Case studies on the utility of sequential carbonate leaching for radiogenic strontium isotope analysis

Accepted Manuscript Case studies on the utility of sequential carbonate leaching for radiogenic strontium isotope analysis

Eric J. Bellefroid, Noah J. Planavsky, Nathaniel R. Miller, Uwe Brand, Chunjiang Wang PII: DOI: Reference:

S0009-2541(18)30422-4 doi:10.1016/j.chemgeo.2018.08.025 CHEMGE 18889

To appear in:

Chemical Geology

Received date: Revised date: Accepted date:

10 December 2017 22 August 2018 26 August 2018

Please cite this article as: Eric J. Bellefroid, Noah J. Planavsky, Nathaniel R. Miller, Uwe Brand, Chunjiang Wang , Case studies on the utility of sequential carbonate leaching for radiogenic strontium isotope analysis. Chemge (2018), doi:10.1016/ j.chemgeo.2018.08.025

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium Isotope Analysis Eric J. Bellefroid1, Noah J. Planavsky1, Nathaniel R. Miller2, Uwe Brand3, Chunjiang Wang4 1

Dept. of Geology and Geophysics, Yale University, New Haven, CT, USA Dept. of Geological Sciences, University of Texas at Austin, Austin, TX, USA 3 Dept. of Earth Sciences, Brock University, St. Catharines, ON, Canada 4 College of Geosciences, China University of Petroleum - Beijing, Beijing, China

T

2

IP

ABSTRACT

CR

Radiogenic strontium isotopes (87Sr/86Sr) have been extensively used as a tool to explore a diversity of Earth system problems, including long-term global weathering rates and global sequence correlation.

US

Strontium isotopes are measured on a range of geological materials (e.g., calcite fossils, barites, limestone micrites), but whole-rock limestones are by far the most abundant of these materials within the geological record for paleo-seawater 87Sr/86Sr work. Whole-rock limestones, however, have a poor track record of 87

Sr/86Sr values. Alteration of the limestone during diagenesis and

AN

recording primary seawater

contamination from detrital aluminosilicate phases during carbonate extraction have been consistent 87

Sr/86Sr

M

problems. Various preparation and quality control methods have been applied to whole-rock

work, yet there remains no consistent framework used to separate and identify both contamination and

ED

alteration simultaneously. The lack of consistent and systematic methods has made it difficult to gauge the accuracy and fidelity of much of the previously generated whole rock limestones

87

Sr/86Sr data,

PT

especially for Precambrian sequences. Building on previous work, we explore a sequential leaching method designed to systematically isolate least-altered carbonate phases from detrital aluminosilicate Sr

CE

contamination and present several case studies that demonstrate the advantages of this approach. In the first case study, we use the Mid-Carboniferous Bird Spring Formation to empirically validate the accuracy of this sequential leaching method. Comparing least-altered sequentially leached whole-rock Sr/86Sr values with well-preserved calcite brachiopod

AC

87

87

Sr/86Sr values from the same section, we find

near identical values. Following this first case, we studied the Neoproterozoic Dhaiqa Formation and the mid-Proterozoic Jixian Group and Muskwa Assemblage to outline a framework for identifying leastaltered leachate fractions in Proterozoic carbonate samples. As a whole, we find that with this method it is possible to better identify whole-rock samples primary or least-altered carbonate fractions, and better account for alteration, providing a means to back-calculate a samples primary and least-altered marine 87

Sr/86Sr value.

Keywords: carbonate extraction, dissolution method, Precambrian, chemostratigraphy, Sr Isotope

1

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

1. INTRODUCTION Radiogenic Sr isotopes (87Sr/86Sr) have proven to be a key tool for solving a wide range of Earth system problems including tracking long-term paleo-weathering rates (e.g., Veizer and Compston, 1976; Raymo and Ruddiman, 1992; Halverson et al., 2007, Kuznetsov et al., 2014), global chemostratigraphic correlation of marine carbonate sequences (McArthur et al., 2012), and distinguishing between marine and non-marine environments (e.g., Veizer et al., 1990; Spencer et al., 1997; Stueeken et al. 2017).

T

Although 87Sr/86Sr values are ideally measured from unaltered primary chemical precipitates (i.e., calcite

IP

and aragonite hard-shelly fossils, marine cements and evaporites; Brand and Brenckle, 2001; Kah et al.,

CR

2001), well-indurated limestone (and dolostone) successions dominate the chemical sedimentary record and are often the only available archives for deciphering seawater compositions in deep time. This is especially true for Precambrian sedimentary successions, which predate the evolution of calcifying

estimates for Precambrian seawater

87

US

organisms and have typically experienced some degree of diagenesis and metamorphism. The current best Sr/86Sr are based on targeting samples with low aluminosilicate

AN

content and assuming the least radiogenic samples to be the least-altered (Shields and Veizer, 2002). However, this interpretation relies on the argument that diagenetic alteration and aluminosilicate

M

contamination exclusively drive carbonate 87Sr/86Sr to more radiogenic values (e.g., Shields and Veizer, 2002), as diagenetic and detrital contamination most commonly source their Sr from aged cratonic

ED

sequences which are Sr-rich and radiogenic (e.g., Miller et al., 2009). In some cases, however, secondary diagenesis has in fact been found to drive carbonate to more unradiogenic values (Connolly et al., 1990; Miller et al., 2008), questioning the validity of this assumption. The utility of carbonate

87

Sr/86Sr is

PT

entirely dependent on the identification of well-preserved, marine carbonates for 87Sr/86Sr measurements. Therefore, it is imperative to establish sample processing methods capable of isolating primary carbonate 87

Sr/86Sr measurements, or at minimum least-altered carbonate fractions, and thereby

CE

fractions for

minimize diagenetic overprinting of primary depositional signatures.

AC

Carbonate 87Sr/86Sr values are susceptible to two main sources of contamination. During early and late diagenesis, primary carbonate minerals (e.g., aragonite, high-Mg calcite) recrystallize to more stable dolomite or low-Mg calcite mineralogies (Morse and McKenzie, 1990), potentially incorporating nonmarine or non-contemporaneous marine Sr from diagenetic fluids (e.g., Brand and Veizer, 1980; 1981; Banner and Hanson, 1990). In addition, during sample preparation, the leaching of detrital aluminosilicate phases (terrigenous clay and silicate minerals) can add an unintended source of non-marine Sr contamination (McArthur, 1994; Banner, 1995; Bailey et al., 2000). Curbing 87Sr/86Sr contamination from diagenetic alteration by any of these scenarios is best achieved by using appropriate filtering methods to clearly identify primary carbonate phases for targeted sampling and analysis (e.g., pristine calcite fossils, 2

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

micro-drilling of pristine whole-rock limestone phases). Commonly, petrographic filters (e.g., cathodoluminescence) are used as a pre-screen to assess primary textural preservation, and geochemical filters (e.g., Mn/Sr, δ18O) are used to identify alteration by secondary non-marine fluids on targeted samples (e.g., Grover and Read, 1983; Morton, 1985; Banner et al., 1988; Machel and Burton, 1991; Brand et al., 2012; Hood and Wallace, 2015). Unlike diagenetic alteration, however, which occurs during lithification and burial, detrital aluminosilicate contamination is most likely a product of lab techniques

T

where aluminosilicate Sr is inadvertently released into solution either by ion exchange or dissolution.

IP

Despite the increasing use of geochemical and petrographic filtering methods to target least-altered whole-rock carbonate samples, many studies still report relatively high aluminium concentrations in 87

Sr/86Sr, suggesting that current dissolution methods may not

CR

carbonate leachates analyzed for

systematically separate carbonate phases from aluminosilicate phases.

US

The most common leaching method for carbonate extractions is bulk leaching, which separates carbonate phases from detrital aluminosilicate phases by selecting an acid and calibrating the volume of acid needed

AN

to only dissolve the carbonate phases. In these methods, a selected acid is added to a powdered carbonate sample and, as carbonate minerals are more reactive and have a much faster dissolution rate than detrital

M

aluminosilicates (Rauch and White, 1977), they will dissolve first leaving insoluble detrital phases behind (Banner et al., 1988; Montanez et al., 1996; Kupecz and Land, 1991; Edwards et al., 2015). Such single-

ED

leachate methods have generally been preferred when authors have focused on large data sets (e.g., Banner; 2004; Li et al., 2011, but with some exceptions; Bailey et al., 2000; Bayon et al., 2002; Wen et al., 2015), but may suffer from a number of potential shortcomings. Whole-rock carbonate samples can

PT

contain variable amounts of insoluble aluminosilicates. As samples are often leached in batches, with a specified acid strength and volume for all samples within a given batch, an acid excess for impure

CE

carbonates risks the dissolution and release of contaminant aluminosilicate-Sr into the leachate solution. Some authors have opted to adjust acid volumes, or sample masses, according to the carbonate content of

AC

individual samples (Miller et al., 2008; 2009; Li et al., 2011; Tostevin et al., 2016). However, an alternative method is to sequentially leach samples, measuring both geochemical contamination proxies (trace elements) and 87Sr/86Sr values on each leach step, in order to identify the individual sample fraction least affected by aluminosilicate contamination (e.g., Bailey et al., 2000). This method negates the need to calibrate acid volumes for each sample, which can introduce errors, and provides an additional benefit in that calcite will leach before dolomite, effectively excluding the influence of secondary, and potentially diagenetic, dolomite on the measured 87Sr/86Sr value. Following from recent work that has explored more systematic carbonate extraction methods (Li et al., 2011; Liu et al., 2013; 2014; Morera-Chavarría et al., 2016; Tostevin et al., 2016), we present a 3

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

streamlined sequential leaching method that effectively isolates bulk carbonate sample fractions from detrital aluminosilicate Sr as well as provides meaningful context on potential sources of alteration missed by single-step leaching methods. Further, we show that application of the sequential leach approach to whole-rock limestone samples can replicate high-precision 87Sr/86Sr values obtained from pristine fossils in the Mid-Carboniferous Bird Spring Fm. (Formation). We then explore the Neoproterozoic Dhaiqa Fm. of the Arabian Peninsula and show that we can systematically identify least altered leachate fractions to 87

Sr/86Sr values than bulk-leached results. Lastly, we assess sample preservation

T

yield more consistent

IP

using this method for samples in two Mesoproterozoic carbonate sequences, the Muskwa Assemblage of north-east British Columbia and the Jixian Group of North China. Although both carbonate successions

CR

contain varying amounts of detrital aluminosilicate and experienced fluid-rich diagenetic alteration, we find that the trends in single sample leachate-to-leachate variations can be used to estimate and project 87

Sr/86Sr values that are consistent with values obtained from better-preserved samples of

US

least-altered

broadly equivalent age. Further, we provide examples of how the sequential method provides a means to

AN

gauge the reliability of 87Sr/86Sr data more accurately than conventional methods for screening diagenetic

M

alteration and detrital aluminosilicate contamination.

