Synergistic effect of Co alloying and surface oxidation on oxygen reduction reaction performance for the Pd electrocatalysts

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

Available at www.sciencedirect.com

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

Synergistic effect of Co alloying and surface oxidation on oxygen reduction reaction performance for the Pd electrocatalysts Yu-Chen Wei a, Chen-Wei Liu a, Hung-Wen Lee a, Shu-Ru Chung b, Sheng-Long Lee a, Ting-Shan Chan c, Jyh-Fu Lee c, Kuan-Wen Wang a,* a

Institute of Materials Science and Engineering, National Central University, No.300, Jhongda Road, Taoyuan 32001, Taiwan Department of Materials Science and Engineering, National Formosa University, Yunlin 63201, Taiwan c National Synchrotron Radiation Research Center (NSRRC), Hsinchu 30076, Taiwan b

article info

abstract

Article history:

The synergistic effect of Co alloying and oxidation treatment induced progressive

Received 30 July 2010

enhancement on oxygen reduction reaction (ORR) activities of Pd/C catalysts is studied.

Received in revised form

Based on the XPS characterization, a new term, degree of surface oxidation (DSO), is

26 November 2010

proposed to illustrate their relationship between ORR activity and surface oxidation extent.

Accepted 19 December 2010

TPR characterization also provides the evolution of surface species within the topmost

Available online 26 January 2011

region. It can be obviously found the optimal temperature for the promotion of ORR activity on various oxidized samples is 520 K. On the other hand, various heat treatment atmo-

Keywords:

spheres (H2 and CO) are applied on PdeCo system without changing their particle size. It is

PdCo/C

clearly evident that the oxidized catalysts can exhibit the superior performance relative to

Oxidation treatment

that of the non-oxidized ones, confirming the improved ORR activity is solely ascribed to

Temperature-programmed

the formation of surface PdO species with 100% DSO value rather than large particle size

reduction (TPR)

effect. Moreover, an explainable model is demonstrated to illustrate the promotional effect

Degree of surface oxidation (DSO)

of ORR performance on the oxidized PdCo/C catalysts.

Oxygen reduction reaction (ORR)

Copyright ª 2010, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

PdO

1.

Introduction

Nanoparticles (NPs) with their unique physical and chemical properties have received growing attention because of applications as electrocatalysts in the polymer electrolyte membrane fuel cells (PEMFCs) [1,2]. Carbon supported platinum (Pt/C) or other noble metals have been extensively used as cathode electrocatalysts for the oxygen reduction reaction (ORR) [3e5]. More recently, due to the rising price of Pt and the sluggish kinetics on ORR, the development of Pd-based or

non-Pt bimetallic catalysts plays a decisive role in the widespread commercialization of this green technology [6e10]. It is unquestionable that the alloying of Pd catalysts with transition metals (M ¼ Fe, Co, Ni, etc) is of great significance for improving their ORR activity owing to the changes of PdePd bond length, modification of the electron configuration, and/or alteration of the surface species and compositions [11e17]. Fernandez et al. [11] have demonstrated the main concept for choosing high performance PdeM alloy catalysts based on the thermodynamic model. As illustrated, the main

* Corresponding author. Tel.: þ886 3 4227151x34906; fax: þ886 3 2805034. E-mail address: [email protected] (K.-W. Wang). 0360-3199/$ e see front matter Copyright ª 2010, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2010.12.098

3790

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

function of the incorporation of active elements (M) such as Co can facilitate the dissociation of OeO bond and thereby the produced CoeOads species may transfer to the Pd site in which the electrochemical reduction reaction takes place immediately. Moreover, Suo et al. [18] have proposed the design consideration for the preparation of PdeCo alloy catalysts. They have indicated that the alloying effect influences the lattice-strain and surface-ligand of the catalysts. As a result, PdeCo bimetallic catalysts with the Pd shell/alloy core structure can display excellent ORR activity. Besides, the post heat treatment may provide alternative ways to improve both the electro-catalytic activity and durability of the catalysts due to the enhancement of alloying extents, the removal of residual surfactants, the formation of some specific core-shell structures, and/or the alteration of the surface states [19e25]. Mayrhofer et al. [25] have reported that by CO induced surface segregation or electrochemical leaching processes, the alloy catalysts with Pt shell/alloy core structure can be achieved and their ORR activity can be modified. With respect to the anode catalysts, the influence of various post heat treatment strategies on the performance enhancement of PtRu/C has attracted numerous concerns [26e29]. Hwang et al. [26] have studied the improvement of methanol oxidation reaction (MOR) activity for PtRu/C electrocatalysts by means of thermal heat treatments. It is noted that H2 treatment is better than O2 treatment owing to the increased surface population of active Pt sites. This result is consistent with the point reported by Babu et al. [30] that annealing PtRu/C in H2 flow can indeed improve the MOR activity and CO tolerance due to the presence of a Ru-rich core and a PteRu alloy overlayer. However, on the other hand, O2 treatment has been performed in PtRu/C system to modify the MOR performance by the formation of crystalline RuO2 and/or modification of the oxygen functional groups of C supports [31e33]. In terms of the cathode catalysts, up to now, various approaches including heat treatment in oxidative, inert or reducing atmospheres have been applied on the Pt, Pt-based and Pd-based electrocatalysts to significantly promote the ORR activity [17,25,34e37]. Jeon et al. [35] have pointed out that Pt and NiO species on the surface of PtNi alloy catalysts can facilitate the catalytic activity after Ar treatment at elevated temperature. In addition, the function of oxide species for the promotion of electro-catalytic activity has been described [38,39]. In SnO2 system, the chemisorption of O2 on the SnO2 2þ surface as O ions produced by 2 may be accelerated by Sn cathode polarization and correspondingly the pronounced increase in ORR activity is noticed [39]. It is of paramount importance to mention the fact that the electro-catalytic properties of alloy NPs toward ORR are greatly dependent on their particle sizes, morphologies, structures and surface compositions [40e45]. It is interestingly noted that by using thermal annealing processes or acid removal of transition metals on Pt-based alloy materials, the surface segregation can be induced and the ORR activity can be enhanced through the formation of a sandwich like Pt-skin structure [46e48]. The original improvement of electro-catalytic activity can be attributed to the variations on chemical compositions of the near surface regions and a following decrease in the metaleoxygen bond strength as compared to

