A surface modification route to nonprecious metal fuel cell catalysts

Article abstract of DOI:10.1149/1.3283100

Nonprecious metal oxygen reduction reaction catalysts have been prepared via a simple chemisorption-based methodology. First, an aminosilane was chemically attached to a carbon surface, followed by coordination of iron and subsequent high temperature pyrolysis. Catalysts were prepared from a series of mono, di, and triaminosilanes. Physical and electrochemical properties of these catalysts were studied using inductively coupled plasma optical emission spectroscopy, X-ray photoelectron spectroscopy, and rotating ring-disk electrode cells. Maximum catalytic activity was achieved for samples prepared from the triaminosilane combined with a more disordered carbon black.

Full text of DOI:10.1149/1.3283100

B370
Journal of The Electrochemical Society, 157 ͑3͒ B370-B375 ͑2010͒
0013-4651/2010/157͑3͒/B370/6/$28.00 © The Electrochemical Society
A Surface Modification Route to Nonprecious Metal Fuel Cell
Catalysts
,z
*
Allen D. Pauric, Andrew W. Pedersen, Tara Andrusiak, and E. Bradley Easton
Faculty of Science, University of Ontario Institute of Technology, Oshawa,
Ontario L1H 7K4, Canada
Nonprecious metal oxygen reduction reaction catalysts have been prepared via a simple chemisorption-based methodology. First,
an aminosilane was chemically attached to a carbon surface, followed by coordination of iron and subsequent high temperature
pyrolysis. Catalysts were prepared from a series of mono, di, and triaminosilanes. Physical and electrochemical properties of these
catalysts were studied using inductively coupled plasma optical emission spectroscopy, X-ray photoelectron spectroscopy, and
rotating ring-disk electrode cells. Maximum catalytic activity was achieved for samples prepared from the triaminosilane com-
bined with a more disordered carbon black.
© 2010 The Electrochemical Society. ͓DOI: 10.1149/1.3283100͔ All rights reserved.
Manuscript submitted October 23, 2009; revised manuscript received November 27, 2009. Published January 19, 2010.
Platinum is presently the best catalyst available for the reduction
lytic activity. One major advantage of this method is that chemisorp-
tion is more specific than physisorption, thereby allowing the attach-
ment to occur at the end of the basal plane. This places the nitrogen
and iron in the location where active sites can form, as opposed to in
the middle of the graphitic sheets where active sites are not believed
to form. The impact of carbon black properties, heat-treatment tem-
perature, and aminosilane selection has been investigated and was
related to the oxygen reduction reaction ͑ORR͒ activity.
of oxygen at the fuel cell cathode. As a consequence of the slow
reaction kinetics surrounding the oxygen reduction of platinum, a
large active catalytic area of platinum is required to maintain high
current densities. Impregnating a high surface area carbon support
with platinum nanoparticles significantly reduces the amount of
platinum needed, though this still constitutes 40% of the cost of a
fuel cell.1 Not only is platinum expensive, but it is also rare. Plati-
num reserves are likely to deplete in the event that the world energy
economy begins a large-scale shift toward polymer electrolyte mem-
brane fuel cells. If the technology is to be made materially sustain-
able, platinum will have to be replaced with more readily available
metals.
Iron–nitrogen complexes have been investigated toward a pos-
sible activity in oxygen reduction. There have been numerous papers
about the synthesis of Fe–N/C catalysts, with the synthetic methods
falling into two main categories. The first and most common method
is the physisorption of iron ͑or cobalt͒ and nitrogen-containing pre-
cursors on a carbon support, followed by pyrolysis at temperatures
between 600 and 1000°C. A vast number of metal precursors, car-
bon supports, and nitrogen sources have been investigated and are
summarized in a couple of recent reviews.2,3 The other category is
physical vapor deposition, which has been used primarily by Dahn’s
group.4-9 In this method, Fe ͑or Co͒ and carbon is cosputtered in a
reactive nitrogen plasma magnetron onto variable substrates, which
are subsequently heat-treated to induce catalytic activity. The cata-
lytic activity is highly dependent on the nitrogen content achieved
after heat-treatment, regardless of the synthetic method.
