Cyclization Strategies in Carbonyl–Olefin Metathesis: An Up-to-Date Review

Abstract

The metathesis reaction between carbonyl compounds and olefins has emerged as a potent strategy for facilitating swift functional group interconversion and the construction of intricate organic structures through the creation of novel carbon–carbon double bonds. To date, significant progress has been made in carbonyl–olefin metathesis reactions, where oxetane, pyrazolidine, 1,3-diol, and metal alkylidene have been proved to be key intermediates. Recently, several reviews have been disclosed, focusing on distinct catalytic approaches for achieving carbonyl–olefin metathesis. However, the summarization of cyclization strategies for constructing aromatic heterocyclic frameworks through carbonyl–olefin metathesis reactions has rarely been reported. Consequently, we present an up-to-date review of the cyclization strategies in carbonyl–olefin metathesis, categorizing them into three main groups: the formation of monocyclic compounds, bicyclic compounds, and polycyclic compounds. This review delves into the underlying mechanism, scope, and applications, offering a comprehensive perspective on the current strength and the limitation of this field.

1. Introduction

Since the groundbreaking awarding of the Nobel Prize in Chemistry to olefin me-tathesis and the palladium-catalyzed cross-coupling reaction in 2005 and 2010, the field of carbon–carbon bond formation has witnessed remarkable advancements [1]. Notably, ole-fin metathesis has found widespread applications across pharmaceutical, agricultural, materials science, and petroleum industries [2,3,4,5,6,7,8,9,10,11]. This includes the synthesis of intricate natural products such as ingenol, (–)-mucocin, coleophomones B, ciguatoxin CTX3C, and amphidinolide A, achieved through ring-closing metathesis macrocyclization [12,13,14,15,16]. Additionally, cross-metathesis degradation and ring-opening metathesis polymerization techniques hold promise for functional rubber synthesis [17,18,19,20,21,22]. In contrast, the carbonyl–olefin metathesis strategy has lagged behind, hindered by the need for stoichiometric amounts of transition metal reagents, photochemical promotion, or a substrate prone to cationic cyclization. Despite these challenges, innovative carbonyl–olefin metathesis protocols have emerged as valuable complements to olefin metathesis, enabling the synthesis of conjugate polymers, biologically active molecules, and complex natural products [23,24,25,26,27,28,29,30,31]. For instance, chain alkenes and enals can be generated through an intermolecular car-bonyl–olefin exchange [32,33,34,35,36,37], while intramolecular reactions facilitate the synthesis of conjugate polymers and natural products [38,39,40,41]. These diverse methodologies can be categorized based on their reaction intermediates: (a) carbonyl–olefin metathesis proceeds via oxetane intermediates (Figure 1a), involving the Paternò–Büchi reaction between alkene and carbonyl under UV light, leading to oxetanes that fragment under acidic condition or high temperature to yield a new alkene and carbonyl [42,43]; (b) carbonyl–olefin metathesis proceeds via pyrazolidine intermediates (Figure 1b), achieved through the 1,3-dipolar cycloaddition of an alkene with azomethine imine generated from hydrazine and an aldehyde, followed by cycloreversion to produce new olefin and ketone [26]; (c) carbonyl–olefin metathesis proceeds via 1,3-diol intermediates (Figure 1c), where the intermolecular addition of a C–O double bond to an alkene-derived radical, prepared via a photoinitiated hole catalytic cycle, forms a reactive 1,3-diol that undergoes Grob fragmentation to yield olefin and ketone [44]; and (d) carbonyl–olefin metathesis proceeds via a metal alkylidene intermediate (Figure 1d), where the olefination of a carbonyl is facilitated by metal alkylidene, with subsequent olefin–olefin metathesis or a direct carbonyl olefination reaction producing the desired product [45,46]. While several recent reviews have focused on different catalytic approaches for the carbonyl–olefin metathesis proceeds via the intermediate oxetane metathesis reaction and complex molecule synthesis [26,27,28,30], a comprehensive overview of the cyclization strategies within this field remains limited.

In this up-to-date review, we summarize the cyclization strategies in carbonyl–olefin metathesis, dividing them into three categories (Figure 2): (1) the formation of monocyclic compounds; (2) the formation of bicyclic compounds; and (3) the formation of polycyclic compounds.

2. The Formation of Monocyclic Compounds

This section includes two major parts: (1) a carbonyl–olefin metathesis cyclization reaction catalyzed by metal; and (2) a carbonyl–olefin metathesis cyclization reaction catalyzed by metal-free condition.

2.1. Carbonyl–Olefin Metathesis Cyclization Reaction Catalyzed by Metal

The earliest reported instance of a carbonyl–olefin metathesis cyclization reaction, catalyzed by a metal to yield five-membered ring scaffolds, dates back to Grubbs’ work in 1993 [47]. When an olefinic ketone 17 served as the substrate, intramolecular cyclization between the olefin and carbonyl functional group occurred, with the aid of stoichiometric amounts of molybdenum alkylidene 18, yielding the cycloalkene 19 using benzene as a solvent (Scheme 1). This approach showcased a versatile substrate scope, enabling the synthesis of diverse unsaturated five-, six-, and seven-membered rings (19a–19c) in good yields. Notably, the carbonyl–olefin metathesis of olefinic ketone 17, featuring α-oxygenation, leads to the formation of cyclic ketone 19d via the isomerization of the cyclic enol ether intermediate, facilitated by 20 (1.0 equiv). Interestingly, 19e bearing an ester functional group could be obtained from functionalized diene as the starting material, utilizing 2 mol% of 18. Drawing from previous studies [48,49], the author hypothesized that the olefin metathesis between olefinic ketone 17 and 18 generated a new alkylidene intermediate, Int 2, containing a carbonyl group, which subsequently underwent intramolecular carbonyl olefination, mediated by metal quaternary ring intermediate Int 3, to produce cycloalkene 19.

