Formation of Chlorinated Intermediate from ... - ACS Publications

Sep 11, 2009 - Surface Saline Water under. Simulated Solar Light Irradiation. HUI LIU, HUIMIN ZHAO, XIE QUAN,*. YAOBIN ZHANG, AND SHUO CHEN...
1 downloads 0 Views 328KB Size
Environ. Sci. Technol. 2009, 43, 7712–7717

Formation of Chlorinated Intermediate from Bisphenol A in Surface Saline Water under Simulated Solar Light Irradiation HUI LIU, HUIMIN ZHAO, XIE QUAN,* YAOBIN ZHANG, AND SHUO CHEN Key Laboratory of Industrial Ecology and Environmental Engineering (Ministry of Education, China), School of Environmental and Biological Science and Technology, Dalian University of Technology, Dalian 116024, China

Received March 17, 2009. Revised manuscript received August 22, 2009. Accepted August 27, 2009.

Chlorinated organic compounds are generally of great concern, but many uncertainties exist regarding how they are generated. To illustrate the possibility of photochemical formation of organochlorine compounds in natural water, the phototransformation of bisphenol A (BPA) in aqueous saline solution containing Fe(III) and fulvic acid (FA), and in coastal seawater under simulated solar light irradiation was investigated. 2-(3-Chloro-4-hydroxyphenyl)-2-(4-hydroxyphenyl) propane (3ClBPA) and 2,2-bis(3-chloro-4-hydroxyphenyl) propane (3,3diClBPA) were the main chlorinated derivatives during the processes. Laser flash photolysis (LFP) and electron spin resonance (ESR) results indicated that the chlorination of BPA was most likely due to the formation of Cl2•- radical as a consequence of Fe(III) irradiation, yielding Cl• and OH• radical species and finally forming Cl2•- radical upon further reaction with chloride. The formation of Fe(III)-FA complex, which is a normal coexistence configuration of Fe(III) and FA in natural water, promoted the BPA chlorination through producing more Cl2•- radical. Moreover, FA had two opposite effects: forming Fe(III)-FA complex to enhance Cl2•- formation and competing radicals with BPA, which resulted in different overall effects at different concentrations: BPA chlorination was enhanced with the increasing of FA concentration ([FA]) when [FA] < 3.2 mg L-1; when the concentration of FA was as high as 10 mg L-1, it slowed down obviously. The described BPA photochlorination process took place from pH 6.3 to 8.5 and increased with the increasing of chloride concentration, indicating it could occur universally in natural saline surface water. These results propose a natural photochemical source for organochlorine compounds.

Introduction Among the many organic chemicals encountered in the environment, chlorinated organic compounds occupy a special position since they have complex diversities of biologic effects, for example, immunotoxicity, endocrine toxicity, reproductive or developmental toxicity, carcinogenesis. Moreover, the twelve persistent organic pollutants (POPs) listed in the Stockholm Convention are all chlorinated organic * Corresponding author phone: +86-411-84706140; fax: +86-41184706263; e-mail: [email protected]. 7712

