Stable silicon isotopes (e.g. δ30Si) in sediment biogenic silica (bSi) are widely used as a paleoproxy for marine silicic acid usage by pelagic diatoms. Despite the growing body of work that uses bSi δ30Si signals, there are a lack of δ30Si data on other reactive pools of Si in sediments. This oversight misses valuable information on early diagenetic products and potentially biases existing sedimentary bSi δ30Si, which only quantified bSi fractions not altered by diagenesis. For the first time, we quantified δ30Si among operationally defined reactive Si pools (using a pre-leach of mild acid prior to alkaline digestion) in Mississippi River plume sediments. We compared the δ30Si signal within these reactive Si pools to a traditional alkaline-only digestion of sedimentary bSi. These data offer proof of concept that δ30Si is a higher throughput approach for quantifying isotopic properties among reactive Si pools marine sediments vs. the more laborious (albeit powerful) examination of natural silicon radioisotopes in these chemical leaches.

Core Sampling

Briefly, samples were acquired from the study area using an Ocean Instruments MC-900 Multi-corer, which preserved the sediment-water interface during recovery. Overlying bottom water was removed, cores were sectioned into 1cm slices, homogenized, packed under N2 gas and frozen at -20o C for further analysis.

Operational Definitions

Operational reactive Si pools have previously been defined by Rahman et al. (2016) but for consistency and clarity with previous literature (DeMaster, 1981; Michalopoulos and Aller, 2004; Qin et al., 2012; Wang et al., 2015; Rahman et al., 2016; Krause et al., 2017) it has been restated here. Therefore we use the following nomenclature;

1. Si-HCl: Mild acid-leachable pre-treatment; Highly reactive silica associated with authigenic clays and metal oxide coatings (Michalopoulos and Aller, 2004).
2. Si-Alk: Mild alkaline-leachable digestion completed after acid pretreatment; Frees reactive silica associated with the biogenic silica pool (Michalopoulos and Aller, 2004).
3. Si-NaOH: Harsh NaOH digestion done after Si-HCl and Si-Alk (Rahman et al., 2016; Rahman et al., 2017); Associated with the reactive lithogenic Si (LSi) pool and the comparatively refractory “dark bSiO2” (e.g. sponge spicules and Rhizaria, Maldonado et al., 2019).
4. T-bSi: Following the traditional definition of biogenic silica (DeMaster, 1981), with no acid pre-treatment.

Reactive Silica Pools

Frozen sediment samples were thawed to room temperature (22o C) and triplicate ~50-100 mg subsamples were immediately weighed into 50 mL polyethylene centrifuge tubes. Samples were never dried or ground before/during extractions. Procedural blanks were also prepared in triplicate. Additional subsamples of sediment were dried at 60o C to obtain correction for water content.

Sequential Extractions

The sequential extraction methodology separates silica into operationally defined pools based on kinetics, reaction conditions and reaction sequence (DeMaster, 1981; Michalopoulos and Aller, 2004; Rahman et al., 2016).

Acid Leachable Silica (Si-HCl)

Sediment extractions occurred at room temperature (22o C) using Honeywell Fluka Trace SELECT 0.1 N HCl for 12 hrs, while keeping particles suspended via constant motion. Following centrifugation, the Si-HCl leachate was removed and neutralized. Remaining sediment was rinsed in triplicate with Milli-Q water to remove any residual acid (Michalopoulos and Aller, 2004). As it had previously been demonstrated by Rahman et al. (2016) that the rinses contained minor amounts of Si these rinses were discarded. A weak HCl molarity was purposely chosen to remove metal coatings, authigenic phases, and activate bSi surfaces while not affecting the sequential Si-Alk digestion (Michalopoulos and Aller, 2004).

Mild Alkaline Leachable Si (Si-Alk)

The remaining sediment from the acid pre-treatment was subsequently digested with 0.1 M Na2CO3 (Fisher Scientific Certified ACS) for 20 mins in a 85o C water bath. Following the 20 min timepoint, samples were placed on ice and neutralized to stop the digestion. Following centrifugation, the Si-Alk leachate was removed and stored for further use. The process was stopped after 20 mins to ensure the absence of lithogenic material (DeMaster, 1981; Michalopoulos and Aller, 2004) and certify that the clear majority of solubilized silica present is biogenic. Fresh 0.1 M Na2CO3 was added to the samples and the digestions were continued for a total of 5 hrs (DeMaster, 1981) to completely remove the bSi phase. Concluding after 5 hrs, samples were placed on ice and neutralized to stop the digestion. Following centrifugation, the leachate was removed and discarded. Remaining sediment was rinsed in triplicate with Milli-Q water to remove any residual Na2CO3 and again the rinses were discarded.

