本實驗研究的元件為鋅錫氧化物(zinc-tin oxide, ZTO)薄膜電晶體(thin film transistor, TFT)，採用SiO2/N+ Si基板，Si基板同時作為閘極，SiO2為介電層，4nm ZTO層採用溶液法製備並利用微影蝕刻法定義面積，汲極與源極則以電子束蒸鍍鋁作為電極，且電極邊緣與ZTO層邊緣貼齊。
本實驗對ZTO TFT施加四種520nm光源相關應力，光源強度10mW/cm2，並施加不同閘極偏壓(VG)條件，應力時間為1600s，四種應力皆於開始前以及應力結束後的0, 200, 400, 600秒會量測轉換特性曲線(transfer characteristic curve, ID-VG curve)和電容電壓曲線(capacitance-voltage curve, C-V curve)，量測之間的空檔VG為0V。四種應力分別為照光應力(illumination stress, IS)，應力期間VG=0V；正閘極偏壓照光應力(positive bias illumination stress, PBIS)，應力期間VG=10V；負閘極偏壓照光應力(negative bias illumination stress, NBIS)，應力期間VG=-10V；以及照光中閘極偏壓掃伏(gate bias sweep during illumination stress, GBSIS)，施加GBSIS期間於0, 500, 1000, 1500秒於照光條件下量測ID-VG或C-V曲線，照光當中量測之間的空檔VG亦為0V。後續又進行五種405nm光源相關應力，其中IS、PBIS、NBIS除了光源波長以外皆與上述520nm相關應力條件相同，另外兩種應力則是在施加405nm IS後分別施加600s的PBS(VG=10V)和NBS(VG=-10V)，由於照射光源能量低於ZTO能隙，無法產生電子電洞對，取而代之是中性氧空缺被激發生成帶正電氧空缺與電子。吾人主要探討受到應力與後續恢復過程中固定電荷(fixed charge)與能隙內能態(subgap states)的變化。
元件內部的固定電荷需要克服能障(energy barrier)才能消除其電荷或移動，其電場導致量測出的曲線發生平移。對於氧化鋅系半導體TFT而言，固定負電荷主要來源是穿隧進入介電層的電子，C-V或ID-VG掃伏時需要施加更大的正偏壓抵銷負電場的效果才能使半導體累積電子形成通道。固定正電荷的來源則為受光激發生成的帶正電氧空缺能態，其正電場會吸引電子，導致C-V或ID-VG掃伏時可以用更低VG達成載子累積與通道開啟。在ID-VG曲線的表現上主要由閾值電壓(threshold voltage, VTH)變化來判斷，C-V曲線則利用電容極大值Cmax對應的VG(Vmax)變化來估計。
對於520nm光源相關應力而言，C-V和ID-VG曲線比較應力結束後0秒，PBIS為正值，而IS、NBIS、GBSIS均為負偏移，其中又以NBIS最顯著。當元件經歷IS時，ZTO的中性氧空缺被激發，生成的帶正電氧空缺又有部分會與通道內的電子複合，最終殘留的帶正電氧空缺變導致IS結束後0秒C-V與ID-VG曲線的負偏移。當元件經歷PBIS時，同時半導體內電子受閘極偏壓吸引，並被介電層捕捉，並且大於殘留的帶正電氧空缺的量導致最終的正偏移。NBIS展現出比IS更加明顯負偏移，因為應力負偏壓排斥半導體內電子，減少帶正電氧空缺複合，使得NBIS結束時帶正電氧空缺的濃度更高，導致更大的負偏移。GBSIS於照光期間掃伏，相當於經歷了短時間PBIS與短時間NBIS，觀察偏移量的絕對值會發現NBIS大於PBIS，造成GBSIS的ID-VG與C-V曲線偏移量介於IS與NBIS偏移量之間，並且可以說明照光當中量測ID-VG或C-V，其變化不單是來自於光反應，而是光與偏壓掃伏的交互作用。當IS、PBIS、NBIS光源改為405nm時，IS和NBIS的負偏移變得更加顯著，PBIS的正偏移則變得幾乎可忽略，上述的變化是來自於405nm可以激發更接近價帶處的中性氧空缺等能隙內能態，生成更多正的固定電荷。