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 20, 2009

compounds, such as polychlorinated dibenzo-p-dioxin (PCDD), polychlorinated dibenzofuran (PCDF), and polychlorinated biphenyls (PCBs), which are highly toxic to humans and the ecosystem (1). Their sources, which were once considered to be predominantly anthropogenic because of widespread use as pesticides and industrial chemicals, improper handling, and accidental spills, are now known to include natural sources (2-6). The biotic and geochemical processes are the main published and focused natural origins of chlorinated organic chemicals. Researchers have identified several living organisms, such as phytoplankton, wood-rotting fungi, and microorganisms, that the abilities to transfer chloride into organochlorine compounds for use in cell adhesion and defense processes through producing certain enzymes (7, 8). And both the chlorinated aliphatics and aromatics, even PCDD/Fs, are testified to be produced during these processes (5, 6, 9). Despite these efforts, relatively little attention has been focused on abiotic, chemical processes for organochlorine production and relatively few examples are known (4, 10). In 2005, Vione (10) and Calza (11) reported formation of chlorophenols by photochemical process occurring in strongly acid marine aerosol or in the presence of semiconductor catalyst (Fe2O3, CdS), respectively. The described processes indicate that oxidation of the chlorine ion to corresponding radicals is responsible for the chlorination reaction. Furthermore, Moore (4) demonstrated that colored dissolved organic matter (CDOM) as the direct precursor yielded CH3Cl through a photochemical process when investigating the source of the methyl moiety of CH3Cl in saline waters. Therefore, it is plausible that the oxidation of chlorine ion happens in natural neutral surface water. But there is little direct evidence for the formation of other persistent and less volatile organochlorine compounds, and the photochemical processes and mechanisms leading to CH3Cl formation in natural neutral surface water are also poorly understood. This paper presents results of photochlorination studies in aqueous solution, using bisphenol A (BPA) as a model molecule. BPA is widely used as monomer in manufacture of polycarbonate and epoxy resins and is listed as one kind of endocrine disruptor often detected in ground and surface waters (12). BPA chlorination was also investigated in the presence of important environmental factors such as Fe(III) and fulvic acid (FA). The study is expected to assess whether such a process is feasible in natural water and to provide more information for understanding the mechanisms of organochlorination.

Materials and Methods Chemicals. BPA (purity grade > 99%) was purchased from Fluka Chemical Corporation (USA). 2-(3-Chloro-4-hydroxyphenyl)-2-(4-hydroxyphenyl)propane (3-ClBPA), 2-bis(3chloro-4-hydroxyl-phenyl) propane (3,3-diClBPA), and 2-(3,5dichloro-4-hydroxyphenyl)-2-(4-hydroxyphenyl) propane (3,5diClBPA) (Figure S1 in the Supporting Information (SI)) were synthesized and purified as described by Fukazawa et al. (13). Phenanthrene D10 used as an internal standard for GC/ MS analyses was purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany). FA was extracted from humic acid (HA, originated from soil, MP Biomedicals, Inc., Eschwege, Germany) according to the method recommended by International Humic Substances Society (IHSS). The details of other chemicals are presented in S1 of the SI. Photochemical Experiments. The photoreaction experiments were performed in two photochemical devices with light intensity of 155 and 12 mW cm-2, respectively; both of 10.1021/es900811c CCC: $40.75

 2009 American Chemical Society

Published on Web 09/11/2009

them were equipped with Xenon lamps and 290 nm cutoff filters to simulate sunlight. Experimental details for photochemical tests are described in S2 of SI. Laser Flash Photolysis Experiments. Laser Flash Photolysis (LFP) experiments were performed using a laser flash photolysis spectrometer (LP920, Edinburgh Instruments Ltd.) with a frequency tripled Q-switched Nd:YAG laser, which provided 355 nm pulse with a duration of 8 ns. The average laser power was 10 mJ/pulse. Experiments were carried out in a 1-cm quartz cell. Electron Spin Resonance Measurements. Electron spin resonance (ESR) samples were prepared in dark at room temperature and loaded into a capillary quartz tube immediately after mixing of solutions and DMPO (final concentration 80 mM), which was used as the spin-trapper for trapping radicals. To minimize experimental errors, the same quartz capillary tube was used for all ESR measurements. ESR spectra were recorded on a Bruker Elexsys 500 ESR spectrometer equipped with an in situ irradiation source (a Quanta-Ray Nd:YAG laser system, λ ) 355 nm). Other operation parameters of the ESR spectrometer were microwave frequency 9.78 GHz, microwave power 12.7 mW, and modulation frequency 100 kHz. Characterization of FA Structure. Elemental analysis of FA was performed on an element analyzer (Vario EL III, Elementar, Germany). The structure of FA was characterized by 1H NMR and pyrolysis-GC/MS (pyr-GC/MS). The 1H NMR spectra were measured on a Varian INOVA plus 400 model instrument. For measurement, FA was freeze-dried, and the resulting powder (50 mg) was then dissolved in 500 µL of D2O. Pyr-GC/MS was analyzed using a PY-2020iD pyrolyzer (Frontier Ltd.) connected with a trace GC Ultra/DSQ II (Thermo Fisher Ltd.). The powdered FA (1 mg) was pyrolyzed at 500 °C for 4.0 s with 0.5 mg of a methylation reagent, tetramethylammonium hydroxide (TMAH). The temperature ramp for the GC was as follows: 40 °C for 3 min, 40-280 at 10.0 °C min-1, and a 280 °C hold for 5 min. Characterization of Fe(III)-FA Complex. Synchronous fluorescence (SyF) of Fe(III)-FA complex was measured using a fluorescence spectrophotometer (model F-4500, Hitachi, Japan), and Fourier transform infrared spectroscopy (FTIR) was recorded with an FTIR spectrometer (Prestige-21, Shimadzu, Japan) to analyze functional groups of the Fe(III)-FA complex. A laser particle analyzer (Zetaplus/90p, Brookhaven, USA) was used to analyze the average particle size of Fe(III)-FA complex. Further details are provided in S3 of SI. Analytical Determinations. The concentration of BPA was determined by HPLC (Waters-2695, photodiode array detector (PDA)-2996, Waters, USA), and the chlorinated derivatives were analyzed using GC/MS (6890/5973, Agilent, USA) after they were extracted by solid-phase extraction (SPE) cartridges (Cleanert PEP-SPE, Agela Technologies, China). Further details of the analytical determinations are in S4 of SI and the GC/MS spectrum of the standard mixture is in Figure S2.

