Electrochemical Impedance of Ethanol Oxidation in Alkaline Media

CHEM. RES CHINESE UNTVERSITTES 2012, 28(1), 19--25Electrochemical Impedance of Ethanol Oxidation inAlkaline mediaDANAEE Iman, JAFARIAN Majid, GOBAL Fereydoon, SHARAFI Mahboobeh' andMAHJANI Mohammad-ghasemL Abadan Faculty of Petroleum Engineering, Petroleum University of Technology, Abadan 63165-619, Iran;2. Department of Chemistry, K. N. Toosi University of Technology, Tehran 15875-4416, iran3. Department of Chemistry, Sharif University of Technology, Tehran 11365-9516, iranAbstract Nickel modified Niooh electrodes were used for the electrocatalytic oxidation of ethanol in alkalinelutions. The electro-oxidation of ethanol in a 1 molL Naoh solution at different concentrations of ethanol was stu-died by ac impedance spectroscopy. Electrooxidation of ethanol on Ni shows negative resistance on impedance plotsThe impedance shows different patterns at different applied anodic potential. The influence of the electrode potentialon impedance was studied and a quantitative explanation for the impedance of ethanol oxidation was given by meansof a proposed mathematical model. At potentials higher than 0.52 V(vs. Ag/AgCI), a pseudoinductive behavior wasobserved, but at those higher than 0.57 V, impedance pattems were reversed to the second and third quadrants. Theconditions required for the reversing of impedance pattern were delineated with the impedance modelKeywords Impedance; Equivalent circuit; Ethanol; Electrocatalytic; NickelArticle ID10059040(2012)-01-019071 Introductiontials, aiming at the elucidation of its reaction mechanism. Theanalysis of the theoretical impedance function provides imporIn the last decade, researches on the electrooxidation of tant information on the kinetic parameters. This informationsmall organic molecules have attracted considerable attentionallows the ElS spectrum simulation and therefore, predicts thedue to the development of direct liquid fuel cells, which re- system behavior with regard to the variation of the overpotenquires highly reactive fuels with high energy density and lowtoxicity]one of the most promising fuels due to its low toxicity, high 2 Methods and Materialsenergy density, biocompatibility and abundant availability)The major challenge for the utilization of ethanol as fuel is itsSodium hydroxide and ethanol used in this work werelow reactivity in the temperature range technically feasible forMerck products of analytical grade and were used without further purification. Doubly distilled water was used throughout.Some of the data on ethanol oxidation in an alkaline elec-Electrochemical studies were carried out in a conventionaltrolyte on platinum have been presented in the work!l in three electrode cell powered by an electrochemical systemwhich the authors observed the strong passivation of platinum comprising of an EG&G model 273 potentiostat/galvanostatand came to the conclusion about the necessity of developingand a Solartron model 1255 frequency response analyzer. Themore acceptable catalyst than platinum. In the furthsystem was run by a personal computer(PC)through M270 andwork(, 13). ethanol oxidation was studied in I mol/L KOHM389 commercial softwares via a general purpose interfacetitanium promoted by Ruo,/Ni system and glassy carbon bus(GPIB)interface. A frequency range of 100 kHz to 15 MHzcoated by RuNi nanoparticles. The activity of these catalystsand a modulation amplitude of 5 mv were employed for im-was low, and the starting potential of alcohol oxidation was pedance measurement. Fitting of experimental impedanceclose to 0.9V vs the reversible hydrogen electrode(RHE)spectroscopy data to the proposed equivalent circuit was doneElectrochemical impedance spectroscopy(EIS)is a good by means of home written least square software based on thetool to analyze the kinetics of electrode reactions. The advan- Marquardt method for the optimization of functions and Mac-tage of EIS over direct current(DC) techniques is that this donald weighting for the real and imaginary parts of the imsteady-state technique is capable of probing relaxation pheno-pedance 20 I. A dual Ag/AgCl-saturated KCl, a Pt wire and amena over a wide frequency range!