The sol-gel-derived nickel-cobalt oxides with high supercapacitor performances

Article abstract of DOI:10.1149/1.3574021

Nickel cobalt (Ni-Co) oxides with various Ni/Co ratios are synthesized by using a sol-gel process. Electrochemical performance, microstructure, morphology and BET surface area are found strongly related to Ni and Co concentration. Increasing Co concentration changes the oxide microstructure from NiO crystal dominate structure (Ni:Co 1:1 and 1:2) to Co₃O₄ dominate structure (Ni:Co 1:4 and 0:1). A maximum specific capacitance of 1539 Fg ^-1 was obtained for Ni-Co (Ni:Co 1:2) oxide at a current density of 1Ag^-1, and this capacitance is similar to that of RuO₂. A systematic study shows that the Ni-Co oxide (1:2) has a mesoporous structure with a high BET surface area of ～ 315 m² g^-1 and porous size of ～ 4.7 nm, which is favorable for the charge/dischargeprocess of a supercapacitor.

Full text of DOI:10.1149/1.3574021

Journal of The Electrochemical Society, 158 (6) A695-A699 (2011)
0
A695
013-4651/2011/158(6)/A695/5/$28.00 VC The Electrochemical Society
The Sol-Gel-Derived Nickel-Cobalt Oxides with High
Supercapacitor Performances
a
Guangxia Hu, Chunhua Tang, Chunxiang Li, Huimin Li, Yu Wang, and Hao Gong
a
a
b
a
a,z
a
Department of Materials Science and Engineering, National University of Singapore, 119260 Singapore
Advanced Materials Technology Centre, Singapore Polytechnic, Singapore 139651
b
Nickel cobalt (Ni-Co) oxides with various Ni/Co ratios are synthesized by using a sol-gel process. Electrochemical performance,
microstructure, morphology and BET surface area are found strongly related to Ni and Co concentration. Increasing Co concentra-
tion changes the oxide microstructure from NiO crystal dominate structure (Ni:Co ¼ 1:1 and 1:2) to Co
3
O
4
dominate structure
ꢀ1
(
Ni:Co ¼ 1:4 and 0:1). A maximum specific capacitance of 1539 Fg was obtained for Ni-Co (Ni:Co ¼ 1:2) oxide at a current
ꢀ
1
2
density of 1Ag , and this capacitance is similar to that of RuO . A systematic study shows that the Ni-Co oxide (1:2) has a meso-
2
ꢀ1
porous structure with a high BET surface area of ꢁ 315 m g and porous size of ꢁ 4.7 nm, which is favorable for the charge/
discharge process of a supercapacitor.
VC 2011 The Electrochemical Society. [DOI: 10.1149/1.3574021] All rights reserved.
Manuscript submitted October 22, 2010; revised manuscript received March 9, 2011. Published April 13, 2011.
1
12
Supercapacitors, also known as electrochemical capacitors, have
thousands times higher capacitance than conventional electrolytic
capacitors, and much higher power density than batteries and fuel
cells. They can complement batteries in electrical energy storage
and harvesting applications, when high power delivery or uptake is
needed. They can even replace batteries, with the great advantage of
much shorter charging time, when the energy density or specific ca-
tance of ꢁ 840 Fgꢀ . More recently, Wei et al. prepared a spinel
nickel cobaltite (NiCo O ) aerogel with the highest specific capaci-
2
4
ꢀ
1
tance of ꢁ 1400 Fg after annealing the as-prepared sample at
ꢃ
200 C for 5 h. However, the synthesis and processing methods
reported so far are not practical for commercial production because
they are either slow in the growth of very thin layer on thick sub-
strates or expensive for large scale industrial production. Furthermore,
a comprehensive understanding of the relationship between the elec-
trochemical behavior and the microstructure of the binary Ni-Co
oxides for supercapactors is still lacking. Efforts are necessary to find
more suitable syntheses methods for cheap and nanostructure Ni-Co
oxides with a high capacitor performance. Furthermore, it is neces-
sary to study the microstructure and performance of Ni-Co oxide
composites with different Ni/Co ratios in depth.