Results and Discussion BPA Photochlorination in the Presence of Fe(III). The preliminary experiment, irradiation of BPA solutions containing Fe(III) and chloride ions at pH 3.0, produced 3-ClBPA and 3,3′-diClBPA as chloroderivatives, thereby testifying the photochlorination of BPA in saline water containing iron. Iron was contained in the system because it is a ubiquitous and crucial element in natural water, and it was reported as an inducer of substitution of phenol in the presence of nitrite or chloride in aerosol conditions (24). Figure 1 shows an example of the time evolution of 87.7 µM BPA, 3-ClBPA, and 3,3′-diClBPA in the presence of 6.6 × 10-4 M Fe(III) and 0.2 M NaCl at pH 3.0. It shows that the formation of 3-ClBPA was favored initially, reaching a maximum concentration of about

FIGURE 1. Photodegradation of BPA (87.7 µM) in the presence of FeCl3 (6.6 × 10-4 M) and Cl- (0.2 M) at pH 3.0 (A), and time evolution of 3-ClBPA and 3,3′-diClBPA (B). 252 nM after 15 min, and then decaying to about 50 nM at 180 min, whereas 3,3′-diClBPA increased over time, reaching a concentration of about 20 times lower than that of 3-ClBPA. These trends were related to the concentration of BPA, which decreased very fast under the irradiation (Figure 1A). An overview of the various transformation intermediates of BPA in the system was obtained by GC/MS analysis of the dichloromethane/methanol extracts. The main identified compounds were 3-ClBPA, p-isopropenylphenol (Figure S3), and other trace chlorine substituted compounds including 3,3′-diClBPA and 4-chlorophenol (not marked). Furthermore, no chloroderivative in dark was found (Figure 1B), indicating the chlorination of BPA was a photoinduced process. There have been reports that the binding affinities of 3-ClBPA and 3,3′-diClBPA for the human estrogen receptor are higher than that of BPA (14). Therefore, the formation of the chlorine derivatives of BPA and its mechanism in photochemical reaction in natural waters are of concern for human health and the environment. Many articles have reported that chlorination of phenol in aerosol was caused by reactive chlorine species (10, 24), however, the reactive chlorine species has not been identified. To explain the process of chlorination of BPA in water, LFP, a useful instrument for transient species research, including radicals and transition state, was used to investigate the reactive chlorine species for the photochlorination of BPA. Transient absorption spectra of LFP are given in Figure 2. The transient absorption spectrum of Fe(III) in saline water (curve a) showed a typical absorption at λ ) 340 nm corresponding to Cl2•- radical (15, 16). Generally, the Cl2•radical is formed through the equilibrium reaction between Cl• and Cl-, and becomes the dominant species in aqueous solution containing Cl-, since the forward reaction of equilibrium 1 is very fast. In the present work the adopted chloride concentration of 0.2 M, reaction 1 resulted in [Cl2•-].[Cl•]. VOL. 43, NO. 20, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7713

