Fermi velocity, interlayer couplings, and magic angle renormalization in twisted bilayer graphene

Through extensive Hartree-Fock calculations, this paper demonstrates that many-body effects in twisted bilayer graphene significantly renormalize the Fermi velocity and interlayer couplings, shifting the magic angle from $0.99^\circ$ to $0.88^\circ$ and challenging the paradigm that maximum superconductivity occurs at the minimum bandwidth.

The Gist

Imagine twisted bilayer graphene as a delicate, two-layered dance floor made of carbon atoms. When you twist the top layer slightly relative to the bottom one, the atoms create a giant, repeating pattern called a "moiré" pattern. At a very specific twist angle, known as the "magic angle," the electrons on this dance floor slow down so much that they get stuck in a "flat" state, moving very little. This flatness is the secret sauce that allows these materials to become superconductors (conducting electricity with zero resistance) or insulators.

For years, scientists have been trying to find the exact "magic angle" to build these materials. They calculated it to be around 0.99 degrees. However, this paper argues that those calculations missed a crucial ingredient: the way electrons talk to each other.

Here is the story of what the authors found, explained simply:

1. The "Crowded Dance Floor" Effect

In the old calculations, scientists treated the electrons like solo dancers who didn't really notice each other. But in reality, electrons are like a crowded dance floor; they bump into each other, push, and pull. This paper uses a sophisticated method (called Hartree-Fock) to simulate this "crowded" environment.

They found that when you account for these electron-to-electron interactions, the "flat" bands where the electrons get stuck actually widen out. It's as if the dance floor suddenly became a little less cramped, giving the electrons a bit more room to move.

2. The Moving Target (The Shifted Magic Angle)

Because the "flatness" of the bands changed, the perfect angle required to get them flat also changed.

• The Old Prediction: The magic angle was thought to be 0.99 degrees.
• The New Prediction: When you include the electron interactions, the magic angle shifts to 0.88 degrees.

Think of it like tuning a guitar. You were aiming for a specific note (0.99°), but once you realized the strings were vibrating against each other (interactions), you had to tighten the tuning peg to a slightly different spot (0.88°) to get the perfect sound.

3. The "Speed Limit" of Electrons

The paper also looked at the Fermi velocity, which is essentially the speed limit of electrons in graphene.

• In normal graphene, electrons zip along at a constant speed.
• In this twisted system, the authors found that the interactions actually speed up the electrons in the flat bands at certain angles, contrary to what you might expect if you just thought about them getting "stuck."

They developed a mathematical "recipe" (analytical formulas) that predicts exactly how much the speed and the connections between the two layers change. They tested this recipe against their massive computer simulations (involving up to 18,000 atoms per cell) and found the recipe worked perfectly.

4. Tuning the System with "Gates"

The authors showed that you can change these results by changing the environment around the graphene.

• If you suspend the graphene in a vacuum (like a floating trampoline), the interactions are strong, and the magic angle shifts a lot.
• If you wrap the graphene in a protective material (like hBN) or put metal gates nearby, the interactions are "screened" or dampened, and the shift is smaller.

This means scientists can actually tune the properties of the material just by changing how they set up their experiment (like changing the distance of metal gates or the surrounding material), rather than having to physically twist the graphene to a new angle.

5. Why This Matters for Superconductivity

The paper suggests a change in how we think about superconductivity in these materials.

• Old Idea: Superconductivity happens exactly at the "magic angle" where the bands are flattest (slowest).
• New Idea: The authors suggest that the best superconductivity might actually happen at a slightly larger angle (around 1.1°), where the bands are not perfectly flat but still have a little bit of "wiggle room" (dispersion).

They propose that at the perfectly flat angle (the new 0.88°), the electrons might be too "jittery" due to quantum fluctuations to form a stable superconducting state. It's like trying to balance a pencil on its tip; if it's too perfectly balanced, it might actually be harder to keep stable than if it's slightly tilted.

Summary

In short, this paper says: "We forgot to count how much the electrons push and pull on each other. Once we do, the 'magic angle' isn't where we thought it was. It's actually a bit smaller, and we can predict exactly how to tune it using the environment around the material."

This helps experimentalists understand why they might be seeing superconductivity at 1.1° instead of the theoretically predicted 0.99°, and gives them a new toolkit to design better quantum materials.

Technical Summary

1. Problem Statement

Twisted Bilayer Graphene (TBG) exhibits a "magic angle" ($\theta_M \approx 1.1^\circ$) where the electronic bands flatten, leading to strongly correlated phenomena such as superconductivity and Mott insulation. However, a significant discrepancy exists between experimental observations (where correlated phases appear near $1.1^\circ$) and ab initio or non-interacting theoretical calculations (which often predict the magic angle closer to $1.0^\circ$).

The central problem addressed is the magnitude of many-body effects (specifically electron-electron interactions) on the electronic structure of TBG. Previous studies suggested these effects might be comparable to lattice relaxation or strain, but a comprehensive, self-consistent treatment of how interactions renormalize the Fermi velocity, interlayer couplings, and the magic angle itself was lacking.

2. Methodology

The authors employ a rigorous self-consistent Hartree-Fock (HF) approach within a tight-binding (TB) model to investigate the charge neutrality point of TBG.

