Nanostructured amorphous MnO2 prepared by reaction of KMnO 4 with triethanolamine

Article abstract of DOI:10.1016/j.jallcom.2010.06.072

Amorphous manganese dioxide is prepared by reaction of potassium permanganate with an organic reductant triethanolamine. The effect of heat-treatment temperature is studied on the characteristics of the materials. Power X-ray diffraction (XRD), scanning electron microscope (SEM) and N ₂ adsorption and desorption measurements are employed to investigate crystalline structure, surface morphology, the specific surface area and the pore size distribution. It is found that when the annealing temperature reaches up to 400 °C, the crystalline convert to α-MnO₂ from amorphous MnO₂. The electrochemical characteristics of the prepared MnO₂ powder are characterized by means of cyclic voltammetry (CV), experiments in 1.0 mol L^-1 Na₂SO₄ electrolyte. The specific capacitance (SC) value is 251 F g^-1 that is obtained from the product annealing at 350 °C at a CV scan rate of 2 mV s ^-1. And charging-discharging measurement reveals the good stability of the prepared material.

Full text of DOI:10.1016/j.jallcom.2010.06.072

Journal of Alloys and Compounds 505 (2010) 555–559
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Nanostructured amorphous MnO2 prepared by reaction of KMnO4 with
triethanolamine
Yan-jing Yang, En-hui Liu∗, Li-min Li, Zheng-zheng Huang, Hai-jie Shen, Xiao-xia Xiang
Key Laboratory of Environmentally Friendly Chemistry and Applications of Ministry of Education, College of Chemistry, Xiangtan University, Xiangtan, Hunan 411105, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Amorphous manganese dioxide is prepared by reaction of potassium permanganate with an organic
reductant triethanolamine. The effect of heat-treatment temperature is studied on the characteristics
of the materials. Power X-ray diffraction (XRD), scanning electron microscope (SEM) and N2 adsorption
and desorption measurements are employed to investigate crystalline structure, surface morphology,
the specific surface area and the pore size distribution. It is found that when the annealing temperature
reaches up to 400 ◦C, the crystalline convert to ␣-MnO2 from amorphous MnO2. The electrochemical
characteristics of the prepared MnO2 powder are characterized by means of cyclic voltammetry (CV),
experiments in 1.0 mol L−1 Na2SO4 electrolyte. The specific capacitance (SC) value is 251 F g−1 that is
obtained from the product annealing at 350 ◦C at a CV scan rate of 2 mV s−1. And charging–discharging
measurement reveals the good stability of the prepared material.
Received 24 November 2009
Received in revised form 9 June 2010
Accepted 11 June 2010
Available online 23 June 2010
Keywords:
Manganese oxides
Heat-treatment
Supercapacitor
Electrochemical properties
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
lation/deintercalation (i/d) of cations (such as Li+, Na+, K+) or/and
protons (H3O+) into the bulk of MnO2, which may be described by
Supercapacitors (or eletrochemical capacitors) have recently
drawn much attention as energy storage devices due to their higher
power density and long cycle life compared to secondary batter-
ies and higher energy density than conventional capacitors [1–3].
According to the charge storage mechanism they can be classi-
fied into two types: One is electrochemical double-layer capacitor
(EDLC) which exhibits a non-faradic reaction with accumulation of
charges at the interface between the electrode and the electrolyte.
The electrodes mainly focus on carbon-based materials [4,5]. The
other is pseudocapacitor which stores energy by the process of
faradic redox reaction [6–8].
the redox reaction:
MnO2 + C+ + e− ꢀ MnOOC(M = Li+, Na+, K+ or H3O+)
The second mechanism is a surface adsorption/desorption (a/d)
process of cations or/and protons from the electrolyte:
(1)
(MnO2)surface + C+ + e−
ꢀ (MnOOC)surface(C+ = Li+, Na+, K+ or H3O+)
The i/d process is expected to take place in the crystalline
MnO2, while the a/d process mainly occurs in the amorphous MnO2
[12,13].
(2)
The psuedocapacitors based on transition oxides (such as
RuO·nH2O, CoOx, MnO2, NiO, etc.) have attracted significant atten-
tion owing to their high specific capacitance, excellent reversibility
and long cycle life. Although supercapacitors based on RuO2·nH2O
exhibit a large specific capacitance of as 750 F g−1 in acid elec-
trolyte [9], its high cost and toxicity limit it from commercialization
[10]. Consequently, researchers have been searching an alternative
for RuO2·nH2O all the time. Metal oxide MnO2 with its low cost,
friendly environmental and richness in nature has been investi-
gated to be a promising candidate as supercapacitor material [11].
