Phase-Modulated Cavity Magnon Polaritons as a Precise Magnetic Field Probe

We describe and operate a spin-magnetometer based on the phase modulation of cavity magnon polaritons. In this scheme a rf magnetic field is detected through the sidebands it induces on a pump, and the experimental configuration allows for a negligible pump noise and a high-frequency readout. The demonstrator setup, based on a copper cavity coupled to an yttrium iron garnet sphere hybrid system, reached a sensitivity of $2.0\mathrm{pT}/\sqrt{\mathrm{Hz}}$ at 220 MHz, evading the pump noise and matching the theoretical predictions. An optimized setup can attain a rf magnetic field sensitivity of about $8\mathrm{fT}/\sqrt{\mathrm{Hz}}$ at room temperature. An orders of magnitude improvement is expected at lower temperatures, making this instrument one of the few magnetometers accessing the subfemtotesla limit.

Figure 1

(a) Scheme of the spin-magnetometer working principle. The gray line represents the transmission spectrum of the PMHS, the dark red line is the pump with the corresponding sideband, the magnetic field effect and its frequency are reported in green. (b) Electronics of an example device. The pump line is filtered with a waveguide with cutoff around ${\omega }_0$ before exciting the ${\omega }_2$ resonance. An antenna (full dot) is used to inject the tone and extract the sideband. The output power is filtered and amplified before the readout.

Figure 2

Scheme of the magnetometer demonstrator. The calibration signal $b_1$ (dark red) is provided by a generator connected to a loop on the holed side of the cavity, and is monitored with an oscilloscope. The cavity (orange) houses a YIG sphere (black), which is biased with the static field $B_0$ supplied by an electromagnet represented with its north (N, blue) and south (S, red) poles. The rf cavity field is extracted by an antenna with variable coupling (gray), and passes through a circulator before being amplified by Amp.1 and Amp.2, while the filter cavity avoids the saturation of the second amplifier. The microwave pumping line delivers to the cavity a monochromatic tone filtered with a waveguide.

Figure 3

The $S_{21}$ spectrum of the system presented in the inset. The two absorption profiles are at the frequencies of the two hybrid modes ${\omega }_1$ and ${\omega }_2$. The waveguide cutoff is roughly at the frequency ${\omega }_0$. The setup in the inset has the same color legend as described in Fig. 2.

Figure 4

(a) Signal spectra obtained with different test loop fields. The $x$ axis is the frequency offset from ${\omega }_1/2\pi$, while the $y$ axis is the voltage read at the spectrum analyser. The top-left spectrum corresponds to a $\mathrm{SNR}=1$ field with a resolution bandwidth of 100 Hz, i.e., an integration time of 10 ms. (b) Test measurement of the spin magnetometer, showing the voltage at ${\omega }_1$ versus the calculated input field generated with the loop. The noise level is subtracted from the data points, which are then fitted to a linear function, giving a slope $m_{\mathrm{fit}}=(1.1\pm 0.3)\times {10}^5\mathrm{V/T}$ and an intercept $q_{\mathrm{fit}}=(-1.9\pm 0.7)\mu \mathrm{V}$.