EDITOR'S PAGE

Better living through plastics

The eyes of the techno-saavy have long been focused on the inexorable miniaturization of microelectronics geometries. Recent advancements signal the acceleration of what some call "the nanotechnology revolution." In our special feature beginning on page 16, "When Micro Meets Nano," senior editor John Conroy assembled a sampler of recent breakthroughs. For example, he reports on a Bell Labs' team which had successfully made an organic transistor out of a cluster of molecules. Shortly after the feature went to the printer, the research group upped the ante by announcing they had created a transistor out of a single molecule.

Although such mind-boggling paradigm obliteration has garnered lots of attention, another manufacturing and application revolution is under way, the commercial benefits of which may be much closer to coming to market than molectronic devices, quantum dots, and the like. I'm talking about the flexibility revolution, where a new generation of chips made on plastic or other nontraditional materials will foster lightweight, low-power yet brighter FPDs and eventually electronic newspapers, bendable instrument panels, and other gee-whiz gadgetry.

One privately financed start-up company exploring this realm is FlexICs (pronounced flex-icks), located in a former Read-Rite facility in Milpitas, CA. "Our mission is to focus on creating an entirely new way of designing electronics on plastic," relates Shyam Dujari, vp of marketing. "Our vision is to set both a technical and economic standard, because once you do it on a flexible substrate, you don't have to use the traditional methods of manufacturing."

Rather than competing with well-entrenched Asian panelmakers, FlexICs is taking the foundry path. "We don't intend to be an IDM (integrated device manufacturer)," notes Dujari. "We intend to be a semiconductor-on-plastic foundry, where our customers will be the display suppliers, memory module suppliers, optical networking component makers, and other subassembly manufacturers. We would essentially manufacture to their specifications based on our process."

FlexICs's proprietary thin-film transistor on plastic process—ultra-low-temperature polysilicon (ULTPS)—was originally developed by the company's four cofounders when they worked at Lawrence Livermore National Laboratory. "We have developed a way to do all of the steps (gate oxidation, dopant activation, polysilicon processing, gate and metallization, contact anneal) at 100°C or less," explains Dujari. "We lay a thick oxide on the plastic substrate, and this oxide acts as a thermal barrier. Then we put amorphous silicon on top of that, on which we shoot an extremely narrow excimer laser pulse. The thick oxide layer protects the bulk of the plastic substrate from the heat (up to 1500°C) that the top silicon surface is subjected to.

"We can use a wide variety of substrates, but we've chosen to use PEN (polyethylenenapthalate) and PET (polyester) primarily because they offer the best chemical properties and resistance to different solvents, they're the least expensive and have high transmissivity, and they have a low to moderate thermal coefficient of expansion, because when you're doing a multilayer process, the registration between the layers becomes a crucial issue. What that gets us is much higher electron mobility (200–400 cm² V/sec) compared to what an amorphous silicon device would get (about 1 cm² V/sec). So with higher mobility, we can actually integrate the drivers (into the substrate), something which amorphous silicon displays cannot do—they have to mount the chips separately and then wire bond them."

FlexICs's pilot line now runs in a traditional batch mode using 6-in. wafers. Dujari says they have almost completed process characterization and already built samples for customer evaluation. But the company's longer-term goal is to fabricate the devices using plastic in a roll-to-roll configuration, what he calls "a whole new way of manufacturing," one which has compelling economic benefits. "Our detailed model shows that our capital expenditures for an equivalent roll-to-roll manufacturing plant versus a panel manufacturing plant would be about a third," or $400 million compared to as much as $1.5 billion.

This new manufacturing approach "is really more in the future as we start to get the process fully debugged," admits Dujari. "We have taken select roll-to-roll manufacturing steps though, and we're doing R&D work to make sure we understand that.... But we really believe that this will enable an entirely new class of electronics, one that could be far more cost effective than the alternatives."

Tom Cheyney
Editor

tom.cheyney@cancom.com

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit