Quantum-informed plasmonics for strong coupling: the role of electron spill-out

TL;DR

This study investigates how nonlocal effects like electron spill-out influence plasmon-exciton coupling in nanoscale systems, finding that local models remain accurate even when nonlocal phenomena are expected to be significant.

Contribution

The paper demonstrates that incorporating surface-response corrections into classical models shows minimal impact on strong coupling, validating local response approximations in quantum plasmonics.

Findings

Nonlocal effects have minimal influence on mode splitting in strong coupling.

Surface-response corrections can be incorporated into classical Mie theory.

Local response approximation remains valid despite spill-out effects.

Abstract

The effect of nonlocality on the optical response of metals lies at the forefront of research in nanoscale physics and, in particular, quantum plasmonics. In alkali metals, nonlocality manifests predominantly as electron density spill-out at the metal boundary, and as surface-enabled Landau damping. For an accurate description of plasmonic modes, these effects need be taken into account in the theoretical modelling of the material. The resulting modal frequency shifts and broadening become particularly relevant when dealing with the strong interaction between plasmons and excitons, where hybrid modes emerge and the way they are affected can reflect modifications of the coupling strength. Both nonlocal phenomena can be incorporated in the classical local theory by applying a surface-response formalism embodied by the Feibelman parameters. Here, we implement surface-response corrections in Mie theory to study the optical response of spherical plasmonic--excitonic composites in core--shell configurations. We investigate sodium, a jellium metal dominated by spill-out, for which it has been anticipated that nonlocal corrections should lead to an observable change in the coupling strength, appearing as a modification of the width of the mode splitting. We show that, contrary to expectations, the influence of nonlocality on the anticrossing is minimal, thus validating the accuracy of the local response approximation in strong-coupling photonics.