See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/258838810

Reduction of the Internal Resistance of Carbon Electrodes for an Electric

Double-layer Capacitor (EDLC)

Conference Paper · May 2012

CITATIONS READS

2 7,193

4 authors, including:

Chunhong Lei Foivos Markoulidis

University of Southampton University of Surrey
69 PUBLICATIONS   1,606 CITATIONS    59 PUBLICATIONS   903 CITATIONS   

SEE PROFILE SEE PROFILE

Constantina Lekakou
University of Surrey
123 PUBLICATIONS   3,707 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

LiSTAr (The Faraday Institution) View project

Development and manufacture of transparent, electrically conductive, flexible plastics (IeMRC/EPSRC) View project

All content following this page was uploaded by Chunhong Lei on 10 July 2014.

The user has requested enhancement of the downloaded file.

 Electrochimica Acta 92 (2013) 183–187

Contents lists available at SciVerse ScienceDirect

Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta

Reduction of porous carbon/Al contact resistance for an electric double-layer

capacitor (EDLC)
Chunhong Lei a , Foivos Markoulidis a , Zenya Ashitaka b , Constantina Lekakou a,∗
a
Division of Mechanical, Medical, and Aerospace Engineering, Faculty of Engineering and Physical Sciences, University of Surrey, Guildford GU2 7XH, UK
b
Toyo Aluminium K.K., 6-8, Kyutaro-machi 3-chome, Chuo-ku, Osaka 541-0056, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Carbonaceous materials are commonly used to fabricate electrodes for electric double-layer capacitors
Received 20 September 2012 (EDLCs) or supercapacitors. The high contact resistance between the carbon active layer and the Al current
Received in revised form collector can decrease capacitor energy and power performance, and shorten the lifetime of the capacitor.
18 December 2012
In this report, the sources of carbon based EDLC internal resistance were explored using electrochemical
Accepted 22 December 2012
impedance spectroscopy (EIS). An equivalent circuit model was coupled with the EIS data for the analyses.
Available online 31 December 2012
The EDLC cells were made from symmetric carbon/Al electrodes and operated in organic electrolyte. The
analysis results showed the effects of pressure and modified Al on the contact resistance, where a novel,
Keywords:
EDLC supercapacitor
carbon modified Al collector with Al4 C3 nano whiskers greatly reduced the contact resistance. Finally the
Carbon/Al contact resistance effect of scale-up on the internal resistance was discussed.
Al4 C3 nano whisker © 2013 Elsevier Ltd. All rights reserved.

1. Introduction being environmentally friendly. A porous carbon electrode consists

of activated carbon powder, conductive additive such as carbon
Electrochemical double layer capacitors (EDLCs) [1] also known black or carbon nanotubes, and a polymer binder which cements
as supercapacitors or ultracapacitors, are fast energy storage and all the components together and onto a metal current collector,
delivery devices, with power and energy density performance usually Al foil.
between traditional dielectric capacitors which have high power The three major contributions to EDLC internal resistance arise
output, and batteries/fuel cells which have high energy storage. from the electrolyte, the active-layer and the contact between
They can be used as alternative or complementary power sources in current-collector and active-layer [10,11]. The electrolyte resis-
various applications including telecommunication devices, standby tance is associated with diffusion of electrolyte ions within the
power systems, portable electronic devices and electric hybrid pores of the active-layer. The resistance within the framework
vehicles [2]. of active-layer is electronic resistance, including resistances from
Since EDLCs accumulate charges on high surface-area conduct- the carbon material, carbon particle–particle contact and poly-
ing materials such as activated carbon via the electrostatic force, mer binder. The direct effect of high internal resistance of the
no faradic reaction in the charge storage process is involved, so carbon electrode is the reduction of capacitor energy output and
that supercapacitors often exhibit higher power capability and power delivery, the heat produced inside the cell and the resulted
much higher recycle lifetime (>100,000 charge/discharge cycles) short lifetime of the capacitor. Methods to reduce the internal
than rechargeable batteries. However, up to now energy density resistance include choosing suitable carbon particle size and com-
of EDLCs is only about one-tenth of the lowest density battery pactness [12], adding conductive additives [10,13], and modifying
[3,4] and hence much of today’s EDLC research is concerned with the current collector to reduce the carbon-metal contact resistance
increasing its energy density at minimum sacrifice to power capa- [11,14].
bility [5,6]. This report explores the sources of internal resistance in an
Carbonaceous materials in the form of aerogel [7], powder [8] EDLC cell, and focuses on the origin and reduction of carbon/Al
and fibre are the most commonly used materials for electrodes in contact resistance. Carbon/Al electrodes were made with activated
EDLC devices due to carbon’s high surface area (up to 2500 m2 /g carbon, carbon black conductive additive and polymer binder
[9]), chemical and temperature stability, relatively low cost and PVDF. Symmetric EDLC cells were made from the electrodes.
1 M tetraethylammonium tetrafluoroborate (TEABF4 ) organic
solution was used as electrolyte. Electrochemical impedance spec-
∗ Corresponding author. troscopy (EIS) and galvanostatic charge–discharge measurements
E-mail address: [email protected] (C. Lekakou). were applied to characterize carbon electrodes in EDLC cells. An

