Radium and Radon Radioisotope Activity Data from Samples Collected Between May 2019 and September 2020 Along the West Florida Shelf (Amberjack and Green Banana Blue Holes)

Two sampling trips in May 2019 (USGS FAN: 2019-328-FA, referred to as 19WFS01) and September 2019 (USGS FAN: 2019-357-FA, referred to as 19WFS03) were made to the Amberjack Hole located approximately 50 km west of Sarasota, Florida on the West Florida Shelf at 27.28740, -83.16030. Another sampling trip in September 2020 (USGS FAN: 2020-317-FA, referred to as 20WFS01) was made to Green Banana, which is located approximately 80 km west of Sarasota at 27.13712, -83.43993. Throughout all field excursions, samples containing radioisotopes of Radon-222, Radium-223, Radium-224, and Radium-226 were collected at select depths for each sampling and control site.

Radon-222 samples were collected by lowering a rosette water sampler to the desired depth, collecting the seawater, and transferring six liters of the sample into an 8 L Nalgene high density polyethylene jerrican (hereafter referred to as jerrican) with a three-port cap fitted with tubing as described by Stringer and Burnett (2004). The planned sampling scheme for the May and September 2019 trips was to collect samples at depths of 1, 10, 20, 30, 60, and 85 meters. However, due to user error during sample collection, Radon-222 samples were collected above Amberjack Hole during 19WFS01 using feet for depth instead of meters, resulting in samples from 1, 10, 20, 30, 60, and 85 feet. On the following day, two deep samples were collected within Amberjack Hole at 60 and 85 meters. Control samples were collected from control site 1 at depths of 1, 18, and 36 m, please refer to figure 1 of Vargas and others (2022) for details. For 19WFS03, samples were collected at the intended depths of 1, 10, 20, 30, 60, and 85 m. Control samples were collected at control site 1 at updated depths of 1, 15, and 30 m. For 20WFS01, samples were collected at Green Banana sink at 2, 20, 40, 40 (duplicate), 60, 80, 100, and 120 m. In-situ waster salinity measurements were collected and averaged for all depths using a YSI EXO2 multiparameter water quality sonde sensor. Control samples for this site were collected at depths of 2, 20, and 40 m from control site 2. Water was transferred from the rosette to jerricans at a slow rate to prevent turbulent outgassing of Radon-222. All water samples were immediately and completely sealed with Hoffman screw-compressor clamps on the tubing to kink the line, thus preventing further Radon-222 loss from the sample. All Radon samples were taken to St. Petersburg Coastal and Marine Science Center for laboratory analysis. Randon-222 samples were analyzed for all trips within 48 hours of sample collection.

Extraction of radium isotopes from surface water can be achieved through multiple techniques, but the one employed by SPCMSC researchers predominantly depends on the target depth of sampling. Two of the most common approaches are direct and passive sampling, both of which are accomplished by extracting dissolved radium in water onto Manganese-impregnated acrylic fiber (Mn-fiber) (Moore and Reid, 1973). When a large volume of water is exposed to Mn-fibers at flow rates less than two liters per minute (L/min), naturally occurring radium isotopes are quantitatively adsorbed onto the high surface area of hydrated manganese-oxides bound to the fibers producing a representative sample of the radium in surficial seawater (Moore and Reid, 1973; Moore and Arnold, 1996; ASTM D8027-17). Passive sampling involves exposure to sampling water in-situ while direct sampling involves the Mn-fiber encountering a known volume of collected water. During trips 19WFS01 and 19WFS03 the target depths were less than 100 meters, therefore samples were collected directly. For direct sampling, 48.11 L of water was pumped across 20 grams (g) Mn-fiber housed in a flow-through column using a bilge pump equipped with a flow valve adapter to adjust the flow to 2 L/min. For sites with target sampling depth less than 30 m, a direct current (DC) electrical bilge pump – fitted with a commercially available hose, flow valve adapter, marine-resistant wiring, and weights – was lowered to the desired depth and used to pass between 18.93 and 50 L of water directly over the Mn-fiber. After the water passed through the column, the volume and water quality parameters were measured in an outflow collection reservoir. When target depths were between 30 and 100 m, a large-volume, high flow pump (for example, Grundfos Redi-flo Pump ©) was lowered to the target depth and water was collected in a 50 L collection reservoir. The volume and water quality parameters of the sample were determined from these reservoirs. Subsequently, an alternating current (AC) or DC bilge pump was lowered into the collection reservoirs and the water was passed across the Mn-fiber as described for shallow water samples. At Amberjack Hole samples were collected at 1, 10, 20,30, 60, and 85 m during both sampling expeditions. At the control site samples were collected 1, 18, and 36 m during 19WFS01, and 1, 15, and 30 m during 19WFS03.

