Conductivity of the Landau levels of two-dimensional Dirac cones and gapped nodal-rings in the quantum limit under impurity-potentials

Abstract: We investigate the dc magnetoconductivity of two-dimensional Dirac cones and gapped nodal rings (GNRs) subjected to a perpendicular magnetic field, which quantises the electronic spectrum into Landau levels (LLs). Working in the ultraquantum limit, where only the lowest LL (LLL) is partially occupied, we employ the Kubo--Bastin formalism to compute the transport coefficients for pointlike, Gaussian, and Yukawa impurity-potentials. For the Dirac case, the longitudinal conductivity is field-independent for pointlike impurities and a monotonic function of $B$ for the Gaussian and Yukawa potentials, while the Hall conductivity vanishes identically owing to the particle-hole symmetry of the two neighbouring LLs. The GNR case is qualitatively different: its non-monotonic stretched-checkmark LL spectrum causes the effective LLL to migrate to successively lower indices as the field increases, producing a pronounced oscillatory structure in both conductivities. The longitudinal response develops resonant peaks at LLL degeneracies, while the Hall conductivity traces out a sawtooth pattern with sharp zero-crossings at these same points. These results establish distinct transport fingerprints for the two systems in the extreme quantum limit, and provide a theoretical framework for interpreting magnetotransport experiments on GNRs.

• The paper uses Kubo–Bastin theory and self-consistent Born approximation to calculate disorder-broadened longitudinal and Hall conductivities for 2D Dirac cones and gapped nodal rings.
• The gapped nodal ring produces resonant longitudinal-conductivity spikes and a sign-alternating Hall sawtooth at Landau-level degeneracies, while the Dirac cone has smooth transport and zero Hall conductivity.
• The results show that impurity density and range control signal amplitude but not resonance positions, enabling high-field magnetotransport to distinguish candidate nodal-ring materials from Dirac systems.

Overview and motivation

This paper computes the dc magnetoconductivity of two two-dimensional (2D) band structures — an isotropic Dirac cone and a gapped nodal-ring (GNR) — in the ultraquantum limit, where only the lowest Landau level (LLL) is partially occupied. The work extends an existing self-consistent Born approximation (SCBA) treatment of disordered Dirac Landau levels (2607.18769) in two directions: it supplies the Hall conductivity for the Dirac cone, which was previously absent, and it provides the first treatment of disorder-broadened dc transport for the 2D GNR, whose Landau-level (LL) problem had been solved only in the clean limit. The central physical distinction is structural: the Dirac LLL is pinned at orbital index $n=0$ at all fields, whereas the GNR's non-monotonic "stretched-checkmark" LL spectrum forces the effective LLL to migrate to successively lower indices as the field increases, producing pronounced oscillatory transport that has no analogue in the Dirac case.

Model Hamiltonians and Landau spectra

Both systems are described by two-band Bloch Hamiltonians. The Dirac cone is $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$, with LL spectrum $E_n = s\,\omega_B\sqrt{n}$ ($\omega_B = v_F\sqrt{2eB}$) plus the anomalous zeroth level at $E_0=0$. The GNR is $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$, where $\Delta$ is a hybridization gap (e.g., from spin-orbit coupling). Its LL spectrum is

$$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$

with $\varepsilon_c = eB/m^*$ and $\varepsilon_r = k_0^2/2m^*$. Because $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$0 enters quadratically through $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$1, this ladder first decreases and then increases with $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$2. The LLL resides at index $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$3, which decreases stepwise as $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$4 grows. Degeneracies occur at special values of $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$5: for odd integers $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$6, levels flanking the LLL are pairwise degenerate; for even integers $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$7, the LLL itself becomes doubly degenerate (indices $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$8 and $$H_{\rm D}(\boldsymbol k) = v_F\,\boldsymbol k \cdot \boldsymbol\sigma$$9), with further pairwise degeneracy among flanking levels.

