🔬 optics

Three-dimensional confinement of light in photonic crystals without bandgaps

This paper demonstrates that three-dimensional light confinement in photonic crystals without a complete bandgap can be achieved by combining a vanishing photonic density of states at symmetry-protected quadratic degeneracies with a point defect whose symmetry mismatch suppresses coupling to propagating bulk modes.

Original authors: Manxi Shi, Sachin Vaidya, Ali Ghorashi, Steven G. Johnson, Marin Soljačić

Published 2026-07-28

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Manxi Shi, Sachin Vaidya, Ali Ghorashi, Steven G. Johnson, Marin Soljačić

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 trying to trap a beam of light inside a tiny box. In the world of physics, light is a restless traveler; it hates being stuck. Usually, to keep light from escaping, scientists build "mirrors" out of special materials called photonic crystals. Think of these crystals like a perfectly ordered forest of tiny pillars. If the pillars are arranged just right, they create a "forbidden zone" (called a bandgap) where light simply cannot exist, forcing it to bounce around inside a small cavity. This is the standard way to build optical traps, but it's like trying to build a fortress with a very specific, difficult-to-mold clay; if the shape isn't perfect, the light leaks out. For years, scientists wondered: Is there another way to trap light that doesn't require building a perfect, impenetrable wall? What if we could trap light not by blocking the road, but by making the road disappear right where the light wants to go?

This is the puzzle a team of researchers at MIT set out to solve. They asked a bold question: Can we trap light in three dimensions even if the "forbidden zone" doesn't exist? Their answer is a resounding "yes," but it comes with a twist. Instead of building a wall, they engineered a situation where the light has nowhere to go because the "traffic" of light waves in the surrounding material suddenly vanishes at a specific frequency. It's like trying to drive a car into a city where, at a specific moment, all the roads simply cease to exist. But there's a catch: if the light is too similar to the surrounding waves, it might still find a way to wiggle through. So, the researchers also dressed the trapped light in a "costume" (a specific symmetry) that makes it invisible to the surrounding waves, ensuring it can't escape even if the road reappears. They didn't just guess this would work; they used powerful computer simulations to prove that light could indeed get stuck in a tiny, invisible trap inside a crystal that has no gaps at all.

The "Ghost" in the Crystal

To understand how this works, let's picture a photonic crystal not as a solid block, but as a giant, 3D dance floor made of invisible waves. In a normal crystal, light waves dance everywhere, spreading out and flowing through the structure. Usually, to stop them, you need a "bandgap"—a frequency range where the dance floor is completely empty, so the light has nowhere to step. But in this new discovery, the dance floor isn't empty; it's just that at one very specific frequency, the music stops for a split second.

The researchers started with a crystal made of four dielectric rods (think of them as stiff, glass-like sticks) arranged in a cube. They tuned the size of these rods so that at a specific frequency (about 0.345 times the speed of light divided by the rod spacing), the "density of states" drops to zero. In everyday terms, this means that at this exact frequency, there are literally no available spots for light waves to exist in the surrounding material. It's as if the entire dance floor vanishes for a single beat. If you try to shine a light at this frequency, it has nowhere to go in the bulk material.

However, vanishing roads aren't enough on their own. If you just put a light source there, it might still leak out because the surrounding waves could still "talk" to it. To fix this, the team introduced a "point defect"—a tiny spherical bump in the middle of the crystal, like a pebble dropped into a pond. They carefully sized this pebble so that it wanted to vibrate at that exact same frequency where the dance floor vanished.

Here is the magic trick: The light trapped inside the pebble has a different "shape" or "symmetry" than the light waves that could exist in the surrounding crystal. Imagine the surrounding waves are all wearing red shirts and dancing in a circle, while the trapped light is wearing a blue shirt and dancing in a square. Because their "outfits" and dance moves are so different, they can't interact. The surrounding waves simply don't recognize the trapped light, so they can't steal its energy. This is called a "symmetry mismatch."

The Result: A Light That Stays Put

When the researchers ran their computer simulations, the results were striking. They created a 5x5x5 grid of these crystals with a single spherical defect in the center. When they tuned the defect to the right size, the light got stuck. It didn't just hover near the center; it was tightly confined in all three dimensions (up-down, left-right, and front-back).

The simulations showed that the light didn't just stop; it decayed in a very specific way. Instead of fading away exponentially (like a sound getting quieter and quieter), the light's electric field faded away algebraically, following a curve like $r^{-2.61}$ . This is a crucial detail. In physics, for a wave to be truly "trapped" or "bound," it must be able to be contained within a finite space. The math showed that this specific type of fading is fast enough to keep the light trapped, making it a "bound state in the continuum." This is a fancy way of saying the light is stuck inside a sea of waves that could exist, but because of the symmetry mismatch and the vanishing density of states, it refuses to leave.

The team measured the "quality factor" (Q-factor) of this trap, which tells us how long the light stays inside before leaking out. They found that as they made the simulation box bigger, the light stayed trapped longer and longer. Specifically, the quality factor grew with the cube of the system size ( $Q \sim n^{3.06}$ ). This is the smoking gun that proves the light is truly bound. If it were just a temporary resonance that would eventually leak out, the quality factor would stop growing once the box got big enough. But because it kept growing, it confirmed that the light was genuinely stuck in a bound state, even though the surrounding material had no bandgap to hide in.

Why This Matters

This discovery is a big deal because it breaks the old rule that you need a complete "forbidden zone" to trap light in 3D. Before this, scientists thought that without a bandgap, light would always find a way to escape. This paper shows that by engineering the "symmetry" of the trap and the "density" of the surrounding waves, you can create a perfect cage without needing a wall.