A passive sampling approach was used to collect radium isotopes during the 20WFS01 field trip due to sampling at depths greater than 100 meters. Unlike direct sampling, the volume of water sampled using a passive approach could not be measured, thus only total radium isotope activity on the fiber was determined. For passive sampling, 20 g of Mn-fiber was placed in a porous bag (for example, a small, 100 square centimeter mesh dive bag). Multiple bags of Mn-fiber were attached at predetermined depths along a down-line mooring. The mooring was deployed for a minimum of 24 hours. The collection date and time when the mooring was retrieved were recorded. Fibers were returned to St. Petersburg Coastal and Marine Science Center for analysis of Radium-223, Radium-224, and Radium-226.

The surface water jerricans were connected to commercially available radon-in-air detectors, Durridge RAD7s, for determination of Radon-222. RAD7s use silicon semiconductor, solid-state alpha detectors to measure alpha particles emitted by Radon-222's daughter isotopes, Polonium-214 and Polonium-218, as they decay. The instrument converts those particle-detector interactions into an electrical signal representing the total counts. Counts are further partitioned to identify the distinct isotope that produced a given signal based on the specific energy of its associated alpha particle. Window A on the RAD7 records the 6.00 mega electron-volt (MeV) signal of Polonium-218 and window C records the 7.69 MeV signal of Pololnium-214. Polonium-218 reaches secular equilibrium with Radon-222 after approximately 30 minutes, therefore, the jerrican water samples were counted in the RAD7’s “Sniff” mode, which determines total Radon-222 based only on Polonium-218 (Durridge, 2017). Due to the half-life of Radon-222 (3.82 days), all measurements were performed within one half-life of sample collection using three different RAD7 detectors. Each RAD7 setup was purged for a minimum of 15 minutes prior to analysis to reduce the instrument’s relative humidity below 10% and dispel any residual gas from previous sample measurements. A relative humidity below 10% is critical for maintaining the best accuracy for the instrument (Durridge, 2017). Thirty-minute background measurements on a five-minute count cycle in “Sniff” mode were run prior to each sample with the jerricans connected to the instrument. The tubing clamps were not yet removed, and the three-way air valves were aligned through the sample bypass column to ensure only background Radon-222 within the system was being measured. For sample analysis, the clamps were removed, and the air valves were adjusted to allow air flow between the RAD7 setup and the jerrican within a closed loop. For the first 60 minutes, the RAD7 continuously pumped air through the tubing setup to purge the dissolved Radon from the water into the headspace within the jerrican and circulate it throughout the entire system volume. After the sample purge, the jerrican was isolated from the air flow of the setup using the three-way valves on the tubing. The Radon-222 purged into the RAD7 counting system was measured on five-minute count cycles in “Sniff” mode for a minimum of three hours to achieve statistically significant counts. After analysis, the sample jerrican tubing was resealed with clamps for at least 30 days to allow all excess Radon-222 to decay away.

CAPTURE software (registered trademark, Durridge, Inc.) was used to download the neccessary raw RAD7 files (.r7raw) that would be used for subsequent data processing to determine Radon-222 activities. Raw files included test number, full date, live time, air temperature, relative humidity, total counts A, and total counts C. Radon in air was calculated by dividing the total counts-per-minute (cpm) in window-A for each detector by the sum of RAD7 efficiencies for that radon detector. The air-water partitioning coefficient (Kw/air) needed to convert radon in air to radon in water was computed using the water temperature, salinity and the six parameters, a1 to b3, in the Weiss equation as determined by the experiments of Schubert and others (2012). The six unitless parameter values used were a1=-76.14, a2=120.36, a3=31.26, b1=-0.2631, b2=0.1673, and b3=-0.027. The combined standard uncertainty (CSU) is the statistical standard deviation of a radiological result. The CSU was calculated by dividing the square root of the total counts in window-A by the total live time, then multiplying by the calculated air-water partitioning coefficient. Additional corrections were made as the ratio of total air volume (volume of tubing/drying column connector, jerrican, and chamber) to chamber air volume. Activity was decay-corrected to time of collection using the Radon-222 half-life (that is, decay/ingrowth correction factor). Sample specific minimum detectable activity and sample specific critical limit were used as metrics to evaluate quantitative meaningfulness of the measured Radon-222 values (Stringer and Burnett, 2004).

Mn-fibers were analyzed for Radium-223 and Radium-224 using the Radium Delayed Coincidence Counter (RaDeCC) and methods described in Moore, 1996. The RaDeCC’s ability to measure Radium-223 and Radium-224 isotopes relies on the short-lived radon daughters, Radon-219 and Radon-220, respectively. Mn-fibers were connected to the RaDeCC in a sample cartridge which is contained within a closed helium-circulated loop. Helium gas strips the short-lived Radon daughters from the fiber and is used a carrier gas to transport these particles to a Zinc-Sulfide coated Lucas cell (Lucas, 1957). Flow of Helium was regulated to allow Radon-219 (half-life of 3.96 seconds) and Radon-220 (half-life of 55.6 seconds) to undergo alpha decay into Polonium-215 and Polonium-216 within the cell. The base of the Lucas cell is coupled to a photomultiplier tube, pre-amplifier, and a high voltage supply sensitive enough to detect photons generated from the interaction of the alpha particles from polonium decay and the scintillator. These analog events and their timing are passed through a delay coincidence circuit that has been tuned to differentiate the mean statistical occurrence (time) of each polonium isotope. The analog counts passing the circuit for each polonium isotope is converted to a digital signal that is then passed to a computer program via a universal serial bus (USB).

