Off-resonant manipulation of spins in diamond via precessing magnetization of a proximal ferromagnet

We report the manipulation of nitrogen vacancy (NV) spins in diamond when a nearby ferrimagnetic insulator, yttrium iron garnet, is driven into precession. The change in NV spin polarization, as measured by changes in photoluminescence, is comparable in magnitude to that from conventional optically detected magnetic resonance, but relies on a distinct mechanism as it occurs at a microwave frequency far away from the magnetic resonance frequency of the NV spin. This observation presents an approach to transferring ferromagnetic spin information into a paramagnet and then transducing the response into a robust optical signal. It also opens strategies for studying ferromagnetism and spin transport at the nanoscale.

Figure 1

Experimental schematic: The sample is a 20-nm-thick single-crystal YIG film with nanodiamonds dispersed on top with a thickness of about 500 nm. To apply microwave fields to the sample, a silver microwire is patterned on the YIG. Green laser light is focused onto nanodiamonds near the wire, and the intensity of the resulting photoluminescence from the NV centers is measured. The inset is an SEM image of dispersed nanodiamonds.

Figure 2

NV PL data while YIG undergoes FMR: (a) Raw data showing the change in the intensity of the PL from the NV centers in nanodiamonds dispersed on top of GGG, with no YIG, as a function of microwave frequency and magnetic field. (b) The same data overlaid with theoretical resonance conditions for the NV centers. The black dashed lines show the resonance condition for an NV center with the magnetic field parallel to the NV axis for the ground and excited states. The gray crosses show the resonance condition for an NV center with the magnetic field perpendicular to the NV axis. (c) Similar data from nanodiamonds dispersed on the YIG sample with the distinct feature corresponding to the YIG FMR condition. (d) The data from (c) with the FMR peaks measured using reflected microwave power (blue dots). Also shown is a fit (blue line) to the calculated dispersion relation for YIG-film FMR with the magnetic field in plane (see main text).

Figure 3

Power dependence of the FMR-induced feature: (a) Frequency sweeps at a magnetic field of 2.36 mT for various microwave (top panel) and laser (bottom panel) powers. (b) The area under the FMR peak [identified in (a), between the dashed lines] is shown as a function of the microwave (red dots, bottom axis) and laser (blue crosses, top axis) powers. A linear guide to the eye is provided which indicates that the FMR-induced feature has a fairly linear response to both laser and microwave power in our measurement range.

Figure 4

NV-center PL data on YIG and on wire above YIG: (a) Data taken with the laser spot focused on top of the $\sim$300-nm-thick patterned microwire. The data shows that the signal corresponding to the YIG FMR is still present, though reduced. The coupling must extend at least 300 nm through the wire. (b) Data from the nanodiamonds directly on top of the YIG for comparison. (c) Line cuts of the data in (a) and (b) at 16.5 mT. The FMR-induced peak is reduced by three times relative to the NV excited-state peak for the case of nanodiamonds on top of the microstrip, compared to the case directly on YIG.