Reaction of Tetrabromobisphenol A (TBBPA) with ... - ACS Publications

May 13, 2009 - Debromination of Tetrabromobisphenol A by Nanoscale Zerovalent Iron: Kinetics, Influencing Factors, and Pathways. Kunde Lin , Jiafeng D...
0 downloads 0 Views 835KB Size
Environ. Sci. Technol. 2009, 43, 4480–4486

Reaction of Tetrabromobisphenol A (TBBPA) with Manganese Dioxide: Kinetics, Products, and Pathways K U N D E L I N , * ,†,‡ W E I P I N G L I U , ‡ A N D JAY GAN† Department of Environmental Sciences, University of California, Riverside, California 92521, and Research Center of Green Chirality, Zhejiang University of Technology, Hangzhou 310032, China

Received December 19, 2008. Revised manuscript received April 29, 2009. Accepted May 1, 2009.

Birnessite (δ-MnO2) is a naturally occurring soil and sediment component that has been shown to oxidize organic compounds containing phenolic or aniline moieties. In this study, for the first time we explored the oxidation reaction of tetrabromobisphenol A (TBBPA), the most heavily used brominated flame retardant, with MnO2. TBBPA rapidly dissipated from the reaction solution and the process was accompanied by the dissolution of Mn2+. Dissipation of 50% of TBBPA occurred in less than 5 min in a system (pH 4.5) containing 625 µM MnO2 and 3.50 µM TBBPA at 21 °C, and the removal further increased to as high as 90% when the reaction was prolonged to 60 min. Analysis of initial reaction kinetics showed that the reaction orders with respect to TBBPA, MnO2, and H+ were 1.0, 0.8, and 0.25, respectively. Higher initial concentrations of TBBPA and MnO2 both enhanced the reaction. In addition, reaction rates increased as pH decreased. A retarded first-order model was found to closely describe the long-term reaction kinetics (R 2 g 0.99), from which initial half-lives of TBBPA under different reaction conditions were estimated. A total of 7 reaction products were identified and a tentative reaction scheme was proposed. This study suggests that oxidative transformation of TBBPA by MnO2 may play an important role in the natural attenuation of TBBPA. The reaction may be further optimized for treatment of TBBPA-containing wastewater or remediation of TBBPApolluted environmental matrices.

Introduction Tetrabromobisphenol A (TBBPA) is the most widely used brominated flame retardant (1). TBBPA is primarily used as a reactive flame retardant in printed circuit boards but also has additive applications in several types of polymers such as acrylonitrile butadiene styrene, epoxy and polycarbonate resins, high impact polystyrene, phenolic resins, and adhesives. The global market demand for TBBPA was over 1.3 × 108 kg in 2002 and 1.7 × 108 kg in 2004 (2). The widespread use of TBBPA has apparently caused its ubiquitous occurrence in the environment. For instance, TBBPA has been frequently detected in air (3-5), dust (5-7), sewage sludge and sediment (8, 9). Furthermore, TBBPA has been detected in serum from computer technicians, electronic assembly * Corresponding author tel: 951-827-3860; fax: 951-827-3993; e-mail: [email protected]. † University of California, Riverside. ‡ Zhejiang University of Technology. 4480

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 12, 2009

workers, laboratory personnel, and the general population (10-12). Although information pertinent to the potential risk of TBBPA is limited, studies have shown that TBBPA can act as an endocrine disruptor (13-16). Given the continuously increasing global use of TBBPA, there is a heightened concern regarding its environmental fate. Owing to its low water solubility (63 µg/L) and high octanol/water partition coefficient (log KOW ) 5.9) (17), TBBPA is expected to sorb onto soil or sediment particles following release into the environment. It is therefore essential to delineate and understand attenuation processes of TBBPA in soils and sediments. TBBPA has been found to partially degrade in soils, sediments, and water under aerobic and anaerobic conditions, with a mean half-life of about 2 months (18-20). Under anaerobic conditions, TBBPA may be reductively dehalogenated to bisphenol A (21-26). Additionally, George and Ha¨ggblom (27) showed that TBBPA could be O-methylated to its mono- and dimethyl ether derivatives by microorganisms in different sediments. While biodegradation of TBBPA in soils or sediments has been well studied, knowledge regarding its abiotic transformations in the environment is rather obscure. The birnessite group of hydrous manganese oxides with an empirical formula of [Na, Ca, MnII] Mn7O14 · 2.8H2O are among the most important naturally occurring reactants or catalysts in soils and sediments (28). For example, with a reduction potential of 1.23 V, δ-MnO2 is capable of selectively oxidizing organic contaminants containing phenolic or aniline moieties (29-36). Li et al. (37) showed that soil Mn (III/IV) oxides contributed greatly to immobilization of amines within organic matters and formation of large aromatic amine polymers. The general postulated reaction mechanism involves the formation of a surface precursor complex on manganese oxide surface followed by electron transfer from an organic reductant to the transitional metal and release of an organic cationic radical (32-35). The addition of an electron-donating group (e.g., -OCH3) on the aromatic ring enhances oxidation by facilitating the electron transfer within the complexes, while addition of an electron withdrawing group (e.g., -NO2) would hinder the electron transfer process (34-36). The primary objective of this study was to explore the reaction of TBBPA with δ-MnO2 by evaluating oxidation kinetics as a function of TBBPA concentrations, pH, and MnO2 concentrations. An empirical retarded first-order model was employed to describe long-term reaction kinetics. Furthermore, the principal reaction products were identified and a detailed reaction scheme was proposed. This study represents the first example where the reaction between TBBPA and manganese oxides is studied. The information will improve our understanding of abiotic processes contributing to TBBPA attenuation in the natural environment.

