Paper Number: abstract number (667), program number (GP-28)
Title of Paper: Effect of low-temperature supercritical fluid technology treatment
on the performance of ZnO TFTs
Name: Min-Chen Chen
Corresponding Author: Ting-Chang Chang
Full Mailing Address:
Department of Physics, National Sun Yat-Sen University, 70 Lien-Hai Road,
Kaohsiung, Taiwan 80424, ROC.
Telephone: +886-7-5252000-3708
Fax: +886-7-5253709
e-mail: tcchang@mail.phys.nsysu.edu.tw
Keywords: ZnO, supercritical fluid
Estimation of the length of the manuscript
Number of Pages
13
x 250
3250
Number of Tables
1
x 150
150
Number of Figures
8
x 150
1200
TOTAL NUMBER OF WORDS =
4600
Effect of low-temperature supercritical fluid
technology treatment on the performance of
ZnO TFTs
Min-Chen Chen(a, Ting-Chang Chang(a,b, Sheng-Yao Huang(a,
Kuan-Chang Chang(c, Hui-Chun Huang (d, Shih-Ching Chen (a, Jin Lu (a,
Der-Shin Gan(d, New-Jin Ho(d, Tai-Fa Young(c, Geng-Wei Jhang(e,
Ya-Hsiang Tai(e
a.
b.
c.
d.
e.
Department of Physics, National Sun Yat-Sen University, Kaohsiung, 804,
Taiwan, R.O.C.
Center for Nanoscience & Nanotechnology, National Sun Yat-Sen University,
Kaohsiung, 804, Taiwan, R.O.C.
Department of Mechanical and Electro-Mechanical Engineering, National Sun
Yat-Sen University, Kaohsiung, 804, Taiwan, R.O.C.
Department of Materials Science and Engineering, National Sun Yat-Sen
University, Kaohsiung, 804, Taiwan, R.O.C.
Department of Display Institute, National Chiao Tung University, Hsin-Chu, 300,
Taiwan, R.O.C.
Abstract
In this paper, the supercritical CO2 (SCCO2) fluid technology is successfully
applied to improve the electrical characteristics of sputtered ZnO TFTs at low
temperature (150 °C). After the treatment of SCCO2 fluids mixed with water, the ZnO
TFT exhibited superior transfer characteristics and lower threshold voltage. According
to X-ray photoelectron spectroscopy (XPS) analyses, the improvements were
attributed to the increase of binding energies of Zn-O bonds, hydrogen-related donors
and the reduction of traps at the grain boundary in ZnO thin films. In addition, the
crystalline quality, physical structure, and transmittance of ZnO thin films were also
investigated by using X-ray diffraction (XRD), transmission electron microscopy
(TEM) and N&K analyzer.
Introduction
In the recent years, thin film transistors (TFTs) based on transparent oxide
semiconductors (TOSs) have attracted much attention for large area electronics due to
their excellent uniformity, great transparency to the visible light, potentially better
device performance and stability compared with the conventional amorphous-Si TFT.
Among several TOSs materials, there have been a number of reports on TFTs that
apply zinc oxide (ZnO) as the active channel layer. Various deposition techniques,
such as pulsed laser deposition [1-2], chemical vapor deposition [3-4], spin-coating
[5], and sputtering [6-7] have been proposed to prepare the ZnO films. The sputtering
deposition is the most favorable among these methods because the fabrication of ZnO
TFT is comparable to the conventional amorphous-Si TFT. However, the
low-temperature deposited ZnO thin films without doping probably exhibits an
insulating property [8-9]. Therefore, a high-temperature post-deposition-annealing [10]
or a high deposition temperature [8] is necessary to further improve the quality of
ZnO films. Nevertheless, the high-temperature process is not suitable to glasses and
plastics substrates due to the low glass transition temperatures (Tg). Therefore, it is
critical to find a low-temperature method to enhance the quality of ZnO thin-film for
fabricating the next generation TFTs.
In our previous works, supercritical CO2 (SCCO2) fluid technology has been
proposed to passivate the defect states in hydrogenated amorphous-silicon thin-film
transistors (a-Si:H TFTs) and improve the dielectric properties of sputter-deposited
hafnium oxide films [11-12]. The supercritical fluid exhibits liquid-like property,
which has good transport capacity, and also holds gas-like properties to diffuse into
the nanoscale structures without damage.
Hence, the SCCO2 fluid can carry the H2O
molecule effectively into the ZnO thin films and to passivate the traps of the deposited
film at low temperature. In this paper, SCCO2 fluid technology, which is a
low-temperature processing set at 150 ℃, is proposed to improve the performance of
sputtered ZnO TFTs. The efficacy of the SCCO2 treatment has also been discussed by
the material analyses and electrical characteristics of the sputtered ZnO films (without
doping).
