[mmega]

Hello,

I am currently developing a residual solvent analysis method by Gas chromatography equipped with a flame ionization detector. My sample is an API with a phosphate salt attached. One of the residual solvents of interest is triethylamine. When my sample is spiked with the residual solvents, all other solvents show recovery except for triethylamine with a recovery of ~75 %. I think it is because of the triethylamine reacting with the phosphate salt in the sample. I have tried adding a base such as NaOH to remove the phosphate salt from the sample first, then adding the triethylamine, but the recovery is still not showing. The sampler used is the headspace sampler and the diluent is NMP.

Please let me know if you have any advice on how to make the recovery better.

Thank you.

[rjb]

Hi Mmega,

This question rings a bell... Did you not ask the same question on a manufacturer (Agilent?) GC forum? I'm guessing you didn't get any response from there that worked?

It seems likely to me that your matrix is having a significant influence on your results for TEA and your attempts to address that haven't resolved the problem. If you can't resolve it, you'll have to work with it, probably by matrix-matching your calibrators rather than deriving your calibration curve from TEA in clean solvent which is what I assume you're doing. The basic idea is that each calibrator is made up in something as close to your matrix as you can get meaning that each calibrant is affected in the same way as your sample. Doing this should get much closer to the 100% 'recovery' you're looking for this way. Failing that, there are alternatives, but these are far from ideal.

As an aside, can I ask what kind of column (i.e. DB-624), column dimensions and GC set-up (injector/oven parameters) you're using please and what's your API?
How does your chromatography look and do you have peak tailing?

R