Bisphenol A in Solid Waste Materials, Leachate Water, and Air

Article pubs.acs.org/est

Bisphenol A in Solid Waste Materials, Leachate Water, and Air Particles from Norwegian Waste-Handling Facilities: Presence and Partitioning Behavior Nicolas Morin,†,‡ Hans Peter H. Arp,*,† and Sarah E. Hale† †

Department of Environmental Engineering, Norwegian Geotechnical Institute, P.O. Box 3930, Ullevål Stadion, N-0806 Oslo, Norway ‡ Department of Chemistry, Linnaeus väg 6, Umeå University, SE-901 87 Umeå, Sweden S Supporting Information *

ABSTRACT: The plastic additive bisphenol A (BPA) is commonly found in landfill leachate at levels exceeding acute toxicity benchmarks. To gain insight into the mechanisms controlling BPA emissions from waste and waste-handling facilities, a comprehensive field and laboratory campaign was conducted to quantify BPA in solid waste materials (glass, combustibles, vehicle fluff, waste electric and electronic equipment (WEEE), plastics, fly ash, bottom ash, and digestate), leachate water, and atmospheric dust from Norwegian sorting, incineration, and landfill facilities. Solid waste concentrations varied from below 0.002 mg/kg (fly ash) to 188 ± 125 mg/kg (plastics). A novel passive sampling method was developed to, for the first time, establish a set of waste-water partition coefficients, KD,waste, for BPA, and to quantify differences between total and freely dissolved concentrations in waste-facility leachate. Log-normalized KD,waste (L/kg) values were similar for all solid waste materials (from 2.4 to 3.1), excluding glass and metals, indicating BPA is readily leachable. Leachate concentrations were similar for landfills and WEEE/ vehicle sorting facilities (from 0.7 to 200 μg/L) and dominated by the freely dissolved fraction, not bound to (plastic) colloids (agreeing with measured KD,waste values). Dust concentrations ranged from 2.3 to 50.7 mg/kgdust. Incineration appears to be an effective way to reduce BPA concentrations in solid waste, dust, and leachate.

■

INTRODUCTION Bisphenol A (BPA, 2,2-(4,4′-dihydroxydiphenyl)propane, CAS Registry No. 80-05-07) is used in vast quantities,1,2 with an estimated 4.6 million tons being produced globally in 2012.3 Its primary use is as a monomer in the production of polycarbonate and epoxy resins. Other uses are as a stabilizing agent in plastics and as an additive in thermal paper or paper coatings. BPA is a known endocrine disruptor. Predicted noeffect concentrations (PNEC) for chronic toxicity of 1.6 μg/L and acute toxicity of 11 μg/L in fresh water have been proposed in a European Union risk assessment,4 as well as a soil chronic PNEC of 3700 μg/kg dry weight.5 With so much BPA being produced for use in consumer products, it is not surprising that BPA is considered ubiquitous in the environment.6 It is commonly included in environmental monitoring studies from various countries (for example, The Netherlands,7,8 China,9 Germany,10,11 Norway,12 Taiwan,13 Japan,14,15 and America16). Klecka et al.17 compiled BPA water monitoring data from Europe and North America and reported that median surface fresh water concentrations were notably below the PNEC, at 0.08 μg/L (n = 1068) and 0.01 μg/L (n = 848) in North America and Europe, respectively. However, © XXXX American Chemical Society

water levels can commonly be found above the acute PNEC in landfill leachate. In Norway, a compilation of landfill leachate data from 2002 to 201212 reported a median of 17 μg/L (interquartile range, IQR, 1−62 μg/L, maximum 692 μg/L). Outside of Norway, landfill leachate concentrations range from 0.1 to 17 200 μg/L in diverse Japanese studies18−23 and from 0.01 to 107 μg/L in four Swedish landfills,24 and exceptionally high leachate concentrations of BPA (4200−25000 μg/L) were reported in a German study.25 One study found that BPA in landfills does not decompose under anaerobic conditions,26 implying that landfills can be a persistent source of BPA. In response, researchers have been prompted to consider remediation options to lower BPA levels in landfill leachate.27 As an alternative to landfilling, incineration has been found to be an effective way to remove BPA from waste, as BPA is prone to thermal degradation above 400 °C.28 A more detailed Received: March 13, 2015 Revised: May 26, 2015 Accepted: June 9, 2015

