Structural studies of electrochemically activated glassy carbon electrode Effects of chloride anion

Electrochimica Acta52(2007)

5907–5913

Structural studies of electrochemically activated glassy carbon electrode: Effects of chloride anion on the redox responses of copper deposition Kang Shi a,?,Kun Hu a,Sheng Wang a,Chung-Yin Lau b,Kwok-Keung Shiu b,??

a Department of Chemistry,Xiamen University,Xiamen361005,China

b Department of Chemistry,Hong Kong Baptist University,Kowloon Tong,Kowloon,Hong Kong,China

Received4December2006;received in revised form8March2007;accepted9March2007

Available online19March2007

Abstract

The redox behaviors of copper species at electrochemically activated glassy carbon electrodes have been investigated in aqueous solutions containing chloride anions.Experimental results showed that the voltammetric responses of copper species were in?uenced by the electrochemical activation means employed.Abnormal copper stripping was observed at electrodes obtained by cyclic polarization.Cyclic polarization would cause changes in the interwoven graphitic crystalline surface of glassy carbon,producing electrode interface with low distribution density of electron transfer sites for the early nucleation of metallic copper in the presence of chloride anion.Potentiostatic activation would generate oxygen-containing functionalities and maintain the basic surface structure of graphitic crystalline with high distribution density of electron transfer sites.

?2007Elsevier Ltd.All rights reserved.

Keywords:Glassy carbon;Electrochemical activation;Surface structure;Copper deposition and stripping;Chloride anions

1.Introduction

Glassy carbon electrode is possibly the most used sp2 hybridized carbon composite electrode in electrochemical research.Its bulk has a complex structure of interwoven graphitic ribbons and the basic structural units of nanometer dimensions are planar aromatic structures with basal and edge planes[1–3]. Usually,glassy carbon electrode undergoes pretreatment pro-cess before use in order to obtain reproducible responses[3]. Simple electrochemical activation is considered a better and quicker in situ pretreatment means[4–9],when compared with other pretreatment methods such as laser activation[10–13] and mechanical polishing[14–17].Electrochemically pretreated glassy carbon electrode(PGCE)usually gives reproducible surface,improved electron transfer and speci?c adsorption behaviors[3,18–20].On the other hand,electrochemical activa-tion might also result in more complex surface structures.Better understanding of the new interfacial structure is necessary and important for its potential applications in many?elds.

?Corresponding author.

??Corresponding author.Tel.:+852********;fax:+852********.

E-mail address:kkshiu@https://www.360docs.net/doc/0f1028718.html,.hk(K.-K.Shiu).

In general,the electrode is electrochemically activated either in basic or acidic(or neutral)https://www.360docs.net/doc/0f1028718.html,rger background cur-rent and higher fraction of oxygen contents have been observed for the electrode pretreated in acidic(or neutral)solutions,and the activated surface has been proposed to be a porous oxi-dized multilayer[4,6,21–23].Raman investigation indicated that electrochemical activation resulted in a high fraction of edge planes and an increasing number of graphitic microcrystalline defects[24].Kinetic studies suggested that electron transfer at the oxidized surface was affected by the electrochemical sur-face structure[1],oxygen-containing functional groups[25], cleanliness[23]and hydrophobicity[26],etc.

Two voltammetric methods are usually employed to activate the glassy carbon electrodes.These include the polarization of the electrode at a highly oxidative potential or the application of potential cycling in a wide potential range[19,20,27,28].Several experimental methods have been used to examine the structures of the graphite oxide?lm.Ellipsometric studies showed the growth of a transparent oxidized?lm when cyclic polarization was employed[29,30].The void fraction of the?lm obtained by potential cycling was estimated to be80%,while the void frac-tion of the?lm obtained by potentiostatic activation was between zero and70%[29,30].Nagaoka et al.[31–38]suggested that electrochemically activated glassy carbon obtained by constant

0013-4686/$–see front matter?2007Elsevier Ltd.All rights reserved. doi:10.1016/j.electacta.2007.03.028

5908K.Shi et al./Electrochimica Acta52(2007)5907–5913

potential oxidation could create micropore structures with the average radius of the surface pores being about10–20?A.We have previously demonstrated that the surface structures of the PGCE were affected by the activation procedures employed[27]. The average size of the void space developed by potentiostatic activation was larger than those obtained by cyclic polarization in sulfuric acid solution.Adsorption studies showed that the rel-ative sizes of both the adsorbent and the void space had strong in?uence on the adsorption behaviors of the PGCE[39,40].

