Efficient Oxidative Debromination of Decabromodiphenyl Ether by

Dec 3, 2012 - Key Laboratory of Catalysis and Materials Science of the State Ethnic Affairs Commission and Ministry of Education, College of Chemistry...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/est

Efficient Oxidative Debromination of Decabromodiphenyl Ether by TiO2‑Mediated Photocatalysis in Aqueous Environment Aizhen Huang,†,‡ Nan Wang,*,†,‡ Ming Lei,† Lihua Zhu,† Yingying Zhang,‡ Zhifen Lin,§ Daqiang Yin,§ and Heqing Tang*,‡ †

College of Chemistry and Chemical Engineering, Huazhong University of Science and Technology, Wuhan 430074, P. R. China Key Laboratory of Catalysis and Materials Science of the State Ethnic Affairs Commission and Ministry of Education, College of Chemistry and Materials Science, South-Central University for Nationalities, Wuhan 430074, P. R. China § State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, P. R. China ‡

S Supporting Information *

ABSTRACT: Direct evidence was first demonstrated for the oxidative degradation of decabromodiphenyl ether (BDE209) in aqueous TiO2 dispersions under UV irradiation (λ > 340 nm). BDE209 was hardly debrominated over TiO2 in UVirradiated acetonitrile dispersions, but the addition of water into the dispersions greatly enhanced its photocatalytic oxidative debromination. The debromination efficiency of BDE209 as high as 95.6% was achieved in aqueous TiO2 dispersions after 12 h of UV irradiation. The photocatalytic oxidation of BDE209 resulted in generation of aromatic ringopening intermediates such as brominated dienoic acids, which were further degraded by prolonging UV irradiation time. The photocatalytic oxidative debromination of BDE209 was further confirmed by the observation that the BDE209 degradation in water−acetonitrile mixtures with different water contents was positively correlated with the formation of •OH radicals, but not photogenerated electrons. The use of water not only avoided the scavenging of reactive radicals by organic solvent but also enhanced the adsorption of BDE209 on the surface of TiO2, both of which favor the contact of BDE209 with photogenerated holes and •OH species. The confirmation of efficient oxidative degradation and debromination of BDE209 is very important for finding new ways to remove polybrominated diphenyl ethers from the environment.



INTRODUCTION Polybrominated diphenyl ethers (PBDEs) are widely used as flame retardants in manufactured materials and electronic circuit boards. Among the typical commercial PBDEs products (such as penta-, octa-, and deca-bromodiphenyl ether), decabromodiphenyl ether (BDE209) is in the greatest demand. As additives, PBDEs can migrate from products and spread into the environment;1 as a result, their levels in environmental samples are overtaking those of polychlorinated biphenyls (PCBs) and becoming a novel class of global contaminants.2 Although the toxicity of PBDEs is not as well understood as that of PCBs, PBDEs are believed to cause liver tumors and neurodevelopmental and thyroid dysfunctions.3 Thus, the elimination of PBDEs attracts increased attention and concern. A possible environmental fate process for PBDEs is photolytic transformation via reductive debromination. The photolysis halflives (t1/2) range from less than 1 h to more than one year, depending on the congener,4 solvent,4,5 substrates,6,7 and quenching characteristics of cosolutes.8 Because of their high lipophilicity, PBDEs are easily adsorbed on solid matrices, which dramatically hindered the photolytic debromination of PBDEs, © 2012 American Chemical Society

due to energy transfer reactions, excited-state quenching, and/or radiation shielding.6 Considering that the organo-halogen pollutants can undergo reductive dehalogenation, some reductive remediation methods have been applied to degrade PBDEs by using sodium borohydride9 and zerovalent iron (ZVI).10,11 The activity of ZVI on dehalogenation was improved by various approaches such as decreasing the particles size to nanoscale, combining a second catalytic metal, and immobilizing on supports.11−13 Recently, Sun et al. reported that the photogenerated electrons of TiO2 could effectively reduce BDE209, leading to a 90% removal of BDE209 (20 μmol L−1) within 8 min in anoxic TiO2/CH3CN dispersions containing isopropyl alcohol as hole scavenger under UV irradiation.14 Sun et al. also found that BDE209 being preloaded on the TiO2 surface underwent efficient reductive debromination in methanol−water (80:20) mixture, but the degradation rate was greatly Received: Revised: Accepted: Published: 518

July 19, 2012 November 27, 2012 December 2, 2012 December 3, 2012 dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525