0013-4686/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.electacta.2012.12.092
184 C. Lei et al. / Electrochimica Acta 92 (2013) 183–187

equivalent circuit model was used to explain the EIS results, with Nyquist plot, two resistances, solution resistance Rs and charge
each component assigned to the origin of the resistance. The effect transfer resistance Rct , can be obtained immediately. Rs comes from
of scale-up of the supercapacitor from 2 cm2 to 18 cm × 18 cm on the bulk solution resistance, its magnitude depends on the elec-
these resistances was investigated. The study clearly showed the trolyte conductivity, the material and thickness of the separator.
origin of the internal resistance and its effect on the energy density Rct is the charge transfer resistance, which can be separated into
of the device. This study may provide guidance for optimizing the two components: the electronic and the ionic resistances [15]. The
fabrication of supercapacitors. electronic resistance includes the intrinsic electronic conductivity
of the carbon particles, the electronic contact between particles,
2. Experimental details and the contact between the active layer and the current collec-
tor. The ionic resistance is the electrolyte ionic resistance inside
2.1. Fabrication of capacitors the pores of the electrode. It depends on the electrolyte conduc-
tivity, porous texture of the electrode and the thickness of active
The composite carbon electrode were prepared from a blend of layer. Cc is the interfacial contact capacitance between carbon and
AC (Sigma–Aldrich, specific surface area 1000 m2 /g), acetylene car- Al foil. The semicircle comes from the parallel RC branch of Cc and
bon black (CB) (Alfa Aesar) and poly(vinylidene fluoride) (PVDF). All Rct . From the Nyquist plot, the internal resistance of the cell, Rint ,
the chemicals were mixed at a weight ratio of AC:CB:PVDF = 90:5:5 can be estimated from the intersection between the low frequency
in l-methyl-2-pyrrolidinone (NMP) to form a slurry, the slurry and more straight section of the Nyquist plot and the X-axis [16,17].
was then coated onto an aluminium foil, and dried in an oven at The capacitance of the cell can be estimated by C = 1/2␲fZim through
150 ◦ C for 2 h. EDLC cells were fabricated using a symmetrical elec- the value of Zim at the low frequency end.
trode configuration. A piece of filter paper soaked with electrolyte GCD tests were carried out on the cells in a potential range of
solution was placed between the carbon electrodes to act as a sep- 0–3 V at various current densities. Fig. 1b displays a typical plot of
arator. The electrode working area was fixed to 2 cm2 in all the the GCD curve of an EDLC cell. From the discharge curve, the device
devices. The device assembly was sealed in a plastic bag by vac- energy density E and power density P can be obtained according to
uum. The electrolyte was 1 M TEABF4 (Sigma–Aldrich, purity 99%) Eq. (1) [18,19] shown in the inset of Fig. 1b. The internal resistance
in PC (Sigma–Aldrich, anhydrous, purity 99.7%). can be determined more reliably from the IR drop at the initiation of
a constant current discharge, or measuring the bounce back of volt-
age at the end of a constant current discharge [13,20]. If there was
2.2. Electrochemical characterization of capacitors
no internal resistance, the IR drop would be zero, and the energy
and power density (E0 , P0 ) would be bigger than (E, P), as shown in
The performance of different EDLC cells was characterized by
Eq. (2) in the inset of Fig. 1b.
electrochemical impedance spectroscopy (EIS) and galvanostatic
charge–discharge (GCD) tests, using a VersaSTAT MC analyser. The
devices were characterized in a 2-terminal configuration; no ref- 3.2. Pressure effect on contact resistance
erence electrode was used in the tests. During the electrochemical
tests, a 2 kg weight was put on the body of EDLC cell to ensure a The carbon/Al contact resistance is due to low surface contact,
fixed gap between the two electrodes. poor adhesion of carbon material to Al surface, and the intrinsic
aluminium oxidation layer formed on the top surface of the Al foil.
3. Results and discussion A common way of reducing the internal resistance is to apply and
keep a compression force on the body of the cell to minimize the
3.1. EIS and GCD tests gap between the two electrodes and improve the carbon/Al contact.
To examine the effect of pressure on the internal resistance, a
The EIS measurements on the cells were carried out at a DC series of weights were placed on the cell body and EIS measure-
bias of 0.2 V with sinusoidal signal of 20 mV over a frequency range ments were carried out under each weight. Fig. 2 shows the Nyquist
from 1 MHz to 0.01 Hz. Fig. 1a shows a typical Nyquist plot of an plots of an EDLC cell at 4 different pressures. The carbon electrode
EDLC cell. Coupled with an equivalent circuit, the individual com- for the cell had a carbon loading of 4.5 mg/cm2 , corresponding to a
ponents of the impedance and origin can be found out. From the thickness of 140 ␮m.

