You are on page 1of 7

doi: 10.1149/2.

057207jes
2012, Volume 159, Issue 7, Pages A1034-A1039. J. Electrochem. Soc.

and Yoshio Ukyo


Nobuhiro Ogihara, Shigehiro Kawauchi, Chikaaki Okuda, Yuichi Itou, Yoji Takeuchi

Impedance Spectroscopy Using a Symmetric Cell


Electrodes for Lithium-Ion Batteries by Electrochemical
Theoretical and Experimental Analysis of Porous
service
Email alerting
click here in the box at the top right corner of the article or
Receive free email alerts when new articles cite this article - sign up
http://jes.ecsdl.org/subscriptions
go to: Journal of The Electrochemical Society To subscribe to
2012 The Electrochemical Society
A1034 Journal of The Electrochemical Society, 159 (7) A1034-A1039 (2012)
0013-4651/2012/159(7)/A1034/6/$28.00 The Electrochemical Society
Theoretical and Experimental Analysis of Porous Electrodes for
Lithium-Ion Batteries by Electrochemical Impedance Spectroscopy
Using a Symmetric Cell
Nobuhiro Ogihara,
,z
Shigehiro Kawauchi, Chikaaki Okuda, Yuichi Itou, Yoji Takeuchi,
and Yoshio Ukyo

Toyota Central R&D Labs., Inc., Nagakute, Aichi 480-1192, Japan


The purpose of this study was to understand the electrochemical behavior of the interface between porous electrodes and electrolytes
of lithium-ion batteries. We propose a new analytical approach that is a combination of the transmission line model theory for cylin-
drical pores and electrochemical impedance spectroscopy using symmetric cells. Mathematical model and experimental impedance
behavior results agree with each other. By mathematically tting the experimental impedance plots, the individual internal resistance
components of the actual porous electrode/electrolyte interface could be described as the following four parameters: electric resis-
tance (R
e
), electrolyte bulk resistance, (R
sol
), ionic resistance in pores (R
ion
), and charge-transfer resistance for lithium intercalation
(R
ct
). In actual electrodes, the R
ion
obtained in this study is a characteristic parameter of the porous electrode/electrolyte interface
that is important to consider for thick electrodes.
2012 The Electrochemical Society. [DOI: 10.1149/2.057207jes] All rights reserved.
Manuscript submitted March 12, 2012; revised manuscript received March 12, 2012. Published July 17, 2012.
The development of rechargeable lithium-ion batteries has had a
substantial impact on society. The development of large-size lithium-
ion batteries is important in automotive applications, such as hy-
brid electric, plug-in hybrid, and electric vehicles. Input-output power
capabilities are an important concern in these applications. Internal
resistance of the lithium-ion cell is a key factor for device power
capability. Here, the purpose of this study was to obtain a detailed
understanding of the internal resistance specic to the porous elec-
trode/electrolyte interface for lithium-ion batteries. Electrochemical
impedance spectroscopy (EIS) is a useful technique for analysis of
the internal resistance.
1
As a preliminary step, a suitable EIS method
is developed to interrogate the true electrode/electrolyte interface. As
shown in Fig. 1, the internal resistances at the porous electrodes could
be expressed by four parameters: electric resistance (R
e
), electrolyte
bulk resistance (R
sol
), ionic resistance in pores (R
ion
), and charge-
transfer resistance for lithium intercalation (R
ct
). To obtain these in-
ternal resistances, we propose a new analytical approach, which is a
combination of a theory based on the transmission line model (TLM)
for cylindrical pores and an EIS analysis technique using symmetric
cells (EIS-SC).
In previous experimental EIS analyzes of lithium-ion batteries,
little consideration has been given to the geometry of porous elec-
trode structures although a relatively large number of simulation
and modeling studies of lithium-ion batteries,
24
electric double
layer capacitors,
5,6
or hydrogen evolution reaction using porous Ni
electrodes
7,8
have been reported. EIS impedance results were mainly
expressed as Nyquist plots (real part: Z

