The Beneficial Role of the Cometals Pd and Au in the ... · ethanol oxidation in alkaline solutions...

Published: July 20, 2011

r 2011 American Chemical Society 15324 dx.doi.org/10.1021/jp200318m | J. Phys. Chem. C 2011, 115, 15324–15334

ARTICLE

pubs.acs.org/JPCC

The Beneficial Role of the Cometals Pd and Au in theCarbon-Supported PtPdAu Catalyst Toward Promoting EthanolOxidation Kinetics in Alkaline Fuel Cells: Temperature Effect andReaction MechanismJayati Datta,*,† Abhijit Dutta,† and Sanjeev Mukherjee‡

†Electrochemistry & Fuel Cell Laboratory, Department of Chemistry, Bengal Engineering and Science University, Shibpur,Howrah�711 103, West Bengal, India‡Department of Chemistry and Chemical Biology, 317 Egan Center, Northeastern University, 360 Huntington Avenue, Boston,Massachusetts 02115, United States

bS Supporting Information

1. INTRODUCTION

The Pt-based alloys and composites are being explored as thealternatives to single Pt for electrocatalysis in fuel cell reactions.1�8

Particularly for the anodic reaction in DEFCs operating in acidicenvironments, several Pt-based alloys have superseded the singlePt with respect to their catalytic activity toward promoting fasterkinetics and complete conversion of the alcohol fuel to the endproducts. The success in electrocatalysis of these alloy materialshas been attributed to the bifunctional mechanism9 and theelectronic effect10�12 induced by the electronic interaction of Ptwith other metals. However, the search for efficient catalysts forselective oxidation in alkaline solutions is scarce. Therefore, amajor thrust may be put to investigate the catalytic activity of thealloy materials in alkaline media for obtaining favorable kineticsand better selectivity toward CO2 production.

13�16

In the present investigation, the carbon-supported singlemetal catalysts, Pt/C, Au/C, and Pd/C, and the ternary

Pt:Pd:Au/C [∼1:1:1] catalysts were synthesized through a bor-ohydride reduction method. The identification of crystal phases,the matrix morphology, and the surface area of the nanoparticleshave been studied by X-ray diffraction (XRD), transmissionelectron microscopy (TEM-EDAX), and the Brunauer�Emmet�Teller (BET) method, respectively. The catalytic activity towardethanol oxidation in alkaline solutions was investigated with thehelp of several electrochemical techniques, such as cyclic voltam-metry (CV), chronoamperometry (CA), and electrochemicalimpedance spectroscopy (EIS), and the temperature effect onethanol oxidation was investigated with Pt/C and PtPdAu/Celectrodes in particular. Apparent activation energies (Ea(app)) forthe oxidation reaction were determined at different potentials

Received: January 11, 2011Revised: April 23, 2011

ABSTRACT:Electrochemical investigations have been carried out to study the oxidationkinetics of ethanol in alkaline solution on carbon-supported ternary alloy catalystsPt�Pd�Auwithin the temperature range of 20�80 �C. To derive a better understandingof the contribution of each of the metallic components toward the catalytic oxidation ofethanol, some of the investigations were extended to the individual noble metals forcomparison, however, at a single temperature (20 �C). The individual metals could barelyshow their catalytic efficiency toward ethanol oxidations when compared to the alloyedcatalyst. The ternary catalyst exhibited much lower values and a larger temperaturedependence of onset potential for ethanol oxidation. With the rise of potential, theapparent activation energy (Ea(app)) for ethanol oxidation on the Pt/C electrodeincreased, whereas a decreasing trend was observed with the Pt30Pd38Au32/C electrode.It was suggested that the Pt30Pd38Au32/C electrode bears an excellent tolerance towardethanolic residues, for the temperature range studied. In correlation with the resultsobtained from the above study, attempts were made to elucidate the oxidation reaction mechanism, and this further evoked interestin extending the work to the estimation of products formed during oxidation of ethanol within the same temperature range throughion chromatographic analysis. The pronounced increase in the quantity of oxidation products, such as acetate and carbonate,obtained over the ternary catalyst as compared to single Pt, substantiates the kinetic enhancement of ethanol oxidation, attributableto the cometal partnership between Pd and Au when incorporated in the Pt matrix. In summary, the multimetallic nanocrystallitescan not only show their capability of extracting the best possible number of electrons from the alcohol fuel in alkaline solutions,harnessing more energy, but also, at the same time, bring down the cost of the catalyst material by reducing the Pt content to aconsiderable extent.

15325 dx.doi.org/10.1021/jp200318m |J. Phys. Chem. C 2011, 115, 15324–15334

The Journal of Physical Chemistry C ARTICLE

within the temperature range of 20�80 �C. The reaction productswere also estimated through ion-chromatographic analyses within thesame temperature regime. On the basis of the information obtained,the kinetics and mechanism of the reaction were elucidated.

2. EXPERIMENTAL SECTION

2.1. Preparation of Catalysts. The nanocatalysts (40 wt %each) of the single metals, Pt, Au, and Pd, and the ternary systemPtPdAu/C were prepared under the NaBH4-reduction schemefrom the chemical bath containing the corresponding precursorsalts (H2PtCl6, PdCl2, AuCl4), Arora Matthey Ltd. and theirmixtures in 0.002 M concentrations of each of the salts. Anappropriate amount of Vulcan XC-72 carbon (Cabot India) wasadded to the 400 mL of milli-Q water (F = 18.2 MΩ 3 cm), andthemixture was ultrasonicated for 30min, followed by stirring for1 h. Subsequently, an excess quantity of 0.05 M NaBH4 (Merck)solution was added, drop by drop, to the mixture with vigorousstirring until the yellowish-brown colloidal solution becamecolorless, containing dispersed black metals particles for thecomplete reduction of metals from the precursor salt. Theresulting heterogeneous colloidal mixtures were allowed to stirfor 1 h at room temperature and centrifuged. Black metal catalystparticles were then isolated by centrifugation, and the super-natant colorless liquid was tested by silver nitrate to detect thepresence of any residual bromide or chloride ion in solution. Thecatalysts were washed with plenty of milli-Q water, and thebromide test was done until the supernatant liquid gave thenegative test of it. The synthesized catalysts were dried in an ovenat 100 �C for 3�4 h.2.2. Surface Characterization and Morphology of the

