Topologically Protected Phonon-Polaritonic BICs

• Topologically protected phonon-polaritonic BICs are nonradiative resonant states emerging from symmetry and topological constraints in anisotropic polaritonic nanostructures.
• These states in isotopically enriched hBN nanoresonators exhibit subwavelength confinement with quality factors ultimately limited by intrinsic phonon damping.
• Numerical, experimental, and theoretical validations collectively demonstrate tunable quasi-BIC formation, paving the way for advanced mid-IR nanophotonic devices.

Searching arXiv for the specified paper and closely related work on phonon-polaritonic/topological BICs. Topologically protected phonon-polaritonic bound states in the continuum (BICs) are non-radiating resonant states embedded within the radiation continuum that arise in anisotropic polaritonic nanostructures through symmetry and topological constraints rather than through accidental parameter tuning. In periodic arrays of cylindrical nanoresonators made of isotopically enriched hexagonal boron nitride ($h^{11}\mathrm{BN}$), such states occur in the lower Reststrahlen band (RB-1), where the material supports surface phonon-polaritons and where radiative leakage can be fully suppressed at the $\Gamma$-point. Gupta et al. theoretically and numerically demonstrated, and experimentally validated, that these BICs are symmetry-protected and topologically charged, while symmetry breaking by oblique incidence converts them into quasi-BICs with strong field confinement and tunable radiation leakage (Gupta et al., 20 Aug 2025).

1. Material platform and polaritonic regime

The relevant platform is a periodic array of cylindrical nanoresonators composed of isotopically enriched $h^{11}\mathrm{BN}$, operated in the lower Reststrahlen band. Hexagonal boron nitride is uniaxial, with dielectric tensor

$$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$

Along each principal axis, the permittivity is described by a Lorentz phonon-oscillator model:

$$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$

Here, $\varepsilon_{\infty,j}$ is the high-frequency permittivity, $\omega_{TO,j}$ and $\omega_{LO,j}$ are the transverse and longitudinal optical-phonon angular frequencies, and $\gamma_j$ is the phonon damping. In RB-1, approximately $760\text{–}830\,\mathrm{cm}^{-1}$, the material satisfies $\Gamma$0 and $\Gamma$1, so surface phonon-polaritons confined at the $\Gamma$2-face can exist (Gupta et al., 20 Aug 2025).

Representative values reported for isotopically enriched $\Gamma$3 are $\Gamma$4, $\Gamma$5, $\Gamma$6, $\Gamma$7, $\Gamma$8, and $\Gamma$9. Within this spectral window, the anisotropy of $h^{11}\mathrm{BN}$0 and the geometry of the cylindrical resonators jointly define the phonon-polaritonic mode structure that permits BIC formation.

The significance of this material choice is explicit in the reported comparison with dielectric metasurfaces. Previous BIC implementations had primarily relied on dielectric metasurfaces and remained limited by the diffraction limit, whereas the phonon-polaritonic implementation in $h^{11}\mathrm{BN}$1 combines subwavelength confinement with a quality factor that is ultimately constrained by intrinsic phonon damping rather than by residual radiative outcoupling.

2. Symmetry protection at the $h^{11}\mathrm{BN}$2-point

At the $h^{11}\mathrm{BN}$3-point, where the in-plane Bloch wavevector satisfies $h^{11}\mathrm{BN}$4, the lattice symmetry of the square array of cylindrical pillars is $h^{11}\mathrm{BN}$5; in the idealized isolated-cylinder limit it is $h^{11}\mathrm{BN}$6. The eigenmodes are classified by irreducible representations of the relevant point group. A symmetry-protected BIC occurs when the resonator mode belongs to an irreducible representation orthogonal to that of the open radiation channel, so that the projection onto outgoing plane waves vanishes by symmetry (Gupta et al., 20 Aug 2025).

