ww.sciencedirect.com

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 en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 0

Available online at w

ScienceDirect

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

Tungsten carbide synthesized by low-temperaturecombustion as gas diffusion electrode catalyst

Ping Li a,*, Zhiwei Liu a, Liqun Cui a, Fuqiang Zhai b, Qi Wan a, Ziliang Li a,Zhigang Zak Fang c, Alex A. Volinsky d, Xuanhui Qu a,*a Institute for Advanced Materials and Technology, University of Science and Technology Beijing, Beijing 100083,

ChinabDepartament Fı́sica Aplicada, EETAC, Universitat Politècnica de Catalunya e BarcelonaTech, 08860 Castelldefels,

SpaincDepartment of Metallurgical Engineering, University of Utah, 135 South 1460 East, Salt Lake City, UT 84112, USAdDepartment of Mechanical Engineering, University of South Florida, Tampa, FL 33620, USA

a r t i c l e i n f o

Article history:

Received 9 December 2013

Received in revised form

18 April 2014

Accepted 23 April 2014

Available online 2 June 2014

Keywords:

Low-temperature combustion syn-

thesis

Tungsten carbide

Electrocatalyst

Gas diffusion electrode

* Corresponding authors. Tel.: þ86 10 823772E-mail addresses: ustbliping@126.com (P.

http://dx.doi.org/10.1016/j.ijhydene.2014.04.10360-3199/Copyright ª 2014, Hydrogen Ener

a b s t r a c t

Tungsten carbide powder, which is used as the catalyst for a gas diffusion electrode, has

been prepared by low-temperature combustion synthesis for the first time. The average

particle size of the prepared tungsten carbide is 200 nm, determined by X-ray diffraction

and field-emission scanning electron microscopy. The effects of the carbon/tungsten (C/W)

molar ratio on the formation of tungsten carbide and carbon content on the complete

carbonization temperature are discussed. The optimal synthesis temperature is 1100 �C,

and the optimal C/W molar ratio is 19/3. The electrocatalytic properties of tungsten carbide

for the oxygen reduction reaction are evaluated through the use of polarization curves and

electrochemical impedance spectroscopy in neutral and alkaline electrolytes. The current

density of the tungsten carbide-based gas diffusion electrode is as high as 350 mA cm�2 at

0.4 V versus Hg/HgO. It is demonstrated that the tungsten carbide catalyst exhibits

excellent electrocatalytic performance, comparable with that of Pt.

Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rightsreserved.

Introduction

In order to improve the adverse effects of fossil fuels burning

on the environment, and reduce dependence on fossil fuels,

development of pure electric and fuel cell vehicles has become

a much sought after goal around the world. The metaleair

battery is an ideal replacement for traditional batteries, and

has attracted attention due to the relative abundance of its

86; fax: þ86 10 62334311.Li), quxh@ustb.edu.cn (X73gy Publications, LLC. Publ

source materials, simple structure, high specific power and

energy density. However, until now, due to the lack of highly

efficient low-cost electrode catalysts for the oxygen reduction

cathode, metaleair batteries have been only used in small-

scale special applications, such as pagers and hearing aids,

indicating that the development of amore effective,metaleair

battery catalyst could have much wider application.

Oxygen reduction cathodes for the metaleair battery are

usually gas diffusion electrodes, and the choice of catalyst for

. Qu).

ished by Elsevier Ltd. All rights reserved.

mailto:ustbliping@126.commailto:quxh@ustb.edu.cnhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.ijhydene.2014.04.173&domain=pdfwww.sciencedirect.com/science/journal/03603199www.elsevier.com/locate/hehttp://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 010912

the gas diffusion electrode is largely responsible for the bat-

tery performance. At present, the catalysts for the gas diffu-

sion electrode mainly include: (1) noble metals, such as Pt [1],

Ag [2], PteAu [3], Pd49Pt47Co4 [4], PteCo [5], and PteRueCo [6];

(2) metal oxides, such as Co3O4 [7], MnOOH [8], and MnO2 [9];

(3) mixedmetal oxides, such as NixCo3�xO4 [10], La0.6Ca0.4CoO3[11], La0.1Ca0.9MnO3 [11], LaNiO3 [11], LaNi0.8Co0.2O3 [12], and

MnxCo3�xO4 [13]; (4) macrocyclic compounds, such as cobalttetramethoxy phenylporphyrin (CoTMPP) [14], cobalt phtha-

locyanine (CoPc) [15], and iron phthalocyanine (FePc) [16].

Among these catalysts, platinum is the most suitable elec-

trocatalyst for the oxygen reduction reaction (ORR) due to its

high electrocatalytic activity and chemical stability. However,

there are several obstacles to utilizing Pt catalysts in most

practical applications, including high cost and easily suscep-

tible to poisoning by carbon monoxide [17]. Improving the

properties of oxygen reduction cathode materials and seeking

inexpensive and efficient catalysts has become a major focal

point in the field of metaleair battery research.

Tungsten carbide (WC) is a promising catalytic material for

the gas diffusion electrode, since its catalytic behavior re-

sembles platinum [18], but its stability [19e21], anti-toxic and

oxidation resistance are much higher than those of platinum

[22]. Mustain et al. [23] studying the stability of Pt/WO3 in acid

media found the sequential electrochemical oxidation of WC

to WOx and WO3 at E >0.8 V followed by the formation and

dissolution of HxWO3. The proposed degradation mechanism

for WC and WO3 demonstrated that as long as the support

surface is exposed to the acidic electrolyte, neither represents

a long-term stable support material for Pt electrocatalysts.

The formed nonconductive WO3 could isolate Pt particles by

coating on their surface, leading to electrochemically inac-

cessible Pt particles. But the presence of Pt on WC surface can

stabilize WC and further inhibit WC oxidation, which is

consistent with the results reported in the literature [24,25].

Furthermore, Mark et al. [26] compared the stability of the

most commonly used carbides in electrochemical applica-

tions: tungsten carbides (WC and W2C) and molybdenum

carbide (MO2C) in electrolytic solutions by varying pH values,

whereWC exhibits the largest region of stability at a relatively

lower pH value. It has been reported that tungsten carbide

exhibits high catalytic activity in electro-catalysis [27], and is a

promising material for hydrogen evolution reactions and

hydrogen oxidation reactions in electro-catalysis [28]. Pt

nanoparticles supported by WC substrate show remarkable

catalytic activity for ORR [29], has anti-poisoning properties

for carbon monoxide in methanol electro-oxidation, and ex-

hibits improved methanol oxidation performance [30]. Addi-

tionally, tungsten carbide particles as a counter-electrode for

dye-sensitized solar cells have been shown to improve cata-

lytic activity for iodide reduction [31,32], and when combined

with titania in nanocomposites, has shown synergistic effects

for electrocatalysts [33]. The infiltrated WCeYSZ (yttrium

stabilized zirconia), as a potential anode for direct methane-

fueled solid oxide fuel cells (SOFCs), performed stably with

no catastrophic degradation at 800e900 �C [34]. However,catalytic activity of WC is lower than Pt.

It was confirmed that the preparation method and pro-

cessing conditions play a critical role in the electrochemical

behavior and chemical stability of tungsten carbide [35,36].

