Chiral Phonon-Polaritons

• Chiral phonon-polaritons are hybrid light–matter excitations defined by circular lattice motions and polarization-selective photon coupling.
• They enable the investigation of magneto-optical effects, topological edge modes, and non-Hermitian band phenomena across diverse material platforms.
• Experimental studies in LiNbO3, hBN, and quartz demonstrate their potential for ultrafast spectroscopy, spin transport, and dynamic symmetry breaking.

Searching arXiv for papers on chiral phonon-polaritons and closely related chiral phonon work. Chiral phonon-polaritons are hybrid light–matter excitations in which the phononic sector carries a definite handedness and the electromagnetic sector either probes, inherits, or selectively amplifies that handedness. In one experimentally realized form, they are propagating phonon-polariton wave packets whose ions execute circular or elliptical motion in the plane perpendicular to propagation, so that the mode has a definite left- or right-handed character; in another, an achiral phonon-polariton medium acquires handedness through strong coupling to a magneto-optical photonic mode with circularly polarized far-field radiation (Biggs et al., 29 Jul 2025, Sun et al., 14 Apr 2026). The subject therefore sits at the intersection of phonon angular momentum, optical activity, magneto-optics, and strong-coupling polariton physics, and its modern formulation depends on a careful distinction between angular momentum, helicity, and pseudo-angular momentum, as well as on symmetry constraints that determine which phonons can legitimately be called chiral (Zhang et al., 28 Mar 2025, Wang et al., 28 Apr 2026).

1. Definition and conceptual scope

A phonon-polariton in a polar dielectric is the mixed excitation obtained when an infrared-active optical phonon couples resonantly to the electromagnetic field. In LiNbO$_3$, for example, the transverse optical E(TO$_1$) phonons couple strongly to THz radiation, and the polariton dispersion follows

$\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$

with the dielectric function written in Lorentz form (Biggs et al., 29 Jul 2025). When the two degenerate in-plane E-mode coordinates $Q_x$ and $Q_y$ are driven with equal amplitude and a quarter-cycle phase offset,

$$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$

the ions execute circular trajectories and the resulting propagating phonon-polariton is chiral in the dynamical sense used in that work (Biggs et al., 29 Jul 2025).

Current usage is not exhausted by that case. In the hBN-based hybrid platform, the phonon resonance itself is not intrinsically chiral; instead, an hBN phonon-polariton couples to a chiral bound state in the continuum of a magneto-optical photonic crystal, and the hybrid state inherits left- or right-circularly polarized photonic character under magnetic bias (Sun et al., 14 Apr 2026). This broader usage makes “chiral phonon-polariton” a family resemblance term: some realizations are chiral because the lattice motion itself is circular and carries angular momentum, while others are chiral because the hybrid mode is handedness-selective under circularly polarized excitation.

A stricter criterion has been formulated for the phonon constituent. “Truly chiral phonons” are defined as lattice eigenmodes that combine broken mirror symmetry with circular atomic motion; conversely, identifying chirality solely through nonzero pseudo-angular momentum or through circular-polarization inversion in Raman scattering is inadequate (Wang et al., 28 Apr 2026, Zhang et al., 28 Mar 2025). For chiral phonon-polaritons, this distinction matters because the photonic component can display handedness even when the lattice part is only mirror-odd, and because true lattice angular momentum is the quantity that couples most directly to spin, magnetization, and dynamical multiferroicity.

2. Symmetry, angular momentum, and helicity

The basic phonon-angular-momentum structure can be expressed either through the mode expectation value of an angular-momentum operator built from ionic displacements and velocities or through a circular-polarization basis for each atom. A symmetry-based high-throughput formulation defines the phonon angular-momentum operator at momentum $\mathbf{q}$ as

$$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$

with components

$$\hat{j}_\gamma(\mathbf{q})=\sum_\mu \Big(\ket{R_{\gamma\mu}(\mathbf{q})}\bra{R_{\gamma\mu}(\mathbf{q})}-\ket{L_{\gamma\mu}(\mathbf{q})}\bra{L_{\gamma\mu}(\mathbf{q})}\Big),$$

and helicity as

$$h(\mathbf{q})=\mathbf{e}_{\mathbf{q}}\cdot \mathbf{J}^{\text{ph}}(\mathbf{q}) \qquad \text{or} \qquad h'(\mathbf{q})=\mathbf{v}(\mathbf{q})\cdot \mathbf{J}^{\text{ph}}(\mathbf{q}) .$$

These quantities transform differently under inversion, time reversal, and improper rotations, and those transformation laws control which crystal classes can support nonzero chiral phonons (Yang et al., 16 Jun 2025).

