🔬 optics

Dual-Polarization Quasi-BIC Refractive Index Sensing via Dielectric Symmetry Breaking in TiO $_2$ -BeS Metasurfaces

This paper numerically demonstrates a dual-polarization TiO $_2$ -BeS metasurface sensor that utilizes dielectric symmetry breaking to simultaneously excite distinct quasi-BIC and magnetic dipole resonances, achieving high sensitivity and selectivity for refractive index sensing in the near-infrared.

Original authors: Shoumik debnath, Sudipta Saha

Published 2026-05-19

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Shoumik debnath, Sudipta Saha

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 tiny, invisible "trap" designed to catch light. This trap is made of two parallel bars of titanium dioxide (a material like white pigment) sitting on a glass slide. Between these two bars, there is a very narrow gap.

In this new study, researchers from Bangladesh University of Engineering and Technology placed a microscopic "bridge" made of a material called Beryllium Sulfide (BeS) right inside that gap. This bridge is only 20 nanometers thick—about 4,000 times thinner than a human hair.

Here is how this simple setup creates a powerful sensor, explained through everyday analogies:

1. The "Two-Track" Light Highway

Usually, if you shine light on a symmetric object (like two identical bars), the light behaves the same way no matter which direction it comes from. But the researchers wanted the light to behave differently depending on its "orientation" (polarization).

Think of the light as a car driving on a road.

The TE Car (Horizontal): When the light waves are oriented horizontally, they hit the BeS bridge and get stuck in a very specific, high-quality "traffic jam" (a resonance) at a specific color (879 nm). This jam is very tight and organized.
The TM Car (Vertical): When the light waves are oriented vertically, they ignore the bridge's special properties and flow around the bars in a broader, looser pattern (a magnetic dipole resonance) at a slightly different color (910 nm).

Because the bridge is made of a special crystal that acts differently for horizontal vs. vertical light, it breaks the symmetry without needing to make the bars themselves look different. It's like having a perfectly symmetrical room where the floor is made of two different types of wood; a ball rolling one way feels the first wood, while a ball rolling the other way feels the second.

2. The "Scent Detector" Analogy

The goal of this device is to sense changes in the air or liquid surrounding it (the "analyte").

Imagine the space between the bars is a sensitive nose.

The Horizontal Sensor (TE): The "nose" for the horizontal light is extremely sensitive. It is located right in the gap where the BeS bridge sits. If the air changes even slightly (like a tiny change in humidity or a gas appearing), this sensor notices immediately. It shifts its "traffic jam" color by a large amount.
The Vertical Sensor (TM): The "nose" for the vertical light is less sensitive because the light spends more time inside the bars and less time in the gap. It still notices the change, but the shift is smaller.

3. The "Fingerprint" Advantage

Most sensors only have one "nose." If two different gases cause the same tiny shift in the sensor, you can't tell them apart.

This new device has two noses that react differently.

If you have a normal gas, the horizontal sensor might shift by 10 steps, and the vertical sensor shifts by 7 steps.
If you have a weird, special gas, the horizontal sensor might shift by 10 steps, but the vertical one might shift by 9 steps.

By looking at the ratio of these two shifts, the device creates a unique "fingerprint." This allows it to tell the difference between things that would look identical to a single-sensor device.

4. What the Numbers Mean

The researchers ran computer simulations (using digital models, not physical experiments yet) and found:

Sensitivity: The horizontal sensor is very sharp, detecting changes as small as 0.00001 units of density (Refractive Index Unit).
Reliability: They tested the design against different computer programs and found the results were consistent. They also checked if the device would break if the bars were built slightly too big or too small (like a factory making a tiny mistake). The device still works well even with small manufacturing errors.
Material Choice: They used Titanium Dioxide because it is transparent and works well with the specific colors of light they are using. Silicon, which is common in electronics, would absorb this light and ruin the sensor.

Summary

The researchers have designed a digital blueprint for a tiny sensor that uses a microscopic bridge to split light into two different behaviors. This allows the sensor to not only detect that something is there, but to get a "second opinion" from a different angle, making it much better at distinguishing between different substances than current single-channel sensors.

Important Note: The paper states this is a computer simulation. The device has not been built or tested in a real lab yet. The results are based on mathematical models and digital experiments.

Technical Summary: Dual-Polarization Quasi-BIC Refractive Index Sensing via Dielectric Symmetry Breaking in TiO2-BeS Metasurfaces

Problem Statement
Conventional surface plasmon resonance (SPR) sensors, while widely used for label-free refractive index (RI) sensing, suffer from Ohmic losses in metal layers that broaden resonance linewidths and limit the quality factor ( $Q$ ). Additionally, local heating effects can be detrimental to biological analytes. While all-dielectric metasurfaces offer a low-loss alternative supporting high- $Q$ Mie resonances, most existing designs rely on a single spectral resonance shift for sensing. This single-channel approach struggles to distinguish between different analytes that possess similar effective refractive indices. Furthermore, many high- $Q$ designs require geometric asymmetry to activate bound states in the continuum (BIC), which can complicate fabrication and reduce tolerance. There is a need for a dual-channel sensing platform that operates in the near-infrared (NIR) with high sensitivity, distinct polarization responses, and robust fabrication tolerance without relying on geometric asymmetry.

