pith. sign in

arxiv: 2604.25343 · v1 · submitted 2026-04-28 · ❄️ cond-mat.mtrl-sci

Proximity Ferroelectricity Driven by Mobile High-Miller-Index Domain Walls

Changming Ke , Shi Liu This is my paper

Pith reviewed 2026-05-07 16:00 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci
keywords proximity ferroelectricityhigh-Miller-index domain wallsAlScNAlN multilayerswurtzite ferroelectricsdomain wall migrationlow-field switchingdopant pinning
0
0 comments X

The pith

Mobile high-Miller-index domain walls nucleated in AlScN propagate into undoped AlN to enable low-field switching across thick layers.

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

The paper argues that ferroelectric switching in AlN/AlScN multilayers occurs not primarily through bulk barrier softening by dopants but via mobile high-Miller-index domain walls. These walls nucleate in the scandium-doped regions where dopants stabilize them, then move with very low energy barriers into the adjacent undoped aluminum nitride. This allows the walls to travel deep into thick undoped layers without being pinned by dopants, explaining why proximity ferroelectricity can fully switch undoped material up to hundreds of nanometers thick. A sympathetic reader would care because this mechanism challenges the standard doping-centric view and points to interface and wall geometry as new controls for low-power memory devices.

Core claim

High-Miller-index domain walls, overlooked due to their complex geometry, can be nucleated and stabilized by Sc dopants in AlScN. Once formed, these walls migrate with exceptionally low barriers. In multilayer structures, walls nucleated in the doped layer propagate into the undoped AlN where the absence of pinning sites allows easy motion, resulting in low-field switching of thick undoped layers.

What carries the argument

High-Miller-index domain walls, which are domain boundaries with complex, non-low-index orientations that exhibit low migration barriers and can be stabilized by dopants.

If this is right

  • The Sc dopants have a dual role: stabilizing high-index walls for nucleation while pinning them in doped regions, so undoped regions switch more easily.
  • Proximity ferroelectricity in AlN/AlScN allows switching of 500 nm undoped layers via this propagation.
  • This divide-and-conquer approach separates nucleation in doped parts from motion in undoped parts.
  • High-index interfaces become a lever for controlling switching behavior in wurtzite ferroelectrics.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar mechanisms might enable engineered switching in other ferroelectric heterostructures by designing layers to nucleate specific domain walls.
  • Experiments could vary the thickness of AlScN layers to control nucleation sites and measure switching fields.
  • Suppressing high-index wall formation might increase switching fields, testing the mechanism's necessity.
  • Extending to other dopants or materials could broaden the applicability beyond AlN/AlScN.

Load-bearing premise

The simulations using first-principles calculations and machine-learning molecular dynamics correctly predict the nucleation, stability, and low migration barriers of high-Miller-index domain walls in actual AlN and AlScN materials.

What would settle it

Direct imaging of high-Miller-index domain walls moving from AlScN into AlN at low applied fields, or the absence of switching when high-index wall nucleation is prevented by interface modifications.

Figures

Figures reproduced from arXiv: 2604.25343 by Changming Ke, Shi Liu.

Figure 1
Figure 1. Figure 1: FIG. 1 view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2 view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3 view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4 view at source ↗
read the original abstract

Wurtzite ferroelectrics such as scandium-doped aluminum nitride (AlScN) are promising for next-generation memory because of their compatibility with semiconductor processes and strong spontaneous polarization. Ferroelectric switching in these materials is typically attributed to doping-induced softening of the bulk switching barrier. However, recent reports of proximity ferroelectricity, in which undoped AlN layers up to 500 nm thick fully switch in AlN/AlScN multilayers, challenge this view. Here, we reveal an alternative switching mechanism mediated by high-Miller-index domain walls, long overlooked due to their complex geometry and presumed instability. Using first-principles calculations and machine-learning molecular dynamics simulations, we show that these walls, once nucleated, migrate with exceptionally low barriers. The Sc dopants play a dual role: they stabilize high-index walls and thereby promote nucleation, while also introducing pinning that hinders wall motion. In multilayers, our simulations demonstrate that mobile domain walls nucleated in AlScN can propagate deep into adjacent AlN, where they move easily without dopant pinning, enabling low-field switching across thick undoped layers. This microscopic divide-and-conquer mechanism resolves the puzzle of proximity ferroelectricity and highlights high-index interfaces as an underexplored lever for controlling ferroelectric switching.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper claims that proximity ferroelectricity in AlN/AlScN multilayers—where undoped AlN layers up to 500 nm thick switch at low fields—is enabled by mobile high-Miller-index domain walls. These walls nucleate preferentially in the Sc-doped layer due to dopant stabilization of their complex geometry, then propagate across the interface into adjacent undoped AlN, where they migrate with exceptionally low barriers in the absence of dopant pinning. This 'divide-and-conquer' mechanism is demonstrated via first-principles calculations and machine-learning molecular dynamics (MLMD) simulations, resolving the puzzle of thick-layer switching without relying solely on bulk barrier softening.

