Journal of The Electrochemical

Society

Ta0.3Ti0.7O2 Electrocatalyst Supports Exhibit Exceptional

Electrochemical Stability
To cite this article: Amod Kumar and Vijay Ramani 2013 J. Electrochem. Soc. 160 F1207

View the article online for updates and enhancements.

This content was downloaded from IP address [Link] on 22/11/2021 at 07:37

 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013) F1207
0013-4651/2013/160(11)/F1207/9/$31.00 © The Electrochemical Society

Ta0.3 Ti0.7 O2 Electrocatalyst Supports Exhibit Exceptional

Electrochemical Stability
Amod Kumar∗ and Vijay Ramani∗∗,z
Center for Electrochemical Science and Engineering, Department of Chemical and Biological Engineering, Illinois
Institute of Technology, Chicago, Illinois 60616, USA

Tantalum (Ta) modified TiO2 was synthesized using a sol-gel procedure and evaluated for use as an electrocatalyst support. Platinum
nanoparticles were deposited on this support via the incipient wetness method. X-ray diffraction (XRD), transmission electron
microscopy (TEM), X-ray energy dispersive spectroscopy (XEDS), X-ray photoelectron spectroscopy (XPS) and nitrogen desorption
were used to investigate their microstructure. Of the formulations studied, Ta0.3 Ti0.7 O2 possessed an adequate powder electronic
conductivity of 0.2 S/cm. The capacitance of this Ta0.3 Ti0.7 O2 support changed by only 12% whereas the pseudocapacitance of a
carbon benchmark changed by over 100% over 10,000 support stability test cycles (support stability test protocol: 1.0-1.5 V vs. RHE
at a scan rate of 500 mV/s in a 0.1 M HClO4 electrolyte at 25◦ C) under identical test conditions. The stability of 20% Pt/Ta0.3 Ti0.7 O2 ,
20% Pt/C, and 46% Pt/C (commercial, TKK) catalysts were evaluated by monitoring the evolution of the electrochemical surface
area (ECSA) under load cycling (load cycling protocol: 0.6-0.95 V vs. RHE with a dwell time of 3 s at either potentials in a 0.1
M HClO4 electrolyte at 25◦ C). The loss in ECSA in 20% Pt/Ta0.3 Ti0.7 O2 was found to be 35% compared to 44-47% for 20% Pt/C
and 46% Pt/C over 10,000 load cycles. 20% Pt/Ta0.3 Ti0.7 O2 yielded the following performance-related metrics: the ECSA of this
electrocatalyst was 41 m2 /g, the mass activity and area-specific activities for the oxygen reduction reaction (ORR) determined at
0.9 V vs. RHE in a 0.1 M HClO4 electrolyte at 25◦ C were 62 mA/mgPt and 151 μA/cm2 Pt , respectively. Koutecky-Levich analysis
carried out to evaluate ORR kinetics suggested that all 3 catalysts (20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C) demonstrated
similar ORR mechanisms. We propose that Ta0.3 Ti0.7 O2 shows outstanding promise as a corrosion-resistant alternative to carbon as
an electrocatalyst support in polymer electrolyte fuel cells.
© 2013 The Electrochemical Society. [DOI: 10.1149/2.038311jes] All rights reserved.

Manuscript submitted July 15, 2013; revised manuscript received August 28, 2013. Published September 18, 2013. This was Paper
1629 presented at the San Francisco, California, Meeting of the Society, October 27-November 1, 2013.

Polymer electrolyte fuel cells (PEFCs) have shown tremendous metal nanoparticles become isolated from the electrode or aggregate
promise for automotive, stationary, and portable applications. There to form larger particles.10
are, however, several challenges that need to be addressed before suc- Several prior studies have investigated alternative electrocatalyst
cessful commercialization of PEFC technology is realized. Among supports. Transition metal nitrides such as TiN were shown to possess
them, hydrogen infrastructure, stack cost, and component durabil- high conductivity and corrosion-resistance. As a catalyst support in
ity are perhaps the most critical. The issue of component durability fuel cell, they were found to have high activity but suffered from low
has been studied widely and various sources/mechanisms of degra- stability during long-term fuel cell operation.15–17 Transition metal
dation in PEFC components have been identified.1 It has been es- carbides such as titanium carbide, boron carbide and silicon carbide
tablished that membranes,2 electrocatalysts,3 and bipolar plates4 all have also exhibited high conductivity and stability.18 Electrocatalysts
suffer from long-term degradation. Currently, platinum supported on using titanium carbide as a support have been shown to have good
carbon (Pt/C) is recognized as the state-of-the-art catalyst in PEFCs.5 catalytic activity,19–21 although they suffer from low surface area,
However, platinum undergoes degradation during fuel cell opera- which prevents high dispersion of metal catalyst.22–24
tion via various mechanisms. Platinum dissolution leads to isola- TiO2 has been widely used in the area of photovoltaics,25
tion of particles, which in turn results in lower performance. Plat- photocatalysis,26 photochromics,27 gas sensors28 and in many other
inum sintering and agglomeration can also occur and result in lower applications.29 It shows high mechanical and chemical stability in
activity.3,6 acidic and oxidative environments and is particularly advantageous
Carbon possesses many properties desired in an electrocatalyst for use as a catalyst support in PEFCs.30 Pure titania, which was
support, such as high electron conductivity and high surface area; shown to be a durable catalyst support,31 is a semiconductor and
the latter allows for high dispersion of catalyst metal and hence high therefore does not possess high electron conductivity - a requirement
utilization of precious metal electrocatalysts.7 Therefore, carbon is for a catalyst support material. Magneli-phase titania which can be
widely used as a catalyst support material in PEFCs. However, carbon obtained by annealing TiO2 at high temperatures of 850◦ C, has higher
suffers from electrochemical degradation under conditions of high conductivities (∼1000 S/cm) due to oxygen defects.32–34 However,
temperature, high humidity, high electrode potential, and low pH.8–10 the viability of maintaining these defect sites in an oxidizing envi-
Carbon corrosion occurs by the following reaction:11 ronment (i.e. at the oxygen electrode) is questionable. Doping rutile
titania with metals such as Zr, Hf, V, Nb, Ta, Cr, Mo, W, Ru, Os, Ir
C + 2H2 O → CO2 + 4H+ + 4e− 0.207 V vs. RHE [1] and Sn can increase its electron conductivity and a few such combina-
The lifetime of PEFC stacks is greatly reduced by carbon sup- tions have been previously evaluated for electrochemical stability.35–40
port corrosion.8,12–14 The reaction rate for the electrooxidation of car- Numerous studies have been carried out using Nbx Ti1-x O2 and have
bon (i.e. carbon corrosion) can increase significantly under certain reported high activity and stability. Park and Seol41 reported an in-
transient operating conditions (e.g., PEFC start-up/shut-down; fuel crease in electron conductivity from 10−6 S/cm to 0.1 S/cm by doping
starvation).8 Furthermore, during the start-up and shutdown of a fuel TiO2 with Nb. Huang et al.42 reported the electron conductivity of
cell, the local cathode potential can escalate to values as high as 1.5 V, niobium-doped titania (Nb0.25 Ti0.75 O2 ) to be around 1.11 S/cm, and
which significantly accelerates the carbon corrosion reaction rate (the also reported comparable catalytic activities for Pt/ Nb0.25 Ti0.75 O2 and
electrochemical rate constant increases exponentially with overpoten- Pt/C catalyst (E-TEK). Moreover, after accelerated durability tests,
tial). Carbon corrosion has been found to accelerate the degradation the activity of Pt/ Nb0.25 Ti0.75 O2 was reported to be 10 times higher
of platinum. As carbon is irreversibly oxidized to carbon dioxide, Pt than that of Pt/C due to carbon corrosion in the Pt/C catalyst. Wang
et al.43 measured the conductivities of Nbx Ti1-x O2 supports, however
the highest conductivity was 7.7 × 10−5 S/cm at room temperature.
∗
Electrochemical Society Student Member.
The low reduction temperature (400◦ C) employed during synthesis of
∗∗
Electrochemical Society Active Member. these supports likely resulted in lower electronic conductivity in this
z
E-mail: ramani@[Link] case.
F1208 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013)