Metallic copper species often showed good reversible response without under-potential deposition at carbon electrode surface[41–45].The redox processes of copper species were strongly affected by the presence of chloride anions in solu-tion[43,45].As suggested by Miller et al.[46,47],the density of intrinsic active sites for nucleation in?uenced the processes involved in the electrochemical deposition and stripping of cop-per.When the metallic portion produced by early nucleation was prematurely oxidized to Cu(I)Cl passivation layer during the anodic process,the electro-conductive connection between metallic crystalloid and the electrode would be broken,causing inhibition to the oxidation of metallic crystalloid.Only after the application of a potential positive enough,the oxidation of copper crystalloid would continue[46,47].The fraction and distribution of electro-conductive metallic copper played a key role in the oxidation of the whole metallic copper deposits.The redox behaviors of copper chloride species re?ected the differ-ence in the distribution density of electron transfer sites within nanometer range.The presence of low distribution density of electron transfer sites would result in an abnormal oxidation behavior of metallic copper in the presence of chloride anion [46,47].

In this report,the PGCE surface was investigated on the basis of the deposition/dissolution reactions of metallic copper,dif-ferent from the soluble electroactive probes employed in most of the previous studies.It has been reported that the nucleation density of metallic copper depositing at the pretreated carbon electrode could be used to quantify the activated properties of carbon electrode surface[41].The early nucleation of metallic copper might re?ect the surface complications of the PGCE.The effects of chloride anion on the deposition/dissolution reactions of metallic copper at the PGCE might offer a new way to exam-ine the interfacial characteristics and structures of the PGCE prepared by different procedures.

2.Experimental

Copper(II)chloride,copper(II)sulfate and sodium chloride were obtained from Aldrich,and were used as received.Cop-per(I)chloride solution was freshly prepared by oxidation of highly pure metal copper wire(99.99%)at a constant poten-tial in sodium chloride solution(pH3.0).All chemicals were of reagent grade and were used as received.Reagent solutions were freshly prepared and degassed for5min before use.The pH of the reagent solutions was adjusted by the addition of hydrochloric acid or sodium hydroxide.Deionized water was obtained by puri?cation through a Millipore system and was used throughout.

V oltammetric measurements were recorded with a CHI-660A V oltammetric Analyzer(CH Instruments Inc.).A conventional three-electrode cell was employed,incorporating a glassy car-bon working electrode,a saturated calomel reference electrode (SCE)and a platinum foil counter electrode.All potentials were quoted versus the SCE reference.Electrode rotation was con-trolled by a Pine MSRX rotator/controller(Pine Instrument Company).

Highly oriented pyrolytic graphite(HOPG)electrode(ZYH grade,Advanced Ceramics)was?xed on a steel plate with car-bon glue and was resurfaced by peeling away surface layers using adhesive tape before use.Electrochemical measurements were carried out in an electrochemical cell similar to that described by Bowler et al.[48].The electrode area exposing to solution was limited to0.165cm2by a silicone O-ring seal. The edge plane of the same HOPG was sealed by epoxy.An elec-trode area of0.04cm2was exposed to solution through either newly cutting or polishing progressively with0.3?m Al2O3 powder.

Both glassy carbon voltammetric electrodes and rotating-disc glassy carbon electrodes of5.0mm in diameter were obtained from Tokai Carbon Company.Glassy carbon electrodes were polished progressively with?ner emery paper,then thoroughly with0.3?m Al2O3powder on polishing cloth.The working electrode was cleaned in an ultrasonicating bath for1min before used.

