The boundary-value problems involving a usual spherical ferromagnetic sample and commonly used cavity geometries are virtually impossible to solve so that perturbation theory must be used to interpret the results of resonance measurements. However, we have succeeded in solving the problem of a spherical sample in a spherical cavity under the assumption of uniform and circularly polarized magnetic field within the sample. Using the results of this calculation, it is possible to measure all components of the intrinsic susceptibility tensor of the medium. Furthermore, they enable the determination of the limits of validity of perturbation theory with regard to sample size, dielectric constant, and susceptibility. The calculation is self-consistent in that uniform and circularly polarized magnetic field within the sample is assumed and verified for a certain range of parameters by results of the calculation itself. Dimensional resonance, characterized by cavity-like resonances within the sample, is found to occur if the sample has a high susceptibility. In the case of ferromagnetic materials with small linewidths, this dimensional-resonance effect can be sufficiently serious as to render the usual interpretation via perturbation theory completely invalid. A measurement technique for determining the magnetic properties of a sample by use of our theory is described. It has been found that use of our solution makes it possible to relax the restriction of very small sample size as implied by perturbation theory. The use of a larger sample is highly desirable, since the change in frequency and Q would be greater, and less refined instruments would be required for making the measurements.
Read full abstract