🔬 optics

Monolithic Barium Titanate Nanophotonics and Electro-optics

This paper demonstrates the design, fabrication, and characterization of high-performance monolithic barium titanate-on-insulator nanophotonic devices, achieving low-loss propagation, high-quality resonators, and an 11 GHz electro-optic bandwidth that enables diverse modulation effects including single sideband modulation and frequency comb generation.

Original authors: Sarah Berman, Sina Dereshgi, David Barton

Published 2026-07-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Sarah Berman, Sina Dereshgi, David Barton

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 super-fast, tiny switch for light, similar to how a light switch controls electricity in your home. But instead of electricity, this switch controls beams of light for computers and communication. The challenge is making this switch small enough to fit on a computer chip, fast enough to handle massive amounts of data, and efficient enough not to waste energy.

This paper describes a breakthrough in building these light switches using a special material called Barium Titanate.

Here is the story of what the researchers did, explained simply:

1. The Problem: The "Heavy" Switch

Most current light switches are made of materials that are either too big, too slow, or require too much power to flip. The researchers wanted to use Barium Titanate because it is like a "super-responsive" material—it reacts very strongly to electricity, meaning you can flip the light switch with very little effort.

However, there was a catch. To make these switches tiny (nanoscale), you have to carve them out of the material. Previous attempts to carve this material resulted in rough, jagged edges, like trying to carve a statue out of sand. These rough edges caused the light to scatter and get lost, ruining the switch's performance.

2. The Solution: The "Smooth Carving" Technique

The team at Northwestern University developed a new way to carve this material. Think of it like switching from a dull, chipping knife to a laser-guided scalpel.

The Process: They used a special "dry etching" technique (using gases instead of liquids) to carve the material.
The Result: They created incredibly smooth walls with a perfect angle (75 degrees). This is like carving a perfect, smooth hallway for light to run through without tripping over bumps.
The Proof: They tested this by building tiny "race tracks" for light (resonators). The light ran around these tracks with almost no loss, proving the carving was clean and precise.

3. The Masterpiece: The "Light Trap" (Photonic Crystals)

Once they could carve smoothly, they built something even more complex: Photonic Crystals.

The Analogy: Imagine a hallway lined with mirrors. If you arrange the mirrors just right, they can trap light in a specific spot or block it from passing through entirely.
What they did: They carved a pattern of tiny holes (like a honeycomb) into the Barium Titanate. This pattern acts as a "traffic cop" for light. It can stop light of certain colors from passing through, creating a "bandgap" (a wall of light).
The Achievement: They built these walls so effectively that they blocked light with over 40 decibels of contrast (a huge difference between "on" and "off"). They also built tiny "cages" (cavities) where light could bounce around thousands of times before escaping, proving the material was high-quality.

4. The Magic Trick: Making Light Dance

The ultimate goal is to use electricity to change the light.

The Test: They applied a small voltage to their device. Because Barium Titanate is "ferroelectric" (it has tiny internal magnets that can be aligned), the electricity lined up these internal magnets.
The Result: This alignment changed how the light moved. They measured that the material was extremely efficient at this, changing the light's path with a very strong "kick" (an effective coefficient of ~154 pm/V). This is much stronger than other common materials used in chips.

5. The Speed Test: The "Highway"

Finally, they tested how fast this switch could work.

The Setup: They sent microwave signals (like radio waves) into the device to see how fast it could flip the light on and off.
The Result: The switch worked at speeds up to 11 GHz (11 billion times per second) for a standard signal, and even up to 21 GHz for a slightly weaker signal.
Why it matters: This speed is determined by the material itself, not by the size of the device. This means they can make the device smaller without losing speed. They also showed that by using the "edges" of their light-trapping patterns, they could create unique effects like generating new colors of light or creating "frequency combs" (a ruler made of light frequencies).

Summary

In short, the researchers figured out how to carve a special crystal (Barium Titanate) so smoothly that they could build tiny, high-speed light traps on a chip. They proved these traps can hold light efficiently, switch it on and off incredibly fast using very little power, and even create complex light patterns. This paves the way for future computers and communication systems that are faster, smaller, and use less energy.

