Journal of

Mechanical
Science and
Technology
Journal of Mechanical Science and Technology 23 (2009) 75~80
www.springerlink.com/content/1738-494x
DOI 10.1007/s12206-008-0911-x

Conductivity of carbon nanofiber/polypyrrole

conducting nanocomposites†
Cheol Kim* and Shuai Zhang
Department of Mechanical Engineering, Kyungpook National University, Daegu, 702-701, Korea

(Manuscript Received April 1, 2008; Revised August 26, 2008; Accepted September 11, 2008)
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Abstract

Carbon nanofiber (CNF) / Polypyrrole (Ppy) composite materials were fabricated by two newly invented processes -
filtering, washing and drying the mixture of CNF dispersion and Ppy-NMP solution (FWP process) and heating an
aqueous solution of CNF (SH process). CNF/Ppy composite materials have never been reported before in any other
research papers. Conductivities of the composite films were obtained by using a four-probe method. To compare the
conductivity of CNF/Ppy with that of a pure single-walled carbon nanotube (SWNT), SWNT films were also fabri-
cated and voltage was measured. SEM images were taken for both a surface and a cross-section of composite samples
fabricated by the two processes. The CNF/Ppy by FWP was a little brittle because of the low solubility of Ppy in the
NMP, and on the other hand, the same material by SH became flexible enough. The conductivity of the pure SWNT
film was as high as double the similar case. The conductivity of the pure SWNT film was 20.11 S/cm and 0.013 cm
thick. The CNF/Ppy composite films with the thicknesses of 0.062 cm and 0.085 cm gave a conductivity of 63.32 S/cm
and 40.57 S/cm, respectively, which are higher than that of the pure SWNT film or SWNT/Polyaniline (PANi) film.
The good conductivity of CNF/Ppy composites shows the improved potential for developing the materials for a small
actuator.
Keywords: Carbon nanofiber; Conductivity; Nanocomposite; Polypyrrole
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

and low actuating force, which restrict the broad ap-

1. Introduction
plication of CPs.
Conducting polymers (CPs) have many attractive In an effort to make the characteristics of CPs more
properties such as electro-chromism, electro-lu- suitable for electromechanical devices, people have
minescence, chemical vapor/gas sensitivity, piezo- started to focus on some other new materials that
sensitivity, electromechanical response, and optical have good electromechanical properties and can be
switching which make them potentially applicable to easily compounded of CPs. Since its first discovery in
artificial muscles or small actuators. CPs may be op- 1991 [4], the carbon nanotube (CNT) has shown re-
erated at less than 10 V and generate strains of 12% markable properties which make it an attractive mate-
[1]. Stresses induced from CP actuators are around 5 rial in a broad range of applications. Due to the excel-
MPa [2], which is much higher than 0.3 MPa [3] from lent strength and stiffness of CNT, it can make up for
a natural muscle. However, there are some disadvan- the weakness of CPs once when it is compounded. In
tages, i.e., low Young’s modulus, low conductivity, addition, its high conductivity also attracts many re-
†This paper was recommended for publication in revised form by searchers for the improvement of conductivity in
Associate Editor Maenghyo Cho some polymers which are non-conducting or low
*
Corresponding author. Tel.: +82 53 950 6586, Fax.: +82 53 950 6550
E-mail address: [email protected] conducting. The CNT itself is also a good candidate
© KSME & Springer 2009 for an actuator. The CP volume change was first re-
76 C. Kim and S. Zhang / Journal of Mechanical Science and Technology 23 (2009) 75~80

