Orbital Altermagnetism

• Orbital altermagnetism is a magnetic order where orbital degrees of freedom, rather than spin, drive momentum-dependent band splitting with no net magnetization.
• Key mechanisms include crystal-field-induced orbital-spin locking, spontaneous antiferro-orbital order, and hybridization-driven effects that break Kramers degeneracy.
• The phenomenon manifests in diverse systems, influencing optical, transport, and piezomagnetic responses with potential applications in spintronics and quantum materials.

Searching arXiv for papers on orbital altermagnetism and closely related orbital-driven altermagnetic phenomena. Orbital altermagnetism denotes a family of symmetry-governed phenomena in which orbital degrees of freedom are not passive labels of electronic states but the microscopic agents that generate, select, or carry altermagnetic order. In one usage, orbital character, crystal-field staggering, orbital order, orbital hybridization, or loop currents produce the momentum-dependent spin splitting of a compensated collinear magnet; in another, the term denotes a symmetry-protected magnetic order of pure orbital degrees of freedom, characterized by ordered anti-parallel orbital magnetic moments in real space and momentum-dependent orbital band splittings in reciprocal space (Vila et al., 2024, Leeb et al., 2023, Pan et al., 1 Oct 2025). Across these usages, the unifying theme is the removal of the equivalence between opposite-spin or opposite-orbital sectors without generating a net magnetization, enabling even-parity $d$-wave-, $g$-wave-, or related band splittings, orbital-spin locking, and orbital-based optical and transport responses (Ghorai et al., 3 May 2026, Che et al., 24 May 2026).

1. Concept and symmetry content

Altermagnetism is defined in the recent literature as a compensated collinear magnetic order with momentum-dependent band splitting, distinct from both ferromagnetism and conventional antiferromagnetism. Orbital altermagnetism extends this framework into the orbital sector. The most explicit definition introduces a symmetry-protected magnetic order of pure orbital degrees of freedom, with anti-parallel orbital magnetic moments in real space but momentum-dependent orbital band splittings, directly analogous to spin altermagnetism (Pan et al., 1 Oct 2025).

In that pure-orbital formulation, the relevant symmetry statement is that orbital altermagnetism arises only when both $\mathcal{PT}$ and $C_{2z}\mathcal{T}$ are absent, since those anti-unitary symmetries would enforce zero orbital magnetization. The microscopic order can originate from staggered loop currents, and the reciprocal-space manifestation is a momentum-dependent orbital splitting with $d$-wave-like orbital-momentum locking (Pan et al., 1 Oct 2025). This makes orbital altermagnetism neither a synonym for spin altermagnetism nor merely a weak spin-orbit-coupling correction to it.

A broader usage emphasizes orbital degrees of freedom as the microscopic origin of conventional spin altermagnetism. In this sense, crystal-field-induced sublattice orbital differentiation, spontaneous antiferro-orbital order, interorbital hybridization, or orbital-selective Mott physics generate the anisotropies that lift Kramers degeneracy in compensated collinear magnets (Vila et al., 2024, Leeb et al., 2023, Cuono et al., 2023). This literature treats orbital physics as indispensable to the onset, classification, and detection of altermagnetism, even when the directly observed splitting is spin-resolved rather than orbital-resolved.

The symmetry setting is correspondingly broader. Interwoven dual-orbital square-lattice configurations generate $d$-wave or $g$-wave altermagnetic states, whereas single-orbital square lattices remain spin-degenerate within the minimal antiferromagnetic model (Che et al., 24 May 2026). Orbital altermagnetism has also been extended beyond even-sublattice settings: on the kagome lattice, non-uniform orbital moments can yield collinear altermagnetic-like states despite the odd number of sublattices (Chakraborty et al., 30 Sep 2025). A further generalization shows that local spin-orbital spontaneous symmetry breaking can realize altermagnetism in amorphous, non-crystalline systems, so that global crystal rotation symmetry is not fundamental in the strong sense assumed by early crystalline formulations (d'Ornellas et al., 11 Apr 2025).

