Dataset: Size-fractionated major particle composition (POC, PIC, bSi) and concentration (SPM) from the US GEOTRACES GP17-ANT cruise in the Amundsen Sea on RVIB Nathaniel B. Palmer (NBP2401) from November 2023 to January 2024

Filter holders and cleaning protocol:
This was the first U.S. GEOTRACES cruise to use McLane vertical intake filter holders (https://mclanelabs.com/wp-content/uploads/2024/11/MRL-WTSLV-Filter-Holders-reference-24.K.06.pdf). This was necessitated by the at-sea loss of the previously-used "mini-MULVFS" filter holders on a prior cruise. The McLane holders are a commercial product based on the "mini-MULVFS" design that was tested and intercalibrated on the 2008 and 2009 GEOTRACES intercalibration cruises (Bishop et al. 2012) and used for all prior U.S. GEOTRACES cruises (Lam et al., 2018; Lam et al., 2015a; Xiang and Lam, 2020). The new holders retained the dimensions of the mini-MULVFS design and generally produced very good particle distributions on the filters. However, modifications in the design necessitated a change in the cleaning protocols, as the McLane version compresses 32 4"-long tubes between two plates, using an O-ring on either end of each tube for a seal. This requires a complete disassembly for cleaning prior to use. For GP17-ANT, ten McLane filter holders had difficult-to-remove red silicone O-rings embedded into a plate, which have poor resistance to HCl. Eight new McLane holders were purchased disassembled and without the embedded O-rings. For the eight new filter holders, we substituted Viton O-rings, which have much better acid-resistance. The following cleaning procedure was used:

Day 1: All disassembled components except the frit were put in a detergent bath (10% Micro) overnight. The frits were placed in a 10% TM-grade HCl bath.

Day 2: All components in detergent were rinsed thoroughly in DI water. All acid-sensitive items (silicone O-rings, components with embedded silicone O-rings, nylon wing nuts, Delrin rods, CPC quick disconnect fittings with embedded O-rings) were placed in a fresh DI bath to soak overnight to further dilute traces of detergent caught between the embedded O-rings and plates. Acid-resistant components without O-rings were placed in a 10% HCl (reagent grade) acid bath overnight. Frits soaking in 10% TM-HCl were removed, rinsed, and placed into a fresh bath of MQ water to soak.

Day 3: All components soaking overnight in 10% reagent-grade HCl from Day 2 were thoroughly rinsed with MQ water and set to dry in a laminar flow environment. Acid-sensitive fasteners (nylon wing nuts, Delrin rods, quick disconnect fittings) from Day 2 soaking in DI water were given a quick (less than 1 hour) soak in the 10% reagent-grade HCl, then rinsed thoroughly in MQ water. Acid-sensitive items that have direct contact with sample (i.e., through which seawater flows) were placed into a tub of 5% TM-grade HCl for a brief (~2 hours) soak, then removed, rinsed thoroughly in MQ water, then placed into a fresh MQ water tub to further dilute traces of acid caught between the embedded O-rings and plates. After several hours, these were given a final MQ rinse and dried under laminar flow conditions.

Frits retain acid that is not easily rinsed out, so they must be rinsed under vacuum suction: a frit was placed into a cleaned filter holder plate with its vacuum quick disconnect fitting in place, a vacuum was applied to the base of the filter holder plate while pouring MQ water onto the frit to rinse all acid out. Rinse water pH can be monitored to ensure all acid is removed. Rinsed frits should be dried in a laminar flow bench.

Day 4 or later: dried pieces can be reassembled. Frits should be packaged separately, as any residual acid fumes could degrade filter holder components in transit. Frits can take several days to dry completely. It is important that they be completely dry before packaging, as there can be mold growth on them otherwise.

Sampling:
Size-fractionated particles were collected using dual-flow McLane Research in-situ pumps (5 Dual Filter Large Volume Water Transfer System--WTS-LVDF and 3 modified Upright Large Volume Water Transfer System--WTS-LVUP). Filter holders were McLane vertical intake filter holders. At most of the stations on the continental shelf, one cast of up to 8 pumps and two filter holders per pump were deployed to collect samples at 8 depths throughout the water column. At stations off the shelf, up to 16-depths were sampled with two casts. The targeted depths of the wire-out were corrected using pressure readings from a self-recording Seabird 19plus CTD at the end of the line. One filter holder/flowpath was loaded with a Sefar polyester mesh prefilter (51 micrometers (μm) pore size, called the "Qp" filter) and paired Whatman QMA quartz fiber filters (1 μm pore size) in series ("QMA-side"). The other filter holder/flowpath was also loaded with a 51 μm prefilter (called the "Sp" filter), but it was followed by paired 0.8 μm Pall Supor800 polyethersulfone filters ("Supor-side"). A 150 μm Sefar polyester mesh was placed underneath all 51 μm prefilters and QMA filters as a support to facilitate filter handling but not analyzed. All filters and filter holders were acid-leached before use based on the recommended methods in the GEOTRACES sample and sample-handling protocols Cutter et al., 2010). QMA filters were pre-combusted at 450 degrees Celsius (°C) for 4 hours after acid leaching.

