[AlphaScent]

If there was an alpha,beta unsaturated ketone and a tri substituted epoxide in a molecule. Would you expect methyllithium to only add to the ketone?

[AlphaScent]

I was on my roommates ipad when I sent that.  Below is the chemistry I am talking about.  I would have thought that though a carbonyl is more reactive, the epoxide will not be affected.  I know nucleophilic reactions are slowed towards hindered epoxides.  But no reaction at all?!?

[kriggy]

Why you dont use grignard? Would it open the epoxy ring rather than add to carbonyl?

[AlphaScent]

cant use a grignard. the magnesium will act as a lewis acid and the epoxide will rearrange to the ketone.  methyl lithium has to be used.  I have come to the conclusion, through discussion with my boss, that the methyl lithium will react much faster than any reaction that takes place at the epoxide. to prevent any rxn at the epoxide I will use 0.95 equivalents of methyl lithium. ill post the results next aeek sometime.  I was just wondering if anyone has tangible evidence to my line of thought.

[orgopete]

I agree with the ketone reacting the fastest. Whether the alkoxide intermediate will react with the epoxide, I was less certain. Because it is tertiary, I think not. I'd keep the reaction cold, just in case.

[AlphaScent]

Yes, cold for sure, -78 is standard conditions for this reaction. 

The chemistry comes from a paper for the synthsis of a pheromone.  They make the trisubstituted epoxide last, even though as you can see from the product above, it would be a diolefin.  In theory the trisubstituted olefin will react "first/faster" than the disubstituted olefin would.  But in practice that is not true.  Using 1 eq mCPBA in DCM at 0 deg C, both the monoepoxide (product above) and the diepoxide were made with the starting diene also left after the oxidation.  I think it is a concentration phenomena.

So we are trying to epoxidize first, then carry through to the last reaction of alkylating with methyllithium.  It wont give us the un-wanted diepoxide.

[clarkstill]

I think you could try the Grignard too, I'm not sure the Lewis acid mediated epoxide rearrangement if favourable enough to compete. These guys:

http://onlinelibrary.wiley.com/doi/10.1002/chem.201302575/pdf

SI: http://onlinelibrary.wiley.com/store/10.1002/chem.201302575/asset/supinfo/chem_201302575_sm_miscellaneous_information.pdf?v=1&s=3b9b8f9b306e55725198825df7a8adf0942e2eff

use both vinyl Grignards and vinyllithium compounds in similar systems.

[zsinger]

Alpha,
One side note on epoxides…..I have found that aqueous acid/base workups can lead to undesired products, so make liberal use of protecting groups if needed.  I have had them get attacked by an undesired Nuc-, and kill my whole reaction.
                -Zack

[AlphaScent]

Zack,

This all new to me on a practical level.  I knew about the possibility of rearrangement under Lewis acid conditions, but my boss has told me it is a likely reaction we want to avoid. 

Clark,

I will most likely for our interests do both reactions and compare results.  It would be easier to use the Grignard since it comes as a 3 M solution as opposed to a 1.6 M for methylithium.

Thanks for the references, will read them this afternoon.