Fig. 1. (a) A typical Nyquist plot for an EDLC cell. Inset diagram is its equivalent circuit. (b) A typical galvanostatic charge–discharge curve of the EDLC cell at constant current
I, between 0 V and 3 V. (E, P) and (E0 , P0 ) are energy and power with and without IR drop respectively.
 C. Lei et al. / Electrochimica Acta 92 (2013) 183–187 185

[21,22]. Too much compression force may alter the pore distribu-
tion/morphology of electrode, resulting in a displacement of the
electrolyte out of the more restricted pores, and reduction in the
accessible surface area of the carbon material due to the restriction
of the transport channels.

3.3. Effect of carbon modified current collector on the contact

resistance

In order to examine the influence of the carbon/Al contact

resistance on the internal resistance of an EDLC device, a carbon
modified Al current collector electrode was applied. The carbon
modified Al foil, Toyal-Carbo® , has a thin carbon black layer coat-
ing held with Al4 C3 nano whiskers (Fig. 3). The Al4 C3 nanometer
scale whiskers were grown from underneath the Al oxide layer,
hence, they not only provide electric conducting channels across
the oxide layer, but also serve as binder to hold carbon particles to
Al surface.
Two groups of electrodes were made with various carbon load-
ings, coated on either bare Al foil or Toyal-Carbo® . Shown in Fig. 4a
are the Nyquist plots of the EDLC cells made from these two groups
Fig. 2. Nyquist plots of EDLC cells with a carbon loading of 4.5 mg/cm2 , obtained of electrodes. For the cells on bare Al foil, the semicircles were
under different pressures. visibly big for all carbon loadings; the diameter of the semicircle
increased with the carbon loading/thickness, indicating increased
charge transfer resistance Rct , as it took longer time for ionic charges
From the Nyquist plots, it can be seen that the semicircle to transfer into/from a thicker carbon electrode. The solution resis-
decreased with increasing pressure, while the Rs was almost tance Rs of these electrodes was almost the same because the only
unchanged. This indicates that the pressure on the cell can reduce difference in these cells was the thickness of active carbon layer. The
the charge transfer resistance, and has little effect on the solution capacitance of these cells estimated from the Nyquist plots shows
resistance. The pressure on the cell may reduce the gap between increasing with the carbon loading. The cells on Toyal-Carbo® cur-
the two electrodes and the thickness of the carbon electrode, but rent collector showed a different pattern. The semicircles are so
this reduction is limited after the initial pressure of 2 kg weight, as small that it is hard to be discerned even at a large carbon loading of
can be seen by the small change on Rs as the weight increased. As 8.4 mg/cm2 , though the Rs is almost the same as on bare Al. The table
the thickness of the carbon layer could hardly change under these in Fig. 4a summarizes the resistances from these two groups of elec-
pressures, the reduction of Rct came mainly from the improved trodes. As can be seen, the carbon modified Al current collector has
carbon/Al contact under pressure. As the weight increased, the car- greatly reduced the overall charge-transfer resistance by lowering
bon/Al contact area increased due to the deformation of Al around the interfacial resistance. The carbon/Al contact on Toyal-Carbo®
the irregular carbon particles, or due to the fact that the electrolyte was thus ohmic contact: the interfacial capacitance diminished as
infiltrated between carbon and Al was squeezed out if the adhesion there was no charge accumulated on the ohmic contact interface.
of carbon layer onto the Al surface was weak. So the RC branch in the equivalent circuit was left with only the
The reduction of carbon/Al contact resistance stopped beyond a Warburg impedance Zw . At the right side of the semicircle, a ∼45◦
certain pressure, 8 kg in this case. This implies that the carbon/Al region was observed, which is the consequence of the frequency
contact resistance cannot be eradicated by increasing the contact dependence of ion diffusion into the bulk of the electrode [23].
area. No matter how good the contact is, the native Al oxidation GCD tests were carried on these cells at various current den-
layer is still there to limit the conductivity. It can be seen that the sities. Fig. 4b displays a charge–discharge curve for each of these
EDLC cell capacitance increased with increasing pressure, possi- cells at a current of 50 mA. All cells showed linear charge–discharge
bly due to improved electric connection between carbon particles. curves, suggesting that the electrodes tested have no obvious redox
However, this increase of capacitance under pressure is limited reaction. The IR drop at the turning point of the charge–discharge

Fig. 3. (a) TEM image of the nano-scale aluminium carbide whiskers on the Toyal-Carbo® foil. (b) Schematic diagram of the composite of the Toyal-Carbo® electrode.
186 C. Lei et al. / Electrochimica Acta 92 (2013) 183–187

Fig. 4. (a) Nyquist plots of two groups of EDLC cells with Al and Toyal-Carbo® current collectors, at different carbon loading. The table below is a list of resistance data of
these cells. (b) Galvanostatic charge–discharge curve of the two groups of EDLC cells at a current of 50 mA, and (c) corresponding Ragone plots of several of the EDLC cells.

line, can be seen due to the internal resistance Rint . The data in the
table in Fig. 4a show that the IR drop changes with carbon loading,
and Rint is much smaller on the Toyal-Carbo® due to the reduced
carbon/Al contact resistance. The Ragone plots of some cells are
shown in Fig. 4c. This plot demonstrates that the energy den-
sity of the cells decreases with increasing carbon loading towards
the high power side of the plot. Comparison between electrodes
with bare Al and Toyal-Carbo® current collector indicates that
reduction of interfacial resistance and hence the internal resis-
tance has increased the energy density in the whole power region,
as indicated by the (E, P)–(E0 , P0 ) relationship in the equation
in Fig. 1b.

