A tunable carbon nanotube resonator

Abstract

Nanoelectromechanical systems (NEMS) have recently been the subjectof much exciting research. They have been proposed for use in various applications such as mass and force detection, RF processing, and investigating quantum effects in the mechanical motion of resonators. Attempts to increase sensitivity for these applications has led to further and further miniaturization of the mechanical devices. When their size reaches the range of hundreds of nanometers, these devices have active masses in the hundreds of the femtograms and operational frequencies in the GHz. An ultimatelimit to this miniaturization is a mechanical resonator based on a single molecule. Such a resonator should not only be able to push the limits of the measurements sensitivities, but can also probe decrease of the quality factor values with size that has so far been attributed to the increase of the surface-to-volume ratios in these resonators. Carbon nanotubes (CNTs), thin tubes of graphene, are light, stiff, strong, and electrically active, which makes them a perfect candidate for a such a NEMS structure.

By employing a capacitive actuation and detection technique, we investigate the performance of a resonator based on a doubly-clamped, suspended CNT in a transistor geometry. We excite vibrations by applying an AC driving voltage to the gate electrode, and we detect them by measuring the current through the CNT device. Controlling the CNT's tension, by applying a downward DC force with a DC voltage on the gate electrodes, enables us to tune the resonant frequency, resulting in the first tunable and self-detecting carbon nanotube resonator.

This setup also allows us to probe the loss mechanisms in these small structures. We systematically study correlation of the quality factor with each of the device characteristics, including electrical resistance, fabrication geometry, and resonant mode harmonic number. We also study dependence of the quality factor on the experimental knobs, such as pressure, temperature, DC gate voltage, and AC driving voltage. We find that the quality factors in CNTs continue the trend previously established by NEMS, and that several dissipation mechanisms must be responsible for losses in this system. We identify coupling to the environment, the thermoelastic effect, and surface-related losses as the three key mechanisms.