Electrochimica Acta 50 (2005) 2499–2506

Determination of the specific capacitance of conducting

polymer/nanotubes composite electrodes
using different cell configurations
V. Khomenkoa,1 , E. Frackowiakb,2 , F. Béguina,∗
a CRMD, CNRS-University, 1B, Rue de la Férollerie, 45071 Orléans Cedex 02, France
b Poznan University of Technology, Ul. Piotrowo 3, 60-965 Poznan, Poland

Received 30 April 2004; received in revised form 9 October 2004; accepted 14 October 2004
Available online 16 December 2004

Abstract

Composite materials containing 20 wt.% of multiwalled carbon nanotubes (MWNTs) and 80 wt.% of chemically formed conducting
polymers (ECP) as polyaniline (PANI) and polypyrrole (PPy) have been prepared and used for supercapacitor electrodes. The well conducting
properties of MWNTs and their available mesoporosity allow a good charge propagation in the composites. Moreover, due to the good resiliency
of MWNTs, an excellent stability of the supercapacitor electrodes is observed. It has been shown that the capacitance values for the composites
strongly depend on the cell construction. In the case of three electrode cells, extremely high values can be found from 250 to 1100 F/g, however
in the two electrode cell much smaller specific capacitance values of 190 F/g for PPy/MWNTs and 360 F/g for PANI/MWNTs have been
measured. It highlights the fact that only two-electrode cells allow a good estimation of materials performance in electrochemical capacitors.
The applied voltage was found to be the key factor influencing the specific capacitance of nanocomposites. For operating each electrode in
its optimal potential range, asymmetric capacitors have been built with PPy/MWNTs as negative and PANI/MWNTs as positive electrodes
giving capacitance values of 320 F/g per electrode material.
© 2004 Elsevier Ltd. All rights reserved.

Keywords: Supercapacitor; Conducting polymers; Carbon nanotubes; Polypyrrole; Polyaniline; Asymmetric capacitor

1. Introduction for this application due to their excellent capacity for energy
storage, easy synthesis, higher conductivity and lower cost
Recently, many attempts have been devoted to the use than many other conducting polymers [2].
of electronically conducting polymers (ECPs) as electrode However, the long-term stability during cycling is a major
materials in electrochemical capacitors, often called superca- demand for an industrial application of ECPs. Swelling and
pacitors [1]. The increasing interest in such polymers is stim- shrinkage of ECPs is well known and may lead to degradation
ulated by the demand of active materials with high specific of the electrode during cycling. It occurs because the dop-
capacitance, for maximum specific energy and specific power ing of polymers requires the insertion/deinsertion of counter
of the device. Especially polyaniline (PANI) and polypyrrole ions, which cause a volume change. Thus, the mechanical
(PPy) have been considered as the most promising materials stress in the polymer film relates directly with the cycle-
life of polymer-based capacitors. This has been overcome
to some extent by the use of composite structures, for ex-
∗ Corresponding author. Tel.: +33 2 3825 5375; fax: +33 2 3863 3796. ample, a combination of ECP and insulating polymers with
E-mail address: [email protected] (F. Béguin).
1 On leave from the Kiev National University of Technology and Design, good mechanical properties such as poly-N(vinyl alcohol)
2, Nemirovich-Danchenko Str., 02011 Kiev, Ukraine. and polystyrene [3,4]. However, in each case, the conductiv-
2 ISE member. ity of the composite materials is lower than in the pristine

0013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.10.078
2500 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506

