Direct template synthesis of mesoporous carbon and its application to supercapacitor electrodes

Article abstract of DOI:10.1016/j.materresbull.2009.04.019

A direct templating method which is facile, inexpensive and suitable for the large scale production of mesoporous carbon is reported herein. A meso-structure surfactant/silicate template was made in a solution phase and resorcinol-formaldehyde as a carbon

Full text of DOI:10.1016/j.materresbull.2009.04.019

Materials Research Bulletin 44 (2009) 1663–1669
Contents lists available at ScienceDirect
Materials Research Bulletin
journal homepage: www.elsevier.com/locate/matresbu
Direct template synthesis of mesoporous carbon and its application to
supercapacitor electrodes
a,
b
a
Songhun Yoon *, Seung M. Oh , Chulwee Lee
a
Advanced Chemical Technology Division, Korea Research Institute of Chemical Technology (KRICT), Sinseongno 19, Yuseong, Daejeon 305-600, Republic of Korea
Research Center for Energy Conversion and Storage (RCECS), School of Chemical and Biological Engineering and Institute of Chemical Process, College of Engineering,
b
Seoul National University, Seoul 151-744, Republic of Korea
A R T I C L E I N F O
A B S T R A C T
Article history:
A direct templating method which is facile, inexpensive and suitable for the large scale production of
mesoporous carbon is reported herein. A meso-structure surfactant/silicate template was made in a
solution phase and resorcinol–formaldehyde as a carbon precursor was incorporated into the template
solution. After aging, carbonization and hydrofluoric acid (HF) etching, mesoporous carbon was
obtained. Using X-ray diffraction, scanning and transmission electron microscopy and nitrogen sorption,
the synthesis mechanism of the mesoporous carbon was elucidated. According to the small angle X-ray
scattering measurements, the surface became smoother after the removal of the silica, indicating that
the silica was mostly located at the pore surface of the carbon. Also, the calculation of the pore volume
demonstrated that the silica was transferred into the pores of the carbon without structural collapse
during HF etching. When the prepared mesoporous carbon was applied to a supercapacitor electrode, the
rectangular shape of the cyclic voltammogram was less collapsed, even at a high scan rate, which is
Received 25 September 2008
Received in revised form 27 March 2009
Accepted 22 April 2009
Available online 9 May 2009
Keywords:
A. Amorphous materials
B. Chemical synthesis
D. Electrochemical properties
indicative of its high rate capability. This was due to the low resistance of the electrolyte in the pores
2
(
3.8
V
cm ), which was smaller than that of conventional activated carbon electrodes and even
comparable to that of ordered mesoporous carbon electrodes. This improved performance was probably
due to the well developed mesoporosity and high pore connectivity of the prepared mesoporous carbon.
ß 2009 Elsevier Ltd. All rights reserved.
1
. Introduction
Various synthesis methods, such as catalytic activation, the
delicate [10]. Furthermore, the OMS has usually been sacrificed
during silica etching using hydrofluoric acid (HF) or NaOH,
resulting in the high preparation cost of the OMC. Hence, it has
become increasingly important to develop a simpler and easier
direct templating method for the synthesis of mesoporous carbon.
Along these lines, much effort has been made to develop effective
direct templating methods [11–16]. Various carbon precursors,
such as divinyl-benzene, phenylene, 1,4-bis (triethoxylsilyl)ben-
zene and block co-polymer surfactants themselves, have been
employed in the direct templating method using a silica/surfactant
template. Nevertheless, silica etching after carbonization to
generate pores remains an inevitable process in all of these
synthesis methods [11–13]. Recently, several examples of OMCs
prepared using only an organic surfactant (soft template) have
been reported [14–16]. In these methods, the interaction between
the carbon precursor and surfactant originates from resol, which is
a partially polymerized oligomeric state of resorcinol–formalde-
hyde (RF) brought about through hydrogen bonding. Using this
interaction, OMC films were prepared after carbonization to
remove the surfactant template.
