🔬 optics

Wavefront Control and Intensity Modulation of Third Harmonic Generation in Nonlocal Metasurfaces

This paper demonstrates a nonlocal phase gradient metasurface that simultaneously achieves efficient third harmonic generation with polarization-dependent wavefront control and tunable intensity modulation via the interference of multiple fundamental beams, effectively merging the high conversion efficiency of nonlocal designs with the versatile control capabilities of local ones for on-chip nonlinear photonic applications.

Original authors: Yu Tian, Nuo Wang, Qi Liu, Shuyuan Xiao, Tingting Liu, Olivier J. F. Martin, Ying Gu

Published 2026-02-03

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Yu Tian, Nuo Wang, Qi Liu, Shuyuan Xiao, Tingting Liu, Olivier J. F. Martin, Ying Gu

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 have a very thin, smart sheet of material (a "metasurface") made of tiny silicon blocks. This sheet has a special superpower: when you shine a specific color of light on it, it can instantly change that light into a new color (specifically, turning red light into blue-violet light, a process called "Third Harmonic Generation").

However, scientists have faced a tricky problem for a long time. Usually, you have to choose between two things:

High Efficiency: Making a lot of the new light (but you can't control where it goes).
Wavefront Control: Steering the new light to go exactly where you want (but you only get a tiny amount of it).

This paper claims to have solved that problem. They built a single sheet that does both at the same time.

Here is a breakdown of how they did it, using simple analogies:

1. The "Super-Resonant" Trampoline (The q-BIC)

Think of the tiny silicon blocks on the sheet as individual trampolines.

Old way (Local Metasurfaces): These were like weak trampolines. If you jumped on them, you didn't go very high. This meant the light conversion was weak.
New way (Nonlocal Metasurfaces): The researchers designed these blocks to act like a "quasi-bound state in the continuum" (q-BIC). Imagine a trampoline that is perfectly tuned so that if you jump on it, the energy gets trapped and amplified, making you bounce incredibly high.
The Result: Because the light gets trapped and amplified in these "super-trampolines," the sheet produces 100 times more of the new colored light than previous designs could.

2. The "Spin-Steering" Wheel (Wavefront Control)

Now, imagine you want to steer that bouncing light.

The Trick: The researchers tilted the tiny elliptical holes in the silicon blocks at different angles, like setting up a row of wind vanes.
The Magic: When the light hits these tilted blocks, it doesn't just bounce straight up. The tilt acts like a steering wheel.
- If you shine Right-Circularly Polarized light (light spinning clockwise) on the sheet, the new light is steered to the right (specifically, into the 2nd and 4th "lanes" or angles).
- If you shine Left-Circularly Polarized light (light spinning counter-clockwise), the new light is steered to the left.
Why it's special: In previous designs, you couldn't do this steering while keeping the high efficiency. This design uses a special "nonlocal" effect (where the blocks talk to each other) to steer the light without losing the power boost.

3. The "Volume Knob" (Intensity Modulation)

Finally, the researchers wanted to see if they could turn the brightness of this new light up and down, like a volume knob, without changing the hardware.

The Setup: They used two beams of light instead of one.
- Beam 1: Shines from the bottom.
- Beam 2: Shines from the top.
The Interference (The Wave Dance): Imagine two people pushing a swing.
- If they push at the exact same time (in sync), the swing goes super high (Constructive Interference).
- If one pushes while the other pulls back, the swing stops (Destructive Interference).
The Control: By slightly changing the timing (phase), the color (polarization), or the strength (intensity) of the second beam, they could make the new light go from completely off to very bright.
The Result: They could adjust the brightness of the new light by a factor of one million (six orders of magnitude). It's like having a dimmer switch that can go from a candle's flame to a stadium spotlight instantly.

Summary

The paper demonstrates a single, silicon-based sheet that:

Amplifies light conversion significantly (like a super-trampoline).
Steers the light to specific directions based on how the light is spinning (like a spin-activated rudder).
Switches the light's brightness on and off using a second beam of light (like a volume knob based on wave interference).

The authors state this paves the way for making smaller, more powerful, and more versatile optical devices that can process signals directly on a chip, essentially creating a "traffic controller" for light that is both fast and efficient.

Technical Summary: Wavefront Control and Intensity Modulation of Third Harmonic Generation in Nonlocal Metasurfaces

Problem Statement
Metasurfaces have emerged as a platform for integrated nonlinear optics, offering control over light properties at the micro/nano scale. However, a significant trade-off exists between conversion efficiency and wavefront control. Local metasurfaces allow for versatile wavefront engineering (e.g., holography, vortex generation) by independently tailoring unit cells, but they typically suffer from low nonlinear conversion efficiencies due to limited quality factors ( $Q < 10^2$ ). Conversely, nonlocal metasurfaces, particularly those supporting quasi-bound states in the continuum (q-BICs), offer strong near-field enhancement and high quality factors ( $Q > 10^3$ ), enabling high conversion efficiencies. Yet, achieving simultaneous high efficiency and flexible wavefront control in nonlocal systems remains challenging because the response of individual unit cells is strongly coupled to neighbors, making independent engineering difficult. Additionally, while all-optical modulation of nonlinear responses is vital for signal processing, existing methods often rely on low-coefficient nonlinear processes (e.g., Kerr effect) or lack universality.

