2.1. Reagents and equipment
All sample preparation and IC-ICPMS analyses were carried out in the St Andrews Isotope Geochemistry laboratory (StAIG). NMR analyses were carried out in the School of Chemistry at the University of St Andrews. Solutions containing chloride, nitrite, nitrate, sulfate, phosphite, phosphate and in some cases hypophosphite and pyrophosphate were prepared in LDPE bottles from pure reagents (NaCl, Sigma-Aldrich p/n 1.06404.0500; KNO2, Fisher Scientific p/n 11328016; NaNO3, Fisher Scientific p/n 10696842; MgSO4, Fisher Scientific p/n 11377658; Na2HPO­2, Fisher Scientific p/n 222791000; Na2HPO3.5H2O, Fisher Scientific p/n 11994281; Na2HPO4, Acros Organics p/n 204855000; K4P2O7, Fisher Scientific p/n 10378860) dissolved in 18.2 MΩ∙cm-1deionized water, which was generated with a Smart2Pure system. During the analysis, a 1 mmol/L NaOH solution was used, which was prepared each day by dilution of a concentrated stock solution (500 g/L, carbonate-free NaOH, VWR p/n 87938.290). This stock bottle was stored and handled with minimal agitation to avoid mixing with atmospheric CO2 gas that may lead to elevated carbonate concentrations. An aliquot of 0.16 ml was pipetted into 2 L of DI-water, and the pipette was dipped as deeply into the stock bottle as possible to avoid carbonate-enriched solution from the upper layer closest to the lid. The bottle with the diluted 1 mmol/L NaOH solution was only shaken up after purging with N2 to remove air. The headspace was then pressurized with N2 to further avoid ingrowth of atmospheric CO2 during the analysis. For the NMR analysis, 10-20 % heavy water (D2O) was used (Sigma-Aldrich p/n 151882-100G) to prepare a 0.6 ml sample solution.
The ion chromatograph used in this study was a Dionex ICS-6000 from Thermo Fisher, equipped with an AS-AP autosampler, a 37.5 µL sample loop, a gradient pump, an eluent generator with an RFIC degasser, a CR-ATC 600, an EGC 500 KOH cartridge, an AG17-C guard column, an AS17-C analytical column, an ADRS 600 2mm suppressor, and a conductivity detector (Thermo Scientific p/n 061830). It was run with a constant flow rate of 0.5 ml/min. KOH was used as an eluent, and its concentration was ramped from 1 mmol/L to 40 mmol/L over the course of each run. Analyses with pyrophosphate were set to last for 55 minutes, and here the KOH was ramped up between 5.5-23 minutes. After 45 minutes of total run time, the KOH concentration was decreased back to 1 mmol/L over a duration of 5 minutes. Analyses of solutions without pyrophosphate were set to last for 27 minutes, the KOH was ramped up between 5.5-23 minutes and ramped back down between 23-26 minutes. Containers used for samples and standards were soaked in hot 2M HCl overnight and rinsed several times with deionized water prior to use.
The ICP-MS was an Element 2 from Thermo Fisher. It was equipped with a Scott quartz spray chamber and a quartz nebulizer rated for a solution flow rate of 1 ml/min. Argon gas flow rates were 16 L/min for the cool gas flow, 0.8 L/min for the auxiliary flow, and 1 L/min for the sample carrier flow. The RF power was set to 1250 W. Prior to the start of the run, the ICP-MS was tuned with a multi-element solution at a concentration of 1 ppb in 5% HNO3 (Thermo Scientific p/n 1099601). The instrument was operated in medium resolution mode (measured resolution was ca. 4200 Δm/m) to avoid HNO interferences with phosphorus at m/z 31. Oxide % measured determined from UO/O was ca. 3-5%.
The NMR used in this study was a Bruker AVIII 500 MHz NMR instrument equipped with nitrogen cooled broadband cryoprobe. It was operated in proton-decoupled mode with 3000-7000 scans per analysis.