carbonization of polymer blend/organic gels, the carbonization of
polymer aerogels and the template method, have been developed
in order to prepare mesoporous carbon materials [1]. Among them,
the template method using porous solid silica templates has been
widely investigated by many researchers. Ordered mesoporous
silica (OMS) materials such as MCM-48, MCM-41, SBA-15, MCF and
HMS have been utilized for the preparation of ordered mesoporous
carbon (OMC) [2–10]. The outstanding properties of OMC have
been observed in its application for gas separation, water
purification, catalyst supports, adsorbents and the electrode
materials of charge storage devices, such as supercapacitors,
lithium ion batteries and fuel cells [2–5]. Although the character-
istic mesoporosity and high pore connectivity of OMCs gave rise to
the fast transport of electrolytes and adsorbents through their
inter-connected mesopores, the preparation of OMS materials and
the templating procedure using OMS remains complex and
In this work, the direct templating method is employed for the
synthesis of mesoporous carbon, whose final form is a powder.
Hence, the preparation method employed herein has the merit of
*
Corresponding author. Tel.: +82 42 860 7199; fax: +82 42 860 7389.
E-mail address: yoonshun@krict.re.kr (S. Yoon).
0025-5408/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2009.04.019

1
664
S. Yoon et al. / Materials Research Bulletin 44 (2009) 1663–1669
easy scale-up for industrial mass production. The meso-structure
of the surfactant/silicate template is made in advance within the
solution phase and the carbon precursor, RF, is incorporated on the
outside of the micelle core. This ordered composite is polymerized,
carbonized and etched to obtain mesoporous carbon. Using small
angle X-ray diffraction (XRD), small angle X-ray scattering (SAXS),
scanning electron microscopy (SEM), transmission electron
microscopy (TEM) and the measurement of the pore size
distribution, the pore formation mechanism in the mesoporous
carbon is clarified. As an application of the prepared mesoporous
carbons, the performance of supercapacitor electrodes is investi-
gated.
carbon specimens carbonized at 600 (C-600), 800 (C-800) and
1000 8C (C-1000) were acquired. For comparison, a control carbon
(C-1) was prepared without the CTAB/SiO template and another
2
control carbon (C-2) was prepared without CTAB.
2.2. Analysis of pores
2
The pore size distribution (PSD) was analyzed by N adsorption
measurements (Micromeritics ASAP 2010). The external morphol-
ogy of the carbon was examined using SEM (JEOL JSM-840A),
whereas the pore image was obtained by TEM (JEOL JEM-2010).
SAXS and small angle XRD were performed at the 4C-1 line of the
Pohang Accelerator Laboratory. The beam flux and size were
1
2
ꢀ2
2
2
. Experimental
10 mm and 0.3 mm , respectively. The data was corrected after
the air and dark scattering.
2.1. Preparation of materials
2.3. Preparation of supercapacitor electrodes
Fig. 1 shows a schematic description of the synthesis process.
After dissolving cetyltrimethylammonium bromide (CTAB) and
sodium hydroxide in deionized water, tetraethylorthosilicate
To prepare the EDLC (electric-double-layer capacitor) electrode,
a mixture of the prepared carbon, polytetrafluoroethylene (PTFE)
binder, and Ketjenblack ECP-600JD (KB) (at a weight ratio of
10:1:1) was dispersed in iso-propyl alcohol and coated on a
1 cm ꢁ 1 cm stainless steel Exmet which was used as the current
collector. It was shown that this amount of conducting materials
(10 wt%) produced much higher electric conductivity than other
ionic conductivity [2]. The resulting electrode plate was pressed
and dried under vacuum at 120 8C for 12 h.
