近年來，由於金屬氧化物半導體擁有諸如：高載子移動率、高均勻度、低製程溫度等的優點，使其成為相當熱門的研究題目。其中高能障的金屬氧化物半導體，由於不吸收可見光，擁有可應用於透明顯示器的潛力，而備受到各界重視，舉例來說如本研究中的非晶態氧化鋅摻銦摻鎵(a-IGZO)。但a-IGZO缺點在於其對環境中氣氛、亮暗以及元件操作偏壓非常敏感，穩定性較差。
在本研究中我們將a-IGZO TFT操作於氧氣與照光的環境中，並探討其在大汲極偏壓下的穩定性。由於應用於主動式有機發光二極體螢幕(AMOLED)的TFT，時常操作於大汲極偏壓，因此大汲極偏壓下的穩定度問題，對IGZO於AMOLED方面的應用非常重要。除了在直流汲極偏壓下的穩定度，我們也更進一步探討交流汲極偏壓下的穩定度問題。另一方面，IGZO在照光下展現的不穩定性，使其擁有應用於光感測器的潛力，我們也探討將IGZO TFT應用於光感測器的可能性。
實驗結果顯示，IGZO當操作於汲極偏壓下時會受到氧氣分子吸附與光產生的電子電洞對影響。而在光感測器研究方面，元件展現出對紫外光有較高的敏感度，在本論文中，IGZO TFT應用於紫外光感測器的操作機制將被充分討論。

Recently, the metal oxide semiconductor is a hot studied subject due to their ad-vanced characteristics, such as high mobility (> 10 cm2 /Vs), good uniformity, and low process temperature. The high band gap semiconductor has excellent transparency, for example amorphous indium gallium zinc oxide (a-IGZO).This property make IGZO can be used for transparent display.But a-IGZO TFTs are very sensitive to atmosphere, illu-mination, and operation bias.
In this study, we discuss about instability of high drain bias in oxygen ambient and under illumination. It is an important issue for application of AMOLED, because TFTs of AMOLED are usually operated in high drain bias. We also study the reliability under AC-drain bias. On the other hand, IGZO unstability of illumination shows good poten-tiality for application of light sensor. We discuss the possibility which IGZO used for sensor.
The experiment results show IGZO TFTs are affect by absorption of oxygen molec-ular and light-induced electron-hole pairs under drain bias stress. The application of light sensor shows a sensitivity of UV. The mechanism of UV sensor operation is clearly study in this paper.

iv
Content
致謝 ............................................................................................................................... i
中文摘要 ...................................................................................................................... ii
English abstract ......................................................................................................... iii
Figure caption ............................................................................................................ vi
Table caption ............................................................................................................ xiii
Chapter1 Introduction ......................................................................................... 1
1-1 Overview ............................................................................................... 1
1-2 Why IGZO? ........................................................................................... 1
1-3 Motivation ............................................................................................. 3
Chapter2 Device fabrication and electrical characterization ............................. 8
1-1 Device fabrication .................................................................................. 8
2-2-1 Device fabrication-Drain Bias Stress .............................................. 8
2-2-2 Device fabrication-for UV sensor operate ....................................... 8
2-3 Electrical characteristics......................................................................... 9
2-3-1 The I-V transfer characteristics ....................................................... 9
2-3-2 The C-V transfer characteristics ................................................... 11
Chapter3 Experiment instrument and parameter extraction ........................... 17
3-1 Instruments .......................................................................................... 17
3-1-1 Drain bias stress ........................................................................... 17
3-1-2 UV sensor .................................................................................... 18
3-2 Parameter extraction ............................................................................ 19
3-2-1 The definitions of normalized drain current and threshold voltage ,
subthreshold swing extraction ...................................................................... 19
3-2-2 The definition of normalized capacitance ..................................... 20
Chapter4 Experiment result and discussion ...................................................... 25
4-1 Drain bias stress ................................................................................... 25
4-1-1 Drain bias stress in oxygen atmosphere ........................................ 25
4-1-2 Drain bias stress under illumination condition .............................. 28
4-2 AC-Drain bias stress ............................................................................ 32
4-3 The study of UV sensor application...................................................... 34
Chapter5 Conclusion .......................................................................................... 67
Chapter6 Reference ............................................................................................ 69

Figure caption
Fig. 1-1-1 Graphical summary of required carrier mobility for future displays. ............. 5
Fig. 1-2-1 Schematic orbital drawings for the carrier transport paths (that is, conduction
band bottoms) in crystalline and amorphous semiconductors. ........................................ 6
Fig. 1-2-2 Amorphous-phase formation of IGZO thin films[6]. ..................................... 6
Fig. 1-2-3 Electron transport properties of IGZO thin films[6]. ..................................... 7
Fig. 2-1-1 The sample structure diagram used in Drain Bias Stress. ............................ 13
Fig. 2-1-2 The sample structure diagram used in UV sensor. In fact, there still has a
passivation layer on the device. ................................................................................... 13
