SOUTH SAN FRANCISCO, Calif. — After several false starts, developers of thin-film transistor circuits last week claimed breakthroughs that they believe will finally push the technology to market. If successful, the efforts could enable a long-promised low-cost class of electronic paper, displays, labels, RFID tags, sensors, smart cards and perhaps even programmable wallpaper.

But will the claims hold up this time, or are they just more hype?

For years, promoters have been billing thin-film transistor circuits--and the broader category of printed or plastic electronics--as the next step beyond wafer-based ICs. The technology uses inkjet printers and polymer inks to print cheap, low-density ICs on a system, display or just about anything else. As such, it promises to make the IC ubiquitous, open- ing a host of new applications.

But despite the millions of dollars that have been invested in R&D over the years, a number of companies--including mighty Intel Corp.--have failed to move thin-film transistor circuits into production. The transistors are working in the lab, but not one company has "sold anything" in the commercial market thus far, said Peter Harrop, chairman of consulting firm IDTechEx Ltd. (Cambridge, U.K.).

At last week's Printed Electronics USA event here, 3T Technologies, Kovio, Orfid, PolyIC, STMicroelectronics, Thin Film Electronics and other concerns tipped details on printed electronics for a range of applications. Vendors claimed to have solved many of the problems associated with the technology,

including lackluster transistor performance, poor mobility and integration issues for the polymer materials.

More than 150 organizations are scrambling to develop the technology in one form or another. Many seek shares of a thin-film transistor and memory market that IDTechEx predicts will grow from nearly zero today to $40 million by 2009 and $8 billion by 2017.

The promising thin-film circuit field is one of several technologies in the broader printed-electronics market. Printed (or plastic) electronics is a catchall phrase that describes the method of depositing simple electronic circuits on a system via inkjet printer or related means.

The broader printed-electronics market, according to IDTechEx, includes conductive inks, electrophoretic displays, organic LEDs, thin-film photovoltaics, and transistor circuits and memories. In total, the market for printed electronics, including organics, inorganics and composites, is projected to grow from $1.8 billion in 2007 to $48.18 billion by 2017, according to IDTechEx.

The fastest-growing market segments are for thin-film transistors and memory, according to the firm. Unlike traditional silicon chips, thin-film transistor circuits do not use crystalline or amorphous silicon; instead, thin-film products are based on organic or inorganic compounds that enable flexible circuits.

"The transistors are the key to huge new markets that the silicon chip will never reach," Harrop wrote in a recent report. "The new transistors can be deposited on low-cost flexible substrates, aluminum or stainless-steel foil. [At present] they are much larger than today's silicon transistors, but they can be one-hundredth of the cost, thinner and lighter in weight."

There are still some downsides. Transistor speeds, for one, are limited to frequencies around tens of megahertz.

"Reliability and stability of organic [thin-film transistors] is the major problem at present," according to a paper from the Fraunhofer Institute for Physical Measurement Techniques (Freiburg, Germany). "Typical problems, which mostly are not yet completely understood, are that the mobility is typically reduced, and the threshold voltage and the subthreshold region of the [transistor] are sensitive to environmental conditions."

For those and other reasons, developers have failed to bring the technology into the commercial markets. In 2001, for example, startup FlexICs attempted to develop semiconductors on plastic substrates for flat-panel display and related applications. But by 2005, FlexICs, which was funded in part by Intel, had folded and put its assets up for auction.

In another ill-fated effort, Opticom (Oslo, Norway) was set up in 1994 to develop multilayer, nonvolatile polymeric memory systems. In 1999, Opticom signed a joint agreement with Intel to commercialize the polymer memory technology. But problems resulting from the introduction of polymeric materials into silicon fab environments resulted in Intel's canceling its collaboration in 2005.

Thin Film Electronics AB, formerly the Swedish research wing of Opticom, was separated from its now-defunct parent in 2006 and has since reapplied itself to the production of low-cost, all-polymer nonvolatile rewritable memories. The company is also pursuing RFID tags, smart labels and other applications, having abandoned the large-scale, silicon-based general-purpose memory it had previously investigated under research contracts with Intel.

It's highly unlikely that thin-film transistors will replace today's semiconductor memories, said Rolf Aberg, an executive board member of Thin Film Electronics. "It will be a complementary technology to semiconductors," he told EE Times. "We think it will open up new applications."

For example, Belgium's Cartamundi Group plans to use Thin Film Electronics' technology to deposit memory circuits on a new class of standalone game cards that would let users play games over the Internet.

Thin Film is putting the manufacturing infrastructure in place to bring its technology to market. Last week, Agfa (Mortsel, Belgium) and Thin Film announced that they intend to enhance the materials for volume production of printed memory devices.

Thin Film last week also signed a deal with InkTec Co. Ltd. (Kyungki-do, South Korea). The joint collaboration is focused on optimizing InkTec's silver inks for Thin Film's memory cell electrodes.

STMicroelectronics Inc. last week claimed to have developed the first entries in a line of standalone thin-film circuits. Using a combination of nano- imprint lithography and inkjet printers, the company has devised a 4-bit arithmetic logic unit, a full adder and a one-time-programmable device, said Luigi Occhipinti, corporate R&D program manager for ST (Geneva).

Occhipinti acknowledged that there are a number of hurdles to clear before the technology enters production. Be- sides lackluster transistor performance, he cited "a lack of stability in n-type organic materials" in a conference presentation last week.

As a result, the first applications in thin-film circuits may not be standalone devices, but offerings in the booming area of RFID. One startup, Kovio Inc. (Sunnyvale, Calif.), came out of stealth mode at the conference with a silicon-based thin-film transistor technology--a low-cost RFID tag (see story, page 16).

Today's RFID tags run 15 cents per unit. "You can't get below 10 cents" with conventional technology, said Kovio CEO Amir Mashkoori. Kovio's technology will lower the cost of RFID tags to 5 cents when it moves into production in 2008, he said. The ultimate goal is to bring the tags down to a penny per unit.

PolyIC GmbH & Co. KG claimed at last week's event that it had beaten rival Kovio to the punch. PolyIC is developing thin-film devices manufactured in an industrial roll-to-roll printing process. The company touts two organic-based chips: PolyID, a 13.56-MHz device equipped with a 4-bit memory, and PolyLogo, geared for "smart objects."

The first products are expected to sample by year's end, said Wolfgang Clemens, head of applications at PolyIC. He would not specify pricing for the company's RFID tags but said the goal is to drive costs to the "few-cent range."

Another emerging market for the technology is displays. Startup Orfid Corp. claims to have begun shipping its first devices for the display market.

Orfid has developed an organic electronic technology called the vertical organic field-effect transistor. Because of its architecture and the use of conductive polymers in its fabrication, the VOFET offers performance characteristics similar to conventional, wafer-based silicon transistors but can be produced at significantly lower cost, the company said.

Another novel approach is being developed by a little-known startup called 3T Technologies Ltd. (Cambridge, U.K.). 3T is working on a generic device platform and process for applying transparent conductive oxide materials to enable what it calls "transparent thin-film technology" or "invisible circuitry."

"Based initially on zinc oxide as the active transparent semiconductor, the device construction and manufacturing approach provide the basic building block for transparent logic circuitry and display pixel switching to support a wide range of display media," according to the company.