Recent advances in alkylidene carbene chemistry

• Users may freely distribute the URL that is used to identify this publication. • Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private study or non-commercial research. • User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) • Users may not further distribute the material nor use it for the purposes of commercial gain.


Introduction
Alkylidene carbenes, also known as alkenylidenes, are reactive intermediates in organic chemistry. Their high reactivity necessitates generation and trapping in situ, but unlike many carbenes, 1 they display high levels of selectivity without the need to be associated with a metal catalyst. Indeed, although many methods for the formation of alkylidene carbenes are metal-based, their reactivity is generally rationalized without invoking a role for the metal.
The most common reactions of alkylidene carbenes are shown below (Scheme 1). 1,2-Migration is the preferred reaction pathway for phenyl and hydrogen substituted alkylidene carbenes (Eq. 1). Since alkylidene carbenes precursors are often derived from aldehydes and ketones, this offers a method of carbon homologation, and is extensively utilized in organic synthesis. Alkyl-substituted alkylidene carbenes generally undergo insertion rather than migration. Intramolecularly, this leads to cyclopentenes through a highly regioselective 1,5 CeH insertion reaction (Eq. 2). This reaction occurs with high levels of chemoselectivity (for more highly substituted CeH bonds) and is stereospecific (retention of absolute configuration at the insertion site), the latter process being consistent with alkylidene carbenes reacting through the singlet state. Coupled with the ability to insert into unfunctionalized CeH bonds, thereby simplifying structural complexity in the starting material, the 1,5 CeH insertion reaction of alkylidene carbenes is particularly useful in organic synthesis. Five-membered unsaturated heterocycle formation occurs when a heteroatom is present in the connecting chain, and is also observed in formal 1,5 insertions into heteroatom bonds, most commonly OeH and OeSiR 3 , to make dihydrofurans (Eq. 3).
This review covers synthetic aspects of both the generation and reactions of alkylidene carbenes, including related organometallic chemistry, which fits into the reactivity patterns shown above. Other reaction pathways and factors affecting selectivity are also discussed, as are applications in target synthesis. Aspects of the chemistry of alkylidene carbenes have been previously reviewed in the literature. A short review on the synthetic potential of alkylidene carbenes appeared in 1997. 2 A comprehensive review by Knorr in 2004 concentrated on the evidence for alkylidene carbenes in a number of reactions. 3 The use of hypervalent alkynyl iodides as precursors to alkylidene carbenes has been previously reviewed, 4 as has the use of a-heteroatom substituted alkenyl lithium reagents. 5 A 2010 review by Koskinen discussed developments in the conversion of carbonyl compounds to alkynes, 6 and as such, the synthetic utility of this transformation will not be extensively covered.

Structure and bonding
In alkylidene carbenes, the carbenic centre is located on the terminus of a carbonecarbon double bond. Ab initio studies have shown that the carbenic carbon is effectively sp hybridized and exists in a singlet state 1 (Fig. 1) with both electrons located in the sp non-bonding orbital (HOMO), and the p orbital (LUMO) empty. 7 The singlet state is lower in energy than the triplet 2 by 48 kcal/ mol, 8 and can react in a single concerted step. Alkylidene carbenes react as either the free carbene 1 or metal carbenoid 3 depending on the generation method employed. Both forms generally give rise to the same product outcome.
Calculations performed by Van Nhien and Postel on alkylidene carbene 4 showed that the HOMO is delocalized over several atoms, but most of the contribution comes from the carbenic carbon ( Fig. 2). 9 The LUMO is also mainly located on the carbenic centre. NBO analysis showed that the carbene is approximately sp 1.5 hybridized. 9 The lone pair is predominantly s in character with 36.5% p mixing.
Retention of absolute configuration in 1,5 CeH insertion reactions of alkylidene carbenes suggests a mechanistic pathway without any long-lived intermediates, such as diradicals or zwitterions, which could result in the loss of stereochemical information. 10 Gilbert proposed that an interaction between the empty p orbital of the carbenic carbon and the s electrons in the CeH bond leads to a transition state such as 5, before the insertion is complete (Fig. 3).
Subsequent calculations by Van Nhien and Postel on the 1,5 CeH insertion reaction of carbene 4 are in agreement with Gilbert's original hypothesis. 9 They reported a half chair-like transition state, with the hydrogen atom, C1, C2 and C5 all occupying the same plane. In this conformation, the CeH s orbital and the carbene's   Please cite this article in press as: Grainger, R. S.; Munro, K. R., Tetrahedron (2015), http://dx.doi.org/10.1016/j.tet.2015.06.053 empty p orbital are periplanar, allowing the orbitals to overlap. 7 This is in contrast to kinetic experiments performed by Gilbert, which suggested a non-linear transition state, with the breaking CeH bond out of the plane of C1, C2 and C5. 10 While acknowledging that a non-linear arrangement would minimize any steric repulsion between the hydrogen and the carbons, Van Nhien and Postel suggested that the long C5eH and C1eC5 distances (1.440 A and 2.245 A, respectively), coupled with the nearly fully-formed C1eH bond (1.168 A), ensure minimal van der Waals repulsive interactions. They also commented that these distances suggest that the new CeH bond forms much faster than the new CeC bond, rendering the insertion process concerted, yet strongly asynchronous. 9 A number of computational studies have been carried out on the isomerization between vinylidene and acetylene (Scheme 2). 11 These suggest that the 1,2-hydride migration proceeds through a transition state such as 6, with a near-linear relationship between the two carbons and the non-migrating hydrogen atom.
While metal carbenoids 3 generally exhibit the same reactivity as the free carbenes, they differ structurally from the linear nature of 1. The LieCl carbenoid 7 has been isolated and analyzed by X-ray crystallography at low temperatures by Boche et al. (Fig. 4). 12 The bond angles at C1 within the carbenoid structure are very different from those expected for an sp 2 -hybridized carbon. The Cl1eC1eC2 angle is 112.6(5) , smaller than the expected trigonal angle of 120 , while the Li1eC1eC2 angle of 137.1(6) is much larger than expected. Another feature was the elongation of the C1eCl1 bond, which at 185.5 (7) pm is longer than the mean C sp 2 eCl length of 172.9 pm. 13 This elongation is in accordance with the conclusions made from analysis of 13 C NMR spectra of simple Li-halogen alkyl carbenoids. 14 Boche et al. demonstrated that similar structural features were also present in magnesium-halogen alkylidene carbenoids 15 . Kimura and Satoh have carried out DFT calculations of organometallic halides of general structure 8 (Fig. 5, Table 1). 16 Vinylmagnesium chlorides bearing a-halo substituents, such as 8a, were found to have substantial vinylidene character (i.e., close to sp hybridization). The carbonehalogen bond lengths were calculated to be longer than in the corresponding vinyl halides 8b, while the C]CeX bond angles were also smaller than those of the vinyl halides. In addition, the C]CeMg angles in a-halo vinylmagnesium chlorides were all larger than the corresponding angle in vinylmagnesium chloride 8c.
Conversely, the CeX bond length of the vinylmagnesium chloride bearing an a-amino substituent 8d, was shown to be similar to that of the corresponding enamine 8e, while the C]CeMg bond angle is smaller than vinylmagnesium chloride. The values for examples bearing a-thio and a-seleno substituents showed similar properties. This suggests that these carbenoids have little or no vinylidene character. NBO analysis demonstrated that in vinylmagnesium chlorides bearing a-halo substituents, the carbon atom of the CeX bond had higher p-character than the corresponding vinyl halides. As such, these carbon atoms were also calculated to have higher s-character in the CeMg bond compared to vinylmagnesium chloride. Replacing the magnesium atom with a lithium atom 8f resulted in an increase in vinylidene character, with the CeCl bond longer than in both the vinyl chloride 8b and its magnesium carbenoid derivative 8a. In addition, the C]CeCl bond angle was smaller and the C]CeLi bond angle larger than the corresponding vinyl chloride and vinyllithium, respectively. The differences in bond angles were also more marked than in the magnesium carbenoid examples.
Kimura and Satoh's calculations are in accordance with the reactivity of these substrates. a-Haloand a-alkoxy-substituted metal carbenoids, which were calculated to have similar geometrical features, display alkylidene carbene reactivity, while those bearing a-amino and a-thio substitution react as carbanions. 3,5,17 In addition, lithium carbenoids are known to be more reactive than their magnesium counterparts. 3,5,17b,c,18 Kimura and Satoh proposed that electron delocalization from the CeMg bond into the CeX antibonding orbital was the key orbital interaction in determining the vinylidene character of a-heteroatom-substituted vinylmagnesium chlorides. The stabilization energy of the s CeMg /s Ã CeX delocalization was found to be much larger for a-halo substituents than for other a-heteroatoms, suggesting that the carbanion is stabilized by this interaction. This interaction is also responsible for the lengthening of the CeX bond.