The exact structure of the catalytically active site is not exactly
known. However, it is presently believed to be Fe ͑or Co͒ coordi-
nated to pyridinic-type and pyrrolic-type nitrogen sites that are em-
bedded into the carbon support at the edge of the basal planes,10
with the pyridinic type believed to be associated with the highest
activity. More recent work from Dodelet’s group further suggests
that iron actually bridges two graphitic sheets via nitrogen
coordination.11-13
Surface modification of electrode structures using organosilane
derivatives has been extensively used to endow an electrode surface
with desired functionalities.14 High surface area carbon blacks pos-
sess surface oxygen functional groups on their surface through
which organosilanes can chemisorb, forming robust linkages. Such
methodologies have been employed in applications such as sensors,
chemically modified electrodes, and modified fuel cell supports.15-17
Here we have grafted nitrogen-containing functional groups to a
carbon black surface via a series of aminosilanes shown in Fig. 1.
After coordination of iron, samples were pyrolyzed to induce cata-
Experimental
Catalysts synthesis.— Modification was performed by using a
combination of two carbon blacks and three aminosilanes. The car-
bon blacks used in this study were Vulcan XC72 and Black Pearls
2000 ͑Cabot͒ and are hereafter referred to as V and BP, respectively.
According to the manufacturer, the specific surface areas of V and
BP carbons were 250 and 1500 m2/g, respectively. The three ami-
nosilanes used for individual modifications were 3-aminopropyl-
trimethoxysilane ͑97%, Sigma͒, N-͓3-͑trimethoxysilyl͒propyl͔-
ethylenediamine ͑97%, Sigma͒, and N -͓3-͑trimethoxysilyl͒-propyl͔
Ј
diethylenetriamine ͑technical grade, Sigma͒, referred to hereafter as
N1, N2, and N3, respectively. Carbon black was modified with an
aminosilane using a procedure similar to that reported by Easton and
Pickup.18 Approximately 500 mg of dry carbon black was placed in
a 200 mL round bottom flask. The system was sealed with a septum
and placed under an inert nitrogen atmosphere. To this, 50 mL of
anhydrous dichloromethane ͑Sigma Aldrich͒ and 5 mL of aminosi-
lane were added via a syringe. Keeping the system under inert at-
mosphere, the mixture was allowed to stir for 2 h, after which the
modified carbon was collected via suction filtration, washed several
times in dichloromethane, and finally dried for 30 min at 110°C.
Iron coordination.— Approximately 400 mg of the dried surface-
modified carbon was ground to a powder with a mortar and pestle
and placed in a conical flask along with 20 mL of 10 mM FeCl3
͑aq͒. This suspension was allowed to sit for 48 h to allow time for
the iron to coordinate to the amino functional groups. After 2 days,
the suspension was filtered over a sintered glass funnel, washed with
deionized water, and collected. This material was then dried at
110°C for 1 h and ground to a fine powder via a mortar and pestle.
The complete uptake of Fe from the solution would result in a final
product with a maximum Fe content of ca. 2 wt %.
Heat-treatment procedure.— Approximately 75 mg of catalyst ma-
terial was placed in an alumina boat and inserted into a tube furnace
͑Barnstead/Thermolyne 21100͒. Samples were heated under flowing
argon at a rate of 20°C/min up to the annealing temperature where
they were held for 2 h.
Sample nomenclature.— To clearly refer to the specific combina-
tions of carbon support, aminosilane, Fe, and heat-treatment tem-
peratures, a short hand nomenclature was developed. For example,
BP-N2-Fe-850 refers to a catalyst produced using a Black Pearls
carbon modified with N2, with Fe coordinated that has been heat-
*
Electrochemical Society Active Member.
z E-mail: brad.easton@uoit.ca
Downloaded on 2015-07-02 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑3͒ B370-B375 ͑2010͒
B371
N1
N2
N3
Table I. Weight percent iron in ORR catalysts as determined by
ICP-OES.
Fe
Sample
͑wt %͒
Pure Vulcan
0.03
2.20
1.24
0.35
0.02
1.05
0.80
0.89
V-N1-Fe-650
V-N2-Fe-650
V-N3-Fe-650
Pure Black Pearls
BP-N1-Fe-700
BP-N2-Fe-700
BP-N3-Fe-700
2
2
2
1
1
1
Figure 1. ͑Color online͒ The chemical structures of the N1, N2, and N3
aminosilanes used in this work.
