Anal. Chem. 2010, 82, 4551–4557
Trace Analytical Methods for Semifluorinated n-Alkanes in Snow, Soil, and Air Merle M. Plassmann* and Urs Berger Department of Applied Environmental Science (ITM), Stockholm University, 106 91 Stockholm, Sweden Semifluorinated n-alkanes (SFAs) are anthropogenic chemicals that are used in ski waxes and, thus, are released directly into the environment, but their subsequent fate and distribution are as yet unknown. Therefore, simple, selective, and sensitive methods were developed for analyzing trace amounts of SFAs in snow/water, soil, and air samples by gas chromatography coupled to electron capture negative ionization mass spectrometry (GC/ECNIMS). Recoveries were generally in the range of 70-120%, depending on the compound and matrix. The analytical sensitivity was higher for SFAs with longer fluorinated chains, and the instrumental limits of detection ranged from 0.3 to 260 pg injected, providing method detection limits of 0.54-311 ng L-1, 0.004-9.86 ng g-1, and 0.4-531 ng m-3 for snow (analyzed as its meltwater), soil, and air samples, respectively. Using the developed procedures, SFAs were found in snow (meltwater) and soil samples from a small cross-country ski area in Sweden at concentrations up to 1.3 µg L-1 and 47 pg g-1, respectively. Diblock semifluorinated n-alkanes (SFAs) are anthropogenic chemicals with the general formula F(CF2)n(CH2)mH (or, briefly, FnHm). They are manufactured with a wide variety of chain lengths, depending on the intended use, by adding an olefin to a perfluoroalkyl iodide followed by reductive dehalogenation.1 These reactions can lead to byproducts such as semifluorinated alkenes2 (SFAenes), with the general formula F(CF2)nCHdCH(CH2)m-2H (or, briefly, FnHmene). Since the 1990s, long-chain SFAs (those with 22 carbon atoms or more) have been frequently applied in ski waxes, since they reduce friction and repel dirt due to their extremely low surface tension,3 while shorter-chain SFAs are used in medicinal applications.4,5 Fluorinated ski waxes are mostly used in crosscountry skiing under wet conditions, in which the water repellence of the fluorinated molecules enhances the glide on the water film generated between the base of the ski and the snow surface. In fluorinated ski waxes, up to 15% of either perfluorocarbons * To whom correspondence should be addressed. E-mail: merle.plassmann@ itm.su.se. Phone: +46 8674 7188. Fax: +46 8674 7637. (1) Napoli, M. J. Fluorine Chem. 1996, 79, 59–69. (2) Coe, L.; Milner, E. J. Organomet. Chem. 1972, 39, 395–402. (3) Rogowski, I.; Leonard, D.; Gauvrit, J. Y.; Lanteri, P. Cold Reg. Sci. Technol. 2007, 49, 145–150. (4) Riess, J. G. Chem. Rev. 2001, 101, 2797–2920. (5) Kirchhof, B.; Wong, D.; Van Meurs, J.; Hilgers, R. D.; Macek, M.; Lois, N.; Schrage, N. F. Am. J. Ophthalmol. 2002, 133, 95–101. 10.1021/ac1005519 2010 American Chemical Society Published on Web 05/06/2010
(perfluorinated alkanes) or SFAs are mixed with normal paraffins.6 No information on production volumes of SFAs is publically available. However, the current worldwide use of fluorinated substances in ski waxes was roughly estimated to several tons per year (personal communication with the ski wax manufacturer Rex, Hartola, Finland). There has also been discussion regarding the possible advantages of including other fluorinated substances, e.g., triblock SFAs7 and tetrakis(perfluoroalkyl)alkane,8 in ski waxes. The waxes are applied to ski bases by ironing or brushing, which can lead to aerosol formation.9 Hence, people waxing skis may inhale fluorinated chemicals, with consequent risks of respiratory distress symptoms10 and decreased CO diffusion capacity.11 Toxicological studies on SFAs are very scarce. The half-life of F6H10ene in rat liver was determined to be 25 days12 and in vitro experiments on F2H4 with rat liver microsomes showed metabolism to the 5-hydroxy derivative.13 Perfluorinated carboxylates (PFCAs) are a group of chemicals that are ubiquitously distributed in the environment, very persistent, and toxic to a certain extent.14,15 Elevated levels of PFCAs (increasing during the course of a ski season) have been found in blood samples from ski service technicians exposed to fluorinated waxes compared to the general population.16,17 These