Reconciling computed band effective masses with measured carrier masses and element-specific structure in amorphous In-Ga-Zn-O: electron-phonon/many-body renormalization and medium-range order

Amorphous In-Ga-Zn-O ($a$-IGZO) is the workhorse semiconductor of oxide thin-film electronics, yet a quantitative gap identified in the founding structure-electronic-structure study of the material remains open. Nomura et al. (Phys. Rev. B 75, 035212, 2007) combined extended x-ray absorption fine structure (EXAFS) measurements with an LDA-relaxed 84-atom amorphous model and concluded that edge-sharing In-O polyhedra survive amorphization and that In 5s orbitals form an extended conduction-band edge with a band effective mass of $m^{*}\approx 0.2m_{e}$, markedly smaller than carrier masses of roughly $0.34m_{e}$ inferred from free-carrier optical analysis of degenerate films, a difference the authors attributed plausibly to polaronic or many-body mass enhancement without computing it. Later first-principles work on larger, independently generated melt-quench models and beyond-LDA functionals keeps the computed band mass at $0.21$-$0.28m_{e}$, while optical and transport experiments continue to return $\sim 0.33$-$0.36m_{e}$ for stoichiometric films, so the discrepancy persists as a mechanism question rather than an artifact of one functional or one cell. The open problem is to give a quantitative, atomistically grounded account of the carrier-mass renormalization in $a$-IGZO (or a demonstrated alternative explanation), and to close the associated structural uncertainty that the 2007 study itself flagged: EXAFS could not unambiguously distinguish Ga from Zn coordination environments or resolve medium-range topology in the multicomponent amorphous network. A solution would fix the correct effective-mass input for transport and device models of $a$-IGZO and settle whether electron-phonon coupling in this polar amorphous oxide produces measurably heavier quasiparticles.

Background

In-Ga-Zn-O is an ionic amorphous oxide semiconductor (IAOS): a multicomponent oxide (In:Ga:Zn roughly 1:1:1 with O) that can be deposited as a thin film at room temperature yet, unlike covalent amorphous semiconductors such as a-Si:H, supports degenerate electron concentrations and large room-temperature mobilities. Deposited in the amorphous phase ($a$-IGZO), it became the channel material of thin-film transistors (TFTs) used in large-area displays after Nomura et al. demonstrated room-temperature fabrication of transparent flexible TFTs with field-effect mobilities of about 10 cm^2 V^-1 s^-1 (Nature 432, 488, 2004). Understanding why an amorphous network can carry high-mobility extended electrons requires connecting three things: the local atomic structure, the conduction-band-edge electronic structure, and the carrier masses measured in transport and optics.

The structural probe in the founding study of $a$-IGZO structure was extended x-ray absorption fine structure (EXAFS): the oscillatory part of the x-ray absorption coefficient beyond an absorption edge, which is sensitive to distances and coordination numbers of the atoms surrounding the absorbing element. By measuring In, Ga, and Zn K-edges and comparing to a crystalline $InGaZnO_{4}$ reference, Nomura et al. (2007) extracted nearest-neighbor metal-oxygen distances and coordination numbers (In-O about 4.5, Zn-O about 4.6, Ga-O about 4.3, versus 6, 5, and 5 in the crystal) and, combined with an atomistic model, concluded that edge-sharing connections between In-O polyhedra survive amorphization. The model was built by classical melt-quench molecular dynamics (MD) in a 1890-atom cell, reduced to an 84-atom cell, and relaxed within density functional theory in the local density approximation (LDA), a semilocal exchange-correlation approximation known to underestimate band gaps and to misdescribe self-trapping (polaronic) effects. On this model, the conduction-band minimum was computed to be dominated by spatially extended In 5s orbitals hybridized with O 2p, with no localized tail states at the edge, and with a band effective mass $m^{*}\approx 0.2m_{e}$ (where $m_{e}$ is the free-electron mass and the effective mass is the parabolic curvature of the band edge that controls carrier acceleration and hence mobility).