Until now, WC powder has been prepared by many methods,

such as chemical precipitation [37], mechanical alloying

[38,39], sonochemical synthesis [40], microwave synthesis

[41e43], a temperature-programmed method [44,45], and hy-

drothermalmethods [46]. However, some of these preparation

methods have deficiencies. For example, it is easy to introduce

the impurities by chemical precipitation method. Mechanical

alloying requires high temperatures and consumes large

amounts of energy, the distribution of the WC particle size is

not uniform by the method. Due to a consequence of the

temperature dependence of dielectric and heating frequency,

the preparation process of WC is not easy to control by Mi-

crowave heating method. Polymeric carbon will be formed on

the surface of WC in the temperature-programmed method,

which can affect the surface activity. So far, preparation of

tungsten carbide catalyst by the low-temperature combustion

synthesis has not been reported. Low-temperature combus-

tion synthesis (LCS) is based on the exothermic redox reaction

between the oxidizer and the appropriate fuel, which could

induce spontaneous redox reaction at much lower tempera-

tures than the actual phase formation temperature. The

products fabricated by the LCS method could have smaller

grains with homogeneous size, since all reactants are mixed

in solution at the molecular level, leading to a faster reaction

rate. Currently, the LCS method is widely used to synthesize

ultrafine ceramic powders of complex oxide compositions and

luminescent materials [47e49]. Its widespread use in this area

is due to the fairly simple equipment needs, short reaction

times and high energy efficiency. In this study, tungsten car-

bide powder was synthesized by means of the LCS method,

and processing conditions, such as temperature and carbon

source contents, were varied in order to determine the

optimal conditions for synthesizing WC with superior elec-

trocatalytic properties. Finally, the catalytic activity of syn-

thesized WC for the ORR was investigated through the use of

polarization curves and electrochemical impedance spec-

troscopy (EIS).

Experimental details

WC catalyst preparation

Ammonium tungstate ((NH4)10W12O41, analytical grade), urea

(CO(NH2)2, analytical grade), nitric acid (HNO3, 65 wt.%), and

glucose (C6H12O6$H2O, analytical grade) were used as raw

materials in the synthesis. In this system, ammonium tung-

state was used as the tungsten source, nitric acid as an

oxidant, urea as fuel, and glucose as the carbon source. The

low-temperature combustion synthesis process is a redox

reaction, thus combustion products are generally oxides, CO2,

N2, and H2O. The combustion reaction can be described as

follows:

3ðNH4Þ10W12O41 þ 30HNO3 þ 10COðNH2Þ2/36WO3 þ 10CO2ðgÞþ 40N2ðgÞ þ 95H2O

(1)

(NH4)10W12O41, CO(NH2)2, and HNO3 were weighed ac-

cording to the 3:30:10 molar ratio in Eq. (1). The amount of

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Fig. 1 e X-ray diffraction patterns of the precursors

synthesized with different C/W molar ratios: (P1)17/3; (P2)

18/3; (P3)19/3; (P4)20/3; (P5)21/3; (P6)22/3.

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 en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 0 10913

C6H12O6$H2O was added according to the C/W molar ratio. In

the experiments, C6H12O6$H2O was added in the C/W molar

ratios of 17/3(P1), 18/3(P2), 19/3(P3), 20/3(P4), 21/3(P5) and 22/

3(P6). The preparation procedure consisted of dissolving

(NH4)10W12O41, CO(NH2)2 and C6H12O6$H2O in deionized water

under stirring. Then, HNO3 was added to the solution and

concentrated in a beaker. The resulting solution was placed in

a closed electric furnace and evaporated to form a viscous gel.

When the viscous gel swelled, it auto-ignited and initiated a

highly exothermic self-contained combustion process, con-

verting raw materials into a loose black mixture. The com-

bustion products were then used as WC precursors. The

precursors already contained some carbon, and were further

carburized to convert to the desired product. The precursor

was ground into powder in a mortar, and transferred into a

tube furnace, where it was carburized under argon flow at

various temperatures, between 800 �C and 1100 �C for 6 h to

study the development of the crystalline phases.

Sample characterization

X-ray diffraction (XRD) patterns of the samples were recorded

with a Rigaku (D/MAX-RB) diffractometer using Cu Ka radia-

tion (l ¼ 1.5406 �A) at a scanning rate of 10� min�1. Themorphology of the powder samples was studied using a field-

emission scanning electron microscope (SEM, Zeiss Ultra 55).

Energy dispersive X-ray spectra (EDS) attached to the SEM and

X-ray photoelectron spectroscopy (XPS, ESCALAB 250 Xi) were

also used to investigate the near surface chemical composi-

tion of the samples.

Preparation of gas diffusion electrodes and electrochemicalmeasurements

Gas diffusion electrodes consisted of three layers: a catalyst

layer, a current collecting layer, and a gas diffusion layer,

prepared based on earlier results from the literature [50e52].

The catalyst layer was made of synthesized tungsten carbide

catalyst, acetylene black and polytetrafluoroethylene. At first,

The WC catalyst, PTFE (60 wt.% suspension) and acetylene

black (WC:PTFE:acetylene black ¼ 3:2:5 in weight ratio) weremixed in deionized water, then added appropriate anhydrous

ethanol under stirring for 5 min. The mixture was placed in a

thermostatic water bath at 80 �C and continually stirred untilthe mixed materials became a ropy and tough paste. At last,

the pastewas pressed into an approximately 0.3mm thick and

1.5 cm diameter disc by using a pellet press. The gas diffusion

layer was prepared as the preparation of the mentioned

catalyst layer. Here in, the nominal molar ratios of acetylene

black: anhydrous sodium sulfate: PTFE were to 1:1:1 (weight

ratio), and then these materials were mixed in absolute ethyl

alcohol by stirring at room temperature for 20 min, and then

themixturewas stirred in the thermostaticwater bath at 80 �Cuntil completelymixed. Themoisture barrierwasmade by the

rolling method. Nickel foam was selected as a current collec-

tor. The catalyst layer and gas diffusion layer were pressed

together with a nickel foam current collector in-between,

under 16 MPa pressure for 1 min. Then, the gas diffusion

electrodes were finished by sintering at 200 �C in N2 flow for30 min.

Electrochemical measurements were performed using a

Parstat 2273 electrochemical workstation at 30 �C. The three-electrode system was utilized for electrochemical analysis

with the nickel foil as the counter-electrode. Hg/HgOwas used

as the reference electrode, which was a 6 M KOH solution

under alkaline conditions. A saturated calomel electrode (SCE)

was used as the reference electrode, which was a 1 M NaCl

solution under neutral pH conditions. The gas diffusion elec-

trode (catalyst layer in contact with electrolyte and its back

exposed to air) was used as aworking electrodewith 0.785 cm2

geometrical exposed area. The steady state polarization curve

measurements were performedwith a scan rate of 0.5 mV s�1.

The electrochemical impedance spectra were recorded in the

10 kHz to 0.01 Hz range. An ac signal amplitude of 5 mV was

used, and Nyquist plots were used to interpret the electro-

chemical performance of the gas diffusion electrode.

