When an electric field is established within a metal sheet, electrons flow through it. In the presence of an applied magnetic field perpendicular to the current, a Lorentz force acts on the moving electrons deviating them transversally, so a local charge imbalance accumulates at either side of the sheet originating an equivalent Hall voltage (whose magnitude equilibrates the Lorentz force), which was early discovered by E. Hall in 1879. The so-called Hall effect has found application for example in magnetic field sensors such as the Hall probes. Moreover, in non-magnetic metals, even in the absence of any applied magnetic field and in the presence of a current flow, as electron scattering is spin dependent, spin-up and spin-down electrons suffer lateral forces of different sign that produce a spin accumulation (up and down) at opposite sides of the samples. This is a pure spin current, as not net flow of charge has occurred because the same amount of charge has been deflected at each side. Therefore magnetization of opposed sign is formed at sample edges and accumulates until the corresponding change of the electrochemical potential equals the magnetic force acting on the charge carriers. This is referred as Spin Hall effect. The same amount of charge is accumulated at each side, and thus no Hall voltage appears. As the amount of accumulated spins is the result of a spin equilibrium resulting from the charge flow, any perturbation on the spin accumulation should result in a modification of the charge flow (this is the so-called "inverse Hall effect") and thus the sample electric resistivity should vary. This is the so-called "Spin Magnetoresistance".
A convenient way to modify the spin accumulation is placing an adjacent insulating magnetic material. If its magnetization is perpendicular to the spin accumulation, then a torque among the magnetization and the accumulated spins should occur that will ultimately modify the magnetization distribution in both materials across the interface (a spin diffusion across will take place). Therefore a change of the resistivity of the metal will be produced. The simplest way to observe this effect is to magnetize the insulating ferromagnetic layers along two perpendicular directions and simultaneously measure the resistivity of the metallic layer. This opens opportunities for creating spin-induced voltages in metals. As spin currents do not dissipate heat, it is expected that pure spin currents could be the basis of a radically new family of spintronic materials, more energy friendly and faster.
In practice, however, when placing a non-magnetic material next to a magnetic material, a proximity-induced ferromagnetism may be induced in the former, which may give rise to a magnetoresistance analogous to that of conventional ferromagnets. Therefore, understanding and controlling pure-spin-current-based devices requires disentangling and excluding any proximity effect.
The goal of the study was to settle whether the observed magnetoresistance in Pt/CoFe2O4 hybrid systems is a signature of spin accumulation at the interface or could arise from a more conventional mechanism due to magnetic moment induced on the Pt metal layer by the proximity effect of the CoFe2O4 ferromagnetic layer. To this end, the authors performed XMCD measurements at the Pt-M3 and Pt-L2,3 edges of the BOREAS beamline at ALBA and the ID12 at ESRF, respectively probing element-specific Pt magnetism in Pt/CoFe2O4 (Pt/CFO) heterostructures. Results showed that the Pt magnetic moment is negligible, i.e. approximately below 0.001 Bohr magneton per atom averaged over the Pt layer, thus ruling out the occurrence of proximity effects in favour of the non-conventional spin-current origin for the observed transport effects.
Fig: Schematics for the magnetic moments originated in Platinum on the case of (left) proximity effects and (right) spin accumulation due to Spin Hall effect.
The study therefore supports the occurrence of spin currents in this novel type of materials, providing a strong experimental evidence that it is possible to engineer spintronic materials and devices using spin current and spin accumulation effects, and in particular Pt and CoFe2O4 insulating ferromagnet interfaces, avoiding interfering proximity effects.
