Degradation on the Current Saturation of Output ...

Degradation on the Current Saturation of Output Characteristics in Amorphous InGaZnO Thin-Film TransistorsIEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO. 8, AUGUST 2018 3243
Degradation on the Current Saturation of Output Characteristics in Amorphous InGaZnO
Thin-Film Transistors Hye Ri Yu, Jun Tae Jang, Daehyun Ko, Sungju Choi , Geumho Ahn, Sung-Jin Choi ,
Dong Myong Kim , Member, IEEE, and Dae Hwan Kim , Senior Member, IEEE
Abstract— Degradation on the current saturation of the output characteristics in amorphous indium– gallium–zinc–oxide (a-IGZO) thin-film transistors with the bottom-gate structure is investigated. As the drain-to- source voltage (VDS) increases at a fixed gate-to-source voltage (VGS), the current from drain to source (IDS) becomes nonsaturated and increases due to the mobility enhancement and IDS also decreases again due to the electron trapping into the gate insulator and/or an interface, which occurs mainly at the channel edge near a drain. Analysis is validated by using the pulsed IDS–VDS measurement. Nonsaturated current is attributed to the Joule heating-assisted mobility enhancement and the thermally activated electron trapping, which is the reason why the nonsaturation becomes more prominent as the IGZO active thin film becomes thinner. Furthermore, it is found that the rate of the increased IDS per increasing VDS gets higher as either the channel width (W ) increases or the channel length (L) decreases; whereas, the rate of the decreased IDS per increasing VDS becomes higher as the device size, i.e., W × L, increases. The former is well correlated with Joule heating while the latter with the self-heating-assisted electron trapping due to a total heat accumulated in an active layer.
Index Terms— Amorphous indium–gallium–zinc–oxide (a-IGZO), charge trapping, nonsaturated current, output characteristic, self-heating, thin-film transistor.
I. INTRODUCTION
THIN-FILM transistors (TFTs) with channel layer made from amorphous indium–gallium–zinc–oxide (a-IGZO)
have been actively researched and begun to be commercialized as the switching/current-driving devices for large-area, high
Manuscript received February 5, 2018; revised April 4, 2018 and May 24, 2018; accepted May 30, 2018. Date of publication June 21, 2018; date of current version July 23, 2018. This work was supported in part by the National Research Foundation of Korea funded by the Korean Government (MSIP) under Grant 2016R1A5A1012966 and in part by the Ministry of Education, Science and Technology (MEST) under Grant 2017R1A2B4006982. The CAD software was supported in part by SILVACO and in part by the IC Design Education Center (IDEC). The review of this paper was arranged by Editor B. Kaczer. (Corresponding author: Dae Hwan Kim.)
The authors are with the School of Electrical Engineering, Kookmin University, Seoul 136-702, South Korea (e-mail: [email protected]).
Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TED.2018.2844862
frame-rate display backplanes of active-matrix liquid crys- tal displays and active-matrix organic light-emitting diodes because they have good electrical and optical characteristics, such as high mobility, large-area uniformity, transparency in the visible light range, low-temperature/cost fabrication processing, and compatibility with flexible substrates [1]. Moreover, a-IGZO TFTs have also been employed very recently in the sensor-embedded flexible circuits for wear- able healthcare and Internet-of-Things applications [2], [3], in which the IGZO TFTs were used not only in the dig- ital logic gates but also used as the analog amplifiers. Therefore, a robust current saturation in output characteristics of IGZO TFTs becomes more important in terms of perfor- mance and reliability of the a-IGZO analog circuitry. However, up to now, the reliability of IGZO TFTs has been evaluated mainly through their transfer characteristics, such as threshold voltage (VT ) and subthreshold slope (SS), rather than output characteristics [4], [5]. Reliability of output characteristics has been rarely investigated compared with its importance.
In this paper, the degradation on the current saturation of the output characteristic of the a-IGZO TFT with the bottom- gate structure is investigated with varying either the thickness of a-IGZO active film or the device size. A large device size is used to exclude short-channel or narrow-width effects. The nonsaturated current is found to result from the Joule heating- assisted mobility enhancement followed by the self-heating- assisted electron trapping into the gate insulator (GI) and/or interface.