In 2007, Rainier’s team elaborated on the exceptional and unprecedented reactivity demonstrated by an in situ-produced reduced titanium ethylidene reagent, which was found to interact with carbonyl functional groups in a unique manner (Scheme 2) [50]. This interaction led to the formation of cyclic enol ether 22 via a ring-closing metathesis reaction, marking a significant advancement in the field. This approach efficiently pro-vided the desired products 22a–22d in moderate yields through a two-step metathesis sequence. Additionally, the gambieric acid A derivative 22e was achieved with a 50% yield via this transformation (Scheme 2c). Specifically, when using dibromomethane as the alkylidene source, the resulting titanium methylidene reagent exclusively produced acyclic enol ether 23.

Inspired by previous work, Roy and colleagues developed a Grubbs–II 25-catalyzed ring-closing carbonyl–olefin metathesis and ring-closing olefin metathesis protocol for the synthesis of unsaturated carbocyclic compounds 26 and 27 (Scheme 3) [51]. When n = 3 in 24, the carbonyl–olefin metathesis product 26a and olefin metathesis product 27a were observed. When n = 4 in 24, only the ring-closing carbonyl–olefin metathesis product 26b was formed in a moderate yield of 20%. In addition, for 24 with n < 4, the olefin metathesis proceeded efficiently using 10 mol% Grubbs-II 25 under reflux conditions, providing a series of bicyclic frameworks, including five–seven (27b) and five–six (27c), in high yields. Further experimental studies have revealed that increasing the loading of 25 favored the ring-closing carbonyl–olefin metathesis of 26.

Later, Duñach’s group reported a Lewis acid-catalyzed intramolecular carbonyl–olefin ring-closing process of phenyl ketone 28, giving cyclohexene 29 (Scheme 4) [52]. Further ESI-MS analysis revealed that the dehydration of the tertiary alcohol Int 4 served as the primary side reaction, contributing to the reduced yield of 29. Conditional screening experiments have demonstrated that Fe(NTf₂)₂ has potential for promoting the metathesis of carbonyl olefin.

Despite the well-established metal-catalyzed carbonyl olefin metathesis reaction, its application has been hindered by the necessity of stoichiometric transition metals or harsh reaction conditions [24,53]. In 2016, Schindler’s group overcame this limitation by demonstrating a catalytic FeCl₃-mediated carbonyl–olefin ring-closing metathesis reaction that efficiently prepared unsaturated carbocyclic olefin in moderate to excellent yields (Scheme 5) [54]. This process used aromatic β-ketoester 30 with a pendant isoprenyl group as a substrate, under mild conditions, facilitating advances in olefin metathesis research. The reaction exhibited high functional group tolerance and a wide substrate scope, ena-bling the conversion of variously substituted aromatic β-ketoesters into the cyclopentene product. Notably, various β-substitutions, including a benzyl ester, ketone, sulfonyl, and amide, were compatible, providing cyclopentenes 31a–31h with up to 87% efficiency. For methyl or p-phenyl methoxy substitution, only thermodynamically stable product 31i was observed. Remarkably, spirocyclic motif 31h, six-membered carbocycle 31k, indene 31l, 1,2-dihydronaphthalene 31m, and tricyclic 31n could also be efficiently synthesized through this carbonyl–olefin metathesis. Based on their findings, Schindler and colleagues proposed two potential mechanisms: a concerted mechanism involving [2+2] cycloaddition and cycloreversion (Scheme 5c) and a carbocation mechanism featuring nucleophilic attack and carbocation trapping (Scheme 5d). Further experiments revealed the absence of a carbocation intermediate, supporting the concerted mechanism (Scheme 5e).

Concurrently, Li and co-workers disclosed an efficient FeCl₃-catalyzed ring-closing carbonyl–olefin metathesis using 1,2-dichloroethane (DCE) as a solvent (Scheme 6) [55]. This method could provide cyclopentene 37a, cyclohexene 37d, tetrahydropyridine 37e, 2,5-dihydro-pyrrole 37f, indene 37g, 1,4-dihydronaphthalene 37h, and 2,5-dihydro-1H-pyrrole 37i, with good yields using 1–20 mol% FeCl₃. Notably, allyl trimethyl silane facilitated key oxetane intermediate formation and benzaldehyde scavenging, which are crucial for the synthesis of five-membered nitrogen heterocycle 37f. The presence of minor side product 37k led to the proposal of a stepwise mechanism (Scheme 6c) [56,57], involving electrophilic cyclization, [2+2] cycloreversion, and FeCl₃ regeneration.

Inspired by previous work, Schindler’s group devised a synthetic route to 3-aryl-2,5-dihydropyrroles via the carbonyl–olefin metathesis of secondary amine 39 (Scheme 7) [58]. This successful transformation was facilitated by the use of 0.5 eq FeCl₃, which inhibited an efficient turnover and attenuated the Lewis basicity of the sulfonamide moiety, thereby overcoming substrate-induced FeCl₃ inhibition. However, the unique reaction pathway and stepwise formation of a carbocation intermediate resulted in low yields of the desired product from prenyl-derived alkene substrate 39 [59]. Notably, the expected product could not be observed using 39 with crotyl alkene or terminal alkene as the starting material. Nevertheless, under optimized conditions, a diverse range of 3-pyrroline derivatives incorporating various substituents (40a–40g) were successfully constructed. Notably, the synthesis of 40g required the aid of allyl trimethyl silane (5.0 eq). Additionally, substrate 39 containing a methylene substituent in the α-position facilitated the formation of the oxetane intermediate. Based on these findings, the authors proposed that the oxygen atom of the sulfonamide and carbonyl group competitively bind to ferric chloride, generating intermediates Int 14 and Int 15. Then, Int 15 served as a key precursor for promoting the formation of oxetane Int 16. Subsequent theoretical investigations demonstrated that an electron-deficient sulfonamide moiety would be beneficial for enabling the turnover of FeCl₃. Furthermore, under mild conditions, the t-butyloxy carbonyl (Boc)-protected derivative of compound 42, along with pyrrolidine-3-one 44, could be efficiently constructed.

