Magnetically Dead Layers in Spinel Films

Spinel films such as (Mg,Al,Fe)3O4 are has recently emerged as ideal candidates for spin-pumping materials, with damping coefficients rivaling those the of yttrium iron garnet (Y₃Fe₅O₁₂), often referred to as YIG. Magnetically dead layers are critical in determining suitability for spin-pumping applications and have long limited the implementation of candidate materials such as YIG into spintronic devices. Here, we characterize the interface quality and magnetic uniformity of a (Mg,Al,Fe)3O4 to determine if intermixing-induced magnetically dead layers are present.

We first begin with a script which models the data as a uniform magnetic layer. The data and fitting script may be downloaded here. If you run the fitting engine for  a few thousand steps (try either the "differential evolution" or Nelder-Mead Simplex" algorithms) and take a look at the spin asymmetry, you'll probably see something that looks like this:

This is not such a good fit to the data. There's a clear phase shift in the oscillations of the spin asymmetry that don't match the oscillations in the actual data - this is a classic sign that there is a dead layer in our film! So why don't we try a model with a magnetically dead layer at the film/substrate interface? The files for that can be found here. If you try this out (remember that you can look in the python scripts for lots of fitting detail and hints!), you will probably end up with a spin asymmetry plot that looks much better:

So now we have learned two things - how to fit with magnetically dead layers and how to decide when a fit doesn't describe the data very well. Take a look at the more complicated examples to learn more!