University of Central Florida University of Central Florida
STARS STARS
Faculty Bibliography 2010s Faculty Bibliography
1-1-2010
Directed assembly of solution processed single-walled carbon Directed assembly of solution processed single-walled carbon
nanotubes via dielectrophoresis: From aligned array to individual nanotubes via dielectrophoresis: From aligned array to individual
nanotube devices nanotube devices
Paul Stokes
University of Central Florida
Saiful I. Khondaker
University of Central Florida
Find similar works at: https://stars.library.ucf.edu/facultybib2010
University of Central Florida Libraries http://library.ucf.edu
This Article is brought to you for free and open access by the Faculty Bibliography at STARS. It has been accepted for
inclusion in Faculty Bibliography 2010s by an authorized administrator of STARS. For more information, please
contact STARS@ucf.edu.
Recommended Citation Recommended Citation
Stokes, Paul and Khondaker, Saiful I., "Directed assembly of solution processed single-walled carbon
nanotubes via dielectrophoresis: From aligned array to individual nanotube devices" (2010).
Faculty
Bibliography 2010s
. 828.
https://stars.library.ucf.edu/facultybib2010/828
Directed assembly of solution processed single-walled carbon nanotubes via
dielectrophoresis: From aligned array to individual nanotube devices
Paul Stokes, and Saiful I. Khondaker
Citation: Journal of Vacuum Science & Technology B 28, C6B7 (2010); doi: 10.1116/1.3501347
View online: https://doi.org/10.1116/1.3501347
View Table of Contents: https://avs.scitation.org/toc/jvb/28/6
Published by the American Vacuum Society
ARTICLES YOU MAY BE INTERESTED IN
High quality solution processed carbon nanotube transistors assembled by dielectrophoresis
Applied Physics Letters 96, 083110 (2010); https://doi.org/10.1063/1.3327521
Solution processed large area field effect transistors from dielectrophoreticly aligned arrays of carbon nanotubes
Applied Physics Letters 94, 113104 (2009); https://doi.org/10.1063/1.3100197
Single- and multi-wall carbon nanotube field-effect transistors
Applied Physics Letters 73, 2447 (1998); https://doi.org/10.1063/1.122477
Low voltage, high performance inkjet printed carbon nanotube transistors with solution processed ZrO
2
gate
insulator
Applied Physics Letters 103, 082119 (2013); https://doi.org/10.1063/1.4819465
The work function of the elements and its periodicity
Journal of Applied Physics 48, 4729 (1977); https://doi.org/10.1063/1.323539
Well-aligned and suspended single-walled carbon nanotube film: Directed self-assembly, patterning, and
characterization
Applied Physics Letters 94, 261903 (2009); https://doi.org/10.1063/1.3151850
Directed assembly of solution processed single-walled carbon nanotubes
via dielectrophoresis: From aligned array to individual nanotube
devices
Paul Stokes and Saiful I. Khondaker
a兲
Nanoscience Technology Center and Department of Physics, University of Central Florida, 12424 Research
Parkway, Orlando, Florida 32826
共Received 9 July 2010; accepted 7 September 2010; published 29 November 2010兲
The authors demonstrate directed assembly of high quality solution processed single-walled carbon
nanotube 共SWNT兲 devices via ac dielectrophoresis using commercially available SWNT solutions.
By controlling the shape of the electrodes, concentration of the solution, and assembly time, the
authors are able to control the assembly of SWNTs from dense arrays down to individual SWNT
devices. Electronic transport studies of individual SWNT devices show field effect mobilities of up
to 1380 cm
2
/ V s for semiconducting SWNTs and saturation currents of up to ⬃15
A for metallic
SWNTs. The field effect mobilities are more than an order of magnitude improvement over previous
solution processed individual SWNT devices and close to the theoretical limit. Field effect
transistors 共FET兲 fabricated from aligned two-dimensional arrays of SWNT show field effect
mobility as high as 123 cm
2
/ V s, which is three orders of magnitude higher than the solution
processed organic FET devices. This study shows promise for commercially available SWNT
solution for the parallel fabrication of high quality nanoelectronic devices. © 2010 American
Vacuum Society. 关DOI: 10.1116/1.3501347兴
I. INTRODUCTION
The unique electronic properties of single-walled carbon
nanotubes 共SWNTs兲 make them promising candidates for fu-
ture nanoelectronic devices.
