Anybody ever experienced contaminated Helium for GCMS?

[jdlh199]

I've been battling with a potential leak on our GCMS system for a little over a week now. I was seeing elevated mass 28 and 32 rates at ~20% and 5% of the He signal (m/z 4) respectively during leak checking.

Nothing I did seemed to get rid of it, including roaming the lab with a He detector, spraying snoop everywhere, taking all the gas lines apart, checking and rechecking all the vacuum seals, changing regulators, and even eventually dragging the He cyclinder down to one machine and connecting with a single line (we have a large gas line system set up which supplies 2 GCMS's and 2 GC-FID's).

Eventually I tried switching to a different He tank (different Lot#) and the problem vanished, mass 28 and 32 went down to ~2% and 0.5% of He signal during leak check within a few hours. I reconnected the original tank and it came back again.

So my question is this, has anybody else ever encountered a contaminated He tank? I've never ever had a bad tank before and was just curious as to how common this is. Both tanks were from the same supplier but different lot numbers. Next time this will be the first thing I check!

[Peter Apps]

I have seen exactly this a couple of years ago (in South Africa but the helium comes from the US as far as I know) and it happened in more than one lab on different sites over a period of 3 - 4 months. Both labs used the same supplier, they replaced the faulty cylinders.

Peter

[JI2002]

I had the same problem a few months ago. Can't imagine what would happen if the helium tank was used for GC/ECD instead of GC/MS.

[HbJ]

The second He cylinder I used in my life was contaminated with hydrocarbons so I got a 'somewhat' erratic baseline (GC/FID).

I took almost my complete setup apart but was unable to find an error. Finally I traded in the cylinders and got a replacement that worked.

A day spent in vain

[Consumer Products Guy]

Not in vain. You learned something.

[Bruce Hamilton]

Not sure about the USA, but here in NZ a decade or two ago, many of the instrument grade gas cylinder ( tanks ) from suppliers had the valve connection changed to incorporate a small internal residual pressure seal, meaning that even a fully open valve would stop delivering gas once the cylinder pressure had dropped to a few 10's of PSI. The user's system gauges still drop to zero, so they don't notice anything different.

The suppliers compensated for the lost gas by slightly increasing the fill presssure. This was to prevent one freqent cause of this problem, some users were leaving valves open after cylinder was removed from system.

If your helium supplier is decanting from larger cylinders, there's always a chance of contamination, but their quality control should catch it. Discuss your results with the supplier, if they go into denial, consider another supplier.

As helium gets increasingly expensive, it's quite likely more major use sites, ( liquid helium-cooled magnets on NMRs etc. ) will start to recycle or return helium to suppliers, providing even more potential for contamination.

If Helium quality is critical, purchase the grade with individual cylinder CoAs. Given the high cost of Helium, the incremental charge is fairly small.

Please keep having fun,

Bruce Hamilton

[AICMM]

Two weeks ago had a helium cylinder that was contaminated with argon. Fortunately customer was not scanning that low and I don't think the level was high enough to affect chromatography so I did not do anything about it. They use the same CGA fitting though.

Best regards.

[jdlh199]

Just thought I'd come back with an update.... I received this email today from our gas supplier. The suspect tanks were confirmed to be contaminated with a nitrogren content of 50ppm!

Root cause: improper evacuation of cylinders and analytical problems.

Corrective actions: the improper evacuation of cylinders occurred during a shift change. Steps have been taken to improve communication between shifts to ensure this does not happen again. Analytical problem is being addressed and that work will be performed at another facility until analytical problem is fixed.

I cut the grovelling apologies...

[Peggsy]

its a bit off track, but i once had problems with a well used and published method for reverse phase HPLC (phosphate and methanol isocratic), in that all peaks were eluting way too early. The wierd thing was that i had just received a new C18 column, and thus threw it back to the suppliers with some choice words. Suffice to say they came to the party with a free replacement, and off i went again, but same problem, Rt's were still very much shortened. To cut to the chase, after wondering what was going on, i ran out of my exisitng HPLC grade methanol and started with a new bottle, same brand and grade. Low and behold, the chromatography came back to what was expected.

Conclusion, the actual bottled HPLC grade methanol turned out to be HPLC grade acetonitrile. Had to send my apologies to the column supplier, whom by the way tested the returned column and found no probelms with retention at all. NEVER assume.

I now sniff test every bottle (just kidding)

[HbJ]

For solvents a GC run (on a DB-624 and a CW column) is my favorite.

Alternatively the RI tells me if the manufacturer made a mistake

But sniffing is also an option