Dataset: GN01 Dissolved and Labile Cobalt

Samples were collected using the ODU GEOTRACES Carousel (GT-C), filtered through 0.2 um Acropak filters in the GEOTRACES clean van and immediately refrigerated. Acid-washed 60 mL LDPE bottles were filled entirely, leaving no headspace. Total dissolved cobalt samples were kept at ambient pH, preserved in a heat-sealed bag containing gas absorbing satchels immediately after sample collection and kept at 4 degrees C until analysis upon return to the Woods Hole laboratory. Labile cobalt were kept at ambient pH, and sampled in separate bottles stored in plastic bags without gas absorbing satchels at 4 degrees C.

The carousel was used to collect samples from surface to near bottom waters. Additional samples were collected from a surface pump when sea ice conditions permitted. Pre-conditioned, teflon-coated 12L Go-Flo sampling bottles (General Oceanics, Miami, FL) were deployed on a polyurethane powder-coated aluminum rosette with titanium pilings and pressure housings (Sea-Bird Electronics, Inc., Bellevue, WA) attached to a Kevlar, non-metallic conducting cable. For more information regarding carousel deployment, please refer to the GEOTRACES Cookbook, cited below and located on the GEOTRACES Program website (https://www.geotraces.org/). Following carousel retrieval, Go-Flo bottles were moved to the GEOTRACES Program class-100 trace metal clean van, and pressurized with HEPA filtered air for sampling in accordance with published methods (Cutter and Bruland 2012). Surface pump samples were collected by zodiac, and a trace metal clean peristaltic pump following the GEOTRACES Program Cookbook sampling recommendations, and in accordance with previous collection protocols (Bruland et al. 2005).

Sample storage and reagent bottles were soaked for >1 week in the acidic detergent Citranox, rinsed thoroughly with 18.2 M-Ohm Milli-Q water (Millipore), filled with 10% HCl to soak for 10 days, and rinsed thoroughly with Milli-Q water adjusted to pH 2 with TM-grade HCl. Reagent purification protocols were identical to those previously published (Saito and Moffett 2001).

Concentrations of total dissolved cobalt and labile cobalt were determined using a previously described cathodic stripping voltammetry (CSV) method (Saito and Moffett 2001, Saito et al. 2004). Measurements were made using the Eco-Chemie µAutolabIII systems connected to Metrohm 663 VA Stands equipped with hanging mercury drop electrodes and Teflon sampling cups. Sample preparation was modified slightly to accommodate use of a Metrohm 837 Sample processor, operated with NOVA 1.8 software (Metrohm Autolab B.V.).

For total dissolved cobalt analyses, samples were UV-irradiated for 1 h prior to analysis in a Metrohm 909 UV digester to degrade the organic ligands that bind cobalt, enabling full chelation by the added electroactive ligand, dimethylglyoxime (DMG). 11 ml of sample was pipetted into 15ml polypropylene tubes. Recrystallized DMG (0.1M in methanol) was added to a final concentration of 400 uM and purified N-(2-hydroxyethyl)piperazine-N-(3-propanesulfonic acid) (EPPS) buffer (0.5 M in Milli-Q water) was added to a final concentration of 3.8 mM. Tubes were inverted several times before being placed in the autosampler queue, where 8.5 ml of the mixture was dosed into the teflon analysis cup using a 800 Dosino automated burrette (Metrohm). 1.5 ml of purified sodium nitrite (1.5M in Milli-Q water) was added directly to the Teflon cup using a dedicated 800 Dosino burrette. Once loaded, samples were purged with high purity (>99.99%) N2 for 3 min and cobalt concentrations were determined by standard addition, with triplicate measurement of the sample followed by four 25 pM cobalt additions. 5 nM Co stock was diluted from a certified 1ppm reference (SPEX) and added to the analysis cup via a third Dosino burrette.

For labile cobalt analyses, 11 ml of sample were pipetted into acid washed teflon vials, preconditioned with a small aliquot of sample water. DMG was added to a concentration of 400 uM and allowed to equilibrate for 8-16hr in the dark prior to analysis (Saito et al. 2004). EPPS was added just prior to measurement using the same standard addition protocol described above. Previously, we determined that natural cobalt ligands in seawater have a conditional stability constant of >10^16.8 (Saito et al. 2005). This suggests that the cobalt is very strongly bound to ligands. Thus, we define labile cobalt as the fraction of total dissolved cobalt that is either bound to weak organic and inorganic ligands in seawater or present as free Co(II), and is then exchangeable with the complexing agent (DMG) used for analysis (Saito et al. 2004, Saito et al. 2005). The difference between the total dissolved cobalt and the labile cobalt can then be used as an estimation of the strong cobalt ligand concentration. The same electrode was used for both total and labile analyses. Before transitioning from labile analyses to total analyses, the autosampler tubing and Teflon cup were rinsed with Milli-Q water and pH2 HCl in Milli-Q water.

The analytical blank was determined by analyzing seawater that had been UV-irradiated for 1 h, equilibrated overnight with prepared Chelex 100 resin beads (Bio-Rad), and UV-irradiated a second time to degrade any leached synthetic ligands. For in lab analyses, mean blank was 2.5 +/- 0.7 pM (n=19), and the detection limit was 2.2 pM.

Intercalibration: GEOTRACES standard seawater and internal standard lab seawater were analyzed periodically. A large batch of UV oligotrophic seawater was generated prior to sample analyses and used as a primary standard during measurements. This standard seawater was run along with each sample batch, as were blanks, and values were consistent for the duration of analyses. Our laboratory has participated in the GEOTRACES intercalibration effort using this electrochemical Co method. Acidified standards were neutralized with concentrated ammonium hydroxide (Seastar), mixing the entire sample between drops, prior to UV digestion. We report our laboratory values for the GEOTRACES and SAFe standard analyses using this electrochemical method to be: SAFe D1 = 47.9 +/- 2.1 (n=3), SAFe D2 = 45.2 +/- 2.1 (n=3), GEOTRACES GSP = 2.4 +/- 1.8 (n=3), GEOTRACES GSC = 77.9 +/- 2.8 (n=3). These results are in good agreement with those from the GEOTRACES intercalibration effort for Co and demonstrate that the methodologies employed to produce this dataset detect concentrations within the standard deviation of current consensus values for UV irradiated samples, which can be found on the International GEOTRACES Program website (https://www.geotraces.org/).