Noble-metal-free Fe-N/C catalyst for highly efficient oxygen reduction reaction under both alkaline and acidic conditions

Article abstract of DOI:10.1021/ja504696r

In this work, we report the synthesis and assessment of a new non-precious-metal oxygen reduction reaction (ORR) catalyst from pyrolysis of an iron-coordinated complex which manifests superior activity in both alkaline and acidic media. 11,11′-bis(dipyrido[3,2-a:2′,3′-c]phenazinyl) (bidppz) was selected as a ligand for the formation of a nitrogen-rich iron-coordinated coordination polymer (Fe-bidppz) which forms a self-supporting catalyst containing high densities of nitrogen and iron doping by pyrolysis. The catalyst pyrolyzed at 800 °C (Fe-N/C-800) shows the highest ORR activity with onset and half-wave potentials of 923 and 809 mV in 0.1 M KOH, respectively, which are comparable to those of Pt/C (half-wave potential 818 mV vs RHE) at the same catalyst loading. Besides, the Fe-N/C-800 catalyst has an excellent ORR activity with onset and half-wave potentials only 38 and 59 mV less than those of the Pt/C catalyst in 0.1 M HClO₄. The optimal Fe-N/C-800 catalyst displays much greater durability and tolerance of methanol than Pt/C. We propose that the Fe-N/C-800 catalyst has a considerably high density of surface active sites because Fe-N/C-800 possesses excellent ORR activity while its specific surface area is not so high. Electrochemical measurements show that the Fe-N/C-800 catalyst in KOH and HClO₄ follows the effective four-electron-transfer pathway.

Full text of DOI:10.1021/ja504696r

Article
pubs.acs.org/JACS
Noble-Metal-Free Fe−N/C Catalyst for Highly Efficient Oxygen
Reduction Reaction under Both Alkaline and Acidic Conditions
Ling Lin, Qing Zhu, and An-Wu Xu*
Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at Microscale, University of Science and
Technology of China, Hefei 230026, China
*
S Supporting Information
ABSTRACT: In this work, we report the synthesis and
assessment of a new non-precious-metal oxygen reduction
reaction (ORR) catalyst from pyrolysis of an iron-coordinated
complex which manifests superior activity in both alkaline and
acidic media. 11,11′-bis(dipyrido[3,2-a:2′,3′-c]phenazinyl)
(
bidppz) was selected as a ligand for the formation of a
nitrogen-rich iron-coordinated coordination polymer (Fe−
bidppz) which forms a self-supporting catalyst containing high
densities of nitrogen and iron doping by pyrolysis. The catalyst
pyrolyzed at 800 °C (Fe−N/C-800) shows the highest ORR
activity with onset and half-wave potentials of 923 and 809 mV
in 0.1 M KOH, respectively, which are comparable to those of
Pt/C (half-wave potential 818 mV vs RHE) at the same
catalyst loading. Besides, the Fe−N/C-800 catalyst has an excellent ORR activity with onset and half-wave potentials only 38 and
5
9 mV less than those of the Pt/C catalyst in 0.1 M HClO . The optimal Fe−N/C-800 catalyst displays much greater durability
4
and tolerance of methanol than Pt/C. We propose that the Fe−N/C-800 catalyst has a considerably high density of surface active
sites because Fe−N/C-800 possesses excellent ORR activity while its specific surface area is not so high. Electrochemical
measurements show that the Fe−N/C-800 catalyst in KOH and HClO follows the effective four-electron-transfer pathway.
4
INTRODUCTION
a series of low-cost nitrogen precursors such as polyacrylonitrile
PAN), polypyrrole (PPy), polyaniline (PANI), ethylenedi-
amine (EN), and cyanamide (CA) were utilized in the
■
(
The oxygen reduction reaction (ORR) at the cathode is an
important process in proton exchange membrane fuel cells
PEMFCs) and is a major limiting factor of energy-conversion
efficiency in PEMFCs. Platinum and its alloys are the most
effective and commercially available ORR catalysts in PEMFCs
under the harsh acidic conditions. However, the high cost and
low abundance of the precious metals limit the wide application
of PEMFCs. Consequently, a range of alternative non-precious-
metal catalysts (NPMCs) have been developed to facilitate the
ORR on electrodes in the past few decades, including metal-
preparation of M−N/C catalysts by doping carbon supports
(
20−24
with these metal-coordinating polymers.
Although py-
rolyzed M−N/C catalysts have high activity and stability, the
exact nature of the active sites of M−N/C in acidic media is still
under debate whether the transition metal acts as the activity
center or whether the transition metal only facilitates the
1
25−28
formation of active nitrogen−carbon functional sites.
Nevertheless, plenty of studies have shown that the choice of
nitrogen precursors, transition-metal species, carbon support
morphology, and pyrolysis temperature greatly influence the
2
−4
free nitrogen-doped carbon (N-doped C),
non-precious-
transition-metal-coordinating
macrocyclic compounds, and transition-metal-coordinating
5
,6
metal oxides and carbides,
13,22
7,8
activity of the catalyst.
