UHV - Ultra High Vacuum
06/07/07 15:19 Filed in: General Science | Work Experiments
Cleanliness is next to godliness.
This is one of the axioms of surface science. In order to study a few atoms at a surface or interface, it really pays not to have a huge number of other atoms running amok.
As soon as a surface is exposed to "normal" conditions, you can bet a huge number of contaminants have impacted the surface and, if possible, stuck to it or altered it in some way. In order to keep it clean, try to get rid of any air and contaminants by doing everything in a vacuum environment. A clean vacuum is the ultimate "insulator," giving you the time needed to do an experiment before anything can impact the sample. An interesting (and important) question is "what is clean?" or in other words, "how good must the vacuum be?"
Pressure in the chamber is directly related to sample integrity. If there is higher pressure, there is more contaminant available and more collisions per second are possible between molecules and the surface. In fact, it's a fun little calculation to estimate how low the pressure needs to be before reasonable experimental conditions are met.
Is 1/1,000th of normal enough? Perhaps 1/1,000,000th is good enough? No and No. In fact, by practical circumstances 1 part in 1 billion of normal atmospheric conditions is only where things become reasonably possible. And in fact, we usually need the pressure much lower. The pressure in the chamber directly relates to the number of random collisions on the surface. It comes out that even at 1 billionth atmospheric pressure, the surface will stay clean for only a second!
Therefore we really need to operate and maintain a base pressure much lower. While the effects we wish to study may in fact be quite fast, the operational speed of the experimenter is set on more "human" timescales. The rule of thumb here tells us that each factor of 10 that we go down from 1 billionth of atmosphere, buys us a factor of 10 in time. Therefore, at a hundred-million-millionth of normal atmospheric pressure we've got a precious day of operation with a clean environment. That's 0.000,000,000,000,01th of normal atmosphere!
Obtaining pressures that small is, frankly, hard to imagine! Pumping out a vacuum chamber involves many tricks, not all of them so intuitive. When it's all done correctly, we can get a chamber from atmosphere to operational pressures within a few days. Most of it isn't so difficult anymore. The normal mechanical "roughing" pump pulls out the first few factors of 10. Once the pressure is low enough, a turbo mechanical pump kicks in spinning its rotary blades at 1-2kHz. The rotary pump usually runs into a wall after getting us about a factor of 10-100 below a million-millionth (that's 10^-10 or 10^-11 of normal atmosphere. Obtaining those last few factors of ten is what really takes some work and crafty tricks.