enhancement of long-term stability of pentacene thin-film transistors encapsulated with transparent...
TRANSCRIPT
www.elsevier.com/locate/apsusc
Applied Surface Science 252 (2005) 1332–1338
Enhancement of long-term stability of pentacene thin-film
transistors encapsulated with transparent SnO2
Woo Jin Kim a, Won Hoe Koo a, Sung Jin Jo a, Chang Su Kim a,Hong Koo Baik a,*, Jiyoul Lee b, Seongil Im b
aDepartment of Metallurgical Engineering, Yonsei University, Seoul 120-749, Republic of Koreab Institute of Physics and Applied Physics, Yonsei University, Seoul 120-749, Republic of Korea
Received 18 January 2005; received in revised form 14 February 2005; accepted 16 February 2005
Available online 11 April 2005
Abstract
The long-term stability of pentacene thin-film transistors (TFTs) encapsulated with a transparent SnO2 thin-film prepared by
ion beam-assisted deposition (IBAD) was investigated. After encapsulation process, our organic thin-film transistors (OTFTs)
showed somewhat degraded field-effect mobility of 0.5 cm2/(V s) that was initially 0.62 cm2/(V s), when a buffer layer of
thermally evaporated 100 nm SnO2 film had been deposited prior to IBAD process. However, the mobility was surprisingly
sustained up to 1 month and then gradually degraded down to 0.35 cm2/(V s) which was still three times higher than that of the
OTFT without any encapsulation layer after 100 days in air ambient. The encapsulated OTFTs also exhibited superior on/off
current ratio of over 105 to that of the unprotected devices (�104) which was reduced from �106 before aging. Therefore, the
enhanced long-term stability of our encapsulated OTFTs should be attributed to well protection of permeation of H2O and O2
into the devices by the IBAD SnO2 thin-film which could be used as an effective inorganic gas barrier for transparent organic
electronic devices.
# 2005 Elsevier B.V. All rights reserved.
Keywords: OTFT; Ion beam-assisted deposition; Encapsulation; Lifetime; Pentacene
1. Introduction
Organic thin-film transistors (OTFTs) have drawn
much attention due to their great compatibility with
plastic substrates available for a flexible display, smart
* Corresponding author. Tel.: +82 2 2123 2838;
fax: +82 2 312 5375.
E-mail address: [email protected] (H.K. Baik).
0169-4332/$ – see front matter # 2005 Elsevier B.V. All rights reserved
doi:10.1016/j.apsusc.2005.02.100
cards and rf tags application over the last decades
[1–4]. Organic semiconductors also make it possible
to fabricate electronic devices at room temperature
through simple methods without using complicated
equipments [5]. Among organic semiconductors
considered as active materials in organic thin-film
transistors, pentacene was widely used for its mobility
as high as that of a-Si or in some cases even exceeding
that value [6,7]. However, when the OTFTs are
.
W.J. Kim et al. / Applied Surface Science 252 (2005) 1332–1338 1333
Fig. 1. (a) Schematic drawing of IBAD process and (b) photo-
graphic plan view (length, L = 100 mm and width, W = 1000 mm)
and cross-section of the device structure.
operated in air ambient, they tend to easily degrade in
the similar way that the lifetime of organic light-
emitting diodes (OLEDs) is limited [8]. In order to
enhance the stability of devices, some encapsulation
layers capable of protecting H2O or O2 in the air from
penetrating through the organic layers have been
employed. SiOx, AlOx, AlOxNy and SiOxNy are
common amorphous oxides used as inorganic gas
barrier [9–11]. But, during the deposition process of
the above-mentioned oxides, some serious damages
might be induced on the organic layers, which would
be attributed to energetic ions, X-rays and electron
beam [12,13]. Encapsulation with an organic material,
such as a vacuum deposited parylene was reported to
have an excellent conformal coating and successfully
protect the organic active layer from solvents in
patterning process [14]. For the sake of more reliable
operation of the devices, it is quite desirable for an
organic layer to be used along with an inorganic oxide,
because an organic material itself could not provide
sufficient capability of blocking permeation of H2O or
O2. Among the transparent amorphous oxides, which
can perform function of gas barrier deposited by
simple thermal evaporation, SiOx and SnO2 would be
taken into account. It was previously reported by our
group that SnO2 was more desirable one due to its high
polarizability resulting in effectively suppressing
permeation of water vapor by strong chemical
interaction, particularly when the film could be made
to have much denser microstructures [15]. Therefore,
the SnO2 thin-film prepared by ion beam-assisted
deposition (IBAD) is a promising candidate for
transparent thin-film encapsulation adaptable for
top-emitting organic light-emitting diodes (TE-
OLEDs) since it has relatively high transmittance
(�85% over 450 nm) as well as quite low water vapor
transmission rate (WVTR) of below 10�3 g/(m2 day).