Methods of generation
The high reactivity of alkylidene carbenes necessitates their in situ generation. The techniques commonly utilized employ the use of diazo olefination reagents, haloalkenes and alkynyliodonium salts. In all cases, the reactive species is generally represented as the free alkylidene carbene, although it is not clear to what extent reactions proceed through this high-energy intermediate. However, where comparisons are possible, the same product outcome is generally observed.

Nitrogen extrusion
One of the most widely employed methods by which to access alkylidene carbenes is the loss of N 2 from a double bond terminus, in particular from 1-diazoalkenes. The simplest, and most commonly used, technique for accessing diazoalkenes is the reaction of aldehydes or ketones in a modified HornereWadswortheEmmons reaction, 19 or in a modified Peterson olefination. 20 These reactions employ the use of diazomethylphosphonate (DAMP) esters 9 (SeyfertheGilbert reagent) and trimethylsilyldiazomethane (10), respectively. While trimethylsilyldiazomethane is commercially available, it is necessary to synthesize the phosphonate esters.
The reaction proceeds through the established mechanism for both the HornereWadswortheEmmons and Peterson reactions, with initial nucleophilic attack on the carbonyl functionality by the deprotonated olefination reagent, with subsequent elimination to give the 1-diazoalkene 11. Unlike a-diazo carbonyls, which often requires a metal catalyst to facilitate the extrusion of nitrogen at a reasonable temperature, 1d 1-diazoalkenes readily undergo the loss of nitrogen to produce alkylidene carbenes (Scheme 3).
The intramolecular CeH insertion reaction of alkylidene carbenes proceeds with remarkable selectivity, producing five-membered rings, irrespective of the substitution pattern present, and generating secondary, tertiary and quaternary centres. Additionally, the stereospecificity of the reaction results in retention of absolute configuration at a stereogenic centre. 21 An early example of the remarkable stereospecificity of the 1,5 CeH insertion reaction of alkylidene carbenes was the synthesis of cyclopentene 14 from the enantiopure ketone 12 by Gilbert in 1985, via alkylidene carbene 13 (Scheme 4). 10 Upon conversion to enone 15, the insertion reaction was found to proceed with a stereoselectivity of at least 99%. This retention of absolute configuration is strong evidence that the alkylidene carbene is reacting through the singlet state.
Wardrop reported the formation of dihydrofurans 18 and 19 through treatment of the a and b isomers of ketone 16 with lithiated trimethylsilyldiazomethane 17 (Scheme 5). 22 This example demonstrates the retention of absolute configuration associated with 1,5 CeH insertion adjacent to heteroatoms.
Diazo olefination reagents have also been used extensively in the synthesis of terminal alkynes from aldehydes, 6,23 and diaryl alkynes from ketones. 24 The use of the modified Peterson conditions has been extended to the synthesis of 2,3-dihydrofurans 21 and 23 from ketones 20 and 22, via formal 1,5 OeH and OeSi insertions, respectively. 25 This appears to be the only case of a direct comparison between OeH and OeSi insertion, and suggests that OeSi insertion is more efficient (Scheme 6).
It is believed that these products form through one of two possible mechanisms (Scheme 7). 26 The alkylidene carbene 24 can undergo insertion into the OeSi bond, forming dihydrofuran 25 directly via a transition state similar to that for CeH insertion. Alternatively, the reaction could proceed through the formation of an oxonium ylide 26, followed by a 1,2-shift (Stevens rearrangement) of the oxygen substituent.
Similar pathways involving intramolecular insertion into NeH bonds or interactions with nitrogen lone pairs to form 2-pyrrolines have also been reported. 27 Diazo olefination reagents have been utilized in the discovery of new alkylidene carbene reaction pathways. Daesung Lee reported the synthesis of silylated cyclopropenes through the generation of alkylidene carbenes from a-silyl ketones using Peterson-type reaction conditions. 28 Treatment of citronellal 27 with trimethylsilyldiazomethane in the presence of catalytic InCl 3 allowed access to a-silyl ketone 28. This was subsequently treated, without purification, with lithiated trimethylsilyldiazomethane 17, giving rise to cyclopropene 29 in 77% yield (Scheme 8). Cyclopentene 30, the expected product of 1,5 CeH insertion, was not observed.
D. Lee proposed that the cyclopropene arose from the insertion of the alkylidene carbene into the nearby CeSi bond, suggesting that the interaction between the empty p-orbital of the carbene and the CeSi s-bond was more favourable than the interaction with the more remote CeH s-bond (Scheme 9).
The reaction was performed on a range of substrates, including those containing straight-and branched-chain alkyl groups,  varying degrees of unsaturation (both cis and trans) and aromatic substituents. Five-membered ring formation was only observed in cases where 1,5 CeH insertion was promoted by an adjacent oxygen atom (vide infra) (Scheme 10). Even so, cyclopropenes remained the major products isolated, suggesting that the 1,3 CeSi insertion pathway is much more favourable.
Alkylidene carbenes can undergo [2þ1] cycloaddition reaction with alkenes, giving rise to bicyclo[3.1.0]hex-1-ene derivatives. For example, alkylidene carbene 33, generated by the reaction of ketone 31 with 17, underwent intramolecular [2þ1] cycloaddition reaction with the tethered alkene, giving rise to the bicyclo[3.1.0] hex-1-ene 34. 29 In situ fragmentation of this strained ring system generated trimethylenemethane diyl 35, 30 which afforded linearly fused triquinane 32 through a stereoselective intramolecular [3þ2] cycloaddition reaction (Scheme 11). 30b Hee-Yoon Lee has expanded his work in this area to generate the initial alkylidene carbene from both a,b-epoxy-N-aziridinylimines 30a and alkynyliodonium salts. 30c,d Bicyclo[3.1.0]hex-1-ene systems generated from alkylidene carbenes can also dimerize via an intermolecular [2þ2] cycloaddition reaction to give polycyclic systems. 31 Alkylidene carbenes can also undergo [2þ1] cycloaddition reaction with the p-system of aromatic rings. 32 Shioiri reported the synthesis of cyclohepta [b]pyrrol-2-ones 37 from the reaction of aketo anilides 36 with 17 (Scheme 12). 32a Insertion of the alkylidene carbene 38 into the aromatic p-system affords norcaradiene 39, which rearranges to the product cycloheptatriene 37. Interestingly, reaction with the p-system occurs in preference to 1,5 CeH insertion into the N-methyl group, as the corresponding N-aryl pyrrol-2-ones were only observed as byproducts in some cases. While modified olefination reagents are undoubtedly useful for the in situ generation of alkylidene carbenes, a potential drawback is the strongly basic conditions required for their use. The thermal decomposition of a,b-epoxy-N-aziridinylimines 40 allows for the generation of alkylidene carbenes under neutral conditions. 33 This approach also offers an additional benefit, allowing for the generation of an alkylidene carbene with a b-hydroxyl group. The decomposition is thought to proceed in a similar fashion to the Eschenmoser fragmentation. 34 Heating 40 in toluene generates diazo compound 41, with concurrent loss of styrene. Opening of the epoxide generates the double bond, and proton transfer from the vinyl carbon to the oxygen in betaine 42 leads to diazonium ylide 43, which extrudes nitrogen, resulting in b-hydroxy alkylidene carbene 44 (Scheme 13). The N-aziridinylimines are readily available by treating the corresponding carbonyl compound with Naziridinylamines.
Kim utilized this methodology to synthesize propargyl alcohols and cyclopentenols, 26a and later expanded the scope of this reaction to include the formation of silylated dihydrofurans through a formal 1,5 OeSi insertion (Scheme 14). 35 Four-membered heterocycles open under similar conditions. 36 Thus, a-oxetanyl-N-aziridinylimines gave dihydrofurans via 1,5 OeH insertion, while a-azetidinyl-N-aziridinylimines gave homopropargyl amines via 1,2-migration (Scheme 15). When heterocyclic systems larger than four-membered rings were employed, then the reaction pathway did not proceed through the alkylidene carbene, suggesting that the relief of ring strain plays a key role in the generation of this reactive intermediate. Related to the loss of nitrogen from 1-diazoalkenes is the fragmentation of tetraazafulvenes 45 to give alkylidene carbenes, with the concurrent loss of two molecules of nitrogen (Scheme 16). Like diazoalkenes, tetraazafulvenes are unstable and it is necessary to generate them in situ from more stable precursors. The use of tetrazoles to generate alkylidene carbenes was first reported by Behringer in 1966, 37 but further development of this chemistry has only occurred in recent years.
In 2005, Van Nhien and Postel described the synthesis of dihydrofuran 47 when cyanomesylate 46 was treated with TMSN 3 in the presence of Bu 2 SnO (Scheme 17). 38 They postulated that the reaction proceeded through initial cycloaddition between the nitrile group and TMSN 3 , affording tetrazole 48. Desilylation and rearrangement to 1H-tetrazole anion 49 is followed by elimination of the mesylate to give the tetraazafulvene 45. The extrusion of two molecules of N 2 gives the alkylidene carbene, which undergoes subsequent 1,5 OeSi insertion (Scheme 18). The Bu 2 SnO is present to accelerate the initial cycloaddition. 39 Two less common reaction pathways of alkylidene carbenes were demonstrated using this methodology. 38 Intermolecular [2þ1] cycloaddition of the alkylidene carbene generated from cyanomesylate 50 gave 51 as a mixture of diastereoisomers, while cyanomesylate 52 was converted to dihydropyran 53, the product of an alkylidene carbene 1,6 CeH insertion reaction, albeit in modest yield (Scheme 19). 38 The product was isolated as a mixture of diastereoisomers. Feldman had previously demonstrated 1,6 alkylidene carbene insertion reactions using naphthol and anthrol systems, suggesting that this reaction can occur in systems where the 1,5 insertion pathway was blocked. 40 Wardrop has also reported the generation of alkylidene carbenes via tetraazafulvenes 45, accessing them through dehydration of 5-hydroxy-1H-tetrazoles 54 with carbodiimides (Scheme 20). 41 The 5-hydroxyalkyl-1H-tetrazoles 54 were readily prepared in two steps from precursor carbonyl compounds via addition of 1allyl-5-tetrazoyllithium 55 followed by Ni-catalyzed N-deallylation in the presence of t BuMgCl (Scheme 21). A range of substrates were prepared this way, mostly in good yields. Treatment of 5-hydroxyalkyl-1H-tetrazoles with a range of carbodiimides afforded alkynes via 1,2-migration of phenyl and hydrogen groups in good yield, with diisopropylcarbodiimide (DIC) giving the best results (Scheme 22).
Unsaturated ring systems could also be synthesized in reasonable yields using this methodology through 1,5 CeH insertion reactions of the intermediate alkylidene carbenes (Scheme 23). The conditions required were more forcing than those for the 1,2-migration reaction, with refluxing DCE needed to ensure conversion.
Nitrogen extrusion from double bond termini is a versatile means for accessing alkylidene carbenes, especially as the necessary precursors are readily available from carbonyl compounds. The development of tetraazafulvene and N-aziridinylimines precursors enable alkylidene carbenes to be generated under neutral conditions. In addition, N-aziridinylimines generate additional alcohol functionality within the target molecule.

