🔬 optics

Artificial Anisotropy Induced Bound States in the Continuum for Integrated Photonic Waveguide

This paper introduces a new design paradigm for integrated photonic waveguides that utilizes artificial optical anisotropy via subwavelength-grating metamaterials to overcome the geometric limitations of traditional bound states in the continuum (BICs), enabling flexible, deterministic control over radiation and field leakage across a broad design space.

Original authors: Jinzhao Wang, Kunrun Lu, Yuanlin Li, Weiming Yao, Yang Feng, Yidi Cao, Wei Liu, Feng He, Jianan Duan, Yi Zou, Yongkang Dong, Xiaochuan Xu

Published 2026-08-24

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Jinzhao Wang, Kunrun Lu, Yuanlin Li, Weiming Yao, Yang Feng, Yidi Cao, Wei Liu, Feng He, Jianan Duan, Yi Zou, Yongkang Dong, Xiaochuan Xu

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

Light usually behaves in predictable ways. When it travels through a clear material, it moves in a straight line until it hits a boundary, where it either bounces back or leaks out. In the world of integrated photonics, which is the science of guiding light through tiny chips to carry information, this leakage is a major problem. Engineers spend a great deal of effort trying to trap light inside narrow channels, but the laws of physics often make it difficult to keep the light confined without it radiating away into the surrounding material. For decades, scientists have known about a strange exception to this rule, a phenomenon where light can be trapped perfectly even though it sits in a spectrum of energy states that should allow it to escape. This counterintuitive state is called a bound state in the continuum. It is like a wave that exists in the middle of a storm but never gets wet. While this concept has been demonstrated in the past, creating it has been a rigid and difficult process, requiring the physical shape of the chip to be tuned with extreme precision. If the dimensions were even slightly off, the light would escape, and the trap would fail.

A team of researchers at the Harbin Institute of Technology and ShanghaiTech University has now found a way to make this trapping mechanism much more flexible and robust. Instead of relying solely on the precise geometry of the chip's structure, they introduced a new method using artificial optical anisotropy. To understand this, imagine the material the light travels through not as a uniform block, but as something that reacts differently depending on the direction the light is moving. The researchers created this effect by carving tiny, repeating patterns into the silicon around the light channel. These patterns are so small that the light cannot see the individual lines; instead, it sees a smooth, engineered material with special properties. By adjusting the spacing and orientation of these tiny patterns, the scientists could control how the light interacts with the surrounding material without having to change the size of the channel itself. This approach transforms the design process from a game of fine-tuning rigid shapes into a more fluid process of tuning the material's internal character.

The researchers built a specific type of waveguide, which is a channel for light, consisting of a central core surrounded by two layers of this engineered material. In a standard setup, light traveling through the core would naturally leak out into the surrounding layers because the surrounding material allows for certain types of light waves to exist at the same speed as the light in the core. This matching of speeds usually causes the light to escape. However, the team discovered that by using their artificial patterns, they could arrange for the light waves leaking out to cancel each other perfectly. When the light hits the boundary between the core and the surrounding layer, it splits into waves that travel in different directions. By carefully designing the surrounding material, the researchers ensured that these split waves arrived back at the core with opposite phases, effectively erasing each other. This destructive interference stops the light from leaking, trapping it inside the core even though it is surrounded by a material that should let it escape.

To prove this worked, the team fabricated devices on a silicon chip and tested them with laser light. They created two sets of waveguides with different patterns in the surrounding layers. In the first set, the patterns were aligned symmetrically on both sides of the channel. When they measured the light passing through, they found that for most widths of the channel, the light was lost, just as expected. But at two very specific widths, the loss dropped dramatically, and the light traveled through with almost no leakage. For one pattern, this happened when the channel was 750 nanometers wide and again at 1500 nanometers. For a different pattern, the low-loss points shifted to 700 nanometers and 1450 nanometers. This shift proved that the researchers could control exactly where the light would get trapped simply by changing the design of the surrounding patterns, rather than being stuck with a single, fixed size. The light that did get trapped traveled with a loss of less than one decibel per centimeter, a level of efficiency that confirms the light was successfully confined.

The team also explored what happened when they broke the symmetry of the design. In previous attempts to create asymmetric traps, scientists had to physically make the two sides of the channel different, which was difficult to manufacture. Here, the researchers simply rotated the tiny patterns on one side relative to the other. This small change altered how the light interacted with the boundary on that side. The result was that the light no longer canceled out perfectly on both sides. Instead, the light leaked out more on one side than the other, creating a controlled, asymmetric flow. This allowed them to tune the amount of light that escaped, turning the device into a variable leak. They found that by changing the angle of the rotation, they could adjust the loss from a few decibels to over ten decibels, giving them a new tool to manage how light escapes from the chip.

This work suggests a new way to build photonic devices that are more adaptable than current technology. The ability to create these trapped states of light without being locked into a single, rigid geometry means engineers can design chips that are more forgiving of manufacturing errors and more versatile in their function. The researchers demonstrated that they could create these traps over a broad range of sizes and that the light could travel through them without losing much energy. They also showed that the trapped light could be made to leak in a controlled, directional way. These findings open the door to more advanced sensors and communication devices that can manipulate light with a level of precision that was previously difficult to achieve. By treating the material around the light channel as a programmable element rather than just a passive container, the researchers have established a general framework for designing the next generation of integrated photonic devices.

