Realizing tunable non-Hermitian skin effects in dynamical quantum systems via the relative phase between multiple time-periodic driving

This paper demonstrates that the relative phase between multiple time-periodic driving fields can act as a tunable switch to reactivate and control the emergence and localization direction of non-Hermitian skin effects in quantum systems, even in regimes where they are prohibited in static conditions.

The Gist

Imagine a crowded dance floor where everyone is trying to move in a specific direction, but the rules of the dance are a bit "lopsided." In the world of quantum physics, this lopsidedness is called a Non-Hermitian Skin Effect (NHSE). Normally, in these systems, all the dancers (particles) get pushed to one edge of the room, piling up against the wall instead of spreading out evenly. This is a strange phenomenon that happens when the system loses energy or when movement isn't the same in both directions.

This paper, by Huan-Yu Wang, introduces a clever way to control this "crowding" using time-periodic driving—think of it as a DJ changing the beat of the music—and, most importantly, the relative phase between different beats.

Here is the breakdown of the paper's findings using simple analogies:

1. The "Off Switch" and the "On Switch"

Imagine you have a static room where the dancers are stuck in a pattern that prevents them from piling up at the wall. This is due to a specific symmetry (called Parity-Time or PT symmetry) that acts like a strict bouncer, keeping everyone spread out.

• The Problem: You want to make them pile up (turn on the Skin Effect), but the bouncer won't let you.
• The Solution: The author shows that if you start shaking the room with a rhythmic beat (time-periodic driving), you can trick the bouncer.
• The Secret Knob: The key is the relative phase. Imagine you have two drummers playing. If they hit their drums at the exact same time (phase = 0), the bouncer stays happy, and the dancers stay spread out (Skin Effect is OFF). But, if you tell the second drummer to hit their drum slightly later (changing the phase), the rhythm gets "out of sync." This breaks the bouncer's rules, and suddenly, all the dancers rush to the wall (Skin Effect turns ON).

2. Changing the Direction of the Crowd

Now, imagine a large, square dance floor (a 2D system). Usually, the dancers might only want to pile up against the bottom wall, ignoring the left, right, or top walls.

• The Magic of Phase: The paper shows that by simply adjusting the timing (phase) between the drummers, you can change which wall the dancers prefer.
• The Analogy: Think of the dancers as water flowing in a channel.
  • With Phase A, the water flows and pools at the bottom.
  • With Phase B, the water suddenly shifts and pools at the left side.
  • With Phase C, the water might pool at both the bottom and the left side simultaneously.
• The paper explains that changing the phase doesn't just turn the effect on or off; it rewrites the "map" of the room (the effective Hamiltonian), creating new paths that guide the crowd to different corners.

3. The "Quench" (A Sudden Stop and Start)

The authors also looked at a different way to change the rhythm, called a "quench." Instead of a smooth, continuous beat, imagine the music stops and starts abruptly in different patterns.

• If the music stops and starts in a specific alternating pattern (like a heartbeat), the crowd piles up on one side.
• If you change the pattern to a steady, even rhythm, the crowd shifts to a different side.
• This proves that you don't need complex, smooth waves to control the effect; even simple, jerky changes in timing can steer the particles exactly where you want them.

Summary of the "Takeaway"

The paper demonstrates that in quantum systems, you don't need to rebuild the entire machine to change how particles behave. You just need to tune the timing between different driving forces.

• No Phase Difference? The crowd stays spread out (or stays in one direction).
• Specific Phase Difference? The crowd piles up (Skin Effect turns on).
• Different Phase Difference? The crowd piles up in a different direction.

The authors conclude that this method is not just a theory; it can be built in real-world labs using shaken optical lattices (lasers holding atoms) or electrical circuits. It gives scientists a new, tunable "remote control" to decide exactly where quantum particles will gather, simply by adjusting the timing of the music they dance to.

Technical Summary

Technical Summary: Realizing Tunable Non-Hermitian Skin Effects in Dynamical Quantum Systems via Relative Phase

Problem Statement
Non-Hermitian skin effects (NHSEs) represent a phenomenon in open quantum systems where bulk eigenmodes accumulate at the boundaries of a lattice, typically driven by non-reciprocal tunneling or onsite dissipation. While NHSEs are well-characterized in static systems, their behavior in time-periodically driven (Floquet) systems offers new avenues for control. A specific challenge addressed in this work is the ability to dynamically switch NHSEs on or off, and to control the spatial localization direction of skin modes in higher dimensions, without altering the fundamental non-reciprocal or dissipative parameters of the system. The authors investigate whether the relative phase between multiple time-periodic driving fields can serve as a control mechanism for these properties.