+9)ni- ckel disk electrode were used as the reference, counter andThe purpose of this work is the analysis of impedanceorking electrodes, respectively. All the studies were carriedcharacteristics of electrooxidation of ethanol on Ni electrode in out at (298+2)Kelectrode was a pure nickel rod中国煤化工nt, the surfaceNaOH solution at different ethanol concentrations and poterCNMHG*Corresponding author. E-mail: mjafarian @kntu ac irReceived April 21, 2011; accepted August 30, 2011CHEM. RES CHINESE UNTVERSITIESVol 28pretreatment of working electrode was performed by hand poli-shing of the electrode surface with 0.05 mm alumina powder on目04a polishing microcloth and rinsed thoroughly with doubly dis-0tilled water prior to modification.0.1030.53 Results and discussionFig. I presents the consecutive cyclic voltammograms(Cv)of a nickel electrode in 1 mol/L Naoh solution recorded at a0potential sweep rate of 100 mV/s. In the first sweep, a pair ofredox peaks appeared at 492 and 429 mV vs. Ag/AgCl that areassigned to the Ni"/Ni redox couple according toFig2 Cyclie voltammograms of Ni electrode inNi(OHh+OHNiOOH+H2O+eI mol/L NaOH solution in the absence(a)In the subsequent cycles, both the anodiand presence(b)of 0.7 mol/L ethanolaks shifted negatively, stably pointing to higher ener.The potential sweep rate was 10 mV/s and the insetgies(potential)required for the nucleation of NiooH in the firstis the initial potential of ethanol oxidationcycle. The enhanced base line current of the first cycle wasThe electrocatalytic oxidation of ethanol occured not onlyassociated with the oxidation of Ni to Niin the anodic half cycle but also continued in the initial stage ofthe cathodic half cycle. Ethanol molecules adsorbed on the nispecies were oxidized at higher potentials, which was parallelto the oxidation of Nito Ni*species. The later process wasin consequence of decreasing the number of sites for ethanoladsorption as the poisoning effect of the products or interme-diates of the reaction tends to decrease the overall rate of etha-nol oxidation. Thus, the anodic current passed through a ma-ximum as the potential was anodically swept. In the reversehalf cycle, the oxidation continued and its cor0.4rent went through the maximum due to the regeneration ofNi(II)species that are active sites for the adsorption of ethanolFig 1 Consecutive cyclic voltammogram of Nias a result of the removal of adsorbed intermediates and pro-oxidation in 1 mol/L NaoH comprised ofducts. Surely, the rate of ethanol oxidation as signified by thefirst(a) and fiftieth (b) cycles at a scan rateanodic current in the cathodic half cycle dropped as the unof 100 mv/sfavorable cathodic potentials were approachedThe current grew with the number of potential scans,Cyclic voltammograms of Ni electrode in the presence ofdicating the progressive enrichment of the accessible electroac-0.5 molL ethanol at various potential sweep rates in a rangetive species Ni and Ni on or near the surface. After prolong 2-500 mV/s are illustrated in Fig 3. The cathodic peak wascycling, the redox peak's potentials were stabilized at 472 and not observed at low scan rates, but was observed upon increa-432 mV(vs. Ag/AgCI)and a shoulder developed on cathodic sing the sweep rate. At higher scan rates, a new oxidation peakpeak at around 350 mV. The changes of the peaks'position and appeared for