Technical Summary: Artificial Anisotropy Induced Bound States in the Continuum for Integrated Photonic Waveguide

Problem Statement
Bound states in the continuum (BICs) offer a mechanism for light confinement without radiation loss, enabling applications such as low-loss waveguiding, sensing, and microlasers. However, existing implementations of BIC waveguides, particularly in cladding–core–cladding architectures, are constrained by a geometry-dependent paradigm. In these conventional designs, the BIC condition relies on precise structural parameters (e.g., core width and etch depth) where geometric tuning simultaneously governs modal birefringence, continuum coupling, and radiation interference. This interdependence limits the design space, restricts flexibility, and makes it difficult to independently engineer the radiative continuum. While natural anisotropic materials (e.g., LiNbO3) offer some control, their fixed birefringence is insufficient, and heterogeneous integration increases fabrication complexity. There is a need for a strategy that allows independent, deterministic reshaping of the radiative continuum to enable flexible BIC formation and control.

Methodology
The authors propose a new design paradigm utilizing artificial optical anisotropy generated by subwavelength-grating (SWG) metamaterials.

Structure: The waveguide consists of a silicon core surrounded by cladding regions engineered with SWG slabs. These slabs behave as homogeneous anisotropic media with continuously tailorable ordinary ( $n_{\parallel}$ ) and extraordinary ( $n_{\perp}$ ) refractive indices, controlled via the grating duty cycle ( $a/\Lambda$ ).
Mechanism: By embedding these anisotropic claddings, the authors decouple the control of the radiative continuum from the core geometry. The SWG cladding allows for polarization-selective dispersion control. Specifically, the effective index of the cladding TE slab mode ( $N_{TE}^{clad}$ ) can be continuously tuned by rotating the optical axis of the SWG or adjusting the duty cycle, while the core TM mode effective index ( $N_{TM}^{core}$ ) remains largely invariant.
BIC Formation: The BIC condition is achieved when the core TM mode, spectrally embedded within the TE slab continuum, experiences destructive interference between TE waves reflected and transmitted at the core–cladding interfaces. The anisotropy allows the phase-matching conditions and the interference pathway to be engineered independently of the core width.
Symmetry Breaking: The platform further enables controllable symmetry breaking by tilting the optical axis of the SWG claddings relative to the waveguide axis. This creates asymmetric radiation channels where leakage can be precisely tailored without altering the geometric structure of the claddings.
Fabrication & Characterization: Devices were fabricated on a standard SOI platform using electron-beam lithography and inductively coupled plasma (ICP) etching. Lateral tapers were implemented to connect SWG slabs to the surrounding silicon, ensuring an open radiation continuum. Transmission loss was measured using TM-polarized light across the 1500–1600 nm range.

Key Results

Deterministic BIC Control: Simulations and experiments demonstrated that the BIC condition (loss minima) can be "steered" across a broad design space by varying the SWG duty cycle. For a duty cycle of 0.6, distinct low-loss states were observed at specific core widths (e.g., ~835 nm and ~1610 nm in simulation; 750 nm and 1500 nm in experiment). While simulation predicted losses as low as 0.04 dB/cm, experimental measurements showed relative-loss values of 0.87 dB and -0.95 dB for the two minima, respectively. The authors note that the negative relative loss value is attributed to fabrication deviations, yet the data confirms the existence of discrete loss minima distinct from the high-loss background.
Broadband Operation: The BIC states exhibited intrinsically broadband low-loss responses due to the weak net dispersion of the phase condition. Conversely, non-BIC states showed high, fluctuating loss, confirming the "all-or-nothing" nature of the interference-stabilized confinement. Specifically, outside the discrete BIC widths, transmission loss exceeded 35 dB, effectively reaching the noise floor of the experimental setup.
Asymmetric Waveguiding: By rotating the SWG optical axis (tilt angles $\theta_r$ ), the authors successfully broke the symmetry of the cladding. This resulted in controllable asymmetric radiation and finite-loss regimes. As the tilt angle increased, the minimum loss initially rose (due to imperfect interference) and then decreased as the radiation channel approached closure. This allowed for the creation of waveguides with tunable leakage and asymmetric field confinement.
Experimental Validation: Measured relative transmission losses confirmed the theoretical predictions, showing distinct loss dips at predicted core widths for different duty cycles (0.6 and 0.8) and tilt angles (10° and 25°). The results validated that metamaterial-induced anisotropy provides a robust degree of freedom for engineering BICs, despite minor offsets attributed to fabrication deviations and the limits of effective medium theory applicability.

Significance and Claims
The paper claims to transform BIC waveguides from geometry-constrained structures into an anisotropy-engineered platform. The primary significance lies in establishing a general framework for programmable radiation engineering.

Design Flexibility: The approach provides an independent degree of freedom (anisotropy) to deterministically reshape the radiative continuum, overcoming the limitations of single-parameter geometric tuning.
Fabrication Compatibility: The method utilizes standard CMOS-compatible lithographic patterning of SWGs, avoiding the complexity of heterogeneous material integration.
Functional Versatility: The platform enables the creation of devices with tailored leakage, including controllable asymmetric radiation and precise field leakage. The authors suggest this paradigm supports applications such as ultra-sensitive sensors, high polarization extinction ratio (PER) polarizers, and broadband TM-operation devices.
New Physics: The ability to engineer anisotropic asymmetry offers a route to explore emergent phenomena in cladding–core–cladding waveguides, such as asymmetric BICs and unidirectional resonances, which are difficult to realize in conventional etched-slab platforms.

In conclusion, the work demonstrates that artificial optical anisotropy serves as a versatile tool for engineering both the radiative continuum and leakage behavior, establishing a practical strategy for next-generation integrated photonic devices.

More like this