🔬 applied physics

High-Q Millimeter-Wave Acoustic Resonators in Thin-Film Lithium Niobate Using Higher-Order Antisymmetric Modes

This paper presents high-quality miniature millimeter-wave acoustic resonators based on thin-film lithium niobate utilizing third-order antisymmetric (A3) modes, achieving a 39.8 GHz device with a figure of merit of 13.7 and demonstrating frequency scalability across the 30–50 GHz range.

Original authors: Vakhtang Chulukhadze, Jack Kramer, Tzu-Hsuan Hsu, Omar Barrera, Ruochen Lu

Published 2026-02-13

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Vakhtang Chulukhadze, Jack Kramer, Tzu-Hsuan Hsu, Omar Barrera, Ruochen Lu

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-fast musical instrument that can hear and process radio signals for your next-generation 5G or 6G phone. This paper is about a team of engineers who built a microscopic drum that can vibrate at incredibly high speeds (millimeter waves) without losing its energy too quickly.

Here is the story of how they did it, explained without the jargon:

1. The Problem: The "Heavy Drum" Issue

For years, engineers have used tiny acoustic resonators (think of them as microscopic tuning forks or drums) to filter radio signals in our phones. These work great for lower frequencies (like standard 4G).

But as we move to millimeter waves (the super-fast 5G/6G frequencies above 30 GHz), things get tricky.

The Analogy: Imagine trying to make a drum vibrate so fast that it's screaming. If the drum skin is too thick or the frame is too heavy, the vibration dies out instantly.
The Reality: Previous designs were either too big (wasting space on the chip) or they lost too much energy (low "Quality" or Q-factor) because the metal parts were too heavy, dragging the vibration down.

2. The Solution: The "Lightweight, High-Pitched" Drum

The team at the University of Texas at Austin decided to change the design of the drum itself. Instead of using the standard "first note" (a low-pitched vibration), they decided to play the third harmonic (a much higher, sharper note).

The Material: They used a super-thin sheet of Lithium Niobate (a special crystal that turns electricity into vibration).
The Trick: They made the "drum skin" incredibly thin (about 1/100th the width of a human hair) and the "drum" itself very small.
The Result: By vibrating at this higher "third-order" mode, they could make the device much smaller while keeping it efficient. It's like switching from a heavy bass drum to a tiny, high-pitched snare drum that can vibrate 40 billion times a second.

3. The "Goldilocks" Design

To make this work, they had to get the details just right:

The Electrodes (The Hammers): They used aluminum electrodes to hit the crystal. But if the aluminum was too thick, it would weigh the drum down. They shaved it down to just 50 nanometers (thinner than a virus). This kept the "drum" light and fast.
The Size: They shrunk the device to the size of a grain of sand (roughly 32 by 44 micrometers). This is so small that hundreds of them could fit on the head of a pin.

4. The Performance: A High-Performance Athlete

How well did it work?

Speed: They tested it at frequencies between 30 GHz and 50 GHz. That's fast enough to handle the data demands of future 6G networks.
Efficiency (The "Q" Factor): In the world of resonators, "Q" is like the ring time of a bell. A low Q means the bell goes thud and stops. A high Q means it goes ding! and rings for a long time.
- Their device had a very high "ring time" (Quality factor) for its size.
- They achieved a "Figure of Merit" (a score combining speed and efficiency) of 13.8, which is a new record for this type of single-layer device.

5. Why This Matters

Think of your phone's radio as a busy highway.

Old filters are like wide, slow lanes that get clogged easily.
This new technology creates a high-speed, dedicated express lane that is tiny enough to fit in your pocket but powerful enough to handle massive amounts of data without crashing.

The Bottom Line

The authors took a complex physics problem (how to make tiny things vibrate fast without losing energy) and solved it by:

Going higher: Using a higher vibration mode (A3 mode).
Going smaller: Shrinking the footprint.
Going lighter: Using ultra-thin metal electrodes.

They proved that you can build a tiny, high-performance acoustic resonator that works perfectly in the "millimeter-wave" zone, paving the way for faster, more efficient 6G communication devices in the future.

1. Problem Statement

The rapid expansion of 5G and the transition toward 6G communication systems require compact front-end signal processing components operating in the millimeter-wave (mmWave) spectrum (above 30 GHz). While Surface Acoustic Wave (SAW) and Bulk Acoustic Wave (BAW) devices based on piezoelectric materials (like AlN, ScAlN, and Lithium Niobate) are successful below 6 GHz, scaling these technologies to mmWave frequencies faces intrinsic limitations:

Solidly Mounted Resonators: Achieve high phase velocities but require expensive substrates (e.g., SiC, diamond) and sub-100 nm feature patterning.
Film Bulk Acoustic Wave Resonators (FBAR): Face challenges with moderate Quality Factors ( $Q$ ) due to the large metal-to-piezoelectric volume ratio required at thin film thicknesses.
Existing XBAR (Laterally Excited Bulk Acoustic Wave) Designs: Recent single-layer Lithium Niobate (LN) XBARs using first-order antisymmetric ( $A_1$ ) modes have demonstrated high electromechanical coupling ( $k_t^2$ ) but suffer from moderate $Q$ factors, resulting in a suboptimal Figure of Merit (FoM = $Q \cdot k_t^2$ ).
Periodically Poled (P3F) Solutions: While P3F structures have achieved high $Q$ , they introduce complex fabrication challenges regarding precise thickness control of multiple layers.

