J. Phys. Chem. A 2010, 114, 6613–6621
6613
Degradation of Atrazine by Electrochemical Advanced Oxidation Processes Using a Boron-Doped Diamond Anode Nu´ria Borra`s,† Ramon Oliver,† Conchita Arias,‡ and Enric Brillas*,‡ Unitat de Quı´mica Industrial, Escola UniVersita`ria d’Enginyeria Te`cnica Industrial de Barcelona, UniVersitat Polite`cnica de Catalunya, Comte d’Urgell 187, 08036 Barcelona, Spain, and Laboratori d’Electroquı´mica dels Materials i del Medi Ambient, Departament de Quı´mica Fı´sica, UniVersitat de Barcelona, Martı´ i Franque`s 1-11, 08028 Barcelona, Spain ReceiVed: April 20, 2010; ReVised Manuscript ReceiVed: May 10, 2010
Solutions of 30 mg L-1 of the herbicide atrazine have been degraded by environmentally friendly electrochemical advanced oxidation processes (EAOPs) such as anodic oxidation (AO), electro-Fenton (EF), and photoelectro-Fenton (PEF) using a small open and cylindrical cell with a boron-doped diamond (BDD) anode. AO has been carried out either with a stainless steel cathode or an O2 diffusion cathode able to generate H2O2. Hydroxyl radicals (•OH) formed at the BDD surface in all EAOPs and in the bulk from Fenton’s reaction between added Fe2+ and electrogenerated H2O2 in EF and PEF are the main oxidants. All treatments yielded almost overall mineralization, although the rate for total organic carbon (TOC) removal is limited by the oxidation of persistent byproducts with •OH at the BDD surface. In AO, TOC abatement is enhanced by parallel electrochemical reduction of organics at the stainless steel cathode, while in PEF, it also increases from additional photolysis of intermediates by UVA light under the synergistic action of •OH in the bulk. The effect of current and pH on the degradative behavior of EAOPs has been examined to determine their optimum values. Atrazine decay always follows a pseudo-first-order reaction, being more rapidly destroyed from •OH in the bulk than at the BDD surface. Aromatic intermediates such as desethylatrazine, desethyldesisopropylatrazine, and cyanuric acid and short linear carboxylic acids such as formic, oxalic, and oxamic have been identified and quantified by reversed-phase and ion-exclusion HPLC, respectively. Released inorganic ions such as Cl-, NO3-, and NH4+ have been followed by ionic chromatography. 1. Introduction Electrochemical advanced oxidation processes (EAOPs) are environmentally friendly technologies that have recently received increasing attention to mineralize persistent organic pollutants (POPs) for water remediation.1-3 These techniques are based on the electrogeneration of different kinds of hydroxyl radicals (•OH), which is the second strongest oxidant known after fluorine, with a high standard reduction potential (E°(•OH/ H2O) ) 2.80 V versus SHE). •OH can then nonselectively react with POPs, giving dehydrogenated or hydroxylated derivatives until total mineralization is achieved. Electrochemical oxidation or anodic oxidation (AO) is the most common EAOP for the treatment of wastewaters with low organic contents.1,2,4-22 This method consists of the use of a high O2 overvoltage anode (M) to oxidize the pollutants contained in the electrolytic cell. When the applied current is high enough, organics are mainly degraded by mediated oxidation with a hydroxyl radical (M(•OH)) formed as an intermediate of O2 evolution from water discharge2,4,5
M + H2O f M(•OH) + H+ + e-
(1)
The oxidation power of AO depends on the reactivity of M(•OH) species generated at the anode surface. Recently, it has been * To whom correspondence should be addressed. Tel: +34 934021223. Fax: +34 934021231. E-mail:
[email protected]. † Universitat Polite`cnica de Catalunya. ‡ Universitat de Barcelona.
