Hard x-ray photoelectron spectroscopy (HX-PES) on heavily doped amorphous In–Ga–Zn–O (a-IGZO) revealed a large occupied subgap density of states (DOS) of order 102010^{20}–1021 cm−310^{21}\,\mathrm{cm^{-3}} immediately above the valence-band maximum (VBM), which was proposed as the reason a-IGZO thin-film transistors show hard saturation in the off state and cannot operate as p-channel inversion devices. Later work has settled adjacent parts of the question — the identity of the near-conduction-band spectral weight (hydrogen occupying oxygen-vacancy sites and donor-like oxygen vacancies), the exclusion of reduced indium as the deep near-VBM feature, and improved cross-section-insensitive density-of-states measurement — but the central question remains open: competing, mutually incompatible atomic-scale attributions for the near-VBM occupied subgap states (oxygen-vacancy-derived deep states, hydrogen-related metal–H–metal bridges and hydroxyl bonds, weakly bound or excess oxygen, undercoordinated oxygen, zinc vacancies, and generic amorphous structural disorder) still coexist in the literature through 2025, no processing route has been shown to reduce these states to levels compatible with p-channel inversion, and no cross-validated absolute calibration of the occupied subgap density exists.

Background

Amorphous In–Ga–Zn–O (a-IGZO) is a transparent amorphous oxide semiconductor, the active layer of the thin-film transistors (TFTs) — field-effect devices in which a gate voltage controls the conductivity of a semiconductor film — that drive current flat-panel displays. Unlike amorphous silicon, a-IGZO has a conduction band derived from spatially extended metal ss orbitals, so electrons retain high mobility (>10 cm2V−1s−1>10\,\mathrm{cm^2V^{-1}s^{-1}}) even in the amorphous network; its valence band, by contrast, is derived from localized oxygen 2p2p orbitals. In 2008 Nomura, Kamiya, Yanagi, Ikenaga, Yang, Kobayashi, Hirano, and Hosono (Applied Physics Letters 92, 202117) combined optical absorption with hard x-ray photoelectron spectroscopy (HX-PES, photoemission excited at photon energies of several keV so that the escape depth of the electrons, a few nanometers, makes the probe sensitive to the film bulk rather than its surface) to study highly doped a-IGZO films. They reported Fermi-edge structures (the sharp photoemission edge of a degenerate electron gas) and extra occupied subgap densities of states (DOS, states lying inside the band gap that should be empty in a perfect semiconductor) with densities above 5×1020 cm−35\times10^{20}\,\mathrm{cm^{-3}} immediately above the valence-band maximum (VBM, the energy of the top of the valence band) and above 5×1019 cm−35\times10^{19}\,\mathrm{cm^{-3}} just below the Fermi level; tail-like optical absorption was traced to these near-VBM states. The authors proposed that the near-VBM subgap DOS explains why a-IGZO TFTs saturate poorly ('hard saturation') in the off state and why a p-channel inversion-mode a-IGZO device — a TFT whose channel is inverted to hole conduction, the complement needed for CMOS-type circuits (complementary metal–oxide–semiconductor logic, which requires both n- and p-channel transistors) — has never been realized. The paper explicitly left open: (i) the absolute density values rely on a calibration factor transferred from a heavily doped conduction-band reference, even though photoionization cross sections and matrix elements differ between the In 5s5s-derived conduction states and the valence-side subgap states; (ii) the microscopic chemical and structural origin of the large occupied near-VBM subgap DOS is unidentified; (iii) the microscopic identity of the extra spectral weight at the conduction-band edge (filled impurity band, heavily doped conduction-band tail, or cluster-related states) is unproven; (iv) thermal annealing reduces the subgap DOS amplitude while leaving its spectral width and energy distribution nearly unchanged, for unknown reasons; and (v) no processing route reduces the valence-side subgap DOS sufficiently to enable p-channel inversion. A decade and a half of follow-up work has constrained this problem in several ways. DFT (density functional theory) studies of the amorphous network proposed oxygen deficiencies forming deep fully-occupied localized states near the VBM (Kamiya, Nomura, Hosono, 2010), undercoordinated oxygen atoms creating levels up to about 1.5 eV1.5\,\mathrm{eV} above the valence-band edge together with metal–metal pairs causing upper-gap levels (Körner, Urban, Elsässer, 2013), and hydrogen pairing at oxygen vacancies into metal–H–metal bridge bonds with filled gap states just above the VBM (Li, Guo, Robertson, 2018). Experimentally, the near-VBM DOS grows with water-vapor or residual-hydrogen exposure during sputtering (Oka et al., 2012; Tang et al., 2015) and with room-temperature hydrogen plasma that incorporates hydrogen as hydroxyl (–OH) bonds (Tang et al., 2017); reduced indium (In+\mathrm{In}^+ lone-pair states) was ruled out as the deep near-VBM feature (Sallis et al., 2015); ultrabroadband photoconduction spectroscopy with DFT assignment found deep acceptor-like peaks — attributed to zinc vacancies — adjacent to the valence-band Urbach tail (the exponentially decaying band-edge disorder tail, probed by Vogt et al., 2020); and a bulk single-crystal InGaZnO4\mathrm{InGaZnO_4} HX-PES comparison found the near-VBM subgap states to be far weaker in the crystal than in the amorphous films, implicating loss of crystallinity itself (Shibata et al., 2025). On the measurement side, constant-final-state yield spectroscopy (CFS-YS, a photoemission-yield method whose signal does not require photoionization-cross-section calibration) now gives the occupied in-gap DOS over six orders of magnitude (Nakazawa et al., 2024), directly addressing the original calibration concern. For the device-level thread, first-principles work showed that holes injected into the valence-band tail states self-trap as polarons (a carrier localized by the lattice distortion it induces) with large self-trapping energies (de Jamblinne de Meux et al., 2017), a hole-inversion layer was sustained in an IGZO-based TFT only with the aid of a high-permittivity tunnel dielectric (Shoute et al., 2016), and the amorphous p-channel transistor finally realized in 2024 used selenium-alloyed tellurium oxide rather than a-IGZO (Liu et al., 2024) — so p-channel operation of a-IGZO itself remains unachieved.

