🔬 optics

Fast and wide-range wavelength tuning of a III-V/Si3N4 external-cavity laser via two-step pulsed heating

This paper presents a hybrid III-V/Si3N4 external-cavity laser that achieves fast, wide-range wavelength tuning (up to ~100 nm) with rates exceeding 8.9 nm/µs by employing a two-step pulsed heating strategy to overcome the speed limitations of traditional thermo-optic tuning.

Original authors: Cong Wang, Fuyi Cao, Xin Xu, Yihan Qi, Dongxin Jiang, Masataka Kobayashi, To-Fan Pan, Zhan Su, Guoen Weng, Hidefumi Akiyama, Shaoqiang Chen

Published 2026-08-14

📖 3 min read☕ Coffee break read

CC BY 4.0

Original authors: Cong Wang, Fuyi Cao, Xin Xu, Yihan Qi, Dongxin Jiang, Masataka Kobayashi, To-Fan Pan, Zhan Su, Guoen Weng, Hidefumi Akiyama, Shaoqiang Chen

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 the internet as a massive, bustling highway where data travels as pulses of light. To keep traffic flowing smoothly, engineers use a clever trick called "wavelength-division multiplexing," which is like having multiple lanes on that highway, each carrying a different color of light. The faster and more agile these light signals can switch lanes, the more data we can send. However, the "traffic cops" that steer these lasers—tiny devices that change the light's color—have a frustrating habit: they are often slow. They rely on heat to work, and just like a heavy metal pan takes time to cool down after being on a stove, these devices struggle to change their mind quickly. This paper tackles that sluggishness, asking a simple question: Can we make these heat-steered lasers change colors as fast as a camera flash, without breaking them?

The researchers behind this study built a special kind of laser that acts like a hybrid vehicle, combining a light-generating engine with a highly efficient, heat-sensitive steering wheel made of silicon nitride. To steer the laser, they use a "Vernier filter," which is a bit like a pair of slightly mismatched gears. When you turn one gear, the teeth don't line up perfectly, creating a unique pattern that only allows one specific color of light to pass through. By heating up tiny parts of these gears, the researchers can shift the pattern, effectively changing the laser's color. The problem is that heating and cooling these gears usually takes time, limiting how fast the laser can switch from one color to another.

In this paper, the team discovered a way to overcome the laws of thermal sluggishness using a "two-step" heating strategy. Instead of gently warming the gears to their target temperature, they blast them with a short, high-power burst of heat—like hitting the gas pedal hard for a split second—before settling into a lower, steady power to hold the new color. They found that this "kick-start" method allows the laser to switch colors incredibly fast. Specifically, they managed to shift the laser's color by 101 nanometers (a red-shift) in just 11.33 microseconds, and by 104 nanometers (a blue-shift) in 10.74 microseconds. That translates to a tuning speed of up to 9.68 nanometers per microsecond.

The paper explicitly rules out the idea that simply waiting for the heat to settle naturally is the best way to go; the data shows that the slow, steady approach is too sluggish for modern needs. They also argue against the notion that you need to completely redesign the physical chip to get these speeds; instead, they show that a clever change in how you apply the power is enough. The authors are quite sure of these results, having measured them repeatedly over two-hour sessions with no damage to the device, though they do note a trade-off: pushing the voltage too high or the pulses too short risks frying the tiny heaters. By using this two-step pulse method, they demonstrated that thermo-optic lasers can be made fast enough for next-generation communication systems and LiDAR, proving that sometimes, to move fast, you just need to hit the gas hard for a moment before cruising.

Technical Summary: Fast and Wide-Range Wavelength Tuning of a III–V/Si3N4 External-Cavity Laser via Two-Step Pulsed Heating

Problem Statement
Rapid, wide-range wavelength switching is critical for frequency-agile optical systems, including telecommunication transceivers, optical interconnects, and LiDAR. While hybrid-integrated external-cavity lasers (ECLs) based on silicon nitride (Si3N4) offer wide tuning ranges (>100 nm) and low propagation loss, their switching speeds are typically limited to the kilohertz order by thermodynamic constraints. Traditional thermo-optic (TO) tuning relies on heating microheaters to shift resonance wavelengths; however, this process is hindered by the small TO coefficient of Si3N4, slow thermal convergence near equilibrium, and long heat diffusion paths caused by thick cladding layers. Existing solutions often trade off tuning range for speed, or vice versa, and high-power two-step pulse heating strategies have not yet been successfully integrated into hybrid-integrated ECLs to overcome these thermal limitations.

