Chlorofluorocarbon Measurements

Section P16S_2005

During the Section P16S_2005, chlorofluorocarbons (CFC-11, CFC-12, and CFC-113) were measured on all 111 stations for a total of 3,078 samples, although the throughput rate of the analytical system necessitated selectively not sampling some Niskin bottles on most casts. The data set was minimally compromised by this procedure by selecting depths in mid-waters of relatively uniform hydrography. The results of this cruise are preliminary and may change by a small percentage after final scrutiny by the principal investigator.

All samples were collected from depth using 10-L Niskin bottles. Bottles had been cleaned prior to the cruise, and all o-rings, seals and taps were removed, washed in deacon solution and propan-2-ol, then baked out in a vacuum oven for 24 h. Of the original 36 bottles initially used, two were lost and replaced, and one was temporarily replaced, repaired and returned. None of the Niskin bottles used showed a CFC contamination during the cruise. All bottles in use remained inside the CTD hanger between casts. All spare bottles were stored on a spare rosette under a tarp, sitting on the main deck. CFC sampling was conducted first at each station, according to WOCE protocol. This reduces contamination by air introduced at the top of the Niskin bottle as water was being removed. A water sample was collected directly from the Niskin bottle petcock using a 100 mL ground glass syringe which was fitted with a three-way stopcock that allowed flushing without removing the syringe from the petcock. Syringes were flushed several times and great care was taken to avoid contamination by air bubbles. Duplicate samples were randomly collected, nominally from every CTD cast. Duplicates were not taken when time was constrained due to a backlog of analyses. Air samples, pumped into the system using an Air Cadet pump, were run about every 2-4 days from a Dekoron air intake hose mounted high on the foremast. These samples were used to check CFC saturation levels in the surface water.

Halocarbon analyses were performed on a GC equipped with an electron capture detector (ECD). Samples were introduced into the GC-ECD via a purge and dual trap system. The samples were purged with nitrogen and the compounds of interest were trapped on a main Porapack N trap held at ~ -20°C with a Vortec Tube cooler. After the sample had been purged and trapped for several min at high flow, the gas stream was stripped of any water vapor via a magnesium perchlorate trap prior to transfer to the main trap. The main trap was isolated and heated by direct resistance to 140°C. The desorbed contents of the main trap were back-flushed and transferred with helium gas over a short period, to a small volume focus trap to improve chromatographic peak shape. The focus trap was also Porapak N and is held at ~ -20°C with a Vortec Tube cooler. The focus trap was flash heated by direct resistance to 155°C to release the compounds of interest onto the analytical pre-column. The analytical precolumn was held in-line with the main analytical column for the first 3 min of the chromatographic run. After 3 min, all of the compounds of interest were on the main column, and the pre-column was switched out of line and back-flushed with a relatively high flow of nitrogen gas. This prevented later eluting compounds from building up on the analytical column, eventually eluting and causing the detector baseline signal to increase.

The syringes were stored in a flow-through seawater bath and analyzed within 8-12 h after collection. Bath temperature was recorded continuously for use in calculating the mass of water analyzed. Every ten measurements were followed by a purge blank and a standard, gas 2.68mL. Time permitting, the surface sample was held after measurement and was sent through the process to "restrip" it to determine the efficiency of the purging process.

For accuracy, the standard, S39, was cross-calibrated to the SIO-98 absolute calibration scale. A 19 point calibration curve was run every 4-9 days for all three halocarbons. Estimated accuracy is ± 2%. Precision for CFC-12, CFC-11 and CFC-113 is better than 1%.

Sample collection and measurement were largely very successful. The integration of the computer software with the GC-EDC system hardware made the procedure almost completely automated. A few problems were encountered initially. Some of the opto-isolator circuitry failed and required replacement. The bow air line filled with moisture transitioning from the warm humid outside air to the cold, dry air-conditioned Main Lab, flooding the magnesium perchlorate trap associated with the pump sample line; this was solved by installing an additional water trap in line just before the magnesium perchlorate trap. The rough seas played havoc with the particular brand of laptop computers integrated with the GC system, causing them to crash several times, which resulted in occasional sample losses. Two of the glass syringes appeared to be contaminated with CFC-11 and CFC-113, respectively, and were removed from service. How they were affected was not discovered, but since no other syringes were contaminated, the situation appeared isolated. To our knowledge, there were no other occurrences of contamination.