a-Elimination of metal carbenoids
Alkylidene carbenes can be generated through the reductive elimination of metal carbenoids, also known as metal vinylidenes (Scheme 24). 5,42 Despite being structurally different to 'naked' alkylidene carbenes, carbenoids display the same reactivity, undergoing 1,2-migration and 1,5 insertion reactions. Despite this, it is unclear to what degree the metal salt is fully dissociated from the carbene prior to reaction. Labelling studies have shown that in the 1,2-migration of diaryl alkylidene carbenoids, the aryl group trans to the halogen preferentially migrates to give the resulting alkyne. 43 Alkylidene carbenoids are readily available through a variety of methods. One of the most widely used is through the base-induced elimination of vinyl halides 44 or vinyl triflates. 23a, 45 Ohira reported the synthesis of cyclopentene 56 from vinyl triflate 55 using KO t Bu as a base (Scheme 25). 45 The most commonly employed tactic for accessing these carbenoids is the deprotonation of vinyl halides. Taber reported the use of NaHMDS to synthesize cyclopentene 58 in good yield from vinyl chloride 57, 44d while Hayes prepared 3-pyrroline 60 through treatment of vinyl bromide 59 with KHMDS, without the need to protect the secondary amine (Scheme 26). 46 In both cases retention of absolute configuration at the 1,5 CeH insertion site adjacent to the heteroatom is observed.
Alternatively, metal carbenoids can be accessed via metalehalogen exchange of a dihaloalkene. Tykwinski employed this strategy in the synthesis of polyynes via a 1,2-migration, including triyne 62 in good yield from dibromoalkene 61 (Scheme 27). 47 This reaction was also the first example of an alkynyl group undergoing 1,2-migration. 47,48 This method has been widely used in the synthesis of polyynes. 49 Tani reported the formation of fused cyclopentene 64 by treating dihaloalkene 63 with excess SmI 2 and HMPA (Scheme 28). 50 While the reaction shown in Scheme 27 is believed to proceed through lithiumehalogen exchange, the transformation in Scheme 28 is more likely to proceed via electron transfer followed by loss of the halide anion from dihaloalkene 65 to give vinyl radical 66. Further reduction by a second equivalent of SmI 2 gives carbenoid 67, which reacts to form the cyclopentene 68 via 1,5 CeH insertion (Scheme 29). 50 The use of vinyl halides and dihaloalkenes as precursors to alkylidene carbenes is attractive as these functionalities are readily available via carbonyl olefination. 44d,47,51 However, other methods of accessing alkylidene carbenoids, which are not reliant on manipulation of a carbonyl functional group, have also been reported. Suzuki described the generation of zirconium alkylidene carbenoids via allylzirconation of iodoalkynes in the presence of methylaluminoxane (MAO). 52 The allylzirconium reagents themselves were accessed via hydrozirconation of allenes (Scheme 30).
This methodology was applied to the synthesis of a range of aryl-substituted enynes from aryl iodoalkynes (Scheme 31).
When the same methodology was applied to alkyl-substituted iodoalkynes, the products observed were again the product of 1,2-migration, even in cases where 1,5 CeH insertion was possible (Scheme 32). Higher temperatures were required for this reaction, and lower yields of the products were observed. NMR labelling studies of the reaction showed a clear difference in how the two substrate classes proceeded through the reaction. When the 13 C-labelled iodoalkyne 69 was subjected to the reaction conditions, the product 70 had the 13 C label directly attached to the naphthyl group, consistent with a 1,2-shift of the aryl group. However, in the reaction with the labelled alkyl iodoalkyne 71, the product 72 showed no change in the position of the 13 C label, suggesting that, in this case, the migrating group was the allyl substituent (Scheme 33).
These results suggest that, in the absence of more suitable migratory groups, the allyl functionality is able to undergo a 1,2migration reaction, giving rise to alkynes. The lack of evidence of any 1,5 CeH insertion products suggests that this process is faster than any competing insertion pathways.
Normant has described the synthesis of alkynes from zinc carbenoids, generated from bismetalloalkenes. 53 Treatment of alkynyllithium 73 with allyl Grignard gave rise to bismetalloalkene 74, which reacts stereoselectively with phenylsulfonyl chloride to give alkylidene carbenoid 75. 54 At low temperatures, 75 can be quenched with a second halogenating reagent. However, if the reaction is warmed to room temperature, 75 cleanly rearranges to give the alkyne 76 (Scheme 34).
Normant was able to demonstrate the stereospecificity of the migration. By quenching the bismetalloalkene 78 with acid, it could be seen that the initial addition of the allyl Grignard to the alkynyllithium 77 occurred with a diastereomeric ratio of 92:8. Since 80 is also formed as a 92:8 mixture of diastereoisomers, the generation of alkylidene carbenoid 79 and its subsequent rearrangement to alkyne 80 occurs in a stereospecific fashion (Scheme 35).
To determine whether the migration had occurred with retention or inversion, 80 was fully reduced and the MEM group removed to give 81. By comparing the sample of 81 synthesized in this manner to a sample of known stereochemistry, it was shown to be the syn isomer. Thus, it was concluded that the migration occurred with complete retention of absolute configuration.
NMR studies were again employed to determine which group underwent migration. 53b Generation of 13 C-labelled carbenoid 82 gave rise to enyne 83, indicating that the ether carbon trans to the chlorine was migrating in this case (Scheme 36).
The absence of the chelating substituent drastically reduced the stereoselectivity of the carbenoid formation, as was seen by the low E/Z ratios of vinyl chlorides 86aec formed when carbenoid 85 was quenched with acid (Scheme 37). However, this appears to have had little or no bearing on the outcome of the reaction, as the formation of enynes occurred with higher selectivity. The reaction of 84a gave 87a and 87b in a ratio of 92:8 indicating that the allyl group preferentially migrates, regardless of whether it is cis or trans to the chloride leaving group. Similarly, the observation that 88a is the major product in the reaction of 84b indicates preferential migration of the substituted allyl group over the cyclohexyl, albeit less so than the allyl group itself. Additionally, the predominance of 89b shows that the octyl group migrates in preference to the substituted allyl, even if it is primarily cis to the chlorine.
The extensive work on zirconium and zinc carbenoids, by Suzuki and Normant, respectively, has demonstrated that under the right conditions, alkyl-substituted alkylidene carbenoids undergo 1,2-migration reactions to form alkynes. It is also important to note that no evidence of 1,5 CeH insertion was reported, despite this generally being the preferred pathway for alkylidene carbenes bearing alkyl groups. While there have been several reports of differing reactivity of alkylidene carbenoids and 'free' carbenes (vide infra), 55 these results suggest that altering the coordinating metal may also affect the reactivity. Normant reported that the corresponding lithium carbenoid gave a complex mixture of products, with the desired alkyne isolated in less than 10% yield. 53a Alkylidene carbenoids have also been generated via the intramolecular cyclization of alkynyllithiums. 56 Harada reported a tandem cyclization process upon treating alkynes with base (Scheme 38). 56b While the reactions were largely carried out with iodides, this transformation could also be performed on sulfonates. NaHMDS or KHMDS served as alternative bases to LDA.
Harada proposed that the reaction proceeded via 5-exo-tet cyclization of the nucleophilic alkynyllithium, generating the alkylidene carbenoid 90. This subsequently underwent CeH insertion, forming the second ring with high diastereoselectivity (Scheme 39). This route stands in contrast to the synthesis of similar systems using alkynyliodonium species, where the alkyne serves as an electrophile (vide infra). 32b,57

