But to make paper-based circuits that can perform calculations or control displays, researchers need to find a way to print transistors. Unfortunately, previous paper transistors perform poorly because the surface of regular paper is bumpy and uneven. For a transistor to perform optimally, electrons have to move easily through super thin layers of conducting and semiconducting materials. These layers are only a couple hundred nanometers thick, while the bumps on the surface of regular paper are tens of micrometers tall. The paper bumps disrupt the flat layers of electronic materials and interrupt the device’s electron flow.
In addition to its rough surface, regular paper’s other limitation is its opaqueness. To produce electronics for transparent displays, researchers need a transparent material, like plastic or glass.
So for printed transistors, Liangbing Hu, a materials scientist at the University of Maryland, College Park, turned to a smooth and transparent kind of paper called nanopaper. Instead of the micrometer-sized cellulose fibers found in regular paper, sheets of this material contain nanoscale fibers that produce an even surface and allow light to pass through.
Hu’s group made their own nanopaper using previously reported methods, which involve treating paper pulp with oxidizing chemicals. The nanopaper has cellulose fibers with an average diameter of 10 nm. “It’s as flat as plastic,” Hu says.
The Maryland researchers then built transistors on the paper by depositing a layer each of three materials: first carbon nanotubes, next an insulating organic molecule, and then a semiconducting organic molecule. To complete the device, the team topped it with electrodes, also made by laying down carbon nanotubes. Besides serving as electrodes for the transistors, the nanotubes provided a structural backbone, preventing excessive wrinkling in the paper after the solvents used in the fabrication process evaporated.
The resulting transistors are about 84% transparent, and their performance decreases only slightly when bent. Still, Hu says the performance of these first transistors is not optimal. He thinks decreasing wrinkling in the nanopaper will improve the devices.
Jeffrey Youngblood, a materials engineer at Purdue University, calls these nanopaper-based transistors “another step down the road to renewable printed electronics.” For such devices to be practical, he says, the researchers will have to find a way to produce the transistors via a scalable process, such as roll-to-roll printing, instead of the tedious layer-by-layer process