nickel oxidation at a lower potential than that ofalso the creation of a new reduction peak were likely due to the the oxidation of ethanol. These phenomena indicate that thechanges in the crystal structures of the nickel hydroxide and the electrooxidation of nickel species to higher valence state isnickel oxyhydroxide constituents ofe electrochemicallymuch faster than the catalytic oxidation of ethanol. This revealsformed surface film 22). It has been reported23, 24) that at theinitial stage of electro-oxidation, a-Ni(OH)h forms that is fur-ther slowly converted to the B-Ni(oH) formFig 2 shows cyclic voltammograms of Ni electrode in1 moV/L NaoH solution in the absence(a) and presence(0. 7 mol/L ethanol at a potential sweep rate of 10 mV/s. At Nielectrode, the oxidation of ethanol appeared as a typical elec-trocatalytic response. The anodic charge increased with respeto that observed for the modified surface in the absence ofethanol, and it was followed by decreasing the cathodic chargeupon increasing the concentration of ethanol in solutionThe decreased cathodic current that ensured the oxidation中国煜仁 Ni electrodeprocess in the reverse cycle indicated that the rate determiningof 0.5 mol/Lstep certainly involved ethanol, and that it was incapable ofCNMHGreducing the entire high valent nickel species formed in theSweep rates from a to L: 2, 5, 10, 20, 30, 40, 50, 75, 100,oxidation cycle200,350and500mv/sDANAEE Iman et aithat the oxidation of ethanol on Ni may belong to a slow frequency regions. The depressed semicircle in high frequencyprocessregion can be related to the combination of charge transferne Fig 4(A)shows the Nyquist diagrams of nickel electrode resistance and the double layer capacitance. The charge transfercorded at 0.59 v(vs. Ag/AgCI)in a concentration range of characteristic appears in the first quadrant, two loops in the0.1-0.7 molL of ethanol. The Nyquist diagrams consist of regions of medium and low frequencies are located in thethree slightly depressed semicircles in high, medium and low second and the third quadrants. Fig 4(B)shows the Nyquistdiagrams of nickel electrode recorded at 0.59 V(vs. Ag/AgCI)1000without ethanol. As can be seen, the Nyquist diagram consistsof one time constant in the first quadrant due to the chargetransfer resistance of nickel(Il) oxidation in parallel with2800double layer capacitance.Fig 5 shows Nyquist plots of the impedance of ethanolnidation at different potentials in 0.7 molL ethanol. At1500the potential below 0.51 V(vs. Ag/AgCI), two large depressed2000capacitive semicircles are observed, revealing a slow reactionZR/Qrate of ethanol oxidation[ Fig. 5(A)]. The semicircles are due to3500}(B)the charge transfer resistance in high frequency region and the3000intermediates adsorption in low frequency regions. Bode phaplots for the same sample are shown in Fig. 6(A). Two over2500lapped peaks are observed in the Bode plots corresponding totwo depressed semicircles on the Nyquist plot. The equivalentcircuit compatible with the Nyquist diagram is depicted inig. 7(A). To obtain a satisfactoryethanol electrooxidation, it is necessary to replace the capacitor,C, with a constant phase element(CPE)@ in the equivalent02000400060008000circuit251. In this electrical equivalent circuit, Rs, CPea and RaZR/Qrepresent solution resistance, a constant phase element corres-Fig, 4 Nyquist diagrams of Ni electrode in the pre ponding to the double layer capacitance and the charge transfersence(A) and absence(B) of different concen-resistance, respectively. CPEads and Rads are the electrical eletrations of ethanol in 1 mol/L NaOHments related to the adsorption of reaction intermediates.DC potential is 0.59 V(vs. Ag/AgCI).(A)c(Ethanoly(molL ): a 0.1;To corroborate equivalent circuit, the experimental datab. 