Cyto- and geno-toxicity of 1,4-dioxane and its transformation products during ultraviolet-driven advanced oxidation processes

Wei Li ab, Elvis Xu c, Daniel Schlenk bc and Haizhou Liu *ab
aDepartment of Chemical and Environmental Engineering, University of California, Riverside, CA 92521, USA. E-mail: haizhou@engr.ucr.edu; Fax: +1 951 827 5696; Tel: +1 951 827 2076
bProgram of Environmental Toxicology, University of California, Riverside, CA 92521, USA
cDepartment of Environmental Sciences, University of California, Riverside, CA 92521, USA

Received 18th February 2018 , Accepted 24th May 2018

First published on 31st May 2018


Ultraviolet-driven advanced oxidation processes (UV/AOPs) are integral steps in water reuse treatment trains. The toxicity of trace organic transformation products during a UV/AOP is critical to its implementation. This study examined the cyto- and geno-toxicity of transformation products of 1,4-dioxane (1,4-D), a trace organic contaminant commonly found in secondary wastewater, in extracts using the CellSensor p53RE-bla HCt-116 cell assay, following UV photolysis at 254 nm with three oxidants, hydrogen peroxide (H2O2), persulfate (S2O82−) and monochloramine (NH2Cl). 1,4-D was transformed into six major oxidation by-products, including ethylene glycol diformate, formaldehyde, glycolaldehyde, glycolic acid, formic acid, and methoxyacetic acid. Formaldehyde and glycolaldehyde were the most geno- and cyto-toxic, while 1,4-D had weak genotoxicity and no cytotoxicity. The order for cytotoxicity on the basis of EC50 values is as follows: glycolaldehyde > formaldehyde > formic acid > glycolic acid > 1,4-D > ethylene glycol diformate ≈ methoxyacetic acid, with glycolaldehyde and formaldehyde showing high genotoxicity. With the three UV/AOPs, genotoxicity expressed as mitomycin equivalency quotient (MEQ) increased significantly by 10 to 100 fold with a UV dosage of 720 mJ cm−2, mainly due to the formation of glycolaldehyde. UV/S2O82− reduced the MEQ with an increased UV dosage of 1440 mJ cm−2, due to the transformation of toxic aldehydes to less toxic organic acids. In contrast, UV/H2O2 increased the MEQ with UV dosage, resulting from the accumulation of aldehyde products. UV/NH2Cl showed the lowest MEQ due to its slow removal of 1,4-D. This study suggests that oxidants and UV dosage can affect the toxicological responses of treatments for recycled water.



Water impact

In this work, we investigated the toxicological responses of 1,4-dioxane – a trace organic solvent widely present in recycled water – during UV-based advanced oxidation processes (UV/AOPs) using cyto- and geno-toxicity bioassays. This is a novel approach to apply quick screening tools to minimize the toxicity response of recycled water, which is critical for potable reuse implementation. This is a timely study considering the occurrence of small and neutrally charged organic molecules in recycled water prior to UV treatment. The manuscript will be of interest to scientists, engineers and practitioners concerned with validation of UV/AOPs for wastewater recycle and potable reuse.

Introduction

Reuse of treated wastewater effluent is critically needed to mitigate fresh water shortages.1–4 Treatment processes typically involve membrane-based pretreatment and reverse osmosis, followed by an advanced oxidation process (AOP).5 Ultraviolet-driven advanced oxidation processes (UV/AOPs) for potable water reuse have been increasingly implemented to remove a variety of trace organic contaminants including pharmaceuticals and personal care products.6–9 However, the formation of oxidation products with potentially high toxicity is of increasing concern. Recently, investigation of toxic UV/AOP by-products has been reported and has received increasing attention.10–14 It is likely that UV/H2O2 produces a set of products that still pose toxicological responses when the parent contaminants are not fully mineralized.10–12