pure Pt [49,50]. On the other hand, besides the variations on surface species induced by various processes, meanwhile, the particle size effect of NPs on the ORR activity is also an important factor needed to be considered. Unfortunately, some controversial results regarding this issue are reported. Takasu et al. [51] have found that the Pt NPs with small size display a decrease in ORR activity due to the strong adsorption of oxygen on Pt surface. In addition, a similar conclusion is demonstrated by Mukerjee et al. [52] that a decline in electrocatalytic activity on small-sized Pt NPs (<5.0 nm) is explicated by a robust adsorption of OH. Apart from this, Mayrhofer et al. [53] also have noted the strong adsorption of OH on small NPs in perchloric acid electrolyte. On the contrary, Maillard et al. [54] have reported that the increased ORR mass activity is found with decreasing Pt particle size from 4.6 to 2.3 nm in a methanol-containing environment, however, the ORR mass activity does not exhibit particle size dependency (from 2.3 to 3.5 nm) in a methanol-free electrolyte. Moreover, Yano et al. [55] have also examined the ORR activity in a series of Pt electrocatalysts with various average diameters ranging from 1.6 to 4.8 nm. The results display that the ORR rate constant is regardless of their particle size. Therefore, the annealing processes on catalysts resulting in the undesirable aggregation or particle growth of NPs may be viewed as a possible factor influencing the ORR activity. Thus, it is a significant issue for us to investigate the size and surface composition effect on the catalytic activity for alloy NPs. Although the post-thermal treatment using either hydrogen or inert atmosphere on catalysts is commonly considered as an indispensible strategy for the promotion of their ORR activity, to the best of our knowledge, there is a paucity of reported studies on the oxidationeactivity relationship of the catalysts. Herein, the main impetus of this study has highlighted the modification of Pd catalysts via the synergistic effect of Co alloying and oxidation treatment. For this purpose, the Pd and PdeCo catalysts have been prepared via the deposition-precipitation (DP) route and various amounts of Pd surface oxides (PdsO) have been produced during the oxidation treatment process at 320e620 K. Additionally, in order to verify the ORR performance enhancement is solely attributed to the formation of PdsO, various heat treatment strategies such as N2, H2 or CO, are used without varying the particle size of the alloy NPs. Their surface compositions, surface species, structures, particle sizes, and ORR performance can be explicitly elucidated. In terms of surface characterization, X-ray photoelectron spectroscopy (XPS) is applied to estimate the surface composition of NPs with particle size of approximately or large than 3.0 nm [56]. Moreover, temperature-programmed reduction (TPR) can be applied to qualitatively provide the view point of surface species on the near surface regions of NPs [57e63].

2.

Experimental section

2.1.

Preparation of catalysts

The PdCo/C alloy catalysts with Pd/Co atomic ratio of 3 to 1 and 20% of metal loading were prepared by the depositionprecipitation (DP) method. Palladium nitrate (Pd(NO3)2) and

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

cobalt nitrate (Co(NO3)2$6H2O) in the desired stoichiometry were co-deposited onto the commercial C black (Vulcan XC72R) at 340 K by controlling the pH value to 9. The solution was stirred for 24 h at room temperature, subsequently filtered extensively with DI water, dried at 320 K for 24 h, reduced in an atmosphere of flowing H2/N2 (10/90 vol.%) gas at 390 K for 1 h and stored as fresh Pd3Co1/C catalysts. Afterward, the oxidation treatment was performed in air at a temperature ranging from 320 to 620 K for 1 h. The catalysts after air heat treatments were referred to as OT (T ¼ the oxidation temperature, 320e620 K). Moreover, the Pd/C catalysts with the same 20% metal loading were also prepared and oxidized in the similar way for the sake of comparison. Meanwhile, the heat treatments in N2, H2 and CO were also applied on the PdeCo catalysts to study the surface composition influence of the ORR activity and the samples are named as N520, H520, and CO520, respectively.

2.2.

Physical characterization of catalysts

ICP: The exact compositions of the alloy catalysts were examined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES, Jobin Yvon JY24). TGA: The metal loadings of the alloy catalysts were measured by thermal gravimetric analysis (TGA, PerkineElmer TGA-7). Initially, a small amount of alloy catalysts was placed into a Pt made basket and then transferred into TGA measurement with a temperature range starting from 300 to 1173 K at a heating rate of 8 K min1 under air atmosphere. XRD: The phase structures of various oxidized Pd and Pd catalysts were accomplished by using Shimadzu X-ray diffractometer (XRD) with the Cu Ka radiation source operating at 40 kV and 25 mA. The XRD patterns were collected by 2q scan from 20 to 70 with the scanning rate of 0.125 per step. XAS: The typical XAS (X-ray absorption spectroscopy) spectra of various alloy NPs were obtained in fluorescence mode at the BL01C1 beam line facility using a double crystal monochromator equipped with Si (111) crystal at the National Synchrotron Radiation Research Center (NSRRC), Taiwan. Si monochromator was employed to adequately select the energy at the Pd K-edge (24,350 eV) and the Co K-edge (7709 eV). In general, alloy NPs were prepared as pallets with appropriate absorption thickness (mx ¼ 1.0, where m is the absorption edge and x is the thickness of the sample) so as to attain the proper edge jump step at the absorption edge region. In order to acquire acceptable quality spectra, each XAS measurement was repeated at least twice and averaged for successive comparison. TEM: The morphologies for various oxidized samples were determined by transmission electron microscopy (TEM, JEOL2100) equipped with a LaB6 electron gun source and operated at 200 kV. Each sample was identically put into a vial with a small amount of catalyst and ultrasonically suspended in 2propanol. The uniform suspension was then immediately dried on an amorphous carbon foil supported on a 200 mesh/ inch copper grid. XPS: An X-ray photoelectron spectroscopy (XPS, Thermo VG Scientific Sigma Probe) using an Al Ka radiation was used to study the surface compositions of the alloy catalysts. All binding energies were calibrated with respect to the C 1s line at 284.6 eV. The surface compositions of the catalysts were

3791

estimated by calculating the integral of each peak. Shirley type background was used to subtract the original peak and then a combination of Lorentzian and Gaussian lines was applied to fit the experimental curve. TPR: In each TPR analysis, a sample of ca. 30 mg was inserted into a tube reactor and pre-oxidized in air at room temperature (300 K) for 1 h. Absorbed samples were reduced by a flow of 20% H2 in N2 at a flow rate of 30 mL min1 while raising the temperature from 220 to 500 K at a heating rate of 7 K min1. Silica gel and molecular sieve absorbents were used for water removal before the flowing gas reached the detector. Temperature profiles of the reducing peaks and the amount of hydrogen consumption were detected by a thermal conductivity detector (TCD).

2.3.

Electrochemical measurements of catalysts

ORR polarization curves: Oxygen reduction current was measured via linear sweep voltammetry (LSV) method with a scan rate of 5 mV1 and a rotation speed of 1600 rpm in a single compartment three-electrode configuration. Rotating disk electrode (RDE) imbedded with a Teflon holder was used as the working electrode. A calomel electrode (SCE) separated from the working compartment by a closed electrolyte bridge for the purpose of avoiding chloride contamination was considered as reference electrode, and the spiral Pt wire was used as counter electrode. Initially, 5 mg catalysts (Pd/C and various oxidized PdeCo alloy catalysts) were ultrasonically suspended in 2-propanol and 5 wt.% Nafion solution (Aldrich) to obtain the catalyst ink, and then a specific amount of 20 mL slurry was spread on the surface of the glassy carbon electrode (area ¼ 0.19625 cm2) leading to a specific metal loading of ca. 100 mg cm2 for the catalysts. After the evaporation of the catalysts ink in the ambient temperature, the electrode was then transferred to the electrochemical experiments carried out in 0.1 M HClO4(aq) saturated with high-purified O2 at room temperature. During each measurement, a moderate O2 gas flow was kept above the electrolyte. The current density at E ¼ 0.7 V, I07, within the mixed kinetic-diffusion region was used for comparison [48,64]. Durability test: The stability of the catalysts was conducted via cyclic voltammograms (CV) characterization on an identical three electrode system as documented by Sarkar et al. [65]. A similar preparation method of catalyst inks was also used as described above. The CV plots were recorded at a scan rate of 20 mV1 in the N2-saturated 0.1 M HClO4 aqueous solution between 0.24 and 1.1 V (vs. SCE). Before recording the voltammograms, the catalyst was cleaned and activated by several times cycling at a scan rate of 50 mV1 between potential of 0.24 and 0.96 V (vs. SCE). All potentials throughout this study are quoted with respect to the normal hydrogen electrode (NHE).