Materials and Methods Chemicals. Analytical standard of TBBPA (97%) was purchased from Sigma-Aldrich (St. Louis, MO). The other chemicals (purity >97%) were purchased from Fisher Scientific or Sigma-Aldrich and used as received. Reagent water (18.3 MΩ-cm resistivity) was prepared using a Barnstead Nanopure water system (Barnstead/Thermolyne, Dubuque, IA). Stock solution of 1.75 mM TBBPA was prepared in HPLC grade methanol and stored at 4 °C prior to use. Synthesis and Characterization of δ-MnO2. Manganese oxide (δ-MnO2) was synthesized according to Murray’s method (38). Briefly, 320 mL of 0.1 M NaMnO4 and 640 mL of 0.1 M NaOH were added to 6.56 L of nitrogen purged 10.1021/es803622t CCC: $40.75

 2009 American Chemical Society

Published on Web 05/13/2009

FIGURE 2. Effect of methanol content in reaction solutions on TBBPA removal from the aqueous by MnO2. Experimental conditions: initial concentrations of MnO2 and TBBPA were 625 µM and 3.5 µM, respectively, pH 4.5, and temperature 22 °C. Data were from reactions quenched by L-ascorbic acid and given as means ( standard errors (n ) 3).

FIGURE 1. Typical time courses of TBBPA reaction with MnO2. (A) Disappearance of TBBPA and formation of products 2 and 4; and (B) Time dependence of total and free TBBPA in reaction mixture. Reaction conditions: (A) [TBBPA]0 ) 17.5 µM, [MnO2]0 ) 1250 µM; (B) [TBBPA]0 ) 3.50 µM, [MnO2]0 ) 625 µM. All experiments were conducted in 10 mM sodium acetate solutions (pH 4.5) at 22 ( 1 °C. Data points are given as means ( standard errors (n ) 3). reagent water, followed by a dropwise addition of 480 mL of 0.1 M MnCl2 while keeping the solution constantly sparged with nitrogen. The formed MnO2 particles were allowed to settle, and the supernatant was decanted and replaced with fresh reagent water several times until the conductivity of supernatant was below 2 µS cm-1. The MnO2 suspensions were stored at 4 °C and were diluted to appropriate concentrations prior to use. A portion of the synthetic MnO2 was air-dried and its surface area was determined to be 236 m2 g-1 by the Brunauer-Emmett-Teller method of N2 adsorption (Micromeritics 2100, Norcross, GA). Power X-ray diffraction analysis showed that the synthetic MnO2 is amorphous (data not shown), consistent with the large surface area determined. Reaction Setup. Experiments were conducted in 100-mL amber borosilicate glass bottles under ambient O2 conditions in the absence of direct sunlight and at room temperature 22 ( 1 °C. Reaction mixtures (40 mL for initial reaction kinetics experiment, 80 mL for the other experiments) were constantly stirred with Teflon-coated magnetic stir bars at 480 rpm. Preliminary MnO2-free control studies showed that only about 50% of spiked TBBPA (at 1.75 µM) could be recovered when using reagent water or a mixture of water/methanol 60/40 (v/v) to prepare the reaction solutions, likely due to the poor water solubility of TBBPA and its adsorption to