(
TEOS) as a silica source was added. The molar ratio of CTAB/
TEOS/NaOH/H O was 2/1/1/200. Under these highly alkaline
pH > 14) conditions, silicate exists in a multi-charged D4R
2
(
(
8
ꢀ
double four-ring, [Si
8
O
20
]
) state and this anionic silicate
interacts electrostatically with the cationic surfactant initially
existing in a micelle and molecule state [8]. This solution was
stirred at 40 8C for 2 h in order to generate a template that exists in
8
ꢀ
as
the phase of the tubular meso-structure (H
shown in Fig. 1 (CTAB/silicate) [8,9]. Another solution of resorcinol
R) and formaldehyde (F) was prepared with Na CO as a catalyst at
a molar ratio of R/F/Na CO /H O = 10/20/1/100. After this RF
0
) of CTAB/[Si
8
O
20
]
2.4. Evaluation of supercapacitor performance
(
2
3
The electrochemical performance of the composite carbon
electrodes was analyzed with a three-electrode configuration in
2
3
2
solution was allowed to react for several tens of minutes, the two
solutions (the homogeneous template solution and the RF solution)
were mixed at a molar ratio of R/TEOS = 1/1, stirred vigorously and
aged at 85 8C for a week. During this period, the silicate transformed
to silica via polymerization, while the RF sol was converted to a gel
through a condensation reaction [17]. The resulting precipitate
2 4
aqueous 2 M H SO electrolyte. A Pt flag and SCE (saturated
calomel electrode) were used as the counter and reference
electrodes, respectively. The inter-electrode gap between the
working electrode and reference one was fixed at 0.5 cm. Cyclic
voltammetry (CV) was conducted using an EG&G PARC 362
potentiostat in the potential range of 0.0–0.7 V (vs. SCE) with scan
2
(CTAB/SiO /RF) was filtrated, dried in air at 85 8C, and then
ꢀ1
carbonized at high temperature under an Ar atmosphere for 1 h
to removethe CTABandcarbonize the RFgel. Asa finalstep, the silica
included in the carbonized composite was removed by HF. Final
rates ranging from 5 to 50 mV s . The current in the cyclic
voltammograms was divided by the scan rate to obtain the
capacitance vs. potential profiles. To calculate the equivalent series
Fig. 1. Schematic explanation of synthesis. Notice that RF polymer was located at the outer part of CTAB/silicate micelles.

S. Yoon et al. / Materials Research Bulletin 44 (2009) 1663–1669
1665
resistance (ESR) of the EDLC, a potential step of 10 mV was applied
to the system after the stabilization of the potential at 0.2 V vs. SCE
to 1 A.
m
3
. Results and discussion
3.1. Mechanism of direct templating
As shown in Fig. 1, under highly alkaline (pH > 14) conditions,
8
ꢀ
silicate exists in a multi-charged D4R (double four-ring, [Si
8
O
20
]
)
state and this anionic silicate interacts electrostatically with the
cationic surfactant (CTAB) initially existing as rod-shaped micelles.
It is well known that this interaction produces a hexagonal tubular
structure (H
0
phase), which is referred to herein as CTAB/silicate
[
9]. As reported by several researchers, a hexagonal tubular
structure (H
0
) phase formed by the condensation between the rod
micelles was observed at a reaction time of 5–6 min at 20 8C and a
distinct hexagonal phase (H ) was observed after 20 h of reaction
17–23]. Similarly, it was observed from the cryo-TEM analysis
that an ordered structure of SBA-15 was developed after 20 min
22]. From these literatures, therefore, it is concluded that the H
1
[
[
0
phase is formed before a reaction time of 2 h under our preparation
conditions. After the addition of the RF oligomers to the solution
2
containing CTAB/silicate, CTAB/SiO /RF was obtained after aging,
filtration, washing and drying.
Fig. 2(a) shows the small angle XRD patterns of CTAB/SiO
and SiO /C. The latter (SiO /C) was acquired from the carbonization
of the former (CTAB/SiO /RF) at 600 8C. The diffraction peak of
CTAB/SiO /RF was observed at 2.58, which was similar to that of the
CTAB/silicate phase, as shown in Fig. 1 (H phase) [8,9]. From the
2
/RF
2
2
2
2
0
TEM analysis in Fig. 3(a), no isolated ordered phase associated
solely with the silica/surfactant was observed, indicating that the
RF oligomers were imbibed into the CTAB/silicate template at the
molecular level. This result was similar to that of Zhao’s group,
where resol oligomers were incorporated into the hydrophilic
portion of a block copolymer at the molecular level to produce
OMC film [15]. In our experiment, an RF oligomer solution with a
brownish color was formed after several tens of minutes. Also, it
was observed that the gelation of the RF solution occurred when
the reaction time exceeded 30 min, which is indicative of network
formation between the RF oligomeric particles. Before gelation, the
RF oligomers were added to the solution, in order to prepare a
homogeneous composite material. It is well known that the
particle size of the RF oligomer is very small, because of its high pH
condition [32]. Since our RF oligomers had a similar structure to
that of the resol oligomers, it was reasonable to assume that the RF
oligomers were incorporated within the CTAB/silicate phase at the
Fig. 2. (a) Small angle X-ray diffraction patterns before carbonization (CTAB/SiO2/
RF) and after carbonization (SiO
of before (SiO /C) and after HF etching (C). The carbonization temperature was
00 8C (solid line), 800 8C (dotted line) and 1000 8C (dashed line). The arrows
represent crossover positions (q ).