The decisive symmetry result is that inversion-symmetric, time-reversal-preserving crystals force phonon angular momentum to vanish identically. In the classification across all 230 space groups, type-I point groups contain inversion and therefore have vanishing phonon angular momentum; type-II groups contain only proper rotations and support $_1$0-wave helicity; type-III groups lack inversion but contain improper rotations and support higher-order helicity textures such as $_1$1-, $_1$2-, or $_1$3-wave patterns (Yang et al., 16 Jun 2025). This immediately separates polariton candidates into classes with uniform handedness and classes with momentum-space sign changes in helicity.

Pseudo-angular momentum adds a different layer. In a symmorphic crystal with an $_1$4-fold rotation,

$_1$5

while in non-symmorphic systems the corresponding screw-rotation eigenvalue makes $_1$6 $_1$7-dependent and generally non-quantized (Zhang et al., 2021). That quantity is indispensable for rotational selection rules, but it is not equivalent to mechanical angular momentum. The systematic symmetry analysis therefore rejects a common simplification: nonzero PAM does not, by itself, certify a chiral phonon, and CPL helicity flipping can arise from mirror symmetry acting as a half-wave plate even when the phonon angular momentum is zero (Zhang et al., 28 Mar 2025).

LiNbO$_1$8 furnishes an important symmetry example for polar crystals. Below $_1$9 K it has the ferroelectric polar corundum structure with space group $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$0, which lacks inversion symmetry but is not chiral as a space group because it contains $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$1-glide planes (Ueda et al., 4 Apr 2025). In that setting, nonzero $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$2 is allowed at some momenta and forbidden at others; at glide-plane momenta only cycloidal modes with $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$3 survive, while at generic momenta true chiral phonons with finite $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$4 are symmetry-allowed (Ueda et al., 4 Apr 2025). For polaritons, this means that handedness can depend not only on mode frequency but also on propagation direction relative to the nonsymmorphic elements.

3. Hybridization mechanisms and effective descriptions

At the level of minimal field theory, a phonon-polariton arises from resonant coupling between a photon mode and an infrared-active phonon. In second-quantized notation, one representative interaction is

$\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$5

and in a chiral crystal this coupling becomes polarization selective because the phonon eigenvectors are complex and circularly polarized (Ohe et al., 2024). In the quartz synthesis, this is explicitly extended to chiral branches $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$6 and circular photon operators $\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$7, suggesting a handedness-resolved interaction of the form

$\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$8

with the crystal chirality fixing which branch corresponds to positive or negative phonon angular momentum (Ohe et al., 2024).

In LiNbO$\frac{c^2 k^2}{\omega^2}=\varepsilon(\omega),$9, the driven lattice coordinate for each point along the E(TO$Q_x$0) polariton dispersion is modeled by

$Q_x$1

and the chiral phonon-polariton magnetic moment is obtained from the circular dynamics through

$Q_x$2

The same work models the electronic inverse Faraday contribution as

$Q_x$3

and concludes that neither phononic nor electronic terms alone reproduce the measured Faraday traces; a linear combination of both is required (Biggs et al., 29 Jul 2025). That result is significant because it places chiral phonon-polaritons in a broader magneto-optical sector rather than treating them as purely lattice excitations.

Two other mechanisms extend the construction. First, nonlinear phonon rectification can create geometric chirality in an otherwise achiral crystal through the cubic potential

$Q_x$4

so that coherent driving of two infrared-active modes $Q_x$5 produces a quasistatic displacement $Q_x$6 along a geometric chiral phonon (Romao et al., 2023). This suggests that a transiently induced chiral lattice can endow an otherwise ordinary phonon-polariton with gyrotropic, handedness-dependent optical response. Second, magnon–phonon coupling can produce nondegenerate elliptically polarized phonon pairs at $Q_x$7, as in FePSe$Q_x$8, where a block-diagonal hybridization matrix couples each phonon helicity only to the magnon of the same helicity (Ning et al., 2024). A plausible implication is that chiral phonon-polaritons can be engineered either from structurally chiral or polar lattices, or from nonchiral lattices whose phonons acquire handedness through nonlinear or magnetic hybridization.

4. Material platforms and experimental observables

Several experimentally distinct platforms now anchor the subject.

