Ultra-thin SWNTs Films with Tunable, Anisotropic Transport Properties
ABSTRACT Directional transport properties at the nanoscale remain a challenge
primarily due to issues associated with control over the underlying anisotropy
and scalability to macroscopic scales. In this letter, we develop a facile
approach based on template-guided fluidic assembly of high mobility building
blocks - single walled carbon nanotubes (SWNTs) - to fabricate ultra-thin and
anisotropic SWNT films. A major advancement is the complete control over the
anisotropy in the assembled nanostructure, realized by three-dimensional
engineering of dip-coated SWNT thin films into alternating hydrophilic and
hydrophobic micro-line patterns with prescribed intra/inter-line widths and
line thicknesses. Variations in the contact line profile results in an
evaporation-controlled assembly mechanism that leads to the formation of an
alternating, and more importantly, contiguous SWNT network. Evidently, the
nanoscopic thickness modulations are direct reflections of the substrate
geometry and chemistry. The nanostructured film exhibits significant anisotropy
in their electrical and thermal transport properties as well as optical
transparency, as revealed by characterization studies. The direct interplay
between the anisotropy and the 3D micro-line patterns of the substrate combined
with the wafer-level scalability of the fluidic assembly allows us to tune the
transport properties for a host of nanoelectronic applications.
DOI: 10.1002/adfm.((please insert DOI)
Ultra-thin SWNTs Films with Tunable, Anisotropic Transport Properties
By Bo Li, Hyun Young Jung, Hailong Wang, Young Lae Kim, Taehoon Kim, Myung Gwan
Hahm, Ahmed Busnaina, Moneesh Upmanyu*, and Yung Joon Jung *
Hahm, Prof. A. Busnaina
Department of Mechanical and Industrial Engineering,
Northeastern University, Boston, MA 02115(USA)
Prof. M. Upmanyu, Corresponding-Author, Dr. H. Wang
Group for Simulation and Theory of Atomic-scale Material Phenomena (stAMP),
Northeastern University, Boston, MA 02115 (USA)
Y. L. Kim
Department of Electrical and Computer Engineering,
Northeastern University, Boston, MA 02115 (USA)
Prof. Y. J. Jung. Corresponding-Author, B. Li, Dr. H. Y. Jung, Dr. T. Kim, Dr. M. G.
Keywords: Single walled carbon nanotube, thin film, anisotropy, transfer.
Directional transport properties at the nanoscale remain a challenge primarily due to issues
associated with control over the underlying anisotropy and scalability to macroscopic scales.
In this letter, we develop a facile approach based on template-guided fluidic assembly of high
mobility building blocks - single walled carbon nanotubes (SWNTs) - to fabricate ultra-thin
and anisotropic SWNT films. A major advancement is the complete control over the
anisotropy in the assembled nanostructure, realized by three-dimensional engineering of dip-
coated SWNT thin films into alternating hydrophilic and hydrophobic micro-line patterns
with prescribed intra/inter-line widths and line thicknesses. Variations in the contact line
profile results in an evaporation-controlled assembly mechanism that leads to the formation of
an alternating, and more importantly, contiguous SWNT network. Evidently, the nanoscopic
thickness modulations are direct reflections of the substrate geometry and chemistry. The
nanostructured film exhibits significant anisotropy in their electrical and thermal transport
properties as well as optical transparency, as revealed by characterization studies. The direct
interplay between the anisotropy and the 3D micro-line patterns of the substrate combined
with the wafer-level scalability of the fluidic assembly allows us to tune the transport
properties for a host of nanoelectronic applications.
