Microscopic identity of the deep acceptor-like charged traps and stress-induced shallow subgap states underlying threshold-voltage instability in room-temperature deposited a-In–Ga–Zn–O thin-film transistors

Amorphous In–Ga–Zn–O (a-IGZO) thin-film transistors (TFTs) deposited at room temperature and annealed in O2 at 400 °C were shown by Nomura, Kamiya, Hirano, and Hosono [Appl. Phys. Lett. (2009), DOI 10.1063/1.3159831] to exhibit two distinct threshold-voltage-instability mechanisms under constant-current stress: a small parallel positive shift in annealed devices, attributed by device simulation to an increased density of deep acceptor-like charged traps about 1.0 eV below the conduction-band edge, and a large shift with subthreshold-swing degradation in unannealed devices, attributed to stress-generated shallow subgap states. Fifteen years of follow-up work has established strong mechanistic candidates — most centrally hydrogen-related defect chemistry passivated by wet-O2 annealing, together with proposed intrinsic centers such as undercoordinated-indium $(\text{In}^{*}\text{–M})^{2-}$ traps, peroxide $\text{O}_{2}^{2-}$ states, and acceptor-like metal vacancies — but no direct atomic-scale identification of either defect class has been confirmed, and competing models coexist. The open problem is to determine the chemical, structural, and spatial identity of the deep acceptor-like charged traps and of the stress-induced shallow subgap states, the atomistic mechanism by which dry and wet O2 annealing suppress them, and the generality of this two-defect picture across a-IGZO compositions, deposition methods, and device architectures.

Background

Amorphous In–Ga–Zn–O (a-IGZO) is an amorphous oxide semiconductor, a disordered cation-sublattice oxide in which carrier transport occurs through spherically symmetric metal ns orbitals and remains band-like despite the lack of crystalline order. TFTs (thin-film transistors, i.e. field-effect transistors built as thin films with a gate dielectric controlling a channel layer) based on a-IGZO have become the standard backplane transistor for high-resolution active-matrix displays because of their high field-effect mobility, low off-current, and the ability to deposit the channel at low temperature. A long-standing practical limitation is the instability of the threshold voltage $V_{\text{th}}$ (the gate voltage at which the channel turns on): under prolonged gate or current stress the transfer curve drifts, degrading display uniformity and, in emerging uses of a-IGZO in dynamic random-access memory (DRAM) and back-end-of-line (BEOL, i.e. chip-level wiring-level) integration, circuit timing.

The source experiment (Nomura et al., 2009) stressed room-temperature-deposited, bottom-gate a-IGZO TFTs with and without 400 °C anneals in dry or wet O2 (wet meaning O2 carrying water vapor) under constant-current stress (a fixed drain current forced through the channel while the gate voltage needed to sustain it is monitored). All devices drifted to more positive $V_{\text{th}}$, and the drift followed a stretched-exponential time dependence $\Delta V_{\text{th}}(t)=\Delta V_{\text{th0}}\{1-\exp[-(t/\tau)^{\beta}]\}$. Annealed devices showed a small (1–2 V) strictly parallel shift of the transfer curve with unchanged subthreshold swing $S$ (the gate-voltage change per decade of drain current, a measure of how sharply the device switches), while unannealed devices showed large shifts (~4–10 V) together with a degradation of $S$. Device simulation using a two-step subgap density-of-states (DOS, the distribution of localized electronic states inside the band gap) model showed that adding acceptor-like deep traps — states that are neutral when empty and negatively charged when they capture an electron — concentrated at 1.0 eV below the conduction-band edge $E_{C}$ reproduces a parallel shift without degrading $S$, whereas the unannealed behavior additionally requires an increase of shallow subgap states near $E_{C}$ that degrade $S$. Thermal desorption spectroscopy (TDS, in which evolved gas species are mass-analyzed while the sample is heated) of companion films showed desorption of Zn–O, hydrogen-related, and –OH-related species up to 400 °C, strongly suppressed by annealing, suggesting weak-bond breaking and desorption as precursors of the stress-generated states. Wet O2 annealing additionally reduced the activation energy $E_{\tau}$ of the stretched-exponential time constant, indicating a more ordered transport structure.