Metal vinylidenes
Catalytic methods for accessing alkylidene carbenoids have also been developed, with metal vinylidene complexes being proposed as intermediates in several carboneheteroatom and carbonecarbon bond forming processes. 58 Yamamoto reported the conversion of aryl-substituted terminal alkynes into functionalized indenes using catalytic PtBr 2 (Scheme 40). 59 Yamamoto proposed that the reaction proceeded via isomerization of the alkyne to the Pt-vinylidene complex, which underwent 1,5 CeH insertion to give the product. Liu 60 and Chatani 61 have reported similar reactions using platinum and ruthenium catalysts, but suggested that the vinylidene intermediate didn't undergo direct CeH insertion, instead proceeding through a 1,5 hydride shift followed by a cycloisomerization. In 2015, Gonz alez utilized iodoalkynes in a similar reaction, using a gold catalyst to synthesize 3-iodoindenes via 1,5 CeH insertion. 62 Zhang and Hashmi independently reported the dual activation of aromatic diynes, 63 using a gold catalyst to access the intermediate gold vinylidene via a carbocyclisation route similar to that described by Harada. 56 Zhang reported the synthesis of a range of fused polycycles from diynes employing a strategy, which utilized the activation of both alkynes (Scheme 41). 63e Zhang proposed that the reaction proceeds through the formation of alkynylgold complex 91, with the N-oxide acting as a base to remove the proton from the terminal alkyne. The internal alkyne is then activated by another molecule of the catalyst, giving 92, which undergoes a 5-endo-dig cyclization, resulting in the formation of gold vinylidene 93. A 1,5 CeH insertion reaction of 93 leads to 94, which gives the product after protodeauration (Scheme 42). This mechanism was supported by computational studies. 63e Propargylic alcohols and amines undergo analogous OeH and NeH insertions, leading to furans and pyrroles.
Hashmi reported the synthesis of similar polycyclic systems from aromatic diynes, again employing a double activation strategy under gold catalysis. 63aed,f Van der Eycken utilized this methodology to synthesize cyclopentapyridinones, 64 while Ohno reported the synthesis of bi-and tricyclic pyrroles from N-propargyl ynamides under similar conditions. 65 Hashmi also reported the synthesis of benzothiophenes and benzofurans using similar chemistry, albeit via a different mechanism. 66 The need for dual activation can be avoided by generating gold alkylidene carbenoids via cyclization onto alkyl sulfonates. 67 Metal vinylidenes such as those described above also undergo reaction with external nucleophiles. Osmium, platinum and ruthenium vinylidenes have been shown to react with H 2 O to form hydroxy carbenes, which then tautomerize to the metal acyl species. 68 These can then undergo elimination to the aldehyde, or extrude CO to form the reduced organic compound and the corresponding metal carbonyl (Scheme 43). Fukumoto reported the synthesis of nitriles via reaction of the intermediate ruthenium vinylidene with N,N-dimethylhydrazine. 69 In 2015, Hashmi combined this reactivity with the dual gold activation of diynes to synthesize functionalised indenes in good yields (Scheme 44). 70 The use of metal carbenoids offer a viable alternative to access alkylidene carbenes, particularly since many can be accessed from the same carbonyl precursors as diazo olefination reagents. The research by Normant and Suzuki into zinc and zirconium carbenoids, respectively, has not only expanded the techniques by which alkylidene carbenoids can be generated, but has also increased the understanding of the reactivity of these useful intermediates. The formation of metal vinylidenes under catalytic conditions is a significant breakthrough as previous reactions involving alkylidene carbenes required stoichiometric generation of the reactive intermediates.