3.

Results and discussion

3.1. The effect of surface oxidation on the promotion of ORR activity and durability The chemical analysis results of Pd/C and PdCo/C catalysts prepared by the DP method are listed in Table S1 (Supplementary data). For the Pd/C and PdCo/C catalysts, the

3792

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

metal loading is about 19.4 and 19.6 wt.%, respectively, and the Pd/Co ratio of PdCo/C is about 3.1. Fig. 1 displays the XRD patterns for the as-prepared and various oxidized PdCo/C catalysts. The vertical dot line represents the peak position of (111) diffraction for metallic Pd (JCPDS 46-1043). Noticeably, the peaks observed at around 40.6, 47.0, and 69.0 are attributed to the reflections of PdeCo alloys. While the oxidation temperature increases from 320 to 420 K, the diffraction peaks suggest no formation of oxide phases. However, on raising the temperature to 470e520 K, the reflections corresponding to metallic PdeCo alloys vanish gradually and the characteristic peaks of PdO (JCPDS 02-1432) are apparently observed. Furthermore, the weak peak belonging to (200) diffraction of CoO (JCPDS 43-1004) phase is also found at 42.4 . For O620 sample, the peak of C becomes insignificant, and peaks of metallic Pd, PdO, and CoO are apparently noted, indicating that the severe heat treatment results in C burning, sintering, oxidation, and phase separation of the alloy catalysts. Moreover, the lattice constants, Co atomic fractions (XCo), and degrees of alloying (Da) [66] of as-prepared and various oxidized PdeCo catalysts are listed in Table S2 of Supplementary data. It is noted that as the heating temperature increases, the Da decreases, suggesting the preferential oxidation of Co and the collapse of the PdeCo alloy structure. On the other hand, similar results can also be found in oxidized Pd/C catalysts as shown in Fig. S1 (see Supplementary data). The presence of PdO phases is clearly detected at the oxidation temperature of 470e520 K. Moreover, intense oxidation treatment on the Pd/C at 620 K gives rise to the aggregation of PdO and the diffraction peaks belonging to metallic Pd are observed, which are the same as those of PdCo/C. The morphologies and particle size distribution histograms of various PdCo/C catalysts based on the TEM characterization are determined in Fig. 2. The inserts in the upper right of Fig. 2 (a), (c), (e), and (g) show the high magnification images of each

C

Pd O

Pd

CoO O620

x0.5 x0.5

XRD intensity

O520 O470 O420

O370 O320 x0.5

Pd3Co1/C

20 25 30 35 40 45 50 55 60 65 70

2θ Fig. 1 e XRD patterns of the as-prepared and various oxidized PdCo/C catalysts. The vertical dot line represents the peak position of (111) diffraction for metallic Pd.

samples. The PdeCo NPs for as-prepared catalysts are uniformly dispersed on the C support with the mean particle size of around 4.2  0.7 nm. Although high oxidation temperature is applied on the PdeCo catalysts, however, the variation in particle size and dispersion is not significant, revealing that the temperature of 470e520 K is not sufficient for generation of severe sintering phenomenon of NPs (the particle size for O470 and O520 is 4.9  1.0 nm and 5.2  1.2 nm, respectively). In contrast, the severe postthermal treatment at 620 K contributes to the dramatic grain growth and sintering of NPs and the size is around 7.9  1.1 nm. It has been reported by Huang et al. [31] that the particle size of PtRu NPs does not change obviously when oxidation temperature is below 520 K, nevertheless, the profound increase in particle size is noted at 620 K. Since the main emphasis of this study is to realize the evolution of surface composition after oxidation treatment at different temperatures, surface characterizations by XPS and TPR are shown below. Typically, the XPS spectra of Pd 3d for various oxidized PdeCo catalysts are depicted and their experimental (black solid line) and fitting results (open circle line) of metallic and oxidized Pd are compared in Fig. 3. For the as-prepared sample, 30 at% of surface Pd exists as the oxide phase which is not observed in the XRD pattern, suggesting that those oxide phase may exist as amorphous state [67]. As increasing heating temperature, the metallic Pd phase gradually diminishes and the PdO phase dominates on the surface. When the oxidation temperature is higher than 520 K, it is unambiguous to note that surface Pd is fully oxidized and consequently only one set of PdO peaks is found. Hereafter, we define a new term, degree of surface oxidation (DSO), based on the percentage of surface PdO (denoted as PdsO) characterized by the peak fitting results of XPS spectra in order to thoroughly realize the correlation between surface oxide species and ORR activities. The DSO of as-prepared, O470, and O520 sample is 30, 81, 100%, respectively. On the other hand, the XPS spectra of Co 2p for various oxidized PdeCo catalysts are compared in Fig. S2. Unlike the XPS spectra of Pd, the XPS spectra of Co are almost identical for as-prepared and oxidized PdeCo catalysts. Moreover, their surface Co with the binding energies of 793 and 778 eV exists mainly as Co2þ w Co3þ species, suggesting that the oxidation treatment does not change the chemical states of surface Co obviously. Since the surface sensitivity of XPS depends on the radiation used, the inelastic mean free path (IMFP) of electrons in this study may be around 2 nm, meaning that XPS analysis may not be completely surface sensitive. Therefore, the TPR characterization can offer relative micro observation on the evolution of topmost surface species during heating in air [57e63]. In terms of PdeCo system, the TPR traces are displayed in Fig. 4. From the profile for the as-prepared PdCo/C, the main hydrogen consumption peak due to Pd hydrogenation is noted at 265 K. The peak then progressively moves to higher temperature (ca. 300 K) and achieves a critical condition on O520 sample in which the surface is mainly composed of PdsO and its DSO value is 100%. On the other hand, the reduction of Co oxide species covered on the surface of catalysts may take place at different temperatures (the reduction temperature for CoOOH and Co3O4 is about 350e500 K and higher than 500 K, respectively) [45,68]. It is demonstrated that

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

3793

Fig. 2 e TEM micrographs and the corresponding particle size distribution histograms for the (a, b) as-prepared, (c, d) O470, (e, f) O520 and (g, h) O620 PdCo/C catalysts.

3794

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

H 2 consumption

Pd PdO

CoOOH

x1/7

O620

x1/7

O520

x1/2

O470 O420

O370 O320

Pd3Co1/C 250

300

350

400

450

500

T (K) Fig. 4 e TPR analyses of the as-prepared and various oxidized PdCo/C catalysts.

