Observation of Erratic Non-Hermitian Skin Effect in Phononic Crystals

This paper reports the experimental observation of the erratic non-Hermitian skin effect (ENHSE) in phononic crystals, demonstrating that disorder-induced localization occurs at the local maxima of cumulative gauge fields and can be controlled by tuning staggered disorder strengths.

The Gist

Imagine you are playing a game of "Follow the Leader" in a massive, crowded hallway. Usually, in physics, waves (like sound or light) behave in predictable ways: they either spread out evenly like a crowd, or they get stuck at the very ends of the hallway (the walls).

This paper describes a strange, new way that waves can get "trapped" in the middle of the hallway, in unpredictable, "erratic" spots. The scientists call this the Erratic Non-Hermitian Skin Effect (ENHSE).

Here is the breakdown of how it works using some everyday analogies:

1. The "One-Way Street" Problem (Non-Hermiticity)

In a normal world, if you push a swing, it moves back and forth equally. But in "Non-Hermitian" physics, the rules are skewed. Imagine a hallway where the floor is a series of one-way escalators. If you try to walk left, it’s easy; if you try to walk right, the escalator fights you. This "unbalanced" energy is what makes the waves behave strangely.

2. The "Drunken Walk" (The Disorder)

Now, imagine these one-way escalators aren't all pointing the same way. Some point left, some point right, and they are scattered randomly.

Think of a person taking a "Drunken Walk" through a city. They take a step left, then two steps right, then one step left. If you track their path, they might wander for a while, but eventually, they will hit a "peak" distance—a point where they went further left or right than ever before.

The researchers discovered that sound waves in their crystal act just like that drunken walker. Instead of hitting the walls of the hallway, the waves get "trapped" at the exact spots where the "random walk" of the energy reaches its extreme high or low points. These are the "Erratic Peaks."

3. The "Remote Control" (The Dimerized Model)

The most exciting part of the paper is that the scientists didn't just observe this randomness; they learned how to steer it.

Imagine the hallway is made of pairs of rooms (this is the "dimerized" part). By slightly changing the "strength" of the one-way escalators in the even-numbered rooms versus the odd-numbered rooms, they can act like a remote control.

• If they tweak the "odd" rooms, the sound waves get trapped in the odd rooms.
• If they tweak the "even" rooms, the sound waves jump over and get trapped in the even rooms.

Why does this matter?

Normally, if you want to trap energy, you have to build a box or a wall. But this research shows that by using "engineered chaos" (smartly designed randomness), we can trap waves anywhere we want inside a material without needing physical boundaries.

In short: The researchers found a way to use "organized randomness" to catch sound waves in mid-air, turning a chaotic "drunken walk" into a precise tool for controlling energy.

Technical Summary

Technical Summary: Observation of Erratic Non-Hermitian Skin Effect in Phononic Crystals

1. Problem Statement

The conventional Non-Hermitian Skin Effect (NHSE) is characterized by the accumulation of bulk eigenstates at the boundaries of a system, a phenomenon that invalidates standard Bloch-band theory and necessitates a generalized bulk-boundary correspondence. While NHSE is well-documented, the interplay between non-reciprocity and disorder introduces a more complex localization regime known as the Erratic Non-Hermitian Skin Effect (ENHSE).

In ENHSE, the localization does not occur at the boundaries but rather at random positions within the bulk. Previous studies have theoretically predicted this, and recent reports have observed it in undimerized lattices. However, the experimental implementation of ENHSE—specifically the ability to actively and selectively manipulate these erratic localization peaks (satellite peaks)—has remained an open challenge due to the difficulty of tuning non-reciprocal couplings while maintaining global reciprocity.

2. Methodology

The researchers employed an acoustic platform to realize and study the ENHSE using two distinct models:

• The Disordered Hatano–Nelson Model: A 1D lattice where non-reciprocal hopping amplitudes ($J_R$ and $J_L$) are modulated by a stochastic imaginary gauge field $\{h_n\}$. This field is drawn from a Bernoulli distribution, creating random non-reciprocal hopping that preserves global reciprocity (mean bias $\bar{h} = 0$).
• The Disordered Su–Schrieffer–Heeger (SSH) Model: A dimerized lattice where disorder is introduced into both intracell and intercell couplings. This allows for two independent channels of stochastic non-reciprocity.
• Experimental Implementation: The acoustic crystal consists of resonant cavities (lattice sites) coupled via connecting tubes (reciprocal coupling $\kappa_0$). Non-reciprocal coupling ($\kappa$) is implemented using a microphone–speaker–amplifier circuit at each site. By controlling the amplifier gain, the researchers can induce unidirectional energy flow between adjacent cavities.
• Analysis Techniques: The team used real-space winding numbers to characterize topology, random-walk extreme-value statistics to predict peak positions, and time-evolution simulations (Crank–Nicolson method) to observe wave packet dynamics.

3. Key Contributions

• Experimental Observation of ENHSE: The first systematic experimental demonstration of ENHSE in a phononic crystal, confirming that localization is driven by the extreme-value statistics of the cumulative imaginary gauge field rather than boundary conditions or topology.
• Validation of the Stochastic Mechanism: They proved that the spatial positions of localized peaks (both main and satellite) correspond precisely to the local maxima of the cumulative random-walk trajectory of the gauge field.
• Active Control of Sublattice Polarization: By utilizing the dimerized SSH geometry, they demonstrated that the "erratic" peaks can be deterministically steered to specific sublattices (even or odd sites) by tuning the statistical bias of the staggered disorder.

4. Results

• Bulk Localization: In the Hatano–Nelson model, acoustic pressure distributions showed that energy accumulates at random bulk sites. Crucially, this localization was independent of the excitation position, confirming it is a property of the eigenstates.
• Agreement with Theory: Experimental acoustic pressure profiles showed excellent agreement with numerical simulations, with peaks appearing at the predicted extremal points of the random-walk trajectories.
• Satellite Peak Manipulation: In the SSH model, the researchers successfully demonstrated "selective manipulation." By adjusting the probability and strength of the disorder in intracell vs. intercell bonds, they could shift the concentration of wave functions from sublattice A to sublattice B. This was verified through histograms of logarithmic peak amplitudes, which showed clear segregation based on the engineered disorder.
• Topology vs. Localization: The study showed that while the system may possess a non-trivial point-gap topology, the ENHSE localization in the weak-bias regime is dominated by stochastic fluctuations rather than topological invariants.

5. Significance

This work represents a significant advancement in non-Hermitian physics by moving beyond the observation of passive phenomena toward the active engineering of disorder.

1. New Physics: It establishes ENHSE as a fundamentally distinct localization mechanism, separate from both Anderson localization (which is uniform/exponential) and standard NHSE (which is boundary-driven).
2. Wave Control: It provides a new route for disorder-engineered wave control, where randomness is not a nuisance to be suppressed but a tool to be utilized for programmable wave confinement and steering.
3. Platform Versatility: The methodology of using active electronic components (amplifiers/microphones) to simulate non-Hermitian gauge fields provides a scalable blueprint for exploring complex non-Hermitian dynamics in other classical wave systems (photonic, electrical, etc.).