At graphene's charge-neutrality (Dirac) point, the minimum conductivity σmin⁡\sigma_{\min} interpolates between the pseudodiffusive ballistic value 4e2/πh4e^2/\pi h set by evanescent modes and a disorder-controlled 'dirty' value near 4e2/h4e^2/h generated by electron-hole puddles from charged substrate impurities. Sajjad, Tseng, Habib and Ghosh (Phys. Rev. B 92, 205408, 2015) mapped this crossover as a 'fan diagram' of σmin⁡\sigma_{\min} versus impurity density nimpn_{\mathrm{imp}} and device aspect ratio W/LW/L, combining a Landauer theory with Fermi-golden-rule scattering (self-consistent Thomas-Fermi screening, puddle-variance fit σE2≈2ℏ2vF2nimp+C\sigma_E^2 \approx 2\hbar^2 v_F^2 n_{\mathrm{imp}} + C) with non-equilibrium Green's function (NEGF) simulations averaged over Gaussian impurity puddles. The study rests on approximations and idealizations identifiable in its text, though not consolidated there into an author-acknowledged limitations list: the low-energy expansion and Thomas-Fermi screening used in the analytics, the idealized Gaussian-correlated disorder model with fixed screening length ζ∼3\zeta \sim 3 nm, the dirty-limit-only validation of the variance fit, finite impurity-configuration sampling in the numerics (the authors attribute analytical/numerical discrepancies to 'the lack of adequate samples'), and a simplified contact model (Fermi level pinned under the contacts, effective doping EF=0.25E_F = 0.25 eV). No later work has established whether the fan diagram — its aspect-ratio flip, its quasi-saturation at ∼4e2/h\sim 4e^2/h, and its variance fit — survives converged configuration averaging, realistic substrate-specific disorder, and controlled contact modeling; the quantitative robustness of the σmin⁡(nimp,W/L)\sigma_{\min}(n_{\mathrm{imp}}, W/L) map stated below therefore remains the open problem.

