Dynamical Control of Non-Hermitian Coupling Between Sub-Threshold Nanolasers Enables Q-Switched Pulse Generation

This paper demonstrates that dynamically tuning non-Hermitian coupling between two sub-threshold photonic crystal nanolasers via asymmetric optical pumping enables the generation of Q-switched optical pulses by controlling collective modal losses in an indium phosphide platform.

The Gist

Imagine you have two tiny, shy light bulbs (nanolasers) sitting next to each other. Individually, they are too weak to shine brightly on their own; they are "sub-threshold," meaning they just glow a little bit but never turn into a full beam of light. They are like two people whispering in a noisy room—they can't be heard clearly.

However, this paper describes a clever trick to make these two weak bulbs work together to shoot out a powerful, super-fast flash of light, like a camera flash or a laser pointer blinking incredibly fast.

Here is how they did it, explained through a simple story:

1. The Setup: Two Whispering Friends

Think of the two nanolasers as two friends standing on opposite sides of a hallway (a waveguide). They are connected by the air in the hallway.

• Friend A is being fed a steady, low-level snack (continuous light pump). They are full enough to whisper, but not loud enough to shout.
• Friend B is being fed snacks in quick, sudden bursts (pulsed light pump).

2. The Secret Sauce: "Non-Hermitian" Magic

In the world of physics, "Hermitian" usually means things are balanced and predictable. "Non-Hermitian" is a fancy way of saying the system is open, messy, and can gain or lose energy in tricky ways.

The researchers used a special kind of connection between the two friends. It's not just a physical wire; it's a phase connection. Imagine if the two friends could hear each other's whispers, but the sound traveled through the hallway in a way that sometimes made the whispers cancel each other out (silence) and sometimes made them amplify each other (loud noise).

By controlling exactly when the sound waves meet, they could switch the system from "Silence Mode" to "Loud Mode" instantly.

3. The Trick: The "Tuning Fork" Effect

Here is the magic moment:

• The Problem: Usually, to get a laser to fire, you need to pump it with so much energy that it breaks its limit. But these tiny bulbs are too small to hold that much energy without breaking.
• The Solution: The researchers kept Friend A steady and suddenly gave Friend B a big burst of energy. This changed the "pitch" (frequency) of Friend B's voice.
• The Magic: For a split second, the pitch of Friend B matched Friend A perfectly. Suddenly, the two friends stopped whispering and started singing in perfect harmony. Because they were in perfect sync, the "noise" (light) that was usually lost to the walls of the hallway suddenly got trapped and amplified between them.

This is called Q-Switching. Imagine a dam holding back water. Usually, the water leaks out slowly. But if you suddenly open a floodgate at the perfect moment, all that stored energy rushes out in one massive, powerful wave.

4. The Result: A Super-Fast Camera Flash

Because the energy was stored in the "shy" bulbs and then released all at once when they finally synced up, the result was a tiny, incredibly bright flash of light.

• Speed: These flashes happen billions of times per second (Gigahertz). It's so fast that if you could see it, it would look like a strobe light for a strobe light.
• Timing: The flashes are incredibly precise, with a "jitter" (timing error) of only about 15 to 40 picoseconds. That's like trying to hit a target with a dart, but your hand only shakes by the width of a single atom.

Why Does This Matter?

Think of your smartphone or the internet. They need to send data as fast as possible.

• Old way: Big, bulky lasers that are hard to fit on a tiny chip.
• New way: This method uses tiny, weak lasers that can't work alone, but when paired up and controlled with this "magic sync" trick, they become powerful, ultra-fast data transmitters.

In a nutshell: The scientists took two weak, useless light sources, forced them to "sing in perfect harmony" for a split second using a special connection, and harvested that moment of harmony to create a super-fast, powerful burst of light. This opens the door to making super-fast, tiny lasers for the next generation of computers and communication networks.

Technical Summary

1. Problem Statement

Generating short, high-peak-power optical pulses on integrated nanophotonic platforms is a significant challenge. Traditional Q-switching relies on modulating the cavity $Q$-factor to store and release energy, but this is difficult to implement in nanocavities due to:

• Low Photon Storage Capacity: Nanocavities have inherently small mode volumes, limiting energy storage.
• Control Mechanisms: Existing methods often rely on thermal or nonlinear effects that are slow or energy-intensive.
• Independent Control: In standard evanescent coupling, reactive coupling (frequency splitting) and dissipative coupling (loss redistribution) are linked, making it difficult to independently tune modal losses and frequencies to achieve Q-switching.
• Sub-Threshold Limitations: Individual nanolasers in this specific platform (InP photonic crystals on SOI) suffer from high external coupling losses, preventing them from lasing efficiently in continuous-wave (CW) operation on their own.

