Anal. Chem. 2009, 81, 4256–4262
Linear and Branched Perfluorooctane Sulfonate Isomers in Technical Product and Environmental Samples by In-Port Derivatization-Gas Chromatography-Mass Spectrometry Shaogang Chu and Robert J. Letcher* Wildlife and Landscape Science Directorate, Science and Technology Branch, Environment Canada, National Wildlife Research Centre, Carleton University, Ottawa, ON, K1A 0H3, Canada Perfluorooctane sulfonate (PFOS) is found globally as an environmental contaminant and is highly bioaccumulative in exposed biota including humans. However, there is a dearth of environmental information on the isomeric profile of PFOS, especially in biological samples, which requires suitable analysis methods for the identification and quantification of ultratrace amounts. In the present study, a novel method was developed that incorporates clean up by solid-phase extraction (SPE) WAX cartridges and in-port derivatization-gas chromatography-mass spectrometry (GC/MS) to identify and quantitatively determine linear PFOS (L-PFOS) and branched (monotrifluoromethyl and bistrifluoromethyl) isomers in PFOS technical product and in environmentally relevant biological samples. Tetrabutylammonium hydroxide (TBAH) was used for derivatization via an in situ pyrolytic alkylation reaction that occurred in the GC injector and generated butyl PFOS isomer derivatives. In addition to L-PFOS, ten branched PFOS isomers were identified in the technical product. The environmental relevance of branched PFOS isomers in addition to L-PFOS was evidenced by the presence of six branched and L-PFOS in representative herring gull and double-crested cormorant egg, and polar bear liver and plasma samples from the Great Lakes and Arctic, respectively. For all PFOS isomers in the technical product and biota samples the method demonstrated high sensitivity with the limit of detection (LOD) ranging from 0.05 to 0.25 ng/mL, with exception of L-PFOS where the LOD was 1.46 ng/mL. For the biotic samples, the method detection limits (MDLs) were slightly higher than the LODs and ranged from 0.09 to 0.46 ng/g wet weight (w.w.) with exception of L-PFOS (MDL ) 6.87 ng/g w.w.). Polyfluoroalkyl and perfluoroalkyl compounds (PFCs) are a scientific and public concern since they have been found in various compartments of the environment.1 Among the environmentally relevant PFCs, perfluorooctane sulfonate (PFOS) is the most prominent in wildlife and has been found globally in many species * To whom correspondence should be addressed. E-mail: robert.letcher@ ec.gc.ca. Phone: +1 613 998-6696. Fax: +1 613 998-0458. (1) Houde, M.; Martin, J. W.; Letcher, R. J.; Solomon, K. R.; Muir, D. C. G. Environ. Sci. Technol. 2006, 40, 3463–3473.
4256
Analytical Chemistry, Vol. 81, No. 11, June 1, 2009