Problem Statement

What is the microscopic chemical and structural origin of the large occupied subgap density of states (of order 1020 cm−310^{20}\,\mathrm{cm^{-3}}) immediately above the valence-band maximum of amorphous In–Ga–Zn–O — the states whose photoemission signature was reported by Nomura et al. (2008) — and can they be quantified and suppressed? Concretely: (1) determine, at the atomic scale, whether these occupied valence-side subgap states arise from oxygen-vacancy-related deep states, hydrogen-related configurations (hydroxyl bonds or hydrogen pairing at oxygen vacancies into metal–H–metal bridge bonds), weakly bound or excess oxygen, undercoordinated oxygen, zinc vacancies, or the amorphous structural disorder itself, and quantitatively account for the measured occupied density; (2) establish an absolute, cross-section-corrected determination of the occupied subgap state density that does not rely on transferring a calibration factor from a heavily doped conduction-band reference, and reconcile hard x-ray photoemission with optical and photoconduction-based density-of-states determinations; (3) identify the processing conditions (annealing ambient, hydrogen and oxygen control during and after deposition) that suppress these states, and explain why thermal annealing changes their amplitude while leaving their spectral width and energy distribution nearly unchanged; and (4) establish whether reducing the valence-side subgap density of states is sufficient — given the competing obstacle of hole self-trapping in the valence-band tail — to enable p-channel inversion operation of a-IGZO thin-film transistors. Parts (1)–(3) are jointly posed and coupled, none of them settled independently of the others; together they constitute the single scientific objective of identifying and controlling the near-valence-band-edge defect states of a-IGZO — the question the 2008 HX-PES study raised and left open and which the field had not closed as of 2025 — while part (4) is a downstream device-level corollary of that objective rather than a substitute for it.

