Nanoscale

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

PAPER

Cite this: DOI: 10.1039/c5nr04929f

View Journal

Nanocrystalline FeFe2O3 particle-deposited

N-doped graphene as an activity-modulated
Pt-free electrocatalyst for oxygen reduction
reaction
Vishal M. Dhavale,a,b Santosh K. Singh,a,b Ayasha Nadeema,a,b Sachin S. Gaikwada
and Sreekumar Kurungot*a,b
The size-controlled growth of nanocrystalline FeFe2O3 particles (23 nm) and their concomitant dispersion on N-doped graphene (FeFe2O3/NGr) could be attained when the mutually assisted redox reaction between NGr and Fe3+ ions could be controlled within the aqueous droplets of a water-in-oil
emulsion. The synergistic interaction existing between FeFe2O3 and NGr helped the system to narrow
down the overpotential for the oxygen reduction reaction (ORR) by bringing a signicant positive shift to
the reduction onset potential, which is just 15 mV higher than its Pt-counterpart. In addition, the halfwave potential (E1/2) of FeFe2O3/NGr is found to be improved by a considerable amount of 135 mV in
comparison to the system formed by dispersing FeFe2O3 nanoparticles on reduced graphene oxide
(FeFe2O3/RGO), which indicates the presence of a higher number of active sites in FeFe2O3/NGr.
Despite this, the ORR kinetics of FeFe2O3/NGr are found to be shifted signicantly to the preferred
4-electron-transfer pathway compared to NGr and FeFe2O3/RGO. Consequently, the H2O2% was found
to be reduced by 78.3% for FeFe2O3/NGr (13.0%) in comparison to FeFe2O3/RGO (51.2%) and NGr
(41.0%) at 0.30 V (vs. Hg/HgO). This dierence in the yield of H2O2 formed between the systems along
with the improvements observed in terms of the oxygen reduction onset and E1/2 in the case of
FeFe2O3/NGr reveals the activity modulation achieved for the latter is due to the coexistence of factors
such as the presence of the mixed valancies of iron nanoparticles, small size and homogeneous distribution of FeFe2O3 nanoparticles and the electronic modications induced by the doped nitrogen in
NGr. A controlled interplay of these factors looks like worked favorably in the case of FeFe2O3/NGr. As a
realistic system level validation, FeFe2O3/NGr was employed as the cathode electrode of a single cell in a

Received 23rd July 2015,

Accepted 26th October 2015

solid alkaline electrolyte membrane fuel cell (AEMFC). The system could display an open circuit voltage

DOI: 10.1039/c5nr04929f

(OCV) of 0.73 V and maximum power and current densities of 54.40 mW cm2 and 200 mA cm2,
respectively, which are comparable to the performance characteristics of a similar system derived by

www.rsc.org/nanoscale

using 40 wt% Pt/C as the cathode electrode.

1.

Introduction

Heteroatom-doped carbon morphologies and their combinations with metals have been gaining world-wide attention in

a
Physical and Materials Chemistry Division, CSIR-National Chemical Laboratory,
Pune 411 008, India. E-mail: k.sreekumar@ncl.res.in; Fax: +91 20-25902636;
Tel: +91 20-25902566
b
Academy of Scientific and Innovative Research (AcSIR), Anusandhan Bhawan,
2 Rafi Marg, New Delhi 110001, India
Electronic supplementary information (ESI) available: Experimental, deconvulated XPS of C 1s, and O 1s of FeFe2O3/RGO, FeFe2O3/NGr, RGO and NGr.
Deconvoluted N 1s of FeFe2O3/NGr and NGr, formulae, CV, LSV at dierent
rpm, and KL plots. See DOI: 10.1039/c5nr04929f
These authors contribute equally.

This journal is The Royal Society of Chemistry 2015

the area of the development of cost-eective electrocatalysts

for energy applications,1 especially as oxygen reduction reaction (ORR) electrocatalysts in fuel cells.1 Various approaches
have been made in the past in order to narrow down the overpotential for the ORR, which subsequently led to the development of heteroatom-based catalysts as possible potential
replacements for the state-of-the-art Pt/C.17 Recently, Fe and
Co supported on graphene/carbon nanotubes/carbon nanohorns (Gr/CNTs/CNHs) have been investigated for ORR and
have shown interesting activity characteristics.49 Despite this,
control over the size of the transition metals and their stability
under operating conditions need to be tackled eectively.47
Therefore, improvement in the surface area of transition
metals such as Fe, Co, Ni, and their eective interaction with

Nanoscale

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

Paper

the support substrate might be a good solution.414 Therefore,

one way to improve the surface area is to narrow down the particle size of the supported metals. Secondly, an eective interaction can be achieved by tuning the surface of the support
substrate without aecting its conductivity. In this context, the
use of heteroatom-doped carbon during the synthesis will be
an eective solution. In view of this, recently, hybrids of Fe, Co
and Mn nanoparticles with NGr/NCNTs (N stands for nitrogen) have been reported to display good performance for
ORR.414 However, these methods often failed to deliver a controlled and uniform particle size with better dispersion on/in
the carbon matrix, due to their intrinsic magnetic characteristics. Therefore, to deal with such magnetic materials,
researchers have focused on the use of surfactants, capping
agents and/or organic/inorganic reducing agents. However,
interference from impurities originating from such agents
sometimes create dead areas with a substantial reduction in
the area of the catalytically active centers.1519 Hence, there is
an urge to develop a simple and easy workup process to restrict
the growth of the magnetic metal nanoparticles and their
oxides and alloys in the nano regime, either in solution and/or
in supported form, and without the interference from the
reagents used.
Here, we report a process, which is free of interference from
the surfactants/capping agents, and which can ensure the sizecontrolled dispersion of nanocrystalline FeFe2O3 particles on
NGr to serve as a potential electrocatalyst for the oxygen
reduction reaction. This has been achieved by the interlocking
of the as-formed FeFe2O3 nanoparticles on NGr by facilitating
the reaction in water-in-oil (W/O) emulsion droplets
(Scheme 1). The emulsion method helps to confine the redox
process within the territory of the tiny emulsion droplets,
giving good control over the size and dispersion of the Fe
Fe2O3 nanoparticles on NGr. In other words, the growth kinetics could be controlled well within the nano regime of the

Nanoscale

emulsion droplets, which virtually act as nanoreactors.20,21

This leads to an eective property modulation that is concomitant with the establishment of electrochemical stability
and a substantially reduced overpotential for ORR.

