Plasmonic drag photocurrent in graphene at extreme nonlocality

TL;DR

This paper explores a novel mechanism called 'plasmonic drag' for photocurrent generation in graphene, driven by non-local electromagnetic field inhomogeneity, with potential for resonant enhancement when wave velocity matches electron Fermi velocity.

Contribution

It introduces a classical kinetic model of non-linear conductivity accounting for non-local effects, revealing resonant enhancement of photocurrent at specific wave velocities in graphene.

Findings

Resonant enhancement of non-linear conductivity when wave phase velocity matches Fermi velocity.

Identification of phase locking between electrons and propagating plasmons.

Upper limit of photocurrent responsivity set by a balance between resonance and electron kinetic energy.

Abstract

It is commonly assumed that photocurrent in two-dimensional systems with centrosymmetric lattice is generated at structural inhomogenities, such as p-n junctions. Here, we study an alternative mechanism of photocurrent generation associated with inhomogenity of the driving electromagnetic field, termed as 'plasmonic drag'. It is associated with direct momentum transfer from field to conduction electrons, and can be characterized by a non-local non-linear conductivity $\sigma^{(2)}({\bf q},\omega)$. By constructing a classical kinetic model of non-linear conductivity with full account of non-locality, we show that it is resonantly enhanced for wave phase velocity coinciding with electron Fermi velocity. The enhancement is interpreted as phase locking between electrons and the wave. We discuss a possible experiment where non-uniform field is created by a propagating graphene plasmon, and find an upper limit of the current responsivity vs plasmon velocity. This limit is set by a competition between resonantly growing $\sigma^{(2)}({\bf q},\omega)$ and diverging kinetic energy of electrons as the wave velocity approaches Fermi velocity.