Fig. 3 e XPS spectra of Pd 3d for various PdCo/C catalysts. The surface compositions of Pd and PdO are also compared.

the synergistic effect of alloying with Co and surface Pd oxidation definitely affects the hydrogen-consumption characteristic of Pd [45]. In other words, the hydrogenation feature of Pd is suppressed by the addition of Co and/or formation of PdsO species. Moreover, most of the Co species, segregating out to the PdeCo alloy surface and existing as Co oxide (CoOOH) at 620 K, may be an undesired species for the ORR. Contrary to the PdeCo catalysts, the representative features of various oxidized Pd/C are shown in Fig. S3. It is also noted that the temperature of major hydrogen uptake gradually moves to 300 K and a corresponding strong peak is observed for O520 sample, indicative of the formation of enriched PdsO species or fully crystalline PdO phase. By means of the combined surface characterizations of XPS and TPR, the evolutions of chemical states for Pd from metallic to oxide phase within individual NPs can be thoroughly illustrated. In addition, the TPR results are consistent with the postulated DSO value as defined by XPS characterization. Besides XPS and TPR, the results of X-ray absorption near edge structure (XANES) can confirm the evolutions of chemical states from metallic Pd to Pd oxide within individual NPs of PdeCo catalysts at T > 320 K. The XANES spectra for Pd Kedge of various catalysts and Pd foil are displayed in Fig. S4(a). It is clearly noted that the increased white line (WL) intensity with increasing heating temperature is an indication of the formation of Pd oxides. The high WL intensity and the obvious

shift of edge absorption energy for O520 can be ascribed to the presence of well-constructed PdO. In the case of XANES spectra of Co K-edge shown in Fig. S4(b), the distinctive absorption hump intensity of all samples is higher than that of the Co foil, suggesting the formation of numerous Co oxides [69], which is consistent with the XPS spectra of Co 2p exhibited in Fig. S2. The electro-catalytic activities for various oxidized Pd and PdeCo catalysts are determined by LSV measurement recorded in O2 saturated 0.1 M HClO4 solution at rotation speed of 1600 rpm. Fig. 5(a) depicts the ORR activity of PdeCo alloy catalysts. Usually, two distinctive potential regions, the diffusion limiting region dominated by mass transportation of oxygenated species starting from 0.05 to 0.55 V and the mixed kinetic-diffusion control region that ideally reflects the catalytic activity of alloy catalysts ranging between 0.6 and 0.8 V, can be apparently detected [48,64]. Besides, the slight difference in diffusion limiting region may be attributed to different degrees of surface roughness induced by several factors such as the preparation routes, heat treatment temperatures, alloy compositions, etc [70e72]. On the other hand, the RDE measurements under various rotation speeds for as-prepared and O520 PdeCo samples are shown in Fig. S5. Based on the KouteckyeLevich equation, the calculated electron transfer numbers (n) at different rotation speeds listed in Table S3 for these two samples are very close to 4, implying that they have similar pathway for the ORR. In order to scrutinize the ORR activityeDSO value relationship, the comparison is summarized in Fig. 6. It is manifestly observed that the ORR activity is correlated to the DSO value. The improvement of ORR activity reaches an optimal value ca. 2.3 mA cm2 for the O520 sample where the complete formation of PdsO and the highest DSO value (100%) are attained. According to the reported literatures, the illustration of size dependent ORR activity for Pt catalysts is quite controversial. However, it is ubiquitously accepted that the increased oxophilicity for relatively small

3795

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

a

0

b

0

-2 -3 -4

0

-1 -2

-4 0.65

0.70

As-PdCo O320 O370 O420 O470 O520 O620

0.75

-5 -6 -7

i / mAcm -2

i / mAcm -2

-1

0

-2 -3

0.4

0.5 0.6 0.7 E / V vs. NHE

0.8

0.9

-4 0.65

0.70

0.75

-4

As-PdCo As-Pd Pd-520 PdCo-520

-5 -6 -7

0.3

-2

0.3

0.4

0.5 0.6 0.7 E / V vs. NHE

0.8

0.9

Fig. 5 e LSV results for (a) as-prepared and various oxidized PdCo/C catalysts and (b) as-prepared and O520 samples of PdCo/ C and Pd/C catalysts in O2 saturated 0.1 M HClO4 solution at a scan rate of 5 mVL1.

particles results in the decreased ORR activity because the strong adsorption of OH and/or oxygenated species on the topmost surface may therefore block the active sites required for the dissociative chemisorption of OeO bond [73]. It seems that the moderate or large NPs size is more effective to exhibit high performance relative to small one. In terms of that, more recently, Wang et al. [41] have prepared monodispersed PteCo NPs with size controlled from 3 to 9 nm for the cathode catalysts and the adequate particle size for displaying high ORR activity is found to be around 4.5 nm. Additionally, Liu et al. [74] have also proposed that for the promotion of activity and durability of PdeCo bimetallic catalysts, the pertinent annealing temperature at 623 K under reducing atmosphere is recommended due to the enhanced degrees of alloying and crystallinity. Here, the promotion of ORR activity on oxidized PdeCo catalysts may be ascribed to either the formation of PdO phase covered on the surface or large particle size effect. In contrast to O520 sample, further heating at high temperature cannot promote their ORR owing to the C burning and sintering of the catalysts. Moreover, from TPR and XRD characterizations, harsh oxidation temperature gives rise to the presence of CoOOH species driven by serious Co segregation and the formation of CoO phase which retards the overall catalysis

3

Pd3Co1/C

100

60

2

40 1

Pd DSO (%)

I 07 / mAcm

-2

80

20

0 0 As

O320 O370

O420

O470

O520

O620

Fig. 6 e The correlation of ORR activity and DSO for various PdCo/C catalysts.

reaction [45]. On the other hand, the similar results on the enhancement of ORR activity for the Pd/C catalysts promoted by oxidation treatment can also be exhibited in Fig. 5(b). It is important to mention that the ORR enhancement of PdeCo due to oxidation treatment is more significant than that of Pd. The durability of as-prepared and O520 samples of Pd/C and PdCo/C catalysts is determined by CV measurement and shown in Fig. 7. For the Pd/C catalysts, the distinguishable difference in hydrogen adsorption/desorption region (Hads/des) of as-prepared and O520 samples is observed, implying that the presence of PdsO species may hinder the dissolution of hydrogen into bulk Pd/C. Besides, the onset potential of cathodic peak on O520 catalyst exhibits a slightly positive shift, relative to the as-prepared one. After 50 cycles, the substantial drop in the cathodic peak current density and Hads/des region is found, nevertheless, the cathodic current density and onset potential of the 50th cycle for the O520 is still larger and more positive than that for the as-prepared one, suggesting the sample after oxidation treatment demonstrates the remarkable durability with respect to the non-treated one. This observation is attributed to the formation of PdsO phase which indirectly influences the chemisorption of OH on the Pd sites at high potentials [75,76]. On the other hand, the similar phenomenon for the oxidized PdCo/C catalysts is also noted. As is known that Co or Co oxides may be stripped in acidic solutions, and the effect of the incorporation of Co may become insignificant and the Pd oxide layer seems to play a dominant role in determining the performance. Therefore, the relatively high electrochemical performance of the oxidized PdeCo sample may be owing to the formation of PdO layer and/or structure. The alterations of surface conditions for various oxidized PdeCo catalysts after the electrochemical measurements catalysts attract our interests and the illustrative XPS traces are presented in Fig. 8. Their experimental (black solid line) and fitting results (open circle line) of metallic and oxidized Pd are also compared. For the as-prepared sample, the amount of PdsO increases from 30 to 57%, suggesting that during the ORR in acid environment, metallic Pd gradually disappears and the concomitant increase in PdO is found in excess [77]. On the contrary, the surface composition of O520 sample completely covered by

3796

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

1

50th cycle

50th cycle

1st cycle

1st cycle

0

i / mAcm

-2

-1

2 -2

0

-2

Pd/C

Pd3Co1/C

O520

O520

-4 0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

E / V vs. NHE Fig. 7 e CV scans for the as-prepared and O520 samples of Pd/C and PdCo/C catalysts in N2-saturated 0.1 M HClO4 solution at a scan rate of 20 mVL1.