What the device-level electrical analysis and simulation could not do was identify the defects themselves: the atomic-scale chemical structure, charge state, and spatial location (bulk network defect, impurity complex, channel/gate-insulator interface state, or back-channel surface state) of the deep acceptor-like charged traps, and the microscopic identity of the nascent shallow states created by stress in unannealed films. Subsequent literature has produced a rich but unresolved set of candidates. Hydrogen, though not intentionally incorporated, emerged as central: some H species passivate deep electron traps (restored by wet-O2 annealing), H occupying oxygen-vacancy sites acts as a shallow donor, neutral H released from defect–hydrogen bonds near the gate insulator diffuses and controls positive-bias-stress kinetics, and hydrogen bistability between donor and trap configurations has been proposed to explain photo-bias-thermal instabilities. First-principles and hybrid-functional calculations have proposed intrinsic candidates: undercoordinated indium atoms forming a DX-like (displaced, deep, negatively charged, lattice-relaxed) electron-trap center $(\text{In}^{*}\text{–M})^{2-}$; hole-mediated peroxide $\text{O}_{2}^{2-}$ defects; deep levels from undercoordinated or chemically disordered oxygen (mainly in the lower half of the gap); and acceptor-like metal (Zn) vacancies. At the same time, state-of-the-art calculations show that the neutral oxygen vacancy in a-IGZO is a shallow donor rather than a deep acceptor, eliminating the most popular folk candidate. Direct spectroscopies applied to a-IGZO — photon-probe methods such as photoexcited charge-collection spectroscopy (PECCS, which maps trap DOS from the photon energy that liberates trapped channel charge), isothermal capacitance transient spectroscopy (ICTS, which resolves trap levels from capacitance transients after a filling pulse), and total photoelectron yield spectroscopy (TPYS, which detects sub-gap occupied states from photoemission thresholds) — measure the energetic distribution and, in some cases, the interfacial localization of traps, but none has delivered an unambiguous atomic-scale structure for the deep acceptor-like traps or the stress-created shallow states. The field's own review (Ide, Nomura, Hosono, and Kamiya, 2019) frames bias-stress instability as the most serious remaining issue of amorphous oxide semiconductor technology and attributes many of the relevant defects to oxygen- and hydrogen-related species, without a settled microscopic identification.

Problem Statement

Determine the microscopic origin of threshold-voltage instability in room-temperature-deposited a-In–Ga–Zn–O thin-film transistors, as posed by the device-level analysis of Nomura, Kamiya, Hirano and Hosono (2009) and left at the level of subgap density-of-states parameters. Specifically, provide direct microscopic identification of:

(1) The deep acceptor-like charged traps — the states held responsible for the small, strictly parallel positive threshold-voltage shifts (with unchanged subthreshold swing) of annealed a-IGZO TFTs under constant-current or positive gate bias stress, modeled in the source work as an interface trap density $N_{\text{ADT}}$ located $\sim 1.0\ \text{eV}$ below the conduction-band edge $E_{C}$ — namely their chemical identity and atomic structure, their charge states, and their spatial distribution (bulk network defects, impurity complexes such as hydrogen-related species, channel/gate-insulator interface states, or back-channel surface states).

(2) The shallow subgap states that are generated or activated by electrical stress in unannealed (room-temperature-deposited) a-IGZO films and that cause the large threshold-voltage shifts with subthreshold-swing degradation — namely the microscopic mechanism by which stress creates them, the role of weak bonds and desorbable Zn–O, H-related and –OH-related precursor species suggested by thermal desorption data, and their relationship to the hydrogen-related defect chemistry established in later work (trap–hydrogen bond dissociation, hydrogen migration, hydrogen passivation).