A special plate was manufactured for one of the pumps that could hold two additional McLane filter holders that were loaded with full filter sets but not connected to plumbing. These 'dipped blank' filters included the full filter sets (51 µm prefilter on top of paired QMA or paired Supor filters) with a 0.2 µm Supor polyethersulfone filter on top of the 51 µm polyester prefilters to exclude all particles from the main filter set. These were processed identically to the regular filters and functioned as full process blanks. A total of 49 dipped blank filters were used for blank subtraction, calculations of uncertainties, and determination of detection limits (Table 2).

In this dataset, data reported from the 51 μm prefilter are referred to with an "LPT" suffix to indicate large particulate total concentrations (>51 μm). Samples could come from the QMA side ("Qp" filter) or the Supor side ("Sp" filter) and are treated equivalently. Data reported from the main filters (QMA—1-51 μm —or Supor—0.8-51 μm) are from the top filter of the pair only, and are referred to with an "SPT" suffix to indicate the small particulate total concentrations.

Analytical Procedures:

Particulate organic carbon (POC) and particulate nitrogen (PN):
For SPT particles, a 25-millimeter (mm) punch of QMA filter was dried at 55°C at sea. LPT (>51 µm) particles were rinsed at sea from the Sp filter using 0.2 µm-filtered seawater onto a 25 mm 0.8 µm Ag filter (Sterlitech) and dried at 55°C at sea. These samples were beta counted for ²³⁴Th at sea by the CafeTh lab at WHOI, then sent to the Lam lab at UCSC for POC and PN analyses. POC and PN sample processing was similar to what was described in Xiang and Lam (2020). Filter samples were dismounted from the beta mounts. Typically, a 22 mm QMA subpunch or the entire 25 mm 0.8 µm Ag filter were fumed in a desiccator with concentrated HCl and dried in the oven at 60 °C overnight, and then pelletized with 30 mm tin discs from EA Consumables. Tin disc encapsulated samples were measured using a Costech 4010 Elemental Analyzer (EA) or ThermoFisher FlashEACE Instruments NC 2500 model Carbon/Nitrogen Analyzer interfaced to a Delta V Plus Continuous Flow Isotope Ratio Mass Spectrometer (IRMS) at the Stable Isotope Ecosystems Laboratory at the University of California, Merced. C and N concentrations and isotopic compositions were calibrated using the USGS 40 and USGS 41a and EA acetanilide (C8H9NO) reference materials. The isotopic data are expressed in the standard delta notation as per mil deviations (‰) with respect to international standards of Pee Dee Belemnite (PDB) and atmospheric nitrogen. The precision of the internal standard (Pugel) analyzed along with the samples in the run is 0.02‰ for d13C and 0.06‰ for d15N.

Particulate inorganic carbon (PIC):
A UIC Carbon dioxide coulometer at the UCSC Marine Analytical Lab was used for PIC measurement (Xiang and Lam, 2020). Briefly, PIC on SPT QMA punches or 1/16 LPT QMA-side prefilter was converted to CO₂ by addition of 2 N sulfuric acid. CO₂ produced is carried by a gas stream into a coulometer cell where CO₂ is quantitatively absorbed by a cathode solution, reacted to form a titratable acid and measured based on the change in current.

Biogenic silica (bSi):
A 1-hour alkaline leach with 0.2 M NaOH at 85°C was used to leach bSi for both size fractions as described in Xiang and Lam (2020) prior to the measurement on a SEAL Analytical 5-channel AA500 segmented flow nutrient analyzer at the UCSC Marine Analytical Lab.

Particulate trace metals (pTM):
pTM leachable concentrations (SPL): A 1/16ᵗʰ slice of Supor filter was leached using the "Berger Leach", a weak acid leach comprising 25%(v/v) Q-grade acetic acid and 0.02M reagent grade hydroxylamine hydrochloride (Berger et al. 2008). Samples are submerged in 2 milliliters (mL) of Berger Leach solution and heated in a 95°C water bath for 10 minutes, cooled at room temperature for about 1.3 hours, and then centrifuged for 30 minutes at 4100 rpm. Total leach time was 2 hours. 1.5 mL of the supernatant was transferred to a 15 mL Savillex vial and dried for over an hour at 110°C. The residue was redissolved and dried in concentrated HNO3 twice and brought up in 5% HNO3 spiked with 1 ppb In. Two certified reference materials (BCR-414 and PACS-2) were leached routinely alongside the samples to assure the quality of each leach (Table 1).

Analysis by ICP-MS: Sample solutions were analyzed using an Element XR high-resolution ICP-MS (Thermo Scientific) at the UCSC Plasma Analytical Facility. Elemental concentrations were standardized using multi-element, external standard curves prepared from NIST atomic absorption-standards in 5% HNO₃. Instrument drift and matrix effects were corrected using the internal 1ppb In standard and monitored using a mixed element run standard. Concentrations were determined using external standard curves of mixed trace elements standards. If a parameter was measured on the ICP-MS, the isotope and resolution (LR=Low Resolution; MR=Medium Resolution) used for the concentration measurement are indicated in Table 2.