Tantalum modified TiO2 has been used previously as a thick film The Brunauer–Emmet–Teller (BET) method was used to calculate
gas sensor44 and as nanowires for solar cells,45 but has hitherto not specific surface area by nitrogen sorption at -196◦ C. X-ray photo-
been studied as an electrocatalyst support. In this study, Ta modified electron spectroscopy (XPS) measurements were made on a X-ray
TiO2 was investigated as a possible corrosion-resistant catalyst support photoelectron spectrometer (model Axis 165, Kratos Analytical, Shi-
for PEFC applications. The main objectives of this study were: a) to madzu Co.) with a monochromatic X-ray source of MgKα.
investigate whether Ta modified TiO2 support possessed adequate
conductivity and higher electrochemical stability than commercial Electronic properties.— To measure the electron conductivity of
carbon black; and b) to compare the stability and activity of a 20% the prepared supports, electrochemical impedance spectroscopy mea-
Pt supported on Ta modified TiO2 catalyst against 20% Pt/C prepared surements were performed on pellet samples using a Gamry potentio-
in-house and against commercial 46% Pt/C (benchmark) in a PEFC. stat over the frequency range 0.1 to 105 Hz with an oscillating voltage
of 10 mV. All experiments were performed at room temperature. Sam-
ple preparation details and the schematic of the conductivity cell are
Experimental
provided in the supplementary information.
Chemicals.— Reagents used during synthesis: Tantalum (V)
ethoxide (Ta[OEt]5 ) – 99.98% - Aldrich, titanium isopropoxide
Electrochemical properties.— All electrochemical studies were
(Ti[OiPr]4 ) – 98% - Acros, absolute ethanol (C2 H5 OH) – 200 proof –
performed in a cell comprising 0.1 M HClO4 electrolyte, a Pt-wire
Fisher Scientific, chloroplatinic acid hexahydrate (H2 PtCl6 .6H2 O) –
counter electrode, and a saturated calomel reference electrode (SCE).
ACS Reagent – Sigma-Aldrich, acetone (C3 H6 O) – 99.9% - Aldrich,
A rotating disk electrode (RDE) with a Pt ring and a glassy carbon
Vulcan carbon XC-72R – Cabot, 46% Pt/C – Tanaka Kikinzoku Kogyo
disk (0.196 cm2 ) was used as the working electrode. 10 mg of the
(TKK), methanol (CH3 OH) – 99.9% - Fisher, Nafion - 5% - Aldrich,
catalyst was added to 3 mL of DI water and 1 mL of isopropanol.
perchloric acid (HClO4 ) – 70%, trace metal grade – Fisher, nitrogen,
100 μL of 5% Nafion dispersion was then added and the mixture
oxygen, 4% hydrogen (balance argon) – purified – Praxair, deionized
was stirred and sonicated for 2 hours. This Nafion loading translated
water (DI H2 O) – prepared fresh daily (18 M) – with Direct Q5
to an ionomer/catalyst weight ratio of 0.43/1 which was consistent
water filtration system by Millipore. All chemicals were used without
with prevalent practice as described in the literature.5 8-10 μL of the
further purification.
catalyst ink was dispersed onto the glassy carbon disk electrode using
Synthesis of Ta modified TiO2 .—Ta modified TiO2 supports were a micropipette. The dispersion was spread across the surface of the
synthesized using a procedure modified from those available in disk and was dried to yield a thin film of the active material.50 Cyclic
literature.44,46 Synthesis of Ta modified TiO2 supports were carried voltammograms (CVs) recorded in nitrogen were used to obtain the
out using a sol-gel technique, employing alkoxides of tantalum and ti- capacitance/pseudocapacitance and the electrochemically active sur-
tanium as precursors in conjunction with a high water: alkoxide ratio. face area (ECSA). To calculate the ECSA, the charge corresponding
0.7 mL of 2.7 mmol of tantalum ethoxide [Ta(OEt)5 ] was mixed with to desorption of a full monolayer of hydrogen from polycrystalline Pt
8 mL of 27 mmol of titanium isopropoxide [Ti(OiPr)4 ] in 92 mL of was assumed to be 210 μC/cm2 . The ECSA of platinum was calculated
absolute ethanol, added dropwise to a ethanol/water 1:1 solution and using the formula:
stirred for 4 hours. The resultant mixture was centrifuged and then
 
