We report a comprehensive study of the maximization of the spin Hall ratio (${\theta }_{\mathrm{SH}}$) in $\mathrm{Pt}$ thin films by the insertion of submonolayer layers of $\mathrm{Ti}$ to decrease carrier lifetime while minimizing the concurrent reduction in the spin Hall conductivity (${\sigma }_{\mathrm{SH}}$). We establish that the intrinsic ${\sigma }_{\mathrm{SH}}$ and carrier lifetime sets a practical upper bound of ${\theta }_{\mathrm{SH}}$ of $\mathrm{Pt}$, while robust against the strain and the moderate interruption of crystal order caused by these insertions, begins to decrease rapidly at high resistivity level because of the shortening carrier lifetime. The unavoidable trade-off between the intrinsic ${\sigma }_{\mathrm{SH}}$ and carrier lifetime sets a practical upper bound of ${\theta }_{\mathrm{SH}}$ ≥ 0.8 for heterogeneous materials where the crystalline $\mathrm{Pt}$ component is the source of the spin Hall effect and the resistivity is increased by shortening the carrier lifetime. This work also establishes a very promising spin Hall metal of [$\mathrm{Pt}$ 0.75 nm/$\mathrm{Ti}$ 0.2 nm]${}_7$/$\mathrm{Pt}$ 0.75 nm for energy-efficient, high-endurance spin-orbit-torque technologies (e.g., memories, oscillators, and logic) due to its combination of a giant ${\theta }_{\mathrm{SH}}$ ≈ 0.8, or equivalently, a dampinglike spin-torque efficiency per unit current density ${\xi }_{\mathrm{DL}}^j\approx 0.35$, with a relatively low resistivity (90 µΩ cm) and high suitability for practical technology integration.

• Received 18 May 2019
• Revised 4 September 2019

DOI:https://doi.org/10.1103/PhysRevApplied.12.051002