Chromatography issues: Thermo GC-MS with helium saver inlet

[shauswirth]

I’m having an issue with poor peak shapes and declining sensitive with a Thermo Trace 1310/ISQ system with a helium saver inlet.
I started a project about a year ago looking at crude oils/tars (up to C38 alkanes+PAHs)—I haven’t run a ton of those samples (< 100 total) and all samples have gone through cleanup (mostly silica gel fractionation). Since starting those analyses, I’ve had the relatively minor issue of loss of sensitivity of especially higher MW alkanes and C3/C4-alkylated PAHs after a number of injections. This issue was always resolved by changing out the inlet liner and/or trimming the column.

In the last few weeks, things have gone steeply downhill. Peaks have gotten very ugly—very wide, “lumpy”, significant tailing, some compounds show peak splitting, some peaks have inconsistent retention times—and the response is getting very low for many peaks (e.g. high MW alkanes) to the point that some aren’t even detectable in any but the highest calibration standards. These symptoms are showing up even for calibration standards.

Last week, I replaced the liner and trimmed a few cm off the column and got one good injection of a standard, before it reverted to being problematic for all subsequent runs. Since then, I have replaced the inlet liner again, trimmed 0.5m off the column, removed and cleaned the entire inlet insert (as per Thermo’s directions), replaced the o-rings in the inlet, leak-checked the inlet, baked-out the column, and cleaned the MS source (tune looks good and the EM is newish). This did not resolve the problem. Nothing else has really changed with the system preceding these problems, with the exception of temporarily having a different model of autosampler syringe (which regularly clogged) installed – I switched back to the original model before the maintenance noted above and have confirmed it’s drawing up sample.

What else could be the issue? Could the column be totally dead? It’s about 5 years old and has seen some dirty samples (soil extracts, oil), but not a lot of throughput. It seems awfully sudden for this to be the problem and RT shifts have been pretty minimal. Or could it be something wrong with the inlet itself? The Thermo helium saver inlet has a piece of “anti-diffusion tubing” that sticks up into the inlet, and I’m wondering if that could have gotten damaged somehow (although, if I understand the Helium saver operation, I think that would affect the amount of N2 going through the column, which would be visible in the air/water check).
Any other ideas of things I should check?
Thanks for any suggestions.

Method details:
Inj 300 or 350deg
Splitless with surge, split valve open at 1.5min, pressure surge @20psi for 0.75 min
Hot needle injection (5sec hold before/after injection)
(Above parameters were optimized for my samples)
Single taper liner with wool
Oven 40 to 300deg with 4deg/min linear ramp
MS source 250deg, transfer line 300deg

[shauswirth]

Adding an update and further details.

I replaced the column (+ added a guard column) and all of the inlet o-rings, and cleaned the injector body as per the manual. Things were better for about 10 injections (though I still wasn't getting the same sensitivity that I used to). I'm now back to the bad peak shapes and inconsistent RTs, even after multiple column trims/liner changes.

A couple strange things:

- Peaks before ~25 minutes (oven T = ~150C) of an 80 min run look fine! Then all peaks after that point look terrible (ragged shark fins a minute wide), with major loss of sensitivity of higher MW compounds.
- Peaks in high concentration samples/standards look okay for the ENTIRE run. Still lower peak areas for high MW compounds and more tailing than ideal, but peaks look like normal peaks.
- Low concentration samples/standards look terrible (after 25min), INCLUDING internal standard compounds that are at the same concentration as in high concentration samples/standards!

What can explain this? Severe activity somewhere in the flow path?

Examples of good and bad peaks (both are the internal standard anthracene-d10, at the same concentration) "bad" is in a low concentration calibration standard., "good" is in a high conc. std.





Thanks for any suggestions you my have.

[James_Ball]

Adding an update and further details.

I replaced the column (+ added a guard column) and all of the inlet o-rings, and cleaned the injector body as per the manual. Things were better for about 10 injections (though I still wasn't getting the same sensitivity that I used to). I'm now back to the bad peak shapes and inconsistent RTs, even after multiple column trims/liner changes.