[k+ ) 2 × 1010 M-1 s-1, k- ) Cl· + Cl- f Cl·2 1.1 × 105 M-1 s-1 (10)]

(1)

It was interesting that the presence of BPA in the solution containing Fe(III) and Cl- decreased the absorption at λ ) 340 nm obviously (curve b), inferring the consumption of Cl2•- by BPA. While the absorption at λ ) 340 nm of the solution containing only BPA and Cl- (curve c) was not obvious, meaning that the amount of Cl2•- was extremely low in the absence of Fe(III), indicating the importance of Fe(III) in the formation of Cl2•-. To explain the important role of Fe(III) on the formation of Cl2•-, as mentioned in the LFP experiment, contribution of Fe(III) species to the generation of Cl2•- under the adopted condition was testified based on the thermodynamic stability constants of iron complexes and ESR DMPO spin-trap technique. The main iron species in the studied system included Fe3+, Fe(OH)2+, FeCl2+, and FeCl2+, and based on equilibrium calculation and the stability constants from literature (10): KFeCl ) a(FeCl2+) a(Fe3+)-1 a(Cl-)-1 ) 30.2, KFeOH ) a(FeOH2+) a(Fe3+)-1 a(OH-)-1 ) 3.47 × 1011, and KFeCl2 ) a(FeCl2+) a(FeCl2+)-1 a(Cl-)-1 ) 4.47 (a ) activity), the concentrations of these main iron species were [Fe3+] ≈ 4 × 10-5 M, [FeOH2+] ≈ 1.44 × 10-4 M, [FeCl2+] ≈ 2.51 × 10-4 M, and [FeCl2+] ≈ 2.24 × 10-4 M in the presence of 6.6 × 10-4 M FeCl3 and 0.2 M Cl- at pH 3.0. Therefore, the prevailing iron complexes were in the following sequence: FeCl2+ > FeCl2+ > FeOH2+ > Fe3+. According to the literature (10, 16, 17), FeCl2+ and FeOH2+ are important to produce Cl• and OH• radical species, respectively, therefore, we deduce there are two pathways contributing to the formation of Cl2•- under the adopted condition. One pathway is through generation of Cl• formed by the photolysis of FeCl2+ complex (10, 17): 2+ FeCl2+ + hν f Fe2+ + Cl· ΦFeCl347nm ) 0.5 (16)

(2) The alternative pathway is through the scavenging of OH• by Cl-, forming ClHO•- radicals, as occurs during Fenton reaction (16), which was proved by the identification of OH• using ESR DMPO spin-trap technique (Figure 3). Cl- + OH· f ClHO·-

(3)

ClHO·- + H+ f Cl· + H2O

(4)

In Figure 3, no signal could be detected in pure water containing DMPO and Fe(III)/Cl- in dark (curves a, b, and c), while the characteristic quartet peaks of the DMPO adduct with a 1:2:2:1 intensity were observed after in situ irradiation

FIGURE 2. Transient absorption spectra photolysis of aqueous solution of (a) 6.6 × × 10-4 M FeCl3 and 87.7 µM BPA, and (c) presence of 0.2 M Cl- at pH 3.0, time delay 7714

9

under laser flash 10-4 M FeCl3, (b) 6.6 87.7 µM BPA in the of 700 ns.

