Fully Transparent Thin-Film Transistor Devices Based on SnO Nanowires 2

Fully Transparent Thin-Film Transistor Devices Based on SnO Nanowires 2
NANO
LETTERS
Fully Transparent Thin-Film Transistor
Devices Based on SnO2 Nanowires
2007
Vol. 7, No. 8
2463-2469
Eric N. Dattoli,† Qing Wan,†,§ Wei Guo,‡ Yanbin Chen,‡ Xiaoqing Pan,‡ and
Wei Lu*,†
Department of Electrical Engineering and Computer Science, and
Department of Materials Science and Engineering, UniVersity of Michigan,
Ann Arbor, Michigan 48109
Received May 23, 2007; Revised Manuscript Received June 6, 2007
ABSTRACT
We report on studies of field-effect transistor (FET) and transparent thin-film transistor (TFT) devices based on lightly Ta-doped SnO2 nanowires. The nanowire-based devices exhibit uniform characteristics with average field-effect mobilities exceeding 100 cm2/V‚s. Prototype nanowire-based TFT (NW−TFT) devices on glass substrates showed excellent optical transparency and transistor performance in terms of
transconductance, bias voltage range, and on/off ratio. High on-currents and field-effect mobilities were obtained from the NW−TFT devices
even at low nanowire coverage. The SnO2 nanowire-based TFT approach offers a number of desirable properties such as low growth cost,
high electron mobility, and optical transparency and low operation voltage, and may lead to large-scale applications of transparent electronics
on diverse substrates.
Semiconductor nanowires (NWs) have attracted increasing
interest in the last a few years and are expected to lead to
novel device applications.1-3 In particular, it has been shown
recently that high-performance nanowire-based thin-film
transistor (TFT) devices can be fabricated on a variety of
substrates including glass and plastics.4-7 The key advantage
of the NW-TFT approach compared to conventional TFT
techniques is the clear separation of the device fabrication
stage from the material growth stage, such that there is no
longer need to be concerned with compatibility with the
deVice substrate during growth, and high growth temperature
can be used to obtain crystalline materials. Specifically, the
synthesis of NW building blocks is first carried out under
conditions optimized to yield high-quality single-crystal
materials, where the desired electronic and/or optoelectronic
properties are defined by material composition, structure, and
size. The NWs are then transferred from the growth substrate
onto the desired device substrate and configured into a thinfilm form, followed by conventional fabrication processes
to produce TFT devices.4,5
In previous studies on NW-TFT devices, the channel was
typically formed by opaque semiconductor nanowires, such
as silicon, and the source, drain, and gate electrodes were
* To whom correspondence should be addressed. E-mail: wluee@
eecs.umich.edu.
† Department of Electrical Engineering and Computer Science, the
University of Michigan.
‡ Department of Materials Science and Engineering, the University of
Michigan.
§ Present address: Micro-Nano Technologies Research Center, Hunan
University, Changsha 410082, People’s Republic of China.
