Ferroelectric Metals

• Ferroelectric metals are materials that display spontaneous, switchable polarization coexisting with mobile charge carriers, challenging conventional metallic screening.
• Local bonding instabilities and short-range interactions drive polar distortions in these metals, as demonstrated in systems like LiOsO₃ and 2D dichalcogenides.
• The interplay with superconductivity, topology, and anisotropic transport makes ferroelectric metals promising for advanced electronics, optoelectronics, and spintronics.

A ferroelectric metal is a material in which a spontaneous, switchable electric polarization coexists with itinerant charge carriers, defying the canonical view that metallic screening inevitably suppresses long-range dipolar ordering. Historically, the Anderson–Blount hypothesis (1965) conjectured that inversion symmetry, and thus a formal polar axis, could emerge from a continuous structural transition in a metal, even as classical ferroelectric signatures (macroscopic switchable P–E loops) would remain inaccessible due to screening. Recent advances—spanning bulk oxides, perovskite-related heterostructures, atomically thin elemental and compound 2D materials, and van der Waals bilayers—have demonstrated that, under the right symmetry and electronic conditions, stable polar distortions and modestly switchable polar states can indeed coexist with metallicity.

1. Fundamental Mechanisms of Ferroelectricity in Metallic Systems

The key mechanism distinguishing ferroelectric metals from insulating ferroelectrics lies in the interplay between screening, bonding-instability, and structural dimensionality.

In classic displacive ferroelectrics (e.g., BaTiO₃), the polar instability arises from the softening of a zone-center transverse optical phonon, stabilized by long-range Coulomb (dipole-dipole) interactions. In the metallic state, itinerant electrons produce Thomas–Fermi screening with a characteristic screening length $\lambda_\mathrm{TF}$,

$$\lambda_\mathrm{TF} = \left[ \frac{\varepsilon_0 \varepsilon_\infty E_F}{3 n e^2} \right]^{1/2},$$

which for high carrier densities (e.g., Cu) is typically $\sim 0.1$ nm. This screens out most long-range electrostatics, suppressing the classic ferroelectric instability.

However, in materials where the polar distortion is driven by local bonding—such as underbonded A-site cations in LiOsO₃ or geometric mismatches in layered perovskites—the instability is insensitive to free-carrier screening. The effective Hamiltonian can then be written in terms of local “soft mode” amplitudes $u_i$,

$$E_\mathrm{tot}[ \{u\} ] = \sum_i (k_2 u_i^2 + a u_i^4) + \frac{1}{2} \sum_{i \neq j} J_{ij} u_i u_j,$$

where $k_2 < 0$, $a > 0$ define a local double-well, and $J_{ij}$ encompasses short-range pair interactions. In metals, it is the short-range component (unscreened for $r \lesssim 2-3$ Å), not dipolar coupling, that fosters long-range polar order (Xiang, 2013, Benedek et al., 2015).

In reduced dimensions (2D crystals, monolayers, π-bilayers), incomplete out-of-plane screening and symmetry constraints (e.g., forbidden mirror/inversion operations by stacking/sliding) enable persistent, sometimes switchable, macroscopic polarization even in systems with carrier densities comparable to conventional metals (Liu et al., 5 Feb 2026, Zhang et al., 2024, Sheng et al., 2023).

2. Archetypal Bulk Ferroelectric Metals: LiOsO₃ and Its Relatives

LiOsO₃ crystallizes at high temperature in a centrosymmetric $R\overline{3}c$ structure; below $T_0 \approx 140$ K, it undergoes a second-order transition to the polar $R3c$ phase, structurally mimicking the $R\overline{3}c \rightarrow R3c$ transition of insulating LiNbO₃ (Shi et al., 2015). The Li ions shift by $\sim0.5$ Å along [001], breaking inversion but leaving the material metallic, with a finite density of states at $E_F$ both above and below $T_0$. Experimental signatures include:

• A $\lambda$-like specific heat anomaly and lattice parameter splitting at $T_0$
• No macroscopic polarization or dielectric hysteresis due to screening
• Resistivity crossover to a Fermi-liquid $T^2$ regime below $T_0$

The polar transition in LiOsO₃ is purely displacive/order-disorder—local bonding and undercoordination of Li drive the instability, with negligible influence from the conduction electrons (Xiang, 2013, Benedek et al., 2015). Monte Carlo simulations on an effective Hamiltonian yield $T_c^\text{sim} \approx 214$ K, and analogous chemistry predicts room-temperature polar metals in MgReO₃ ($T_c \sim 600$ K).