In 2019, the same group expanded their work by disclosing a Lewis acidic super-electrophile-mediated carbonyl–olefin metathesis of aliphatic ketone 45 (Scheme 8) [60]. Specifically, allyltrimethylsilane was the most promising additive for restricting this process as a competing product inhibition of the aldehyde by-product was formed, which enabled the carbonyl–olefin metathesis of aliphatic ketones bearing styrene. Diverse substrates, including electron-rich and -deficient aryl, thiophene, cyclopropyl, alkene, and β-aliphatic moieties, were tolerated, giving the expected products 46a–46f. However, β-ketoester containing aliphatic ketone showed low reactivity (46f–46g). Interestingly, the oxetane compound 46h could be isolated in 52% yield. Combining DFT calculations, kinetic studies, Raman spectroscopy, and IR data, the researcher proposed a mechanism involving FeCl₃ coordination, Lewis acid super-electrophile generation, substrate polarization, and fragmentation to form cyclopentene 46a. Firstly, Int 17 was afforded by the coordinate of FeCl₃ with 45a. Subsequently, a second equivalent of FeCl₃ was used for the generation of stronger Lewis acid which functioned as a super-electrophile capable of lowering the energy of the transition state to provide sufficient activation for carbonyl–olefin metathesis. Ultimately, the fragmentation of Int 19 formed by substrate polarization resulted in cyclopentene 46a.

In the same year, it was reported that polycyclic substrate 48 underwent a transan-nular carbonyl–olefin metathesis process, catalyzed by FeCl₃, to afford multi-substituted cyclopentadiene derivative 49 (Scheme 9) [29]. This conversion was effective for a variety of 9- and 10-membered ring systems, affording the desired metathesis products 49a–49e in moderate yields. Notably, both carbonyl-ene and carbonyl–olefin metathesis pathways competed under FeCl₃ catalysis, with metathesis prevailing as the thermodynamic product. Importantly, this transformation revealed that either a transannular carbonyl-ene product or carbonyl–olefin metathesis product could be furnished, utilizing a distinct Lewis acid catalyst [61]. Further mechanistic studies indicated that the retro-[2+2]-cycloaddition of Int 21, formed via a unique reaction pathway involving Int 20, offered the expected transannular carbonyl–olefin metathesis product 49e.

In 2020, a FeCl₃-mediated carbonyl–olefin metathesis utilizing amino acid 51 as a chiral pool reagent was employed for the synthesis of substituted tetrahydropyridine 52, accompanied by the formation of by-product ketone 53 (Scheme 10) [62]. Enhanced yields of 52 were realized when using electron-deficient sulfonamide-bearing amino acids 52 as substrates (52b, 52c). The experiments demonstrated that aryl ketone 51 with a prenyl sub-stituent (51a) performed optimally, and amino acids featuring alanine, thienyl alanine, and thienyl ketone were well tolerated. Notably, the challenging target compound 52a was successfully obtained from unsubstituted glycine-derived aryl ketone 51. Further-more, 1,2,3,6-tetrahydropyridine 55 was accessible from 52b through deprotection and exposure to Boc₂O at 50 °C.

In 2019, Bour, Gandon, and their teams developed a series of tandem carbonyl–olefin metathesis and transfer hydrogenation protocols, catalyzed by [IPr·GaCl₂] [SbF₆] and uti-lizing 1,4-cyclohexadiene (1,4-CHD) as a H₂ surrogate, for the synthesis of 1,2-cis-disubstituted cyclopentanes and various cyclohexanes (Scheme 11) [63]. This method allowed for the production of various cyclopentanes (57a, 57b, 57g, and 57h), tricyclic compound 57d, and cyclohexane 57i. Notably, a key advancement was the ability to transform ketone 56 devoid of β-substituents into the desired products, contrasting with previous work [54]. While tricyclic 57e was not observed under optimized conditions, the metathesis product of 57e was isolated using 1,2-dichloroethane (1,2-DCE) as a solvent. Aliphatic ketones failed to participate in this transformation (57f). Building on prior research [54,64,65], it was proposed that the metathesis product Int 23 originated from [2+2] cycloadduct Int 22, proceeding through two distinct pathways to yield 57j, either via the hydride transfer of carbocation Int 24 and proto-degallation of Int 25 or through the activation of Int 28 by a homodimeric Ga-complex Int 27 coordinated by [IPr·GaCl₂][SbF₆] and 1,4-CHD. Finally, the DFT calculations demonstrated path 2 was an advantageous pathway.

Shortly thereafter, Schindler and co-workers disclosed a highly efficient methodology for constructing five-, six-, and seven-membered rings via an aluminum (III)-ion pair-promoted carbonyl–olefin ring-closing metathesis reaction (Scheme 12) [66]. In-depth investigations revealed several key findings: (1) the heterobimetallic ion pairs [AlCl₂]⁺[SbF₆]^–, formed by AlCl₃ and AgSbF₆, served as the critical active catalytic species; (2) the unexpected such as benzaldehyde and acetophenone hindered the reaction’s progress; (3) aryl ketones (59) containing an oxygen atom in the carbon chain or a β-hydrogen atom were well tolerated (60b, 60e, and 60f); (4) a previously inaccessible seven-membered ring (60g) was prepared by this method; and (5) no products stemming from trapped carbocation intermediates were detected through isolation or high-resolution mass spectrometry. Finally, DFT calculations suggested two plausible pathways for the formation of 60h, involving either a Lewis acid–base complex-mediated cyclization followed by a retro-[2+2] cycloaddition or an intramolecular addition of a terminal alkene via an oxetane intermediate.

