Magnetic and electric Purcell factor control through geometry optimization of high index dielectric nanostructures

TL;DR

This paper presents the design of silicon nanostructures optimized via evolutionary algorithms to significantly enhance magnetic and electric emission rates, with a focus on understanding the physical mechanisms involved.

Contribution

It introduces a combined approach of evolutionary optimization, modal analysis, and finite element simulations to control the magnetic and electric Purcell factors in dielectric nanostructures.

Findings

Achieved up to 2000-fold enhancement of magnetic Purcell factor in europium ions.

Demonstrated the effectiveness of evolutionary algorithms in optimizing nanostructure geometries.

Provided physical insights into the origin of Purcell factor enhancement through modal analysis.

Abstract

We design planar silicon antennas for controlling the emission rate of magnetic or electric dipolar emitters. Evolutionary algorithms coupled to the Green Dyadic Method lead to different optimized geometries which depend on the nature and orientation of the dipoles. We discuss the physical origin of the obtained configurations thanks to modal analysis but also emphasize the role of nanoscale design of the LDOS. We complete our study using finite element method and demonstrate an enhancement up to 2000 of the magnetic Purcell factor in europium ions. Our work brings together random optimizations to explore geometric parameters without constraint, a first order deterministic approach to understand the optimized designs and a modal analysis which clarifies the physical origin of the exaltation of the magnetic Purcell effect.