In this work, Ni-Co oxides are prepared by a sol-gel method that
can be suitable for large scale commercial production. It will be
shown that the sol-gel process enables to prepare mesoporous oxides
with high specific surface area and 3D network having an average
pore size of several nanometers. The effect of Ni/Co molar ratio on
the crystalline structure and capacitive behaviors of the oxides are
investigated. Ni-Co oxide with a high specific surface area (maxi-
1
pacitance is significantly increased. According to energy storage
mechanisms, supercapacitors can be divided into two types: electro-
chemical double-layer capacitors (EDLC) and pseudo-capacitors. In
the former, energy storage arises mainly from the separation of elec-
tronic and ionic charges at the interface between the electrode mate-
rials, typically carbon, and the electrolyte solution. In the latter, fast
Faradic reactions take place at the electrode materials, typically
metal oxides, at characteristic potentials like in batteries, and give
rise to what is called pseudo-capacitance. Although the well devel-
ꢀ
1
oped carbon materials (ꢁ 200 Fg ) for EDLC are commercially
available, other materials with higher capacitance are always desira-
ble. Transition metal oxides with various oxidation states are very
2
promising candidates for the next generation high capacitance
supercapacitors. They can have high pseudo-capacitance, high
power density, energy density, mass density and cycle stability.
2
ꢀ1
mum 315 m g ) is achieved when the Ni/Co molar ratio is 1:2.
The resulting Ni-Co oxide exhibits a high maximum specific capaci-
tance of 1539 F g under constant current discharge (current density
x
Among these transition metal oxides, ruthenium oxide (RuO ) with
ꢀ
1
a specific high capacitance up to 1580 Fg (Ref. 3) is widely stud-
ꢀ1
4
ied in the past 30 years. However the high cost of RuO limits it
ꢀ1
at 1 Ag ) in 1 M KOH electrolyte.
x
from commercialization except in some special applications. It is of
great interest in having cheaper metal oxides for high capacitance
supercapacitors.
Experimental
Ni-Co oxides preparation.— All the chemicals used are of ana-
lytical grade without further purification. An aqueous solution of
To find cheaper metal oxides, various transition metal oxides
have been investigated, such as Co O , MnO , NiO, V O , etc.
3
4
2
2 5
Na
3
C
6
H
5
O
7
ꢂ2H
2
O, NaOOCCH
3
, NiSO
4
ꢂ6H
2
O and CoSO
4
2
ꢂ7H O
Among them, nickel oxide has shown very high specific capaci-
tance, low cost, low toxicity and environment friendliness, but with
relatively low cyclic reversibility. Cobalt oxide, on the other hand,
(
Ni:Co in solution, Sample C:1:2 cationic ratio of Ni:Co in solution,
Sample A: pure NiSO ꢂ6H O, Sample B: 1:1 cationic ratio of
4
2
5
Sample D:1:4 cationic ratio of Ni:Co in solution, Sample E: pure
ꢃ
has high redox activity and good reversibility, but the specific ca-
pacitance is relatively low. Nowadays because of more and more
6
CoSO
4
ꢂ7H
2
O) was kept at 80 C in water bath under constant stir-
ring. 3M KOH was added dropwiselyinto the solution until the pH
advantages of Nickel and Cobalt, both of them attracted more atten-
tion. For instance, Zhang and Fang found a general strategy to pre-
pare Pt 3d-transition metal (Co, Ni) nanotubes for use in fuel cells.
ꢃ
value reached 12. The solution was kept at 80 C for 2 h, and then at
room temperature for 24 h. The resulting suspension was centri-
fuged and rinsed repeatedly with deionized water and ethanol. The
7
Meanwhile previous research results indicated that doping Co could
enhance NiO electrochemical reversibility and conductivity. Hu et
ꢃ
8
gel was then heated to 70 C and held there for 12 h, and then
ꢃ
9
al. prepared hydrous Ni-Co oxide with an amorphous structure
grinded into fine powders and kept at 200 C in air for 12 h. The
obtained Ni-Co oxides for various cationic ratio Ni/Co solutions are
referred to as Ni-Co oxides (the cationic ratio of Ni/Co in solution);
for example, Ni-Co oxide (1:2) indicates that the Ni-Co oxide com-
posite is prepared with a 1:2 cationic ratio of Ni:Co in the solution.
