Logo Beilstein Institut

Exploring the photoleakage current and photoinduced negative bias instability in amorphous InGaZnO thin-film transistors with various active layer thicknesses

  1. 1 ORCID Logo and
  2. 2,3 ORCID Logo
1Key Laboratory of Applied Surface and Colloid Chemistry, Ministry of Education; Shaanxi Key Laboratory for Advanced Energy Devices; Shaanxi Engineering Lab for Advanced Energy Technology, School of Materials Science and Engineering, Shaanxi Normal University, Xi’an 710119, China
2School of Environmental Science and Engineering, Kochi University of Technology, Kami, Kochi 782-8502, Japan
3Center for Nanotechnology in Research Institute, Kochi University of Technology, Kami, Kochi 782-8502, Japan
  1. Corresponding author email
Associate Editor: E. Meyer
Beilstein J. Nanotechnol. 2018, 9, 2573–2580. https://doi.org/10.3762/bjnano.9.239
Received 09 May 2018, Accepted 07 Sep 2018, Published 26 Sep 2018
Full Research Paper
cc by logo
PDF
Album
Cite

Abstract

The photoleakage current and the negative bias and illumination stress (NBIS)-induced instability in amorphous InGaZnO thin-film transistors (a-IGZO TFTs) with various active layer thicknesses (TIGZO) were investigated. The photoleakage current was found to gradually increase in a-IGZO TFTs irrespective of the TIGZO when the photon energy of visible light irradiation exceeded ≈2.7 eV. Furthermore, the influence of the TIGZO on NBIS-induced instability in a-IGZO TFTs was explored by the combination of current–voltage measurements in double-sweeping VGS mode and capacitance–voltage measurements. The NBIS-induced hysteresis was quantitatively analyzed using a positive gate pulse mode. When the TIGZO was close to the Debye length, the trapped electrons at the etch-stopper/IGZO interface, the trapped holes at the IGZO/gate insulator interface, and the generation of donor-like states in an a-IGZO layer were especially prominent during NBIS.

Keywords: active layer thickness; gate bias; illumination stress; InGaZnO; photoleakage current; thin-film transistors

Introduction

Over the last decade, the amorphous oxide-based semiconductor thin-film transistors (AOS TFTs) have attracted global attention for use in advanced display technologies due to their outstanding properties such as high electron mobility, good transparency to visible light, and low process temperature with good uniformity [1-4]. Among the numerous AOS materials, indium gallium zinc oxide (IGZO) is one of the most promising candidates used as the active layer because of its excellent electrical and optical properties [5-8]. Although the band gap of IGZO (≈3.1 eV) is higher than the photon energy of visible light, photoinduced leakage current under visible-light irradiation can be detected in the oxide-based TFTs [9,10]. This is due to the fact that the electrons are excited from the trapped states existing near the valence band (EV). In addition, the a-IGZO TFTs inevitably suffer electrical and optical stresses during practical operation conditions, especially for the negative bias and illumination stress (NBIS) tests [11-16], which leads to device instability and restricts the development of oxide TFTs for commercial products.

In our previous study, a double-sweeping VGS mode was proposed to investigate the origin of NBIS-induced hysteresis of a-IGZO TFTs [17]. A promising method to suppress NBIS degradation was also considered by applying a large negative VDS bias of VDS < VGS during NBIS [18]. These studies imply that the fabrication parameters for the active layer should be well taken into account to improve the reliability of oxide TFTs. The active layer thickness is a key parameter to modify the performance of a-IGZO TFTs. Some works have highlighted that the electrical properties of the device (for both the initial and after stress conditions) such as threshold voltage, on/off ratio, and field effect mobility, can be effectively adjusted by controlling the active layer thickness [19-23]. Up to now, the impact of the active layer thickness (TIGZO) on the photoleakage current and NBIS-induced instability in a-IGZO TFTs has been rarely reported. The NBIS-induced degradation of a-IGZO TFTs with various active layer thicknesses has also rarely been discussed.