In doped and strained perovskites (e.g., Nb-doped EuTiO₃, electron-doped CaTiO₃), robust ferroelectricity persists so long as the conduction bands derive from orbitals (e.g., $d_{xy}$ in-plane) spatially decoupled from the polar phonon mode (Xu et al., 2022). This orbital selectivity is critical for preserving the local distortion in a metallic background—an explicit realization of the Anderson–Blount “decoupled-electron” criterion.

3. Two-Dimensional, van der Waals, and Sliding Ferroelectric Metals

Quasi-2D materials circumvent classical screening by geometrical and electronic means:

• Bilayer and π-bilayer dichalcogenides (e.g., PtTe₂, MoTe₂): Composed of two monolayers related by $C_{2z}$ rotation and interlayer slip, these break inversion and develop switchable out-of-plane $P \sim 0.25$–0.5 pC/m via interlayer charge transfer, with barriers as low as 3–56 meV (Sheng et al., 2023). The vertical polarization and metallicity both originate from spatially extended Te-$p_z$ states across the vdW gap; carrier doping tunes both magnitude and sign of $P$.
• Sliding FE metals with magnetism: In AFM π-bilayers derived from Fe₅GeTe₂, in-plane sliding induces simultaneous emergent polarization $P_z$, net ferrimagnetic moment $M$, and anomalous Hall conductivity, with a robust tri-state switching realized through a single sliding path (Guo et al., 11 Aug 2025). Here, Landau theory reveals a bilinear coupling between sliding displacement and polarization, leading to electrically reconfigurable magnetoelectric devices.
• Atomically thin TaNiTe₅: Surface-limited polar distortions confine the ferroelectric order to a $\sim1.3$ nm monolayer, while the mobile carrier density ($n \sim 10^{15}$ cm⁻², order of $10^{22}$ cm⁻³) matches elemental metals (Liu et al., 5 Feb 2026). The internal “dead-layer” model ensures that, despite a tiny $\lambda_{TF}$, the polarization (revealed by PFM and phase contrast) remains unscreened at the surface and is switchable by local fields.
• Switchable polarization in WTe₂ and 2D bimetal phosphates: Few-layer WTe₂ demonstrates gate-switchable out-of-plane polarization ($P \sim 0.1$–$0.3$ μC/cm²) at carrier densities $n \sim 10^{19}$–$10^{20}$ cm$^{-3}$, with bistability persisting to room temperature (Fei et al., 2018). Extensive high-throughput searches and machine learning have uncovered a family of 2D ferroelectric metals (e.g., MMP₂X₆), where spatial separation of conducting electrons (on one face) and the ionic/electronic polarization (on the other) allows the coexistence of high-conductivity and robust, switchable out-of-plane $P$ (Ma et al., 2020).

4. Experimental Signatures and Detection Methodologies

Switchable ferroelectric order in a metal cannot be probed directly via macroscopic P–E loops due to screening; alternative electrical and optical probes are required:

• Nonlinear Hall effect and Berry curvature dipole: In polar metals, the onset of a polar axis generates a finite Berry curvature dipole, leading to a nonlinear Hall response under ac excitation (Xiao et al., 2020). The tensor $\mathbf{D}$ is symmetry-constrained and directly proportional to the polar displacement. Experimental detection schemes yield mV-scale transverse voltages, with sign-reversal on polarization switching, and generalize to a broad class of noncentrosymmetric metals.
• Bulk photovoltaic (shift current) effect: Ferroelectric metals with nontrivial band topology (e.g., PtBi₂, EuAuBi) exhibit giant shift current conductivity, with $\sigma_{xx}^z \sim 370$–$400~\mu$A/V², much larger than in oxide ferroelectrics (Yang et al., 2024, Tan et al., 10 Jul 2025). The response is inherently polarization-dependent and robustly reverses with P-switching, enabling direct optical tests of FE-metal character in photoresponse experiments.
• Flexoelectric switching: While electric-field reversal is prohibited by screening, mechanical strain gradients (flexoelectricity) can efficiently switch the polar distortion in polar metals like LiOsO₃. The critical bending radius for switching is on the nanometer scale (Zabalo et al., 2020).
• Van der Waals heterostructure engineering: Polarization reversal in a 2D FE metal layer at the interface with a semiconductor van der Waals contact modulates Schottky barrier heights, junction types, and vertical transport, with large “giant electroresistance” arising from P-dependent Fermi-level pinning (Ma et al., 2020).