2
/C) at 600 8C. (b) Small angle X-ray scattering data
2
6
c
molecular level. After the carbonization of CTAB/SiO
temperatures ranging from 600 to 1000 8C, the silica/carbon
composite (SiO /C) was obtained. No diffraction peak for SiO /C
2
/RF at
temperatures ranging from 600 to 1000 8C and then the silica
was removed by HF etching to acquire porous carbon materials (C).
2
2
was observed, which is indicative of the collapse of the ordered
structure. This disappearance of the ordered structure was
probably due to the high thermal stress resulting from the large
volume contraction (about 50%) during the carbonization of the RF
polymer [24].
Although the ordered structure collapsed during the carboniza-
tion process, our templating method is easier and simpler when
compared with the template method using solid porous silica. Also,
since the mesoporous carbon was prepared inside a glass bottle on
the bulk scale of several hundred grams, our preparation process of
mesoporous carbon is suitable for scale-up.
2
Fig. 2(b) presents the SAXS data of SiO /C and C prepared at three
different carbonization temperatures (600, 800 and 1000 8C). Note
that the carbonization temperatures were included in the figure.
From the SAXS experiment, it was found that the relationship
between the measured intensity (I) and scattering vector (q)
follows a power law; I(q) / qꢀ
(6 ꢀ D)
= q [25–27]. As q increases, z
ꢀz
ꢀ2
varies from the scattering of the mass fractal (z = 2 from I / q ) to
the scattering of the surface fractal (Porod regime; z = 4 from
ꢀ4
I / q ) [26]. According to Bale and Schmidt, the fractal dimension
ꢀ
D
indicating the surface roughness can be represented by n = N
0
r
.
Here, D, N , r and n are the fractal (Hausdorff) dimension, the
0
characteristic constant of the fractal surface, the length of the
surface edge and the number of cubes comprising the surface,
respectively [25]. Theoretically, D should be 2 for a perfectly
smooth surface, while the larger the deviation from 2, the rougher
the surface; z should have a value of 4 for perfectly smooth
3
.2. Small angle X-ray scattering (SAXS)
To investigate the change in the pore structure with the
2
carbonization temperature, CTAB/SiO /RF was carbonized at

1
666
S. Yoon et al. / Materials Research Bulletin 44 (2009) 1663–1669
2 2 2 2
Fig. 3. TEM image: CTAB/SiO /RF (a). FE-SEM images: SiO /C (b) and C-1000 (c). TEM image: C-1000 (d). SiO /C was obtained by carbonizing CTAB/SiO /RF at 1000 8C.
surfaces, but z < 4 for rough ones [26]. Hence, the pore surface
roughness can be estimated from the measured value of z (z ) in
temperature, which is indicative of a smaller particle size and
negligible sintering (less particle growth).
s
the surface scattering regime (Porod regime: 0.1 < q < 0.25) [27].
Also, the size of the particles near the pore surface can be
3.3. Analysis of pore size distribution
c
compared qualitatively from the crossover q value (q ), at which
the change from scattering by the particles (mass scattering) to
that by the particle surface (surface scattering in Porod regime)
occurs [25].
Fig. 3(a) and (b) presents the morphologies of SiO
respectively. Interestingly, similar tubular-shaped particles were
observed in both SiO /C and C [28]. In Fig. 3(d), the TEM image of
2
/C and C,
2
In Fig. 2(b), the relationships of I(q) vs. q are plotted with
sintering temperature and HF etching. In Table 1, the measured
carbon(C) indicated that it had a highly mesoporous structure,
where the bright part corresponds to the mesopores. The
observed pore shape was a wormhole-like one, which was
assumed to be due to the collapse of the ordered structure after
carbonization [29,30]. From the tubular particle shape and highly
mesoporous structure of the resulting carbon (C), it was
concluded that the CTAB/silicate played the role of a template
to determine the particle shape of the resulting carbon and
concomitantly to generate the pores in the carbon, as explained
in Fig. 1.
values of z
s c s 2
and q are listed. Here, z increases from 3.5 for SiO /C to
4
.2 for C, irrespective of the carbonization temperature, indicating
that the surface becomes smoother after the removal of the silica.
