Back to Table of Contents

40904-GB1
New Applications of Cobalt-Alkyne Complexes in Organic Synthesis

Kevin M. Shea, Smith College

The final report for this PRF award will describe the progress we have made on two projects outlined in earlier reports and one project that we initiated over the past year.  The theme of our PRF-funded research continues to be the development of new synthetic applications of cobalt-alkyne complexes in Diels-Alder reactions.  Specifically, we continued our investigations into the synthesis and reactivity of dienes adjacent to cobalt-complexed alkynes, and we developed cationic Diels-Alder dienes stabilized by cobalt-alkyne complexes.

Earlier reports described our strategies for the synthesis of dienes 1 and 2, which we hoped would be highly reactive in Diels-Alder reactions.  As previously summarized, diene 1 could be prepared quickly and efficiently; however, it was unreactive with nearly all dienophiles.  We predicted that diene 2 would have a better reactivity profile, but we were unable to successfully achieve a useful synthesis.

Consequently, we turned our focus to the synthesis of cyclic dienes adjacent to cobalt-complexed alkynes.  We predicted that these would be easier to prepare and more reactive in Diels-Alder reactions.  We synthesized diene 5 in 2 steps from 5-bromofuraldehyde (3), and we could easily cobalt-complex the alkyne in 5 to yield diene 7.

Diels-Alder reactions of dienes 5 and 7 with the highly reactive dienophile PTAD (6) supported our hypothesis that cobalt-alkyne complexes adjacent to dienes in the Diels-Alder reaction will enhance their reactivity.  Uncomplexed diene 5 is unreactive with PTAD at -78 °C in the presence of boron trifluoride.  However, cobalt-complexed diene 7 reacts completely in 5 minutes under identical conditions.  Isolation of cycloadduct 8 proved problematic under the reaction conditions; we believe that this cyclic ether undergoes undesired fragmentation or rearrangement process in the presence of boron trifluoride.  Nevertheless, these results clearly demonstrate that a cobalt-alkyne adjacent to a diene has a dramatic rate enhancement effect in the Diels-Alder reaction.  Studies are currently underway to investigate similar reactions of substituted cyclohexadienes that should yield Diels-Alder products that are more easily isolable. 

The second project that has been previously described in PRF annual reports focuses on a tandem intramolecular Diels-Alder/Pauson-Khand strategy for the synthesis of the tetracyclic steroid skeleton.  Unfortunately, we have been unable to investigate the key Diels-Alder/Pauson-Khand steps because of the failure of one crucial coupling reaction (9 + 10 → 11).

We explored numerous options for the construction of this critical carbon-carbon bond.  In addition to the route shown here, we initially attempted to couple the allylic acetate analog of aldehyde 9 with Grignard 10 in the presence of a copper catalyst.  The second generation approach pictured above focused on the direct addition of organometallic 10 to the aldehyde of 9.  However, all of our attempts proved fruitless.  No improvement was observed when varying the nucleophiles from a Grignard reagent, to an organozinc compound, or to an organocerium reagent.  We plan to address this problem by synthesizing the Weinreb amide analog of aldehyde 9 in hopes that it will react as desired with organometallic reagents 10.

We have recently initiated an investigation into the behavior of cationic Diels-Alder dienophiles stabilized by adjacent cobalt-complexed alkynes.  Gassman demonstrated that alkenes conjugated to oxonium ions serve as excellent dienophiles, and we hoped to achieve similar results with alkenes conjugated to cobalt-complexed alkyne-stabilized carbocations (i.e., 12).

We planned to generate the requisite carbocation by the ionization of a cyclic ether which could reform the ether after participating in the Diels-Alder reaction.  After a thorough literature search for known molecules with the desired enyne ether functionality we desired, we identified cyclic ether 13 as our dienophile precursor.  We predicted that 13 would react with Lewis and/or protic acids to provide cationic dienophile 14.

Following a synthetic procedure from Trost's lab, we converted cinnamoyl chloride (15) into dienyne 19 in three steps.

Alkyne 19 was easily converted into cobalt-alkyne complex 13 upon exposure to dicobalt octacarbonyl.  Treatment of 13 with boron trifluoride generates cationic dienophile 20 that combines with pyrrole in a Diels-Alder reaction to provide cycloadduct 21.  We are currently investigating the stereochemical outcome of this reaction.

Excited by this result, investigations are underway with numerous other dienes as well as attempts to make a variety of similar dienophiles.  One interesting application, shown below, involves a tandem combination of the Diels-Alder and Pauson-Khand reactions.  For example, reaction of 1,3,6-heptatriene with cobalt-complexed alkyne 22 should yield tetracycle 23 after a tandem Diels-Alder/Pauson-Khand process.  We hope to explore this reaction in the near future.

Back to top