washed with deionized water and acetone followed by drying in an Q H −adsor ption (C)
oven at 60◦ C for 8 hours. This resulted in the formation of 10% Ta EC S A (m 2 g −1 ) = 105
210 μC cm −2 L Pt (mg cm −2 )A g (cm 2 )
modified TiO2 . Similarly, 20% and 30% Ta modified TiO2 were also [2]
synthesized and tagged as Tax Ti1-x O2 (where x = molar ratio of Ta to where, QH-adsorption was the H2 adsorption/desorption charge calculated
Ti). Unmodified TiO2 using the above procedure was also synthesized from the CV, LPt was the loading of Pt (mgPt cm−2 ) on the working
for comparison. The samples were annealed at 850◦ C in a tube furnace electrode and Ag (cm2 ) was the geometric area of the working elec-
in a reducing atmosphere of 4%H2 in Argon for 3 hours. trode. The change in ECSA as a function of simulated load cycling was
Synthesis of Pt supported on Ta modified TiO2 .—Pt deposition was car- evaluated from the cyclic voltammograms obtained during accelerated
ried out using an improved incipient wetness method similar to those testing of the catalysts. The electrolyte was purged with oxygen for 30
available in the literature.47,48 To obtain a nominal loading of 20wt% min prior to oxygen reduction reaction studies. The oxygen reduction
Pt, the following steps were followed. 0.1327 g of H2 PtCl6 .6H2 O was linear sweep voltammograms were recorded at rotation rates of 400,
added to 30 mL of ethanol and stirred for 30 min. Then, 0.2 g of 900, 1600 and 2500 rpm. The oxygen reduction current was calculated
Ta0.3 Ti0.7 O2 was added to the mixture and sonicated for 1 hour. The from the difference between currents measured in the nitrogen- and
suspension was stirred continuously at 60◦ C to evaporate the solvent. oxygen-saturated electrolytes.
After a few minutes, the resultant thick and smooth slurry was dried The accelerated test protocols for measuring support and electro-
at 80◦ C and ground using a mortar and pestle. To ensure complete catalyst stability are shown in Figure S2 (under supplementary infor-
reduction of the Pt precursor to metallic platinum, the powder was mation). The support stability protocol was designed to aggressively
placed in a ceramic boat and heated in a tube furnace at 120◦ C under simulate start-up/shut-down conditions that occur during fuel cell op-
flowing H2 /Ar. A similar procedure was also used to prepare a 20% erations. Under start-stop conditions, severe potential excursions are
Pt deposited on Vulcan XC-72 carbon. known to occur. The support (typically carbon) is extensively corroded
Microstructure.—Transmission electron microscopy (TEM) micro- during these transients. This support stability protocol cycles the elec-
graphs of the various samples were obtained using a JEOL- 3010F trode between 1 V and 1.5 V vs. RHE at a scan rate of 500 mV/s in a
electron microscope operating at 300 kV. Elemental analysis was per- triangular wave-form to mimic the start-stop cycle (the test runs over
formed for all samples using X-ray energy dispersive spectroscopy 10,000 cycles). It has been established by automotive OEMs that this
(XEDS). X-ray powder diffraction patterns were measured on a Bruker cycle effectively mimics the potential transients seen at the cathode
D2 Phaser diffractometer equipped with a LynxEye position-sensitive during fuel cell start-up and shut-down when mixed fuel-air streams
detector. Samples were either mounted in standard sample holders, are present at the anode.51–53 The capacitance/pseudocapacitance of
or as thin layers mounted in zero-background cells using double- the support was calculated from the cyclic voltammograms (CV) ob-
sided adhesive tape. The experimental conditions were: 20-130 2θ, tained periodically during this test.
0.0202144◦ steps, 0.5 sec/step, 0.6 mm divergence slit, 2.5E Soller To evaluate electrocatalyst stability, a test protocol was designed to
slits, 3 mm scatter screen height. Quantitative phase analysis of the simulate no-load to full-load transitions during fuel cell operation, also
crystalline phases was carried out by the Rietveld method using the referred to as the load cycling protocol.51–53 This protocol involved
General Structure Analysis System (GSAS).49 Surface area measure- cycling the material between 0.6 and 0.95 V vs. RHE, dwelling at
ments were performed using a Quantachrome NOVA 2200 analyzer. each potential for 3 seconds, for a total of 10,000 cycles The evolution
 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013) F1209

of ECSA under simulated load cycling was monitored to evaluate the

stability of the platinum catalyst on the various supports.
Electrocatalytic activity.—Calculation of mass and specific activity:
The ORR polarization curves were corrected by subtracting back-
ground current measured under identical conditions under N2 atmo-
sphere. Then, I was taken as the value of the measured current at E
= 0.90 V and Ilim was taken as the value of measured current at E
= 0.40 V versus RHE following the procedure given in reference.50
The mass- and area-specific activities were calculated after using
the mass-transport correction for thin-film RDEs:5
(Ilim ∗ I )
Ik = [3]
Ilim − 1
where: Ik (mA) is the kinetic current and Ilim (mA) is the measured
limiting current. The Pt area-specific activities were estimated by
calculating Ik and normalizing to the Pt ECSA. The Pt mass-specific
activities were estimated by calculating Ik and normalizing this value
with the Pt loading on the disk electrode.

