Photosensitized Degradation of Bisphenol A by Dissolved Organic

Amon, R. M. W.; Benner, R. Photochemical and microbial consumption of dissolved organic carbon and dissolved oxygen in the Amazon River system. Geochi...
0 downloads 0 Views 208KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 5888-5894

Photosensitized Degradation of Bisphenol A by Dissolved Organic Matter† Y U - P I N G C H I N , * ,‡,§ P E N N E Y L . M I L L E R , | LINGKE ZENG,‡ KAELIN CAWLEY,‡ AND LINDA K. WEAVERS§ Department of Geological Sciences and Department of Civil and Environmental Engineering and Geodetic Science, The Ohio State University, Columbus, Ohio 43210, and Department of Chemistry, Rose-Hulman Institute of Technology, Terre Haute, Indiana 47803

The direct and indirect photolysis of bisphenol A (BPA) was investigated using a solar simulator in the absence and presence of dissolved organic matter (DOM). BPA degradation by direct photolysis was significantly slower than its rate in the presence of DOM. In natural waters, the direct photolytic pathway would be even less important due to light screening effects. Surprisingly, differences in the rate of indirect BPA photolysis were relatively small between DOM samples. Two of the DOM samples represented terrestrial (Suwannee River fulvic acid) and autochthonous (Lake Fryxell) geochemical endmembers. The third DOM (Fulton County, Ohio) was derived from a temperate artificial wetland. We were unable to correlate BPA photoreactivity to the structural components of DOM or its extinction coefficient at 280 nm. The addition of methanol, a hydroxyl radical scavenger, to reaction solutions slowed but did not completely quench the indirect photolysis of BPA. This observation suggests that BPA photodegrades via multiple pathways involving other transients formed by the photolysis of DOM. Competitive experiments using 2,4,6trimethylphenol also reduce the reaction rate of BPA by DOM and imply that other DOM-derived phototransients (e.g., excited triplet state DOM) are involved in the reaction. The reaction rate coefficients reported under solar-simulated irradiance in the presence of DOM are significantly faster than those reported for the microbial degradation of BPA. Thus, in natural surface waters photosensitized transformation of BPA by dissolved organic matter may be as important as biodegradation.

Introduction Bisphenol A (BPA) is used in the production of polycarbonate plastics, coatings in canning operations, and other industrial applications (1). This compound is comprised of two phenols connected by a propyl bridge. BPA has documented estrogenmimicking properties (2, 3) and has been detected in wastewater effluents (4) and, to a lesser degree, in surface †

This paper is part of the Walter J. Weber Jr. tribute issue. * Corresponding author phone: (614)292-6953; fax: (614)292-7688; e-mail: [email protected]. ‡ Department of Geological Sciences, The Ohio State University. § Department of Civil and Environmental Engineering and Geodetic Science, The Ohio State University. | Department of Chemistry, Rose-Hulman Institute of Technology. 5888

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004

waters (5). Many of the studies done to date (6-9) have shown that BPA can be degraded through aerobic microbial processes with reported half-lives of days to weeks, which may explain its relatively low abundance in the natural environment (5). Thus, it is widely believed that the fate of BPA in surface waters is largely controlled by microbial degradation pathways. For many organic pollutants, photolysis is an important transformation process in surface waters. Many of these compounds lack chromophores capable of absorbing light at wavelengths present in sunlight, but they can undergo transformation by indirect pathways in the presence of a photosensitizer. Dissolved organic matter (DOM) is ubiquitous to surface waters and known to sensitize the transformation of pollutants by generating reactive photooxidants comprised of both reactive oxygen species (ROS) (e.g., hydroxyl radicals (OH•), peroxyl radicals (ROO•), singlet oxygen (1O2), etc.) and other non-ROS transients (10-15). Thus, depending upon the target pollutant and the chromophoric nature of the DOM, a number of possible indirect photolytic pathways may exist. Photolysis of BPA may be an important but overlooked process in surface waters. In the past, alkyl phenols have been widely used as compounds for probing DOM photoreactivity (16-19). The presence of electron-donating alkyl groups make them specifically reactive to certain photoproduced transients (e.g., ROO•, 1O2, and triplet DOM (3DOM)) (10, 16). Moreover, for certain alkyl phenols the kinetics of this transformation pathway is particularly fast (t0.5 on the order of minutes to hours) (16-19). This research focuses on the phototransformation of BPA in the presence of DOM isolated from various sources. We hypothesize that BPA can undergo significant indirect photolysis in the presence of DOM and that the degradation kinetics will be dictated, in part, by the type of DOM present. Therefore, our focus in this work was to correlate the kinetics of BPA degradation to the properties of the target DOM and elucidate the possible pathways of BPA degradation through the use of transient specific scavengers.

