A novel transparent air-stable printable n-type semiconductor technology using ZnO nanoparticles

Abstract

We report on a novel, air-stable, printable, transparent, NMOS semiconductor technology using soluble ZnO nanoparticles. We demonstrate solution-processed transistors with mobility > 0.1 cm²/V·s, which is the highest solution-processed NMOS mobility reported to date. The air-stability and transparency make this device an ideal candidate for low-cost printed displays and CMOS circuitry.

Full text

Page 1

A novel transparent air-stable printable n-type semiconductor
technology using ZnO nanoparticles

Steven K. Volkman, Brian A. Mattis, Steven E. Molesa, Josephine B. Lee, Alejandro de
la Fuente Vornbrock, Teymur Bakhishev, and Vivek Subramanian

Department of Electrical Engineering and Computer Sciences, University of California, Berkeley
Berkeley, CA 94720-1770, USA

Abstract

We report on a novel, air-stable, printable, transparent,
NMOS semiconductor technology using soluble ZnO
nanoparticles. We demonstrate solution-processed transistors
with mobility > 0.1cm2/V·s, which is the highest solution-
processed NMOS mobility reported to date. The air-stability
and transparency make this device an ideal candidate for low-
cost printed displays and CMOS circuitry.

Introduction

There is great interest in printing for realizing low-cost
electronics. Based on various reported cost models, printed
electronics is expected to be two to three orders of magnitude
cheaper per unit area than conventional semiconductor
manufacturing flows, albeit at a higher cost per transistor (1).
Therefore, for area-constrained applications such as displays
and low-frequency RFID tags, printed electronics has
garnered substantial interest. Most printed transistors to date
make use of organic semiconductors with mobilities between
0.01 and 1 cm2/V·s, which is adequate for some displays, and
approaching the realm of performance required for RFID tags
(2). Several deficiencies remain, however. Most printable
semiconductors today are p-type; available n-type
semiconductors have mobilities <10-2 cm2/V·s, which is
generally considered to be too low for most printed
electronics applications. This prevents the use of low-power
CMOS circuits and also increases transistor count and circuit
complexity. Most printable semiconductors also have poor
air-stability, complicating packaging and degraded device
lifetime. Additionally, many printable semiconductors,
including organics and chalcogenides are toxic, and their
integration into disposable circuits is problematic. Finally,
for display applications, none of these materials is
transparent. This is problematic, since, due to the low
mobility of printed devices, it is generally desirable to use
multi-transistor pixels or wide pixel transistors to maximize
display brightness and contrast. Since all printed devices to
date are opaque or semi-transparent at best, pixel aperture
ratio is sacrificed when using any of these techniques. This
in turn degrades display brightness and energy efficiency.

The need therefore exists for a printable NMOS material
offering improved mobility, air-stability, low toxicity, and
optical transparency. For the first time, we report on such a
material. We have developed a novel printable
semiconductor using 3nm zinc oxide nanoparticles. We take
advantage of the reduction in sintering temperature observed
for nanoparticles in order to produce thin films at reduced
temperatures (3). These particles have a 105ºC melting point
and are soluble and printable. By annealing solution-
processed films at plastic-compatible
semiconducting ZnO (a comparatively benign material
routinely used in antifungal ointments) films are formed.
NMOS TFTs fabricated using these films, which are optically
transparent, have mobilities > 0.1cm2/V·s. Due to their
transparency, they may be sized without brightness tradeoffs
in flexible display applications. Because the material is an
oxide and is therefore unreactive in air, ambient exposure
over long periods of time has no effect on performance; this
is in great contrast to most other printable semiconductors.
This therefore represents a major step towards the realization
of printed CMOS integrated circuits for low-cost electronics.

Experimental Details

ZnO nanoparticles were synthesized by reacting zinc acetate
with NaOH in 2-propanol. After 15 minutes, dodecanethiol
encapsulant is added. After 2 hours, the resulting alkanethiol-
encapsulated ZnO nanoparticles are collected and purified.
The resulting particles consist of ~3nm ZnO crystals
surrounded by a monolayer of dodecanethiol encapsulant.
The size of the particle is determined by the time of addition
of the encapsulant and the relative concentrations of the
encapsulant and metallic precursor. The encapsulant serves
to ensure the particle growth is self-limited to 3nm, prevents
subsequent particle agglomeration,
solubilization of the particles in numerous common organic
solvents. The process for particle synthesis is shown
schematically below (Fig. 1)
temperatures,
and allows the
Zn
O
O
OO
+
NaOH
1-dodecanethiol
ZnO nanoparticle
Fig. 1: Scheme used for synthesis of ZnO nanoparticles.

