Fluorotelomer Carboxylic Acids and PFOS in Rainwater from an Urban

Carboxylates from the Consumption of Store-Bought and Restaurant Foods for the ..... Critical Reviews in Food Science and Nutrition 2011, 51 (7) ,...
0 downloads 0 Views 292KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 2944-2951

Fluorotelomer Carboxylic Acids and PFOS in Rainwater from an Urban Center in Canada M A R K L O E W E N , †,‡ T H O R H A L L D O R S O N , † F E I Y U E W A N G , ‡ A N D G R E G G T O M Y * ,†,‡ Department of Fisheries & Oceans Canada, Freshwater Institute, Winnipeg, Manitoba, Canada and Department of Chemistry, University of Manitoba, Winnipeg, Manitoba, Canada

A method based on LC/MS/MS analysis of fluorotelomer carboxylic acids (FTCAs: CnF2n+1CH2COOH, n ) 6, 8, and 10) and fluorotelomer unsaturated carboxylic acids (FTUCAs: CnF2nCHCOOH, n ) 6, 8, and 10) in rainwater using negative ionization electrospray multiple reaction monitoring conditions is described. These compounds are thought to be oxidative products of atmospherically transported fluorotelomer alcohols (FTOHs: CnF2n+1CH2CH2OH). Preconcentration from rainwater samples collected in Winnipeg, Manitoba, Canada, was achieved using solidphase extraction on C18 sorbent. Low parts per trillion levels of the C10- and C12-FTCAs and FTUCAs were detected, suggesting that one possible pathway of removing FTOHs from the atmosphere is through oxidation and wet deposition. Perfluorocarboxylic acids (PFCAs) and perfluorooctane sulfonate (PFOS) were simultaneously analyzed in the rainwater samples using established LC/MS/MS methods. PFOS was deposited in rainwater with a concentration of 0.59 ng/L while PFCAs were not detected above their respective method detection limits.

Introduction Fluorinated organic compounds (FOCs) are a diverse class of compounds used in a myriad of consumer and industrial applications (1, 2). They are used as refrigerants, agrochemicals, chemical catalysts/reagents, and surfactants (2). They also have medicinal applications which include treatment of obesity and clinical depression (2). Although FOCs have been manufactured for over 50 years, there are no published reports on their total production volumes. Recent interest has been on the perfluorinated class of surfactants. These are chemicals characterized by chains of carbon atoms of varying lengths in which all the hydrogen atoms are substituted by fluorine atoms. Perfluorooctane sulfonate (PFOS, C8F17SO3-) and perfluorooctanoate (PFOA, C7F15COO-) are two of the perfluorinated surfactants that have received the most attention. Much of the concern surrounds the ubiquitous presence of both compounds in the environment. PFOS and PFOA have been detected in human serum (3), freshwater and marine biota (4, 5), and surface water (6, 7). The stability that makes fluorinated * Corresponding author phone: 204-983-5167; fax: 204-984-2403; e-mail: [email protected]. † Freshwater Institute. ‡ University of Manitoba. 2944

