Chiral Phonon Edge Modes

• Chiral phonon edge modes are boundary-confined vibrational excitations that propagate in a fixed direction due to broken symmetries or nonzero topological invariants.
• They manifest as strict Chern edge states in two-dimensional systems, surface arcs in Weyl materials, or hybrid modes coupled with magnons and valley channels.
• Experimental diagnostics via spectroscopy, thermal transport, and engineered boundary conditions validate their existence and potential for controlled directional energy flow.

Chiral phonon edge modes are boundary-confined vibrational excitations whose propagation acquires a definite handedness through broken symmetries, Berry-curvature structure, or boundary-induced mode selection. In the strict topological sense, they are phonon boundary states mandated by bulk–boundary correspondence after a phonon band acquires a nonzero Chern number or is connected to Weyl charges in three dimensions. In broader contemporary usage, the same phrase can also denote valley-locked boundary channels, phonon-dominated hybrid edge modes, or bosonized edge collective modes whose transport is phononic or entropy-wave-like. The common thread is directional boundary propagation tied to a bulk or boundary mechanism that suppresses the opposite-moving partner on the same edge (Ji et al., 2016).

1. Definition, terminology, and scope

The term “chiral” is used in at least two distinct but related senses. One sense is propagation chirality: an edge or surface mode has a definite sign of group velocity along a boundary segment or closed loop in the surface Brillouin zone. This is the sense relevant to Chern phonons in a two-dimensional Wigner crystal, to surface phonon arcs in Weyl systems such as EuPtSi and NbSi$_2$, and to phonon-dominated edge magneto-polarons. Another sense is polarization chirality: the atomic displacement field executes circular or elliptical motion with nonzero phonon angular momentum $L_z$, as in the square–octagon spring–mass model with an explicit TRS-breaking gyroscopic term. These two senses frequently coexist but are not identical (Mahraj et al., 2024).

The term “edge mode” is likewise non-uniform. In the strongest usage it denotes a mode exponentially localized at a physical boundary and traversing a bulk spectral gap. In weaker or analogical usages it can refer to modes whose existence is boundary-induced but whose spatial profile remains bulk-like, as in boundary induced chiral anomaly bulk states in a trivial phononic crystal slab, or to collective bosonic channels on the edge of an electronic fluid, as in chiral second-sound and BF-theory edge bosons. This broader terminology is useful, but it makes conceptual distinctions essential: topological protection, edge localization, and phononic character need not coincide in every construction (Wang et al., 2022).

2. Chern phonon edge modes in a two-dimensional Wigner crystal

A concrete microscopic realization of chiral phonon edge modes was proposed for a two-dimensional Wigner crystal of electrons on a triangular lattice. The low-energy degrees of freedom are the lattice displacements $\mathbf{u}(\mathbf{l})$, their conjugate momenta $\hat{\mathbf P}(\mathbf l)$, and spins that are treated in mean field because ferromagnetism and Zeeman splitting freeze spin fluctuations. Spin–orbit coupling enters the kinetic momentum through the electrostatic force generated by the Coulomb crystal itself, so that phonon dynamics acquires an effective TR-breaking antisymmetric contribution. After Fourier transformation, the phonon problem can be written as

$$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$

with

$$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$

The low-energy mass term near $K$ is

$$M=\omega G(\mathbf K)=\omega\left(\omega_c-2\alpha_0\sigma\omega_0^2\right),$$

so a topological transition occurs when $M$ changes sign, approximately at

$$2\alpha_0\omega_0=-\mathrm{sgn}(g^\ast)\frac{\omega_c}{\omega_0}.$$

This immediately imposes the condition $L_z$0 for the transition (Ji et al., 2016).

Topology is diagnosed through a phonon Berry curvature defined with the modified bosonic metric and a Chern number

$L_z$1

Numerically, the upper phonon band changes from $L_z$2 to $L_z$3 after the $L_z$4 band inversion. In a strip geometry with $L_z$5, periodic boundary conditions along $L_z$6, and open boundaries along $L_z$7, the full-gap topological phase exhibits two edge branches on each edge, both with the same sign of slope on a given edge. These are chiral phonon edge modes in the strict Chern-insulator sense: they connect the lower and upper bulk phonon bands, are localized within a few lattice constants of the boundary, and their multiplicity matches the bulk Chern number. The same study also emphasized an important negative result: in ordinary semiconductor Wigner crystals the native spin–orbit coupling is far too weak to drive an observable topological phonon transition, motivating the search for correlated Wigner crystals with emergent effective spin–orbit coupling, particularly composite-fermion Wigner crystals (Ji et al., 2016).

