Long-range electron-electron interactions in graphene make its electrodynamics nonlocal

B. Rosenstein, H. C. Kao, M. Lewkowicz

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

Graphene, a one-layer-thick hexagonal array of carbon atoms, when undoped, exhibits a curious mixture of properties pertinent to either metals or insulators. On the one hand, despite near absence of both charge carriers and impurities, it has a finite conductivity like a metal. On the other hand, the Coulomb interaction between electrons is unscreened like in a dielectric and hence is long range. The chemical potential is pinned right between the conical valence and conduction bands causing quasiparticles to move like massless relativistic particles. We demonstrate at small coupling that the electrodynamics of graphene exhibits nonlocality on a macroscopic level due to the combination of the long-range interactions and the linear dispersion relation. The frequency and wave vector k-dependent conductivity tensor, in addition to a local pseudo-Ohmic part σTδij, possesses a nonlocal contribution σnlkikj/k2. While the coefficient of the local part is σT≈e2/4, the coefficient of the nonlocal part is proportional to the Coulomb interaction strength α, σnl=σTα. This leads to several remarkable effects in transport and optical response. In particular, the resistance of the graphene flake depends on the location and the geometry of the source, drain, and probes. A voltage perpendicular to the current appears in a time-reversal symmetric situation and the polarization of reflected and transmitted light is modified, without either the magnetic field (like in the Faraday effect) or anisotropy.

Original languageEnglish
Article number045137
JournalPhysical Review B - Condensed Matter and Materials Physics
Volume90
Issue number4
DOIs
StatePublished - 28 Jul 2014

Fingerprint

Dive into the research topics of 'Long-range electron-electron interactions in graphene make its electrodynamics nonlocal'. Together they form a unique fingerprint.

Cite this