Hypervalent iodine
The use of hypervalent alkynyliodides is a complementary method for alkylidene carbene generation. 4a,b,d In contrast to the methods above, which typically rely on carbonyl precursors, the use of hypervalent halides allows access to heteroatom-substituted alkylidene carbenes.
Alkylidene carbenes are generated via the reductive elimination of iodonium ylides. Similar to the deprotonation of vinyl halides, treating hypervalent vinyl iodonium salts with base leads to an iodonium ylide. Ochiai utilized this approach in the synthesis of alkynyl sulfides and sulfoxides. 71 This process demonstrated the A much more common approach to hypervalent vinyl iodonium ylides is the addition of nucleophiles to alkynylaryliodonium salts, also referred to as 1-alkynyl(aryl)-l 3 -iodanes. 4a,b,d Aryliodine(III) groups are strongly electron-withdrawing, 72 as well as demonstrating considerable leaving group ability, 73 making them ideal for this kind of transformation. The initial Michael addition to the iodonium species results in an iodonium ylide, which reductively eliminates the aryl iodide to give the corresponding carbene, with the initial nucleophile being incorporated into the final product (Scheme 46). This is a versatile methodology, allowing for introduction of functionalities which may otherwise be difficult to incorporate. For example, Stang reported the synthesis of cyclic vinyl azide 96 through the treatment of alkynyliodonium tosylate 95 with sodium azide (Scheme 47). 74 A wide range of other nucleophiles have been used in intermolecular reactions to generate alkylidene carbenes from alkynyliodonium salts, including sulfonamides, 32b phenoxides, 75 sulfinates, 76 1,3-dicarbonyls, 77 tropolone, 78 thiocyanates, 79 tellurides and selenides. 80 This methodology has been developed by Wipf for the synthesis of 1,3-thiazoles, employing thioamides as the nucleophile with alkynyliodonium mesylates (Scheme 48). 81 While Wipf proposed two possible mechanisms, 81 Ochiai later demonstrated that the reaction proceeded through initial Michael addition of the thioamide to the alkynyliodonium salt, to afford the alkylidene carbene 97, which underwent a subsequent 1,2- migration to form alkyne 98. This alkyne then undergoes a 5-endodig cyclization to form the thiazole (Scheme 49). 82 Ochiai proposed that the known migratory aptitude of sulfenyl groups made this mechanism feasible, 71 and was able to provide evidence of this mechanism by performing the reaction with N,Ndimethylthiobenzamide. After the proposed alkyne formation, the thiobenzimidonium salt 99 generated would be unable to cyclize and its isolation provided strong evidence for 98 being an intermediate in thiazole formation (Scheme 50).
Kitamura reported the alkynylation of benzotriazole with alkynyliodonium triflates, where the initial conjugate addition is followed by 1,2-migration of the intermediate alkylidene carbene (Scheme 51). 83 There is a preference for alkynylation at N-2 with the silylated alkynes, and at N-1 in the absence of silicon, believed to be due to unfavourable steric interaction between the benzotriazole and the silyl group, or iodonium substituents, in the event of nucleophilic attack from N-1 (Fig. 6). These interactions are minimized in attack at N-2. Such interactions are absent in the terminal alkyne, and so attack from N-1 is preferred. The use of alkynyliodonium salts has also led to the discovery of new alkylidene carbene reactivity. It has been demonstrated that the 1,5 CeH insertion is not limited to sp 3 carbon centres. 75,78,84 Kitamura reported the formation of benzofuran systems upon treating alkynyliodonium salts with sodium phenoxide (Scheme 52). 84b It is interesting to note that the formation of benzofurans is preferred over insertion into the alkyl side chain, with no formation of cyclic enol ethers seen. This selectivity is believed to be due to an interaction between the empty p orbital of the alkylidene carbene and the p-system of the aromatic ring, resulting in aromatic 1,5 CeH insertion becoming more favourable than aliphatic 1,5 CeH insertion. 84b Croatt described the generation of cyanocarbenes via an intermediate alkylidene carbene. 85 Treating alkynyliodonium salts 100 and 101 with tetrabutylammonium azide gave rise to cyanocyclopropanes 102e104 after reaction with the alkene solvent (Scheme 53).
The reaction is thought to proceed through initial formation of the alkylidene carbene 105 via nucleophilic attack on the b-position of the alkynyliodonium species, as described previously. Subsequent 1,2-migration results in the formation of alkynyl azide 106, which extrudes N 2 to form the cyanocarbene 107. This is then trapped by the alkene to give the cyclopropane 102 (Scheme 54). Feldman reported formal 1,5 CeO insertion in the reaction between alkynyliodonium salts 108 and sodium para-toluenesulfinate (Scheme 55, Table 2). 26b,76b The reaction is believed to proceed via  initial oxonium ylide 109 formation followed by subsequent migration to give the disubstituted dihydrofurans 110. Feldman reported that only groups, which offered substantial carbocation stabilization underwent migration, as evidenced by the lack of reaction of benzyl and allyl ethers. In addition, a significant amount of the protonated dihydrofuran 111 was observed in most cases.
The formation of 111 in the reaction of 108c suggested that the oxonium ylide 109 was formed but reacted via a different pathway other than the desired migration. In particular, the isolation of small quantities of p-TolSO 2 CH 2 Ph suggested nucleophilic attack a to the oxonium ylide interfered with the desired 1,2-shift. In addition, the reactions of 108d and 108e suggest that the migration occurs largely with retention of absolute configuration.
Intramolecular nucleophilic attack to form alkylidene carbenes from alkynyliodonium salts has also been reported. 57 Examples will be discussed in the section on diastereoselectivity.
There are a number of methods that can be employed in order to synthesize alkynyliodonium salts. 77c,86 The most commonly employed route is the reaction of alkynylstannanes with Stang's reagent (PhI(CN)OTf). 86g,h This method allows for the synthesis of alkynyliodonium triflates incorporating a range of functionalities including alkyl, aryl, heteroaryl, halogens, carbonyls and sulfones, in yields ranging from 42% to 89%. 86h However, a major drawback of the route is the release of stoichiometric HCN. In addition, it is also necessary to prepare Stang's reagent prior to use. 29c,30b,77c,87 Alternative routes have been developed that allow for the synthesis of alkynyliodonium salts using PhI(OAc) 2 in the presence of TfOH, Tf 2 O or TFA. 88 These developments have also allowed for the use of alkynylsilanes and terminal alkynes instead of alkynylstannanes.
Olofsson reported the synthesis of alkynyliodonium salts directly from terminal alkynes 89 and alkynyl boronates. 90 Moran used this methodology to explore the effect of the aromatic iodine species on migration and insertion reactions of alkylidene carbenes. 91 Superior results were achieved using a 2-anisyl substituent (Scheme 56). These results can be contrasted with those of Carroll, 88b who reported altering the aromatic substituent had minimal effect on the 1,5 CeH insertion into an aromatic system. As part of the same study, Moran reported that altering the counter ion had no discernible effect on the formation of alkynyl sulfones. 91 Waser reported the use of ethynyl-1,2-benziodoxol-3(1H)-one 113 (EBX) as a reagent for the alkynylation of electron-deficient esters. 92 EBX was generated in situ by treating TMS-EBX 112 with TBAF. Subsequent reaction with the ester led to a-alkynylation (Scheme 57). TBAF also acts as a base within the reaction.
TMS-EBX 112 can be prepared simply by partial oxidation of 2iodobenzoic acid with NaIO 4 , followed by reaction with bis(trimethylsilyl)acetylene in the presence of TMSOTf. 93 The reaction was shown to proceed through an alkylidene carbene by the use of 13 C labelling (Scheme 58). By labelling the alkyne carbon bonded to the silicon in TIPS-EBX 114, and determining its presence in the final product as being directly bonded to the aposition of the ketone, it suggests that initial nucleophilic attack  occurs on the b-carbon of the alkynyliodonium species, leading to alkylidene carbene 115. This also suggests that the hydrogen undergoes migration to form the final product. This reaction has been rendered asymmetric, albeit with modest enantioselectivity, through use of a chiral phase transfer catalyst. 92 TIPS-EBX 114 has also been used for the synthesis of alkynyl sulfides, 94 although the mechanism is not believed to proceed through an alkylidene carbene intermediate. 95 While the use of alkynyliodonium salts has allowed for the synthesis of cyclopentenes and alkynes with a wide range of functionalities, it usually requires the pre-functionalization of an alkyne with the hypervalent iodine. Recently, Muniz has reported the synthesis of ynamides and cyclic enamides from terminal alkynes using a range of iodine(III) reagents, which are readily available from PhI(OAc) 2 . 96 The reaction was shown to be successful in synthesizing a range of bis-tosyl ynamides from the corresponding acetylenes, using hypervalent iodine reagent 116 (Scheme 59).
The methodology has also been applied to the synthesis of cyclic enamides. Treating alkyl alkynes with 116 gave cyclization products in good yields (Scheme 60).
The proposed mechanism is as follows: the initial coordination of the electrophilic iodine(III) to the acetylene, forming complex 117, increases the acidity of the alkyne CeH bond. Deprotonation of the alkyne leads to alkynyliodonium 118. Nucleophilic attack on the b-carbon by the tosylamide anion, results in alkylidene The relative ease of access to alkynyliodonium species, coupled with the wide range of functionalities that can be incorporated into the alkylidene carbenes generated from them, makes them important reagents in the use of alkylidene carbenes in organic synthesis. The work of Kitamura and Carroll has allowed for the generation of these versatile reagents under milder, safer conditions, while the recent developments by Muniz have allowed for the in situ generation and reaction of alkylidene carbenes directly from the terminal alkyne, eliminating the need to prepare alkynylstannanes or alkynyliodonium species.

Hypervalent bromine
Expansion on the extensive work carried out on alkynyliodonium species as alkylidene carbene precursors has led to the investigation of the use of related alkynylbromonium, or 1alkynyl(aryl)-l 3 -bromane, substrates in a similar fashion. 97 l 3 -Bromanyl groups have a greater Hammett substituent constant than the equivalent l 3 -iodanyl (s p ¼1.63 for PhBrBF 4 compared to s p ¼1.37 for PhIBF 4 ), suggesting that 1-alkynyl-l 3 -bromanes may be better Michael acceptors. 98 Ochiai reported the synthesis of a number of alkynylbromonium tetrafluoroborates. Treatment of aryl silane 120 with BrF 3 gave rise to 121, which subsequently reacted with alkynylstannanes to give the alkynylbromonium tetrafluoroborates 122aeg (Scheme 62). 99 Like the hypervalent iodine reagents described above, alkynylbromonium reagents undergo reaction with sulfinate nucleophiles to give the corresponding cyclopentenes. 100 However, Ochiai demonstrated less reactive nucleophiles such as tosylates and mesylates react to give alkynes 123aed and cyclopentenes 124a,b (Scheme 63). 99 Reaction with weak nucleophiles such as sulfonate anions, shows that these reagents are considerably more reactive than the corresponding iodonium reagents. Sulfonate anions do not react with alkynyliodonium salts. 74,86e The use of sodium trifluoromethanesulfinate (triflinate) as a nucleophile gave rise to a mixture of vinyl triflone 125 and alkynyl triflone 126 in good yield. A similar reaction between 127 and sodium triflinate gave the bicycle 128 with excellent selectivity (Scheme 64).
The range of nucleophiles that underwent reaction with alkynylbromonium species was expanded to include metallated alkynes. Ochiai reported that treating alkynylstannanes with 121 resulted in an oxidative homocoupling (Scheme 65). 101 It is thought that the reaction proceeds through the initial formation of 122, followed by the formation of bromonium ylide 130 and subsequent elimination to the alkylidene carbene.
The methodology was applied to various alkyneealkyne oxidative cross-couplings between alkynylbromoniums and various alkynylstannanes to give unsymmetrical diynes. 101 Ochiai also reported the synthesis of alkynyl triflate 131, by reacting alkynylbromonium 122b with a large excess of sodium triflate (Scheme 66). This example in particular demonstrates the propensity of alkynylbromonium to act as Michael acceptors, as triflate anions are generally held to be non-nucleophilic in nature.
Alkynylbromonium species holds great promise as alkylidene carbene precursors, particularly in the use of poor nucleophiles such as tosylates and triflates; however, their application is currently limited by the preparation of difluorobromonium reagents, due to the high toxicity and extreme reactivity of the BrF 3 required for their formation. 102

