Non-Hermitian Multipole Skin Effects Challenge Localization

This paper demonstrates that while quenched disorder can induce a transition from a non-Hermitian skin effect to many-body localization in systems conserving only U(1) charge, the skin effect remains robust against arbitrary disorder in systems with conserved dipole or higher multipole moments, ensuring delocalization regardless of disorder strength.

The Gist

The Big Picture: A Crowd in a Stormy Room

Imagine a crowded room full of people (these are quantum particles). Usually, if you shake the room up (add disorder or noise), people get confused and get stuck in random spots. They can't move around; they are "localized." This is a famous concept in physics called Anderson Localization.

Now, imagine this room is a bit weird. It's not a normal room; it's a "non-Hermitian" room. In this room, the rules of movement are unfair. If you try to walk to the right, the floor is slippery and you slide easily. If you try to walk to the left, the floor is sticky and you get stuck. This unfairness is called non-reciprocity.

In a clean version of this weird room, something amazing happens: The Skin Effect. Because the floor is slippery to the right, everyone slides to the right wall and piles up there. They don't stay in the middle; they all crowd the edge.

The Big Question: What happens if we shake the room up with a storm (disorder)?

1. Does the storm knock everyone off the wall and scatter them randomly (Localization)?
2. Or does the slippery floor win, and everyone still piles up at the wall (Skin Effect)?

This paper answers that question, but with a twist: it looks at two different types of "people" in the crowd.

---

Scenario 1: The Regular Crowd (Conserving Charge)

The Analogy: Imagine a crowd of individual people. They can move around, but they can't carry heavy objects. They just want to get to the exit.

• The Setup: The room has a slippery floor to the right (non-reciprocity) and a storm shaking the room (disorder).
• The Result: It's a tug-of-war.
  • Weak Storm: The slippery floor wins. Everyone slides to the right wall. This is the Skin Effect.
  • Strong Storm: The chaos of the storm wins. People get stuck in random spots, ignoring the slippery floor. They are Localized.
• The Takeaway: For regular particles, disorder can kill the skin effect. If the noise is loud enough, the crowd stops piling up at the wall and gets scattered.

Scenario 2: The "Dipole" Crowd (Conserving Multipole Moments)

The Analogy: Now, imagine the people in the room are holding hands in pairs (dipoles). They are stuck together. They can't just move one person; they have to move the pair. Furthermore, the room has a rule: "You cannot change the total balance of these pairs."

This is the Multipole Skin Effect. The paper looks at what happens when these "hand-holding pairs" are in the stormy, slippery room.

• The Twist: In the regular crowd, disorder wins easily. But for these pairs, the physics is different.
• The Result: The slippery floor ALWAYS wins.
  • No matter how hard the storm shakes the room (even if the disorder is massive), the pairs always slide to the edges.
  • They don't just pile up; they stretch out. The pairs on the left stretch to the left wall, and the pairs on the right stretch to the right wall.
• The Counter-Intuitive Surprise: Usually, if you add "kinetic constraints" (like being stuck in pairs) and "disorder" (a storm), you expect the system to freeze completely. You'd think, "If they can't move easily and the room is chaotic, they must be stuck!"
  • But the paper shows the opposite: The non-reciprocal nature of the room (the slippery floor) is so powerful that it breaks the "frozen" state. The system remains delocalized (moving) and the pairs always find the walls, regardless of how strong the storm is.

Why Does This Happen? (The "Super-Sticky" Analogy)

The authors explain this using a mathematical trick called a Similarity Transformation. Think of it like a special pair of glasses.

1. For Regular People: If you put on the glasses, the slippery floor disappears, and you just see a normal room with a storm. In a normal room with a storm, people get stuck. So, the skin effect breaks.
2. For the Pairs (Dipoles): When you put on the glasses for the pairs, the "slippery floor" doesn't just disappear; it turns into a super-powerful magnet that pulls the pairs to the edges.
   • The paper argues that for these pairs, the "pull" to the edge grows so fast (mathematically, it's "super-exponential") that it crushes the chaos of the storm.
   • Even if the storm is a hurricane, the magnet is a black hole. The pairs must go to the edge.