ED

2. SAMPLE GEOLOGY

2.1 Bird Spring Formation, Arrow Canyon, Nevada

PT

The Bird Spring Fm. in southern Nevada and southeastern California is a shallow marine carbonate shelf sequence that was deposited along the western Pangean margin (Fig. 1; Stevens and Stone, 2007), with

CE

the Arrow Canyon section (north of Las Vegas, NV) recognized as the Global Boundary Stratotype Section and Point (GSSP; Lane et al., 1999) for the Mid-Carboniferous (Pennsylvanian-Mississippian). Moderately overprinted by Mesozoic Sevier thrusting, the sequence was exposed during Neogene uplift

AC

and remains relatively undeformed and well-preserved (Page and Dixon, 1997). The Bird Spring Fm. is a thick and nearly continuous carbonate section that preserves low energy intertidal to high energy beach and shoal facies (mainly aragonitic; Brand et al., 2012), with minor siliciclastic contents (Lane et al., 1999; Barnett and Wright, 2007; Bishop et al., 2009). Global sea-level during this time period fluctuated rapidly in response to glacioeustatic forcings, with the development of thin microkarst surfaces indicative of exposures during sea level low-stands (Bishop et al., 2009). Although these thin microkarst surfaces indicate minimal soil formation suggesting an arid climate and short exposure times, these low-stands may have driven meteoric fluid flow and diagenesis (Bishop et al., 2009; 2010). Indeed, detailed petrographic work on this specific section documents the presence of 4

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

secondary meteoric calcite that fills remaining pore space and minor fractures and replaces precursor evaporite minerals (Bishop et al., 2009; Brand et al., 2012). Thus, although the sequence is relatively well-preserved in terms of depositional facies, it has undergone some degree of meteoric diagenesis (Brand, 2004; Brand et al., 2012). A key global boundary location, Brand and Brenckle (2001) identified and measured 87Sr/86Sr on pristine low-Mg calcite brachiopods within the sequences and found a consistent value of 0.708082 ± 0.000040

T

throughout the sequence. Work by Brand (2004) and Brand et al. (2012) used the site to test diagenetic

IP

screening methods and found that bulk dissolved whole-rock samples from the same sequence had

CR

systematically more radiogenic 87Sr/86Sr values. Following up on these studies, we applied the sequential leaching method to 13 splits of whole-rock samples from the Brand (2004) and Brand et al. (2012)

US

studies. 2.2 Dhaiqa Formation, Jibalah Group

AN

Located along the western margin of the Arabian shield, the Ediacaran Dhaiqa Formation is a several hundred meter-thick limestone unit bearing metazoan trace fossils (Miller et al., 2008) that is preserved within the Dhaiqa Basin, a mildly deformed and metamorphosed pull-apart basin related to the Najd shear

M

zone (Johnson et al., 2011; 2013; Vickers-Rich et al., 2013; Nettle et al., 2014). Dhaiqa Fm. limestones

ED

unconformably overlie metasediments and metavolcanics of the Baydan Group, which are uniquely exposed in the roughly 3 km by 1 km Dhaiqa Basin (Fig. 2; Bryan, 1985; Miller et al., 2008). The uppermost unit of the dominantly volcanic and siliciclastic Jibalah Group, the Dhaiqa Fm. is primarily

PT

composed of platy and undulate laminated microbialite limestones with some siltstone (Miller et al., 2008). Although dominantly limestone, the basal 100 m is partially dolomitized, and most of the unit

CE

contains minor late-stage ferroan dolomite (Miller et al., 2008). Given that the basin is well-exposed, fairly accessible and contains metazoan trace fossils, geochemical and sedimentological work correlating

AC

these successions globally has attracted attention recently (e.g., Johnson et al., 2013; Al-Husseini, 2014; Nettle et al., 2014). From this initial work, it remains unclear whether the Dhaiqa Fm. and its equivalents record a broad and continuously deposited sequence well-connected to the oceans, or whether these all represent coeval units deposited independently within regionally or locally restricted basins (for review see Johnson et al., 2013). To place the Dhaiqa Fm. within a global context, Miller et al. (2008) measured

87

Sr/86Sr on

petrographically-informed targeted micro-drilled whole-rock limestones but found 87Sr/86Sr values from 0.70394 to 0.70603, far below typical Ediacaran seawater

87

Sr/86Sr values (Halverson et al., 2010).

Preserved within and surrounded by the Baydan Group, a mixed rhyolite and basalt metavolcanic 5

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

sequence (Bryan, 1985), the low

87

Bellefroid et al.

Sr/86Sr values of the Dhaiqa Fm. have been suggested to reflect

restricted basin conditions during deposition, where water column

87

Sr/86Sr values reflect a mix of

seawater and locally weathered basalt. Alternatively, Miller et al. (2008) found that many samples were partially altered (Mn/Sr >1, high Al/Ca), thus these low

87

Sr/86Sr values could be a product of late

diagenesis by diagenetic fluids carrying low 87Sr/86Sr signatures derived from alteration of local basalts. Differentiating between an altered signal and a restricted basin using

87

Sr/86Sr can be difficult as

T

aluminosilicate dissolution during leaching will lead to more radiogenic values while secondary

IP

alteration, in this case, would drive carbonate 87Sr/86Sr values to less radiogenic values. Thus, isolating the least-altered carbonate fraction of each sample for 87Sr/86Sr analysis is imperative to better understand the

CR

depositional and diagenetic history of this basin.

Using petrographic and geochemical results from Miller et al. (2008), we selected seven of the same

method.

US

samples for analysis and micro-drilled thin-section billets for powders using the sequential leaching The objective here was to see if sequential leachate fractions could provide additional

AN

information about the source of the nonradiogenic Dhaiqa Fm. 87Sr/86Sr values. 2.3 Muskwa Assemblage, NE British Columbia and the Jixian Group, North China

M

The Mesoproterozoic portion of the Proterozoic seawater 87Sr/86Sr record contains sparse data compiled

ED

from only a few key basins around the world (Shields and Veizer, 2002; Prokoph et al., 2008). The poor preservation and lack of skeletal carbonate in pre-Ediacaran strata have made it difficult to delimit Proterozoic marine

87

Sr/86Sr evolution. Well-preserved limestone samples are ideal candidates for

Sr/86Sr analysis, yet they are rare in the Proterozoic and Archean rock record compared to dolomitic

PT

87

lithologies (Ronov, 1964). To improve the Mesoproterozoic

87

Sr/86Sr record, we targeted the Muskwa

CE

Assemblage, a little-studied Mid-Proterozoic basin located in northeastern British Columbia, and the Jixian Group in Northern China, both of which are substantially comprised of dolomite and partially

AC

dolomitized limestone.

Dominantly carbonate and well-exposed, the 6 km-thick Muskwa Assemblage is a Mid-Proterozoic mixed carbonate-siliciclastic sequence exposed in the Rocky Mountains of northeastern British Columbia (Bell, 1968; Taylor and Stott, 1973; LeCheminant and Heaman, 1994; Ross et al., 2001). Consisting of a carbonate ramp seaward of a large western-draining delta, the sequence was deposited over Paleoproterozoic granitic basement and subsequently exposed during Late Cretaceous-Paleogene Laramide thrusting (Bell, 1968; Taylor and Stott, 1973; Ross et al., 2001; Cook et al., 2004). The assemblage experienced deformation and sub-greenschist grade metamorphism during thrusting (Bell, 1968; Taylor and Stott, 1973). We targeted thin limestone beds exposed in the middle Muskwa sequence 6

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

(upper George Fm.) for

87

Bellefroid et al.

Sr/86Sr work. Although mainly limestone, the sampled beds are partially

dolomitized and thus constitute an important test of the sequential leach approach for isolating leastaltered (non-dolomitic) fractions that are most likely to approach primary marine 87Sr/86Sr values. Based on bulk leach geochemistry and petrographic analysis, we selected 21 samples from Bellefroid et al. (in review) for 87Sr/86Sr analysis. The Jixian Group is a well-dated 1.6-1.4 Ga Mesoproterozoic carbonate passive margin sequence within

T

the North China craton that is exposed north of Beijing, China (Qiao et al., 2007; Lu et al. 2008; Li et al.,

IP

2010; Su et al., 2010: Meng et al., 2011). The basal three formations of the Jixian Group (Yangzhuang

CR

Fm., Wumishan Fm. and Hongshuizhuang Fm.) consist of dolomite stromatolites, microbialites, and dolomicrites, interbedded within thin and bituminous shales, and rare limestone microbialites (Chu et al., 2007; Zhang et al., 2007). The upper Jixian Group (Tieling Fm.) is a mixed limestone and dolomite

US

sequence (Lu et al., 2008; Guo et al., 2013). Exposed in a series of structurally rotated and deformed blocks (Lu et al., 2008), the deformation and preservation of Jixian Group lithologies vary significantly in

AN

each block. In addition, the Jixian Group is well-known for its large and abundant syn-sedimentary ferromanganese ore deposits (mainly Mn-carbonate minerals: rhodochrosite and chamosite; Fan and

M

Yang, 1999; Fan et al., 1999a; 1999b; Yeh et al., 1999), which are thought to have formed in a carbonate ramp setting in association with weakly oxic, Mn-rich seawater (Fan and Yang, 1999).

ED

A previous study of Jixian Group carbonate lithologies found 87Sr/86Sr between 0.7071 and 0.7089 (Hong et al., 1994), far more radiogenic than current best estimates for typical Mesoproterozoic seawater

PT

(Shields and Veizer, 2002). This early study used bulk-leaching methods without a pre-leaching step. To improve our understanding of Mesoproterozoic seawater 87Sr/86Sr and explore the potential for sequential 87

Sr/86Sr values, we selected 15 samples from a

CE

leaching methods to identify least-altered carbonate

combination of drill cores and outcrop samples from near Dahebei, western Lingyuan, Liaoning province,

AC

and near Kuancheng, Hebei province.

3. GEOCHEMICAL METHODS Using a method similar to Liu et al. (2013; 2014), we streamlined a multi-step sequential leaching procedure that systematically and iteratively removes weakly bound and sorbed ions sourced from aluminosilicate phases, in order to isolate sample fractions preferentially enriched in Sr associated with carbonate phases. Based on trial and effort, we find that effective separation between aluminosilicate and carbonate phases can be obtained in 7 leaching steps (vs.15 leaching steps in Liu et al., 2013, 2014).

