Phonon Magnetic Moments

• Phonon magnetic moments are magnetic responses from chiral lattice vibrations, arising from both intrinsic ionic motion and electron–phonon coupling.
• The mechanisms involve circular ionic motion, emergent gauge fields, and spin–phonon interactions that lead to measurable Zeeman splitting and Berry curvature effects.
• Experimental studies in materials such as MoS2 and KTaO3 reveal how hybrid exciton–phonon dynamics and topological contributions amplify effective phonon magnetic moments.

Phonon magnetic moments are magnetic responses associated with lattice vibrations that carry angular momentum. In the narrowest sense, they arise from the orbital motion of charged ions in a circularly polarized phonon; in current condensed-matter literature, the term also encompasses larger effective moments generated through electron–phonon coupling, electronic orbital magnetization, Hall responses, exciton-assisted Raman processes, and phonon–magnon hybridization. This broader usage explains why “bare” ionic estimates are typically tiny, whereas effective moments extracted from Zeeman splittings or electron-mediated responses can reach the scale of the Bohr magneton in specific materials (Ren et al., 2021, Chen et al., 14 May 2025, Wang et al., 13 Mar 2025).

1. Definition and conceptual scope

A chiral phonon is a circularly polarized lattice vibration. For a degenerate optical mode with coordinates $\boldsymbol{q}=(q_1,q_2)$, circular motion implies a phonon angular momentum

$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$

which is a pseudovector, even under spatial inversion and odd under time reversal (Geilhufe et al., 2022). In an operational Zeeman description, the phonon magnetic moment is the coefficient of the field-induced energy shift,

$\hat H_Z=-\mathbf B\cdot \mathbf m,$

and, in a long-wavelength description, can be written as

$$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$

with $\lambda_{ij}$ determined by electronic response (Xue et al., 6 Jan 2025).

The classical ionic estimate follows immediately from circular motion of charge. For a particle of charge $e$, mass $m_{\mathrm{ion}}$, and angular momentum $\hbar$, one obtains

$$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$

so the moment is suppressed by $m_e/m_{\mathrm{ion}}$ relative to the Bohr magneton and is therefore typically of the order of the nuclear magneton, often $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$0 or smaller per ion or mode (Geilhufe et al., 2022). Related literature on nonmagnetic insulators gives intrinsic phonon magnetic moments of order $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$1, with relative phonon splittings $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$2 in accessible fields (Cheng et al., 2019).

Recent work uses several distinct but related notions. One is the direct ionic orbital moment of the lattice motion. A second is an electronic contribution, in which chiral lattice motion induces orbital or spin polarization of electrons, yielding a much larger magnetic response (Geilhufe et al., 2022). A third is an effective phonon magnetic moment extracted from the linear Zeeman splitting of a spectroscopic line, as in

$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$3

where the inferred $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$4 characterizes the magnetic response of a coupled phonon–electron or phonon–exciton system rather than only the ionic loop current (Tang et al., 2024).

A central consequence is that “phonon magnetic moment” is not a single universally fixed object. It may denote an intrinsic ionic quantity, an electronically renormalized orbital response, or an experimentally inferred effective moment. The literature consistently distinguishes these cases, even when the same terminology is used.

2. Microscopic mechanisms

The traditional framework treats the magnetic moment as the orbital moment of Born-effective-charge motion. That approach is refined by quantum-geometric theories in which lattice displacements $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$5 act as slow parameters of the Bloch Hamiltonian $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$6. In that setting, the phonon-induced electronic orbital magnetization is governed by a second Chern form in the combined $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$7 parameter space, and the conventional Born effective charge must be replaced by a $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$8-resolved Born effective charge $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$9 (Ren et al., 2021). The same work identifies an additional contribution involving the phonon-modified electronic energy and momentum-space Berry curvature, so the total magnetic response is not exhausted by the classical circulating-charge picture.

A distinct route is a spin-rotation-like coupling between phonon angular momentum and electronic angular momentum. In KTaO$\hat H_Z=-\mathbf B\cdot \mathbf m,$0, the symmetry-allowed interaction

$\hat H_Z=-\mathbf B\cdot \mathbf m,$1

is recast microscopically as an effective Hamiltonian $\hat H_Z=-\mathbf B\cdot \mathbf m,$2, where $\hat H_Z=-\mathbf B\cdot \mathbf m,$3 is the angular velocity associated with the chiral ionic motion (Geilhufe et al., 2022). Because $\hat H_Z=-\mathbf B\cdot \mathbf m,$4 and spin–orbit coupling on Ta $\hat H_Z=-\mathbf B\cdot \mathbf m,$5 states is strong, this term transiently splits spin-up and spin-down channels and produces a sizable electronic magnetization.

