Leak in GC/MS

[MrGC]

Hello everyone,

I'm facing a leak in my GC-MS system. The values obtained for oxygen and nitrogen in the manual tune menu are ~160 for nitrogen and ~50 for oxygen (at a flow rate of 0.6ml/min) in relation to the mass of 69. I experienced a flow rate dependency. At elevated flow rates, ie 2ml/min, these values decrease significantly and go down to 1.6 for oxygen and 3.5 for nitrogen. I used an electronic leak detector but could not find anything. I already tightened the nuts that go to the inlet and the one that goes to the transfer line. The displayed value for the vacuum is absolutely fine.
I use two columns that are connected by means of an agilent column connector. I have not started any heat cycle yet.

Does anyone have any suggestions?

By the way I use Helium as carrier gas.

Kind regards and thank you very much in advance!

[Consumer Products Guy]

I suggest isolating first the MSD, see if the contaminants go down, sounds like that would be the case.

So for me, I'd try spraying "canned air" at the inlet, all connections, etc, while the MSD was running. I'd start with the inlet and column connector.

[MrGC]

I will get one and try it, thank you for the advice! Anyway, isn't it peculiar that the abundance of oxygen and nitrogen is that flow dependent or is it always like this?

[Peter Apps]

Are the two columns connected in series i.e. end to end ?

It is usual for the contamination to decrease with flow rate when you have a leak in the carrier gas path. How are you operating the inlet, split or splitless ? and does it make any difference if you change from one to the other or change the split ratio ?.

That you cannot find a leak suggests that it is somewhere with pressure below atmospheric.

Peter

[MrGC]

Are the two columns connected in series i.e. end to end ?

It is usual for the contamination to decrease with flow rate when you have a leak in the carrier gas path. How are you operating the inlet, split or splitless ? and does it make any difference if you change from one to the other or change the split ratio ?.

That you cannot find a leak suggests that it is somewhere with pressure below atmospheric.

Peter

Hello Peter,

thank you for your reply! The two columns I'm using are in series. I've exclusively used the splitless setting. Furthermore, I was told the split injection does not work with this system so unfortunately I cannot try it.
I'm using agilent column connectors. I read that they must be heated to seal two columns tightly. Is there any truth to it? I used them before and they sealed perfectly fine even without heating the oven...

Kind regards

Edit: I switched to split mode. Nothing changes...

[Peter Apps]

Are the two columns connected in series i.e. end to end ?

It is usual for the contamination to decrease with flow rate when you have a leak in the carrier gas path. How are you operating the inlet, split or splitless ? and does it make any difference if you change from one to the other or change the split ratio ?.

That you cannot find a leak suggests that it is somewhere with pressure below atmospheric.

Peter

Hello Peter,

thank you for your reply! The two columns I'm using are in series. I've exclusively used the splitless setting. Furthermore, I was told the split injection does not work with this system so unfortunately I cannot try it.
I'm using agilent column connectors. I read that they must be heated to seal two columns tightly. Is there any truth to it? I used them before and they sealed perfectly fine even without heating the oven...

Kind regards

Edit: I switched to split mode. Nothing changes...

If the split mode doesn't work than its use for diagnostics is doubtful - usually if the air goes down in split mode it means a leak at the inlet septum.

If the length and diameter of the second column is such that connector between the two columns is below atmospheric then that is the prime candidate for the leak. Remove the column form the transfer line (as suggested earlier), plug the connection and check the air.

Also plug the end of the column with a septum and check at the connector with a leak seeker - with no vacuum at the end of the column you will get helium leaking out which the leak seeker will detect.

Peter

[MrGC]

I plugged the transferline with a nut and a ferrule without a hole. The contaminants surprisingly even rise although the displayed vacuum is lower. With a connected column I'm normally somewhere in the range 10^-6 to 10^-5 Torr. Right now, despite higher contaminants I'm in the range of 10^-7.
I will vent my system and clean the o-ring and remove eventual dust particles on it. Is there anything further I could do to spot my MS leak?

Kind regards

[dblux_]

Before opening MS chamber try to locate possible leak with argon or canned air as was previously advised.

[MrGC]

Before opening MS chamber try to locate possible leak with argon or canned air as was previously advised.

The problem is right now that I don't have either canned air nor argon at the moment. So trying this would not be possible before monday next week (probably).
So timewise I think it's better to vent my system and clean the o-ring in hope of an improvement. If this alone does not help I will definitely try the argon method.

What baffles me a bit is that increasing the flow rate helped lower the contaminants. As Peter Apps said this normally occurs with a leak in the carrier gas flow rate which makes sense.
Would a leak at the o-ring also cause a flow rate dependent contaminant abundance?

Btw. thank you very much to all the people who reply to me and try to find a solution. My GC work has previously been limited to rather trouble-free routine work. It's quite interesting to glance beneath the surface and look more closely on the actual working mechanism.

[Consumer Products Guy]

Splitless essentially "returns" to split mode conditions after a specified time, so it's curious that you say the unit only works in split mode and not in splitless. At least that's been my experience with 6890 GCs.

[MrGC]

Splitless essentially "returns" to split mode conditions after a specified time, so it's curious that you say the unit only works in split mode and not in splitless. At least that's been my experience with 6890 GCs.

The system allegedly just works in splitless mode and NOT in split mode. From what I know someone set the split to a certain ratio and made a sample dilution that should match the ratio. The obtained results were way off from what I was told. However, they might have used a liner with an outer diameter too big.
I use the model 7890A.

Is there a more advisable way of plugging the MS than with a ferrule without a hole?

[dblux_]

...
Is there a more advisable way of plugging the MS than with a ferrule without a hole?

No. This one is the most reliable and the most advisible.

[MrGC]

The leak seemed to be at the o-ring of the analyzer chamber. There were several dust particles visible. After cleaning I got a proper vacuum.

Upon removing the nut with the ferrule without a hole, would you vent the system or is it okay to just work quickly and exchange the nut with the column? I read on the internet that some people don't vent the system before to save time. Others, on the other hand, said not venting might ruin the turbo pump... any ideas?

[dblux_]

... Upon removing the nut with the ferrule without a hole, would you vent the system or is it okay to just work quickly and exchange the nut with the column? I read on the internet that some people don't vent the system before to save time. Others, on the other hand, said not venting might ruin the turbo pump... any ideas?

To be on the safe side vent MS prior to removing the nut.

[Consumer Products Guy]

Vent.

If you had a capillary column attached - which would restrict any air flow in - then I'd say different.