3. Surface chiral phonon modes in Weyl and multifold phonon systems

Three-dimensional chiral crystals realize the same idea in a surface rather than strip geometry. In EuPtSi, which crystallizes in space group $L_z$8, the phonon spectrum contains a spin-1 Weyl point at $L_z$9 and a charge-2 Dirac point at $\mathbf{u}(\mathbf{l})$0, each with Chern number $\mathbf{u}(\mathbf{l})$1. On the (001) surface, $\mathbf{u}(\mathbf{l})$2 and $\mathbf{u}(\mathbf{l})$3 project to $\mathbf{u}(\mathbf{l})$4 and $\mathbf{u}(\mathbf{l})$5, and the surface phonon spectral function shows Fermi-arc-like phonon modes connecting these projections. Constant-frequency cuts at approximately $\mathbf{u}(\mathbf{l})$6 THz, $\mathbf{u}(\mathbf{l})$7 THz, and $\mathbf{u}(\mathbf{l})$8 THz reveal arc connectivity, while loop spectra around $\mathbf{u}(\mathbf{l})$9 and $\hat{\mathbf P}(\mathbf l)$0 show unidirectional surface propagation. In this setting, chirality is encoded in the monotonic winding of surface branches around projected topological nodes; the number of arcs equals $\hat{\mathbf P}(\mathbf l)$1, and the sign of the Chern number fixes the propagation sense (Mahraj et al., 2024).

NbSi$\hat{\mathbf P}(\mathbf l)$2, in the chiral noncentrosymmetric space group $\hat{\mathbf P}(\mathbf l)$3, provides a related but more intricate example. The relevant Weyl physics occurs in the crossing of the 25th and 26th phonon branches, including a multifold Weyl point at $\hat{\mathbf P}(\mathbf l)$4 with $\hat{\mathbf P}(\mathbf l)$5, several $\hat{\mathbf P}(\mathbf l)$6 Weyl points near $\hat{\mathbf P}(\mathbf l)$7, and, notably, three $\hat{\mathbf P}(\mathbf l)$8 Weyl points arranged around $\hat{\mathbf P}(\mathbf l)$9. On the (001) surface, the spectral function at $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$0 THz displays unconventional phononic Fermi arcs: the projected Weyl-point connectivity near $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$1 and $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$2 mimics an effective $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$3, even though the bulk charge is distributed among several nodes at different $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$4. The resulting surface modes are called chiral edge modes because their dispersion along closed loops in the surface Brillouin zone is unidirectional, not because the paper explicitly analyzes real-space circular polarization (Mahraj et al., 18 Jul 2025).

These three-dimensional systems clarify a central point. Chiral phonon edge modes need not be restricted to two-dimensional Chern bands; they can equally emerge as surface phonon states required by Weyl or multifold topological charges. In that case the natural signatures are surface arcs, loop chirality, and localization within the outermost atomic layers rather than a single strip-edge dispersion.

4. Boundary-induced and hybrid routes

Not every directional boundary phonon channel is a Chern edge state. A distinct route was experimentally demonstrated in a topologically trivial hexagonal phononic-crystal waveguide bounded by hard walls aligned with mirror planes. The bulk crystal supports even and odd valley bands crossing at $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$5, but the hard-wall boundary condition enforces zero normal velocity and therefore filters out the odd-parity sector. The projected strip spectrum then develops a pseudogap containing a single surviving valley-locked branch. These boundary induced chiral anomaly bulk states are robust to moderate disorder and bends, and they support experimentally observed valley-selected transport, complete valley state conversion, and valley focusing. Yet they are explicitly bulk-like and width-independent rather than edge-localized, so they are better regarded as boundary-engineered analogues of chiral edge transport than as topological edge modes in the strict sense (Wang et al., 2022).

A second route uses topological magnons to endow phonons with chirality. In a honeycomb ferromagnet with next-nearest-neighbor Dzyaloshinskii–Moriya coupling, topological magnon edge states hybridize with Rayleigh-like armchair edge phonons through magnetoelastic coupling. Near the avoided crossing, a phonon-dominated edge magneto-polaron inherits the unidirectionality of the magnon sector. The resulting coherent phonon response is quantified by

$$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$6

where $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$7 and $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$8 are elastic powers detected by right and left transducers. Here the chiral phonon edge mode is not a bare phonon band carrying its own primary topology; it is a hybrid edge excitation whose phononic weight remains dominant in the relevant frequency window (Thingstad et al., 2018).

A third route arises from strong coupling between electrons and a single chiral valley phonon in a honeycomb lattice with broken inversion symmetry. The phonon carries momentum $$\hat h(\mathbf{k})\mathbf{u}(\mathbf{k})=\omega^2\mathbf{u}(\mathbf{k}),\qquad \hat h(\mathbf{k})=\Phi(\mathbf{k})+\omega G(\mathbf{k})\sigma_2,$$9 and produces inelastic Umklapp scattering between valleys. In a non-perturbative Fock-space treatment, this generates a valley-selective pseudogap near $$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$0, bridged by hybrid electron–phonon edge states. These states are copropagating on opposite edges and coexist with a bulk continuum from ungapped valleys. Their existence therefore differs both from ordinary quantum-Hall edge states and from purely phononic Chern modes: they are dynamical hybrid boundary channels produced by chiral-phonon-mediated replica coupling (Dueñas et al., 2021).