found that boron-doped diamond (BDD) thin-film electrodes are preferable for this technique, owing to their greater oxidation ability than other conventional anodes such as PbO2,6,19 Ti/IrO2,6 RuO2,12 and Pt.11,16,20,21 BDD can completely mineralize many aromatics and carboxylic acids4-21 because it possesses an inert surface with low adsorption properties, remarkable corrosion stability, and extremely wide potential windows in an aqueous medium, giving rise to a very weak electrode-•OH interaction that results in a higher O2 overvoltage than other anodes and an enhancement of POP degradation with reactive BDD(•OH) formed from reaction 1. In an undivided cell, the mineralization power of AO can be affected by the reduction of POPs on the cathode.2,3 This is not feasible when a carbon polytetrafluoroethylene (PTFE) O2 diffusion cathode is utilized since it mainly supplies continuously H2O2 to the contaminated solution from the two-electron reduction of O2 gas directly injected into it16,20
O2(g) + 2H+ + 2e- f H2O2
(2)
This method is called AO with electrogenerated H2O2 (AOH2O2) and involves the degradation of POPs by different species such as M(•OH) and other weaker oxidizing agents like H2O2 and a hydroperoxyl radical (HO2•) produced from its oxidation at the anode3,11
10.1021/jp1035647 2010 American Chemical Society Published on Web 05/27/2010
6614
J. Phys. Chem. A, Vol. 114, No. 24, 2010
H2O2 f HO•2 + H+ + e-
Borra`s et al.
(3)
The oxidation power of electrogenerated H2O2 can be enhanced by the electro-Fenton (EF) method.3,16,20,23-33 This EAOP consists of the addition of a small amount of catalytic Fe2+ to the acidic contaminated solution to react with hydrogen peroxide generating •OH and Fe3+ in the bulk from Fenton’s reaction34
Fe2+ + H2O2 f Fe3+ + •OH + OH-
(4)
Typical cathodes for EF are carbon-felt25,28-33 and O2 diffusion16,20,23,26-28 electrodes allowing an efficient O2 reduction from reaction 2. An advantage of this procedure is the regeneration of Fe2+ from the cathodic reduction of Fe3+, thereby promoting reaction 4 and favoring the continuous production of oxidant • OH for POPs destruction.28,29 A variant of the EF process is photoelectro-Fenton (PEF), where the contaminated solution is irradiated by an UVA light of λmax ) 360 nm,16,20,23,26,27 which photolyzes the Fe(OH)2+ complex, the predominant species of Fe3+ in the pH range of 2.5-5.0, as follows34,35
Fe(OH)2+ + hν f Fe2+ + •OH
(5)
In PEF, a greater amount of •OH is directly formed from reaction 5, and also, a greater quantity of Fe2+ is regenerated to accelerate Fenton’s reaction 4. Moreover, complexes between Fe(III) and the intermediates can be rapidly photodecomposed, increasing the oxidation ability of this method. Atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine) has been the most widely used pre- and postemergency herbicide in the world for combating grassy and broadleaf weeds in sorghum, corn, rangeland, and sugar cane. This compound has a low solubility of near 34 mg L-1 at 20 °C in neutral aqueous medium and has been found in ground and surface waters because its half-life is as long as about 100 days.22,33 Its presence in the aquatic environment is a potential danger for public health since it is considered an endocrine disruptor.36 For this reason, atrazine belongs to the priority list of 76 substances of the Water Framework of the Directive 2000/60/ EC of the European Parliament, and it has been banned in several countries,37 although its presence and that of its metabolites in natural waters will continue to last for several years.33 Many research efforts have been made to develop potent and environmentally friendly techniques for atrazine removal from aqueous media. The degradation of this POP has been studied by biological methods,38,39 chemical and photochemical processes,40 photodegradation,41 and chemical, photochemical, and photocatalytic advanced oxidation processes (AOPs) with •OH generation including O3/H2O2,42 TiO2/UV,43-47 H2O2/UV,48,49 H2O2/Fe2+,49,50 and H2O2/Fe2+/UV or H2O2/Fe3+/UV.49,51 However, no total mineralization of atrazine is reached using the above methods since cyanuric acid (2,4,6-trihydroxy-1,3,5triazine) is detected as the predominant byproduct, along with other substituted triazines. The same behavior has been reported by several EAOPs. Thus, AO with a Pt anode only allows the conversion of atrazine to cyanuric acid without any formation of CO2.22 In contrast, the latter persistent compound can be mineralized by AO using an undivided cell with a BDD anode and a stainless steel (SS) cathode,9 occurring more rapidly with
