Annealing induced spin reorientation transition (SRT)

From the above discussions, we understand that the phase transition, film thickness, and the ratio of Ni% to Co% in the subsurface and bulk regions of Ni-Co-Pt alloy affect the depression of Curie temperature. In the calculation of Pt concentration for 1 ML Ni/1 ML Co/Pt(111) and 1 ML Co/1 ML Ni/Pt(111) The question arising now is what difference does it make if the surface composition becomes less homogeneous? In order to investigate this problem, we prepared two mirror systems of 2 ML Co/2 ML Ni/Pt(111) and 2 ML Ni/2 ML Co/Pt(111). The ferromagnetic films of theses samples are thicker than those of 1 ML Co/1 ML Ni/Pt(111) and 1 ML Ni/1 ML Co/Pt(111), thus the difference in the ratio of Ni% to Co% between the subsurface region and bulk region becomes more obvious than those of 1 ML Co/1 ML Ni/Pt(111) and 1 ML Ni/1 ML Co/Pt(111). This helps us to investigate the influence of the ratio of Ni% to Co% of the subsurface region.

Fig. 6.13(a) shows the polar and longitudinal hysteresis loops of 2 ML Ni/2 ML Co/Pt(111) after annealing at each T_a for 20 minutes. All the loops were measured at room temperature. Only polar Kerr signal can be observed unless annealing the sam-ple at a temperature higher than 675 K. When the annealing temperature is 675 K, both the polar and longitudinal Kerr signals can be observed, after which only the longitudinal can be detected. The easy axis of the magnetization is initially in the out-of-plane direc-tion, it changes from the out-of plane to the in-plane after annealing at a temperature higher than 675 K. That means a spin reorientation transition occurred by the annealing.

The annealing temperature dependence of Kerr intensities is shown in Fig. 6.13(b). The polar Kerr intensity increases a little when the annealing temperature increases from 300 K to 525 K, after which it starts to decrease and approaches a value of zero when the

Figure 6.13: (a) Polar and longitudinal hysteresis loops of 2 ML Ni/2 ML Co/Pt(111).

Each curve was measured at room temperature after annealing at T_a for 20 minutes. (b) The annealing temperature dependences of polar and longitudinal Kerr intensities. (c) The evolutions of Ni 848 eV, Co 656 eV, and Pt 237 eV AES signals versus annealing temperature.

annealing temperature is higher than 675 K. The longitudinal Kerr intensity stays zero when 300 K ≤ Ta ≤ 600 K, after which it starts to increase except for a decrease at a temperature higher than 850 K. The Auger intensities of Pt 237 eV, Ni 848 eV and Co 656 eV were measured to monitor the diffusion process of 2 ML Ni/2 ML Co/Pt(111) and the result is shown in Fig. 6.13(c). The Ni 848 eV AES intensity starts to decrease at 525 K while the Co 656 eV AES intensity starts to increase at the same annealing temperature.

No significant increase of Pt 237 eV AES intensity is observed unless T_a≥ 800 K in which the Co 656 eV AES intensity starts to depress. From the AES studies in sec. 5.2, we know

that the intensity of Pt 237 eV will not change when the intermixing process of Ni and Co atoms occurred. Therefore, the Ni and Co atoms of 2 ML Ni/2 ML Co/Pt(111) inter-mix to each other when 300 K≤ Ta ≤ 525 K. Comparing Fig. 6.13(b) with (c), we know that the Ni-Co intermixing process decreases the perpendicular magnetic anisotropy when 525 K≤ Ta≤ 750 K. The Ni-Co intermixing layer starts to diffuse into the Pt substrate when T_a ≥ 800 K. The interaction between neighboring Co and Pt decreases and the interactions between neighboring Ni and Pt or Ni and Co increase when the intermixing process occurred. Hence, the perpendicular magnetic anisotropy deceases.

Figure 6.14: (a) Polar and longitudinal hysteresis loops of 2 ML Co/2 ML Ni/Pt(111).

