Synthesis of 1,3-Enynes by Iron-Catalyzed Propargylic C–H Functionalization: An Alkyne Analogue for the Eschenmoser Methenylation

A two-step protocol for the conversion of alkyl-substituted alkynes to 1,3-enynes is reported. In this α-methenylation process, an iron-catalyzed propargylic C–H functionalization delivers tetramethylpiperidine-derived homopropargylic amines which undergo facile Cope elimination upon N-oxidation to afford the enyne products. A range of aryl alkyl and dialkyl acetylenes were found to be suitable substrates for this process, which constitutes an alkyne analogue for the Eschenmoser methenylation of carbonyl derivatives. In addition, a new bench-stable precatalyst for iron-catalyzed propargylic C–H functionalization is reported.

C onjugated enynes are prominent and fundamental substructures found in a range of natural products, pharmaceuticals, and other bioactive molecules (Figure 1). 1 Additionally, 1,3-enynes serve as versatile building blocks for the synthesis of polysubstituted aromatic compounds, conjugated dienes, and chiral allenes. 2 A number of general approaches convert difunctional starting materials to 1,3-enynes through standard functional group interconversions.These approaches include dehydration of propargylic alcohols, 3 Wittig olefination of conjugated ynones, and Corey−Fuchs alkynylation of conjugated enals. 4part from these transformations, transition metal catalysis has also enabled the assembly of 1,3-enynes through C−C bond forming coupling reactions.The most common method to access 1,3-enynes is transition metal catalyzed crosscoupling reactions between alkynyl and alkenyl precursors, with Pd-catalyzed transformations being the most versatile and well-developed (Scheme 1A). 5 The Sonogashira reaction between terminal alkynes and alkenyl (pseudo)halides is the most widely used among them, due to the convenience and availability of the starting materials. 6Palladium catalysis has also been successfully applied to the synthesis of 1,3-enynes by the three-component coupling of aryl iodides, internal alkynes, and alkynylsilanes, 7 by Heck-type coupling of vinylarenes and alkynyl bromides, 8 and by the oxidative coupling of terminal alkynes with alkenes or vinylmetals. 9As Earth-abundant alternatives to palladium catalysis, iron-10 or copper-11 based catalysts have also been successfully employed in cross-coupling reactions for the synthesis of 1,3-enynes.Additionally, Cu or Cu/Fe-mixed catalysts have also been used in the coupling of vinylmetal species with alkynyl halides. 12Transition metal catalyzed dimerization 13 or trimerization 14 of alkynes can also be highly effective approaches for 1,3-enyne synthesis (Scheme 1B).In some special cases, the coupling of two distinct alkynes through Pdor Co-catalysis has also been achieved. 15The transition metal catalyzed coupling of terminal alkynes with carbene precursors has also proved to be a highly general approach for the synthesis of enynes (Scheme 1C). 16hese approaches generally involve the coupling of two building blocks, which are often prefunctionalized, or they require the manipulation of di-or trifunctionalized starting materials.The direct installation of a methylene group at the α-position of an aryl alkyl or dialkyl acetylene would constitute a more direct approach for the synthesis of 4substituted or 2,4-disubstituted 1,3-enynes.Such a process would resemble the celebrated Eschenmosher methenylation of carbonyl derivatives for the synthesis of conjugated enones. 17Due to the relatively high acidity of the C−H bonds α to a carbonyl group (pK a ≈ 20 to 25), selective deprotonation to form an enolate is possible using various lithium amide reagents.Subsequent aminomethylation with an iminium reagent (Eschenmoser salt) would afford a Mannich base, which could then be converted to the corresponding α-methylene carbonyl compound by Cope or Hofmann elimination (Scheme 1D).We wondered whether a similar process could be developed for the more weakly acidic (pK a ≈ 35 to 40) propargylic position of alkynes.
Previously our group developed a catalytic method for the C−H functionalization of unsaturated hydrocarbons by employing cationic iron complexes for π-activation to increase the acidity of the propargylic, allylic, or allenic C−H bonds and enable their deprotonation by weak amine or pyridine bases to generate nucleophilic organoiron species.These organometallic nucleophiles undergo subsequent functionalization with carbonyl and iminium electrophiles to generate α-C−H functionalization products. 18We hypothesized that allenyliron intermediates generated from alkyne substrates could react with an Eschenmosher salt to give homopropargylic amine products.These adducts could then undergo elimination of the pendent dialkylamino group to afford 1,3enynes, allowing for the development of a protocol that is formally and mechanistically analogous to the Eschenmoser methenylation (Scheme 1E).In this Communication, we report the successful development of this process through the use of an in situ generated iminium electrophile to give homopropargylic 2,2,6,6-tetramethylpiperidine derivatives.Upon N-oxidation, spontaneous Cope elimination occurred to deliver the desired 1,3-enyne products.