Hydrogen peroxide (H2O2) is the most commonly used UV/AOP oxidant in potable water reuse, with persulfate (S2O82−) and monochloramine (NH2Cl) also being relevant.15–17 NH2Cl is frequently used as a membrane anti-foulant in water reuse treatment trains and can be used as a carry-over oxidant.18 Each oxidant produces a unique set of reactive radicals in UV photolysis. For instance, H2O2 produces HO˙, S2O82− produces SO4˙ and HO˙, and NH2Cl generates Cl˙, Cl2˙ and HO˙.19

1,4-Dioxane (1,4-D) is ubiquitously present in municipal wastewater effluent, and it is not well rejected by RO membranes because it is a small and neutral molecule. Listed as a class B carcinogen by the USEPA with a notification level of 1 μg L−1 in California,20,21 1,4-D is used as a surrogate to validate UV/AOP efficiency.22,23 Efforts have been made to remove 1,4-D and other neutral molecules in RO permeate using UV/AOPs.18,21 However, less attention has been paid to the formation of oxidation products of 1,4-D during water reuse treatment and the associated toxicity implications. There is an urgent need for a better understanding of the occurrence of oxidation products and comparing their toxicity levels with that of 1,4-D. Therefore, the objectives of this research are to identify the formation and distribution of oxidation products of 1,4-D during treatment by UV photolysis of H2O2, S2O82− and NH2Cl, and compare the toxicity of the transformation products to that of the original contaminant using human cell-based bioassays.

Materials and methods

UV/AOP treatments

All chemicals used in this study were of reagent grade or higher and obtained from Sigma Aldrich or Fisher Scientific. All cell culture supplies and chemicals were obtained from Life Technologies or Fisher Scientific and stored under appropriate conditions as instructed. The working solution contained 5 mM oxidant (e.g. H2O2, S2O82− or NH2Cl) and 1 mM 1,4-D at pH 8. This pH was typical of recycled water and all oxidants were chemically stable at this pH. To avoid potential interference of background chemicals in toxicity response, all experiments were conducted in Milli-Q water. A 50 mM NH2Cl stock solution was prepared daily by adding a NaOCl stock solution to (NH4)2SO4 with a N[thin space (1/6-em)]:[thin space (1/6-em)]Cl molar ratio of 1.2 and buffered at pH 8 using borate. The solutions were then transferred to multiple 8 mL quartz tubes and placed in a carousel in a UV chamber (ACE Glass). The samples were illuminated with a low-pressure monochromatic mercury UV lamp (λ = 254 nm) at an intensity of 1.2 mW cm−2 (Phillips TUV6T5), which was cooled by circulating water. The UV fluence was measured using a multimeter equipped with a thermopile 919P sensor (Newport Power meter). Samples were collected every 5 min and quenched with 10 mM sodium bisulfite, followed by chemical analysis.

Analytical methods

Concentrations of H2O2 and S2O82− were measured by the potassium iodine colorimetric method,24 and NH2Cl were determined using DPD titration.25 Concentrations of 1,4-D and ethylene glycol diformate were directly measured with an Agilent 1200 liquid chromatograph (Text S1). Formaldehyde and glycolaldehyde were derivatized with 2,4-dinitrophenylhydrazine (DNPH) and analyzed by HPLC-UV. Carbonyl compounds such as aldehydes and ketones react with DNPH to form a complex in the presence of a strong acid. The complex then can be read using a UV-spectrometer at λ = 360 nm.26 Concentrations of glycolic acid, formic acid and methoxyacetic acid were quantified using a Dionex 1000 Ion Chromatography system (Text S1).

