Altermagnetic Spintronics

• Altermagnetic spintronics is defined by collinear, compensated magnetic order that yields momentum-dependent even-parity spin splitting without net magnetization.
• It enables robust spin-polarized transport, anomalous Hall effects, and efficient charge-to-spin conversion in advanced device architectures.
• Research focuses on symmetry control, interface engineering, and ligand-assisted coupling to tailor spin filtering and fast spin dynamics.

Searching arXiv for recent and foundational papers on altermagnetic spintronics, device concepts, and control mechanisms. Altermagnetic spintronics is the branch of spintronics that exploits the distinctive combination of collinear compensated magnetic order, zero net magnetization, and strong momentum-dependent non-relativistic spin splitting in altermagnets. In the symmetry-based formulation, altermagnetism is neither conventional ferromagnetism nor conventional collinear antiferromagnetism: opposite-spin sublattices are related by crystal rotations or mirrors rather than by simple translation or inversion, so the electronic structure acquires even-parity spin splitting of $d$-, $g$-, or $i$-wave type while retaining magnetic compensation. This permits strongly spin-polarized transport, anomalous Hall and magneto-optical responses, spin filtering, and fast spin dynamics without ferromagnetic stray fields, and it has motivated a rapid expansion of device concepts based on tunnel junctions, bilayers, heterostructures, ferroelectric coupling, superconducting hybrids, and electrically controlled phase transitions (Jungwirth et al., 13 Aug 2025).

1. Symmetry framework and defining characteristics

Altermagnetism is defined by spontaneous breaking of combined spin–space–crystal symmetry in a collinear compensated magnet. In the non-relativistic limit, the ordering breaks spin-space rotation symmetry and crystal rotation symmetry, but preserves a combined symmetry that couples a spin rotation to a real-space rotation. The resulting electronic structure is spin split at generic $\mathbf{k}$ despite zero net magnetization. A standard representation is

$$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$

with $\Delta(\mathbf{k})$ changing sign across symmetry-enforced nodal surfaces; for a $d_{x^2-y^2}$-type case in two dimensions, $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$ (Jungwirth et al., 13 Aug 2025).

This symmetry structure distinguishes altermagnets from ferromagnets, where spin splitting is essentially $s$-wave and tied to nonzero $\mathbf{M}$, and from conventional collinear antiferromagnets, where translation or inversion combined with time reversal restores spin degeneracy at each $g$0. In altermagnets, the local moments remain collinear and compensated, but the momentum-space spin polarization is anisotropic and sign alternating. This makes spin-up and spin-down good quantum numbers in the non-relativistic limit while still permitting strong $g$1-breaking transport signatures (Jungwirth et al., 13 Aug 2025).

Although the original formulation emphasized collinear two-sublattice states, the concept has also been extended to accommodate multiple spin and local-structure variations, including non-collinear spins. In that extended formulation, altermagnetic behavior is organized by the combined pattern of spin directions and alternating local structural directors, and the classification into type-I, type-II, and type-III altermagnets specifies whether ferromagnetic-like responses appear spontaneously or only under external perturbations that preserve $g$2 (Cheong et al., 2024).

2. Spin transport, charge-to-spin conversion, and real-space spin dynamics

The central transport consequence of altermagnetism is spin-polarized current without net magnetization. In linear response, the spin current takes the form

$g$3

and the conductivity tensor is tightly constrained by the spin point group. In the quasi-two-dimensional $g$4-wave altermagnet KV$g$5Se$g$6O, first-principles Kubo–Bastin calculations at 300 K yield room-temperature non-relativistic charge-to-spin conversion with maximum longitudinal spin polarization $g$7 and maximum spin Hall angle also exceeding $g$8; both quantities show a $g$9 angular dependence for an in-plane electric field rotated by angle $i$0, directly reflecting the underlying $i$1-wave symmetry (Zhang et al., 23 Dec 2025).

