🔬 mesoscale physics

Twisted bilayer graphene as a terahertz plasmonic crystal

This paper demonstrates that minimally-twisted gapped bilayer graphene with a triangular network of partial dislocations functions as a plasmonic crystal supporting unique features like flat bands and dissipationless modes, which are analyzed through a novel network-based formalism and simulated for terahertz nano-imaging applications.

Original authors: Brian S. Vermilyea, Michael M. Fogler

Published 2026-05-05

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Brian S. Vermilyea, Michael M. Fogler

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 you have two sheets of graphene (a material made of a single layer of carbon atoms, like chicken wire) stacked on top of each other. If you twist them slightly, just a tiny bit, something magical happens. The atoms don't line up perfectly anymore; instead, they create a giant, repeating pattern called a "moiré pattern," similar to the wavy lines you see when you overlap two window screens.

In this specific paper, the authors look at a version of this twisted material where the twist is very small. This creates a landscape of tiny triangular "rooms" (domains) separated by narrow "hallways" (domain walls).

Here is the simple breakdown of what the paper discovers:

1. The "Hallways" are Special

In the middle of the triangular rooms, the material acts like an insulator (it blocks electricity). But in the narrow hallways separating these rooms, electricity flows freely. Even better, these hallways are "topologically protected," meaning the electrons are like cars on a one-way street that cannot easily turn around or crash. They are forced to flow in specific directions depending on their "valley" (a quantum property).

2. The "Plasmonic Crystal"

The authors study how waves of electricity (called plasmons) move through this network of hallways. Think of these plasmons not as individual cars, but as a synchronized wave of traffic.

They found that this twisted graphene acts like a crystal made of light and electricity. Just as a crystal has a rigid structure that affects how sound travels through it, this network of hallways affects how these electrical waves travel.

3. The "Train Station" Analogy

Imagine the hallways meeting at intersections. These intersections are like busy train stations.

The Links: The hallways are the tracks.
The Nodes: The intersections where three hallways meet are the stations.
The Scattering: When a wave of electricity hits a station, it has to decide which track to take next.

The authors created a mathematical model to predict exactly how these waves behave when they hit these stations. They treated the whole system like a giant electrical circuit board.

4. The Surprising Results

When they calculated how these waves move, they found some very cool, unique behaviors:

Flat Bands: Sometimes, the waves get "stuck" in a specific rhythm. They don't speed up or slow down as they move; they just sit there with a constant energy. It's like a train that is stuck on a specific speed limit no matter what.
Gapless Branches: The waves can flow without needing a "push" to get started. They can exist at almost zero energy.
Dissipationless Modes: At certain perfect spots in the pattern (called high-symmetry points), the waves travel without losing any energy. It's like a frictionless slide where the wave never slows down.

5. Two Ways to Look at It

The paper compares two different ways of understanding this system:

The "Perfect World" View (RPA): This assumes the electrons are perfectly coordinated and don't lose energy to chaos. It predicts very sharp, clear waves.
The "Real World" View (Network Model): This assumes the electrons get a bit messy and lose energy as they scatter at the stations. This model predicts that the waves are "damped" (they fade out faster), except for those special frictionless spots mentioned above.

The authors show that while the "Perfect World" view is good for a general idea, the "Real World" view is more accurate for describing how these waves actually behave in a messy, real-life environment.

6. Seeing the Invisible

Finally, the paper simulates what would happen if you tried to "see" these waves using a special microscope (called a near-field imager). They predict that if you shine a light on a tiny speck on the material, the waves would ripple out in a specific pattern, creating interference patterns (like ripples in a pond hitting a rock). This gives scientists a roadmap for how to actually photograph these invisible waves in a lab.

In a nutshell: The paper shows that twisting two sheets of graphene just a tiny bit creates a natural, built-in circuit board for electricity waves. This circuit has unique properties, like frictionless paths and "stuck" energy levels, which could be useful for future technologies that need to handle terahertz frequencies (a type of high-speed signal between radio waves and light).

Technical Summary: Twisted Bilayer Graphene as a Terahertz Plasmonic Crystal

Problem Statement
Minimally-twisted bilayer graphene (mTBG) undergoes lattice reconstruction into a triangular superlattice of AB and BA stacking domains separated by narrow domain walls. When an out-of-plane electric field is applied, the AB and BA regions develop a bandgap, while the domain walls host topologically protected, gapless, one-dimensional (1D) helical electronic states. These states form a triangular network of conducting channels. While the single-particle electronic band structure of this network has been described by scattering models, the collective dynamics—specifically the behavior of surface plasmons in this 1D network—remains an open question. Understanding these plasmons is crucial because low-energy excitations in 1D systems are collective in nature, rendering single-particle dispersion unobservable. The authors aim to characterize the plasmonic band structure of mTBG, determine its validity as a "plasmonic crystal," and identify its unique spectral features in the terahertz (THz) domain.