=
+
͓2͔
id ik idl
where idl is the diffusion-limited current density, which is given by
idl = 0.62nFCO DO3/2
␯
␻
͓3͔
−1/6 1/2
2
2
treated at 850°C. Similarly, V-N3-Fe-700 refers to a catalyst pro-
duced using a Vulcan XC72 carbon modified with N3, with Fe co-
ordinated that has been heat-treated at 700°C.
where n is the number of electrons, F is Faraday’s constant, CO is
2
the concentration of oxygen, DO is the diffusion coefficient of oxy-
2
gen, ␯ is the kinematic viscosity, and ␻ is the rotation rate of the
−1/2
electrode. Thus, plots of 1/id vs ␻
yield an intercept of 1/ik. The
Electrochemical measurements.— Catalyst ink preparation.—
Electrochemical measurements were obtained through the creation
and analysis of electrode inks. A series of inks were prepared by
dispersing ca. 10 mg of catalyst in 500 ␮L of solution containing
Nafion, water, and isopropyl alcohol. Ink ͑4 ␮L͒ was deposited
onto the glassy carbon disk surface ͑5 mm in diameter͒ of a rotating
ring-disk electrode ͑RRDE͒ using a microsyringe. The deposited ink
was allowed to air-dry for 30 min before electrochemical measure-
ments were performed.
The electrocatalytic behavior of each deposited ink was studied
using cyclic voltammetry ͑CV͒ with an RRDE. The RRDE consisted
of an ink-coated glassy carbon disk ͑5 mm diameter͒ surrounded by
a Pt ring ͑7.5 and 8.5 mm inside and outside diameters, respec-
tively͒. Measurements were made in a one-compartment cell
equipped with a Pt wire counter electrode and a saturated calomel
electrode ͑SCE͒ reference electrode. The potential of the SCE refer-
ence electrode was calibrated vs the reversible hydrogen electrode
͑RHE͒. All the data presented in this work were corrected to the
RHE potential. Measurements were made at room temperature using
0.5 M H2SO4 ͑aq͒ as the electrolyte.
values of CO and DO employed here were 8.263 mg/L and 2.20
ϫ 10−5 cm2/2s, respect2ively.
Material characterization.— Inductively coupled plasma-optical
emission spectrometry ͑ICP-OES͒ was utilized for a quantitative
determination of iron content in synthesized catalysts. A Varian
Vista-MPX CCD simultaneous ICP-OES was set to an analytical
wavelength of 238.204 nm to look for the characteristic iron emis-
sion line. The sample preparation was similar to that reported by
Sirk et al. and involved refluxing ca. 15 mg of catalyst in 10 mL of
10% HNO3 for 24 h.22 Subsequent gravity filtration over Whatman
42 ashless filter paper and volumetric dilution yielded a solution of
sufficient purity to be analyzed by ICP-OES. Standardization was
performed with four iron͑III͒ chloride external standards ranging
from approximately 1–10 ppm. These standards contained approxi-
mately 2% nitric acid to ensure the complete dissolution and keep
both sample and standard matrices equivalent.
X-ray photoelectron spectroscopy ͑XPS͒ was performed using an
Omicron EA-125 energy analyzer and a multichannel detector
equipped with
a monochromatic Mg K␣ X-ray source ͑h␯
All the electrochemical measurements were made with the Pine
Instruments AFCBP1 bipotentiostat controlled using the PineChem
software. The electrolyte was purged with nitrogen and the disk
electrode was cycled from 0 to 1 V vs RHE at a sweep rate of 100
mV/s until a steady-state CV was obtained. After a steady state was
obtained, a final CV at 10 mV/s was obtained to determine the
nonfaradaic current. Afterward, the electrolyte was purged with oxy-
gen and subjected to the same sweep conditions as the nitrogen
purge. The electrode was then rotated at various rotation rates to
determine limiting currents. While under rotation, the ring was held
at a potential of 1.2 V vs RHE to oxidize any generated peroxide.19
The %H2O2 produced was determined from the following
equation19-21
= 1253.6 eV͒. The N 1s peaks were fitted using a mixed Gaussian–
Lorentzian-shaped profile with a Shirley background. Identical pa-
rameters were used for all the observed N 1s peaks. The peak broad-
ening was constrained to a maximum of 2 eV.
Results
Material characterization.— ICP-OES was used to quantita-
tively determine the amount of iron present in all the samples. The
results are compiled in Table I. For the Vulcan-based samples, a
trend is observed in which iron content decreases as nitrogen content
increases. Nevertheless, these values are still above the minimum
0.2 wt % Fe that is necessary for the activity.23 The iron content of
Black-Pearls-based catalysts were all quite similar, ranging between
0.8 and 1.0 wt %, showing no dependence on the nitrogen content of
the aminosilane.
XPS data.— Figure 2 compares the N 1s XPS spectra obtained for
V-N3-Fe, V-N3-Fe-650, and BP-N3-Fe-700 catalysts. The spectra
obtained from V-N3-Fe can be fitted well with two peaks centered
on 400.0 and 401.2 eV, which we attribute to the secondary and
primary amine groups, respectively, that originated from the N3-
aminosilane. The atomic ratio of the secondary to primary amines
was determined to be ca. 2.3, which is close to the expected value.