This transport regime is closely related to the non-relativistic spin-splitter effect discussed in the broader review literature. For a model $i$2-wave altermagnet, an electric field along one crystalline axis produces a longitudinally spin-polarized current, whereas a field along the in-plane diagonal can generate a transverse pure spin current because spin-up and spin-down carriers are deflected in opposite directions. The review emphasizes that this spin-splitter effect is formally analogous to the relativistic spin Hall effect but is $i$3-breaking, non-relativistic, and can exceed typical spin Hall current amplitudes by two orders of magnitude (Jungwirth et al., 13 Aug 2025).

Altermagnetic transport can also manifest as real-space spin precession. In a multi-terminal $i$4-wave altermagnet, the momentum-space splitting $i$5 produces a phase difference between spin-up and spin-down propagators, yielding spatial modulation

$i$6

The resulting Hall voltage oscillates along the device, and the oscillation period directly measures the altermagnetic splitting strength. When the splitting is electrically tunable, the same platform functions as a spin transistor; the paper further reports robustness against dephasing and crystalline warping (Liu et al., 7 Nov 2025).

3. Materials platforms and microscopic origin of spin splitting

The materials landscape of altermagnetic spintronics now spans three-dimensional inorganic compounds, quasi-two-dimensional vanadium chalcogenide oxides, transition-metal dichalcogenide derivatives, oxides, fluorides, and layered heterostructures. Experimentally discussed compounds include MnTe, CrSb, KV$i$7Se$i$8O, RbV$i$9Te$\mathbf{k}$0O, RuO$\mathbf{k}$1, Mn$\mathbf{k}$2Si$\mathbf{k}$3, VNb$\mathbf{k}$4S$\mathbf{k}$5, FeS, TmFeO$\mathbf{k}$6, and La$\mathbf{k}$7CuO$\mathbf{k}$8; direct band-structure probes, spontaneous anomalous Hall measurements, and domain imaging have all entered the field (Jungwirth et al., 13 Aug 2025).

At the microscopic level, recent work has moved beyond symmetry-allowed $\mathbf{k}$9 terms to identify the dominant real-space bonding processes. In the $$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$0-wave altermagnet Co$$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$1NbSe$$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$2, first-principles Wannier Hamiltonian engineering shows that the spin splitting is captured by a short-range tight-binding model and that the dominant contribution does not come from direct magnetic-ion hopping. Instead, ligand-mediated hybridization is decisive: Co–Se and Nb–Se pathways transfer anisotropy from localized magnetic sites to itinerant Nb bands, establishing ligand-assisted coupling as the key mechanism of altermagnetic spin splitting (Camerano et al., 20 May 2026).

A complementary route is interaction-driven itinerant altermagnetism. In a two-orbital Hubbard model with strong orbital anisotropy and van Hove singularities, altermagnetism emerges over a broad range of interactions and dopings when each orbital develops ferromagnetic-like polarization but the two orbitals on the same site polarize oppositely, $$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$3, so the total site moment vanishes while the quasiparticle bands remain spin split. The work identifies tunability of the spin-charge conversion ratio as a central consequence of this correlation-driven mechanism (Giuli et al., 2024).

Atomic-scale studies of MnTe have further shown that real altermagnets may depart substantially from ideal high-symmetry limits. Combining HAADF-STEM and EMCD, recent work finds that MnTe is not an ideal uniform $$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$4 $$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$5-wave altermagnet at the atomic scale, but hosts ubiquitous inversion-symmetry-breaking distortions that lower the spin-space-group symmetry, admit $$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$6-wave altermagnetic components, and in lower-symmetry regimes even allow $$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$7-wave spin splitting. In that picture, local lattice symmetry becomes a direct control knob for spin splitting, spin-current generation, and multiferroic functionality (Ren et al., 26 May 2026).

4. Device architectures

A wide range of device concepts has emerged by translating momentum-space spin splitting into measurable tunnel, interface, and mesoscopic transport signatures.