Results and discussion

Influence of the C/W molar ratio on the formation of WCcatalyst

The molar ratio of C/W has an obvious effect on the crystal-

linity of the precursors as seen in Fig. 1, showing XRD patterns

of the as-prepared precursors with different C/Wmolar ratios.

The XRD trace shows the presence of crystalline phases for

the tungsten trioxide and a residue of amorphous elemental

carbon, namely carbon black. With an increase of the C/W

molar ratio, the intensity of diffraction peak for the tungsten

trioxide phase decreases, suggesting reduced crystallization.

The morphology and microstructure of the precursors were

investigated in the field-emission scanning electron micro-

scope (SEM), and the results are shown in Fig. 2, revealing

highly porous agglomerates with many faceted grains. The

specimen exhibits a small amount of porosity, which results

from gases being released during the combustion process [53].

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Fig. 2 e SEM micrographs of: (a) precursor P2; (b) precursor P3; (c) precursor P4; (d) precursor P5.

Fig. 3 e SEM micrographs and EDS spectra of the precursor

P3. Elemental spectra corresponds to the Spectrum 1 point

on the SEM micrograph.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 010914

The particles size and shape do not differ significantly from

one sample to another.

Fig. 3 presents the EDS spectrum of the P3 precursor. The

elemental spectrum in Fig. 3 shows that the chemical com-

ponents of the sample are carbon, oxygen and tungsten,

which is consistent with the XRD results [54]. The proportion

of various elements has been obtained by the EDS spectrum,

as shown in Table 1. This shows that precursors prepared by

the LCS method mainly consist of tungsten trioxide and car-

bon, while the tungsten trioxide is dispersed in carbon. The

carbon content of the P3 precursor with a 19/3 C/Wmolar ratio

is 14.6 wt.%, as shown in Table 2. Meanwhile, in order to

further prove the composition of P3 precursor analyzed from

the EDS results, Fig. 4 shows the XPS spectrum for the surface

Table 1e The element content of the P3 precursor via EDSanalysis.

Element Weight% Atomic%

C 28.16 53.54

O 28.81 41.12

W 43.02 5.34

Total 100.00 100.00

Table 2 e Phases and C wt% of carburized product withdifferent C/W ratio in raw materials.

C/W ratio inraw materials

C wt% inthe precursor

Phases of thefinal product

C wt%in the finalproduct

18/3 12.9 W, W2C, WC 5.9

19/3 14.6 WC 7.1

20/3 15.6 WC 7.8

21/3 17.6 W2C, WC 8.6

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Fig. 4 e XPS spectra of the precursor P3: (a) W 4f peak; (b) O

1s peak; (c) C 1s peak.

Fig. 5 e XRD patterns of different powders prepared from

the precursors (a) P1; (b) P2; (c) P3; (d) P4; (e) P5; (f) P6.

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 en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 0 10915

composition of the P3 precursor. Fig. 4a shows the photo-

emission spectrum ofW 4f at 35.7 eV, 37.9 eV, and 41.7 eV, and

Fig. 4b shows the photoemission spectrum of O1s at 530.6 eV,

which both indicate the tested composition corresponds to

WO3. The C 1s peak is dominated by that of carbon at 284.8 eV,

as shown in Fig. 4c. The XPS results are consistent with the

EDSmeasurement, indicating that the P3 precursor consists of

WO3 and carbon.

To convert WO3 to WC, precursors synthesized by the LCS

method with various C/W molar ratios are reduced and

carburized in argon for 6 h at 1100 �C. The X-ray diffractionpatterns for these carburized powders are shown in Fig. 5.

There are relatively stronger W peaks and weaker W2C peaks

in Fig. 5a, suggesting that the carbonization reaction had

occurred. Only the WC peaks can be observed when C/W

molar ratios are 19/3 and 20/3, which suggest that the WO3 of

the precursors completely transformed into tungsten carbide

during the carbonization process. This can be observed in

Fig. 5c and d. The desiredWC emerges as themajor phasewith

further increase of the C/W molar ratio, while the secondary

W2C phase also appears in Fig. 5e and f. Thus, both excessive

and insufficient molar ratios of C/W in the raw materials

affect the final WC phase generation. For pure WC, the

appropriate C/W molar ratio is determined to be 19/3.

The carbon content of tungsten carbide improves with an

increase of the C/W molar ratio, as shown in Table 2. When

the C/W molar ratio is 18/3, the precursor after carbonization

consists of W, W2C, and WC phases. The theoretical carbon

content value is C/WC ¼ 6.1 wt.%, while for the 18/3 sample itis 5.9 wt.%, lower than the theoretical value. Complete

carbonization may only occur with carbon content higher

than 6.1 wt.%, so that a singleWC phase is obtained. However,

when the carbon content is higher, i.e. the C/W molar ratio is

21/3, the W2C phase will be generated when the carbon con-

tent is 8.6 wt.%.

Fig. 6 displays SEMmicrographs of reduced and carburized

powders prepared from different C/W molar ratio precursors.

The powders consist of homogeneous nano-sized particles

with 200 nm average particle size. In Fig. 6d, one can clearly

see residual carbon in the powders, due to the high C/Wmolar

ratio in the P5 precursor.

Carbon content effect on the complete carbonizationtemperature

In order to study the effect of carbon content on the complete

carbonization temperature, the P3 precursorwith the 19/3C/W

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Fig. 7 e XRD patterns of the products of the precursor (a) P3

and (b) P4 carburized at 800 �C; 900 �C; 1000 �C; 1100 �C for6 h.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 010916

molar ratio and the P4 precursor with the 20/3 C/Wmolar ratio

are reduced and carburized at various temperatures for 6 h in

argon. X-ray diffraction patterns from the 19/3 molar ratio

carburized powders are shown in Fig. 7a. The precursor pre-

pared by low-temperature combustion synthesis contains the

W2C phase at 800 �C, which may be due to the low carbon-ization temperature and incomplete carbonization. When the

temperature is higher than 900 �C, the product after the car-bon reduction reaction is based on the single-phase tungsten

carbide particles. Analyzing the XRD spectrum of the 20/3 C/W

molar ratio carbon tungsten powder, there are relatively

strong W peaks at 800 �C, suggesting that WO2 has beenreduced to elementalW at the beginning of carbonization. It is

clear that the intensity of the WC reflections begins to in-

crease at 900 �C, as shown in Fig. 7b. The W2C phase isobserved below 1000 �C, and pure WC powder is obtainedwhen the carbonization temperature reaches 1100 �C. Thus,the optimal carbonization temperature is 1100 �C.

The SEM micrographs of powders carburized at 800 �C and900 �Care shown in Fig. 8. As to the carbide powderwith the 19/3 C/Wmolar ratio at 800 �C, it appears that part of the particlesare coated with free carbon [55,56], based on the SEM micro-

graphs. This suggests that some portion of the carbon is not

combined with tungsten. When the temperature increases to

900 �C, the phase after carbonization appears to be a singleWCphase. Comparing resulting morphology from the SEM mi-

crographs of the 20/3 C/W molar ratio powder after carbon-

ization, carbonization reaction at 1100 �C proceeds morecompletely than at 800 �C, indicating that a higher carboniza-tion temperature promotes precursor carbonization.

