Evaluation of Perfluorooctane Surfactants in a Wastewater Treatment

Influent, effluent, and river water at the point of discharge for a 6-MGD wastewater treatment plant (WWTP) were screened for eight perfluorooctane su...
0 downloads 7 Views 129KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 5524-5530

Evaluation of Perfluorooctane Surfactants in a Wastewater Treatment System and in a Commercial Surface Protection Product B R Y A N B O U L A N G E R , †,‡ J O H N D . V A R G O , § JERALD L. SCHNOOR,† AND K E R I C . H O R N B U C K L E * ,† Department of Civil and Environmental Engineering, University of Iowa, SC 4105, Iowa City, Iowa 52240, and University Hygienic Laboratory, University of Iowa, 102 Oakdale Campus, Iowa City, Iowa 52242

The origin and amount of perfluorooctane surfactants in wastewater treatment systems, and the transformation these compounds may undergo during treatment, were evaluated through measurement and experiment. Influent, effluent, and river water at the point of discharge for a 6-MGD wastewater treatment plant (WWTP) were screened for eight perfluorooctane surfactants. N-EtFOSAA was quantified in influent (5.1 ( 0.8 ng/L), effluent (3.6 ( 0.2 ng/ L), and river water samples (1.2 ( 0.3 ng/L), while PFOS and PFOA were quantified in effluent (26 ( 2.0 and 22 ( 2.1 ng/L, respectively) and river water (23 ( 1.5 and 8.7 ( 0.8 ng/L, respectively). Signals for PFOS and PFOA were observed in influent samples, but exact quantitative determination could not be made due to low recoveries of these two compounds in field spike samples. Although the source of PFOS and PFOA observed in WWTP effluents is not clear, two hypotheses were examined: (1) the highly substituted perfluorooctane surfactants that constitute commercially available fabric protectors can be transformed to PFOS and PFOA during biological treatment in wastewater treatment systems, and (2) the end products themselves are directly introduced to WWTPs because they are present as residual in the commercial mixtures. Biotransformation experiments of 96 h were run to determine whether N-EtFOSE (a primary monomer used in 3M’s polymer surface protection products) could be transformed to lesser-substituted perfluorooctane compounds in bioreactors amended with aerobic and anaerobic sludge from the sampled plant. At the end of the aerobic biotransformation experiment, N-EtFOSAA and PFOSulfinate were the only two metabolites formed and each accounted for 23 ( 5.0% and 5.3 ( 0.8% of the transformed parent on a molar basis, respectively. Transformation of N-EtFOSE was not observed under anaerobic conditions. A sample of a commercially available surface protection product from 1994 was analyzed and found to contain six of the * Corresponding author e-mail: [email protected]; phone: 319-384-0789; fax: 319-335-566. † Department of Civil and Environmental Engineering. ‡ Current Address: U. S. Environmental Protection Agency, National Risk Management Research Laboratory, MS421, 26 W. Martin Luther King Drive, Cincinnati, OH 45268. § University Hygienic Laboratory. 5524

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

targeted perfluorinated surfactants, including PFOS and PFOA. These findings suggest transformation of precursors within wastewater treatment is not an important source of these compounds compared to direct use and disposal of products containing the end products as residual amounts.

Introduction Commercial and industrial usage of fluorinated surfactants has led to the observed accumulation of perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) across environmental matrixes. These surfactants are environmentally ubiquitous and have been reported in air, water, and biota samples worldwide (1-12). PFOS and PFOA are the end metabolic products of the two most common perfluorooctane surfactant classes: perfluorooctane sulfonamides and fluorotelomer alcohols (13, 14). Despite their global environmental presence, little is understood concerning the major environmental pathways of these compounds. Wastewater treatment plants are a major source of perfluorooctane surfactants to natural waters. In a previous paper, we reported a mass budget of eight perfluorooctane surfactants in Lake Ontario. Wastewater treatment plant (WWTP) effluent discharge was determined to be the largest source of these surfactants to this Great Lake (15). Our estimate of inputs from WWTP drew on a 1999 3M (St. Paul, MN) study of perfluorooctane surfactant concentrations in wastewater treatment effluent and wastewater treatment biosolids. 3M reported concentrations of PFOS and PFOA in the effluents of six WWTPs (16). Mean ( standard deviation concentrations of PFOS (1140 ( 1823 ng/L) and PFOA (549 ( 840 ng/L) were highly variable in the six plants. Examination of dried aerobic sludge at the WWTPs also showed variable, but detectable, levels of PFOS in all but one plant, while PFOA was present in the sludge of four plants (16). The origin of PFOS, PFOA, and related compounds in wastewater treatment is unclear. We have previously hypothesized that this compound class is introduced into WWTPs through the cleaning and care of surface treated products (from clothing to carpets), their use in industrial processes, and from the treatment of wastewater influent containing discarded product. Surprisingly, the presence of PFOS, PFOA, and other perfluorooctane surfactants in any consumer products has not been evaluated in the peer review literature. 3M, however, did perform preliminary experiments to determine if higher-substituted perfluorooctane surfactants could undergo biological transformation during aerobic wastewater treatment. These tests showed that biotransformation of higher-substituted perfluorooctane surfactants, such as N-ethyl perfluorooctanesulfonamido ethanol (NEtFOSE), to PFOS occurred within 35 days. Because PFOS is known not to further degrade (17), these results indicated that biological transformation within WWTPs could potentially lead to the environmental release of PFOS from transformed precursor compounds. However, the duration of the 3M test exceeds the hydraulic and solids retention time of normal WWTPs and only evaluates aerobic pathways, making the results difficult to extrapolate to typical WWTP operations. The study reported here expands upon 3M’s early efforts by evaluating the concentrations of PFOS, PFOA, and six other perfluorooctane surfactants in effluent discharge from a wastewater treatment plant servicing an area with no known manufacture or industrial application of perfluorooctane surfactants. Aerobic and anaerobic sludge systems from the 10.1021/es050213u CCC: $30.25

