Article pubs.acs.org/est
Electrochemical Treatment of the Antibiotic Sulfachloropyridazine: Kinetics, Reaction Pathways, and Toxicity Evolution Ahmad Dirany,† Ignasi Sirés,‡ Nihal Oturan,† Ali Ö zcan,§ and Mehmet A. Oturan*,† †
Université Paris-Est, Laboratoire Géomatériaux et Environnement (LGE), EA 4508, 5 Bd Descartes, 77454 Marne-la-Vallée Cedex 2, France ‡ Laboratori d’Electroquímica dels Materials i del Medi Ambient, Departament de Química Física, Facultat de Química, Universitat de Barcelona, Martí i Franquès 1-11, 08028 Barcelona, Spain § Anadolu University, Faculty of Science, Department of Chemistry, 26470 Eskişehir, Turkey S Supporting Information *
ABSTRACT: The electro-Fenton treatment of sulfachloropyridazine (SCP), a model for sulfonamide antibiotics that are widespread in waters, was performed using cells with a carbonfelt cathode and Pt or boron-doped diamond (BDD) anode, aiming to present an integral assessment of the kinetics, electrodegradation byproducts, and toxicity evolution. H2O2 electrogeneration in the presence of Fe2+ yielded •OH in the solution bulk, which acted concomitantly with •OH adsorbed at the anode (BDD(•OH)) to promote the oxidative degradation of SCP (kabs,SCP = (1.58 ± 0.02) × 109 M−1 s−1) and its byproducts. A detailed scheme for the complete mineralization was elucidated. On the basis of the action of • OH onto four different SCP sites, the pathways leading to total decontamination includes fifteen cyclic byproducts identified by HPLC and GC-MS, five aliphatic carboxylic acids, and a mixture of Cl−, SO42−, NH4+, and NO3− that accounted for 90−100% of initial Cl, S, and N. The time course of byproducts was satisfactorily correlated with the toxicity profiles determined from inhibition of Vibrio f ischeri luminescence. 3-Amino-6-chloropyridazine and p-benzoquinone were responsible for the increased toxicity during the first stages. Independent electrolyses revealed that their toxicity trends were close to those of SCP. The formation of the carboxylic acids involved a sharp toxicity decrease, thus ensuring overall detoxification.
■
environmental mobility.14 As a result, traces of sulfonamides are present in almost all surface waters,1,15 mainly coming from waste dump leachates,16 animal manure effluents,13,17 and manure waste lagoons from swine farms.18 In particular, veterinary sulfonamides are typically administered through water,13,19 which dramatically increases their environmental impact and toxicity. Moreover, they can be accumulated in various organisms of the food chain.1,17,20 Sulfonamides are resistant to biological degradation;21,22 for example, Ternes et al.23 have reported that only 24% of sulfamethoxazole could be removed upon soil-aquifer treatment. Sulfachloropyridazine (SCP) is a broad spectrum sulfonamide used against both Gram-positive and Gram-negative aerobic bacteria, also being effective against Chlamydia, and it is considered a common pollutant in surface and groundwater. In
INTRODUCTION Antibiotics are among the most commonly detected pharmaceuticals in the aquatic environment because their antibacterial nature prevents effective removal in sewage treatment plants. The occurrence and fate of antibiotics and their transformation products in water streams is recognized as one of the emerging issues in environmental chemistry.1−5 Their ubiquity is particularly alarming due to the potential adverse effects on aquatic ecology6 and human health.4,5,7 Furthermore, they may cause the proliferation of antibiotic resistance of microbial populations in humans and wildlife due to continuous exposure to the low amounts contained in water supplies,3,5,8−11 including tap water.12 This significantly increases the complexity of treating infections and requires the intake on new generation, expensive antibiotics. Sulfonamides are synthetic antimicrobial agents, derivatives of sulfanilamide, which are among the most widely used veterinary antibiotics in aquaculture and animal husbandry and also in human medicine. They are polar amphoteric compounds that are readily soluble in water,13 exhibiting high © 2012 American Chemical Society
