Plasmon-plasmon interaction and the role of buffer in epitaxial graphene micro-flakes

TL;DR

This study explores the origin of plasmon behavior in epitaxial graphene micro-flakes, revealing how buffer layers and plasmon interactions influence the optical properties and resonance shifts, supported by experimental and theoretical analysis.

Contribution

It introduces a comprehensive model explaining plasmon resonance shifts in epitaxial graphene, accounting for buffer effects and plasmon-plasmon interactions, supported by experimental data.

Findings

Plasmon absorption is independent of light polarization relative to terraces.

Large redshift in plasmon resonance explained by plasmon-plasmon interaction and buffer effects.

Buffer layers influence Coulomb screening and plasmon behavior in graphene.

Abstract

We investigate the origin of the translational symmetry breaking in epitaxially grown single-layer graphene. Despite the surface morphology of homogeneous graphene films influenced by the presence of mutually parallel SiC surface terraces, the far-infrared magneto-plasmon absorption is almost independent of the angle between the probing light polarization and the orientation of terraces. Based on a detailed analysis of the plasmon absorption lineshape and its behavior in the magnetic field, supported by confocal Raman mapping and atomic force microscopy, we explain this discrepancy by spontaneously formed graphene micro flakes. We further support our conclusions using data collected on artificially created graphene nanoribbons: we recognize similar plasmon origin in artificial ribbons and naturally formed grains. An unexpectedly large plasmon resonance redshift was observed in nanoribbons. In a hydrogen-intercalated sample (which does not contain the buffer), this redshift is quantitatively taken into account by a plasmon-plasmon interaction. In non-intercalated samples featuring a buffer layer, this redshift is due to an interplay between the plasmon-plasmon coupling and Coulomb screening by the buffer-induced interface states. This model determines the density of interface states in good agreement with experimentally reported values.