II. EXPERIMENT
A schematic of the fabricated a-IGZO TFT with a bottom-gate structure is illustrated in Fig. 1(a). The TFTs were fabricated on highly doped p-type silicon substrates with thermally grown SiO2 layers. The highly doped p-type silicon and SiO2 acted as the gate electrode and GI, respectively, with the GI thickness (Tox) = 50 nm. The a-IGZO layer was then RF sputter-deposited in an Ar/O2 mixture (3/0.1 sccm) at room temperature. In this process, the active-layer thickness (Tact) was designed from 15 to 100 nm in order to investigate the effect of IGZO film thickness [see insets of Fig. 1(c) and (d)]. Before depositing the source/drain (S/D) electrode, the device was treated with Ar plasma (150 W, 0.1 torr) to improve
0018-9383 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
3244 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO. 8, AUGUST 2018
Fig. 1. (a) Schematic 3-D view of the fabricated a-IGZO TFTs with an inverted staggered bottom-gate structure. (b) IDS–VGS transfer char- acteristics at VDS = 10 V with various Tacts. The IDS–VGS trans- fer characteristics of (c) thin-active device with Tact = 15 nm and (d) thick-active device with Tact = 100 nm. Insets: cross-sectional scanning electron microscope images of deposited a-IGZO film on the thermally grown SiO2. Tox = 50 nm and TS/D = 80 nm.
its performance [6]. Then, a Ti layer was deposited with an electron-beam evaporator and was patterned to form the S/D electrode with the thickness of S/D electrodes TS/D = 80 nm. Finally, the samples were annealed at 523 K for 1 h to improve their electrical properties.
The channel width (W ) and length (L) were 200 and 100 μm, respectively. The dc current–voltage (I–V ) and pulsed I–V characteristics were measured at room temperature under dark ambient using an Agilent 4156C precision semi- conductor parameter analyzer and a Keithley 4200 system, respectively. A technology computer-aided design (TCAD) simulation was carried out by using Atlas 2-D of Silvaco.
III. RESULTS AND DISCUSSION
The measured transfer characteristics, i.e., the current from drain to source (IDS) versus the applied gate-to-source voltage (VGS) at fixed drain-to-source voltage (VDS) = 10 V, with varying Tact are shown in Fig. 1(b). The detailed transfer characteristics at VDS = 0.5 and 10 V are shown in Fig. 1(c) [Tact = 15 nm] and Fig. 1(d) [Tact = 100 nm], respectively. The parameters of thin devices (Tact = 15 nm) are as follows: VT = 1.22 V, SS = 0.18 V/dec, ON-current (ION) = 5.08 μA, ON/OFF current ratio (ION/IOFF) = 1.50 × 106, and the field-effect mobility (μFE) = 9.74 cm2/Vs, while those of thick devices (Tact = 100 nm) are VT = −1.69 V, SS = 0.27 V/dec, ION = 4.70 μA, ION/IOFF = 0.83 × 106, and μFE = 7.80 cm2/Versus. It is clearly observed that both VT and ION increase as the active layer gets thinner, as shown in Fig. 1(b)–(d).
Although these tendencies are consistent with those in [7]–[10], it should be noted that the electrostatic and quantitative explanation of the active thickness dependence
of the transfer characteristic was insufficient in most of the previous studies.
In order to explain the case where the process and the material are the same and only Tact is different, it is reasonable to assume that the trap density per unit volume of IGZO film is the same regardless of Tact. However, a total amount of charged traps in the active layer will be larger if the active layer is thicker. Under the subthreshold regime, a channel layer that is difficult to control by VGS is formed easier especially from the bottom of the active layer as Tact becomes thicker. Therefore, the subthreshold current increases under the same VGS as Tact gets larger. If Tact is large, the total amount of charged traps in the active film is also large, so that the charge for compensating the electric field generated by VGS can be coped enough with localized charge, i.e., a voltage drop occurs across the entire depth direction of the active film, and the accumulation layer (conduction channel) would be formed only thinly on the surface between the IGZO active and GI. Accordingly, the accumulation layer can be formed on the surface of the active layer with a lower VGS. That is, VT is lowered with the increase of Tact as shown in Fig. 1(b)–(d).
On the other hand, in the above-mentioned threshold regime (the ON condition of TFT), the total amount of charged traps in the active layer decreases as the active thickness becomes thinner, so the charge for compensating the electric field generated by VGS must be covered by the accumulated charge by electrons. That is, the thinner the active layer is, the more the accumulation occurs in the active layer. This means that the thickness of the conduction channel in the ON-state becomes rather thick, which means that the channel resistance is lowered [8]. Therefore, as the active layer becomes thinner, ION increases as shown in Fig. 1(b)–(d). In other word, the surface accumulation condition changes into the volume accumulation condition as the active film gets thinner.
In order to verify the observed Tact dependences of VT , SS, and ION, the TCAD simulation was carried out as shown in Fig. 2.