In order to prevent possible damages induced from
energetic ions during IBAD process, it is also desir-
able that a proper buffer layer should be employed
[16].
In this paper, we report on the investigation of the
long-term stability of our OTFTs in terms of electrical
properties associated with materials characterization
by X-ray photoemission spectroscopy (XPS) when the
TFT was encapsulated with the IBAD SnO2 thin-film
having a buffer layer of same kind of material on top of
the device.
2. Experiment
Our OTFTs with the pentacene thin-film as an
active layer were fabricated on SiO2/p+-Si substrate.
The thickness of thermally grown SiO2 was 100 nm.
Prior to the deposition of pentacene, the surface of
SiO2 (to be used as a gate dielectric) was ultrasonically
cleaned with trichloroethylene, acetone, isopropylal-
cohol and deionized water sequentially. Our thermal
evaporation chamber was evacuated to a base pressure
of 1 � 10�7 Torr. We then deposited the pentacene
film through a shadow mask by heating the effusion
cell containing pure pentacene powder (99.8%,
Aldrich Chem. Co.) at a rate of 0.5 A/s at room
temperature (RT). The resulting film thickness of the
pentacene film was adjusted to 50 nm as measured
with a quartz crystal monitor. The source and drain
contacts were patterned by a second shadow mask,
through which Au was thermally evaporated to the
thickness of about 100 nm. The channel length,
defined as the spacing between the source and drain
pads, was 100 mm and the channel width was
1000 mm. SnO2 (to be used as a buffer layer) was
W.J. Kim et al. / Applied Surface Science 252 (2005) 1332–13381334
then deposited on top of the device by thermal
evaporation through a third shadow mask to the film
thickness of about 100 nm. Finally, another 80 nm
SnO2 thin-film (to be used as an encapsulation layer)
was consecutively deposited without breaking vacuum
by IBAD process shown in Fig. 1(a). The end-hall ion
gun was used to increase SnO2 ad-atom mobility. Ar+
ion energy and current density were 150 eV and
45 mA/cm2, and the deposition rate was 1 nm/s. In all
cases of ion beam irradiation conditions, the incident
angle of the ion beam was 458. Fig. 1(b) shows
photographic plan view and schematic cross sectional
view of our pentacene OTFT with encapsulation.
Fig. 2. (a) The output characteristics of the reference OTFTwithout
any encapsulation and (b) degradation of the drain current in the
saturation regime when VG is �40 V after 100 days in air.
3. Results
The I–V measurements of our OTFTs were
performed with an HP 4145B parameter analyzer in
the dark at RT. Fig. 2(a) displays the output
characteristics of our OTFT without encapsulation
and one can see the typical drain current–voltage (ID–
VD) curves of the FET devices. The drain current in the
saturation regime at a gate bias of �40 V was around
�120 mA and the leakage current measured between
the gate and source electrode (IG) was below 2 nA. It is
then clear that the measured drain current could be
used for evaluation of the electrical performance of
our OTFTs. The curves in Fig. 2(b) are representing
the drain current of the OTFT without encapsulation
(open circle) and with encapsulation (close circle)
when the gate bias of VG is�40 V. According to Fig. 2
(b), the saturation current of the unprotected OTFT is
decreased to �35 mA from 120 mA after 100 days in
air ambient, whereas the encapsulated device showing
a relatively low initial current of�100 mAmaintained
considerable current level of �60 mA. The initial
degradation of the OTFTwith encapsulation might be
attributed to some possible damages induced on the
pentacene active layer during the encapsulation
process in which small portion of the energetic ions
could penetrate through the thermally evaporated
buffer layer [16].
