Flexible Displays Spring Into Action

Several companies are ramping up production of flexible displays, expecting to hit the market in 2009. The launch of Plastic Logic’s magazine-sized e-reader is expected in the first half of the year. (Source: Plastic Logic Ltd.)

Also in the running for near-term launches of flexible displays are Prime View International Co. Ltd. (PVI, Hsinchu, Taiwan) and LG Display Co. Ltd. (Seoul, South Korea). Most companies working in the field are readying monochrome flexible displays, with later plans for color, but LG has focused on color from the beginning, showing off a 14 in. flexible full-color screen in 2007.

Also, most players have their eyes set on handheld devices, but LG and Sony Corp. (Tokyo) are looking further into the future when flat-panel TVs will be replaced by thin, flexible panels consumers can simply hang on a wall. “It is not an unreachable target; it can be achieved within 10 years,” said Jiro Kasahara, R&D director at Sony's Fusion Domain Laboratory. “But only with these technologies can we have a wall-hanging display. That is the big difference between this technology and LCD and plasma.”

PVI has already carved out a niche for itself in today's generation of e-readers, which are based on rigid displays. Both the Amazon Kindle and the Sony eReader use PVI's current technology. But the company is working on flexible versions, having licensed technology from Philips Research that it will be able to make in its existing TFT-LCD fabs.

Each of the players in the flexible display industry has its own blend of display and control technologies, differences that even permeate through to choice of substrate material. But the first generation of flexible displays will be built mostly on existing LCD equipment, which reduces risk, according to Thomas van der Zijden, vice president of marketing for Polymer Vision. “We make use of standard processes from the LCD and semiconductor industries. We did that on purpose because coming up with a completely new technology introduces a lot of uncertainties.”

Konrad Herre, vice president of manufacturing at Plastic Logic, said the company looked at more than 200 locations before it settled on Dresden, Germany, to build its first commercial-scale production plant. The decision was based on several factors, Herre said, including local support and development grants. But existing experience in production and R&D from the semiconductor industry was also key, since Dresden is home to fabs owned by Advanced Micro Devices (AMD), Infineon Technologies and Qimonda, among others.

Aiming to build 10 in. displays, Plastic Logic decided to base its production line on Generation 3.5 LCD equipment. “The factory is designed to be able to produce more than 1 million 10 in. displays next year, and we can upgrade the facilities to double capacity,” Herre said.

Using a substrate measuring 650 × 780 mm, Plastic Logic can squeeze nine display modules onto the same substrate. Twenty backplanes go into a cassette that can be moved between tool clusters in the fab by automated guided vehicles. “The cassette weighs 60 kg,” Herre said. “It is not possible to handle them manually.”

One problem that affects the display makers in Europe is that, because the continent lost all its flat-panel display manufacturing to the Far East, there is no local infrastructure for support. Shortly before Plastic Logic announced its decision to locate in Dresden, Hermann Hauser of Amadeus Partners, one of the lead investors in the company, said support from the LCD toolmakers in Europe was a concern. And it was not just a concern for Plastic Logic, but for the nascent plastic electronics business on the continent as well.

One key problem for European manufacturers is that the LCD tools they buy are not “CE marked,” Herre noted, meaning that they have not been through the standard approval process needed to sell electronics within the European Union. The marking process is something that Plastic Logic itself must go through.

Van der Zijden acknowledged the concerns, but said Innos (Southampton, UK), which Polymer Vision acquired last year, had successfully built up manufacturing expertise in organic semiconductors and displays. “We chose Innos on purpose because we are confident in the organization there, and the support available around Europe,” he said. “We see this market as a fantastic opportunity for companies in Europe, and to keep that manufacturing capacity in Europe.”

For the time being, manufacturers have developed their own processes that they use in combination with those supported by the LCD manufacturing tools. “It is a little like the situation for the semiconductor industry in the mid-1960s,” Herre said. “We had to develop some of the equipment ourselves.” Some 40 years later, though, semiconductor manufacturing equipment supports standard processes, he added.