Nanoelectronic devices that rely on the superior electronic transport in individual single-
walled carbon nanotubes (SWNTs) are typically beleaguered by reliability and scalability
issues, the primary reason for the ongoing paradigm shift towards SWNTs thin films as active
elements. Recent success in wafer-scale synthesis of these thin films, either by SWNT
growth or by assembly, [1-6] has emerged as the driving force for several applications including
next-generation flexible and transparent thin-film electrodes, high mobility transistors, and a
suite of robust sensors. Expectedly, engineering the carbon nanotube-scale nanostructure as
well as the overall film architecture is the key in tailoring electronic transport through the
films. Unfortunately, it remains a limitation as the existing fabrication techniques typically
yield one- or two-dimensional (1D/2D) isotropic SWNT network film structures.[3, 7, 8]
Compounding matters is the fact that the constituent SWNTs are usually a mixture of
semiconducting and metallic nanotubes that cannot be easily separated.[9, 10] Interestingly,
though, the interplay between the intrinsic electrical heterogeneity and the nanoscale
structure/topology of the assembled SWNT networks can result in drastically different
transport characteristics[11-13] and therefore is a promising route for engineering highly
functional and integrated film systems with prescribed electrical, thermal, and optical
Tailoring the nanotube-scale network in turn requires complete control over film density,
morphology and architecture. In this letter we present a facile route for the synthesis of ultra-
thin SWNT film architectures which permit engineering of the (in-plane) directionality in the
electrical and thermal transport, i.e. the anisotropy. The enabling component unique to our
approach is the fluidic assembly of dispersed SWNTs onto suitably micro-patterned and
chemically heterogeneous three-dimensional (3D) substrates. The resultant SWNT films have
several features that allow control over the extent of the anisotropy. First, it is immediately
clear that the assembled SWNTs inherit the topography of 3D patterned SiO2 and photoresist
(PR) micro-line structures, forming quasi-wavy, ultra-thin, and transparent films. Second, the
chemical heterogeneity of the substrate due to the alternating hydrophobic (PR) and
hydrophilic (SiO2) micro-patterns allow us to control the local thickness of SWNT films,
thereby allowing control over the electrical and thermal properties. Lastly, these 3D SWNT
films can be easily transferred onto polymer matrices, a critical step in synthesis of flexible
and transparent SWNT network-polymer composite films for flexible, transparent and
anisotropic components in electrical devices.
2. Results and Discussion
2.1. Morphology of 3D SWNT Films
Figure 1 shows representative scanning electron microscopy (SEM) and optical images
as well as schematics of our assembled SWNT films. To this end, the underlying SiO2 or
quartz substrates are first subjected to plasma treatment to improve their affinity to SWNT-
deionized (DI) water solution.[14-16] Then, chemical heterogeneity is introduced through the
fabrication of hydrophobic PR micro-line structures (6 µm in width with a spacing of 9 µm)
with controlled PR thicknesses (HPR=323, 360, and 480 nm) using photolithography
processes, resulting in a 3D hydrophobic-hydrophilic surface architecture. Finally, the fluidic
assembly of SWNTs is realized by vertically dip-coating the substrate into 0.23 wt% SWNT-
DI water solution and gradually lifting at a controlled pulling velocity of V=0.1 mm/min.
SEM images (Fig. 1a and 1b) clearly show that the assembled SWNT network forms a
continuous thin film which inherits the 3D morphology of hydrophobic and hydrophilic
micro-lines on the substrate, i.e. the film morphology is quasi-wavy as illustrated
schematically in Fig. 1c. This assembled SWNTs film is transparent as well, evident from the
optical microscope image shown in Fig. 1d. We were also able to assemble these ultra-thin 3D
SWNT films on PR/quartz substrates. Figure 1e shows the optical image of the sample after
removing the PR, further illustrating the transparent nature of our assembled SWNT films.
