Biodegradation of Polyfluoroalkyl Phosphates as a Source of

Mar 31, 2010 - Wastewater treatment plants (WWTPs) have been identified as a major source of perfluorocarboxylates (PFCAs) to aqueous environments. Th...
0 downloads 16 Views 1MB Size
Environ. Sci. Technol. 2010, 44, 3305–3310

Biodegradation of Polyfluoroalkyl Phosphates as a Source of Perfluorinated Acids to the Environment HOLLY LEE, JESSICA D’EON, AND SCOTT A. MABURY* Department of Chemistry, University of Toronto, 80 St. George St., Toronto, Ontario, Canada M5S 3H6

Received September 17, 2009. Revised manuscript received March 10, 2010. Accepted March 12, 2010.

Wastewater treatment plants (WWTPs) have been identified as a major source of perfluorocarboxylates (PFCAs) to aqueous environments. The observed increase in PFCA mass flows from WWTP influent to effluent suggests the biodegradation of commercial fluorinated materials within the WWTP. Commercial fluorinated surfactants are used as greaseproofing agents in food-contactpaperproductsaswellaslevelingandwettingagents. As WWTPs are likely the major fate of these surfactants, their biodegradation may be a source of PFCA production. One class of commercial surfactants, the polyfluoroalkyl phosphates (PAPs), have been observed in WWTP sludge. While PAPs have been shown to degrade into PFCAs in a rat model, the presentstudyinvestigatestheirmicrobialfatetodeterminewhether the biodegradation of PAPs within a WWTP-simulated system will contribute to the load of PFCAs released. PAPs are applied commercially in mixed formulations of different chain lengths and substitution at the phosphate center. The effect of chain length and phosphate substitution on the biodegradation of PAPs was investigated by incubating mixtures of 4:2, 6:2, 8:2, and 10:2 monosubstituted PAPs (monoPAPs) in an aerobic microbial system and by separately incubating the 6:2 monoPAP and 6:2 disubstituted PAP (diPAP) for 92 days. Headspace sampling revealed production of the fluorotelomer alcohols (FTOHs) from the hydrolysis of the PAP phosphate ester linkages. Analysis of the aqueous phase revealed microbial transformation of the PAPs to the final PFCA products was possible. The majority of the oxidation products observed were consistent with previous investigations that have suggested fluorotelomer precursor compounds degrade predominantly via a β-oxidation-like mechanism. However, in this study, the detection of odd-chain PFCAs suggests that other pathways may be important. The present study demonstrated microbially mediated biodegradation of PAPs to PFCAs. This observation, together with the diPAP concentrations observed in WWTP sludge, suggest PAPs-containing commercial products may be a significant contributor to the increased PFCA mass flows observed in WWTP effluents.

Introduction In near-source regions, perfluorinated carboxylic acids (PFCAs) emitted from wastewater treatment plants (WWTPs) * Corresponding author phone: (416)978-1780; fax: (416)978-3596; e-mail: [email protected]. 10.1021/es9028183

 2010 American Chemical Society

Published on Web 03/31/2010

have been identified as a major source of PFCA contamination to aqueous environments (1, 2). PFCA concentrations have also been observed to increase from WWTP influent to effluent (3-5). In one of two WWTPs studied in-depth, Sinclair and Kannan (5) found strong correlations between the concentrations of perfluorooctanoate (PFOA) and perfluorononanoate (PFNA) and between perfluorodecanoate (PFDA) and perfluoroundecanoate (PFUnA), with higher concentrations of the even chain length PFCA observed as compared to the odd chain lengths. This PFCA congener profile is consistent with the biological production of PFCAs from fluorotelomer-based materials (6-11) and appears to result from biodegradation within the WWTP, as no correlation was observed between PFDA and PFUnA before activated sludge treatment. These studies together suggest that the biodegradation of fluorotelomer-based materials within WWTPs may be a source of PFCAs to the environment. Biotransformation of fluorotelomer alcohols (FTOHs) to PFCAs has been observed in mixed microbial systems and WWTP sludge (6-8), soil (9), rat hepatocytes and microsomes (10, 11), and whole rat models (12). FTOHs have no known direct commercial application but are instead used as building blocks in the synthesis of fluorinated polymers and fluorinated surfactants, which are themselves incorporated in the final sales products (13). Some evidence of polymeric degradation into PFCAs was recently reported in two soil biodegradation studies of a fluorotelomer acrylate polymer although the importance of this pathway is still widely debated (14, 15). Fluorinated surfactants may also be potential precursors to PFCAs, as the surfactants are expected to be less sterically constrained to microbial attack as compared to the polymers. The polyfluoroalkyl phosphates (PAPs) are commercial fluorinated surfactants used primarily in food-contact paper products and as leveling and wetting agents (16-19). PAPs have been identified as a potential source of human PFCA exposure as these chemicals can leach out of food packaging into food (20, 21). Biotransformation of PAPs to PFCAs was observed in a rat model (22), and human exposure was confirmed by recent measurements of PAPs in human sera at µgL-1 (ppb) concentrations (23). After consumer use, PAPs-containing products may be released into WWTPs. Recent detection of the disubstituted PAP (diPAPs) in WWTP sludge at levels (i.e., 50-200 ng/g) comparable to perfluorooctane sulfonic acid (PFOS) and far exceeding the PFCAs demonstrates the potential for these chemicals to contribute to the PFCA contamination observed in WWTPs (23). The present study investigated microbial transformation of PAPs to PFCAs by incubating in-house synthesized monosubstituted and disubstituted PAPs (monoPAPs and diPAPs) with aerobic microbes collected from a local WWTP. For oil repellency applications, PAPs are generally applied as a mixture of varying fluoroalkyl chain lengths as well as the mono-, di-, and trisubstituted phosphate congeners (21, 24). As a result, two studies were performed. The first, hereinafter called the “substitution experiment”, was performed to compare the effect of substitution at the phosphate center on PAPs biodegradation and involved separate incubations of the 6:2 monoPAP and 6:2 diPAP. The second, hereinafter called the “chain length experiment”, was performed to compare the effect of chain length on PAPs biodegradation and involved the incubation of monoPAPs of four different chain lengths (4:2, 6:2, 8:2, and 10:2). The PAP phosphate ester linkage is expected to undergo microbially mediated hydrolysis to produce the corresponding FTOH, which, based on previous investigations, is expected to oxidize to the PFCAs. VOL. 44, NO. 9, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3305

