Dataset: CICLOPS Dissolved Cobalt

Samples were collected along the coastal Antarctic Shelf from the Amundsen Sea, Ross Sea, and Terra Nova Bay during the CICLOPS expedition on the RVIB Nathanial B. Palmer (cruise ID NBP-1801) from December 11, 2017 to March 3, 2018. Dissolved seawater was collected from full-depth station profiles using a trace metal clean sampling rosette deployed on a conducting synthetic line, both supplied by the U.S. Antarctic Program (USAP), and equipped with twelve 8-liter (L) X-Niskin bottles (Ocean Test Equipment), supplied by the Saito Laboratory. Real-time trace metal rosette operations allowed for the careful collection of seawater from 10 and 20 meters (m) above the ocean floor to study sediment-water interactions within a potential nepheloid layer. After deployment, the X-Niskin bottles were transported to a trace metal clean van and pressurized with high-purity N₂ gas. Seawater samples for dCo analysis were then filtered through acid-washed 0.2-micromole (µM) Supor polyethersulfone membrane filters (Pall Corporation, 142-millimeter (mm) diameter) within 3 hours of rosette recovery.

To minimize metal contamination of samples, all sample bottles were prepared using trace metal clean procedures, including soaking sample bottles for ~1 week in Citranox, an acidic detergent, rinsing with Milli-Q water (Millipore), soaking sample bottles for ~2 weeks in 10% trace metal grade HCl (Optima, Fisher Scientific), and rinsing with lightly acidic Milli-Q water (> 0.1% HCl).

Samples for dCo analysis were collected in 60-milliliter (mL) low-density polyethylene (LDPE) bottles and stored at 4⁰C until analysis. Duplicate dCo samples were collected: one for at-sea analysis of labile dCo and total dCo, and another for preservation and total dCo analysis in the laboratory after the expedition. Preserved total dCo samples were stored with oxygen-absorbing satchels (Mitsubishi Gas Chemical, model RP-3K), which preserve the sample for long-term storage and future analysis (Noble et al. 2017; Bundy et al. 2020). Preserved dCo samples were stored in groups of 6 within an open (unsealed) plastic bag, which was then placed into a gas-impermeable plastic bag (Ampac) with one oxygen-absorbing satchel per 60 mL dCo sample. The outer bag was then heat-sealed and stored at 4°C until analysis.

Total dCo – the combined fractions of labile and ligand-bound dCo, hereafter simply dCo – and labile dCo concentrations were analyzed via cathodic stripping voltammetry (CSV) as described by Saito and Moffett (2001) and modified by Saito et al. (2010) and Hawco et al. (2016). CSV analysis was conducted using a Metrohm 663 VA and µAutolabIII systems equipped with a hanging mercury drop working electrode. All reagents were prepared as described in Chmiel et al. (2022). Most samples were analyzed at sea within 3 weeks of sample collection, and stations 57 and 60 were analyzed for labile dCo at sea and their duplicate preserved samples were analyzed for total dCo in November 2019 in the laboratory.

To measure total dCo concentrations, filtered seawater samples were first UV-irradiated in quartz tubes for one hour in a Metrohm 705 UV Digester to destroy natural ligand-bound Co complexes. 11 mL of sample was then added to a 15 mL trace metal clean polypropylene vial, and 100 microliters (µL) of 0.1 molar (M) dimethyglyoxime (DMG; Sigma Aldrich) ligand and 130 µL of 0.5 M N-(2-hydroxyethyl)piperazine-N-(3-propanesulfonic acid) (EPPS, Sigma Aldrich) buffer was added to each sample vial. A Metrohm 858 Sample Processor then loaded 8.5 mL of each sample into the electrode's Teflon cup and added 1.5 mL of 1.5 M NaNO₂ reagent (Merck). The mercury electrode performed a fast linear sweep from -1.4 volts (V) to -0.6 V at a rate of 5 volts per second (V s⁻¹) and produced a cobalt reduction peak at -1.15 V, the voltage at which the Co(DMG)₂ complex is reduced from Co(II) to Co(0) (Saito and Moffett 2001). The height of the Co reduction peak is linearly proportional to the amount of total dCo present in the sample. Peak heights were determined by NOVA 1.10 software. A standard curve was created with 4 additions of 25 picomoles(pM) dCo to each sample, and a type-I linear regression of the addition standard curve performed by the LINEST function in Microsoft Excel allowed for the calculation of the initial amount of Co present in the sample.

When analyzing labile dCo concentrations, samples were not UV-irradiated so as to only quantify the free or weakly bound dCo not bound to strong organic ligands. 11 mL of labile samples were instead allowed to equilibrate with the DMG ligand and EPPS reagent overnight (~8 hours) before analysis so as to allow time for the labile dCo present in the sample to bind to the DMG ligand via competitive ligand exchange (K > 10^16.8). Labile dCo samples were then loaded onto the Sample Processor and analyzed electrochemically using identical methods as described above for total dCo samples.

Known Issues/Problems:
Many of the dCo and labile dCo values measured were unusually low and below the analytical detection limit of 4 pM. In cases where no dCo or labile dCo were detected (i.e. when no peak was measurable and/or the dCo value predicted was < 0 pM), values of 0 pM were assigned for the purposes of plotting and select statistical analysis and were flagged as not detected (n.d.) as well as <DL in the dataset; although these concentrations are not detectable with our methodology, we believe the incredibly low concentrations of dCo and labile dCo observed on this expedition were meaningful, and that removing these values from our analysis misrepresents the data and would skew the results to appear higher than was observed.