2. Experimental
2.1.

Materials

Ferric nitrate (Fe(NO3)3, graphite, potassium permanganate

(KMnO4), phosphoric acid (H3PO4), melamine, sulfuric acid
(H2SO4), hydrogen peroxide (H2O2), sodium borohydride
(NaBH4), potassium hydroxide (KOH), perchloric acid (HClO4),
silicone oil and hexane were purchased from Sigma-Aldrich.
All the chemicals were used as such without any treatment.
2.2.

Synthesis of reduced graphene oxide (RGO)

Initially, an improved Hummers method was employed for

the synthesis of graphene oxide (GO).22 GO was further subjected to heating at 900 C in a tubular furnace under
Ar atmosphere in order to prepare the reduced graphene oxide
(RGO). The temperature was maintained for 3 h in order to
achieve the required level of reduction. Later, the furnace was
cooled down naturally under the same atmosphere to get the
final RGO sample.
2.3.

Synthesis of N-doped Gr (NGr)

For the synthesis of N-doped Gr (NGr), the graphene oxide

(GO) was mixed with melamine in the ratio of 1 : 6 in DI water
(100 ml) and the mixture was subjected to stirring for 24 h. In
the next step, the GOmelamine mixture was kept on stirring
at 70 C until complete evaporation of the water is achieved.
Subsequently, the dry GOmelamine powder was subjected to
heating at 900 C for 3 h under Ar-atmosphere. Following this
process, the temperature of the furnace was allowed to cool
naturally in the same atmosphere and the obtained NGr was
collected.
2.4. Synthesis of nanocrystalline FeFe2O3 particles
supported on nitrogen-doped graphene (FeFe2O3/NGr)

Scheme 1 Schematic illustration of the formation of FeFe2O3/NGr

electrocatalyst using a simple water-in-oil (W/O) emulsion technique.

Nanoscale

For the synthesis of FeFe2O3/NGr, 4 mg of the as-synthesized

NGr was mixed with 2 ml of aq. 0.1 M Fe(NO3)3 and the
mixture was sonicated for 30 min to form a homogeneous
solution. Afterwards, the obtained solution was added dropwise (at room temperature) into previously taken silicone oil
(25 ml) with continuous stirring (750 rpm). The stirring was
continued even after the addition for another 5 h. Subsequently, the mixture was transferred into a Petri dish, and
was subjected for heating at 80 C for 80 h in an oven. During
this process, evaporation of water and crystallization of the
solid components are expected to take place simultaneously.
Finally, the formed FeFe2O3/NGr was separated by centrifugation (at 10 000 rpm for 15 min) before being washed with
hexane. Subsequently, the obtained product was dried under a
lamp. Scheme 1 represents the illustration of the processes
involved in the synthesis of FeFe2O3/NGr.

This journal is The Royal Society of Chemistry 2015

View Article Online

Nanoscale

2.5. Synthesis of nanocrystalline FeFe2O3 particles

supported on reduced graphene oxide (FeFe2O3/RGO)
FeFe2O3/RGO was prepared by using the same procedure
adopted for synthesizing FeFe2O3/NGr, just by replacing NGr
with RGO.

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

2.6.

Structural characterization

For imaging the prepared electrocatalysts, HR-TEM analysis

was carried out. The sample for HR-TEM was prepared by
drop-coating the well dispersed sample in isopropanol (1 mg
of sample in 5 ml solvent) on a Cu grid (200 mesh). Subsequently, the coated surface of the electrode was dried in a
vacuum prior to imaging. The imaging was done on an FEI,
TECNAI G2 F20 instrument (conditions: accelerated voltage =
200 kV, Cs = 0.6 mm, and resolution = 0.17 nm). X-ray diraction (XRD) patterns were recorded using a PANalytical instrument (conditions: radiation Cu K ( = 1.54 ), scan rate
2 min1, step size 0.02 in 2. For carrying out the X-ray
photoelectron spectroscopy (XPS) measurements, a VG Micro
Tech ESCA 300 instrument was used by maintaining a
pressure of >1 109 Torr, pass energy of 50 eV, electron takeo angle of 60 and a resolution of 0.1 eV.
2.7.

