Superconducting Graphene — Superconductivity Through Geometric Rotation

Discovery

Technical Explanation

1. The moiré pattern and superlattice. When two graphene layers are stacked with a twist of θ ≈ 1.1°, the geometric interference between the two hexagonal lattices produces a moiré pattern with a period λ = a/(2 sin(θ/2)) ≈ 13 nm (where a = 0.246 nm is the graphene lattice parameter). This pattern acts as an artificial crystalline superlattice with a unit cell ~50× larger than that of monolayer graphene.

2. Flat bands. In monolayer graphene, electrons at the Dirac points have a Fermi velocity v_F ≈ 10⁶ m/s and a linear dispersion relation. In tBLG at the magic angle, interlayer hybridization flattens the energy bands near the Fermi level. The effective Fermi velocity drops to v_F* ≈ 10⁴ m/s — a ~100-fold reduction. The electron kinetic energy (∝ v_F²) becomes negligible compared to the Coulomb interaction energy (~e²/εd, where d ~ 13 nm). The system enters a strongly correlated regime, analogous to heavy-fermion systems or cuprates.

3. The phase diagram. By tuning the carrier density via an electrostatic gate, the system reveals a rich phase diagram: at filling ν = −2 (2 fewer electrons per moiré unit cell), a correlated insulator appears (analogous to a Mott insulator). Slight doping around ν = −2 ± δ triggers the emergence of superconductivity. This behavior — Mott insulator → superconductor upon doping — is the hallmark of cuprates (YBa₂Cu₃O₇, La₂CuO₄), suggesting an unconventional pairing mechanism, possibly mediated by magnetic fluctuations rather than phonons (classical BCS).

4. Fabrication and angular control. The "tear-and-stack" technique involves exfoliating a graphene monolayer, tearing it in two using an AFM tip, then restacking the halves with the desired rotation via a PPC (propylene polycarbonate) polymer on a heated substrate. Angular control must be precise to ±0.1° — a deviation of 0.3° suppresses superconductivity entirely. Encapsulation in hBN (hexagonal boron nitride) shields the graphene and reduces disorder. Electrical contacts are patterned via electron-beam lithography (e-beam, ~10 nm resolution).

Why It Worked

The theoretical prediction by Bistritzer and MacDonald (2011) had gone largely unnoticed: no one believed it possible to control the twist angle to ±0.1°. The MIT team developed exceptional mastery of the tear-and-stack technique, combined with ultra-low-temperature transport measurements (dilution refrigerator, base T ~50 mK). The surprise is not that graphene becomes superconducting at 1.7 K (a very low temperature), but that a 2D material this simple, controlled solely by its geometry, reproduces the physics of the cuprates — which require complex chemical compositions like YBa₂Cu₃O₇₋ₓ.

Causal Chain

Graphene isolated by Geim & Novoselov (2004) → Bistritzer-MacDonald flat band theory (2011) → Tear-and-stack technique perfected (MIT, 2016–2017) → Observation of correlated insulator in magic-angle tBLG (Nature, March 2018, paper 1) → Observation of superconductivity in tBLG (Nature, March 2018, paper 2) → "Twistronics" founded — electronic properties controlled by geometry → Extensions: twisted trilayer graphene (2021, T_c ≈ 2.9 K), WSe₂/WS₂ moiré systems

Anecdote: Yuan Cao, the MIT PhD student and first author of both 2018 Nature papers, published these two landmark discoveries at the age of 21. He had begun his undergraduate studies at the University of Science and Technology of China (USTC) at 14. Both papers were published on the same day — an exceedingly rare event in Nature.

Sources

References verified during the August 2026 fact-checking audit: these are the pages
against which this bulletin's claims were checked.

1. Unconventional superconductivity in magic-angle graphene superlattices — Cao et al., Nature 556, 43-50 (2018)
2. How Magical Is Magic-Angle Graphene? — revue de référence, Matter
3. Supraconductivité accordable dans le tricouche torsadé — Graphene-Info