Epitaxial superlattices that interleave single formula units of the charge-ordered, ferrimagnetic oxide LuFe2O4 within the improper ferroelectric hexagonal LuFeO3 were shown in 2016 to be magnetoelectric multiferroics with a magnetic transition at 281 K and electric-field control of the ferrimagnetic state at 200 K. The unresolved objective is to turn this near-room-temperature coexistence into deterministic, quantitative, physically understood control: establish the microscopic mechanism by which the ferroelectric lutetium 'trimer' distortion couples to the ferrimagnetic order, determine the spin and charge-ordering ground state of the LuFe2O4 layers as a function of composition, disentangle epitaxial strain from oxygen-stoichiometry effects, correct the quantitative failure of the DFT+U-derived Ising description of the transition temperatures, measure the high-temperature ferroelectric transitions and domain-wall transport of the high-m members, and thereby achieve robust electric-field switching of the ferrimagnetic state at 300 K and above. Later work (2020-2026) has narrowed the charge-ordering pattern of the m=3 member and revealed oxygen-vacancy ordering, but the central coupling mechanism and room-temperature deterministic switching remain open.

Background

A multiferroic is a material that develops two ferroic orders simultaneously - here a spontaneous electric polarization and magnetic order - and a magnetoelectric multiferroic is one in which an electric field can control the magnetic state. Such control would enable memory devices that write magnetism with voltages instead of currents, but single-phase materials combining displacive ferroelectricity with strong magnetism are extremely rare, because the empty dd orbitals favoured by conventional ferroelectricity oppose the partially filled dd orbitals needed for magnetism, and because most known candidates order magnetically only far below room temperature. Hexagonal LuFeO3 (hh-LuFeO3) is an 'improper' or geometric ferroelectric: its polarization is a by-product of a nonpolar 'trimer' distortion, an up-up-down rumpling of the Lu-O planes (the same K3K_3-type trimerization known from the related hexagonal rare-earth manganites), which persists above 1000 K; its antiferromagnetic order, however, sets in only near 147 K, with a weak canting-induced moment. LuFe2O4 is in several ways the complement: its Fe-O bilayers contain both Fe2+ and Fe3+ that charge-order (a candidate 'electronic' ferroelectricity, itself contested), and it is a ferrimagnet near 240 K - a strong moment but at too low a temperature, and with a frustrated, still-debated spin/charge arrangement.

Mundy and co-workers (Nature 537, 523-527; 2016) combined the two by molecular-beam epitaxy (MBE, a technique that deposits elemental beams layer by layer under ultra-high vacuum, allowing monolayer control): they inserted single formula-unit-thick LuFe2O4 slabs into a LuFeO3 matrix, forming (LuFeO3)m/(LuFe2O4)1 superlattices with 1≤m≤101\leq m\leq 10 on yttria-stabilized zirconia substrates. The LuFeO3 matrix forces its trimer rumpling onto the LuFe2O4 slabs, breaking inversion symmetry in them, reducing the magnetic frustration of the Fe-O bilayers, and raising the ferrimagnetic transition from 240 K (bulk LuFe2O4) to 281 K for m = 9, with ferroelectric order that is electrically switchable by piezoresponse force microscopy (PFM, a scanning-probe technique that writes and reads ferroelectric domains through their local electromechanical response) and persists to approximately 550 K (m = 3), 500 K (m = 5), and beyond the measurement range for m = 9. Writing ferroelectric domains at 300 K and imaging the magnetic response by X-ray magnetic circular dichroism photoemission electron microscopy (XMCD-PEEM, an element-specific magnetic imaging technique) at 200 K showed that the ferroelectric pattern controls the ferrimagnetic contrast - i.e. genuine magnetoelectric coupling, but demonstrated only at 200 K. The theory side relied on density functional theory with Hubbard corrections (DFT+U) mapping the Fe-O-Fe super-exchange constants JijJ_{ij} (the magnetic coupling between neighbouring iron moments mediated by the intervening oxygen ion) onto an Ising Hamiltonian simulated by Monte Carlo; this reproduced trends but overestimated the absolute Curie temperature of the charge-ordered Ising (COI) model at roughly 500 K against 240 K experiment, requiring an empirical scale factor of about 0.48. The calculations also found the ferroelectric charge-ordered LuFe2O4 state (COII) only about 4 meV per formula unit above the antiferroelectric ground state, and predicted self-doped, charged domain-wall configurations with conducting Fe 3d3d states.

