Dataset: Radiochemistry data for fluids sampled from the Lost City Hydrothermal Field at the Atlantis Massif in 2018

Fluid samples were collected in September 2018 during the AT42-01 expedition of the R/V Atlantis with the remotely operated vehicle (ROV) Jason II using the Hydrothermal Organic Geochemistry (HOG) sampler that shunts fluid from a titanium intake into pre-cleaned Kynar® bags (Lang and Benitez-Nelson, 2021). Fluid samples were collected into either 2-liter (L) or 11-liter (L) kynar bags that were acid washed (10% HCl soak overnight, followed by thorough rinsing, overnight Milli-Q soak, and further rinsing).

As an initial exploratory test, two fluid aliquots (0.500 L each) from Marker C (J2-1110-LV16) and Sombrero (J2-1111-LV16) were processed at sea before freezing. Radium in the fluid (Raw) was directly co-precipitated with 5 milliliters (mL) of 1M barium nitrate and 25 mL of saturated sodium sulfate, which quantitatively removes radium with the barium sulfate (BaSO4) precipitate from seawater (van der Loeff and Moore, 1999). Although the samples were not filtered, no significant particles were observed prior to precipitation. The BaSO4 precipitates were stored and measured onshore multiple times by gamma-ray spectrometry beginning two weeks later.

On shore, six additional vent fluid samples (0.100 to 0.500 L) were selected for analysis: a duplicate sample from Sombrero (J2-1111-LV16) and single samples from Marker 2 (J2-1107-Chem7), Beehive (J2-1108-Chem3), Marker 8 (J2-1110-Chem5), Marker 3 (J2-1111-Chem4), and Calypso (J2-1108-Chem7). The frozen samples were thawed and the liquid was transferred to a beaker. In some cases, particles that settled quickly were observed in the bottom of the beaker. These particles were not quantified or analyzed. After decanting, the liquid volume was measured and the samples were pH-adjusted to ~7 with 1N HCl as necessary. These six samples were slowly filtered (< 1 L min-1) by gravity through a column containing 15 g of manganese dioxide-coated acrylic fiber (Mn-fiber), which preferentially removes radium, thorium, and actinium from the fluid (Moore 1976, 2008). The efficiency of removal was evaluated by passing each filtrate through a second column of new Mn-fiber.

The two exploratory fluid samples processed at sea were measured in the lab by gamma spectrometry. The activity (decays per minute or dpm) of 223Ra from the sample from Marker C was identified based on characteristic 223Ra and progeny gamma peaks at 144, 155, 269, 271, and 402 keV (Firestone, 1999). This sample was recounted four times over the next 60 days to confirm the presence of dissolved 223Ra (223Raw) based on its 11.4 day half-life. Each measurement agreed within 5% of the others when corrected for 223Ra decay. Final 223Ra activity of this sample is based on the mean of five decay-corrected measurements (16 to 60 days since collection) of the combined 269+271 keV peaks; the error of the measurement is the standard deviation of these measurements. The five counting periods fit an exponential curve (activity versus time) with a half-life of 11.5 days. The exploratory sample from Sombrero was recounted one time. The 223Raw is based on the decay-corrected first measurement; the error is based on the net number of counts recorded in the 269+271 keV peak. The gamma detector was calibrated using a 227Ac standard precipitated with BaSO4 prepared from a solution described by Le Roy et al. (2019). There was no measurable peak at 238 keV where the 228Th - 224Ra couple has a characteristic gamma peak. Because 223Ra has daughters that produce gamma rays in the 351 keV region, where normally 214Pb is measured as a 226Ra proxy, samples were aged for more than three months until all 223Ra had decayed before remeasurement for 226Ra. Measurements made after three months had no activity in the 269+271 keV region, meaning 227Ac (half-life = 21.7 y), the grandparent of 223Ra, was below detection in the two exploratory samples.

The 223Raw activities of six additional samples were determined using a Radium Delayed Coincidence Counter System (RaDeCC system; Moore and Arnold, 1996). This system identifies 224Ra (half-life = 3.66 d) and 223Ra based on the coincidence timing of their decay products. The 223Ra efficiency was determined by the procedure of Moore and Cai (2013) and was verified using a 227Ac standard described by Le Roy et al. (2019). After extraction of the radium, the Mn-fibers were rinsed with de-ionized water and dried with compressed air to achieve an approximate 1:1 fiber to moisture weight ratio. The six fluid samples were counted within a month of collection. Only 223Ra was determined due to 224Ra decay prior to analysis. Most samples were recounted for 900 to 1200 minutes two to four times over the next two weeks to follow the characteristic 223Ra decay. For samples measured multiple times, the decay-corrected mean of these measurements was taken as the final activity; the error of the measurement was taken as the standard deviation of the multiple decay-corrected measurements. One sample recorded zero counts for 223Ra during a 1200-minute period. The 223Ra for this sample is therefore considered to be below the detection limit (bdl). None of the reprocessed samples contained more than 2% of the activity of the initial column, indicating the first extraction was close to 100%. After these initial counts, the Mn-fiber samples were stored for five months to allow all initial 223Ra to decay and were then measured again to determine 227Ac and 228Th (Shaw and Moore, 2002).

After RaDeCC analysis was completed, Ra was leached from the Mn-fiber with a solution of 1N hydroxylamine hydrochloride and 1N hydrochloric acid (van der Loeff and Moore, 1999). Radium was co-precipitated with BaSO4 as described for the initial fluid samples and measured by gamma ray spectrometry (Moore, 1984) to quantify long-lived Ra isotopes, similar to the process followed for the exploratory fluid samples. The low activities required five to eight-day counts. We only used the 351 keV peak to calculate 226Ra. No peaks above background appeared in the 228Ra, 228Th, or 227Ac regions of the spectra. The gamma detector was calibrated with NIST standard solutions for 226Ra and 228Ra in a BaSO4 matrix.