Large enhancement of the spin Hall effect in Au by side-jump scattering on Ta impurities

We present measurements of the spin Hall effect (SHE) in AuW and AuTa alloys for a large range of W or Ta concentrations by combining experiments on lateral spin valves and ferromagnetic-resonance/spin-pumping techniques. The main result is the identification of a large enhancement of the spin Hall angle (SHA) by the side-jump mechanism on Ta impurities, with a SHA as high as $+0.5$ (i.e., $50\%$) for about 10% of Ta. In contrast, the SHA in AuW does not exceed $+0.15$ and can be explained by intrinsic SHE of the alloy without significant extrinsic contribution from skew or side-jump scattering by W impurities. The AuTa alloys, as they combine a very large SHA with a moderate resistivity (smaller than $85\mu \mathrm{\Omega }\text{cm}$), are promising for spintronic devices exploiting the SHE.

Figure 1

(a) SEM image of a nonlocal lateral spin valve structure with inserted AuW nanowire. (b) Typical nonlocal spin signals in experiments of spin absorption by the SHE material (AuW): reference device without AuW (blue) and device with AuW (red). (c) Field dependence of the signal induced by inverse spin Hall effect at different orientations of the field. The insets in (b) and (c) display the respective nonlocal probe configurations.

Figure 2

(a) Spin diffusion lengths extracted using spin-absorption experiments in LSVs (green) or spin-pumping voltage in FMR (red). Inset: The $\rho \times {\lambda }_{sf}$ product remains approximately constant in the whole experimental range. (b) Dependence of the spin Hall angles derived by LSVs and SP-FMR techniques on the AuW concentration/resistivity. The dashed line represents the expected intrinsic contribution derived by averaging data on Au films in the $4\text{–}19\mu \mathrm{\Omega }\text{cm}$ resistivity range (see text). (c) Same as in (b) for the product of the spin Hall angle and the spin diffusion length characterizing the efficiency of the spin-charge conversion [17, 26].

Figure 3

(a) and (b) Ferromagnetic resonance (FMR) spectra (a) and inverse spin Hall effect signal (b) of the $\text{AuTa}2.5\%\left(40\text{nm}\right)\setminus \text{NiFe}\left(15\text{nm}\right)$ sample, measured in parallel and antiparallel positions (relative to the external magnetic field) in cavity, with $\text{rf}$ power of 50 mW. (c) Schematic representation of the experiment geometry. (d) AuTa thickness dependencies of charge current production by ISHE for Ta concentrations of $2.5\%$, $5\%$, $7.5\%$, and $10\%$ and normalized in units of rf field excitation, $\mu \text{A}/{\text{G}}^2$. Lines correspond to fits using Eq. (1).

Figure 4

(a) Spin diffusion lengths extracted from the thickness dependence of the SP-FMR of Fig. 3 for the AuTa alloy. Inset evidences that $\rho {\lambda }_{sf}$ remains constant. (b) Spin Hall angle of AuTa alloys for different thickness of AuTa as a function of the Ta concentration (or AuTa resistivity). (c) Variation of the product of the spin Hall angle by the spin diffusion length characterizing the efficiency of the spin-charge conversion [17, 26].