In the April 6, 2006 issue of the respected British journal Nature, Seiko Epson Corp. and JSR Corp. announced they have succeeded in creating the world’s first high-quality silicon film with liquid-coating and inkjet-patterning processes. The companies demonstrated that low-temperature polysilicon thin-film transistors (LTPS-TFTs) produced with a spin-coated silicon film have performance that is comparable to that of LTPS-TFTs made from the conventional chemical vapor deposition (CVD) method - which has not been used on substrates larger than Gen 4.

Ken Werner
Senior Analyst and Editor
of MDTV Retailer

Why does this matter? The existing process for manufacturing the TFTs used in LCDs requires enormous vacuum devices to form the film and expensive photolithography equipment to pattern in the film, which is expensive and time consuming.

One way around this would be to apply the semiconductor material directly to the substrate using some kind of precise printing process - say, ink-jet printing. This would avoid the need for both vacuum deposition and photolithography equipment and save substantial energy and processing time. But until recently it seemed that the only way to form TFTs from liquids would be to use organic semiconductors. And while there has been a lot of research in that direction, the performance and reliability of existing organic TFTs are not yet sufficient for high-volume display manufacturing. The Epson-JSR process allows the ink-jet approach to be used with tried-and-true silicon, and to make high-performance silicon film from a liquid material, which, says Epson, could be the key to solving all of these issues at once.

The new material announced by Epson and JSR is a high-order silane compound - silicon-hydrogen chains dissolved in an organic solvent. It forms a silicon film when spin coated on the substrate and baked in an inert atmosphere. When a TFT prototype using such a spin-coated film that is processed like a conventional LTPS TFT, the electrons achieved a mobility of 108 cm2/V-s, which is similar to a conventional LTPS TFT with silicon film deposited by CVD.

Epson also tried depositing the material by ink-jetting and then processing it conventionally. In this case, the electrons mobility was only 6.5 cm2/V-s. That’s not as good as it needs to be, but it’s still several times more than amorphous silicon that is used in the majority of active-matrix LCDs sold today. Certainly, Epson believes that printed silicon TFTs show sufficient potential to justify further development.

This is a potentially significant R&D development, but it’s a long way from being a manufacturing process. Will organic semiconductors get to market first, and if they do, will that make liquid-silicon development less attractive? We could see the ink-jetted TFT world breaking into (at least) two parts. The first, liquid silicon for high performance at lower cost than is currently available; and the second, organic, for highly flexible displays where image-quality requirements are less stringent. But we’re not at the product stage yet.