7

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

For the Bird Spring Fm., we took powdered samples splits which were previously studied in Brand (2004) and Brand et al., (2012). Brand et al. (2012) billeted whole-rock samples from the Bird Spring Fm. for petrographic screening and then mechanically separated samples prior to powdering (for a detailed description see Brand et al., 2012). For the Dhaiqa Fm., billeted whole-rock samples from Miller et al., (2008) were petrographically assessed for alteration and then micro-drilled, avoiding clear signs of alteration (e.g., secondary veining and micro-fractures). Whole-rock samples from the Jixian Group and

T

the Muskwa Assemblage were billeted and a subset of samples were petrographically assessed before

IP

being micro-drilled.

CR

Sample powders (100 mg) were transferred to 15 ml centrifuge tubes to carry out the seven leachate steps. Successive leaching treatments were performed on remaining solid phase sample material. To remove weakly clay-sorbed ions, samples were first pre-leached using 7 ml of 1 N ammonium acetate, an ion-

US

exchange salt solution (Banner et al., 1988; Kupecz and Land, 1991; Montanez et al., 1996; Bailey et al., 2000). Pre-leached samples were then dissolved in four successive 8 ml steps of 0.04 M acetic acid (S1-

AN

S4, pH ~3.1), followed by two 8 ml steps of 0.175 M acetic acid (S5-S6, pH ~ 2.7). For each step, samples were reacted for 15 minutes in an ultrasonic bath and then centrifuged. Following centrifugation,

M

the supernatant (leachate) was transferred to an acid-cleaned Teflon beaker, dried down and dissolved in 5% HNO3 for trace element analysis. The insoluble residue was then dried and weighed to calculate the

ED

fraction of dissolved carbonate.

Major and trace element concentrations, as well as Sr isotopes for each sample leachate fraction, were

PT

determined in a Pico-Trace clean laboratory at the Yale Metal Geochemistry Center. Sequentially leached samples were dissolved in acid-cleaned 15ml plastic centrifuge tubes using Trace metal grade acetic acid.

CE

A 5% split from each dissolved leachate was dried in clean Teflon beakers and then reloaded with distilled ultra-pure 5% HNO3 to a dilution factor of 1000 for major and trace element analysis on a Thermo Element XR ICP-MS. Calibration standards were made in-house using certified single element

AC

standards and were mixed to match typical carbonate concentration ranges. Geological standards (USGS COQ-1 carbonatite standard and USGS BHVO-2 basalt standard) were run alongside samples as external reference standards. Both external geological standards were within 10% of reported values for each element, with most elements within 5% of reported values. Long-term average machine errors and blank values for each element are included with trace element data (see supplemental material). Though all samples were pre-leached using ammonium acetate, only a fraction of pre-leach steps where measured for major and trace-element concentrations and 87Sr/86Sr. Samples selected for Sr isotope analysis were purified by column chromatography using a Sr-Ca separation protocol on an ESI PrepFast SC-4DXS (Romaniello et al., 2015), using in-house distilled acids 8

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

for elution. Strontium isotopes were measured on a Thermo Neptune MC-ICP-MS on LR mode using a standard-sample bracketing method with NIST SRM-987. Four Sr isotopes were measured simultaneously (84Sr, 86Sr, 87Sr, 88Sr) as well as 85Rb and 84Kr to track potential interferences. In addition, USGS COQ-1 and an in-house modern carbonate standard were used as procedural standards. Runs were linearly corrected to a NIST SRM-987 value of 0.710260. Over the measurement interval for sample unknowns, multi-run standard averages for NIST SRM-987 were 0.710260 ± 0.0000049, and for USGS

T

COQ-1 were 0.703319 ± 0.000015, which agree with findings from other authors (Shields and Veizer,

IP

2002; McArthur et al., 2012). Single day standard averages for each data set are included in each table (see supplemental material). Although USGS COQ-1 has not been measured previously, the Oka

CR

carbonatite complex from which COQ-1 was sampled has a very homogeneous 87Sr/86Sr value of 0.70331 ± 0.00002 (Grünenfelder and Tilton, 1986).

US

4. RESULTS

AN

4.1 Bird Spring Formation, Arrow Canyon, NV

Sequential leach fractions for Bird Spring Fm. carbonate samples yield a wide range of 87Sr/86Sr (range: 0.70818 - 0.70901), but we found that leachate fractions obtained lower minimum 87Sr/86Sr values than

M

were measured in similar samples by a single bulk leach approach (Brand et al., 2012) (Table 1, see

ED

supplemental material for complete results). The 87Sr/86Sr values obtained on successive leachate fractions show generally U-shaped or broadly flat profiles for middle steps with only small shifts (e.g., Fig. 3A).

PT

Lowest 87Sr/86Sr values were typically obtained in steps 3 and 4, with some sample to sample variation.

CE

Similar to 87Sr/86Sr values, minimum Rb/Ca values occur in steps 3 and 4. In the initial pre-leaching (N1) and initial acid leaching steps (S1-S3), Rb/Ca is typically high but shows a decreasing pattern towards nadir levels in steps S3 and S4. Later steps (S5-S6, for this example S6 only) typically increase in Rb/Ca, 87

Sr/86Sr. Mn/Sr ratios, a proxy used to track diagenesis from meteoric fluids (Brand and

AC

Al/Ca and

Veizer, 1980), are uniformly low across the entire suite of samples, with the exception of sample A40A1M, which has systematically higher

87

Sr/86Sr in all leachate fractions. Outside of this sample, Mn/Sr

values across the sample suites are all below 1.0, a cut-off value often used to identify diagenetic alteration (e.g., Montanez et al., 1996; Montanez et al., 2000). Bird Spring Fm. samples generally have low Mg/Ca values (mean 0.0083, range 0.0026 to 0.09), with only a few samples showing increased Mg/Ca values in their terminal leaching steps (S5, S6). However Mg/Ca values are higher than typical aragonite ([Mg] = ~700 ppm). Nadir

87

Sr/86Sr values for each sample show strong correlations with

corresponding low Mg/Ca values. 9

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Though nadir

87

Bellefroid et al.

Sr/86Sr values for each sample are slightly more radiogenic than in coeval brachiopods

(Brand and Brenckle, 2001), which are thought to be the best material for preserving primary 87Sr/86Sr values (McArthur et al., 1993), these results are a significant and systematic improvement in

87

Sr/86Sr

values for whole-rock carbonate samples relative to bulk leaching methods. In addition, these nadir 87

Sr/86Sr values have a much narrower range (0.70818 - 0.70834) than found in bulk-rocks from the same

section (Brand and Brenckle, 2001; Brand, 2004; Brand et al., 2012), excluding one outlier (0.70894)

T

with uniformly elevated Mn/Sr ratios.

Mg/Ca (ug/ug) 0.005 0.005 0.005 0.005 0.004 0.004 0.006 0.013 0.007 0.010

US

Rb/Ca (ug/ug) 3.34E-07 3.52E-07 1.72E-07 2.34E-06 7.66E-07 7.07E-07 2.65E-06 9.74E-07 1.29E-07 5.95E-07

AN

Mn/Sr (ug/ug) 0.17 0.17 0.20 0.56 0.47 0.55 1.09 0.84 0.53 0.18

M

Leaching Step S5 S4 S5 S3 S5 S5 S3 S6 S6 S5

87

Sr/86Sr 0.70823 0.70829 0.70825 0.70829 0.70822 0.70818 0.70834 0.70829 0.70821 0.70829

Error (2-σ) 1.55E-05 2.93E-05 1.87E-05 1.88E-05 2.06E-05 3.90E-05 5.91E-05 1.30E-05 1.50E-05 3.22E-05

ED

Sample AC4-2M AC5-3M AC6-4M AC9-22M AC13-3M AC15-2M A31-1M A42-1M A43-1M A63-22M

CR

IP

Table 1: Trace element ratios and 87Sr/86Sr values for the least-altered step from each Bird Spring Fm. sample (excluding A40A-1M outlier). See supplemental material for complete data set.

4.2 Dhaiqa Formation, Jibalah Group

87

PT

Similar to the Bird Spring Fm., sequential leachate fractions for Dhaiqa Fm. samples have highly variable Sr/86Sr values (range: 0.70397 to 0.70688), but with V-shaped 87Sr/86Sr and Rb/Ca profiles (Fig. 3B) and 87

Sr/86Sr values for individual sample nadir steps (0.70397 - 0.70517) (Table 2, see

CE

a narrow range of

supplemental material for complete Rb/Ca results). In addition, all 87Sr/86Sr values are lower than reported

AC

bulk leached values for the same samples (Miller et al., 2008), except for sample Dhaiqa-1 and Dhaiqa4A, which have near identical value (Table 2). Unlike the Bird Spring Fm., both Rb/Ca and Al/Ca ratios show V-shaped patterns across successive leach steps for individual samples, with values differing by up to two orders of magnitude. Mn/Sr ratios are also much higher and show a much greater range (0.16 – 8.42) with most samples above the traditional 1.0 cut-off. Mg/Ca values are generally low and consistent across all samples with only rare increases above typical calcite range in terminal leaching steps (range: 0.0055-0.1080, mean: 0.0240).

10

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Table 2: A comparison of

87

Bellefroid et al.

Sr/86Sr values between least-altered steps in sequentially leached samples,

bulk-leached whole-rock samples and back-projected values using sequentially leached data. See supplemental material for complete data set.

87

Sr/86Sr 0.703966 0.704030 0.704411 0.704960 0.704778 0.705167 0.705070

Error (2-σ) 0.80 x10-5 4.61 x10-5 0.68 x10-5 1.21 x10-5 0.82 x10-5 0.77 x10-5 1.04 x10-5

87

US

4.3 Jixian Group

Sr/86Sr 0.703968 0.704207 0.704423 0.704898 0.704841 0.705151 0.705118

T

Step S2 S6 S3 S2 S2 S1 S2

Back-Projected values

IP

Sample Dhaiqa - 1 Dhaiqa - 2 Dhaiqa - 4A Dhaiqa - 6 Dhaiqa - 7B Dhaiqa - 9 Dhaiqa - 10

Bulk Leached Whole-rock Miller et al., (2008) 87 Sr/86Sr Error (2-σ) 0.703944 2.20x10-5 0.704166 1.60 x10-5 0.704429 2.20 x10-5 0.705071 1.90 x10-5 0.705157 2.80 x10-5 0.705356 1.90 x10-5 0.705354 3.40 x10-5

CR

Least-Altered Seq. Leaching Steps

Error 1.98 x10-5 5.16 x10-5 3.01 x10-5 4.84 x10-5 6.55 x10-5 2.63 x10-4 8.31 x10-5

Samples from the Wumishan Fm. show a relatively narrow range of 87Sr/86Sr values across all leachate

AN

fractions (0.70442 - 0.70563) and show V-shaped profiles, but a tight range of

87

Sr/86Sr values among

nadir leachate values (0.70434 - 0.7044) (See supplemental material for complete results). Wumichan Fm.