increases could be attributed to the emission and take-up of fluorinated substances (residual PFCAs in the waxes or other fluorinated substances that are transformed during heating and/ or are metabolized after uptake). Abrasion of SFAs from the ski sole during skiing leaves them on the snow surface. This could lead to high amounts of SFAs and possible degradation products in skiing areas. However, no (6) Charonnat, N. www.rideandglide.bizland.com/fluoro_waxing.htm (accessed January 5, 2010). (7) Conte, L.; Zaggia, A.; Sassi, A.; Seraglia, R. J. Fluorine Chem. 2007, 128, 493–499. (8) Gambaretto, G.; Conte, L.; Fornasieri, G.; Zarantonello, C.; Tonei, D.; Sassi, A.; Bertani, R. J. Fluorine Chem. 2003, 121, 57–63. (9) Liesivuori, J.; Kiviranta, H.; Laitinen, J.; Hesso, A.; Ha¨meila¨, M.; Tornaeus, J.; Pfa¨ffli, P.; Savolainen, H. Ann. Occup. Hyg. 1994, 38, 931–937. (10) Bracco, D.; Favre, J. B. Ann. Emerg. Med. 1998, 32, 616–619. (11) Kno ¨pfli, B.; Guntensperger, U.; Schibler, A.; Villiger, B. Schweiz. Rundsch. Med./Prax. 1992, 81, 884–887. (12) Zarif, L.; Postel, M.; Septe, B.; Trevino, L.; Riess, J. G.; Mahe´, A.-M.; Follana, R. Pharm. Res. 1994, 11, 122–127. (13) Baker, M. H.; Foster, A. B.; Hegedus, L.; Jarman, M.; Rowlands, M. G. Biomed. Mass Spectrom. 1984, 11, 512–521. (14) Giesy, J. P.; Kannan, K. Environ. Sci. Technol. 2002, 36, 146A–152A. (15) Kissa, E. Fluorinated surfactants and repellents, Surfactant Science Series, 2nd ed.; Marcel Dekker: New York, 2001; Vol. 97. (16) Nilsson, H.; Ka¨rrman, A.; Westberg, H.; Bavel, B. v.; Lindstro¨m, G. Organohalogen Compd. 2008, 40. (17) Nilsson, H.; Ka¨rrman, A.; Westberg, H.; Rotander, A.; Bavel, B. v.; Lindstro ¨m, G. Environ. Sci. Technol. 2010, 6, 2150–2155.
Analytical Chemistry, Vol. 82, No. 11, June 1, 2010
4551
information has been published regarding the environmental distribution and fate of SFAs. Therefore, the aim of this study was to develop methods with sufficiently high selectivity and sensitivity for analyzing trace amounts of SFAs in snow, soil, and indoor air. In a first step, ski waxes were analyzed and their SFA composition was determined using mass spectrometry (MS) applying electron capture negative ionization (ECNI) and electron ionization (EI) techniques. The latter was also applied in the only previously reported MS based analysis of short-chain SFAs.18 Methods for extracting SFAs from samples of relevant environmental compartments, including snow/water, soil, and air were then developed. Finally, these methods were applied to snow and soil samples, and the extracted SFAs were analyzed by gas chromatography (GC) coupled to ECNI-MS, thus providing the first data on concentrations of SFAs in the environment. EXPERIMENTAL SECTION Chemicals. SFAs of chain lengths F6H16, F10H16, F12H14, F12H16, and F12H16ene (all of purity >95% according to 1H- and 19 F-NMR analysis) were custom-synthesized by Synthon-Lab Ltd. (St. Petersburg, Russia). F6H8, F6H14, F8H10, and F8H16 (unspecified purity) were acquired from ABCR (Karlsruhe, Germany), and F10H2 (97% purity) was purchased from Apollo Scientific (Stockport, England). Four ski waxes with the brand names Ski Go HF Blue, Ski Go LF White, Swix LF4 Green, and Toko LF Diblock Yellow were purchased in local sport shops. Two types of fluorinated ski wax raw materials, a mixture of perfluorocarbons (RM-PFC), and a mixture of SFAs (RM-SFA) were kindly provided by the ski wax producer Rex (Hartola, Finland). A mixture of mass-labeled internal standards of perfluorinated carboxylates (PFCAs) and sulfonates (PFSAs) (MPFAC-MXA) and mixtures of native PFCAs and PFSAs were purchased from Wellington Laboratories Inc. (Guelph, Canada). Cyclohexane (reagent grade) was purchased from Scharlau (Barcelona, Spain), and methanol (licrosolv) was from Merck (Darmstadt, Germany). Characterization of SFA Profiles in Ski Waxes. In preparation for qualitative analysis, each of the four ski waxes and two raw materials were dissolved in cyclohexane to a concentration of 1 mg mL-1 by sonication for half an hour at room temperature. Then, the mixtures were centrifuged (10 min at 2500 rpm) to remove undissolved material. The supernatants were diluted ten- to a hundred-fold with cyclohexane prior to analysis by GC/MS, as described