Electrochemical studies

The electrochemical properties were investigated by using a set

of electrochemical methodologies including cyclic voltammetry (CV), a rotating disc electrode (RDE) and a rotating ring
disk electrode (RRDE) (Pine Instruments) in a three-electrode
electrochemical cell fitted with a platinum flag as the counter
electrode and Hg/HgO as the reference electrode on a BioLogic instrument (Model: VMP-3). For the working electrode
preparation, first, the catalyst ink was prepared by mixing the
electrocatalyst (5 mg) in 1 ml ethanolwater (3 : 2) solution.
Subsequently, the ink was drop-coated on the surface of the
RDE (0.19625 cm2) and the RRDE. The RRDE consists of a
glassy carbon (GC) disk (diameter = 6 mm) and a Pt ring (outer
diameter = 9 mm, inner diameter = 7.5 mm). The coating was
done to attain a constant mass (total catalyst) loading of 50 g.
Special care was taken to maintain the same loading of the
catalyst in all the samples. Prior to coating the RDE and RRDE
surfaces, the surfaces of the electrodes were polished with
0.3 m alumina slurry followed by washing with acetone and
DI water. After the material coating, the electrode was dried
under a lamp for 4 h. This was used as the working electrode
for the electrochemical study. An aqueous 0.1 M KOH solution
(de-aerated with nitrogen gas) was used as an electrolyte for
the CV, RDE and RRDE studies. Moreover, in order to screen
the ORR activity of FeFe2O3/NGr in acid medium, another set
of experiments was carried out by using HClO4 as the electrolyte, Ag/AgCl as the reference electrode and a platinum flag as
the counter electrode. These results dealing with the activity
evaluation under acid conditions are discussed in the ESI.
Compared to alkaline conditions, the performance of the
system under an acidic environment is relatively low. Hence,
the detailed investigation of the present catalyst is restricted to

This journal is The Royal Society of Chemistry 2015

Paper

0.1 M KOH as the electrolyte. To get the current density plots,

the measured current was divided by the geometrical area of
the electrode. Mainly, the RDE technique has been used to get
the oxygen reduction polarization data, and conclusions on
the intrinsic ORR activities of the systems are obtained by
generating KouteckyLevich (KL) plots. The KL plot involves
a plot of the inverse of the current density j (i.e. 1/j ) versus
1/1/2, where is the angular rotation of the electrode. The
obtained current density ( j ) during the measurement is the
combination of three dierent types of current densities,
namely, the diusion-limiting current density ( jd), the kinetic
current density ( jk), and the film diusion current density
( jf ). Here, the contribution of jf is neglected owing to the very
low concentration of binder used during the electrode preparation. Hence, the obtained current density can be written in
the form of eqn (1), and using the expanded forms of the
corresponding parameters eqn (1) can be written as eqn (2),
1 1 1

j jk jd

1
1
1

j nFkCo* 0:62nFC* Do2=3 1=6 1=2

where, j is the current density of the glassy carbon disk electrode, n is the number of electrons involved during the ORR,
A is the geometrical area of the electrode (0.19625 cm2), C*0
is the oxygen concentration in the bulk (1.22 106 mol cm3),
k is the rate constant, F is the Faraday constant (96 500 C),
is the kinematic viscosity of the electrolyte (0.01 cm2 s1),
Do is the diusion coecient of oxygen (1.93 105 cm2 s1)
and lastly, is the angular rotation rate of the electrode in
radians s1 [ = 2f = (2*rpm)/60]. Eqn (2) provides a linear
relation of 1/j and 1/1/2, and the intercept of the graph gives
the kinetic current density jk. Moreover, the RRDE technique
has been used to determine the exact behavior (the 2- and
4-electron oxygen reduction pathways) of the electrocatalyst
during the ORR. Prior to that, the collection eciency (N) of
the RRDE was obtained by using a standard ferrocene couple.
N is found to be 0.37 in the present case.20,21 During the RRDE
investigation, the ring potential was set to 0.60 V (vs. Hg/HgO).
The electron pathway of the electrocatalysts towards ORR was
investigated by calculating the yield of hydrogen peroxide
(H2O2) by using eqn (3) and the number of electrons transferred during the process by using eqn (4),23



jR =N
H2 O2 % 200
3
jD jR =N
2

6 4jD
7
 7
n6
4
jR 5
jD
N

where, jR and jD are the Faradaic ring current density and

disc current density, respectively. N is the collection eciency
of the ring electrode (0.37) and the n is the number of transferred electrons.

Nanoscale

View Article Online

Paper

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

2.8.

Nanoscale

Fuel cell study

Single-cell performance investigation of the FeFe2O3/NGr

electrocatalyst has been carried out by using an FCT test
station (Fuel Cell Technologies Inc., USA). Lab-view software
was used to control the cell parameters. A FumaTech FAA
membrane was used as the solid anion exchange membrane.
Prior to use, the membrane was soaked in 1 M KOH solution
for 24 h, after which the wet and swelled membrane was
washed with DI water. The cathode electrode (4 cm2 active
area) was fabricated by maintaining a catalyst loading of
3.00 mg cm2 of FeFe2O3/NGr on a porous carbon paper. For
comparison purposes, another cathode electrode was also fabricated with 40 wt% Pt/C by keeping a constant Pt-loading of
0.35 mg cm2. In all the cases the Pt/C-coated carbon paper
(0.35 mg cm2) was used as the anode electrode. The membrane electrode assembly (MEA) was fabricated by sandwiching the Pt/C anode and FeFe2O3/NGr cathode with the
pretreated anion exchange membrane. The MEA was fabricated
at 25 C with an applied pressure of 0.3 ton. In addition,
during the MEA fabrication, 250- and 140-m-thick Kapton
and FRT were used as the sub-gasket and main gasket, respectively. The control MEA also was similarly fabricated by using
the Pt-coated carbon paper electrode as the cathode. Finally,
the MEAs (area: 4 cm2) were tested by passing hydrogen at the
anode and oxygen at the cathode side at a flow rate of 0.2 standard liters per minute (slpm) and with a relative humidity of
100%.