increasing current and pH, although saturated solutions of atrazine only undergo about 70 and 90% total organic carbon (TOC) decay operating in acid and neutral media at 50 mA cm-2, respectively, without overall removal of cyanuric acid. No better results have been described by applying EF to treat 250 mL of a 42 mg L-1 atrazine solution with 0.1 mM Fe3+ (pH 3) as the catalyst using an undivided cell with a carbonfelt cathode since at 250 mA, 63% TOC decay was found with a Pt anode, which raised to 82% with a BDD anode.33 More research is then required to show if this POP can be efficiently mineralized by EAOPs with a BDD anode. This paper reports a comparative study on the degradation of 30 mg L-1 atrazine solutions by AO, EF, and PEF with an undivided cell containing a BDD anode in order to clarify the oxidative action of the different hydroxyl radicals produced in these EAOPs. The former technique was tested with a stainless steel (AO-SS) or O2 diffusion (AO-H2O2) cathode to ascertain if electrochemical reduction can affect the degradative process. The effect of applied current and pH on the degradation power of each method was examined. The decay of the herbicide and the evolution of its byproducts, with special attention for cyanuric acid as the most persistent aromatic, were determined by chromatographic techniques. 2. Experimental Section 2.1. Chemicals. Atrazine and five of its possible primary byproducts22,33 such as desethylatrazine, desisopropylatrazine, desethyldesisopropylatrazine, desethyldesisopropyl-2-hydroxyatrazine, and cyanuric acid were reagent-grade supplied by Sigma. Formic, oxalic, and oxamic acids were reagent-grade purchased from Panreac. Sulfuric acid, anhydrous sodium sulfate, ferrous sulfate heptahydrate, and sodium hydroxide were analyticalgrade from Merck and Fluka. Solutions were prepared with highpurity water obtained from a Millipore Milli-Q system with resistivity > 18 MΩ cm at 25 °C. Organic solvents and other chemicals used were either HPLC or analytical-grade supplied by Aldrich, Lancaster, Merck, and Panreac. 2.2. Electrolytic Systems. All electrolyses were conducted in an open and undivided cylindrical cell containing 100 mL of solution stirred with a magnetic bar at 800 rpm to ensure its homogenization and the transport of reactants toward/from the electrodes. The solution temperature was regulated at 35 °C by circulating external thermostatted water through the double jacket of the cell. The anode was a BDD thin-film electrode provided by Adamant Technologies (La-Chaux-de-Fonds, Switzerland) and synthesized by the hot filament chemical vapor deposition technique on single-crystal p-type Si(100) wafers (0.1 Ω cm, Siltronix). Before its use in the electrolytic assays, it was polarized for 60 min in a 0.05 M Na2SO4 solution at 300 mA to remove the impurities of its surface. The cathode was either a stainless steel (AISI 304 grade) sheet for AO-SS or a carbon-PTFE from E-TEK (Somerset, NJ, USA) for AO-H2O2, EF, and PEF. The preparation of the latter cathode is described elsewhere,23 and it was fed with pure O2 at 12 mL min-1 for continuous H2O2 generation from reaction 2. The geometric area of all electrodes in contact with the solution was 3 cm2, and the interelectrode gap was about 1 cm. PEF trials were made under irradiation with a Philips TL/6W/08 fluorescent black light blue tube of λmax ) 360 nm, placed at the top of the open cell at 7 cm above the solution. It yielded a photoionization energy input of 1.4 W m-2, as detected with a NRC 820 laser power meter. Comparative degradations of 100 mL of solutions containing 30 mg L-1 of atrazine in 0.05 M Na2SO4 were performed with