The effective mass is also measurable: in degenerately doped films, free-carrier (Drude) optical absorption and the Burstein-Moss shift of the optical absorption edge (a filling of conduction-band states that blue-shifts the edge) both yield the carrier mass, giving values around $0.33$-$0.36m_{e}$ for $a$-IGZO, essentially the same as in crystalline $InGaZnO_{4}$. The 2007 study noted that this measured value is substantially larger than its computed band mass and attributed the difference plausibly to enhancement of the carrier mass by electron-phonon and many-body interactions: in a polar (partly ionic) material an electron couples to longitudinal-optical (LO) lattice vibrations through the Fröhlich interaction, and a large polaron (an electron dressed by a self-induced polarization cloud, still mobile and dispersive) is heavier than the bare band electron. The paper did not compute this renormalization, and it also acknowledged that EXAFS in a multicomponent amorphous system is intrinsically ambiguous for medium-range order (ordering beyond the first coordination shell) and could not unambiguously distinguish the subtly different Zn and Ga coordination environments, and that the 84-atom LDA-relaxed cell is a finite-size, semilocal-functional approximation.

Later literature has substantially rebuilt the modeling side: ensembles of independent melt-quench supercells computed with generalized-gradient (PBE), LDA+U, and hybrid (HSE) functionals reproduce the extended In-5s conduction edge and give band masses of $0.21$-$0.28m_{e}$, so the small computed band mass is robust across methods, while experiments on nonstoichiometric films return masses up to three times the stoichiometric value of about $0.34m_{e}$, consistently between optical and thermopower/Hall analyses. What has not been done is a quantitative first-principles calculation of electron-phonon and many-body renormalization for a realistic amorphous IGZO model (methods of this type have recently been demonstrated on crystalline anatase TiO$_2$, giving a polaron mass enhancement factor of 1.6), or an experiment that resolves Ga- versus Zn-specific coordination and medium-range topology in sputtered $a$-IGZO with an element-specific probe such as anomalous x-ray scattering, which has so far been applied to the related ternary a-In-Ga-O. Candidate mechanisms for the discrepancy include polaronic and many-body mass renormalization, conduction-band nonparabolicity at the degenerate carrier densities of the measured films, disorder-induced occupation of tail states, and systematics of the free-carrier optical analysis; none of these contributions has been quantified separately for $a$-IGZO.

Problem Statement

For amorphous In-Ga-Zn-O, close the quantitative gap between the atomistically validated band structure and the measured carrier response that the founding EXAFS-plus-ab-initio study of the material identified: (1) determine whether the carrier mass inferred from optical and transport measurements ($\sim 0.33$-$0.36m_{e}$ in stoichiometric films) is genuinely enhanced relative to the computed conduction-band mass ($0.21$-$0.28m_{e}$), and by how much - i.e., whether conduction electrons in this polar amorphous oxide are large polarons with heavier quasiparticle masses than the bare In-5s band electrons - and identify the mechanism that quantitatively accounts for the full magnitude of the measured-versus-computed discrepancy; and (2) resolve the structural ambiguity that limits the structural side of the structure-mass connection, namely the Ga- versus Zn-specific coordination environments and the medium-range topology (persistence and connectivity of edge-sharing In-O polyhedra) of sputtered $a$-IGZO, to a resolution beyond what the 2007 EXAFS analysis could establish. The two parts are posed jointly and bear on the same structure-mass connection that the 2007 study left open.