3. Results and discussion

Fig. 1a and 1b show the good dispersion of FeFe2O3 nanoparticles on the surface of NGr (FeFe2O3/NGr). Here, Fig. 1b is
an exposed image of the part highlighted in the red rectangle,
which indicates the presence of well separated FeFe2O3 particles having a homogeneous dispersion on the surface of NGr,
which is quite rare in the case of magnetic nanoparticles.
Fig. 1c shows the uniform size and controlled geometry of the
FeFe2O3 particles with a size in the range of 23 nm by eectively preventing particle agglomeration. Fig. 1d displays the
HR-TEM image of FeFe2O3/NGr and Fig. 1e is the highlighted
portion of Fig. 1d (marked with a green-colored rectangle)
showing the well separated particles with clear lattice fringes.
The FeFe2O3 particles in Fig. 1d (two particles are marked
with yellow-colored dashed circles) indicate the uniform size
distribution with a controlled size of 3.0 nm. More interestingly, Fig. 1e displays crossing of the lattice fringes with lattice
constant values of 0.23 and 0.24 nm, which indicates the presence of mixed faceted FeFe2O3 particles. This mixed faceted
particle growth might have originated due to the eective
maintenance of the thermochemical relationship between Fe
(0) and Fe(II) during the synthesis.24 The SAED pattern provides information on the crystallinity of FeFe2O3/NGr
(Fig. 1f ). Moreover, the particle size histogram of FeFe2O3/
NGr (Fig. S1a) shows an average particle size of 3.0 nm. In
comparison, FeFe2O3/RGO is found to show an average par-

Nanoscale

Fig. 1 TEM images of FeFe2O3/NGr: (a and c) displaying the images

taken at dierent magnications, (b) enlarged view of a portion of (a)
which is marked with the red-colored dashed rectangle, (d) high resolution image of the system indicating the well separated FeFe2O3 nanoparticles with lattice fringes, where the particles marked with yellow
circles have sizes of 3 nm, (e) highlighted portion of (d) which is
marked with the green-colored rectangle, where the crossing of lattice
fringes is marked with yellow lines showing the lattice constant values
of around 0.23 and 0.24 nm, and (f ) the selected area electron diraction (SAED) pattern.

ticle size of 5.4 nm (Fig. S1b and S2). This reduction in the
size and better dispersion of FeFe2O3 nanoparticles on NGr
in comparison with the FeFe2O3/RGO could be attributed to
the important role played by the nitrogen of NGr and the high
pressure exerted by the water vapors within the infinitesimal
reaction area within the W/O nanoreactor.20,21
Moreover, the crystalline nature of the material was confirmed through X-ray diraction (XRD) study. Comparative
XRD profiles of FeFe2O3/RGO, FeFe2O3/NGr, RGO, and NGr
are provided in Fig. 2. The clear dierence (lower shift) in the
G(002) diraction peak of NGr in comparison with RGO indicates the eective nitrogen doping within the graphene
network.20,21,2426 In addition, the eective interaction
between the FeFe2O3 nanoparticles with NGr further shifts
the G(002) peak towards lower 2. The peaks at 12.5, 19.0,
21.2, 25.0, 28.6, 31.9, 35.0, 41.2, 51.9, 59.2, and 63.1 are
assigned to the Fe2O3 (102), (220), (014), (012), (104), (220),
(110), (006), (116), (122), and (300) planes, respectively (Fig. 2).
In addition, the characteristic diraction peaks of Fe0 appear
at 2 values of 40.2 and 51.1, which could be assigned to the
(011) and (200) planes, respectively.27 Thus, the obtained 2
values highlight the presence of two distinct phases of iron,

This journal is The Royal Society of Chemistry 2015

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

Nanoscale

Fig. 2 Comparative XRD proles of RGO, NGr, FeFe2O3/RGO, and Fe

Fe2O3/NGr.

i.e., the Fe-phase and the Fe2O3-phase. (JCPDS card number

06-0669, 39-1346.27)
Furthermore, the surface electronic configuration of
FeFe2O3/RGO and FeFe2O3/NGr was investigated with the aid
of X-ray photoelectron spectroscopy (XPS). Fig. 3a and 3b show
the Fe 2p spectra of FeFe2O3/RGO and FeFe2O3/NGr, respectively. Both the spectra could be deconvoluted into five common
peaks at 707.06, 710.49, 719.10, 721.89, and 723.13 eV. The
photoelectron peaks shown at 707.06 and 710.49 eV correspond to the binding energies of 2p3/2 of Fe(0) and Fe(III) moieties, respectively.2832 The peak at 721.89 and 723.13 eV can

Fig. 3 Deconvulated Fe 2p spectra of (a) FeFe2O3/NGr, (b) FeFe2O3/

RGO, and (c and d) deconvulated XPS of C 1s and N 1s of FeFe2O3/
NGr, respectively.

