Chromatography Forum

I need help with a column/mobile phase/api chemistry interaction theory involving the USP hplc method for pseudoephedrine hcl. This is a current debate in my lab as we attempt to verify and optimize the method for our use.

The column is an eclipse xbd phenyl and the mobile phase is a 90/10 triethylamine buffer pH=6.8/methanol.

I was not able to achieve the resolution I needed so I added water through the switching valve that ended up in 97/3 (buffer/meoh)/water. This decreased the ratio of triethylamine to methanol. In doing this, the resolution was increased and the peak shape held up.

My hunch to do this was based on my understanding of the chemistry involved with the amine. The amine competes with the pseudo for silica binding sights. In adding water, I decreased the amount of the triethylamine (minor component), thus increasing the probability of the pseudo interacting with the silica and increasing retention.

Is this what is going on and is this why I was able to achieve better resolution?

Another chemist working on this did the same thing, but instead of adding water, she just added more buffer in the same amount as I added water. I said that with the ratio of triethylamine to meoh increasing, she would see worse resolution and peak shape. This is exactly what we saw. Of course she disagrees with this explaination of the result and is pushing to purchase more columns.

Can somebody help me with this chemistry problem? Thanks a bunch.

My guess: this is an old method, the TEA is there to "mask" unwanted interactions of the stat phase with your ephedrine. You changed the pH advantagously, the other chemist got a lesser pH change and apparently to the worse.

I don't have the current USP, but silica columns were used recently, along with 97:3 alcohol : aqueous 0.4% ammonium Acetate. The EP used a phenyl column with 94:6 ammonium acetate buffer pH4 with methanol for related substances.

I'd change the method.

Bruce Hamilton

As a gross oversimplification, most of the retention in reversed-phase chromatography comes from hydrophobicity. Secondary interactions with residual silanols are usually just that, "secondary". They can tweak retention allowing separation of two compounds of similar hydrophobicity.

Reversed-phase LC is surprisingly well explained by boy-scout-merit-badge level theory: "like dissolves like, and like sticks to like". You have a non-polar stationary phase (phenyl groups). As you make your mobile phase more non-polar (i.e., increase the methanol/water ratio), it becomes more effective at washing your sample off the stationary phase.

What both you and your colleague did was to decrease the amount of the strong solvent (methanol) in the mobile phase. The result was, predictably, longer retention.

Tom, maybe I misunderstood, but to me the original question seems to indicate that the results with H2O and buffer were sort of opposite.

I think too, it might be some pH contributions. Or….it just might be a case of not fully controlled experimentation – e.g. different columns, equipments, separation temperature etc.

Best Regards

Am I correct that your buffer is only phosphoric acid used to control the pH at 6.8? If so then what you are saying is correct, the TEA is merely blocking active sites on the silica surface. The psuedo increases in retention based on its secondary reaction with the silica surface. I bet if you look closely at your data the psuedo shows an increase in k' and also a slight increase in tailing due to the extra retention the silica provides. You can test this further by cutting the TEA addition in half or even further, then you will really see the effect.

Tom, maybe I misunderstood, but to me the original question seems to indicate that the results with H2O and buffer were sort of opposite.

You're right, Hans. I misread it.

Well, the original question is not that clear, for instance does adding "buffer" mean that TEA was added?
Anyway, there are two reasons for assuming that the TEA concentration is not the main reason for the observation:
1. it would be surprising that the column has a strong deleterious
secondary effect on the analyte
2. if there was such an effect an increase of TEA would improve peak shape, thus resolution, not the other way around. One used to add amines to the mobile phase so that tailing of amine analytes was eliminated.

It's a little off topic, but the xdb series is one of those that was developed for basic analytes, so I doubt the TEA is even needed. There are now numerous methods for pseudoephedrine and related compounds that don't need TEA, and many no longer even need ion pairing agents.

It's a shame that USP methods are 20 years out of date, and yet labs are forced to use them for legal and regulatory reasons.