 2005 American Chemical Society Published on Web 06/30/2005

sampled plant are also evaluated for their ability to transform N-EtFOSE, the most highly substituted perfluorooctane sulfonamide present in commercial fabric treatment products, to lesser substituted metabolites within a time period relevant to wastewater treatment systems. Finally, a sample of a commercial fabric protector from a 1994 can of Scotchgard was analyzed and the amount of PFOS, PFOA, and six other perfluorooctane surfactants present as residual compound in the polymeric product was quantified.

Materials and Methods Chemicals and Standards. Standards of 2(N-ethyl-perfluorooctanesulfonamido) ethyl alcohol (N-EtFOSE, chemical purity 97.7%), 2-(N-ethyl-perfluorooctanesulfonamido) acetic acid (N-EtFOSAA, 98.6%), perfluorooctanesulfonylethylamide (N-EtFOSA, 99.3%), 2-(perfluorooctanesulfonamido) acetic acid (PFOSAA, 99.6%), perfluorooctane sulfonylamide (FOSA, 98.94%), perfluorooctane sulfonate potassium salt (PFOS, 86.9%), perfluorooctane sulfinate potassium salt (PFOSulfinate, 97.2%), and perfluorooctanoate postassium salt (PFOA, 95%) were provided by 3M. Structures of each compound discussed in the study and information on all chemicals used for the analysis of perfluorooctane surfactants in wastewater, the commercial product, and for the biotransformation experiments can be found in the Supporting Information. Wastewater Analysis. Sampling. Influent and effluent water samples were collected in triplicate on July 14, 2004 from the South WWTP located in Iowa City, IA. Triplicate river water samples just downstream of the effluent discharge were collected on the same day. The sampled plant operates at 6 million gallons per day (MGD) and receives domestic and industrial wastewater. A description of the sampled plant’s treatment processes is presented in the Supporting Information. The plant was selected for sampling because of its close proximity to the University of Iowa and lack of known perfluorooctane surfactants manufacture and production within its service area. In the United States approximately 2600 wastewater treatment facilities operate at a flow rate between 1.0 and 10.0 MGD and approximately 13 800 facilities incorporate secondary treatment processes, making the sampled plant representative of a large number of facilities in terms of design flow and treatment capabilities (18). Six 1-L grab samples of influent, effluent, and river water immediately downstream from the outfall were collected in methanol-rinsed and air-dried polycarbonate narrow-mouth bottles. Grab samples were collected at the river and influent locations by attaching the 1-L polycarbonate bottle to an aluminum pole. Influent samples were acquired prior to screening. Effluent samples were taken from a collection port in the plant post dechlorination. Three bottles from each station were immediately spiked to 100 ng/mL for each analyte measured in the study. The bottles were capped and carried to the lab for processing on the day of collection. Sample Extraction. All wastewater samples were extracted using solid-phase extraction (SPE) cartridges (Waters Oasis HLB, 1-g capacity) using previously described methods (3) that are summarized in detail in the Supporting Information. For influent and spiked influent samples 0.3 L of sample was processed. For effluent and river water samples and their spiked counterparts, 1.0 L was processed through the SPE cartridges. Analysis. A Micromass Quattro tandem LC/MS/MS system equipped with an electrospray interface operated in negative ion-monitoring mode was used to analyze the environmental sample extracts. Samples were introduced into the LC/MS/ MS system using a Waters Alliance 2695 autosampler/gradient pump system operated in a reversed-phase separation mode. A 10-µL aliquot of extract was injected onto a 3.0 × 150 mm (5 µm) Zorbax SB C8 column. The mobile phase consisted of (A) 50% (v/v) MeOH/ACN, 0.15% acetic acid and (B) 0.15%

