Today was a very frustrating day, but may have led me to a better process for loading the MOT with Sr. First off, I should mention that I realized that the ion pump magnet probably has not been degraded since it was operating fine before we turned it off and did not heat it again when it began operating poorly. Most likely the pump was contaminated with Sr-oxide and we’re seeing some vapor pressure issues because the temperature of the arc when running is going to spew more oxide until it redistributes to other parts of the chamber.
This morning I set about assembling the vapor cell with my lead window protectors. I couldn’t seem to get a lead foil thin enough to my liking in the lab, so I went upstairs to Wes’ machine shop and used a real anvil and a metal working hammer. I was able to get a 1/4” dia. 2” long piece of lead spread out enough to cover 2 2.75” CF gaskets. I cut the foil to the right size by placing a gasket on top and scoring the inner diameter with a sharp tool. I used a wire cutter to cut out a slightly larger circle than I’d marked and cleaned both the foil and gasket with spectroscopic grade propanyl (isopropyl alcohol). It actually wasn’t feasible to use a soldering iron to tack the foil onto the copper, but it staid well enough by pressing the foil onto the copper by hand. I was careful that the lead did not extend to the point where the knife edge contacts the copper.
When I vented the chamber for the first time to install a window it seems like some atmosphere may have flowed through but I don’t think it could have been significant enough to worry about. After both windows were installed and the cell insulated again, I started heating.
Because lead melts at 327 C I set the windows to 300 C and left the valve at 400 C. I must have positioned the fan for the cartridge heater better this time though because I was able to actually get the sheath to 724 C while the lead end was 265 C. It did require more power this time (88 VAC) because the windows were contributing significantly less heat to the cell. Since, we didn’t have a sensitive enough gauge on the cell to witness a pressure bump we simply left it at full heat for ~ 1 hour. It seemed like that should be a pretty good amount of time, however, I have nothing to base that on since I’ve never done this before. We figured if we were spraying Sr during any part of that time it would mostly stay inside the cell, and the back of the envelope calculation we did for Sr transfer to other parts of the cell seemed pretty slow.
I set the cell to cool down right at the end of the day so we’ll see tomorrow if we managed to evaporate all of the Sr oxide. If not we have to open the cell up and redo the foil shields if we want to get it to work.
When I came back from lunch today, it was discovered that the window on the MOT had catastrophically failed. The glass was cracked from several points on the out rim extending toward the center and the ion gauge had automatically shut down to protect itself. We immediately closed the gate valve between the chambers, cooled down the ion pump, and vented the chamber with argon. I should mention that when I came in this morning the first thing I did was set the window, oven, and nozzle to cool in anticipation of putting the new heater on the oven so those were already cool.
When we unwrapped the foil on the heater we discovered that a previous group member had put the thermocouple on the outside of the band heater as opposed to between the heater and window so the window was likely hotter and maybe MUCH hotter than we ever thought. I’m guessing that the stresses of being over heated finally overcame the window given the way it had failed. I’ve had windows fail because of heating or cooling to fast before and never did the glass crack. In those situations the glass-kovar seal was compromised. For the glass to break like that it had to have been because the difference in thermal expansion between the glass and metal finally over came the tensile strength of the glass.
It seemed like we would never get that window off of the MOT either. After some serious pounding with a hammer the zero length adapter finally came off the chamber flange. I was able to then put that in a vice and separate off the window by pushing from the inside with a hammer and screwdriver on the gasket. The adapter was not damaged and we’ll reuse it.
We put a blank on with a viton gasket and pumped down the chamber again. With the ion pump back up to temperature the pressure in the chamber was only at 2x10^-6 T. Not good. Hopefully, this will be better tomorrow and we won’t have to bake the entire chamber again. Although, maybe it’s a good idea. Since the window was already exhibiting some signs of Sr-oxide coating from when the oven temp ran away I had the idea of leaving a cold blank on in place of the slower window while evaporating Sr-oxide in the MOT. We could then vent the chamber and quickly put on the window, resulting in a much thinner oxide layer to evaporate the second time. When we do that we’ll be able see how long it takes to actually evaporate the layer.
It would probably be a good idea to bake out the whole chamber (except for the blank) when we evaporate the oxide now that I think about it. I am going to calculate how much Sr we want to load the oven with since I didn’t put in very much at all the first time and we’ll make sure we have what we want in there before we dive into all the work.
No comments:
Post a Comment