose nitrate membrane filter papers or centrifuged after each leaching step and rinsed thoroughly with 18 MΩ water. Samples were then dried and weighed to determine the percentage of original sample lost during the leaching step. Dissolution of all samples was performed using 10% HCl in 18 MΩ water (Table 2). To vary the possibility for oxidation, dissolution was completed under one of three atmospheric conditions. A subset of samples was dissolved under normal atmospheric conditions in open beakers, with HCl added dropwise via pipette. Another subset was dissolved in Erlenmeyer flasks using a constant N2 purge in the flask headspace. In the third subset, flask headspace was only initially purged with N2. For experiments using N2 purge and headspace, the system was kept closed using a bubbler apparatus to release CO2 gas evolved during dissolution (Fig. 1). HCl in both the N2 flush and N2 headspace setup was added dropwise via a dropping flask. The dropping flask was closed before addition of the final ~0.5–1 mL of HCl to prevent atmosphere from entering the sample flask. Samples were left for 1.5 h to complete dissolution. Samples were typically dissolved in HCl in a ratio of 1:15 (g/mL). After dissolution, sample flasks were opened to atmosphere and samples were either centrifuged or filtered, depending on the experimental variables (Table 2). The filtrate was heated to 75°–80 °C under atmospheric conditions and 8.5% BaCl2 in a volume of approximately 10% of the filtrate volume was added dropwise using a pipette. The resulting solution was allowed to sit at this temperature for 3 h, then cooled to room temperature to crystallize BaSO4 for at least 72 h. We do not add ammonia solution to the filtrate before BaSO4 precipitation to produce an alkaline solution that will precipitate iron because we would not expect appreciable soluble iron in samples after leaching,

Table 1 Molarity of compounds in 18 MΩ deionized water used to generate synthetic CaCO3-CAS and mass of CaCO3-CAS produced. Synthetic CaCO3-CAS stock #

Molarity CaCl2

Molarity NaHCO3

Molarity Na2SO4

Mass CaCO3-CAS produced (g)

1 2

0.551 0.551

0.232 0.331

0.063 0.063

15.11137 15.44985

and because phosphate, a possible contaminant, is insoluble in acid solutions and soluble in alkaline solutions (especially true of barium phosphate).

2.3. Forced pyrite oxidation After comparing the reproducibility of each method and characterizing impurities in the precipitate and isotopic fractionation between extraction methods (Table 2), we generated a preferred protocol for non-CAS sulfate leaching and CAS extraction (see Discussion). It was important to test whether the pyrite additive was oxidized at any stage using our protocol. As a result, we compared extractions of synthetic CAS, CAS with a pyrite additive, and CAS with pyrite that we attempted to oxidize during extraction. Although additional sulfate may be produced from pyrite either by microbial oxidation with molecular oxygen (Eq. (1)) or inorganically by Fe3+ (Eq. (3)) we chose the latter as more easily controllable in the laboratory. Therefore, anaerobic and abiotic oxidation of the pyrite was performed by adding 14 times the stoichiometric equivalent mass of solid FeCl3 · 6H2O as the initial pyrite mass to acidified samples (as defined by Eq. (3)), simulating the isotopic mixing between CAS and sulfate generated from pyrite oxidation if the protocol fails.

2.4. Geologic samples To test the reproducibility of our suggested protocol in natural samples, we chose samples from the mixed mineralogy Monterey Formation (Fm) at Naples Beach (Fig. 2), deposited on a low-gradient deep water slope during the middle Miocene (Issacs et al., 2001), and one high pyrite concentration sample from the early Jurassic Jet Rock Formation (≥30% pyrite; Raiswell, 1976). The Monterey Formation is a lithologically diverse hemipelagic to pelagic succession is rich in organic matter, carbonate, phosphate, and biosilica (e.g. Bramlette, 1946; Garrison and Douglas, 1981; Barron, 1986; Galloway, 1998; Behl, 1999; Issacs et al., 2001; Föllimi et al., 2005). Three facies were chosen from this succession based on the stratigraphy of Föllimi et al. (2005) to represent a wide range of sedimentological characteristics and concentrations of carbonate, organic matter, pyrite, and phosphate: phosphatic “sand”, red mudstone, and carbonate concretion. The Jet Rock Formation is a Toarcian age sedimentary deposit and comprises fine-grained, well-laminated organic-rich shales in which there are many horizons containing pyrite-rimmed carbonate concretions (Howarth, 1962; Raiswell, 1976). Due to the large insoluble component of both the Monterey and Jet Rock sediments, acidified samples were first centrifuged and the supernate was filtered to ensure the most accurate calculation of % carbonate and % insolubles. Loyd et al. (2012) report b 0.2 wt.% pyrite for samples from this locality.

2.5. Yield measurement and precipitate purity The concentration of CAS from analyzed samples is calculated as parts per million (ppm) CAS relative to the total original sample mass (Tables 2–5). During isotopic analysis, samples are converted to SO2 gas by combustion and measured as a peak area in volt · seconds (V · s) by the mass spectrometer. We define the slope and intercept of the line defined by the chromatographic peak area (in V · s) versus the mass SO2− 4 for a range of masses of BaSO4 standard, assuming that the standard is stoichiometrically 100% BaSO4. We calculate a theoretical mass of SO2− 4 for each sample analyzed based on the mass of the precipitate analyzed and use the equation produced by analyses of the BaSO4 standard to calculate the actual mass of SO2− 4 in the sample precipitate that is combusted and measured in the same manner as the standard. The precipitate 2− purity is calculated as % SO2− 4 sample/% SO4 stoichiometric BaSO4 ∗ 100%.

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

39

Table 2 Analytical data from experiments 5–7, in which we compare precipitate yield and purity, δ34S, and δ18O data generated by CAS-only and pyrite-added samples using variable methods of CAS extraction. Headspace Sample Mass CaCO3 Mass FeS2 NaCl CAS (g)a (mg)b leaching conditions ID

Precipitate δ34S δ18O Total mass Precipitate Mass BaSO4 Mass BaSO4 ppmc SO2− 4 precipitate yield (%)d for δ34S (mg) purity (%)e (‰, VCDT) for δ18O (mg) (‰, VSMOW) generated (g)