, imaginary part: Z

), which
give no detailed information on the porous-electrode structure. Un-
doubtedly, the actual electrodes of lithium-ion cells have several
types of porous structures consisting of a mixture of active material,
carbon-conducting agent, and binder. This mixture involves several
internal resistances caused by various factors including geometric ef-
fects. Especially, the authors believe that lithium ion movement in the
electrolyte-lled pores inuences the internal resistance in the thick
electrodes. When charge transfer reaction occurs, the charge-transfer
resistance has been often interpreted to include the ionic resistance
in pores. In order to consider the internal resistance essentially, those
two resistances should be distinguished. TLMhas been widely used to
describe such porous electrodes as a cylindrical pore for both faradaic
(ideally non-polarized electrode) and non-faradaic (ideally polarized
electrode) processes.
5,6,911
TLM for cylindrical pores is applied ei-
ther to estimate Nyquist plot proles or to nd individual internal
resistances by experimental data tting.

Electrochemical Society Active Member.


z
E-mail: ogihara@mosk.tytlabs.co.jp
Figure 1. Schematic illustration of respective internal resistances at porous
electrodes for lithium-ion batteries.
In the EIS-SC technique, two identical electrodes are prepared at
the same potential before measurement and used as both electrodes in
the symmetric cell.
12,13
In conventional EIS measurements for positive
or negative electrodes, lithium metal is often used as a counter elec-
trode. The resulting impedance spectrum includes the impedance be-
havior of the lithium-metal electrode and consequently gives a mixed
impedance prole (Z() = Z
+
() + Z
Li
(), as shown in Fig. 2a).
Therefore, these EIS measurements are not suitable for expressing the
accurate internal resistance of the actual porous electrode. In contrast
to the conventional EIS, EIS-SCprovides the impedance spectra of the
true electrode/electrolyte interface without interference from counter
electrodes (Z() = 2Z
+
(), illustrated in Fig. 2b, 2c).
From the temperature dependence of the respective internal resis-
tances obtained by experimental, a kinetics interpretation for faradaic
processes is discussed, and comparisons between the charge-transfer
reaction at the interface and the lithium ion conduction in the bulk
electrolyte and in the electrolyte-lled pores are made.
Mathematical Background
Impedance theory for cylindrical pores by use of the transmis-
sion line model. In the case of a non-faradaic process, the overall
impedance is expressed as Equation (1).
Z

R
i on,L
j C
dl, A
2r
coth

R
i on,L
j C
dl, A
2r L (1)
Figure 3a shows the Nyquist plot for a cylindrical pore calculated
using Eq. (1) when numerical values are given for each parameter. In
the high frequency region, the plot is linear with a 45-degree slope
from the real axis, and transitions to a constant Z

value at low
Journal of The Electrochemical Society, 159 (7) A1034-A1039 (2012) A1035
Figure 2. Schematic cell congurations: (a) asymmetric cell for conventional and classical impedance measurements, (b) symmetric cell without separator
including electrolyte for electric resistance (R
e
) measurements, and (c) symmetric cell with separator including electrolyte for ionic resistance in pores (R
ion
) and
charge-transfer resistance of lithium intercalation reaction (R
ct
) measurements.
frequencies. As shown in Fig. 3a, the limiting values of the real (Z

)
and imaginary parts (Z

) as 0 are shown as follows,


11
Z

0
=
R
i on
3
(2)
Z

0
=
1
C
dl
(3)
where R
ion
is a characteristic parameter that expresses the mobility of
lithium ions inside the porous electrodes.
In the case of a faradaic process, The overall impedance is then
calculated by Eq. (4).
Z

R
i on,L
R
ct, A
(1 + j R
ct, A
C
dl, A
) 2r
coth

R
i on,L
(1 + j R
ct, A
C
dl, A
)2r
R
ct, A
L (4)
The Nyquist plots calculated with Eq. (4) are shown in Fig. 3b
and 3c for R
ct, A
values of 300 cm
2
and 3000 cm
2
, respectively.
At high frequencies, the plots display linear behavior with a 45-degree
slope, and showsemi-circle behavior in the lowfrequency region. The
results in Fig. 3 show that if the same R
ion, L
value is used, Eq. (1) and
(4) give the same result in the high frequency region. The limiting
values of the Z

and Z

components of Eq. (4) as 0 are written


as follows,
11
Z

0
=
R
i on
3
+ R
ct
(5)
Z

0
= 0 (6)
The internal resistances shown in Fig. 3a and 3b are related through
equations (2) and (5). Therefore, the meaning of R
ion
and R
ct
can be
easily understood from the prole of the Nyquist plots. If the value
of R
ion, L
is the same in both cases, when R
ct, A
is increased to large
A1036 Journal of The Electrochemical Society, 159 (7) A1034-A1039 (2012)
Figure 3. Simulated Nyquist plots for a cylindrical pore, L = 1 cm,
r = 0.5 cm, C
dl
= 0.1 mF cm
2
, as predicted by: (a) Eq. (1), R
ion,L
= 100 cm
1
, (b) Eq. (4), R
ion,L
= 100 cm
1
, R
ct,A
= 300 cm
2
,
(c) Eq. (4), R
ion,L
= 100 cm
1
, R
ct,A
= 3000 cm
2
, Frequency range:
100 kHz100 mHz.
values (as shown in Fig. 3c), the Nyquist plot calculated using Eq. (4)
becomes similar to that calculated with Eq. (1).
Experimental Methods
LiNiO
2
-based positive electrodes were prepared by coating a dis-
persion composed of LiNi
0.75
Co
0.15
Al
0.05
Mg
0.05
O
2
, carbon black (as a
conducting agent), and polyvinylidene uoride (PVDF) (as a binder,
Kureha Corp. Japan) (85:10:5 weight ratio) in N-methyl-2-pyrrolidone
(NMP) on aluminum foil. LiNi
0.75
Co
0.15
Al
0.05
Mg
0.05
O
2
was prepared
by a coprecipitation method.
1
The surface morphology of the posi-
tive electrode was examined by scanning electron microscope (SEM,
Hitachi, S-4300). Figure 4 shows the SEMimages of the positive elec-
trode. The SEM images of the electrode indicate that heterogeneous
porous structures are formed by mainly active material (Fig. 4a) or
carbon black (Fig. 4b) particles. Graphite-based negative electrodes
were prepared using the same procedure as for the positive electrodes.
The mixture of graphite and PVDF (90:10 weight ratio) was coated
on a copper foil. Both electrodes were dried at 120

C under vacuum
for at least 10 hrs before construction of the electrochemical cell.
Typical loadings for the positive and negative active materials were
7.0 and 5.0 mg cm
2
, respectively. The electrolyte used in the experi-
ments was 1.0 M LiPF
6
dissolved in a solution of ethylene carbonate
(EC), dimethyl carbonate (DMC), and ethyl methyl carbonate (EMC)
(30/40/30 volume ratio, respectively). A microporous polypropylene
lm was used as a separator.
All electrochemical measurements were performed using
laminate-type pouch cells assembled in an argon-lled drybox. For
EIS measurements of R
e
(Fig. 2b), two identical electrodes were as-
sembled in direct contact (e.g., positive-positive electrodes) without
the separator (electrically non-blocking condition). For EIS measure-
ments of non-faradaic processes to determine R
ion
, identical electrodes
were assembled with the polypropylene separator (electrically block-
ing condition). Prior to EIS measurements of faradaic processes to
determine R
ct
(Fig. 2c), cells with positive and negative electrodes
were prepared with the separator, cycled between 4.1 and 3.0 V at
least once at a low current density of 0.05 mA cm
2
, and maintained
at 3.685 V for an additional 2 hrs for setting the state of charge (SOC)
at 50%. The symmetric cells were then prepared with electrodes taken
from this cell.
Figure 4. SEM images of the LiNiO
2
-based positive electrode surface.
AC impedance measurements (Solatron 1260/1286) were per-
formed at open circuit potential. The frequency was varied from
100 kHz to 100 mHz with a perturbation amplitude of 10 mV (peak
to peak), while the measurements were performed between 30 and
60