10.1021/nl0712217 CCC: $37.00
Published on Web 06/27/2007
© 2007 American Chemical Society
made of normal metals such as Au or Ni.4-6 On the other
hand, high-performance optically transparent TFTs will lead
to new frontiers such as “invisible electronics” and are of
great current interest. Here we report fully transparent
TFT devices based on SnO2 nanowire films and indium tin
oxide (ITO) electrodes. Key performance metrics including
mobility, on/off ratio, and operating voltages of the SnO2
NW-TFT devices fabricated on glass substrates are similar
to those of individual nanowires and are better than those of
conventional metal oxide TFTs fabricated near room temperature.8,9 The performance metrics are also comparable
with or better than conventional transparent TFTs using
single-crystalline channel materials, which typically require
high-temperature growth/annealing and expensive singlecrystalline substrates.10-14
SnO2 nanowires were chosen in our studies due to the low
growth cost, high optical transmittance, and the ease of
obtaining Ohmic contacts with conventional transparent
conducting oxide films. Particularly, single-crystalline SnO2
nanowires may be grown using a simple vapor transport
synthesis method that allows for the growth of large
quantities of nanowires at a cost lower than that of indiumbased metal oxide materials15 or silicon nanowires and carbon
nanotubes.16 Previous studies on SnO2 nanowire or nanobelt
structures, however, have focused on undoped materials in
which the carriers (electrons) were provided by deviation
from stoichiometry (in the form of oxygen vacancies) or
unintentional doping by (uncontrolled) impurities in the
growth facility.17-19 The undoped nanowire samples typically
exhibit low carrier concentrations and are very sensitive to
changes in ambient conditions. These properties, desirable
for environmental sensors,19 are nonetheless detrimental for
high-performance transistor device applications. To address
this problem, we have recently developed a reliable in situ
doping process that has been shown to dramatically affect
the nanowires’ electrical properties in a controlled fashion.16,20 For instance, degenerately Sb-doped SnO2 nanowires
have been demonstrated to exhibit resisitivites as low as the
best thin-film samples and show metal-like behavior.16 In
this work, we focus on lightly Ta-doped SnO2 nanowires
that serve as the channel material in nanowire FET and TFT
devices. The Ta-doped SnO2 nanowires were synthesized on
(100) Si substrates by a catalyst-mediated vapor-liquidsolid (VLS) process16,20,21 in which the Ta and Sn source
materials were provided by a vapor transport method.16,20
Briefly, high-purity (99.99%) powders containing the source
materials (Sn:Ta ) 95:5 in weight ratio) were first mixed
thoroughly and loaded in an alumina boat. The (100) Si
growth substrates were sputter-deposited with 5 nm of Au
film, serving as the catalysts in the VLS process, and were
placed on top of the boat. The alumina boat was then loaded
into an alumina reactor tube positioned inside a horizontal
tube furnace. During growth, the furnace was heated from
room temperature to 900 °C at a rate of 20 °C/min under Ar
flow (500 sccm) with a trace amount of oxygen. The growth
time was 1 h at 900 °C, followed by cool down to room
temperature at a rate of 5 °C/min.
Figure 1a shows a low-magnification scanning electron
microscopy (SEM) image of the as-synthesized sample.
Large quantities of nanowires were observed only in the areas
covered with Au catalysts. The nanowires have a mean lateral
size of 55 nm and are typically tens of micrometers long.
Structural properties of the nanowires were further investigated with transmission electron microscopy (TEM) studies.
Figure 1b shows a low-magnification TEM image of a single
Ta-doped SnO2 nanowire with lateral size of ca. 60 nm,
illustrating the uniform lateral size and lack of tapering along
the growth direction. The crystallography of the nanowires
was studied by select area electron diffraction (SAED) (upper
inset, Figure 1b) and HRTEM imaging (lower inset, Figure
1b). These results show that the Ta-doped SnO2 nanowires
have a tetragonal rutile crystal structure (a ) 0.474 nm, and
c ) 0.318 nm). HRTEM studies also verify that each Tadoped SnO2 nanowire is a perfect single crystal with no
visible dislocations and amorphous surface overcoating. The
growth direction of the Ta-doped SnO2 nanowire shown in
Figure 1b was determined to be [101] based on the analysis
of the diffraction pattern.
After growth, the nanowires were transferred to the
corresponding device substrates and a number of device
structures were fabricated and tested. Unless noted otherwise,
all electrical measurements were carried out in air at room
temperature. Single-nanowire-based FETs with a standard
back-gated structure were studied first to investigate the
intrinsic electrical properties of the Ta-doped SnO2 nanowires. The device fabrication involves removing the nanowires from the Si growth substrate by sonication in isopropyl
2464
Figure 1. Structural characterization of the Ta-doped SnO2
nanowires. (a) SEM image of the as-grown nanowires displaying
the high-growth density. Scale bar: 2 µm. (b) Low-magnification
TEM image of a single Ta-doped SnO2 nanowire with lateral size
of ca. 60 nm. Scale bar: 100 nm. Top inset: select area electron
diffraction (SAED) pattern of the SnO2 nanowire. Bottom inset:
HRTEM image of the same SnO2 nanowire. Scale bar: 5 nm.
alcohol and nanowire deposition onto a degenerately doped
n+ silicon substrate capped with a 50 nm silicon dioxide
(SiO2) layer by drop drying. Photolithography was then used
to define pairs of source/drain electrodes to contact each
nanowire, followed by the metal deposition of Ti/Au
(10/100 nm) by electron beam evaporation to complete the
device structure, with the n+ Si substrate serving as the back
gate (Figure 2a, inset).