Two different voltammetric activation methods were employed in0.5M H2SO4solutions.For potentiostatic activa-tion,the glassy carbon electrode was anodized at+2.0V for a short period of time(from0.5to3.0min)and then polarized at?1.0V for1min.The potential was then cycled between +0.8and?0.5V at a scan rate of0.1V/s for two cycles and ended at?0.5V.Cyclic polarization could also be employed for the activation of glassy carbon electrodes.The electrode was cycled between?0.3and+2.0V at0.1V/s for8–10voltammet-ric cycles until stable voltammograms were obtained.

3.Results and discussion

3.1.Voltammetric responses of copper(II)chloride at

PGCE obtained by different pretreatment means

Fig.1(a)shows the cyclic voltammogram of polished glassy carbon in0.5mM CuSO4+0.1M Na2SO4(pH3.0).A broad cathodic wave centred at?0.16V was observed.On the reverse scan,a sharp anodic peak appeared at+0.02V.The effects of chloride anion on the redox processes of metal copper deposi-tion at glassy carbon electrodes were investigated.The pH of the test solution was adjusted to3.0in order to avoid possible precip-itation as Cu(OH)1.5Cl0.5[43].Fig.1(b)shows the voltammetric response of polished glassy carbon in a0.5mM CuCl2+0.1M NaCl solution,when cycled between+0.6and?0.6V.Two cathodic peaks were observed at+0.06and?0.32V,and two anodic peaks appeared at?0.09and+0.12V in the follow-ing anodic scan.The voltammetric behaviors were similar to those reported at graphite and gold electrodes[42,46,49]. Three reactions were involved within such potential range

K.Shi et al./Electrochimica Acta52(2007)5907–5913

5909

Fig.1.V oltammetric responses of glassy carbon electrodes in copper solutions.

(a)Polished GCE in0.5mM CuSO4+0.1M Na2SO4(pH3.0);polished GCE

(b)and PGCE obtained by potentiostatic activation with anodization time of 10s(c);45s(d);90s(e)in0.5mM CuCl2+0.1M NaCl(pH3.0).Scan rate: 20mV/s.

[46]:

Cu2++Cl?+e?→CuCl(1) CuCl+Cl?→CuCl2?(2) CuCl+e?→Cu+Cl?(3) When the potential was scanned between+0.6and?0.1V,only the?rst cathodic peak at+0.06V and the second anodic peak at+0.12V were observed.These two redox peaks appeared to be a reversible redox couple,corresponding to the combined reaction involving both reactions(1)and(2)at high chloride ion concentration,and can be expressed as reaction(4)[43,46,50]:

Cu2++2Cl?+e?→CuCl2?(4) Copper metallic phase began to deposit at potentials more neg-ative than?0.26V and the corresponding cathodic peak was observed at?0.32V,as shown in Fig.1(b).The deposition pro-cesses might involve many factors including electron transfer kinetics,early nucleation of metal and interfacial structure of electrode[51,52].

According to the literature reports[46,50],metallic copper was?rstly oxidized to the barely conductive CuCl precipitate ?lm in chloride solution,as described by the reverse reac-tion of reaction(3).When the chloride ion concentration was high enough to cause dissolution of the CuCl?lm as described by reaction(2),the underneath metallic copper would become exposed and then be continuously oxidized[49,50].The anodic peak observed at?0.09V in the reverse scan was resulted from both reactions(2)and(3).The soluble CuCl2?species pro-duced were subsequently oxidized at more positive potential,Table1

V oltammetric responses of glassy carbon electrodes in0.5mM CuCl2+0.1M NaCl(pH3.0)

Electrode E pc(1)(V)E pc(2)(V)E pa(3)(V)E pa(4)(V) Polished GCE+0.06?0.32?0.09+0.12 PGCE(potentiostatic activation)

10s+0.05?0.31?0.09+0.13

45s+0.03?0.32?0.10+0.14

90s+0.02?0.32?0.10+0.15 PGCE(cyclic polarization)

1cycle+0.05?0.35?0.09+0.12,+0.04 3cycles+0.04?0.36?0.10+0.12,+0.05 9cycles+0.02?0.38?0.11+0.10(IIIa) corresponding to the second anodic peak at+0.12V described by reaction(4).

