Effect of picosecond strain pulses on thin layers of the ferromagnetic semiconductor (Ga,Mn)(As,P)

The effect of picosecond acoustic strain pulses (ps-ASP) on a thin layer of (Ga,Mn)As codoped with phosphorus was probed using magneto-optical Kerr effect (MOKE). A transient MOKE signal followed by low-amplitude oscillations was evidenced, with a strong dependence on applied magnetic field, temperature, and ps-ASP amplitude. Careful interferometric measurement of the layer’s thickness variation induced by the ps-ASP allowed us to model very accurately the resulting signal, and interpret it as the strain modulated reflectivity (differing for ${\sigma }_{\pm }$ probe polarizations), independently from dynamic magnetization effects.

Figure 1

(Color online) (a) Schematics of the setup with the probe beam on Al for the acoustics characterization, or directly on GaMnAsP for MOKE measurements. Definitions of the reflection, and transmission coefficients. (b) Interferometry characterization of the acoustic strain pulse ($T=58\text{K}$, ${\lambda }_{probe}=780\text{nm}$, and $I=280\mu \text{J}{\text{cm}}^{-2}$): surface displacement $\text{Im}\left(\Delta r/r\right)$ and corresponding strain incident on the GaMnAsP layer. Inset: strain spectrum.

Figure 2

(Color online) (a) Schematics of the optical bridge placed on the path of the reflected beam and reflectivity seen by one of the diodes ($B=63\text{G}$, $I=280\mu \text{J}{\text{cm}}^{-2}$, ${\lambda }_{probe}=780\text{nm}$, and $T=80\text{K}$). (b) Transient Kerr signal under opposite applied fields ($I=280\mu \text{J}{\text{cm}}^{-2}$, ${\lambda }_{probe}=705\text{nm}$, and $T=58\text{K}$). Curves were spaced equally for clarity and $t=0$ set arbitrarily.

Figure 3

(Color online) (a) Transient MOKE signal with varying $B_{\perp }$. ($T=58\text{K}$, $I=280\mu \text{J}{\text{cm}}^{-2}$, and $\lambda =705\text{nm}$). Curves were spaced equally for clarity, and $t=0$ set arbitrarily. (b) Transient and static MOKE versus applied field.

Figure 4

(Color online) Varying pump fluences, $I=60–280\mu \text{J}{\text{cm}}^{-2}$. $T=54\pm 4\text{K}$, ${\lambda }_{probe}=705\text{nm}$, and $t=0$ set arbitrarily. (a) Transient MOKE signal at $B=63\text{G}$. Inset: peak amplitude versus fluence. (b) GaMnAsP layer thickness variation estimated from the Sagnac interferometry data.

Figure 5

(Color online) [(a), (b), and (c)] Normalized simulated MOKE signal using a multilayer reflectivity model. See within text for description of the different terms. (d) Comparison of the full simulated MOKE signal with the experimental data ($I=280\mu \text{J}{\text{cm}}^{-2}$, $T=58\text{K}$, ${\lambda }_{probe}=705\text{nm}$, and $B=-63\text{G}$).