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

surfactants so desirable appears to preclude any degradation or metabolism and contributes to the global bioaccumulation and persistence of PFOS and PFOA. While it is unlikely that PFOS and PFOA are atmospherically transported in their vapor phase due to low vapor pressures, they appear in regions with no direct sources of these contaminants. It has been hypothesized that these chemicals are atmospherically transported to remote areas as volatile precursors and then deposited by some mechanism (4). Perfluoroalkylethanols are straight chain FOCs having the general formula CnF2n+1CH2CH2OH (n ) 2, 6, 8,...14). The even number of carbons in the fluoroalkyl chain is a result of the manufacturing process (8). Also referred to as fluorotelomer alcohols (FTOHs), these compounds are used as surfactants and intermediates in processing of other products which themselves have many uses, including polymers, adhesives, and paints (1). The worldwide production of FTOHs for the period of 2000-2002 was estimated to be between 5 and 6.5 × 106 kg year-1, of which 40% is produced in North America (9). Although the processes that lead to the release of FTOHs into the environment are still unclear (8), recent measurements in the troposphere imply that FTOHs are escaping from the products in which they are incorporated (8, 10). Atmospheric lifetimes of some of the shorter chain length FTOHs also suggests that transport to remote regions is highly likely (8, 11). The relatively high vapor pressure [254 Pa for the 2-perfluorooctylethanol and 144 Pa for the 2-perfluorodecylethanol (12)] and low water solubility of FTOH (8) are likely to deter wet or dry deposition from the atmosphere. The high tendency of FTOHs to remain in the gaseous phase is supported by their detection in the troposphere (13). Oxidation appears to be a degradation route of FTOHs in the atmosphere (8). Smog chamber reaction studies of FTOH with chlorine radicals showed that 26% of the starting material could be transformed into the corresponding or shorter chain length 2H,2H-perfluoroalkanoic acids or fluorotelomer acids (FTCA: CnF2n+1CH2COOH) (8). Compared to FTOHs, the wet deposition of these acids should be more favorable because of their increased water solubility and lower vapor pressure. However, no information is currently available to support or refute this. Metabolic biotransformation of FTOHs in biota is also thought to yield FTCAs (14). Perfluorooctylethanol (C8F17CH2CH2OH) has been shown to be metabolized to PFOA in adult male rats (14). An intermediate in the metabolism was 2H,2H-perfluorooctanoic acid (C8F17CH2COOH). It was suggested that the dehydrofluorinated form of the 2H,2Hperfluorooctanoic acid, i.e., 2H-perfluoro-2-octenoic acid (C6F12CHCOOH), was also a likely intermediate. A study by Lange (2002) in which a mixture of FTOHs was biodegraded to FTCAs and FTUCAs by microbes in sewage sludge provides further evidence that biotransformation of FTOH is an important degradation pathway (15). To our knowledge, there are no reports on the detection of FTCAs or FTUCAs in the environment. This is perhaps due, in part, to a lack of a robust method for the detection and analysis of these compounds. To investigate whether high-performance liquid chromatography (HPLC) might be useful for the analysis of these compounds we developed a method based on HPLC coupled with electrospray (ES) introduction tandem mass spectrometry (MS/MS). On the basis of the presence of FTOHs in the troposphere, it was hypothesized that rainwater would be an environmental 10.1021/es048635b CCC: $30.25

 2005 American Chemical Society Published on Web 03/17/2005

TABLE 1. Molecular Formula and Weights of the Fluorinated Acids Examined in This Studya mol formula and weights

a

compound

native

13C-labeled

2H,2H-perfluorooctanoic acid (6:2 FTCA) 2H,2H-perfluorodecanoic acid (8:2 FTCA) 2H,2H-perfluorododecanoic acid (10:2 FTCA) 2H-perfluoro-2-octenoic acid (6:2 FTUCA) 2H-perfluoro-2-decenoic acid (8:2 FTUCA) 2H-perfluoro-2-dodecenoic acid (10:2 FTUCA)

C6F13CH2COOH (378) C8F17CH2COOH (478) C10F21CH2COOH (578) C6F12CHCOOH (358) C8F16CHCOOH (458) C10F20CHCOOH (558)

C6F1313CH213COOH (380) C8F1713CH213COOH (480) C10F2113CH213COOH (580) C6F1213CH13COOH (360) C8F1613CH13COOH (460) C10F2013CH13COOH (560)

Molecular weights in parentheses.

compartment containing FTCAs and FTUCAs. We report here for the first time concentrations of FTCAs, FTUCAs, and PFOS in rainwater from Winnipeg, Manitoba, Canada. Preconcentration from rainwater was achieved using solid-phase extraction on an octadecyl silica gel (C18) stationary phase.

