GCMS leak? High N2 and low O2

[wan]

We have a new Agilent GCMS. After my colleague changed the liner and septum, she did tuning and obtained high N2 % but low O2 %. N2 % hovers at about 46-49%.

She tighten the inlet, did manual tuning, but obtained the same results.

She also replace the liner and septum again, tighten, did manual tuning and the results are still the same.

I did a search in this forum and found that it could be due to a bad Helium cylinder. However, we have another older GCMS connected to the same gas cylinders this new GCMS is connected to. We checked that older GCMS and everything is okay.

Did anyone else experienced the same thing before?

[bhuvfe]

Is there a chance that your colleague changed also the gas bottle?

If you don't see O2 it might mean you have air in the lines before the oxygen trap of your MS. If you disconnect the line and let it vent helium for 45s-1 min (but sealing the trap and preventing O2 to enter) you should get rid of the problem.

[Peter Apps]

I have seen this under two sets of circumstances: when helium with air in it passes through an oxygen scrubber (as bhuvfe mentioned) and with no other apparent cause than stationary phase bleed.

Peter

[carras]

I had the same symptom when my helium bottle was fully emptied during a weekend and the line filled with air. I have two GC/MS connected to that line but only the upstream one was filled with nitrogen (as bhuvfe and Peter said, the filter scrubbed off the oxygen). I purged thoroughly the line as close downstream the GC/MS as possible and had it running at a split flow of 100 ml/min at least. I checked for nitrogen in the system from time to time, when it was back to normal I reverted to a normal split flow of 10 ml/min.

[fsistere]

Hi
Check the conection of the column and the interface. If you use vespel ferules sometimes you have to retight the nut a little

[JTM]

make sure the column is sealed, the vent port is completely sealed, and the MS chamber makes a perfect seal.

the easiest way to do this is to have the MS scanning (in the MS Params window) and then spray "dust off" on everything. you'll see a huge spike when you hit the leak.

keep in mind, if you spray it near the injection port and the leak is there, it may take a few seconds (but not very long).

[wan]

Thank you for all the replies.

I checked the helium cylinders and they're almost empty. So I switch the valve to use the new cylinders and ran a split flow of 400 ml/min.

It helped to reduce the N2 % a little, but my colleagues had to changed the columns after that. We went ahead with that and while leaving the MS to pump down, I turn up the split flow again.

I don't see high N2 % and low O2%, but there are obviously gas leaks. So we checked and tighten where we could.

The last time I checked, both N2 and O2 % are at approximately 9% and 3% respectively (it was at 70% and 20% before we checked for leaks). So I left both GCMS to pump down and can only check them on Monday again.

Thanks for all the help rendered.

[Bigbear]

Don't forget these systems are a " coontrolled leak" so the air values will never be 0. Agilent says if your <20% your good to go.
You will get less air in your autotunes if you tune have your injector in split mode.

Looks like your'e good to go. It will look better Monday.

[Justinoz10]

I use these systems daily and although the agilent limits are <20 and <10% we usually run at <2 and <1 (anything else and there is a problem).

[che313]

u can troubleshoot this easly
isolate ms from gc
then pump down ur ms
and check for air/water leak
if it is ok then ur leak from gc
and if it is not then ur leak from chamber seal

[aldehyde]

Optimally things should be <1% but over time as instruments age small leaks tend to form and it is possible from instrument to instrument to see variation as high as 8% (for older instruments). However, when I see anything higher than 5-10% and the instrument has equilibrated at least 24 hours then it is worth fixing immediately. As long as water, O2, N2, CO2 are low and vacuum is OK you can run for a bit.

You may want to try using canned compressed air (looking for difluoroethane, you'll have to look up the mass yourself depending on the CAS# listed on the can) or throw a chunk of dry ice into a plastic squirt bottle and squirt the CO2 gas around common leak spots (ion 44). Go into manual tune and run a profile or scan with the mass of interest set so that it compares it against peak 69.

When you spray the leak check gas over the leak you'll see a big jump. This is not helpful for inlet or gas supply leak checking because the gas would have to go through the column before being detected and there would be a time delay. For inlet you should remove column from inlet, put in a blocking nut. Also put a swagelok nut on the septum purge port on the top of the GC, then set inlet pressure to 25 and wait for it to stabilize. Turn it off (dont set it to 0, turn it off) and watch for a pressure drop. If the inlet is at constant temperature then you should observe only a tiny drop, a solid inlet won't leak at all. If it drops a bit initially (~0.5 psi) and then stabilizes that is OK.

To check for leaks on your gas supply manifold use a 50:50 mixture of water and methanol to drip on the connections, then immediately wipe off when a chem wipe.

If I saw a big nitrogen leak but not much oxygen I would suspect something with the manifold, with the oxygen getting stuck in the trap. If you have a ~3:1 N2:O2 leak that is atmospheric and is something after the trap. Start at your inlet, work your way to the MS.

[Don_Hilton]

What grade of helium are you using? If you use the less expensive stuff - you will find backtround in it - like N2

[thohry]

What grade of helium are you using? If you use the less expensive stuff - you will find backtround in it - like N2

I agree with this. I myself met this problem already. The only way is to replace with a better grade He cylinder.
One other prosibility is that there is leak before the O2 trap.

[wan]

Hi,

Don_Hilton and thohry: We used 99.999% He. I believe that's the highest grade offered by the gas company.

che313: We did just that and found a leakage on the GC inlet on the following Monday, which was promptly fixed. Thanks!

aldehyde: Thanks for the tips! I'll know better what to do in future.

So far we're getting about 5% N2 and 1% or less O2 and analyses looks okay.

One thing strange about one of the GCMS is the inlet loosen after about 100 injections, leading to loss of pressure. We don't have that problem with the other GCMS. The Agilent enginner can't do anything about it, so it is very frustrating, when we have sequences as long as 300.

Any idea what I can do?

[Don_Hilton]

You did not say what part of the inlet was leaking - so I will make a guess: the ferrule holding the column in the inlet? And the reason I make the guess is that it is not uncommon for this ferrule to become loose over a number of GC runs. My experience is that it will become loose fairly quickly after you use a new ferrule, but after you tighten it down one or two times, the ferrule will hold quite well over long sequences of runs.

And if you can, avoid graphite ferrules. Graphite/Vespel ferrules deform more slowly, giving a longer time beteween having to tighten the column nut. Graphite ferrules look the same as graphite/Vespel ferrules -and have been know to be installed in their place. (And, said chromatogapher finally learned to store the two types of ferrules separately.)