From our results we can conclude the following:
- We have constructed an effective semiclassical (spin localized) Heisenberg Hamiltonian, passing the magnetic information from "ab-initio" calculations to the atomistic level.
- We have shown that the magnetic moment in Co thin films is affected by the existence of the surface and the capping. In the case of
thin films we observed a polarization of the Ag layer. However, the modifications of the spin and the orbital magnetic moment of Ag are small.
- We have obtained the layer resolved effective anisotropy constant and observed that it has experienced a huge increment due to the breaking of symmetry on the surface. Also we have shown that its character could be changed by the surface capping. A transformation from a system with easy axis perpendicular to the plane to the one with easy plane anisotropy by one or two layers of Ag capping has been predicted.
- The symmetry breaking of the surface induces an increment in the value of the exchange interaction. Also it leads to an asymmetry of the exchange tensor interaction in the direction Z perpendicular to the surface. This, together with the Dzyaloshinsky-Moriya interaction contributes to an overall macroscopic anisotropy.
- In the case of a thin film
we have discussed different exchange interaction contributions, and the temperature dependence of the macroscopic magnetization and anisotropy. We have observed that the magnetization dependence on the temperature can be reproduced both with the truncated exchange tensor and within the isotropic exchange constant approximation . Nevertheless the dependence of the anisotropy on the temperature is highly dependent on the correct account of the exchange interaction. This indicates that the IEC is not able to provide a correct temperature dependence of the magnetic properties of the system, in this case we should use a total exchange matrix.
Rocio Yanes