Transition-metal-catalyzed domino reactions of strained bicyclic alkenes

Introduction

A well-orchestrated sequence of events – cascade, also known as domino, tandem, and sequential reactions, constitutes a fascinating branch of organic chemistry dedicated to the synthesis of highly functionalized products through sequential transformations in a single reaction. Classically, a domino reaction has been defined by Tietze as a reaction involving two or more bond-forming transformations that take place under the same reaction conditions, without adding additional reagents and catalysts, and in which the subsequent reactions result as a consequence of the functionality formed in previous steps [1].

Bicyclic alkenes, a family of strained ring systems, have seen widespread applications in organic synthesis in the last 20 years [2-6]. Broadly speaking, bicyclic alkenes can be classified into two groups: homobicyclic and heterobicyclic alkenes. Homobicyclic alkenes are hydrocarbons, like norbornadiene, while heterobicyclic alkenes contain at least one heteroatom in the bicyclic framework. Typically, reactions involving these strained bicyclic alkenes are thermodynamically driven forward with the release of ring-strain energy (Figure 1) [7,8]. Intuitively, increasing the number of olefin moieties in the bicyclic system from zero, one, and two, increases the ring-strain energy. Moreover, the introduction of a bridging heteroatom increases the ring-strain energy of the system, conceptualized by the decrease in bond distances. Typically, there are two modes for ring-strain release. First, functionalization of the double bond mildly alleviates the ring strain by relieving nonoptimal bond angles enforced by the rigid bicyclic framework. Secondly, through ring opening of the bicyclic framework; the C–X bond of a heterobicyclic alkene is much weaker than the corresponding C–C bond of a homobicyclic alkene, which allows the C–X bond to be readily cleaved over the course of a reaction.

The stereochemically well-defined and rigid nature of these bicyclic alkenes creates two diastereotopic faces, namely the endo and exo face (Figure 2). The exo face is sterically less congested than the endo face; therefore, the exo face will typically interact with metal catalysts through side-on coordination of the olefin, and in the case of heterobicyclic alkenes, the heteroatom. This preferential exo coordination is not always the case, as norbornadiene derivatives are known to preferentially form chelated endo complexes which can change the stereochemical outcome of the reaction. Nevertheless, the predisposition of metal catalysts towards coordination on the exo face biases the reaction outcome towards exo-selective functionalization.

Upon exo coordination of a metal catalyst with the π system and subsequent migratory insertion, the resulting alkyl metal intermediate is quite limited in how it can propagate. In the case of a carbobicyclic system (Figure 2b), the rigidity of the bicyclic framework restricts β-H elimination. The inability to rotate to achieve an optimal synperiplanar geometry restricts efficient elimination (Figure 2b, H_a). Bridgehead protons are in a more favorable geometry for β-H elimination (Figure 2b, H_b); however, their elimination would generate a highly unstable alkene at the bridgehead, violating Bredt’s rule [9]. For these reasons, carbobicyclic alkenes have been exploited as propagation mediators, as seen in Catellani-type reactions [10-12]. In this review, we will focus on the functionalization of the bicyclic framework itself rather than its use as a transient mediator for domino reactions [13-17]; however, we point the reader to several excellent reviews. The “trapped” alkyl metal intermediate can undergo subsequent migratory insertion steps with other π systems or can be intercepted by an electrophile.

The introduction of a bridging heteroatom into the bicyclic scaffold can dramatically alter the reactivity (Figure 2c). Besides the apparent increase in the ring strain (vide supra), their potential propagation steps are more complex. After an exo coordination of a metal catalyst with the π system and migratory insertion, the resulting heterobicyclic alkyl metal intermediate is not as kinetically stable as its carbocyclic counterpart. While β-H elimination is still limited, these heterobicyclic alkenes will often undergo β-heteroatom elimination to generate ring-opened intermediates (Figure 2c). Fortunately, the metal center can be tuned to promote or hamper the β-H elimination, providing two routes for reaction propagation: ring opening and interception of the ring-opened intermediate or functionalization of the alkyl metal intermediate.

Throughout the past decade, research efforts have demonstrated a broad range of strained bicyclic alkenes can be exploited in domino reactions to selectively generate highly functionalized ring systems. Over the years, several different metal catalysts have been used, each allowing for a breadth of unique coupling partners to either propagate the reaction or to terminate the process.

This review presents a comprehensive examination of domino reactions involving strained bicyclic alkenes. Rather than being exhaustive in the range of potential difunctionalization processes covered, the review will be limited to domino reactions which include at least two distinct reactions. The review is divided on the basis of the transition-metal catalyst used in the reaction and will not cover metal-free methods. The literature is covered up to and including January 2023. For reasons of clarity, newly formed bonds are sketched in red, with newly formed cyclic structures being highlighted.

Review

Earth-abundant metals

Among the transition metal used in organic synthesis, the late transition metals like rhodium, palladium, and iridium have taken center stage when it comes to methodology development. Although these late-stage transition metals have contributed immensely to synthetic organic and organometallic chemistry, increasing societal awareness in terms of sustainable developments and resource management has prompted chemists to explore the use of environmentally benign, inexpensive, and earth-abundant metals [18-27]. In this section, we summarize recent progress in Ni, Fe, Cu, and Co-catalyzed domino reactions of strained bicyclic alkenes.

Nickel-catalyzed reactions

Without close inspection, nickel might seem like the peculiar younger sibling of palladium within the field of transition-metal catalysis. Nickel lies directly above palladium in the periodic table, as such, it readily performs many of the same elementary reactions. Because of their reactive commonalties, nickel is often seen as the budget-friendly replacement; however, this misconception will clearly be refuted in this section, showcasing several diverse nickel-catalyzed domino reactions.

In 2001, Rayabarapu and co-workers investigated the Ni-catalyzed ring-opening/cyclization cascade of heterobicyclic alkenes 1 with alkyl propiolates 2 for the synthesis of coumarin derivatives 3 (Scheme 1) [28]. The reaction initiates with the in situ reduction of Ni(II) to Ni(0) followed by the side-on coordination of the alkene and alkyne substrates to the metal center with subsequent oxidative cyclometallation to form a nickel metallacycle, similar to several reported Ni-catalyzed [2 + 2] cycloadditions [29,30]. Rather than undergoing reductive elimination to afford to [2 + 2] adduct, β-oxygen elimination followed by E/Z isomerization and intramolecular lactonization generates the annulated coumarin scaffold. In 2003, the Cheng lab extended on this Ni-catalyzed ring-opening strategy [31]. It was noted the addition of 1.5 equivalents of water interrupted the cyclization step and led entirely to reductively coupled alkenylated ring-opened products. Interestingly, when this methodology was applied to the ester-bearing oxabicyclic 1a, the anticipated reductive coupling product was not detected; instead, bicyclic γ-lactone 4 was solely observed (Scheme 1). This unprecedented lactone is presumed to be generated through the expected reductive coupling to generate the ring-opened intermediate 5 which undergoes subsequent intramolecular lactonization with the distal ester group. In the same year, Cheng and co-workers observed the identical reactivity when exploring the Pd- and Ni-catalyzed asymmetric reductive ring opening of heterobicyclic alkenes, ultimately generating the bicyclic product 7 (Scheme 1) [32].

In 2003, the Cheng laboratory continued studying Ni-catalyzed routes towards coumarin cores through the Ni-catalyzed ring-opening/cyclization cascade of heterobicyclic alkenes 1 with β-iodo-(Z)-propenoates and o-iodobenzoates 9 (Scheme 2) [33]. The authors noted the ring-opening/cyclization cascade proceeded smoothly for a variety of heterobicyclic alkenes including both oxa- and azabenzonorbornadienes as well as oxanorbornenes; however, the latter two substrates did not undergo dehydrogenation, generating cis-selective annulated coumarins (10b and 10d). In 2006, the same group applied this methodology for the total synthesis of arnottin I (10h), a coumarin-type natural product isolated from the bark of the Xanthoxylum arnottianum Maxim which possesses some antibiotic properties [34]. Mechanistically, the authors proposed the reaction begins with the in situ reduction of Ni(II) to Ni(0) by zinc to generate Ni(0) which undergoes oxidative addition with the organo iodide to yield Ni(II) intermediate 11. Coordination of 11 to the bicyclic alkene followed by migratory insertion affords intermediate 12 which undergoes β-oxygen elimination to form 13. Rearrangement of 13 via β-hydride elimination and enolization generates a 1-naphthol species which undergoes intramolecular cyclization with the ester to form the final product 10. The selectivity for the non-dehydrogenated coumarins 10d is not understood, but 10b likely does not undergo dehydrogenation because there is no formation of aromaticity to drive the reaction forward. When the bicyclic alkene is substituted unsymmetrically at the bridgehead position, the reaction is entirely regioselective for the formation of a 1,2,4-trisubstituted pattern. The observed regioselectivity arises from the preferential migratory insertion of the aryl group distal to the bridgehead substituent.

In 2010, Ogata and Fukuzawa explored the Ni-catalyzed two- and three-component difunctionalization of norbornene derivatives 15 with alkynes (Scheme 3) [35]. It was noted the reaction is amenable to both electron-donating groups (EDGs) and electron-withdrawing groups (EWGs); however, yields were diminished with increasing electron deficiency. Moreover, the use of the bulkier tert-butyldimethylsilyl-protecting group resulted in the corresponding 1,5-enyne only being produced in a 33% yield. Several different norbornene derivatives were explored and gave the anticipated exo,exo-difunctionalized product in good yield. In contrast, when using an ethylene-bridged bicycloalkene to generate the product 19c, the latter was obtained in a greatly reduced yield, perhaps due to less ring strain providing a thermodynamic driving force.