glassware. A mixture of water/methanol 50/50 (v/v) provided >95% recoveries and was subsequently employed to prepare reaction solutions. Reaction solutions were maintained at constant pH with 10 mM buffer: acetic acid/sodium acetate for pH 4-6, 4-morpholinepropanesulfonic acid and its sodium salt for pH 6.5-7.5, and 2-(cyclohexylamino)ethanesulfonic acid and its sodium salt for pH 8.6. A certain amount of NaCl was also added to the solution to maintain the reaction solution at a constant ionic strength (0.01 M). Reactions were initiated by adding a small amount of concentrated MnO2 solution into a continuously stirred buffer solution containing TBBPA. For kinetics experiments, aliquots of 1.0 mL of reaction mixture were periodically withdrawn, transferred to 2-mL HPLC vials containing 5 µL of 50 mg mL-1 L-ascorbic acid solution, and the mixture was immediately vortexed for 10 s on a Genie 2 Fisher Vortex (Fisher Scientific, Bohemia, NY). The typical quench times for 1250, 625, 312, and 156 µM MnO2 were about 5, 3, 2, and 2 s, respectively. Dissolution of MnO2 released absorbed TBBPA and reaction products. One set of experiments was quenched by centrifuging the collected samples at 10500g for 2 min on an Eppendorf microcentrifuge 5415D (Brinkmann Instruments, Westbury, NY) and the supernatants were transferred to HPLC vials for analysis. All samples were analyzed within 24 h. Samples quenched by L-ascorbic acid and centrifugation were used to determine total and free concentrations of TBBPA, respectively. All treatments were conducted in triplicates. Chemical Analysis. Disappearance of TBBPA over time in the reaction mixture was determined using a reverse-phase HPLC (Dionex, Sunnyvale, CA) coupled with a PDA-100 photodiode array detector (Dionex) with absorbance detection at 207 nm. A Dionex Acclaim 120 C18 column (4.6 × 250 mm, 5 µm) was employed for the separation. The isocratic mobile phase consisted of 75% acetonitrile and 25% 60 mM acetic acid solution at a flow rate of 1 mL min-1. The injection volume was 25 µL. The ultraviolet (UV) spectra of TBBPA and reaction products were scanned by the online PDA detector. Under these conditions, the typical retention time for TBBPA in the HPLC system was 8.9 min. For Mn2+ ion measurement, 5-mL samples were withdrawn and passed through a Whatman 0.2 µM polyvinylidene fluoride membrane filter (Whatman, Florham Park, NJ). The filtrate was then acidified to contain 1% HNO3 and the VOL. 43, NO. 12, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4481

reagent, the vial was crimped with a cap with Teflon-lined septum and the silylation reaction was held for 2 h at 60 °C. Benzene (400 µL) was added into the vial to dissolve the silylated products. Aliquots (2-µL) of silylated and nonsilylated extract were analyzed on an Agilent 6890 gas chromatograph (GC) coupled with an Agilent 5975 mass spectrometer (MS) (Agilent Technologies, Wilmington, DE) for chemical structural analysis. Further operation parameters for GC-MS analysis are given in Text S1 in the Supporting Information. Details on the identification of PA and PB shown in the HPLC analysis are given in Text S2.

Results and Discussion

FIGURE 3. Determination of reaction orders with respect to (A) TBBPA and (B) MnO2 in 10 mM sodium acetate buffered solutions (pH 4.5) at 22 ( 1 °C. Data points are given as means ( standard errors (n ) 3). All regression coefficients are significant with p < 0.05. emission of Mn2+ was determined using an atomic absorption spectrometer (Perkin-Elmer AAnalyst 800, Ueberlingen, Germany). For reaction product identification, 17.5 µM TBBPA and 625 µM MnO2 were reacted in 100 mL of pH 4.50 sodium acetate solutions (methanol/H2O, 50/50, v/v), and the reaction was quenched after 15 min, 4 h, or 24 h by adding L-ascorbic acid. The quenched solution was subsequently extracted with 50 mL of methylene chloride and 50 mL of benzene. The combined organic phases were dehydrated by passing through a filter paper filled with anhydrous sodium sulfate, and concentrated to near dryness on a vacuum rotary evaporator. The residue was redissolved in 1.0 mL of benzene and transferred to a 1.5-mL vial. For derivatization, the final extract was further evaporated to dryness under a gentle nitrogen stream, and 100 µL of silylation reagent N,Obis(trimethylsilyl)trifluoroacetamide with trimethylchlorosilane (BSTFA+TMCS, 99:1) was added to silylate polar products. Immediately after the addition of the silylation 4482

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 12, 2009

Reaction of TBBPA with MnO2. A typical reaction profile of TBBPA with MnO2 is given in Figure 1. No appreciable amount of TBBPA was degraded in the absence of MnO2 under the used experimental conditions in 24 h (data not shown). However, TBBPA was rapidly transformed in the presence of MnO2. For example, 50% of TBBPA was removed in less than 5 min under experimental conditions of pH 4.5 and 3.5 µM TBBPA and 625 µM MnO2 initial concentrations. Up to 90% of TBBPA disappeared as the reaction was prolonged to 60 min (Figure S1). The disappearance of TBBPA over reaction time apparently deviated from a pseudo-first-order kinetics (Figure 1). Reactions of MnO2 with other organic compounds also commonly deviated from the first-order kinetics (29-36). It is generally accepted that the oxide surface changes over the course of reaction due to adsorption of reaction products and shift in the distribution of surface sites toward those that are less reactive (29-36). Adsorption of reaction products on the oxide surface may limit the accessibility of oxide surface for organic reactants. In this study, the bound amount of TBBPA on oxide surfaces decreased as reaction proceeded, as it was evident that the free amount of TBBPA in reaction solution steadily increased from about 48% at 0.5 min to 93% at 60 min (Figure 1B). Previous studies showed that the oxidation of organic compounds by MnO2 was initiated by the formation of a precursor complex between the organic reductant and the oxide bound MnIV, and followed by electron transfer (29-36). In addition, the redox potential of MnO2 decreased as more Mn2+ was continuously formed and released into the solution. Transformation of TBBPA was apparently accompanied by the reduction of MnIV to Mn2+. Overall, Mn2+ concentrations declined during the first few minutes and then steadily increased until the reaction was quenched (Figure 1A). A similar Mn2+ formation pattern was observed in the oxidation of oxytetracycline by a hydrous manganese oxide (33). Interestingly, the observed minimum Mn2+ concentrations occurred right after the detected maximum product concentrations in both this (Figure 1A) and a previous study (33). Cosolvent Effect. The amount of methanol added in the reaction solutions apparently influenced the removal of TBBPA by MnO2. After 30 min of reaction, more than 80% percent of TBBPA was removed in solutions with 50% of methanol, while it decreased to approximately 50% and 10% for solutions containing methanol at 75% and 100%, respectively (Figure 2). Such an inhibitive effect by methanol may be attributable to the fact that methanol in the solutions affects the dissociation of TBBPA, and thus hinders the electron transfer process to be discussed below. Therefore, the reactivity of TBBPA with MnO2 in solutions without methanol may be stronger than that measured in this study. Initial Reaction Kinetics. Initial reaction rates were determined from the slope of the TBBPA disappearance curve by using the first few data points for which the reactant concentrations changed linearly (35). The initial reaction order was determined with respect to TBBPA and MnO2 at