The device structure and thermal conductivities of materials used in our simulation are shown in Fig. 2(a) [11]–[14]. The measured Tact dependence of transfer characteristic [Fig. 1(b)–(d)] is reproduced well by the device simulation [Fig. 2(b)], which suggests that the electron concentration in channel increases (decreases) at VGS sufficiently above VT
(below VT ) with thinner IGZO film. The measured IDS–VDS output characteristics at VGS =
VT + 15 V are shown in Fig. 3(a) and (b). While a current saturation is clearly observed at VDS < 20 V [VDS,max = 20 V in Fig. 3(a)], IDS becomes nonsaturated and continues increasing at VDS > 20 V [VDS,max = 40 V in Fig. 3(b)], where VDS,max means the maximum value of swept VDS. In the case of Tact = 15 nm, IDS again decreases at VDS > 30 V [Fig. 3(b)]. The Tact-dependent nonsaturated current observed is schematically illustrated in Fig. 3(c), where the region of IDS increase and IDS decrease is symbolized by I and II, respectively.
In order to figure out the origin of the nonsaturated IDS and its Tact dependence, the output curve of thin device
YU et al.: DEGRADATION ON CURRENT SATURATION OF OUTPUT CHARACTERISTICS IN a-IGZO TFTs 3245
Fig. 2. (a) Schematics illustrating the device structure and thermal conductivities [11]–[14] used in TCAD simulation. (b) Simulated transfer characteristics. (c) Schematic views of vertical and lateral energy band diagrams with indicating the position along the current path. (d) Simulated vertical and lateral distribution of electric field. The electric field increases with thinner Tact.
Fig. 3. Measured output characteristics of a-IGZO TFTs with different Tacts in the VDS sweep range. (a) From 0 to 20 V (VDS,max = 20 V) and
(b) from 0 to 40 V (VDS,max = 40 V) with VGS = VT +15 V. (c) Schematic illustrating the Tact-dependent output characteristic of a-IGZO TFT.
(Tact = 15 nm) is measured in the pulsed I–V mode, as shown in Fig. 4, with varying the conditions of VDS pulse applied at a fixed VGS = VT + 15 V. The conditions of VDS pulse, such as
Fig. 4. Measured pulsed IDS–VDS characteristic of thin devices with variations of (a) duty and (b) tON/tOFF of VDS pulse.
TABLE I MEASUREMENT CONDITION OF PULSED I–V WITH VARIATION OF DUTY
TABLE II MEASUREMENT CONDITION OF PULSED I–V
WITH VARIATION OF tON /tOFF
the duty, ON-time (tON), and OFF-time (tOFF), are controlled as summarized in Tables I and II. They are intended for modulating the extent to which a self-heating or Joule heating occurs because the device would be heated (cooled down) during tON (tOFF). Note that the time scale of each VDS step on the order of magnitude of milliseconds is the enough heating time [15], whereas the heat in IGZO film accumulated during the current-flowing period cannot be efficiently dissipated out of active film due to relatively low thermal conductivities of IGZO and SiO2 as shown in Fig. 2(a) [12], [13].
It is clearly observed in Fig. 4(a) that the degradation on IDS saturation in output characteristic gains prominence as the duty increases. This suggests that the nonsaturated IDS, i.e., the IDS increase followed by the IDS decrease, is attributed to the self-heating because it becomes more significant with the reduction of time for heat release (higher duty). On the other hand, the IDS increase dominates at longer tON and the IDS decrease becomes more activated at longer tOFF [see Fig. 4(b)]. The temperature in the channel is greatly increased because of the longer heat generation time; while in the latter case, the temperature change is relatively small since there is sufficient time to release the heat.
3246 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO. 8, AUGUST 2018
Therefore, the IDS increase at high VDS [indicated by I in Fig. 3(c)] originates from self-heating in the IGZO active layer. Self-heating not only enhances the carrier mobil- ity in the trap-limited conduction regime but lowers VT as well [16]–[18], which results in the IDS increase. Our explana- tion is validated in a TCAD-based device simulation as shown in Fig. 2(b). In the simulation results, the higher IDS of the thin device is consistent with more significant self-heating and reproduces more activated IDS increase compared with that of the thick device [Fig. 3(c)].
On the other hand, when the electrons gain sufficient kinetic energy, they inject into the GI as illustrated in Fig. 2(c). This injection by thermionic field emission and Fowler–Nordheim tunneling occurs especially around the drain, along the path (1) → (2) → (3) in Fig. 2(c), and the trapped electrons increase VT and consequently decrease IDS as VDS increases. Thus, as VDS increases further, the consecutive IDS decrease [indicated by II in Fig. 3(c)] is observed because electrons become trapped in the GI and/or interface. The simulation results in Fig. 2(d) suggest that the vertical and/or lateral electric field becomes higher and the electrons gain more kinetic energy in the thin device than in the thick device. Therefore, the electron trapping, i.e., the IDS decrease, would begin to dominantly occur at lower VDS as the IGZO active layer becomes thinner [see Fig. 3(b) and (c)].
