| |
Thanks very much for the reference article (see below) showing the work done by SELETE/ASET to push optical lithography below 100 nm by using top surface imaging techniques. As has been discussed on other threads, the key issue is to produce smaller L/S to reduce die size (cost issue) in addition to just smaller gate (performance issue). Can the technologists comment on the viability of this (resurrected) process approach? If viable, then the lasermakers have quite a run before them.
Thanks
TI2
The article from EE Times:
ÿ
Direct Article Link | New Search | Search the Web
January 12, 1998, Issue: 988
Section: News
------------------------------------------------------------------------
ArF laser produces 0.04-micron isolated line -- Lithography advance cuts
patterns below 0.1 micro
Yoshiko Hara
Tokyo - A process technology that cuts chip patterns to the fineness of
a 0.04-micron isolated line has been established by a research
consortium here.
The Association of Super-advanced Electronics Technologies (ASET) took
an argon-fluoride (ArF) excimer laser beyond its wavelength limitation
to cut the finer patterns. More than seven of those finer lines could be
etched in the space taken today by the slimmest state-of-the-art line.
"No optical laser could have cut patterns less than 100 nm," said Masaru
Sasago, research manager at ASET's Hyper-fine Optical Lithography
Laboratory of ASET. But a process advance allowed ASET to "show the
potential of how far ArF can go."
ASET, an R&D organization, runs nine front-edge projects directly
supervised by Japan's Ministry of International Trade and Industry .
ASET researchers recently fabricated a 0.09-micron contact hole and a
0.04-micron isolated line using an ArF excimer-laser lithography
combined with a Top-Surface Imaging (TSI) process. TSI provided the key
to overcoming a resolution-limitation barrier beyond the laser's 193-nm
wavelength.
Compared with a conventional single-layer resist process, TSI requires
silylation-the implanting of a silicon compound agent-after the
exposure. Then the surface is dry-etched with carbon-dioxide plasma. In
this process, areas that were not exposed are etched. Even if the
exposed pattern is somewhat vague, remaining areas form finer patterns
by the etching.
The idea of TSI technology has existed for about 15 years. Progress in
materials and dry-etching opened the possibility to its use in volume
production, Sasago said. TSI is applicable to krypton-fluoride and other
potential optical lithography such as 13-nm wavelength extreme UV as
well.
Using the ArF laser with conventional single-layer resists, ASET
researchers cut patterns as fine as 0.13 micron.
ASET had already used TSI to fabricate 0.09-micron line-and-space
patterns last March. Now the researchers proved ArF could cut isolated
patterns-considered to be difficult with laser lithography-as fine as a
0.04-micron isolated line and a 0.09-micron contact hole. That's roughly
the equivalent of a process for a microprocessor with 160 million
transistors/cm2, according to the Semiconductor Industry Association
(SIA) technical road map. The density is more than 20 times that of a
microprocessor on a 0.25-micron process. The 0.09-micron line-and-space
patterns correspond to a 16-Gbit DRAM process.
"It may be possible for ArF lithography to go as deep as a 0.07-micron
L/S pattern," said Sasago. "I believe that ArF can be used for at least
three generations, in terms of DRAM: the second generation of 1-Gbit,
4-Gbit and 16-Gbit devices."
ASET's ArF lithography project started in February 1996. "At the
beginning, the target was 0.18-micron process," Sasago said. "But as KrF
lithography is now extending into 0.18-micron level, ArF is targeting
0.15 micron and deeper."
Sematech, a non-profit consortium of 11 semiconductor manufacturers
based in Austin, Texas, is running a similar project, Sasago said.
"Japan has strength in resist and process technologies, while Sematech
has strength in glass materials. As a result, ASET and Sematech could
work cooperatively and complementarily," he said.
ASET researchers are now working to establish the basic infrastructure
of ArF-based production by 1998 so that ArF-based mass production can be
launched by 2000 or 2001.
Sample shipments of resist have already started from multiple Japanese
manufacturers and prototype steppers will be shipped this year from
Nikon and Canon. "Except for mask technology, which is slightly behind,
all other technologies are in progress as scheduled," said Sasago.
ASET's result will be transferred to Selete Inc., an R&D company
established by 10 leading Japanese semiconductor manufacturers.
Selete will begin development and evaluation of ArF lithography systems
from next April focusing on practical application of the ArF lithography
in production lines. Selete set the target date for practical technology
to realize a 0.13-micron to 0.1-micron process by the year 2000.
ASET, a five-year project that began in 1996, is being financed by MITI.
Three American companies-Texas Instruments, IBM and Merck-are
participating in some of the research projects.
Among the research programs are efforts to develop technology for future
generations of disk drives and liquid-crystal-display screens as well as ICs. |
|
