Skip to Main Content

Researchers Create Printable, Flexible Transistors

Technology Uses Carbon Nanotubes

Asst. Prof. Xuejun Lu, right, and graduate student Jarrod Vaillancourt.

By For more information, contact or 978-934-3224

Imagine high-speed transistors as thin and light as a sheet of paper that can be bent and shaped to follow the contours of a person’s body, an aircraft’s wing or fuselage, or practically any product packaging. Such electronic devices could help monitor one’s health, check the jet engines’ performance, or track the whereabouts of merchandise.

Although the working principle behind this technology has been known for more than a decade, the process of manufacturing printable, flexible transistors on a commercial scale has just become one step closer to reality, thanks to a collaborative effort between researchers from UMass Lowell and Brewer Science Inc., a major high-tech semiconductor/microelectronics innovator headquartered in Rolla, Mo.

Asst. Prof. Xuejun Lu of the Department of Electrical and Computer Engineering and graduate student Jarrod Vaillancourt, working with Drs. Xuliang Han and Daniel Janzen of Brewer Science, came up with a breakthrough process that enables them to mass-produce flexible high-speed thin-film transistors (TFTs) cheaper, faster, and over large surface areas, without the need for clean-room facilities or special lithographic and etching equipment. Such technology has a broad range of applications, including radio-frequency identification tags, electronic papers, “smart” skins, wraparound panel displays, transparent-transistor window panes, and other uses yet to be imagined.

According to Lu, the printing of TFTs at room temperature directly on a flexible substrate (base) using various organic semiconducting polymers has already been demonstrated. “The mobility, or flow, of electrons through polymers, however, is relatively slow, limiting a device’s operating speed to only a few kilohertz [thousands of cycles per second],” he says. In comparison, today’s desktop computers have processors running in the gigahertz (billions of cycles per second) range.

Studies have shown that carbon nanotubes (CNTs) ߝ billionth-of-a-meter-size cylindrical structures made up entirely of carbon atoms ߝ can be used to manufacture high-speed transistors. But growing CNTs requires extremely high temperatures, typically more than 900 degrees Celsius, which poses a major obstacle to fabricating electronic devices on flexible substrates.

“Moreover, transistors made with individual CNTs or low-density CNT thin films can carry only a very limited amount of current, just in the microampere range or even less,” says Lu. This is because the sidewalls of individual nanotubes get covered by amorphous carbon “soot,” which is a very common byproduct of CNT production. This soot forms barriers within the tubes, dramatically restricting the electrons’ mobility and generating heating effects throughout the circuit. “That’s why such transistors can’t be used for high-speed applications beyond 100 megahertz [millions of cycles per second],” he adds.

With the help of Prof. William Goodhue’s group at UML’s Photonics Center, Lu and his colleagues were able to fabricate a CNT thin-film transistor on a regular plastic transparency film using a printing process. The transistor was formed at room temperature by dispensing a tiny droplet ߝ approximately 0.05 milliliter of ultrapure, high-density CNT solution developed by Brewer Science ߝ with a syringe, similar to the process used in a typical inkjet printer.

The group’s findings, which were published in the December 2007 issue of Micro & Nano Letters, are being funded though a one-year, $150,000 grant from the National Science Foundation. Of this, $80,000 was awarded to UML, while the rest went to Brewer Science. The company is partnering with the University in patenting and commercializing this technology.

“Tests on our transistor indicated an operating speed as high as 312 MHz and a current-carrying capacity of more than 20 milliamperes,” says Lu. “The speed and current load are limited not by the transistor itself but by the external circuits we were using, which were not optimized. We expect to attain gigahertz speeds and higher currents soon.”