In the same year, Wang’s group realized the AuCl₃-catalyzed intramolecular ring-closing carbonyl–olefin metathesis of 62, featuring a pendant isoprenyl group, to produce N-heterocycle (63a) and cyclopentene (63b, 63c) in good to excellent yields (Scheme 13) [67]. This transformation was compatible with diverse benzene ring backbones containing electron-donating or electron-withdrawing groups, a furan ring, and a β-naphthalene ring at R¹. However, switching β-naphthalene with an α-naphthalene ring could not give the corresponding cyclopentene derivative 63d. Notably, the isomer 63d is generated using 4-methoxyphenyl-substituted β-ketoester as a substrate, which is more selective compared to the approach mediated by FeCl₃ or Host-HCl [54,68]. Density functional theory (DFT) calculations suggested a [2+2] cycloreversion mechanism of fused oxetane Int 34, leading to the desired product 63.

Subsequently, Greb’s group employed β-ketoester 64 as a substrate for directly access to cyclopentene 66, promoted by 65-(CH₃CN)₂ at room temperature in a highly efficient manner, which indicated that the formation of an acetone by-product did not hinder catalytic activity (Scheme 14) [69]. Remarkably, this transformation could not proceed with replacing 65-(CH₃CN)₂ with 65-(acetone)₂, which emphasized the critical role of the displacement of the first Lewis base in a bis-adduct of 65.

In 2023, structurally diverse cyclopentenes 68 and 69 were accessed via ring-closing carbonyl–olefin metathesis, catalyzed by 10 mol% InCl₃ with the assistance of microwave irradiation (Scheme 15) [70]. Generally, 67 featuring a 1,3-diketone functional group delivered the preferential cyclization of electron-rich aromatic ketones (68a–68b). While the acetyl and benzoyl groups were contained in the substrate, 68c was formed in a 23% yield. Mechanistic studies revealed that the tandem [2+2] cycloaddition/cycloreversion process facilitated the construction of 68 and 69. Notably, the coordination of 67a with InCl₃ could occur at the carbonyl group, leading to intermediates Int 35 and/or Int 36, simultaneously forming oxetanes Int 37 and Int 38. Consequently, this reaction provided a mixture of 67e and 69b.

2.2. Carbonyl–Olefin Metathesis Cyclization Reaction Catalyzed by Metal-Free Condition

The field of metal-free carbonyl–olefin metathesis emerged as a pivotal branch of research [43,71]. In 2006, the treatment of (E)-3β,17β-diacetoxy-5,10-secoandrost-1(10)-en-5-one 70 with BF₃·Et₂O was demonstrated by Khripach’s research team, resulting in the cleavage of the macrocyclic structure and the subsequent formation of a novel compound featuring cyclopentenone ring 71 (Scheme 16) [72]. Based on extensive DFT calculations, a plausible reaction mechanism was proposed, involving an intramolecular Lewis acid-promoted [2+2] cycloaddition of a carbonyl with olefin, which was followed by the cycloreversion of the intermediate oxetane species.

Subsequently, Tiefenbacher and his colleagues pioneered the first instance of Brønsted acid-mediated carbonyl–olefin metathesis within a self-assembled supramolecular host (Scheme 17) [68]. Their study featured 72 adorned with α-methyl styryl, β-ketoamide, an unsaturated ketone, a rigid skeleton, and bis-substitution at the α-position, which proved adept at accessing structurally intricate unsaturated carbon ring frameworks 74a–74c, 31b, and 31f in the presence of a co-catalytic system comprising hexamer 73₆(H₂O)₈ and HCl. It is worth noting that the expected product 74 could not be formed without either 73₆(H₂O)₈ or HCl. This catalytic system has demonstrated yields that are comparable to, or even surpass, those achieved by the current benchmark Lewis acid. Moreover, the control experiments conducted conclusively demonstrated that the supramolecular host and Brønsted acid collaborate synergistically to promote the metathesis reaction, with the reaction taking place within the confines of the host structure’s cavity [73,74,75]. Under the standard reaction conditions, when probe 74a was subjected to the process, heterocycle 74c was exclusively isolated as the sole product of the reaction, providing strong evidence for a stepwise oxetane formation. Compared with previous work [59], it appears likely that the mechanism of the oxetane formation depends on the employed substrate type.

In 2022, the same group successfully demonstrated the efficient synthesis of several 3-aryl-2,5-dihydropyrroles (77), utilizing the combination of 73₆(H₂O)_8/HCl (Scheme 18) [76]. This catalytic system was found to be compatible with Lewis basic protecting groups. Additionally, the size of the substrates (76) significantly influenced the yield and selectivity of the reaction, which indicated that the carbonyl–olefin metathesis reaction took place inside the capsule exclusively, combined with the result of the control experiments. Intriguingly, the presence of 73₆(H₂O)_8/HCl triggered the deallylation of chiral amino acid derivatives.

