Chromatography Forum

I am trying to separate the polymer impurities. the basic structure is poly ethylene glycol (PEG) chains linked by small molecule about 150(MW). I may have four of them. All are 30K PEG with different links. SEC probably unlikely to separate them. So i am willing to try any new methods such as Rp, Np, HILIC, graphite, multiple phase adn so on.

Any suggestion will be very appreciated.

The obvious problem is that you essentially have a very small difference in a moderately large molecule. The real question is how those small-molecule links differ from one another. If the differences are such as to orient the PEG chains differently, then something like graphite mike work surprisingly well. On the other hand, if the PEG attachment part of those molecules is similar, and there are small differences elsewhere, you might have a real problem.

Is there anything you can do to "chemically amplify" the differences (e.g., make derivatives)?

If you links are ionizable fragments you can try to use combination of ion-exchange/ion-exclusion with size exclusion.
We did something similar for one of our customers when links contained amino groups.

regards,

Vlad

Tom and Vlad,
Thanks for your reply.
these impuritys consist of some combinations of some limited number of functional groups such as PEG-A-(BCOOH)-PEG, PEG-A-B-(BCOOH)-PEG, PEG-A-PEG-BCOOH. therefore there might 1 PEG, two PEG, ot more, there could be one A, or more, one B or more.
BUt the PEG is 10K Mw, and the A or B only about 150.

I will try graphite.

I am not sure how ion exchange works since all COOH is connceted by the same chain such as always ~COCH2CH2COOH so the pka difference among these COOH will be very very small.

I am also wondering how we can reduce the effect of the huge PEG and enhance the effect the link? I guess one idea is to compact the PEG. Unfortunately, PEG is soluble both in water and in organic and i do not know what solvent can compact the PEG.

SEC can tell you just how similar your products are regarding MW and structure. I would start with SEC here, rather than assuming that all products are similar in the mentioned characteristics.
In protein chemistry very similar species are sometimes separated on RP using a gradient, I think it was Uwe who mentioned in another threat that this is done for polymers, etc., as well.

Maybe not as difficult as you feared.

First off, you have to be able to separate based on different numbers of PEG units. SEC should work for this, but you might also see pore-size effect with a reversed-phase column of "standard" (e.g., 100 Ã… or so) pore size.

Within each "PEG units" family, you may be able to separate on the basis of number of COOH groups using something like ion-pair chromatography (actually, an ion-exchanger of appropriate pore size should do the second sort of thing). Whether you can accomplish that on a single column or whether you will need to do {SEC first, collect fractions, run IP on each fraction} is an open question.