• Model System: They use the FORGE package to perform calculations on commensurate TBG structures with up to $\sim 18,000$ atoms per moiré unit cell.
• Interactions: The electron-electron interaction is modeled via a double-gated Coulomb potential $V(r)$, simulating a sample suspended between metallic gates (distance $\xi = 10$ nm). They include an on-site Hubbard term ($U=4$ eV) and treat the dielectric environment using a static Random Phase Approximation (RPA) value ($\epsilon \approx 10$).
• Symmetry Constraints: A crucial aspect of their methodology is the explicit enforcement of all relevant symmetries during the self-consistent loop: lattice translations, crystallographic symmetries ($C_{6v}$), time-reversal, spin rotation, and emergent valley charge conservation. This prevents convergence to symmetry-broken states that are not the ground state.
• Analytical Derivation: To interpret the numerical results, they derive analytical expressions for the renormalization of key parameters to first order in the Coulomb interaction ($\alpha$) and interlayer couplings. They project the mean-field Hamiltonian onto a plane-wave basis to extract renormalized parameters.

3. Key Contributions

A. Renormalization of the Magic Angle

The study demonstrates that many-body effects cause a significant widening of the flat bands. Consequently, the condition for minimum bandwidth (the definition of the magic angle) shifts.

• Result: The magic angle is renormalized from the ab initio value of $0.99^\circ$ to a lower value of $0.88^\circ$ for a suspended sample with specific gate geometry.
• Implication: This shift explains why experiments observe correlated physics at $\sim 1.1^\circ$ (which is now effectively "further" from the renormalized magic angle) while theory predicts the flat band condition at lower angles.

B. Analytical Expressions for Renormalized Parameters

The authors derive closed-form analytical formulas for the renormalization of:

1. Fermi Velocity ($v_F$): They confirm the logarithmic enhancement of $v_F$ due to Coulomb interactions, modified by the presence of metallic gates (screening).
2. Interlayer Couplings ($w_0, w_1$): They derive expressions for the corrections $\delta w_0$ and $\delta w_1$ to first order.
   • $\delta w_0$ and $\delta w_1$ depend on the twist angle (via momentum transfer $q_1$), the interlayer distance $d$, and the gate distance $\xi$.
   • The analytical predictions show remarkable agreement with the numerical HF results.

C. Generalized Bistritzer-MacDonald (BM) Model

The paper proposes a generalized BM model where the band structure can be accurately reproduced by renormalizing only a few parameters:

• The Fermi velocity $v_F$.
• The interlayer potential $T(\mathbf{r})$ (via $w_0, w_1$).
• The intralayer pseudo-gauge field $A_\ell(\mathbf{r})$.
• Finding: Including the momentum dependence of these renormalized couplings (non-local effects) allows for an almost exact reproduction of the full numerical band structure with discrepancies $< 1$ meV.

4. Key Results

• Bandwidth Widening: Many-body interactions increase the bandwidth of the flat bands. This is a competing effect to lattice relaxation, which typically narrows bands.
• Gate Tunability: The renormalization is highly sensitive to the dielectric environment.
  • Suspended samples ($\epsilon \approx 10, \xi = 60$ nm) show a larger shift in the magic angle compared to hBN-encapsulated samples ($\epsilon \approx 15, \xi = 10$ nm).
  • This suggests that experimentalists can tune the flat band properties by changing the gate geometry or dielectric environment.
• Fermi Velocity Enhancement: At intermediate twist angles (away from the magic angle), the flat-band Fermi velocity is significantly enhanced relative to the bare value. The authors propose measuring this velocity as a direct probe of the effective couplings.
• Paradigm Shift in Superconductivity:
  • The results suggest that the maximum critical temperature ($T_c$) for superconductivity may not correspond to the condition of minimum bandwidth (the strict magic angle).
  • Instead, superconductivity might be optimal at a slightly larger twist angle ($\sim 1.1^\circ$) where the bandwidth is small but not minimal, avoiding the suppression caused by phase fluctuations (Berezinskii-Kosterlitz-Thouless scenario) that dominate at the renormalized magic angle ($\lesssim 1.0^\circ$).

5. Significance

• Resolving the Angle Discrepancy: The work provides a theoretical mechanism to resolve the long-standing tension between ab initio predictions ($\theta \approx 1.0^\circ$) and experimental observations ($\theta \approx 1.1^\circ$) by showing that interactions shift the "true" magic angle to lower values.
• Predictive Power: The derived analytical formulas allow for the prediction of TBG properties in various experimental setups without needing computationally expensive self-consistent calculations for every configuration.
• Experimental Guidance: The paper offers concrete experimental proposals:
  • Tuning the dielectric environment to shift the magic angle.
  • Measuring the Fermi velocity at intermediate angles to extract effective coupling constants.
• Theoretical Framework: It establishes that a low-energy effective theory for interacting TBG must start from a renormalized single-particle state (incorporating non-local exchange) rather than a non-interacting one, fundamentally changing how the low-energy physics of moiré materials should be modeled.

In summary, this paper demonstrates that many-body effects are not merely perturbative corrections but are central to defining the electronic landscape of TBG, shifting the magic angle, and potentially dictating the optimal conditions for superconductivity.