The researchers didn't just stop at one crystal shape. They looked at a massive list of possible crystal structures (space groups) and found that many of them could support this kind of trap. They created a "menu" of 200+ crystal types that could potentially work, suggesting that this isn't a fluke of one specific design but a whole new way to think about trapping light.

Of course, this is currently a result from computer simulations. The authors are careful to note that they haven't built this in a lab yet. They suggest that the next step is to use "inverse design" (a method where computers figure out the best shape for you) to find real-world materials that can do this with easier-to-make structures. But the path is now clear: we don't need to build perfect, gap-filled fortresses to trap light anymore. We just need to make the light wear the right outfit and dance to the right beat.

Technical Summary: Three-Dimensional Confinement of Light in Photonic Crystals Without Bandgaps

Problem Statement
Achieving strong three-dimensional (3D) confinement of light in photonic crystals (PhCs) has traditionally relied on the existence of a complete photonic bandgap, which forbids propagation within a specific frequency range. However, realizing such gaps often requires complex geometries that are difficult to fabricate and limit practical applicability. Alternative confinement mechanisms, such as index guiding or plasmonic confinement, typically rely on momentum conservation, resulting in modes that are extended in at least one dimension. While Anderson localization offers a path to 3D confinement without momentum conservation, it lacks deterministic control over spatial location and frequency, and recent studies suggest that longitudinal electric fields may prevent such disorder-induced localization in 3D dielectric media. Consequently, the possibility of achieving full 3D confinement in ordered, all-dielectric systems without a complete bandgap remains an open question.

Methodology
The authors propose a mechanism for light confinement that exploits symmetry-protected quadratic degeneracies in the bulk band structure of a 3D PhC, where the photonic density of states (DOS) vanishes at an isolated frequency. The study utilizes a specific 3D PhC based on a simple cubic lattice (space group #224, $P\bar{3}m$ ) containing four dielectric rods oriented along the body diagonals.

Bulk Structure Analysis: The authors identify a triply degenerate quadratic band touching at the $R$ point of the Brillouin zone. At this frequency ( $\omega_0$ ), the DOS vanishes as $\sqrt{|\omega - \omega_0|}$ .
Defect Engineering: A spherical point defect is introduced at the center of a supercell. This defect breaks the nonsymmorphic symmetries of the original lattice, reducing the space group to #195 ( $P23$ ). Crucially, the reduced symmetry still supports three-dimensional irreducible representations (irreps) at the $\bar{R}$ point, preserving the triply degenerate bulk bands.
Symmetry Mismatch: The defect geometry is tuned (specifically the defect radius, $r_d$ ) so that a singly degenerate defect mode (transforming under a 1D irrep) becomes degenerate in frequency with the triply degenerate bulk modes (transforming under a 3D irrep). Because the irreps are incompatible, symmetry forbids hybridization (coupling) between the defect mode and the bulk modes.
Numerical Validation: The authors employ two numerical methods:
- Plane Wave Expansion (PWE): Using MIT Photonic Bands (MPB) to calculate supercell band structures and verify the absence of avoided crossings between the defect and bulk modes.
- Finite-Difference Time-Domain (FDTD): Using MEEP to simulate the time-domain response, extract complex eigenfrequencies, and calculate quality factors ( $Q$ ) and field profiles.

Key Results

Bound States in the Continuum (BIC): The combination of a vanishing DOS at $\omega_0$ and symmetry-enforced decoupling yields a strongly localized bound defect mode, despite the absence of a spectral gap. This mode is identified as a bound state in the continuum.
Algebraic Localization: Unlike conventional cavity modes that exhibit exponential decay, the defect mode decays algebraically. Field profiles show the electric field amplitude decaying as $E \sim |r|^{-2.61}$ to $|r|^{-2.78}$ . Theoretical analysis using the Lippmann–Schwinger equation explains this: the leading $r^{-1}$ tail of the Green's function vanishes because the overlap integrals between the defect mode and the bulk modes are zero due to symmetry mismatch. The resulting $r^{-5/2}$ decay (implied by the $r^{-2}$ or faster asymptotic behavior) ensures the mode is normalizable in 3D.
Quality Factor Scaling: The quality factor ( $Q$ ) of the defect mode scales cubically with the system size ( $Q \sim L^3$ ) when the defect is tuned to the confinement frequency. This scaling confirms the bound nature of the mode, distinguishing it from extended resonances where $Q$ saturates with system size. In modest system sizes, $Q$ values approach $10^5$ .
Generality: The authors identify a broad set of space groups (summarized in Figure 4) that can support this mechanism. These are groups that, when stripped of nonsymmorphic symmetries by a spherical defect, yield subgroups capable of hosting both multidimensional irreps (for bulk degeneracy) and one-dimensional irreps (for the defect mode).

Significance and Claims
The paper claims to demonstrate a novel route for designing optical cavities in 3D photonic crystals that does not rely on complete photonic bandgaps, momentum conservation, or disorder. By engineering the photonic environment around a defect—specifically by matching a defect mode to a symmetry-protected quadratic degeneracy with a vanishing DOS—the authors achieve strong 3D confinement.

The authors position this work as a general design principle for photonic crystal nanocavities. They emphasize that this approach expands the design space beyond conventional bandgap-based methods and provides a deterministic mechanism for creating high- $Q$ cavities in gapless systems. The work suggests that light trapping can be achieved not by globally eliminating states, but by modifying the spectral and symmetry properties of the cavity's environment. The authors note that while they identified necessary symmetry conditions, the realization of these structures at lower dielectric contrasts and specific geometries remains a target for future inverse design efforts.

The "Ghost" in the Crystal

The Result: A Light That Stays Put

Why This Matters

Technical Summary: Three-Dimensional Confinement of Light in Photonic Crystals Without Bandgaps

More like this