ECP. Rather high energy and power densities have been re- cal method. The electrochemical method is limited in terms
ported when ECPs are deposited as a thin coating on some of mass production of the materials and also it is not suit-
supporting substrate such as carbon paper and cloth [5–7]. able for preparing composites with a large proportion of
Hence, the most interesting solution is to use carbon mate- polymer.
rials to improve the mechanical properties of the electrodes. Composites were prepared by immersion of the MWNTs
Moreover, the presence of carbon in the bulk of ECPs al- (0.1 g) into an aqueous solution of monomer and by addi-
lows to ensure a good electronic conduction in the electrode tion of an oxidant to this solution. Electroconducting PPy
when the polymer is in its insulating state. However, electro- was deposited on MWNTs by chemical polymerization of
conducting carbon additives generally provide a low specific 0.5 ml pyrrole with 1.2 g of FeCl3 in 50 ml of 0.1 mol l−1 HCl.
capacity in comparison to ECPs. Therefore, it is necessary to The PANI/MWNT nanocomposite was synthesized by poly-
minimize the carbon content in the composite in order to ob- condensation of 0.4 ml aniline by 0.4 g K2 Cr2 O7 in 50 ml of
tain a high value of specific capacitance and energy per total 1 mol l−1 HCl. The nanocomposites materials were then fil-
mass of supercapacitor electrode. For these reasons, it can be tered and washed with a large amount of water and several
profitable to use a moderate amount of nanotubular materials times with ethanol to remove the residual oxidant. At the end,
as surface area-enhancing component and as electronic con- all composites were washed with acetone and dried at 60 ◦ C.
ductor in ECP-based electrodes. Different kinds of nanotubes The content of MWNTs in the composites was determined
and their composites have been already proposed as electrode by weighing of the dry composite materials.
materials for supercapacitors [8–16]. Thin layers of ECP can The electrodes were pellets obtained by pressing the com-
be deposited on the nanotubes, while still keeping the advan- posite material at 500 kg cm−2 . The diameter of the pellets
tage of their open mesoporous network that allows an easy was 1 cm. It was experimentally found that it is possible to
diffusion of ions. Hence, electrodes containing an important form electrodes without any polymer binder in the case of the
mass of ECP can be prepared from such nanocomposites and nanocomposite materials. These electrodes have good me-
used in supercapacitors. chanical properties preserved after long cycling. Scanning
This work focuses on the combination of the complemen- electron microscopy (SEM, HITACHI S 4200) has been used
tary properties of ECPs and multiwalled carbon nanotubes for the observation of the obtained composite materials and
(MWNTs) prepared by catalytic decomposition of acetylene of the pellet capacitor electrodes.
on a Cox Mg1−x O solid solution [17]. Considerable attention The electrodes were tested by cyclic voltammetry and
is paid to the determination of the specific capacitance of galvanostatic charge–discharge cycling in 1 mol l−1 H2 SO4
materials based on two kinds of conducting polymer, i.e. as electrolyte using a VMP multichannel potentio-
polypyrrole (PPy) and polyaniline (PANI). The values of spe- stat–galvanostat (Biologic, France). Two- and three-electrode
cific capacitance reported in literature are mainly quoted in cells were used for the measurements. The two-electrode
terms of number of farads per gram (F/g) of active material. cells were built with a glassy fibrous separator and gold
Rather, different values of specific capacitance are reported current collectors using a sandwich type construction (elec-
for the conducting polymers and their composites due to the trode/separator/electrode) and Swagelok® cylinder cells
variable properties of ECPs and the difference in the experi- from Teflon. In the case of the three-electrode cell, the work-
mental methods used for capacitance determination. For ex- ing electrode was the MWNT/ECP composite, a platinum
ample, with PPy, capacitance values ranging from 100 to wire was the counter electrode, and Hg/Hg2 SO4 was used as
500 F/g are reported using different experimental techniques reference electrode.
[5–7]. The reasons for this phenomenon are not clearly under-
stood. For better understanding of this inconsistency, special
attention is paid in this study to specify the electrochemical 3. Results and discussion
technique used, and to verify the results obtained from the
two- and three-electrode cell set up. A comparison of the 3.1. Morphological characteristics
capacitance values for different ECP/MWNTs composites is
presented emphasizing the crucial role of nanotubes and the The surface morphology of the PANI/MWNTs and
kind of cell construction on the results obtained. PPy/MWNTs nanocomposites containing 80 wt.% of ECP
is shown in Fig. 1a and b. The SEM picture 1a shows that
a homogeneous layer of PANI covers the surface of indi-
2. Experimental vidual MWNTs. The average diameter of the nanotubes af-
ter chemical deposition of PANI are up to 80 nm. For com-
For preparing the nanocomposite materials, we used parison, it should be noted that the diameter of the pristine
MWNTs prepared by acetylene decomposition on MgO sup- MWNTs is mainly ranging from 10 to 15 nm [17]. In the case
ported cobalt particles at 600 ◦ C [17]. ECP/MWNTs com- of PPy/MWNTs, a non-uniform layer of composite was ob-
posites are fabricated by chemical oxidative polymeriza- served (Fig. 1b). After an initial period of MWNTs coating,
tion of a monomer on the surface of MWNTs in order to PPy tends to preferentially grow on itself rather than on the
overcome the problems encountered with the electrochemi- carbon nanotubes. In the PPy composite materials, a large
 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506 2501

