Chemosphere 97 (2014) 26–33

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Coupling of solar photoelectro-Fenton with a BDD anode and solarheterogeneous photocatalysis for the mineralization of the herbicideatrazine

0045-6535/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.chemosphere.2013.10.044

⇑ Corresponding author. Tel.: +52 8183294000x3432.E-mail address: edgar.ruizrz@uanl.edu.mx (E.J. Ruiz-Ruiz).URL: http://www.uanl.mx (E.J. Ruiz-Ruiz).

Benjamín R. Garza-Campos a, Jorge Luis Guzmán-Mar a, Laura Hinojosa Reyes a, Enric Brillas b,Aracely Hernández-Ramírez a, Edgar J. Ruiz-Ruiz a,⇑a Universidad Autónoma de Nuevo León, Facultad de Ciencias Químicas, Av. Pedro de Alba, Ciudad Universitaria, San Nicolás de los Garza, Nuevo León, Mexicob Laboratori d’Electroquímica dels Materials i del Medi Ambient, Departament de Química Física, Facultat de Química, Universitat de Barcelona, Martí i Franqués 1-11,08028 Barcelona, Spain

h i g h l i g h t s

� Uniform and adherent TiO2 films prepared onto glass spheres by sol–gel dip-coating.� Very low removal of atrazine by solar photocatalysis with TiO2 catalyst.� Atrazine partly mineralized by solar photoelectro-Fenton using a BDD/BDD cell.� Faster and greater mineralization by coupled solar photoelectro-Fenton-solar photocatalysis.� Atrazine mineralization inhibited by cyanuric acid formation with maximum 65% conversion.

a r t i c l e i n f o

Article history:Received 13 August 2013Received in revised form 15 October 2013Accepted 16 October 2013Available online 11 November 2013

Keywords:Anodic oxidationAtrazineSolar heterogeneous photocatalysisSolar photoelectro-Fenton

a b s t r a c t

Here, the synergetic effect of coupling solar photoelectro-Fenton (SPEF) and solar heterogeneousphotocatalysis (SPC) on the mineralization of 200 mL of a 20 mg L�1 atrazine solution, prepared fromthe commercial herbicide Gesaprim, at pH 3.0 was studied. Uniform, homogeneous and adherent ana-tase-TiO2 films onto glass spheres of 5 mm diameter were prepared by the sol–gel dip-coating methodand used as catalyst for SPC. However, this procedure yielded a poor removal of the substrate becauseof the low oxidation ability of positive holes and �OH formed at the catalyst surface to destroy it. Atrazinedecay was improved using anodic oxidation (AO), electro-Fenton (EF), SPEF and coupled SPEF-SPC at100 mA. The electrolytic cell contained a boron-doped diamond (BDD) anode and H2O2 was generatedat a BDD cathode fed with an air flow. The removal and mineralization of atrazine increased when moreoxidizing agents were generated in the sequence AO < EF < SPEF < coupled SPEF-SPC. Organics weredestroyed by �OH formed from water oxidation at the BDD anode in AO, along with �OH formed from Fen-ton’s reaction between added Fe2+ and generated H2O2 in EF. In SPEF, solar radiation produced higheramounts of �OH induced from the photolysis of Fe(III) species and photodecomposed intermediates likeFe(III)-carboxylate complexes. The synergistic action of sunlight in the most potent coupled SPEF-SPCwas ascribed to the additional quick removal of several intermediates with the oxidizing agents formedat the TiO2 surface. After 300 min of this treatment, 80% mineralization, 9% mineralization current effi-ciency and 1.93 kW h g�1 TOC energy cost were obtained. The mineralization of atrazine was inhibitedby the production of cyanuric acid, which was the main byproduct detected at the end of the coupledSPEF-SPC process.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Advanced oxidation processes (AOPs) are environmentallyfriendly chemical, photochemical, photocatalytic, electrochemical

and photoelectrochemical technologies based on the productionof �OH to remove toxic and/or biorefractory organic pollutants fromwastewaters (Malato et al., 2009; Martínez-Huitle and Brillas,2009; Ahmed et al., 2011). The high standard reduction potentialof �OH (E�(�OH/H2O) = 2.8 V vs SHE) confers to this radical a greatoxidizing power for the non-selective destruction of most organicsgiving dehydrogenated or hydroxylated derivatives, which can be

B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33 27

in turn totally mineralized to CO2, water and inorganic ions.Among these methods, electrochemical AOPs (EAOPs) havereceived great attention by their environmental compatibility, ver-satility, high energy efficiency, amenability of automation andsafety because they operate at mild conditions (Brillas et al.,2009; Panizza and Cerisola, 2009).

The simplest EAOP is anodic oxidation (AO), where organics areoxidized by hydroxyl radicals generated from water discharge atthe surface of a high O2-overpotential anode including Pt, PbO2,mixed metal oxides such as RuO2, TiO2, IrO2 and SnO2, and more re-cently boron-doped diamond (BDD) thin-film electrodes (Panizzaand Cerisola, 2009; Zhou et al., 2011; Ramírez et al., 2013). BDDis considered the best anode for AO due to their wide electrochem-ical window, low adsorption ability, large chemical stability andhigher O2-overpotential than other electrodes (Panizza and Ceriso-la, 2009). Thanks to these properties, organics can be incineratedby the great amounts of weakly adsorbed hydroxyl radicals (de-noted as BDD(�OH)) produced from water oxidation at high currentas follows (Marselli et al., 2003):

BDDþH2O! BDDð�OHÞ þHþ þ e� ð1Þ

EAOPs based on Fenton’s reaction chemistry such as electro-Fenton (EF) and photoelectro-Fenton (PEF) are being developedto treat acidic wastewaters (Brillas et al., 2009). They are character-ized by the continuous supply of H2O2 to a contaminated solutionby two-electron reduction of injected O2 by the following reaction:

O2 þ 2Hþ þ 2e� ! H2O2 ð2Þ

Good efficiencies for H2O2 generation from reaction (2) havebeen shown for carbonaceous cathodes such as carbon nano-tubes-polytetrafluoroethylene (PTFE) (Iranifam et al., 2011;Khataee et al., 2012), carbon-felt (Sirés et al., 2007; Özcan et al.,2008; Dirany et al., 2012), carbon sponge (Özcan et al., 2008),graphite-felt (Panizza and Oturan, 2011), carbon-PTFE gas (O2 orair) diffusion electrodes (Flox et al., 2007; Isarain-Chávez et al.,2010; Ruiz et al., 2011a,b) and BDD electrodes (Cruz-Gonzálezet al., 2010; Ramírez et al., 2013).