This distinctive change of the surface roughness corresponds to the
fact that the silica (SiO
surface and is removed by HF etching (as shown in Fig. 1) to
generate pores. The change of the crossover position (q ) in Table 1
indicates that the of SiO /C decreased with increasing
carbonization temperature. Because of the increase in the sintering
with increasing temperature, the SiO located at the pore surface
in SiO /C was sintered to form larger particles (smaller q ).
Furthermore, the q of porous carbon (C) was larger than that of
SiO /C and remained invariant, irrespective of the carbonization
2 2
) in SiO /C is mostly located at the pore
c
q
c
2
In Fig. 4, the variations in the pore size distributions (PSD) of
2
2
SiO /C (empty circles) and C (filled circles) with the carbonization
2
c
temperature (600, 800 and 1000 8C) are presented. The Barrett–
Joyner–Halenda (BJH) method based on the adsorption branch was
utilized to obtain the pore size distribution. Also, the surface area
c
2
and pore volume of SiO
respectively. Apparently, the pore volume and surface area were
increased significantly by removing the silica of SiO /C for all
carbonization temperatures. For comparison, a control carbon
C-1) was prepared without CTAB/SiO and it showed nonporous
2
/C and C are listed in Tables 1 and 2,
Table 1
The small angle X-ray scattering data and surface area of SiO
the carbonization temperature.
2
/C and C according to
2
SiO
2
/C
C
(
2
2
ꢀ1
properties (surface area; 10 m g ), because the pore structure
completely collapsed after carbonization (carbon xerogel). This
nonporous property was attributed to the very small primary
particle size of the RF polymer and lack of cross-linking induced by
Carbonization
temperature/8C
Slope in Porod
600
800
3.53
0.025
108
1000
3.51
0.02
100
600
800
4.25
0.12
598
1000
4.18
0.12
568
3.54
0.06
4.21
0.12
756
regime, z
s
Crossover
+
H ions under high pH conditions [17,31,32] shown in Fig. 4 is the
position, q
BET surface area,
c
131
bimodal pore distribution of carbon. Although the reason for it is
not clear, this bimodal distribution may originate from the
2
ꢀ1
S
BET/m g

S. Yoon et al. / Materials Research Bulletin 44 (2009) 1663–1669
1667
3.4. Estimation of pore conversion during HF etching
To clarify the pore formation of C after the HF etching of SiO
2
/C,
the pore generation from SiO
as follows:
2
/C into C was analyzed quantitatively
mea
pore b
cal
SiO
mea
V
þ V
C
V
pore b
xSiO2
cal
pore b
2
b
1
V
¼
¼
þ x
(1)
x
x
C
C
dSiO2
cal
Here V
(
is the calculated specific volume of SiO
2
in SiO
2
/C
SiO
b
2
mea
before HF etching), V
pore b
is the measured specific pore volume in
cal
pore a
SiO
volume in C (after HF etching), xSiO2 is the weight fraction of SiO
in SiO /C, x is the weight fraction of carbon in SiO /C and dSiO2 is
density of SiO
From the thermogravimetric analysis (TGA) under Ar and air, it
was found that SiO /C was composed of 52 wt% carbon and 48 wt%
silica (x was compared
¼ 0:48, x
= 0.52). Using Eq. (1), Vcal
2
/C (before HF etching), V
is the calculated specific pore
2
2
C
2
3
2
2
in SiO /C (2.2 g/cm ) [27].
2
SiO
mea
pore a
C
pore a
2
with V
, and the results are presented in Table 2. Irrespective of
cal
pore a
were very similar to V pmo er ea a,
the carbonization temperature, V
2
indicating that the silica was located within the inner space of SiO /
C and was transferred into the pores of C without structural
collapse during HF etching, showing the validity of the proposed
mechanism in Fig. 1.