Hence, recent research on M−N/C
9
−11
ORR catalysts has mainly focused on the above several aspects.
In this study, we have developed an iron-coordinated
nitrogen-doped carbon catalyst (Fe−N/C) using a nitrogen-
rich ligand, 11,11′-bis(dipyrido[3,2-a:2′,3′-c]phenazinyl)
nitrogen-doped carbon catalysts (M−N/C).
Up to now,
only the M−N/C catalysts with Fe and/or Co have been found
to possess good activity and durability in an acidic
1
0,12,13
medium.
Unfortunately, M−N/C catalysts still suffer
(
bidppz) as the nitrogen-rich precursor, the structure of
from inactivation because of the protonization of N atoms and
1
4,15
which is two dipyrido[3,2-a:2′,3′-c]phenazine (dppz) moieties
linked by a C−C single bond. When coordinated with iron,
bidppz molecules form a coordination polymer (Fe−bidppz)
consisting of alternating chains of ligand and iron which hinders
anion bonding to active sites.
The use of M−N/C catalysts as NPMCs began in the 1960s
due to the discovery of cobalt phthalocyanine with ORR
16
activity in alkaline conditions, followed by the investigation of
pyrolysis of these transition-metal macrocycles to improve their
ORR catalytic activity and stability, especially in acidic
Received: May 11, 2014
Published: July 24, 2014
1
7−19
electrolytes.
In recent years, in addition to macrocycles,
©
2014 American Chemical Society
11027
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033

Journal of the American Chemical Society
Article
3
2
the agglomeration of iron species during the subsequent heat
treatment. Furthermore, the C−C single bond connecting two
dppz groups is readily rotatable, which results in numerous
configurations of bidppz and a network structure of Fe−bidppz.
Unlike the liner polymers previously used for M−N/C
catalysts, this network complex can be utilized as a self-
supporting catalyst excluding the use of a carbon sup-
aromatic ring structures in bidppz are highly thermally stable,
promoting graphitization during carbonization at a high
3
3,34
temperature,
which is favorable for the high activity of
the catalyst. The Fe−bidppz precursor was pyrolyzed in argon
at 800 °C for 1.5 h, and the product was ultrasonically leached
in 6 M hydrochloric acid (HCl) solution to obtain the Fe−N/
C-800 self-supporting catalysts (Figure 1c). The aim of the
leaching process is to remove the inactive iron particles and
other iron species produced from agglomeration during
pyrolysis.
Previous studies showed that the pyrolysis temperature has
an important influence on the catalytic activity of Fe−N/C
catalysts, since calcination at a low temperature may facilitate
9
,20−22,29
port.
The nearly 20 wt % concentration of highly
thermally stable pyridinic nitrogen in bidppz allows for a high
degree of Fe−N coordination at high pyrolysis temperatures.
We demonstrate that our catalyst (Fe−N/C-800) prepared by
heating the Fe−bidppz precursor to 800 °C exhibits superior
ORR catalytic activity, long-term stability, and higher methanol
tolerance than the 20 wt % Pt/C (hereafter “Pt/C”) catalyst in
both alkaline and acidic media at a catalyst loading of 0.1 mg
9
the incorporation of nitrogen and iron into the carbon matrix
while a high pyrolysis temperature can increase the electrical
conductivity of the carbon material. To find an optimum
pyrolysis temperature for our self-supporting Fe−N/C catalyst,
the Fe−bidppz precursor was pyrolyzed at various temperatures
−2
cm . Overall, this self-supporting Fe−N/C-800 catalyst is a
potential candidate as a highly efficient, stable, and low-cost
ORR catalyst.
(
700−900 °C). The X-ray diffraction (XRD) patterns shown in
RESULTS AND DISCUSSION
Figure 1 depicts the chemical structure of bidppz, the synthetic
process of the Fe−N/C-800 catalyst, and the corresponding
Figure S1a (Supporting Information) display the structures of
the catalysts prepared at different temperatures. It is clearly
seen that the peak around 25° becomes stronger and narrower
with increasing temperature, indicating that a higher pyrolysis
temperature may lead to more ordered graphitic carbon. Figure
S1b compares the Raman spectra of Fe−N/C catalysts
synthesized at different pyrolysis temperatures. The ratios of
the D band to G band integrated intensities (I /I ) are 3.38
■
D
G
(
1
700 °C), 3.13 (750 °C), 2.87 (800 °C), 2.44 (850 °C), and
.78 (900 °C). As the highly ordered graphite shows a very
weak D-band, the decreasing value of I /IG with increasing
D
temperature proves the formation of more and more ordered
35
graphitic structure domains during heat treatment. A more
ordered structure with graphitic structure domains is benificial
for increasing the electrical conductivity of the derived carbon.
