Structural Phase Engineering of (α + β)-W for a Large Spin Hall Angle and Spin Diffusion Length

Abstract

Spin−orbit coupling (SOC) plays a crucial role in spin-to-charge interconversion in ferromagnet (FM)−heavy metal (HM) heterostructures. The SOC of a heavy metal strongly depends on its crystalline phase. Here, we report a systematic study of the effect of growth rate (D r ) on SOC-related parameters such as the spin-mixing conductance (g ↓↑ ), spin diffusion length (λ sd ), spin transparency, and spin Hall angle (θ SH ) in (α + β)-W. X-ray diffraction (XRD) results for W sputtered at different D r values show that the adatom energy is critical in determining the structural phase. The thickness dependence of the XRD results shows that all films sputtered at 0.06 nm/s nucleate as (α + β)-W on the Si substrate. Four-probe resistivity measurements for (α + β)-W (10) show that the resistivity is linearly proportional to D r with a negative slope. A ferromagnetic resonance (FMR)-based spin pumping study shows that the line width (ΔH) is enhanced due to spin current transport across the Py/(α + β)-W interface. The g ↓↑ shows a maximum of 10.1 × 10 18 m −2 and a minimum of 4.2 × 10 18 m −2 at D r of 0.08 and 0.06 nm/s, respectively. The extracted spin diffusion length for (α + β)-W is 0.98 ± 0.01 nm. The calculated spin transparency for (α + β)-W is 0.74. The inverse spin Hall effect (ISHE) is demonstrated for Py(20)/(α + β)-W(10) bilayer heterostructures. From ISHE measurements, the calculated spin Hall angle for (α + β)-W is found to be θ SH = −0.24 ± 0.08. It is observed that D r strongly influences the spin diffusion length, spin transparency, and spin Hall angle. Our systematic study may open new possibilities for tuning spin transparency and spin Hall angle for developing low-power memory and logic devices.