In 2013, Mannathan et al. discussed a Ni-catalyzed intermolecular three-component difunctionalization of oxabicyclic alkenes 1 with organoboronic acids 20 and alkynes 17 (Scheme 4) [36]. While broadly successful, when electron-deficient arylboronic acids were used, slightly diminished yields were observed. Moreover, when 3-hexyne was used, the reaction failed to afford any product. The reaction likely begins similarly to Cheng’s 2003 report (Scheme 1) [31] where the coordination of the alkyne 17 and alkene 1 to the Ni(0) center, followed by oxidative cyclometallation, yields the following nickelocycle 24. Unlike Cheng’s 2003 report, which proposes subsequent β-oxygen elimination (Scheme 1) [31], alkoholysis by MeOH affords an alkyl(methoxy)nickel intermediate 25. Transmetalation of 25 with the organoboronic acid gives intermediate 26, which upon reductive elimination affords the difunctionalized product 21 and regenerates the Ni(0) catalyst.

In 2019, the Stanley laboratory explored the Ni-catalyzed intermolecular three-component carboacylation of norbornene derivatives 15 using imides 27 and tetraarylborates 28 (Scheme 5) [37]. The method utilizes C–N bond activation to trigger the reaction. The authors demonstrated a broad reaction scope. Electron-deficient amides were shown to perform worse than their electron-rich counterparts with the p-trifluoromethyl substituent forming the ketone product in <10% yield. While substitution of the norbornene was tolerated, both EWGs and EDGs hindered the reaction. Upon several mechanistic studies, the authors proposed the catalytic cycle begins with the oxidative addition of the active Ni(0) catalyst to imide 27 to afford the acyl–Ni(II)–amido intermediate 30. Side-on coordination followed by migratory insertion of the bicyclic alkene selectively generates the exo-alkyl–Ni(II)–amido complex 31. Transmetalation with triarylborane affords 32 which undergoes reductive elimination to form the carboacylated product 29 as well as regenerates the Ni(0) catalyst. In 2022, the Tobisu group explored a two-component carboacylation of norbornene derivatives. Exploiting a Ni/NHC system, the authors were able to develop an entirely atom-economic carboacylation process utilizing N-indoyl arylamides [38].

In 2019, Gutierrez and Molander reported the coupling 4-alkyl-1,4-dihydropyridines 31 with heterobicyclic alkenes 30 under photoredox/Ni dual catalysis (Scheme 6) [39]. In contrast to other photoredox-mediated transformations, the authors utilized the inexpensive organic photosensitizer 4-CzIPN (Scheme 6 and Scheme 7) instead of the more commonly, and expensive, metal-based photocatalysts. While broadly successful, tertiary radicals failed to deliver any desired product. Of note, the reaction was amenable to a broad scope of derivatized heterobicyclic alkenes with mono- and disubstituted bridgeheads having little effect on the reaction (32b) with reactions involving unsymmetrically substituted bicyclic alkenes demonstrating complete regioselectivity for either 1,2,3- or 1,2,4-trisubstitued products (32a, 32f). DFT calculations were used to explain the syn-1,2-substitution experimentally observed rather than the possible syn-1,4-substituted product. It was found the reductive elimination transition state leading to the 1,4-disubstituted product TS_33-P1 would require an increase in distortion energy compared to TS_35-P2 which contributes to an overall greater kinetic barrier.

The following year, Lautens and Renaud expanded the scope of the photoredox/Ni dual-catalyzed coupling of alkyl nucleophiles 36 with heterobicyclic alkenes 30 to include α-amino radicals (Scheme 7) [40]. The authors noted the electron-rich oxabenzonorbornadiene derivatives provided the corresponding ring-opened adducts in good yields (63–68% yield) while those bearing EWG led to poor product formation. Unlike Gutierrez and Molander’s work (Scheme 6) [39], it was found mono- and disubstituted bridgeheads affected the efficacy of the reaction with the demethylated bridgehead oxabenzonorbornadiene only delivering the product in a 20% yield. Although yields were slightly diminished, unsymmetrical bridgehead-monosubstituted oxabenzonorbornadiene led solely to the 1,2,4-trisubstituted regioisomer (Scheme 7), similar to that observed by Gutierrez and Molander [39]. Selected substituents on the aniline motif were found to hamper reactivity with a few examples failing to provide the desired product when 4-CzIPN was used as the photocatalyst; however, the products were isolated when [Ir(dF(CF₃)ppy)₂(bpy)]PF₆ was used. Based on experimental observations and control reactions, the authors proposed the reaction begins with the photoexcitation of the photosensitizer 43 to form 44 which can oxidize aniline 36a to give radical cation 46 (Scheme 7). Deprotonation by DBU produces the radical 40. The radical anion photosensitizer 45 can reduce Ni(I) to Ni(0), closing the first catalytic cycle. The Ni(0) complex can undergo oxidative addition into the C–O bond of the oxabicyclic alkene 30a to afford the σ-allyl intermediate 38 which can isomerize to the more stable π-allyl intermediate 39. Addition of the α-amino radical to the Ni(II) center generates the Ni(III) complex 41. Reductive elimination, followed by protodemetalation, leads to the final ring-opened adduct 37.

Copper-catalyzed reactions

In 2009, Pineschi and co-workers explored the Cu-catalyzed rearrangement/allylic alkylation of 2,3-diazabicyclo[2.2.1]heptenes 47 with Grignard reagents 48 (Scheme 8) [41]. The reaction is thought to proceed via the Lewis acid-catalyzed [3,4]-sigmatropic rearrangement of the diazabicycle 47 to form the allylic carbazate intermediate 51. Nucleophilic attack of an organomagnesium, or organocuprate, in an anti S_N2’ fashion on 52 furnish the final ring-opened product 49. The authors note the use of a carbamate protecting group was crucial for the success of the reaction, hypothesizing it inhibited the classical [3,3]-sigmatropic Lewis acid-catalyzed rearrangement often observed. Both alkyl and aryl Grignard reagents were amenable to the reaction; however, heteroaryl Grignard reagents resulted in poor conversion.

The Cu-catalyzed borylative difunctionalization of π-systems is a power tool for the facile synthesis of complex boronate-containing compounds [42]. Generally, these reactions proceed through the generation of a Cu–boryl species via σ-bond metathesis, followed by migratory insertion with a π-system. The subsequent alkyl–Cu intermediate is intercepted by an electrophile to generate the difunctionalized system. This methodology has been applied several times to strained bicyclic alkenes with a variety of electrophiles.

In 2015, Hirano and Miura developed a Cu-catalyzed aminoboration of bicyclic alkenes 1 with bis(pinacolato)diboron (B₂pin₂) (53) and O-benzoylhydroxylamine derivatives 54 (Scheme 9) [43]. While the scope of bicyclic alkenes was quite extensive with aza-, carbo-, and oxabicyclic alkenes being amenable to the reaction, electron-deficient substrates resulted in lowered yields. Of note, the reaction is highly regioselective with the unsymmetrically methyl-substituted bicyclic alkene producing a single regioisomer 55a. The authors noted the aminoborylated products bearing a BPin moiety were not always stable upon isolation, so they were either converted into the more stable Bdan (dan = 1,8-diaminonaphthalenyl) or Bpin-Bdan was used directly which showed comparable yields. The authors also reported preliminary results for an asymmetric variant of the reaction using (R,R)-Ph-BPE as a chiral ligand. Although the use of the chiral phosphine ligand resulted in slightly diminished yields, the authors were able to achieve ees up to 88%. The authors proposed the reaction begins with the generation of the tert-butoxide Cu salt which undergoes σ-bond metathesis with B₂Pin₂ generating the Cu–boryl species 59 (Scheme 9). Side-on coordination on the exo face of the bicyclic alkene followed by migratory insertion generates the alkyl–Cu species 60 which after electrophilic amination with the O-benzoylhydroxylamine 54 liberates the final aminoborylated product 55 and a benzoyl–Cu complex 61. To close the catalytic cycle a transmetalation of 61 with LiOt-Bu regenerates the active catalyst.

In 2017, Xiao and Fu studied the Cu-catalyzed borylalkynylation of oxabenzonorbornadiene (30b) with B₂pin₂ (53) and bromoalkynes 62 (Scheme 10) [44]. The scope of the reaction was limited to only two examples of bromoalkynes reacting with oxabenzonorbornadiene (30b). Notably, the yield of the reaction dramatically diminished when the terminal triisopropylsilyl (TIPS) group in 63a was swapped for a Ph (63b). Mechanistically, the reaction operates in a similar manner reported by Hirano and Miura (Scheme 9) [43]; however, the alkyl–Cu species 60 is intercepted by the bromoalkyne rather than an O-benzoylhydroxylamine.

In the same year, the Brown laboratory investigated the Cu-catalyzed borylacylation of bicyclic alkenes 1 (Scheme 11) [45]. Like the previous borylative difunctionalization reactions, it was found the reaction generated a single exo,exo diastereomer. A brief investigation into an enantioselective variant of the borylacylation was investigated; however, the methodology was not applied to bicyclic alkenes.

