Photonic crystal cavities based on suspended yttrium iron garnet nanobeams

This paper reports the fabrication and optical characterization of air-suspended yttrium iron garnet (YIG) photonic crystal nanobeam cavities via focused-ion-beam milling, demonstrating a critical step toward on-chip integration for future coupled photon-phonon-magnon quantum technologies.

The Gist

Imagine you are trying to build a tiny, super-efficient factory inside a single grain of sand. This factory needs to handle three very different types of "workers" at the same time:

1. Light particles (Photons): Like messengers carrying information at the speed of light.
2. Sound vibrations (Phonons): Like tiny mechanical springs shaking back and forth.
3. Magnetic spins (Magnons): Like tiny compass needles spinning in unison.

In the world of quantum technology, getting these three workers to talk to each other is the "Holy Grail." If they can collaborate, we could build super-fast computers that talk to our current internet, or sensors so sensitive they can detect a single atom.

The Problem:
The material usually chosen for the "magnetic" workers is called Yttrium Iron Garnet (YIG). It's fantastic at handling magnetism and doesn't absorb much light. However, it's notoriously difficult to shape. Think of YIG like a block of diamond-hard cheese. You can't easily carve it into the tiny, intricate shapes needed for a quantum factory using standard tools. Previous attempts resulted in big, clumsy blocks (like marbles) where the workers couldn't really see or touch each other, making them inefficient.

The Solution:
The researchers in this paper decided to try something new. They used a Focused Ion Beam (FIB), which is essentially a super-precise, microscopic "laser cutter" made of heavy ions. Think of it as a CNC machine for the nanoscale world.

They took a thin slice of YIG and carved a suspended nanobeam out of it.

• The Shape: It looks like a tiny, floating bridge with a series of perfectly spaced holes drilled into it (like a Swiss cheese or a guitar string with holes).
• The Suspension: They cut the bridge so it hangs in the air, not touching the ground. This is crucial because it allows the bridge to vibrate freely (the "sound" worker) without losing energy to the floor.

What They Did:

1. Design: They used computer simulations to design the perfect pattern of holes. This pattern acts like a "trap" or a cage that forces light, sound, and magnetic waves to stay in the exact same small spot.
2. Fabrication: They used the ion beam to carve this structure. To protect the delicate YIG from getting damaged by the heat and debris of the cutting process, they put a temporary "sacrificial" aluminum coat on it (like a protective shell). After carving, they dissolved the aluminum, leaving a clean, floating YIG bridge.
3. Testing: They shone a laser at the bridge.

The Results:

• Success: The bridge worked! It trapped the light and created a specific "resonance" (a musical note) at a wavelength of 1634.8 nanometers. This proves they successfully made a quantum factory out of YIG.
• The Hiccup: The "quality" of the sound in the factory wasn't perfect yet. The light bounced around a bit too much and got lost (the "Quality Factor" was lower than expected).
  • Analogy: Imagine trying to sing in a concert hall, but the walls are a bit rough and uneven, so the sound echoes and fades quickly instead of ringing out clearly.
  • Cause: The tiny imperfections left by the ion beam (rough edges) and slight misalignments in the holes caused the light to leak out.

Why This Matters (The Future):
Even though the current version isn't perfect, this is a massive breakthrough.

• Before: We had to use big, separate devices for light, sound, and magnetism. They were like three different languages spoken in different rooms.
• Now: We have proven we can carve all three into a single, tiny, suspended piece of material.

The Next Steps:
The team knows why the light is leaking (rough edges). In the next version, they plan to:

1. Polish the edges to make the "concert hall" walls smooth.
2. Tune the size of the bridge so the "sound" worker and the "magnetic" worker vibrate at the exact same frequency.

The Big Picture:
If they succeed, this tiny device could act as a universal translator for the quantum internet. It could take a microwave signal from a quantum computer (which is very cold and fragile) and translate it into a light signal that can travel through fiber-optic cables across the world, without losing any information.

In short: They built the first prototype of a "quantum translator" out of a material that was previously too hard to work with. It's a bit rough around the edges right now, but the blueprint is proven, and the future of connected quantum networks just got a whole lot brighter.

Technical Summary

1. Problem Statement

Hybrid quantum systems integrating optical, mechanical, and magnetic (magnonic) degrees of freedom are promising for quantum technologies, such as coherent transduction between microwave and optical photons. Yttrium Iron Garnet (YIG) is the premier material for magnonic systems due to its ultra-low Gilbert damping and low optical absorption in the infrared. However, a significant barrier exists:

• Fabrication Limitations: YIG is incompatible with standard semiconductor fabrication processes (e.g., lithography and etching) used for silicon or diamond optomechanical crystals.
• Scale Mismatch: Prior YIG research has focused on millimeter-scale spheres or simple microstructures. These large geometries result in poor mode overlap between phonons (mechanical vibrations) and magnons (spin waves), limiting the single phonon-magnon coupling rate to the millihertz range—too weak for practical quantum applications.
• Lack of Integration: Existing YIG photonic crystals are typically planar multilayer stacks that cannot confine mechanical modes in a suspended geometry, preventing dynamic interactions between photons, phonons, and magnons.