(v/v) acetic acid/HPLC grade water. The flow rate was set at 0.6 mL/min through the column and was split postcolumn to send ∼0.2 mL/min to the electrospray interface. The separation proceeded from initial conditions of 15% A to 100% A at 10 min with a linear gradient, held at 100% A until 13 min, dropped to 15% A at 13.5 min, and held at 15% A until 19.1 min. The column temperature was held constant at 30 °C. The MS/MS parameters were optimized for the transmission of the deprotonated molecular ion, with the exception of N-EtFOSE which was monitored as the acetate ion adduct, through the first quadrupole set and its most abundant product ion fragments subsequently formed in the collision cell. The capillary voltage was set at 3.5 kV. The cone and collision cell voltages were optimized for each monitored ion pair. Quantification of analytes was accomplished through multiple reaction monitoring (MRM) of the following ion pair transitions: 499 f 80 for PFOS, 630 f 59 for N-EtFOSE, 584 f 419 for N-EtFOSAA, 556 f 498 for PFOSAA, 526 f 169 for N-EtFOSA, 498 f 78 for FOSA, 483 f 419 for PFOSulfinate, and 413 f 369 for PFOA. External calibration was used to quantify analyte values for LC/MS/MS analyses using multicomponent standards. Each multicomponent standard contained every analyte at the same concentration. The calibration curve ranged from a low concentration of 0.5 ng/mL or 1 ng/mL (depending upon sensitivity of the individual analyte) to a high concentration of 250 ng/mL. Quadratic curve fits were used for all analytes. Sludge Microcosm Studies. One liter of aerobic fresh liquor suspended solids was taken from a sampling point inside the plant’s mixed liquor solids aeration basin. One liter of mesophilic anaerobic sludge solids (approximately 10 days old) was collected from a sampling port. Sludge samples were collected in Pyrex glass bottles, capped, and transported to the lab. Both aerobic and anaerobic solids from the treatment plant were allowed to settle for 24 h at room temperature before use in the microcosm studies. Microcosm Experimental Setup. Eleven reactor treatments were created in triplicate at the start of the experiment. These reactor types included aerobic and anaerobic: bioreactors, abiotic controls, media controls, sterilized controls, and sludge blanks (Table 1). Triplicate positive anaerobic degradation controls were also created. All aerobic reactor samples were prepared on the benchtop in 125-mL Nalgene polycarbonate Erlenmeyer culture flasks. Anaerobic samples were prepared in 125-mL high-density polyethylene (HDPE) bottles within an anaerobic glovebox. Prior to removing anaerobic samples from the glovebox, the HDPE bottles were sealed with a rubber stopper that was then crimped. Aerobic and anaerobic bioreactors received 57 mL of designated media (composition of aerobic and anaerobic media used in this study is described in detail in the Supporting Information), contained 3 mL of sludge solids, and were spiked with 50 µL of concentrated N-EtFOSE stock solution to achieve a 1.50 mg/L N-EtFOSE concentration in each bioreactor. Abiotic controls contained 60 mL of aerobic media and received the 50 µL addition of concentrated N-EtFOSE stock solution to achieve a 1.50 mg/L N-EtFOSE concentration in the abiotic controls. Media controls contained only media. Sludge blanks contained 57 mL of designated media and 3 mL of solids. Sterilized controls were prepared with 57 mL of designated media and 3 mL of solids and then were autoclaved. Once removed from the autoclave, both 50 µL of concentrated N-EtFOSE stock solution (to achieve a concentration of 1.50 mg/L N-EtFOSE in the sterilized control) and mercuric chloride (0.2%) were spiked into the reactors. Prepared anaerobic bioreactors were spiked with nitrobenzene to a final concentration of 5 mg/L in the sample to serve as positive anaerobic degradation controls. NitrobenVOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5525

TABLE 1. Batch Reactor Treatments Used in the Laboratory N-EtFOSE Biotransformation Experimentsa additions

Aerobic Reactors ANF AAC

sample size aerobic media aerobic sludge solids N-EtFOSE spike 0.2% mercuric chloride autoclaved

3 x x x

AMC

ASC

ASB

3 x

3 x x x x x

3 x x

NSC

NSB

NPC

3 x

3 x x

3 x x

3 x x

Anaerobic Reactors NNF NAC NMC sample size anaerobic media anaerobic sludge solids N-EtFOSE spike 0.2% mercuric chloride autoclaved nitrobenzene spike

3 x x x

3 x x

3 x

x x x

x

Triplicate reactors (n ) 3) were incubated at 25 °C and 120 rpm for 96 h. Abbreviations: ANF, aerobic non-fed reactors; NNF, anaerobic non-fed reactors; AAC, aerobic abiotic control; NAC, anaerobic abiotic control; AMC, aerobic media control; NMC, anaerobic media control; ASC, aerobic sterile control; NSC, anaerobic sterile control; ASB, aerobic sludge blank; NSB, anaerobic sludge blank; x, analyzed as described in the text; NPC, anaerobic positive degradation control. a

zene was selected because it degrades readily to aniline under anaerobic conditions but will not transform under aerobic conditions. All media and spike solutions were filter-sterilized prior to use. Oxygen levels in the aerobic bioreactors were not monitored and the solids concentration of the 3 mL of settled biosolids added to the reactors was not measured during the experiment. Microcosm Sampling. All reactors were initially sampled at time 0.5 h. Immediately following the initial sample, the reactors were placed in the dark on a shaking incubator set at 120 revolutions per minute (rpm) and 25 °C. Aerobic bioreactors were sampled at 6, 12, 18, 24, 36, 48, 72, and 96 h. Anaerobic bioreactors were sampled at 24, 48, and 96 h. Aerobic and anaerobic controls were sampled at 0.5, 48, and 96 h. For sampling, aerobic reactors were thoroughly mixed using a vortex mixer to resuspend settled solids and were sampled by taking a 2 mL aliquot from the reactor using sterilized polypropylene pipet tips at each sampling time from each reactor type. The anaerobic reactors were mixed and then sampled in two aliquots (2 mL for bioreactor samples and 1 mL for the positive degradation control) using a polyethylene syringe with a 20-gage needle to puncture the rubber stopper. Positive anaerobic degradation controls were sampled and analyzed initially and at the conclusion of the experiment. Sludge Microcosm Extraction. Extraction of microcosm aliquots followed the procedure described for the wastewater samples except Alltech SEP-VAC C18 6 cm3 SPE cartridges were used to extract all samples. A 5-mL BD syringe was placed on top of the SPE cartridge to hold the wash solutions, the sample aliquot, and the eluent solutions during the extraction. Matrix spikes of the extracts were prepared by spiking 2 mL of the extract with a multicomponent standard such that the final concentration of the added spike was 100 ng/mL. Analysis of Perfluorooctane Compounds. Because the microcosm studies were run at high concentrations of analyte (samples spiked to an initial concentration of 1.50 mg/L), a dedicated instrument was used to do all the analyses for these experiments. This instrument has always been used only to perform highly concentrated biotransformation studies. All analytes were separated and quantified using an 5526

