SOT Parameter SOT_xi(or EpsilonPrime) for Pt(8nm)/Py(2nm) SHNO in Mumax3

[Parabi Datta]

Dear Mumax3 Community,

I am currently working on simulating a Spin Hall Nano-Oscillator (SHNO) in Mumax3, consisting of a Pt (8 nm) / Py (2 nm).
So far, I have been using spin-transfer torque (STT) to define the oscillations, but now I would like to implement spin–orbit torque (SOT) following the process provided in the Mumax3 tutorial.

For this, I need to specify the value of SOT_xi (or EpsilonPrime), which represents the spin Hall angle–related efficiency parameter. However, we do not know the exact value for our Pt/Py system. What would be your suggestion on which value we should use?

Thanks and Regards,
Parabi Datta

[Josh Lauzier]

Hi,

When you're using SOT, Epsilonprime has two parts. SOTxi, which relates to the relative strengths between the dampinglike and fieldlike torques, and alpha_H, which is the spin hall angle. Unfortunately, for both parameters I don't think there is a good way to guess what they should be, it is something that needs to be measured. If you cannot do so, I would suggest finding papers in the literature with similar films, and using that value. Luckily Pt/Py is popular enough that you can probably find something comparable. 

A warning: for values of alpha_H in particular, be a bit careful as reported values can vary wildly depending on the type of measurement and how exactly the film is made. I seem to recall values of around 0.07-0.15 being commonly used, but double check that. Last I had heard there is not firm agreement on the true value, there are complications due to losses at the interface.

See for instance: 
https://journals.aps.org/prb/abstract/10.1103/PhysRevB.98.024402
https://www.science.org/doi/10.1126/sciadv.aat1670

Best,
Josh L.