Nanoelectromechanical (NEM) resonators, which are nanosized slabs of material vibrating at a certain frequency, can weigh tiny particles with extraordinary sensitivity. They do this by measuring the change in frequency of the resonator when an object is placed on top of it. Indeed, these cantilevers can weigh extremely light objects, including single molecules and even single protons. A resonator is limited, however, by its ability to resolve background or thermal fluctuations that add unwanted noise to the measurement signal. Researchers have always thought that this noise was intrinsic to a NEM and could not be avoided, but the practical limit does appear to be much larger than theory predicts.

To better understand this phenomenon, a team led by Sébastien Hentz of CEA LETI and the University of Grenoble-Alpes has tested a monocrystalline silicon resonator by making it vibrate at frequencies other than its natural vibrating frequency, but close to it. Using this technique, the researchers were able to show that, as well as the expected theoretical background noise signal, there was another parasitic effect that limits the mass-detection resolution of the system and that its resonant frequency varies slightly.

Extreme sensitivity

“This fluctuation in the resonance frequency comes from the extreme sensitivity of these systems – they are able to detect tiny masses and forces but they are also exceptionally sensitive to the slightest variation in temperature in their surroundings or, for instance, to molecules sliding along their surfaces,” explains Hentz. “We can no longer ignore these effects at such small length scales because they limit the performance of nanoresonators. For example, a tiny change in external temperature can alter the frequency at which a resonator vibrates and these variations can occur rapidly and in a random fashion.”

The researchers say that they can now use their experimental technique, which has been patented, to calculate whether losses in mass-detection resolution come from the intrinsic limits of the system itself or from external fluctuations. Once the sources of these losses have been identified, they can be corrected for, says Hentz.

The team, which includes researchers from the EPFL in Switzerland, the Indian Institute of Sciences in Bangalore and Caltech in the US, says it is now busy tying to better understand where these fluctuations actually come from in an effort to improve the overall mass-detection performance of nanoresonators.

The research is detailed in Nature Nanotechnology.