Frustrating air leak in new GC, help?

[Peter Apps]

Are you helium "nozzles" these things or something similar:

https://www.swagelok.com/en/catalog/Qui ... anguage=en&

?

If they are be sure that you have a stem with a shut off, otherwise the tube fills with air very time you disconnect, and then if you reconnect without flow the air diffuses back into the rest of the piping.

Peter

[RogerBardsley]

Hi MA,

How does the High Vacuum Pressure compare between the different MS's. Following on one of the other post asking about the transfer line, have you used argon or duster spray for all of the connections and potential openings around the MS, including the rough pump connection and the ion gauge on the back of the MS?

Best regards,
Roger Bardsley

[MAForensics]

They are like faucets, all with swagelock connections. Even if we had some air diffuse during disconnect/reconnect, there's no way it should persist over multiple days at a flow rate of 2 ml/min. I would think that would be gone in a few hours. And as I said, we don't get any issue hooking up a different system to the same nozzle.

I mean, a 40% N2 would lead me to believe we have a fairly large gaping hole somewhere that should be obvious, but we can't find ANYTHING.

[MAForensics]

Hi MA,

How does the High Vacuum Pressure compare between the different MS's. Following on one of the other post asking about the transfer line, have you used argon or duster spray for all of the connections and potential openings around the MS, including the rough pump connection and the ion gauge on the back of the MS?

Best regards,
Roger Bardsley

5977 (new system): 1.39e-5 Torr
5975: 9.99e9 Torr (this doesn't seem right..)
5975: 3.16e-6 Torr

I can't get the vacuum pressure on the other system we used to test the gas line, as it's a 5973 and doesn't give a reading in MS parameters.

I have sprayed all the GC fittings, including the interface nut and transfer line with DFE and gotten no response. We've also put a blanking nut on the interface and have no air/water tuning the MS by itself without the GC.

[Peter Apps]

They are like faucets, all with swagelock connections. Even if we had some air diffuse during disconnect/reconnect, there's no way it should persist over multiple days at a flow rate of 2 ml/min. I would think that would be gone in a few hours. And as I said, we don't get any issue hooking up a different system to the same nozzle.

I mean, a 40% N2 would lead me to believe we have a fairly large gaping hole somewhere that should be obvious, but we can't find ANYTHING.

If the "faucets" are on-off valves and you are attaching and detaching swageloks on their downstream sides then you are getting air into the pipe every time. Also, repeated make and break of compression fitting is itself a probably source of leaks.

At 2ml/min it will take forever to flush a gas line - you only have exponential dilution going on. However, I suspect that you have more than 2ml/min going through the lines - what is the total flow setting on the GC ?

I use a different make of MS, but on the Agilents I doubt that a 40% nitrogen reading means that 40% of the carrier going to the MS is nitrogen, so you should not be looking for a major leak - the air/nitrogen is getting in by diffusion where there is a tiny helium leak, so tiny that it needs very careful work with a leak seeker to find it.

Peter

[MAForensics]

Yes, the 40% is in relation to the 69 ion.

They are not on/off type valves, more like faucets or hood gas lines with a twist knob to open. Whenever I change lines I purge them by letting them run helium disconnected before attaching them.

Agilent was back again today, they still want to blame our gas lines...

Currently, our new instrument is set up in a T configuration with our older system. The older system has a nitrogen leak of about 3% when hooked up by itself to it's own gas spigot.

Both the new instrument, and the older instrument it's T'd off of, currently BOTH have a leak of 20% N2. So this leak pretty much has to be coming from the new instrument, no?!

Ugh, headaches.

[Peter Apps]

Yes, the 40% is in relation to the 69 ion.

They are not on/off type valves, more like faucets or hood gas lines with a twist knob to open. Whenever I change lines I purge them by letting them run helium disconnected before attaching them.

Agilent was back again today, they still want to blame our gas lines...

Currently, our new instrument is set up in a T configuration with our older system. The older system has a nitrogen leak of about 3% when hooked up by itself to it's own gas spigot.

Both the new instrument, and the older instrument it's T'd off of, currently BOTH have a leak of 20% N2. So this leak pretty much has to be coming from the new instrument, no?!

Ugh, headaches.

