Nonlinear dynamics in magnonic Fabry-Pérot resonators: Low-power neuron-like activation and transmission suppression

Abstract

We report on nonlinear spin-wave dynamics in magnonic Fabry-Pérot resonators composed of yttrium iron garnet (YIG) films coupled to CoFeB nanostripes. Using super-Nyquist sampling magneto-optical Kerr effect microscopy and micromagnetic simulations, we observe a systematic downshift of the spin-wave transmission gaps as the excitation power increases. This nonlinear behavior occurs at low power levels, reduced by a strong spatial concentration of spin waves within the resonator. The resulting power-dependent transmission enables neuron-like activation behavior and frequency-selective nonlinear spin-wave absorption. Our results highlight magnonic Fabry-Pérot resonators as compact low-power nonlinear elements for neuromorphic magnonic computing architectures.

Figures (5)

• Figure 1: Microscopy image of the sample and the measurement geometry. The Fabry-Pérot resonator, consisting of a 30-nm-thick CoFeB nanostripe patterned on an 85-nm-thick YIG film, is positioned 25 $\upmu$m from a 1.5-$\upmu$m-wide microwave antenna. SNS-MOKE measurements were performed 2 $\upmu$m behind the resonator center, as indicated by the red circle.
• Figure 2: Measured spin-wave transmission spectra for parallel (a,b) and antiparallel (c,d) YIG-CoFeB magnetization alignments are shown for excitation powers between $-15$ and 5 dBm. In (a) and (c), the curves are vertically offset for clarity, and the dashed lines highlight the power-dependent frequency shift of the transmission gaps. The same shifts are visible as the dark contrast in the spectral maps shown in (b) and (d). The insets show the orientations of the YIG and CoFeB magnetizations, with spin waves propagating from left to right.
• Figure 3: (a) Simulated and (b) measured spin-wave transmission maps for an 850-nm-wide Fabry-Pérot resonator under excitation powers from $-15$ to $-4$ dBm. The suppression of the spin-wave signal corresponds to the $n=2$ resonance. The simulations reproduce the experimental results with exception of a minor frequency offset, likely arising from small differences between the measured system and its model. (c) Transmitted signal as a function of the excitation power, demonstrating neuron-like threshold activation behavior at $f=0.833$ GHz (simulation) and $f=0.850$ GHz (experiment). (d) Excitation-power dependence of the transmitted signal, showing nonlinear spin-wave suppression at $f=0.807$ GHz (simulation) and $f=0.805$ GHz (experiment).
• Figure 4: (a) Simulated spin-wave transmission as a function of the incident amplitude $A_0$ for waves crossing an 850-nm-wide resonator, obtained by comparing simulations performed with and without the CoFeB nanostripe. Line color corresponds to the excitation power, ranging from $-15$ to $-4$ dBm. The values of the incident amplitude corresponding to different values of the excitation power (in dBm) are depicted in the inset. (b) Simulated neuron-like threshold activation (red) and nonlinear spin-wave suppression (black) in the isolated magnonic Fabry-Pérot resonator at $f=0.838$ GHz and $f=0.812$ GHz, respectively.
• Figure 5: Profiles of the Fourier magnitudes calculated from the simulated in-plane, $|m_x|$, and out-of-plane, $|m_z|$, magnetization components are shown as a function of frequency in the left and right panels, respectively. The $|m_x|$ and $|m_z|$ values for the top surface of the YIG film in a 850-nm-wide magnonic Fabry-Pérot resonator in the parallel configuration are normalized by their corresponding values obtained for a bare YIG film. The vertical yellow dashed lines mark the edges of the resonator. The line scans show the maximum values of $|m_x|$ and $|m_z|$ within the resonator. Results for the antiparallel configuration as well as a 350-nm-wide resonator are provided in Figs. S3 to S5 of the supplementary material.