Catalyst Matrix. The characteristics of the crystalline structureof the supported catalysts were determined by using the powderXRD technique. The data were determined by using the Seifert2000 diffractometer operating with CuKR radiation (λ =0.1540 nm) generated at 35 kV and 30 mA. Scans were recordedat 1�min�1 for 2θ values between 2 and 90�. Scherrer and Braggformulas were employed to calculate the mean diameter and thelattice parameter of the catalysts. The recorded XRD patternswere analyzed following the JCPDS file. To get the morphologyand particle size, TEM images of catalysts were obtained by usingan FEI model STWIN operated at an accelerating voltage of 200kV. Specimens for TEM analysis were prepared by ultrasonicallysuspending the particles in alcohol. A drop of the suspension wasdeposited onto a standard carbon-coated Cu grid and allowed todry before being inserted into the microscope. The particle sizedistribution histograms were obtained by the observation ofabout 300 particles from the four different locations of the

corresponding TEM images. An electron-dispersive X-ray de-tecting system, EDAX, coupled with the TEM instrument, wasused to analyze the chemical composition of the alloy, consider-ing several regions of the catalyst particles throughout thesurface. The spatial resolution of the EDAX probe is 0.35 nm.The adsorption isotherm of the catalysts was recorded at 77 K bythe BET method using nitrogen as the adsorbate in a Quanta-chrome Autosorb instrument (model AS1-CT). Prior to themeasurements, samples were degassed at 150 �C for 30 min. Thesurface area was evaluated according to the BET theory,17 usingthe multipoint method at relative pressures up to P/P0 = 0.2, andthe total volume (Vtot) was calculated from the volume ofnitrogen adsorbed at near to saturation pressure, P/P0 = 0.99.The volume of micropores (Vm) was derived from the so-calledt-plot dependence of the nitrogen sorption isotherm.18 The poresize distribution was derived from the adsorption isotherm accord-ing to Barrett�Joyner�Halenda (BJH) model calculations.2.3. Fabrication of the Electrodes. A catalyst ink was

prepared by taking an appropriate amount of catalyst in 5 wt %Nafion ionomer (Eloctrochem. Inc., USA) and 2-propanol (GRgrade, Merck), and the mixture was sonicated. A calculatedamount of this slurry was micropipetted out onto the graphitesupport (0.65 cm2) to maintain a constant catalyst loading of0.77 mg cm�2. A graphite plate (GLM grade, Graphite IndiaLtd.) with a thickness of approximately 2 mm was used as thesupport for the working electrode.2.4. Electrochemical Measurements. Electrochemical mea-

surements were conducted using a computer-controlled poten-tiostat/galvanostat with PG stat 30 and FRA modules (EcoChemie B.V., The Netherlands). Ethanol (AR grade, Merck)was added to the nitrogen (XL grade, 99.999%, BOC India Ltd.)saturated 0.5 M NaOH electrolyte to make a 1.0 M EtOHconcentration. All these solutions were prepared using Mili-Qwater. The geometrical area of the electrocatalyst exposed to thesolution was always 0.65 cm2. Voltammetry was carried out in aglass cell using a conventional three-electrode setup incorporat-ing a mercury/mercurios oxide (Hg�HgO) reference electrodeand a bright Pt foil (10 mm � 10 mm) counter electrode alongwith the working electrode fabricated with the synthesizedcatalysts. Steady-state potantiostatic data were recorded afterpolarization for 5 min. Electrochemical impedance measure-ments were performed with an amplitude of 5 mV in thefrequency range of 30 kHz to 30 mHz at a potential of �0.3 V.2.5. Ion-Chromatographic Analysis. Aliquots from the

working electrolyte were analyzed by ion chromatography afterpolarization at constant potential for 2 h at selected temperaturesin the range of 20�80 �C. Metrohm’s Advanced Modular IonChromatography systems, aMetrosep A Supp 5-250 column and

Table 1. Results of EDAX, XRD and Voltammetric Analysis of the Single and Ternary Catalsyts

average crystallite size (nm) ECSA (m2/g)

electrocatalysts Pt/Pd/Au

(atomic ratio)

EDAX

composition (at. %)

interplanar

distance (Å)

lattice

parameter (Å)

XRD TEM H2 ads/des

region

reduction of adsorbed

O2 region

Pt/C 1:0:0 Pt: 100 2.260 3.925 3.84 3.50 27.3 33.1

Pd/C 0:1:0 Pd: 100 2.231 3.893 3.40 4.00 19.8 24.9

Au/C 0:0:1 Au: 100 2.240 4.081 3.22 3.50 11.3 09.5

PtPdAu/C 1:1:1 Pt: 30 2.165 3.903 3.90 3.50 69.3 73.7

Pd: 38

Au: 32



a Metrosep A Supp 4-550 organic acid column, were used foranalyzing the acetate and carbonate ions. A 1mL sample solutionwas diluted with 19mL of filteredMilli-Qwater, and 25 μL of thissample was injected into the ion chromatograph.

3. RESULT AND DISCUSSION

3.1. Morphology and Structural Characterization ofPtPdAu/C Electrocatalyst. It has been found in an earlier work

that optimization of the catalyst loading with respect to thecarbon content in the catalyst matrix, for a fuel cell reaction, is anessential prerequisite to deriving the best catalytic performanceof the nanoparticles toward the respective red-ox processes (seeFigures S1�S3 in the Supporting Information). In this regard, a40% catalyst loading was found to be the choice for catalystsupport combination.The EDS measurement was recorded to derive the atomic

composition of the PtPdAu/C catalysts prepared from thechemical bath containing the metal precursors, all having aequimolar concentration, and the corresponding atomic ratiosare given in Table 1.The XRD patterns in Figure 1 clearly demonstrate that the Pt/