below. Quantitative Analysis of PFCAs and PFSAs in Ski Waxes. PFCAs and PFSAs were quantitatively analyzed in each of the four ski waxes and two fluorinated raw materials RM-PFC and RMSFA. They were extracted by dissolving 2 mg of each material in 2 mL of methanol, sonicating for 20 min, leaving the mixture overnight at room temperature, and sonicating again for 20 min. After centrifugation (10 min at 2500 rpm), 1 mL of each supernatant was removed, spiked with 25 µL of the MPFAC-MXA internal standard mixture (44 pg µL-1 of each analyte in methanol), and concentrated to 100 µL. After adding 50 µL of 3,7-dimethyl branched perfluorodecanoic acid (20 pg µL-1 in methanol) as a volumetric standard and 150 µL of 4 mM (18) Napoli, M.; Krotz, L.; Conte, L.; Seraglia, R.; Traldi, P. Rapid Commun. Mass Spectrom. 1993, 7, 1012–1016.
4552
Analytical Chemistry, Vol. 82, No. 11, June 1, 2010
aqueous ammonium acetate, the mixture was analyzed using a high performance liquid chromatograph (Waters Alliance 2695; Milford, MA) coupled to a negative ion electrospray tandem mass spectrometer (Micromass Quattro II; Altrincham, UK). Details of the instrumental method are given elsewhere.19 Analysis of SFAs in Indoor Air. Isolute ENV+ solid phase extraction (SPE) cartridges (200 mg, 6 mL) from Biotage (Uppsala, Sweden) were used to enrich SFAs from air samples, as follows. The cartridges were rinsed with 3 mL of cyclohexane and dried by drawing air through them for 2 min using a SPE vacuum manifold. Air (approximately 0.3 m3) was then drawn through them at a flow rate of 1.25 L min-1 for 4 h using a Leland Legacy Sample Pump from SKC Inc. (Eighty Four, USA). SFAs were subsequently eluted from the cartridges with 10 mL of cyclohexane, and the resulting extracts were concentrated to 0.5 mL at 30 °C under a gentle flow of nitrogen. F8H10 (5 µL of a 50 ng µL-1 solution in cyclohexane) was added as volumetric standard prior to analysis by GC/MS, as described below. Analysis of SFAs in Snow/Water Samples. Water samples (500 mL of tap water for method development) were extracted twice with 30 mL of cyclohexane by liquid-liquid extraction (LLE) using a 1 L separation funnel. Before adding the cyclohexane to the water sample in the funnel, it was used to rinse the sample bottle. The combined extract was concentrated under nitrogen to about 0.5 mL and cleaned up using Silica SPE cartridges (1 g, 6 mL) from Biotage. The cartridges were conditioned with 3 mL of cyclohexane; the sample was applied, and the SFAs were eluted with 5 mL of cyclohexane. The resulting extract was concentrated again to 0.5 mL, and F8H10 (5 µL of a 50 ng µL-1 solution in cyclohexane) was added as volumetric standard, prior to analysis by GC/MS (see below). To test the applicability of the method to real samples, snow was collected from a ski area in Sa¨fsen, Sweden (approximately 270 km northwest of Stockholm), in March 2009. Samples were taken at the starting points of two different cross-country skiing tracks (designated sites A and B) and in front of the ski rental station (designated site C) by filling a 1 L polypropylene bottle with the top layer (approximately 1 cm depth) of the snow. The snow was then allowed to melt at room temperature, yielding 300-400 mL of water, containing abundant particles. Finally, SFAs in the meltwater, including the particles, were extracted and analyzed using the method described above. Analysis of SFAs in Soil Samples. Samples of approximately 5 g of soil (sandy loam purchased from Lufa, Speyer, Germany, for method development) were placed in a 50 mL PP tube, and 10 mL of cyclohexane was added. The resulting suspension was vortex-mixed for 1 min, sonicated for 30 min at room temperature, and centrifuged for 10 min at 2500 rpm. The supernatant was carefully removed, and the extraction was repeated with 10 mL of cyclohexane. The combined extract was concentrated to 0.5 mL under nitrogen and cleaned up using the Silica SPE method applied to the water samples (see above). The final extract was concentrated to 100 µL, and F8H10 (1 µL of a 50 ng µL-1 solution in cyclohexane) was added as a volumetric standard, prior to analysis by GC/MS (see below). (19) McLachlan, M. S.; Holmström, K. E.; Reth, M.; Berger, U. Environ. Sci. Technol. 2007, 41, 7260–7265.