3.4. Contact resistance in scaled up EDLC

EDLC cells with larger electrode area were made on bare Al foil. A
18 cm × 18 cm cell was clamped between two stainless steel plates
by steel beams with bolts. The tightening was by hand. Although
the pressure exerted on the cell body was unknown, the resulted
EIS measurements showed that the device parameters became
unchanged beyond a certain point of pressure.
In the Nyquist plot shown in Fig. 5, all the resistances were
very small, the semicircle cannot be identified. This is due to the
fact that the resistance is reversely proportional to the electrode
area. So that if every resistance is normalized by multiplying by
Fig. 5. Nyquist plot of a 18 cm × 18 cm EDLC cell with a carbon loading of 4.5 mg/cm2
the single electrode area S, it should be constant no matter what
on Al. The table underneath is a list of resistance data of the 2 cm2 cell and the
the area is. Shown in the table in Fig. 5 are the normalized resis- 18 cm × 18 cm cell.
tances of the EDLC cells with an area of 2 cm2 and 18 cm × 18 cm,
both had the same carbon loading. The Rct ·S for 18 cm × 18 cm cell
cannot be defined, while Rs ·S from EIS and Rint ·S from GCD mea- 4. Conclusions
surements were bigger in the 18 cm × 18 cm than in 2 cm2 cell. This
is because as the electrode area was enlarged, the coating unifor- The internal resistance of the EDLC devices comes from the ionic
mity was unavoidably compromised. Since the gap between carbon resistance in electrolyte and the separator, the electronic resistance
electrodes, where electrolyte reservoir and separator were located, in the carbon active-layer, and the electronic resistance at the inter-
was uneven in the large area cell, this resulted in an increased Rs ·S. face between current collector and active layer. The existence of
In addition, the pressure on the enlarged cell could also be non- internal resistance reduces the device energy and power perfor-
uniform, The RC circuit thus consisted of many RC branches with mance. EIS combined with an equivalent circuit model was used to
various R and C values as they are dependent on both the carbon analyze the effects of every component of the device on the internal
layer thickness and the pressure on it. All these RC branches were resistance, aimed to reduce the internal resistance. The carbon/Al
connected in parallel, and resulted in a smeared semicircle in the contact resistance was shown to be greatly reduced by a carbon
Nyquist plot. modified Al current collector, where Al4 C3 nano whiskers provided
 C. Lei et al. / Electrochimica Acta 92 (2013) 183–187 187