C, Pd/C, Au/C, and Pt30Pd38Au32/C alloy catalysts have face-centered cubic (fcc) structures. All the catalysts exhibited diffrac-tion peaks of (111), (200), (220), (311), and (222) planescorresponding to an fcc crystal structure (JCPDS 04-0802). The2θ values of the (111) planes were found to be shifted slightlyfrom the values of Pt/C at 39.763�40.101 for Pd/C, 38.184 forAu/C, and 40.069 for the ternary system PtPdAu. Assuming alloyformation among Pt, Pd, and Au based on a substantial solidsolution, such a shift can be attributed to the difference in atomicsizes of the order of Pd > Pt > Au. Seemingly, the gold catalystexhibited the lowest 2θ value, being significantly different fromthat of the other catalysts.19 The lattice parameters of thesecatalysts were obtained from the XRD patterns displayed inTable 1. It was found that the addition of Pd and Au to the Ptmatrix barely makes any effect on the crystalline structure of Pt/C because Pt, Pd, and Au have almost the same fcc crystalline

Figure 1. X-ray diffraction patterns of Pt/C, Pd/C, Au/C, and Pt-basedternary catalysts. Inset: magnified view of the (111) peak.

Figure 2. (a) TEM image of the Pt30Pd38Au32/C catalyst. (b) Particle size distribution. (c) Fringe pattern and the FFT images. (d) Point-resolved EDSspectrum of the PtPdAu/C catalyst.



structure. The diffraction position of the ternary catalyst ispositively shifted compared with that of the other catalysts,resulting in the decreased lattice parameter of 0.3903 forPt30Pd38Au32/C compared with 0.3925 nm of pure Pt/C.Low-resolution TEM images of Pt30Pd38Au32/C catalysts and

the corresponding particle size distribution histograms are shownin Figure 2a,b, respectively, and the data from TEM analysis arelisted in Table 1. There is clear tradeoff in particle size with theincorporation of Au in the ternary system. The average sizes ofthe particles for all the catalysts calculated by using the Debye�Scherrer equation, agreed well with the TEM results, and thesize distribution is found to be narrower with almost 80% ofthe catalyst particles in the 3�5 nm range. The fringe patterns forthe Pt30Pd38Au32/C catalyst are shown in Figure 2c, where thecrystalline plane is calculated with the help of FFT analysis(Figure 2c, inset). The d-spacing of Pt30Pd38Au32/C is shifted to0.2165 nm from 0.226 nm, the normal value of Pt (111) (JCPDS04-0802), indicating alloy formation, which is also supported byXRD data, discussed earlier. The overall impression is such thatthe narrow size distribution and better dispersion are achieveddue to the existence of Au in the ternary matrix. A representativepoint-resolved EDS spectrum (Figure 2d) collected from anumber of individual particles reveals that the ternary catalystmatrix is grown with an ensemble of atoms of the individualmetals Pt, Pd, and Au forming the nanoparticles.A typical reversible curve representative of a type III nitrogen

adsorption and desorption isotherm and the Vulcan XC-72[C(V72)] carbon reversible curve representative of a type IV/III are exhibited in Figure 3a. The single and ternary catalystswere studied according to the Brunauer�Deming�Deming�Teller (BDDT) classification,20 and the types of hysteresis wereattributed to cylindrical pores. Pore volume and pore sizedistributions (PSDs) were calculated on the basis of the BJHmodel performed from the adsorption branch of the isotherm, asshown in the inset of Figure 3a.The adsorption isotherms associated with the structure of the

single and ternary catalysts are essentially those of the combinationofmicropores andmesopores and as further clarified from the poresize distribution (PSD) curves. Small micropores (<2 nm) in thecarbon-supported catalyst decreases the catalyst utilization be-cause the mass transport of reactants and products becomeinsignificant for these tiny micropores. When the macropore sizeis larger than 50 nm, the surface area becomes smaller and theelectrical resistance increases.21 Ternary catalysts onVulcan XC72with optimum pore sizes in the range of 2�50 nm (as shown inFigure 3a, inset) become useful as they can render a better abilityto enhance both the dispersion and the utilization of metals in thisrange. The relevant structural parameters of all the catalystsdetermined on the basis of the isotherms18 are revealed inFigure 3b, where it is observed that, with the incorporation ofPd and Au into the Pt matrix, the total BET surface area increasesalong with the external surface area and the micropore area of thecatalyst surface. This indicates that Pt30Pd38Au32/C catalysts notonly intercalate into themicropore of the graphite support but alsoare grown on the external surface area of the support. The largequantity of micropores can enhance the adsorbability of thecatalyst and extend the active sites. On the other hand, highersurface areas and larger pore volumes allow a higher degree ofcatalyst dispersion as well as lead to highly integrated intercon-nected pore systems with periodic order and thus enable betterutilization of catalyst particles by efficient transport of reactantsand products through the catalyst matrix.

3.2. Electrochemical Investigation and the TemperatureEffect on the Electrode Kinetics.Typical voltammograms of Pt,Pd, Au, and PtPdAu catalysts in 0.5 M NaOH are shown inFigure 4. The well-established features of hydrogen adsorptionand desorption, double layer charging, oxide formation, andoxide reduction are evident from the voltammograms of all the

Figure 3. (a) Nitrogen adsorption (filled) and desorption (unfilled)isotherms for single and ternary catalysts. Inset: pore size distributionscalculated from the nitrogen adsorption�desorption isotherms. (b)Different surface areas for Pt/C, Pd/C, Au/C, and Pt30Pd38Au32/C.