Platform	Chiral ingredient	Observable
$Q_x$9-quartz	Thermally driven phonon angular momentum tied to crystal handedness	ISHE voltage in W/Pt
LiNbO$Q_y$0	Propagating E(TO$Q_y$1) chiral phonon-polaritons	Ultrafast Faraday rotation
LiNbO$Q_y$2	Direct chiral phonons in a polar ferroelectric	O K-edge RIXS circular dichroism
hBN + magneto-optical photonic crystal	hBN phonon coupled to chiral BIC	Handedness-selective absorption
SiC nanoparticle chain	SSH-type topological phonon-polariton edge modes	Complex-band edge resonances and LDOS

In $Q_y$3-quartz, thermally driven phonons along the $Q_y$4-axis carry chirality-dependent angular momentum, and the sign of the net phonon angular momentum flips between $Q_y$5 and $Q_y$6 enantiomers (Ohe et al., 2024). The experimentally measured conversion coefficient

$Q_y$7

reverses sign between left- and right-handed quartz and changes sign again when W is replaced by Pt, consistent with inverse spin Hall detection of angular momentum transferred from phonons to electron spins (Ohe et al., 2024). Although this study does not construct polaritons, it establishes the structural-to-dynamical chirality map that a quartz-based chiral phonon-polariton would inherit.

LiNbO$Q_y$8 provides both a direct chiral-phonon and a direct chiral-phonon-polariton realization. In the phonon-polariton experiment, two broadband THz pulses with orthogonal polarizations and controlled delay create an approximately circular THz driving field that excites a continuum of wavevectors along the E(TO$Q_y$9) polariton branches, yielding a propagating three-dimensional chiral polariton wavepacket (Biggs et al., 29 Jul 2025). Dual chopping isolates the nonlinear chiral signal through

$$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$0

and the isolated Faraday signal reverses sign when the THz helicity is switched and peaks near the most circular ellipticities (Biggs et al., 29 Jul 2025). In the same material, O K-edge RIXS detects chiral phonons directly at $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$1 meV, $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$2 meV, and $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$3 meV through circular dichroism, with the signal vanishing on a glide plane and reversing between glide-related momenta, exactly as required by the $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$4 symmetry (Ueda et al., 4 Apr 2025).

A third experimental protocol resolves “truly chiral phonons” by symmetry-selective nonlinear THz spectroscopy. In $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$5-quartz and $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$6-TeO$$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$7, the azimuthally averaged chiral response isolates mirror-odd tensor elements such as $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$8, and vector-field detection directly reveals time-dependent polarization rotation of the emitted THz field, thereby separating phonon chirality from mere circular-polarization inversion (Wang et al., 28 Apr 2026). This is especially consequential for polariton work because it provides a tabletop, phase-resolved route to identify whether the phononic constituent of a hybrid mode truly carries real-space angular momentum.

Finally, the magnetically tunable hBN platform realizes chiral phonon-polaritons through strong coupling between an hBN phonon and a chiral quasi-BIC supported by a magneto-optical photonic crystal (Sun et al., 14 Apr 2026). The bare photonic crystal has $$Q_x(t)=Q_0\cos(\omega t),\qquad Q_y(t)=Q_0\sin(\omega t),$$9 at $\mathbf{q}$0 before symmetry breaking; with an asymmetry $\mathbf{q}$1 nm and a 15 nm hBN layer, the hybridization produces a Rabi splitting $\mathbf{q}$2 meV and coupling strength $\mathbf{q}$3 meV (Sun et al., 14 Apr 2026). Under magnetic bias the hybrid states show handedness-selective absorption for circularly polarized excitation.

5. Topological, non-Hermitian, and magnetic extensions

The chiral phonon-polariton problem extends naturally into non-Hermitian topology. In one dimension, dimerized SiC nanoparticle chains realize topological phonon polaritons that mimic the SSH model while retaining full radiative and absorptive non-Hermiticity (Wang et al., 2018). The Bloch problem is written as a $\mathbf{q}$4 effective Hamiltonian with complex matrix elements, and for longitudinal modes the topology remains classified by a quantized complex Zak phase even though the full Hamiltonian lacks exact chiral symmetry because of equal diagonal terms. Finite chains with $\mathbf{q}$5 support edge-localized topological phonon-polariton modes, while transverse modes exhibit a non-Hermitian skin effect and require a modified complex Zak phase to restore bulk-boundary correspondence (Wang et al., 2018). These modes are “chiral” in the SSH sense of sublattice winding rather than in the dynamical sense of circular ionic motion, but they show that phonon-polariton chirality can also be topological and non-Hermitian.