A notable feature, revealed by atomic force microscopy (AFM) characterization in Figure
2, is that the local thickness of these assembled SWNT films differs dramatically between the
PR and SiO2 micro-lines, resulting in alternating thin and thick SWNT film strips. The inset
schematic in Fig. 2 shows the cross sectional view of the 3D SWNT films on the patterned
substrate. A key aspect that allows us to engineer the SWNT film morphology is that the
thickness of SWNT films on photoresist (HSWNT/PR) is strongly influenced by HPR. More
specifically, lower HPR (323 nm) results in a higher HSWNT/PR (40 nm). Conversely for higher
HPR (480 nm), the film thickness decreases to a few nanometers, HSWNT/PR=5.5 nm, while the
thickness of SWNTs film on SiO2 (HSWNT/SiO2) is relative inert to HPR.
2.2. Characterization of Anisotropic Transport Properties
Controlled modulations in the local thickness of the assembled SWNTs films can be
exploited to tailor the directionality of the electrical transport through SWNTs film. To this
end, the 100×100 µm2 square gold contact pads were patterned on the SWNTs film with
100µm spacing, as shown in Figure 3a (inset). The electrical resistances were measured in the
vertical V-V’ (Rvertical) and in the horizontal H-H’ direction (Rhorizontal) by two-point probe
method. The dependence of the electrical resistance (Rvertical, red line and Rhorizontal, green line)
as well as the electrical anisotropy (HSWNT/SiO2/HSWNT/PR, blue line) on thickness ratio of
SWNT film (HSWNT/SiO2/HSWNT/PR) are shown in Fig. 3. For all samples, Rhorizontal is higher
than Rvertical, strongly indicative of the in-plane electrical anisotropy in these films.
Furthermore, the electrical resistance in both directions increased with the increase in the ratio
HSWNT/SiO2/HSWNT/PR, or increase in the PR thickness. However, Rhorizontal increases more
rapidly such that the electrical anisotropy, expressed as Rhorizontal/Rvertical, also increases with
HSWNT/SiO2/HSWNT/PR. The results clearly demonstrate that the electrical anisotropy can be
controlled by changing the thickness of photoresist HPR. Simple percolation based simulations
of randomly assembled SWNT networks corroborate this effect, as shown in Figure S1. The
inverse number of possible conductive paths (1/paths), related to the resistance of the film,
increase with the increase of inverse number of stacking layer (1/layers), or the thickness of
film. Decreasing film thickness leads to a non-linear decrease of possible conductive path,
and thus a non-linear increase of resistance. The effect is a reflection of the reduction in
effective conduction pathway of metallic SWNTs in assembled SWNTs films. In addition,
owing to the electrical heterogeneity of the SWCNT mix, the metallic SWCNTs get
effectively shielded such that the nature of electrical conduction becomes increasingly
semiconducting at small thicknesses. It follows then that in the 3D anisotropic system with
variations in film thickness, the electrons pass the thin and thick SWNT strips in parallel
along the vertical direction and the thicker films dominate the electron transportation.
However, in the horizontal direction, the electrons pass alternating thin and thick strips in
series such that thinner SWNT strips determine the overall electrical transport property of
these SWNTs films. In each case, the vertical direction (V-V’) has enhanced conduction
pathways relative to the horizontal direction.