Received: Revised: Accepted: Published: 4074
December 23, 2011 February 10, 2012 February 14, 2012 February 14, 2012 dx.doi.org/10.1021/es204621q | Environ. Sci. Technol. 2012, 46, 4074−4082
Environmental Science & Technology
Article
follower. The performance of Pt/carbon-felt and BDD/carbonfelt cells was tested under electro-Fenton conditions, using either a cylindrical Pt mesh (4.5 cm height, i.d. = 3.1 cm) or a 25 cm2 thin-film BDD electrode onto a Nb substrate (CONDIAS GmbH, Germany) as the anode and a large surface area (14.2 cm × 4.3 cm each side, 0.5 cm width) carbon felt (Carbon-Lorraine) as the cathode. The anode was centered in the electrolytic cell, surrounded by the cathode that covered the inner wall of the cell. H2O2 was produced from reduction of O2 dissolved in the solution. Continuous saturation of this gas at atmospheric pressure was ensured by bubbling compressed air. Solutions of 220 mL containing up to 0.21 mM SCP (i.e., 25 mg/L TOC) with 50 mM Na2SO4 and 0.2 mM Fe2+ at the optimum pH value of 3.0 have been electrolyzed at constant current (30−350 mA), always at room temperature.30 Analytical Procedures. Current was supplied by a Hameg HM8040 triple power supply. The mineralization of SCP solutions was assessed from their TOC abatement, which was determined on a Shimadzu VCSH TOC analyzer (see Text S1 of the Supporting Information). The time course of the concentration of SCP and other cyclic compounds including its oxidation byproducts was followed by reversed-phase HPLC using a Merck Lachrom chromatograph equipped with an L-7100 pump, fitted with a Purospher RP-18 5 μm, 25 cm × 4.6 mm (i.d.), column at 40 °C, and coupled with an L-7455 photodiode array detector selected at optimum wavelengths of 299, 245, 254, and 270 nm for ACP, BZQ, pHB, and SCP, respectively. The analyses were carried out isocratically using a methanol/water (1% H3PO4) 20:80 (v/v) mixture as the mobile phase at a flow rate of 0.8 mL/min. The corresponding retention times (tR) for ACP, BZQ, p-HB, and SCP were 3.1, 7.5, 12.8, and 16.0 min, respectively (see Text S2 in Supporting Information). GC-MS analysis was performed using a Thermo Finnigan PolarisQ analyzer equipped with a TRB-5-MS column from Teknokroma. The temperature ramp was 40 °C for 2 min and 10 °C/min up to 280 °C, and the hold time was for 4 min. The temperature of the injection part and transfer line was 250 °C. Helium was used as carrier gas at a flow rate of 1.5 mL/min. In general, the samples for GC-MS were obtained by extraction of the electrolyzed solutions employing 100 mL of ethyl acetate three times, followed by drying of the organic fraction, filtration, and final concentration with a rotary evaporator. The remaining solid extract was dissolved with ethyl acetate for injection. A derivatization step was included in some cases (see Text S3 in Supporting Information). Generated aliphatic carboxylic acids were identified and quantified by ion-exclusion HPLC using an Alltech chromatograph equipped with a Model 426 pump, fitted with a Supelco Supelcogel H 9 μm, 25 cm × 4.6 mm (i.d.), column at room temperature, and coupled with a Dionex AD20 UV detector selected at λ = 220 nm. A 0.1% H3PO4 solution at a flow rate of 0.5 mL/min was used as the mobile phase (see Text S2 in Supporting Information). The inorganic ions released in the treated solutions were determined by IC upon injection into a Dionex ICS-1000 Basic Ion Chromatograph System, using an anion-exchange or cation-exchange column (see Text S4 in Supporting Information). Toxicity Measurements. The toxicity of SCP and of two of its oxidation byproducts (ACP and BZQ) were evaluated on samples collected from solutions at different electrolysis times. Measurements were performed by means of the Microtox method, based on determining the inhibition of the bio-