M

samples also have low Rb/Ca values in initial and middle leaching steps (S1-S4) and Mn/Sr values below 1.0 for all sample leachate fractions. Nadir values for the Wumishan Fm. are below lowest

87

Sr/86Sr

ED

values from other units of this age (Shields and Veizer, 2002; Kah et al., 2007; Kuznetsov et al., 2008). By comparison, leachate fractions for samples of the younger Tieling Fm. have a broader range of Sr/86Sr values (0.70480 - 0.71822), including nadir values (0.70351 - 0.71245), with V-shaped Rb/Ca

PT

87

profiles across leaching steps for each sample (Figure 3C; See supplemental material for complete

CE

results). Mn/Sr values among sample leaching steps are also relatively high, ranging from 0.85 to 19.8, with values >1 for most samples.

AC

4.4 Muskwa Assemblage

Sequentially leachate fractions for samples of the Muskwa Assemblage exhibit a large range of 87Sr/86Sr (0.70520 - 0.73243), including a large range for individual sample nadir steps (See supplemental material for complete results). Both Rb/Ca and Al/Ca are higher than in sample leachates of the Bird Spring and Dhaiqa Fm. samples, and show V-shaped patterns across sequential leachate steps for each sample. Although Mn/Sr values are generally low in nadir steps, many samples are partially dolomitized as shown by much higher Mg/Ca values in the later leaching steps of most samples (S5-S6). Samples with the lowest

87

Sr/86Sr are scattered, but within the range typically observed for the mid-Proterozoic period

11

R2 0.99 0.84 0.78 0.98 0.99 0.94 0.92

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

(Shields and Veizer, 2002). Higher

87

Bellefroid et al.

Sr/86Sr values broadly correlate with higher Mn/Sr and Mg/Ca

leachate values.

5. DISCUSSION 5.1 Tracking aluminosilicate contamination using Rb

IP

T

Traditionally, detrital Sr contamination in leachates for 87Sr/86Sr analysis has been tracked using Al and Rb concentrations (e.g., Montanez et al., 1996; Kah et al., 1999). An insoluble element, Al is concentrated

CR

in detrital aluminosilicate phases and thus high Al concentrations in solution are a strong sign of aluminosilicate dissolution during leaching, indicating the potential for detrital Sr contamination. In

US

addition to dissolution, however, weakly bound aluminosilicate-Sr (clay minerals) may also be released by ion exchange during initial stages of leaching without significant dissolution of aluminosilicate phases (Morton, 1985; Banner et al., 1988; Gao, 1990; Montanez et al., 1996). Therefore, Al may not be a

AN

sensitive proxy of Sr released from clay surfaces.

In comparison to aluminium, Rb tracks both the contamination of weak clay-surface bound ions and

M

aluminosilicate dissolution (Banner et al., 1988; Bailey et al., 2000). Using samples from the Dhaiqa Fm. as an example and assuming that each sample had a similar background detrital aluminosilicate source

ED

with Rb and Sr released in similar proportions when leached, Rb/Ca shows a strong positive correlation with 87Sr/86Sr, unlike Al/Ca (Fig. 4A). In particular, when comparing between early leaching steps (S1 -

PT

S3) and later leaching steps (S4 - S6, Fig. 4B), Al/Ca shows a weak correlation with Rb/Ca for steps S1 S3 as Al less exclusively tracks the release of weak clay-bound surface ions.

CE

Although it is common practice to include a pre-treatment step to remove weak clay-surface bound ions (e.g., Li et al., 2011), the use of Rb allows for a secondary check that weakly bound ions have been

AC

substantially removed from the residual carbonate fraction. Aluminum remains an obvious tool to track the dissolution of detrital phases during leaching, however, it should be complemented by Rb to better track the leaching of weak clay-bound ions.

5.1 Effective Isolation of Least-Altered Carbonate Phases: Bird Spring Formation Case Study Bulk leaching methods extract carbonate while leaving behind detrital aluminosilicate minerals by using the exact volume of acid needed to dissolve only the carbonate phases (e.g., Banner; 2004; Li et al., 2011). This simple method, however, has a drawback that unless acid volumes are precisely calibrated for 12

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

each sample, samples may be overleached and detrital aluminosilicates could be dissolved releasing radiogenic Sr. Therefore, using a bulk leaching method, well-preserved primary carbonate samples may still be contaminated if samples are overleached. In contrast, sequential leaching methods are exempt from the need to “calibrate” acid leaching volumes for each sample as the progression of each sample leach is monitored using the trace element chemistry measured on each leaching step. Thus, it is possible to evaluate each step for detrital Sr contamination and identify the purest carbonate steps from each

T

sample for interpreting 87Sr/86Sr results.

IP

Using samples from the Bird Spring Fm. as a case example, we compared 87Sr/86Sr values for sequential

CR

leached whole-rock samples to those derived from bulk leaching measurements of well-preserved low Mg calcite brachiopods from the same section (Brand and Brenckle, 2001; Brand, 2004; Brand et al. 2012). The latter is considered the most robust mineralogy for preserving primary marine

87

Sr/86Sr values

US

(McArthur et al., 1993). To assess aluminosilicate contamination, we compared Rb/Ca values between successively leached sample fractions. Using sample AC15-2M as an example (Fig 3A), Rb/Ca displays a

AN

U-shaped pattern across leaching steps, where steps N1 and S6 record higher Rb/Ca values relative to intermediate steps S1-S5. High Rb/Ca in step N1 is expected and likely indicates weak clay surface-

M

bound Sr released during the initial pH-neutral ion-exchange step (Bailey et al., 2000). The drop in Rb/Ca values between steps N1 and S1 and generally flat Rb/Ca pattern between steps S1 - S5 likely indicates

ED

that the initial N1 pre-cleaning step was effective at removing the majority of weak clay surface-bound Sr, as reflected in the decreasing 87Sr/86Sr values from N1 to steps S1 - S5. The high Rb/Ca values for S6, in addition to a sudden drop in Ca concentration for this step, likely indicates exhaustion of available

PT

carbonate and partial dissolution of residual aluminosilicate. For this sample, individual step values mirror the Rb/Ca pattern, with more radiogenic

87

Sr/86Sr

87

Sr/86Sr values associated with highest Rb/Ca

CE

leachate fraction. Thus, in this example aluminosilicate contamination has the strongest control on 87

Sr/86Sr variability. We find that aluminosilicate free leaching steps from each Bird Spring Fm. sample

AC

have less radiogenic 87Sr/86Sr values than comparable bulk-leached samples from the same section (Brand et al., 2012), and suggest that the lower

87

Sr/86Sr values obtained should better approximate a primary

seawater composition.

To assess potential sources of alteration, it is useful to evaluate samples as a suite, for example, by comparing

87

Sr/86Sr to alteration proxies (e.g., Mn/Sr, Mg/Ca) to see if they correlate (e.g., Marshall,

1992). However, for samples that are both diagenetically altered and contain aluminosilicate phases, a correlation between alteration proxies and 87Sr/86Sr may be masked by aluminosilicate contamination. As can be seen from the case example of the Bird Spring Fm., sequentially leached whole-rock carbonates

13

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

can effectively isolate least-altered carbonate from detrital contamination for each individual sample, effectively excluding detrital Sr contamination from other potential sources of alteration. To evaluate whole-rock Bird Spring Fm. samples for diagenetic alteration, we evaluated each individual sample for contamination and selected one detrital-free leaching step for each sample as a representative of the carbonate phase for that respective sample (Table 1, see supplemental material for complete results). To evaluate each sample for burial and meteoric diagenesis, we compare these representative 87

Sr/86Sr values with trace element alteration proxies. Bird Spring Fm. samples have a

T

leaching steps

IP

narrow range of Mn/Sr values between 0.17 – 0.84, except for samples A40A and A31, which both have

CR

Mn/Sr values above 1.0 and have much higher 87Sr/86Sr values. The higher Mn/Sr values suggest these two anomalous samples underwent minor meteoric diagenesis, as interpreted by Brand et al. (2012), and are thus excluded from further analysis. We evaluated the remaining samples for dolomitization, which 87

Sr/86Sr. Dolomitizing fluids often include Mg-rich basinal

US

can drive significant alteration of primary

fluids which are likely to contain much higher 87Sr/86Sr values (e.g., Stueber et al., 1984; Connolly et al.,

AN

1990). Using Mg/Ca to assess alteration from dolomitizing basinal and mixed marine sources fluids, Bird Spring Fm. samples show a clear trend between Mg/Ca and 87Sr/86Sr (Fig. 5). From this trend, dolomite87

Sr/86Sr values between 0.708101 and

M

free Bird Spring carbonate with Mn/Sr < 1 extrapolates to

0.708219 (0.708160 ± 0.000059) assuming an original aragonite mineralogy with Mg concentrations of

ED

~700 ppm (Mg/Ca ~ 0.0014 at 35% Ca).

Our estimate of seawater 87Sr/86Sr using sequential leaching and back-calculations from Bird Spring Fm.

PT

whole-rock carbonate samples represent an improvement on previous bulk leached 87Sr/86Sr values of at least 1-2x10-4 (Brand, 2004; Fig. 9 Brand et al., 2012). To place such an offset within context, when using Sr/86Sr to date Cenozoic sediment an error of this magnitude could lead to an age miscalculation of up to

CE

87

10 million years. Therefore, this is an example of where detrital aluminosilicate phases are a source of contamination and the sequential leaching method allows us to effectively isolate the least-altered

AC

carbonate fraction from each sample. From this, it further allows us to use a suite of trace-element proxies to effectively evaluate and extrapolate dolomite-free

87

Sr/86Sr values that should more closely

approximate primary seawater 87Sr/86Sr. 5.2 Removing Detrital Contamination: Dhaiqa Fm., Jibalah Group Ideal whole-rock carbonate samples for 87Sr/86Sr analysis should have negligible aluminosilicate contents where sequential leaching allows for the clear separation and identification of a sample’s least-altered carbonate fraction, such as the Bird Spring Fm. However, the geological record may not always be so accommodating, and for some localities, the best-preserved samples for 14

87

Sr/86Sr analysis may contain

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

small fractions of detrital aluminosilicates (e.g., Stueeken et al., 2017). While non-ideal, it may still be possible to back-calculate least-altered carbonate

87

Sr/86Sr values, effectively removing the effects of

aluminosilicates leaching. Using samples from the Dhaiqa Fm. as a case study, we first assessed each sample for detrital contamination. Unlike samples from the Bird Spring Fm., samples from the Dhaiqa Fm. show V-shaped Rb/Ca patterns across each sample’s leaching steps, with highest values in final leachate fractions (Fig. 3B). The aluminosilicate-rich matrix of Dhaiqa Fm. samples likely exceeded the

T

initial pre-cleaning step (N1) exchange capacity and thus weak clay surface-bound Sr continued to be

IP

released in step S1. In addition, the lower fraction of carbonate minerals in the sample likely means carbonate was exhausted more rapidly during the leach, explaining the rapid increase in Rb/Ca values in

CR

step S3 - S6. As there is no basal plateau in Rb/Ca value across multiple leaching steps (cf. Bird Spring Fm. Fig. 3A), we cannot assume that the lowest Rb/Ca of step S2 is free of detrital Sr contamination.