Methods The DOM used in this study was comprised of fulvic acids from the Suwannee River (IHSS reference grade) and Lake Fryxell, Antarctica, and an ultrafiltration isolate taken from an artificial agricultural wetland located in northeast Ohio (Fulton County, OH). The Suwannee River fulvic acid (SRFA) represents DOM comprised of predominantly terrestrial precursors, while Lake Fryxell fulvic acid (LF) is derived entirely from autochthonous sources. The Fulton County DOM (FC) is unique in that it is a eutrophic wetland that also receives terrestrial organic matter inputs from the surrounding farmland. Extinction coefficients for the three DOMs were determined by normalizing absorbances measured via UVVis spectrophotometry to the total organic carbon content. Scans of reconstituted DOM in water were obtained from 200 to 600 nm on a Varian Cary 1 UV-Vis dual beam spectrophotometer, while organic carbon was measured using a Shimadzu TOC-5000. The 13C NMR derived properties of these three materials have been compiled from data reported in Mash (20). For the photolysis experiments, solutions of BPA or BPA and DOM were placed into 13 × 100 mm Pyrex reaction tubes and irradiated in a Suntest CPS+ (Atlas Devices) solar simulator. The initial BPA concentration was 10 µM for all experiments, and DOM solutions contained from 5 to 6 mg/L DOM as carbon (reflecting average surface water values). 10.1021/es0496569 CCC: $27.50

 2004 American Chemical Society Published on Web 10/07/2004

FIGURE 1. Direct and indirect bisphenol A photosensitized degradation by SRFA, LF, and FC DOM under simulated solar irradiance.

TABLE 1. Pseudo-First-Order Rate Coefficients for Direct and Indirect Phototransformation of BPA Reported ( 1 SDa

c

DOM type

[DOC] (mg/L)

kobs (s-1)

kobs,800b (s-1)

half-life (h)c

carbon normalized k*obs (s‚mg of C/L)-1

light control SRFA LF DOM FC DOM

0 5.78 6.00 4.35

(8.07 ( 0.8) × 10-7 (8.62 ( 0.2) × 10-6 (1.11 ( 0.7) × 10-5 (5.98 ( 0.2) × 10-6

not applicable (1.15 ( 0.1) × 10-5 (1.40 ( 0.1) × 10-5 (8.83 ( 0.1) × 10-6

235 22.0 17.4 32.2

not applicable 1.48 × 10-6 1.79 × 10-6 1.41 × 10-6

a Rate coefficients were corrected for light screening (22). Half-lives based on kobs.

b

kobs,800 is the rate coefficient calculated from the initial 800 min of the experiment.

Solution pH ranged from 3.72 to 5.29 (depending upon the DOM used) and is significantly lower than the reported BPA pKa values of 9.6 and 10.2 (3). Thus, we assumed that all the BPA existed in the un-ionized form. Actinometry experiments using p-nitroanisole and pyridine (21) were run concurrent with our BPA rate studies. The pyridine concentration was adjusted to gauge lamp stability over the time span of the BPA experiments (∼40 h in most cases) and revealed that our light source was stable over the duration of a time course. The measured intensity of the light source was, on average, 82% of that predicted for summer sunlight at 40° N (n ) 9; RSD ) 30%) (21), assuming the lamp output has a similar spectral distribution to sunlight. Light (target compound in water with no DOM) and dark (target compound with DOM and no light) samples were run concurrently. Dark controls were double-foil wrapped and placed within the solar simulator with the other samples. Temperature in the reactor was kept at 30° C. BPA was analyzed by reverse-phase HPLC (Waters, Inc.) with a 15 cm Waters Nova-Pak C-18 column using an isocratic acetonitrile/water mixture (45:55% v/v at 1 mL/min) and detected by fluorescence spectroscopy (Waters 470 with excitation/emission wavelengths of 225/ 305 nm, respectively). Rate coefficients for the observed reaction were calculated from a least-squares fit of the degradation data to the pseudo-first-order kinetics model. Finally, all the rate coefficients were corrected for light screening factors over the range of 290-400 nm using the method of Miller and Chin (22). The percentage of photons absorbed by the solutions was 15%, 4%, and 3% for SRFA, LF, and FC, respectively. Hydroxyl radical scavenging experiments were conducted with 5 mM MeOH as the OH• quencher. White (23) demonstrated that methanol becomes the dominant OH• scavenger at this concentration and the DOM levels used in the experiments. Moreover, we also conducted competition