Methodology
The authors employ two distinct theoretical approaches to model the plasmon dynamics, distinguished by the relationship between the network lattice constant $L$ and the single-particle phase coherence length $L_\phi$ :

Random Phase Approximation (RPA): Applied in the coherent regime ( $L_\phi \gg L$ ). The authors construct a single-particle network model based on forward scattering probabilities at the nodes (junctions) of the domain walls. They calculate the dielectric function and plasmon dispersion by solving for the poles of the inverse dielectric function, incorporating electron-electron interactions via a 2D interaction kernel. This approach captures fine features like particle-hole continua and Landau damping.
Plasmon Network Model (PNM): Applied in the incoherent regime ( $L_\phi \ll L$ ), where single-particle phases are randomized. This approach treats the plasmons as collective modes governed by Tomonaga-Luttinger liquid (TLL) theory. The authors model the scattering of plasmons at a single junction (a node where six leads meet) using an effective circuit analogy. They derive a 6 $\times$ 6 admittance matrix and a scattering matrix $S$ that relates incoming and outgoing plasmon currents. The scattering phase shifts depend on the TLL interaction parameter $K$ . The global plasmon spectrum is then obtained by solving an eigenvalue problem for the network, analogous to the single-particle case but using the plasmon scattering matrix. The authors also develop a non-local extension of the PNM to account for long-range Coulomb interactions between distant links.

Key Contributions and Results

Plasmonic Crystal Behavior: The study demonstrates that mTBG functions as a natural plasmonic crystal. Its 2D plasmon modes are constructed from 1D plasmons propagating along the domain walls and scattering at the AA-stacking nodes.
Unique Band Structure Features: The calculated plasmon dispersion exhibits several notable features:
- Multiple Gapless Branches: The spectrum contains multiple acoustic branches, including a fast diffusive mode and a slower, anisotropic diffusive mode associated with conserved valley polarization.
- Flat Bands: The model predicts the existence of flat (non-dispersive) bands at specific frequencies ( $\omega_\triangle$ and $\omega_\diamond$ ). These correspond to localized "Wannier" states confined to triangular or rhombic loops within the network.
- Dissipationless Modes: At high-symmetry points ( $\Gamma$ , $K$ , and $K'$ ) in the mini-Brillouin zone, the authors identify dissipationless modes. These arise from the conservation of total charge (at $\Gamma$ ) and valley polarization (at $K$ and $K'$ ), which suppresses backscattering in specific angular momentum channels ( $m=0$ and $m=3$ ).
Regime Comparison: The RPA and PNM yield spectra with similar gross structures. However, the RPA predicts sharper modes with numerous narrow particle-hole continua, whereas the PNM predicts more uniform damping. In the asymptotic low-frequency limit, the PNM predicts all modes become overdamped, while the RPA maintains sharp modes.
Interaction Effects: The authors show that Coulomb interactions renormalize the scattering amplitudes at the junctions. In the strong interaction limit ( $K \to 0$ ), the scattering matrix becomes unitary, and the system supports purely dissipationless flat bands and propagating modes.
Near-Field Imaging Simulations: The authors simulate near-field scattering-type scanning near-field optical microscopy (s-SNOM) signals. They show that tip-launched plasmons produce interference patterns with the period of the moiré superlattice, while impurity-launched plasmons exhibit direction-dependent wavelengths and decay behaviors (logarithmic or exponential) depending on whether the frequency lies in a bandgap or resonates with a specific mode.

Significance and Claims
The paper claims that mTBG represents a unique, naturally occurring plasmonic crystal that does not require advanced nanofabrication (unlike artificial plasmonic crystals with patterned gates). The authors assert that for typical experimental parameters (twist angles $\theta \lesssim 1^\circ$ , bandgaps $\sim 250$ meV), the plasmon frequencies lie within the terahertz domain, making the system potentially relevant for nanophotonics applications.

The work provides a theoretical framework for understanding the collective response of topological 1D networks. It highlights that the interplay between the network topology and electron-electron interactions leads to a rich plasmonic spectrum characterized by flat bands and dissipationless modes at high-symmetry points. The authors note that their simplified model, while capturing essential attributes, relies on parameters (such as forward scattering probability and interaction strength) that may vary with experimental conditions like the twist angle and displacement field. They conclude that the network physics becomes relevant at twist angles below approximately $1^\circ$ , provided the Fermi level is tuned into the bandgap and the frequency is low enough to avoid Landau damping.