3.5. Performance of supercapacitor electrodes
In order to investigate the EDLC performance, electrodes was
fabricated using the prepared mesoporous carbons. Typically, an
EDLC utilizes the charge accumulated in the electric-double layer
of the porous carbon electrodes [33]. The superior properties of
EDLCs compared to batteries lie in their high rate performance and
good cyclability. In previous papers, it has been revealed that the
pore structure, such as the size and connectivity, significantly
influence the rate performance of EDLCs, which is related to the
electrolyte transport through the pores of the carbon electrodes
2
Fig. 4. The pore size distributions calculated by BJH method in N adsorption
profiles. The carbonization temperature was (a) 600 8C, (b) 800 8C and (c) 1000 8C.
The empty circle and filled one are the data measured before and after HF etching,
respectively.
[2–4]. Especially, because of their mesopores and high connectiv-
ity, electrodes fabricated from OMCs prepared using OMS
templates such as MCM-48 and HMS showed superior rate
performance to those made of conventional microporous activated
carbon. Herein, the EDLC electrode performance of the mesoporous
carbon prepared at 1000 8C (C-1000) was investigated. For
comparison, another control carbon (C-2) was prepared without
CTAB and HF-etched to remove the silica (pore diameter: ꢂ1 nm,
characteristic agglomeration of the CTAB/silicate micelles or
2
partial phase separation between CTAB/SiO and RF.
From further experiments, it was found that the pore size
increased with increasing reaction time of CTAB/silicate before the
addition of RF (not shown here). In addition, the particle size of the
mesoporous carbon became smaller with decreasing CTAB
surfactant concentration. This control of the mesoporous carbon
porosity by changing the preparation conditions in the direct
templating method will be addressed in a future report.
2
ꢀ1
measured SBET = 292 m g ), showing microporous properties.
The EDLC performances of the electrodes using the two carbons
were compared.
Fig. 5 shows the capacitance vs. voltage profiles and chron-
oamperograms for the two carbon electrodes. Here, the capaci-
tance vs. voltage profiles were acquired by dividing the measured
current by the scan rate (5 and 50 mV/s) in the cyclic
voltammograms [2,3]. From these profiles, the estimated specific
Table 2
The comparison between the calculated specific pore volume and the measured one
2
in C after HF etching of SiO /C.
ꢀ
1
Carbonization temperature/8C
capacitances (Csp) of C-2 and C-1000 were 20 and 50 Fg
,
respectively, indicating that Csp was proportional to SBET [34]. In an
ideal capacitor, the capacitance vs. voltage profile should have a
rectangular shape, irrespective of the scan rate. The collapse of the
rectangular shape with increasing scan rate is highly influenced by
6
00
800
1000
SiO
2
/C
mea
pore
cal
=cm3
=cm3
g
g
ꢀ1
ꢀ1
a
0.18
0.22
0.11
0.22
0.10
0.22
V
b
b
b
V
SiO
2
the time constant (
following relation between
resistance);
= ESR ꢁ Csp. Importantly, the rate performance of
EDLC electrodes is inversely proportional to [2,3,35,36]. As seen
t
) of the EDLC electrode, which has the
C
V
V
C
sp and ESR (equivalent series
cal
=cm3
=cm3
ꢀ1
ꢀ1
c
g
g
0.77
0.76
0.63
0.66
0.62
0.59
pore
mea
pore
a
a
d
t
t
a
Measured specific pore volume in SiO
2
/C (before HF etching) based on BJH
in Fig. 5(a) and (c), the rectangular shape of the C-1000 electrodes
collapsed less severely than that of C-2. When considering the
higher Csp of the C-1000 electrodes, it is probable that the observed
lesser collapse of the rectangular shape was due to the smaller ESR.
method.
b
Calculated specific volume of SiO
2
in SiO /C (before HF etching).
2
c
Calculated specific pore volume in C (after HF etching).
d
Measured specific pore volume in C (after HF etching) based on BJH method.

1
668
S. Yoon et al. / Materials Research Bulletin 44 (2009) 1663–1669
Fig. 5. (a and c) Capacitance vs. voltage profile obtained by cyclic voltametry. (b and d) Typical current transient from a chronoamperometry experiment. Here,
DE, R and C are
potential step value, resistance (ESR) and capacitance, respectively. (a and b) Mesoporous carbon prepared direct template methods (C-1000). (c and d) Control one (C-2).
