04 October 2002

Making Plastic Vacuum Chambers

by Joseph DiVerdi

I've been working for some time now on a project to build a few large area ionizing radiation detectors (inspired by an article by Shawn in an old Amateur Scientist column). A good deal of that time has been involved with vacuum chamber construction including developing reliable seals, keeping the leakage rate extremely low, and developing a satisfactory (un)fill port.

To work out some of the practical details of building vacuum chambers I have built a number of small vacuum chambers made of acrylic plastic, approximately 15 x 15 x 3 cm on the outside, with a 13 x 13 x 1 cm vacuum space on the inside fitted with a dial-type vacuum gauge and connector for attachment to a vacuum line. An example of one such chamber is shown in the following figure.

Test Vacuum Chamber

Click image to enlarge

My local vendor of plastic suggested that I use "acrylic solvent" to assemble and seal my chambers. The solvent consists of several interesting components (methylene chloride, trichloroethylene, and others) which soften and dissolve the acrylic plastic. When two plastic pieces in such a softened state are placed in contact and allowed to harden they form a bond which is strong, permanent, gas-tight, and almost indistinguishable from the bulk plastic. In typical practice, the pieces to be joined are fitted together dry and the very mobile fluid is applied to the joint with a fine dropper or pipette where it is rapidly drawn into the joint through capillary action. After a few minutes, the joint is made but full strength usually takes a few hours to develop.

Joints made in this fashion are as gas-tight as the bulk plastic and assemblies made in this fashion are more than adequate for many soft vacuum applications. However, these joints possess the unfortunate tendency to out-gas, that is for solvent in the body of the joint to slowly and continuously evaporate and contribute to gas pressure in the chamber. This can be a serious problem for certain applications where a volume bound by a relatively long length of joint must be maintained at a fixed, soft vacuum for months or even years at a time without external intervention. Protracted periods of pumping and mild baking help reduce the out-gassing but do not eliminate it.

Good joints which do not out-gas can be made using an RTV-style silicone sealant as described in Shawn's article. Unfortunately, I have found that, at least in my hands, these joints lack mechanical strength and must be supplemented by the addition of a series of nuts and screws which run along the joints as can be seen in the previous figure. To build these assemblies sealant is generously applied to all the joints and the components are fitted together with the nuts and screws. As the nuts are tightened the sealant conforms to and fills the joints and some will ooze out of the joints. Judging the amount of sealant and the degree of tightness requires some experimentation. I typically advance to hand-tight with a screwdriver-type nut driver. The joints become gas-tight as the sealant cures over several hours.

Once a gas-tight chamber is obtained the issue of developing an external vacuum connector remains. While quarter-turn brass stopcocks are often used in this application, I sought a better solution which is inexpensive, offers no opportunity to accidentally break the vacuum and contains as little radiation obscuring material as possible. These goals were achieved with the following scheme. The following components are required:

The connector's male-NPT threads are smeared with silicon sealant and screwed into the acrylic chamber body. The polyethylene tubing is fitted with the proper plastic ferrule, wrapped with a bit of PTFE tape, and inserted into the compression fitting which is then firmly tightened. This tubing is attached to the vacuum line. After the chamber is filled to the desired pressure with the desired gas mixture two screw clamps are securely applied to the tubing - one, a few centimeters from the compression fitting and the second, a few centimeters further along. The chamber, vacuum line, and a small length of tubing are now isolated from each other. A small flame is applied to the tubing between the clamps to melt it. The molten tubing is drawn apart and the molten end is squeezed flat with the pliers. The tubing is thoroughly cooled to room temperature and then the screw clamp can be removed.

Vacuum Chamber Seal

Click image to enlarge

Besides meeting the design goals mentioned above, this sealing method offers the distinct advantage of being tamperproof and permanent but replaceable. The tubing can be removed and replaced at a later date and as needed. A completed seal can be seen in the following figure.

This completed chamber has been evaluated for leakage and has successfully maintained a pressure of around 16 cm of Hg (170 mBar) for several weeks at 20°C and occasional exposure to 4°C without material change in the pressure. Exposure to sub-0°C resulted in total loss of vacuum but no permanent damage to the assembly. After replacement of the seal and a short session with the vacuum line the chamber maintains a vacuum as well as before the low temperature soak.

One area of continuing work for me is the search for sheet material which can serve as a gasket and replace the silicone sealant. Then there would be no concern with oozing of the sealant. Any thoughts from the readership in this area will be greatly appreciated.

The many helpful discussions and generous loan of equipment from Dan Storey (dstorey@ionedge.com) and Gary Turell (gary@pressurepointer.com) in the course of this work are gratefully acknowledged.

Joseph DiVerdi enjoys building hardware, writing software, and generally making measurements. He can be reached at diverdi@xtrsystems.com.