0.3; c 0.5; d 0. 7. Inset of (A): high frequency semicircles.were fitted to equivalent circuit and the circuit elements were200(4)400(C)180800-120100020003000Z/Q2Zo/QFig s Experimental Nyquist diagrams as a function of applied potential for ethanol electrooxidation on Ni electrodein 0.7 molL ethanol at different potentialsPotentials/V(w. Ag/AgCl):(A)a.049;b.0.50c.0.51;(B)a.0.53;b.0.54;c0.55:(Ca0.57;b.0.58;c.0.5920-80Ig(w/Hz)g(w/Hz)中国煤化工Fig 6 Experimental phase shift plots as a function of applied potential forCNMHGelectrodein 0.7 molL ethanol at different potentialsPotentials/v(v, AgAgCl:(A)a.049b.0.50c0:51;(B)a.053:b.0.54;c0.55:(C)a0.57;b.0.58;c.0.5922CHEM. RES CHINESE UNTVERSITIESVol 28from Table 1, increasing potential decreases the diameters ofthe two semicirclesAt the potential between 0.53-0.55 V, a typicdoinductive behavior was observed, as shown in Fig. S(B).here large semicircles at high and medium frequenciesaccompanied by a small arc in the forth quadrant at low fre-quency with all the diameters decreasing sharply with increa-sing potential This inductive behavior occurs for the relaxationphenomenon characteristics of the generation of further activeQ2sites5, I6 and the further adsorption of electroactive constituents, ethanol,on active sites 6. The negative peak is also ob.served on the phase shift plot due to inductive behavior[ Fig. 6( B)]. As potential arrives at 0.57 V, a change of the im-pedance plots happens where the two loops in the regionsmedium and low frequencies reverse to those in the second andFig7 Equivalent circuits compatible with the expe-the third quadrants due to the passivation of electrode sur.rimental impedance data in Fig. S(A)andface, IS), as shown in Fig. (C). The phase shift(Fig. 6(C))ofFig 6(B)for ethanol electrooxidation on Niexperimental impedance data shows an abrupt jump betweenelectrodethe positive and negative values of phase angle, indicating theobtained Table 1 illustrates the equivalent circuit parameters change of the rate determining step of elecrooxidation of etha-for the impedance spectra of ethanol oxidation. As can be seen nol on Ni in a potential range of 0.49-0.59VTable 1 equivalent circuit parameters of electrooxidation of 0.7 mol/L ethanol on Ni electrode in1 mol/L NaOH solution obtained from Fig 4n23208.43.517.22.00808.00.538.03.010.l0.786.0100.780.585.0The equivalent circuit compatible with the Nyquist diaNi*+Intermediategram recorded at a potential higher than 0.53 V is depicted inFig.7(B). Table 1 illustrates the equivalent circuit parameterNia--ethanolk4 ni+-Intermediate+efor the impedance spectra of ethanol oxidation at differentNi*-Intermediate-Ni*-Products+e (6)lied potentialsAcetaldehyde CH3 29,30), CH3 Co and Co3l-33)haveAs previously mentioned the first depressed semicircle in been reported as the intermediates in the oxidationthe entire potential range is due to the charge transfer resistanceEquations (3)and (4)are according to Fleischmann me-and the inductive behavior at higher potentials is due to the chanism 26,27 and in Equations(5)and(6), Ni*is used as ac-adsorption of ethanol. The EIS results indicate that the ethanol tive surface for ethanol oxidation. Observation of a high currentelectrooxidation on Ni catalyst at various potentials shows density in the presence of ethanol in comparison to the Ni(ohdifferent impedance behavior.is according to Equations(5)and(6). On this basis the kineticsIn order to analyze the reaction mechanism of ethanol of the oxidation reaction can be expressedelectrooxidation on Ni catalyst, according to the high currentdensity in the presence of ethanol in comparison to the Ni(OH)=2(4-a)sp1-2cpwe assume that part of the current is due to ethanol oxidationby Niooh due to the disappearance of the NiOOh reduction2=kc(1-)peak