kinetics experiments using 2,4,6-trimethylphenol (TMP) as a competing scavenger for DOM-derived phototransients (11, 17). Initial TMP concentrations were identical to that used for BPA. TMP was quantified using the same procedure as BPA.

Results and Discussion Reactions in DOM Solutions We observed some direct photolysis of BPA in our Milli-Q (no DOM) control experiments (Figure 1 and Table 1). Conversely, no direct dark transformation of BPA occurred in the presence of DOM alone. With respect to the direct phototransformation process, light screening by BPA itself at 10 µM was approximately 1% and corrected for self-absorption results in a half- life of roughly 9.8 d at constant irradiance in the solar simulator. Because this half-life is greater than that reported for the biodegradation of BPA (t1/2 of 3-6 d) (6), we believe that direct photolysis of BPA will be insignificant in natural surface waters. Moreover, variations in photon flux in natural systems over a diurnal cycle coupled with larger screening factors from DOM would significantly increase the half-life of BPA via this pathway. Significant BPA transformation occurred in all samples containing DOM, (Figure 1 and Table 1). All the rate data fitted reasonably well (r2 values ranged from 0.81 to 0.99 with an average for the entire data set of 0.94 for n ) 23) to the pseudo-first-order rate expression (i.e., d[BPA]/dt ) -kobs[BPA]); however, there was a noticeably faster rate of transformation in the first 12 h. Indeed fitting the data to the first 800 min of each experiment increased the pseudo-firstorder rate constants appreciably (Table 1). Reasons why indirect photolysis of BPA by DOM diminishes over time are discussed elsewhere. More importantly, the half-life for BPA degradation in the presence of DOM under natural sunlight VOL. 38, NO. 22, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5889

FIGURE 2. Effect of 5 mM methanol on the indirect photolytic rate of bisphenol A.

TABLE 2. Composition (as % of total carbon) of DOM Isolates Used in This Study As Determined by 13C NMR Spectroscopy and Extinction Coefficients (E) Measured at 280 nm 0-60a 60-90a 120-160a 160-190a 190-220a E (L/mol aliphatic I aliphatic II aromatic carboxyl ketones of C‚cm)c

DOM SRFAa FC a LF b a 13C

26.3 33.9 48.4

21.1 29.4 12.2

NMR shift in ppm.

19.0 11.0 11.9 b

15.0 14.7 20.7

3.46 1.08 4.40

From ref 20. c From ref 27.

d

468 204 216 This study.

may result in degradation rate coefficients that are significantly higher than, or at least on the order of, those reported for biodegradation. For example, we would predict BPA to have a half-life of 7.2 (( 0.3) d if irradiated by noon sunlight at 40° N for 8 h a day based upon our measured light intensity and the rate coefficient measured for the FC DOM. Depending upon the latitude of the water body, the season, and the DOM levels present in the water body, this translates into half-lives on the order of or faster than degradation by microorganisms. Surprisingly, we did not observe large differences in rates of BPA transformation between the different DOMs studied when the pseudo-first-order rate coefficients were normalized to the total organic carbon content for each experiment (Table 1). The algal derived Lake Fryxell fulvic acid exhibited the highest reactivity for BPA, while the DOM isolated by ultrafiltration from the agricultural wetland was slightly slower than the SRFA sample (Table 1). On the basis of the properties of the DOMs used by this study (Table 2), there appears to be little correlation between reactivity and DOM structure, although DOM reactivity loosely corresponded to carboxyl content. Typically aromatic and colored DOMs are very photoreactive (24-26), and we expected highly colored DOM to possess the highest reactivity toward BPA. In our case this would be the Suwannee River fulvic acid sample, which is derived from a “blackwater” environment that has significant terrestrial inputs with little organic matter derived from pelagic primary production. In general, samples reflecting terrestrial DOM sources have high aromatic contents and high molar extinction coefficients at 280 nm (29). The most reactive DOM in this study, however, was the Antarctic LF fulvic acid comprised entirely of organic matter that origi5890