5-1

1.00836

0

Yes

N2 flush

0.00897

2968

0.9

5-2 5-3 5-4 6-1

1.00478 1.02653 1.00465 1.0024

0.0184 0.1737 1.0287 0

Yes Yes Yes Yes

N2 flush 0.00021 N2 flush 0.00056 N2 flush N2 headspace 0.00448

86 225

b0.05 0.1

1250

0.4

6-2 6-3 7A 7B

1.00497 1.00482 0.75677 0.75625

0.0274 0.0976 0 0

Yes Yes Yes No

N2 flush N2 headspace Atmosphere Atmosphere

0.00069 0.0008 0.00052 0.04988

291 155 283 21,258

0.1 0.1 0.1 6.6

7C 7D

0.75563 0.75582

0 0

Yes No

N2 flush N2 flush

0.00015 0.04765

82 17,130

b0.05 6.3

7E 7F

0.75649 0.75879

0 0

Yes No

N2 headspace 0.00045 N2 headspace 0.04483

245 23,722

0.1 5.9

7G 7H

0.75372 0.75778

0.1499 0.1499

Yes No

Atmosphere Atmosphere

0.00062 0.05029

339 20,735

0.1 6.6

7I

0.75579

0.1546

Yes

N2 flush

0.00298

1050

0.4

7J

0.75456

0.1514

No

N2 flush

0.04777

17,965

6.3

7K 7L

0.75533 0.75901

0.1509 0.1515

Yes No

N2 headspace 0.00177 N2 headspace 0.04876

527 21,008

0.2 6.4

7M

0.7103

0.1516

Yes

N2 headspace 0.00133

798

0.2

0.5726 0.5575 0.5663

91.6 98.2 53.2

2.9 3.4 1.1

0.1538

16.4

0.1645

17.8

0.5068 0.4797 0.5306 0.0930 0.2147

37.9 81.4 84.4 102.9 47.2

1.5 2.6 2.8 3.0 1.9

0.1296 0.1730

16.9 17.2

0.4959 0.5286 0.5591

91.5 91.9 51.3

0.6 0.8 −0.3

0.1710 0.1383

12.5 13.2

0.4820 0.4695 0.4848

58.6 77.0 62.2

0.4 0.8 0.3

0.1464 0.1362

14.3 15.1

0.4947 0.4895 0.4948

89.8 108.7 93.9

0.8 2.2 1.1

0.1598 0.1539

13.2 14.1

0.5740 0.5801 0.5086 0.4916 0.4722 0.5434 0.5223 0.4848 0.5159 0.5292 0.5846 0.5085 0.4038

95.3 60.7 71.5 45.1 84.2 93.5 62.3 50.8 54.6 67.5 113.8 56.8 103.5

0.8 0.3 −0.5 1.2 2.5 0.8 0.1 −1.0 2.0 0.2 1.2 −0.4 2.9

0.1366 0.1469

13.5 13.6

0.1269 0.1554 0.1706 0.1337

15.1 15.0 13.7 14.2

0.1285 0.1773 0.1492

16.3 14.1 13.8

a

Standard error on mass of synthetic carbonates is 0.5% relative to the desired sample mass. b Standard error on mass of pyrite additive is 0.4% relative to the desired mass. c d 2− ppm SO2− 4 calculation incorporates the measured precipitate purity, where data is available. When purity is not available, ppm SO4 is a maximum estimate, assuming 100% stoichiometric BaSO4 precipitate. d Precipitate yield = mass BaSO4 produced/mass original powdered sample *100%. e Precipitate purity = % SO2− of BaSO4 calculated by normalizing to chromatographic peak area (description in text). 4

2.6. Isotopic analyses Isotopic analyses of sulfur were performed by weighing 0.550 mg (+/− 0.050 mg) BaSO4 into tin capsules, which were then combusted using a Costech Elemental Combustion System (Valencia, CA, USA) and the resulting SO2 gas was measured via CF-IRMS using a Thermo MAT 253 mass spectrometer (Bremen, Germany) at the Jet Propulsion Laboratory, California (JPL). In the case of low precipitate yield, we attempted to analyze samples as low as 0.093 mg with internal BaSO4 standards of similarly low mass for comparison. δ34S is reported in δ notation in per mil (‰) relative to Vienna Canyon Diablo Troilite (VCDT). Reproducibility (1σ standard deviation) of δ34S measurements was monitored each run by the isotopic ratios of laboratory standards included in the sample run; BaSO4, Ag2S, and elemental S, measured at least in triplicate over the course of each run (1σ standard deviation ≤ 0.3‰). Laboratory standards are calibrated to national and international standards NBS 127, IAEA S-2, and IAEA S-1, each having a 1σ standard deviation of ≤ 0.2‰ for each run. Oxygen isotopic analyses were performed on 0.150 mg (+/− 0.050 mg) BaSO4 weighed into silver capsules, reduced by glassy carbon in a Thermo Finnigan High Temperature Conversion Elemental Analyzer (TC/EA) and measured via CF-IRMS on a MAT 253 mass spectrometer (Bremen, Germany) (Kornexl et al., 1999). δ18O is reported in

δ notation as ‰ relative to Vienna Standard Mean Ocean Water (VSMOW). Reproducibility (1σ) of δ18O measurements was checked each run using at least three replicate analyses each of the laboratory

CO2 release

dropping funnel

water level N2 gas

Erlenmeyer flask Fig. 1. Apparatus for dissolution of samples with an N2 headspace. See text for additional description.

40

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

A

B 166

166

1

ifo

l Ca rn

101

Santa Barbara

ia

Naples Beach

5

1 101

Los Angeles 1

405

Fig. 2. Location of the Naples Beach outcrop of the Monterey Formation.

standard BaSO4 and national and international BaSO4 standards NBS 127 and IAEA SO-6, with a reproducibility (1σ) of ≤ 0.5‰. δ18O of our laboratory BaSO4 standard was calibrated using IAEA 601, NBS 127 and IAEA SO-6, with a standard deviation of ≤ 0.5‰. 2.7. ESEM and EDAX analyses We evaluated precipitate purity for several samples to test whether calculation of purity using chromatograph peak area was an accurate reflection of purity. Therefore, we utilized images from a Perkin Elmer environmental scanning electron microscope (ESEM) and standardless energy dispersive x-ray spectroscopy (EDX) analyses of several BaSO4 precipitates to both semi-quantitatively verify calculated purity from chromatography at JPL and chemically and/or physically characterize those precipitates that did not produce SO2 peaks. The precipitate was imaged and analyzed on an aluminum stage, rather than adhering to carbon tape, to assess possible carbon content. EDX analyses are

reported in weight percent (wt. %) of a given element, where the sum of all elements analyzed is normalized to 100%. Precipitates analyzed include two synthetic samples measured via chromatography as ~ 100% purity, two from synthetic samples of less than 100% purity, one from a synthetic sample in which an oxidant was used during extraction (experiment 9), two from geologic samples calculated as ~100% purity, one from a geologic sample b 100% purity, and one geologic sample that produced a red–brown precipitate with no measurable SO2 peak (therefore calculated at 0% purity). Due to the mixed and unknown mineralogy of natural samples, additional elements such as C, Na, Mg, Si, P, Cl, Ca, Mn, and Fe were monitored for Monterey Fm precipitates. 3. Results All samples are described in terms of the replicate mean ± 1σ standard deviation, when replicates were available. If no replicates were

Table 3 Analytical data from experiment 8, in which we compare precipitate yield and purity, δ34S, and δ18O data generated by CAS-only and pyrite-added samples using our suggested protocol. All samples in experiment 8 are leached with 10% NaCl and dissolved under N2 headspace. Sample ID

Mass CaCO3 CAS (g)a

Mass FeS2 (mg)b

Total mass precipitate generated (g)

ppmc SO24

Precipitate yield (%)d

Mass BaSO4 for δ34S (mg)

Precipitate purity (%)e

δ34S (‰, VCDT)

Mass BaSO4 for δ18O (mg)

δ18O (‰, VSMOW)

8A

2.50446

0

0.04608

988

1.8

8B

2.50276

0

0.03688

828

1.5

0

0.03992

912

1.6

0.1242 0.0956 0.1272 0.1061

15.5 13.3 13.4 13.2

8D

2.50095

0

0.03991

764

1.6

0.1231 0.0989 0.1085

13.9 13.7 15.2

8E

2.50045

0.3059

0.0452

973

1.8

8F

2.50008

0.3007

0.0412

880

1.6

0.0971 0.1028 0.1033 0.1174 0.1163

13.4 13.0 14.2 14.2 14.4

8G

2.50089

0.3009

0.03636

745

1.5

8H

2.50016

0.2935

0.03922

863

1.6

1.1 0.9 0.9 1.2 1.2 1.0 1.1 1.0 1.3 1.2 1.3 1.2 0.8 1.1 1.3 1.1 1.2 1.1 0.8 1.0 1.3 0.9 1.2 1.2 0.9

