Available online at www.sciencedirect.com

www.elsevier.com/locate/chemosphere

Chemosphere 70 (2008) 1629–1636

Performance of Gd-doped Ti-based Sb-SnO2 anodesfor electrochemical destruction of phenol

Yujie Feng a,*, Yuhong Cui a, Bruce Logan b, Zhengqian Liu a

a Department of Environmental Science and Engineering, Harbin Institute of Technology, No. 202, Haihe Road, Nangang District, Harbin 150090, Chinab Department of Civil and Environmental Engineering, Penn State University, 231Q Sackett Building, University Park, PA 16802, USA

Received 4 April 2007; received in revised form 28 July 2007; accepted 31 July 2007Available online 24 October 2007

Abstract

The performance of electrodes for the electro-catalytic decomposition of a model pollutant (phenol) was enhanced using Gd-dopedTi/SnO2-Sb electrodes prepared by a thermal deposition method. Phenol degradation followed first-order rate kinetics, with the maxi-mum rate achieved using a 2% Gd doping level (molar ratio based on Gd:Sn) for tests conducted over a doping range of 1–10%. The first-order rate constant with 2% Gd was 0.044 min�1, versus 0.026 min�1 obtained with the control (plain Ti/SnO2-Sb). TOC removal andUV scans revealed that different intermediates were produced for different Gd contents, and that destruction efficiencies of these inter-mediates also varied with Gd doping levels of 1–5%. Electrodes were characterized by scanning electron microscopy, X-ray diffraction,electron dispersive spectrometry, and X-ray photon–electron spectroscopy. It is suggested that the state of specific active sites on theelectrode surface and the oxygen transfer activity at the electrode/electrolyte interface affect the performance of anodes with differentcompositions.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Gd; SnO2; DSA electrode; Phenol; Electrochemical degradation

1. Introduction

Electrochemical (EC) processes are promising, versatilemethods for degrading pollutants in low-volume wastewa-ter applications (Kotz et al., 1991; Walsh and Mills, 1993;Esplugas et al., 1994; Brillas et al., 1998, 2000; Belhadj andSavall, 1999). The feasibility of the technology is dependenton the development of anodes that have high stability,high activity with targeted chemicals, and low cost. Thechemical composition and structure of the coating on theelectrode surface are the main factors affecting the elec-tro-catalytic characteristics and stability of the electrodes.Some researchers have found that electrodes based on

0045-6535/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.chemosphere.2007.07.083

* Corresponding author. Tel.: +86 451 86283068; mobile:+13069891017; fax: +86 451 82373516.

E-mail addresses: yujief@hit.edu.cn, yujief@yahoo.com (Y. Feng).

SnO2, or minor components added to SnO2 in a DSA(dimensionally stable anode), achieved good reduction ofrecalcitrant and toxic chemicals based on their high oxygenevolution over-potential (Correalozano et al., 1997; Houket al., 1998; Vicent et al., 1998). SnO2 in its pure form isan n-type semiconductor with a wide band gap (3.87–4.3 eV), with a high electrical conductivity that is due toa modest level of non-stoichiometric impurities (i.e., anO:Sn ratio <2). Addition of small amounts of dopants,such as Sb, Cl, F, and Br, can enhance the electrical con-ductivity of the electrode and its performance in pollutantdegradation (Supothina and De Guire, 2000). The mostpromising electrodes are Sb-doped Ti-based SnO2 anodes(Ti/SnO2-Sb), based on their electro-catalytic activitiesand organic oxidation rates (Kotz et al., 1991; Stuckiet al., 1991; Polcaro et al., 1999).