The elemental compositions of Fe−N/C samples were
determined by X-ray photoelectron spectroscopy (XPS)
(
Table S1, Supporting Information). The content of carbon
increases with the pyrolysis temperature, while the contents of
nitrogen and iron are reduced due to high temperatures. The
content of iron decreases sharply from 0.51 to 0.28 atom %
when the pyrolysis temperature increases from 750 to 800 °C,
indicating the higher the pyrolysis temperature, the greater the
degree of agglomeration. Electrochemical measurements
Figure 1. Schematic illustration of the synthetic process of the Fe−N/
C-800 catalyst: (a) chemical structure and photograph of bidppz, (b)
chemical structure and photograph of the corresponding Fe−bidppz
precursor, (c) photograph of the obtained Fe−N/C-800 catalyst.
(
Figure S2, Supporting Information) confirm that the catalyst
with the best ORR activity is the one obtained by pyrolyzing
Fe−bidppz at 800 °C (Fe−N/C-800) in both alkaline and
acidic conditions, implying that the catalyst obtained at 800 °C
has optimal comprehensive conditions considering the effective
factors, including the content of nitrogen and iron and the
structure and electrical conductivity of carbon.
photographs. The rotatable C−C single bond linking two dppz
moieties in the bidppz molecule (Figure 1a) is marked in red;
the rotation of the C−C single bond produces different
configurations of bidppz. A simple solvothermal in situ metal/
ligand reaction was used to synthesize the Fe−bidppz
precursor. The solvothermal technique is able to not only
overcome the difficulties resulting from different solubilities of
organic and inorganic precursors, but also produce products
that are inaccessible or not easily obtainable by conventional
The scanning electron microscopy (SEM) image of the Fe−
bidppz precursor (Figure 2a) shows that the precursor has an
elongated sheetlike morphology and the precursor sheets broke
into small pieces during pyrolysis at 800 °C (Figure 2b,c). The
Brunauer−Emmet−Teller (BET) surface areas of Fe−bidppz
3
0,31
methods.
The structure of the obtained Fe−bidppz
2
−1
complex is shown in Figure 1b, in which two different planar
configurations of bidppz are displayed. Nevertheless, other
nonplanar configurations may also appear in the Fe−bidppz
complex, which leads to a network coordination polymer. This
coordination polymer structure is effective at separating one
iron site from another and preventing the iron sites from
agglomeration during heat treatment, facilitating the formation
of well-dispersed active sites. In addition, the quintuple
and Fe−N/C-800 are 77 and 56 m g , respectively (Figure
S3a,b, Supporting Information). This decrease of surface area
may be attributed to the collapsing of pores in the Fe−bidppz
porous network. The Barrett−Joyner−Halenda (BJH) analysis
indicates that the Fe−N/C-800 catalyst contains mesopores
∼3.9 and 6.6 nm in diameter (Figure S3c). It is most likely that
these mesopores are helpful for the ORR activity of Fe−N/C-
800 catalyst. The transmission electron microscopy (TEM)
1
1028
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033

Journal of the American Chemical Society
Article
Figure 2. SEM image of the Fe−bidppz precursor (a). SEM image (b) and TEM image (c) of the Fe−N/C-800 catalyst. N 1s XPS spectrum (d) and
Fe 2p XPS spectrum (e) of the Fe−N/C-800 catalyst.
Figure 3. Cyclic voltammograms (a) and RDE voltammograms (b) of Pt/C and Fe−N/C-800 in O -saturated 0.1 M KOH. Cyclic voltammograms
2
(
c) and RDE voltammograms (d) of Pt/C and Fe−N/C-800 in O -saturated 0.1 M HClO . Endurance test of the Fe−N/C-800 catalyst for 10 000
2
4
−2 −1
cycles in O -saturated 0.1 M KOH (e) and 0.1 M HClO (f). For all tests, the catalyst loading is 0.1 mg cm and the scan rate is 10 mV s .
2
4
1
1029
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033

Journal of the American Chemical Society
Article
image (Figure 2c) and XRD pattern of Fe−N/C-800 (Figure
S1a, Supporting Information) reveal that no iron particles can
be found in the Fe−N/C-800 sample, indicating that the acid
leaching process removes most of the inactive iron particles
since the XRD pattern and TEM image of the pyrolyzed
product before acid leaching show the existence of iron
particles (Figure S4, Supporting Information). Electrochemical
measurements show the leaching process is significant to the
ORR activity of the catalyst. From Figure S5 (Supporting
Information), it can be seen that the catalytic activity of Fe−N/
C-800 is greatly enhanced after acid leaching because the bulk
iron particles formed during pyrolysis are inactive and block the
surface active sites.
importantly, the ORR peak current in 0.1 M HClO for Fe−N/
4
C-800 is much larger than that in 0.1 M KOH, reaching 124.2
μA, while the peak current of Pt/C in an acid is only 65.0 μA.
The higher capacitance currents of Fe−N/C-800 in CV suggest
that Fe−N/C-800 promotes O transport within the catalyst
2
44
layer due to the mesopores in the Fe−N/C-800 catalyst.
