Frequency Filtering with a Magnonic Crystal Based on Nanometer-Thick Yttrium Iron Garnet Films

Hugo Merbouche, Martin Collet, Michael Evelt, Vladislav E. Demidov, José Luis Prieto, Manuel Muñoz, Jamal Ben Youssef, Grégoire de Loubens, Olivier Klein, Stéphane Xavier, Olivier D’Allivy Kelly, Paolo Bortolotti, Vincent Cros, Abdelmadjid Anane and Sergej O. Demokritov
ACS Appl. Nano Mater. 2021, 4, 1, 121–128

Magnonics rely on the wave nature of the magnetic excitations to process information, an approach that is common to many fields such as photonics, phononics, and plasmonics. Nevertheless, magnons, the quanta of spin-wave excitations, have the unique advantage to be at frequencies that are lying between a few GHz to tens of GHz, that is, in the technologically relevant radio-frequency bands for 4G and 5G telecommunications. Furthermore, their typical wavelengths are compatible with on-chip integration. Here, we demonstrate radio-frequency signal filtering by a micron-scale magnonic crystal (MC) based on a nanopatterned 20 nm-thick film of yttrium iron garnet with a minimum feature size of 100 nm where the Bragg vector is set to be kB = 2.1 μm–1. We map the intensity and the phase of spin waves (SWs) propagating in the periodic magnetic structure using phase-resolved microfocus Brillouin light-scattering spectroscopy. Based on these maps, we obtain the SW dispersion and the attenuation characteristics. Efficient filtering is obtained with a frequency selectivity of 20 MHz at an operating frequency of 4.9 GHz. The results are analyzed by performing time- and frequency-resolved full-scale micromagnetic simulations of the MC that reproduce quantitatively the complexity of the harmonic response across the magnonic band gap and allow the identification of the relevant SW-quantized modes, thereby providing an in-depth insight into the physics of SW propagation in periodically modulated nanoscale structures.


This work was supported in part by the Deutsche Forschungsgemeinschaft, and the ANR Maestro (ANR-18- CE24-0021). M.M acknowledges the financial support from the project MAT2017-87072-C4-4-P. M.C. acknowledges DGA for financial support. We acknowledge C. Carrétéro, E. Jacquet, and R. Lebourgeois for their contribution to sample preparation.