Background

Graphene, a single atomic layer of carbon, has electrons whose low-energy band structure is described by a massless Dirac Hamiltonian: the energy varies linearly with momentum, E=±ℏvF∣k∣E = \pm\hbar v_F |k|, with pseudospin (a sublattice-isospin degree of freedom locked to the momentum direction). At the charge-neutrality point (the Dirac point) the conduction and valence bands touch and the density of states vanishes, so transport is unusual. In a short, wide, perfectly clean sample contacted by metals, conduction at the Dirac point is carried by evanescent (exponentially decaying) modes, giving a 'pseudodiffusive' minimum conductivity σmin⁡=4e2/πh\sigma_{\min} = 4e^2/\pi h that mimics Ohmic scaling (Tworzydło et al., 2006). In real samples on substrates such as SiO2_2, charged impurities trapped in the substrate dope the sheet inhomogeneously, breaking it into spatially fluctuating electron and hole regions ('charge puddles'); averaging the linear Dirac density of states over a Gaussian distribution of local Dirac-point shifts with variance σE2\sigma_E^2 produces a finite minimum density of states and hence a finite conductivity. A self-consistent Boltzmann-plus-puddles theory (Adam, Hwang, Galitski and Das Sarma, 2007; Galitski, Adam and Das Sarma, 2007; Li, Hwang and Das Sarma, 2011) and experiments (Chen et al., 2008) established that the residual conductivity near the Dirac point scales with impurity density and is measured to be of order 4e2/h4e^2/h; the review by Das Sarma, Adam, Hwang and Rossi (2011) presents this 'standard model' as quantitatively adequate. Within this landscape, Sajjad, Tseng, Habib and Ghosh (2015) computed the full crossover — a 'fan diagram' of σmin⁡\sigma_{\min} versus impurity density nimpn_{\mathrm{imp}} and device aspect ratio W/LW/L (width over length) — by combining Landauer transport theory (conductance as a sum over transverse-mode transmission probabilities) with scattering rates from Fermi's golden rule for a Thomas-Fermi-screened Coulomb potential (a coarse static screening approximation with wave vector qTFq_{TF} determined self-consistently from the puddle-enhanced density of states), and, independently, atomistic NEGF (non-equilibrium Green's function, a quantum-transport simulation method treating contacts via self-energies) simulations on lattices decorated with Gaussian impurity potentials of width ζ∼3\zeta \sim 3 nm and amplitude standard deviation 100 meV, configuration-averaged over roughly 800 realizations. They found an aspect-ratio-dependent flip in the trend of σmin⁡\sigma_{\min} versus nimpn_{\mathrm{imp}} (increasing for W/L>1W/L > 1, decreasing for W/L<1W/L < 1), universal convergence toward 128e2/π3h≈4e2/h128e^2/\pi^3 h \approx 4e^2/h in the dirty limit, a simple fitted relation σE2≈2ℏ2vF2nimp+C\sigma_E^2 \approx 2\hbar^2 v_F^2 n_{\mathrm{imp}} + C with C=0.027C = 0.027 eV2^2 (valid mainly in the dirty limit), and — in the ballistic, contact-doped (pinned Fermi level, effective doping EF=0.25E_F = 0.25 eV) geometry — a striking absence of Fabry-Pérot interference oscillations (conductance oscillations from quantum interference between multiply reflected waves at the contact interfaces) in the lowest transverse mode of single-layer graphene, interpreted as a signature of Klein tunneling (perfect transmission through a potential barrier at normal incidence, enforced by pseudospin conservation). The model's quantitative limits are identifiable in the source text, although the paper does not consolidate them into an acknowledged limitations list: the analytics rely on a low-energy expansion E≪ℏvFκE \ll \hbar v_F \kappa (with κ\kappa the screening wave vector) and on Thomas-Fermi screening; the disorder model is Gaussian-correlated disorder with fixed screening length ζ∼3\zeta \sim 3 nm and without substrate specificity; the variance fit is validated only in the dirty limit; the NEGF configuration averaging is statistically under-converged, the authors themselves attributing analytical/numerical discrepancies to 'the lack of adequate samples'; and the contact model pins the Fermi level under the contacts at an effective doping EF=0.25E_F = 0.25 eV.

Problem Statement

Determine, with quantitatively controlled and converged calculations, whether the graphene minimum-conductivity fan diagram of Sajjad et al. (2015) — the map of σmin⁡\sigma_{\min} versus charged-impurity density nimpn_{\mathrm{imp}} and device aspect ratio W/LW/L, with its aspect-ratio-dependent trend flip, its quasi-saturation toward ∼4e2/h\sim 4e^2/h (128e2/π3h128e^2/\pi^3 h) in the dirty limit, and its puddle-variance fit σE2≈2ℏ2vF2nimp+C\sigma_E^2 \approx 2\hbar^2 v_F^2 n_{\mathrm{imp}} + C — survives beyond the approximations in which it was derived. For each of the three headline features, report the deviation from the corresponding source value and, within a stated uncertainty band — the configuration-averaged sampling error combined with the spread over the screening, disorder, and contact variants — whether the feature survives. The source values are: for the trend flip, the sign of dσmin⁡/dnimp\mathrm{d}\sigma_{\min}/\mathrm{d}n_{\mathrm{imp}} (increasing for W/L>1W/L > 1, decreasing for W/L<1W/L < 1); for the quasi-saturation, 128e2/π3h128e^2/\pi^3 h; and for the variance fit, σE2≈2ℏ2vF2nimp+C\sigma_E^2 \approx 2\hbar^2 v_F^2 n_{\mathrm{imp}} + C with C=0.027C = 0.027 eV2^2 in the dirty limit. Concretely: (a) establish the regime of validity of the analytic derivation, in particular the low-energy expansion E≪ℏvFκE \ll \hbar v_F \kappa and Thomas-Fermi screening, and determine whether the variance fit remains accurate at low-to-moderate nimpn_{\mathrm{imp}} once screening is treated beyond the self-consistent Thomas-Fermi approximation; (b) test the sensitivity of the map to the disorder model by replacing the idealized Gaussian puddles with substrate-specific impurity realizations for at least one representative experimentally characterized substrate (e.g., SiO2_2), with the impurity spatial statistics taken from representative scanning-probe (puddle-spectroscopy) measurements, selected and cited in the calculation, rather than assumed Gaussian; (c) redo the configuration-averaged NEGF (or equivalent quantum-transport) numerics with demonstrably converged statistical sampling (source baseline: roughly 800 random impurity configurations), resolving whether the reported analytical/numerical discrepancies — which the authors attribute to 'the lack of adequate samples' — are purely finite-sampling artifacts; and (d) quantify how the source's contact-doping model (Fermi level pinned under the contacts, effective doping EF=0.25E_F = 0.25 eV) shifts the fan diagram, comparing the resulting map against the available measured σmin⁡\sigma_{\min} data, which currently span impurity density and geometry only partially (e.g., Chen et al. 2008 covers the dirty regime). All four items (a)–(d) are required. The substrate-realism objective is confined to the representative-substrate form of item (b); a requirement that the disorder model be adequate for all experimental substrates and impurity realizations would have no natural completion boundary.