In the reported system, free-space plane waves at normal incidence transform according to the $h^{11}\mathrm{BN}$7 representation of $h^{11}\mathrm{BN}$8. The fundamental dipolar phonon-polariton of the cylinder carries an out-of-plane $h^{11}\mathrm{BN}$9-dipole moment with angular dependence $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$0 and $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$1. Under a $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$2 mirror $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$3, this $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$4 mode acquires a minus sign if it is antisymmetric in $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$5 across the mirror plane, corresponding to an irreducible representation such as $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$6 or $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$7. Because the normal-incidence plane wave is $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$8, the overlap integral between the mode and the radiation channel vanishes by parity mismatch.

This vanishing overlap makes the $$\varepsilon(\omega)=\mathrm{diag}\!\left[\varepsilon_\perp(\omega),\,\varepsilon_\perp(\omega),\,\varepsilon_\parallel(\omega)\right].$$9-point mode dark. In the terminology of the paper, it therefore forms a true symmetry-protected BIC. The physical content of the protection is not merely weak leakage but complete suppression of radiative losses at normal incidence, provided the symmetry is preserved.

3. Far-field topology and topological charge

The reported BIC is not only symmetry-protected but also topologically protected through the structure of its far-field polarization in reciprocal space. When the far-field $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$0 amplitude and phase are tracked around $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$1, the phase forms a vortex. Defining

$$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$2

the topological charge, or winding number, is

$$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$3

with $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$4 any loop encircling $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$5. The paper reports $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$6 (Gupta et al., 20 Aug 2025).

Because this winding number is nonzero, it is a topological invariant. In the reported interpretation, this forces $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$7 and guarantees robustness of the zero-coupling condition at $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$8. The role of topology here is therefore complementary to group-theoretical symmetry arguments: symmetry identifies the forbidden coupling channel at the high-symmetry point, while the nontrivial far-field winding encodes the robustness of that zero under continuous deformations that preserve the relevant structure of the mode.

A common misconception is to treat all BICs as purely symmetry-forbidden states with no topological content. The reported result directly rejects that reduction for this system: the $$\varepsilon_j(\omega)=\varepsilon_{\infty,j}\left[1+\frac{\omega_{LO,j}^2-\omega_{TO,j}^2}{\omega_{TO,j}^2-\omega^2-i\gamma_j\omega}\right], \qquad j\in\{\perp,\parallel\}.$$9-point state is both symmetry-protected and characterized by a nonzero topological charge in the far-field polarization. This suggests that the protection is best understood as a combined symmetry-topology mechanism rather than as a purely algebraic selection rule.

4. Quality factor, intrinsic loss, and quasi-BIC formation

The total quality factor is decomposed as

$\varepsilon_{\infty,j}$0

where $\varepsilon_{\infty,j}$1 is set by radiative leakage and $\varepsilon_{\infty,j}$2 is the intrinsic material limit imposed by phonon damping. For an ideal BIC at $\varepsilon_{\infty,j}$3, $\varepsilon_{\infty,j}$4, so the total quality factor approaches the absorption-limited value,

$\varepsilon_{\infty,j}$5

Using $\varepsilon_{\infty,j}$6 and $\varepsilon_{\infty,j}$7, the paper reports $\varepsilon_{\infty,j}$8, and also notes the consistently scaled estimate $\varepsilon_{\infty,j}$9 in rad/s units (Gupta et al., 20 Aug 2025).

When cylindrical symmetry is broken by tilting the incident beam by an angle $\omega_{TO,j}$0, or equivalently by taking $\omega_{TO,j}$1, the BIC evolves into a quasi-BIC. To lowest order in $\omega_{TO,j}$2, the radiative coupling matrix element satisfies $\omega_{TO,j}$3, so the radiative width follows

$\omega_{TO,j}$4

Correspondingly, the total linewidth at small tilt is

$\omega_{TO,j}$5

and the phenomenological quality factor satisfies

$\omega_{TO,j}$6

once $\omega_{TO,j}$7. The paper states that this $\omega_{TO,j}$8-law was explicitly verified in both simulated and experimental $\omega_{TO,j}$9 versus $\omega_{LO,j}$0 curves.