14.6 14.2 14.0

2.5013

98.2 85.3 96.2 96.2 95.8 99.3 96.5 96.3 98.3 99.0 96.5 99.2 81.9 91.3 92.7 94.6 92.7 96.9 80.7 94.8 90.1 77.7 99.0 97.6 85.5

0.1237 0.1428 0.1069

8C

0.5061 0.5251 0.4981 0.4911 0.5134 0.5076 0.4919 0.4905 0.5243 0.4952 0.5126 0.4877 0.492 0.4917 0.4958 0.4887 0.5005 0.4999 0.5081 0.4968 0.5016 0.4958 0.4983 0.5093 0.5004

0.1215 0.1396 0.1047 0.1099 0.1150 0.0957

14.4 14.8 13.6 14.4 14.5 13.4

a

Standard error on mass of synthetic carbonates is 0.5% relative to the desired sample mass. Standard error on mass of pyrite additive is 0.4% relative to the desired mass. d 2− ppm SO2− 4 calculation incorporates the measured precipitate purity, where data is available. When purity is not available, ppm SO4 is a maximum estimate, assuming 100% stoichiometric BaSO4 precipitate. d Precipitate yield = mass BaSO4 produced/mass original powdered sample *100%. e Precipitate purity = % SO2− of BaSO4 calculated by normalizing to chromatographic peak area (description in text). 4 b c

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

41

Table 4 Analytical data from experiment 9, in which we promote oxidation of included pyrite using FeCl3 · 6H2O. All samples in experiment 9 are leached with 10% NaCl and dissolved under N2 headspace before adding oxidant. Sample ID

Mass CaCO3 CAS (g)a

Mass FeS2 (mg)b

Total mass precipitate generated (g)

ppmc SO24

Precipitate yield (%)d

Mass BaSO4 for δ34S (mg)

Precipitate purity (%)e

δ34S (‰, VCDT)

Mass BaSO4 for δ18O (mg)

δ18O (‰, VSMOW)

9A

2.50359

0.3002

0.04514

2893

1.8

0.4906

94.5

1.0

9B

2.50717

0.3045

0.04171

2564

1.7

2.50669

0.3026

0.04664

2763

1.9

9D

2.50194

0.2992

0.03616

2179

1.4

90.2 92.4 97.1 83.5 92.2 92.8 77.6 91.9 91.3 93.7 78.4

1.0 1.0 1.1 0.9 1.0 1.1 0.7 1.2 1.5 1.4 0.7

13.1 6.7 14.3 12.7 14.8 14.1

9C

0.5171 0.5031 0.5080 0.4911 0.4974 0.5029 0.4854 0.4904 0.5098 0.4881 0.5080

0.1045 0.1280 0.1080 0.1010 0.1075 0.1057 0.1179 0.1219 0.0999 0.1049 0.1192 0.0952

11.0 14.2 14.3 10.8 14.7 12.7

a

Standard error on mass of synthetic carbonates is 0.5% relative to the desired sample mass. Standard error on mass of pyrite additive is 0.4% relative to the desired mass. ppm SO2− calculation incorporates the measured precipitated purity, where data is available. When purity is not available, ppm SO2− is a maximum estimate, assuming 100% 4 4 stoichiometric BaSO4 precipitate. d Precipitate yield = mass BaSO4 produced / mass original powdered sample ∗ 100%. e Precipitate purity = % SO2− of BaSO4 calculated by normalizing to chromatographic peak area (description in text). 4 b c

available, the sample is noted as a single value. We discuss precipitates from experiments 8 and 9 separately because experiments 8 and 9 were used to check the analytical uncertainty of our proposed method. δ34S from Na2SO4 used to generate CaCO3-CAS average − 2.4‰ ± 0.1‰. δ18O for Na2SO4 used to precipitate CaCO3-CAS averages 12.7‰ ± 2.0‰.

3.1. Leached vs. unleached CAS concentrations average 20,303 ppm for unleached samples and 638 ppm for leached samples (Fig. 3, Table 2). Precipitate purity for leached samples average 87.6% ± 15.2% and unleached samples average 77.6% ± 19.9%. All except one unleached sample demonstrate lower δ34S values than leached samples (Fig. 5, Table 2). δ34S for unleached samples is 0.6‰ ± 0.3‰, while δ34S for leached samples is 1.4‰ ± 0.6‰. δ18O for unleached samples average 13.6‰ ± 0.7‰. δ18O for leached samples is 14.1‰ ± 1.8‰.

3.2. CAS only vs. pyrite added Purity could not be measured for the test CAS-only precipitates due to the low mass of precipitate generated. Precipitate purity of pyriteadded samples is 71.9% ± 26.9% (Table 2, Fig. 5). δ34S for samples with pyrite average 0.8‰ ± 1.2‰ and without pyrite average 0.5‰ ± 0.4‰. δ18O for samples without pyrite average 13.7‰ ± 1.0‰ and 14.4‰ ± 0.9‰ for samples with added pyrite.

3.3. Headspace experiments Purity of precipitates generated by dissolution under atmospheric conditions is 77.0% ± 18.6%, whereas N2 flush and N2 headspace precipitates are 73.3% ± 20.3% and 78.3% ± 25.2%, respectively (Table 2, Fig. 5). Atmospheric headspace conditions produce the lowest isotopic ratios of 0.3‰ ± 0.6‰ for δ34S and 13.1‰ ± 0.7‰ for δ18O. δ34S for N2

Table 5 Analytical data from CAS extractions of geologic samples from the Monterey Formation, California. All samples are leached 1–3 times with 10% NaCl and dissolved under N2 headspace. # leaching steps

% acid soluble

Total mass precipitate generated (g)

ppma SO2− 4

67.3558 82.2407

3 3

98.1 75.8

0.00054 0.01098

3 43

82.8179

1

84.9

0.01513

52

Concretion Concretion Red mudstone

85.9141 101.7334 60.1700

3 1 3

2.7 0.0 13.0

0.00000 0.00000 0.01340

0 0 86

1301-11b

Red mudstone

50.3500

1

11.2

0.01022

76

1301-12

Phosphatic sand

49.9121

3

6.9

0.00957

80

1301-14 1301-14b

Phosphatic sand Phosphatic sand

32.9271 42.3943

3 1

20.9 3.4

0.00056 0.00083

7 8

Sample ID

Facies

1301-05 1301-07A

Concretion interior Concretion

1301-07Ab

Concretion

1301-10 1301-10b 1301-11

a b

Whole-rock powder mass (g)

Mass BaSO4 for δ34S (mg)

Precipitate purity (%)b

δ34S (‰, VCDT)

Mass BaSO4 for δ18O (mg)

δ18O (‰, VSMOW)

0.5376 0.6029 0.5693 0.5872 0.6099 0.5928

85.6 73.2 76.7 68.6 59.9 80.7

18.5 17.1 17.9 18.3 17.7 19.0

0.1451 0.1502 0.1498 0.149 0.1545 0.1445

13.7 13.8 12.3 12.4 12.8 10.9

0.5673 0.5304 0.5491 0.5691 0.5494 0.5542 0.5372 0.5553

90.1 100.2 92.1 85.0 82.1 107.3 108.1 94.3

11.5 12.5 12.6 14.1 13.4 15.6 13.4 12.9

0.155 0.1554 0.1494 0.1506 0.1565 0.1519 0.1407 0.1448 0.0904 0.1008

3.2 3.9 5.5 5.2 2.7 4.7 6.0 5.4

ppm SO2− = μg SO2− in BaSO4 produced/g original powdered sample. 4 4 Precipitate purity = % SO2− of BaSO4 calculated by normalizing to chromatographic peak area (description in text). 4