1
For practical applications in na-
noelectronics, it is important that SWNTs are assembled at
selected positions of the circuit with high yield. Chemical
vapor deposition 共CVD兲 growth of SWNTs using litho-
graphically patterned catalytic islands and then making elec-
trical contact to them has been used for the parallel fabrica-
tion of SWNT devices.
2,3
Although CVD grown SWNT has
shown good device properties, high growth temperature
共900 °C兲 is a major bottleneck to make them compatible
with current complementary metal-oxide-semiconductor
共CMOS兲 fabrication technologies.
An attractive alternative to CVD growth techniques for
the high throughput assembly of SWNT electronic devices at
selected positions of the circuit is from postsynthesis fabri-
cation using solution processed SWNTs.
4
Solution process-
ing could be advantageous due to its ease of processing at
room temperature, CMOS compatibility, and potential for
scaled up manufacturing of SWNT devices on various sub-
strates. Several assembly techniques from solution include
chemical and biological patterning,
5,6
flow assisted
alignment,
7
Langmuir–Blodgett assembly,
8
bubble blown
films,
9
contact printing,
10
spin coating assisted alignment,
11
and evaporation driven self-assembly.
12
However, most of
these techniques are used either for only large area devices or
only single nanotube devices. In addition, most of these as-
sembly techniques require postetching to remove excess
SWNTs in the circuit.
Dielectrophoresis 共DEP兲 offers a convenient way in which
SWNTs can be precisely positioned from solution at room
temperature using a nonuniform ac electric field on prepat-
terned electrodes.
13–24
DEP can be advantageous over other
solution processed techniques because it allows for the posi-
tioning from large areas to individual SWNTs at predefined
coordinates of the circuit and does not require the need of
postetching or transfer printing. One crucial aspect of the
DEP process is the quality of the SWNT solution. The solu-
tion should be free of catalytic particles, contain mostly in-
dividual SWNTs, and be stable for long periods of time.
Catalytic particles in the solution tend to make their way into
the electrode gap with the SWNTs during assembly process
due to their highly conducive nature which can disrupt their
device performance. Solutions containing bundles make it
difficult to only obtain individual SWNTs reproducibly into
the electrode gap as the DEP force will likely select the
larger bundles due to their higher dielectric constant and con-
ductivity. Additionally, avoiding degradation of the SWNTs
from processing is extremely important to maintain their ex-
cellent electrical properties.
25
In this article, we used a clean commercially available,
surfactant-free SWNT solution combined with the DEP tech-
nique to achieve directed assembly of high quality SWNT
devices with high yield. By optimizing the device design,
concentration of the solution, and assembly time, we are able
to control the assembly of SWNTs from large scale arrays
down to individual devices. Comparison of the assembly
from commercial solution of SWNT with other homemade
solutions in common organic solvents show that the clean
a兲
Author to whom correspondence should be addressed; electronic mail:
saiful@mail.ucf.edu
C6B7 C6B7J. Vac. Sci. Technol. B 28„6…, Nov/Dec 2010 1071-1023/2010/28„6…/C6B7/6/$30.00 ©2010 American Vacuum Society
assembly was obtained from the commercial solution. Elec-
tronic transport properties of individual semiconducting
SWNTs show field effect mobilities up to 1380 cm
2
/ V s and
saturation currents up to ⬃15
A for metallic SWNTs. The
field effect mobilities are more than an order of magnitude
improvement over previous solution processed individual
SWNT devices and close to the theoretical limit. Addition-
ally, field effect transistors 共FET兲 fabricated from aligned
two-dimensional 共2D兲 arrays of SWNT show field effect mo-
bilities as high as 123 cm
2
/ V s, which is three orders of
magnitude higher than solution processed organic FET de-
vices.
II. EXPERIMENTAL DETAILS
A. Electrode design and fabrication
Devices were fabricated on heavily doped silicon sub-
strates capped with a thermally grown 250 nm thick SiO
2
layer. The electrode patterns were fabricated by a combina-
tion of optical and electron beam lithography 共EBL兲. First,
contact pads and electron beam markers were fabricated with
optical lithography using double layer resists 共LOR 3A/
Shipley 1813兲 developing in CD26, thermal evaporation of 3
nm Cr and 50 nm Au followed by lift-off. Smaller electrode
patterns were fabricated with EBL using single layer PMMA
resists and then developing in 共1:3兲 methyl isobutyl keto-
ne:isopropal alchohol 共MIBK:IPA兲. After defining the pat-
terns, 2 nm Cr and 25 nm thick Pd were deposited using
electron beam deposition followed by lift-off in warm ac-
etone. Pd was used because it is known to make the best
electrical contact to SWNTs.
