Photonic-Phononic Integration

• Photonic-Phononic Integration is the targeted co-localization of optical and mechanical modes in nanoscale structures, enabling robust photon–phonon interactions.
• Engineered periodic structures like nanobeams, 2D crystals, and waveguides facilitate simultaneous bandgap confinement and efficient optomechanical coupling.
• Applications include narrowband RF filtering, quantum transduction, and optomechanical sensing with high Q-factors and enhanced conversion efficiencies.

Photonic-Phononic Integration refers to the intentional co-localization and mutual engineering of electromagnetic (photonic) and mechanical (phononic) modes within a single micro- or nanoscale structure, yielding strong photon–phonon interactions for classical and quantum information processing, signal transduction, and metrological applications. This coupling is realized via engineered periodic or aperiodic material patterns (e.g., nanobeams, 2D lattices, waveguides), achieving bandgap or modal confinement of both infrared photons (typically near 200 THz, λ ≈ 1.5 μm) and hypersonic phonons (0.1–10 GHz). Integration approaches are underpinned by optomechanical Hamiltonians, and advance beyond previous architectures by leveraging wavelength-scale confinement, simultaneous photonic-phononic bandgaps, and engineered spatial overlap, leading to unprecedented sensitivity, bandwidth, and multi-functionality in planar chip-scale devices.

1. Foundational Physical Principles and Hamiltonian Formalism

The fundamental interaction governing photonic-phononic integration is encapsulated in the canonical cavity optomechanical Hamiltonian:

$$H = \hbar \omega_o a^\dagger a + \hbar \Omega_m b^\dagger b - \hbar g_0 a^\dagger a (b + b^\dagger)$$

where $a$ ($a^\dagger$) and $b$ ($b^\dagger$) are the annihilation and creation operators for the optical and mechanical cavity modes with frequencies $\omega_o$ and $\Omega_m$, respectively, and $g_0$ is the vacuum optomechanical coupling rate. The vacuum coupling is given by $$g_0 = (\partial\omega_o/\partial x) x_{\mathrm{zpf}}$$ where $$x_{\mathrm{zpf}} = \sqrt{\hbar/2 m_\mathrm{eff} \Omega_m}$$ is the mechanical zero-point amplitude and $a$0 is the effective mass of the mechanical mode.

Optomechanical coupling arises from two predominant mechanisms: (i) the moving-boundary (MB) effect, where deformations of the dielectric boundary shift the optical resonance, and (ii) the photoelastic (PE) effect, wherein elastic strain modulates the refractive index. The frequency shift is decomposed as

$a$1

$a$2

where $a$3 is the displacement, $a$4 is the permittivity, $a$5 the electric displacement, and $a$6 encodes the strain-induced permittivity perturbation. Both contributions scale with the overlap of optical fields and mechanical displacement, forming the basis for strong optomechanical interaction when high spatial co-localization is established (Safavi-Naeini et al., 2018).

2. Principal Architectures: Crystals, Waveguides, and Hybrid Platforms

2.1 Photonic-Phononic Crystals and Cavities

Structures with periodic modulation at sub-wavelength scales—such as 1D nanobeam crystals (0906.1236), 2D snowflake lattices (Safavi-Naeini et al., 2010, Safavi-Naeini et al., 2014), and programmable hexagonal phoxonic lattices (Bharadwaj et al., 5 Aug 2025)—enable simultaneous photonic and phononic bandgaps. In these, TE-polarized optical modes (ω/2π ≈ 200 THz) and GHz-band acoustic modes are both confined by Bragg scattering. For instance, in a planar SOI nanobeam (w ≈ 1.4 μm, t = 220 nm, Λ ≈ 362 nm), engineered quadratic modulation of the lattice yields Hermite-Gauss photon and phonon defect modes, co-localized with sub-(λ/n)^3 optical and (λ_m)^3 mechanical mode volumes (0906.1236). Effective mass $a$7 as low as 50–1000 fg and optomechanical coupling lengths $a$8 approaching 2.9 μm have been achieved.

In 2D snowflake crystals, tuning parameters such as snowflake arm width and radius controls the simultaneous in-plane bandgaps for both photonic (≈190–216 THz) and phononic modes (≈6.5–12 GHz) (Safavi-Naeini et al., 2010). Quadratic tapering of defect waveguides enables localization of both optical and mechanical modes, with computed $a$9 up to ≈292 kHz (Safavi-Naeini et al., 2010). Programmatic control over geometric parameters in hexagonal lattices enables independent tuning of photonic and phononic gaps by ≈20% (Bharadwaj et al., 5 Aug 2025).