TABLE 1. Retarded Rate Model Parameters and Estimated Initial Half-Lives for the Reactions of TBBPA with MnO2 in 10 mM Sodium Acetate Solutions (pH 4.5) at 22 ± 1 °C (Data are Given as Means ± Standard Errors (n = 3))

a

[TBBPA]0, µMa

kinit, min-1

3.50 1.75 0.85 0.42

0.510 ( 0.022 0.522 ( 0.048 0.567 ( 0.044 0.540 ( 0.073

3.50 1.75 0.85 0.42

0.223 ( 0.018 0.230 ( 0.012 0.243 ( 0.013 0.271 ( 0.034

3.50 1.75 0.85 0.42

0.110 ( 0.010 0.116 ( 0.012 0.127 ( 0.010 0.131 ( 0.013

3.50 1.75 0.85 0.42

0.036 ( 0.007 0.040 ( 0.009 0.045 ( 0.007 0.047 ( 0.009

Initial concentration of TBBPA.

b

(, min-1 [MnO2]0 ) 1250 µMc 0.387 ( 0.012 0.435 ( 0.029 0.467 ( 0.026 0.440 ( 0.041 MnO2]0 ) 625 µM 0.247 ( 0.014 0.269 ( 0.010 0.317 ( 0.012 0.418 ( 0.038 [MnO2]0 ) 312 µM 0.224 ( 0.014 0.234 ( 0.016 0.284 ( 0.016 0.308 ( 0.023 [MnO2]0 ) 156 µM 0.076 ( 0.011 0.102 ( 0.016 0.140 ( 0.016 0.179 ( 0.024

t1/2,init, minb

0.99 0.99 0.99 0.99

1.79 1.80 1.65 1.73

0.99 0.99 0.99 0.99

4.68 4.64 4.64 4.58

0.99 0.99 0.99 0.99

13.84 13.03 13.07 13.32

0.99 0.99 0.99 0.99

43.69 47.61 54.57 72.69

Initial half-life is calculated as t1/2,init ) (2R/kinit - 1)/(R). c Initial concentration of MnO2.

pH 4.5 (Table S1). Reaction order for TBBPA transformation at pH 4.5 was calculated to be 1.0 and 0.8 with respect to TBBPA and MnO2, respectively (Figure 3). The initial reaction rate constant at pH 4.5, kinit,pH4.5, was 9.88 ((0.36) × 10-4 µM-0.8 · min-1 (n ) 16) as calculated from kinit,pH 4.5 )

model fit (R 2)

-rinit [TBBPA]1.0 × [MnO2]0.8

(1)

The similar reaction orders for TBBPA and MnO2 suggest that they are consumed at a nearly equal rate and around two electrons were transferred to MnO2 for each molecule TBBPA transformed or Mn2+ ion released. The observed nonelementary reaction order for MnO2 suggests that the concentration terms were not directly related to stoichiometry in the reaction. Noninteger reaction orders were also observed previously in the reactions of other organic compounds with MnO2 (29-36). Zhang and Huang (30) suggested that the noninteger reaction order was likely due to the weak sorption limiting precursor complex formation. The initial reaction rate was further pH dependent (Figure S2). Overall, the initial reaction rates decreased with increasing pH in the range of pH 4.5-8.6. The apparent pH dependence of the initial reaction rate may be attributed to the fact that pH influences sorption of TBBPA to oxide surfaces and the electron transfer processes. On the other hand, pH influences the redox potential of MnO2 (1/2MnO2(s) + 2 H+ + e f 1/2 Mn2+(aq) + H2O). For example, when pH was decreased from 8.0 to 4.0, the reduction potential for MnO2 increased from 0.76 to 0.99 V (39). In addition, pH governs the speciation distribution of TBBPA. For instance, undissociated TBBPA dominates at pH < 6.0, the monoanion is the primary form between pH 7.5 and 8.5, while dibasic TBPPA is prevalent at pH > 8.5 (Figure S2). The reaction potential of TBBPA with MnO2 may vary with different TBBPA species. For instance, the protonated form (ArOH) is considerably less reactive than the phenolate anion (ArO-) in the reaction of substituted phenols by MnO2 (34). These combined effects may have contributed to the observation that the initial reaction rate of TBBPA with MnO2 at pH 6.5 did not decrease as expected. Based on the reaction orders determined with respect to TBBPA, MnO2, and H+, the overall reaction rate of TBBPA