Experimental Section Chemicals. Native and 13C2-labeled FTCAs [C6F13CH2COOH (6:2 FTCA), C8F17CH2COOH (8:2 FTCA), and C10F21CH2COOH (10:2 FTCA)] and FTUCAs [C6F12CHCOOH (6:2 FTUCA), C8F16CHCOOH (8:2 FTUCA), and C10F20CHCOOH (10:2 FTUCA)] (see Table 1) used in this study were supplied by Wellington Laboratories Inc. (Guelph, ON, Canada). The nomenclature of these compounds has been described earlier (16). The suite of perfluorinated compounds perfluorooctane sulfonate (PFOS), perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), perfluoroundecanoic acid (PFUA), perfluorododecanoic acid (PFDoA), and perfluorodimethyloctanoic acid (PFMe2OA) were obtained from Aldrich (Oakville, ON, Canada) and SynQuest Labs (Alachua, FL). HPLC Optima-grade methanol and water were obtained from Fisher Scientific (Nepean, ON, Canada). Octadecyl (C18) functionalized silica gel and polyethylene tubes (60 mL) were obtained from Sigma-Aldrich (Oakville, ON, Canada) Liquid Chromatography. Separations were performed on a Discovery C18 analytical column (5.0 cm × 2.1 mm i.d., 5 µm particle size; Supelco, Oakville, ON, Canada). An Agilent 1100 Series HPLC system (Agilent Technologies, Palo Alto, CA) equipped with a vacuum degasser, binary pump, and autosampler was used for all analyses. The mobile phase system used consisted of water and methanol; a mobile phase flow rate of 300 µL/min was utilized, and the sample injection volume was 3 µL. The gradient employed started at 20% methanol increasing to 95% in 9.5 min and was held for 2 min. Thereafter the mobile phase composition was returned to starting conditions in 5 min. The column was allowed to equilibrate for 5 min between runs. Mass Spectrometry. A Sciex API 2000 triple-quadrupole mass spectrometer (Applied Biosystems, Foster City, CA) was used in the electrospray (ES) negative-ion mode. The collision cell was upgraded (mSPEC, Concord, ON, Canada) to improve the instrument sensitivity in the multiple reaction mode (MRM). All gas supplies were provided from a nitrogen generator (Peak Scientific, Renfrewshire, Scotland, U.K.). Infusion experiments on FTCAs and FTUCAs utilized the built-in Harvard syringe pump operating at a flow rate of 20 µL/min using standards in the concentration range of 100-500 pg/ µL. A scan using the first quadrupole (Q1) was initially performed to determine the presence of a stable [M - H]parent ion. The Q1 scan range was set to 10 mass units above and below the expected [M - H]- for each compound. For MS acquisition Q1 was operated with unit resolution with a

scan time of 0.5 s. Source and MS parameters were optimized for each compound independently under Q1 conditions. After the Q1 scans multiple reaction monitoring (MRM) conditions were determined for each compound by first performing a product ion scan (PIS) of the generated([M H]-. The most abundant transition observed under PIS conditions was then optimized under MRM conditions and used for quantitation. If present, a second transition of abundance was also optimized under MRM conditions and used for confirmation. Transitions characteristic of each compound class were observed, and the transition(s) used for MRM optimizations and ultimately MS/MS detection are shown in Table 1. MRMs were determined utilizing unit resolution on the first and third quadrupoles and a 200 ms dwell time. Flow injection analyses (FIA) for each compound class were then performed in order to optimize source and collision-activated dissociated (CAD) gas pressure under simulated LC conditions. Unlike MS/MS parameters determined specifically for each MRM transition listed above, source and CAD parameters are fixed for all monitored transitions throughout the analyses. Both 8:2 FTCA and 8:2 FTUCA were chosen to optimize FIA methods for the FTCAs and FTUCAs, respectively, as these compounds were intermediate in carbon chain length for the available standards used in this study. The MRM transitions used in the analysis of PFCAs have been reported previously (6). The same ion transitions were used for both PFDA and PFMe2OA; the latter is a branched isomer of PFDA that elutes earlier off the analytical column. Collection and Extraction of Rainwater Samples. Rainwater samples were collected in triplicate in July 2004 into 20-L stainless steel cans prerinsed with methanol. The samples were collected at the Freshwater Institute in Winnipeg. The entire rain event was captured. The unfiltered samples were spiked with C10F2113CH213COOH, C10F2013CH13COOH, and PFMe2OA (10 µL of a 2-5 ng/µL solution) recovery internal standards (RIS) to monitor extraction and cleanup efficiencies of the method. PFMe2OA was used in this study to recovery correct PFOS and the perfluorinated carboxylic acids (PFCAs). While not an ideal RIS, PFMe2OA was chosen because it had similar affinities for the C18 material as PFOS and PFCAs based on its elution off the analytical column. A polyethylene tube (60 mL) containing 5 g of C18 silica gel was preconditioned with 50 mL of Optima-grade methanol followed by 20 mL of Optima-grade water. Within 24 h of the rainfall event the acids were extracted (and analyzed) from water by pumping 4 L of rainwater through the extraction tube at approximately 100 mL/min. The tube was briefly air-dried and then eluted with 50 mL of Optima-grade methanol. Samples were reduced by rotary evaporation, transferred to a polyethylene tube, and dried under a gentle stream of nitrogen to 1 mL. The samples were spiked with mass-labeled C6F1313CH213COOH and C6F1213CH13COOH (10 µL of a 2-5 ng/µL solution) instrument performance internal standards (to monitor fluctuations of the instrument between VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2945