in the negative sweep/26, and part of the current is due tov2=360ethanol oxidation on the surface of oxide layer. The redox tran-sition of nickel species present in the filmv=kCKN(1-0,)exp(10Ni*+e(2)中国煤化工and ethanol is oxidized on the modified surface via the following reaCNMHGNi+EthanolNi*+Intermediatehere k(i-l, -1, 2, 3)is rate constant, bF1, 2, 3)the TafelDANAEE Iman et al23slope, 0, the fractional coverage of Ni, 0 the surface co-ae aeverage of reaction intermediates, Cg the ethanol concentrationa0 a8dEand v, the reaction rate. Based on the proposed mechanism, theFaradaic current density(lf)can be written asIF(vr+v4+vs)Fa/aIIn the steady state, the Faradaic current density(F)Eethanol electrooxidation can be expressed as a function of elec-trode potential and two state variable, fractional coverage of66Ni(e.) and the surface coveBof reaction interme-0八八(a人aediates(0)IF =f(E, O, 0)(13)The ratio AIf/AE is defined as the Faradaic admittanceY (inverse of the Faradaic impedance), thusIn accordance with the above when ar0 andaJnY△R八(a人△E八a八E(14)aE/<0, positive resistance and when gE0, negative resistance both appear in the Nyquistximation of Tailor's series expansion around the steadythe Faradaic admittance can be written asdiagrams. On the basis of reaction mechanism, theoretical im-pedance was obtained[ Equation( 16)] and the kinetic constant is-(R丿+(B(15) determined by fitting the model to the experimental Nyquistdiagram(Table 2)whereTable 2 Values of rate constant calculated from Equation(16) and Fig 4 for electrooxidation of 0.7 moVL ethanolon Ni electrode in l molL NaOH solution10-k/mol-cm-'-s- 10kI/(moI-cm2s-) 10k/(mol-cm2-s-) 10 k /(mol-cm2s- 10 k/(mol-cm"")10k/(mol-cm"")19531,,+waoa j+wao 41-a,wa,i(16) diates as a function of potential can be obtained from the model4=-KN-CR-KONe-R, CRN-k, enapof ni+ and the intermediates rise. AlNi coverage approaches unity at a higher applied anodica2=-2-2k1+kC(-)-4a3=kC2(1-日)-+C1(-日)-k4Fn-w(G=风C-k8÷kC9(-)b0.4ba=C→+k022k(-6)2k马=-2-2k+kC(1-日)+kFig8 Calculated surface coverages of Ni(a)and in-a Fcas aFa,a, a, F'a,a, asF'a2astermediate(b) for 0.7 moUL ethanol electro-91929192oxidation on Ni electrodeFa. aFa92(i In the low potential region(0.5 V), assumed reaction(S)上is rate-determining step and vs0q1 q2F(av./aE-a,/d the impedance da2+w'aj中国煤化工where qi and q2 are charges required for complete Ni"or com- show capacitiveplete adsorption of intermediate on unit surfacemicircles appear。CNMHGcapacitive se-The corresponding coverages of Niand the interme- tion with one adsorbed intermediate! 5. The EIS data are in lineCHEM. RES CHINESE UNIVERSITIESVol 28equivalent circuit preseanalyses from EIS indicate that in the low potential region, equivalent circuit(Table 1), it is observed that with the otherreaction(5)might be rate-determining step, meaning that the elements remaining positive, the only parameter that will causeoxidation of intermediates is fast in comparison with the gene. the reversal of impedance pattern to those in the second and theration of intermediatesthird quadrants is R2, the value of which is determined by elec-(i)When ethanol is electrooxidized in intermediate poten- trode potential. At potentials lower than 0.54 V(vs. Ag/AgCI)tial range(0.54 V), increasing the potential enhances the rate of R2 decreases with the increase of applied anodic potential but atreaction (5),but not enough to exceed that of reaction(6). In higher potentials R2 jumps to very negative values. Furtherthis case the rate-determining step of ethanol electrooxidation increase of potential will lead to the decrease of absolute valueis in a transition region. Thus(al /00 ) <0,(ae /aE <0 and of R2. By comparing the potential dependence of simulatedimpedance patten, it is found that if R2 is