🔬 optics

Toward scalable and bias-stable optical phased arrays on lithium tantalate

This paper demonstrates that lithium tantalate optical phased arrays overcome the critical phase drift bottleneck of ferroelectric photonic circuits by leveraging intrinsically low carrier drift, achieving unprecedented bias stability and scalability for diverse applications ranging from quantum computing to LiDAR.

Original authors: Gongcheng Yue, Xuqiang Wang, Yihan Miao, Bowen Chen, Yangming Zhan, Weiran Zhou, Phatham loahavilai, Jiachen Cai, Siyuan Yu, Chengli Wang, Xin Ou, Yang Li

Published 2026-03-25

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Gongcheng Yue, Xuqiang Wang, Yihan Miao, Bowen Chen, Yangming Zhan, Weiran Zhou, Phatham loahavilai, Jiachen Cai, Siyuan Yu, Chengli Wang, Xin Ou, Yang Li

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

The Big Picture: The "Drifting" Problem

Imagine you are trying to conduct a massive orchestra of 16 musicians (these are the Optical Phased Arrays, or OPAs). Your goal is to get them all to play the exact same note at the exact same time so that the sound waves combine to create one incredibly loud, focused beam of sound (or light) in a specific direction.

In the world of high-tech optics, we use special crystals (called ferroelectric materials) to control these "musicians." We apply a voltage to tell them when to play. Ideally, if you set the voltage once, the crystal should hold that setting perfectly forever.

But here's the problem: In most of these crystals (like the popular Lithium Niobate), there are tiny, invisible "ghosts" called electrons or carriers. When you apply a voltage, these ghosts start to wander around (drift) and get stuck in the wrong places.

Think of it like this: You set a thermostat to 70°F. But because the house has a leak, cold air slowly seeps in, and the heater has to work harder and harder just to stay at 70°F. Eventually, the temperature drifts away from your setting. In our optical orchestra, this "leak" causes the musicians to fall out of sync. The beautiful, focused beam of light starts to blur, scatter, and turn into a messy mess. This is called phase drift.

The Solution: Finding the "Stable" Crystal

The researchers in this paper asked: "Is there a crystal where these 'ghosts' don't wander as much?"

They found the answer in Lithium Tantalate (LT).

The Analogy: Imagine Lithium Niobate (the old standard) is a muddy field. If you try to walk across it, your feet get stuck, and you drag mud everywhere, creating a mess. Lithium Tantalate is like a smooth, polished marble floor. The "ghosts" (electrons) can't get stuck or wander easily. They stay put.

By using Lithium Tantalate, the researchers built an optical device that stays perfectly in tune for a very long time.

The Two Versions: The "Glass House" vs. The "Open Air"

The team built two versions of their device to test just how stable this new material is.

1. The Cladded Version (The Glass House)
First, they wrapped the crystal in a layer of glass (silicon dioxide).

The Metaphor: Even though the crystal floor is smooth, the glass walls around it had some sticky spots. The "ghosts" could still get stuck on the walls.
The Result: The device stayed in tune for about 16 minutes. That's good, but not great for long-term applications.

2. The Cladding-Free Version (The Open Air)
Then, they removed the glass wrapping entirely, leaving the crystal exposed.

The Metaphor: They took away the sticky walls. Now, the "ghosts" had nowhere to get stuck. The crystal was in its purest, most stable state.
The Result: The device stayed perfectly in tune for over 4 hours.
Why this matters: Previous state-of-the-art devices (using the older material) could only hold their focus for a few minutes. This new device is 100 times (two orders of magnitude) more stable. It's like the difference between a candle that flickers out in a minute and a lighthouse beam that shines steadily for hours.

What Can We Do With This?

Because this device is so stable, it can do things that were previously impossible or very difficult:

The "Slow-Motion" Laser: Usually, lasers move fast. But this device is so stable it can move a beam of light incredibly slowly—like a snail (0.1 Hz). Imagine a laser pointer that can hold a spot perfectly still on a wall for 10 seconds, then slowly trace a triangle, then a square, then a circle.
Real-World Superpowers:
- Optical Tweezers: Using light to grab and move tiny cells or atoms without touching them.
- LiDAR & Self-Driving Cars: Scanning the environment with a laser that doesn't get blurry over time.
- Space Communications: Sending laser signals to satellites or other planets without the beam spreading out and getting lost.
- Quantum Computers: Trapping ions (charged atoms) to build super-fast computers.

The Bottom Line

This paper is a breakthrough because it solved a fundamental "leak" in the technology. By switching to a better crystal (Lithium Tantalate) and removing the sticky layers around it, the researchers created a "super-stable" optical brain.

This means we can now build much larger, more complex optical systems that don't fall apart after a few minutes. It opens the door to a future where our computers, sensors, and communication systems are faster, more reliable, and capable of handling massive amounts of data using light instead of electricity.

1. Problem Statement

The paper addresses a fundamental bottleneck in Ferroelectric Photonic Integrated Circuits (PICs): DC phase drift (also known as carrier drift).

