450 mm Wafers: Contamination and Litho Focus

Although plenty of people will tell
you that the switch from 300 to 450 mm wafers is going to have to
happen at some point, and most likely around the 2012 timeframe,
the conversation is far from over. I understand the reluctance to
embrace this change from a tool manufacturer’s point of view,
and probably even more so from a wafer supplier’s point of
view, but I hadn’t thought about the impact that 450 mm
wafers would have when it comes to particle contamination and the
already shrinking lithography process window.

I was out at Sematech’s
Surface Preparation and Cleaning Conference (SPCC) last week in
Austin, and had an interesting conversation with a wet cleans
engineer at a major memory manufacturer. The point he made was
about how much more important it becomes with a 450 mm plate to
avoid backside contamination.

Already, with 300 mm wafers,
particles making their way between a wafer and a chuck can distort
a wafer, bringing some regions out of best focus. With a bigger
wafer, those problems would just be compounded. Say, for example, a
particle gets under a wafer. By the time you reach the edge of the
wafer, it could be pretty out of focus. If that wafer were a 450 mm
rather than a 300 mm wafer, it would be even more out of focus. I
put together a quick sketch of this example. Although this is a
very simplistic view of the problem, I just wanted to illustrate
the difference in height at the edge of the 450 mm wafer vs. the
height at the edge of the 300 mm wafer.

Another issue the engineer mentioned
was that throughput trumped everything for memory manufacturers,
because DRAM is so cheap. The idea of scrapping a 450 mm wafer vs.
a 300 mm wafer sounded horrifying to him.

What other issues are there on the
300 vs. 450 debate within the world of lithography? What are your
thoughts and insights about wafer sizes, productivity,
contamination, etc.? I’d really like to understand the issues
more thoroughly, and would love to get your feedback.