🔬 optics

High-Efficiency Acousto-Optic Modulation on Non-Suspended Thin-Film Lithium Tantalate

This paper establishes lithium tantalate on insulator (LTOI) as a scalable, non-suspended platform for integrated acousto-optics by demonstrating record-low modulation efficiency in Mach-Zehnder interferometers and racetrack resonators, thereby enabling robust microwave-to-optical conversion for communications and quantum technologies.

Original authors: Weiran Zhou, Chengli Wang, Xuqiang Wang, Bowen Chen, Jiachen Cai, Tianyao Yang, Dongchen Sui, Xinjian Ke, Yang Chen, Xudong Wang, Ailun Yi, Shibin Zhang, Chengjie Zuo, Xin Ou

Published 2026-04-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Weiran Zhou, Chengli Wang, Xuqiang Wang, Bowen Chen, Jiachen Cai, Tianyao Yang, Dongchen Sui, Xinjian Ke, Yang Chen, Xudong Wang, Ailun Yi, Shibin Zhang, Chengjie Zuo, Xin Ou

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 have a super-fast internet cable (light) and a standard radio signal (microwaves). Usually, these two speak completely different languages and can't really talk to each other without a bulky, expensive translator. This paper introduces a new, tiny, and incredibly efficient "translator" chip that lets them chat seamlessly.

Here is the story of how they did it, broken down into simple concepts:

1. The Problem: The "Glass" vs. The "Metal"

In the world of tiny computer chips (photonics), we use Lithium Niobate (a special crystal) to make things that control light. It's like the "gold standard" material. However, it has a flaw: if you shine a really bright light on it, it gets confused (a phenomenon called the photorefractive effect) and stops working well. It's like a glass window that gets foggy when the sun hits it too hard.

The scientists in this paper decided to try a "cousin" crystal called Lithium Tantalate.

The Analogy: Think of Lithium Niobate as a high-performance sports car that breaks down if you drive it in the rain. Lithium Tantalate is the same sports car, but with a waterproof roof and a stronger engine. It handles heat and bright light much better, and it's already being mass-produced for 5G phones, making it cheap and easy to get.

2. The Magic Trick: The "Acoustic Wave"

How do you make light and microwaves talk? You use sound.

The Analogy: Imagine a long, thin rope (the light waveguide). If you shake one end of the rope, a wave travels down it. Now, imagine you have a tiny speaker (the microwave) that vibrates the rope.
In this chip, they send a microwave signal into the crystal. This creates a Sound Wave (specifically a Surface Acoustic Wave) that ripples across the surface of the chip. As this sound wave ripples, it physically squeezes and stretches the material, which changes how light travels through it.
The Result: The sound wave acts like a gatekeeper, turning the light on and off or changing its color, effectively translating the microwave signal into an optical one.

3. The Breakthrough: "Non-Suspended" and "Anisotropy"

Previous attempts at this were tricky. To make the sound waves strong enough, scientists often had to carve the chip into a floating bridge (suspended structure).

The Analogy: It's like trying to play a violin by suspending the strings in mid-air so they vibrate freely. It works, but the violin is fragile and hard to build in a factory.
This Paper's Solution: They kept the chip solid (non-suspended), like a sturdy floorboard. Usually, sound waves get lost in a solid floor. But these scientists discovered a secret: The crystal has a "sweet spot" direction.
The Analogy: Think of the crystal like a wooden floorboard. If you tap it one way, the sound dies out. If you tap it at a specific angle (like 90 degrees), the sound travels perfectly and bounces back and forth, getting louder and louder. They figured out exactly which angle to tap the crystal to make the sound waves super-efficient without needing to float the chip in the air.

4. The "Race Track" vs. The "Straight Road"

They built two types of devices to test this:

The Straight Road (Mach-Zehnder Interferometer): Light goes down two paths, and the sound wave pushes one path slightly more than the other. It works well.
The Race Track (Resonator): They bent the light into a tiny loop (a race track).
- The Analogy: Imagine a runner on a straight track vs. a runner on a circular track. On the circle, the runner passes the same point over and over again. If you push them every time they pass, they get faster and faster.
- By using this "race track" design, the light and sound interact many times in a tiny space. This made their device incredibly efficient. They achieved a record-breaking level of efficiency, meaning they needed very little power to control the light.

5. Why Should You Care?

This isn't just a lab experiment; it's a blueprint for the future of technology.

Faster Internet: It allows us to send data from microwave towers directly into fiber-optic cables with almost no loss.
Quantum Computers: It helps translate signals between the quantum world (which uses microwaves) and the internet (which uses light).
Stability: Because they didn't have to suspend the chip (make it float), these devices are tough, stable, and can be mass-produced in factories right now.

The Bottom Line

The team took a material that was already great at making phone filters (Lithium Tantalate), figured out the perfect angle to vibrate it, and built a tiny "race track" for light. The result is a super-efficient, sturdy, and cheap chip that can translate between radio waves and light waves better than almost anything else currently available, all without needing fragile, floating structures. It's like upgrading from a bicycle messenger to a high-speed bullet train for your data.

1. Problem Statement

Acousto-optic (AO) interactions are critical for bridging the microwave and optical domains, enabling functionalities like optical switching, non-reciprocal propagation, and efficient microwave-to-optical transduction. While Thin-Film Lithium Niobate (TFLN) has been the dominant platform for integrated AO devices due to its strong piezoelectric response, it suffers from limitations:

Stability Issues: TFLN is prone to photorefractive effects and thermal drift, especially under high optical intensities.
Manufacturing Scalability: While TFLN is promising, Lithium Tantalate (LT) offers a more mature industrial ecosystem with high-volume manufacturing capabilities (Piezoelectric-on-Insulator or POI), yet AO interactions on LT platforms had remained largely unexplored.
Structural Complexity: Many high-performance AO devices rely on suspended structures to confine acoustic waves and reduce loss. However, suspended structures compromise mechanical stability and power handling, making them less suitable for robust, large-scale integration.