Harsh NaOH Digestion (Si-NaOH)

The remaining sediment from the Si-Alk treatment was subsequently digested with Honeywell Fluka 4 M NaOH for 2 hrs in a 85o C water bath. After 2 hrs, samples were placed on ice and neutralized to stop the digestion. Following centrifugation, the Si- NaOH leachate was removed, the remaining sediment was rinsed with Milli-Q water to remove any residual leachate and this rinse was added to the Si-NaOH leachate and stored for further analysis (Rahman et al., 2016).

Traditional bSi Digestion (T-bSi)

Additionally, a second treatment following the traditional definition of biogenic silica (DeMaster, 1981), with no acid pre-treatment was used to derive δ30Si from traditional bSi measurements. New subsamples of sediment were weighed out. 0.1 M Na2CO3 was added to samples and heated in a 85o C water bath for 20 mins to remove the bSi phase. Following the 20 min timepoint, samples were placed on ice and neutralized to stop the digestion. Following centrifugation, leachate was removed and stored for further use. Similar to the Si-Alk digestions, the process was stopped after 20 mins to ensure the absence of lithogenic material.

A 1 ml aliquot of each resulting liquid (Si-HCl, Si-Alk, Si-NaOH and T-bSi) was analyzed for dissolved SiOH4 concentration (dSi) as described by Brzezinski and Nelson (Brzezinski and Nelson, 1986) using the molybdate-blue method on a Genesys 10S UV-Vis Spectrophotometer. The remaining supernatants were concentrated via evaporation at 100o C and stored following DeMaster (1980) in preparation for stable isotope analysis.

Stable Isotope Analysis

Sample purification and isotope analysis were carried out at the University of Bristol Isotope Group laboratories. Concentrated sample fluids were purified via cation ion exchange chromatography (Bio-Rad AG50W-X12, 200-400 mesh cation exchange resin in H+ form). Purified solutions were analyzed in duplicate for Si isotopes (28Si, 29Si, 30Si) using a multi collector-inductively coupled plasma-mass spectrometer (MC-ICP-MS, Finnigan Neptune s/n 1002), equipped with CETAC PFA spray chamber and PFA nebulizer (100ul/min). A standard-sample-standard bracketing procedure with Mg doping following Cardinal et al., (2003) was used to correct for both instrumental mass bias and matrix effects. Additionally, sample and standard solutions were both doped with 0.1 M H2SO4 (ROMIL UpA) and 1 M HCl (in-house distilled) to reduce any matrix effects from anion loading and guarantee matrix matching between sample and standard (Hughes et al., 2011). All isotopic composition results are expressed as δ30Si, corresponding to the silicon isotopic abundances in samples relative to the international reference standard NBS-28 (NIST RM8546, purified quartz sand). Reference standards Diatomite (Reynolds et al., 2007) and LMG08 (sponge) (Hendry et al., 2011) were run in tandem with samples to assess long- term reproducibility. Average measured values are reported as +1.27 ± 0.09‰ (n=75) and -3.47 ± 0.16‰ (n=27) (±SD) respectively, which are well within agreement with published values (Reynolds et al., 2007; Hendry et al., 2011). All samples and standards are consistent with the kinetic mass fractionation law (Reynolds et al., 2007) with the δ29Si = 0.518xδ30Si. Procedural blanks were lower than the detection limit and thus considered negligible on δ30Si of the samples.

Major Metal Compositions and Corrections

Additional thawed/wet sediment subsamples were used for duplicate sequential extractions and digestions (Si-HCl, Si-Alk, Si-NaOH and T-bSi) run as previously described. Supernatants were concentrated via evaporation at 100o C and fluids were reconstituted in 2% HNO3 (in-house distilled) to determine major ion concentrations on an Agilent 7700 Series ICP-MS. The instrument was calibrated using a blank and seven matrix-matched, mixed standards. Internal standardization during analysis was monitored via the addition of (50 μl, 10,000 ppb) 115In and 4Be to all standards and samples. Using Aluminum (Al):Si corrections (Kamatani and Oku, 2000; Ragueneau et al., 2005), both Si-Alk and T-bSi δ30Si signals (‰) and mass of Si released (μmol/g) were adjusted for bias from lithogenic material (however, this was more important for the mass of Si, as isotopic content was derived from 30-minute digestions, opposed to 5 hour digestions for the former).

Organic Matter

Sediment total organic carbon (TOC) and total organic nitrogen (TON) content were analyzed at the Dauphin Island Sea Lab using a Costech elemental combustion system (4010 ECS) following vapor phase acidification to remove carbonates. Briefly, dried sediment samples were placed in a glass desiccator and reacted with reagent-grade 12N HCl vapor for 24 hrs at room temperature. Samples were then dried at 60o C overnight to remove remaining HCl and water content before TOC/TON analyses (Yamamuro and Kayanne, 1995).