(3) The atomistic mechanism by which 400 °C O2 annealing — especially wet O2 — suppresses both defect classes and produces the more ordered transport structure (smaller activation energy $E_{\tau}$ of the stress time constant), and whether the resulting two-defect-class picture generalizes across a-IGZO compositions, deposition methods, and device architectures (including passivated versus unpassivated back channels).

The scientific objective is to replace the device-level, effective subgap-DOS description of a-IGZO bias-stress instability with a verified atomic-scale defect model: to identify the defect structures — including the charge states and spatial distribution of the deep acceptor-like traps — and their formation/passivation reactions, to establish whether the resulting two-defect-class picture generalizes across a-IGZO compositions, deposition methods, and device architectures, and to show that the identified defects quantitatively reproduce the observed stress and annealing behavior in at least the original (In:Ga:Zn = 1:1:1, room-temperature-deposited, 400 °C O2-annealed) system. The related engineering goal of achieving comparable stability at processing temperatures compatible with flexible substrates has been addressed empirically in later literature and is part of the lineage rather than the target of this problem.

References

1. Kenji Nomura, Toshio Kamiya, Masahiro Hirano, Hideo Hosono, 'Origins of threshold voltage shifts in room-temperature deposited and annealed a-In–Ga–Zn–O thin-film transistors', Applied Physics Letters (2009), DOI 10.1063/1.3159831, https://doi.org/10.1063/1.3159831
2. Kenji Nomura, Toshio Kamiya, Hiromichi Ohta, Masahiro Hirano, Hideo Hosono, 'Defect passivation and homogenization of amorphous oxide thin-film transistor by wet O2 annealing', Applied Physics Letters (2008), DOI 10.1063/1.3020714, https://doi.org/10.1063/1.3020714
3. Kenji Nomura, Toshio Kamiya, Hideo 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 (2010), DOI 10.1889/jsid18.10.789, https://doi.org/10.1889/jsid18.10.789
4. Kenji Nomura, Toshio Kamiya, Hideo Hosono, 'Highly stable amorphous In-Ga-Zn-O thin-film transistors produced by eliminating deep subgap defects', Applied Physics Letters (2011), DOI 10.1063/1.3622121, https://doi.org/10.1063/1.3622121
5. Kenji Nomura, Toshio Kamiya, Hideo Hosono, 'Effects of Diffusion of Hydrogen and Oxygen on Electrical Properties of Amorphous Oxide Semiconductor, In-Ga-Zn-O', ECS Journal of Solid State Science and Technology (2012), DOI 10.1149/2.011301jss, https://doi.org/10.1149/2.011301jss
6. Yuichiro Hanyu, Kay Domen, Kenji Nomura, Hidenori Hiramatsu, Hideya Kumomi, Hideo Hosono, Toshio Kamiya, 'Hydrogen passivation of electron trap in amorphous In-Ga-Zn-O thin-film transistors', Applied Physics Letters 103, 202114 (2013), DOI 10.1063/1.4832076, https://doi.org/10.1063/1.4832076
7. Kay Domen, Takaya Miyase, Katsumi Abe, Hideo Hosono, Toshio Kamiya, 'Positive Gate Bias Instability Induced by Diffusion of Neutral Hydrogen in Amorphous In-Ga–Zn-O Thin-Film Transistor', IEEE Electron Device Letters (2014), DOI 10.1109/led.2014.2327234, https://doi.org/10.1109/led.2014.2327234
