A User's Guide to Vacuum Pumps Part 2: Quiet Pumps

In the previous post on this topic, I discussed the various types of noisy vacuum pumps, both clean and dirty varieties. This time out, we'll deal with the quiet pumps, the ones that don't deafen people working in the lab.


Quiet and Dirty: The quintessential quiet and dirty pump is an oil diffusion pump. These have no moving parts, but operate by heating a low-vapor-pressure oil inside a series of baffles so that it sprays out in a downward jet. The oil jet will collide with air molecules in the way, and force them back into the reservoir region of the pump, where they can be pumped away by a mechanical backing pump.

Properly designed diffusion pumps can have very high pumping speeds, but they rely on contact between the oil and the gas being pumped in order to work, which means that they're very messy. The oil tends to creap out and coat absolutely everything if you're not careful, and even if you are, you'll still get oil on the inside of your vacuum chamber. Diffusion pump oil is nasty stuff, too-- you don't want it on your hands, if you can avoid it.

On the bright side, diff pumps are essentially free-- people who start out using them tend to switch to turbo pumps as soon as they have the funding, and are usually perfectly happy to have somebody take the used diffusion pump off their hands.

Straddling the line between Quiet and Dirty and Quiet and Clean pumps are titanium sublimation pumps and other types of getter pumps. "gettering" consists of putting a reactive substance (barium and cesium are popular choices) inside the vacuum chamber, where it will bind with reactive gases, forming compounds that coat the walls of the chamber, and removing the reactive gases from the vapor.

These are completely passive systems, and thus very quiet, and they don't generally involve enough material to contaminate the lab outside the vacuum chamber. They do, however, involve introducing extra stuff to the inside of the chamber, and reactive stuff at that, so if you accidentally vent the chamber to atmosphere, you end up with all sorts of oxides and other crap compounds in there, that need to be cleaned out. This can be a real pain, and puts them just barely into the "dirty" quadrant.

Quiet and Clean: The simplest clean and quiet pump is a sorption pump, which is basically a bucket of sand. The way it works is that you attach a container of zeolite to your vacuum system with a valve, and then cool the zeolite down to liquid nitrogen temperature, generally by dunking it in a bucket of liquid nitrogen.

When you open the valve between the sorption pump and a high-pressure chamber, the gas from the chamber will rush into the sorption pump, and freeze onto the surface of the zeolite, removing it from the vapor in the chamber. This lowers the pressure in the chamber as long as the zeolite is cold, and if you close the valve, you've removed those gases from the system entirely. Then you can let the zeolite warm up (or actively heat it), and drive those gases off into the general atmosphere, at which point it's ready to go again.

Sorption pumps are roughing pumps, useful only for getting a system down from atmosphere to a lower pressure in order to start some other sort of pump (a turbo or an ion pump, usually). They're very clean and quiet, though, and extremely robust.

The best example of a clean and quiet pump for continuous operation is an Ion Pump, which starts by ionizing a small amount of gas, and trapping the free electrons created in a large magnetic field between titanium electrodes. The electrons will collide with gas atoms in the vacuum chamber, ionizing them, and causing them to be pulled into the walls. Reactive gases will bind with the titanium, and be removed that way, while stuff that doesn't react chemically with titanium can hit the electrode hard enough to get physically buried in the electrode, which still counts as being pumped away.

Ion pumps have a lot of nice features: they have no moving parts, so they're very quiet and robust, and you can use the current flowing through the discharge to measure the pressure-- the more gas there is, the more ions you'll create, and the more current will flow between the electrodes. They're fairly cheap, and any idiot can run one.

The disadvantages of ion pumps are that they can't be run anywhere near atmospheric pressure, so you need some sort of roughing pump to get the pressure down to where they will start (sorption pumps are good for that). They also involve big magnets, which can screw up any measurements that depend on knowing the magnetic field well (they're permanent magnets, so you can shield the field out, but it's still a source of uncertainty). Most critically, for my puposes, they don't actually do that well at pumping noble gases-- they'll remove the trace amounts that are in air, but if you want to deal with krypton, argon, and neon, as I do, you can't really use an ion pump to get the job done.

And that's the wonderful world of vacuum pumps. This is probably more information than you really wanted, but now, at least, if you get waylaid by an experimental physicist who insists on talking about vacuum hardware, you'll have some idea what he's babbling about.

More like this

At my high school, the science labs had compressed air spigots at every station, fed by a common compressor. Nothing similar is done for suction?

By Johan Larson (not verified) on 11 Mar 2008 #permalink

At my high school, the science labs had compressed air spigots at every station, fed by a common compressor. Nothing similar is done for suction?

The lab in which I'm currently working has a similar system -- but we never use it. One problem is that if you've got system 1 pumping down and you want to start evacuating system 2, opening the valve to system 2 raises the pressure in system 1 until it can all be pumped out again. You could, of course, interrupt the pumping of system 1 by closing its gate valve until the pressure in system 2 is as low as that in system 1, but then you don't gain the full benefit of being able to work in parallel.

Ah............ an all-glass, triple-stage mercury diffusion pump........... it takes me back to those happy days in graduate school. The pump could be dirty if, you smashed it, or if the water flow to the condensor dropped. Either case, you ended up with a lab full of mercury vapor.

My company manufactures ion pumps. There are two main types of ion pumps, diode (positive polarity) and triode (negative polarity). If you have noble gases such as argon, you will find that the triode does not experience "argon instability" and this pump is best if argon is a problem, although the "noble" diode counteracts the argon problem but is more expensive than the other pumps as it involves a tantalum cathode. For pumping of systems which do not involve a great amount of noble gas pumping,i.e. argon etching, a diode can be used.The diode gives a better speed at low pressures.