Toxicity assays

HCT-116 human colorectal carcinoma cells were cultured in a 5% CO2 humidified incubator at 37 °C and collected after the fourth passage. The cyto-toxicity assay (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)) and geno-toxicity assay (CellSensor p53RE-bla geno-toxicity assay) were subsequently conducted. These two assays were chosen because they were the most popular and robust toxicity assays to examine DNA damage and cell viability.27 Details on both bioassays are provided in Text S2. The chemicals employed to treat the cells were prepared from known standards to avoid toxicity interference from other potential toxins including oxidants and quenching reagents. Standards of 1,4-D and its transformation products were mixed based on the concentration distribution observed in the UV/AOP experiments (Table S1). Although the concentrations of the chemicals used in the bioassay were higher than the concentrations detected in recycled water, the data provided insight into the toxicity in comparison with 1,4-D and its oxidation products. For the mixture toxicity, the chemical standards were mixed together based on the experimentally determined product compositions after UV/AOP treatment with a UV dosage up to 1440 mJ cm−2, which is within the typical UV dosage in water reuse. After chemical exposure, the fluorescence of the mixture in the bioassay was recorded using a Victor 2 plate reader (Perkin Elmer, Shelton, CT). Concentration response curves were then plotted and the EC50 values were calculated using GraphPad Prism 7. Both theoretical and experimentally observed genotoxicity mitomycin equivalency quotients (MEQs) were calculated to determine the evolution of toxicity from 1,4-dioxane during UV/AOP treatments (Text S3).

Results and discussion

Oxidation products of 1,4-dioxane in UV/S2O82−, UV/H2O2 and UV/NH2Cl

UV/S2O82− exhibited the fastest kinetics with respect to 1,4-D removal (Fig. S3), because S2O82− had a higher quantum yield than H2O2 (0.7 vs. 0.5), and produced both SO4˙ and HO˙ as reactive radicals.19 Although UV/NH2Cl produced HO˙ through the transformation of Cl˙ and Cl2˙, the major radicals Cl˙ and Cl2˙ are less reactive with 1,4-D as compared to SO4˙ and HO˙.18,19 The three UV/AOPs produced a variety of 1,4-D oxidation products, which included ethylene glycol diformate, formaldehyde, glycolaldehyde, glycolic acid, formic acid, and methoxyacetic acid (Fig. 1 and Table S1). Similar products have been identified in other oxidation processes.28–31 SO4˙, HO˙, and Cl2˙ likely oxidized 1,4-D through H atom abstraction to initially form 1,4-dioxanyl radicals, and then reacted with O2 to generate peroxyl radicals (Scheme S1).28 The peroxyl radicals were finally terminated to produce tetroxide. A study suggested that the decomposition of tetroxide led to the formation of an oxyl radical, which was the precursor for formaldehyde and ethylene glycol diformate.28 Formaldehyde was then oxidized by radicals to formic acid. SO4˙ favored the generation of methoxyacetic acid, and HO˙ favored the formation of glycolic acid. In UV/NH2Cl, only aldehyde products were observed, because an insufficient number of radicals were produced to degrade aldehyde products to carboxylic acids. The identified chemicals accounted for the majority of the transformation products, adding up to 80–90% of the initial 1,4-dioxane dosage.
image file: c8ew00107c-f1.tif
Fig. 1 1,4-D degradation product evolution during the UV photolysis of (A) S2O82−, (B) H2O2 and (C) NH2Cl as the oxidants. Initial [oxidant] = 5 mM, initial [1,4-D] = 1 mM, pH = 8. The standard deviation of each data point was based on triplicate measurements.