The Ta-doped SnO2 nanowire FETs were found to exhibit
standard n-type transistor behavior. Doping of the SnO2
nanowires also helps to reduce contact resistance in the FET
devices, as evidenced by the apparent absence of Schottky
behavior in the current-voltage (Ids-Vds) characteristics
(Figure 2b). In contrast, our control experiments on undoped
SnO2 nanowires showed pronounced Schottky barrier behavior (Figure S1, Supporting Information), consistent with
Nano Lett., Vol. 7, No. 8, 2007
Figure 2. Transistor characteristics of back-gated Ta-doped SnO2 nanowire FET devices on silicon substrates. (a) Current (Ids) vs gate
voltage (Vgs) curve at linear region (Vds ) 1V) for a back-gated NW-FET. Inset: schematic of the device. (b) Family of Ids-Vds curves for
the same device with Vgs ) 6 to -4 V in -2 V steps from top to bottom. (c) Ids-Vgs curve in log scale at Vds ) 1V. The subthreshold slope
S was estimated to be 270 mV/decade. Inset, SEM image of the device. Scale bar: 2 µm. (d) Histogram of the extracted field-effect
mobilities (µfe) for 75 devices.
earlier studies.18,19 The field-effect mobility µfe of the device
in Figure 2 was estimated using the equation
gm )
µfeCg
Vds
L2
(1)
in the linear operation regime. Here gm ) dIds/dVds is the
linear-region transconductance, and L ) 4.95 µm is the
nanowire device channel length. Cg is the capacitance of the
back gate and can be estimated by utilizing the cylinderon-plane model:
Cg )
2πr0
L
2h + d
cosh-1
d
(
)
(2)
where 0 is the vacuum dielectric constant, h ) 50 nm is
the thickness of the SiO2 layer, and d ) 87 nm is the lateral
size of the nanowire. r is the relative dielectric constant and
was chosen to be 2.5, which is the average of air (1) and
SiO2 (3.9). We note eq 2 evolves into the commonly used
capacitance model22 Cg ) 2πr0/ln(4h/d) when h . d, and
the estimated capacitance value of 100 aF/µm using eq 2
Nano Lett., Vol. 7, No. 8, 2007
also agrees well with results obtained from finite element
simulations. Using the estimated capacitance value and
measured device parameters, µfe of the device in Figure 2
was estimated to be 120 cm2/(V‚s) in the linear bias region.
Other device parameters were also obtained in transport
studies. The subthreshold slope S was estimated to be 270
mV/decade and independent of the bias voltage. An on/off
ratio of >105 was achieved within the 10 V bias range.
Finally, the maximum transconductance was obtained in
saturation and was found to be 2.94 µS at Vds ) 10 V. The
histogram of the extracted µfe for 75 devices was plotted in
Figure 2d with an average µfe value of 156 cm2/V‚s. We
note this value is consistent with those obtained in planar
single-crystalline SnO2 samples23 and other metal oxide
nanowires.24
To demonstrate the potential of Ta-doped SnO2 nanowires
as transparent devices, we further fabricated nanowire
transistors on glass substrates in which the back-gate, source,
and drain electrodes were replaced with transparent conducting Sn-doped In2O3 (ITO) films (Figure 3a, inset). Briefly,
an ITO film with thickness of ca. 250 nm was first deposited
on a glass substrate (Fisherbrand, 2.5 cm × 2.5 cm, 250 µm
thick) by pulsed laser deposition (PLD) at 400 °C, followed
by the SiO2 gate dielectric layer (75 nm thick) deposition
2465
Figure 3. Transistor characteristics of a back-gated transparent Ta-doped SnO2 nanowire FET on glass with ITO contacts. (a) Ids-Vgs
curve in the linear region (Vds ) 0.1 V). The subthreshold slope S was estimated to be 312 mV/decade, and µfe was estimated to be 178
cm2/V‚s. Inset, schematic of the device. (b) Family of Ids-Vds curves for the same device with Vgs ) 6 to -8 V in -2 V steps from top
to bottom. Inset: SEM image of the device. Scale bar: 2 µm.
via plasma-enhanced chemical vapor deposition (PECVD).
