I have prepared a short summary of my thesis trying to make it understandable for everyone. For the more accurate, better written but loooonger version, click here. Or if you feel a bit lazy and would rather listen to a 9 minute podcast about my thesis, go ahead here.   

The magnetic skyrmion

Considering a bidimensional ferromagnetic background, the magnetic skyrmion is a particular spin texture. 

These spin textures have properties that make them behave like particles. They can be moved by electrical currents and act as bits of information. This is why the skyrmion is a promising candidate for information storage.

Without going into too much detail, the main drawback of the skyrmion motion is their deviation in the trajectory with respect to the current direction. There is a force that deviates the skyrmion. As a result, skyrmions can be annihilated at the edges of a nanotrack resulting in an information loss.

A solution for this is to couple two skyrmions cancelling the deviating force. For this we use synthetic antiferromagnets (SAF).

SAF skyrmion motion

The SAF skyrmions flow along the current. Moreover, faster speeds were expected compared to the standard skyrmion velocities in ferromagnets.

Challenges with SAF skyrmions: the nucleation

The standard recipe to nucleate ferromangetic skyrmions can not be used for SAF skyrmions. They are not sensitive to external magnetic fields. One of my challenges was to reliably nucleate SAF skyrmions finding new approaches.

Electrical detection of skyrmions

The electrical detection of skyrmions had mainly been performed using the anomalous Hall effect (AHE) technique. However, the detected signals due to the presence of the skyrmion are not unambiguous enough.

We proposed to integrate the skyrmion in the magnetic tunnel junction device to obtain a larger electrical signal using the tunneling magnetoresistance (TMR).

The magnetic tunnel junction (MTJ)

Magnetic tunnel junctions (MTJs) serve as the basic elements for magnetic memories. It consists of two ferromagnetic layers sandwiching a thin insulator. Electrons can tunnel from one ferromagnet into the other. The Figure a shows the MTJ device. The FeCoB ferromagnetic layers are sandwiching  the MgO insulator. 

The magnetization of the bottom FeCoB is pinned by the Co/Pt layers below. The top FeCoB layer is free to switch (up and down), changing the relative orientation with respect to the pinned layer. The orientation of both FeCoB layer has an impact on the tunneling of the electrons. When they are antiparallel (parallel), we can measure a high (low) resistance, as shown in Figure b. The relative orientation between the FeCoB layers can be changed by an external magnetic field.

A skyrmion in a MTJ

Those figures show the top view of an MTJ device (circular pillar) using a x-ray microscopy technique (STXM). The contrast is showng the Fe content of our sample. In Figure a, we see both layers with Fe content in parallel configuration, we have a dark contrast because both layers are aligned (low resistance state). When the layers are in AP orientation, we did not see any contrast because the contributions cancel out.

I don't want to go to much details to how we created the skyrmion in tha MTJ. Starting from a parallel state, we applied a voltage pulse of 10 ns and saw how the resistance changed. This resistance change was associated with the presence of the skyrmion in the free layer (Figure b). The change of resistance about 500 Ohm gives us a large readout signal compared to the AHE.

The fast current-induced motion of SAF skyrmions

We saw befre that the motion of FM skyrmions via current has drawbacks such as their trajectory deviation and the limited speeds. Using SAF skyrmions we can expect the cancelation of the trajectory with respect to the applied current direction as well as much faster speeds.

We used magnetic force microscopy monitor the motion of SAF skyrmions after current injections and measured skyrmion velocities of up to 900 m/s along the current direction.