Inverse design of magnonic filter

Monochromatic spin wave (SW) is the substrates for the realizing the practical function of magnonic devices, however the generating perfect monochromatic SW is still a unsolved but fundamental problems. In our latest work (Xing et al., 2022), the resonant tunneling effect was found in a magnonic crystals (MCs), which demonstrates a similar effect of Fabry–Pérot interferometer, and shows a promising application of band-pass magnonic filter (MF) for generating perfect monochromatic SW. The optimization of such MF is a complex problem, due to that the natural nonlinearity of spin-wave scattering behavior enhances sensitivity to device parameters. In this work, the inverse design (ID) approach is developed for designing the magnonic filters (MFs). To determine the optimizable variables, the fundamental properties of antiferromagnetically coupled heterojunction based MCs were theoretically investigated. The particle swarm algorithms (PSA) is used for automatically finding MFs which meet the inputting requirements. Here, the input of algorithms is a vector of SW frequency and output is a vector of parameters of MFs. Besides, the PSA is further optimized to be self-adapting to required accuracies, showing a higher computing efficiency than the convectional one. Finally, the output data of predicted MFs with various filtering characteristics are briefly summarized in a table as an alternative proposal for the future experimental design. These encouraging results show high potential of using ID approach in magnetism and boosting optimization of magnetic multilayer heterojunction devices, such as magnon transistor and magnon space–time crystals.