In this study, a-IGZO films with various active layer thicknesses were prepared by magnetron sputtering. The initial electrical properties and the photoleakage current of a-IGZO TFTs with various active layer thicknesses were investigated. The subthreshold value slightly increased while the threshold voltage (Vth) and mobility (μ) decreased with increasing TIGZO. The photoleakage current increased in all TFTs when the wavelength of visible-light irradiation was shorter than 460 nm. Moreover, the photoleakage current increased with an increase in the TIGZO. Furthermore, the impact of the active layer thickness on the NBIS-induced instability in a-IGZO TFTs was explored by combining the current–voltage (I–V) measurements in double-sweeping VGS mode and capacitance–voltage (C–V) measurements. The NBIS-induced hysteresis was quantitatively analyzed using a positive gate pulse mode. The I–V and C–V results revealed that the trapped holes at the etch-stopper/IGZO interface, the trapped holes at the IGZO/gate insulator interface, and the generation of donor-like states in a-IGZO layer were particularly prominent after NBIS tests when the active layer thickness was close to the Debye length.

Experimental

A schematic cross-sectional view of a bottom-gate a-IGZO TFT is shown in Figure 1a. The detailed fabrication procedure for the a-IGZO TFT was described in our previous publication [24]. After the fabrications of a chromium (Cr) gate electrode and a SiOx gate insulator (150 nm), the a-IGZO layer with thicknesses of 25, 45, 75, and 100 nm respectively were deposited at 160 °C from a sintered IGZO ceramic target by DC magnetron sputtering with a mixed gas of Ar/O2 = 29.4/0.6 sccm at a deposition pressure of 1 Pa. After patterning of the IGZO films as the active channel, a SiOx etch-stopper (200 nm), source and drain electrodes, and a 200 nm-thick SiOx passivation layer were sequentially formed. Following the preparation of a-IGZO TFTs, all devices were annealed in N2 environment at 350 °C for 1 h before the electrical measurements. The channel width (W) and length (L) of the IGZO TFTs were 50 and 20 μm, respectively. All of the I–V characteristics were measured using an Agilent 4156C precision semiconductor parameter analyzer.

[2190-4286-9-239-1]

Figure 1: (a) Schematic cross-sectional view and (b) the initial transfer characteristics of a-IGZO TFTs with various active layer thicknesses (TIGZO) measured at VDS = 20.1 V.

For the photoleakage current test, monochromatic light irradiation was supplied by a Xe lamp with a band pass filter (FWHM of 10 nm) at an intensity of 0.2 mW/cm2. The wavelength of the light was in the range of 400–530 nm and was introduced to the back-channel side of a-IGZO TFTs. For the NBIS test, blue light with a wavelength of 460 nm and a gate voltage (VGS) of −30 V was simultaneously applied to all devices during the stress test. The NBIS was interrupted briefly when the transfer characteristics were measured in double-sweeping VGS mode in darkness at VDS = 0.1 V, and then NBIS was reapplied up to a stress time of 104 s. In terms of the double-sweeping VGS mode, the transfer characteristics were measured with VGS = −10–20 V (denoted hereafter as forward measurement), and then scanned instantly back to −10 V (denoted hereafter as reverse measurement). The C–V measurements for the channel capacitance were measured at 1 kHz and an AC level of 100 mV. All of the measurements were carried out at room temperature in ambient air.

Results and Discussion

The initial transfer characteristics (IDS–VGS) of a-IGZO TFTs with various active layer thicknesses (TIGZO) measured at VDS = 20.1 V are shown in Figure 1b. Table 1 summarizes the electrical properties, such as field effect mobility in the saturation region (μsat), threshold voltage Vth (VGS at IDS of 1 nA), hysteresis of the transfer curves (the difference of VGS at IDS of 1 pA between the forward and reverse sweeps), subthreshold swing (SS = dVGS/dlog10(IDS)), and the maximum area density of state (Nt).

Table 1: The initial electrical properties of a-IGZO TFTs with various active layer thicknesses (TIGZO).