Recent studies indicate that co-doped SnO2-Sb–x elec-trodes (x represents Ir, Fe, and Pt, etc.) get better perfor-mance in an EC process than Sb-doped SnO2 electrodes

0 30 60 90 120 1504

5

6

7

8

-5

-4

-3

-2

-1

0

-5

-4

-3

-2

-1

0

c

pH v

alue

Gd 5% Gd 3.3% Gd 2% Gd 1% Gd 0%

Time (min)

b

-ln[

TO

C]/

[TO

C] 0

Gd 5% Gd 3.3% Gd 2% Gd 1% Gd 0%

Gd 10% Gd 5% Gd 3.3% Gd 2% Gd 1% Gd 0%

-lnC

/C0

a

Fig. 1. Phenol, TOC removal and pH changes as a function of time fordifferent composition of Gd-doped Ti/SnO2-Sb electrodes. Initial phe-nol = 100 mg l�1, I = 0.12 A, (a) phenol removal; (b) TOC removal; and(c) pH changes.

1630 Y. Feng et al. / Chemosphere 70 (2008) 1629–1636

(Chen et al., 2002; He and Mho, 2004; Montilla et al.,2005). Previous research by our group has shown thatmetal doping of the anode changes the concentration ofoxygen vacancies in the SnO2 crystal lattice, and influencesits electro-catalytic performance (Cui and Feng, 2005).Doping the anode with rare earth (RE) metals of gadolin-ium (Gd) improves the EC oxidation rate of phenol. Forexample, phenol degradation rate was improved by 27%when Gd was added to the Ti-based Ti/Sb-Sn-RuO2 anode(Feng and Li, 2003).

It has been reported that RE metals enhance chemicalcatalytic processes by either acting as a catalyst or byassisting catalytic processes. RE metals have been success-fully applied in the petrochemical industry for oil refiningand showed a potential for use in fuel cell applications.Nevertheless, there are few studies on the use of RE metalsfor electro-catalytic degradation of pollutants. Someresearchers (Wang et al., 1995; Zou et al., 2001) have foundthat doping of RE metals changes some characteristic ofthe electro-catalytic electrodes, such as electric conductiv-ity, pyrolysis temperature, or oxygen evolution over-poten-tial. In the present study Gd was used as a RE metaldopant with Ti/SnO2-Sb electrodes to examine the degra-dation of a model pollutant (phenol). Characteristics ofelectrodes produced by the doping process that were exam-ined included morphology and physical structure of theelectrode surface.

2. Methods

2.1. Electrode preparation

Titanium plates, which served as the base metal for alloxide-coated electrodes, were polished using 40-grit and320-grit sand papers, degreased in hot 40% NaOH at80 �C for 2 h, and then etched in 6 M HCl at 90 �C for1 h followed by thorough washing with distilled water.The inner layers of Gd-doped Ti/SnO2-Sb electrodes wereprepared by electro-deposition of Sn followed by thermaloxidation. Taking the pretreated Ti plate as the cathodeand Ti/RuO2 as anode, electro-deposition of metal Snwas achieved at constant current of 0.12 A for 25 minin an ethanol (100 ml) solution containing 17.5 gSnCl4 Æ 5H2O, 0.73 g Sb2O3, and 2 ml of concentrated(37%) HCl. The electrodes were then dried and heatedunder air at an annealing temperature of 400 �C for 2 h.The outer layer of the electrode was prepared by dippingthe electrode into a solution, which consisted of 30 gSnCl4 Æ 5H2O, 0.8 g Sb2O3, 2.5 ml of concentrated HCland a variable concentration of Gd(NO3)3 (molar ratiosof Gd/Sn of 1:100, 2:100, 3.3:100, 5:100, and 10:100,respectively) in 50 ml n-butanol. The electrode was thendried in an infrared oven. After five cycles of both dippingand drying, the Ti plates were heated in a muffle oven(450 �C for 20 min) for coating pyrolysis. This whole pro-cess (dipping, drying and pyrolysis) was repeated 3–5 timesand finally, the electrodes were heated at 650 �C for 3 h.