Results and Discussion Figure 1. X-ray diffraction patterns for Tax Ti1-x O2 samples with different Ta
doping levels. The inset shows the position of the (110) peak.
Stoichiometric rutile and anatase TiO2 are semiconductors with
band gaps of 3.0 and 3.2 eV at room temperature, respectively.25
Unmodified and modified TiO2 samples annealed at or above 850◦ C
existed almost entirely in the rutile phase (lower bandgap). Higher DFT studies.55 The diffraction peaks obtained (Figure 2a) from the
temperature processing (> 850◦ C) resulted in materials with lower XRD of Ta0.3 Ti0.7 O2 could be completely indexed to a tetragonal unit
BET surface area; therefore 3 modified samples of Tax Ti1-x O2 (x = 0.1, cell, and were found to be shifted slightly from that of pure rutile TiO2
0.2 and 0.3) were synthesized and subsequently annealed at 850◦ C. (JCPDS no. 88-1175) due to Ta doping. The presence of well-defined
TEM and XEDS images (Figures S3-S6 shown under supplementary fringes at ∼3.2 Å (Figure 2b, inset) corresponding to the spacing of
information) revealed the presence of modified TiO2 with an average the (110) plane of rutile TiO2 indicated a high degree of crystallinity in
particle size in the range 70-85 nm and also confirmed the presence of the Ta0.3 Ti0.7 O2 material, which correlated well with the XRD pattern.
both Ti and Ta in the support. XEDS measurements (Figures S3, S4 The BET surface area, electron conductivity and mean particle sizes
and S5), indicated a Ti:Ta atomic ratio of 89.72:10.28, 81.40:18.60 and of all samples are tabulated in Table II. The slight increase in BET
68.85:31.15 (the model used was Materials-Biological Thin Specimen surface area of the Ta-modified samples was attributed to their smaller
(MBTS)), which agreed closely with the expected atomic ratio of particle size. Growth in particles sizes was expected in all the samples
90:10, 80:20 and 70:30 for Ta0.1 Ti0.9 O2 , Ta0.2 Ti0.8 O2 and Ta0.3 Ti0.7 O2 . as they were treated at the same temperature of 850◦ C. However due
This suggested that the composition of the (Ti:Ta) modified metal to doping, the growth in particle sizes of the modified TiO2 metal
oxide could be controlled in a facile manner by adjusting the ratios oxides were slightly inhibited due to distortion in the crystal lattice.56
of Ta(OEt)5 and Ti(OiPr)4 in the precursor solutions without the need The incorporation of Ta significantly increased the electrical conduc-
for a template. tivity of the particles. The maximum conductivity at room tempera-
Ta is an n-type dopant for TiO2 and can be substitutionally incor- ture of the 30% Ta modified sample (Ta0.3 Ti0.7 O2 ), which approached
porated within the rutile lattice to occupy at least 30% of the cation 0.2 S/cm, was remarkably high for a wide bandgap semiconductor
sites, as evidenced by the X-ray diffractograms shown in Figure 1. The such as titania.
strong diffraction peaks at about 27◦ , 36◦ and 55◦ arose from diffrac- The key requirement for a PEFC electrocatalyst support of electron
tions from rutile TiO2 (110), (101) and (211) planes, respectively. An conductivity in excess of 0.1 S/cm57,58 was met only by Ta0.3 Ti0.7 O2
increase in the Ta (dopant) content resulted in a shift of the (110), (101) among the formulations studied. This recommended support conduc-
and (200) reflections to lower angles (Figure 1, inset), corresponding tivity value of 0.1 S/cm represents a value that is 5 times in ex-
to a unit cell expansion of about 1%, 3% and 5% for the Ta0.1 Ti0.9 O2 , cess of the maximum conductivity of the ionomer in the electrode
Ta0.2 Ti0.8 O2 and Ta0.3 Ti0.7 O2 particles, respectively, as shown in (ca. 0.02 S/cm – see references 57 and 58) and is deemed adequate
Table I. This expansion was due to the replacement of Ti4+ ions (radius to prepare acceptable porous electrodes. Prior work in literature has
60.5 pm) by the slightly larger Ta5+ ions (radius 64 pm) in accordance also demonstrated that it is possible to achieve relatively high elec-
with Vegard’s law.54 Indeed, Ta formed a complete range of solid so- trocatalyst performance with support conductivity on the order of
lutions with TiO2 . Furthermore, Ta was completely incorporated into 0.1 S/cm.59,60 Hence, only Ta0.3 Ti0.7 O2 was further evaluated and Pt
the TiO2 structure - no diffraction peaks from tantalum oxide (TaO2 was deposited on it using an improved incipient wetness technique47,48
and/or Ta2 O5 ) were observed in any of the Tax Ti1−x O2 samples. For and evaluated for electrochemical performance.
all the 3 supports, Ta doping caused a slight lattice expansion of the The diffraction pattern of Pt supported on Ta0.3 Ti0.7 O2 is shown
TiO2 crystallite and resulted in increased electrical conductivity by in Figure 3a. The presence of Pt nanoparticles was confirmed by
shifting the Fermi level to the unfilled bands in TiO2 as predicted by the peak present at 2θ≈40◦ corresponding to the (111) phase of the

Table I. Lattice parameters of the Ta modified TiO2 supports and 20% Pt/Ta0.3 Ti0.7 O2 .

Samples TiO2 Ta0.1 Ti0.9 O2 Ta0.2 Ti0.8 O2 Ta0.3 Ti0.7 O2 20% Pt/Ta0.3 Ti0.7 O2
a, Å 4.59404(4) 4.61909(13) 4.63944(8) 4.6638(2) 3.9175(4)
c, Å 2.95956(3) 2.97368(8) 2.99419(5) 3.0148(1) -
V, Å3 62.462(2) 63.048(7) 64.487(3) 65.576(7) -
Strain,% 0.13(1) 0.50(1) 0.74(1) 0.94(1) -
Pt, wt% - - - - 19.9(1)
F1210 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013)

Figure 2. (a) X-ray diffraction pattern of Ta0.3 Ti0.7 O2 . (b) TEM image of Ta0.3 Ti0.7 O2 .

Table II. BET surface area, electron conductivity and mean on Ta0.3 Ti0.7 O2 was observed at a binding energy (BE) of 71.40 eV,
particle size of the various samples. which was ca. 600 meV lower when compared to the corresponding
signal for Pt deposited on carbon (∼72.0 eV). This difference in the Pt
Electron 4f BE was attributed to a local increase in the electron density at the Pt
BET surface Conductivity Mean particle site when it was deposited on Ta0.3 Ti0.7 O2 and confirmed that strong
Sample area (m2 /g) (mS/cm) size (nm) metal support interactions (SMSI) existed in 20% Pt/Ta0.3 Ti0.7 O2 . A
similar decrease in Pt 4f BE has also been reported for Pt deposited on
TiO2 9±2 0.03 ± 0.07 84
Ta0.1 Ti0.9 O2 12 ± 2 1.2 ± 3 80
other oxides of titanium.61–63 Likewise, the increased electron density
Ta0.2 Ti0.8 O2 18 ± 1 5.78 ± 2 72 has been attributed to SMSI in these studies.
Ta0.3 Ti0.7 O2 26 ± 3 209 ± 31 67 The stability of the supports was estimated from the change in
capacitance/pseudocapacitance over 10,000 cycles of an accelerated
fcc lattice of metallic platinum. The average crystallite size of Pt, support stability protocol designed to mimic electrode potential tran-
calculated using the Scherrer equation, was found to be 4.0 nm. The sients seen during fuel cell start-up / shut-down and under fuel
low intensities and broadening of the peaks was attributed to the small starvation conditions. The CVs recorded for baseline Vulcan XC-
particle size of Pt nanoparticles, as confirmed by TEM (Figure 3b). 72R carbon during an identical accelerated support stability test (1–
The fringes with a lattice spacing of 2.2 Å were indexed as the (111) 1.5 V vs. RHE; 10000 cycles) showed evidence of substantial sur-
plane of fcc Pt (Figure 3b, inset), indicating the existence of good face oxide formation through increased peak current assigned to the
crystallinity in the Pt nanoparticles anchored onto the Ta0.3 Ti0.7 O2 hydroquinone−quinone (HQ−Q) redox couple of carbon13,64 (Figure
support. Quantitative phase analysis of the crystalline phases carried S7a, supplementary information). This indicated significant carbon
out by the Rietveld method using the General Structure Analysis oxidation occurred during the accelerated test. On the other hand,
System (GSAS) confirmed that the targeted Pt loading of 20 wt% was the CVs for the Ta0.3 Ti0.7 O2 support (see Figure S7d, supplementary
achieved on the Ta0.3 Ti0.7 O2 support (Table I). Besides, the TEM image information), showed little change in capacitance upon cycling up
shown in Figure 3b also revealed the presence of well-dispersed Pt to 10,000 cycles. Although carbon showed a change in pseudoca-
particles on this support, with particle sizes ranging from 2-5 nm. XPS pacitance of more than 100%, the observed change in the capacitance
data, shown in Figure 4, revealed that the Pt 4f signal for Pt deposited (normalized to initial value) was only 12% for Ta0.3 Ti0.7 O2 (Figure 5a).

