>>40521593
A tiny YIG sphere might not be the best bet, and I say that because the intensity of the field has to be matched with the intrinsic frequency of the resonator. I posted the comparative table and formulae earlier. Scroll up to Table.png and Formula5.png.
>bigger resonator, lower frequency, weaker field
>smaller resonator, higher frequency, stronger field
>ideal resonator for this looks like a foamed metamaterial that acts as a very high-Q whispering gallery mode resonator with collective resonance
>bulk metal resonator still possible, but has to be fine-tuned
>the higher the cavity Q-factor, the greater the topological charge density, the better
>any damping or leakage spoils it
>even phonon scattering is a problem
>the resonator has to be nearly perfect
>ideally, the cavity wall itself would be a superconductor to avoid ohmic losses in the plasmon oscillations