Fin Lithium Niobate Acoustic Resonator

• FinLN acoustic resonators are microfabricated electromechanical devices with a high-aspect-ratio fin geometry in lithium niobate, enabling enhanced acoustic confinement.
• They employ layered substrates and sidewall interdigitated electrodes to achieve strong lateral, vertical, and longitudinal confinement for well-defined resonant modes.
• Experimental results reveal a 1.8× improvement in electromechanical coupling over planar SAW devices, promising advances in RF MEMS, acousto-optic modulation, and nonlinear phononics.

Fin lithium niobate (FinLN) acoustic resonators are microfabricated electromechanical devices employing a high-aspect-ratio, three-dimensional fin geometry within single-crystal lithium niobate (LiNbO₃) platforms. This architecture enhances acoustic confinement, maximizes electromechanical coupling, and enables compact device integration, outperforming conventional planar surface acoustic wave (SAW) structures in electromechanical efficiency, scalability, and versatility for radio-frequency (RF) MEMS and emerging piezo-optomechanical applications (Ni et al., 25 Jan 2026, Mayor et al., 2020).

1. Device Architecture and Physical Principles

FinLN resonators utilize thick lithium niobate on insulator (LNOI) or thin-film lithium niobate on sapphire substrates for device fabrication. The canonical structure comprises a high-aspect-ratio fin (W = 2 μm, H = 5 μm) with a pitch equal to the acoustic wavelength (λ, e.g., 12 μm for 310 MHz resonance) patterned into the LiNbO₃ layer. A buried SiO₂ or sapphire substrate serves to reflect acoustic energy, increasing vertical confinement and suppressing substrate leakage. Sidewall interdigitated transducer (IDT) electrodes (e.g., 15 nm Ti / 200 nm Al) are conformally deposited on opposing fin faces, forming differential excitation across the fin width.

Three-dimensional acoustic confinement in FinLN derives from:

• Lateral confinement: Fin sidewalls and periodic electrode structures define a waveguide, suppressing transverse modes otherwise permitted in planar devices.
• Vertical confinement: Acoustic impedance contrast between LiNbO₃ and SiO₂/Si or sapphire reflects energy into the fin.
• Longitudinal confinement: The finite fin length and periodic IDT implement a standing-wave cavity for the fundamental longitudinal mode.

Index guiding, analogous to photonic rib waveguides, further facilitates strong mode localization within the fin for thin-film variants on sapphire (Mayor et al., 2020).

2. Mathematical Models and Key Performance Equations

The FinLN resonant acoustic mode satisfies the modified elastodynamic eigenproblem for anisotropic piezoelectric media: $$\nabla\!\cdot\!\boldsymbol{\sigma} + \rho\,\omega^2\,\mathbf{u} = 0, \quad \boldsymbol{\sigma} = \mathbf{c}:\nabla_s\mathbf{u} + \mathbf{e}^T\mathbf{E}, \quad \mathbf{D} = \mathbf{e}:\nabla_s\mathbf{u} + \boldsymbol{\varepsilon}\,\mathbf{E}$$ where $\mathbf{u}$ is the displacement field, $\mathbf{c}$ the elastic tensor, $\mathbf{e}$ the piezoelectric tensor, and $\boldsymbol{\varepsilon}$ the permittivity tensor. The fundamental mode occurs at: $$f_s \approx \frac{v_\mathrm{eff}}{\lambda}, \quad v_\mathrm{eff} = \sqrt{c_\mathrm{eff}/\rho}$$ Electromechanical coupling, quantified by $k_{\mathrm{eff}}^2$, is computed via the modified Butterworth–Van Dyke (mBVD) model: $$k_{\mathrm{eff}}^2 = \frac{\pi^2}{8} \frac{C_m}{C_0} = \frac{\pi^2}{8} \frac{f_p^2 - f_s^2}{f_s^2}$$ Mechanical quality factor ($Q_m$): $$Q_m = \frac{\omega_0}{\Delta\omega}, \quad \omega_0 = 2\pi f_s$$

For thin-film FinLN resonators on sapphire, index-guided acoustic modes are described by an effective acoustic index: $\mathbf{u}$0 Resonance in racetrack architectures follows the round-trip phase condition: $\mathbf{u}$1 with free spectral range: $\mathbf{u}$2

3. Fabrication Methodologies and Critical Tolerances

The LNOI-based FinLN fabrication sequence relies on:

1. Patterning the fin geometry by photolithography and deep Ar⁺ ion etching (e.g., ~5.3 μm depth for full isolation).
2. Secondary lithography to define sidewall IDT regions.
3. Glancing-angle e-beam evaporation (GLAD) to deposit metal (Ti/Al) conformally on sidewalls.
4. Liftoff processing to confine electrodes to fin surfaces.