Environmental Science & Technology

Article

decreased to zero with increasing water content to 100%.15 It was noted that these reduction technologies were effective for the decomposition of parent compound BDE209, but they are difficult to further debrominate the degradation products of BDE209.10,14,16 For example, about 50% and 45% of BDE209 was respectively transformed into hexa- and penta-BDEs after 10 days in the system of resin-bound ZVI nanoparticles,11 and the accumulation of hepta- and hexa-BDEs respectively corresponded to 25% and 65% of BDE209 in the TiO2/CH3CN system even after 24 h of UV irradiation.14 Relative to the fast disappearance of generated nona- and octa-BDEs, the considerably higher accumulation of PBDEs with 5−7 atoms indicates that the debrominated intermediates of BDE209 with less Br atoms are much more resistant to further reductive debromination than that with more Br atoms. A more serious problem was noted in the reductive debromination of PBDEs: the accumulation of less brominated products still possesses a great threat to the environment, because some less brominated congeners are believed to be more toxic and persistent in the atmosphere.4,14,17 Luo et al. reported a two-stage reduction/ subsequent oxidation treatment of BDE47, which consisted of Fe−Ag reduction and Fenton-like oxidation.18 The TiO2 immobilized montmorillonite was applied in the preadsorption and the sequential photocatalytic destruction of BDE209 in 5% tetrahydrofuran (THF) aqueous dispersions, where the BDE209 degradation involved the photochemical reduction of highly brominated BDEs and the •OH attacking of less brominated products with 1−3 Br atoms.16 These two reports confirmed that • OH radicals are capable of reacting with low brominated BDEs, leading to increased debromination of PBDEs.16,18 However, few feasible oxidative systems have been reported for the degradation of BDE209, to our best knowledge. We noted that although perfluorocarboxylates (PFCs) exhibits extremely chemical stability because of the great strength of the carbon−fluorine bond (CH3−F, 116 kcal/mol), a number of photochemical oxidative methods such as UV−Fenton,19 persulfate photolysis,20,21 and TiO2 photocatalysis22,23 have been demonstrated to be efficient for the removal of PFCs. The addition of the •OH radical at the carbon−fluorine position has been observed in the defluorination of fluoroquinolone pharmaceuticals24 and hexafluobenzene.25 It is also worthy to note that the oxidative degradation of PFCs was carried out in aqueous media, whereas almost all the studies on the degradation of BDE209 was carried out in organic solvents or mixtures containing a high content of organic solvent because of the high hydrophobicity of BDE209. In organic solvents, the reactive species such as •OH radicals will be more likely scavenged by the organic solvent that is more abundant and reactive than the target pollutant. For example, even addition of a small amount of alcohol greatly inhibited the oxidative degradation of organic pollutants in the system of Fenton catalysis or TiO2 photocatalysis.26−28 Moreover, •OH is considerably less reactive in dipolar, aprotic solvents such as acetonitrile than in water.29,30 Poole et al. reported that the reactions of •OH with benzene were faster in water than in acetonitrile by a factor of 65.29 Thus, it may underestimate the power of photogenerated holes and reactive oxygen species to destruct PBDEs in organic solvents. Our opinion is that PBDEs should be oxidatively degradable in aqueous environments. To confirm this, the present work investigated the photocatalytic oxidative degradation of BDE209 over TiO2 in water and/or acetonitrile matrix. Here, acetonitrile was chosen as the organic solvent because it is relatively stable and gives results usable for comparison with that reported in

literature. It was very interesting that a fast and complete debromination of BDE209 was observed over TiO2 in aqueous solution but not in acetonitrile under UV irradiation. The origin of the beneficial effect of water on the photocatalytic debromination of BDE209 over TiO2 and the debromination mechanism was explored.