2.2 Photonic-Phononic Waveguides and Hybrid Materials

Travelling-wave architectures extend integration beyond resonators, enabling wideband and multi-channel operations. Suspended silicon nanowires (Laer et al., 2014) with sub-μm cross-sections and minimal oxide pillars (15 nm) facilitate strong co-confinement of near-infrared TE modes and 10 GHz phonons, yielding measured forward SBS gain γ_SBS ≈ 3.2×10³ W⁻¹ m⁻¹ in straight wires and rings. The SBS process is governed by the spatial overlap of optically induced force (electrostriction + radiation pressure) and acoustic displacement.

On piezoelectric platforms, such as GaN-on-sapphire (Zhang et al., 2 Mar 2025) or TFLN hybrid waveguides (Yang et al., 12 Sep 2025), simultaneous optical (λ = 1.55 μm) and acoustic (≈1 GHz) confinement is achieved without suspended geometries, using refractive index and acoustic impedance mismatches for guidance. In GaN, IDTs permit direct RF-to-acoustic transduction, facilitating multi-tone, multi-channel operation with Q_ac up to 1.2×10⁴ and measured extinction/crosstalk >30 dB/channel (Zhang et al., 2 Mar 2025). In TFLN devices, nine simultaneous channels across 40 nm of optical bandwidth and 250 MHz of RF bandwidth are realized, with internal photon–phonon conversion efficiencies up to 2.2% (Yang et al., 12 Sep 2025).

3. Coupling Mechanisms and Figures of Merit

The salient interaction is governed by the Hamiltonian

$a^\dagger$0

for cavity modes, and by a distributed Brillouin Hamiltonian (travelling-wave):

$a^\dagger$1

where $a^\dagger$2 represents the local coupling rate (Yang et al., 12 Sep 2025).

Key performance metrics:

Parameter	Typical Value / Range	Device Type
Vacuum coupling rate $a^\dagger$3	220 kHz – 1 MHz	Si crystals (Safavi-Naeini et al., 2014), snowflake (Safavi-Naeini et al., 2010)
Mechanical Q-factor $a^\dagger$4	$a^\dagger$5 – $a^\dagger$6	Crystals, waveguides, memories
SBS gain coefficient $a^\dagger$7	$a^\dagger$8–$a^\dagger$9 W$b$0m$b$1	Si wires (Laer et al., 2014), chalcogenide (Merklein et al., 2016)
Bandwidth (filter, memory, etc.)	1–10 MHz (narrow), 1–2 GHz (wide)	PPER filter (Kittlaus et al., 2017), TFLN (Yang et al., 12 Sep 2025)
Optical Q-factor $b$2	$b$3 – $b$4	Crystals, cavities

High cooperativity is achieved via small effective mass, large spatial overlap of modes, and high acoustic and optical $b$5-factors. In high-$b$6 nanobeams or membranes, strong back-action supports phenomena such as ground-state cooling and quantum-limited measurement (0906.1236, Safavi-Naeini et al., 2014).

4. Device Implementations and Experimental Characterization

4.1 Nanobeam and 2D Crystal Devices

In the canonical silicon nanobeam structure, photonic and phononic bands are simulated via FEM (COMSOL) and plane-wave (MPB) methods for the relevant geometric parameters (t, w, Λ, hole shapes). Cavity formation leverages defect modulation of unit cell period or hole radii, yielding simulated and measured $b$7, $b$8 (air), and $b$9 up to 2 kHz for fundamental modes (0906.1236).

In 2D snowflake OMCs, defect cavities are formed by tapering hole radii over ~14 unit cells, yielding measured $b^\dagger$0 kHz, optical $b^\dagger$1, and mechanical $b^\dagger$2. Two-tone optical spectroscopy (e.g., EIT-type) is used to extract coupling rates, mode volumes, and dissipation coefficients (Safavi-Naeini et al., 2014).

4.2 Emit-Receive and Multi-Channel Waveguides

Photonic-phononic emit-receive (PPER) architectures employ coupled waveguides and phononic crystal membranes to achieve reflectionless, traveling-wave, narrowband MHz-scale bandpass or serial filter banks. For example, parallel suspended Si waveguides (≃500 nm) sharing a high-$b^\dagger$3 (Q ≈ 820 – $b^\dagger$4) Lamb-wave resonance support device lengths up to several centimeters, internal conversion efficiencies >10%, and dynamic range SFDR >99 dB Hz$b^\dagger$5 (Kittlaus et al., 2017, Gertler et al., 2019).