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004

nated from autochthonous precursors. Autochthonous DOMs have lower molar absorptivites at 280 nm and higher aliphatic contents (29). The least reactive sample is the FC DOM, which has properties that more closely resemble the LF sample even though it is derived from a water body that receives allochthonous materials. On the basis of the similarity of the 13C NMR and extinction coefficient data between FC and LF DOMs (Table 2), the contribution of autochthonous organic matter in the FC DOM is significantly higher than the terrestrial materials entering this artificial wetland (29). Given the autochthonous contribution to the FC DOM, we expected it to exhibit a higher reactivity toward BPA degradation if one does not differentiate between different types of autochthonous materials. It is possible that the algal assemblage in the Fulton County site might have created a DOM that is less reactive than the LF material. The algal assemblage in Lake Fryxell is comprised of eukaryotic and prokaryotic photoautotrophs and heterotrophs that have developed in a unique ecosystem in Antarctica (27). These organisms are very different than organisms in a temperate agricultural wetland that receives copious amounts of nutrient-rich runoff. Thus, DOM that may look similar when analyzed by bulk methods such as 13C NMR and UV-Vis spectroscopy are actually different at the molecular level. Intense photobleaching at the FC wetland site during the summer months (our sample was collected in August) may also explain its lower reactivity. The Fulton County wetland is used to treat agricultural runoff and located in a setting devoid of shade. Moreover, it is constructed in a manner, which maximizes its hydraulic residence time. Thus, it is possible that photoreactive moieties in the DOM matrix from our sample may have been rendered less reactive over the course of the summer. Conversely, the highly pristine Lake Fryxell is ice-covered year round with the exception of a “moat” that is formed during the austral summer (27). Photobleaching would be nearly nonexistent in this system as UV-light penetration to the surface water is significantly impaired by ice cover. Chemical Probing of DOM Promoted Reactions In an effort to delineate between OH• and non-hydroxyl-mediated photopathways, we conducted competitive scavenging experiments using reactive probes. Methanol has been used in several studies as a specific OH• probe (14, 23) since it has been reported to be relatively unreactive with other ROS (14). BPA and DOM in the presence of 5 mM MeOH revealed

TABLE 3. Effect of Methanol (5 mM) on BPA’s Indirect Photolytic Degradation Rate by DOMa DOM SRFA LF DOM FC DOM a

5.91 5.60 4.35

carbon normalized k*obs,MeOH (s‚mg of C/L)-1

kobs,MeOH (s-1)

[DOC] (mg/L)

10-6

10-7

(4.31 ( 0.2) × (3.33 ( 0.2) × 10-6 (4.08 ( 0.3) × 10-6

7.29 × 5.95 × 10-7 9.38 × 10-7

kOH (s‚mg of C/L)-1 10-7

7.51 × 1.20 × 10-6 4.72 × 10-7

kOH•/kobs 0.50 0.66 0.33

The rate constant kOH• is calculated from eq 1 where kother ) kobs,MeOH for this specific scenario.

