Spin and Charge Interconversion in Dirac-Semimetal Thin Films

We use spin torque ferromagnetic resonance and ferromagnetic-resonance-driven spin pumping to detect spin-charge interconversion at room temperature in heterostructure devices that interface an archetypal Dirac semimetal, ${\mathrm{Cd}}_3{\mathrm{As}}_2$, with a metallic ferromagnet, ${\mathrm{Ni}}_{0.80}{\mathrm{Fe}}_{0.20}$ (permalloy). Angle-resolved photoemission directly reveals the Dirac-semimetal nature of the samples prior to device fabrication and high-resolution transmission electron microscopy is used to characterize the crystalline structure and the relevant heterointerfaces. We find that the spin-charge interconversion efficiency in ${\mathrm{Cd}}_3{\mathrm{As}}_2/$permalloy heterostructures is comparable to that in heavy metals and that it is enhanced by the presence of an interfacial oxide. Spin torque ferromagnetic resonance measurements reveal an in-plane spin polarization regardless of an oxidized or pristine interface. We discuss the underlying mechanisms for spin-charge interconversion by comparing our results with first principles calculations and conclude that extrinsic mechanisms dominate the observed phenomena. Our results indicate a need for caution in interpretations of spin-transport and spin-charge conversion experiments in ${\mathrm{Cd}}_3{\mathrm{As}}_2$ devices that seek to invoke the role of topological Dirac and Fermi arc states.

Figure 1

RHEED pattern of the ${\mathrm{Cd}}_3{\mathrm{As}}_2$ film along the (a) $[\overline{1}\overline{1}1]$ and (b) $[1\overline{1}0]$ crystal directions. (c) X-ray diffraction $2\theta$ scan of ${\mathrm{Cd}}_3{\mathrm{As}}_2$ films. (d) AFM image of the ${\mathrm{Cd}}_3{\mathrm{As}}_2$ surface. High-angle annular dark-field STEM image showing the different layers in (e) a complete device stack and (f) just the ${\mathrm{Cd}}_3{\mathrm{As}}_2$ layer.

Figure 2

(a) Electronic band structure of ${\mathrm{Cd}}_3{\mathrm{As}}_2$ thin films obtained by ARPES. The measurements are taken at $T=300$ K along the $K$–$\mathrm{\Gamma }$–$K$ direction, projected on the (112) surface. (b) Second derivative of the spectrum shown in (a). (c) Calculated band structure and spin Hall conductivity (SHC) of bulk ${\mathrm{Cd}}_3{\mathrm{As}}_2$ The color in the band structure shows the spin Berry curvature (${\mathrm{\Omega }}_{xy}^z$) with $z$ along the [001] direction. (d) Three independent components of SHC with $z$ along the [112] direction as a function of chemical potential.

Figure 3

(a) Illustration of charge-to-spin conversion due to ST-FMR. (b) Optical microscope image of the ST-FMR device. ST-FMR mixing voltage signal measured in (c) a $\mathrm{Py}(4\mathrm{nm})/{\mathrm{Cd}}_3{\mathrm{As}}_2(12\mathrm{nm})$ heterostructure with ${\mathrm{Cd}\mathrm{O}}_x$ and (d) in an in vacuo grown $\mathrm{Py}(5\mathrm{nm})/{\mathrm{Cd}}_3{\mathrm{As}}_2(8\mathrm{nm})$ heterostructure without ${\mathrm{Cd}\mathrm{O}}_x$. We used a microwave power of 20 dBm and swept the magnetic field at different frequencies (data have been offset in both cases for clarity). (e),(f) ST-FMR spectra of the same devices shown in (c),(d), respectively, at 4 GHz showing the measured data and the fit needed to extract the symmetric ($S$) and antisymmetric ($A$) components of the torque. Angle dependence of the symmetric and antisymmetric components of the torque fitted with $\sin (\varphi )\cos (\varphi {)}^2$ for (g) the sample with ${\mathrm{Cd}\mathrm{O}}_x$ and (h) for the completely in vacuo grown sample.