2.2. Electrolysis

Taken phenol as model organic compounds, the elec-trolysis cells were a series of 100 ml beakers made of glass.For each cell, a 6 cm2 (2 cm · 3 cm) anode, prepared asdescribed above, and a stainless steel cathode having thesame area were placed in the beaker at a spacing of15 mm between the electrodes. A DC potentiostat with avoltage range of 0–30 V was used as the power supply fororganic degradation studies. Phenol (100 mg l�1, 80 ml)was placed in each cell with electrolyte (Na2SO4, 0.25 M).Electrolysis was performed under galvanostatic control at0.12 A. At the end of each run, based on a pre-determinedtime of electrolysis, the solution in the cell was analyzed forphenol, total organic carbon (TOC), pH, and UVabsorbance.

2.3. Analysis methods

Phenol concentration was measured using a gas chro-matograph (HP6890, Hewlett Packard) equipped with aflame ionization detector and a 10 m · 0.53 mm HP-FFAPfused-silica capillary column. Samples were filteredthrough a 0.2 lm filter, and acidified using formic acid.The temperature of the injector and detector were bothconstant (200 �C), and helium was used as the carrier gas.

Y. Feng et al. / Chemosphere 70 (2008) 1629–1636 1631

TOC was analyzed using a TOC Analyzer (Shimadu5000A). UV absorbance was measured using a UV/visspectrometer (Lambda12, Perkin Elmer).

2.4. Surface characterization of electrodes

Several different types of analyses were conducted toanalyze the chemical composition, morphology, crystalstructure, and distribution of elements on the modifiedelectrodes. The morphology and element analysis of theelectrodes were obtained using a scanning electron micro-scope (KYKY-AMRAY MODEL-1000B) equipped withan energy dispersive spectrometer (EDS). X-ray dot-map-ping was used to analyze the distribution of coating ele-ments. X-ray diffraction (XRD) patterns of the coating

Table 1Kinetic parameters of phenol degradation in electrochemical process byGd-doped Ti/SnO2-Sb anode

Gd content (Gd:Sn, molar ratio) k (min�1) R2

10:100 0.004 ± 0.001 0.87795.0:100 0.014 ± 0.001 0.97573.3:100 0.023 ± 0.002 0.98612.0:100 0.044 ± 0.005 0.97321.0:100 0.024 ± 0.001 0.99750.0:100 0.026 ± 0.003 0.9801

Abs

.A

bs.

a

b

Fig. 2. UV scan curves of electrolytes for: (a) phenol and some possible intermUV absorbency at 269 nm and (d) at 290 nm of electrolytes as a function of time

films on these electrodes were recorded on a D/max-rBXRD instrument (Rigaku, Japan), using Cu Ka radiationand a graphite monochromator, with an operating voltageof 45 kV and current of 50 mA. The average particle size ofthe SnO2 crystals was calculated from XRD data using theScherrer equation (Kitazawa et al., 2006). X-ray photoelec-tron spectroscopy (XPS) measurements were carried outusing a PHI5700 spectrometer with Al Ka radiation(ht = 1486.6 eV).

3. Results and discussion

3.1. The performance of Gd-doped Ti/SnO2-Sb anodes

A maximum rate of phenol removal for Gd-doped Ti/SnO2-Sb electrodes was achieved using a Gd/Sn molarratio of 2% (Fig. 1a). Observed rates of phenol removalwere fitted with first-order rate constants (Table 1). Therate with a 2% Gd-doped electrode was 41% higher thanthe control (0% Gd) and 46% higher than using an elec-trode with 1% Gd/Sn. Slower phenol degradation rateswere also produced with Gd/Sn ratios of 3.3%, 5%, and10%.

The same general trend in the rate of removal of allorganic species in solution (indicated by TOC concentra-tions) with the Gd/Sn content of the electrode was initially

0 30 60 90 120 1500.0

0.1

0.2

0.3

0.4

0.5

0.6

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Gd 5% Gd 3.3% Gd 2% Gd 1% Gd 0%

Time (min)

d Gd 5% Gd 3.3% Gd 2% Gd 1% Gd 0%

c

ediates; (b) UV scan of electrolytes of 2% Gd-doped Ti/SnO2-Sb anode; (c)for different composition of Gd-doped Ti/SnO2-Sb electrodes, I = 0.12 A.

Fig. 3. Micrographs (·1000) of coating films on: (a) 2 mol% Gd-dopedTi/SnO2-Sb electrode, and (b) 0 mol% Gd-doped Ti/SnO2-Sb electrode.

1632 Y. Feng et al. / Chemosphere 70 (2008) 1629–1636

observed (i.e., first-order kinetics) (Fig. 1b). However, thedegradation rate began to increase later during the ECprocess. As a result, the overall rate of TOC removalwas not a first-order process over time, suggesting thatthere was formation of different chemical intermediatesduring phenol oxidation that were subsequently degradedat different rates by the electrodes. There was very poorperformance of the electrode containing 10% Gd/Sn,and therefore this electrode was not examined further insubsequent tests. For the 2 mol% Gd/Sn electrode, 74%of the TOC was removed in 90 min versus only 64% withthe control. After 2.5 h, only 1% TOC remained in solu-tion for the 2% Gd/Sn electrode, but 7% TOC remainedin the control solution.

The pH of the electrolyte initially decreased, but thenincreased to values equal to or larger than the initialpH (Fig. 1c). The final pH for the 2 mol% Gd-dopedTi/SnO2-Sb anode was 7.0 after 2.5 h, which is onlyslightly higher than the initial pH of 5.7. The formationof acidic intermediates and the further degradation ofthese acids were thought to be the main reason for thepH change. Based on previous studies, chemical interme-diates formed during phenol degradation likely containcarboxylic acids, which lower the solution pH (Comninel-lis and Pulgarin, 1991; Lund and Baizer, 1991). As theseintermediates are consumed, the pH return to values sim-ilar to the initial condition. Thus, a cycle of pH fluctua-tions with TOC removal should occur, consistent withthat observed here.

3.2. EC degradation of phenol on different Gd-dopedTi/SnO2-Sb anodes

UV scans (200–400 nm) were used to further investigatepossible intermediates formed in solution during the ECdegradation of phenol with Gd-doped Ti/SnO2-Sb anodes.Previous research on EC degradation of phenol has shownthat phenol is first oxidized to benzoquinone, and then thering is opened to form other small organics such as car-boxylic acids (Comninellis and Pulgarin, 1991; Lund andBaizer, 1991; Gattrell and Kirk, 1993). UV scans of phe-nol and three different potential intermediates (1,4-benzo-quinone, maleic acid and malonic acid) are shown inFig. 2a, along with scans of solutions taken at differenttimes (Fig. 2b). The peak at 269 nm indicates the presenceof the aromatic ring of phenol (Feng and Li, 2003), so theloss in peak height indicates a loss of phenol due to ringcleavage. This decrease in the 269 nm peak was accompa-nied by an increase in peak height at 290 nm, suggestingthe formation of benzoquinone. For the 5% Gd/Sn elec-trode, the 290 nm peak height was greatest compared tothe other electrodes, showing that more benzoquinoneaccumulated in solution due to the high Gd content ofthe electrode (Fig. 2d). For the 2 mol% Gd-doped elec-trode, the absorbance at 269 and 290 nm were amongthe lowest of the four electrodes examined over most ofthe reaction times (Fig. 2c and d). These UV scan results

therefore indicate that the Gd content of the electrodeinfluenced both the rate of the EC phenol degradationprocess as well as the intermediates formation (e.g.,benzoquinone).