Surface soil samples (approximately 3 cm deep) were taken from Sa¨fsen, Sweden, in September 2009 at two of the sites (A and C), where the snow had been sampled 6 months earlier. A depth profile (four layers, each of 3 cm depth) and surface soil samples from points approximately 1 and 5 m away from the track were also taken at site A. The samples were air-dried for 2 days at room temperature and sieved to 0.996 (Table 3). The analytes were quantified by external calibration following normalization of peak areas to that of F8H10, the volumetric
Figure 2. Comparison of chromatograms (GC/ECNI-MS, TIC of SIM windows) of a standard, the RM-SFA solution, the snow sample extract of site A, track, and the soil sample extract of site A 1 (see Table 4 for quantified concentrations). For better clarity, only the part of the soil sample extract chromatogram containing detectable SFAs is shown. Authentic standards of F10H16ene, F14H16, and F16H16 were not available. These compounds were assigned on the basis of ion mass-to-charge ratios and expected retention times. The highest peak in each chromatogram was normalized to 100% relative abundance. Analytical Chemistry, Vol. 82, No. 11, June 1, 2010
4555
Table 3. Method Validation Parameters for All Target SFAs and Examined Environmental Matrixes F10H2
F6H8
LOD/LOQ [pg absolute] linear detector range [ng absolute] MDL/MQL air method [ng/m3] MDL/MQL water method [ng/L] MDL/MQL soil method [ng/g]
2.4/8.0 0.008-5
mix 1 mix 2 mix 3
58 (3.4) 58 (3.8) 49 (8.8)
mix 1 mix 2 mix 3
Water Method 9 (65) 45 (60) 6 (39) 28 (43) 24 (4.7) 24 (42)
mix 1 mix 2 mix 3
nd nd nd
a
F6H14
F6H16
72/241 35/116 215/716 0.24-200 0.12-200 0.72-200
11.3/37.6 131/437 22.6/75.3 114/379 ndb nd
49.5/165 50.0/167 nd
346/1150 229/763 5.59/18.6
F8H16a
F12H14 0.5/1.5 0.002-2
F12H16 0.6/1.9 0.002-2
F12H16ene 0.3/0.8 0.001-2
531/1770 2.80/9.34 2.58/8.59 0.98/3.26 0.40/1.34 311/1040 4.5/15.0 0.54/1.79 0.83/2.76 0.67/2.24 9.86/32.9 0.45/1.50 0.007/0.024 0.007/0.023 0.004/0.012
Air Method Recoveries (Standard Deviation), n 99 (4.3) 97 (6.5) 93 (6.3) 88 (6.3) 97 (6) 95 (3.9) 93 (4) 97 (2.9) 91 (4.6) 87 (6.1) 87 (2.6) 105 (5.1) Recoveries 85 (8.3) 67 (15) 58 (26)
F10H16
260/867 2.9/9.5 0.87-100 0.01-5
) 4 [%] 90 (5.6) 110 (3.5) 113 (5.4)
(Standard Deviation), n ) 4 [%] 88 (8.3) 82 (9.6) 82 (12) 74 (11.4) 78 (13) 82 (12) 67 (22) 76 (18) 83 (18)
Soil Method Recoveries (Standard Deviation), n nd nd 110 (6.3) 74 (3.1) nd nd 124 (22.5) 78 (6.6) nd nd 113 (22.9) 88 (4.2)
) 4 [%] 74 (3.2) 95 (8.3) 125 (5.2)
92 (5.3) 114 (4.1) 112 (2.9)
92 (4.9) 126 (4.8) 122 (2.2)
97 (4.2) 135 (5.6) 115 (5.4)
83 (9.5) 79 (13) 89 (31)
84 (12) 88 (14) 91 (23)
86 (12) 90 (12) 87 (20)
76 (4.2) 112 (10.3) 137 (5.7)
70 (4.7) 111 (11.7) 141 (5)
74 (6.6) 107 (13.3) 119 (9.1)
Standard impure. b nd: not determined.