There are usually two charge/discharge mechanisms in
MnO2-based supercapacitor in aqueous electrolyte. One is interca-
Manganese oxides have been prepared by many different
methods such as hydrothermal synthesis [14,15], sol–gel [16,17],
thermal decomposition [18], coprecipitation [19], electrochemi-
cal deposition [20,21] and so on. Liang and Hwang [9] reported
that a capacitance of 244 F g−1 and good stability was obtained of
hydrous manganese oxide in Na2SO4 solution. Huang et al. [15]
prepared ␤-MnO2 by hydrothermal synthesis and obtained an ini-
tial discharge capacity of 251 mAh g−1 when applied in lithium
battery. Babakhani and Ivey [21] obtained manganese oxide with
rod-like structures and researched its properties as supercapacitor
electrode material, the results showed a capacitance of 185 F g−1
.
Among these methods, the main advantage of hydrothermal
synthesis is that interesting nanostructure (such as nanowires,
nanorods, nanochins and so forth) can be tailored by controlling
the temperature and time of the reaction, but the SC value is lower.
Although sol–gel and coprecipitation methods can gain high SC val-
∗
Corresponding author. Tel.: +86 731 58292229; fax: +86 731 58292477.
E-mail address: liuenhui99@sina.com.cn (E.-h. Liu).
0925-8388/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2010.06.072

556
Y.-j. Yang et al. / Journal of Alloys and Compounds 505 (2010) 555–559
ues materials, the multiple prepared steps are required. It is thin
films electrodes that are usually obtained by electrochemical depo-
sition, which may suffer from poor energy density.
In this paper, we prepared MnO2 through a simple precipitation
technique. Poor crystalline nanostructured MnO2 was synthesized
by reduction of KMnO4 with an organic reductant triethanolamine
and annealed at 300, 350, 400 ◦C, respectively. The structure,
surface morphology, specific surface area and electrochemical per-
formance of prepared products are presented and discussed. To
the best of our knowledge, it has not been reported that using
triethanolamine as reducing agent to synthesis manganese diox-
ide. Furthermore, KMnO4 is the only one starting manganese
precursor in our approach, the organic reductant used here has
the merits of simple steps, less sample consumption and high
yield, which are advantages when considered for commercial
purposes.
2. Experimental
2.1. Synthesis of manganese dioxide material
Fig. 1. XRD patterns of MN100 (a), MN300 (b), MN350 (c) and MN400 (d).
All chemicals were analytical grade and were used without further purification.
All aqueous solutions were prepared using distilled water.
In a typical synthesis process, 1.7 mL triethanolamine was diluted in 100 mL
distilled water followed by 200 mL 0.03 mol L−1 KMnO4 added drop-wise with vig-
orous stirring. After stirring for a period of time at room temperature, the stirring
was terminated and a brown precipitate was obtained. The precipitate was filtered
and washed with distilled water until the pH of the washed water is 7.
The reaction of synthesis MnO2 was displayed below:
Surface morphology images of the synthesized MnO2 materials
were examined by SEM microscope. Fig. 2 depicts the photomi-
crographs of MN100, MN300, MN350 and MN400. As shown in
Fig. 2a, a mass of particles agglomerates together and forms big
conglomerates on the MN100 surface. For the annealed samples,
the macro-agglomerates have been dispersed to a certain degree.
The morphology of MN300 (Fig. 2b) is composed of nanoparticles
with an average diameter of about 11 nm. It is found that the surface
of MN350 (Fig. 2c) is characterized by abundant particles connect-
ing with each other and forming the porous properties. It provides
the locale for cations reaching into MnO2 bulk when the process
of intercalation/deintercalation took place. The diameters of the
nanoparticles ranged from 17–32 nm with an average of 25 nm
are observed for MN400 (Fig. 2d). The heat-treatment changed the
morphology and increased the average diameter of the MnO2 pow-
ders.
N2 adsorption–desorption measurement was employed to
determine the specific surface area and the pore size distribution of
the prepared samples. The specific surface area, total pore volume,
and specific capacitance of synthesized MnO2 powders are indi-
cated in Table 1. The BET surface area was obtained to be 65 m2 g−1
for the MN100. And there is a little decrease in specific surface area
of annealed materials, which are 49, 57 and 48 m2 g−1 for MN300,
MN350 and MN400, respectively. Fig. 3 shows the N2 adsorption
and desorption of samples heat-treated at different temperature.