A couple strange things:

- Peaks before ~25 minutes (oven T = ~150C) of an 80 min run look fine! Then all peaks after that point look terrible (ragged shark fins a minute wide), with major loss of sensitivity of higher MW compounds.
- Peaks in high concentration samples/standards look okay for the ENTIRE run. Still lower peak areas for high MW compounds and more tailing than ideal, but peaks look like normal peaks.
- Low concentration samples/standards look terrible (after 25min), INCLUDING internal standard compounds that are at the same concentration as in high concentration samples/standards!

What can explain this? Severe activity somewhere in the flow path?

Examples of good and bad peaks (both are the internal standard anthracene-d10, at the same concentration) "bad" is in a low concentration calibration standard., "good" is in a high conc. std.





Thanks for any suggestions you my have.

I am not familiar with the Thermo instrument, but similar things can happen with Agilent if the Split Vent line becomes contaminated. They can be removed and flushed with solvent and I have seen that come out completely black even after running fairly clean samples for long periods of time. Usually on the Agilent we just replace that portion of copper tubing.

It does sound as some type of adsorption is happening causing problems with the heavier compounds getting onto the column efficiently, and when you overload the adsorbing component then you get better peaks. Any carbonized material in the inlet flow paths can cause it.

[shauswirth]

Thanks for the reply. There's not really a split "line" on the instrument (more a hole in the inlet block). I'll try to flush those out and replace the split vent filter while I'm at it.

[sales01@ql-spe.com]

I’m having an issue with poor peak shapes and declining sensitive with a Thermo Trace 1310/ISQ system with a helium saver inlet.
I started a project about a year ago looking at crude oils/tars (up to C38 alkanes+PAHs)—I haven’t run a ton of those samples (< 100 total) and all samples have gone through cleanup (mostly silica gel fractionation). Since starting those analyses, I’ve had the relatively minor issue of loss of sensitivity of especially higher MW alkanes and C3/C4-alkylated PAHs after a number of injections. This issue was always resolved by changing out the inlet liner and/or trimming the column.

In the last few weeks, things have gone steeply downhill. Peaks have gotten very ugly—very wide, “lumpy”, significant tailing, some compounds show peak splitting, some peaks have inconsistent retention times—and the response is getting very low for many peaks (e.g. high MW alkanes) to the point that some aren’t even detectable in any but the highest calibration standards. These symptoms are showing up even for calibration standards.

Last week, I replaced the liner and trimmed a few cm off the column and got one good injection of a standard, before it reverted to being problematic for all subsequent runs. Since then, I have replaced the inlet liner again, trimmed 0.5m off the column, removed and cleaned the entire inlet insert (as per Thermo’s directions), replaced the o-rings in the inlet, leak-checked the inlet, baked-out the column, and cleaned the MS source (tune looks good and the EM is newish). This did not resolve the problem. Nothing else has really changed with the system preceding these problems, with the exception of temporarily having a different model of autosampler syringe (which regularly clogged) installed – I switched back to the original model before the maintenance noted above and have confirmed it’s drawing up sample.

What else could be the issue? Could the column be totally dead? It’s about 5 years old and has seen some dirty samples (soil extracts, oil), but not a lot of throughput. It seems awfully sudden for this to be the problem and RT shifts have been pretty minimal. Or could it be something wrong with the inlet itself? The Thermo helium saver inlet has a piece of “anti-diffusion tubing” that sticks up into the inlet, and I’m wondering if that could have gotten damaged somehow (although, if I understand the Helium saver operation, I think that would affect the amount of N2 going through the column, which would be visible in the air/water check).
Any other ideas of things I should check?
Thanks for any suggestions.

Method details:
Inj 300 or 350deg
Splitless with surge, split valve open at 1.5min, pressure surge @20psi for 0.75 min
Hot needle injection (5sec hold before/after injection)
(Above parameters were optimized for my samples)
Single taper liner with wool
Oven 40 to 300deg with 4deg/min linear ramp
MS source 250deg, transfer line 300deg

But helium is very expensive gas,i do think hydrogen is better economy solution to replace helium.onsite hydrogen generator https://smtlabtech.com/ql-300-pem-hydrogen-generator/ is handy and low cost.

[venki]

Useful info