3.6. Estimation of equivalent series resistance (ESR)
low Rpore in the C-1000 electrodes. The low ESR of C-1000 was
comparable to that of OMC electrodes using HMS or MCM-48 silica
2
In order to estimate the ESR, chronoamperometry was
as a template (2.0–2.9
V
cm ), and is indicative of the presence of
conducted. For these measurements, a potential step (10 mV)
was applied from 0.2 V and the subsequent current transient was
recorded every 0.1 s (Fig. 5(b) and (d)) [2,3,37]. The base potential
well developed mesopores and the high connectivity of the pores
in our mesoporous carbon [2,3,30]. For the conventional activated
carbon, MSC-25 (molecular sieving carbon), the measured ESR was
2
(0.2 V) that was used was selected because the effect of the
4.92
V
cm [2]. Hence, the high rate capability of our mesoporous
pseudocapacitance arising from quinone–hydroquinone couples is
negligible in this potential range [35]. The exponentially decaying
current transients were fitted with the equation indicated in the
inset of Fig. 5(b) and the ESR was obtained from this fitting. In the
figure, the filled circles represent the sampled current, whereas the
solid line corresponds to the best fitted one.
carbon electrode was certainly attributable to its well developed
mesoporosity and high pore connectivity, which survived even
after the collapse of the ordered structure. Additionally, the direct
templating method proposed herein is simpler and cheaper than
the conventional template method using porous silica.
As expressed in the inset of Fig. 5(b), the ESR (equivalent series
4. Conclusions
resistance) can be represented by ESR = Relectrode + Rpore + Rbulk
.
Here, Relectrode, Rpore and Rbulk are the electronic resistance of the
electrode, the electrolyte resistance in the pores and the bulk ionic
resistance between the working and reference electrodes, respec-
tively. After adding more than 7 wt% of carbon black as a
A silicate/surfactant template in the solution phase was utilized
as a direct template for the preparation of mesoporous carbon.
From the small angle XRD, SEM and TEM analyses, the synthesis
mechanism of the mesoporous carbon was clarified. The structural
change with the carbonization temperature was elucidated from
the SAXS experiment. Because of its well developed mesoporosity,
the prepared carbon exhibited improved rate capability, which
was comparable to that of OMS electrodes.
conducting material, Relectrode was calculated to be as small as
2
about 0.02
V
cm , which was negligible in the total ESR [2]. From
ꢀ
1
the bulk electrolyte conductivity (0.68 S cm ) and inter-electrode
gap between the working and reference electrodes (0.5 cm), Rbulk
2
was calculated to be 0.73
V
cm [3]. Because the same cell
structure was employed for both of them, Relectrode and Rbulk were
invariant for the two carbon electrodes. After the extraction of Rbulk
Acknowledgement
and Relectrode from the ESR, Rpore was calculated. The estimated
2
values of Rpore for C-1000 and C-2 were 3.8 and 9.4
V
cm ,
This research was supported by a grant from the Fundamental
R & D Program for Core Technology of Materials funded by the
Ministry of Knowledge Economy, Republic of Korea.
respectively. Hence, the lesser collapse of the rectangular shape in
the capacitance vs. voltage profile was mostly attributable to the

S. Yoon et al. / Materials Research Bulletin 44 (2009) 1663–1669
1669
References
[19] Q. Huo, D.I. Margolese, U. Ciesla, D.G. Demuth, P. Feng, T.E. Gier, P. Sieger, A.
Firouzi, B.F. Chmelka, F. Schuth, G.D. Stucky, Chem. Mater. 6 (1994) 1176.
[
[
[
[
20] Y. Wan, H. Yang, D. Zhao, Acc. Chem. Res. 39 (2006) 423.
21] Y. Wan, Y. Shi, D. Zhao, Chem. Commun. (2007) 897.
22] Y. Wan, D. Zhao, Chem. Rev. 107 (2007) 2821.
23] R. Liu, Y. Shi, Y. Wan, Y. Meng, F. Zhang, D. Gu, Z. Chen, B. Tu, D. Zhao, J. Am. Chem.
