Sublattice polarization and filamentary superconductivity in strained graphene

This study reveals that in periodically strained graphene, strain-induced sublattice polarization suppresses superconductivity in high-density flat-band regions, forcing the pairing to instead emerge as robust quasi-one-dimensional filaments at geometric nodes where sublattice symmetry is restored.

The Gist

Imagine a sheet of graphene (a material made of a single layer of carbon atoms) not as a flat, smooth table, but as a wavy, rippled surface, like a piece of fabric that has been crumpled and then smoothed out just enough to leave deep, regular folds.

This paper explores what happens when you try to make electricity flow through this wavy sheet without resistance—a state called superconductivity.

Here is the story of their discovery, broken down into simple concepts:

1. The Setup: The "Magic" Waves

Usually, scientists think that if you can pack a lot of electrons into a small space (a high "density of states"), they will easily pair up and start superconducting. It's like thinking that if you crowd a dance floor with people, they will naturally start dancing together.

In this experiment, the researchers used "strain engineering." They stretched and compressed the graphene in a specific, repeating pattern. This created "pseudo-magnetic fields"—invisible forces that act like magnets but are actually caused by the physical bending of the material. These forces created "flat bands," which are regions where electrons get stuck and pile up, creating that high density of electrons everyone loves.

2. The Surprise: The Crowd is Stuck in Separate Rooms

The researchers expected these crowded "flat band" regions to become superconducting hotspots. Instead, they found the opposite.

The Analogy: Imagine the graphene sheet is a dance floor with two types of dancers: Team A and Team B. To dance together (superconduct), a Team A dancer needs to hold hands with a Team B dancer.

In the crowded, flat-band regions created by the waves, the "pseudo-magnetic fields" act like a strict bouncer. They force all the Team A dancers into one corner of the room and all the Team B dancers into the opposite corner. They are so far apart that they can't reach each other. Even though the room is packed with people, no one can dance because the two teams are completely separated.

This is called sublattice polarization. The electrons are so segregated by the strain that they can't form the pairs needed for superconductivity.

3. The Twist: Superconductivity Hides in the "Valleys"

So, if the crowded flat areas are dead zones for superconductivity, where does it happen?

The researchers found that as they increased the "push" to make electrons pair up, the superconductivity didn't get stronger in the crowded spots. Instead, it jumped to the geometric "nodes"—the specific points where the waves cross or flatten out (the crests and troughs of the ripples).

The Analogy: Think of the wave pattern as a series of hills and valleys.

• The Hills/Valleys (Flat Bands): The dancers are segregated and can't dance.
• The Nodes (The flat spots between waves): Here, the "bouncer" steps aside. Team A and Team B are allowed to mix freely. Because they can finally hold hands, superconductivity ignites here.

But it doesn't light up the whole floor. It forms thin, glowing filaments (like glowing wires) running along these specific nodes. The superconductivity is no longer a blanket covering the sheet; it's a set of narrow, one-dimensional highways.

4. The "Fingerprint" of Filamentary Superconductivity

To prove this strange new state exists, the researchers simulated dropping a small, non-magnetic "pebble" (an impurity) into the superconducting filament.

In normal superconductors, dropping a pebble creates a messy, blurry disturbance. But in these narrow, one-dimensional filaments, the pebble acts like a wall in a hallway. The electrons bounce back and forth between the pebble and the walls of the filament.

The Result: Instead of a mess, this creates a very sharp, distinct "echo" or resonance exactly at the edge of the energy gap. The paper suggests that if scientists can measure this specific "echo" in a real experiment, it will be the smoking gun proof that superconductivity is happening in these thin, filamentary lines rather than across the whole sheet.

5. Why This Matters (According to the Paper)

The paper concludes that in these strained materials, geometry is king. You can't just look at how many electrons are present; you have to look at where they are allowed to sit.

The strain acts like a "spatial filter." It separates the electrons based on their "team" (sublattice), effectively turning off superconductivity in the crowded areas and turning it on only in the specific geometric spots where the teams can mix. This creates a unique, filamentary form of superconductivity that is fundamentally different from what we see in other flat-band systems.

In short: Straining graphene creates a dance floor where the crowd is too segregated to dance, forcing the "dance" (superconductivity) to happen only in narrow, specific lanes where the segregation is lifted.