Electrochemically activated glassy carbon electrode(PGCE) obtained by potentiostatic activation was also studied.The extent of activation for PGCE obtained by potentiostatic activation would depend on the anodization time[18–20].In practice, the anodization time used should be less than90s to avoid memory effects and poor reproducibility[19].Curves(c),(d) and(e)of Fig.1show the voltammetric responses of PGCE obtained by potentiostatic activation with different anodization time.The cyclic voltammograms were characterized by two cathodic peaks and two anodic peaks,very similar to those for the polished glassy carbon.The corresponding voltammetric data are tabulated in https://www.360docs.net/doc/0f1028718.html,ually,the peak current increased with increasing extent of activation.This might result from the increases in the actual electrode area and the background current. Similar to those observed for the polished GCE,the peak cur-rent of the second anodic peak at PGCE was always smaller than the?rst anodic peak.Meanwhile,an additional anodic shoulder peak was observed at around zero potential for the PGCE with anodization time longer than45s,as shown in Fig.1(d)and(e).

Electrochemical activation of glassy carbon electrodes can also be conducted by cyclic polarization https://www.360docs.net/doc/0f1028718.html,ually, a stable activation voltammogram was resulted after a nine-cycle activation[18,20].Fig.2shows the typical voltammetric responses of PGCE,obtained by cyclic polarization with differ-ent extent of activation,in0.5mM CuCl2+0.1M NaCl solution at pH3.0.The corresponding voltammetric data are also shown in https://www.360docs.net/doc/0f1028718.html,paring with the PGCE obtained by potentio-static activation,deposition of metallic copper occurred at more negative potential when cyclic polarization was employed for electrochemical activation.The peak current for the?rst anodic wave decreased obviously as the extent of activation increased. The most obvious change was that a new anodic peak(IIIa) appeared at around+0.10V.This new anodic peak(IIIa)even overlapped with the redox peak for reaction(4),as shown in Fig.2(c).The electrochemical behaviors were very similar to those observed at nitrogen-incorporated tetrahedral amorphous carbon electrodes(taC:N)reported by Miller and coworkers [46,47].

Experimental results shown in Figs.1and2indicated that there were differences in the redox processes of copper and chlo-

5910K.Shi et al./Electrochimica Acta52(2007)

5907–5913

Fig.2.Cyclic voltammograms of PGCE obtained by cyclic polarization with different voltammetric cycles.(a)One cycle;(b)three cycles;(c)nine cycles. Electrolyte:0.5mM CuCl2+0.1M NaCl(pH3.0).Scan rate:20mV/s.

ride ions at the electrodes obtained by different pretreatment procedures.Since the solution was of identical composition, the changes in voltammetric behaviors should be related to the difference in the surface characteristics.Different pretreatment means would produce different electroactive surfaces,resulting in different voltammetric responses,especially at higher extent of activation.

3.2.Voltammetric responses in copper(I)chloride solution

The PGCEs prepared by different activation procedures have been found to have different adsorption characteristics [3,18–20].Meanwhile,the newly deposited metal copper phase probably reacted with copper(II)species in solution to form cop-per(I)species[50].The anodic peak(IIIa)nearly overlapped with the redox peaks corresponding to reaction(4).This anodic peak might arise from the oxidation of the preferred accumula-tion of Cu(I)intermediate species at the new surface.The nature of the IIIa peak was investigated in NaCl solution containing Cu(I)Cl2?complex ions so as to limit the presence of copper(II) cations.

