Environ. Sci. Technol. 2008, 42, 465–470
Hexachlorocyclohexanes (HCHs) In the Canadian Archipelago. 2. Air-Water Gas Exchange of r- and γ-HCH L I I S A M . J A N T U N E N , * ,† P A U L A . H E L M , ‡,| HENRIK KYLIN,§ AND TERRY F. BIDLEMAN† Centre for Atmospheric Research Experiments, Environment Canada, 6248 Eighth Line, Egbert ON, L0L 1N0, Canada, Chemical Engineering and Applied Chemistry, University of Toronto, 200 College St., Toronto, Ontario, M5S 3E5 Canada, and Norwegian Institute for Air Research, Polar Environmental Centre, No-9296 Tromsø, Norway; Department of Environmental Assessment, Swedish University of Agricultural Sciences, P.O. Box 7050, SE-705-07, Uppsala, Sweden
Received July 04, 2007. Revised manuscript received October 04, 2007. Accepted October 05, 2007.
Air and water were sampled in the Canadian Archipelago during summer on the Tundra Northwest 1999 (TNW-99) expedition and air was sampled at Resolute Bay (RB), Nunavut, to determine the gas exchange of R- and γ-hexachlorocyclohexanes (HCHs) and the enantiomers of R-HCH. Air concentrations of ΣHCH during TNW-99 and at RB were similar, averaging 55 and 53 pg m-3, respectively. The net gas exchange direction was volatilization for R-HCH and near equilibrium or deposition for γ-HCH, whereas actual fluxes depended on the fraction of open water. Enantiomer fractions, EF ) (+)/[(+) + (-)] of R-HCH in air sampled from shipboard were significantly correlated to those in surface water for events with >90% open water, but were closer to racemic and not correlated to EFs in water for events with 0–50% open water. Levels of R-HCH in air at RB averaged 37 ( 9 pg m-3 from June to early July, and EFs were close to racemic (0.496 ( 0.004). In mid-July the ice pack broke up around RB. From this point through August, air concentrations increased significantly to 53 ( 5 pg m-3, and the mean EF decreased significantly to 0.483 ( 0.009. Air concentrations of γ-HCH at RB did not differ significantly before (8.0 ( 3.7 pg m-3) and after (6.6 ( 0.76 pg m-3) ice breakup. Results show that R-HCH enantiomers are sensitive tracers for following the impact of ice cover loss on gas exchange in the Arctic.
Introduction Technical hexachlorocyclohexane (HCH) is a mixture of several isomers, the most abundant being the R-HCH * Corresponding author fax: 705-458-3301; e-mail:
[email protected]. † Environment Canada. ‡ University of Toronto. | Current Address: Environmental Monitoring and Reporting Branch, Ontario Ministry of the Environment, 125 Resources Road, West Wing, Toronto, Ontario, M9P 3V6 Canada. § Norwegian Institute for Air Research and Swedish University of Agricultural Sciences. 10.1021/es071646v CCC: $40.75
Published on Web 12/14/2007
2008 American Chemical Society
(60–70%), β-HCH (5–12%), and γ-HCH (10–15%) (1). Although γ-HCH is the only isomer with insecticidal properties, and β-HCH is generally the most bioaccumulating (2), all three HCHs are found in arctic biota. Differences in the relative body burdens of the HCH isomers among species result from selective metabolism and varying concentration distributions in Arctic Ocean water masses (3). HCH is the most abundant organochlorine pesticide in arctic air (4) and water (5). The transport and mass balance of HCHs in the Arctic have been discussed in recent reports (3, 6, 7). Technical HCH was heavily used in Asian countries until it was banned or heavily restricted by China, the former Soviet Union, and India between the mid-1980s and 1990 (8, 9). Concentrations of R-HCH in arctic air responded quickly to these large-scale usage changes and declined by an order of magnitude from the early 1980s to mid-1990s in steps that closely matched global usage (8) and emission estimates (9). As a consequence, the direction of net gas exchange in arctic waters reversed from deposition in the 1980s to air-water equilibrium or volatilization in the mid-1990s (10–13). Although HCHs have been reported in arctic air and water in the Bering-Chukchi seas (10, 11), the northern Canada Basin (11), the Eurasian Arctic Ocean (12, 14), and in air at arctic monitoring stations (15), few studies have been done in the main portion of the Canadian Archipelago. HCH concentrations in central Archipelago water were measured at Resolute Bay on Cornwallis Island, NV (RB, 74.68°N, 94.90°W) in 1992 (16) and seasonal air-water exchange at RB was followed in 1993 (17), using air data from the monitoring station at Alert, NV (82.50°N, 62.33°W). Air concentrations of HCHs were reported at Kinngait, NV (64.22°N, 76.53°W) in the eastern Archipelago for 2000–2001, but no co-located water measurements were made (15). An assessment of surface water concentrations and air-water exchange in the Arctic Ocean and adjoining seas was recently made for γ-HCH (18). The purpose of this investigation was to determine the spatial distribution and air-water gas exchange of HCHs across the Canadian Archipelago. Parallel water and air samples were collected in the Archipelago on a SwedishCanadian expedition, Tundra Northwest 1999 (TNW-99), with additional air samples collected at RB. Spatial variability and pathways of HCHs in surface water are discussed in a separate paper (19). Here we report air-sea exchange of R-HCH and γ-HCH as affected by seasonal ice cover and the use of R-HCH enantiomers to follow the exchange.