Fig. 2-2-1 A normal I d versus V ds characteristic curve measured from TFT. ................. 14
Fig. 2-2-2 The red color region and arrows are symbols of electron path. .................... 14
Fig. 2-2-3 This is I ds versus V gs curve. In this picture, we can observe three stages,
which are A) accumulation stage, B) depletion stage and C) inversion stage................ 15
Fig. 2-2-4 This is gate-to-drain capacitance-voltage curve, which the measurement
frequency is 1MHz. ..................................................................................................... 15
Fig. 2-2-5 This picture shows the capacitance which extend from drain to source when
the channel is going to turn on..................................................................................... 16
Fig. 3-1-1 Microscope, LCD display, and TTP-6 probe station. ................................... 21
Fig. 3-1-2 The picture shows I-V property analyzers, which are Agilent 4156C and
Agilent B1500A, LCR meter Agilent 4980A, and the switchings, which are Agilent
4980A and AgilentB2201A ......................................................................................... 21
Fig. 3-1-3 The connecting of the semiconductor parameter analyzer, TTP-6 probe
station and computer. .................................................................................................. 22
Fig. 3-1-4 Microscope, LCD display, dark box and Cascade Microtech M150 probe
station. ........................................................................................................................ 22
Fig. 3-1-5 The connecting of the Agilent B1500A, Cascade Microtech M150 probe
station and computer. .................................................................................................. 23
Fig. 3-2-1 Threshold voltage determination by the linear extrapolation technique.[19] 23
Fig. 3-2-2 Threshold voltage is determined by constant NI d . ....................................... 24
Fig. 3-2-3 The different connections of C-V measurement, C GC , C GD and C GS . ........... 24
Fig. 4-1-1 The NI d -V gs curves are shown in picture before and after Drain Bias Stress in
vacuum ....................................................................................................................... 39
Fig. 4-1-2 The NI d -V gs curves are shown in picture before and after Drain Bias Stress in
oxygen. ....................................................................................................................... 39
Fig. 4-1-3 The V ds -I ds curves are shown in picture before and after stress.The (a) is
shown the degradation in vacuum and (b)(c) are shown in oxygen. The difference
between (b) and (c) is that the former sweep voltage is applied on drain electrode, the
latter on source electrode. ............................................................................................ 40
Fig. 4-1-4 The band gap diagram and electron behavior is shown in picture during the
sweep voltage applied on drain or source. ................................................................... 41
Fig. 4-1-5 The CGC curves are shown in picture before and after Drain Bias Stress. ... 42
Fig. 4-1-6 The single side capacitance curves are shown in picture before and after
stress. The right side (b)(d) is CGD curves and the left side (a)(c) is CGS curves. We
can observe the oxygen induced effect in (c)(d). .......................................................... 42
Fig. 4-1-7 The NI d -V gs curves are shown in picture before and after Drain Bias
Illumination Stress in (a) vacuum and (b) oxygen ambient. ......................................... 43
Fig. 4-1-8 The I d -V ds curves are shown in picture before and after Drain Bias
Illumination Stress in (a) vacuum and (b) oxygen ambient. ......................................... 44
Fig. 4-1-9 The CGC curves are shown in picture before and after Drain Bias
Illumination Stress in (a) vacuum and (b) oxygen ambient. ......................................... 45
Fig. 4-1-10 The single side capacitance curves are shown in picture before and after
stress. The right side (b)(d) is CGD curves and the left side (a)(c) is CGS curves. We
can observe the hump effect in CGD curves and the oxygen induced effect in (c)(d). .. 45
Fig. 4-1-11 The CGD curves are classified by environment. The hump effect is only
take place under illumination condition (b)(d) and the oxygen induced distortion effect
is shown both in darkness and illumination (c)(d). ....................................................... 46
Fig. 4-1-12 The CGC curves are shown in picture. The hump effect disappears after rest for 4hr. ........................................................................................................................ 46
Fig. 4-1-13 (a)At initial stage, the device is operated at subthreshold region.(b) During
stress stage, the device is operated at turn-on region. ................................................... 47
Fig. 4-1-14 (a)At initial stage, vertical electrical field drive hole trap at interface.(b)
During stress, the trapping keep going at drain region, but does not appear at center. .. 47
Fig. 4-1-15 With the accumulation of trapped holes, obvious energy barrier-lowing can
be obtained and the hump effect of CGC and CGD curve occurred. ............................ 48
Fig. 4-1-16 With the accumulation of trapped holes, obvious energy barrier-lowing can
be obtained and the hump effect of CGC and CGD curve occurred. ............................ 48
Fig. 4-1-17 We measured CGC curves as a function of applied stress time. First, the
negative Vth shift occurs for 10 s. Then the hump effect starts to appear after the DBIS
for 100s. ...................................................................................................................... 49
Fig. 4-1-18 The simulation results are shown in picture. All the electrical characteristics