In the vertical direction, the electrical resistance is mainly determined by the thick SWNT
strips on SiO2 alongwith the thin SWNT strips on PR that provide extra conductive paths in
parallel. Thus, as the height ratio increases, the number of conducting paths reduce as the the
metallic SWCTs in the heterogeneous SWNT mixture are shielded such that the resistance
along the PR trenches is increasingly controlled by semiconducting SWNTs. The
combination of geometry and nature of electrical conduction drives the resistance of the PR
film ; as mentioned earlier, the increase is highly non-linear (see Supplementary Documents
and Ref). Consequently, the resistance of the SWCT film on the PR becomes significant
and drives the overall resistance. This is evident in Fig. 2 wherein the decrease in HSWNTs/PR
due to increasing PR height results in higher vertical resistance. Additional effects due to
variations in the SiO2 trench geometry and film height as well as non-trivial structures at the
PR-SiO2 interface cannot be ruled out. Their characterization, however, is beyond the scope of
We also investigated thermal transport properties of the assembled SWNTs film
having the thickness ratio (HSWNT/SiO2/HSWNT/PR) of 1.44. The measurement of thermal
conductivity of the anisotropic SWNTs film was performed by utilizing a self-heating 3ω
technique (see Figs. 3b and 3c). The 3ω signal correlates with thermal conductivity through
where L, R, and S are the distance between contacts, electrical resistance, and cross sectional
area of the sample, respectively. R´=(δR/δT) is the temperature gradient of the resistance at
the chosen temperature and k is the thermal conductivity.[17, 18] The 3ω method was utilized by
the four-point-probe third harmonic characterization to eliminate the contact resistance and to
avoid related spurious signals. Specifically, the resistance of the assembled SWNTs film and
its temperature dependency in the vertical and horizontal directions was measured in the
temperature range of 21 ~ 25 °C which lies close to the measurement temperature of the 3ω
signal. The measured resistance and the temperature coefficient are 422 Ω and 31 Ω/°C along
the vertical direction (V-V’), and 469 Ω and 25 Ω/°C along the horizontal direction (H-H’)
(Fig. 3b). Again, the lower temperature coefficient in horizontal direction results from the
reduction in effective conduction pathway of metallic SWNTs. The third-harmonic voltage
measured at the frequency of 1000 Hz is shown in Fig. 3c. The thermal conductivities
calculated with Eq.1 were 22 W/mK and 4 W/mK along the vertical and horizontal directions,
respectively. We assume that the decrease in the effective thermal conductivity across the
horizontal direction is due to the reduced conduction pathways relative to the vertical
direction, and additional phonon scattering at the interfaces between the alternating thin and
thick film strips.
2.3. Transfer of the 3D SWNT film
We also demonstrate the ability to transfer assembled SWNT films on flexible
polymeric substrates, a crucial capability for the synthesis of scalable yet tailored integrated
functional flexible systems. Figure 4a shows the schematic of developed transfer process:
first, a layer of poly(methyl methacrylate) (PMMA) is spin-coated on the SWNT films;
second, the underlying SiO2 layer is etched by HF solution; finally, the PR/SWNT/PMMA
film is peeled off from the original substrate and can be readily transferred onto any other
target substrate. Due to the unique 3D morphology, our assembled films can be transferred
together with arrays of PR micro-lines inserted into the trenches of 3D SWNT films. Figure
4b shows the SEM image of the transferred SWNT film, in which the black color regions are
exposed to SWNT film while the bright color regions are PR micro-lines. A well defined
boundary between SiO2 and PR micro-lines can be seen in Fig. 4c. Finally as confirmation,
Fig. 4d shows the optical image of the transferred PR/SWNT/ PMMA hybrid film.
2.4. Formation Mechanism of 3D SWNT films
To elucidate the formation mechanism of anisotropic 3D SWNT films, we have first
quantified the chemical heterogeneity of the substrate with respect to the SWCNT-DI solution
via micron-scale static contact angle measurements. Our results reveal (Fig. 6a) that the PR is
significantly hydrophobic (θ=60° for PR-solution) compared to the plasma treated SiO2
(θ<5°). In our experiments, the SWNTs deposit along the receding liquid-air contact line.
