A New Method of Carbon‐Nanotube Patterning Using Reduction Potentials

Full text

Page 1

A New Method of Carbon-Nanotube Patterning Using
Reduction Potentials
By Jong Hak Lee, Jun Ho Shin, Yu Hee Kim, Sung Min Park, P. S. Alegaonkar,
and Ji-Beom Yoo*
Carbon nanotubes (CNTs) have attracted considerable interest as
high-stiffness materials and for use in nanodevices, on account of
their extraordinary mechanical, chemical, and electrical proper-
ties.[1]There have been several examples of CNTs being used to
realize devices and composites.[2,3]However, several hurdles
remain to be overcome before CNT-based technology can be
deployed on a commercial scale. Among these hurdles are
locating and patterning technologies. So far, films of CNT
networks have been patterned in the micrometer regime by a
variety of techniques, including using a poly(methyl methacry-
late) (PMMA) or poly(dimethylsiloxane) (PDMS) stamp, CO2
snow-jet etching, and O2-plasma etching.[4]However, these
methods have limitations in commercial applications, such as
lack of scalability, low resolution, and low reliability. This paper
reportsthatnoble metals,such asAu,Pt,and Ag,canpromotethe
oxidation of CNTs at a relatively low temperature (3508C),
because of the reduction potential of the CNTs (in this study,
oxidation means decomposition to CO2). Based on these
phenomena, a nanometer-sized, patterned, random network of
CNTs is fabricated. This study also examines the difference in the
reduction potentials of single-walled and multiwalled CNTs
(SWCNTs and MWCNTs, respectively).
Scheme 1 summarizes the experimental procedure used to
determine the reduction potential of the CNTs and to obtain the
novel nanometer-sized patterns of SWCNTs. In order to make a
CNT–metal junction, very thin metal films were deposited on a
CNT film using electron-beam evaporation. The samples were
then annealed in a box furnace at 3508C in ambient air. In a
reactive environment, a material system can be considered a
‘‘galvanic cell’’ if there is electrical contact between two materials
with different reduction potentials, and they are in the same
electrolyte. Under these conditions, the corrosion rate of the
material with the lower reduction potential can be faster than that
of the material with the higher reduction potential.[5]The novel
method for patterning CNT films was developed by applying this
phenomenon. In order to confine the CNTs to a selected area,
lines and numbers were drawn with Ag (15nm thick) using
photolithography, before depositing the CNT film on the
substrate. Figure 1a and b shows images of the patterned CNT
films after annealing. The CNTs are confined to the areas without
Ag. Only the CNT patterns remained after an aqua-regia
treatment. The widths of the lines patterned by the CNTs were
560 and 750nm. Photolithography and metal deposition are
well-developed techniques, with which CNTpatterns several tens
of nanometers in size can be obtained.
Recently, several applications based on the interaction between
CNTs and metal contacts were reported,[6]but there has been no
concise explanation for the phenomenon. A better understanding
of the interaction between CNTs and metals is needed, since
metals are used as electrodes, substrates, and/or as the main
components of CNT-based devices and composites. In particular,
in nanoscale materials, a small degree of metamorphosis can
cause serious problems, as these materials are quite small,
containing very few atoms.Therefore, for practical applications of
CNTs, it is very important to obtain information on their basic
parameters, such as their reduction potential. The reduction
potentials were estimated by producing galvanic coupling
between various metals and CNTs.
Figure 2 shows scanning electron microscopy (SEM) images of
metal-coated MWCNTs after annealing at 3508C for 4h. The
density of the CNTs shown in Figure 2a was almost the same as
that of the CNTs shown in Figure 2d. All CNTs were totally
eliminated in Figure 2b; and only a few remained in Figure 2c.
According to other reports, MWCNTs are thermally stable at
3508C in ambient air.[7]It is well-known that CNTs have excellent
thermal stability.[8]In some reports, in situ Raman spectroscopy
COMMUNICATION
www.advmat.de
Scheme 1. A flow diagram of the method used to estimate the CNT
reduction potentials and to pattern the CNT films. PR: photoresist.
[*] Prof. J.-B. Yoo, P. S. Alegaonkar, J. H. Shin, Y. H. Kim
School of Advanced Materials
Science & Engineering (BK21)
Sungkyunkwan University
Suwon 440-746 (Republic of Korea)
E-mail: jbyoo@skku.edu
J. H. Lee, S. M. Park
SKKU Advanced Institute of Nanotechnology (SAINT)
Sungkyunkwan University
Suwon 440-746 (Republic of Korea)
DOI: 10.1002/adma.200802507
Adv. Mater. 2009, 21, 1257–1260
? 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1257

