In this post I show how we made a small homemade titanium sublimation pump for an 8” Kimball Physics spherical octagon UHV vacuum chamber.
Our little chamber has a 60 l/s ion pump, but even with baking (both IR and UVC), we were able to get only into the low 10-8 to high 10-9 Torr range. However, using the little titanium sublimation pump, which we “Frankensteined” together using parts we had readily available, allowed us to get in the low 10-9 to high 10-10 Torr range, a factor of ten improvement.
A titanium sublimation pump works by heating a titanium filament wire to about 1300 degrees C. That is hot enough to create titanium gas molecules (sublimate) but not so hot that the filament wire melts. The sublimated titanium deposits on the wall of the chamber (or preferably on a shield wall) and forms a thin film. This layer of titanium is very reactive and will bond with other molecules in the vacuum chamber such as CO and O2. Disassociated hydrogen and water vapor also diffuse into the titanium layer.
The reactivity of the titanium film is increased with lower temperatures, but most titanium sublimation pumps are operated at room temperature. Over time, the titanium film will become coated and need to be replenished. All commercial titanium sublimation pumps have 3 to 4 filaments so that when one filament burns out you can switch to another. Those filaments are also relatively thick in diameter at 12 gauge (.080”) and need about 50 amps of current to operate.
For our homemade titanium sublimation pump, we used 24 gauge (.020”) so that we could operate at a much lower current of 4 amps. We also only have one filament.
Before I show how we made our homemade titanium sublimation pump, here are links to some videos on how TSPs work:
The first thing that we needed was a 2-pin electrical feedthrough on a 2.75” CF flange. For that we used a Getter pump flange from a PHI 04-303 ion gun as shown below.
We then needed to somehow support and electrically isolate the TSP filament wire. To do that we used a coupler and some little shoulder washers.
The top part of the getter pump assembly is conveniently designed to allow gas molecules to pass through but also block direct deposition of titanium into the vacuum chamber.
Next we added a few turns into the titanium wire so that it would have a little bit of a spring to it. Then we connected the wire to the flange and support assembly. We have only one filament and so by effectively doubling the length of the wire we could also double the amount of titanium that we would be sublimating.
For a chamber wall we used a 2.75” nipple that has a tube ID of 1.6” and a length of 4”. The larger the surface area the better, but for the size chamber that we have, we are limited to a small 2.75” nipple. We mounted this nipple on our chamber horizontally so that any flakes that form will not get into the chamber or ion pump.
For a power supply, we used a 30 volt 5 amp Lavolta.
After installing our homemade titanium sublimation pump into the chamber we pumped down and were ready to operate the TSP.
To operate our titanium sublimation pump, we slowly increased the power supply current while observing the color of the light coming off the TSP filament and also monitoring the chamber pressure. The filament needs to be orange for the titanium to sublimate. Too hot and the lifetime of the filament will be reduced. Too low and the pumping effect is reduced. By experimenting we determined that about 3.8 amps DC was the correct amount of current. Once that was determined, we could just periodically turn the TSP on for about 2 minutes at a time. We did that 3 times over a 6-hour period and then let the chamber pump overnight. The next morning we were in the high 10-10 Torr range. Success!
- It is possible to make a small titanium sublimation pump using off-the-shelf components that will operate with less than 5 amps of DC current.
- Adding a titanium sublimation pump to a small chamber can help to get from HV to UHV.