Pith. sign in

REVIEW 2 major objections 8 minor 44 references

Large-scale first-principle simulations of amorphous indium oxide

T0 review · 2 major / 8 minor · reviewed 2026-07-10 · glm-5.2

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Peroxide bonds in amorphous indium oxide act as built-in dopants

desk verdict Solid large-scale MLIP study of a-In2O3 with one load-bearing gap: O-O bond energetics are not independently validated at the scale where they matter most. read the letter →

arxiv 2607.08617 v1 pith:NGS763KN submitted 2026-07-09 cond-mat.mtrl-sci cond-mat.dis-nn

Matthew Bousquet , Francois Gygi , Giulia Galli This is my paper · ORCID
classification cond-mat.mtrl-scicond-mat.dis-nn
keywords amorphousindiumoxideperoxidebondsmachine-learnedinteratomicpotentialintrinsicn-typedopingelectronmobilityRamanspectroscopyphotoanodewateroxidation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Amorphous indium oxide is a widely used semiconductor that retains high electron mobility even when its crystalline order is lost — a behavior that is unusual and poorly understood. This paper uses a machine-learned interatomic potential, trained on first-principles molecular dynamics data, to simulate amorphous structures of up to 5120 atoms, far larger than previous studies. At this scale, the authors identify two structural features they argue are central to the material's properties. First, they find extended chains of edge-sharing indium–oxygen polyhedra that span the simulation cell, providing a structural pathway for the high electron mobility that has been observed experimentally but never connected to a specific atomic-scale motif. Second, they find that oxygen–oxygen peroxide-like bonds, about 1.5 Å long, form readily in the amorphous network. These bonds introduce filled electronic states just below the conduction band edge, functioning as an intrinsic source of n-type doping — meaning the material can generate charge carriers without any external impurity. The bonds also enhance absorption of below-gap light, which matters for using the material as a photoanode to split water. The authors predict that these bonds produce a sharp Raman signal near 850 cm⁻¹ that is absent from infrared spectra, offering a way to detect and quantify them experimentally. They also find that the amorphous phase is only about 0.7 eV per formula unit above the crystal in energy, consistent with the experimental observation that indium oxide is a poor glass former prone to quasi-crystalline regions.

What carries the argument

The argument works through a chain of scale and observation: (1) a MACE machine-learned potential, trained on first-principles molecular dynamics trajectories, enables simulations at 5120 atoms — large enough to see percolating chains of polyhedra that 80-atom cells cannot resolve; (2) at these scales, O–O peroxide bonds appear in the majority of independently generated samples, with a mean length of 1.5 Å; (3) electronic structure calculations at both SCAN and hybrid DDH levels show these bonds introduce filled states just below the conduction band minimum; (4) inverse participation ratio analysis confirms the in-gap states are localized on the peroxide bond while the associated donor state

What would settle it

Measure Raman spectra of amorphous In₂O₃ films deposited under controlled conditions. If no peak appears near 850 cm⁻¹, or if the peak's intensity does not correlate with carrier density, the claim that peroxide bonds are the source of intrinsic n-type doping is undermined.

Watch

Extended reading notes

Core claim

The central claim is that amorphous indium oxide contains two previously unresolved structural motifs — extended chains of edge-sharing InO_k polyhedra and short O–O peroxide-like bonds — and that these motifs respectively explain the material's anomalously high electron mobility and its intrinsic n-type conductivity. The peroxide bonds introduce localized in-gap states near the conduction band minimum, acting as self-dopants, and produce a detectable Raman signature at 850 cm⁻¹ that is absent from IR spectra.

Load-bearing premise

The machine-learned potential, trained primarily on 80-atom first-principles trajectories, is assumed to accurately reproduce the energetics of O–O peroxide bonds when extrapolated to cells 64–640 times larger. The energy difference between structures with and without these bonds is only about 0.015 eV per formula unit, so even small errors in the potential's description of the O–O interaction could change the conclusion about whether these bonds are genuinely prevalent or an