Vinyl aminosulfoxonium salts
Gais reported the use of chiral vinyl aminosulfoxonium salts as precursors to alkylidene carbenes in the synthesis of enantiomerically enriched homopropargylic alcohols. 103 Treatment of the salts with base gave the corresponding sulfur ylide 132, from which loss of sulfinamide 133 gave the alkylidene carbene 134. Subsequent 1,2-migration gave the chiral homopropargylic silyl ethers 135 (Scheme 67).
Gais also employed this methodology for the asymmetric synthesis of dihydrofurans 137 via 1,5 OeSi insertion of the  104 The use of a sulfoximine as a both chiral auxiliary (e.g. in the synthesis of 136) and as an alkylidene carbene precursor is an attractive feature of this methodology, particularly as the chiral sulfonamide by-product can be recovered in excellent yields (in excess of 85%) and recycled for use in subsequent enantiopure sulfoximine synthesis. 105

Substrate effects on selectivity
Due to carbenes being strongly electron-deficient, they are highly electrophilic and so react faster with functionalities which are relatively electron-rich. Such is the sensitivity of alkylidene carbenes to these electronic effects, they preferentially undergo 1,5 CeH insertion reactions with tertiary CeH bonds in the presence of less substituted 1,5 CeH insertion sites. 106 The relative rates of reaction have been demonstrated to be 1:30:240 (primary/secondary/tertiary). 106 It has also been shown that heteroatoms adjacent to CeH bonds accelerate reaction at that site. 10,44e,55a,107 Insertions into OeSi bonds have been shown to occur faster than CeH insertion reactions, with Kim reporting the formation of dihydrofuran 139 from azirindinylimine 138 with no evidence of 140 seen. However, this is only true for the formation of fivemembered heterocycles as employing azirindinylimines 141 in the same reaction gave mainly cyclopentenes 142 via 1,5 CeH insertion, with heterocycles 143a and 143b, formed via 1,6 OeSi insertion and 1,7 OeSi insertion, respectively, isolated in only low yields (Scheme 69). 26a Feldman has also demonstrated that the formation of fivemembered heterocycles via formal insertions into heteroatomehydrogen bonds, occurs in preference to competing 1,5 CeH insertion reactions. 108 While alkylidene carbenes bearing aryl substituents usually undergo 1,2-migration reactions in preference to 1,5 CeH insertion reactions, Shioiri demonstrated that 1,5 OeSi insertion reactions can compete with the migration of phenyl groups, as shown by the reaction of lithiated trimethylsilyldiazomethane 17 with ketone 144 giving a mixture of dihydrofuran 145 and alkyne 146 (Scheme 70). 25 Shioiri 25 and Kim 26a both reported that 1,2 hydrogen migration occurs in preference to 1,5 OeSi insertion.
Despite the tendency for alkyl-substituted alkylidene carbenes to undergo insertions reactions, Wills demonstrated that it is possible for 1,2-alkyl shifts to occur and even predominate. 109 The reaction of ketone 147a with lithiated trimethylsilyldiazomethane 17 gave alkyne 148a as the major product with the corresponding dihydrofuran 149a isolated in only low yield (Scheme 71). The methyl ketone 147b gave the dihydrofuran 149b exclusively, and it was postulated that the increased electron-donating capabilities of the isopropyl group resulted in its preferential migration to form the alkyne. 109 Substrate effects can also have a significant influence on alkylidene carbene insertion reactions. Gilbert  Gilbert proposed that diazoalkene 156, from the initial Hor-nereWadswortheEmmons reaction between 150 and 151, would be close to planar in order to maintain p-delocalization throughout the system. Rationalizing that the s-trans conformation 156a would be more favourable than the s-cis 156b, due to decreased dipoleedipole interactions, this would result in an alkylidene carbene with a suitable CeH insertion site in close proximity. Conversely, due to more favourable steric interactions, the more conformationally flexible diazoalkene 157 is likely to exist in a conformation such as 157a, which lacks a proximal site for CeH insertion, allowing intermolecular reaction with the solvent to dominate (Scheme 72).
Performing the reaction on unsymmetrically substituted amides gave a mixture of 3-pyrrol-2-ones 159 and 160, with insertion into the N-methyl group preferred over insertion into the N-methylene, along with ynamides 161 (Scheme 73). 106 The reaction between 158 and 151 leads to diazoalkene 162, which can exist in two rotameric forms, both of which are in a suitable conformation for efficient CeH insertion (Scheme 73). Gilbert proposed that the increased steric hindrance between the larger N-alkyl substituent and the diazoalkenyl group resulted in 162b being less favoured, allowing insertion into the primary CeH bond to dominate, and hence the reverse of typical selectivity.
Grainger and Owoare reported preferential CeH insertion over the normally favoured OeSi insertion in synthesizing tetracycle 164 from vinyl chloride 163 (Scheme 74). 111 It was proposed that in order for OeSi insertion to occur, there must be an interaction between the oxygen lone pair and the empty p-orbital on the alkylidene carbene. In this example, the bulky TMS group is expected to be orientated exo to the ring system in order to minimize any unfavourable steric interactions, as in 165. This would result in both oxygen lone pairs siting in an endo orientation, unable to interact with the empty p-orbital. In order to achieve the necessary orbital overlap, the molecule must adopt a much higher energy conformation, such as 166 or 167 (Scheme 75). As a result, the 1,5 CeH insertion pathway dominates.
However, further work on the reactivity of alkylidene carbenes on 8-oxabicyclo[3.2.1]oct-6-ene systems by Grainger and Munro demonstrated that this selectivity was not absolute. A comparison As expected, the 2-benzyloxy group controls the regioselectivity of the CeH insertion reaction; however, its introduction also leads to OeSi insertion. Moreover, the ratio of CeH insertion to OeSi insertion is contingent on the method employed to generate the alkylidene carbene. The reaction of vinyl chloride 173 with NaHMDS gave the same mixture of products, this time in a w1:1 ratio (Scheme 77). 55b The reaction of the dimethyl analogue 174 again gave cyclopentene 169 as the only product. 111 The change in product ratio in the reactions of 170 and 173 can be attributed to the nature of the carbene generated; 170 leads to a reactive free alkylidene carbene as opposed to the less reactive carbenoid formed from 173 (vide infra), however, it is unclear why such a drastic change in reactivity occurs upon the introduction of the benzyloxy substituent. Further examples of the effects of reaction conditions on the selectivity of alkylidene carbenes will be discussed later in this review.
H.-Y. Lee has shown that it is possible to override the preference for alkylidene carbenes to insert into tertiary CeH bonds over secondary CeH bonds (Scheme 78). 112 The alkylidene carbene generated from ketone 175 was shown to react exclusively with the tertiary CeH bond giving tetracycle 176 in good yield. However, the epimer 177 shows preferential insertion into the secondary CeH bond, giving 178. H.-Y. Lee suggested that an unfavourable steric interaction between the OTMS group and the ring system in 177 disfavours the conformation for insertion into the tertiary CeH bond, allowing insertion to occur at the electronically less favoured methylene site.
Feldman demonstrated how the reactivity of the intermediate alkylidene carbene could be controlled through substrate modification (Scheme 79). 113 Treatment of alkynylstannane 179 with Stang's reagent, followed by reaction of the crude iodonium salt with NaSO 2 Tol, gave a mixture of vinyl sulfone 180 and dihydrofuran 181 (formed via initial reaction between the alkylidene carbene and the oxygen lone pair) in poor yield, with 181 as the major product. 113 It was postulated that replacing the gem-dimethyl group with a more electron-withdrawing substituent would reduce the reactivity of the oxygen lone pairs towards the alkylidene carbene, allowing the desired CeH insertion reaction to occur preferentially. Thus, alkynylstannane 182 was synthesized and subjected to the same reaction conditions, leading to a mixture of products. Feldman proposed that bis-sulfone 185 arose from vinyl sulfone 183, and so the ratio of CeH insertion to reaction at oxygen had increased from 1:3 to 1.5:1. Alkynylsulfone 186 was also isolated in low yield.
When the N-benzyl substituent in 182 was replaced with N-Boc a mixture of vinyl sulfone 188 and alkynylsulfone 189 was isolated. Feldman proposed that employing the electron-withdrawing Boc group results in the nitrogen lone pair being delocalized primarily into the carbamate moiety, and the oxygen lone pair being delocalized into the oxazolidinone carbonyl. As such, the oxygen lone pair is less available to interact with the alkylidene carbene, suppressing reaction at oxygen. However, the introduction of the Boc group most likely accounts for the increased formation of 189, as the CeH bond at which insertion is desired is now more electrondeficient and so CeH insertion occurs more slowly, allowing the normally slower 1,2-sulfonyl shift to compete. 113 Cox reported an interesting change in regioselectivity of a 1,5 CeH insertion reaction adjacent to nitrogen upon decomplexation of a Fe(CO) 3 moiety from a tethered diene (Scheme 80). 114 Performing the reaction on the iron complex 190 gave a mixture pyrrolizidines 191 and 192 in moderate yield, with preferential insertion into the tertiary CeH bond. However, reaction of the free diene 193 led to a 1:1 mixture of 194 and 195 in better yield. Cox postulated that the Fe(CO) 3 group led to increased steric crowding of the methine CeH bond, resulting in a longer and weaker bond. He also suggested that anchimeric assistance from the iron complex would electronically activate the tertiary CeH bond towards insertion, resulting in improved regioselectivity. Both 192 and 195 were formed diastereoselectively, with the diene substituent on the exo face of the [3.3.0] bicyclic ring system. Although preferential insertion into CeH bonds adjacent to heteroatoms was well known in the literature prior to 2009, a comprehensive study by D. Lee and co-workers clearly demonstrated that 1,5 CeH insertion reactions adjacent to oxygen are subject to a subtle combinations of steric, electronic, conformational and stereoelectronic effects. 115 Indeed, in a systematic study of reactions of lithium trimethylsilyldiazomethane 17 with various oxygenated ketones, it was shown that the preference for CeH insertion next to oxygen could in fact be overridden. The following conclusions could be made: While adjacent oxygen atoms increase the rate of 1,5 CeH insertion, this is more pronounced in systems where the oxygen atom is part of the newly formed ring system. The increased activation of endocyclic oxygen is relatively subtle and the use of mildly electron-withdrawing substituents can diminish its effect, allowing competing insertion next to exocyclic oxygen atoms to dominate. 1,5 CeH insertion into more conformationally constrained systems is slower than reaction in more flexible systems. 1,5 CeH insertion does not occur into the CeH of an epoxide. The steric environment of the CeH bond undergoing insertion can affect the rate at which reaction occurs. However, these effects are less significant than electronic and conformational effects.
Of particular note is how stereoelectronic effects influence the selectivity of 1,5 CeH insertion. 115 In alkylidene carbene 196, D. Lee proposed that the most important orbital interaction would be between the axial non-bonding orbital of the oxygen and the s* The preferred formation of 202c and 202d through insertion into the apparently less activated CeH b bond was deemed to be due to the stereoelectronic effect that D. Lee had previously proposed. 115 The formation of allenylsilanes 203aed was thought to be due to an intermolecular process that occurs because of the lower reactivity of CeH bonds in conformationally constrained systems compared with those in more flexible systems. 115b D. Lee was able to selectively synthesize these interesting functionalities by altering the reaction conditions (vide infra).
This work by D. Lee et al. has offered significant insight into the role played by oxygen in affecting the selectivity of 1,5 CeH insertion reactions, as well as addressing how various other subtle changes can drastically alter how this reactive intermediate will behave.