[
denoted as a-(Co þ Ni)(OH)
2
ꢂnH
2
O] by using anodic deposition
ꢀ
1
method and obtained a specific capacitance of ꢁ 730 Fg . He
1
et al. and Liu et al. prepared Ni-Co oxide/Al-layered double
0
11
hydroxides composite composites, Ni-Co oxide/TiO nanotube com-
2
posites and Ni-Co oxide/carbon nanotube composites by using
anodic deposition method and obtained the highest specific capaci-
Electrode preparation.— 70 wt % of the prepared active Ni/Co
oxide powder, 25 wt % of acetylene black, and 5 wt % of polytetra-
fluoroethylene (PTFE) were mixed together in ethanol, and ultra-
sonic was used to ensure uniform mixing. Then the mixed slurry
z
E-mail: msegongh@nus.edu.sg
Downloaded on 2013-11-22 to IP 128.230.168.102 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

A696
Journal of The Electrochemical Society, 158 (6) A695-A699 (2011)
was spread onto a piece of Ni foam of weight (m
ꢁ
1
) within an area of
2
0.25 cm . The electrode was then dried at 70 C in air for 2 h,
ꢃ
ꢃ
pressed at 8 Mpa, kept at 140 C in air for 12 h, and weighed (m ).
2
The weight of the active material was calculated by (m
which was about 4–6 mg for each of our samples.
2
-m
1
) ꢄ 0.7,
Characterization and electrochemical tests.— Transmission elec-
tron microscopy (TEM, JEOL 2000FX) with energy dispersive
X-ray (EDX) analyzer was used to characterize the crystal structure
and composition of the materials. The surface morphology was
determined by scanning electron microscope (SEM, Zeiss SUPRA
4
0). Specific surface area, pore volume and width were measured by
nitrogen adsorption/desorption using Brunauer-Emmet-Tell (BET,
Micromeritics ASAP2020) method. All electrochemical measure-
ments were carried out in a three-electrode arrangement with the
prepared electrode as working electrode, a platinum plate as counter
electrode and a saturated calomel electrode (SCE) as reference elec-
trode in 1 M KOH electrolyte. Electrochemical performance was
evaluated by cyclic voltammetry (CV) and galvanostatic charge/dis-
charge tests (Solartron S1 1287).
Results and Discussion
Figure 1. (Color online) XRD patterns of pure NiO, Ni-Co oxdie (1:1), Ni-
Co oxdie (1:2), Ni-Co oxide (1:4), and pure Co
3 4
O .
Characterization of the Ni-Co oxides.— The molar ratios of Ni/
Co in these Ni-Co oxides were determined by EDX. The results are
shown in Table I. It is noted that for Ni-Co oxide samples 1:1, 1:2
and 1:4, the actual Ni/Co molar ratios are 1:0.8, 1:1.4 and 1:4.3,
respectively.
NiO (a ¼ b ¼ c ¼ 0.4176 nm) and cubic structure Co
O
3 4
(a ¼ b ¼ c ¼ 0.8085 nm), the radii of 111, 200, 220 and 311 rings of
NiO are only 1% different from 311, 400, 440 and 533 rings of
Co O , respectively. We notice that the 111, 200, 220 and 311 NiO
3 4
rings for Ni-Co oxides (1:2) shown in Fig. 2c are slightly brighter
than those rings of NiO (Fig. 2a), which may suggest the co-exis-
Figure 1 shows XRD patterns of five samples. From the XRD
pattern, we can find that it is Co
3
O
4
structure (PDF#09-0418: space
group Fd-3m, Cubic, lattice constant 0.8084 nm) for the Ni:Co ¼ 0:1
sample. Meanwhile the diffraction peaks of the Ni-Co oxides (1:4)
tence of 311, 400, 440 and 533 rings of Co
terns. The 311, 400, 440 and 533 rings are slightly brighter in Fig.