of simulation are same with DBIS experiment. ............................................................ 50
Fig. 4-1-19 The NI d -V gs curves are shown in picture before and after Drain Bias
Illumination Stress with source floating under (a) dark and (b) light. ........................... 51
Fig. 4-1-20 The CGC curves are shown in picture before and after Drain Bias
Illumination Stress with source floating under (a) dark and (b) light. ........................... 52
Fig. 4-1-21 The hump effect is disappear in C-V characteristics. (c)(d) Under illumination condition only have V th shift. (a)(b) In contrast, the properties are very
stable under dark. ........................................................................................................ 53
Fig. 4-1-22 The energy band diagram is shown in picture. The vertical electrical filed
drive holes trap at insulator/channel interface in entire channel. .................................. 53
Fig. 4-2-1 The AC stress condition is shown in picture. The AC voltage is applied in
drain electron and waveform of AC is two steps level. ................................................ 54
Fig. 4-2-2 The CGD transfer characteristics don’t be observed hump effect after
AC-DBIS, which duty ratio is 1 and etch peak/base period is (a) 10us, (b) 100us and (c)
1000us. ....................................................................................................................... 54
Fig. 4-2-3 The CGD transfer characteristics of a-IGZO TFT after AC-DBIS. The base
time widths are 100us, 10us and 1us and the peak time width is 1000us. ..................... 55
Fig. 4-2-4 The CGD transfer characteristics of a-IGZO TFT after AC-DBIS. The peak
time widths are 1000us, 100us and 10us and the base time width is 1us. ..................... 55
Fig. 4-2-5 The vertical cross-section energy band diagram of hole-trapping and
hole-electron recombine mechanism near the drain region during the (a) peak time and
(b)base time of AC-DBIS. ........................................................................................... 56
Fig. 4-2-6 The CGD transfer characteristics of a-IGZO TFT after AC-DBIS under
different light illumination, which are (a) 20000 lux, (b) 10000 lux and (c) 5000 lux. All
AC-DBIS have proceeded in duty ratios of 10, 100 and 1000. ..................................... 57
Fig. 4-3-1 shows asymmetric structure, which is photographed by microscope, and
cross section of sample from drain to source. .............................................................. 57
Fig. 4-3-2 The definitions of LED wave length are (a) red[660nm], (b) blue[470nm] &
green[525nm], and (c) UV[375nm]. ............................................................................ 58
Fig. 4-3-3 The absolute drain current versus gate voltage curves (AI d -V gs ) are show in
picture. The AI d -V gs curves are stable under (a) red, (b) blue, and (c) green. ................ 58
Fig. 4-3-4 We can observe light-induced leakage current under UV. ............................ 59
Fig. 4-3-5 The energy band diagram show that (a) barrier lowering induced by
accumulation of holes and (b) drain current increase with growth of drain voltage [22].
................................................................................................................................... 59
Fig. 4-3-6 Leakage current has different value under intensity of UV are various from
100lux to 12600lux.(a) AI d -V gs curves and (b) sensor operated characteristic properties.
................................................................................................................................... 60
Fig. 4-3-7 The Persistent Photo-leakage (PPC) effect is shown in picture. ................... 61
Fig. 4-3-8 The Persistent Photo-leakage (PPC) is resulted from accumulation of holes
and slightly holes trapping at channel/oxide interface. ................................................. 61
Fig. 4-3-9 We apply a pulse to reset sensor after illuminated by UV light. ................... 62
Fig. 4-3-10 We set pulse time as 100us. Picture shows gate voltage of pulse are various
from 10V to 40V. The PPC decrease is in proportion to gate voltage. .......................... 62
Fig. 4-3-11 We vary the drain voltage from 0V to 30V.The decay of PPC is proportional
to the increase of drain voltage. ................................................................................... 63
Fig. 4-3-12 The energy band diagram of applied by gate pulse is illustrated in picture. 63
Fig. 4-3-13 The lateral field attracts electrons into channel and electrons recombine with
holes which are trapped at defect of interface. ............................................................. 64
Fig. 4-3-14 We change the length of pulse time as 1s, 0.5s, 100us, and 1us.The PPC rise
with pulse time decay. ................................................................................................. 64
Fig. 4-3-15 We can observe PPCs are separated by gate voltage of pulse..................... 65
Fig. 4-3-16 The AI d -V gs characteristics is same whether drain is at I type or U type
under dark and UV. ..................................................................................................... 65
Fig. 4-3-17 We can observe PPCs are decreased when drain is U type electrode at same
pulse condition. ........................................................................................................... 66
Fig. 4-3-18 The most stronger electrical field is at I type electrode whether the drain
voltage is applied on I type electrode or U type electrode. ........................................... 66

Table caption
Table 1-1-1 Comparison of a-Si TFT, poly-Si and Oxide-based TFT............................. 5

[1] T. Kamiya, K. Nomura and H. Hosono. Sci. Technol. Adv. Mater. (2010), Vol. 11, pp.
