Skip to main content
Physics· 14-page report· 1 figure

Graphene Tight-Binding Dispersion & Dirac Physics

Compute graphene tight-binding band structure and visualize Dirac-cone physics near the Fermi level.

What this research found

Graphene's electronic structure was derived analytically and then checked numerically using the minimal tight-binding model: one π orbital per carbon atom, a lattice constant of 2.46 Å, and a nearest-neighbour hopping integral of 2.7 eV. The calculation reproduces the two Dirac points at the Brillouin-zone corners, a Fermi velocity of 8.7390 × 10⁵ m/s, and a density of states that vanishes linearly at the Dirac energy. Adding next-nearest-neighbour hopping shifts the band touching to −0.81 eV and breaks the electron–hole symmetry of the spectrum.

  • The conduction and valence bands touch exactly at the two inequivalent zone corners K and K′. Computed gaps of 2.472 × 10⁻¹⁵ eV and 2.449 × 10⁻¹⁵ eV are zero to floating-point precision, matching the analytical proof that the structure factor vanishes there.
  • The closed-form Fermi velocity √3ta/2ℏ evaluates to 8.739040 × 10⁵ m/s, and extracting the same slope numerically from the computed band gave 8.739035 × 10⁵ m/s — agreement to a relative error of 5.7 × 10⁻⁵ percent.
  • The density of states rises linearly from zero at the Dirac energy, the signature of a two-dimensional Dirac semimetal. A fit over 0.1–1.0 eV gives a slope of 0.05189 per eV² against the analytic value of 0.0504 per eV², with R² = 0.9991, and van Hove singularities appear at ±2.7 eV with band edges at ±8.1 eV.
  • A next-nearest-neighbour hopping of −0.27 eV opens no gap but moves the Dirac point to −0.81 eV and makes the sum of the two bands k-dependent, so no energy exists about which the spectrum is symmetric. The L1 asymmetry of the density of states rises from 0 to 1.78, and the van Hove peaks split to −3.24 and +2.16 eV.
  • The bare nearest-neighbour Fermi velocity sits 12.61% below the accepted experimental value of 1.0 × 10⁶ m/s. The shortfall is attributed chiefly to Coulomb renormalisation of the Dirac cone, with the note that a hopping of roughly 3.1 eV would largely remove it.

How it was done

The two-atom honeycomb basis was used to build the 2 × 2 Bloch Hamiltonian, giving the closed-form dispersion E± = ±t|f(k)|, plotted along the high-symmetry path Γ → K → M → Γ at 200 points per segment. Degeneracy was verified by evaluating the structure factor at both zone corners, and the Fermi velocity was obtained twice over: analytically by linearising about K, and numerically by a central finite difference on the computed band at a radial offset of 0.001 per Å. The density of states was built by sampling both bands on a uniform 1000 × 1000 grid across the reciprocal cell with 0.02 eV Gaussian broadening, then checked against the requirement that it integrate to 2.0000 states per unit cell. The model was extended with a next-nearest-neighbour diagonal term and the resulting spectral asymmetry quantified with an L1 metric, and the whole treatment was written up as a 14-page technical report with derivations, band-structure figures, and a table of raw numerical outputs.

Data sources

  • Wallace, Physical Review 71:622 (1947) — the original honeycomb tight-binding derivation
  • Castro Neto, Guinea, Peres, Novoselov & Geim, Reviews of Modern Physics 81:109 (2009)
  • Reich, Maultzsch, Thomsen & Ordejón, Physical Review B 66:035412 (2002) — tight-binding parameters
  • Elias et al., Nature Physics 7:701 (2011) — Fermi velocity measured in suspended graphene

Limitations

The single-particle model omits electron–electron interactions, finite orbital overlap, and substrate effects, which is why its Fermi velocity falls short of experiment by roughly one part in eight. The linear-cone result holds only very close to the Dirac point, before trigonal warping and band curvature set in.

How this research was produced

K-Dense Web planned and ran this physics investigation end to end — gathering the sources, carrying out the analysis, producing the figures, and drafting the report. The full session transcript, including every intermediate step, is available to view.

Share:
Physics

Simulating and reconstructing PLATON scintillator tracks

Simulate PLATON scintillator particle tracks and validate reconstruction performance against known geometries.

Physics

AME2020 SEMF Binding Energy Fit

Fit the semi-empirical mass formula to AME2020 binding energies and map residuals across the nuclear chart.

Physics

ATLAS Higgs ML Challenge Classifier

Train and evaluate classifiers on the ATLAS Higgs ML Challenge with a documented end-to-end methodology pipeline.

Run this kind of analysis on your own question

Try K-Dense Web free and see how an AI co-scientist accelerates your research.