SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Cymer (CYMI) -- Ignore unavailable to you. Want to Upgrade?


To: Jerome who wrote (22507)7/13/1999 1:45:00 PM
From: BillyG  Respond to of 25960
 
IBM preps R&D blitz for next-generation masks
eetimes.com

By David Lammers
EE Times
(07/12/99, 10:33 a.m. EDT)

BURLINGTON, Vt. — With a helping hand from Washington, IBM
Corp.'s Microelectronics Division and Photronics Inc. (Jupiter, Fla.) will
convert IBM's existing X-ray mask facility in Burlington, Vt., into a
research and development center in masks for next-generation lithography.
Technologies developed at this "mask center of competency" will be
commercialized at some point by Photronics, for use by any device
manufacturer.

The effort is envisioned as an "open shop" that will allow participants in the
semiconductor industry to join in devising mask schemes that cut across
lithography technology borders, including design-data preparation, mask
writing, metrology and defect detection and repair.

Though the work could be applied to any of the proposed next-generation
lithography (NGL) solutions, the center has a contract in hand from Lucent
Technologies to further the development of masks for the Scalpel scanning
electron-beam system. Scalpel, developed at Bell Labs, is a close cousin to
an e-beam system being developed by IBM and slated for
commercialization by Nikon Corp.

NGL masks are expected to leave today's chrome-on-glass substrate
behind for more exotic materials. Extreme-ultraviolet (EUV) lithography,
for example — the type favored by Intel Corp. — uses reflective masks
that will need 60 or more monotonic layers of different materials to gain the
level of reflectivity required. Like Scalpel and projection ion-beam
technologies, EUV will also require a silicon substrate. The difficulty of
creating such exotic masks has prompted many to say that the NGL
challenge lies not in lithography, but in mask making.

Seeded by a $9 million federal research contract Photronics won for the
1999-to-2001 time frame, the Burlington center will zero in on non-optical,
NGL mask technologies. Sen. Patrick Leahy, D-Vt., a member of the
defense appropriations subcommittee, spearheaded the effort to gain
federal support for the project. The award will be administered by Naval
Air Command Systems.

Schemes that emerge from the center's work could be applied not just to
e-beam systems like Scalpel, but also to the extreme-ultraviolet technology
championed by Intel and other members of the EUV LLC consortium, as
well as to the X-ray lithography IBM has long been incubating.
Representatives of IBM and Photronics said they will be meeting soon with
leaders of the EUV LLC in hopes of gaining their participation. The
partners also intend to drum up support at this week's Semicon West
show, which opens Monday (July 12) in San Francisco and San Jose,
Calif.

In the past year, momentum has built for 157-nanometer optical lithography
using a fluorine-based laser, and much of the litho research community has
homed in on 157-nm masks, including phase-shift and optical proximity
technologies.

For example, DuPont Photomasks Inc. (Round Rock, Texas) established a
$75 million advanced-research center two years ago, sharing costs with
Micron Technology, Advanced Micro Devices and Motorola. The partners
want to extend optical mask technologies through the 193- and 157-nm
generations.

The ascendancy of 157-nm optical research has left precious few funds
available for NGL mask R&D. "All of the major stepper vendors, including
ASML and Nikon, are working on what they think will be the
next-generation lithography solution," said Jack Moneta, senior vice
president of planning at Photronics' technology center in Allen, Texas. "But
nothing is really being done in the mask end of the business. That is a black
hole so far."

X-ray leverage

Indeed, with research dollars stretched among 157 nm and the various
NGL candidates, IBM came to realize that NGL mask development was
lagging, said Jaga Jagannathan, project manager for mask-technology
development at IBM Microelectronics. Jagannathan said the company
wants to continue its X-ray mask development effort and "leverage what
IBM has learned from X-ray masks across all of the NGL candidates."

For the 1:1 proximity masks required for X-ray lithography, IBM
developed a rapid e-beam mask-writing tool, the EL4 Plus, that could be
used to write any kind of NGL mask. And KLA Instruments created a
mask-inspection tool, called SemSpec, that is also in place at the Burlington
facility. "We want to gather all of these core competencies under one roof
to apply to the NGL mask challenge," Jagannathan said. "No one company
can afford to pay for this work [alone]."

About 20 people, most of them from IBM and Photronics, are expected to
soon start work at the new center here. IBM will contribute the facility,
people and certain patents, but will not directly fund the center's activities, a
spokesman said.

Photronics will manage the facility, which Moneta called "very unique, an
effort that will benefit the entire industry." The center will have an advisory
board that is open to other semiconductor vendors. However, because the
Department of Defense wants to keep advanced electronics for defense
applications in the United States, foreign participation will not be an option,
according to a Photronics spokesman. Moneta said participation by
DuPont Photomasks would not be welcomed, either.

That prompted Ed Reigler, executive vice president of planning at DuPont,
to say — perhaps only half in jest — that he might write his Congressman
to see why federal dollars were going to a competitor. DuPont
Photomasks, slightly larger than Photronics overall, has chemical giant
DuPont as a 50 percent owner, and benefits from materials research issuing
from the parent company's research center in Wilmington, Del.

The mask center will get up and running at the tail end of a dry spell for a
mask industry hit hard by the three-year semiconductor slump. In the first
six months of this year, Photronics' revenue was down 9 percent from the
same period a year earlier, to $102 million. About 15 percent of revenue
now comes from mask sets created for 0.25-micron geometries and below,
a ratio that will increase rapidly over the next year, the company said.

Photronics' Moneta and IBM's Jagannathan said their companies will spend
significant amounts on 157-nm mask technologies, leaving limited funds
available for independent NGL mask research. "The timing of 157-nm is
unsure," Jagannathan said, adding that if an affordable NGL mask solution
emerges it might "make a difference here."

The competition between 157-nm optical — with its complex phase-shift
and optical proximity correction (OPC) masks based on fluorine-doped
quartz materials — and the yet-to-be-tried NGL solutions will come to a
head at the 70-nm node. But some in the industry believe that even there,
optical techniques will have enough juice.

Reigler said many of DuPont's customers want to extend optical lithography
as long as possible. Many now believe that the krypton-fluoride-based
248-nm scanners, with phase-shift masks and OPC technologies, may be
extendable to the 0.13-micron generation, he said.

A few dozen commercial-use 193-nm scanners, based on an
argon-fluoride excimer-laser source, will ship this year to semiconductor
vendors. Some industry watchers believe that such scanners — outfitted
with phase shifting, OPC and chemical mechanical polishing — may extend
optical lithography to the 90-nm generation; 157-nm tools could be used at
the 70-nm node, this camp maintains.

"Our customers see their way clear to 70 nanometers. They are familiar
with the behavior of optics, and they see no need for next-generation
lithography solutions for seven to 10 years," Reigler said. He added that
Micron Technology (Boise, Idaho), which needs advanced lithography to
shrink the die sizes of its DRAMs, "has shown absolutely no interest in
NGL research at this time."

At the DuPont-AMD-Micron-Motorola Reticle Technology Center at
Round Rock, work is under way on developing masks for the Scalpel
system, under a contract with Lucent Technologies. But Reigler said the
financial stress in the mask industry has reduced companies' direct R&D
funding to only 5 percent or so.