Editorial extensions

If this is right

  • If the Raman feature at 850 cm⁻¹ is confirmed experimentally, it would provide a direct, quantitative diagnostic for O–O bond concentration in deposited amorphous films, linking deposition conditions to carrier density.
  • The identification of percolating edge-sharing polyhedra as the mobility pathway suggests that deposition protocols favoring longer chains could further boost electron transport.
  • The self-doping mechanism via peroxide bonds offers a way to tune n-type carrier density without introducing external impurities, which could reduce scattering and improve mobility.
  • If amorphization reliably activates optically dark transitions present in the crystal, a-In₂O₃ becomes a more viable photoanode candidate for solar water splitting.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The energy difference between structures with and without O–O bonds is only ~0.015 eV per formula unit, meaning that whether these bonds form could be highly sensitive to deposition conditions, annealing history, and oxygen partial pressure — suggesting a route to controllable doping.
  • If peroxide bonds are a general feature of amorphous wide-bandgap oxides (as hinted by the authors' reference to other systems), the self-doping mechanism identified here may extend beyond indium oxide to materials like amorphous zinc oxide or gallium oxide.
  • The quasi-crystalline regions found in the amorphous samples raise the question of whether the high mobility actually requires a fully amorphous phase or whether it is sustained by these crystalline remnants — a distinction that matters for device engineering.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 8 minor

Summary. This manuscript presents large-scale simulations of amorphous In₂O₃ (a-In₂O₃) using a MACE-based machine-learned interatomic potential (MLIP) trained on first-principles molecular dynamics (FPMD) data. The authors generate amorphous structures of up to 5120 atoms—two orders of magnitude larger than typical ab initio studies—and report X-ray structure factors in quantitative agreement with experiment. The paper makes three principal claims: (i) extended chains of edge-sharing InOₖ polyhedra provide a structural basis for high electron mobility; (ii) O–O peroxide-like bonds (mean length ~1.5 Å) form in the amorphous network, introducing localized in-gap states near the conduction band minimum that act as intrinsic n-type self-doping; and (iii) these bonds produce a distinct Raman feature near 850 cm⁻¹ absent in the IR spectrum, offering an experimentally testable prediction. The combination of structural, vibrational, and electronic-structure analysis, together with the scale of the simulations, represents a substantial contribution to the understanding of a-In₂O₃.

Significance. The paper addresses a long-standing debate about the structure and transport properties of a-In₂O₃, a technologically important semiconductor. The key strengths include: (1) the achievement of simulation scales (5120 atoms) that reveal percolating polyhedral chains inaccessible to prior ab initio work; (2) quantitative agreement with experimental X-ray scattering data (Figure 3C), providing external validation of the structural model; (3) a falsifiable experimental prediction—the Raman-active O–O stretch near 850 cm⁻¹—that is directly testable; and (4) electronic-structure calculations at both SCAN and dielectric-dependent hybrid (DDH) levels connecting the O–O bonds to in-gap states and sub-gap optical absorption. The structure–property picture is coherent and the photoelectrochemical implications are well-motivated.