Thickness, TIGZO (nm) 25 45 75 100
μsat (cm2∙V−1∙s−1) 13.61 13.47 12.43 12.20
VGS at IDS = 1 nA (V) 3.35 1.24 1.16 0.74
hysteresis ΔVH (V) 0.47 0.42 0.35 0.30
subthreshold swing (mV/dec.) 264 273 316 352
trap density (1011 cm−2) 5.06 5.29 6.35 7.24

Compared to a previous publication [25], the electrical properties of a-IGZO TFTs with various TIGZO exhibit the identical tendency. It is suggested that the devices exhibit great repeatability for the same kind of material under the same fabrication process. In addition, the Vth results demonstrate that the free carrier numbers in the bulk of the active layer are gradually increased with increasing TIGZO. Moreover, since the fabrication condition for the IGZO films are exactly identical, except the deposition duration, the results suggest that the variation in the SS value mainly originates from the density of defect states in the active layer. Correspondingly, the obtained results indicate that the increase in the Nt majorly stems from the increase of the IGZO bulk traps because of the identical a-IGZO/GI interfaces.

Figure 2a–d shows the variation in the transfer characteristics of a-IGZO TFTs with various TIGZO under monochromatic light irradiation at the excitation wavelengths of 400, 430, 460, 490, and 530 nm. The transfer characteristics measured in the dark are also shown as a reference. The transfer characteristics of all devices are measured with a VDS of 10.1 V and a VGS scanned from the on-to-off direction. It is found that the photoleakage current increases in all TFTs when the irradiation wavelength is shorter than 460 nm. In addition, the photoleakage current increases with increasing TIGZO. Figure 2e exhibits the photoleakage current of a-IGZO TFTs with various TIGZO as a function of the photon energy of incident light. When the photon energy exceeds ≈2.7 eV (460 nm), the photoleakage current starts to increase and increases gradually with increasing photon energy. Note that the photoleakage current increases dramatically in the photon energy range of >2.7 eV for the TFT with the thicker TIGZO. These results indicate that the electrons are excited from the trapped states existing near the valence band (EV) to the conduction band (EC) even though the photon energy is smaller than the band gap of IGZO. In terms of the a-IGZO material, the high-density electron traps exist at (EC − E) of over 2.7 eV [9], which affect the photoleakage current of a-IGZO TFTs. The total amount of trapped electrons increase with an increase in the TIGZO. The oxygen-related defects, such as oxygen vacancies (VO), may be the origin of high-density electron traps near the EV in a-IGZO TFTs, which occupy the region near the valence band maximum with an energy width of ≈1.5 eV [26,27].

[2190-4286-9-239-2]

Figure 2: Variation in the transfer characteristics of a-IGZO TFTs with (a) TIGZO = 25, (b) 45, (c) 75, and (d) 100 nm under monochromatic light irradiation, and (e) photoleakage current (VGS = −10 V and VDS = 10 V) of a-IGZO TFTs with various thicknesses as a function of photon energy of the incident light.

Figure 3a–d shows the variation in the transfer characteristics of a-IGZO TFTs with various TIGZO under NBIS for the forward measurements. It is found that for the NBIS duration of 1000 s, the transfer curves of all TFTs shift in the negative VGS direction without SS degradation. When the NBIS duration exceeds 1000 s, a shift in the positive VGS direction as well as the appearance of a hump with SS degradation in the transfer curves is observed, and this phenomenon gradually increases with increasing NBIS duration. It is noted that the phenomenon, combined the positive shift and the hump effect, is weakened as the TIGZO increases. For the reverse measurements, as shown in Figure 3e–h, the transfer curves of all TFTs shift parallel in the positive VGS direction without SS degradation during the NBIS duration of >1000 s. Noticeably, the abnormal phenomenon of hump appearance observed in the forward measurements is hardly observed in the reverse measurements. The positive shift of Vth without SS degradation is well fitted to the commonly used stretched-exponential equation [28]. The obtained results suggest that electron trapping at the back-channel interface between a-IGZO and etch-stopper layers occurs because a negative gate bias is performed during NBIS.