Degradation of Atrazine by EAOPs
J. Phys. Chem. A, Vol. 114, No. 24, 2010 6615
AO-SS, AO-H2O2, EF, and PEF. The influence of constant applied current between 100 and 450 mA and pH in the range of 2.0-7.0 on the oxidation power of each process was examined. In EF and PEF, 0.5 mM Fe2+ was employed as the catalyst since this content was found optimal for analogous treatments of aromatic pollutants.3 2.3. Apparatus and Product Analysis Procedures. The solution pH was measured with a Crison GLP 22 pH meter. Electrolyses were performed with an Amel 2053 potentiostatgalvanostat. Prior to analysis of aliquots withdrawn from electrolyzed solutions, they were filtered with 0.45 µm PTFE filters from Whatman. The solution TOC was determined with a Shimadzu VCSN total organic carbon analyzer. Reproducible TOC values with an accuracy of (1% were found by injecting 50 µL aliquots into the analyzer. These data allowed estimating the mineralization current efficiency (MCE, in %) for each treated solution at a given time t (h) by the equation16,20
MCE )
nFVs∆(TOC)exp 4.32 × 107mIt
× 100
(6)
where F is the Faraday constant (96487 C mol-1), Vs is the solution volume (L), ∆(TOC)exp is the solution TOC decay (mg L-1), 4.32 × 107 is a conversion factor to homogenize units (3600 s h-1 × 12000 mg mol-1), m is the number of carbon atoms of atrazine (eight C atoms), and I is the applied current (A). The number of electrons (n) consumed per atrazine molecule was taken as 70, assuming that it is mineralized to CO2 with the release of Cl- and NO3- as the main primary ions, as will be discussed below. This total process can be written according to the following reaction
C8H14ClN5 + 31H2O f 8CO2 + Cl- + 5NO3 + 76H+ + 70e- (7) Atrazine decay and the evolution of its aromatic byproducts were followed by reversed-phase HPLC using a Waters 600 liquid chromatograph fitted with a Spherisorb ODS 5 µm, 150 mm × 4.6 mm (i.d.), column at room temperature, coupled with a Waters 996 photodiode array detector selected at λ ) 280 nm. Carboxylic acids were detected by ion-exclusion HPLC using the above liquid chromatograph fitted with a Bio-Rad Aminex HPX 87H, 300 mm × 7.8 mm (i.d.), column at 35 °C and the photodiode array detector selected at λ ) 210 nm. In all HPLC measurements, 20 µL aliquots were injected into the chromatograph. The mobile phase was a 40:60 (v/v) acetonitrile/ water mixture at 0.8 mL min-1 and 4 mM H2SO4 at 0.6 mL min-1 for reversed-phase and ion-exclusion HPLC, respectively. Inorganic ions were quantified by ionic chromatography with a Shimadzu 10 Avp HPLC coupled with a Shimadzu CDD 10 Avp conductivity detector by injecting 25 µL aliquots. A Shodex IC YK-421, 125 mm × 4.6 mm (i.d.), cation column under circulation of a mobile phase composed of 5.0 mM tartaric acid, 2.0 mM dipicolinic acid, 24.2 mM boric acid, and 15.0 mM corona ether at 1.0 mL min-1 and 40 °C was employed to quantify the NH4+ ion at a retention time (tr) of 5.0 min. A Shim-Pack IC-A1S, 100 mm × 4.6 mm (i.d.), anion column and 1.0 mM p-hydroxybenzoic acid and 1.1 mM N,N-diethylethanolamine of pH ) 7.9 as the mobile phase at 1.5 mL min-1 and 40 °C were used for the measurements of Cl- (tr ) 3.2 min) and NO3- (tr ) 7.2 min) contents.
Figure 1. TOC removal with consumed specific charge for the degradation of 100 mL of a 30 mg L-1 atrazine solution in 0.05 M Na2SO4 of pH 3.0 using an open and undivided cylindrical cell with a 3 cm2 boron-doped diamond (BDD) anode at 300 mA and 35 °C. (0) Anodic oxidation with a 3 cm2 stainless steel cathode (AO-SS). The other three methods used a 3 cm2 O2 diffusion cathode: (O) anodic oxidation with electrogenerated H2O2 (AO-H2O2), (b) electro-Fenton with 0.5 mM Fe2+ (EF), and (2) photoelectro-Fenton with 0.5 mM Fe2+ under 6 W UVA irradiation of λmax ) 360 nm (PEF).