The carbon content has a major influence on the purity of

tungsten carbide after carbonization. Since the reaction tem-

perature of pure tungsten carbide with the 20/3 C/W molar

ratio after carbonization is 1100 �C,which is higher than 900 �Cfor the 19/3 C/W molar ratio, it can be concluded that the re-

action temperature of pure tungsten carbide increases with

Fig. 6 e SEM micrographs of powders prepared from the precursors (a) P2; (b) P3; (c) P4; (d) P5.

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Fig. 8 e SEMmicrographs of the products of the precursor P3 carburized at (a) 800 �C; (b) 900 �C and P4 carburized at (c) 800 �C;(d) 1100 �C for 6 h.

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 en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 0 10917

the carbon content. The above studies have shown that car-

bon content has a functional effect on the complete carbon-

ization temperature. With increased carbon content in the

precursor, the complete carbonization temperature rises

significantly. Therefore, when the carbide products of the 21/3

C/W molar ratio are sintered at 1100 �C, the W2C phase ap-pears, since the sintering temperature is too low.

Electrocatalytic activity of WC

In order to investigate the electrocatalytic activity of the syn-

thesized tungsten carbide samples, the 20/3-WC and 19/3-WC

samples are used as catalysts to prepare gas diffusion elec-

trodes (S1 and S2) and the samples are characterized using

polarization curves and electrochemical impedance spectros-

copy. Low-cost activated carbon is commonly used as the

cathode material for small load metaleair batteries. Thus, an

activated carbon gas diffusion electrode (C) is also character-

ized to compare its electrocatalytic activity with the synthe-

sized tungsten carbide. Fig. 9 shows the steady-state linear

polarization curvesofdifferentgasdiffusionelectrodesat 30 �Cwith a scan rate of 0.5 mV s�1. Based on Fig. 9a, the low po-tential area of each electrode polarization curve is almost the

same (1 M NaCl solution). In the high potential area, the elec-

trode polarization current density with tungsten carbide

catalyst is higher than activated carbon gas diffusion electrode

(C). When the electrode potential is high than 0.35 V, the po-

larization current densities achieved by the S2 gas diffusion

electrode is higher than those of the S1 gas diffusion electrode

in high over potential region. It could also be due to different

porosity, gas diffusivity, mass transport limitations, residual

carbon.

As seen in Fig. 9b, all gas diffusion electrodes exhibit no

difference in polarization between �0.2 V and 0 V (vs. Hg/

HgO). However, electrodes containing the WC catalyst exhibit

smaller polarization than those without the WC catalyst

below �0.2 V, with the S2 gas diffusion electrode polarizationhaving the lowest value. The current density of the S2 gas

diffusion electrode at �0.4 V is approximately 350 mA cm�2.Comparing this value with the literature result for the Pt

catalyst [57]: at �0.4 V (vs. Hg/HgO), the measured currentdensity of the gas diffusion electrode for the Pt catalyst,

loaded by the method of evaporation to dryness and adsorp-

tion, is similar to the literature value of about 400 mA cm�2.These results imply that the 19/3-WC catalyst is an active

component of the ORR in alkaline solutions and is a viable

alternative to expensive Pt for the electrocatalytic cathode in

metaleair batteries.

In order to gain additional information on ORR, the EIS

characteristics of the S1 and S2 electrodes, where the catalysts

are 20/3-WC and 19/3-WC, respectively, are investigated.

Again, an activated carbon gas diffusion electrode (C) is also

characterized to compare its electrocatalytic activity with the

synthesized tungsten carbide. Fig. 10 exhibits the corre-

sponding Nyquist plots at 0.05 V (vs. SCE) with the test fre-

quency between 10 kHz and 0.01 Hz. As shown in Fig. 10, both

impedances comprise a semicircle in the frequency range,

which represent the ORR charge transfer resistance (Rct) cor-

responding to electron and ion transfer processes occurring at

interfaces [58]. Thus, the charge transfer step is the ORR rate-

determining step. Considering that the semicircle in EIS

curves represents the ORR charge transfer resistance (Rct), the

corresponding Rct in the case of S2 electrode is the smallest

among the three electrodes at the tested potential. Therefore,

it is reasonable to believe that the ORR on S2 electrode is more

favorable compared with that on the other two electrodes and

the 19/3-WC catalyst shows better catalytic properties

resulting from the enhanced reaction kinetics.

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Fig. 9 e Linear polarization curves of gas diffusion

electrodes with different catalysts (C) activated carbon; (S1)

20/3-WC; (S2)19/3-WC; (a) neutral solution; (b) alkaline

solution.

Fig. 10 e Impedance spectra of gas diffusion electrodes

with different catalysts (C) activated carbon; (S1) 20/3-WC;

(S2) 19/3-WC; (a) neutral solution; (b) alkaline solution.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 010918

Summary

Tungsten carbide catalysts for gas diffusion electrodes have

been successfully synthesized by the low-temperature com-

bustion synthesis for the first time. In this work ammonium

tungstate is used as a tungsten source, and glucose as a carbon

source. The combustion products are in the form of loose

agglomerates, which are easily ground to fine powders, and

mainly consist of tungsten trioxide and carbon. Concentra-

tions both above and below the C/W molar ratios needed for

stoichiometry of the raw materials are tested, and affect the

finalWC phases’ generation. Carbon content has a clear effect

on the complete carbonization temperature. When the C/W

molar ratio is equal to 19/3, efficientWC catalyst is obtained by

carburizing at 1100 �C for 6 h. The particle size of the finalsynthesized tungsten carbide is about 200 nm. The synthe-

sized 19/3-WC catalyst exhibits small polarization values, and

the charge transfer is the ORR rate-determining step. The 19/

3-WC catalyst exhibits sufficient electrochemical values,

making it a realistic alternative to the more expensive Pt as

the cathode catalyst in metaleair batteries.

Acknowledgments

Ping Li and Fuqiang Zhai thanks China Scholarship Council for

providing the scholarship.

http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

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 en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 0 10919

r e f e r e n c e s

[1] Pozio A, De Francesco M, Cemmi A, Cardellini F, Giorgi L.Comparison of high surface Pt/C catalysts by cyclicvoltammetry. J Power Sources 2002;105:13e9.

[2] Wang T, Kaempgen M, Nopphawan P, Wee G,Mhaisalkar S, Srinivasan M. Silver nanoparticle-decorated carbon nanotubes as bifunctional gas-diffusionelectrodes for zinc-air batteries. J Power Sources2010;195:4350e5.

[3] Lu YC, Xu Z, Gasteiger HA, Chen S, Hamad-Schifferli K, Shao-Horn Y. Platinum-gold nanoparticles: a highly activebifunctional electrocatalyst for rechargeable lithium-airbatteries. J Am Chem Soc 2010;132:12170e1.

[4] Antolini E, Zignani SC, Santos SF, Gonzalez ER. Palladium-based electrodes: a way to reduce platinum content inpolymer electrolyte membrane fuel cells. Electrochim Acta2011;56:2299e305.

[5] Hsieh CT, Chen WY, Chen IL, Roy AK. Deposition and activitystability of Pt-Co catalysts on carbon nanotube-basedelectrodes prepared by microwave-assisted synthesis. JPower Sources 2012;199:94e102.