[2190-4286-9-239-3]

Figure 3: Variation in the transfer characteristics in the forward measurement for a-IGZO TFTs with (a) TIGZO = 25, (b) 45, (c) 75, and (d) 100 nm, and in the reverse measurement with (e) TIGZO = 25, (f) 45, (g) 75, and (h) 100 nm as a function of the stress duration for 104 s under −30 V VGS NBIS. Transfer characteristics of a-IGZO TFTs with (i) TIGZO = 25, (j) 45, (k) 75, and (l) 100 nm after 104 s NBIS without and with a gate pulse of 1 ms width and 10 V high.

On the basis of the photoleakage current results, when the photon energy of the light irradiation exceeds ≈2.7 eV, the photoleakage current of TFTs increases. In this study, a photon energy of ≈2.7 eV is set for the incident light. The previous publication indicates that a photon energy of ≈2.0 and ≈2.3 eV is required for the transition from VO to VO+ and VO2+, respectively. Moreover, the ionized oxygen vacancies of VO+ and VO2+ are located near the mid-gap and the bottom of the EC [15,29], respectively. It is sufficient to excite high-density VO defects to VO+/VO2+ and then to generate free electrons to EC. Simultaneously, the electron–hole pairs are photoexcited from EV, which leads to the neutralization between the ionized VO+/VO2+ and the generated electrons, contributing to free holes in EV [18]. During the NBIS duration with VGS = −30 V, a vertical electric field is exerted along the growth direction of the active layer. In general, the electric potential exponentially declines inside the active layer and has a maximum transfer length called the Debye length. In terms of a-IGZO TFT, a Debye length of ≈40 nm is calculated based on a previous publication [30]. In case of a-IGZO TFT with the TIGZO = 25 nm, the channel layer is totally depleted under the negative VGS bias since the TIGZO is less than the Debye length. Therefore, the photoexcited electrons and holes will be respectively accumulated and trapped at the IGZO/etch-stopper and the GI/IGZO interfaces. Meanwhile, the defect states are generated, which originate from the photoexcited VO+/VO2+. In the forward measurement, the transfer curves exhibit a positive shift in the VGS direction with a hump at the turn-on voltage region when the NBIS duration exceeds 1000 s, which is attributed to the synergistic effects of the generated defect states and the trapped holes at the front-channel interface. After the forward measurement with VGS = −10–20 V, the ionized VO+/VO2+ would be gradually neutralized by capturing electrons, and the trapped holes at the front-channel interface are completely de-trapped due to the vertical electric field induced by the positive VGS. Consequently, the abnormal hump observed in the forward measurement disappears in the reverse measurement, suggesting that the donor-like defect states, located near the Fermi level (EF) at VGS of the turn-on voltage, are generated and stabilized in the IGZO layer. It is noted that the trapped electrons at the back-channel interface are hardly de-trapped even when the positive VGS is applied [17]. As a result, the transfer curves in the reverse measurement exhibit a parallel shift of 10.03 V without SS degradation in the positive VGS direction after the NBIS duration of 104 s.

When the TIGZO is increased to 45 nm, which is close to the Debye length, the whole channel layer is almost depleted under the negative VGS bias. During the NBIS duration, more electrons and holes are excited and trapped at the back-channel and the front-channel interfaces. Simultaneously, the high-density defect states are generated due to the increase in the photoexcited VO+/VO2+. As a consequence, after the 104 s NBIS duration, the transfer curves show a significant shift in the positive VGS direction with a prominent hump for the forward measurement and display a distinct change of 12.46 V in the positive VGS direction without SS degradation for the reverse measurement, as shown in Figure 4a. When the TIGZO is further increased to 75 and 100 nm, which is larger than the Debye length, the electric potential exponentially decreases inside the active layer under −30 V VGS bias. Although some amount of electrons are photoexcited to EC, they are partly accumulated and trapped at the back-channel interface due to the weaker vertical electric field. As a result, the excited hole in EV and the ionized VO+/VO2+ near EF at VGS of the turn-on voltage would be neutralized by the free electrons. Therefore, after the NBIS duration of 104 s, the transfer curves exhibit a small shift in the positive VGS direction with a weak hump for the forward measurement, and show the small shift of 9.45 and 9.96 V in the positive VGS direction for the reverse measurement corresponding to TIGZO = 75 and 100 nm, respectively.