US

Instead, we can treat each leaching step as a mixing line between carbonate and detrital material, where changes in 87Sr/86Sr values across a single sample represent varying proportions between carbonate and

AN

detrital Sr end-members. Assuming that Rb/Ca tracks the degree of detrital contamination in each leachate fraction, we performed a linear regression between

87

Sr/86Sr and Rb/Ca to derive the

87

Sr/86Sr

value for each sample’s carbonate end-member where Rb/Ca = 0. In addition, for each sample, we

M

calculated the standard error of the regression at a Rb/Ca value of 0 to determine the precision of each linear regression and then calculated the offset between back-calculated and nadir-measured

87

Sr/86Sr

ED

values. It must be noted, that a key requirement is that samples do not have a third mineral phase, such as dolomite, which would cause a deviation between Rb/Ca and 87Sr/86Sr values. In the case of the Dhaiqa

PT

Fm. samples, Mg/Ca values are generally low and a significant impact from a dolomite component can be ignored.

CE

We find that 87Sr/86Sr is strongly correlated to Rb/Ca for almost all sample dissolution steps (R 2 for each sample between 0.787 - 0.997) and that the offset between back-calculated and nadir-measured 87Sr/86Sr

AC

values is very low (<0.00006) (Table 2, see supplemental material for complete results). Once each sample was back-corrected for detrital contamination, we followed a procedure similar to that applied to the Bird Spring Fm. and applied traditional trace element constraints to evaluate the potential influence of burial and meteoric diagenetic alteration. Back-calculated 87Sr/86Sr values for Rb/Ca = 0 are concentrated in two sample groups (Fig. 6). Dhaiqa-1, 2 and 4A all have back-projected 87Sr/86Sr between 0.70397 ± 0.00002 and 0.70442 ± 0.00003, whereas Dhaiqa- 6, 7B, 9, 10 have more radiogenic 87Sr/86Sr between 0.70484 ± 0.000066 and 0.7052 ± 0.00026. These two groups correspond to different Mn/Sr values, samples Dhaiqa-6, 7B, 9 and 10 all have Mn/Sr below 1.15, whereas Dhaiqa-1,2 and 4A have values above 1.4 (Fig. 6). The simplest explanation for these trends is that samples Dhaiqa-1, 2 and 4A are affected by secondary nonradiogenic meteoric or burial fluid-rich alteration. Back-calculated 15

87

Sr/86Sr

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

values for each sample do appear to show a mixing line with Mn/Sr, which is likely between a “primary” carbonate value and a meteoric fluid (Fig. 7). Alternatively, the increase in

87

Sr/86Sr values up-section

may reflect a secular change in basinal 87Sr/86Sr where the higher Mn in the lower section reflects higher dissolved Mn concentrations potentially due to anoxic conditions. In such a scenario, the progressive opening of the basin to marine incursion would lead to an increase in 87Sr/86Sr value, though this would require a strongly anoxic basin given much of the Dhaiqa Fm. is shallow water carbonate above storm

T

wave base (Miller et al., 2008). In either case, there are no underlying trends which suggest that the

IP

Dhaiqa Fm. may have preserved marine 87Sr/86Sr and our best estimate of primary 87Sr/86Sr values for the Dhaiqa Fm. fall between 0. 7048 - 0.7052, which is far below typical Neoproterozoic

87

Sr/86Sr values

CR

(Halverson et al., 2010). This suggests that the Dhaiqa Fm. formed in a semi- to fully restricted basin and could have received considerable Sr from locally eroding unradiogenic basalts. In addition, a subset of 87

Sr/86Sr values below our best estimate of primary basinal

US

samples appear to be altered and record

Sr/86Sr values, suggesting post-depositional alteration towards unradiogenic values (Fig. 7).

AN

87

These results contrast from the generally held assumption that alteration typically drives 87Sr/86Sr values

M

towards more radiogenic ones (cf. Brand et al., 2010). This case study best illustrates that if detrital contamination can be clearly separated, previously unrecognisable trends in alteration proxies can be

ED

discerned (Fig. 6).

PT

5.3 Improving the Precambrian Sr isotope record: The Muskwa Assemblage and the Jixian Group Precambrian carbonate successions are commonly dolomitized, and the few preserved limestone

CE

successions are often diagenetically altered and/or have considerable aluminosilicate content, making them difficult to analyse. As a result, our best estimates of Precambrian seawater

87

Sr/86Sr generally

assume that alteration drives marine carbonate toward more radiogenic 87Sr/86Sr values, and therefore the

AC

lowest preserved 87Sr/86Sr values are least-altered (Shields and Veizer, 2002). However, as outlined above this assumption may not always hold true (Conelly, 1990; Miller et al., 2008; this study). Thus, it is imperative that more systematic leaching procedures, which can identify the direction of alteration, need to be applied to better estimate primary seawater

87

Sr/86Sr values. In addition, assessing Precambrian

carbonate samples for meteoric diagenesis is difficult given the poor constraints on seawater Mn concentrations at that time. Mn concentrations in carbonates are a key alteration proxy for assessing potential meteoric and burial diagenesis as Mn2+ has a similar radius to Sr2+ and there is a strong contrast between current seawater and meteoric Mn concentrations (Banner and Hanson, 1990). It has been empirically observed that the progressive alteration of carbonates by meteoric diagenesis often decreases 16

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Sr while increasing Mn concentrations (Brand and Veizer, 1980). It is likely that Mn concentrations were somewhat higher in the Precambrian ocean than the modern ocean, especially given the much lower oxygen concentrations in the shallow ocean (Wallace et al., 2017). However, seawater Mn concentrations were unlikely to be higher than meteoric fluids with the exception of rare local outliers, such as Mn mineral deposits (Maynard, 2014). A strong correlation between 87Sr/86Sr and Mn/Sr is best explained by alteration, thus Mn and Mn/Sr should remain useful proxies, but absolute values of either Mn or Mn/Sr

T

should be treated cautiously, especially in the Precambrian. To explore the utility of sequential leaching

IP

methods in the yet-poorly resolved Mesoproterozoic and address potential issues surrounding the use of values and better constrain Precambrian seawater 87Sr/86Sr.

87

Sr/86Sr

CR

Mn/Sr, we selected two Mid-Proterozoic successions to evaluate their least-altered carbonate

Sequential leachates for samples from the Mesoproterozoic Wumishan Fm. generally exhibit V-shaped

US

Rb/Ca profiles (see supplemental material for complete results). Following the same back-correction procedure as applied to Dhaiqa Fm., sample projections to Rb/Ca = 0 correspond to

87

Sr/86Sr values

corrected and measured nadir

AN

tightly clustered between 0.70434 ± 0.000095 and 0.7044 ± 0.00013, only a small offset between back87

Sr/86Sr values (<0.0003). This close agreement suggests that detrital

primary marine carbonate

M

contamination in these sample leachate fractions are minimal, and therefore may best approach the 87

Sr/86Sr value. This potential is reinforced by tightly clustered low Mn/Sr 87

Sr/86Sr values among leachate fractions for our

ED

values between 0.32 - 0.55. We find that lowest

Wumishan Fm. samples are significantly below previous 87Sr/86Sr measurements for the Wumishan Fm. (Hong et al., 1994), which were measured from a single bulk sample leaching step, without an initial ion-

PT

exchange step (ammonium acetate). The narrow range of Mn/Sr values among Wumishan Fm. samples from this study may indicate a primary signal, however, it also makes it difficult to determine whether

CE

these samples suffered minor diagenetic alteration and whether that alteration drove 87Sr/86Sr values more radiogenic. Thus, despite the limited aluminosilicate contamination, these

87

Sr/86Sr values should be

AC

viewed with caution.

Sequentially leached samples from the Muskwa assemblage show similar V-shaped Rb/Ca and 87Sr/86Sr profiles (see supplemental material for complete results), again posing the potential problem of leachate fraction contamination driving 87Sr/86Sr values. In addition to high aluminosilicate content, many Muskwa assemblage limestones are also partially dolomitized, thus Sr imported from dolomitizing fluids may impart a second influence on leachate (non-dolomite) carbonate

87

87

Sr/86Sr. To reliably back-calculate detrital contamination-free

Sr/86Sr values, we exclude leachate fractions indicative of dolomite

dissolution (Mg/Ca > 10%) from further consideration. This reduces the number of data-points per sample for each linear regression, increasing the uncertainty of the estimates. Using Rb/Ca as a proxy for 17

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

aluminosilicate contamination, detrital Al-silicate free carbonate 87Sr/86Sr compositions for the Muskwa assemblage extrapolate to values between 0.7049 ± 0.00036 and 0.723 ± 0.00290 (Fig. 8). Offsets between these back-calculated and measured

87

Sr/86Sr values are substantially larger than in other data

sets (up to 0.005) and show very large error on the back-calculation (up to 0.0029), suggesting that something other than leaching-related Al-silicate contamination may impede our ability to back-calculate primary carbonate 87Sr/86Sr values. In this regard, a positive correlation between 87Sr/86Sr and Mn/Sr for 87

Sr/86Sr values. This

T

these samples (Fig. 8), is consistent with diagenetic alteration driving higher

lower

87

Sr/86Sr values. Our lowest measured

IP

relationship supports that least diagenetically altered Muskwa assemblage samples should have both 87

Sr/86Sr value for the ~1.5 Ga Muskwa assemblage

CR

carbonates is 0.70520 ± 0.0000164, which is broadly consistent with current best estimates of midProterozoic 87Sr/86Sr values (0.7046 - 0.7048; Shields and Veizer, 2002).