5. Interplay with Superconductivity, Topology, and Correlated Phases

The coupling of polar ground states with metallicity enables new correlated and topological phenomena:

• Superconductivity: Electron–polar phonon coupling in a FE metal introduces an additional attractive channel (arising from the linear polarization–phonon cross-term), yielding an enhanced $T_c$ in the ferroelectric phase. Theoretical models find that both singlet and triplet pairing channels are promoted near the ferroelectric quantum critical point, with $T_c$ substantially exceeding the bare BCS estimate (Zyuzin et al., 2022, Klein et al., 2022).
• Transport anisotropy and non-Fermi liquid response: Linear-in-T, anisotropic resistivity emerges in the ferroelectric phase due to one-phonon scattering off the polar distortion, and transverse voltage drops are predicted even at zero magnetic field—a defined “smoking-gun” for FE metallicity (Zyuzin et al., 2022, Klein et al., 2022).
• Topological edge and bulk states: In certain 2D elemental and binary FE metals, stacking and polar symmetry tuning can induce topological nontrivial band inversions, quantified by $\mathbb{Z}_2=1$ and the presence of protected edge states. These “topological ferroelectric metals” unite switchable polarity, metallic transport, and protected edge currents in the same phase (Zhang et al., 2023, Yang et al., 2024).

6. Design Principles and Materials Classes

Designing robust, switchable FE metals requires incorporating:

• Geometric/chemical drivers: Favor local bonding instabilities over cross-gap charge-transfer mechanisms. Examples include undercoordinated A-sites in layered perovskites, or polar stacking sequences of buckled honeycomb monolayers (Bi, Sn, Ge).
• Electronic decoupling: Ensure conduction electrons populate orbitals spatially orthogonal to soft-mode displacements—commonly achieved in $t_{2g}$-derived perovskites or via out-of-plane $p_z$ orbitals in dichalcogenides.
• Symmetry selection: As polar metals must break inversion ($\mathcal{P}$) and potentially mirror ($M_z$) symmetry, stacking, interlayer sliding, or compositional asymmetry (A-site disorder; vdW registry) can be used to engineer noncentrosymmetric metallic states (Zhang et al., 2024).
• Low screening regime: 2D and ultrathin superlattices, or interfaces with van der Waals heterostructures, maximize the field penetration and allow polar switching with experimentally accessible fields or mechanical means.
• Barrier engineering: Target moderate double-well barriers ($\sim 30$–$140$ meV/f.u.) to balance the demands of nonvolatility and electrically/mechanically feasible switching.

7. Outlook and Applications

The establishment of Fe metals brings forth a new paradigm for multifunctional devices. Applications include:

• Nonvolatile memories and logic: Electrically switchable, metallic channels allow memory elements with both conductance and polarization encoding.
• Multiferroic and spintronic architectures: Simultaneous polarization, magnetization, and anomalous Hall effect, tunable by field or sliding, realize operation modes inaccessible to conventional materials (Guo et al., 11 Aug 2025).
• Photovoltaics and optoelectronics: Giant bulk photovoltaic effect, tunable by polarization, opens a pathway to above-bandgap photovoltages and polarization-encoded photodetectors (Yang et al., 2024, Tan et al., 10 Jul 2025).
• Van der Waals devices: Control over interfacial charge and band alignment through nonvolatile polarization can enable transistor-like operation, reconfigurable Schottky/Ohmic contacts, and topological switches in stacked architectures (Ma et al., 2020).

Continued expansion of materials families using machine learning, stacking engineering, and rational chemical design presage accelerated discovery of FE metals across both atomic and extended inorganic systems. The intersection of topology, superconductivity, and ferroelectric order in the metallic state forms a fertile landscape for emergent quantum materials and devices (Ma et al., 2020, Filippetti et al., 2015, Yang et al., 2024).