This journal is The Royal Society of Chemistry 2015

Paper

be assigned to the 2p1/2 of Fe(0) and Fe(III) ions,

respectively.2832 The peak at 719.10 eV has been identified as
a satellite peak corresponding to Fe 2p3/2.2832 Interestingly, in
the case of FeFe2O3/RGO and FeFe2O3/NGr, the dierence
between the two characteristic peaks of Fe(0) and Fe(III) is
found to be 14.83 and 12.64 eV, respectively (Fig. 3a and 3b),
which matches with the previously reported values.33
Hence, it can be concluded that the iron oxide phase
retains Fe in two dierent oxidation states, i.e. Fe0 and Fe3+, in
both the RGO- and NGr-based systems.33 In conjunction,
Fig. 3a shows the presence of extra peaks at higher binding
energies of 712.90 and 725.12 eV, which provide information
on the existence of the FeN bonding in FeFe2O3/NGr.33 Furthermore, the quantitative estimation of the elements present
in FeFe2O3/RGO and FeFe2O3/NGr systems has been carried
out from the XPS results. The wt% of Fe is found to be 48.69
and 47.22 for FeFe2O3/NGr and FeFe2O3/RGO, respectively.
The C and N wt% values are found to be 47.69 and 3.62 in the
case of FeFe2O3/NGr, and 52.78 and 0.0 wt% in the case of
FeFe2O3/RGO. In detail, the C 1s spectra of FeFe2O3/NGr
(Fig. 3c) show a peak corresponding to the interaction of C
and Fe at the lower binding energy of 283.64 eV, whereas the
peaks appearing at 284.40 and 285.20 eV are credited to the
presence of CvC (sp2) and CC (sp3) states.20,21,33 However,
the peaks at the higher binding energy values of 286.19 and
287.78 eV indicate the presence of the oxygen functional
groups on the carbon surface and interaction of the CN
bonding, respectively.33 Looking very closely, the C 1s (Fig. 3c)
and N 1s (Fig. 3d) spectra of FeFe2O3/NGr show the distinctive
peaks at 283.64, 287.83 and 397.04 eV. This, along with the
previous observations on the existence of the FeN bonding,
unambiguously confirms the coordination of Fe with C and N,
which suggests that the moieties are in the form of MNC
and MC (M = Fe) (Fig. 3a, 3c and 3d). On the other hand, in
the case of FeFe2O3/RGO, it is in just the MC (M = Fe) format
(Fig. 3b and S4) as there is no nitrogen entity in the system to
facilitate an MNC coordination. Fig. 3d (N 1s of FeFe2O3/
NGr) shows the peaks at 397.04, 398.52, 400.16, 402.0 and
403.2 eV, which are assigned to the presence of NFe, pyridinic-N, pyrrolic-N, quaternary-N and NO bonding, respectively.33 Looking very closely at the N 1s of FeFe2O3/NGr
(Fig. 3d), we found that the percentages of pyridinic-N and pyrrolic-N are more than those of the contributions by the quaternary-N and pyridinic-N-oxide (NO). Out of the total 3.62 wt%
of nitrogen, contributions (%) of the pyridinic-N and pyrrolic-N are found to be 42.50 and 33.35%, respectively. By contrast, the quaternary-N and NO percentages are found to be
around 18.75 and 5.4%, respectively. Importantly, the peaks
corresponding to the dierent types of nitrogen appear at
slightly lower binding energy values than observed in pristine
NGr (Fig. S6), which is attributed to the interaction of the
FeFe2O3 nanoparticles with NGr in FeFe2O3/NGr. Therefore,
the shift and the existence of new characteristic peaks (i.e.
397.04 eV) in FeFe2O3/NGr give strong evidence of the mutual
charge transfer between the Fe and NGr entities, which helps
to tune the properties of FeFe2O3/NGr to serve as a better elec-

Nanoscale

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

Paper

trocatalyst for oxygen reduction.34,35 In addition to the vital

role played by NGr in modulating the electrochemical properties of the system, NGr is also expected to impart additional
stability to FeFe2O3/NGr due to its corrosion resistivity under
the electrochemical environment in comparison to
RGO.20,21,2426 Details on the corrosion resistivity of RGO and
NGr are given in a following section. Despite this, the d-orbitals of the metals interact with the s-orbital of NGr (C/N 1s)
and partial covalent bonding is formed, which also can contribute towards the attainment of better stability of the Fe
Fe2O3/NGr system.914 Therefore, these internal structural and
electronic modulations in FeFe2O3/NGr could make the
system more useful and active towards ORR compared to Fe
Fe2O3/RGO. For comparison, XPS data of C 1s and O 1s of
RGO, NGr, FeFe2O3/RGO, O 1s of FeFe2O3/NGr, and N 1s of
NGr have been provided in the ESI (Fig. S3S9).
Taking advantage of the smaller size of the FeFe2O3 nanoparticles evenly distributed on NGr and the presence of two
dierent oxidation states within the system, we examined the
material for ORR in 0.1 M KOH and found it to be very
ecient compared to all the other prepared samples. The presence of mixed valancies is being considered as an advantageous parameter in a system which can help for better
electron transfer during electrocatalysis.35 Along with this, the
system benefits from the small size and homogeneous distribution of the FeFe2O3 nanoparticles, and the activity modulations induced through the electrostatic interactions
facilitated by the doped nitrogen in NGr.
Initially, the cyclic voltammograms (CVs) were recorded in a
potential window of 0.80 to 0.40 V vs. Hg/HgO in N2-saturated 0.1 M KOH with a scan rate of 20 mV s1 (Fig. S10). Furthermore, to fully screen the ORR activity, linear sweep
voltammograms (LSVs) were recorded in a potential window of
0.30 to 0.50 V vs. Hg/HgO with dierent rotation rates (in
rotations per minute, rpm) of the electrocatalyst-coated electrode (Fig. S11). The increase in the current with respect to
the rpm of the electrode originates from the eective mass
transfer on the electrode surface. Furthermore, the comparative LSVs of all the samples with 20 wt% Pt/C are presented in
Fig. 4a. The onset potential of the ORR is in the order of Pt/C
(0.09 V) > FeFe2O3/NGr (0.075 V) > NGr (0.05 V) > FeFe2O3/
RGO (0.07 V) > RGO (0.15 V). The onset potential of Fe
Fe2O3/NGr is found to be shifted to the desired positive side of
the potential window by 145 mV in comparison to FeFe2O3/
RGO and by 125 and 225 mV with respect to NGr and RGO,
respectively. The clear positive shift in the case of FeFe2O3/
NGr thus confirms the better ORR activity achieved by the
system. This favourable activity modulation has helped to
narrow down the dierence between the performances of the
present system with the state-of-the-art Pt/C, where the onset
potential dierence is found to be only 15 mV. Whereas the
improved onset potential dierence confirms the better intrinsic activity achieved by the system, an investigation on the
half-wave potential (E1/2) has been considered as a reliable
method to understand the role of the active center density in
defining the performance of the system under the current

Nanoscale

Nanoscale

Fig. 4 (a) Comparative LSVs recorded at 1600 rpm in O2-saturated 0.1

M KOH with a scan rate of 10 mV s1 and (b) KouteckyLevich plots
obtained at 0.30 V (vs. Hg/HgO) of the catalysts.