3
Figures 1共a兲 and 1共b兲 show a
cartoon of the electrode patterns for the DEP assembly of
individual SWNTs and parallel arrays of SWNTs, respec-
tively. The electrode patterns for the alignment of individual
tubes use a pair of adjacent taper shaped electrodes with
sharp tips separated by 1
m, whereas the electrode patters
for aligned arrays of SWNTs were done using 200
m long
parallel electrodes with 5
m spacing.
B. Solution preparation
We used three different SWNT solutions for the DEP as-
sembly: 共i兲 A homemade dimethylformamide 共DFM兲 solu-
tion, 共ii兲 homemade dichloroethane 共DCE兲 solution, and 共iii兲
already suspended, surfactant-free aqueous SWNT solution
purchased from Brewer Science Inc.
26
The DMF-SWNT sus-
pension was made by ultrasonically dispersing HiPCO
grown SWNTs 共Carbon Nanotechnologies Inc.兲 in ⬃5ml
DMF and 1 ml trifluoroacetic acid 共TFA兲. TFA was used to
dissolve any unwanted catalytic particles and amorphous car-
bon from the bulk material. After dissolving the SWNTs in
DMF/TFA, the solution was centrifuged, the supernatant was
decanted, and the solid is then redispersed for further
dispersion/centrifugation/decantation cycles. The final solu-
tion was diluted until it became clear and then sonicated for
several minutes before assembly. The DCE mixture was
made by simply adding a very small pinch of the HiPCO
SWNT soot to ⬃4 ml of DCE and then sonicating for
⬃5 – 10 min before the assembly. The commercial solution
has an original SWNT concentration of ⬃50
g/ ml and was
diluted using de-ionized 共DI兲 water to a desired concentra-
tion.
C. Dielectrophoretic assembly
The directed assembly of SWNTs at predefined electrode
positions was done in a probe station under ambient condi-
tions. A small drop of SWNT solution was cast onto the chip
containing the electrode pairs. An ac voltage of 1 MHz,
5V
p-p
was applied using a function generator between the
source and drain electrodes or by a simultaneous deposition
technique
14,22
共between source and gate兲 to align at several
electrode pairs simultaneously. The ac voltage gives rise to a
time averaged dielectrophoretic force. For an elongated ob-
ject, it is given by F
DEP
⬀
m
Re关K
f
兴ⵜE
rms
2
, K
f
=共
p
ⴱ
−
m
ⴱ
兲/
m
ⴱ
,
and
p,m
ⴱ
=
p,m
−i共
p,m
/
兲, where
p
and
m
are the permit-
tivities of the nanotube and, solvent respectively, K
f
is the
Claussius–Mossotti factor,
is the conductivity, and
=2
f is the frequency of the applied ac voltage.
27
The in-
duced dipole moment of the nanotube interacting with the
strong electric field causes the nanotubes to move in a trans-
lational motion along the electric field gradient.
Figures 1共c兲 and 1共d兲 show a simulation of the electric
field around the electrode gap for the adjacent taper shaped
electrode and parallel plate electrode, respectively. The simu-
lations were done using a commercially available software
共
FLEX PDE兲 assuming that the potential phasor is real and
therefore using the electrostatic form of the Laplace equation
共ⵜ
2
⌽=0兲.
14
Hence we can set the effective potential of the
electrodes to ⌽ = ⫾V
p-p
/ 2 for our simulation. From Fig.
1共c兲, it can be seen that the strongest electric field lines are
confined at the sharp tips. This increases the probability of
aligning individual SWNTs. For the parallel plate geometry,
the electric field is uniform throughout the electrode gap al-
lowing for many nanotubes to align parallel to one another
throughout the gap. After the assembly, the function genera-
FIG.1. 共Color online兲共a兲 Cartoon of the electrode patterns for DEP assem-
bly of 共a兲 individual SWNTs and 共b兲 2D arrays of SWNTs. 2D simulated
electric field around the electrode gap for 共c兲 the taper shaped electrodes and
共d兲 parallel plate electrodes.
C6B8 Paul Stokes and Saiful I. Khondaker: Directed assembly of solution processed SWNTs C6B8
J. Vac. Sci. Technol. B, Vol. 28, No. 6, Nov/Dec 2010
tor was turned off and the sample was blown dry by a stream
of nitrogen gas.
