Strongly Correlated Superconductivity in Twisted Bilayer Graphene: A Gutzwiller Study

This study employs a variational Gutzwiller wavefunction approach on an 8-band model to reveal that magic-angle twisted bilayer graphene exhibits a dome-shaped Fermi liquid phase separating weakly and strongly correlated superconducting regimes, with the latter characterized by a nematic, nodal-gap state stabilized by interaction-driven gap reconstruction and distinct from conventional Mott insulators.

The Gist

The Big Picture: The "Magic" Dance Floor

Imagine two sheets of graphene (a material made of carbon atoms, like a chicken wire fence) stacked on top of each other. If you twist them at a very specific, "magic" angle, something incredible happens: the electrons on this dance floor stop behaving like individual people and start acting like a single, synchronized crowd.

This is Magic-Angle Twisted Bilayer Graphene (MATBG). Scientists have found that when you add just the right amount of electricity (doping) to this twisted stack, it becomes a superconductor. This means electricity can flow through it with zero resistance, like a ghost gliding through a wall.

But here's the mystery: Usually, superconductors are explained by electrons pairing up gently (like a slow dance). But in this magic graphene, the electrons are "strongly correlated," meaning they are constantly bumping into each other, arguing, and pushing away. It's more like a mosh pit than a ballroom. The big question was: How do these chaotic electrons manage to pair up and dance perfectly without tripping over each other?

The Tool: The "Gutzwiller Filter"

To solve this, the authors used a mathematical tool called the Gutzwiller Approximation.

Think of the electrons in this material as a chaotic crowd of people in a room. Some people are "heavy" and stuck in place (the f-orbitals), while others are "light" and run around freely (the c-orbitals).

• The Problem: If you try to calculate how everyone moves, the math gets too heavy for any computer because the "heavy" people interact so violently with each other.
• The Solution: The authors built a special filter (the Gutzwiller projector). This filter looks at every possible way the heavy people could arrange themselves and says, "No, that arrangement is too messy and costs too much energy. Let's only keep the arrangements that make sense."

Crucially, they tweaked this filter to allow for superconductivity. Usually, this filter assumes electrons are just sitting there. But they modified it to allow the electrons to form pairs (like holding hands) even while they are being squeezed by the crowd.

The Discovery: Three Stages of the Dance

By running their calculations, the authors mapped out a "phase diagram," which is like a weather map for the electrons. They found three distinct "seasons" or states, depending on how strong the electron repulsion (the "pushing away" force) is:

1. The Weak Push (Small Repulsion): The "Standard Ballroom"

When the electrons don't push each other too hard, they behave like a standard superconductor (BCS theory).

• Analogy: Imagine a calm ballroom where couples are dancing a slow waltz. Everyone is paired up, and the music is smooth. This is the BCS-like state.

2. The Strong Push (Large Repulsion): The "Mosh Pit Superconductor"

When the electrons push each other very hard (strong correlation), something weird happens. They don't just dance; they form a Strongly Correlated Superconductor (SC-SC).

• Analogy: Imagine a massive, chaotic mosh pit. Everyone is pushing and shoving. But somehow, in the middle of the chaos, everyone suddenly locks arms and moves in a perfect circle.
• The Magic: The authors found that the "heavy" electrons (the ones in the mosh pit) stop moving around individually (their charge fluctuations are suppressed) but still manage to hold hands and pair up. They become a "heavy" superconductor. It's like a heavy metal band that suddenly starts playing a perfect symphony.

3. The "Small" Liquid: The "Hidden Room"

The most surprising discovery is a new state called the small Fermi Liquid (sFL).

• Analogy: Imagine a large concert hall (the normal state) where everyone is standing around. Suddenly, a secret, smaller room opens up inside the hall. The "heavy" electrons hide inside this small room, forming a tight, quiet group (a local singlet), while the "light" electrons run around the outside.
• Why it matters: The authors found that this "small room" state is actually the parent of the superconducting state when the repulsion is very strong. It's like the secret room is the training ground where the electrons learn to dance before they explode onto the main floor as superconductors.

The "Nematic" Twist: Breaking the Rules

The paper also found that the superconducting state isn't perfectly round; it's nematic.

