Environ. Sci. Technol. 2005, 39, 6086-6092
Kinetics of Aqueous Ozone-Induced Oxidation of Some Endocrine Disruptors M A R I E D E B O R D E , * ,† S Y L V I E R A B O U A N , †,‡ JEAN-PIERRE DUGUET,§ AND BERNARD LEGUBE† Laboratoire de Chimie de l’Eau et de l’Environnement, UMR CNRS 6008 Ecole Supe´rieure d’Inge´nieurs de Poitiers - 40, avenue du Recteur Pineau - 86022 Poitiers Cedex, France, Faculte´ de Me´decine et Pharmacie - 34, rue du Jardin des Plantes, BP 199 - 86005 Poitiers Cedex, France, and Direction Qualite´ et Environnement - Eau de Paris - SAGEP - 9, rue Victor Schoelcher - 75675 Paris Cedex 14, France
This study investigated aqueous ozone-induced oxidation of six endocrine disruptors (EDs: 4-n-nonylphenol, bisphenol A, 17R-ethinylestradiol, 17β-estradiol, estrone, and estriol). In the first part, ED ozonation kinetics were studied over a pH range of 2.5-10.5 at 20 ( 2 °C and in the presence of tert-butyl alcohol. Under these conditions, for each studied compound, the apparent ozone rates presented minima at acidic pH (pH < 5) and maxima at basic pH (pH > 10). In the second part, to explain this pH dependence, elementary reactions, i.e., reactions of ozone with neutral and ionized ED species, were proposed, and the intrinsic constants of each of them were calculated. The reactivity of ozone with ionized EDs (i.e. 1.06 × 1096.83 × 109 M-1 s-1) was found to be 104-105 times higher than with neutral EDs (i.e. 1.68 × 104 M-1 s-1-2.21 × 105 M-1 s-1). At pH > 5, ozone reacted to the greatest extent with dissociated ED forms. Finally, to assess the potential of ozone for inducing ED oxidation in water treatment conditions, the expected removal rates for each of the studied EDs were determined on the basis of the kinetic study at pH ) 7 and 20 ( 2 °C. For all EDs considered, O3 exposures of only ∼2 × 10-3 mg min L-1 were calculated to achieve g95% removal efficiency. The ozonation process could thus highly oxidize the studied EDs under water treatment conditions.
Introduction Endocrine disruptors (ED) are defined as “exogenous substances that cause adverse health effects in an intact organism, or its progeny, consequent to changes in endocrine function” (1, 2). They include (i) natural compounds such as steroid hormones naturally secreted by humans and animals and (ii) anthropogenic compounds such as synthetic hormones used for contraception or in the management of menstrual and menopausal disorders or agricultural and industrial chemicals (pesticides, alkylphenols, bisphenol A, phthalate plasticizers, etc.). In recent years, there has been * Corresponding author phone: 335 49 45 44 74; fax: 335 49 45 37 68; e-mail:
[email protected]. † Laboratoire de Chimie de l’Eau et de l’Environnement. ‡ Faculte ´ de Me´decine et de Pharmacie. § Direction Qualite ´ et Environnement - Eau de Paris - SAGEP. 6086
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 16, 2005
growing concern about the presence of EDs in the aquatic environment. Actually, several studies have recently documented a great variety of EDs in surface waters (2, 3). Moreover, the feminization of male fish has been detected in several countries over the past 10 years (4, 5). Most EDs are believed to reach the aquatic environment via sewage effluent (6-9). Concerning hormones, estrogens produced daily by men and women or synthetic steroids ingested by humans are mainly excreted as conjugates of sulfuric and glucuronic acids in urine (10). These biologically inactive compounds are then discharged in the municipal sewage system where they are deconjugated by bacteria (11, 12) and partially removed (13), before being released into effluents and surface waters as active estrogens. There is high variability in ED concentrations in surface waters. However, at some sites, concentrations have been detected in the ng L-1 range for natural or synthetic hormones (1416) and in the µg L-1 range for alkylphenols and bisphenol A (17, 18). In laboratory studies, such concentrations were found to induce feminization phenomena in male fish. In 1997, for nonylphenol, Gray and Metcalfe (19) reported the intersex induction at 50 µg L-1. In 2001, similar effects were observed at 10 ng L-1 for estradiol or estrone and 0.1 ng L-1 for ethinylestradiol (20). In the same way, several studies have reported a correlation between reproductive abnormalities in animals and exposure to EDs in real situations (21-23). Concerning the potential action of EDs on human health, apart from accidental situations (24, 25), no clear correlations between ED exposure and reproductive disturbance have been shown. However, an increase in estrogendependent cancers and a reduction in human sperm production have been noticed over the last 10 years (26, 27). Drinking water produced from surface waters could be a common means of human exposure to EDs. Hence, it is important to assess water treatment processes with regard to their ED removal potential. Concerning oxidation processes, the French Department of Health and Human Services allows the use of ozone as a clarifying agent or disinfectant. To avoid potential byproduct formation, ozonation