2. Microscopic routes from orbitals to altermagnetic splitting

A central microscopic mechanism is crystal-field-induced orbital-sublattice-spin coupling. In a minimal model for $d$-wave altermagnets, the Hamiltonian is written as

$$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$

with

$$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$

$g$0

Here the crystal field alternates sign between sublattices, locks $g$1 and $g$2 orbital character to the sublattice degree of freedom, and the antiferromagnetic exchange locks the sublattice to spin, thereby establishing orbital-spin locking (Vila et al., 2024). The mechanism is explicitly momentum dependent, and anisotropic hopping $g$3 is essential; with isotropic hopping, neither orbital-spin locking nor altermagnetic order emerges in that model (Vila et al., 2024).

A second route is interaction-driven orbital ordering. A two-orbital square-lattice model with staggered antiferromagnetic order and staggered orbital order at $g$4 shows that altermagnetism can arise even when crystallographic sublattice anisotropy is absent. In that setting, orbital order spontaneously breaks lattice rotation symmetry, the two spin sublattices are related only by $g$5 rotation combined with orbital exchange, and the coexistence of staggered magnetization $g$6 and staggered orbital order $g$7 produces spin-split Fermi surfaces with zero net magnetization (Leeb et al., 2023). This establishes an entirely interaction-driven origin of altermagnetism.

A closely related mechanism appears in correlation-driven metallic two-dimensional systems. Antiferro-orbital ordering between $g$8 and $g$9 breaks the equivalence of opposite-spin magnetic sublattices and yields a symmetry-enforced altermagnetic spin texture. For a square-lattice description, the splitting takes the form

$\mathcal{PT}$0

which is explicitly $\mathcal{PT}$1-wave-like and vanishes on the Brillouin-zone diagonals (Jana et al., 26 Mar 2026). In that work, Fermi-surface nesting is identified as the driver of the antiferro-orbital instability.

A third route emphasizes orbital hybridization rather than orbital order alone. A many-body comparison of MnF$\mathcal{PT}$2, MnTe, and RuO$\mathcal{PT}$3 distinguishes a strongly localized Mott regime with weak hybridization, a strongly correlated regime with robust ligand hybridization, and an itinerant regime with significant hybridization. The analysis concludes that strong local electron correlations and judicious ligand selection to promote orbital hybridization are both key prerequisites for realizing altermagnetism in strongly correlated systems (Kang et al., 14 May 2026). This is a corrective to any simplified view that stronger correlation automatically implies stronger altermagnetic splitting.

3. Reciprocal-space structure, orbital locking, and model classes

The reciprocal-space structure of orbital altermagnetism is governed by orbital anisotropy. In correlation-driven $\mathcal{PT}$4/$\mathcal{PT}$5 systems, the difference between orbital dispersions,

$\mathcal{PT}$6

directly generates the $\mathcal{PT}$7-wave spin splitting once staggered orbital polarization is established (Jana et al., 26 Mar 2026). In the crystal-field picture, the same orbital asymmetry becomes an orbital-spin-locking problem: $\mathcal{PT}$8-polarized light couples to $\mathcal{PT}$9, $C_{2z}\mathcal{T}$0-polarized light to $C_{2z}\mathcal{T}$1, and the spin character follows from the orbital-sublattice-spin chain (Vila et al., 2024). In both formulations, even-parity momentum structure is not an auxiliary detail but the defining observable consequence of the orbital mechanism.

The square-lattice classification makes this dependence especially transparent. Within the minimal antiferromagnetic square-lattice model, single-orbital lattices remain spin-degenerate, whereas interwoven dual-orbital configurations lift Kramers degeneracy and realize either $C_{2z}\mathcal{T}$2-wave or $C_{2z}\mathcal{T}$3-wave altermagnetic states. The splitting originates from orbital anisotropy in same-spin hopping channels, so the wavefunction character itself becomes a design parameter for altermagnetism (Che et al., 24 May 2026). In this framework, $C_{2z}\mathcal{T}$4 interweaving produces $C_{2z}\mathcal{T}$5-wave behavior, while suitable $C_{2z}\mathcal{T}$6-orbital combinations generate $C_{2z}\mathcal{T}$7-wave structure.