In doped Dirac materials, two geometric mechanisms have been isolated. One treats phonons as emergent gauge fields, leading to a magnetic moment proportional to the electronic Hall conductivity. The other treats phonons as frame fields, or emergent gravitational fields, producing a magnetic moment proportional to Hall viscosity (Chen et al., 14 May 2025). In the gauge-theory formulation, the induced Hall-viscosity term in the phonon Lagrangian,

$\hat H_Z=-\mathbf B\cdot \mathbf m,$6

splits left- and right-circular optical phonons and yields

$\hat H_Z=-\mathbf B\cdot \mathbf m,$7

so the phonon magnetic moment is directly proportional to Hall conductivity and equivalently to phonon Hall viscosity (Chen et al., 2024).

Beyond adiabatic theories, a non-Hermitian density-matrix formalism incorporating relaxation identifies extrinsic skew-scattering and side-jump contributions to the phonon magnetic moment (Xue et al., 6 Jan 2025). In that framework, the nonadiabatic term $\hat H_Z=-\mathbf B\cdot \mathbf m,$8 is controlled by a nonlocal molecular Berry curvature and vanishes only in the relaxation-adiabatic limit. This establishes a close analogy with extrinsic anomalous Hall physics in electronic transport.

The phonon magnetic moment also emerges in magnetic systems through hybridization with spin excitations. In ferrimagnetic Fe$\hat H_Z=-\mathbf B\cdot \mathbf m,$9Zn$$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$0Mo$$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$1O$$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$2, the effective moment is attributed to angular-momentum-selective phonon–magnon coupling, summarized by the proportionality

$$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$3

with $$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$4 the phonon–magnon coupling and $$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$5 the detuning (Wu et al., 18 Jan 2025). A related, though conceptually broader, extension is phonon-mediated spin–spin interaction: integrating out phonons produces long-range symmetric anisotropic exchange for generic phonons and antisymmetric Dzyaloshinskii–Moriya-like exchange when inversion symmetry is broken (Fransson, 16 Apr 2026). This does not define a phonon magnetic moment in the narrow dipolar sense, but it shows how phonons can act as carriers of effective magnetic chirality.

3. Symmetry, chirality, and momentum-space structure

Degeneracy is essential. A doubly degenerate pair of linear modes can be recombined into left- and right-handed circular states, often denoted $$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$6 or $$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$7, and these carry opposite angular momentum. In monolayer MoS$$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$8, the relevant zone-center $$m_i=\lambda_{ij}(\mathbf u^a\times \dot{\mathbf u}^a)_j,$$9 mode is explicitly described as a superposition of orthogonal LO and TO components, producing two degenerate chiral phonons $\lambda_{ij}$0 and $\lambda_{ij}$1 (Tang et al., 2024). In Fe$\lambda_{ij}$2Zn$\lambda_{ij}$3Mo$\lambda_{ij}$4O$\lambda_{ij}$5, Raman selection rules under $\lambda_{ij}$6 symmetry assign pseudo-angular momentum $\lambda_{ij}$7 to the chiral branches, and the long-wavelength limit identifies the sign of the pseudo-angular momentum with the sign of the actual angular momentum (Wu et al., 18 Jan 2025).

Symmetry determines whether this angular momentum can survive after summing all contributions. If combined $\lambda_{ij}$8 symmetry is present, the phonon angular momentum projection $\lambda_{ij}$9 must vanish, producing nonaxial phonons with zero net phonon magnetic moment (Wang et al., 8 Dec 2025). Once $e$0 is broken, magnetic point-group analysis yields several families of magneto-axial phonons, including even-wave and odd-wave patterns. The resulting momentum-space textures can display nodal structures analogous to $e$1-, $e$2-, $e$3-, up to $e$4-wave multipoles of phonon angular momentum (Wang et al., 8 Dec 2025).

This symmetry-based language clarifies a common source of confusion. A crystal may support local circular ionic motion while still exhibiting zero net phonon angular momentum or zero net magnetic moment because symmetry pairs opposite chiralities. Conversely, broken inversion or broken time-reversal symmetry can lift the degeneracy and generate observable Zeeman splittings or circular dichroism. In the phonon-mediated spin-interaction problem, the same distinction appears differently: symmetric anisotropic exchange survives generically, whereas antisymmetric Dzyaloshinskii–Moriya-like coupling requires inversion breaking because the relevant integrand must avoid cancellation under $e$5 (Fransson, 16 Apr 2026).

A further distinction is between “truly chiral” and “falsely chiral” phonons. In the Raman study of ferrimagnetic Fe$e$6Zn$e$7Mo$e$8O$e$9, backscattering gives a small but finite phonon wavevector along $m_{\mathrm{ion}}$0, and because the phonon angular momentum is also along $m_{\mathrm{ion}}$1, the observed propagating modes satisfy $m_{\mathrm{ion}}$2 and are classified as truly chiral (Wu et al., 18 Jan 2025). This definition is stricter than merely having circular atomic trajectories at $m_{\mathrm{ion}}$3.