Experimental
We have fabricated the bottom-gate and bottom-contact type TFTs having a
undoped ZnO thin films as the active channel layer. The schematic cross-section view
of the undoped ZnO TFTs is shown in Fig. 1. The channel width (W) and length (L)
of the TFTs were 50  m and 8  m, respectively. The device was fabricated by the
following deposition procedure. First, a 300-nm-thick MoW gate electrode was
deposited and patterned on the glass substrate. For the gate insulator, a 300-nm-thick
silicon nitride (SiNx) was deposited by PECVD (Plasma Enhanced Chemical Vapor
Deposition). Then, a 100-nm-thick ITO film was deposited and patterned for forming
the source/drain electrodes. Finally, a 100-nm-thick undoped ZnO film was formed by
radio frequency (RF) magnetron sputtering system at room temperature (RT). The
active layer was defined using photolithography and wet etching. Subsequently,
different post-treatment methods were performed on the undoped ZnO films to
improve the electric characteristics. The first group of ZnO TFTs labeled as “H2O
vapor” was placed in a pressure-proof stainless steel chamber at 150 °C for 1 h, and a
pure H2O vapor was immersed into ambience. For the second group, device labeled as
“SCCO2” was placed in a supercritical fluid system at 150 °C for 1 h, where it was
injected with a 3000 psi SCCO2 fluids mixed with 5ml pure H2O. In addition, the
third group of ZnO TFTs without any treatment was taken as the control sample,
marked as “standard.” After these different treatments, the electrical characteristics of
device were carried out using Agilent 4156-C semiconductor parameter analyzer in
the dark environment at room temperature. The crystalline quality and crystal
orientation of ZnO thin films deposited on Si wafer were investigated by the X-ray
diffraction (XRD). In addition, the physical structure and chemical functional bonding
of these treated-ZnO films were analyzed by transmission electron microscopy (TEM)
and X-ray photoelectron spectroscopy (XPS).
Results and discussion
Figure 2 shows the typical transfer characteristics for the ZnO TFTs that have
gone through RT deposition and subsequent post-treatment under two different
conditions. As shown in the figure, the characteristics of undoped ZnO TFTs are
much affected by the post-treatment. For the standard ZnO TFT, the device
characteristics exhibit a largest threshold voltage (VT) and a lowest drain current (ID).
The VT of ZnO TFTs was extracted from the plot of (ID)1/2 versus gate voltage (VG),
as the derivative in the linear region of the plot and the intersection with the VG axis.
The result indicated that the standard ZnO device with undoped ZnO films as an
active channel layer does not show acceptable transistor characteristics. Comparing
the standard device, as the device after H2O vapor treatment, the ID and subthreshold
slope (S.S) of ZnO device are slightly improved. However, a large VT still exists.
Clearly, for the SCCO2 treatment, the VT, ID and on/off current ration are significantly
improved compared to the two devices. From the above results, SCCO2 treatment is
the most effective method to improve the device property.
The electrical properties can be influenced by the film structure, especially the
grain sizes and shapes. The XRD spectra for the ZnO films before and after
post-treatments were shown in Fig. 3. As can be seen, all samples exhibit a similar
peak located at about 34 degree is the (002) plane with the crystallized phase of the
wurtzite structure, indicating a preferred orientation with the c-axis perpendicularly to
the substrate surface [13]. Figure 4 shows the cross-sectional TEM images of the ZnO
films under various treatment conditions: (a) standard, (b) H2O vapor treated, and (c)
SCCO2 treated, and it is observed that the thickness are 107 nm, 110 nm, and 100 nm,
respectively. The inset of Fig. 4a-c shows diffraction patterns taken from a region of
the ZnO thin films. The ring pattern indicated that the ZnO films are polycrystalline
wurtzite structure, containing many crystalline grains [14]. Even though after
post-treatments, the grain size, crystallized phase and orientation are approximately
the same to standard sample. Therefore, the improvement of device characteristics is
not from the film structure, maybe from the dangling bonds (or traps) passivated.
Because the as-deposited ZnO film is consisted of numbers of ZnO grains, the
carriers have to transport from source to drain via the grain boundaries as the device is
operated. Therefore, the traps in grain boundaries may cause the potential barrier
resulting in a low ID and a high VT. Although the current is enhanced when after H2O
vapor treatment, a large VT still existed, and it means that only partial traps were
passivated. Therefore, the most traps at grain boundaries were passivated by SCCO2
treatment result in the device characteristics significantly improved.