Figure 3. (a) X-ray diffraction pattern of 20% Pt/Ta0.3 Ti0.7 O2 . The Ta0.3 Ti0.7 O2 support is also shown for comparison. The peaks marked by asterisks (*)
correspond to the (111), (200), (220) and (311) of the face-centered cubic (fcc) Pt. (b) TEM image of the 20% Pt/Ta0.3 Ti0.7 O2 catalyst.
 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013) F1211

metal support interactions (SMSI) as confirmed from XPS results (see

Figure 4). Note that this result was very much in line with expecta-
tions. The degradation of the platinum catalyst occurs through many
mechanisms, most of which are independent from the degradation
of the electrocatalyst support. Several mechanisms have been identi-
fied in literature to explain the reduction in ECSA caused during load
cycling.66,67 In one commonly proposed mechanism, the platinum par-
ticles present are postulated to dissolve (Pt -> Pt2+ +2e), migrate, and
get re-deposited on other platinum particles, resulting in particle size
growth via Ostwald ripening. It has also been shown that at potentials
below 1 V, the presence of adsorbed oxygen species on Pt (Pt-OH)
could lead to carbon support corrosion.68 Pt agglomeration caused by
support corrosion can also cause the decrease in ECSA. Realistically,
the latter mode of ECSA loss is the only mode that can be mitigated by
using a corrosion resistant support. The results obtained suggested that
the loss in ECSA during load cycling was modestly lowered by using
a corrosion resistant support. The lower surface free energy of these
larger Pt nanoparticles in 20% Pt/Ta0.3 Ti0.7 O2 (compared to the 3 nm
average diameter particles in the 46% Pt/C baseline) would also lead
to lower rates of platinum dissolution and hence lower ECSA loss.
Thus, the apparent increase in stability relative to 46% Pt/C could also
have arisen due to the intrinsically larger size of the platinum particle.
Figure 4. XPS of the Pt 4f region in 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46%
Pt/C. Figure 5b clearly demonstrates the impact of using the Ta0.3 Ti0.7 O2
versus a traditional carbon support on catalyst stability. The ECSA
loss was significantly larger in 20% Pt/C than in 20% Pt/Ta0.3 Ti0.7 O2 .
TiO2 is known to be stable in acidic environments over a broad po- TEM images of pre and post cycled 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and
tential range,65 which explains the very modest change in capacitance 46% Pt/C catalysts (Figure 6) confirmed the higher stability of 20%
during cycling. The support stability test on the 20% Pt/Ta0.3 Ti0.7 O2 Pt/Ta0.3 Ti0.7 O2 under load cycling. Intensive growth of Pt particles was
sample (Figure 5a) was carried out to investigate whether addition of observed on both 46% Pt/C and 20% Pt/C, which was consistent with
Pt resulted in a change in Ta0.3 Ti0.7 O2 stability (i.e. did Pt addition prior reports in the literature.69,70 On the other hand, the Pt particle
catalyze support corrosion?); the results confirmed that the support size in 20% Pt/Ta0.3 Ti0.7 O2 were almost unchanged. This observation
was equally stable even after addition of platinum, demonstrating that confirmed that the observed enhancement in catalyst stability on the
the added platinum did not in any way accelerate the corrosion of the non-carbon support was truly due to metal-support interactions, and
support. not due to extraneous factors such as differences in initial particle size
The stability of the electrocatalyst was estimated from the change or inter-particle distance (arising from different platinum loadings).
in the ECSA over 10,000 cycles of an accelerated load cycling test de- The positive interactions between the Pt particles and the oxide sup-
signed to mimic full load-no load transients in an operating PEFC. The port (which we also demonstrated using XPS) helped enhance catalyst
response from load cycling tests performed on 20% Pt/Ta0.3 Ti0.7 O2 , stability under accelerated load cycling.
20% Pt/C and 46% Pt/C electrocatalysts (shown in Figure 5b) re- The activity of the 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C
vealed that the ECSA of the 20% Pt/C as well as the commercial toward the oxygen reduction reaction (ORR) was evaluated in a RDE
46% Pt/C catalyst decreased substantially, with a 44% and 47% loss apparatus. The results are shown in Figure 7. The ECSA, mass-specific
after 10000 cycles respectively. The 20% Pt/Ta0.3 Ti0.7 O2 electrocata- activity and area-specific activity of 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C
lyst exhibited a lower, but still significant, ECSA loss of ∼35% over and 46% Pt/C catalysts are all reported in Table III. The lower value of
10,000 cycles. The lower extent of degradation was attributed to strong ECSA of 20% Pt/Ta0.3 Ti0.7 O2 and 20% Pt/C compared to commercial

Figure 5. (a) Normalized capacitance/pseudocapacitance of carbon, Ta0.3 Ti0.7 O2 , and 20% Pt/Ta0.3 Ti0.7 O2 as a function of cycle number under an accelerated
support stability test protocol designed to test electrocatalyst support stability. (b) Normalized ECSA of 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C as a function
of cycle number under an accelerated load cycling test designed to test electrocatalyst stability.
F1212 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013)

Figure 6. TEM images of the catalyst particles of 20%

Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C before and after
10000 load cycles.

