Extracting the oscillatory RKKY-coupling

The strength of the RKKY-coupling determines the required saturation field that is needed to fully align the magnetic moments of the two Co layers. As shown in section 2.5, in the absence of anisotropy the saturation field and the RKKY-coupling strength are directly related via

J_RKKY = −µ₀H_sM t_F 2

for which tF = 3 nm and M = 1055 kA/m (as experimentally obtained). In order to investigate the oscillatory behavior of the RKKY coupling, the saturation fields for different spacer layer thicknesses are calculated. The saturation field values are

calculated by intersecting a lineair fit with the saturation magnetization, as is shown in figure 4.3. All samples containing a wedged spacer layer are analysed using the MOKE setup and as will be shown in section 4.2 it was not always possible to reach saturation fields with this setup. In order to extract the saturation field values for these measurements, the saturation magnetization is calculated when the Co layers are coupled ferromagnetically. For measurements where saturation is not reached, the lineair fit is extrapolated and intersected with this obtained saturation magnetization in order to calculate the saturation field values.

A plot of the saturation field and the RKKY-coupling strength is shown in figure 4.4. The magnetization is measured locally using the MOKE setup, and a scan of the sample is done as explained in section 3.2.1. At spacer layer thicknesses between 0.13 and 1.46 nm, the saturation field value increases and the magnetic layers are in an antiferromagnetically coupled state. The peak in AF coupling is found at tRu = 0.8 nm and the width of the AF region is 1.33 nm. Both values are significantly higher than the values Parkin published in his article^[25]. Parkin mentions that the first AF peak is found at tRu= 0.3nm with a width of the AF region of 0.3 nm. Furthermore, Parkin found more than one AF-peak present for a Co/Ru/Co SAF at tRu < 4nm.

This is not observed in our study. A maximum saturation field of 236 mT is found, corresponding to JRKKY = 0.37 ∗ 10⁻³ J/m² which is significantly lower than the J_RKKY = 5 ∗ 10⁻³ J/m² as reported by Parkin.

In his article, Parkin shows that besides Ru, other spacer materials such as Cr and Ir can result in RKKY-coupling as well^[25][26]. In order to investigate this, mea-surements are done on magnetic multilayers using Cr and Ir as spacer layers. The Cr layer does not show any AF coupling. A plot of the saturation field versus the thick-ness for the Ir sample is shown in figure 4.5. An AF coupling is found at tIr = 0.9 nm and the width of the AF region is 0.57 nm with JRKKY = 0.049 ∗ 10⁻³ J/m². A second AF peak is found for thicknesses between 2.34 and 3.09 nm with a maximum coupling strength JRKKY = 0.006 ∗ 10⁻³ J/m². Table 4.1 shows a comparison of important parameters as experimentally obtained and as reported by Parkin^[25] .

Parkin Experiment

Spacer material Ru Ir Cr Ru Ir Cr

t_{AF −peak} (10⁻⁹ m) 0.3 0.4 0.7 0.8 0.9

-∆t_{AF −region} (10⁻⁹ m) 0.3 0.3 0.7 1.33 0.57

-No. AF regions 3+ 3+ 3+ 1 2 0

J_RKKY (10⁻³ J/m²) 5 1.85 0.24 0.37 0.049

-Table 4.1: Overview of the different parameter values as reported by Parkin^[25] and experimentally obtained.

Figure 4.4: Plot of the saturation field and the RKKY-coupling strength versus the Ru spacer layer thickness for a Ta(2)/Co(3)/Ru(wedged 0-4 nm)/Co(3)/Ta(5) sample. An AF coupling state is found for spacer layer thickness between 0.13 and 1.46 nm. The samples are grown under standard deposition conditions and analysed using the MOKE setup at room temperature.

Figure 4.5: Saturation field and RKKY-coupling strength versus spacer layer thick-ness for a Ta(2)/Co(3)//Ir(wedged 0-4)/Co(3)/Ta(5) sample. An AF coupling state is found for spacer layer thicknesses between 0.63 and 1.20 nm and 2.34 and 3.09 nm. The Ta and Co layers are grown at standard conditions and the Ir wedge is deposited at 7 sccm 20 W. The sample is measured at room temperature using the MOKE setup.

The experimentally obtained values do not agree with the values Parkin pub-lished. The coupling strengths are much weaker, and for Ru and Ir only 1 and 2 AF regions are found respectively. The AF peaks are found at larger values and the AF regions are much wider. An explanation could be that the quality of the samples is much worse than the quality of the samples Parkin used. When intermixing of the atomic layers occurs at the Co/Ru, Co/Ir and Co/Cr interfaces, the interfaces are not well-defined, leading to a decrease of JRKKY as explained in section 2.4.5. The RKKY-coupling decreases rapidly for increasing spacer-layer thickness^[25]. When the RKKY-coupling is already weak due to intermixing of the atomic layers at the inter-faces, the oscillatory RKKY-coupling will damp out rapidly and less oscillations can be observed. Furthermore, When the Co and Ru layers are intermixed and roughness is present at the interface, the spacer layer thickness is not well defined. This leads to a wider region where AF-coupling is found. Finally, The Co/Ru layer Parkin used consists of a hexagonal close packed crystal structure, whereas samples investigated in this thesis are polycrystalline or with a fcc-crystalline structure of the Co layer.

This results in a more random growth of the layers, creating domains and grains which result in a lower quality of the layers.