and radical oxidation are not allowed in order to focus specifically on the oxidation of certain organic compounds (such as pesticides and related compounds). Note that chlorination is not also used for the removal of organic pollutants. However, when such pollutants are present in treated water, it is important to know their reactivity with disinfectants. Little is currently known about this question. However, the effects of chlorination on some EDs were investigated in some recent studies. It was demonstrated that chlorination was effective in oxiding numerous EDs but sometimes also induced some biologically active first byproducts (28-31). In previous studies (32, 33) on the chlorination kinetics of seven EDs (β-estradiol, estrone, estriol, progesterone, 17Rethinylestradiol, 4-n-nonylphenol, and bisphenol A), it was shown that all molecules including a phenolic ring in their structure were oxidized by chlorine, whereas progesterone remained unchanged. For reactive compounds, regardless of the pH, a second-order reaction rate was observed. At pH 7, the apparent second-order rate constants were calculated to be 12.6 M-1 s-1 for 4-n-nonylphenol, 61.8 M-1 s-1 for bisphenol A, and 112-131 M-1 s-1 for hormones. For a total chlorine concentration of 1 mg L-1, the corresponding halflife times were then calculated to be about 65 min for 4-nnonylphenol, 13 min for bisphenol A, and 6-8 min for hormones. Concerning ozonation, only limited information is available about the efficacy of ED oxidation by this process. However, in their study on oxidation of pharmaceuticals 10.1021/es0501619 CCC: $30.25
2005 American Chemical Society Published on Web 07/15/2005
TABLE 1: Selected Endocrine Disruptors
*Carbon numbers are similar for all studied hormones.
during ozonation, Huber et al. (34) reported rapid removal of 17R-ethinylestradiol by ozone with a second-order rate constant of 3 × 106 M-1 s-1 at pH 7 and 20 °C. On the basis of chemical characteristics, they also estimated the potential removal and second-order rate constants of estradiol (106 M-1 s-1), 4-nonylphenol (1-10 × 106 M-1 s-1), bisphenol A (1-10 × 106 M-1 s-1), and testosterone (105 M-1 s-1) at pH 7 and 20 °C. The objective of this study was to assess the potential of ozone to directly induce the oxidation of some EDs, i.e., to determine, for each compound, the rate constant of each elementary reaction in order to be able to calculate the apparent rate constant for a given pH. The ozonation kinetics of six EDs were thus studied in the presence of tert-butyl alcohol. These compounds (reported in Table 1) were chosen because they are commonly found in the environment. They include natural estrogens (17β-estradiol (E2), estrone (E1) and estriol (E3)), synthetic hormone (17R-ethinylestradiol (EE2)), and chemicals used in industry (4-n-nonylphenol (NP) and bisphenol A (BPA)). In the first part, oxidation reactions with ozone were assessed in pure aqueous solution at 20 ( 2 °C and different pH levels. The apparent ozone rate constants were determined for each molecule. In the second part, the rate constants of each elementary reaction were determined on the basis of the pH dependence of the rate constants and the speciation of each ED. Finally, from the kinetic rates
obtained, the ozone efficiency for ED oxidation was predicted under water treatment conditions.
Experimental Section Standards and Reagents. All endocrine disruptors (E2, E1, E3, EE2, NP, and BPA) were supplied by Aldrich (purity g 97%). Other reagents (NaOH, H2SO4, phosphate, and carbonate) were analytical grade and used without further purification. All stock solutions were prepared with ultrapurified water (18 MΩ cm) obtained from Milli RO-Milli Q Millipore system. Ozone stock solutions were produced by sparging O3containing oxygen through Milli-Q water, which was cooled at 4 °C (35). Kinetic Experiments. For all studied compounds, some preliminary ozonation tests revealed very fast reaction rates. Hence, all ozone rate constants were determined by a competitive kinetic method. The competitor used was phenol because similar reaction mechanisms and rate constants were expected. For each compound, ozonation experiments were performed with Milli-Q purified water, in a batch reactor, at 20 ( 2 °C and different pH (range 2.5-10.5). The pH of tested aqueous solutions was adjusted with H2SO4 for pH < 6 and buffered with phosphate or carbonate salts (5 mM) for 6 e pH < 9 and pH g 9, respectively. An OH radical scavenger VOL. 39, NO. 16, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6087
TABLE 2: Mobile Phase Composition and Elution Used To Analyze Phenol and EDs by High-Performance Liquid Chromatography analyzed compounds phenol and EE2, E2, or E1 phenol and BPA phenol and E3 NP (in the presence of phenol) phenol (in the presence of NP) a
mobile phase composition methanol/water (v/v) from to kept
elution used
55/45 (initial composition) 70/30 (after 15 min) 70/30 (for 5 min) 55/45 50/50 90/10 55/45a
slightly convex gradient isocratic isocratic isocratic isocratic
Analysis followed by a slightly convex gradient to 90/10 to eliminate all trace of NP.