Cytotoxicity and genotoxicity of 1,4-dioxane and its oxidation products

Fig. 2A presents a concentration–response curve of cell viability for 1,4-D, formaldehyde, and glycolaldehyde in HCT-116 cells. Other compounds with negligible cytotoxicity are shown in Fig. S4A. The EC50 concentrations are reported in Table 1, with glycolaldehyde being the most cytotoxic (EC50 = 155 mM), followed by formaldehyde (EC50 = 613 mM). The rank order for cytotoxicity of 1,4-D and its oxidation products based on their EC50 values was: glycolaldehyde > formaldehyde > formic acid > glycolic acid > 1,4-dioxane > ethylene glycol diformate ≈ methoxyacetic acid. This trend showed that aldehydes in general exhibited a high cytotoxicity. Glycolaldehyde was highly cytotoxic to HK-2 cells and caused depletion of adenosine triphosphate (ATP), release of lactate dehydrogenase (LDH), and degradation of enzymes as well as selected phospholipids.32 Glycolaldehyde also induced growth inhibition and oxidative stress in human breast cancer cells.33 Formaldehyde is known to be highly reactive with proteins and DNA which induces cytotoxicity.34 Formaldehyde-induced cytotoxicity inhibited mitochondrial respiration, decreased ATP depletion, and generated reactive oxygen species which contributed to oxidative stress and cell lysis in isolated rat hepatocytes.35 Ethylene glycol diformate has two carbonyl groups; however, when applied to HCT-116 cells, cell viability was not reduced in the present study. Although studies on the toxicity of ethylene glycol diformate have not been reported, our data suggests that ethylene glycol diformate might be quickly metabolized to downstream products that are not cytotoxic.
image file: c8ew00107c-f2.tif
Fig. 2 (A) Cytotoxicity and (B) genotoxicity dose–response curves of 1,4-dioxane, formaldehyde and glycolaldehyde. Cell viability represented percent of viable cells compared to the controls based on the MTT assay. Response values were calculated based on the ratio of stimulated cells vs. unstimulated cells obtained from the CellSensor p53RE-bla HCT-116 assay. Each value represents the mean of replicates ± standard deviation.
Table 1 EC50 and relative effect potency (REP) values of 1,4-dioxane and its degradation products in the P53RE-bla CT-116 cell line after 16 hours of exposure
Toxicity-level ranking Chemicals Cytotoxicity EC50 (μM) Genotoxicity EC50 (μM) REPa
a REP is the relative effect potency, REP = EC50(mitomycin)/EC50(i), where i is a specific degradation product. Details of the calculation are provided in Text S3 in the ESI.
1 Glycolaldehyde (1.6 ± 0.2) × 102 (7.1 ± 1.0) × 101 6.7 × 10−2
2 Formaldehyde (6.1 ± 0.9) × 102 (4.0 ± 2.5) × 102 1.2 × 10−2
3 Formic acid (1.3 ± 3.0) × 1014
4 Glycolic acid (7.8 ± 3.1) × 1015
5 1,4-Dioxane (1.1 ± 5.2) × 1029 >(2.0 ± 7.9) × 104 2.4 × 10−4


Previous studies reported that aldehydes are highly reactive electrophilic molecules that damage DNA through the formation of aldehyde-derived DNA adducts.36,37 The genotoxicity assay using the P53-GeneBLAzer assay indicated that aldehyde compounds were highly genotoxic (Fig. 2B) compared to 1,4-D and its carboxylic acid oxidation products (Fig. S4B). The EC50 concentrations were 71 μM for glycolaldehyde and 395 μM for formaldehyde (Table 1). 1,4-D showed relatively low genotoxicity with an EC50 > 20[thin space (1/6-em)]000 μM. Glycolaldehyde has been reported to cause DNA-protein crosslinks and DNA single-strand breaks in human peripheral mononuclear blood cells.38 Similarly, formaldehyde formed adducts with DNA and proteins, which resulted in chromosome loss due to formaldehyde-induced defects in the mitotic apparatus.34,39 In agreement with our observation, 1,4-D produced negative genotoxicity responses in several in vitro assays.40,41 A few studies reported that 1,4-dioxane increased chromosomal breaks and DNA repairs in rats or mice with chronic injection of 1,4-D.42,43 Our results indicated that 1,4-D is a weak genotoxicant to human cells. In contrast, its aldehyde oxidation products were extremely toxic to human cells.