China, it was detected in swine farm and fishpond samples with concentrations up to 47 μg/L, which suggests its usefulness to reveal livestock source pollution.11 SCP is known to be persistent to conventional bioremediation.24 Some physical methods tested, like adsorption into organophilic zeolite25 or membrane nanofiltration,26 did not lead either to positive results. It is therefore crucial to develop and/or apply more efficient water treatment technologies for the removal of SCP and other sulfonamides from polluted effluents. Recently, the electrochemical advanced oxidation processes (EAOPs) have demonstrated their great ability to efficiently destroy a large variety of toxic and/or biorefractory organic pollutants.27−34 These methods are based on the electrochemical generation of hydroxyl radicals (•OH), either directly by means of water oxidation at the anode or indirectly by total or partial on-site production of the Fenton’s reagent (H2O2 + Fe2+). This radical is a very powerful oxidizing agent that reacts nonselectively with organic molecules leading to their oxidation until reaching a high mineralization degree.35−40 In particular, Fenton-based EAOPs34,41−43 and anodic oxidation42−45 processes have been successfully applied to the treatment of pharmaceutical pollutants in recent years. However, the electrochemical treatment of SCP has not been addressed yet. The present contribution investigates the performance of the electro-Fenton process for the removal of SCP, chosen as a model sulfonamide, from water, using Pt/carbon-felt and boron-doped diamond (BDD)/carbon-felt cells. The effect of operating parameters such as applied current and SCP concentration on the decay kinetics of SCP was first surveyed. The mineralization of all the solutions was simultaneously assessed from the removal of the total organic carbon (TOC) abatement. Cyclic byproducts, short-chain aliphatic carboxylic acids, and released inorganic ions were identified by highperformance liquid chromatography (HPLC), ion chromatography (IC), and gas chromatography−mass spectrometry (GCMS). This has allowed a detailed elucidation of the oxidation pathways that justify the complete mineralization of SCP. Finally, the solution toxicity profiles obtained from the inhibition of Vibrio f ischeri luminescence have been correlated with the time course of byproducts, since the oxidation pathways may lead to the formation of intermediates showing a higher or lower toxicity than the initial pollutant.46−48
■
EXPERIMENTAL SECTION Chemicals. The antibiotic SCP (4-amino-N-(6-chloro-3pyridazinyl)benzenesulfonamide, C10H9ClN4O2S, ≥98%) was reagent grade from Sigma-Aldrich and was used as received. 3Amino-6-chloropyridazine (ACP) and p-benzoquinone (BZQ) were reagent grade from Sigma-Aldrich (≥98%). Reagent grade p-hydroxybenzoic acid (p-HB) from Acros Organics was used as the competition substrate in some kinetic experiments. Oxalic, maleic, malic, glyoxylic, pyruvic, phosphoric, and sulfuric acids; anhydrous sodium sulfate (background electrolyte) and iron(II) sulfate heptahydrate used as the Fe2+ source (catalyst); and ammonium oxalate and sodium nitrate, nitrite, chloride, and sulfate employed in ion chromatography were analytical grade from Fluka, Merck, and Acros Organics. All solutions were prepared with ultrapure water obtained from a Millipore Milli−Q system with resistivity >18 MΩ cm at room temperature. Electrolytic System. All experiments were conducted in an open, cylindrical, undivided glass cell of 6 cm diameter and 250 mL capacity, with vigorous stirring by a magnetic PTFE 4075
dx.doi.org/10.1021/es204621q | Environ. Sci. Technol. 2012, 46, 4074−4082
Environmental Science & Technology
Article
luminescence of the bacteria V. f ischeri. A luminometer Berthold Autolumat Plus LB 953 was employed, according to the international procedure (OIN 11348-3). Bioluminescent bacteria and the activation reagent LCK 487 LUMISTOX were provided by Hach Lange France SAS. The bioluminescence measurements were carried out on blank solutions as well as on electrolyzed solutions initially containing SCP or cyclic byproducts at different concentrations. In all cases, the bioluminescence intensity of V. f ischeri bacteria was measured after 15 min of exposition to the samples, at 15 °C. See description of the toxicity measurement protocol in Supporting Information Text S5.