8. Takaya Miyase, Ken Watanabe, Isao Sakaguchi, Naoki Ohashi, Kay Domen, Kenji Nomura, Hidenori Hiramatsu, Hideya Kumomi, Hideo Hosono, Toshio Kamiya, 'Roles of Hydrogen in Amorphous Oxide Semiconductor In-Ga-Zn-O: Comparison of Conventional and Ultra-High-Vacuum Sputtering', ECS Journal of Solid State Science and Technology (2014), DOI 10.1149/2.015409jss, https://doi.org/10.1149/2.015409jss
9. Ken Watanabe, Dong-Hee Lee, Isao Sakaguchi, Kenji Nomura, Toshio Kamiya, Hajime Haneda, Hideo Hosono, Naoki Ohashi, 'Surface reactivity and oxygen migration in amorphous indium-gallium-zinc oxide films annealed in humid atmosphere', Applied Physics Letters (2013), DOI 10.1063/1.4829996, https://doi.org/10.1063/1.4829996
10. Kazushi Hayashi, Aya Hino, Shinya Morita, Satoshi Yasuno, Hiroshi Okada, Toshihiro Kugimiya, 'Electron traps in amorphous In–Ga–Zn–O thin films studied by isothermal capacitance transient spectroscopy', Applied Physics Letters (2012), DOI 10.1063/1.3691923, https://doi.org/10.1063/1.3691923
11. Kazushi Hayashi, Aya Hino, Hiroaki Tao, Mototaka Ochi, Hiroshi Goto, Toshihiro Kugimiya, 'Photoelectron emission yield experiments on evolution of sub-gap states in amorphous In-Ga-Zn-O thin films with post deposition hydrogen treatment', Applied Physics Letters (2015), DOI 10.1063/1.4931143, https://doi.org/10.1063/1.4931143
12. Wolfgang Körner, Daniel F. Urban, Christian Elsässer, 'Origin of subgap states in amorphous In-Ga-Zn-O', Applied Physics Letters (2013), DOI 10.1063/1.4826895, https://doi.org/10.1063/1.4826895
13. S. Sallis, K. T. Butler, N. F. Quackenbush, D. S. Williams, M. Junda, D. A. Fischer, J. C. Woicik, N. J. Podraza, B. E. White, A. Walsh, L. F. J. Piper, 'Origin of deep subgap states in amorphous indium gallium zinc oxide: Chemically disordered coordination of oxygen', Applied Physics Letters (2014), DOI 10.1063/1.4883257, https://doi.org/10.1063/1.4883257
14. Ho-Hyun Nahm, Yong-Sung Kim, 'Undercoordinated indium as an intrinsic electron-trap center in amorphous InGaZnO4', NPG Asia Materials (2014), DOI 10.1038/am.2014.103, https://doi.org/10.1038/am.2014.103
15. Ho-Hyun Nahm, Yong-Sung Kim, Dae Hwan Kim, 'Instability of amorphous oxide semiconductors via carrier-mediated structural transition between disorder and peroxide state', Physica Status Solidi B (2012), DOI 10.1002/pssb.201147557, https://doi.org/10.1002/pssb.201147557
16. Youngho Kang, Byung Du Ahn, Ji Hun Song, Yeon Gon Mo, Ho-Hyun Nahm, Seungwu Han, Jae Kyeong Jeong, 'Hydrogen Bistability as the Origin of Photo-Bias-Thermal Instabilities in Amorphous Oxide Semiconductors', Advanced Electronic Materials (2015), DOI 10.1002/aelm.201400006, https://doi.org/10.1002/aelm.201400006
17. Motoki Nakashima, Masashi Oota, Noritaka Ishihara, Yusuke Nonaka, Takuya Hirohashi, Masahiro Takahashi, Shunpei Yamazaki, Toshimitsu Obonai, Yasuharu Hosaka, Junichi Koezuka, 'Origin of major donor states in In–Ga–Zn oxide', Applied Physics Letters (2014), DOI 10.1063/1.4902859, https://doi.org/10.1063/1.4902859
