The Plastic Processor
Take a bow, flexible chip. This week at the International Solid-State Circuits Conference, in San Francisco, European researchers will introduce the world’s first microprocessor made with organic semiconductors. The 4000-transistor, 8-bit logic circuit has the processing power of only a 1970s-era silicon model, but it has a key advantage—it can bend. The device’s designers say the chip could lead the way to cheaper flexible displays and sensors. Wrapped around pipes, for example, sheets of sensors with these processors could record average water pressure, and wrapped around food and pharmaceuticals, they might indicate that your tuna is rancid or that you forgot to take your pills.
The key to the chip’s design was taming the somewhat unruly organic transistor, says Jan Genoe, a polymer and molecular electronics researcher at Belgian nanotech research center Imec, in Leuven, who led the research with colleague Kris Myny. One advantage silicon has over organics is its monocrystalline structure, which allows for well-behaved switches. If you increase the transistor gate’s voltage above a known threshold, the current turns on. But today’s organic transistors—which swap silicon for a polymer—are unpredictable. Each one can have a slightly different switching threshold.
In applications where organic transistors are already taking hold, such as turning pixels on or off in some e-reader displays, a few transistors don’t affect the overall performance. Yet in logic circuits, a single transistor can stop the show. ”If only one is a little bit off, then nothing works,” Genoe says.
So Genoe’s team built an extra gate into the back of each organic transistor. He says this back gate allows the researchers to better control the electric field in the semiconductor, and thus avoid accidental switching.
Fabricating the 25-micrometer-thick chip starts with a substrate made from polyethylene naphthalate—a plastic. ”You could compare it to the material that you use to wrap your sandwiches,” says Genoe. ”It’s very flexible.” On top, the team placed a 25-nanometer-thick layer of gold, patterned to make the circuit. Above that sits an organic dielectric, followed by a second patterned gold layer, and finally the organic semiconductor, made of pentacene.
After fabricating the chip, Genoe’s team tested it by running a 16-line program to average changing input values with those stored in memory, the software for which they had hardwired into a second flexible chip. The processor, he says, could execute about six instructions per second.
Genoe hopes such chips can be made at a tenth of the cost of a similar silicon circuit But to realize that promise, manufacturers will need to translate the IMEC researchers’ carefully controlled, photolithography-based, laboratory-scale fabrication technique into a commercial one—such as those being used for large-area, printed electronics.
”It’s not as difficult as one might think,” says Dan Gamota, cofounder and president of the electronics printing company Printovate Technologies, in Palatine, Ill. Gamota, who was not involved in the research, taught commercial printing press operators how to modify their traditional ink-on-paper printing techniques to manufacture an early printed electronics display while a director at Motorola in the late 2000s.
Still, he says, printing logic circuits will have some special requirements. For today’s printed electronics, such as those proposed for lighting devices, he says, the thickness of the materials is crucial, but for logic circuits, manufacturers will also need to align the circuit’s layers more precisely. That will require both new measuring tools and new reliability training programs for printing press operators. ”A printed electronics operator is like a mechanic who knows how to work on a Ferrari,” Gamota says, ”while a traditional printer knows how to fix a Ford.”
Though manufacturing will improve, Gamota says, he doesn’t believe organic logic circuits will ever have the hundreds of millions of transistors found in today’s silicon chips. Instead, he says, many in the field look to use organics as a relatively dim-witted sidekick for silicon processors. As an example, he describes shopping for a new pair of pants by using your smartphone to communicate directly with plastic circuits inside the clothing. The circuits will tell you how the pants will look on you so you can try the trousers on virtually.
Like Gamota, Gerwin Gelinck, who worked on the IMEC chip, also believes that organics will make their start as a complement for silicon. Gelinck is a program manager at the Holst Centre, in Eindhoven, Netherlands, a research organization with commercial partners that include the display companies Polymer Vision and Panasonic. He believes that eventually more-complex organic logic may replace ”peripheral” silicon chips in devices like displays, to lower these gadgets’ cost and size.