More in detail, among four tON/tOFF conditions such as 0.1/0.9, 0.9/0.1, 2/0.5, and 0.5/2 ms, the critical VDS, i.e., VDS,crt (the VDS value at which the current ceases to saturate and begins to rise) is ∼20 V (tON/tOFF = 0.9/0.1 ms), ∼23 V (tON/tOFF = 2/0.5 ms), and ∼27 V (tON/tOFF = 0.1/0.9 and/or 0.5/2 ms) as shown in Fig. 4. Here, the IDS decrease is prominently observed at VDS < 40 V only when tON/tOFF = 0.9/0.1 ms while the IDS decrease is not observed at VDS < 40 V only when tON/tOFF = 2/0.5 ms. Compared with cases of tON/tOFF = 0.9/0.1 and 2/0.5 ms, the extent to which IDS becomes nonsaturated is relatively weak in cases of tON/tOFF = 0.1/0.9 and 0.5/2 ms. In addition, the IDS decrease occurs very weakly when tON/tOFF = 0.5/2 ms.
Therefore, at tON > tOFF, the longer tON is compared to tOFF, the current ceases to saturate and starts to rise at the lower VDS (VDS,crt becomes lower). Also, as the IDS increase begins at the lower VDS, the IDS decrease begins again at the lower VDS value. This means that the larger the degrees of mobility enhancement and VT lowering are, the more active the charge trapping phenomenon is. This is evidence that charge trapping is also related to self-heating. In addition, VDS,crt decreases sensitively as the tON/tOFF ratio increases.
On the other hand, when tON<tOFF, the degradation of current saturation is relatively less. That is, VDS,crt further increases. In the case of tON<tOFF, compared with tON>tOFF, the IDS increase is weak, the IDS decrease is also weak, and VDS,crt is relatively insensitive to the tON/tOFF ratio. Furthermore, the longer the duration of tON is, the IDS decrease due to charge trapping is observed at the lower VDS.
In Figs. 3 and 4, we emphasize that charge trapping can be accelerated by the self-heating associated with thermionic-field emission [19], [20]. As a result, in a structure with a high
Fig. 5. (a) Measured output characteristics of a-IGZO TFTs with different Tacts at VGS = VT + 15 V. (b) Changes of VDS,crt and VDS,dec with the variation of Tact.
current flow, heat generation due to power consumption is enhanced, and thereby, electron trapping is promoted.
In order to investigate the active thickness dependence of the current saturation degradation more in detail, both VDS,crt and VDS,dec, i.e., the VDS value at which the current begins to decrease in region II, are taken for the cases of Tact = 15, 45, 70, and 100 nm as shown in Fig. 5. The parameters are as follows: VDS,crt = 20.4 V (Tact = 15 nm), 21.4 V (Tact = 45 nm), 29.8 V (Tact = 70 nm), and 30.4 V (Tact = 100 nm), VDS,dec = 35 V (Tact = 15 nm), and 35.7 V (Tact = 45 nm), respectively. These results suggest that our explanation of the Tact dependence of current saturation degradation is reasonable and reproducible.
In addition, in order to determine whether the electron trap- ping happens locally near the drain region, we switch the loca- tion of S/D to each other after sweeping VDS up to VDS,max. In the transfer characteristics shown in Fig. 6(a) and (b), the forward mode means the same S/D location as that when sweeping VDS up to VDS,max, whereas the reverse mode corresponds to switching the location of S/D to each other compared with the forward mode. We also confirm the transfer characteristics in the linear and saturation regions by reading out IDS by setting the constant VDS to 0.5 and 10 V. In the thin device, VT shifts in the positive direction when VDS,max = 40 V (max 40), but remains unchanged compared with the initial (Init.) VT when VDS,max = 20 V (max 20) [Fig. 6(c)]. Therefore, it is certain that the IDS decrease in output characteristic results from the electron trapping followed by the VT increase. Furthermore, it is noteworthy that VT slightly increases only when the forward mode is carried out at VDS = 10 V [Fig. 6(c)]. This suggests that the electrons are locally trapped in GI near the drain rather than the source, as illustrated by energy band diagram with different energy barriers in Fig. 6(e). Among the read-out conditions, the energy barrier is lowered only under the condition of VDS = 10 V in the forward mode, resulting in a relatively small VT compared to others. In contrast, VT lowers with increasing VDS,max in the thick device, as shown in Fig. 6(d). It validates that self-heating followed by VT lowering causes the abnormal IDS increase in the output characteristic [16].
In order to investigate the influence of device size on the degradation of current saturation, the output curves are also measured with varying W and L, as shown in Fig. 7(a).
