SOUND is an integral part of the human existence. It propagates through the environment at various frequencies, allowing us to hear music from our radios and voices through our cell phones. But not all sound is audible to the human ear. Some acoustic waves have terahertz frequencies—that is, they oscillate at 1012 cycles per second. Sound in this range is too high for humans to hear, but researchers are finding that these high-frequency waves are exceedingly useful for scientific research.
In collaboration with Los Alamos National Laboratory and Nitronex Corporation, Livermore physicists Evan Reed and Michael Armstrong have discovered that by propagating acoustic waves through materials with different piezoelectric coefficients, they can transform waves at terahertz frequency into electromagnetic radiation of the same frequency. “We have developed a fundamentally new technological pathway to get into terahertz regimes,” says Reed. “We first predicted this phenomenon using molecular dynamics simulations.” With the help of an ultrafast laser and piezoelectric micrometer-thick heterostructures, they have become the first to observe the predicted behavior.
Reed and Armstrong, both of whom work in the Laboratory’s Science and Technology Principal Directorate, want to measure acoustic waves up to approximately 10 terahertz—the frequencies predicted to occur at the front of shock waves. Funded by Livermore’s Laboratory Directed Research and Development Program, their research is primarily geared toward developing high-resolution diagnostics for examining the shock and strain that materials undergo during laser experiments. Their new terahertz radiation generation and detection method is sparking interest outside the Laboratory as well. Within the semiconductor industry, it could serve as an improved, more direct approach for investigating the structural properties of thin films used to make computer chips.
Blazing a New Trail
For the experiments, Nitronex Corporation in Durham, North Carolina, supplied the Livermore team with silicon substrates coated with a layer of gallium nitride (GaN). The team sputter-coated each substrate with a 260- to 700-nanometer-thick layer of aluminum. An ultrafast laser then generates a 100-femtosecond-long pump pulse (where a femtosecond is one-quadrillionth of a second) with an 800-nanometer wavelength and approximately 1 millijoule of power and fires it at each substrate. The aluminum absorbed the energy from each pulse, causing that layer to heat and expand. This surface expansion created strain in the material, and the resulting acoustic wave propagated through the aluminum to the interface between the aluminum and GaN layers. At that boundary, material compression from the acoustic wave generated polarization currents through the piezoelectric effect, producing terahertz radiation, or light, which was then emitted from the material.
The team applied a standard technique known as electro-optic sampling to detect the radiation from a distance of a few millimeters. “Basically, we use a nonlinear optical process in which we write the terahertz radiation onto an optical pulse and then read the wave off the pulse,” says Reed. A brief terahertz signal produced by fast, nonlinear processes that occur when the laser pulse hits the aluminum layer denotes the time the acoustic wave was generated. After this wave transits the aluminum layer, it travels through the interface, generating terahertz radiation that provides the wave’s time history.
The laser pulse power is set low enough to be nondestructive to the material. Thus, to obtain an accurate estimate of time history, Reed and Armstrong had to average the signals produced by many pulses hitting one substrate. “Ultimately, we want the same results using a single shot,” says Armstrong.
X-ray ellipsometry, a technique that measures the polarization of light reflected from a surface, is a prominent method in the semiconductor industry for characterizing thin films. According to Armstrong, ellipsometry is an indirect method that models a thin film’s optical properties and then compares the results to actual data. Reed and Armstrong’s approach is a more direct way to determine layer thickness. “Characterizing thin films is just one application for this type of acoustic wave measurement,” says Armstrong.
Although the characterization method has promise for the semiconductor industry, it is first and foremost applicable to mission-related research at the Laboratory. It may enable scientists to better understand how materials act under extremely high pressures and how much pressure can be applied before a material is damaged. It may also provide a better way to evaluate strain and stress in materials used in shock and ramp-compression laser experiments, where pressure is applied incrementally to a sample. (See S&TR, June 2009, A Laser Look inside Planets.)
Beefing Up Security
Reed and Armstrong already have future experiments planned. “We tested this process using piezoelectric materials,” says Reed, “but we want to evaluate it with other materials as well.” By further exploring their technique, the researchers may find other applications for their research, demonstrating that “probing” into basic science can sometimes yield unexpected and fruitful results.
Key Words: acoustic wave, piezoelectric material, ramp compression, shock wave, semiconductor, sound wave, strain, terahertz frequency, ultrafast laser.
Lawrence Livermore National Laboratory
Privacy & Legal Notice | UCRL-TR-52000-09-9 | September 4, 2009