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

Agilent 1100 series HPLC/MSD system using previously described methods (3). Six multicomponent calibration standards containing all analytes (10, 50, 100, 500, 1000, and 2500 ng/mL) were used to form the external calibration curve for N-EtFOSE. Only four of these calibration standards, 10 to 500 ng/mL, were used to quantify the metabolites. Quadratic curve fits were used for all analytes. Nitrobenzene Analysis. Nitrobenzene was analyzed by HPLC with UV detection. A summary of the method is presented in the Supporting Information. Three calibration standards (1, 5, and 10 mg/L) were used to form the calibration curve for nitrobenzene. Solid Separation Experiment. To evaluate partitioning of compounds to the solids in the sludge microcosms, triplicate aerobic sterilized controls were spiked with 250 ng of each analyte and allowed to incubate. After one week the samples were centrifuged for 10 minutes at 7500 rpm. The supernatant was placed on a washed SPE and eluted in 5 mL of methanol. An aliquot was analyzed. The separated solids were resuspended in 2 mL of the appropriate media using a vortex mixer and placed on the SPE cartridge. The cartridge was then eluted with 5 mL of methanol and an aliquot was analyzed. SPE Analyte Retention Experiment. Because the reactors were only spiked with N-EtFOSE, a 50-µL spike of concentrated stock solution containing 250 ng of each target analyte was added to prepared aerobic and anaerobic microcosms. The microcosms were immediately extracted to determine the recovery efficiency of each compound on the C18 SPE cartridge. The C18 cartridges were eluted with 5 mL of methanol and analyzed using the tandem quadrupole LC/ MS/MS system. Analysis of Surface Protection Product. A 0.832-g sample of Scotchgard from 1994 was sprayed into a dry 30-mL HDPE bottle and dissolved into 20 mL of methanol. An aliquot of this solution mixture was analyzed using the previously reported LC/MS method (3). To confirm our single quad findings, a 10:1 dilution of the solution mixture was analyzed using the LC/MS/MS method described earlier.

Results Wastewater Analysis. Quality Control. Background contamination was evaluated by analyzing HPLC water, conditioned SPE cartridges, and methanol-soaked sample bottles. All studies were completed in triplicate. PFOS (6.2 ( 2.0 ng/ L) and PFOA (2.4 ( 1.6 ng/L) were detected when 1 l of HPLC-grade water was extracted and analyzed. None of the other analytes were detected in the HPLC-grade water blank. SPE blanks, consisting of Oasis HLB SPE cartridges that were prerinsed and directly eluted with 5 mL of methanol without running any water through, were also analyzed for potential contaminants. No analytes were found in the SPE blank extracts. Background contamination from our sample bottles was evaluated by adding 100 mL of methanol to the polycarbonate labware vessels used in the study and allowed to sit overnight. The 100-mL sample was then evaporated down to dryness, reconstituted in 300 µL of methanol, and analyzed. No analyte signals were observed. Therefore, the source of PFOS and PFOA in the blank was from the HPLC grade water itself (Alfa Aesar Lot #K06N23) and was not a function of the method. Because PFOS and PFOA concentrations in the wastewater and river samples were at least 3.5 times greater than the HPLC-grade water blank concentrations, quantification of the sample concentrations for these two compounds was not affected by the background. The top section of Table 2 reports field spiked sample recoveries for each compound measured in the study. Analyte concentrations were reported only if their field spike recoveries were greater than 70% and if the observed analyte concentration was at least 3.5 times the concentration

TABLE 2. Mean Field Spike Recoveries and Concentrations (Mean ( Standard Deviation) for WWTP Influent Samples, WWTP Effluent Samples, and River Water Samples at the Point of Dischargea

blankb influent effluent river

blankb influent effluent river

PFOS

N-EtFOSE

79 34 74 76

68 10 62 56

PFOS

N-EtFOSE

6.2 ( 2.0 [>400]d 26 ( 2.0 23 ( 1.5

ndc nd nd nd

Field Spike Recovery (%) N-EtFOSAA PFOSAA N-EtFOSA 74 82 70 70

64 17 65 66

66 17 57 56

Sample Concentration (ng/L) N-EtFOSAA PFOSAA N-EtFOSA nd 5.1 ( 0.8 3.6 ( 0.2 1.2 ( 0.3

nd nd nd nd

nd nd nd nd

FOSA

PFOSulfinate

PFOA

71 2 66 70

80 14 76 79

86 16 80 79

FOSA

PFOSulfinate

PFOA

nd nd nd nd

nd nd nd nd

2.4 ( 1.6 [>4]d 22 ( 2.1 8.7 ( 0.8

a Reported concentration values have sample to blank mass ratios greater than 3.5, a field spike recovery greater than 70%, and represent triplicate samples unless otherwise noted. b Blank consists of 1 L of HPLC-grade water. c nd ) not detected in sample. d Values in brackets represent estimated concentrations corresponding to observed signals in the influent samples. However, the given values had less than a 40% field spike recovery and exhibited strong signal suppression.

