Angular Dependence of the Spin Photocurrent in a $\mathrm{Co}\text{−}\mathrm{Fe}\text{−}\mathrm{B}/\mathrm{MgO}/n\text{−}i\text{−}p$ GaAs Quantum-Well Structure

We evidence at room temperature the detection of photogenerated spin currents by using a magnetic electrode without the need of an external magnetic field. The device is based on a semiconductor $(\mathrm{Al},\mathrm{Ga})\mathrm{As}/\mathrm{GaAs}$ quantum well embedded in a $p\text{−}i\text{−}n$ junction. The spin filtering is performed owing to a $\mathrm{Co}\text{−}\mathrm{Fe}\text{−}\mathrm{B}/\mathrm{MgO}$ electrode with in-plane magnetization. We observe a helicity-dependent photocurrent when the device is excited under oblique incidence with circularly polarized light. The helicity-dependent photocurrent is explored as a function of the incident and azimuth angles of the incoming light wave vector with respect to the magnetization direction of the magnetic electrode. The results are interpreted as a consequence of the photogenerated average electron spin under oblique incidence in a quantum well governed by optical selection rules involving electron-heavy-hole and electron-light-hole transitions. A systematic study of the helicity asymmetry as a function of the photon energy and applied bias is performed. It demonstrates that this asymmetry is at its maximum close to the GaAs quantum well and (Al,Ga)As bulk optical transitions. The asymmetry can be controlled by an external bias on the structure. Finally, we show that this asymmetry decreases when the temperature increases.

Figure 1

Sample structure and photocurrent measurement. (a) The schematic diagram of the cross section of the device. The black bold arrows in Co-Fe-B denote the orientation of the remanent magnetic moment. (b) HRTEM image of the $\mathrm{MgO}/\mathrm{Co}\text{−}\mathrm{Fe}\text{−}\mathrm{B}/\mathrm{Ta}$ spin detector on (Al,Ga)As-based LED structure. (c) Schematic of the difference of photocurrent between right and left circularly polarized light with oblique incidence thanks to a selective spin filtering by the $\mathrm{Co}\text{−}\mathrm{Fe}\text{−}\mathrm{B}/\mathrm{MgO}$ layer. The red and blue horizontal arrows correspond to $S_x$, the projection along $x$ of the photogenerated average spin $S$ for ${\sigma }^{-}$ and ${\sigma }^{+}$, respectively. (d) Schematic diagram of the electrical measurement under bias.

Figure 2

Total photocurrents and helicity asymmetry as a function of the azimuth and incidence angle. (a) The total photoinduced direct currents and (b) the helicity asymmetry of the photocurrents changing with the azimuth angle for the incident angle $\theta =30°$ under three different bias voltages at 300 K. The solid lines correspond to cosine fitting. (c) The total photoinduced currents and (d) the helicity asymmetry changing with incidence angle with the azimuth angle $\alpha =0°$ under three different bias voltages at 300 K. The solid lines correspond to the sine fitting. The insets in (b) and (d) illustrate the definition of the azimuth angle and the incident angle, respectively.

Figure 3

Bias voltage dependence of the spectra of total photocurrents and helicity asymmetry at 80 K. (a) Spectra of the total photoinduced direct currents. Inset: Enlargement of the quantum-well transition. (b) Spectra of the helicity asymmetry at different bias voltages at 80 K. Note that the spectra in (a) and (b) are intentionally shifted for clarity, and the color coding is the same for (a) and (b). (c) The helicity asymmetry of the photocurrent and the total photoinduced direct current as a function of bias voltage corresponding to the transition of the $i\text{−}\mathrm{GaAs}$ QW, respectively. The azimuth and incident angles in these experiments are fixed to be 0° and 15°, respectively.

Figure 4

Schematic diagrams of the energy band under bias. (a) Tilted bands under a negative-bias voltage with electron’s killing channel (radiative recombination) almost closed. (b) Almost flat band conditions under a positive-bias voltage with the electron’s killing channel (radiative recombination) open.

Figure 5

Temperature dependence of the spectra of total photocurrents and helicity asymmetry under -0.163-V bias. (a) Spectra of the common photoinduced direct currents, whose color coding is the same as Fig. 5. The inset is a schematic diagram of the cryostat for temperature control. (b) Spectra of the helicity asymmetry at different temperatures. Please note that the spectra are intentionally shifted for clarity. (c) The amplitude of helicity asymmetry as a function of the temperature corresponding to the transition of the $i\text{−}\mathrm{GaAs}$ QW. The bias voltage and the azimuth and incident angles in these experiments are fixed to be $-0.163\mathrm{V}$, 0°, and 15°, respectively.