5. Transport, spectroscopy, and experimental diagnostics

The observable content of chiral phonon edge modes depends strongly on the realization. In acoustic phononic waveguides, direct field mapping and Fourier-space reconstruction have already resolved the boundary-engineered valley-locked modes. In the hexagonal phononic-crystal slab, measured spectra in the $$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$1–$$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$2 kHz window showed only $$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$3-valley intensity in the straight-waveguide geometry, while bends and locally modified boundaries implemented either transmission blocking, $$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$4 forwarding, or complete $$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$5 valley conversion. A sharply tapered geometry concentrated intensity into a narrow strip, producing valley focusing. These observations established that directional boundary transport can be engineered even in a bulk-topologically trivial medium when mirror parity and hard-wall constraints are used as the control knob (Wang et al., 2022).

In systems where the modes are genuinely edge-localized, the natural observables are thermal and spectroscopic. The Wigner-crystal proposal identified phonon thermal Hall response and edge-sensitive probes of vibrational spectra as possible signatures, although it simultaneously concluded that the estimated gaps and critical fields in ordinary semiconductors are many orders of magnitude too small for clean observation. The magnetoelastic honeycomb proposal instead suggested an elastic-transducer protocol: interdigital transducers on an armchair edge selectively excite Rayleigh-like phonons at the wave vector where they anticross a chiral magnon edge mode, producing a large asymmetry in left- and right-going elastic power (Ji et al., 2016).

A related but conceptually broader heat-transport realization appears in Berry-curved electronic fluids. In gapped undoped graphene on hBN, chiral second-sound modes propagate along the edge with valley-selective directionality fixed by the sign of the Berry curvature. Their localization length follows from cancellation of normal and anomalous heat currents at the boundary, and at low temperature the single-particle transverse thermal conductance is exponentially suppressed while the edge second-sound channel contributes

$$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$6

These are not lattice phonons, but they are phonon-like neutral entropy waves and therefore occupy an important adjacent position in the broader literature on chiral bosonic edge transport (Principi et al., 2017).

Bulk chirality diagnostics can also matter even when edge spectra remain uncomputed. In the square–octagon lattice, a TRS-breaking term of the form $$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$7 generates bulk chiral phonons with nonzero angular momentum

$$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$8

The associated infrared circular dichroism has the derivative-like form

$$G(\mathbf{k})=\omega_c-\alpha_0\sigma\left[\Phi_{xx}(\mathbf{k})+\Phi_{yy}(\mathbf{k})\right].$$9

Because that work did not compute Berry curvature or edge spectra, it does not demonstrate chiral phonon edge modes; however, it shows how bulk chirality and optical helicity selection can be diagnosed in parameter regimes that may later support topological edge states (Kiran et al., 29 Oct 2025).

6. Conceptual boundaries, constraints, and open problems

Several recurring misconceptions are resolved by the literature itself. First, chiral does not always mean the same thing. In Wigner crystals, EuPtSi, and NbSi$K$0, it refers primarily to boundary propagation and bulk topological charge. In the square–octagon model it refers to circular or elliptical polarization with $K$1. In trivial phononic-crystal waveguides it can denote valley-locked transport without edge localization. Second, robustness is mechanism-dependent. Chern and Weyl phonon boundary states inherit bulk–boundary correspondence, whereas boundary induced chiral anomaly bulk states rely on parity filtering and suppressed intervalley scattering rather than on a bulk topological invariant (Wang et al., 2022).

Third, the word phonon itself spans literal lattice vibrations, electron-crystal oscillations in a Wigner crystal, phonon-dominated magneto-polarons, and entropy-wave analogues such as second sound. A still more abstract extension appears in the abelian BF description of topological edges, where the boundary theory reduces to two scalar fields that decouple into Luttinger actions for chiral bosons. In that framework the edge velocities become local functions of the induced boundary metric, so the edge excitations are “accelerated”; positivity of the Hamiltonian rules out the case where two edge modes propagate in the same direction, while time-reversal symmetry uniquely selects equal and opposite velocities, the helical case appropriate to topological insulators (Bertolini et al., 2022).

The main open technical divide is between bulk chirality diagnostics and fully established edge topology. The square–octagon study supplies a detailed TRS-breaking dynamical matrix, tunable gaps, flat-band anomalies, anisotropic sound propagation, and infrared circular dichroism, but it leaves Berry curvature, Chern numbers, and ribbon spectra for future work. The Wigner-crystal study supplies a complete topological mechanism and edge calculation, but its semiconductor realization is impractical without much stronger effective spin–orbit coupling. The boundary-engineered phononic-crystal work demonstrates transport phenomena experimentally, but precisely by showing that edge-like chirality can arise in a topologically trivial bulk. Taken together, these results suggest that the field is no longer organized by a single paradigm. “Chiral phonon edge modes” now names a family of boundary phenomena ranging from strict Chern and Weyl phonon states to symmetry-filtered, hybrid, and bosonized analogues, with the decisive questions always being what is localized, what is topological, and what quantity actually carries the chiral current (Ji et al., 2016).