7.3

42

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

A

B

C

D

E

F

Fig. 3. Photomicrographs of synthetic precipitates. White horizontal scale bars in each figure represent 20 μm. A) Sample 8C, calculated at 98% purity (refer to Section 2.5 for purity determination). B) Sample 8D, calculated at 94% purity. C) Sample 9AS, in which added pyrite was forced to oxidize during acidification, calculated at 95% precipitate purity. D) Original synthetic CaCO3-CAS used for extraction experiments 5–9. E) Sample 7I, calculated at 65% purity. F) Sample 7 K, calculated at 55% purity.

flush are 1.8‰ ± 1.2‰ and for N2 headspace are 1.6‰ ± 0.9‰. δ18O values average 15.4‰ ± 1.6‰ for dissolution under N2 flush, whereas N2 headspace average 15.1‰ ± 1.7‰. 3.4. Precision and forced pyrite oxidation experiments Experiment 8 purity average 93.3% ± 6.3%. δ34S for experiment 8 is 1.1‰ ± 0.2‰, and δ18O is 14.1‰ ± 0.7‰ (Table 3, Fig. 6). Experiment 9 purity averages 89.6% ± 6.3%. δ34S values average 1.1‰ ± 0.2‰ for experiment 9. δ18O values for experiment 9 average 12.8‰ ± 2.4‰ (Table 4, Fig. 6). 3.5. Geologic samples A subset of Monterey Fm sample analyses was replicated to compare the amount of precipitate generated, purity, and isotopic ratios when leached once versus three times, shown in Table 5. Purity of 1301-07A is 78.5% ± 6.4% when leached three times and 69.8% ± 10.4‰ when

leached once. δ18O of 1301-07A samples leached three times is 13.3‰ ± 0.8‰, whereas 1301-07A leached once is 12.0‰ ± 1.0‰. δ34S for 1301-07A leached three times average 17.8‰ ± 0.7‰, and 18.3‰ ± 0.6‰ when leached once. No precipitate was produced from 1301-10 (concretion). The δ18O of these samples are 4.2‰ ± 1.2‰ for samples leached three times and 4.2‰ ± 1.3‰ for samples leached once. δ34S for 1301-11 is 12.2‰ ± 0.7‰ for samples leached three times and 14.4‰ ± 1.1‰ for samples leached once. 1301-14 generated chromatographic peaks below our instrument detection limits and was therefore discarded for δ34S analysis. δ18O for 1301-14 leached once yielded only enough precipitate for one analysis and is 7.3‰. 1301-05 and 1301-12 were extracted only after leaching three times each. Neither δ18O nor δ34S are measured for 1301-05 due to low purity based on EDX results (Table 6). 1301-12 (phosphatic sand) purity is 101.2% ± 9.7%. δ34S for 1301-12 is 13.2‰ ± 0.4‰ and for δ18O is 5.7‰ ± 0.4‰. One Jet Rock concretion sample was processed to determine whether our protocol can be used for samples with very high pyrite concentrations. The Jet Rock sample demonstrates significant

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

43

Table 6 EDX analyses for selected synthetic CaCO3-CAS extractions from experiments 7–9 and geologic samples from the Monterey Formation, California. Sample

wt.% O

wt.% Na

wt.% Mg

wt.% Si

wt.% P

wt.% S

wt.% Cl

wt.% Ca

wt.% Mn

wt.% Fe

wt.% Ba

Synthetic precipitates 7I 11.4 7K 28.6 8C 4.6 8D 3.5 9A 3.4

wt.% C

14.8 16.7 19.4 19.0 16.2

– 1.2 – – 1.4

– 1.4 0.8 – 0.8

– 0.9 – – 0.6

– 0.5 – – 0.5

12.6 9.3 13.8 13.2 13.0

2.0 1.7 1.0 1.0 2.1

0.5 0.3 0.2 0.3 –

– 0.7 – – 0.5

– 0.5 – – 0.5

55.3 41.4 61.0 63.5 61.0

Geologic precipitates 1301-05 25.2 1301-07A 4.6 1301-11 15.3 1301-12 5.1

31.7 18.8 19.9 17.3

0.6 – 2.0 1.5

0.9 1.0 3.6 1.7

18.4 2.1 2.8 1.2

10.7 2.4 0.7 0.7

1.5 12.5 11.1 12.8

0.7 1.4 1.4 1.6

0.4 0.3 – –

0.5 0.4 1.3 0.8

2.1 0.5 1.2 0.8

7.3 58.3 40.5 56.6

sulfate is produced while leaching under atmospheric conditions, and that this sulfate concentration decreased rapidly after the sample is leached under anoxic conditions (Table 7). 3.6. ESEM and EDX analysis of precipitates Synthetic and geologic samples calculated via chromatography to be ~100% purity exhibit tabular crystals arranged into b 10 μm diameter rosettes (Figs. 3A, B, C, 4C, D). EDX analyses are shown in Table 6. Several precipitates demonstrate the presence of additional phases, comprised of a range of elements, including carbon, sodium, magnesium, silicon, phosphorous, chlorine, iron, and manganese. Synthetic and geologic samples calculated as b 100% purity exhibit varied crystal habit ranging from single or aggregate tabular crystals to amorphous (Figs. 3E, F, 4A, B). EDX of b100% precipitates demonstrate more non-BaSO4 phases than ~100% purity precipitates (Table 6). 4. Discussion 4.1. Precipitate purity Comparison of purity calculated from chromatography with EDX analyses indicates that ~100% purity precipitates demonstrate a consistently tabular-rosette crystal habit with few additional, quantifiable phases identified (Figs. 3 and 4). Several inconsistencies occur between precipitate purity based on chromatography and identification of additional phases via EDX analyses. Although EDX analyses are only semi-quantitative, one clear inconsistency in purity is that sample 1301-11 (94% purity) shows a high wt.% C (15%). Other inconsistencies are those samples that are calculated with N 100% and those with b100% Table 7 The amount of sulfate (g) recovered from successive leachates of the pyrite-rich Jet Rock concretion. Sulfate recovery from the leachate decreases rapidly and ceases after the sample is leached under an N2 atmosphere. Times leached

Headspace conditions

SO2− recovered from 4 leachate (g)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere Atmosphere N2 headspace N2 headspace N2 headspace N2 headspace N2 headspace

