[ashmus]

So, I'm synthesizing phosphoramidates and came across something I can't rationalize. Scheme 1 (see image) shows the removal of a 2-nitrobenzyl group in an anhydrous polar aprotic solvent to form the desired phosphoramidate product. If I however do this same reaction in a polar protic solvent, I find I exclusively make phosphate esters. After doing a literature search, I came across a paper (Tetrahedron Letters 1997, 38, 5285-5288) that gives an example of doing a similar reaction in a triethylammonium acetate buffer (essentially water) in the presence of dithiothrietol to make the unprotected phosphoramidate.

Now my question is, how does this additive (dithiothrietol) prevent the formation of the phosphate ester? If possible, could you help explain it 'mechanistically'? I also included the general mechanistic scheme for photolytic cleavage of the 2-nitrobenzyl group.

[Babcock_Hall]

What is the source of H⁺ in your reaction scheme?

[ashmus]

There is no acid in the reactions. If you're referring to the mechanistic scheme that shows loss of proton, I copied the scheme from a book chapter...it's more likely that there's an intermolecular hydrogen transfer with another molecule of itself or the phosphoramide leaves with a negative charge stabilized through resonance, which can then pick up the proton found on the formed nitroso group.

[ashmus]

For those viewing, would it help at all if I show the step-by-step mechanism that I came up with instead of the general mechanistic scheme?

[Babcock_Hall]

This is just a WAG.  Free phosphoramidates are labile in acid, and if the solution became acidic, I could see more solvolysis happening.  That might explain one reason for the presence of triethylammonium acetate (it is buffering).  Phosphoramidate esters can be hydrolyzed to phosphate esters in formic acid: Zamecnik, 1988  http://www.pnas.org/content/85/19/7079.full.pdf

Another way I might envision activating a phosphoramidate toward solvolysis is by oxidizing the nitrogen (I don't have a literature precent for this, although I seem to recall activation of phosphorothioates by oxidation).  Dithiothreitol is a good reductant (E°' is about -0.3 volts).  I can also imagine DTT scavenging electrophiles.

[ashmus]

I'm a bit skeptical of the acidity playing a role here as I have experience removing protecting groups in the presence of the unprotected phosphoramidate using acidic conditions, but I'll look into it further since it involves different conditions and substrates. The nitrogen oxidation is an interesting concept as I do recall reading a paper that reports a radical byproduct forming instead of the 2-nitrosobenzylaldehyde and use of dithioerythritol (DTE) as a radical scavenger (wish I could remember what paper/journal). I assumed it was to prevent unwanted reactions with the protein in their experiment, but perhaps it was also to promote the phosphoramidate formation.

Anyway, thanks for your input Babcock_Hall...your input was definitely helpful.

[Dan]

Could DTT be reducing N-O bonds?

For example, if the endocyclic N-O bond in the aminal intermediate was broken, I would expect faster fragmentation to the phosphoramidite via the hemiaminal - i.e. it might speed up what is possibly the RDS.