Return to this week's Bulletin
See The "Plumber's Special" Sorption Pump by Roy Schmaus.
07 December 2001
Working in a Vacuum
by Shawn Carlson
adapted from The Amateur Scientist, October 1996.
THERE'S NO WAY AROUND IT. Sooner or later, every serious amateur needs a vacuum system. Vacuums are crucial if you ever want to experiment with particle beams or make your own optical filters or radiometers, to name a few projects. The systems, however, have a reputation for being complex and costly, discouraging many amateurs from bringing vacuum techniques into their laboratories. But this need not be the case. Vacuum systems adequate for many scientific needs can be easily built and inexpensively maintained. Here's how to construct a system capable of achieving pressures as low as one ten-millionth of an atmosphere.
Figure 1: EVACUATING A GLASS CANNING JAR is achieved with molecular sieve pellets. A plastic shield or a doubled-over pillowcase protects in case of implosion. (click image to enlarge)When it comes to vacuum vessels, think small. Low volumes are easier to seal and pump down. A smooth glass canning jar (having no designs, artwork or scratches, which can weaken the glass) makes an adequate chamber for the vacuum. From a scrap-metal yard, purchase a one-inch-thick aluminum plate to serve as a base. It should be larger than the jar's lid. Secure the lid to the base plate with a generous helping of aluminized epoxy. (If your local hardware stores don't carry it, call Devcon in Danvers, Mass., at 508-777-1100, for the nearest distributor.) The epoxy should ooze out evenly from around the lid when the lid is pressed into place under the weight of a few old books. Wipe away the excess and let the epoxy set.
Next, drill a hole one quarter inch in diameter through the center of the lid and the base plate. If possible, tap the hole to give it threads. Obtain a one-quarter-inch-wide threaded pipe from a hardware store. Coat its threads with epoxy, then screw it through the bottom of the base plate. If you can't tap the hole, just glue in an unthreaded pipe. Draw a bead of epoxy around the pipe as it is inserted to make sure the gap is completely filled with epoxy.
Cut a half-inch-wide hole in an old card table and rest the base plate on the table so that the pipe hangs down through the hole. The pipe's end should be about 10 inches from the floor. If the pipe's end has threads, cut them off and file the edge smooth.
Canning jars are designed to hold a vacuum, so you will most likely be able to screw the jar right into its lid. If you need pressures approaching 10 millionths of an atmosphere, you may want to take special precautions against tiny leaks. You can place a layer of Teflon tape (check your local hardware store) over the threads on the jar's lip before screwing it in. It may be necessary first to put a bead of vacuum grease along the rim of the jar's mouth to ensure an airtight seal. The grease is available from Duniway Stockroom Corporation in Mountain View, Calif. (800-446-8811 or 415-969-8811).
Precautions are needed in case the jar implodes. (It eventually will if you conduct enough vacuum experiments or if the jar has some structural weakness.) On implosion, small glass fragments could hurtle out at nearly the speed of sound! It is therefore absolutely vital that you always keep your vessel under a protective shield whenever you pump it down. If you don't need to see inside, a doubled pillowcase affords the necessary protection. Otherwise, cover the jar with a clear, thick-walled plastic container, such as a three-liter plastic soft-drink bottle with its neck cut off. Additionally, Ace Glass in Vineland, N.J. (800-223-4524 or 609-692-3333; catalogue no. 13100-10), sells a protective plastic coating that will hold the glass together in case of a catastrophe. Half a liter will run you about $28 and is well worth the cost for the protection. Use it in addition to, not in lieu of, a shield.
For many applications, sorption pumps are the vehicles of choice for creating a good vacuum. They have no moving parts; instead they work by chilling a type of substance, called a sorbent, to a temperature at which it absorbs gases. Activated charcoal works, but a molecular sieve is better. Molecular sieves are little pellets with so many microscopic nooks and crannies that they have fantastically large surface areas; a one-gram pellet may have more than 1,000 square meters of surface.
Figure 2: HEATING THE MOLECULAR SIEVE drives off any moisture in the pellets. (click image to enlarge)When chilled, air molecules get caught in these microchasms. A 50-gram supply can pump a one-liter volume down to 10 millitorr in 20 minutes. (Atmospheric pressure is about 760 torr.) Half a gallon of molecular sieve from Duniway Stockroom sells for about $35.
To hold the sorbent, you need to obtain a Pyrex bulb approximately one inch in diameter and three and a half inches long, with a one-quarter-inch glass tube neck. A local glassblowing shop will probably make you one for less than $30. Fill it with the sorbent, then stuff in a little glass wool on top to keep the molecular sieve in place. Over the neck of the glass tube, slip a short length of flexible tubing, called Tygon tubing (check your local hardware store).
Before it can be used, the molecular sieve must first be activated—that is, it must be baked. Wrap the bulb with heating tape, available from Omega Engineering in Stamford, Conn. (800-826-6342 or 203-359-1660; model no. FGS0031-010). The 12-inch-long piece sells for $20. Or cannibalize an old toaster for its heating element. In either case, be sure that the heater does not cross over itself and that all of it touches the bulb. Wire in a dimmer switch to control the temperature of the heater.
To monitor the temperature, use a thermocouple probe (Omega, model no. 5TC-GG-J-30-36, $33) wired to a digital voltmeter. Place the probe against the bulb between windings of the heating tape and then wrap the bulb with aluminum foil. Safely secure the bulb so that the neck points downward and turn on the current. Adjust the current so that the voltage from the thermocouple increases by 18 millivolts, the signal that the sieve has reached the correct baking temperature of 350 degrees Celsius. The heat drives off the trapped molecules, including water vapor, which will condense on the bulb's neck and drip out. Leave the heater on until the neck is completely dry. Turn off the heater and pinch off the Tygon tubing to prevent the sieve from absorbing moisture from the air while the bulb cools. And you're ready to connect it to your vessel.
You will need to chill the sorbent with liquid nitrogen. Don't worry--liquid nitrogen is inexpensive (less than $1 per liter) and easy to obtain (try the Yellow Pages under "Welder's Supplies"). It can be safely handled if you exercise some common sense. Store it in a large plastic drink cooler--10 liters will last a weekend. Make sure the container does not have a spigot at the bottom. Do not put the lid on tight, or else pressure from the boiling nitrogen will build up inside and burst the container.
To pump the air out of the canning jar, immerse the Pyrex bulb in the liquid nitrogen. The molecular sieve will suck the air out of the glass chamber, producing a vacuum as low as 10 millitorr.
A few hints. Thoroughly wash and dry the vacuum-vessel assembly before using it, making certain not to touch the inside with your fingers. I'm told that a fingerprint can outgas (evaporate under low pressure) for years if not removed. To drive off moisture, bake the vessel above 100 degrees C for an hour. The epoxy will also outgas, as will any plastic seals in the lid of the canning jar and any coating on the inside of the lid. Minimize the surface area of these materials exposed to the vacuum. If more than about one square centimeter of any of the materials is exposed, consider coating it with vacuum grease, which outgases at a much lower rate.
You can insert a vacuum gauge between the sorption pump and vessel. To measure pressure in the tens of millitorr range, you'll want a thermocouple gauge or a Pirani gauge. These devices exploit the fact that the thermal conductivity of a gas drops sharply from a constant at about one torr to essentially zero at one millitorr. You can purchase a complete thermocouple gauge from Kurt J. Lesker Company in Clairton, Pa. (call 800-245-1656 or 412-233-4200) for about $200. The electronically inclined can save about $150 by buying a type 531 thermocouple vacuum tube for $45 (part no. KJL5311) and then building a simple power supply and amplifier circuit. Pirani gauges, however, are much more versatile and are quite easy and inexpensive to build.
As a service to Scientific American readers, the Society for Amateur Scientists is offering a complete sorption pump kit, including a Pyrex flask packed with a molecular sieve, heating tape, a small liquid-nitrogen holder and Tygon tubing. A vacuum vessel is not included. The cost is $60 for domestic orders, $70 for international ones (shipping included). This offer expires September 30, 1997.
For more about vacuum systems, visit the SAS World Wide Web site and the Bell Jar's site I gratefully acknowledge insightful conversations with George Schmermund, an amateur scientist from Vista, Calif., and with Steve Hansen, editor of the Bell Jar, a newsletter of vacuum experiments and the best amateur science quarterly I've seen.
Further Reading
Procedures in Experimental Physics. John Strong. Lindsay Publications, Bradley, Ill., 1986 (originally published in 1938).
Building Scientific Apparatus. J. H. Moore, C. C. Davis and M. A. Coplan. Addison-Wesley, 1989.
An Experimenter's Introduction to Vacuum Technology. Steve Hansen. Lindsay Publications, 1995.
The Bell Jar. Quarterly journal. Edited by Steve Hansen (35 Windsor Drive, Amherst, NH 03031). $20 per year in U.S., $23 in Mexico and Canada, $29 elsewhere.
