Growth, structure and magnetic properties of magnetron sputtered FePt thin films

The L10 FePt phase belongs to the most promising hard ferromagnetic materials for high density recording media. The main challenges for thin FePt films are: (i) to lower the process temperature for the transition from the soft magnetic A1 to the hard magnetic L10 phase, (ii) to realize c-axes preferential oriented layers independently from the substrate nature and (iii) to control layer morphology supporting the formation of FePt - L10 self-organized isolated nanoislands towards an increase of the signal-to-noise ratio. In this study, dc magnetron sputtered FePt thin films on amorphous substrates were inve-stigated. The work is focalized on the correlation between structural and magnetic properties with respect to the influence of deposition parameters like growth mode (co-sputtering vs. layer – by - layer) and the variation of the deposition gas (Ar, Xe) or pressure (0.3 - 3 Pa). In low-pressure Ar discharges, high energetic particle impacts support vacancies formation during layer growth lowering the phase transition temperature to (320 +/- 20)°C. By reducing the particle kinetic energy in Xe discharges, highly (001) preferential oriented L10 - FePt films were obtained on a-SiO2 after vacuum annealing. L10 - FePt nano-island formation was supported by the introduction of an Ag matrix, or by random ballistic aggregation and atomic self shadowing realized by FePt depositions at very high pressure (3 Pa). The high coercivity (1.5 T) of granular, magnetic isotropic FePt layers, deposited in Ar discharges, was measured with SQUID magnetometer hysteresis loops. For non-granular films with (001) preferential orientation the coercivity decreased (0.6 T) together with an enhancement of the out-of- plane anisotropy. Nanoislands show a coercive field close to the values obtained for granular layers but exhibit an in-plane easy axis due to shape anisotropy effects. An extensive study with different synchrotron X-ray scattering techniques, mainly performed at the ESRF, BM-20 (ROBL-Beamline), pointed out the importance of in-situ investigations to clearly understand the kinetic mechanism of the A1 to L10 transition and ordering and to control FePt nanoclusters evolution