APPLIED PHYSICS LETTERS
VOLUME 85, NUMBER 15
11 OCTOBER 2004
Organic thin-film transistors with nanocomposite dielectric gate insulator
Fang-Chung Chen, Chih-Wei Chu, Jun He, and Yang Yanga)
Department of Materials Science and Engineering, University of California, Los Angeles, Los Angeles,
California 90095
Jen-Lien Lin
Union Chemical Laboratories, Industrial Technology Research Institute (ITRI), Hsinchu 300, Taiwan
(Received 1 April 2004; accepted 23 August 2004)
High-performance organic thin-film transistors (OTFTs) with a nanoparticle composite dielectric
layer have been demonstrated. The dielectric layer consists of cross-linked poly-4-vinylphenol
(PVP) and high-dielectric titanium dioxide sTiO2d nanoparticles. Because of the nanosize of TiO2,
it disperses well in the organic solvent, which makes it possible to use solution-processable methods
to prepare the dielectric layer. OTFTs with pentacene as the semiconducting layers have been
demonstrated; it was found that the OTFTs with the nanocomposite dielectric layer have higher
field-induced current than that of conventional devices because the dielectric constant of the gate
insulator is increased. This finding opens an interesting direction for the preparation of
high-performance OTFTs without complicated sputtering of high-k dielectric materials. © 2004
American Institute of Physics. [DOI: 10.1063/1.1806283]
There has been great interest in thin-film transistors
made of organic compounds, since organic thin-film transistors (OTFTs) have many unique advantages, such as light
weight, flexibility, low cost of fabrication, and solution
processability.1–5 However, traditional OTFTs often suffer
from high operating voltage due to the low charge carrier
mobilities of organic semiconductors. Hence, for the applications that require high current output, such as switching of
organic light-emitting diodes, OTFTs are still not the suitable
candidates.6 Since the field-induced current is proportional to
the field-induced charge density and carrier mobility, one
way to overcome this problem is to use high-dielectricconstant (high-k) gate insulators,7,8 which can enhance the
field-induced carrier density. However, most high-k materials used for OTFTs so far are based on ceramics and hence
are usually brittle and expensive to prepare. The poor mechanical properties of these materials makes it highly challenging to be realistic in the flexible electronics. In addition,
the preparation of these high-k materials requires a hightemperature annealing process, which is not compatible with
plastic substrates. Consequently, it is necessary to develop a
cheap and easy way (e.g., a solution-processable method) to
fabricate gate insulators with both a high dielectric constant
and mechanical flexibility.
In this work, nanocomposite dielectric layers, consisting
of cross-linked poly-4-vinylphenol (PVP) and titanium dioxide sTiO2d nanoparticles, a high-k material sk = 80d,9 were
prepared for OTFT gate insulators. Because of the nanosize
of TiO2, it disperses well in the organic solvent, which
makes it possible to use solution-processable methods, such
as spin coating, to prepare the devices. With pentacene as the
semiconducting layer, it was found that the OTFTs with the
nanocomposite dielectric layers have a higher field-induced
current than that of convention devices because the dielectric
constant of the gate insulator is increased.
Pentacene, poly-4-vinylphenol (PVP, M w = 20 000),
poly(melamine-co-formaldehyde) methylated sM n = 511d and
a)
Electronic mail: yangy@ucla.edu
propylene glycol monomethyl ether acetate (PGMEA) were
purchased from Aldrich, and used as received without purification. TiO2 nanoparticles were obtained from Nanophase
Technologies. Figure 1 shows the device structure for a typical OTFT fabricated in this study. The channel length sLd and
width sWd are 100 and 6000 mm, respectively. ITO patterned
on a glass substrate was used as the gate electrode. More
than ten OTFTs were made for each experiment; reproducible data were obtained, and typical device characteristics are
shown in this work. We found that an additional layer of
30 nm 3,4-polyethylenedioxythiophene-polystyrenesulfonate
(PEDOT), on the ITO could enhance the OTFT on-off ratio
from 103 to 104, presumably due to a decrease in surface
roughness of the ITO surface by the PEDOT layer. PVP
s11 wt.%d and poly(melamine-co-formaldehyde) methylated
s4 wt.%d was dissolved in PGMEA, and blended with different concentrations of TiO2 nanoparticles. After stirring the
solution for more than 24 h, a stable TiO2 dispersed solution
was obtained. No apparent precipitation was found even after
24 h in the solutions for all the concentrations of TiO2 used
in this study. The solution could also easily go through
1.0 mm filters. The solution was then spin coated onto the
PEDOT/ITO substrates. The substrate was then prebaked at
120° C for five min followed by baking at 200° C for 20 min
to cross-link the polymer. Pentacene was then thermally deposited as the semiconducting layer. The gold source and
drain electrodes were then thermally evaporated through a
shadow mask. The film thicknesses were measured by Dektak 3030 profilometer. The current–voltage sI – Vd character-
FIG. 1. The device structure of the OTFTs.