Fig.3shows the voltammetric responses of PGCE in1mM Cu(I)Cl2?species in0.1M NaCl(pH3.0).When the potential was scanned negatively from zero to?0.6V,metallic copper deposition was observed and a sharp deposition peak was appar-ent at?0.25V at the surface obtained by potentiostatic activation for90s,as shown in Fig.3(a).In the anodic scan,the?rst anodic peak was observed at?0.06V,with a peak current of 139?A.The second smaller anodic peak appeared at+0.14V. The reversible cathodic peak was observed at+0.06V.There were no particular differences in the voltammetric responses between Cu(I)and Cu(II)chloride species at PGCE obtained by potentiostatic

activation.Fig.3.Cyclic voltammograms of PGCE in1mM Cu(I)Cl2?+0.1M NaCl(pH 3.0).(a)PGCE obtained by potentiostatic activation of90s;(b)and(c)PGCE obtained by cyclic polarization of nine cycles.Scan rate:20mV/s.

Fig.3(b)shows that copper metal started to deposit from ?0.27V and the cathodic peak was subsequently observed at ?0.38V on the surface obtained by cyclic polarization for nine

cycles.In the reverse anodic scan,a very broad anodic peak was observed at around?0.12V and the new anodic peak(IIIa) appeared at+0.11V.However,when the potential was cycled between?0.1and+0.6V,only a couple of redox peaks for reaction(4)was observed at+0.05and+0.14V,as shown in Fig.3(c).Obviously,Cu(I)intermediates did not accumulate at the electrode surface.It indicated that the presence(or absence) of the new anodic peak(IIIa)was dependent on whether metallic copper had been deposited or not.Consequently,the new anodic peak(IIIa)should correspond to an abnormal oxidation process of copper metal deposits at the PGCE prepared by cyclic polar-ization,as compared to the normal oxidation processes observed at polished glassy carbon and metal electrodes.

3.3.Redox reactions of copper species at rotating-disc PGCE

A rotating-disc electrode(RDE)was employed to study the oxidation peak(IIIa)under slow scan rate to examine whether soluble or precipitate species were involved in the redox pro-cesses.Fig.4(a)shows the voltammogram of rotating-disc glassy carbon electrode,obtained by potentiostatic activation for90s,in a solution containing0.5mM CuCl2+0.1M NaCl (pH3.0).The?rst cathodic plateau current was observed from +0.03V and the current increased with increasing rotating rate. Metallic copper started to deposit at a potential more negative than?0.25V and a current loop was observed in the reverse scan.On the other hand,metallic copper usually deposited from zero potential at the same RDE in CuSO4+Na2SO4solution,

K.Shi et al./Electrochimica Acta52(2007)5907–5913

5911

Fig.4.V oltammetric responses of rotating-disc PGCE obtained by potentio-static activation for90s.(a)0.5mM CuCl2+0.1M NaCl(pH3.0);(b)0.5mM CuSO4+0.05M Na2SO4(pH3.0).Scan rate:5mV/s.Rotating rate:500rpm. as shown in Fig.4(b).It suggested that reduction of CuCl2?anion was involved during the early stage of nucleation.The application of rotating-disc technique at a slow scan rate of 5mV/s signi?cantly improved the mass transport of copper species to the electrode surface.The amount of metallic copper deposits increased signi?cantly,as evidenced by the presence of a thick brown copper?lm at the electrode surface.During the anodic process,the?rst anodic peak appeared at?0.04V, about50mV more positive than that observed in Fig.1(e).No oxidation peak relating to reaction(4)was observed within the potential range examined.The stirring of the rotating-disc elec-trode pushed the CuCl2?anion species away from the electrode surface.