YU et al.: DEGRADATION ON CURRENT SATURATION OF OUTPUT CHARACTERISTICS IN a-IGZO TFTs 3247
Fig. 6. Measured transfer characteristics of the forward and reverse modes at initial state and after observing output characteristics with conditions of Fig. 3 of (a) thin and (b) thick devices. (c) VT taken from (a) in the thin device. (d) VT taken from (b) in the thick device. (e) Schematics illustrating the lateral electric potential profiles according to read-out mode, i.e., forward and reverse.
The slopes (dIDS/dVDS) are extracted from the output char- acteristics to indicate the extent to which the nonsaturated current phenomena happen and the maximum value among their magnitudes, i.e., (dIDS/dVDS)max, is taken in the IDS decrease (II) as well as the IDS increase (I) regions, respec- tively, as seen in Fig. 7(b). Extracted (dIDS/dVDS)max is summarized as the functions of W and L, as shown in Fig. 7(c). (dIDS/dVDS)max in region I becomes larger as either W increases or L decreases. This suggests that the IDS increase in output characteristics is observed more prominently as IDS itself increases, which is strongly reminiscent of Joule heating.
On the other hand, (d IDS/dVDS)max in region II becomes larger as the device size, i.e., W × L, increases, which is strongly correlated with self-heating-assisted electron trap- ping because a total heat accumulated in an active layer with low thermal conductivity increases with increasing the device size [12], [19]. The W and L dependences of the VT shift (VT ) after read-out of the output characteristics at VDS,max = 40 V are also shown in Fig. 7(d). VT is the largest at the case of maximum W × L value, which suggests VT
is also correlated with self-heating. In Fig. 7(c), it is noteworthy that the case
of W × L = 50 × 150 μm2 is not applicable to the discussion on the device size dependence of (dIDS/dVDS)max in region II
Fig. 7. (a) Device size dependence of the output character- istics. (b) Enlarged schematic describing the defined parameter (dIDS/dVDS)max. (c) Extracted (dIDS/dVDS)max from the output charac- teristics shown in (a). (d) ΔVT taken from the read-out transfer curves at initial state and after sweeping output curve in devices with different sizes (W/L).
because in the case of device with W /L = 50/150 μm, any IDS decrease is not observed after read-out of the output characteristics at VDS,max = 40 V, i.e., (dIDS/dVDS)max in region II ∼ 0.
Furthermore, in Fig. 7(d), it is clearly observed that VT
after read-out of the output characteristics at VDS,max = 40 V is the largest in W /L = 200/150 μm among W /L = 200/100, 200/150, and 50/150 μm. VT in W /L = 200/150 μm is larger than that in W /L = 200/100 μm although ION is lower in W /L = 200/150 μm than that in W /L = 200/100 μm, which suggests that the W × L value as well as W /L value affects VT .
In order for the IDS increase (negative VT ) to occur, and then the IDS decrease (positive VT ) to be observed, it is necessary to compensate the mobility enhancement and VT
lowering by the charge trapping. However, since the device with W/L = 50/150 μm has little IDS increase itself, there is no negative VT to compensate. Therefore, the larger VT
in W /L = 50/150 μm than that in W /L = 200/100 μm can be observed, as shown in Fig. 7(d).
Finally, it needs to be discussed whether a different active layer thickness would have a possibility to cause the heat dissipation difference. If ION is independent of the active thickness Tact, a total heat generated in an active film can be roughly proportional to a total volume of active film, i.e., W × L × Tact. This total volume would strongly affect the heat dissipation. However, as mentioned earlier, ION increases as Tact becomes thinner. Thus, the self-heating definitely becomes more prominent as Tact becomes thinner, which makes the Tact aspect simpler. It should be noted that from the viewpoint of Tact, the self-heating is dominated by a
3248 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO. 8, AUGUST 2018
high ION rather than by the surface area or the volume of heat dissipation.
If ION decreases as Tact becomes thinner, the thickness aspect may become very complicated because the Joule heat- ing and the heat dissipation will move in an opposite direction. However, it is not our case. Quantitative analysis is necessary as further study.
Consequently, the factors that determine nonsaturation in output characteristics are largely related to the dimensions of the device structure, which affect the current, the area in which heat is generated, and the electric field. These comprehensive measured and simulation results support two mechanisms on the nonsaturated current in output characteris- tics of IGZO TFTs under high current flowing conditions and their dependence on the dimensions. The first mechanism is Joule heating followed by VT lowering and mobility enhance- ment, and the second is self-heating-assisted (thermionic field emission and Fowler–Nordheim tunneling) electron trapping in the GI near the drain region followed by VT increase.
IV. CONCLUSION
Degradation on the current saturation in output character- istics in IGZO TFTs is observed and its origins are exper- imentally analyzed and verified by TCAD simulation. The influence of IGZO active thickness on this nonsaturation is also investigated. IDS becomes nonsaturated and gradually increases at a high VDS and then decreases at a higher VDS when the output curve is measured. IDS increases due to Joule-heating followed by VT lowering and mobility enhance- ment, which gains prominence as the active layer becomes thinner. These phenomena were also more activated as either W increases or L decreases. In addition, the IDS decrease results from the self-heating-assisted electron trapping into GI near the drain region followed by VT increase and gains significance as the active layer becomes thinner and as the device size, i.e., W × L, increases.
Our results suggest that the nonsaturated output charac- teristics and their related mechanisms need to be carefully considered as main reliability issues when designing ana- log circuitry or high-current driving applications based on IGZO TFTs where a robust current saturation and large output resistance are critical.
REFERENCES
[1] J. C. Park et al., “Highly stable transparent amorphous oxide semicon- ductor thin-film transistors having double-stacked active layers,” Adv. Mater., vol. 22, no. 48, pp. 5512–5516, Dec. 2010, doi: 10.1002/adma. 201002397.
[2] K. Takei, “High performance, flexible CMOS circuits and sensors toward wearable healthcare applications,” in IEDM Tech. Dig., Dec. 2016, pp. 6.1.1–6.1.4, doi: 10.1109/IEDM.2016.7838358.
[3] P. Heremans et al., “Flexible metal-oxide thin film transistor cir- cuits for RFID and health patches,” in IEDM Tech. Dig., Dec. 2016, pp. 6.3.1–6.3.4, doi: 10.1109/IEDM.2016.7838360.
[4] M. Mativenga, S. Hong, and J. Jang, “High current stress effects in amorphous-InGaZnO4 thin-film transistors,” Appl. Phys. Lett., vol. 102, no. 2, p. 023503, Jan. 2013, doi: 10.1063/1.4775694.
[5] J.-M. Lee, I.-T. Cho, J.-H. Lee, and H.-I. Kwon, “Bias-stress-induced stretched-exponential time dependence of threshold voltage shift in InGaZnO thin film transistors,” Appl. Phys. Lett., vol. 93, no. 9, pp. 093504-1–093504-3, 2008, doi: 10.1063/1.2977865.
[6] J.-S. Park, J. K. Jeong, Y.-G. Mo, H. D. Kim, and S.-I. Kim, “Improve- ments in the device characteristics of amorphous indium gallium zinc oxide thin-film transistors by Ar plasma treatment,” Appl. Phys. Lett., vol. 90, pp. 262106-1–262106-3, Jun. 2007, doi: 10.1063/1.2753107.
[7] C.-S. Hwang et al., “Effects of active thickness in oxide semicon- ductor TFTs,” in SID Int. Symp. Dig. Tech. Papers, vol. 40, 2009, pp. 1107–1109, doi: 10.1889/1.3256478.
[8] Y. W. Jeon et al., “Subgap density-of-states-based amorphous oxide thin film transistor simulator (DeAOTS),” IEEE Trans. Electron Devices, vol. 57, no. 11, pp. 2988–3000, Nov. 2010, doi: 10.1109/TED.2010. 2072926.
[9] D. Kong et al., “P-202L: Late-News Poster: Density-of-states based analysis on the effect of active thin-film thickness on current stress- induced instability in amorphous InGaZnO AMOLED driver TFTs,” in SID Int. Symp. Dig. Tech. Papers, vol. 42, 2011, pp. 1223–1226, doi: 10.1889/1.3621052.
[10] D. Kong et al., “The effect of the active layer thickness on the negative bias stress-induced instability in amorphous InGaZnO thin-film transistors,” IEEE Electron Device Lett., vol. 32, no. 10, pp. 1388–1390, Oct. 2011, doi: 10.1109/LED.2011.2161746.
[11] D. Setoyama et al., “Thermal properties of titanium hydrides,” J. Nucl. Mater., vol. 344, nos. 1–3, pp. 298–300, Sep. 2005, doi: 10.1016/j. jnucmat.2005.04.059.
[12] T. Yoshikawa et al., “Thermal conductivity of amorphous indium– gallium–zinc oxide thin films,” Appl. Phys. Express, vol. 6, no. 2, p. 021101, Jan. 2013, doi: 10.7567/APEX.6.021101.
[13] T. Yamane, N. Nagai, S. Katayama, and M. Todoki, “Measurement of thermal conductivity of silicon dioxide thin films using a 3ω method,” J. Appl. Phys., vol. 91, no. 12, pp. 9772–9776, May 2002, doi: 10.1063/ 1.1481958.
[14] H. R. Shanks, P. D. Maycock, P. H. Sidles, and G. C. Danielson, “Ther- mal conductivity of silicon from 300 to 1400 K,” Phys. Rev., vol. 130, no. 5, pp. 1743–1748, Jun. 1963, doi: 10.1103/PhysRev.130.1743.