US

Lastly, samples from the Tieling Fm. show similar V-shaped Rb/Ca and 87Sr/86Sr profiles (Fig. 3C), with exceptionally high Rb/Ca values for all samples and steps, suggesting that nadir 87Sr/86Sr values may be

AN

contaminated by aluminosilicates (see supplemental material for complete results). A similar backcorrection of Tieling Fm. samples to detrital contamination-free carbonate compositions (Rb/Ca = 0) 87

Sr/86Sr values with high uncertainties, between 0.7035 ± 0.000072 and

M

extrapolates to a vast range

0.7125 ± 0.0002. In contrast to the Muskwa Assemblage, back-corrected 87Sr/86Sr values from the Teiling 87

Sr/86Sr value of

ED

Fm. show a broad negative correlation with Mn/Sr (Fig. 9), with a lowest measured

0.70480 ± 0.000015. That Mn/Sr values are very high (0.9 - 19.8) suggests that alteration in this sequence may have driven 87Sr/86Sr to less radiogenic values. The high Mn concentrations are suspect and

PT

could indicate ambient anoxic conditions within the water column (e.g., Black Sea; German et al., 1991), but there is also is no reason to expect a correlation with

87

Sr/86Sr other than alteration. Thus, although

CE

nadir Tieling Fm. 87Sr/86Sr values are close to our best estimate of Proterozoic seawater 87Sr/86Sr values for this time period and would significantly supplement the current paucity of measured basins in this

AC

time period (0.7046 - 0.7048; Shields and Veizer, 2002), we are unable to determine the origin(s) of their alteration. This example is a clear case where the lowest observed 87Sr/86Sr may not necessarily reflect a primary value and caution must be taken when assessing partially altered carbonate sequences for primary 87

Sr/86Sr values.

5.4 Application and Future Work These case studies demonstrate the utility of the sequential leaching method. This method can be used to provide a means to measure shorter timescale shifts in 87Sr/86Sr through the Precambrian, a way to check 18

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

long-standing conclusions about the long-term evolution of the Precambrian Sr isotope record, and provide a path forward in obtaining Sr isotope values in carbonates with siliciclastic contamination. Reconstruction of the Phanerozoic seawater

87

Sr/86Sr record has substantially benefitted from the

abundance of well-dated and well-preserved fossil records, for which petrographic and geochemical screening have relevant analogues for interpreting sample preservation. Constraining the Precambrian 87

Sr/86Sr evolution curve is substantially more challenging. Given the lack of skeletal fauna, decreased

T

availability of carbonate successions (particularly pre-Neoproterozoic), and lower degrees of preservation

IP

with respect to diagenesis and metamorphism (e.g., Ronov, 1964), geochemists are ultimately far more

CR

reliant on Precambrian whole-rock limestones (e.g., Stueecken et al., 2017). Many Phanerozoic limestone sequences also suffer from these issues, where well-preserved fossils may be uncommon or spatially restricted. Thus to compensate, available material will need to be prepared and analysed using modified

US

geochemical methods, like sequential leaching, which can geochemically isolate least-altered carbonate phases and back-calculate 87Sr/86Sr primary values when necessary. Sample selection methods, including

AN

petrographic and geochemical pre-screens to select the best-preserved samples and target least-altered phases using micro-drilling, will, of course, remain critically important. Yet the systematic application of

drives

M

sequential leaching methods to isolate least-altered carbonate phases and identify whether alteration 87

Sr/86Sr values more or less radiogenic can provide a path toward a more complete Sr isotope

ED

record and complement current best practices to robustly measure primary 87Sr/86Sr.

PT

6. CONCLUSION

CE

Through the case study examples explored in this study, we have shown how systematic application of a sequential leaching approach to whole rock limestone samples results in identification of a carbonate sample fraction that is least contaminated by Sr derived from leaching of aluminosilicate phases and

between

87

AC

therefore is more likely to retain a marine

87

Sr/86Sr signature. Recognition of systematic relationships

Sr/86Sr and proxies for the influence of aluminosilicate input (e.g., Rb/Ca) and/or diagenesis

(e.g., Mg/Ca; Mn/Sr) among successive leachate fractions in some data sets, can further define the direction of related

87

Sr/86Sr shift, from which the 87Sr/86Sr compositions of unaffected samples can be

extrapolated. The reasonable proximity of extrapolated (back-corrected) compositions to seawater 87

Sr/86Sr values over the same time periods supports the potential of this approach for minimizing the

effects of diagenetic and/or sample-processing related overprinting on primary marine

87

Sr/86Sr values.

This approach should improve the definition of the seawater Sr isotope evolution curve, or application of

19

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Sr-isotope chronostratigraphy, over time intervals and/or settings where whole-rock limestones are the only available sample materials.

ACKNOWLEDGEMENTS The authors wish to thank A. v.S. Hood and J. Katchinoff for comments which have greatly improved the

T

manuscript. E.J. Bellefroid acknowledges support from the Natural Sciences and Engineering Research

IP

Council of Canada Grant PGS D3 471503 2015, the Geological Society of America and the American

CR

Philosophical Society. C.Wang acknowledges support from the National Natural Science Foundation of China grant 40972021.U. Brand acknowledges Natural Sciences and Engineering Research Council of Canada Discovery grant 7969-2015. This work was supported by the Alternate Earths NASA

AN

US

Astrobiology Institute.

REFERENCES

M

Al-Husseini, M.I. (2014). Ediacaran–Cambrian Middle East geologic time scale 2014: Proposed correlation of Oman’s Abu Mahara Supergroup and Saudi Arabia’s Jibalah Group. GeoArabia 19:2, 107132.

ED

Bailey, T., McArthur, J., Prince, H. and Thirlwall, M. (2000) Dissolution methods for strontium isotope stratigraphy: whole rock analysis. Chem Geol 167, 313-319.

PT

Banner, J.L. (1995) Application of the trace element and isotope geochemistry of strontium to studies of carbonate diagenesis. Sedimentology 42, 805-824.

CE

Banner, J.L. (2004) Radiogenic isotopes: systematics and applications to earth surface processes and chemical stratigraphy. Earth-Science Reviews 65, 141-194.

AC

Banner, J.L. and Hanson, G.N. (1990) Calculation of simultaneous isotopic and trace element variations during water-rock interaction with applications to carbonate diagenesis. Geochim Cosmochim Ac 54, 3123-3137. Banner, J.L., Hanson, G.N. and Meyers, W.J. (1988) Determination of initial Sr isotopic compositions of dolostones from the Burlington-Keokuk Formation (Mississippian): Constraints from cathodoluminescence, glauconite paragenesis and analytical methods. Journal of Sedimentary Research 58. Barnett, A.J. and Wright, V.P. (2008) A sedimentological and cyclostratigraphic evaluation of the completeness of the Mississippian–Pennsylvanian (Mid-Carboniferous) Global Stratotype Section and Point, Arrow Canyon, Nevada, USA. Journal of the Geological Society 165, 859-873.

20

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Bayon, G., German, C.R., Boella, R.M., Milton, J.A., Taylor, R.N. and Nesbitt, R.W. (2002) An improved method for extracting marine sediment fractions and its application to Sr and Nd isotopic analysis. Chem Geol 187, 179-199. Bell, R.T. (1968) Proterozoic Stratigraphy of North-Eastern British Columbia. Geological Survey of Canada Paper 67-68.

T

Bellefroid, E.J., Planavsky, N.J., Hood, A.v.S., Halverson, G.P. and Spokas, K. (submitted) Shallow water redox conditions of the Mid-Proterozoic Muskwa Assemblage, British Columbia, Canada.

CR

IP

Bishop, J.W., Montañez, I.P., Gulbranson, E.L. and Brenckle, P.L. (2009) The onset of midCarboniferous glacio-eustasy: Sedimentologic and diagenetic constraints, Arrow Canyon, Nevada. Palaeogeography, Palaeoclimatology, Palaeoecology 276, 217-243. Bishop, J.W., Montañez, I.P. and Osleger, D.A. (2010) Dynamic Carboniferous climate change, Arrow Canyon, Nevada. Geosphere 6, 1-34.

US

Brand, U. (2004) Carbon, oxygen and strontium isotopes in Paleozoic carbonate components: an evaluation of original seawater-chemistry proxies. Chem Geol 204, 23-44.

AN

Brand, U. and Brenckle, P. (2001) Chemostratigraphy of the Mid-Carboniferous boundary global stratotype section and point (GSSP), Bird Spring Formation, Arrow Canyon, Nevada, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 165, 321-347.

M

Brand, U., Azmy, K., Tazawa, J.I., Sano, H., Buhl, D. (2010). Hydrothermal diagenesis of Paleozoic seamount carbonate components. Chem Geol 278, 173-185.

PT

ED

Brand, U., Jiang, G., Azmy, K., Bishop, J. and Montañez, I.P. (2012) Diagenetic evaluation of a Pennsylvanian carbonate succession (Bird Spring Formation, Arrow Canyon, Nevada, U.S.A.) — 1: Brachiopod and whole rock comparison. Chem Geol 308–309, 26-39.

CE

Brand, U. and Veizer, J. (1980) Chemical diagenesis of a multicomponent carbonate system; 1, Trace elements. Journal of Sedimentary Research 50, 1219-1236. Brand, U. and Veizer, J. (1981) Chemical diagenesis of a multicomponent carbonate system; 2, Stable isotopes. Journal of Sedimentary Research 51, 987-997.

AC

Bryan, D. (1985) Geologic map of the Al-Wajh quadrangle, sheet 26B, Kingdom of Saudi Arabia, Ministry of Petroleum and Mineral Resources, Deputy Ministry of Mineral Resources. Map GM 83, 27. Chu, X., Zhang, T., Zhang, Q. and Lyons, T.W. (2007) Sulfur and carbon isotope records from 1700 to 800 Ma carbonates of the Jixian section, northern China: Implications for secular isotope variations in Proterozoic seawater and relationships to global supercontinental events. Geochim Cosmochim Ac 71, 4668-4692. Connolly, C.A., Walter, L.M., Baadsgaard, H. and Longstaffe, F.J. (1990) Origin and evolution of formation waters, Alberta Basin, Western Canada Sedimentary Basin. II. Isotope systematics and water mixing. Applied Geochemistry 5, 397-413.

21

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Cook, F.A., Clowes, R.M., Snyder, D.B., van der Velden, A.J., Hall, K.W., Erdmer, P. and Evenchick, C.A. (2004) Precambrian crust beneath the Mesozoic northern Canadian Cordillera discovered by Lithoprobe seismic reflection profiling. Tectonics 23. Edwards, C.T., Saltzman, M.R., Leslie, S.A., Bergström, S.M., Sedlacek, A.R.C., Howard, A., Bauer, J.A., Sweet, W.C. and Young, S.A. (2015) Strontium isotope (87Sr/86Sr) stratigraphy of Ordovician bulk carbonate: Implications for preservation of primary seawater values. Geological Society of America Bulletin 127, 1275-1289.