In EF, the oxidation power of H2O2 is enhanced by adding smallamounts of Fe2+ to produce Fe3+ and �OH in the bulk from Fenton’sreaction (3), with an optimum pH of 2.8 (Brillas et al., 2009). Theadvantages of this EAOP compared with chemical Fenton’s reagentis the in situ H2O2 generation and the regeneration of Fe2+ catalystat the cathode from Fe3+ reduction by reaction (4) (Sirés et al.,2007):

Fe2þ þH2O2 þHþ ! Fe3þ þH2Oþ �OH ð3Þ

Fe3þ þ e� ! Fe2þ ð4Þ

When an undivided cell with a BDD anode is used, organics areoxidized by both, BDD(�OH) formed from reaction (1) and �OH pro-duced in the bulk from Fenton’s reaction (3) (Brillas et al., 2009;Ruiz et al., 2011a).

The PEF process involves the UV irradiation of the contaminatedsolution under EF treatment. The main problem of this EAOP is theexcessive economical cost of the artificial UV lamps used and forthis reason, in our laboratories we have tested the alternative useof sunlight (k > 300 nm) as a free and renewable energy source inthe so-called solar PEF (SPEF) process (Flox et al., 2007; Almeidaet al., 2011; Ruiz et al., 2011a,b; Salazar et al., 2012). The UV lightfavors the degradation of pollutants because it promotes the fasterFe2+ regeneration and �OH production from Fe(OH)2+ photoreduc-tion, which is the predominant Fe3+ species at pH near 3, by

reaction (5) and the photolysis of Fe(III) complexes with generatedcarboxylic acids from reaction (6) (Garcia-Segura and Brillas,2011):

FeðOHÞ2þ þ hm! Fe2þ þ �OH ð5Þ

FeðOOCRÞ2þ þ hm! Fe2þ þ CO2 þ R� ð6Þ

On the other hand, heterogeneous photocatalysis (PC) is consid-ered one of the most promising AOPs for the destruction of lowamounts of water-soluble organic pollutants (Gaya and Abdullah,2008; Malato et al., 2009; Pelaez et al., 2012). Although differentsemiconductors have been tested as active materials for PC, TiO2

nanoparticles, mainly crystallized in the anatase form, yield thebest performance (Miranda-García et al., 2010). When TiO2 is illu-minated with UV photons of k < 380 nm, an electron from the filledvalence band is promoted to the empty conduction band ðe�cbÞ withan energy gap of 3.2 eV generating a positively charged vacancy orhole ðhþvbÞ by reaction (7). The holes thus produced at the TiO2 sur-face can oxidize either water or OH� giving �OH from reactions (8)and (9), respectively, which can subsequently attack the organicspecies (Pelaez et al., 2012).

TiO2 þ hm! e�cb þ hþvb ð7Þ

hþvb þH2O! �OHþHþ ð8Þ

hþvb þ OH� ! �OH ð9Þ

The major loss of efficiency in PC is due to the recombination ofelectrons promoted to the conduction band either with unreactedholes by reaction (10) or with adsorbed �OH by reaction (11):

e�cb þ hþvb ! TiO2 þ heat ð10Þ

e�cb þ �OH! OH� ð11Þ

When sunlight is used as energy source, the procedure is so-called solar PC (SPC) (Klamerth et al., 2009; Malato et al., 2009).In PC and SPC, the TiO2 catalyst is commonly used as suspendedpowder and has to be separated from the solution at the end ofthe process, which represents a serious limitation for large scaleapplications. This disadvantage can be solved by immobilizingthe TiO2 catalyst onto a substrate like activated carbon, glass tube,fiberglass and silica, primordially by sol–gel dip-coating, sputteringand electrophoretic deposition. The sol–gel dip-coating methodhas been found an appropriate technology for the synthesis oftransparent porous TiO2 films on borosilicate glass, because thecatalyst is well adhered at the support which has a suitable trans-parency (�300–400 nm) for the TiO2 activation range (Miranda-García et al., 2010). Although TiO2 films generally exhibit asignificant reduction of its photocatalytic efficiency, several papershave reported that TiO2 coatings can show more efficiency for or-ganic removal than uncoated catalyst (Gaya and Abdullah, 2008).

The coupling of different AOPs has been proposed for improvingthe oxidation ability of individual treatments. In this way, diverseemerging pollutants were mineralized by coupling of SPC andphoto-Fenton (Klamerth et al., 2009). Also, several papers de-scribed the combined use of PEF and PC with TiO2 (Khataee et al.,2012) or ZnO (Iranifam et al., 2011) supported on glass substratesunder artificial UV radiation to degrade dyes. As far as we know,the coupling of SPEF with SPC has not been previously reportedin the literature. For this reason, we have undertaken a study toclarify the characteristics of this novel coupled process forwastewater treatment. The reactions involved in this EAOP are

28 B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33

schematized in Fig. 1a. The mineralization of a very recalcitrantpollutant such as atrazine (2-chloro-4-ethylamino-6-isopropyla-mino-1,3,5-triazine) was tested. Atrazine has a solubility of34 mg L�1 at pH 7 and 20 �C and has been the most commonpre- and post-emergency used s-triazine herbicide in the worldfor combating weeds in sorghum, corn, rangeland, sugar and grassyand broadleaf, largely being accumulated in soils, sediments andwaters (Abate et al., 2004). In some Latin-American countries, atra-zine is the most widely used herbicide in many commercial formu-lations (Malpass et al., 2012), despite it belongs to the priority listof 76 substances of the Water Framework of the Directive 2000/60/EC of the European Parliament and it has been banned in severalcountries (Chan and Chu, 2003). Atrazine is a potential dangerfor public health since it is an endocrine disruptor (Lawton et al.,2006). This compound has been removed by photodegradation(Sun et al., 2011), AOPs like UV/H2O2 (Beltrán et al., 1993) and Fen-ton (Chan and Chu, 2003) and EAOPs such as AO with Pt, metal oxi-des and BDD (Malpass et al., 2012) and EF and PEF with BDD (Balciet al., 2009; Borràs et al., 2010; Oturan et al., 2012). These worksshowed a hard mineralization of atrazine due to the formation ofthe very stable product cyanuric acid (2,4,6-trihydroxy-1,3,5-tri-azine), which can only be slowly destroyed at a BDD anode byapplying a high current.