major comments (2)
  1. §III.B, Figure 5: The central claim that O–O peroxide-like bonds form in a-In₂O₃ rests on the MLIP correctly capturing O–O bond energetics in 640- and 5120-atom cells. The reported energy difference between structures with and without O–O bonds is only ~0.015 eV/f.u., which is comparable to the expected accuracy of the MLIP relative to DFT. The paper does not report any independent DFT validation (e.g., single-point DFT calculations) on MLIP-generated 640-atom structures with and without O–O bonds to confirm that DFT agrees with the MLIP on the sign and magnitude of this energy difference. Given that the entire electronic-structure argument (in-gap states, n-type self-doping, enhanced sub-gap absorption) depends on O–O bonds being thermodynamically accessible and not merely kinetically trapped, at least one direct DFT cross-check on a representative 640-atom structure should be performed
  2. §III.C, Figure 8: The identification of extended chains of edge-sharing InOₖ polyhedra is a headline result, but the statistics are based on a single 5120-atom sample. The reported average chain length (12.18 ± 2.41 In atoms) carries an uncertainty derived from one configuration. Since this result is presented as a structural basis for high electron mobility, generating at least one additional independent 5120-atom sample (or several 640-atom samples) to confirm the persistence and typical length of these chains would substantially strengthen the claim. The authors should also clarify what fraction of In atoms belongs to percolating chains versus isolated clusters.
minor comments (8)
  1. The title reads 'first-principle simulations' but should read 'first-principles simulations.'
  2. Several cross-references in the manuscript are broken (e.g., 'Figure ??' appears multiple times in §III.B and §III.C, and 'section ??' in §III.C). These should be fixed before publication.
  3. §II.A: The DDH exact exchange fraction α is stated to be determined from ε∞ = 4.2 evaluated at the hybrid SE-RSH level for crystalline In₂O₃. It would be helpful to state the resulting value of α explicitly for clarity.
  4. §III.A: The comparison of total g(r) with previous works (Figure 2) notes quantitative discrepancies in In–O bond distances attributed to the use of SCAN versus PBE. A brief comment on whether these differences are expected from known SCAN versus PBE trends in oxide bond lengths would strengthen this argument.
  5. §IV: The 80-atom DDH calculations (Figure 10) yield a band gap of 2.27 eV, while the 640-atom SCAN calculations (Figure 9) are used for the EDOS analysis. The authors note finite-size effects on the conduction density of states. A brief discussion of whether the gap reduction and in-gap state positions are expected to change at the hybrid level for the 640-atom cells would aid comparison.
  6. Figure 6B: The Raman spectrum is computed at the LDA level for an 80-atom cell. The authors note that VDOS calculations with SCAN and LDA yield consistent results, but the Raman intensities depend on the Raman tensor. A brief justification for using LDA for Raman intensities, or a note on the expected level of agreement, would be helpful.
  7. §III.B: The defect formation energy of 0.04 eV/f.u. for the peroxide interstitial in the recovered crystalline structure is reported as an average over 10 snapshots. The spread of these values and the structural definition of the defect should be briefly described for direct comparison.
  8. References [10] and [38] are cited with 2026 dates. If these are accepted/in-press manuscripts, the DOI and journal information should be complete and not placeholder dates.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: central claims are independently benchmarked against experiment and prior DFT; one self-citation is present but not load-bearing.

full rationale

The paper's central claims about O–O peroxide bonds, in-gap states, and the 850 cm⁻¹ Raman feature are derived from a chain that is not circular. The MLIP is trained on FPMD trajectories (not fitted to the target results), and its structural predictions are validated against experimental X-ray structure factors (Ref. 38) and prior computational work (Refs. 9, 10, 35, 36). The electronic-structure consequences of O–O bonds are computed independently at both SCAN and DDH levels on representative snapshots. The Raman prediction is externally falsifiable. The one self-citation (Ref. 16, prior work by the same authors on crystalline In₂O₃ photoelectrodes) is used only to establish that the band positions of crystalline In₂O₃ are favorable for water oxidation and that the VBM→CBM transition is optically dark—background motivation, not a load-bearing premise for the amorphous-phase results. The skeptic's concern about MLIP extrapolation accuracy for O–O energetics is a correctness/validation risk, not a circularity issue: the paper does not define its inputs in terms of its outputs, nor does it rename a fitted parameter as a prediction. No step in the derivation chain reduces to its own inputs by construction.

Assumptions & free parameters 3 free parameters · 3 assumptions · 1 invented entities

The paper introduces no fundamentally new physical entities; the O–O bonds are a structural motif whose existence is the central question being investigated, not a postulate.