and especially in birds and mammals at the top of aquatic food webs.1-3 PFOS is similarly persistent and bioaccumulative as other well documented, environmentally persistent organic pollutants (POPs) such as the insecticide 1,1-bis(4-chlorophenyl),2,2,2trichloroethane (DDT) and polychlorinated biphenyl (PCBs). The hydrophilic and lipophilic (amphiphilic) nature of perfluorinated sulfonates such as PFOS means that the physio-chemical properties that govern their bioaccumulation and pharmacokinetics are different from lipophilic POPs whose persistence and bioaccumulation can be predicted from, for example, octanol-water partition coefficients (Kow). PFCs including PFOS show preferential protein binding (e.g., serum and fatty acid binding proteins) and partitioning into protein-rich tissues such as the liver and blood.4 On the basis of various configurations of the C8 hydrocarbon chain, hundreds of PFOS isomers are theoretically possible; however, thus far a maximum of eleven isomers have reported in technical PFOS products.5-9 Very recently, Houde et al.5 reported the separation but not the specific structural identification of three monotrifluoromethyl and three bistrifluoromethyl PFOS isomers in addition to the major isomer, linear PFOS (L-PFOS), in technical PFOS product, as well as in sediment, water zooplankton and fish samples from Lake Ontario. Factors such as the food web and different sources were shown to contribute to PFOS isomer fractionation relative to a technical PFOS standard. Apart for the aquatic food web study of Houde et al.,5 investigations of PFOS isomers and fractionation in wildlife are to our knowledge non-existent. (2) Martin, J. W.; Whittle, D. M.; Muir, D. C. G.; Mabury, S. A. Environ. Sci. Technol. 2004, 38, 5379–5385. (3) Tomy, G. T.; Budakowski, W.; Halldorson, T.; Helm, P. A.; Stern, G. A.; Friesen, K.; Pepper, K.; Tittlemier, S. A.; Fisk, A. T. Environ. Sci. Technol. 2004, 38, 6475–6481. (4) Jones, P. D.; Hu, W.; De Coen, W.; Newsted, J. L.; Giesy, J. P. Environ. Toxicol. Chem. 2003, 22, 2639–2649. (5) Houde, M.; Czub, G.; Small, J. M.; Backus, S.; Wang, X.; Alaee, M.; Muir, D. C. G. Environ. Sci. Technol. 2008, 42, 9397–9403. (6) Vyas, S. M.; Kania-Korwel, I.; Lehmler, H. J. J. Environ. Sci. Health Part A 2007, 42, 249–255. (7) Langlois, I.; Berger, U.; Zecak, Z.; Oehme, M. Rapid Commun. Mass Spec. 2007, 21, 3547–3553. (8) Benskin, P. J.; Bataineh, M.; Martin, J. W. Anal. Chem. 2007, 79, 6455– 6464. (9) Arsenault, G.; Chittim, B.; Gu, J.; McAlees, A.; McCrindle, R.; Robertson, V. Chemosphere 2008, 73, S53–S59.
10.1021/ac8027273 CCC: $40.75 Published 2009 by the American Chemical Society Published on Web 04/29/2009
Table 1. Chemical Formulas and Names of PFOS Isomers and Their Abbreviations
a
abbreviation
chemical formulas
chemical name
L-PFOS P1MHpS P2MHpS P3MHpS P4MHpS P5MHpS P6MHpS P35DMHxS P44DMHxS P45DMHxS P55DMHxS MPFOSa
CF3CF2CF2CF2CF2CF2CF2CF2SO3H CF3CF2CF2CF2CF2CF2CF(CF3)SO3H CF3CF2CF2CF2CF2CF(CF3)CF2SO3H CF3CF2CF2CF2CF(CF3)CF2CF2SO3H CF3CF2CF2CF(CF3)CF2CF2CF2SO3H CF3CF2CF(CF3)CF2CF2CF2CF2SO3H CF3CF(CF3)CF2CF2CF2CF2CF2SO3H CF3CF(CF3)CF2CF(CF3)CF2CF2SO3H CF3CF2C(CF3)2CF2CF2CF2SO3H CF3CF(CF3)CF(CF3)CF2CF2CF2SO3H CF3C(CF3)2CF2CF2CF2CF2SO3H CF3CF2CF2CF213CF213CF213CF213CF2SO3H
n-perfluoro-1-octanesulfonate perfluoro-1-methyl-heptanesulfonate perfluoro-2-methyl-heptanesulfonate perfluoro-3-methyl-heptanesulfonate perfluoro-4-methyl-heptanesulfonate perfluoro-5-methyl-heptanesulfonate perfluor-5-methyl-heptanesulfonate perfluoro-3,5-dimethyl-hexanesulfonate perfluoro-4,4-dimethyl-hexanesulfonate perfluoro-4,5-dimethyl-hexanesulfonate perfluoro-5,5-dimethyl-hexanesulfonate n-perfluoro-1-[1,2,3,4-13C4]octanesulfonate
Internal standard.