(a)

(b)

Fig. 1. (a) Schematic representation of the processes taking place in the coupledsolar photoelectro-Fenton (SPEF)–solar heterogenous photocatalysis (SPC) processused for atrazine degradation. (b) Sketch of the experimental setup of the coupledsystem.

This paper presents a study on the comparative mineralizationof 20 mg L�1 atrazine at pH 3.0 by SPC, AO, EF, SPEF and coupledSPEF-SPC processes. In SPC, TiO2 nanoparticles synthetized bysol–gel and supported on glass spheres by dip-coating were usedas catalyst. In the EAOPs, a BDD electrode was used as anode andH2O2 was generated at a BDD cathode. The solution was directlyexposed to sunlight as source of UV radiation. The atrazine decayand the evolution of cyanuric acid and generated carboxylic acidswere followed by high-performance liquid chromatography(HPLC).

2. Experimental

2.1. Chemicals

The commercial formulation Gesaprim containing 90% atrazinewas purchased from Syngent. Cyanuric acid (98% purity), sodiumdihydrogenphosphate (99% purity), disodium hydrogenphosphate(99% purity), HPLC grade acetonitrile and titanium oxysulfate(99.99% purity) were supplied by Sigma–Aldrich. Standards of atra-zine, and formic, oxalic and oxamic acids were of analytical gradepurchased from Fermont. Anhydrous sodium sulfate was of analyt-ical grade supplied by LeMont. Sulfuric acid was of analytical gradepurchased from Baker. Ferrous sulfate heptahydrate (99% purity)was purchased from Jalmek. Titanium butoxide Ti (97% purity)and sec-butyl alcohol (99.5% purity) were supplied by Fluka andTedia respectively. All the solutions were prepared with doublydistilled deionized water.

2.2. Synthesis of TiO2 and deposition on glass spheres

TiO2 was synthesized by the sol–gel method using titaniumbutoxide as precursor and sec-butyl alcohol as a solvent (Ramos-Delgado et al., 2013) Under continuous stirring, 14.7 mL of precur-sor was added to 45 mL of alcohol and the pH was adjusted at 3 byadding glacial acetic acid drop wise. A solution of 3 mL of water in40 mL of alcohol was separately prepared and added drop wise tothe above solution. The resulting solution was stirred for 24 h giv-ing a sol, which was used to prepare TiO2 coatings on borosilicateglass spheres of 5 mm diameter purchased from Marienfeld bythe dip-coating technique, followed by calcination at 500 �C during1 h using a Thermoline 47900 furnace with a heating rate of10 �C min�1. Deposits were made layer by layer every 24 h untilobtain 10 layers of coated catalyst. The morphology and composi-tion of TiO2 coatings were analyzed by scanning electron micros-copy (SEM) using a FEI-Nova nanosem 200 system and X-raydiffraction (XRD) using a Siemens D500 system type Bragg–Brent-ano h/2h by applying a Cu Ka1+2 radiation (k(a1) = 0.154060 nmand k(a2) = 0.154443 nm) at 40 kV and 30 mA current.

2.3. Photocatalytic and electrolytic systems

A solution with 20 mg L�1 atrazine was prepared by dissolving22.5 mg of commercial Gesaprim in 1 L of ultrapure water for 3 hunder vigorous stirring. All degradation experiments were con-ducted in an open, undivided and cylindrical tank reactor contain-ing 200 mL of the above atrazine solution in 0.05 M Na2SO4 at pH3.0 adjusted by addition of concentrated H2SO4 and at room tem-perature (about 25 �C) under vigorous stirring with a magneticbar at 400 rpm. The SPC and SPEF-SPC assays were made by placing180 spheres of TiO2-coated glass (corresponding to about 0.3 g L�1

of TiO2) sustained by a nylon mesh near the bottom of the cell, asshown in the experimental setup of Fig. 1b. These trials were runafter keeping the spheres in the solution for 1 h in the darknessto achieve the absorption equilibrium of atrazine onto the TiO2

B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33 29

coating. After each treatment, the solution was filtered to recoverthe spheres, which were washed with distilled water and reusedconsecutively at least 5 times without significant changes in theirdegradation power. For all the electrolytic experiments, the anodewas a BDD thin-film electrode on a conducting p-Si substrate sup-plied by Adamant Technologies. For AO, the cathode was a Pt mesh,whereas for EF, SPEF and the coupled SPEF-SPC, it was a BDD thin-film electrode on Nb purchased from Metakem. The geometric areaof all electrodes in contact with the solution was 7.5 cm2 (with ageometric area of 3 cm � 2.5 cm � 0.1 cm) and the interelectrodegap was about 0.5 cm. In EF, SPEF and SPEF-SPC, 0.1 mM Fe2+

was added as the Fenton catalyst. The applied current was always100 mA provided by a MPL-1303 power source. Solar experimentswere made under an average irradiation of 1150 W m�2 in Monter-rey, N.L., Mexico, related to an UV average irradiation of 34 W m�2.H2O2 was produced from the cathodic reduction of O2 dissolved inthe solution by bubbling compressed air at 1 L min�1, starting30 min before electrolysis until the final of trials to kept O2 satura-tion. The concentration of H2O2 accumulated in the medium wasdetermined by electrolyzing a 0.05 M Na2SO4 solution of pH 3.0in a BDD/BDD cell.

Fig. 2. SEM images of the TiO2 deposits obtained on glass spheres by dip-coatingtechnique. (a) Morphology and particle size and (b) lateral view of the TiO2 coating.Magnification 240000�.

10 20 30 40 60 70 80 90

Inte

nsity

(a.u

)

(101

)

(103

)(0

04)

(200

)

(105

)(2

11)

2 θ / degrees

Fig. 3. X-ray diffraction patterns of TiO2 synthesized by sol–gel method followed by

2.4. Instruments and analytical procedures

The solution pH was measured with a Thermo Scientific Orion 3Star pH meter. H2O2 concentration was determined from the lightabsorption of its colored complex with titanium oxysulfate atk = 409 nm, measured on a Varian Cary-50 UV–Vis spectrophotom-eter (Welcher, 1975). The aliquots were filtered with 0.45 lmNylon filters purchased from Phenex before HPLC and TOC analysis.The dissolved organic carbon of atrazine solutions was determinedas TOC using a Shimadzu VCSH TOC analyzer. Reproducible TOCvalues with an accuracy of ±1% were found by injecting 50 lL ali-quots to the analyzer.

