- Email: [email protected]
Fate of 14C–triclocarban in biosolids-amended soils
Science of the Total Environment 408 (2010) 2726–2732
Contents lists available at ScienceDirect
Science of the Total Environment j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i t o t e n v
Fate of
14
C–triclocarban in biosolids-amended soils
Elizabeth Hodges Snyder a,b,⁎, George A. O'Connor c, Drew C. McAvoy d a
Soil and Water Science Department, University of Florida, 408 Newell Hall, Gainesville, Florida, 32611, USA Department of Health Sciences, University of Alaska Anchorage, DPL 404, 3211 Providence Drive, Anchorage, AK 99508-4614, USA c Soil and Water Science Department, P.O. Box 110510, University of Florida, Gainesville, FL 32611-01519, USA d Environmental Safety Department, P.O. Box 538707, The Procter & Gamble Company, Cincinnati, OH, 45253-8707, USA b
a r t i c l e
i n f o
Article history: Received 14 September 2009 Received in revised form 24 December 2009 Accepted 5 January 2010 Available online 2 April 2010 Keywords: Triclocarban Antibacterial Biosolids Sludge Biodegradation
a b s t r a c t Triclocarban (TCC) is an antibacterial compound commonly detected in biosolids at parts-per-million concentrations. Approximately half of the biosolids produced in the United States are land-applied, resulting in a systematic release of TCC into the soil environment. The extent of biosolids-borne TCC environmental transport and potential human/ecological exposures will be greatly affected by its bioavailability and the rate of degradation in amended soils. To investigate these factors, radiolabeled TCC (14C–TCC) was incorporated into anaerobically digested biosolids, amended to two soils, and incubated under aerobic conditions. The evolution of 14CO2 (biodegradation) and changes in chemical extractability (bioavailability) was measured over time. Water extractable TCC over the study period was low and significantly decreased over the first 3 weeks of the study (from 14% to 4% in a fine sand soil and from 3 to < 1% in a silty clay loam soil). Mineralization (i.e. ultimate degradation), as measured by evolution of 14CO2, was < 4% over 7.5 months. Methanol extracts of the amended soils were analyzed by radiolabel thin-layer chromatography (RAD-TLC), but no intermediate degradation products were detected. Approximately 20% and 50% of the radioactivity in the amended fine sand and silty clay loam soils, respectively, was converted to bound residue as measured by solids combustion. These results indicate that biosolids-borne TCC becomes less bioavailable over time and biodegrades at a very slow rate. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Triclocarban (Fig. 1) is a common constituent of domestic wastewater due primarily to its use as an active ingredient in antibacterial bar soaps. Up to 98% of TCC present in wastewater influent is removed from the aqueous phase by activated sludge treatment (TCC Consortium, 2002; Heidler et al., 2006), and approximately 75% is concentrated in the solid fraction (i.e. sludge) (Heidler et al., 2006). Sludge is often converted, via a variety of methods to meet regulations in the US EPA 40CFR503 (Standards for the Use or Disposal of Sewage Sludge), to biosolids approved for land application. The USEPA estimates 8 × 106 dry metric tons of biosolids are produced annually (USEPA, 2006), and approximately half are land-applied (NRC, 2002). Documented biosolids-borne TCC concentrations range from 5 to 51 mg kg− 1 (Heidler et al., 2006; Chu and Metcalfe, 2007; Halden, 2007; Sapkota et al., 2007), with an apparent mean of approximately 20 mg kg− 1. The recently published Targeted National Sewage Sludge Survey (TNSSS)
⁎ Corresponding author. Department of Health Sciences, University of Alaska Anchorage, DPL 404, 3211 Providence Drive, Anchorage, AK 99508-4614, USA. Tel.: +1 907 786 6540; fax: +1 907 786 6572. E-mail addresses: lizah@ufl.edu, [email protected] (E.H. Snyder), gao@ufl.edu (G.A. O'Connor), [email protected] (D.C. McAvoy). 0048-9697/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2010.01.005
quantified TCC concentrations up to 441 mg kg− 1 in wastewater treatment sludge and/or biosolids (mean and standard deviation: 39± 60 mg kg− 1) (USEPA, 2009a). Little is known about the persistence and bioavailability (i.e. the degree of ability to be absorbed and ready to interact in organism metabolism; USEPA, 2009b) of TCC in the terrestrial environment following land application of biosolids. Efficient biodegradation of any organic contaminant requires that the compound be bioavailable to organisms capable of degradation. The limited solubility (0.045 mg L− 1) and moderate log Kow (3.5) (Snyder et al., in review) suggests TCC will preferentially partition to the biosolids matrix, potentially limiting bioavailability. Migration into or onto other soil fractions following land application of biosolids could subsequently alter bioavailability over time. Any process that acts to lower the effective solution concentration of a soil contaminant could be expected to affect the rate of biodegradation (Alexander, 1999) and, potentially, the effective toxicity of the contaminant. Few studies have characterized TCC degradation in the combined environmental matrices of wastewater treatment products and soil. These studies are generally limited to sludge-, soil-, or sediment-only systems (Gledhill, 1975; Ying et al., 2007; Miller et al., 2008), do not attempt to isolate aerobic biodegradation intermediates (Xia et al., 2008; Wu et al., 2009), or quantify degradation as loss of extractable parent compound over time (Ying et al., 2007; Xia et al., 2008; Wu et al., 2009), which could be attributable to the formation of bound
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
2727
Fig. 1. Chemical structure and select physicochemical properties of triclocarban (TCC; N-(4-chlorophenyl)-N′-(3,4-dichlorophenyl) urea).
residues rather than actual primary and/or ultimate degradation. Aerobic degradation products identified in a bench-top continuous flow activated sludge (CFAS) system included 3,4-dichloroaniline (DCA), and to a lesser extent, p-chloroaniline (PCA), aniline, and TCC condensation products (i.e. 4-chloro-4′-(4-chloroaniline)-azobenzene and 4,4′-dichloro-azobenzene) (Gledhill, 1975). The aforementioned studies provide useful information for the assessment of TCC environmental fate and transport, but multiple gaps in our understanding of land-applied biosolids-borne TCC fate remain. In particular, better information is needed on bioavailability, biodegradation, and metabolite formation in soils. Further, the sorption of TCC to organic matter and the potential impacts of partitioning on degradation necessitate the inclusion of biosolids as the source of TCC in an assessment of its fate following introduction into the terrestrial environment. Therefore, we characterized biodegradation of TCC using a respirometry test, in which the microbial metabolism of 14C-labeled TCC was quantified by monitoring the evolution of 14CO2 from spiked biosolids amended to soil. The test was also coupled with methods to isolate potential degradation products, and to characterize changes in chemical extractability (which can be related to changes inbioavailability) over the study period. Results were used to characterize the influence of soil type on biosolids-borne TCC persistence and extractability. This paper is one in a planned series of papers characterizing multiple components of a human and ecological risk assessment of land-applied, biosolids-borne TCC.