Soc. 128 (2009) 11652.
24] G. Savage, Carbon–Carbon Composites, Chapman & Hall, London, 1993, p. 121.
25] H.D. Bale, P.W. Schmidt, Phys. Rev. Lett. 53 (1984) 596.
26] V. Bock, A. Emmerling, J. Fricke, J. Non-Cryst. Solids 225 (1998) 69.
27] T.P. Reiker, S. Misono, F. Ehrburger-Dolle, Langmuir 15 (1999) 914.
28] A. Sayari, J. Am. Chem. Soc. 122 (2000) 6504.
29] C.C. Landry, S.H. Tolbert, K.W. Gallis, A. Monnier, G.D. Stucky, P. Nordy, J.C.
Hanson, Chem. Mater. 13 (2001) 1600.
30] J. Lee, J. Kim, Y. Lee, S. Yoon, S.M. Oh, T. Hyeon, Chem. Mater. 16 (2004) 3323.
31] R.W. Pekala, J. Mater. Sci. 24 (1989) 3221.
32] C. Lin, J.A. Ritter, Carbon 35 (1997) 1271.
33] K. Kinishita, Carbon: Electrochemical and Physicochemical Properties, John Wiley
[
[
[
[
[
[
1] T. Kyotani, Carbon 38 (2000) 269.
2] S. Yoon, J. Lee, T. Hyeon, S.M. Oh, J. Electrochem. Soc. 147 (2000) 2507.
3] J. Lee, S. Yoon, S.M. Oh, T. Hyeon, Adv. Mater. 12 (2000) 359.
4] J. Lee, S. Yoon, S.M. Oh, T. Hyeon, Chem. Commun. (1999) 2177.
5] S.H. Joo, S.J. Choi, I. Oh, J. Kwak, Z. Liu, O. Terasaki, Nature 412 (2001) 169.
6] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992)
[
[
[
[
[
[
7
10.
[
7] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, G.H. Chimelka, G.D. Stucky,
Science 279 (1998) 548.
8] J.Y. Ying, C.P. Mehnert, M.S. Wong, Angew. Chem. Int. Ed. 38 (1999) 56.
9] J. Xu, Z. Luan, H. He, W. Zhou, L. Kevan, Chem. Mater. 10 (1998) 3690.
[
[
[
[
10] J. Lee, J. Kim, T. Hyeon, Adv. Mater. 18 (2006) 2073.
11] I. Moriguchi, Y. Koga, R. Matsukura, Y. Teraoka, M. Kodama, Chem. Commun.
[
[
[
[
(2002) 1844.
[
[
[
[
12] F. Zhang, D. Gu, Y. Yan, C. Yu, B. Tu, D. Zhao, J. Am. Chem. Soc. 127 (2005) 13508.
13] B.-H. Han, W. Zhou, A. Sayari, J. Am. Chem. Soc. 125 (2003) 3444.
14] S. Tanaka, N. Nishiyama, Y. Egashira, K. Ueyama, Chem. Commun. (2005) 2125.
15] Y. Meng, D. Gu, F. Zhang, Y. Shi, H. Yang, Z. Li, C. Yu, B. Tu, D. Zhao, Angew. Chem.
Int. Ed. 44 (2005) 7053.
&
Sons, New York, 1998, p. 12.
34] B.E. Conway,ElectrochemicalSupercapacitors,KluwerAcademic/PlenumPublisher,
999 ,, p. 125.
[
1
[
[
[
35] S. Yoon, J.H. Jang, B.H. Ka, S.M. Oh, Electrochim. Acta 50 (2005) 2255.
36] L. Li, H. Song, Z. Chen, Electrochim. Acta 51 (2006) 5715.
37] A.J. Bard, L.R. Fulkner, Electrochemical Methods: Fundamentals and Applications,
John Wiley & Sons, New York, 1988, p. 15.
[
16] C. Liang, K. Hong, G.A. Guiocho, J.W. Mays, S. Dai, Angew. Chem. Int. Ed. 43 (2004)
5
785.
[
[
17] R.W. Pekala, D.W. Schaefer, Macromolecules 26 (1993) 5487.
18] J. Zhang, D. Goldfarb, Micropor. Mesopor. Mater. 48 (2001) 143.