Atrazine decay was followed by reversed-phase HPLC using aPerkin Elmer 200 LC fitted with a C18 Phenomenex Hyper-CloneODS 5 lm, 250 mm � 4.6 mm (id), column, coupled with aPerkin–Elmer 200 UV–Vis detector selected at k = 223 nm. The evo-lution of cyanuric acid was followed by reversed-phase HPLC usinga Waters 2695 LC fitted with the same C18 column and coupledwith a Waters 2996 photodiode array detector set at k = 213 nm.Generated carboxylic acids were detected by ion-exclusion HPLCusing the above LC fitted with a Bio-Rad Aminex HPX 87H,300 mm � 7.8 mm (id) column and the photodiode array detectorselected at k = 210 nm. For atrazine analysis, 20 lL aliquots wereinjected into the chromatograph, whereas for cyanuric acid andcarboxylic acid analysis, 100 lL samples were injected. The mobilephase was a 45:55 (v/v) acetonitrile/water mixture at 1 mL min�1

for atrazine, 95:5 (v/v) phosphate buffer/methanol mixture at0.8 mL min�1 for cyanuric acid and 4 mM H2SO4 at 0.8 mL min�1

for short-linear carboxylic acids. The phosphate buffer was com-posed of 0.00625 M NaH2PO4 and 0.0125 M Na2HPO4.

calcination at 500 �C for 1 h.

3. Results and discussion

3.1. SEM and XRD characterization of TiO2 coating onto glass spheres

The excellent deposit obtained for the synthesized TiO2 ontoglass spheres is depicted in the SEM image of Fig. 2a. The coatingshowed a homogeneous morphology and was composed of roughlyspherical TiO2 nanoparticles of about 40 nm average diameter,which is a typical particle size (nm) for this material when synthe-sized by the sol–gel method. Fig. 2b illustrates a SEM image of a cutof a glass sphere showing a lateral view of a deposited TiO2 film. Auniform and homogenous coating with an average thickness close

to 400–450 nm can be observed, as expected for the consecutivedeposition of 10 layers of TiO2 nanoparticles with an average sizeof about 40 nm. The coating was well adhered and it was not re-leased from the glass surface during the SPC and coupled SPEF-SPC treatments performed to the atrazine solution.

Fig. 3 shows the XRD pattern of a sample of TiO2 powder ob-tained after 1 h of heat treatment at 500 �C. The main diffractionpeaks related to the 2h angles of 25.5�, 37.3�, 38.1�, 48.2�, 54.2�and 55.2� can be indexed to the (101), (103), (004), (200),(105) and (211) crystallographic phases of anatase TiO2 (Suet al., 2004; Liu et al., 2007). All these peaks were very sharp in

7

0

20

40

60

80

0 60 120 180 240 300 360

100

(C /

C0) x

100

(a)

30 B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33

agreement with the formation of nanoparticles. Note that if theTiO2 powder was heated to temperatures above 500 �C, a mixtureof its anatase and rutile forms was obtained, since the anatasephase is a metastable polymorphic form of TiO2 with respect tothe rutile one (Liu et al., 2007). The XRD results of Fig. 3 confirmthat the prepared TiO2 coating onto glass spheres is predominantlycomposed of the anatase phase, which has much higher photocat-alytic activity than the rutile one (Liu et al., 2007), and hence it isappropriate for the SPC and coupled SPEF-SPC assays made in thiswork and reported below.

3.2. H2O2 accumulation in the BDD/BDD tank reactor

The ability of the undivided BDD/BDD tank reactor to produceH2O2 from reaction (2) and accumulate it was tested by electrolyz-ing 200 mL of a 0.05 M Na2SO4 solution of pH 3.0 at constant cur-rent of 50 and 100 mA and 25 �C for 120 min under air bubbling at1 L min�1. As can be seen in Fig. 4, the concentration of accumu-lated H2O2 increased gradually up to reach a quasi-steady valueclose to 11.5 and 20 mg L�1 after 50–60 min of electrolysis at 50and 100 mA, respectively. The rise in H2O2 concentration at highercurrent can be simply related to the concomitant increase in rate ofreaction (2). The plateau attained in both cases is indicative of theparallel destruction of electrogenerated H2O2, which became fasteras the concentration of this species rose, so that, the quasi-steadycontent in each current was reached just when its electrogenera-tion and destruction rate are equal. The quick removal of H2O2 inthe electrolytic system can be accounted for by its decompositionto O2 at the BDD anode surface via formation of the hydroperoxylradical ðHO�2Þ by reactions (12) and (13) (Flox et al., 2007; Brillaset al., 2009):

BDDþH2O2 ! BDDðHO�2Þ þHþ þ e� ð12Þ

BDDðHO�2Þ ! BDDþ O2 þHþ þ e� ð13Þ

The fact that the quasi-steady H2O2 concentration is roughlydirectly proportional to the applied current agrees with the fara-daic behavior expected for reactions (2), (12), and (13). Note thata small quantity of H2O2 can also be formed from the dimerizationof BDD(�OH) generated at the BDD anode from reaction (1) as fol-lows (Skoumal et al., 2008):

2BDDð�OHÞ ! 2BDDþH2O2 ð14Þ

The above findings evidence that the undivided BDD/BDD tankreactor is able to continuously produce H2O2 that can be utilized to

0

5

10

15

20

25

0 20 40 60 80 100 120 140

[H2O

2] / m

g L-1

time / min

Fig. 4. Concentration of hydrogen peroxide accumulated in 200 mL of a 0.05 MNa2SO4 solution at pH 3.0 using a BDD/BDD stirred tank reactor by applying: ( )50 mA and ( ) 100 mA.

generate oxidant �OH in the bulk from Fenton’s reaction (3) in theEF, SPEF and coupled SPEF-SPC processes of atrazine, as discussedbelow.

3.3. Atrazine removal by SPC and EAOPs

The reaction of atrazine with the oxidants generated in SPC andthe EAOPs methods tested was followed by reversed-phase HPLC,where it displayed a well-defined peak at retention time (tr) of8.0 min. Previous tests of a 20 mg L�1 herbicide solution in0.05 M Na2SO4 of pH 3.0 in the darkness, covering the catalystspheres with a black plastic, demonstrated an insignificant adsorp-tion of atrazine onto the TiO2 coating. Besides, direct exposition ofthe above solution to sunlight did not cause any atrazine decay,indicating that it was not photolyzed under the experimental con-ditions checked.

