getting retention on an SCX column

[MestizoJoe]

Hi everyone. I have a sample which is a weak base and I can't get it to retain more than k = 0.85. I started with 5 mM aqueous sodium phosphate pH = 2.1 .

I thought that decreasing the salt concentration by running at 1.25 mM phosphate buffer pH 2.1 would improve retention and found that the sample eluted in the void volume.

Today I tried 0.2 % TFA adjusted to pH 2.25 with sodium hydroxide and retention was worse.

My next thought is using lithium phosphate but we have none in house. Am I approaching this the right way?

Thanks,

[tom jupille]

Yes, you're going in the right direction. If memory serves, the order of driving strength is Li < H < Na < NJ4 < K.

You might want to try just dilute phosphoric acid (1 mM ?) while you're waiting for the lithium phosphate to arrive, but it may be that you simply need a higher ion-exchange-capacity (i.e., more retentive) column.

[pph31]

Hi everyone. I have a sample which is a weak base and I can't get it to retain more than k = 0.85. I started with 5 mM aqueous sodium phosphate pH = 2.1 .

I thought that decreasing the salt concentration by running at 1.25 mM phosphate buffer pH 2.1 would improve retention and found that the sample eluted in the void volume.

Today I tried 0.2 % TFA adjusted to pH 2.25 with sodium hydroxide and retention was worse.

My next thought is using lithium phosphate but we have none in house. Am I approaching this the right way?

Thanks,

why not try SAX at pH8?

[tom jupille]

why not try SAX at pH8?

Because the free base would not be charged ?

[pph31]

why not try SAX at pH8?

Because the free base would not be charged ?

I guess no retention at low pH on a SCX column may indicate the protein has a low pI, maybe nagatively charged. The nagative charged protein was not retained on SCX column.

[MestizoJoe]

The sample is a small molecule with an amine functional group and may possibly have a carboxylic acid group. I thought about increasing pH to 3 using formic acid as the buffer, but I figured if there is a carboxylic acid group on the molecule it will retain less. I may try this today just to see what happens.

[tom jupille]

If you do have an acidic functional group on there, then anion exchange is, in fact, a possibility. If the pKa's overlap so that it's a zwitterion, then something like HILIC might be a better bet.

[HW Mueller]

What are you trying to do? Characterize this molecule, is it pure or are you trying to isolate it from a matrix? You don´t have access to IR, UV, NMR, MS spectrometry to check on functional groups, etc.?
Now, just a quick over the thumb estimation, the pH of 5 mM H3PO4 should be higher than pH 2.1, a 1 mM H3PO4 would be expected to give a pH near 3. What is LiHxPO4 going to bring? You don´t know what the dissociation constant of the base is, but it looks like it is hardly, if at all, ionized at the pH you have really had.

[MestizoJoe]

The goal is to characterize. The sample is in a matrix of tablet.

I'm not sure what you mean when you say pH should be higher. For my 5mM solution of phosphate I pH adjusted it to 2.1.

The Lithium in the lithium phosphate is a weaker exchanger than sodium. The idea is that the molecule will retain longer on the column using lithium instead of sodium.

The pKa of the base is around 9 since it's an amine. It ought to be fully ionized.

Tom, I tried 5mM TFA pH 2.2 and 1mM phosphoric acid and got better resolution but not much better.

I like the idea of using HILIC but I'm certain no one else here will. Thanks for the help, unfortunately I have to move forward with this.

[carls]

If your compound is a base it should be well retained with these low ionic strength mobile phases. Are you sure its not an amide? If not, you may need to clean the column with high salt concentration. A flush starting with 10 column volumes (CV) of ~15mM phosphoric acid in water followed by 25 CV 0.1M KH2PO4 (no pH adjustment needed) then 25 CV 0.5M KH2PO4 (no pH adj needed) and finally 10 CV ~15mM phosphoric acid in water should do the trick but check with the column manufacturer for more specific suggestions

[HW Mueller]

Ok, I checked

http://www.liv.ac.uk/buffers/buffercalc.html

It turns out that I was a bit off on my guess as to the pH of 0.005 M H3PO4. One needs about half the moles in H3PO4 and half the moles in NaH2PO4 to get a pH of 2.1. The point I was trying to make is that you have such a low M+ concentration that you will not gain anything by using Li+, there is obviously something wrong here. First you say that you have a weak base which could make one to think that it is hardly ionized at your pH, then you mention that you expect a ~normal (pKa = 9) amine, which would be totally ionized and retained as stated by Carls already (I am a bit more careful about predictions today, but still it seems that even a very low capacity ion exchange should do the trick).
So, if the goal is to characterize, it might be time to give up the idea that you have an amine?
(Incidentally, it might be of relevance that Beynon of the above link thinks that the calculator should refuse answers at considerable lower pH and/or concentrations).

[Peter Apps]

The goal is to characterize. The sample is in a matrix of tablet.

Please tell us that this is not reverse engineering.

Peter

[MestizoJoe]

I don't know if this is 'reverse engineering.' There is a deg that has appeared in a certain drug and needs to be characterized to move forward.

My best results so far are using 5 mM TFA (aq), 5 mM H3PO4 (aq) which seem to be equivalent. I'm moving forward with TFA so the method will be MS friendly.

I see what you were trying to say before. This deg comes from a drug that has an amine, but that doesn't mean the deg has to have an amine.

I may have to approach this using reverse phase.