Skip to main content
Nanomaterials

Nanomaterials

Nanomechanics weighs in

10 Apr 2004

Nano-scale devices can measure masses with a precision of one attogram, which is three orders of magnitude better than the previous record

We tend to associate mass with the quantity or size of an object. At the microscopic scale, however, mass measurements are a powerful tool that can provide information about the molecular and atomic composition of an object. While measurements of macroscopic masses traditionally rely on the fact that the gravitational force is directly proportional to the mass of the object, the Earth’s gravitational field is too weak to produce reliable measurements of the force in the case of molecular-scale objects.

A better way to measure the mass of a microscopic sample is to quantify the sample’s inertia as it is forced into motion. This is the principle behind mass spectroscopy, in which the trajectory of an ionized particle in a strong electromagnetic field provides a precise measure of the particle’s inertia, and therefore a measure of its mass. Mass spectroscopy is able to distinguish ionized particles that differ by a single atomic-mass unit – about 1.66 x 10-27kg, or 1.66 yoctograms. However, many researchers have pondered whether a less complex and more versatile measurement technique could be devised that has a similar level of sensitivity.

Harold Craighead and co-workers at Cornell University in the US have now built a mechanical device that suggests such an alternative could eventually be possible. Following the basic idea of placing particles in a strong field, they studied how the inertia of a nanoelectromechanical cantilever changed when it was loaded with small masses. As heavier objects were added the oscillation frequency of the device decreased, which enabled the Cornell team to measure the mass of a particle with a precision of one attogram (10-18 g). This is three orders of magnitude more sensitive than the previous record (R Ilic et al. 2004 J. Appl. Phys. at press).

Mass oscillation

Nanoelectromechanical devices (NEMS) are tiny structures that have mechanical degrees of freedom. They can be batch fabricated in a similar way to electronic chips, and a typical NEMS device used for mass measurements looks a bit like a diving board. This structure resonates at a frequency that is precisely defined by its stiffness, mass and geometry. Any additional mass that is added to the suspended portion of the device tends to slow down this oscillatory motion.

In 1992 researchers at Simon Fraser University in Canada demonstrated for the first time that such resonators can be used to detect masses less than 0.5 ng. Michael Roukes of the California Institute of Technology and co-workers developed NEMS devices further with the goal of detecting masses in the femtogram regime (see “Nanoelectromechanical systems face the future”). These studies showed that the sensitivity of NEMS devices could reach the level of a single atomic-mass unit. Recently Roukes and co-workers also demonstrated attogram-mass sensitivity, although their experiments were performed in an ultrahigh vacuum and at cryogenic temperatures (arXiv.org/abs/cond-mat/0402528).

A useful rule of thumb in the NEMS mass-detection business is that the smallest theoretically detectable mass roughly corresponds to a millionth of the mass of the cantilever. This reflects the fact that the smallest meaningful change in the resonance frequency of a nanomechanical device at room temperature is limited by thermal noise. While predictions of very high sensitivity using nanomechanical resonators have abounded in the literature, experimental NEMS that can operate under “normal” ambient conditions have clearly lagged behind the theory. It appears that approaching the theoretical limit is a very challenging experimental task.

One of the main reasons for this is the difficulty in reading-out the NEMS data. The frequency of nanomechanical resonators is generally determined by reflecting a laser off the resonator. As a result, reading-out the data becomes increasingly complicated as the widths of the resonators become smaller than the wavelength of visible light. While UV lasers with smaller wavelengths could be used, they tend to be less experimentally friendly than visible lasers and much more expensive.

The operation of mass-sensitive NEMS also needs to be optimized, which involves a difficult trade-off between several factors. For instance, thicker and stiffer resonators are less susceptible to thermal noise, but they are also more difficult to read-out because they tend to oscillate with smaller amplitudes and at higher frequencies. Resonators as thin as 50 nm and just several microns wide offer a reasonable compromise between the requirements of mass sensitivity and suitability for an optical read-out.

In 2003 the present authors took all these requirements into account and used a technique called focused ion milling to carve tiny slivers out of much larger commercially available silicon cantilevers (top figure). The result was a number of lightweight resonators that could undergo readily measurable changes in resonance frequency when they were loaded with a few femtograms of an organic material (2003 Appl. Phys. Lett. 82 2697-2699).

New order

Craighead and co-workers have now taken the optimization of NEMS for mass detection to new levels. Using their extensive experience with nanofabrication techniques, and by experimenting with various types of NEMS devices, the Cornell team has managed to improve the limit of mass detection by three orders of magnitude compared with our group’s previous world best. The resonators were made of silicon, were 4 µm long and 0.5 µm wide, and had a suspended mass of less than one nanogram (bottom figure). This allowed the researchers to achieve attogram-mass sensitivity, which is consistent with the rule of thumb of approximately one millionth of the mass of the resonator. Despite this minute mass, the devices were still large enough to accommodate a very convenient optical read-out.

Attogram-sensitive NEMS have immediate applications for novel chemical and biological sensors. Real-time trace analysis of highly hazardous agents, such as toxins, explosives and pathogens, may finally become possible without the need for expensive instrumentation, and this would allow hazardous agents to be detected before dangerous amounts of them accumulate. Such devices could also be used to detect large protein molecules, and to differentiate between individual viruses simply by weighing them. In short, an attogram-sensitive NEMS is an excellent addition to the toolkit of any scientist interested in studying interactions at the level of individual molecules.

To improve the sensitivity of mass-detecting NEMS still further, we will have to rely on even smaller resonators. But Craighead and co-workers have good reason to be optimistic. They have built a variety of NEMS devices with features as small as 20 nm, including the world’s smallest guitar (see Physics World December 2003 p3).

The next milestone in nanoelectromechanical mass detection is achieving zeptogram (10-21 g) sensitivity, which will prove whether nanomechanical mass spectroscopy is feasible. We anticipate this prize will attract even more researchers to join the mass-sensitive NEMS community in the next few years.

Copyright © 2024 by IOP Publishing Ltd and individual contributors