Ah, no leaks

[Doug Coulter]

See pic. Water you will always have with you, of course. This is what my system looks like about a week after the last fusor run. Props to anyone who figures out what the line at 43 is...

Mass spectrum of main system.

Looks like no leaks though. 1.6e-8 mbar net pressure now. There was some scramble there - I added a 1/4 wave antenna for 2.45 ghz, a new grid for the secondary ion source, and an an antenna for broadband LF RF.
Thought I'd try putting a little twirling motion on my ions and stuff just to see what happens.

[chrismb]

43 ... that's a mystery.

Have you been using any cleaning/processing agents/experiments with calcium or sodium in them?

[Doug Coulter]

No. Well, what just happened was I replaced all the glass and quartz in my main feedthrough, added an antenna (welded in type N connector) for 2.45 ghz, and changed the 2nd grid/ion source to one made of tantalum. So, there's sodium in the glass...I think (some of it is pyrex), but as far as I know, that's it. When I first put it back together (and fought some new leaking) there were some lines up in the 80s-90s I think due to using diff pump oil to lube the O ring that seals the glass on the FT, but that's all gone now.

The glass is well protected now inside copper pipe in the tank - the copper extends past the end of the glass (and is grounded), so as to protect the glass ends from being reduced by hot D ions hitting it when it takes on a little charge - which seems to be working as a scheme. I slid 1" ID copper tubing over the glass, then near the grid a 1" to 1.25" adapter that flares out and comes within about a cm of the grid end, with the glass about 2 cm back from the end of that - it's pretty protected, in other words. Whatever this is I think it's a gas...hmmm.

I tossed this up on my science group on G+ too, and no one there has a clue either. This is pretty new - it didn't used to show up at all. Water, yeah, CO2 etc, yeah, a few light things, always. But not 43.
Note, no argon at all. That's what really tells me "no leaks". If the remaining O2 and N2 weren't matched by a proper percent of Ar, I'd be thinking "leak", but being non noble, they just stick in there longer even if there's no leak at all.

So, it's still a mystery. Not that it seems to be hurting anything at the moment, but I hate not to have a real solid baseline before I do the next set of experiments of putting in some RF to spin things, find plasma resonances, drive recirculation, whatever - just looking to see what effect it might have and maybe learn something.

[johnf]

Ok
Its a fraction of a bigger molecule pentane probably

following stolen from the interjerk

Consider pentane, C5H12.

Mass Spectrum of Pentane

m Fragment loss of
72 CH3CH2CH2CH2CH3
57 .CH2CH2CH2CH3 CH3 (m-15)
43 .CH2CH2CH3 CH2 (m-14)
29 .CH2CH3 CH2 (m-14)
The highest m/z peak at 72 corresponds to the molecular mass of the pentane, CH3CH2CH2CH2CH3, molecule.

MM (C5H12) = (5 x 12) + (12 x 1) = 60 + 12 = 72
The peak at 57 represents the loss of a CH3 (m=15) fragment from pentane, ie 72 - 15 = 57.
57 is the mass of the CH2CH2CH2CH3 remaining fragment.

The peak at 43 represents the loss of a CH2 fragment (m=14), ie, 57 - 14 = 43.
43 is the mass of the CH2CH2CH3 remaining fragment.

The peak at 29 also represents the loss of a CH2 fragment (m=14), ie, 43 - 14 = 29.
29 is the mass of the CH2CH3 remaining fragment.

[Doug Coulter]

That's the most plausible one I've heard yet, John. There is carbon in the tank as part of my grid, but of course what mostly hits it hard is D, not H. We might suppose that the "pentane" came from the tiny bit of diffoil-20 I used to lube the O-ring for the new glass on the feedthrough, though. When I get a chance today, I'll run it up to mass 200 and see if there's any lines up at the big masses - maybe even turn on the internal quartz heaters and see if that makes something heavy show up - normally that's what it takes to see anything over 50 or so.

[Doug Coulter]

Ok, here's a full sweep after about 10 minutes running the mass spec (and nothing else). My noise floor is usually around e-14, but today it appears to be a little higher - maybe my shop air is a little damp today.
I've turned on the heat and will add another screen capture when that gets stable again.

Full 0-200 amu sweep, no heat, cold lab too.

Here's after just a couple of minutes of heat - tank isn't really hot. The heaters are two quartz 500w lamp/heaters, but in series so they are only yellow, not UV, hot. Pretty quick change though.

After only a couple minutes of gentle heat

And lookee - all that high mass stuff just came out of nowhere. Looks like you were right, John.

[Doug Coulter]

Since this is going along pretty swimmingly for a change, here's two more sets of data. For the first, I used the quartz heaters in parallel - 1kw into the tank, lots of UV, and this makes the quartz bulbs themselves red-orange hot (so I don't do it that often, it makes some things in the tank VERY hot - as in it'll melt any solder in there).

High heat, only a few minutes. Tank warm to the touch, through the lead. Hot in there.

For what it's worth, this boosted my basic pressure by about a factor of 10, up to a few times e-7 mbar.

Then I let it cool down for awhile, not all the way back to room temp since that takes hours, but enough, and here's what that looks like. Looks like my noise floor dropped too, but I wouldn't put much faith in anything much below e-14 here.

Mostly cooled off. Basic tank pressure back down to 3e-8 mbar or so - a little higher than at the start of the test series.

So yes, 43 (and probably a couple others) are most likely breakdown products of heavy hydrocarbons, in other words, that tenth of a drop of diffoil-20 on an o-ring. Always better to measure than to guess!