2. Materials and methods 2.1. Chemicals, biosolids, and soils 14 C–triclocarban labeled on the 4-chloroaniline ring [specific activity: 75 mCi/mmol; purity: 98.5%], which is expected to be the most rapidly degraded of the two rings (Gledhill, 1975), was synthesized by Amersham Life Science and supplied by the Procter & Gamble Company. Barium chloride (BaCl2), chloroform (CHCl3), hydrochloric acid (HCl), methanol (MeOH), sodium hydroxide (NaOH), and sodium azide (NaN3) were purchased from Fisher Scientific. EcoScint A liquid scintillation cocktail was purchased from National Diagnostics (Atlanta, GA). Anaerobically digested, Class B dewatered cake biosolids (identification code: CFBC; indigenous TCC concentration: 40 mg kg− 1) was collected from the belt-press of a wastewater treatment plant (WWTP) serving primarily single family homes and a dye factory in Ohio. The biosolids-borne TCC concentration was previously determined using extraction and analysis methodology adapted from Halden and Paull (2005) and Chu and Metcalfe (2007) (Snyder et al., in review). Samples of two soils, a Florida Immokalee fine sand (sandy, siliceous, hyperthermic Arenic Alaquods) (clay: 10 g kg− 1, OC: 5.5 g kg− 1) and an Ohio Genesee silty clay loam (fine-loamy, mixed, superactive, mesic Fluventic Eutrudepts) (clay: 330 g kg− 1, OC: 23 g kg− 1) (Walkley and Black, 1934)
were collected from sites with no known history of receiving landapplied biosolids or sludge. 2.2. Study design The study design was adapted from the USEPA Office of Prevention, Pesticides, and Toxic Substances (OPPTS) harmonized test guideline Soil Biodegradation (835.3300)1 (USEPA, 1998). The OPPTS Guidelines were developed through blending USEPA Office of Pollution Prevention and Toxics (OPPT), USEPA Office of Pesticide Programs (OPP), and Organization for Economic Cooperation and Development (OECD) test guidance to produce standardized methods for generating a uniform and consistent chemical properties database. Glass, round-bottom, 30 mL centrifuge tubes were used to prepare 140 samples [2 soils×1 biosolids×1 rate×4 replicates×8 sampling periods×2 treatments (biotic/inhibited)+12 controls]. Biosolids samples (0.10 g d.w. equivalent) were loaded into the centrifuge tubes, spiked drop-wise with 1.3 × 106 dpm 14C-TCC g biosolids− 1 (d.w. equivalent) in methanol, allowed to equilibrate 24 h, and amended with 10 g (d.w. equivalent) of soil to simulate a realistic field loading rate (22 Mg ha− 1 or ∼10 tons acre− 1). The loaded centrifuge tubes were capped and vortexed for 30 s to mix the biosolids and soil, and avoid potential removal of radioactivity through use of a stirring rod. The radioisotope spike increased the effective biosolids-borne TCC concentration to ∼65 mg TCC kg− 1, and the total TCC load in amended soil was ∼0.65 mg TCC kg amended soil− 1. The inhibited treatment was included in the study design to facilitate differentiation between biotically-induced and non-biotically-induced effects on TCC degradation and extractability with time. The inhibited samples were prepared by vortexing 1000 μg g− 1 sodium azide (0.1%) in methanol carrier solvent into a subset of the biosolids-amended soils. Sodium azide is effective against bacteria (Lichstein and Soule, 1944) and fungi (Morton et al., 1993; Seki et al., 1998), but the term “inhibited,” rather than “abiotic,” is used to describe the treatment because of the differential effects of sodium azide on gram negative and gram positive microorganisms. The chemical primarily acts as a bacteriostatic agent against gram negative microoorganisms, and is less effective against gram positives (Lichstein and Soule, 1944). Each centrifuge tube was fitted with a silicon/Teflon septum and connected, via Dow Corning Silastic laboratory tubing, to a series of glass trapping vials containing 5 mL 0.2 M potassium hydroxide. The base traps were used to collect evolved 14CO2 and CO2 as a measure of 14 C–TCC mineralization and soil microbial respiration, respectively. The first positioned vial remained empty to prevent backflow into the centrifuge tubes if pump failure were to occur. An oil-less pump 1 Additional relevant guidelines for assessment of aerobic and anaerobic biodegradation have since been published, including Anaerobic Biodegradability of Organic Compounds in Digested Sludge: By Measurement of Gas Production (835.3420), Aerobic Soil Metabolism–Anaerobic Soil Metabolism (835.4100-835.4200).
2728
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
aerated samples with humidified, CO2-free air by first pumping air through 1 M potassium hydroxide, and then deionized water. Although not prescribed by the OPPTS guideline, CO2 was removed from incoming air to facilitate calculations of total CO2 evolved from the amended-soil samples (as a measure of biotic activity). The experiment was conducted in the dark at ∼ 23 °C. Samples were weighed weekly and sterile water was added, via syringe needle through the tube septum as needed, to maintain a soil moisture content equivalent to field capacity [i.e. 10% w/w (fine sand) and 20% w/w (silty clay loam)]. 2.3. Sample extraction and liquid scintillation counting (LSC) analysis Once a week, position-1 base traps were removed, remaining traps were moved forward, and a fresh trap was added to the newly open position-3. Four replicates of each biosolids-amended-soil treatment were periodically removed and sequentially extracted (sonication, 1 h) with 20 mL of deionized water (twice), 20 mL of methanol (twice), and 20 mL of 1 M NaOH (once). The water and methanol extractions served as estimates of TCC bioavailability in biosolidsamended soils (Kelsey et al., 1997; Jussara et al., 2006), and the NaOH extraction estimated the extent of TCC incorporation into the humic fraction (Schnitzer, 1982). A subsample of each sequentially extracted amended-soil replicate was removed, weighed, and combusted on a sample oxidizer to complete a mass balance of the 14C–TCC spike and to estimate the fraction of 14C converted to bound residue. Sample extracts and base traps were aliquoted into 15 mL Ecoscint A liquid scintillation cocktail and stored 24 h prior to liquid scintillation counting (LSC) analysis. The Tukey's Studentized Range Test was used to assess differences in 14C extractability and 14CO2 production with time and across treatments. Alpha was set to 0.05. 2.4. Sample radiological thin-layer-chromatography analysis (RAD-TLC) Extracts containing adequate radioactivity to allow radiological thinlayer-chromatography (RAD-TLC) analysis were concentrated to a known volume under nitrogen gas and spotted (30–60 µL) onto Whatman thin-layer chromatography plates (Partisil LK5D, Silica Gel 150 Å, 20x20 cm; Piscataway, NJ). Plates were developed in a sealed glass chamber containing 100 mL of 98:2 chloroform:methanol and read on an automated RAD-TLC scanner. Appearance of new chromatographic peaks with time would indicate formation of 14C–TCC degradation products. 2.5. Total carbon dioxide evolution analysis The initial sampling schedule prescribed amended-soil replicate removal at 0, 3, 6, 9, 12, 15, 18, and 21 weeks (T0–T7, respectively). However, because 14CO2 evolution rates and RAD-TLC extract analyses of initial samplings indicated slow 14C–TCC mineralization, sampling periods T5, T6, and T7 were extended to 16, 24, and 30 weeks, respectively. At week 19, the remaining samples received a second biosolids amendment of 0.10 g (d.w. equivalent), raising the effective amended-soil TCC concentration to ∼ 1 mg kg− 1. Biosolids were added to the samples to compensate for potential microbial die-off due to depletion of nutrients over the course of the study, and as an attempt to increase 14C–TCC biodegradation rates. Further, to confirm 14 C–TCC addition at study initiation did not inhibit microbial activity in the biotic samples, total CO2 (i.e. “cold” plus radioactive) evolution was determined for representative replicates (using an adaptation of Anderson, 1982). A solution of barium chloride was used to first precipitate the carbonates (representing trapped CO2) in a known volume of base trap, which was subsequently centrifuged at ∼2000 × g for 10 min. One mL of the supernatant was transferred to a glass 20 mL scintillation vial, treated with phenolphthalein indicator, and titrated to neutrality with 0.1 M HCl. The 1:1 relationship between
HCl required to achieve neutrality and the unused potassium hydroxide remaining in the base trap was used to calculate the moles of potassium hydroxide neutralized by evolved CO2. 3. Results and discussion 3.1. Spike recovery and formation of bound residues Biotic fine sand (BFS) and inhibited fine sand (IFS) total percent recoveries for T0–T7 averaged 91 ± 5% and 109 ± 6%, respectively. Percent recoveries in the biotic silty clay loam (BCL) and inhibited silty clay loam (ICL) for T0–T7 averaged 107 ± 8% and 109 ± 9%, respectively. There was no trend, either decreasing or increasing, in total percent recoveries over time in any treatment. Across sampling times, water and methanol extractability of 14C was significantly less in the BFS samples than in the IFS samples (Fig. 2a and b: normalized for 100% total spike recovery), suggesting a biotic effect on reductions in extractability over time. However, recoveries in the NaOH and combusted fractions were not statistically different between the two fine sand treatments. At every sampling, radioactivity recovered by water and methanol extractions was significantly greater in the fine sand than in the silty clay loam, and recovery in the combusted fraction was always greater in the silty clay loam than in the fine sand (both biotic and inhibited treatments). The difference is likely associated with the greater clay and OC content of the silty clay loam (clay: 330 g kg− 1, OC: 23 g kg− 1) as compared to the fine sand (clay: 10 g kg− 1, OC: 5.5 g kg− 1). With the exception of no significant difference between water extractability in the BCL and ICL samples, relationships between the recoveries in the sequential extracts of the BCL and ICL samples (Fig. 2c and d; normalized for 100% recovery) were similar to the relationships between the BFS and IFS samples. Water extractability in the BCL samples, although not statistically significant, appeared to be less than in the ICL samples. Water extractability of 14C in the BFS and BCL samples decreased significantly from T0 to T7 (14% to 3%, and 3% to ∼0.6%, respectively), and cumulative 14CO2 evolution continued to increase (Fig. 2a and c, respectively). Although no statistically significant changes in extractability occurred over time in the remaining fractions of any of the four treatments, methanol extractability tended to decrease, and the radioactivity in the combusted fraction tended to increase. 3.2. Mineralization Total 14CO2 production was significantly greater in the BCL samples than in the BFS samples (Fig. 2), indicating greater mineralization of the 14C–TCC in the biotic silty clay loam than in the biotic fine sand. Despite addition of biosolids to samples remaining after the T5 sampling period, no increase in 14CO2 production occurred in the BFS samples between T5 and T6, and only a small increase (from 2% to 4%) occurred in the BCL samples. The greatest 14CO2 production (prior to the second biosolids addition) occurred between T0 and T1 in the BCL samples and between T0 and T1 or T2 in the BFS samples. The total percent of added 14C recovered as 14CO2 increased slightly with subsequent samplings, but most mineralization occurred within the first 2 weeks of the experiment. All weekly base traps contained radioactivity on the order of fractions of a percent of total spiked, suggesting that 14C–TCC continued to slowly degrade. After 30 weeks of incubation (T7 samples), approximately 3% and 5% of applied 14C–TCC was captured in BFS and BCL base traps, respectively. Correcting for 14C–TCC captured by control base traps, approximately 2% (BFS) and 4% (BCL) of 14C–TCC was mineralized in 7.5 months. However, given the purity of the radiolabel (98.5%), cross-contamination of radioactivity into base traps connected to control (i.e. non-spiked) samples (25–50% of evolved 14 CO2), and the variation in total percent recoveries following sequential extraction, trends in degradation could be artifacts of
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
2729
Fig. 2. Radiolabeled triclocarban (14C–TCC) spike recoveries as a function of treatment, fraction (water, methanol, sodium hydroxide, 14-carbon dioxide, and combusted), and time (normalized for 100% total spike recovery).