For MN300 and MN350, the N2 adsorption and desorption curves
form hysteresis loops, exhibiting a H3-type of IV-type according to
IUPAC, which is characteristic of mesoporous solids. For MN100
and MN400, the fraction of mesopores is not so large because
the hysteresis loops are not so clear and with less amount of N2
absorbed [22]. Pore size distribution curves of MN100, MN300,
MN350 and MN400 are shown in Fig. 4. According to IUPAC nomen-
clature, micropores are less than 2 nm in diameter, mesopores are
between 2 and 50 nm in diameter, and macropores are greater than
50 nm in diameter. For MN100, the peaks with mean pore widths
of around 1.3–4 nm some around 10 nm were emerged (Fig. 4).
As temperature increasing, the peaks of MN100’s pore size dis-
tribution curve changed to increase. The enhanced peaks may be
attributed to exposure of new pores that were occupied by water
and some small organic molecules before heat-treatment. And for
MN300 and MN350, a broad distribution was found in the meso-
pore region, and both of the two curves have low broad peaks
centered at about 10 nm. But due to the point-to-point nature of
the experimental data in Fig. 3, the curves in Fig. 4 are not smooth.
4KMnO4 + N(CH2CH2OH)3 → 4MnO2 + K2C2O4 + 2CH3COOK + NH3 + 3H2O
KMnO4, the only source here containing manganese and oxidizing agent, was
reduced by triethanolamine, which lead to the formation of manganese dioxide.
Then the precipitate was dried at 100 ◦C for 24 h. Subsequently, part of the pow-
der was annealed 2 h at 300, 350 and 400 ◦C, respectively. The samples were denoted
as MN100, MN300, MN350, and MN400 in the following paragraphs.
2.2. Characterization
The XRD was employed to study the crystallographic characteristics of syn-
thesized samples by a diffractometer (D/MAX-3C) with Cu K␣ (ꢀ = 1.5406 Å) as
radiation. SEM studies were examined on a LEO 1525 microscopy to investigate
the surface morphology of the prepared materials. The surface area of MnO2 was
studied by BET measurement by nitrogen gas adsorption–desorption method at 77 K
using a NOVA2200 (Quantachrome) automatic adsorption unit.
2.3. Preparation of the electrodes
The electrode consisted of 80% active material (MN100, MN300, MN350,
or MN400) powder, 10% acetylene black (AB) as conducting agent, and 10%
polyvinylidene fluoride (PVDF) as a binder. First, PVDF was dissolved in N-methyl-
2-pyrrolidone (NMP), then AB and active material were mixed orderly to form a
homogeneous slurry, which was subsequently blush-coated onto nickel foam cur-
rent collectors (˚ = 10 mm) and dried at 100 ◦C for 12 h to fabricate electrodes. Mass
of the active materials of each electrode was about 10 mg.
All the electrochemical measurements were completed in 1.0 mol L−1 Na2SO4
electrolyte, and electrochemical characteristics were carried out at CHI660A elec-
trochemical workstation (CH Instruments, USA).
3. Results and discussion
3.1. Material characteristics
Power XRD was employed to analyze the structures. Fig. 1 shows
the XRD patterns of MN100, MN300, MN350 and MN400. Fig. 1a–c
are similar and only a few broad peaks are appeared. Those broad-
ening peak features indicate the amorphous nature of MN100,
MN300 and MN350. Broad peaks at 2ꢁ = 36.9◦ and 66.3◦ are present.
There is a transformation in crystallinity when post-treat tempera-
ture reached up to 400 ◦C with the appearance of clear sharp peaks.
As can be see from Fig. 1d, the sharp peaks at 2ꢁ = 12.7◦, 18.0◦,
28.6◦, 37.5◦, 41.9◦, 49.7◦, 60.1◦, and 69.3◦ corresponding to ␣-MnO2
matched with JCPDS 44-0141 [6,12].

Y.-j. Yang et al. / Journal of Alloys and Compounds 505 (2010) 555–559
557
Fig. 2. SEM images of MN100 (a), MN300 (b), MN350 (c), and MN400 (d).
Table 1
Specific surface area, total pore volume and specific capacitance of prepared MnO2.
