Thermal control of the magnon-photon coupling in a notch filter coupled to a yttrium iron garnet/platinum system

We report thermal control of mode hybridization between the ferromagnetic resonance and a planar resonator (notch filter) working at 4.74 GHz. The chosen magnetic material is a ferrimagnetic insulator (yttrium iron garnet: YIG) covered by 6 nm of platinum (Pt). A current-induced heating method has been used in order to enhance the temperature of the YIG/Pt system. The device permits us to control the transmission spectra and the magnon-photon coupling strength at room temperature. These experimental findings reveal a potentially applicable tunable microwave filtering function.

Figure 1

(a) Experimental setup: a vector network analyzer (VNA) is connected to a 50-$\mathrm{\Omega }$ microstrip line which is capacitively coupled to the resonator. The hybrid YIG/Pt system is placed at 0.$25\lambda$. $(+)$ and $(-)$ are the electrical contacts which permit to detect (inject) a dc charge current from (at) the edges of the Pt layer. (b) Frequency dependence of $S$ parameters with and without the YIG/Pt system. (c) Magnetic field dependence of the frequency: observation of the strong coupling regime via the anticrossing fingerprint. The color map illustrates the magnitude of $S21$ from 0 to $-12$ dB. (d) Frequency dependence of $S21$ at different applied magnetic fields using the analytic solution from Ref. [9].

Figure 2

Injection: (a) frequency dependence of the $S$ parameters at the resonant condition $\left(H=H_{\mathrm{res}}\right)$ for different applied charge current. Labels 1–7 correspond to the magnitude of the applied charge current as indicated in (b). (b) Charge current dependence of the frequency measured at the resonant condition. The color map is associated to the magnitude of $S21$ from 0 to $-12$ dB. Solid cyan curves correspond to the current dependence of the hybridized frequencies based on Eq. (1), including the temperature dependence of $F_r$. Red and blue triangles indicate the position of $F_1$ and $F_2$. (c) Magnetic field dependence of the frequency: the correlation between magnetic field and charge current dependence of $F_1$ and $F_2$. Inset shows a schematic of the measurement setup. Dependence of Pt strip resistance (d), the temperature increase of the Pt strip (e), and $\mathrm{\Delta }{F}_r$ (f) with respect to the injected current $\left(I^2\right)$.

Figure 3

(a)–(c) Calculated magnetic field dependence of the microwave transmission spectrum based on Ref. [9] for different values of the damping parameter $\alpha$. (d)–(f) Measured magnetic field dependence of the frequency as function of the injected current magnitude. The color map of (a)–(f) is associated to the magnitude of the transmission parameter $S21$ from 0 to $-3$ dB. (g) Calculated dependence of $S21$ (color map from 0 to $-12$ dB) as function of the frequency and the damping parameter. Representation of $F_1,F_2$, and $F_0$. (h) Thermal annihilation of the coupling strength measured (blue stars) and extracted from (g) (dashed dotted red line).

Figure 4

Detection: simultaneous detection of (a) $S21$ and (b) the dc voltage generated by the conversion of a spin current into a charge current via inverse spin Hall effect (ISHE) at resonant condition for the positive and negative configurations of the magnetic field. (c) Measured voltage $V_{\text{ISHE}}$ as function of the frequency for different applied magnetic field.

Figure 5

Charge current dependence of the frequency measured at $H=1128$ Oe (off resonance). The color map is associated to the magnitude of $S21$ from 0 to $-12$ dB. Solid black curves correspond to the current dependence of the hybridized frequencies based on Eq. (1), including the temperature dependence of $F_r$ with $k=0.110$.