oxidation by MnO2 between pH 4.5 and 6.0 can be described by the following equation: rinit )

d[TBBPA] ) -kinit[TBBPA]1.0[MnO2]0.8[H+]0.25 dt (2)

From this rate law, the third-order rate constant kinit was calculated to be 4.52 ((0.54) × 10-4 µM-1.05 · min-1 (n ) 12). Long-Term Reaction Kinetics. Although the initial reaction kinetics can provide some important information on the reaction mechanism, it fails to predict the disappearance of TBBPA for an extended period that has a time scale of environmental relevance. Generally, the reaction rate of TBBPA with MnO2 decreased with time when the reactions were allowed to proceed for several min or longer (Figure 1). To describe the reaction of TBBPA with MnO2 under declining rate conditions, the data were fitted into a retarded firstorder rate equation (40): Ct ) C0(1 + Rt)-kinit/R

(3)

where C0 and Ct (µM) are TBBPA concentrations in the initial (time zero) and time t reaction mixtures, respectively. The initial rate constant kinit is analogous to a first-order rate constant, and the retardation factor R describes the decline of the reaction rate with time (40). A more detailed explanation of the retarded first-order model is given in Text S3. According to eq 3, the initial half-life t1/2,init of TBBPA in the reaction solution can be calculated as t1/2,init )

2R/kinit - 1 R

(4)

It should be noted that, unlike the half-life calculation from the first-order model, t1/2,init here can just be used to estimate the initial first half disappearance of TBBPA because of the decline of the reaction rate over time. All reactions of TBBPA with MnO2 under the experimental conditions employed in this study were well fit to the retarded rate model with correlation coefficient (R2) being >0.98 (Table 1 and Figure S3). Using the generated model parameters kinit and R, the estimated initial half-lives for TBBPA under various conditions are summarized in Table 1. It is evident that the VOL. 43, NO. 12, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4483

FIGURE 4. Proposed reaction schemes for oxidation of TBBPA by MnO2. The symbol “>” denotes bonds between surface metal centers and the oxide lattices. initial half-lives of TBBPA were strongly MnO2 concentration dependent while initial TBBPA concentrations had minor impact on t1/2, init. For instance, the initial half-lives of TBBPA with concentrations ranging from 0.42 to 3.50 µM in the presence of 312 µM MnO2 were less than 14 min, and the initial half-lives were further reduced to less than 2 min when MnO2 concentration was increased to 1250 µM. However, the effect of initial TBBPA concentration on the reaction rate became as prominent as the initial concentration of MnO2 as the reactions proceeded (Figure S3). For instance, the 4484

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 12, 2009

estimated time for 95% disappearance of TBBPA in reaction with 156 µM MnO2 in pH 4.5 solutions was increased from about 122 h for initial TBBPA concentration of 3.5 µM to 2802 h for initial TBBPA concentration of 0.42 µM. Overall, at a fixed initial concentration of MnO2, R and kinit increased as initial TBBPA concentration decreased. Likewise, R and kinit increased as initial MnO2 concentration increased when initial TBBPA concentration was kept unchanged. In addition, kinit correlated well with the initial concentrations of MnO2 (R2 > 0.99) (Figure S4), suggesting

