I have made a fair number of Langmuir probes during the time I have been running experiments for my thesis. Some of the plasma facing material is aluminum, which I used because it was easy enough to machine that I could do it myself. The benefit of being able to do this myself is that the probe is built more quickly and it is easier to modify compared to the process of having the machine shop do the work. Granted, they do much better work than I possibly can, but for these probes my result is good enough even if it is not perfect.
One problem with using aluminum to make these probes is evident in the figure below. These are photographs of a probe after removing it from the plasma (it ran for a couple days inside the vacuum chamber).
If you noticed the black marks near the center of both images then you have spotted the problem. Those images show the “top” and “bottom” of my planar triple probe. There is no absolute orientation for the probe, so that is essentially a view of both burned sides.
The following figure is a diagram, from the triple probe construction download, that illustrates the basic design features of these probes. The “Head Holder” is the piece that is made from aluminum. It is used to provide centering alignment for the probe head. On a side note, the conducting tips of the probe head are electrically insulated from the probe shaft by alumina. Not to be confused with aluminum, alumina is a hard ceramic with good insulation and vacuum properties. It held its form even under the scorching laid upon it by the aluminum centering piece (head holder).
What Caused the Alumina to Burn?
The burn marks on the probe are likely caused by sputtering from the aluminum. Ions in the plasma impact the aluminum with enough force to knock some of the aluminum atoms out of the solid. In effect, it appears as though the aluminum is arcing out into the ceramic. The problem area is small, as shown in the following image. Notice how far away the burn area is from the probe tips themselves (≈ 7 cm). This is good because the further away the burn is occurring, the less likely there is an effect on the acquired probe signal. While there have been no obvious signal effects, the aluminum pieces are being replaced with stainless steel that is unlikely to demonstrate this behavior. If it does, then I will have to think of something else to remedy the problem.
Significant sputtering does not seem to be a problem for probes in which the volume between the aluminum and ceramic is minimal. The photograph below shows a probe for which there is little to no space between the protruding ceramic and aluminum centering piece. This suggests that using a different material may not solve the problem. It is possible that plasma entering through the gap between the centering piece and ceramic is related to the burn effect.
The lesson is, whether it is the root of my problem or not, that aluminum is not a good material for plasma facing components. Unless, of course, your intention is to sputter, in which case that is a great choice. If there is any interesting (i.e., burn mark) behavior from the stainless pieces I will put some pictures here.