520nm相關應力結束後600秒相對於應力前，仍有殘存的偏移量，除了IS C-V例外，由應力結束後0秒的負偏移，到600秒變為正偏移，正偏移來源是後續掃伏過程正偏壓導致的電子捕捉。其餘三種應力的應力偏移量相對大，掃伏當中電子被捕捉的效應不明顯，其中PBIS因為被捕捉的電子回到半導體與帶正電氧空缺複合兩種效應相互抵銷導致最慢的偏移回復速率，NBIS因為帶正電氧空缺於長時間負偏壓下移動到介面處難以複合次之，GBSIS由於生成帶正電氧空缺仍處於容易複合的半導體內，而有最快的偏移回復速度。405nm相關應力當中，若ZTO TFT單純施加IS，則後續經過600秒後的回復並不明顯，若IS後接續PBS 600s後負偏移則有顯著的消除，除了有帶正電氧空缺加速複合的因素外，還有電子被捕捉在介電層內負電場的效應，IS後接續NBS 600s則使負偏移稍微加劇，這是由於帶正電氧空缺被吸引至半導體與介電層介面，其電場不被自由電子屏蔽而增強。
經歷過520nm相關應力後能隙內的缺陷能態帶電狀態在ID-VG曲線主要由(VTH-Von)的變化量來判斷，其中Von為TFT開關電壓(turn-on voltage)。在C-V曲線則利用與應力結束C-V曲線相對於應力開始前的C-V線型伸展(stretch out)程度來計算出能隙內能態密度來估計帶電缺陷能態密度變化量。Von時ID-VG曲線處於電流最低點，代表此時半導體層內自由電子濃度最低，VTH時自由電子數量則足以使通道開啟，隨著VG增加，越來越多電子注入半導體內，注入的電子一部分增加自由電子濃度，另一部分則會佔據極靠近導帶的淺層能隙內能態。假設TFT ID-VG掃伏時通道開啟時自由電子的濃度為定值，則經歷應力前後(VTH-Von)的變化量會與填入能隙內能態電子數目呈正比。Δ(VTH-Von)是應力後(VTH-Von)減掉應力前(VTH-Von)，當應力結束後0秒，四種應力的Δ(VTH-Von)皆為正值，代表於照光相關應力過後的ID-VG量測過程當中更多電子被填入半導體的淺層能隙內能態，需要多填入的電子則可能來自於EC以上帶正電氧空缺能態與導帶共振(resonant)衍伸出容納光電子的能態，其中NBIS的Δ(VTH-Von)於應力結束達600秒相對應力結束0秒變化最少，由於帶正電氧空缺在NBIS期間移動至介面難以複合，上述的淺層能態也因此繼續存在。
吾人將應力結束後測量的C-V曲線平移以抵銷其偏移量。受限於ZTO本身極薄的厚度使得VG分配在ZTO層的壓降極低。平帶(flat band)狀況下EF就相當接近EC，量測到的能態密度(density of states, DOS)約為EC以下0.05至0.2eV的淺層能態。量測VG偏負的區間，應力電容值增加，計算出DOS為正值，代表有能態不易被電子回填，可能是部分能量較低的帶正電氧空缺能態；VG偏正時電容值則減少，對應一極短區間DOS為負值，代表部分ID-VG 曲線Δ(VTH-Von) 偵測到的極淺層能態，代表一部份可被電子佔據的能態。量測的DOS並不隨應力結束時間拉長呈現有趨勢的變化。

The devices investigated in this work was zinc-tin oxide (ZTO) thin film transistor (TFT). 4nm-thick ZTO was fabricated as active layer on SiO2/n+ Si substrate by solution process and was patterned by photolithography. SiO2 acted as gate insulator, and n+ Si acted as gate terminal. Al was fabricated on ZTO by e-beam evaporation and patterned with hard mask. Al acted as source/drain electrodes, the edges of which were aligned with the edges of active layer.