Then, the interplay between the chemistry and morphology of the substrate and the
microfluidics of the receding contact line becomes important,[20, 21] evident from the snapshot
of the assembly during the dip-coat shown in Fig. 5b. The SEM image above the reservoir is
along the film normal while that below is a schematic illustration. The wavy contact line,
consisting of liquid bridges across the PR lines that are curved along the dip-coat direction, is
a result of a solution that wets into the PR, strongly suggestive of “cross-talk” between the
micro-lines, i.e. modification of the contact line dynamics due to chemical heterogeneity of
the abutting lines. While its origin follows from the simple fact that the width of the
hydrophobic PR pattern lines (~6 µm) is much smaller than the capillary length of the SWNT
solution (see Supplementary Documents for a discussion), the effect on the final film
thickness on both PR and SiO2 micro-lines requires an understanding of the microfluidics
during the dip-coat. To illustrate the salient effects, consider a cartesian coordinate system
with its xy plane along the substrate (Fig. 5c): x denotes the direction of plane withdrawal
(streamwise direction) and z is the direction normal to the plane. The large curvature
transverse to the pattern lines and the change in contact angle (~5° to 60°) at the chemically
heterogeneous interface between the patterns modify the excess pressure at the liquid-air
interface and therefore the shape of the meniscus h(x, y) associated with the entrained liquid.
The excess pressure is central to the extent of the cross-talk between the patterns and for small
slopes it can be approximated as δp=γ(hxx + hyy). Near the reservoir, the much larger hxx is
controlled by the width of the hydrophobic pattern lines W and scales as hxx~δ/W2 at the center
of PR pattern, where δ is a length scale along the streamwise direction. Evidently, large
capillary pressures are required to balance the excess pressure and drive the enhanced liquid
pick-up into the PR micro-lines forming a wetting layer well above the horizontal of the
reservoir (Fig. 5b and 5c).
The above analysis elucidates the basis for the formation of continuous SWNT thin films,
yet the fluidics that controls the local thickness of the film is slightly different in that, i) for
the set of deposition variables employed in this study the film thickness is evaporation-
controlled, and ii) the variation in the PR step thickness (Fig. 5c) introduces an additional
transverse curvature hzz due to the out-of plane variation in the thickness profile, h≡h(y, z). In
the absence of cross-talk, the film thickness following evaporation is H≅Φθld, where Φ is
strength of the solution and ld is the drying length, the length along the streamwise direction at
which the SWNT flux due to motion of the contact line is exactly balanced by the evaporation
flux. Following Berteloot, a one-dimensional flux balance yields ld=(J0/Vθ)2, where J0 is
evaporation constant for the solvent (approximately that of water) which depends on ambient
conditions such as vapor saturation concentration, diffusion constant, mass density, etc. Under
standard conditions, J0=10-9 m3/2s-1 for a millimeter sized water droplet. For the dip-coating
parameters in our experiments, this yields a drying length of ld ~ 4.7 µm and ld ~ 0.3 µm for
the SiO2 and PR surfaces, respectively; the corresponding film thicknesses on SiO2 and PR
are ~ 9.5 nm and ~ 0.8 nm (see Supplementary Documents). The estimates for film
thicknesses represent lower bounds as they correspond to infinite, homogeneous surfaces. The
larger thicknesses reported in Fig. 5c, especially on PR lines, are in part an indicator of the
cross-talk which leads to continuous thin film formation. The traverse flow decreases the
effective velocity of the receding contact line on the PR lines, thereby increasing the drying
length and the evaporation-controlled film thickness (∼1/V2). The change in the thickness of
the SiO2 lines is minimal as the contact line profile lags away from the reservoir such that the
transverse flow away from the line occurs at a region where viscous forces become important.
The observed trends in the film thickness are due to an out-of-plane curvature that increases
with PR step height and works against the in-plane transverse curvature, therefore opposing
the liquid pick-up from the SiO2 lines onto the PR surface (see Fig. 5c). In the evaporation-
controlled limit, this leads to a reduction in the mean local velocity of the entrained liquid
relative to the PR surface. As discussed before, we expect minimal changes in the thickness
deposited on SiO2 lines, consistent with our data. In summary, the thickness of the PR region,
HPR, emerges as a robust knob that offers complete control over the SWNTs film thickness,
and the form and extent of the heterogeneity in film morphology.