0.03929 0.01452 0.17122 0.10197 0.15673 0.03903 0.07060 0.04239 0.05781 0.04171 0.06740 0.01263 0.00626 0.00416 0.00040 0.00000

purity that show very low contribution from additional phases. These results suggest that the chromatography method of purity calculation is subject to error, depending heavily on the efficiency of combustion and chromatographic separation. As an example, a few replicates from experiments 8 and 9 that show ~ 10% lower calculated purity than previous replicates (Tables 3 and 4) were analyzed at the end of the lifespan of the combustion/reduction column. Furthermore, it is clear from Figs. 5 and 6 that the precipitate purity and δ34S demonstrate a positive linear correlation, whereby lower purity precipitates have lower δ34S values. Therefore, we recommend careful tracking of the number of analyses on each combustion/reduction column for the most accurate purity assessment. It is unclear why some samples produced a non-BaSO4 precipitate following a standard gravimetric method for measuring BaSO4. NonBaSO4 samples are identified as having no chromatographic SO2 peak or a peak smaller than expected (and therefore low calculated purity), and in rare cases by a dark red or brown colored precipitate. EDX analyses of these samples confirm contributions from additional phases containing carbon, silicon, and phosphorus (Table 6). A precipitate of BaPO4 is a reasonable product of Monterey Fm samples, given that the sampled section of the Monterey Fm is phosphate-rich (Föllimi et al., 2005). Although the mechanism for precipitation of non-BaSO4 precipitates is unclear, non-BaSO4 precipitates are generated by samples with very little/no CAS. 4.2. Leached vs. unleached A 6% decrease in precipitate yield from unleached to leached samples suggests that either non-CAS SO24 − (possibly adsorbed onto CaCO3-CAS crystal surfaces) or a non-BaSO4 precipitate contributed to the overall precipitate mass for unleached samples. The most likely candidate for a non-BaSO4 precipitate is CaCl2. In a few early extractions, we observed that the acidified sample + BaCl2 generated crystallized CaCl2 along with BaSO4. CaCl2 was easily removed due to its high solubility. To avoid contamination by CaCl2, we rinsed and agitated BaSO4 precipitates with 18 MΩ water three times to remove any watersoluble salts. Indeed, low wt.% Ca and Cl from EDX analyses (Table 6) are insufficient to produce the large observed increase in precipitate yield between leached and unleached samples, suggesting our protocol effectively removes CaCl2. Therefore, we suggest that the yield discrepancy reflects contamination by a non-CAS SO24 − component and that comparison of leached to unleached precipitates serves as an analogue for sufficient leaching and removal of non-CAS SO2− 4 . Experiments 5–7 demonstrate a statistically significant 1.8‰ increase (between average values) in δ34S for leached samples as opposed to unleached samples (Table 2). Both leached and unleached samples demonstrate a positive linear relationship between precipitate purity and δ34S (Fig. 5). We interpret this offset in average isotopic value arises from addition of surface adsorbed non-CAS SO24 −in unleached extractions, so that δ34S from unleached samples reflects mixing of CAS-only 34 δ34S and non-CAS SO24 δ S. These data agree well with predictions

44

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

A

B

C

D

Fig. 4. Photomicrographs of precipitates from geologic samples of the Monterey Formation. White horizontal scale bars in each figure represent 20 μm. A) Sample 1301-05, hypothesized to have low purity based on red color of precipitate. EDX analysis confirms low wt.% O and very low wt.% S and wt.% Ba (Table 6). B) Sample 1301-07A, with 79% purity EDX analysis confirms near stoichiometric wt.% Ba, but low wt.% S and O. C) Sample 1301-11, with 94% purity. EDX analysis estimates lower than expected wt.% O, S, and Ba. D) Sample 1301-12, with 101% purity, confirmed by estimates from EDX analysis.

that the heavier isotope is preferentially incorporated into a crystal lattice given its greater bond-strength (Urey, 1947). We suggest that the contribution of a lower δ34S source is from surface adsorbed SO2− 4 (Usdowski and Hoefs, 1993; Gabitov et al., 2012). δ18O for unleached samples are statistically significant and average 2.6‰ lower than leached samples (Fig. 5, Table 2), suggesting similar fractionation processes for oxygen isotopes and sulfur isotopes govern incorporation of sulfate into calcite from the initial solution, confirming component in unleached synthetic CAS precipitates. a non-CAS SO2− 4 Therefore, we interpret that samples with a lower δ34S and δ18O than the leached CAS-only value indicate incomplete removal of non-CAS or contribution from oxidized pyrite (in pyrite-added samples). SO2− 4 To ensure removal of leached non-CAS SO2− 4 , we rinsed and centrifuged samples three times with 18 MΩ water after each leaching step. Due to the significant isotopic artifact demonstrated in unleached synthetic samples, all further discussion of isotopic data from variable extraction methodologies focuses on leached samples unless otherwise noted. 4.3. Clean CAS vs. pyrite added Pyrite added to synthetic CaCO3-CAS is the national standard NIST 8455 (bioleaching pyrite ore). Although NIST 8455 is not a δ34S standard, we consistently measure NIST 8455 at 2.7‰ ± 0.5‰ (average ± 1σ standard deviation). Neither average δ34S nor δ18O for leached CAS-only and pyrite-added samples are significantly different (Tables 2 and 3). Given the similarity of CAS-only δ34S to NIST 8455 δ34S, it is unlikely that oxidation of the added pyrite would produce a significant artifact δ34S. Similar δ18O values between all CAS-only and pyrite-added samples also suggest no significant oxidization of included pyrite. This result does not take into account the headspace conditions

for dissolution, however, which we hypothesized would be the critical variable in extraction of CAS. 4.4. Headspace conditions Due to extremely low mass of precipitate generated for all leached samples extracted under typical atmospheric conditions, we cannot assess the combined effect of leaching and atmospheric dissolution on δ34S and δ18O values (Table 2, Figs. 4, 5, and 6). It is unclear why atmospheric dissolution generated the lowest precipitate masses, though this result was consistently reproduced. As a result, we use the combined results of leached and unleached samples as a proxy for variable headspace extractions. We favor dissolution under an N2 environment because unleached samples dissolved under atmospheric conditions produced δ34S (Tables 2 and 3, Figs. 5C and 6) and δ18O values (Tables 2 and 3, Figs. 5F and 6) lower than the average CAS-only values. Samples extracted using an N2 headspace versus a constant N2 flush are within analytical error for both δ18O and δ34S, but are on average closer to CAS-only values than samples dissolved under atmospheric conditions. We favor N2 headspace dissolution because it requires less N2 gas, making it the more economical method. 4.5. Test of proposed method In experiments 8–9, we compare our proposed protocol between four replicates each of CAS-only and pyrite added sample extractions to test the accuracy and precision of the protocol (Tables 3 and 4, Fig. 6). The lack of any difference in δ34S between CAS-only and pyrite added samples (both are 1.1‰ ± 0.2‰) suggest that our proposed protocol sufficiently removes non-CAS SO2− 4 and will not oxidize associated

δ34S (‰,VCDT)

3.5

y = 0.0328x - 0.183 r2 = 0.76

2.5 y = 0.314x - 1.97 r2 = 0.71

1.5 0.5 -0.5 20

A

average δ18O (‰,VSMOW)

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

45

18

16

14

12 40

60 80 100 precipitate purity (%)

120

D

50

60 70 80 90 average precipitate purity (%)

100

50

60 70 80 90 average precipitate purity (%)

100

50

60 70 80 90 average precipitate purity (%)

100

average δ18O (‰,VSMOW)

18 δ34S (‰,VCDT)

2.5

1.5

0.5

-0.5 20

40

δ34S (‰,VCDT)

C

12

E

2.5

1.5

0.5

-0.5 20

14

120

average δ18O (‰,VSMOW)

B

60 80 100 precipitate purity (%)

16

18

16

14

12 40

60 80 100 precipitate purity (%)

120

F

Symbols leached

unleached

CAS-only

atmospheric headspace

pyrite-added

N2 flush

N2 headspace

Fig. 5. A–C) Plots of purity from synthetic CaCO3-CAS versus δ34S for experiments 5–7, in which extraction techniques are compared. D–F) Since purity is calculated during δ34S analyses, we use average purity and average δ18O to compare δ18O data from experiments 5–7. Error bars represent the 2σ on replicates of the average purity and average δ18O, when available.