Experimental

Pd:PdO

Fitting line

0:100

O620

0:100

XPS intensity

O520

13:87

O470 O420

29:71 O370

37:63 O320

52:48

43:57

As 334

336

338

340

342

344

Binding energy / eV

Fig. 8 e XPS spectra of Pd 3d for various PdCo/C catalysts after the ORR electrochemical measurements.

PdsO still remains unchanged owing to the high stability of the metal oxide. Very recently, due to the high dissolution resistance of TiO2 in acid environment, growing efforts have been focused on the development of TiO2 as an enabling support with the aim to improve the long-term stability on cathode catalysts [78,79]. In our study, the PdsO species may play a resembled role as the metal oxide support that can resist the corrosion in acid electrolyte and thereby enhance the durability. Thus, the PdsO may not only enhance the ORR activity but protect the catalysts from structural changes, and their durability can be also promoted accordingly. An explicative model based on our results for the enhancement of ORR activity caused by the synergistic effect of Co alloying and surface oxidation is provided in Fig. 9. Based on the thermodynamic model, it is suggested that the incorporation of Co will facilitate the dissociative adsorption of O2 and the thus-produced Oads can migrate from the Co site to the Pd site [11,18]. When the PdeCo is immersed into the O2saturated electrolyte, the oxygen molecules in the electrolyte diffuse and chemisorb onto Co surface oxide through the following equation [59,68]. 1 CoOx þ O2 /CoOx eOads 2

(1)

The surface oxides can subtly enhance the dissociation of OeO bond and the CoOxeOads species spillover instantly to the PdsOx site where the ORR occurs [11,18]. CoOx eOads þ Pds Ox /CoOx þ Pds Ox eOads

(2)

Pds Ox eOads þ 2Hþ þ 2e /Pds Ox þ H2 O

(3)

3797

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

Fig. 9 e An explicative model of the improved ORR activity arisen from the synergistic effect of Co alloying and surface oxidation.

3.2. Relationship between surface species and ORR activity on various heat treated PdCo/C catalysts According to the reported literatures, the influence of NPs size on their ORR activity must be taken into account. In other words, it is of great significance to distinguish that if the promotion effect arises from the presence of PdsO or from the large size. Moreover, since the oxidation treatment performed in air has positive effect on the ORR performance of PdeCo catalysts, the effect of heat treatment by N2 cannot be negligible. Hence, the PdeCo alloy catalysts are subjected into the following annealing treatments at 520 K under CO or H2 atmospheres without changing their particle size significantly. The phases and structures of various heat treated samples are determined by XRD characterization as shown in Fig. 10. Obviously, for N520 sample, its (111) diffraction peak is relatively sharp when compared with CO520 and H520 samples, indicative of the meaningful increase in crystallinity and size of NPs. By using Scherrer’s equation, the grain size for CO520, H520 and N520 is 4.1, 4.3 and 6.8 nm, respectively [83]. Besides, the mean grain size for O520 is around 4.4 nm, which is close to that for H520 and CO520, suggesting that those samples (except N520) can be used to illustrate the correlation between surface composition and ORR activity of the catalysts. To further assess the morphologies and size distribution histograms of CO520 and H520 catalysts, the TEM characterization is presented in Fig. 11. The inserts in the upper right of Fig. 11(a) and (c) show the high magnification images

of each sample. It can be clearly seen that all the alloy particles are well-dispersed onto the carbon support along with narrow size distribution after the heat treatment. Besides, the normal particle diameter of the CO520 and H520 catalysts is about 5.1  0.9 and 5.3  1.1 nm, respectively, which is determined via measuring up to more than 200 randomly chosen particles in TEM images. The particle size and distribution for these two samples is in a good agreement with that for the oxidized one. Besides, the microstructure of N520 sample is also provided in Fig. S6 and its average particle diameter is about 7.2  1.3 nm. Similarly, both the XPS and TPR techniques are performed on H520 and CO520 catalysts to investigate their surface

C

PdCo

PdO

CoO

O520

XRD intensity

Moreover, it has been reported that the addition of some promoters and/or oxygen containing species into the cathode catalysts can facilitate the supply of oxygen species, accelerate the adsorption and desorption of oxygen, and lead to an enhanced ORR performance [80e82]. Here, the surface oxides species may be recognized as the medium as mentioned above to provide the sufficient oxygen which is prerequisite for the catalysis reaction. Consequently, it can be inferred that the pronounced enhancement of ORR arises from the combination of spillover effect and oxygen containing ability of oxide species.

x2 CO520

x2 H520

N520 20

25

30

35

40

45

50

55

60

65

70

2θ Fig. 10 e XRD patterns of the various atmospheres heat treated PdCo/C catalysts at 520 K.

Fig. 11 e TEM micrographs and the corresponding particle size distribution histograms for the (a, b) CO520 and (c, d) H520 PdCo/C catalysts.

a

b

Experimental Fitting line

O520

Pd

PdO

CoOOH

Pd:PdO 0:100

H520

x10

60:40

H 2 consumption

XPS intensity

O520

H520

CO520

72:28

x15

332 334 336 338 340 342 344 346 348

Binding energy / eV

CO520 250

300

350

400

450

500

T (K)

Fig. 12 e Surface characterizations probed by the techniques of XPS and TPR. (a) XPS spectra of Pd 3d and (b) TPR analyses of various heat treated PdCo/C catalysts.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

0 CO520 H520 O520

-1

i / mAcm

-2

-2 -3 -4 -5 -6 -7 0.3

0.4

0.5

0.6

0.7

0.8

0.9

E / V vs. NHE Fig. 13 e LSV results for various PdCo/C catalysts treated under CO, H2 and O2 atmospheres at 520 K in O2 saturated 0.1 M HClO4 solution at a scan rate of 5 mVL1.

compositions. As evidenced from XPS spectra of Pd 3d depicted in Fig. 12 (a), the DSO value for H520 and CO520 sample is 40 and 28%, respectively, implying that surface Pd exists as metallic state mainly instead of oxide. Our previous study shows that the introduction of H2 treatment on PteAu catalysts can result in not only the surface segregation of Au but also the enrichment of metallic Pt or Au on the surface as compared to nontreated ones [22]. Here, once the PdeCo catalysts are subjected to reducing atmosphere, the enriched metallic Pd on the surface can be found. On the other side, the TPR features shown in Fig. 12(b) exhibit that the surface of H520 and CO520 sample is comprised of Pd and PdO species, whereas, the topmost surface of O520 one almost completely consists of PdO species, confirming that only the oxidation treatment can generate the PdO species enriched on the surface of NPs. Besides, it should be mentioned that those analyses are under ex-situ conditions, and theoretically the oxide species may not be found after H2 treatment. However, during sample handling in air, some surface oxide may be formed on those samples. Therefore, the in-situ measurements can get insight into the effect of different treatments on the surfaces and structures of catalysts precisely and perfectly [84e86]. Fig. 13 shows the typical ORR polarization curves for various heat treated PdeCo catalysts in O2 saturated 0.1 M HClO4. It can be inspected within the ORR kinetic region that the activity on O520 sample is slightly higher than that of CO520 and H520, clearly implying that the oxide phase is favorable for accelerating the ORR through the synergistic mechanism proposed above. Here, the promotion effect on ORR activity for oxidized samples by the size effect is excluded, therefore, it can be deduced that the modification of ORR activity is dominantly attributed to the presence of PdsO species with high DSO value.

