Cyclic Siloxane Contamination

[Travisdog2]

Hello,

I've just started working on developing a splitless method using an Agilent 7000 GC-MS/MS equipped with a DB5MS-UI (Agilent) 30m 0.25/0.25 column using a mixture of 4 analytes dissolved in IPA (starting initially) at 10 PPM.

A couple of runs into the work, a bunch of peaks (mostly cyclic siloxane derivatives) began to appear in the chromatogram which to my eye at least appeared to be most likely to have come from the septum and became progressively worse as more samples were run. Fairly obviously I changed the septum, but as the peaks were still present, I figured that perhaps the liner had become contaminated with septum particles, so I swapped it out for a new one expecting to see a beautiful baseline. I was much disappointed when I saw nothing but a very clean liner and contaminant peaks that were still of a similar height to one of my analyte peaks!

I did a 300C ‘bakeout’ of the injector (as recommended) which reduced the peak sizes temporarily, but they came back with a vengeance fairly rapidly. To my mind, the peaks are most likely to be from a septum (injection port or vial) so my plan over the next few days is to try to determine this by doing some blank runs without an injection, to see if the problem is injector based, then try runs without a vial septum to see if that is the problem as well as checking the needle for any snags. I have to admit that I have done a few runs taking sample from the same vial multiple times over a couple of days (not great practise), but IPA is pretty innocuous stuff, so I can’t imagine that it’s breaking down the material (Silicone/PTFE) and causing issues.

Before I start going mad and changing out the gold seal and everything else I can think of, does anyone have any further suggestions please?

[70 eV]

Throw a Merlin microseal in and change the liner/gold seal.

[James_Ball]

Have you trimmed any from the column?

If a small septa particle gets into the head of the column you will see similar problems and it will continue until it is removed.

[Travisdog2]

Thanks for the replies so far...

Having done a couple of test runs today I think I have isolated the problem to the vial septa. A run without injection yields a nice flat(ish) and peak free baseline which suggests that the liner and injector are free of contamination. A solvent run without vial septum yields similar, so I reckon this is some kind of problem with the vial septa/syringe.

I haven't had the chance to get the syringe off to have a look at the needle to check for any deformation which might be pulling bits of vial septum off (which somebody suggested in a post on this forum back in about 2012) - I did notice that the vial seems to be pulled up from the auto sampler when first injected which I haven't seen previously with this instrument - Is this unusual? I suspect that this might be the issue but have put on some runs with various different types of vial septa in case I have a bad batch... Not a problem I have ever seen before - Anyone else?

Muchos Gracias

R

[Travisdog2]

Hello,

Thanks again for the responses.
Just to end this with something potentially useful to a future chromatograher with issues...It looks like the issue is a combination of syringe needle and vial septa. The following are images after 5 injections using 2 different vial septa brands. In the case of the 'thermo' septum, the contamination present in TIC runs was present from the first injection onwards and became progressively worse, in the case of the Agilent vial, its barely there at all in the first couple of injections, but then becomes more pronounced.

Agilent cap after 5 injections - Not even obvious its been penetrated...


Agilent cap inside surface after 5 injections


Thermo cap after 5 injections - Not great and possible evidence of coring with a little worm of materials obvious


Thermo cap inside surface after 5 injections - A worm of silicone trying to escape!


TIC result after 3 runs - 'Thermo' Cap



Thanks

TD2

[Peggsy]

@Travisdog2

Great write up, thanks for taking the time to note your approach and outcomes.

cheers
Peggsy

[DR]

Have you tried pre-split septa?

[indium]

IPA will extract siloxanes from the vial cap silicone septa even with a PTFE barrier. I see this frequently in LC-MS methods that use IPA sample solvent. The only solution is to limit the number of injections per vial to one. You might get two or three injections with little to no siloxanes, but it can vary from septum to septum. There are pierceable PP vial caps but evaporation could be a problem.

[Travisdog2]

I had no idea that was the case, I was always under the misapprehension that IPA was fairly kind to silicones; useful to know for future reference.

On the plus side the project has moved to the next stage and so the optimised method is now SIM based, meaning that siloxanes are no longer a problem that I have to worry too much about

The next challenge is moving back to scan and a change of solvent to MTBE... I suspect that'll be an interesting one when it comes to septa compatibility!