B01 - Ultrafast spin currents and spin torques studied by (non-)linear magneto-optics

Principal Investigator:

A central ingredient for ultrafast spintronics is the interaction of femtosecond spin-current pulses with magnetic nanostructures. In ferromagnet (FM)/non-magnetic (NM) bilayer systems, the non-thermal spin-dependent Seebeck effect allows one to generate high-amplitude pulses by optical means. The spatio-temporal properties of the spin-transfer torque (STT) exerted by ultrashort spin-current pulses on the FM, provide access to exchange-dominated spin waves in the (sub) THz frequency range.

This project aims at the fundamental understanding of elementary processes in spin-dependent scattering of hot (non-equilibrium) charge carriers in metallic multilayer structures and at the investigation of spin transfer torque effects generated by femtosecond spin-current pulses on the magnetization of itinerant ferro- and ferrimagnets. In FM/NM/FM structures, these pulses propagating across the NM layer are optically excited at the interface with a FM layer (emitter). STT can thus be exerted on a second ultrathin FM layer (collector), located on the other side of the NM layer. A suitable technique to monitor spin currents in a time-resolved fashion is the magneto-induced second harmonic generation. The interaction of the spin-current pulses with a FM collector results in a STT-induced perturbation of the magnetization close to the NM/FM interface and the excitation of standing spin-wave modes in the sub THz frequency range.

One goal of the project is to optimize the generation of ultrashort, femtosecond spin-current pulses in metallic multilayers by varying the thickness and material of both emitter and spacer, as well as the photon energy. Further, we will study their propagation across FM layers, NM layers, and interactions at interfaces. We intend to use the spin-current pulses to excite directly the high-frequency exchange mode in ferrimagnetic materials. In particular, we would like to test the possibility of spin-current induced switching with ferrimagnetic collector layers near their compensation point. Furthermore, we will try to control and shape the polarization of spincurrent pulses. The main material system we intend to use is Fe/Au/Fe, epitaxially grown on MgO(001) substrates. In addition to Fe, also FeCo and ferrimagnetic rare earth-NiFe alloys will be used as collector materials.

The STT-induced spin dynamics will be studied by varying the collector thickness and thus the frequencies of the excited perpendicular standing spin-wave modes. In particular, we plan to achieve an enhanced excitation of only the first mode with relatively high and variable frequency in order to resonantly couple its excitation into exchange-coupled ferromagnetic and antiferromagnetic insulators, such as EuO, Fe3O4, or NiO. Another interesting aspect here is the damping of the high-frequency modes, which we will study directly in the time domain on the ultrafast timescale.

Ultrashort spin-current pulses may also be generated by the spin Hall effect (SHE) using THz pulses to drive charge-current pulses in metallic wires. This second set of experiments will allow us to investigate the SHE on the ultrafast time scale in heavy elements such as Au, Pt, and Ta layers as well as NM/FM bi-layers at the sub-picosecond time scale. For this, the SHE-induced spin accumulation at the surface of the metal will be monitored by linear and nonlinear magneto-optics. We will verify how the measured signals correlate with the spin Hall angles measured at dc and study the dynamics of the build-up of the SHE-induced spin accumulation on an ultrafast timescale directly within the normal metal. Finally, we will explore the SHE even at optical frequencies (400-800 THz) by analyzing its contribution to linear and nonlinear magneto-optical effects.