3. Results and Discussion 3.1. Comparative Degradation by EAOPs. To clarify the oxidation power of the different methods tested, a first series of trials was made by electrolyzing 30 mg L-1 atrazine solutions (corresponding to 13.4 mg L-1 TOC) in 0.05 M Na2SO4 of pH 3.0 at 300 mA (current density, 100 mA cm-2). In all cases, the solution remained colorless, and its pH decreased to 2.6-2.8 at the end of 360 min of electrolysis. The TOC removal against consumed specific charge (Q, in Ah L-1) obtained for these experiments is depicted in Figure 1. A gradual decay of TOC can always be observed up to 12 Ah L-1 (240 min), whereupon the solution becomes hardly degraded, attaining 90-92% of final decontamination. That means that almost total mineralization can only be reached for all EAOPs because of the formation of very stable byproducts in small extent (10% as the maximum) that cannot be destroyed under the action of electrogenerated oxidants, direct electron transfer at the BDD anode, and/or direct electrochemical reduction at the SS cathode. Results of Figure 1 also evidence that the degradation ability of EAOPs tested follows the sequence AO-H2O2 ≈ EF < AOSS ≈ PEF. This trend is rather surprising if one takes into account that previous work on aromatic pollutants reports a faster degradation by EF than that by AO, which is even strongly increased using PEF.3,16,20,26-28 The similar degradation rate obtained for the AO-H2O2 and EF treatments of the heteroaromatic atrazine can then be related to the persistent nature of most byproducts that are mineralized by BDD(•OH) formed from reaction 1 but not with •OH produced by Fenton’s reaction 4. Partial electro-oxidation of atrazine and its intermediates (e.g., cyanuric acid) by direct electron transfer at the BDD anode is also possible.2,3 This behavior also discards a significant participation in the degradation process of weaker oxidants such as H2O2 electrogenerated by reaction 2 and HO2• formed from reaction 3, as well as other produced oxidizing agents at the BDD anode such as the peroxodisulfate ion (S2O82-) by reaction 8 and ozone by reaction 9.2,5 22SO24 f S2O8 + 2e
(8)
6616
J. Phys. Chem. A, Vol. 114, No. 24, 2010
3H2O f O3(g) + 6H+ + 6e-
Borra`s et al.
(9)
The fact that AO-SS is slightly more potent than AO-H2O2 can be explained by the parallel direct electrochemical reduction of some intermediates at the SS cathode giving species that are more easily destroyed and mineralized with BDD(•OH). As can be seen in Figure 1, the maximum difference between these methods is reached at 4.5 Ah L-1 (90 min), with TOC removals of 57% for AO-SS and 40% for AO-H2O2. On the other hand, the superiority of the PEF process compared with EF and AOH2O2 can be due to the additional photodecomposition of some intermediates that enhances the mineralization of the organic matter. This point will be more extensively discussed below from the detection and quantification of the main stable byproducts. Note that the small mineralization action of •OH during EF is also indicative of little influence of the greater production of this radical from reaction 5 on the oxidation power of PEF. 3.2. Effect of Current and pH on the Degradation Rate and Mineralization Current Efficiency. The applied current is an important experimental parameter that can affect the oxidation ability of EAOPs tested.2,3 To check this possibility, comparative degradation of 30 mg L-1 atrazine solutions at pH 3.0 were made for AO-SS, AO-H2O2, EF, and PEF at 100, 300, and 450 mA. Figure 2a and b exemplifies the corresponding TOC decay with electrolysis time and specific charge, respectively, in the case of AO-H2O2 and PEF treatments. At 100 mA, a quite similar and slow degradation rate can be observed in Figure 2a for both methods, reaching close to 85% TOC removal at 540 min of electrolysis. Similar TOC trends were found for the AO-SS and EF techniques at this current. This behavior indicates that the mineralization process of all EAOPs at 100 mA is limited by the slow destruction of very persistent intermediates with BDD(•OH), without any significant participation of other electrogenerated oxidants and/or reduction processes. Figure 2a shows a high increase in TOC removal when the current rises from 100 to 300 mA, attaining ∼90-92% mineralization in 360 min, as expected from the concomitant production of a greater amount of BDD(•OH) that oxidizes much more rapidly the limiting persistent byproduct. At 300 mA, a greater generation of •OH from reactions 4 and/or 5 also takes place because of the electrogeneration of greater H2O2 concentration.11,26 This can produce higher contents of species that can be photodecomposed by UVA light, thus explaining the slightly higher oxidation power of PEF than that of AOH2O2 and EF under these conditions (see Figures 1 and 2a). Compared with 100 mA, the larger acceleration of the electrochemical reduction of some byproducts on the SS cathode at 300 mA can be accounted for by the quicker TOC removal for AO-SS than that for AO-H2O2 at this current (see Figure 1). In contrast, when 450 mA is applied, only a slight increase in the degradation rate with respect to that for 300 mA is found in all cases (see Figure 2a), indicating that the process is controlled by mass transfer of persistent byproducts toward the anode. Then, the excess of current is consumed in parasitic reactions of hydroxyl radicals involving mainly the anodic oxidation of BDD(•OH) to O2 via reaction 10 and the dimerization of • OH to H2O2 by reaction 11 or its destruction with H2O2 and Fe2+ from reactions 12 and 13, respectively,2,3 as well as in the production of more S2O82- from reaction 8 and O3 from reaction 9.