4. Representative realizations and reported magnitudes

The recent literature spans several experimentally and theoretically distinct realizations. The following examples illustrate how the reported moment depends on mechanism, material class, and operational definition.

System	Mechanism or observable	Reported scale
KTaO$m_{\mathrm{ion}}$4	Electronic magnetization from $m_{\mathrm{ion}}$5 coupling	$m_{\mathrm{ion}}$6 per unit cell
Monolayer MoS$m_{\mathrm{ion}}$7	Exciton-activated $m_{\mathrm{ion}}$8 phonon Zeeman splitting	$m_{\mathrm{ion}}$9
Cd$\hbar$0As$\hbar$1	Cyclotron-coupled effective phonon magnetic moment	$\hbar$2
Gated bilayer graphene	Quantum topological plus perturbative electronic contribution	$\hbar$3
Fe$\hbar$4Zn$\hbar$5Mo$\hbar$6O$\hbar$7	Phonon–magnon hybridization	$\hbar$8 at 15 K; $\hbar$9 near $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$0

In KTaO$$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$1, a circularly polarized THz pulse excites the lowest infrared-active soft mode, and the resulting phonon angular momentum couples to Ta $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$2 electrons through $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$3. For light electron doping or elevated electronic temperature, the spin-resolved density of states becomes imbalanced, producing an electronic magnetic moment of $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$4 per unit cell, about three orders of magnitude larger than the purely ionic estimate in the same material (Geilhufe et al., 2022).

In monolayer MoS$$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$5, helicity-resolved magneto-Raman spectroscopy identifies a zone-center $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$6 mode near $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$7 that is activated only under resonant excitation of the A exciton. The measured Zeeman slopes are $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$8 and $$\mu_{\mathrm{phonon}}=\frac{e\hbar}{2m_{\mathrm{ion}}},$$9, corresponding to $m_e/m_{\mathrm{ion}}$0 and $m_e/m_{\mathrm{ion}}$1, respectively (Tang et al., 2024). The work explicitly interprets this as an effective phonon magnetic moment of the coupled exciton–phonon system.

In Cd$m_e/m_{\mathrm{ion}}$2As$m_e/m_{\mathrm{ion}}$3, magnetoterahertz spectroscopy reveals a low-frequency $m_e/m_{\mathrm{ion}}$4 optical phonon near $m_e/m_{\mathrm{ion}}$5 that hybridizes selectively with cyclotron resonance. The splitting between right- and left-circular channels reaches $m_e/m_{\mathrm{ion}}$6, and the effective phonon magnetic moment extracted from the low-field slope is $m_e/m_{\mathrm{ion}}$7, almost four orders of magnitude larger than ab initio predictions for intrinsic ionic moments in nonmagnetic insulators (Cheng et al., 2019).

In gated bilayer graphene, the classical Born-effective-charge theory predicts an almost vanishing moment because the carbon atoms have equal masses and the electronic and nuclear classical contributions cancel. A fully quantum calculation instead finds a chiral shear-mode moment $m_e/m_{\mathrm{ion}}$8 at $m_e/m_{\mathrm{ion}}$9, with the perturbative electronic contribution dominating the total (Zhang et al., 2023). The moment is highly gate-tunable because the perturbative term decreases approximately as the inverse square of the gap.

In ferrimagnetic Fe$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$00Zn$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$01Mo$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$02O$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$03, polarization-resolved magneto-Raman spectroscopy shows a pair of chiral phonons split by about $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$04 of the phonon frequency due to spontaneously broken time-reversal symmetry. The P1a branch exhibits an effective phonon magnetic moment of $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$05 at $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$06, and the low-field slope near the critical temperature yields $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$07, exceeding the moment of a magnon according to the authors’ comparison (Wu et al., 18 Jan 2025).

5. Topological and many-body extensions

Topological formulations place phonon magnetic moments within the same response hierarchy as Hall conductivity, Hall viscosity, Berry curvature, and Chern numbers. In doped Dirac systems, the phonon magnetic moment can be written as a Hall-conductivity response induced by an emergent gauge field, or as a Hall-viscosity response induced by an emergent frame field (Chen et al., 2024, Chen et al., 14 May 2025). This makes phonon dynamics a potential probe of electronic Hall viscosity: in the frame-field channel, the circular splitting

$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$08

directly measures $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$09, and the corresponding

$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$10

becomes field-independent in the weak-field regime because $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$11 (Chen et al., 14 May 2025).