C.
Results and Discussion
Impedance of porous electrodes of the symmetric cell. The
impedance behavior of the porous positive electrodes was investigated
in detail. Because the formation of an SEI lmleads to a complex inter-
face on the graphite negative electrodes, the positive electrodes were
selected for use in preliminary impedance experiments. The symmet-
ric cell results for positive electrodes with SOC = 0 and 50% were
characterized as being due to non-faradaic and faradaic processes,
respectively. Figure 5 shows a comparison of the Nyquist plots ob-
tained by experiment and calculation. The experimental data are in
agreement with theory and indicate that the EIS-SC method is not
only able to isolate positive and negative electrode effects, as reported
previously,
12,13
but is also able to produce an ideal impedance spec-
trum that can be expressed by the TLM for cylindrical pores, which
describes the true porous electrode/electrolyte interface.
At SOC = 0% (Fig. 5a), a linear plot with a 45-degree slope
from the real axis was observed in the high frequency region. The
imaginary part of the impedance increases at low frequencies and
the plot approaches a vertical line, which indicates typical electrical
blocking behavior. This behavior suggests that only the formation of
an electrical double layer occurs. No charge-transfer reactions (i.e.
lithium intercalation) occur at the interface despite using the charge-
transfer-active LiNiO
2
electrode. This also means that the electrode is
in an electrochemical equilibrium state without charge-transfer. This
result reects the lithium ion conduction in the bulk electrolyte and in
the electrolyte-lled pores, which are expressed as the electrolyte bulk
resistance (R
sol
) and the ionic resistance in pores (R
ion
), respectively.
At low frequencies, the Nyquist plot does not behave as a perfectly
Journal of The Electrochemical Society, 159 (7) A1034-A1039 (2012) A1037
Figure 5. Nyquist plots for symmetric cells using two positive electrodes in
1.0 M LiPF
6
in EC/DMC/EMC (30/40/30) at 20

C. Electrodes prepared at:


(a) SOC = 0% (squares) and (b) SOC = 50% (circles). The solid lines are the
best-tted results with the equivalent circuits using Eq. (1) and (4), for (a) and
(b), respectively.
vertical line. This may be due to a small leakage current. Previous
work has shown that the leakage current inuence appears in the low
frequency region of the electrical double layer formation and is due
to the non-uniform current distributions caused by pore geometry or
side reactions.
6,8
If the leakage current effect is large, the slope at low
frequencies decreases. From this point of view, the leakage current
effect seems to be negligible for the Nyquist plot obtained in this
study.
While the Nyquist plot at SOC =50% (Fig. 5b) is almost the same
as that seen for SOC = 0% (Fig. 5a) in the frequency region greater
than 20 Hz, the plot appears as a semi-circle at lowfrequencies. Based
on the consideration of theoretical modeling in Fig. 4, the Nyquist plot
contains information on R
sol
and R
ion
in the high frequency region as
well as the charge-transfer resistance for the lithium-intercalation re-
action (R
ct
) (a faradaic process) in the low frequency region. The
overlap of the Nyquist plots in the high frequency region for SOC
=0 and 50% is an interesting and important experimental result. This
result suggests that R
ion
exists independently and can be separated
using the results from R
ct
in the porous electrodes for faradaic pro-
cesses. Hence, R
sol
and R
ion
can be treated independently using the
proposed analytical method in this study. Consequently, the electro-
chemical behavior of faradaic processes for the porous electrode can
be examined in detail. Furthermore, the potential dependence of R
ct
can be accurately determined by controlling the electrode potential
before EIS-SC.
Temperature dependence of individual internal resistances of
porous electrodes. In this section, the temperature dependence of
Nyquist plots is used to build a kinetics interpretation of the individ-
ual internal resistances. Figure 6 shows the temperature dependence
of Nyquist plots between 20 and 60