Fig.5(a)shows the corresponding voltammogram for the rotating-disc PGCE obtained by cyclic polarization for nine cycles in CuCl2+NaCl solution.The?rst cathodic plateau cur-rent became apparent at+0.06V,and the deposition of metallic copper occurred at?0.31V.Correspondingly,metallic copper started to deposit from zero potential in CuSO4+Na2SO4solu-tion at the same RDE surface,as shown in Fig.5(b).In the reverse anodic scan,a small shoulder anodic peak appeared at?0.06V, while the broad anodic peak(IIIa)appeared at around+0.10V. Since the soluble CuCl2?species were forced to diffuse away from the RDE surface,the large anodic peak(IIIa)should be resulted from the oxidation of some precipitate species adhering to the rotating electrode surface,consistent with those observed in Fig.3.The above observations indicated that chloride ions were responsible for the abnormal voltammetric behaviors of metallic copper deposition at the PGCE obtained by cyclic polar-ization.Thus,these should be the differences in the interfacial structure,where the early nucleation of copper metal occurs, which cause the abnormal oxidation

behavior.Fig.5.V oltammetric responses of rotating-disc PGCE obtained by cyclic polar-ization for nine cycles.(a)0.5mM CuCl2+0.1M NaCl(pH3.0);(b)0.5mM CuSO4+0.05M Na2SO4(pH3.0).Scan rate:5mV/s.Rotating rate:500rpm.

3.4.Redox reactions of copper species at HOPG electrodes

As one type of sp2hybridized carbon,glassy carbon contains

both edge and basal planes.The different PGCE surface struc-

tures might be resulted from the difference in the ratios between

edge planes and basal planes.Highly oriented pyrolytic graphite

(HOPG)possesses the simplest structure of sp2hybridized car-

bon electrodes.Examination of the voltammetric responses of

the copper deposition reactions at HOPG electrode might pro-

vide useful information for the elucidation of the redox reactions

involved,especially for the nature of the anodic peak(IIIa).

Fig.6(a)and(b)shows the respective voltammetric responses

of the polished edge plane and the newly resurfaced basal plane

of HOPG electrode in1mM CuCl2+0.1M NaCl(pH3.0).Two

cathodic peaks and two anodic peaks were observed,similar

to those observed at the polished glassy carbon.The redox

wave corresponding to reaction(4)appeared to be reversible

at the edge plane,as shown in Fig.6(a).Copper deposition

was observed at around?0.36V for the edge plane and at ?0.39V for the basal plane.In the reverse anodic scan,the?rst anodic peak was observed at?0.06and?0.04V,respectively.

No anodic peak(IIIa)appeared at both surfaces.It indicated that

the anodic peak(IIIa)was not arisen from the difference in the

interfacial characteristics and structures of the two planes.

Based on semi-quantitative analysis,the edge plane and the

PGCE prepared by potentiostatic activation provided compara-

ble characteristics for redox reactions of copper chloride,includ-

ing low deposition overpotential,high electron transfer kinetics

rate for reaction(4)and normal oxidation process of metal-

lic copper deposits.It was consistent with our previous report

that potentiostatic activation created void space large enough

to expose both the edge planes and the electronically perturbed

5912K.Shi et al./Electrochimica Acta52(2007)

5907–5913

Fig.6.V oltammetric response of HOPG electrodes in1mM CuCl2+0.1M NaCl(pH3.0).(a)Polished edge plane and(b)the newly resurfaced basal plane. Scan rate:20mV/s.

basal planes(edge step)in the interior of the graphite oxide?lm [27,39].Interestingly,neither edge nor basal planes offered inter-facial characteristics resembling the PGCE activated by cyclic polarization,especially for the abnormal oxidation behavior.It suggested that there were changes in the fundamental structure of glassy carbon during cyclic polarization.

