[Babcock_Hall]

We are interested in making an acrylamide from an aldehyde and a diethyl phosphonate bearing a carboxamide in the beta-position.  Our typical Horner Wadsworth Emmons conditions are similar to what Masamune, Roush, and their collaborators did (Blanchette et al., Tetrahedron 1984 25(21) 2183-86), and similar conditions have since been used by many groups.  The typical conditions are LiX and a tertiary amine base.  Esters are tolerated, but I was wondering about the acidity of the hydrogen on the nitrogen atom and whether this might be a problem. In the 1984 paper there was one aldehyde which had a secondary amide elsewhere in the molecule.  And the 1995 paper (see next paragraph) is also suggestive of success.  So maybe a carboxamide is fine.

Although my literature searching was not as extensive as it might have been, I did turn up one paper that used similar chemistry in the solid phase.  The solid support was attached to the nitrogen atom of the phosphonate.  They used a 10-fold excess of aldehyde, and they used P-31 NMR to detect (Johnson CR, Zhang, B, 1995 Tet Lett 36(51) 9253-9256).  We would not work in the solid phase.  Most conditions that we have used for solution HWE reactions have used a slight excess of phosphonate over the aldehyde, although the opposite was true for one reaction. 

[rolnor]

Since a tertiary amine is sufficient as base for the reaction the carboxamide should not interfere I think. But its good that you spot this potential problem. If you make a classic Wittig-reagent with triphenylphosphine this will be even more easy to de-protonate but maybe you have reasons for using a phosphonate like wanting the E, Z-ratio to be in a certain way?

[Babcock_Hall]

We can buy the phosphonate for the Horner reaction and we have experience with it.  I just had not thought about Wittig chemistry (which is my failure of imagination).  We have always aimed for trans-stereochemistry, and for now I would want to keep that constant.

[rolnor]

If you can buy the carboxamide with a bromine in the possition where you want triphenylphosphine its easy to make the Wittig-reagent, just mix them i toluene and reflux for one hour, usually the phosphonium salt crystallizes upon cooling and this can be used directly, just add TEA as base to create the ylide. Even pyridine is probably basic enough to make the ylide.

[Babcock_Hall]

I am continuing my review of the Horner Wadsworth Emmons reaction with β-carboxamidophosphonates.  I have found several other which use variations of the conditions in Blanchette (1984).  The stoichiometry of the phosphonate, the base (often DBU), and the aldehyde is different in the hands of various workers.  Perez, Charles Rubert et al., Chemical Biology & Drug Design, 79(3), 260-269; 2012 used 1:2:0.625 (or 1.6:3.2:1, normalizing to the aldehyde).  Forster, Michael et al., Journal of Medicinal Chemistry, 61(12), 5350-5366; 2018 used 1.5:1.5:1.

[rolnor]

I think it is not so good to use a large excess phosphonate, the product is a Michael-acceptor and the phoshonate could attack the double bond, both the alpha-carbon or the amide-nitrogen could potentially do this. This would be no problem with a Wittig-reagent.

[Babcock_Hall]

I cannot see the point of having a large excess of phosphonate, either.  We were able to produce some acrylamide with 1.2 equivalents of phosphonate, although our yield was unimpressive.  We will repeat the reaction and purification.

[rolnor]

Maybe you can monosilylate the amide-nitrogen and thereby make it less nukleophilic?
Heat the amide with bis-trimethylsilylacetamide first then apply vacuum+heat to remove excess reagent and acetamide formed as by-product from the reagent. The mono-silylated amide can be analyzed by NMR in CDCl3 (not MeOD).

[Babcock_Hall]

Thank you.  Our second synthesis went with even less impressive yield.  I am also looking into possible loss of material in the extraction from the quench with aqueous ammonium chloride.  How would one remove the silyl group?

[rolnor]

The silyl group is removed just by adding water or MeOH to the reaction mixture, its very unstable.

[Babcock_Hall]

I extracted the aqueous layer with ethyl acetate last night, and I obtained a small amount of solid.  My hypothesis that our poor yield could be fixed by a different extraction solvent was not supported.  One other hypothesis is that the ammonium chloride is protonating the nitrogen of the pyridine ring, by mass action.

[rolnor]

Can you monitor the reaction with LC-MS? That would be very good.

[Babcock_Hall]

I don't have access to that capability at the moment.  We are trying a third reaction today with an aliphatic aldehyde, one with a couple of protecting groups on it.  We will see what happens.

[rolnor]

Good Luck!

[Babcock_Hall]

Both Horner Wadsworth Emmons reactions went mainly to the desired product, although I lost a little product due to a handling error in one instance.