Control of Terahertz Emission by Ultrafast Spin-Charge Current Conversion at Rashba Interfaces

We show that a femtosecond spin-current pulse can generate terahertz (THz) transients at Rashba interfaces between two nonmagnetic materials. Our results unambiguously demonstrate the importance of the interface in this conversion process that we interpret in terms of the inverse Rashba Edelstein effect, in contrast to the THz emission in the bulk conversion process via the inverse spin-Hall effect. Furthermore, we show that at Rashba interfaces the THz-field amplitude can be controlled by the helicity of the light. The optical generation of electric photocurrents by these interfacial effects in the femtosecond regime will open up new opportunities in ultrafast spintronics.

Figure 1

Experimental schematics and sketch of THz emission in (a) bilayer samples and (b) trilayer samples, where the spin-to-charge-current conversion occurs at the Rashba interface between Ag and Bi. The MgO substrate is facing perpendicularly to the incident femtosecond laser pulse, and the magnetization $M$ is lying perpendicularly to the $x$ component of the THz-electric field $E_x$. The amplitude of $E_x$ generated from (a) the bilayer emitter is smaller than for (b) the trilayer emitter. (c) Illustration of the inverse Rashba Edelstein effect: The generation of a charge current carried by the interfacial states is due to a nonzero spin density induced by an ultrafast, pulsed spin-current injection.

Figure 2

(a) Comparison of a $\mathrm{CoFeB}/\mathrm{Ag}(2)/\mathrm{Bi}$ trilayer with bilayer control samples. An enhancement of the signal strength is observed when Bi is deposited onto Ag. The inset shows that the signal of $\mathrm{CoFeB}/\mathrm{Ag}(2)/\mathrm{Bi}$ is inverted when the magnetization $M$ is reversed. Please note that the signal amplitudes have been normalized to the $\mathrm{CoFeB}/\mathrm{Ag}(2)/\mathrm{Bi}$ sample. The raw data time traces can be found in Supplemental Material [30]; see Fig. S4. (b) Corresponding Fourier transform of the time-resolved measurements shown in (a). The inset shows the THz intensity of the control samples on a magnified scale.

Figure 3

(a) Comparison of THz amplitude for different sample stacks. The signal strength of the benchmark spin-to-charge converter Pt exceeds the one of the IREE-based trilayer system. An enhancement of the signal strength with respect to the control samples ($\mathrm{CoFeB}/\mathrm{Ag}/\mathrm{Al}$ denoted as Ag, $\mathrm{CoFeB}/\mathrm{Bi}$ denoted as Bi and $\mathrm{CoFeB}/\mathrm{Al}$ denoted as $ø$) is observed for the $\mathrm{Ag}/\mathrm{Bi}$ system. (b) Comparison of the temporal evolution of the THz-electric field $E_x$ in $x$-direction for three different Ag thicknesses. (c) The peak THz-electric field as a function of the rotation of the laser polarization. Please note that the signal amplitudes have been normalized to the $\mathrm{CoFeB}/\mathrm{Ag}(2)/\mathrm{Bi}$ sample in (a-c). The raw data time traces can be found in the Supplemental Material, Fig. S4. (d) Parametric plot of THz-electric field amplitude. The generated THz signal is mainly polarized along the $x$-direction.

Figure 4

(a) Illustration of $x$ and $y$ components of the THz-electric field. (b) Sketch of the helicity-dependent $y$ component of the THz amplitude. $E_y$ is oriented parallel to the magnetization $M$. The magnitude of $E_y$ is exaggerated with respect to $E_x$ for illustration purpose. (c) Time traces of the $y$ component of THz-electric field $E_y$ for right-circularity and left-circularly polarized laser light, respectively. The result was obtained from $\mathrm{CoFeB}/\mathrm{Ag}(2)/\mathrm{Bi}$. Please note that the signal amplitude has been normalized to the maximum of the THz-electric field in the $x$ direction.