Photoexcited carrier dynamics of thin film ${\mathrm{Cd}}_3{\mathrm{As}}_2$ grown on a $\mathrm{GaAs}\left(111\right)B$ substrate by molecular beam epitaxy

The topological Dirac semimetal ${\mathrm{Cd}}_3{\mathrm{As}}_2$ has drawn great attention for its novel physics and technical applications in optoelectronic devices operating in the infrared and THz regimes. Exploring the photoexcited carrier dynamics in ${\mathrm{Cd}}_3{\mathrm{As}}_2$ is helpful to understand the transient carrier occupation and cooling processes that are closely associated with its unconventional band characteristic. Here, the photoexcited carrier dynamics in the ${\mathrm{Cd}}_3{\mathrm{As}}_2$ thin film epitaxially grown on the $\mathrm{GaAs}\left(111\right)B$ substrate was investigated by employing the time-resolved midinfrared transient reflectance measurements, along with a theoretical modeling concerning the electron-polar optical ($e$-PO) phonon interaction. The measured photoexcited electron relaxation time of ${\mathrm{Cd}}_3{\mathrm{As}}_2$ increases with the probe wavelength, and increases with temperature at temperatures higher than 30 K. Theoretical modeling under the framework of the two-temperature model can give qualitatively good description on both the temperature and probe wavelength dependence of the carrier relaxation time. Additionally, theoretical modeling shows that charge screening of the interaction of electrons with polar optical phonons can greatly reduce the hot carrier relaxation rate. The measured transient reflectance analyses indicate that the hot phonon effect plays a negligible role on the carrier relaxation in ${\mathrm{Cd}}_3{\mathrm{As}}_2$. Our findings provide further insight into the photoexcited carrier dynamics in ${\mathrm{Cd}}_3{\mathrm{As}}_2$ and valuable reference for developing its high-performance infrared optoelectronic devices.

Figure 1

The transient reflectance $\mathrm{\Delta }R/R$ response of 40-nm ${\mathrm{Cd}}_3{\mathrm{As}}_2$ film (a) measured at 4.5 K and a fixed probe wavelength of 4000 nm, by varying pump power from 18 to 88 μW and (b) measured at 300 K and a fixed probe wavelength of 4000 nm, by varying pump power from 18 to 88 μW. The insets of both figures show a saturated behavior for the amplitude of $\mathrm{\Delta }R/R$ response. (c) Measured at a fixed probe wavelength of 4000 nm by varying $T_{\mathrm{L}}$ from 4.5 to 300 K, under the pump power of 35 μW. (d) Measured at a fixed temperature of 4.5 K by varying probe wavelength from 3000 to 4000 nm, under the pump power of 35 μW. The inset shows that the absolute peak value of $\mathrm{\Delta }R/R$ increases with probe wavelength. The dash curves are guides to the eyes. All the $\mathrm{\Delta }R/R$ responses were normalized with respect to their minimum values. The data are plotted with a vertical offset interval of 0.1 for clearer view along with their fitting curves (solid lines).

Figure 2

(a) The deduced effective falling time ${\tau }_{\mathrm{fall}}$ of $\mathrm{\Delta }R/R$ as a function of $T_{\mathrm{L}}$ for the three different probe wavelengths, measured under a fixed pump power of 35 μW. (b) The deduced ${\tau }_{\mathrm{fall}}$ as a function of pump power, measured with probe wavelength of 4000 nm at 4.5 and 300 K, respectively. (c) The extracted decay constants ${\tau }_{\mathrm{decay}}$ as a function of $T_{\mathrm{L}}$ for the three different probe wavelengths, under a fixed pump power of 35 μW. The dashed lines are guides to the eyes. The inset shows the detailed variation of ${\tau }_{\mathrm{decay}}$ probed at 4000 nm at $T_{\mathrm{L}}<120\mathrm{K}$.

Figure 3

(a) The calculated decay curves of $T_{\mathrm{e}}$ (namely $T_{\mathrm{e}}-T_{\mathrm{L}}$) with (solid curves) and without (dashed curves) the static charge screening on the $e$-PO interaction at different $T_{\mathrm{L}}$. The initial $T_{\mathrm{e}}=1600\mathrm{K}$. The calculated dynamic decay of $T_{\mathrm{e}}$ within $t<10\mathrm{ps}$ when including the charge screening and that within $t<5\mathrm{ps}$ when excluding the charge screening, can be approximately fitted with straight lines. The slopes are shown in the inset. (b) The calculated $\mathrm{\Delta }R/R$ responses (curves) for the three different probe wavelengths based on the decay process of $T_{\mathrm{e}}$ at $T_{\mathrm{L}}=300\mathrm{K}$, namely the gray solid curve in (a), and the measured $\mathrm{\Delta }R/R$ (circles) at $T_{\mathrm{L}}=300\mathrm{K}$ under the pump power of 35 μW. Here all the $\mathrm{\Delta }R/R$ are normalized with respect to their minimum values at zero decay time.

Figure 4

The raw $\mathrm{\Delta }R/R$ response data comparison under the pump power of 18 μW (red curve) and 88 μW (blue curve). The $T_{\mathrm{L}}$ is at 4.5 K (a) and 300 K (b). The measured $\mathrm{\Delta }R/R$ response (the dash curve) under the pump power of 88 μW is shifted to the solid curve in order to compare with the decay part of $\mathrm{\Delta }R/R$ measured at 18 μW. Starting from $\mathrm{\Delta }R/R\approx -0.025$, the decay rate of $\mathrm{\Delta }R/R$ under both the high- and low-power pumping is almost the same at 4.5 K (a), while the decay rate of $\mathrm{\Delta }R/R$ under high pumping is slightly larger than that of low pumping at 300 K (b).