Agilent 6890 GC analysis of diethylenetriamine

[Mike H.]

I'm trying to assay diethylenetriamine. I'm using an Agilent DB-35 column (30m x 0.53 mm x 1.0 µm) with a base deactivate single taper liner using base deactivated wool. I've encountered a couple problems that I wanted advice on.

#1 When injecting 100ppm DETA standards, I never seen any peaks until I inject a 1000 ppm standard first. Is it common for this analyte to have to prime the system before low level injections produce peaks?

#2 After injecting a 1000 ppm standard I can achieve 0.999 linearity using 0, 500ppm and 1000 ppm standards and achieve an <10% RSD for 5 injections of the 100 ppm standard, but I'm seeing what appears to be much greater tailing than a comparable chromatogram from a Restek method http://www.restek.com/chromatogram/view ... 7/111-40-0
I don't know how this could be so when all the contacts points ie. liner, column are base deactivated. Where else in the GC flow path could the DETA be sticking to?

Thanks for any ideas.

[James_Ball]

Liners are a protective mechanism in an injection port but not all of your injected analyte remains in either the liner or column. When you think about things being in a gas phase, and how fast those molecules are moving, there are certainly some that move outside the liner before entering the column, even as far as the split line. We have found when doing splitless injections that even a dirty split vent line can cause tailing and breakdown of analytes.

If you are using an Agilent GC, make sure that you remove the split vent line and clean or replace it often. Also be sure to clean the port where it attaches to the injector to remove any contaminates there.

Column insertion depth and the smoothness of the column cut can also affect peak shape.

[Mike H.]

Thank you. I'll mention this the next time the Agilent tech comes to do a PM.

[Yama001]

We have noticed the 6890s onwards have significant interaction with the upstream plumbing as well. This happens less with a split injection, if you can afford the sensitivity loss.

If stuck with a splitless injection, solvent rinsing the wldment gas lines helps for a few days. A pressure pulse injection may help as well.

[AICMM]

Mike H.,

You might consider two other things (in a discussion with someone at Restek for example.) Consider something like a Volamine column which is tailored to this type of analysis and consider the pinched liner with the hole at the top so that the analyte is not exposed to the metal surfaces at the bottom of the injection port.

I've never been a fan of shocking the system since you don't know where the passivation has worn off in your sequence (right after you low standard????)

Best regards,

AICMM

[mrussell]

Liners are a protective mechanism in an injection port but not all of your injected analyte remains in either the liner or column. When you think about things being in a gas phase, and how fast those molecules are moving, there are certainly some that move outside the liner before entering the column, even as far as the split line. We have found when doing splitless injections that even a dirty split vent line can cause tailing and breakdown of analytes.

If you are using an Agilent GC, make sure that you remove the split vent line and clean or replace it often. Also be sure to clean the port where it attaches to the injector to remove any contaminates there.

Column insertion depth and the smoothness of the column cut can also affect peak shape.

And of course the split vent trap.

For amines I use resteks sky deactivate liners and siltek inlet seals and either a base deactivated wax, db 5 amine or my current favorite, rtx-volatile amine and the similar agilent volamine column. (With an NPD with a blos bead. I like helium as carrier and makeup to deliver a constant flow to the detector, of a constant gas makeup). For particularly bad ones a direct injection liner can help. If you getting degradation in the syringe itself, she makes a series of low dead volume autosampler syringes that can be fitted with needles with hydrophobic or hydrophillic coatings.

Some people passivate the instrument with a shot of pyradine or

such. On some analyses Ive tried adding 50uL of pyradine to the samples themselves. If I have good separation then it seems to work OK and it pasivates every injection. Another way I've done it is to simply have two methods, and do a shot of pyradine, with a run that takes maybe a min, with a high oven temperature and flow, then a sample, with my run conditions, then repeat.

The blue liners and the volamine type columns have eliminated most of that. Agilent has a whole line of ultra inert stuff, from coated deactivated metal ferrules, to ultra inert liners and inlet seals. My take away from the webanar though was that theirs is better, but their competition is almost as good, and in fact usually good enough at a substantially lower cost. In fact, the number on place I'd go if I really needed to push the limits of sensitivity would be to a direct injection liner. It takes the metal in the inlet out of the equation for the most part. (Restek makes an entire deactivated inlet, as well as a deactivated for the detector. I'm skeptical about the jet though. You get less than a half cm of contact between eluent and jet, before its in the plane, and honestly, suppose it breaks down on the Hort metal surface, sine I've not seen my jets clogged with carbon, it must get flushed out, which means its still burned, which means I still get signal. It might cause some tailing, but if it does, it has not been enough to notice above all the other amine issues.

One problem I've had is that some methods collect on a phosphoric acid coated sampler and then later neutralize it with hydroxide. The excess hydroxide can quickly foul your liner. (Like 5 or 10 injections)


If I were in your position, not having the things I have around to deal with these analytes, I'd try passivating with pyradine. I'd also call restek and get a rtx volamine and get a direct injection liner.
.

[mrussell]

I'm trying to assay diethylenetriamine. I'm using an Agilent DB-35 column (30m x 0.53 mm x 1.0 µm) with a base deactivate single taper liner using base deactivated wool. I've encountered a couple problems that I wanted advice on.

#1 When injecting 100ppm DETA standards, I never seen any peaks until I inject a 1000 ppm standard first. Is it common for this analyte to have to prime the system before low level injections produce peaks?

#2 After injecting a 1000 ppm standard I can achieve 0.999 linearity using 0, 500ppm and 1000 ppm standards and achieve an <10% RSD for 5 injections of the 100 ppm standard, but I'm seeing what appears to be much greater tailing than a comparable chromatogram from a Restek method http://www.restek.com/chromatogram/view ... 7/111-40-0
I don't know how this could be so when all the contacts points ie. liner, column are base deactivated. Where else in the GC flow path could the DETA be sticking to?

Thanks for any ideas.

The column itself. Your column is a db35. Theirs is a db35 amine, which is specially deactivated. I think they might have pulled a few more tricks too.. Rtx volatile amine column, as well as agilents volamine equivalent are even more refined. The used a bunch of tricks that weren't ready for prime time when the 5 and 35 amines came out.