Attempts have been made to quantify branched PFOS isomers in technical product by using 19F-NMR spectroscopy,6,9 although such methods are not suitable for complex environmental matrixes since there is inherently poor sensitivity and the presence of other fluorinated (and 19F-containing) interferents. The determination of PFCs, including PFOS, in environmental samples has been possible mainly because of advances in liquid chromatographyelectrospray tandem mass spectrometry (HPLC/MS/MS).1,10 However, HPLC has shown relatively poor separation of branched PFOS isomers. Even with the advent of ultrahigh pressure liquid chromatography (UPLC), attempts in separating all branched PFOS isomers still has not been optimally achieved.5,8,11-14 In environmental samples PFOS has almost exclusively been measured as a total L-PFOS concentration using HPLC/MS/MS based methods, where an alkyl reversed-phase column has been used.1,5,10 In using such methods it has been assumed that all PFOS isomers have the same MS response as L-PFOS. Furthermore, it has also been assumed that environmental behavior is the same for all PFOS isomers, as well as the toxicology and effects in exposed organisms.15 Compared with HPLC, GC is a superior approach with ability to more effectively separate PFOS isomers, although a derivatization process is necessary.7 In contrast to other common hydrocarbon compounds, the perfluorinated alkyl skeleton of PFOS has been shown to present a thermal and/or kinetic challenge in derivatization reactions. Langlois et al.7 used catalyzed esterification methods to derivatize PFOS, and found that the derivatization reaction required 12 h or more (with shaking), which still resulted in poor quantitative derivatization. Some GC/ MS methods have been used to determine perfluorocarboxylic acids in environmental samples.16,17 However, to date and to our (10) Martin, J. W.; Kannan, K.; Berger, U.; de Voogt, P.; Field, J.; Franklin, J.; Giesy, J. P.; Harner, T.; Muir, D. C. G.; Scott, B.; Kaiser, M.; Jarnberg, U.; Jones, K. C.; Mabury, S. A.; Schroeder, H.; Simcik, M.; Sottani, C.; van Bavel, B.; Karrman, A.; Lindstrom, G.; van Leeuwen, S. Environ. Sci. Technol. 2004, 38, 248A–255A. (11) Arsenault, G.; Chittim, B.; McAlees, A.; McCrindle, R.; Riddell, N.; Yeo, B. Chemosphere 2008, 70, 616–625. (12) Ka¨rrman, A.; Langlois, I.; van Bavel, B.; Lindstrom, G.; Oehme, M. Environ. Int. 2007, 33, 782–788. (13) Langlois, I.; Oehme, M. Rapid Commun. Mass Spectrom. 2006, 20, 844– 850. (14) Ochoa-Herrera, V.; Sierra-Alvarez, R.; Somogyi, A.; Jacobsen, N. E.; Wysocki, V. H.; Field, J. A. Environ. Sci. Technol. 2008, 42, 3260–3264. (15) de Voogt, P.; Saez, M. TrAC-Trends Anal. Chem. 2006, 25, 326–342. (16) Alzaga, R.; Bayona, J. M. J. Chromatogr., A. 2004, 1042, 155–162.
knowledge there have been no published reports on a totally successful methodology (including HPLC/MS, GC/MS, 19F-NMR based) for the quantitative determination in environmental samples of the maximum number of branched PFOS isomers. Optimal analytical methods are required to determine PFOS isomers in environmental samples and particularly in wildlife and their food webs as well as in humans to elucidate sources and pathways, accumulation, and fate. The goal of the present study was to develop a rapid, sensitive, and reliable analysis method to maximally identify and accurately quantify branched and linear PFOS isomers in both technical product and environmentally relevant biological samples. EXPERIMENTAL SECTION Chemicals and Reagents. All the standard solutions of branched and linear PFOS isomers were purchased from Wellington Laboratories (Guelph, ON, Canada). The molecular formula of the analytical standard compounds, including the mass labeled internal standard, and their abbreviations are given in Table 1. The concentrations of target compound in the individual standard solutions of all PFOS isomers were 1.00 µg/mL in methanol. A standard mixture solution that contained P45DMHxS and P35DMHxS was also used where the concentrations were 1.00 and 0.50 µg/mL, respectively. In these standard solutions, with exception of P1MHpS, branched PFOA isomers were also present. For example, in the P2MHpS standard solution there was perfluoro-2-methylheptanoic acid with a concentration of 1.05 µg/ mL. No details were provided by Wellington Laboratories as to whether other impurities were present in the solutions. The concentration of standard solutions of sodium linear perfluorooctane sulfonate (L-PFOS) and sodium perfluoro-1-[1,2,3,413 C4]octane sulfonate (MPFOS) were 50 µg/mL in methanol. The solution of technical potassium perfluorooctane sulfonate product (T-PFOS) was also obtained from Wellington Laboratories (Guelph, ON, Canada) at a concentration of 50 µg/mL in methanol. It should be noted that the concentration of pure potassium PFOS salts in the mixture was 40 µg/mL, and the concentration of pure PFOS was recalculated based on its free acid form. Therefore, the actual total concentration of the free acid form of PFOS in this standard solution was 37.2 µg/mL. Tetrabutylammonium hydroxide 30 hydrate (TBAH) was purchased from Sigma-Aldrich (Oakville, ON, Canada). Formic (17) Belisle, J.; Hagen, D F. Anal. Biochem. 1980, 101, 369–376.