observed in the HPLC water blank. Field spike recoveries for effluent and river samples were above 70% for PFOS, N-EtFOSAA, PFOSulfinate, FOSA, and PFOA. Field spike recoveries for influent samples were 35% or less for all compounds except N-EtFOSAA, whose recovery was 82%. To test for matrix suppression of the analyte signals, known amounts of each analyte were added to the final influent extract via a standard addition just prior to analysis. Analyte signal suppression was observed for all analytes in the influent extracts except for N-EtFOSAA. Standard addition spikes were also made to the final extracts of effluent and river water samples. No enhancement or suppression of the analyte signals was observed. Duplicate analysis of individual extracts and calibration check standards was run every fifteenth and twentieth sample, respectively. Instrument reproducibility was within 5% for every duplicate and calibration check run. Reported concentrations have not been corrected. Effluent Water, River Water, and Influent Water Concentrations. Concentrations are presented in the bottom part of Table 2. Effluent concentrations for PFOS and PFOA were 26 ( 2.0 and 22 ( 2.1 ng/L, respectively. PFOS and PFOA effluent concentrations at the Iowa City WWTP are 1/2 to 1/3 of reported concentrations of PFOS and PFOA at two “control locations” sampled in the 1999 3M study (51 ( 9 and 60 ( 22 ng/L respectively). The “control locations” sampled by 3M were both also located in service areas with no known source production facilities. Mean concentrations of effluents from three of 3M’s “test locations”scities with known sources of perfluorooctane surfactant productionswere 22 times higher for PFOS (586 ( 243 ng/L) and 14 times higher for PFOA (298 ( 284 ng/L) compared to the Iowa City WWTP (16). The highest reported effluent concentrations in the 3M study were reported from Decatur Alabama’s WWTP and were 192 times higher for PFOS (4980 ( 438 ng/L) and 104 times higher for PFOA (2280 ( 198) than the Iowa City WWTP effluent concentrations (16). The high variability of concentrations between the six 3M sample locations and our location highlights how important knowing an individual plant’s effluent concentration is to determining the flux of perfluorooctane surfactants originating from the plant when its effluent is discharged. Reducing loadings from select plants that have significant discharge of perfluorooctane surfactants may help to bring down environmental concentrations. PFOA concentrations in river samples (8.7 ( 0.8 ng/L) were lower than PFOA effluent concentrations (22 ( 2.1 ng/ L), but PFOS river concentrations (23 ( 1.5 ng/L) were not (26 ( 2.0 ng/L in effluent). Because upstream sources of PFOA and PFOS are unknown, we are unable to conclude

what this difference in relative concentrations of PFOS and PFOA means. The concentrations reported are similar to previously reported concentrations for river water (9, 19) potentially indicating a high background concentration of these compounds in river water from upstream sources. There are 293 registered wastewater discharge points in the Iowa River watershed upstream of our studied plant, representing about 200 000 people (20). On the basis of our findings for the Iowa City plant, we would expect that the upstream plants are also contributing to the PFOS and PFOA in the river. PFOS and PFOA were detected in influent WWTP samples and exhibited high concentrations (estimated minimum concentrations exceeded 400 ng/L and 4 ng/L, respectively). However, analytical uncertainties did not allow for exact quantitative determination in this matrix using our method. PFOS field spike recoveries were only 34% and spiked additions to the extracts showed strong signal suppression for PFOS (>50% suppression). Signal suppression of greater than 80% was observed for PFOA during analysis of the influent sample extracts and the field spike recovery was 16%. To reduce the signal suppression observed for PFOS and PFOA, extracts were filtered through a 0.45-µm syringe filter prior to analysis. This additional filter step eliminated signal suppression effects for all analytes from matrix spiked samples, but also retained PFOS. Efforts to improve recovery of influent field spikes by adding additional elution steps during solid-phase extraction were also unsuccessful. Although exact influent concentrations of PFOS and PFOA are not reportable due to poor recovery of the compounds from the field spikes using our method, these compounds are present at elevated concentrations in influent. N-EtFOSAA was the only compound that could be quantified in all three stages of treatment. No signal suppression was observed for N-EtFOSAA in any of the samples and N-EtFOSAA field spike recovery was always above 70%. N-EtFOSAA concentrations decreased during treatment, starting at a mean concentration of 5.1 ( 0.8 ng/L in influent to 3.6 ( 0.2 ng/L in effluent. This is an overall WWTP removal efficiency of 29%. N-EtFOSE, PFOSAA, N-EtFOSA, FOSA, and PFOSulfinate were not detected in any of the WWTP system samples, although their presence in influent samples cannot be ruled out due to their low recovery by this method. Biotransformation Experiments. Analyte Detection Limits, Recoveries, and Matrix Spikes. The instrument detection limit (IDL) was established by running seven replicate samples of the lowest possible concentration resulting in a signal-to-noise ratio of between 5 and 20 (21) (Table 3). The signal-to-noise ratio was defined as the average concentration VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5527

