The constantly developing computer technology requires an extensive search for materials to increase the data density. One of the factors influencing the applicability of a given material in magnetic recording media is strong magnetocrystalline anisotropy. This property is observed, for example, in the L10 FePt phase [1].
For this reason, its magnetic properties have been repeatedly analyzed using the density functional theory (DFT) [2]. We decided to perform further research, emphasizing the full potential methods, including the full-potential local-orbital (FPLO) scheme. We used disordered local moments theory and coherent potential approximation to obtain Curie temperatures of the bulk L10 FePt. We paid particular attention to the dependence of the magnetocrystalline anisotropy energy (MAE) on the value of the magnetic moment (m) of the system using the fully relativistic fixed spin moment (FSM) method for various exchange-correlation potentials. This dependency allowed us to explain the differences between the results obtained earlier using different exchange-correlation potentials and determine the potential ways of system modification to increase the MAE of the system. Interestingly, the determined dependencies of MAE(m) for different correlation-exchange potentials are almost identical near the equilibrium point [3]. We also determined the magnetocrystalline anisotropy constants $K_1$ and $K_2$, the magnetic moments and the Curie temperature.
The basic magnetic parameters [4] were studied further for a decreasing number of atomic layers, down to a single atomic bilayer of L10 FePt. A complete structural optimization preceded the calculations. We derived the dependence of the in-plane lattice parameter and the c/a ratio in a single unit cell on the multiatomic layer thickness. Furthermore, the dependence of the magnetic moment versus the distance to the interface was obtained.