Hybrid waveguide devices on TFLN enable multi-channel operation (up to 9 simultaneously demonstrated) with 2.2% internal and $b^\dagger$6 system photon–phonon conversion efficiency, single-channel bandwidths ≈3.5 MHz, and continuous optical tuning >40 nm (Yang et al., 12 Sep 2025).

5. Bandgap Engineering and Modal Co-Localization Strategies

Photonic-phononic bandgap engineering exploits Bloch–Floquet periodicity and eigenmode analysis to realize simultaneous energy gaps for optical and elastic waves in the same geometry. In programmable hexagonal or snowflake lattices, geometric knobs—sector radius $b^\dagger$7, tether length $b^\dagger$8, snowflake arm width etc.—control both photon and phonon gaps independently. For instance, tuning $b^\dagger$9 in a hexagonal lattice shifts photonic band edges by 18.2% and phononic edges by 21.4%, facilitating robust dual-mode confinement (Bharadwaj et al., 5 Aug 2025).

Overlap of gaps enables the formation of defect regions supporting co-localized high-$\omega_o$0 photon and phonon modes, essential for maximizing optomechanical interaction. Alternatively, decoupling modal regions by tuning $\omega_o$1 or $\omega_o$2 outside gap overlap can serve vibration isolation purposes for photonic circuitry without suppressing targeted phonon transmission (Bharadwaj et al., 5 Aug 2025).

6. Signal Processing, Memory, and Quantum Transduction Applications

Photonic-phononic integration enables:

• Narrowband RF-Photonic Filtering: Simultaneous acoustic and optical resonances enable MHz-wide, high-rejection filters with low insertion loss, essential for next-generation RF communications and radar. Cascading PPER sections yields higher-order filter responses with >70 dB out-of-band rejection (Shin et al., 2014, Kittlaus et al., 2017, Gertler et al., 2019).
• Coherent Photonic-Phononic Memory: Optical information is coherently mapped to GHz-frequency hypersound phonons, yielding phase-preserving variable delays up to ≈10 ns and multi-wavelength, low-crosstalk operation (Merklein et al., 2016).
• Quantum Transduction: Strongly coupled, high-$\omega_o$3 optomechanical systems are under development for conversion between microwave and optical photons for quantum information, with recent multi-channel conversion architectures realizing 2.2% internal efficiency at room temperature and projected >50% at cryogenic conditions (Yang et al., 12 Sep 2025).
• Sensing and Metrology: Small $\omega_o$4, large $\omega_o$5, and spectrally isolated phonon modes support mass sensing down to zeptogram scales (e.g., Δf ≈ 700 Hz for hemoglobin binding), quantum-limited displacement detection, and gravitational sensor platforms (0906.1236).
• Integrated Nonreciprocal Devices: By exploiting traveling-wave photon–phonon coupling, devices such as isolators, circulators, and delay lines with broad bandwidth and low loss are achievable (Safavi-Naeini et al., 2018).

7. Prospects, Scalability, and Fabrication Considerations

Large-scale photonic-phononic integration requires platforms supporting high optical and acoustic confinement, low-loss operation, field-appropriate piezoelectricity, and CMOS-compatible process flows. Recent demonstrations in GaN-on-sapphire and TFLN-on-sapphire leverage acoustic index mismatch (e.g., $\omega_o$6) for solid-on-insulator routing, eliminating the need for fragile suspended structures and supporting dense, robust, multi-layer circuit integration (Zhang et al., 2 Mar 2025, Yang et al., 12 Sep 2025).

Wafer-scale processing and integration with on-chip lasers, detectors, and RF electronics, together with piezoelectric actuation and high-$\omega_o$7 phononic crystal engineering, are paving the way toward hybrid quantum–classical photonic-phononic circuit platforms with reconfigurable, multi-channel functionality, robust thermal and fabrication tolerances, and compatibility with quantum information nodes (Zhang et al., 2 Mar 2025, Yang et al., 12 Sep 2025).

Open challenges include further reduction of acoustic loss, advanced topological design for broader simultaneous gaps, low-insertion-loss optical coupling, and suppression of thermal and quantum noise for applications in quantum networks and precision measurement. Nonetheless, the unification of photonic and phononic engineering principles at the wavelength scale now enables on-chip, high-cooperativity optomechanics, scalable signal processing, and future quantum interfaces (Safavi-Naeini et al., 2018, 0906.1236).