free parameters (3)
  • DDH exact exchange fraction α = 0.238 (1/ε∞, ε∞=4.2)
    Determined from the inverse of the macroscopic electronic dielectric constant computed at the SE-RSH level for crystalline In₂O₃ (§II.A). Consistent with experimental values 4.0–4.1.
  • MACE model hyperparameters = Not specified in main text
    Training procedure and hyperparameters are referenced to the SI, which is not available. The foundation model is fine-tuned, but the specific architecture parameters are unstated in the main text.
  • Quench rates = 100–450 K/ps
    Chosen to generate amorphous structures; the range and specific values are computational choices not derived from physical constraints.
assumptions (3)
  • domain assumption SCAN meta-GGA accurately describes the In₂O₃ potential energy surface including O–O bond formation
    All FPMD training data and electronic structure calculations use SCAN (§II.A). The accuracy of SCAN for O–O peroxide bonds in this system is assumed but not independently validated against higher-level methods (e.g., CCSD(T) on cluster models).
  • domain assumption The MACE MLIP trained on 80-atom FPMD trajectories extrapolates reliably to 5120-atom cells
    The central structural and O–O bond results depend on this extrapolation. The paper validates short-range RDFs between 80 and 5120 atom cells (Figure 3A) but does not independently validate the potential's accuracy for rare motifs at large scale.
  • domain assumption LDA adequately describes Raman intensities for O–O stretching modes
    Raman spectra were computed at LDA level 'for simplicity' (§II.D, §III.B). The paper notes VDOS is consistent between SCAN and LDA but does not validate Raman tensor accuracy at LDA vs. SCAN.
invented entities (1)
  • O–O peroxide-like bonds in a-In₂O₃ independent evidence
    purpose: Explains in-gap states, n-type self-doping, and sub-gap optical absorption
    The paper provides a falsifiable Raman prediction at 850 cm⁻¹ and structural evidence (bond length ~1.5 Å) consistent with known peroxide chemistry. The existence is debated in prior literature but supported by multiple independent samples here.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Large-scale first-principle simulations of amorphous indium oxide." pith.science (2026). https://pith.science/paper/NGS763KN

@misc{pith2026260708617,
  author       = {Pith},
  title        = {Pith review of: Large-scale first-principle simulations of amorphous indium oxide},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NGS763KN}},
  note         = {Machine review of arXiv:2607.08617}
}
abstract

Amorphous indium oxide (a-In$_2$O$_3$) is a high-electron-mobility semiconductor of central importance in thin-film transistors and a promising photoanode for solar-driven water oxidation. Despite sustained experimental and computational investigations, the structural motifs underlying its unusual transport properties and the existence of O-O peroxide-like bonds within its network have remained unresolved. Here we develop a MACE-based machine-learned interatomic potential trained on first-principles molecular dynamics trajectories and use it to generate and analyze amorphous structures containing up to 5120 atoms, two orders of magnitude larger than those adopted in typical ab initio studies. We find X-ray structure factors in excellent quantitative agreement with experiment and we confirm that In$_2$O$_3$ is a poor glass former, with the likely presence of quasi-crystalline regions in amorphous samples. Our large-scale structural analysis reveals extended chains of edge-sharing InO$_k$ polyhedra providing a concrete structural basis for the high electron mobility of a-In$_2$O$_3$. Our results strongly support the formation of O-O peroxide-like bonds in the amorphous network, with a mean length of 1.5 \AA{}. We show that these bonds introduce localized in-gap states near the conduction band minimum, acting as a source of intrinsic n-type self-doping and enhancing sub-gap optical absorption. These effects are detectable via a distinct Raman feature near 850 cm$^{-1}$ that is absent in the IR spectrum. Overall, our results establish a comprehensive structure-property picture of a-In$_2$O$_3$, provide directly testable experimental predictions, and suggest that controlled amorphization is a viable strategy for improving the photoelectrochemical activity of a-In$_2$O$_3$.

Figures

Figures reproduced from arXiv: 2607.08617 by the authors.

Figure 1
Figure 1. FIG. 1. Partial radial distribution functions, g [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Comparison of the computed total correlation func [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Histogram of O-O bond lengths for samples generated [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Panel A shows the vibrational density of states [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Distribution of MLIP energies for 640-atom systems, [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 7
Figure 7. Figure 7: FIG. 7. Computed angular distribution functions (upper [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Longest chain of InO edge sharing polyhedra detected [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9. Electronic Density of States (EDOS) for a-In [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10. Electronic Density of States (EDOS) computed for [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 12
Figure 12. Figure 12: FIG. 12. Dipole transition strengths from occupied orbitals [PITH_FULL_IMAGE:figures/full_fig_p011_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 44 canonical work pages

  1. [1]

    However, the valence band maximum (VBM) to conduction band minimum (CBM) transition is optically dark, and strategies to lower the optical gap are required

    showed that the bands of crystalline In 2O3 are well positioned to catalyze the two-electron water oxidation reaction (WOR). However, the valence band maximum (VBM) to conduction band minimum (CBM) transition is optically dark, and strategies to lower the optical gap are required. Amorphization of the material is a possi- ble strategy to decrease the onse...

    work page 2017

  2. [2]