TABLE 3. Limit of Quantification (LOQ) and Mean Recovery of Individual Compounds Measured in the Biotransformation Study

compound PFOA PFOSulfinate FOSA PFOS N-EtFOSA PFOSAA N-EtFOSAA N-EtFOSE

LOQ (pg/mL in extracts)

Mean Recoverya (%) bulk sample solids only aqueous only t ) 0 days t ) 7 days t ) 7 days

6 2 2 3 2 8 6 2

87 91 95 91 85 88 91 81

ndb nd nd nd nd nd nd nd

82 86 89 89 88 93 92 72

a Recovery is for triplicate aerobic and triplicate anaerobic reactors (n ) 6). b nd ) not detected in sample.

TABLE 4. Mean ( Standard Deviation of N-EtFOSE (mg/L) in Each Reactor Type (n ) 3 per type) at Each Sampling Timea,b Sample Time (hrs) 0.5

48

96

media control sludge blank media spike sterile control spike

aerobic controls ndc nd nd nd 1.16 ( 0.25 1.20 ( 0.17 1.34 ( 0.21 1.27 ( 0.07

nd nd 1.18 ( 0.16 1.24 ( 0.11

media control sludge blank media spike sterile control spike

anaerobic controls nd nd nd nd 1.42 ( 0.12 1.43 ( 0.28 1.39 ( 0.14 1.55 ( 0.19

nd nd 1.51 ( 0.16 1.63 ( 0.23

FIGURE 1. Standardized anaerobic bioreactor N-EtFOSE concentration measured over the course of the biotransformation experiment. Bioreactors were spiked to 1.50 mg/L N-EtFOSE at time 0. Error bars represent one standard deviation about the mean of triplicate samples for each time point. Formation of metabolites was not observed under anaerobic conditions.

a Note: PFOA, PFOS, PFOSulfinate, N-EtFOSAA, N-EtFOSA, PFOSAA, and FOSA were not detected in any control sample. a Initial amount spiked into media and sterile control spike samples at time ) 0 was 1.5 ( 0.05 mg/L. c nd ) not detected.

of 7 replicate analyzed samples divided by the standard deviation of the concentration of the 7 replicate analyzed samples. The IDL was defined as 3.143 times the standard deviation of the measured replicates. For aerobic and anaerobic media controls and sludge blanks, none of the compounds measured in the study were detected above the IDL. Therefore, the IDL was used as the experiment limit of quantification (LOQ). All analytes measured in the biodegradation studies were easily recovered from the bioreactor system (Table 3). Immediate analyte recovery from prepared bioreactors was above 80% for all analytes. After one week, recovery for all analytes from the spiked aerobic sterile controls was above 70%. Recovery of N-EtFOSE spiked into aerobic and anaerobic media controls throughout the duration of the experiment exceeded 77% for all bioreactors (Table 4). Aerobic and anaerobic sterilized control N-EtFOSE spiked sample recoveries were above 83% for the entire experiment. All nonspiked controls did not have detectable quantities of any of the measured analytes. Spiked additions were used to identify matrix effects on the ion signals. Matrix spiked additions of the biotransformation reactor extracts did not display signal enhancement or suppression for any of the measured compounds. Anaerobic Transformation. Under anaerobic conditions, N-EtFOSE did not transform to lesser-substituted perfluorooctane compounds (Figure 1) and formation of metabolites during the experiment was not observed. Nitrobenzene was detected in positive controls at the beginning of the experiment, but was not detected in any of the samples at the end of the experiment. Therefore, the lack of N-EtFOSE degradation in the anaerobic reactors was not a function of killed or dysfunctional microbial bioreactor populations. 5528

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

FIGURE 2. Concentrations of parent compound and metabolites in aerobic bioreactors. (a) Standardized aerobic bioreactor N-EtFOSE concentration measured over the course of the biotransformation experiment. Bioreactors were spiked to 1.50 mg/L N-EtFOSE at time 0. (b) Metabolites formed from the biotransformation of N-EtFOSE during the experiment. Error bars represent one standard deviation about the mean of triplicate samples for each time point. Aerobic Transformation. Aerobic transformation of NEtFOSE resulted in the formation of two primary metabolites: N-EtFOSAA and PFOSulfinate (Figures 2 and 3). After 96 h 40 ( 5.9% of the parent remained. N-EtFOSAA and PFOSulfinate each accounted for 23 ( 5.0% and 5.3 ( 0.8% of the transformed parent on a molar basis, respectively. The remaining 32% of transformed parent was not identified and was likely transformed to other lesser-substituted compounds not analyzed in this study. In the 3M degradation study, N-EtFOSAA and PFOSAA were the primary metabolites

