🔬 applied physics

FinLN: 3D Fin Lithium Niobate Acoustic Resonators

This paper presents the fabrication and characterization of a three-dimensional fin lithium niobate (FinLN) acoustic resonator that achieves strong acoustic confinement and a 1.8x enhancement in electromechanical coupling compared to planar counterparts, establishing it as a promising platform for compact, high-performance RF MEMS and piezo-optomechanical systems.

Original authors: Haorui Ni, Sunil Bhave

Published 2026-01-27

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Haorui Ni, Sunil Bhave

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to build a tiny, super-efficient musical instrument that vibrates at incredibly high speeds to process radio signals. For years, engineers have been trying to make these instruments smaller and more powerful, but they've hit a wall: making them too flat makes them weak, and making them thicker makes them too big and expensive.

This paper introduces a clever new design called FinLN (Fin Lithium Niobate) that solves this problem by changing the shape of the instrument entirely.

The Problem: Flat vs. Deep

Think of traditional acoustic devices like a flat sheet of paper lying on a table. To make the sound (or vibration) stay in one spot and not leak out to the sides, you have to make the paper very wide. This takes up a lot of space on your circuit board, like trying to park a huge truck in a tiny garage.

Also, making these "sheets" out of high-quality crystal (Lithium Niobate) usually requires expensive, high-tech slicing methods that are hard to do in bulk.

The Solution: The "Fin" Shape

The researchers at Purdue University decided to stop thinking flat and start thinking 3D. They took inspiration from computer chips, where engineers moved from flat transistors to "FinFETs" (tall, thin fins) to get better control.

They did the same thing with their acoustic device:

The Material: Instead of using expensive, ultra-thin slices of crystal, they used a thick block of Lithium Niobate (about 5 micrometers thick). They made this using a cheaper, more reliable method involving gluing and polishing, rather than expensive ion-slicing.
The Shape: They used deep etching to carve this thick block into tall, thin walls (fins), standing up like a row of tiny dominoes or the fins on a shark.
The Electrodes: Instead of putting metal contacts on top, they wrapped metal electrodes around the sides of these fins.

How It Works: The "Acoustic Cage"

Here is the magic of the FinLN design:

The Cage Effect: In a flat device, sound waves can easily escape sideways. In the FinLN, the tall, thin shape acts like a 3D cage. The sound waves are trapped inside the fin, bouncing back and forth between the walls. This "confinement" makes the vibration much stronger and more efficient.
The Side-Step: Because the metal electrodes are on the sides, they push and pull the fin from the sides, creating a very tight squeeze that generates strong vibrations without needing a huge surface area.
The Result: They got a device that is 1.8 times more efficient at converting electricity into sound (and vice versa) compared to a traditional flat device made from the exact same material.

The Performance

The team built these devices and tested them:

Speed: They vibrate at about 300 MHz (300 million times a second), which is perfect for modern radio and signal processing.
Efficiency: They achieved a coupling coefficient of 6.2%, which is a very high number for this type of technology.
Quality: The vibrations are very clean, with a "quality factor" of 430, meaning the sound rings true with very little energy loss.

Why This Matters (According to the Paper)

The paper claims this is a big deal because:

It's Compact: You get high performance in a tiny footprint, allowing for denser electronics.
It's Cheaper: By using thick, bonded crystals instead of expensive ion-sliced films, the manufacturing process is more scalable and cost-effective.
It's Versatile: The deep, etched shape is naturally compatible with optical devices (lasers and light), suggesting it could be used to build systems that mix sound and light (piezo-optomechanical systems) in the future.

In short, the researchers took a thick block of crystal, carved it into tall, thin fins, and wrapped the sides with metal. This simple shape change turned a standard acoustic device into a high-performance, space-saving powerhouse.