As expected, the heat-treatment induces significant structural
200IR
%H2O2 =
͓1͔
NID + IR
where IR is the current obtained from the ring, ID is the current
obtained from the disk, and N is the collection efficiency, which was
determined to be 0.1933 for our RRDE system by a ferricyanide
calibration.
Kinetic current densities ͑ik͒ for each catalyst were determined
using the Koutecky–Levich theory, which expresses the disk current
density ͑id͒ as
Downloaded on 2015-07-02 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

B372
Journal of The Electrochemical Society, 157 ͑3͒ B370-B375 ͑2010͒
(a)
(b)
(c)
Figure 2. ͑Color online͒ XPS N 1s nar-
row scan spectra obtained for ͑a͒ V-N3-
Fe, ͑b͒ V-N3-Fe-650, and ͑c͒ BP-N3-Fe-
700 catalysts.
406 404 402 400 398 396 394
406 404 402 400 398 396 394
406 404 402 400 398 396 394
Binding Energy (eV)
changes, as shown in Fig. 2b and c. The data sets for both heat-
treated catalysts were fit with three peaks that are known to originate
from pyridinic ͑398 eV͒-, pyrrolic ͑400 eV͒-, and graphene ͑402
eV͒-type nitrogen.10 The relative contribution of each are listed in
Table II. Pyrrolic N is the largest component of the V-N3-Fe-650
catalyst, almost twice that of pyridinic N. Pyrrolic N is also the
largest component of the BP-N3-Fe-650 catalyst, although it is only
slightly larger than the pyridinic content. Both heated catalysts have
a surprisingly large signal due to the graphene-type N.
V-N3-Fe catalyst system. This clearly shows that the ideal pyrolysis
temperature is 650°C for this system, with a sharp decline in activity
occurring at higher ͑and lower͒ temperatures.
To separately evaluate the electrochemical activity before and
after the addition of iron, Vulcan-based catalyst systems were char-
acterized at various stages of development in the absence of iron.
Figure 4 shows the oxygen reduction current produced by a pure
Vulcan, V-N3, and V-N3-650 ͑no iron͒. This clearly demonstrates
that the ORR activity is quite poor in the absence of iron.
Given the clear importance of both the presence of iron and
heat-treatment temperature in producing catalysts active toward the
ORR, the influence of aminosilane precursor on the ORR activity
was investigated. Figure 5 compares the ORR activity achieved with
V-N1-Fe-650, V-N2-Fe-650, and V-N3-Fe-650. The percent of H2O2
produced as a function of disk potential for the catalysts are shown
in Fig. 6. The ORR activity increases with increasing nitrogen con-
tent, with maximum activity achieved with the N3-based catalyst
system. This agrees with the trend of catalytic activity depending
upon nitrogen content as reported by others.6,10 Furthermore, perox-
ide production decreased as the nitrogen content increases. The
V-N3-Fe-650 catalyst, in particular, demonstrated a very low percent
peroxide production ͑ca. 3% at 0.3 V͒.
Electrochemical characterization: Vulcan-based samples.— A
survey of the RDE behavior for V-N3-Fe catalysts at various heat-
treatment temperatures was conducted to determine the optimum
temperature for further investigation. The onset potential for the
reduction of oxygen, VORR, is defined as the potential at which the
current for ORR is first observed24 and is the commonly used pa-
rameter to gauge the catalytic activity. VORR was determined from
the intersection of the lines tangential to the baseline and steepest
part of the ORR wave in the RDE voltammogram. Figure 3 plots
VORR as a function of the heat-treatment temperature for the
To further investigate the electrochemical kinetics, Koutecky–
Levich plots were constructed for V-N2-Fe-650 and V-N3-Fe-650
catalysts systems, which are shown in Fig. 7 ͑these plots could not
be constructed reliably for V-N1-Fe-650 as its ORR activity was too
low.͒. This allowed for the determination of the kinetic current den-
sities ͑ik͒ for each catalyst system, which are listed in Table I. The
maximum kinetic current density was achieved for the V-N3-Fe-650
catalyst, which was substantially larger than that for V-N2-Fe-650.
Figure 8 shows the oxygen-free voltammograms obtained for
V-N3-based catalysts at various stages of development. The CVs are
featureless, with no discernable peaks in the absence of Fe. How-
Table II. The relative concentration of nitrogen species present as
determined by XPS.
