Manufacturing progress key to flexible electronics’ success
By Tom Cheyney
As the first commercial flexible electronics reach consumers, many significant manufacturing and technological obstacles must be overcome for the market to reach its multi-billion-dollar potential over the next five to ten years. Pioneering efforts by Polymer Vision, Plastic Logic, Nanoident (see “Nanoident opens facility for organic semiconductors,” page 41), and other companies signal the first moves from the lab or pilot-line stage to volume production.
Polymer Vision, which spun off from Philips in late 2006, has announced it will ramp up a Southampton, U.K., manufacturing facility (in partnership with Innos) for its ultra-thin-film transistor polymer display modules. The company has also entered into an agreement with Telecom Italia to “bring the ‘cellular book’ to market.” Models of what the company touts as the “world’s first commercial rollable display product” were unveiled at the 3GSM World conference in Barcelona in February.
Edzer Huitema, Polymer Vision’s CTO, says a blend of refurbished and new AMLCD equipment will be deployed, as well as a proprietary lamination/delamination tool, in its Class 100 production facility (scheduled to come on line later this year). He adds that they have achieved field-effect mobility and driving voltages comparable to those of conventional TFT devices. Yields appear to be sustainable throughout the process flow, and defect sources, which are “comparable to those found in LCD manufacturing,” are “under control.” Ongoing quality-control work at Polymer Vision is focusing on materials purity and various types of insulator layers, according to Huitema.
Polymer Vision provides a five-inch display in Readius, a product the size of a cell phone.
Bolstered by a recent funding round of $100 million, Plastic Logic plans to build and equip a green-field factory site in Dresden, says Simon Jones, VP of product development. The company expects to have “product-quality modules” of its “take anywhere, read anywhere...thin, light, robust e-paper displays” by mid-2008, with a production target of more than 1 million 10-inch-equivalent units for 2009.
Plastic Logic’s direct-write, room-temperature process requires no mask alignment and can be scaled to a large substrate size, says Jones. The company “measures contrast and yield on every panel” and has “captured a huge amount of defect data,” which is “essential for the move from R&D to production.”
For the emerging flexible thin-film, organic, and printed electronics markets to flourish, most industry professionals agree that roll-to-roll (R2R) processing must be implemented on the factory floor. Indeed, thin-film photovoltaics, OLEDs, and RFIDs are already in pilot or volume production using R2R techniques.
But for the manufacturing of large-area and conformable displays, paperlike e-books, and high-performance solid-state lighting to take place, it remains an open question whether inkjet, thermal laser imaging, or other printing technologies; optical, imprint, or digital lithography; adapted semiconductor and LCD processing methods; inorganic or organic material; or a combination of the various methodologies and chemistries will be leveraged into successful, scalable R2R approaches.
Why R2R? The main reason is cost, which must be cut by at least 50% compared with batch-processed components, according to DisplaySearch’s vice president, Barry Young. Hans Maidhof, senior product manager at Applied Materials, echoes the cost-cutting sentiment, saying that “flex will only succeed if it’s cheaper.”
Maidhof says, “Thin films reduce the cost of semiconductor materials, continuous fabrication increases utilization and reduces production costs, and application of industrial processes simplifies production while providing high manufacturing rates. ...R2R [also] leverages form factor to lower overall customer costs.”
The challenges that face those trying to commercialize continuous processing are cultural as well as technical. “R2R and Web-converting folks are not used to semiconductor requirements and vice versa,” says Carl Taussig, program manager at Hewlett-Packard Labs. “We pretty much build everything, and building [your own] tools slows development.”
Included among the equipment that Taussig and his team have built is the self-aligned imprint lithography (SAIL) system.
This high-resolution tool, along with stamping, mastering, and other proprietary technologies, has allowed HP (with its partner PowerFilm Solar) to build what he claims is “the first flexible TFTs and [active-matrix] backplanes made fully with R2R processes.”
A seventh-generation dry-process, maskless production tool for printing thermal color filters resides “at a customer site in Asia for evaluation,” according to Eran Elizur of Kodak’s graphic communications group. The system handles 2250 x 2250mm substrates, employs five 5-micron-resolution laser heads, has 3-micron imaging accuracy with an imaging speed of up to 2 meters per second, and supposedly decreases manufacturing costs per panel by 30%.
For R2R processes to be consistent, efficient, and high-yielding, there must be reduced contamination and defectivity levels, precise endpoint control, very high uniformity, subnanometer-level surface roughness, and assured reliability.
Applied’s Maidhof admits that, although his company has a cleanroom-compatible vacuum web coater system, the “particle issue is most important” and “not totally solved in our tools.”
“With an endless process, where’s the endpoint?” quips Taussig of HP Labs. For their amorphous-silicon TFT process, his group uses interferometry to assess the endpoint of back-channel etching and fluorescence techniques to monitor and control the thickness of the polymer mask etch.
“Surface roughness is still an issue with flex and is not good enough for making transistors,” explains Bob Street, senior research fellow at Palo Alto Research Center. The roughness average needs to be a few angstroms, but that plastic substrates remain “five to ten times rougher than glass.” Pointing out the susceptibility of flexible substrates to scratching, Street says the plastics people “need to learn how to improve quality.”
Dan Gamota, director of printed electronics group at Motorola, believes the key challenge facing printed electronics is “how to [perform] quality control and characterization [tests] on rolls many hundreds or thousands of feet long. ...How do you check individual transistor device mobility on 2,000 feet of film?” (Motorola has produced “more than 60 miles of printed electronics” and is “close to getting dielectric layers 1 to 5 microns thick.”)
Display Search’s Young points out the limitations of current R2R technology: “Do you have a process where a particular display is going to stay constant from roll to roll? There’s not a lot of flexibility in roll-to-roll manufacturing.”