Update FixedLayer direction based on selected region for STT simulation

[Gary Young]

Dear all,

I am trying to simulate STT switching in a system with two ferromagnetic layers. Specifically, I would like the spin polarization direction (i.e. the FixedLayer) acting on the second layer (region 2) to be dynamically determined by the magnetization of the first layer (region 1).

Since the magnetization of the first layer is itself evolving during the simulation due to Zhang-Li STT, I need the FixedLayer of the second layer to continuously update according to the instantaneous magnetization of the bottom layer by SlonczewskiTorque.

Is there any way in MuMax3 to dynamically use the magnetization of one region (e.g. region 1) as the spin polarization vector for another region (e.g. region 2)?

Thank you very much for your time and advice.

Best regards,

Gary

[Josh Lauzier]

Hi,

I don't know if there's a straightforward way to do this using the built in FixedLayer. I think you'd end up having to update it each step in a for-loop. 

A more straightforward way is probably to use a custom field term to handle the Slonczewski term, and you can use the magnetization of the first layer as the effective "fixed layer" term in Slonczewski STT equation. The mumax tutorial has some examples in session 4 on how to do custom fields. It's not quite the same as usecase as what you're trying to do, but close enough it is probably a useful guide. There, they implement SOT, but you can do something similar for the normal Slonczewski STT, just following the equation for the STT instead. (If you do this, make sure to disable the built-in STT with DisableSlonczewskiTorque=true)

Best,
Josh L.

[Gary Young]

Hi Josh,

Many thanks for the pointers—super helpful! I’ll go the custom-field route for the Slonczewski term. I’ll keep Zhang–Li STT enabled (built-in) and disable the built-in Slonczewski torque while adding my own custom field for it. This matches my use-case (helical order in the free layer), so I’ll implement the Slonczewski term via AddFieldTerm(...) and mask it to the free layer, using the bottom layer’s local m as p.

Please correct me if I’m wrong, but in my understanding both Zhang–Li and Slonczewski STT can coexist in a device, as long as the fixed layer acts as a polarizer and the free layer has a non-uniform magnetization.

Best regards,
Gary

[Josh Lauzier]

Hi,

As far as I'm aware, you can have both (in particular, you can easily have both if your electrical current happens to be in the z direction). They correspond to different physical mechanisms, so there is no reason you can't have both. The difficult part is getting a set up where they're both meaningful, since they tend to want different things. There are a few discussions here from people doing both: [1] and [2]

Oh, one other note that I should have mentioned previously: the built-in Slonczewski only cares about the current in the z-direction. This is a hard limitation, you can't do anything about it (other than change the Fixedlayer from top or bottom). This shouldn't matter for the Custom Field approach though

Best,
Josh L.

[Gary Young]

Hi Josh,

Thanks a lot for clarifying — that makes perfect sense now. Really appreciate you pointing out the limitation with the built-in Slonczewski term and the note about current direction. The references you shared are also very helpful.

Best,
Chenghao

Josh Lauzier <joshl...@gmail.com> 于2025年9月6日周六 07:59写道：

--
You received this message because you are subscribed to a topic in the Google Groups "mumax2" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mumax2/CJr6bKUx25Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mumax2+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/mumax2/5f8d94d4-0dbf-4d14-8268-59446222dacdn%40googlegroups.com.