4.

Conclusion

The effect of oxidation treatment on the improvement of ORR activity and durability for the Pd/C and PdCo/C catalysts has

3799

been systematically investigated. By virtue of the XPS technique, an index, DSO has been defined to explicitly illustrate the relationship between electrochemical property and the oxidation temperature. On the other hand, the TPR technique further provides the insight into the surface species of near surface regions so as to realize the evolution of chemical states for Pd and PdeCo during heating in air. From our observations, the catalytic activity and durability of Pd catalysts can be well-promoted not only by Co alloying but also by oxidation treatment. The optimum oxidation condition for the enhancement of ORR performance is at 520 K, where the surface of catalysts is mainly covered by PdO and the DSO is 100%. Besides, a model is provided according to our experimental results to illustrate this correlation between oxidative thermal treatment and ORR performance for PdeCo alloy catalysts. In order to verify the ORR enhancement is solely attributed to the PdsO species, various heat treatment processes by either utilizing inert or reducing atmospheres are applied on the PdeCo system without changing their particle size. While the PdeCo catalysts are subjected into air, CO and H2 at 520 K, the particle size of catalysts does not change significantly and the size is about 5.1 nm. The air-treated PdCo/C catalysts show the superior performance to CO and H2-treated ones, suggesting that the promotion of ORR is mainly ascribed to the surface species effect instead of size effect. Thus, for the modification of PdeCo catalysts, the oxidative thermal treatment is a promising process to attain high ORR activity and durability.

Acknowledgment This work was supported by the National Science Council of Taiwan under contract no. NSC-99-2221-E-008-058.

Appendix. Supplementary data Supplementary data related to this article can be found online at doi:10.1016/j.ijhydene.2010.12.098.

references

[1] Appleby AJ, Foulkes FR. Fuel cell handbook. New York: Van Nostrand Reinhold; 1898. [2] Schmid G, editor. Clusters and colloids from theory to applications. Weinheim, Germany: VCH; 1994. [3] Shao MH, Huang T, Liu P, Zhang J, Sasaki K, Vukmirovic MB, et al. Palladium monolayer and palladium alloy electrocatalysts for oxygen reduction. Langmuir 2006;22: 10409e15. [4] Shao MH, Sasaki K, Adzic RR. PdeFe nanoparticles as electrocatalysts for oxygen reduction. J Am Chem Soc 2006; 128:3526e7. [5] Stamenkovic VR, Fowler B, Mun BS, Wang G, Ross PN, Lucas CA, et al. Improved oxygen reduction activity on Pt3Ni (111) via increased surface site availability. Science 2007;315: 493e7.

3800

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

[6] Gasteiger HA, Kocha SS, Sompalli B, Wagner FT. Activity benchmarks and requirements for Pt, Pt-alloy, and non-Pt oxygen reduction catalysts for PEMFCs. Appl Catal B Environ 2005;56:9e35. [7] Jarvi TD, Stuve S, Sriramulu EM. Potential dependence of the yield of carbon dioxide from electrocatalytic oxidation of methanol on platinum (100). J Phys Chem B 1997;101: 3649e52. [8] Kadirgan F, Kannan AM, Atilan T, Beyhan S, Ozenler SS, Suzer S, et al. Carbon supported nano-sized PtePd and PteCo electrocatalysts for proton exchange membrane fuel cells. Int J Hydrogen Energy 2009;34:9450e60. [9] Xu JB, Zhao TS, Yang WW, Shen SY. Effect of surface composition of PteAu alloy cathode catalyst on the performance of direct methanol fuel cells. Int J Hydrogen Energy 2010;35:8699e706. [10] Grigoriev SA, Lyutikova EK, Martemianovb S, Fateev VN. On the possibility of replacement of Pt by Pd in a hydrogen electrode of PEM fuel cells. Int J Hydrogen Energy 2007;32: 4438e42. [11] Fernandez JL, Walsh DA, Bard AJ. Thermodynamic guidelines for the design of bimetallic catalysts for oxygen electroreduction and rapid screening by scanning electrochemical microscopy. MeCo (M: Pd, Ag, Au). J Am Chem Soc 2005;127:357e65. [12] Liu H, Manthiram A. Tuning the electrocatalytic activity and durability of low cost Pd70Co30 nanoalloy for oxygen reduction reaction in fuel cells. Electrochem Commun 2008; 10:740e4. [13] Liu H, Manthiram A. Controlled synthesis and characterization of carbon-supported Pd4Co nanoalloy electrocatalysts for oxygen reduction reaction in fuel cells. Energy Environ Sci 2009;2:124e32. [14] Tominaka S, Momma T, Osaka T. Electrodeposited PdeCo catalyst for direct methanol fuel cell electrodes: preparation and characterization. Electrochim Acta 2008;53:4679e86. [15] Shao M, Liu P, Zhang J, Adzic R. Origin of enhanced activity in palladium alloy electrocatalysts for oxygen reduction reaction. J Phys Chem B 2007;111:6772e5. [16] Lee K, Savadogo O, Ishihara A, Mitsushima S, Kamiya N, Ota KI. Methanol-tolerant oxygen reduction electrocatalysts based on Pd-3D transition metal alloys for direct methanol fuel cells. J Electrochem Soc 2006;153:A20e4. [17] Zhang L, Lee K, Zhang J. The effect of heat treatment on nanoparticle size and ORR activity for carbon-supported PdeCo alloy electrocatalysts. Electrochim Acta 2007;52: 3088e94. [18] Suo Y, Zhuang L, Lu J. First-principles considerations in the design of Pd-alloy catalysts for oxygen reduction. Angew Chem Int Ed 2007;46:2862e4. [19] Wang KW, Huang SY, Yeh CT. Promotion of carbonsupported platinumeruthenium catalyst for electrodecomposition of methanol. J Phys Chem C 2007;111: 5096e100. [20] Wei YC, Liu CW, Wang KW. Activityestructure correlation of Pt/Ru catalysts for the electrode composition of methanol: the importance of RuO2 and PtRu alloying. ChemPhysChem 2009;10:1230e7. [21] Li X, Hsing IM. Surfactant-stabilized PtRu colloidal catalysts with good control of composition and size for methanol oxidation. Electrochim Acta 2006;52:1358e65. [22] Liu CW, Wei YC, Wang KW. Promotion of ceria-modified PteAu/C cathode catalysts for oxygen reduction reaction by H2-induced surface segregation. Chem Commun 2010;46: 2483e5. [23] Tsiouvaras N, Martinez-Huerta MV, Paschos O, Stimming U, Fierro JLG, Pena MA. PtRuMo/C catalysts for direct methanol fuel cells: effect of the pretreatment on the structural