One significant problem with flexible displays, particularly those made on plastic, is that the substrates stretch when they are handled. This makes registration between layers very difficult, and demands that manufacturers find ways to compensate, especially if they are using mask-based lithography to define circuit features.

Plastic Logic decided to compensate for distortion dynamically, opting for a direct-write form of lithography for most steps. An initial step exposes the bottom circuit elements in one step, using conventional lithography. Subsequent steps involve coating the surface and then removing portions of it using laser ablation or by using an inkjet printer to deposit materials. The computer that controls the laser measures the substrate at each step to ensure that it maintains precise registration against the source pattern.

To deal with the problem of thermal expansion, LG has used metal foil as the display's substrate rather than plastic. “Plastic is transparent and light and reliable. But it has poor thermal stability and permeability,” said Jong Kwon Lee, chief research engineer at LG. “So we used metal foil.”

However, the move to metal or metal-foil substrates brings it own problems, the main one being surface roughness. Lee said LG developed a multilayer barrier that smoothes over the surface defects to provide a better substrate for the successive steps in the process.

More choices face manufacturers as they move up from the substrate to the active electronics and the actual display elements. Companies in the Far East, as well as U.S.-based organizations such as Universal Display Corp. (Ewing, N.J.), have focused primarily on combining conventional silicon processes with flexible displays. The two main European players, in contrast, have picked plastic electronics to drive the display elements.

LG has had to use amorphous silicon to keep processing temperatures down to <100°C, according to Lee. The use of amorphous silicon means that switching speeds are comparatively low, but the process steps are well understood. Using glass-based LCDs, Sharp Electronics has demonstrated that it is possible to use laser treatment with a catalyst to anneal silicon into a more regular crystalline form that offers much better electron mobility. The technique made it possible to build microprocessors onto a glass substrate, but the technique has yet to be tried on flexible displays.

Polymer transistors are generally slower. The electron mobility of the transistors used in Polymer Vision's displays is ~0.2 cm2/Vsec, according to CTO Edzer Huitema. By using alternative polymers, researchers have found ways to achieve electron mobilities that approach those of amorphous silicon (~1 cm2/Vsec). However, to be able to display video and animation, most LCDs use polycrystalline with electron mobility at least 10× that of amorphous silicon. The transistors in the Polymer Vision display allow the display to be refreshed in about half a second, Huitema said.

Some researchers, such as Professor Henning Sirringhaus of the University of Cambridge (Cambridge, UK), believe that it is possible to boost performance beyond 1 cm2/Vsec by not only improving the core carrier mobility, but also by optimizing structures. One possible direction is to exploit oxide semiconductors or hybrids of organic and inorganic materials rather than just carbon-based materials, typically a derivative of pentacene.

“We have done a lot of work with oxide semiconductors and nanoparticle dispersions,” said Sirringhaus, who also co-founded Plastic Logic. “We have got quite decent performance of the order of 1.5 cm2/Vsec. However, it has been hard to get the processing temperature below 150°C.”

Sirringhaus said his research team at Cambridge has been optimizing the self-aligned process it created for plastic displays. The technique works by depositing a molecular monolayer onto an electrode formed using lithography. A second ink used to define a contact flows off the monolayer so that it does not form an electrical contact. “We can control that down to 50 nm to 400 nm,” he said. “We are using inkjet printing, but reaching submicron dimensions.”

By reducing feature sizes, Sirringhaus said, it was possible to build transistors that could switch at up to 1 MHz. “We have made some simple circuits, such as ring oscillators. But the bulk of our work is going on to understand whether this processing could achieve the uniformity that is needed,” he said. “That goes out of our comfort zone, so this is a collaborative project with Plastic Logic — to understand whether this process can realize transistors with sufficient performance and yield. But we have made arrays with a couple of thousand transistors. It is not a process where one out of two doesn't work; it is where one out of a hundred doesn't work.”