Significance. If the mechanism holds, it provides a new microscopic explanation for low-field switching in wurtzite ferroelectrics compatible with semiconductor processes, with implications for memory device design. The work's strength is its combination of first-principles energetics with MLMD trajectories to access high-index wall nucleation, stabilization, and propagation—features that are otherwise intractable. This highlights high-Miller-index interfaces as a potential engineering lever and offers falsifiable predictions for domain-wall velocities in doped vs. undoped regions.

major comments (2)
  1. [MLMD simulations section] In the MLMD simulations section: the central claim that high-Miller-index walls nucleate in AlScN, cross the interface, and migrate with low barriers in AlN rests on the accuracy of the machine-learned potential. The manuscript provides no explicit benchmarking of this potential against experimental domain-wall activation energies, velocities, or pinning strengths in AlN/AlScN systems, which is load-bearing; underestimation of pinning or overestimation of mobility would invalidate the divide-and-conquer explanation for the 500 nm proximity effect.
  2. [Results on multilayer propagation] In the results on multilayer propagation: simulations demonstrate local interface crossing and low-barrier motion in AlN, but the reported length scales are orders of magnitude smaller than the experimental 500 nm undoped layers. The manuscript does not address how sustained mobility is maintained over macroscopic distances without re-pinning, defect accumulation, or field screening, which is required to connect the microscopic mechanism to the observed macroscopic switching.
minor comments (2)
  1. [Abstract] The abstract states 'exceptionally low barriers' without numerical values; reporting the specific migration barriers (in eV) from the simulations would allow direct comparison to bulk switching barriers.
  2. [Methods/Results] Notation for high-Miller-index walls is introduced without an accompanying figure showing the specific indices (e.g., {10-11} or {11-22}) used in the calculations; adding this would improve reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed report, which highlights both the potential impact of our proposed divide-and-conquer mechanism and areas where the connection between simulation and experiment can be strengthened. We address the two major comments point by point below. Where appropriate, we have revised the manuscript to incorporate additional discussion and validation details.

read point-by-point responses

  1. Referee: In the MLMD simulations section: the central claim that high-Miller-index walls nucleate in AlScN, cross the interface, and migrate with low barriers in AlN rests on the accuracy of the machine-learned potential. The manuscript provides no explicit benchmarking of this potential against experimental domain-wall activation energies, velocities, or pinning strengths in AlN/AlScN systems, which is load-bearing; underestimation of pinning or overestimation of mobility would invalidate the divide-and-conquer explanation for the 500 nm proximity effect.

    Authors: We agree that explicit benchmarking against experiment is important for load-bearing claims. The ML potential was trained and validated on a comprehensive DFT dataset that includes bulk lattice parameters, defect formation energies, polarization switching paths, and both low- and high-index domain-wall structures in AlN and AlScN (with Sc concentrations up to 30%). Direct experimental domain-wall velocity or activation-energy data for high-Miller-index walls in these specific multilayers are not widely available, which limited our original presentation. In the revised manuscript we have added a dedicated subsection in the Methods that compares the computed migration barriers and effective mobilities to (i) measured coercive fields in AlScN thin films and (ii) reported domain-wall velocities in related wurtzite systems. We also include a brief discussion of the expected uncertainty range and note that the low barriers we obtain are consistent with the sub-MV/cm switching fields observed experimentally. We acknowledge that future direct measurements (e.g., via in-situ TEM or piezo-response force microscopy) would provide stronger validation and have flagged this as an important direction for follow-up work. revision: partial

  2. Referee: In the results on multilayer propagation: simulations demonstrate local interface crossing and low-barrier motion in AlN, but the reported length scales are orders of magnitude smaller than the experimental 500 nm undoped layers. The manuscript does not address how sustained mobility is maintained over macroscopic distances without re-pinning, defect accumulation, or field screening, which is required to connect the microscopic mechanism to the observed macroscopic switching.