Experimental Section Sample Collection and Preparation. The TNW-99 expedition platform was on board the CCGS Louis S. St-Laurent. The ship traveled from Nuuk, Greenland across the Davis Strait to Iqaluit, Canada then traversed the Archipelago from Hudson Strait to RB (Leg 1), along a southern route to Tuktoyaktuk and the southern Beaufort Sea (no samples taken), and returned along a northern route to RB, Ellef Ringnes Island, over Devon Island, through Baffin Bay and the Davis Strait (Leg 2). A map of the cruise track is shown in Supporting Information (SI) Figure S1 and in ref 19. Air and surface water samples on Leg 1 between Iqaluit and Resolute Bay were collected by Environment Canada (EC) personnel while those on Leg 2 were collected by the Swedish University for Agricultural Sciences (SLU). Additional air samples were collected at RB from June 7 to August 14, 1999. Collection and processing methods for air samples are given below. Dates and locations are summarized in Table 1 and provided in detail in SI Table S1. Collection of water samples is described in ref 19. VOL. 42, NO. 2, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
465
TABLE 1. Atmospheric Concentrations of HCHs during TNW-99 and at Resolute Bay (RB)a, pg m-3
Leg 1 TNW-99 Leg 2 TNW-99 RB, before ice break up RB, after ice break up a
date (1999)
r-HCH mean
s.d.
γ-HCH mean
s.d.
r-HCH/γ-HCH
r-HCH EF
s.d.
June 30 - July 11 August 1 - 30 June 7 - July 17 July 19 - Aug 14
42 48 37 53
9.0 14 9.2 5.0
11 8.8 8.0 6.6
2 4.7 3.7 0.76
4.0 6.3 5.1 8.2
0.475 0.460 0.496 0.483
0.015 0.019 0.0035 0.0088
Details are given in Supporting Information S1 and S3.
EC: High volume air samples were continuously collected on the bow of the ship. When possible, the bow was into the wind to avoid contamination. At RB, the air sampling site was located inland from Lancaster Sound, 6 km from the village of Resolute and 3 km from the local airport behind a ridge of hills. Air samples were collected using sampling trains consisting of a glass fiber filter (EPM 2000 20 × 25 cm Whatman, Ltd., Maidstone, England) followed by two polyurethane foam (PUF) plugs (8.0 cm diameter × 7.5 cm). Air volumes ranged from 462–689 m3 and 846–1652 m3 for TNW99 and RB, respectively. Air sampling media were prepared and processed after sample collection as described in ref 20. Extract volumes were brought to 1 mL (quantitative analysis) or 100 µL (enantiomer analysis) through volume reduction under a nitrogen stream and solvent exchange into isooctane. SLU: Air sampling was done on the deck above the bridge using the same sampling apparatus and PUF size as EC but with a 15 cm diameter filter. Air volumes ranged from 520–710 m3. PUFs were Soxhlet extracted in acetone. Air sample extracts were cleaned up on a gel permeation column with Biobeads SX-3, eluted with 1:1 DCM:ethyl acetate to remove some of the larger molecules and then fractionated on deactivated silica (21). Analysis was done using capillary GC-electron capture negative ion mass spectrometry. A DB-5 column was used for quantitative work and chiral-phase columns for analysis of R-HCH enantiomers, details are provided in ref 19.