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 20, 2009

FIGURE 3. ESR spectra of radicals trapped by DMPO (a) in pure water in dark, (b) under irradiation, (c) in the presence of 1.32 × 10-4 M FeCl3 and 0.2 M Cl- in dark, (d) under irradiation of 80 s, and (e) under irradiation of 160 s. of 80 s under 355 nm irradiation (curve d), consistant with the similar spectra reported by others for the DMPO-OH• adduct (18, 19), indicating the generation of OH• radical. Furthermore, the signal intensity increased with the irradiation time (curve e). The results suggest Fe(III) in the adopted system contributes to the formation of OH•. Indeed, FeOH2+ is the most photoactive compound to produce OH• radicals (reaction 5) in aqueous solution at pH 2.5-5 (33): 2+ FeOH2+ + hν f Fe2+ + OH· ΦFeOH347nm ) 0.21 (16) (5)

Therefore, we can deduce that the photolysis of FeCl2+ and FeOH2+ in the experimental system, forming Cl• and OH• radical species, plays important roles in the generation of Cl2•-. Noticeably, as an important oxidant, OH• in this system not only facilitates the Cl2•- formation, but also causes the photodegradation of BPA obviously, which is the reason for the difference between the decrease of BPA and the formation of 3-ClBPA and 3,3′-diClBPA in Figure 1. However, it is undeniable that the photochlorination of BPA occurs. Both Cl2•- and Cl• radical species are strong oxidants with E(Cl2•-/Cl-) ) 2.0 V and E(Cl•/Cl-) ) 2.4 V vs NHE (20) that can be involved in electron-transfer, hydrogen abstraction, or addition reactions with organic compounds (20, 21). For example, Alegre and Gerone´s (20) reported the addition rates of Cl• and Cl2•- to aromatic ring were near 1010 M-1 s-1 and 105 M-1 s-1, respectively. Therefore, BPA undergoes a reaction with Cl•/Cl2•- and consequently forms a considerable amount of 3-ClBPA and 3,3′-diClBPA (Figure 1B). Considering the Cl2•- radical is one of the dominant species in aqueous solution containing Cl-, and according to the pathway of phenol chlorination (10), the photochlorination of BPA with Cl2•- is suggested to have one-electron oxidation and hydrogen abstraction to yield bisphenoxyl A radical (BPA•), which then reacts with Cl2•- to yield ClBPA:

But, unfortunately, we did not find the method to identify the existence of BPA•. BPA Photochlorination in the Presence of Fe(III) and FA. In this section, the effect of FA on the photochlorination in simulated natural water was investigated for three reasons: (1) FA is the most photoreactive component of dissolved organic matter (DOM), which plays an important role in the natural aquatic environment for its ubiquitous diffusion and

findings of Gu et al. (25), which inferred that the appearance of a strong absorption band at about 1400 cm-1 means the interaction of natural organic matter with Fe(III) to form the Fe(III)-FA complex, the band at 1402 cm-1 could be assigned to -COO-Fe, which was evidence for the Fe(III)-FA complex. SyF of FA in the presence of Fe(III) in Figure S5 showed the degree of fluorescence quenching with increasing quantities of Fe(III), which was another indicator of formation of Fe(III)-FA complex (26). Due to the formation of Fe(III)-FA complex, light excited FA could effectively inject electrons into Fe(III) leading to the production of Fe(II) and an easy cycle of Fe(III)/Fe(II), herein a continuous production of OH• (reactions 6-9):