that a higher initial concentration of MnO2 accelerated the oxidation of TBBPA. On the other hand, a faster reaction caused a quicker loss of reactive sites on MnO2 surface and the release of Mn2+, which in turn inhibited the oxidation process. Therefore, a faster retarded reaction was observed, as evidenced by the higher retarded factor R shown in Table 1. Therefore, a lower initial TBBPA concentration or a higher initial MnO2 concentration might facilitate the deviation of the reaction from the first-order model. Similar to this study, long-term kinetics of oxidation of oxyteracycline with MnO2 were also successfully described by eq 3 (33). These findings together suggest that the empirical retarded rate model may be used to describe the long-term kinetics reaction of MnO2 with organic compounds in general. Reaction Products and Scheme. Reaction of TBBPA with MnO2 formed several transformation products. Two major products (designated as PA and PB) were detected in HPLCUV system (Figure S5A). Both products of PA and PB exhibited similar UV spectra to that of TBBPA, by showing a small shift in their major absorbance peak near 207 nm (Figure S5B-D). We further monitored the evolution of PA and PB by semiquantitatively calculating their concentrations via the standard calibration curve of TBBPA (Figure 1A). It was evident that PA and PB exhibited similar evolution profiles. Both PA and PB concentrations increased as the reaction proceeded and reached their corresponding plateaus at 12 min, after which both PA and PB declined slowly. The decline of PA and PB suggested that both products were further transformed to other products. It should be noted that the actual amounts of PA and PB were unknown because the response factors of PA and PB likely differed from that of TBBPA. Further characterization of PA and PB with GC-MS verified that PA and PB were products 2 and 4 (Figure 4), respectively (Figure S6H and I). To gain more chemical structure information on the reaction products, reaction solution was subsequently extracted and possible polar products were silylated with BSFTA, and then subjected to GC-MS analysis. The silylated derivatives are formed by the replacement of the active hydrogen (-OH) in phenolic reaction products with the trimethylsilyl group -Si(CH3)3. One product (retention time 29.56 min) was identified in the nonsilylated extract (Figure S6A), while six products (retention times at 34.12, 34.51, 39.25, 62.94, 67.11, and 68.13 min, respectively) were identified in silylated derivative forms (Figure S6B-G). By analysis of the MS spectra, the detected products were identified to be 1 and silylated derivatives of 2-7 in Figure 4, respectively. Authentic standards of products 1-7 are not available. Therefore, structurally related compounds 2,5-dibromohydorquinone and TBBPA were used as references to interpret the MS spectra (Figure S6J and K). Detailed major ion fragments of the detected compounds are assigned in the inset of Figure S6. The amount of products detected varied with reaction time. For example, products 5 and 7 were not detected at 15 min, while all the products were detected at 4 and 24 h. Because of the lack of certificated standards, we could not quantify the absolute amounts of the formed products. On the basis of the identified oxidative products and similar reactions of MnO2 with other organic compounds (28-36), tentative pathways for the reaction of TBBPA with MnO2 are proposed (Figure 4). Initially TBBPA is bound to the surface MnIV to produce a precursor complex. The phenol moiety of TBBPA is then oxidized by MnIV to lose an electron, forming a phenoxy radical (pathway I). The phenoxy radical is stabilized by resonance and forms R1 (pathway II). Radical R1 undergoes β-scission and releases a new radical R2 and an alkene 1 (pathway III). Further oxidation of R2 by MnO2 generates 3 (pathway VI). Radical couplings are an important reaction in the formation of other products. For instance, coupling of R1 forms 6 (pathway IX). Coupling of R1 with R2

forms 6 and R3 by the elimination of tertiary carbonation (pathway IV) (41). Subsequently, the cationic intermediate R3 is subjected to a series of substitution or elimination reactions and generates more other products. For example, as indicated in pathway VII, R3 may substitute a proton of water to form 4. The carbonation intermediate R3 may also eliminate a proton to form 2 (pathway V). Such substitutions can further occur on products from above reactions. For example, 2 and 3 from pathways V and VI may further react with R3 and form 5 and 7, respectively (pathways VIII and X). Environmental Implications. This study demonstrated high susceptibility of TBBPA to MnO2 oxidation, a mineral oxide ubiquitously present in soils and sediments. The disappearance rate of TBBPA depended strongly on the oxide surface sites available for reaction and pH of the reaction solution. Although the general abundance of manganese oxides in the environment such as soils and sediments is lower than that employed in this study, enrichment can occur in the vicinity of oxic/anoxic interfaces. For example, the average content of manganese oxide in sediment/water interface from Lake Zurich was as high as 7 µM (42). Like the inhibitive effect of Mn2+ on the reaction rate, other organic cosolutes (e.g., humic acid) and cations (e.g., Ca2+, Zn2+, etc.) may hinder the reaction of MnO2 with organic compounds (33, 36). Therefore, we cannot simply extrapolate our results to what might happen in the natural environment. Nevertheless, this study suggests that the abiotic degradation by MnO2 oxidation in soils and sediments may significantly impact the environment fate of TBBPA, contributing to the overall environmental attenuation of this important anthropogenic contaminant. The rapid transformation of TBBPA by MnO2 suggests that the process might be potentially used for treatment of TBBPA-contaminated wastewater or remediation of polluted environmental matrices. As the reactions proceed, it is expected that most of intermediates may further react and form polymer-like products, which may reduce their bioavailability and toxicity. However, given the complexity of the reaction and the formation of multiple intermediates, it is important to better understand the toxicity and fate of the predominant reaction products before the reaction is considered for decontamination treatment.

Acknowledgments This work was supported by the California State Water Resource Control Board.