Inspired by these works, Nguyen, Koenigs, and colleagues demonstrated a hydrogen bonding-assisted Brønsted acid catalysis strategy using hexafluoroisopropanol (HFIP) solvent in conjunction with para-toluenesulfonic acid (p-TSA), which controlled the spatial arrangement of reactants and stabilized the transition state, enabling carbonyl–olefin metathesis reactions to yield a mixture of compound 82 and its isomerized alkene 83 (Scheme 19) [77]. This approach prepared a diverse range of cyclopentenes (82a and 82b), five-membered N-heterocycles (82c and 82d), indene (82g), naphthalene (82h), and six-membered cyclic derivatives (82e, 82f, and 82j), with good to excellent yields. Moreover, the study demonstrated the feasibility of 6-endo-trig, carbonyl-ene, and 5-exo-trig cyclization processes, affording desired products 84a and 85a in moderate yields. The theoretical analysis revealed the following: (1) HFIP played an important role in the formation of hydrogen bond networks; (2) the proton migration process resulted in Int 43, which delivered classic carbonyl-ene product 84a via a stepwise mechanism; (3) when n = 1 in 81a, the high ring strain of the putative 1-oxo-bicyclo-[2.2.0]-hexane intermediate could not be formed smoothly, while the 6-endo dig cyclization reaction giving pyrane 85a was a good choice; and (4) Int 44 without an activated carbonyl group underwent 6-endo-trig and 5-exo-trig cyclization, leading to the formation of 85a. While there was an aromatic substituent at the alpha position of 81a, cyclization in a Friedel–Crafts alkylation fashion worked smoothly, affording tetrahydronaphthalene 83c.

Very recently, a homogeneous Brønsted acid-catalyzed intramolecular ring-closing protocol employing nitromethane as a solvent was established by the same group (Scheme 20) [78]. It is worth noting that a negligible isomerization product was observed in this transformation. Diverse cyclopentenes (82a, 82b, 87a, and 87d) and five-membered heterocycle derivatives (87e and 87f) were successfully constructed. Compared with 82a and 87d, the substituent at the α-position of 86 has a significant impact on the yield of the isomerization. Further investigations revealed that protonated carbonyl species (Int 45) derived from triflic acid transformation underwent a stepwise mechanism involving two nucleophilic addition steps, leading to the formation of the oxetane intermediate Int 47 via an ion pair intermediate Int 46. This was followed by a retro-[2+2]-cycloaddition of Int 47, resulting in the cycloalkene product 82b.

In 2018, Nguyen introduced a novel approach leveraging tropylium ions (90) to foster carbon–carbon bond formation, enabling both intramolecular and intermolecular carbonyl–olefin metathesis, as well as ring-opening metathesis (Scheme 21) [79]. This method accommodated a diverse array of substrates, giving the anticipated products (91), albeit with the inevitable isomerization of newly formed double bonds (91b and 91c). Notably, 89 featuring methyl substituents at the α position relative to the original carbonyl group typically undergoes a rearrangement to offer thermodynamically stable olefin product 91d. These results of the control experiments suggested that the volatility of acetone may play a role in driving the reaction equilibrium towards the formation of 91. Mechanistically, the Lewis acidity of tropylium 90 was instrumental in catalyzing these metathesis reactions.

Later, carbonyl–olefin metathesis promoted by molecular iodine was unveiled by Nguyen and co-workers (Scheme 22) [80]. When non-nitrogen-containing substrate 93 underwent this transformation, a mixture of regioisomeric products (94 and 95) was observed due to trace Brønsted acids in the reaction mixture. Notably, the functional group at the α-position of the new C–C double bond have an influence on the ratio of 94 to 95. For example, for substrates containing Brønsted basic nitrogen centers or double substitution at the α-position of 93, isomerization would not occur (87f, 94e). Moreover, the conjugation between the aromatic ring and the carbonyl group is crucial for this transformation, which could explain the low yield of 94e. Increasing the loading to 25%, the corresponding product 94c could be obtained through the metathesis of 93, without using a substitution on the carbon between the carbonyl and the nitrogen centers as a substrate. With different carbon chain linkers between the carbonyl and the olefin moieties, the target compound changed vastly (Scheme 22c). Based on these controlled experiments, DFT calculations and kinetic studies, Nguyen’s group hypothesized that I⁺ ion produced by a very endergonic splitting of molecular iodine due to an unfavorable charge separation could have served as a Lewis acid to activate the carbonyl group providing Int 48, which triggered the formation of oxetane Int 50, followed by a rearrangement, generating 94f.

At the same time, a similar experiment where N-iodosuccimide (NIS) and iodine monochloride (ICl) were engaged as iodonium sources to promote the intramolecular carbonyl–olefin metathesis of 97 was achieved (Scheme 23) [81], further expanding the toolbox for these transformative reactions. Interestingly, the noticeable regioisomerization of the cyclized product 98 was not observed in this conversion. The control experiment validated that the iodonium catalytic pathway of molecular iodine promoted this process.

3. The Formation of Bicyclic Compounds

3.1. The Formation of Bicyclic Compounds Catalyzed by Metal

In 1984, a 30% yield of 7,8-dimethyl-1,2,4,5,6,8-hexahydroazulene 100 was achieved through the ring-closing process of 99, involving the loss of acetone, in the presence of a specific ratio of MeAlCl₂ (2/3 eq) and Me₂AlCl (1/3 eq) (Scheme 24) [82], which opened the door to the synthesis of bicyclic structures through a Lewis acid-catalyzed carbonyl–olefin metathesis reaction. Notably, when 99 was treated with SnCl₄, LiAlH₄, and AlCl₃, a spirocyclic product, 102, was obtained instead [83], highlighting the significant influence of the chosen Lewis acid on product formation.

Inspired by Snider’s work, various bicyclic frameworks were provided, mediated by FeCl₃ and AlCl₃ (56, 57, and 69). In 2020, Wang and co-workers described an efficient way to construct bicyclic aromatic skeletons catalyzed by AuCl₃ (Scheme 25) [67]. Moreover, 1H-indene 63e, 1,2-dihydronaphthalene 63f, and 6,7-dihydro-5H-benzo [7] annulene 63g were prepared through this process. Unfortunately, indole 63h, benzofuran 63h, and 2H-chromene 63i could not be obtained through carbon–oxygen and carbon–nitrogen bond cleavages promoted by AuCl₃. Importantly, polycycle derivatives (63j–63l) were delivered in good yields.