shows the effect of the applied current for the electro-Fenton treatment of 59.2 mg/L (0.21 mM) SCP using Pt/carbon-felt and BDD/carbon-felt cells. Solutions remained colorless along all the electrolyses. Note that a complete destruction of the antibiotic was achieved in all cases, requiring shorter or longer times depending on the current value and the anode nature. According to decays of Figure 1, the reaction of SCP with hydroxyl radicals can be described by pseudofirst-order kinetics, assuming a quasi-stationary state for •OH concentration: v=−
■
(4)
RESULTS AND DISCUSSION Kinetic Analysis of the Electrochemical Destruction of SCP. The performance of the •OH-mediated oxidative degradation of SCP by electro-Fenton process mainly relies on the indirect electrogeneration of •OH in the bulk solution via Fenton’s reaction: Fe2 + + H2O2 + H+ → Fe3 + + H2O + •OH
The pseudofirst-order rate constant (kapp,SCP) values were calculated from linear regression of the corresponding semilogarithmic plots, yielding correlation coefficients higher than 0.98. Table 1 summarizes the values obtained for the cells with Pt and BDD. Table 1. Apparent Rate Constants for the Electro-Fenton Degradation of Sulfachloropyridazine (SCP) under the Action of •OH, Assuming Pseudofirst-Order Reaction Kinetics from the SCP Decays Shown in Figures 1A,B and S1 (Supporting Information)
(1)
Such radicals are able to react with SCP leading to the formation of oxidation byproducts, whose oxidative cleavage by • OH may cause even the total mineralization of the polluted solutions: SCP + •OH → By‐products
d[SCP] = kabs,SCP[•OH][SCP] = k app[SCP] dt
[SCP] (mM)
[Fe2+] (mM)
I (mA)
kapp, SCP × 101 (min−1)
Pt/carbon felt
0.082 0.21
0.2 0.2
BDD/carbon felt
0.21
0.2
300 30 60 120 200 300 350 60
8.80 0.50 0.78 1.33 1.70 6.00 6.10 1.15
120 200 300
1.40 2.30 6.20
cell
(2)
By‐products + •OH → → → CO2 + H2O + inorganic ions (3)
In the present electro-Fenton system, Fenton’s reagents (H2O2 + Fe2+) were continuously electro(re)generated from two simultaneous cathodic reactions: (i) reduction of ferric ions, initially introduced as a Fe2+ salt, and (ii) reduction of dissolved O2 from compressed air.30 The decay of the SCP concentration under different operating conditions was monitored by HPLC, where it displayed a well-defined peak at tR = 16.0 min. Figure 1
Figure 1A shows the SCP decays for the Pt/carbon-felt cell in the range of 30−350 mA. The total disappearance of SCP became faster at a higher current, as expected from the progressively quicker (re)generation of H2O2 and Fe2+ that led to a higher •OH concentration at a given time. Thus, the kapp,SCP values increased as the current went from 30 to 300 mA, as shown in Table 1. However, a higher current value like 350 mA did not increase the removal efficiency, which can be explained by the progressive enhancement of the parasitic reactions; mainly, the H2 evolution at the cathode, increasingly favored over the H2O2 electrogeneration. In fact, such contribution of the wasting reactions was already given from the lower current values, as demonstrated by the nonproportional increase of kapp,SCP as current was raised. The influence of the initial SCP concentration was also studied using the Pt/carbon-felt cell, by working at 300 mA (Figure S1 in Supporting Information). A much more rapid disappearance was obtained for the treatment of 23.3 mg/L SCP, with total destruction after only 5 min. The longer time required as the initial concentration increases is then due to the presence of a larger number of SCP molecules to be oxidized for the same •OH generation rate, which in turn generates a
Figure 1. Effect of the applied current on the time course of the SCP concentration for the electro-Fenton treatment of 59.2 mg/L (0.21 mM) SCP using (A) Pt/carbon-felt and (B) BDD/carbon-felt cells. 4076
dx.doi.org/10.1021/es204621q | Environ. Sci. Technol. 2012, 46, 4074−4082
Environmental Science & Technology
Article
electrolysis time under conditions of Figure 1, with starting solutions containing 25 mg/L TOC. The influence of the applied current in the cell with Pt can be observed in Figure 2A.
