🔬 mesoscale physics

Dirac fermions in non-Hermitian magnetic fields: Zero modes and index theorem

This paper demonstrates that spatially modulated non-Hermitian parameters in Lorentz symmetric Dirac theories act as non-Hermitian gauge fields generating magnetic fluxes that induce localized zero-energy modes, a phenomenon numerically verified on graphene and proposed as a foundation for exploring non-Hermitian magnetic catalysis and topological phases.

Original authors: Christopher A. Leong, Bitan Roy

Published 2026-06-26

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Christopher A. Leong, Bitan Roy

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where the usual rules of physics get a little "fuzzy." In our normal world, if you push a ball, it moves predictably. But in the strange realm of non-Hermitian physics (the focus of this paper), things don't always behave symmetrically; energy can seem to appear or disappear, and the math describing them is a bit more complex.

This paper explores what happens when we take massless Dirac fermions—which are like tiny, ultra-fast particles that act like light but have mass (think of them as the "ghosts" of electrons found in materials like graphene)—and put them into a special kind of magnetic field.

Here is the breakdown of their discovery using simple analogies:

1. The Setup: The "Ghost" Magnetic Field

Usually, when you put a particle in a magnetic field, it gets trapped in specific energy levels (like rungs on a ladder). The authors created a theoretical "magnetic field" that isn't made of normal magnetism but is instead a non-Hermitian (NH) field.

Think of this field not as a magnet, but as a wind that blows differently depending on where you are. This wind is "mass-like," meaning it interacts with the particles in a way that mimics giving them weight, even though they are supposed to be weightless.

2. The Discovery: The "Zero-Energy" Traps

The researchers found that when these fast-moving particles enter this special wind, something magical happens: they get stuck in zero-energy states.

The Analogy: Imagine a flat, frictionless floor (the particles' normal state). Now, imagine a gentle, invisible slope that suddenly appears. The particles roll down and get stuck in a perfectly flat valley at the very bottom. They stop moving (zero energy) but don't disappear.
The Twist: In this new world, these stuck particles come in two flavors: Left-handed and Right-handed.
- If the "wind" blows one way, only the Left-handed particles get stuck in the middle of the room (the bulk).
- If the wind blows the opposite way, only the Right-handed particles get stuck in the middle.
- The other flavor (the one not stuck in the middle) gets pushed all the way to the walls (the edges).

3. The Rulebook: Counting the Trapped Particles

The paper proves a very specific rule about how many of these particles get stuck. It turns out the number of trapped particles depends entirely on how much "wind" (magnetic flux) is trapped inside the system.

The Analogy: Think of the magnetic field as a bucket of water. The size of the bucket (the total amount of water/flux) determines exactly how many fish (particles) can fit inside the "zero-energy" zone. It doesn't matter if the water is splashed unevenly or poured smoothly; if the total amount is the same, the number of trapped fish is exactly the same. This is a modern update to a famous physics rule called the Aharonov-Casher Index Theorem.

4. Testing it on a Grid (Graphene)

To make sure this wasn't just a math trick, the authors simulated this on a honeycomb grid (like the structure of graphene, a super-thin sheet of carbon).

Type I Field: They created a field where the "wind" pushes particles to the center. The simulation showed the particles gathering right in the middle, just like the math predicted.
Type II Field: They created a slightly different field. Here, the particles gathered in a ring near the edge, but not on the very edge.
The Result: The computer simulations matched their math perfectly. They counted the trapped particles and found the number grew in a straight line as they increased the "amount of wind" in the system.

5. What This Means for the Future

The authors suggest that while this is currently a theoretical discovery, it opens the door to building new types of materials.

The "Catalyst": They call this phenomenon "Non-Hermitian Magnetic Catalysis." Imagine a catalyst that forces particles to stick together and form new, stable structures just because they are trapped in this zero-energy zone.
Where to look: They mention that we don't need to wait for new quantum computers to see this. We can build "artificial atoms" using light (photonic lattices) or electrical circuits that mimic these honeycomb grids. Since we can already control how light or electricity hops between these artificial atoms, we could potentially create these "ghost magnetic fields" in a lab tomorrow.

In short: The paper shows that by creating a special kind of "fuzzy" magnetic field, we can trap fast-moving particles in a zero-energy state. The number of trapped particles is strictly determined by the strength of the field, and we can see this happening in simulations of graphene-like grids. This could lead to new ways of controlling matter in future electronic or optical devices.

Technical Summary: Dirac Fermions in Non-Hermitian Magnetic Fields

Problem Statement
The paper addresses the behavior of massless Dirac fermions when subjected to non-Hermitian (NH) magnetic fields (NHMFs). In standard Hermitian Lorentz-invariant Dirac theories, strong external magnetic fields induce a topologically protected degenerate manifold of zero-energy states (the zeroth Landau level), a phenomenon guaranteed by the Aharonov-Casher index theorem (ACIT). The authors investigate whether this topological protection and the resulting zero-energy manifold persist when the system is extended to a Lorentz-invariant non-Hermitian framework. Specifically, they explore the coupling of Dirac fermions to "masslike" anti-Hermitian operators that act as NH gauge fields, asking if such systems support localized zero-energy eigenmodes and how the chirality (left vs. right) of these modes is determined.