Technical Summary

Technical Summary: Sublattice Polarization and Filamentary Superconductivity in Strained Graphene

Problem Statement
While periodic strain engineering in monolayer graphene is known to generate colossal pseudo-magnetic fields (PMF) and flat pseudo-Landau levels (PLLs), the nature of the superconducting ground state in these systems remains unexplored. Conventional wisdom suggests that the divergent density of states (DOS) associated with flat bands universally enhances superconducting pairing. However, it is unclear whether this holds true for strain-engineered Dirac materials, where the unique interplay between strain-induced gauge fields and the lattice's sublattice structure may introduce novel constraints. Specifically, the question arises: does the high DOS in strain-modulated graphene enhance superconductivity as it does in twisted bilayer systems, or do geometric factors inhibit it?

Methodology
The authors investigate the superconducting phase of periodically corrugated graphene using a self-consistent Bogoliubov-de Gennes (BdG) framework.

• Model: The system is modeled as a tight-binding Hamiltonian with strain-modulated hopping integrals. The graphene lattice is subjected to a unidirectional sinusoidal ripple $z(x) = h \sin(2\pi x/L)$, which generates a periodic PMF.
• Pairing Mechanism: A phenomenological on-site attractive interaction $V$ is employed to mimic conventional s-wave pairing. This choice isolates the geometric effects of the strain-induced gauge fields from the complexities of unconventional pairing symmetries, provided the pairing is spin-singlet and requires inter-sublattice coherence.
• Simulation: The BdG equations are solved self-consistently at finite temperature ($T = 10^{-5}$) with the chemical potential set to the Dirac point ($\mu = 0$). The study utilizes a large supercell ($L = 500$ unit cells) with a corrugation ratio $r = 0.16$, ensuring the system remains uniform along the $y$-direction.

Key Results
The study reveals a striking, interaction-driven spatial crossover in the superconducting order parameter, governed by strain-induced sublattice polarization:

1. Sublattice Polarization in Flat Bands: In the normal state, the zeroth pseudo-Landau levels ($n=0$ PLLs) exhibit extreme sublattice polarization. The wavefunctions are spatially segregated: states from the $K$ valley localize exclusively on the A sublattice, while those from the $K'$ valley localize on the B sublattice. This results in a high DOS at the Fermi level that is hosted almost entirely by a single sublattice in specific spatial regions.
2. Suppression of Pairing in Flat-Band Regions: Contrary to the expectation that high DOS promotes superconductivity, the macroscopic coherence is significantly hindered in these flat-band regions. The spatial disjointedness between the A and B sublattices prevents the necessary inter-sublattice wavefunction overlap required for a robust s-wave condensate. Consequently, the pairing amplitude remains weak despite the divergent DOS.
3. Emergence of Filamentary Superconductivity: As the pairing interaction strength ($V$) increases, the system undergoes a spatial reconstruction. Robust superconductivity does not emerge in the high-DOS flat-band regions but instead relocates to the geometric nodes of the corrugation (where the local strain is zero and sublattice symmetry is restored). At these nodes, the pairing amplitude increases by more than an order of magnitude, forming robust, quasi-one-dimensional superconducting filaments.
4. Impurity Response: The introduction of non-magnetic impurities within these filamentary regions does not induce in-gap bound states (as seen in conventional 2D systems). Instead, strong impurities generate intense, symmetric resonant peaks precisely at the gap boundaries. This is attributed to the quasi-one-dimensional confinement, which acts as a hard-wall barrier, pushing resonance modes to the gap edges via multiple Andreev reflections.

Significance and Claims
The paper claims to identify a fundamental physical distinction between strain-induced flat bands and moiré flat bands (e.g., in magic-angle twisted bilayer graphene). While the latter possess a non-zero quantum metric that supports superfluid weight even with vanishing bandwidth, the strain-induced flat bands in this study lack such geometric protection due to extreme real-space sublattice segregation.

The authors conclude that in strain-engineered Dirac materials, the spatial distribution of wavefunctions, governed by sublattice degrees of freedom, is a decisive factor in determining superconducting properties. The work proposes a generalized principle where non-uniform strain acts as a "spatial phase filter," segregating regions capable of coherent pairing (nodes) from those where spatial disjointedness suppresses the condensate (flat bands). This leads to the emergence of filamentary superconductivity, a phenomenon that may offer a phenomenological analogy for understanding strain-induced superconductivity in other correlated systems, such as nickelates, where strain differentially tunes orbital overlaps. The predicted gap-edge resonances are proposed as a definitive local experimental signature for identifying this filamentary state.