3.3. Morphology and structure analysis of Gd-dopedSnO2-Sb oxides film

Freshly prepared metal oxide coatings of Gd-doped Ti/SnO2-Sb electrodes which had not been used for electroly-sis were examined for morphology by scanning electronmicroscopy. The surface coating of the Gd-doped Ti/SnO2-Sb electrode (2 mol%, Gd/Sn) was much morecompact and smooth than that of the control (0 mol%)(Fig. 3). A smoother surface favors resistance towardsEC corrosion, and therefore helps to prolong the service-life time of the electrode. EDS confirmed the presence ofSn, Sb, Gd, and Ti in the Gd-modified electrodes(Fig. 4a). X-ray dot-mapping, which was used for analyz-ing the metal distributions, found that elements of Ti, Sn,Sb, and Gd were all well distributed on the base Ti surface(Fig. 4b).

Fig. 4. Elements analysis in the oxides electrode film: (a) EDS results; (b) X-ray dot-mapping of oxides film, distribution of Ti, Sn, Sb and Gd.

Y. Feng et al. / Chemosphere 70 (2008) 1629–1636 1633

XRD patterns of these two electrodes showed a seriesof diffraction peaks corresponding to rutile-type SnO2 forboth surfaces. TiO2 diffraction peaks were especiallymore evident in the control electrode (Fig. 5). The averageparticle size of the SnO2 crystals, calculated fromXRD data using the Scherrer equation, ranged from 8 to13 nm.

3.4. Proposed function of Gd doping

Based on the performance of the electrodes and changesin electrode morphology at different Gd doping levels, it is

thought that the introduction of Gd into the Ti/SnO2-Sbanode structure changed the SnO2 crystal formation pro-cess (Fig. 5) and thus improved its performance. The radiusof the Gd3+ cation (94 pm) is much larger than that of theSn4+ cation (71 pm), and thus it would be difficult for alladded Gd atoms to fit within the SnO2 crystal lattice. Lat-tice parameters of SnO2 were calculated from XRD data(Fig. 5) in order to better understand the doping state ofGd3+ present on the electrode. XRD data show that therutile SnO2 lattice was slightly expanded with the 2 mol%Gd-doped electrode (a = b = 0.478 and c = 0.319 nm)compared with the control lacking Gd (a = b = 0.472 and

0 20 40 60 80 100

2 θ (degrees)

TiO2

SnO2

Gd 2 mol%

Gd 0 mol%

Inte

nsity

(a.

u.)

Fig. 5. XRD patterns of coating films on 2 and 0 mol% Gd-dopedTi/SnO2-Sb electrodes.

525 530 535 540 545

Inte

nsity

(a.

u.)

Binding energy (eV)

Gd 0 mol%

Gd 2 mol%

Fig. 6. XPS spectra of O 1s of coating films on 2 and 0 mol% Gd-dopedTi/SnO2-Sb electrodes.

1634 Y. Feng et al. / Chemosphere 70 (2008) 1629–1636

c = 0.320 nm). A reasonable explanation for the latticeexpansion is that Gd3+ ions entered the unit cell of theSnO2, with the larger Gd3+ ions replacing the smallerSn4+ ions.

It is generally believed that during EC pollutant degra-dation organic compounds in aqueous solution are oxi-dized by both direct electron transfer and by indirectoxygen atom transfer (Chiang et al., 1995; Polcaro et al.,1999). In the direct oxidation process, organics areadsorbed on the anode surface and they transfer electronsdirectly to the anode. With an indirect process, hydroxylradicals adsorbed on the anode surface readily react withthe organic molecules adsorbed on (or in the vicinity of)the anode. In both cases the specific sites on the anode sur-face which can adsorb either organic molecules or hydroxylradicals produce effective degradation reactions (Belhadjand Savall, 1999). In all likelihood these active sites areidentical for electrodes with and without Gd doping, butcertain levels of Gd doping enhance this activity.