was used to avoid radical reactions. Since the rate constants of the competitor used were determined in the presence of tert-butyl alcohol (36), tert-butyl alcohol was also chosen as OH radical scavenger under our experimental conditions. For each experiment, the initial tert-butyl alcohol concentration ([tert-but]0 ) 4 mM) was calculated by eq 1, so the OH radicals mainly reacted with tert-butyl alcohol
kOH/tert-but [tert-but]0 > 50 (kOH/ED [ED]T,0 + kOH/Phenol [Phenol]T,0) (1) where [ED]T,0 and [Phenol]T,0 are the total initial ED and phenol concentrations, respectively, and kOH/tert-but ) 6.0 × 108 M-1 s-1 (37); kOH/Phenol ) 1.4 × 1010 M-1 s-1 (37); kOH/ED ) 1 × 1010 M-1 s-1. Note that, for the studied EDs, only the kinetic rate constant of the OH radical with EE2 at pH 7 has been given in the literature (kOH/EE2 ) 9.8 ((1.2) × 109 M-1 s-1) (34). By considering a similar reactivity of the OH radical with each studied compound, a rate constant kOH/ED of 1.0 × 1010 M-1 s-1 was then used for the calculation. For each compound and pH, experiments were performed by adding, with a glass syringe, different substoichiometric concentrations of ozone (range 0.2-3 µM) to a series of reactors (7 or 8) containing equal initial ED and competitor concentrations (1 µM) and vigorously stirred. The remaining ED and phenol concentrations of each reactor were then analyzed by HPLC. Analytical Methods. For each ozonation experiment, ED and phenol used as competitors were analyzed by highperformance liquid chromatography (HPLC) equipped with an automatic Waters 717 plus autosampler injector and a Waters 600E pump. All compounds were detected with a Waters 484 Tunable absorbance detector set at 270 nm. For each sample, a 200 µL injection volume was used, and all analytes were separated using a 250 × 4.6 Kromasil C18 column. A water-methanol mixture was used as the mobile phase with a flow-rate of 1 mL min-1. Depending on the compounds and experiments, isocratic or gradient elutions were used by varying the eluent ratio. Table 2 summarizes the mobile phase composition and elution used to analyze phenol and EDs for each experiment. For the NP kinetic experiments, due to the very different NP and phenol affinities for the mobile and stationary phases, these two compounds were analyzed separately. The dissolved ozone concentration was determined with the indigo method (35). The absorptivity of the different samples at 600 nm was determined using a SAFAS spectrophotometer UV-visible 320. pH was measured with a Tacussel LPH330T pH-meter equipped with a Radiometer Analytical combined electrode and previously calibrated with standard buffers (pH 4, 7, and 10).
Results and Discussion Apparent Rate Constants for the Reaction of Ozone with EDs. All studied EDs were rapidly oxidized by ozone under experimental conditions. Regardless of the pH and ED 6088
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 16, 2005
FIGURE 1. Determination of the kappBPA/kappPhenol ratio for simultaneous reactions of ozone with phenol and bisphenol A at 20 ( 2 °C and different pH levels. studied, kinetic experiments were performed in the presence of phenol as competitor. Under these conditions, if ozone reactions with ED and phenol exhibit second-order reaction rates, first-order relative to ozone ([O3]), and first-order relative to the total ED or phenol ([ED]T and [Phenol]T, respectively), the rate laws can be formulated as
d[ED]T ) kappED[O3] [ED]T dt
(2)
d[Phenol]T ) kappPhenol[O3] [phenol]T dt
(3)
v)and
v)-
with kappED and kappPhenol representing the ED and phenol apparent second-order kinetic constants for a given pH. When eq 2 is divided by eq 3, the kinetics for simultaneous reactions of ozone with ED and phenol can be expressed as follows:
d[ED]T d[Phenol]T
)
kappED [ED]T kappPhenol[Phenol]T
(4)
or after integration of eq 4:
ln
[ED]T,0 [ED]T,t
)
kappED
[Phenol]T,0 ln kappPhenol [Phenol]T,t
(5)
Figure 1 presents the graph of ln [ED]T,0/[ED]T,t versus ln [Phenol]T,0/[Phenol]T,t in the case of simultaneous reactions of ozone with BPA and phenol at pH 2.78, 4.07, 6.64, and 8.86. It shows, irrespective of the pH, linear plots with correlation coefficients always higher than 0.99. Similar results were obtained for other compounds (r2 > 0.97 except for NP, r2 > 0.95). Equation 5 was then followed irrespective
O3 + ED- f byproducts
k2
(11)
O3 + ED2- f byproducts
k3
(12)
and for BPA
FIGURE 2. pH-dependence of the apparent second-order rate constants for the reaction of ozone with bisphenol A, 4-nnonylphenol, and estrone at 20 ( 2 °C. Symbols represent measured data and lines represent the model calculations. of the compounds and the pH considered. For each experiment, from the slope of the straight line (kappED/kappPhenol) and the calculated phenol apparent second-order rate constants (eq 6), the apparent second-order rate constant of ED was then determined -pH