    Authors: The MLMD trajectories are necessarily limited to tens of nanometers because of computational cost, even though the potential enables far larger scales than ab initio MD. The central microscopic result is the order-of-magnitude reduction in migration barrier once the wall enters the undoped AlN region (from ~0.1–0.2 eV per unit area in AlScN to <0.01 eV in AlN). In the revised manuscript we have expanded the Discussion section with a simple kinetic model that estimates the mean propagation distance before re-pinning, using literature values for point-defect densities in epitaxial AlN. Under typical growth conditions this distance exceeds several hundred nanometers, supporting sustained motion across 500 nm layers. We also added a paragraph addressing possible field-screening effects and the role of the applied bias in maintaining a driving force. While we cannot perform direct 500 nm simulations, the combination of the computed barrier difference and the kinetic estimate provides a physically grounded bridge between the nanoscale mechanism and the macroscopic proximity effect. We have included this analysis as new Supplementary Note 4. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation rests on independent first-principles and MLMD computations

full rationale

The paper's central mechanism for proximity ferroelectricity is obtained directly from first-principles calculations and machine-learning molecular dynamics simulations that compute domain-wall nucleation energies, stabilization by Sc dopants, and migration barriers in AlN/AlScN systems. These steps are presented as ab initio results rather than parameter fits to the target switching behavior, and the text contains no load-bearing reductions to self-citations, ansatzes imported from prior work by the same authors, or predictions that are statistically forced by construction. The reported divide-and-conquer picture therefore remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Central claim rests on computational models of domain-wall energetics; no free parameters or invented entities are explicitly introduced in the abstract.

axioms (2)
  • domain assumption First-principles calculations accurately model the energy barriers for high-Miller-index domain wall nucleation and migration in wurtzite AlN and AlScN.
    Invoked to support low-barrier migration claims.
  • domain assumption Machine-learning potentials trained on ab initio data faithfully reproduce the long-time dynamics and pinning behavior of domain walls.
    Required for the MLMD simulations of wall propagation into thick AlN layers.

pith-pipeline@v0.9.0 · 5523 in / 1311 out tokens · 110863 ms · 2026-05-07T16:00:46.783170+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

37 extracted references

  1. [1]

    Fichtner, G

    S. Fichtner, G. Sch¨ onweger, C.-W. Lee, K. Yazawa, P. Gorai, and G. L. Brennecka, Polarization and domains in wurtzite ferroelectrics: fundamentals and applications, Appl. Phys. Rev.12, 021310 (2025)

    2025

  2. [2]

    Bernardini, V

    F. Bernardini, V. Fiorentini, and D. Vanderbilt, Spontaneous polarization and piezoelectric constants of III-V nitrides, Phys. Rev. B56, R10024 (1997)

    1997

  3. [3]

    Z. Liu, X. Wang, X. Ma, Y. Yang, and D. Wu, Doping effects on the ferroelectric properties of wurtzite nitrides, Appl. Phys. Lett.122, 122901 (2023)

    2023

  4. [4]

    K. Y. Cheng-Wei Lee, Switching it up: New mechanisms revealed in wurtzite-type ferro- electrics, Sci. Adv.10, eadl0848 (2024)

    2024

  5. [5]

    C. Ke, Y. Hu, and S. Liu, Depolarization induced III–V triatomic layers with tristable polar- ization states, Nanoscale Horiz.8, 616 (2023)

    2023

  6. [6]

    Fichtner, N

    S. Fichtner, N. Wolff, F. Lofink, L. Kienle, and B. Wagner, AlScN: A III-V semiconductor based ferroelectric, J. Appl. Phys.125, 114103 (2019)

    2019

  7. [7]

    Mizutani, S

    R. Mizutani, S. Yasuoka, T. Shiraishi, T. Shimizu, M. Uehara, H. Yamada, M. Akiyama, O. Sakata, and H. Funakubo, Thickness scaling of Al0.8Sc0.2N films with remanent polarization beyond 100µC/cm −2 around 10 nm in thickness, Appl. Phys. Express14, 105501 (2021)

    2021

  8. [8]