Technical Summary: 3D Fin Lithium Niobate Acoustic Resonators (FinLN)

Problem Statement
Microacoustic devices operating in the hundreds of MHz to GHz range are critical for sensing, signal processing, and quantum technologies. While thin-film lithium niobate (LN) has emerged as a superior piezoelectric platform due to its single-crystal quality and high electromechanical coefficients, current high-performance implementations face significant limitations. Devices relying on sub-micrometer LN films typically require ion-slicing technologies (e.g., Smart-Cut), which are costly, process-intensive, and limited in wafer availability. Conversely, thick piezoelectric films (>5 µm) offer lower cost and better manufacturability but have historically been difficult to utilize for high-performance resonators without the complex processing associated with thin films. Furthermore, conventional planar surface acoustic wave (SAW) devices often require large apertures (20× the acoustic wavelength) to suppress transverse modes, resulting in large device footprints that hinder dense integration.

Methodology
The authors introduce a three-dimensional fin-based lithium niobate (FinLN) acoustic resonator, inspired by the transition from planar MOSFETs to FinFET architectures. The methodology involves:

Material Platform: Utilization of a thick Z-cut lithium niobate on insulator (LNOI) wafer featuring a 5 µm LN device layer, a 2 µm buried SiO2 layer, and a Si handle substrate. This thick layer is achieved via wafer bonding and mechanical lapping rather than ion-slicing.
Fabrication: The process employs deep Ar ion etching to define high-aspect-ratio LN fin structures (approx. 5 µm deep). Sidewall electrodes are patterned using e-beam evaporation with glancing-angle deposition (GLAD) to ensure conformal metallization of the fin sidewalls. The electrodes consist of 15 nm Ti and 200 nm Al, arranged as interdigital transducers (IDTs) in a ground–signal configuration.
Design & Simulation: Three-dimensional finite-element simulations (COMSOL Multiphysics) were conducted to optimize design parameters. The study systematically varied fin width and acoustic wavelength to evaluate the electromechanical coupling coefficient ( $k_t^2$ ) and resonance frequency ( $f_s$ ).
Characterization: Devices were characterized using a network analyzer and ground–signal–ground (GSG) probes. A planar SAW reference device was fabricated on the same wafer to isolate the performance benefits of the 3D geometry.

Key Contributions

Novel Architecture: Demonstration of a FinLN resonator that leverages a high-aspect-ratio fin cavity to provide strong lateral acoustic confinement, eliminating the need for wide IDT apertures required in planar SAW devices.
Sidewall Excitation: Implementation of conformal sidewall electrodes that orient the electric field along the fin length, suppressing fringing fields and enhancing spatial overlap with the confined acoustic mode.
Process Integration: A fabrication strategy that utilizes cost-effective, thick LNOI wafers (bonded and lapped) combined with deep ion etching, avoiding the complexity and cost of ion-sliced thin films.

Results

Performance Metrics: The fabricated FinLN devices operate near 300 MHz (specifically ~310 MHz for the $\lambda=12$ µm device). The device with a 2 µm fin width and 12 µm acoustic wavelength achieved an effective electromechanical coupling coefficient ( $k_{eff}^2$ ) of 6.2% and a mechanical quality factor ( $Q_m$ ) of 430.
Design Trends: Experimental results confirmed simulation predictions: $k_{eff}^2$ exhibits a monotonic increase as the acoustic wavelength decreases (from 0.5% at $\lambda=20$ µm to 6.2% at $\lambda=12$ µm).
Comparative Analysis: When compared to a planar SAW resonator fabricated on the same wafer (which yielded $k_{eff}^2 \approx 3.4\%$ ), the FinLN configuration demonstrated an 1.8× enhancement in effective electromechanical coupling.
Mode Behavior: The devices exhibit a weak free spectral range (FSR) attributed to longitudinal overtone modes, but the fundamental mode is clearly defined with a distinct resonance peak.

Significance
The paper establishes FinLN as a promising platform for low-cost, compact, and high-performance RF MEMS and piezo-optomechanical systems. By utilizing thick LN layers and 3D geometric confinement, the architecture achieves enhanced electromechanical coupling and a compact footprint simultaneously. The authors note that the deep-etched geometry is compatible with optical waveguide fabrication and that the sidewall electrode configuration may enhance acousto-optic interaction strength, suggesting potential for integrated acousto-optic devices and hybrid photonic–phononic systems. The work validates that thick-film LN, traditionally limited by processing constraints, can be engineered into high-performance resonators through 3D structuring, offering a scalable alternative to ion-sliced thin-film technologies.