Critical tolerances include:

• Etch depth uniformity (±0.1 μm) for reproducible boundaries and resonance.
• Fin width control (±0.1 μm), influencing H/W aspect ratio and $\mathbf{u}$3 (higher H/W yields stronger coupling).
• Electrode alignment and thickness (<10 nm variation) to ensure electric field–acoustic mode overlap.

Thin-film sapphire-based FinLN resonators employ rib waveguide lithography, selective etching, and surface passivation for Q optimization (Mayor et al., 2020).

4. Experimental Characterization and Comparative Metrics

Empirical results for thick-film Z-cut FinLN resonators (λ = 12 μm) demonstrate:

• Resonance frequency $\mathbf{u}$4 MHz.
• Effective coupling $\mathbf{u}$5.
• Mechanical quality factor $\mathbf{u}$6.

Planar SAW reference resonators on identical LNOI wafers (same thickness, orientation, pitch) yield $\mathbf{u}$7, confirming a $\mathbf{u}$8 enhancement in FinLN due to 3D confinement.

Thin-film FinLN devices (w ≈ 1 μm, t ≈ 300 nm on sapphire) demonstrate group velocity $\mathbf{u}$9 m/s, ring resonance Q factors up to $\mathbf{c}$0 at 4 K, and insertion losses dominated by scattering and phonon–phonon interactions (Mayor et al., 2020).

5. Nonlinear Phononics and Integrated Device Applications

FinLN architectures enable strong nonlinear phononic interactions, notably four-wave mixing (FWM). Experimentally measured modal nonlinear coefficient $\mathbf{c}$1 mW⁻¹ mm⁻¹, parametric threshold $\mathbf{c}$2 mW, and FWM efficiency $\mathbf{c}$3 with $\mathbf{c}$4–$\mathbf{c}$5 (mW²)⁻¹. Negligible phase modulation (SPM/XPM) affirms mechanical nonlinearity dominance.

Potential uses encompass:

• RF MEMS: filters, duplexers (0.1–1 GHz).
• Voltage-controlled oscillators (low phase noise).
• Piezo-optomechanical systems: integration with optical waveguides enables acousto-optic modulation and microwave–optical transduction.

6. Design Optimization and Frequency Scaling

To maximize FinLN performance:

• Frequency targeting: $\mathbf{c}$6, e.g., for 1 GHz operation, $\mathbf{c}$7 μm (Z-cut LN $\mathbf{c}$8–$\mathbf{c}$9 km/s).
• Aspect ratio ($\mathbf{e}$0) and wavelength ($\mathbf{e}$1): $\mathbf{e}$2 increases monotonically with $\mathbf{e}$3 and decreases with $\mathbf{e}$4; approximate scaling $\mathbf{e}$5.
• Waveguide width/height: For index guiding and reduced scattering, maintain $\mathbf{e}$6 large relative to sidewall roughness.
• Surface treatment: Employ O₂ plasma/piranha descum and optional oxide cladding to reduce phonon losses and optimize Q.
• Electrode shape/taper: Slant sidewalls to minimize mode reflection and optimize overlap.
• Resonator coupling: Adjust coupling gap $\mathbf{e}$7 and length $\mathbf{e}$8 for critical ($\mathbf{e}$9) resonance.
• Nonlinear enhancement: Increase effective length $\boldsymbol{\varepsilon}$0; operate at low temperature for Q improvement.

High-Q ($\boldsymbol{\varepsilon}$1), low-loss ($\boldsymbol{\varepsilon}$2 dB/mm), and sub-milliwatt nonlinear parametric thresholds are achievable with high-quality films, smooth etch profiles, matched IDTs, and cryogenic operation (Mayor et al., 2020).

7. Significance, Limitations, and Forward Perspectives

FinLN resonators represent a versatile platform for compact RF MEMS and advanced hybrid photonic–phononic systems. Principal advantages include:

• Enhanced electromechanical coupling via 3D acoustic confinement.
• Significant footprint reduction ($\boldsymbol{\varepsilon}$3m $\boldsymbol{\varepsilon}$4 $\boldsymbol{\varepsilon}$5 unit cell) versus planar SAW devices needing wide apertures (%%%%39$\mathbf{c}$740%%%%).
• Compatibility with thick LNOI simplifies process integration compared to thin-film ion slicing.

Limitations currently include increased fabrication complexity (deep etch, precision sidewall metallization), appearance of weak FSR overtones, and uncharacterized high-power thermal/mechanical stability. A plausible implication is that further process refinement and material optimization could increase Q, decrease loss, and expand high-frequency operation or nonlinear conversion capabilities.

Ongoing research demonstrates promising paths toward scalable, high-Q, and high-efficiency acoustic resonators, bridging RF electronics and integrated microwave–optical systems for quantum technologies (Ni et al., 25 Jan 2026, Mayor et al., 2020).