Prior to the analysis of Radium-223 and Radium-224, each system’s background radiation level was measured by circulating ambient air through the system. Mn-fiber samples were rinsed with radium-free, Milli-Q© deionized water to remove sea salts and the fibers’ masses were adjusted to roughly 50-100% total moisture content to ensure maximum emanation of Radon from the Mn-fibers (Sun and Torgeren, 1998). The weight was recorded, and fibers were gently pulled apart to increase surface area and placed into sample cartridges where Helium was pumped throughout the system until the flow rate was approximately 0.4 liters per minute. Helium is superior to ambient air as a carrier gas in the counting chamber because more alpha particles from Radon decay can reach the walls of the chamber and interact with the Zinc-sulfide, thus dramatically increasing the system’s efficiency (Moore, 2008). Once the correct flow rate was reached, the helium hose was removed and quickly replaced with the system output hose to create a closed-loop system with the sample and the instrument. The pump was turned on and the analysis was started on the RaDeCC software. After a five-minute delay to allow the radium-radon-polonium decay series to reach secular equilibrium, the RaDeCC software test was reset and left to count until Radium-220 counts reached a minimum of 300 counts or until the test reached 240 minutes.

Raw Radium-223 and Radium-224 data were exported off the RaDeCC software (Scientific Computer Instruments, 2016) into a text file and included: background count rates, count time, Radon-219 and Radon-220 count rates, Radon-219 and Radon-220 total counts, and combined total counts Determination of Radium-223 and Radium-224 via the delayed coincidence counter follows the procedure outlined by Moore and Arnold (1996) and Scholten and others (2010). The count rates for the Radon-219 and Radon-220 circuits had to be corrected for chance coincidence events using the formula found in Griffin and others, 1963. Griffin bases this formula on total count rates and the time each circuit is open. Corrected Radon-219 and Radon-220 count rates were then determined by subtracting the correction factor from the raw count rates. Another adjustment to the Radon-220 data was required, due to Radon-219 and its daughter, Polonium-215. It is possible that two Radon-219 molecules decay while the Radon-220 window is open. This leaves the second Radon-219 and the Polonium-215 decays to be recorded in the Radon-220 channel. A second order correction must be made to produce a final Radon-220 count rate (Moore and Arnold, 1996). This second order correction is not needed for the Radon-219 channel. Radium-223 and Radium-224 count rates were calculated using the final Radon-219 and Radon-220 count rates, respectively, and adjusting for decay, system efficiency, sample volume, and background measurements using the formula: Cx = (CRFinal x / [ DF * Ex]) * (100/V). Cx is the concentration of the desired isotope, CRFinal x is the final count rate of the respective circuit, DF is the decay factor, Ex is the system efficiency, and V is the sample volume.

Following the Radium-223 and Radium-224 analyses, the Mn-fibers were weighed to ensure the total moisture content was between 50-100%. If the weight was too low, additional Milli-Q deionized water was added to the Mn-fiber. Fibers were then placed into 0.5 L borosilicate glass vessels and sealed with Hoffman screw-compressor clamps on the tubing to kink the line, thus preventing gas loss. Vessels were stored for approximately 30 days (8 half-lives of Radon-222) to ensure Radon-222 and Radium-226 were in secular equilibrium. After 30 days, the tubing from the vessels were connected to the input and the pump port of the RAD7 to start analysis for Radon-222. Precise ingrowth time for quantification of the ingrowth correction, which approaches 1.00 at 30 days, was determined from the timestamp the vessel was last sealed and the timestamp at the start of analysis. Decay during analysis was accounted for through a correction factor based on the total count time. Purging the instrument and background measurements were performed as described for analysis of the jerricans with the RAD7. Following the initial purge and the background measurement, the clamps were removed from the sample and the airflow was sent through the vessel and a sample purge was performed for 15 minutes to distribute Radon-222 throughout the entire system. After the sample purge, the test was started and left to count for a minimum of three hours.

CAPTURE software (registered trademark, Durridge, Inc.) was used to download the raw RAD7 files (.r7raw), which included the test number, total counts, live time, % counts A, and counts A. Radium-226 total activity (unit: disintegrations per minute or dpm) was determined as the net count rate (total count rate minus background count rate) divided by the system efficiency (dpm per cpm - provided by Durridge) multiplied by a decay-to-ingrowth correction ratio (approximately 1.00 with this procedure) and air volume correction factor (ratio of total volume of system to chamber volume). Radium-226 specific activity (unit: dpm per 100 L of seawater) was computed by dividing total Radium-226 by the total volume of seawater. Combined standard uncertainty was determined by propagating, using standard statistical procedures, measurement uncertainties and reported instrument uncertainties. Additionally, all samples were run a minimum of two times, each following 30 days to establish secular equilibrium, to evaluate overall reproducibility. A third analysis would be conducted if two measured activities for one sample fell outside the range of analytical error.