In 2019, the Yang lab examined the Cu-catalyzed diastereoselective 1,2-difunctionalization of oxabenzonorbornadienes 30 for the synthesis of β-thiocyanato thioethers 68 (Scheme 12) [46]. In contrast to the previous difunctionalization reactions, the authors noted the reaction was stereoselective for the trans-addition product. Mechanistically, the authors proposed the reaction begins with the Cu-mediated substitution reaction of iodobenzene (66a) with KSCN to afford phenyl thiocyanate (70). The Cu complex can then undergo oxidative addition into the S–C bond of the thiocyanate 70 to afford intermediate 71 which can side-on coordinate to the exo face of 30b. Subsequently, the thiocyanate attacks the olefin from the endo face via 72 to give complex 73. Reductive elimination furnishes the final difunctionalized product and regenerates the active Cu(I) catalyst. The reaction was broadly successful with the steric and electronic nature of the aryl iodide having little effect on the reaction.

Iron-catalyzed reactions

Being the most earth-abundant d-block element, as well as orders of magnitude less expensive than other transition-metal catalysts, iron is bringing a renaissance to the idea of sustainable, green catalysis. In 2011, Ito et al. reported a diastereoselective Fe-catalyzed carbozincation of heterobicyclic alkenes 1 with diphenylzinc (74a) (Scheme 13) [47]. Using an ortho-phenylene diphosphine ligand L3, the authors were able to suppress β-heteroatom elimination enabling sequential electrophilic trapping of the alkylzinc complex. Although this reaction would more closely fall under the definition of a telescoped reaction than a strict domino reaction, this methodology allowed for the synthesis of difunctionalized strained alkenes. While broadly successful, strongly electron-withdrawing groups lowered the yield of the reaction. In 2021, Isozaki and Nakamura reinvestigated the reaction and established an asymmetric variant of the Fe-catalyzed carbozincation of azabicyclic alkenes 77 (Scheme 13) [48]. Using (S,S)-chiraphos, the authors were able to achieve enantioselectivities of up to 99%. Unfortunately, only two examples of electrophilic capturing were explored, using CD₃CO₂D to give deuterated products and I₂. Most reports simply underwent protodemetalation upon quenching to afford the monosubstituted bicyclic alkene. The catalytic cycle starts with a diaryl Fe(II)–(S,S)-chiraphos complex 80 being generated through the reduction of FeCl₃ with excess diarylzinc in the presence of the phosphine ligand. Side-on coordination to the exo face of the azabicycle 77a generates 81 where subsequent migratory insertion affords the alkyl–Fe(II) complex 82. Transmetalation with an organozinc produces 78a which can be trapped by an electrophile to generate the final product 79a.

Cobalt-catalyzed reactions

In 2014, the Yoshikai lab investigated the Co-catalyzed addition of arylzinc reagents 83 of norbornene derivatives 15 (Scheme 14) [49]. In contrast to the 1,2-difunctionalization of bicyclic alkenes via arylzinc reagents reported by Nakamura under Fe catalysis (Scheme 13) [48], this reaction is considered to undergo a 1,4-Co migration ultimately generating 1,4-difunctionalization species. Mechanistically, the reaction likely proceeds similarly to Nakamura’s Fe-catalyzed methodology (Scheme 13) [48].

In 2017, the Cheng laboratory investigated the Co-catalyzed ring-opening/dehydration of oxabicyclic alkenes via the C–H activation of arenes (Scheme 15) [50]. First, the group explored the ortho-naphthylation of N-pyrimidinylindole derivatives 85. The reaction was amenable for both electron-rich and deficient indoles. When the reaction was attempted on electron-deficient oxabicyclic alkene derivatives, it was observed the reaction did not undergo dehydration to give the 2-naphthyl product, rather the ring-opened 1,2-hydroxy adduct. When the Lewis acid cocatalyst AgSbF₆ was removed from the reaction mixture, it was noted only ring-opened 1,2-hydroxy adducts were formed, so it is likely the Lewis acid is required for dehydration. In contrast, when N-pyrimidinylbenzimidazole derivatives were used, the 1,2-C–H addition product was observed exclusively. By slightly altering the reaction conditions, 2-arylpyridines 85a were able to undergo the ring-opening/dehydration reaction with oxabicyclic alkenes to afford ortho-naphthylated products 86a.

Concurrently, the Li group investigated the same ortho-naphthylation of N-pyrimidinylindole derivatives 85 (Scheme 15) [51]. In contrast to Cheng’s report, it’s noted the addition of AcOH rather than CsOAc enabled the same ring-opening/dehydration cascade to occur; however, acidic conditions seem to require less energy to drive the dehydration step.

In 2019, the Zhai Group investigated the Co-catalyzed [3 + 2] annulation/ring-opening/dehydration domino reaction of oxabicyclic alkenes 30 with 2-(1-methylhydrazinyl)pyridine (MHP) directed arenes 87 for the synthesis of benzo[b]fluorenones 88 (Scheme 16) [52]. C–H bond functionalization with heterobicyclic alkenes as annulation partners has received considerable attention in recent years. Several different arene and directing groups have been investigated; however, they typically result in the exo-selective addition product with the bridge heteroatom intact. Although this limits the applicability of the reaction, the authors noted the use of 5.0 equivalents of Cs₂CO₃ provided the naphthalene core via sequential dehydration. Based on preliminary mechanistic experiments, the authors proposed the reaction begins with the oxidation of Co(II) to Co(III) by O₂. MHP-directed C–H activation of the ortho-C–H position generates 90 which can coordinate to the bicyclic alkene forming 91. Migratory insertion of the olefin affords 92 which undergoes intramolecular nucleophilic addition followed by protodemetalation and elimination of MHP to afford 94. Base-mediated ring opening of the bridging ether generates 95 which undergoes an elimination reaction to afford the naphthalene product 88a.

Inspired by Zhao’s seminal report on the racemic carboamination of bicyclic alkenes [53], the Cramer laboratory studied the Co-catalyzed enantioselective carboamination of bicyclic alkenes 1 via C–H functionalization in 2021 (Scheme 17) [54]. The authors noted decreasing the steric bulk of the amide moiety of the substrate from isopropyl to ethyl to methyl decreased the enantioselectivity of the reaction. Carbon- and nitrogen-bridging bicyclic alkenes were also identified as competent substrates. In this respect, norbornadiene was found to give the desired carboaminated product in slightly diminished yields while azabicyclic alkenes generated the targeted products in excellent yield, albeit with slightly reduced enantioselectivity. To showcase the synthetic capabilities of this methodology, the authors synthesized the non-natural amino acid derivative 98j in good diastereoselectivity.

Ruthenium-catalyzed reactions

In 2006, the Tam laboratory investigated the Ru-catalyzed cyclization of oxabenzonorbornene derivatives 30 with propargylic alcohols 99 for the synthesis of isochromenes 100 (Scheme 18) [55]. After coordination of the Ru-center to the exo face of 30b, oxidative cyclization can afford the ruthenacycle 101. Unlike previous works studying Ru-catalyzed cyclizations involving bicyclic alkenes and alkynes [56-59], the reaction preferentially undergoes β-hydride elimination to generate 102 rather than reductive elimination which would afford the [2 + 2] adduct. Hydroruthenation of the allene produces 103 which can either undergo reductive elimination to afford the cyclopropanated bicyclic alkene or undergo a [2 + 2] cycloreversion to generate the Ru–carbene 104. The Ru–carbene 104 can rearrange to 100 through a 1,3-migration of the alkoxy group which can finally reductively eliminate the isochromene product. Based on control reactions, the authors proposed the active catalytic species is cationic, as the use of the cationic precatalyst [Cp*Ru(CH₃CN)₃]PF₆ in THF afforded the isochromene as the major product, suggesting a similar cationic species may be generated in MeOH [60].

In 2011, Tenaglia and co-workers investigated the Ru-catalyzed coupling of oxabenzonorbornene derivatives 30 with propargylic alcohols and ethers 106 to access benzonorcaradienes 107 (Scheme 19) [61]. While discriminating between the neutral and cationic active ruthenium species, the authors noted the use of [Cp*Ru(CH₃CN)₃]PF₆ as the precatalyst produced the cyclopropanated bicyclic alkene adducts exclusively. This contrasts with Tam’s report (Scheme 18) [55] which found cationic Ru species formed the isochromene 100 preferentially which may be attributed to the solvent playing a more impactful role in the reaction than previously anticipated. Of note, the reaction was amenable to a broad scope of derivatized heterobicyclic alkenes. Electron-deficient bicyclic alkenes were found to react much slower, ultimately affording products in diminished yields. Mono- and disubstituted bridgehead variants were applicable, but with reduced efficacy with the former producing a dihydronaphthofuran 107i as the major product.

In 2019, the Cramer group continued studying this reaction and developed an enantioselective variant utilizing a chiral Cp* derivative (Scheme 19) [62]. Similar reactivity trends were observed in both accounts. Mechanistically, the transformation was proposed to begin with the coordination of Cp*RuI to the exo face of the bicyclic alkene. Oxidative addition into the C–O bond, which is proposed to be the enantiodetermining transition state, followed by coordination to the alkyne generates intermediate 109. Migratory insertion of the alkyne results in the ruthenacycle 110. Subsequent reductive elimination generates putative allyl vinyl ether 111 and regenerates the active ruthenium complex. The allyl vinyl ether intermediate undergoes a Claisen rearrangement to afford the endo-isomer 112. Thermal isomerization of 113a by a 6π-electrocyclic ring-opening/closing cascade leads to the to the final exo-isomer 107.