Moreover, it was realized that ester-substituted norbornenes (104), in conjunction with Titanamethylene complex 103, served as substrates to synthesize bicyclo[3.2.0]heptene ring 105 via cyclobutane intermediate Int 52 (Scheme 26) [46]. Enhancing the steric bulk of the ester group not only fortified steric protection around the carbonyl, limiting methylene transfer, but also intensified vicinal steric interactions between the ester groups, promoting the norbornene metathesis process (105a–105c). Additionally, the carbonyl–olefin metathesis of unsymmetrical norbornene substrates facilitated the production of desired products (105e–105g) in moderate yields, without a competing reaction between the norbornene double bond and endo-ester. Moreover, the construction of OPQ, UVW, and JKL ring systems found in maitotoxin was achieved through the carbonyl–olefin metathesis reactions of 106, 109, and 111, followed by a reduction in the resulting carbon–carbon double bonds catalyzed by Tebbe reagent 107 or Petasis reagent 112 (Scheme 27) [84].

Bennasar’s group explored an innovative two-step procedure that involved the titanium-mediated methylation of N-acylamides (114), followed by ruthenium-catalyzed ring-closing metathesis (RCM) to give enamide 115 (Scheme 28) [85,86]. This protocol efficiently facilitated the synthesis of various benzo-fused five- and six-membered cyclic enamides (115a–115f) (including indole, 1,4-dihydroquinoline, and 1,2-dihydroisoquinoline) from an amide derived from ortho-alkenyl aniline or benzylamine. While there was potential for an extension to higher homologues, this process was constrained by the interference of alkene isomerization during the sluggish RCM step. Notably, direct annulation was observed in the titanium-mediated step, particularly with certain substrates, such as styrene derivatives, likely occurring through an olefin metathesis–intramolecular olefination tandem mechanism.

Drawing inspiration from Clark’s work [87,88], Grubbs and Rainier’s teams devised a method for synthesizing benzofuran derivative 117 and cyclic enol ethers (119 and 121) (Scheme 29) [89,90]. This method involved the use of stoichiometric titanium reagents to convert esters into acyclic olefinic enol ethers, which were subsequently transformed into the desired products through catalytic ring-closing olefin metathesis utilizing molybdenum alkylidene complex 18.

Excitingly, Rainier and colleagues demonstrated that Takai–Utimoto-reduced titanium alkylidene could promote the ring-closing reaction of olefinic esters (122), directly producing bicyclic compound 123 in high yields (Scheme 30) [91]. Notably, the steric environment of the ester or olefin influences the ratio of cyclic enol ethers 123 to 124. Moreover, the olefin metathesis of 122 could proceed under standard conditions. Moreover, this conversion was realized via the Ti intermediate Int 54 generated in situ. Given the advantages of this method, it was applied for the synthesis of key molecular building blocks (125, 127, and 129) to construct Brevenal, which demonstrated the practicality of the olefinic ester cyclization reactions (Scheme 31) [39].

As a continuous study, olefinic lactone cyclization to macrocycle 131 under similar conditions was developed in 2009 (Scheme 32) [92]. Employing 13-, 16- or 17-membered lactones as starting materials, the desired products 131a–131d were constructed. In addition, the oxidation, reduction, and ring-opening process of 131 proceeded smoothly to give macrocyclic ketal 132, macrocyclic pyran 133, and macrocyclic ketone 134 in good yields. Furthermore, natural products (R)-(−)-muscone 135 and (R)-(+)-muscopyridine 136 could be delivered through the key intermediate 131d.

The use of in situ-generated titanium ethylidene to react with olefinic amide or olefinic lactam, facilitating indole 138a, dihydroquinoline 138b, and fused seven-membered ring 138c, has been reported by the same group (Scheme 33) [93]. The carbonyl–olefina tion pathway of nonaromatic olefinic actam 137 could proceed to provide 138d–138i. Notably, these transformations were accompanied by an olefin metathesis reaction to give 139, which could be inhibited by using dihydroquinone as an additive [94]. Moreover, the treatment of 138d with aqueous acid would convert it into cyclobutanone 140. Bromoaminal 141 was afforded by the olefin functionalization of 138d. Further experiments illustrating the carbonyl-ene alkylation of 137 could not furnish the desired product, 138, under optimized reaction conditions.

3.2. The Formation of Bicyclic Compounds Catalyzed by Metal-Free Conditions

In 2009, Kutateladze’s research team disclosed a highly efficient methodology involving a one-pot, photoinduced transformation of endo-aroyl and heteroaroyl Diels–Alder adducts (142), facilitating the construction of novel bicyclic aldehydes (143) or their corresponding hemiacetals (144), with satisfactory yields (Scheme 34) [24]. Bicyclic aldehyde 143 adorned with carbo- and heterocyclic compounds could be obtained through this transformation. The targeted product, 143, was attained through a sequence comprising the Paternò–Büchi reaction, a subsequent Grob-like fragmentation of the initially ring-opened carbocationic intermediate Int 56, and ultimately, acidic conditions promoting the formation of hemiacetal 144 via a secondary electrophilic addition. Notably, trace amounts of HCl present in solvents like dichloromethane could modulate the cycloreversion process. Furthermore, the oxetane intermediate Int 55 could be diverted towards aldehydes or hemiacetals, contingent upon the utilization of Lewis acid or protic acid.

