ARTICLE pubs.acs.org/est
Using Chemical Reactivity To Provide Insights into Environmental Transformations of Priority Organic Substances: The Fe0-Mediated Reduction of Acid Blue 129 Salma Shirin† and Vimal K. Balakrishnan*,† †
Water Science and Technology Directorate, Aquatic Ecosystems Protection Research Division, Environment Canada, 867 Lakeshore Road, Burlington, ON, Canada L7R 4A6
bS Supporting Information ABSTRACT: Sulfonated anthracenedione dyes are medium priority organic compounds targeted for environmental assessment under the Government of Canada’s Chemical Management Plan (CMP). Since organic compounds undergo transformations in environmental matrices, understanding these transformations is critical for a proper assessment of their environmental fate. In the current study, we used zero-valent iron (ZVI) to provide insight into reductive transformation processes available to the anthracenedione dye, Acid Blue 129 (AB 129), a dye which is used in the textile industry. At high temperatures, we found that AB 129 was rapidly reduced (within 3 h) after being adsorbed onto the ZVI-surface, whereupon decomposition took place via multiple competitive and consecutive reaction pathways. Reaction products were identified using state-of-the-art accurate mass Liquid Chromatography-Quadrupole Time of Flight-Mass Spectroscopy (LC-QToF-MS). Five transformation products were identified, including a genotoxic (and thus, potentially carcinogenic) end-product, 2,4,6-trimethylaniline. The same products were found at room temperature, demonstrating that the transformation pathways revealed here could plausibly arise from biological and/or environmental reductions of AB 129. Our results demonstrate the importance of identifying reaction product arising from priority substances as part of the environmental risk assessment process.
’ INTRODUCTION Among the many organic pollutants known to be present in aquatic ecosystems, synthetic dyestuffs are a growing concern due to apprehensions that they could pose a serious threat to both environmental and human health.16 Synthetic dyes are extensively used in many fields, including paper production, leather tanning, food coloring, and personal care products (e.g., hair color, deodorant, etc.) as well as in textiles.1,3 Treatment options for the removal of dyes from wastewater effluents include coagulation, adsorption, chemical decomposition (advanced oxidation), and biological treatments.1,7,8 However, these methods have proven inadequate to the task of fully removing these priority organic pollutants, resulting in approximately 20% of the target dyes entering aquatic ecosystems,1 where very little is known about their ultimate fate and impact. Upon entering aquatic ecosystems, small organic molecules such as dyes are expected to partition to sediments that contain organic matter9 where they undergo chemical, biological, or environmental transformations. For example, Maguire found Disperse Blue 79 and its reduced moiety (2-bromo-4,6-dinitroaniline) in the water and sediment near the textile mills at Yamaska river in Quebec, Canada,10 which was attributed to anaerobic degradation of the parent compound in the sediment. Later, Weber and Adams verified this degradation route in their kinetic study of Published 2011 by the American Chemical Society
Disperse Blue 79 reduction.11 In their study, Weber and Adams demonstrated that 2-bromo-4,6-dinitroaniline was produced upon the chemical reduction of Disperse Blue 79 by sodium thiosulfate and also when anoxic sediments were exposed to Disperse Blue 79.11 The use of chemistry to provide insight into environmental or biological transformations has also been examined in cases such as the following: the chemical reduction of Sudan III,5 the oxidative degradation of Orange G using bimetallic nanoparticles,12 the reduction of Orange II using a photoinitiator,13 and the oxidation of Methyl Orange using Fenton’s reagent.14 Of particular significance was the observation that the parent dye molecules underwent transformations that led to the formation of genotoxic aromatic amines.2,5,6,1517 For example, Pielesz et al. showed the formation of benzidine from Acid Red 85 and Direct Blue 7 and the formation of substituted anilines from Sudan III and Disperse Yellow 7.5 The majority of these studies focused on azo dye compounds, leaving a paucity of information relating to the fate of anthracenedione (also known as anthraquinone) dyes beyond the fact that the initial decomposition Received: March 16, 2011 Accepted: November 3, 2011 Revised: October 24, 2011 Published: November 03, 2011 10369
dx.doi.org/10.1021/es202780r | Environ. Sci. Technol. 2011, 45, 10369–10377
