Boundary-sensitive non-Hermiticity of Floquet Hamiltonian: spectral transition and scale-free localization

This paper reports a novel mechanism in one-dimensional Floquet systems where a time-periodic driving protocol induces boundary-sensitive PT symmetry breaking and scale-free localization, characterized by a spectral transition triggered when the quasienergy bandwidth covers the entire frequency Brillouin zone rather than by band touching.

The Gist

Imagine you have a long line of people (a chain of atoms) holding hands. In the world of quantum physics, how these people interact determines the "mood" or energy of the whole group. Usually, if you look at the group from the outside (ignoring the ends), everything looks balanced and predictable. But if you look at the ends, things get weird.

This paper introduces a new, surprising way to make a quantum system behave strangely, but only when you look at its edges. Here is the story in simple terms:

1. The Magic Trick: The "Time-Periodic" Dance

The researchers designed a system where the rules of the game change back and forth in a rhythmic cycle, like a dance.

• Step 1: Everyone passes a note to their left neighbor.
• Step 2: Everyone passes a note to their right neighbor.
• Repeat: They keep doing this over and over.

In physics, this is called a Floquet system. Because the rules change with time, the "average" rulebook (the Hamiltonian) is different from just adding Step 1 and Step 2 together.

2. The Two Different Worlds: The Ring vs. The Line

The most fascinating part of this discovery is how the system behaves depending on how you arrange the people.

• World A: The Ring (Periodic Boundary Conditions)
  Imagine the line of people is actually a circle. The last person holds hands with the first person. In this world, the "left" and "right" moves are perfectly balanced. The system is Hermitian, which is a fancy way of saying it's perfectly stable, balanced, and predictable. Nothing weird happens here.

• World B: The Line (Open Boundary Conditions)
  Now, imagine the circle is cut open. You have a start and an end. The first person has no one to their left, and the last person has no one to their right.
  Because the "left" and "right" moves happen at different times, they don't cancel each other out perfectly at the ends. This creates a glitch or a friction specifically at the boundaries.
  Suddenly, the system becomes Non-Hermitian. It's no longer perfectly balanced. It's like the ends of the line are now "leaking" energy or acting strangely, even though the middle of the line is fine.

3. The "Tipping Point" (PT Symmetry Breaking)

In this weird, unbalanced state, the system has a special "tipping point."

• Below the Tipping Point: The system is still calm. Even though the ends are glitchy, the overall energy levels remain real and stable.
• Above the Tipping Point: The glitch gets too strong. The system snaps into a new phase called PT Symmetry Breaking.
  • The Analogy: Think of a tightrope walker. As long as the wind is light, they stay balanced. But once the wind (the glitch at the ends) gets too strong, they lose balance and start spinning wildly. In physics terms, the energy levels turn "imaginary" (mathematically complex), meaning the system is no longer stable in the usual sense.

The Big Discovery:
Usually, in static systems, this "snap" happens when two energy bands crash into each other. But here, the snap happens because the energy "band" gets so wide that it wraps around the entire universe of possibilities (the frequency zone) and hits itself. It's like a snake eating its own tail; the energy spectrum loops around and collides with itself, causing the instability.

4. The "Scale-Free" Ghosts (Localization)

Once the system snaps into this unstable phase, something magical happens to the people (the quantum states).

• Normal Localization: Usually, if you push a crowd, people pile up at one end.
• Scale-Free Localization: In this new system, the people pile up at the ends, but in a very specific way. If you double the size of the line, the pile-up doesn't just get bigger; it stretches out perfectly to fit the new size.
  • The Metaphor: Imagine a shadow cast by a streetlamp. If you move the streetlamp twice as far away, the shadow stretches to be twice as long, but it looks exactly the same, just bigger. The shape of the "pile-up" at the ends doesn't care about the total size of the system. It is scale-free. This is a unique fingerprint of this new type of physics.

Why Does This Matter?

