BY HANK HOGAN
AUSTIN, TX-With the second quarter of 2005 already underway, research consortium SEMATECH (www.sematech.org) is looking ahead to technical challenges facing the semiconductor industry in 2006. From lithography and various device components to metrology, manufacturing effectiveness, and environment, safety, and health (ES&H), the consortium sees significant contamination-control concerns-particularly with lithography and new materials.
Texas Instruments Inc. (TI; www.ti.com) and other semiconductor manufacturers believe that 193-nm lithography immersion can take the industry down through the 45- and 32-nm nodes, which are considerably smaller than the current 90-nm node. With 193-nm immersion, water or another fluid-even oil-is used instead of air for the final optical element. Because the refractive index of the fluid is greater than that of air, it’s possible to resolve finer linewidths.
TI is conducting feasibility studies to see if it’s possible to go below 32 nm with 193-nm lithography immersion techniques.
“We’re used to putting wafers in water,” explains Bob Doering, technology strategy manager at TI. “But every geometrical situation is a little different.” Water must be injected and removed, with an exposure to laser light in between, and Doering says particulates could be one contamination problem to address as nodes get smaller. He’s optimistic, however, that overcoming the challenge won’t be difficult.
But a more challenging situation could arise if immersion lithography uses an even higher refractive index fluid such as oil. Squirting oil onto a wafer and then rapidly removing it could lead to new contamination-control issues. At this point, however, Doering says the use of oil is only in the discussion stage.
Other more realistic contamination-control challenges for future semiconductor processes, says SEMATECH, could involve the introduction of new materials. With the pace of new material inclusion accelerating, there is concern about cross-contamination. In addition, resists are getting increasingly sensitive to airborne molecular contaminants (AMC), with the result being an increase in the use of chemical filters and other means to scrub the air.
The problem, comments SEMATECH Lithography Manager Jeff Mute, is finding metrology methods that can determine when the air is clean enough. “You have to have analytical techniques capable to some parts-per-trillion level,” he says. And existing technologies may require sampling times of 48 hours or longer, and would be relatively expensive.
Neither characteristic is desirable for instruments in a manufacturing area, and the alternative to such imperfect instrumentation is to use the resist and lithography process itself as a test.
One benefit of this resist sensitivity is that those concerned with ES&H get a complete signature of the air inside a fab. So, for at least one aspect of contamination control, SEMATECH’s consensus is that ES&H won’t show up in next-gen semiconductor technology cleanrooms.
Walter Work, a SEMATECH Fellow who specializes in ES&H, sums, “We feel very confident of our environment as far as personnel goes.” III