In 2018, the Zhang lab investigated the Ru-catalyzed ring-opening/dehydration of oxabicyclic alkenes 30 via the C–H activation of anilides 114 (Scheme 20) [63]. When the optimized conditions were applied to azabenzonorbornadiene derivatives, the dehydrative naphthylation sequence did not occur with the reaction being exclusive for exo-ring-opened products, similar to that observed in a typical Rh-catalyzed ring-opening reaction (vide infra). The reaction seems to be sensitive to the steric bulk of the amide functionality with n-propyl and isopropylamides having diminished yields. While the scope of anilides was quite extensive, electron-deficient substrates resulted in lowered yields.

In 2022, the Jeganmohan group investigated the Ru-catalyzed ring-opening/lactamization of azabenzonorbornadiene derivatives 30 with arylamides 116 (Scheme 21) [64]. Weinreb amides outperformed other arylamides, likely serving as a better directing group for the initial aryl-C–H activation. While the scope of functionalized aryl Weinreb amides was quite wide, including different EWGs and EDGs, as well as heterocycles, ortho-substitution was not tolerated. The authors applied the methodology for the synthesis of biologically important benzo[c]phenanthridine derivatives 117. Through methylation and subsequent aromatization of the phenanthridinones produced, the authors were able to quickly afford novel fagaronine 117j and nitidine 117k derivatives.

Rhodium-catalyzed reactions

In 2002, the Lautens laboratory reported a tandem cyclization of arylboronate esters 118 with a variety of bicyclic alkenes 15 using a water-soluble Rh-catalytic system (Scheme 22) [65]. The authors reported the reaction proceeded smoothly with a limited variety of substituted norbornenes and boronate esters.

In 2004, the same group expanded this Rh-catalyzed cyclization to heterobicyclic alkenes 1 with arylboronate esters 118 for the synthesis of a variety of functionalized indanes 120 (Scheme 22) [66]. This reaction proceeded smoothly with a broad range of [2.2.1] and [3.2.1]-bicyclic alkenes; however, doubly bridgehead-substituted bicyclic alkenes exclusively produced an undesirable demetalated aryl ester byproduct. The authors attributed this to a steric prevention of the attack of the arylrhodium nucleophile to the alkene. Azabicyclic alkenes also proved difficult and failed to react. Mechanistically the authors proposed the arylboronate ester 118 first undergoes a transmetalation with the Rh(I) complex producing 122 which performs an exo-carborhodation with the bicyclic substrate to produce 123. A 5-exo-trig ring closure of 123 produces 124 followed by a rapid protodemetalation with water releasing the final indane product 119a and regenerating the active Rh(I) species. The authors proposed that the origin of the diastereoselectivity is due to significant steric interactions between the –COR group on the pendant alkene and the bridging group of the bicyclic alkene in 123.

In 2006, the Lautens lab continued to extend this reaction to include dienylboronate esters 125 and found that an unexpected vinylcyclopropane product 126 was generated (Scheme 23) [67]. Again, nitrogen-containing bicyclic alkenes proved difficult, as diazabicyclic alkenes produced the desired product in low yields while azabenzonorbornadienes failed to react entirely. It was found the introduction of a methyl group α to the boron on the dienylboronate caused the selectivity to be shifted to the 1,4-addition producing a cyclopentene product leading to the conclusion that the substitution pattern on the boronate ester played a significant role in the selectivity between 1,6-addition and 1,4-addition. The mechanism proposed by the authors initially begins in the same manner as Scheme 22 with the transmetalation of the boronate ester with Rh(I) producing 127 which undergoes an exo-carborhodation with the bicyclic substrate 15a producing 128. The reaction path diverges from the previous mechanism undergoing a 1,6-addition resulting in 129. A rapid protodemetalation with water then occurs releasing the final vinylcyclopropane product 126a and regenerating the active Rh(I) species. A later 2009 investigation revealed methyl groups α to the ester produced a hydrofunctionalization product [68]. Dienylboronate esters bearing methyl groups β to the ester group produced vinylcyclopropane products 126 while dienylboronate esters bearing methyl groups at the δ or γ position resulted in cyclopentene products.

In 2009, the Lautens lab continued investigating the reactions of bicyclic alkenes 130 with a novel range of heteroaromatic boronate esters 131 (Scheme 23) [69]. This has previously been a challenging transformation due to the propensity of these systems to produce non-cyclized hydroarylation products due to an unproductive rhodium 1,4-migration on heteroaromatic moieties. The use of benzothiophene, benzofurans, and benzopyrrole boronate esters in this investigation prevented this shift as these systems lack the hydrogen to participate in this shift. This reaction proceeded smoothly with a variety of bicyclic alkenes although diazabicyclic alkenes had little to no reactivity. Moreover, benzofuran and benzopyrrole boronate esters resulted in low yields.

In 2011, the Lautens lab reported the Rh-catalyzed domino reaction of doubly bridgehead-substituted oxabicyclic alkenes 134 with secondary amine nucleophiles 135 for the synthesis of bicyclo[2.2.2]lactones 136 (Scheme 24) [70]. This reaction proceeded smoothly with a variety of secondary amine nucleophiles, including those with hydrocarbon, ether, acetal, and ester functionalities; although, aniline nucleophiles only resulted in the one step asymmetric ring-opening (ARO) product under the standard reaction conditions. Fortunately, the authors noted the addition of triethylamine allowed for aniline nucleophiles to undergo the domino reaction, generating the desired bicyclo[2.2.2]lactone 136. The authors proposed the reaction first takes place through an ARO of the doubly bridgehead-substituted oxabicyclic alkene with the secondary amine nucleophile ultimately producing 137. The Rh(I) catalyst then facilitates the allylic alcohol isomerization in 137 resulting in the aldehyde 138. This aldehyde, in close proximity to the tertiary alcohol, leads to the production of the hemiacetal 139 which can finally undergo an oxidation producing the final bicyclo[2.2.2]lactone product 136.

In 2011, the Radhakrishnan laboratory reported the carboannulation of diazabicyclic alkenes 130a with 2-cyanophenylboronic acid (140) and 2-formylphenylboronic acid (142) for the synthesis of indanones 141 (Scheme 25) [71]. This reaction proceeded smoothly with a variety of substituted diazabicyclic alkenes including a variety of ester substituents on the nitrogens and sterically more hindered tricyclic adducts. Mechanistically, the authors proposed the reaction begins with a transmetalation of 2-cyanophenylboronic acid with the Rh(I) species resulting in 143. Upon association of 143 with the diazabicyclic alkene 132a a syn exo-addition occurs producing 144. Subsequently, coordination of the Rh(I) to the electrophilic cyano group leads to an intramolecular addition producing 145. The imine undergoes a hydrolysis releasing the final carboannulated product 141 as well as regeneration of the active Rh(I) catalyst. A similar mechanism can be envisioned for the carbonannulation of diazabicyclic alkenes with 2-formylphenylboronic acid up to the last step which likely operates through a β-hydride elimination of the Rh(I) alkoxide, furnishing the final carbonyl-containing product.

In 2013, Lautens reported the synthesis of oxazolidinone scaffolds 147 through a domino ARO reaction followed by cyclization reaction of oxabicyclic alkenes 30 with sodium cyanate (146) (Scheme 26) [72]. This reaction proceeded smoothly with electron-rich oxabenzonorbornadiene derivatives; however, electron-poor oxabenzonorbornadiene derivatives resulted in reduced yield and enantioselectivity. Bridgehead-substituted, non-benzo-fused oxabicycles, as well as azabicyclic alkenes failed to produce the desired product. When the benzo-fused moiety was unsymmetrically substituted, little regioselectivity was observed. Based on X-ray crystallographic data for their final product, and previously reported Rh-catalyzed ARO reactions, the authors hypothesized the reaction begins with the oxidative addition of the Rh(I) catalyst into the bridgehead C–O bond of the oxabenzonorbornadiene producing 148 which is considered the enantiodetermining step. The isocyanate anion then nucleophillically attacks the alkene in an S_N2’ fashion producing the trans-isocyanate 149. Subsequently, insertion of the Rh–O bond into the isocyanate results in 150. Finally, protonolysis produces the oxazolidinone product 147e as well as regenerates the active Rh(I) catalyst.

In 2013, the Radhakrishnan laboratory reported the Rh-catalyzed oxidative coupling of salicylaldehyde derivatives 151 with diazabicyclic alkenes 130a producing fused chromanone derivatives 152 (Scheme 27) [73]. It was determined alkyl- and methoxy-substituted salicylaldehydes resulted in a minor reduction of yield while salicylaldehydes with EWGs failed to react. The authors hypothesized the reaction mechanism begins with the association of the Rh(III) catalyst with the hydroxy group of salicylaldehyde (151a) resulting in a selective cleavage of the aldehyde C–H bond producing the rhodocycle 153 which side-on coordinates with the alkene of the azabicycle producing 154. A C–N bond cleavage occurs creating π-allylrhodium 155. Subsequently, the phenol oxygen then adds to the π–allyl species in a cis fashion, furnishing 156 which is proposed to be the enantiodetermining step. The carbonyl–rhodium species 156 inserts into the alkene to produce 157. Following this, β-hydride elimination occurs yielding the final product 152 and a Rh(I) species which is oxidized back to its active Rh(III) state by Cu(OAc)₂·H₂O.

