Gaplessness from disorder and quantum geometry in gapped superconductors

This paper demonstrates that in sign-changing, fully gapped superconductors, the Fubini-Study metric of electronic Bloch wavefunctions governs the localization length of disorder-induced Andreev bound states, where an increased metric promotes delocalization and drives the system toward a gapless, dirty nodal superconductor state relevant to moiré graphene experiments.

The Gist

Imagine a superconductor as a perfectly organized dance floor where electrons pair up and move in perfect sync, creating a frictionless flow of electricity. Usually, this dance floor is smooth and has no "holes" in the energy levels; it's a solid, gapped system.

However, real-world materials are messy. They have impurities and disorder, like rocks scattered on that dance floor. In certain types of superconductors, these rocks can create tiny, isolated pockets where the dance rules flip upside down. At the borders of these pockets (called $\pi$-junctions), the electrons get stuck in a holding pattern, forming what physicists call Andreev bound states. Think of these as dancers trapped in a small, isolated corner of the room, unable to join the main flow. Usually, these trapped dancers stay put; they are "localized."

The Big Discovery
This paper asks a simple question: What if we could change the "shape" of the space these electrons live in?

The authors introduce a concept called Quantum Geometry. To use an analogy, imagine the electrons aren't just points on a map, but fuzzy clouds. In a normal material, these clouds are very tight and small. But in this specific type of material (inspired by "moiré" graphene, which is like stacking two sheets of honeycomb-patterned paper at a slight angle), the "clouds" of the electrons are naturally more spread out. The paper calls the measure of this spread the Fubini-Study metric.

The Mechanism: Stretching the Trap
The researchers found that when this "spread" (the quantum geometry) is increased, something amazing happens to those trapped dancers at the borders:

1. The Trap Gets Bigger: The "localization length" (the size of the corner where the dancer is stuck) gets longer. It's as if the corner of the room expands, giving the trapped dancer more room to move.
2. They Start Talking: Because the trapped states are now larger, they start to overlap with their neighbors. Instead of being isolated islands, they begin to "hybridize" or merge, creating a connected network.
3. The Result: Even though the material is supposed to be fully "gapped" (with no low-energy movement allowed), these expanded, overlapping trapped states create a new, low-energy path. The system starts behaving as if it has no gap at all, acting like a "dirty" superconductor with free-moving particles, even though the underlying material is technically gapped.

What They Measured
To prove this, the team ran computer simulations (like a digital twin of the material) and looked at three main things:

• The "Spread" of the Wave: They measured how much the electron waves were spread out. As the quantum geometry increased, the waves spread over more of the material, confirming they were becoming less "trapped."
• Stiffness (The Dance Floor's Rigidity): They measured how hard it is to twist the flow of the supercurrent. In a perfect superconductor, this is very stiff. In their "messy" system, as the quantum geometry increased, the stiffness dropped in a specific way that mimics a material with no energy gap.
• The "Fermi Surface": In a normal metal, electrons fill up a specific shape of energy levels called a Fermi surface. In a gapped superconductor, this surface disappears. However, the authors found that in their disordered system, these trapped states reassembled to form a "Bogoliubov Fermi surface"—a ghostly, gapless structure that looks like a metal, even though the material is a superconductor.

The Real-World Connection
The paper connects this theory to recent experiments with moiré graphene superconductors. These are real materials where scientists have observed strange, gapless behaviors that didn't fit the standard models. The authors suggest that these experiments might not be seeing "true" gapless superconductors (where the gap is naturally zero), but rather gapped superconductors where disorder and quantum geometry have combined to create a "fake" gapless state by stretching out the trapped electron states.

In Summary
The paper demonstrates that disorder (messiness) combined with quantum geometry (the natural spread of electron clouds) can turn a perfectly gapped superconductor into a system that behaves like it has no gap. The "trapped" states at the boundaries of disorder don't just stay stuck; they stretch out, connect, and create a low-energy highway for electrons, fundamentally changing how the material conducts electricity and heat.

Technical Summary

Technical Summary: Gaplessness from Disorder and Quantum Geometry in Gapped Superconductors

Problem Statement
Disorder in sign-changing superconductors is known to induce low-energy Andreev bound states (ABS) at $\pi$-junctions, which are typically localized. While flat-band superconductivity exhibits robustness against disorder, the interplay between quenched disorder, electronic interactions, and quantum geometry (QG) in two-dimensional superconductors remains largely unexplored. Specifically, it is unclear how the intrinsic quantum geometric properties of electronic Bloch wavefunctions—specifically the Fubini-Study metric—affect the localization length of impurity-induced subgap states and the resulting low-energy thermodynamic and spectral properties of an intrinsically gapped superconductor.

