The model below shows how one face of the alkene (orange)
is sterically shielded.
(c). This is a tough model to build. Me2CuLi is a dimeric/oligomeric species. Its possible the Li binds to the O of the epoxide,
and possibly the Me2Cu unit to the alkene, but the true origin of the regiospecificity is likely to be more complex.
The Zimmerman-Traxler Transition state requires metal chelation of the two oxygens;
which for diastereomer (a) induces a close steric contact
not present for diastereomer (b)
(f) Diels-Alder reactions are normally endo because of stereoelectronic effects. In this case, using a Molecular Mechanics
model (ie no electronic component to the model), the lowest energy isomer is clearly that observed, which suggests that in this case,
a steric argument alone can explain the specificity.