Similar performances of the ORR activities of Pt/C and Fe−N/
C-800 catalysts were observed by rotating disk electrode
−1
(RDE) experiments at 1600 rpm and 10 mV s . Figure 3b
shows that the onset potential for the Fe−N/C-800 catalyst in
alkaline conditions (923 mV) is ∼33 mV lower than that of Pt/
C (956 mV) and the half-wave potentials (E ) for Fe−N/C-
1/2
800 and Pt/C are 809 and 818 mV, respectively. The
polarization curves performed in acidic conditions are displayed
in Figure 3d. It can be seen that the value of the onset potential
and the half-wave potential on an Fe−N/C-800 catalyst
electrode are only 38 mV and 59 mV more negative than
XPS spectra reveal the elemental information on the Fe−N/
C-800 catalyst. From the survey scan of Fe−N/C-800 shown in
Figure S6 (Supporting Information), it can be seen that the
Fe−N/C-800 catalyst contains four kinds of elements, carbon
(
(
87.25%), nitrogen (9.62%), oxygen (2.85%), and iron
0.28%). The N 1s spectrum (Figure 2d) can be divided into
those for Pt/C. The limiting current densities (j ) of Fe−N/C-
L
800 are 6.06 and 6.09 mA cm− in KOH and HClO at a
2
4
several peaks at 398.4, 399.1, 399.9, 401.0, and 402.7 eV, which
can be assigned to pyridinic N (24.73%), pyrrolic N (24.23%),
rotating speed of 1600 rpm, respectively, which are close to the
theoretical value determined by the Koutecky−Levich equation.
These electrochemical values are close to those of the
previously reported active ORR catalysts; thus, our Fe−N/C-
800 catalyst has a high ORR activity not only in an alkaline
electrolyte but also in an acidic electrolyte. The modest specific
+
−
graphitic N (30.27%), and quarternary N −O (20.77%) on
3
6,37
the basis of the respective binding energies.
The graphitic
38
N was reported to play a crucial role in oxygen reduction;
besides, pyridinic N and pyrrolic N can serve as metal-
coordination sites due to their lone-pair electrons. These three
kinds of ORR active nitrogen are of high content in our Fe−N/
C-800 elecrocatalyst, which leads to a high catalytic activity.
Figure 2e shows the Fe 2p spectrum. According to previous
2
−1
surface area (56 m g ) and the high ORR activity of Fe−N/
C-800 comparable to those of other kinds of M−N/C catalysts
with specific surface areas as high as hundreds of square meters
per gram imply that our Fe−N/C-800 catalyst may have a
much higher density of surface active sites than previously
3
9,40
reports,
the binding energies of the 2p3/2 orbitals of Fe and Fe
species, respectively. For the 2p1/2 band, the peak at 723.2 eV is
the peaks at 710.5 and 714.4 eV can be assigned to
2
+
3+
10,22,41
reported M−N/C catalysts.
Because of iron doping in the Fe−N/C-800 catalyst, there is
a question of whether iron species serve as the active site.
Previous works have proposed two different hypotheses to
2
+
3+
attributed to the binding energy of Fe , and the peak for Fe
is observed at 725.4 eV. The peak at 719.6 eV is a satellite peak.
The XPS survey scan suggests a trace amount of S as well,
describe the active sites in the M−N/C ORR catalyst. One is
4
1
44,45
which is likely to be present in the form of S−C and SO .
A
the M−N species,
and the other is the N atoms doping the
x
x
46
control experiment shows that the trace amount of S has no
obvious contribution to the high activity of our catalyst (details
in the Supporting Information, Figure S7).
carbon matrix. The latter is the same as the nonmetal N/C
4,47
catalyst. The ORR activity of Fe−N/C-800 is most probably
considered to follow the former hypothesis that the Fe−N may
x
To examine the electrocatalytic activity of the Fe−N/C-800
catalyst, cyclic voltammetry (CV) measurements were carried
out for Pt/C and Fe−N/C-800 catalysts in O - or N -saturated
act as the major active sites. A control experiment by pyrolyzing
the pure bidppz was conducted to confirm the viewpoint that
Fe−N plays a key role in the high catalytic activity of Fe−N/
2
2
x
0
.1 M KOH (Figure 3a) and 0.1 M HClO (Figure 3c) at the
C-800. The pyrolyzed pure bidppz catalyst has a N content
(10.15%) higher than that of the Fe−N/C-800 catalyst (Table
S2, Supporting Information), while the ORR activities of the
pyrolyzed pure bidppz catalyst are much poorer than those of
the Fe−N/C-800 catalyst in both electrolytes, especially in
acidic conditions (Figure S7c,d, Supporting Information).