The poor performance of the electrodes with a high Gdcontent (5 or 10 mol%) likely arises from the inability of allof the added Gd atoms to fit within the SnO2 lattice. Exces-sive addition of Gd would result in Gd deposition on theelectrode surface, decreasing the number of active sites.Another possible reason for the activity decrease of anodeswith a high Gd content could be related to the SnO2 grainsize. Our XRD data indicate that the SnO2 grain sizeincreased with Gd content from 8.8 nm (no Gd) to9.3 nm (2 mol% Gd) and 12.7 nm (5 mol% Gd). As is wellknown for a heterogeneous catalysis (Pirkanniemi and Sil-lanpaa, 2002), a smaller grain size means a larger interfaceand more active sites. Thus, we conclude that over-dopingof Gd led to less active sites and reduced catalyticperformance.

The improved performance of the 2%-doped Gd elec-trode likely resulted from indirect phenol oxidation dueto an increased hydroxyl radical activity. SnO2 is a semi-

conducting metal oxide, so its properties depend stronglyon deviation from the stoichiometric composition (nativedefects) and on the nature and concentrations of foreignatoms incorporated into the crystal lattice. Oxygen vacancyis the predominant atomic defect for SnO2 and the concen-tration of these defects determines the electrical propertiesof this material (Safonova et al., 2000). Some researchershave suggested that possible oxygen transition may occurfrom adsorbed hydroxyl radicals to the crystal lattice ofthe oxide anode, forming so-called higher metal oxides(MOx+1), and that these higher oxide species also partici-pate in organic oxidation processes (Comninellis, 1994).The mechanism can be detailed as shown below, whereMOx stands for the oxide anode and R stands for theorganic pollutant

MOx + H2O!MOx [�OH] + Hþ + e� ð1Þ

MOx[�OH] + R! CO2 + MOx ð2Þ

MOx [�OH]!MOxþ1 þHþ + e� ð3Þ

MOxþ1 þR! ROþMOx ð4Þ

The existence of a number of oxygen vacancies on ametal oxide anode is essential for the formation ofMOx+1. Thus, the concentration of oxygen vacancies isthought to influence the ratio of MOx[�OH] versusMOx+1 on the anode surface, and therefore the catalyticperformance. The presence of MOx[�OH] favors EC com-bustion of organic species, while MOx+1 favors EC conver-sion, which means more intermediates production.

XPS was employed to study the surface composition ofthe electrode samples to determine if the oxygen content ofthe lattice was increased as a result of Gd doping. A com-parison of the O 1s spectra of freshly prepared 2 mol% Gd-doped sample and a control (0 mol%) is shown in Fig. 6. Itis clear that the O 1s binding energy of the 2 mol%

Y. Feng et al. / Chemosphere 70 (2008) 1629–1636 1635

Gd-doped sample is lower than that of the sample lackingGd, indicating there was more lattice oxygen formed in theGd-doped sample than that of the control (Montilla et al.,2004). Therefore, it appears likely that the enhanced per-formance of the Gd-doped electrode arose from anincreased activity of adsorbed hydroxyl radicals due tothe higher oxygen content of the crystal lattice in the pres-ence of the RE metals.

The valance state of Sb was determined to be +5 by ourXPS analysis, as the banding energy of Sb 3d5/2 is around530.9 eV for all of the samples, in agreement with otherresearchers (Correalozano et al., 1996). Because SnO2 isan n-type semiconducting metal oxide, either oxygenvacancies or interstitial tin atoms, are expected to bedonors in pure SnO2. The higher valance of Sb5+ ions thanSn4+ ions could provide more excessive electrons and func-tion as dominant donors for the Sb-doped SnO2 semicon-ductor, improving conductivity of the coating electrode.On the other hand, the natural donors of pure SnO2, oxy-gen vacancies or interstitial tin atoms are inhibited as aresult of the introduced foreign donors.

It is worth noting that there is a direct correlationbetween energy band and catalysis activity for RE metaloxides. There are many 4f-electron orbits for rare-earthatoms and, when the rare earth element was doped intothe DSA coating materials, additional energy bands couldbe introduced into the structure of the coating metal oxi-des. This might help to develop convenient channels forelectron transition and also help to enhance the electro-catalytic characteristics of the anodes.