C at SOC = 0%. As seen in


Fig. 5a, at all temperatures, straight lines with 45-degree slopes are
observed at high frequencies, and near-vertical lines are observed at
low frequencies. The temperature-dependent proles show the typical
electric blocking behavior and can be described by the TLMmodel for
cylindrical pores for non-faradaic processes. With decreasing temper-
ature, the real impedance component at 100 kHz, which determines
the value of R
sol
, shifts toward more positive values and the length of
the 45-degree-slope line, which reects R
ion
, increases. This indicates
that these two internal resistances gradually increase with decreas-
ing temperature. The frequency where the slope changes from 45 to
Figure 6. Nyquist plots for symmetric cells using two positive electrodes at
SOC = 0% in 1.0 M LiPF
6
in EC/DMC/EMC (30/40/30) at various
temperatures.
90 degrees is nearly equal (around 20 Hz) for all temperatures. This
means that the time constant for lithium ion conduction in the bulk
electrolyte and inside pores does not vary with changes in temperature.
Figure 7 shows the temperature dependence of Nyquist plots mea-
sured at SOC = 50%. Similar behaviors seen in Fig. 5b are also
observed at all temperatures. The plots in Fig. 7 can be described
by TLM for cylindrical pores for faradaic processes at all tempera-
tures. The 45-degree-slope lines in Fig. 7 show similar behavior to
those in Fig. 6. This means that R
ion
in faradaic processes exists in-
dependently from R
ct
at all temperatures. The semi-circle in the low
frequency region, which corresponds to R
ct
, becomes larger with de-
creasing temperature. It seems that the temperature dependence of R
ct
is larger than that of R
sol
or R
ion
. In addition, at temperatures higher
than 30

C, a 45-degree-slope, straight line appears at low frequencies


ranging from 100 mHz to 1 Hz. This straight line at very low fre-
quencies corresponds to typical semi-innite diffusion for long-range
ionic diffusion in the bulk of active-electrode materials.
14,15
Figure 7. Nyquist plots for symmetric cells using two positive electrodes at
SOC = 50% in 1.0 M LiPF
6
in EC/DMC/EMC (30/40/30) at various
temperatures.
A1038 Journal of The Electrochemical Society, 159 (7) A1034-A1039 (2012)
Figure 8. Temperature dependences of each internal resistance: (a) electric
resistance (R
e
), (b) electrolyte bulk resistance (R
sol
), (c) ionic resistance in
pores (R
ion
), and (d) charge-transfer resistance of lithiumintercalation reaction
(R
ct
). The data were obtained by tting with Eq. (1) and (4). R
ct
is calculated
using Eq. (5).
Equivalent circuits fromthe TLMfor cylindrical pores model were
used to obtain the quantitative internal resistance parameters from
the experimentally recorded Nyquist plots. As shown in Fig. 5, the
impedance proles could t well with the equivalent circuits using
Eq. (1) and (4). R
sol
, R
ion
, and R
ct
can be obtained separately from
the Nyquist plots using Eq. (5). The inverse of the internal resistances
obtained by tting is plotted as a function of temperature in Fig. 8. R
e
is obtained from measurements performed under the same conditions
using the cell conguration shown in Fig. 2b. R
e
is smaller by 2-4 or-
ders of magnitude when compared with other internal resistances, and
can therefore be ignored. R
ion
is larger than R
sol
, but the temperature
dependency of R
sol
and R
ion
is similar. The relationship between R
sol
and R
ion
is expressed in Eq. (7).
R
i on
R
sol
=
S
nr
2

L
l
(7)
This relationship strongly depends on the ratio of the distance of
lithium ion movement to the contact area. R
ct
is the largest of these
four internal resistances for temperatures between room temperature
and 30

C. However, R
ct
shows a strong temperature dependence,
and becomes smaller than R
sol
and R
ion
above room temperature.
Kinetics interpretation of individual internal resistances of porous
electrodes. To determine the kinetic parameters of interfacial phe-
nomena at porous electrodes for faradaic processes, the activation
energies of each resistance (R
x
) were evaluated using the Arrhenius
equation,
1
R
x
= A exp