Genotoxicity comparison among UV/S2O82−, UV/H2O2 and UV/NH2Cl

Genotoxicity expressed as mitomycin equivalency quotient (MEQ) was compared for the three oxidants with UV exposures of 0, 720 and 1440 mJ cm−2, respectively (Fig. 3). At the beginning of the treatment, 1,4-D was the only chemical present in the system with a MEQ of 2.4 × 10−4. After 720 mJ cm−2 of irradiation, UV/S2O82−, UV/H2O2, and UV/NH2Cl increased the MEQs from 2.4 × 10−4 to 1.9 × 10−2, 1.6 × 10−2 and 4.0 × 10−3, respectively (Fig. 3). Glycolaldehyde was consistently the major contributor to the MEQs in all the three UV/AOPs (Table 1 and S1). After 1440 mJ cm−2 of irradiation, the MEQ in UV/S2O82− decreased from 1.9 × 10−2 to 1.2 × 10−2, mainly due to the oxidation of glycolaldehyde to non-toxic carboxylic acids (Table S1). In contrast, H2O2 and NH2Cl further increased the MEQs 2 to 3 times at a UV dosage of 1440 mJ cm−2 (Fig. 3), which was consistent with the 23-fold increase of glycolaldehyde concentrations. The results suggest that UV/S2O82− oxidizes 1,4-D and further degrades its partial oxidation products, while UV/H2O2 and UV/NH2Cl degrade 1,4-D at slower rates, resulting in accumulation of toxic glycolaldehyde. The observed MEQ and theoretical MEQ did not statistically differ from each other in all the three UV/AOP treatments, indicating that the mixture of the analytes produced an additive effect rather than a synergistic effect. In addition, the data suggested that the identified transformation products (i.e., accounted for >80% of the product distribution) are the major contributors to the observed overall toxicity.
image file: c8ew00107c-f3.tif
Fig. 3 Mitomycin equivalency quotient (MEQ) of genotoxicity evolution during UV/AOP treatments. The observed MEQ was calculated based on the EC50 of the mixture of 1,4-dioxane and six identified transformation products (Fig. S5B). The concentration of each analyte was determined based on the experimental observations (Table S1). The theoretical MEQ was calculated based on the EC50 of each individual analyte (Fig. 2B, S4B and Text S3). Error bars represent one standard deviation. UV/S2O82−: two-way ANOVA test showed no difference between calculated and observed MEQs (P = 0.677) and a significant difference between MEQs with different UV doses of 0, 720 and 1440 mJ cm−2 (P = 0.014). UV/H2O2: two-way ANOVA test showed no difference between calculated and observed MEQs (P = 0.254) and a significant difference between MEQs with different UV doses of 0, 720 and 1440 mJ cm−2 (P = 0.0043). UV/NH2Cl: two-way ANOVA test showed no difference between calculated and observed MEQs (P = 0.250) and a significant difference between MEQs with different UV doses of 0, 720 and 1440 mJ cm−2 (P = 0.005).

Engineering implications

Our study addressed the concerns over the formation of more toxic oxidation products from UV/AOP treatments of recycled wastewater for potable water reuse. The degradation of 1,4-dioxane by UV/AOPs can generate glycolaldehyde and formaldehyde which induce toxicological responses 100 times higher than 1,4-dioxane itself. Validation of UV/AOPs for water reuse applications requires at least 0.5[thin space (1/6-em)]log of 1,4-D removal,44 which corresponds to the extent of removal achieved after 15 minutes of UV/AOP under experimental conditions in this study. For all the three UV/AOPs, glycolaldehyde remained as the major product species. For many aromatic compounds, the initial oxidation steps usually lead to the formation of aldehydes as intermediates with higher geno- and cyto-toxicity. Despite the low-level existence of oxidation products in highly treated wastewater, an accurate assessment of potential human health risks from long-term exposure to these products is needed. Although human health risk assessment of oxidation product mixtures is complex, our study demonstrates that the risk may be evaluated using cost-effective bioassay screening tools to identify causative agents in the mixture. Although additional treatment steps such as groundwater infiltration for indirect potable reuse may remove oxidation products, results from this study are important for prioritizing future toxicological assessment for potable water reuse, preparing the water industry for additional chemical detection methods and assisting the design of effective UV/AOPs that minimize the formation of toxic products.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by grants to H. L. from the National Science Foundation (CHE-1611306) and to W. L. from the National Science Foundation Graduate Research Fellowship, IGERT Water Sense Fellowship, and UC Riverside One Health Fellowship. We thank Shane Snyder at the University of Arizona for providing the HCT-116 cell line.