3.2. Electrochemical studies
The cyclic voltammetry (CV) is usually considered to be a
powerful tool to measure whether the material is suitable for
supercapacitor or battery electrode. Fig. 5a shows the CV curves
of prepared MnO2 electrodes. The CV curves were measured
in 1.0 mol L−1 Na2SO4 electrolyte and a potential range from 0
to 1.0 V at a scan rate of 2 mV s−1. The rectangular CV pro-
file is a fingerprint indicating ideal capacitive behavior [24]. For
MN100, MN300 and MN350, the voltammograms are approxi-
mately rectangular in shape. Among these samples, the highest
current density is obtained for MN350. The CV curve of MN100
mingles with MN300, implying the SC values of MN100 and
MN300 are similar. While the CV shape of MN400 shows a
Sample
Specific surface
Total pore
Specific capacitance
area (m2 g−1
)
volume (cc g−1
)
(F g−1
)
MN100
MN300
MN350
MN400
65
49
57
48
0.172
0.166
0.192
0.116
169
182
251
153
And when temperature reached up to 400 ◦C, pore size distribu-
tion was scattered and surface area decreased, the reason may be
due to that the pores collapsed during the high-temperature heat-
treatment [23].
Fig. 4. Pore size distribution of MN100, MN300, MN350 and MN400.
Fig. 3. N2 adsorption–desorption results of MN100, MN300, MN350 and MN400.

558
Y.-j. Yang et al. / Journal of Alloys and Compounds 505 (2010) 555–559
Fig. 5. (a) CV curves of prepared MnO2 electrodes at a sweep rate of 2 mV s−1 in
1.0 mol L−1 Na2SO4 electrolyte. (b) Effect of post-treat temperature and CV scan rate
on specific capacitance of MnO2 electrodes.
Fig. 6. (a) Galvanostatic charge–discharge curves of prepared MnO2 electrodes
in 1.0 mol L−1 Na2SO4 electrolyte between 0 and 1.0 V at a current density of
500 mA g−1. (b) Charge–discharge cycles of MN350 electrode at a current density
of 200 mA g−1 in 1.0 mol L−1 Na2SO4 solution.
deformed rectangle in the potential range studied. The current
density of discharging is lower than that of charging at 1 V
or so. The specific capacitance is calculated by integrating the
area of the CV curves. The best result of specific capacitance
value is 251 F g−1 that is obtained from the sample of MN350
at a scan rate of 2 mV s−1. The value is comparable to that of
previous reported result that powder MnO2-based electrode capac-
itance is 150–300 F g−1 [25]. And the specific capacitances of
MN100, MN300 and MN400 are 169, 182 and 153 F g−1, respec-
tively.
Fig. 5b shows the specific capacitance values of the samples
at different CV scan rates. It is clearly observed that there is a
decrease in SC values with an increase in sweep rate. For instance,
the SC value of MN350 is 251 F g−1 at a scan rate of 2 mV s−1 and
145 F g−1 retains when the scan rate increases to 100 mV s−1. That
may be due to a loss in efficiency of utilization of active material
at high sweep rate. Because there is no enough time for cations
reaching into MnO2 matrix, only reaching the outer surface at
a high scan rate. Particular in a redox supercapacitor, the effec-
tive use of the material is directly bound up with the capacitance
behavior.
capacitance is calculated from the formulation:
It
C =
× 2
mꢂV
where
C
is the specific capacitance (F g−1),
I is the charg-
ing/discharging current (A), t is the charging/discharging time (s),
m is the mass of one electrode active material (g) and ꢂV is the
potential range. The SC values calculated from the above formula-
tion for MN100, MN300, MN350, and MN400 are147, 175, 221 and
110 F g−1 at a current density of 500 mA g−1, respectively, which
approximately match with the results calculated from CV test.
Fig. 6b shows the first several charging–discharging cycles for
MN350 at a current density of 200 mA g−1 in 1.0 mol L−1 Na2SO4
solution. The charging time and discharging time are almost the
same indicating a high charging–discharging efficiency. The spe-
cific capacitance calculated here is 247 F g−1 at a current density
of 200 mA g−1 and there is hardly any decay in capacitance after
several cycles.