Section P16N_2006

Approximately 900 samples were drawn and analyzed for CFC during Section P16N_2006 Leg 1. In addition, 120 samples were analyzed for sulfur hexafluoride (SF₆). The precision of the CFC analysis, based on replicate pairs, is estimated to be the greater of 1% or 0.005 pmol/kg.

The CFC analysis was based on the work of Bullister and Weiss (1988). CFC samples were drawn from the Niskin bottles into glass syringes to prevent contamination from air. A 30 mL aliquot was injected into a glass-fritted reservoir, and clean nitrogen was bubbled through the water to remove the CFCs, which were dried over magnesium perchlorate and concentrated on a trap of Porapak N at -20°C. The trap was subsequently heated and the gases swept off of the trap with nitrogen and injected onto a precolumn of Porasil C (70°C). Once the gases of interest had passed through the precolumn, the remaining gases were vented while the CFCs passed to the 19 main analytical columns (carbograph 1AC, 70°C). The gases were detected by a Hewlett Packard ECD.

During Section P16N_2006 Leg 2, samples for the analysis of dissolved CFC-11, CFC-12, and CFC-113 were drawn from 960 of the 1300 water samples. Specially designed 12-L water sample bottles were used on the cruise to reduce CFC contamination. These bottles have the same outer diameter as standard 10-L Niskin bottles, but use a modified end-cap design to minimize the contact of the water sample with the end-cap O-rings after closing. The O-rings used in these water sample bottles were vacuum-baked prior to the first station. Stainless steel springs covered with a nylon powder coat were substituted for the internal elastic tubing provided with standard Niskin bottles. When taken, water samples for CFC analysis were the first samples drawn from the 12-L bottles. Care was taken to coordinate the sampling of CFCs with other samples to minimize the time between the initial opening of each bottle and the completion of sample drawing. In most cases, helium-3, dissolved oxygen, alkalinity and pH samples were collected within several minutes of the initial opening of each bottle. To minimize contact with air, the CFC samples were drawn directly through the stopcocks of the 12-L bottles into 100 mL precision glass syringes equipped with 3-way plastic stopcocks. The syringes were immersed in a holding bath of freshwater until analyzed.

For air sampling, a ~100-m length of 3/8-in. outside diameter (OD) Dekaron tubing was run from the main laboratory to the bow of the ship. A flow of air was drawn through this line into the CFC van using an Air Cadet pump. The air was compressed in the pump, with the downstream pressure held at ~1.5 atm using a back-pressure regulator. A tee allowed a flow (100 mL/min) of the compressed air to be directed to the gas sample valves of the CFC and SF6 analytical systems, while the bulk flow of the air (>7 L/min) was vented through the back pressure regulator. Air samples were generally analyzed when the ship was on station and the relative wind direction was within 60° of the bow of the ship to reduce the possibility of shipboard contamination. The pump was run for approximately 45 min prior to analysis to ensure that the air inlet lines and pump were thoroughly flushed. The average atmospheric concentrations determined during the cruise (from a set of 5 measurements analyzed approximately once per day, n=23) were 252.9 ± 4.4 parts per trillion (ppt) for CFC-11, 547.2 ± 5.0 ppt for CFC-12, and 76.3 ± 1.9 ppt for CFC-113. Concentrations of CFC-11 and CFC-12, and CFC-113 in air samples, seawater, and gas standards were measured by shipboard ECD-GC using techniques modified from those described by Bullister and Weiss (1988). For seawater analyses, water was transferred from a glass syringe to a fixed volume chamber (~30 mL). The contents of the chamber were then injected into a glass sparging chamber. The dissolved gases in the seawater sample were extracted by passing a supply of CFC-free purge gas through the sparging chamber for 4 min at 70 mL/min. Water vapor was removed from the purge gas during passage through an 18 cm long, 3 in diameter glass tube packed with the desiccant magnesium perchlorate. The sample gases were concentrated on a cold-trap consisting of a 1/8-in OD stainless steel tube with a ~10 cm section packed tightly with Porapak N (60-80 mesh). A vortex cooler, using compressed air at 95 psi, was used to cool the trap, to approximately -20°C. After 4 min of purging, the trap was isolated, and the trap was heated electrically to ~100°C. The sample gases held in the trap were then injected onto a precolumn (~25 cm of 1/8-in OD stainless steel tubing packed with 80 to 100 mesh Porasil C, held at 70°C) for the initial separation of CFC-12, CFC-11 and CFC-113 from other compounds. After the CFCs had passed from the pre-column into the main analytical column (~183 cm of 1/8-in OD stainless steel tubing packed with Carbograph 1AC, 80-100 mesh, held at 70°C) of GC1 (a HP 5890 Series II gas chromatograph with ECD), the flow through the pre-column was reversed to backflush slower 23 eluting compounds. Both of the analytical systems were calibrated frequently using a standard gas of known CFC composition. Gas sample loops of known volume were thoroughly flushed with standard gas and injected into the system. The temperature and pressure was recorded so that the amount of gas injected could be calculated. The procedures used to transfer the standard gas to the trap, precolumn, main chromatographic column, and ECD were similar to those used for analyzing water samples. Two sizes of gas sample loops were used. Multiple injections of these loop volumes could be made to allow the system to be calibrated over a relatively wide range of concentrations. Air samples and system blanks (injections of loops of CFC-free gas) were injected and analyzed in a similar manner. The typical analysis time for seawater, air, standard, or blank samples was ~10.5 min.