slower BPA transformation kinetics and demonstrate the importance of OH• as a reactive transient (Figure 2). In past studies, we observed complete quenching of the reactant in the presence of methanol when OH• was believed to be the principal reactive transient (22). In this study, however, we did not observe complete quenching of the BPA reaction by methanol, suggesting that other parallel reactions are important in the photofate of BPA. Under conditions of constant irradiance and assuming that reactive phototransients reach a steady-state concentration, we estimated the relative contribution of the hydroxyl radical by partitioning the overall transformation rate coefficient into a sum of the relevant pseudo-first-order rate coefficients, i.e.:

kobs ) kOH• +

∑k

other

(1)

where kobs is the overall rate coefficient, ∑kother is the rate coefficient measured for the phototransformation of BPA by DOM in the presence of methanol and accounts for the sum of all non-hydroxyl radical phototransients, and kOH• is the reaction rate coefficient attributable only to OH•, regardless of its source. Besides DOM, trace levels of transition metals such as iron may contribute to hydroxyl radical production in solutions (22, 23, 28). Our analysis of the data using both the BPA experimental results in the presence and absence of methanol revealed that 50-66% of the total transformation of BPA could be attributed to the hydroxyl radical for SRFA and LF, respectively (Table 3). The effect of OH• on BPA phototransformation by FC DOM was significantly less accounting for 33% of the overall transformation. Such calculations may actually overestimate the contribution of direct reaction of BPA with OH• to kobs because it ignores the possibility that OH• may initiate the formation of peroxyl radicals by abstracting hydrogen from DOM (11). Thus, MeOH scavenging of OH• may also decrease ∑kother in eq 1. The role of OH• in transforming BPA through photosensitized pathways by DOM appears to be highly variable and is in part due to the balance between OH• production and scavenging by DOM. In natural surface waters, the presence of other photosensitizers such as nitrate and OH• scavengers (e.g., carbonate and bicarbonate) would also influence the importance of the hydroxyl radical mediated degradation of BPA in sunlight (12). One important non-hydroxyl radical reaction pathway for alkyl phenols is through reaction with triplet state DOM. Canonica et al. (17) observed that the reaction between alkyl phenols and excited triplet state benzophenone occurs at near diffusion-controlled rates (k ) 2.6-5.6 × 109 M-1 s-1). They hypothesized that such moieties may also be responsible for the observed high degree of reactivity between 3DOM and alkyl phenols by electron abstraction from the phenol (to form a phenoxyl radical) and/or hydrogen transfer. Accurate measurements of ketone and quinone moieties by 13C NMR are difficult to make due to the poor signal-tonoise ratio and resolution of the peak. Because of these inaccuracies, we were unable to establish a clear relationship between the ketone content of our DOM samples and BPA reactivity. Experiments were conducted using TMP in an effort to elucidate the importance of 3DOM (18) and/or DOM-derived

peroxyl radicals (11) as photooxidants. Because of the limited amount of FC and LF DOM in our possession, we could only use SRFA to conduct a limited number of TMP and TMPBPA competition experiments. We chose TMP as a competitive scavenger to use with the SRFA because Canonica et al. (16) showed that photodegradation of TMP in a SRFA solution occurred almost exclusively via 3DOM and/or DOM-derived peroxyl radicals when irradiated with light g320 nm originating from a medium-pressure Hg vapor lamp. We observed fast TMP kinetics with a half-life of less than 2 h for SRFA in our light source. It is important to note that in a separate control experiment (see Supporting Information), we observed that methanol did not affect the reaction kinetics of TMP in the SRFA sample. Therefore, we believe hydroxyl radical contributes very little to TMP degradation in SRFA DOM, an observation that corroborates the results of Canonica et al. (16), who observed no effect of 1-propanol (another OH• scavenger) on the DOM photosensitized reaction of TMP and dimethoxyphenol. Thus, it appears that TMP may be transformed by reaction with 3DOM and/or DOM-derived peroxyl radicals as hypothesized by Canonica and co-workers (16-19). With respect to 1O2, several investigators (30, 31) demonstrated significant alkyphenol reactivity with humic substances that have the highest singlet oxygen quantum yield. Conversely, others (10, 16, 18) observed limited singlet oxygen reactivity to a variety of alkylphenols. In the study by Canonica et al. (16, 18), the singlet oxygen pathway was thought to account for, at most, 12% of the TMP photodegradation in the presence of SRFA, and assumed to be an unimportant reactant in their system. It is possible that 1O2 could be more important when irradiated by natural or simulated sunlight because of wavelengths not present in the light source used by Canonica et al. (16, 17). Moreover, recent work by Paul et al. (32) has shown a high degree of seasonal variability with singlet oxygen quantum yields. Typically DOM sampled in the spring generated up to four times more 1O2 than that sampled in the fall from the identical site. Thus, the limited observed reactivity between alkylphenols and 1O2 may in part be a function of temporal variability in DOM composition. We observed roughly a 28% increase in the half-life of the SRFA photosensitized transformation of BPA in the presence of TMP at an equivalent initial concentration (Figure 3) relative to experiments in the absence of TMP. Because TMP reacts so quickly in the presence of DOM, we assume that any competitive effects with BPA for the 3DOM would decrease significantly after 3 or 4 half-lives (roughly 8 h), and we expect BPA’s reaction rate to increase once TMP has been consumed. Moreover, TMP’s relatively low solubility in water (in comparison to methanol) precludes the addition of significantly greater amounts of this competitive probe relative to BPA to the point where it dominates the relevant photooxidation pathways. Nonetheless, on the basis of our limited data set, we did not observe any increase in BPA’s reaction rate beyond 8 h. These results demonstrate that one of the BPA’s transformation pathways probably involves the triplet state DOM and/or peroxyl radicals, but at this stage we cannot quantitatively determine the contribution of either pathway to BPA’s overall photosensitized transVOL. 38, NO. 22, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5891