Since 2016, follow-up work has clarified pieces of this picture but also complicated it. Site-specific optical absorption and magnetic-circular-dichroism spectroscopy combined with DFT (Fan et al., 2020) traced the enhanced Curie temperature to a Lu-layer distortion that selectively strengthens the spin-up Fe2+ to Fe3+ charge-transfer channel and hence the exchange, and identified a non-polar, self-doped charge-ordering arrangement in the m = 3 member; atomically resolved theory (Li et al., 2021) instead finds ferroelectric charge-ordered states generally favoured in the superlattices; atom probe tomography (a microscopy technique that reconstructs the three-dimensional positions and elemental identities of individual atoms; Hunnestad et al., 2024) revealed a three-dimensional order of oxygen vacancies coincident with the LuFe2O4 layers that electron-dopes them and stabilizes charged domain walls - an extrinsic, defect-driven contribution to the very properties the original work attributed to the intrinsic trimer distortion. What is still missing is a single, composition-resolved, quantitatively validated account of the coupling, and the deterministic switching of the magnetic state at room temperature that motivated the design in the first place.

Problem Statement

For the epitaxially engineered (LuFeO3)m/(LuFe2O4)1 superlattices (grown by molecular-beam epitaxy, with formula-unit-thick ferrimagnetic LuFe2O4 layers embedded in an improper-ferroelectric LuFeO3 matrix), establish a complete, composition-resolved, quantitative account of the magnetoelectric coupling and use it to achieve deterministic electric-field control of the ferrimagnetic state at room temperature (300 K and above). This single objective comprises the following scientific parts: (1) the microscopic mechanism by which the ferroelectric Lu 'trimer' distortion couples to the ferrimagnetic order parameter - including the relative weights of the intrinsic exchange enhancement (Lu-layer distortion strengthening the spin-up Fe2+ to Fe3+ charge-transfer channel) and the extrinsic, oxygen-vacancy-driven electron doping of the LuFe2O4 layers, and the conditions under which the coupling permits deterministic, symmetry-equivalent reversal of the ferrimagnetic state at or above 300 K rather than only at 200 K; (2) the spin and charge-ordering ground state of the LuFe2O4-derived layers as a function of superlattice periodicity and domain state, currently constrained only by near-degenerate DFT candidates (differing by a few meV per formula unit) and indirect spectroscopic/neutron evidence; (3) an experimental and/or improved-theoretical disentangling of epitaxial strain from oxygen stoichiometry (vacancy ordering) as causes of the enhanced magnetic transition temperatures, particularly for the lower-mm compositions (termed region I in the source work, where the magnetic transition is enhanced only to about 250 K); (4) an explanation of the quantitative failure of the DFT+U-derived Ising model, which overestimates the magnetic transition temperature by roughly a factor of two (about 500 K calculated versus about 240 K measured) and needed an empirical normalization, in terms of missing electronic correlations, relativistic interactions, or other physics; (5) the measurement of the high-temperature ferroelectric transitions of the high-m superlattices (for example m = 9, whose transition lies beyond the temperature range reachable by the variable-temperature X-ray linear dichroism (an orbital-anisotropy probe used to track the ferroelectric distortion) and electron-microscopy measurements used to date) and of the predicted evolution of domain-wall conductivity and charge screening with superlattice periodicity; and (6) a demonstrated, reproducible, electrically switched ferrimagnetic state at 300 K or above in this family. The mechanistic account of (1) must be consistent with the ground state, defect, and transition-temperature facts above, and the electrically switched ferrimagnetic state of (6) must be demonstrated at 300 K or above in this family.