dragging conditions. The E1/2 values follow the order of Pt/C

(0.00 V) > FeFe2O3/NGr (0.045 V) > NGr (0.12 V) > Fe
Fe2O3/RGO (0.18 V) > RGO (0.25 V). Thus, the improvement
acquired in terms of the onset potential as well as E1/2 of Fe
Fe2O3/NGr compared to the rest of the prepared systems
clearly stands out as valid evidence of the higher intrinsic ORR
activity as well as the active reaction center density of the
system. This result, thus, goes well with the previous observation that the system is expected to attain activity modulation
due to factors such as the presence of the mixed valancies of
iron, the small size and homogeneous distribution of the Fe
Fe2O3 nanoparticles and the activity modulations induced by
the doped nitrogen in NGr. A controlled interplay of these
factors appears to work favourably in the present case. In
addition, to confirm the corrosion resistivity of RGO and NGr
in alkaline conditions, an accelerated durability test (ADT) was
performed (Fig. S12 and S13). RGO displays a 200 mV negative E1/2 shift after the 5000 potential cycles (i.e., ADT) in a
potential window of 0.0 to 0.4 V (vs. Hg/HgO) (Fig. S12). On
the other hand, NGr displays 122 mV of negative E1/2 shift
(Fig. S13). The obtained results can be directly correlated to
the extent of corrosion resistivity of RGO and NGr in alkaline
conditions, and, thus, it can be confirmed that the stability of
NGr is better than that of RGO. Along with this advantage of
having the higher corrosion resistivity of the NGr support, the
eective interaction between the Fe and N of NGr make the
FeFe2O3/NGr system more viable in comparison to the Fe
Fe2O3/RGO system because, the wt% of Fe is the same in both

This journal is The Royal Society of Chemistry 2015

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

Nanoscale

cases and the only the dierence is in the content of N. This

could lead to the formation of eective active centres, i.e., Fe
NC and FeC. Overall, this can give a clear benefit of having
the presence of N and its favourable role in tuning the activity
of the Pt-free systems.
A more insightful analysis on the ORR kinetics has been
made by obtaining KouteckyLevich (KL) plots (Fig. 4b and
S14).23 The straight line plots (Fig. 4b and S14) lead to the
calculation of the number of electrons taking part during the
ORR process and the values were found to be 3.8, 3.6, 3.5 and
2.2 for FeFe2O3/NGr, FeFe2O3/RGO, NGr and RGO, respectively. Moreover, the kinetics of ORR can be judged from the
kinetic current density values, which are found to be in the
order of Pt/C (12.12 mA cm2) > FeFe2O3/NGr (9.48 mA cm2)
> NGr (6.49 mA cm2) > FeFe2O3/RGO (4.90 mA cm2) > RGO
(2.58 mA cm2). This unravels a nice interplay of the substrateactive species interaction, wherein the FeFe2O3 nanoparticles supported on NGr attain more contribution from the
preferred 4-electron reduction pathway. In addition, a rotating
ring disk electrode (RRDE) study was carried out to investigate
the exact percentage of H2O2 formed during the ORR process
by the 2-electron reduction process, which is often expected to
accompany the preferred 4-electron route. It has been documented that the Pt-free and carbon-based materials show a
high H2O2 yield compared to the Pt-based electrodes.3640
During the RRDE investigation (Fig. 5a), the ring potential
was kept constant at 0.60 V (vs. Hg/HgO) and the disk and ring
currents were monitored separately and used to calculate the
amount of H2O2 formed during the process (Fig. 5a). It can be
noted from the ring current profile that the current steadily
increases in the kinetic region and displays a gradual increase
in the diusion region (i.e. <0.20 V) (Fig. 5a). Fig. 5b and

Fig. 5 (a) Comparative rotating ring disk electrode (RRDE) studies of

FeFe2O3/RGO, FeFe2O3/NGr, RGO, NGr and Pt/C (rotation of the
working electrode: 1600 rpm; electrolyte: O2-saturated 0.1 M KOH; scan
rate: 10 mV s1; ring potential: constant 0.60 V (vs. Hg/HgO)). Here, jD
and jR are the recorded disk and ring currents, respectively. (b) H2O2%
with respect to the potential of the disk electrode and (d) H2O2% at
0.30 V (vs. Hg/HgO).

This journal is The Royal Society of Chemistry 2015

Paper

S15 show the variation in the H2O2% and the number of electrons transferred with respect to the potential of the disc electrode during the ORR process. FeFe2O3/RGO shows 51.2% of
H2O2, at 0.30 V (vs. Hg/HgO), which gets reduced to 13.0% for
FeFe2O3/NGr (Fig. 5c). Conversely, RGO and NGr show 60.0
and 41.0% of the H2O2 yield, respectively. This leads to a conclusion that when FeFe2O3 is supported on RGO, the H2O2% is
reduced (Fig. 5b and 5c) by 14.7% in comparison to RGO alone.
Contrary to this, the FeFe2O3 dispersion on NGr leads to a
reduction in the production of H2O2 by 69%, in comparison to
NGr alone (Fig. 5b and 5c). Interestingly, looking at the above
two scenarios, it can be seen that the FeFe2O3/NGr catalyst
shows an overall 78.3% reduction in the production of H2O2 in
comparison with FeFe2O3/RGO. The observed better activity
of FeFe2O3/NGr in comparison to the all other prepared catalysts is assigned to the doped nitrogen and their strong interaction with the metal counterpart. This is because the % of Fe
is found to be same in both the FeFe2O3/RGO and FeFe2O3/
NGr systems and the only dierence is in the nitrogen content
and its coordination (i.e., FeNC and FeC) between the
systems. The obtained activity of FeFe2O3/NGr has been compared with the literature reports, as presented in Table S1.
Furthermore, in a realistic system level validation, the solid
alkaline electrolyte membrane fuel cell (AEMFC) performance
was evaluated by employing the FeFe2O3/NGr as the ORR catalyst and a FumaTech-FAA membrane as the anion exchange
electrolyte. Fig. 6 compares the IV characteristics of the 4 cm2
membrane electrode assembly (MEA) in single-cell mode with
the performance characteristics of a similar system derived by
using the commercially available 40 wt% Pt/C as the cathode.
The FeFe2O3/NGr fabricated single cell shows an OCV of
0.73 V and subsequently it delivers the maximum power and
current densities of 54.4 mW cm2 and 200 mA cm2, respectively, which are comparable to the values reported for the

