Controlling energy spectra and skin effect via boundary conditions in non-Hermitian lattices

This paper demonstrates that generalized boundary conditions with complex hopping amplitudes in the Hatano-Nelson model can be precisely tuned to control the non-Hermitian skin effect, determine the emergence of exceptional points, and dictate whether the energy spectrum remains real or becomes complex.

The Gist

Imagine you have a long, straight hallway made of tiles. On this hallway, there are little balls (representing particles or energy) that can hop from one tile to the next.

In the normal world (what physicists call "Hermitian" systems), if you throw a ball down the hallway, it behaves predictably. It might bounce off the walls, but it generally spreads out evenly, and the energy it has is a simple, real number (like 5 joules, not 5 plus a mystery number).

But in this paper, the authors are playing with a strange, "non-Hermitian" hallway. In this version, the rules are different:

1. The "One-Way" Street: The balls prefer to hop to the right more easily than to the left. It's like a hallway with a strong wind blowing one way.
2. The "Skin Effect": Because of this one-way preference, if you let the balls loose in a hallway with open ends, they don't spread out. Instead, they all get sucked to one end of the hallway and pile up there. This is called the Non-Hermitian Skin Effect. It's like a crowd of people all rushing to the exit and getting stuck in the doorway.

The Big Discovery: The "Magic Door"

Usually, physicists think you can only control this behavior by changing the whole hallway (making it a loop or leaving it open). But this paper says: "No, you just need to tweak the door."

The authors looked at what happens at the very ends of the hallway (the boundary). They asked: What if the connection between the last tile and the first tile isn't just "open" or "closed," but something in between?

They introduced a "Magic Door" (Generalized Boundary Conditions) that can be:

• Open: The balls fall off the edge.
• Closed (Loop): The balls go from the end back to the start.
• The "Magic" Middle: A door that is slightly ajar, or maybe slightly locked, or has a weird color (a complex phase) to it.

The Three Cool Things They Found

1. The "Ghost" vs. The "Real" Energy
In these strange hallways, the energy of the balls can sometimes become "imaginary" (a mathematical concept that feels like a ghost—unreal and unstable).

• The Finding: By adjusting the "Magic Door" just right, they can force the energy to become real and stable again, even if the hallway has that one-way wind.
• The Analogy: Imagine a spinning top that is wobbling so much it's about to fall over (unstable/ghost energy). By tightening a specific screw on the base (the boundary condition), the top suddenly spins perfectly straight and stable (real energy).

2. The "Traffic Jam" Switch (Controlling the Skin Effect)
Remember how the balls usually pile up at one end?

• The Finding: The authors showed that by turning the "Magic Door" knob, they can decide where the balls pile up. They can make them pile up on the left, on the right, or make them spread out evenly like normal.
• The Analogy: Think of a crowd of people in a hallway. Usually, they all rush to the exit. But if you change the color of the exit sign (the boundary condition), you can make them rush to the entrance instead, or make them stand calmly in the middle. You are controlling the crowd's behavior with a single dial.

3. The "Size Matters" Surprise
Usually, in physics, if you have a huge hallway, the rules are the same as a tiny hallway.

• The Finding: Here, the rules change depending on how many tiles (how big the system is) you have. The "Magic Door" settings that work for a hallway with 10 tiles might not work for 100 tiles.
• The Analogy: It's like a magic trick that only works if you have exactly 10 cards in your deck. If you add an 11th card, the trick fails. This means the size of the device matters a lot for these quantum effects.

Why Does This Matter?

This isn't just about math games. It's about building better quantum devices.

Imagine you are building a new type of computer or a sensor. You want to control exactly where the information (the energy) goes.

• If you want the signal to stay in one spot (like a memory), you can tune the boundary to create a "skin effect" pile-up.
• If you want the signal to be stable and not glitch out, you can tune the boundary to make the energy "real."

In short: The authors found that you don't need to rebuild the whole machine to change how it works. You just need to tweak the door at the end. By turning a single knob on the boundary, you can switch between stable and unstable states, and decide exactly where the energy lives. It's like having a master key that controls the entire quantum world just by adjusting the lock on the front door.

Technical Summary

1. Problem Statement

Non-Hermitian systems exhibit unique phenomena such as the non-Hermitian skin effect (NHSE) and exceptional points (EPs), where eigenvalues and eigenvectors coalesce. While the behavior of these systems under standard Periodic Boundary Conditions (PBC) and Open Boundary Conditions (OBC) is well-understood (e.g., the Hatano-Nelson model shows complex spectra under PBC and real spectra with skin localization under OBC), the impact of Generalized Boundary Conditions (GBC) remains largely unexplored.

The central problem addressed is: How can generalized boundary conditions, specifically those involving complex hopping amplitudes at the lattice edges, be used to control the reality of the energy spectrum, the emergence of exceptional points, and the localization properties (skin effect) of eigenmodes?

2. Methodology

The authors analyze the Hatano-Nelson model, a paradigmatic one-dimensional non-Hermitian lattice where non-Hermiticity arises from asymmetric hopping amplitudes ($t_L \neq t_R$).