Effect of reaction conditions on selectivity
While significant changes in the selectivity of alkylidene carbenes can generally only be brought about through substrate modification, there have been reports where altering the reaction conditions under which the alkylidene carbene is generated have led to changes in selectivity. Gilbert reported that increasing the polarity of the solvent used for the reaction of 204 with 151, resulted in an increase in the formation of alkyne 206 via migration, compared to CeH insertion to form 3-pyrrol-2-one 205, with an eventual reversal of selectivity occurring in water (Scheme 84). 110 Taber demonstrated a similar change in chemoselectivity in a competition between NeH insertion and 1,2-aryl migration, the ratio of which could be altered by changing the reaction solvent. 116 Increasing the polarity of the solvent resulted in the increased formation of indole 208, in comparison to alkyne 207, when generating the alkylidene carbene from 206 (Scheme 85).
An earlier study by Garcia-Gabiray on alkyl carbenes suggested that less polar solvents could give rise to the triplet state, and thus differing reactivity. 117 However, Taber believed that both 207 and 208 arose from the singlet alkylidene carbene and so suggested that the change in reactivity was more likely as a result of a solventinduced change in the hybridization of the alkylidene carbene.
The effect of altering the solvent and temperature on alkylidene carbene selectivity was seen by Hayes while investigating the reaction of vinyl bromide 209 with KHMDS as a route to (þ)-lactacystin (Scheme 86). 118 The ratio of 1,5 CeH insertion to 1,2migration increased by performing the reaction in THF instead of Et 2 O. Further increases could be achieved by performing the reaction at lower temperatures. Wardrop had previously investigated the same reaction in his formal synthesis of (þ)-lactacystin and described the ratio of products obtained was also dependant of the geometry of the starting vinyl bromide. 119 Treating a mixture of (E/Z)-209 with KHMDS in ether, gave a mixture of dihydropyrrole 210 and alkyne 211 in a w1.4:1 ratio. However, when a pure sample of (E)-209 was reacted under the same conditions the efficiency of the insertion reaction improved, giving the mixture of 210 and 211 in a w2.7:1 ratio. Conversely, a pure sample of (Z)-209 gave preferential formation of alkyne 211, as a 3:1 mixture with 210. 119 The latter result is somewhat surprising as it has been shown that the alkene geometry has little effect on the efficiency of CeH insertion reactions. 21a Wardrop offered no explanation for this unusual phenomenon, however, subsequent work by D. Lee suggests that insertion into the desired CeH bond may be retarded due to the conformation rigidity of the dioxane system, 115 coupled with the proximity of the bulky i Pr group, allowing the alkyne formation to compete. It remains unclear why the efficiency of this reaction should depend on the geometry of the starting vinyl bromide.
In 2012, Xi also demonstrated a profound change in reactivity when altering the solvent in which the reaction of dibromoalkenes with n BuLi was performed (Scheme 87). 120 When the reaction was carried out in hexane, the alkyne was formed via 1,2 aryl migration. However, when THF was used, the butatriene (mainly E) was Xi suggested that the difference in polarity and coordination ability of the solvents was responsible for this significant change in reactivity. 120 Taber reported an increase in the regioselectivity of 1,5 CeH insertion reactions by altering the manner in which the alkylidene carbene was generated. 55a Treating ketone 212 with TMSCHN 2 and MeLi resulted in a mixture of dihydrofuran 213 and cyclopentene 214, with a slight preference for insertion into the primary CeH over the tertiary CeH. When the same ketone was converted to the vinyl chloride and treated with KHMDS, the ratio of 213 to 214 increased, again in favour of the dihydrofuran, in accordance with the increased activation effect seen from endo oxygen over exo oxygen (Scheme 88).
Taber suggested that carbenes generated via a-elimination are more carbenoid in nature, and their reactivity is tempered by the coordination of metal salts, resulting in increased regioselectivity in the CeH insertion reaction.
As has already been described (Scheme 83), D. Lee reported the formation of allenylsilanes in the reactions of substrates where the CeH required for insertion was relatively unreactive. By increasing the relative amount of TMSCHN 2 and n BuLi in the reaction, as well as increasing the reaction concentration, D. Lee was able to form the allenylsilane preferentially. Hence the formation of 216 from 215 could be achieved in good yield, whereas 217 gave only the product of 1,5 CeH insertion due to the activating effect of the adjacent oxygen (Scheme 89).
D. Lee proposed that the allenylsilane arose from an intermolecular reaction between the alkylidene carbene 219 and trimethyldiazomethane to form 220. Subsequent elimination of nitrogen would result in the formation of allenylsilane 221 (Scheme 90).