In order to analyzer the composition of the ZnO film after post-treatment, the
XPS analysis is performed, and the binding energies have been calibrated by taking
the carbon C 1s peak (285.0 eV) as reference. The Zn 2p3/2 peaks of ZnO thin films
after different treatments are shown in Fig. 5, and the energy state of standard sample
at about 1021.6 eV is corresponding to Zn-O bonding [15]. After the treatment of H2O
vapor and SCCO2 treatment, the binding energy shifts to 1021.9 and 1022.1 eV,
respectively. The stronger Zn-O bonding indicates that Zn atoms in ZnO film have
been oxidized more completely with the H2O molecules. In addition, the SCCO2
treatment is more efficient than H2O vapor treatment in transporting H2O molecules
into the ZnO film.
The O 1s peaks of various treatments ZnO thin films are shown in Fig. 6. The O
1s peaks have a shoulder at higher binding energy, and it increases obviously after the
treatment of H2O vapor and SCCO2. The lower energy peak detected at 530.0 eV is
corresponded to O-Zn bonding, and the higher energy peak detected at 531.6 eV is
attributed to O-H bonding which can be due to the H2O molecules adsorption in the
ZnO films [16]. However, the standard sample without any post-treatment also exhibit
the high-energy peak of O-H bonding, which originated from the surface of ZnO films
with adsorbed H2O molecules as it is exposed to the air. In addition, the H2O absorbed
in ZnO thin films was dissociated into H and OH groups, and rapidly diffuse into the
ZnO film [17]. As previous study reported, the hydrogen atoms can be incorporated
into ZnO to form hydrogen-related donors, such as the substitutional hydrogen bound
to zinc atoms and the interstitial hydrogen bound to oxygen atoms, and improve the
electrical conductivity of ZnO film [18-19]. Otherwise, the dangling bonds (or traps)
in the grain boundaries can be passivated by OH and form Zn-OH bonds. Therefore,
the dangling bonds (or traps) induced potential barrier can be suppressed. By using
the integrated peak area with the sensitivity factors, the atomic concentrations of Zn
and O were listed in Table 1. The O-H bonding for the standard ZnO sample and the
ZnO film after H2O vapor are 15.02% and 16.36%, while the O-H bonding after
SCCO2 treatment rapidly increases to 46.13%. The results of the O 1s spectra indicate
that H2O molecules can be effectively carried into the ZnO film by the
highly-diffusing SCCO2 fluid and passivate the traps.
To investigate the effect of various treatments for transmittance of ZnO films and
avoid the influence of Si substrates, the ZnO films deposited on the glass substrates
under the same deposited conditions were prepared. Figure 7 shows the transmittance
spectra of the ZnO films after post-treatments. Clearly, the transmittance for all
samples is almost the same and exhibits the high transmittance above 80% within the
wavelength range from 430 nm to 700 nm.
For test the electrical stability of the ZnO TFTs after SCCO2 treatment, a
prolonged gate bias stress was applied to investigate the degradation of the current.
The device was stress at VGS=20V+VT for 1000s, while the source and drain were
grounded. Figure 8 shows the shifts of the transfer characteristic after gate bias stress
and the recovery characteristic (it is measured after 1000s standing). It is evident that
VT after stress is shifted apparently to positive direction, but the shape of the transfer
curve is nearly unchanged compared to initial curve. Because the S.S is not varied, the
degradation is mainly attributed to charge trapping in the bulk dielectric layers or at
the channel/dielectric interface [20-21]. Otherwise, the stressed transfer characteristic
is rapidly recovered to initial within 1000 sec due to the charge de-trapping. Thus, the
degradation is caused by the effect of a transient charge trapping.
Conclusion
The effects of low-temperature SCCO2 treatments on the ZnO films have been
demonstrated experimentally. The preliminary improvement is obtained by H2O vapor
treatment attribute to the H2O absorbed in ZnO thin films, and the hydrogen atoms
can be incorporated into ZnO to form hydrogen-related donors. Besides, the dangling
bonds (or traps) in the grain boundaries can be passivated by OH and form Zn-OH
bonds. A further investigation verifies that traps passivation efficiency can be
optimized by the SCCO2 treatment, because SCCO2 fluid could more effectively
transport H2O molecules into ZnO films. Therefore, lower potential barrier at grain
boundaries resulting in the device characteristics significantly improved. From XRD
and TEM, the grain size and crystal orientation are approximately unchanged which
means SCCO2 fluids can diffuse into the nanoscale structures without damage.
Moreover, the ZnO films exhibits a high transmittance above 80% within the visible
light range. The changes of VT under gate bias stress do not require any thermal/bias
annealing for the devices to recover their initial state, which is caused by the effect of
a transient charge trapping in the bulk dielectric or at the channel/dielectric interface.
Acknowledgment
This work was performed at National Science Council Core Facilities Laboratory
for Nano-Science and Nano-Technology in Kaohsiung-Pingtung area and was
supported by the National Science Council of the Republic of China under Contract
Nos.NSC-97-3114-M-110-001 and NSC 97-2112-M-110-009-MY3.
