Helicity-dependent photocurrent in a $\mathbf{B}{\mathbf{i}}_{\mathbf{2}}\mathbf{S}{\mathbf{e}}_{\mathbf{3}}$ thin film probed by terahertz emission spectroscopy

We investigate photocarrier dynamics in a $\mathrm{B}{\mathrm{i}}_2\mathrm{S}{\mathrm{e}}_3$ thin film via terahertz (THz) emission spectroscopy. We report amplitude and phase modulations of the emitted THz wave under the variations in a polarization of an input laser beam and sample azimuth. Photon-helicity-dependent responses are manifested in both time and frequency domains, and azimuth dependence of such circular anisotropy exhibits a clear threefold periodicity. We analyze these results based on the symmetry of the bulk and the surface of $\mathrm{B}{\mathrm{i}}_2\mathrm{S}{\mathrm{e}}_3$ and demonstrate that the observed helicity-dependent photocurrent originates from the circular photon drag effect, namely, linear and angular momentum transfers from photons to photocarriers.

Figure 1

(a) Schematic of the THz emission spectroscopy experiment. (b) X-ray-diffraction pattern of the $\mathrm{B}{\mathrm{i}}_2\mathrm{S}{\mathrm{e}}_3$ film. The inset shows an enlarged (006) diffraction peak. (c) and (d) THz waves in time and frequency domains, respectively, emitted from the $\mathrm{B}{\mathrm{i}}_2\mathrm{S}{\mathrm{e}}_3$ film at representative azimuth angles (Φ) under illumination of linearly polarized light. The inset in (d) displays THz-wave magnitudes averaged for the shaded regions in time (3.4–3.87-ps) and frequency (0.23–0.46-THz) domains as a function of Φ where each of them is described as blank circle (left) and filled triangle (right) symbols, respectively. A solid line indicates the fitting result.

Figure 2

(a) and (b) THz waves in time and frequency domains, respectively, under illumination of left-handed (black) and right-handed (red) circularly polarized light where the azimuth (Φ) is set to 30°. (c) THz-wave amplitudes averaged for the shaded regions in time (blank circle) and frequency (filled square) domains as a function of the quarter-wave-plate angle (φ). The symbol and solid line indicate the experimental data and fitting result, respectively.

Figure 3

(a)–(c) THz-wave time profiles emitted from the $\mathrm{B}{\mathrm{i}}_2\mathrm{S}{\mathrm{e}}_3$ film excited by linearly and circularly polarized light when Φ equals 0°, 90°, 180°, and 270°. The RC, L, and LC indicate the right-handed circularly, linearly, and left-handed circularly polarized light injections, respectively. (d)–(f) THz-wave amplitudes averaged for shaded regions in (a)–(c), respectively, as a function of sample azimuth angle Φ. Here, the symbol and the line indicate the experimental data and the fitting result, respectively. The fitting is performed using onefold and threefold periodic functions together with a constant offset, and fitting curves except for the onefold periodic component are drawn in the bottom part of each figure.

Figure 4

(a) Difference between THz amplitudes (symbol) under the pumping with LC and RC laser pulses. The fitting curve is shown with a solid line. (b) THz amplitudes versus the power of the incident laser beam. The helicity-dependent term is displayed together with the result obtained with a linearly polarized light injection.

Figure 5

Magnitudes of susceptibility tensor components. All the values of linear contributions ($L_2, l_4$, and $l_7$) are normalized by the circular photon drag effect (CPDE) contribution $c_7$. $t_s$ and $t_p$ are transmission Fresnel coefficients for $s$- and $p$-polarized light, respectively. The circle (square) denotes the result obtained by assuming a circularly (elliptically) polarized beam in the $\mathrm{B}{\mathrm{i}}_2\mathrm{S}{\mathrm{e}}_3$ film. The uncertainty in the fitting functions for the results in Fig. 3 leads to finite errors in the magnitude of susceptibility tensor components.