We employed a hydride abstraction strategy for in situ formation of the requisite iminium intermediate. 19A mixture of tritylium tetrafluoroborate (Ph 3 C + BF 4 − ) and 1,2,2,6,6pentamethylpiperidine (2) was first stirred at rt for 1 h in toluene to generate iminium salt 2′.18a,b,e Using [Cp*Fe- ) as the catalyst and applying previously reported conditions, desired aminomethylation product 3a was observed in 10% yield by NMR analysis (Table 1, entry 1).While switching the solvent to trifluorotoluene was found to be beneficial (entry 2), the addition of BF 3 •Et 2 O as an additive resulted in a dramatic (and unexpected) improvement of the yield to 80% (entry 3).Other Lewis acids were, therefore, explored as additives.The use of a substoichiometric amount of zinc bistriflimide (Zn(NTf 2 ) 2 ) in place of BF 3 •Et 2 O was found to further enhance the yield (entries 5−6), while use of zinc triflate gave a poor outcome, and other metal bistriflimide salts were less effective (entries 8−12). 20Finally, the reaction conditions were further optimized by additional adjustments of Lewis acid and base stoichiometries (entry 7), delivering amine 3a in 86% isolated yield.Although 3.0 and 4.0 equiv of TMPH gave identical yields for model substrate 1a, 4.0 equiv was found to be more general for more challenging substrates and was therefore chosen as the standard condition for subsequent investigations of the substrate scope.experiments showed the necessity of both base and catalyst in this reaction (see Supporting Information for details).
To form the enyne, purified samples of 3a were then subjected to oxidation with 3-chloroperoxybenzoic acid (m-CPBA) to generate the N-oxide.Pleasingly, the amine was found to undergo oxidation and subsequent Cope elimination under very mild conditions (0 °C, THF, 1 h) to deliver 4a in a 78% isolated yield over two steps (Scheme 2).Little (<5%) to no overoxidation to the epoxide was observed under these optimized conditions.It was subsequently found that after filtration through SiO 2 gel to remove excess base and iron residue, crude products 3 could be subjected directly to oxidation and Cope elimination to minimize material loss and improve the overall yield of the 1,3-enyne.
During the course of mechanistic investigations (see the Supporting Information for the results of some preliminary studies on the role of the secondary Lewis acid), we explored some alternative strategies for accessing the cationic iron species.For instance, we considered the one-electron oxidation of [Fp*] 2 dimer 10b with AgBF 4 . 21Using this approach, the desired reactivity was observed, giving a 53% NMR yield of 3a under standard reaction conditions (Scheme 4A).This encouraging result prompted us to explore other potentially bench-stable catalyst precursors for this transformation.While tetrahydrofuran complex 10a demonstrates excellent catalytic activity, its water sensitivity and lability under dynamic vacuum complicate its synthesis and isolation.We turned to pyridine complexes of [Fp*] + as alternatives, given their tunability and potentially improved robustness.While the unsubstituted pyridine complex 10c was inactive, 2,6-difluoropyridine complex 10d was found to exhibit reactivity similar to that of 10a for three representative substrates (Scheme 4B).Moreover, catalyst samples that were stored on the benchtop for more than one month retained their full catalytic activity and did not show signs of degradation, either visually or by 1 H NMR analysis.As such, complex 10d may serve as a more user-friendly catalyst for accessing [Fp*] + for the current protocol as well as other synthetic applications.
In conclusion, we have developed a method for the methenylation of propargylic C−H bonds using inexpensive and readily prepared cyclopentadienyliron(II) dicarbonyl complexes as catalysts.Further investigations toward the formation of strategic C−C bonds using this approach are ongoing and will be reported in due course.

■ ASSOCIATED CONTENT
Experimental procedures and spectroscopic data for the substrates and products (PDF)

Table 1 .
Control Optimization of the Fe-Catalyzed α-C−H Functionalization Step a