ZTO TFT underwent four types of 520nm illumination-associated stress in this work, and transfer characteristic curves (ID-VG curves) and capacitance-voltage curves (C-V curves) were measured before and after the stress at 0, 200, 400 and 600s, with VG maintained 0V between measurement. The light source was at power density of 10mW/cm2, and the duration time of illumination was 1600s. Four types of stress were illumination stress (IS) with gate voltage (VG) = 0V during illumination, positive bias illumination stress (PBIS) with VG = 10V during illumination, negative bias illumination stress (NBIS) with VG = -10V during illumination, and gate bias sweep during illumination stress (GBSIS) with ID-VG and C-V curves measured at 0, 500, 1000, 1500s and VG = 0V between measurement during illumination. ZTO TFT also underwent five types of 405nm (3.06eV) illumination-associated stress. The conditions of IS, PBIS and NBIS were the same as corresponding 520nm-associated stress except light source wavelength. The other two kinds were 405nm IS following by 600s positive bias stress (PBS, VG=10V)/negative bias stress (NBS, VG=-10V). Electron-hole pairs would not be generated during 520nm illumination since the photon energy (2.38, 3.06eV) was lower than the bandgap of ZTO (3.7eV). Instead, neutral oxygen vacancies would be formed positively charged oxygen vacancies and electrons. We mainly investigated the variation of fixed charges and subgap density of states affected by the stress and recovery behavior after stress.
The fixed-charge-induced electric field caused measured curves to shift. The main source of negative fixed charges of ZnO-based material TFT was trapped electrons tunneling from semiconductor into gate insulator. In order to cancel out negative electric field from trapped electrons, higher VG should be applied so that electrons could accumulate, and the channel could be turned on during ID¬-VG or C-V measurement, which made ID¬-VG or C-V curve shift positively. The major origin of positive fixed charge was positively oxygen vacancies created from photoexcitation of neutral oxygen vacancies. The electric field of positively charged oxygen vacancies could attract electrons. Less VG were needed to make electrons accumulate and turn on the channel, which caused ID¬-VG or C-V curve shift negatively. The shift of ID-VG curves could be judged by threshold voltage change (ΔVTH); The shift of C-V curves could be judged by the change of VG corresponding to maximum capacitance (ΔVmax).
Comparing ΔVTH of four type of 520nm-associated stress between ID-VG curves before stress and after stress at 0s. The ΔVTH of PBIS was positive, while the ΔVTH of IS, NBIS, GBSIS were negative. ID-VG and C-V curves of the device experienced NBIS exhibit the most significant negative shift. When the device underwent IS, neutral oxygen vacancies would be transformed into positively charged oxygen vacancies and electrons. Part of generated positively charged oxygen vacancies would recombine with electrons. Residue positively charged oxygen vacancies led to negative shift of C-V and ID-VG curves after IS at 0s. For PBIS condition, electrons in the ZTO layer were attracted by the positive bias and trapped in insulators. Eventually, the concentration of trapped electrons was higher than that of positively charged oxygen vacancies, which led to final positive shift of ID-VG and C-V curves. For NBIS condition, lower electrons concentration induced by negative gate bias would reduce the recombination rate between positively charged oxygen vacancies and electrons. As a result, the device after NBIS left more positively charged oxygen vacancies than the device after IS did and showed more negative shift in C-V and ID-VG measurement. For the GBSIS condition, the device equivalently experienced short-term PBIS and short-term NBIS during VG sweeping. The absolute value of ΔVTH from ID-VG curves and ΔVmax from C-V curves of NBIS were larger than those of PBIS. ΔVTH and ΔVmax of GBSIS were consequently between those of NBIS and IS respectively. As a consequence, the change of ID-VG and C-V curves were not only the photoreponse of semiconductor. When the light source of IS, PBIS and NBIS changed into 405nm, the negative shift of C-V and ID-VG curves caused by IS and NBIS became more apparent, while the positive shift caused by PBIS become almost ignorable because the deeper subgap states at ionized by 405nm laser and created more positive fixed charge.
ΔVTH from ID-VG curves and ΔVmax from C-V curve at 600s implied that residue effect of 520nm-associated stress existed except C-V curves from IS condition. ΔVmax was negative after IS at 0s while ΔVmax became positive after the end of IS at 600s, which might be caused by the electrons trapping in the longer time spending on positive VG regime in C-V measurement. The electrons trapping effect from positive VG regime in C-V or ID-VG measurement could be neglected for other three types of stress since the shift were large enough. For PBIS condition, since trapped electrons escaping and positively charged oxygen vacancies recombination has opposite effect, the device exhibited the lowest recovery rate. For NBIS condition, positively charged oxygen vacancies moved to semiconductor-insulator interface because of long-term negative bias during NBIS and had lower probability of recombination with. For GBSIS condition, positively charged oxygen vacancies only underwent short-term negative bias and still stayed in the semiconductor and had higher probability to recombine with electrons in semiconductor. GBSIS condition exhibit higher recovery rate than PBIS and NBIS conditions. For 405nm-associated stress, when ZTO TFT simply underwent IS, the recovery behavior was not apparent after IS 600s. When 405nm IS were followed by 600s PBS, the negative shift caused by IS could be eliminated because the positive charged oxygen vacancies recombination was boosted, and some electrons was trapped into insulator. When 405nm IS were followed by 600s NBS, the negative shift caused by IS would become larger since some of the positively charged oxygen were attracted to the interface and the positive electric field was no longer screened by free electrons IS.