2d for Ni-Co oxides (1:4) than those of Co (Fig. 2e), which may
suggest the coexistence of the 111, 200, 220 and 311 NiO rings in
Fig. 1c. Therefore, the possibility of the existence of Co crystal
3 4
O in the diffraction pat-
also indicate that it has the similar Co
Ni-Co oxides (1:2), the diffraction peaks are strongly different from
the Co structure, and they can be indexed as NiO phase
PDF#65-2901, space group Fm-3m, Cubic, lattice constant 0.4195
nm). For the Ni-Co oxides (1:1), there are inconspicuous humps
instead of obvious peaks. NiO and Ni(OH) (PDF#02-1112, space
3 4
O structure. However for the
O
3 4
3 4
O
(
O
3 4
in Ni-Co oxide (1:1 and 1:2) and NiO crystal in Ni-Co oxide (1:4)
cannot be excluded.
2
group P-3m1, hexagonal, lattice constants a ¼ b ¼ 0.311 nm and
c ¼ 0.466 nm) peaks fall in the ranges of the humps. The same result
could be obtained from the XRD pattern of the Ni:Co ¼ 1:0 sample.
Additionally the sharp peaks of the XRD pattern of pure Ni sample
may be caused by impurity noises. In order to further understand the
structure of these oxides, we also take TEM images and diffraction
patterns.
The dark field TEM images, which were taken using the smallest
objective aperture at a small portion of the strongest diffraction pat-
tern rings, reveal that the sizes of bright patches in Figs. 2b–2d are
smaller than those in Figs. 2a and 2e. Because one diffraction spot
corresponds to a series parallel crystal planes, the size of a bright
patch in the dark-field image can represent the size of a crystallite
with straight lattice planes. Such results suggest that crystallite sizes
Figure 2 shows electron diffraction patterns and TEM images of
the five samples. The diffraction rings of the corresponding crystal
planes are indexed in the patterns. The ring patterns demonstrate
polycrystalline nature of the five samples. The selected area diffrac-
tion pattern of Ni:Co ¼ 1:0 sample correspond to NiO. For samples
of Ni-Co oxides (1:1 and 1:2), the diffraction patterns are very simi-
lar to that of NiO, and the rings belong to NiO (111), (200), (220),
in the Ni-Co oxides are smaller than those in Ni(OH)
Co samples, which may have effect on supercapacitor
performance.
2
/NiO and
O
3 4
Electrochemical Testing of Ni-Co oxides.— Cyclic voltammetry
tests within a 0.0–0.5V potential range (vs. SCE) at a scan rate of
ꢀ1
1
mV s were performed in 1M KOH electrolyte at room tempera-
ture. Figure 3 shows the CV curves of the five samples. A pair of re-
dox peaks can be observed in each CV curve, which indicates a
pseudo-capacitance characteristic. The CV curve of pure NiO shows
two obvious peaks. The cathodic peak occurs at around 0.36 V and
(
311) and (222). The diffraction pattern of Ni-Co oxide (1:4) is close
to that of Co , and the rings appear belonging to Co (111),
220), (311)/(222), (400), (422), (511), (440), (620)/(533)/(622)/
444), (642). These results are in agreement with the XRD results.
3
O
4
3 4
O
(
(
the anodic peak occurs at around 0.22 V. These two peak positions
are in consistent with those of Ni(OH) /NiO reported previously
and the corresponding Faradic redox reaction is
Based on theoretical calculation of the ring radius of cubic structure
13
2
Table I. Ni/Co Molar ratio, specific surface area, pore width and
cumulative pore volume of five samples Ni-Co oxides 1:0, 1:1, 1:2,
NiO þ OH $ NiOOH þ eꢀ
ꢀ
[1]
[2]
1
:4 and 0:1.
ꢀ
ꢀ
NiðOHÞ þ OH $ NiOOH þ H2O þ e
2
Ni/Co Molar Ratio in solution
Ni/Co Molar Ratio in sample
1:0
1:0
1:1
1:2
1:4
0:1
0:1
3 4
For pure Co O , a pair of broad and less obvious peaks can be seen
at around 0.39 and 0.35 V, in agreement with those of Co
reported in literature and can be attributed to the Faradic redox
reaction
3 4
O
1:0.8 1:1.4 1:4.3
14
2
Specific Surface Area (m g
ꢀ1
)
377.68 329.7 315.10 99.42 61.26
29.80 31.69 47.65 41.28 55.76
˚
Average Pore Width (A)
Cumulative Pore Volume
ꢀ1
0.32
0.38
0.55
0.17 0.12
3
cm g
ꢀ
ꢀ
(
)
Co3O4þOH þH2O $ 3CoOOH þ e
[3]
Downloaded on 2013-11-22 to IP 128.230.168.102 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

Journal of The Electrochemical Society, 158 (6) A695-A699 (2011)
A697
Figure 3. (Color online) Cyclic voltammograms of pure NiO, Ni-Co oxide
(
KOH solution at a scan rate of 1 mVs
1:1), Ni-Co oxide (1:2), Ni-Co oxide (1:4), and pure Co O recorded in 1 M
3 4
ꢀ1
.
