New method - new impurity

[Mattias]

Dear all,

The USP and Ph.Eur method for 5-aminosalicylic acid (5-ASA) is a ion-pair, reversed phase method (C18 column). I wanted to analyse the substance with LC-MS so I tried a Primesep 100 column, and got good retention with a simple mobile phase (0.05% TFA in 30% ACN).

The thing is that I now see a "gigantic" impurity (5-20 area%) eluting a couple of minutes before the main peak. The impurity is present in all tested 5-ASA batches, including the newly purchased CRS standard from USP. The LC-MS and UV data suggests that the impurity is related to 5-ASA. Mass (M+H): 304 (154 for 5-ASA), UV max (three) shifted towards longer wavelengths. The mass is not mentioned in any of the references I have found about 5-ASA degradation.

The peak is not present in the blank injections, or in the injections of the known impurities.

I have isolated the peak, but it is not stable after isolation (t½ about 30 min). And it is converting back to 5-ASA! I have tested to dissolve the substance in many solvents/buffers: no difference

Right now I don't know what to do with this information. Is it a true impurity, or could it be some kind of artefact still? Some kind of unknown equilibrium in the aqueous phase?? If it would be a "real" impurity it should make the original method unprecise, but I never see any problem there. Any good suggestions?

[tom jupille]

Mattias, when you say it's converting back to 5-ASA, how are you measuring that? By the same method you used to generate it?

And does it return to that same ratio of "impurity" to 5-ASA?

My thought on first reading was a dimer or aggregate, but that doesn;t explain the 3 unit difference, so I'm out of ideas.

You might also try cross-posting this in the LC-MS section and see if some of the more MS-oriented people might have some suggestions.

[SIELC_Tech]

Hi Mattias,

We run this method with various mobile phases (UV and LC/MS):

http://www.sielc.com/compound_286.html

I need to look at individual injections and see if we observed something similar.
I would suggest reproducing method from the link using formic acid instead of TFA. May be TFA catalyzes some kind of condensation/decarboxylation.
There is nothing unusual to see some additional impurities with Primesep columns. We once analyzed “pure standardâ€

[Mattias]

Tom> I have collected the peak manually (LC-vial in the outlet of the detector). Then I reinjected it in the same system. First injection gives only one peak ("impurity"), then two peaks (impurity +5-ASA), at last only 5-ASA.

Vlad> Thanks for the link! I will try formic acid and see how it looks. The detection is at about 300 nm, so there should be no problem.

I'll let you know how it goes.