References

  1. K. Nomura, T. Kamiya, H. Yanagi, E. Ikenaga, K. Yang, K. Kobayashi, M. Hirano, H. Hosono, 'Subgap states in transparent amorphous oxide semiconductor, In–Ga–Zn–O, observed by bulk sensitive x-ray photoelectron spectroscopy', Applied Physics Letters 92, 202117 (2008). DOI: 10.1063/1.2927306, https://doi.org/10.1063/1.2927306
  2. K. Nomura, T. Kamiya, E. Ikenaga, H. Yanagi, K. Kobayashi, H. Hosono, 'Depth analysis of subgap electronic states in amorphous oxide semiconductor, a-In-Ga-Zn-O, studied by hard x-ray photoelectron spectroscopy', Journal of Applied Physics 109, 073726 (2011). DOI: 10.1063/1.3560769, https://doi.org/10.1063/1.3560769
  3. T. Kamiya, K. Nomura, H. Hosono, 'Present status of amorphous In–Ga–Zn–O thin-film transistors', Science and Technology of Advanced Materials 11, 044305 (2010). DOI: 10.1088/1468-6996/11/4/044305, https://doi.org/10.1088/1468-6996/11/4/044305
  4. T. Kamiya, K. Nomura, H. Hosono, 'Subgap states, doping and defect formation energies in amorphous oxide semiconductor a-InGaZnO4 studied by density functional theory', physica status solidi (a) 207, 1698 (2010). DOI: 10.1002/pssa.200983772, https://doi.org/10.1002/pssa.200983772
  5. W. Körner, D. F. Urban, C. Elsässer, 'Origin of subgap states in amorphous In-Ga-Zn-O', Journal of Applied Physics 114, 163704 (2013). DOI: 10.1063/1.4826895, https://doi.org/10.1063/1.4826895
  6. N. Oka, T. Aoi, R. Hayashi, H. Kumomi, Y. Shigesato, 'Electronic State of Amorphous Indium Gallium Zinc Oxide Films Deposited by DC Magnetron Sputtering with Water Vapor Introduction', Applied Physics Express 5, 075802 (2012). DOI: 10.1143/apex.5.075802, https://doi.org/10.1143/apex.5.075802
  7. H. Tang, K. Ishikawa, K. Ide, H. Hiramatsu, S. Ueda, N. Ohashi, H. Kumomi, H. Hosono, T. Kamiya, 'Effects of residual hydrogen in sputtering atmosphere on structures and properties of amorphous In-Ga-Zn-O thin films', Journal of Applied Physics 118, 205703 (2015). DOI: 10.1063/1.4936552, https://doi.org/10.1063/1.4936552
  8. H. Tang, Y. Kishida, K. Ide, Y. Toda, H. Hiramatsu, S. Matsuishi, S. Ueda, N. Ohashi, H. Kumomi, H. Hosono, T. Kamiya, 'Multiple Roles of Hydrogen Treatments in Amorphous In–Ga–Zn–O Films', ECS Journal of Solid State Science and Technology 6, P365 (2017). DOI: 10.1149/2.0071707jss, https://doi.org/10.1149/2.0071707jss
  9. S. Sallis, N. F. Quackenbush, D. S. Williams, M. Senger, J. C. Woicik, B. E. White, L. F. J. Piper, 'Deep subgap feature in amorphous indium gallium zinc oxide: Evidence against reduced indium', physica status solidi (a) 212, 1471 (2015). DOI: 10.1002/pssa.201431806, https://doi.org/10.1002/pssa.201431806
  10. P. Migliorato, Md. D. H. Chowdhury, J. G. Um, M. Seok, J. Jang, 'Light/negative bias stress instabilities in indium gallium zinc oxide thin film transistors explained by creation of a double donor', Applied Physics Letters 101, 123502 (2012). DOI: 10.1063/1.4752238, https://doi.org/10.1063/1.4752238
  11. M. Nakashima, M. Oota, N. Ishihara, Y. Nonaka, T. Hirohashi, M. Takahashi, S. Yamazaki, T. Obonai, Y. Hosaka, J. Koezuka, 'Origin of major donor states in In–Ga–Zn oxide', Journal of Applied Physics 116, 213703 (2014). DOI: 10.1063/1.4902859, https://doi.org/10.1063/1.4902859
  12. K. T. Vogt, C. E. Malmberg, J. C. Buchanan, G. W. Mattson, G. M. Brandt, D. B. Fast, P. H.-Y. Cheong, J. F. Wager, M. 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
  13. H. Li, Y. Guo, J. Robertson, 'Oxygen vacancies and hydrogen in amorphous In-Ga-Zn-O and ZnO', Physical Review Materials 2, 074601 (2018). DOI: 10.1103/PhysRevMaterials.2.074601, https://doi.org/10.1103/PhysRevMaterials.2.074601
  14. R. Nakazawa, A. Matsuzaki, K. Shimizu, I. Nakamura, E. Kawashima, S. Makita, K. Tanaka, S. Yasuno, H. Sato, H. Yoshida, M. Abdi-Jalebi, S. D. Stranks, S. Tadano, P. Krüger, Y. Tanaka, H. Tokairin, H. Ishii, 'Reliable measurement of the density of states including occupied in-gap states of an amorphous In–Ga–Zn–O thin film via photoemission spectroscopies: Direct observation of light-induced in-gap states', Journal of Applied Physics 135, 085301 (2024). DOI: 10.1063/5.0185405, https://doi.org/10.1063/5.0185405
  15. G. Shibata, Y. Takahashi, M. Okawa, A. Yasui, Y. Takagi, Y. Kawamura, N. Miyakawa, N. Kase, N. Hamada, T. Saitoh, 'Hard x-ray photoemission study of bulk single-crystalline InGaZnO4', Applied Physics Letters 126, 241902 (2025). DOI: 10.1063/5.0271655, https://doi.org/10.1063/5.0271655
  16. A. de Jamblinne de Meux, G. Pourtois, J. Genoe, P. Heremans, 'Effects of hole self-trapping by polarons on transport and negative bias illumination stress in amorphous-IGZO', Journal of Applied Physics 123, 161513 (2017). DOI: 10.1063/1.4986180, https://doi.org/10.1063/1.4986180
  17. G. Shoute, A. Afshar, T. Muneshwar, K. Cadien, D. Barlage, 'Sustained hole inversion layer in a wide-bandgap metal-oxide semiconductor with enhanced tunnel current', Nature Communications 7, 10632 (2016). DOI: 10.1038/ncomms10632, https://doi.org/10.1038/ncomms10632
  18. A. Liu, Y.-S. Kim, M. G. Kim, Y. Reo, T. Zou, T. Choi, S. Bai, H. Zhu, Y.-Y. Noh, 'Selenium-alloyed tellurium oxide for amorphous p-channel transistors', Nature 629, 798 (2024). DOI: 10.1038/s41586-024-07360-w, https://doi.org/10.1038/s41586-024-07360-w