Figure 4. Cyclic voltammograms of Pt/C, Pd/C, Au/C, andPt30Pd38Au32/C in 0.5 M NaOH. Scan rate = 50 mV s�1. Inset:magnified view of the oxide reduction peak region and the hydrogenadsorption�desorption region.



catalysts studied. A measure of the real electrochemical surfacearea Ar (cm

2 Pt) can be obtained from the voltammogram usingthe following equation

Ar ¼ Q h=Qm ð1Þ

where Qh is the saturated hydrogen coverage on the electrode(μC) andQm is the charge associated with monolayer adsorptionof hydrogen (μC cm�2 Pt). It is assumed that each surfaceplatinum atom is associated with one chemisorbed hydrogenatom, allowing the charge corresponding to the area under thehydrogen adsorption peaks, Qh, to be converted to the realelectrochemical surface area. The charge associated with amonolayer of hydrogen is commonly taken as 210 μC cm�2

for Pt22,23 after double layer correction. However, for Pd-con-taining catalysts, the method based on the hydrogen adsorptioncharge seems to be less reliable because Pd can absorb largequantities of hydrogen (up to 900 times its own volume). Thereal ECSA of the Pd-containing PtPdAu/C catalyst is determinedby considering the Columbic charges corresponding to the oxidereduction peak of the alloyed catalyst. However, it is not possibleto measure the ECSA value accurately by using this methodbecause the reduction peaks may be ascribed to the reduction ofthe oxides of Pd, Pt, and Au formed on the surface of thePtPdAu/C catalyst during the positive scan.24 The chargesrequired for the reduction of PdO (qPdO-red), PtO (qPtO-red), and Au2O3 (qAu2O3-red) monolayers were taken to be405, 420, and 400 μC cm�2, respectively, as reported in theliterature.25�27 The mean value of charges required for the oxidereduction of the alloyed catalyst with the atomic composition ofPt30Pd38Au32 was calculated to be 409.1 μC cm�2 and used forevaluating the ECSA of the alloyed catalyst, summarized inTable 1. The value of ECSA for Pt30Pd38Au32/C was found tobe far higher than the value obtained for pure metal catalysts,reflecting the higher catalytic surface area for the ternary alloyedcatalyst.The oxidation of Pt30Pd38Au32/C surfaces commences at

more negative potential than those observed for other catalysts,and the double layer charging region and beyond is far moreexuberated, indicating dipolar interactions, followed by oxideformation, in this region.With the incorporation of Pd and Au into the Pt matrix, the

oxidation current rises in the potential region of �750 to +250mV, as shown in Figure 5, almost 4 times that of Pt and in theorder of PtPdAu > Pd > Pt > Au.The forward peak current in the anodic sweep and the reversed

peak current in the cathodic sweep are compared in Figure 6, anda significant increase in both the current densities are observedwith the ternary catalyst compared with the single metals. Thus,the possibility of using the single metal as an effective catalyst forethanol oxidation becomes less important. Hence, for the rest ofthe investigations involving the reaction kinetics at differenttemperatures, the studies on Au and Pd were abandoned andthe results of electrode kinetics were compared for the ternarycatalyst and the single Pt only.Figure 7a,b represents the cyclic voltammograms for the EOR

on Pt/C and Pt30Pd38Au32/C in 0.5 M NaOH containing 1.0 Methanol at the temperatures of 20, 40, 60, and 80 �C. The Pt/Cand Pt30Pd38Au32/C catalysts show significantly different CVprofiles, particularly at higher temperatures. The ethanol oxida-tion peak current densities obtained with Pt/C in the forwardsweep were raised by a factor of 3 as the temperature increased

from 20 to 80 �C (shown in Figure 7a), whereas the negativegoing sweeps display a sharp vertical increase of the reverseoxidation peak, which are more prominent at higher tempera-tures. This sharp rise in current on Pt/C is indicative of the factthat the ethanolic residues remaining unoxidized on the surfaceduring the forward sweep can only be oxidized by the Pt oxideformed at higher potentials. As the temperature was increased,the potential differences between forward and reverse peaks werereduced, suggesting that surface activation of Pt/C by OHspecies is facilitated at higher temperatures. The influence oftemperature on Pt30Pd38Au32/C toward the ethanol oxidation isfar more significant. Considering the voltammograms of theternary system, as shown in Figure 7b obtained in the sametemperature range, the oxidation peak current densities wasfound to increase by almost 4 times, when the temperature wasincreased from 20� to 80 �C. In contrast to the Pt catalyst, thediminution in the reverse oxidation current with the ternarysystem is noted and a considerable rise in anodic peak intensity isobserved, particularly at high temperature. This points toward afair catalytic performance of the Pt30Pd38Au32/C system com-pared to other catalysts.

Figure 5. (a) Cyclic voltammograms of Pt/C, Pd/C, Au/C, andPt30Pd38Au32/C in 0.5 M NaOH containing 1 M ethanol. Scan rate =50mV s�1. (b)Magnified view of onset region of cyclic voltammograms.(c) Bar diagram of respective oxidation potential for all the catalysts at3.00 mA cm�2.

Figure 6. Variation of the forward and backward peak current ratio (IF/IB) and forward peak current (IF) vs backward peak current (IB) plots(inset) for the electrooxidation of ethanol on Pt/C, Pd/C, Au/C, andPt30Pd38Au32/C.



The pronounced influence of temperature on the electrocata-lysis of the ternary system as compared to Pt alone is furtherclarified with the variation of ips with temperature, as shown inFigure 8a. For the Pt/C catalyst, the anodic peak current (ip) ispractically less invariant with temperature, whereas a substantialincrease in anodic peak current is observed for the Pt30Pd38Au32/C catalyst. A considerable negative shift in the onset (Eonset)potential was observed in the alloyed electrodes, as shownin Figure 8b. A constant slope of �3.10 mV/�C for thePt30Pd38Au32/C catalyst as compared with �0.4 mV/�C forPt/C clearly demonstrates that the catalytic activity of the formeris improved to a reasonable extent with a moderate increase intemperature. Thus, it is suggested that the effect of introducingAu into Pt or PtPd may cause a monotonic decrease of reductionpotential and thereby increase the tendency of surface oxideformation.28

Figure 9a shows chronoamperometric experiments in 0.5 MNaOH containing 1.0 M ethanol on Pt/C and Pt30Pd38Au32/C at different temperatures to determine the stability of thecatalysts. A decrease of current density was recorded atthe beginning of each experiment for both Pt/C andPt30Pd38Au32/C catalysts at each temperature. The Pt/Celectrode showed rapid current decay and very low initialand final current densities for all the temperatures. Thisindicates that the Pt/C surface is prone to poisoning byethanolic residues. For Pt30Pd38Au32/C catalysts, currentdecay was comparatively low and reasonably high initialas well as final current densities were observed. Thus,