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

characteristics and methanol electrooxidation. Int J Hydrogen Energy 2010;35:11478e88. Kim DB, Chun HJ, Lee YK, Kwon HH, Lee HI. Preparation of Pt/ NiOeC electrocatalyst and heat-treatment effect on its electrocatalytic performance for methanol oxidation. Int J Hydrogen Energy 2010;35:313e20. Mayrhofer KJJ, Juhart V, Hartl K, Hanzlik M, Arenz M. Adsorbate-induced surface segregation for coreeshell nanocatalysts. Angew Chem Int Ed 2009;48:3529e31. Hwang BJ, Sarma LS, Wang GR, Chen CH, Liu DG, Sheu HS, et al. Heat-induced alterations in the surface population of metal sites in bimetallic nanoparticles. Chem Eur J 2007;13: 6255e64. Lee KS, Park HY, Cho YH, Park IS, Yoo SJ, Sung YE. Modified polyol synthesis of PtRu/C for high metal loading and effect of post-treatment. J Power Sources 2010;195:1031e7. Raman RK, Shukla AK, Gayen A, Hegde MS, Priolkar KR, Sarode PR, et al. Tailoring a PteRu catalyst for enhanced methanol electro-oxidation. J Power Sources 2006;157:45e55. Wei YC, Liu CW, Chang WJ, Wang KW. Promotion of PteRu/C catalysts driven by heat treated induced surface segregation for methanol oxidation reaction. J Alloys Compd 2011;509: 535e41. Babu PK, Kim HS, Kuk ST, Chung JH, Oldfield E, Wieckowski A. Activation of nanoparticle PteRu fuel cell catalysts by heat treatment: a 195Pt NMR and electrochemical study. J Phys Chem B 2005;109:17192e6. Huang SY, Chang SM, Lin CL, Chen CH, Yeh CT. Promotion of the electrochemical activity of a bimetallic platinumeruthenium catalyst by oxidation-induced segregation. J Phys Chem B 2006;110:23300e5. Huang SY, Chang CM, Wang KW, Yeh CT. Promotion of platinumeruthenium catalyst for electrooxidation of methanol by crystalline ruthenium dioxide. ChemPhysChem 2007;8:1774e7. Yeh TK, Tsai CH, Wu MT. The effect of oxygen pretreatment on the morphological and electrochemical characteristics of a direct methanol fuel cell anode of different structures. Mater Chem Phys 2009;114:199e203. Sellin R, Grolleau C, Arrii-Clacens S, Pronier S, Clacens JM, Coutanceau C, et al. Effects of temperature and atmosphere on carbon-supported platinum fuel cell catalysts. J Phys Chem C 2009;113:21735e44. Jeon TY, Yoo SJ, Cho YH, Lee KS, Kang SH, Sung YE. Influence of oxide on the oxygen reduction reaction of carbonsupported PteNi alloy nanoparticles. J Phys Chem C 2009;113: 19732e9. Min MK, Cho JH, Cho KW, Kim H. Particle size and alloying effects of Pt-based alloy catalysts for fuel cell applications. Electrochim Acta 2000;45:4211e7. Shukla AK, Neergat M, Bera P, Jayaram V, Hegde MS. An XPS study on binary and ternary alloys of transition metals with platinized carbon and its bearing upon oxygen electroreduction in direct methanol fuel cells. J Electroanal Chem 2001;504:111e9. Chen W, Ny D, Chen S. SnO2eAu hybrid nanoparticles as effective catalysts for oxygen electroreduction in alkaline media. J Power Sources 2010;195:412e8. Baker WS, Pietron JJ, Teliska ME, Bouwman PJ, Ramaker DE, Swider-Lyons KE. Enhanced oxygen reduction activity in acid by tin-oxide supported Au nanoparticle catalysts. J Electrochem Soc 2006;153:A1702e7. Hwang BJ, Kumar SMS, Chen CH, Monalisa, Cheng MY, Liu DG, et al. An investigation of structureecatalytic activity relationship for PteCo/C bimetallic nanoparticles toward the oxygen reduction reaction. J Phys Chem C 2007;111:15267e76. Wang C, Vliet DVD, Chang KC, You H, Strmcnik D, Schlueter JA, et al. Monodisperse Pt3Co nanoparticles as

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56] [57]

[58]

[59]

a catalyst for the oxygen reduction reaction: size-dependent activity. J Phys Chem C 2009;113:19365e8. Nores-Pondal FJ, Vilella IMJ, Troiani H, Granada M, DeMiguel SR, Scelza OA, et al. Catalytic activity vs. size correlation in platinum catalysts of PEM fuel cells prepared on carbon black by different methods. Int J Hydrogen Energy 2009;34:8193e203. Axnanda S, Cummins KD, He T, Goodman DW, Soriaga MP. Structural, compositional and electrochemical characterization of PteCo oxygen-reduction catalysts. ChemPhysChem 2010;11:1468e75. Chen S, Ferreira PJ, Sheng W, Yabuuchi N, Allard LF, Horn YS. Enhanced activity for oxygen reduction reaction on “Pt3Co” nanoparticles: direct evidence of percolated and sandwichsegregation structures. J Am Chem Soc 2008;130:13818e9. Wei YC, Liu CW, Chang YW, Wang KW, Lai CM, Lim PY, et al. The structureeactivity relationship of PdeCo/C electrocatalysts for oxygen reduction reaction. Int J Hydrogen Energy 2010;35:1864e71. Stamenkovic V, Schmidt TJ, Ross PN, Markovic NM. Surface segregation effects in electrocatalysis: kinetics of oxygen reduction reaction on polycrystalline Pt3Ni alloy surfaces. J Electroanal Chem 2003;554e555:191e9. Ruban AV, Skriver HL, Norskov JK. Surface segregation energies in transition-metal alloys. Phys Rev B 1999;59: 15990e6000. Stamenkovic VR, Mun BS, Arenz M, Mayrhofer KJJ, Lucas CA, Wang G, et al. Trends in electrocatalysis on extended and nanoscale Pt-bimetallic alloy surfaces. Nat Mater 2007;6:241e7. Norskov JK, Rossmeisl J, Logadottir A, Lindqvist L, Kitchin JR, Bligaard T, et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J Phys Chem B 2004;108: 17886e92. Xu Y, Ruban AV, Mavrikakis M. Adsorption and dissociation of O2 on PteCo and PteFe alloys. J Am Chem Soc 2004;126: 4717e25. Takasu Y, Ohashi N, Zhang XG, Murakami Y, Minagawa H, Sato S, et al. Size effects of platinum particles on the electroreduction of oxygen. Electrochim Acta 1996;41: 2595e600. Mukerjee S, Mc-Breen J. Effect of particle size on the electrocatalysis by carbon-supported Pt electrocatalysts: an in situ XAS investigation. J Electroanal Chem 1998;448: 163e71. Mayrhofer KJJ, Blizanac BB, Arenz M, Stamenkovic VR, Ross PN, Markovic NM. The impact of geometric and surface electronic properties of Pt-catalysts on the particle size effect in electrocatalysis. J Phys Chem B 2005;109:14433e40. Maillard F, Martin M, Gloaguen F, Leger JM. Oxygen electroreduction on carbon-supported platinum catalysts: particle-size effect on the tolerance to methanol competition. Electrochim Acta 2002;47:3431e40. Yano H, Inukai J, Uchida H, Watanabe M, Babu PK, Kobayashi T, et al. Particle-size effect of nanoscale platinum catalysts in oxygen reduction reaction: an electrochemical and 195Pt EC-NMR study. Phys Chem Chem Phys 2006;8: 4932e9. Briggs D, Seah MP. Practical surface analysis, vol. 1. New York: Wiley; 1990. Huang SY, Chang SM, Yeh CT. Characterization of surface composition of platinum and ruthenium nanoalloys dispersed on active carbon. J Phys Chem B 2006;110:234e9. Liu CW, Wei YC, Wang KW. Preparation and surface characterization of PteAu/C cathode catalysts with ceria modification for oxygen reduction reaction. Electrochem Commun 2009;11:1362e4. Chou CW, Chu SJ, Chiang HJ, Huang CY, Lee CJ, Sheen SR, et al. Temperature-programmed reduction study on