In 2013, the same group reported that conformationally pliable aroyl methyl chromophore-decorated bicyclic scaffolds (145) were harnessed to construct innovative bicyclic frameworks 146 and 147, leveraging photoinduced intramolecular cyclo-penetration and photo-protolytic oxametathesis reactions (Scheme 35) [95]. Subsequent investigations highlighted several crucial insights as follows: (1) the steric hindrance of 145 significantly influenced the ratio of intermediates Int 57 to Int 61; (2) the epimerization of 146A and 146B via enolization, bringing a formyl group into the exo-position, was restricted in the presence of ethylene glycol; (3) β-hydrogen abstraction from the CH₂ group by an excited benzoyl moiety proved pivotal for the formation of 149; (4) this unique photorearrangement/cyclo-penetration paradigm involving double bonds emerged as an effective strategy for synthesizing fused cyclopentanol. Notably, Int 68 was converted from Int 65, involving a 1,2-radical shift of the bicyclic core facilitated by enol formation, followed by the keto-enol equilibrium, second excitation with H abstraction, and an interrupted cyclization to give 149.

Recently, trifluoromethanesulfonic acid (TfOH)/MeNO₂ was proved to be an efficient combination to promote the carbonyl–olefin metathesis (COM) reaction of 150, producing various polyaromatic products (151a–151d and 152a–152e) (Scheme 36) [78]. Notably, small amounts of carbonyl-ene-type by-products (152) were inevitably generated.

Additionally, Schmalz and his collaborators established a protocol involving boron trifluoride etherate-mediated intramolecular carbonyl–olefin metathesis (Scheme 37) [96]. In this process, acetophenone or benzophenone derivative 153 underwent Prins-type (exo-trig) cyclization, forming a tertiary carbenium ion, Int 70, which subsequently isomerized to Int 72 via oxetane intermediate Int 71. Ultimately, the fragmentation of Int 72 produced the desired product 154, with yields reaching 93%, accompanied by ketone 58 as a by-product.

Subsequently, 4-phenyl-diphenylmethylium ion 156 proved to be an adept catalyst for promoting the ring-closing metathesis of aldehyde and olefin, leading to the synthesis of indene 157 in good to excellent yields (Scheme 38) [97]. The loading of 156 could be reduced to 2 mol%, and these products (157) could be isolated in good yields, with high purity and selectivity, which is different from previous work. Utilizing the developed method, a series of functionalized indene derivatives with various substituents at any position could be easily prepared. Importantly, the decomposition of product 157 was observed under optimized reaction conditions, and the benzylic functionalization of enal 155 was compatible with this transformation, facilitated by the addition of just 2 mol% of 156.

Later, the synthesis of 2H-chromene 161 was successfully achieved through a novel [3+2]/retro-[3+2] metathesis sequence, utilizing O-allyl salicylaldehyde 159 as the substrate, catalyzed by 10 mol% of [2.2.1]-bicyclic hydrazine 160 (Scheme 39) [98]. Among the various allyl groups screened, 3,3-diethylallyl emerged as the optimal fragment for this transformation, effectively suppressing undesired deallylation side reactions. Notably, compound 159 with substitutions at positions 6, 7, and 8 was suitable for this transformation, enhancing the versatility of this process. Increasing the loading of 166 to 20 mol% and replacing ethanol with isopropanol enable the smooth formation of 161e. Critically, a formal high-throughput intermolecular screening method was implemented to assess the compatibility of various functional groups, demonstrating robust tolerance towards ketones, unprotected alcohols, and carboxylic acids [99]. A plausible reaction mechanism was then proposed, containing the condensation of aldehyde 159a with hydrazine 160 to form hydrazonium intermediate Int 73, which subsequently underwent 1,3-dipolar cycloaddition with an olefin to offer Int 74. Following proton migration and cycloreversion, 2H-chromenes (161a) were furnished, with hydrazine 160 being regenerated through the hydrolysis of Int 76. Additional DFT calculations underscored the importance of allyl substitution patterns in sterically hindered systems, where cycloreversion was facilitated by the relief of the syn–pentane strain in the diethylidene moiety.

In 2020, the same group utilized a similar strategy to construct 1,2-dihydroquinoline 164, utilizing N-prenylated 2-aminobenzaldehyde 163 as the substrate (Scheme 40) [100]. While the cycloreversion step was found to proceed more rapidly compared to the O-allyl salicylaldehyde system, the formation of the cyclization intermediate was hindered by the strong electron-donating capability of the nitrogen atom. Employing t-butyloxycarbonyl (Boc) or acetyl (Ac) as N-protecting groups led to the formation of quinoline derivatives. However, a pronounced steric clash between the tosyl group and methyl substituent negatively impacted the yield of 164e. Additionally, the presence of aniline and alkyl halide additives was detrimental to the product yield, attributed to their competitive reactions with 160 or an aldehyde. Nonetheless, the optimization of reaction conditions followed by Lewis acid-mediated cycloaddition allowed for the isolation of the azapterocarpan analog 167 in a respectable 52% yield [101].

4. The Formation of Polycyclic Compounds

4.1. The Formation of Polycyclic Compounds Catalyzed by Metal

The synthesis of polycyclic structures has been a formidable challenge in chemistry. In 1986, Grubbs pioneered an effective approach to accessing capnellane, wherein substrate 168 underwent an intricate intramolecular cycloaddition, ring-opening, and olefin metathesis sequence in the presence of titanium reagent 169 and DMAP, providing the polycyclic product 171 (Scheme 41) [102]. Subsequently, capnellane was derived from 171 through a five-step synthesis. Further developments indicated olefin esters could be directly converted into enol ethers featuring polycyclic structures with six- or seven-membered rings in high yields, facilitated by Tebbe (169) or Petasis reagents (112) (Scheme 42) [45]. The proposed mechanism commenced with the formation of titanacyclobutane, resulting from the olefination of enol ether 172 with 169 or 112. This intermediate then underwent fragmentation, intramolecular cycloaddition, and regioselective fragmentation, ultimately leading to the desired product 173. Additionally, hexacyclic polyethers could be synthesized to form 173d through three steps.

