No Arabic abstract
In the normal metal/ferromagnetic insulator bilayer (such as Pt/Y$_{3}$Fe$_{5}$O$_{12}$) and the normal metal/ferromagnetic metal/oxide trilayer (such as Pt/Co/AlO$_{x}$) where spin injection and ejection are achieved by the spin Hall effect in the normal metal, we propose a minimal model based on quantum tunneling of spins to explain the spin-transfer torque and spin pumping caused by the spin Hall effect. The ratio of their damping-like to field-like component depends on the tunneling wave function that is strongly influenced by generic material properties such as interface $s-d$ coupling, insulating gap, and layer thickness, yet the spin relaxation plays a minor role. The quantified result renders our minimal model an inexpensive tool for searching for appropriate materials.
We have proposed a method to synchronize multiple spin-transfer torque oscillators based on spin pumping, inverse spin Hall, and spin Hall effects. The proposed oscillator system consists of a series of nano-magnets in junction with a normal metal with high spin-orbit coupling, and an accumulative feedback loop. We conduct simulations to demonstrate the effect of modulated charge currents in the normal metal due to spin pumping from each nano-magnet. We show that the interplay between the spin Hall effect and inverse spin Hall effect results in synchronization of the nano-magnets.
The polarization of the spin current pumped by a precessing ferromagnet into an adjacent normal metal has a constant component parallel to the precession axis and a rotating one normal to the magnetization. The former component is now routinely detected in the form of a DC voltage induced by the inverse spin Hall effect (ISHE). Here we compute AC-ISHE voltages much larger than the DC signals for various material combinations and discuss optimal conditions to observe the effect. Including the backflow of spins is essential for distilling parameters such as the spin Hall angle from ISHE-detected spin pumping experiments.
We demonstrate that the spin Hall effect in a thin film with strong spin-orbit scattering can excite magnetic precession in an adjacent ferromagnetic film. The flow of alternating current through a Pt/NiFe bilayer generates an oscillating transverse spin current in the Pt, and the resultant transfer of spin angular momentum to the NiFe induces ferromagnetic resonance (FMR) dynamics. The Oersted field from the current also generates an FMR signal but with a different symmetry. The ratio of these two signals allows a quantitative determination of the spin current and the spin Hall angle.
We theoretically examine the spin-transfer torque in the presence of spin-orbit interaction (SOI) at impurities in a ferromagnetic metal on the basis of linear response theory. We obtained, in addition to the usual spin-transfer torque, a new contributioin $sim {bm j}_{rm SH}^{phantom{dagger}} cdot abla {bm n}$ in the first order in SOI, where ${bm j}_{rm SH}^{phantom{dagger}}$ is the spin Hall current driven by an external electric field. This is a reaction to inverse spin Hall effect driven by spin motive force in a ferromagnet.
Efficient generation of spin-orbit torques (SOTs) is central for the exciting field of spin-orbitronics. Platinum, the archetypal spin Hall material, has the potential to be an outstanding provider for spin-orbit torques due to its giant spin Hall conductivity, low resistivity, high stabilities, and the ability to be compatible with CMOS circuits. However, pure clean-limit Pt with low resistivity still provides a low damping-like spin-orbit torque efficiency, which limits its practical applications. The efficiency of spin-orbit torque in Pt-based magnetic heterostructures can be improved considerably by increasing the spin Hall ratio of Pt and spin transmissivity of the interfaces. Here we reviews recent advances in understanding the physics of spin current generation, interfacial spin transport, and the metrology of spin-orbit torques, and summarize progress towards the goal of Pt-based spin-orbit torque memories and logic that are fast, efficient, reliable, scalable, and non-volatile.