Chemistry 324 - Spring 2008

Today’s paper contains a number of interesting issues. Rather than take several days to create one big long post, I am going to slice my ideas into small chunks and invite your comments.

Here’s one issue: diastereoselectivity in the reactions of 8 with dienes (Table I). Since this reaction involves three distinct steps, there are multiple points at which to stop and ponder selectivity.

• Conversion of 8 (diazo ester) into a metal carbenoid
• Cycloaddition of carbenoid and diene making a divinylcyclopropane
• Tautomerization of divinylcyclopropane to benzofuran

Step 1: 8 does not contain any chiral atoms. The carbenoid does (the ligand is asymmetric), but there is no selectivity that we are aware of (or can monitor).

Step 2: The cycloaddition can give a mixture of diastereomers: cis and trans divinylcyclopropanes might result from two different orientations of the diene relative to 8 (see Scheme 4). By the way, you should see an analogy between formation of the cis product and endo selectivity that is observed in Diels-Alder cycloadditions. In both cases, there is what Alder called, “an accumulation of unsaturation,” that is, an unsaturated group on one reactant lines up opposite an unsaturated group on the other reactant. We discussed possible reasons for the orientation drawn in Scheme 4 in class.

Step 3: The most puzzling selectivity is seen in the tautomerization reaction (18 → 13). My first thought was this must be a base (possibly acid) catalyzed reaction and 13 must be the most stable stereoisomer. Although this is the normal way for compounds to tautomerize, it cannot be operating here!

Proof of a “strange” mechanism lies in the selective formation of compounds 13 (cis), 14 (trans), and 15 (S).

Suppose a base-catalyzed process was responsible for the tautomerization leading to 13 and 14. If this occurred, there would be a single intermediate (see below), and this would produce the same product or mixture of products. In fact, 13 and 14 cannot be in equilibrium at all because 13 is obtained from trans-piperylene and 14 from cis-piperylene. The tautomerization must accomplish a formal suprafacial [1,3]-H migration. This is a forbidden reaction (Huckel, 4e).

The case of 15 provides what might be an even greater mental shock. 15 contains only one chiral atom, so the tautomerization is enantioselective, not diastereoselective. Again, the tautomerization must occur by way of a formal suprafacial [1,3]-H migration.

Since 15 and its enantiomer are produced in ~ 95:5 ratio, the reaction pathway cannot involve achiral intermediates or racemic precursors. Therefore, I have drawn the precursor as a single enantiomer. The rationale for making a single enantiomer of 15’s precursor lies back with the stereochemistry of divinylcyclopropane formation. The chiral Rh catalyst produces one enantiomer of the cis divinylcyclopropane in excess and Cope rearrangement is necessarily selective because it must occur exclusively via a single boat transition state. This is not enough to guarantee selective formation of 15, however, unless the subsequent tautomerization is also stereoselective, i.e., H must migrate suprafacially.

This raises several tough questions:

• How does tautomerization occur without involving an acid or base?
• Why is [1,3]-H migration stereoselective (and enantioselective in the case of 15)?
• Why is the migration suprafacial, i.e., why does a forbidden reaction occur?

All of these questions point to a catalyzed stepwise process involving a chiral catalyst. Where can we find a chiral catalyst? The Rh(R-DOSP)₄ complex, of course. It turns out that transition metals routinely catalyze 1,3-H migrations (also called alkene migrations or alkene isomerizations) by reversibly inserting in allylic CH bonds. The insertion and its reverse occur only when the metal interacts with the pi electrons of the neighboring alkene, so the metal must necessarily deliver H suprafacially to its new location. (Stop by my office and I will draw you a picture.)