[6] Lu G, Cooper JS, McGinn PJ. SECM characterization of Pt-Ru-WC and Pt-Ru-Co ternary thin film combinatorial libraries asanode electrocatalysts for PEMFC. J Power Sources2006;161:106e14.

[7] Liang Y, Li Y, Wang H, Zhou J, Wang J, Regier T. Co3O4nanocrystals on graphene as a synergistic catalyst foroxygen reduction reaction. Nat Mater 2011;10:780e6.

[8] Sun W, Hsu A, Chen R. Carbon-supported tetragonal MnOOHcatalysts for oxygen reduction reaction in alkaline media. JPower Sources 2011;196:627e35.

[9] Yang CC, Hsu ST, Chien WC, Shih MC, Chiu SJ, Lee KT, et al.Electrochemical properties of air electrodes based on MnO2catalysts supported on binary carbons. Int J Hydrogen Energy2006;31:2076e87.

[10] Nguyen-Cong H, De La Garza Guadarrama V, Gautier JL,Chartier P. Oxygen reduction on NixCo3�xO4 spinel particles/polypyrrole composite electrodes: hydrogen peroxideformation. Electrochim Acta 2003;48:2389e95.

[11] Bursell M, Pirjamali M, Kiros Y. La0.6Ca0.4CoO3,La0.1Ca0.9MnO3 and LaNiO3 as bifunctional oxygenelectrodes. Electrochim Acta 2002;47:1651e60.

[12] Yang X, Li S, Liu Y, Wei X, Liu Y. LaNi0.8Co0.2O3 as a cathodecatalyst for a direct borohydride fuel cell. J Power Sources2011;196:4992e5.

[13] Restovic A, Rios E, Barbato S, Ortiz J, Gautier L. Oxygenreduction in alkaline medium at thin MnxCo3�xO4 (0 � x � 1)spinel films prepared by spray pyrolysis. Effect of oxidecation composition on the reaction kinetics. J ElectroanalChem 2002;522:141e51.

[14] Li Zhu A, Wang H, Qu W, Li X, Jong Z, Li H. Low temperaturepyrolyzed cobalt tetramethoxy phenylporphyrin catalyst andits applications as an improved catalyst for metal airbatteries. J Power Sources 2010;195:5587e95.

[15] Tamizhmani G, Dodelet JP, Guay D, Lalande G.Electrocatalytic activity of Nafion-impregnated pyrolyzedcobalt phthalocyanine a correlative study between rotatingdisk and solid polymer electrolyte fuel-cell electrodes. JElectrochem Soc 1994;141:41e5.

[16] Yuan Y, Ahmed J, Kim S. Polyaniline/carbon blackcomposite-supported iron phthalocyanine as an oxygenreduction catalyst for microbial fuel cells. J Power Sources2011;196:1103e6.

[17] Acres GJK, Frost JC, Hards GA, Poter RJ, Ralph TR,Thompsett D, et al. Electrocatalysts for fuel cells. Catal Today1997;38:393e400.

[18] Levy RB, Boudart M. Platinum-like behavior of tungstencarbide in surface catalysis. Science 1973;181:547e9.

[19] Mellinger ZJ, Kelly TG, Chen JG. Pd-modified tungstencarbide for methanol electro-oxidation: from surface sciencestudies to electrochemical evaluation. ACS Catal2012;2:751e8.

[20] Zellner MB, Chen JG. Supporting monolayer Pt onW (110) andC/W (110): modification effects on the reaction pathways ofcyclohexene. J Catal 2005;235:393e402.

[21] Chhina H, Campbell S, Kesler O. Thermal andelectrochemical stability of tungsten carbide catalystsupports. J Power Sources 2007;164:431e40.

[22] Torabi A, Etsell TH, Semagina N, Sarkar P. Electrochemicalbehaviour of tungsten carbide-based materials as candidateanodes for solid oxide fuel cells. Electrochim Acta2012;67:172e80.

[23] LiuY, Shrestha S,MustainWE. Synthesis of nanosize tungstenoxide and its evaluation as an electrocatalyst support foroxygen reduction in acid media. ACS Catal 2012;2:456e63.

[24] Weigert EC, Esposito DV, Chen JG. Cyclic voltammetry and X-ray photoelectron spectroscopy studies of electrochemicalstability of clean and Pt-modified tungsten and molybdenumcarbide (WC and Mo2C) electrocatalysts. J Power Sources2009;193:501e6.

[25] Lee K, Ishihara A, Mitsushima S, Kamiya N, Ota K. Stabilityand electrocatalytic activity for oxygen reduction in WC þ Tacatalyst. Electrochim Acta 2004;49:3479e85.

[26] Weidman MC, Esposito DV, Hsu YC, Chen JG. Comparison ofelectrochemical stability of transition metal carbides (WC,W2C, Mo2C) over a wide pH range. J Power Sources2012;202:11e7.

[27] Weigert EC, Stottlemyer AL, Zellner MB, Chen JG. Tungstenmonocarbide as potential replacement of platinum formethanol electrooxidation. J PhysChemC2007;111:14617e20.

[28] Liu Y, Mustain WE. Evaluation of tungsten carbide as theelectrocatalyst support for platinum hydrogen evolution/oxidation catalysts. Int J Hydrogen Energy 2012;37:8929e38.

[29] Elezovi�c NR, Babi�c BM, Gaji�c-Krstaji�c L, Ercius P,Radmilovi�c VR, Krstaji�c NV, et al. Pt supported on nano-tungsten carbide as a beneficial catalyst for the oxygenreduction reaction in alkaline solution. Electrochim Acta2012;69:239e46.

[30] Cui G, Shen PK, Meng H, Zhao J, Wu G. Tungsten carbideas supports for Pt electrocatalysts with improved COtolerance in methanol oxidation. J Power Sources2011;196:6125e30.

[31] Ko AR, Oh JK, Lee YW, Han SB, Park KW. Characterizations oftungsten carbide as a non-Pt counter electrode in dye-sensitized solar cells. Mater Lett 2011;65:2220e3.

[32] Wu M, Lin X, Hagfeldt A, Ma T. Low-cost molybdenumcarbide and tungsten carbide counter electrodes for dye-sensitized solar cells. Angew Chem Int Ed Engl2011;50:3520e4.

[33] Hu S, Shi B, Yao G, Li G, Ma C. Preparation andelectrocatalytic activity of tungsten carbide and titaniananocomposite. Mater Res Bull 2011;46:1738e45.

[34] Torabi A, Etsell TH. Tungsten carbide-based anodes for solidoxide fuel cells: preparation, performance and challenges. JPower Sources 2012;212:47e56.

[35] Fleischmann R, Böhm H. Water oxidation on versatiletungsten carbide materials. Electrochim Acta1977;22:1123e8.

[36] Nikolov I, Vitanov T, Nikolova V. The effect of the method ofpreparation on the catalytic activity of tungsten carbide forhydrogen evolution. J Power Sources 1980;5:197e206.