impossible to find experimental evidence to reveal the reasons for
the phenomena observed, the experimental observations and the
proposed possible reasons may be good starting points in further
research for a deeper understanding of such a materials system and
the electrochemical behavior. It is further interesting to notice that
the curves of Ni-Co oxides, in comparison with the CV curve of
Ni(OH) /NiO, are closer to rectangular shape that is the characteris-
2
tic shape of carbon and the expensive RuO supercapacitors. Fur-
2
thermore, the area enclosed by the CV curve is much bigger than
that of Co O . Such results suggest that the Ni-Co oxides are more
3 4
favorable for supercapacitor application. The exact capacitance val-
ues are determined from discharging curve shown below.
Figure 4 shows the charge/discharge curves at a current density
ꢀ
1
of 1 A g within the potential window of 0.0–0.4 V in 1 M KOH
electrolyte for the five samples. When the electrode starts to dis-
charge, a potential drop can be seen from the curves. Such a phe-
nomenon is due to the electrode polarization at a high current den-
sity. The specific capacitance can be calculated using Eq. 4
I ꢄ Dt
C ¼
[4]
mDV
Figure 2. Electron diffraction patterns and TEM images of (a, a’) pure NiO;
b, b’) Ni-Co oxide (1:1); (c, c’) Ni-Co oxide (1:2); (d, d’) Ni-Co oxide (1:4)
and (e, e’) pure Co
(
3 4
O .
In the CV curves of the three Ni-Co oxides (1:1, 1:2, and 1:4), the
redox peaks are broadened with an increase in Co, and both anodic
and cathodic peaks shift to the negative direction. It is better to
admit that we are still unable to provide an explanation on the
broadening and left shift of the redox peaks, however, such phenom-
ena may suggest that the samples are not a simple mixture of NiO
3 4
and Co O . Changes in valence of metal ions or/and oxygen content
in the compound, which cannot be obtained from XRD and TEM,
may also be important to cause redox peaks shift and broadening.
However, XPS characterization of our samples does not show
obvious difference in the peak positions for Co and Ni atoms/ions
for different samples, therefore, XPS results cannot reveal the cause
of the redox peak shift and broadening. Another possible reason for
the redox peak shift and broadening may be because Ni is partly
substituted by Co with different valence number so that the elec-
tronic state and the electrochemical properties of oxide materials are
changed. However, to determine the substitution is not an easy task
and we are unable to perform such a measurement. Although it is
Figure 4. (Color online) The charge/discharge curves of Ni(OH)
2
/NiO, Ni-
Co oxide (1:1), Ni-Co oxide (1:2), Ni-Co oxide (1:4), and pure Co O
3
4
ꢀ1
recorded in 1 M KOH solution at a current density of 1 Ag
.