The core challenge addressed by this work is achieving high-efficiency AO modulation on a non-suspended, scalable Lithium Tantalate on Insulator (LTOI) platform without sacrificing performance, while leveraging the material's intrinsic anisotropy.

2. Methodology

The researchers employed a systematic approach combining material engineering, device design, and experimental characterization:

Platform: Devices were fabricated on LTOI wafers consisting of a 600 nm X-cut Lithium Tantalate (LiTaO₃) thin film, a 4.7 µm buried oxide (BOX) layer, and a silicon substrate. This non-suspended structure ensures mechanical robustness.
Device Design:
- Mach-Zehnder Interferometers (MZI): Designed with a push-pull mechanism where the distance between arms corresponds to an odd multiple of half the acoustic wavelength to induce differential phase shifts.
- Racetrack Resonators: Utilized to enhance photon-phonon interaction via optical resonance.
- Acoustic Cavity: Interdigital Transducers (IDTs) were flanked by acoustic grating reflectors to form a Fabry-Perot acoustic cavity, confining acoustic energy near the optical waveguide.
Anisotropy Engineering: A key innovation was the systematic variation of the propagation angle ( $\theta$ ) of the Surface Acoustic Waves (SAW) relative to the crystal axes. The team utilized Finite Element Method (FEM) simulations to map the electromechanical coupling coefficient ( $k^2$ ) and acoustic quality factors ( $Q_m$ ) across different angles.
Fabrication: Standard processes including electron-beam lithography (EBL), ion-beam etching (IBE), and lift-off for Cr/Au electrodes were used.

3. Key Contributions

First Demonstration of AO on LTOI: This is the first report of acousto-optic modulation on the LTOI platform, establishing it as a viable alternative to TFLN.
Exploitation of Crystalline Anisotropy: The authors revealed a direct correlation between AO modulation efficiency and the electromechanical coupling coefficient ( $k^2$ ) of specific acoustic modes (Rayleigh $R_0, R_1$ and Shear-Horizontal $SH_0$ ). They demonstrated that exciting acoustic waves along the crystal Z-axis ( $\theta = 90^\circ$ ) significantly enhances the higher-order $R_1$ mode, yielding the highest efficiency.
Non-Suspended High Performance: The work proves that suspended structures are not strictly necessary for high-efficiency AO modulation. By optimizing the acoustic cavity and mode selection, they achieved record-breaking performance on a mechanically stable, non-suspended platform.
Multi-Mode Excitation: The platform supports the simultaneous excitation of multiple acoustic modes ( $R_0, R_1, SH_1$ ), enabling potential multi-band modulation and wavelength-division multiplexing.

4. Key Results

Modulation Efficiency (MZI): The Mach-Zehnder interferometers achieved a half-wave voltage-length product ( $V_\pi L$ ) of 0.68 V·cm.
Modulation Efficiency (Racetrack): The racetrack resonators achieved a record-low $V_\pi L$ of 0.022 V·cm (specifically 0.024 V·cm in initial tests, refined to 0.022 V·cm in the abstract). This is the lowest $V_\pi L$ reported to date for non-suspended ferroelectric platforms.
Acoustic Quality: The non-suspended LTOI structures demonstrated high acoustic quality factors ( $Q_m$ ), reaching up to 8,750 for the $R_0$ mode and 3,214 for the racetrack resonator, indicating low acoustic loss despite the lack of suspension.
Microwave-to-Optical Conversion: The devices successfully demonstrated efficient microwave-to-optical transduction. In a racetrack resonator driven at 1.44 GHz, the system generated clear optical sidebands, confirming high conversion efficiency without suspended structures.
Angle Dependence: The study confirmed that the $R_1$ mode at $\theta = 90^\circ$ provides the optimal overlap with the optical mode and the highest $k^2$ , validating the theoretical predictions regarding the anisotropy of X-cut LT.

5. Significance and Impact

Scalability and Robustness: By utilizing LTOI, which already has a mature high-volume manufacturing base (supplying millions of wafers annually for RF filters), this work paves the way for the mass production of integrated AO devices. The non-suspended design offers superior mechanical stability and power handling compared to suspended TFLN devices.
Overcoming Material Limitations: The use of Lithium Tantalate mitigates the photorefractive effects and thermal drift issues associated with Lithium Niobate, making the devices more stable for long-term operation.
Applications: The demonstrated high efficiency and low $V_\pi L$ $V_{π} L$ enable advanced applications in:
- Microwave Photonics: Low-loss long-haul signal transport and on-chip signal processing.
- Quantum Information: Efficient transduction between microwave qubits and optical photons.
- Communications: High-speed optical switching and beam steering.
Future Outlook: The paper suggests that further miniaturization (e.g., using ring resonators) and integration with high-acoustic-velocity materials could push performance even further, establishing LTOI as a versatile platform for next-generation microwave photonic systems.

In summary, this paper successfully bridges the gap between high-performance integrated photonics and industrial-scale manufacturing by demonstrating that non-suspended Lithium Tantalate can outperform or match suspended Lithium Niobate in AO modulation efficiency, provided that the material's anisotropic properties are correctly engineered.