The pure-orbital minimal model on the square-kagome lattice realizes a related but distinct phenomenon. Complex nearest-neighbor hoppings induce staggered loop currents, producing compensated orbital moments in real space and momentum-dependent orbital splitting in reciprocal space. The resulting orbital response has $C_{2z}\mathcal{T}$8-wave-like orbital-momentum locking (Pan et al., 1 Oct 2025). Here the altermagnetic object is the orbital magnetic moment itself, not a spin splitting inherited from orbital order.

Other lattices realize modified angular structures. In single-layer Sr$C_{2z}\mathcal{T}$9RuO$d$0 with octahedral rotations, the effective $d$1 model yields an orbital-selective $d$2-wave altermagnetic phase, driven primarily by second- and third-nearest-neighbor interorbital hybridizations in the $d$3 sector; in the bulk, interlayer hopping lowers the symmetry and produces a $d$4-wave altermagnet while retaining orbital selectiveness (Autieri et al., 24 Jan 2025). On the honeycomb lattice, strained monolayer VCl$d$5 hosts a nematic $d$6-wave altermagnetic spin splitting with $d$7 and two nodal lines, because orbital order lowers the lattice symmetry from the pristine threefold structure (Camerano et al., 25 Mar 2025).

4. Material realizations and orbital selectivity

The materials literature shows that orbital altermagnetism is not confined to a single microscopic class. The following examples organize the diversity of currently proposed realizations.

System or class	Orbital ingredient	Reported outcome
Ca$d$8RuO$d$9, YVO$d$0	Orbital-selective Mott physics or robust orbital order	Orbital-selective altermagnetism; correlation-enhanced spin splitting
Sr$d$1RuO$d$2	Octahedral-rotation-enabled interorbital hybridization	Orbital-selective $d$3-wave single layer and $d$4-wave bulk altermagnetism
Sr$d$5Cr$d$6O$d$7	Spontaneous orbital ordering in layered chromates	Odd $d$8 altermagnetic, even $d$9 anti-altermagnetic, strain-tunable
Monolayer YbMn$g$0Ge$g$1	Correlation-driven antiferro-orbital order	Stable 2D metallic altermagnet with spin splitting of order 1 eV
LaTiO$g$2	Ordered $g$3 occupation with specific $g$4 character	Fragile altermagnetism destroyed by orbital disorder
MnTe	Strong Mn-$g$5–Te-$g$6 hybridization and dominant orbital magnetization	Robust altermagnetic splitting with dominant orbital magnetization

In Ca$g$7RuO$g$8 and YVO$g$9, first-principles calculations identify orbital-selective altermagnetism in the $d$0 sector. In Ca$d$1RuO$d$2, the quasi-two-dimensional $d$3 orbital is protected by a mirror-plane symmetry that suppresses altermagnetic splitting, whereas $d$4 states display pronounced splitting. In YVO$d$5, altermagnetism is present for A-, C-, and G-type magnetic orders, while the magnitude of the non-relativistic splitting increases by a factor of 4–5 as the Coulomb repulsion $d$6 is increased (Cuono et al., 2023).

Sr$d$7RuO$d$8 adds a structurally controlled version of the same theme. The altermagnetic phase appears for a range of finite octahedral rotations; the $d$9 sector dominates because the $$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$0 sector requires much longer-range intraorbital hopping to participate, making the effect strongly orbital-selective. When spin-orbit coupling is included, an effective staggered Dzyaloshinskii–Moriya interaction generates weak ferromagnetism (Autieri et al., 24 Jan 2025).

In the Ruddlesden–Popper chromates Sr$$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$1Cr$$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$2O$$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$3, spontaneous orbital ordering rather than fixed crystal symmetry distinguishes aligned altermagnetic and compensated anti-altermagnetic phases. If spin and orbital orders align in adjacent layers, the system displays net spin splitting; if either reverses from layer to layer, local altermagnetism persists but the global splitting is compensated. Odd $$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$4 compounds can host altermagnetic phases, whereas even $$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$5 systems and the perovskite limit remain anti-altermagnetic; increasing $$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$6 favors metallicity, and epitaxial strain can promote the altermagnetic phase (Meier et al., 3 Feb 2025).