Formalism

The conductivities are computed via the Kubo–Bastin formula combined with an SCBA treatment of disorder, neglecting ladder vertex corrections. Within the LL-overlap approximation at $E_n = s\,\omega_B\sqrt{n}$0, the longitudinal and Hall conductivities reduce to sums over products of spectral functions $E_n = s\,\omega_B\sqrt{n}$1 and dispersive functions $E_n = s\,\omega_B\sqrt{n}$2 evaluated at the chemical potential, weighted by velocity matrix elements. Three impurity potentials are treated: pointlike (white-noise), Gaussian, and Yukawa (screened Coulomb). For the Gaussian potential, the momentum integrals over Girvin–Jach form factors are evaluated in closed form using Jacobi-polynomial identities; for the Yukawa potential they are reduced to one-dimensional Laguerre-polynomial integrals evaluated numerically.

A notable methodological correction is made here: the self-energy of the neighbouring level $E_n = s\,\omega_B\sqrt{n}$3 should not be solved self-consistently, since $E_n = s\,\omega_B\sqrt{n}$4 cuts the LLL rather than $E_n = s\,\omega_B\sqrt{n}$5; its linewidth is generated by scattering into the broadened LLL spectrum. The authors state explicitly that this point was missed in the earlier Dirac-cone calculation (2607.18769), which produced an incorrect neighbour self-energy; correcting it changes numerical prefactors while preserving asymptotic field scalings.

Selection rules in the quantum limit

Velocity matrix elements impose a strict constraint ("vv selection rule"): only pairs of LLs with $E_n = s\,\omega_B\sqrt{n}$6 contribute to either conductivity component. Moreover, the identity $E_n = s\,\omega_B\sqrt{n}$7 fixes the algebraic structure of $E_n = s\,\omega_B\sqrt{n}$8. Because the ultraquantum-limit truncation makes all relevant self-energies purely imaginary, the LLL's dispersive kernel drops out of the Hall response entirely, leaving a compact expression involving only $E_n = s\,\omega_B\sqrt{n}$9 weighted by $\omega_B = v_F\sqrt{2eB}$0.

The participating levels are determined by comparing three energy spacings: $\omega_B = v_F\sqrt{2eB}$1 and $\omega_B = v_F\sqrt{2eB}$2 (the gap to the hole branch). Four regimes result: (i) when $\omega_B = v_F\sqrt{2eB}$3, the negative-energy branch participates — but for non-degenerate LLLs the vv rule kills all terms, giving exactly zero conductivity; (ii)–(iv) otherwise, one or both positive-energy neighbours contribute, depending on whether $\omega_B = v_F\sqrt{2eB}$4, $\omega_B = v_F\sqrt{2eB}$5, or the neighbours are degenerate. This selection mechanism depends only on the spectrum parameters $\omega_B = v_F\sqrt{2eB}$6, $\omega_B = v_F\sqrt{2eB}$7, $\omega_B = v_F\sqrt{2eB}$8 and is completely insensitive to disorder type.

Longitudinal conductivity of the GNR

The computed $\omega_B = v_F\sqrt{2eB}$9 exhibits resonant spikes precisely at the fields where $E_0=0$0, i.e., where a neighbouring level sweeps through degeneracy with the LLL. At such points both spectral factors saturate at their Lorentzian maxima, so peak heights scale approximately as $E_0=0$1.

Pointlike disorder: the white-noise correlator yields an index-independent linewidth $E_0=0$2, and crucially $E_0=0$3 holds exactly, so peaks scale as $E_0=0$4. Quantitatively, the tallest peak falls from $E_0=0$5 at $E_0=0$6 to $E_0=0$7 at $E_0=0$8 and $E_0=0$9 at $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$0. The envelope rises monotonically with $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$1, driven by the explicit $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$2 prefactor in the velocity matrix element, only partly offset by the decreasing LLL index.

Long-ranged disorder: for Gaussian and Yukawa potentials, the linewidth decreases with increasing LL index (high-$$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$3 orbitals average over more of a smooth potential), so $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$4 grows substantially as $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$5 increases and $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$6 falls toward zero. This produces a steeper low-to-high-field rise than in the pointlike case, with weak conductivity at low $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$7. The range parameter $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$8 enhances peak heights in both models (larger $$H_{\rm GNR}(\boldsymbol k) = [(k^2-k_0^2)/2m^*]\,\sigma_x + \Delta\,\sigma_z$$9 means smoother Gaussian or more strongly screened Yukawa potentials, suppressing backscattering and narrowing $\Delta$0): e.g., Gaussian peaks grow from $\Delta$1 to $\Delta$2 as $\Delta$3 goes from $\Delta$4 to $\Delta$5 at fixed $\Delta$6. Across all three disorder types, increasing $\Delta$7 rescales the entire curve downward by roughly $\Delta$8 without shifting resonance positions, which are set purely by the level-crossing condition.