Methodology
The authors employ a dual approach combining continuum field theory and microscopic lattice modeling:

Continuum Theory Construction:
- They define a Lorentz-invariant NH Dirac operator, $H^{\text{NH}}_{\text{Dir}}(k) = H_{\text{Dir}}(k) + \alpha M H_{\text{Dir}}(k)$ , where $H_{\text{Dir}}$ is the canonical kinetic Hamiltonian, $M$ is a Hermitian mass matrix, and $\alpha$ is a real parameter controlling non-Hermiticity.
- To introduce NHMFs, they minimally couple the fermions to a NH vector potential $A(r)$ within the anti-Hermitian term, effectively replacing momentum $k$ with $k + \alpha M A(r)$ .
- In two dimensions ( $d=2$ ), they utilize a similarity transformation to relate the NH Hamiltonian with a gauge potential to the original Hermitian Hamiltonian without one. This allows them to analytically derive the form of the left ( $\Psi_L$ ) and right ( $\Psi_R$ ) zero-energy eigenmodes.
- They analyze the normalization conditions of these modes, linking the number of normalizable zero modes to the total NH magnetic flux $\Phi$ enclosed by the system.
Lattice Model Realization:
- To validate the continuum predictions, the authors construct a microscopic model on graphene's honeycomb lattice.
- They identify specific mass matrices ( $M$ ) that preserve translational symmetry, leading to two distinct realizations of NHMFs: Type-I (associated with $\gamma_{12}$ , representing Haldane-like currents) and Type-II (associated with $\gamma_0$ , representing Semenoff-like staggered potentials).
- They numerically diagonalize the lattice-regularized NH operators for both types under periodic and open boundary conditions.
- The study also extends to a hyperbolic lattice ({10, 3} Schl"afli symbol) to test the robustness of the findings in negatively curved spaces.

Key Contributions and Results

Existence of Chiral Zero Modes: The study demonstrates that gapless planar Dirac fermions coupled to masslike NHMFs support a manifold of spatially localized zero-energy eigenmodes. Unlike the Hermitian case where modes are typically degenerate and bidirectional, the NHMFs induce a chiral separation: the system supports either normalizable right eigenmodes or normalizable left eigenmodes, but not both simultaneously for a given flux direction.
Non-Hermitian Index Theorem: The authors establish that the number of self-normalizable zero-energy modes ( $N$ ) is exactly equal to the number of NH magnetic flux quanta ( $N = \Phi / 2\pi$ ) enclosed by the system. This result holds irrespective of the spatial profile of the NHMF, mirroring the ACIT in Hermitian systems.
Type-I vs. Type-II Realizations:
- Type-I NHMFs: Support normalizable left zero-energy modes in the bulk while the corresponding right modes are non-normalizable and reside at the boundaries (or vice versa depending on flux direction).
- Type-II NHMFs: Due to the valley degree of freedom in graphene, Type-I fields support normalizable modes of one chirality near both valleys. In contrast, Type-II fields support normalizable left modes near one valley and normalizable right modes near the complementary valley. Numerical results show that for Type-II, the eigenmodes appear as a superposition of normalizable and non-normalizable components, peaking near the boundary but not strictly on the edges.
Lattice Verification: Numerical diagonalization on the honeycomb lattice confirms the continuum predictions. The density of states for near-zero-energy modes shows sharp localization consistent with the analytical forms derived. Furthermore, the number of near-zero-energy modes scales linearly with the number of plaquettes (flux), confirming the index theorem on the lattice.
Hyperbolic Extension: The authors extend the analysis to a hyperbolic lattice, finding that the spatial variation of near-zero-energy modes under Type-I and Type-II NHMFs remains qualitatively similar to the Euclidean case.

Significance and Claims
The paper claims that the introduction of masslike NHMFs opens a new theoretical framework for exploring non-Hermitian topological phenomena. The primary significance lies in the demonstration that the topological protection of zero-energy states (the ACIT) survives the transition to non-Hermitian systems, albeit with a modified chiral structure where modes are strictly left- or right-handed.

The authors highlight several potential future directions grounded in their findings:

NH Magnetic Catalysis: The existence of a degenerate near-zero-energy manifold suggests the system is susceptible to various orderings at weak coupling.
Symmetry Preserving Topological Phases: The robustness of the Type-I zero-energy manifold against spontaneous symmetry breaking (due to the definition of vacuum expectation values in NH systems) could facilitate the realization of strong interaction-driven topological phases.
NH Chiral Anomaly: For 3D Dirac or Weyl fermions coupled to z-directional NHMFs, the dispersion of zero modes in the field direction raises the possibility of an NHMF-induced chiral anomaly when an electric field is applied.
Experimental Platforms: The authors propose that these effects can be tested in highly tunable classical metamaterials, specifically topolectric circuits and photonic lattices, where complex and non-reciprocal hopping (Peierls phases) can be engineered to mimic the required NHMFs. Optical lattices of neutral atoms are also suggested as a platform for testing these predictions in noninteracting and interacting regimes.

The work is presented as a foundational step toward understanding relativistic landscapes immersed in masslike NH gauge fields, bridging concepts from condensed matter and high-energy physics.