References

  1. R. N. Sajjad, F. Tseng, K. M. Masum Habib, A. W. Ghosh, "Quantum transport at the Dirac point: Mapping out the minimum conductivity from pristine to disordered graphene", Phys. Rev. B 92, 205408 (2015). DOI: 10.1103/PhysRevB.92.205408; arXiv:1507.01257, https://arxiv.org/abs/1507.01257 (source paper)
  2. J. Tworzydło, B. Trauzettel, M. Titov, A. Rycerz, C. W. J. Beenakker, "Sub-Poissonian shot noise in graphene", Phys. Rev. Lett. 96, 246802 (2006). DOI: 10.1103/PhysRevLett.96.246802, https://doi.org/10.1103/PhysRevLett.96.246802
  3. S. Adam, E. H. Hwang, V. M. Galitski, S. Das Sarma, "A self-consistent theory for graphene transport", Proc. Natl. Acad. Sci. USA 104, 18392 (2007). DOI: 10.1073/pnas.0704772104, https://doi.org/10.1073/pnas.0704772104
  4. V. M. Galitski, S. Adam, S. Das Sarma, "Statistics of random voltage fluctuations and the low-density residual conductivity of graphene", Phys. Rev. B 76, 245405 (2007). DOI: 10.1103/PhysRevB.76.245405, https://doi.org/10.1103/PhysRevB.76.245405
  5. Q. Li, E. H. Hwang, S. Das Sarma, "Disorder-induced temperature-dependent transport in graphene: Puddles, impurities, activation, and diffusion", Phys. Rev. B 84, 115442 (2011). DOI: 10.1103/PhysRevB.84.115442, https://doi.org/10.1103/PhysRevB.84.115442
  6. J.-H. Chen, C. Jang, S. Adam, M. S. Fuhrer, E. D. Williams, M. Ishigami, "Charged-impurity scattering in graphene", Nat. Phys. 4, 377 (2008). DOI: 10.1038/nphys935, https://doi.org/10.1038/nphys935
  7. S. Das Sarma, S. Adam, E. H. Hwang, E. Rossi, "Electronic transport in two-dimensional graphene", Rev. Mod. Phys. 83, 407 (2011). DOI: 10.1103/RevModPhys.83.407, https://doi.org/10.1103/RevModPhys.83.407
  8. Z. Shi, B. Sbierski, P. W. Brouwer, "Disorder correction to the minimal conductance of a nodal-point semimetal", Phys. Rev. B 102, 024204 (2020). DOI: 10.1103/PhysRevB.102.024204; arXiv:2002.10353, https://arxiv.org/abs/2002.10353
  9. A. Benyahia, R. Bouamrane, "Modelling the minimum conductivity of graphene using random resistor networks", Physica A 626, 129078 (2023). DOI: 10.1016/j.physa.2023.129078, https://doi.org/10.1016/j.physa.2023.129078