Downloaded on 2013-11-22 to IP 128.230.168.102 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

A698
Journal of The Electrochemical Society, 158 (6) A695-A699 (2011)
ꢀ
1
where C (Fg ) is specific capacitance, I (A) is discharge current, Dt
s) is discharge time, m(g) is mass of the active electrode material,
(
and DV (V) is potential change during discharge (set as 0.4 V). The
specific capacitance is calculated to be 1289, 1410, 1539, 1060 and
ꢀ
61 F g for Ni(OH) /NiO, Ni-Co oxide (1:1), Ni-Co oxide (1:2),
2
1
3
Ni-Co oxide (1:4), and pure Co O , respectively. Such results are in
3
4
consistent with the IV curves shown in Fig. 3. When the ratio of Ni
and Co is 1:2, we can obtain the maximum specific capacitance
ꢀ
1
value 1539 F g , which is much closer to the value of RuO
x
. How-
ever based on Eqs. 1 and 3, the theoretical specific capacitance is
calculated to be 3229 and 1001 F g for NiO and Co O , respec-
ꢀ
1
3
4
tively. Therefore the specific capacitance of Ni-Co oxide should
reduce with increasing Co concentration. However, the experiment
results show that the Ni-Co oxide (1:2) has the highest specific ca-
pacitance among the five samples, which will be discussed further
later. Also interestingly, the discharge curves become more symmet-
rical to their charge curves and more linear with an increment of Co
in the Ni-Co oxide. The linear discharging curve is characteristic of
4
supercapacitor behavior in contrast to the battery behavior. For
sample Ni-Co oxides 1:2 and 1:4, the CV curve shape is close to
those of carbon and RuO , and the specific capacitance is much
2
Figure 6. (Color online) BJH adsorption pore distribution of samples with
Ni:Co equal to 1:0 (Ni(OH) /NiO), 1:1, 1:2, 1:4 and 0:1 (Co ).
2
3 4
O
2
higher than that of carbon and similar to that of RuO . Due to the
high specific capacitance and the much lower price of Ni and Co
than Ru, the Ni-Co oxide supercapacitor has its unique advantage
and may be a very promising candidate for the next generation of
commercial high capacitance supercapacitors.
SEM images show porous morphologies of these samples. For
the Ni:Co ¼ 1:0 NiO sample (Fig. 5a), we can find that most of
pores are mesopores, which can possibly be one reason why
insertion and extraction, and thus increases the material utilization
in supercapacitor functioning. Pore size distribution maximum shifts
to a lager pore size with increasing Co concentration. The specific
surface area and cumulative pore volume, determined from BET
calculation, for different samples are listed in Table I. It is seen that
the specific surface area decreases with an increase in Co content in
the Ni-Co oxides sample, while the cumulative pore increase from
sample 1:0 to 1:4, and then decreases. The trend in mean pore width
may possibly be due to the co-existence of big and small pores in
the Ni-Co oxides samples.
Ni(OH)
tance than that of Co O With increasing of Co concentration, the
2
/NiO has a much higher surface area and specific capaci-
3
4.
ratio of macrospores is increased (Figs. 5b, 5c, and 5d), which may
contribute to the decreasing of specific surface area of samples.
However SEM cannot tell quantitative size distribution of the mi-
crospores in these samples, further characterization was carried out.
In order to give quantitative distribution of the microspores in
these samples, nitrogen adsorption/desorption experiment was car-
The reasons for Ni-Co oxide (1:2) having the maximum specific
capacitance value are further discussed here. If we only consider the
electrochemical double-layer mechanism or the adsorption of cati-
15
ons on the surface of the active material, the capacitance should
follow the trend of surface area, the sample Ni:Co ¼ 1:0 should then
have the highest specific capacitance according to the data in
Table I. Furthermore, NiO has a much higher theoretical specific ca-
pacitance than that of Co O , also indicating that the sample
2
ried out. BJH (Barrett-Joyner-Halenda) calculation with N adsorp-
tion–desorption isotherm desorption branch was adopted. Figure 6
shows the BJH adsorption pore distribution of samples Ni-Co (1:0),
Ni-Co (1:1), Ni-Co (1:2), Ni-Co (1:4), and Ni-Co (0:1). The precise
positions and distributions of the 3–4 nm peaks may be doubtful due
to the use of desorption branch for the calculation, but the qualita-
tive results and trend are meaningful. The majority of the pores fall
in the size range of 2–10 nm, which is favorable for supercapacitor
applications due to the interconnected pores forming channels easy
3
4
Ni:Co ¼ 1:0 should have the highest capacitance. However, sample