1. New Physics: It shows that you don't need a permanent "broken" system to get weird quantum effects. You can create them just by dancing (driving the system with time) and looking at the edges.
2. Sensors: Because the system is so sensitive to the edges, it could be used to build incredibly sensitive sensors. A tiny change at the boundary could trigger a massive change in the whole system.
3. Control: It gives scientists a new "recipe" to build quantum materials that act differently depending on whether they are connected in a loop or a line.

In a nutshell: The authors found a way to make a quantum system act perfectly normal in a circle, but act wild and unstable at the edges of a line. This instability creates a unique "ghostly" pile-up of particles at the ends that scales perfectly with the size of the system, offering a new way to control and detect quantum states.

Technical Summary

1. Problem Statement

The paper addresses a fundamental gap in the understanding of non-Hermitian physics within periodically driven (Floquet) systems. While the Non-Hermitian Skin Effect (NHSE) and Parity-Time (PT) symmetry breaking are well-studied in static systems, their interplay in Floquet systems remains underexplored.

• The Challenge: In static non-Hermitian systems, PT symmetry breaking typically requires band touching (exceptional points) induced by increasing non-Hermitian strength. In contrast, the authors seek to identify a mechanism where PT symmetry breaking is driven uniquely by boundary conditions in a system that is Hermitian under Periodic Boundary Conditions (PBC) but becomes non-Hermitian under Open Boundary Conditions (OBC).
• The Goal: To construct a Floquet system where the effective Hamiltonian is Hermitian under PBC but acquires non-Hermitian boundary terms under OBC due to the non-commutativity of time-dependent driving steps, leading to a novel PT symmetry breaking transition and a unique localization phenomenon.

2. Methodology

The authors employ a combination of analytical derivation, numerical simulation, and perturbation theory:

• Model Construction: They design a minimal 1D Floquet system driven by a time-periodic Hamiltonian $H(\tau)$ consisting of two distinct steps ($H_1$ and $H_2$) within a period $T$.
  • $H_1$ and $H_2$ are defined using shift operators ($\hat{L}$ and $\hat{R}$) with a boundary control parameter $\eta$ ($\eta=1$ for PBC, $\eta=0$ for OBC).
  • The driving protocol ensures that under PBC, the operators commute, yielding a Hermitian effective Hamiltonian. Under OBC, the non-commutativity of $\hat{L}$ and $\hat{R}$ at the boundaries generates non-Hermitian terms.
• Theoretical Framework:
  • Floquet Theory: The system is analyzed via the stroboscopic Floquet operator $U_F = e^{-iH_F T}$.
  • Baker-Campbell-Hausdorff (BCH) Expansion: The effective Floquet Hamiltonian $H_F$ is expanded in powers of a dimensionless parameter $\lambda = tT/2$. The authors analyze the commutators $[H_1, H_2]$ to identify the origin of non-Hermitian terms.
  • Perturbation Theory: They treat the non-Hermitian terms arising from OBC as a perturbation to the Hermitian bulk Hamiltonian.
• Generalization: They derive a set of general conditions (involving commutators of hopping matrices) to construct multi-band models that exhibit similar boundary-induced non-Hermiticity.

3. Key Contributions

A. Boundary-Sensitive PT Symmetry Breaking Mechanism

The paper establishes a novel mechanism where PT symmetry breaking is solely induced by boundary conditions.

• Hermitian PBC vs. Non-Hermitian OBC: Under PBC, the system is Hermitian with a purely real quasienergy spectrum. Under OBC, the non-commutativity of driving steps generates effective non-Hermitian boundary terms (specifically, terms proportional to $|1\rangle\langle 1| - |N\rangle\langle N|$).
• Transition Criterion: Unlike static systems where band touching is required, the PT symmetry breaking in this Floquet system occurs when the quasienergy bandwidth expands to cover the entire frequency Brillouin zone ($2\pi/T$).
  • When the bandwidth $< 2\pi/T$, the spectrum is real.
  • When the bandwidth $\ge 2\pi/T$, the band edges fold back into the Brillouin zone due to quasienergy periodicity, creating level crossings. The non-Hermitian boundary terms then induce Exceptional Points (EPs), leading to complex eigenvalues.