[37] Nie M, Tang HL, Wei ZD, Jiang SP, Shen PK. Highly efficientAuPd-WC/C electrocatalyst for ethanol oxidation.Electrochem Commun 2007;9:2375e9.

http://refhub.elsevier.com/S0360-3199(14)01231-2/sref1http://refhub.elsevier.com/S0360-3199(14)01231-2/sref1http://refhub.elsevier.com/S0360-3199(14)01231-2/sref1http://refhub.elsevier.com/S0360-3199(14)01231-2/sref1http://refhub.elsevier.com/S0360-3199(14)01231-2/sref2http://refhub.elsevier.com/S0360-3199(14)01231-2/sref2http://refhub.elsevier.com/S0360-3199(14)01231-2/sref2http://refhub.elsevier.com/S0360-3199(14)01231-2/sref2http://refhub.elsevier.com/S0360-3199(14)01231-2/sref2http://refhub.elsevier.com/S0360-3199(14)01231-2/sref2http://refhub.elsevier.com/S0360-3199(14)01231-2/sref3http://refhub.elsevier.com/S0360-3199(14)01231-2/sref3http://refhub.elsevier.com/S0360-3199(14)01231-2/sref3http://refhub.elsevier.com/S0360-3199(14)01231-2/sref3http://refhub.elsevier.com/S0360-3199(14)01231-2/sref3http://refhub.elsevier.com/S0360-3199(14)01231-2/sref4http://refhub.elsevier.com/S0360-3199(14)01231-2/sref4http://refhub.elsevier.com/S0360-3199(14)01231-2/sref4http://refhub.elsevier.com/S0360-3199(14)01231-2/sref4http://refhub.elsevier.com/S0360-3199(14)01231-2/sref4http://refhub.elsevier.com/S0360-3199(14)01231-2/sref5http://refhub.elsevier.com/S0360-3199(14)01231-2/sref5http://refhub.elsevier.com/S0360-3199(14)01231-2/sref5http://refhub.elsevier.com/S0360-3199(14)01231-2/sref5http://refhub.elsevier.com/S0360-3199(14)01231-2/sref5http://refhub.elsevier.com/S0360-3199(14)01231-2/sref6http://refhub.elsevier.com/S0360-3199(14)01231-2/sref6http://refhub.elsevier.com/S0360-3199(14)01231-2/sref6http://refhub.elsevier.com/S0360-3199(14)01231-2/sref6http://refhub.elsevier.com/S0360-3199(14)01231-2/sref6http://refhub.elsevier.com/S0360-3199(14)01231-2/sref7http://refhub.elsevier.com/S0360-3199(14)01231-2/sref7http://refhub.elsevier.com/S0360-3199(14)01231-2/sref7http://refhub.elsevier.com/S0360-3199(14)01231-2/sref7http://refhub.elsevier.com/S0360-3199(14)01231-2/sref7http://refhub.elsevier.com/S0360-3199(14)01231-2/sref8http://refhub.elsevier.com/S0360-3199(14)01231-2/sref8http://refhub.elsevier.com/S0360-3199(14)01231-2/sref8http://refhub.elsevier.com/S0360-3199(14)01231-2/sref8http://refhub.elsevier.com/S0360-3199(14)01231-2/sref9http://refhub.elsevier.com/S0360-3199(14)01231-2/sref9http://refhub.elsevier.com/S0360-3199(14)01231-2/sref9http://refhub.elsevier.com/S0360-3199(14)01231-2/sref9http://refhub.elsevier.com/S0360-3199(14)01231-2/sref9http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref10http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref11http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref12http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref13http://refhub.elsevier.com/S0360-3199(14)01231-2/sref14http://refhub.elsevier.com/S0360-3199(14)01231-2/sref14http://refhub.elsevier.com/S0360-3199(14)01231-2/sref14http://refhub.elsevier.com/S0360-3199(14)01231-2/sref14http://refhub.elsevier.com/S0360-3199(14)01231-2/sref14http://refhub.elsevier.com/S0360-3199(14)01231-2/sref15http://refhub.elsevier.com/S0360-3199(14)01231-2/sref15http://refhub.elsevier.com/S0360-3199(14)01231-2/sref15http://refhub.elsevier.com/S0360-3199(14)01231-2/sref15http://refhub.elsevier.com/S0360-3199(14)01231-2/sref15http://refhub.elsevier.com/S0360-3199(14)01231-2/sref15http://refhub.elsevier.com/S0360-3199(14)01231-2/sref16http://refhub.elsevier.com/S0360-3199(14)01231-2/sref16http://refhub.elsevier.com/S0360-3199(14)01231-2/sref16http://refhub.elsevier.com/S0360-3199(14)01231-2/sref16http://refhub.elsevier.com/S0360-3199(14)01231-2/sref16http://refhub.elsevier.com/S0360-3199(14)01231-2/sref17http://refhub.elsevier.com/S0360-3199(14)01231-2/sref17http://refhub.elsevier.com/S0360-3199(14)01231-2/sref17http://refhub.elsevier.com/S0360-3199(14)01231-2/sref17http://refhub.elsevier.com/S0360-3199(14)01231-2/sref18http://refhub.elsevier.com/S0360-3199(14)01231-2/sref18http://refhub.elsevier.com/S0360-3199(14)01231-2/sref18http://refhub.elsevier.com/S0360-3199(14)01231-2/sref19http://refhub.elsevier.com/S0360-3199(14)01231-2/sref19http://refhub.elsevier.com/S0360-3199(14)01231-2/sref19http://refhub.elsevier.com/S0360-3199(14)01231-2/sref19http://refhub.elsevier.com/S0360-3199(14)01231-2/sref19http://refhub.elsevier.com/S0360-3199(14)01231-2/sref20http://refhub.elsevier.com/S0360-3199(14)01231-2/sref20http://refhub.elsevier.com/S0360-3199(14)01231-2/sref20http://refhub.elsevier.com/S0360-3199(14)01231-2/sref20http://refhub.elsevier.com/S0360-3199(14)01231-2/sref21http://refhub.elsevier.com/S0360-3199(14)01231-2/sref21http://refhub.elsevier.com/S0360-3199(14)01231-2/sref21http://refhub.elsevier.com/S0360-3199(14)01231-2/sref21http://refhub.elsevier.com/S0360-3199(14)01231-2/sref22http://refhub.elsevier.com/S0360-3199(14)01231-2/sref22http://refhub.elsevier.com/S0360-3199(14)01231-2/sref22http://refhub.elsevier.com/S0360-3199(14)01231-2/sref22http://refhub.elsevier.com/S0360-3199(14)01231-2/sref22http://refhub.elsevier.com/S0360-3199(14)01231-2/sref23http://refhub.elsevier.com/S0360-3199(14)01231-2/sref23http://refhub.elsevier.com/S0360-3199(14)01231-2/sref23http://refhub.elsevier.com/S0360-3199(14)01231-2/sref23http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref24http://refhub.elsevier.com/S0360-3199(14)01231-2/sref25http://refhub.elsevier.com/S0360-3199(14)01231-2/sref25http://refhub.elsevier.com/S0360-3199(14)01231-2/sref25http://refhub.elsevier.com/S0360-3199(14)01231-2/sref25http://refhub.elsevier.com/S0360-3199(14)01231-2/sref25http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref26http://refhub.elsevier.com/S0360-3199(14)01231-2/sref27http://refhub.elsevier.com/S0360-3199(14)01231-2/sref27http://refhub.elsevier.com/S0360-3199(14)01231-2/sref27http://refhub.elsevier.com/S0360-3199(14)01231-2/sref27http://refhub.elsevier.com/S0360-3199(14)01231-2/sref28http://refhub.elsevier.com/S0360-3199(14)01231-2/sref28http://refhub.elsevier.com/S0360-3199(14)01231-2/sref28http://refhub.elsevier.com/S0360-3199(14)01231-2/sref28http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref29http://refhub.elsevier.com/S0360-3199(14)01231-2/sref30http://refhub.elsevier.com/S0360-3199(14)01231-2/sref30http://refhub.elsevier.com/S0360-3199(14)01231-2/sref30http://refhub.elsevier.com/S0360-3199(14)01231-2/sref30http://refhub.elsevier.com/S0360-3199(14)01231-2/sref30http://refhub.elsevier.com/S0360-3199(14)01231-2/sref31http://refhub.elsevier.com/S0360-3199(14)01231-2/sref31http://refhub.elsevier.com/S0360-3199(14)01231-2/sref31http://refhub.elsevier.com/S0360-3199(14)01231-2/sref31http://refhub.elsevier.com/S0360-3199(14)01231-2/sref32http://refhub.elsevier.com/S0360-3199(14)01231-2/sref32http://refhub.elsevier.com/S0360-3199(14)01231-2/sref32http://refhub.elsevier.com/S0360-3199(14)01231-2/sref32http://refhub.elsevier.com/S0360-3199(14)01231-2/sref32http://refhub.elsevier.com/S0360-3199(14)01231-2/sref33http://refhub.elsevier.com/S0360-3199(14)01231-2/sref33http://refhub.elsevier.com/S0360-3199(14)01231-2/sref33http://refhub.elsevier.com/S0360-3199(14)01231-2/sref33http://refhub.elsevier.com/S0360-3199(14)01231-2/sref34http://refhub.elsevier.com/S0360-3199(14)01231-2/sref34http://refhub.elsevier.com/S0360-3199(14)01231-2/sref34http://refhub.elsevier.com/S0360-3199(14)01231-2/sref34http://refhub.elsevier.com/S0360-3199(14)01231-2/sref35http://refhub.elsevier.com/S0360-3199(14)01231-2/sref35http://refhub.elsevier.com/S0360-3199(14)01231-2/sref35http://refhub.elsevier.com/S0360-3199(14)01231-2/sref35http://refhub.elsevier.com/S0360-3199(14)01231-2/sref36http://refhub.elsevier.com/S0360-3199(14)01231-2/sref36http://refhub.elsevier.com/S0360-3199(14)01231-2/sref36http://refhub.elsevier.com/S0360-3199(14)01231-2/sref36http://refhub.elsevier.com/S0360-3199(14)01231-2/sref37http://refhub.elsevier.com/S0360-3199(14)01231-2/sref37http://refhub.elsevier.com/S0360-3199(14)01231-2/sref37http://refhub.elsevier.com/S0360-3199(14)01231-2/sref37http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 0 9 1 1e1 0 9 2 010920