Turn-on voltage (Von) and VTH corresponded respectively to lowest point of the ID-VG curve when the free electrons were at concentration close to zero in active layer and the point that the electron concentration were high enough to form the carrier channel. As VG increase, more electrons were injected into active layer. some injected electrons become free electrons, while the other occupied shallow subgap states close to conduction band minima. Assuming that the free electrons concentration was constant when the carrier channel formed, then the change of (VTH-Von) between ID-VG curves before and after stress, denoted asΔ(VTH-Von) would be proportional to the electrons concentration occupied shallow subgap states. The values of Δ(VTH-Von) are all positive for all four types of stress after the stress at 0s. It implies that more electrons filled in the shallow subgap states during ID-VG measurement after illumination-associated stress. The phenomenon mentioned above might be associated with the states accommodating photoexcited electrons, and the states were created by resonance between positively charged oxygen vacancies and conduction band. The NBIS condition shows lowest recovery rate basin on Δ(VTH-Von) at 600s after the end of stress. The positively charged oxygen vacancies moved to the semiconductor-insulator interface under NBIS and had lowest recombination rate. As a result, the shallow states mentioned above had lowest elimination rate for ZTO TFT after NBIS.
For 520nm-associated stress, to eliminate the C-V curves shift first, C-V curves after stress should be shifted by -ΔVmax before subgap density of states (DOS) calculation. Extremely thin ZTO resulted in low potential drop of ZTO while VG applied, and EF was very close to EC in flat band condition. As a result, the DOS calculated from C-V measurement was associated with shallow states in the energy interval of 0.05 to 0.2eV below EC. In negative VG regime of C-V curves, capacitance values after stress were usually higher than that before stress, and it was corresponded to positive DOS which implied the existence of subgap states that couldn’t be occupied by electrons after stress and might be composed of part of low energy positively charged oxygen vacancies states. In positive VG regime of C-V curves, capacitance values after stress usually were lower than that before stress in spite of stress type, which was corresponded to negative DOS, which implied the existence of subgap states occupied by photo-generated electrons in extreme narrow interval of energy close to EC.

1. K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano and H. Hosono, Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors, Nature 432, 488-492 (2004).
2. H. Hosono, Ionic amorphous oxide semiconductors: Material design, carrier transport, and device application, Journal of Non-Crystalline Solids 352, 851-858 (2006).
3. J. Robertson, Disorder and instability processes in amorphous conducting oxides, physica status solidi (b) 245, 1026-1032 (2008).
4. H. Bae, H. Choi, S. Jun, C. Jo, Y. H. Kim, J. S. Hwang, J. Ahn, S. Oh, J.-U. Bae, S.-J. Choi, D. H. Kim and D. M. Kim, Single-Scan Monochromatic Photonic Capacitance-Voltage Technique for Extraction of Subgap DOS Over the Bandgap in Amorphous Semiconductor TFTs, IEEE Electron Device Lett. 34, 1524-1526 (2013).
5. J. T. Jang, S.-J. Choi, D. M. Kim and D. H. Kim, The Calculation of Negative Bias Illumination Stress-Induced Instability of Amorphous InGaZnO Thin-Film Transistors for Instability-Aware Design, IEEE Trans. Electron Devices 65, 1002-1008 (2018).
6. A. Janotti and C. G. Van de Walle, Oxygen vacancies in ZnO, Appl. Phys. Lett. 87, 122102 (2005).
7. I. Tanaka, F. Oba, K. Tatsumi, M. Kunisu, M. Nakano and H. Adachi, Theoretical Formation Energy of Oxygen-Vacancies in Oxides, MATERIALS TRANSACTIONS 43, 1426-1429 (2002).