FIGURE 3. Changes in bisphenol A half-lives in the presence of 2,4,6-trimethylphenol (TMP). The smallest bar is the half-life of TMP alone in the presence of SRFA and simulated solar irradiance.

FIGURE 4. (a) Carbon normalized rate coefficients (kobs*) determined for indirect bisphenol A photolysis in the presence of SRFA at 4, 12, and 36 h durations. Light bars are the same experiments conducted in the presence of 5 mM methanol. (b) Changes in the measured UV-Vis absorbance of SRFA over the course of the experiment.

formation. Finally, we do not know how alkyl phenoxyl radicals generated from TMP’s photooxidation by 3DOM can influence BPA’s reactivity relative to other phototransients. Possible Causes for the Observed Change in Reaction Kinetics. Previously, we reported a rapid decrease in BPA concentration in the first 800 min of irradiation followed by 5892

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004

a decrease in the reaction rate. While the pseudo-first-order fit for the entire rate data were reasonably good for all experiments, the deviation we observed for the initial rate of BPA degradation demonstrates that this model fails to adequately describe the actual kinetics of the reaction. We suspect that photobleaching of the DOM over time may

destroy some fraction of the reactive chromophores. This phenomenon could lead to transients that are no longer at steady-state, which would result in the observed deviation from first-order kinetic behavior. We conducted BPA indirect photolysis experiments using SRFA at 4, 12, and 36 h, respectively, to probe how the observed rate coefficient changes as a function of irradiation time. A significant decrease in the carbon normalized rate coefficients occurred between the 4 h experiments and those conducted for 12 or 36 h (Figure 4a, dark bars) and is consistent with our hypothesis that certain transients do not reach steady-state levels but diminish over time. The decrease in BPA rate coefficients mirrors a similar decrease in the SRFA absorbance monitored from 200 to 500 nm over the same time span (Figure 4b). We also monitored the molar extinction coefficient () at 336 nm for the SRFA and observed a decrease during the course of the reaction (see Supporting Information, Figure S2). These measurements suggest that the composition of the chromophores changed as a result of photobleaching. Overall, we believe the effect of exposing SRFA solutions for the duration of the experiment resulted in compromising its photosensitizing ability with respect to BPA’s phototransformation. A similar series of experiments were conducted in the presence of 5 mM methanol and BPA in solutions of SRFA. Surprisingly, we observed no statistical difference in the carbon normalized rate coefficients at 4, 12, or 36 h (Figure 4a, light bars). Indeed, inspection of the BPA concentrations measured in 36 h experiments in the presence of methanol (Figure 2) for all the DOM isolates do not exhibit the rapid decrease in BPA concentration that occurs within the first 12 h of irradiation when MeOH is not present (Figure 1). Thus, it appears that the photobleaching process is selective whereby those moieties in the DOM matrix responsible for the production of OH•, or from other reactive transients through reaction with OH•, are compromised over time, while BPA-reactive DOM components that are independent of OH•mediated pathways (e.g., triplet states) are more resistant to photolysis. Our results support our hypothesis that BPA can be transformed by photooxidants generated by the irradiation of DOM at wavelengths present in natural sunlight; however, we did not observe large differences in BPA reactivity for the various DOMs used in this study. The hydroxyl radical appears to be one important photooxidant initiating the observed degradation of BPA based upon our methanol quenching experiments, but its role in the phototransformation of BPA is highly dependent upon the type of DOM used. Similarly, it appears likely that direct reaction of BPA with excited triplet state DOM or peroxyl radicals may also be important. Our data suggest that BPA reaction pathways mediated by OH• may be affected by photobleaching. The results from our experiments using a solar simulator suggest that that indirect photolysis may be as important a pathway (if not more so) as aerobic microbial degradation in influencing BPA fate in natural surface waters but needs to be corroborated by fieldwork.