References

  1. B. Jiménez and T. Asano, Water reuse: an international survey of current practice, issues and needs, IWA, London, 2008 Search PubMed .
  2. G. W. Miller, Desalination, 2006, 187, 65–75 CrossRef .
  3. V. Lazarova, B. Levine, J. Sack, G. Cirelli, P. Jeffrey, H. Muntau, M. Salgot and F. Brissaud, Water Sci. Technol., 2001, 43, 25–33 CrossRef PubMed .
  4. T. Asano, Water Sci. Technol., 2002, 45, 23–33 CrossRef .
  5. D. Gerrity, B. Pecson, R. S. Trussell and R. R. Trussell, J. Water Supply: Res. Technol.--AQUA, 2013, 62, 321–338 CrossRef .
  6. P. Westerhoff, H. Moon, D. Minakata and J. Crittenden, Water Res., 2009, 43, 3992–3998 CrossRef PubMed .
  7. B. A. Wols and C. H. M. Hofman-Caris, Water Res., 2012, 46, 2815–2827 CrossRef PubMed .
  8. Y. Lester, I. Ferrer, E. M. Thurman and K. G. Linden, J. Hazard. Mater., 2014, 280, 104–110 CrossRef PubMed .
  9. V. J. Pereira, K. G. Linden and H. S. Weinberg, Water Res., 2007, 41, 4413–4423 CrossRef PubMed .
  10. A. Oliveira, M. G. Maniero and J. R. Guimaraes, J. Adv. Oxid. Technol., 2015, 18, 211–220 Search PubMed .
  11. A. R. Fernández-Alba, D. Hernando, A. Agüera, J. Cáceres and S. Malato, Water Res., 2002, 36, 4255–4262 CrossRef .
  12. L. A. de Luna, T. H. da Silva, R. F. P. Nogueira, F. Kummrow and G. A. Umbuzeiro, J. Hazard. Mater., 2014, 276, 332–338 CrossRef PubMed .
  13. P. B. Chang and T. M. Young, Water Res., 2000, 34, 2233–2240 CrossRef .
  14. P. J. Chen, K. G. Linden, D. E. Hinton, S. Kashiwada, E. J. Rosenfeldt and S. W. Kullman, Chemosphere, 2006, 65, 1094–1102 CrossRef PubMed .
  15. N. A. Ballester and J. P. Malley, J. - Am. Water Works Assoc., 2004, 96, 97–103 CrossRef .
  16. B. A. Lyon, A. D. Dotson, K. G. Linden and H. S. Weinberg, Water Res., 2012, 46, 4653–4664 CrossRef PubMed .
  17. C. Yang, Y. R. Xu, K. C. Teo, N. K. Goh, L. S. Chia and R. J. Xie, Chemosphere, 2005, 59, 441–445 CrossRef PubMed .
  18. S. Patton, W. Li, K. D. Couch, S. P. Mezyk, K. P. Ishida and H. Liu, Environ. Sci. Technol. Lett., 2016, 4, 26–30 CrossRef .
  19. W. Li, T. Jain, K. Ishida and H. Liu, Environ. Sci.: Water Res. Technol., 2017, 3, 128–138 Search PubMed .
  20. S. M. Simonich, P. Sun, K. Casteel, S. Dyer, D. Wernery, K. Garber, G. Carr and T. Federle, Integr. Environ. Assess. Manage., 2013, 9, 554–559 CrossRef PubMed .
  21. T. K. Mohr, J. A. Stickney and W. H. DiGuiseppi, Environmental investigation and remediation: 1, 4-dioxane and other solvent stabilizers, CRC Press, 2016 Search PubMed .
  22. M. Patel, UV/AOP A Key Part of the Groundwater Replenishment System, Water Environment Federation, Alexandria, 2011 Search PubMed.
  23. T. Fujioka, S. Masaki, H. Kodamatani and K. Ikehata, Curr. Pollut. Rep., 2017, 1–9 Search PubMed .
  24. W. Li, R. Orozco, N. Camargos and H. Liu, Environ. Sci. Technol., 2017, 51, 3948–3959 CrossRef PubMed .