The "Periodic Room" (No Walls)

What if the room has no walls? (Periodic Boundary Conditions).

• Regular Crowd: If there are no walls, the slippery floor just makes everyone run in a circle. But if the storm is strong, they stop running and get stuck in place.
• The Pairs: Even with no walls, the pairs never get stuck. They keep running in a circle forever, creating a persistent current. The storm cannot stop them.

Why Should We Care?

1. New Physics: It challenges our understanding of how disorder works. We thought disorder always freezes things. This paper says, "Not if you have these special symmetry rules!"
2. Real-World Applications:
   • Cold Atoms: Scientists can create these "slippery floors" in labs using lasers and magnetic fields. They can test if these "unfreezable" currents really exist.
   • Circuits: Engineers are building "topoelectric circuits" (circuits that act like these quantum systems). This research suggests that even if the circuit is messy or has defects, the signal might still flow perfectly in one direction because of these skin effects.

Summary in One Sentence

While a stormy, messy room can scatter a crowd of regular people, a crowd of "hand-holding pairs" in a room with unfair, slippery floors will always slide to the edges, no matter how chaotic the storm gets, effectively defying the laws of localization.

Technical Summary

1. Problem Statement

The paper investigates the interplay between disorder, interactions, and non-Hermiticity in quantum systems. Specifically, it addresses the fate of the Non-Hermitian Skin Effect (NHSE) in the presence of quenched disorder and many-body interactions.

• Context: In Hermitian systems, disorder leads to Anderson localization (non-interacting) or Many-Body Localization (MBL, interacting), where eigenstates are localized and transport is suppressed. In non-Hermitian systems, the NHSE causes an extensive number of eigenstates to accumulate at boundaries due to non-reciprocal hopping.
• The Gap: While the stability of the NHSE against disorder is understood for non-interacting systems (Hatano-Nelson model), the behavior of interacting systems and, crucially, systems with multipole conservation (e.g., dipole or quadrupole conservation) remains unclear. Multipole-conserving systems are intrinsically interacting and possess kinetic constraints that hinder particle motion.
• Core Question: Does disorder induce a transition from a skin-effect phase to a localized (MBL) phase in multipole-conserving non-Hermitian systems, similar to the charge-conserving case?

2. Methodology

The authors employ a combination of analytical arguments based on similarity transformations and numerical simulations (Exact Diagonalization).

• Analytical Framework:

  • Similarity Transformation: The authors generalize the Hatano-Nelson argument. They utilize a similarity transformation $S$ that maps the non-Hermitian Hamiltonian $H(g)$ to a Hermitian counterpart $H(0)$ in Open Boundary Conditions (OBC).
  • Local Integrals of Motion (LIOMs): For interacting systems, they analyze the structure of LIOMs. They apply the similarity transformation to the Hermitian LIOMs of the MBL phase to derive the structure of non-Hermitian LIOMs.
  • Scaling Arguments: They examine how the non-Hermitian prefactors (e.g., $e^{g j}$ for charge, $e^{g j^2}$ for dipoles) compete with the exponential decay of localized wavefunctions ($e^{-|j|/\xi}$).

• Numerical Models:

  • Charge-Conserving Model: A generalized Hatano-Nelson model with nearest-neighbor hopping and on-site disorder, extended to include nearest-neighbor interactions ($V \sum n_j n_{j+1}$).
  • Dipole-Conserving Model: A model with constrained hopping (e.g., $c^\dagger_j c_{j+1} c_{j+2} c^\dagger_{j+3}$) that conserves both particle number $N$ and dipole moment $P$.
  • Observables:
    • Order Parameters: Renormalized dipole/quadrupole moments (for OBC) and the fraction of eigenvalues with non-zero imaginary parts ($f_{imag}$) (for Periodic Boundary Conditions, PBC).
    • Dynamics: Time evolution of the charge imbalance $I(t)$ starting from a Néel state.
    • Entanglement: Bipartite von Neumann entropy to distinguish between area-law and volume-law scaling.