Derived parameters:

SPM (PARTICLEMASS):
Direct comparisons between gravimetric and chemical dry weight have been made in the equatorial Atlantic, and they were found to be quite similar to each other (Bishop et al., 1977). The chemical dry weight is used to estimate SPM for each size fraction in this paper, as the sum of all major particle composition, which is the sum of POM, CaCO₃, opal, lithogenic material (Litho), and Fe and Mn (oxyhydr)oxides.

SPM [μg/L] = POM [μg/L] + CaCO₃ [μg/L] + OPAL [μg/L] + LITHO [μg/L]

+ Fe(OH)₃ [μg/L] + MnO₂ [μg/L]

Note that the resolution of this data is dictated by the lowest resolution of the component parts.

Errors in derived parameters are calculated based on rules of error propagation.

Particulate organic matter (POM):
Particulate organic matter (POM) was calculated from POC by multiplying a weight ratio of 1.88 g POM/g POC (Lam et al., 2011; Lam et al., 2017; Lam et al., 2015b), determined by Nuclear Magnetic Resonance (NMR) measurements of the phytoplankton biochemical composition (Hedges et al., 2002).

POM [μg/L] = POC [μM] × 12 [μg POC/μmol POC] × 1.88 [g POM/g POC]

Calcium carbonate (CaCO₃):
The mass concentration of CaCO₃ was calculated from PIC using a constant weight ratio:

CaCO₃ [μg/L] = PIC [μM] × 100.08 [(μg CaCO₃)/μmol PIC]

Opal:
A hydrated form of silica as SiO₂∙(0.4 H₂O) was assumed in order to calculate the mass concentrations of opal (Mortlock and Froelich, 1989):

Opal [μg/L] = bSi [μM] × 67.2 [μg opal/μmol bSi]

Lithogenic particles (Litho):
Aluminum (Al) is highly abundant in the crust and relatively invariant between two main lithogenic sources, upper continental crust (UCC Al = 8.04 wt%) and bulk continental crust (BCC Al = 8.41 wt%) (Taylor and McLennan, 1995), serving as an appropriate tracer for lithogenic particles. Total particulate Al was determined by ICP-MS after digestion (cf. Xiang and Lam 2020) and provided by our collaborator, Dr. Daniel Ohnemus, for the calculation of lithogenic particle mass. He will upload total particulate trace metals separately. Errors and quality flags for Litho_SPT and Litho_LPT will be added in the future, once Dr. Ohnemus' dataset is submitted. We assumed that most of particulate Al originates from lithogenic sources and use UCC Al wt% to calculate concentrations of lithogenic particles.

LITHO [μg/L] = Al [nmol/L] × 27×10^-3 [μg/nmol] / 0.0804 [μg Al/μg UCC]

Fe and Mn (oxy)hydroxide (Fe(OH)3 & MnO2):
For the small size fraction, these were calculated from the leachable Fe and Mn concentrations using the "Berger Leach" (Berger et al., 2008). Fe oxyhydroxides are treated as Fe(OH)₃ (ferrihydrite approximation) and Mn oxides are as MnO₂ (birnessite approximation). We apply the formula weights of 106.9 g Fe(OH)₃/mol Fe and 86.9 g MnO₂/mol Mn, respectively.

Fe(OH)₃ [μg/L] = (leachFe [nM]) × 106.9 [(ng Fe(OH)₃)/nmol Fe] × [10^-3 μg/ng]

MnO₂ [μg/L] = ( leachMn [nM] × 86.9 [(ng MnO₂)/nmol Mn] × [10^-3 μg/ng] )

For the large size fraction, where Berger Leaches were not conducted, excess Fe and Mn beyond their lithogenic contributions were assumed to be Fe and Mn (oxy)hydroxide and calculated by subtracting the lithogenic Fe and Mn from the totals. The UCC Fe/Al (0.2106) and Mn/Al (0.00367) ratios were used in the lithogenic corrections. Total particulate Al, Fe, Mn were determined by ICP-MS after digestion (cf. Xiang and Lam 2020) and provided by our collaborator, Dr. Daniel Ohnemus, for the calculation of lithogenic Al, Fe, and Mn. He will upload total particulate trace metals separately. Errors and quality flags for Fe(OH)3_LPT and MnO2_LPT will be added in the future, once Dr. Ohnemus' dataset is submitted. Fe oxyhydroxides are treated as Fe(OH)₃ (ferrihydrite approximation) and Mn oxides are as MnO₂ (birnessite approximation). We apply the formula weights of 106.9 g Fe(OH)₃/mol Fe and 86.9 g MnO₂/mol Mn, respectively. Negative numbers were set to 0.

Fe(OH)₃ [μg/L] = ( Fe [nM] - (Al [nM] × 0.2106 [nmol Fe/nmol Al]) )

×106.9 [(ng Fe(OH)₃)/nmol Fe] × [10^-3 μg/ng]

MnO₂ [μg /L] = ( Mn [nM] - (Al [nM] × 0.00367 [nmol Mn /nmol Al]) )

× 86.9 [(ng MnO₂)/nmol Mn] × [10^-3 μg/ng]