• Analogy: Imagine a round table where everyone is sitting. In a normal superconductor, the table is perfectly round. In this "nematic" state, the table gets squashed into an oval. The electrons prefer to dance in one direction over another. This happens because the "heavy" electrons are fighting with the "light" ones, creating a tension that stretches the dance floor.

The "Gap" Mystery: The V-Shape

When they looked at the energy gap (the energy needed to break the electron pairs), they saw a V-shape.

• Analogy: In a normal superconductor, the gap is like a smooth hill you have to climb. In this material, the gap looks like a sharp "V" (like a valley). This suggests that the electrons are pairing up in a very complex, unconventional way, likely driven by the vibrations of the carbon atoms (phonons) acting like a glue.

The Takeaway: Why This Matters

This paper is a breakthrough because it provides a blueprint for how to calculate these chaotic systems.

1. It solves the puzzle: It shows that even when electrons are pushing each other apart violently, they can still form a superconductor by suppressing their individual movement and focusing on pairing.
2. It finds the parent: It identifies the "small Fermi liquid" as the likely starting point for this superconductivity, giving scientists a new target to look for in experiments.
3. It's a universal tool: The mathematical method they used (the Gutzwiller framework) can now be applied to other weird, strongly correlated materials, not just graphene.

In short: The authors built a mathematical filter that allowed them to see how a chaotic crowd of electrons, when squeezed tight enough, can spontaneously organize into a perfect, friction-free dance. They discovered that before the dance begins, the electrons hide in a "small room" to prepare, and the dance floor itself gets stretched into an oval shape. This explains the mysterious superconductivity of magic-angle graphene and opens the door to understanding other exotic materials.

Technical Summary

1. Problem Statement

Magic-angle twisted bilayer graphene (MATBG) exhibits superconductivity (SC) with phenomenology strikingly similar to high-$T_c$ cuprates, including a dome-shaped phase diagram, linear-in-$T$ resistivity, and unconventional pairing signatures (e.g., nodal gaps, nematicity). However, a quantitative, self-consistent theoretical understanding of the SC state within a realistic lattice model remains elusive. Key challenges include:

• Strong Correlations: The interplay between strong Coulomb repulsion ($U$) and pairing mechanisms is not fully understood. Recent experiments suggest that increasing interactions (reducing screening) actually suppresses $T_c$, contradicting early screening-enhanced theories.
• Unconventional Mechanisms: Evidence points to mechanisms beyond standard electron-phonon coupling, including anti-Hund's coupling mediated by moiré phonons and resonating valence bond (RVB) physics.
• Theoretical Gap: Existing methods often fail to capture the simultaneous strong correlations and superconducting order in a multi-orbital, topologically non-trivial system like MATBG.

2. Methodology

The authors employ a variational Gutzwiller wavefunction approach applied to an 8-band tight-binding model specifically designed for MATBG.

• Model Hamiltonian:
  • Orbitals: The model consists of two correlated $f$-orbitals (localized, heavy fermions) and six uncorrelated $c$-orbitals (itinerant) per spin-valley, capturing the "topological heavy fermion" nature of MATBG.
  • Interactions: The Hamiltonian includes:
    • Onsite Coulomb repulsion ($\hat{H}_U$).
    • Anti-Hund's coupling ($\hat{H}_{J_A}$): Mediated by inter-valley phonons, favoring inter-valley spin-singlet pairing.
    • Hund's coupling ($\hat{H}_{J_H}$): Intra-orbital coupling favoring $d$-wave pairing.
    • Hartree terms ($\hat{H}_W, \hat{H}_V$) to adjust orbital energy levels and ensure realistic filling.
• Variational Ansatz:
  • The ground state is constructed as $|\Psi_G\rangle = \prod_R \hat{P}_R |\Phi_0\rangle$, where $|\Phi_0\rangle$ is a Bardeen-Cooper-Schrieffer (BCS) wavefunction and $\hat{P}_R$ is a Gutzwiller projector acting on $f$-orbitals.
  • Key Innovation: The projector $\hat{P}_R$ is allowed to break charge $U(1)$ symmetry. This enables the description of superconducting states within the Gutzwiller framework, where the off-diagonal (pairing) components of the projector are variational parameters.
  • Symmetry Handling: The variational parameters are classified by discrete symmetries ($T, C_{2z}, C_{2x}$) and continuous symmetries (spin-$SU(2)$, valley-$U(1)$). The authors explicitly allow $C_{3z}$ breaking to study nematicity.
• Numerical Implementation:
  • The ground state energy is minimized using the Gutzwiller Approximation (GA), which becomes exact in the limit of infinite coordination number.
  • To handle the large parameter space (513 independent real parameters), the authors treat the uncorrelated reduced density matrix ($\varrho_0$) as an independent variational variable alongside the projector parameters.
  • They derive analytical Jacobians for the energy functional derivatives, significantly improving convergence speed and stability compared to previous self-consistent schemes.

