Terahertz devices may spur advances in counterterrorism technology like airport scanners, agile-frequency devices, and satellite communications. They may also provide medical equipment with non-ionizing radiation that does not harm DNA. For instance, they could allow for high-resolution, soft-tissue scanners or replace x-rays for dental work. Currently, the best devices available use a femtosecond laser beam incident upon a conventional semiconductor such as p-type indium arsenide or zinc telluride (see Figure 1). By optically rectifying the laser beam, researchers can generate emissions near 1 or 2THz.
Our work has produced emissions in the 2–10THz frequency range. We collaborated with researchers at Vilnius University, Aveiro University, and the Jozef Stefan Institute1 to generate intense terahertz emission (see Figure 2) from an array of ferroelectric oxide nanotubes of lead zirconate-titanate (PZT). PZT is the most popular commercial actuator material, but has never been used for terahertz devices. To accomplish this task, we used an array of approximately 10 million PZT nanotubes, which are about 100μm in height, 1–4μm in outside diameter, and 40nm in wall thickness. The bottom 10% of each tube is embedded into a silicon substrate. This configuration maximizes the surface-to-volume ratio, and enables some coherent antenna effects.
Figure 1. (top) Femtosecond laser beams can generate terahertz emission from lead zirconate-titanate (PZT) nanotubes. (bottom) Scanning electron micrograph of the actual ferroelectric nanotube array.
Figure 2. PZT on silicon (7a) shows a greater response versus frequency than bare silicon (5a). a.u.: Arbitrary units.
In order to be a good terahertz emitter, substances need an extremely large charge density gradient near their surface. For example, indium arsenide normally has an accumulation layer, rather than a depletion layer. PZT is ideal in this respect because its surface charge density is typically 3×1021/cm3, whereas it is nearly a thousand times less only 20nm into the film. The second requirement is that the piezoelectric coupling constant d33 should be very large. PZT has one of the largest piezoelectric coefficients known in nature, making it attractive as a terahertz emitter. Prototype PZT terahertz devices are being manufactured by Cambridge Nano-Electronics.
PZT emitters have an additional advantage. Although the initial results were done with optical femtosecond laser excitation, future work aims to generate terahertz radiation in PZT merely by reversing the polarization. We can reverse polarity in these nanotubes with a 10V battery. This kind of nonstandard semiconductor heterostructure, as advocated by Ryzhii,2 has a wide bandgap oxide (Eg= 3.6eV) replacing the III-V or II-VI semiconductor.
Elimination of laser excitation would make terahertz emitters cheaper, smaller, and lighter. It would also permit field operation with hand-held devices. Studies from the group of Tonouchi3 make it clear that polarizations can be read from the terahertz spectra for other ferroelectric oxides like bismuth ferrite. Thus, a terahertz-accessed computer memory may also be possible.4 As next steps we plan to vary the diameters, shape (round or square), and separation of our nanotubes. In particular, we would like to see whether the nanotubes act as a coherent antenna array. We and others will also be making a serious effort to demonstrate terahertz emission without laser excitation. Because these PZT nanotubes switch polarization with application of dc voltages 10V across the wall of the tube,5,6 there is good reason to suspect that they will function as unconventional terahertz heterojunction emitters.
Earth Sciences and Physics
University of Cambridge
James Scott is a professor of ferroics at Cambridge University. An American educated at Harvard, he spent his early career at Bell Labs and the University of Colorado. In 2008 he was elected a Fellow of the Royal Society and in December received the Materials Research Society Medal for oxide devices.