MATERIALS AND METHODS

Materials. BDE209 (purity >98%) was purchased from Aldrich Chemical Co. (Milwaukee, WI), the composition of which was identified as 98.1% BDE209, 1.91% nona-BDE, and trace octa-BDE ( 340 nm) for 2 h, more than 90% of the added BDE209 was able to be recovered in the TiO2/CH3CN system, but only 53.6% was recovered in the TiO2/H2O system. Since BDE209 could be degraded to generate potentially more persistent and/or less toxic brominated congeners,4,14,17 the favorable elimination methods are expected to achieve a full debromination. We evaluated the degradation of BDE209 by monitoring the formation of Br− ions. After 12 h of UV irradiation, the weak direct photolysis of BDE209 in either CH3CN or water without TiO2 released 1.82 and 1.77 mg L−1 of Br−, corresponding to a debromination efficiency (YBr−) of 21.8% and 21.2%, respectively (curves 1 and 2 in Figure 1b). The UV irradiation of BDE209 over TiO2 in CH3CN released 1.28 mg L−1 of Br−, corresponding to a YBr− of 15.4% (curve 3 in Figure 1b), implying that the addition of TiO2 did not stimulate the degradation of BDE209 in CH3CN under UV irradiation. This is well in agreement with the report of Sun et al. that the degradation of BDE209 was barely observed in the air-saturated TiO2 suspensions with acetonitrile as solvent under UV irradiation.14 However, if the photocatalytic degradation of BDE209 over TiO2 was carried out in the aqueous solution, the concentration of Br− was gradually increased with increasing UV irradiation time, and it reached 7.96 mg L−1 at 12 h of UV irradiation, corresponding to a YBr− of 95.6% (curve 4 in Figure 1b). Dark adsorption experiments in both water and acetonitrile dispersions showed no appearance of Br− ions. These comparisons imply that the simultaneous inclusion of TiO2 as photocatalysts, UV light as excitation source, and water as solvent is critical in the debromination of BDE209 under air-saturated conditions. Effect of TiO2 Load. The debromination of BDE209 is significantly dependent on TiO2 load (Figure 2). With increasing TiO2 load from 0 to 0.1 g L−1, the debromination efficiency of BDE209 within 12 h is greatly increased from 21.2% to 95.6%. The debromination efficiency is decreased to 41.1% with further increasing the TiO2 amount to 1.0 g L−1. This is due to the dual roles of TiO2. On one hand, increasing the TiO2 load increases the concentration of active species. On the other hand, higher TiO2 load will increase the turbidity of dispersion and decrease the penetration depth of UV irradiation, which reduces light utilization efficiency and limits the photo excitation of TiO2. According to our measurement with a UV-A irradiation meter, as TiO2 load increased from 0.025 to 0.1 and then 1.0 g L−1, the transmitted light intensity after passing through TiO2 dispersions decreased from 11.0 to 2.25 and then 0.028 mW cm−2. Once the

(1)

where [Br−]t is the concentration of Br− at exposure time t (mg L−1) and [BDE209]0 is the initial concentration of BDE209 (mg L−1). To identify the intermediate products, the degradation experiments in THF-free systems were carried out as described above. The sampled TiO2 dispersions were separated by centrifugation, and the intermediates adsorbed on the surface of TiO2 particles and dissolved in the solution were analyzed independently. The polar degradation products dissolved in reaction supernatant were monitored by HPLC (or LC-MS) and FT-IR, while the other possible intermediates were identified by using GC-MS (see Section S4 for the details). To investigate the formation of •OH radicals or the photoinduced generation of electrons, either coumarin or DNBODIPY as a hydrophobic fluorescent probe instead of BDE209 was added into the TiO2 dispersions under similar air-saturated and anoxic conditions, respectively. In the anoxic reductive experiments, methanol (0.05 mol L−1) as the electron donor was added into TiO2 dispersions, followed by N2 purging for 30 min to remove O2 prior to UV irradiation. The atmosphere was protected with N2 purging during the UV irradiation. On a JASCO FP6200 spectrofluorometer, the fluorescence intensity was monitored at 456 nm with excitation at 346 nm for the coumarin system and at 510 nm with excitation at 470 nm for the DN-BODIPY system, respectively. To avoid any solvent effect on the fluorescence intensity, the 0.5 mL sample was diluted with 0.5 mL of pure solvent or a mixture of water and acetonitrile with different ratios. Thus the final solvent was composed of 0.5 mL of water and 0.5 mL of acetonitrile.



RESULTS AND DISCUSSION Photocatalytic Oxidative Debromination of BDE209. Figure 1a compares the amount of BDE209 dissolved in the liquid and adsorbed on the TiO2 surface in the dark and under UV irradiation. In dark conditions, only about 3.3% of the added BDE209 is adsorbed on the surface of TiO2 with acetonitrile as solvent, while the adsorption of BDE209 over TiO2 takes up about 95% in the water dispersions. This is due to the extreme hydrophobic BDE209 partition from the aqueous phase onto the surface of TiO2. It should be pointed out that no BDE209 was 520

dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525

Environmental Science & Technology

Article

significantly increased from 15.4% to 95.6%. This is in contrast with the photocatatlytic reduction of BDE209 observed by Sun et al., who reported that 1% water in anoxic acetonitrile could stop the BDE209 removal over TiO2 under the UV irradiation in the presence of isopropyl alcohol as hole scavenger,14 but well in agreement with the solvent effect on the photocatalytic oxidation of benzene over TiO2 pillared clay, where the activity was increased by adding water in acetonitrile−water mixtures.32 To understand the solvent effect on the BDE209 degradation, the role of water on the generation of electrons and •OH radicals was investigated by using fluorescent probes. When nonfluorescent DN-BODIPY or coumarin instead of BDE209 was added, DN-BODIPY can trap photogenerated electrons to produce fluorescent 4-hydroxyamino-3-nitrophenyl-BODIPY (HN-BODIPY, eq 4),31,33 while coumarin reacts with •OH to generate fluorescent 7-hydroxy coumarin (7-HC, eq 5).34

Figure 2. Effect of TiO2 load on debromination efficiency (YBr−) of BDE209 in UV-irradiated aqueous dispersions under air-saturated conditions.