Methodology
The authors propose and simulate a silicon-based nonlocal phase gradient metasurface (NPGM) designed to operate at a q-BIC resonance.

Structure: The unit cell consists of a silicon slab (260 nm thick) with elliptical holes etched into it, placed on a glass substrate. The holes have a major axis of 260 nm and a minor axis of 70 nm. The rotation angle ( $\theta$ ) of these elliptical holes is varied across the metasurface to encode a geometric phase gradient.
Resonance: The structure is designed to support a q-BIC at a fundamental wavelength of approximately 1530 nm, yielding a quality factor of $\sim 120$ .
Simulation: Numerical analysis was performed using the finite element method (COMSOL Multiphysics) in the frequency domain. The study examines linear transmission/reflection spectra, near-field distributions, and third harmonic generation (THG) efficiency under various pump intensities (up to 1 GW/cm²) and polarizations.
Modulation Strategy: To achieve intensity modulation, the study introduces a secondary fundamental beam incident from the opposite side (air side) relative to the primary beam (substrate side). The interference between these two beams is analyzed by varying their relative phase, intensity, and polarization states.

Key Contributions

Nonlocal Nonlinear Geometric Phase: The paper establishes that nonlocal metasurfaces at q-BICs impart a geometric phase to the generated third harmonic light that is distinct from local metasurfaces. Specifically, the phase accumulation is $4\theta\sigma$ for co-polarized components and $8\theta\sigma$ for cross-polarized components (where $\theta$ is the rotation angle and $\sigma$ represents circular polarization handedness). This is exactly twice the phase accumulation found in local metasurfaces due to the unique symmetry properties of the nonlocal resonance.
Simultaneous Efficiency and Control: The authors demonstrate that this nonlocal nonlinear geometric phase enables polarization-dependent wavefront control without sacrificing the high conversion efficiency provided by the q-BIC.
Interference-Based Intensity Modulation: The study introduces a method to modulate THG efficiency via linear interference between two fundamental beams. By overlapping the TH light generated by a primary beam and a secondary beam, the conversion efficiency can be dynamically tuned.

Results

High Efficiency THG: Under a pump intensity of 1 GW/cm², the NPGM achieves a THG conversion efficiency of $1.45 \times 10^{-4}$ at a TH wavelength of 509 nm. This represents an improvement of nearly two orders of magnitude compared to local metasurfaces.
Wavefront Control: The TH light is deflected into specific diffraction orders based on the incident polarization.
- Right-circularly polarized (RCP) incidence deflects TH light into the $+2^{nd}$ and $+4^{th}$ diffraction orders.
- Left-circularly polarized (LCP) incidence deflects TH light into the $-2^{nd}$ and $-4^{th}$ diffraction orders.
Intensity Modulation: By introducing a secondary beam and varying its parameters, the THG efficiency is modulated over a range of six orders of magnitude, from $3.9 \times 10^{-9}$ $3.9 \times 1 0^{- 9}$ to $5.5 \times 10^{-3}$ $5.5 \times 1 0^{- 3}$ .
- Phase Control: Varying the relative phase ( $\Delta\phi$ ) between the two beams allows for switching between constructive and destructive interference, achieving a near-unity modulation depth.
- Polarization Control: Adjusting the polarization state (orientation and ellipticity) of the secondary beam modulates the amplitude of the interfering LCP component, further enabling control over the THG intensity.

Significance
The paper claims that these results effectively address the challenge of simultaneously realizing high conversion efficiency and flexible wavefront control in a single metasurface. By leveraging the advantages of both local (wavefront engineering) and nonlocal (high efficiency) metasurfaces, the proposed NPGM paves the way for compact, multifunctional nonlinear photonic devices. Furthermore, the demonstrated interference-based modulation of nonlinear conversion efficiency, which does not rely on specific low-coefficient nonlinear processes, offers a promising route for on-chip nonlinear signal processing and optical switching. The work suggests that such platforms could also be extended to quantum nonlinear processes, such as spontaneous parametric down-conversion, with wavefront control capabilities.

1. The "Super-Resonant" Trampoline (The q-BIC)

2. The "Spin-Steering" Wheel (Wavefront Control)

3. The "Volume Knob" (Intensity Modulation)

Summary

More like this