The bottom ITO layer and SiO2 layer serve as the back-gate
and dielectric, respectively. Single-nanowire transistor devices were then fabricated on the SiO2/ITO/glass substrate
in the same fashion as the back-gated device in Figure 2,
except that the source/drain electrodes were replaced by 200
nm thick PLD deposited ITO films.
The devices on glass substrates show transmittances of
∼80% in the visible light range of 380-800 nm (Figure S2,
Supporting Information). Significantly, no performance
degradation was observed on devices fabricated on glass
substrates with ITO electrodes. As seen in Figure 3b, linear
Ids-Vds behavior was still observed at small bias, indicating
Ohmic contacts with ITO source/drain electrodes. At high
bias, “hard saturation” was once again observed for the
transparent FET, with a saturation transconductance gm )
2.27 µS measured at Vds ) 10 V. The on/off ratio and
subthreshold slope S were inferred from Figure 3a to be 105
and 312 mV/decade, respectively. The field-effect mobility
µfe was estimated to be 179 cm2/V‚s in the linear region.
These values are consistent with those obtained on devices
fabricated on silicon substrates with Ti/Au S/D electrodes
(e.g., the device in Figure 2) and clearly demonstrate the
potential of SnO2 nanowires in fully transparent electronics
applications.
Following studies on single-nanowire devices, fully transparent thin-film transistors were fabricated on glass substrates
using arrays of parallel Ta-doped SnO2 nanowires as the
transistor channel with a staggered top-gated TFT design
(Figure 4b). Aligned Ta-doped SnO2 nanowires were transferred onto a Pyrex glass substrate (500 µm thick) with
surface coverage of at least 25% (Figure 4a) using a physical
transfer method.5 Conventional sputter, photolithography, and
liftoff processes were used to pattern the device structure.
Several improvements were made compared to the backgated devices to improve the device performance and to
facilitate large-scale applications. First, we note that PLD
likely will not be a scalable technique to fabricate the ITO
electrodes for large scale TFT devices. Second, the deposition
2466
temperatures of the ITO films and the SiO2 dielectric (both
at 400 °C) are too high for device applications on plastic
substrates. To address these issues, sputter-deposited ITO
films and SiO2 dielectric were used. The ITO films were
deposited by RF sputtering using an In2O3/SnO2 target
(90/10 by wt) at 400 W without substrate heating. Afterward the ITO films were annealed at 250 °C inside a furnace to improve the conductivity and transparency. The
SiO2 dielectric was deposited by RF sputtering using a SiO2
target at 700 W without substrate heating or annealing.
The TFT devices fabricated using this approach are still
highly transparent in the visible range, as shown in Figure 4c,d, with a transmittance of ∼70%, including the pyrex
glass substrate. Significantly, the highest temperature in
the TFT fabrication process was limited to 250 °C in our
approach, making this approach compatible with certain
plastic substrates for applications of transparent, flexible
electronics.
The single-crystalline channel of the nanowire-based TFT
devices affords excellent performance metrics. Electrical
characterizations obtained on a top-gated NW-TFT device
with a W/L ratio of 12.6 (W ) 48 µm, L ) 3.8 µm) are
shown in Figure 5. The device displays enhancement-mode
n-type transistor behavior, with clear linear and saturation
regions observed in the Ids-Vds output curves (Figure 5b).
Significantly, a large on current Ion ) 71 µA, transconductance of 49 µS, and on/off ratio ∼103 can be obtained within
a small Vdd bias window of 2.5 V (with Vgs from 3.5 to 6 V
and Vds ) 2.5 V). Compared with amorphous or polycrystalline transparent TFT devices fabricated near room temperature, the NW-TFT device shows better performance
within a smaller Vdd bias window.8,9 The subthreshold slope
S was measured to be 538 mV/decade from Figure 5c. This
value is comparable with that obtained in silicon nanowire
TFTs4 and may be further improved by improving the
nanowire/dielectric interface or using high-k dielectrics.15,25,26
Figure 5c also shows that the device exhibits a small gate
leakage of less than 10 nA within the bias window, and
negligible drain-induced barrier-lowering (DIBL) effects.