[2190-4286-9-239-4]

Figure 4: (a) Variation in Vth of the transfer curves of a-IGZO TFTs with various TIGZO in the reverse measurement as a function of NBIS duration and (b) hysteresis of a-IGZO TFTs with various TIGZO as a function of NBIS duration.

The combination of the transfer curves in the forward and reverse measurements after the NBIS duration of 104 s is shown in Figure 4b. The NBIS-induced hysteresis increases remarkably from 11.94 V for the TFT with TIGZO = 25 nm to 13.47 V for the TFT with TIGZO = 45 nm, and decreases drastically to 9.54 and 9.93 V when the TIGZO further increases to 75 and 100 nm. For further quantitative analysis of the origin of the NBIS-induced hysteresis, a positive gate pulse mode is carried out just after the NBIS duration of 104 s. Based on our previous publication [17], the optimized condition of a positive gate pulse with pulse width of 1 ms and pulse height of 10 V is enough to neutralize the ionized VO+/VO2+-induced donor-like defect states while it has no influence on the trapped holes at the front-channel interface, as shown in Figure 3i–l. It is found that the trapped hole-induced hysteresis are 3.41, 5.58, 4.39, and 4.62 V corresponding to the IGZO TFTs with TIGZO = 25, 45, 75, and 100 nm, respectively.

To further reveal the mechanism of the NBIS-induced hump and transfer curve shift in a-IGZO TFTs with various TIGZO, C–V analyses before and after the NBIS duration of 104 s are measured, as shown in Figure 5. In the initial stage, all C–V curves without distortion are observed, which are in agreement with the initial I–V curves. After the 104 s NBIS duration, all C–V curves shift in a positive VGS direction with distortion near the turn-on region. The C–V results suggest that NBIS-induced defect states are uniform in the whole channel layer near EF at VGS of the turn-on voltage. In case of a-IGZO TFT with TIGZO = 25 nm, the C–V curve shifts 11.4 V in the positive VGS direction with a hump in the off-state. On the basis of the C–V results, the energy-band diagrams for the IGZO TFTs with various TIGZO under NBIS are illustrated in Figure 6. The energy band at the front-channel is remarkably bent upward under the negative VGS bias when the TIGZO is less than the Debye length, as shown in Figure 6a. A hump observed at the turn-on region of the C–V curve indicates that the energy level of the generated defect states is located near EF at VGS of the turn-on voltage. The positive shift of the C–V curve demonstrates that electrons are trapped at the back-channel interface due to the vertical electric fields in the channel. When the TIGZO is increased to 45 nm, the C–V curve exhibits a large shift of 15.1 V in the positive VGS direction with a distinct hump at the turn-on region. The obtained results suggest that because the TIGZO is close to the Debye length, the high-density defect states are generated in the whole channel layer and more electrons are photoexcited and trapped at the back-channel interface. When the TIGZO is further increased to 75 and 100 nm, the C–V curves exhibit smaller shifts of 8.7 and 9.3 V in the positive VGS direction with a weaker hump in the off-state compared to the 45 nm-thick channel layer case. The energy band at the front-channel is slightly bent upward as the TIGZO is much larger than the Debye length, as shown in Figure 6b. The weakened hump near the turn-on region illustrates that the photoexcited VO+/VO2+ would be neutralized by the free electrons in EC, contributing to the low-density defect states near EF at VGS of turn-on voltage. The small shift of the C–V curves demonstrates that the fewer electrons are accumulated and trapped at the back-channel interface. The obtained C–V results are correlated with the results of the I–V measurements.

[2190-4286-9-239-5]

Figure 5: C−V curves before and after the NBIS duration of 104 s with (a) TIGZO = 25, (b) 45, (c) 75, and (d) 100 nm, respectively.