Hall conductivity of the GNR

In contrast to the positive comb of longitudinal spikes, $\Delta$9 forms a sawtooth pattern oscillating about zero, with sharp zero-crossings aligned exactly with the $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$0 degeneracy points. The mechanism differs fundamentally from the longitudinal case: away from crossings, only one neighbour contributes, and $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$1 fixes each tooth's sign; at a crossing, the two nearly degenerate neighbours enter with equal weight and opposite sign and cancel exactly. The sign-alternating capability stems from the dispersive factor $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$2 changing sign as $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$3 crosses $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$4, whereas the product of two Lorentzians in $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$5 is always positive. Tooth heights follow the same disorder trends as the longitudinal peaks (e.g., tallest tooth $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$6 across $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$7 for pointlike disorder), and zero-crossing positions are unaffected by any disorder parameter.

Contrast with the Dirac cone

For the Dirac cone, the LLL is pinned at $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$8 with $$E_n = \sqrt{\frac{\varepsilon_c^2}{4}(2n+1-\rho)^2 + \Delta^2}, \qquad \rho = \frac{2\varepsilon_r}{\varepsilon_c},$$9, and particle-hole symmetry between the two adjacent levels ($\varepsilon_c = eB/m^*$0 at $\varepsilon_c = eB/m^*$1) forces their Hall contributions to cancel identically: $\varepsilon_c = eB/m^*$2 for all three impurity models. The longitudinal resistivity is smooth and monotonic, with model-dependent scaling: field-independent for pointlike disorder; $\varepsilon_c = eB/m^*$3 for Gaussian potentials with $\varepsilon_c = eB/m^*$4 but constant-plus-linear-in-$\varepsilon_c = eB/m^*$5 (linear magnetoresistivity) for $\varepsilon_c = eB/m^*$6; and $\varepsilon_c = eB/m^*$7 for generic fixed-screening Yukawa disorder versus constant-plus-$\varepsilon_c = eB/m^*$8 when the screening length is fixed self-consistently via RPA. The paper thus establishes a clean diagnostic dichotomy: monotonic longitudinal transport with vanishing Hall response indicates a Dirac node, while oscillatory longitudinal spikes with a sign-alternating Hall sawtooth indicate a GNR. Candidate GNR materials include Kagome-honeycomb lattices, Be$\varepsilon_c = eB/m^*$9C and BeH$\varepsilon_r = k_0^2/2m^*$0 monolayers, and MX compounds (M = Pd, Pt; X = S, Se, Te).

Limitations and open questions

Several assumptions bound the results. The calculation is single-band and non-interacting; electron-electron interactions and momentum-dependent scattering rates are omitted, and the authors note it remains open how much of the oscillatory structure survives their inclusion. Ladder vertex corrections are neglected throughout, so intraband vertex renormalisation due to disorder geometry lies outside the treatment. Tilt and anisotropy of the band structure — generic in real materials and known to modify Dirac-cone response — are not considered, and their effect on the sawtooth pattern and resonance positions is unresolved. The analysis assumes well-separated LLs with $\varepsilon_r = k_0^2/2m^*$1 and weak disorder justifying SCBA; the regime of strong disorder or overlapping levels is unaddressed. Finally, extension to three-dimensional nodal-ring semimetals with toroidal Fermi surfaces is identified as a natural but unexplored direction.

Conclusion

Using the Kubo–Bastin formalism with a corrected SCBA disorder treatment, this work establishes distinct ultraquantum-limit transport fingerprints for 2D Dirac cones and gapped nodal rings. The Dirac case shows smooth, disorder-model-dependent magnetoresistance and an identically vanishing Hall conductivity enforced by particle-hole symmetry. The GNR instead displays resonant longitudinal spikes and a sign-alternating Hall sawtooth, both anchored to the field-driven migration of the effective LLL through its non-monotonic spectrum — features robust against the choice of impurity potential, whose range and density shape only the envelope and amplitude. These signatures provide a concrete theoretical framework for interpreting high-field magnetotransport experiments on candidate nodal-ring materials.