18. Youn-Gyoung Chang, Dae Hwan Kim, Gunwoo Ko, Kimoon Lee, Jae Hoon Kim, Seongil Im, 'Trap Density of States Measured by Photon Probe on Amorphous-InGaZnO Thin-Film Transistors', IEEE Electron Device Letters (2011), DOI 10.1109/LED.2010.2102739, https://doi.org/10.1109/LED.2010.2102739
19. Md Delwar Hossain Chowdhury, Piero Migliorato, Jin Jang, 'Light induced instabilities in amorphous indium–gallium–zinc-oxide thin-film transistors', Applied Physics Letters (2010), DOI 10.1063/1.3503971, https://doi.org/10.1063/1.3503971
20. Keisuke Ide, Kenji Nomura, Hideo Hosono, Toshio Kamiya, 'Electronic Defects in Amorphous Oxide Semiconductors: A Review', Physica Status Solidi A (2019), DOI 10.1002/pssa.201800372, https://doi.org/10.1002/pssa.201800372
21. Kyle T. Vogt, Christopher E. Malmberg, Jacob C. Buchanan, George W. Mattson, G. Mirek Brandt, Dylan B. Fast, Paul H.-Y. Cheong, John F. Wager, Matt W. Graham, 'Ultrabroadband density of states of amorphous In-Ga-Zn-O', Physical Review Research 2, 033358 (2020), DOI 10.1103/PhysRevResearch.2.033358, https://doi.org/10.1103/PhysRevResearch.2.033358
22. Hochul Song, Gijae Kang, Youngho Kang, Seungwu Han, 'The Nature of the Oxygen Vacancy in Amorphous Oxide Semiconductors: Shallow Versus Deep', Physica Status Solidi B (2018), DOI 10.1002/pssb.201800486, https://doi.org/10.1002/pssb.201800486
23. A. de Jamblinne de Meux, G. Pourtois, J. Genoe, P. Heremans, 'Defects in Amorphous Semiconductors: The Case of Amorphous Indium Gallium Zinc Oxide', Physical Review Applied 9, 054039 (2018), DOI 10.1103/PhysRevApplied.9.054039, https://doi.org/10.1103/PhysRevApplied.9.054039
24. Chen Wang, Wenmo Lu, Fengnan Li, Haiyue Ning, Fei Ma, 'Persistent photoconductivity in a-IGZO thin films induced by trapped electrons and metastable donors', Applied Physics Letters (2022), DOI 10.1063/5.0086772, https://doi.org/10.1063/5.0086772
25. Michael Paul A. Jallorina, Juan Paolo S. Bermundo, Mami N. Fujii, Yasuaki Ishikawa, Yukiharu Uraoka, 'Significant mobility improvement of amorphous In-Ga-Zn-O thin-film transistors annealed in a low temperature wet ambient environment', Applied Physics Letters (2018), DOI 10.1063/1.5026515, https://doi.org/10.1063/1.5026515
26. A. Kruv, M. J. van Setten, A. Chasin, D. Matsubayashi, H. F. W. Dekkers, A. Pavel, Y. Wan, K. Trivedi, N. Rassoul, J. Li, Y. Jiang, S. Subhechha, G. Pourtois, A. Belmonte, G. Sankar Kar, 'In-poor IGZO: superior resilience to hydrogen in forming gas anneal and PBTI' (2024), arXiv:2412.07362, DOI 10.48550/arXiv.2412.07362, https://arxiv.org/abs/2412.07362
27. Hyunsung Cho, Minseok Moon, Jaehoon Kim, Eunkyung Koh, Hyeon-Deuk Kim, Rokyeon Kim, Gyehyun Park, Seungwu Han, Youngho Kang, 'Atomistic insights into hydrogen migration in IGZO from machine-learning interatomic potential: linking atomic diffusion to device performance' (2025), arXiv:2508.19674, DOI 10.48550/arXiv.2508.19674, https://arxiv.org/abs/2508.19674
28. Jie Luo, Yunjiao Bao, Yanyu Yang, Yupeng Lu, Guilei Wang, Gaobo Xu, Huaxiang Yin, Chao Zhao, Jun Luo, 'The Investigation of the Hysteresis and Reliability Mechanism of Amorphous Oxide Semiconductor Thin-Film Transistors Applied in Dynamic Random Access Memory', IEEE Transactions on Electron Devices (2025), DOI 10.1109/TED.2025.3545007, https://doi.org/10.1109/TED.2025.3545007