positive, the impe-it can be concluded that(alF/a0 (ae /aE >0. The inductive dance will show the patten similar to Fig. 5(A)and(B).If R2 isehavior observed in our experiments and that shown in negative, the reversing impedance pattern, as in Fig. S(C), carFig. (B)is predicted. The Eis data can be simulated using the be observed. The other element in equivalent circuit remainsequivalent circuit presented in Fig.7(B). In general, the condi- positive and decreases with increasing applied potentialtion of the occurrence of an inductive behavior in the NyquistAccording to the mechanism of the electrooxidation on Niplot is(a/a0 )(ae/aE]>0. This implies that if the variation electrode proposed by Fleischmann et al. 26), no negative resis-of the electrode potential causes a variation of the Faradaic tance was obtained in the derivation of impedance equationcurrent density not only through its effect on the strength of thTherefore the mechanism proposed by Fleischmann et al. is noelectric field in the double layer but also through its effect on complete for the electrooxidation of ethanol on Ni surface. Theanother variable and both the effects act in the same direction, theoretical impedance diagrams obtained accordingthen an inductive component is expected in the impedance posed electrooxidation mechanism are in agreement with nega-pattern). According to impedance parameters(al/a0 )andtive resistance observed on the ex(ae,/aE), inductive behavior in ethanol electrooxidation re-therefore the proposed mechanism is a complete mechanism forthe electrooxidation of ethanol on Ni electrodeveals that the coverage of the intermediates decreases withincreasing potential, leading to an increase of Faradaic curent. 4 ConclusionsApparently, with increasing potential, large amounts of Iareformed on the electrode surface that react with intermediates toThe nickel oxide film was formed electrochemicallydecrease their coverage and also higher active sites are availanickel electrode by cyclic voltammetry and tested for elecble for reactions(5)and (6). Meanwhile, decreasing the surfacetrooxidation of ethanol in alkaline media, Electrochemical im.coverage of intermediates will contribute to the adsorption of pedance studies of ethanol oxidation on Ni electrode demonethanol on free sites which subsequently enhances the Faradaicstrate the potentialities of this method as a tool for investigatinghe mechanism of ethanol oxidation. Different impedance pat-current. So, in the intermediate potential range,crease of potential, the transition from capacitive behavior toinductive behavior indicates that the rate-determining step is nol on Ni shows negative resistance on impedance plots. Thechanginimpedance data analyzed at different potentials show evidence(iii) In the high potential range(0. 58 V), reaction(6)for two processes occurring at the interface: one is associatecan be assumed as rate-determining step k>>k and vs>v6.with the ethanol electrooxidation leading to intermediates formation on the surface, and the other is assigned to the oxidationIn this case (al/0 )<0,(a0/aE>0 and thus of intermediates. A theoretical impedance model based on ki(a /00 a0/aE <0. The capacitive arc at the intermediate netics is proposed which captures and explains all the featuresfrequencies of the Nyquist plot will flip to the second and theof the potential dependence ofthird quadrants with the real component of the impedance be-impedance behavior in different potential regions reveals thatcoming negative. This means that the passivation of electrodethe mechanism and rate-determining step in ethanol electro-surface has occurred. The eis data can also be simulated viaoxidation vary with potential. In the low potential range, ethathe equivalent circuit of Fig.7(B). Melnick et al. 