Optical bound states in the continuum in subwavelength gratings made of an epitaxial van der Waals material

This study demonstrates the fabrication and characterization of an ultrathin subwavelength grating made from epitaxially grown MoSe$_2$ that supports optical bound states in the continuum and enhances third-harmonic generation efficiency by over three orders of magnitude, paving the way for compact near-infrared photonic devices.

The Gist

Imagine you are trying to trap a fly inside a room made of glass. Usually, the fly will buzz around and eventually fly out through the open windows. But what if you could design the room so that the fly gets stuck in a specific spot, vibrating intensely, yet never touches the walls or escapes, even though the room is technically "open" to the outside world?

In the world of light, this impossible-sounding trap is called a Bound State in the Continuum (BIC). It's a special state where light gets stuck in a tiny space, vibrating with incredible energy and staying there forever (theoretically), even though it's surrounded by open space where light usually flies away.

This paper is about building the perfect "fly trap" for light, but instead of using glass, the researchers used a very special, ultra-thin material called Molybdenum Diselenide (MoSe₂).

Here is the story of how they did it, broken down into simple steps:

1. The Problem: Finding the Right "Glass"

To trap light, you need a material that bends light very strongly (a high "refractive index"). Think of this like a dense forest that slows down a runner. Most materials are either too transparent (the light runs right through) or they absorb the light (the runner gets tired and stops).

The researchers looked at a huge list of materials (like silicon, glass, and various crystals) and found that MoSe₂ is a superstar. It's a "van der Waals" material, which means it's made of layers that are stuck together loosely, like a stack of sticky notes. It has an incredibly high ability to bend light, but it's also very thin and flexible.

2. The Design: The "Comb" Trap

You can't just put a flat sheet of MoSe₂ down and expect to trap light. You need to carve it into a specific shape. The researchers designed a subwavelength grating.

Imagine taking a piece of MoSe₂ and carving it into a tiny, microscopic comb or a series of parallel ridges.

• The Comb: The ridges are so close together (smaller than the wavelength of the light) that the light can't just pass through.
• The Trap: When light hits this comb, it gets caught in the "teeth" of the comb. It bounces back and forth so perfectly that it creates a standing wave. Because of the specific geometry, the light doesn't leak out the top or bottom. It's trapped in a "perfect storm" of vibration.

3. The Manufacturing: Growing and Polishing

Making this isn't easy. You can't just buy a sheet of MoSe₂ this perfect.

• Growing: The team used a high-tech oven (Molecular Beam Epitaxy) to grow a giant, uniform sheet of MoSe₂ on a sapphire crystal. It's like growing a perfect crystal snowflake, but on a scale of a few inches wide.
• Polishing: The grown surface was a bit bumpy (like a rough road). They had to gently polish it down with a silk tissue until it was as smooth as a mirror.
• Carving: Then, they used an electron beam (like a super-fine laser pen) to draw the "comb" pattern and etched away the extra material, leaving only the tiny ridges.

4. The Discovery: Catching the Light

When they shined light on their new "comb," something magical happened.

• The Vortex: At a specific angle and color of light (near-infrared, around 1100 nm), the light got stuck. It didn't reflect away normally. Instead, the polarization of the light (the direction it wiggles) started spinning in a circle, creating a vortex. This spinning pattern is the "fingerprint" that proves the light is trapped in a Bound State in the Continuum.
• The Proof: Even though the material isn't perfectly pure (it has tiny impurities), the trap was so good that the light stayed put.

5. The Superpower: Making Light Brighter

Why do we care about trapping light? Because trapped light is loud.
When light is trapped, it interacts with the material much more strongly. The researchers tested this by trying to create a "Third Harmonic" signal.

• The Analogy: Imagine hitting a drum. If you hit it once, you hear a quiet thump. But if you hit it while it's already vibrating in a perfect resonance, the sound explodes.
• The Result: When they shined a laser on their trapped-light comb, the material generated a new color of light (the third harmonic) that was 1,650 times brighter than if they had just shined the laser on a flat, uncarved sheet of the same material.

Why This Matters

This paper is a big deal for two reasons:

1. It's the thinnest trap yet: They made this work with a layer of material only about 40 nanometers thick (that's 1,000 times thinner than a human hair).
2. It's scalable: Unlike previous experiments that used tiny, fragile flakes of material (like tearing a piece of paper), they grew a large, uniform sheet. This means we could potentially mass-produce these "light traps" for real-world devices.

In summary: The team built a microscopic, ultra-smooth comb out of a special crystal. They proved that this comb can trap light in a perfect, non-leaking vortex. This trapped light acts like a super-charged amplifier, making the material glow much brighter and more efficiently. This could lead to better lasers, faster computers, and more sensitive sensors in the future.

Technical Summary

1. Problem Statement

Optical Bound States in the Continuum (BICs) are non-radiating electromagnetic resonant states that exist within a continuous spectrum of unbounded states. They are highly desirable for applications like low-threshold nanolasers, single-photon sources, and enhanced light-matter interactions due to their theoretically infinite quality (Q) factors.