Analytical Chemistry, Vol. 81, No. 11, June 1, 2009
4257
acid (98-100%), ammonium hydroxide (28-30% w/v), and other chemicals were analytical grade or better and also purchased from Sigma-Aldrich (Oakville, ON, Canada). Oasis WAX SPE cartridges (3 mL, 60 mg, 30 µm) were from Waters (Mississauga, ON, Canada). All the solvents used in the experiment were HPLC grade from Fisher Scientific (Ottawa, ON, Canada). Samples. Commercially available pork liver was used for the determination of recovery efficiencies and method detection limits (MDLs) of the PFOS isomers. Herring gull egg samples were collected in 1989 from Laurentian Great Lakes colonies. A large number of the 1989 collected eggs were pooled to produce a reference pool sample that is routinely used at NWRC as an inhouse reference material for bioaccumulative POPs (and recently including PFCs) in the eggs of birds.18 Another in-house reference material at NWRC is pooled double-crested cormorant (Phalacrocorax auritus) (DCCO) egg sample, which was prepared from large scale egg collections made in 2003 and taken from Laurentian Great Lakes colonies. Composed of a sample set of male and female adults, polar bear liver samples were collected in 2007-2008 from the Canadian Arctic (Nunavut), and the plasma samples were collected in 2007 from bears in Svalbard, Norway. These polar bear samples have been used in circumpolar trend studies on PFCs, including PFOS, as well as other chlorinated and brominated POPs in polar bears (unpublished data, Letcher et al.). All samples were well homogenized and stored under -80 °C until isomeric PFOS analysis. Sample Extraction and Cleanup. The sample preparation process included extraction of target compounds from the matrix and cleanup using SPE, which can also be used for determination of other PFCs compounds by LC/MS/MS, and is briefly described as follows. An amount of 0.1 to 1 g of sample was weighed into a polypropylene centrifuge tube, and 100 µL of a solution of sodium perfluoro-1-[1,2,3,4-13C4]octane sulfonate (MPFOS) at a concentration of 100 ng/mL was spiked into the sample. The sample was extracted with 3 mL of 10 mM KOH acetonitrile/water (80:20, v/v) solution after homogenizing for 1 min. After centrifugation the supernatant was transferred to a new tube. The extraction process was repeated 3 times and the supernatants were combined. A volume of 2 mL of the combined supernatant was then transferred into another tube, diluted with 8 mL water, and adjusted to pH ) 4 with 2% aqueous formic acid solution. The enrichment and cleanup of the PFCs in the extract were performed using Oasis WAX cartridges (3 mL, 60 mg, 30 µm). The cartridges were preconditioned by passage of 3 mL of methanol, followed by 3 mL of water with flow rate about 2 mL/ min. The sample was then loaded onto the cartridge, and the cartridge was then washed with 1 mL of 2% aqueous formic acid, and then with 2 × 1 mL of water and 2 × 1 mL of methanol. The target compounds (included other perfluorinated sulfonates and perfluorinated carboxylates) were then eluted from the cartridge with 2 × 1 mL solution of ammonium hydroxide (28-30%) in methanol (1/9 v/v). Sample Preparation for in-Port Derivatization. An amount of 0.5 g of TBAH and 5 mL of diethyl ether was added into a (18) Gauthier, L. T.; Potter, D.; Hebert, C. E.; Letcher, R. J. Environ. Sci. Technol. 2009, 43, 312–317.