[15] K.-H. Liu et al., “Investigation of channel width-dependent threshold voltage variation in a-InGaZnO thin-film transistors,” Appl. Phys. Lett., vol. 104, no. 13, p. 133503, Mar. 2014, doi: 10.1063/1.4868430.
[16] S. Lee et al., “Temperature dependent electron transport in amor- phous oxide semiconductor thin film transistors,” in IEDM Tech. Dig., Dec. 2011, pp. 14.6.1–14.6.4, doi: 10.1109/IEDM.2011.6131554.
[17] J. I. Kim et al., “Thermoreflectance microscopy analysis on self- heating effect of short-channel amorphous In-Ga-Zn-O thin film tran- sistors,” Appl. Phys. Lett., vol. 105, no. 4, p. 043501, Jul. 2014, doi: 10.1063/1.4891644.
[18] M. Fujii et al., “Thermal analysis of degradation in Ga2O3–In2O3–ZnO thin-film transistors,” Jpn. J. Appl. Phys., vol. 47, no. 8R, pp. 6236–6240, Aug. 2008, doi: 10.1143/JJAP.47.6236.
[19] T.-C. Chen et al., “Self-heating enhanced charge trapping effect for InGaZnO thin film transistor,” Appl. Phys. Lett., vol. 101, no. 4, p. 042101, 2012, doi: 10.1063/1.4733617.
[20] T.-Y. Hsieh et al., “Self-heating-effect-induced degradation behaviors in a-InGaZnO thin-film transistors,” IEEE Electron Device Lett., vol. 34, no. 1, pp. 63–65, Jan. 2013, doi: 10.1109/LED.2012.2223654.
Hye Ri Yu received the B.S. degree in electri- cal engineering from Kookmin University, Seoul, South Korea, in 2017, where she is currently pursuing the M.S. degree with the Department of Electrical Engineering.
Jun Tae Jang received the B.S. and M.S. degrees in electrical engineering from Kookmin University, Seoul, South Korea, in 2016, where he is currently pursuing the Ph.D. degree with the Department of Electrical Engineering.
YU et al.: DEGRADATION ON CURRENT SATURATION OF OUTPUT CHARACTERISTICS IN a-IGZO TFTs 3249
Daehyun Ko received the B.S. and M.S. degrees in electrical engineering from Kookmin University, Seoul, South Korea, in 2016 and 2018, respectively.
He is currently an Analog and Digital Circuit Design Engineer at Silicon Works, Seoul.
Sungju Choi received the B.S. and M.S. degrees in electrical engineering from Kookmin University, Seoul, South Korea, in 2016, where he is currently pursuing the Ph.D. degree with the Department of Electrical Engineering.
Geumho Ahn received the B.S. degree in electrical engineering from Kookmin University, Seoul, South Korea, in 2017, where he is cur- rently pursuing the M.S. degree with the Depart- ment of Electrical Engineering.
Sung-Jin Choi received the M.S. and Ph.D. degrees in electrical engineering from the Korea Advanced Institute of Science and Technology, Daejeon, South Korea, in 2012.
He is currently an Assistant Professor with the School of Electrical Engineering, Kookmin University, Seoul, South Korea.
Dong Myong Kim (S’86–M’88) received the B.S. (magna cum laude) and M.S. degrees in electronics engineering from Seoul National Uni- versity, Seoul, South Korea, in 1986 and 1988, respectively, and the Ph.D. degree in electrical engineering from the University of Minnesota Twin Cities, Minneapolis, MN, USA, in 1993.
He is currently a Professor with the School of Electrical Engineering, Kookmin University, Seoul.
Dae Hwan Kim (M’08–SM’12) received the B.S., M.S., and Ph.D. degrees in electrical engineer- ing from Seoul National University, Seoul, South Korea, in 1996, 1998, and 2002, respectively.
He is currently a Professor with the School of Electrical Engineering, Kookmin University, Seoul. His current research interests include nanoCMOS, oxide and organic thin-film transis- tors, biosensors, and neuromorphic devices.