2BDD(•OH) f 2BDD + O2(g) + 2H+ + 2e-
(10)
2•OH f H2O2
(11)
H2O2 + •OH f HO2• + H2O
(12)
Fe2+ + •OH f Fe3+ + OH-
(13)
The action of these side reactions can be more clearly observed in Figure 2b for AO-H2O2 and PEF treatments, where the corresponding TOC decay against Q at 450 mA is much slower than those found at 100 and 300 mA. That means that increasing the current from 300 to 450 mA causes a larger consumption of the applied specific charge by the acceleration of waste reactions like reactions 8-13, giving a relatively lower quantity of organic oxidation events. This phenomenon is also associated with a loss in efficiency of processes. Figure 2c depicts the mineralization current efficiency calculated from eq 6 versus Q for the experiments of Figure
Figure 2. Effect of applied current on TOC abatement with (a) electrolysis time and (b) consumed specific charge for the degradation of 100 mL of a 30 mg L-1 atrazine solution in 0.05 M Na2SO4 of pH 3.0 at 35 °C by (4, O, 3) AO-H2O2 and ([, 2, 1) PEF. Current: (4, [) 100 mA, (O, 2) 300 mA, and (3, 1) 450 mA. Plot (c) gives the corresponding mineralization current efficiency calculated from eq 6.
Degradation of Atrazine by EAOPs
Figure 3. Effect of pH on TOC removal for the degradation of 100 mL of a 30 mg L-1 atrazine solution in 0.05 M Na2SO4 at 300 mA and 35 °C using (a) AO-H2O2 and (b) PEF. Solution pH: (3, b) 2.0, (O, 2) 3.0, (9) 4.0, (0, 1) 6.0, and ([) 7.0. In the three latter media, the pH was continuously adjusted to its initial value by adding small volumes of 0.05 M NaOH.
2b. Low MCE values are reached for atrazine degradation because of the slow reaction of its intermediates with electrogenerated oxidants. Decreasing maximum efficiencies of 3.0, 2.5, and 2.0% with increasing currents of 100, 300, and 450 mA can be seen at the first stages of AO-H2O2. Under these conditions, maximum efficiencies of 3.1, 3.1, and 2.0% are determined for PEF. These trends agree with the formation of smaller relative amounts of oxidants like BDD(•OH) and •OH with raising current because of the concomitant acceleration of waste reactions. The existence of a maximum value for MCE is indicative of a fast conversion to CO2 of several byproducts at short electrolysis time and the parallel formation of persistent intermediates that are slowly destroyed at longer time. Moreover, the presence of less organic matter in solution with prolonging electrolysis also causes a drop in efficiency.2 The above results indicate that a current of 300 mA (current density 100 mA cm-2) is optimal for atrazine degradation since it allows reaching of the higher degradation rate with an efficiency similar to that of lower currents. This current was then used to investigate the influence of solution pH in the range of 2.0-7.0 on all processes. Solutions starting from pH > 4.0 were quickly acidified during electrolysis, and for this reason, these trials were made with continuous pH regulation to their initial value by adding 0.05 M NaOH. In contrast, the treatments for initial pH 2.0 and 3.0 were carried out without pH regulation because the solution pH dropped slowly to final values of 1.9 and 2.7. Figure 3a illustrates the small effect of the pH on the degradation of atrazine in AO-H2O. Similar TOC-Q plots can be observed for pH 2.0 and 3.0, while slightly smaller TOC decay takes place at pH 6.0, although in all cases, about 85% mineralization is already attained at 12 Ah L-1 (240 min). This slight deceleration in TOC removal at pH 6.0 can be associated with the formation of byproducts that react more slowly with BDD(•OH) since the concentration of this electrogenerated
