[bobsacamano]

I am interested in doing a Wittig with MePPh3Br and an aldehyde that is the side chain of an amino acid.  What base would be best to use that would not cause racemization of the amino acid?  Is there a base other than BuLi or KHMDS that I could use for deprotonation?

[lavoisier]

When I did Wittig reactions, I used to generate the ylide before adding the electrophile.
If you plan to do the same and if you don't use excess base, the base you actually have in solution when you add your aldehyde is the ylide itself, a weaker base than the original BuLi.

And by all means, any other base you might chose would need to be at least as strong as H2C=PPh3, and you're back to square one...

As in many other cases, I think you must rely on chemoselectivity, i.e. you hope your ylide will behave more as a nucleophile than as a base. Aldehydes are very good electrophiles, you shouldn't have particular problems.

Btw, how did you protect the amino-acid moiety, if I may ask you?

[bobsacamano]

Thanks for the advice, I was thinking that also.  Just was wondering if anyone had success running a Wittig with a weaker base.  The amino acid that I am using is a Di-Boc-Guanidine with a trimethylsilylethyl ester at the alpha carboxylic acid and I have already converted the gamma to the aldehyde.

[bobsacamano]

Sorry, doing too many things at once.  Di-Boc-Glutamic acid trimethylsilylethyl ester is what I meant to say there.  Not Di-Boc-Guanidine.

[lavoisier]

I  may be wrong, but I guess you can use weaker bases only if your ylide is particularly stable.
There are even Wittig reactions done in aqueous NaOH (see for instance the synthesis of certain cephalosporins with a CH=CH2 chain in 3), but there you really need very acidic phosphonium salts and a non-base-sensitive aldehyde.

In your case I'm afraid you will need to generate the ylide stoichiometrically at very low temperature.

But again, I may be wrong and there may be Wittig reactions with unstabilised ylides where only a catalytic amount of weak base is needed.

[bobsacamano]

I found an example of a Wittig done on an amino acid aldehyde very similar to mine.  No racemization is reported.  They used KHMDS as their base and they actually used an excess of the MePPh3Br over their base when generating the ylide.  Rxn was run at r.t.  Only problem I see is the poor yield, they report only 40%.

[lavoisier]

I'm surprised that they do it at RT, it's interesting. I wonder if they do so because the reaction is too slow at low T, or if there's some unexplained reason...
The yield is low indeed, but you never know, your compound may give a better result.
Maybe you could let us know what you find.

[mkaplan4]

BuLi is a good choice for the deprotonation to form the ylide, however I like to use Sodium Hydride, b/c the only reaction by product is Hydrogen gas which leaves the system.