Researchers at the National Institute of Standards and Technology (NIST) and the University of Colorado at Boulder have taken an important step toward making nanoscale resonators that could be used in communications devices. The researchers have grown gallium-nitride nanowires that display properties much better suited to such uses than other nanostructures of similar sizes.
Resonators are an integral part of radio receivers and cell phones. Typically made of quartz crystals, these devices perform the critical function of picking out the frequency of the relevant radio signal from the cacophony of transmissions in the airwaves. While quartz crystals perform exceedingly well, they are bulky. “If you look at the chips in cell phones, resonators are huge compared to the rest of the circuitry,” says NIST researcher Kris Bertness, a coauthor of the Applied Physics Letters paper that outlines the new work. Crystal resonators take up areas of millimeters squared, while control electronics occupy square micrometers, she says.
Researchers have been trying to build micro- and nanoscale devices to replace quartz resonators. The problem is, as resonators shrink in size, they don’t work as well. In the past, researchers have made resonators using silicon nanostrings and carbon nanotubes; the nanowires grown by the NIST/Colorado team work at least 10 times better than any of these.
Resonators in radio receivers and cell phones work by vibrating within a narrow band of frequencies, vibrating the most at the band’s central frequency, called the resonance frequency. To determine how well a resonator works, engineers measure its quality factor, or Q factor. This depends on the width of this frequency band: the narrower it is, the higher the Q factor and the better a resonator is at filtering out a particular radio frequency from the neighboring signals. Quartz crystals have high Q factors, ranging from 10,000 to 1,000,000.
Efforts to build smaller resonators from silicon and carbon nanotubes have been hampered by simple physics: as the devices shrink, their Q factors go down. This is because at the nanoscale, even the smallest impurities or defects in the device surface affect its vibrations. Even gas molecules sticking to the surface can change the nanostructure’s mass, damping its vibrations and reducing the Q factor.
The new gallium-nitride nanowires, however, overcome some of the limitations that nanostructures face. Bertness and her colleagues grow the hexagonal nanowires on a silicon substrate using a cheap, easy method compatible with techniques used to manufacture microchips; replacing quartz resonators with nanowires grown this way might decrease manufacturing costs for cell phones. The nanowires have diameters between 30 and 500 nanometers, and lengths of 5 to 20 micrometers. The wires have no crystal defects, and they have very low chemical impurities, Bertness says. As a result, “they tend not to pick up a lot of junk from the environment, and they’re very smooth.” Because of this, they vibrate stably at their resonant frequencies and have high Q values.
To measure the effectiveness of the new nanowires, the researchers used a piezoelectric device–one that converts electrical signals into mechanical vibrations–to shake the nanowires at different frequencies. They then used a scanning electron microscope (SEM) to observe the wire’s vibration and calculate its resonance frequency and Q factor. The Q values ranged from 2,700 to 60,000–up to 10 times higher than measured for previous experimental nanoscale resonators.
The widely varying values are a result of limitations in the SEM measurement technique, Bertness says. Indeed, the Q values changed with different measurements even on the same wire. Bertness says that this is because the intense electron beam causes carbon molecules in the air to deposit on the nanowire, damping its vibrations.
Hong Tang, an electrical-engineering professor at Yale University, who is also working on nanoscale resonators, is skeptical about the researchers’ results. He says that combining a piezoelectric shaking with SEM detection artificially raises the Q value. Because SEM uses a tightly focused electron beam, he says, if the nanowire vibrates more than the beam spot size, the measurement of the wire’s displacement is not accurate. Tang’s guess is that the actual Q factors are probably lower than the reported values, although they are still likely to be higher than those reported for silicon-based nanowires, which have been around 1,000. He says that the researchers would have to use other measurement methods to verify their nanowires’ Q factors.
Bertness acknowledges the need for better measurements, adding that the nano resonator is far from practical right now. To be used in a cell-phone receiver, the nanowire will have to be driven by an electrical signal, not a mechanical shaking. Because gallium nitride is piezoelectric, the researchers believe that this should be possible, she says, and they are now trying to prove that theory.