8. S. Lany and A. Zunger, Anion vacancies as a source of persistent photoconductivity in II-VI and chalcopyrite semiconductors, Physical Review B 72, 035215 (2005).
9. T. Oga, Y. Izawa, K. Kuriyama, K. Kushida and Q. Xu, Persistent Photoconductivity in electron-irradiated ZnO bulk single crystals: Evaluation of the metastable conductive state by the dual light illumination, Solid State Commun. 151, 1700-1703 (2011).
10. U. K. Kim, S. H. Rha, J. H. Kim, Y. J. Chung, J. Jung, E. S. Hwang, J. Lee, T. J. Park, J.-H. Choi and C. S. Hwang, Study on the defects in metal–organic chemical vapor deposited zinc tin oxide thin films using negative bias illumination stability analysis, Journal of Materials Chemistry C 1, 6695-6702 (2013).
11. B. Ryu, H.-K. Noh, E.-A. Choi and K. J. Chang, O-vacancy as the origin of negative bias illumination stress instability in amorphous In–Ga–Zn–O thin film transistors, Appl. Phys. Lett. 97, 022108 (2010).
12. H.-H. Hsieh, T. Kamiya, K. Nomura, H. Hosono and C.-C. Wu, Modeling of amorphous InGaZnO4 thin film transistors and their subgap density of states, Appl. Phys. Lett. 92, 133503 (2008).
13. K. Hayashi, A. Hino, S. Morita, S. Yasuno, H. Okada and T. Kugimiya, Electron traps in amorphous In–Ga–Zn–O thin films studied by isothermal capacitance transient spectroscopy, Appl. Phys. Lett. 100, 102106 (2012).
14. B. Li, Y. Adachi, J. Li, H. Okushi, I. Sakaguchi, S. Ueda, H. Yoshikawa, Y. Yamashita, S. Senju, K. Kobayashi, M. Sumiya, H. Haneda and N. Ohashi, Defects in ZnO transparent conductors studied by capacitance transients at ZnO/Si interface, Appl. Phys. Lett. 98, 082101 (2011).
15. W. H. Han and K. J. Chang, Subgap States near the Conduction-Band Edge Due to Undercoordinated Cations in Amorphous In-Ga-Zn-O and Zn-Sn-O Semiconductors, Physical Review Applied 6, 044011 (2016).
16. W. Körner and C. Elsässer, Density-functional theory study of stability and subgap states of crystalline and amorphous Zn–Sn–O, Thin Solid Films 555, 81-86 (2014).
17. M. G. Yun, S. H. Kim, C. H. Ahn, S. W. Cho and H. K. Cho, Effects of channel thickness on electrical properties and stability of zinc tin oxide thin-film transistors, Journal of Physics D: Applied Physics 46, 475106 (2013).
18. T.-C. Chen, T.-C. Chang, T.-Y. Hsieh, M.-Y. Tsai, C.-T. Tsai, S.-C. Chen, C.-S. Lin and F.-Y. Jian, Analyzing the effects of ambient dependence for InGaZnO TFTs under illuminated bias stress, Surface and Coatings Technology 231, 465-470 (2013).
19. K. Nomura, T. Kamiya and H. Hosono, Interface and bulk effects for bias—light-illumination instability in amorphous-In—Ga—Zn—O thin-film transistors, Journal of the Society for Information Display 18, 789-795 (2010).
20. H. Oh, S.-M. Yoon, M. K. Ryu, C.-S. Hwang, S. Yang and S.-H. K. Park, Photon-accelerated negative bias instability involving subgap states creation in amorphous In–Ga–Zn–O thin film transistor, Appl. Phys. Lett. 97, 183502 (2010).
21. Y. Ueoka, Y. Ishikawa, J. P. Bermundo, H. Yamazaki, S. Urakawa, M. Fujii, M. Horita and Y. Uraoka, Density of States in Amorphous In-Ga-Zn-O Thin-Film Transistor under Negative Bias Illumination Stress, ECS J. Solid State Sc. 3, Q3001-Q3004 (2014).
22. J. G. Um, M. Mativenga and J. Jang, Mechanism of positive bias stress-assisted recovery in amorphous-indium-gallium-zinc-oxide thin-film transistors from negative bias under illumination stress, Appl. Phys. Lett. 103, 033501 (2013).
