A 2025 first-principles analysis (Cañas, Yassine, and Ambacher, Phys. Rev. B 112, L081201) defined a spontaneous polarization for zincblende III-V semiconductors of roughly 2.5-4.3 C/m^2 - about three times the wurtzite values - as the formal-polarization difference between the zincblende phase and the centrosymmetric rocksalt structure reached along the ⟨111⟩\langle 111\rangle switching path between opposite polarities. The computed switching barriers (about 0.5-0.6 eV/atom) imply coercive fields exceeding 40 MV/cm, far above the breakdown fields of these materials, so this polarization has never been observed. The open problem is to make it experimentally accessible: identify and validate zincblende-derived alloys or heterostructures in which polarization reversal along ⟨111⟩\langle 111\rangle is achievable below breakdown while an insulating gap is preserved along the whole path; determine whether group-III transition-metal alloying can lower both the layered and the rocksalt energy maxima that set the coercive field; and establish whether proximity switching by an adjacent ferroelectric wurtzite nitride can reverse a zincblende layer and under what interfacial conditions.

Background

The zincblende (ZB) structure, in which two interpenetrating face-centered-cubic sublattices of a group-III and a group-V element are offset by a quarter body diagonal, is the common crystal form of mainstream III-V semiconductors such as GaAs and AlAs, and of the metastable cubic polytype of GaN; the wurtzite (WZ) structure is its hexagonal counterpart built from the same tetrahedral bonding. Under the modern theory of polarization, the 'formal polarization' of a crystal is the sum of an ionic term (point charges at the nuclear positions) and an electronic term computed as a Berry phase (a quantum-geometric phase accumulated by the occupied Bloch wavefunctions over the Brillouin zone); because the value depends on the choice of primitive cell, the formal polarization of an infinite crystal is multivalued, defined only modulo a 'polarization quantum' Pq=eR/ΩP_{q}=eR/\Omega set by the lattice. The 'spontaneous polarization' PsP_{s} is not the formal polarization itself but the difference in formal polarization between a polar crystal and the non-polar centrosymmetric structure at the midpoint of its ferroelectric switching path; equivalently it is half the formal-polarization difference between the two opposite polarities, so it is defined through switching.

For wurtzite nitrides this definition became experimentally consequential when ferroelectricity was discovered in Sc-alloyed AlN (Fichtner et al. 2019): the measured switched polarization agrees with calculations referenced to the centrosymmetric layered-hexagonal structure (Dreyer et al. 2016; Yassine et al. 2024), not to the zincblende reference historically used in the nitride community (Bernardini et al. 2001). Cañas, Yassine and Ambacher (2025) extended this switching-based definition to zincblende III-Vs. They constructed the transformation that reverses ZB polarity along the ⟨111⟩\langle 111\rangle axis - N-polar ZB, an N-polar layered structure, the centrosymmetric rocksalt structure, a metal-polar layered structure, metal-polar ZB - and computed Berry-phase polarizations and total energies along it. Defining PsP_{s} relative to rocksalt, they obtained Ps=2.521P_{s}=2.521 C/m2^2 for GaAs, 2.5342.534 C/m2^2 for AlAs, 4.0244.024 C/m2^2 for GaN and 4.2934.293 C/m2^2 for AlN, about three times the wurtzite values, refuting the longstanding assumption that ZB spontaneous polarization is zero. In their branch convention the ZB formal polarization is −0.25Pq-0.25P_{q} (group-V polar) or 1.25Pq1.25P_{q} (group-III polar) and the rocksalt value is 0.5Pq0.5P_{q}, with Pq=e/AbP_{q}=e/A_{b} set by the basal-plane area; notably, no stoichiometric rocksalt cell can have a formal polarization smaller than Pq/2P_{q}/2 along ⟨111⟩\langle 111\rangle, and such large formal polarizations are not observed on (111)-oriented finite crystals because surface reconstructions, partial coverage and extrinsic passivation compensate them.

Switching a zincblende III-V has, however, never been achieved. The computed energy barriers along the path are about 0.5 eV/atom for GaAs, AlAs and AlN and above 0.6 eV/atom for GaN - roughly twice those of binary wurtzite nitrides, whose AlN member already requires a coercive field near its ∼10\sim 10 MV/cm critical (breakdown) field. The estimated ZB coercive fields exceed 40 MV/cm, far beyond breakdown; worse, the rocksalt intermediate has a smaller band gap and hence a smaller critical field. The paper sketches two escape routes, each open. First, alloying: in wurtzite, group-III transition-metal ternary alloys (AlScN being the canonical case, Wang et al. 2021) lower the rocksalt energy and the switching barrier, but it is unknown whether such alloying also lowers the layered structures that may host the path's energy maximum in ZB. Second, proximity switching: Skidmore et al. (Nature 2025) demonstrated 'proximity ferroelectricity', in which a non-ferroelectric polar layer (AlN or ZnO, both wurtzite) is reversed by an adjacent ferroelectric layer (Al1−xBxN\mathrm{Al}_{1-x}\mathrm{B}_{x}\mathrm{N}, Al1−xScxN\mathrm{Al}_{1-x}\mathrm{Sc}_{x}\mathrm{N}, or Zn1−xMgxO\mathrm{Zn}_{1-x}\mathrm{Mg}_{x}\mathrm{O}) whose antipolar nuclei and propagating domain walls lower the switching barrier beyond the interface; extending this to a zincblende layer is proposed but unmodeled.