Supporting Information Available Text S1-S3, Table S1, and Figures S1-S6 as described in this paper. This information is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Birnbaum, L. S.; Staskal, D. F. Brominated flame retardants: Cause for concern. Environ. Health Persp. 2004, 112, 9–17. (2) Bromine Science and Environmental Forum (SBEF). Available at http://www.bsef.com. (3) Sjodin, A.; Carlsson, H.; Thuresson, K.; Sjolin, S.; Bergman, A.; Ostman, C. Flame retardants in indoor air at an electronics recycling plant and at other work environments. Environ. Sci. Technol. 2001, 35, 448–454. (4) Tollback, J.; Crescenzi, C.; Dyremark, E. Determination of the flame retardant tetrabromobisphenol A in air samples by liquid chromatography-mass spectrometry. J. Chromatogr. A 2006, 1104, 106–112. (5) Abdallah, M.; Harrad, S.; Covaci, A. Hexabromocyclododecanes and tetrabromobisphenol-A in indoor air and dust in Birmingham, UK: implications for human exposure. Environ. Sci. Technol. 2008, 42, 6885–6861. (6) Santillo, D.; Labunska, I.; Davidson, H.; Johnston, P.; Strutt, M.; Knowles, O. Consuming Chemicals: Hazardous chemicals in VOL. 43, NO. 12, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4485

(7)

(8)

(9)

(10)

(11)

(12) (13)

(14)

(15)

(16)

(17) (18) (19)

(20) (21)

(22)

4486

house dust as an indicator of chemical exposure in the home; Technical Note 01/2003; GRL-TN-01-2003; Greenpeace Research Laboratories, 2003; Available at http://www.greenpeace.to/ publications/housedust_uk_2003.pdf. Takigami, H.; Suzuki, G.; Hirai, Y.; Sakai, S. Comparison of brominated flame retardants in indoor air and dust samples from two homes in Japan. Organohalogen Compd. 2007, 69, 2785–2788. Morris, S.; Allchin, C.; Zegers, B. N.; Haftka, J. H.; Boon, J. P.; Belpaire, C.; Leonards, P. G.; Van Leeuwen, S. J.; de Boer, J. Distribution and fate of HBCD and TBBP-A flame retardants in North Sea estuaries and aquatic food webs. Environ. Sci. Technol. 2004, 38, 5497–5504. Chu, S.; Haffner, G.; Letcher, R. Simultaneous determination of tetrabromobisphenol A, tetrachlorobisphenol A, bisphenol A and other halogenated analogues in sediment and sludge by high performance liquid chromatography-electrospray tandem mass spectrometry. J. Chromatogr. A 2005, 1097, 25–32. Hagmar, L.; Jakobsson, K.; Thuresson, L.; Rylander, A.; Sjo¨din, A.; Bergman, Å. Computer technicians are occupationally exposed to polybrominated diphenyl ethers and tetrabromobisphenol-A. Organohalogen Compd. 2000, 47, 202–205. Jakobsson, K.; Thuresson, K.; Rylander, L.; Sjo¨din, A.; Hagmar, L.; Bergman, Å. Exposure to polybrominated diphenyl ethers and tetrabromobisphenol-A among computer technicians. Chemosphere 2002, 46, 709–716. Thomsen, C.; Lundanes, E.; Becher, G. Brominated flame retardants in plasma samples from three different occupational groups in Norway. J. Environ. Monit. 2001, 3, 66–370. Meerts, I. A. T. M.; van Zanden, J. J.; Luijks, E. A. C.; van LeeuwenBol, I.; Marsh, G.; Jakobsson, E. Potent competitive interactions of some brominated flame retardants and related compounds with human transthyretin in vitro. Toxicol. Sci. 2000, 56, 95– 104. Kitamura, S.; Jinno, N.; Ohta, S.; Kuroki, H.; Fujuimoto, N. Thyroid hormonal activity of the flame retardants tetrabromobisphenol A and tetrachlorobisphenol A. Biochem. Biophys. Res. Commun. 2002, 293, 554–559. Kitamura, S.; Suzuki, T.; Sanoh, S.; Kohta, R.; Jinno, N.; Sugihara, K.; Yoshihara, S.; Fujimoto, N.; Watanabe, H.; Ohta, S. Comparative study of the endocrine-disrupting activity of bisphenol A and 19 related compounds. Toxicol. Sci. 2005, 84, 249–259. Olsen, C. M.; Meuseen-Elholm, E. T. M.; Samuelsen, M.; Holme, J. A.; Hongslo, J. K. Effects of the environmental oestrogens bisphenol A, tetrachlorobisphenol A, tetrabromobisphenol A, 4-hydroxybiphenyl and 4,42-dihydroxybiphenyl on estrogen receptor binding, cell proliferation and regulation of estrogen sensitive proteins in the human breast cancer cell line MCF-7. Pharmacol. Toxicol. 2003, 92, 180–188. European Union. Risk Assessment Report TBBP-A, Part II-Human Health; 2006. Fackler, P. Bioconcentration and elimination of 14C-residues by Eastern oysters (Crassostrea virginica) exposed to tetrabromobisphenolA; Springborn Life Sciences, Inc.: Wareham, MA, 1989. Fackler, P. Bioconcentration and elimination of 14C-residues by fathead minnows (Pimephales promelas) exposed to tetrabromobisphenol A; Springborn Life Sciences, Inc.: Wareham, MA, 1989. Fackler, P. Determination of the biodegradability of tetrabromobisphenol A in soil under aerobic conditions; Springborn Life Sciences, Inc.: Wareham, MA, 1989. Voordeckers, J. W.; Fennell, D. E.; Jones, K.; Ha¨ggblom, M. M. Anaerobic biotransformation of tetrabromobisphenol A, tetrachlorobisphenol A, and bisphenol A in estuarine sediments. Environ. Sci. Technol. 2002, 36, 696–701. Ravit, B.; Ha¨ggblom, M. M.; Ehrenfeld, J. G. Salt marsh rhizosphere affects microbial transformation of the widespread halogenated contaminant tetrabromobisphenol A (TBBPA). Soil Biol. Biochem. 2005, 37, 1049–1057.