electric conducting channels and served as binder to hold carbon [9] A. Tanimura, A. Kovalenko, F. Hirata, Molecular theory of an electrochemical
particles to the Al surface. The reduction of contact resistance has double layer in a nanoporous carbon supercapacitor, Chemical Physics Letters
378 (2003) 638.
improved the cell energy and power performance. The solution [10] C. Dangler, M.R.-Fondacaro, T.S. Devarajan, S. Higashiya, Jeremy Snyder,
resistance Rs and charge transfer resistance Rct was shown to be Pradeep Haldar, Role of conducting carbon in electrodes for electric double
dependent on the thickness and pressure, respectively. As the EDLC layer capacitors, Materials Letters 65 (2011) 300.
[11] H.-C. Wu, Y.-P. Lin, E. Lee, W.-T. Lin, J.-K. Hu, H.-C. Chen, N.-L. Wu, High-
cell was scaled up, its normalized resistances were also increased performance carbon-based supercapacitors using Al current-collector with
as a result of non-uniformity in electrode thickness and pressure conformal carbon coating, Materials Chemistry and Physics 117 (2009)
on the cell’s body. This method of analyzing internal resistance is 294.
[12] A.G. Pandolfo, G.J. Wilson, T.D. Huynh, A.F. Hollenkamp, The Influence of Con-
useful in seeking suitable materials for the EDLC device.
ductive Additives and Inter-Particle Voids in Carbon EDLC Electrodes, Fuel Cells
10 (2010) 856.
Acknowledgments [13] Y.J. Kim, Y.A. Kim, T. Chino, H. Suezaki, M. Endo, M.S. Dresselhaus, Chemically
Modified Multiwalled Carbon Nanotubes as an Additive for Supercapacitors,
Small 2 (2006) 339.
This research was funded by the EPSRC and the European Com- [14] C. Portet, P.L. Taberna, P. Simon, C. Laberty-Robert, Modification of Al current
mission FP7 project “AUTOSUPERCAP”. The authors would like to collector surface by sol–gel deposit for carbon–carbon supercapacitor applica-
tions, Electrochimica Acta 49 (2004) 905.
thank Professor Joe Keddie’s laboratory in the Physics Department
[15] P.L. Taberna, C. Portet, P. Simon, Electrode surface treatment and electrochem-
at the University of Surrey for the courtesy of facility usage. ical impedance spectroscopy study on carbon/carbon supercapacitors, Applied
Physics A82 (2006) 639.
[16] M.D. Stoller, R.S. Ruoff, Best Practice Methods for Determining an Electrode
References
Material’s Performance for Ultracapacitors, Energy and Environmental Science
3 (2010) 1294.
[1] V. Khomenko, E. Raymundo-Pi˜nero, F. B’eguin, High-energy density [17] S. Zhao, F. Wu, L. Yang, L. Gao, A.F. Burke, A measurement method for determina-