How are you purging the line between the "faucet" and the GC ? - opening the "faucet" to atmosphere will purge only the line upstream of the "faucet", not the line form the "faucet" to the GC. Can you post a picture of these "faucets"; anything like a fume hood gas tap is nowhere near good enough for GC-MS helium. If they are not on-off valves how are you closing them to stop the flow of helium when there is nothing connected ?

The only thing that the old and new instruments have in common if they are t-eed off the same gas line is the gas - so in this case Agilent is correct - you have a problem with the helium that is coming out of that particular faucet. This is not to say that there is not also a problem with the new instrument.

I am still waiting for you to tell me the total flow setting on the new instrument.

Peter

[MAForensics]

Here's a pic of what our helium outlets look like.

The total flow of the new system is 105.

As of this morning, we are still T'd off from the older system. The older system has an N2 of ~3%. The new system has an N2 of ~16%.

[Peter Apps]

Since Friday the old instrument has improved from 20% to 3% while the new one has improved only to 16%.

What differences are there in flows, split ratios, temperatures, connections, scrubbers, overall sensitivity etc etc between the 2 instruments ?

Can you also post pics of all the connections between the helium valve and the GC.

Peter

[MAForensics]

Ok, so I hadn't realized the Agilent tech changed the inlet type from splitless to split while he was here on Friday, and that the temps were all turned down. So, remedied that for a better comparison. All our methods are splitless injection.

So:

The older system is an Agilent 6890N/5973 GC/MS with EI source. It's gas line is piped from the helium outlet, then two scrubbers (a universal trap and a really old, unlabeled trap, maybe hydrocarbon), then to the EPC. The total flow on the older system is 20. Column flow is 2 ml/min. Inlet temp is 280, oven temp is 60, aux temp is 300. Source and quad are at 230 and 150. Nitrogen level is ~2.5% of the 69 ion right now.

New system is a 7890B/5977 with HES source. Gas is from the helium outlet next to the older system, T'd off, with no scrubbers. Total flow is 25. Column flow is 2 ml/min. Inlet temp 280, oven temp 60, aux temp 300, source and quad are 230 and 150. Nitrogen level is currently ~25% of the 69 ion.

Pictures:



So, from the helium tap to a T-connection. The left goes to the two scrubbers, the right goes to the new instrument.



They didn't have enough line to make it all the way to the EPC. So.. extension.



And then straight to the EPC.

Agilent was just here and found a very small leak in the T going to the new instrument (first pic). So he remedied that and now I'm giving it some time to purge out.

[Peter Apps]

"Agilent was just here and found a very small leak in the T going to the new instrument (first pic). So he remedied that and now I'm giving it some time to purge out."

That would do it.

Peter

[dblux_]

If I only see it properly the big universal trap is not connected according to Agilent's recommendations (horizontal position is not accepted, maybe your trap is not connected with respect to flow direction marked on the label - check it).

[MAForensics]

"Agilent was just here and found a very small leak in the T going to the new instrument (first pic). So he remedied that and now I'm giving it some time to purge out."

That would do it.

Peter

It didn't do it. New system was still holding steady around 15% after purging for 4 hours. Purge flow was set to 900 and got it down to 9%. As soon as I set the flow back to normal (15) it rose back up to ~15%. Pretty positive the gas is fine at this point, it's been tested about 5 different ways.

And yes, the universal trap is horizontal, which is not recommended but is really a moot issue since none of the traps capture nitrogen.

[MAForensics]

"Agilent was just here and found a very small leak in the T going to the new instrument (first pic). So he remedied that and now I'm giving it some time to purge out."

That would do it.

Peter

It didn't do it. New system was still holding steady around 15% after purging for 4 hours. Purge flow was set to 900 and got it down to 9%. As soon as I set the flow back to normal (15) it rose back up to ~15%. Pretty positive the gas is fine at this point, it's been tested about 5 different ways.

And yes, the universal trap is horizontal, which is not recommended but is really a moot issue since none of the traps capture nitrogen.

[MAForensics]

"Agilent was just here and found a very small leak in the T going to the new instrument (first pic). So he remedied that and now I'm giving it some time to purge out."

That would do it.

Peter

It didn't do it. New system was still holding steady around 15% after purging for 4 hours. Purge flow was set to 900 and got it down to 9%. As soon as I set the flow back to normal (15) it rose back up to ~15%. Pretty positive the gas is fine at this point, it's been tested about 5 different ways.

And yes, the universal trap is horizontal, which is not recommended but is really a moot issue since none of the traps capture nitrogen.