3. Key Contributions

1. Development of a SC-Gutzwiller Framework: The paper establishes a robust variational framework capable of handling strong correlations and superconducting order simultaneously in multi-orbital systems, specifically addressing the gauge freedom issues in SC Gutzwiller states.
2. Identification of a Strongly Correlated SC (SC-SC) Phase: The study reveals a distinct superconducting regime at large $U$ that is qualitatively different from weak-coupling BCS superconductivity.
3. Discovery of a Novel "Small Fermi Liquid" (sFL): The authors identify a metastable normal state with a reduced Fermi surface volume ($\nu + 2$) that acts as the parent state for the SC-SC phase at large $U$.
4. Mechanism of Gap Reconstruction: The study elucidates how interaction-driven reconstruction leads to nodal, V-shaped density of states (DOS) and nematicity.

4. Key Results

The phase diagram is mapped at filling $\nu = 2.5$ as a function of Coulomb repulsion $U$ and anti-Hund's coupling $J_A$:

• Phase Diagram Structure:

  • BCS-SC (Weak Coupling): At small $U$, the system exhibits a BCS-like superconducting state.
  • FL Phase: A dome-shaped Fermi liquid (FL) phase separates the weak and strong coupling regimes.
  • SC-SC (Strong Coupling): At large $U$ ($\gtrsim 40$ meV), a Strongly Correlated Superconducting phase emerges.
    • Mechanism: The off-diagonal components of the Gutzwiller projector strongly suppress $f$-orbital charge fluctuations (reducing Coulomb energy) while maintaining finite pairing order and a sizeable quasiparticle weight $Z$. This distinguishes it from a conventional Mott insulator.
    • Nematicity: Competition between $J_A$ and $J_H$ leads to a mixed $s+d$-wave pairing state that breaks $C_{3z}$ rotational symmetry, resulting in a nematic SC state.
    • Gap Structure: At large $U$, interaction-driven reconstruction transforms the gap into a nodal structure with a V-shaped DOS.

• The Small Fermi Liquid (sFL):

  • In the large-$U$ regime, a novel normal state (sFL) becomes energetically competitive with the conventional FL.
  • Characteristics: The sFL is characterized by $[\hat{P}_R, \hat{N}_R] = 2\hat{P}_R$, implying a local singlet formation that reduces the effective Fermi surface volume to $\nu + 2$.
  • Role: The sFL is identified as the likely parent state of the SC-SC phase at large $U$, whereas the conventional FL is the parent at intermediate $U$.

• Gap Reconstruction:

  • As $U$ increases, the SC gap undergoes reconstruction. A kink in the quasiparticle weight $Z$ at $U \approx 51$ meV signals enhanced Kondo screening.
  • At $U \approx 54-55$ meV, a second reconstruction occurs, leading to the emergence of nodal lines and the V-shaped DOS, consistent with experimental tunneling spectra.

5. Significance

• Resolution of Experimental Contradictions: The results support the experimental finding that stronger interactions (larger $U$) can be detrimental to $T_c$ in certain regimes, while simultaneously explaining the emergence of unconventional pairing features (nematicity, nodal gaps) in the strong-coupling limit.
• Unified Picture of MATBG: The study provides a unified framework linking the heavy-fermion picture, phonon-mediated anti-Hund's coupling, and strong correlation effects. It suggests that MATBG superconductivity is a crossover from BCS-like to a strongly correlated state driven by the suppression of charge fluctuations.
• General Applicability: The developed Gutzwiller framework, particularly the handling of charge symmetry breaking and analytical derivatives, is versatile and applicable to other strongly correlated superconductors beyond graphene.
• Theoretical Insight: The identification of the sFL state offers a new perspective on the "parent state" of high-temperature superconductivity in moiré systems, suggesting that the reduction of Fermi surface volume via local singlet formation is a key precursor to strong-coupling superconductivity.