3.5.Effects of surface morphology on the redox behaviors

Experimental results indicated that different voltammetric behaviors were observed at glassy carbon electrodes obtained by different pretreatment processes and were likely related to the difference in the surface characteristics produced by electro-chemical activation.The observation of the abnormal oxidation behavior of metallic copper in the presence of chloride anion for PGCE obtained by cyclic polarization re?ected that this type of PGCE interface apparently had the lower distribution den-sity of electron transfer sites for the early nucleation.Cyclic polarization might cause changes in the surface structure of interwoven graphitic crystallites of glassy carbon throughout, while polishing or potentiostatic polarization means did not. Cyclic polarization was considered a weak activation means not able to remove carbon oxide effectively when compared with potentiostatic activation.Oxygen-containing groups pro-duced by cyclic polarization might accumulate at the whole graphitic crystallite surface and penetrate the crystallite to some extent with increasing extent of polarization.The accumulation of oxygen-containing functionalities might cause decreases in the distribution density of electron transfer sites.On the other hand,potentiostatic activation was an extensive erosion means capable of removing the excrescent carbon oxide at the graphitic crystallite surface to create larger void space without changing the basic surface structure of graphitic crystalline.4.Conclusions

In aqueous solutions containing chloride anions,the redox behaviors of copper species at glassy carbon electrode varied with the interfacial surface prepared by different electrochem-ical activation procedures.Carbon electrodes were exposed to extreme potentials for a longer time when potentiostatic activation was employed.The application of the more neg-ative potential of?1.0V resulted in more effective removal of surface oxides.Potentiostatic activation would generate oxygen-containing functionalities and maintain the basic surface structure of graphitic crystalline with high distribution density of electron transfer sites.On the other hand,cyclic polarization was considered a weak activation means not able to remove car-bon oxide effectively.Cyclic polarization might cause changes in the interwoven graphitic crystalline surface of glassy carbon, producing electrode interface with low distribution density of electron transfer sites for the early nucleation of metallic copper.

Acknowledgements

This work was partially supported by the Research Grants Council of Hong Kong(HKBU2049/01)and the Natural Sci-ence Foundation of Fujian Province,China(E0310006).

References

[1]G.M.Jenkins,K.Kawamura,Polymeric-Carbons,Carbon Fiber,Glass and

Carbon,Cambridge University Press,England,1976.

[2]A.Oberlin,in:B.A.Thrower(Ed.),Chemistry and Physics of Carbon,

Marcel Dekker,New York,1989.

[3]R.L.McCreery,in:A.J.Bard(Ed.),Electroanalytical Chemistry,vol.13,

Marcel Dekker,New York,1991.

[4]R.J.Bowling,R.T.Packard,R.L.McCreery,J.Am.Chem.Soc.111(1989)

1217.

[5]R.C.Engstrom,Anal.Chem.54(1982)2310.

[6]N.Cenas-Rozgaite,A.Pocius,J.Kulys,J.Electroanal.Chem.154(1983)

121.

[7]K.K.Shiu,F.Y.Song,H.P.Dai,Electroanalysis8(1996)1160.

[8]H.P.Dai,K.K.Shiu,J.Electroanal.Chem.419(1996)7.

[9]H.M.Wu,R.P.Riourg,P.Vernereau,J.Electroanal.Chem.162(1984)225.

[10]G.E.Cabaniss,A.A.Diamanties,W.R.Murrphy,R.Wlinton,T.J.Meyer,

J.Am.Chem.Soc.107(1985)1845.

[11]E.Hershenhart,R.L.McCreery,R.D.Knight,Anal.Chem.56(1984)2256.

[12]T.G.Strein,A.G.Ewing,Anal.Chem.63(1991)194.

[13]N.M.Pontikos,R.L.McCreery,J.Electroanal.Chem.224(1992)229.

[14]H.Zhang,L.A.Coury,Anal.Chem.65(1993)1552.

[15]G.N.Kamau,W.S.Wills,J.F.Rusling,Anal.Chem.57(1985)545.

[16]I.F.Hu,D.H.Karweik,T.Kuwana,J.Electroanal.Chem.188(1985)59.

[17]R.M.Wightman,M.R.Deakin,P.M.Kovach,W.G.Kuhr,K.J.Stutts,J.

Elechem.Soc.131(1984)1578.

[18]K.K.Shiu,K.Shi,Electroanalysis10(1998)959.

[19]K.K.Shiu,K.Shi,Electroanalysis12(2000)85.

