Device for acoustic pulse echo experiments in the subterahertz range

We present experimental results on the light-to-sound transduction by a superlattice embedded in an optical microcavity, using picosecond ultrasonics. Experiments on the ability to generate and detect phonons at 0.3 and 0.6 THz were performed at low temperature. Finally, we demonstrate that such a device can detect the acoustic burst that it has generated and which has propagated through the substrate, been reflected on the backside of the substrate, and has returned to the device. We study the efficiency of this detection by varying independently the laser wavelength of the pump and probe, by means of asynchronous optical sampling.

Figure 1

(a) Scheme of the sample B, constituted by the acoustic SL inside an optical microcavity, whose mirrors are the distributed Bragg reflectors (DBR). The laser beams can be focused on the structure, which is the usual experimental configuration (1), or in acoustic transmission geometry with the pump beam on the back of the substrate (2). (b) Reflectivity of the sample versus light wavelength: measurements at room temperature (red solid line with a spectrometer, red open symbols $\square$ with the femtosecond laser) and at 15 K (shifted for clarity) (blue open symbols $\circ$ with the femtosecond laser). The calculation done with the real structure inside the cavity is in good agreement (solid lines).

Figure 2

In the plain SL (sample A), transient reflectivity corresponding to the first echo as a function of time delay (${\mathrm{t}}_0\sim 144\mathrm{ns}$ is the time needed for the acoustic burst to make a round trip through the substrate, as sketched in the inset). The dotted curve is a low pass filter ($<20$ GHz); the red curve is a band pass filter about 400 GHz.

Figure 3

FT amplitude of the change of reflectivity, compared to the SL dispersion curve (blue line, right scale), in transmission geometry [configuration (2)] as sketched in (a). (b): in sample A (776 nm) (c): in sample B (754 nm). In sample B, $q=0$ modes detection is enhanced.

Figure 4

In sample B, (a) FT amplitude of the change of reflectivity just after time delay zero (red line), compared to the SL dispersion curve (blue line, right scale). The laser wavelength is set to 750 nm; (b) $q=0$ and $q=2k$ mode amplitude at 250 and 290 GHz as a function of the laser wavelength. The reflectivity curve (right scale) is reminded [experimental points ($\circ$) and calculation (solid curve)].

Figure 5

In sample B, (a) transient reflectivity $\mathrm{Re}(\mathrm{\Delta }r/r$) measured by the $X$ and $Y$ lock-in amplifier components (the $X$ component is in phase with the first echo detected about a time delay of 3500 ps, after a propagation time of 152.7 ns). The laser wavelength is 758 nm (cavity mode). The arrow indicates the first echo. Inset: zoom of the echo at another laser wavelength (746 nm) with a band pass filter between 288 and 295 GHz corresponding to $q=0$ mode frequency. (b) Windowed Fourier transform amplitude over 500 ps interval of the signal, as a function of the time delay. The mode $q=0$ at 293 GHz dominates the $q=2k$ modes (Brillouin at 48 GHz and 253 GHz), at the beginning but also in the first echo appearing after 3 ns.

Figure 6

In sample B, windowed Fourier transform amplitude over 200 ps interval of the signal, as a function of the time. The red line indicates the expected arrival time.

Figure 7

In sample B, $q=0$ and $q=2k$ mode amplitude at 253 and 290 GHz as a function of the laser wavelength (the laser intensity was set to 13 mW). The reflectivity curve (right scale) is reminded [experimental points ($\circ$) and calculation (solid curve)].

Figure 8

In sample B, FT amplitude of the change of reflectivity due to the echo (real part) for $q=0$ at 293 GHz and $q=2k$ lower mode ($\times 3$), as a function of the pump laser wavelength. ${\lambda }_{\text{probe}}=744\mathrm{nm}$. In dotted line, electric field square modulus calculated inside the cavity (arb. units), and in dotted dashed line, the corresponding calculated reflectivity (right scale).

Figure 9

In sample B, FT amplitude of the change of reflectivity due to the echo (real part) for $q=0$ at 293 GHz as a function of the probe laser wavelength. The pump beam was set at a wavelength ${\lambda }_{\text{pump}}=756.7\mathrm{nm}$ (red square) or ${\lambda }_{\text{pump}}=745\mathrm{nm}$ (open blue square, $\times 1.8$). In dotted dashed line, the calculated reflectivity (right scale).

Figure 10

In sample B, (a) FT amplitude of the $q=0$ and $q=2k$ components of the imaginary part of the change of reflectivity due to the echo (open symbols) as a function of the probe laser wavelength. The pump beam was set at a wavelength ${\lambda }_{\text{pump}}=756.7\mathrm{nm}$. Results for the real part ($q=0$) are rescaled (red square) for comparison; the maxima are indicated by dotted vertical lines. Right scale: measured reflectivity. (b) Results of a simple model for the real and imaginary parts of the FT amplitude of the reflectivity change. Right scale: reflectivity curve (calculations in solid line, measurements in symbols).