[lemonoman]

Okay, here's the deal:

There's a reaction scheme attached to this post (I drew it in Paint! lol...)...it should be at the bottom of the message...

In the third step, I want to carry out an enolate alkylation, to give the methylated product (top).  But I'm worried that there might be a strain-releasing elimination to give the enone (bottom).  That second one seems more likely to me, just because the ring is SO strained...

The carbons attached to the strained oxygen are nucleophilic too, I believe...that might cause problems too...

Any comments?

[hereafter]

In anhydrous system, how can the ring-open reaction happen?

Just do it, I think it is fine.

[lemonoman]

Organic texts tell me that LDA won't break open the ring...so I assume I^- (which is produced in this reaction) can't either...?

[movies]

Diisopropylamine won't be a problem in this reaction.  Your elimination theory could be, however.

I doubt that it would happen though because there is poor orbital overlap between the p-orbital of the enolate carbon and the sigma^* orbital of the carbon-oxygen bond that would fragment.  If this were an acyclic system, however, the presence of an alkoxide leaving group beta to a carbonyl would rapidly lead to the enone with expulsion of alkoxide.