In 2017, Radhakrishnan reported a Rh-catalyzed annulation of O-acetyl ketoximes 159 or N-methoxybenzamides 161 with [2.3.1]-bicyclic alkenes 158 for the synthesis of isoquinoline (160) or isoquinolone-fused bicycles 162 (Scheme 28) [74]. Compared to their previous C–H functionalization reaction (Scheme 27) [73], no ring opening was observed. This reaction with O-acetyl ketoximes was amenable to a variety of para-substituents including methoxy and halide groups; however, O-acetyl ketoximes with ortho- or meta-substituents failed to react. A small number of substituted [2.2.1]diazabicyclic alkenes 130a were successfully employed, albeit with slightly lower yields. In the reaction with N-methoxybenzamides 161, the same substituent trends were seen as that with the reaction with O-acetyl ketoximes. Mechanistically, the reaction begins when the Rh(III) catalyst is converted to an active Rh(III) species, by AgSbF₆ and Cu(OAc)₂, which oxidatively inserts into the ortho C–H bond forming 163. Migratory insertion of the alkene forms 164. Next, cleavage of the N–O bond followed by an oxidative addition of the Rh(III) to the N–O bond forms intermediate 165 which can finally undergo reductive elimination giving the final product 160a.

In 2013, Li reported the domino coupling reaction of 2-phenylpyridines 165 with oxa- and azabicyclic alkenes 30 (Scheme 29) [75]. When investigating the reaction with oxabenzonorbornadiene (30b), the resulting product was found to exclusively be the dehydrated 2-naphthalene derivative 166. It was found that the addition of pivalic acid greatly improved the yield, likely due to its facilitation of C–H activation as well as its involvement in the dehydration process. This reaction proceeded smoothly with a variety of both EWGs and EDGs on the 2-phenylpyridine. Interestingly, when swapping the pyridine directing group for thiophene or furan, yields were improved although quinolinyl and pyrimidyl directing groups, despite reacting, resulted in a mixture of mono- and diarylation products. When investigating substituted oxabenzonorbornadienes both mono- and diarylated products were formed with only moderate yield. When azabenzonorbornadienes 30 were investigated in the same redox-neutral conditions no reaction occurred; however, upon the addition of AgOAc a cis-fused dihydrocarbazole product was formed (Scheme 29). Mechanistically this reaction was proposed to proceed through first a conversion of the Rh(III) catalyst to the active Rh(III) species by AgSbF₆. This active Rh(III) catalyzes the cleavage of the ortho-C–H bond of 2-phenylpyridine furnishing 168. This is followed by the cis addition of 168 to the oxabenzonorbornadiene producing 169 whereby subsequent β-oxygen elimination affords 170, followed by protonolysis producing 171 and regenerating the active Rh(III) species. Finally, a dehydration occurs furnishing the final product 166. In terms of the azabicyclic substrates, following the β-eliminated heteroatom, a second round of C–H activation/reductive elimination occurs to generate the annulated product 167.

In 2014, Chen and Li reported the Rh-catalyzed domino dehydrative naphthylation of oxabenzonorbornadienes 30 with N-sulfonyl 2-aminobenzaldehydes 172 (Scheme 30) [76]. This reaction was amenable to a variety of EDG, EWG, as well as a broad scope of sulfonyl groups. Surprisingly, this reaction also proceeded smoothly with nitro substituents on the benzene ring which are typically problematic in C–H activation reactions. Through mechanistic studies, the authors proposed the rate limiting step for this reaction is the C–H cleavage.

In 2015, Miura and co-workers reported the Rh-catalyzed domino dehydrative naphthylation of oxabenzonorbornadienes 30 with arylphosphine derivatives 174 (Scheme 31) [77]. The reaction was amenable to a wide range of substituted arylphosphine derivatives. Moreover, the reaction could be extended to include various phosphinate, phosphonate, and phosphonamide derivatives. The use of triarylphosphine oxides required the reaction to be performed at a 2:1 ratio with oxabenzonorbornadienes 30 to prevent multiarylated products from being formed. Arylphosphine sulfides were also investigated but gave unimpressive yields (8%); however, upon a substitution of the AgOAc for 3 equiv of AcOH moderate yields were obtained (39%). Mechanistically, this reaction likely operates in a similar manner to the previously discussed C–H activation/dehydration domino reactions.

In 2020, a similar method for the construction of 2-aryl-substituted naphthalene cores was discussed by Wang and co-workers who disclosed a Rh-catalyzed C–H bond naphthylation of anilides and benzamides with azabenzonorbornadienes [78]. Interestingly, the dehydration step occurred smoothly with an aza-leaving group rather than the more common oxa-leaving group discussed above.

In 2013, the Radhakrishnan laboratory reported the Rh-catalyzed domino ring-opening coupling reaction of azaspirotricyclic alkenes 176 using arylboronic acids 177 (Scheme 32) [79]. This reaction proceeded well with a variety of ester substituents on the nitrogens of the azaspirotricyclic alkenes. The authors proposed this reaction proceeds first through a transmetalation of the arylboronic acid 177a with the Rh(I) catalyst producing 179 which undergoes a cis addition to the azaspirotricyclic alkene resulting in intermediate 180. C–H cleavage at the ortho-position followed by an intramolecular reductive elimination affords in 182. Unlike previous reports [80], this arylrhodium complex has a long enough lifetime to propagate further. A subsequent migratory insertion into a second azaspirotricyclic alkene furnishes 183. Finally, the anion from the catalyst attacks 183 causing a ring opening, forming the final product 178d and regenerating the Rh(I) catalyst. Keeping with other mechanisms, the Rh(I) may also undergo an anti-β-nitrogen elimination to furnish the ring-opened intermediate [80].

In 2016, Liu reported the Rh(III)/Sc(III)-catalyzed domino reaction of oxabenzonorbornadienes 30 with alkynols 184 directed by a transient hemiketal group (Scheme 33) [81]. The use of a transient directing group avoids the tedious process of installation and then removal of directing groups which is a common issue. A variety of substituents were tolerated on both the alkynols 184 and oxabenzonorbornadienes 30; however, substituted oxabenzonorbornadiene derivatives typically had diminished reactivity. Expansion of the bicyclic scope was limited as other bicycles such as norbornene failed to react. The authors propose the catalytic cycle begins with the Rh(III)-catalyzed hydration of the alkynol to produce 186 followed by a Sc(III)-catalyzed addition to form the transient hemiketal 187. Ortho-C–H activation generates 188 which can undergo migratory insertion with the Sc(III)-coordinated oxabicyclic alkene 189 to form 190. β-Oxygen elimination, likely assisted by the Sc(III) Lewis acid, produces 191 which subsequently undergoes a protonolysis forming 192 and regenerating the Rh(III) and Sc(III) catalysts. Next, 192 is dehydrated producing 193 which finally undergoes a Prins-type cyclization to afford the final product 185.

In 2018, the Fan laboratory reported the Rh-catalyzed asymmetric cyclization/addition domino reaction of 1,6-enynes 194 with oxa/azabenzonorbornadienes 30 (Scheme 34) [82]. Both oxa- and azabenzonorbornadienes 30 worked well; however, the authors noted the latter produced better enantioselectivities while sterically bulky substituents led to both reduced yield and enantioselectivities. The authors proposed the reaction mechanistically occurs though the coordination and reaction of the Rh(I) species with the 1,6-enyne 194a producing 196 which undergoes an oxidative cyclization leading to 197. Subsequent β-hydride elimination forms 198 which side-on coordinates with azabenzonorbornadiene 30c forming 199. Migratory insertion of the olefin followed by reductive elimination of the hydride affords the final product 195a.

In 2019, the Shao group reported the Rh/Zn-catalyzed domino ARO/cyclization of oxabenzonorbornadienes 30 with phosphorus ylides 201 (Scheme 35) [83]. Despite the difficulty of using phosphorus ylides as nucleophiles in metal-catalyzed reactions due to their ability to strongly bind transition metals, this reaction proceeded smoothly with a broad range of ester-, ketone-, and amide-stabilized phosphorus ylides. Oxabenzonorbornadienes bearing both EWG and EDG substituents worked well including bridgehead-substituted substrates which only experienced a slight reduction in yield. Similar to other ARO reactions, the catalytic cycle is proposed to begin with the oxidative insertion of the Rh(I) catalyst into the bridgehead C–O bond producing 204. The phosphorus ylide attacks 204 in an S_N2’ fashion on the endo face giving the ring-opened 205 as well as regenerating the Rh(I) catalyst after dissociation. Alternatively, 205 can undergo a ring closure followed by a subsequent C–P-bond cleavage causing a ring opening resulting in 207. Intramolecular S_N2’ and elimination of the phosphine oxide generates the final product 202e which the authors propose is stereoselective due to significant steric interactions between the carbonyl and aryl groups. The authors proposed that the Zn(OTf)₂ Lewis acid cocatalyst may activate the bridging oxygen of the oxabenzonorbornadiene lowering the kinetic barrier of C–O oxidative addition.

In 2019, Lautens reported the Rh-catalyzed domino ring opening/lactonization of oxabenzonorbornadienes 30 with nosyl-protected amino acid-derived nucleophiles 208 (Scheme 36) [84]. This reaction proceeded smoothly with a range of amino acid derivatives; however, the authors noted that increased steric bulk of the nucleophiles reduced the yields which they attributed to the lactonization being disfavored on steric grounds. In contrast to other ARO reactions, substituents on the oxabicycles were not tolerated well and only two derivatized substrates successfully reacted with greatly diminished yields. Moreover, amino acid derivatives without α-substituents failed to react, leading the authors to conclude that α-substitution is required to make lactonization kinetically feasible.

