Contribution of remote bands to orbital magnetization in twisted bilayer graphene

This paper develops a gauge-invariant Hartree-Fock framework to compute orbital magnetization in magic-angle twisted bilayer graphene, demonstrating that remote bands make substantial contributions to both orbital magnetization and self-rotation terms, necessitating careful convergence for accurate evaluation of magnetic responses in correlated moiré systems.

The Gist

Imagine you have a piece of paper with a honeycomb pattern (graphene). Now, take a second piece of paper, stack it on top, and twist it slightly. This creates a giant, repeating pattern called a "moiré pattern," kind of like the shimmering interference you see when two window screens overlap. Scientists call this Twisted Bilayer Graphene (TBG).

When you twist it at a very specific "magic angle," the electrons inside this material get stuck in a slow-motion dance. They stop zooming around and start interacting heavily with each other, creating strange new states of matter, like superconductors (materials with zero electrical resistance) or insulators that act like magnets.

This paper is about understanding one specific property of these electrons: Orbital Magnetization.

The Big Problem: The "Remote Band" Blind Spot

To understand the magnetism of these electrons, scientists usually look at the "active" electrons—the ones currently dancing in the lowest energy levels. It's like trying to understand the mood of a party by only looking at the people on the dance floor.

However, this paper argues that this approach is incomplete. The authors discovered that the remote bands (the electrons in the higher energy levels, far away from the dance floor) play a massive, hidden role in creating the magnetism.

The Analogy:
Imagine a concert.

• The Active Bands are the lead singers on stage.
• The Remote Bands are the backup singers and the massive sound system in the back.

In the past, physicists thought you only needed to listen to the lead singers to understand the song's volume (magnetism). This paper says, "No! The backup singers and the sound system are actually contributing 50% of the volume." If you ignore them, your calculation of the magnetism is wrong.

The New Tool: A "Gauge-Invariant" Calculator

The authors developed a new mathematical framework (a "gauge-invariant framework") to calculate this magnetism correctly.

The Analogy:
Think of calculating the total weight of a suitcase.

• Old Method: You weigh the clothes you can see (the active bands) and guess the weight of the hidden items. This often leads to errors because the "hidden" items (remote bands) are heavy.
• New Method: The authors built a special scale that weighs everything inside the suitcase, including the hidden layers, but does it in a way that doesn't get confused by how the suitcase is rotated or labeled (this is the "gauge-invariant" part).

They found that to get an accurate answer, you have to include about 20 layers of these "remote" electrons in your calculation. If you stop at 5 layers, your result is still wobbling and inaccurate.

What They Found: The "Chern Insulator" Mystery

They applied this new tool to specific states of the twisted graphene, specifically at "fillings" of $\nu = \pm 3$ and $\nu = \pm 1$. (Think of "filling" as how many electrons are packed into the system).

1. At $\nu = \pm 3$: The material becomes a "Chern Insulator." This is a fancy term for a material that acts like a magnet without needing an external magnet, and it conducts electricity in a very specific, one-way loop (the Quantum Anomalous Hall effect).

   • The Discovery: The magnetism here is huge. The "remote bands" contribute significantly to this magnetic strength. Without them, the material wouldn't be as magnetic as experiments show.

2. At $\nu = \pm 1$: There was a debate about what state the material was in. Was it a "Chern Insulator" (strong magnet) or an "Intervalley Coherent" state (a different kind of order)?

   • The Discovery: By calculating the magnetism correctly (including the remote bands), they found that the Chern Insulator state has a much stronger magnetic response. This explains why, when you apply a magnetic field in experiments, the material switches to this state. The "remote bands" help tip the scales in favor of the magnetic state.

Why Does This Matter?

• Fixing the Math: It corrects a major oversight in how physicists calculate magnetism in these complex materials. You can't just look at the "top" of the energy stack; you have to look deep into the "remote" layers.
• Designing New Tech: Understanding exactly how these materials become magnetic helps engineers design better sensors, faster computers, and new types of quantum devices.
• The "Self-Rotation" Clue: They also calculated something called "self-rotation" ($m_{SR}$). Imagine the electrons spinning on their own axes while orbiting. This specific type of rotation is what scientists can actually see in experiments using light (magnetic circular dichroism). The paper shows that this visible signal is largely driven by those hidden remote bands.

In a Nutshell

This paper is like realizing that to understand why a car is so fast, you can't just look at the engine (the active bands); you have to account for the aerodynamics, the tires, and the weight of the passengers (the remote bands). By building a better way to measure all these factors together, the authors finally explained why twisted bilayer graphene acts like such a strong magnet, solving a puzzle that had been confusing scientists for a while.

Technical Summary

Here is a detailed technical summary of the paper "Contribution of remote bands to orbital magnetization in twisted bilayer graphene."

1. Problem Statement

Twisted bilayer graphene (TBG) at the "magic angle" exhibits a rich phase diagram of correlated insulating states and unconventional superconductivity. Recent experiments have observed Quantum Anomalous Hall (QAH) effects and strong magnetic responses in these states, which are theoretically linked to orbital magnetization ($M_{orb}$).

While the non-interacting Bistritzer–MacDonald (BM) model successfully describes the topology (Chern numbers) and Berry curvature of the flat bands, it fails to capture the full physics of the interacting system. In TBG, electron-electron interactions are comparable to or larger than the bandwidth, leading to self-consistent Hartree-Fock (HF) band reconstructions.