measurement variability. In contrast, no mineralization occurred in the inhibited samples. 3.3. Impact on microbial respiration The circles in Fig. 3 indicate the first base trap sampled following the second biosolids addition at week 19. As expected, CO2 production
increased following supplementation with additional organic material. Biotic sample total CO2 evolution throughout the study was comparable to total CO2 evolution measured from poultry litter amended soils (∼ 1250 mg CO2 evolved g− 1 amendment in 90 days) (Khalil et al., 2005), but were less than rates measured from a biosolids-amended sandy clay loam (∼ 9000 mg CO2 evolved g− 1 amendment in 98 days) (Franco-Hernandez and Dendooven, 2006). The biosolids used in the Franco-Hernandez and Dendooven (2006) study, however, were processed in a bioreactor and treated with slake-lime (biosolids pH = 12). Lime treatment can solubilize organic matter and increase bioavailability, leading to increased mineralization rates (Kemmitt et al., 2006). Despite the addition of sodium azide to inhibited samples at T0 and again at the time of biosolids reapplication, microbial activity continued in the inhibited samples throughout the course of the study, albeit at a reduced rate compared to the biotic samples.
3.4. Transformation products
Fig. 3. Cumulative CO2 production by biosolids-amended-soil samples in the triclocarban biodegradation experiment.
14
C-
The methanol extracts were the only fractions containing adequate radioactivity to perform RAD-TLC analyses. Fig. 4a (bottom) illustrates a typical RAD-TLC chromatograph for the 14C–TCC standard solution. Fig. 4b (bottom) illustrates a typical RAD-TLC chromatograph for methanol extracts obtained from both soils during T0–T7 sampling periods. Chromatographs of methanol extracts of both soils and treatments (biotic and inhibited) from T0 to T7 were similar, suggesting no change in the composition of methanol extracts with time (i.e. no methanol-extractable degradation products formed between T0 and T7). Distinct regions of radioactivity above background are numbered in Fig. 4a (bottom) and Fig. 4b (bottom). Regions 7 and 9 in chromatographs from T0 to T7 methanol extracts were not as pronounced as Regions 7 and 9 in chromatographs from the 14C–TCC standard solution, suggesting that the radiolabel impurities were not methanol-extractable or their retention factor
2730
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
Fig. 5. Percent of radiolabel remaining (100%–%14CO2) in biotic biosolids-amended fine sand over time (identical data given on two scales to emphasize minimal degradation).
Fig. 4. Radio-thin-layer chromatography (RAD-TLC) of a 14C–TCC standard (a) and a biotic biosolids-amended-soil sample MeOH extract (b) representative of all T0–T7 chromatographs (bottom) and corresponding RAD-TLC “fingerprints” (top).
life in the biosolids-amended soils is problematic. The data collected herein might represent such early stages in the TCC degradation process that the true shape of the degradation curve cannot be accurately predicted. However, it is clear that the persistence of 14C– TCC in the biosolids-amended-soil degradation experiment is considerably greater than the 120-day soil half-life predicted using the QSAR analyses software EPA Persistent Bioaccumulative Toxic (PBT) Profiler (USEPA, 2009c) and the “weeks” to “months” predicted by BIOWIN in EPI Suite (which incorporates the same estimation software packages as PBT Profiler) (United States Environmental Protection Agency (USEPA), 2009d). The differences between the software-based and laboratory-based chemical behavior estimates illustrate the value of measured data for accurately estimating human and environmental exposure risks and regulatory impact. The persistence of biosolids-borne TCC in amended field soils might be less than anticipated if anaerobic degradation occurs in nonaerated micro- or macro-sites. The study described herein, however, scrubbed CO2 from the incoming air and could have prevented the conditions necessary for anaerobic degradation to proceed. Postulated anaerobic TCC degradation products dichlorocarbanilide (DCC), monochlorocarbanilide (MCC), and nonchlorinated carbanilide (NCC) were identified in sediment cores from a Chesapeake Bay tributary that receives WWTP effluent (Miller et al., 2008), but additional research using adapted OPPTS guidelines (i.e. 835.3400 Anaerobic Biodegradability of Organic Chemicals; 835.3420 Anaerobic Biodegradability of Organic Compounds in Digested Sludge; 835.5154 Anaerobic Biodegradation in the Subsurface) would be required to characterize the influence of anaerobic conditions on TCC loss in biosolids-amended soils.
was affected by the extract matrix. Low sensitivity, poor analyte separation, and/or presence of degradation products in the other extract fractions (albeit at low concentrations) could have contributed to the failure to identify 14C–TCC intermediates in the methanol extracts. 3.5. Persistence Figs. 5 and 6 illustrate the percent of radiolabel remaining in the biotic biosolids-amended-soil samples with time, and reflect the extremely slow mineralization of the 14C–TCC spike. No more than 4% of the spiked 14C–TCC mineralized in 7.5 months, suggesting ultimate degradation of TCC in land-applied biosolids will require several years. Because the 7.5 month degradation experiment did not reach the first 14C–TCC half-life, and the contribution of radiolabeled impurities to 14CO2 production was uncertain, predicting TCC half-
Fig. 6. Percent of radiolabel remaining (100%–%14CO2) in biotic biosolids-amended silty clay loam over time (identical data given on two scales to emphasize minimal degradation).