A

DOI: 10.1021/acs.est.5b01307 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

of 1, 10, 50, 100, and 1000 μg/L were prepared in 500 mL glass flasks with glass stoppers by adding Milli-Q water and spiking with a solution of BPA in ethyl acetate (such that the co-solvent did not exceed 0.2% of the total volume). To calibrate for two thicknesses, a 76 μm strip (0.2 g, CS Hyde USA) and a 55 μm strip39 (0.2 g) were introduced to the flasks; they are referred to as POM-76 and POM-55, respectively. The flasks were equilibrated by shaking end-over-end at 13 rpm for a period of 8 weeks in the dark at room temperature. The POM strips were removed, and CPOM and Cwater were determined for BPA as described below. A kinetic experiment was also carried out for POM-76 at Cwater = 100 μg/L under the same conditions by placing 10 POM-76 pieces (0.2 g each) in the flasks as above and removing duplicate strips for BPA quantification at days 3, 7, 15, 21, and 28. BPA spiked into blank control flasks (no POM added, 100 μg/L spike, triplicate) showed an average mass loss of BPA of 17 ± 4% over 28 days, which was corrected for in the kinetic uptake experiments. Field Campaigns. Waste-handling facilities were chosen to provide a broad range of waste-handling methods and types of solid waste fractions. Twelve different facilities in southeastern Norway were sampled during two or three sampling campaigns, June−October 2013, October−December 2013, or March− June 2014. The facilities included three specialized landfills (accepting bottom ash, fly ash, and sewage-sludge digestate for composting, though all containing municipal/industrial waste), two combustible waste-sorting facilities (municipal/industrial waste), and seven waste electric and electronic equipment (WEEE)/vehicle shredding and sorting facilities. Due to requests from some site owners to keep the data anonymous, the locations are referred to as Landfill A−C, Incineration/ Sorting A,B, and WEEE/Vehicle A−E (only five WEEE/vehicle locations are assigned, as two sets of two individual facilities shared water drainage and therefore leachate drainage and air emissions). Based on logistics or feasibility, solid waste, leachate water, and air were sampled from these facilities. More details related to the field sites and sampling campaign are presented in the SI (Table S2 and Figure S1). Sampling. Solid waste samples (4−12 kg) were collected by hand (while wearing nitrile gloves) into 4 L polyurethane bags from random locations within each facility. Samples were collected such that they were visually homogeneous and representative of a particular waste fraction (e.g., coarse/fine ash, coarse/fine combustibles, cable plastics, etc.). Samples were transported back to the laboratory and stored at 4 °C until further processing. Descriptions of the waste fractions sampled are presented in the SI (Table S3). Grab (active) sampling was used to obtain total leachate concentrations, and POM passive sampling was used to obtain freely dissolved leachate concentrations. The grab samples were obtained by submerging a pre-sterilized 1 L green-tinted glass bottle in the leachate water (either an open stream or inside a culvert or manhole) on the first day of the relevant field campaigns. The bottles were wrapped in aluminum foil and transported cool (4 °C) to the laboratory. The same day, 2 g of sodium azide (Sigma-Aldrich, USA) was added to the water samples to prevent microbial degradation of BPA; they were also spiked with BPA-d6 (used as a recovery standard, TRC, Canada) to check the degradation/extraction recovery and frozen until analysis. Leachate passive samples were obtained by deploying POM76 samplers housed in stainless steel frame into the leachate water for the entirety of the sampling campaigns (ca. 2−3

