Topological and fractal defect states in non-Hermitian lattices

This paper establishes a universal framework linking spectral winding topology, fractal structures, and defect-localized states in arbitrary-dimensional non-Hermitian lattices, demonstrating that such states emerge only when the winding number exceeds a defect-size-dependent threshold and exhibit amplified responses under external driving.

The Gist

Imagine a giant, multi-story hotel where the guests are tiny particles of energy. In a normal, "fair" hotel (what physicists call a Hermitian system), guests can move freely between rooms. If you knock on a door, the sound travels evenly, and everyone eventually settles down in a balanced way.

But this paper explores a very strange, "unfair" hotel (a Non-Hermitian system). In this hotel, the doors are rigged. Some doors are wide open, letting guests rush through easily, while others are sticky and hard to open. This creates a one-way flow, like a giant escalator that only goes up.

Here is the simple story of what the researchers discovered in this strange hotel:

1. The "Skin" Effect: Everyone Crowds the Exit

In this rigged hotel, if you have a long hallway with these one-way doors, almost every single guest gets pushed to one end of the hall. They pile up at the exit, ignoring the rest of the building. Physicists call this the Non-Hermitian Skin Effect. It's like a crowd of people all rushing to the emergency exit because the other doors are locked.

2. The New Discovery: The "Defect" Trap

Usually, scientists thought this crowding only happened at the very edges of the building. But this paper asks: What happens if there is a broken wall or a missing room right in the middle of the hallway?

The researchers found that if the "one-way-ness" of the hotel is strong enough, the guests don't just crowd the exit. They also get trapped around the broken wall in the middle. These are the "Skin Defect States."

3. The Secret Code: The "Winding Number" vs. The "Broken Wall Size"

The most exciting part of the paper is the rule they found for when this trapping happens. It's a game of numbers:

• The Winding Number: Imagine the "twist" or "strength" of the one-way flow in the hotel. Let's call this the Twist Score.
• The Broken Wall Size: Imagine the broken wall in the middle is made of several missing rooms. Let's call this the Gap Size.

The Rule: The guests will only get trapped around the broken wall if the Twist Score is higher than the Gap Size.

• If the flow is weak (low Twist Score), the guests ignore the broken wall and just pile up at the exit.
• If the flow is strong (high Twist Score), the guests get stuck around the broken wall.

It's like a river: If the current is gentle, a small rock in the middle won't stop the water. But if the current is a raging torrent, the water will swirl violently around the rock.

4. The Fractal Connection: The "Crazy" Broken Wall

The researchers took this a step further. What if the "broken wall" isn't just a straight line, but a weird, jagged, fractal shape (like a coastline or a snowflake)?

They found that the shape of the broken wall matters. The more "jagged" or "fractal" the wall is, the more "twist" (Twist Score) you need to trap the guests. They created a mathematical formula that links the complexity of the shape directly to the strength of the flow needed to trap the energy.

5. The Amplifier: Making Signals Louder

Finally, they showed why this is useful. If you send a signal (like a whisper or a radio wave) into this hotel, and the "Twist Score" is high enough, the signal doesn't just get trapped; it gets amplified.

Imagine shouting into a tunnel. In a normal tunnel, your voice fades. In this rigged hotel, if you shout near the broken wall, the "one-way" flow catches your voice, bounces it around the defect, and makes it louder and louder as it travels. This could be used to build super-sensitive sensors or powerful amplifiers in the future.

Summary

In short, this paper discovered a new rule for how energy behaves in strange, one-way systems:

1. Strong flows can trap energy around holes in the middle of a system, not just at the edges.
2. Whether this happens depends on a simple math game: Is the flow stronger than the size of the hole?
3. This works even if the hole has a fractal, jagged shape.
4. This effect can be used to boost signals, making it a potential tool for new technologies in optics, electronics, and quantum computing.

It's like realizing that if you turn the water pressure up high enough, a tiny crack in a dam can become the most powerful place in the entire system.

Technical Summary

1. Problem Statement

Non-Hermitian systems exhibit unique topological phases characterized by a spectral winding number, leading to phenomena like the Non-Hermitian Skin Effect (NHSE), where eigenstates accumulate at boundaries. While the NHSE is well-understood in 1D systems with open boundaries, its behavior in higher-dimensional (mD) systems with defects remains unclear.

• The Gap: Existing studies often treat the spectral winding number as a binary indicator (zero vs. non-zero) or focus on boundaries. It is unknown how specific integer values of the winding number relate to the emergence of states localized specifically at defects (rather than just boundaries) in arbitrary dimensions.
• The Question: Can the spectral winding topology be precisely mapped to the existence and characteristics of defect-localized states, and can this relationship be used to characterize the fractal geometry of the defects themselves?