044305.
[2] K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, and H. Hosono. Nature (2004),
432, 488-492.
[3] C. T. Tsai, T. C. Chang, S. C. Chen, I. Lo, S. W. Tsao, M. C. Hung, J. J.Chang, C. Y.
Wu, and C. Y. Huang. Appl, Phys. Lett. (2010),Vol. 96 242105
[4] E. Fortunato, P. Barquinha, A. Pimentel, A. Goncalves, A. Marques, L. Pereira, and R.
Martins. Adv. Mater (Weinheim Ger) (2005), 17 590.
[5] H. N Lee, J. Kyung, S.K. Kang, D. Y. Kim, M. C Sung, S. J. Kim, C. N. Kim, H. G.
Kim, S. T. Kim, Tech. Dig. SID, California, USA, (2007) 1826.
[6] T. Kamiya and H. Hosono. NPG Asia Material (2010), Vol. 2, pp. 15-22.
[7] S.Y. Huang, T.C. Chang, M.C. Chen, S.C. Chen, T.C. Chen and F.Y. Jian . Electro-
chem. Solid-State Lett. (2012), Vol.15(5) H161-H163.
[8] W.F. Chung, T.C. Chang, H.W. Li, C.W. Chen, Y.C. Chen, S.C. Chen, T.Y. Tseng and
Y.H. Tai, Electrochem. Solid-State Lett. (2011), Vol.14(3) H114-H116.
[9] G.W. Chang, T. C. Chang, J. C. Jhu, T. M. Tsai, Y. E. Syu, K. C. Chang, Y. H. Tai, F.
Y. Jian, Y. C. Hung. IEEE Electron Device Lett. (2012), Vol. 3 540-542.
[10] W. F. Chung, T. C. Chang, H. W. Li, S. C. Chen, Y. C. Chen, T. Y. Tseng and Y. H. Tai. Appl. Phys. Lett. (2011), Vol.98 152109.
[11] T.Y. Hsieh, T.C. Chang, T.C. Chen, Y.T. Chen, M.Y. Tsai, A.K. Chu,Y.C. Chung,
H.C. Ting, and C.Y. Chen. IEEE Electron Device Lett. (2013), Vol.34 63-65
[12] M.Y. Tsai, T.C. Chang, A.K. Chu, T.C. Chen, T.Y. Hsieh and Y.T. Che. ECS Trans.
(2012), Vol.45 119-131.
[13] X. Huang, C. Wu, H. Lu, F. Ren, Q. Xu, H. Ou, R. Zhang ,and Y. Zheng. Appl.
Phys. Lett. (2012), Vol.100 243505.
[14] S.Y. Huang, T. C. Chang, M. C. Yang, L. W. Lin, M.H. Wu, K. H. Yang, M. C.
Chen, Y. J. Chiu, and Bo-Liang Yeh. Appl. Phys. Lett. (2012), Vol.101 253502.
[15] N. Yamaguchi, S. Taniguchi, T. Miyajima, and M. Ikeda. J. Vac. Sci. Technol. B.
(2009), Vol.27 1746.
[16] K. Takechi, M. Nakata, T. Eguchi, H. Yamaguchi, and S. Kaneko, Appl. Phys. Lett.
(2010), Vol.97 192103.
[17] M. Fujii, Y. Hiroshi, H. Tomoaki, U. Yukiharu, F. Takashi, J. J. Sim, and K. J. Yeon,
Jpn. J. Appl. Phys. (2008), 47(8) 6236–6240.
[18] Schroeder DK. Semiconductor material and device characterization, 2nd ed. New
York: Wiley; 1998.
[19] D. Kang, H. Lim, C. Kim, I. Song, J. Park, and Y. Park. Appl. Phys. Lett. (2007),
90:192101.
[20]Y. Kamada, S. Fujita, T. Hiramatsu, T. Matsuda, M. Furuta, T. Hirao. Solid-State
Electron. (2010), 54 1392–1397.