Technical Summary: Monolithic Barium Titanate Nanophotonics and Electro-optics

Problem Statement
Electro-optic (EO) integrated photonic devices are critical for low-energy, broadband optical signal modulation in communications, microwave photonics, and quantum information. While the Pockels effect in non-centrosymmetric materials offers broadband modulation without the loss associated with carrier injection, achieving high-speed (>10 GHz), low-drive-voltage (<1 V), and compact modulators in CMOS-compatible formats remains challenging. Barium Titanate (BTO) is a promising candidate due to its large bulk Pockels coefficient ( $r_{42} > 1200$ pm/V) and CMOS compatibility. However, existing BTO-on-insulator (BTOI) devices often rely on ridge waveguides where the optical mode is primarily localized in a passive cladding (SiN or Si), limiting the effective EO interaction. Furthermore, while photonic crystals (PhCs) can enhance light-matter interaction and reduce device capacitance, fabricating low-loss, monolithic sub-wavelength nanostructures in BTOI has been difficult due to challenges in achieving anisotropic etching and high-fidelity pattern transfer without excessive scattering loss.

Methodology
The authors developed a nanofabrication process to create monolithic one-dimensional photonic crystals and Fabry–Pérot cavities directly in thin-film BTOI.

Fabrication: Starting with 330-nm thick BTO-on-insulator films, the team utilized electron beam lithography to define features (~~100 nm) and inductively coupled plasma reactive ion etching (using Ar $^+$ and CHF $_3$ gases) to achieve anisotropic sidewalls (~~75°) and low surface roughness. A multi-step cleaning and annealing process was employed to remove redeposition and improve quality factors.
Device Design: The team designed single-mode waveguides, ring/racetrack resonators, and 1-D PhCs with tunable unit cell dimensions (elliptical holes) to engineer bandgaps. Fabry–Pérot cavities were formed by etching PhC mirrors on either side of a straight waveguide.
Characterization: Devices were characterized using transmission spectroscopy to extract quality factors ( $Q$ ) and propagation losses. Ferroelectric domain alignment was verified via confocal second-harmonic generation (SHG) microscopy. DC electro-optic tuning was measured by monitoring resonance shifts under bias. AC modulation performance was assessed using a vector network analyzer (VNA) and RF signal generators to measure bandwidth, sideband generation, and frequency comb formation.

Key Results

Fabrication Quality: The process yielded anisotropic sidewalls (~75°) and low optical loss. Racetrack resonators demonstrated intrinsic quality factors ( $Q_i$ ) exceeding 1 million and propagation losses of approximately 0.45 ± 0.077 dB/cm.
Photonic Crystals: The fabricated 1-D PhCs exhibited geometry-tunable bandgaps with contrasts greater than 40 dB. The bandgap center and width could be adjusted by varying the lattice period and hole dimensions.
Fabry–Pérot Cavities: By utilizing PhC mirrors, the authors constructed cavities with loaded quality factors ( $Q_L$ ) up to 230k. The $Q$ -factor scaled systematically with mirror reflectivity and cavity length, confirming that performance was limited by mirror design rather than scattering loss.
Electro-Optic Performance:
- DC Tuning: After poling (200 V across a 6 µm gap for 10 minutes), the devices exhibited a stable linear modulation range. The effective Pockels coefficient ( $r_{eff}$ ) was extracted to be ~154 pm/V for racetrack resonators and ~145 pm/V for Fabry–Pérot cavities.
- AC Bandwidth: Probing the microwave response at the PhC band edge (where bandwidth is set by material response rather than cavity lifetime), the team measured a 3-dB electro-optic bandwidth of 11 GHz and a 6-dB bandwidth of 21 GHz. This roll-off aligns with the frequency-dependent behavior of BTO's $r_{42}$ coefficient near 10 GHz.
- Modulation Effects: The study demonstrated various modulation regimes, including asymmetric sideband generation and quasi-single sideband modulation at PhC band edges due to dispersion. Under high-power RF driving, asymmetric frequency combs were generated. In resonant cavities, the authors observed transitions between sideband-resolved and sideband-unresolved regimes depending on the relationship between the modulation frequency and the cavity linewidth.

Significance and Claims
The paper claims to demonstrate a viable route to high-performance, sub-wavelength nanostructures in monolithic BTOI, overcoming previous limitations in low-loss etching. By co-locating the electro-optic material and the optical mode, the authors achieved compact devices with high tuning efficiency and low propagation loss. The work establishes that monolithic BTOI devices can potentially outperform lithium niobate-on-insulator (LNOI) in terms of energy consumption and size. The demonstrated low-loss, high-contrast nanostructures open pathways for visible-wavelength EO devices, dispersion-engineered cavities for nonlinear optics, and compact quantum photonic components. Additionally, the ability to reprogram ferroelectric domains in these nanostructures suggests potential for non-volatile memory and reconfigurable photonic circuits, though the authors note that further improvements in growth and strain engineering could further enhance EO strength and bandwidth.