III. RESULTS AND DISCUSSIONS
A. Controlling the assembly of individual SWNTs
Figure 2 shows the effect of SWNT concentration on the
DEP assembly using the commercial solution diluted in DI
water. We use a simultaneous deposition technique
14,22
in
this case, applying the ac field between source and gate for 3
min. Figure 2共a兲 shows an atomic force microscopy 共AFM兲
image of a device after the assembly for a SWNT concentra-
tion of 1000 ng/ml. Dilution of the solution by ten times to
100 ng/ml concentration yield less SWNTs in the gap, as
shown in Fig. 2共b兲. It can be seen here in both cases that the
SWNTs mimic the electric field lines around the electrode
gap, as simulated in Fig. 1共c兲. The yield for the 1000 ng/ml
and 100 ng/ml concentration is ⬎95%. By diluting the solu-
tion to 10 ng/ml, we obtained an individual SWNT in the gap
关Fig. 2共c兲兴. The diameter of this individual SWNT is
⬃2.0 nm, measured by AFM. Figure 2共d兲 shows a histogram
of ⬃100 individual SWNTs giving an average diameter of
2.0⫾0.2 nm. Approximately 10% of the diameters are
greater than 3.5 nm, which is an indication of the possible
presence of some double-walled nanotubes or large diameter
SWNTs.
28
The total yield of individual SWNTs at low con-
centration is ⬃20% on average and as high as 35% for a
single chip. Figure 3 shows a number of SEM images of
individual nanotubes assembled by this technique. It can be
seen here that the devices are free of bundles and catalytic
particles which stems from the quality of the commercial
solution.
B. Comparison of solutions
We investigated the effect of the different solutions at low
concentration 共⬃10 ng/ ml兲 on the DEP assembly. In con-
trast to the commercial solution in water, for the DMF and
DCE solutions, we apply the ac voltage between one pair at
a time for ⬃5 s. This is done because long trapping times
are more complex to use as DMF and DCE evaporate
quickly in air. Figure 4共a兲 shows a representative AFM im-
age after the assembly for the DMF solution. As can be
clearly seen here, the resulting SWNT deposition contains a
bundle and catalytic particles shown by the arrows. At cer-
tain areas along the SWNT, the diameter is as large as 30–40
nm. Figure 4共b兲 shows the representative AFM image of a
device after assembling the SWNTs using DCE. The result-
ing device also contains catalytic particles near the electrode
tip on the right and shows diameters up to 10 nm along the
SWNT. Figure 4共c兲 shows an AFM image of a device after
assembly by using the Brewer Science SWNT solution. It is
clear from the AFM image that the SWNT is individual and
does not contain any catalytic particles. The commercial so-
lution turned out to be stable for months, therefore increasing
the reproducibility. For the DCE solution we were able to
assemble ⬃10 individual SWNT devices out of ⬃80 tries,
however, all of them contained catalytic particles 共with
diameter⬎10 nm兲 attached to the tubes. Another problem
that arises when using DCE is that it evaporates in air very
quickly, it is highly toxic, volatile, and did not remain stable
for more than a few hours. In the DMF case, the solutions
also only remained stable for a short period of time and the
results often came with larger diameter bundles ⬎15 nm out
of ⬃50 tries.
As can be concluded here, the commercial solution
yielded the best results for the assembly of clean and indi-
vidual SWNT devices. The good results from the commercial
solution are due to several reasons. First, its very low density
of impurity particles 共less than 50 ppb兲
26
is particularly ad-
vantageous in the DEP process because catalytic particles in
the solution tend to make their way into the electrode gap
during DEP due to their high conductivity. This is displayed
FIG.2. 共Color online兲 AFM image of nanotubes assembled between the
electrodes with SWNT concentrations of 共a兲 1000 ng/ml 共b兲 100 ng/ml, and
共c兲 10 ng/ml in the solution. Scale bar: 1
m in all images. 共d兲 Histogram of
diameters for over 100 nanotube devices.
FIG.3. 关共a兲-共l兲兴 SEMs of several individual SWNT assembled via DEP from
commercial solution at a concentration of ⬃10 ng/ml. Gap between the
electrodes is 1
m.
FIG.4. 共Color online兲 Representative AFM images of SWNTs assembled via
DEP 共a兲 from DMF solution, 共b兲 from DCE solution, and 共c兲 the Brewer
Science solution.
C6B9 Paul Stokes and Saiful I. Khondaker: Directed assembly of solution processed SWNTs C6B9
JVSTB-MicroelectronicsandNanometer Structures