Pictures of the Future - Polymer Chips

Wolfgang Mildner places a plastic yogurt container on his desk and points to the barcode. "We’re going to replace these with plastic chips," he says. Mildner is the Managing Director of PolyIC, a start-up located in Erlangen, Germany. Barcodes are on all products today, but the only significant data they contain is price. To identify products individually (for example, by expiration date or other information) what’s needed are so-called intelligent labels that use RFID (radio frequency identification) technology. These radio chips, which are affixed to products, are opening up new possibilities in delivery, inventory management and labeling, especially because they can be read from a distance. In theory, this means a company could identify all the products it has in stock at the push of a button—and determine their exact location. Another conceivable application for this technology is the automatic check-out line, where customers would simply move their shopping carts past a radio scanner that automatically registers everything in the wagon.

Major retailers such as WalMart, Tesco and Metro are currently testing RFID systems that use conventional silicon chips, which cost 1 € or more each. "That’s so expensive that they’re only suitable for use with expensive merchandise," Mildner says. Even with great effort, a silicon-based RFID chip will never cost less than five to ten cents per unit, even over the long term. Silicon does offer advantages, though. The chips feature very high performance and are also fast. "In this respect, we’re not even competing with silicon by using plastic," says Mildner.

A study by the German market research firm Soreon forecasts that total market volume for RFID systems in Europe alone will increase from 400 mill. € this year to 2.5 bill. € in 2008. "This market will be open for plastic chips too because we’re sharply cutting costs," says Mildner. "Over the long term, we’ll get them down to one cent per unit—or even less." To do this, the start-up company is relying on revolutionary production technology. The idea is to print circuits made of organic polymers onto foils—like a newspaper is printed on paper. For this, PolyIC is taking advantage of the expertise of Leonhard Kurz GmbH & Co. KG. Together with Siemens Automation & Drives, Kurz, a leading manufacturer of stamping foils, established PolyIC in November 2003.

"Basically, we have to invent a plastic-based silicon semiconductor technology," says Mildner. But this is doable, as Dr. Wolfgang Clemens, head of Applications, demonstrated with several successes. Even back when he was a project manager at Siemens Corporate Technology, for example, Clemens and his team had already built key microelectronics components—transistors and rectifiers—from polymers (see Pictures of the Future, Fall 2002, Chips on a Roll). Today, PolyIC has about a dozen developers and is much further along with the technology. "We’ve used different printing techniques to create very stable circuits that conduct logical operations," says Clemens. No other research group has done this using printing techniques. One polymer chip, for example, ran continuously for more than ten months. The chips also function after two days stored at a temperature of 60 °C and at 100 % humidity, and they’ll work in a heat chamber until temperatures exceed the 120 °C mark. PolyIC also holds the world record for the highest frequency for a polymer circuit: One of the company’s plastic ring oscillators achieved 200,000 cycles per second. That’s more than enough for processing data in an RFID chip. But it’s not so easy to achieve the standard carrier frequency for radio communication—around 13 MHz—using organic components.

Developers must solve several problems before polymer chips can be printed on foils like newspaper is printed on paper (see box). And silicon-chip designs can’t just be applied to polymer chips; silicon has entirely different material properties that are the basis for optimizing production processes. "That’s why we develop special simulation models to create new circuit layouts compatible with printing procedures," says Walter Fix, head of Chip Design. "But compared with silicon technology, we can implement new chip generations more rapidly." In fact, it only takes a few days to set up the layout on the computer, produce the masks and then make the chip.

The first product that PolyIC plans to launch on the market in two years will be a simple RFID polymer chip with a few hundred transistors, for use in applications such as forgery-proof labeling. "The next step will be a 32-bit chip that will usher in the first applications for such units in the logistics sector," says Mildner. This chip type will enable companies to establish internal standards for their inventory management systems. Then, in five years, it might be possible to introduce the electronic product code containing several thousand transistors onto the market. The storage capacity of these chips will range from 64 to 128 bits, which means they have the potential to replace barcodes, which usually stores only 44 bits of data. "We’re working on printing chips directly onto packaging, the way barcodes are printed today," says Mildner. "Inexpensive chips could also serve as logic circuits for electronic advertisements, to be placed in areas never considered before." Mildner is referring to displays that could be mounted on packaging to show different types of product information, or small digital display units for blood, urine, and pregnancy test kits. But even 128 bits sounds modest when you consider the gigabit storage capacities of silicon. So will we ever see computer chips made of plastic? "From today’s perspective, that seems a bit utopian," says Mildner. "But then again, never say never…"

Norbert Aschenbrenner