Spintronics is mainly based on the phenomenon of spin accumulation, which is inherent to the circulation of an electric current at the interfaces between ferromagnetic and non-magnetic materials. These accumulations are conventionally obtained in multilayers for which the thicknesses of the layers are smaller than the characteristic lengths of the spin-dependent transport. It is thus possible to generate in these multilayers magnetoresistances or spin transfer effects.
The development of nanofabrication processes makes it nowadays possible to create nanodevices whose lateral dimensions are less than the characteristic lengths of the spin-dependent transport, and thus to bring into play these same phenomena. During my thesis I studied different F / N lateral nanostructures, showing that it is possible to take advantage of the three-dimensional geometry of the structures, and of the different possible orientations of the injected spins. In particular, transport studies have been carried out in collinear and non-collinear regimes, in order to study the consequences of the non-collinearity on the spin accumulations and magnetoresistances.
After an introductory chapter on spin-dependent electron transport, the second chapter demonstrates the interest of the CoFe alloy in lateral structures. The third chapter explores the new opportunities offered by lateral structures in the case of collinear transport. The non-collinear case of spin transport through a ferromagnetic material is then examined using spin absorption measurements and Hanle measurements. Finally, the exploitation of pure spin currents in order to realize functional devices is studied in Chapters V and VI. In particular, new nanostructures whose geometry takes advantage of the three directions of space (based on both vertical and lateral spin transport) are presented in Chapter VI.