[2190-4286-9-239-6]

Figure 6: Schematic diagram of NBIS-induced degradation mechanism in a-IGZO TFTs with (a) TIGZO < 45 and (b) over the Debye length (75 and 100 nm).

On the basis of the above discussion, it is demonstrated that the TIGZO is one of the critical parameters to modify the electrical properties of the device. Besides the active layer thickness, the intrinsic characteristics of a-IGZO and the front- and back-channel interfaces of the TFT also play a vital role for the high-performance devices. Moreover, to reduce the density of oxygen vacancies in the bulk of the IGZO for the enhancement of electrical properties and stress stability of the TFTs, the following two aspects should be mainly considered: (i) oxidizing the densities of the defect state of oxide semiconductors to suppress charge trapping, for example by oxygen annealing and N2O plasma treatment [31]; and (ii) inactivating the defects in the semiconductor by means of introducing new elements to form stable chemical bonds with the defects, for example by fluoride ion implantation and nitrogen annealing [32,33].

Conclusion

The impact of the TIGZO on the photoleakage current and the NBIS-induced instability in a-IGZO TFTs were systematically investigated. It was found that when the photon energy of the light irradiation exceeds ≈2.7 eV, the photoleakage current increases in all TFTs irrespective of the TIGZO due to the high-density electron traps existing at an (EC − E) of ≈2.7 eV. Because the total amount of trapped electrons increases with increasing TIGZO, the photoleakage current gradually increases with increasing TIGZO. On the basis of the photoleakage current results, the influence of the TIGZO on NBIS with a photon energy of ≈2.7 eV in a-IGZO TFTs is clarified by the I–V and C–V measurements. In addition, the NBIS-induced hysteresis is quantitatively evaluated through a positive gate pulse mode, contributing to the separation of the trapped holes at the front-channel interface and the generation of donor-like defect states in a-IGZO layer. The obtained I–V and C–V results indicate that when the TIGZO is close to the Debye length, the trapped holes at the front-channel interface, the trapped electrons at the back-channel interface, and the generated donor-like defect states in a-IGZO are distinctly prominent during NBIS. This study suggests that to improve the reliability of oxide TFTs under light irradiation and gate bias stresses, the quality of the active layer and interface engineering should be taken into account.

Acknowledgements

The authors acknowledge all support from the National Key Research and Development Program of China (2016YFA0202403), the National Nature Science Foundation of China (61674098, 91733301), the Natural Science Foundation of Shaanxi Provincial Department of Education (2017KW-023, 2017JM6020), the Fundamental Research Funds for the Central Universities (GK201603053, GK201702003, GK201601010), the Changjiang Scholar and the Innovative Research Team (IRT_14R33), the 111 Project (B14041), and the Chinese National 1000-talent-plan program (Grant No. 111001034). This work was partly supported by JSPS KAKENHI Grant Number 16K06309.