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
4. Conclusions The large amount of radioactivity extractable in the water and methanol fractions of biotic treatments, paired with minimal 14CO2 evolution, suggests three possible scenarios (none of which are mutually exclusive): (1) the soil microorganisms did not have the capability to degrade the radiolabel before bioavailability was reduced, (2) water and methanol are not good predictors of TCC bioavailability to soil microorganisms, or (3) the radiolabel spike killed or otherwise inhibited the potential degraders. Scenarios one and two depend on the conservative assumption that 14CO2 collected from the biotic treatments was a product of 14C–TCC mineralization (as opposed to mineralization of spike impurities). The small amount of 14CO2 collected from the biotic treatments suggests that some fraction of the radiolabeled parent compound was bioavailable. Thus, scenario one could be eliminated, though very little 14 CO2 evolved over the study period. However, decreasing bioavailability with time may have affected degradation, as the majority of cumulative evolved 14CO2 was collected in the first 3 weeks of the experiment. Scenario number two is only partially accurate. Water seemed to be a better predictor of bioavailability in the BCL samples than in the BFS samples. In the first 3 weeks in the biotic silty clay loam samples, the fraction of water extractable radioactivity decreased by approximately the same percentage as the amount mineralized. However, the water extractable fraction less accurately predicted mineralization in the BFS samples. Water extractability in the inhibited treatments for both soils remained largely unchanged over time, suggesting the decrease in water extractability in the biotic samples was microbiallymediated. The large percent recoveries and the absence of statistically significant decreases in extractability with time in the methanol fractions of both soils did not reflect the limited rates of radiolabel mineralization. Recoveries in the water and methanol fractions in both biotic soils were inconsistently correlated with 14CO2 evolution across sampling periods, thus preventing development of a mathematical model relating the two parameters (i.e. 14CO2 evolution and extractant recovery). Scenario number three is thought to be unlikely. The microorganisms found to be capable of degrading TCC in activated sludge are related to the genus Burkholderia (Miller et al., 2008), gram negative organisms that are not expected to be adversely affected by TCC at the concentrations present in the 14C–TCC biodegradation study. Further, the increasing cumulative CO2 evolution from biotic samples with time confirmed that the addition of TCC did not sterilize the system, nor did it depress total respiration. Sodium hydroxide solution extracted little of the radioactivity from the amended soils in all treatments, and extractability did not significantly change with time. The limited extractability of 14C–TCC by NaOH indicates only a small fraction of the radiolabel was associated with the humic fraction, though some of the radioactivity in the combusted fraction could have been organically bound. The apparent rapid irreversible sorption of 14C–TCC to the silty clay loam (∼30% in combusted fraction even at T0) is hypothesized to represent conversion of TCC in biosolids extractable by water and methanol to recalcitrant forms associated with the clay fraction in the amended soil. In all treatments, recovery by combustion appeared to increase with time, suggesting the formation of bound residues. Combustion and 14CO2 recoveries for the silty clay loam were consistently (and significantly) greater than for the fine sand, and suggested that greater amounts of bound residues were formed in the silty clay loam, but also that microbial communities in the silty clay loam might be better equipped to degrade the fraction of bioavailable 14C–TCC. Triclocarban appears to persist when applied to soils as a component of biosolids, with only 2–4% mineralization over the 7.5 month study. The results presented herein indicate biosolids-
2731
borne TCC becomes less bioavailable over time, which is expected to correspond with reduced uptake by biota and leaching through the soil profile. Additional research to characterize land-applied biosolidsborne TCC bioavailability, toxicity, and environmental mobility is needed to understand the implications of TCC persistence in amended-soil systems and facilitate an integrated human and ecological health risk assessment. Acknowledgments We thank Drs. Margaret James (Chair, Medicinal Chemistry Department, UF), Greg MacDonald (Agronomy Department, UF), John Thomas (Soil and Water Science Department, UF), and Robert Querns (Agronomy Department, UF) for generously providing work space and access to analytical equipment. We also thank Erin Schwab and Nina Itrich (The Procter & Gamble Company) for assistance in test apparatus design and radiological analysis techniques. The research was made possible through a grant from the USEPA Office of Wastewater Management (CP-83288701.9) and collaboration with The Procter & Gamble Company. References Alexander M. Biodegradation and bioremediation. San Diego, California: Academic Press; 1999. 168 pp. Anderson JP. Soil respiration. In: Page AL, Miller RH, Keeney DR, editors. Methods of soil analysis, part 2 — chemical and microbiological properties, Second ed. 1982. 842–843. Chu S, Metcalfe C. Simultaneous determination of triclocarban and triclosan in municipal biosolids by liquid chromatography tandem mass spectrometry. J Chromatogr 2007;1164:212–8. Franco-Hernandez O, Dendooven L. Dynamics of C, N and P in soil amended with biosolids from a pharmaceutical industry producing cephalosporines or third generation antibiotics: a laboratory study. Biores Technol 2006;97:1563–71. Gledhill WE. Biodegradation of 3, 4, 4′-trichlorocarbanilide, TCC, in sewage and activated sludge. Water Res 1975;9:649–54. Halden R. Emerging knowledge on emerging contaminants. Northeast Water Science Forum, 8 August–9 August 2007, Portland, Maine; 2007. http://www.neiwpcc.org/ ppcpconference/ppcpPresentations07.asp, accessed March 1, 2009. Halden RU, Paull DH. Co-occurrence of triclocarban and triclosan in U.W. water resources. Environ Sci Technol 2005;39:1420–6. Heidler J, Sapkota A, Halden R. Partitioning, persistence, and accumulation in digested sludge of the topical antiseptic triclocarban during wastewater treatment. Environ Sci Technol 2006;40:3634–9. Jussara R, Koskinen W, Sadowsky M. Influence of soil aging on sorption–desorption and bioavailability of simazine. J Agric Food Chem 2006;54:1373–9. Kelsey J, Kottler BD, Alexander M. Selective chemical extractants to predict bioavailability of soil-aged organic chemicals. Environ Sci Technol 1997;31:214–7. Kemmitt S, Wright D, Goulding K, Jones D. pH regulation of carbon and nitrogen dynamics in two agricultural soils. Soil Biol Biochem 2006;38:898–911. Khalil M, Hossain M, Schmidhalter U. Carbon and nitrogen mineralization in different upland soils of the subtropics treated with organic materials. Soil Biol Biochem 2005;37:1507–18. Lichstein H, Soule M. Studies of the effect of sodium azide on microbic growth and respiration. J Bacteriology 1944;47:221–30. Miller T, Heidler J, Chillrud S, DeLaquil A, Ritchie J, Mihalic J, et al. Fate of triclosan and triclocarban in estuarine sediment. Environ Sci Technol 2008;42:4570–6. Morton JB, Bentivenga SP, Wheeler WW. Germ plasm in the International Collection of Arbuscular and Vesicular–Arbuscular Mycorrhizal Fungi (INVAM) and procedures for culture development, documentation, and storage. Mycotaxon 1993;48: 491–528. National Research Council (NRC). Biosolids applied to land: advancing standards and practices. United States Environmental Protection Agency, Committee on Toxicants and Pathogens in Biosolids Applied to Land. Washington, D.C: National Academies Press; 2002. 13 pp. Sapkota A, Heidler J, Halden R. Detection of triclocarban and two co-contaminating chlorocarbanilides in US aquatic environments using isotope dilution liquid chromatography tandem mass spectrometry. Environ Res 2007;103:21–9. Schnitzer M. Organic matter characterization. In: Page AL, Miller RH, Keeney DR, editors. Methods of soil analysis, part 2 — chemical and microbiological properties. Second ed. Soil Science Society of America; 1982. p. 581–94. Seki K, Miyazaki T, Nakano M. Effects of microorganisms on hydraulic conductivity decrease infiltration. Eur J Soil Sci 1998;49:231–6. Snyder, E.H., O'Connor, G.A., McAvoy, D.C., Measured physicochemical characteristics and biosolids-borne concentrations of the antimicrobial triclocarban (TCC). Sci Tot Environ, in review. TCC Consortium. IUCLID data set, report 201-14186B, 2002. http://www.epa.gov/ chemrtk/tricloca/cl14186rs.pdf, accessed March 1, 2009.
2732
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
United States Environmental Protection Agency (USEPA). Product properties test guidelines, OPPTS 835.3300, soil biodegradation. Washington, D.C.: OPPTS; 1998. United States Environmental Protection Agency (USEPA). Emerging technologies for biosolids management; EPA 832-R-06-005. Washington, D.C: USEPA; 2006. United States Environmental Protection Agency (USEPA). Terms of environment: glossary, abbreviations and acronyms; 2009a. http://www.epa.gov/OCEPAterms/ bterms.html, accessed December 20, 2009. United States Environmental Protection Agency (USEPA). Targeted National Sewage Sludge Survey Statistical Analysis Report (“statistical report”, EPA-822-R-08-018). Washington, D.C: USEPA; 2009b. United States Environmental Protection Agency (USEPA). Persistent bioaccumulative toxic (PBT) profiler, v 1.203; 2009c. United States Environmental Protection Agency (USEPA). Estimation Programs Interface Suite™ for Microsoft® Windows, v 3.20. Washington, D.C., USA: United States Environmental Protection Agency; 2009d.
Walkley A, Black IA. An examination of Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Sci 1934;37:29–37. Wu C, Spongberg AL, Witter JD. Adsorption and degradation of triclosan and triclocarban in soils and biosolids-amended soils. J Agric Food Chem 2009;57: 4900–5. Xia K, Kwon J-W, Armbrust K. Transformation and extractability of antibacterial agents triclosan and triclocarban in soils and biosolids-amended soils. Proceedings from the 2008 Joint Annual Meeting, 5 October–9 October 2008, Houston, Texas; 2008. http://a-c-s.confex.com/crops/2008am/webprogram/Paper50897.html, accessed January 5, 2009. Ying G, Yu X, Kookana R. Biological degradation of triclocarban and triclosan in a soil under aerobic and anaerobic conditions and comparison with environmental fate modeling. Environ Pollut 2007;150:300–5.