Fig. 6 Comparative single-cell polarization curves obtained from the

MEAs in single-cell mode by using FeFe2O3/NGr and 40 wt% Pt/C as
the cathode catalysts. The Pt/C catalyst was used as the anode catalyst
in both cases. Cell temperature: 60 C, Pt-loading: 0.35 mg cm2, Fe
Fe2O3/NGr loading: 3.0 mg cm2, relative humidity: 100%, and feed
composition: anode (H2, 0.2 slpm) and cathode (O2, 0.2 slpm).

Nanoscale

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

Paper

Nanoscale

Fe/N/C catalysts (Table S2).41,42 On the other hand, the Pt/C

fabricated single cell displays an OCV of 0.75 V and delivers
the maximum power and current densities of 70.0 mW cm2
and 250 mA cm2, respectively. The observed single-cell data
of Pt/C are in good agreement with the literature reports.41,42
Even though the Pt/C-based cell outperforms the FeFe2O3/
NGr-based cell, it should be noted that the basic features covering the activation, ohmic and concentration polarization
regions of the IV plots essentially follow a similar type of behaviour. Also, the performance of the FeFe2O3/NGr fabricated
MEA is quite good compared to many of the systems reported
in the literature (Table S2).43,44 The relatively good AEMFC
performance obtained in the present case goes well alongside
the intrinsic activity characteristics of the catalyst explored
through the various electro-analytical methodologies as discussed in the previous sections. It also should be noted that
the AEMFC performance is largely dependent upon how well
the anion exchange membrane performs its role as a solid
electrolyte for facilitating the OH ion transport. Perhaps this
issue exists as one of the biggest bottlenecks in obtaining the
maximum benefits from various non-Pt electrocatalysts which
display good intrinsic ORR activity under alkaline conditions.
Also, the catalyst ink preparation for the electrode fabrication
and the procedures involved for making the MEA require
careful optimization of many parameters, which really aect
the actual performance. In the present study, we followed
similar electrode fabrication protocols while using our catalyst
and Pt/C despite their dierences in terms of the carbon
content and active centres. All these things need further
optimization to achieve an even better performance, and our
current research activities are also aligned in the direction of
developing a better protocol which can overcome some of the
existing issues.

4.

Conclusions

To sum up, we have synthesized a system consisting of nanocrystalline FeFe2O3 particles (3 nm) supported on NGr (Fe
Fe2O3/NGr) at low temperature through a W/O emulsion
process by utilizing the multi-functional properties of NGr as
the reducing agent, support material, and nucleating site (i.e.
nitrogen of NGr). The in situ redox reaction, which is well controlled within the boundary of the W/O emulsion droplets,19,20
provides concomitant assurance of the size-controlled dispersion of FeFe2O3 nanoparticles on NGr and synergistic electronic interactions to tune the property of the material to serve
as a potential ORR catalyst in alkaline condition. The electrochemical investigation was targeted on understanding the
oxygen reduction kinetics and discloses the key role played by
the interaction of FeFe2O3 nanoparticles with NGr in modulating the property of the overall system to drive the ORR with
a greater contribution from the desired 4-electron pathway. As
an ultimate benefit of this favorable supportactive site interaction, FeFe2O3/NGr could narrow down the overpotential for
ORR by bringing a significant positive shift to the reduction

Nanoscale

onset potential, which is just 15 mV higher than its Pt-based

counterpart. The E1/2 is also found to be significantly improved
for FeFe2O3/NGr by 135 mV in comparison to FeFe2O3/RGO,
suggesting the greater number density of the active sites
acquired by the system. The FeFe2O3/NGr cathode-based MEA
of AEMFC displayed good IV characteristics in single-cell
mode. The cell could deliver 54.4 mW cm2 and 200 mA cm2
of power- and current-density values, respectively.

Acknowledgements
This study was funded by Council of Scientific and Industrial
Research (CSIR), New Delhi, India, by research fellowship to
V.M.D. and project funding (CSC0122) to S.K. Authors are
thankful to Dr K. R. Patil for the XPS and Dr Gopinathan Anil,
Noritake Ltd, Japan, for the TEM analyses.