Page 2

COMMUNICATION
www.advmat.de
analysis of nanotube oxidation showed a decrease in the intensity
of the D band beginning at approximately 3708C,[7]indicating
oxidation of the amorphous carbon deposited on the CNTs. The
temperature for the heating-induced purification of CNTs has
been reported to range from 370 to 4008C. The G band has been
reported to increase or remain constant until 5808C.[7]This
suggests that most CNTs are thermally stable, while a few that
present some defects are eliminated below 4008C. There is a
distinct difference between MWCNTs (Fig. 2a) and MWCNTs
coated with various metals, such as Au, Pt, Ag, Cu, and W (the
results in the case of Pt, Ag, Cu can be seen in the Supporting
Information). In the case of the metal-coated MWCNTs, there are
many contacts between the CNTs and the metals, which have
different reduction potentials from the tubes. Moreover, air can
act as an electrolyte by supplying moisture.[9]
Therefore, this structure can be called a
‘‘galvanic cell’’. Although its appearance is
similar to that of the CNT-only films, some
reactions, such as reduction or oxidation, take
place in corrosive environments, such as
high-temperature air. The oxidation rate of a
material with a low reduction potential will be
fast, beingdrivenby the difference inreduction
potential between the two materials. If the
CNTs have a higher reduction potential than
the metal, the state of the CNT will remain
unchanged, while the metal will be oxidized. In
the opposite case, the CNTwill be oxidized and
ultimately disappear. Therefore, the reduction
potential of the MWCNTs can be estimated
from these results. In the case of Ni, all
nanotubes remained after annealing. This
suggests that the Ni thin film was oxidized
and the CNTs remained in their as-deposited
state. This result was also observed in the case
of Ti (the SEM images can be seen in the
Supporting Information). Therefore, as shown
in Figure 4, the reduction potential of the
MWCNTs at 3508C is located somewhere
between that of W and Ni.
Different results
SWCNTs. As shown in Figure 3, the SWCNTs
with no metal contact and those with In
remainedintact.OnlyafewSWCNTsremained
in the case of Ni, and they disappeared
completely when in contact with metals such
as Cu, W, and Ni after annealing at 3508C for
4h (SEM images can be seen in the Supporting
Information). In the case of In and Ti, all the
SWCNTs remained after annealing. Therefore,
as shown in Figure 4, the reduction potential of
the SWCNTs at 3508C is located somewhere
betweenthoseofNiandIn.Thismeansthatthe
reduction potential of SWCNTs is lower than
that of MWCNTs. This higher reactivity can
possibly be attributed to the increased strain on
the aromatic structure of the nonplanar
benzene subunits that make up the nanotubes.
MWCNTs are generally more stable on account
of the steric hindrance of the outer walls, as well as the lower
strain that exists on these outer walls because of their larger
diameters. This steric hindrance affects the reduction potential,
indicating why the burning temperature of SWCNTs[10]is lower
than that of MWCNTs,[11]despite the fact that the mechanical
strength of the former is higher than that of the latter,[12]and that
SWCNTs have fewer defects than MWCNTs. Based on these
experiments, it can be concluded that the oxidation of SWCNTs
and MWCNTs is dependent on their reduction potentials. The
reactions between the metal and the CNTs can be written as
follows
wereobservedfor
aM þ bO2! kMxOy: DG1
kMxOyþ bCCNT! aM þ bCO2: DG2
Figure 1. a,b) Patterned CNT films obtained after annealing with Ag. c,d) The films after acid
treatment to remove the Ag (all scale bars are 1mm).
Figure 2. Field-emission SEM (FESEM) images of: a) CNT, b) Au(3nm)–CNT, c) W(1nm)–CNT,
and d) Ni(3nm)–CNT films.
1258
? 2009 WILEY-VCH Verlag GmbH & Co. KGaA, WeinheimAdv. Mater. 2009, 21, 1257–1260