Magnetic materials introduce another route to topological phonon bands. A linear-response ab initio theory of EPC-induced TR-breaking phonon self-energies finds large phonon Zeeman splittings in magnetic metals and shows that these gaps can carry nonzero phonon Chern numbers (Wang et al., 13 Mar 2025). In the representative material EuSi$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$12, the EPC-induced Zeeman splitting is reported as $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$13, the linewidth as $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$14, and the observable phonon gap as $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$15 (Wang et al., 13 Mar 2025). The same work identifies GdSi$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$16, EuGa$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$17, GdGa$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$18, and CrB$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$19 as candidate topological phonon magnets.

Thermally generated phonon angular momentum can also be amplified by many-body feedback within the phonon sector itself. In a theory motivated by the orbital Seebeck effect in $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$20-quartz, a Heisenberg-type long-range L–L coupling between phonon angular momenta yields a self-consistent response

$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$21

so the amplification factor is $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$22 (Sun et al., 26 Mar 2026). As the threshold $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$23 is approached, the total phonon angular momentum is enhanced by up to nearly two orders of magnitude, and the corresponding phonon magnetic response in $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$24-quartz is estimated as $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$25 or a few $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$26 (Sun et al., 26 Mar 2026). This suggests that giant effective phonon magnetic moments need not be purely electronic; collective phononic feedback can also amplify a small bare moment.

The many-body perspective extends further to insulating and semiconducting magnets. Phonon-mediated spin–spin interactions in such systems decay as $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$27 for acoustic phonons, with oscillatory factors and a temperature dependence set by $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$28, becoming nearly temperature-independent at low temperature and approximately linear in $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$29 at high temperature (Fransson, 16 Apr 2026). Although this mechanism concerns effective exchange rather than a dipolar phonon moment, it reinforces a general point: phonons can transmit angular momentum and magnetic anisotropy over unexpectedly long distances.

6. Experimental signatures, misconceptions, and open issues

The dominant experimental signatures are phonon Zeeman splitting, helicity-dependent Raman or THz response, Kerr or Faraday rotation induced by transient magnetization, and circular or linear phonon dichroism. In monolayer MoS$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$30, the key evidence is helicity-resolved Raman splitting of the $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$31 mode under out-of-plane magnetic field (Tang et al., 2024). In Cd$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$32As$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$33, it is the field-dependent difference between right- and left-circular magnetoterahertz transmission near the cyclotron–phonon resonance (Cheng et al., 2019). In KTaO$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$34, the predicted signal is a transient electronic magnetization following the chiral soft mode under a resonant THz drive (Geilhufe et al., 2022).

A frequent misconception is that a reported “phonon magnetic moment” always measures the orbital loop current of the ions alone. The literature does not support that identification. In KTaO$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$35, the dominant moment is an electronic magnetization generated by phonon-driven spin splitting (Geilhufe et al., 2022). In MoS$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$36, the measured value is explicitly an effective moment inferred from Zeeman splitting of an exciton-activated phonon Raman line (Tang et al., 2024). In Cd$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$37As$$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$38, the giant effective moment is controlled by electron–phonon coupling to cyclotron motion (Cheng et al., 2019). These examples show that effective phonon moments can encode hybrid electron–phonon dynamics rather than only ionic motion.

A second misconception is that giant values necessarily contradict the classical theory. The more precise statement is that they exceed the classical ionic estimate because additional channels become active: electronic topological magnetization, Hall-conductivity and Hall-viscosity responses, nonadiabatic skew-scattering or side-jump terms, exciton dressing, or phonon–magnon hybridization (Ren et al., 2021, Xue et al., 6 Jan 2025). The discrepancy is therefore diagnostic of missing physics, not evidence that phonon magnetism is ill-defined.

Several limitations remain standard across the field. Semiclassical phonon treatments are appropriate for strong coherent driving but do not capture quantum fluctuations or multiphonon processes. Adiabatic electronic treatments can fail near resonances or in gapless systems unless relaxation is included explicitly (Geilhufe et al., 2022, Xue et al., 6 Jan 2025). Effective couplings such as $$\boldsymbol{L}^{\mathrm{phonon}}=\boldsymbol{q}\times\dot{\boldsymbol{q}},$$39 or emergent-gauge-field mappings are symmetry-grounded and often quantitatively useful, but they are still reduced descriptions of a more complicated electron–phonon problem. In the topological setting, mode linewidths and disorder can compete with the Zeeman gap, so the relevant experimental criterion is not only a large splitting but a splitting that exceeds the broadening (Wang et al., 13 Mar 2025).

The overall trajectory of the subject is clear. The direct ionic picture remains the reference point, but the modern theory of phonon magnetic moments is increasingly a theory of how chiral lattice dynamics couple to quantum geometry, Hall responses, spin polarization, and collective magnetic order. This suggests that the most favorable platforms are not merely crystals with circular ionic motion, but materials combining suitable phonon degeneracies with strong spin–orbit coupling, small electronic gaps or high carrier mobility, broken inversion or broken time-reversal symmetry, and, in some cases, strong magnetic or excitonic correlations.