The plots offfiffiffiffiffiffiffiffiffi
�IDp
versus VG and log10 (� ID)
versus VG with and without encapsulation are
presented in Fig. 3(b and c), respectively. The field-
effect mobility of holes in the pentacene layer was
obtained from the slope in VG versusffiffiffiffiffiffiffiffiffi
�IDp
plot in the
saturation regime where the source–drain voltage of
VD = �40 V was applied. The initial field-effect
mobility of the encapsulated and unprotected OTFT
was calculated to be 0.5 and 0.62 cm2/(V s), respec-
tively. As expected from serious degradation of
saturation current of the OTFT without encapsulation
in Fig. 2(b), the mobility of the device decreased
sharply from 0.62 to 0.10 cm2/(V s) after 100 days.
On/off current ratio and sub-threshold slope also
showed a similar degradation trend according to the
log10 (� ID) versus VG plots. The on/off current ratio
decreased down to �104 from initial value of �106
with a very inferior sub-threshold slope of 7.2 V/dec.
In the mean while, the encapsulated device maintained
an on/off current ratio of �105 even after 100 days
with similar sub-threshold slope to the initial value. It
is also interesting to note in theffiffiffiffiffiffiffiffiffi
�IDp
versus VG plot
that the threshold voltage (VT) changes toward positive
W.J. Kim et al. / Applied Surface Science 252 (2005) 1332–1338 1335
gate bias showing early turn-on upon degree of
degradation. Although the initial VT values of both
devices were almost same as approx. �2.5 V, the
unprotected device displayed about 7 V as shown in
Fig. 3(a), while the same-period-aged but encapsulated
one showed only �0.1 V of VT after 100 days in air.
Since the pentacene channel/SiO2 interfaces should
be in the same condition for all OTFTs, we suspected
if water molecules exist in the channel layer. Horowitz
demonstrates that the surface of gate dielectrics in
contact with organic channel layer has so many
defects that it would strongly aggravate the charge
transport between the source and drain electrode [17].
Fig. 3. The plots offfiffiffiffiffiffiffiffiffi
�IDp
vs. VG and log10 (� ID) vs. VG. for the
OTFTs without encapsulation (a) and with encapsulation (b).
Hence, we should take into account the organic
molecular layer beyond the first monolayer in the
channel region to find out what would be responsible
for the total charge transport. The water molecules in
air are known to be adsorbed on pentacene surface or
even penetrate into the channel layer during aging
period [18]. Then, they could be one of the main
leakage sources or active deep level traps when the
OTFT is particularly under a high electric field of
depletion state [19].
The property-endurance limit or aging effect of our
OTFTs in terms of field-effect mobility and on/off
current ratio upon time are shown in Fig. 4(a and b),
respectively. The encapsulated TFT shows the best
performance even after a long time exposure to air.
The initial degradation of mobility from 0.62 to
0.5 cm2/(V s) is the expected behavior in line with
Fig. 4. The property-endurance limit or aging effect with time of our
OTFTs in terms of field-effectmobility (a) and on/off current ratio (b).
W.J. Kim et al. / Applied Surface Science 252 (2005) 1332–13381336
reduction of the drain current of the encapsulated
OTFT in Fig. 2(b). However, the SnO2 film was
deposited by using IBAD on top of the device without
any buffer layer, the device was found to seriously
degrade due to direct penetration of energetic ions into
the organic active layer [16]. In order to avoid
deterioration of the device during IBAD process, it is
strongly required to employ a proper buffer layer of
which is thermally evaporated SnO2 in this study.
Unfortunately, thermally evaporated SnO2 film had
quite open structure resulting in a loose microstructure
and this would also be known from the fact that, when
being used alone, it could not effectively suppress H2O
permeation into the device as indicated in Fig. 4(a).
Therefore, the buffer layer itself could not play an
important role in blocking permeation of water vapor
and it is the IBAD SnO2 layer that performs this
function.
4. Discussion
It was once reported that the main reason for
degradation of the OTFTs with time was adsorption
and penetration of H2O into the active layer of a
pentacene film [18]. Fig. 5 represents O 1s peaks that
we obtained from the both surfaces of aged and un-
aged pentacene films as they were scanned by X-ray
photoemission spectroscopy. All the peak energies
were calibrated with C 1s peak. Since OH molecular
groups in pentacene are thought to be involved during
Fig. 5. XPS spectra for O 1s peaks obtained from the pentacene
thin-films before and after exposed to air for 100 days.
pentacene film deposition, the O 1s peaks from the OH
groups in aged and un-aged pentacene should be the
same each other in intensity. It is then clear that the
100-day-aged pentacene film has much higher density
of H2O than the un-aged film. Total amounts of H2O
and OH group inside of the pentacene film could be
estimated from corresponding O 1s peaks and
calculated to be �1.7 and �5% for the un-aged and
aged one, respectively. It is thus, considered that the
water molecules are crucial clues for the observed
property-degradation.