Page 3

COMMUNICATION
www.advmat.de
resulting in the overall chemical reaction
aM þ bCCNTþ bO2! aM þ bCO2
(1)
These results can be theoretically explained using the Nernst
equation. The Nernst equation can determine the reduction process
and trends in a chemical reaction.[13]Two solutions can be obtained
using this equation. One isDG ¼ DGCNT!CO2? nFDEo, where
DGCNT!CO2is the Gibbs free energy of the CNTpyrolysis, n is the
number of electrons in the reaction, F is Faraday’s constant, and
DEois the difference in the standard reduction potentials
between the metal and CNT (DEo¼ Eo
the noble-metal–CNT system, DEohas a positive value and the
free energy of the reaction (DG) is lower than that of the isolated
CNTsystem. This suggests that a metal contact, which has a high
reduction potential, can promote the oxidation of CNTs. The
other equation is, dDG ¼ ?DSCNT!CO2dT ? nFdDEo, where
DSCNT!CO2is the entropy of the reaction and T is the absolute
temperature. At equilibrium, dDG approaches zero, and the
equilibrium temperature can be affected by the difference in
reduction potential. Consequently, pyrolysis or oxidation can be
induced more easily with increasing difference in reduction
potential and temperature. When two different materials are
coupled together, the one with lower reduction potential has
excess electron activity, where the electrons are transferred to the
material with the more-positive reduction potential.
Figure 5 shows SEM images of elaborately patterned CNT
films. In order to prevent aggregation and migration of Ag
nanoparticles during the annealing process, a 5nm Ti layer was
deposited before the Ag film. This result indicates that CNT
patterns several tens of nanometers in size can be obtained as
long as this size of metal nanopatterens can be obtained.
In summary, a nanometer-sized pattern of a CNT film was
successfully fabricated by employing the difference in reduction
potential between the CNTs and different metals. Moreover, this
study examined the reduction potentials of CNTs using a very
simple method. Noble metals (such as Au, Pt, and Ag) can induce
the oxidation of CNTs at a relatively low temperature in air. The
reduction potential of MWCNTs is located between those of W
andNi,whilethatofSWCNTsarelocatedbetween thoseofNiand
In.Overall,thesefindingscanbeusedasguidelinesfordesigning
CNT-based devices.
M? Eo
CNT). In the case of
Experimental
Sample Preparation and Annealing Process: MWCNTs and SWCNTs
were obtained from ILJIN Nanotech Co., Ltd. (Korea). The CNTs were
Figure 3. Images of various CNT films annealed for 4h: a) CNT,
b) Ni(1nm)–CNT, c) In(2nm)–CNT.
Figure 4. Table of the standard reduction potentials (SRPs) of various
metals and the inferred reduction potentials of MWCNTs and SWCNTs at
3508C.
Figure 5. The elaborately patterned CNT films with Ti/Ag (all scale bars
indicate 100nm).
Adv. Mater. 2009, 21, 1257–1260
? 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1259