Results and Discussion Quality Control. EC and SLU Air Samples. Field blanks were done by placing a clean PUF and filter into the air sampling apparatus and drawing air for 30 s. No peaks were detected in the blank PUFs; blank filters were not analyzed. Each PUF was spiked with R-HCH-d6 prior to extraction, recoveries averaged, 80 ( 15%, (n ) 60, EC) and 92 ( 12% (n ) 15, SLU). Sample concentrations have been adjusted for mean recoveries. The saturation state of the PUF was checked by analyzing the front and back PUFs separately. About half of the samples showed a low level of breakthrough for R-HCH, ranging from 0.5 to 4%, whereas no breakthrough was seen for γ-HCH. Enantiomer fractions of R-HCH, EF ) quantities of (+)/[(+) + (-)] enantiomers, were determined on two dissimilar chiral-phase columns for EC samples on TNW-99 (n ) 9) and at RB (n ) 19), the average difference in EFs was 0.32%. A single chiral column was used by SLU (19). Quality control for quantitative and enantiomer analyses of water samples is reported in ref. HCHs in Air. Air concentrations of R-HCH from TNW-99 averaged 46 ( 13 pg m-3 (n ) 24) and compared very well with those at RB, 44 ( 11 pg m-3 (n ) 19). For γ-HCH, results from TNW-99 averaged 9.5 ( 3.9 pg m-3, which is not significantly different from the RB mean of 7.4 ( 2.9 pg m-3 (p >0.05) (Table 1, and SI Tables S1 and S3). Concentrations of R-HCH and γ-HCH were correlated on TNW-99 (r2 ) 0.62, p < 0.001). The relationship at RB was not significant (p ) 0.10) and skewed by four samples that were elevated for γ-HCH but not for R-HCH (SI Table S3: samples 1, 2, 9, and 10). When these four samples were removed, the correlation was significant with r2 ) 0.48 (p ) 0.004). 466
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 2, 2008
Annual mean concentrations of R-HCH at Alert and Kinngait during 2000–2001 were 22 and 28 pg m-3, whereas those of γ-HCH were 5.6 and 4.1 pg m-3, respectively (15). Slightly higher annual means were found at Alert in 1999, 34 and 5.8 pg m-3 for R-HCH and γ-HCH (Su, Y. and Hung, H., Environment Canada, personal communication). Means for R-HCH and γ-HCH at western arctic and subarctic stations in 2001–2002 were 19 and 2.7 pg m-3 at Point Barrow, AK (71.30°N, 156.60°W), and 48 and 4.5 pg m-3 at Little Fox Lake, YK (61.35°N, 135.63°W) (15). Shen et al. (22) deployed passive samplers throughout North America, including the Canadian Arctic, during 2000–2001. Concentrations in the Arctic ranged from 34–95 pg m-3 for R-HCH and 5.3–15 pg m-3 for γ-HCH, with the highest levels on the north shore of Baffin Island and the lowest at Alert. The concentrations found on TNW99 and at RB fall within the ranges in these studies. Hung et al. (4) found that R-HCH and γ-HCH declined in air at Alert with times for 50% reduction (half-times) of 9.1 and 5.7 y between 1993–1999. The decline of R-HCH in arctic air is slower than at stations on the Great Lakes, where half-times between 1996–2003 ranged from 1.6 to 4.2 y (23). Half-times for γ-HCH at Great Lakes stations ranged from 4.2 to 10 y (23), spanning the rate at Alert. The slower decline of R-HCH in the Arctic may be due to regional buffering of air concentrations by revolatilization during times of open water, see below. Air-Water Gas Exchange. Fugacity Ratios. Water samples collected during TNW-99 allowed estimates of air-water exchange to be made. Concentrations of R-HCH and γ-HCH in surface water of the Archipelago are summarized in Table 2 and details are given in Table S1 in the Supporting Information of ref 19. Examination of spatial trends showed higher concentrations of HCHs in the southern Beaufort Sea, western and northern Archipelago, which decreased to the east and south of Barrow Strait near RB. EFs’ of R-HCH averaged 0.438 west and north of Barrow Strait and increased to a mean of 0.452 in the eastern Archipelago. A plausible explanation for these trends is dilution of water advected from west of Barrow Strait with Arctic Ocean water entering northern Baffin Bay via the Nares Strait and Atlantic Ocean water entering southern Baffin Bay through the Davis Strait (19). The water/air fugacity ratios (FR) of R-HCH and γ-HCH were calculated from FR )
CWH CARTA
(1)
where CW and CA are water and air concentrations (mol m-3), H is the Henry’s law constant (Pa m3 mol-1) at the water temperature, TA is the air temperature (K), and R is the gas constant (8.314 Pa m3 mol-1 K-1) (13, 20). FR )1.0, >1.0, and 1.0 at p