46% Pt/C was attributed to the differences in platinum particle size effect. However, this result was not observed; the 20% Pt/Ta0.3 Ti0.7 O2
(around 4-5 nm average for 20% Pt/Ta0.3 Ti0.7 O2 as well as 20% Pt/C and 20% Pt/C catalysts both had ∼25% lower area-specific activities
and 3 nm average for 46% Pt/C catalyst) and to the non-optimized than 46% Pt/C.
ionomer/support ratio in the 20% Pt/Ta0.3 Ti0.7 O2 electrode. The mass
and specific activities of 20% Pt/Ta0.3 Ti0.7 O2 and 20% Pt/C were ORR kinetics: Koutecky–Levich (K-L) analysis.— Figure 8a-8c
nearly identical, but were both lower than the corresponding values show the oxygen reduction currents measured at four different ro-
for 46% Pt/C for the reasons mentioned above. The mass activity tation rates, 400, 900, 1600 and 2500 rpm for 20% Pt/Ta0.3 Ti0.7 O2 ,
of platinum depends considerably on the platinum particle size and 20% Pt/C and 46% Pt/C respectively. These currents were used to
morphology,71 while the specific activity is postulated to be nearly construct the K-L plot for each catalyst. Figure 8d-8f show the K-L
independent of, and perhaps even increases with crystallite size (crys- plot of i−1 and ω−1/2 for 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C
tallite size effect).72 Lower availability of surface platinum atoms for various disk potentials from 0.85 V to 0.65 V. It can be seen that the
due to poor dispersion leads to poor mass activity in electrocatalysts constant overpotential lines were parallel and distinct for potentials
with larger platinum crystallites. 46% Pt/C, having smaller platinum 0.65 V and above. Information on the kinetic parameters and number
particles (see TEM results), demonstrated superior mass activity and of electrons transferred (n) during the oxygen reduction reaction at
turnover compared to 20% Pt/C and 20% Pt/Ta0.3 Ti0.7 O2 . Since the various potentials were extracted from the K–L plot. The kinetic cur-
ECSA was lower for 20% Pt/Ta0.3 Ti0.7 O2 and for 20% Pt/C, it was ex- rent density denotes the intrinsic activity of the catalyst surface toward
pected that their area-specific activity would be equivalent or perhaps the ORR and was extracted from the intercept of the K–L plot for all
a bit larger than that of 46% Pt/C in accordance with the crystallite size catalysts at various overpotentials spread within 200 mV from onset
 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013) F1213

Table III. Comparison of figures of merit obtained from RDE

studies.

Im (mA/mgPt ) Is (μA/cm2 Pt )
Sample ECSA (m2 /g) @0.9 V @0.9 V
20% Pt/Ta0.3 Ti0.7 O2 41 ± 1 62 ± 5 151 ± 7
20% Pt/C 43 ± 2 65 ± 3 147 ± 4
46% Pt/C 72 ± 3 145 ± 6 201 ± 11
46% Pt/Ca 86 ± 7 160 210

a Extractedfrom literature, Gasteiger et al.5 these RDE experiments

were carried out at 60◦ C in 0.1 M HClO4 at a sweep rate of 20 mV/s.

potential (Eo = 0.95 V). Figure 8g illustrates the relation between the
kinetic current and the overpotential in all catalysts. Both the Tafel
slope and exchange current density (i0 ), were calculated from this plot.
Figure 7. Oxygen reduction reaction polarization curves of 20% We recognize that there is a lot of scope for error in estimating the ex-
Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C obtained on a thin-film RDE at
1600 rpm in 0.1 M HClO4 at 5 mV/s.
change current density due to extrapolation along a logarithmic scale.
The Tafel slopes were found to be nearly identical in all three cata-

Figure 8. Oxygen reduction currents measured at four different rotation rates, 400, 900, 1600 and 2500 rpm for (a) 20% Pt/Ta0.3 Ti0.7 O2 , (b) 20% Pt/C and (c)
46% Pt/C. Koutecky–Levich analysis of ORR data for (d) 20% Pt/Ta0.3 Ti0.7 O2 , (e) 20% Pt/C and (f) 46% Pt/C. (g) Kinetic currents for 20% Pt/Ta0.3 Ti0.7 O2 , 20%
Pt/C and 46% Pt/C at various overpotentials (Tafel analysis).
F1214 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013)

