[chysce]

Hi there. I am doing some solution phase peptide synthesis. I'm using a standard coupling procedure:
1 eq of amino acid ester, 1,1 eq of NBoc protected amino acid, 1,5 eq of HBTU, 1,5 eq of HOBt and 2 eq of DIPEA or Et3N, at room temperature overnight. I'm conducting reaction in DCM rather than DMF for easier workup...

However I am almost always getting a mixtures of diastereoisomers which I can't separate. I know it's a common issue in peptide synthesis and hopefuly some of you have experience with this.

Thanks

[rolnor]

This all seems strange, racemization is more a problem if you make long peptidechains.
How do you analyze your product, If you use NMR perhaps you see different conformations, it depends on what amino acids you use?

[chysce]

I made esters of various amino acids with a chiral alcohol (all are enantiopure compounds) and now I'm trying to add some more amino acids to the peptide... and every time i get diastereoisomers (even tho compounds have only 2 amino acids and a chiral alcohol on one end).

I use NMR for characterization.. and almost all of the signals are in something like 8:2 ratio or so. C13 also gives double signals. I thought those could be rotamers but even when I remove boc group... double signals remain.

I used leucine, glycine, tryptophane, methionine, phenylalanine and proline.

[rolnor]

The only thing I can think of is maby the alcohol reacts when you put on more amino acids? Not a very good explanation.

[kriggy]

Did you try different coupling conditions?

[TheUnassuming]

To piggy back on kriggy, it might take you screening a few different coupling conditions before you find one that is best for your needs.  It wasn't an amino acid, but for one coupling I did that tended to racimize I found using propylphosphonic acid cyclic anhydride gave the best results. 

Is you free acid coupling partner the same in all the reactions you have tried?

[kriggy]

Yeah, thats what I ment. My friends had a great experience with t3p, I usualy make acyl chloride but my acid is aromatic so it cant racemize via ketene formation. Also we do a lot of solid phase chemistry that relies on amino acid amide synthesis, usualy they use HOBT, DMAP, DIC (1eq) in DCM/DMF mixture.

Also, did you try the reaction in DMF? It seems to me that you are following lit procedure but you changed the solvent so it might be the reason why its not working

[phth]

Yeah, thats what I ment. My friends had a great experience with t3p, I usualy make acyl chloride but my acid is aromatic so it cant racemize via ketene formation. Also we do a lot of solid phase chemistry that relies on amino acid amide synthesis, usualy they use HOBT, DMAP, DIC (1eq) in DCM/DMF mixture.

Also, did you try the reaction in DMF? It seems to me that you are following lit procedure but you changed the solvent so it might be the reason why its not working

T3P works best with pyridine.  The side product is ketene formation 1-2 % run at 0º.  20 % epimerization seems like much. OP I would suggest learning about things like HOBt.  DMF doesn't matter if you extract correctly because it is a water soluble impurity.  A common mistake is to evaporate DMF.  Dilute 4x DMF volume EtOAc, and extract 4x DMF volume water, and you will get an emulsion the first time.

[chysce]

Thank you for advices.

I don't have T3P in my lab. I have EDCI though and will try couple of test reactions. One in DMF with previously mentioned conditions and one in DCM with EDCI. Those are my options at the moment.

I'll report back with the results

[clarkstill]

From your description, it sounds like you're doing a coupling on something like this:



I don't know the numbers, but I would guess that, once the carboxylic acid is activated the alpha hydrogen is considerably more acidic here than in a normal polypeptide, due to the adjacent electronegative oxygen? So your molecule might be especially prone to epimerization via ketene formation. Or maybe I'm incorrect in the sort of structure you are looking at, in which case never mind!

A slight aside, but there is a very simple NMR method to check whether pairs of peaks are due to rotamer formation or a mixture of non-interconverting species. If they are indeed rotamers, then if you do a 1D gradient nOe experiment and irradiate a particular peak, the corresponding peak in the other rotamer gives a peak of the same phase. Much quicker and simpler than running VT experiments etc.

Steve Ley talks about it in JOC 2012, 5198.

[rolnor]

Is it perhaps possible to replace TEA with pyridine to avoid basic conditions?

[clarkstill]

Someone with much more experience of peptide chemistry can probably advise, but you could definitely try it out.

Also, I realized about 5 mins after posting that what I said/the structure I gave was stupid - you're probably using an Fmoc or Boc strategy and growing from the N- not the C-terminus?

I need some coffee this morning :S

[chysce]

Just to avoid confusion..

Here's what I'm working with:

(I can't see the image for some reason so here's the link   http://imgur.com/54kUyYG   )

P.S. Thanks clarkstill for that JACS reference.. it will help a lot

[clarkstill]

It doesn't look especially vulnerable to racemization... aapptec have a decent guide to avoiding such problems here: https://www.aapptec.com/aggregation-racemization-side-reactions-i-494.html (primarily oxazolone formation) but it looks like the main suggestion is to add HOBt, which you're already doing.

[BobfromNC]

Esters are easier to racemize.   I would try a faster coupling agent, like HATU, and then run the reaction as short a time as possible, then work it up and get it away from the base as fast a possible.  Letting it stir overnight with any base is a risk, most amide couplings are done in a few minutes if done right, so leaving it to stir overnight will lead to racemization, even a slow process can go a long way overnight.