Primary Secondary
Sample
amine
amine
Pyridinic Pyrrolic Graphene
V-N3-Fe
V-N3-Fe-650
BP-N3-Fe-700
28.6%
—
—
71.4%
—
—
—
27.4%
33.3%
—
50.3%
37.8%
—
22.4%
28.9%
0.65
0.60
0.55
0.50
0.45
0.40
0.35
Pure Vulcan
V-N3
V-N3-650 (without Fe)
0
-1
-2
-3
-4
-5
-6
0 rpm
100 rpm
400 rpm
900 rpm
1600 rpm
0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8
0.0 0.2 0.4 0.6 0.8
Disk Potential (V vs. RHE)
500 550 600 650 700 750 800 850 900 950 1000
Temperature (°C)
Figure 4. ͑Color online͒ Disk current densities for the Vulcan-based cata-
lysts at different stages of development in the absence of iron. Measurements
were made at a sweep rate of 10 mV/s in O2-saturated 0.5 M H2SO4 at room
temperature.
Figure 3. Variation in VORR with heat-treatment temperature for V-N3-Fe-
based catalysts.
Downloaded on 2015-07-02 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑3͒ B370-B375 ͑2010͒
B373
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
N1
0.2
0.1
0.0
N3
N2
N2
N3
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
ω-½ (rpm-½)
Disk Potential (V vs. RHE)
Figure 5. ͑Color online͒ Comparison of the disk current densities obtained
for V-N1-Fe-650, V-N2-Fe-650, and V-N3-Fe-650 catalysts. Measurements
were made at a sweep rate of 10 mV/s in O2-saturated 0.5 M H2SO4 at room
temperature while the electrode was rotated at 900 rpm.
Figure 7. ͑Color online͒ Koutecky–Levich plot for V-N2-Fe-650 and V-N3-
Fe-650 catalysts at 0.3 V vs RHE. Measurements were made in O2-saturated
0.5 M H2SO4 at room temperature.
A preliminary analysis was performed on BP-N3-based catalysts
before the addition of iron to gauge the ORR activity of the as-
received and aminosilane-modified carbon blacks, which is shown
in Fig. 10. Even before heat-treatment, the as-received Black Pearls
carbon black shows a mild ORR activity. Modification with N3
leads to a small increase in activity before heat-treatment. However,
heat-treating the N3-modified Black Pearls leads to a substantial
ever, a stable and well-defined redox wave centered around 0.6 V vs
RHE was observed after heat-treatment in the presence of Fe. This
corresponds to the only V-N3-based catalyst with a high ORR ac-
tivity and may possibly be related to the formation of the active site
as waves at higher potentials are sometimes ͑but not always͒ ob-
served in Fe–N/C-based catalysts.2 Thermal treatment has led to a
significant increase in the double-layer capacitance of the sample in
both the presence and absence of Fe. This is somewhat surprising
because we would expect the surface area of carbon, and hence its
double-layer capacitance,25 to decrease upon heating in the presence
of a graphitization catalyst such as Fe.
1.0
(a)
(b)
0.5
0.0
Electrochemical characterization: Black-Pearls-based sam-
ples.— As with the Vulcan-based ORR catalysts, the temperature
dependence of the BP-N3-Fe catalysts system was first surveyed,
which is shown in Fig. 9. Unlike the Vulcan-based samples, the
Black-Pearls-based samples displayed a much broader temperature
range under which the optimal activity could be achieved. This can
be attributed to the higher surface area and more disordered nature
of Black Pearls compared to Vulcan.
-0.5
-1.0
-1.5
V
V-N3
V-N3-Fe
V-N3-650oC
V-N3-Fe-650oC
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Disk Potential (V vs. RHE)
45
40
35
30
25
Figure 8. ͑Color online͒ Steady-state CVs obtained for the Vulcan-based
catalysts at different stages of development ͑a͒ in the absence and ͑b͒ in the
presence of iron. Measurements were made at a sweep rate of 100 mV/s in
N2-saturated 0.5 M H2SO4 at room temperature.
0.70
0.68
0.66
0.64
20
N1
N2
15
N3
10
5
0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Disk Potential (V vs. RHE)
550 600 650 700 750 800 850 900 950 1000
Temperature (°C)
Figure 6. ͑Color online͒ Percent of H2O2 produced as a function disk po-
tential for V-N1-Fe-650, V-N2-Fe-650, and V-N3-Fe-650 catalysts. Measure-
ments were made at a sweep rate of 10 mV/s in O2-saturated 0.5 M H2SO4
at room temperature while the electrode was rotated at 900 rpm.
