Japanese startup recruits backers to unique 100nm – 70nm litho approach

By Paula Doe

WaferNews Contributing Editor

A bold little Japanese startup has big plans to challenge all the big names with its own alternative approach to next generation lithography. LEEPL Corp., Machida, said it will have two beta tools out by the end of the year that will use a simple, low-cost combination of conventional e-beam and proximity x-ray lithography technology to write 100nm to 70nm lines through a 1:1 stencil mask.

“No one can believe we can make this 1x mask,” said LEEPL President Nobuo Shimazu. “When we show it, then maybe 5% will believe it.” NTT Advanced Technology (NTT-AT) made the demonstration mask with a conventional e-beam writer on a four-inch silicon wafer topped with a polycrystalline diamond thin film. “It was only one layer,” he explained, “So it was no problem. NTT-AT made it in three weeks.”

Shimazu himself used to be project leader for NTT-AT’s e-beam writer.

If nothing else, the new approach certainly knocks the socks off other N.G.L. approaches with the lightning speed at which it has moved so far. LEEPL started from zero a year ago, made a proof-of-concept tool and mask in four months, then convinced Japan’s major suppliers of masks and resists to all sign on to help develop the infrastructure needed for the new approach, and also persuaded chipmakers Sony, NEC and Rohm to try it out. TSK is developing a mask inspection tool for its stencil masks, under its new brand name Accretech.

Sony will take the first beta tool, while the other will go into TSK’s cleanroom for use by other companies working on developing the process. Along with Sony, NEC and Rohm, maskmakers Dai Nippon Printing, Toppan Printing, and HOYA, and resist suppliers Tokyo Ohka, JSR, and Fuji Film Arch have so far joined the LEEPL consortium to develop the process. Participants are not putting up any money to directly fund development of the technology, but are working together to share information.

However, the LEEPL tool is a vacuum system with proximity exposure. It will likely be a bit less challenging than 1X X-ray, but it will have to overcome many of the same problems. The mask problems may be different, but not easier. A stencil mask may be suitable for some things, including contacts, but mechanical stability may be hard to prove.

At Semicon West 2001, LEEPL was showing photos from Sony of 100nm resist patterns from the LEEPL tool, which looked to about match the quality of those from a leading 50kV direct-write e-beam tool, with roughness remaining along the line edges. Shimazu noted, however, that he’s still using regular e-beam photo resist, while the process needs a thinner resist film. Resist makers in the consortium are working on developing more appropriate resists. He also claimed LEEPL has gotten resolution down to 70nm and expects to be able to extend the process down to 35nm.

LEEPL says that its Low Energy E-beam ProjectionLithograhy (LEEPL) approach has higher throughput than the conventional-beam approach because it avoids space charge effect problems by using proximity printing, and its low-energy 2kV source eliminates problems with the proximity effect. The company claims it gets throughput of 40 300mm wafers per hour, about what F2 systems may get, and not too far off the 60 wph rate of advanced KrF and ArF systems.

LEEPL figures its first-generation system will cost about $6.5 million (?800million), compared to about $10 million for ArF tools, while its second-generation tool will be about $8 million (?1 billion), compared to perhaps as much as $20 million for F2 lithography tools. TSK owns just under half of the LEEPL venture; the remainder is controlled by companies owned by Shimazu and Takao Utsumi, founder and chairman.

“The LEEPL system is very, very simple,” said Shimazu. “The conventional way is no use, so revolutionary steps are needed.”


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.