3.1 Motivation

Colleagues at the Cavendish Laboratory had earlier performed a series of measurements on the system formed by growing a $0.5ML$ $Mn/6ML$ $Co/Cu(001)$ multilayer [18,20,21]. Low-energy electron diffraction (LEED) revealed, through the existence of $(\frac{1}{2}\frac{1}{2})$ order spots (figure 3.1,) a $I_{ci}$ (chessboard-like) superstructure in the plane of the film.

Figure 3.1: LEED Pattern from $0.5ML$ $Mn/6ML$ $Co/Cu(001)$ System (Left Panel.) The Existence of the $\left( \frac{1}{2}\frac{1}{2}\right)$ Spots Indicates that a $I_{ci}$ Superstructure Has Formed in the Plane of the Film. Plot of Intensity of $\left( \frac{1}{2}\frac{1}{2}\right)$ Spot Relative to Specular $R_{ij}$ Spot as a Function of $Mn$ Coverage (Right Panel.) Figure Reproduced from [20].

Three possible explanations were advanced for the $I_{ci}$ superstructure[18,20,21]:

• Inter-diffusion of copper atoms through the cobalt layer may have led to the formation of a $MnCu$ surface alloy [13,15]. However, these experiments were performed at room temperature, which is too cold to permit significant inter-diffusion through $6ML$ $Co$ [12]. This suggestion was therefore ruled out.

• The manganese may have had an internal $I_{ci}$ anti-ferromagnetic configuration. However, the measured $(\frac{1}{2}\frac{1}{2})$ spot intensity was $\sim 10$ times that predicted for magnetic diffraction of this type by Tamura et al. [7]. This phenomenon is therefore incapable of providing a complete explanation of the observed superstructure.

• An ordered $MnCo$ surface alloy may have been formed. Given the evidence against the other two models, the observed diffraction pattern was attributed to $MnCo$ alloy formation.

Figure 3.2 shows the results of magneto-optical Kerr effect (MOKE) measurements taken by colleagues at the Cavendish Laboratory as part of the same project. The $[110]$ direction, easy axis Kerr signal is seen to increase with the addition of manganese in the regime where the alloy exists. The immediate reaction to this is that the system's magnetic moment is increasing, i.e. that the manganese atoms which are being added are ferromagnetically aligned both with each other and with the cobalt. However, the simultaneous drop in $[100]$ Kerr signal does not paint the same picture, and it may be that a change in magneto-optical response, rather than a genuine addition of magnetic moments, is responsible for the increased $[110]$ signal. The latter is a possibility which is in particular need of investigation given the tight-binding calculation by Noguera et al. which predicts a $c(2x2)$ anti-ferromagnetic configuration for the manganese atoms, although this calculation neglected the possibility of the structural alloying observed by LEED, which seems likely to have a very significant effect on the magnetism.

Figure 3.2: Amplitude of Magneto-Optical Kerr Effect (MOKE) Signal, an Indicator of Magnetization, as a Function of $Mn$ Thickness in a $Mn/6ML\: Co/Cu(001)$ System (Left Panel.) Coercive Field as a Function of $Mn$ Thickness in the Same System(Right Panel.) Figure Reproduced from [20].

The polarization of a reflected electron beam from the sample surface provides an alternative to MOKE for measuring the magnetization through the difference, for the two electron spins, in density of states at a particular energy created by the exchange splitting. To this end, the polarizations of reflected electron beams at three energies from a $10ML\: Cu/0.5ML\: Mn/7ML\: Co/Cu(001)$ sample have been measured.