My PhD work is dedicated to the spin transfer induced self-sustained dynamics of two coupled vortices, in nano-pillars spin-valves structures (Py/Cu/Py). A first objective was to understand the spin-polarized transport processes as well as spin transfer mechanisms associated to highly non-homogeneous magnetic configurations. This study allows me to identify and then precisely tune the vortex based magnetic configurations, and notably to observe the influence of spin transfer on reversal mechanisms of the vortex core. Combining analytical calculations and micro-magnetic simulations, we determine the conditions on relative parameters for the two vortices (chiralities and polarities) necessary to obtain self-sustained gyrotropic oscillations of the coupled vortices in a single pillar. A very interesting case is predicted for the pillars with larger diameters (typically over 200nm) for which the critical current is reduced to zero. The experimental results confirm the predictions that a coupled dynamics exists with linewidths as narrow as 200kHz, that is a record at zero field (corresponding to a quality factor Q ≈ 5000, an order of magnitude over the self-sustained oscillations of a single vortex), and even down to 50kHz under external field.A second objective was to investigate the synchronization of two vortex based spin transfer oscillators. We demonstrate theoretically that the phase locking through dipolar coupling of two identical oscillators can be achieved for any parameters of the two vortex. However, the coupling is three times stronger when vortices have opposite core polarities. From an experimental point of view, the synchronization capability for two oscillators having a frequency mismatch reaching up to 10 % of the auto-oscillation frequency has been demonstrated. This work, being part of the research effort made to improve the rf properties of spin transfer nano-oscillators emphasizes how the excitation of coupled magnetizations modes is important to reach lower and lower linewidths.
