Generalizing Normal Mode Expansion of Electromagnetic Green's Tensor to Open Systems

Parry Y. Chen, David J. Bergman, Yonatan Sivan

Research output: Contribution to journalArticlepeer-review

22 Scopus citations

Abstract

We generalize normal mode expansion of Green's tensor G(r,r′) to resonators in open systems, resolving a long-standing open challenge. We obtain a simple yet robust formulation whereby radiation of energy to infinity is captured by a complete, discrete set of modes rather than a continuum. This enables rapid simulations by providing the spatial variation of G(r,r′) over both r and r′ in one simulation. Systems with or without material losses can be treated. Few eigenmodes are often necessary for nanostructures, facilitating both analytic calculations and unified insight into computationally intensive phenomena such as Purcell enhancement, radiative heat transfer, van der Waals forces, and Förster resonance energy transfer. We bypass all implementation and completeness issues associated with the alternative quasinormal eigenmode methods by defining modes with permittivity rather than frequency as the eigenvalue. We obtain true stationary modes that decay rather than diverge at infinity, and are trivially normalized. Completeness is achieved both for sources located within the inclusion and the background through use of the Lippmann-Schwinger equation. Modes are defined by a linear eigenvalue problem, readily implemented with any numerical method. We demonstrate its simple implementation with comsol multiphysics using the default inbuilt tools. The results are validated against direct scattering simulations, including analytic Mie theory, attaining arbitrarily accurate agreement regardless of source location or detuning from resonance.

Original languageEnglish
Article number044018
JournalPhysical Review Applied
Volume11
Issue number4
DOIs
StatePublished - 5 Apr 2019

Fingerprint

Dive into the research topics of 'Generalizing Normal Mode Expansion of Electromagnetic Green's Tensor to Open Systems'. Together they form a unique fingerprint.

Cite this