kappPhenol ) kO3/Phenol
10 + 10-pKa + 10-pH
10-pKa kO3/Phenolate -pKa (6) 10 + 10-pH
with kO3/Phenol ) 1.3 ((0.2) × kO3/Phenolate ) 1.4 ((0.4) × 109 M-1 s-1, and pKa ) 9.9 (36). Thereby, the kappED values were determined for about 1015 pH values for each ED. pH Dependence Profile and Rate Constants of Elementary Reactions. Figure 2 presents the pH profile of the apparent second-order rate constants for BPA, NP, and E1. Similar profiles were obtained for E3, E2, and EE2. These results show a pH dependence of the rate constants, with minima for low pH values (pH < 5) and maxima for high pH values (pH > 10), irrespective of the compound considered. Chemically, all studied compounds include one (or two) phenolic rings. They are then characterized by one (or two) acidity constants (Table 1) and present in aqueous solution in two (or three) forms (reactions 7 and 8): 104
M-1
ED a ED- + H+
s-1,
Ka1ED
(7)
and for BPA
ED- a ED2- + H+ Ka2ED
(8)
[ED]T ) [ED] + [ED-] (+ [ED2-] for BPA)
(9)
with
Under our experimental conditions, tert-butyl alcohol was used as an OH radical scavenger. Only the ozone (O3) action was then studied. Hence, for a given pH, the global ozone reaction could be explained by the reactions of ozone with neutral and ionized EDs (reactions 10-12):
O3 + ED f byproducts
k1
(10)
where k1, k2, and k3 represent the rate constants of each elementary reaction and are constant irrespective of the pH considered. At acidic pH (pH < 5), neutral EDs were the predominante ED species. The nearly constant reactivity of EDs with ozone would thus be explained by the reaction between O3 and neutral ED species (reaction 10). In the 5-10 pH range, the ionized ED fraction increased with increasing pH values. The reaction between O3 and ionized EDs (reactions 11 and 12) would thus gradually increase, while the reaction of O3 with the neutral fraction would decrease. In this pH range, there was an increase in ED reactivity, as shown in Figure 2. If the global ozone reaction could be explained by reactions 10-12, then the reaction between O3 and ionized EDs would be faster than that of O3 with neutral EDs (i.e. k1 < k2 and k3). Considering that global ozone reaction involves reactions 10-12, the ozone-induced oxidation kinetics of EDs can be expressed as follows:
d[ED]T ) dt [O3] (k1[ED] + k2[ED-] ( + k3[ED2-] for BPA)) (13)
v)-
Replacing [ED], [ED-], (and [ED2-] for BPA) by their expressions as ratios of [ED]T, the rate of EDs disappearance is
v)-
d[ED]T ) dt [O3][ED]T
(k1R1 + k2R2 (+ k3R3 for BPA)) (14)
with R1 ) neutral ED fraction, R2 ) ionized ED- fraction, and R3 ) ionized ED2- fraction. Based on eq 14, the kappED term in eq 2 can then be rewritten as
kappED )
k1[H+] + k2Ka1ED
(15)
[H+] + Ka1ED
and for BPA
kappED )
k1[H+]2 + k2Ka1ED[H+] + k3Ka1EDKa2ED [H+]2 + Ka1ED[H+] + Ka1EDKa2ED
(16)
For each studied compound, from eqs 15 or 16, the intrinsic constants k1, k2, and (k3 for BPA) were calculated by multiple regression of the experimental kappED data. Values of k1, k2, and k3 were obtained for the lowest quadratic mean deviation, defined as Σ((kappED.exp - kappED.theo)2)/(kappED.exp)2, where kappED.exp and kappED.theo represent the experimental and theoretical apparent second-order rate constants, respectively. The ED pKa values used for the calculation were derived from the literature and are reported in Table 1. For E2 and E1, the pKa values were estimated according to Perrin (41) based on the analogy of their structure with EE2 and E3. Besides the experimental kappED values, Figure 2 represents the theoretical pH profiles obtained with the model in the case of BPA, NP, and E1. It shows a good fit of the experimental pH profile and the proposed model for each of these compounds. Similar correlations were obtained for E3, E2, and EE2. Hence, the previous elementary reactions (reactions 10-12) are consistent with the data, irrespective of the ED considered. VOL. 39, NO. 16, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6089
TABLE 3: Second-Order Rate Constants Calculated for Elementary Reactions of Ozone with the Investigated EDs (Reactions 10-12) at 20 ( 2 °C compounds
k1 ((σ) (M-1 s-1)
k2 ((σ) (M-1 s-1)
k3 ((σ) (M-1 s-1)
bisphenol A 4-n-nonylphenol 17R-ethinylestradiol 17β-estradiol estrone estriol
1.68 (0.21) × 104 3.80 (0.83) × 104 1.83 (0.38) × 105 2.21 (0.50) × 105 1.53 (0.29) × 105 1.01 (0.26) × 105
1.06 (0.15) × 109 6.83 (0.90) × 109 3.65 (0.46) × 109 3.69 (0.44) × 109 4.24 (0.62) × 109 3.89 (0.55) × 109
1.11 (0.52) × 109