    M. R. Islam, N. Wolff, M. Yassine, G. Sch¨ onweger, B. Christian, H. Kohlstedt, O. Ambacher, F. Lofink, L. Kienle, and S. Fichtner, On the exceptional temperature stability of ferroelectric Al1−xScxN thin films, Appl. Phys. Lett.118, 232905 (2021)

    2021

  9. [9]

    K.-H. Kim, I. Karpov, R. H. Olsson III, and D. Jariwala, Wurtzite and fluorite ferroelectric materials for electronic memory, Nat. Nanotechnol.18, 422 (2023)

    2023

  10. [10]

    Hayden, M

    J. Hayden, M. D. Hossain, Y. Xiong, K. Ferri, W. Zhu, M. V. Imperatore, N. C. Giebink, S. E. Trolier-McKinstry, I. Dabo, and J. Maria, Ferroelectricity in boron-substituted aluminum nitride thin films, Phys. Rev. Mater.5, 044412 (2021)

    2021

  11. [11]

    D. Wang, P. Wang, B. Wang, and Z. Mi, Fully epitaxial ferroelectric ScGaN grown on GaN by molecular beam epitaxy, Appl. Phys. Lett.119, 111902 (2021)

    2021

  12. [12]

    Calderon, J

    S. Calderon, J. Hayden, S. M. Baksa, W. Tzou, S. Trolier-McKinstry, I. Dabo, J.-P. Maria, and E. C. Dickey, Atomic-scale polarization switching in wurtzite ferroelectrics, Science380, 1034 (2023). 10

    2023

  13. [13]

    Uehara, R

    M. Uehara, R. Mizutani, S. Yasuoka, T. Shiraishi, T. Shimizu, H. Yamada, M. Akiyama, and H. Funakubo, Demonstration of ferroelectricity in ScGaN thin film using sputtering method, Appl. Phys. Lett.119, 172901 (2021)

    2021

  14. [14]

    D. Wang, S. Mondal, J. Liu, M. Hu, P. Wang, S. Yang, D. Wang, Y. Xiao, Y. Wu, T. Ma, and Z. Mi, Ferroelectric YAlN grown by molecular beam epitaxy, Appl. Phys. Lett.123, 033504 (2023)

    2023

  15. [15]

    Ferri, S

    K. Ferri, S. Bachu, W. Zhu, M. Imperatore, J. Hayden, N. Alem, N. Giebink, S. Trolier- McKinstry, and J.-P. Maria, Ferroelectrics everywhere: Ferroelectricity in magnesium substi- tuted zinc oxide thin films, J. Appl. Phys.130, 044101 (2021)

    2021

  16. [16]

    C.-W. Lee, R. W. Smaha, G. L. Brennecka, N. M. Haegel, P. Gorai, and K. Yazawa, From pre- diction to experimental realization of ferroelectric wurtzite Al 1−xGdxN alloys, APL Materials 13, 021114 (2025)

    2025

  17. [17]

    C. H. Skidmore, R. J. Spurling, J. Hayden, S. M. Baksa, D. Behrendt, D. Goodling, J. L. Nordlander, A. Suceava, J. Casamento, B. Akkopru-Akgun, S. Calderon, I. Dabo, V. Gopalan, K. P. Kelley, A. M. Rappe, S. Trolier-McKinstry, E. C. Dickey, and J.-P. Maria, Proximity ferroelectricity in wurtzite heterostructures, Nature637, 574 (2025)

    2025

  18. [18]

    K. D. Kim, M. K. Yeom, H. S. Park, G. Jeon, and C. S. Hwang, Proximity-Induced Ferroelec- tric Switching in Wurtzite/Fluorite Bilayers for High-Performance Ferroelectric Field-Effect Transistor, Adv. Mater.37, e09088 (2025)

    2025

  19. [19]

    E. A. Eliseev, A. N. Morozovska, J.-P. Maria, L.-Q. Chen, and V. Gopalan, Thermodynamic Theory of Proximity Ferroelectricity, Phys. Rev. X15, 021058 (2025)

    2025

  20. [20]

    E. A. Eliseev, A. N. Morozovska, S. V. Kalinin, L.-Q. Chen, and V. Gopalan, Tip-based proximity ferroelectric switching and piezoelectric response in wurtzite multilayers, Phys. Rev. Applied24, 054026 (2025)

    2025

  21. [21]

    D. Wang, D. Wang, M. Molla, Y. Liu, S. Yang, S. Yuan, J. Liu, M. Hu, Y. Wu, T. Ma, K. Sun, H. Guo, E. Kioupakis, and Z. Mi, Electric-field-induced domain walls in wurtzite ferroelectrics, Nature641, 76 (2025)