4
−2
same catalyst loading of 0.1 mg cm , and the Pt loading for the
Pt/C catalyst was 20 μg cm . In an alkaline electrolyte, Fe−N/
C-800 shows a quasi-rectangular double-layer capacity current
−2
4
2
in N2, and a well-defined ORR peak in O with a peak
2
potential of 851 mV is observed, which is a little negative
compared to that of Pt/C (868 mV). However, the ORR peak
current of Fe−N/C-800 is 90.2 μA, much higher than that of
Pt/C (60.8 μA). In an acidic electrolyte, Fe−N/C-800 displays
a pair of symmetric peaks at about 750 mV, which indicates a
redox reaction occurs between Fe(II) and Fe(III) during this
process. A small oxidation peak also appears at 780 mV in
Although the Fe−N active sites need to be further confirmed
x
by other characterization methods, the obvious improvement of
ORR activity by introduction of iron suggests that the Fe−Nx
species make a major contribution to the activity of Fe−N/C-
2
+
3+
800. Particularly, according to the peak area of Fe and Fe
2
+
from the Fe 2p XPS spectrum, it is found that the use of Fe as
the iron precursor results in a quite high percentage of Fe in
our Fe−N/C-800 catalyst. Since Fe −N can serve as the
active sites according to a previous work, abundant Fe
2
+
saturated O for Fe−N/C-800 in acidic solution, indicating iron
2
2+
doping in the carbon material may play the role of a mediator
in the electron-transfer processes during the oxygen reduction
4
44
2+
43
reaction occurring in acidic conditions. A previous study
species may play a major role in superior ORR activity of the
Fe−N/C-800 catalyst.
Although some reported NPMCs exhibit relatively high ORR
activities, durability is of significance for practical applications.
proposed an oxygen reduction mechanism for CoFeN /C in an
x
acidic electrolyte in which transition-metal ions can act as a
mediator in the electron-transfer process. The ORR peak
potential for Fe−N/C-800 in an acid is 687 mV, which is
slightly more positive than that for Pt/C (672 mV). More
Cycling durability tests in O -saturated electrolytes were carried
2
out to test the stability of Fe−N/C-800 for 10 000 cycles. The
1
1030
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033

Journal of the American Chemical Society
Article
Figure 4. RRDE voltammograms (a) and peroxide yield (black) with regard to the total oxygen reduction products and the electron-transfer number
n) (red) (b) of Fe−N/C-800 in O -saturated 0.1 M KOH. RRDE voltammograms (c) and peroxide yield (black) with regard to the total oxygen
(
2
−2
reduction products and n (red) (d) of Fe−N/C-800 in O -saturated 0.1 M HClO . For all tests, the catalyst loading is 0.1 mg cm , the rotating
speed is 900 rpm, and the scan rate is 10 mV s .
2
4
−1
polarization curves recorded after 10 000 cycles show a negative
shift of E1/2 of 30.4 and 28.5 mV in alkaline and acidic
electrolytes (Figure 3e,f), respectively, indicating its durability is
superior to that of Pt/C (Figure S8, Supporting Information).
The obvious inactivation behavior of Pt/C (large negative shifts
of onset potential and E , but no obvious decrease in j ) is
process that produces H O , in addition, H O formed during
2
2
2
2
the 2e process is poisonous to catalysts. ORR polarization
curves of Fe−N/C-800 at rotation rates from 400 to 1600 rpm
are shown in Figure S10a,c (Supporting Information). On the
1
0
−1
−1/2
basis of the Koutecky−Levich equation, four J versus ω
plots were obtained (Figure S10b,d). Good linear relationships
1
/2
L
between J− and ω
1
−1/2
are observed, indicating first-order
mainly ascribed to the dissolution of Pt nanoparticles at high
4
8
potentials. Different from Pt/C, the simultaneous decrease in
reaction kinetics with respect to the O concentration. We
2
E1/2 and j of Fe−N/C-800 is attributed to the oxidation of the
calculated the electron-transfer number (n) from the slope of
each line. For an alkaline electrolyte, the average value is 4.15,
while the calculated value is 4.09 for an acidic electrolyte.
Therefore, the oxygen reduction reaction on the Fe−N/C-800
catalyst in both electrolytes follows the four-electron-transfer
pathway, corresponding to the complete reduction of oxygen
into water. To further validate this conclusion, rotating ring−
disk electrode (RRDE) measurements were performed to
determine the four-electron selectivity of the Fe−N/C-800
catalyst. Parts a and c of Figure 4 show the disk and ring
currents measured at 900 rpm in 0.1 M KOH and 0.1 M
HClO , respectively. The onset potential for H O generation
L
carbon support as well as the oxidation of active sites, as active
site oxidation will only lead to a negative shift of the onset
43
potential and E1/2
. The limiting currents only decrease by 5%
and 6.3% after 10 000 cycles in 0.1 M KOH and HClO4,
respectively, implying that the self-supporting carbon in our
catalyst is a good candidate for the carbon support. It is worth
noting that our Fe−N/C-800 catalyst shows superior ORR
activity and high stability in an acidic electrolyte, proving that
the Fe−N sites in the Fe−N/C-800 catalyst are quite stable in
x
an acidic environment. Notably, the Fe−N/C-800 catalyst
shows better tolerance toward methanol than Pt/C. As shown
in Figure S9 (Supporting Information), a small decrease of the
ORR peak is observed for Fe−N/C-800, indicating little
influence of methanol on the ORR activity of our Fe−N/C-800
catalyst. In contrast, the ORR peaks of Pt/C disappear and
methanol oxidation peaks appear because the electro-oxidation
of methanol hinders the ORR process. These results indicate
that the Fe−N/C-800 catalyst is superior to Pt/C as an ORR
catalyst for direct methanol fuel cells.