[20]K.Shi,K.K.Shiu,Electroanalysis13(2001)1319.

[21]P.K.Upadhyay,J.Electroanal.Chem.271(1989)339.

[22]R.L.McCarley,E.A.Irene,R.W.Murray,J.Phys.Chem.95(1991)2492.

[23]R.C.Engstrom,V.A.Strasser,Anal.Chem.57(1985)136.

[24]Y.C.Liu,R.L.McCreery,J.Am.Chem.Soc.67(1995)3115.

[25]T.C.Kuo,R.L.McCreery,Anal.Chem.71(1999)1553.

[26]I.F.Hu,T.Kuwana,Anal.Chem.58(1986)3235.

[27]K.Shi,K.K.Shiu,Anal.Chem.74(2002)879.

[28]D.M.Anjo,M.K.M.Khodahakhsh,S.Nowinski,M.Wonger,Anal.Chem.

61(1989)2603.

K.Shi et al./Electrochimica Acta52(2007)5907–59135913

[29]L.J.Kepley,A.J.Bard,Anal.Chem.60(1988)1459.

[30]C.Barbero,R.Kotz,J.Electrochem.Soc.140(1993)1.

[31]T.Nagaoka,T.Yoshino,Anal.Chem.58(1986)1037.

[32]T.Nagaoka,T.Fukunaga,T.Yoshino,J.Electroanal.Chem.217(1987)

453.

[33]T.Nagaoka,T.Sakai,K.Ogura,T.Yoshino,J.Chem.Soc.,Faraday Trans.

183(1987)1823.

[34]T.Nagaoka,T.Fukunaga,Anal.Chem.60(1988)2766.

[35]T.Nagaoka,Y.Uchida,K.Ogura,Anal.Chim.Acta222(1989)127.

[36]T.Nagaoka,Y.Uchida,K.Ogura,Anal.Chim.Acta220(1989)269.

[37]T.Nagaoka,Y.Uchida,K.Ogura,J.Chem.Soc.,Faraday Trans.185(1989)

3757.

[38]T.Nagaoka,M.Katayama,M.Fujimoto,H.Nakao,K.Ogura,Anal.Chim.

Acta285(1994)135.

[39]K.Shi,K.K.Shiu,J.Electroanal.Chem.574(2004)63.

[40]K.Shi,K.K.Shiu,Electrochim.Acta51(2005)2631.

[41]L.Bodalbhai,A.Brajter-Toth,Anal.Chim.Acta231(1990)191.[42]R.Srinivasan,P.Gopalan,Surf.Sci.338(1995)31.

[43]C.Nila,I.Gonz′a lez,J.Electroanal.Chem.401(1996)171.

[44]A.Ramos,M.Miranda-Hernandez,I.Gonz′a lez,J.Electrochem.Soc.148

(2001)C315.

[45]R.Drissi-Daoudi,A.Irhzo,A.Darchen,J.Appl.Electrochem.33(2003)

339.

[46]K.Yoo,https://www.360docs.net/doc/0f1028718.html,ler,X.Shi,R.Kalish,J.Electrochem.Soc.148(2001)

C95.

[47]J.J.Lee,https://www.360docs.net/doc/0f1028718.html,ler,X.Shi,R.Kalish,K.A.Wheeler,J.Eletrochem.Soc.

148(2001)C183.

[48]R.Bowler,T.J.Davies,M.E.Hyde,https://www.360docs.net/doc/0f1028718.html,pton,Anal.Chem.77(2005)

1916.

[49]D.Krznaric,T.Goricnik,Langmuir17(2001)4347.

[50]G.G.L′a ng,M.Ujv′a ri,G.Hor′a nyi,J.Electroanal.Chem.522(2002)179.

[51]S.Fletcher,C.S.Halliday,T.Lwin,G.Nelson,J.Electroanal.Chem.159

(1983)267.

[52]M.V.Mirkin,A.P.Nilov,J.Electroanal.Chem.283(1990)35.