greater amount of byproduct molecules that cause the additional consumption of •OH. The efficiency of the electro-Fenton process strongly depends on the nature of the electrode materials.30 Carbon felt is ideal as the cathode because it favors the Fe2+ regeneration. On the other hand, large O2-overpotential anodes are being widely used in electro-oxidation systems due to the promotion of more active •OH adsorbed at the anode surface. Therefore, BDD was tested as a high oxidation power anode, so that the effect of current was studied in the range of 60−300 mA for the BDD/carbon-felt cell, as depicted in Figure 1B. In this system, the oxidation process of SCP was also accelerated by increasing the applied current because of the higher •OH concentration. By comparing Figure 1A,B, it is evident that the use of BDD causes the acceleration of the SCP destruction. This phenomenon arises from the higher oxidation power of BDD compared to Pt, which is accounted for by the production of more actively adsorbed •OH. The loosely bound BDD(•OH) formed at the anode surface (eq 5) can readily react with organic matter, in contrast to the chemisorbed radicals typically formed at Pt. BDD + H2O → BDD(•OH) + H+ + e−
(5)
The positive effect of increasing current and using BDD instead of Pt is clearly confirmed from the higher values of kapp,SCP in Table 1. The use of the BDD/carbon-felt cell then allows the production of additional amounts of hydroxyl radicals (“heterogeneous •OH”) that act concomitantly with those generated in the bulk via Fenton’s reaction (homogeneous • OH). It can be observed that the comparatively greater performance of BDD is more pronounced at low current values. In fact, higher values than 200 mA minimize significantly the positive role of BDD favoring generation of •OH in the bulk, and the decay rate becomes independent of the anode’s nature. See for example, in Table 1, the very similar kapp,SCP value for BDD and Pt cells at 300 mA. This can be explained by a progressively higher (re)generation rate of Fe2+ and H2O2, so that the amount of •OH accumulated in the bulk tends to be much more relevant than that adsorbed at the anode. Furthermore, high current values favor the self-destruction of the “heterogeneous •OH”. In order to determine the absolute (second order) rate constant (kabs,SCP) for the reaction between SCP and hydroxyl radicals, which has not been reported before, competition kinetics experiments were performed at 60 mA with solutions containing identical concentrations of SCP and p-HB as the competing substrate with a well-known absolute rate constant (kabs,pHB = 2.19 × 109 M−1 s−1). Decays are shown in Supporting Information Figure S2A.49 According to eq 6 and using the linear plot of Figure S2B in Supporting Information,50,51 the value of kabs,SCP was determined to be (1.58 ± 0.02) × 109 M−1 s−1. kabs,SCP = k abs,pHB ×
Figure 2. TOC decay vs electrolysis time for the mineralization of SCP solutions by electro-Fenton with (A) Pt and (B) BDD anode, under the same conditions as those in Figure 1A,B, respectively.
When current increased from 30 to 300 mA, the percentage of TOC removal was raised from 34% to 48%, 67%, 72%, and 79% at 240 min. As expected, the higher accumulation of •OH in the bulk led to a greater mineralization rate because of the simultaneous degradation of SCP and its byproducts. After 600 min, 72%, 82%, 89%, and 92% TOC removal was achieved at 30, 60, 120, and 200 mA, respectively, with almost overall mineralization (≥95% TOC removal) only reached at 300 mA. A similar behavior has been observed for the treatment of other organic pollutants.41,43,51 As current is raised, the higher electrogenerated H2O2 concentration is beneficial for yielding larger amounts of •OH from Fenton’s reaction, but it turns out to be detrimental in terms of current efficiency because of its parallel destruction reaction by •OH.41,42 That is to say, the mineralization current efficiency decreases at higher current values. The optimal value for SCP mineralization was then reached