2. Methodology

The authors propose and demonstrate a novel approach using phase-controlled non-Hermitian coupling between two waveguide-coupled photonic crystal nanolasers.

• System Architecture: Two InGaAs quantum well nanocavities are integrated onto a Silicon-on-Insulator (SOI) waveguide, separated by $\sim 22 \, \mu\text{m}$. This distance allows for waveguide-mediated coupling with a controllable phase shift ($\phi$).

• Operational Regime: Both nanolasers are operated below their individual lasing thresholds.

  • NL1: Continuously pumped (CW) at a sub-threshold level.
  • NL2: Periodically pumped with short optical pulses (modulated via an Electro-Optical Modulator).

• Mechanism:

  1. Asymmetric Pumping: Modulating the pump power of NL2 changes its carrier density ($n_2$), creating a transient carrier density difference ($\Delta n$) relative to NL1.
  2. Non-Hermitian Coupling: Due to the non-zero Henry factor ($\alpha_h$), this carrier density difference induces a frequency detuning. When the system crosses a specific resonance condition ($\Delta n \approx 0$), the interference between the coupling pathways changes.
  3. Q-Factor Modulation: This dynamic tuning shifts the system from a high-loss state to a low-loss, high-gain state for one of the collective supermodes (the $+$ mode).
  4. Pulse Release: The rapid transition triggers the release of stored carrier energy as a short optical pulse (Q-switching).

• Modeling: The authors developed a coupled rate-equation model (Class-B semiconductor laser model) incorporating carrier dynamics, spontaneous emission, and carrier saturation to simulate the system's behavior.

3. Key Contributions

• Demonstration of Non-Hermitian Q-Switching: This is the first demonstration of active Q-switching in nanolasers driven purely by the dynamical control of non-Hermitian coupling, rather than direct intracavity loss modulation.
• Enabling Sub-Threshold Lasing: The system generates pulses from cavities that cannot lase efficiently in CW mode individually. The collective mode gain enhancement allows lasing only during the transient resonance.
• Decoupling of Parameters: By using waveguide-mediated distant coupling, the authors achieve independent control over frequency detuning and modal losses, overcoming the limitations of direct evanescent coupling.
• High-Frequency Operation: The system demonstrates stable pulse generation at multi-gigahertz repetition rates, a critical metric for integrated photonic applications.

4. Key Results

• Pulse Generation: Experiments on an Indium Phosphide (InP) platform successfully generated optical pulses with durations ranging from 70 ps to 300 ps.
• Threshold Crossing: The pulses are generated only when the pump power of NL2 is modulated to bring the system to the resonant condition. Below this threshold, the output remains weak spontaneous emission.
• Timing Characteristics:
  • Build-up Time: The delay between the pump pulse and the optical pulse decreases from $\sim 2.5$ ns to $0.2$ ns as the pump power increases.
  • Jitter: The timing jitter is extremely low, ranging from 15 to 40 ps, indicating high temporal stability.
• Repetition Rate:
  • At low frequencies ($< 2.7$ GHz), the system produces two pulses per cycle (one on the rising edge and one on the falling edge of the pump pulse).
  • At higher frequencies, the system transitions to a single-pulse-per-cycle regime due to incomplete carrier recovery between pulses.
  • The effective modulation bandwidth (3-dB cutoff) was measured at $10.5 \pm 2.5$ GHz.
• Model Validation: The numerical simulations based on the rate-equation model accurately reproduced the experimental observations, including the pulse build-up time, amplitude, and the transition from double-pulse to single-pulse regimes.

5. Significance

This work establishes a new paradigm for ultrafast pulse generation in integrated photonics:

• New Functionalities: It proves that non-Hermitian coupling can be used as a dynamic "switch" to control effective cavity losses in time, enabling functionalities like Q-switching in systems where traditional methods fail.
• Scalability: The approach is compatible with standard semiconductor fabrication (InP on SOI), making it suitable for mass-producible on-chip light sources.
• Applications: The ability to generate stable, high-repetition-rate pulses from sub-threshold nanolasers opens new avenues for:
  • High-speed optical communications: As compact transmitters for data transmission.
  • Neuromorphic computing: Providing the necessary spiking dynamics for photonic neural networks.
  • Precision metrology: Offering compact sources for frequency combs or timing synchronization.

In summary, the paper demonstrates that by exploiting the interplay of gain, loss, and interference in non-Hermitian systems, one can engineer ultrafast, high-performance pulse sources from nanocavities that are otherwise inefficient, bridging the gap between fundamental non-Hermitian physics and practical integrated photonic applications.