ported by Burgmayer and Murray [5]. Ionic move- agglomerates of SWNTs. From these figures, it is
ment and subsequent volume changes in CPs are the shown that the length of each SWNT is about 200nm
fundamental of the motion in CP actuators. and the diameter is less than 10nm. The agglomerates
Tahhan et al. [6] compounded single-walled carbon can be formed by bonds between SWNTs during the
nanotubes (SWNTs)/Polyaniline (PANi) composite process.
materials and tested their properties for actuators.
Kim and Liu [7] developed the strain-voltage rela- 2.2 Fabrication of CNF/Ppy composites
tionship of a CNT/CP composite film-type actuator
and improved the fabrication process of a SWNTs/ Straight type CNF with a purity of 95% was dis-
PANi composite actuator. solved into a TritonX100 1% (w/w) surfactant solu-
Carbon nanofiber (CNF) has similar properties to tion, then sonicated with the sonic dismembrator for
CNT, and is much cheaper than CNT, owing to the more than 6 hours and the powder of Ppy was dis-
manufacturing process. Therefore, the CNF can be a solved in NMP (N-methyl Pyrrolidine) to make a
good substitute for CNT. The remarkable properties solution. Two solutions were mixed together with the
of CNF such as high stiffness and strength, large weight ratio 3:1 of Ppy (150 mg) and CNF (50 mg)
thermal and electrical conductions were reported re- and then the mixture was filtered through a previously
cently [8, 9]. In the present work, a pure SWNT film wetted PVDF membrane in an ethanol/distilled water
was fabricated and its conductivity was obtained by a mixture (50:50). The methanol was also applied for
four-probe method. A CNF/ Polypyrrole (Ppy) com- removing the TritonX100 and a filtration assembly
posite film was then made by a chemical blending and a vacuum pump were applied for filtering and
method and its electrical conductivity was investi- washing processes. The extracted sample film was
gated and compared with that of a pure SWNT film. then soaked in HCl (1mol/L) to make a form of com-
Up to now, the combined CNF and Ppy composite posites. Thereafter, the film was dried in the air be-
material has not been compounded because Ppy is fore peeling off. This whole fabrication process,
hard to dissolve into a solution. Naturally the conduc-
tivity of CNF/Ppy has been rarely investigated, even
though CNF/Ppy is potentially a good material for a
small actuator.

2. Fabrication of films
2.1 Fabrication of pure SWNT films

SWNTs with the purity of 90% were utilized.

SWNTs (80mg) and 1g of TritonX100 1% (w/w)
surfactant solution was added to 50ml of distilled
water. This dispersion was sonicated by a sonic dis- Fig. 1. SEM image of the surface of a pure CNT film.
membrator for more than 6 hours. The purpose of
sonication is to release SWNTs from twisting each
other. The dispersion was filtered through a previ-
ously wetted PVDF (Polyvinylidene Fluoride)
membrane in an ethanol/distilled water mixture
(50:50). Then, methanol was applied for removing the
TritonX100. The filtering and washing processes
were done by using a filtration assembly and a vac-
uum pump. Finally, the SWNT film was dried natu-
rally before peeling off.
Figs. 1 and 2 show SEM pictures of the surface and
cross-section of the pure SWNT film. The slender and
fiber-like one is an SWNT and large particles are Fig. 2. SEM image of the cross-section of a pure CNT film.
 C. Kim and S. Zhang / Journal of Mechanical Science and Technology 23 (2009) 75~80 77

Fig. 3. CNF/Ppy composite samples fabricated by an SH

process. Fig. 4. SEM image of the surface of a composite film fabri-
cated by an FWP process.

which is basically based on filtering and washing with

powder (FWP) called a FWP process takes at least 70
hours in a laboratory. The composite films fabricated
by the process seemed to be a little brittle. The com-
posite sample was split into several pieces of itself
after being dried in an ambient air at room tempera-
ture. This phenomenon is due to the fact that Ppy is
usually difficult to dissolve in any solvent. The solu-
tion of Ppy in NMP is more like a dispersion.
In order to improve the FWP process which takes
more than 70 hours and results in a little brittle mate-
rial, an improved process based on a solution and Fig. 5. SEM image of the cross section of a composite film
heating (SH) process was invented. Because of the fabricated by an FWP process.
low solubility of Ppy, a Ppy solution (100ml, 5%,
1g/ml) was used instead of powder. The PPy solution
100ml purchased from Aldrich and the sonicated
CNF dispersion (with 1.67g CNF) were blended in a
beaker together and then heated on a hotplate at the
temperature of 50OC. After water evaporation in the
aqueous solution, flexible (not fragile) composite
films as shown in Fig. 3 were left at the bottom.
However, composite materials were broken into
pieces if heating temperature was over 60 OC.
Figs. 4 and 5 show the surface and the cross-section
of a composite film fabricated by FWP, respectively.
Fig. 6. SEM image of the surface of a composite film fabri-
As shown in the figures, the CNFs what look like cated by an SH process.
long threads have a diameter of more than 100 nm.
Compared with Figs. 1 and 2, the diameter of CNF is
3. Conductivity
much larger than that of SWNT, nearly 20 times. The
white agglomerates like particles are Ppy matrix. It is The voltage was measured with a four-probe or
clearly shown that in the CNF/Ppy composite, CNFs Kelvin probe method when a current was applied [10].
play the roll of fibers and Ppy acts like a matrix. Figs. Each sample may have a different voltage reading
6 and 7 show the surface and the cross-section of a due to different electric resistance in materials. This is
composite film fabricated by the improved SH proc- the most common way to obtain conductivity.
ess, respectively. The SEM micro-structures of the As shown in Fig. 8, there are four probes arranged
materials were similar to those by FWP shown in Figs. in line and the distances between the adjacent probes
4 and 5. are equal to each other. A current source is used for
78 C. Kim and S. Zhang / Journal of Mechanical Science and Technology 23 (2009) 75~80