overview of monitoring levels of BPA is presented in the Supporting Information (SI) and Table S1. The focus of BPA emission and exposure research has been on food and product packaging, with packaging materials made from polycarbonate plastics and epoxy−resin-lined containers identified as substantial sources of exposure, along with thermal paper.29−33 When these materials are disposed of, they enter the waste stream to form bulk waste fractions that are rich in BPA, such as combustible waste (plastic sub-fractions)20 and incineration residues.18,23 How these different types of waste fractions contribute to BPA leachate concentrations at landfills and other waste facilities remains unclear. Compared to food packaging, comparatively limited research has been carried out to investigate the mechanisms from which BPA can be released from bulk waste fractions. Further, it remains unknown how environmental concentrations around landfilling facilities compare to other kinds of waste-handling facilities, such as fragmenting, sorting, incineration, and recycling facilities. A Japanese survey reported that BPA concentrations in leachate appeared independent of waste composition at landfill sites.22 To gain new insight into the sources and mechanisms regulating BPA concentrations at waste-handling facilities, we conducted a comprehensive field and laboratory campaign comprising 12 different facilities and eight types of waste categories to study their presence and partitioning behavior. A key novel aspect of the presented investigation is the development and utilization of a passive sampling method to specifically target the freely dissolved concentrations in water, which allows for measuring the waste-water partitioning behavior of these eight waste categories, as well as a comparison of the total and freely dissolved concentration in landfill leachate. Freely dissolved concentrations are more appropriate to consider when describing partitioning behavior of contaminants, as they more closely regulate environmental fate and bioavailability.34,35 The water-phase passive sampling material used was polyoxymethylene (POM), which is slightly polar and therefore appropriate for BPA.36,37 The hypotheses we set out to test in this study were the following: (1) substantial amounts of BPA in landfill leachate originate from plastic-containing waste fractions; (2) BPA leachate concentrations are primarily freely dissolved (and not bound to plastic particulates or colloids); and (3) BPA concentrations in air and water from waste-sorting and incineration facilities are lower than from landfills.

■

MATERIALS AND METHODS Bisphenol A. BPA is moderately hydrophobic (log Kow = 3.4), weakly acidic (pKa = 9.8 and 11.3), and soluble in water (solubility of 300 mg/L), with recommended environmental half-lives of 4 h in air, 4.5 days in water, and 300 days in soil.38 Calibration of Polyoxymethylene Samplers. A novel method using POM passive samplers36 for quantifying the freely dissolved fraction of BPA in leachate water and for determining waste-water partition coefficients was developed in this study. For this method, a reliable understanding of BPA uptake kinetics into POM, as well as the POM−water partition coefficient, KPOM, is needed: KPOM(L/kg) = C POM /Cwater

(1)

where CPOM is the equilibrium concentration in the POM phase (μg/kg) and Cwater is the equilibrium freely dissolved concentration in the water phase (μg/L). To quantify KPOM over a concentration range for landfill leachate, BPA solutions B


