PRINCETON, N.J. A group of Princeton University researchers is taking its amorphous silicon process into the commercial arena with a startup called Printed Transistor Inc. The company plans to initially offer design and prototyping service for the process, which can print transistors on a wide variety of substrates, including plastic and flexible foil.
The Princeton group has been working since 1990 to develop a class of materials and processes aimed at merging conventional printing technology with semiconductor electronics to create a new type of "macroelectronics." Low-cost displays and image sensor arrays are typical application areas that could benefit from the macroelectronic techniques the startup has assembled.
To get the enterprise off the ground, Printed Transistor is offering a fast prototyping service for large-area circuits. "This technology is particularly useful for prototyping because you can turn something around very quickly," said Matthias Wagner, business development manager for the company. The initial service will make use of laser printers to print the masks needed to define circuits. By eliminating conventional mask-making and silicon fab runs, the thin-film process will allow clients to quickly and cheaply redesign prototypes.
The company eventually envisions a technology similar to current digital printing services. A customer would design a circuit and transfer the digital file to a circuit printer, which would run off a large number of copies.
"We have a good portfolio of intellectual property that we are licensing from Princeton University. The idea is to build expertise on our team through customer projects and let the market meet us as we develop our core technologies versus trying to push a particular technology onto the market," Wagner said. The two critical properties of the technique circuit definition on a wide variety of rugged, low-cost substrates and the potential low-cost use of printing technology to manufacture circuits could address a wide number of applications.
Large-area X-ray sensors provide an example of how the technology could be applied, Wagner said. "Active-matrix plates for very high-resolution X-ray sensing have been tried. The concept of being able to make very large arrays is interesting, both for coarse-resolution X-rays that you would use to check luggage, or for field use, where you wouldn't have to worry about it breaking if it were on a large plate of glass," he said.
The technique could also be applied to laying down additional active thin-film layers on conventional ASIC circuits. The technique could be used to build low-cost image sensors, for example. The flexible substrate aspect could be employed to build sensor membranes that could be bonded to the exterior of machines or consumer products.
"We have been hearing a lot lately about plastic transistors, and that is certainly a promising technology, but we are planning to leverage the experience that the Princeton group has in amorphous silicon. We think that technology still has a long way to go," Wagner said. One founder of the company is Sigurd Wagner, who heads the Macroelectronics Group at Princeton's department of electrical engineering. Formerly, the research division was known as the Amorphous Silicon Group, but the processes have gone beyond that initial area of expertise. The amorphous silicon expertise resulted from the origins of the research group in silicon solar-cell technology.
Amorphous silicon originally was viewed as a low-grade form of the semiconductor that would be better matched to the mass production of solar cells, which use the basic p-n junction that underpins high-performance crystalline silicon microelectronics. However, the junctions do not need critical characteristics such as high speed or low noise to be effective at solar electric generation. But the process has to be amenable to the high-volume manufacture of large-area integrated circuits. Thus, the solar cell industry has found methods for the low-temperature deposition of silicon films.
While solar cells have traditionally supplied the major application area for amorphous silicon, researchers have also developed thin-film transistor and circuit technology for the medium. Although such circuits cannot compare to single-crystal silicon ICs in terms of performance, they are still fast enough for human interface components, such as displays.
Two characteristic problems of the technology are stress-induced fracturing of the films and the adhesion of the silicon to a substrate. The Princeton group has therefore spent many years studying these essentially mechanical problems, according to Matthias Wagner. Solving those problems is crucial to applying the technology to large-area substrates or in rugged applications such as smart credit cards or sensor membranes.
Starting with the basic process of building thin-film transistors (TFTs) on glass substrates using hydrogenated amorphous silicon, the researchers have branched out into a variety of substrates and materials. The group can now routinely make TFTs by depositing amorphous silicon on plastic substrates at 150C. At higher temperature deposition, it is possible to create a better-performing material called microcystalline silicon that has been used to build both n- and p-channel transistors. Using steel substrates and a 750 process, the group has demonstrated TFTs with unusually high electron mobility. Another process has created a silicon conductor by depositing microcrystalline silicon with chlorine.
Along with the performance upgrades for traditional amorphous silicon TFTs, the Princeton group discovered some novel aspects of the mechanical problems associated with silicon on diverse substrates. One tactic to emerge from that research is the transfer of stress from the silicon film to the substrate. Since the silicon is usually more rigid than the material on which it is deposited, the technique of building "compliant substrates" gives the system an added design dimension.
Recently, there have been breakthroughs in applying printing techniques to the definition of circuits. The researchers have developed a process for jet-printing copper contacts onto the source and drain regions of transistors. The goal is a set of circuit definition techniques that could be implemented with known printing processes.
The silicon thin-film technology will be complemented by work by another group at Princeton that is using organic materials to build transistors. The director of that effort, James Sturm, is also a cofounder of Printed Transistor. Sturm's group has integrated organic light-emitting diodes (LEDs) with thin-film silicon transistors. The organic LEDs use large dye molecules as the active elements. Current is driven through a medium containing the dye molecules that respond by emitting light.
The technique has demonstrated that the underlying transistors can make good enough contact with the organic materials to drive viable light emitters. The result is a prototype rugged display on a 3-inch2 steel foil substrate with thin-film transistors integrated with organic LEDs. Operating at 5 V, the devices produce a bright display.