IP

T

Fan, D. and Yang, P. (1999) Introduction to and classification of manganese deposits of China. Ore Geology Reviews 15, 1-13.

CR

Fan, D., Yang, P. and Wang, R. (1999) Characteristics and origin of the Middle Proterozoic Dongshuichang chambersite deposit, Jixian, Tianjin, China. Ore Geology Reviews 15, 15-29.

US

Fan, D., Ye, J. and Li, J. (1999) Geology, mineralogy, and geochemistry of the Middle Proterozoic Wafangzi ferromanganese deposit, Liaoning Province, China. Ore Geology Reviews 15, 31-53.

AN

Gao, G. (1990) Geochemical and isotopic constraints on the diagenetic history of a massive stratal, late Cambrian (Royer) dolomite, Lower Arbuckle Group, Slick Hills, SW Oklahoma, USA. Geochim Cosmochim Ac 54, 1979-1989.

M

German, C.R., Holliday, B.P. and Elderfield, H. (1991) Redox cycling of rare earth elements in the suboxic zone of the Black Sea. Geochim Cosmochim Ac 55, 3553-3558.

ED

Grover Jr, G. and Read, J. (1983) Paleoaquifer and deep burial related cements defined by regional cathodoluminescent patterns, Middle Ordovician carbonates, Virginia. AAPG Bulletin 67, 1275-1303.

PT

Grünenfelder, M.H., Tilton, G.R., Bell, K. and Blenkinsop, J. (1986) Lead and strontium isotope relationships in the Oka carbonatite complex, Quebec. Geochim Cosmochim Ac 50, 461-468.

CE

Guo, H., Du, Y., Kah, L.C., Huang, J., Hu, C., Huang, H. and Yu, W. (2013) Isotopic composition of organic and inorganic carbon from the Mesoproterozoic Jixian Group, North China: Implications for biological and oceanic evolution. Precambrian Research 224, 169-183.

AC

Halverson, G.P., Dudás, F.Ö., Maloof, A.C. and Bowring, S.A. (2007) Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology 256, 103129. Halverson, G.P., Wade, B.P., Hurtgen, M.T. and Barovich, K.M. (2010) Neoproterozoic chemostratigraphy. Precambrian Research 182, 337-350. Hong, L.H.C. and Bo, Q.Z.L. (1994) Strontium isotope compositions of Jixian Middle-Upper Proterozoic stratatype section and their significances. Acta Geosicientia Sinica. Hood, A.v.S. and Wallace, M.W. (2015) Extreme ocean anoxia during the Late Cryogenian recorded in reefal carbonates of Southern Australia. Precambrian Research 261, 96-111. Johnson, P., Halverson, G., Kusky, T., Stern, R. and Pease, V. (2013) Volcanosedimentary Basins in the Arabian-Nubian Shield: Markers of Repeated Exhumation and Denudation in a Neoproterozoic Accretionary Orogen. Geosciences 3, 389. 22

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Johnson, P.R., Andresen, A., Collins, A.S., Fowler, A.R., Fritz, H., Ghebreab, W., Kusky, T. and Stern, R.J. (2011) Late Cryogenian–Ediacaran history of the Arabian–Nubian Shield: A review of depositional, plutonic, structural, and tectonic events in the closing stages of the northern East African Orogen. Journal of African Earth Sciences 61, 167-232.

T

Kah, L.C., Crawford, D.C., Bartley, J.K., Kozlov, V.I., Sergeeva, N.D. and Puchkov, V.N. (2007) C-and Sr-isotope chemostratigraphy as a tool for verifying age of Riphean deposits in the Kama-Belaya aulacogen, the east European platform. Stratigraphy and Geological Correlation 15, 12-29.

IP

Kah, L.C., Lyons, T.W. and Chesley, J.T. (2001) Geochemistry of a 1.2 Ga carbonate-evaporite succession, northern Baffin and Bylot Islands: implications for Mesoproterozoic marine evolution. Precambrian Research 111, 203-234.

US

CR

Kah, L.C., Sherman, A.G., Narbonne, G.M., Knoll, A.H. and Kaufman, A.J. (1999) δ13C stratigraphy of the Proterozoic Bylot Supergroup, Baffin Island, Canada: implications for regional lithostratigraphic correlations. Canadian Journal of Earth Sciences 36, 313-332.

AN

Kupecz, J.A. and Land, L.S. (1991) Late-stage dolomitization of the Lower Ordovician Ellenburger Group, West Texas. Journal of Sedimentary Research 61, 551-574.

M

Kuznetsov, A.B., Semikhatov, M.A. and Gorokhov, I.M. (2014) The Sr isotope chemostratigraphy as a tool for solving stratigraphic problems of the Upper Proterozoic (Riphean and Vendian). Stratigraphy and Geological Correlation 22, 553-575.

ED

Lane, H.R., Brenckle, P.L., Baesemann, J. and Richards, B. (1999) The IUGS boundary in the middle of the Carboniferous: Arrow Canyon, Nevada, USA. Episodes 22, 272-283.

PT

LeCheminant, A. and Heaman, L.M. (1994) 779 Ma mafic magmatism in northwest Canadian Shield and Northern Cordilerra: A new regional time marker., Eighth International Conference on Geochronology, Cosmochronology and Isotope Geology. USGS, Berkeley, California.

CE

Li, D., Shields-Zhou, G.A., Ling, H.-F. and Thirlwall, M. (2011) Dissolution methods for strontium isotope stratigraphy: Guidelines for the use of bulk carbonate and phosphorite rocks. Chem Geol 290, 133-144.

AC

Li, H., Zhu, S., Xiang, Z., Su, W., Lu, S., Zhou, H., Geng, J., Li, S. and Yang, F. (2010) Zircon U–Pb dating on tuff bed from Gaoyuzhuang Formation in Yanqing, Beijing: further constraints on the new subdivision of the Mesoproterozoic stratigraphy in the northern North China Craton. Acta Petrologica Sinica 26, 2131-2140. Liu, C., Wang, Z. and Raub, T.D. (2013) Geochemical constraints on the origin of Marinoan cap dolostones from Nuccaleena Formation, South Australia. Chem Geol 351, 95-104. Liu, C., Wang, Z., Raub, T.D., Macdonald, F.A. and Evans, D.A. (2014) Neoproterozoic cap-dolostone deposition in stratified glacial meltwater plume. Earth Planet Sc Lett 404, 22-32. Lu, S., Zhao, G., Wang, H. and Hao, G. (2008) Precambrian metamorphic basement and sedimentary cover of the North China Craton: A review. Precambrian Research 160, 77-93.

23

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Machel, H.G. and Burton, E.A. (1991) Factors governing cathodoluminescence in calcite and dolomite, and their implications for studies of carbonate diagenesis. SEPM. Marshall, J.D. (1992) Climatic and oceanographic isotopic signals from the carbonate rock record and their preservation. Geological Magazine 129, 143-160. Maynard, J.B. (2014) 9.11 - Manganiferous Sediments, Rocks, and Ores A2 - Holland, Heinrich D, in: Turekian, K.K. (Ed.), Treatise on Geochemistry (Second Edition). Elsevier, Oxford, pp. 327-349.

IP

T

McArthur, J., Howarth, R. and Shields, G. (2012) Strontium isotope stratigraphy, The Geologic Time Scale, pp. 127-144.

CR

McArthur, J., Thirlwall, M., Gale, A., Kennedy, W., Burnett, J., Mattey, D. and Lord, A. (1993) Strontium isotope stratigraphy for the Late Cretaceous: a new curve, based on the English Chalk. Geological Society, London, Special Publications 70, 195-209.

US

McArthur, J.M. (1994) Recent trends in strontium isotope stratigraphy. Terra Nova 6, 331-358.

AN

Meng, Q.-R., Wei, H.-H., Qu, Y.-Q. and Ma, S.-X. (2011) Stratigraphic and sedimentary records of the rift to drift evolution of the northern North China craton at the Paleo- to Mesoproterozoic transition. Gondwana Research 20, 205-218.

M

Miller, N., Johnson, P.R. and Stern, R.J. (2008) Marine Versus Non-Marine Environments for the Jibalah Group, NW Arabian Shield:--A Sedimentologic and Geochemical Survey and Report of Possible Metazoa in the Dhaiqa Formation. Arabian Journal for Science and Engineering 33, 55-78.

ED

Miller, N.R., Stern, R.J., Avigad, D., Beyth, M. and Schilman, B. (2009) Cryogenian slate-carbonate sequences of the Tambien Group, Northern Ethiopia (I): Pre-“Sturtian” chemostratigraphy and regional correlations. Precambrian Research 170, 129-156.

CE

PT

Montañez, I.P., Banner, J.L., Osleger, D.A., Borg, L.E. and Bosserman, P.J. (1996) Integrated Sr isotope variations and sea-level history of Middle to Upper Cambrian platform carbonates: Implications for the evolution of Cambrian seawater 87Sr/86Sr. Geology 24, 917-920. Montañez, I.P., Osleger, D.A., Banner, J.L., Mack, L.E. and Musgrove, M. (2000) Evolution of the Sr and C isotope composition of Cambrian oceans. GSA today 10, 1-7.

AC

Morera-Chavarría, A., Griffioen, J. and Behrends, T. (2016) Optimized sequential extraction for carbonates: Quantification and δ13C analysis of calcite, dolomite and siderite. Chem Geol 443, 146-157. Morse, J.W. and Mackenzie, F.T. (1990) Geochemistry of sedimentary carbonates. Elsevier. Morton, J.P. (1985) Rb-Sr dating of diagenesis and source age of clays in Upper Devonian black shales of Texas. Geological Society of America Bulletin 96, 1043-1049. Nettle, D., Halverson, G.P., Cox, G.M., Collins, A.S., Schmitz, M., Gehling, J., Johnson, P.R. and Kadi, K. (2014) A middle–late Ediacaran volcano-sedimentary record from the eastern Arabian-Nubian shield. Terra Nova 26, 120-129. Page, W.R. and Dixon, G. (1997) Geologic framework of the Arrow Canyon area, Clark County, Nevada.

24

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Prokoph, A., Shields, G. and Veizer, J. (2008) Compilation and time-series analysis of a marine carbonate δ18O, δ 13C, 87Sr/86Sr and δ34S database through Earth history. Earth-Science Reviews 87, 113-133. Qiao, X., Gao, L. and Zhang, C. (2007) New idea of the Meso-and Neoproterozoic chronostratigraphic chart and tectonic environment in Sino-Korean Plate. Geological bulletin of China 26, 503-509. Rauch, H.W. and White, W.B. (1977) Dissolution kinetics of carbonate rocks: 1. Effects of lithology on dissolution rate. Water Resources Research 13, 381-394.