Contents lists available at ScienceDirect
Science of the Total Environment j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i t o t e n v
Fate of
14
C–triclocarban in biosolids-amended soils
Elizabeth Hodges Snyder a,b,⁎, George A. O'Connor c, Drew C. McAvoy d a
Soil and Water Science Department, University of Florida, 408 Newell Hall, Gainesville, Florida, 32611, USA Department of Health Sciences, University of Alaska Anchorage, DPL 404, 3211 Providence Drive, Anchorage, AK 99508-4614, USA c Soil and Water Science Department, P.O. Box 110510, University of Florida, Gainesville, FL 32611-01519, USA d Environmental Safety Department, P.O. Box 538707, The Procter & Gamble Company, Cincinnati, OH, 45253-8707, USA b
a r t i c l e
i n f o
Article history: Received 14 September 2009 Received in revised form 24 December 2009 Accepted 5 January 2010 Available online 2 April 2010 Keywords: Triclocarban Antibacterial Biosolids Sludge Biodegradation
a b s t r a c t Triclocarban (TCC) is an antibacterial compound commonly detected in biosolids at parts-per-million concentrations. Approximately half of the biosolids produced in the United States are land-applied, resulting in a systematic release of TCC into the soil environment. The extent of biosolids-borne TCC environmental transport and potential human/ecological exposures will be greatly affected by its bioavailability and the rate of degradation in amended soils. To investigate these factors, radiolabeled TCC (14C–TCC) was incorporated into anaerobically digested biosolids, amended to two soils, and incubated under aerobic conditions. The evolution of 14CO2 (biodegradation) and changes in chemical extractability (bioavailability) was measured over time. Water extractable TCC over the study period was low and significantly decreased over the first 3 weeks of the study (from 14% to 4% in a fine sand soil and from 3 to < 1% in a silty clay loam soil). Mineralization (i.e. ultimate degradation), as measured by evolution of 14CO2, was < 4% over 7.5 months. Methanol extracts of the amended soils were analyzed by radiolabel thin-layer chromatography (RAD-TLC), but no intermediate degradation products were detected. Approximately 20% and 50% of the radioactivity in the amended fine sand and silty clay loam soils, respectively, was converted to bound residue as measured by solids combustion. These results indicate that biosolids-borne TCC becomes less bioavailable over time and biodegrades at a very slow rate. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Triclocarban (Fig. 1) is a common constituent of domestic wastewater due primarily to its use as an active ingredient in antibacterial bar soaps. Up to 98% of TCC present in wastewater influent is removed from the aqueous phase by activated sludge treatment (TCC Consortium, 2002; Heidler et al., 2006), and approximately 75% is concentrated in the solid fraction (i.e. sludge) (Heidler et al., 2006). Sludge is often converted, via a variety of methods to meet regulations in the US EPA 40CFR503 (Standards for the Use or Disposal of Sewage Sludge), to biosolids approved for land application. The USEPA estimates 8 × 106 dry metric tons of biosolids are produced annually (USEPA, 2006), and approximately half are land-applied (NRC, 2002). Documented biosolids-borne TCC concentrations range from 5 to 51 mg kg− 1 (Heidler et al., 2006; Chu and Metcalfe, 2007; Halden, 2007; Sapkota et al., 2007), with an apparent mean of approximately 20 mg kg− 1. The recently published Targeted National Sewage Sludge Survey (TNSSS)
⁎ Corresponding author. Department of Health Sciences, University of Alaska Anchorage, DPL 404, 3211 Providence Drive, Anchorage, AK 99508-4614, USA. Tel.: +1 907 786 6540; fax: +1 907 786 6572. E-mail addresses: lizah@ufl.edu, [email protected] (E.H. Snyder), gao@ufl.edu (G.A. O'Connor), [email protected] (D.C. McAvoy). 0048-9697/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2010.01.005
quantified TCC concentrations up to 441 mg kg− 1 in wastewater treatment sludge and/or biosolids (mean and standard deviation: 39± 60 mg kg− 1) (USEPA, 2009a). Little is known about the persistence and bioavailability (i.e. the degree of ability to be absorbed and ready to interact in organism metabolism; USEPA, 2009b) of TCC in the terrestrial environment following land application of biosolids. Efficient biodegradation of any organic contaminant requires that the compound be bioavailable to organisms capable of degradation. The limited solubility (0.045 mg L− 1) and moderate log Kow (3.5) (Snyder et al., in review) suggests TCC will preferentially partition to the biosolids matrix, potentially limiting bioavailability. Migration into or onto other soil fractions following land application of biosolids could subsequently alter bioavailability over time. Any process that acts to lower the effective solution concentration of a soil contaminant could be expected to affect the rate of biodegradation (Alexander, 1999) and, potentially, the effective toxicity of the contaminant. Few studies have characterized TCC degradation in the combined environmental matrices of wastewater treatment products and soil. These studies are generally limited to sludge-, soil-, or sediment-only systems (Gledhill, 1975; Ying et al., 2007; Miller et al., 2008), do not attempt to isolate aerobic biodegradation intermediates (Xia et al., 2008; Wu et al., 2009), or quantify degradation as loss of extractable parent compound over time (Ying et al., 2007; Xia et al., 2008; Wu et al., 2009), which could be attributable to the formation of bound
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
2727
Fig. 1. Chemical structure and select physicochemical properties of triclocarban (TCC; N-(4-chlorophenyl)-N′-(3,4-dichlorophenyl) urea).
residues rather than actual primary and/or ultimate degradation. Aerobic degradation products identified in a bench-top continuous flow activated sludge (CFAS) system included 3,4-dichloroaniline (DCA), and to a lesser extent, p-chloroaniline (PCA), aniline, and TCC condensation products (i.e. 4-chloro-4′-(4-chloroaniline)-azobenzene and 4,4′-dichloro-azobenzene) (Gledhill, 1975). The aforementioned studies provide useful information for the assessment of TCC environmental fate and transport, but multiple gaps in our understanding of land-applied biosolids-borne TCC fate remain. In particular, better information is needed on bioavailability, biodegradation, and metabolite formation in soils. Further, the sorption of TCC to organic matter and the potential impacts of partitioning on degradation necessitate the inclusion of biosolids as the source of TCC in an assessment of its fate following introduction into the terrestrial environment. Therefore, we characterized biodegradation of TCC using a respirometry test, in which the microbial metabolism of 14C-labeled TCC was quantified by monitoring the evolution of 14CO2 from spiked biosolids amended to soil. The test was also coupled with methods to isolate potential degradation products, and to characterize changes in chemical extractability (which can be related to changes inbioavailability) over the study period. Results were used to characterize the influence of soil type on biosolids-borne TCC persistence and extractability. This paper is one in a planned series of papers characterizing multiple components of a human and ecological risk assessment of land-applied, biosolids-borne TCC.
