by Robert P. Donovan
The first thing we do: Let's get rid of all the dirty people. Human beings have a bad reputation in the cleanroom world. They are viewed as unpredictable, moving sources of contaminantshard to model and harder still to control.
Control technology exists but implementation remains a struggle between the good guys in management, located a floor or two from the scene, and the cantankerous, rebellious masses in the bunny suits.
Let me review some of the problems people pose for a cleanroom contamination officer. People can drool over wafers, as documented in well traveled micrographs of spittle on a chip published by Bob Thomas from Rome AFB a score of years ago.
People emit particles just by breathing. Smokers in general exhale a lot more particles in the hour or so after their last puff than those who don't smoke. When walking, hurrying or otherwise moving through the cleanroom, all people emit particles at a higher rate than when sitting and also introduce turbulence that upsets the desired, air-cleansing unidirectional flow streams of the cleanroom.
Particle emissions from people are not going to go away. However, the standard cleanroom garb the bunny suit significantly reduces the particle burden contributed by cleanroom personnel. I recall being impressed several years ago when I could not register increased particle counts on a particle counter even as I jumped and danced in my bunny suit. Snapping my uncovered fingers in this same vicinity increased the particle count almost immediately. This experience convinced me that bunny suits work.
Fab managers have issued behavior guidelines (walk slowly, shield wafers from face, etc.) thought to minimize people particle emissions. Many companies impose constraints on where and when smokers can light up. The growing popularity of fully enclosed helmets for cleanroom personnel in which exhaled breath is discharged into the room through a HEPA filter addresses the particle problem of the exhaled breath.
Human vapor emissions, on the other hand, are not eliminated by this arrangement. I was once the guinea pig in a test series to measure the effectiveness of bunny suits in suppressing particle emissions from the wearer. The test chamber used was enclosed and sealed to isolate the inside environment from the laboratory ambient. The only air entering came through HEPA filters located in the roof of this box-like chamber and the only exhaust was through the air sampling lines leading to particle counters outside the enclosure.
I stepped in fully suited. Coworkers outside the chamber monitored the particle counters for evidence of increasing particle concentration within. I was presumed to be the only particle source in the chamber so that any particles detected would be interpreted as leaks or shortcomings in the protective garment I was wearing.
This worked fine in that few particles were detected until I began the exercise portion of the test. I jogged for 10 minutes at which time the particle concentration skyrocketed. My first reaction was that this particular bunny suit wasn't performing very well until I noticed water running down the inside of the chamber walls. It became clear that what the particle counter was registering was the condensed vapor that was pouring out of my overtaxed, sweating body. This really wasn't what we were looking to measure and the measurement plan based on this particular apparatus and test protocol was abandoned. Anthropogenic, condensing moisture, however, remains a potential source of particles in a cleanroom.
Despite these examples, people emissions are no longer thought to be a primary particle source in a state-of-the-art cleanroom fab see my column in the October 1999 issue of CleanRooms. Nonetheless, the long-term solution to this now admittedly second-tier particle problem would seem to be to eliminate all people from the fab, and indeed this movement has been underway for some time. But, alas, the robots that replace people can also emit particles so a new set of design rules and control tactics arise. Bunny suits somehow seem inappropriate for robots.
|
Robert P. Donovan is a process engineer assigned to the Sandia National Laboratories as a contract employee by L & M Technologies Inc., Albuquerque, NM. His Sandia project work is developing technology for recycling spent rinse waters from semiconductor wet benches.