at 300 mA, since a higher current like 350 mA did not accelerate the TOC abatement. The time course of TOC for the BDD/carbon-felt system is depicted in Figure 2B. As in the case of SCP disappearance in Figure 1, the use of BDD gives rise to a remarkable acceleration of the TOC removal at low current values. For example, at 60 mA, 50% mineralization was achieved in only 190 min, thus confirming the great oxidation ability of BDD(•OH), not only toward SCP but also for degrading its byproducts. Current increase led to the expected faster mineralization, so that shorter times of 140, 120, and 100 min were needed for attaining 50% TOC removal at 120, 200, and 300 mA, respectively. However, the most remarkable feature in Figure 2B is that, at 300 mA, the cell with BDD allowed >95% mineralization after only 480 min instead of 600 min needed with Pt. BDD anode contributes to the degradation of all the byproducts, even the most refractory ones. It can then be concluded that the electro-Fenton technology with BDD anode
(kapp,SCP) (kapp,pHB)
(6)
Interestingly, this value is very close to that recently reported for the hydroxylation of another sulfonamide antibiotic, namely sulfamethoxazole (1.6 × 109 M−1 s−1).43 Effect of the Applied Current and Anode Material on the Mineralization of SCP Solutions. The oxidization ability of the Pt/carbon-felt and BDD/carbon-felt cells to mineralize SCP solutions was assessed from their TOC abatement over 4077
dx.doi.org/10.1021/es204621q | Environ. Sci. Technol. 2012, 46, 4074−4082
Environmental Science & Technology
Article
and derivatized samples. Three kinds of byproduct can be observed: (i) benzenesulfonamide derivatives, exhibiting two cyclic rings, (ii) pyridazine derivatives, and (iii) benzenic derivatives. HPLC analyses of samples from solutions of 0.21 mM SCP treated by electro-Fenton at 60 and 300 mA using the Pt/ carbon-felt cell demonstrated the presence of two distinctive peaks in the chromatograms, revealing the formation of BZQ and ACP as the two pre-eminent cyclic byproducts. The time course of their concentration is represented in Figure 3B. A typical accumulation−destruction cycle can be seen in all cases, due to the greater generation from SCP destruction during the first stages followed by the predominant degradation by •OH in the later stages. As in the case of SCP in Figure 1A, a faster disappearance of both byproducts was achieved at the highest current; BZQ was totally destroyed at 40 and 10 min, whereas ACP required 90 and 25 min at 60 and 300 min, respectively. ACP was then more refractory to oxidation, being still present in the solutions once SCP had disappeared (see Figures 1A and 3B). Maximum concentrations of 0.012 and 0.014 mM were attained for BZQ and ACP, regardless of the current tested. It is well-known that the cyclic intermediates undergo oxidation to form polyhydroxylated and/or quinone forms, whose cleavage leads to the formation of short-chain aliphatic carboxylic acids, which are usually harder to oxidize than their parent compounds.30,41 SCP solutions were electrolyzed under the conditions of Figure 2A,B, at 60 mA, and then analyzed by ion-exclusion HPLC. The resulting chromatograms displayed five well-defined peaks corresponding to oxalic, maleic, pyruvic, glyoxylic, and malic acids at tR = 2.9, 3.3, 4.4, 4.7, and 8.5 min, respectively. Their evolution in the Pt/carbon-felt and BDD/ carbon-felt cells is depicted in Figure 4A,B. They were formed
is a very effective technology for the treatment of aqueous SCP solutions. Upon bond cleavage of the SCP molecules, the Cl, N, and S atoms are released into the solution as inorganic ions. Figure 3A shows the evolution of the concentration of the ions
Figure 3. (A) Evolution of the concentration of the inorganic ions released during the electro-Fenton treatments with Pt and BDD anode, under the conditions of Figure 2A,B at 300 mA. (B) Time course of the concentration of the main cyclic byproducts accumulated during the treatments with Pt anode at 60 and 300 mA shown in Figure 2A.