B. Scale-Free Localization

In the PT-broken phase, the authors identify a phenomenon termed "scale-free localization."

• Definition: Unlike standard NHSE where eigenstates localize exponentially at boundaries with a localization length independent of system size, here the localization profile is invariant under length-scale transformations.
• Scaling Laws:
  • The imaginary part of the quasienergies scales as $O(1/N)$, where $N$ is the system size.
  • The complex momenta $\beta$ deviate from the unit circle as $|\beta| \approx 1 + \alpha/N$.
  • The wavefunction profile follows $|\psi(x)| \sim e^{\alpha x/N}$, meaning the shape of the eigenstate remains self-similar regardless of the system size $N$.

C. General Construction Framework

The authors provide a "recipe" for constructing a broad class of Floquet systems exhibiting this behavior. They identify three necessary conditions for the driving Hamiltonians $H_1$ and $H_2$:

1. PT Symmetry: The driving steps must satisfy $\hat{P}\hat{K}H_{1,2}\hat{K}^{-1}\hat{P}^{-1} = H_{2,1}$.
2. PBC Hermiticity: The sum $H_1 + H_2$ must be Hermitian, and $[H_1, H_2] = 0$ under PBC.
3. OBC Non-Hermiticity: The commutator $[H_1, H_2]$ must be non-zero under OBC, specifically generating boundary terms that survive the BCH expansion.

4. Key Results

• Critical Threshold: For the minimal model, the critical parameter for PT breaking is $\lambda_c = \pi/2$. This corresponds exactly to the point where the bandwidth of the PBC spectrum reaches $2\pi/T$.
• Finite-Size Scaling: Numerical simulations confirm that the critical threshold $\lambda_c$ asymptotically approaches $\pi/2$ as $N \to \infty$.
• Eigenstate Behavior:
  • In the PT-broken phase, eigenstates exhibit asymmetric localization.
  • The mean position of eigenstates $\langle x \rangle_n$ deviates from the center ($N/2$), and the fraction of localized states remains constant across different system sizes, confirming scale invariance.
• Multi-Band Dynamics:
  • Type-I Models (Commuting bands): PT breaking occurs only when an individual band's width reaches $2\pi/T$.
  • Type-II Models (Non-commuting bands): Inter-band mixing allows PT breaking to occur when the total spectral width reaches $2\pi/T$, triggered by the intersection of the top of the upper band and the bottom of the lower band.

5. Significance and Implications

• Theoretical Advancement: This work challenges the conventional view that non-Hermiticity must be intrinsic to the bulk Hamiltonian. It demonstrates that effective non-Hermiticity can emerge purely from the non-commutativity of time-dependent driving steps at boundaries.
• Distinction from Static Systems: It highlights a fundamental difference between static and Floquet non-Hermitian systems. In Floquet systems, the periodicity of the quasienergy spectrum allows a single band to self-intersect across the Brillouin zone boundary, enabling EPs without the need for distinct band touching in the static limit.
• Experimental Relevance: The "scale-free localization" provides a robust, size-independent signature for experimental detection. The authors suggest that this phenomenon is accessible in current platforms such as photonic quantum walks and synthetic photonic lattices.
• Dynamical Signatures: The paper predicts that an initial wave packet will be progressively amplified and accumulate at the boundaries over timescales proportional to the system size, offering a clear dynamical observable for verifying the PT phase transition.

In summary, the paper uncovers a new class of boundary-induced phase transitions in driven quantum systems, linking the topology of the quasienergy spectrum (winding) directly to the emergence of non-Hermitian physics and scale-invariant localization phenomena.