lysts – 69± 2, 69 ± 2 and 67 ± 2mV/decade for 20% Pt/Ta0.3 Ti0.7 O2 , 4. D. P. Davies, P. L. Adcock, M. Turpin, and S. J. Rowen, Journal of Applied Electro-
20% Pt/C and 46% Pt/C respectively, while the calculated exchange chemistry, 30, 101 (1999).
5. H. A. Gasteiger, S. Kocha, B. Sompalli, and F. T. Wagner, Appl. Catal., B, 56, 9
current densities were found to be 0.016 ± 0.001, 0.017 ± 0.002 and (2005).
0.019 ± 0.002 mA/cm2 for 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% 6. M. Cai, M. S. Ruthkosky, B. Merzougui, S. Swathirajan, M. P. Balogh, and S. H. Oh,
Pt/C respectively. The similarity of Tafel slopes and in the number of Journal of Power Sources, 160, 977 (2006).
electrons transferred in 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% Pt/C 7. T. J. Schmidt, H. A. Gasteiger, G. D. Stab, P. M. Urban, D. M. Kolb, and R. J. Behm,
J. Electrochem. Soc., 145, 2354 (1998).
catalysts for ORR (3.84, 3.80 and 3.90, respectively) revealed that all 8. C. A. Reiser, L. Bregoli, T. W. Patterson, J. S. Yi, J. D. Yang, M. L. Perry, and
three catalysts had similar ORR mechanisms. While it is tempting to T. D. Jarvi, Electrochem. Solid-State Lett., 8, A273 (2005).
conclude that 46% Pt/C exhibited slightly higher ORR activity than 9. X. Wang, W. Li, Z. Chen, M. Waje, and Y. Yan, J. Power Sources, 158, 154 (2006).
20% Pt/C and 20% Pt/Ta0.3 Ti0.7 O2 based on the trends in the exchange 10. S. Maass, F. Finsterwalder, G. Frank, R. Hartmann, and C. Merten, Journal of Power
Sources, 176, 444 (2008).
current densities, one must temper this conclusion with the likelihood 11. K. Kinoshita and J. A. S. Bett, Carbon, 11, 403 (1975).
for error during estimation of these values. 12. L. M. Roen, C. H. Paik, and T. D. Jarvi, Electrochem. Solid-State Lett., 7, A19 (2004).
13. K. H. Kangasniemi, D. A. Condit, and T. D. Jarvi, J. Electrochem. Soc., 151, E125
(2004).
Conclusions.— Ta modified TiO2 supports were successfully syn- 14. E. Antolini and E. R. Gonzalez, Solid State Ionics, 180, 746 (2009).
thesized, characterized and evaluated in an electrochemical environ- 15. S. A. G. Evans, J. G. Terry, N. Plank, A. Walton, L. Keane, C. Campbell, P. Ghazal,
ment. 3 different samples of Tax Ti1-x O2 (x = 0.1, 0.2 and 0.3) were J. Beattie, T. Su, J. Crain, and A. Mount, Electrochem. Commun., 7, 125 (2005).
prepared and XRD confirmed that Ta was completely incorporated 16. O. T. M. Musthafa and S. Sampath, Chemical Communications, 67 (2008).
17. B. Avasarala and P. Haldar, Electrochimica Acta, 55, 9024 (2010).
within the TiO2 rutile lattice with an increase in expansion in the TiO2 18. Y.-J. Wang, D. P. Wilkinson, and J. Zhang, Chemical Reviews, 111, 7625 (2011).
unit cell with increasing Ta content. Initial work was directed toward 19. N. Liu, K. Kourtakis, J. C. Figueroa, and J. G. Chen, J. Catal., 215, 254 (2003).
meeting the conductivity and surface area requirements for a catalyst 20. H. Chhina, S. Campbell, and O. Kesler, J. Power Sources, 164, 431 (2007).
support in a PEFC. Unlike Ta0.1 Ti0.9 O2 and Ta0.2 Ti0.8 O2, Ta0.3 Ti0.7 O2 21. M. B. Zellner and J. G. Chen, Catal. Today, 99, 299 (2005).
22. R. Ganesan, D. J. Ham, and J. S. Lee, Electrochem. Commun., 9, 2576 (2007).
was found to have adequate electron conductivity (0.2 S/cm) and 23. Y. Hara, N. Minami, and H. Itagaki, Appl. Catal., A, 323, 86 (2007).
sufficient BET surface area (26 m2 /g). Pt supported on Ta0.3 Ti0.7 O2 24. Y. Hara, N. Minami, H. Matsumoto, and H. Itagaki, Appl. Catal., A, 332, 289 (2007).
was synthesized for further evaluation. The durability and catalytic 25. M. Gratzel, Nature (London, U. K.), 414, 338 (2001).
activity of this material was measured and benchmarked against the 26. A. L. Linsebigler, G. Lu, and J. T. Yates Jr., Chem. Rev. (Washington, D. C.), 95, 735
(1995).
state-of-the-art 46% Pt/C TKK electrocatalyst as well as 20% Pt/C 27. A. Nakazawa and T. Shinoda, Photochromics and photochromic materials based on
catalyst prepared in-house. Ta0.3 Ti0.7 O2 exhibited an order of mag- composite oxide films and transparent conductors, and methods for their manufac-
nitude higher electrochemical stability than the benchmark carbon turing, in, p. 16 pp., Fujitsu, Limited, Japan. (2003).
material when examined using an aggressive accelerated test protocol 28. L. D. Birkefeld, A. M. Azad, and S. A. Akbar, J. Am. Ceram. Soc., 75, 2964 (1992).
29. X. Chen and S. S. Mao, Chem. Rev. (Washington, DC, U. S.), 107, 2891 (2007).
that simulated 10000 start-up and shut-down cycles (support stability 30. D. Wang, C. V. Subban, H. Wang, E. Rus, F. J. DiSalvo, and H. D. Abruňa, J. Am.
test protocol: 1.0-1.5 V vs. RHE at a scan rate of 500 mV/s in a 0.1 Chem. Soc., 132, 10218 (2010).
M HClO4 electrolyte at 25◦ C). 20% Pt/Ta0.3 Ti0.7 O2 also demonstrated 31. S. Y. Huang, P. Ganesan, S. Park, and B. N. Popov, J. Am. Chem. Soc., 131, 13898
greater electrochemical stability than both Pt/C catalysts, as estimated (2009).
32. T. Ioroi, Z. Siroma, N. Fujiwara, S. Yamazaki, and K. Yasuda, Electrochem. Commun.,
using the load cycling protocol, showing a 35% vs. a 44-47% loss in 7, 183 (2005).
ECSA over 10,000 load cycles (load cycling test protocol: 0.6-0.95 V 33. K. Kolbrecka and J. Przylski, Electrochim. Acta, 39, 1591 (1994).
vs. RHE with a dwell time of 3 s at either potentials in a 0.1 M HClO4 34. G. Chen, S. R. Bare, and T. E. Mallouk, J. Electrochem. Soc., 149, A1092 (2003).
electrolyte at 25◦ C). Pre- and post-test (load cycle test) TEM images 35. L. L. Y. Chang, Journal of the American Ceramic Society, 51, 295 (1968).
36. J. Horkans and M. W. Shafer, Journal of The Electrochemical Society, 124, 1202
and pre-test XPS spectra of 20% Pt/Ta0.3 Ti0.7 O2 , 20% Pt/C and 46% (1977).
Pt/C revealed the presence of strong metal support interaction (SMSI) 37. L. Pejryd, Journal of the American Ceramic Society, 69, 717 (1986).
in 20% Pt/Ta0.3 Ti0.7 O2 . Koutecky-Levich analysis suggested similar 38. G. Chen, C. C. Waraksa, H. Cho, D. D. Macdonald, and T. E. Mallouka, Journal of
oxygen reduction reaction (ORR) pathways in the 20% Pt/Ta0.3 Ti0.7 O2 , The Electrochemical Society, 150, E423 (2003).
39. B. L. Garcı́a, R. Fuentes, and J. W. Weidner, Electrochemical and Solid-State Letters,
20% Pt/C and the 46% Pt/C, deduced from the similarity in Tafel slope 10, B108 (2007).
(69 mV/decade, 69 mV/decade and 67 mV/decade respectively) and in 40. K. Sasaki, L. Zhang, and R. R. Adzic, Physical Chemistry Chemical Physics, 10, 159
the number of electrons transferred (3.84, 3.80 and 3.90 respectively). (2008).
The mass activity, determined at 0.9 V vs. RHE in a 0.1 M HClO4 41. K.-W. Park and K.-S. Seol, Electrochem. Commun., 9, 2256 (2007).
electrolyte at 25◦ C, of the 20% Pt/Ta0.3 Ti0.7 O2 (62 mA/mgPt ) presently 42. S.-Y. Huang, P. Ganesan, and B. N. Popov, Applied Catalysis B: Environmental, 96,
224 (2010).
falls short of the benchmark 46% Pt/C (145 mA/mgPt ), however there 43. Y.-J. Wang, D. P. Wilkinson, and J. Zhang, Dalton Transactions, 41, 1187 (2012).
is adequate room for improvement via detailed optimization of the 44. E. Traversa, M. L. Di Vona, S. Licoccia, M. Sacerdoti, M. C. Carotta, M. Gallana,
various steps involved in support and electrocatalyst preparation. In and G. Martinelli, Journal of Sol-Gel Science and Technology, 19, 193 (2000).
45. X. Feng, K. Shankar, M. Paulose, and C. A. Grimes, Angewandte Chemie Interna-
closing, Ta0.3 Ti0.7 O2 is an excellent corrosion-resistant support and is tional Edition, 48, 8095 (2009).
a viable alternative to carbon, particularly for automotive applications. 46. D. C. Hague and M. J. Mayo, Journal of the American Ceramic Society, 77, 1957
(1994).
47. B. Yang, Q. Lu, Y. Wang, L. Zhuang, J. Lu, P. Liu, J. Wang, and R. Wang, Chemistry
Acknowledgments of Materials, 15, 3552 (2003).
48. D. Wang, L. Zhuang, and J. Lu, The Journal of Physical Chemistry C, 111, 16416
The authors acknowledge Prof. James Kaduk, Department of Bi- (2007).
ological and Chemical Sciences, Illinois Institute of Technology for 49. B. H. Toby, J. Appl. Crystallogr., 34, 210 (2001).
powder X-ray diffractograms and Rietveld analyzes. This work was 50. Y. Garsany, O. A. Baturina, K. E. Swider-Lyons, and S. S. Kocha, Analytical Chem-
istry, 82, 6321 (2010).
supported by NSF Award # 0847030. 51. C. Wang, N. Dale, and K. Adjemian, ECS Trans., 41, 2245 (2011).
52. C.-P. Lo and V. Ramani, ACS Appl. Mater. Interfaces, 4, 6109 (2012).
53. A. Kumar and V. K. Ramani, Applied Catalysis B: Environmental, 138–139, 43
References (2013).
54. Y. Liu, J. M. Szeifert, J. M. Feckl, B. Mandlmeier, J. Rathousky, O. Hayden,
1. R. Borup, J. Meyers, B. Pivovar, Y. S. Kim, R. Mukundan, N. Garland, D. Myers, D. Fattakhova-Rohlfing, and T. Bein, ACS Nano, 4, 5373 (2010).
M. Wilson, F. Garzon, D. Wood, P. Zelenay, K. More, K. Stroh, T. Zawodzinski, 55. E. Dy, R. Hui, J. Zhang, Z.-S. Liu, and Z. Shi, The Journal of Physical Chemistry C,
J. Boncella, J. E. McGrath, M. Inaba, K. Miyatake, M. Hori, K. Ota, Z. Ogumi, 114, 13162 (2010).
S. Miyata, A. Nishikata, Z. Siroma, Y. Uchimoto, K. Yasuda, K.-i. Kimijima, and 56. J. D. Bryan and D. R. Gamelin, Prog. Inorg. Chem., 54, 47 (2005).
N. Iwashita, Chemical Reviews, 107, 3904 (2007). 57. H. A. Gasteiger, Electrochemistry (Tokyo, Jpn.), 75, 103 (2007).
2. A. Collier, H. J. Wang, X. Z. Yuan, J. J. Zhang, and D. P. Wilkinson, International 58. P. T. Yu, W. Gu, J. Zhang, R. Makharia, F. T. Wagner, and H. A. Gasteiger, Polymer
Journal of Hydrogen Energy, 31, 1838 (2006). Electrolyte Fuel Cell Durability (2009).
3. D. A. Stevens, M. T. Hicks, G. M. Haugen, and J. R. Dahn, Journal of the Electro- 59. C. V. Subban, Q. Zhou, A. Hu, T. E. Moylan, F. T. Wagner, and F. J. DiSalvo, Journal
chemical Society, 152, A2309 (2005). of the American Chemical Society, 132, 17531 (2010).
 Journal of The Electrochemical Society, 160 (11) F1207-F1215 (2013) F1215