Reference
1.
S. Masuda, K. Kitamura, Y. Okumura, S. Miyatake, H. Tabata, and T. Kawai, J.
Appl. Phys. 93, 1624 (2003).
2.
V. Craciun, J. Elders, J.G.E. Gardenievs and I.W. Boyd, Appl. Phys. Lett. 65,
2963 (1994).
3.
K. Minegishi, Y. Koiwai, Y. Kikuchi, K. Yano, M. Kasuga, and A.Shimizu, Jpn. J.
Appl. Phys., Part 2 36, 1453 (1997).
4.
M. Purica, E. Budianu, E. Rusu, M. Danila, R. Gavrila, Thin Solid Films
403-404, 485 (2002).
5.
B. J. Norris, J. Anderson, J. F. Wager, and D. A. Keszler, J. Phys. D 36, L105
(2003).
6.
G. Xiong, J. Wilkinson, B. Mischuck, S. Tuzemen, K. B. Ucer, and R. T.
Williams, Appl. Phys. Lett. 80, 1195 (2002)
7.
K. B. Sundaram and A. Khan, Thin Solid Films 295, 87 (1997).
8.
H. S. Bae, J. H. Kim, and S. Im, Electrochem. Solid-State Lett. 7, G279 (2004).
9.
J. Y. Lee, Y. S. Choi, W. H. Choi, H. W. Yeom, Y.K. Yoon, J. H. Kim, and S. Im,
Thin Solid Films 420-421, 112 (2002).
10. H. Q. Chiang, J. F. Wager, R. L. Hoffman, J. Jeong, and D. A. Keszler, Appl.
Phys. Lett. 86, 13503-1 (2005).
11. C. T. Tsai, P. T. Liu, T. C. Chang, C. W. Wang, P. Y. Yang, and F. S. Yeh, IEEE
Trans. Electron Device Letters, vol. 28, no. 7, p.584 (2007)
12. C. T. Tsai, T. C. Chang, P. T. Liu, P. Y. Yang, Y. C. Kuo, K. T. Kin, P. L. Chang,
and F. S. Yeh, Appl. Phys. Lett. 91, 012109 (2007).
13. L.W. Lai and C.T. Lee, Mater Chem Phys 110, 393 (2008).
14. S.J. Henley, M.N.R. Ashfold and D. Cherns, Surf. Coat. Tech. 177, 271(2004).
15. C.X. Xu, X.W. Sun, X.H. Zhang, L. Ke and S.J. Chua, Nanotechnology 15, 856
(2004).
16. L. J. Meng, C. P. Moreira, and M. P. d. Santos, Appl. Surf. Sci. 78, 57 (1994)
17. Y. F. Lu, H. Q. Ni, Z. H. Mai, and Z. M. Ren, J. Appl. Phys. 88, 498 (2000).
18. Min-Suk Oh, Dae-Kue Hwang, Jae-Hong Lim, Yong-Seok Choi, and Seong-Ju
Park, Appl. Phys. Lett. 91, 212102 (2007).
19. H. S. Bae, J. H. Kim, and S. Im, Electrochem. Solid-State Lett. 7, G279 (2004).
20. Suresh A, Wellenius P and Muth J F Tech. Dig.—Int. Electron Devices Meet. p
587 (2007)
21. M. J. Power, C. V. Berkel, A. R. Franklin, S. C. Deane, and W. I. Milne, Phys.
Rev. B 45, 4160 (1992).
Figure Captions:
Figure 1 Schematic diagram of the fabricated ZnO TFT structure.
Figure 2 Transfer
characteristics
of
undoped
ZnO
TFT before
and
after
post-treatments.
Figure 3 X-ray diffraction of ZnO films before and after post-treatments.
Figure 4 Cross-sectional TEM image and diffraction pattern (inset) of the ZnO thin
films after various treatment conditions, including (a) standard, (b) H2O
vapor, (c) SCCO2.
Figure 5 XPS of Zn 2p3/2 spectra for the ZnO thin films after different treatments.
Figure 6 XPS of O 1s spectra with two resolved O bonding components for the ZnO
thin films after various treatment conditions, including (a) standard, (b) H2O
vapor, (c) SCCO2.
Figure 7 Transmittance spectrum of ZnO films on a glass substrate under various
treatments.
Figure 8 Comparison of the transfer curves of the ZnO TFTs after SCCO2 treatment
measure before bias stress (t = 0 s), after bias stress (t = 1000 s) and after
relaxation (t = 2000 s).
Table:
Table 1 Atomic concentrations calculated from XPS of the ZnO films after various
treatment conditions.