The core challenge is to develop a single-layer mmWave acoustic resonator platform that achieves both high $Q$ and high $k_t^2$ with a compact footprint, without relying on complex multi-layer poling or expensive substrates.

2. Methodology

The authors propose a design strategy utilizing higher-order antisymmetric ( $A_3$ ) modes in a single-layer thin-film Lithium Niobate (LN) platform.

Device Architecture:
- Stack: A suspended structure consisting of a 128Y-cut LN film (110–160 nm thick) on an amorphous silicon/sapphire stack.
- Mode Selection: Instead of the conventional $A_1$ mode, the design targets the third-order antisymmetric ( $A_3$ ) Lamb mode. This allows for operation at higher frequencies with smaller lateral wavelengths.
- Lateral Wavelength ( $\lambda_{Lateral}$ ): The design reduces $\lambda_{Lateral}$ to 3.15 $\mu$ m and 4 $\mu$ m (compared to larger wavelengths in prior $A_1$ designs). This reduction enables a compact footprint and minimizes surface losses.
- Electrode Optimization: Aluminum (Al) electrodes were selected for low mass-loading. Finite Element Analysis (FEA) using COMSOL was used to optimize electrode thickness. A 50 nm Al thickness was chosen to minimize mechanical damping losses, which disproportionately affect the series resonance quality factor ( $Q_s$ ).
- Aperture: A minimized device aperture of 8 $\lambda_{Lateral}$ was selected to reduce routing resistance.
Fabrication:
- The process follows standard suspended LN fabrication flows (similar to prior work [31]).
- Devices were fabricated with varying LN film thicknesses (110–160 nm) to demonstrate frequency scalability across the 30–50 GHz range while keeping the lateral layout constant.
Characterization:
- Devices were measured using a Keysight Vector Network Analyzer (VNA) calibrated with a GGB CS-5 substrate.
- Data was fitted using a modified Butterworth-Van-Dyke (mBVD) model including series routing resistance ( $R_s$ ) and inductance ( $L_s$ ) to extract accurate $k_t^2$ and $Q$ values.

3. Key Contributions

Higher-Order Mode Utilization: The paper successfully demonstrates that the $A_3$ mode in single-layer LN is a viable and superior alternative to $A_1$ for mmWave applications, offering a better trade-off between footprint and performance.
Compact Footprint: By reducing the lateral wavelength and aperture, the authors achieved a significantly smaller device size (e.g., 32 $\times$ 44 $\mu$ m $^2$ ) compared to previous mmWave resonators.
Loss Mechanism Mitigation: The study highlights the impact of electrode mechanical damping on $Q_s$ . By thinning the electrodes to 50 nm, they significantly reduced these losses, a factor often underappreciated in series resonance analysis.
Frequency Scalability: The work validates a design that can cover a wide mmWave spectrum (30–50 GHz) simply by varying the piezoelectric film thickness, without redesigning the lateral layout.

4. Key Results

The experimental results demonstrate state-of-the-art performance for single-layer LN resonators:

Peak Performance Device (39.83 GHz):
- Frequency ( $f_s$ ): 39.83 GHz
- Series Resonance $Q$ ( $Q_s$ ): 97
- Parallel Resonance $Q$ ( $Q_p$ ): 342
- Electromechanical Coupling ( $k_t^2$ ): 4% (extracted)
- Figure of Merit (FoM): 13.8 ( $Q_p \cdot k_t^2$ )
- Footprint: 32 $\times$ 44 $\mu$ m $^2$
Broadband Performance (30–50 GHz):
- Across the tested frequency span, $Q_s$ ranged from 33 to 102, and $Q_p$ ranged from 210 to 464.
- The FoM ranged from 8 to 22.3.
- The $f \cdot Q$ product reached up to $1.4 \times 10^{13}$ , significantly outperforming prior single-layer LN XBARs and competing with more complex P3F structures.
Comparison: The proposed devices exceed the state-of-the-art in terms of the $f \cdot Q$ product while maintaining a normalized footprint that allows for 50 $\Omega$ impedance matching at series resonance.

5. Significance

This work represents a critical step forward in mmWave acoustic resonator technology:

Enabling 6G Front-Ends: It provides a pathway to realizing high-performance, miniature filters and resonators for 6G communication systems operating above 30 GHz, where current technologies struggle with size and loss.
Simplification of Fabrication: By achieving high $Q$ with a single-layer LN structure (avoiding the complex P3F process), the technology is more compatible with standard semiconductor manufacturing processes, potentially lowering costs and improving yield.
Design Paradigm Shift: The successful transition to higher-order modes ( $A_3$ ) with reduced lateral wavelengths challenges the previous reliance on $A_1$ modes for XBARs, opening new design spaces for ultra-compact, high-frequency acoustic devices.
Loss Understanding: The paper clarifies the role of electrode mechanical damping in limiting $Q_s$ , providing a clear roadmap for future optimizations (e.g., electrode material selection, annealing, and geometry tuning).

In conclusion, Chulukhadze et al. have demonstrated that single-layer Lithium Niobate XBARs utilizing higher-order antisymmetric modes can achieve high $Q$ and high coupling at mmWave frequencies, offering a scalable and manufacturable solution for next-generation wireless communication.