Chromatography Forum

I am trying to separate some extremly polar zwitterionic compounds (containing two sulfonates and two primary amines - bis-sulfonated diphenylethylene diamine if anyone is interested).
My question is are zwitterionic compounds retained better on reverse phase columns than species with an overall charge? Or is the major factor simply the total number of charged sites on a molecule?
Also if anyone has any tips on how best to go about this separation it would be great. Only problem is I need my method to be suitable for prep, and the only prep columns I have access to are C8 and C18.

My question is are zwitterionic compounds retained better on reverse phase columns than species with an overall charge? Or is the major factor simply the total number of charged sites on a molecule?

To a first approximation, it's the total number of charged sites.

Also if anyone has any tips on how best to go about this separation it would be great. Only problem is I need my method to be suitable for prep, and the only prep columns I have access to are C8 and C18.

I suspect the only way you would get that kind of compound retained on a C8/C18 column is by using an ion-pairing reagent, and then you would be faced with getting rid of it.

Maybe you can buffer it at the isoelectric point, use a low organic mobile phase and hope for the best?

Probably most zwitters show their highest polarity at the isoelectric point.
Chances are that you will have to invest in a NP column (HILIC). (It could be very instructive to get some info from people who have done HILIC on a prep scale).

Actually, zwitterions are least soluble in water at their isoelectric points. Since this is a prep method, ion-pairing is out: cation exchange would be at pH ~1, and anion exchange at pH ~10. I vote for HILIC. Actually a plain old silica column is a good HILIC phase.

As we are on for alternative to clasical silica bonded C18 stationary phases, a nearly perfect (but very expensive) stationary phase for this compound would be Hypercarb (i.e. porous graphitic carbon) due to it's high affinity for phenyl containing compounds (pi-pi interactions) and it's unique ability (for a reversed phase material) to retain anionic compounds.

Mark, why should zwitterions generally be less soluble in H2O when they are the most polar version of the molecule? I know of one possible exception, though: Baclofen(R), a chlorophenyl GABA derivative, had the highest retention at its isoelectric (in the meanwhile I wonder whether I made a mistake while arriving at this finding).

Here are two application for similar sulphonated zwittre-ions on Primesep columns:

http://www.sielc.com/application_137.html
http://www.sielc.com/pdf/SIELC_Taurine.pdf

Your compound might retain similar. If you send us compound we will run it in the lab and try to develop a method for you (prep compatible).
Contact me if you have questions.

Thanks everyone for the replies

Sielc_tech: I see that the examples are for primesep A, D, SB and 100. I have an analytical primesep C, is this a suitabe column, or would one of the others be better in this case? I am currently trying to get retention on this column, if we do get a sep on any column we would be willing to get a prep version. I can send some sample to Sielc, although would it have to go to the US? I am based in the UK.

MarkTracey /HW Mueller: When you say a silica column, do you mean normal phase? (Forgive my ignorance, but before now I have only dealt with chiral NP separations). And do you need particular types of silica columns to run aqueous eluents?

Silica columns are usually run under NP conditions (no water) or HILIC conditions (<25% water).

To me, it seems like an ion exchange would be your best bet to separate the species mentioned. I'm not sure how RPC would work, but it's worth giving a try. Just note that it doesn't really matter if it's a C8 or C18 column. The degree of carbon doesn't really make a difference. It's more of a marketing tool. The functionality and selectivity is more to do with the ligand than anything else.

What scale are you working at?

Before I forget, you can run normal buffer conditions on silica, it just depends mostly on what you're working with. If it's nonpolar organics, the you want to go with solvent. If it's globular protein, you can go with a buffered mobile phase (more applicable to other modes than RPC).

TosohTechman, what ligands are you talking about? What evidence do you have that there is no difference between a C-8 and C-18?
Your second entry: With "solvent" you mean organic solvent? With other words you are restating Mark´s contribution? (Apparently, I am having enormous problems with modern (?, slang?) terminology.

My apologies for the confusion. What I should have said is, in terms of selectivity, you can't assume that a C18 is going to work better than a C8 column. Selectivity is relative to the interaction of the specific molecule in question with the support. When I say ligand, in this case, I'm referring to the R group (C8, C18) on the resin. I started to speak in general terms a little when addressing cleaning procedures as they relate to all types of resin. Also, when I say solvent, yes I do mean organic solvent.

TosohTechman: I am currently working with analytical loadings, but I will want to separate up to 1g of material when I get a suitable method.

I will be encountering this problem again I am sure, albeit with less dramatically hydrophilic compounds (compounds with multiple amine and carboxylate groups, and metal complexes).