2. Methodology

The authors employ a combination of analytical derivation and numerical simulations to investigate $m$-dimensional non-Hermitian lattices containing $(m-1)$-dimensional defects.

• Model Construction:

  • They start with a minimal $m$D lattice with nearest-neighbor asymmetric hoppings ($t_R, t_L$ along the $x$-direction and $j_R, j_L$ along $y$-directions).
  • Defects are introduced by removing hopping terms between specific sites ($x_d$ and $x_d+1$), effectively creating a "cut" or line defect in the lattice.
  • The system is mapped from an $m$D lattice to an equivalent 1D Hamiltonian ($H_{1D}$) where the transverse dimensions ($y$) are treated as internal degrees of freedom (unit cell components).

• Analytical Framework:

  • Diagonalization: The coupling matrix $J$ (representing transverse hoppings) is diagonalized using a transformation matrix $Q$.
  • Formal Solution: The bulk eigenstates are solved using the non-Bloch band theory (Hatano-Nelson model formalism), yielding solutions of the form $\beta^x$.
  • Boundary Conditions: The authors derive a matrix equation $M\Gamma = 0$ by applying boundary conditions at the defect.
  • Rank Analysis: They analyze the rank of the matrix $M$ in the thermodynamic limit ($N_x \to \infty$). They establish that defect states emerge only when the system is underdetermined, i.e., when the number of unknowns exceeds the rank of the constraint matrix.

• Key Assumption: The derivation assumes the matrix $Q$ is totally non-singular (all its minors are non-zero). This condition is generally satisfied in disordered systems or when the lattice size in the $y$-direction is a prime number.

3. Key Contributions & Results

A. The Threshold Condition for Skin Defect States

The paper establishes a precise correspondence between the spectral winding number ($W_x$) and the defect size.

• The Criterion: Defect-localized states (skin defect states) emerge if and only if the average spectral winding number exceeds a critical threshold determined by the defect geometry:
  $$|W_x| > \frac{L_0}{N_y}$$
  Where:
  • $L_0$ is the length of the lattice along the $y$-direction without defects (i.e., $N_y - L_d$, where $L_d$ is the total defect length).
  • $N_y$ is the total system size in the $y$-direction.
• Significance: Unlike previous studies where the winding number only determined the direction of accumulation, this work shows that the magnitude of the winding number dictates whether states localize at the defect. If $|W_x|$ is too small, the states remain extended.

B. Characterization of Fractal Defects

The authors extend this framework to fractal defects (specifically generalized Cantor sets).

• Fractal Dimension Relation: They derive a direct relationship between the fractal dimension ($D_f$) of the defect and the critical winding number ($|W_x^c|$):
  $$|W_x^c| = 1 - r^{1-D_f}$$
  Where $r$ is the scaling factor of the fractal iteration.
• Result: This allows the topological invariant (winding number) to serve as a probe to measure the fractal dimension of the defect, linking topology directly to geometric complexity.

C. Amplified Steady-State Response

The paper demonstrates a physical observable signature for these states using the Green's function ($G = (E_r - H)^{-1}$).

• Signal Amplification: When the system is driven by an external field, the response at the defect sites is exponentially amplified only when $|W_x| > |W_x^c|$.
• Scaling: The amplification ratio scales as $e^{\kappa N_x}$, where $\kappa$ is positive in the topological regime (defect states) and negative or zero in the trivial regime (extended states). This provides a clear experimental signature for detecting both the topological phase and the fractal nature of the defect.

4. Significance and Impact

1. Universal Framework: The work provides a universal framework for understanding defect-localized states in arbitrary dimensions, independent of specific crystal symmetries, provided the coupling matrix satisfies the non-singularity condition.
2. Topological-Fractal Link: It is the first to establish a rigorous mathematical link between spectral winding topology and the fractal dimension of defects, suggesting that topological invariants can quantify geometric complexity in non-Hermitian systems.
3. Experimental Relevance: The prediction of amplified signal responses at defects offers a practical method for experimental detection. This is applicable to various platforms realizing non-Hermitian physics, including photonic crystals, electrical circuits, and cold atom systems.
4. Beyond Boundaries: It shifts the focus from boundary-localized skin effects to defect-localized effects, revealing that internal geometric imperfections can host robust topological states with specific amplification properties.

In summary, the paper resolves the ambiguity of how spectral winding numbers manifest in defective higher-dimensional systems, proving that defect states emerge only when the topological invariant exceeds a geometric threshold, and demonstrating that this threshold encodes the fractal dimension of the defect itself.