FIGURE 3. (a) Modeled bisphenol A, 4-n-nonylphenol, and estrone removal as function of O3 exposure at pH ) 7 and 20 ( 2 °C. (b) Modeled bisphenol A, 4-n-nonylphenol, and estrone removal as function of chlorine exposure at pH ) 7 and 20 ( 2 °C. This graph was drawn up on the basis of data from refs 32 and 33. Table 3 summarizes, for each studied compound, the fitted values of the intrinsic rate constants, i.e., k1 (O3 reaction with ED), k2 (O3 reaction with ED-), and k3 (O3 reaction with ED2in the case of BPA). For EE2, the second-order rate constant of O3 with ionized EE2 species (k2) was comparable to the one reported by Huber et al. (34) (≈7 × 109 M-1 s-1 instead of 3.65 ( 0.46 × 109 M-1 s-1 obtained in this study). As for ED chlorination (32, 33), the reactivity of ozone with neutral ED species was 104-105 times weaker than with the ionized forms. Attenuation of the electron-donor character of the hydroxyl group, i.e., change of the O- group in OH, thus decreased the reaction rates. For pH < 5, ozone reactions with EDs were mainly controlled by the reaction between O3 and neutral species. At higher pH, the reaction between O3 and ionized species prevailed. Finally, the rate constants of O3 with dissociated forms were found to be similar regardless of the ED considered (Table 3). They ranged from 1.06 ((0.15) to 6.83 ((0.90) × 109 M-1 s-1, which is in the same order of magnitude as the corresponding rate constant of phenol (36). This suggested that phenol moieties could be the main ozone reaction site of the ionized ED species. It also suggests that there could be a similar reactivity for other phenolic EDs (e.g. octylphenol, equilin, equilenin, diethylstilbestrol, etc.) at pH > 5. In contrast, among the studied compounds, only the rate constants of O3 with neutral species of BPA and NP (i.e. 1.683.80 × 104 M-1 s-1) were similar to the corresponding rate constant of phenol. Rate constants for the neutral studied hormones (EE2, E2, E1, E3) are higher (i.e. 1.01-2.21 × 105 M-1 s-1) and about 10-fold that of phenol (36). Therefore, the chemical structure of these hormones seems to enhance the ozone reactivity with neutral species. As neutral species of ED were predominant at pH < 5, this suggested that hormones with a phenolic ring (EE2, E2, E1, E3) have a higher reactivity than the other studied phenolic compounds (BPA, 6090
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 16, 2005
NP) at acidic pH. Although the studied hormones have a more complex chemical structure than BPA, NP, or phenol, it is difficult to explain the significant difference between the observed rate constants at acidic pH. Moreover, this cannot be explained by the difference between pKa values (Table 1). However, if the carbon 2 of the studied hormones (Table 1) has a slightly higher atom partial charge than the corresponding carbon of BPA or NP, then ozone could attack hormones faster at acidic pH. This effect would be attenuated at elevated pH due to the stronger electron-donor character of the O- group. ED Ozonation in Water Treatment Conditions. During ozonation, under water treatment conditions, OH radicals are formed as a consequence of O3 decay. Micropollutants can then be oxidized either by O3 directly or by OH radicals. O3 is a very selective oxidant that attacks certain functional groups, whereas the OH radical reacts very fast with a large number of moieties. In natural water, most OH radicals are then scavenged by the water matrix (42). In this work, the studied EDs were shown to react very fast with O3 at neutral pH. Based on the known OH radical reactivity with EE2 (34), very fast reactions between OH radicals and the studied EDs could also be expected. However, due to the higher selectivity of O3, EDs should be mainly oxidized by O3 under water treatment conditions. Therefore, to study ED ozonation in water treatment conditions, we only focused on O3 oxidation at pH 7. For each studied ED, the theoretical apparent secondorder rate constants can be calculated for a given pH from the rate constants of each elementary reaction (Table 3) and eqs 15 and 16. Based on the assumption that reactions of O3 with EDs exhibit second-order reactions (eq 2), the removal of each studied compound as function of O3 exposure could then be modeled. To assess the effect of ozone disinfection on EDs in water treatment conditions, Figure 3a represents