Concentrations of the CFCs in air, seawater samples and gas standards are reported relative to the SIO98 calibration scale (Prinn et al. 2000). Concentrations in air and standard gas are reported in units of mole fraction CFC in dry gas, and are typically in the parts per trillion (ppt) range. Dissolved CFC concentrations are given in units of picomoles per kilogram seawater (pmol/kg). CFC concentrations in air and seawater samples were determined by fitting their chromatographic peak areas to multi-point calibration curves, generated by injecting multiple sample loops of gas from a working standard (UW cylinder 45191 for CFC-11: 386.94 ppt, CFC-12: 200.92 ppt, and CFC-113: 105.4 ppt) into the analytical instrument. The response of the detector to the range of moles of CFC-12 and CFC-113 passing through the detector remained relatively constant during the cruise. A thorough baking of the column and trap after a power outage during trapping of a seawater sample introduced an unknown contaminant into the column changed the response of the detector to CFC-11. Full-range calibration curves were run at intervals of 10 days during the cruise. These were supplemented with occasional injections of multiple aliquots of the standard gas at more frequent intervals. Single injections of a fixed volume of standard gas at one atmosphere were run much more frequently (at intervals of ~90 min) to monitor short-term changes in detector sensitivity. The CFC-113 peak was often on a small bump on the baseline, resulting in a large dependence of the peak area on the choice of endpoints for integration. The height of the peak was instead used to provide better precision. The precisions of measurements of the standard gas in the fixed volume (n=395) were ± 0.44% for CFC-12, 0.56% for CFC-11, and 3.0% for CFC-113.

The efficiency of the purging process was evaluated periodically by re-stripping high concentration surface water samples and comparing the residual concentrations to initial values. These re-strip values were approximately <1% of the initial sample concentration for all 3 compounds. A fit of the re-strip efficiency as a function of temperature will be applied to the final data set. The determination of a blank due to sampling and analysis of CFC-free waters was hampered by a contamination peak that co-eluted with CFC-11 and varied greatly in size during this leg. The size of the peak decreased exponentially with time, but jumped to very high values (0.05 pmol/kg) after each of the four power outages encountered during Leg 2. Further investigation needs to be done to understand the origin of this contamination. CFC-113 and CFC-12 sampling blanks were less than 0.005 pmol/kg.

During the expedition, the precisions (1 standard deviation) of 0.45% or 0.004 pmol/kg (whichever is greater) for dissolved CFC-11, 0.36% or 0.003 pmol/kg for CFC-12 measurements, and 0.004 pmol/kg for CFC-113 was estimated based on the analysis of 38 duplicate samples. A very small number of water samples had anomalously high CFC concentrations relative to adjacent samples. These samples occurred sporadically during the cruise and were not clearly associated with other features in the water column (e.g. anomalous dissolved oxygen, salinity or temperature features). This suggests that these samples were probably contaminated with CFCs during the sampling or analysis processes. Measured concentrations for these anomalous samples are included in the data, but are given a quality flag value of either 3 (questionable measurement) or 4 (bad measurement). A quality flag of 5 was assigned to samples that were drawn from the rosette but never analyzed due to a variety of reasons (e.g., power outage during analysis).