Alterferrimagnetism: Compensated Magnetic Phases

• Alterferrimagnetism is a symmetry‐protected phase characterized by compensated magnetic sublattices and momentum‐dependent, alternating spin splitting.
• It utilizes spin-space group formalism to link crystal symmetry with band structure, revealing distinct split regions and nodal planes.
• Experimental signatures such as ARPES-detected spin splitting and anisotropic transport highlight its potential for advanced spintronic applications.

Alterferrimagnetism is a generalization of altermagnetism that incorporates features of ferrimagnetism, extending the notion of momentum-dependent, symmetry-protected spin splitting to systems with multiple magnetic species and compensated or weakly uncompensated magnetic sublattices. It fundamentally connects crystallographic and spin-space group symmetry to the existence of novel magnetic phases, with distinct signatures in the band structure and transport properties that are not present in conventional antiferromagnets or ferrimagnets. This class of phases is defined both by its unique group-theoretical conditions and by its broad phenomenology, which includes alternating spin polarization, tunable coercivity, and distinct responses to relativistic and orbital effects (Barman et al., 29 Dec 2025, Cheong et al., 20 Mar 2025, Ali et al., 30 Jan 2026, Mineev, 21 Jan 2026).

1. Definition and Theoretical Foundation

Alterferrimagnetism (AFiM) refers to a fully compensated or weakly uncompensated magnetic phase in which two or more distinct magnetic species form collinear sublattices, each with vanishing net spin moment, but whose overall configuration yields momentum-dependent, anisotropic spin splitting in the electronic structure. In the spin-space group (SSG) classification, AFiMs are realized as Type-III groups where the symmetry operation exchanging spin sublattices involves real-space rotation or reflection (not inversion or translation), and magnetic species occupy structurally distinct Wyckoff positions compatible with altermagnetism.

Collinear altermagnets (AMs) are characterized by zero net magnetization and by $\pi$-spin-degenerate, momentum-dependent splitting protected by spatial symmetries other than inversion or translation. Ferrimagnets (FiMs) typically exhibit net moments due to unequal opposing sublattices, resulting in momentum-independent splitting. In contrast, AFiMs consist of multiple compensated sublattices (per species), but their combined spin group symmetry leads to alternating, momentum-dependent spin splitting and vanishing or nearly vanishing net magnetization (Barman et al., 29 Dec 2025).

The formal group-theoretical condition for the existence of an altermagnetic or alterferrimagnetic phase is encapsulated in the Fundamental Lemma of Altermagnetism (FLAM): a Wyckoff position is AM-compatible if and only if (i) its multiplicity is even and (ii) its site-symmetry group $W$ is a subgroup of the halving subgroup $H$ (of index 2) in a parent crystallographic space group $G$ (Barman et al., 29 Dec 2025).

2. Spin-Space Group Formalism and Band Structure

The full classification of AFiMs involves a direct product structure on the spin group:

$\bar{G} = \{\mathbb{1}, \mathcal{C}_2\} \otimes G'$

where $G' = \cap_{i} H_i$ is the intersection of all halving subgroups $H_i$ corresponding to the respective Wyckoff positions of each magnetic species. The nontrivial operation $[\mathcal{C}_2 \| R]$ exchanges spin sublattices via real-space symmetry $R \in G - H$, producing momentum-dependent spin splitting according to:

$$[\mathcal{C}_2 \| R]\ \varepsilon_j(\sigma,\mathbf{k}) = \varepsilon_j(-\sigma, R\mathbf{k})$$

The resulting spin polarization $$P_s(\mathbf{k}) = [n_\uparrow(\mathbf{k}) - n_\downarrow(\mathbf{k})]/[n_\uparrow(\mathbf{k}) + n_\downarrow(\mathbf{k})]$$ alternates sign on $R$-related directions, giving rise to nodal planes where spin degeneracy is restored and split regions elsewhere. First-principles calculations demonstrate splittings up to $\sim170\,\mathrm{meV}$ in representative compounds (Barman et al., 29 Dec 2025).

The phenomenological band model for alterferrimagnets incorporates distinct sublattice moments $m_1$, $m_2$ (with $|m_1|\neq|m_2|$), yielding

$$H(\mathbf{k}) = \varepsilon_0(\mathbf{k})\,I + [\gamma_1 \sin k_x a\,\sin k_y a + \Delta](\sigma_x + \sigma_y) + \gamma_2 [\sin k_x a + \sin k_y a] \sin k_z c\,\sigma_z$$

with $\Delta\propto (m_1 - m_2)$. The Berry curvature $\Omega_z^\pm(\mathbf{k})$ vanishes by symmetry, ensuring the absence of a zero-field anomalous Hall effect (AHE) despite finite momentum-dependent splitting, in contrast to weak ferromagnetic states (Mineev, 21 Jan 2026).