the extent of potential removal as function of O3 exposure modeled for BPA, NP, and E1 at pH ) 7 and 20 °C. As EE2, E2, and E3 present intrinsic rate constants similar to those of E1, a similar removal efficiency could then be expected. Under these conditions, ozone prompted rapid elimination of all the studied EDs with O3 exposures of less than 4 × 10-4 mg min L-1 such that half of the initial ED was removed. For all compounds, an O3 exposure of only ∼2 × 10-3 mg min L-1 was necessary to achieve g95% removal efficiency. In the presence of chlorine, and under the same conditions, i.e., pH ) 7 and 20 °C, the second-order rate constants of the same EDs ranged from 12.6 to 131.1 M-1 s-1 (32, 33), which was significantly weaker than for ozone. As shown in Figure 3b, chlorine exposure of about 65 mg min L-1 for NP, 13 mg min L-1 for BPA, and 6-7 mg min L-1 for hormones could be necessary to remove half of the initial ED. Finally, chlorine exposure of about 27-32 mg min L-1 for hormones, 57 mg min L-1 for BPA, and 281 mg min L-1 for NP could be necessary to achieve 95% removal efficiency. Ozonation was therefore much more efficient than chlorination for oxiding the selected EDs: under usual drinking water treatment conditions, only hormones and BPA should be efficiently oxidized (g90%) by chlorine, whereas all the studied EDs would be eliminated by ozone. To assess the overall effect of desinfection treatments, byproducts of parent EDs should be identified, and their reactivity with oxidants and their biological activity should be evaluated. Concerning chlorination, numerous byproducts were identified during BPA, NP, E2, and EE2 chlorination. Most of these byproducts seemed to react with chlorine. However, some of them could remain in the water for several hours in the presence of a residual chlorine concentration (28-30). Estrogenic activity was also investigated during these four chlorination EDs. These studies showed that estrogenic activity markedly decreased for long chlorine contact times (24 h) irrespective of the compounds considered (43). Nevertheless, in the case of BPA and hormones, residual estrogenic activity of more than 1 h was observed in the presence of a residual chlorine concentration. This biological activity could be explained by the first chlorinated byproducts which exhibited estrogenic activity similar or superior to that of their parent compounds (28, 29, 31). Concerning ozonation, little is currently known about the studied ED ozonation byproducts. However, it is well-known that phenol ozonation induces some byproducts that are less reactive with ozone (44, 45). As all the studied EDs include a phenolic ring which probably reacts with O3, similar mechanisms and byproducts could thus also be expected. For EE2, E2, and E1, the reactivity of O3 with the phenol moiety has been recently documented. Hence, O3 would attack at position 2 or 4 (Table 1) of the phenol ring to quickly induce the formation of catechol, orthoquinone, and muconic acid derivatives as major intermediates. These intermediates would still be reactive to O3 and could thus undergo further reactions. Moreover, for these hormones, in addition to the phenol ring, other reactive moieties such as the ethinyl group for EE2 or alcohol groups for E2 would also be attacked by O3. Contrary to chlorination byproducts, the chemical structures of the ozonation byproducts were then rapidly and significantly altered as compared to the parent compounds. This drastic transformation of the molecule was shown to prompt a significant decrease in estrogenic activity during EE2 ozonation. In fact, according to Huber et al., O3 doses typically applied for disinfection of drinking water were sufficient to reduce estrogenicity by a factor of more than 200 (46). In France, to avoid potential byproduct induction, ozonation and radical oxidation are not usually allowed so as to only specifically oxidize organic compounds. However, depending on the specific organic pollutant to remove and the health impact of its potential byproducts, the use of ozonation and
radical oxidation such as O3/H2O2, can require special permission. Concerning hormone ozonation, this study indicated that hormones could be rapidly removed under water treatment conditions. Moreover, according to Huber et al., a significant decrease in estrogenic activity would be expected for O3 doses typically applied for disinfection of drinking water. Ozonation could thus be efficient for removal of the studied hormones from drinking water. For NP and BPA, this study also indicated that they could be rapidly removed under water treatment conditions. However, little is currently known about byproduct induction and estrogenic activity. Further identifications, kinetic studies, and evaluations of the biological activity of ozonated byproducts are then required to be able to model the fate of these compounds in the environment and assess the health impact of drinking tap water.
Acknowledgments The authors gratefully thank SAGEP for the financial support. The authors also thank Prof. Marc Arnaudon, Mathematics section, University of Poitiers, for his advice and help concerning the statistical calculations.