injections) and transferred to a polyethylene vial for LC injection. QA/QC. Triplicate blanks of 1 L of Optima-grade water were prepared in an identical fashion to the rainwater samples. They were analyzed for the presence of the native compounds of interest as well as extraction efficiency of the mass-labeled compounds. Native standards were freshly prepared in methanol ( 0.99) spanning concentrations from 1 to 50 pg/µL for each analyte. Instrument performance and recovery internal standards were included in the calibration standards. Concentrations of all analytes were corrected based on the responses of both the appropriate instrument performance and recovery internal standards. Additionally, method blanks were run with each batch and used to identify any potential interferences arising from both the method and the LC system. Potential for ion suppression arising from the matrix was assessed by comparing the responses of the native compounds in a methanol solution to the method blanks that were intentionally fortified with each compound. There was good agreement between the two measurements (>80%), suggesting that ion suppression is not a significant cause of uncertainty in our analytical measurement. For argument purposes, we assumed that Optima-grade water is a suitable surrogate for rainwater. This is unlikely to be the case for more complex aqueous samples such as lake water. In that case, standard addition is probably the preferred approach for quantitation. Extraction efficiencies for FTCAs were between 50% and 75%, while FTUCAs were between 61% and 82%. Average PFMe2OA recoveries were 47 ( 6% (arithmetic mean ( 1 × standard error). Method detection limits (MDLs), defined as the mean blank signal (n ) 3) for each analyte plus 3 times the standard deviation (17), for 8:2 and 10:2 FTCAs were 0.17 and 0.08 ng/L, respectively. Respective MDLs for 8:2 and 10:2 FTUCAs were 0.07 and 0.04 ng/L (n ) 3). PFOS MDL was found to be 0.39 ng/L (n ) 3), while the MDLs of the perfluorocarboxylic acids PFOA, PFNA, PFDA, PFUA, and PFDoA were 7.2, 3.7, 1.7, 1.2, and 1.1 ng/L, respectively. Degradation of FTCAs in Methanol and Water. The possibility of HF loss from FTCAs and transformation to FTUCAs was examined in solutions of methanol and water and is a critical analytical consideration in evaluating the stability of FTCAs in both standards and analyzed samples. Crystals of 6:2, 8:2, and 10:2 FTCA (∼2 mg) were weighed and dissolved in Optima-grade methanol to give a stock solution of ∼1 mg/mL. Degradation studies were conducted in solutions of Optima-grade methanol (n ) 4) and water (n ) 4) by diluting the stock solution to a concentration of 600 pg/µL in methanol or water. Experimental blanks consisted of Optima-grade methanol (n ) 3) and water (n ) 3) with no FTCAs. One of the FTCAs solutions in methanol and water was kept at 4 °C in the dark. All other solutions were kept at ambient room temperature and under fluorescent lighting (∼8 h each day). Solutions were injected directly (1 µL) onto the analytical column to follow the transformation of FTCAs to FTUCAs.

Results and Discussion Optimization of ES-MS/MS Conditions. Standard solutions of all acids (native and labeled) were prepared in methanol (∼150 pg/µL) and used to optimize the MS under negative-ion ES conditions. Q1 scans for the native and labeled acids with a range of m/z from 350 to 400 (6:2 FTCA and 6:2 FTUCA), 450 to 500 (8:2 FTCA and 8:2 2946