We have successfully demonstrated a scalable yet facile route for fabricating highly
engineered ultra-thin anisotropic SWNT films formed on 3D micro-patterned substrates using
a highly controlled template guided fluidic assembly technique. Our assembled ultra thin
SWNT film structures show continuous but alternating arrangement of SWNT line patterns
with differing thicknesses. The thickness ratio can be controlled to at least an order of
magnitude using our assembly technique. A simple scaling analysis explains the fundamental
assembly mechanism that allows us to control the nanostructural heterogeneity. The electrical
and thermal anisotropy is a direct result of the anisotropy in the nanostructure, and therefore
can be tuned by controlling the 3D structure of the substrate. Moreover, by using a polymer
casting transfer method, the films can be transferred onto the PMMA substrate to form
heterogeneous composite films with multiple polymeric materials. The anisotropic and
composite SWNT films with tunable electrical and thermal transport properties are expected
to have broad implications for the development of next-generation flexible and transparent
thin-film electrode, transistors and sensors, with minimized device-to-device variation and
Measuring the thickness of 3D SWNTs films. Atomic force microscopy (PSIA XE-150, Park
Systems Inc., USA) was employed to measure the thickness of 3D SWNTs films. First, the
profile of photoresist (PR) patterns on silicon dioxide (SiO2) substrate was determined and the
height of photoresist (HPR) was obtained prior to the fabrication of SWNTs films. After
depositing the 3D SWNTs film, the part of SWNTs film was carefully removed by a tungsten
probe to expose SiO2 substrate without any damage on the substrate. The exposed SiO2
surface was used as the reference plane. Then, a large area scanning (30 µm x 15 µm) was
performed to determine the thickness of SWNT film on SiO2 surface (HSWNT/SiO2) as well as
the height of the PR with coated SWNTs film (HPR+ SWNT/PR). Finally, the thickness of SWNT
film on PR (HSWNT/PR) was obtained by subtracting HPR from the HPR+SWNT/PR.
Thermal conductive measurement. Heat transport equation can be expressed in terms of the
third harmonic voltage signal induced by an AC current of the form V3ω =Io sinωt passing
through the sample at low frequencies. The AC current with frequency ω creates a
temperature fluctuation at 2ω, which further causes a third harmonic voltage signal. The 3ω
signals can be used for measuring thermal conductivity of anisotropic SWNT film. A lock-in
amplifier (Stanford Research System SR850) was used for obtaining 3ω signals by
amplifying the small voltage and removing the noise. An AC current source (Keithley 6221)
was used to provide a stable current supply. All the measurements including resistance,
temperature, and 3ω signals were done under high vacuum (P<10-5 Torr) in a Janis Research
ST-500 cryogenic probe station to reduce radial heat losses through gas convection. The
temperature coefficient of the resistance is also measured in order to obtain the thermal
conductivity based on Eq.1. The thermal conductivities calculated with Eq.1 were obtained to
be 22 W m-1 K-1 and 4 W m-1 K-1 for vertical and horizontal direction sample, respectively.
We would like to thank financial support from NSF CMMI Nanomanufacturing Program
(0927088), Center for High-Rate Nanomanufacturing in Northeastern University, and
Fundamental R&D Program for Core Technology of materials by Ministry of Knowledge
Economy, Republic of Korea. We are also grateful to Brewer Science Corp. for kindly
donating the relevant SWNT solutions. HW and MU are grateful for support from Structural
Metallics Program, ONR -N000141010866.
Received: ((will be filled in by the editorial staff))
Revised: ((will be filled in by the editorial staff))
Published online: ((will be filled in by the editorial staff))
 Q. Cao, J. A. Rogers, Adv. Mater. 2009, 21, 29.
 S. Rao, L. Huang, W. Setyawan, S. Hong, Nature 2003, 425, 36.
 Z. Wu, Z. Chen, X. Du, J. Logan, J. Sippel, M. Nikolou, K. Kamaras, J. Reynolds, D.
Tanner, A. Hebard, Science 2004, 305, 1273.
