Finding ways to make very sensitive measurements of biological molecules has long been of interest to Scott Manalis's group at the Massachusetts Institute of Technology in Cambridge. But they didn't want to use fluorescent or radioactive labelling techniques that require multistep sample preparation methods and relatively large sample volumes. “We wanted to develop methods for label-free detection that could be as sensitive as fluorescence,” says Manalis. One way of detecting molecules is by their mass. But how do you weigh really tiny things?

Nanoscale mechanical resonators can measure the mass of particles weighing as little as several zeptograms (a zeptogram is 10−21 grams). These instruments are designed to vibrate at a given frequency, known as the resonant frequency. When a molecule lands on the resonator's surface, the resonant frequency changes by an amount that correlates to the mass of the molecule. However, these instruments do not work as well when they are placed in a solution, because fluids dampen the mechanical vibrations. This often limits their biological applications, because these frequently require fluid.

In 2002, Manalis and his team came up with an approach to overcome the problem. Why not try putting the fluid in microchannels inside the resonator? Thomas Burg, a graduate student in Manalis's lab at the time, set about making a prototype. Although it worked, it didn't take very sensitive or reliable measurements. This was partly because Manalis and Burg had not been able to make it in such a way that it could be contained within a vacuum, one of the requirements for measuring really small weights. “A well-known challenge in the MEMs [micro-electro-mechanical systems] field is packaging,” says Manalis. “In many cases, the details of packaging are known only by the graduate student who developed the process. Once the student graduates, it can be difficult to advance the project beyond the initial demonstration.”

To prevent this from happening, Manalis joined forces with the Santa Barbara-based labs of Innovative Micro Technology (IMT), a Californian company with expertise in manufacturing MEMs. “Collaborating with IMT has given us access to state-of-the-art packaging and microfluidic processes that have allowed us to develop highly robust and sensitive mass detectors,” says Manalis. Burg, who had by this time almost completed his thesis, decided to continue on the project as a postdoc. After about three years' further development, the group had a vacuum-packaged resonator that could weigh individual nanoparticles, single bacteria and protein monolayers in solution with a resolution of 10−15 grams (see page 1066).

In addition to weighing particles that bind to the sides of the channels, the instrument can measure samples as they flow through it. The idea of designing the instrument in this way came about almost by accident. “Once the device had been designed with three-micrometre-tall channels, it occurred to us that we could flow bacteria through them and weigh individual cells one by one,” says Manalis. Using this flow-through mode, they could weigh a wide variety of particles, as ways to bind particles to the instrument's surface were not needed.

The instrument works better than Manalis and his co-workers could have ever anticipated, but the proof lies in what it will be able to do. “We know we can weigh nanoparticles and cells. Now we need to focus on useful applications,” says Manalis.