FIGURE 4. (A) Photoformation of 3-ClBPA during the irradiation of 87.7 µM BPA under different conditions (the concentration of FA was 3.2 mg L-1), and (B) time evolution of 3-ClBPA as a function of FA added in the presence of Cl- and FeCl3. Conditions: 0.2 M Cl-, 1.32 × 10-4 M FeCl3, pH 6.3 ( 0.1. ability to absorb sunlight when photoinduced processes are considered (22); (2) FA has strong affinity toward metal cations and therefore can serve as effective chelator for Fe(III) ions to form Fe(III)-FA complex, which affects the phototransformation of organic compounds to a great extent (23); (3) CDOM acted as the direct precursor for the yield of CH3Cl through a photochemical process (4). Moreover, to be more coincident with natural water, all experiments in this section were performed under almost neutral condition. The BPA chlorination conducted in different solutions at pH 6.3 under Xe lamp irradiation is summarized in Figure 4. Both 3-ClBPA and 3,3′-diClBPA were identified by GC/MS in all the solutions during the reaction, but only the concentration of 3-ClBPA during the reaction is given in Figure 4, since the concentration of 3,3′-diClBPA is below its quantitative detection limit (QDL ) 10 S/N ) 0.4 ng L-1). It is worth noting that in simulated neutral saline water, the coexistence of Fe(III) and FA promoted the chlorination of BPA significantly (Figure 4A), the maximum concentration of 3-ClBPA was near 16 nM and decreased to 10 nM after 5 h irradiation. While compared among 3-ClBPA concentrations during the irradiation at pH 6.3, it was found that the presence of Fe(III) (1.32 × 10-4 M) did not change the trend of BPA chlorination, and the presence of FA (3.2 mg L-1 as total organic carbon (TOC)) even inhibited BPA chlorination. Therefore, there must be certain combined effects between Fe(III) and FA, which promote BPA chlorination significantly. To investigate the interaction of Fe(III) and FA, FTIR spectra were measured. FTIR has long been used for the structural analysis, particularly for exploring the interactions between metal and ligands according to variations of peaks. Detail descriptions of FTIR spectra (Figure S4) are in S5 of SI. All changes in FTIR spectra were observed as a result of the formation of Fe(III)-FA complex. According to the

FA + O2 + hν f HO·2 /O·2 + products

(6)

HO·2 /O·2 f H2O2

(7)

Fe(III)-FA + hν f FA·+ + Fe(II)

(8)

Fe(II) + H2O2 f Fe(III) + OH· + OH-

(9)

Many articles support the above suggestion. For example, Ou et al. (23) reported similar results that more OH• was produced with the formation of Fe(III)-HA complex, promoting atrazine photodegradation. The scavenging of OH• by chloride forming Cl• and Cl2•- occurs not only in acidic solution as mentioned above (reactions 3 and 4), but also in neutral chloride solution, which is very fast and assigned to the local acidity in the solution (27). In addition, according to reports by Calza et al. (11) that the oxidation of Cl- to Cl• radicals occurs directly on the surface of Fe(III) lattice atom (symbol: >Fe(III) in hematite, the oxidation of Cl- to Cl• radicals direct on Fe(III) colloids is the alternative pathway for Cl• generation (reaction 10 (30)), because under neutral conditions, Fe(III) and Fe(III)-FA are mostly in colloidal form (28) containing many iron atoms on the surface (29): >FeIII(:OH-) + Cl- + hν f Fe(II) + OH- + Cl·

(10)

Thereby, as the results of OH• formation by Fe(III)-FA complex and Cl- direct oxidation on the surface of Fe(III) colloids, BPA chlorination with the coexistence Fe(III) and FA observed in Figure 4A was increased. The formation of Cl• in the presence of Fe(III)-FA complex is proposed in Scheme S1. Effect of FA Concentration. The effect of FA concentration ([FA]) on BPA chlorination was investigated to further illustrate the influences of Fe(III)-FA complex. It is plausible that the BPA chlorination increases as the Fe(III)-FA complex increases, however, BPA chlorination did not increase with the increasing of [FA] in single direction. Figure 4B shows the formation of 3-ClBPA at initial [FA] of 0, 1.0, 1.6, 3.2, and 10 mg L-1 (as TOC). It is interesting to note that when [FA] < 3.2 mg L-1, BPA chlorination was enhanced with the increasing of [FA]. While at high concentration (10 mg L-1), BPA chlorination slowed down obviously. It is understandable that at lower concentration of FA, more Fe(III)-FA complexes are responsible to the enhancement of Cl2•- formation, which is benefit for the formation of 3-ClBPA. Additionally, the average particle sizes of Fe(III) colloids at [FA] of 0, 1.0, 1.6, and 3.2 mg L-1 are 315, 82.8, 58.5, and below 10 nm, respectively, which means that increasing [FA] can increase the proportion of iron in small size and dissolved fraction (