In 2019, the Punniyamurthy lab reported the Rh-catalyzed domino C–C/C–N bond formation of azabenzonorbornadienes 30 with aryl-2H-indazoles 210 (Scheme 37) [85]. This reaction was amenable to both EWGs and EDGs; however, it was noted that an azabenzonorbornadiene bearing a pyridine-2-sulfonyl protecting group only produced a trace amount of product which was attributed by the authors to an unproductive chelation of the Rh(III) by the pyridine nitrogen. Furthermore, aryl-2H-indazoles with para-substituents failed to react which the authors attributed to both electronic and steric effects.

In 2020, Bian and Wang reported the Rh/Pd-catalyzed domino reaction of indole derivatives with 2-(phenylethynyl)anilines 212 and oxabenzonorbornadienes 30 (Scheme 38) [86]. In this reaction, both EWG and EDG substitutions were tolerated; although, the authors noted the latter reduced the yield and enantioselectivity of the final product. These indole derivatives are widely present in many nonsteroidal anti-inflammatory drugs such as indomethacin. The indole derivatives synthesized were subjected to virtual screenings for their anti-inflammatory properties and three of them (213a, 213b, and 213c) showed better results than indomethacin. Mechanistically, this transformation proceeds initially through a Rh-catalyzed ARO via the aromatic amine nucleophile followed by an Pd-catalyzed cyclization.

In 2021, He and Tian reported the Rh-catalyzed domino 1,2-carborhodation of heterobicyclic alkenes 30 with B₂pin₂ (53) (Scheme 39) [87]. EDGs and EWGs were well tolerated on the benzo-fused moiety; however, bridgehead substituents shutdown the reaction. Carbocyclic alkenes, like benzonorbornadiene, failed to produce the desired product leading the authors to conclude the bridging heteroatom of oxa- and azabenzonorbornadiene played a vital role in the carboboration reaction.

In 2021, Ellman reported a Rh(III)-catalyzed three-component 1,2-carboamidation reaction of bicyclic alkenes 30 with aromatic and heteroaromatic C–H substrates 215 and dioxazolones 216 (Scheme 40) [88]. This reaction was successful with a wide range of directing groups and substituents on the heteroaromatic C–H substrate and a broad range of bicyclic alkenes. Bicyclic diene derivatives like norbornadiene failed to react, likely due to non-productive complexation to the catalyst. Using a chiral cyclopentadiene ligand, the authors showcased an asymmetric variant of the reaction producing 5 enantioenriched products with an average enantiomeric excess of 80% ee. The authors proposed the reaction begins with a concerted metalation–deprotonation of the aromatic C–H substrate 215a with the Rh(III) catalyst yielding 218. Migratory insertion of the olefin of 15a to 218 produces 219. Subsequently, nitrene insertion of the dioxazolone 216a to 219 furnishes 220, which after protodemetalation yields the final product 217e.

Palladium-catalyzed reactions

In 1998, Kosugi and co-workers explored the Pd-catalyzed diarylation and dialkenylation reactions of norbornene derivatives 8 (Scheme 41) [89]. The authors noted the use of chloroacetone was crucial to the reaction as it acted as an exogenous oxidant. Although not perfect, alkenyl stereochemistry was retained for the majority of examples. In the case of (Z)-tributylstannylacrylate, the exclusive product was the exo-cis-(E,Z)-difunctionalized product. Albeit low yielding, heterobicyclic alkenes were tolerated and produced both diarylated and dialkenylated products 222. On the other hand, benzo-fused heterobicyclic alkenes failed to give the difunctionalized product with the corresponding monofunctionalized ring-opened species being the sole product. Concurrently, the Kang laboratory investigated similar reactivity, disclosing an alternative method for diarylated norbornene derivatives through the three-component coupling of bicyclic alkenes and iodonium salts or diazonium salts with organostannanes, or sodium tetraphenylborate [90]. In 2021, Liu and Chen investigated the use of organoammonium salts and organoboronic compounds as a simple method for the synthesis of diarylated norbornene derivatives [91]. The reaction was also applicable for the addition of benzyl and allyl groups via the organoammonium species.

In 2008, the Liu laboratory explored the Pd-catalyzed three-component 1,2-arylalkynylation of oxabenzonorbornadiene derivatives 30 (Scheme 42) [92]. Unlike previous reports disclosing the coupling of aryl halides and oxabicyclic alkenes, the authors disclosed the use of 5 M aqueous NaOH to hinder unwanted β-oxygen elimination, promoting difunctionalization of the olefin. The use of the phase-transfer catalyst was paramount, as its removal resulted in little to no conversion. Aryl, alkynyl and alkenyl iodide derivatives, as well as methyl iodide, were shown to operate in the reaction; however, only aryl iodide derivatives routinely gave the desired product in appreciable yield.

In 2022, Wan and Chen explored similar reactivity using aryl triflates (Scheme 42) [93]. The scope of aryl triflates was expansive with derivatives from biologically relevant compounds, like vanillin (227b) and eugenol (227a), being applicable. Unfortunately, the authors did not expand their scope beyond carbobicyclic frameworks; however, it would be expected the difunctionalization likely does not occur with heterobicyclic alkenes as β-heteroatom elimination could likely be the predominate pathway.

In 2023, Ji and Liu expanded on the Pd-catalyzed three-component arylalkynylation of oxabenzonorbornadiene derivatives (Scheme 43) [94]. Initially reported by Liu and co-workers in 2006 [95], present conditions were altered to avoid aqueous NaOH, opting for Cs₂CO₃. Interestingly, the reaction was applicable to a variety of functional groups, including esters, chlorines, and bromines. In line with similar reports, the scope of bicyclic alkenes was limited with all but a single example being performed on norbornene. DFT calculations were used to explore the reaction mechanism which involves the oxidative addition of the C–I bond, coordination, migratory insertion, transmetalation, and reductive elimination. The authors determined the migratory insertion via TS_232–233 is the rate-determining step for the catalytic cycle.

In 2014, Ma and Wang reported the Pd-catalyzed three-component coupling of N-tosylhydrazones, aryl halides, and norbornene (Scheme 44) [96]. The reaction tolerated small substituents on the N-tosylhydrazone and aryl halide coupling partners, but the reaction was quite sensitive to ortho-substitution and steric bulk. Generally, the reaction gave the corresponding product in good yield and excellent diastereoselectivity; however, a few substrates produced diastereomeric ratios of 3:1. As the propensity for an exo-selective migratory insertion is well understood, it is surprising some products displayed such poor selectivity. As such, this may indicate some form of stereoisomerization rather than a poorly selective migratory insertion. In the following year, Xu and Liang reported a reaction involving the same three coupling partners [97]. By altering the reaction conditions, the authors observed the first palladium-catalyzed ring opening of norbornene to prepare methylenecyclopentane derivatives via an unusual β-carbon elimination.

In 2016, the Song laboratory reported the Pd-catalyzed arylboration of norbornene derivatives (Scheme 45) [98]. Generally, electron-rich aryl halides afforded the product in a higher yield than those bearing electron-withdrawing groups. Moreover, the reaction was amenable to heteroaromatic iodides, but yields were diminished. The authors showed aryl bromides were tolerated albeit with slightly diminished yields relative to their iodide-containing counterparts. The scope of bicyclic alkenes was mainly limited to norbornene with a single example using norbornadiene.

In 2019, Jiang and co-workers expanded on this chemistry and revealed allyl carboxylates can be used as the nucleophilic partner in carboborylation difunctionalization reactions (Scheme 45) [99]. Besides allyl acetates, the authors revealed formates, propionates, and butanoates were able to afford the desired product; however, allyl bromides and chlorides failed. Unfortunately, the reaction was sensitive to the bicyclic alkene used; norbornadiene and 2,3-diester-substituted norbornene were unable to undergo the transformation. Surprisingly, oxabenzonorbornadiene was amenable and afforded the difunctionalized product in 44% yield rather than a ring-opened product.

In 2018, Fu and Chen reported the Pd-catalyzed, three-component annulation of aryl iodides 66, alkenyl bromides 241, and bicyclic alkenes 1 (Scheme 46) [100]. Similar reports by the Lautens [101] and Perumal [102] laboratories have demonstrated the use of norbornene derivatives for the synthesis of tetrasubstituted olefins; however, limited work has been done for the synthesis of trisubstituted olefins. The authors noted ortho-substituted iodobenzenes delivered products in a greater yield compared to their strictly meta- or para-substituted counterparts like due to the elimination of complex byproducts. Typically, reactions gave products with very high Z stereoselectivity. The authors demonstrated the methodology could be applied towards the synthesis of tetrasubstituted olefins as well, giving the desired product in moderate to good yields. This methodology avoided the use of highly substituted internal alkynes, a substrate which can be more difficult to synthesis than its alkenyl bromide counterpart. The reaction is applicable to other bicyclic alkenes although with slightly diminished yields compared to norbornene. Unsymmetrically substituted bicyclic alkenes bearing relatively sensitive functionalities, such as -CHO and -CN, worked, albeit with no regioselectivity.