From these different specific capacitance values, we can find that
the post-treat temperature, which changed the crystalline struc-
ture, surface morphology, specific surface area value and the pore
size distribution of the material, is an important factor to the mate-
rial in electrochemical performance. There has been an increase in
capacitance with the increase in annealed temperature, while the
capacitance decreased when the temperature exceeds 350 ◦C. It is
the advantage that large pores in mass transport and the cations
in the electrolyte can transport in the macropores and meso-
Galvanostatic charging–discharging plots of the MnO2 elec-
trodes measured at a current density of 500 mA g−1 in a 1.0 mol L−1
Na2SO4 electrolyte are shown in Fig. 6a. The linear and symmetric
characteristic is another identification indicating ideal capacitance
behavior for electrode materials. It can be seen from Fig. 6a, for
all the electrodes, the variation of potential with time is approxi-
mately linear in both charging and discharging process. The specific

Y.-j. Yang et al. / Journal of Alloys and Compounds 505 (2010) 555–559
559
frequency region, the slopes of imaginary parts of the impedance
plots, associated with the capacitive behavior, increased with the
annealed temperature. For an ideal capacitor, the impedance plots
should be a vertical line parallel to the imaginary axis. From Fig. 6,
the plot of MN350 is almost perpendicular to the real axis, indicat-
ing lower diffusive resistance for intercalation/deintercalation of
cations than that of others. The EIS results reveal the better prop-
erties of MN350 that are in step with the CV results. Although the
plot of MN400 nearly coincides with MN350, the electrochemical
active sites would decrease when the heat-treatment temperature
≥400 ◦C. That may be another reason for the decreased specific
capacitance of MN400. The result is in agreement with Sang-Eun
Chun and his co-workers [29] who researched the effect of heat-
treatment to the mechanism of charging/discharging in neutral
solution by impedance spectroscopy. All the results are indicative
of MN350 suitable for electrode material.
4. Conclusions
Fig. 7. Impedance plots of electrodes at 1.0 mol L−1 Na2SO4 electrolyte, frequency
Amorphous MnO2 powders have been successfully prepared
by reaction of potassium permanganate with an organic reduc-
tant triethanolamine. And one of the advantages of this method
is that KMnO4 is the only source containing manganese. The effect
of heat-treatment was studied. The XRD results show a crystalline
convert to ␣-MnO2 from amorphous MnO2 when annealed at
400 ◦C. N2 adsorptions and desorption studies show higher specific
surface area and total pore volume and wider pore size distri-
bution for sample annealed at 350 ◦C. MnO2 annealed at 350 ◦C
exhibits a better specific capacitance of 251 F g−1 in 1.0 mol L−1
Na2SO4 electrolyte, which increased 48.5% compared to MN100.
The results show that the pseudo-capacitance is obtained depend-
ing on the intercalation/deintercalation of cations into the porous
bulk of MnO2 annealed at 350 ◦C. Good cycle stability characteris-
tic for MnO2 annealed at 350 ◦C is also obtained by galvanostatic
charging–discharging test.
range from 105 to 0.01 Hz (inset is a local enlargement).
pores more easily and rapidly than in micropores. For the whole
charge storage process of MnO2 involving transport and insertion
of cations in the electrolyte, large pore size and a wide pore size dis-
tribution in the mesopore region insure that most surface of MN300
and MN350 are immediately accessible for cations. As for the spe-
cific capacitance of MN300 lower than MN350, the only reason is
the slightly higher specific surface area for MN350. For MN100
and MN400, the existence of micropores goes against the access
for cations and may slow down the transport of cations. Although
the specific surface area of MN350 (57 m2 g−1) is little lower than
MN100 (65 m2 g−1), the pore distribution is wider and total pore
volume is higher than that of MN100. That is key factors when
discussing the electrochemical properties [6]. And the MN350 pro-
vides more advantageous porous structure for deintercalation upon
oxidation and intercalation upon reduction process of cations into
the MnO2 matrix. The BET surface area value is not the only fac-
tor influenced the electrochemical properties, so the capacitance
does not absolutely follow the trend of the BET surface area value.
Similar results that higher specific capacitance with lower specific
surface area were obtained in the literature [6,26,27]. The alter-
ation in crystalline structure by high-temperature heat-treatment
may lead to the variation of physical or chemical properties and
cause the change in specific capacitance [28]. Lower specific sur-
face area value and total pore volume and more narrow pore size
distribution were obtained for MN400 compared to MN350.
In order to further understand the electrochemical properties of
the MnO2 electrodes, EIS was test in 1.0 mol L−1 Na2SO4 electrolyte.
Fig. 7 displays the Nyquist plots for the prepared MnO2 materi-
als performed in a frequency range from 105 to 0.01 Hz. It can be
observed from Fig. 6 that the impedance plots of all the electrodes
are consist of a distorted semi-circle at the high-frequency region
and an almost vertical linear spike at the low-frequency region.