60. V. T. T. Ho, C.-J. Pan, J. Rick, W.-N. Su, and B.-J. Hwang, Journal of the American 67. E. Antolini, J. Mater. Sci., 38, 2995 (2003).
Chemical Society, 133, 11716 (2011). 68. B. Merzougui and S. Swathirajan, Journal of The Electrochemical Society, 153,
61. J. A. Horsley, Journal of the American Chemical Society, 101, 2870 (1979). A2220 (2006).
62. B. H. Chen and J. M. White, The Journal of Physical Chemistry, 86, 3534 (1982). 69. Y. Shao-Horn, W. C. Sheng, S. Chen, P. J. Ferreira, E. F. Holby, and D. Morgan, Top.
63. A. Lewera, L. Timperman, A. Roguska, and N. Alonso-Vante, The Journal of Physical Catal., 46, 285 (2007).
Chemistry C, 115, 20153 (2011). 70. K. Yasuda, A. Taniguchi, T. Akita, T. Ioroi, and Z. Siroma, Physical Chemistry
64. K. Kinoshita, Carbon : electrochemical and physicochemical properties, p. xiii, Chemical Physics, 8, 746 (2006).
Wiley, New York (1988). 71. M. Nesselberger, S. Ashton, J. C. Meier, I. Katsounaros, K. J. J. Mayrhofer, and
65. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions (1974). M. Arenz, Journal of the American Chemical Society, 133, 17428 (2011).
66. P. J. Ferreira, G. J. la O, Y. Shao-Horn, D. Morgan, R. Makharia, S. Kocha, and 72. H. R. Kunz and G. A. Gruver, Journal of The Electrochemical Society, 122, 1279
H. A. Gasteiger, J. Electrochem. Soc., 152, A2256 (2005). (1975).