Automated design of nonreciprocal thermal emitters via Bayesian optimization
This addresses the problem of inefficient design for thermal and energy applications, offering an automated method that is incremental in improving upon existing intuitive approaches.
The study tackled the design of nonreciprocal thermal emitters by developing a Bayesian optimization approach to maximize broadband nonreciprocity, resulting in structures that achieve emission from 5 to 40 micrometers with fewer layers, outperforming current intuitive designs.
Nonreciprocal thermal emitters that break Kirchhoff's law of thermal radiation promise exciting applications for thermal and energy applications. The design of the bandwidth and angular range of the nonreciprocal effect, which directly affects the performance of nonreciprocal emitters, typically relies on physical intuition. In this study, we present a general numerical approach to maximize the nonreciprocal effect. We choose doped magneto-optic materials and magnetic Weyl semimetal materials as model materials and focus on pattern-free multilayer structures. The optimization randomly starts from a less effective structure and incrementally improves the broadband nonreciprocity through the combination of Bayesian optimization and reparameterization. Optimization results show that the proposed approach can discover structures that can achieve broadband nonreciprocal emission at wavelengths from 5 to 40 micrometers using only a fewer layers, significantly outperforming current state-of-the-art designs based on intuition in terms of both performance and simplicity.