Acknowledgments We thank Amanda Grannas for helping us with the photochemical experiments and data interpretation. This work was supported by the NSF funded Environmental Molecular Science Institute at the Ohio State University (CHE0089147).

Supporting Information Available Text and two figures showing the lack of reactivity between TMP and OH• and changes in the molar extinction coefficient

of SRFA over the time course of the experiments. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Alexander, H. C.; Dill, D. C.; Smith, L. W.; Gulney, P. D.; Dorn, P. B. Bisphenol A: Acute aquatic toxicity. Environ. Toxicol. Chem. 1988, 7, 19-26. (2) Atkinson, A.; Roy, D. In vitro conversion of environmental estrogenic chemical bisphenol A to DNA binding metabolite(s). Biochem. Biophys. Res. Commun. 1995, 210, 424-433. (3) Staples, C.; Dorn, P. H.; Klecka, G. M.; O’Block S. T.; Harris L. H. A review of the environmental fate, effects, and exposures of bisphenol A. Chemosphere 1998, 36, 2149-2173. (4) Fromme, H.; Kuchler, T.; Otto, T.; Pilz, K.; Mueller, J.; Wenzel, A. Occurrence of phthalates and bisphenol A and F in the environment. Water Res. 2002, 36, 1429-1438. (5) Belfroid, A.; Velzen, M.; van der Horst, B.; Vethaak, D. Occurrence of bisphenol A in surface water and uptake in fish: Evaluation of field measurements. Chemosphere 2002, 49, 97-103. (6) Kang, J.-H.; Kondo, F. Bisphenol A degradation by bacteria isolated from river water. Arch. Environ. Contam. Toxicol. 2002, 43, 265-269. (7) Ike, M.; Jin, C. S.; Fujita, M. Biodegradation of bisphenol A in the aquatic environment. Water Sci. Technol. 2000, 42 (7-8), 31-38. (8) Kang, J.; Kondo, F. Effects of bacterial counts and temperature on the biodegradation of bisphenol A in river water. Chemosphere 2002, 49, 493-498. (9) Klecka, G. M.; Gonsior, S. J.; West, R. J.; Goodwin, P. A.; Markham, D. A. Biodegradation of bisphenol A in aquatic environments: River die-away. Environ. Toxicol. Chem. 2001, 20, 2725-2735. (10) Faust, B.; Hoigne, J. Sensitized photooxidation of phenols by fulvic-acid and in natural-waters. Environ. Sci. Technol. 1987, 21, 957-964. (11) Blough, N. V.; Zepp, R. G. Active Oxygen: Reactive Oxygen Species in Chemistry; Foote, C., Valentine, J., Greenberg, A., Liebman, J. F., Eds.; Chapman and Hill: New York, 1995; pp 280-332. (12) Brezonik, P. L.; Fulkerson-Brekken, J. Nitrate induced photolysis in natural waters: Controls on hydroxyl radical photointermediates by natural scavenging agents. Environ. Sci. Technol. 1998, 32, 3004-3010. (13) Mabury, S. A.; Crosby, D. G. Pesticide reactivity toward hydroxyl and its relationship to field persistence. J. Agric. Food Chem. 1996, 44, 1920-1924. (14) Zhou, X.; Mopper, K. Determination of photochemically produced hydroxyl radicals in seawater and fresh-water. Mar. Chem. 1990, 30, 71-88. (15) Zafiriou, O. C. Jousset-Dubien, J.; Zepp, R. G.; Zika, R. G. Photochemistry of natural-waters. Environ. Sci. Technol. 1984, 18, 358A-371A. (16) Canonica, S.; Jans, U.; Stemmler, K.; Hoigne., J. Transformation kinetics of phenols in water: Photosensitization by dissolved natural organic matter and aromatic ketones. Environ. Sci. Technol. 1995, 29, 1822-1831. (17) Canonica, S.; Hellrung, B.; Wirz, J. Oxidation of phenols by triplet aromatic ketones in aqueous solution. J. Phys. Chem. 2000, 104, 1226-1232. (18) Canonica, S.; Freiburghaus, M. Electron-rich phenols for probing the photochemical reactivity of freshwaters. Environ. Sci. Technol. 2001, 35, 690-695. (19) Canonica, S.; Hoigne, J. Enhanced oxidation of methoxy phenols at micromolar concentration photosensitized by dissolved natural organic material. Chemosphere 1995, 30, 2365-2374. (20) Mash, H. An assessment of electroanalytical techniques for the measurement of copper in the presence of natural aquatic organic ligands. Ph.D. Dissertation, The Ohio State University, 2001. (21) Dulin, D.; Mill, T. Development and evaluation of sunlight actinometers. Environ. Sci. Technol. 1982, 16, 815-820. (22) Miller, P. L.; Chin, Y. P. Photoinduced degradation of carbaryl in wetland surface water. J. Agric. Food Chem. 2002, 50, 67586765. (23) White, E. M. Determination of photochemical production of hydroxyl radical by dissolved organic matter and associated iron complexes in natural waters. M.S. Thesis, The Ohio State University, 2001. (24) Vialaton D.; Richard C. Phototransformation of aromatic pollutants in solar light: Photolysis versus photosensitized reactions under natural water conditions. Aquat. Sci. 2002, 64, 207-215. VOL. 38, NO. 22, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5893