With application of encapsulation to the OTFTs,
the long-term stability in terms of the electrical
properties was tremendously enhanced. This should be
attributed to well protection of the devices from water
vapor or oxygen in air due to our IBAD SnO2 film.
Fig. 6 displays XPS O 1s spectrum of the IBAD SnO2
film comprised of two main peaks which were
originated from the OH groups at around 531.6 eV
and the O 1s peak corresponding to the Sn4+ at around
530.3 V while the inset in Fig. 6 represents the peak
from Sn 3d3/2 (495.3 eV) to 3d5/2 (486.9 eV). It was
clearly shown that the SnO2 film had O 1s peak from
OH group inside of the film, which is a result of higher
polarization and refractive index of the SnO2 film [20].
The H2O has a large dipole moment and lone-pair
electrons, thus is a good donor. Molecular adsorption
occurs by acid/base interaction between water vapor
and the oxides [21]. Therefore, the continuous
transmission of water vapor is restrained by the
interaction between the formed OH groups and water
vapor. However, although the thermally evaporated
SnO2 film has the high refractive index of about 1.9
and considerable amount of OH groups inside of film,
it is obvious that this layer could not prevent
permeation of H2O effectively as shown in
Fig. 4(a). Here, we should consider the influence of
the film density on the water vapor permeation,
because if the pore size and density are too large and
high, the permeating water vapor would be expected to
penetrate the films without interaction through the
central pore space away from the surface OH groups
within pores [15]. Consequently, when encapsulated
with the IBAD SnO2, the device displays much
enhanced performance in terms of long-term stability
due to its quite low WVTR, which should come from
strong chemical interaction between penetrating polar
water vapor and hydroxyl elements inside of the film
W.J. Kim et al. / Applied Surface Science 252 (2005) 1332–1338 1337
Fig. 6. XPS spectra for O 1s peaks and Sn 3d2/3 and 3d2/5 spectra (inset) from the SnO2 thin-film prepared by IBAD.
in conjunction with a dense microstructure by IBAD
process.
However, even the encapsulated devices finally
began to degrade its field-effect mobility with time
after 1 month. This might be attributed to our
pentacene films with a very rough surface of rms
9.3 nm as measured by atomic force microscope
(AFM) in the channel region shown in Fig. 7(a). Since
the IBAD SnO2 thin-film with a buffer layer was
deposited on the rough pentacene surface, the
encapsulation layer was also found to be rough
following underlying layer. Fig. 7(b) shows the top
surface of the encapsulation layer prepared by IBAD
Fig. 7. AFM images of 50 nm pentacene thin-film (a) and 80 nm IBAD
and according to a section analysis of AFM profile, the
vertical distance between marks was found to reach as
high as 50 nm where the thickness of the IBAD SnO2
layer was 80 nm. Therefore, in addition to a dense
microstructure, the barrier film should have a flat
surface because vapor permeation should be mainly
dominated by diffusion via the shortest path in the
film. For this reason, a thin-film encapsulation
consisted of multi-layer was more desirable to
enhance the barrier properties and some polymer
smoothing layers could be used to make a surface flat
on which the inorganic barrier film will be deposited.
Hence, a further study would be required to make our
SnO2 encapsulation layer on top of 100 nm SnO2 buffer layer (b).
W.J. Kim et al. / Applied Surface Science 252 (2005) 1332–13381338
encapsulation meet the requirements needed for stable
operation of a real device in air ambient.
5. Conclusion
We have fabricated pentacene TFTs with the
transparent SnO2 encapsulation layer prepared by
IBAD to enhance long-term stability of the devices.
