Leak in a Thermo Trace1310 GCMS after shutting down.

[hypervalent_iodine]

I'm a lot of problems with a semi-new Trace1310 GC from Thermo today and I am hoping that someone can she'd some light on the issue. I am relatively new at the more advanced aspects of maintaining and using GCMS equipment and so this may be a silly question with an obvious solution. Yesterday while using the machine we were having trouble deleting a sequence from the acquisition queue on Xcaliber and so one of my colleagues turned the machine off at the power switch at the back and reset the computer. It cleared the problem in Xcaliber, but since then there seems to be a very large air leak. The auto-tune file from yesterday:



And now this morning it looks like this:




After changing the septum it got worse:



The carrier gas tank seems to be okay and our other Shimadzu GCMS, which is connected to the same tank, is fine. I haven't played with the columns or any of the connections at all except earlier today to see if they were loose (they weren't). I am wondering if shutting the machine off in the way that we did has caused problems and perhaps more importantly, what I can do to fix it?

Edit:

I performed a quick check scan. There is no difference with the Cal gas on or off.



I have no idea why and how there is so much oxygen and argon in there.

[Klaus I.]

I haven't played with the columns or any of the connections at all except earlier today to see if they were loose (they weren't).

Was methanol involved?

[hypervalent_iodine]

Was it involved how? The leak happened before I tightened everything, so I'm not sure what you're getting at. After it started showing the leak and before I realised what was going on there was one sample that had a methanol wash, but other than that the only thing that went in it was hexane. I finally managed to get hold of one of our reps yesterday, who was about as stumped as I was about the air/water check. Is there any reason that I should be getting so much oxygen? We have no traps in the system at all. Come Monday I am going to do a bit more investigating and look at the mass spec without any column. Hopefully it will have fixed itself by then.

[BMU_VMW]

Klaus is asking about methanol because there is a lot of m/z 32 in the spectrum.
With a large air leak you will see 32 (oxygen) and 28 (nitrogen) but you would expect to see 18 (water) as well and this one is missing.
Air leaking in from the lab will always contain water. Since there is no water the 32 might not be oxygen but MeOH.

[X]

Your foreline pressure increased from 37 to 53-54 mTorr, which indicated a big leak in mass spec.

[James_Ball]

Your foreline pressure increased from 37 to 53-54 mTorr, which indicated a big leak in mass spec.

Also note that the analyzer pressure (ion gauge) went up a factor of 10x (10^-6 to 10^-5)

I am also curious why the tune says you have 70% and more 18/69 when running a tune check while when you have the cal gas closed you have not 18 at all? Was the cal gas vial cleaned at any time? If so with what?

If the cal gas vial was refilled, did you allow it to purge to vacuum for a while before running the tune check?

[hypervalent_iodine]

The leak was not with the MS. On Monday I fitted a blanking nut to where the column enters the transfer line to check if that was the issue and it was fine as per the air/water check, indicating an issue with the column. I disassembled the column and discovered that the column had somehow snapped at the other end of the nut inside the SSL inlet. I have no idea how since no one touched it until after the problems began, but reassembling the column and baking it overnight has since fixed it. Could it have snapped off during a poorly executed manual injection with a syringe or SPME fibre? Someone had performed these potentially around the same time the problems began, though I would have thought the liner would have broken first.

BMU: thanks for the clarification, I should have realised that was what he was talking about. Particularly since I myself had wondered about whether there was MeOH when I first the air/water check. I could think of no possible sources for it and dismissed it pretty quickly. If the 32 peak were MeOH, why would I still be getting no water and a lot of nitrogen?

Cal gas was probably last done when we had some engineers up from Thermo a few months ago, at which point I wasn't in the lab full time and so I couldn't say how they did it. Unlike our Shimadzu engineers, they don't always break more than they fix when they stop by so I trust they did it correctly.

Even having now fixed the problem, I am still a little unsure of how what seems to have happened led to the weird results I was getting. The magical loss of water is confusing. The Thermo rep I spoke to said the peak at 27 could be the result of a lot of nitrogen reaching the detector, but this didn't make very much sense to me. If true, I suppose it could explain the strange 28:32 ratio when taking into account the accumulated intensity of the 27 and 28 peak, though it doesn't explain where the water went.

ETA: I retract my first sentence. I changed the nut and ferrule when reassembling the column at the transfer line end, so the leak may have been there and the column could have snapped upon removing it from the inlet.