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78]

3801

calcination of nano-palladium. J Phys Chem B 2001;105: 9113e7. Wang KW, Chung SR, Liu CW. Surface segregation of PdxNi100x alloy nanoparticles. J Phys Chem C 2008;112: 10242e6. Liu CW, Wei YC, Wang KW. Preparation and characterization of carbon-supported PteAu cathode catalysts for oxygen reduction reaction. J Colloid Interface Sci 2009;336:654e7. Chang YW, Liu CW, Wei YC, Wang KW. Promotion of PtRu/C anode catalysts for ethanol oxidation reaction by addition of Sn modifier. Electrochem Commun 2009;11:2161e4. Wei YC, Liu CW, Wang KW. Surface species alteration and oxygen reduction reaction enhancement of PdeCo/C electrocatalysts induced by ceria modification. ChemPhysChem 2010;11:3078e85. Wang W, Zheng D, Du C, Zou Z, Zhang X, Xia B, et al. Carbon supported PdeCo bimetallic nanoparticles as electrocatalysts for the oxygen reduction reaction. J Power Sources 2007;167:243e9. Sarkar A, Murugan AV, Manthiram A. Synthesis and characterization of nanostructured PdeMo electrocatalysts for oxygen reduction reaction in fuel cells. J Phys Chem C 2008;112:12037e43. Raghuveer V, Ferreira PJ, Manthiram A. Comparison of PdeCoeAu electrocatalysts prepared by conventional borohydride and microemulsion methods for oxygen reduction in fuel cells. Electrochem Commun 2006;8:807e14. Vogel W, He W, Huang QH, Zou Z, Zhang XG, Yang H. Palladium nanoparticles “breathe” hydrogen; a surgical view with X-ray diffraction. Int J Hydrogen Energy 2010;35: 8609e20. Wu RJ, Wu JG, Tsai TK, Yeh CT. Use of cobalt oxide CoOOH in a carbon monoxide sensor operating at low temperatures. Sens Actuators B 2006;120:104e9. Lai FJ, Su WN, Sarma LS, Liu DG, Hsieh CA, Lee JF, et al. Chemical dealloying mechanism of bimetallic PteCo nanoparticles and enhancement of catalytic activity toward oxygen reduction. Chem Eur J 2010;16:4602e11. Kant R, Rangarajan SKJ. Effect of surface roughness on diffusion-limited charge transfer. J Electroanal Chem 1994; 368:1e21. Luo J, Njoki PN, Lin Y, Wang L, Zhong CJ. Activity composition correlation of AuPt alloy nanoparticle catalysts in electrocatalytic reduction of oxygen. Electrochem Commun 2006;8:581e7. Kumar SS, Phani KLN. Exploration of unalloyed bimetallic AuePt/C nanoparticles for oxygen reduction reaction. J Power Sources 2009;187:19e24. Mayrhofer KJJ, Strmcnik D, Blizanac BB, Stamenkovic V, Arenz M, Markovic NM. Measurement of oxygen reduction activities via the rotating disc electrode method: from Pt model surfaces to carbon-supported high surface area catalysts. Electrochim Acta 2008;53:3181e8. Liu H, Li W, Manthiram A. Factors influencing the electrocatalytic activity of Pd100  xCox (0 & x & 50) nanoalloys for oxygen reduction reaction in fuel cells. Appl Catal B Environ 2009;90:184e94. Li X, Huang Q, Zou Z, Xia B, Yang H. Low temperature preparation of carbon-supported PdeCo alloy electrocatalysts for methanol-tolerant oxygen reduction reaction. Electrochim Acta 2008;53:6662e7. Anderson AB. O2 reduction and CO oxidation at the Ptelectrolyte interface. The role of H2O and OH adsorption bond strengths. Electrochim Acta 2002;47:3759e63. Burke LD, Casey JK. An examination of the electrochemical behavior of palladium electrodes in acid. J Electrochem Soc 1993;140:1284e91. Huang SY, Ganesan P, Park S, Popov BN. Development of a titanium dioxide-supported platinum catalyst with

3802

[79]

[80]

[81]

[82]

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 7 8 9 e3 8 0 2

ultrahigh stability for polymer electrolyte membrane fuel cell applications. J Am Chem Soc 2009;131:13898e9. Lim DH, Lee WJ, Wheldon J, Macy NL, Smyrl WH. Electrochemical characterization and durability of sputtered Pt Catalysts on TiO2 nanotube arrays as a cathode material for PEFCs. J Electrochem Soc 2010;157:B862e7. Lim DH, Lee WD, Choi DH, Kwon HH, Lee HI. The effect of cerium oxide nanoparticles on a Pt/C electrocatalyst synthesized by a continuous two-step process for lowtemperature fuel cell. Electrochem Commun 2008;10:592e6. Yu HB, Kim JH, Lee HI, Scibioh MA, Lee J, Han J, et al. Development of nanophase CeO2ePt/C cathode catalyst for direct methanol fuel cell. J Power Sources 2005;140:59e65. Lee KH, Kwon K, Roev V, Yoo DY, Chang H, Seung D. Synthesis and characterization of nanostructured

[83] [84]

[85]

[86]

PtCoeCeOx/C for oxygen reduction reaction. J Power Sources 2008;185:871e5. Warren BE. X-ray diffraction. New York: Dover Publications, Inc.; 1990. Vogel W, Rhun VL, Garnier E, Alonso-Vante N. Ru clusters synthesized chemically from dissolved carbonyl: in situ study of a novel electrocatalyst in the gas phase and in electrochemical environment. J Phys Chem B 2001;105:5238e43. Vogel W. Size contraction in Pt/C and PtRu/C commercial ETEK electrocatalysts: an in situ X-ray diffraction study. J Phys Chem C 2008;112:13475e82. Vogel W, Kaghazchi P, Jacob T, Alonso-Vante N. Genesis of RuxSey nanoparticles by pyrolysis of Ru4Se2(CO)11: a combined X-ray in situ and DFT study. J Phys Chem C 2007; 111:3908e13.