T

Raymo, M. and Ruddiman, W.F. (1992) Tectonic forcing of late Cenozoic climate. Nature 359, 117-122.

CR

IP

Romaniello, S.J., Field, M.P., Smith, H.B., Gordon, G.W., Kim, M.H. and Anbar, A.D. (2015) Fully automated chromatographic purification of Sr and Ca for isotopic analysis. Journal of Analytical Atomic Spectrometry 30, 1906-1912.

US

Ronov, A. (1964) Common tendencies in chemical evolution of Earths crust, ocean and atmosphere. Geochemistry International USSR, 713-&.

AN

Ross, G., Villeneuve, M. and Theriault, R. (2001) Isotopic provenance of the lower Muskwa assemblage (Mesoproterozoic, Rocky Mountains, British Columbia): New clues to correlation and source areas. Precambrian Research 111, 57-77.

M

Shields, G. and Veizer, J. (2002) Precambrian marine carbonate isotope database: Version 1.1. Geochemistry, Geophysics, Geosystems 3, 1 of 12-12 of 12.

ED

Spencer, J.E. and Jonathan Patchett, P. (1997) Sr isotope evidence for a lacustrine origin for the upper Miocene to Pliocene Bouse Formation, lower Colorado River trough, and implications for timing of Colorado Plateau uplift. GSA Bulletin 109, 767-778.

PT

Stevens, C.H. and Stone, P. (2007) The Pennsylvanian–Early Permian Bird Spring Carbonate Shelf, Southeastern California: Fusulinid Biostratigraphy, Paleogeographic Evolution, and Tectonic Implications. Geological Society of America Special Papers 429, 1-82.

CE

Stueber, A.M., Pushkar, P. and Hetherington, E.A. (1984) A strontium isotopic study of Smackover brines and associated solids, southern Arkansas. Geochim Cosmochim Ac 48, 1637-1649.

AC

Stueeken, E.E., Bellefroid, E., Prave, A.R., Asael, D., Planavsky, N. and Lyons, T. (2017) Not so nonmarine? Revisiting the Stoer Group and the Mesoproterozoic biosphere. Geochemical Perspectives Letters. Su, W., Li, H., Huff, W.D., Ettensohn, F.R., Zhang, S., Zhou, H. and Wan, Y. (2010) SHRIMP U-Pb dating for a K-bentonite bed in the Tieling Formation, North China Chin Sci Bull. 55, 3312-3323. Taylor, G.C. and Stott, D.F. (1973) Tuchodi Lakes map-area, British Columbia (94K). Dept. of Energy, Mines and Resources, Gov. of Canada. Tostevin, R., Shields, G.A., Tarbuck, G.M., He, T., Clarkson, M.O. and Wood, R.A. (2016) Effective use of cerium anomalies as a redox proxy in carbonate-dominated marine settings. Chem Geol 438, 146-162.

25

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Veizer, J., Clayton, R.N., Hlnton, R.W., Von Brunn, V., Mason, T.R., Buck, S.G. and Hoefs, J. (1990) Geochemistry of Precambrian carbonates: 3-shelf seas and non-marine environments of the Archean. Geochim Cosmochim Ac 54, 2717-2729. Veizer, J. and Compston, W. (1976) 87Sr/86Sr in Precambrian carbonates as an index of crustal evolution. Geochim Cosmochim Ac 40, 905-914.

IP

T

Vickers-Rich, P., Ivantsov, A., Kattan, F.H., Johnson, P.R., Al Qubsani, A., Kashghari, W., Leonov, M., Rich, T., Linnemann, U. and Hofmann, M. (2013) In Search of the Kingdom’s Ediacarans: The First Genuine Metazoans (Macroscopic Body and Trace Fossils) from the Neoproterozoic Jibalah Group (Vendian/Ediacaran) on the Arabian Shield. Saudi Geological Survey: Jiddah, Saudi Arabia.

CR

Wallace, M.W., Shuster, A., Greig, A., Planavsky, N.J. and Reed, C.P. (2017) Oxygenation history of the Neoproterozoic to early Phanerozoic and the rise of land plants. Earth Planet Sc Lett 466, 12-19.

US

Wen, B., Evans, D.A.D., Li, Y.-X., Wang, Z. and Liu, C. (2015) Newly discovered Neoproterozoic diamictite and cap carbonate (DCC) couplet in Tarim Craton, NW China: Stratigraphy, geochemistry, and paleoenvironment. Precambrian Research 271, 278-294.

AN

Yeh, H.-W., Hein, J.R., Ye, J. and Fan, D. (1999) Stable isotope, chemical, and mineral compositions of the Middle Proterozoic Lijiaying Mn deposit, Shaanxi Province, China. Ore Geology Reviews 15, 55-69.

ED

M

Zhang, C., Wu, Z., Gao, L., Wang, W., Tian, Y. and Ma, C. (2007) Earthquake-induced soft-sediment deformation structures in the Mesoproterozoic Wumishan Formation, North China, and their geologic implications. Science in China Series D: Earth Sciences 50, 350-358.

Figure Captions

PT

Figure 1: General geologic map and stratigraphic column (in meters) of the Bird Spring Fm. type section in Arrow

CE

Canyon (modified from Bishop et al., 2009, 2010 and Brand et al. 2012).

Figure 2: Generalized map of the Al-Warazi quadrangle. The Jibalah basin, located in the southeastern portion of the

AC

map, is juxtaposed onto Proterozoic aged granites but is surrounded on its southern edge by the Baydan Group, a bimodal volcanic unit containing abundant primitive volcanics in the area (Bryan, 1985). Map modified from Bryan (1985) and stratigraphic column modified from Miller et al. (2008).

Fig. 3. A comparison of 87Sr/86Sr (circles) and Rb/Ca (squares) across sequential leaching profiles for different case study units. A: Sequentially leached carbonate samples AC15-2M (filled) and AC63-22M from the Bird Spring Fm. in Arrow Canyon, NV. Both 87Sr/86Sr (circles) and Rb/Ca (squares) show a U shaped pattern across leaching steps N1-S6 in both samples. We suggest that steps N1 and S1 are contaminated by weak clay-surface bound Sr, and that step S6 contains dissolved detrital Sr. Steps S2-S5, however, with lower Rb/Ca are consistent with less detrital

26

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

contamination and are more likely to record a primary carbonate 87Sr/86Sr value. B: Sequentially leached carbonate sample Dhaiqa-6 from the Dhaiqa Fm. Both 87Sr/86Sr and Rb/Ca show a V-shaped pattern across leaching steps S1 S6. We interpret this pattern to suggest that step S1 is contaminated by weak clay-surface bound Sr, and that steps S3 - S6 contain dissolved detrital Sr. Step S2 would appear to be the cleanest step, however it is difficult to determine whether it is contamination free. C: Sequentially leached carbonate sample JQ2-6 from the Tieling Fm. Similar to sample Dhaiqa-6, sample JQ2-6 shows a V-shaped leaching profile.

T

Fig. 4: Comparison of 87Sr/86Sr, Rb/Ca and Al/Ca values for all leaching steps for Dhaiqa Fm samples. A:

IP

Comparing Rb/Ca and Al/Ca with 87Sr/86Sr, Al/Ca does not appear to be effectively tracking detrital contamination at low values. This is potentially a concern because Al is only dissolved during clay and silicate dissolution but not

CR

during leaching. B: Comparing which steps show a poor correlation between 87Sr/86Sr and Al/Ca, steps S1-S3, where

US

weak clay-bound Sr is most likely to be released, Al/Ca clearly does not pick-up this signal.

Fig. 5: Plotted is the Mg/Ca ratio and the 87Sr/86Sr for the cleanest leach from each Bird Spring Fm. sample (circles). Disregarding two samples due to high Mn/Sr values, likely indicating diagenetic alteration, the sample suite can be

AN

back-corrected to an aragonite Mg/Ca of roughly 0.0014 (square). This “calculated primary” 87Sr/86Sr value for this operation is 0.70816, nearly identical to the cleanest leach sample of 0.70818. Though these values are slightly more

M

radiogenic than 87Sr/86Sr measured from pristine brachiopod, this method marks a considerable improvement on

ED

previous bulk-leaching methods (Brand et al., 2012).

Fig. 6: Back-correcting the silicate overprint on each individual sample from the Dhaiqa Fm. using Rb/Ca, there is a

PT

clear distinction between altered (Mn/Sr > 1.4) and unaltered samples (Mn/Sr < 1.15). These values suggest that though the Dhaiqa Fm. was likely deposited in a restricted setting, it was also altered by unradiogenic-sourced

CE

fluids.

AC

Fig. 7: We compare the back-corrected 87Sr/86Sr values of each Dhaiqa Fm. sample with its Mn/Sr value. Though not strong, there appears to be some trend towards a diagenetic fluid composition of 0.704, a value that is well within the range of typical basaltic 87Sr/86Sr values. Errors on each data point represent 2 standard deviation uncertainty on that regression.

Fig. 8: Individual samples from the Muskwa Assemblage have been back-corrected for silicate overprint to a Rb/Ca intercept of 0, and the back-calculated 87Sr/86Sr value are plotted against Mn/Sr ratio. Errors in 87Sr/86Sr are for uncertainty on those regressions (95% confidence). Though the data scatter is too great to reasonably estimate primary carbonate 87Sr/86Sr values, the increase in 87Sr/86Sr with increasing Mn/Sr is indicative that alteration in the Muskwa assemblage drives carbonate samples more radiogenic. Therefore, we can confidently say that the lowest

27

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

measured 87Sr/86Sr values are an upper bound in primary seawater 87Sr/86Sr. Outlier above 0.7180 are not included in the figure.

Fig. 9: Back-corrected 87Sr/86Sr values to Rb/Ca=0 for each sample from the Tieling Fm. compared to average Mn/Sr values. Though no samples show Mn/Sr value below 1.0, the decrease in 87Sr/86Sr values with increasing Mn/Sr clearly shows that Mn/Sr is an unreliable proxy for this sample set and that an Mn/Sr cut-off will not reliably

AC

CE

PT

ED

M

AN

US

CR

IP

T

identify pristine carbonates samples.

28

ACCEPTED MANUSCRIPT Case Studies on the Utility of Sequential Carbonate Leaching for Radiogenic Strontium …

Bellefroid et al.

Highlights

- Carbonate sequential leaching method can identify least-altered 87Sr/86Sr fraction - Method can better separate carbonate fraction from secondary silicate contamination

AC

CE

PT

ED

M

AN

US

CR

IP

T

- Provides a path towards building a more detailed Precambrian 87Sr/86Sr curve

29

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9