Probing Floquet topological phases via non-Hermitian skin effect of reflected waves

This paper demonstrates that the non-Hermitian skin effect of reflected waves in periodically driven Floquet Chern insulators manifests as a gap-dependent Goos-Hänchen shift, which can be quantitatively integrated to directly measure the system's bulk Floquet topological invariants.

The Gist

Imagine a quantum world where the rules of physics are constantly being tweaked by a rhythmic, repeating beat. This is the world of Floquet systems. Think of it like a dance floor where the music (the energy of the system) changes every few seconds. Because the music keeps looping, the dancers (the particles) can form patterns that are impossible in a static room. Some of these patterns are "topological," meaning they are robust and have special properties, like a knot that can't be untied.

The paper by Fangqiao Ye and Haiping Hu explores a specific type of these dancing systems called Floquet Chern insulators. Here is the core discovery, broken down into simple concepts:

1. The Mystery of the "Ghost" Dancers

In normal, static systems, you can tell if a dance floor has a special topological pattern by looking at the crowd in the middle (the "bulk"). But in these rhythmic, time-driven systems, the middle might look completely empty and boring, while the edges are buzzing with special "chiral edge states" (dancers moving in a circle).

The problem is: How do you know the system is special if the middle looks normal? Usually, scientists have to map out the entire dance floor to find the answer, which is hard to do.

2. The Bouncing Ball Experiment

The authors propose a simpler way: Throw a ball at the wall and see how it bounces back.

In their experiment, they imagine sending a wave (like a ripple of water or a sound wave) toward the edge of this rhythmic system. They don't look at the middle; they only watch the reflected wave.

They discovered something strange happens to this reflected wave, which they call the Non-Hermitian Skin Effect (NHSE).

• The Analogy: Imagine throwing a ball at a wall. In a normal room, it bounces straight back. In this special rhythmic room, the ball hits the wall, but instead of bouncing straight back, it gets "sucked" along the wall to one specific corner before finally bouncing back.
• The Result: The reflected wave doesn't just bounce; it gets "skinny" and piles up at the corners of the boundary. This happens because the rhythmic driving of the system creates a one-way street for the wave along the edge.

3. The "Gap" Matters

The system has different "energy gaps" (like different lanes on a highway). The authors found that whether the wave gets "sucked" to the corner or bounces normally depends entirely on which lane (energy gap) the wave is traveling in.

• If the wave is in a "trivial" lane, it bounces normally.
• If the wave is in a "topological" lane, it gets sucked to the corner.

4. Measuring the "Goos-Hänchen" Shift

The paper introduces a way to measure this effect using something called the Goos-Hänchen (GH) shift.

• The Analogy: Imagine you are sliding a puck across a table. If the table is perfectly smooth, it goes straight. But if there is a hidden, invisible current, the puck might slide a few inches to the left or right before it even hits the wall.
• In this study, when the wave hits the boundary, it doesn't reflect from the exact spot it hit. It reflects from a spot slightly shifted to the side.
• The Magic: The authors show that if you add up all these tiny sideways shifts for waves coming in from different angles, the total number you get is a perfect code. It tells you exactly what the topological "knot" is in the middle of the system, even if the middle looks empty.

5. Why This is a Big Deal

Usually, to find out if a system is topologically special, you have to look at the whole system in a complex way (like taking a 3D scan of the whole dance floor).

This paper offers a real-space shortcut. You don't need to see the whole system. You just need to:

1. Send a wave at the edge.
2. Measure how much it shifts sideways when it bounces back.
3. Do the math on that shift.

If the shift adds up to a specific number, you know the system has a special topological phase. This works even for the weirdest "anomalous" phases where the middle of the system looks completely boring.

Summary

The paper reveals that in rhythmic quantum systems, the way a wave bounces off the edge is a secret message. The wave gets "sucked" to the corners and shifts sideways in a specific way. By measuring this shift, you can decode the hidden topological secrets of the system without ever needing to look inside the bulk. It turns a complex quantum puzzle into a simple game of "throw a ball and see where it lands."