4. Conclusions

The addition of rare earth Gd improved the morphol-ogy and performance of Ti/SnO2-Sb electrodes. Basedon the doping technique used here, it was found that per-formance of the anode was optimized using a 2 mol%addition of Gd/Sn. Both phenol and intermediate prod-ucts (e.g., benzoquinone) were decomposed more rapidlyat a 2% doping for electrodes examined over a range of0–10% doping levels. At the higher Gd doping levels thedata suggest that many Gd atoms could not fit into thecrystal lattice and deposit on the electrode surface, thusdecreasing the number of active catalytic sites. IntroducingGd atoms may enhance pollutant degradation by chang-ing the concentration of oxygen vacancies in the SnO2

crystal lattice, and thus influencing the transition processof oxygen atoms from adsorbed hydroxyl radicals to thecrystal lattice, leading to different behavior of the anodesin the EC process.

Acknowledgements

This research was supported by the National NatureScience Foundation of China (Projects 50278022and 50638020) and the Chinese 973 Key Project(2004CB41850). The research was also supported by Pro-

gram for Changjiang Scholars and Innovative ResearchTeam in University of China (IRT0424).

References

Belhadj, T.N., Savall, A., 1999. Electrochemical degradation of phenol inaqueous solution on bismuth doped lead dioxide: a comparison of theactivities of various electrode formulations. J. Appl. Electrochem. 29,277–283.

Brillas, E., Calpe, J.C., Casado, J., 2000. Mineralization of 2, 4-D byadvanced electrochemical oxidation processes. Water Res. 34, 2253–2262.

Brillas, E., Mur, E., Sauleda, R., Sanchez, L., Peral, J., Domenech, X.,Casado, J., 1998. Aniline mineralization by AOP’s: anodic oxidation,photocatalysis, electro-Fenton and photoelectro-Fenton processes.Appl. Catal. B: Environ. 16, 31–42.

Chiang, L.C., Chang, J.E., Wen, T.C., 1995. Indirect oxidation effect inelectrochemical oxidation treatment of landfill leachate. Water Res. 29,671–678.

Comninellis, Ch., 1994. Electrocatalysis in the electrochemical conversion/combustion of organic pollutants for waste-water treatment. Electro-chim. Acta 39, 1857–1862.

Comninellis, Ch., Pulgarin, C., 1991. Anodic oxidation of phenol forwaste-water treatment. J. Appl. Electrochem. 21, 703–708.

Correalozano, B., Comninellis, Ch., De Battisti, A., 1996. Physicochem-ical properties of SnO2–Sb2O5 films prepared by the spray pyrolysistechnique. J. Electrochem. Soc. 143, 203–209.

Correalozano, B., Comninellis, Ch., De Battisti, A., 1997. Service life ofTi/SnO2–Sb2O5 anodes. J. Appl. Electrochem. 27, 970–974.

Chen, G.H., Chen, X.M., Yue, P.L., 2002. Electrochemical behavior ofnovel Ti/IrOx–Sb2O5–SnO2 anodes. J. Phys. Chem. B 106, 4364–4369.

Cui, Y.H., Feng, Y.J., 2005. EPR study on Sb doped Ti-base SnO2

electrodes. J. Mater. Sci. 40, 4695–4697.Esplugas, S., Yue, P.L., Pervez, M.I., 1994. Degradation of 4-chlorophe-

nol by photolytic oxidation. Water Res. 28, 1323–1328.Feng, Y.J., Li, X.Y., 2003. Electro-catalytic oxidation of phenol on

several metal-oxide electrodes in aqueous solution. Water Res. 37,2399–2407.

Gattrell, M., Kirk, D.W., 1993. A study of the oxidation of phenol onplatinum and preoxidized platinum surfaces. J. Electrochem. Soc. 140,1534–1540.

