Water treatment is 'nontrivial,' says researcher

Despite the buzz surrounding the rapid development of immersion lithography (IML), employing the technique may not be as easy as water falling off a duck's back. For instance, using DI water is not enough to ensure clean optical surfaces in IML, says Michael Switkes. His research group at MIT Lincoln Laboratory, which is exploring water for use in immersion lithography, told participants at a workshop sponsored by International Sematech in late January that the group has uncovered several water quality issues. In addition to residue from DI water, these concerns include the interaction between water and the optics, gas content, particulates, and bubbles.

BUBBLE TROUBLE: Simulation of the effect of an air bubble suspended in water shows a shadowing-type effect, a potential defect cause. SIMULATION IMAGE BY MICHAEL YEUNG, BOSTON UNIVERSITY; USED WITH PERMISSION OF MIT LINCOLN LAB

Water treatment, for instance, is "nontrivial," Switkes concluded. Asked to clarify, he replied, "We tried pretty hard to get water that was clean enough and that didn't leave residue the first time around. We thought we had done a good job. We measured all the things we thought we ought to measure." However, referring to a presentation slide that shows two red arrows pointing to areas of optical contamination, Switkes says, "We measured resistivity to look at ion content. We measured so-called TOC [total organic carbon], and it looked very good from all of those points of view. But it still left residue on our optics."

The team explored further. One step they took was to measure nonvolatile residue, "which is by no means a trivial measurement," Switkes says. With its industrial partner, Mykrolis, the researchers examined different points in the water treatment system. Referring to a later slide of the same optic, Switkes points out that the spots are missing, "meaning that the water was leaving a film deposit behind before we got rid of it. It's not a trivial exercise; it took careful measurement."

Bubbles are potentially another "novel source of defects. We've been looking at a couple of places where bubbles might come from," particularly in 2200 outgassing images using TOK resist. He says that the presentation at the workshop focused on the possibility of bubbles from outgassing. "So far, we've tested only one kind of model of resists. It's not an enormous problem. It might be an issue with other resists."

Another concern is whether moving a wafer very quickly underneath the water will cause "some air or gas ambient. . . to be entrained into that space." He notes that Kodak has considerable experience in this area because of its expertise in film coating. "We're just starting to understand how to transfer that knowledge. We don't yet have a sense whether it's totally ridiculous to even worry about or whether it's a major issue that needs to be dealt with. It's probably not a showstopping issue, or we would have heard about it from the people who are sliding wafers fast under the water."—JC