Plasmon Polaritons in Disordered Nanoparticle Assemblies

Abstract

Multilayer assemblies of metal nanoparticles can act as photonic structures, where collective plasmon resonances hybridize with cavity modes to create plasmon-polariton states. For sufficiently strong coupling, plasmon polaritons qualitatively alter the optical properties of light-matter systems, with applications ranging from sensing to solar energy. However, results from experimental studies have raised questions about the role of nanoparticle structural disorder in plasmon-polariton formation and the strength of light-matter coupling in plasmonic assemblies. Understanding how disorder affects optical properties has practical implications since methods for assembling low-defect nanoparticle superlattices are slow and scale poorly. Modeling realistic disorder requires large system sizes, which is challenging using conventional electromagnetic simulations. We employ Brownian dynamics simulations to construct large-scale nanoparticle multilayers with controlled structural order. We investigate their far- and near-field optical response using a superposition T-matrix method with two-dimensional periodic boundary conditions. We find that while structural disorder broadens the polaritonic stop band and the near-field hot-spot distribution, the polariton dispersion and coupling strength remain unaltered. To understand the effects of nanoparticle composition, we consider assemblies with Drude model particles mimicking gold or tin-doped indium oxide (ITO) nanocrystals. Assemblies of ITO nanocrystals, which have lower carrier concentrations, exhibit ultrastrong coupling, in contrast to Au nanoparticle multilayers that display deep strong coupling. Finally, we demonstrate that while computationally efficient mutual polarization method calculations employing the quasistatic approximation modestly overestimate the strength of collective plasmon coupling in these assemblies, they reproduce the polariton dispersion relations determined by electrodynamic simulations.