Efficient intrinsic spin-to-charge current conversion in an all-epitaxial single-crystal perovskite-oxide heterostructure of $\mathrm{L}{\mathrm{a}}_{0.67}\mathrm{S}{\mathrm{r}}_{0.33}\mathrm{Mn}{\mathrm{O}}_3/\mathrm{LaAl}{\mathrm{O}}_3/\mathrm{SrTi}{\mathrm{O}}_3$

We demonstrate efficient intrinsic spin-to-charge current conversion in a two-dimensional electron gas using an all-epitaxial single-crystal heterostructure of $\mathrm{LaSrMn}{\mathrm{O}}_3/\mathrm{LaAl}{\mathrm{O}}_3/\mathrm{SrTi}{\mathrm{O}}_3$, which can suppress spin scattering and give us an ideal environment to investigate intrinsic spin-charge conversion. With decreasing temperature to 20 K, the spin-to-charge conversion efficiency is drastically enhanced to $+6.7\mathrm{nm}$. Our band-structure calculation well reproduces this behavior and predicts further enhancement by controlling the density and relaxation time of the carriers.

Figure 1

(a) Schematic illustration of the (001)-oriented full-epitaxial multilayer structure of LSMO/LAO grown on an STO (001) substrate. The sample size is $2\times 1\mathrm{mm}$. In the electron-spin-resonance system, a radio-frequency magnetic field $h_{\mathrm{rf}}$ was applied along the $\left[\overline{1}10\right]$ ($x$) direction of the sample. The static magnetic field ${\mu }_{0}H$ was applied along the [110] ($y$) axis. Here, $M$ represents the magnetization of LSMO. (b) Principle of spin-to-charge conversion via the inverse Edelstein effect; the spin current injected into the LAO/STO interface moves the outer and inner Fermi circles, generating a charge current in the $x$ direction. Here, the dotted and solid lines are the original Fermi circles and the ones after a spin current is injected, respectively. (c) Scanning-transmission-electron microscope-lattice image of the LSMO (30 u.c.)/ LAO (2 u.c.)/STO heterostructure (sample A) projected along the [010] axis. (d) Atomic-force-microscope image of the surface of sample A, in which atomic steps are observed.

Figure 2

(a) Temperature dependence of the magnetization of LSMO (30 u.c.)/LAO (2 u.c.)/STO (sample A). (b) Temperature dependence of the sheet resistance of the reference samples of LAO (8 u.c.)/STO with the ratio $c$ of La to Al of $83\%$ (sample ref-A) and 101% (sample ref-B). (c) Temperature dependences of the mobility and the carrier density measured for sample ref-A $(c=83\%)$. (d) Temperature dependence of the sheet resistance of the LSMO (30 u.c.)/LAO(2 u.c.)/STO samples with a 2DEG (sample A, $c=83\%$) and without a 2DEG (sample B, $c=101\%$). The dotted curve is the sheet resistance reported for LSMO/STO, which is reproduced from Ref. [27] assuming the film thickness to be the same as that of our LSMO layer.

Figure 3

(a)–(c) Magnetic-field ${\mu }_{0}H$ dependences of (a) the microwave absorption derivative, (b) EMF, and (c) $j_{\mathrm{c}}^{\mathrm{A}}$ measured for sample A (LSMO/LAO/STO with a 2DEG) at various temperatures. The used microwave power is 30 mW.

Figure 4

(a) Comparison between $j_{\mathrm{c}}^{\mathrm{A}}$ (with a 2DEG) and $j_{\mathrm{c}}^{\mathrm{B}}$ (without a 2DEG) as a function of temperature. (b) Temperature dependences of experimental ${\lambda }_{\mathrm{IEE}}$ and predicted ${\lambda }_{\mathrm{IEE}}$ for ${\mathrm{\Delta }}_{\mathrm{ASO}}=4$, 5, and 6 meV. (c) Dispersion relation of the 2DEG at the LAO/STO interface (blue curves). The red dotted lines, from bottom to top, are the estimated $E_{\mathrm{F}}$ positions at 20, 40, 60, 80, 100, and 140 K in our sample. (d) Fermi surface of the 2DEG when the carrier density is $2.1\times {10}^{14}\mathrm{c}{\mathrm{m}}^{-2}$ and $E_{\mathrm{F}}=210\mathrm{meV}$, which corresponds to the measurement condition of the IEE at 20 K. The color scale represents the absolute value of the group velocity in the $x$ direction. (e) Details of the Fermi surface with the spin orientations (see arrows). The color scale represents $F_x\left(\mathbf{k}\right)$, which is an indicator of the contribution of each state to the electron flow when a spin current is injected. In (c)–(e), ${\mathrm{\Delta }}_{\mathrm{ASO}}$ is assumed to be 5 meV. Note the large difference in the scale of the axes between (d) and (e).

Figure 5

(a) Obtained damping constant ${\alpha }_{\mathrm{A}}$ and (b) $j_{\mathrm{s}}$ as a function of temperature.

Figure 6

Calculated $j_{\mathrm{c}}^{2\mathrm{D}}$/$\delta s$ as a function of $E_{\mathrm{F}}$ when ${\mathrm{\Delta }}_{\mathrm{ASO}}$ is 4, 5, 6, and 10 meV. The dotted vertical lines correspond to the carrier concentrations at 20 and 140 K in our sample (see Fig. S4 in SM [28]). We can see that $j_{\mathrm{c}}^{2\mathrm{D}}$/$\delta s$ can be negative for ${\mathrm{\Delta }}_{\mathrm{ASO}}=10\mathrm{meV}$, which may be the origin of the negative ${\lambda }_{\mathrm{IEE}}$ reported in Ref. [13].