latter effect of TiO2 concentration prevails over the former one, the degradation of BDE209 would decrease with increasing TiO2 load. Effect of Solvent. Experiments were carried out in water− acetonitrile mixtures with various amounts of water. As shown in Figure 3a, the debromination of BDE209 over TiO2 within 12 h of UV illumination was negligible in air-saturated acetonitrile after deducting the direct photolysis. However, the addition of water improved its photocatalytic debromination. As the volume content of water was increased from 0% to 100%, the debromination efficiency within 12 h of UV irradiation was

TiO2 → e− + h+

(2)

h+ + H 2O(OH−) → •OH + H+

(3)

DN‐BODIPY (nonfluorescent) + e− → HN‐BODIPY (fluorescent)

(4)

coumarin (nonfluorescent) + •OH(h+, H 2O) → 7‐HC (fluorescent)

(5)

Figure 3. Effects of water content in water−CH3CN mixtures on (a) debromination efficiency (YBr−) of BDE209 within 12 h, and the fluorescent intensity of (b) DN-BODIPY and (c) coumarin in the TiO2-mediated photocatalysis. (d) Relationship between the YBr− within 12 h and the initial apparent formation rate constant of 7-HC (circle, i.e., k•OH) within 60 min or of HN-BODIPY within 20 min (triangle, i.e., ke‑). Star symbols (★) in panel a represented YBr− of BDE209 in UV-irradiated TiO2-free solutions. 521

dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525

Environmental Science & Technology

Article

Figure 4. (a) FTIR spectra and (b) HPLC chromatograms of reaction supernatants with different UV irradiation times. (c, d) GC-MS TIC chromatograms of the reaction extracts from (c) the separated TiO2 (m/z 200−1000) and (d) supernatant (m/z 50−800) with different UV irradiation times. A TG-5MS and TG-35MS capillary column was used in c and d, respectively, and the analysis details are illustrated in Section S4. All the photocatalytic degradations of BDE209 were carried out in THF-free systems. The spectrum of BDE209 was given as a reference in panel a, and the fact that no signals of BDE209 were observed at 0 h is consistent with the complete adsorption of BDE209 on the TiO2 particle surface.

CH3OH + •OH/h+ → → CO2 + H 2O

(6)

O2 + e− → O2•−

(7)

scavenger (eq 6) to the UV/TiO2/H2O system greatly inhibited the debromination of BDE209 (curve 5 in Figure 1b). The use of water as solvent not only enhances the formation of •OH radicals but also improves the reactivity of •OH radicals. Therefore, the photocatalytic debromination of BDE209 over TiO2 is increased with increasing water content in the water−acetonitrile mixture. Degradation Intermediates. The variations in the chemical compositions of degradation systems were checked with FT-IR, HPLC, and LC-MS analysis. Compared with pure BDE209 and the control sample at dark, the reaction supernatants after UV irradiation show IR absorption bands at 1720, 1580, and 1405 cm−1, which are typical features of the stretching vibration of C O, asymmetric vibration of OC−O−, and symmetric vibration of OC−O− groups, respectively (Figure 4a). This signals the formation of some carboxylic derivatives. In the HPLC measurements, a broad peak with shoulders appears at 1.2 min when detecting at 210 nm during the BDE209 degradation. As a control, no such changes were observed in similar photocatalytic systems without BDE209 (Figure S3b). By using LC-MS with ESI negative mode detection, the HPLC peak observed in the 0.5 h UV-irradiated sample possesses mass fragment ions at m/z 374/376/378 ([M − 2H]•−), 328/330/332 ([M − CHO − OH]−), 295/297([M − Br − 2H]−), 79/81([Br]−), and 45 ([COO]−), which may be assigned to 2,3,5-tribromo-4-hydroxy6-oxahexa-2,4-dienoic acid or 2,4,5-tribromo-3-hydroxy-6-oxahexa-2,4-dienoic acid or their isomers (C6H3Br3O4, Mr. 378.8, Figure S5). As the UV irradiation time was increased, the

As shown in Figure 3b, the fluorescence intensity of generated HN-BODIPY in acetonitrile dispersions is greatly depressed by adding 0.1% water, and completely inhibited by adding 1% water. This is well consistent with Sun’s observation that 1% water in CH3CN stopped the photocatalytic reduction of BDE209 over TiO2 within isopropyl alcohol as hole scavenger.14 This also suggests that the amount of photogenerated electrons in aqueous TiO2 dispersions is very limited and hence the photocatalytic reduction of BDE209 is negligible. In contrast, the fluorescent intensity of formed 7-HC in TiO2 dispersions was greatly increased with increasing water content, suggesting that water has a drastic positive impact on the reaction of coumarin with • OH. This can be explained by the following two reasons: (1) the presence of CH3CN decreases the adsorption of water molecules on the TiO2 surface, and then minimizes the conversion of holes to •OH (eq 2);35 (2) the reaction of •OH with aromatic hydrocarbons is much slower in CH3CN than in water.29,30 Moreover, the debromination efficiency of BDE209 has a roughly positive relationship with the apparent formation rate constant of 7-HC (R2 = 0.987), but it is not related to that of HNBODIPY (Figure 3d). These comparisons indicate that the debromination of BDE209 here is achieved not by a reductive process but mainly through an oxidative process, which is further evidenced by the fact that the addition of methanol as •OH 522

dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525

Environmental Science & Technology

Article

Figure 5. Possible pathways for the photocatalytic oxidative degradation of BDE209 over TiO2 in aqueous dispersions. The brominated dienoic acids being represented in bold characters were detected in LC-MS measurements. More details can be found in Figure S9.

and 29 (Figure S7), further indicating that the products are small carboxylic derivatives or hydroxylated compounds. Since it is impossible to crack the aromatic cores in the reductive reactions, the formation of brominated dienoic acids implies that a distinct oxidative degradation process must occur in the present work. Mechanism for Photocatalytic Oxidation of BDE209. The above results confirmed that BDE209 was degraded in aqueous TiO2 dispersions under UV irradiation. The oxidative degradation of aromatic pollutants usually proceeds through three stages as parent pollutant → aromatic intermediates → aliphatic intermediates → CO2 + H2O. However, we did not find any aromatic intermediates during the photocatlaytic oxidation of BDE209. In a consecutive reaction, the concentration of intermediates is dependent on the rate of their generation from the parent and their transformation to daughter compounds. If the degradation of aromatic intermediates is much faster than

retention time of the above-mentioned HPLC peak was slightly shortened and its mass signal intensity was decreased, implying that the brominated dienoic acids are broken to shorter chain carboxyl derivatives. To identify other possible intermediates, the components extracted from the recovered TiO2 and the reaction supernatants were further analyzed by GC-MS. As shown in Figure 4c, except for the peaks for parent BDE209 at 9.5 min and trace impurities at 7.75−8.81 min and 7.0−7.1 min (Section S1), there were no other PBDEs or brominated compounds detected in the GS-MS spectra of extracted solutions from the recovered TiO2. However, several new peaks appeared in the GC-MS spectra of the extracted solutions from the degradation supernatants (Figure 4d and Figure S6). These mass fragment ions are mostly present in the region of m/z lower than 200, and the prominent feature is a loss of specific number of m/z units like 16, 17, 28, 523

dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525

Environmental Science & Technology

Article

montmorillonite, and thus the reductive sites could facilitate the electron transfer from TiO2 to BDE209.16 This made BDE209 and the generated hydrophobic products be far away from oxidative sites (TiO2 surface), being unfavorable to the h+/•OH-induced reactions in comparison with the substrates localized on the TiO2 surface. In the present system, polar water as solvent repels BDE209 and the hydrophobic intermediates, which makes these organic pollutants adhere strongly on the TiO2 surface, leading to a favorable reaction environment for the decomposition and debromination of BDE209. This study confirmed efficient oxidative debromination of BDE209 by TiO2-mediated photocatalysis in an aqueous environment and provided insights to the mechanism of BDE209 oxidation. After 12 h of UV irradiation, the TiO2/ H2O system caused a BDE209 removal of 96.8% with debromination efficiency of 95.6%. As we know, proper reduction methods may rapidly remove BDE209 (a degradation removal of about 100%) within 8 min but provide a debromination efficiency of only 43% within 24 h (Table S1). From the viewpoint of full debromination and mineralization of BDE209, the photocatalytic oxidation method is a promising method for degrading BDE209. Moreover, highly toxic PBDEs congeners and aromatic intermediates were hardly observed in the UV/TiO2/H2O system, which also suggests that the findings in the present work will provide an alternative approach to developing green and effective methods for removing BDE209 and its family members.

