Ferromagnetic Ferroelectricity due to Orbital Ordering

Realization of ferromagnetic ferroelectricity, combining two ferroic orders in a single phase, is the longstanding problem of great practical importance. One of the difficulties is that ferromagnetism alone cannot break inversion symmetry $\mathcal{I}$. Therefore, such a phase cannon be obtained by purely magnetic means. Here, we show how it can be designed by making orbital degrees of freedom active. The idea can be traced back to a basic principle of interatomic exchange, which states that an alternation of occupied orbitals along a bond (i.e., antiferro orbital order) favors ferromagnetic coupling. Moreover, the antiferro orbital order breaks $\mathcal{I}$, so that the bond becomes not simply ferromagnetic but also ferroelectric. Then, we formulate main principles governing the realization of such a state in solids, namely: (i) The magnetic atoms should not be located in inversion centers, as in the honeycomb lattice; (ii) The orbitals should be flexible enough to adjust they shape and minimize the energy of exchange interactions; (iii) This flexibility can be achieved by intraatomic interactions, which are responsible for Hund’s second rule and compete with the crystal field splitting; (iv) For octahedrally coordinated transition-metal compounds, the most promising candidates appear to be iodides with a $d^{2}$ configuration and relatively weak $d$-$p$ hybridization. Such a situation is realized in the van der Walls compound VI$_3$, which we expect to be ferromagnetic ferroelectric.

💡 Research Summary

The manuscript tackles the long‑standing challenge of realizing a single‑phase material that simultaneously exhibits ferromagnetism (FM) and ferroelectricity (FE). Conventional multiferroics achieve magnetoelectric coupling by breaking inversion symmetry (𝑰) through non‑collinear spin textures (e.g., cycloidal spirals) or by employing stereochemically active lone‑pair cations. However, these mechanisms inevitably involve antiferromagnetic (AF) components; a pure FM order alone cannot break 𝑰 because the FM configuration is invariant under inversion. Consequently, a different degree of freedom must be invoked to obtain a genuine FM‑FE state.

The authors propose that orbital degrees of freedom provide precisely the missing ingredient. Building on the classic Goodenough‑Kanamori‑Anderson (GKA) rules, they point out that an antiferro‑orbital (AFO) arrangement—where neighboring magnetic ions occupy different d‑orbitals—favors ferromagnetic superexchange. Crucially, the same AFO pattern also makes the two sites inequivalent with respect to the inversion center, thereby breaking 𝑰 at the bond level. In other words, a single bond can simultaneously host a ferromagnetic exchange and generate an electric dipole.

To formalize this idea, the paper invokes modern theories of polarization: the Berry‑phase expression in k‑space and the real‑space Wannier‑function formulation. Since the position operator does not act on spin, any magnetization dependence of the polarization must arise from the spatial asymmetry of the occupied Wannier functions. Within second‑order perturbation theory, the authors show that the occupied Wannier function on site i acquires “tails” on neighboring sites j proportional to hopping integrals t_{ij} between occupied and empty orbitals. When the occupied orbitals differ on the two sites (AFO), the hopping amplitudes become directionally asymmetric, leading to a net shift of electronic charge and thus a finite polarization 𝑷.

The superexchange analysis proceeds from the atomic limit with on‑site Coulomb repulsion U and intra‑atomic Hund’s exchange J_H. The effective exchange constant J_{ij} is derived as

J_{ij} =