Selective Carbon-Carbon bond Activation: A Wellspring of Untapped Reactivity

The distant and selective activation of unreactive C−H and C−C bonds is one of important challenges in organic chemistry. In recent years, the development of remote functionalization has received a growing interest as it allows for the activation of rather challenging C−H and C−C bonds distant from the initiation point by means of a “metal-walk. A “metal-walk” or “chain-walk” is defined by an iterative series of consecutive 1,2- or 1,3-hydride shifts of a metal complex along a single hydrocarbon chain. With this approach, simple building blocks or mixtures thereof can be transformed into complex scaffolds in a convergent and unified strategy. Synthetic organic strategies that enable the catalytic and rapid assembly of a large array of organic compounds that possess multiple stereocenters in acyclic systems are somewhat rare, especially when it comes to reaching today’s high standards of efficiency and selectivity. In particular, the catalytic preparation of a three-dimensional molecular layout of a simple acyclic hydrocarbon skeleton that possesses several stereocenters from simple and readily available reagents still represents a vastly uncharted domain. In this research work, a rapid, modular, stereodivergent and diversity-oriented unified strategy to construct acyclic molecular frameworks that bear up to four contiguous and congested stereogenic elements, with remarkably high levels of stereocontrol and in only few catalytic steps from commercially available starting materials. One key-elements of all the strategies investigated relied on selective C–C bond cleavage of a cyclopropane, easily prepared by catalytic enantioselective carbometallation of cyclopropenes.

Our research project entitled “Selective Carbon-Carbon Bond Activation: A Wellspring of Untapped Reactivity” under the acronym of “CMetC” was supported by the European Research Council from November 1st, 2013 to October 30, 2018.

After a short period of initiation, we started our research program intensively as planned in our proposal. We initially could show that any w-ene cyclopropane species possessing a remote double bond or alkylidenecyclopropanes could easily be transformed into acyclic molecular fragments possessing two stereogenic centers including challenging quaternary carbon stereocenter. This process goes through a unique merging of zirconocene-mediated allylic C-H activations with highly selective carbon-carbon bond activations/fragmentations (Angew. Chem. Int. Ed. 2015, 54, 414). Due to the presence of the quaternary center, the resulting bifunctional nucleophilic species are further derivatized with two different electrophiles to give more complex molecular architectures arising from explicitly simple starting materials (Nature 2014, 505, 199). Particularly interesting is the high 1,4-diastereocontrol in the functionalization of w-ene cyclopropane species. As highly enantiomerically enriched cyclopropane derivatives are easily accessible (Chem. Sci. 2018, 9, 6503 and Chem. Rev. 2018, 118, 8415), this reaction represents an important entry to the formation of enantiomerically enriched quaternary and tertiary carbon stereocenters in acyclic systems and give important scientific basis for remote functionalization of alkenes (Chem. Sci. 2015, 6, 2770). Computational studies revealed the original way to achieve this high stereocontrol by having a plethora of equilibria feeding a preferred reactive channel leading to the major isomer through dynamic thermodynamic control (Nature Protocols, 2016, 12, 74). We were then interested to extend this approach to a special case of three-membered ring namely the w-ene alkenyl spiro[2.2]pentanes. These smallest encountered spirocarbocycles have a higher ring strain energy (63 kcal/mol) than two cyclopropanes (27.5 kcal/mol) suggesting an even higher propensity towards fragmentation. Following this initial assessments, w-ene alkenyl spiro[2.2]pentanes could undergo selective ring-opening in good to excellent diastereoisomeric ratios . When these starting materials were treated with the Negishi reagent, we were gratified to observe a rapid and clean isomerization reaction followed by a selective single C-C bond cleavage to provide, after reaction with acetone and subsequent hydrolysis, a single E-isomer without any trace of second C-C bond cleavage product (J. Org. Chem. 2018, 83, 3947). It is interesting to note that the diastereospecificity of the reaction is quantitative (ds > 98:2). The presence of the remaining sp3 carbon-zirconium could then easily be functionalized (Org. Biomol Chem. 2016, 14, 10325). This tandem sequence was applied to various w-ene alkenyl spiro[2.2]pentanes and as previously discussed, longer alkyl chain between the strained ring –namely here the spiro[2.2]pentane unit- and the unsaturation did not seem to alter the transformation. Yields resulting from the isomerization followed by the selective mono- C-C bond cleavage with n = 3, 4 and 5 respectively were indeed comparable. Following the development of this straightforward approach for a selective mono- C-C bond cleavage, we then turned our attention to promote the second C-C bond cleavage. As already mentioned, the reaction proceeded invariably of the chain length and affords the doubly ring-opened products after reaction with ketone. Beside the isomerization, this study provides a unique, flexible and selective multiple C-C bond cleavage of w-ene spiro[2.2]pentanes leading either to a selective single or double ring-opening selectively. This transformation