Fig. 1. Scanning electron micrographs showing the morphology of (a) PANI/MWNTs and (b) PPy/MWNTs composite materials, which contain 80 wt.% of
conducting polymer.

amount of polymer globules is observed on the surface of MWNTs presented a texture similar to that of the pure poly-
MWNTs. mer. From the electrochemical characterization, it was found
The pellet composite electrodes were also examined by that such dense pellet electrodes demonstrate a high initial
SEM. From Fig. 2a, it can be seen that all the polymer is resistance and a small value of specific capacitance. Hence,
dispersed on the nanotubes, keeping the advantage of the en- under the limit of 20 wt.% of MWNTs, the electrodes have a
tangled network of the nanotubes that allows a good access limited permeability for the electrolyte and are not applicable
of the electrolyte to the active polymer material. There is no for high performance capacitors.
doubt that such a texture of the capacitor electrodes is opti-
mal for the fast ion diffusion and migration in the polymer 3.2. Electrochemical characterization
so that the electrode performance should be improved. By
contrast, as can be seen from Fig. 2b, the electrodes, which The obtained composite electrodes containing 80 wt.% of
are pressed from pure ECP, are not porous. The pellets of ECP were characterized in three-electrode cells by cyclic
composite electrodes, which contained less than 20 wt.% of voltammetry and galvanostatic discharge. Cyclic voltammo-

Fig. 2. SEM micrographs showing the morphology of a cross-section of pellets (a) PANI/MWNTs and (b) pure PANI.
2502 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506

Fig. 3. Cyclic voltammogram (three-electrode cell) of a PPy/MWNTs pellet Fig. 5. Galvanostatic discharge curve (three-electrode cell) at 2 mA of a
electrode; mass of composite, 9.6 mg; electrolyte, 1 mol l−1 H2 SO4 ; scan PPy/MWNTs pellet electrode; mass of composite, 9.6 mg.
rate, 5 mV/s.
The specific capacitance of the composite electrodes was
grams (CV) for PPy and PANI composite electrodes are also determined by galvanostatic discharge. The potential
shown in Figs. 3 and 4, respectively. The CV curves show the range for this experiment was chosen taking into account
reversible charge/discharge behavior of the composite elec- the CV curves. It should be noted that all ECPs may supply a
trodes. The shape of the obtained curve is very similar to the charge within the potential window, which is determined by
typical CV of thin ECP films on different substrates shown in the polymer degradation at positive potential and the switch-
other investigations [18,19]. This means that the properties ing to an isolating state (undoped state) at negative potential
of the composite electrodes are generally determined by the during the discharge of the electrode. According to Fig. 3,
electrochemical behavior of ECP. It is noteworthy that the the maximum positive value for the PPy composite electrode
pure MWNTs supply a capacitance of ca. 20 F/g [20]. Addi- is ca. 0.2 V versus Hg/Hg2 SO4 . Potentials more positive than
tionally, the pellets contain only 20 wt.% of MWNTs against 0.2 V are related with a fast increase of the current response
80 wt.% of ECP. Hence, the double-layer current due to the due to an irreversible oxidation of the polymer. The negative
capacitance of MWNTs is negligible compared to the charge value is −0.6 V versus Hg/Hg2 SO4 , which corresponds to
stored by the ECP. the potential in the fully undoped state of the PPy electrode.
The capacitance value was obtained from the CV curves Hence, the maximum useful potential range for the PPy elec-
according to the equation: trode is 0.8 V. In the case of electrodes based on PANI, it was
necessary to operate with a smaller voltage range because the
i
C= (1) potential of switching to the isolating state is more positive
s (−0.4 V versus Hg/Hg2 SO4 according to Fig. 4) than in the
where i is the average current and s is the potential sweep rate. case of the PPy composite electrode.
According to this formula, the cathodic charge determined Taking into account the accurate potential range, the gal-
by integration of the negative sweep of the CV curves is very vanostatic and CV measurements provide very close values
high for such composite electrodes. The capacitance is 670 of specific capacitance for the composite electrodes. As it
and 506 F/g of electrode for the PANI and PPy composite is shown in Figs. 3 and 4, the ECP/MWNTs composite ex-
electrodes, respectively. hibits a non-constant current response in cyclic voltamme-
try with one maximum for PPy or several maxima for PANI
instead of the typical rectangular voltammogram expected
for an ideal double layer capacitor. The maxima on the CV
curve are related with bigger values of capacitance in this
range of potential. Hence, the galvanostatic discharge curves
(Figs. 5 and 6) are not linear in the total range of potential, and
they are characterized by varying slopes during discharge.
Since the capacitance of the electrode can be evaluated from
the galvanostatic discharge curve by the equation C = I/tg(α),
where α is the angle between the discharge curve and the time
scale; the composite electrodes show a higher capacitance in
the negative than in the positive potential range. In the case
of the PPy composite electrode, the evaluated specific capac-
Fig. 4. Cyclic voltammogram (three-electrode cell) of a PANI/MWNTs pel-
itance is 250 F/g in the potential range from 0.2 to −0.3 V
let electrode; mass of composite, 5.4 mg; electrolyte 1 mol l−1 H2 SO4 ; scan versus Hg/Hg2 SO4 and 903 F/g for the range from −0.3 to
rate, 2 mV/s. −0.6 V (Fig. 5). The average value is 495 F/g for the total
 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506 2503