Pt30Pd38Au32/C catalysts conform to a greater tolerance toethanolic residues within this temperature range.Long-term poisoning rates (δ) were obtained by measuring

the linear decay of the current at times greater than 500 s usingthe following equation29

δ ¼ 100i0

� didt

� �t > 500 s

ð% per sÞ ð2Þ

where (di/dt)t>500 s is the slope of the linear portion of thecurrent decay and i0 is the current at the start of polarizationback-extrapolated from the linear current decay. It is evidentfrom Figure 9b that the long-term poisoning effect for ethanoloxidation can be controlled on Pt30Pd38Au32/C as compared tothe Pt/C electrode. This is in agreement with the results of thecyclic voltammetric experiments (Figure 7a,b). It is establishedthat the ethanol oxidation reaction proceeds through the forma-tion of reactive intermediates (CH3COads) and the ultimatepoisoning species (COads) on the Pt group of metals.30 However,in the present case, it appears that the reaction in its course doesnot follow the same pathway for the Pt catalyst and the ternaryalloy. The above observations were further clarified by thenegative deviations from Cottrell’s law (i R t�1/2) due toblocking of the surface area in contrast to the normal positivedeviations as a result of unwanted stirring of the solution.31

Moreover, a modification of that equation, which is known asAnson’s plot (Q R t1/2), allows the determination of the amountof charge involved in the process for the different catalysts underinvestigation. Here, we have taken the advantage of applyingchronocoulometry (CC) (the study of species adsorbed to thesurface of the working electrode) over chronoamperometry(CA), and it is possible to separate the charge due to theelectrolysis of the adsorbed molecules from the charge due tothe electrolysis of molecules in solution and the double layercharge. Figure 9c demonstrates the high level of charge on theternary electrodes, even at 40 �C, which is much higher than thatproduced by Pt at 80 �C. This is better understood when wefollowed the Arrhenius plots to derive the activation energy ofthe reaction over the temperature range with Pt andPt30Pd38Au32/C catalysts.Figure 10a shows Arrhenius plots for ethanol oxidation on the

Pt/C and Pt30Pd38Au32/C catalysts at different potentials. Linear

Figure 7. Cyclic voltammograms for ethanol electrooxidation on (a)Pt/C and (b) Pt30Pd38Au32 catalysts in a solution of 0.5 M NaOH and1.0 M ethanol at different temperatures. Sweep rate = 50 mV s�1.

Figure 8. Variation of (a) anodic peak current ip, (b) onset potential(Eonset), and (c) anodic peak potential Ep for the electrooxidation ofethanol on Pt/C and Pt30Pd38Au32 catalysts with different temperatures.



relationships between log i and 1/T are maintained throughoutthe temperature range for each of the potentials. The apparentactivation energy (Figure 10b) for ethanol oxidation on Pt/Cincreases from 9.04 to13.55 kJ mol�1, whereas on Pt30Pd38Au32/

C, the same decreases from 5.44 to 4.09 kJ mol�1 during the shiftto potential from �0.4 to 0.2 V. Therefore, it can be suggestedthat alloying Pd and Au with Pt helps to lower the activationenergy for ethanol oxidation. The lower values of Ea(app) forethanol oxidation on Pt30Pd38Au32/C catalysts leads to anincrease in the ethanol oxidation rate on that catalyst. It isunderstandable that conversion of ethanol to CO2 alwaysrequires transfer of additional oxygen species from other activa-tion species, such as H2O and OH�, so as to overcome thelimitations of sluggish kinetics as associated with the Pt catalyst.The cometals Pd and Au, in almost equal proportions, hold astrong partnership with the base metal Pt, providing their bestindividual capabilities of activating the surface with the formationof (OH)ads, as in the case of Pd, through image dipolar inter-actions32 and a ready supply of oxygen to the ultimate carbonac-eous species by Au through surface oxide formation.The reaction on Pt presumably keeps on proceeding toward

ultimate poisoning by Pt �CtO formation, thus exhibiting anunfavorable kinetics for the complete oxidation of ethanol.However, the reaction of COads with OHads/OH

� can occur athigh activation energy only when PtOH is formed at higherpotentials, producing carbonates and thus alleviating theproblem of surface poisoning to some extent at elevated

Figure 10. (a) Arrhenius plots of ethanol oxidation current densities.(b) Dependence of apparent activation energy with potentials on Pt/Cand Pt30Pd38Au32/C catalysts in a solution of 0.5 M NaOH and 1.0 Methanol on Pt/C and Pt30Pd38Au32/C catalysts.

Figure 9. (a) Chronoamperometric studies of ethanol oxidation on Pt/C and Pt30Pd38Au32/C catalysts. (b) The long-term poisoning rate vsdifferent temperatures. (c) Charge density vs t1/2 plots of Pt andPt30Pd38Au32/C at 40� and 80 �C in a solution of 0.5 M NaOH and1.0 M ethanol at a constant potential of �300 mV (vs Hg�HgO).



temperatures. On the other hand, kinetic control is establishedon Pt30Pd38Au32/C by the reaction of the carbonaceous inter-mediates (CH3CO)ads with OHads, leading to the formation ofCH3CO

� ions as well as formation of carbonates by facile C�Cbond cleavage at a lower overpotential. These two anionic speciesexit the pre-electrode layers and diffused to the bulk withsufficient ease, thus rendering the surface free from blockage.Eventually, an appreciable reduction in the activation energyEa(app) of ethanol oxidation was observed on the Pt30Pd38Au32/Ccatalyst, which translates to the higher intrinsic activity for thiselectrode compared with Pt and validates its potentiality foralkaline DEFCs.The electrochemical impedance spectra, used to investigate