At the high-frequency region, the X-intercept yields the internal
resistance Rs, while the diameter of the semi-circle provides the
charge-transfer resistance (Rct) of the electrode/electrolyte inter-
face. Since we used the same electrolyte, the Rs values of MN100,
MN300, MN350 and MN400 are similar to each other, which are
1.9, 1.7, 1.5 and 1.6 ꢃ, respectively. The charge-transfer resistance
of MN100 (0.68 ꢃ) is higher than that of annealed materials, and
it is noticeable that Rct decreased as the post-treat temperature
increased, following the order: MN300 > MN350 ∼ MN400. It can
be concluded that heat-treatment can reduce the charge-transfer
resistance, which is advantageous for electrodes. At the low-
References
[1] H. Adelkhani, M. Ghaemi, J. Alloys Compd. 493 (2010) 175–178.
[2] J. Li, E.H. Liu, W. Li, X.Y. Meng, S.T. Tan, J. Alloys Compd. 478 (2009) 371–374.
[3] E. Frackowiak, V. Khomenko, K. Jurewicz, K. Lota, F. Beguin, J. Power Sources
153 (2006) 413–418.
[4] W.Y. Kima, M.Y. Kang, J.H. Yi, et al., J. Power Sources 195 (2010) 2125–2129.
[5] H.F. An, Y. Wang, X.Y. Wang, et al., J. Power Sources 195 (2010) 6964–6969.
[6] V. Subramanian, H.W. Zhu, R. Vajtai, P.M. Ajayan, J. Phys. Chem. B 109 (2005)
20207–20214.
[7] R. Liu, S-I.I. Cho, S.B. Lee, Nanotechnology 19 (2008) 215710, 8.
[8] L.M. Li, E.H. Liu, J. Li, Y.J. Yang, J. Power Sources 195 (2010) 1516–1521.
[9] C.H. Liang, C.S. Hwang, J. Alloys Compd. 500 (2010) 102–107.
[10] Y.J. Yang, E.H. Liu, L.M. Li, J. Alloys Compd. 487 (2009) 564–567.
[11] D.P. Dubal, D.S. Dhawalea, C.D. Lokhandea, et al., J. Alloys Compd. 484 (2009)
218–221.
[12] M. Toupin, T. Brousse, Chem. Mater. 16 (2004) 3184–3190.
[13] P. Ragupathy, H.N. Vasan, J. Electrochem. Soc. 155 (2008) A34–A40.
[14] Y. Chen, Y.Z. Hong, Y.P. Ma, et al., J. Alloys Compd. 490 (2010) 331–335.
[15] X.K. Huang, D.P. Lvb, Q.S. Zhang, et al., Electrochim. Acta 55 (2010) 4915–4920.
[16] S.L. Zhao, H.Y. Chen, J. Alloys Compd. 474 (2009) 473–476.
[17] H.J. Guo, X.H. Li, Z.X. Wang, W.J. Peng, X. Cao, H.F. Li, J. Power Sources 189 (2009)
95–100.
[18] L.L. Zhang, T.X. Wei, Micropor. Mesopor. Mater. 123 (2009) 260–267.
[19] F.J. Liu, J. Power Sources 182 (2008) 383–388.
[20] K.W. Nam, C.W. Lee, J. Power Sources 188 (2009) 323–331.
[21] B. Babakhani, D.G. Ivey, J. Power Sources 195 (2010) 2110–2117.
[22] X.L. Hong, G.Y. Zhang, Y.Y. Zhu, Mater. Res. Bull. 38 (2003) 1695–1703.
[23] S. Devaraj, N. Munichandraiah, J. Phys. Chem. C 112 (2008) 4406–4417.
[24] S.B. Ma, Y.H. Lee, K.Y. Ahn, C.M. Kim, K.H. Oh, J. Electrochem. Soc. 153 (2006)
C27–C32.
[25] O. Ghodbane, J.L. Pascal, F. Favier, Appl. Mater. Interfaces 1 (2009) 1130–1139.
[26] R.N. Reddy, R.G. Reddy, J. Power Sources 124 (2003) 330–337.
[27] R.N. Reddy, R.G. Reddy, J. Power Sources 132 (2004) 315–320.
[28] J.K. Chang, Y.L. Chen, W.T. Tsai, J. Power Sources 135 (2004) 344–353.
[29] S.E. Chun, Su-I.I. Pyun, G.J. Lee, Electrochim. Acta 51 (2006) 6479–6486.