(25) Gao, H.; Zepp, R. G. Factors influencing photoreactions of dissolved organic matter in a coastal river of the Southeastern United States. Environ. Sci. Technol. 1998, 32, 2940-2946. (26) Amon, R. M. W.; Benner, R. Photochemical and microbial consumption of dissolved organic carbon and dissolved oxygen in the Amazon River system. Geochim. Cosmochim. Acta 1996, 60, 1783-1792. (27) McKnight, D. M.; Aiken, G. R.; Smith, R. L. Aquatic fulvic-acids in microbially based ecosystemssresults from 2 desert lakes in Antarctica. Limnol. Oceanogr. 1991, 36, 998-1006. (28) Southworth, B. A.; Voelker, B. M. Hydroxyl radical production via the photo-Fenton reaction in the presence of a fulvic acid. Environ. Sci. Technol. 2003, 37, 1130-1136. (29) Chin, Y. P.; Aiken, G.; O’Loughlin, E. Molecular weight, polydispersity, and spectroscopic properties of aquatic humic substances. Environ. Sci. Technol. 1994, 28, 1853-1858. (30) Aguer, J. P.; Richard, C.; Andreux, F. Comparison of photoinduced properties of commercial, synthetic and soil extracted

5894

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004

humic substances. J. Photochem. Photobiol. A 1997, 103, 163168. (31) Traytnek, P. G.; Hoigne, J. Photooxidation of 2,4,6-trimethylphenol in natural waters and laboratory systems: kinetics of reaction with singlet oxygen. J. Photochem. Photobiol. A 1994, 84, 154-160. (32) Paul, H.; Hackbarth, S.; Vogt, R. D. Roeder, B.; Burnison, K. B.; Steinberg, C. E. W. Photogeneration of singlet oxygen by humic substances: Comparison of humic substances of aquatic and terrestrial origin. Photochem. Photobiol. Sci. 2004, 3, 273280.

Received for review March 3, 2004. Revised manuscript received August 10, 2004. Accepted August 17, 2004. ES0496569