Table 1. Conductivity of a pure CNT film

h I ∆V C
Thickness Current Voltage Conductivity
(cm) (A) (V) (S/cm)
0.4 0.34 19.98
0.6 0.50 20.38
0.013
0.8 0.68 19.98
Average 20.11

Fig. 7. SEM image of the cross section of a composite film Fig. 8 shows the experimental schematic of the
fabricated by an SH process. measuring method. Fig. 9 shows the setup for voltage
measurement in which the DC power supply ADPS-
503D supplies the current and is connected to outer
two probes and Agilent 34401A is used to measure
the voltage between the inner two probes. The con-
ductivity is calculated by following two relationships:

∆V
R = 4.53 h ( ) (1)
I
1
C = (2)
Fig. 8. Four-probe resistivity method. R

where R is the volume resistivity (Ω-cm), ∆V is the

measured voltage potential (volts), I is the source
current (amperes), and h is the sample thickness (cm).
It is noticed from the relationship that conductivity is
inversely proportional to the thickness of materials.
As the thickness increases, the conductivity decreases.
The conductivities of CNF/Ppy (1:3) composite
films and a pure SWNT film were calculated, based
on Eqs. (1) and (2), and are listed in Tables 1, 2, and 3.
The thicknesses of samples in the tables are different
because of difficulties in obtaining the desired thick-
ness in a controlled way. The film materials were
retrieved from materials left at the bottom of a beaker
after filtering or heating an aqueous mixed solution of
Fig. 9. Experimental setup for conductivity measurement. CNF and Ppy therefore, the thickness is uncontrolla-
ble.
supplying an electric current through the outer two The conductivity of the present pure SWNT film
probes and a voltmeter measures the voltage between with a thickness of 0.013 cm was 20.11 S/cm (see
the inner two probes. When differently specified cur- Table 1) and that of the similar SWNT samples with
rents are supplied to the same sample material several 0.0049 cm thick [6] was 11.70 S/cm. When compared
times, the voltages are measured. Then, the conduc- simply, our number itself is about as high as double
tivities may be obtained from Eqs. (1) and (2). This the other even if the thickness is different. The con-
type of measurement usually results in errors due to ductivity is inversely proportional to the thickness of
the probe resistance, the spreading resistance under materials, as shown in Eq. (1). Therefore, the thick-
each probe, and the contact resistance between each nesses can be adjusted equivalently to each other. If
metal probe and the composite film surface. the thickness, 0.0049 cm, was increased to 0.013
 C. Kim and S. Zhang / Journal of Mechanical Science and Technology 23 (2009) 75~80 79