the kinetics of ethanol oxidation on Pt/C and Pt30Pd38Au32/C,were recorded at �300 mV, in the solution containing 0.5 MNaOH and 1.0 M EtOH at temperatures of 20�80 �C. Thecomplex-plane impedance spectra for Pt30Pd38Au32/C and Pt/Ccatalysts correspond to theNyquist plots inFigure 11a,b.A semicircleis observed at all temperatures, which indicates the presence of aresistive component in the integrated circuit. The diameter ofthis semicircle decreased as the temperature increased. Pseu-doinductive behavior (positive Z// values) at low frequencies isobserved when the temperature is raised above 20 �C. A similarresult was obtained for methanol oxidation,33 where the induc-tive effect was attributed to the kinetics of the electrooxidationprocess, in particular, to CO oxidation.34�37 Bode plots areshown in the insets of Figure 11a,b. These plots show that thephase angle (Φ) is almost independent of the temperature in thefrequency range of 30 kHz to 30 mHz. At low frequency ranges,the phase angle shows a maximum; its value depends on thetemperature as well as the frequency. The maximum value of thephase angle is obtained at the highest temperature studied. As thetemperature is increased, the maximum in the phase angledecreased and shifted to higher frequencies. In the case ofresponses, obtained for temperatures above 40 �C, the phaseangle values cross the frequency axis and move toward positivevalues of the phase angle with a displacement toward higherfrequencies where the phase angle nears zero. The increasingfrequency values as a function of the temperature indicates thepredominance of the pseudoinductive behavior, which translatesto an increase in the reaction rate with increasing temperature.This indicates that the reaction activity increased with increasingtemperature, which is in concordance with the results of voltam-metry. The electrooxidation of ethanol on both the catalystsshows inductive behavior only at higher temperature, which isindicative of the possible oxidation of the accumulated reactionintermediates on the catalyst surface at elevated temperatures.Therefore, with increasing temperature, the reaction kinetics isenhanced and governed by partial oxidation of ethanol, which, ineffect, forms more COads. The impedance data were analyzedusing the software available with the potentiostat (Autolab 12,FRA ,version 4.9), resulting in the equivalent circuits shown inthe insets of Figure 11a,b. In this circuit Rs, CPE, and Rct arerepresentative of the solution resistance, constant phase element,and charge-transfer resistance associated with the ethanol oxida-tions primarily involving the dissociative adsorption of thealcohol on the surface. In another circuit, Rs, CPE, and Rctrepresent the solution resistance, a constant phase elementcorresponding to the double layer capacitance, and the charge-transfer resistance associated with the electrooxidation of etha-nol. For the ternary catalyst, the circuit included the inductiveresistance (Ro) and the inductance (L) associated with the COads

oxidation kinetics. The equivalent circuit parameters are sum-marized in Table 2. It is evident from Table 2 that very feeblecharge-transfer resistance is offered by Pt30Pd38Au32/C, which isalmost 30 times less than those offered by Pt alone at thistemperature range (compared at 60 �C). There is a significant fallin the Rct values for this catalyst, even at a temperature as low as20 �C, and is clearly shown in Table 2. The appearance of L andits subsequent lowering at high temperature (80 �C) for theternary system reflects the successful oxidative removal ofcarbonaceous intermediates from the surface. Thus, it appearsthat the cometals provide tremendous kinetic input in the

Figure 11. Impedance spectra and Bode plots (inset) for ethanoloxidation on (a) Pt/C and (b) Pt30Pd38Au32/C catalysts at differenttemperatures in a solution of 0.5 M NaOH and 1.0 M ethanol at �300mV (vs Hg�HgO). Equivalent circuit diagrams (insets) represent the fitto the impedance data.

Table 2. Values of the Electrical Elements in the EquivalentCircuits Employed for the Fittings of the Impedance Spectraat �300 mV for Pt/C and Pt30Pd38Au32/C at DifferentTemperatures

Pt/C Pt30Pd38Au32/C

EC parameters 20 �C 40 �C 60 �C 80 �C 20 �C 40 �C 60 �C 80 �C

Rs/Ω 5.79 4.91 4.99 5.91 8.84 6.29 5.31 3.39

Rct/Ω 140 89.5 86.3 38.9 10.48 5.36 2.76 1.87

CPE �102/F cm�2

7.98 7.77 7.18 7.01 1.28 1.41 1.45 1.85

Lx 102 (H) 1.93 1.65 1.49 1.12 0.19 0.13



reaction sequences for ethanol oxidation on the Pt�Pd�Ausurface as the dehydrogenation process gets immediately coupledwith faster oxygen transfer on the preactivated surface and theoxidation tends to proceed toward completion. This outstandingcatalytic performance of the Pt30Pd38Au32/C system toward thecharge-transfer kinetics of ethanol oxidation establishes the super-iority of this electrode over Pt in an alkaline medium.3.3. Estimation of Oxidation Products Formed during

Ethanol Electrooxidation and Its Mechanistic Overview.Figure 12a,b shows the ion chromatograms for the productanalysis of the constant potential (�300 mV vs Hg�HgO)oxidation of ethanol carried out on Pt/C and Pt30Pd38Au32/Celectrodes for a period of 1 h at temperatures in the range of20��80 �C. The respective products formed, acetate andcarbonate, during the oxidation of ethanol were estimated andverified under the influence of temperature. The estimatedquantities of acetate and carbonate are extremely high on thePt30Pd38Au32/C as compared to Pt/C which, however, showed agreater temperature dependence for ethanol oxidation as thetemperature was raised from 20� to 80 �C.

The overall oxidation reaction equation for ethanol in analkaline medium is supposed to be