Nano Lett., Vol. 7, No. 8, 2007
Figure 4. Structure of the NW-TFT devices. (a) Dark-field optical microscope image of a SnO2 nanowire film obtained through the
physical transfer method. Scale bar: 20 µm. (b) Schematic of the transparent SnO2 nanowire-based TFT device, showing the ITO source/
drain and gate electrodes, and the staggered transistor structure. (c) Digital photograph of a 200-device array of NW-TFTs fabricated on
500 µm thick Pyrex glass. The logo of University of Michigan can be clearly seen through the NW-TFT array. The dashed region illustrates
the device array area. (d) Optical transmittance spectrum of the NW-TFT device array on glass substrate (solid line) and the glass substrate
alone (dashed line).
The field-effect mobility (µfe) of this TFT device can be
estimated using eq 1 by noting that Cg is now the total gate
capacitance of the TFT device. A conservative estimate of
Cg ) C0 × (W × L) was used in our estimate, corresponding
to the ideal case of full nanowire coverage. Here C0 is the
capacitance per unit area using a parallel plate model C0 )
r0/d ) 265.5aF/µm2, and W ) 48 µm, L ) 3.8 µm are the
physical device width and length, respectively. With the
estimated Cg values and measured device parameters, µfe was
estimated to be 145 cm2/V‚s in the linear region and 112
cm2/V‚s in the saturation region.
The mobility values estimated using the conservative
parallel-plate model are in fact very close to the average
mobility value obtained on individual NW devices. This
result was unexpected at first sight considering the capacitance obtained from the parallel-plate model assuming 100%
nanowire coverage may overestimate the actual gate capacitance due to the ∼25% nanowire coverage observed for the
device in Figure 5. Here the term nanowire coverage denotes
the ratio of the area covered by the transferred nanowires to
that of the physical channel area (W × L). To address this
apparent discrepancy, we performed electrostatic simulations
(Figure 5d) using measured parameters of the fabricated
NW-TFT device structure (130 nm thick SiO2 gate dielectric
was used for the device in Figure 5). Our study verified that
Nano Lett., Vol. 7, No. 8, 2007
the estimated Cg value using the parallel plate model is in
fact within 89% of the more exact Cg value obtained from
electrostatic simulations when the nanowire coverage exceeds
25%. This observation can be attributed to the fact that, above
a certain threshold, the nanowire array can effectively screen
the field lines from the gate just as in the continuous film
case (Figure 5d).27 Because the TFT drain current (Ids) and
transconductance (gm) are directly related to Cg, one can
conclude that increasing the nanowire coverage above 25%
in the present device structure will have little effect on the
TFT performance. Looking at this problem from a different
angle, the measured transconductance gm (therefore µfe) for
a NW-TFT composed of N parallel NWs can be rewritten
as the sum of the transconductance from N individual NW
devices:
µfeCw
Vds
L2
gm ) N
(3)
assuming the nanowires are identical. Here Cw corresponds
to the effective gate capacitance for a single-nanowire (cf.
eq 1). The unexpected device performance can be explained
by noticing the fact that the effective gate capacitance per
wire Cw decreases with nanowire coverage. Qualitatively,
in the extreme case when only one wire forms the channel,
2467
Figure 5. Electrical characterization of a transparent NW-TFT device on glass. (a) Ids-Vgs curve at Vds ) 0.1V. µfe was estimated to be
145 cm2/V‚s. Inset, optical dark-field microscope image of the device. The dashed lines highlight the source and drain electrodes. Scale
bar: 10 µm. (b) Family of Ids-Vds curves for the same device with Vgs ) 6 to 3.5 V in -0.5 V steps. (c) Ids vs Vgs curve in log scale for
Vds ) 0.1 V (blue solid line) and 3 V (black dashed line). The gate leakage current Igs at Vds ) 1 V was also plotted (red dash-dotted line).