Page 4

COMMUNICATION
www.advmat.de
dispersed in a 1% sodium dodecylsulfate solution dissolved in water and
sonicated using a tip-sonication instrument. In order to obtain a
well-dispersed CNT solution, the sonicated solution was centrifuged at
12000g (g: 10ms?2). A uniform CNT film was obtained by vacuum
filtration; and transferred [14] to a silicon wafer with thermally grown SiO2
(500nm). All the metals were deposited onto the CNT films to a thickness
of 3nm to examine the elimination of CNT. With Au and Ag, nanoparticle
formation began rapidly. Therefore, the thickness of the Au and Ag films
was increased to 15nm in order to completely remove the CNTs. After
preparation, the samples were annealed at 3508C for 4h in air.
CNT-Film Patterning: Photoresists (PRs) were patterned using a Nikon
stepper. Ag films 15nm thick were deposited onto the nanometer-sized
PR-patterned substrate. After a lift-off process to remove the PR, a uniform
CNT film was transferred onto the substrate and annealed at 3508C for 4h
in air. The Ag nanoparticles were removed by dipping the samples in aqua
regia for 30min; followed by rinsing with deionized water.
Fine CNT-Film Patterning: The photolithography process and annealing
conditions were the same as those used in CNT-film patterning. The only
difference was that a 5nm Ti film was deposited as a binding layer before
Ag-film deposition.
Acknowledgements
This research was funded by the KOSEF through the CNNC (Center for
Nanotubes and Nano structured Composites) at Sungkyunkwan Uni-
versity;andtheBK21ProjectoftheSchoolofAdvancedMaterialsScience&
Engineering. Supporting Information is available online from Wiley
InterScience or from the author.
Received: August 27, 2008
Revised: October 7, 2008
Published online: January 14, 2009
[1] a) R. H. Baughman, A. A. Zakhidov, W. A. de Heer, Science 2002, 297, 787.
b) M. Motta, A. Moisala, I. A. Kinloch, A. H. Windle, Adv. Mater. 2007, 19,
3721. c) M. Hughes, M. S. P. Shaffer, A. C. Renouf, C. Singh, G. Z. Chen,
D. J. Fray, A. H. Windle, Adv. Mater. 2002, 14, 382.
[2] S. N. Kim, J. F. Rusling, F. Papadimitrakopoulos, Adv. Mater. 2007, 20, 363.
[3] A. A. Gusev, O. Guseva, Adv. Mater. 2007, 19, 2672.
[4] a)Y.X.Zhou,L.B.Hu,G.Gru ¨ner,Appl.Phys.Lett.2006,88,123109.b)E.S.
Snow, J. P. Novak, P. M. Campbell, D. Park, Appl. Phys. Lett. 2003, 82, 2145.
c) T. Ozel, A. Gaur, J. A. Rogers, M. Shim, Nano Lett. 2005, 5, 905.
[5] D. A. Jones, Principles and Prevention of Corrosion, 2nd Edn., Prentice-Hall,
Englewood Chiffs, NJ 1996, p. 168.
[6] a) W. Xia, V. Hagen, S. Kundu, Y. Wang, C. Somsen, G. Eggeler, G. Sun, G.
Grundmeier,M.Stratmann, M.Muhler,Adv.Mater.2007,19,3648. b)L.Ci,
S. M. Manikoth, X. Li, R. Vajtai, P. M. Ajayan, Adv. Mater. 2007, 19, 3300.
[7] S. Osswald, E. Flahaut, Y. Gogotsi, Chem. Mater. 2006, 18, 1525.
[8] S. T. Purcell, P. Vincent, C. Journet, V. T. Binh, Phys. Rev. Lett. 2002, 88,
105502.
[9] http://www.moldmakingtechnology.com/articles/040302.html (accessed
November 2008).
[10] A. G. Rinzler, J. Liu, H. Dai, P. Nikolaev, C. B. Huffman, F. J. Rodrı ´guez-
Macı ´as, P. J. Boul, A. H. Lu, D. Heymann, D. T. Colbert, R. S. Lee, J. E.
Fischer, A. M. Rao, P. C. Eklund, R. E. Smalley, Appl. Phys. A. 1998, 67, 29.
[11] D. Bom, R. Andrews, D. Jacques, J. Anthony, B. Chen, M. S. Meier, J. P.
Selegue, Nano Lett. 2002, 2, 615.
[12] a)T.W.Ebbesen, H.J.Lezec,H.Hiura,J.W.Bennett,H.F.Ghaemi,T.Thio,
Nature 1996, 382, 54. b) J. E. Fischer, H. Dai, A. Thess, R. Lee, N. M.
Hanjani, D. L. Dehaas, R. E. Smalley, Phys. Rev. B. 1997, 55, R4921.
[13] D. A. Jones, Principles and Prevention of Corrosion, 2nd Edn. Prentice-Hall,
Englewood Chiffs, NJ 1996, p. 45.
[14] Z. Wu, Z. Chen, X. Du, J. M. Logan, J. Sippel, M. Nikolou, K. Kamaras, J. R.
Reynolds, D.B.Tanner, A.F. Hebard, A.G. Rinzler,Science 2004,305,1273.
1260
? 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2009, 21, 1257–1260