J. Phys. Chem. A, Vol. 114, No. 24, 2010 6617 species is expected to be pH-independent.2,8,11 The same behavior was obtained for the AO-SS and EF processes. A more significant influence of pH can be seen in Figure 3b for the PEF treatment, where TOC undergoes an analogous abatement starting from pH 2.0 and 3.0, which is progressively lower as the pH raises from 3.0 to 7.0. For example, mineralization degrees of 80% at pH 2.0, 77% at pH 3.0, 63% at pH 4.0, 53% at pH 6.0, and 50% at pH 7.0 are determined at 6 Ah L-1 (120 min). This tendency can be related to the gradual smaller production of •OH from Fenton’s reaction 4, which is maximum at pH 2.8.34 Although this radical does not act as the limiting oxidizing species, it can favor the generation of intermediates that can be photodecomposed by UVA light in PEF, thereby enhancing the oxidation power of this method compared with that of AO-H2O2 and EF (see Figure 1). Evidence on this action of the •OH radical will be discussed below. Accordingly, the decay in •OH concentration with increasing pH leads to lower amounts of byproducts that can be photolyzed, decreasing the degradation rate of the process. Figure 3b also shows that about 92-94% mineralization was reached for all PEF treatments at 18 Ah L-1, indicating that about 6-8% of the persistent intermediates cannot be destroyed by this method. Results of Figure 3a and b allow one to establish that pH values of 2.0-3.0 are optimal for the treatment of atrazine solutions by EAOPs with a BDD anode. 3.3. Kinetics for Atrazine Decay. The role of electrogenerated oxidants such as BDD(•OH) and •OH in the destruction of atrazine was analyzed by reversed-phase HPLC, where it displayed a well-defined absorption peak at tr ) 7.4 min. Blank experiments performed at pH 3.0 with and without 20 mM H2O2 under UVA irradiation did not show any significant removal of the herbicide. This indicates that it is not directly photolyzed and can only be attacked by the different hydroxyl radicals generated in the EAOPs. Figure 4a shows the abatement of atrazine concentration at the optimum pH 3.0 and 100 mA for all methods tested. The herbicide is removed at a similar rate for AO-SS and AO-H2O2, to disappear from solution in 100 min, as expected if it is only attacked by BDD(•OH) formed from reaction 1 under these conditions. A much faster and quite analogous disappearance of atrazine takes place for EF and PEF, with overall removal in only 40 min. The enhancement of atrazine destruction in the two latter EAOPs can be accounted for by its parallel reaction with •OH produced from Fenton’s reaction 4. The similar rate for atrazine decay in EF and PEF is indicative of little participation of photocatalytic reaction 5 in •OH generation and corroborates that it is not directly photolyzed by UVA light, as pointed out above from the blank experiments. These findings evidence that •OH is able to rapidly oxidize heteroaromatics, although this radical is so inefficient to destroy some persistent byproducts that it plays a very poor role in the total mineralization process (see Figure 1). The above concentration decays were fitted to kinetic equations with different simple reaction orders. Excellent linear correlations were only obtained considering a pseudo-first-order reaction, as can be seen in the inset panel of Figure 4a. From this analysis, a pseudo-first-order rate constant (k1) of 6.4 × 10-4 s-1 (square regression coefficient (R2) ) 0.982) for AOSS and 6.1 × 10-4 s-1 (R2 ) 0.980) for AO-H2O2 was found, suggesting a constant production of BDD(•OH) from reaction 1. The apparent rate constant increases ∼3.6 times for EF, k1 ) 2.3 × 10-3 s-1 (R2 ) 0.992), and PEF, k1 ) 2.2 × 10-3 s-1
6618
J. Phys. Chem. A, Vol. 114, No. 24, 2010
Borra`s et al.
Figure 4. Decay of atrazine concentration with electrolysis time during the treatment of 100 mL of a 30 mg L-1 herbicide solution in 0.05 M Na2SO4 of pH 3.0 at 35 °C. In plot (a), (]) AO-SS, (O) AO-H2O2, (b) EF, and (3) PEF at 100 mA. In plot (b), AO-SS at (]) 100, (0) 300, and (4) 450 mA. Each inset panel presents the corresponding kinetic analysis, assuming that atrazine disappears following a pseudo-firstorder reaction.