2. Materials and methods 2.1. Chemicals, biosolids, and soils 14 C–triclocarban labeled on the 4-chloroaniline ring [specific activity: 75 mCi/mmol; purity: 98.5%], which is expected to be the most rapidly degraded of the two rings (Gledhill, 1975), was synthesized by Amersham Life Science and supplied by the Procter & Gamble Company. Barium chloride (BaCl2), chloroform (CHCl3), hydrochloric acid (HCl), methanol (MeOH), sodium hydroxide (NaOH), and sodium azide (NaN3) were purchased from Fisher Scientific. EcoScint A liquid scintillation cocktail was purchased from National Diagnostics (Atlanta, GA). Anaerobically digested, Class B dewatered cake biosolids (identification code: CFBC; indigenous TCC concentration: 40 mg kg− 1) was collected from the belt-press of a wastewater treatment plant (WWTP) serving primarily single family homes and a dye factory in Ohio. The biosolids-borne TCC concentration was previously determined using extraction and analysis methodology adapted from Halden and Paull (2005) and Chu and Metcalfe (2007) (Snyder et al., in review). Samples of two soils, a Florida Immokalee fine sand (sandy, siliceous, hyperthermic Arenic Alaquods) (clay: 10 g kg− 1, OC: 5.5 g kg− 1) and an Ohio Genesee silty clay loam (fine-loamy, mixed, superactive, mesic Fluventic Eutrudepts) (clay: 330 g kg− 1, OC: 23 g kg− 1) (Walkley and Black, 1934)
were collected from sites with no known history of receiving landapplied biosolids or sludge. 2.2. Study design The study design was adapted from the USEPA Office of Prevention, Pesticides, and Toxic Substances (OPPTS) harmonized test guideline Soil Biodegradation (835.3300)1 (USEPA, 1998). The OPPTS Guidelines were developed through blending USEPA Office of Pollution Prevention and Toxics (OPPT), USEPA Office of Pesticide Programs (OPP), and Organization for Economic Cooperation and Development (OECD) test guidance to produce standardized methods for generating a uniform and consistent chemical properties database. Glass, round-bottom, 30 mL centrifuge tubes were used to prepare 140 samples [2 soils×1 biosolids×1 rate×4 replicates×8 sampling periods×2 treatments (biotic/inhibited)+12 controls]. Biosolids samples (0.10 g d.w. equivalent) were loaded into the centrifuge tubes, spiked drop-wise with 1.3 × 106 dpm 14C-TCC g biosolids− 1 (d.w. equivalent) in methanol, allowed to equilibrate 24 h, and amended with 10 g (d.w. equivalent) of soil to simulate a realistic field loading rate (22 Mg ha− 1 or ∼10 tons acre− 1). The loaded centrifuge tubes were capped and vortexed for 30 s to mix the biosolids and soil, and avoid potential removal of radioactivity through use of a stirring rod. The radioisotope spike increased the effective biosolids-borne TCC concentration to ∼65 mg TCC kg− 1, and the total TCC load in amended soil was ∼0.65 mg TCC kg amended soil− 1. The inhibited treatment was included in the study design to facilitate differentiation between biotically-induced and non-biotically-induced effects on TCC degradation and extractability with time. The inhibited samples were prepared by vortexing 1000 μg g− 1 sodium azide (0.1%) in methanol carrier solvent into a subset of the biosolids-amended soils. Sodium azide is effective against bacteria (Lichstein and Soule, 1944) and fungi (Morton et al., 1993; Seki et al., 1998), but the term “inhibited,” rather than “abiotic,” is used to describe the treatment because of the differential effects of sodium azide on gram negative and gram positive microorganisms. The chemical primarily acts as a bacteriostatic agent against gram negative microoorganisms, and is less effective against gram positives (Lichstein and Soule, 1944). Each centrifuge tube was fitted with a silicon/Teflon septum and connected, via Dow Corning Silastic laboratory tubing, to a series of glass trapping vials containing 5 mL 0.2 M potassium hydroxide. The base traps were used to collect evolved 14CO2 and CO2 as a measure of 14 C–TCC mineralization and soil microbial respiration, respectively. The first positioned vial remained empty to prevent backflow into the centrifuge tubes if pump failure were to occur. An oil-less pump 1 Additional relevant guidelines for assessment of aerobic and anaerobic biodegradation have since been published, including Anaerobic Biodegradability of Organic Compounds in Digested Sludge: By Measurement of Gas Production (835.3420), Aerobic Soil Metabolism–Anaerobic Soil Metabolism (835.4100-835.4200).
2728
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
aerated samples with humidified, CO2-free air by first pumping air through 1 M potassium hydroxide, and then deionized water. Although not prescribed by the OPPTS guideline, CO2 was removed from incoming air to facilitate calculations of total CO2 evolved from the amended-soil samples (as a measure of biotic activity). The experiment was conducted in the dark at ∼ 23 °C. Samples were weighed weekly and sterile water was added, via syringe needle through the tube septum as needed, to maintain a soil moisture content equivalent to field capacity [i.e. 10% w/w (fine sand) and 20% w/w (silty clay loam)]. 2.3. Sample extraction and liquid scintillation counting (LSC) analysis Once a week, position-1 base traps were removed, remaining traps were moved forward, and a fresh trap was added to the newly open position-3. Four replicates of each biosolids-amended-soil treatment were periodically removed and sequentially extracted (sonication, 1 h) with 20 mL of deionized water (twice), 20 mL of methanol (twice), and 20 mL of 1 M NaOH (once). The water and methanol extractions served as estimates of TCC bioavailability in biosolidsamended soils (Kelsey et al., 1997; Jussara et al., 2006), and the NaOH extraction estimated the extent of TCC incorporation into the humic fraction (Schnitzer, 1982). A subsample of each sequentially extracted amended-soil replicate was removed, weighed, and combusted on a sample oxidizer to complete a mass balance of the 14C–TCC spike and to estimate the fraction of 14C converted to bound residue. Sample extracts and base traps were aliquoted into 15 mL Ecoscint A liquid scintillation cocktail and stored 24 h prior to liquid scintillation counting (LSC) analysis. The Tukey's Studentized Range Test was used to assess differences in 14C extractability and 14CO2 production with time and across treatments. Alpha was set to 0.05. 2.4. Sample radiological thin-layer-chromatography analysis (RAD-TLC) Extracts containing adequate radioactivity to allow radiological thinlayer-chromatography (RAD-TLC) analysis were concentrated to a known volume under nitrogen gas and spotted (30–60 µL) onto Whatman thin-layer chromatography plates (Partisil LK5D, Silica Gel 150 Å, 20x20 cm; Piscataway, NJ). Plates were developed in a sealed glass chamber containing 100 mL of 98:2 chloroform:methanol and read on an automated RAD-TLC scanner. Appearance of new chromatographic peaks with time would indicate formation of 14C–TCC degradation products. 2.5. Total carbon dioxide evolution analysis The initial sampling schedule prescribed amended-soil replicate removal at 0, 3, 6, 9, 12, 15, 18, and 21 weeks (T0–T7, respectively). However, because 14CO2 evolution rates and RAD-TLC extract analyses of initial samplings indicated slow 14C–TCC mineralization, sampling periods T5, T6, and T7 were extended to 16, 24, and 30 weeks, respectively. At week 19, the remaining samples received a second biosolids amendment of 0.10 g (d.w. equivalent), raising the effective amended-soil TCC concentration to ∼ 1 mg kg− 1. Biosolids were added to the samples to compensate for potential microbial die-off due to depletion of nutrients over the course of the study, and as an attempt to increase 14C–TCC biodegradation rates. Further, to confirm 14 C–TCC addition at study initiation did not inhibit microbial activity in the biotic samples, total CO2 (i.e. “cold” plus radioactive) evolution was determined for representative replicates (using an adaptation of Anderson, 1982). A solution of barium chloride was used to first precipitate the carbonates (representing trapped CO2) in a known volume of base trap, which was subsequently centrifuged at ∼2000 × g for 10 min. One mL of the supernatant was transferred to a glass 20 mL scintillation vial, treated with phenolphthalein indicator, and titrated to neutrality with 0.1 M HCl. The 1:1 relationship between
HCl required to achieve neutrality and the unused potassium hydroxide remaining in the base trap was used to calculate the moles of potassium hydroxide neutralized by evolved CO2. 3. Results and discussion 3.1. Spike recovery and formation of bound residues Biotic fine sand (BFS) and inhibited fine sand (IFS) total percent recoveries for T0–T7 averaged 91 ± 5% and 109 ± 6%, respectively. Percent recoveries in the biotic silty clay loam (BCL) and inhibited silty clay loam (ICL) for T0–T7 averaged 107 ± 8% and 109 ± 9%, respectively. There was no trend, either decreasing or increasing, in total percent recoveries over time in any treatment. Across sampling times, water and methanol extractability of 14C was significantly less in the BFS samples than in the IFS samples (Fig. 2a and b: normalized for 100% total spike recovery), suggesting a biotic effect on reductions in extractability over time. However, recoveries in the NaOH and combusted fractions were not statistically different between the two fine sand treatments. At every sampling, radioactivity recovered by water and methanol extractions was significantly greater in the fine sand than in the silty clay loam, and recovery in the combusted fraction was always greater in the silty clay loam than in the fine sand (both biotic and inhibited treatments). The difference is likely associated with the greater clay and OC content of the silty clay loam (clay: 330 g kg− 1, OC: 23 g kg− 1) as compared to the fine sand (clay: 10 g kg− 1, OC: 5.5 g kg− 1). With the exception of no significant difference between water extractability in the BCL and ICL samples, relationships between the recoveries in the sequential extracts of the BCL and ICL samples (Fig. 2c and d; normalized for 100% recovery) were similar to the relationships between the BFS and IFS samples. Water extractability in the BCL samples, although not statistically significant, appeared to be less than in the ICL samples. Water extractability of 14C in the BFS and BCL samples decreased significantly from T0 to T7 (14% to 3%, and 3% to ∼0.6%, respectively), and cumulative 14CO2 evolution continued to increase (Fig. 2a and c, respectively). Although no statistically significant changes in extractability occurred over time in the remaining fractions of any of the four treatments, methanol extractability tended to decrease, and the radioactivity in the combusted fraction tended to increase. 3.2. Mineralization Total 14CO2 production was significantly greater in the BCL samples than in the BFS samples (Fig. 2), indicating greater mineralization of the 14C–TCC in the biotic silty clay loam than in the biotic fine sand. Despite addition of biosolids to samples remaining after the T5 sampling period, no increase in 14CO2 production occurred in the BFS samples between T5 and T6, and only a small increase (from 2% to 4%) occurred in the BCL samples. The greatest 14CO2 production (prior to the second biosolids addition) occurred between T0 and T1 in the BCL samples and between T0 and T1 or T2 in the BFS samples. The total percent of added 14C recovered as 14CO2 increased slightly with subsequent samplings, but most mineralization occurred within the first 2 weeks of the experiment. All weekly base traps contained radioactivity on the order of fractions of a percent of total spiked, suggesting that 14C–TCC continued to slowly degrade. After 30 weeks of incubation (T7 samples), approximately 3% and 5% of applied 14C–TCC was captured in BFS and BCL base traps, respectively. Correcting for 14C–TCC captured by control base traps, approximately 2% (BFS) and 4% (BCL) of 14C–TCC was mineralized in 7.5 months. However, given the purity of the radiolabel (98.5%), cross-contamination of radioactivity into base traps connected to control (i.e. non-spiked) samples (25–50% of evolved 14 CO2), and the variation in total percent recoveries following sequential extraction, trends in degradation could be artifacts of
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
2729
Fig. 2. Radiolabeled triclocarban (14C–TCC) spike recoveries as a function of treatment, fraction (water, methanol, sodium hydroxide, 14-carbon dioxide, and combusted), and time (normalized for 100% total spike recovery).