In 2019, Perumal and Cho reported a Pd-catalyzed double insertion/annulation reaction for synthesizing tetrasubstituted olefins (Scheme 47) [103]. Mechanistically, the transformation involves the formation of four new C–C bonds through three consecutive carbopalladations and a C–H activation. Unlike the anticipated exo migratory insertion seen almost exclusively in these types of systems, the authors noted the second norbornene moiety undergoes an endo-selective reaction, as confirmed through X-ray crystallography. The diastereoselectivity of the reaction was limited due to the production of the unanticipated exo-endo adduct 244/247 which was formed as the major product while the thermodynamically more stable exo-exo product 244/248 was only ever achieved in minor yields. The methodology was explored with a broad scope of aryl substituents revealing the robustness of the reaction. Additionally, heteroaromatic alkynes 246 were found to be tolerable but gave slightly diminished yields.

In 2013, van Vranken and co-workers reported the Pd-catalyzed aminocyclopropanation of bicyclic alkenes 1 with 5-iodopent-4-enylamine derivatives 249 (Scheme 48) [104]. The reaction was effective for a range of N-substituted derivatives 249; however, the reaction was sensitive to steric bulk. With large groups, like N-adamantyl, only modest yields of the desired pyrrolidine product were obtained, owing to the formation of the vinylcyclopropane side product. Other bicyclic alkenes were amenable, including an example with an oxabicyclic alkene which underwent the desired reaction rather than the anticipated β-oxygen elimination side reaction. The mechanism for this transformation involves the oxidative addition of the alkenyl iodide to the Pd(0) and side-on coordination to the olefin 252, followed by the migratory insertion of the bicyclic alkene to afford complex 253. Aminopalladation of the olefin affords 254 which undergoes a reductive elimination to generate the final product 250. In the case of the vinylcyclopropane side product, complex 253 preferentially undergoes a carbopalladation to generate a cyclopropane intermediate 255 which undergoes a β-hydride elimination to give 256.

In 2017, Wu and Jiang reported a Pd-catalyzed, three-component coupling of alkynyl bromides 62, norbornene derivatives 15, with electrophilic trapping agents (Scheme 49) [105]. Mechanistically, the transformation begins with the oxidative addition of the alkynyl bromide to the Pd(0) catalyst. From here, four consecutive carbopalladation reactions ultimately end up producing an alkylpalladium intermediate which undergoes a β-carbon elimination to afford a Pd–π-allyl species. First, the authors captured this π-allyl species with N-tosylhydrazone derivatives 236. Notably, alkynyl bromides 62 bearing electron-withdrawing groups typically afforded the desired product in greater yield. The scope of the N-tosylhydrazones 236 was expansive with electronic substituents having little effect on the reaction. Heteroaromatic N-tosylhydrazones 236 were applicable but gave diminished yields. Moving on, the authors showed the Pd–π-allyl species can be trapped with boronic acids 20. Like the N-tosylhydrazones 236, the substituents on the boronic acid had little effect on the reaction. Lastly, the authors demonstrated the use of B₂pin₂ 53 to capture the Pd–π-allyl species.

In 2006, the Cheng group investigated the Pd-catalyzed intramolecular cyclization/ring-opening reaction of heterobicyclic alkenes 30 with 2-iodophenoxyallenes 260 (Scheme 50) [106]. Surprisingly, the efficacy of the reaction was more susceptible to derivatization of the benzo-fused moiety with sterically demanding functionalities rather than altering the electronics, as seen with severely diminished yields with phenanthrene-fused oxabicyclic alkenes. The reaction was unaffected by the identity of the bridging heteroatom with both oxa- and aza-bridging atoms performing equally as well; although, the latter was only explored a single time. Altering the tether length of the allene moiety seemed to mildly affect the reaction with 5-membered rings being formed in slightly greater yields compared to their 6-membered counterparts. Mechanistically, this reaction operates similarly to other cyclization/capture chemistry seminally presented by Griggs [107,108]. First, the Pd(II) catalyst is reduced to the Pd(0) active catalyst with Zn metal. Oxidative addition of the aryl iodide 260a to Pd(0) gives 262 which can side-on coordinate with the allenyl group. Intramolecular migratory insertion affords the Pd–π-allyl species 263 which can side-on coordinate to the exo face of the bicyclic alkene 264. Rather than dissociation of the iodide ligand to generate a cationic Pd center, it has also been proposed the loss of a phosphine ligand could allow for the generation of a free coordination site. Migratory insertion affords intermediate 265 which undergoes a β-oxygen elimination to 266. Transmetalation with ZnCl₂ affords the zinc alkoxide 267 which is hydrolyzed to give the final product 261a. Alternatively, Zn metal could reduce Pd(II) intermediate 266 to Pd(0) directly, bypassing the transmetalation step.

In 2019, the Fan group explored the Pd-catalyzed dimer- and trimerization of oxabenzonorbornadiene derivatives 30 with anhydrides 268 (Scheme 51) [109]. The authors noted electron-deficient oxabenzonorbornadiene derivatives resulted in diminished product yields. When electron-rich bicyclic alkenes were used, the trimer 270 to dimer 269 ratio was increased. When applied to unsymmetrically substituted bicyclic alkenes, the authors propose the dimerized product was formed as a single regioisomer, as evaluated by ¹H NMR, with no trimerization observed.

In 2013, the Hu laboratory developed a method to form annulated xanthones 272 using norbornadiene (15b), 3-iodochromones 271, and aryl iodides 66 via a Catellani-type reaction (Scheme 52) [110]. The authors proposed a mechanism beginning with the oxidative addition of Pd(0) to 271a, followed by migratory insertion across norbornadiene (15b) and alkenyl C–H activation of the chromone ring, furnishing the palladacycle 274. The oxidative addition of 274 to the aryl iodide 66b yields a Pd(IV) species 275 that can undergo reductive elimination by either an sp²–sp³ or sp²–sp² coupling event. The authors probed the regioselectivity of this step using p-iodotoluene and, based on the product, concluded that only sp²–sp³ coupling occurred. The resulting intermediate 276 undergoes an aryl C–H activation step and a subsequent reductive elimination yielding a norbornadiene-fused xanthone derivative 277, which forms the final product 272a via a retro-Diels–Alder reaction. The reaction was generally tolerant of substituted 3-iodochromones; however, substituted aryl iodides were shown to have significant effects. Electron-donating para-substituents and bulky ortho-substituents resulted in lower yields while bulky meta-substituents could be used to influence the regioselectivity of the C–H activation step. The scope was limited to dienes because of the necessity for a retro-Diels–Alder to furnish the desired product but a norbornane-fused xanthone was also produced in 82% yield.

In 2017, Vijayan et al. investigated both the Pd-catalyzed hydroarylation and annulation of urea-derived bicyclic alkenes 158 using aryl iodides 66 (Scheme 53) [111]. In both reactions, the 1,2-migratory insertion of palladium across the olefin resulted in a palladacycle intermediate that was further reacted, either via hydride-donor or an ortho-directing group, to form the hydroarylated 280 or annulated products 282, respectively. For this reason, the formic acid additive was necessary in the hydroarylation but was left out in the annulation to promote capture by the phenolic directing group. The hydroarylation gave moderate to good yields with EWGs and EDGs alike, as well as accommodating ortho-substituents. It was also tolerant of spiro-, furan-derived, and N-protected pyrrole-derived bicyclic alkenes, all giving similar yields. The heterobicyclic alkenes were shown to be compatible with the annulation as well, though they resulted in slightly reduced yields compared to the carbocyclic examples. Although the authors focused on the use of an alcohol directing group for the annulation to furnish dihydrobenzofurans, they also provided a simple example using methyl bromide and nitrile directing groups giving indane and indanone products in similar yields.

In 2018, the Chen laboratory explored a Pd/norbornene cocatalyzed Catellani annulation reaction of phenyl iodides 66 and NH-sulfoximines 283 in an attempt to produce dibenzothiazines [112]. Though they were successful in this effort, they also reported accessing eight-membered sulfoximine heterocycles when norbornene was not extruded, which was accomplished in two distinct ways (Scheme 54) [112]. The first requires aryl iodides with meta-EWGs, which was shown by DFT calculation to favor sp²–sp³ coupling over sp²–sp² coupling. This coupling step prevents the extrusion of norbornane later without restricting the Pd catalyst’s access to the sulfoximine directing group, thus allowing the formation of the 8-membered heterocyclic product 284. The other method requires slightly modified conditions, mainly by increasing the equivalents of NH-sulfoximines 283, and for the phenyl iodides have two ortho-hydrogens. The second hydrogen allows for sequential C–H activation after the standard sp²–sp² coupling, again preventing the extrusion of norbornene, and creating a Pd(II) species that undergoes oxidative addition with the extra sulfoximines provided, eventually forming a heterocycle bearing two sulfoximine moieties 285. Understandably, the presented examples are limited, as these products were of secondary interest to the authors but yields of up to 94% for product 284 and from 42% to 64% for product 285 were reported.

Six years after the work of Hu and co-workers producing annulated xanthones, Yang et al. completed a 2,2-bifunctionalization of bicyclic alkenes 1 to produce xanthone derivatives bearing spirobicyclic moieties 277 (Scheme 55) [113]. This was achieved via a Pd-catalyzed [2 + 3 + 1] annulation of 3-iodochromones 271, bromoacetones 276, and bicyclic alkenes 1. The reaction generally afforded good yield and diastereoselectivity even across the wide swathe of functionalized substrates and few bicyclic alkenes tested and provided a good yield (71%) at the gram scale.