References

1. K. Nomura, T. Kamiya, H. Ohta, T. Uruga, M. Hirano, H. Hosono, Local coordination structure and electronic structure of the large electron mobility amorphous oxide semiconductor In-Ga-Zn-O: Experiment and ab initio calculations, Phys. Rev. B 75, 035212 (2007). DOI: 10.1103/PhysRevB.75.035212, https://doi.org/10.1103/PhysRevB.75.035212
2. K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, H. Hosono, Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors, Nature 432, 488-492 (2004). DOI: 10.1038/nature03090, https://doi.org/10.1038/nature03090
3. A. de Jamblinne de Meux, G. Pourtois, J. Genoe, P. Heremans, Comparison of the electronic structure of amorphous versus crystalline indium gallium zinc oxide semiconductor: structure, tail states and strain effects, J. Phys. D: Appl. Phys. 48, 435104 (2015). DOI: 10.1088/0022-3727/48/43/435104, https://doi.org/10.1088/0022-3727/48/43/435104
4. J. Srivastava, S. Nahas, S. Bhowmick, A. Gaur, Electronic structure and transport in amorphous metal oxide and amorphous metal oxynitride semiconductors, J. Appl. Phys. 126, 125702 (2019). DOI: 10.1063/1.5096042, https://doi.org/10.1063/1.5096042
5. X. Zhu, Y. Yang, X.-D. Liu, Z.-Q. Li, Electron effective mass and electronic structure in nonstoichiometric amorphous Indium Gallium Zinc Oxide films, J. Alloys Compd. 813, 152183 (2020). DOI: 10.1016/j.jallcom.2019.152183, https://doi.org/10.1016/j.jallcom.2019.152183
6. Y. Kang, Y. Cho, S. Han, Cation disorder as the major electron scattering source in crystalline InGaZnO, Appl. Phys. Lett. 102, 152104 (2013). DOI: 10.1063/1.4802093, https://doi.org/10.1063/1.4802093
7. X. Wang, L. F. Register, A. Dodabalapur, Redefining the Mobility Edge in Thin-Film Transistors, Phys. Rev. Applied 11, 064039 (2019). DOI: 10.1103/PhysRevApplied.11.064039, https://doi.org/10.1103/PhysRevApplied.11.064039
8. 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, J. Appl. Phys. 123, 161513 (2017). DOI: 10.1063/1.4986180, https://doi.org/10.1063/1.4986180 (preprint: arXiv:1709.03728, https://arxiv.org/abs/1709.03728)
9. Y. Kang, H. Song, H.-H. Nahm, S. H. Jeon, Y. Cho, S. Han, Intrinsic nature of visible-light absorption in amorphous semiconducting oxides, APL Mater. 2, 032108 (2014). DOI: 10.1063/1.4868175, https://doi.org/10.1063/1.4868175
10. S. L. Moffitt, Q. Zhu, Q. Ma, A. F. Falduto, D. B. Buchholz, R. P. H. Chang, T. O. Mason, J. E. Medvedeva, T. J. Marks, M. J. Bedzyk, Probing the Unique Role of Gallium in Amorphous Oxide Semiconductors through Structure-Property Relationships, Adv. Electron. Mater. 3, 1700189 (2017). DOI: 10.1002/aelm.201700189, https://doi.org/10.1002/aelm.201700189
11. C. Revenant, M. Benwadih, M. Maret, Self-organized nanoclusters in solution-processed mesoporous In-Ga-Zn-O thin films, Chem. Commun. 51, 1218-1221 (2015). DOI: 10.1039/c4cc08521c, https://doi.org/10.1039/c4cc08521c
12. H.-H. Nahm, Y.-S. Kim, D. H. Kim, Instability of amorphous oxide semiconductors via carrier-mediated structural transition between disorder and peroxide state, phys. status solidi (b) 249, 1277-1281 (2012). DOI: 10.1002/pssb.201147557, https://doi.org/10.1002/pssb.201147557
13. K. Iordanidou, C. Persson, Optoelectronic properties of coexisting InGaZnO4 structures, Mater. Sci. Semicond. Process. 121, 105297 (2021). DOI: 10.1016/j.mssp.2020.105297, https://doi.org/10.1016/j.mssp.2020.105297
14. H. Hosono, 68.3: Invited Paper: Transparent Amorphous Oxide Semiconductors for High Performance TFT, SID Symposium Digest of Technical Papers 38, 1830-1833 (2007). DOI: 10.1889/1.2785692, https://doi.org/10.1889/1.2785692
15. F. Liu, Z. Yang, Y. Luo, S. Guo, C. Zhang, S. Choo, X. Xu, X. Wang, K. A. Mkhoyan, M. Bernardi, B. Jalan, Understanding Polaronic Transport in Anatase TiO2 Films by Combining Precise Synthesis and First-Principles Many-Body Theory (2025). arXiv:2510.07097, https://arxiv.org/abs/2510.07097