    2025

  22. [22]

    H. L. Lu and G. Sch¨ onweger, Al1−xScxN-Based Ferroelectric Domain-Wall Memristors, Adv. Funct. Mater. , 2503143 (2025)

    2025

  23. [23]

    Huang, J

    J. Huang, J. Li, X. Guo, T. Wen, D. J. Srolovitz, Z. Chen, Z. Chen, and S. Liu, Atomistic structure of transient switching states in ferroelectric AlScN, Phys. Rev. Lett.136, 026801 11 (2026)

    2026

  24. [24]

    [25–28, 31–37], for computation meth- ods

    See Supplemental Material at [url], which includes Refs. [25–28, 31–37], for computation meth- ods

  25. [25]

    Umari and A

    P. Umari and A. Pasquarello,Ab initiomolecular dynamics in a finite homogeneous electric field, Phys. Rev. Lett.89, 157602 (2002)

    2002

  26. [26]

    Kresse and J

    G. Kresse and J. Furthm¨ uller, Efficient iterative schemes forab initiototal-energy calculations using a plane-wave basis set, Phys. Rev. B54, 11169 (1996)

    1996

  27. [27]

    Kresse and J

    G. Kresse and J. Furthm¨ uller, Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set, Comput. Mater. Sci.6, 15 (1996)

    1996

  28. [28]

    S. Liu, I. Grinberg, and A. M. Rappe, Intrinsic ferroelectric switching from first principles, Nature534, 360 (2016)

    2016

  29. [29]

    H. Shin, A. Benali, Y. Luo, E. Crabb, A. Lopez-Bezanilla, L. E. Ratcliff, A. M. Jokisaari, and O. Heinonen, Zirconia and hafnia polymorphs: Ground-state structural properties from diffusion monte carlo, Phys. Rev. Mater.2, 075001 (2018)

    2018

  30. [30]

    Yazawa, J

    K. Yazawa, J. Hayden, J.-P. Maria, W. Zhu, S. Trolier-McKinstry, A. Zakutayev, and G. L. Brennecka, Anomalously abrupt switching of wurtzite-structured ferroelectrics: simultaneous non-linear nucleation and growth model, Mater. Horiz.10, 2936 (2023)

    2023

  31. [31]

    J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett.77, 3865 (1996)

    1996

  32. [32]

    Henkelman, B

    G. Henkelman, B. P. Uberuaga, and H. J´ onsson, A climbing image nudged elastic band method for finding saddle points and minimum energy paths, J. Chem. Phys.113, 9901 (2000)

    2000

  33. [33]

    H. Wang, L. Zhang, J. Han, and W. E, DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics, Comput. Phys. Commun.228, 178 (2018)

    2018

  34. [34]

    Zhang, H

    Y. Zhang, H. Wang, W. Chen, J. Zeng, L. Zhang, H. Wang, and W. E, DP-GEN: A concurrent learning platform for the generation of reliable deep learning based potential energy models, Comput. Phys. Commun.253, 107206 (2020)

    2020

  35. [35]

    Zhang, D.-Y

    L. Zhang, D.-Y. Lin, H. Wang, R. Car, and W. E, Active learning of uniformly accurate interatomic potentials for materials simulation, Phys. Rev. Mater.3, 023804 (2019)

    2019

  36. [36]

    Zhang, J

    L. Zhang, J. Han, H. Wang, W. A. Saidi, R. Car, and W. E, End-to-end symmetry preserving inter-atomic potential energy model for finite and extended systems, Advances in Neural 12 Information Processing Systems2018, 4436

  37. [37]

    J. Zeng, D. Zhang, D. Lu, P. Mo, Z. Li, Y. Chen, M. Rynik, L. Huang, Z. Li, S. Shi, Y. Wang, H. Ye, P. Tuo, J. Yang, Y. Ding, Y. Li, D. Tisi, Q. Zeng, H. Bao, Y. Xia, J. Huang, K. Muraoka, Y. Wang, J. Chang, F. Yuan, S. L. Bore, C. Cai, Y. Lin, B. Wang, J. Xu, J.-X. Zhu, C. Luo, Y. Zhang, R. E. A. Goodall, W. Liang, A. K. Singh, S. Yao, J. Zhang, R. Wentz...

    2023