The Mechanism: In ferroelectric waveguides (e.g., Lithium Niobate, LN), defects like oxygen vacancies and impurities generate free carriers. Under a static bias voltage, these carriers drift and redistribute, creating an internal electric field that opposes the applied field. This reduces the total effective electric field, causing a time-dependent change in the refractive index and, consequently, a continuous phase drift at the output.
The Consequence: Most reconfigurable PICs (such as Optical Phased Arrays or OPAs) rely on precise interference. Phase drift disrupts this interference, causing beam diffusion and loss of focus.
Scalability Issue: As OPA arrays scale up (more channels), the time required to optimize the phase distribution exceeds the relaxation time of the phase drift. This makes it impossible to achieve or maintain beam focusing in large-scale systems, severely limiting the scalability of ferroelectric PICs for applications like LiDAR, quantum computing, and free-space communications.

2. Methodology

The authors propose a material and structural solution to suppress phase drift by leveraging the intrinsic properties of Lithium Tantalate (LT) and optimizing the device architecture.

Material Selection (LT vs. LN):
- The authors chose LT over the industry-standard Lithium Niobate (LN).
- Physics: LT has a wider bandgap (reducing photon-induced carrier generation) and different defect dynamics (anti-site tantalum ions have different trapping/releasing dynamics compared to anti-site niobium ions in LN). This results in significantly slower carrier accumulation and a weaker opposing internal electric field.
Structural Optimization (Cladding-Free Design):
- The authors identified that the interface between the ferroelectric waveguide and the standard Silicon Dioxide ( $SiO_2$ ) cladding introduces additional donor and trapping energy levels, facilitating carrier migration.
- Solution: They fabricated two versions of the device:
  1. $SiO_2$ -Cladded LT OPA: Standard fabrication with a top cladding layer.
  2. Cladding-Free LT OPA: The top $SiO_2$ layer was removed to minimize carrier trapping at the interface, thereby weakening the build-up of the internal electric field.
Device Fabrication:
- Fabricated on an x-cut TFLT-on-insulator (Tantalate Thin Film on Insulator) wafer using a standard planar process.
- Structure: A 16-channel integrated OPA with a 600-nm LT membrane, cascaded multimode interferometers for splitting, 1.5-cm long phase modulators, and a slab grating for beam emission.
- Electrodes: Dual-layer lift-off process with Au/Ti electrodes.

3. Key Contributions

Demonstration of Intrinsic Stability: The paper experimentally proves that LT possesses intrinsically lower DC drift compared to LN, making it a superior platform for bias-stable PICs.
Architectural Breakthrough: The introduction of a cladding-free design significantly mitigates interface-induced carrier trapping, pushing the stability limits of ferroelectric OPAs.
Scalability Validation: Through numerical simulations and experimental data, the authors demonstrate that suppressing phase drift is a prerequisite for scaling ferroelectric OPAs to large channel counts (e.g., >120 channels) where optimization times would otherwise exceed drift relaxation times.
Ultra-Low Frequency Operation: The high stability enables the generation of arbitrary spatial-temporal waveforms with modulation frequencies as low as 0.1 Hz, a regime previously inaccessible for stable ferroelectric beam steering.

4. Key Results

Stability Performance:
- Cladded LT OPA: Maintained a far-field mainlobe 8 dB higher than sidelobes for ~16 minutes.
- Cladding-Free LT OPA: Maintained the same performance (8 dB mainlobe-to-sidelobe ratio) for over 4 hours.
- Comparison: This represents an improvement of at least two orders of magnitude compared to state-of-the-art integrated LN OPAs (which typically stabilize for only a few minutes).
Beam Steering & Control:
- Achieved beam steering over a 40° angular range ( $\theta_x$ ) via phase modulation and a 9° range ( $\theta_y$ ) via wavelength tuning.
- Successfully generated arbitrary waveforms (triangular, square, sinusoidal) with temporal periods up to 10 seconds (0.1 Hz).
Power Independence: The stability was observed to be largely independent of optical power (tested at 14 dBm, 0 dBm, and -14 dBm), confirming that the drift is primarily driven by carrier dynamics rather than photorefractive effects.
Device Metrics:
- Half-wave voltage ( $V_\pi$ ): 4.4 V.
- Beam width (FWHM): 2.6° (phase direction) and 1.6° (wavelength direction).

5. Significance and Impact

Enabling Large-Scale Systems: By solving the phase drift bottleneck, this work unlocks the potential for large-scale, high-performance ferroelectric PICs. This is critical for applications requiring massive channel counts, such as optical neural networks, quantum computing (trapped-ion control), and high-resolution LiDAR.
Broad Application Spectrum: The demonstrated stability enables practical deployment in diverse fields:
- Optical Tweezers & Trapped-Ion Quantum Computers: Requiring long-term, static beam trapping.
- Adaptive Optics & Astronomy: Needing stable wavefront correction.
- AR/VR & 3D Printing: Requiring precise, stable beam positioning.
- Free-Space Optical Communications: Demanding reliable beam pointing.
Future Roadmap: The paper establishes a baseline for "pristine" phase stability. Future work can focus on integrating this LT platform with advanced low-drift passivation materials and heterogeneous integration (combining with lasers and detectors) to create fully monolithic, bias-stable photonic computing and sensing systems.

In summary, this paper shifts the paradigm for ferroelectric photonics from managing inevitable drift to exploiting material properties (LT) and structural engineering (cladding-free) to achieve bias-stable, scalable optical systems.