4
2
2
in 0.1 M KOH is lower than the onset potential of ORR, which
49
is similar to the behavior of Pt catalysts. The behavior of the
ring current curve in 0.1 M HClO is different from that in 0.1
4
M KOH, and the ring current in acidic conditions is about 10
times smaller than that in alkaline conditions under low
potentials. The hydrogen peroxide yield (%) and the electron-
transfer number (n) curves (Figure 4b,d) can be directly
calculated from the disk currents (i ) and the ring currents (i ).
d
r
ORR polarization curves of the Fe−N/C-800 catalyst at
different rotation rates from 400 to 1600 rpm were measured to
determine whether the ORR of Fe−N/C-800 occurs by four-
As shown in Figure 4b, the peroxide species yield in alkaline
solution is less than 4% in the measured potential range from
1.1 to 0.1 V and the average n is ∼3.96. For an acidic
electrolyte, Figure 4d shows that the value of n is almost
constant at 4.00 from 0 to 0.5 V and gradually decreases to
electron (4e) reduction of oxygen to produce H O. Generally,
2
the 4e process is believed to be more efficient than the 2e
1
1031
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033

Journal of the American Chemical Society
3.92 at potentials higher than 0.5 V. The average n obtained
Article
∼
ACKNOWLEDGMENTS
■
from the curve is ∼3.99, which is slightly larger than that in
KOH. The electron-transfer number calculated from RRDE
measurements is in agreement with the electron-transfer
number based on the Koutecky−Levich plots, signaling mainly
the 4e pathway for the Fe−N/C-800-catalyzed ORR process.
Previous studies have demonstrated that the higher over-
potential for ORR in acidic electrolytes is due to the hydrogen
peroxide intermediate (H O ) in acidic media being less stable
Support from the National Basic Research Program of China
Grants 2011CB933700 and 2010CB934700) and the National
Natural Science Foundation of China (Grant 21271165) is
gratefully acknowledged.
(
REFERENCES
■
(
(
1) Wu, J. B.; Yang, H. Acc. Chem. Res. 2013, 46, 1848−1857.
2) Chen, T.; Cai, Z.; Yang, Z.; Li, L.; Sun, X.; Huang, T.; Yu, A.; Kia,
H. G.; Peng, H. Adv. Mater. 2011, 23, 4620−4625.
(3) Zheng, Y.; Jiao, Y.; Chen, J.; Liu, J.; Liang, J.; Du, A.; Zhang, W.;
Zhu, Z.; Smith, S. C.; Jaroniec, M.; Lu, G. Q.; Qiao, S. Z. J. Am. Chem.
Soc. 2011, 133, 20116−20119.
2
2
2+
−
on Fe −N/C active sites than the Lewis basic form (HO ) in
2
2+
alkaline media. The destabilization of H O on Fe −N/C
2
2
active sites in an acid results in the requirement of a secondary
active site to reduce or disproportionate H O . The high
2
2
(
4) Deng, D.; Pan, X.; Yu, L.; Cui, Y.; Jiang, Y.; Qi, J.; Li, W.-X.; Fu,
Q.; Ma, X.; Xue, Q.; Sun, G.; Bao, X. Chem. Mater. 2011, 23, 1188−
193.
5) Su, H.-Y.; Gorlin, Y.; Man, I. C.; Calle-Vallejo, F.; Nørskov, J. K.;
Jaramillo, T. F.; Rossmeisl, J. R. Phys. Chem. Chem. Phys. 2012, 14,
4010−14022.
6) Esposito, D. V.; Chen, J. G. Energy Environ. Sci. 2011, 4, 3900−
912.
7) Carver, C. T.; Matson, B. D.; Mayer, J. M. J. Am. Chem. Soc. 2012,
34, 5444−5447.
(8) Samanta, S.; Sengupta, K.; Mittra, K.; Bandyopadhyay, S.; Dey, A.
Chem. Commun. 2012, 48, 7631−7633.
9) Zhao, Y.; Watanabe, K.; Hashimoto, K. J. Am. Chem. Soc. 2012,
34, 19528−19531.
10) Peng, H.; Mo, Z.; Liao, S.; Liang, H.; Yang, L.; Luo, F.; Song,
H.; Zhong, Y.; Zhang, B. Sci. Rep. 2013, 3, 1765.