<< /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Gray Gamma 2.2) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Warning /CompatibilityLevel 1.4 /CompressObjects /Off /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.0000 /ColorConversionStrategy /sRGB /DoThumbnails true /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams true /MaxSubsetPct 100 /Optimize true /OPM 0 /ParseDSCComments false /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo false /PreserveFlatness true /PreserveHalftoneInfo true /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts false /TransferFunctionInfo /Remove /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true /Arial-Black /Arial-BoldItalicMT /Arial-BoldMT /Arial-ItalicMT /ArialMT /ArialNarrow /ArialNarrow-Bold /ArialNarrow-BoldItalic /ArialNarrow-Italic /ArialUnicodeMS /BookAntiqua /BookAntiqua-Bold /BookAntiqua-BoldItalic /BookAntiqua-Italic /BookmanOldStyle /BookmanOldStyle-Bold /BookmanOldStyle-BoldItalic /BookmanOldStyle-Italic /BookshelfSymbolSeven /Century /CenturyGothic /CenturyGothic-Bold /CenturyGothic-BoldItalic /CenturyGothic-Italic /CenturySchoolbook /CenturySchoolbook-Bold /CenturySchoolbook-BoldItalic /CenturySchoolbook-Italic /ComicSansMS /ComicSansMS-Bold /CourierNewPS-BoldItalicMT /CourierNewPS-BoldMT /CourierNewPS-ItalicMT /CourierNewPSMT /EstrangeloEdessa /FranklinGothic-Medium /FranklinGothic-MediumItalic /Garamond /Garamond-Bold /Garamond-Italic /Gautami /Georgia /Georgia-Bold /Georgia-BoldItalic /Georgia-Italic /Haettenschweiler /Impact /Kartika /Latha /LetterGothicMT /LetterGothicMT-Bold /LetterGothicMT-BoldOblique /LetterGothicMT-Oblique /LucidaConsole /LucidaSans /LucidaSans-Demi /LucidaSans-DemiItalic /LucidaSans-Italic /LucidaSansUnicode /Mangal-Regular /MicrosoftSansSerif /MonotypeCorsiva /MSReferenceSansSerif /MSReferenceSpecialty /MVBoli /PalatinoLinotype-Bold /PalatinoLinotype-BoldItalic /PalatinoLinotype-Italic /PalatinoLinotype-Roman /Raavi /Shruti /Sylfaen /SymbolMT /Tahoma /Tahoma-Bold /TimesNewRomanMT-ExtraBold /TimesNewRomanPS-BoldItalicMT /TimesNewRomanPS-BoldMT /TimesNewRomanPS-ItalicMT /TimesNewRomanPSMT /Trebuchet-BoldItalic /TrebuchetMS /TrebuchetMS-Bold /TrebuchetMS-Italic /Tunga-Regular /Verdana /Verdana-Bold /Verdana-BoldItalic /Verdana-Italic /Vrinda /Webdings /Wingdings2 /Wingdings3 /Wingdings-Regular /ZWAdobeF ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 150 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 600 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages false /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 600 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 400 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /CreateJDFFile false /Description << /CHS <FEFF4f7f75288fd94e9b8bbe5b9a521b5efa7684002000410064006f006200650020005000440046002065876863900275284e8e55464e1a65876863768467e5770b548c62535370300260a853ef4ee54f7f75280020004100630072006f0062006100740020548c002000410064006f00620065002000520065006100640065007200200035002e003000204ee553ca66f49ad87248672c676562535f00521b5efa768400200050004400460020658768633002> /CHT <FEFF4f7f752890194e9b8a2d7f6e5efa7acb7684002000410064006f006200650020005000440046002065874ef69069752865bc666e901a554652d965874ef6768467e5770b548c52175370300260a853ef4ee54f7f75280020004100630072006f0062006100740020548c002000410064006f00620065002000520065006100640065007200200035002e003000204ee553ca66f49ad87248672c4f86958b555f5df25efa7acb76840020005000440046002065874ef63002> /DAN <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> /DEU <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> /ESP <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> /FRA <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> /ITA (Utilizzare queste impostazioni per creare documenti Adobe PDF adatti per visualizzare e stampare documenti aziendali in modo affidabile. I documenti PDF creati possono essere aperti con Acrobat e Adobe Reader 5.0 e versioni successive.) /JPN <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> /KOR <FEFFc7740020c124c815c7440020c0acc6a9d558c5ec0020be44c988b2c8c2a40020bb38c11cb97c0020c548c815c801c73cb85c0020bcf4ace00020c778c1c4d558b2940020b3700020ac00c7a50020c801d569d55c002000410064006f0062006500200050004400460020bb38c11cb97c0020c791c131d569b2c8b2e4002e0020c774b807ac8c0020c791c131b41c00200050004400460020bb38c11cb2940020004100630072006f0062006100740020bc0f002000410064006f00620065002000520065006100640065007200200035002e00300020c774c0c1c5d0c11c0020c5f40020c2180020c788c2b5b2c8b2e4002e> /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken waarmee zakelijke documenten betrouwbaar kunnen worden weergegeven en afgedrukt. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR <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> /PTB <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> /SUO <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> /SVE <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> /ENU (Use these settings to create PDFs that match the "Required" settings for PDF Specification 4.0) >> >> setdistillerparams << /HWResolution [600 600] /PageSize [612.000 792.000] >> setpagedevice