measurement variability. In contrast, no mineralization occurred in the inhibited samples. 3.3. Impact on microbial respiration The circles in Fig. 3 indicate the first base trap sampled following the second biosolids addition at week 19. As expected, CO2 production
increased following supplementation with additional organic material. Biotic sample total CO2 evolution throughout the study was comparable to total CO2 evolution measured from poultry litter amended soils (∼ 1250 mg CO2 evolved g− 1 amendment in 90 days) (Khalil et al., 2005), but were less than rates measured from a biosolids-amended sandy clay loam (∼ 9000 mg CO2 evolved g− 1 amendment in 98 days) (Franco-Hernandez and Dendooven, 2006). The biosolids used in the Franco-Hernandez and Dendooven (2006) study, however, were processed in a bioreactor and treated with slake-lime (biosolids pH = 12). Lime treatment can solubilize organic matter and increase bioavailability, leading to increased mineralization rates (Kemmitt et al., 2006). Despite the addition of sodium azide to inhibited samples at T0 and again at the time of biosolids reapplication, microbial activity continued in the inhibited samples throughout the course of the study, albeit at a reduced rate compared to the biotic samples.
3.4. Transformation products
Fig. 3. Cumulative CO2 production by biosolids-amended-soil samples in the triclocarban biodegradation experiment.
14
C-
The methanol extracts were the only fractions containing adequate radioactivity to perform RAD-TLC analyses. Fig. 4a (bottom) illustrates a typical RAD-TLC chromatograph for the 14C–TCC standard solution. Fig. 4b (bottom) illustrates a typical RAD-TLC chromatograph for methanol extracts obtained from both soils during T0–T7 sampling periods. Chromatographs of methanol extracts of both soils and treatments (biotic and inhibited) from T0 to T7 were similar, suggesting no change in the composition of methanol extracts with time (i.e. no methanol-extractable degradation products formed between T0 and T7). Distinct regions of radioactivity above background are numbered in Fig. 4a (bottom) and Fig. 4b (bottom). Regions 7 and 9 in chromatographs from T0 to T7 methanol extracts were not as pronounced as Regions 7 and 9 in chromatographs from the 14C–TCC standard solution, suggesting that the radiolabel impurities were not methanol-extractable or their retention factor
2730
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
Fig. 5. Percent of radiolabel remaining (100%–%14CO2) in biotic biosolids-amended fine sand over time (identical data given on two scales to emphasize minimal degradation).
Fig. 4. Radio-thin-layer chromatography (RAD-TLC) of a 14C–TCC standard (a) and a biotic biosolids-amended-soil sample MeOH extract (b) representative of all T0–T7 chromatographs (bottom) and corresponding RAD-TLC “fingerprints” (top).
life in the biosolids-amended soils is problematic. The data collected herein might represent such early stages in the TCC degradation process that the true shape of the degradation curve cannot be accurately predicted. However, it is clear that the persistence of 14C– TCC in the biosolids-amended-soil degradation experiment is considerably greater than the 120-day soil half-life predicted using the QSAR analyses software EPA Persistent Bioaccumulative Toxic (PBT) Profiler (USEPA, 2009c) and the “weeks” to “months” predicted by BIOWIN in EPI Suite (which incorporates the same estimation software packages as PBT Profiler) (United States Environmental Protection Agency (USEPA), 2009d). The differences between the software-based and laboratory-based chemical behavior estimates illustrate the value of measured data for accurately estimating human and environmental exposure risks and regulatory impact. The persistence of biosolids-borne TCC in amended field soils might be less than anticipated if anaerobic degradation occurs in nonaerated micro- or macro-sites. The study described herein, however, scrubbed CO2 from the incoming air and could have prevented the conditions necessary for anaerobic degradation to proceed. Postulated anaerobic TCC degradation products dichlorocarbanilide (DCC), monochlorocarbanilide (MCC), and nonchlorinated carbanilide (NCC) were identified in sediment cores from a Chesapeake Bay tributary that receives WWTP effluent (Miller et al., 2008), but additional research using adapted OPPTS guidelines (i.e. 835.3400 Anaerobic Biodegradability of Organic Chemicals; 835.3420 Anaerobic Biodegradability of Organic Compounds in Digested Sludge; 835.5154 Anaerobic Biodegradation in the Subsurface) would be required to characterize the influence of anaerobic conditions on TCC loss in biosolids-amended soils.
was affected by the extract matrix. Low sensitivity, poor analyte separation, and/or presence of degradation products in the other extract fractions (albeit at low concentrations) could have contributed to the failure to identify 14C–TCC intermediates in the methanol extracts. 3.5. Persistence Figs. 5 and 6 illustrate the percent of radiolabel remaining in the biotic biosolids-amended-soil samples with time, and reflect the extremely slow mineralization of the 14C–TCC spike. No more than 4% of the spiked 14C–TCC mineralized in 7.5 months, suggesting ultimate degradation of TCC in land-applied biosolids will require several years. Because the 7.5 month degradation experiment did not reach the first 14C–TCC half-life, and the contribution of radiolabeled impurities to 14CO2 production was uncertain, predicting TCC half-
Fig. 6. Percent of radiolabel remaining (100%–%14CO2) in biotic biosolids-amended silty clay loam over time (identical data given on two scales to emphasize minimal degradation).