Fig. 5a shows the comparative decay in atrazine concentrationfor 200 mL of a 20 mg L�1 herbicide solution in 0.05 M Na2SO4 ofpH 3.0 during 300 min of the different processes. For all the EAOPs,a current of 100 mA was applied, whereas in EF, SPEF and coupledSPEF-SPC, 0.1 mM Fe2+ was added to the treated solution since thisconcentration of metal catalyst was taken as optimal for the anal-ogous EF treatment using a carbon-felt cathode (Oturan et al.,2012). In all these trials, the solution pH remained practically un-changed with electrolysis time. As can be seen in Fig. 5a, the useof SPC led to a very slow decay of atrazine up to a final reductiononly of 30%, as expected if low quantities of reactive holes and�OH are formed from reactions (7) and (8), probably due to theirquick inactivation by reactions (10) and (11). In contrast, all theEAOPs yielded a much faster removal of the herbicide up to its totaldisappearance, with an increasing oxidation ability in the sequenceAO < EF < SPEF < coupled SPEF-SPC. The slower disappearance of

0

1

2

3

4

5

6

ln (C

0 / C

)

time / min

(b)

0 30 60 90 120 150 180 210

Fig. 5. (a) Normalized atrazine concentration decay for 200 mL of a 20 mg L�1

substrate solution in 0.05 M Na2SO4 at pH 3.0 and (b) kinetic analysis assuming apseudo-first-order reaction for the herbicide. Method: ( ) SPC, ( ) anodicoxidation (AO), (N) electro-Fenton (EF), ( ) SPEF and ( ) coupled SPEF-SPCprocess. In the three latter processes, 0.1 mM Fe2+ was added to the solution.Current applied in the electrochemical treatments: 100 mA.

B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33 31

atrazine by AO can be ascribed to the only action of BDD(�OH) pro-duced by reaction (1). The larger abatement found in EF can be re-lated to the additional generation of �OH from Fenton’s reaction (3),which is enhanced in SPEF due to the higher production of this rad-ical induced by the photolytic reaction (5) because of the formationof Fe3+ from reaction (3). The more potent process is the coupledSPEF-SPC, where atrazine can be simultaneously attacked by alloxidants, i.e., BDD(�OH), �OH in the bulk and holes and �OH at theTiO2 surface. That means that the immobilized catalyst onto theglass spheres is effectively active under the UV radiation of sun-light contributing to the oxidation of organics in the coupledprocess.

The above concentration decays for the EAOPs were analyzed bymeans of kinetic equations related to several simple order reac-tions. Fig. 5b highlights that they were well fitted to a pseudo-first-order reaction. As expected, the apparent rate constant (k1)thus obtained increased with increasing the oxidation ability ofthe EAOP, being of 1.2 � 10�2 s�1 (R2 = 0.996) for AO,1.4 � 10�2 s�1 (R2 = 0.981) for EF, 2.8 � 10�2 s�1 (R2 = 0.995) forSPEF and 3.4 � 10�2 s�1 (R2 = 0.987) for the coupled SPEF-SPC. Thisbehavior suggests that atrazine is attacked with a constant amountof each oxidant in each process because they are continuously pro-duced under steady conditions.

3.4. Evolution of cyanuric acid and generated carboxylic acids

It is well-known that the oxidation of atrazine involves dealky-lation, deamination and/or hydroxylation reactions generating aro-matic products like desethylatrazine, desisopropylatrazine,desethyldesisopropylatrazine, desethyldesisopropy-l,2-hydrox-yatrazine and cyanuric acid (Borràs et al., 2010; Malpass et al.,2012). The latter compound is the most persistent aromaticbyproduct using AOPs because it cannot be directly destroyed by�OH, although it can be slowly removed when large quantities ofreactive BDD(�OH) are formed at high currents (Oturan et al.,2012). Taking this into mind, the production of cyanuric acid inthe SPC and EAOPs tested was followed by reversed-phase HPLC.These chromatograms exhibited a well defined peak for this com-pound at tr = 4.1 min.

Fig. 6 illustrates that for the trials reported in Fig. 5, cyanuricacid was accumulated in larger extent as the oxidation ability ofthe process rose. Thus, after 300 min of treatment, 1.1, 3.0, 5.1,7.1 and 7.8 mg L�1 of this product were found for SPC, AO, EF, SPEFand coupled SPEF-SPC, respectively. Taking into account that20 mg L�1 of atrazine can be completely oxidized to 12.0 mg L�1

of cyanuric acid, one can infer a total conversion of 9.2% (30% withrespect to atrazine removed), 25%, 42%, 59% and 65% in the aboveassays. This indicates that the increasing generation of moreoxidizing species favors the degradation of precedent aromatic

0

2

4

6

8

0 60 120 180 240 300 360

[Cya

nuric

aci

d] /

mg

L-1

time / min

Fig. 6. Evolution of cyanuric acid concentration during the treatments of Fig. 5.

intermediates to cyanuric acid. At the low applied current of100 mA, however, this ultimate aromatic cannot be removed byBDD(�OH).

Ion-exclusion chromatograms of the above electrolyzed solu-tions revealed the generation of short-linear carboxylic acids likeoxalic (tr = 5.0 min), oxamic (tr = 7.6 min) and formic (tr = 10.1 -min). While oxalic and oxamic acids can proceed from the oxida-tion of the lateral ethylamine and isopropylamine groups ofatrazine, formic acid can be formed from the degradation of all car-bon atoms. These acids were detected at very low concentrations(<1 mg L�1) under nonsolar conditions (EF) because they yieldedFe(III)-carboxylate complexes that are hardly attacked byBDD(�OH) and �OH (Brillas et al., 2009; Garcia-Segura and Brillas,2011). In contrast, these complexes were rapidly photolyzed byreaction (6) and disappeared completely at 60 min of SPEF. Whenthe coupled SPEF-SPC process was applied, oxamic acid was notdetected, probably because Fe(III)-oxamate complexes are not onlyquickly photolyzed, but also destroyed by the holes formed at theTiO2 surface in a synergistic process.