    7 1.3 1.4 1.5 1.6 1.7 O–O distance (Å) Prob

    Results from Ref.[10] are also shown. 7 1.3 1.4 1.5 1.6 1.7 O–O distance (Å) Prob. Density 640 atoms 80 atoms 5120 atoms FIG. 4. Histogram of O-O bond lengths for samples generated by our MLIP atρ= 7.11 g·cm −3. The mean distance is 1.49±0.04 ˚A for 80 and 640 atom cells and 1.50±0.05 ˚A for the 5120 cell. -4132.60 -4132.54 -4132.48 MACE Energy / f.u. (eV...

    work page 2022

  3. [3]

    Charnas, Z

    A. Charnas, Z. Zhang, Z. Lin, D. Zheng, J. Zhang, M. Si, and P. D. Ye, Review—extremely thin amor- phous indium oxide transistors, Advanced Materials36, 10.1002/adma.202304044 (2023)

    work page doi:10.1002/adma.202304044 2023

  4. [4]

    B. K. Yap, Z. Zhang, G. S. H. Thien, K.-Y. Chan, and C. Y. Tan, Recent advances of in2o3-based thin-film tran- sistors: A review, Applied Surface Science Advances16, 100423 (2023)

    work page 2023

  5. [5]

    M. Si, Z. Lin, Z. Chen, X. Sun, H. Wang, and P. D. Ye, Scaled indium oxide transistors fabricated using atomic layer deposition, Nature Electronics5, 164–170 (2022)

    work page 2022

  6. [6]

    S. K. Kim, Atomically thin indium oxide transistors, Na- ture Electronics5, 129–130 (2022)

    work page 2022

  7. [7]

    Yuvaraja, H

    S. Yuvaraja, H. Faber, M. Kumar, N. Xiao, G. I. Ma- ciel Garc´ ıa, X. Tang, T. D. Anthopoulos, and X. Li, Three-dimensional integrated metal-oxide transistors, Nature Electronics7, 768–776 (2024)

    work page 2024

  8. [8]

    Seong, T

    S. Seong, T. Ha, S. Jung, W. Jeong, and Y. Chung, Re- cent progress and opportunities in oxide semiconductor devices for in-memory and neuromorphic computing, Ad- vanced Electronic Materials12, 10.1002/aelm.202500521 (2025)

    work page doi:10.1002/aelm.202500521 2025

Show all 44 references
  1. [9]

    Q. Hu, S. Zhu, Y. Zhu, C. Gu, S. Liu, R. Huang, and Y. Wu, Amorphous indium tin oxide transistors for power amplification above 10 ghz, Nature Electronics8, 803–809 (2025)

    2025

  2. [10]

    D. B. Buchholz, L. Zeng, M. J. Bedzyk, and R. P. Chang, Differences between amorphous indium oxide thin films, Progress in Natural Science: Materials International23, 475–480 (2013)

    2013

  3. [11]

    D. B. Buchholz, Q. Ma, D. Alducin, A. Ponce, M. Jose- Yacaman, R. Khanal, J. E. Medvedeva, and R. P. H. Chang, The structure and properties of amorphous indium oxide, Chemistry of Materials26, 5401–5411 (2014)

    2014

  4. [12]

    Jankousky, D

    M. Jankousky, D. Pashov, J. H. Mazo, R. E. Larsen, V. Dobrosavljevi´ c, M. van Schilfgaarde, and V. Ste- vanovi´ c, Effective bands and band-like electron transport in amorphous solids, Nature Physics22, 88–93 (2026)

    2026

  5. [13]

    Aliano, A

    A. Aliano, A. Catellani, and G. Cicero, Characteriza- tion of amorphous in2o3: An ab initio molecular dynam- ics study, Applied Physics Letters99, 10.1063/1.3664224 (2011)

    2011 doi

  6. [14]

    J. E. Medvedeva, I. A. Zhuravlev, C. Burris, D. B. Buch- holz, M. Grayson, and R. P. H. Chang, Origin of high carrier concentration in amorphous wide-bandgap ox- ides: Role of disorder in defect formation and electron localization in in2o3-x, Journal of Applied Physics127, 10....