Each curve was measured at room temperature after annealing at T_a for 20 minutes. (b) The annealing temperature dependences of polar and longitudinal Kerr intensities. (c) The evolutions of Ni 848 eV, Co 656 eV, and Pt 237 eV AES signals versus annealing temperature.

Fig. 6.14(a) shows the polar and longitudinal hysteresis loops of 2 ML Co/2 ML Ni/Pt(111) after annealing at each T_a for 20 minutes. All the loops were measured at room temperature. Only longitudinal Kerr signal can be observed unless annealing the sample at a temperature higher than 850 K. When the annealing temperature is higher than 850 K, only polar Kerr signal can be observed. The easy axis of the magnetization is initially in the in-plane direction, it changes from the in-plane to the out-of plane after annealing at a temperature higher than 850 K. That means, as 2 ML Ni/2 ML Co/Pt(111), a spin reorientation transition occurred by the annealing. The annealing temperature dependence of Kerr intensities is shown in Fig. 6.14(b). When the annealing temperature is between 300 K and 600 K, the longitudinal Kerr intensity does not change and so does the polar Kerr intensity. The longitudinal Kerr intensity increase with T_a when 600 K≤ Ta ≤ 750 K, after which it starts to decrease and reach a value of zero at 850 K. It is interesting that the polar Kerr intensity appears abruptly when the annealing temperature is 850 K. The annealing temperature dependence of Pt 237 eV, Ni 848 eV and Co 656 eV AES signals are shown in Fig. 6.14(c). The Co 656 eV intensity starts to decrease at 600 K while the Ni 848 eV intensity starts to increase at the same annealing temperature. No significant increase of Pt 237 eV AES intensity is observed unless T_a≥ 800 K. The Ni and Co atoms intermix to each other when 600 K ≤ Ta ≤ 800 K. The Ni-Co intermixing layer diffuses into the Pt substrate increase the perpendicular magnetic anisotropy when 800 K≤ Ta ≤ 900 K. Comparing the annealing induced SRT between 2 ML Ni/2 ML Co/Pt(111) and 2 ML Co/2 ML Ni/Pt(111), the change of the surface composition by annealing may change the effective anisotropy of the ferromagnetic film thus causes the spin reorientation transition.

The effective anisotropies the as-deposited films of 2 ML Ni/2 ML Coi/Pt(111) and 2 ML Co/2 ML Ni/Pt(111) can be expresses as eq. (6.4) and eq. (6.5), respectively. The two groups marked as (1) and (2) have the same value when the buffer layer changes from Co films to Ni films. The values of K_s^{Co−P t}, K_s^{N i−P t}, K_s^Co−vacuum, and K_s^{N i−vacuum} are 0.57 mJ/m² [101], 0.17 mJ/m² [16], −0.28 mJ/m² [59], and −0.48 mJ/m² [60], respec-tively. The value of K_s^{Co−P t} is about 3.5 times the value of K_s^{N i−P t}, but the value of K_s^{N i−vacuum} is only 1.5 times that of K_s^Co−vacuum. Therefore, it is reasonable that the direction of the magnetization easy axis changes from the out-of-plane direction to the in-plane direction when the buffer layer changes from Co films to Ni films. (out-of plane:

K_{ef f} > 0; in-plane: K_{ef f} < 0)

After high temperature annealing, the Co and Ni atoms diffuse into the Pt substrate.

It is straightforward that, after higher temperature annealing, the ratio of Ni% to Co%

of 2 ML Ni/2 ML Co/Pt(111) is higher than that of 2 ML Co/2 ML Ni/Pt(111) in the subsurface region. Alternatively, in the bulk region, the ratio of Ni% to Co% of 2 ML Ni/2 ML Co/Pt(111) must be smaller than that of 2 ML Co/2 ML Ni/Pt(111). The schematic diagrams of the surface compositions are shown in Fig. 6.15. The ratio of the black region is lower than that of the white region. Comparing the spin reorientation transitions of 2 ML Ni/2 ML Co/Pt(111) and 2 ML Co/2 ML Ni/Pt(111), it seems that the sample favors an in-plane easy axis when the ratio is larger in the top surface region and smaller in the bulk region. But it is interesting that no SRT was observed for the 1 ML Ni/1 ML Co/Pt(111) and 1 ML Co/1 ML Ni/Pt(111) systems. Unlike the one-ML mirror systems, the ratios of Ni% to Co% for 2 ML Ni/2 ML Co/Pt(111) and 2 ML Co/2 ML Ni/Pt(111) systems do not keep constant during high temperature (not shown here). This implies that, after high temperature annealing, the surface compositions of the two-ML mirror systems are less homogeneous than one-ML mirror systems. This may be the reason that causes the phenomenon of SRT after high temperature annealing.