FIGURE 3. Structural relationship of primary metabolites formed during the 96-h aerobic biotransformation experiments. formed from the transformation of N-EtFOSE after 35 days, accounting for 33 ( 10 and 46 ( 11% of the initial parent, respectively (Supporting Information Figure S1) (16). However, significant formation of PFOSAA in their system did not appear until 144 h. After 65 h in their experiment (the closest sampling time point to our studies completion time), 45 ( 3.1% of their parent compound remained, and NEtFOSAA, PFOSAA, PFOSulfinate, FOSA and PFOS accounted for 22 ( 1.1, 4.5 ( 0.2, 2.7 ( 0.1, 1.0 ( 0.1, and 0.6 ( 0.1% of the transformed parent, respectively. These findings were in qualitative agreement with our results for N-EtFOSE, N-EtFOSAA, and PFOSulfinate after 96 h (Figure 4). However, PFOSAA, FOSA, and PFOS were not observed in our reactors. In both experiments only N-EtFOSAA and PFOSulfinate are formed during the first 24 h of reaction and no other metabolites were detected. N-EtFOSE is dealklyated to form N-EtFOSAA and either of these two higher-substituted compounds may be deaminated to form PFOSulfinate (Figure

3). The lack of observed intermediate metabolites between N-EtFOSAA and PFOSulfinate suggests that deamination is occurring. Partitioning to Solids during Wastewater Treatment. No target analytes were recovered from the microcosm sludge solids (Table 3). This result suggests that the hydraulic residence time of the WWTP will control the amount of time these compounds can react during treatment. Because of this time dependence, transformation of PFOS from NEtFOSE or N-EtFOSAA (the first metabolite formed) is unlikely to occur within WWTPs during the plant’s hydraulic retention time. Therefore, any presence of PFOS in effluent is from PFOS coming into the WWTP in influent as was observed, but were not able to quantify, in our measured influent samples. While a possibility exists that other compounds such as PFOSAA, N-EtFOSA, PFOSulfinate, or FOSA may be transformed to PFOS during a plant’s hydraulic retention time, these compounds were not observed in our analyzed WWTP effluent samples (Table 2) or during the biotransformation study. Analysis of Surface Protector Product. Direct release of perfluorinated surfactants to municipal wastewater may occur through production, use, and disposal of consumer products that contain the surfactants. The analysis of a 1994 Scotchgard sample shows that all of our target compounds, except for N-EtFOSAA and PFOSAA, are present in the product. The concentrations of the six compounds detected are low but easily detectable, suggesting that these compounds are residual monomers in the formula. The chromatograms for the 1994 Scotchgard sample analysis are included in the Supporting Information (Figures S2 and S3). The results of the LC/MS/MS quantitative analysis of the sample are summarized in Table 5. While Scotchgard is primarily composed of polymeric fluorinated surfactants,

FIGURE 4. Comparison of N-EtFOSE biotransformation experimental results from the 35-day 3M test (16) and this study. (a) N-EtFOSE concentration, (b) metabolite formation (3M), (c) metabolite formation (this study). VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5529

TABLE 5. Perfluorooctane Surfactants Compounds in a Scotchgard Sample Analyzed by Tandem Mass Spectrometry compound PFOA FOSA N-EtFOSE PFOSulfinate PFOS N-EtFOSA Σtotal

masscompound/massproduct [g/g]

% composition

1.34E-05 1.27E-06 6.33E-07 4.67E-07 2.21E-07 1.53E-07 1.62E-05

1.34E-03 1.27E-04 6.33E-05 4.67E-05 2.21E-05 1.53E-05 1.62E-03

all compounds monitored in the study were monomeric. Compound residual may be susceptible to wash-off at higher rates than the polymer applied product. While it is unknown what effect aging of the can had on the product, the can itself had not been previously used, suggesting that the contents may be representative of other cans produced prior to the phase out. When the mass contribution of each perfluorooctane surfactant measured in the sample was summed (n ) 8), PFOA, a compound usually not associated with Scotchgard, accounted for 82% of the total summed mass. FOSA, N-EtFOSE, PFOSulfinate, and PFOS accounted for 8, 4, 3, and 1% of the total summed mass, respectively. All other compounds were present at less than 1% of the total summed mass. While each compound present in the Scotchgard sample was present at less than 1/100th of a percent of the total product mass, the sheer volume of Scotchgard product globally produced could result in high amounts of monomer compound introduction to the environment. PFOS and PFOA were both identified in the influent, effluent, and river water samples analyzed in this study and have been identified in Great Lakes waters (3). N-EtFOSE itself was not found in the wastewater samples, but was shown to readily transform to N-EtFOSAA and PFOSulfinate during aerobic biological treatment. Both of these two metabolites, along with FOSA, were present in water samples in the Great Lakes (3). Presence of perfluorooctane surfactant compounds as monomers in the consumer product and in effluent supports the proposition that care and cleaning of treated surfaces is a source of these compounds to the WWTPs from down-the-drain disposal. Further development of a method to quantitate perfluorooctane surfactants in influent samples would add support to this conclusion. The findings reported in this paper suggest transformation of precursors within wastewater treatment is not an important source of these compounds compared to direct use and disposal of products containing residual amounts.

Acknowledgments We thank Collin Just and Craig Just of the University of Iowa Environmental Science and Engineering Laboratories for their help and discussion concerning our analytical methods. We also thank Steve Julius at the Iowa City South wastewater treatment plant for his help in acquiring samples, Jamie Nivala for supplying all necessary information concerning the plant’s design and operation, and Mark Weldon for analysis of upstream WWTPs on the Iowa River. We also acknowledge the University of Iowa Environmental Health Science Research Center (National Institute of Environmental Health Sciences Grant P30 ES05605), the National Science Foundation Research Training Grant in Gene Expression and Bioremediation, and the University of Iowa Center for Global and Regional Environmental Research for funding and supporting this study.