E
a
RT

(8)
where A, E
a
, R, and T are the frequency factor, activation energy, gas
constant, and absolute temperature, respectively. All internal resis-
tances display Arrhenius behavior as shown in Fig. 8. The activation
energies for the individual internal resistances are calculated from
Eq. (8) and are listed in Table I. The activation energies for R
sol
(14.9 kJ mol
1
) and R
ion
(16.3 kJ mol
1
) are similar. Although the
individual resistances are different, the concordance of the activation
energies for these resistances implies analogous kinetic processes. The
activationenergies for R
sol
andR
ion
are reective of lithiumionconduc-
Table I. Summary of Activation Energies for the Internal
Resistance at Porous Electrodes.
Internal resistance Activation energy (kJmol
1
)
R
e
0.84
R
sol
14.9
R
ion
16.3
R
ct
57.6
tion (or transport) in the bulk electrolyte and in the electrolyte-lled
pores, respectively. This assignment is based on our experimentally
determined activation energy of 16 kJ mol
1
for the ionic conductivity
in the bulk electrolyte. Previous studies reported that the activation
energies for lithium-ion conduction in a crystalline solid electrolyte,
La
0.55
Li
0.35
TiO
3
, and in a Li-Al-Ti-phosphate-based glass electrolyte
were 2836 kJ mol
1
.
16
Additionally, the lithium ion migration pro-
cesses associated with diffusion of naked lithium ions through an
SEI lm layer on graphite negative electrodes was shown to have an
activation energy of 2025 kJ mol
1
.
17
Therefore, the values of the
activation energies for R
sol
and R
ion
obtained in this experiment are at-
tributable to lithium ion conduction in the liquid phase. Furthermore,
the agreement of the R
sol
and R
ion
activation energies indicates a nearly
energy-free kinetic barrier for transport of lithium ions at the interface
between the bulk electrolyte and inside the electrolyte-lled pores.
The activation energy for R
ct
(57.6 kJ mol
1
) is higher than those of
other resistances. This large value is consistent with previous reports
that lithium ion transfer at several solid/liquid interfaces is affected
by lithium ion desolvation.
18,19
Thus, a higher kinetic barrier exists at
the solid/liquid interface than exists for moving from the electrolyte
bulk to electrolyte inside the pore. Simple comparison of the internal
resistance components at porous positive electrodes reveals that the
ordering of the activation energies is as follows:
E
a,ct
> E
a,ion
E
a,sol
>> E
a,e
.
Importance of R
ion
on electrochemical behavior in porous
electrodes. Based on the above results, R
ct
is the dominant com-
ponent of the total internal resistance. However, R
ion
is one of the
characteristic parameters of porous electrodes that is expected to be
dependent on the electrode geometry. Here, we consider thick porous
electrodes based on the TLM theory for cylindrical pores. If the re-
sistance per unit area (R
ion, L
and R
ct, A
) remains constant and L is
increased, R
ion
increases while R
ct
decreases (see list of symbols).
Furthermore, from Eq. (7), R
ion
becomes greater than R
sol
if the elec-
trode thickness (L) becomes larger than the characteristic lithium ion
movement distance in the bulk electrolyte (l). This prediction suggests
that R
ion
will be expected to become the largest resistance component
of the examined internal resistances, and is expected to have a large
effect on the total internal resistance in thicker electrodes. We strongly
believe that R
ion
should be a signicant factor for porous electrodes
thicker than those examined in the current study. Such effects on the
thick electrode will be discussed in detail in subsequent reports.