Methodology
The authors employ a theoretical framework combining Floquet theory with non-Hermitian topological analysis. They analyze two primary models:

1. 1D Bipartite Quantum Chain: A static model with Parity-Time (PT) symmetry is constructed, where the constraints inherently prohibit NHSEs (resulting in $|\beta|=1$). The system is subjected to multiple time-periodic drivings on intra-cell tunneling parameters ($t_1$ and $r_1$).
2. 2D Modified Non-Hermitian Qi-Wu-Zhang (NH-QWZ) Model: A two-dimensional lattice model is analyzed to study directional localization.

The core methodology involves:

• Floquet Evolution: Calculating the time-evolution operator $U(T)$ and the effective static Hamiltonian $H_{eff}$ to determine the quasi-energy spectrum and topological invariants.
• Symmetry Analysis: Investigating how the relative phase $\phi$ between driving terms affects the global PT symmetry of the Floquet operator. The authors derive conditions under which the PT symmetry of the instantaneous Hamiltonian is preserved or broken in the effective Floquet operator.
• Topological Invariants: Using the winding number $C$ of the momentum-space complex energy spectrum (point gap topology) to identify the presence of NHSEs.
• Magnus Expansion: In the high-frequency limit, the authors use the Magnus expansion to analytically derive the time-independent effective Hamiltonian and demonstrate how the relative phase alters its spatial structure.
• Quench Dynamics: The study also explores "quench" driving formalisms with different temporal tempos (e.g., Tempo A vs. Tempo B) to simulate the effects of relative phase changes.

Key Contributions and Results

1. Switching NHSE via Relative Phase in 1D:

   • In a static PT-symmetric chain, NHSEs are prohibited. The authors demonstrate that applying time-periodic driving can reactivate NHSEs.
   • Crucially, the relative phase $\phi$ between multiple driving fields acts as a switch.
   • When $\phi = p\pi$ (where $p$ is an integer), the global PT symmetry of the Floquet operator is preserved, and NHSEs are switched off (all bulk modes are delocalized).
   • When $\phi \neq p\pi$ (e.g., $\phi = \pi/2$), the temporal symmetry constraints are broken, the winding number becomes non-zero ($C \neq 0$), and NHSEs are switched on.
   • This mechanism relies on breaking the temporal symmetry of the Floquet operator, a condition that does not hold for systems with particle-hole or time-reversal symmetry in the same manner.

2. Tunable Localization Direction in Higher Dimensions:

   • In 2D systems, NHSEs often exhibit a "favorable localization direction" determined by the lattice geometry or boundary conditions (e.g., localization only on the bottom edge).
   • The authors show that the relative phase $\phi$ can alter the spatial structure of the long-time averaged effective Hamiltonian.
   • For specific phases (e.g., $\phi = p\pi$), the system retains a directional preference (e.g., $C_y \neq 0, C_x = 0$).
   • By changing the phase to a general value (e.g., $\phi = \pi/3$), the effective Hamiltonian's spatial symmetry is modified such that winding numbers become non-zero in both directions ($C_x \neq 0, C_y \neq 0$). This results in skin modes localizing in multiple directions simultaneously, effectively changing the skin density profile.

3. Quench Dynamics and Tempo Control:

   • The study confirms that these effects are not limited to sinusoidal driving. By varying the "tempo" (timing sequence) of quench dynamics between different parameter sets, the authors can replicate the effects of changing the relative phase, interchanging the favored localization direction (e.g., from x-direction to y-direction).

Significance and Claims
The paper claims to provide a general formalism for realizing tunable skin density profiles in dynamical quantum systems. The primary significance lies in demonstrating that the relative phase between driving fields is a powerful control knob that can:

• Artificially reactivate or prohibit NHSEs in systems where static symmetries would otherwise forbid them.
• Dictate the spatial orientation of skin mode localization in higher dimensions without changing the underlying lattice geometry or dissipation rates.

The authors state that their formalisms are generally realizable in diverse optical and mechanical platforms exhibiting Floquet dynamics. Specifically, they suggest feasibility in:

• Shaken dissipative optical lattices: Using dynamical superlattice potentials where the relative phase can be tuned via acousto-optic modulators.
• Discrete time quantum walks.
• Circuit lattices.

The work concludes by paving the way for the artificial design of skin density patterns, offering a method to manipulate non-Hermitian topological phenomena through dynamical control rather than static parameter engineering.