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 12, 2009

(23) Ronen, Z.; Abeliovich, A. Anaerobic-aerobic process for microbial degradation of tetratbromobisphenol-A. Appl. Environ. Microbiol. 2000, 66, 2372–2377. (24) Arbeli, Z.; Ronen, Z. Enrichment of a microbial culture capable of reductive debromination of the flame retardant tetrabromobisphenol-A, and identification of the intermediate metabolites produced in the process. Biodegradation 2003, 14, 385– 395. (25) Gerecke, A. C.; Giger, W.; Hartmann, P. C.; Heeb, N. V.; Kohler, H.-P.; Schmid, P.; Zenneg, M.; Kohler, M. Anaerobic degradation of brominated flame retardants in sewage sludge. Chemosphere 2006, 64, 311–317. (26) Arbeli, Z.; Ronen, Z.; Diaz-Baez, M. C. Reductive dehalogenation of tetrabromobisphenol-A by sediment from a contaminated ephemeral streambed and an enrichment culture. Chemosphere 2006, 64, 1472–1478. (27) George, K. W.; Ha¨ggblom, M. M. Microbial O-methylation of the flame retardant tetrabromobisphenol-A. Environ. Sci. Technol. 2008, 42, 5555–5561. (28) Post, J. E. Manganese oxide minerals: Crystal structures and economic and environmental significance. Proc. Natl. Acad. Sci. 1999, 96, 3447–3454. (29) Zhang, H. C.; Chen, W. R.; Huang, C. H. Kinetic modeling of oxidation of antibacterial agents by manganese oxide. Environ. Sci. Technol. 2008, 42, 5548–5554. (30) Zhang, H. C.; Huang, C. H. Reactivity and transformation of antibacterial N-oxides in the presence of manganese oxide. Environ. Sci. Technol. 2005, 39, 593–601. (31) Zhang, H. C.; Huang, C. H. Oxidative transformation of fluroquinolone antibacterial agents and structurally related amines by manganese oxide. Environ. Sci. Technol. 2005, 39, 4474–4483. (32) Zhang, H. C.; Huang, C. H. Oxidative transformation of triclosan and chlorophene by manganese oxides. Environ. Sci. Technol. 2003, 37, 2421–2430. (33) Rubert, K. F.; Pedersen, J. A. Kinetics of oxyteracycline reaction with a hydrous manganese oxide. Environ. Sci. Technol. 2006, 40, 7216–7221. (34) Stone, A. T. Reductive dissolution of manganese (III/IV) oxides by substituted phenols. Environ. Sci. Technol. 1987, 21, 979– 988. (35) Laha, S.; Luthy, R. G. Oxidation of aniline and other primary aromatic amines by manganese dioxide. Environ. Sci. Technol. 1990, 24, 363–373. (36) Klausen, J.; Haderlern, S. B.; Schwarzenbach, R. P. Oxidation of substituted anilines by aqueous MnO2: effect of co-solutes on initial and quasi-steady-state kinetics. Environ. Sci. Technol. 1997, 31, 2642–2649. (37) Li, H.; Lee, L. S.; Schulze, D. G.; Guest, C. A. Role of soil manganese in the oxidation of aromatic amines. Environ. Sci. Technol. 2003, 37, 2686–2693. (38) Murray, J. W. Surface chemistry of hydrous manganese dioxide. J. Colloid Interface Sci. 1974, 46, 357–371. (39) Stumm, W.; Morgan, J. J. Aquatic Chemistry, 3rd ed.; Wiley: New York, 1996. (40) Weber, W. J., Jr. Environmental Systems and Processes: Principles, Modeling, and Design; Wiley-Interscience: New York, 2001. (41) Huang, Q. G.; Weber, W. J. Transformation and removal of bisphenol A from aqueous phase via peroxidase-mediated oxidative coupling reactions: efficacy, products, and pathways. Environ. Sci. Technol. 2005, 39, 6029–6036. (42) Sigg, L.; Sturm, M.; Kisteler, D. Vertical transport of heavy-metals by setting particles in Lake Zurich. Limnol. Oceanogr. 1987, 32, 112–130.

ES803622T