• The Gap: A systematic, gauge-invariant method to calculate bulk orbital magnetization in these interaction-driven correlated phases was lacking.
• The Specific Challenge: It was unclear how interaction-induced band reconstruction, valley polarization, and contributions from remote bands (bands far from the Fermi level) quantitatively modify orbital magnetization compared to non-interacting predictions. Standard calculations often truncate the Hilbert space, potentially missing critical contributions from these remote bands.

2. Methodology

The authors developed a gauge-invariant framework to compute orbital magnetization within the self-consistent Hartree-Fock (HF) approximation for magic-angle TBG.

• Model: They started with the continuum BM model projected onto the relevant bands, incorporating Coulomb interactions. They performed self-consistent HF calculations in momentum space to determine symmetry-broken ground states at integer fillings ($\nu$).
• Theoretical Framework: Instead of relying on Bloch wavefunction derivatives (which suffer from gauge ambiguities), they utilized the projector formulation of orbital magnetization theory (based on recent developments by Kang et al.).
  • They expressed the total orbital magnetization ($M_{orb}$) and the self-rotation contribution ($m_{SR}$) directly in terms of projection operators ($P$ and $Q$) acting on the HF Hamiltonian ($H_{HF}$).
  • Formulas:
    $$M_{orb} = -\frac{e}{\hbar} \sum_k \text{Im} \left[ W_{xy}(k) - N_{xy}(k) \right]$$
    $$m_{SR} = \frac{e}{\hbar} \sum_k \text{Im} \left[ W_{xy}(k) + N_{xy}(k) \right]$$
    Where $W$ and $N$ involve traces of products of projection operators, their derivatives, and the Hamiltonian.
• Handling Remote Bands: A key innovation was the explicit treatment of remote bands. The authors defined truncated projection operators ($P_{ncut}$ and $Q_{ncut}$) that include the active flat bands plus a specific number ($n_{cut}$) of occupied and empty remote bands. They demonstrated that physical quantities converge only when a sufficient number of remote bands are included.

3. Key Contributions

1. Gauge-Invariant HF Framework: Established a robust, gauge-invariant method to calculate $M_{orb}$ and $m_{SR}$ directly from the self-consistent HF Hamiltonian, avoiding the ambiguities associated with Bloch wavefunction derivatives in topological systems.
2. Discovery of Remote Band Sensitivity: Demonstrated that, unlike topological invariants (e.g., Chern numbers) which are determined solely by the occupied bands, orbital magnetization is highly sensitive to remote bands.
   • Even though remote bands may remain time-reversal symmetric, their momentum derivatives overlap with active bands once symmetry is broken in the active sector, leading to substantial contributions to $M_{orb}$ and $m_{SR}$.
   • Numerical convergence requires including at least 20 pairs of remote bands ($n_{cut} \approx 20$).
3. Systematic Calculation for Correlated Phases: Applied this method to integer fillings ($\nu = \pm 1, \pm 3$) to evaluate the magnetic response of competing correlated phases.

4. Key Results

A. Fillings $\nu = \pm 3$ (Chern Insulators)

• Ground State: The system forms a spin-valley polarized Chern insulator with Chern number $C = \pm 1$ (breaking $C_2T$ symmetry).
• Convergence: Both $M_{orb}$ and $m_{SR}$ converge rapidly once $n_{cut} \approx 20$.
• Symmetry: The results respect particle-hole symmetry: $M_{orb}(\nu=3) = M_{orb}(\nu=-3)$ and $m_{SR}(\nu=3) = -m_{SR}(\nu=-3)$.
• Gap Behavior: Inside the insulating gap:
  • $M_{orb}$ shows a linear dependence on the chemical potential ($\mu$), with a slope proportional to the Chern number ($\partial M_{orb}/\partial \mu = eS/h \cdot C$). It changes sign across the gap.
  • $m_{SR}$ (self-rotation) remains constant within the gap, with a large magnitude ($\approx 40 \mu_B$), making it accessible via magneto-optical probes like Magnetic Circular Dichroism (MCD).

B. Fillings $\nu = \pm 1$ (Competing Phases)

• Competing States: Two nearly degenerate states were identified:
  1. Correlated Chern Insulator (CCI): Three valence bands break $C_2T$ symmetry, yielding a total Chern number $C=3$.
  2. Intervalley-Coherent (IVC) State: A mixed phase with $C=1$.
• Magnetization Comparison:
  • The CCI state possesses a significantly larger orbital magnetization (both magnitude and slope) compared to the IVC state, consistent with its higher Chern number.
  • The large $M_{orb}$ of the CCI explains why external magnetic fields stabilize this phase over the IVC phase (energetically favored in a field).
  • The self-rotation contribution $m_{SR}$ is also substantially larger for the CCI state.

5. Significance

• Microscopic Origin of Magnetic Response: The work clarifies that the strong magnetic response observed in TBG experiments is not just a property of the flat bands but is critically enhanced by interactions and remote band contributions.
• Experimental Guidance: The prediction of a large, constant $m_{SR}$ in the gap suggests that magneto-optical measurements (MCD) are a viable and powerful tool to distinguish between competing correlated phases (e.g., CCI vs. IVC) in TBG.
• Methodological Advance: The projector-based, gauge-invariant approach provides a systematic and controlled route to evaluate orbital magnetization in any interaction-driven moiré system, correcting the common misconception that remote bands can be ignored in magnetic property calculations.
• Thermodynamic Consistency: The results confirm the thermodynamic relation between orbital magnetization and the Chern number (via the Streda formula), validating the theoretical framework against fundamental principles.

In summary, this paper establishes that accurate modeling of orbital magnetization in correlated moiré materials requires a rigorous inclusion of remote bands and a gauge-invariant formalism, providing a new lens through which to understand the magnetic properties of twisted bilayer graphene.