He, D., Mho, S.I., 2004. Electrocatalytic reactions of phenolic compoundsat ferric ion co-doped SnO2:Sb5+ electrodes. J. Electroanal. Chem. 568,19–27.

Houk, L.L., Johnson, S.K., Feng, J., Houk, R.S., Johnson, D.C., 1998.Electrochemical incineration of benzoquinone in aqueous media usinga quaternary metal oxide electrode in the absence of a solublesupporting electrolyte. J. Appl. Electrochem. 28, 1167–1177.

Kitazawa, S., Choi, Y.S., Yamamoto, S.Y., Yamaki, T.Y., 2006. Rutileand anatase mixed cryatal TiO2 thin films prepared by pulsed laserdeposition. Thin Solid Films 515, 1901–1904.

Kotz, R., Stucki, S., Carcer, B., 1991. Electrochemical wastewatertreatment using high overvoltage anodes, 1. Physical and electrochem-ical properties of SnO2 anodes. J. Appl. Electrochem. 21, 14–20.

Lund, H., Baizer, M.M., 1991. Organic Electrochemistry: An Introductionand a Guide. Marcel Dekker, New York.

Montilla, F., Morallon, E., De Battisti, A., Barison, S., Daolio, S.,Vazquez, J.L., 2004. Preparation and characterization of antimony-doped tin dioxide electrodes. 3. XPS and SIMS characterization. J.Phys. Chem. B 108, 15976–15981.

Montilla, F., Morallon, E., Vazquez, J.L., 2005. Evaluation of theelectrocatalytic activity of antimony-doped tin dioxide anodes towardthe oxidation of phenol in aqueous solutions. J. Electrochem. Soc. 152,B421–B427.

Pirkanniemi, K., Sillanpaa, M., 2002. Heterogeneous water phase catalysisas an environmental application: a review. Chemosphere 48, 1047–1060.

1636 Y. Feng et al. / Chemosphere 70 (2008) 1629–1636

Polcaro, A.M., Palmas, S., Renoldi, F., Mascia, M., 1999. On theperformance of Ti/SnO2 and Ti/PbO2 anodes in electrochemicaldegradation of 2-chlorophenol for wastewater treatment. J. Appl.Electrochem. 29, 147–151.

Safonova, O.V., Rumyantseva, M.N., Kozlov, R.I., Labeau, M., Dela-bouglise, G., Ryabova, L.I., Gaskov, A.M., 2000. Two successiveeffects in the interaction of nanocrystalline SnO2 thin film withreducing gases. Mater. Sci. Eng. B 77, 159–166.

Stucki, S., Kotz, R., Carcer, B., Suter, W., 1991. Electrochemical waste-water treatment using high overvoltage anodes. 2. Anode performanceand applications. J. Appl. Electrochem. 21, 99–104.

Supothina, S., De Guire, M.R., 2000. Characterization of SnO2 thin filmsgrown from aqueous solutions. Thin Solid Films 371, 1–9.

Vicent, F., Morallon, E., Quijada, C., Vazquez, J.L., Aldaz, A., Cases, F.,1998. Characterization and stability of doped SnO2 anodes. J. Appl.Electrochem. 28, 607–612.

Walsh, F., Mills, G., 1993. Electrochemical techniques for a cleanerenvironment. Chem. Ind. 8, 576–580.

Wang, Y.G., Lin, J.Y., Chen, K.N., 1995. A study of metallic anodecoatings with rare earth compounds (I). J. East China Normal U. (Nat.Sci.) 4, 60–64 (in Chinese).

Zou, Z., Li, J., Ding, F.Q., Zhang, W.G., Xiao, J., Ye, S.L., Liu, Y.X.,2001. Effect of doping with rare earth Eu on electrocatalysis ofmetal oxide anode coating. Chin. J. Nonferr. Met. 11, 91–94, inChinese.