  25. E. Rice, R. Baird, A. Eaton and L. Cleceri, Standard Methods for the Examination of Water and Wastewater, 22nd edn, 2012 Search PubMed .
  26. US Environmental Protection Agency, Method 8315A, Determination of carbonyl compounds by high performance liquid chromatography.
  27. B. I. Escher, M. Allinson, R. Altenburger, P. A. Bain, P. Balaguer, W. Busch and A. R. Humpage, Environ. Sci. Technol., 2013, 48, 1940–1956 CrossRef PubMed .
  28. M. I. Stefan and J. R. Bolton, Environ. Sci. Technol., 1998, 32, 1588–1595 CrossRef .
  29. N. Merayo, D. Hermosilla, L. Cortijo and Á. Blanco, J. Hazard. Mater., 2014, 268, 102–109 CrossRef PubMed .
  30. V. Maurino, P. Calza, C. Minero, E. Pelizzetti and M. Vincenti, Chemosphere, 1997, 35(11), 2675–2688 CrossRef .
  31. M. Otto and S. Nagaraja, Biorem. J., 2007, 17(3), 81–88 CrossRef .
  32. V. Poldelski, A. Johnson, S. Wright, V. D. Rosa and R. A. Zager, Am. J. Kidney Dis., 2001, 38, 339–348 CrossRef PubMed .
  33. K. S. Al-Enezi, M. Alkhalaf and L. T. Benov, Free Radical Biol. Med., 2006, 40, 1144–1151 CrossRef PubMed .
  34. J. Shaham, Y. Bomstein, A. Meltzer, Z. Kaufman, E. Palma and J. Ribak, Carcinogenesis, 1996, 17, 121–126 CrossRef PubMed .
  35. S. Teng, K. Beard, J. Pourahmad, M. Moridani, E. Easson, R. Poon and P. J. O'Brien, Chem.-Biol. Interact., 2001, 130, 285–296 CrossRef PubMed .
  36. V. Vasiliou, A. Pappa and D. R. Petersen, Chem.-Biol. Interact., 2000, 129, 1–19 CrossRef PubMed .
  37. G. P. Voulgaridou, I. Anestopoulos, R. Franco, M. I. Panayiotidis and A. Pappa, Mutat. Res., Fundam. Mol. Mech. Mutagen., 2011, 711, 13–27 CrossRef PubMed .
  38. J. G. Hengstler, J. Fuchs, S. Gebhard and F. Oesch, Mutat. Res., Fundam. Mol. Mech. Mutagen., 1994, 304, 229–234 CrossRef PubMed .
  39. T. Orsiere, I. Sari-Minodier, G. Iarmarcovai and A. Botta, Mutat. Res., Genet. Toxicol. Environ. Mutagen., 2006, 605, 30–41 CrossRef PubMed .
  40. T. Morita and M. Hayashi, Environ. Mol. Mutagen., 1998, 32, 269–280 CrossRef PubMed .
  41. T. L. Goldsworthy, T. M. Monticello, K. T. Morgan, E. Bermudez, D. M. Wilson, R. Jäckh and B. E. Butterworth, Arch. Toxicol., 1991, 65, 1–9 CrossRef PubMed .
  42. W. T. Stott, J. F. Quast and P. G. Watanabe, Toxicol. Appl. Pharmacol., 1981, 60, 287–300 CrossRef PubMed .
  43. S. K. Roy, A. K. Thilagar and D. A. Eastmond, Mutat. Res., Genet. Toxicol. Environ. Mutagen., 2005, 586, 28–37 CrossRef PubMed .
  44. California State Water Resources Control Board, Regulations Related to Recycled Water, Title 22 California Code of Regulations.

Footnote

Electronic supplementary information (ESI) available: Additional texts, figures and tables of analytical methods, cell bioassays and 1,4-D degradation and product distribution. See DOI: 10.1039/c8ew00107c

This journal is © The Royal Society of Chemistry 2018