3-1 The leakage of solution-processed metal-oxide TFTs
In our experiment, a 100-nm-thick layer of thermal silicon nitride (SiNx) was
grown onto silicon wafers to serve as the dielectric and the capacitance was
5.31×10-08 F/cm2. To avoid producing a large current, the metal-oxide films were
isolated by a wet etching process with dilute hydrochloric acid (HCl(aq): DIW=1:5). If
the active layer were not isolated, the leakage current path from drain electrode or
source electrode to gate electrode by the whole active layer result a large leakage
current and ruin the TFTs device. In the ideal condition, ID+IS+IG＝0. Decrease the
gate leakage would induce the origin of drain current or source current.
Fig. 3.1 showed the transfer characteristics of the zinc-oxide TFT at a drain-source
voltage VDS of 40V, the annealing temperate was 600℃, exhibiting a typical
n-channel enhancement field-effect behavior. It clear sees that the IG current in black
dash line is larger than the red dash line. And the IG current maintained at 10-9(A)
even in large gate voltage. The sub-threshold swing, S.S, decreased from
5.54(V/decade) to 0.64(V/decade), without wet etching and after wet etching,
respectively. The good performance for pattern TFTs avoided large leakage currents
and increased the TFTs device stability.
22
3-2 Bottom gate top contact zinc-oxide TFTs
Metal-oxide is a flexible material with easily tunable phases that can crystal to
amorphous or crystallization. Oxygen vacancy is an accepted electron donor in metal
-oxide film. Therefore, the oxygen quantity significantly influences the conductivity
or carrier concentration in metal oxide. As regard to the application on
semi-conductive metal-oxide, the oxygen ratio controlling seems critical and
important. In solution processed metal-oxide TFTs, There are some experiment
parameters to controlled oxygen ratio, such as adjusted the process atmosphere (N2,
O2), adjusted the process annealing temperature and adjusted the precursor solutions.
A set of experiment parameters could determine the final oxygen ratio or carrier
concentration in metal-oxide films. In our study, in order to get the best performance
of TFTs, we tested above methods and process conventional TFTs. This study was
expected to provide a useful guideline to do other device structure.
3-2.1 Zinc-oxide TFTs characteristics in different precursor solution stirred atmosphere
In order to get high mobility, low sub-threshold swing and near zero threshold
voltage TFT, we adjusted different precursor solutions atmosphere to fabricate ideal
metal-oxide film. In the section, the annealing temperature for zinc-oxide 3-layers
was hold at 600℃. From the beginning, we produced colorless, clear solutions with
molar concentration 0.30M (ZnO) in glove box, but the threshold voltage was -19.77V.
23
The negative threshold voltage, beyond we expect (See in Fig. 3.2 and Table 3-1).
Y. S. et al. reported that the precursor solution stirred in air would enhance the IPCE
of organic solar cell. [30] We tested this method for further investigation. In
fabricating the zinc-oxide precursor (the concentration for precursor solution was
0.43M), we placed the precursor onto the hot-stirring (stirred rate was 250rpm/s) in
air atmosphere at room temperature for 12 h for the hydrolysis reaction in air. The Voff
and VT we estimated at this moment was slightly positive shift than early, moved to
-2.15V. When zinc-oxide precursor solution stirred rate raise to 400rpm/s, the
threshold voltage moved unobvious. Saturation mobility from 1.76 cm2V-1s-1 slight
increased to 2.24 cm2V-1s-1, and VT positive shifted so slightly that we did not find
almost. We raise the film annealing temperature up to 60℃ (placed in hot-stirring
with 250rpm/s and 60℃), the mobility decreased slightly, but the VT and Voff became
more positive, even though the on current, on/off ratio, was small than the devices we
produced than before, but the positive threshold voltage seemed important to further
work. The negative VT in such case no matter zinc-oxide thin-film transistor and
indium-gallium-zinc-oxide thin-film transistor may be observed, there might have
some mechanism in it. We considered that there was one or all of following reasons.
First, when the solutions were stirred at high temperature in air atmosphere, the
solvent of 2-MOE might be evaporated, resulting the thickness of zinc-oxide film
24
would thicker than we expected. Second, stirring at high temperature might result the
change of metal-oxide precursors. In order to realize that first reason was right or
wrong, we designed to vary different active layers to make the zinc-oxide TFTs. From
Fig. 3.3, the thicker thickness may result the more negative VT. This result told us
evaporating of solvent could not induce the threshold voltage shifted to positive. The
phenomenon for negative threshold voltage shift was conjectured because when the
thickness of zinc-oxide film increased, we must need more negative voltage to
suppress the TFT. [10] Contrary to the thickness of metal-oxide film, the primarily
cause for negative shift of zinc-oxide precursor solution stirred in glove box is
because the different ratio for oxygen vacancies. Since the precursor stirred in glove
box was at a atmosphere of lock of O2, the lock of O2 induced the incomplete of
precursor and result the conducted was lead by metal ion stacking in metal-oxide film,
the performance of TFTs was because of rising a larger number of oxygen vacancies
and zinc interstitials. The precursor solutions stirred in air atmosphere or at high
stirred temperature seemed increased the rate of hydrolysis of zinc-oxide precursor
complete so that we only need small gate voltage, the TFT devices turn on quickly.
Besides, the highest mobility we extracted was
2.24 cm2V-1s-1 (the precursor solution stirred vigorously in air at room temperature)
(see Tab 3-1).
25
We observed similar result in AFM (see Fig. 3.4). When the zinc-oxide precursor
solution stirred in air at 60℃, the roughness for zinc-oxide film seem different from
zinc-oxide precursor solution stirred at room temperature. The large crystal for
zinc-oxide film supported our hypothesis so that we could adjust the precursor
solution atmosphere to control the threshold voltage we want.