Article


extraction recovery. The samples were extracted using a solid phase extraction method with Strata-SDB L cartridges (500 mg, 6 mL, 100 μm, 260 Å, USA). The cartridges were first conditioned with 5 mL of ethyl acetate and 5 mL of preextracted distilled water. Following this, between 1 and 50 mL samples of leachate water were loaded onto the cartridges. The cartridges were then rinsed with 10 mL of pre-extracted distilled water and dried using a vacuum pump. BPA was eluted from the cartridges using 10 mL of ethyl acetate. This extract was evaporated to dryness and treated as above before GC-MS analysis. POM passive samplers were placed in an extraction vial and spiked with the BPA-d16 RS. They were then cold extracted for 7 days with ethyl acetate (20 mL) by shaking end-over-end at 13 rpm and then handled as the solid waste samples. Quality Assurance and Control. All the solvents used were GC grade or Emsure quality. All stock solutions were prepared from pure BPA powder dissolved in GC-grade ethyl acetate and were stored at −20 °C. All glassware was rinsed with acetone, washed in a laboratory dishwasher, and pyrolyzed at 450 °C (except the volumetric glassware used for standards, which did not undergo the pyrolysation step) prior to use. Before storage, water samples were spiked with sodium azide to prevent microbial degradation and with BPA-d6 to check the degradation/extraction recovery. In addition all samples were also spiked with BPA-d16 prior to extraction. All raw data results were corrected on the basis of the recovery percentages of all these standards. During a GC-MS sequence, the calibration standards (1−100 μg/L) were injected at least twice to take into account the possibility of signal drift with time. The quantification was done with the internal calibration technique (internal standard PCB77). All the samples were quantified within the range of the calibration standard (1−100 μg/L). Solvent blank samples were analyzed regularly to check for potential contamination from the GC-MS. All analytes were quantified using a quantification and a confirmation transition from the MS. The quantification transition was chosen as the most intense peak, and the ratio of the confirmation/ quantification transition was used when it was in the same range (±20%) as that of the calibration standards. GC-MS Analysis. The concentrations of BPA in the derivatized sample extracts and standards were quantified using a gas chromatograph 6850 coupled to a mass spectrometer 5973 (Agilent, USA). The chromatographic column was a SLB-5ms fused silica capillary column 30 m × 0.25 mm × 0.25 μm (Supelco, USA). A five-point calibration was made at 1, 5, 10, 50, and 100 μg/L BPA concentration standards, in which BPA-d16, BPA-d6, and PCB 77 were added. Further GC-MS method details are presented in the SI (section S4). Waste-water Partition Coefficients. A batch-shake method was used to obtain waste-water partition coefficients, KD,waste:

months, for logistic reasons and to ensure equilibrium partitioning), then wrapped in aluminum foil, placed in glass jars, and transported at 4 °C back to the laboratory where they were stored at −20 °C until CPOM analysis. Further details of both grab and passive sampling of leachate are provided in the SI (section S2). At selected waste-handling facilities, ambient air particulate matter (PM) samples were obtained at the site of most activity, either next to a shredder, waste sorter, loading dock or in a central location, using a high-volume (HighVol) air sampler (Digitel, Switzerland), which was equipped with a PM10 cutoff, 150 mm diameter glass fiber filter (GFF, Sigma-Aldrich, USA), to quantify the air-particle-associated BPA concentration. The HighVol was deployed for 1−5 days. Due to intense particle loadings in some areas, the cutoff may have been compromised, and particles larger than PM10 may have entered the HighVol. Air passive samplers consisting of XAD-2 resin beads contained in a stainless tube, as designed by Wania et al.,40 were deployed in central locations, but not directly next to particle shredders or areas were dust was visible. These samplers were deployed for the entirety of a given sampling campaign (2−3 months). Both GFF and XAD-2 tubes were wrapped in aluminum foil after deployment and transported at 4 °C back to the laboratory where they were stored at −20 °C until analysis. Further details are provided in the SI (section S3). Sample Preparation. In the laboratory, solid waste samples were further homogenized in the polyurethane bags by shaking or manual mixing, before 20−400 g was randomly sampled from within the bag for grinding. All samples were ground until they could pass through a 2 or 4 mm sieve (depending on the material, as indicated in Table S2). Crushing was carried out using either a BB100 Retsch jaw crusher (VWR, Norway) (typically for glass and coarse ashes), a kitchen handblender (Braun or Phillips), a hand-powered malt mill (Bryggeland, Norway) (typically for fluff and plastic), a mortar and pestle, or simply by sieving through the appropriate mesh. Hard plastics and metal materials (>4 mm) were the most difficult to crush and sieve, and thus for the four samples of this consistency (two WEEE samples from the site “WEEE/Vehicle B”, one vehicle fluff and one vehicle plastic sample from the site “WEEE/Vehicle E”), the original mass fraction of these materials may be slightly misrepresented in the mixed, crushed sample that was used for analysis. Quantification of BPA. Solid waste samples, including the GFF filters and the XAD-2 resins, were extracted using a Soxhlet method (Behr Labor-Technik, Germany) with 100 mL of ethyl acetate (GC-MS grade, 99.8% purity, Merck, KGaA, Germany) for 12 h at 105 °C. The samples were spiked with a recovery standard (RS; BPA-d16, 99.9% purity, Supelco, USA) prior to extraction to check extraction recovery. Following extraction, between 10 μL and 10 mL of the solvent was evaporated to dryness using a vacuum centrifuge (Vacuubrand 2C, Vakuum Service AS, Germany). The residue was then dissolved in 950 μL of ethyl acetate that contained an internal standard (IS) at 50 μg/L to check for matrix effects (PCB-77, 99.97% purity, Fluka, Switzerland) and 50 μL of a derivatization reagent (MTBSTFA, at 60 °C for 30 min, > 97% purity, SigmaAldrich, USA) in order to allow for GC-MS quantification. All solid concentration data are presented on a dry weight basis (d.w.). Frozen grab leachate water samples were thawed in the dark and then spiked with BPA-d16 to act as a RS to check for the