References

  1. Nomura, K.; Ohta, H.; Takagi, A.; Kamiya, T.; Hirano, M.; Hosono, H. Nature 2004, 432, 488–492. doi:10.1038/nature03090
    Return to citation in text: [1]
  2. Zhang, J.; Yang, J.; Li, Y.; Wilson, J.; Ma, X.; Xin, Q.; Song, A. Materials 2017, 10, 319. doi:10.3390/ma10030319
    Return to citation in text: [1]
  3. Hirao, T.; Furuta, M.; Hiramatsu, T.; Matsuda, T.; Li, C.; Furuta, H.; Hokari, H.; Yoshida, M.; Ishii, H.; Kakegawa, M. IEEE Trans. Electron Devices 2008, 55, 3136–3142. doi:10.1109/ted.2008.2003330
    Return to citation in text: [1]
  4. Tsai, C.-T.; Chang, T.-C.; Chen, S.-C.; Lo, I.; Tsao, S.-W.; Hung, M.-C.; Chang, J.-J.; Wu, C.-Y.; Huang, C.-Y. Appl. Phys. Lett. 2010, 96, 242105. doi:10.1063/1.3453870
    Return to citation in text: [1]
  5. Kim, K. S.; Ahn, C. H.; Kang, W. J.; Cho, S. W.; Jung, S. H.; Yoon, D. H.; Cho, H. K. Materials 2017, 10, 530. doi:10.3390/ma10050530
    Return to citation in text: [1]
  6. Liu, A.; Liu, G.; Zhu, C.; Zhu, H.; Fortunato, E.; Martins, R.; Shan, F. Adv. Electron. Mater. 2016, 2, 1600140. doi:10.1002/aelm.201600140
    Return to citation in text: [1]
  7. Kim, Y. G.; Kim, T.; Avis, C.; Lee, S.-H.; Jang, J. IEEE Trans. Electron Devices 2016, 63, 1078–1084. doi:10.1109/ted.2016.2518703
    Return to citation in text: [1]
  8. Liu, X.; Jiang, L.; Zou, X.; Xiao, X.; Guo, S.; Jiang, C.; Liu, X.; Fan, Z.; Hu, W.; Chen, X.; Lu, W.; Hu, W.; Liao, L. Adv. Mater. 2014, 26, 2919–2924. doi:10.1002/adma.201305073
    Return to citation in text: [1]
  9. Shimakawa, S.-i.; Kamada, Y.; Kawaharamura, T.; Wang, D.; Li, C.; Fujita, S.; Hirao, T.; Furuta, M. Jpn. J. Appl. Phys. 2012, 51, 03CB04. doi:10.7567/jjap.51.03cb04
    Return to citation in text: [1] [2]
  10. Shimakawa, S.-i.; Wang, D.; Furuta, M. Jpn. J. Appl. Phys. 2012, 51, 108003. doi:10.7567/jjap.51.108003
    Return to citation in text: [1]
  11. Chang, Y.-G.; Moon, T.-W.; Kim, D.-H.; Lee, H. S.; Kim, J. H.; Park, K.-s.; Kim, C.-D.; Im, S. IEEE Electron Device Lett. 2011, 32, 1704–1706. doi:10.1109/led.2011.2167736
    Return to citation in text: [1]
  12. Oh, H.; Yoon, S.-M.; Ryu, M. K.; Hwang, C.-S.; Yang, S.; Park, S.-H. K. Appl. Phys. Lett. 2011, 98, 033504. doi:10.1063/1.3540500
    Return to citation in text: [1]
  13. Kim, B.; Chong, E.; Hyung Kim, D.; Woo Jeon, Y.; Hwan Kim, D.; Yeol Lee, S. Appl. Phys. Lett. 2011, 99, 062108. doi:10.1063/1.3615304
    Return to citation in text: [1]
  14. Chen, T. C.; Chang, T. C.; Hsieh, T. Y.; Tsai, C. T.; Chen, S. C.; Lin, C. S.; Jian, F. Y.; Tsai, M. Y. Thin Solid Films 2011, 520, 1422–1426. doi:10.1016/j.tsf.2011.09.002
    Return to citation in text: [1]
  15. Oh, H.; Yoon, S.-M.; Ryu, M. K.; Hwang, C.-S.; Yang, S.; Park, S.-H. K. Appl. Phys. Lett. 2010, 97, 183502. doi:10.1063/1.3510471
    Return to citation in text: [1] [2]
  16. Huang, S.-Y.; Chang, T.-C.; Lin, L.-W.; Yang, M.-C.; Chen, M.-C.; Jhu, J.-C.; Jian, F.-Y. Appl. Phys. Lett. 2012, 100, 222901. doi:10.1063/1.4722787
    Return to citation in text: [1]
  17. Wang, D.; Hung, M. P.; Jiang, J.; Toda, T.; Furuta, M. ACS Appl. Mater. Interfaces 2014, 6, 5713–5718. doi:10.1021/am500300g