graphite/AC capacitor in organic electrolyte, Journal of Power Sources tion of dc internal resistance of batteries and supercapacitors, Electrochemistry
177 (2008) 643. Communications 12 (2010) 242.
[2] A. Burke, Ultracapacitors: why, how, and where is the technology, Journal of [18] T. Bordjiba, M. Mohamedi, L.H. Dao, Synthesis and electrochemical capacitance
Power Sources 91 (2000) 37. of binderless nanocomposite electrodes formed by dispersion of carbon nano-
[3] Y. Zhang, H. Feng, X. Wu, L. Wang, A. Zhang, T. Xia, H. Dong, X. Li, L. Zhang, tubes and carbon aerogels, Journal of Power Sources 172 (2007) 991.
Progress of electrochemical capacitor electrode materials: A review, Interna- [19] C. Arbizzani, M. Mastragostino, F. Soavi, New trends in electrochemical super-
tional Journal of Hydrogen Energy 34 (2009) 4889. capacitors, Journal of Power Sources 100 (2000) 164.
[4] L.L. Zhang, X.S. Zhao, Carbon-Based Materials as Supercapacitor Electrodes, [20] A. Izadi-Najafabadi, S. Yasuda, K. Kobashi, T. Yamada, D.N. Futaba, H. Hatori,
Chemical Society Reviews 38 (2009) 2520. M. Yumura, S. Iijima, K. Hata, Extracting the Full Potential of Single-Walled
[5] A. Burke, R&D Considerations for the Performance and Application of Electro- Carbon Nanotubes as Durable Supercapacitor Electrodes Operable at 4 V with
chemical Capacitors, Electrochimica Acta 53 (2007) 1083. High Power and Energy Density, Advanced Materials 22 (2010) E235.
[6] C. Lei, P. Wilson, C. Lekakou, Effect of poly(3,4-ethylenedioxythiophene) [21] G. Gourdin, T. Jiang, P. Smith, D. Qu, The effects of cell assembly compression on
(PEDOT) in carbon-based composite electrodes for electrochemical supercap- the performance of carbon electrochemical double-layer capacitor electrodes,
acitors, Journal of Power Sources 196 (2011) 7823. Journal of Power Sources 215 (2012) 179.
[7] B. Fang, L. Binder, A modified activated carbon aerogel for high-energy storage [22] X. Li, J. Rong, B. Wei, Electrochemical behavior of single-walled carbon nanotube
in electric double layer capacitors, Journal of Power Sources 163 (2006) 616. supercapacitors under compressive stress, ACS Nano 4 (2010) 6039.
[8] C.-C. Hu, C.-C. Wang, F.-C. Wu, R.-L. Tseng, Characterization of pistachio shell- [23] C. Masarapu, H.F. Zeng, K.H. Hung, B. Wei, Effect of temperature on the capaci-
derived carbons activated by a combination of KOH and CO2 for electric double- tance of carbon nanotube supercapacitors, ACS Nano 3 (2009) 2199.
layer capacitors, Electrochimica Acta 52 (2007) 2498.

View publication stats