    2020 doi

  7. [15]

    Arhammar, A

    C. Arhammar, A. Pietzsch, N. Bock, E. Holmstr¨ om, C. M. Araujo, J. Grasj¨ o, S. Zhao, S. Green, T. Peery, F. Hennies, S. Amerioun, A. F¨ ohlisch, J. Schlappa, T. Schmitt, V. N. Strocov, G. A. Niklasson, D. C. Wal- lace, J.-E. Rubensson, B. Johansson, and R. Ahuja, Un- veiling...

    2011

  8. [17]

    J. H. Skone, M. Govoni, and G. Galli, Self-consistent hy- brid functional for condensed systems, Physical Review B89, 10.1103/physrevb.89.195112 (2014)

    2014 doi

  9. [18]

    Bousquet, J

    M. Bousquet, J. Zhan, C. Luo, A. B. Martinson, F. Gygi, and G. Galli, Computational study of indium oxide pho- toelectrodes, The Journal of Physical Chemistry C129, 8395–8403 (2025)

    2025

  10. [19]

    Gygi, Architecture of qbox: A scalable first-principles molecular dynamics code, IBM Journal of Research and Development52, 137–144 (2008)

    F. Gygi, Architecture of qbox: A scalable first-principles molecular dynamics code, IBM Journal of Research and Development52, 137–144 (2008)

    2008

  11. [20]

    Schlipf and F

    M. Schlipf and F. Gygi, Optimization algorithm for the generation of oncv pseudopotentials, Computer Physics Communications196, 36–44 (2015)

    2015

  12. [21]

    Bussi, D

    G. Bussi, D. Donadio, and M. Parrinello, Canonical sam- pling through velocity rescaling, The Journal of Chemical Physics126, 10.1063/1.2408420 (2007)

    2007 doi

  13. [22]

    J. Zhan, M. Govoni, and G. Galli, Dielectric-dependent range-separated hybrid functional calculations for metal oxides, Physical Review Materials9, 10.1103/physrev- materials.9.053808 (2025)

    2025 doi

  14. [23]

    Feneberg, J

    M. Feneberg, J. Nixdorf, C. Lidig, R. Goldhahn, Z. Galazka, O. Bierwagen, and J. S. Speck, Many-electron effects on the dielectric function of cubic in2o3: Effective electron mass, band nonparabolicity, band gap renormal- ization, and burstein-moss shift, Physical Review B93, ...

    2016 doi

  15. [24]

    Sundaresh, S

    S. Sundaresh, S. D. Nehate, and K. B. Sundaram, Elec- trical and optical studies of reactively sputtered indium oxide thin films, ECS Journal of Solid State Science and Technology10, 065016 (2021)

    2021

  16. [25]

    Batatia, D

    I. Batatia, D. P. Kovacs, G. Simm, C. Ortner, and G. Cs´ anyi, Mace: Higher order equivariant message pass- ing neural networks for fast and accurate force fields, Advances in Neural Information Processing Systems35, 11423 (2022)

    2022

  17. [26]

    Batatia, C

    I. Batatia, C. Lin, J. Hart, E. Kasoar, A. M. Elena, S. W. Norwood, T. Wolf, and G. Cs´ anyi, Cross learning be- tween electronic structure theories for unifying molecu- lar, surface, and inorganic crystal foundation force fields, arXiv preprint arXiv:2510.25380 (2025)

    2025

  18. [27]

    Giannozzi, S

    P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococ- cioni, I. Dabo, A. Dal Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F. ...

    2009

  19. [28]

    Giannozzi, O

    P. Giannozzi, O. Baseggio, P. Bonf` a, D. Brunato, R. Car, I. Carnimeo, C. Cavazzoni, S. de Gironcoli, P. Delugas, F. Ferrari Ruffino, A. Ferretti, N. Marzari, I. Timrov, A. Urru, and S. Baroni, Q¡scp¿uantum¡/scp¿ espresso to- ward the exascale, The Journal of Chemical Physics...