(R2 ) 0.996), indicating that atrazine reacts more quickly with • OH, which is also generated at a constant rate from Fenton’s reaction 4. The influence of applied current on atrazine decay for the AO-SS method was also examined. Results of Figure 4b show that the herbicide abatement is strongly accelerated as the current rises, disappearing in shorter times of 100, 25, and 15 min at increasing 100, 300, and 450 mA. The inset panel of Figure 4b depicts the good straight lines obtained considering a pseudofirst-order reaction, leading to a k1 value of 6.4 × 10-4 s-1 (R2 ) 0.982) at 100 mA, 2.8 × 10-3 s-1 (R2 ) 0.980) at 300 mA, and 4.9 × 10-3 s-1 (R2 ) 0.982) at 450 mA. Compared with 100 mA, the rate constant rises 4.4 and 7.6 times, while the current only increases 3 and 4.5 times for 300 and 450 mA, respectively. This larger enhancement in k1 cannot be related to the concomitant production of more oxidant BDD(•OH), which is expected to increase to a lesser extent than the ratio of currents, as pointed out above. This suggests the existence of parallel electrochemical reduction of atrazine at the SS cathode that takes place to a larger extent as the current rises from 100 mA, in agreement with the faster TOC removal found for AOSS than that for AO-H2O2 at 300 mA (see Figure 1). 3.4. Evolution of Aromatic Intermediates. Reversed-phase chromatograms of electrolyzed solutions displayed a few additional absorption peaks than that of atrazine. They were analyzed to attempt to detect possible aromatic byproducts coming from dealkylation, deamination, and/or hydroxylation reactions under the action of hydroxyl radicals such as desethylatrazine, desisopropylatrazine, desethyldesisopropylatrazine, desethyldesisopropyl-2-hydroxyatrazine, and cyanuric acid.22,33 Among these derivatives, only the formation of desethylatrazine (tr ) 3.1 min), desethyldesisopropylatrazine (tr ) 2.1 min), and cyanuric acid (tr ) 1,8 min) was confirmed by this technique
Figure 5. Evolution of the main aromatic intermediates detected during the degradation of 100 mL of a 30 mg L-1 atrazine solution in 0.05 M Na2SO4 at pH 3.0 and 35 °C by means of (a) AO-SS, (b) EF, and (c) PEF. Compounds: ([) desethylatrazine at 100 mA and cyanuric acid at (]) 100, (0) 300, and (4) 450 mA.
from comparison of their retention times and UV-vis spectra, measured on the diode array detector, with those of pure compounds. The concentrations of these byproducts were then determined via calibration with external standards. Figure 5 illustrates the evolution of cyanuric acid for AOSS, EF, and PEF. This compound is the last and most persistent aromatic byproduct detected in the degradation of atrazine by AOPs because it cannot be directly destroyed by •OH.42-51 Polcaro et al.9 showed that cyanuric acid can be partially removed in acid medium by AO-SS operating at 50 mA cm-2. This behavior can also be observed in Figure 5a for this technique at 100 mA (current density, 33 mA cm-2), where this acid attains a maximum of 0.8 mg L-1 between 240 and 360 min, dropping to 0.4 mg L-1 at the end of electrolysis. In contrast, it is completely removed in 180-300 min by this EAOP operating at higher currents of 300 (current density, 100 mA cm-2) and 450 mA (current density, 150 mA cm-2), after reaching maximum concentrations of 2.0 and 3.9 mg L-1, respectively, at 60 min. The larger accumulation of cyanuric acid as the current rises can be related to the faster destruction of atrazine (see Figure 4b) and the precedent byproducts with the higher amounts of BDD(•OH) produced from reaction 1. When such species disappear at 300 and 450 mA, cyanuric acid can be completely destroyed by BDD(•OH). Its presence at the
Degradation of Atrazine by EAOPs
J. Phys. Chem. A, Vol. 114, No. 24, 2010 6619
Figure 6. Proposed reaction sequence for the degradation of atrazine to cyanuric acid by EAOPs with a BDD anode.
end of electrolysis at 100 mA is then indicative of a slower removal of byproducts that are not completely converted into cyanuric acid. This fact is corroborated from the EF and PEF treatments at 100 mA. Thus, Figure 5b depicts a maximum of 4.6 mg L-1 of cyanuric acid at 90 min of EF, whereupon it decays to disappear in 540 min. This behavior can be accounted for by the parallel quicker destruction of atrazine (see Figure 4a) and the precedent byproducts with •OH formed from Fenton’s reaction 4, accelerating the accumulation of cyanuric acid that reacts slowly with BDD(•OH). This effect is less significant in PEF, where the rapid photolysis of some byproducts with UVA light (probably Fe(III) complexes) leads to a smaller accumulation of cyanuric acid, only 0.4 mg L-1 as the maximum at 180 min, disappearing in 540 min, that is, the same time as that in EF. The different influence of oxidants and UVA light on the degradation of atrazine can also be deduced from the evolution of desethylatrazine working at 100 mA. This dealkylated derivative was detected in very small contents (