The subthreshold slope S was estimated to be 538 mV/decade. (d) Distribution of the electrical fields in the
cross-section of a NW-TFT with 25% nanowire coverage obtained through electrostatic simulation (Maxwell Ansoft 2D, version 11) at
Vgs ) 1 V.
field lines from the NW extends over the whole gate area
and Cw is determined by eq 2 from the cylinder-on-plane
model. In other words, the effective capacitor area A is much
larger than the size of the nanowire itself. As N and the
nanowire coverage increases, the field lines from each
nanowire are confined to smaller areas (e.g., Figure 5d). As
a result, the effective capacitor area A for each nanowire
decreases, resulting in smaller Cw. Above a certain nanowire
coverage, the effects of increasing N and decreasing Cw
almost completely cancel each other, and the performance
metrics of the NW-TFT device will no longer improve with
increased nanowire coverage.
This observation has important consequences in nanowirebased electronics, as it shows that nanowire arrays produced
using less-than-optimal assembly techniques with relatively
low coverage can still act effectively as thin-film devices.
Figure 6 shows the calculated gate capacitance Cg as a
function of the nanowire coverage at different oxide thickness
conditions. For relatively thick gate dielectric thicknesses
(blue squares) that can be readily produced with inexpensive,
scalable techniques such as sputtering, a nanowire film with
low surface coverage will behave similarly to a continuous
planar single-crystalline thin-film of the same material and
thickness. To take advantage of higher coverage (denser)
nanowires as the TFT channel, thinner gate dielectrics or
2468
Figure 6. Gate capacitance of the NW-TFT as a function of the
nanowire coverage in the channel region at different dielectric
thicknesses. The gate capacitance is normalized to the maximum
value when the nanowire coverage is 100%.
high-k materials will have to be used so that the gate is more
effectively coupled to the nanowire array (red circles, Figure
6).
In conclusion, single-crystalline Ta-doped SnO2 nanowires
were obtained using a simple low-cost growth method.
Nano Lett., Vol. 7, No. 8, 2007
Electrical characterizations on single-nanowire devices show
that the nanowires can serve as channel materials in
transparent transistor devices with field-effect mobilities over
100 cm2/V‚s. Transparent NW-TFT devices were produced
near room temperature and were found to exhibit similar
high-mobility values even at low nanowire coverage. This
study may lead to large-scale applications of high-performance transparent nanowire-based thin-film devices on
diverse substrates such as flexible electronics on plastics.
Acknowledgment. This work was supported in part by
the National Science Foundation (ECS-0601478). This work
used the Michigan Nanofabrication Facility (MNF) at UM,
a member of the National Nanotechnology Infrastructure
Network (NNIN) funded by the NSF.
Supporting Information Available: Ids-Vds curve of
undoped SnO2 devices, optical transmittance spectrum of an
array of bottom-gated single-nanowire transistor device on
glass, and optical micrograph of array of the top-gated NWTFT structure. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) Lieber, C. M. MRS Bull. 2003, 28, 486.
(2) Thelander, C.; Agarwal, P.; Brongersma, S.; Eymery, J.; Feiner, L.
F.; Forchel, A.; Scheffler, M.; Riess, W.; Ohlsson, B. J.; Gosele, U.;
Samuelson, L. Mater. Today 2006, 9, 28.
(3) Wang, Z. L. Mater. Today 2004, 7, 26.
(4) Duan, X. F.; Niu, C. M.; Sahi, V.; Chen, J.; Parce, J. W.; Empedocles,
S.; Goldman, J. L. Nature 2003, 425, 274.
(5) Javey, A.; Nam, S.-W.; Friedman, R. S.; Yan, H.; Lieber, C. M. Nano
Lett. 2007, 7, 773.
(6) McAlpine, M. C.; Friedman, R. S.; Jin, S.; Lin, K.-h.; Wang, W. U.;
Lieber, C. M. Nano Lett. 2003, 3, 1531.