This transition from true BIC to quasi-BIC provides a controlled mechanism for trading lifetime against outcoupling. The BIC is non-radiative at normal incidence, whereas the quasi-BIC retains strong confinement while allowing angle-dependent access to the mode.

5. Numerical and experimental realization

The reported realization concerns the RB-1 quasi-BIC in a structure with $\omega_{LO,j}$1, $\omega_{LO,j}$2, and $\omega_{LO,j}$3. The simulated transmittance shows no dip at $\omega_{LO,j}$4, consistent with a true BIC. At $\omega_{LO,j}$5, a deep dip appears at $\omega_{LO,j}$6, with $\omega_{LO,j}$7 in simulation and $\omega_{LO,j}$8 in experiment. At $\omega_{LO,j}$9, the dip remains near $\gamma_j$0, with $\gamma_j$1 in simulation and $\gamma_j$2 in experiment (Gupta et al., 20 Aug 2025).

The work also reports a comparison with a bright mode in RB-2 at $\gamma_j$3, for which the experimental quality factor is $\gamma_j$4 and nearly independent of $\gamma_j$5. This contrast separates the angle-sensitive quasi-BIC behavior in RB-1 from a bright-mode response lacking the same symmetry-protected mechanism.

Field localization is correspondingly strong. In simulations, $\gamma_j$6 in the cylindrical gap region exceeds the incident field by more than $\gamma_j$7, and the mode volume is reported as $\gamma_j$8. The paper further reports good agreement between HFSS-driven-mode predictions and FTIR-microscope transmittance measurements after tilting the substrate by $\gamma_j$9 and $760\text{–}830\,\mathrm{cm}^{-1}$0.

Taken together, these results establish not only the existence of the state but the consistency of the theoretical, numerical, and experimental descriptions. The absence of a normal-incidence dip, the emergence of angle-dependent sharp resonances, the measured quality factors, and the agreement between HFSS and FTIR all support the interpretation of a symmetry-protected BIC transitioning into a tunable quasi-BIC.

6. Scientific implications and application space

The reported demonstration opens several avenues in mid-infrared nanophotonics. The paper identifies ultra-narrow-band mid-IR filters and modulators whose resonance can be tuned by angle or by local refractive-index changes, low-threshold polariton lasers exploiting the high $760\text{–}830\,\mathrm{cm}^{-1}$1 and slow-light enhancement of nonlinearities, surface-enhanced infrared absorption spectroscopy with extremely high field confinement, and integration with quantum emitters such as molecular vibrational lines and defects in 2D materials for Purcell-enhanced single-photon sources at $760\text{–}830\,\mathrm{cm}^{-1}$2 (Gupta et al., 20 Aug 2025).

It also points to scalable, wafer-scale metasurfaces in the Reststrahlen band of hBN, benefiting from CMOS-compatible substrate processes. The broader significance is the combination of low-loss phonon-polaritonic response, deeply subwavelength confinement, and topological plus symmetry protection in a single platform. Within the formulation of the paper, this combination enables high-$760\text{–}830\,\mathrm{cm}^{-1}$3 mid-IR resonators with minimal radiation leakage and tunable outcoupling.

A plausible implication is that these structures occupy a distinct regime relative to conventional dielectric BIC metasurfaces: they inherit the BIC framework of symmetry-protected non-radiating states, yet they do so in a polaritonic material system where the characteristic scale is not fixed by the diffraction limit and where the ultimate linewidth is bounded by phonon damping. In that sense, topologically protected phonon-polaritonic BICs connect the theory of BICs, the physics of hyperbolic or anisotropic phonon-polaritons, and the engineering of mid-infrared resonant nanostructures into a unified research direction.