On the Possibility of Obtaining Large Amplitude Resonance in Very Thin Ferrimagnetic Disks

Abstract

A theoretical prediction is made that the uniform resonance and the spinwaves in a very thin ferrimagnetic disk can be made nondegenerate. If this condition can be achieved experimentally, the usual second order spinwave instabilities will be forbidden and cannot limit the component of the magnetization occurring at the resonance frequency (ω0). The precession cone angle (θ) should then increase, in response to the usual rf driving field, until limited by third order instabilities involving spinwaves with frequencies equal to 32 ω0. The state of nondegeneracy is to be achieved in the following manner: A thin disk, magnetized in a direction perpendicular to the plane, is excited by two transverse positive-circularly-polarized pulsed rf magnetic fields of specified amplitude. One of these fields has a frequency equal to ω0, the other a frequency detuned from ω0 by an amount comparable to the width of resonance. As θ increases from zero during the ensuing transient, there will occur a point at which the familar second order coupled spinwaves become unstable (the state of nondegeneracy has not yet been reached). Since there is a time lag before these modes can effect the uniform precession, the cone angle will continue to increase. When some larger value of θ has been attained the degeneracy between the spin waves and the resonance is removed; the previously unstable spin waves are then decoupled from the uniform resonance. The degeneracy is removed because, as the author has already shown theoretically, the resonance frequency of the thin disk should lie below the main spinwave band (volume band) provided θ> (2Nt)½, where Nt (the transverse demagnetizing factor) is by assumption very small. This level of excitation would create the desired state of nondegeneracy, if it were not for the fact that a range of degenerate spinwaves (separate from the main band) still exists. These degenerate modes may be destroyed, however, by introducing the additional transverse rf field. Large amplitude resonance would clearly be desirable for a variety of reasons. Optical modulation experiments, to give only one example, would be of great interest. The necessity for using thin disks or films is, in this case, a fortuitous circumstance. It also follows that any nonlinear device involving harmonic generation, frequency mixing, etc. would become more efficient.