Chromatography Forum

Last week I plugged off the MS and it looked great. I added in the column, plugged off the inlet end and that also looked great; "looked great" meaning pumped down and had very low air and water. I performed an inlet leak check and no extra gas flow was leaking out. I then performed an inlet pressure decay test and that did not lose pressure, even over the course of 20 minutes. I have changed the septum, liner and o-ring, changed He tanks, and checked all of the gas line connections. I have changed the reducing nut and gold seal. Each change and check has not resulted in restoring full pressure and flow. A new split vent trap will arrive soon so I will replace that, but I doubt that actually has anything to do with this leak issue. I will also be installing an universal trap into the gas line to see if the contamination is coming from the He. The pressure to the inlet is hovering just slightly below the set point (~0.02 psi below) and the flow is also approximately the same difference in flow from actual to set point. I did try the dust off, and nothing showed up. I am not sure where else to look.

You described in detail all the material and time waste You did, but did not describe the problem at all.

Why You did start the investigation in the first place? It was because the method was working perfectly started to behave strangely? Air and water appeared suddenly in the tune report?

Or You just did not like the pressure number on the display? Then You mention changing trap because of contaminants which were not mentioned among Your problems. How did You measure the flow? GC does not control the flow - it is calculated from the pressure and column dimensions - it can not be "perfectly" correct as the column usually is cut shorter during the use. Electronic pressure control adjusts the pressure according to room atmospheric pressure; sensors also must be checked and reset after some time periods - it might be off slightly - that means "wrong number" might not mean there is a leak....

Please describe the situation properly if You want any effective help.
Are You new on this particular instrument?
Implementing new method?
Old method behaving odd? How?
Details on the column/temperatures/flows/injection/sample?
What is wrong - changing areas, times no/extra peaks, baseline problems?

Only with these details explanation of the steps You made will make any sense.

I am puzzled as to why you thnk that you have a problem - the discrepancy in pressure and calculated flow is close to the resolution of the readout and is so small that it will not make any practical difference. Is the GC showing that it is ready to run ?, or does it stay not ready with a flow/pressure error ? Are you seeing air and water in the MS ?

You need to invest in a leak seeker - the built in pressure decay tests are almost useless for actually locating leaks, and they are vulnerable to malfunctions in the EPC.

Peter

Sorry about not including the symptoms. That was a copy/paste from an email. The symptoms are very high air, to the point that the 28 peak is the main peak in the scan in Manual Tune (Abs: 1.2 million, 69 Abs at 300,000), though the peaks at 69, 219, and 502 look great in profile. There is some water in the system and a proportional peak of oxygen to the nitrogen to reflect air. Column dimensions: 60m, 0.25, 0.25. I started this whole troubleshooting process because the high air peaks popped up. When the problem initially came up the 28 peak was as high as 8 million. I have yet to change the split vent trap or include the universal trap because they are not here yet. Each step that I described in the original post have yet to fix the leak issue. I thought that the pressure was an issue because it was of concern to the tech I was talking to. I have been working with this instrument for about 9 months now with really no prior experience with GCMS, but have worked with other analytical instrumentation. I had a contamination issue about a month ago, and that was solved by changing the gold seal. I then had a small leak issue 3 weeks ago at the MS interface which was resolved by cleaning the threads of the interface with a die that had been well cleaned in methylene chloride and replacing the nut. I then ran samples over the weekend and the following week (last week) this major leak became the new issue. There is the background information. I will run a standard sample on our essential oils protocol which has an inlet of 250, oven: 70 (5 min), ramp 7/min, 320 (10 min), 1 uL injection, splitless, 1 mL/min through the column, with a purge flow of 60 mL/min after 0.1 min. I have been in contact with both Agilent techs and a tech from the company we purchased the instrument from through this whole process and they are to the point that they think someone needs to come in to look at it.

Is this a 5973?

There is almost certainly a leak in the inlet, or upstream of it. The inlet passes its pressure decay test, so upstream is the place to look initially. Do not be tempted to just tighten every connection - they leak when they are overtightened as well as when they are loose. And do not use soapy water to look for bubbles - Snoop is soapy water. You really do need a leak seeker.

Peter

Did the air peaks appear suddenly during instrument operation or after shutdown/maintenance?

I think massive air peaks indicate problem at MSD end rather than at inlet area.
I presume You did shutdown MSD several times during earlier described test with caping and reconfiguration? What I mean it could happen that during these start/shutdown cycles leak might be appearing/resealing randomly.

Also the leak at the transfer line You mentioned looks suspicious. Just cleaning the thread normally should not change sealing capabilities. Problem could be in the lost elasticity of rubber seals or microscopic crack in metal parts, or overtightening of the connections.

I think You should try to test the seals in MSD and transfer line. Earlier that was easily done by using freon sprays. Procedure is to have MSD on and monitoring ion present in the "test material" - then You spray small gas portions around the seals (one at the time) and follow at which location peak abundance increases (keep in mind that gas needs some time to travel from the seal to ion source so wait some time before proceeding to the next spot). You can try to use vapours of dichloromethane or something similar (be careful with the transfer line and other hot parts).

I was able to borrow a leak detector and found a small leak with the He line connection to the back of the GC. I recut the copper line and used a new ferrule, washer, and nut. Now no leak is picked up there. I checked all other connections that would have positive pressure behind them and found nothing. I checked all the connections at the vacuum end of the instrument using Ar and a small peak popped up around the PFTBA tube. I'll try topping that off and see if that makes a difference. Either way I have yet to find the source of this massive leak. Could it be that some settings got changed somewhere? Thank you all for your help.

What MS is attachted to the 6890?

5973.

There is no leak if the column is connected to the MS and capped off at the inlet with a septum. The problem cannot be with the MS. But the inlet passed both the inlet leak test and pressure decay test, so it cannot be the inlet either. I have also switched tanks of He, so it cannot be the He either. I have checked both ends (pressure side/GC, and vacuum side/MS) with the leak detector and argon with no telling results. There was a small leak at the He connection to the back of the GC which is now fixed. Argon did get in somewhere and now there is just a constant small peak, but I don't know where it came from. There was no jump in the Argon peak when I checked areas and waited a minute before moving to the next spot.

Is your inlet in split mode when you do these tests? If not the "leak" is the split vent.

If you are still convinced it is comming from your inlet, do this:
Set up your ms to look for 28 and 44 click on repeat profile. Watch for a while then using the GC control reduce your column flow by half. If its coming from thinlet the 28 and 44 should reduce by about half.

50% Isopropanol water solution is great for checking the gas lines and won't be a big contaminant like soap. For the connections in the oven/inlet I spray compressed air at it while looking at a manual tune scan with pftba closed for a 52 ion (from the propellant). It should appear instantly and quickly subside if there is a leak where you sprayed.

It could be at the transfer line weldment o-ring or the side pannel.

m.finnBMBM, at this point I think you can be confident you have corrected that leak in the He line. There's always a small quantity of Ar in the system along with oxygen and CO2 because they are part of our atmosphere. If you spray with Ar and there is a leak the Ar peak will jump.