Methodology
The authors numerically investigate a dual-polarization dielectric metasurface composed of TiO $_2$ nanobar pairs separated by a 60 nm gap, resting on a SiO $_2$ substrate. The core innovation is the insertion of a 20 nm Beryllium Sulfide (BeS) bar symmetrically within the gap. Unlike traditional approaches that break symmetry via geometric distortion, this design utilizes dielectric symmetry breaking.

Material Properties: TiO $_2$ is chosen for its high refractive index ( $n \approx 2.35$ ) and transparency in the NIR. BeS, a II–VI semiconductor, is utilized for its thin-film birefringence ( $\Delta n \approx 0.10\text{--}0.12$ ) and negligible extinction coefficients in the NIR. The optical constants for BeS are derived from first-principles Density Functional Theory (DFT) calculations.
Simulation: The study employs Finite-Difference Time-Domain (FDTD) simulations using Lumerical FDTD. Periodic boundary conditions are applied in the $x$ and $y$ directions, with Perfectly Matched Layers (PML) in $z$ . A broadband plane wave (850–950 nm) is used to excite the structure under both TE (0 $^\circ$ , $E$ along the gap) and TM (90 $^\circ$ , $E$ along the bar axis) polarizations.
Validation: To ensure numerical robustness, the results are cross-validated using the Tidy3D solver. Mesh convergence tests confirm that resonance shifts due to mesh refinement are negligible compared to RI-induced shifts. Structural tolerance is analyzed by sweeping geometric parameters ( $\pm 10$ nm) and perturbing BeS optical constants ( $\pm 5\%$ ).

Key Contributions

Dielectric Symmetry Breaking Mechanism: The paper demonstrates that a symmetric geometric structure can support two distinct, polarization-selective resonances solely through the introduction of a birefringent dielectric insert (BeS). This eliminates the need for geometric asymmetry (e.g., varying bar lengths or widths) to activate quasi-BIC modes.
Dual-Channel Sensing Architecture: The design enables simultaneous tracking of two independent sensing channels (TE and TM) within a single measurement window. The TE channel excites a quasi-BIC resonance, while the TM channel excites a magnetic dipole (MD) mode, separated spectrally by 31.6 nm.
Polarization-Dependent Spectral Fingerprint: By exploiting the different field distributions of the two modes, the sensor generates a "polarization fingerprint." Isotropic analytes follow a fixed slope in the dual-channel shift space, while anisotropic analytes or those with specific surface interactions deviate from this line, offering a potential pathway for chemical discrimination beyond simple RI measurement.
Fabrication Tolerance and Robustness: The study confirms that the resonance behavior remains stable under moderate dimensional variations and that the sensing metrics are independent of the specific FDTD solver used.

Results

Resonance Characteristics: Under TE illumination, the structure supports a quasi-BIC resonance at 879.2 nm with a quality factor $Q = 128$ and a narrow linewidth (FWHM = 6.87 nm). Under TM illumination, a broader magnetic dipole resonance occurs at 910.8 nm with $Q = 36$ (FWHM = 25.55 nm).
Sensing Performance: For background refractive indices ranging from 1.00 to 1.05:
- TE Channel: Sensitivity ( $S$ ) = 243.1 nm/RIU, Figure of Merit (FOM) = 35.4 RIU $^{-1}$ , Limit of Detection (LoD) = $4.1 \times 10^{-5}$ RIU.
- TM Channel: Sensitivity ( $S$ ) = 178.8 nm/RIU, FOM = 7.1 RIU $^{-1}$ , LoD = $5.6 \times 10^{-5}$ RIU.
Field Distribution: Near-field analysis reveals that the TE mode is strongly confined within the gap region (where the BeS insert resides), leading to high analyte overlap and sensitivity. In contrast, the TM mode stores energy primarily within the TiO $_2$ bar volume, resulting in lower sensitivity.
Tolerance: Variations in bar height and unit-cell period ( $\pm 10$ nm) cause the largest resonance shifts ( $\sim 20\text{--}40$ nm), but the nominal operating point lies near a local minimum in the relative shift map, suggesting partial compensation for correlated errors. Perturbations in BeS optical constants ( $\pm 5\%$ ) result in resonance shifts $< 2.1$ nm, confirming the design's stability against material property uncertainties.

Significance and Claims
The paper claims that this platform offers a distinct advantage over conventional single-channel dielectric sensors by providing polarization-dependent selectivity. The dual-resonance output allows for the differentiation of analytes based on their optical anisotropy or specific surface interactions, a capability not available in single-resonance sensors.

The authors position the work as a step toward practical NIR biosensing, noting that TiO $_2$ is transparent in the 879–911 nm range where silicon absorbs, making it a superior material choice for this specific wavelength regime. While the $Q$ factors (36–128) are lower than some geometrically asymmetric silicon quasi-BIC designs (which can reach $Q \sim 10^4$ ), the authors argue that the dual-channel output and the NIR operating wavelength are the primary advantages.

The study explicitly states that the work is entirely computational. It does not claim experimental demonstration of the device or direct modeling of specific analyte adsorption in this iteration. However, it highlights that the optical constants for BeS interacting with seven specific lung-disease biomarkers (e.g., 4HHE, isoprene) are already available from prior DFT studies, suggesting a clear, data-ready pathway for future experimental extensions toward analyte-specific gas discrimination. The design is noted to be compatible with standard nanofabrication techniques, specifically Atomic Layer Deposition (ALD) and Electron-Beam Lithography (EBL).

1. The "Two-Track" Light Highway

2. The "Scent Detector" Analogy

3. The "Fingerprint" Advantage

4. What the Numbers Mean

Summary

More like this