their generation, there will not be any accumulation of aromatic products. To validate this presumption, we compared the photocatalytic degradation of BDE209 and pentabromophenol (as the most probable aromatic product, see Section S5) over TiO2 in their mixture solutions (Figure S8a). It was found that about 30% of the added BDE209 (10 mg L−1, 10.4 μmol L−1) was removed within 4 h, while all the added pentabromophenol (10.4 μmol L−1) rapidly disappeared within 0.5 h during the degradation. This strongly suggests that pentabromophenol is much more susceptible to oxidation than BDE209. It was worthy of noting that the production rate of Br− ions is well matched the degradation rate of BDE209 in the UV/TiO2/H2O system (Figure S8b). By summing organic bromine in residual BDE209 and the generated inorganic Br− ions, the obtained total bromine elements were slightly decreased at the initial reaction stage, and then increased almost up to the added initial amount after 3 h UV irradiation, suggesting that a few bromine-containing intermediates were formed and then rapidly disappeared. On the basis of the above results, Figure 5 tentatively proposes pathways for the photocatalytic oxidative degradation of BDE209. Since BDE209 tends to locate on the TiO2 surface in aqueous solution, both photogenerated holes and •OH radicals participate in the BDE209 degradation. The addition of •OH to BDE209 leads to an aromatic-OH adduct radical. The hole can directly oxidize adsorbed BDE209, resulting in the formation of a radical cation, which will react quickly with water, leading to the same OH-adduct radical after a subsequent deprotonation (Figure S9a). Because the carbons bound to the ether oxygen (namely, ipso carbons) have the highest electron density due to the electron-donating of oxygen and the para carbons exhibit the least steric hindrance, the initial attack of h+/•OH appears to be either at the ipso position, followed by the cleavage of ether linkage to form pentabromophenol and pentabromophenoxyl radical, or at the para carbon to form bromohydrin intermediate, and then to BDE-phenoxy radical by rapidly losing HBr. After undergoing several steps of h+/•OH additions and the subsequent HBr/Br eliminations, BDE-phenoxy radical is oxidized to brominated benzoquinone and/or benzenediol derivatives or pentabromophenol (Section S5). A similar OH addition and HBr/Br elimination pathway has been observed for the reaction of •OH with fluoroquinolone pharmaceuticals24 and hexafluorobenzene.25 The reactivity of more brominated PBDEs toward •OH radical is lower than that of less brominated congeners and the corresponding hydroxylated molecules.36,37 According to the above-mentioned results for the photocatalytic degradation of BDE209 and pentabromophenol and that for the Br balance (Figure S8), it is rational to propose that the initial attack of h+/•OH on BDE209 is much slower than the following oxidation of the generated pentabromophenol and other possible less-brominated hydroxylated PBDEs, so that almost all aromatic intermediates are rapidly consumed in situ as soon as they are formed on the TiO2 surface. However, after the aromatic ring is opened, the generated aliphatic carboxylic acids exhibit strong hydrophilicity and will diffuse away from the TiO2 surface into the bulk solution, being unfavorable to their decomposition by the attack of h+/• OH species. Thus, we observed the accumulations of aliphatic carboxylic derivatives dissolved in solution. It is worthy to note that the main degradation intermediates are very different from the previously reported debrominated and hydroxylation products in the system of TiO2 immobilized montmorillonite,16 possibly because the different localization of substrate influences the reaction pathways. In that system, BDE209 was mainly adsorbed on hydrophobic



ASSOCIATED CONTENT

S Supporting Information *

Supporting sections (Section S1−S5), figures (Figure S1−S9), and table (Table S1), as mentioned in the text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Phone/fax: +86-27-87543632 (N.T.); +86-27- 67843323 (H.T.). E-mail: [email protected] (N.W.); hqtang62@ yahoo.com.cn (H.T.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors acknowledge financial support from the National Natural Science Foundation (Grants Nos. 21077037, 21107027, and 21177044), the National High Technology Research and Development Program of China (863 Program) (Grant No. 2012AA06A304), China Postdoctoral Science Foundation (Nos. 2011M501207 and 2012T50649), and the Foundation of State Key Laboratory of Pollution Control and Resource Reuse (Tongji University) of China (No. PCRRV10001). The authors also thank Dr. Jianxia Lv (Thermo Fisher Scientific Inc., Beijing) for her kind help in the GC-MS measurements.



REFERENCES

(1) La Guardia, M. J.; Hale, R. C.; Harvey, E. Detailed polybrominated diphenyl ether (PBDE) congener composition of the widely used penta-, octa-, and deca-PBDE technical flame-retardant mixtures. Environ. Sci. Technol. 2006, 40 (20), 6247−6254. (2) Rahman, F.; Langford, K. H.; Scrimshaw, M. D.; Lester, J. N. Polybrominated diphenyl ether (PBDE) flame retardants. Sci. Total Environ. 2001, 275 (1−3), 1−17.

