Oliver B. Wright
Department of Applied Physics
Faculty of Engineering
Hokkaido University
Sapporo 060 Japan
Popular version of paper 2aPA13
Presented Tuesday morning, December 3, 1996
3rd Joint ASA/ASJ Meeting, Honolulu, Hawaii
Embargoed until December 3, 1996
The human ear can detect sounds up to about 20 kHz (20,000 cycles/second). Anything above this is termed ultrasonic. Bats' vocal chords and ears can cope with frequencies up to 200 kHz. They sense their environment acoustically by emitting sound and listening for the reflected echo, hopefully from their insect dinner.
We can mimic bats with man-made sound emitters and receivers, like those used in medical imaging and in looking for submarines. The same principle is adopted here but with vastly greater ultrasonic frequencies: the bats are put to shame by creating and picking up ultrasound as shrill as one terahertz or more (1 THz=one million million Hertz).
The detail we can visualize with sound depends on its wavelength, that is, the distance between adjacent crests of the wave. The higher the sound frequency the smaller is the wavelength. For bats this is as short as 2 mm, enough to locate an insect, whereas the best human ear does well to manage a paltry few centimetres.
With extremely brief laser pulses, lasting for only a picosecond or less (one million millionth part of a second), we can produce and detect terahertz ultrasound, which has a corresponding wavelength as short as a few nanometers (1 nanometer=10^-9 m). Just like a miniature sonar system we can use this 'nanoear' technology to eavesdrop on sound passing through atomic cracks, ultrathin films and minute nanostructures. Since the whole measurement typically only takes a few tens of picoseconds, apt names that have been coined for this method are ultrafast laser acoustics, laser picosecond acoustics, or picosecond ultrasonics.
What is done is to focus a beam of light consisting of a periodic train of ultrashort pulses onto an opaque sample. This light is converted into bursts of sound, an effect first noted at lower frequencies in 1881 by Alexander Graham Bell when he shone a modulated beam of sunlight onto selenium. The sudden heating produced by the light expands (or contracts) the solid in the region of illumination, and, just like a small piston, the hot volume of solid near the surface drives a sound wave into the material. This sound, reflected from defects or interfaces in the interior, can be picked up as echoes on arrival at the surface. As with sonar or medical echography, we can then assess the sub-surface structure.
Listening to such tiny sound waves at the surface requires an acutely sensitive detection combined with the same ultrashort time resolution as the sound production. The solution is to sense the sound with another beam of light pulses, identical in form to the beam that produced the sound but delayed in time. The delayed beam, made to travel further than the other before hitting the sample, is used to measure the sound that has returned to the surface. Introducing such delays is simple if both beams are derived from the same laser. A few millimeters of optical path difference correspond to a few tens of picoseconds.
It is then a matter of measuring the fluctuations in laser beam direction induced by the returning sound wave. Such angular deflection is caused by the tiny protrusion produced when the sound wave is reflected back to the surface of the material, very much like the bulge in a jelly that appears after being hit on the opposite side with a spoon. Vibrations of amplitude in the sub-nanometer range can be sensed. To help you imagine these minuscule vibrations better: if the illuminated region was the area of Australia, the scaled up displacement at the centre would only be about one meter. It is also possible to sense the sound by monitoring the ppm order changes in beam intensity.
To understand what determines the pitch of the echoes, the first step is to examine the sound-generation processes, the 'vocal chords' of the set-up. This depends on the light-sound conversion mechanisms particular to the solid in question. Metals and semiconductors show different behaviour depending on their electronic structure. Light absorption in solids occurs by knocking electrons, that are whizzing round the atoms of the solid, into higher energy levels. Depending on the colour of the light, the availability of the electrons to be knocked, and their motion afterwards, the sound production can be smeared out in space by electron diffusion or delayed in time by slow electron relaxation that inhibits sample heating. These processes tend to lower the frequency of the sound wave generated. However, depending on the material, the jumping or relaxing electrons may also cause higher, terahertz-frequency sound bursts to be emitted. Studying the pitch and harmonic content of the echoes therefore tells us how the electrons behave on very short time scales.
The echo pitch also depends on the reflection and scattering of sound from inhomogeneities inside the material. We can thus measure intrinsic properties of the solid, such as its ultrasonic attenuation. The time of echo arrival allows us the gauge the sample thickness or its sound velocity. The echo amplitude tells us about the acoustic hardness of the material from which the sound was reflected inside the solid.
Complex artificial nanoscale structures provide an ideal testing ground for this method. When jolted with light pulses such structures will vibrate at certain discrete frequencies determined by their shape and make-up. The resonances inside periodic semiconductor sandwich structures (known as 'superlattices'), with layer thicknesses approaching atomic dimensions is one example that has been investigated. Another is the ringing of microscopic gold ingots. Even nanobells - nicknamed, of course, Little Ben - could soon be tolling the virtues of the kingdom of the ultrasmall with ultrahigh-pitched sound.
Some references
Grahn, H. T., Maris, H. J. and Tauc, J., 'Picosecond ultrasonics,' IEEE J. Quantum Electron. 25, pp. 2562-2569 (1989)
Wright, O. B. and Kawashima, K., 'Coherent phonon detection from ultrafast surface vibrations,' Phys. Rev. Lett. 69, 1668-1671 (1992)
Akmanov, A. and Gusev, V. E., 'Laser excitation of ultrashort acoustic pulses: New possibilities in solid-state spectroscopy, diagnostics of fast processes, and nonlinear acoustics,' Sov. Phys. Usp. 35, pp. 153-191, (1992)
Ippen, E. P. and Shank, C. V., 'Techniques for measurement,' in Ultrashort Light Pulses, edited by S. L. Shapiro (Springer, Berlin) Ch. 3. pp. 83-122 (1984)