Inverse design of radiative thermal meta-sources via discrete dipole approximation model

Abstract

The radiation from conventional thermal infrared sources is incoherent and unpolarized. It has been recently shown that patterning plasmonic materials into metasurfaces can enhance the coherence of thermal radiation through thermal excitation of localized resonant modes in meta-atoms, enabling realization of thermal emitting metasurfaces - metasources - with high directivity and spectral selectivity. Toward achieving flexible modulation of radiative heat using metasurfaces and realization of complex functionalities, the design rules and modeling tools should be modified to take into account the incoherent nature of thermal radiation. In this paper, we utilize a robust formulation based on discrete dipole approximation within the framework of fluctuation dissipation theorem, which can efficiently characterize the thermal emission from large-area finite thermal metasources consisting of nanostructured materials. We employ an evolutionary lattice approach for the inverse design of radiative thermal metasources through topology optimization without relying on the phase analysis based on the assumption of coherency. To this end, the interaction matrix of dipolar equations is initialized for a lattice of subwavelength blocks and the arrangement of the blocks is subsequently optimized via a genetic algorithm toward achieving a certain spatial emission pattern with minimal computational cost. This approach enables fast and efficient design of large-scale thermal metasources with arbitrary functionalities. In particular, we investigate steering, focusing, and arbitrary shaping of partially coherent thermal radiation via finite silicon carbide gratings supporting surface phonon polaritons. Moreover, we demonstrate that the use of electro-optical materials such as graphene enables active tuning of thermal emission and realization of geometrically fixed multifunctional thermal metasources.