Figure 9. Variation in VORR with heat-treatment temperature for BP-N3-Fe-
based catalysts.
Downloaded on 2015-07-02 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

B374
Journal of The Electrochemical Society, 157 ͑3͒ B370-B375 ͑2010͒
Unmodified Black Pearls
BP-N3
BP-N3-700 (without Fe)
50
40
30
20
10
0
0
-1
-2
-3
-4
-5
N1
N2
N3
0 rpm
100 rpm
400 rpm
900 rpm
1600 rpm
0.0 0.2 0.4 0.6 0.8
0.0 0.2 0.4 0.6 0.8
0.0 0.2 0.4 0.6 0.8
Disk Potential (V vs. RHE)
Figure 10. Disk current densities for the Black-Pearls-based catalysts at
different stages of development in the absence of iron. Measurements were
made at a sweep rate of 10 mV/s in O2-saturated 0.5 M H2SO4 at room
temperature.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Disk Potential (V vs. RHE)
increase in the ORR activity ͑VORR = 576 mV͒, which is compa-
rable to other results in the literature for metal-free ORR catalysts.26
However, because the as-received Black Pearls do contain a mea-
surable quantity of Fe ͑0.02 wt %͒, these are not truly metal-free but
are certainly below the point of Fe saturation of active sites.23 Fur-
ther enhancements in the iron-free activity may be achievable by
varying the heat-treatment conditions.
The influence of aminosilane precursor on the ORR activity was
investigated for Black-Pearls-based catalysts with adsorbed iron,
followed by heat-treatment at the optimal conditions ͑700°C͒. Fig-
ure 11 compares the ORR activity achieved with BP-N1-Fe-700,
BP-N2-Fe-700, and BP-N3-Fe-700 catalysts. The percent of H2O2
produced as a function of disk potential for these catalysts are
shown in Fig. 12. This clearly shows that the N3-based catalysts
have higher activity than either the N1- or N2-based catalysts, al-
though it did produce slightly higher amounts of H2O2 than the
N2-based catalyst ͑ca. 9 vs 13% at 0.3 V͒.
Koutecky–Levich plots were constructed for BP-N1-Fe-700, BP-
N2-Fe-700, and BP-N3-Fe-700 catalysts systems, which are shown
in Fig. 13. Kinetic current densities were found to increase with the
amount of nitrogen in the aminosilane precursors, with the N3-based
catalyst having a substantially larger iK than the N1- and N2-based
catalysts.
Figure 14 shows the oxygen-free voltammograms obtained for
BP-N3-based catalysts at various stages of development. Much like
the Vulcan-based samples, the CVs are featureless, with no discern-
Figure 12. ͑Color online͒ H2O2 produced as a function disk potential for
BP-N1-Fe-700, BP-N2-Fe-700, and BP-N3-Fe-700 catalysts. Measurements
were made at a sweep rate of 10 mV/s in O2-saturated 0.5 M H2SO4 at room
temperature while the electrode was rotated at 900 rpm.
able peaks except for the catalyst heat-treated in the presence of Fe.
This corresponds to the most active BP-N3-based catalyst with a
high ORR activity.
Comparing the Black-Pearls- and Vulcan-based samples.— A
summary of the key electrochemical parameters for the Vulcan- and
Black-Pearls-based catalysts are complied in Table III. The results
above show that the N3-based catalysts on both Vulcan and Black
Pearls are the most active, with the Black-Pearls-based catalysts
significantly more active. The best Vulcan-based catalysts have an
onset potential that is ca. 100 mV less than that obtained with the
best Black-Pearls-based catalysts. Furthermore, the kinetic current
densities determined for the most active Black-Pearls-based was ca.
3 times that of the most active Vulcan-based catalyst. This is not
surprising given the fact that Black Pearls have a much higher sur-
face area and a higher content of disordered carbon, the latter of
which has been shown to yield more active catalysts.13 Such differ-
ences in activity between Vulcan- and Black-Pearls-based catalysts
have also been observed by others.27 In addition, the XPS data show
that the BP-N3-Fe-700 catalyst has a significantly higher ratio of
pyridinic- to pyrrolic-type N compared to V-N3-Fe-650 ͑0.88 vs
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
-3.0
N1
N2
-3.5
N3
N1
0.2
0.1
0.0
N2
-4.0
N3
-4.5
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Disk Potential (V vs. RHE)
ω-½ (rpm-½)
Figure 11. ͑Color online͒ Comparison of the disk current densities obtained
for BP-N1-Fe-700, BP-N2-Fe-700, and BP-N3-Fe-700 catalysts. Measure-
ments were made at a sweep rate of 10 mV/s in O2-saturated 0.5 M H2SO4
at room temperature while the electrode was rotated at 900 rpm.