Literature Cited (1) Lutz, I.; Kloas, W. Amphibians as a model to study endocrine disruptors: I. Environmental pollution and estrogen receptor binding. Sci. Total Environ. 1999, 225, 49-57. (2) European Commission. In European Workshop on the Impact of Endocrine Disrupters on the Human Health and Wildlife; Report EUR 17549; Report of the Proceedings, December 2-4, 1996, Weybridge, U.K., 1996. (3) Kolpin, D. W.; Furlong, E. T.; Meyer, M. T.; Thurman, E. M.; Zaugg, S. D.; Barber, L. B.; Buxton, H. T. Pharmaceuticals, hormones and other organic wastewater contaminants in U.S. streams, 1999-2000: a national reconnaissance. Environ. Sci. Technol. 2002, 36, 1202-1211. (4) Jobling, S.; Nolan, M.; Tyler, C. R.; Brighty, G.; Sumpter, J. P. Widespread sexual disruption in wild fish. Environ. Sci. Technol. 1998, 32, 2498-2506. (5) Larsson, D. G. J.; Adolfsson-Erici, M.; Parkkonen, J.; Pettersson, M.; Berg, A. H.; Olsson, P. E.; Fo¨rlin, L Ethinyl oestradiol - an undesired fish contraceptive? Aquat. Toxicol. 1999, 45, 91-97. (6) Aherne, G. W.; Briggs, R. J. The relevance of the presence of certain synthetic steroids in the aquatic environment. J. Pharm. Pharmacol. 1989, 41, 735-736. (7) Knight, W. M. In Estrogens in the Environment; Elsevier-NorthHolland: Amsterdam, 1980. (8) Tabak, H. H.; Bloomhuff, R. N.; Bunch, R. L. Steroid hormones as water pollutants II. Dev. Ind. Microbiol. 1981, 22, 497-519. (9) Shore, L. S.; Gurevitz, M.; Shemesh, M. Estrogen as an environmental pollutant. Bull. Environ. Contam. Toxicol. 1993, 51, 361-366. (10) D’Ascenzo, G.; Di Corcia, A.; Gentili, A.; Mancini, R.; Mastropasqua, R.; Nazzari, M.; Samperi, R. Fate of natural estrogen conjugates in municipal sewage transport and treatment facilities. Sci. Total Environ. 2003, 302, 199-209. (11) Ternes, T. A.; Kreckel, P.; Mueller, J. Behavior and occurrence of oestrogens in municipal sewage treatment plants - II. Aerobic batch experiments with actived sludge. Sci. Total Environ. 1999, 225, 91-99. (12) Panter, G. H.; Thompson, R. S.; Beresford, N.; Sumpter, J. P. Transformation of non oestrogenic steroid metabolite to an oestrogenically active substance by minimal bacterial activity. Chemosphere 1999, 38, 3579-3596. (13) Baronti, C., Curini, R.; D’Ascenzo, G.; Di Corcia, A.; Gentili, A.; Samperi, R. Monitoring natural and synthetic estrogens at actived sludge sewage treatment plants and in a receiving river water. Environ. Sci. Technol. 2000, 34, 5059-5066. (14) Ying, G. G.; Kookana, R. S.; Ru, Y. J. Occurrence and fate of hormone steroids in the environment. Environ. Int. 2002, 28, 545-551. (15) Xiao, X. Y.; McCalley, D. V.; McEvoy, J. Analysis of estrogens in river water and effluents using solid-phase extraction and gas chromatography-negative chemical ionisation mass spectrometry of pentafluorobenzoyl derivatives. J. Chromatogr. A 2001, 923, 195-204. (16) Belfroid, A. C.; Van der Horst, A.; Vethaak, A. D.; Scha¨fer, A. J.; Rijs, G. B. J.; Wegener, J.; Cofino, W. P. Analysis and occurrence of estrogenic hormones and their glucuronides in surface water VOL. 39, NO. 16, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6091
(17) (18)
(19) (20)
(21)
(22) (23)
(24) (25) (26) (27) (28) (29) (30) (31)
and wastewater in The Netherlands. Sci. Tot. Environ. 1999, 225, 101-108. Blackburn, M. A.; Waldock, M. J. Concentrations of alkylphenols in rivers and estuaries in England and Wales. Water Res. 1995, 29, 1623-1629. Ahel, M.; Giger, W.; Schaffner, C. Behaviour of alkylphenol polyethoxylate surfactants in the aquatic environment - II. Occurrence and transformation in rivers. Water Res. 1994, 28, 1143-1152. Gray, M. A.; Metcalfe, C. D. Induction of testis-ova in Japanese medaka (Oryzias latipes) exposed to p-nonylphenol. Environ. Toxicol. Chem. 1997, 16, 1082-1086. Metcalfe, C. D.; Metcalfe, T. L.; Kiparissis, Y.; Koenig, B. G.; Khan, C.; Hughes, R. J.; Croley, T. R.; March, R. E.; Potter, T. Estrogenic potency of chemicals detected in sewage treatment plant effluents as determined by in vivo assays with Japanese medaka. Environ. Toxicol Chem. 2001, 20, 297-308. Guillette, L. J., Jr.; Gross, T. S.; Masson, G. R.; Matter, J. M.; Percival, H. F.; Woodward, A. R. Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contamined and control lakes. Environ. Health Perspect. 