The core challenge is to realize a nanofabricated, air-suspended YIG optomechanical crystal (OMC) that can simultaneously confine optical, mechanical, and magnonic modes within a shared nanoscale volume.

2. Methodology

The authors employed a combination of finite-element simulations and focused-ion-beam (FIB) milling to design and fabricate the device.

• Design:

  • The device is a suspended rectangular YIG nanobeam (20 µm length, 1.25 µm width) patterned with elliptical holes to form a photonic crystal waveguide.
  • Simulations: COMSOL Multiphysics was used to optimize the optical bandgap (targeting ~187 THz) and mechanical bandgap (targeting ~1.52 GHz). The design ensures confinement of a single TE-like optical mode and an x-symmetric mechanical "flapping" mode.
  • Magnonic Simulation: MuMax3 was used to simulate spin-wave dynamics under an in-plane magnetic field, predicting a backward-volume magnonic mode at ~11.59 GHz.

• Fabrication (FIB Milling):

  • Substrate: An 840 nm YIG film (111 orientation) on a Gadolinium Gallium Garnet (GGG) substrate.
  • Sacrificial Layer: A 50 nm Aluminum (Al) layer was deposited on the YIG surface. This served three purposes: acting as a hard mask, conducting heat to prevent thermal damage, and capturing redeposited material to keep the YIG surface clean.
  • Milling Process: A Xe plasma FIB was used (to avoid Ga contamination) in a three-step process:
    1. Coarse trench milling (1–3 nA) to define the bridge.
    2. Intermediate shaping (100–300 pA).
    3. Polishing and hole drilling (~30 pA) for high precision.
  • Release: The Al layer was removed via wet etching in heated KOH (80°C), followed by critical point drying to prevent the collapse of the suspended nanostructure.

• Characterization:

  • Optical properties were measured using a tunable laser (1600 nm range) coupled via a tapered, dimpled optical fiber. Transmission spectra were analyzed to determine resonance wavelengths and quality factors.

3. Key Contributions

• First Nanofabricated Suspended YIG OMC: This work reports the first observation of optical resonance in a nanofabricated, air-suspended YIG structure designed to host three bosonic quasiparticles (photons, phonons, magnons).
• Process Innovation: The development of a specific FIB milling protocol using a sacrificial Aluminum layer and Xe plasma to mitigate ion implantation, thermal damage, and surface roughness in YIG.
• Multiphysics Design: A unified design that theoretically supports overlapping optical, mechanical, and magnonic modes, bridging the gap between optomechanical and magnomechanical systems.

4. Results

• Optical Performance:

  • A resonance was observed at $\lambda = 1634.8$ nm.
  • The measured internal quality factor ($Q$) was $\approx 2 \times 10^3$.
  • Discrepancy Analysis: The measured $Q$ is significantly lower than the simulated radiation-limited value ($\sim 10^6$). The authors attribute this primarily to lateral misalignment of the elliptical hole row (approx. 100 nm offset) and aspect ratio deviations, which open radiation pathways. Secondary factors include surface roughness and FIB-induced damage.
  • The device operates in the unresolved sideband regime ($\kappa \gg \Omega_m$), preventing the optical readout of the mechanical mode.

• Theoretical Predictions (Simulations):

  • Mechanical Mode: A flapping-type mode at $\Omega_m/2\pi \approx 1.52$ GHz.
  • Magnonic Mode: A backward-volume spin-wave mode at $\omega_{mag}/2\pi = 11.59$ GHz under a 400 mT in-plane bias field.
  • Coupling Rates:
    • Single photon-phonon coupling ($g_0$) is estimated at 50 kHz (limited by YIG's low photoelastic tensor).
    • Linear magnon-phonon coupling ($g_{mb}$) is estimated at $\sim 37$ MHz (assuming perfect mode matching), though the current device is off-resonant.

5. Significance and Future Outlook

• Foundational Step: Despite the low optical $Q$, this device demonstrates the feasibility of creating suspended YIG nanostructures. It lays the groundwork for magneto-optomechanical systems capable of interacting photons, phonons, and magnons simultaneously.
• Quantum Transduction: The platform holds the potential for high-efficiency coherent conversion between microwave and optical photons, a critical requirement for connecting superconducting quantum processors to optical communication networks.
• Future Directions:
  • Fabrication Optimization: Future iterations will replace curved elliptical holes with straight-groove lattices to reduce sensitivity to lateral misalignment and improve $Q$.
  • Resonant Matching: Designs will be adjusted to bring mechanical modes (currently ~1.5 GHz) closer to the magnonic spectrum (10–12 GHz) to enable resonant, strong linear coupling.
  • Microwave Integration: Incorporating on-chip planar microwave resonators will allow for direct magnon characterization and cavity-magnomechanical measurements, even before the optical sideband is resolved.

In conclusion, this paper successfully bridges the gap between theoretical hybrid quantum systems and experimental reality by overcoming the fabrication challenges of YIG, paving the way for advanced quantum transducers and sensors.