Chromatography Forum

Hello,
I would like to determine chromatographic purity of trifluoroacetic anhydride. I use direct injection, HP5 column (30m x 0.53mmID x 1.5mcm), injector T 250st, detector T FID 280st, Split 10:1; oven T 40st-5st/min-250st, flow 3ml/min. However precision is low. Does anybody know for the reason?
Thank's for suggestions
ana

Ana,

Do you work in a humid environment?

Are you diluting your neat sample for injection?

How much are you injecting?

Is your precision equally bad for other samples when you inject splitting at a ratio of 10:1?

Have you tried varying your oven or injection temperature?

Are you using steel needles manually or are you using an autosampler?

Are you injecting slowly or in burst mode?

Do you have a side port needle?

Are you using an internal standard?

Are you using area percent as your measurement of purity?

Are your absolute peak areas imprecise, or the area percent values imprecise?

There are lots of questions that must be answered before a good (guess) answer can be formulated.

best wishes,

Rod

Hello Rod,
In my environment relative humidity is about 40%.
My sample is not diluted, because every diluent has a small amount of water, which can transfer the sample into the acid. From the same reason I do not use internal standard.
At first we injected 0.1mcl, later 0.5mcl (5mcl syringe), but there was no impact. The injection mode was fast and we use avtosampler. Ussualy when we split at 1:10 the precision is very good. We did not vary oven and injection temperature.
We are using area percent method and absolute peak area is imprecise and also the area percent. For example, 5 among the 6 peaks (6 Runs) of trifluoroacetic anhydride has the peak area of about 30.000, one has 9000.

Thank's for the answer
ana

It boils at < 4OC, I'd be a little surprised if you obtain repeatable injections without dissolving it in a less volatile anhydrous solvent.

I suppose you could try plunger-in-needle syringes, but I think the method needs revising, at least to accomodate the low boiling point.

You could check if you can obtain repeatable injections with dichloromethane or pentane - and they don't react with moisture vapour.

Obviously, after the injection, any left will react with vapour in the air, so by having it in a diluent should also reduce that effect, and facilitate the addition of an internal standard.

I assume you would be measuring any acid by titration, so you could correct for that - if necessary, but I suspect that dilution will be the solution.

Bruce Hamilton

Hello,
I am still in troubles with the method-the same problems. I also try with another column, another instrument etc. I was looking again for the diluent however every diluent contains a bit of water which can convert anhydride to the acid. 'My' method is our vendor's method and I wonder if they validate the method. Maybe they just inject once the sample directly. I can not ask them because I would not get the answer.
Does anyone know for the other good method for the assay determination?

Thank's
ana

Putting the anhydride into vials and using an autosampler to inject submits the anhydride to exposure of water and forms the free acid.

This makes any accurate analysis difficult to achieve.

However, if you dry your diluent, vials, syringes, and other materials that may contact your sample thoroughly, and inject onto your injector and column before the actual analysis attempt you should be able to get a fair estimate of purity.

Spectroscopy and wet chemical methods will be of great value in determining the purity of your anhydride.

Call Sigma Chemical and ask them how they verify the purity of their anhydride. Their technical service assistance is free and usually most helpful.

best wishes,

Rod

Thank's for the answer Rod. We've allready asked Sigma and hope to get something.
Bye
ana