Environmental Science & Technology of compounds such as Acid Blue 25 involves a desulfonation step,18 with no information on the nature of the reaction products generated thereafter. While a variety of techniques have previously been used for the analysis of dyes and their degradation products,5,6,1114,1922 recent advances in instrumentation allow for more comprehensive product analyses. For example, liquid chromatographs (LC) coupled to high resolution Quadrupole Time-of-Flight Mass Spectrometry (QToF-MS) enable structural elucidation of unknowns from a complex mixture with a very high degree of accuracy by generating accurate molecular masses, isotopic abundances, and accurate mass measurements of mass fragmentation patterns.23 Currently, the North American risk assessment/risk management (RA/RM) process for environmental pollutants relies on parameters intrinsic to the parent compound: persistence, bioavailability, and inherent toxicity. Degradation products arising from environmental and/or biological transformations of the parent molecule are not considered, owing to the lack of information concerning their formation. Regulatory agencies24 are working to overcome this shortfall by identifying possible degradation products arising from environmental contaminants. For example, the Chemicals Management Plan (CMP) is a Government of Canada program under which priority was attached to developing information on contaminants (including dyestuffs) possessing aromatic azo, anthracenedione, and benzidine-based structures. Reduction processes are particularly important in understanding the fate of anthracenedione dyes, since this category of dyestuff cannot be fixed to fabric without first being reduced to its leuco-form.25 Many studies have demonstrated that zero-valent iron (ZVI) effectively induces the reductive decomposition of a variety of both inorganic and organic compounds, including nitrates,26 chlorinated alkanes and alkenes,27,28 polychlorinated dioxins,29 explosives,30,31 herbicides,32 and even azo dyes.11,12,14,3335 Accordingly, in this work, we present the reduction of Acid Blue 129 (AB 129) by ZVI under oxic conditions and use state-of-theart high resolution LC-QToF-MS to identify novel reaction intermediates and products arising from the transformation process.
’ EXPERIMENTAL SECTION Chemicals. Acid Blue 129 (1-amino-4-(2,4,6-trimethylanilino)anthraquinone-2-sulfonate, 25% dye content, 75% inert, proprietary filler material; CAS 6397-02-0) and zero-valent iron (99% pure, < 212 μm; CAS 7439-89-6) were purchased from Sigma-Aldrich (Oakville, Canada) and used without further treatments or purification. 2,4,6-Trimethylamine (98% pure; CAS 88-05-1) was obtained from Alfa-Aesar (Oakville, Canada), and 1,4-diaminoanthraquinone (90% pure; CAS 128-95-0) and FeSO4 3 7H2O (g99.0% pure; CAS 7782-63-0) were obtained from Sigma-Aldrich and used as received. Acetonitrile (MeCN, HPLC grade, 99.8%) and methanol (MeOH, HPLC grade, 99.8%) were purchased from Caledon, while formic acid (FA, 88%) was obtained from J.T Baker. For solutions preparation and chromatographic purposes, ultrapure water (Milli-Q water, Millipore) was used. Mobile phases were filtered through 0.2 μm ultrafiltration prior to use. Degradation of Acid Blue 129. A. At 100 °C. Aqueous solutions of Acid Blue 129 (22.2 μmol/L) were prepared, and 50.0 mL was transferred to a two-neck round-bottom flask. Zerovalent iron (5.0 g) was added into the flask without any pretreatment. The reaction mixture was then heated under reflux
ARTICLE
conditions and sampled at regular time intervals, collecting both an aqueous and an iron phase. The iron phase was extracted using MeOH (2 25 mL) after which the extract was analyzed by LCQToF-MS, whereas the aqueous phase was analyzed without any further treatment. Two control experiments were also performed. The first, a positive control, involved an aqueous solution of Acid Blue 129 being refluxed in the absence of Fe, while the second, a negative control, involved refluxing Fe in the absence of Acid Blue 129. To investigate the role of iron in the reaction process, two additional experiments were conducted: (i) a 50 mL solution of Acid Blue 129 (22.2 μmol/L) was prepared in dry acetonitrile and refluxed with 5.0 g of zero-valent iron; and (ii) a 50 mL aqueous solution of Acid Blue 129 (22.2 μmol/L) was added to FeSO4 3 7H2O (7.5 mg; 27 μmol) and refluxed. In each case, sampling was performed at regular time intervals, followed by LC-QToF-MS analyses. B. At 22 °C. Into a 20 mL amber glass vial was transferred 10.0 mL of an aqueous solution of Acid Blue 129 (22.2 μmol/L) and 1.0 g of ZVI. The molar ratio of AB129 to ZVI was thus the same as that used in the high temperature study. For each time point, duplicate sample vials were prepared along with two controls (one containing only AB129 without ZVI and the other containing 1.0 g ZVI and 10.0 mL water). These vials were sealed with Teflon-coated screw cap septa and placed in a rotary shaking device (New Brunswick Scientific Co. Inc., Model: TC-7), where the vials underwent end-over-end rotation at 20 rpm at room temperature (22 ( 1 °C). At regular intervals, the vials