Using a Ti reagent, a precursor 174 adorned with a silicon ring could undergo ether–olefin ring-closing metathesis in the presence of 25, giving the expected product 175 in an 80% yield (Scheme 43) [103]. Moreover, 177, containing double seven-membered rings, was also produced by this methodology in a 50% yield utilizing Schrock catalyst 18. In 2009, the same research group disclosed a novel two-directional olefinic ester ring-closing metathesis promoted by reduced Ti alkylidenes (Scheme 44) [104], enabling the synthesis of tri-, penta-, and heptacyclic scaffolds (179a–179d). When the starting material (180) possessed two carbonyl groups, direct metathesis between the alkene and carbonyl moieties yielded 48% of 182, with the absence of competitive product 181 attributed to the high torsional strain in the bridged structure (Scheme 45) [105]. Moreover, (–) huperzine could be prepared through four steps with 182 as the starting material.

In 2017, Schindler and colleagues introduced a potent method for constructing polycyclic aromatic hydrocarbon 184 (Scheme 46) [106]. The starting material (183) underwent carbonyl–olefin metathesis via an oxetane intermediate, affording the corresponding metathesis product 184 with yields up to 99%, compatible with both ketones and aldehydes. Notably, Brønsted acids like HCl and p-TsOH could not facilitate this transformation. An alkene evaluation disclosed the following: (1) the E- and Z-isomer of olefin functional groups in 183 had no adverse effect on the reactivity; (2) benzaldehyde could not impede the reaction progress; (3) there was no reaction employing 183 featuring a terminal alkene as a substrate; and (4) nonstyrenyl substrate 183 could promote the competing carbonyl-ene reaction pathway, leading to a low yield of 184a. In addition, various polycyclic skeletons, such as phenanthrene 184a and benzothiophene 184h, could be obtained. Interestingly, two carbonyl–olefin metathesis processes catalyzed by FeCl₃ could proceed to offer the expected product, 184j. Notably, only oxetane 184k was isolated using the prenylated analog of 183 as a substrate. Later, angucycline 188 and tetrangulol skeleton 189 were furnished in yields of 60% and 67%, respectively, through this transformation, improving the catalytic loading to 20–27 mol% (Scheme 47) [107].

Recently, a metathesis reaction involving ketone–olefin 190, aldehyde–olefin 190, acetal–alcohol 191, and carbonyl–alcohol 192 was successfully catalyzed by the Brønsted acid H₃PMo₁₂O₄₀ (Scheme 48) [108]. Both naphthyl-based and heteroaromatic ring-derived substrates were found to be compatible with this catalytic process. Nonetheless, benzo[h]quinoline scaffold 193b could not be synthesized due to competitive binding with the pyridine skeleton, which hindered the reaction progression. Furthermore, the inherently sluggish reactivity of esters and amides (where R = MeO or NH₂) resulted in the failure to generate the expected product under optimized reaction conditions. Intriguingly, when an E/Z mixture of olefin was used as the substrate, the corresponding product was successfully furnished. Additionally, both the double in situ metathesis of acetal–alcohol and a sequential dehydration/in situ metathesis of carbonyl–alcohols proceeded smoothly, affording the desired product with a good yield. Notably, the metathesis of carbonyl–alcohols did not proceed when a traditional Lewis acid was employed. These mechanistic studies indicated that the reaction proceeded via a stepwise mechanism (Scheme 48c), involving cation intermediate Int 78, which was stabilized by a vast anion cavity structure centered around a tetrahedral (PO₄) unit, surrounded by four octahedral (MoO₆) metal units.

4.2. The Formation of Polycyclic Compounds Catalyzed by Metal-Free Conditions

As a supplement to previous research, Lambert’s team achieved a remarkable breakthrough by catalyzing the ring-closing carbonyl–olefin metathesis of biaryl alkenyl aldehyde 195 using hydrazine as a catalyst. This process efficiently yielded polycyclic heteroaromatic compound 196, with an excellent yield achieved using just 5 mol% of 160 (Scheme 49) [109]. A diverse range of anticipated products, including benzo[h]isoquinolines (196a–196e), phenanthrene (196f), thiophene-fused polycyclic heteroaromatics (196h–196l), naphthofuran (196m), benzocarbazole (196n), and naphthoimidazole (196o), were constructed in moderate yields. An olefin substitution screen revealed that 2-methyl-1-propenyl emerged as the optimal substituent for this transformation. Furthermore, DFT calculations were performed, elucidating the reaction mechanism. The condensation between the starting material (195) and [2.2.1]-hydrazinium 160 yielded the intermediate Int 80. Subsequently, an intramolecular cycloaddition between the olefin and the 1,3-dipole occurred, leading to the formation of Int 81. This intermediate then underwent a cycloreversion process via Int 82, ultimately producing the desired product, 196a.

5. Conclusions

In summary, we provided an up-to-date review of the cyclization of carbonyl–olefin metathesis for the synthesis of monocyclic skeletons, bicyclic scaffolds, and polycyclic frameworks. These complicated frameworks were furnished through the fragmentation of oxetane, the cycloreversion of pyrazolidine formed by the [3+2]-cycloaddition of a carbonyl with hydrazine, and a metal alkylidene intermediate. Since the carbonyl–olefin metathesis transformation was disclosed in 1960, significant progress has been made within the last ten years. However, there remain many challenging tasks. Firstly, the functional group tolerance and broadening the substrate scope should be solved to enable efficient applications in the construction of complex molecules. Meanwhile, the protocol tolerating exceedingly Lewis basic sites and producing medium- and larger-sized rings should be prioritized for development. Moreover, the isomerization of newly formed olefins is commonly presented in current methods. In addition, enhancing the distinct reactivity mode of a carbonyl with alkene functionalities was a key branch in this field. Photocatalytic carbonyl–olefin metathesis is still in its infancy. Notably, the interrupted cycloreversion of oxetane has become a current frontier to afford heterocyclic and polycyclic compound skeletons.