Relying on standard materials and manufacturing techniques would make it much easier to incorporate a nanosensor with the electronics inside a handheld device, says chemist Mark Reed of Yale University, co-author of a report in this week's Nature detailing the technology. "This has the ability to scale in power and cost, just like regular electronics," he says.
Reed and his colleagues coated their 30-nanometer-wide wires in antibodies or other biological molecules capable of latching onto certain proteins. These receptors plucked their matching proteins from a solution washed over the sensor, which detected the change because the electric charges on the amassed proteins easily disrupted the current flowing through the wires. Reed likens the effect to the way that stepping on a flimsy garden hose (but not a tough fire hose) would block its flow.
The device detected as few as 30,000 free-floating proteins in a cubic millimeter of fluid in a matter of seconds, which Reed says compares favorably with other nanowire sensors. It also recognized immune cells by the acid they emit when they bind to antibodies, the group reports. Co-author and Yale bioengineer Tarek Fahmy adds, "There's no other way to do this rapidly [with] high throughput. This is what we're really excited about."
The researchers carved their device from a high quality wafer of insulating material topped with a thin layer of silicon. They used standard techniques to build a stencil (shaped like the device they wanted), which they placed on the wafer. They then poured a solvent on top that etched away the exposed silicon.
Normally such a process would leave relatively thick wires, so to reduce the wires to nanosize they removed the stencil and let the etching continue. Reed says the combination of a good wafer and a slow-acting solvent gave them smoother, more precise nanowires than other groups have achieved by etching.
How long for a version you can hold in your hand? Reed will not speculate, but he says, "this is something I will see in my lifetime."