Figure 13. ͑Color online͒ Koutecky–Levich plot for BP-N1-Fe-700, BP-N2-
Fe-700, and BP-N3-Fe-700 catalysts at 0.3 V vs RHE. Measurements were
made in O2-saturated 0.5 M H2SO4 at room temperature.
Downloaded on 2015-07-02 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑3͒ B370-B375 ͑2010͒
Conclusions
B375
4
2
(a)
(b)
A chemisorption approach to nonprecious metal fuel cell cata-
lysts was successful in producing catalysts with a decent ORR ac-
tivity. Many variables were manipulated to produce the most highly
active catalysts, including nitrogen content, carbon type, and heat-
treatment temperature. The results showed that the catalysts pre-
pared with organosilanes with higher nitrogen content produced
more highly active catalysts. Heat-treatment temperature played a
significant role with an optimum temperature being critical to ob-
taining a highly active catalyst. However, one of the most substantial
changes in activity came from the choice of carbon black. Black
Pearls, due to its significantly higher surface area, more oxidized
surface, and more disordered nature, possessed significantly more
activity toward the ORR than Vulcan even before heat-treatment
with iron.
0
-2
-4
-6
BP
BP-N3-Fe
BP-N3
BP-N3-Fe-700oC
BP-N3-700oC
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Disk Potential (V vs. RHE)
Figure 14. ͑Color online͒ Steady-state CVs obtained for the Black-Pearls-
based catalysts at different stages of development ͑a͒ in the absence and ͑b͒
presence of iron. Measurements were made at a sweep rate of 100 mV/s in
N2-saturated 0.5 M H2SO4 at room temperature.
Acknowledgments
This work was supported by the Natural Sciences and Engineer-
ing Research Council of Canada ͑NSERC͒ and UOIT. The authors
thank Grzegorz Szymanski ͑University of Guelph͒ for the assistance
with XPS measurements.
0.54͒. This is of note because high pyridinic-N contents have been
related to more active catalysts.21 Furthermore, ͑semiquantitative͒
XPS survey scans indicate that the surface nitrogen-to-carbon ratio
of BP-N3-Fe-700 was more than double that of the V-N3-Fe-650
͑1:8 vs 1:20͒. Assuming that a complete monolayer of aminosilane
was achieved for both Vulcan- and Black-Pearls-based catalysts, the
Black Pearls would adsorb ca. 6 times the amount of aminosilane
per gram of carbon due to its higher specific surface areas. Samples
with higher initial nitrogen content have been shown to retain more
nitrogen after heat-treatment, which is likely one of the main rea-
sons for the higher activity.6,28
The differences in activity may also be related to the higher iron
content in the Black-Pearls-based catalysts compared to the Vulcan-
based samples. For example, both the exchange current density and
the Fe content obtained for BP-N3-Fe-700 was ca. 3 times that
obtained for V-N3-Fe-650. Thus, the activity of each sample was
very similar on the basis of Fe loading.
University of Ontario Institute of Technology assisted in meeting the
publication costs of this article.
References
1. H. Tsuchiya and O. Kobayashi, Int. J. Hydrogen Energy, 29, 985 ͑2004͒.
2. J. P. Dodelet, in N4-Macrocyclic Metal Complexes, Chap. 3, J. Zagal, F. Bedioui,
and J. P. Dodelet, Editors, Springer, New York ͑2006͒.
3. C. W. B. Bezerra, L. Zhang, K. Lee, H. Liu, A. L. B. Marques, E. P. Marques, H.
Wang, and J. Zhang, Electrochim. Acta, 53, 4937 ͑2008͒.
4. E. B. Easton, A. Bonakdarpour, and J. R. Dahn, Electrochem. Solid-State Lett., 9,
A463 ͑2006͒.
5. R. Yang, A. Bonakdarpour, E. B. Easton, and J. R. Dahn, J. Electrochem. Soc.,
154, A275 ͑2007͒.
6. E. B. Easton, A. Bonakdarpour, R. Yang, D. A. Stevens, and J. R. Dahn, J. Elec-
trochem. Soc., 155, B547 ͑2008͒.
7. E. B. Easton, R. Yang, A. Bonakdarpour, and J. R. Dahn, Electrochem. Solid-State
Lett., 10, B6 ͑2007͒.