The diameter of the resulting particle is verified to be
approximately 3nm by transmission electron micrograph
(Fig. 2).
0-7803-8684-1/04/$20.00 ©2004 IEEE

Page 2

Fig. 2: TEM of zinc-oxide nanoparticles, showing tight size distribution. The
particles have an average diameter of 3nm, and are encapsulated with 1-
dodecanethiol.

Heating experiments were performed to verify that the
encapsulant evaporates and the particles sinter at ~105ºC in
air; this material is therefore potentially plastic-compatible,
assuming the remainder of the process is also optimized for
use on plastic. Electrical performance was evaluated by
fabricated bottom-gated transistors with N+ silicon gates,
100nm thermal SiO2 gate dielectrics, and evaporated gold
S/D pads. The ZnO particles were spun-cast in chloroform
and annealed at 150ºC. A conventional 400ºC forming-gas
anneal was performed to passivate dangling bonds; this step
may be replaced with plasma hydrogenation in a plastic
compatible process. The final thickness of the zinc oxide is
40nm, while the initial thickness was 80nm (Fig. 3), as
measured by profilometry. This thickness reduction is
evidence of the sublimation of the encapsulant and the
sintering of the particles to form a film. Physically, the film
also undergoes substantial changes during this sintering
process, going from a powdery film that is easily washed off
in solvents to a brittle, glassy-film that is impervious to
solvent treatments.
0
0.02
0.04
0.06
0.08
0.1
0.12
100 120140160180 200220240 260280 300
Scan Distance (µ µm)
Height (µ µm)
Pre-anneal
Post-anneal
Fig. 3: Profilometry scans showing thickness reduction after annealing; this
thickness reduction is caused by the encapsulant removal and particle
sintering processes.

Results

Zinc oxide is an n-type semiconductor with poor inversion
behavior. Therefore, the transistors are operated as
accumulation-mode NMOS devices, and show excellent
electrical characteristics (Figs 4, 5).
0.0E+00
1.0E-07
2.0E-07
3.0E-07
4.0E-07
5.0E-07
6.0E-07
7.0E-07
0510 1520 2530
VD (V)
ID (A)

Fig. 4: Output characteristics of a typical device (W/L=20/10 µm), showing
excellent characteristics. The offset in the zero-intercept is due to gate
leakage and S/D barriers.
1.0E-12
1.0E-11
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
-10 -50510 15202530
VG (V)
ID (A)
-1.E-08
0.E+00
1.E-08
2.E-08
3.E-08
4.E-08
5.E-08
6.E-08
7.E-08
8.E-08

Fig. 5: Transfer Characteristics of a typical device (W/L=20/10 µm). The
lack of saturation in the mobility is due to the large contact resistances in the
device.

The on-off ratio is >103 and the field-effect mobility is
typically in the range of 0.1- 0.2 cm2/V-s. These are the
highest mobilities ever reported from a solution-processed
NMOSFET. This high performance is achieved despite the
large-barrier expected to exist (and apparent from the
electrical characteristics) between gold and ZnO. The large
bandgap of ZnO vs. the large workfunction of Au indicates
that a large barrier should exist at the S/D electrodes due to
the formation of ZnO/Au schottky junctions. Evidence for
this is seen in the convex turn-on characteristics in the low-
VDS portion of the ID-VD curves. This in turn dramatically
degrades performance. Use of appropriate contacting
materials should enhance performance even further.
0-7803-8684-1/04/$20.00 ©2004 IEEE

Page 3

The viability of accumulation-mode devices may be analyzed
using scaling. Since accumulation mode devices do not
provide a large barrier to current flow in the bulk of the
semiconducting film, they are prone to short channel effects.
The devices herein show some VT roll-off (Fig 6), but
generally show good characteristics down to 5µm (Fig 7, 8).
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
0123456789 10
Channel Length, L (µm)
VT (V)
Fig. 6: VT Roll-off characteristics. The long-channel VT is ~12V.