CH3CH2OH + 12OH� f 2CO2 + 9H2O + 12e� ð3ÞHowever, the electrooxidation reaction of ethanol is consid-

ered to be a complex reaction that proceeds through multiplesteps. Several reaction products and intermediates are expectedto be formed on the catalyst surface. Some of the adsorbedintermediates converted into the end products, such as acetateand carbonate, and the rest being retained on the catalyst surfacemay cause considerable reduction in the fuel efficiency. Table 3shows the yield of acetate and carbonate ions produced at eachtemperature for Pt30Pd38Au32/C and Pt/C catalysts. The re-spective yield of these two anionic products at both the catalystsincreased approximately linearly with temperature.Liang et al. have recently reported on the mechanism of

ethanol oxidation on a Pd electrode in alkaline solution basedupon the qualitative assessment through voltammetric analysisand showed that acetate is the final product of the alcoholoxidation.38 This is possibly due to the ability of Pd to adsorbethoxy ion directly in an alkaline medium and release the acetateas the main product. In a very recent work,39 we have also studiedthe reaction path of the ethanol electrooxidation on the PtPd/Ccatalyst and estimated the oxidation products through ionexchange chromatography where also acetate was found to bethe major product. In the present investigation, we have tried tojustify the quantitative yield of the reaction products throughmechanistic interpretation of ethanol oxidation on the ternarysystem. In contrast to the Pt/C electrode, the higher yield ofcarbonate ion in oxidized solution for the Pt30Pd38Au32/Ccatalyst is possibly due to better coverage of the ternary catalystsurface by CH3COads and OHads species. However, the appre-ciably high content of acetate with the ternary system is alsoindicative of a parallel pathway that limits the oxidation to 4e�

transfer processes, to some extent. However, the voluminousproduction of CO3

�2 on the ternary alloy takes the lead inpropagating the oxidation reaction mostly toward completion.The presence of Pd and Au in the Pt30Pd38Au32/C catalyst notonly prohibits the building up of COads but also activates thesurface with the formation of (OH)ads through image dipolarinteractions and provides an oxygen supply for the ultimateconversion of ethanol to the end product on this catalyst surface,resulting in a higher carbonate ion content in the alkalinemedium. The Pt30Pd38Au32/C catalyst showed remarkable im-provement in catalytic activity toward ethanol oxidation com-pared with the Pt/C catalyst at 20 and 80 �C. The findings fromthe product analysis also corroborate with the results obtainedfrom the electrochemical measurements.A schematic of mechanistic pathways for the ethanol oxidation

process on the Pt30Pd38Au32/C catalyst in alkaline media has

Figure 12. Ion chromatograms for (a) acetate and (b) carbonateproduced by the electrooxidation of ethanol in a solution of 0.5 MNaOH and 1.0 M ethanol on Pt30Pd38Au32/C at �300 mV (vsHg�HgO) and at 20�80 �C.

Table 3. Results of the Reaction Product Analysis afterOxidation of Ethanol on Pt30Pd38Au32/C Catalyst in a Solu-tion Containing 0.5 M NaOH and 1.0 M Ethanol at�300 mV(vs MMO) for 1 h

Pt/C Pt30Pd38Au32/C

products 20 �C 40 �C 60 �C 80 �C 20 �C 40 �C 60 �C 80 �C

acetate (in ppm) 13.4 28.9 68.9 82.6 225 393 625 680

carbonate (in ppm) 26.8 43.3 90.8 128.4 355 449 698 951



been proposed (Scheme 1) with the presumption that Pd and Auincorporation into the Pt matrix makes a combined effort to

facilitate the adsorption of OH� and also supply the oxygenspecies at the electrode surface.

4. CONCLUSION

For ethanol oxidation, the onset potentials on thePt30Pd38Au32/C catalyst show a considerable negative shift withtemperature, and the values are much lower than that obtainedwith the Pt/C catalyst. The decay in oxidation current density isminimal for the Pt30Pd38Au32/C catalyst at all temperatures,whereas the Pt/C catalyst shows rapid current decay, indicating ahigher rate of catalyst poisoning. Proportionate addition of Pdand Au into the Pt matrix drastically reduces the activationenergies and the charge-transfer resistance and thereby promotesthe electrocatalytic activity toward ethanol oxidation. The out-standing behavior toward electrocatalysis of the optimal ternarysystem is further reflected by the very high yield of acetate andcarbonate during the course of ethanol oxidation in an alkalinemedium.

’ASSOCIATED CONTENT

bS Supporting Information. Additional data of electro-chemistry. This material is available free of charge via the Internetat http://pubs.acs.org.

’AUTHOR INFORMATION

Corresponding Author*Tel: 033 2668 4561 � 63 (x 514). Fax: 033 2668 2916. E-mail:[email protected].

’ACKNOWLEDGMENT

The authors gratefully acknowledge the financial support ofthe Defense Research and Development Organization (DRDO),New Delhi, India. A.D. wishes to acknowledge the SeniorResearch Fellowship from the Council of Scientific and IndustrialResearch (CSIR), New Delhi, India. The authors also acknowl-edge the Ministry of Human Resource and Development(MHRD), New Delhi, India, for the instrumental facilitiesprovided to the Department of Chemistry, BESUS, India.

’REFERENCES

(1) Liu, Z.; Shamsuzzoha, M.; Ada, E. T.; Reichert, W. M.; Nikles,D. E. Synthesis and activation of Pt nanoparticles with controlled size forfuel cell electrocatalysts. J. Power Sources 2007, 164, 472.

(2) Du, H. Y.; Wang, C. H.; Hsu, H. C.; Chang, S. T.; Chen, U. S.;Yen, S. C.; Chen, L. C.; Shih, H. C.; Chen, K. H. Controlled platinumnanoparticles uniformly dispersed on nitrogen-doped carbon nanotubesfor methanol oxidation. Diamond Relat. Mater. 2008, 17, 535.

(3) Park, S.; Xie, Y.; Weaver, M. J. Electrocatalytic pathways oncarbon-supported platinum nanoparticles: Comparison of particle-size-dependent rates of methanol, formic acid, and formaldehyde electro-oxidation. Langmuir 2002, 18, 5792.

(4) Nores Pondal, F. J.; Vilella, I. M. J.; Troiani, H.; Granada deMigue, M. S. R; Scelza, O. A.; Corti, H. R. Catalytic activity vs. sizecorrelation in platinum catalysts of PEM fuel cells prepared on carbonblack by different methods. Int. J. Hydrogen Energy 2009, 19, 8193.

(5) Narayanan, R.; El-Sayed, M. A. Shape-dependent catalyticactivity of platinum nanoparticles in colloidal solution. Nano Lett.2004, 4, 343.