Figure 4

(a) Illustration of spin-to-charge conversion due to SP. (b) Voltage signal measured in the ${\mathrm{Cd}}_3{\mathrm{As}}_2(40\mathrm{nm})/\mathrm{Py}(30\mathrm{nm})$ heterostructure at room temperature with an applied microwave power of 20 dBm. (c) SP voltage signal $(V_{\mathrm{SP}})$ and FMR absorption spectra $(dI/dH)$ measured at 3 GHz and 23 dBm. (d) Charge current ($J_c$) as a function of power for ${\mathrm{Cd}}_3{\mathrm{As}}_2(40\mathrm{nm})/\mathrm{Py}(30\mathrm{nm})$ and $\mathrm{Py}$(30 nm)/$\mathrm{Pt}$(20 nm) heterostructures showing linear behavior.

Figure 5

Band structure, spin Berry curvature, and spin Hall conductivity along the (001) direction. Calculated band structure and spin Hall conductivity of bulk ${\mathrm{Cd}}_3{\mathrm{As}}_2$. The color in the band structure shows the spin Berry curvature (${\mathrm{\Omega }}_{xy}^z$). The right panel presents three independent values of spin Hall conductivity calculated in the $⟨001⟩$ direction as a function of chemical potential.

Figure 6

(a) Theoretical angle dependent functional form of the ST-FMR signal with spin polarization along different axis. (b) Image of one of our samples showing the coordinate axis. Angle dependence of the (b) symmetric and (c) antisymmetric components of the torque fitted with spin polarization along the $y$ axis (red) and all axes (black).

Figure 7

(a) Voltage signal measured in a $\mathrm{Py}(30\mathrm{nm})/{\mathrm{Cd}}_3{\mathrm{As}}_2(40\mathrm{nm})$ heterostructure as we sweep the magnetic field. Magnitude of the (b) antisymmetric and (c) symmetric components of the voltage signal at resonance condition as a function of power.

Figure 8

STEM-EDX elemental map of the ${\mathrm{Cd}}_3{\mathrm{As}}_2/\mathrm{Py}$ interface. (a) HAADF-STEM image and elemental maps of $\mathrm{Ni}$, $\mathrm{Fe}$, $\mathrm{O}$, $\mathrm{Cd}$, and $\mathrm{As}$. Scale bar is 10 nm. (b) Concentration of the elements in (a) across the interface. Note the approximately 1 nm amorphous layer of ${\mathrm{Cd}\mathrm{O}}_x$ at the interface.

Figure 9

(a) HAADF-STEM image of the top layers of the device. The scale bar is 10 nm. (b) Concentration of the elements in (a) across the device interface layers shown. (c) Elemental maps of $\mathrm{O}$, $\mathrm{Al}$, $\mathrm{Ni}$, $\mathrm{Fe}$, $\mathrm{As}$, $\mathrm{Cd}$, $\mathrm{Ga}$ and $\mathrm{Sb}$. Scale bar is 10 nm.

Figure 10

(a) ST-FMR mixing voltage measured at different frequencies in a completely in vacuo grown heterostructure. (b) Angle dependance of the symmetric (red) and antisymmetric (blue) components of the torque fitted with an in plane spin polarization (continuous line) and a three-dimensional spin polarization (dotted line).

Figure 11

(a) Spin-dependent PNR data from a ${\mathrm{Cd}}_3{\mathrm{As}}_2/\mathrm{Py}$ bilayer at room temperature in an applied field of 2 T alongside a theoretical fit. We show the Fresnel reflectivity, where the measured reflectivity is normalized by the $Q$-dependent theoretical reflectivity of the $\mathrm{Ga}\mathrm{As}$ substrate. (b) Measured spin asymmetry, defined as ($R^{\uparrow \uparrow }-R^{\downarrow \downarrow })/(R^{\uparrow \uparrow }+R^{\downarrow \downarrow }$) alongside theoretical fit. (c) Best fit nuclear and magnetic scattering length density (SLD) profiles used to generate the fits shown. The right axis shows the magnetization which is equivalent to a given magnetic SLD value.

Figure 12

(a) Spin Hall angle determined from ST-FMR at different frequencies in different samples. (b) The $J_c/J_s$ ratio and spin mixing conductance $(g_{\uparrow \downarrow })$ obtained from SP in heterostructures with 30 nm of $\mathrm{Ni}\mathrm{Fe}$ and variable ${\mathrm{Cd}}_3{\mathrm{As}}_2$ thickness.