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

FTUCA), and 550 to 600 (10:2 FTCA and 10:2 FTUCA) were recorded separately for each compound using the multiple count addition (MCA) mode by directly infusing each at a flow rate of 20 µL/min. The Q1 spectra of the acids examined in this study were all dominated by their respective [M - H]ions. Selected ion monitoring (SIM) mode was then employed to optimize the Q1 parameters. In general, changing the ion spray (IS) voltage had the most pronounced effect on the ion signal of all the acids. There was almost a 2-fold increase in intensity going from an IS voltage of -4000 to -1500V (not shown). Once all the source parameters had been optimized, product ion scans were recorded. In this mode the first stage of the mass analysis (Q1) is used to isolate specific ions and transmit them to the second stage (Q2), where the ions are excited by means of a collision gas (18). This excitation leads to formation of fragment ions that are detected by Q3. As an example, the product ion spectrum of 8:2 FTCA is shown in Figure 1. The top panel, obtained at a collision energy (CE) of -5.0 V, shows many fragment ions. The peak at m/z 433.1 is thought to arise from a loss of CO2 from the [M - H]- ion. The peak at m/z 412.7 arises from concomitant losses of CO2 and HF from the parent ion. Consistent with its structure, the peak at m/z 392.6 arises from the loss of CO2 and 2HF molecules from the parent ion. The low mass peak at m/z 62.9 is thought to be that of the C2HF2-. Product ion scans of 6:2 FTCA and 10:2 FTCA were also dominated by ions arising from losses of CO2, CO2/HF, and CO2/2HF from their respective molecular ions. The C2HF2- ion was also present in both spectra. The appearance of the product ion spectra of the 13C2labeled FTCA surrogates was similar to their native analogues at high m/z values. Similar losses of CO2, CO2/HF, and CO2/ 2HF were seen in all three labeled surrogates, and respective m/z values of these fragment ions were 1 amu higher than their respective native analogue. An interesting feature in the product ion spectra of the 13C2-labeled compounds is the m/z value of the low-mass peak thought to be C2HF2-. The peak corresponding to this ion was at m/z 64, 1 amu higher than in the native compounds. This implies that this fragment must contain one 13C atom and suggests that the carbon next to the carboxylic group must constitute one of the carbon atoms to this ion. To increase instrument sensitivity, the MS/MS was run in the multiple reaction monitoring (MRM) mode. In this configuration a selected precursor ion(s) is transmitted into the collision cell (Q2) and only a selected product ion(s) is monitored by Q3. This mode is particularly sensitive because no mass scanning takes place by either Q1 or Q3 (18). Specificity can also be achieved by selecting more than one product ion using Q3. For 8:2 FTCA, increasing the CE energy from -5 to -30 V resulted in only two major product ions being formed (bottom, Figure 2). Similar observations were made for 6:2 and 10:2 FTCAs. (The small peak at m/z 242, observed under higher collision energy conditions for all FTCAs, is thought to be from C6F9-). Quantitation of 8:2 FTCA was achieved by recording the MRM ion signal of the m/z 477 [M - H]- to 393 [M - COOH - 2HF]- transition. For confirmation purposes, the ion signal from the m/z 477-63 [C2HF2-] reaction was used. Table 2 lists the MRM transitions for each of the FTCAs. The FTUCAs were optimized in a manner similar to the FTCAs. As an example, the product ion spectrum (CE ) -10 V) of 6:2 FTUCAs is shown in Figure 2. Under lower CE voltages the product ion spectra were dominated by the [M - H]- and [M - COOH - HF]- ions. Consistent with their structure, ions arising by loss of two HF molecules were not observed. By increasing the CE energy to -50 V, ions at m/z

FIGURE 1. Product ion spectrum of 8:2 FTCA at -5 (top) and -30 V (bottom). 63, 119, 169, and 219 belonging to the CnF2n+1 series characteristic of fluoroalkanes arise (19). The ions at m/z 193, 143, and 93 arise from successive losses of CF2 from the [M - HF - CO2H]- ion at m/z 293. Quantitation of FTUCAs was achieved by recording the respective ion signals from the MRM transition of [M - H]- to [M - COOH - HF]-. Despite our best efforts, other MRM transitions from the FTUCAs (e.g., m/z 357 to 243) generated an ion signal significantly lower (usually less than 2%) than the [M - H]to [M - COOH - HF]- ion transition. At low analyte concentration no other significant MRM signal other than the [M - H]- to f [M - COOH - HF]- ion transition was observed. Therefore, confirmation ions were not included

for the FTUCAs. Table 2 lists the MRM transitions for each of the FTUCAs. Degradation of FTCAs in Methanol and Water Solutions. Figure 3A-C shows the kinetics of formation of 6:2, 8:2, and 10:2 FTUCAs over time from their corresponding FTCAs in solutions of methanol and water. There is very little degradation of the FTCAs in methanol at t ) 0 days; degradation appears to be rapid after this time and then tapers off by day 10. This finding demonstrates the necessity of handling both native and labeled FTCAs with extreme care when analyzing for these compounds as well as FTUCAs due to this potential confounding effect. Mildly basic conditions should favor HF loss from FTCAs. At t ) 10 days VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2947

