Cool idea for sensor technology

From Technology Review:

Cheap Chemical Sensors
Electronic "noses" made from printed electronics could detect toxic chemicals inexpensively.

Technology already exists that can sniff out chemicals in the air and water -- but the detecting devices are expensive, limiting their use. Now Vivek Subramanian, electrical engineering professor at the University of California, Berkeley, has made arrays of sensors cheap enough that they could be widely distributed for monitoring toxins in the environment.

The goal is to "identify environmental problems before they become severe, then react to them," says Subramanian. "One of the major requirements, if we want to do this, is ultra-low cost," he says. Subramanian makes his array of inexpensive chemical sensors using organic semiconductors and inkjet printing technology. The first generation of his devices, which would still rely on costly silicon-based technology to process signals from the sensors, would run about 30 cents a piece, Subramanian estimates. That's a bargain compared with several hundred dollars for today's sensors, he says. Subramanian reported on his work at the Materials Research Society meeting in Boston this week.

Organic transistors tend to degrade, especially when exposed to air, chemicals, or water. Yet this reactivity can also make them good sensors. For one thing, different chemicals affect the rate at which the sensors degrade. Subramanian's innovation is to use an array of different organic semiconductors, each responding slightly differently to different chemicals. The signals from this array then create a distinctive pattern -- a sort of electronic fingerprint of a particular chemical.

A bit more:

Subramanian prints the different kinds of organic semiconductors in arrays using the multiple nozzles of a inkjet printer. So far, he has only produced arrays of transistors using five different semiconductors -- still more than enough to detect the difference between good and spoiled wine, for instance. His arrays can also detect different types of organic solvents in industrial processes.

I love it! The downside is that the sensors are degrading so they will need to be replaced but I am betting that this is not a frequent event. Once/year for low concentrations maybe. Using a C.O.T.S. Ink-Jet printer to make the organic transistors is a stroke of genius. Very cool! (C.O.T.S. is the mantra of the quick-and-dirty engineering crowd. It stands for something that is Commercial and Off-The-Shelf -- ie: readily available and cheap.)