Notes and references

1 S. Yang, L. Zhi, K. Tang, X. Feng, J. Maier and K. Mullen,
Adv. Funct. Mater., 2012, 22, 3634.
2 Z. Yang, H. Nie, X. Chen, X. Chen and S. Huang, J. Power
Sources, 2013, 236, 238.
3 L. Wang, A. Ambrosi and M. Pumera, Angew. Chem., Int.
Ed., 2013, 52, 13818.
4 Y. Hou, T. Huang, Z. Wen, S. Mao, S. Ciu and J. Chen, Adv.
Energy Mater., 2014, 4, 1400337.
5 Q. Yi, Y. Zhang, X. Liu, B. Xiang and Y. Yang, J. Mater. Sci.,
2014, 49, 729.
6 N. A. Aquayo, Z. Liu, E. Bertran and J. Yang, J. Phys. Chem.
C, 2013, 117, 19167.
7 L. M. A. Monzon, K. Rode, M. Venkatesan and
J. M. D. Coey, Chem. Mater., 2012, 24, 3878.
8 A. Muthukrishnan, Y. Nabae, T. Okajima and T. Ohsaka,
ACS Catal., 2015, 5, 5194.
9 F. Banhart, Nanoscale, 2009, 1, 201.
10 S. Guo, S. Zhang, L. Wu and S. Sun, Angew. Chem., 2012,
124, 11940.
11 S. Guo, S. Zhang, L. Wu and S. Sun, Angew. Chem., Int. Ed.,
2012, 51, 11770.
12 J. Duan, Y. Zheng, S. Chen, Y. Tang, M. Jaronie and S. Qiao,
Chem. Commun., 2013, 49, 7705.
13 K. Parvez, S. Yang, Y. Hernandez, A. Winter, A. Turchanin,
X. Feng and K. Mullen, ACS Nano, 2012, 6, 9541.
14 Y. Lim, M. Gong, Y. Liang, J. Feng, J.-E. Kim, H. Wang,
G. Hong, B. Zhang and H. Dai, Nat. Commun., 2013, 4,
1805.
15 J. Turkevich, P. C. Stevenson and J. Hillier, Discuss. Faraday
Soc., 1951, 11, 55.
16 M. Brust, M. Walker, D. Bethell, D. J. Schirin and
R. Whyman, J. Chem. Soc., Chem. Commun., 1994, 801.
17 H. Kawasaki, Nanotechnol. Rev., 2013, 2, 5.
18 C. Yang, H. B. Zhao, Y. L. Hou and D. Ma, J. Am. Chem.
Soc., 2012, 134, 15814.

This journal is The Royal Society of Chemistry 2015

View Article Online

Published on 28 October 2015. Downloaded by University of Ulsan on 23/11/2015 02:32:40.

Nanoscale

19 V. Amendola, P. Riello and M. Meneghetti, J. Phys. Chem. C,

2011, 115, 5140.
20 V. M. Dhavale, S. S. Gaikwad, L. George, R. N. Devi and
S. Kurungot, Nanoscale, 2014, 6, 13179.
21 V. M. Dhavale, S. S. Gaikwad and S. Kurungot, J. Mater.
Chem. A, 2014, 2, 1383.
22 D. C. Marcano, D. V. Kosynkin, A. Berlin, Z. Sinitskii,
A. Sun, L. B. Slesarev, W. Alemany and J. M. Tour, ACS
Nano, 2010, 4, 4806.
23 A. J. Bard and L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, Wiley, 2nd edn, 2002, ISBN:
978-81-265-0807-5.
24 N. Luciano, C. Lourdes, L. Agostinho and F. C. Bernadete,
J. Chil. Chem. Soc., 2013, 58, 1609, DOI: 10.4067/S071797072013000100023.
25 X. X. Wang, Z. H. Tan, M. Zeng and J. N. Wang, Sci. Rep.,
2013, 4, 1.
26 N. W. Khun and E. Liu, J. Electrochem. Soc., 2010, 157,
C269.
27 M. Ruguang, M. Wang, P. Tao, Y. Wang, C. Cao, G. Shan,
S. Yang, L. Xi, J. C. Y. Chung and Z. Lu, J. Mater. Chem. A,
2013, 1, 15060.
28 Z. Li, H. Chen, H. Bao and M. Gao, Chem. Mater., 2004, 16,
1391.
29 P. Mills and J. L. Sullivan, J. Phys. D: Appl. Phys., 1983, 16,
723.
30 R. V. Siriwardane and J. M. Cook, J. Colloid Interface Sci.,
1985, 108, 414.
31 P. Marcus and J. M. Grimal, Corros. Sci., 1992, 33, 805.

This journal is The Royal Society of Chemistry 2015

Paper

32 X. Wang, S. Zhang, J. Li, J. Xub and X. Wang, Inorg. Chem.

Front., 2014, 1, 641.
33 J. F. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben,
in Handbook of X-ray Photoelectron Spectroscopy: A Reference
Book of Standard Spectra for Identification and Interpretation
of XPS Data, ed. J. C. Prairie and E. Minn, Physical
Electronics Division, Perkin-Elmer Corp., 2nd edn,
1992.
34 W. Zhou, J. Zhou, Y. Zhou, J. Lu, K. Zhou, L. Yang, Z. Tang,
L. Li and S. Chen, Chem. Mater., 2015, 27, 2026.
35 S. K. Singh, V. M. Dhavale and S. Kurungot, ACS Appl.
Mater. Interfaces, 2015, 7, 442.
36 T. Okada, T. Abe and M. Kaneko, Springer series in Material
Science 111, Spinger-verlag, Berlin, 2009.
37 L. Zhang and Z. Xia, J. Phys. Chem. C, 2011, 115, 11170.
38 K. Gong, F. Du, Z. Xia, M. Durstock and L. Dai, Science,
2009, 323, 760.
39 R. L. Liu, D. Q. Wu, X. L. Feng and K. Mullen, Angew.
Chem., Int. Ed., 2010, 49, 2565.
40 S. Wang, L. Zhang, Z. Xia, A. Roy, D. W. Chang, J. B. Baek
and L. Dai, Angew. Chem., Int. Ed., 2012, 51, 4209.
41 J. Sanetuntikul and S. Shanmugam, Nanoscale, 2015, 7,
7644.
42 S. M. Unni, S. Ramadas, R. Illathvalappil, S. N. Bhange and
S. Kurungot, J. Mater. Chem. A, 2015, 3, 4361.
43 T. Palaniselvam, M. O. Vlappil, R. Illathvalappil and
S. Kurungot, Energy Environ. Sci., 2014, 7, 1059.
44 C. V. Rao and Y. Ishikawa, J. Phys. Chem. C, 2012, 116,
4340.

Nanoscale