    2020 doi

  20. [29]

    J. P. Perdew and A. Zunger, Self-interaction correction to density-functional approximations for many-electron systems, Physical Review B23, 5048–5079 (1981). 13

    1981

  21. [30]

    D. M. Ceperley and B. J. Alder, Ground state of the elec- tron gas by a stochastic method, Physical Review Letters 45, 566–569 (1980)

    1980

  22. [31]

    Kohn and L

    W. Kohn and L. J. Sham, Self-consistent equations in- cluding exchange and correlation effects, Physical Review 140, A1133–A1138 (1965)

    1965

  23. [32]

    J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Physical Review Letters77, 3865 (1996)

    1996

  24. [33]

    D. T. Cromer and J. B. Mann, X-ray scattering factors computed from numerical Hartree–Fock wave functions, Acta Crystallographica Section A24, 321 (1968)

    1968

  25. [34]

    P. J. Brown, A. G. Fox, E. N. Maslen, M. A. O’Keefe, and B. T. M. Willis, Intensity of diffracted intensities, inInternational Tables for Crystallography(International Union of Crystallography, 2006) p. 554–595

    2006

  26. [35]

    T. E. Faber and J. M. Ziman, A theory of the electrical properties of liquid metals. III. the resistivity of binary alloys, Philosophical Magazine11, 153 (1965)

    1965

  27. [36]

    Lazzeri and F

    M. Lazzeri and F. Mauri, First-principles calculation of vibrational raman spectra in large systems: Signature of small rings in crystalline sio2, Physical Review Letters 90, 10.1103/physrevlett.90.036401 (2003)

    2003 doi

  28. [37]

    J. E. Medvedeva, D. B. Buchholz, and R. P. H. Chang, Recent advances in understanding the structure and properties of amorphous oxide semiconductors, Advanced Electronic Materials3, 10.1002/aelm.201700082 (2017)

    2017 doi

  29. [38]

    J. E. Medvedeva, E. Caputa-Hatley, and I. Zhuravlev, Metallic networks and hydrogen compensation in highly nonstoichiometric amorphous in2o3-x, Physical Review Materials6, 10.1103/physrevmaterials.6.025601 (2022)

    2022 doi

  30. [39]

    Khanal, D

    R. Khanal, D. B. Buchholz, R. P. H. Chang, and J. E. Medvedeva, Composition-dependent structural and transport properties of amorphous transparent con- ducting oxides, Physical Review B91, 10.1103/phys- revb.91.205203 (2015)

    2015 doi

  31. [40]

    Utsuno, H

    F. Utsuno, H. Inoue, I. Yasui, Y. Shimane, S. Tomai, S. Matsuzaki, K. Inoue, I. Hirosawa, M. Sato, and T. Honma, Structural study of amorphous in2o3 film by grazing incidence x-ray scattering (gixs) with syn- chrotron radiation, Thin Solid Films496, 95–98 (2006)

    2006

  32. [41]

    Ellsworth, A

    S. Ellsworth, A. Navrotsky, and R. C. Ewing, Energetics of radiation damage in natural zircon (zrsio4), Physics and Chemistry of Minerals21, 140–149 (1994)

    1994

  33. [42]

    Molodetsky, A

    I. Molodetsky, A. Navrotsky, M. J. Paskowitz, V. J. Lep- pert, and S. H. Risbud, Energetics of x-ray-amorphous zirconia and the role of surface energy in its formation, Journal of Non-Crystalline Solids262, 106–113 (2000)

    2000

  34. [43]

    Aykol and K

    M. Aykol and K. A. Persson, Oxidation protection with amorphous surface oxides: Thermodynamic insights from ab initio simulations on aluminum, ACS Applied Mate- rials and Interfaces10, 3039–3045 (2018)

    2018

  35. [44]

    Kranert, R

    C. Kranert, R. Schmidt-Grund, and M. Grund- mann, Raman active phonon modes of cubic in¡sub¿2¡/sub¿o¡sub¿3¡/sub¿, physica status solidi (RRL) – Rapid Research Letters8, 554–559 (2014)

    2014

  36. [45]

    Stokey, R

    M. Stokey, R. Korlacki, S. Knight, A. Ruder, M. Hil- fiker, Z. Galazka, K. Irmscher, Y. Zhang, H. Zhao, V. Darakchieva, and M. Schubert, Optical phonon modes, static and high-frequency dielectric constants, and effective electron mass parameter in cubic in2o3, Journal of Appli...

    2021 doi

Pith tools

Reviewed July 10, 2026 · model on record in the stance chip above.