identified during the treatments shown in Figure 2 at 300 mA. Four peaks were found by IC, corresponding to Cl−, SO42−, NH4+, and NO3−. The final concentration of NO3− and SO42− was slightly higher using BDD, which agrees with the faster mineralization mentioned in Figure 2. Also, the amount of NH4+ tended to be higher with BDD, although at the end of the treatments almost identical values were achieved for both cells. The total nitrogen content after 600 min, corresponding to the sum of NH4+ and NO3−, represented 97% and 90% of the initial N content for the BDD and Pt cells, respectively, which confirms the greater performance of the former system. The release of SO42− accounted for more than 90% of the initial S content. A particular behavior was observed for Cl−. In the cell with Pt, a progressive accumulation of this ion was observed, until reaching the expected maximum value of ca. 0.2 mM. In contrast, the release of Cl− was somewhat greater in the first stages with BDD, reaching a maximum concentration after 120 min, whereupon it decreased until a final value of 0.04 mM at 600 min. This event can be explained by the oxidation of Cl− to Cl2 and/or ClO−/HClO, ClO3−, and ClO4− by BDD(•OH) and/or •OH in the bulk.52 The TOC removal values along with a quasi-quantitative release of inorganic ions constitute glaring evidence of the high mineralization degree achieved for the SCP solutions under examination. Identification and Evolution of Oxidation Byproducts. Up to fifteen cyclic byproducts formed upon electro-Fenton treatment of SCP were identified by GC-MS and/or HPLC analysis. Their structures and mass fragmentation data are summarized in Supporting Information Table S1. Figure S3 in the Supporting Information illustrates the total ion chromatograms (TIC) obtained upon GC-MS analysis of underivatized
Figure 4. Time course of the concentration of the main short-chain aliphatic carboxylic acid byproducts formed during the electro-Fenton treatments with (A) Pt and (B) BDD anode, under the conditions of Figure 2A,B at 60 mA, respectively.
since the beginning of the electrolyses and also followed accumulation−destruction cycles. In general, much longer times were needed for their destruction compared to cyclic compounds such as SCP, ACP, and BZQ, due to their much lower reactivity with •OH (maleic is the sole exception). The concentration profiles and maximum values were similar in both systems, but a slower disappearance was observed for 4078
dx.doi.org/10.1021/es204621q | Environ. Sci. Technol. 2012, 46, 4074−4082
Environmental Science & Technology
Article
Figure 5. Reaction pathways for the total mineralization of SCP by electro-Fenton.
carboxylic acids, along with Cl−, SO42−, NH4+, and NO3−. On the basis of the trends shown in Figure 2, the complete transformation of the organic matter into CO2 is proposed. Toxicity Evolution upon Electro-Fenton Treatment of SCP Solutions. The evolution of toxicity of solutions with 0.21 mM SCP over electrolysis time was studied for the Pt/carbonfelt cell at different current values in the range of 30−300 mA, based on the V. f ischeri bacteria luminescence inhibition caused by the presence of SCP and/or its reaction byproducts. The four experimental curves corresponding to the percentage of inhibition, which are depicted in Figure 6A, presented analogous features. In all cases, a strong initial increase of luminescence inhibition percentage occurred, indicating a considerably higher toxicity of the primary SCP byproducts. In the first stages of the electro-Fenton treatment, cyclic compounds are predominant, and they are responsible for maximum values of 99−100% inhibition attained. Such maximum inhibition lasted for a longer time when the applied current was lower, due to a smaller •OH amount that causes a slower destruction of the organic matter. For example, the maximum value remained for 7 and 12 min at 300 and 30 mA, respectively. This was followed by a sharp decrease, suggesting the toxicity decay, and suddenly increased again from 10 min (300 mA) and 15 min (30 mA). The maximum value, which was about 40−60% lower than the first maximum but still was higher than the initial SCP inhibition percentage, was attained at a time that matched quite well with the maximum concentrations of ACP and BZQ in Figure 3B. Note that these secondary peaks were also shifted toward longer time at a