Figure 6.15: Surface structures after high temperature annealing for 2 ML Ni/2 ML Co/Pt(111) and 2 ML Co/2 ML Ni/Pt(111)

The effective anisotropy constant K_{ef f}^alloy of a binary ferromagnetic alloy A_xB1−x can be calculated approximately from eq. (6.6) [16, 115]

K_{ef f}^alloy= xK^A+ (1− x)K^B (6.6)

, where K^A, and K^B are the anisotropy constant of ferromagnetic A and B metals.

Lin et al. [116] have been studied the alloying and strain relaxation effects on spin reorientation transitions in Co_xNi1−x/Cu₃Au(100). They indicated that, from the view of K_s, increasing the Co concentration x will shifts K_{ef f} upward, and in the view of K_v, increasing x will shifts K_{ef f} downward. The alloy effects on K_s and K_v are opposite to each other. Although the alloy composition of our two-ML mirror systems is not as homogeneous as that in the report of Lin et al. [116], the alloy effect on K_{ef f} still can be seen. The ratio is higher in the subsurface region and lower in the bulk region causes a decrease of perpendicular magnetic anisotropy. In other words, after the Ni-Co-Pt alloy formation of 2 ML Ni/2 ML Co/Pt(111) occurs, K_{ef f}^{N iCoP t} shifts upward from negative to positive by increasing the Ni concentration in the subsurface region and the Co concentration in the bulk region, this causes the easy axis of magnetization changes its direction from the out-of-plane to the in-plane. Alternatively, after the Ni-Co-Pt alloy formation of 2 ML Co/2 ML Ni/Pt(111) occurs, K_{ef f} shifts downward from positive to negative by decreasing the Ni concentration in the subsurface region and the Co concentration in the bulk region, this causes the easy axis of magnetization changes its direction from the in-plane to the out-of-plane. Concluding speaking, the direction of the magnetization easy axis can be adjusted by varying the concentration of Ni in the subsurface region and the Co concentration in the bulk region.

Summary

1. At room temperature, 1-24 ML Ni/1 ML Co/Pt(111) system exhibits strong per-pendicular magnetic anisotropy.

2. Whether the Co atoms contact with the Pt atoms is very important to the obser-vation of PMA of d_Co/1 ML Ni/Pt(111).

3. The surface magnetic structure changes from a 2D-like to a 3D-like when the surface alloy changes from Ni_xCo_1−xPt to Ni_xCo_1−xPt₃.

4. The depression of Curie temperature becomes more rapidly when the surface alloy changes from Ni_xCo_1−xPt to Ni_xCo_1−xPt₃.

5. The Curie temperature is mainly dominated by three factors: (1) the change of surface structure from Ni_xCo_1−xPt to Ni_xCo_1−xPt₃; (2) the thickness of the ferro-magnetic film; (3) the ratio of Ni% to Co% in the subsurface region and bulk region of Ni-Co-Pt alloy.

6. It is evident that the enhancement of the coercivity for 3 ML Co/1 ML Ni/Pt(111) is strongly related to the concentration of Pt of the surface alloy. The Pt atoms act as domain wall pinning sites.

7. The surface compositions of the two-ML mirror systems are less homogeneous than one-ML mirror systems. This may be the reason that causes the phenomenon of SRT after high temperature annealing.

8. The direction of the magnetization easy axis can be adjusted by varying the con-centrations of the Ni and the Co.