[brettlawrie]

Can someone briefly explain to me why the cis isomer in the provided attachments is not the favoured product in that specific Wittig mechanism?

I have a feeling it has something to do with steric hindrance, delocalization and stability but I am quite confused when it comes to those topics. I understand that having all those big phenyl groups close together on the cis (A), cis (B), and even trans (B) diagrams in the "Wittig Intermediate" attachment can't be good...but I can't really formulate an answer (as to why cis can't be produced) that I am confident in. Thank you for your help.

[discodermolide]

I don't think the mechanism is concerted, it is stepwise allowing rotation before the 4-membered ring transition state is formed thus putting the phenyl groups in a stereochemically favored trans configuration.

[Arctic-Nation]

Unstabilized ylids are supposed to give cis-alkenes exclusively in the Wittig reaction. The idea is that in the formation of the intermediate betaine (if one is actually formed; the literature is conflicted, to say the least), the large substituents will stay as far from each other as possible, but end up right next to each other when the molecule rotates to form the oxaphosphetane.

However, a quick google search has shown that some amount of trans-stilbenes seem to be formed even in the classical Wittig reaction. Couldn't find an explanation, though (don't have easy access to any literature), and I don't think the Wittig-Horner mechanism is applicable here. Some kind of pi-stacking between the two phenyl groups, perhaps?

[orgopete]

In my book, I suggested it was due to a difference in the rates of the addition vs cyclization. If the addition gives the intermediate that cyclizes to the cis product, then the cis product can result. If the cyclization/elimination reaction is slower than the addition, then the addition may revert and the product of the cyclization/elimination will be found.

I show the schemes here, http://orgo.curvedarrow.com/punbb/viewtopic.php?id=379

I concede in reading what I had written some time ago, I don't think my description is satisfactory, except as a device to remember which product may have formed. I need a better analysis should include an asymmetric induction analysis, Cram, Cornforth, etc. The argument should be similar to the inductive vs resonance effects I argued (in this case, addition preference vs product stability), but I don't want to go there!