3. Key Contributions and Results

A. Charge-Conserving Systems (U(1) Symmetry)

The authors successfully generalize the Hatano-Nelson argument to the interacting case.

• Phase Transition: In systems conserving only U(1) charge, a transition exists between:
  1. Skin Effect Phase (Weak Disorder): Eigenstates cluster at the boundary (OBC) or exhibit persistent currents (PBC).
  2. Many-Body Localized (MBL) Phase (Strong Disorder): Eigenstates localize at random positions, and transport is frozen.
• Order Parameter: The renormalized dipole moment serves as an order parameter. It is extensive in the skin phase but vanishes in the MBL phase.
• Entanglement Transition: The transition corresponds to a change in entanglement dynamics. The MBL phase exhibits volume-law entanglement (due to interaction-induced dephasing), while the skin phase exhibits area-law entanglement (due to the projection onto maximal dipole states).
• PBC Behavior: In the skin phase, PBC leads to delocalization (complex energy spectrum). In the MBL phase, PBC states remain real and localized, unaffected by boundary conditions.

B. Multipole-Conserving Systems (Dipole and Higher)

The most significant finding concerns systems conserving higher multipoles (e.g., dipole moment).

• Stability Against Disorder: Unlike charge-conserving systems, the multipole skin effect is stable against arbitrary disorder strength. There is no transition to an MBL phase.
• Mechanism:
  • In the dipole-conserving case, the similarity transformation involves the quadrupole moment ($S \sim e^{g \sum j^2 n_j}$).
  • The non-Hermitian prefactor scales as $e^{g j^2}$, which grows super-exponentially with distance from the center.
  • Even though disorder induces exponential localization ($e^{-|j|/\xi}$) in the Hermitian limit, the super-exponential growth of the non-Hermitian factor dominates for any $g > 0$.
  • Consequently, LIOMs are pushed to the boundaries, preventing Anderson/MBL localization.
• Delocalization: Under Periodic Boundary Conditions (PBC), the system is always delocalized with a persistent dipole current, regardless of disorder strength. This is counterintuitive because multipole conservation usually imposes kinetic constraints that hinder motion, and disorder usually freezes motion; here, non-Hermiticity overcomes both.

C. Generalization to Higher Multipoles

The authors argue that for any $m$-pole conserving system, the skin effect involves maximizing the $(m+1)$-pole moment. The similarity transformation scales as $e^{g \sum j^{m+1} n_j}$. For $m \geq 1$, this super-exponential scaling always overwhelms the disorder-induced localization, ensuring the system remains delocalized.

4. Significance and Implications

• Fundamental Physics: The work challenges the conventional wisdom that strong disorder inevitably leads to localization in interacting systems. It demonstrates that non-Hermiticity can act as a "delocalizing agent" that is robust against both disorder and kinetic constraints (multipole conservation).
• New Phase of Matter: It identifies a robust "Multipole Skin Phase" that exists even in the presence of strong disorder, distinct from the standard NHSE and MBL phases.
• Experimental Relevance: The results are relevant for cold atom experiments (where dipole conservation is approximate but long-lived) and metamaterials (topoelectric circuits, acoustic lattices). The authors suggest that the stability of the dipole skin effect makes it a promising candidate for experimental observation, as it does not require fine-tuning of disorder strength.
• Entanglement Dynamics: The paper highlights a unique entanglement transition (Area-to-Volume law) driven by the competition between disorder and non-Hermitian skin effects, offering a new diagnostic tool for non-Hermitian many-body physics.

5. Conclusion

The paper establishes that while disorder can destroy the skin effect in charge-conserving non-Hermitian systems (leading to MBL), it cannot destroy the skin effect in systems with multipole conservation. The super-exponential nature of the multipole skin effect ensures that the system remains delocalized and supports persistent currents, even in the presence of strong disorder and stringent kinetic constraints. This reveals a profound robustness of non-Hermitian topology in interacting, constrained quantum systems.