GC peak shape troubleshooting

[harringg]

I ran a set of samples and check standard a week ago on "column1", total injections to date=190,and peak shape was good. I came back after the weekend and ran on the same column and have severe tailing. I put in a different (used) column, "column2", total injections to date=350, and same poor chromatography, which at last run was acceptable.

Agilent 6890/Autospec

Here's what I've done.

Finished one project.
Remove column from that project.
Replace GC liner, injector septa, gold plated injector seal.
Install column1
Run samples, all is good, including last check standard after run (valley between co-eluting peaks = 6%).
No one else uses this instrument and it sat over the weekend as routinely does.
Run standards, poor peak shape. (no instrument/column changes made between "good" and "bad" runs) (valley between co-eluting peaks = 15%)
Replace retention gap (deactivated FS, ~1M), Union and 1 loop of column1 (DB-5MS, 15M, remove ~0.5M) (valley between co-eluting peaks = 18%, no improvement)
Run standard, same poor peak shape (earlier RT due to column trimming).
Visually inspect liner and injector seal for septum or other debris, none noted.
Install column2
Run standard, same poor peak shape as on column1. (valley between co-eluting peaks = 15%)

I'm not having any air leak issues, using 18/28/32 m/z ratios as diagnostic as well as argon leak testing the column fittings/union. Pressure is holding fine. Even with the poor peak shape, retention time is stable (except when 1 loop of column1 was removed of course).

The GC shares a He tank with a second GC and that instrument is running fine. Just trying to rule out a "bad" GC tank as well.

I don't have access to a new column, but will try a third used column that ran good too when last run. But it's like everything "fell off" after that good run, and I'm trying to figure out what else to look for.

Thanks,
Grant

[chromatographer1]

Are you using freshly made samples and standard solution?

best wishes,

Rod

[harringg]

Are you using freshly made samples and standard solution?

best wishes,

Rod

Yes. The first standard I used was made the same day it was run (good run) and the "bad" run was a new standard as well. To clarify, the standards are in a stock bottle and just pipetted out the day of the run to a GC vial. So for all things being equal, the good run and bad run used the same standard. And I've been using these standards for awhile now.

I changed liner, injector septum and re-ran and no improvement, and no worse. I'm going to switch to the previous column (DB-1MS) from the project before I switched and see what that looks like tomorrow. If that's good, it will point to the current columns. If it's bad, it will point to something with the GC system or MS transfer lines.

I checked the injector syringe and that looks fine, and we have solvent rinses before and after each injection, and there is solvent in the rinse bottles. Trying to cover all the obvious points, but if I've overlooking a troubleshooting item, please advise.

Thanks

[Don_Hilton]

The question is what kind of samples and what inlet temperature? If the column goes bad after injections and inactivity, I am wondering if some stuff sits in the inlet and slowly chars or otherwise cooks to become something that kills the column. If this is the case, a change in inlet temperature may help -- or you may need to pull the liner at the end of a sample sequence.

[Peter Apps]

When you change columns do you also change the retention gap ? What connection do you have between retention gap and columns ? You mention "MS transfer lines" - what are these on your setup ?

Have a close look at peak shapes and retention times - the poor resolution can be a symptom of; symmetrical broadening of one or both peaks, tailing of the first peak, or a very small relative retention shift.

What solvent are you using, how much sample do you inject and with what split ratio, and what is the starting temperature of the oven programme ? What compounds are in the affected peaks ?

Peter

[harringg]

I switched to the DB-1MS and it was just as good as before. So that helped me rule out something in the "system". Liner/injector seal/injector septum/syringe/etc... all had to be good.

I'll followup with a third DB-5MS column next week and see how things look. It may be as simple as a dirty sample caused damage to the column, but it was so "on/off", looked good then was bad that it caused me try to figure out what happened.

Thanks to all for the feedback.

p.s. MS transfer lines are the heated region between the GC oven and the inlet to source of the Mass Spec.

[Bigbear]

Most poor chromatography is due to the inlet. With that said, seeing all you did to the inlet leaves me wondering how long this inlet has been in service? You may need to either replace the injector weldment or gun brush it ( check Agilent.com for instructions). Don't forget the split vent line and filter ( replace them).