oxalic and glyoxylic acids in the Pt cell, in agreement with the above-mentioned lower oxidation ability of such system. In fact, these two acids always reached the highest concentrations and exhibited the largest persistence, suggesting that they can be formed from different routes to become the ultimate reaction byproducts. Their presence may partly explain the residual TOC remaining at the end of treatments (see Figure 2). On the basis of the findings discussed above, a scheme including various oxidation pathways leading to the total mineralization of SCP is proposed in Figure 5. The attack of • OH onto four different reaction sites of SCP is suggested, which yields different primary cyclic byproducts according to pathways A−D. Routes A and B involve the formation of five benzenesulfonamides caused by consecutive hydroxylation steps with or without Cl− release. This occurs simultaneously to cleavage of the SCP structure, which leads to the formation of up to ten pyridazine and benzenic derivatives as shown in pathways C and D. The former involves attack of •OH onto the S position to first form ACP and 4-amino-3-hydroxybenzenesulfonic acid, whereas pathway D leads to the appearance of ACP and p-aminophenol as primary byproducts, with SO42− release. Note that, similarly, also the five benzenesulfonamides could participate in the formation of compounds arising from routes C and D. ACP and BZQ have been highlighted because they were the most abundant byproducts, thus emphasizing the importance of their respective degradation routes. Successive hydroxylation onto different positions of the three primary molecules formed in C and D makes them fragile and promotes their final ring breakage to yield the five above-mentioned 4079
dx.doi.org/10.1021/es204621q | Environ. Sci. Technol. 2012, 46, 4074−4082
Environmental Science & Technology
Article
higher toxicity than SCP, which can justify the presence of a strong initial luminescence inhibition plateau at short times. No EC50 values are found for ACP in the literature. On the other hand, the much larger EC50 values going from 250 to 450 mg/L as reported elsewhere56 for the short-chain carboxylic acids such as maleic, fumaric, acetic, oxalic, and formic completely agree with the weaker luminescence inhibition values observed in the final stages of Figure 6A,B. After demonstrating in previous sections the fast, total destruction of SCP, as well as the almost overall mineralization of its reaction byproducts, the great ability of the electro-Fenton process to ensure the total detoxification has been confirmed, being evident that it can be an interesting alternative for the remediation of sulfonamide antibiotics in waters.
■
Figure 6. Inhibition of luminescence of V. f ischeri bacteria, after an exposure time of 15 min, observed for the electro-Fenton mineralization of (A) solutions with 0.21 mM SCP at different current values and (B) solutions with 0.014 mM ACP or 0.012 mM BZQ at 60 mA, using the Pt/carbon-felt cell.
ASSOCIATED CONTENT
S Supporting Information *
Text S1−S4, Table S1, and Figures S1−S3. This information is available free of charge via the Internet at http://pubs.acs.org.
■
lower current. The toxicity decayed again and, after 120 min, the inhibition percentage reached its minimum value, which can be explained by the destruction of SCP and its cyclic byproducts; only the aliphatic carboxylic acids remained in the solution. Such acids present a known to exert very low toxicity toward V. fischeri bacteria. In order to complete the toxicity evaluation, the behavior of solutions of the two major cyclic byproducts upon electro-Fenton treatment was further studied by means of independent electrolyses. Contribution of the Major Cyclic Byproducts (ACP and BZQ) to the Toxicity Evolution Profiles. The treatment of organic pollutants exhibiting cyclic/aromatic structures by advanced oxidation processes may give rise to oxidation byproducts being more toxic than the initial molecules, and thus, it is important to assess the behavior of such intermediates.46,48,53 With this aim, the evolution of toxicity was studied for solutions containing BZQ or ACP using the Pt/carbon-felt cell, at 60 mA, with initial concentrations of 0.012 and 0.014 mM, respectively, which corresponded to the maximum amounts quantified during the electro-Fenton treatment of 0.21 mM SCP (see Figure 3B). The percentage of luminescence inhibition obtained in both cases can be seen in Figure 6B. The toxicity profiles were characterized by a strong and wide initial luminescence inhibition plateau, comprised between about 0 and 30 min. This was followed by a relatively rapid, almost complete decrease taking place within ca. 20 min for both compounds, in agreement with their destruction at slightly different rates. These profiles suggest that BZQ and ACP possess toxicity trends that are relatively close to those of SCP. It is interesting to note that the decrease of toxicity of ACP and BZQ solutions took place within electrolysis times comparable to those observed for the disappearance of both intermediates (Figure 3B), confirming that these two byproducts were, at least partly, responsible for the second toxicity peak during the SCP treatment of Figure 6A. Furthermore, comparison of EC50 values of SCP, BZQ, and ACP corroborates the obtained results. Indeed, the EC50 value for SCP was reported to be 1.96 mg/L toward Escherichia coli54 and 53.7 mg/L toward V. f ischeri,55 which differs significantly from the value of