[38] Bolokang S, Banganayi C, Phasha M. Effect of C and millingparameters on the synthesis of WC powders by mechanicalalloying. Int J Refract Met Hard Mater 2010;28:211e6.

[39] Zakeri M, Rahimipour MR. Effect of cup and ball types onalumina-tungsten carbide nanocomposite powdersynthesized by mechanical alloying. Adv Powder Technol2012;23:31e4.

[40] Kim J, Jang JH, Lee YH, Kwon YU. Enhancement ofelectrocatalytic activity of platinum for hydrogen oxidationreaction by sonochemically synthesized WC1�xnanoparticles. J Power Sources 2009;193:441e6.

[41] Lu JL, Li ZH, Jiang SP, Shen PK. Nanostructured tungstencarbide/carbon composites synthesized by a microwaveheating method as supports of platinum catalysts formethanol oxidation. J Power Sources 2012;202:56e62.

[42] Rees EJ, Essaki K, Brady CDA, Burstein GT. Hydrogenelectrocatalysts from microwave-synthesisednanoparticulate carbides. J Power Sources 2009;188:75e81.

[43] Yin S, Cai M, Wang C, Shen PK. Tungsten carbide promotedPd-Fe as alcohol-tolerant electrocatalysts for oxygenreduction reactions. Energy Environ Sci 2011;4:558e63.

[44] Brady CDA, Rees EJ, Burstein GT. Electrocatalysis bynanocrystalline tungsten carbides and the effects ofcodeposited Silver. J Power Sources 2008;179:17e26.

[45] Kelly TG, Stottlemyer AL, Ren H, Chen JG. Comparison ofOeH, CeH, and CeO bond scission sequence of methanol ontungsten carbide surfaces modified by Ni, Rh, and Au. J PhysChem C 2011;115:6644e50.

[46] d’Arbigny JB, Taillades G, Marrony M, Jones DJ, Rozière J.Hollow microspheres with a tungsten carbide kernel forPEMFC application. Chem Commun 2011;47:7950e2.

[47] Zhang Z, Wang W, Shang M, Yin W. Low-temperaturecombustion synthesis of Bi2WO6 nanoparticles as a visible-light-driven photocatalyst. J Hazard Mater 2010;177:1013e8.

[48] Ahmed IS, Shama SA, Dessouki HA, Ali AA. Low temperaturecombustion synthesis of CoxMg1�xAl2O4 nano pigmentsusing oxalyldihydrazide as a fuel. Mater Chem Phys2011;125:326e33.

[49] Bansal NP, Zhong ZM. Combustion synthesis ofSm0.5Sr0.5CoO3�x and La0.6Sr0.4CoO3�x nanopowders for solidoxide fuel cell cathodes. J Power Sources 2006;158:148e53.

[50] Drillet JF, Bueb H, Dettlaff-Weglikowska U, Dittmeyer R,Roth S. Development of a self-supported single-wall carbonnanotube-based gas diffusion electrode with spatially well-defined reaction and diffusion layers. J Power Sources2010;195:8084e8.

[51] Thangamuthu R, Lin CW. Preparation of gas diffusionelectrodes using PEG/SiO2 hybrid materials and the effect oftheir composition onmicrostructure of the catalyst layer andon fuel cell performance. J Power Sources 2006;161:160e7.

[52] Tran C, Yang XQ, Qu D. Investigation of the gas-diffusion-electrode used as lithium/air cathode in non-aqueouselectrolyte and the importance of carbon material porosity. JPower Sources 2010;195:2057e63.

[53] Shih FY, Fung KZ, Lin HC, Chen GJ. Low-temperaturesynthesis of nanocrystalline NiO-YSZ powders by succinicacid-assisted combustion. J Power Sources 2006;160:148e54.

[54] Obradovi�c MD, Gojkovi�c SL, Elezovi�c NR, Ercius P,Radmilovi�c VR, Vra�car Lj D, et al. The kinetics of thehydrogen oxidation reaction on WC/Pt catalyst with lowcontent of Pt nano-particles. J Electroanal Chem2012;671:24e32.

[55] Lemaitre J, Vidick B, Delmon B. Control of the catalyticactivity of tungsten carbides: I. Preparation of highlydispersed tungsten carbides. J Catal 1986;99:415e27.