(6) Arenz, M.; Mayrhofer, K. J. J.; Stamenkovic, V.; Blizanac, B. B.;Tomoyuki, T.; Ross, P. N.;Markovic, N.M. The effect of the particle size

Scheme 1. Suggested Electrooxidation of Ethanol on theTernary Alloy (PtPdAu) Surface Enriched with (OH)ads



on the kinetics of CO electrooxidation on high surface area Pt catalysts.J. Am. Chem. Soc. 2005, 127, 6819–6829.(7) Wang, H.; Jusys, Z.; Behm, R. J. J. Phys. Chem. B 2004,

108, 19413.(8) Camara, G. A.; Iwasita, T. J. Electroanal. Chem. 2005, 578, 315.(9) Wang, J. X.; Marinkovic, N. S.; Zajonz, H.; Ocko, B. M.; Adzic,

R. R. J. Phys. Chem. 2001, 105, 2809.(10) Sasaki, K.; Wang, J. X.; Balasubramanian, M.; McBreen, J.;

Uribe, F.; Adzic, R. R. Electrochim. Acta 2004, 49, 3873.(11) Iwasita, T. J. Brazil. Chem. Soc. 2002, 13, 401.(12) Xia, X. H.; Liess, H. D.; Iwasita, T. J. Electroanal. Chem. 1997,

437, 233.(13) Delime, F.; Leger, J. M.; Lamy, C. J. Appl. Electrochem. 1999,

29, 1249.(14) Lamy, C.; Lima, A.; LeRhum, V.; Delime, F.; Coutanceau, C.;

Leger, J. M. J. Power Sources 2002, 105, 283.(15) Santos, V. P.; Tremiliosi-Filho, G. J. Electroanal. Chem. 2003,

395, 554.(16) Antolini, E.; Gonzalez, E. R. J. Power Sources 2010, 195, 3431.(17) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of gases in

multimolecular layers. J. Am. Chem. Soc. 1938, 60, 309–319.(18) Lowel, S.; Shields, J. E.; Thomas, M. A.; Thommes, M.

Characterization of Porous Solids and Powders: Surface Area, Pore Sizeand Density; Kluwer Academic Publishers: Dordrecht, The Netherlands,2004; ISBN 1-4020-2302-2.(19) Irissou, E.; Laplante, F.; Garbarino, S.; Chaker, M.; Guay, D.

J. Phys. Chem. C 2010, 114, 2192–2199.(20) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.;

Peirotti, R. A.; Rouquerol, J. IUPAC commission on colloid and surfacechemistry including catalysis. Pure Appl. Chem. 1985, 57, 603.(21) Chan, K. Y.; Ding, J.; Ren, J.; Cheng, S.; Tsang, K. Y. J. Mater.

Chem. 2004, 14, 505.(22) Shepperd, S.-A.; Campbell, S. A.; Smith, J. R.; Lloyd, G. W.;

Ralph, T. R.; Walsh, F. C. Electrochemical and microscopic character-ization of platinum-coated perfluorosulphonic acid (Nafion 117) mate-rials. Analyst 1998, 123, 1923.(23) Bagotzky, V. S.; Vassilyev, Y. B.Mechanism of electro-oxidation

of methanol on the platinum electrode. Electrochim. Acta 1967, 12, 1323.(24) Yi, Q.; Wu, H.; Liu, X.; Xu, G.; Zhou, Z.; Chen, A. Electro-

activity of titanium-supported nanoporous Pd�Pt catalysts towardsformic acid oxidation. J. Electroanal. Chem. 2008, 619�620, 197–205.(25) Singh, R. N.; Singh, A.; Anindita Electrocatalytic activity of

binary and ternary composite films of Pd, MWCNT and Ni. Part II:Methanol electrooxidation in 1 M KOH. Int. J. Hydrogen Energy 2009,34, 2052.(26) Trasatti, S.; Petrii, O. A. Real surface area measurements in

electrochemistry. Pure Appl. Chem. 1991, 63, 711.(27) Wang, K. L.; Lu, J. T.; Zhuang, L. Direct determination of

diffusion coefficient for borohydride anions in alkaline solutions usingchronoamperometry with spherical Au electrodes. J. Electroanal. Chem.2005, 585, 191.(28) Guerin, S.; Hayden, B. E.; Lee, C. E.; Mormiche, C.; Russell,

A. F. J. Phys. Chem. B 2006, 110, 14355.(29) Guo, J. W.; Zhao, T. S.; Prabhuram, J.; Chen, R.; Wong, C. W.

J. Power Sources 2006, 156, 345.(30) Muller, J. T.; Urban, P. M.; Holderich, W. F. J. J. Power Sources

1999, 84, 157.(31) Salazar-Banda, G. R.; Suffredini, H. B.; Calegaro, M. L.; Tani-

moto, S. T.; Avaca, L. A. J. Power Sources 2006, 162, 9.(32) Bockris, J.; Reddy, A. K. N.; Aldeco,M.Modern Electrochemistry,

2nd ed.; Vol. 2A; Kluwer Academic Publisher: 2002, New York, USA.(33) Liu, Y. C.; Qiu, X. P.; Zhu, W. T.; Wu, G. S. J. Power Sources

2003, 114, 10.(34) Melnick, R. E.; Palmore, G. T. R. J. Phys. Chem. B 2001,

105, 1012.(35) Otomo, J.; Li, X.; Kobayashi, T.; Wen, C.; Nagamoto, H.;

Takahashi, H. J. Electroanal. Chem. 2004, 573, 99.

(36) Lu, J.; Lu, S.; Wang, D.; Yang, M.; Liu, Z.; Xu, C.; Jiang, S. P.Electrochim. Acta 2009, 54, 5486.

(37) Fang, X.; Wang, L.; Shen, P. K.; Cui, G.; Bianchini, C. J. PowerSources 2010, 195, 1375.

(38) Liang, Z. X.; Zhao, T. S.; Xu, J. B.; Zhu, L. D. Electrochim. Acta2009, 54, 2203.

(39) Mahapatra, S. S.; Dutta, A.; Datta, J. Electrochim. Acta 2010,55, 9097–9104.

The Beneficial Role of the Cometals Pd and Au in the ... · ethanol oxidation in alkaline solutions...

Documents

Transcript of The Beneficial Role of the Cometals Pd and Au in the ... · ethanol oxidation in alkaline solutions...