• The Challenge: Realizing BICs typically requires subwavelength periodic structures (gratings) made of high-refractive-index materials. However, many candidate materials suffer from absorption, low refractive indices, or fabrication difficulties.
• The Gap: While Transition Metal Dichalcogenides (TMDs) like MoSe$_2$ possess exceptionally high refractive indices (up to 4.5–5), they have not been successfully utilized to create subwavelength gratings hosting BICs. Previous attempts were limited to theoretical predictions or relied on mechanically exfoliated flakes, which lack the size homogeneity and thickness control required for scalable device fabrication.

2. Methodology

The authors employed a combined approach of theoretical modeling, epitaxial growth, nanofabrication, and optical characterization.

• Theoretical Design:

  • Used the Plane-Wave Admittance Method (PWAM) and Plane-Wave Reflection Transformation Method (PWRTM) to simulate optical modes.
  • Modeled a MoSe$_2$ subwavelength grating on an Al$_2$O$_3$ (sapphire) substrate.
  • Optimized geometric parameters: Period ($L$), height ($h$), and fill factor ($F$). The design targeted the TE$_{10}$ (antisymmetric) mode, which is predicted to host a BIC, while avoiding the TE$_{20}$ (symmetric) mode which leaks energy.
  • Selected parameters: $L = 500$ nm, $h \approx 40\text{--}42$ nm, and $F \approx 0.8$.

• Material Synthesis (Growth):

  • Grown large-area (few cm$^2$), homogeneous MoSe$_2$ layers using Molecular Beam Epitaxy (MBE) on sapphire substrates.
  • Growth rate was kept low ($\sim$1 monolayer/hour) with intermediate annealing steps to ensure high optical quality and prevent vertical island formation.
  • Post-growth mechanical polishing (using silk tissue) reduced surface roughness from $S_q = 14.6$ nm to $2.0$ nm.
  • Characterized optical constants (real $n$ and imaginary $k$) via variable-angle spectroscopic ellipsometry, confirming a high real refractive index ($n_{xy} \approx 4.5$) and negligible absorption in the near-infrared (NIR).

• Fabrication:

  • Created one-dimensional subwavelength gratings using electron-beam lithography (EBL) and dry etching (Reactive Ion Etching with O$_2$ and SF$_6$).
  • The process avoided metal deposition and lift-off steps, utilizing a direct etch of the MoSe$_2$ layer.

• Characterization:

  • Angle-resolved reflectivity: Measured in k-space to map dispersion relations and identify BIC signatures (linewidth narrowing and vanishing at $k=0$).
  • Polarization analysis: Measured Stokes parameters to detect polarization vortices, a topological signature of BICs.
  • Nonlinear Optics: Measured Third Harmonic Generation (THG) to assess light-matter interaction enhancement.

3. Key Contributions

1. First Experimental Realization: This is the first demonstration of optical BICs in subwavelength gratings made entirely of a TMD semiconductor (MoSe$_2$).
2. Scalable Epitaxial Growth: The authors successfully grew large-area, homogeneous MoSe$_2$ layers via MBE, overcoming the size and uniformity limitations of exfoliated flakes.
3. Ultra-Thin BIC Structures: The fabricated gratings are only $\sim$42 nm thick, representing the thinnest BIC-hosting structures reported to date.
4. Direct Integration: Unlike previous works where TMDs were deposited onto conventional gratings, this work demonstrates BICs and nonlinear enhancement directly within the nanostructured van der Waals material itself.

4. Results

• Optical Resonance: Angle-resolved reflectivity measurements showed a distinct optical mode at $\sim$1100 nm. As the in-plane momentum ($k$) approached zero, the linewidth of the TE$_{10}$ mode narrowed significantly and vanished, confirming the non-radiating nature of the BIC.
• Topological Confirmation: A polarization vortex was observed in the far-field radiation pattern around the BIC point ($k=0$). The rotation of the polarization vector confirmed the topological charge associated with the BIC.
• Robustness to Absorption: Despite the MoSe$_2$ layer having a non-zero imaginary refractive index (absorption), the BIC state was successfully formed, proving that internal losses do not preclude the existence of BICs in this geometry.
• Nonlinear Enhancement:
  • The structure was used to enhance Third Harmonic Generation (THG).
  • At an incident angle of 27° (where coupling is possible), the THG signal was enhanced by a factor of 1650 compared to a uniform, unstructured MoSe$_2$ layer.
  • This represents an enhancement of over three orders of magnitude, demonstrating the ability of the BIC to confine light and boost light-matter interactions even in ultra-thin films.

5. Significance and Outlook

• Material Potential: The study highlights MoSe$_2$ as a superior candidate for photonic devices due to its exceptionally high refractive index and low absorption in the NIR, properties that were previously underutilized.
• Scalability: The use of MBE for growth and standard lithography for patterning suggests a pathway toward industrial-scale fabrication of ultrathin photonic structures, a significant step forward from the limited size of exfoliated 2D materials.
• Future Applications: The work opens avenues for:
  • Ultra-compact, low-threshold nanolasers.
  • Efficient single-photon sources.
  • Enhanced nonlinear frequency conversion devices.
  • Integration of single-photon emitters deterministically within the grating structure.

In conclusion, this paper bridges the gap between theoretical predictions of BICs in 2D materials and practical, scalable device implementation, leveraging the unique optical properties of epitaxial MoSe$_2$ to achieve record-breaking light confinement and nonlinear enhancement in ultra-thin films.