Andreev Reflection at a Graphene Normal-Superconductor Interface: A Lattice-Continuum Comparison
Abstract
We compare quadratic continuum and nearest-neighbour tight-binding models of Andreev reflection at a clean graphene normal-superconductor interface. For abrupt changes in chemical potential and pairing potential, with unchanged hopping, the lattice model yields four independent boundary conditions. Its reflection probabilities agree with independent Kwant calculations to numerical precision for the channels examined. We investigate trigonal warping and its effect on the relation between momentum and group velocity. For (μN,μS,Δ0,E)/t0=(0.5,1,0.1,0.05), the Andreev-reflection probability for the ξ=+1 valley peaks at 19.9∘ in the lattice model, versus 3.4∘ in the quadratic continuum model. These parameters lie outside the expected validity regime of the low-energy approximation. As all energy scales are reduced at fixed ratios, the central angular response and conductance become more consistent with lattice results. The quadratic continuum boundary problem also requires short-wavelength auxiliary modes outside the validity range of the valley expansion. Thus, differences cannot be attributed to bulk dispersion alone. A small excitation energy relative to chemical potential does not ensure continuum validity; relevant energy scales, modes in both regions, and interface matching assumptions must also be considered.
Related
- Equal-Spin local and nonlocal Andreev reflection in altermagnet/superconductor/altermagnet junction
- Fermiology and the Candidate Chiral Superconductor in Rhombohedral Tetralayer Graphene
- Topological and kinetic origins of fractional thermal conductance at the topological insulator--superconductor interface
- Effects of interface regularity on the bulk-edge correspondence in continuum photonic systems
- Intercalation of Ultrathin Ferromagnetic Iron Films at the Graphene/SiC Interface
BibTeX
@misc{haq2026andreev,
title = {{Andreev Reflection at a Graphene Normal-Superconductor Interface: A Lattice-Continuum Comparison}},
author = {Dar Faheem Ul Haq and S. S. Z. Ashraf},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.34270},
eprint = {2610.34270},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.34270},
}
RIS
TY - UNPB
TI - Andreev Reflection at a Graphene Normal-Superconductor Interface: A Lattice-Continuum Comparison
AU - Dar Faheem Ul Haq
AU - S. S. Z. Ashraf
PY - 2026
DA - 2026/10/25
PB - arXiv
DO - 10.48550/arxiv.2610.34270
UR - https://arxiv.org/abs/2610.34270
AB - We compare quadratic continuum and nearest-neighbour tight-binding models of Andreev reflection at a clean graphene normal-superconductor interface. For abrupt changes in chemical potential and pairing potential, with unchanged hopping, the lattice model yields four independent boundary conditions. Its reflection probabilities agree with independent Kwant calculations to numerical precision for the channels examined. We investigate trigonal warping and its effect on the relation between momentum and group velocity. For (μ_N,μ_S,Δ_0,E)/t_0=(0.5,1,0.1,0.05), the Andreev-reflection probability for the ξ=+1 valley peaks at 19.9^∘ in the lattice model, versus 3.4^∘ in the quadratic continuum model. These parameters lie outside the expected validity regime of the low-energy approximation. As all energy scales are reduced at fixed ratios, the central angular response and conductance become more consistent with lattice results. The quadratic continuum boundary problem also requires short-wavelength auxiliary modes outside the validity range of the valley expansion. Thus, differences cannot be attributed to bulk dispersion alone. A small excitation energy relative to chemical potential does not ensure continuum validity; relevant energy scales, modes in both regions, and interface matching assumptions must also be considered.
ER -