Internal Combustion Used to Make Thin-Film Transistors
Combustible mix of metal and oxygen could be path to flexible electronics
Photo: Northwestern University
18 April 2011—A team of scientists has developed chemical solutions that use their own internal heat to fuse metal and oxygen atoms and form semiconducting films at low temperatures. This method could pave the way for cheaper next-generation thin-film and flexible electronics. The work, which describes the results for films of several different compositions, appeared on Sunday in the journal Nature Materials.
The thin-film electronics behind today’s flat-panel displays are made of chaotically structured, or amorphous, silicon. But amorphous silicon is reaching its performance limits, and a new class of materials—amorphous oxides— will soon be making its commercial debut. Electrons in amorphous oxides can zoom through the material dozens of times as fast as they do in amorphous silicon, making for faster electronics. And unlike amorphous silicon, oxides carry current the same way in every direction, making them better candidates for bendable electronics like flexible solar arrays and roll-up displays.
To make these thin films, engineers primarily rely on ”sputtering,” in which vaporized material is flung at its target inside a vacuum chamber. This process would potentially be less costly if the material could be applied as a solution instead. But fans of the solution-based method have had to confront some inconvenient physics. Heat must be applied to condense the metal oxide, and the material performs best after it has been heated above 300 °C, which is about 100 degrees too hot for most flexible plastics.
Now Mercouri Kanatazidis, Tobin Marks, Antonio Facchetti, and Myung-Gil Kim of Northwestern University, in Evanston, Ill., think they’ve come up with a fix: replacing the external heat of an oven with the internal heat of a chemical reaction.
Many thin-film metal oxide solutions are made using water and metal-containing salts. When the temperature rises high enough, the oxygen atoms bind with the metal to form a chaotic tangle of metal-oxygen bonds. The team found that if they included a fuel like acetylacetone or urea in the mix, they could raise the internal energy of the mixture. Boosting the temperature to just 200 °C triggered a combustion reaction and enough self-generated heat to anneal metal-oxide films.
One of the team’s biggest challenges was finding a way to deal with structural changes created through the combustion process. The internal heat can create voids in the resulting films. These voids are useful for sensors and catalysts that require a lot of surface area, says team member Facchetti, who is also affiliated with Polyera Corp., in Skokie, Ill. But the gaps are counterproductive for thin-film electronics because they reduce the overlap between the atoms’ electron clouds and thus hinder the ability to transport current. ”One of the major challenges was to make sure we could make a film that is very dense,” Facchetti says. The team ultimately found it could circumvent the void problem by alternately depositing and annealing thin layers to build up the film.
One device made using the technique—an indium oxide transistor—boasted an electron mobility of 6 square centimeters per volt second, roughly 10 times that of thin-film amorphous silicon devices. That’s a heartening figure but one that will have to be backed up with more experiments, says John Wager of Oregon State University, in Corvallis, Ore. ”If the extra energy from combustion-based synthesis really does give you better-performing devices at low temperature, then that’s really nice,” Wager says. (See the feature article by John Wager and Randy Hoffman in the May 2011 issue of IEEE Spectrum for more on amorphous oxide semiconductors.)
One big question that will need to be answered in future work, Wager says, is how stable these devices can be. The threshold voltage needed to turn on thin-film transistors tends to drift with use, and that behavior ”tends to be more problematic at low temperature,” he says. ”If their combustion synthesis leads to more stable transistors, that could be really big.”