References

  1. Mundy, J. A., Brooks, C. M., Holtz, M. E., Moyer, J. A., Das, H., Rebola, A. F., Heron, J. T., Clarkson, J. D., Disseler, S. M., Liu, Z., Farhan, A., Held, R., Hovden, R., Padgett, E., Mao, Q., Paik, H., Misra, R., Kourkoutis, L. F., Arenholz, E., Scholl, A., Borchers, J. A., Ratcliff, W. D., Ramesh, R., Fennie, C. J., Schiffer, P., Muller, D. A., Schlom, D. G. Atomically engineered ferroic layers yield a room-temperature magnetoelectric multiferroic. Nature 537, 523-527 (2016). DOI 10.1038/nature19343, https://doi.org/10.1038/nature19343
  2. Fiebig, M. Condensed-matter physics: Multitasking materials from atomic templates. Nature 537, 499-500 (2016). DOI 10.1038/537499a, https://doi.org/10.1038/537499a (News and Views accompanying the source paper)
  3. Fan, S., Das, H., Rebola, A., Smith, K. A., Mundy, J., Brooks, C., Holtz, M. E., Muller, D. A., Fennie, C. J., Ramesh, R., Schlom, D. G., McGill, S., Musfeldt, J. L. Site-specific spectroscopic measurement of spin and charge in (LuFeO3)m/(LuFe2O4)1 multiferroic superlattices. Nature Communications 11, 5582 (2020). DOI 10.1038/s41467-020-19285-9, https://doi.org/10.1038/s41467-020-19285-9
  4. Holtz, M. E., Padgett, E. S., Steinhardt, R., Brooks, C. M., Meier, D., Schlom, D. G., Muller, D. A., Mundy, J. A. Dimensionality-Induced Change in Topological Order in Multiferroic Oxide Superlattices. Physical Review Letters 126, 157601 (2021). DOI 10.1103/PhysRevLett.126.157601, https://doi.org/10.1103/PhysRevLett.126.157601 (arXiv:2007.03872, https://arxiv.org/abs/2007.03872)
  5. Hunnestad, K. A., Das, H., Hatzoglou, C., Holtz, M., Brooks, C. M., van Helvoort, A. T. J., Muller, D. A., Schlom, D. G., Mundy, J. A., Meier, D. 3D oxygen vacancy distribution and defect-property relations in an oxide heterostructure. Nature Communications 15, 5400 (2024). DOI 10.1038/s41467-024-49437-0, https://doi.org/10.1038/s41467-024-49437-0 (study of (LuFeO3)9/(LuFe2O4)1)
  6. Li, M., Cheng, S., Wang, W., Li, X., Wang, N., Zhu, Y. Interplay between charge ordering and geometric ferroelectricity in LuFe2O4/LuFeO3 superlattices. Physical Review Materials 5, 094412 (2021). DOI 10.1103/PhysRevMaterials.5.094412, https://doi.org/10.1103/PhysRevMaterials.5.094412
  7. Das, H. Coupling between improper ferroelectricity and ferrimagnetism in the hexagonal ferrite LuFeO3. Physical Review Research 5, 013007 (2023). DOI 10.1103/PhysRevResearch.5.013007, https://doi.org/10.1103/PhysRevResearch.5.013007
  8. Liu, F., Hao, Y., Ni, J., Zhao, Y., Zhang, D., Fabbris, G., Haskel, D., Cheng, S., Xu, X., Yin, L., Xiang, H., Zhao, J., Lü, X., Wang, W., Shen, J., Yang, W. Pressure-induced charge orders and their postulated coupling to magnetism in hexagonal multiferroic LuFe2O4. npj Quantum Materials 8, 1 (2023). DOI 10.1038/s41535-022-00522-x, https://doi.org/10.1038/s41535-022-00522-x
  9. Li, X., Yun, Y., Xu, X. Recent progress on multiferroic hexagonal rare-earth ferrites (h-RFeO3, R = Y, Dy-Lu). Journal of Physics D: Applied Physics 58, 073003 (2025). DOI 10.1088/1361-6463/ad97c5, https://doi.org/10.1088/1361-6463/ad97c5
  10. Spaldin, N. A. The chemistry of magnetoelectric multiferroics. Nature Reviews Chemistry 10, 415-426 (2026). DOI 10.1038/s41570-026-00827-6, https://doi.org/10.1038/s41570-026-00827-6
  11. Fert, A., Ramesh, R., Garcia, V., Casanova, F., Bibes, M. Electrical control of magnetism by electric field and current-induced torques. Reviews of Modern Physics 96, 015005 (2024). DOI 10.1103/RevModPhys.96.015005, https://doi.org/10.1103/RevModPhys.96.015005
  12. Qin, Y. Preparation and physical properties of LuFe2O4±x polycrystalline samples with different oxygen contents. Next Materials 13, 103040 (2026). DOI 10.1016/j.nxmate.2026.103040, https://doi.org/10.1016/j.nxmate.2026.103040