Fig. 6. Galvanostatic discharge curve (three-electrode cell) at 2 mA of a

Fig. 8. Cyclic voltammogram (2 mV/s) of a symmetric capacitor based on
PANI/MWNTs pellet electrode; mass of composite, 5.4 mg.
PANI/MWNTs composite electrodes (two-electrode cell); mass of each elec-
trode, 10.4 mg.
range from 0.2 to −0.6 V. The same experimental result was
obtained with the three-electrode cell from PANI/MWNTs.
According to Fig. 6, it is possible to estimate an extremely
high value of 1110 F/g for the potential range from −0.2 to
−0.4 V and much lower value of 410 F/g for the range from
0.2 to −0.2 V.
On the basis of the above results, it should be concluded
that the values of specific capacitance are determined by
the potential range used for testing the composite electrode.
Therefore, the performance of the prepared composite elec-
trodes must be evaluated in real supercapacitors constituted as
a two-electrode cell. For this reason, we prepared two sym-
metric capacitors (Type I according to the Ref. [5]) using
the PANI and PPy composite electrodes. The electrochemi- Fig. 9. Galvanostatic charge/discharge characteristics of a symmetric ca-
cal properties were evaluated by the same techniques, which pacitor based on PPy/MWNTs composite electrodes (two-electrode cell);
were applied in the above investigation of the three-electrode current, 2 mA; mass of each electrode, 8.6 mg.
cells. The CV curves of two-electrode capacitors using PPy
culated according to the equation:
and PANI composite electrodes are shown in Figs. 7 and 8,
respectively. They present a rectangular box-like shape char- C
Cspec = 2 (2)
acteristic for an ideal capacitor with good charge propagation. m
The capacitance was obtained using Eq. (1) by integration
where C is the experimental measured capacitance of the
of the negative current in the CV curves. Also galvanostatic
supercapacitor, m is the mass of one composite electrode.
discharge curves (Figs. 9 and 10) were used to determine
A more detailed estimation of the electrochemical prop-
the capacitance of these polymer composite electrodes. The
erties of the PPy/MWNTs and PANI/MWNTs-based super-
specific capacitance per mass of one pellet electrode was cal-
capacitors can be carried out by analysis of the impedance

Fig. 7. Cyclic voltammogram (5 mV/s) of a symmetric capacitor, based on Fig. 10. Galvanostatic charge/discharge characteristics of a symmetric ca-
PPy/MWNTs composite electrodes (two-electrode cell); mass of each elec- pacitor based on PANI/MWNTs composite electrodes (two-electrode cell);
trode, 8.6 mg. current, 2 mA; mass of each electrode, 10.4 mg.
2504 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506

Table 1
Average value of specific capacitance (F/g) of composite electrodes depend-
ing on the measurement technique
ECP in the Three-electrode Two-electrode
composite electrode cell cell
CV Galvanostatic CV Galvanostatic
discharge discharge
PPy 506 495 192 200
PANI 670 650 344 360

tance of the PPy composite materials according to Fig. 5, i.e.