Table 2. Conductivity of a CNF/Ppy composite film fabri- esses on a laboratory level for possible applications in
cated by an FWP process small sensors and actuators. Especially, the fabrica-
tion of CNF/Ppy composite films has not been re-
h I ∆V C
Thickness Current Voltage Conductivity ported in any other papers before. The conductivities
(cm) (A) (V) (S/cm) of the film samples were investigated experimentally.
0.4 0.023 61.92 The pure SWNT film which is 0.013 cm thick
0.6 0.033 64.74
showed a conductivity of 20.11 S/cm, which is as
0.062 high as double the value in Ref. [6]. The CNF/Ppy
0.8 0.045 63.30
composite film fabricated by the FWP process was
Average 63.32 0.062 cm thick and the conductivity was 63.32 S/cm.
The flexible CNF/Ppy (0.085 cm thick) fabricated by
Table 3. Conductivity of a CNF/Ppy composite film fabricated the SH process showed similar converted conductiv-
by an SH process
ity of 55.6 S/cm after the thickness conversion (40.57
h I ∆V C S/cm measured). The newly invented aqueous solu-
Thickness Current Voltage Conductivity tion-based fabrication process (SH) newly invented
(cm) (A) (V) (S/cm) produced more flexible CNF/Ppy composite films
0.4 0.026 39.95 than FWP. The newly developed CNF/Ppy materials
0.6 0.038 41.01 showed better conductivity than SWNT/PANi [6].
0.085 The electrical conductivity is one of the key factors
0.8 0.051 40.74
required for sensors or actuators. The higher conduc-
Average 40.57 tivity of CNF/Ppy has a good potential for a small
actuator material.
cm, the conductivity would be converted to (0.0049/
0.013) 11.70 = 4.41 S/cm. In this way, ours becomes
higher 4.5 times.
Acknowledgment
The conductivity of SWNT/PANi was 53 S/cm This work was supported by Korea Research
(0.0169 cm thick) [6] and that of our CNF/Ppy by the Foundation Grant (KRF-D00092 (KOSEF R05-2003-
FWP process (see Table 2) is 63.3 S/cm (0.0620 cm 000-12350-0) ).
thick), which is even higher than that of SWNT films.
If Ref. [6] was converted to the thickness of 0.0620
References
cm, the equivalent conductivity would be down to
14.44 S/cm. The conductivity of CNF/Ppy by the SH [1] L. Bay, K. West, P. Sommer-Larsen, P. Skaarup
process (see Table 3) is 40.57 S/cm (0.085 cm thick). and M. Benslimane, A conducting polymer artificial
If it is converted equivalently to 0.0620 cm thick, the muscle with 12% linear strain, Adv. Mater. 15
conductivity becomes 55.6 S/cm which is a little (2003) 310-313.
lower than 63.3 S/cm (by FWP). A little drop in con- [2] A. S. Hutchison, T. W. Lewis, S. E. Moulton, G. M.
ductivity is because TritonX 100 which hinders con- Spinks and G. G. Wallace, Development of
duction cannot be removed completely by heating. polypyrrole based electromechanical actuators,
The CNF/Ppy composite films were never fabricated Synth. Met. 113 (2000) 121-127.
before and show a better conductivity than [3] I. W. Hunter and S. Lafontaine, A comparison of
SWNT/PANi. The improved conductivity seems due muscle with artificial actuators, Proc. of Technical
to the enhanced connectivity of long carbon nanofi- Digest IEEE Solid-State Sensor Actuator Workshop,
bers. This simple improvement of CNF/Ppy may lead Hilton Head, South Carolina, USA. (1992) 178-185.
to better conducting polymer actuators and expanded [4] S. Iijima, Helical microtubules of graphitic carbon,
applications. Nature, 354 (1991) 56-58.
[5] P. Burgmayer and R. W. Murray, An ion gate
membrane: electrochemical control of ion perme-
4. Conclusions
ability through a membrane with an embedded elec-
Pure SWNT and CNF/Ppy composite (1:3) films trode, J. American Chemical Society, 104 (1982)
were fabricated by powder- and solution-based proc- 6139-6140.
80 C. Kim and S. Zhang / Journal of Mechanical Science and Technology 23 (2009) 75~80

[6] M. Tahhan, V. Truong, G. M. Spinks and G. G. probe impedance spectroscopy, Solid State Ionics,
Wallace, Carbon nanotube and polyaniline compos- 119 (1) (1999) 115-125.
ite actuators, Smart Materials and Structures, 12
(2003) 626-632.
[7] C. Kim and X. Y. Liu, Electromechanical behaviors Cheol Kim received a B.S.
and application of carbon nanotube composite ac- degree in Mechanical Engi-
tuator consisting of bundles and mats, J. The Ko- neering from Yonsei Univer-
rean Society for Composite Materials, 18 (2005) sity in 1985 after his military
34-39. service. He then received his
[8] B. Kim, L. Lee and I. Yu, Electrical properties of M.S. from Georgia Tech in
1989 and Ph.D. from U. of
single-wall carbon nanotube and epoxy composites,
Illinois at Urbana-Champaign
J. Appl. Phys. 94 (2003) 6724-6728. in 1994. Dr. Kim is currently a
[9] J. Xua, J. P. Donohoeb and C. U. Pittman Jr., Prepa- Professor at the School of Mechanical Engineering
ration, electrical and mechanical properties of vapor at Kyungpook National University (KNU) in Daegu,
grown carbon fiber (VGCF)/vinyl ester composites, Korea. He is currently serving as an associate editor
Composites: Part A, 35 (2004) 693-701. of the Transactions of the KSME A.
[10] J. R. Dygas, G. Fafil and M. W. Breiter, Study of Shuai Zhang is a Ph.D. candidate. He received his B.S. in
grain boundary polarization by two-probe and four- 2003 from HIT in China and M.S. in 2006 from KNU.