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
4. Conclusions The large amount of radioactivity extractable in the water and methanol fractions of biotic treatments, paired with minimal 14CO2 evolution, suggests three possible scenarios (none of which are mutually exclusive): (1) the soil microorganisms did not have the capability to degrade the radiolabel before bioavailability was reduced, (2) water and methanol are not good predictors of TCC bioavailability to soil microorganisms, or (3) the radiolabel spike killed or otherwise inhibited the potential degraders. Scenarios one and two depend on the conservative assumption that 14CO2 collected from the biotic treatments was a product of 14C–TCC mineralization (as opposed to mineralization of spike impurities). The small amount of 14CO2 collected from the biotic treatments suggests that some fraction of the radiolabeled parent compound was bioavailable. Thus, scenario one could be eliminated, though very little 14 CO2 evolved over the study period. However, decreasing bioavailability with time may have affected degradation, as the majority of cumulative evolved 14CO2 was collected in the first 3 weeks of the experiment. Scenario number two is only partially accurate. Water seemed to be a better predictor of bioavailability in the BCL samples than in the BFS samples. In the first 3 weeks in the biotic silty clay loam samples, the fraction of water extractable radioactivity decreased by approximately the same percentage as the amount mineralized. However, the water extractable fraction less accurately predicted mineralization in the BFS samples. Water extractability in the inhibited treatments for both soils remained largely unchanged over time, suggesting the decrease in water extractability in the biotic samples was microbiallymediated. The large percent recoveries and the absence of statistically significant decreases in extractability with time in the methanol fractions of both soils did not reflect the limited rates of radiolabel mineralization. Recoveries in the water and methanol fractions in both biotic soils were inconsistently correlated with 14CO2 evolution across sampling periods, thus preventing development of a mathematical model relating the two parameters (i.e. 14CO2 evolution and extractant recovery). Scenario number three is thought to be unlikely. The microorganisms found to be capable of degrading TCC in activated sludge are related to the genus Burkholderia (Miller et al., 2008), gram negative organisms that are not expected to be adversely affected by TCC at the concentrations present in the 14C–TCC biodegradation study. Further, the increasing cumulative CO2 evolution from biotic samples with time confirmed that the addition of TCC did not sterilize the system, nor did it depress total respiration. Sodium hydroxide solution extracted little of the radioactivity from the amended soils in all treatments, and extractability did not significantly change with time. The limited extractability of 14C–TCC by NaOH indicates only a small fraction of the radiolabel was associated with the humic fraction, though some of the radioactivity in the combusted fraction could have been organically bound. The apparent rapid irreversible sorption of 14C–TCC to the silty clay loam (∼30% in combusted fraction even at T0) is hypothesized to represent conversion of TCC in biosolids extractable by water and methanol to recalcitrant forms associated with the clay fraction in the amended soil. In all treatments, recovery by combustion appeared to increase with time, suggesting the formation of bound residues. Combustion and 14CO2 recoveries for the silty clay loam were consistently (and significantly) greater than for the fine sand, and suggested that greater amounts of bound residues were formed in the silty clay loam, but also that microbial communities in the silty clay loam might be better equipped to degrade the fraction of bioavailable 14C–TCC. Triclocarban appears to persist when applied to soils as a component of biosolids, with only 2–4% mineralization over the 7.5 month study. The results presented herein indicate biosolids-
2731
borne TCC becomes less bioavailable over time, which is expected to correspond with reduced uptake by biota and leaching through the soil profile. Additional research to characterize land-applied biosolidsborne TCC bioavailability, toxicity, and environmental mobility is needed to understand the implications of TCC persistence in amended-soil systems and facilitate an integrated human and ecological health risk assessment. Acknowledgments We thank Drs. Margaret James (Chair, Medicinal Chemistry Department, UF), Greg MacDonald (Agronomy Department, UF), John Thomas (Soil and Water Science Department, UF), and Robert Querns (Agronomy Department, UF) for generously providing work space and access to analytical equipment. We also thank Erin Schwab and Nina Itrich (The Procter & Gamble Company) for assistance in test apparatus design and radiological analysis techniques. The research was made possible through a grant from the USEPA Office of Wastewater Management (CP-83288701.9) and collaboration with The Procter & Gamble Company. References Alexander M. Biodegradation and bioremediation. San Diego, California: Academic Press; 1999. 168 pp. Anderson JP. Soil respiration. In: Page AL, Miller RH, Keeney DR, editors. Methods of soil analysis, part 2 — chemical and microbiological properties, Second ed. 1982. 842–843. Chu S, Metcalfe C. Simultaneous determination of triclocarban and triclosan in municipal biosolids by liquid chromatography tandem mass spectrometry. J Chromatogr 2007;1164:212–8. Franco-Hernandez O, Dendooven L. Dynamics of C, N and P in soil amended with biosolids from a pharmaceutical industry producing cephalosporines or third generation antibiotics: a laboratory study. Biores Technol 2006;97:1563–71. Gledhill WE. Biodegradation of 3, 4, 4′-trichlorocarbanilide, TCC, in sewage and activated sludge. Water Res 1975;9:649–54. Halden R. Emerging knowledge on emerging contaminants. Northeast Water Science Forum, 8 August–9 August 2007, Portland, Maine; 2007. http://www.neiwpcc.org/ ppcpconference/ppcpPresentations07.asp, accessed March 1, 2009. Halden RU, Paull DH. Co-occurrence of triclocarban and triclosan in U.W. water resources. Environ Sci Technol 2005;39:1420–6. Heidler J, Sapkota A, Halden R. Partitioning, persistence, and accumulation in digested sludge of the topical antiseptic triclocarban during wastewater treatment. Environ Sci Technol 2006;40:3634–9. Jussara R, Koskinen W, Sadowsky M. Influence of soil aging on sorption–desorption and bioavailability of simazine. J Agric Food Chem 2006;54:1373–9. Kelsey J, Kottler BD, Alexander M. Selective chemical extractants to predict bioavailability of soil-aged organic chemicals. Environ Sci Technol 1997;31:214–7. Kemmitt S, Wright D, Goulding K, Jones D. pH regulation of carbon and nitrogen dynamics in two agricultural soils. Soil Biol Biochem 2006;38:898–911. Khalil M, Hossain M, Schmidhalter U. Carbon and nitrogen mineralization in different upland soils of the subtropics treated with organic materials. Soil Biol Biochem 2005;37:1507–18. Lichstein H, Soule M. Studies of the effect of sodium azide on microbic growth and respiration. J Bacteriology 1944;47:221–30. Miller T, Heidler J, Chillrud S, DeLaquil A, Ritchie J, Mihalic J, et al. Fate of triclosan and triclocarban in estuarine sediment. Environ Sci Technol 2008;42:4570–6. Morton JB, Bentivenga SP, Wheeler WW. Germ plasm in the International Collection of Arbuscular and Vesicular–Arbuscular Mycorrhizal Fungi (INVAM) and procedures for culture development, documentation, and storage. Mycotaxon 1993;48: 491–528. National Research Council (NRC). Biosolids applied to land: advancing standards and practices. United States Environmental Protection Agency, Committee on Toxicants and Pathogens in Biosolids Applied to Land. Washington, D.C: National Academies Press; 2002. 13 pp. Sapkota A, Heidler J, Halden R. Detection of triclocarban and two co-contaminating chlorocarbanilides in US aquatic environments using isotope dilution liquid chromatography tandem mass spectrometry. Environ Res 2007;103:21–9. Schnitzer M. Organic matter characterization. In: Page AL, Miller RH, Keeney DR, editors. Methods of soil analysis, part 2 — chemical and microbiological properties. Second ed. Soil Science Society of America; 1982. p. 581–94. Seki K, Miyazaki T, Nakano M. Effects of microorganisms on hydraulic conductivity decrease infiltration. Eur J Soil Sci 1998;49:231–6. Snyder, E.H., O'Connor, G.A., McAvoy, D.C., Measured physicochemical characteristics and biosolids-borne concentrations of the antimicrobial triclocarban (TCC). Sci Tot Environ, in review. TCC Consortium. IUCLID data set, report 201-14186B, 2002. http://www.epa.gov/ chemrtk/tricloca/cl14186rs.pdf, accessed March 1, 2009.
2732
E.H. Snyder et al. / Science of the Total Environment 408 (2010) 2726–2732
United States Environmental Protection Agency (USEPA). Product properties test guidelines, OPPTS 835.3300, soil biodegradation. Washington, D.C.: OPPTS; 1998. United States Environmental Protection Agency (USEPA). Emerging technologies for biosolids management; EPA 832-R-06-005. Washington, D.C: USEPA; 2006. United States Environmental Protection Agency (USEPA). Terms of environment: glossary, abbreviations and acronyms; 2009a. http://www.epa.gov/OCEPAterms/ bterms.html, accessed December 20, 2009. United States Environmental Protection Agency (USEPA). Targeted National Sewage Sludge Survey Statistical Analysis Report (“statistical report”, EPA-822-R-08-018). Washington, D.C: USEPA; 2009b. United States Environmental Protection Agency (USEPA). Persistent bioaccumulative toxic (PBT) profiler, v 1.203; 2009c. United States Environmental Protection Agency (USEPA). Estimation Programs Interface Suite™ for Microsoft® Windows, v 3.20. Washington, D.C., USA: United States Environmental Protection Agency; 2009d.
Walkley A, Black IA. An examination of Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Sci 1934;37:29–37. Wu C, Spongberg AL, Witter JD. Adsorption and degradation of triclosan and triclocarban in soils and biosolids-amended soils. J Agric Food Chem 2009;57: 4900–5. Xia K, Kwon J-W, Armbrust K. Transformation and extractability of antibacterial agents triclosan and triclocarban in soils and biosolids-amended soils. Proceedings from the 2008 Joint Annual Meeting, 5 October–9 October 2008, Houston, Texas; 2008. http://a-c-s.confex.com/crops/2008am/webprogram/Paper50897.html, accessed January 5, 2009. Ying G, Yu X, Kookana R. Biological degradation of triclocarban and triclosan in a soil under aerobic and anaerobic conditions and comparison with environmental fate modeling. Environ Pollut 2007;150:300–5.