Problem Statement

Achieve and demonstrate ferroelectric switching of a zincblende III-V semiconductor, thereby making the giant spontaneous polarization defined by Cañas, Yassine and Ambacher (2025) experimentally observable. Building on the ⟨111⟩\langle 111\rangle switching path that reverses zincblende polarity through the N-polar layered, centrosymmetric rocksalt, and metal-polar layered structures, answer the following linked questions, which share the single objective of connecting the rigorous bulk definition of spontaneous polarization to realizable switching:

(1) Alloying: can incorporating group-III transition metals (e.g. Sc, the canonical alloying element in the wurtzite case) into a zincblende III-V lower the energy barrier along the ⟨111⟩\langle 111\rangle path enough to enable switching below breakdown - and specifically, does such alloying lower not only the rocksalt configuration but also the layered configurations that may host the highest-energy maximum determining the coercive field?

(2) Proximity switching: can a zincblende layer coupled to an adjacent ferroelectric wurtzite nitride undergo proximity-induced polarization reversal, as demonstrated for non-ferroelectric AlN and ZnO in wurtzite stacks, and what coupling strengths and interfacial conditions are necessary?

(3) System identification: identify specific alloy compositions or heterostructure designs that simultaneously preserve an insulating gap along the entire switching pathway and keep the coercive field below the material's breakdown limit, and validate them computationally and/or experimentally so that the defined spontaneous polarization can be directly observed in a zincblende-derived system.

The sub-questions are independently answerable: a well-supported negative answer to (1) or (2) is a complete resolution of that sub-question, and the task is resolved once all three are answered - with (3) satisfied either by specific designs identified and validated or by a well-supported demonstration that no zincblende-derived system meets the gap and coercive-field requirements together.

References

  1. J. Cañas, M. Yassine, and O. Ambacher, 'Giant spontaneous polarization in zinc-blende III-V semiconductors', Physical Review B 112, L081201 (2025), DOI 10.1103/q7yw-hnmy, https://doi.org/10.1103/q7yw-hnmy (preprint arXiv:2505.19778, https://arxiv.org/abs/2505.19778, where the title is spelled 'zincblende').
  2. 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', Nature 637, 574-579 (2025), DOI 10.1038/s41586-024-08295-y, https://doi.org/10.1038/s41586-024-08295-y
  3. S. Fichtner, N. Wolff, F. Lofink, L. Kienle, and B. Wagner, 'AlScN: A III-V semiconductor based ferroelectric', Journal of Applied Physics 125, 114103 (2019), DOI 10.1063/1.5084945, https://doi.org/10.1063/1.5084945
  4. C. E. Dreyer, A. Janotti, C. G. Van de Walle, and D. Vanderbilt, 'Correct Implementation of Polarization Constants in Wurtzite Materials and Impact on III-Nitrides', Physical Review X 6, 021038 (2016), DOI 10.1103/PhysRevX.6.021038, https://doi.org/10.1103/PhysRevX.6.021038
  5. F. Bernardini, V. Fiorentini, and D. Vanderbilt, 'Accurate calculation of polarization-related quantities in semiconductors', Physical Review B 63, 193201 (2001), DOI 10.1103/PhysRevB.63.193201, https://doi.org/10.1103/PhysRevB.63.193201
  6. H. Wang, N. Adamski, S. Mu, and C. G. Van de Walle, 'Piezoelectric effect and polarization switching in Al1-xScxN', Journal of Applied Physics 130, 104101 (2021), DOI 10.1063/5.0056485, https://doi.org/10.1063/5.0056485
  7. N. L. Adamski, C. E. Dreyer, and C. G. Van de Walle, 'Giant polarization charge density at lattice-matched GaN/ScN interfaces', Applied Physics Letters 115, 232103 (2019), DOI 10.1063/1.5126717, https://doi.org/10.1063/1.5126717
  8. M. Yassine, A. Yassine, A. Nair, B. Sundarapandian, N. Afshar, L. Kirste, S. Fichtner, and O. Ambacher, 'Modeling of polarization reversal-induced interface sheet charge in wurtzite-type AlScN/GaN heterostructures', Journal of Applied Physics 135, 155702 (2024), DOI 10.1063/5.0190885, https://doi.org/10.1063/5.0190885
  9. J. Grümbel, C. Harms, D. V. Dinh, O. Brandt, R. Goldhahn, and M. Feneberg, 'Comprehensive analysis of the anisotropic dielectric response of wurtzite (0001) ScxAl1-xN for 0.1 <= x <= 0.3', Journal of Applied Physics 140, 044102 (2026), DOI 10.1063/5.0343345, https://doi.org/10.1063/5.0343345
  10. E. A. Eliseev, A. N. Morozovska, J.-P. Maria, L.-Q. Chen, and V. Gopalan, 'A Thermodynamic Theory of Proximity Ferroelectricity', arXiv:2501.05219 (2025), https://arxiv.org/abs/2501.05219
  11. C. Ke and S. Liu, 'Proximity Ferroelectricity Driven by Mobile High-Miller-Index Domain Walls', arXiv:2604.25343 (2026), https://arxiv.org/abs/2604.25343
  12. A. Mercado Tejerina, P. Chen, K. Yazawa, A. Zakutayev, L. Bellaiche, and C. Paillard, 'Systematic comparison of approximations and functionals in first-principle calculations of aluminum-based III-V ferroelectric nitrides', arXiv:2605.24334 (2026), https://arxiv.org/abs/2605.24334