250 F/g calculated for the positive range of potential from 0.2
to −0.3 V and 903 F/g for the negative range from −0.3 to
−0.6 V using a three-electrode cell, the overall capacitance
will be the following:
C1 × C2 250 × 903
C= = = 196
Fig. 11. Nyquist plots of MWNT/PANI (䊉) and MWNT/PPy (
)-based C1 + C 2 250 + 903
supercapacitors (two-electrode cells).
This value of capacitance (196 F/g) is well consistent with
spectra. The Nyquist plot of the supercapacitors (Fig. 11) the experimental value of specific capacitance obtained in the
shows a good capacitor-like behavior with a small diffusion two-electrode cell (192 F/g). In addition, it should be noted
limitation. The internal resistances at 10 kHz (real Z axis) that the value of C obtained in a two-electrode cell will de-
for capacitors built from PPy- and PANI-based composites pend on the equilibrium potentials of both electrodes. Thus,
were 0.86 and 0.48  cm2 , respectively. the value of specific capacitance should be increased when
The results of the capacitance measurements are summa- the equilibrium potentials of both electrodes will be shifted in
rized in Table 1. It is important to notice that for both types of negative potential direction. However, in the practice, one can
cells, there is a good correlation between the values calculated observe an opposite trend, especially with increasing the su-
from CV and galvanostatic measurements. The comparison percapacitor operating voltage. The potential of the positive
of the data obtained from the three-, and two-electrode cells electrode is mainly shifted in the positive direction where a
clearly demonstrate a marked difference between the values smaller capacity is observed. For example, this fact was con-
of specific capacity. The trend for higher values in the case firmed experimentally in a symmetric capacitor with pressed
of three-electrode cells might be understood on the basis of PANI composite electrodes in 1 mol l−1 H2 SO4 . The OCV
the above experimental results. In the real capacitor, the same potential of both PANI-based electrodes was −0.25 V ver-
polymer material works in different potential range for each sus Hg/Hg2 SO4. According to Fig. 6, it is obvious that the
electrode. Hence, the positive electrode operates in the posi- electrode polarization above −0.25 V deals with a lower ca-
tive range of potential and the negative electrode in the nega- pacitance than below −0.25 V. Hence, the positive electrode
tive one. According to the measurements in a three-electrode gives a lower specific capacitance than the negative electrode.
cell, the capacitance of composite materials at negative elec- The value of C was about 360 and 210 F/g at cell voltages of
trode potentials behaves differently from positive potentials. 0.4 and 0.6 V, respectively. The same trend was observed in
The capacitance of electrodes based on ECP decreases with the case of a symmetric capacitor with PPy composite elec-
the increase of positive potential. Thus, the positive electrode trodes, where the specific capacitance of the PPy composite
will have less capacitance than the negative one. Taking into electrode decreased to 180 F/g after increasing the operating
account the equivalent circuit of a two-electrode supercapac- voltage to 0.8 V. This is because the OCV of both PPy-based
itor: electrodes was −0.38 V. According to Fig. 5, it is obvious that
the electrode polarization above −0.38 V supplies a lower ca-
pacitance than below −0.38 V. Thus, the positive electrode
the overall capacitance (C) is given by the equation: based on PPy is characterized by a lower specific capaci-
tance than the negative electrode. It should be mentioned
1 1 1
= + (3) that the cycling life of capacitor decreases dramatically af-
C C1 C2 ter increasing the operating voltage. The cycle characteriza-
where C1 and C2 are the capacitance of the positive and neg- tion of supercapacitor was evaluated using constant current
ative electrodes, respectively. charge–discharge cycling over 1000 cycles. From experimen-
Hence, according to Eq. (3), the total capacitance of the su- tal data obtained using lower operating voltage, a total loss
percapacitor (C) will be essentially determined by the small- of about 15% for PPy and 10% for PANI of initial discharge
est of the two components, i.e. C1 . For example, based on capacitance is found after 50 cycles, but a fairly steady capac-
Eq. (3) and taking into account the values of specific capaci- ity is maintained thereafter. On the contrary, upon increasing
 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506 2505