pyrite. Samples in which we promote oxidation of pyrite after dissolution are not significantly different from unoxidized samples in terms of δ34S (1.0‰ ± 0.2‰). We suggest that the lack of change in δ34S from oxidized samples results from the small difference between average CAS-only δ34S (0.5‰) and the pyrite additive (2.7‰).

average δ18O (‰,VSMOW)

δ34S (‰,VCDT)

1.5

Similarly, δ18O for samples with and without added pyrite (13.8‰ ± 0.6‰ and 13.9‰ ± 0.5‰, respectively) confirm no oxidation of pyrite occurs using our proposed method. In contrast, purposefully oxidized samples demonstrate a statistically significant (95% confidence) 1.3‰ decrease in δ18O and dramatically reduced 1σ precision (2.4‰), due to

1.3 1.1 0.9 0.7 0.5 75

80

85 90 95 precipitate purity (%)

16 15 14 13 12 11 10

100

85

90 95 average precipitate purity (%)

100

Symbols CAS-only

pyrite-added

forced oxidation

Fig. 6. δ18O versus δ34S for experiments 8–9, comparing isotopic analyses of CAS-only samples (open circles), pyrite-added samples (filled circles), and samples in which pyrite is forced to anaerobically and abiotically oxidize (x), while using our suggested CAS protocol. Error bars represent the 2σ on replicates of the average purity and average δ18O, when available.

46

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

variable extents of oxidation. We demonstrated in earlier leaching experiments (Section 4.2) that our leaching method successfully removes non-CAS sulfate. Rennie and Turchyn (2014b) rigorously tested the exchange potential between oxygen isotopes of sulfate and low pH water, demonstrating that no exchange will occur on timescales of laboratory preparation of samples (days–weeks). Therefore, the observed oxygen isotope artifact must result from oxidation of included pyrite. During pyrite oxidation, pyrite sulfur combines with water oxygen as shown in Eq. (3) (δ18O water Pasadena, CA averages − 8.5‰; IAEA/WMO, 2014) to generate lower δ18O sulfate. The subsequent combination of CAS with δ18O 13.9‰ and lower δ18O sulfate generated by pyrite oxidation serves to decrease the δ18O of BaSO4 in oxidation experiments (Table 4). Our experiment that promotes oxidation by Fe3+ is not only important to describe potential isotopic mixing of CAS and oxidized sulfides, but also as a simulation of oxidation by release of Fe3+ from associated iron-bearing minerals including iron-oxyhydroxides (e.g. goethite). Recent studies have utilized the addition of SnCl2 to reduce free Fe3+ (Planavsky et al., 2012; Rennie and Turchyn, 2014a). While this method offers a possible solution to oxidation of pyrite (and other sulfides) by Fe3 +, it may not be sufficient. First, the concentration of Fe3 + in the sample must be known in order to add the amount of SnCl2 stoichiometrically required to reduce the Fe3+ to Fe2+. However, experimental reduction of acid volatile sulfides demonstrate that some reduction does occur, but do not quantify the completeness of the reaction (Chanton and Martens, 1985). Even if we assume that the reduction of Fe3+ by adding SnCl2 during acidification is complete, the resulting Fe2+ is subject to reoxidation. In the presence of oxygen, Fe2+ in an acidic solution (typically pH = 2 for acidified solution during CAS extraction) will rapidly be reoxidized to Fe3+ (Eq. (2)). We suggest that the use of SnCl2 as a reductant may be useful for samples with high concentrations of Fe3+bearing minerals, however the reaction will require removal of O2. 4.6. Testing on geologic samples: Monterey Formation and Jet Rock Formation Four of the six facies subsampled to test our proposed CAS extraction protocol produced a quantifiable amount of precipitate (Table 5). Due to the mixed mineralogy of natural samples, we replicated extractions using a single leaching/rinsing step for one sample group and three leaching/rinsing steps for a second sample group. One versus three leaching steps produced no measurable change in precipitate yield for the red mudstone and concretion samples (Table 5). Likewise, purity does not change significantly between samples, suggesting that one for these facies. δ34S leaching step sufficiently removed non-CAS SO2− 4 does not show a significant difference between one versus three leaching steps for 1301-07A, but does show a difference between leaching steps for 1301-11 (Table 5). This suggests that some samples may benefit from more than one leaching step. We suggest that a trial leaching experiment be conducted on each facies type identified in the sampled succession to determine whether all non-CAS sulfate is removed during leaching. δ18O values between samples leached one versus three times also show no measurable difference. We suggest that higher uncertainty in δ34S and δ18O evident in some Monterey Fm samples results from sample heterogeneity rather than oxidation of pyrite or other sulfides. The high organic content of Monterey Formation samples provides an additional check on our proposed method. If organics were oxidized during our extraction procedure, we would see a shift in δ34S to very low values and mixing of oxygen isotopes similar to that expected for sulfide oxidation. However, these samples do not demonstrate isotopic mixing of such endmembers, suggesting that our extraction procedure does not oxidize associated organics. Test leaching of the Jet Rock concretion sample however, demonstrates significant oxidation of pyrite during leaching (Table 7). Sulfate was consistently produced under subsequent leaching in normal, oxic atmospheric conditions. However, BaSO4 produced from leaching

under N2 conditions decreased almost exponentially. Based on this observation, we recommend one of the following approaches for samples of similarly high pyrite concentration. If analysis of both pyrite and CAS are desired, the sample must be prepared entirely under anoxic conditions, including leaching, filtering, acidification, and precipitation of BaSO4. If only the CAS is required, we recommend promoting oxidation of the sample sulfides during the leaching step. This can be completed simply by leaching under an oxic atmosphere or by addition of FeCl3 · 6H2O during leaching. This step should be repeated until no BaSO4 is precipitated from the leachate. 4.7. Suggested protocol for CAS extraction Based on these results, we suggest the following protocol to preserve both CAS and associated sulfides: 1. Conduct precursor leaching test on a small subset of samples based on mineralogical characteristics to determine whether additional leaching steps are necessary to remove non-CAS SO2− 4 . For samples with high concentrations of Fe3+-bearing minerals such as siderite and goethite, the leaching process must be done entirely under an N2 atmosphere. 2. Leach samples a specified number of times based on the results from step 1 in 10% NaCl in 18 MΩ deionized water. 3. Centrifuge and rinse samples in 18 MΩ water 3 times after leaching to ensure removal of non-CAS SO2− 4 . 4. Dissolution set-up (Fig. 1): Transfer sample to an Erlenmeyer flask with vacuum/gas attachment. A double-holed stopper should be inserted on the top of the flask. The stem of the dropping flasks is inserted into one hole, allowing 10% HCl to be added dropwise. A tube attachment exits the second hole and extends to a waterfilled flask, allowing CO2 gas to escape without allowing for backpressure of atmosphere (Fig. 1). 5. Allow headspace to flush with N2 gas for 5 min and turn off gas, leaving the system isolated from the atmosphere. 6. Add 10% HCl in 18 MΩ deionized water dropwise to the sample. Allow the sample to complete acidification for at least 1.5 h, agitating occasionally to ensure all sample has reacted with the acid. 7. Given the large volume of acid + sample, centrifuge, then filter the sample. 8. Heat filtrate to 72 °C–80 °C and add BaCl2 in 18 MΩ deionized water dropwise at ~10% of the volume of the filtrate. 9. Heat the sample for 3 h, then remove from heat and allow to cool for at least 72 h. 10. Centrifuge precipitate and rinse three times in 18 MΩ water. 5. Conclusions 1. These results demonstrate that leaching with 10% NaCl and rinsing in 18 MΩ water sufficiently removes non-CAS SO2− 4 . 2. CAS extraction using N2 headspace during the dissolution by 10% HCl shows no isotopic artifact due to pyrite oxidation. 3. Our proposed protocol produces accurate δ34S and δ18O analyses on synthetically generated CAS precipitates with a 1σ standard deviation of 0.1‰ and 0.3‰, respectively, compared with 1σ standard deviation of 0.2‰ and 0.3‰ for standards run in the same way. Our proposed protocol demonstrates that included pyrite does not oxidize during the leaching and acidification steps. 4. The 1σ standard deviation on replicates of mixed mineralogy and low pyrite concentration geologic samples from the Monterey Formation is 0.4‰–0.7‰ for δ34S and 0.8‰–1.3‰ for δ18O, resulting from sample heterogeneity. 5. Leaching experiments on mixed mineralogy geologic samples demonstrate that leaching under oxic conditions oxidized high pyrite concentration samples from the Jet Rock Formation, and that oxidation ceased as the sample was leached under an N2 atmosphere.