536l indicatedoxidation is the rate-deterthat the passivation of the Pt electrode during methanol elec-potential range, the oxidation and removal of adsorbed inter-trooxidation is probably due to the reversible formation ofmediates become the rate-determining step. Meanwhile, atoxide species. Meanwhile, with reaction (6) being theintermediate potential, the rate-determining step in ethanolrate-determining step, the oxidation of intermediates with Nielectrooxidation is in transition range. The theoretical impeis much slower than the generation of intermediates by reactionce diagrams obtained according to our proposed electro-5)and the passivation at higher potentials can be explainedoxidation mechanism are in agreement with experimental im-the formation of a large amount of intermediates on the surface pedance plots.中国煤化工of the catalyst, Therefore, the adsorption of ethanol is inhibitedCNMHGdue to an increase of the coverage of intermediates and theReererenceselectrooxidation rate shows almost no significant increase. As[l] Wang H T, Yin Y. C Gao S J, Chem. Res. Chinese Universities,No.DANAEE Iman et al2008,24(5)632[19] Zhuang Q C, Fan x.Y, XuJ M, Wei G. Z, Dong Q. F, Sun S G,[2] Xu D, Wang Y, Zhang Y, Zhang G, Shao K, Li S w. Na H,Chem.Chem. Res Chinese Universities, 2008, 24(4). 511Res. Chinese Universities, 2010, 26(6). 1031[20] Danace L, Jafarian M, Forouzandeh F, Gobal F, Mahjani M. G,[3]Wang Z.B., Yin G P, Shi P F, Carbon, 2006, 44, 133[4] Spinace E. V, Neto A. O, Linardi M, J. Power Sources, 2003, 124, [21] Jafarian M, Gobal F, Danaee L, Mahjani M. G, Electrochim. Acta,2007,52,5437[s]Zhou w.J. Song S. Q, Li w. Z, Zhou Z. H, Sun G. Q Xin Q- [22] El-Shafei A. A, J. Electroanal. Chem., 1999. 471, 89Douvartzides S. Tsiakaras P. J. Power Sources, 2005. 140. 50[23] Danaee I, Jafarian M, Mirzapoor A, Gobal F, Mahjani M G[6]Wang J, Wasmus S, Savinell R F, J. Electrochem. Soc, 1995, 142Electrochim. Acta. 2010. 55. 20934218[24] Hahn F, Beden B, CroissarLamy C, Electrochim. Acta[7 Fujiwara N, Friedrich K A, Stimming U, J. Electroanal. Chem.1999,472,120[25] Maritan A Toigo F, Electrochim Acta, 1990, 35, 141[8] Lamy C, Belgsir E M, Leger J M, J Appl. Electrochem, 2001, 31, [26] Fleischmann M, Korinek K, Pletcher D, J. Chem Soc. PerkinTans,1972,2,1396[9]Ke X, Deng LL, Shen P K, Cui G. F. Chem. Res. Chinese [27] Fleischmann M, Korinek K, Pletcher D, J.Electroanal.Chem.Universities, 2010, 26(3)4431971,3,39[10] Wang J, L, Liu Z M, Cao H. Y, Gong M. C, Chen Y D, Chen Y. [28] Danaee L, Jafarian M, Forouzandeh F, Gobal F, Mahjani M. G, IntQ, Chem Res Chinese Universities, 2009, 25(1)81J. Hydrogen Enengy, 2008, 33, 436[11] Park S M, Chen N C, Doddapaneni N, J. Electrochem. Soc., 1995,[29] Ribeiro J, Purgato F L S, Kokoh K. B, Leger J M, de Andrade A[12]Kim J W, Park S M, J. Electrochem. Soc., 1999, 146, 1075[30] Vigier F, Coutanceau C, Hahn F, Belgsir E. M, Lamy C.,J.[13] Kim J. w Park S. M, J. Electrochem. Solid State Lett, 2000, 3, 385Electroanal Chem. 2004.. 81[14] Amstrong R D J. Electroanal. Chem., 1972, 34, 387[1] Liang Z. X, Zhao T. S, Xu J B, Zhu L. D, Electrochim. Acta, 2009[15] Amstrong R. D, Henderson M, J. Electroanal. Chem., 1972, 39, 8154,2203[16] Seland F. Tunold R, Harrington D. A, Electrochim. Acta, 2006, 51, [32] Liu C. W, Chang Y. W, Wei Y C Wang K W, Electrochim. Acta,201,56,2574[17] Danaee L, Jafarian M, Forouzandeh F, Gobal F, Mahjani M. G, int [33] Lima F H. B, Gonzalez E R, Electrochim. Acta, 2008, 53, 2963J. Hydrogen Energy, 2009, 34, 859[34]Cao C N, Electrochim Acta, 1990, 35, 83[18] Danae L, Jafarian M, Forouzandeh F, Gobal F, Mahjani M. G.,J. [35] Bagotzky V.s. Vassilyew Y B, Electrochim. Acta, 1967, 12, 1323Phys. Chem. B,2008,1l2,15933[36] Melnick R E, Palmore G. T. R,. Phys. Chem. B, 2001, 105, 9449H中国煤化工CNMHG