Technical Summary

Technical Summary: Probing Floquet Topological Phases via Non-Hermitian Skin Effect of Reflected Waves

Problem Statement
Periodically driven (Floquet) systems host topological phases with no static analogs, such as the anomalous Floquet Chern insulator, where chiral edge states exist even when all bulk Chern numbers vanish. A central challenge in this field is diagnosing these non-equilibrium topological properties. Existing experimental methods, such as momentum-space state tomography or circular dichroism, require extensive measurements across the entire Brillouin zone and precise quantum state manipulation. While scattering formalisms have been used to encode bulk topology in static systems via the non-Hermitian skin effect (NHSE) of reflection matrices, the reflection characteristics of periodically driven systems remain largely unexplored due to their dynamical nature and the cyclic structure of the quasienergy spectrum.

Methodology
The authors investigate the scattering problem of a Floquet Chern insulator using a discrete-time scattering formalism. The study employs a prototypical multi-step driving lattice model (Rudner et al.) defined on a two-dimensional bipartite square lattice with a time-periodic Hamiltonian $H(t) = H(t+\tau)$.

1. Scattering Setup: The driven lattice is coupled to semi-infinite, topologically trivial leads. The scattering matrix $S(\epsilon)$ is derived by relating incoming and outgoing wave amplitudes through a self-consistent condition imposed by the stroboscopic dynamics of the Floquet operator $U$.
2. Reflection Matrix Analysis: The study focuses on the reflection matrix $R(\epsilon, k_y)$ for waves incident from the left lead. The authors analyze the eigenvalues of this matrix in the complex plane under both periodic (PBC) and open (OBC) boundary conditions in the transverse direction.
3. Topological Invariants: The non-Hermitian winding number $\nu(\epsilon)$ of the reflection matrix is calculated and compared against the bulk Floquet invariant $W_3(\epsilon)$, defined in the (2+1)-dimensional momentum-time space.
4. Goos-Hänchen (GH) Shift: Using the stationary phase approximation, the authors derive the transverse spatial shift (GH shift) of a reflected wave packet as a function of incident momentum. They numerically simulate wave-packet dynamics to calculate the momentum-integrated GH shift.

Key Contributions and Results

• Equivalence of Reflection Winding and Bulk Invariant: The paper establishes an exact correspondence between the non-Hermitian winding number of the reflection matrix, $\nu(\epsilon)$, and the bulk Floquet topological invariant, $W_3(\epsilon)$, for a specific quasienergy gap. This equivalence is mediated by boundary resonances: the poles of the reflection matrix in the complex plane correspond to the chiral edge states of the bulk system.
• Gap-Dependent NHSE: The study reveals that the reflection matrix exhibits a non-Hermitian skin effect (NHSE) that is dependent on the quasienergy gap of the incident wave.
  • In the trivial phase ($J=0.5\pi/\tau$), the reflection spectrum lies on the unit circle for both the 0-gap and $\pi$-gap, with no NHSE.
  • In the Floquet Chern insulator phase ($J=1.5\pi/\tau$), the NHSE (spectral collapse inside the unit circle and eigenstate localization at boundaries) appears only for the $\pi$-gap ($\nu(\pi)=1$), while the 0-gap remains trivial.
  • In the anomalous phase ($J=2.5\pi/\tau$), the NHSE appears for both gaps ($\nu(0)=\nu(\pi)=1$), despite the bulk Chern numbers being zero.
• Real-Space Topological Probe: The authors demonstrate that the momentum-integrated Goos-Hänchen shift, $\int \Delta_{GH} dk_{y0}$, quantitatively yields the bulk Floquet invariant $W_3(\epsilon)$. Specifically, the integral equals $-2\pi \nu(\epsilon)$. This provides a direct real-space measurement of the topological invariant without requiring momentum-space tomography.
• Mechanism: The NHSE and the associated GH shift arise from the unidirectional transport of wave amplitude along the interface via chiral edge states before the wave reflects back into the lead. This transverse transport reduces the reflection amplitude, pulling eigenvalues inside the unit circle and causing the spatial shift.

Significance
The paper claims to provide a reliable, real-space scattering approach to identify non-equilibrium topology in driven systems. By linking the reflection phase winding to the bulk Floquet invariant, the work offers a method to probe topological phases, including the anomalous phase, through boundary measurements alone. The authors highlight that this approach is particularly advantageous for complex driving protocols where bulk momentum-space reconstruction is difficult. They suggest that the method is feasible in experimental platforms such as ultracold atoms, photonic crystals, and acoustic metamaterials, where wave packet preparation and spatial shift detection are accessible. Furthermore, because the method relies on real-space scattering, it is potentially applicable to disordered systems where bulk momentum is not a good quantum number.