(7) Tao, A.; Kim, F.; Hess, C.; Goldberger, J.; He, R.; Sun, Y.; Xia, Y.;
Yang, P. Nano Lett. 2003, 3, 1229.
(8) Nomura, K.; Ohta, H.; Takagi, A.; Kamiya, T.; Hirano, M.; Hosono,
H. Nature 2004, 432, 488.
Nano Lett., Vol. 7, No. 8, 2007
(9) Fortunato, E. M. C.; Barquinha, P. M. C.; Pimentel, A. C. M. B. G.;
Gonçlves, A. M. F.; Marques, A. J. S.; Pereira, L. M. N.; Martins,
R. F. P. AdV. Mater. 2005, 17, 590.
(10) Dehuff, N. L.; Kettenring, E. S.; Hong, D.; Chiang, H. Q.; Wager, J.
F.; Hoffman, R. L.; Park, C.-H.; Keszler, D. A. J. Appl. Phys. 2005,
97, 064505.
(11) Chiang, H. Q.; Wager, J. F.; Hoffman, R. L.; Jeong, J.; Keszler, D.
A. Appl. Phys. Lett. 2005, 86, 013503.
(12) Presley, R. E.; Munsee, C. L.; Park, C.-H.; Hong, D.; Wager, J. F.;
Keszler, D. A. J. Phys. D 2004, 37, 2810.
(13) Hoffman, R. L.; Norris, B. J.; Wager, J. F. Appl. Phys. Lett. 2003,
82, 733.
(14) Nomura, K.; Ohta, H.; Ueda, K.; Kamiya, T.; Hirano, M.; Hosono,
H. Science 2003, 300, 1269.
(15) Wang, L.; Yoon, M.-H.; Lu, G.; Yang, Y.; Facchetti, A.; Marks, T.
J. Nat. Mater. 2006, 5, 893.
(16) Wan, Q.; Dattoli, E. N.; Lu, W. Appl. Phys. Lett. 2007, 90, 222107.
(17) Liu, Z.; Zhang, D.; Han, S.; Li, C.; Tang, T.; Jin, W.; Liu, X.; Lei,
B.; Zhou, C. AdV. Mater. 2003, 15, 1754.
(18) Hernandez-Ramirez, F.; Tarancon, A.; Casals, O.; Rodriguez, J.;
Romano-Rodriguez, A.; Morante, J. R.; Barth, S.; Mathur, S.; Choi,
T. Y.; Poulikakos, D.; Callegari, V.; Nellen, P. M. Nanotechnology
2006, 17, 5577.
(19) Arnold, M. S.; Avouris, P.; Pan, Z. W.; Wang, Z. L. J. Phys. Chem.
B 2003, 107, 659.
(20) Wan, Q.; Dattoli, E. N.; Fung, W. Y.; Guo, W.; Chen, Y.; Pan, X.;
Lu, W. Nano Lett. 2006, 6, 2909.
(21) Lu, J. G.; Chang, P. C.; Fan, Z. Y. Mater. Sci. Eng. R 2006, 52, 49.
(22) Cui, Y. Y.; Duan, X.; Hu, J.; Lieber, C. M. J. Phys. Chem. B 2000,
104, 5213.
(23) Gordon, R. G. MRS Bull. 2000, 25, 52.
(24) Li, C.; Zhang, D. H.; Han, S.; Liu, X. L.; Tang, T.; Zhou, C. W.
AdV. Mater. 2003, 15, 143.
(25) Klauk, H.; Zschieschang, U.; Pflaum, J.; Halik, M. Nature 2007, 445,
745.
(26) Xiang, J.; Lu, W.; Hu, Y.; Wu, Y.; Yan, H.; Lieber, C. M. Nature
2006, 441, 489.
(27) We note that after we finished drafting the manuscript and performed
all the experimental and analytical work, an independent study was
published discussing similar effects on arrays of aligned carbon
nanotubes (Kang, S. J.; Kocabas, C.; Ozel, T.; Shim, M.; Pimparkar,
N.; Alam, M. A.; Rotkin, S. V.; Rogers, J. A. Nat. Nanotechnol.
2007, 2, 230).
NL0712217
2469
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising