Agilent 6890/5975B shorter than expected column lifetime

[bisnettrj2]

All:

I'm running an Agilent 6890N with a 5975B MSD using a Varian VF-5ms 35mx0.25mmx0.25um column. This chromatogram was obtained using the following conditions:

Injector set at 250C
0.2uL injection of air (capped vial with nothing in it)
35 psi pressure pulse
1 min pulse time
0.9 min purge time
80mL/min purge flow
50mL/min gas saver on at 5 min
1.3 mL/min helium column flow, constant flow mode
Transfer line at 290C
MSD Quad at 150C, Source at 230C, vacuum reads at approximately 9.8e-06 during the run (after the pressure pulse ends, that is)

Oven - 50 C for 4 min, 30C/min to 150C, 5C/min to 290C, 120C/min to 310C, hold 2 min

Detector - 5975B MSD scanning 10-900 mz - spectrum zoomed in to show main spectral components (m/z 10-70)



The humps in the baseline have a spectra consisting primarily of m/z 28. In three consecutive runs with this setup, including a 2uL injection of DCM and a 2uL injection of a standard, the humps occurred, but not at reproducible retention times (but within the same relative time frame, like 12-20 min and 25-32 min, with two humps in each chromatogram and about 12-16 min between the apex of each hump in the chromatogram). We don't normally scan below 45, so I don't know if this has always been occurring, or if it's recent. We have been experiencing shorter-than-expected column lifetime with more extreme maintenance necessary when the chromatography degrades (e.g. needing to trim 2 meters or more off a column to correct tailing or the loss of active compounds like 4-nitrophenol or 2,4-dinitrophenol).

My question - has anyone seen baseline humps like this, are they detrimental to the column or system, and if they are something to worry about, how do I get rid of them?

[Hornet]

It seems you have an air leak burning your column, it's potential damage is probably amplified by the pressure pulse injection. How is the air/water check?
Stationary phase molecules are at a higher m/z range (207, 281, etc) that's nitrogen.

[bisnettrj2]

Air/water always checks out fine in tuning, less than 1% for oxygen and water. Nitrogen is normally around 3%.

[Hornet]

You could try air/water check while pressure pulse is active.

[cleh]

Check your air/H2O manually by doing a profile scan in tune with 18 in one window and 28 in the second window. You should be able to see mass 14 in the 18 window. If that is high, you have a major leak and it will damage your column and filament.

[bisnettrj2]

Called Agilent today - they said they've seen that sort of behavior when oxygen and moisture filters are saturated and they start to "burp". We hooked up a stand-alone tank instead of using our house gas manifold and got a very nice, flat baseline. We're going to be changing the manifold gas filters in the morning.

[Hornet]

Called Agilent today - they said they've seen that sort of behavior when oxygen and moisture filters are saturated and they start to "burp". We hooked up a stand-alone tank instead of using our house gas manifold and got a very nice, flat baseline. We're going to be changing the manifold gas filters in the morning.

Excuse me, how is it possible that your air/water check was fine if your filters were saturated? There was clearly a lot of air in your system from that chromatogram so your air/water check is either useless or working wrong, think about it. This could lead to more serious problem in the future if you'll have a bigger leak and you won't be able to detect it.

Anyway i'm happy that you solved the problem.

[bisnettrj2]

I'm not really sure. Multiple air water checks never revealed a lot of oxygen or water. Nitrogen, sure, but oxygen was always low. I'm not 100% convinced that the filter change is going to be the solution, I will be re-running the previous test once the new filters are installed, and we'll be installing an indicating filter in-line after our normal filters to have a visual cue as to the state of the filters. We have other GCMS systems (Varian ion traps, but I think full scan should show somewhat comparable results) on the helium manifold line too, and I think I'm going to try to do the same test on all of them to confirm the lack of these broad baseline peaks once all the filters have been changed.

[cleh]

I have not used the air/water check since the 5973 was released because it doesn't show you the whole picture. If there is a major air leak, the nitrogen (28) appears normal because it is mostly mass 14.

[gjwkodak]

I see wide humps of air like that when I change a Helium cylinder (scanning 10 - 800). But only once, and on all instruments at the same time. So I don't run anything within 15 minutes of a tank change.

But you are seeing multiple humps? And the times vary, so.. it is not the GCMS, it is likely from outside the GCMS, something else is introducing a slug of air. What else shares that Helium line from the tank? And yes, putting on a separate tank would solve the issue, the cause of the problem is now off-line.

Is there a device that uses Helium, then does not use any, in a ~15 minute cycle?? A small air leak in it's plumbing would allow air to diffuse back into the main line, then when it starts using Helium it pull the leaked air/He mix into itself for awhile... and repeats.... as long as it is running...

Your tune/air check looks fine.. as long as you are between "humps" when you do the check....

I question blaming it on your filters - they do not trap Nitrogen, so could not release it .....?

[bisnettrj2]

I'm not sure what might be doing it. I'm on a manifold with 8 semivolatile GCs and I think three or four volatiles GCs, so it could be a combination of all of them. I've also installed an oxygen/moisture trap at my instrument, so if there's some leak somewhere downstream of our filters, hopefully I'll catch it there. The scan was re-run Tuesday, and no baseline disturbances were seen. I hope to repeat the test on a semi-regular basis to see if it is intermittent, or if it becomes more prevalent as performance degrades.

[gjwkodak]

Are the Volatiles instruments upstream (on the Helium supply line) from your problem instrument? Do other semivol instruments downstream also have this issue when you run the same test?

When you reran the test Tuesday, were the Volatiles intruments running? In particular, the autosamplers? What about the first time you saw this issue?

Typically VOA autosamplers use Helium for carrier, but also purge and other flush lines internally, which are not on continiously. Try the test when all VOA instruments are running, and when all are not running. If the autosamplers are the cause, run each one individually, then check that one carefully for air leaks in the Helium lines.

The indicating Oxygen trap will stop the Oxygen - but not the Nitrogen - from getting into your column, and the MS. So it will stop the column degradation, but you will still see the hump if you scan from 10-900.

[bisnettrj2]

At this point, I'm happy I'm not seeing the hump and the associated elevated oxygen. If I start seeing it again, I'll try to keep in mind the volatiles instruments when diagnosing the problem, along with the filter saturation. However, they are on a separate line from the manifold and have their own filtering setup, so I'm relatively certain they aren't contributing, but again I'll try to keep it in mind if it becomes an issue again.

[Bigbear]

What do you get if you repeat your first experiment after taking out the syringe ( no injection run)?