In 2019, Zhong et al. reported a method to produce phenanthrene derivatives 279 using 2-iodotoluenes 66, o-bromobenzoyl chlorides 278, and norbornadiene (15b). This method proceeds through a Pd-catalyzed Catellani reaction; however, norbornadiene extrusion is avoided via the loss of the acyl chloride group as CO, allowing the formation of a 7-membered palladacycle, reductive elimination, and subsequent retro-Diels–Alder reaction to the phenanthrene 279 (Scheme 56) [114]. This was an improvement over past methods that used less reactive ortho-haloaryl carboxylic acids which required harsher conditions and longer reaction times to optimally perform. The reaction was shown to be tolerant of diverse functionality, providing excellent yields barring a couple notable examples; 1-iodonaphthalene (64%) and o-iodonitrobenzene (75%). The authors were also able to demonstrate its efficacy at the gram scale with a yield of 88%.

In 2020, Zhang and colleagues explored a three-component Pd-catalyzed annulation reaction furnishing norbornane-fused indanes 281 (Scheme 57) [115]. This reaction sees an aryl iodide 66 coupled to a bicyclic alkene 30 to produce a 5-membered palladacycle intermediate that is then captured by the third reagent, either methylene bromide (280) or an α-diazoester 282. A reduced yield was seen in the absence of iPrOH, so it was kept as an additive with the authors proposing it functions as a reductant, reducing Pd(II) to the active catalyst Pd(0). A great variety of examples using methylene bromide (280) were reported, including using a few different bicyclic alkenes 30, with up to 96% yield. A similar variety of examples with similar yields were shown using α-diazoesters 282, however, only norbornene proved suitable in this case with heterobicyclic alkenes unable to afford the desired product.

In 2009, the Radhakrishnan laboratory investigated a Pd-catalyzed annulation of diazabicyclic alkenes 130a, and 2-iodophenols 284 or 2-iodoaniline (286) towards fused benzofuran 285 or indole 287 products (Scheme 58) [116]. The reaction begins with the oxidative addition of Pd(0) into the aryl iodide 284a, followed by migratory insertion across the bicyclic alkene to form 289. Base-assisted addition of the alcohol and β-nitrogen elimination forms a ring-opened cyclopentene intermediate 290 which then undergoes oxypalladation and β-hydride elimination, furnishing the benzofuran product 284a. The authors noted that in the absence of the Bu₄NCl additive the reaction did not work. The authors hypothesized the chloride ions are important for regenerating and stabilizing the Pd(0) species. While only a handful of examples were reported, it was demonstrated that diazabicyclic alkenes with bulkier ester groups caused reduced yields.

One year later, the Gilbertson laboratory expanded on this annulation reaction, increasing its efficiency and significantly decreasing the reaction time using tweaked conditions and microwave irradiation (Scheme 59) [117]. They also significantly increased the scope of the reaction, providing many examples with up to 98% yield, and utilizing N-substituted anilines to create N-substituted indoles 284. The authors were also able to apply their methodology to an acetal-protected vanillin derivative, producing the corresponding benzofuran with 90% yield.

Two years after their previous work, the Radhakrishnan group explored a non-ring-opening annulation utilizing 2-iodobenzonitrile (292) and 2-formylphenylboronic acids 142 to access diazabicyclic-fused indanones 293 and indanols 294 (Scheme 60) [71]. The authors noted the addition of base increased the yield of 2-iodobenzonitrile (292) reactions but reduced it for those with 2-formylphenylboronic acids 142. Only a few examples producing indanones 293 were presented showing very small changes in yield with different diazabicyclic esters. Different N-substituted triazolinedione-derived bicyclic alkenes were also tested but failed, likely due to their base sensitivity. The annulation reaction yielding indanols 294 was seen to produce the 3,4-disubstituted cyclopentene 295 in ratios of about 1:9 when the diazabicyclic alkenes 130a were used. However, when using the N-substituted triazolinedione-derived bicyclic alkenes the 3,4-disubstituted cyclopentene 294 could be produced exclusively in yields of up to 90%.

In 2013, Pihko and Radhakrishnan revisited their 2009 annulation reaction using 2-iodophenols 284 and 2-iodoaniline (286) in an attempt to access larger polycyclic compounds 296 through the use of spirotricyclic olefins 176 (Scheme 61) [118]. It is proposed that the reaction follows a similar ring-opening/ring-closing mechanism to their 2009 report (Scheme 58), but the cyclopropane moiety allows a second ring opening and the subsequent generation of a π-allyl–palladium complex. This complex undergoes an intramolecular nucleophilic attack by hydrazine, forming the fourth fused ring. When the methodology was applied to 2-iodoaniline (286), the anticipated polycyclic product was not formed; instead, trans-disubstituted spiro[2,4]hept-4-enes 297 were formed. A variety of substituted 2-iodophenols 284 were tested showing significantly reduced yields with para-EWGs, emphasizing the importance of an electron-rich alcohol directing group.

In 2017, the Radhakrishnan group investigated another ring-opening/ring-closing reaction of diazabicyclic alkenes 130a, synthesizing cyclopenta[b]pyrroline derivatives 299 using aromatic enamides 298 (Scheme 62) [119]. Since the reaction begins with an alkenyl C–H activation, forming a 6-membered palladacycle intermediate with amide oxygen chelation, Cu(OAc)₂ was added as an oxidant to regenerate Pd(II). Afterwards, the transformation progresses similarly to their 2009 report (Scheme 58). The 6-membered palladacycle will undergo migratory insertion into the diazabicyclic alkene 130a which after a β-nitrogen elimination, adds to the amide via the nitrogen atom. Aminopalladation forms the C–N bond that produces the fused pyrroline moiety in the product 299. The authors reported several examples using substituted aromatic enamides finding that EWGs were well tolerated while EDGs significantly reduced yields. When applied at the gram scale, the desired product was produced with 60% yield.

In 2018, Radhakrishnan and colleagues again expanded on their past work, attempting to produce 3,4,5-trisubstituted cyclopentenes 300 from diazabicyclic alkenes 130a and 2-iodobenzoates 9 (Scheme 63) [120]. The authors proposed a charged fused-oxane intermediate is produced after the ring-opening/ring-closing sequence, as anticipated in 2009 (Scheme 58), whose eventual breakdown furnishes a π-allyl–palladium complex which undergoes nucleophilic attack by the acetate or azide anion. Several examples were reported, ranging a 60–85% yield, showing minimal electronic influence by 2-iodobenzoate substituents. However, another reaction path was observed when 2-iodo-3-methylbenzoate (9a) was used, producing a cyclopentene-fused indane 302. The authors suggested that the mechanism of this reaction follows the same steps until the formation of the π-allyl–palladium complex, which can undergo cyclopalladation via benzylic C–H activation of the 3-methyl group, and subsequently reductive elimination to yield the fused indane product 302.

In 2012, Ge et al. investigated a palladacycle-catalyzed reaction furnishing highly substituted fused furans 304 using bicyclic alkenes 1 and terminal ynones 302 (Scheme 64) [121]. The authors noted the reaction was sensitive to the identity of both basic and acidic additives, as bases tended to slow down reactions while stronger acids typically produced only a trace amount of the desired product 304. Eventually, the authors discovered their goldilocks additive, settling on p-methoxybenzoic acid which showed a significant increase in yield. The reaction was generally tolerant of a variety of substituted terminal ynones 303 and bicyclic alkenes 1, as well as norbornene (15a) and norbornadiene (15b). Two years later, this methodology was expanded by the same group, using terminal alkynyl imines 305 to access polycyclic 5H-benzo[b]azepines 306 (Scheme 64) [122]. The authors reported low yields when R² or R³ were weak EWGs and no reaction with strong EWGs at R³, somewhat restricting the scope of the reaction.

In 2018, the Jiang laboratory explored a Pd-catalyzed carboesterification reaction, using bicyclic alkenes 15 and alkynoates 307, ynamides 309, and alkynols 310 to produce α-methylene γ-lactone 308 and tetrahydrofuran derivatives 311 (Scheme 65) [123]. The reaction was shown to be functionally tolerant, boasting a large number of high yielding examples. Largely, the authors noted substitution of the ester or the amide moiety had little influence on the reaction. Only two examples were reported for the reaction of the alkynol 310, albeit in good yields.

Conclusion

Over the last two decades, there has been remarkable progress in transition-metal-catalyzed domino reactions of homo- and heterobicyclic alkenes. Bicyclic alkenes can be exploited in two ways. Firstly, through the release of ring-strain energy which drives the reaction forward under milder conditions compared to strainless alkene counterparts. Secondly, the stereochemically well-defined, dual-faced nature of these systems can be exploited to synthesize highly stereoselective products.

Multicomponent domino reactions can be challenging due to selectivity issues, but recent advancements have provided straightforward protocols for the construction of complex molecules with multiple carbon–carbon and carbon–heteroatom bonds in a single step. When participating in a well-orchestrated domino sequence, these bicyclic alkenes can quickly generate highly functionalized products with extreme stereo-, regio-, and enantioselectivity.

Currently, a majority of transition-metal-catalyzed domino reactions use simple carbocyclic alkenes, such as norbornene, as the propagative π-system of choice, limiting its relevance. To see further advancements in this field, it is necessary to expand the scope to include more heterobicyclic alkenes and understand their fundamental reactivity. As heterobicyclic alkenes have the tendency to undergo some form a β-heteroatom elimination which can prematurely terminate a cascade, their use requires more thought. However, altering coupling partners, reaction conditions, and the metal center have all been used to promote difunctionalization of heterobicyclic alkenes while hindering β-heteroatom elimination. We hope this comprehensive overview of bicyclic alkene chemistry will drive further advancements in the area of transition-metal-catalyzed domino reactions.