524

dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525

Environmental Science & Technology

Article

(3) Siddiqi, M. A.; Laessig, R. H.; Reed, K. D. Polybrominated diphenyl ethers (PBDEs): new pollutants-old diseases. Clin. Med. Res. 2003, 1 (4), 281−90. (4) Eriksson, J.; Green, N.; Marsh, G.; Bergman, A. Photochemical decomposition of 15 polybrominated diphenyl ether congeners in methanol/water. Environ. Sci. Technol. 2004, 38 (11), 3119−3125. (5) Davis, E. F.; Stapleton, H. M. Photodegradation pathways of nonabrominated diphenyl ethers, 2-ethylhexyltetrabromopenzoate and di(2-ethylhexyl)tetrabromophthalate: identifying potential markers of photodegradation. Environ. Sci. Technol. 2009, 43 (15), 5739−5746. (6) Söderström, G.; Sellström, U.; De Wit, C. A.; Tysklind, M. Photolytic debromination of decabromodiphenyl ether (BDE 209). Environ. Sci. Technol. 2004, 38 (1), 127−132. (7) Ahn, M. Y.; Filley, T. R.; Jafvert, C. T.; Nies, L.; Hua, I.; BezaresCruz, J. Photodegradation of decabromodiphenyl ether adsorbed onto clay minerals, metal oxides, and sediment. Environ. Sci. Technol. 2006, 40 (1), 215−220. (8) Hua, I.; Kang, N.; Jafvert, C. T.; Fabrega-Duque, J. R. Heterogeneous photochemical reactions of decabromodiphenyl ether. Environ. Toxicol. Chem. 2003, 22 (4), 798−804. (9) Granelli, L.; Eriksson, J.; Athanasiadou, M.; Bergman, A. Reductive debromination of nonabrominated diphenyl ethers by sodium borohydride and identification of octabrominated diphenyl ether products. Chemosphere 2011, 82 (6), 839−846. (10) Keum, Y. S.; Li, Q. X. Reductive debromination of polybrominated diphenyl ethers by zerovalent iron. Environ. Sci. Technol. 2005, 39 (7), 2280−2286. (11) Li, A.; Tai, C.; Zhao, Z. S.; Wang, Y. W.; Zhang, Q. H.; Jiang, G. B.; Hu, J. T. Debromination of decabrominated diphenyl ether by resinbound iron nanoparticles. Environ. Sci. Technol. 2005, 41 (19), 6841− 6846. (12) Zhuang, Y.; Ahn, S.; Seyfferth, A. L.; Masue-Slowey, Y.; Fendorf, S.; Luthy, R. G. Dehalogenation of polybrominated diphenyl ethers and polychlorinated biphenyl by bimetallic, impregnated, and nanoscale zerovalent iron. Environ. Sci. Technol. 2011, 45 (11), 4896−4903. (13) Zhu, B. W.; Lim, T. T.; Feng, J. Reductive dechlorination of 1,2,4trichlorobenzene with palladized nanoscale Fe0 particles supported on chitosan and silica. Chemosphere 2006, 65 (7), 1137−1145. (14) Sun, C. Y.; Zhao, D.; Chen, C. C.; Ma, W. H.; Zhao, J. C. TiO2mediated photocatalytic debromination of decabromodiphenyl ether: kinetics and intermediates. Environ. Sci. Technol. 2009, 43 (1), 157−162. (15) Sun, C. Y.; Zhao, J. C.; Ji, H. W.; Ma, W. H.; Chen, C. C. Photocatalytic debromination of preloaded decabromodiphenyl ether on the TiO2 surface in aqueous system. Chemosphere 2012, 89 (4), 420− 425. (16) An, T. C.; Chen, J. X.; Li, G. Y.; Ding, X. J.; Sheng, G. Y.; Fu, J. M.; Mai, B. X.; O’Shea, K. E. Characterization and the photocatalytic activity of TiO2 immobilized hydrophobic montmorillonite photocatalysts degradation of decabromodiphenyl ether (BDE 209). Catal. Today 2008, 139 (1−2), 69−76. (17) He, J. Z.; Robrock, K. R.; Alvarez-Cohen, L. Microbial reductive debromination of polybrominated diphenyl ethers (PBDEs). Environ. Sci. Technol. 2006, 40 (14), 4429−4434. (18) Luo, S.; Yang, S. G.; Xue, Y. G.; Liang, F.; Sun, C. Two-stage reduction/subsequent oxidation treatment of 2,2′,4,4′-tetrabromodiphenyl ether in aqueous solutions: kinetic, pathway and toxicity. J. Hazard. Mater. 2011, 192 (3), 1795−1803. (19) Tang, H. Q.; Xiang, Q. Q.; Lei, M.; Yan, J. C.; Zhu, L. H.; Zou, J. Efficient degradation of perfluorooctanoic acid by UV-Fenton process. Chem. Eng. J. 2012, 184, 156−162. (20) Hori, H.; Yamamoto, A.; Hayakawa, E.; Taniyasu, S.; Yamashita, N.; Kutsuna, S. Efficient decomposition of environmentally persistent perfluorocarboxylic acids by use of persulfate as a photochemical oxidant. Environ. Sci. Technol. 2005, 39 (7), 2383−2388. (21) Rayne, S.; Forest, K. Perfluoroalkyl sulfonic and carboxylic acids: A critical review of physicochemical properties, levels and patterns in waters and wastewaters, and treatment methods. J. Environ. Sci. Heal. A. 2009, 44 (12), 1145−1199.

