Dataset: Measurement of dissolved CFC-11, CFC-12, CFC-113, and SF6  from Niskin bottles on the US GEOTRACES GP-17 OCE cruise RR2214 on R/V Roger Revelle from December 2022 to January 2023

Samples were collected first from the non-trace metal rosette. A sampling tube was connected to the Niskin petcock. The petcock was opened followed by opening the Niskin air vent. The sampling tube was held upright so that when water flow started all bubbles were pushed out. The tube was then immersed to the bottom of a 300 milliliter (ml) BOD bottle that was placed inside a 1-liter (L) overflow container. Water was allowed to flow until it started running out of the overflow container, the tube was removed from the 300 ml BOD bottle while water was still flowing, the Niskin bottle petcock was closed, then the stopper was placed into the BOD bottle while it was still immersed in the water in the overflow bottle. The BOD bottle was inspected for the presence of bubbles. If there were bubbles, the procedure was repeated.

The UM (University of Miami) tracer lab uses a custom-built purge and trap capillary column gas chromatography method with electron capture detection to measure all five compounds in a single run. The gas chromatograph was an SRI 8610. This system has been used successfully on many CLIVAR, GO-SHIP, and GEOTRACES cruises with an older Shimadzu gas chromatograph. Right before this cruise, the custom-built system was modernized with a new gas chromatograph and new instrument control and data acquisition software. N2O peaks in the chromatogram can interfere with SF6 peaks and by adding a 6-meter length of a mole sieve 5a capillary column to the end of the 60m Gas Pro main column, the elution of N2O was delayed so that it is after the elution of the other four compounds. However, the attempt to quantitate the N2O peak was not successful. This system can measure all compounds in gas standards, air, and water.

Blanks are determined by running a water method without loading a water sample into the sampling loop. The blank peak areas are subtracted from unknown samples peak areas before concentrations are calculated. One or two blanks are run for each station analyzed. Blank values are typically very consistent and very small compared to unknown peak sizes.

One duplicate water sample is collected during each station sampled. Precision is 2% or better for each parameter measured.

Standards are in the gas phase, which is a tank of compressed nitrogen containing all compounds. Standards used on this cruise were prepared at PMEL (Pacific Marine and Environmental Laboratory) and calibrated against the SIO (Scripps Institute of Oceanography) absolute calibration scale. All US-based tracer laboratories use these standards for calibration. One standard is used, and different amounts of that standard are analyzed, using a combination of five different gas sample loops to generate a calibration curve covering the range of unknows analyzed. Typical curves consist of 20 to 30 data points so that the non-liner response of the Electron Capture detector can be determined well. Calibration curves are generated at the beginning of the cruise. The detector response changes that may occur with time are monitored by analyzing a single volume of the standard, two or three times during the analysis of each station. Any drift in response is used to correct the final concentration determination.

Additional calibration curves will be generated at the end of the cruise and during the cruise if instrument conditions change.

Another check on the accuracy of the analytical technique is to measure the concentration of all the compounds in air during the cruise. The air concentration is known, and air measurements should be close to the known values. Air samples are analyzed as time permits, but at least several times during each cruise.

Surface water concentrations should be close to equilibrium with the air concentration at the time of the cruise. These values are typically within 5% of the temperature- and salinity-dependent saturation values.