Conclusions
In this paper, the individual internal resistance components of
the porous electrode/electrolyte interface have been investigated in
detail by combining the transmission line model (TLM) theory for
cylindrical pores with electrochemical impedance spectroscopy using
symmetric cells (EIS-SC). With this approach, we have succeeded
in separating four internal resistances at the porous electrode as the
following parameters: R
e
, R
sol
, R
ion
, and R
ct
. From the temperature
dependence of these resistances, a kinetics interpretation is given to
the comparison between the charge-transfer reaction at the interface
and the lithium ion conduction inside pores for faradaic processes.
The R
ion
obtained is a characteristic parameter of the actual porous
electrode/electrolyte interface that is important to consider for thick
electrodes. The results show that despite its simplicity, our approach
Journal of The Electrochemical Society, 159 (7) A1034-A1039 (2012) A1039
will give reasonably accurate results and is sufcient for analysis
of the electrochemical characteristics of actual porous electrodes for
lithium-ion batteries.
List of Symbols
L unit pore length, cm
L characteristic distance of lithium ion movement in the bulk
electrolyte, cm
r pore radius, cm
electrolyte resistance, cm
S electrode surface area, cm
2
n number of pores per unit electrode surface area
Z
i, L
interfacial impedance per unit pore length, cm
Z
i, A
interfacial impedance per unit surface area, cm
2
R
e
electric bulk resistance,
R
sol
electrolyte resistance (R
sol
= l / S ),
R
ion, L
ionic resistance per unit pore length (R
ion, L
=/r
2
), cm
1
R
ion
ionic resistance in pores (R
ion
= R
ion, L
L / n),
R
ct, A
charge-transfer resistance per unit surface area, cm
2
R
ct
total charge-transfer resistance (R
ct
= R
ct, A
/ 2rL),
C
dl, A
electric double layer capacitance per unit surface area, Fcm
2
C
dl
total electric double layer capacitance (C
dl
= C
dl, A
2rL), F
References
1. H. Kondo, Y. Takeuchi, T. Sasaki, S. Kawauchi, Y. Itou, O. Hiruta, C. Okuda,
M. Yonemura, T. Kamiyama, and Y. Ukyo, J. Power Sources, 174, 1131 (2007).
2. M. Doyle, J. P. Meyers, and J. Newman, J. Electrochem. Soc., 147, 99 (2000).
3. J. P. Meyers, M. Doyle, R. M. Darling, and J. Newman, J. Electrochem. Soc., 147,
2930 (2000).
4. S. Devan, V. R. Subramanian, and R. E. White, J. Electrochem. Soc., 151, A905
(2004).
5. J. H. Jang and S. M. Oh, J. Electrochem. Soc., 151, A571 (2004).
6. J. H. Jang, S. Yoon, B. H. Ka, Y.-H. Jung, and S. M. Oh, J. Electrochem. Soc., 152,
A1418 (2005).
7. P. Los, A. Lasia, H. Menard, and L. Brossard, J. Electroanal. Chem., 360, 101 (1993).
8. C. Hitz and A. Lasia, J. Electroanal. Chem., 500, 213 (2001).
9. R. D. Levie, Electrochim. Acta, 8, 751 (1963).
10. R. D. Levie, Electrochim. Acta, 9, 1231 (1964).
11. M. Itagaki, S. Suzuki, I. Shitanda, and K. Watanabe, Electrochemistry, 75, 649 (2007).
12. C. H. Chen, J. Liu, and K. Amin, J. Power Sources, 96, 321 (2001).
13. C. H. Chen, J. Liu, and K. Amin, Electrochem. Commun., 3, 44 (2001).
14. F. Nobili, F. Croce, B. Scrosati, and R. Marassi, Chem. Mater., 13, 1642 (2001).
15. Q.-C. Zhuang, T. Wei, L.-L. Du, Y.-L. Cui, L. Fang, and S.-G. Sun, J. Phys. Chem.
C, 114, 8614 (2010).
16. T. Abe, F. Sagane, M. Ohtsuka, Y. Iriyama, and Z. Ogumi, J. Electrochem. Soc., 152,
A2151 (2005).
17. K. Xu, Y. Lam, S. S. Zhang, T. R. Jow, and T. B. Curtis, J. Phys. Chem. C, 111, 7411
(2007).
18. I. Yamada, T. Abe, Y. Iriyama, and Z. Ogumi, Electrochem. Commun., 5, 502 (2003).
19. Y. Yamada, Y. Iriyama, T. Abe, and Z. Ogumi, Langmuir, 25, 12766 (2009).

You might also like