[56] Yang X, Kimmel YC, Fu J, Koel BE, Chen JG. Activation oftungsten carbide catalysts by use of an oxygen plasmapretreatment. ACS Catal 2012;2:765e9.

[57] Huang H, Zhang W, Li M, Gan Y, Chen J, Kuang Y. Carbonnanotubes as a secondary support of a catalyst layer in a gasdiffusion electrode for metal air batteries. J Colloid InterfaceSci 2005;284:593e9.

[58] Sen Gupta S, Mahapatra SS, Datta J. A potential anodematerial for the direct alcohol fuel cell. J Power Sources2004;131:169e74.

http://refhub.elsevier.com/S0360-3199(14)01231-2/sref38http://refhub.elsevier.com/S0360-3199(14)01231-2/sref38http://refhub.elsevier.com/S0360-3199(14)01231-2/sref38http://refhub.elsevier.com/S0360-3199(14)01231-2/sref38http://refhub.elsevier.com/S0360-3199(14)01231-2/sref39http://refhub.elsevier.com/S0360-3199(14)01231-2/sref39http://refhub.elsevier.com/S0360-3199(14)01231-2/sref39http://refhub.elsevier.com/S0360-3199(14)01231-2/sref39http://refhub.elsevier.com/S0360-3199(14)01231-2/sref39http://refhub.elsevier.com/S0360-3199(14)01231-2/sref40http://refhub.elsevier.com/S0360-3199(14)01231-2/sref40http://refhub.elsevier.com/S0360-3199(14)01231-2/sref40http://refhub.elsevier.com/S0360-3199(14)01231-2/sref40http://refhub.elsevier.com/S0360-3199(14)01231-2/sref40http://refhub.elsevier.com/S0360-3199(14)01231-2/sref40http://refhub.elsevier.com/S0360-3199(14)01231-2/sref41http://refhub.elsevier.com/S0360-3199(14)01231-2/sref41http://refhub.elsevier.com/S0360-3199(14)01231-2/sref41http://refhub.elsevier.com/S0360-3199(14)01231-2/sref41http://refhub.elsevier.com/S0360-3199(14)01231-2/sref41http://refhub.elsevier.com/S0360-3199(14)01231-2/sref42http://refhub.elsevier.com/S0360-3199(14)01231-2/sref42http://refhub.elsevier.com/S0360-3199(14)01231-2/sref42http://refhub.elsevier.com/S0360-3199(14)01231-2/sref42http://refhub.elsevier.com/S0360-3199(14)01231-2/sref43http://refhub.elsevier.com/S0360-3199(14)01231-2/sref43http://refhub.elsevier.com/S0360-3199(14)01231-2/sref43http://refhub.elsevier.com/S0360-3199(14)01231-2/sref43http://refhub.elsevier.com/S0360-3199(14)01231-2/sref44http://refhub.elsevier.com/S0360-3199(14)01231-2/sref44http://refhub.elsevier.com/S0360-3199(14)01231-2/sref44http://refhub.elsevier.com/S0360-3199(14)01231-2/sref44http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref45http://refhub.elsevier.com/S0360-3199(14)01231-2/sref46http://refhub.elsevier.com/S0360-3199(14)01231-2/sref46http://refhub.elsevier.com/S0360-3199(14)01231-2/sref46http://refhub.elsevier.com/S0360-3199(14)01231-2/sref46http://refhub.elsevier.com/S0360-3199(14)01231-2/sref47http://refhub.elsevier.com/S0360-3199(14)01231-2/sref47http://refhub.elsevier.com/S0360-3199(14)01231-2/sref47http://refhub.elsevier.com/S0360-3199(14)01231-2/sref47http://refhub.elsevier.com/S0360-3199(14)01231-2/sref47http://refhub.elsevier.com/S0360-3199(14)01231-2/sref47http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref48http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref49http://refhub.elsevier.com/S0360-3199(14)01231-2/sref50http://refhub.elsevier.com/S0360-3199(14)01231-2/sref50http://refhub.elsevier.com/S0360-3199(14)01231-2/sref50http://refhub.elsevier.com/S0360-3199(14)01231-2/sref50http://refhub.elsevier.com/S0360-3199(14)01231-2/sref50http://refhub.elsevier.com/S0360-3199(14)01231-2/sref50http://refhub.elsevier.com/S0360-3199(14)01231-2/sref51http://refhub.elsevier.com/S0360-3199(14)01231-2/sref51http://refhub.elsevier.com/S0360-3199(14)01231-2/sref51http://refhub.elsevier.com/S0360-3199(14)01231-2/sref51http://refhub.elsevier.com/S0360-3199(14)01231-2/sref51http://refhub.elsevier.com/S0360-3199(14)01231-2/sref51http://refhub.elsevier.com/S0360-3199(14)01231-2/sref52http://refhub.elsevier.com/S0360-3199(14)01231-2/sref52http://refhub.elsevier.com/S0360-3199(14)01231-2/sref52http://refhub.elsevier.com/S0360-3199(14)01231-2/sref52http://refhub.elsevier.com/S0360-3199(14)01231-2/sref52http://refhub.elsevier.com/S0360-3199(14)01231-2/sref53http://refhub.elsevier.com/S0360-3199(14)01231-2/sref53http://refhub.elsevier.com/S0360-3199(14)01231-2/sref53http://refhub.elsevier.com/S0360-3199(14)01231-2/sref53http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref54http://refhub.elsevier.com/S0360-3199(14)01231-2/sref55http://refhub.elsevier.com/S0360-3199(14)01231-2/sref55http://refhub.elsevier.com/S0360-3199(14)01231-2/sref55http://refhub.elsevier.com/S0360-3199(14)01231-2/sref55http://refhub.elsevier.com/S0360-3199(14)01231-2/sref56http://refhub.elsevier.com/S0360-3199(14)01231-2/sref56http://refhub.elsevier.com/S0360-3199(14)01231-2/sref56http://refhub.elsevier.com/S0360-3199(14)01231-2/sref56http://refhub.elsevier.com/S0360-3199(14)01231-2/sref57http://refhub.elsevier.com/S0360-3199(14)01231-2/sref57http://refhub.elsevier.com/S0360-3199(14)01231-2/sref57http://refhub.elsevier.com/S0360-3199(14)01231-2/sref57http://refhub.elsevier.com/S0360-3199(14)01231-2/sref57http://refhub.elsevier.com/S0360-3199(14)01231-2/sref58http://refhub.elsevier.com/S0360-3199(14)01231-2/sref58http://refhub.elsevier.com/S0360-3199(14)01231-2/sref58http://refhub.elsevier.com/S0360-3199(14)01231-2/sref58http://dx.doi.org/10.1016/j.ijhydene.2014.04.173http://dx.doi.org/10.1016/j.ijhydene.2014.04.173

Tungsten carbide synthesized by low-temperature combustion as gas diffusion electrode catalystIntroductionExperimental detailsWC catalyst preparationSample characterizationPreparation of gas diffusion electrodes and electrochemical measurements

Results and discussionInfluence of the C/W molar ratio on the formation of WC catalystCarbon content effect on the complete carbonization temperatureElectrocatalytic activity of WC

SummaryAcknowledgmentsReferences