As with the first batch of e-readers from Amazon, iRex Technologies (Eindhoven, Netherlands) and Sony, the unifying link between the various first-generation flexible displays lies in their use of e-paper. The e-paper displays are electrophoretic — they work by moving charged colored particles around a capsule by applying a voltage difference to contacts in front of and behind the display. Typically, dark particles are designed to carry an opposite charge to white particles, so they move in opposite directions.

This type of display has two advantages that are not confined to flexible displays. One is that the electronic ink needs very little power to maintain an image. The other is that the display is reflective: it uses ambient light rather than light emission to form an image. Again, that reduces power. It also makes the display look more like paper, which should reduce eyestrain.

Most first-generation flexible displays use e-paper, with electrophoretic technology for low-power capabilities. Although alternatives are available, most commercial products use e-paper materials supplied by E Ink. (Source: LG Display Co. Ltd.)

Turn on the color

The e-paper displays are, for the most part, monochrome because color displays require an additional filter layer. However, LG opted for the e-paper approach for its prototype color display rather than the organic light-emitting diodes (OLEDs) picked by some other manufacturers keen to introduce color displays. “When color comes out, it will massively increase the value of the market,” said Jennifer Colegrove, senior analyst for displays at iSuppli (El Segundo, Calif.).

There are several possible approaches to making a color electrophoretic display, one of which is to put different colors into the capsules. The problem comes when trying to assemble them into a display. The cheapest way to make an electrophoretic display is simply to scatter the capsules across the display — the electrodes underneath determine the pixel size. To make a color display, you need to place the different colored capsules with high precision.

Most e-paper displays are monochrome because color displays require an additional filter layer, but LG is using an e-paper approach for its prototype color display. (Source: LG Display Co. Ltd.)

One option, Lee said, is to put color filters over conventional monochrome electrophoretic capsules. “But the filter absorbs light, so you need to improve the reflectivity and the contrast factor,” he said.

Another option is to use electrodes of varying sizes to selectively control the spread of pigment particles across the microcapsules. “But it is hard to match the shutters with the pixels,” Lee said.

For its prototype 14 in. color display, LG decided to work with color filters, despite the contrast penalty, and use electronic effects to improve the apparent color performance. The basic color image does not give good color reproduction, according to Lee. But, by increasing the apparent color saturation using image-processing techniques, it was possible to get a better looking image. “E-ink only has 256 levels of color, so we use dithering to reduce the stairstep effect,” he said.

The current display is about the size of an A4 sheet of paper. “In the future, we will make large-sized public-display applications,” Lee said.

OLEDs have the advantage of being able to support color displays, and a coming generation of displays from suppliers such as Sony and Universal Display will use OLEDs. This is likely to be the technology favored for wall-hanging TVs, although Universal Display is designing its OLED-based modules for military applications.

OLED displays present manufacturers with the choice of sticking with silicon transistors or moving over to polymer electronics. Sony has decided to stack the OLED elements on organic transistors, while Universal Display is using an amorphous silicon backplane underneath the OLED layer.

Seeing the activity in flexible displays in other parts of the world, the U.S. Display Consortium (San Jose) has accelerated its collaborative research programs and renamed itself to reflect its new focus, becoming the FlexTech Alliance in early July.

“The EU has to be complimented and congratulated for having a lot of foresight with its science and technology programs,” said Michael Ciesinski, FlexTech CEO. “They have put an enormous amount of time into their framework research programs to support flexible electronics. And we are seeing R&D efforts by governments in other parts of the world, such as Singapore and Taiwan. There is significant recognition of flexible electronics outside the U.S. That is why the FlexTech Alliance should have a strong advocacy role to support this industry in the U.S.”

As countries ramp up R&D in flexible electronics, the focus will be on the early entrants to see how they overcome difficulties with yield, performance and manufacturability to drive a new class of electronic display.