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

We suggest that high pyrite concentration samples be leached and acidified under an N2 atmosphere to preserve pyrite during CAS extraction.

Acknowledgments We would like to thank Jim Boles for his assistance collecting samples from the Monterey Formation and Sean Loyd for his guidance in accessing the Naples Beach outcrops of the Monterey Formation. We are grateful for the comments and suggestions from Simon Bottrell and an anonymous reviewer, which considerably improved the focus of the manuscript. This research was carried out at the Jet Propulsion Laboratory (JPL), California Institute of Technology, under contract with the National Aeronautics and Space Administration (NASA), and supported by the NASA Astrobiology Institute (CAN 6, Wisconsin Astrobiology Research Consortium). References Antler, G., Turchyn, A.V., Rennie, V., Herut, B., Sivan, O., 2013. Coupled sulfur and oxygen isotope insight into bacterial sulfate reduction in the natural environment. Geochim. Cosmochim. Acta 118, 98–117. Barron, J.A., 1986. Paleoceanographic and tectonic controls on deposition of the Monterey Formation and related siliceous rocks in California. Palaeogeography, Palaeoclimatology, Palaeoecology 53, 27–45. Behl, R.J., Bramlette, Since, 1946. The Miocene Monterey Formation of California revisited. Geological Society of America Special Papers 338 (1999), 301–338. Böttcher, M.E., Thamdrump, B., 2001. Anaerobic sulfide oxidation and stable isotope fractionation associated with bacterial sulfur disproportionation in the presence of MnO2. Geochim. Cosmochim. Acta 65, 1573–1581. Bottrell, S.H., Newton, R.J., 2006. Reconstruction of changes in global sulfur cycling from marine sulfate isotopes. 75, 59–83. Bramlette, M.N., 1946. The Monterey Formation of California and the origin of its siliceous rocks. USGS Professional Paper 212. Brunner, B., Bernasconi, S.M., Kleikemper, J., Schroth, M.H., 2005. A model for oxygen and sulfur isotope fractionation in sulfate during bacterial sulfate reduction processes. Geochim. Cosmochim. Acta 69, 4773–4785. Burdett, J.W., Arthur, M.A., Richardson, M., 1989. A Neogene seawater sulfur isotope age curve from calcareous pelagic microfossils. Earth Planet. Sci. Lett. 94, 189–198. Chanton, J.P., Martens, C.S., 1985. The effects of heat and stannous chloride addition on the active distillation of acid volatile sulfide from pyrite-rich marine sediment samples. Biogeochemistry 1, 375–382. Claypool, G.E., Holser, W.T., Kaplan, I.R., Sakai, H., Zak, I., 1980. The age curve of sulfur and oxygen isotopes in marine sulfate and their mutual interpretation. Chem. Geol. 28, 190–260. Fike, D.A., Grotzinger, J.P., 2008. A paired sulphate–pyrite δ34S approach to understanding the evolution of the Ediacaran–Cambrian sulphur cycle. Geochim. Cosmochim. Acta 72, 2636–2648. Fike, D.A., Grotzinger, J.P., Pratt, L.M., Summons, R.E., 2006. Oxidation of the Ediacaran Ocean. Nature 444, 744–747. Föllimi, K.B., Badertscher, C., de Kaenel, E., John, C.M., Adatte, T., Steinmann, P., 2005. Phosphogenesis and organic-carbon preservation in the Miocene Monterey Formation at Naples Beach, California—The Monterey hypothesis revisited. Geological Society of America Bulletin 117, 589–619. Fritz, P., Basharmal, G.M., Drimmie, R.J., Ibsen, J., Qureshi, R.M., 1989. Oxygen isotope exchange between sulfate and water during bacterial reduction of sulfate. Chem. Geol. 79, 99–105. Gabitov, R.E., Watson, E.B., Sadekov, A., 2012. Oxygen isotope fractionation between calcite and fluid as a function of growth rate and temperature: an in situ study. Chem. Geol. 306–307, 92–102. Garrison, R.E., Douglas, R.G., 1981. The Monterey Formation and related siliceous rocks of California. Los Angeles, Pacific Section, Society of Economic Paleontologists and Mineralogists 327pp. Gill, B.C., Lyons, T.W., Frank, T.D., 2011. A global perturbation to the sulfur cycle during the Toarcian Oceanic Anoxic Event. Earth Planet. Sci. Lett. 312, 484–496. Holland, H.D., 1984. The Chemical Evolution of the Atmosphere and Oceans. Princeton University Press, Princeton, New Jersey (582 pp.). Holser, W.T., 1992. Stable isotope geochemistry of sulfate and chloride rocks. In: Clauer, N., Chaudhuri, S. (Eds.), Isotopic Signatures and Sedimentary Records. Lecture Notes Earth Science v.43, pp. 153–176. Howarth, M.K., 1962. The Jet Rock Series and Alum Shale Series of the Yorkshire Coast. Proc. Yorks. Geol. Soc. 33, 381–422. Hurtgen, M.T., Arthur, M.A., Suits, N.S., Kaufman, A.J., 2002. The sulfur isotopic composition of Neoproterozoic seawater sulfate: implications for a snowball Earth? Earth Planet. Sci. Lett. 203, 413–429. IAEA/WMO, 2014. Global Network of Isotopes in Precipitation. The GNIP Database Accessible at: http://www.iaea.org/water. Issacs, C.M., Pisciotto, K.A., Garrion, R.E. (Eds.), 2001. The Monterey Formation: From Rocks to Molecules. New York, Columbia University Press (608 pp.).

47

Kampschulte, A., Bruckschen, P., Strauss, H., 2001. The sulfur isotopic composition of trace sulphates in Carboniferous brachiopods: implications for coeval seawater, correlation with other geochemical cycles and isotope stratigraphy. Chem. Geol. 175, 149–173. Kamschulte, A., Strauss, H., 2004. The sulfur isotopic evolution of Phanerozoic seawater based on the analysis of structurally substituted sulfate in carbonates. Chem. Geol. 204, 255–286. Kornexl, B.E., Gehre, M., Hofling, R., Werner, R.A., 1999. On-line δ18O measurement of organic and inorganic substances. Rapid Commun. Mass Spectrom. 13, 1685–1693. Krouse, H.R., Mayer, B., 2000. Sulfur and oxygen isotopes in sulfate. In: Cook, P., Herczeg, A.L. (Eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Norwell, Massachusetts, pp. 195–232. Loyd, S.J., Berelson, W.M., Lyons, T.W., Hammond, D.E., Corsetti, F.A., 2012. Constraining pathways of microbial mediation for carbonate concretions of the Miocene Monterey Formation using carbonate-associated sulfate. Geochim. Cosmochim. Acta 78, 77–98. Luo, G., Kump, L.R., Wang, Y., Tong, J., Arthur, M.A., Yang, H., Huang, J., Yin, H., Xie, S., 2010. Isotopic evidence for an anomalously low oceanic sulfate concentration following end-Permian mass extinction. Earth Planet. Sci. Lett. 300, 101–111. Marenco, P.J., Corsetti, F.A., Hammond, D.E., Kaufman, A.J., Bottjer, D.J., 2008. Oxidation of pyrite during extraction of carbonate associated sulphate. Chem. Geol. 247, 124–132. Mitzutani, Y., Rafter, T.A., 1973. Isotopic behaviours of sulphate oxygen in the bacterial reduction of sulphate. Geochem. J. 6, 183–191. Nardelli, M.P., Barras, C., Metzger, E., Mouret, A., Filipsson, H.L., Jorissen, F., Geslin, E., 2014. Experimental evidence for foraminiferal calcification under anoxia. Biogeosciences 11, 4029–4038. Newton, R.J., Pevitt, E.L., Wignall, P.B., Bottrell, S.H., 2004. Large shifts in the isotopic composition of seawater sulphate across the Permo–Triassic boundary in northern Italy. Geology 39, 7–10. Newton, R.J., Reeves, E.P., Kafousia, N., Wignall, P.B., Bottrell, S.H., Sha, J.-G., 2011. Low marine sulfate concentrations and the isolation of the European epicontinental sea during the Early Jurassic. Earth Planet. Sci. Lett. 218, 331–345. Nielsen, H., 1989. Local and global aspects of the sulphur isotope age curve of oceanic sulphate in. In: Brimblecombe, P. (Ed.), Evolution of the Global Biogeochemical Sulphur Cycle. Wiley, New York, pp. 57–64. Nordstrom, D.K., Southam, G., 1997. Geomicrobiology of sulfide mineral oxidation. In: Banfield, J.F., Nealson, K.H. (Eds.), Geomicrobiology. Reviews in Mineralogy v. 35. Mineralogical Society of America, Washington, pp. 361–390. Ohkouchi, N., Kawamura, K., Kajiwaea, Y., Wada, E., Okada, M., Kanamatsu, T., Taira, A., 1999. Sulfur isotope records around Livello Bonarelli (northern Apennines, Italy) black shale at the Cenomanian–Turonian boundary. Geology 27, 535–538. Paris, G., Fehrenbacher, J.S., Sessions, A.L., Spero, H.J., Adkins, J.F., 2014. Experimental determination of carbonate-associated sulfate δ34S in foraminifera shells. Geochem. Geophys. Geosyst. 15, 1452–1461. Paytan, A., Kastner, M., Campbell, D., Thiemens, M.H., 1998. Sulfur isotopic composition of Cenozoic seawater sulfate. Science 282, 1459–1462. Peng, Y., Bao, H., Pratt, L.M., Kaufmann, A.J., Jiang, G., Boyd, D., Wang, Q., Zhou, C., Yuan, X., Xiao, S., Loyd, S., 2014. Widespread contamination of carbonate-associated sulfate by present-day secondary atmospheric sulfate: evidence from triple oxygen isotopes. Geology 42, 815–818. Planavsky, N.J., Bekker, A., Hofmann, A., Owens, J.D., Lyons, T.W., 2012. Sulfur record of rising and falling marine oxygen and sulfate levels during the Lomagundi event. Proc. Natl. Acad. Sci. 109, 18300–18305. Raiswell, R., 1976. The microbiological formation of carbonate concretions in the upper Lias of NE England. Chem. Geol. 18, 227–244. Rennie, V.C., Turchyn, A.V., 2014a. The preservation of δ34SSO4 and δ18OSO4 in carbonateassociated sulfate during marine diagenesis: a 25 Myr test case using marine sediments. Earth Planet. Sci. Lett. 395, 13–23. Rennie, V.C., Turchyn, A.V., 2014b. Controls on the abiotic exchange between aqueous sulfate and water under laboratory conditions. Limnol. Oceanogr. Methods 12, 166–173. Shen, B., Xiao, S., Bao, H., Kaufman, A.J., Zhou, C., Yuan, X., 2011. Carbon, sulfur, and oxygen isotope evidence for a strong depth gradient and oceanic oxidation after the Ediacaran Hankalchough glaciations. Geochim. Cosmochim. Acta 75, 1357–1373. Staudt, W.J., Reeder, R.J., Schoonen, M.A.A., 1994. Surface structural controls on composiand SeO2− in synthetic single crystals. Geochim. Cosmochim. tional zoning of SO2− 4 4 Acta 58, 2087–2098. Strauss, H., 1997. The isotopic composition of sedimentary sulfur through time. Palaeogeogr. Palaeoclimatol. Palaeoecol. 132, 97–118. Strauss, H., 1999. Geological evolution from isotope proxy signals: sulfur. Chem. Geol. 161, 89–101. Takano, B., 1985. Geochemical implications of sulfate in sedimentary carbonates. Chem. Geol. 49, 393–403. Taylor, B.E., Wheeler, M.C., Nordstrom, D.K., 1984. Stable isotope geochemistry of acid mine drainage: experimental oxidation of pyrite. Geochim. Cosmochim. Acta 48, 2669–2678. Thode, H.G., Kleerekoper, H., McElcheran, D., 1951. Isotope fractionation in the bacterial reduction of sulfate. Research, Lond. 4, 581–582. Turchyn, A.V., Schrag, D.P., 2004. Oxygen isotope constraints on the sulfur cycle over the past 10 million years. Science 303, 2004–2007. Urey, H.C., 1947. The thermodynamic properties of isotopic substances. J. Chem. Soc. 562–581. http://dx.doi.org/10.1039/JR9470000562. Usdowski, E., Hoefs, J., 1993. Oxygen isotope exchange between carbonic acid, bicarbonate, carbonate, and water: a re-examination of the data of McCrea (1950) and an expression for the overall partitioning of oxygen isotopes between the carbonate species and water. Geochim. Cosmochim. Acta 57, 3815–3818. van Stempvoort, D.R., Krouse, H.R., 1994. Controls of δ18O in sulfate—review of experimental data and applications to specific environments. In: Alpers, C.N., Blowes,

48

B.P. Theiling, M. Coleman / Chemical Geology 411 (2015) 36–48

D.W. (Eds.), Environmental geochemistry of sulfide oxidation. American Chemical Society Symposium Series 550, pp. 446–480. Wotte, T., Shields-Zhou, G.A., Strauss, H., 2012a. Carbonate-associated sulfate: experimental comparisons of common extraction methods and recommendations toward a standard analytical protocol. Chem. Geol. 326–327, 132–144. Wotte, T., Strauss, H., Fugmann, A., Garbe-Schönberg, D., 2012b. Paired δ34S data from carbonate-associated sulfate and chromium-reducible sulfur across the traditional

Lower–Middle Cambrian boundary of W-Gondwana. Geochim. Cosmochim. Acta 85, 228–253. Xiao, S., McFadden, K.A., Peek, S., Kaufman, A.J., Zhou, C., Jiang, G., Hu, J., 2012. Integrated chemostratigraphy of the Doushantuo Formation at the northern Xiaofenghe section (Yangtze Gorges, South China) and its implication for Ediacaran stratigraphic correlation and ocean redox models. Precambrian Res. 192–195, 125-141.