Ni-Co oxide (1:2) has the highest capacitance, even though it does
not have the biggest surface area and the highest NiO or Ni(OH)2
concentration. This phenomenon should be related to the utilization
of these oxides during the charge/discharge process. Pore size and
pore volume are also important factors for determining the surface
utilization besides the specific surface area. If the pore size is big,
the surface area can be small and thus the capacitance is low; how-
ever, too small a pore width will limit ions penetrating into the ma-
1
2
for electrolyte to diffuse into the material. An increase in the elec-
trode-electrolyte interfacial area enables more activities of OH
ꢀ
16
terial to cause a low capacitance as well. As shown in Table I, Ni-
Co oxide (1:2) has large pore width and cumulative pore volume
among the five samples. The large pore width and volume may
allow ions penetrating into the pores and react with the material
more easily and gives rise to an increase in the surface utilization. In
addition, pore size distribution in Ni-Co oxides may also play roles,
the capillary force would hold the electrolyte and prevent it from
17
over flooding due to the small size of the pores (<10nm). Further-
more, Pseudo-capacitance mechanism can be another factor in
affecting capacitor performance and capacitance. Both the curved
instead of linear shape of the charge/discharge profiles and the redox
peaks in CV curves suggest a pseudo-capacitive characteristic. Re-
dox peaks in CV curves are caused by the Faradaic reactions (elec-
trochemical reactions) between active materials in electrodes and
electrolytes. The capacitance from Faradaic reactions belongs to a
pseudo-capacitance. Therefore, in addition to surface utilization,
materials reactivity also play important role. The improvement in
the capacitance by adding certain amount of Co ions into the NiO
Figure 5. SEM images of (a) Ni(OH)
2
/NiO (b) Ni-Co oxide (1:2) (c) Ni-Co
3
þ 2þ
oxide (1:4) (d) pure Co
3
O
4
.
matrix may also may be that Co takes some sites of Ni in NiO
Downloaded on 2013-11-22 to IP 128.230.168.102 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

Journal of The Electrochemical Society, 158 (6) A695-A699 (2011)
A699
cycle test. One is that we want to see the supercapacitor perfor-
mance at a high charge/discharge current, because it is well known
that a high charge/discharge current is very harmful to the life of a
battery. Another reason is that 2000 charge/discharge cycles at 1 A/g
takes too long (ꢁ 670 h) due to the high capacitance or high charge
storage ability of the sample, and it takes only ꢁ 110 h for 2000
cycles at 5 A/g. The results are shown in Fig. 8. It can be seen that
the discharge specific capacitance increases sharply in the initial 240
cycles. This initial increase is due to the mesoporous structure,
because electrodes were needed to activate fully. Similar phenomena
12,17
17
were reported by other groups.
Luo et al. observed a capaci-
tance increase in the initial 300 cycles for Mn-Ni-Co oxide compo-
sites and they explained that such a phenomenon was caused by a
gradual penetration of electrolyte into the active material. After reach-
ing the maximum at around 240 cycles, it begins to decrease and sta-
bilizes after around 1000 cycles. The capacitance retention after 2000
cycles is about 47%. The fade of capacitance of Ni/Co oxide is
mainly due to the peel off the active materials form electrode during
charging and discharging process, as precipitates can be found in the
bottom of container.
Figure 7. (Color online) Charge/discharge curves of Ni-Co oxide (1:2)
recorded in 1 M KOH electrolyte at different current densities of 1, 2, and 5
ꢀ1
A g
.
Conclusions
In this work, Ni-Co oxides with different compositions were suc-
cessfully prepared by a sol-gel method. The microstructure and
nanostructure of these marterials are found strongly influenced by
Ni and Co concentrations. Increasing Co concentration changes the
sample from NiO dominant structure (Ni: Co ¼ 1:2) to Co O domi-
crystal lattice, which induces additional positive defection. The
probability of protons transporting is thus increased and electrons
can escape from Ni much more easily, so that the chemical active-
3
þ
ness of electrode material is improved. Furthermore, Co could
also improve the conductivity of electrodes, reduce the inner resist-
ance of the chemical reactions, and improve reversibility of the elec-
3
4
nant structure (Ni: Co ¼ 1:4) with a reduced BET surface area. The
electrochemical properties of the Ni-Co oxides are found being
strongly affected by the Ni/Co ratio, microstructure, morphology,
18
trode reactions.