Although the encapsulated OTFTs showed somewhat
degraded field-effect mobility of 0.5 cm2/(V s), the
devices exhibited much enhanced electrical perfor-
mances sustaining the mobility up to 1 month. On/off
current ratio also prevails in the case of the
encapsulated devices showing around 105 that is just
one order of magnitude lower than that of its initial
value of�106, while the unprotected OTFT shows on/
off current ratio of �104 after 100 days. The IBAD
SnO2 was found to effectively prevent H2O or O2 from
permeating into the devices by means of densification
of microstructures through IBAD process and che-
mical interaction between water vapor and SnO2
which resulted in quite low WVTR value accounting
for enhancement of the long-term stability of the
OTFTs. It is thus, concluded that a transparent SnO2
thin-film prepared by IBAD is a promising material
that could be used as an inorganic gas barrier for
transparent organic electronic devices with an appro-
priate buffer layer including some smoothing layers.
References
[1] A.R. Brown, A. Pomp, C.M. Hart, D.M. de Leeuw, Science 270
(1995) 972.
[2] A.R. Brown, C.P. Jarret, D.M. de Leeuw, M. Matters, Synth.
Met. 88 (1997) 37.
[3] P.F. Baude, D.A. Ender, M.A. Haase, T.W. Kelly, D.V. Muyres,
S.D. Theiss, Appl. Phys. Lett. 82 (2003) 3964.
[4] C.D. Sheraw, L. Zhou, J.R. Huang, D.J. Gundlach, T.N.
Jackson, M.G. Kane, I.G. Hill, M.S. Hammond, J. Campi,
B.K. Greening, J. Francl, J. West, Appl. Phys. Lett. 80 (2002)
1088.
[5] F. Garnier, R. Hajlaoui, A. Yassar, P. Strivastava, Science 265
(1994) 1684.
[6] P.V. Necliudov, M.S. Shur, D.J. Gundlach, T.N. Jackson, J.
Appl. Phys. 88 (2000) 6594.
[7] H. Klauk, D.J. Gundlach, J.A. Nichols, T.N. Jackson, IEEE
Trans. Electron. Devices 46 (1999) 1258.
[8] H. Aziz, Z. Popovic, C. Tripp, N.-X. Hu, A.-M. Hor, G. Xu,
Appl. Phys. Lett. 72 (1998) 2642.
[9] M. Vogt, R. Hauptmann, Surf. Coat. Technol. 676 (1995) 74.
[10] A.G. Erlat, B.M. Henry, J.J. Ingram, D.B. Mountain, A.
McGuigan, R.P. Howson, C.R.M. Grovenor, G.A.D. Briggs,
Y. Tsukahara, Thin Solid Films 388 (2001) 78.
[11] A.G. Erlat, B.M. Henry, J.J. Ingram, C.R.M. Grovenor, G.A.D.
Briggs, R.J. Chater, Y. Tsukahara, J. Phys. Chem. B 883 (2004)
108.
[12] V. Bulovic, P. Tian, P. Burrows, M.R. Gokhale, S.R. Forres,
M.E. Thompson, Appl. Phys. Lett. 70 (1997) 2954.
[13] L.S. Hung, C.W. Tang, M.G. Manson, P. Raychudhuri, J.
Madathil, Appl. Phys. Lett. 74 (2001) 544.
[14] I. Kymissis, C.D. Dimitrakopoulos, S. Purushothaman, J. Vac.
Sci. Technol. B 20 (3) (2002) 956.
[15] W.H. Koo, S.M. Jeong, S.H. Choi, S.M. Lee, S.J. Lee, H.K.
Baik, J. Phys. Chem. B 108 (2004) 18884.
[16] S.M. Jeong,W.H. Koo, S.H. Choi, S.J. Jo, S.J. Lee, K.M. Song,
H.K. Baik, Appl. Phys. Lett. 85 (2004) 1051.
[17] G. Horowitz, J. Mater. Res. 19 (7) (2004) 1946.
[18] Y. Qiu, Y. Hu, G. Dong, L. Wang, J. Xie, Y. Ma, Appl. Phys.
Lett. 83 (2003) 1644.
[19] J. Lee, K.B. Kim, J.H. Kim, S.I. Im, D.Y. Jung, Appl. Phys.
Lett. 82 (2003) 4169.
[20] V. Dimitrov, T. Komastsu, J. Solid State Chem. 100 (2003) 163.
[21] V.E. Henrich, P.A. Cox, The Surface Science of Metal Oxides,
Cambridge University Press, New York, 1996.