standard. F8H10 was chosen as a volumetric standard since it is not known to be used for purposes other than research investigations of monolayers and membranes,23 it was not found in any of the ski waxes, and it is not expected to be present in environmental samples. The recoveries and reproducibility of all three extraction methods were found to be very good (with the exception of recoveries for two short-chain analytes in water, see Table 3 and discussion below), which was essential since external quantification was applied in the study, and no recovery correction of quantified results was performed. Analysis of SFAs in Indoor Air. This method was mainly developed for planned laboratory experiments. It is, therefore, based on sampling of relatively low air volumes (0.3 m3) at room temperature and could also be used for workplace measurements or in conjunction with personal samplers to evaluate the exposure of people waxing skis. Good recoveries were obtained for SFAs using Isolute ENV+ cartridges with three different spike levels (Table 3). Repeatability was excellent, with standard deviations below 8.8%. MDLs/MQLs for the entire air method were in the low to high ng m-3 range (Table 3). No detectable traces of any of the target analytes were found in blank extracts, and no breakthrough was observed in tests with two cartridges sequentially coupled; i.e., no SFAs were found in the downstream cartridge. Tests with a substantially higher volume (1.5 m3 instead of 0.3 m3) were also conducted by sampling for 20 h instead of 4 h. This resulted in slightly lower recoveries, especially for the shorter-chain SFAs, but recoveries for F6H16 and longerchain SFAs were still in the range of 60-90%. Thus, there is potential for up-scaling the method for analyzing higher air volumes, which would lead to lower MDLs. The potential utility of small polyurethane foam (PUF) plugs and Oasis HLB cartridges (Waters) for sampling air was also tested. Use of the former resulted in total losses of F6H16 and (21) Stock, N. L.; Ellis, D. A.; Deleebeeck, L.; Muir, D. C. G.; Mabury, S. A. Environ. Sci. Technol. 2004, 38, 1693–1699. (22) Martin, J. W.; Muir, D. C. G.; Moody, C. A.; Ellis, D. A.; Kwan, W. C.; Solomon, K. R.; Mabury, S. A. Anal. Chem. 2002, 74, 584–590. (23) Krafft, M. P.; Riess, J. G. Chem. Rev. 2009, 109, 1714–1792.
4556
Analytical Chemistry, Vol. 82, No. 11, June 1, 2010
shorter-chain SFAs and breakthrough of some longer-chain SFAs after sampling of 0.15 m3 air. The Oasis matrix provided similar recoveries to Isolute ENV+, but Isolute ENV+ cartridges were chosen since they have also been shown to give good results for the enrichment of other fluorinated substances, e.g., fluorotelomer alcohols from indoor air.24 Analysis of SFAs in Snow/Water. LLE was chosen over other possible extraction methods, e.g., SPE, since it can also extract analytes adsorbed to the sample bottle. This proved to be important for SFAs, since lower recoveries were obtained when the sample bottle was not rinsed with cyclohexane. Particles coextracted during LLE were removed from the extracts in the following Silica SPE cleanup step. Florisil columns were also tested for cleanup and provided good recoveries when a fluorinated solvent was used for elution. Silica SPE cleanup was chosen since it worked well with cyclohexane and provided better recoveries than florisil for the longer-chain SFAs. Recoveries, MDLs, and MQLs for the water extraction method can be found in Table 3. MDLs/MQLs were in the low to high ng L-1 range for the target compounds. No extraction blank contamination was present at the given MDLs. Recoveries for two of the short-chain SFAs, F10H2 and F6H8, were not satisfactory (6-45%), probably due to volatilization during concentration of the cyclohexane extract with nitrogen. The amount of solvent used was, therefore, kept to a minimum. No further method optimization was attempted for F10H2 and F6H8, since these compounds are not used in ski waxes and were, therefore, not relevant for environmental samples. SFAs were detected in all five environmental snow samples. The quantified concentrations can be found in Table 4. The values for F10H16ene, F14H16, and F16H16 have to be considered semiquantitative, since no standards were available for these analytes. The method used to estimate rrf for F10H16ene, F14H16, and F16H16 is described in the Experimental Section. In addition (24) Jahnke, A.; Huber, S.; Temme, C.; Kylin, H.; Berger, U. J. Chromatogr., A 2007, 1164, 1–9.
Table 4. Concentrations of SFAs Found in Snow and Soil Samples F6H16 site site site site site
Ab, track A, next to track B, track 1 B, track 2 C
795