Figure 7 shows the Ni-Co oxide (1:2) charge/discharge curves at
different current densities within the potential window of 0.0–0.4 V
in 1 M KOH solution. The specific capacitance is calculated to be
ꢀ1
and mesoporosity. A maximum specific capacitance of 1539 Fg
was obtained for Ni-Co oxide (1:2) at a charge/discharge current
density of 1 Ag , such a capacitance is similar to the maximum
ꢀ1
ꢀ ꢀ1
1
1
539, 1447 and 1136 F g at 1, 2, and 5 Ag , respectively. An
2
value of the expensive RuO . Its cyclic reversibility and perform-
increase in the current density leads to a decrease in specific capaci-
tance as expected, because some of the ions have not participated in
the reaction due to the slow diffusion rate and electron transfer rate
at a high current density. In order to get a better understanding of
the cyclic behavior of Ni-Co oxide (1:2), another experiment was
carried out with the electrode subjected to charge/discharge over
ance under large discharge current are also good.
Acknowledgment
Funding supports of A*Star SERC grant 0721330044/R284-000-
67-592 and MOE Tier 1 (FRC) grant R284-000-071-112 of Singa-
0
pore are appreciated.
ꢀ
1
2
000 cycles at a current density of 5 Ag between 0.0 and 0.4 V at
National University of Singapore assisted in meeting the publication
costs of this article.
room temperature. There are two reasons for our using a high
charge/discharge current of 5 A/g, rather than 1 A/g, in the 2000
References
1
.
P. Simon and Y. Gogotsi, Nature Mater., 7, 845 (2008); Vasile V. N. Obreja, Phys-
ica E (Amsterdam), 40, 2596 (2008).
L. L. Zhang and X. S. Zhao, Chem. Soc. Rev., 38, 2520 (2009).
C. C. Hu, W. C. Chen, and K. H. Chang, J. Electrochem. Soc., 151, A281 (2004).
B. E. Conway, Electrochemical Supercapacitors: Scientific Fundamentals and Tech-
nological Applications, Kluwer Academic/Plenum, New York (1999); J. Zheng,
P. Cygan, and T. Zow, J. Electrochem. Soc., 142, 2699 (1995).
2.
3.
4.
5
.
.
J. Cheng, G. Cao, and Y. Yang, J. Power Sources, 159, 734 (2006); G. X. Hu, C.
Li, and H. Gong, J. Power Sources, 195, 6977 (2010).
Y. Gao, S. Chen, D. Cao, G. Wang, and J. Yin. J. Power Sources, 195, 1757
6
(2010).
J. Zhang and J. Y. Fang J. Am. Chem. Soc., 131, 51 (2009).
J. Chen, D. H. Bradhurst, S. X. Dou, and. H. K. Liu, J. Electrochem. Soc., 146,
7
8
.
.
3606 (1999).
C. Hu and C. Cheng. Electrochem Solid-State Lett. 5, A43–A46 (2002).
9.
10. K. He, Q. Wu, X. Zhang, and X. Wang, J. Electrochem. Soc., 153, A1568 (2006);
K. He, X. Zhang, and J. Li, Electrochim. Acta., 51, 1289 (2006).
11. X. Liu, Y. Zhang, X. Zhang, and S. Fu, Electrochim. Acta., 49, 3137 (2004).
12. T. Wei, C. Chen, H. C. Lu, and C. Hu, Adv. Mater., 22, 347 (2010).
13. J. Deng, J. Deng, Z. Liu, H. Deng, and B. Liu, J Solid State Electrochem., 13, 1387
(
2009).
4. Z. Fan, J. Chen, K. Cui, F. Sun, Y. Xu, and Y. Kuang, Electrochim. Acta., 52, 2965
2007).
1
(
1
1
1
5. S. Devaraj and N. Munichandraiah, J. Phys. Chem. C, 112, 4406 (2008).
6. L. Cao, Y. Zhou, M. Lu, and H. Li, Chin. Sci. Bull., 48, 1212 (2003).
7. J. Luo, B. Gao, and X. Zhang, Mater Research Bull., 43, 1119 (2008).
Figure 8. Dependence of specific capacitance of Ni-Co oxide (1:2) on num-
ber of cycles.
18. W. Yu, X. Yang, and G. Wang, Chin. Sci. Bull., 49, 1446 (2004).
Downloaded on 2013-11-22 to IP 128.230.168.102 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use