A particularly strong two-dimensional realization is monolayer YbMn$$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$7Ge$$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$8, where spontaneous orbital order driven by correlations and Fermi-surface nesting yields giant nonrelativistic spin splitting of order 1 eV and a gate-tunable transverse spin conductivity. The reported Fermi-surface contribution reaches $$\mathcal{H}=\mathcal{H}_0+\mathcal{H}_{CF}+\mathcal{H}_{AFM},$$9 S per monolayer (Jana et al., 26 Mar 2026). This provides an explicit metallic 2D case in which orbital order is the primary microscopic route to altermagnetism.

LaTiO$$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$0 illustrates the opposite limit, namely fragility. Its altermagnetic ground state requires Ti-site occupations with $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$1 and $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$2 on paired sites. When the single electron is distributed almost equally among two or three $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$3 orbitals, the spin splitting is suppressed and the system reverts to ordinary antiferromagnetism; spin-orbit coupling is discussed as a source of orbital disorder that can damage the altermagnetic state (Maznichenko et al., 2024).

MnTe has become a key test case for the orbital sector itself. First-principles calculations find a net orbital magnetization of $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$4 per unit cell along $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$5, compared with a spin magnetization of $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$6 in the same direction, while a many-body study identifies MnTe as the correlated-hybridized regime in which robust Mn-$$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$7–Te-$$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$8 orbital hybridization preserves large spin-band splitting of about 1 eV (Ye et al., 13 May 2025, Kang et al., 14 May 2026). Together, these works show that orbital magnetization is not a negligible correction in at least some altermagnets.

5. Optical, transport, and piezomagnetic responses

Orbital altermagnetism has a characteristic response phenomenology because orbital texture couples directly to polarization, strain, and current. In the orbital-spin-locking model for $$\mathcal{H}_0=\sigma_x\left(T_x\cos k_x a+T_y\cos k_y a\right)s_0,$$9-wave altermagnets, the optical matrix elements are

$g$00

with $g$01 (Vila et al., 2024). Since $g$02-polarized light couples to $g$03 and $g$04-polarized light to $g$05, orbital-spin locking makes the absorption momentum dependent and spin selective; changing photon energy can նույնիսկ reverse which spin is excited for a given polarization. The predicted magneto-optical signature is magnetic linear dichroism, which appears only when a magnetic field is applied parallel to the Néel vector and vanishes for a perpendicular field (Vila et al., 2024).

Light can also induce large orbital magnetism in altermagnets. A first-principles study of rutile RuO$g$06 and CoF$g$07 shows that in the non-relativistic limit, symmetry-allowed sublattice-resolved orbital responses are strongly canted, while spin-orbit coupling enhances the total effect by up to two orders of magnitude, introduces strong polarization anisotropy, and suppresses the canting. The induced orbital moment in RuO$g$08 can reach approximately $g$09 per formula unit under the specified laser conditions, and linearly polarized light in light altermagnets can induce moments exceeding those predicted for heavy ferromagnets under circularly polarized light (Adamantopoulos et al., 2024).

Transport theory has likewise moved beyond spin-only descriptions. The orbital-splitter current is defined as the orbital analogue of the spin-splitter current, namely a pure transverse charge-neutral orbital angular momentum current generated by a longitudinal charge current. Within a density-matrix framework, the response decomposes as

$g$10

with Drude and orbital-Berry-curvature contributions (Ghorai et al., 3 May 2026). In FeSb$g$11, mirror symmetries force the orbital magnetic moment to vanish throughout the Brillouin zone, suppress the Drude channel, and leave a purely intrinsic orbital-splitter current. For selected field orientations, the orbital response exceeds the spin-splitter current by nearly a factor of four; in an altermagnet–ferromagnet heterostructure, the combined orbital and spin splitter currents reduce the simulated switching time by about a factor of three (Ghorai et al., 3 May 2026).