KD,waste(L/kg) = Cwaste/Cwater

(2)

where Cwaste is the concentration in the waste at equilibrium (μg/kg). KD,waste was determined here using an approach that was adapted from a standard method for metals in waste materials (EN 12457), by using POM to quantify Cwater and by increasing the duration of shake time from 1 day to 28 days to ensure equilibrium. Between 0.5 and 2 g of ground solid waste material (≤4 mm), along with 0.1 g of pre-cleaned POM, were C


Article


inferred that the deployment time used here (2−3 months) would be more than sufficient to reach equilibrium if BPA leachate concentrations were time independent. In areas of fluctuating concentrations, the polar organic chemical integrative sampler (POCIS), which has been used to measure the concentration of BPA in spiked tap water, surface water, wastewater, and estuary water41−45 would be more appropriate as it does not need to reach equilibrium in order to determine water concentrations. However, in areas of stable concentration, POM is advantageous compared to POCIS, as extrapolation of Cwater is not dependent on (multi-phase) uptake rates, and sample handling time is reduced.46 Solid and Leachable Waste Concentrations. The total and leachable BPA concentrations quantified for the different solid waste categories and sub-fractions are shown in Table S4, with waste category results plotted in Figure 2. Note that the waste category results presented in Figure 2 are weighted averages based on the annual mass produced in Norway of the various waste sub-fractions belonging to a waste category (as presented in Tables S3 and S4). As an example, the waste category WEEE contains the sub-fractions “BFR plastic”, “cable plastic”, “other plastics”, and “metals”, which are reported to be generated at 2, 15, 48, and 80 megatons/year in Norway, respectively (Table S4).38 These WEEE sub-fractions were measured in our study to have BPA concentrations of 84400, 29100, 200500, and 1170 μg/kg, respectively. Thus, the weighted average of BPA in WEEE was 71100 μg/kg = {(2×84400 μg/kgBFR plastic + 15×29100 μg/kgcable plastic + 48×200500 μg/kgother plastics + 80×1170 μg/kgmetals)/(2 + 15 + 48 + 80)}. From Figure 2, the waste category with the largest concentrations of BPA was plastics (weighted average 188000 ± 125000 μg/kg), followed by the plastic-rich waste fractions WEEE (71200 ± 46700 μg/kg) and vehicle fluff (6490 ± 3350 μg/kg). The lowest concentrations were found in fly ash ( WEEE > vehicles > combustibles ≈ digestate > bottom ash > glass > fly ash). Leachable concentrations at L/S 10 ranged from < LOQ for fly ashes to 1970 μg/kg for plastics (Figure 2 and Table S4), roughly corresponding to 1% of the total BPA leaching into the water phase for most waste samples (0.6−1.6%), except for glass samples (30.6%) and fly ash (80% sorption equilibrium being achieved within 7 days of shaking (Figure 1). A follow-up experiment with more sampling events in the first 7 days would be recommended to better characterize the uptake kinetics. For determining waste-water partitioning coefficients, the 28 day shaking test is therefore conservative regarding POM uptake, though it is still recommended to account for potential slow desorption kinetics from waste fractions. This kinetic system is not representative of that encountered when fielddeploying POM into flowing leachate water, but it can be D