3.5. TOC removal, mineralization current efficiency and energy cost forthe EAOPs

All the treated solutions contained 20 mg L�1 of atrazine, asconfirmed by reversed-phase HPLC, which corresponded to8.9 mg L�1 TOC. However, they were prepared from the commer-cial herbicide Gesaprim with 90% atrazine, leading to a higher ini-tial TOC of 13.4 mg L�1 due to the organic load of unknownexcipients and stabilizers of such product. Thus, the percentageof residual TOC shown in Fig. 7a corresponds to the experimentaldata obtained for all the EAOPs from 13.4 mg L�1 TOC.

The SPF process alone did not yield a significant TOC reductionof the 20 mg L�1 atrazine solution due to the low mineralizationability of the oxidizing agents formed at the TiO2 surface under so-lar radiation, as reflected by the fact that about 6 mg L�1 of the her-bicide were removed in 300 min (see Fig. 5a) giving rise to1.1 mg L�1 cyanuric acid (30% conversion, see Fig. 6). In contrast,all the EAOPs led to a gradual TOC decay with partial mineraliza-tion of the atrazine solution, which rose as their oxidation abilityincreased because of the production of more oxidizing species. At300 min of electrolysis, for example, Fig. 7a shows a TOC reductionof 46% for AO, 50% for PEF, 70% for SPEF and 80% for SPEF-SPC. Fromthe data of Fig. 6, one can then determine that the final solutioncontained 11%, 21%, 48% and 82% TOC related to generated cyan-uric acid in the above assays. This means that the other productsare more rapidly destroyed when more oxidizing species are pro-duced in the EAOP, yielding species like the stable cyanuric acidand carboxylic acids that can be mineralized. For the most potentprocess, the coupled SPEF-SPC, cyanuric acid was the predominantbyproduct in the final treated solution, indicating that the methodcan remove the major part of all the other intermediates formed.This can be related to the synergistic action of sunlight that en-hances the oxidative characteristics of the EF process by meansof the photolytic reactions (5) and (6), as well as the generationof holes and �OH at the TiO2 coating from reactions (7) and (9),respectively. Our results indicate that the latter oxidizing speciesimprove the removal rate of low contents of some intermediatesrespect to the SPEF process, like oxamic acid, and their synergisticcontribution allows the quicker mineralization of all byproductsexcept cyanuric acid. However, a 20% of residual TOC was finallyobtained for the coupled SPEF-SPC containing cyanuric acid relatedto the conversion of 65% of the initial herbicide. The remainingsolution also contained the inorganic ions coming from the parallelmineralization of 35% of the 20 mg L�1 atrazine. According to pre-vious work (Borràs et al., 2010), all the organic Cl was transformedinto Cl� ion, which is subsequently oxidized to Cl2 at the BDD

0

20

40

60

80

100%

resi

dual

TO

C

0

5

10

15

% M

CE

0

1

2

3

4

0 60 120 180 240 300 360

ECTO

C /

kWh

g-1 T

OC

time / min

(a)

(b)

(c)

Fig. 7. (a) Percentage of residual TOC, (b) mineralization current efficiency and (c)energy cost per unit TOC mass for the: ( ) AO, (N) EF, ( ) SPEF and ( ) coupledSPEF-SPC processes given in Fig. 5.

32 B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33

anode, whereas the organic N from mineralized matter was pre-eminently released as NO�3 ion and in smaller proportion, as NHþ4ion.

From the above TOC data, the mineralization current efficiency(MCE) at a time t (h) of the EAOPs assays was estimated from Eq.(15)(Skoumal et al., 2008):

MCE ð%Þ ¼nFV sDðTOCÞexp

4:32� 107mIt100 ð15Þ

where F is the Faraday constant (96487 C mol�1), Vs is thesolution volume (L), D(TOC)exp is the experimental TOC decay(mg L�1), 4.32 � 107 is an homogenization factor (3600 s h�1 �12000 mg mol�1), m is the number of carbon atoms of atrazine (8carbon atoms) and I is the applied current (A). The number ofelectrons (n) consumed per atrazine molecule was taken as 70,considering that its mineralization to CO2 involves the release ofCl� and NO�3 as the main primary ions (Borràs et al., 2010) viareaction (16):

C8H14ClN5 þ 31H2O! 8CO2 þ Cl� þ 5NO�3 þ 76Hþ þ 70e� ð16Þ

Fig. 7b evidences that the maximum MCE values were found atthe beginning of each EAOP. As expected, the coupled SPEF-SPCtreatment gave the best efficiency, which dropped from 18% at60 min to ca. 9% at 300 min. This drastic fall in MCE can be relatedto the quick decay of products that can be mineralized, along with

the persistence of the ultimate cyanuric acid. The same trend canbe observed for SPEF since intermediates were largely mineralized.The lower mineralization rate in AO and EF, however, was reflectedin a much smaller efficiency with slower reduction. It should benote worthy that only 3% of MCE was reported as maximal forthe removal of 100 mL of 20 mg L�1 atrazine by PEF with a BDD an-ode at 100 mA (Borràs et al., 2010), a value lower than those ob-tained in the present work for the EAOPs tested.

Besides, to better analyze the viability of the more potent cou-pled SPEF-PC treatment, the energy cost per unit TOC mass (ECTOC)for the trials was calculated from Eq. (17)(Ruiz et al., 2011a).

ECTOC ðkW h g�1TOCÞ ¼ EcellItDðTOCÞexpV s

ð17Þ

where Ecell is the average potential difference of the cell (10.4, 9.2,8.7 and 8.6 V for AO, EF, SPEF and SPEF-SPC respectively). Fig. 7c re-veals that all the atrazine treatments by EAOPs required very highECTOC values because of is slow mineralization rate as a result ofthe large persistence of intermediates and accumulation of cyanuricacid. The deceleration of the mineralization process with loss inMCE as electrolysis time rose was also reflected by a gradual greaterECTOC value in all cases. For example, at 300 min of electrolysis,4.20 kW h g�1 TOC were consumed in AO, which dropped to3.59 kW h g�1 TOC (15% reduction) in EF. The synergistic action ofsolar radiation led to a lower cost of 2.31 kW h g�1 TOC (45% reduc-tion) in SPEF and even smaller, of 1.93 kW h g�1 TOC (54% reduc-tion) in the coupled SPEF-SPC. This confirms that the latterprocess is more viable from an economical point of view consider-ing that the TiO2 catalyst remained stable and is then reusable inconsecutive treatments.

