🔬 mesoscale physics

Mobile Exceptional Points Generate Momentum-Space Switching Domains

This paper demonstrates that mobile exceptional points under cyclic modulation generate momentum-space switching domains that partition the Brillouin zone into regions with distinct band-switching behaviors, a phenomenon theoretically characterized by a band-permutation invariant and experimentally verified in a lossy photonic crystal.

Original authors: Jung-Wan Ryu, Chang-Hwan Yi

Published 2026-05-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Jung-Wan Ryu, Chang-Hwan Yi

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 are walking through a vast, flat city called the Brillouin Zone. In this city, there are two types of "citizens" (which physicists call eigenmodes or wave patterns). Usually, if you walk in a circle and come back to where you started, you expect to find the exact same citizens you left behind.

But in this paper, the authors discover a strange phenomenon where the citizens can swap places just because you walked in a circle.

Here is the story of how they found this out, using simple analogies:

1. The "Ghost" Meeting Points (Exceptional Points)

In normal physics, two things can get close but never truly become the same. But in this special "non-Hermitian" world (think of it as a city with some foggy, leaky corners), there are special spots called Exceptional Points (EPs).

Think of an EP as a magical meeting point where two citizens merge into one. If you walk around this meeting point in a circle, something weird happens: when you return to your starting spot, the two citizens have swapped identities. The one who was "A" is now "B," and the one who was "B" is now "A."

2. The Moving Ghosts

Previous studies only looked at these meeting points as if they were fixed statues in the city. You would walk around a statue, and the swap would happen.

But this paper asks: What if the statues themselves start moving?

The authors imagine a control knob (a parameter called $\phi$ ) that they turn in a circle. As they turn this knob, the "meeting points" (EPs) don't stay still; they march through the city. They trace out a path, like a ghost walking a specific route.

3. The Switching Zones

Here is the big discovery: Because these ghosts are moving, they create zones or districts in the city.

Inside the Ghost's Path: If you are a citizen standing inside the loop that the ghost walked, and you watch the ghost go around once, you will find that you have swapped identities with your neighbor. You are now in a different state.
Outside the Ghost's Path: If you are standing outside that loop, you watch the ghost walk by, but when it returns, you are exactly the same as you were before. No swap happened.

The paper calls these areas "Switching Domains." The boundary between the "Swap Zone" and the "No-Swap Zone" is exactly where the ghost walked.

4. Turning Up the Volume (Modulation Strength)

The authors also found that they can control how big these zones are by turning up the "strength" of the knob (modulation amplitude).

Low Strength: The ghost walks a small, tight circle. Only a few people in the center of the city swap places.
Medium Strength: The ghost walks a wider circle. The "Swap Zone" gets bigger, swallowing up more of the city.
High Strength: The ghost walks so wide that it covers the entire city. Now, everyone swaps places. The whole city has undergone a "Global Band Switching."

5. Testing it with Light

To prove this isn't just math on a page, the authors built a model using light (photonic crystals) and materials that absorb some of the light (lossy materials).

They shone light through this crystal and watched how the light waves behaved. Just like their theory predicted, they saw that depending on where the light was in the "city" (its momentum), the light waves either swapped identities or stayed the same after a cycle. They even saw the "Swap Zones" appear exactly where the math said the moving ghosts would walk.

The Big Picture

In simple terms, this paper shows that if you make the "special meeting points" of a system move around in a circle, you can carve up the entire system into regions where things change and regions where they stay the same.

It's like a dance floor where the DJ (the moving EP) creates a circle. Anyone inside the circle swaps dance partners when the song ends; anyone outside keeps dancing with the same partner. By changing how wildly the DJ moves, you can make the whole dance floor swap partners or just a tiny corner of it.

This gives scientists a new tool: instead of just looking for where the "ghosts" are, they can now engineer the movement of these ghosts to control exactly which parts of a system change and which parts stay stable.

1. Problem Statement

Non-Hermitian systems are characterized by Exceptional Points (EPs), spectral degeneracies where both eigenvalues and eigenvectors coalesce. A well-known phenomenon in non-Hermitian physics is eigenmode switching: when system parameters trace a closed loop encircling a fixed EP, the system's eigenmodes exchange after one cycle.

However, previous research has largely focused on fixed EPs in parameter space, where the central question is whether a specific parameter loop encloses an EP. This paper addresses a distinct and unexplored regime: what happens when EPs themselves become mobile due to the periodic modulation of a control parameter in a spatially extended system (a lattice)? The authors investigate how the motion of EPs across the Brillouin zone alters the global band topology and whether this motion creates new topological structures that partition momentum space.

2. Methodology

The authors employ a combination of analytical modeling, topological invariants, and numerical simulations on realistic physical systems.