0003-6951/2004/85(15)/3295/3/$22.00
3295
© 2004 American Institute of Physics
Downloaded 19 Oct 2004 to 164.67.193.59. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
3296
Chen et al.
Appl. Phys. Lett., Vol. 85, No. 15, 11 October 2004
FIG. 4. Plots of IDS and sIDSd1/2 versus VG at constant VDS = −40 V for
device A (dashed line) and device B (solid line). The condition is the same
as that in Fig. 3.
FIG. 2. Dielectric constant vs frequency plots of different dielectric insulators used for OTFTs. The dielectric constants increase with the amount of
TiO2 nanoparticles blended into the dielectric films.
istics of OTFTs were measured by HP 4155 semiconductor
parameter analyzer. The devices for capacitance measurements consist of different dielectric layers sandwiched between ITO and Al electrodes. The capacitance measurements
were conducted with a HP 4284A Precision LCR meter. The
dielectric constants were calculated following the equation
C = k«0A/t,
where C is the measured capacitance, «0 is the permittivity of
free space, A is the area of the capacitor, and t is the thickness of the dielectric.
Figure 2 shows the frequency-dependent dielectric constants of dielectric layers with different nanoparticle concentrations. For a pure cross-linked PVP film, the dielectric constant is 3.5 at 1 MHz, which is close to the value reported
earlier.10 After inserting a 30 nm PEDOT film, the dielectric
constant increases to 3.9. In lower frequency regions, the
dielectric constant increases rapidly, which may be due in
part to the free ions in the PEDOT layer. From Fig. 2, we can
see that the dielectric constant increases with the amount of
TiO2 nanoparticles embedded in the thin films. For the di-
electric film with 7 wt.% of TiO2 nanoparticles, the dielectric
constant increased to 5.4. From the data shown previously, it
can be concluded that the cross-linked PVP films blended
with high-k nanoparticles have higher dielectric constants
than a pure polymer film. The dielectric constant can be further increased if nanoparticles with an even higher dielectric
constant are used.
Figure 3(a) shows the drain–source current sIDSd versus
drain–source voltage sVDSd of a typical OTFT with a crosslinked PVP gate insulator (device A) at different gate voltages sVGd. The corresponding plot of sIDSd1/2 versus VGS for
the device is shown in Fig. 4 (dashed line). The carrier mobility was calculated at the saturation regions with the following equation:
IDS = sWCi/2LdmsVG − VTd2 ,
s1d
where Ci is the capacitance per unit area of the insulator, and
VT is the threshold voltage. The mobility and the threshold
voltage of the OTFT were 0.20 cm2 V−1 s−1 and −7.8 V, respectively. The on-off ratio was more than 104. On the other
hand, the device with 7 wt.% TiO2 nanoparticles PVP gate
insulator (device B) had almost twice the field-induced current at the same gate voltage as shown in Fig. 3(b). The
calculated mobility from Fig. 4 was 0.24 cm2 V−1 s−1 and
the threshold voltage decreased slightly to −7.0 V. From Eq.
(1), since the mobility is almost unchanged for both devices,
the enhanced output current should be attributed to the
higher field-induced charge carrier density in the conducting
channel. Consequently, Fig. 3 clearly demonstrates that the
OTFT with a nanocomposite dielectric layer gives a higher
current output due to a gate insulator embedded with high
dielectric constant material. Additionally, this has also been
TABLE I. Electrical parameters of the OTFTs in this study
Conc. of TiO2
nanoparticles
swt.%d
0 (w/o PEDOT)
0
3
5
7
Thickness Dielectric
Mobility
(nm)
constant scm2 V−1 s−1d
600
630
700
700
700
3.5
3.9
4.4
5.1
5.4
0.23
0.20
0.25
0.24
0.24
Threshold
voltage On-off
(V)
ratio
−7.5
−7.8
−7.5
−7.2
−7.0
103
104
103
103
103
FIG. 3. The current–voltage characteristics of OTFTs with (a) 670 nm
cross-linked PVP or (b) 700 nm nanocomposite s7 wt.% TiO2d dielectric
gate insulators. The thickness of pentacence for all devices in this study is
40 nm.