were sacrificed, collecting both the aqueous and the iron phases. The iron phase was extracted using MeOH (2 25 mL) prior to analysis by LC-QToF-MS, whereas the aqueous phase was analyzed without any further treatment. Instrumental Analysis. All samples were analyzed using an Agilent 1200 HPLC (SL Rapid Resolution), interfaced with an Agilent 6520 high resolution Q-TOF (4 GHz) mass spectrometer equipped with a positive electrospray ionization source. Based on operational requirements, analyses were performed using both scanning and multiple reaction monitoring (MRM) modes. For each run, 5 μL of sample was injected, and samples were eluted at 0.3 mL/min through a C18 column (Kinetex, Phenomenex; 100 mm 2.10 mm; 3.5 μm particle size, pore size 100 Å), using a gradient elution program. Initially, the mobile phase was 95:5 v/v % in 0.1% FA (aq):MeCN, which was then ramped to 5:95 v/v FA:MeCN within 6 min, where it was held for 4 min. Thereafter, the initial conditions were restored within 1 min and the column was allowed to re-equilibrate for a further 3 min. The total run time was thus 13 min. MS conditions were as follows: ESI (+); capillary voltage, 3.0 kV; Fragmentor, 175 V; Skimmer, 65 V; nebulizer pressure 15 psi, drying gas, N2 at a flow rate of 8 L/min; drying gas temperature, 325 °C. For scanning mode: spectral acquisition, 1 scan/s, mass range, 501100 m/z; for MRM: collision energy, 2050 V. Positive reference ions: m/z 121.050873 and 922.009798. Chromatographic and spectral analyses were performed using Agilent MassHunter Workstation software, Qualitative Analysis, Version B.03.01. Using this software, mass spectral patterns were obtained for the chromatographic peak of choice, and empirical molecular formulas were generated for selected spectrum peaks where the experimentally obtained {[(M + n) + H]+, n = 0,1,2,3...}values were compared to their theoretical values up to 5 decimal points. The [M + H]+ ions of unknown reaction 10370
dx.doi.org/10.1021/es202780r |Environ. Sci. Technol. 2011, 45, 10369–10377
Environmental Science & Technology
ARTICLE
intermediates/products of AB 129 were used as the precursor ions to obtain fragment ions for structural identification.
’ RESULTS AND DISCUSSION Degradation of Acid Blue 129. In textile industries, dyeing processes typically require temperatures up to, or even in excess of, 100 °C.36 In order to be readily fixed onto fabric, anthraquinone dyes such as Acid Blue 129 must first undergo reduction to their leuco-form via reaction with a sodium dithionite solution.25 Given that the wastewater effluent generated by textile mills is subsequently released into the aquatic environment, ascertaining the stability and the expected product distribution of AB 129 under reducing conditions at such elevated temperatures is of environmental consequence. However, using sodium dithionite (in pseudofirst order excess) as a reducing agent poses two concerns: first, using dithionite (a soluble reducing agent) would necessitate the use of a quenching agent, and, second, injection of a dithionite salt (+ quenching agent) into the LC-QToF-MS would result in poor chromatography and great susceptibility to matrix effects. By contrast, the use of Fe0 (ZVI) as a reducing agent enables the reaction to be quenched by filtering the solution to remove Fe0, which prevents the reducing agent from being injected into the LC-QToF-MS, thereby improving chromatography and sensitivity. High temperature investigations are also well-known to provide insight into the fate of organic contaminants under more environmentally relevant conditions.37 Accordingly, in this work we examined the reaction between AB 129 and Fe0 at elevated temperatures (100 °C) and, to establish whether our findings could be extrapolated to more environmentally relevant conditions, also at room temperature (22 °C). Figure 1A shows that at 100 °C, Acid Blue 129 completely degraded within 3.5 h of reaction with Fe0 (under aerobic conditions). During the first hour we were able to detect AB 129 on the surface of the ZVI phase, which was consistent with fast initial decline in the aqueous AB 129 concentrations and was attributed to AB 129 partitioning between both the ZVI and solution phases. Within 3 h, AB 129 had completely disappeared from both phases. Meanwhile, Figure 1B demonstrates that at room temperature, the degradation proceeded in the presence of ZVI with a half-life of approximately 22 days. However, at both temperatures, AB 129 was stable in control reactions conducted in the absence of ZVI (Figure 1), demonstrating that the disappearance of AB 129 is a ZVI mediated chemical degradation process. Role of Fe. In the Fe0-H2O-organic contaminant system, there are three main reactions upon the aerobic corrosion of the metal38 2Fe0 þ O þ 2H O f 2Fe2þ þ 4OHð1Þ 2
2
Fe0 þ AB 129 f Fe2þ þ Products
ð2Þ
Fe2þ þ AB 129 f Fe3þ þ Products
ð3Þ
Since the oxidation of Fe0 to FeII in water is concomitant with the release of hydroxide ions (eq 1), it was not surprising that solution pH increased from an initial value of 5.5 to a value of 7.5 within 1 h of reaction, whereupon it remained constant. Furthermore, it was only after the first hour of contact between AB 129 and ZVI that the dye compound began to disappear (Figure 1a). During the first hour, we suggest that the dissolved
oxygen initially present in the oxic system is consumed (eq 1), after which the system became anoxic and eq 1 ceased to be operative. Thereafter, Acid Blue 129, the less reactive oxidizer, served as the electron sink for either Fe0 or Fe2+ or both. Although Fe2+ can be oxidized further to form Fe3+ (eq 3), under aerobic conditions the electrons generated from this oneelectron process will be trapped by triplet oxygen, TO2 (a diradical in the ground state that acts as an electron “sponge”) instead of initiating AB 129 destruction. Indeed, when AB 129 (22 μmol/L) was mixed with FeSO4 (0.540 mMol) at 100 °C (a temperature at which the concentration of free radical scavenging O2 is only 0.153 mM,39 well below the initial FeSO4 concentration), we found that no degradation took place (data not shown), demonstrating that the reaction shown in eq 3 was not operative in this study. However, when we examined the impact of iron powder on refluxed solutions of AB 129 (22.2 μmol/L) in dry acetonitrile (MeCN, boiling point = 82 °C), where Fe2+ oxidation to Fe 3+ is unfavorable,40 we found that AB 129 disappeared, albeit more slowly than it did in water (Figure 2). The discrepancy in rate can be attributed to the decreased polarity of MeCN compared to water, effectively destabilizing the transition states,41 leading to increased activation energies and thus, slower reaction rates. Once again, a lag time was observed. The increased length of the lag time in MeCN compared to that in water (4 h vs 1 h) is attributed to the greater solubility of oxygen in nitrogen-containing solvents.39 The fact that reduction occurred under these circumstances allowed us to further rule out the possibility that ZVI surface-bound Fe2+ triggered the reaction. From these experiments, we concluded that Fe0 served as the reducing agent initiating a direct two-electron degradation of AB 129, as posited by eq 2. Degradation Pathways. Identifying Intermediates by LCQToF-MS. During the course of the high temperature degradation of AB 129 (Figure 1), we observed the formation and disappearance of a variety of intermediates, suggestive of a series of competitive and consecutive pathways. LC-MS analyses (Figure 3) revealed the formation of compounds with masses of 135 Da (A; retention time = 0.9 min), 238 Da (B; rt = 1.09 min), 356 Da (C; rt = 9.3 min), 359 Da (D; rt = 9.1 min), and 391 Da (E; rt = 1.3 min). These unknowns were not observed in the parent solution or in any control reaction. Figure 4 depicts the behavior of compounds A-E throughout the time course study as well as that of AB 129. For illustrative purposes, the peak area of each individual compound is normalized against the maximum area observed for that compound during the time course. Figure 4 reveals that compounds B and C attained maxima at approximately 3.5 h of reaction, by which time AB 129 had disappeared from solution. Therefore, B and C could only be generated directly from AB 129. Both B and C continued to react, and once AB 129 was depleted from solution, there was no further source from which B or C could form, resulting in the appearance of peak maxima at 3.5 h for both B and C. The fact that both B and C were observed indicates that their degradation rates were slower than their formation rates from Acid Blue 129. Since AB 129 (the starting material) cannot be entirely depleted before the products arising directly from AB 129 can even be formed (and hence, detected), we deduced that none of A, D, and E were directly produced from AB 129, arising instead from consecutive reactions. Finally, the fact that A did not disappear during the reaction allowed us to conclude that A was a final product of the degradation. 10371
dx.doi.org/10.1021/es202780r |Environ. Sci. Technol. 2011, 45, 10369–10377
Environmental Science & Technology
ARTICLE
Figure 1. Aerobic reduction of an aqueous solution of Acid Blue 129 (22.2 μM) in the presence of ZVI (a) under reflux conditions (top) and (b) at room temperature (bottom). Data points are the mean of 3 replicates, with error bars corresponding to the standard deviation about the mean. O denotes control reactions, in which Acid Blue 129 is heated in the absence of ZVI.
We then analyzed the chromatographic peaks for compounds A-E using QToF-MS. First, we were able to generate empirical formulas for the experimentally observed precursor masses. These masses were then compared to the theoretical masses associated with the formulas, and based on the close agreement between the experimental and theoretical masses (
