Spectral Riemann Sheet Topology of Gapped Non-Hermitian Systems

This paper demonstrates that threading exceptional points across the Brillouin zone boundary in gapped non-Hermitian systems annihilates them to form non-trivially closed branch cuts, thereby establishing distinct topological configurations of complex energy spectra that can only be interchanged by closing the spectral gap.

The Gist

The Big Idea: A Map with Hidden Doors

Imagine you are looking at a map of a strange, magical world. In normal physics (Hermitian systems), this map is flat and simple: every location has a clear, single height. But in Non-Hermitian systems (the subject of this paper), the map is more like a multi-layered cake or a spiral staircase. The "height" of the land isn't just a number; it's a complex value that can twist and turn.

Usually, on this twisted map, there are special "knots" or "tangles" called Exceptional Points (EPs). If you walk around these knots, the layers of your map swap places. In the past, scientists focused on these knots.

This paper, however, asks a different question: What happens if we untie the knots but leave the twist in the map?

The authors show that even after the knots (EPs) disappear, the map can remain "twisted" in a way that is topologically protected. They call these twists Closed Fermi Cuts.

The Story of the Thread and the Donut

To understand how this works, imagine the map is drawn on the surface of a donut (a torus). This donut has two holes: one going through the middle, and one going around the outside.

1. Creating the Knots: First, the scientists create a pair of "knots" (EPs) on the map. These knots are connected by a red line called a Fermi cut. Think of this line as a zipper that separates the two layers of the map. As long as the knots exist, the zipper is stuck open.
2. The Journey: Now, imagine dragging one of the knots across the entire donut, all the way around the hole, and bringing it back to meet its partner on the other side of the boundary.
3. The Snap: When the two knots meet, they annihilate each other and vanish. In a normal situation, the zipper (the Fermi cut) would disappear too, and the map would flatten out.
4. The Surprise: But because the knot traveled all the way around the donut's hole, the zipper doesn't disappear. Instead, it snaps shut into a closed loop that circles the donut.

Now, the map has no knots (it is "gapped" and smooth), but it still has a permanent, unbreakable loop of a zipper running around it. You cannot remove this loop without tearing the map or closing the gap. This is the Closed Fermi Cut.

The Four Possible Worlds

The authors discovered that for systems with a specific symmetry (Time-Reversal Symmetry), there are only four distinct ways this map can be twisted. They compare this to a famous puzzle in computer science called the Toric Code.

• The Toric Code Analogy: Imagine a giant chessboard wrapped around a donut. You can flip the colors of the squares along a line that goes around the donut. You can do this for the "horizontal" loop, the "vertical" loop, both, or neither. This creates four unique, stable patterns.
• The Physics Analogy: The four patterns in this paper are defined by whether the "zipper" (Fermi cut) runs around the horizontal hole, the vertical hole, both, or neither.
  • Pattern 1: No zippers (0,0).
  • Pattern 2: A zipper around the horizontal hole (1,0).
  • Pattern 3: A zipper around the vertical hole (0,1).
  • Pattern 4: Zippers around both holes (1,1).

You cannot change from one pattern to another smoothly. To switch from "No Zippers" to "Horizontal Zipper," you must temporarily create the knots (EPs), drag them around, and let them vanish. This is like having to break the donut to change its shape.

Fragile vs. Strong

The paper also highlights a difference between "Fermi Arcs" and "Fermi Cuts."

• Fermi Arcs are like a piece of string lying on the table. If you blow on it (a tiny perturbation), it blows away. They are fragile.
• Fermi Cuts (the ones in this paper) are like a ring of steel welded around the donut. You cannot remove them with a tiny push. They are topologically protected.

How to See This in Real Life

The authors suggest we can build these "twisted maps" in the real world using:

1. Metasurfaces: Tiny, engineered surfaces (like a grid of nano-antennas) that control light or sound. By adjusting how these antennas lose energy (dissipation), we can create the non-Hermitian conditions.
2. Single-Photon Interferometry: Using single particles of light in a controlled setup.
3. Acoustic Metasurfaces: The paper specifically mentions using a grid of metal cavities (like tiny rooms) with speakers. By adjusting the feedback from the speakers, they can tune the "energy" of the sound waves to create these twisted maps and watch the "zippers" appear and disappear.

Summary

In short, this paper discovers a new type of "twist" in the energy maps of certain materials. Even when the messy knots (EPs) are gone, the map can retain a permanent, unbreakable loop (a Closed Fermi Cut) that wraps around the system. There are four distinct versions of this twist, and they act like a protected code, similar to the ground states of a quantum computer's error-correction system. This gives scientists a new way to classify and potentially use non-Hermitian systems.

Technical Summary

Technical Summary: Spectral Riemann Sheet Topology of Gapped Non-Hermitian Systems

Problem Statement
While extensive research has focused on exceptional points (EPs) and spectral winding numbers in non-Hermitian systems, the topological classification of fully non-degenerate, gapped non-Hermitian spectra remains underexplored. Conventional topological classifications often rely on eigenstates or require the presence of degeneracies (EPs) that close the energy gap. This paper addresses the challenge of defining and classifying topologically distinct configurations in non-Hermitian systems where the complex energy gap remains open, specifically investigating how the topology of the spectral Riemann sheets can be non-trivial even in the absence of EPs.