FIGURE 2. Product ion spectrum of 6:2 FTUCAs at -5 (top) and -50 V (bottom). it is likely that the concentration of HF in solution is sufficient to inhibit further degradation. Little degradation occurs in water throughout the 10-day period. It may be difficult to extrapolate the results of the degradation study in Optima-grade water to other aqueous media as Optima-grade water may not be a wholly suitable surrogate. Aqueous pH, temperature (see below), and presence of humic acids (in lake water, for example) are all factors that can affect degradation of FTCAs. 2948

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

The effect of temperature on the rate of degradation is shown in Figure 4 a and b. At 4 °C the amount of degradation is much less than that at ambient temperature of approximately 20 °C. This is significant and suggests that care should be taken when preparing and storing external standard solutions. In addition to reduced temperatures, mildly acidic conditions may also reduce the rate of HF loss. Work is ongoing in our laboratory to test this possibility.

TABLE 2. Ion Transitions Used in MRM Mode quantitation and confirmationa MRM transitions compound

native (m/z)

6:2 FTCA 8:2 FTCA 10:2 FTCA 6:2 FTUCAb 8:2 FTUCAb 10:2 FTUCAb

377-293 (63) 477-393 (63) 577-493 (63) 357-293 457-393 557-493

13C-labeled

(m/z)

379-295 (64) 479-395 (64) 579-495 (64) 359-294 459-394 559-494

a Confirmation ion shown in parentheses. See text for proposed structures of the fragment ions. b No confirmation MRM transitions were possible for the FTUCA (see text for details).

FIGURE 4. Degradation of FTCAs in methanol at 4 °C (top panel) and ambient temperature (∼20 °C, bottom panel).

FIGURE 3. Amount (pmol) of (A) 6:2, (B) 8:2, and (C) 10:2 FTUCAs formed with time by degradation of their corresponding FTCAs in solutions of methanol (O) and water (b). Each data point represents the arithmetic mean (n ) 3) ( 1 × standard error.

Concentrations of FTCAs, FTUCAs, and PFOS in Rainwater. The described method was applied to the analysis of rainwater collected in Winnipeg. Selected reaction monitoring chromatograms of FTCAs and FTUCAs in rainwater samples from Winnipeg are shown in Figures 5 and 6, respectively. Recovery and blank corrected concentrations of FTCAs were found to be 1.00 ( 0.08 ng/L (n ) 3) for 8:2 FTCA and 0.30 ( 0.04 ng/L (n ) 3) for 10:2 FTCA. The 8:2 FTUCA concentrations were 0.12 ( 0.02 ng/L (n ) 3) and 10:2 FTUCA 0.12 ( 0.01 ng/L. The concentration of PFOS was 1.5 times higher than the MDL [0.59 ( 0.04 ng/L (n ) 3)]. None of the PFCAs were detected in the rainwater samples. Although only one rainfall event was captured and analyzed in this study, there appears to be some correlation between the rank order of 8:2 and 10:2 FTOH reported by Stock et al. to the corresponding FTCAs reported here (13). Stock et al. reported concentrations for the 8:2 FTOH of approximately 10 pg/m3 in Winnipeg. The 10:2 FTOH was not detected above the method detection limit. While the concentrations of the 8:2 and 10:2 FTOHs are assumed to fluctuate over time, it is interesting to note that the FTCAs in rainwater were analogously dominated by 8:2 FTCA with a smaller amount of the 10:2 FTCA. The 8:2 and 10:2 FTUCAs were detected at the same concentration in the rainwater samples. One would assume a similarly high concentration of the 8:2 FTUCA relative to the 10:2 FTUCA if the FTUCAs are formed from degradation of the FTCAs. As this is not observed, the findings may be a result of differing transformation rates of the 8:2 and 10:2 FTCA to the corresponding FTUCAs. We hypothesize that both FTCA and FTUCA are abiotically formed, but it is unclear whether FTUCA is being formed in the atmosphere or in the rainwater. It would be interesting to monitor the change in FTCA concentrations over time in rainwater. In 15 minute smog chamber degradation experiments of 8:2 FTOH, PFCA compounds were found to contribute VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2949

FIGURE 6. Selected reaction monitoring chromatogram of 6:2, 8:2, and 10:2 FTUCAs in (A) method blank, (B) 4 L rainwater samples collected in Winnipeg, MB, and (C) external standard

FIGURE 5. Selected reaction monitoring chromatogram of 6:2, 8:2, and 10:2 FTCAs in (A) method blank, (B) 4 L rainwater samples collected in Winnipeg, MB, and (C) external standard. approximately 5% to the mass balance of products (20). Wet deposition of FOC acids is likely to be highly efficient due to their low vapor pressures and relatively high water solubility. The fact that PFCAs were not detected in our rainwater samples could be due to a combination of insufficient atmospheric concentrations of PFCAs and relatively high method detection limits. Although we are not certain about the atmospheric stability of FTCAs and FTUCAs we may assume that PFCAs will increase in the atmosphere over time given the appropriate conditions. As our samples were collected during a relatively wet summer cycle we hypothesise that wet deposition of FTCAs may dominate during relatively high summer precipitation cycles while wet deposition of PFCAs may become a more significant pathway when the time between rain events is relatively longer given appropriate atmospheric conditions. In order to properly study this phenomenon, method detection limits of PFCAs in water need to be improved. PFOS is believed to be formed from both biotic and abiotic degradation of neutral precursors much like PFCAs. While PFCAs appear to be atmospherically transported as the corresponding chain length FTOH, PFOS has many volatile precursors which may be transported atmospherically and then deposited as PFOS. In this study it is unclear whether PFOS precursors are being transported and subsequently wet deposited followed by degradation to PFOS or atmospherically degraded followed by wet deposition. While FTOHs are extremely volatile, known PFOS precursors such as N-ethyl2950

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

perfluorooctane sulfonamide (21) appear to be somewhat less volatile (longer retention times on DB5 GC column) and may themselves be subject to wet deposition. This warrants further investigation. However, this study does provide evidence for the abiotic degradation of PFOS precursors. Work is continuing to test whether there is a seasonal variability in the concentrations of FTCAs, FTUCAs, and PFOS in precipitation from Winnipeg. The method described in this study is a necessary first step in understanding the behavior of these compounds in the environment.

Acknowledgments We are grateful to Gilles Arsenault and Brock Chittim (Wellington Laboratories Inc., Guelph, ON) for the native and 13C2-labeled FTCA and FTUCA solutions. Sheryl Tittlemier (Health Canada, Ottawa, ON) and Paul Helm (Ministry of Environment, Toronto, ON) provided insightful comments on an earlier version of the manuscript.

Note Added after ASAP Publication This paper was released ASAP on March 17, 2005. The caption to Figure 4 was incorrect. The corrected version was published on April 28, 2005.

Literature Cited (1) Kissa, E. Fluorinated surfactants and repellents, 2nd ed.; Marcel Dekker Inc.: New York, 2001. (2) Key, B. D.; Howell, R. D.; Criddle, C. S. Fluorinated organics in the biosphere Environ. Sci. Technol. 1997, 31, 2445-2454. (3) Olsen, G. W.; Burris, J. M.; Mandel, J. H.; Zobel, L. R. Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluorochemical production employees. J. Occup. Environ. Med. 1999, 41, 799-806.

