Optical Signatures of Band Flatness and Anisotropic Quantum Geometry in Magic-Angle Twisted Bilayer Graphene

Technical Summary: Optical Signatures of Band Flatness and Anisotropic Quantum Geometry in Magic-Angle Twisted Bilayer Graphene

Problem Statement
Moiré superlattice materials, particularly magic-angle twisted bilayer graphene (TBG), serve as ideal platforms for engineering strongly correlated phases such as flat band superconductivity and fractional Chern insulators (FCI). While the emergence of nearly flat bands is central to these phenomena, a systematic quantification of the "degree of band flatness" and the associated anisotropic quantum geometry remains lacking. Existing experimental techniques like angle-resolved photoemission spectroscopy (ARPES) and scanning tunneling microscopy (STM) can observe flat band features and high density of states, respectively, but they do not provide a unified framework to measure the precise conditions required for the saturation of quantum geometric inequalities or the transition to ideal flat bands. Furthermore, the relationship between band flatness, emergent symmetries, and the specific optical signatures of anisotropic quantum geometry requires further theoretical elucidation.

Methodology
The authors employ a continuum model of TBG-hBN heterostructures based on the Bistritzer-MacDonald (BM) Hamiltonian, incorporating lattice relaxation via a parameter $\kappa = w_0/w_1$ (the ratio of AA to AB interlayer hopping). To break gapless points and ensure well-defined quantum geometric tensors, a sublattice staggered potential ($\Delta$) is introduced to simulate the presence of an hBN substrate.

The study utilizes linearly polarized optical conductivity calculated via the Kubo-Greenwood formula. The authors connect the real part of the optical conductivity to the quantum geometric tensor (comprising the quantum metric $g$ and Berry curvature $\Omega$) through the interband contribution. A central tool in this analysis is the "optical bound" (an inequality derived from the trace condition of quantum geometry):
$$K_{OP}(\omega) = 2[\Re\sigma_{xx}(\omega) + \Re\sigma_{yy}(\omega)] \geq 2\Im\hat{\sigma}_{xy}(\omega)$$
where $\Im\hat{\sigma}_{xy}(\omega)$ is the imaginary part of the generalized optical Hall conductivity. The authors systematically vary the twist angle ($\theta$) and the lattice relaxation parameter ($\kappa$) to map the evolution of band structures, optical bounds, and quantum geometric properties from narrow bands to ideal flat bands.

Key Contributions and Results

1. Optical Signatures of Band Flatness Regions:
   The study categorizes the twist angle into three distinct regions based on optical conductivity features:

   • Region 1 ($\theta_c1 > \theta \geq 1.2^\circ$): The bandwidth is larger than the Coulomb interaction strength (~25 meV). Optical conductivity shows a broad zone and low isolated peaks, indicating less localized electrons and a lower likelihood of superconductivity.
   • Region 2 ($1.2^\circ > \theta > 1.0^\circ$): This region corresponds to the experimentally observed superconducting window. The optical conductivity exhibits a narrow, isolated peak in the low-energy region, indicating a bandwidth smaller than the electron interaction. This narrowness implies a large effective mass and high localization, satisfying the critical condition for flat band superconductivity.
   • Region 3 ($\theta_c2 < \theta \leq 1.0^\circ$): The gap between the nearly flat band and the dispersive band becomes very small. The low-energy peak is high, and transitions overlap, suggesting a smaller effective mass and a gap that may hinder superconductivity due to high kinetic energy from dispersive bands, though it may facilitate other phases.

2. Optical Bounds and Ideal Flat Bands:
   By tuning lattice relaxation ($\kappa$), the authors demonstrate the evolutionary process toward an ideal flat band. As $\kappa$ decreases:

   • The first peak of the optical bound decreases.
   • The trace condition (TC) tends to saturate.
   • The optical bound ($K_{OP}$) and the imaginary part of the generalized optical Hall conductivity ($2\Im\hat{\sigma}_{xy}$) increasingly overlap.
   • Complete overlap signifies the saturation of the TC, a hallmark of the ideal flat band limit.

3. Berry Curvature and the Refined Trace-Determinant Inequality (TDI):
   The study investigates the sign-definiteness of the Berry curvature. Using the refined TDI ($\text{tr}g \geq 2\sqrt{\det g} \geq |\Omega|$), the authors show that as the system approaches the ideal flat band limit (specifically when $\kappa \approx 0.6$), the negative component of the Berry curvature vanishes, becoming entirely positive (or negative).

   • For the valence flat band, the ratio of negative to positive Berry curvature components saturates to zero.
   • For all occupied bands, the ratio remains slightly non-zero because not all bands are perfectly flat.
   • This vanishing of the negative component is a necessary condition for the emergence of FCI phases.

4. Anisotropic Quantum Geometry and Saturation Conditions:
   The paper distinguishes between isotropic and anisotropic systems:

   • Isotropic Case: The saturation of the trace condition is enforced by the vanishing of flat band velocities and emergent chiral symmetry ($\kappa=0$). This leads to $g_{xx} = g_{yy}$ and $g_{xy} = 0$.
   • Anisotropic Case: In anisotropic systems, the trace condition is not saturated; instead, the determinant condition (DC) can be saturated. Here, the quantum metric and Berry curvature are related by a "saturation matrix" involving a complex constant $c$.
   • The constant $c$, representing intrinsic anisotropy or lattice deformation, can be experimentally measured via the Souza-Wilkens-Martin sum rule using optical weights ($W_{aa}$) and the generalized optical Hall conductivity ($C$).
   • The non-vanishing difference between the optical bound and the imaginary part of the generalized optical Hall conductivity serves as a specific signature of anisotropy in FCI phases.

Significance
The paper claims to provide a unified perspective for studying flat band systems by establishing optical conductivity and optical bounds as experimental probes for the degree of band flatness and anisotropic quantum geometry. It demonstrates that:

• Optical absorption peaks and their widths can distinguish between regions conducive to superconductivity versus those conducive to fractional Chern insulating phases.
• The saturation of the trace and determinant conditions, driven by emergent chiral symmetry and vanishing velocities, can be directly linked to measurable optical quantities.
• The "saturation matrix" and the constant $c$ offer a method to quantify intrinsic anisotropy and lattice deformation in moiré materials without relying solely on structural characterization.

The authors assert that these findings are directly applicable to other moiré superlattice materials and kagome materials, offering a theoretical framework to interpret optical data in the context of topological and correlated phases.