Carbon nanotubes hold promise as a material for making thin, flexible electronics like displays and solar cells. But one stumbling block to making transistors out of them has been achieving the right combination of electrical properties in the nanotubes. Now a simple chemical processing method developed by researchers at Cornell University and DuPont overcomes this obstacle and provides a path toward low-cost, commercially viable electronic inks.
Carbon nanotubes can be thought of as sheets of graphite rolled up in different ways to form tubes. Depending on how they fold, the nanotubes can either have metallic or semiconducting properties, says physicist Graciela Blanchet, a research fellow at DuPont. Carbon-nanotube mixtures are usually two-thirds semiconducting and one-third metallic. The metallic nanotubes provide a conducting path that causes a drain on the circuit, “like putting a wire between two electrodes in a transistor,” Blanchet says. “You cannot turn it off, and you need to be able to switch the transistor off.”
Researchers have different approaches to dealing with this problem. Metallic and semiconducting nanotubes can be separated based on their chirality, which involves the subtleties of how the carbon atoms attach to each other. “That’s been successful at the lab scale but not on an industrial scale,” says Michael Strano, an associate professor of chemical engineering at MIT. Some researchers deposit a mixture of carbon nanotubes into a thin film and then burn away the metallic ones by applying a high current–a process that is also difficult to scale up and control. Another method involves chemically treating a surface so that semiconducting nanotubes deposit preferentially, says Zhenan Bao, an associate professor of chemical engineering at Stanford University.
The Cornell and DuPont team reacted carbon nanotubes with tetrafluoroethylene, an inexpensive chemical. The addition of fluorine molecules to the metallic nanotubes either destroys their conductivity or converts them into semiconducting nanotubes. Blanchet, along with Cornell materials scientist George Malliaras and their colleagues, described the method last week in the journal Science.
“This method looks quite promising. I think it’s very important work because getting rid of the metallic tubes for high-performance transistors is a big challenge,” Bao says. “This is a totally different approach and [it] achieved very good results.”
Blanchet says that other groups have tried chemically modifying nanotubes, but the electrical properties degraded, and the materials became unstable. “We were very lucky in some ways,” she says. “We started doing the functionalization very slowly, so we were tracking the electrical properties at the same time. We found a window in which it worked.”
Blanchet and her colleagues also demonstrated that an electronic ink could be made out of the semiconducting nanotubes. They dispersed the nanotubes in a dichlorobenzene solvent and used it to print working transistors.
Smaller design teams can now prototype and deploy faster.