Methodology
The authors analyze two-dimensional periodic non-Hermitian systems described by Bloch Hamiltonians $H(\mathbf{k})$ over a toroidal Brillouin zone. The methodology involves:

1. Riemann Sheet Analysis: Examining the complex-valued energy spectrum $\varepsilon(\mathbf{k})$ as a multi-sheeted covering of the Brillouin zone.
2. Fermi Cut Definition: Defining "Fermi cuts" as branch cuts connecting EPs, identified with lines of degenerate real (or imaginary) eigenenergies.
3. Threading Procedure: Utilizing a topological operation where a pair of EPs is created, separated, and then threaded across the periodic boundary of the Brillouin zone before being recombined (annihilated). This process is performed without closing the complex energy gap in the final state.
4. Topological Invariants: Constructing invariants based on the permutation of eigenenergies along non-contractible loops ($C_x, C_y$) in the Brillouin zone. The authors distinguish between connected and disconnected Riemann sheet structures.
5. Symmetry Constraints: Applying time-reversal symmetry ($T H^*(\mathbf{k}) T^{-1} = H(-\mathbf{k})$) to restrict the emergence of EPs to time-reversal invariant momenta, thereby simplifying the classification.
6. Analogy Construction: Drawing a structural analogy between the resulting topological configurations and the ground state manifold of Kitaev's toric code.

Key Contributions and Results

• Closed Fermi Cuts in Gapped Systems: The authors demonstrate that while EPs typically close the spectral gap, threading an EP pair across the Brillouin zone boundary and recombining them can result in a fully gapped, non-degenerate system containing a "closed Fermi cut." This cut is a non-contractible branch cut that persists even after the EPs have annihilated.
• $Z_2 \times Z_2$ Topological Classification: For time-reversal-symmetric non-Hermitian two-band systems, the authors identify exactly four topologically distinct configurations of the Riemann sheets. These are characterized by a $Z_2 \times Z_2$ topological invariant $(m_x, m_y)$, where $m_\alpha \in \{0, 1\}$ indicates whether the Riemann sheets are connected (swapped) or disconnected along the $k_\alpha$ direction ($\alpha \in \{x, y\}$).
  • The invariants are formulated algebraically using the discriminant of the Bloch Hamiltonian, $\Delta(\mathbf{k})$, evaluated at time-reversal invariant momenta.
• Distinction from Fermi Arcs: The paper clarifies that non-contractible Fermi arcs (lines of band touching without EPs) are fragile and can be removed by infinitesimal perturbations without closing the gap. In contrast, closed Fermi cuts in gapped systems are topologically protected by the non-trivial braid structure of the eigenenergies around the holes of the toroidal Brillouin zone.
• Toric Code Analogy: The four distinct gapped configurations are shown to be structurally analogous to the four ground states of the toric code on a torus.
  • Non-contractible closed Fermi cuts correspond to non-contractible loops of flipped spins.
  • The presence or absence of these cuts encodes a logical qubit pair, protected by the $Z_2 \times Z_2$ order.
• EPs as Excitations: The authors propose interpreting EPs as excitations within this framework. Similar to electric and magnetic charges in the toric code, EPs appear in pairs connected by open Fermi cuts (flux lines). However, unlike the anyonic excitations of the quantum toric code, these EPs are classical spectral degeneracies. In multi-band systems, this analogy extends to allow for multiple types of excitations (e.g., EP3s) and a larger space of topological configurations.

Significance and Claims
The paper claims to establish a novel topological classification for gapped non-Hermitian systems based on the connectivity of their spectral Riemann sheets. The primary significance lies in:

1. Extending Topology to Gapped Regimes: Providing a framework to define topological order in non-Hermitian systems without requiring the spectral gap to be closed, a condition often necessary for defining conventional topological invariants in these systems.
2. Robustness: Demonstrating that these configurations are stable against perturbations as long as the complex energy gap remains open, distinguishing them from fragile spectral features like Fermi arcs.
3. Experimental Feasibility: The authors outline potential experimental realizations using optical, plasmonic, and mechanical metasurfaces (specifically acoustic metasurfaces with tunable feedback) and single-photon interferometry. They argue that these platforms possess the necessary parametric tunability to create, thread, and annihilate EPs to transition between the distinct topological states.
4. Conceptual Bridge: Offering a concrete link between non-Hermitian spectral topology and the well-understood topological order of the toric code, suggesting that non-Hermitian systems can host topological structures analogous to quantum error-correcting codes, albeit realized as classical spectral objects.

The authors conclude that this framework offers a new approach to utilizing the inherent topological features of non-Hermitian systems, moving beyond the focus on EPs as singularities to viewing them as tools for manipulating the global topology of the energy spectrum.