Fig. 12. Galvanostatic charge/discharge characteristics (two-electrode cell)

of an asymmetric capacitor based on PANI/MWNTs as a positive electrode
Fig. 14. Specific capacity during galvanostatic cycles (two-electrode cell)
and PPy/MWNTs as a negative electrode; current, 1 mA; mass of each elec-
for the: PANI/MWNTs supercapacitor operating from 0 to 0.4 V (䊉); the
trode, 8.2 mg.
asymmetric PANI/MWNTs-PPy/MWNTs supercapacitor operating from 0
to 0.6 V (); the PPy/MWNTs supercapacitor working from 0 to 0.6 V().
the operating voltage, the decrease of capacity reaches about
40% after 1000 cycles’ performance. higher value of 320 F/g has been found. The box-like shape
From the above data, it is possible to assume that the to- of the voltammogram (Fig. 13) suggests pure capacitance be-
tal capacitance of a symmetric capacitor will be higher, if haviour for both electrodes in this kind of asymmetric capac-
the mass of the positive electrode will be increased. How- itor. Hence, we showed the possibility to increase the specific
ever, it will not change significantly the value of specific capacitance using an asymmetric capacitor with ECP com-
capacitance per material, because the total mass of active posite electrodes (Type II). The value of capacitance obtained
materials in the cell will be increased. Therefore, we decided from the asymmetric capacitor, ca. 305 F/g is higher than
to check another possibility of increasing the value of spe- for the symmetric capacitor with PPy composite electrodes
cific capacitance by using the same mass of PANI/MWNTs (192 F/g), but slightly lower than for the similar system with
and PPy/MWNTs composites as positive and negative elec- PANI composite electrodes (360 F/g). Nevertheless, higher
trodes, respectively, forming an asymmetric supercapacitor values of voltage are reachable with the asymmetric con-
(Type II). It is based on the determination presented above, figuration (−) PPy/MWNTs|PANI/MWNTs (+) using 0.6 V
which shows that the PANI composite electrode demonstrates operating voltage against 0.4 V with the symmetric capacitor
a higher value of capacitance than the PPy composite in the based on PANI composite electrodes, that is an advantage
positive range of potentials. Galvanostatic charge/discharge in term of stored energy. The cyclic characterization of the
from 0 to 0.6 V (Fig. 12) and cyclic voltammetry with a asymmetric configuration was evaluated using galvanostatic
potential scan rate of 5 mV/s (Fig. 13) were performed to charge–discharge over 1000 cycles. From the experimental
estimate the value of capacitance. The linear galvanostatic data presented in Fig. 14, a total loss of about 8% of the initial
discharge (Fig. 12) proves the perfect capacitive behaviour discharge capacitance is found after 50 cycles, but a fairly sta-
of this asymmetric capacitor, and the capacitance calculated ble capacity is maintained thereafter. Thus, the asymmetric
according to Eq. (2) reaches a value of about 305 F/g per system (−) PPy/MWNTs|PANI/MWNTs (+) presents a bet-
pellet electrode. From the voltammetry measurement, even a ter cyclic performance than the symmetric supercapacitors
where both electrodes are from the same composite.

4. Conclusion

Nanocomposites containing 80 wt.% of ECP materials

were obtained by chemical deposition of PPy and PANI on
MWNTs. The present study indicated that pellet electrodes
obtained by simply pressing the nanocomposite materials can
be employed for electrochemical capacitors, especially when
an asymmetric configuration (Type II) is realized. The spe-
cific capacitance of a Type II capacitor with PPy/MWNT
composite as negative and PANI/MWNT as positive elec-
Fig. 13. Cyclic voltammogram (two-electrode cell) at 5 mV/s of an trode achieved a value up to 320 F/g.
asymmetric capacitor based on PANI/MWNTs positive electrode and The presented paper provides guidelines about the
PPy/MWNTs negative electrode; mass of each electrode, 8.2 mg. methodology for determining the specific capacitance of
2506 V. Khomenko et al. / Electrochimica Acta 50 (2005) 2499–2506