TABLE 1. Structures, Names, and Acronyms of the Target Analytes in This Study

Experimental Section Chemicals. The synthesis of the PAPs is described elsewhere (22). A list of all chemicals used in this study is provided in the Supporting Information (SI). All target analytes are listed in Table 1. Purging Control Experiment. Purging has been demonstrated to be an effective technique for removing unreacted FTOHs from the synthesis of commercial fluorinated materials dissolved in the aqueous phase (25). Analysis of the PAPs used in this study revealed the presence of FTOHs at significant quantities in the starting monoPAPs and 6:2 diPAP. As a result, it was important to reduce the levels of FTOHs in the starting PAP materials as much as possible before microbial inoculation so that any FTOH or PFCA production in the biodegradation experiments could be attributed to the degradation of PAPs. As PAPs are highly surface active, the effect of purging on the aqueous concentrations of PAPs over time was investigated. The experiment was performed in a purge-and-trap system described elsewhere (25), and illustrated in Figure S1. Briefly, 400 µg of 4:2, 6:2, 8:2, 10:2 monoPAPs, and 6:2 diPAP was spiked into two sets of polypropylene bottles containing 400 mL of deionized water, with one set purged with air for 6-7 days, and the other set left to stand. The aqueous phase was routinely sampled. At the end of the experiment, the gas diffuser tubes (only in the purging bottles), septa, bottle caps, and bottles were sonicated in methanol at 60 °C for 1 h. Experimental setup, extraction, and chromatographic analysis are described in the SI. Biodegradation Experiments Using Aerobic WWTP Microbes. Mixed liquor, a mixture of raw wastewater and sewage sludge, was collected from Ashbridges Bay WWTP (Toronto, ON). Prior to being used as inocula or autoclaved as sterile controls, the mixed liquor was aerated with inhouse air to maintain viability. Both the chain length and substitution experiments were performed in a purge-andtrap system with polypropylene bottles containing a total volume of 400 mL of aqueous phase. The setup included the following: (1) “mixed liquor only” control bottles (n ) 2), 3306

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 9, 2010

with 10% v/v of washed mixed liquor in mineral media, were included to monitor any production of FTOHs or PFCAs from potential fluorinated materials present in the WWTP mixed liquor; (2) “sterile” control bottles (n ) 2), with 10% v/v of autoclaved mixed liquor, 400 µg of monoPAPs for the chain length experiment or 400 µg of 6:2 monoPAP and diPAP for the substitution experiment, 300 mg of Hg2Cl2, and mineral media were included to quantify any nonmicrobial-mediated degradation; (3) “PAPs only” control bottles (n ) 2), with 400 µg of monoPAPs for the chain length experiment or 400 µg of 6:2 monoPAP and diPAP for the substitution experiment, and mineral media were included to quantify abiotic degradation; (4) “experimental” bottles (n ) 3), with 10% v/v of washed mixed liquor, 400 µg of monoPAPs for the chain length study or 400 µg of 6:2 monoPAP and diPAP, and media (Table S2). Prior to microbial inoculation, each bottle spiked with PAPs was purged for 5 days to strip the system of residual unreacted FTOHs that may have carried through the synthesis. After purging, the FTOHs present in the starting PAPs were reduced to within their detection limits. After microbial inoculation, each bottle was continuously purged with air for 92 days to strip volatile products (e.g., FTOHs) from the system. FTOHs were collected using XAD-2 cartridges. The aqueous phase was sampled to monitor the production of nonvolatile metabolites and disappearance of PAPs. Preparation of the mineral media, washing procedures of the WWTP mixed liquor, extraction procedures, and chromatographic and instrumental conditions are described in detail in the SI. Quality Assurance of Data. PFCAs and the saturated and unsaturated fluorotelomer carboxylates (FTCAs and FTUCAs) were quantified using internal calibration (Table S3). Due to the lack of native and internal standards for 3:3, 5:3, and 9:3 FTCAs, these analytes were quantified using 4:2, 6:2, and 10:2 FTCAs as surrogate standards. PAPs were quantified by external calibration as no appropriate internal standards were available. To confirm that external calibration was appropriate, the PAPs were spiked in mineral media treated with autoclaved mixed liquor and analyzed by both standard addition and external calibration for comparison. Details are discussed in the SI. Recoveries for the FTOHs were in the range of 58-91% (Table S4). The XAD cartridges used included a second XAD plug that acted as a breakthrough to determine any potential FTOH loss. The breakthroughs in all the vessels contained