4258
Analytical Chemistry, Vol. 81, No. 11, June 1, 2009
centrifuge test tube and mixed by vortex mixer for 1 min, and sonicated for half an hour to ensure dissolution of the TBAH. After centrifugation the upper phase (TBAH in diethyl ether) was used as the in-port derivatization reagent solution. The eluant from the SPE cleanup process was evaporated to dryness using a nitrogen evaporator, and the residue was dissolved in 100 µL of water. A volume of 100 µL of TBAH diethyl ether solution was added. The tube was capped, and the solution was well mixed by vortex mixer. Then 1 mL of diethyl ether was added and mixed again. After it was centrifuged and frozen at -20 °C, the diethyl ether phase was separated from aqueous phase (ice) and transferred to another centrifuge tube. The sample solution was evaporated to dryness, reconstituted in 100 µL diethyl ether, and transferred to a vial with 100 µL insert for GC/MS determination. GC/MS Determination. In-port derivatization, gas chromatography-mass spectrometric analysis was performed using an Agilent 6890 gas chromatograph (GC) coupled to a Agilent-5973N quadrupole mass spectrometer (MS) detector (Mississauga, ON, Canada). The GC column was a 30 m × 0.25 mm i.d. DB-5 capillary column with a film thickness of 0.25 µm. Helium was used as carrier gas at a constant flow of 1 mL/min. Five microliters of sample solution were injected into GC/MS by a 25 µL syringe with splitless mode, and a double taper deactivated splitless inlet linear was used. The injector temperature was 300 °C, the purge time was 1 min after injection, and the vent flow was 50 mL/ min. The oven temperature was programmed as follows: 50 °C, hold for 2 min, then at 5 °C/min to 120 °C. Ionization was performed in electron capture negative ionization (ECNI) mode using methane as reagent gas. The transfer line, source, and quadrupole temperatures are 280, 180, and 150 °C, respectively. It should be mentioned that although all target compounds could be eluted from the column during this process, this was not true for the residual matrix in the samples. Therefore, just after the sample run was over, 12.5 µL solution of TBAH in diethyl ether was injected with split mode (split ratio:100/1) and oven temperature held at 300 °C for ten min, and the solvent delay time was set at 9.95 min to avoid contaminating the MS source. This injection also reduced any sample-to-sample carry-over. Individual branched PFOS isomer standards were prepared and injected into GC/MS operating in the full scan mode (m/z 50-650). The relative abundances of molecular ions and the fragment ions of the derivatized PFOS isomers that were obtained from full scan MS with an ECNI source are listed in Table 2. To increase the sensitivity, selected ion-monitoring mode (SIM) was used for quantitative analysis, and the ions of m/z 137, 480, 499, and 503 (for the MPFOS internal standard) were monitored. Identification of PFOS isomers in technical PFOS product (TPFOS) was accomplished by comparing the relative abundance of monitored ions and retention time with corresponding standards. Quantitative determination of the individual PFOS isomers in this technical PFOS product solution was performed using an isotope dilution method with a five point calibration curve of individual PFOS isomers standard solutions, in which the concentrations spanned the range of anticipated analyte concentrations in the technical PFOS product solution. Since the concentration of individual isomers was measured in the technical PFOS solution (Table 3), for the environmental samples, identification and quantitative determination of individual PFOS isomers were
Table 2. Relative Retention Times (RRTs) and Relative Abundances of the Fragments Ions of Derivatized PFOS Isomers Obtained from Full Scan MS with an ECNI Sourcea relative abundance of ions (m/z amu)
a
no.
compound
RRT
556
499
480
400
269
137
1 2 3 4 5 6 7 8 9 10 11
P1MHpS P2MHpS P3MHpS P35DMHxS P4MHpS L-PFOS P5MHpS P6MHpS P45DMHxS P55DMHxS P44DMHxS
0.927 0.958 0.978 0.986 0.992 1.000 1.018 1.023 1.035 1.054 1.059
0 0 10 6 11 0 2 7 0 0 0
6 6 6 1 8 37 6 6 3 26 16
2 0 100 100 100 11 100 100 100 4 2
100 78 28 9 14 27 7 10 4 0 0
0 0 0 0 1 0 0 0 2 0 100
67 100 44 35 13 100 21 19 11 100 94
The ions with corresponding percent abundances in bold were used for quantitative analysis purposes.
Table 3. Mean Percentage (( %RSD) of Branched PFOS Isomers in Technical PFOS (T-PFOS) Standard Solution from Wellington Lab and in Selected Biota Samples no.
1 2 3 4 5 6 7 8 9 10 11 ∑PFOS
sample
P1MHpS P2MHpS P3MHpS P35DMHxS P4MHpS L-PFOS P5MHpS P6MHpS P45DMHxS P55DMHxS P44DMHxS (ng/g w.w.)
T-PFOS
0.85 ± 0.08 1.13 ± 1.13 7.21 ± 7.21 0.59 ± 0.04 5.57 ± 0.25 65.04 ± 1.76 6.94 ± 0.33 11.33 ± 0.56 0.69 ± 0.03 0.51 ± 0.06 0.13 ± 0.02
herring gull egg
polar bear liver
polar bear plasma
DCCO egg
Great Lakes, Canada (pooled, n ) 6) 0.23 ± 0.03 0.13 ± 0.02 0.60 ± 0.13
Nunavut, Canada (n ) 8) 0.3 ± 0.1 0.4 ± 0.1 0.8 ± 0.1