11) Chung, H. T.; Won, J. H.; Zelenay, P. Nat. Commun. 2013, 4,
surface active site density of the Fe−N/C-800 catalyst will
facilitate this process because of the high probability for H O
2
2
1
(
to be adsorbed by a secondary active site immediately, leading
to an onset potential only 38 mV negative of that for Pt/C and
an electron-transfer number as high as 3.99 in acidic media.
1
(
3
(
CONCLUSIONS
■
In this work, we have developed a highly efficient self-
supporting M−N/C ORR electrocatalyst in both alkaline and
acidic conditions by pyrolysis of a complex precursor, Fe−
bidppz, in argon. The pyrolysis temperature largely affects the
ORR activity, and the optimal catalyst pyrolyzed at 800 °C
exhibits the highest ORR activity comparable to that of
commercial Pt/C. Although the specific surface area of Fe−N/
C-800 is not high, the activity of the Fe−N/C-800 catalyst is
comparable to that of the Pt/C catalyst in both alkaline and
acidic media, indicating a high active site density of our Fe−N/
C-800 catalyst. We propose that the major active sites in the
1
(
1
(
(
1
922.
(12) Chen, Z.; Higgins, D.; Yu, A.; Zhang, L.; Zhang, J. Energy
Environ. Sci. 2011, 4, 3167−3192.
(13) Jaouen, F.; Proietti, E.; Lefevre, M.; Chenitz, R.; Dodelet, J. P.;
Wu, G.; Chung, H. T.; Johnston, C. M.; Zelenay, P. Energy Environ. Sci.
011, 4, 114−130.
14) Oberst, J. L.; Thorum, M. S.; Gewirth, A. A. J. Phys. Chem. C
012, 116, 25257−25261.
15) Herranz, J.; Jaouen, F.; Lefevre, M.; Kramm, U. I.; Proietti, E.;
Dodelet, J.-P.; Bogdanoff, P.; Fiechter, S.; Abs-Wurmbach, I.; Bertrand,
P.; Arruda, T.; Mukerjee, S. J. Phys. Chem. C 2011, 115, 16087−16097.
(16) Jasinski, R. Nature 1964, 201, 1212−1213.
Fe−N/C-800 catalyst are the N-binding iron species (Fe−N ),
x
2
(
2
(
as the introduction of iron into the catalyst greatly enhances the
ORR activity. Moreover, the Fe−N/C-800 catalyst shows
better stability and tolerance of methanol than the commercial
Pt/C catalyst. There is a very small decrease of the limiting
current after 10 000 cycles, demonstrating that the self-
supporting carbon in the Fe−N/C-800 catalyst is a good
candidate for the carbon support. More notably, the oxygen
reduction reaction on the Fe−N/C-800 catalyst in alkaline and
acidic media follows the efficient four-electron-transfer path-
way, indicating the complete reduction of oxygen into water.
Taken together, our developed self-supporting Fe−N/C-800
catalyst could find potential applications in PEMFCs with
superior ORR performance and high stability.
(
(
1
(
17) Wiesener, K. Electrochim. Acta 1986, 31, 1073−1078.
18) van Veen, J. A. R.; Colijn, H. A.; van Baar, J. F. Electrochim. Acta
988, 33, 801−804.
19) Bezerra, C. W. B.; Zhang, L.; Lee, K. C.; Liu, H. S.; Marques, A.
L. B.; Marques, E. P.; Wang, H. J.; Zhang, J. J. Electrochim. Acta 2008,
3, 4937−4951.
20) Gupta, S.; Tryk, D.; Bae, I.; Aldred, W.; Yeager, E. J. Appl.
5
(
Electrochem. 1989, 19, 19−27.
21) Bashyam, R.; Zelenay, P. Nature 2006, 443, 63−66.
(
(
22) Wu, G.; More, K. L.; Johnston, C. M.; Zelenay, P. Science 2011,
ASSOCIATED CONTENT
■
332, 443−447.
*
S
Supporting Information
(23) Choi, J.; Hsu, R.; Chen, Z. J. Phys. Chem. C 2010, 114, 8048−
8
(
2
(
053.
Experimental details, XRD patterns, Raman spectra, elemental
compositions from XPS, nitrogen adsorption/desorption
isotherms, and supporting electrochemical analysis. This
material is available free of charge via the Internet at http://
pubs.acs.org.
24) Chung, H. T.; Johnston, C. M.; Zelenay, P. ECS Trans. 2009,
4, 485−492.
25) Lefevre, M.; Proietti, E.; Jaouen, F.; Dodelet, J. P. Science 2009,
24, 71−74.
(26) Wu, G.; Chen, Z.; Artyushkova, K.; Garzon, F. H.; Zelenay, P.
ECS Trans. 2008, 16, 159−170.
27) Ferrandon, M.; Kropf, A. J.; Myers, D. J.; Artyushkova, K.;
Kramm, U.; Bogdanoff, P.; Wu, G.; Johnston, C. M.; Zelenay, P. J.
Phys. Chem. C 2012, 116, 16001−16013.