(4) Renner, R. Growing concern over perfluorinated chemicals. Environ. Sci. Technol. 2001, 35, 154A-160A. (5) Kannan, K.; Koistinen, J.; Beckmen, K.; Evans, T.; Gorzelany, J. F.; Hansen, K. J.; Jones, P. D.; Helle, E.; Nyman, M.; Giesy, J. P. Accumulation of perfluorooctane sulfonate in marine mammals. Environ. Sci. Technol. 2001, 35, 1593-1598. (6) Moody, C. A.; Kwan, W. C.; Martin, J. W.; Muir, D. C. G.; Mabury, S. A. Determination of perfluorinated surfactants in surface water samples by two independent analytical techniques: liquid chromatography/tandem mass spectrometry and 19F NMR. Anal. Chem. 2001, 73, 2200-2206. (7) Moody, C. A.; Martin, J. W.; Kwan, W. C.; Muir, D. C. G.; Mabury, S. A. Monitoring perfluorinated surfactants in biota and surface water samples following an accidental release of fire-fighting foam into Etobicoke Creek. Environ. Sci. Technol. 2002, 36, 545551. (8) Ellis, D. A.; Martin, J. W.; Mabury, S. A.; Hurley, M. D.; Sulbaek Andersen, M. P.; Wallington, T. J. Atmospheric lifetime of fluorotelomer alcohols. Environ. Sci. Technol. 2003, 37, 38163820. (9) Telomer Research Program (TRP) Update; Presented to US EPAOPPT; AR226-1141; 2002. (10) Martin, J. W.; Muir, D. C. G.; Moody, C. A.; Ellis, D. A.; Kwan, W. C.; Solomon, K. R.; Mabury, S. A. Collection of airborne fluorinated organics and analysis by gas chromatography/ chemical ionization mass spectrometry. Anal. Chem. 2002, 74, 584-590. (11) Betts, K. S. Potential explanation for fluorinated compounds’ persistence. Environ. Sci. Technol. 2003, 37, 312A-313A. (12) Stock, N. L.; Ellis, D. A.; Deleebeeck, L.; Muir, D. C. G.; Mabury, S. A. Vapor pressures of the fluorinated telomer alcoholss limitations of estimation methods. Environ. Sci. Technol. 2004, 38, 1693-1699. (13) Stock, N. L.; Lau, F. K.; Ellis, D. A.; Martin, J. W.; Muir, D. C. G.; Mabury, S. A. Polyfluorinated telomer alcohols and sulfonamides in the North American troposphere. Environ. Sci. Technol. 2004, 38, 991-996.

(14) Hagen, D. F.; Belisle, J.; Johnson, J. D.; Venkateswarlu, P. Characterization of fluorinated metabolites by a gas chromatographic-helium microwave plasma detector- the biotransformation of 1H,1H,2H,2H-perfluorodecanol to perfluorooctanoate. Anal. Biochem. 1981, 118, 336-343. (15) Lange, C. C. Biodegradation Screen study for telomer-type alcohols; AR226-1149; 3M Environmental Laboratory: St. Paul, MN, 2002. (16) Dinglasan, M. J. A.; Ye, Y.; Edwards, E. A.; Mabury, S. A. Fluorotelomer alcohol biodegradation yields poly- and perfluorinated acids. Environ. Sci. Technol. 2004, 38, 28572864. (17) Winefordner, J. D.; Long, G. L. Limit of Detection. A Closer look at the IUPAC definition. Anal. Chem. 1983, 55, 712A-724A. (18) Willoughby, R.; Sheehan, E.; Mitrovich, S. A Global View of LC/ MS. How to Solve your most Challenging Analytical Problems, 2nd ed.; Global View Publishing: Pittsburgh, PA, 2003. (19) McLafferty, F. W. Interpretation of Mass Spectra, 3rd ed.; University Science Books: Mill Valley, CA, 1980; pp 1-303. (20) Ellis, D. A.; Martin, J. W.; De Silva, A. O.; Mabury, S. A.; Hurley, M. D.; Sulbaek Andersen, M. P.; Wallington, T. J. Degradation of fluorotelomer alcohols: a likely atmospheric source of perfluorinated carboxylic acids. Environ. Sci. Technol. 2004, 38, 3316-3321. (21) Tomy, G. T.; Tittlemier, S. A.; Palace, V. P.; Budakowski, W. R.; Braekevelt, E.; Brinkworth, L.; Friesen, K. J. Biotransformation of N-Ethyl Perfluoroooctane Sulfonamide (N-EtPFOSA) by Rainbow Trout (Onchorhynchus mykiss) Liver Microsomes. Environ. Sci. Technol. 2004, 38, 758-762.

Received for review September 1, 2004. Revised manuscript received January 28, 2005. Accepted February 8, 2005. ES048635B

VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2951