• Hamiltonian Formulation: They introduce a Hamiltonian $H$ with generalized boundary conditions defined by complex parameters $\alpha_L$ and $\alpha_R$ at the link between sites $N$ and $1$:
  $$H = \sum_{n=1}^{N-1} (t_R c^\dagger_n c_{n+1} + t_L c^\dagger_{n+1} c_n) + (\alpha_R t_R c^\dagger_N c_1 + \alpha_L t_L c^\dagger_1 c_N)$$
  Here, $\alpha_{L,R}$ interpolate between PBC ($\alpha=1$), OBC ($\alpha=0$), and anti-periodic conditions ($\alpha=-1$), or introduce "defective" links with arbitrary complex phases.

• Similarity Transformation: The core analytical tool is a similarity transformation defined by $c_n \to e^{-\frac{1}{2}qn} \tilde{c}_n$ and $c^\dagger_n \to e^{\frac{1}{2}qn} \tilde{c}^\dagger_n$, where $t_R/t_L = e^q$ ($q \in \mathbb{C}$).

  • This maps the non-Hermitian Hamiltonian $H$ to a transformed Hamiltonian $\tilde{H}$.
  • Crucially, $H$ and $\tilde{H}$ are isospectral (share the same eigenvalues) but not unitarily equivalent.
  • The transformation allows the authors to determine conditions under which $\tilde{H}$ becomes Hermitian, thereby guaranteeing a real spectrum for $H$.

• Parameter Space Analysis: The study systematically varies the boundary parameters $\alpha_L, \alpha_R$ (parameterized by magnitude $\rho$ and phase $\phi$) and system size $N$ to map the transition between real and complex spectra and analyze eigenmode localization.

3. Key Contributions

1. Generalized Boundary Control: The paper establishes that GBCs, implemented via a single tunable parameter (the boundary hopping amplitude), can precisely control the transition between real and complex energy spectra.
2. Exact Conditions for Real Spectra: The authors derive an exact condition (Eq. 4) for the spectrum to remain real despite the system being non-Hermitian:
   $$\alpha_L = 1/\alpha_R = e^{i\phi} e^{\frac{1}{2}qN}$$
   This condition effectively "cancels" the non-Hermiticity at the boundary, rendering the transformed Hamiltonian $\tilde{H}$ Hermitian.
3. Size-Dependent Physics: A novel finding is that the condition for a real spectrum is size-dependent ($N$-dependent). This implies that the spectral properties of non-Hermitian systems are highly sensitive to finite-size effects, a feature that vanishes in the thermodynamic limit ($N \to \infty$) where $\alpha \to 0$.
4. Skin Effect Reversal: The study demonstrates that the direction and presence of the skin effect can be reversed or suppressed by tuning the boundary parameters, specifically the parameter $\rho$ in the relation $\alpha_L = 1/\alpha_R = e^{i\phi} e^{\frac{1}{2}\rho qN}$.

4. Key Results

• Exceptional Points (EPs): The transition from real to complex spectra occurs at exceptional points. For finite lattices, these EPs appear as square-root singularities in the energy spectrum as a function of the boundary hopping amplitudes. The splitting of eigenvalues at these points is controlled solely by boundary parameters.
• Spectral Trajectories:
  • When $\rho = 1$ (satisfying the real spectrum condition), the energy spectrum is purely real.
  • When $\rho \neq 1$, the spectrum becomes complex. The trajectories of eigenvalues in the complex plane swirl around the origin, with the rotation angle increasing with the imaginary part of $q$.
  • The topology of these trajectories depends on the phase $\phi$.
• Localization (Skin Effect):
  • Hamiltonian $H$: Exhibits skin effect (exponential localization at edges) for any $\rho \neq 0$. The localization is at the left or right boundary depending on the sign of $\text{Re}(q)$.
  • Hamiltonian $\tilde{H}$: Exhibits skin effect for any $\rho \neq 1$.
  • Reversal: By tuning $\rho$ across the value 1, the localization of the skin modes in $\tilde{H}$ can be reversed (e.g., from right-edge to left-edge localization) by adjusting the hopping between only two sites.
• Degeneracy: The spectra show specific degeneracies (e.g., double degeneracy for $N=10$) and symmetries under the transformation $\rho \to 2-\rho$ when $\phi = 0$ or $\pi$.

5. Significance and Implications

• Engineering Quantum Devices: The findings provide a framework for engineering quantum lattice models with tailored spectral and localization features. Since the control requires tuning only one parameter (the boundary link), it is highly feasible for experimental realization in photonic, acoustic, or cold-atom systems.
• Fundamental Understanding: The work reveals that non-Hermitian phenomena (like the skin effect and PT-symmetry breaking) are not intrinsic solely to the bulk but are extremely sensitive to boundary conditions and system size.
• Defective Links: The model offers a theoretical description of a single impurity in an otherwise translationally invariant system, showing how a "defective" link can induce or suppress skin effects and control spectral reality.
• Finite-Size Effects: The size-dependence of the real-spectrum condition highlights a unique regime for finite non-Hermitian systems that is distinct from the thermodynamic limit, opening new avenues for studying non-Hermitian topological phases in mesoscopic systems.

In summary, this paper demonstrates that boundary conditions are a powerful, tunable knob for manipulating the fundamental properties of non-Hermitian systems, allowing for the precise control of spectral reality and eigenmode localization through simple adjustments of boundary hopping amplitudes.