Downloaded 19 Oct 2004 to 164.67.193.59. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
Chen et al.
Appl. Phys. Lett., Vol. 85, No. 15, 11 October 2004
3297
FIG. 5. The AFM height-mode images for a neat cross-linked PVP film (left panel) and a cross-linked PVP film blended with 7 wt.% TiO2 nanoparticles (right
panel). The images were taken by a Digital Instruments NanoScope III with a 125 mm etched Si cantilever in tapping mode.
confirmed from the increase of output current due to the
increasing concentrations of TiO2 nanoparticles embedded
in the gate insulators. The electrical characteristics of the
OTFTs with different amounts of TiO2 nanoparticles embedded in the gate insulators are summarized in Table I.
On the other hand, the on-off ratio of device B decreases
to 103, about one order smaller than that of device A (Fig. 4).
This is probably due to higher leakage current from the
nanocomposite film. From the atomic force microscopy
(AFM) images of the surfaces of the dielectric films (Fig. 5),
it can be seen that the surface of the dielectric layer of device
B is rough; in contrast to a smooth surface obtained in device
A. The rms roughness of the dielectric film with 7 wt.%
TiO2 nanoparticles is 12.8 nm, which is much rougher than
that of the pure crossed-linked PVP film s0.3 nmd. This suggests that there are probably some defects present in the gate
insulator, which may cause the current leakage. This problem
is likely to be resolved in the future by engineering the nanoparticle with side groups soluble in organic solvent. The
morphology of the polymer thin film will be smoothened,
subsequently reducing the leakage current.
In summary, we have shown that the field-induced current of an OTFT can be enhanced by blending high-k nanoparticles into the dielectric gate insulator. The nanocomposite
can also be used to tune the dielectric constant of the gate
dielectric layer. This method offers an easy and cheap way to
prepare gate insulators of OTFTs with higher dielectric constants. Although enhanced current output was observed for
the OTFTs with the nanocomposite dielectric gate insulator,
apparently, the device performance is still limited by the
solubility of TiO2 nanoparticles in solutions. The weight ratio of TiO2 to cross-linked polymer in the composite layer is
less than 0.5. With a higher amount of TiO2 nanoparticles in
the composite film, a higher dielectric constant and, subsequently, a higher current output are expected. In the future,
the current leakage problem may be overcome with a better
film morphology of the dielectric layer. In addition, the surface of the nanoparticles should be modified to enhance the
solubility in common organic solvent.
The authors are indebted to the financial support from
the Air Force Office of Scientific Research (F49620-01-10427, program manager Dr. Charles Lee) and the Office of
Naval of Research (N00014-01-1-0136, program manager
Dr. Paul Armistead).
1
L. Torsi, A. Dodabalapur, L. J. Rothberg, A. W. P. Fung, and H. E. Katz,
Science 272, 1462 (1996).
2
Y. Y. Lin, D. J. Gundlach, S. F. Nelson, and T. N. Jackson, IEEE Trans.
Electron Devices 44, 320 (1997).
3
H. Sirringhaus, N. Tessler, and R. H. Friend, Science 280, 1741 (1998).
4
C. D. Sheraw, L. Zhou, J. R. Haung, D. J. Gundlach, T. N. Jackson, M. G.
Kane, I. G. Hill, M. S. Hammond, J. Campi, B. K. Greening, J. Francl, and
J. West, Appl. Phys. Lett. 80, 1088 (2002).
5
M. Shtein, J. Mapei, J. B. Benziger, and S. R. Forrest, Appl. Phys. Lett.
81, 268 (2002).
6
F. Garnier, R. Hajlaoui, and M. E. Kassmi, Appl. Phys. Lett. 73, 1721
(1998).
7
C. D. Dimitrakopoulos, S. Purushothaman, J. Kymissis, A. Callegari, and
J. M. Shaw, Science 283, 822 (1999).
8
G. Velu, C. Legrand, O. Tharaud, A. Chapoton, D. Remiens, and G.
Horowitz, Appl. Phys. Lett. 79, 659 (2001).
9
G. D. Wilk, R. M. Wallace, and J. M. Anthony, J. Appl. Phys. 89, 5243
(2001).
10
M. Halik, H. Klauk, U. Zschieschang, G. Schmid, W. Radlik, and W.
Weber, Adv. Mater. (Weinheim, Ger.) 14, 1717 (2002).
Downloaded 19 Oct 2004 to 164.67.193.59. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp