Agilent GC/MS high air (but not water) check

[chris taylor]

Good afternoon,

First off, my GC/MS conditions:
He at flow rate of 1.8 ml/min
Cool on-column injection, inlet temp mirrors oven temp
DB5-MS, 30 m X 0.25mm
MS transfer line 280dC
EI mode with quad set to 150dC and source set to 230dC

I am having a problem that I come across more often than I want to, which causes dowtime on my instrument. When I have to change my carrier gas, after running the GC/MS system overnight I am still getting Air (actually N2, m/z 28) values at around 10% or even higher. Strangely, my water (m/z 18) value is well below 1% and the ratio of N2 to O2 is much higher than in atmospheric air. The new He tank had been unused prior to installation. I have also read and heard differing specifications. In the help guide, the spec for N2 and water are <5% and <10%, respectively. BUT in the tune evaluation, the N2 spec is <10%. Folks at Agilent tech services have also told me that N2 < 10% is acceptable. SO I am unsure which applies.

Interestingly, when I remove the septum inlet the N2 level falls to below 5%. I've gone thru the system from the He tank, copper tubing, molecular sieve, inlet, column, MS transfer inlet and can find no apparent leaks. I've sprayed DUST-OFF in the GC oven and around the MS door and did not see the corresponding ion for Difluoroethane, so a leak in the MS is unlikely. I've determined that the issue has to be on the GC side, but if there were a leak and the system were pulling atmospheric air, wouldn't the N2/O2 ratio reflect this?

Or does the system just need that much time to pump the air thru? It's been nearly 30 hours since I've changed the tank. Unless of course, the tank is bad...which would be a gamble every time I have to change carrier gas tanks.

Has anyone seen anything like this, and if so what to do about it?

[Yama001]

We have had a spate of helium tanks with too much nitrogen and argon in them lately. This has never happpened before, but our current supplier (Praxis maybe?) is clearly having issues.

[rb6banjo]

I too have had a tank or 2 that have higher N2 in them than they should. I changed the tanks, and the N2 went away.

[soccerdad007]

I agree that the tank is a likely cause.

I was told once that if there was a leak between the tank and the oxygen trap then air could be getting in to the system but the O2 removed by the trap so the n2/o2 ratio would be wrong even though it is an air leak. I suspect that most of us use the combination traps now but I think the same thing would apply

Kevin

[chris taylor]

The tech at Agilent had mentioned a helium shortage as well as a recent spate of He tanks with higher levels of N2 in them. Airgas is our current supplier, so perhaps they are having the same issue as well.

BUT if there were a leak in the copper tubing between the tank and the molecular sieve, how would air be pulled in if the tubing is under pressure (~80 psi)?

I changed the regulator as well as the tubing leading to the trap and will have it pump down overnight. Hopefully in the AM the levels will have dropped to an acceptable level.

What kinds of specs do you use for your H2O and N2 levels?

Thanks for the responses.

[Don_Hilton]

How would air get into a line? Diffusion works in both directions.

One other thing to be aware of on checking system air/water: If your head pressure in the GC is low enough, air will diffuse in through the split vent. Be sure the instrument is in splitless mode before running the leak check! I have one instrument that does not show much 18, but does show 28 and 32 with the split vent open - and I have learned to close the split vent before trying to find a leak at the regulator. In fact I close the split vent before checking to see that the system is leak free.

A corred septum will allow air into the instrument as well.

[gcguy]

We had a similar problem. High nitrogen but low water and oxygen. Turns out it was the base plate of the gas purifier. There seems to have been a bit of messing about with these with Agilent's purchase of Varian. Agilent now supply the Varian (Chrompack) style filters, rebranded with Agilent, not their own old ones. These don't seem to fit very well on the old Agilent base plates. We moved to using their "big" trap which is directly conneted in-line using swagelok fittings. The problem went immediately.

I seem to remember a note from Varian many years ago warning about a change in their filter design and poor fitting of their cartridges, although I may be wrong.

GCguy

[Peter Apps]

The simplest explanation for low water when there is an air leak is very low ambient humidity.

If you are relying on pushing the air that gets into the regulator all the way through your gas lines and out through the GC-MS you will wear out your gas scrubbers much more quickly, and ultimately contaminate your system with all the other interesting substances that float around in the air. At a minimum I would add a purge valve just downstream of the regulator so that you can blow clean gas through the regulator after a cylinder change without it getting into the gas lines. You can also get fancy back-purge and switch over rigs from the GC suppliers and independents, but they are pricey.

CH

[Bigbear]

I've been told by Agilent to have the inlet in split mode ( at least 20:1) while doing an air/water check. That way there is so much He going out the split vent, little or no air can be pulled in.

[akar]

to make shure if it is a leak in MS part or in GC part you need to vent system, get out capilary from transfer line, then block transfer line with "no hole' ferrule. Turn on MS, after night make tests. If you find air components - probably leak at MS. Very often reason is a side plate seal: it must be cleaned with no dust material.

[Peter Apps]

to make shure if it is a leak in MS part or in GC part you need to vent system, get out capilary from transfer line, then block transfer line with "no hole' ferrule. Turn on MS, after night make tests. If you find air components - probably leak at MS. Very often reason is a side plate seal: it must be cleaned with no dust material.

A leak anywhere would produce a high air signal all the time, not just when cylinders are changed.

Peter

[Don_Hilton]

Looking at Bigbear's post - yes. I turned it around in splitless mode you still have air workign its way into the sytem.

[Peter Apps]

With back-pressure regulated inlets the total flow into and out of the inlet is the same in splitless and in split. In splitless with a sensible split ratio setting (set as total flow) nearly all the carrier is redirected through a flow path that does not take it through the inlet liner. The only way the gas and vaporized sample can get out of the liner is onto the column. In split mode all the gas goes through the liner and most of it with most of the sample goes out through the split vent. I am ignoring the septum purge here. If there is a leak at the inlet - most likely at the septum, gold seal (on an Agilent) or column connection - the air that diffuses in is the same in both split and splitless, but in split the majority of it gets flushed out down the split and so the air and water check on the MS is better. This is a convenient way of narrowing down the site of a leak, with a leak anywhere else changing from split to splitless makes no difference to the air levels in the MS.

Peter

[James_Ball]

We often run with N2 at 5% or more, as long as the O2 and Water are low. The N2 won't damage the column or oxidize the source and is so low it really doesn't affect chromatography.

Of course we are also doing purge and trap so we always have moisture being passed through the column, and yet our columns last over a year, sometimes two while running almost 24 hours per day each week.

When changing tanks I always crack open the tank valve before completely tightening the regulator to the tank which purges the gap between the tank and the regulator so less air gets passed through the system. Also be sure to change tanks once they get down to 500psi left in the tank or you will begin to see more N2/O2/Ar/Water as these will partition to the bottom of the tank when full, then begin to diffuse out when the pressure is low.

[Bigbear]

An quick way to check if the air is comming through the column is to check it after cutting the column flow in half.