electrodes based on ECP. Our data demonstrate that the References

three-electrode cell is only useful for the investigation of
the electrochemical behavior of a single electrode. How- [1] B.E. Conway, Electrochemical Supercapacitors, Kluwer Academic,
ever, the value of specific capacitance obtained in this Plenum Publishers, New York, 1999.
[2] S. Gottesfeld, A. Redondo, S.W. Feldberg, J. Electrochem. Soc. 134
way cannot be simply compared with the value for a two- (1994) 271.
electrode cell. Moreover, it is shown that this value can- [3] E. Rubenstein, J.S. Park, J. Appl. Polym. Sci. 42 (1991) 925.
not be achieved in a practical symmetric capacitor due to [4] R.S. Campomanes, E. Bittencourt, J.S.G. Gampos, Synth. Met. 102
the pseudocapacitance behavior of ECP during the dop- (1999) 1230.
ing/dedoping process. The capacitance of ECP depends on [5] A. Rudge, J. Davey, I. Raistrick, S. Gottesfeld, J. Power Sources 47
(1994) 89.
the range of potential window where it is investigated. In [6] A. Rudge, I. Raistrick, S. Gottesfeld, J.P. Ferraris, Electrochim. Acta
a two-electrode cell, the same ECP material is working 39 (1994) 273.
in different potential ranges. In this case, the total capac- [7] J.H. Park, J.M. Ko, O.O. Park, D. Kim, J. Power Sources 105 (2002)
itance of the supercapacitor is determined by the smaller 20.
capacitance of the positive electrode. For a comparison of [8] E. Frackowiak, K. Jurewicz, S. Delpeux, F. Béguin, J. Power Sources
97–98 (2001) 822.
different ECP materials, it is more reasonable to compare [9] E. Frackowiak, F. Béguin, Carbon 39 (2001) 937.
the values of specific capacitance obtained in two-electrode [10] K. Jurewicz, S. Delpeux, V. Bertagna, F. Béguin, E. Frackowiak,
cells, especially by the galvanostatic charge–discharge test. Chem. Phys. Lett. 347 (2001) 36.
Such an estimation is more close to the practical work of a [11] J.H. Chen, Z.D. Huang, D.Z. Wang, S.X. Yang, W.Z. Li, J.G. Wan,
capacitor. Z.F. Ren, Synth. Met. 125 (2002) 289.
[12] M. Hughes, G.Z. Chen, M.S.P. Schaffer, D.J. Fray, A.H. Windle,
The freshly prepared ECP, depending on the synthesis con- Chem. Mater. 14 (2002) 1610.
ditions, is in the doping state that determines the value of its [13] M. Hughes, M.S.P. Schaffer, A.C. Renouf, C. Singh, G.Z. Chen, D.J.
rest potential Er . Er will affect the values of working potential Fray, A.H. Windle, Adv. Mater. 14 (2002) 382.
range for the positive and negative electrodes and in conse- [14] K.H. An, K.K. Jeon, J.K. Heo, S.C. Lim, D.J. Bae, Y.H. Lee, J.
quence, the capacitance of the supercapacitor, taking into ac- Electrochem. Soc. 149 (2002) 1058.
[15] Y. Zhou, B. He, W. Zhou, J. Huang, X. Li, B. Wu, H. Li, Elec-
count that always the electrode with the smallest capacitance trochim. Acta 49 (2004) 257.
will decide on the total capacitance. [16] J. Huang, X. Li, J. Xu, H. Li, Carbon 41 (2003) 2731.
[17] S. Delpeux, K. Szostak, E. Frackowiak, S. Bonnamy, F. Béguin, J.
Nanosci. Nanotechnol. 2 (2002) 481.
Acknowledgements [18] G. Zotti, S. Cattarin, M. Conisso, J. Electroanal. Chem. 235 (1987)
259.
[19] Y. Qiu, J.R. Reynolds, Polym. Eng. Sci. 31 (1991) 417.
This work was supported by an INTAS grant no. YSF-00- [20] E. Frackowiak, K. Jurewicz, K. Szostak, S. Delpeux, F. Béguin, Fuel
114 and the NATO Science for Peace Programme no. 973849. Process. Technol. 77–78 (2002) 213.