Architecture	Representative platform	Reported functionality
AFMTJ	KV$$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$8Se$$E_{\uparrow}(\mathbf{k}) = E_0(\mathbf{k}) + \Delta(\mathbf{k}),\qquad E_{\downarrow}(\mathbf{k}) = E_0(\mathbf{k}) - \Delta(\mathbf{k}),$$9O / SrTiO$\Delta(\mathbf{k})$0 / KV$\Delta(\mathbf{k})$1Se$\Delta(\mathbf{k})$2O (Zhang et al., 23 Dec 2025)	TMR $\Delta(\mathbf{k})$3, robust to Fermi-level shifts
All-altermagnetic tunnel junction	RuO$\Delta(\mathbf{k})$4 / NiF$\Delta(\mathbf{k})$5 / RuO$\Delta(\mathbf{k})$6 (Zhang et al., 27 Oct 2025)	TMR $\Delta(\mathbf{k})$7; spin filtering $\Delta(\mathbf{k})$8 in both spin channels
Schottky injector	Altermagnet/semiconductor contact (Ang, 2023)	Spin-polarized thermionic injection without ferromagnetism; $\Delta(\mathbf{k})$9
Electrically gated heterostructure	Normal metal / 2D $d_{x^2-y^2}$0-wave altermagnet (Fu et al., 5 Jun 2025)	Gate-tunable full spin polarization in strong altermagnets; double-gated spin valve
Multi-terminal transistor	$d_{x^2-y^2}$1-wave altermagnet with Hall probes (Liu et al., 7 Nov 2025)	Hall-voltage oscillation and altermagnetic spin transistor
Open-orbit platform	Lifshitz-transitioned 2D altermagnet (Kang et al., 3 Jun 2026)	Magnetic spin filtering and persistent spin currents

The AFMTJ proposal based on KV$d_{x^2-y^2}$2Se$d_{x^2-y^2}$3O attributes giant TMR to a momentum-space spin filter formed by quasi-two-dimensional, strongly anisotropic, spin-resolved Fermi surfaces in the electrodes and a barrier decay-rate profile in SrTiO$d_{x^2-y^2}$4. The calculated Fermi-level transmissions differ by roughly twelve orders of magnitude between parallel and antiparallel Néel-vector configurations, and the TMR remains above $d_{x^2-y^2}$5 for Fermi-level shifts between $d_{x^2-y^2}$6 and $d_{x^2-y^2}$7 eV (Zhang et al., 23 Dec 2025).

The all-altermagnetic tunnel-junction proposal RuO$d_{x^2-y^2}$8/NiF$d_{x^2-y^2}$9/RuO$\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$0 replaces both ferromagnetic electrodes and nonmagnetic barriers with altermagnets. The key mechanism is the synergistic or antagonistic alignment of momentum-dependent spin splitting in the metallic electrodes and insulating barrier. Depending on the magnetic configuration, the same stack can operate as a strong spin-up filter, a strong spin-down filter, or a low-transmission state, enabling multistate magnetoresistance and spin filtering without stray fields (Zhang et al., 27 Oct 2025).

At interfaces, the altermagnetic Schottky contact extends spintronics into semiconductor injection. For a two-dimensional altermagnet with anisotropic spin-contrasting Fermi surfaces, thermal injection across an altermagnet/semiconductor barrier is spin polarized even though the electrode has zero net magnetization. The model predicts analytic orientation dependence of the thermionic current and a maximum spin polarization of approximately $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$1 at $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$2 and $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$3, identifying interface orientation as a primary design parameter (Ang, 2023).

5. Electrical, optical, ferroelectric, and strain control

A defining feature of altermagnetic spintronics is the growing set of nonmagnetic control knobs for spin polarization. In a CrS bilayer with C-type antiferromagnetic stacking, layer-resolved analysis reveals hidden layer-spin locking: each layer is individually altermagnetic, but the global band structure is spin degenerate because $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$4 symmetry is restored. An out-of-plane electric field lifts the layer degeneracy and fully reverses the current spin polarization when the field polarity is reversed, with sign-reversible polarization up to $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$5 at room temperature (Peng et al., 2024).

Optical control has been formulated in a bilayer altermagnetic Mott insulator described by an SU(4) flavor-wave theory. There, a counter-flow in-plane electric field induces a polarization current, which in turn drives a spin current in each layer. The polarization current is isotropic, but the spin current is anisotropic and can reverse sign as the photon energy is tuned through flavor-wave resonances, establishing a route to electrically and optically tunable spin transport without spin–orbit coupling (Sicheler et al., 9 Aug 2025).