Magnetic control introduces a second extension. In the hBN quasi-BIC system, the effective photonic Hamiltonian is

$\mathbf{q}$6

with magneto-optical bias entering through the gyrotropic refractive tensor

$\mathbf{q}$7

and the full hybrid phonon–BIC problem is promoted to a $\mathbf{q}$8 polariton Hamiltonian (Sun et al., 14 Apr 2026). The resulting phonon-proportion analysis shows that the upper-branch phonon fraction changes from about $\mathbf{q}$9 at $$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$0 to about $$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$1 at $$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$2, while the lower branch changes from about $$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$3 to about $$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$4 (Sun et al., 14 Apr 2026). This is a direct demonstration that chiral phonon-polariton composition itself can be magnetically tuned.

A third extension comes from Berry-curved magnetic hybrids. In inversion-symmetric ferromagnetic bcc Fe, chirality-selective magnon–phonon coupling produces propagating truly chiral phonons, finite zero-point phonon angular momentum, and large anomalous Hall responses linked to finite Berry curvatures (Weißenhofer et al., 2024). This suggests that if analogous chiral optical phonons are made infrared-active and strongly coupled to light, the resulting phonon-polaritons should inherit Berry-curvature hot spots and Hall-like transport anomalies, even when the underlying chirality originates in magnetic rather than structural symmetry breaking.

6. Limitations, controversies, and open directions

Several constraints and controversies currently define the field. The most visible concerns interpretation of ultrafast magneto-optical signals. In LiNbO$$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$5, Faraday rotation data combined with a static Verdet constant yield an induced field estimate of $$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$6–$$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$7 T at the most circular ellipticity, but the work explicitly warns that this should be regarded as an “effective magneto-optical field strength,” not an unambiguous Maxwellian magnetic field, because linearly polarized THz excitation also rotates the probe through Raman-like processes (Biggs et al., 29 Jul 2025). The same study therefore argues that direct probes of magnetic order or Zeeman splittings are needed to separate true induced magnetic fields from nonmagnetic polarization-rotation channels.

A second limitation is definitional. The comprehensive symmetry study shows that neither nonzero pseudo-angular momentum nor CPL polarization inversion is a universal benchmark for phonon chirality, and the nonlinear THz study formalizes the stricter criterion that “truly chiral phonons require the simultaneous presence of chirality and real-space angular momentum” (Zhang et al., 28 Mar 2025, Wang et al., 28 Apr 2026). For chiral phonon-polaritons, this means that handedness-selective optical absorption or helicity conversion is insufficient evidence unless the lattice part is independently shown to carry finite angular momentum.

A third limitation is the gap between phonon demonstrations and full polariton theories. The quartz thermal-transport work establishes generation, propagation, and spin transfer of phonon angular momentum, but it contains no direct optical measurements and no explicit photon–phonon Hamiltonian or polariton-band calculation (Ohe et al., 2024). The geometric-chirality work provides a microscopic mechanism for inducing a transient chiral lattice through nonlinear phonon rectification, yet it also stops short of computing the resulting polariton bands (Romao et al., 2023). This suggests that the next theoretical layer should combine lattice dynamics, Born effective charges, gyrotropic dielectric tensors, and full electromagnetic mode calculations in one framework.

A fourth limitation is materials identification at scale. The recent catalogue of 11,614 compounds reports 2,738 materials with chiral phonon modes and a shortlist of 170 especially promising candidates, organized by symmetry class and helicity texture in an open-access database (Yang et al., 16 Jun 2025). A plausible implication is that the discovery bottleneck has shifted from symmetry screening to mode-specific optical viability: the decisive question is no longer only whether a material hosts chiral phonons, but whether those phonons are infrared-active, sufficiently long-lived, and embedded in a photonic environment that preserves or enhances their handedness.

The emerging agenda therefore has three convergent fronts. One is direct spectroscopy of polariton dispersion and polarization, especially in THz and mid-IR geometries that can resolve handedness-dependent branches. Another is device engineering: ferroelectric domain control in LiNbO$$\hat{J}^{\text{ph}}(\mathbf{q})=\big(\hat{j}_x(\mathbf{q}),\hat{j}_y(\mathbf{q}),\hat{j}_z(\mathbf{q})\big),$$8, magneto-optical BIC coupling in hBN-based heterostructures, and topological waveguiding in phonon-polaritonic nanostructures all provide distinct control knobs (Ueda et al., 4 Apr 2025, Sun et al., 14 Apr 2026, Wang et al., 2018). The third is a more microscopic theory of angular-momentum transfer, linking the phonon operator content, photon helicity, interfacial spin conversion, and Berry-curvature responses in one hybrid description. Current results already establish that chiral phonon-polaritons are not a single mechanism but a symmetry-governed class of hybrid excitations whose handedness can originate in structural chirality, ferroelectric polarity, magnetic order, nonlinear lattice rectification, or photonic chirality itself.