We hope that you find this Newsletter
interesting & informative reading. Please feel free to
contact me with any questions or comments. If you wish to
"un-subscribe" from future mailings, please use my e-mail
link and type "UN-SUBSCRIBE" in the subject line, and your
name will be removed from future mailings.
Steve Silverman, President
Bartlett
Bay Consulting
___________________________
Coming
soon from Bartlett Bay Consulting!
(Back
to Top)
___________________________
|
Contamination Control
Questions:
What are the biggest contamination control
problems in the semiconductor
industry? |
| |
As I see it, the biggest problem is with ignorance of
the specifications: not only of their existence and
where to find them, but also, their meaning. Take a very
simple case: “PWP” (Particles/wafer/pass) referring to
the number of particles added to the wafer by a
particular operation or process tool. The question is
exactly which operations are involved: the robot? The
LoadPort? The Pre-Aligner? The process tool itself? The
list goes on, but whenever one looks at a PWP spec, it
is not clear how/where it is measured (Particles added
to the backside is in an even worse situation). SEMI
E14-93 was some help, but that has been withdrawn from
the SEMI Standards. The good news is that there are
resources that can offer help, but generally, they are
not focused in any one place. It has been my hope,
incidentally, that this newsletter will provide some of
this focus, and become such a resource.
(Back
to Top)
|
Will
the contamination control specs continue to decrease?
Is there a “limit”? |
| |
Yes, specs will continue to decrease because chip
design/lithographic parameters must keep in step with
Moore’s Law (doubling chip speed/capacity every 18
months). Consequently, as each new chip generation gets
smaller and more sensitive, their sensitivity to defects
gets more acute. This has been the basis for all
contamination roadmaps. So if we believe that Moore’s
Law will continue, and scientists think it will be valid
for at least another 5-6 years, then the contamination
control specs will also continue to track these changes
for this time period. The limits, however, are quite
another story: not only do we have to be concerned that
both airborne and PWP (particles/wafer/pass) demands
change by about an order of magnitude (!) every 4-5
years making achieving these limits a Herculean task,
but we also have to be concerned about whether the
measurement instruments will be able to measure these
small particles. Right now, for example particle
measurements and specifications in the air and on wafers
are in the range of 80-100nm, and laser measurement
tools are available for that, but in the near future,
particle sizes will be in the 40-50nm range, and this
increasing sensitivity places considerable pressure on
the instrumentation manufacturers and the
testers.
(Back
to Top)
|
How does one best
characterize/measure a minienvironment?
By airborne
particles? By PWP? Or other
factors? |
| |
Actually, there are a number of factors that should
be considered. While the ultimate test is the PWP, this
test is so expensive and time consuming that a
contamination control engineer would not want to perform
the test unless there is a high probability it would
pass; consequently, airborne measurements, and others,
are performed first. Such measurements include
differential pressure, and air velocity. Each is an
indicator of the overall design of the minienvironment.
In addition, a good contamination control engineer will
be able to assess other important, but non-quantifiable
items, such as design inadequacies, which could
contribute to high PWP values.
(Back
to Top)
|
| Can I use a ceiling HEPA rather
than an dedicated Filter Fan Unit (FFU) for my
minienvironment? |
| |
Yes. BUT there are potential problems. Factors
which should be considered are:
1. Does the minienvironment have a good seal to
the ceiling:
Too large a gap between the top of
the minienvironment and the filter can allow leaks at
this gap resulting in “dirty” air entering into the
minienvironment. Minienvironments with built-in FFUs are
designed/sealed to specifically prevent this problem
2. Does the Filter provide enough
air:
Air velocity at the FFU filter face in a
minienvironment usually is set at 0.4-0.5m/sec such that
the velocity at the wafer level is in the 0.2-0.3m/sec
range; this may be hard to do with some ceiling
heights.
3. Will the velocity at the filter change because
of changes in room design/load:
This is a very
common problem: new cleanrooms with adequate air volume,
for example, often change, as increasing demands on air
handlers in the ceiling can result in the filters
providing less air volume.
Net: If I had a choice between using existing
filters or a dedicated one, I would go for the dedicated
one as there is much more control and a larger margin of
error.
(Back
to Top)
|
| Is my
cleanroom really clean or are there areas which could be
problematic and how can I find
them. |
| |
If the cleanroom has been certified by a qualified
company/organization, then it is probably clean, and any
trouble spots would have been found, identified, and
probably fixed during the certification. However,
cleanrooms change (this is one of the strengths of
minienvironments: that it is a fixed environment) due to
new tools or tools being moved around, exhaust changes,
etc. If your cleanroom hasn’t been measured/certified
for 3-4 years, it is probably time to do it again.
(Back
to Top) |
|