4. Conclusions

Uniform, homogeneous and adherent TiO2 films onto glassspheres were prepared by the sol–gel dip-coating method. TheTiO2 coating was predominantly composed of its anatase form, asconfirmed by XRD, which is the most photoactive phase of this cat-alyst. The SPC treatment of 20 mg L�1 atrazine at pH 3.0 performedwith such TiO2 coating led to a poor removal of the substrate dueto the low oxidation ability of holes and �OH produced at the cata-lyst surface under solar radiation. A much quicker and completeatrazine decay was found using EAOPs with a BDD anode at100 mA, which was gradually enhanced when more oxidizingagents were generated in the sequence AO < EF < SPEF < coupledSPEF-SPC. The latter treatment with a BDD/BDD tank reactor wasalso the most potent degradation process of the 20 mg L�1 atrazinesolution giving 80% mineralization, 9% mineralization current effi-ciency and 1.93 kW h g�1 TOC energy cost after 300 min of elec-trolysis. Organics were attacked by BDD(�OH) in AO, along withby �OH formed from Fenton’s reaction (3) in EF. The synergistic ac-tion of sunlight in SPEF-SPC can be explained by the quicker oxida-tion of intermediates with the higher amounts of �OH induced fromthe photolytic reaction (5) and their additional destruction withthe oxidizing agents formed at the TiO2 surface. Besides, severalcompounds like Fe(III)-carboxylate complexes can be photolyzedby solar radiation, also enhancing the degradation process. Themineralization of atrazine was inhibited by the formation of theultimate and stable cyanuric acid, which was the major byproductpresent in the final solution treated by the coupled SPEF-SPC, cor-responding to a 65% conversion of the initial herbicide. The cou-pling of SPEF and SPC then represents the best viable EAOP forthe effective mineralization of atrazine and other unknown addi-tives contained in the commercial herbicide Gesaprim.

B.R. Garza-Campos et al. / Chemosphere 97 (2014) 26–33 33

Acknowledgments

The authors thank the financial support from the PAICYT pro-gram of the Universidad Autónoma de Nuevo León and Conacyt-Red (Project: 193883). B.R. Garza-Campos acknowledges thegranted scholarship awarded by CONACyT (Consejo Nacional deCiencia y Tecnología, Mexico).

References

Abate, G., Penteado, J.C., Cuzzi, J.D., Vitti, G.C., Lichtig, J., Masini, J.C., 2004. Influenceof humic acid on adsorption and desorption of atrazine, hydroxyatrazine,deethylatrazine, and deisopropylatrazine onto a clay-rich soil sample. J. Agric.Food Chem. 52, 6747–6754.

Ahmed, B., Limem, E., Abdel-Wahab, A., Nasr, B., 2011. Photo-Fenton treatment ofactual agro-industrial wastewaters. Ind. Eng. Chem. Res. 50, 6673–6680.

Almeida, L.C., Garcia-Segura, S., Bocchi, N., Brillas, E., 2011. Solar photoelectro-Fenton degradation of paracetamol using a flow plant with a Pt/air-diffusioncell coupled with a compound parabolic collector: process optimization byresponse surface methodology. Appl. Catal. B – Environ. 103, 21–30.

Balci, B., Oturan, N., Cherrier, R., Oturan, M.A., 2009. Degradation of atrazine inaqueous medium by electrocatalytically generated hydroxyl radicals. A kineticand mechanistic study. Water Res. 43, 1924–1934.

Beltrán, F.J., Ovejero, G., Acedo, B., 1993. Oxidation of atrazine in water byultraviolet radiation combined with hydrogen peroxide. Water Res. 27, 1013–1021.

Borràs, N., Oliver, R., Arias, C., Brillas, E., 2010. Degradation of atrazine byelectrochemical advanced oxidation processes using a boron-doped diamondanode. J. Phys. Chem. A 114, 6613–6621.

Brillas, E., Sirés, I., Oturan, M.A., 2009. Electro-Fenton process and relatedelectrochemical technologies based on Fenton’s reaction chemistry. Chem.Rev. 109, 6570–6631.

Chan, K.H., Chu, W., 2003. Modeling the reaction kinetics of Fenton’s process on theremoval of atrazine. Chemosphere 51, 305–311.

Cruz-González, K., Torres-López, O., García-León, A., Guzmán-Mar, J.L., Reyes, L.H.,Hernández-Ramírez, A., Peralta-Hernández, J.M., 2010. Determination ofoptimum operating parameters for Acid Yellow 36 decolorization by electro-Fenton process using BDD cathode. Chem. Eng. J. 160, 199–206.

Dirany, A., Sirés, I., Oturan, N., Özcan, A., Oturan, M.A., 2012. Electrochemicaltreatment of the antibiotic sulfachloropyridazine: kinetics, reaction pathways,and toxicity evolution. Environ. Sci. Technol. 46, 4074–4082.

Flox, C., Garrido, J.A., Rodríguez, R.M., Cabot, P.-L., Centellas, F., Arias, C., Brillas, E.,2007. Mineralization of herbicide mecoprop by photoelectro-Fenton with UVAand solar light. Catal. Today 129, 29–36.

Garcia-Segura, S., Brillas, E., 2011. Mineralization of the recalcitrant oxalic andoxamic acids by electrochemical advanced oxidation processes using a boron-doped diamond anode. Water Res. 45, 2975–2984.

Gaya, U.I., Abdullah, A.I., 2008. Heterogeneous photocatalytic degradation of organiccontaminants over titanium dioxide: a review of fundamentals, progress andproblems. J. Photochem. Photobiol. C 9, 1–12.

Iranifam, M., Zarei, M., Khataee, A.R., 2011. Decolorization of CI Basic Yellow 28solution using supported ZnO nanoparticles coupled with photoelectro-Fentonprocess. J. Electroanal. Chem. 659, 107–112.

Isarain-Chávez, E., Arias, C., Cabot, P.L., Centellas, F., Rodríguez, R.M., Garrido, J.A.,Brillas, E., 2010. Mineralization of the drug beta-blocker atenolol by electro-Fenton and photoelectro-Fenton using an air-diffusion cathode for H2O2

electrogeneration combined with a carbon-felt cathode for Fe2+ regeneration.Appl. Catal. B – Environ. 96, 361–369.

Khataee, A.R., Safarpour, M., Naseri, A., Zarei, M., 2012. Photoelectro-Fenton/nanophotocatalysis decolorization of three textile dyes mixture: responsesurface modeling and multivariate calibration procedure for simultaneousdetermination. J. Electroanal. Chem. 672, 53–62.