The strain response defines another orbital diagnostic. For pure insulating two-dimensional altermagnets, microscopic expressions for orbital magnetization under strain show that $g$12-wave altermagnets possess a linear orbital piezomagnetic polarizability, whereas the $g$13-wave case requires a nonlinear response. In all three analyzed tetragonal models, the polarizability is tied to the Berry curvature of the occupied bands (Bell et al., 10 Feb 2026). This gives orbital altermagnetism a direct piezomagnetic signature even when net spin magnetization is symmetry forbidden.

A complementary emergent-field formulation introduces the orbital Néel vector as a dynamical variable alongside the spin Néel vector. The extended Hamiltonian yields orbital and spin emergent electromagnetic fields, as well as orbital and magnetic multipole currents. In that framework, lattice anisotropy is required for some charge and octupole responses, and dynamic lattice distortion can generate non-vanishing emergent electric fields even for simplified spin and orbital textures (Choi et al., 7 Apr 2026). The formalism is explicitly generalized beyond $g$14-wave systems.

6. Extensions, analogues, and unresolved issues

One major extension is the claim that orbital altermagnetism need not be a crystalline phenomenon. A model on an amorphous lattice with two orthogonal orbitals per site and spin-orbital interactions shows that local spontaneous symmetry breaking in orbital and spin space can generate altermagnetism without global rotational symmetry. The resulting phase exhibits anisotropic spin transport and anisotropic spin spectral functions despite the non-crystalline setting (d'Ornellas et al., 11 Apr 2025). This substantially broadens the conceptual scope of orbital-driven altermagnetism.

A second extension is bosonic. The first experimental realization of an orbital altermagnetic photonic crystal uses an antiunitary $g$15 symmetry to enforce a correspondence between a local $g$16-orbital $g$17 doublet and crystal momentum. The measured band structure and iso-frequency contours display momentum-dependent pseudospin splitting with alternating pseudospin polarization and a $g$18-wave form factor, together with pseudospin-selective transport, pseudospin splitting, and pseudospin filtering of electromagnetic waves (Qiu et al., 27 May 2026). This result suggests that orbital altermagnetic design principles are not restricted to electronic matter.

Kagome systems introduce a third generalization. In a model relevant to $g$19V$g$20Sb$g$21, intertwined charge-density-wave and loop-current instabilities near the van Hove singularity yield orbital ferromagnetic, antiferromagnetic, and altermagnetic phases inside the charge-ordered state. In the altermagnetic case, the compensated orbital moments are related by point-group symmetry rather than translation, and spin-orbit coupling transfers the orbital order into nodal, $g$22-wave-like spin splitting in the electronic structure (Chakraborty et al., 30 Sep 2025). This extends collinear altermagnetic-like states to odd-sublattice lattices via non-uniform orbital moments.

Several open issues remain visible across the literature. One is robustness: LaTiO$g$23 shows that orbital disorder can destroy altermagnetism, while MnF$g$24 shows that strong local correlations can suppress spin-band splitting when hybridization is weak (Maznichenko et al., 2024, Kang et al., 14 May 2026). Another is interpretation of experiments: in MnTe, the orbital contribution to magnetization is much larger than the spin contribution, implying that Hall and magneto-optical responses cannot always be interpreted in spin-only terms (Ye et al., 13 May 2025). A final issue is scope: the crystal-field analysis of orbital-spin locking was presented as general to other altermagnetic classes beyond the $g$25-wave case, and later work on square lattices, kagome systems, amorphous models, and photonic crystals supports that broader view (Vila et al., 2024, Che et al., 24 May 2026).

Taken together, these developments establish orbital altermagnetism as both a microscopic mechanism and, in the pure-orbital sense, a distinct magnetic order parameter. The field now encompasses crystal-field-driven orbital-spin locking, spontaneous orbital ordering, hybridization-enabled correlated altermagnetism, staggered loop-current phases, orbital-selective responses, and bosonic analogues, all linked by the same organizing principle: compensated real-space order with symmetry-protected, momentum-dependent orbital or spin differentiation in reciprocal space.