Although the ORR activity of these catalysts report here are re-
spectable, their activity is less than the best Fe–N/C catalysts re-
ported, such as that reported in Ref. 21. One major reason for this
difference is that our catalysts were not heat-treated in the presence
of ammonia. Heat-treatment in an ammonia atmosphere is com-
monly used to increase nitrogen content, which significantly im-
proves the ORR activity. In this work, we have intentionally chosen
not to employ ammonia so that the impact of the organosilane struc-
ture could be clearly demonstrated. Future work on this topic will
revolve around both characterizing and increasing the nitrogen con-
tent of the catalyst through both ammonia treatment and surface
modification. For example, attachment of a species that contained
pyridinic functional groups ͑like the ones believed to be in the active
site͒ via a chemisorbed linkage may prove interesting.
8. G. C.-K. Liu and J. R. Dahn, Appl. Catal., A, 347, 43 ͑2008͒.
9. R. Yang, K. Stevens, A. Bonakdarpour, and J. R. Dahn, J. Electrochem. Soc., 154,
B893 ͑2007͒.
10. D. Villers, X. Jacques-Bedard, and J. P. Dodelet, J. Electrochem. Soc., 151, A1507
͑2004͒.
11. M. Lefevre, E. Proietti, F. Jaouen, and J. P. Dodelet, Science, 324, 71 ͑2009͒.
12. F. Jaouen, M. Lefevre, J. P. Dodelet, and M. Cai, J. Phys. Chem. B, 110, 5553
͑2006͒.
13. F. Charreteur, S. Ruggeri, F. Jaouen, and J. P. Dodelet, Electrochim. Acta, 53, 6881
͑2008͒.
14. R. W. Murray, Acc. Chem. Res., 13, 135 ͑1980͒.
15. O. Lev, Z. Wu, S. Bharathi, V. Glezer, A. Modestov, J. Gun, L. Rabinovich, and S.
Sampath, Chem. Mater., 9, 2354 ͑1997͒.
16. A. Walcarius, Chem. Mater., 13, 3351 ͑2001͒.
17. E. B. Easton, Z. G. Qi, A. Kaufman, and P. G. Pickup, Electrochem. Solid-State
Lett., 4, A59 ͑2001͒.
18. E. B. Easton and P. G. Pickup, Electrochem. Solid-State Lett., 3, 359 ͑2000͒.
19. O. Antoine and R. Durand, J. Appl. Electrochem., 30, 839 ͑2000͒.
20. U. A. Paulus, T. J. Schmidt, H. A. Gasteiger, and R. J. Behm, J. Electroanal.
Chem., 495, 134 ͑2001͒.
21. C. Medard, M. Lefevre, J. P. Dodelet, F. Jaouen, and G. Lindbergh, Electrochim.
Table III. Summary of the electrochemical properties of the Fe–
N/C catalysts in this study.
Acta, 51, 3202 ͑2006͒.
22. A. H. C. Sirk, S. A. Campbell, and V. I. Birss, Electrochem. Solid-State Lett., 8,
A104 ͑2005͒.
23. F. Jaouen and J. P. Dodelet, Electrochim. Acta, 52, 5975 ͑2007͒.
24. L. Zhang, J. Zhang, D. P. Wilkinson, and H. Wang, J. Power Sources, 156, 171
͑2006͒.
ik ͑mA/cm2͒
VORR
͑mV vs RHE͒
% H2O2 at 0.3 V
Sample
at 0.3 V
͑900 rpm͒
25. D. Lozano-Castelló, D. Cazorla-Amoros, A. Linares-Solano, S. Shiraishi, H. Kuri-
hara, and A. Oya, Carbon, 41, 1765 ͑2003͒.
26. R. A. Sidik, A. B. Anderson, N. P. Subramanian, S. P. Kumaraguru, and B. N.
Popov, J. Phys. Chem. B, 110, 1787 ͑2006͒.
27. M. Bron, J. Radnik, M. Fieber-Erdmann, P. Bogdanoff, and S. Fiechter, J. Elec-
troanal. Chem., 535, 113 ͑2002͒.
V-N1-Fe-650
V-N2-Fe-650
V-N3-Fe-650
BP-N1-Fe-700
BP-N2-Fe-700
BP-N3-Fe-700
237.6
454.7
603.9
639.2
644.8
701.2
N/A
5.68
11.01
11.62
16.57
29.52
N/A
41.1
3.5
26.4
9.3
13.4
28. F. Jaouen, F. Charreteur, and J. P. Dodelet, J. Electrochem. Soc., 153, A689 ͑2006͒.
Downloaded on 2015-07-02 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).