(22) Dillert, R.; Bahnemann, D.; Hidaka, H. Light-induced degradation of perfluorocarboxylic acids in the presence of titanium dioxide. Chemosphere 2007, 67 (4), 785−792. (23) Estrellan, C. R.; Salim, C.; Hinode, H. Photocatalytic decomposition of perfluorooctanoic acid by iron and niobium codoped titanium dioxide. J. Hazard. Mater. 2010, 179 (1−3), 79−83. (24) Santoke, H.; Song, W. H.; Cooper, W. J.; Greaves, J.; Miller, G. E. Free-radical-induced oxidative and reductive degradation of fluoroquinolone pharmaceuticals: kinetic studies and degradation mechanism. J. Phys. Chem. A. 2009, 113 (27), 7846−7851. (25) Kobrina, L. S. Radical reactions of polyfluoroarenes. Fluorine Notes 2012, 2 (81), http://notes.fluorine1.ru/public/2012/2_2012/ retro/index.html. (26) Wang, M. Q.; Wang, N.; Tang, H. Q.; Cao, M. J.; She, Y. B.; Zhu, L. H. Surface modification of nano-Fe3O4 with EDTA and its use in H2O2 activation for removing organic pollutants. Catal. Sci. Technol. 2012, 2 (1), 187−194. (27) Wang, N.; Zhu, L. H.; Wang, D. L.; Wang, M. Q.; Lin, Z. F.; Tang, H. Q. Sono-assisted preparation of highly-efficient peroxidase-like Fe3O4 magnetic nanoparticles for catalytic removal of organic pollutants with H2O2. Ultrason. Sonochem. 2010, 17 (3), 526−533. (28) Kim, S.; Choi, W. Kinetics and mechanisms of photocatalytic degradation of (CH3)nNH4‑n+ (0 ≤ n ≤ 4) in TiO2 suspension: The role of OH radicals. Environ. Sci. Technol. 2002, 36 (9), 2019−2025. (29) DeMatteo, M. P.; Poole, J. S.; Shi, X. F.; Sachdeva, R.; Hatcher, P. G.; Hadad, C. M.; Platz, M. S. On the electrophilicity of hydroxyl radical: A laser flash photolysis and computational study. J. Am. Chem. Soc. 2005, 127 (19), 7094−7109. (30) Poole, J. S.; Shi, X. F.; Hadad, C. M.; Platz, M. S. Reaction of hydroxyl radical with aromatic hydrocarbons in nonaqueous solutions: A laser flash photolysis study in acetonitrile. J. Phys. Chem. A. 2005, 109 (11), 2547−2551. (31) Tachikawa, T.; Wang, N.; Yamashita, S.; Cui, S. C.; Majima, T. Design of a highly sensitive fluorescent probe for interfacial electron transfer on a TiO2 surface. Angew. Chem., Int. Ed. 2010, 49 (46), 8593− 8597. (32) Shimizu, K. I.; Kaneko, T.; Fujishima, T.; Kodama, T.; Yoshida, H.; Kitayama, Y. Selective oxidation of liquid hydrocarbons over photoirradiated TiO2 pillared clays. Appl. Catal., A 2002, 225 (1−2), 185−191. (33) Wang, N.; Tachikawa, T.; Majima, T. Single-molecule, singleparticle observation of size-dependent photocatalytic activity in Au/ TiO2 nanocomposites. Chem. Sci. 2011, 2 (5), 891−900. (34) Wang, N.; Chen, Z. F.; Zhu, L. H.; Jiang, X.; Lv, B.; Tang, H. Q. Synergistic effects of cupric and fluoride ions on photocatalytic degradation of phenol. J. Photochem. Photobiol., A 2007, 191 (2−3), 193−200. (35) El-Morsi, T. M.; Budakowski, W. R.; Abd-El-Aziz, A. S.; Friesen, K. J. Photocatalytic degradation of 1,10-dichlorodecane in aqueous suspensions of TiO2: A reaction of adsorbed chlorinated alkane with surface hydroxyl radicals. Environ. Sci. Technol. 2000, 34 (6), 1018− 1022. (36) Raff, J. D.; Hites, R. A. Deposition versus photochemical removal of PBDEs from lake superior air. Environ. Sci. Technol. 2007, 41 (19), 6725−6731. (37) Zhou, J.; Chen, J. W.; Liang, C. H.; Xie, Q.; Wang, Y. N.; Zhang, S. Y.; Qiao, X. L.; Li, X. H. Quantum chemical investigation on the mechanism and kinetics of PBDE photooxidation by •OH: A case study for BDE-15. Environ. Sci. Technol. 2011, 45 (11), 4839−4845.

525

dx.doi.org/10.1021/es302935e | Environ. Sci. Technol. 2013, 47, 518−525