0.0E+00
5.0E-07
1.0E-06
1.5E-06
2.0E-06
2.5E-06
3.0E-06
3.5E-06
4.0E-06
05 1015202530
VD (V)
ID (A)
Fig. 7: Output characteristics for a 100µm/5µm device

1.0E-12
1.0E-11
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
-10-5051015 2025 30
VG (V)
ID (A)
Fig. 8: Transfer characteristics for a 100µm/5µm device

Only when the channel length was reduced to 3µm did we see
significant scaling related degradation (Figs 9, 10). These
reasonably good scaling characteristics of these devices are
likely due to two reasons. First, the use of a relatively thin
channel films (~50nm) generally suppresses sub-surface
leakage. Second, the relatively large source-side barrier
likely suppresses DIBL, albeit at the expense of drive current
and transconductance.
0.0E+00
5.0E-06
1.0E-05
1.5E-05
2.0E-05
2.5E-05
3.0E-05
3.5E-05
4.0E-05
4.5E-05
05101520 2530
VD (V)
ID (A)
Fig. 9: Transfer characteristics for a 100µm/3µm device

1.0E-12
1.0E-11
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
1.0E-04
-10-50510 15 20 2530
VG (V)
ID (A)
Fig. 10: Transfer characteristics for a 100µm/3µm device

Optical transparency of the zinc-oxide films was measured
and found to be >93% in all samples. No degradation in
performance was seen over a three month period of the
devices being exposed to air. It should also be noted that all
relevant device processing was performed in air, unlike most
other printed electronics technologies, which require the use
of inert ambients.

Discussion

The availability of a viable NMOS printable semiconductor is
very importance for printed electronics. Currently, virtually
all circuit implementations make use of PMOS only,
complicating design and increasing power consumption. The
few demonstrated CMOS implementations typically suffer
0-7803-8684-1/04/$20.00 ©2004 IEEE

Page 4

from substantial performance degradation due to the typically
weak NMOS transistor This is problematic, since most low-
cost electronics applications will be portable (such as low-
cost displays) or power starved (such as RFID tags). The
availability of a suitable NMOS material will allow the
realization of printed CMOS circuits, offering design
simplicity and reduced power consumption. The transparency
of zinc oxide also has an important benefit for displays.
Given the low mobility of printed semiconductors, printed
displays typically make use of pixel architectures involving
wide transistors and/or multiple transistor pixel schemes.
Such designs reduce the pixel emissive area, since the
transistors themselves block a portion of the light. By using a
transparent device, it is possible to achieve high currents
without sacrificing brightness. To date only printed
transparent conductors and
demonstrated; semiconductors have been a missing element.
The technology demonstrated here fills this gap. Since the
conduction band edge of ZnO is reasonably well-aligned with
ITO, it is expected that high performance transparent devices
should be realizable using this technology. This will enable
the demonstration of active matrix displays offering high-
brightness at low power while still maintaining low-
fabrication cost through printing. Furthermore the stability of
the compound in air when compared to many other printable
semiconductor materials makes it attractive for even general
applications. Unlike organic-based printed electronics and
also unlike previously reported nanoparticles-based printed
electronics systems (such as CdSe, etc.), the ZnO system is
inherently oxygen-compatible and air-stable. Therefore, all
processing may be performed in air and the need for highly
oxygen exclusive barrier and encapsulation layers is also
reduced. This should substantially simplify the overall
process while increasing device stability. Given that poor
device stability has historically been one of the major
concerns hindering the deployment of printed electronics, the
ZnO system is a particularly attractive candidate for printed
electronics in this regard. Further optimization of the
process, in particular the contacts and the annealing
methodology should enable the achievement of substantially
improved performance and the realization of a plastic-
compatible process flow.

dielectrics have been
Conclusion

We have demonstrated the highest mobility solution-
processed NMOS devices to date using soluble ZnO
nanoparticles. Mobilities as high as 0.2cm2/V·s have been
realized using this air-stable, non-toxic and transparent
material. This should enable the realization of all-printed
CMOS low-cost electronics and low-cost flexible displays.

References

(1) V. Subramanian, “Towards Printed Low-Cost RFID Tags: Device,
Materials and Circuit Technologies”, 2nd Advanced Technology
Workshop on Printing an Intelligent Future: Printed Organic and
Molecular Electronic Technologies, Boston, MA, March 16-19, 2003.
(2) C.D. Dimitrakopoulos and D. J. Mascaro, “Organic thin-film transistors:
A review of recent advances,” IBM Res. & Dev., vol. 45, No. 1, 2001.
(3) D. Huang, F. Liao, S. Molesa, D. Redinger, and V. Subramanian,
“Plastic-compatible low-resistance printable gold nanoparticle
conductors for flexible electronics”, Journal of the electrochemical
society, Vol. 150, pp. 412, 2003.
0-7803-8684-1/04/$20.00 ©2004 IEEE