Methodology
The authors developed a hybrid-integrated III–V/Si3N4 ECL featuring a reflective semiconductor optical amplifier (RSOA) coupled to a Si3N4 waveguide containing a dual-microring resonator (MRR) Vernier filter. The device utilizes two MRRs with slightly different radii (FSRs of 1.96 nm and 1.99 nm) to generate a Vernier envelope with a free spectral range (FSR) of approximately 130 nm, covering the S, C, and part of the L bands.

To achieve fast switching, the study employs a "two-step pulsed thermal pumping" strategy. This approach exploits the initial quasi-linear heating transient of the microheaters. Instead of applying a constant voltage to reach a target temperature (which follows a slow exponential curve), the method applies a short, high-amplitude pre-pump pulse to rapidly drive the system through the linear heating phase, followed immediately by a lower-amplitude main pulse to sustain the target wavelength at thermal equilibrium. The device was characterized using a 2 kHz pulse repetition rate, with switching times defined as the duration required for the output to transition between 90% of the initial and final steady-state levels.

Key Contributions

Integration of Two-Step Heating in ECLs: The study successfully adapts the two-step pulse heating technique, previously used in passive filters, to a hybrid-integrated active ECL system.
Structure-Agnostic Approach: The method relies on waveform control rather than complex structural modifications, offering a practical route to speed up TO-tuned devices without altering the waveguide geometry.
Predictive Modeling: The authors demonstrate that the TO-tuning dynamics can be accurately reproduced using a simple first-order thermal model, allowing for precise prediction of switching behavior and optimization of pulse parameters.

Results
The device achieved stable single-mode lasing across a static tuning range of 1486–1614 nm (127 nm total) with an output power >10 mW and a side-mode suppression ratio (SMSR) exceeding 60 dB.

Static Performance: The thermal tuning coefficients were measured at $k_1 = 0.163$ nm/mW (red-shift) and $k_2 = -0.155$ nm/mW (blue-shift).
Standard Pulsed Switching: Under standard single-pulse heating (25 V, 25 µs), the laser achieved a switching range of 81 nm with a tuning rate of 3.06 nm/µs.
Two-Step Pulsed Switching: By applying a high-voltage pre-pump (40 V, 10 µs) followed by a hold pulse, the switching performance was significantly enhanced:
- Red-Shift: A 101 nm wavelength shift was achieved in 11.33 µs, yielding a quasi-linear tuning rate of 8.91 nm/µs.
- Blue-Shift: A 104 nm wavelength shift was achieved in 10.74 µs, yielding a quasi-linear tuning rate of 9.68 nm/µs.
Thermal Dynamics: The first-order thermal fit showed excellent agreement with experimental data. The extracted thermal time constants ( $\tau$ ) increased with pump amplitude during heating but remained relatively constant during cooling, indicating that the high-power pre-pump effectively utilizes the initial linear transient before thermal inertia slows the response.

Significance and Claims
The paper claims that this two-step pulsed thermal pumping strategy provides a practical and effective route to overcoming the inherent speed limitations of thermo-optic tuning in Si3N4-based ECLs. By leveraging the high initial slope of the thermal transient, the method achieves wide-range switching (over 100 nm) with rates approaching 10 nm/µs, which is significantly faster than previously reported TO-tuned Si3N4 ECLs.

The authors emphasize that this approach does not reduce the intrinsic thermal time constant but rather accelerates the transition by operating in the quasi-linear regime of the thermal response. While the method introduces a trade-off between switching speed, tuning range, and device reliability (due to high peak power and thermal stress), it successfully demonstrates that TO-tuned ECLs can be made frequency-agile enough for applications such as wavelength-division multiplexed (WDM) transceivers and LiDAR systems. The study concludes that this structure-agnostic pumping scheme is a viable path toward extending the applicability of hybrid-integrated lasers in high-speed photonic systems.

More like this