 M. Zhang, S. Fang, A. Zakhidov, S. Lee, A. Aliev, C. Williams, K. Atkinson, R.
Baughman, Science 2005, 309, 1215.
 S. J. Kang, C. Kocabas, T. Ozel, M. Shim, N. Pimparkar, M. A. Alam, S. V. Rotkin, J.
A. Rogers, Nat. Nanotechnol. 2007, 2, 230.
 M. LeMieux, M. Roberts, S. Barman, Y. Jin, J. Kim, Z. Bao, Science 2008, 321, 101.
 E. Artukovic, M. Kaempgen, D. Hecht, S. Roth, G. Grüner, Nano Lett. 2005, 5, 757.
 Q. Cao, H. S. Kim, N. Pimparkar, J. P. Kulkarni, C. J. Wang, M. Shim, K. Roy, M. A.
Alam, J. A. Rogers, Nature 2008, 454, 495.
 R. Krupke, F. Hennrich, H. Lohneysen, M. Kappes, Science 2003, 301, 344
 Z. Chen, X. Du, M. Du, C. Rancken, H. Cheng, A. Rinzler, Nano Lett. 2003, 3, 1245.
 C. Kocabas, N. Pimparkar, O. Yesilyurt, S. J. Kang, M. A. Alam, J. A. Rogers, Nano
Lett. 2007, 7, 1195.
 S. J. Kang, C. Kocabas, H. S. Kim, Q. Cao, M. A. Meitl, D. Y. Khang, J. A. Rogers,
Nano Lett. 2007, 7, 3343.
 S. Somu, H. Wang, Y. Kim, L. Jaberansari, M. Hahm, B. Li, T. Kim, X. Xiong, Y.
Jung, M. Upmanyu, ACS Nano 2010, 4, 4142.
 Y. L. Kim, B. Li, X. H. An, M. G. Hahm, L. Chen, M. Washington, P. M. Ajayan, S. K.
Nayak, A. Busnaina, S. Kar, Y. J. Jung, ACS Nano 2009, 3, 2818.
 L. Jaber-Ansari, M. Hahm, S. Somu, Y. Sanz, A. Busnaina, Y. Jung, J. Am. Chem. Soc.
2009, 131, 804.
 X. Xiong, L. Jaberansari, M. Hahm, A. Busnaina, Y. Jung, Small 2007, 3, 2006.
 L. Lu, W. Yi, D. L. Zhang, Rev. Sci. Instrum. 2001, 72, 2996.
 T. Choi, D. Poulikakos, J. Tharian, U. Sennhauser, Nano Lett. 2006, 6, 1589.
 A. Aliev, M. Lima, E. Silverman, R. Baughman, Nanotechnology 2010, 21, 035709.
 H. Shimoda, S. Oh, H. Geng, R. Walker, X. Zhang, L. McNeil, O. Zhou, Adv. Mater.
2002, 14, 899.
 A. Darhuber, S. Troian, J. Davis, S. Miller, S. Wagner, J. Appl. Phys. 2000, 88, 5119.
 G. Berteloot, C. Pham, A. Daerr, F. Lequeux, L. Limat, Europhys. Lett. 2008, 83,
Figure 1. (a) A tilted SEM of assembled SWNTs film formed on photoresist (PR) micro-lines/
SiO2 substrate. The scale bar is 5 µm and the thickness of PR is 360 nm. (b) An enlarged SEM
image showing quasi-wavy 3D morphology of ultra-thin SWNTs film. The scale bar is 5 µm.
(c) Schematic illustrating the anisotropic and nanostructured of SWNT film. (d) Optical
microscopy image of the SWNT film formed on PR/SiO2 substrate. Note that the background
colors can be observed directly (pink and blue colors representing the PR and SiO2 strips,
respectively) due to the transparent nature of the developed SWNT film. The scale bar is 10
µm. (e) An optical picture of SWNT film assembled on the quartz substrate after removing the