could be extended to remote double bonds (n = 5) thanks to the facile migration of the zirconacyclopropane to easily walk along the hydrocarbon chain. Based on these initial achievements, we then extended our concept to the catalytic C-C bond cleavage (Nature Chemistry, 2016, 8, 2019). In this context, we have reported that a single transition metal based-catalytic system successively mediated a (i) regioselective arylation of the terminal olefin, (ii) a chain-walking, (iii) a selective C–C bond cleavage of the cyclopropane, (iv) additional chain-walking and (v) release the carbonyl moiety to regenerate the active catalytic species. All these consecutive elementary steps proceeded in a one-pot operation with impressive levels of selectivities. During this study, we could uncover the highly selective C–C bond cleavage conditioned by the chain-walking directionality. As a result, the strategically positioning of the cyclopropane ‘vault’ enabled the consecutive introduction of a tertiary and a quaternary stereocenter at unrelated positions between two distant functional groups into a linear hydrocarbon skeleton (Nat. Commun. 2017, 8, 14200). This contrasts with catalytic chain-walking reactions that are generally perceived as proceeding exclusively across methylene chains. Additionally, the ease of synthesis of such diastereomerically pure w-ene cyclopropyl derivatives associated with this proposed versatile and mild procedure allowed the downstream functionalization into further sophisticated acyclic hydrocarbon frameworks (Nat. Rev. Chem. 2017, 1, 0035). We subsequently identified the unique ability of a Ru-based “alkene zipper” catalyst to efficiently trigger the remote ring-opening of alkenyl cyclopropyl carbinol through a long-range alkene isomerization combined with a retro-ene reaction (Adv. Synth. and Catal. 2018, 360, 1389). The ring-opening event remarkably turned out to selectively yield linear unconjugated ene-aldehydes bearing quaternary stereocenter. This methodology provided an operationally simple access to functionalized acyclic polyfunctional molecular fragments (possessing an aldehyde, a quaternary stereocenter and an adjacent E-double bond) using a commercially available complex. Mechanistic studies, involving kinetic and isotopic labelling experiments provided several critical insights prompting us to suggest a mechanism. Finally, in this particular field of remote functionalization (ACS Central Science, 2018, 46, 1659), we designed a strategy that entails a succession of highly stereoselective carbon–carbon bond-forming reactions catalyzed by a transition metal complex. As such, complicated and otherwise difficult- to-access acyclic fragments have been prepared from simple and readily available reagents. We have illustrated that various acyclic hydrocarbon motifs could be synthesized in a stepwise manner by the following chemo- and stereoselective C–C forming bond processes: Rh-catalysed [2+ 1] cycloaddition, Cu-catalysed carbomagnesiation/ Pd-catalysed cross-coupling followed by ester reduction and regio and diastereoselective Pd-catalysed oxidative Heck migratory insertion to trigger the selective unfolding of a cyclopropane intermediate. Notably, products that bear adjacent quaternary carbon stereogenic centers were synthesized from readily accessible terminal alkynes through just three catalytic reactions. The stereodivergence of this new approach was highlighted through the preparation of four different stereoisomers of compounds that represent the same constitutional isomer (Nat. Chem. 2018, 10, 1164). The investigations described above demonstrate that, by the judicious implementation of distinct transformations carried out in a particular sequence, synthesis routes may be conceived that are of considerable utility and render streamlined access to previously unknown stereoisomeric motifs a realistic possibility (Chem. Eur. J. 2018, 24, 8553). One missing part for remote functionalization was to transform w-ene alkyl ether into vinyl metal species; this could be achieved by combining the Ru-metal walk with the Ni-catalyzed cross-coupling reaction of enol ethers (Angew. Chem. Int. Ed. 2018, 57, 8012). Having in hands, cyclopropyl methanol, we could also develop a novel approach for the preparation of acyclic hydrocarbons bearing one quaternary and one tertiary stereocenter in a 1,4-relationship. The method was based on the diastereoselective biomimetic ring-opening of cyclopropane methanol derivatives by using ambiphilic organoalane nucleophiles. The strategy was then applied to the total synthesis of botryococcene and epibotryococcene (Angew. Chem. Int. Ed. 2018, 57, 13237). In a related part of this project, we were interested to extend this concept to the ring-opening of cyclobutane. However, to have a powerful approach in organic synthesis, one first needs to develop an easy and efficient access to the starting materials. We have therefore reported the diastereoselective carbometalation of cyclobutenes en route to diatsreoisomerically pure polysubstituted cyclobutanes (Chemical Sciences 2017, 8, 334). In the second part of the proposal, we were interested to develop the diastereoselective and enantioselective carbometalation reactions of alkenes (Chem. Soc. Rev. 2016, 45, 4552), eventually combined with selective ring cleavage. As regards to the enantioselective carbometalation reaction, we were delighted to develop a new and successful general approach to diastereoisomerically pure and enantiomerically enriched cyclopropanes through catalytic asymmetric carbometalation of cyclopropenes (Angew. Chem. Int. Ed. 2018, 57, 3682 and Org. Lett. 2017, 19, 3970). Indeed, as we were more and more interested in the exploitation of the multifold reactivity of organometallic species to merge allylic C-H and C-C bond activations of pure hydrocarbons, we soon realized that the diastereo- and enantioselective preparation of cyclopropanes possessing only hydrocarbon chains (therefore composed of C and H) required tedious manipulations of any functionalized cyclopropanes. It became clear that a catalytic method leading to the formation of variously enantioenriched substituted cyclopropanes composed of only hydrocarbon chains as single diastereo- and enantiomer was in its complete infancy and therefore highly desirable. In this context, the copper-catalyzed enantioselective carbozincation and carbomagnesiation reactions of cyclopropenes has been initially developed very successfully to afford highly enantioenriched configurationally stable cyclopropylzinc reagents (J. Am. Chem. Soc. 2015, 137, 15414 and Angew. Chem. Int. Ed. 2017, 56, 6783). This carbometalation reaction of cyclopropanes could be used either as a starting point for oxidation (Angew. Chem. Int. Ed. 2018, 57, 1543) but also for reaction with acylsilanes to promote a Brook rearrangement and selective C-C bond activation (Synthesis, 2016,48, 3279 and Angew. Chem. Int. Ed. 2016, 55, 714) or for a zinc homologation and selective ring opening reaction (Chemical Sciences, 2016, 7, 5989). This new sequence of diastereoselective carbometalation – zinc-homologation – C-C bond cleavage allows the easy transformation of enantiomerically enriched cyclopropenyl esters into acyclic allylic moiety bearing the challenging quaternary carbon stereocenters in a single-pot reaction through the formation of two new C-C bonds. As the carbometalation reaction may lead to two different diastereoisomers according to the nature of the solvent, this strategy paved the way to the diastereodivergent synthesis of both enantioenriched final products at will. The zinc homologation as well as reaction of acylsilane triggered the development of a new approach to enantiomerically enriched allylzinc species, unreported to date. When enantiomerically enriched a-hydroxy allylsilanes are treated with Et2Zn, an allyl Zn-Brook rearrangement occurs to give configurationally stable allylzinc intermediates that react with acyl chloride and methyl chloroformate with a complete transfer of chirality. Experimental data as well as quantum mechanical calculations confirm that the product is obtained through the formation of a chiral allylzinc intermediate bypassing the classical [1,2]-Brook rearrangement (Angew. Chem. Int. Ed. 2016, 55, 6057). Subtle changes in the experimental conditions or in the structure of the starting cyclopropanes delivered either polysubstituted cyclobutenes (J. Am. Chem. Soc. 2017, 139, 8364) or acyclic systems (Chem. Eur. J. 2018, In press). Finally, the selective oxidation reaction for the preparation of reactive intermediates has been also investigated in this period of research particularly as a new source of stereodefined polysubstituted enolate species (Chem. Commun. 2014, 50, 12597). One clear evidence of the importance of a carbonyl group in organic synthesis is that it represents one of the most versatile and broadly utilized functional groups, particularly for the preparation of enolate in carbon-carbon and carbon-heteroatom bonds formation. In view of the tremendous utility of such enolates in synthesis, application of this chemistry to the enantioselective synthesis of quaternary carbon stereocenters from stereochemically defined acyclic trisubstituted enolates of ketones was highly valuable. However, the main problem that limited the formation of such quaternary stereocenters, still considered as a significant challenge in organic synthesis, was the paucity of methods for the preparation of acyclic trisubstituted enolates of ketones and amides as single isomer (Angew. Chem. Int. Ed. 2015, 54, 14933). In this research, we could describe a general approach leading to the formation of several new stereogenic centers—including the quaternary one—via either a combined metalation–addition of a carbonyl–carbamoyl transfer to in-situ reveal stereodefined trisubstituted enolates of ketone as a single isomer (Angew. Chem. Int. Ed. 2016, 55, 5517) or via a combined carbometalation reaction oxidation and silylation as a new entry to stereodefined fully substituted silyl enol ether used in the Mukaiyama reaction (Angew. Chem. Int. Ed. 2015, 54, 14933) or via a Thischenko reduction (Tetrahedron 2018, 74, 6761). We could then use these new approaches to generate a series of aldol and Mannich products from enol carbamate or amide enolates with excellent diastereomeric ratios and in a single pot operation. The resulting stereodefined enolate could react with various electrophilic reagents such as alkylating agent (Chemical Sciences, 2017, 8, 627), electrophilic fluorinating agent (Org. Biomol. Chem. 2018, 16, 1079) or electrophilic oxidant (Eur. J. Org. Chem. 2018, 614)

Overall, we could demonstrate the feasibility of new synthetic approaches that exploit multifold reactivity of a single substrate so as to furnish advanced molecular scaffolds through a unique merging of otherwise challenging transformations. Even more exciting is the number of new perspectives that this work brought to light and to finally show that a completely different approach, opposite to the more classical mainstream thought, could solve synthetic problems and lead to new approaches to solve synthetic problems.