Supporting Information Available Chromatograms, compound names, abbreviations, and structural information. This material is available free of charge via the Internet at http://pubs.acs.org. 5530

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

Literature Cited (1) Saito, N.; Harada, K.; Inoue, K.; Sasaki, K.; Yoshinaga, T.; Koizumi, A. Perfluorooctanoate and Perfluorooctane Sulfonate Concentrations in Surface Water in Japan. J. Occup. Health 2004, 46, 49-59. (2) Olsen, G. W.; Church, T. R.; Larson, E. B.; van Belle, G.; Lundberg, J. K.; Hansen, K. J.; Burris, J. M.; Mandel, J. H.; Zobel, L. R. Serum Concentrations of Perfluorooctanesulfonate and Other Fluorochemicals in an Elderly Population from Seattle, Washington. Chemosphere 2004, 54, 1599-1611. (3) Boulanger, B.; Vargo, J.; Schnoor, J. L.; Hornbuckle, K. C. Detection of perfluorooctane surfactants in Great Lakes water. Environ. Sci. Technol. 2004, 38, 4064-4070. (4) Saito, N.; Sasaki, K.; Nakatome, K.; Harada, K.; Yoshinaga, T.; Koizumi, A. Perfluorooctane Sulfonate Concentrations in Surface Water in Japan. Arch. Environ. Contam. Toxicol. 2003, 45, 149158. (5) Kannan, K.; Hansen, K. J.; Wade, T. L.; Giesy, J. P. Perfluorooctane Sulfonate in Oysters, Crassostrea virginica, from the Gulf of Mexico and the Chesapeake Bay, USA. Arch. Environ. Contam. Toxicol. 2002, 42, 313-318. (6) Kannan, K.; Choi, J. W.; Iseki, N.; Senthilkumar, K.; Kim, D. H.; Masunaga, S.; Giesy, J. P. Concentrations of Perfluorinated Acids in Livers of Birds from Japan and Korea. Chemosphere 2002, 49, 225-231. (7) Kannan, K.; Newsted, J.; Halbrook, R. S.; Giesy, J. P. Perfluorooctanesulfonate and Related Fluorinated Hydrocarbons in Mink and River Otters from the United States. Environ. Sci. Technol. 2002, 36, 2566-2571. (8) Kannan, K.; Corsolini, S.; Falandysz, J.; Fillmann, G.; Kumar, K. S.; Loganathan, B. G.; Mohd, M. A.; Olivero, J.; Van Wouwe, N.; Yang, J. H.; Aldous, K. M. Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Env. Sci. Technol. 2004, 38, 4489-4495. (9) Hansen, K. J.; Johnson, H. O.; Eldridge, J. S.; Butenhoff, J. L.; Dick, L. A. Quantitative Characterization of Trace Levels of PFOS and PFOA in the Tennessee River. Environ. Sci. Technol. 2002, 36, 1681-1685. (10) Kannan, K.; Franson, J. C.; Bowerman, W. W.; Hansen, K. J.; Jones, J. D.; Giesy, J. P. Perfluorooctane Sulfonate in Fish-Eating Water Birds Including Bald Eagles and Albatrosses. Environ. Sci. Technol. 2001, 35, 3065-3070. (11) 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. (12) Giesy, J. P.; Kannan, K. Global Distribution of Perfluorooctane Sulfonate in Wildlife. Environ. Sci. Technol. 2001, 35, 13391342. (13) Moody, C. A.; Field, J. A. Perfluorinated Surfactants and the Environmental Implications of Their Use in Fire-Fighting Foams. Environ. Sci. Technol. 2000, 34, 3864-3870. (14) Schultz, M. M.; Barofsky, D. F.; Field, J. A. Fluorinated Alkyl Surfactants. Environ. Eng. Sci. 2003, 20, 487-501. (15) Boulanger, B.; Peck, A. M.; Schnoor, J. L.; Hornbuckle, K. C. Mass budget of perfluorooctane surfactant in Lake Ontario. Environ. Sci. Technol. 2005, 39, 74-79. (16) 3M. Docket AR-226. U. S. Environmental Protection Agency, Office of Pollution and Prevention and Toxic Substances: Washington, DC, 2000. (17) Key, B. D.; Howell, R. D.; Criddle, C. S. Defluorination of Organofluorine Sulfur Compounds by Pseudomonas sp. strain D2. Environ. Sci. Technol. 1998, 32, 2283-2287. (18) Metcalf & Eddy, Inc., Ed. Wastewater Engineering: Treatment and Reuse, 4th edition; McGraw-Hill: Boston, 2003. (19) 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, 545-551. (20) Iowa Department of Natural Resources, 2005. http://www.igsb.uiowa.edu/nrgislibx/. (21) Definition and Procedure for the Determination of Method Detection Limit, Revision 1.11. Code of Federal Regulations, Part 136, Appendix B; Fed. Regist. 1984, 49 (209), 43430-43431.

Received for review January 31, 2005. Revised manuscript received May 1, 2005. Accepted June 1, 2005. ES050213U