    Return to citation in text: [1] [2] [3]
  18. Hung, M. P.; Wang, D.; Jiang, J.; Furuta, M. ECS Solid State Lett. 2014, 3, Q13–Q16. doi:10.1149/2.010403ssl
    Return to citation in text: [1] [2]
  19. Park, H.-W.; Park, K.; Kwon, J.-Y.; Choi, D.; Chung, K.-B. IEEE Trans. Electron Devices 2017, 64, 159–163. doi:10.1109/ted.2016.2630043
    Return to citation in text: [1]
  20. Park, J.; Kim, Y. S.; Kim, J. H.; Park, K.; Park, Y. C.; Kim, H.-S. J. Alloys Compd. 2016, 688, 666–671. doi:10.1016/j.jallcom.2016.07.245
    Return to citation in text: [1]
  21. Yang, Z.; Yang, J.; Meng, T.; Qu, M.; Zhang, Q. Mater. Lett. 2016, 166, 46–50. doi:10.1016/j.matlet.2015.12.029
    Return to citation in text: [1]
  22. Avis, C.; Hwang, H. R.; Jang, J. ACS Appl. Mater. Interfaces 2014, 6, 10941–10945. doi:10.1021/am501153w
    Return to citation in text: [1]
  23. Jeong, J.; Hong, Y. IEEE Trans. Electron Devices 2012, 59, 710–714. doi:10.1109/ted.2011.2180908
    Return to citation in text: [1]
  24. Wang, D.; Hung, M. P.; Jiang, J.; Toda, T.; Li, C.; Furuta, M. Jpn. J. Appl. Phys. 2014, 53, 03CC01. doi:10.7567/jjap.53.03cc01
    Return to citation in text: [1]
  25. Wang, D.; Zhao, W.; Li, H.; Furuta, M. Materials 2018, 11, 559. doi:10.3390/ma11040559
    Return to citation in text: [1]
  26. Nomura, K.; Kamiya, T.; Yanagi, H.; Ikenaga, E.; Yang, K.; Kobayashi, K.; Hirano, M.; Hosono, H. Appl. Phys. Lett. 2008, 92, 202117. doi:10.1063/1.2927306
    Return to citation in text: [1]
  27. Kamiya, T.; Nomura, K.; Hosono, H. J. Disp. Technol. 2009, 5, 468–483. doi:10.1109/jdt.2009.2034559
    Return to citation in text: [1]
  28. Libsch, F. R.; Kanicki, J. Appl. Phys. Lett. 1993, 62, 1286–1288. doi:10.1063/1.108709
    Return to citation in text: [1]
  29. Janotti, A.; Van de Walle, C. G. Appl. Phys. Lett. 2005, 87, 122102. doi:10.1063/1.2053360
    Return to citation in text: [1]
  30. Li, Y.; Pei, Y. L.; Hu, R. Q.; Chen, Z. M.; Zhao, Y.; Shen, Z.; Fan, B. F.; Liang, J.; Wang, G. Curr. Appl. Phys. 2014, 14, 941–945. doi:10.1016/j.cap.2014.04.011
    Return to citation in text: [1]
  31. Chang, G.-W.; Chang, T.-C.; Jhu, J.-C.; Tsai, T.-M.; Syu, Y.-E.; Chang, K.-C.; Tai, Y.-H.; Jian, F.-Y.; Hung, Y.-C. Appl. Phys. Lett. 2012, 100, 182103. doi:10.1063/1.4709417
    Return to citation in text: [1]
  32. Wang, D.; Jiang, J.; Furuta, M. J. Disp. Technol. 2016, 12, 258–262. doi:10.1109/jdt.2015.2472981
    Return to citation in text: [1]
  33. Yoon, S.; Tak, Y. J.; Yoon, D. H.; Choi, U. H.; Park, J.-S.; Ahn, B. D.; Kim, H. J. ACS Appl. Mater. Interfaces 2014, 6, 13496–13501. doi:10.1021/am502571w
    Return to citation in text: [1]

© 2018 Wang and Furuta; licensee Beilstein-Institut.
This is an Open Access article under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0). Please note that the reuse, redistribution and reproduction in particular requires that the authors and source are credited.
The license is subject to the Beilstein Journal of Nanotechnology terms and conditions: (https://www.beilstein-journals.org/bjnano)

Other Beilstein-Institut Open Science Activities
Journal Logo
Institut Logo
Preprint Logo
Symposia Logo