Minimal Lattice Model:
They introduce a minimal two-band non-Hermitian lattice model defined by the Hamiltonian $H(\phi, \mathbf{k})$ , where $\mathbf{k} = (k_x, k_y)$ is the crystal momentum and $\phi \in [0, 2\pi]$ is a cyclic control parameter. The model includes onsite potential differences, hopping amplitudes ( $t_x, t_y$ ), and a phase-modulated coupling term $\mu e^{i\phi}$ .
Band-Permutation Invariant ( $W(\mathbf{k})$ ):
To characterize the switching behavior, they define a topological invariant:
$W(\mathbf{k}) = \frac{1}{2\pi i} \oint_0^{2\pi} \partial_\phi \log[\Delta^2(\mathbf{k}, \phi)] d\phi$
where $\Delta(\mathbf{k}, \phi)$ $Δ (k, ϕ)$ is the square root of the discriminant of the Hamiltonian. $W(\mathbf{k})$ $W (k)$ measures the winding number of the squared spectral gap around the origin in the complex plane.
- $W(\mathbf{k}) = 1$ : Eigenmode exchange occurs (switching).
- $W(\mathbf{k}) = 0$ : No exchange occurs.
Extended Parameter Space Analysis:
The authors analyze the system in an extended 3D parameter space $(k_x, k_y, \phi)$ . They derive the condition for EPs ( $\Delta = 0$ ) and project the resulting EP trajectories onto the 2D Brillouin zone to identify the boundaries between switching and non-switching regions.
Photonic Implementation:
To validate the theory, they simulate a 2D photonic crystal with lossy dielectric materials (complex refractive indices). They solve the 2D Helmholtz equation using the boundary element method to compute complex eigenfrequencies and demonstrate the switching domains in a realistic wave system.

3. Key Contributions

Discovery of Momentum-Space Switching Domains: The paper establishes that mobile EPs do not just cause local switching but generate domains in momentum space. The Brillouin zone is partitioned into regions with distinct dynamical behaviors (switching vs. non-switching).
Geometric Origin of Boundaries: The boundaries between these domains are identified as the projections of EP trajectories from the extended parameter space $(k_x, k_y, \phi)$ onto the Brillouin zone. This provides a geometric organizing principle for non-Hermitian band topology.
Generalization of Eigenmode Switching: The work reformulates the question of eigenmode switching from "Does the parameter loop enclose an EP?" to "For which crystal momenta does the spectrum switch branches during a modulation cycle?"
Extension to Higher Windings: In complex band structures (like the photonic crystal), the authors show that overlapping EP trajectories can lead to higher winding numbers ( $W=2$ ), where eigenvalues braid non-trivially but return to their original labels (trivial net permutation).

4. Key Results

Domain Formation: For small modulation strengths ( $\mu$ ), switching occurs only in isolated regions of the Brillouin zone. As $\mu$ increases, these regions expand, merge, and eventually wrap around the Brillouin zone torus.
Global Band Switching: At sufficiently large modulation strengths, the switching domains can cover the entire Brillouin zone ( $W(\mathbf{k})=1$ everywhere), resulting in global band switching for all crystal momenta. In this regime, EPs may no longer exist in the extended parameter space, yet the topological switching persists due to the global spectral topology.
Analytical Boundary Condition: The boundary between switching ( $W=1$ ) and non-switching ( $W=0$ ) regions is analytically derived as:
$|\epsilon_0^2 + (t_x + t_y)(t_x e^{ik_x} + t_y e^{ik_y})| = |t_x + t_y|\mu$
This equation defines the closed curves in the Brillouin zone that separate the domains.
Photonic Realization: The photonic crystal simulation confirms the theoretical predictions. The complex eigenfrequency trajectories show clear braiding for points inside the switching domain ( $W=1$ ) and trivial evolution for points outside ( $W=0$ ). The domain boundaries in the simulation perfectly match the projected EP trajectories.

5. Significance

New Avenue in Non-Hermitian Topology: This work opens a new direction in non-Hermitian physics by shifting focus from static EPs to dynamically moving EPs. It demonstrates that the controlled motion of EPs can be used to engineer spectral topology.
Engineering Control: The findings suggest a mechanism to control wave transport and mode switching in wave systems (photonic, acoustic, electronic) by tuning the modulation strength and frequency. This could lead to novel devices for chiral mode switching, robust signal routing, and topological lasers.
Fundamental Insight: The study bridges the gap between the local topology of EPs and the global topology of the Brillouin zone, showing how the geometric motion of singularities dictates the global permutation of eigenmodes.

In summary, the paper reveals that mobile EPs act as topological organizers in momentum space, creating distinct switching domains whose boundaries are determined by the projection of EP trajectories. This provides a powerful framework for designing non-Hermitian systems with tailored dynamical responses.