3. Mechanisms for Weak Magnetization: DMI and $g$-Tensor Anisotropy

Alterferrimagnetism supports multiple microscopic origins for a weak net moment:

• In structures where the Dzyaloshinskii–Moriya interaction (DMI) is symmetry-allowed (specifically, where nearest-neighbor magnetic ions are related by rotation or roto-translation rather than inversion), staggered DMI can cant otherwise collinear sublattices, inducing a weak ferromagnetic (WFM) moment perpendicular to the Néel vector in centrosymmetric crystals and a weak ferrimagnetic moment parallel to the Néel vector in noncentrosymmetric crystals. In the latter case, distinct canting angles for inequivalent sites yield $M_\parallel \propto D^2/J$ and $M_\perp \propto D/J$ (Autieri et al., 2023).
• When DMI is symmetry-forbidden, alternating $g$-tensor anisotropy provides a purely orbital mechanism. Site-dependent $g$-tensors, arising from local crystal field environments and spin-orbit coupling, yield a spontaneous orbital moment:

$$\mathbf m_\mathrm{tot} = -\frac{\mu_B}{\hbar} \sum_i \mathbf{g}^i \cdot \langle \mathbf{S}^i \rangle$$

The net orbital magnetization grows linearly in the spin–orbit coupling and crystal-field splitting, and can be manipulated by strain, chemical substitution, or stacking—providing direct links to the anomalous Hall effect and orbitronic control (Jo et al., 2024, Cheong et al., 20 Mar 2025).

4. Thermodynamics, Compensation, and Extended Néel Diagram

Alterferrimagnets naturally arise at a critical point in the extended Néel diagram, where the sublattice moment imbalance and exchange ratios are precisely tuned such that complete compensation of net magnetization ($M_\mathrm{tot}(T) = 0$) is maintained up to the ordering temperature $T_C$. The two key conditions at the critical point are:

• $N_A \mu_A = N_B \mu_B$ (zero net moment at $T=0$)
• $2(J_{AA} - J_{BB})/(-J_{AB}) = N_A - N_B$ (zero net moment in the linear response limit near $T_C$)

Materials such as Gd(Co$_5$–$_x$Ni$_{0.5x}$Fe$_{0.5x}$) exemplify this regime, showing multi-eV spin splittings in DFT calculations and enhancement of coercive fields as $M_\mathrm{tot}$ approaches zero, with the coercivity diverging in a Stoner-Wohlfarth model and remaining large over an extended $T$-range (Ali et al., 30 Jan 2026).

Phase Type	Net Magnetization	Band Splitting
Altermagnet (AM)	0	$k$-dependent
Ferrimagnet (FiM)	$\neq 0$	$k$-independent (isotropic)
Alterferrimagnet	0 or $\ll 1$	$k$-dependent (alternating)

5. Magnetic Point Group Classification and Broken Symmetries

According to the SAM classification, alterferrimagnets (M-type altermagnets) belong to one of the 31 ferromagnetic magnetic point groups that break time-reversal ($T$), break parity–time ($PT$), and preserve space inversion ($P$). This constraint guarantees the persistence of a net orbital magnetization despite full spin compensation, and sets the conditions under which linear AHE, even-order kinetomagnetism, and other responses are symmetry-allowed (Cheong et al., 20 Mar 2025).

6. Experimental Manifestations and Materials Design

Key experimental signatures of alterferrimagnetism include:

• ARPES detection of spin band splitting that alternates under real-space symmetry operations and vanishes along nodal planes prescribed by the SSG (Barman et al., 29 Dec 2025).
• Momentum-dependent anisotropic transport: Magnetotransport experiments on Ti-doped hematite reveal sixfold angular dependence of longitudinal and Hall conductivities, with sign changes every 60°, matching the symmetry of the underlying spin texture and thus confirming altermagnetic order in a hopping-conductivity regime (Galindez-Ruales et al., 2023).
• Enhanced coercivity and spintronic potential: Compounds tuned to the extended compensation critical point exhibit both altermagnetic-like dynamics and robust coercivity, beneficial for ultrafast and permanent-magnet applications (Ali et al., 30 Jan 2026).

Materials realization strategies involve tailoring Wyckoff positions (satisfying $W \subseteq H$ and even multiplicity), tuning sublattice local moments by adjusting chemical composition, and manipulating exchange constants to achieve persistent compensation (Barman et al., 29 Dec 2025, Ali et al., 30 Jan 2026).

7. Hall Effect, Kinetomagnetism, and Distinction from Other Phases

Unlike weak ferromagnetic altermagnets, alterferrimagnetism does not generically support an intrinsic anomalous Hall effect in zero field—the combined action of mirror and rotation symmetries enforces vanishing Berry curvature despite bulk magnetization. However, the presence of spontaneous orbital moments permits a linear AHE proportional to $M^\mathrm{orb}$ under appropriate symmetry conditions (Mineev, 21 Jan 2026, Cheong et al., 20 Mar 2025). Kinetomagnetism—the current-induced generation of magnetization—emerges in all altermagnetic classes upon breaking $PT$, with M-type alterferrimagnets distinguished by their spontaneous zeroth-order orbital moment.

---

Alterferrimagnetism thus represents a symmetry-protected, multi-sublattice magnetic state that unifies the band-structure anisotropy of altermagnets with the sublattice complexity of ferrimagnets. Its emergence is dictated by crystallographic–group-theoretical criteria, and its experimental detection exploits the unique interplay of symmetry, spin–orbit coupling, and orbital character—leading to new classes of magnetic materials with applications ranging from anisotropic spintronics to orbitronics (Barman et al., 29 Dec 2025, Cheong et al., 20 Mar 2025, Ali et al., 30 Jan 2026, Jo et al., 2024, Mineev, 21 Jan 2026, Galindez-Ruales et al., 2023, Autieri et al., 2023).