28) Wu, G.; Johnston, C. M.; Mack, N. H.; Artyushkova, K.;
3
(
AUTHOR INFORMATION
■
Corresponding Author
(
anwuxu@ustc.edu.cn
Ferrandon, M.; Nelson, M.; Lezama-Pacheco, J. S.; Conradson, S. D.;
More, K. L.; Myers, D. J.; Zelenay, P. J. Mater. Chem. 2011, 21,
11392−11405.
Notes
The authors declare no competing financial interest.
1
1032
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033

Journal of the American Chemical Society
Article
(
29) Yuan, X.; Zeng, X.; Zhang, H.-J.; Ma, Z.-F.; Wang, C.-Y. J. Am.
Chem. Soc. 2010, 132, 1754−1755.
(
(
30) Chen, X.-M.; Tong, M.-L. Acc. Chem. Res. 2007, 40, 162−170.
31) Zhu, Q.; Peng, Y.; Lin, L.; Fan, C.-M.; Gao, G.-Q.; Wang, R.-X.;
Xu, A.-W. J. Mater. Chem. A 2014, 2, 4429−4437.
32) Johns, I. B.; McElhill, E. A.; Smith, J. O. Ind. Eng. Chem. Prod.
Res. Dev. 1962, 1, 2−6.
33) Yang, X.; Li, Z.; Zhi, J.; Ma, J.; Hu, A. Langmuir 2010, 26,
1244−11248.
34) Kim, T.-W.; Park, I.-S.; Ryoo, R. Angew. Chem., Int. Ed. 2003, 42,
375−4379.
35) Sadezky, A.; Muckenhuber, H.; Grothe, H.; Niessner, R.; Po
U. Carbon 2005, 43, 1731−1742.
36) Zhang, Y.; Ge, J.; Wang, L.; Wang, D.; Ding, F.; Tao, X.; Chen,
W. Sci. Rep. 2013, 3, 2771.
37) Shin, D.; Jeong, B.; Mun, B. S.; Jeon, H.; Shin, H.-J.; Baik, J.;
Lee, J. J. Phys. Chem. C 2013, 117, 11619−11624.
38) Liu, R.; Wu, D.; Feng, X.; Mullen, K. Angew. Chem., Int. Ed.
010, 49, 2565−2569.
39) Gupta, A.; Kumar, A.; Waghmare, U. V.; Hegde, M. S. Chem.
Mater. 2009, 21, 4880−4891.
40) Kong, H.; Song, J.; Jang, J. Chem. Commun. 2010, 46, 6735−
737.
41) Chang, Y. Q.; Hong, F.; He, C. X.; Zhang, Q. L.; Liu, J. H. Adv.
Mater. 2013, 25, 4794−4799.
42) Yang, S. B.; Zhi, L. J.; Tang, K.; Feng, X. L.; Maier, J.; Mu
(
(
1
(
4
(
̈
schl,
(
(
(
̈
2
(
(
6
(
(
̈
llen, K.
Adv. Funct. Mater. 2012, 22, 3634−3640.
(
(
43) Jiang, R.; Chu, D. J. Power Sources 2014, 245, 352−361.
44) Ramaswamy, N.; Tylus, U.; Jia, Q.; Mukerjee, S. J. Am. Chem.
Soc. 2010, 135, 15443−15449.
45) Lefevre, M.; Dodelet, J. P.; Bertrand, P. J. Phys. Chem. B 2002,
06, 8705−8713.
46) Nabae, Y.; Moriya, S.; Matsubayashi, K.; Lyth, S. M.; Malon, M.;
(
1
(
Wu, L.; Islam, N. M.; Koshigoe, Y.; Kuroki, S.; Kakimoto, M.-a.;
Miyata, S.; Ozaki, J.-i. Carbon 2010, 48, 2613−2624.
(
47) Silva, R.; Voiry, D.; Chhowalla, M.; Asefa, T. J. Am. Chem. Soc.
2
(
013, 135, 7823−7826.
48) Borup, R.; Meyers, J.; Pivovar, B.; Kim, Y. S.; Mukundan, R.;
Garland, N.; Myers, D.; Wilson, M.; Garzon, F.; Wood, D.; Zelenay,
P.; More, K.; Stroh, K.; Zawodzinski, T.; Boncella, J.; McGrath, J. E.;
Inaba, M.; Miyatake, K.; Hori, M.; Ota, K.; Ogumi, Z.; Miyata, S.;
Nishikata, A.; Siroma, Z.; Uchimoto, Y.; Yasuda, K.; Kimijima, K.;
Iwashita, N. Chem. Rev. 2007, 107, 3904−3951.
(
49) Schmidt, T. J.; Stamenkovic, V.; Arenz, M.; Markovic, N. M.;
Ross, P. N., Jr. Electrochim. Acta 2002, 47, 3765−3776.
1
1033
dx.doi.org/10.1021/ja504696r | J. Am. Chem. Soc. 2014, 136, 11027−11033