Article


Figure 2. Waste category total BPA concentrations, Cwaste (left n-value), and leachable BPA concentrations, Cleachable (right n-value) at a liquid/solid (L/S) ratio of 10 of the types of solid waste fractions considered in this study. Data presented are weighted averages based on the annual mass produced in Norway of the various waste sub-fractions belonging to a waste category. The waste categories are organized from the smallest to largest BPA concentrations. Also presented is the predicted no-effect concentration (PNEC) of BPA in soil of 3700 μg/kg d.w.2

latter observation could be due to BPA typically not being associated within these materials, though trace levels could be on their surface as coatings or labels (similar to our results for glass). Regarding incinerator ashes, a Japanese survey of leachate from ash-landfills concluded that bottom ash was a minor source of BPA in leachate, but “solidified fly ash” and incombustibles were major sources of BPA.23 Our study agrees with the former but not with the latter conclusion; we speculate that this disagreement may be related to the solidification process of the fly ash in this Japanese survey.23 Waste-water Partitioning. The solid waste-water partitioning coefficients (KD,waste) for all waste categories and subfractions are shown in Table S4. All waste sub-fractions have similar log KD,waste values (ranging from 2.1 ± 0.6 to 3.1 ± 1.0) except for glass waste (log KD = 1.5 ± 0.3), where BPA concentrations were close to the LOQ, and WEEE metals (log KD = 1.8 ± 0.4). The mass fraction of total organic carbon in the waste, f TOC, was used to normalize KD,waste, as this parameter is generally correlated with KD values of organic compounds, according to

quantification, and also to BPA residue being mostly on the glass surface (e.g., from epoxy−resin coatings or labels). Additionally, the average pH for the leachable BPA samples for the glass (pH 9.9) were slightly higher than the pKa1 (pKa1 = 9.8) for BPA, which means that about 50% of BPA was present in its single negatively charged form. For bottom ashes, the BPA was present at about 50% in its single- and 50% in its double-negatively charged form (pH 10.8 and pKa2 = 11.3). For fly ashes, the pH was 12.2, which is higher than pKa2, meaning that if any BPA was present it would be in its double-negatively charged form. A recent survey of contaminants in Norwegian vehicle fluff48 measured an average BPA concentration of 5000 μg/kg (±5000 μg/kg; n = 10), which is comparable to the result determined here (6492 ± 3350 μg/kg; n = 12). Another Norwegian survey reported median BPA concentrations of 536 ± 446 μg/kg (n = 32) in sludge samples from 8 different water treatment plants10 (SI, section S1), which is consistent with the digestate samples (888 ± 401 μg/kg; n = 8). Several studies have quantified concentrations of BPA in plastic waste and products. Yamamoto and Yasuhara49 measured the total BPA concentration in 17 plastic waste samples. Ten of their samples had concentrations below the limit of detection, while the other seven had concentrations between 71000 and 1280000 μg/kg (average 605000 μg/kg). Xu et al.50 reported total BPA concentrations between 1600 and 12100 μg/kg for five different plastic wastes, and Biles et al.51 quantified BPA concentrations in PC bottles to be between 7000 and 58000 μg/kg. Thus, concentrations of BPA in plastics are highly variable and quite dependent on the type of material (this is further discussed in the SI, section S1); however, the results measured here are well within the range reported in these other studies. Regarding leachable concentrations, a previous Japanese study reported that less than 3% of the total BPA was leached from nine plastic samples,49 in agreement with our average of 1%. Xu et al.50 reported a similar level of BPA leaching from polycarbonate and polyethylene (

Bisphenol A in Solid Waste Materials, Leachate Water, and Air

Recommend Documents