23. S. Jeon, S. E. Ahn, I. Song, C. J. Kim, U. I. Chung, E. Lee, I. Yoo, A. Nathan, S. Lee, J. Robertson and K. Kim, Gated three-terminal device architecture to eliminate persistent photoconductivity in oxide semiconductor photosensor arrays, Nat Mater 11, 301-305 (2012).
24. M. M. Billah and J. Jang, Millisecond Positive Bias Recovery of Negative Bias Illumination Stressed Amorphous InGaZnO Thin-Film Transistors, IEEE Electron Device Lett. 38, 477-480 (2017).
25. A. Wadsworth, 安捷倫科技參數量測手冊, 3 ed. (2013), pp.154-158.
26. M. Kimura, T. Nakanishi, K. Nomura, T. Kamiya and H. Hosono, Trap densities in amorphous-InGaZnO4 thin-film transistors, Appl. Phys. Lett. 92, 133512 (2008).
27. S. Lee, Y. W. Jeon, S. Kim, D. Kong, D. H. Kim and D. M. Kim, Comparative study of quasi-static and normal capacitance–voltage characteristics in amorphous Indium-Gallium-Zinc-Oxide thin film transistors, Solid-State Electronics 56, 95-99 (2011).
28. H. Choi, J. Lee, H. Bae, S.-J. Choi, D. H. Kim and D. M. Kim, Bias-Dependent Effective Channel Length for Extraction of Subgap DOS by Capacitance–Voltage Characteristics in Amorphous Semiconductor TFTs, IEEE Trans. Electron Devices 62, 2689-2694 (2015).
29. E. H. Nicollian and J. R. Brews, MOS(Metal-Oxide-Semiconductor) Physics and Technology. (John Wiley & Sons, New Jersey, 1982), pp.58-71.
30. D. A. Neamen, Semiconductor Physics and Device: Basic Principle, 4 ed. (McGraw-Hill Science Engineering, US, 2012), pp.385-389.
31. J. F. Wager, D. A. Keszler and R. E. Presley, Transparent Electronics. (Springer Science+Business Media, LLC, New York, 2008), pp.107-109.
32. D. Hong, G. Yerubandi, H. Q. Chiang, M. C. Spiegelberg and J. F. Wager, Electrical Modeling of Thin-Film Transistors, Crit. Rev. Solid State Mater. Sci. 33, 101-132 (2008).
33. T. Kamiya, K. Nomura and H. Hosono, Origins of High Mobility and Low Operation Voltage of Amorphous Oxide TFTs: Electronic Structure, Electron Transport, Defects and Doping, Journal of Display Technology 5, 468-483 (2009).
34. K. Jeon, C. Kim, I. Song, J. Park, S. Kim, S. Kim, Y. Park, J.-H. Park, S. Lee, D. M. Kim and D. H. Kim, Modeling of amorphous InGaZnO thin-film transistors based on the density of states extracted from the optical response of capacitance-voltage characteristics, Appl. Phys. Lett. 93, 182102 (2008).
35. P. Jun-Hyun, J. Kichan, L. Sangwon, K. Sunil, K. Sangwook, S. Ihun, K. Chang Jung, P. Jaechul, P. Youngsoo, K. Dong Myong and K. Dae Hwan, Extraction of Density of States in Amorphous GaInZnO Thin-Film Transistors by Combining an Optical Charge Pumping and Capacitance–Voltage Characteristics, IEEE Electron Device Lett. 29, 1292-1295 (2008).
36. L.-C. Liu, J.-S. Chen and J.-S. Jeng, Role of oxygen vacancies on the bias illumination stress stability of solution-processed zinc tin oxide thin film transistors, Appl. Phys. Lett. 105, 023509 (2014).
37. X. Zhou, Y. Shao, L. Zhang, H. Lu, H. He, D. Han, Y. Wang and S. Zhang, Oxygen Interstitial Creation in a-IGZO Thin-Film Transistors Under Positive Gate-Bias Stress, IEEE Electron Device Lett. 38, 1252-1255 (2017).
38. J. T. Jang, J. Park, B. D. Ahn, D. M. Kim, S. J. Choi, H. S. Kim and D. H. Kim, Study on the photoresponse of amorphous In-Ga-Zn-O and zinc oxynitride semiconductor devices by the extraction of sub-gap-state distribution and device simulation, ACS Appl Mater Interfaces 7, 15570-15577 (2015).