1994, 102, 680-688. Andersson, T.; Forlin, L.; Harig, I.; Larsson, A. Physiological disturbances in fish living in coastal water polluted with bleached kraft mill effluents. Can. J. Fish. Aquat. Sci. 1988, 45, 1525-1536. Morrison, P. A.; Leatherland, J. F.; Sonstegard, R. A. Plasma cortisol and sex steroid levels in great lakes coho salmon (Oncorhynchus kisutch walbaum) in relation to fecundity and egg survival. Comp. Biochem. Physiol. 1985, 80A, 61-68. Herbst, A. L.; Senekjian, E. K.; Frey, K. Abortion and pregnancy loss among diethylstilbestrol-exposed women. Semin. Reprod. Endocrinol. 1989, 7, 124-129. Mocarelli, P.; Brambilla, P.; Gerthoux, P. M.; Patterson, D. G., Jr.; Needham, L. L. Change in sex ratio with exposure to dioxin. Lancet 1996, 348, 401-416. Swan, S. H.; Elkin, E. P.; Fenster, L. Have sperm densities declined? A reanalysis of global trend data. Environ. Health Perspect. 1997, 105, 1228-1232. WHO. In The world health report; World Health Organization: Geneva, 1997. Hu, J. Y.; Aizawa, T.; Ookubo, S. Products of aqueous chlorination of bisphenol A and their estrogenic activity. Environ. Sci. Technol. 2002, 36, 1980-1987. Hu, J. Y.; Cheng, S.; Aizawa, T.; Terao, Y.; Kunikane, S. Products of aqueous chlorination of 17β-estradiol and their activities. Environ. Sci. Technol. 2003, 37, 5665-5670. Hu, J. Y.; Xie, G. H.; Aizawa, T. Products of aqueous chlorination of 4-nonylphenol and their estrogenic activity. Environ. Toxicol. Chem. 2002, 21, 2034-2039. Moriyama, K.; Matsufuji, H.; Chino, M.; Takeda, M. Identification and behaviour of reaction products formed by chlorination of ethynylestradiol. Chemosphere 2004, 55, 839-847.
6092
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 16, 2005
(32) Gallard, H.; Leclercq, A.; Croue´, J. P. Chlorination of bisphenol A: kinetics and byproducts formation. Chemosphere 2004, 56, 465-473. (33) Deborde, M.; Rabouan, S.; Gallard, H.; Legube, B. Aqueous chlorination kinetics of some endocrine disruptors. Environ. Sci. Technol. 2004, 38, 5577-5583. (34) Huber, M. M.; Canonica, S.; Park, G. Y.; von Gunten, U. Oxidation of pharmaceuticals during ozonation and advanced oxidation processes. Environ. Sci. Technol. 2003, 37, 1016-1024. (35) Bader, H.; Hoigne´, J. Determination of ozone in water by the Indigo method. Water Res. 1981, 15, 449-456. (36) Hoigne´, J.; Bader, H. Rate constants of reaction of ozone with organic and inorganic compounds in water - I. dissociating organic compounds. Water Res. 1983, 17, 185-194. (37) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (•OH/O•-) in aqueous solution. J. Phys. Chem. Ref. Data 1988, 17, 513-886. (38) Kosky, P. E.; Silva, J. M.; Guggenheim, E. A. The aqueous phase in the interfacial synthesis of polycarbonates. 1. Ionic equilibria and experimental solubilities in the BPA-NaOH-H2O system. Ind. Eng. Chem. Res. 1991, 30, 462-467. (39) Maguire, R. J. Review of the persistence of nonylphenol and nonylphenolethoxylates in environments. Water Qual. Res. J. Can. 1999, 34, 37-78. (40) Clarke, E. G. C.; Moffat, A. C. In Clark’s Isolation and Identification of Drugs; The Pharmaceutical Press: London, 1986. (41) Perrin, D. D. In Physical Chemical Properties of drugs; Yalkowsky, S. H., Sinkula, A. A., Valvani, S. C., Eds.; Marcel Dekker Inc.: New York and Basel, 1980; pp 1-48. (42) von Gunten, U. Ozonation of drinking water: Part I. Oxidation kinetics and product formation. Water Res. 2003, 37, 1443-1467. (43) Lee, B. C.; Kamata, M.; Akatsuka, Y.; Takeda, M.; Ohn, K.; Kamei, T.; Magara, Y. Effects of chorine on the decrease of estrogenic chemicals. Water Res. 2004, 38, 733-739. (44) Yamamoto, Y.; Niki, E.; Shiokawa, H.; Kamiya, Y. Ozonation of organic compounds. 2. Ozonation of phenol in water. J. Org. Chem. 1979, 44, 2137-2142. (45) Legube, B.; Langlais, B.; Dore´, M. Reactions of ozone with aromatics in dilute aqueous solution: reactivity and biodegradability of oxidation products. Prog. Water Technol. 1980, 12, 553-570. (46) Huber, M. M.; Terns, T. A.; von Gunten, U. Removal of estrogenic activity and formation of oxidation products during ozonation of 17R-ethinylestradiol. Environ. Sci. Technol. 2004, 39, 51775186.
Received for review January 25, 2005. Revised manuscript received June 10, 2005. Accepted June 14, 2005. ES0501619