Klamerth, N., Miranda, N., Malato, S., Agüera, A., Fernández-Alba, A.R., Maldonado,M.I., Coronado, J.M., 2009. Degradation of emerging contaminants at lowconcentrations in MWTPs effluents with mild solar photo-Fenton and TiO2.Catal. Today 144, 124–130.

Lawton, J.C., Pennington, P.L., Chung, K.W., Scott, G.I., 2006. Toxicity of atrazine tothe juvenile hard clam, Mercenaria mercenaria. Ecotox. Environ. Saf. 65, 388–394.

Liu, Z., Zhang, X., Nishimoto, S., Jin, M., Tryk, D.A., Murakami, T., Fujishima, A., 2007.Highly ordered TiO2 nanotube arrays with controllable length forphotoelectrocatalytic degradation of phenol. J. Phys. Chem. C 112, 253–259.

Malato, S., Fernández-Ibáñez, P., Maldonado, M.I., Blanco, J., Gernjak, W., 2009.Decontamination and disinfection of water by solar photocatalysis: recentoverview and trends. Catal. Today 147, 1–59.

Malpass, G.R.P., Salazar-Banda, G.R., Miwa, D.W., Machado, S.A.S., Motheo, A.J., 2012.Comparing atrazine and cyanuric acid electro-oxidation on mixed oxide andboron-doped diamond electrodes. Environ. Technol., 1–9.

Marselli, B., Garcia-Gomez, J., Michaud, P.-A., Rodrigo, M.A., Comninellis, C., 2003.Electrogeneration of hydroxyl radicals on boron-doped diamond electrodes. J.Electrochem. Soc. 150, D79–D83.

Martínez-Huitle, C.A., Brillas, E., 2009. Decontamination of wastewaters containingsynthetic organic dyes by electrochemical methods. A general review. Appl.Catal. B – Environ. 87, 105–145.

Miranda-García, N., Maldonado, M.I., Coronado, J.M., Malato, S., 2010. Degradationstudy of 15 emerging contaminants at low concentration by immobilized TiO2

in a pilot plant. Catal. Today 151, 107–113.Oturan, N., Brillas, E., Oturan, M., 2012. Unprecedented total mineralization of

atrazine and cyanuric acid by anodic oxidation and electro-Fenton with aboron-doped diamond anode. Environ. Chem. Lett. 10, 165–170.

Özcan, A., Sahin, Y., Koparal, S.A., Oturan, M.A., 2008. Carbon sponge as a newmaterial for the electro-Fenton process. Comparison with carbon felt cathodeand application to degradation of synthetic dye Basic Blue 3 in aqueousmedium. J. Electroanal. Chem. 616, 71–78.

Panizza, M., Cerisola, G., 2009. Direct and mediated anodic oxidation of organicpollutants. Chem. Rev. 109, 6541–6569.

Panizza, M., Oturan, M.A., 2011. Degradation of Alizarin Red by electro-Fentonprocess using a graphite-felt cathode. Electrochim. Acta 56, 7084–7087.

Pelaez, M., Nolan, N.T., Pillai, S.C., Seery, M.K., Falaras, P., Kontos, A.G., Dunlop,P.S.M., Hamilton, J.W.J., Byrne, J.A., O’Shea, K., Entezari, M.H., Dionysiou, D.D.,2012. A review on the visible light active titanium dioxide photocatalysts forenvironmental applications. Appl. Catal. B – Environ. 125, 331–349.

Ramírez, C., Saldaña, A., Hernández, B., Acero, R., Guerra, A., Garcia-Segura, S., Brillas,E., Peralta-Hernández, J.M., 2013. Electrochemical oxidation of methyl orangeazo dye at pilot flow plant using BDD technology. J. Ind. Eng. Chem. 19, 571–579.

Ramos-Delgado, N.A., Hinojosa-Reyes, L., Guzmán-Mar, J.L., Gracia-Pinilla, M.A.,Hernández-Ramírez, A., 2013. Synthesis by sol–gel of WO3/TiO2 for solarphotocatalytic degradation of malathion pesticide. Catal. Today 209, 35–40.

Ruiz, E.J., Arias, C., Brillas, E., Hernández-Ramírez, A., Peralta-Hernández, J.M.,2011a. Mineralization of Acid Yellow 36 azo dye by electro-Fenton and solarphotoelectro-Fenton processes with a boron-doped diamond anode.Chemosphere 82, 495–501.

Ruiz, E.J., Hernández-Ramírez, A., Peralta-Hernández, J.M., Arias, C., Brillas, E.,2011b. Application of solar photoelectro-Fenton technology to azo dyesmineralization: effect of current density, Fe2+ and dye concentrations. Chem.Eng. J. 171, 385–392.

Salazar, R., Brillas, E., Sirés, I., 2012. Finding the best Fe2+/Cu2+ combination for thesolar photoelectro-Fenton treatment of simulated wastewater containing theindustrial textile dye Disperse Blue 3. Appl. Catal. B – Environ. 115–116, 107–116.

Sirés, I., Oturan, N., Oturan, M.A., Rodríguez, R.M., Garrido, J.A., Brillas, E., 2007.Electro-Fenton degradation of antimicrobials triclosan and triclocarban.Electrochim. Acta 52, 5493–5503.

Skoumal, M., Arias, C., Cabot, P.L., Centellas, F., Garrido, J.A., Rodríguez, R.M., Brillas,E., 2008. Mineralization of the biocide chloroxylenol by electrochemicaladvanced oxidation processes. Chemosphere 71, 1718–1729.

Su, C., Hong, B.Y., Tseng, C.M., 2004. Sol–gel preparation and photocatalysis oftitanium dioxide. Catal. Today 96, 119–126.

Sun, X., Liu, H., Zhang, Y., Zhao, Y., Quan, X., 2011. Effects of Cu(II) and humic acid onatrazine photodegradation. J. Environ. Sci. 23, 773–777.

Welcher, F.J. (Ed.), 1975. Standard Methods of Chemical Analysis, vol. 2, Part B, 6thed., R.E. Krieger Publishers, Huntington, New York, p. 1827.

Zhou, M., Särkkä, H., Sillanpää, M., 2011. A comparative experimental study onmethyl orange degradation by electrochemical oxidation on BDD and MMOelectrodes. Sep. Purif. Technol. 78, 290–297.