arXiv News

First-hand information for everyone

Site

AboutPrivacy PolicyTerms of Use

© 2026 arXiv News

arXiv News

EnglishJapanese

Switch language

EnglishJapanese
Loading account…

First-hand information for everyone

Back to front page
MathematicsEnglishPublished

Trace formula for magic angles in a wider class of twisted bilayer graphene models

September 14, 2026arXiv: 2609.13080v1

This note extends a striking mathematical identity about “magic angles” in the chiral model of twisted bilayer graphene. Earlier work by Becker et al. showed that, for the exact Bistritzer–MacDonald potential, the sum of the fourth powers of the generalized magic angles equals 8π/√3. The author proves a version of that identity for a larger family of potentials that share the lattice and rotational symmetries of the model.

The paper works with the chiral continuum model. In that model one studies a differential operator that depends on a complex potential U(z) and a parameter α that is essentially the inverse of the physical twist angle. Magic angles are special values of α at which the operator develops flat electronic bands. Mathematically, these magic values appear as inverse eigenvalues of a compact operator built from U(z) and the derivative operators. The author assumes the potential respects the model symmetries and can be written with finitely many Fourier modes of a specific form.

Under these assumptions the main result is an explicit trace formula for the fourth power sum of the inverse magic angles. Writing the potential in the chosen Fourier basis with real coefficients c_n and a lattice parameter K, the paper shows

sum_{α in A} α^{-4} = 8 sqrt(3) π K^4 \sum_{n,m} (c_n^2 c_m^2 + 2 c_n^2 c_m c_{-n-m})/(n^2 + n m + m^2).

Here the left sum counts each magic value with its algebraic multiplicity. The c_n are the Fourier coefficients that describe the allowed potential, and the integer K comes from the dual lattice used in the model. For the exact Bistritzer–MacDonald potential, choosing the corresponding coefficients recovers the earlier 8π/√3 identity reported by Becker et al.

How did the author get this formula? At a high level the argument turns the spectral question into a trace computation. The relevant compact operator has only off-diagonal parts, so odd powers have zero trace and the fourth-power trace is the first nontrivial case. The paper computes that trace by expanding in a Fourier basis and then evaluating a finite sum of residues coming from meromorphic integrals. Symmetries of the potential are used repeatedly to simplify and group terms.

Important caveats follow. The result holds for the idealized chiral model, not for the full physical Hamiltonian that includes additional terms. The allowed potentials are restricted: they must obey the lattice and threefold rotation symmetries and be expressible with finitely many specified Fourier modes. The magic-angle sum is taken with algebraic multiplicity, and several technical steps use trace-class and meromorphicity properties of the operators. The note therefore gives an exact mathematical identity in a clear setting, but it does not by itself settle questions about what happens under generic perturbations or in more complete physical models.

Source paper

Trace formulas for TBG

MathematicsMathematicsMathematics
View on arXiv
Trace formula for magic angles in a wider class of twisted bilayer graphene models | arXiv News