GC Acetone peak area issues

[sianq]

Afternoon all, here's an issue from a new user of this forum.

I've recently been running an assay to determine levels of Triethylamine in an API product we make. The method calls for an internal standard to be used - acetone. The solvent is DMF. Sample prep - 50mg sample, 3mL internal standard, mix, sonicate until dissolved.
Oven at 70C, injector and detector at 150C, column Unisole 10T + KOH (20+4%) / Uniport C, 80 – 100mesh, 3mm i.d. x 3m, glass tube.

We are observing random low peak areas for the acetone peaks, sometimes only 10% lower than expected, other times up to 60% lower. It happens to blanks, standards and samples. The TEA peaks don't behave the same way when these anomalies are observed so we've eliminated injector/instrument repeatability problems (plus the %RSD check before the run is always excellent)

Just wondering if anyone out there has ever come across such a problem and if so, what causes it and what steps can be taken to stop it.

[chromatographer1]

Your problem I believe is in your sample preparation.

Acetone is leaving the API solution before you are injecting it into the GC.

Rethink how you are making or handling your samples.

Dissolve your API *before* you add your IStd.

best wishes,

Rod

[sianq]

Hi Rod,

The sample is dissolved in the Internal Standard solution which contains the Acetone (1mL acetone to 50mL with DMF, then a 1/100 dilution scaled to the amount of solution required).

It is very difficult to understand why this is happening as all samples and standards are made up using the same IS solution and all samples are treated in the same way.

Any other thoughts are appreciated!
Sian

[chromatographer1]

Rethink your sample prep. Don't add a volatile Istd and then do a lot of manipulations diluting.

Do not add IStd in the dissolution solvent and then sonicate it.

DO NOT.

Use a non-volatile Istd if you cannot change your prep.

best wishes,

Rod

[sianq]

Unfortunately I have very little choice in the matter as this is a method transfer. The sending lab did not have these problems during method validation or QC testing and don't know why we're seeing it.

I was just asking if anyone else had seen it and if they had, what I could do to reduce it.

Thanks!

Sian

[sianq]

I should also say that the standards (which are not sonicated) also exhibit low internal standard peak areas, but not as significant (maybe 10% less than expected).

We carried out a brief test on the effect of sonication on sample prep as I suspected that the sonication was 'boiling' off the acetone. We sonicated one sample just long enough to dissolve it and one for 30 minutes extra. Our results showed no difference in acetone peak area at all!

During our subsequent investigations, samples prepped together, same sonication time etc. randomly showed loss of acetone in one prep only. It's baffling!

I inject a blank each time the assay is run.
I have seen in the past (only twice though) a blank showing low acetone peak areas. On revialling of the solution (from the same flask, no manipulation) the acetone peak area was as expected. Obvious loss of acetone going in to the vial....but how?! The same solution used to prep samples and standards had the expected peak area.

[chromatographer1]

What kind of septa are you using in your vials? Butyl rubber?

Try silicone rubber coated with teflon if possible or aluminum foil coated septa.

Good luck.

Rod

[chromatographer1]

Are you using an autosampler?

Some are known to store samples in an vial holder that gets warm from the heat of the injection port. This causes undue warming of the sample vials.

Another issue is that the vials may not be smooth and flat at the septa sealing surface causing the vial to leak.

Good hunting.

Rod