Ferroelectricity supplies nonvolatile control. In bilayer CuF$\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$6, interlayer sliding generates out-of-plane ferroelectric polarization $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$7 and $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$8, and this sliding ferroelectricity directly reverses the $\Delta(\mathbf{k}) \propto \cos k_x - \cos k_y$9-wave altermagnetic spin splitting. Because the splitting is layer locked, the ferroelectric states encode simultaneous changes in polarization, spin polarization, and layer localization; in quadrilayer CuF$s$0, four polarization states are identified for multistate logic (Peng et al., 11 Mar 2026).

A more general symmetry-engineering route sandwiches a conventional antiferromagnetic monolayer between two identical ferroelectric layers. In the In$s$1Se$s$2/MnPTe$s$3/In$s$4Se$s$5 trilayer, the ferroelectric polarization breaks inversion and induces altermagnetic splitting in the central layer, while global $s$6 symmetry ensures that reversing polarization exactly inverts the spin-splitting pattern. The anomalous Hall conductivity then flips sign under ferroelectric switching, providing an electrical fingerprint of the two altermagnetic states (Zhao et al., 30 Jun 2026).

Strain-mediated electric-field control has also been demonstrated near a magnetic phase transition. In MnTe/PMN-PT heterostructures, $s$7 induces piezoelectric strain that raises the Néel temperature from $s$8 to $s$9 K. Around the transition, the altermagnetic spin splitting is reversibly switched “on” and “off,” and the resistance modulation reaches $\mathbf{M}$0. The same electrically programmable resistance states were mapped onto a Hopfield network, which achieved $\mathbf{M}$1 pattern recognition accuracy at noise levels $\mathbf{M}$2 (Duan et al., 11 Dec 2025).

6. Experimental status, ambiguities, and future directions

The experimental basis of altermagnetic spintronics now includes ARPES observation of broken Kramers degeneracy, XMCD-PEEM imaging of domains and vortices, spontaneous anomalous Hall effects in compensated magnets, spin-charge conversion in thin films, and emerging tunnel-junction demonstrations. The review literature frames the field as one that combines ferromagnet-like spin-polarized transport with antiferromagnet-like speed, scalability, and field robustness, while also stressing outstanding challenges in materials control, interface quality, and unambiguous separation of non-relativistic and relativistic mechanisms (Jungwirth et al., 13 Aug 2025).

One major ambiguity concerns the distinction between spin splitting and directly usable spin polarization. In $\mathbf{M}$3-MnTe, high-resolution SARPES and first-principles calculations reveal a $\mathbf{M}$4-independent Rashba-like spin texture primarily governed by spin–orbit coupling, while the magnetic order mainly contributes to band splitting and a multi-domain structure suppresses the ideal $\mathbf{M}$5-wave spin polarization expected for altermagnetic spintronics. The work argues that materials with fewer symmetry-related antiferromagnetic domains should be favored when the goal is direct exploitation of momentum-space spin polarization (Zeng et al., 4 Nov 2025).

Atomic-scale MnTe studies sharpen this point further by showing that local inversion-breaking polar distortions, orthorhombicity, and coexistence with ferroelectric signatures are not secondary imperfections but active determinants of the spin-space-group class and the allowed form of spin splitting. This suggests that practical altermagnetic device design will depend not only on identifying the correct bulk magnetic order, but also on controlling local structural distortions, strain fields, domain topology, and interface-induced symmetry breaking (Ren et al., 26 May 2026).

A plausible implication is that the most robust altermagnetic technologies will be those that deliberately combine symmetry control, Fermi-surface engineering, and electrically addressable ferroic order. Current proposals already span giant-TMR antiferromagnetic tunnel junctions, all-electric spin filters and spin valves, spin-precession transistors, open-orbit spintronics, multiferroic switching, acoustic spin splitters, and superconducting hybrids. The remaining bottlenecks are the same ones repeatedly identified across the literature: deterministic domain control, room-temperature stability in thin films and heterostructures, quantitative disentanglement of SOC and Zeeman backgrounds from intrinsic altermagnetic responses, and integration of these symmetry-driven spin functionalities into scalable device stacks (Jungwirth et al., 13 Aug 2025).