Chromatography Forum

Please help,
I am working on identifying an unknown compound eluting from a drug intermediate sample. It has been thoroughly identified as a contaminant in the intermediate however as stated it is my task to identify the structure. The column is a water's atlantis 4.6x100mm. The mobile phase is 0.1% Phosphoric Acid in water (A) and Acetonitrile (B). The original method was not developed by me but has been used for quite some time in identifying impurity limits.

The basics of the problem are this, at higher concentrations of the sample two compounds have been identified one with a mw = 546 and the other eluting slightly later with a mw = 458. After a number of experiments we are leaning towards teh possibility of water causing some reaction that forms the earlier species (mw = 546). However due to water in the mobile phase I can not be 100% certain. My question is this, is there a way of testing that water is causing this reaction without switching to a normal phase column? as this would require more time and effort than I am allotted.

At higher concentrations of intermediate sample in (NEW) DMSO only mw = 458 is observed at T = 0. (Sample was prepped and run immediately) Over 72 hrs mw = 546 starts to appear. Similar study in acetone reveals the formation of 546 at lower sample concentrations. Note, in no study has compound with mw = 458 been eliminated.


I know this post is unusually long but I appreciate any responses. Thanks!

Unless you suspect (for structural reasons) that reaction with water would be very rapid, I don't think the presence of an aqueous mobile phase would, per se, be a problem. If a compound reacts "on-column", the result is usually a misshapen peak (anywhere from tailing, through a split peak, to a "blob").

Do you have any other way of monitoring the putative conversio (e.g., spectral changes)?

I did a fairly extensive literature search last night and found that it is also unlikely to have any reaction after injection, ie. as a result of the mobile phase, which on retrospect makes sense. I am in the process of collecting enough sample to have identified via 2D NMR as well as MS/MS. Both compounds are in very small concentrations (<0.30% AUC). The next step is to control any reactions from taking place, I have a few ideas including pH and Temperature, any other suggestions?

1) if you want bigger amount of the impurites you want to identify, you may do forced degradation using the conditions you know accelerate the reations;
2) if you want to know how much the impurity(ies) is generated before and during the HPLC, you may prepare the sample in your mobile and inject to LC ASAP and one time after some intermitter such as at 0.5, 1, 4, 8 , 16, 32, 64. As it's low concentration, the rate should be at linear area so you can infer the amount generated during the injection.