Methodology
The authors employ a combination of analytical low-energy theory and self-consistent mean-field numerical simulations on a microscopic lattice model.

1. Analytical Framework: They derive a one-dimensional Bogoliubov–de Gennes (BdG) Hamiltonian for a single $\pi$-junction. The quantum geometry is modeled via a momentum-dependent pair function, $\Delta(k) \approx \Delta_0(1 - \zeta^2 a^2 k^2)$, where $\zeta$ parameterizes the quantum geometric tensor and $a$ is a lattice scale. This allows for the derivation of an analytical expression for the localization length $\xi$ of zero-energy bound states.
2. Microscopic Model: The study utilizes a two-dimensional interacting electron model on a square lattice with two orbitals. The Hamiltonian includes a flat-band component ($H_{flat}$) controlled by the parameter $\zeta$ (which sets the quantum metric but keeps the band dispersion flat) and a dispersive component ($H_{disp}$) controlled by $t'$ to introduce a finite Fermi velocity.
3. Disorder Implementation: Inhomogeneity is introduced by promoting the nearest-neighbor attractive interaction $V(r)$ to a random variable taking values $V_1$ and $V_2$ with exponentially decaying spatial correlations (correlation length $\ell$). This creates domains of sign-changing order parameters ($+\Delta_1, -\Delta_2$), effectively generating a network of $\pi$-junctions.
4. Numerical Analysis: The authors perform self-consistent mean-field calculations on lattices (e.g., $18 \times 18$ and $22 \times 22$) to compute disorder-averaged quantities, including the Inverse Participation Ratio (IPR), superfluid stiffness $\rho_s(T)$, and momentum-resolved spectral functions $A(k, \omega)$.

Key Contributions and Results

• Control of Localization Length by Quantum Geometry: The authors demonstrate that the localization length $\xi$ of ABS at $\pi$-junctions is not solely determined by the intrinsic coherence length $\xi_0 \sim v^*_F/\Delta$ but is significantly enhanced by the quantum geometric contribution. The derived relation is:
  $$\xi = \frac{1}{2} \left( \xi_0 + \sqrt{\xi_0^2 + 4\zeta^2 a^2} \right)$$
  Numerical results confirm that increasing the Fubini-Study metric parameter $\zeta$ increases the localization length, even at fixed disorder strength.
• Delocalization Tendency: Analysis of the Inverse Participation Ratio (IPR) reveals that while the states remain fundamentally localized (as expected for Class CI systems in 2D), increasing $\zeta$ leads to a stronger tendency toward delocalization. The wavefunctions spread over a larger fraction of the lattice sites, and the IPR shows a decreasing trend with system size before saturation.
• Gapless Thermodynamic Signatures: The network of disorder-induced $\pi$-junctions, combined with the enhanced localization length, results in low-energy properties resembling a dirty nodal superconductor. The superfluid stiffness $\rho_s(T)$ exhibits a power-law suppression at low temperatures (exponents ranging from $\approx 3.7$ to $2.0$ depending on $\zeta$), indicative of gapless excitations, rather than the exponential suppression expected for a fully gapped system.
• Spectral Function and "Bogoliubov Fermi Surface": The momentum-resolved spectral function $A(k, \omega)$ shows that disorder-induced subgap states form a "Bogoliubov Fermi surface" along the original electronic Fermi surface. At small $\zeta$, these appear as isolated impurity states; at large $\zeta$, they mimic an impurity "band" due to enhanced hybridization, yet remain localized below the bulk gap. The spatially resolved density of states indicates that the system appears gapped deep within individual domains, with the global gaplessness arising solely from the Andreev bound states.

Significance and Claims
The paper posits that quantum geometry provides a nontrivial mechanism to control the spatial extent of disorder-induced bound states in superconductors. The primary significance lies in the finding that a nominally gapped superconductor with sign-changing order parameters can exhibit low-energy properties indistinguishable from a gapless (nodal) superconductor if the quantum geometric contribution to the localization length is sufficiently large.

The authors place these results in the context of recent experiments on moiré graphene superconductors, where unconventional superconductivity and intrinsic inhomogeneities are observed. They suggest that the observed "gapless" behavior in such materials might not necessarily imply intrinsic nodal quasiparticles but could instead arise from disorder-induced bound states whose localization is enhanced by quantum geometry. The work highlights the challenge of experimentally distinguishing between a clean nodal superconductor and a gapped superconductor with disorder-induced, geometrically enhanced delocalized bound states. The authors modestly note that while their results rely on mean-field theory, future work involving numerically exact quantum Monte Carlo computations and analysis of optical/thermal conductivity could further elucidate these effects.