Diastereoselectivity
Alkylidene carbenes can sometimes undergo highly diastereoselective reactions, although the diastereoselectivity has not always been rationalized. Examples previously reported in this review include the formation of 53 (Scheme 19) and the 1,5 CeH insertion of alkylidene carbenoid 90 (Scheme 39). 38,56 In the formation of fused bicyclic systems, the cis ring junction is usually formed, 57a,b,112,113,115,121 however trans formation is clearly possible as seen in the formation of 176 (Scheme 78) and to some extent 201c and 202c (Scheme 83). 112,115 In 1994, Taber reported the first diastereoselective 1,5 CeH insertion reaction, synthesizing cyclopentenes 222 and 223 as a 4.4:1 mixture of diastereoisomers (Scheme 91). 122 Taber proposed that this selectivity arose from more favourable interactions within the transition state of the insertion reaction (Scheme 92). It was calculated (using an MM2 force field and a 'weak bond' between the carbenic atom and the H atom) that transition state 224, with the methyl group in a pseudoequatorial position, was 1.5 kcal/mol lower in energy than 225, resulting in the formation of the anti-isomer 222 being more favourable than the syn isomer 223. 7 Taber also reported 1,3 diastereomeric induction in the reaction of ketone 226 to form cyclopentenes 227 and 228 in a 2.4:1 ratio (Scheme 93). 123 The benzyl group was removed to give the natural product a-necrodol, and its cis diastereoisomer. 123 Taber proposed four possible transition states 230e233 for the 1,5 CeH insertion reaction (Fig. 7). Calculations (using an MM2 force field and a 'weak bond' between the carbenic atom and the H atom) showed that 230 and 231, which would lead to the trans product, were lower in energy than either 232 or 233, which would give the cis product. It is notable that 232, with two pseudoequatorial substituents, is thought to be less stable that either 230 or 231, which each contain one pseudoaxial substituent. This is believed to be due to the increased gauche interactions in 232 between the gem-dimethyl group and both pseudoequatorial substituents. 123 Interestingly, the major product in the reaction above was cyclopropene 229, the result of 1,3 CeH insertion (Scheme 93).
Taber postulated that 1,5 CeH insertion would be retarded by the boxygen substituent, while 1,3 CeH insertion could be accelerated due to it being a tertiary CeH bond as opposed to the secondary CeH undergoing 1,5 CeH insertion, resulting in the cyclopropene being the major product.
Wills also reported 1,3-diastereomeric induction in the synthesis of dihydrofurans 235 and 236 from ketone 234 (Scheme 94 Proposing similar transition states 237e240 to Taber,7,123 Wills argued that the formation of 235 was more favourable due to the diminished steric interactions present in 237 by having both the CH 2 OTBS and phenyl groups in pseudoequatorial positions (Fig. 8). 124 Van Nhien and Postel reported diastereoselective induction in 1,5 CeH insertion reactions on a sugar-derived system. 9,125 Treatment of cyano mesylate 241 with TMSN 3 and Bu 2 SnO gave trans isomer 242 as the major product (Scheme 95). 9 Calculations (B3YLP with 6-31(d,p) basis set) showed the transition state leading to 243 was higher in energy by w2.5 kcal/mol than the transition state leading to 242, due to electrostatic repulsion between the dioxolane and benzoyl groups. 9 Treatment of sulfonamides 244 with base resulted in the bicyclic alkaloid skeleton 245, via intramolecular nucleophilic attack on the alkynyliodonium salt and subsequent 1,5 CeH insertion of the resulting alkylidene carbene (Scheme 96). 57a The diastereoselectivity of the reaction was rationalized in terms of the transition state 246, 10,57a where the steric interactions between the non-hydrogen group on C-5 and the rest of the molecule are minimized, leading to the major product (Scheme 97).
When the sulfonamides were replaced with alcohols, the resulting fused tetrahydrofurans were formed with higher diastereoselectivity (Scheme 98). 57b Feldman argued that the alkylidene carbene generated in the oxygen system was less electrophilic than that in the nitrogen examples, due to the absence of the electron-withdrawing tosyl substituent, rendering it less reactive towards CeH insertion. This results in a later, more product-like transition state where the steric interactions would be more apparent, resulting in higher diastereoselectivity.
Grainger and Munro reported the first diastereotopic group selective 1,5 CeH insertion reactions of alkylidene carbenes, with the thermolysis of a,b-epoxy-N-aziridinylimine 247 giving a mixture of cyclopentenols 248 and 249 in a 2:1 ratio, with a preference for insertion into CeH a (Scheme 99). 55b Protecting the tertiary alcohol as the silyl ether 250 reversed the observed selectivity, with insertion into CeH b , to give 252, preferred.   It was postulated that the diastereoselectivity arose from interactions within the newly formed vicinal diol (Scheme 100). 55b When the tertiary alcohol is unprotected, hydrogen bonding within the diol 253 would be expected to direct the alkylidene carbene towards reaction at CeH a . Conversely, protecting the alcohol prevents this hydrogen bonding, and so steric repulsions between the silyl ether and the secondary alcohol promote insertion into the CeH b bond.
These results demonstrate the sensitivity alkylidene carbenes have to the environment in which a CeH bond is located. The calculated energy differences between the transition states proposed by Taber, Van Nhien and Postel are by no means excessive, and yet reasonable diastereoselectivities are obtained. Similarly, the work of Feldman and Grainger show have shown how relatively minor substrate modifications can have profound effects on diastereoselectivity in 1,5 CeH insertion reactions.

Alkylidene carbenes in target synthesis
Due to their versatility and ease of generation, coupled with their high levels of selectivity, alkylidene carbenes have seen widespread use in target synthesis, including natural products. 2,7,21e,22,107c,121,126 While the conversion of aldehydes to alkynes has not been discussed in detail in this review, it is nonetheless a powerful tool for the synthetic organic chemist. 6 In particular, the Ohir-aeBestmann reagent 254 (Fig. 9) has been utilized extensively in natural product synthesis. Li employed this reagent in the first total synthesis of (AE)-d-rubromycin, an anti-tumour compound (Scheme 101). 127 Similarly, Reddy utilized this transformation in the synthesis oplopandiol, oploxyne A and (À)-oploxyne B (Scheme 102), all of which show activity towards a range of cancer cell lines. 128 The formation of cyclopentene rings via 1,5 CeH insertion reactions has also proven to be a useful transformation in target synthesis. Several approaches to nitrogen containing natural products utilizing the stereospecific 1,5 CeH insertion adjacent to nitrogen have been reported. 21b,d,46,118,129 Hayes reported the synthesis of cyclopentene 256, an intermediate in the formal syntheses of (À)-TAN1251A 129a,h and (À)-FR901483, 129e by treating vinyl chloride 255 with KHMDS (Scheme 103).
Hayes employed a related approach to the synthesis of another versatile intermediate for natural product synthesis. Reaction between ketone 257 and lithiated (trimethylsilyl)diazomethane gave  cyclopentene 258, which was subsequently converted to several natural products including (1S,3R)-ACPD, (2R)-hydroxymethyl glutamic acid and (þ)-lactacystin (Scheme 104). 46,129b,i Hayes has utilized alkylidene carbene 1,5 CeH insertion reactions in the enantioselective syntheses of several other natural products including omuralide (and its C-7 epimer), 118 and (À)-cephalotaxine. 129f,g An earlier approach to (þ)-lactacystin, utilizing the 1,5 CeH insertion reaction of bromoalkene 209 (Scheme 86), was independently reported by Hayes 118 and Wardrop. 119 The reaction between alkynyliodonium salts and soft nucleophiles has also been utilized in natural product synthesis, 4d particularly the use of sodium sulfinates as nucleophiles. 76c, 113,130 Wardrop employed this strategy in the synthesis of (AE)-magnofargesin, a biologically active lignan (Scheme 105). 76c While the above examples utilize the common generation methods and reactions of alkylidene carbenes in natural product synthesis, a number of alternative transformations have also been employed. Wardrop applied his tetraazafulvene methodology in a total synthesis of combretastatin A4 (Scheme 106). 41 Tykwinski's demonstration that alkynyl groups undergo 1,2migrations (Scheme 27) 47,48,131 has led to significant developments in the synthesis of polyynes, and has been utilized in investigations into the potential structure and properties of the proposed carbon allotrope carbyne. 49 The insertion of alkylidene carbenes into carbonecarbon double bonds has proven to be a useful tool in the synthesis of complex natural product frameworks. Feldman utilized this approach in the synthesis of paraeitropone, with the formation of polycycle 259 and subsequent ring expansion to 260 a key step in this synthesis (Scheme 107). 130a,132 Similarly, H.-Y. Lee's conversion of a,b-epoxy-N-aziridinylimine 261 into triquinane 262 via the [2þ1] cycloaddition, fragmentation and [3þ2] cycloaddition process described previously (Scheme 12), was used in a synthesis of hirsutene (Scheme 108). 30a In addition to the use of new alkylidene carbene reactions and generation techniques, increased knowledge of their selectivity has been exploited to achieve the synthesis of natural products. Feldman's study into the effects of the N-substituent (Scheme 79) was adapted for the total synthesis of (À)-agelastatin A and (À)-agelastatin B (Scheme 109). 113 Scheme 107. Taber has extensively utilized the stereospecific 1,5 CeH insertion reaction of alkylidene carbenes in natural product synthesis, 7,21a,e,107c,121,123,126g,h,134 frequently exploiting the two-step transformation of the resulting cyclopentenes into cyclohexenones through oxidative cleavage followed by intramolecular aldol condensation and dehydration. Representative examples are in his syntheses of oxo-T-cadinol and (À)-morphine (Scheme 111). 121 In both cases, cis [6,5] ring junctions are formed in the alkylidene carbene step, although trans formation is clearly possible, as described above.

Conclusions
The use of alkylidene carbenes in synthesis has a reached a level where reasonable prediction can be made regarding chemoselectivity, regioselectivity and to a certain extent diastereoselectivity. The mild methods of alkylidene carbene generation and stereospecificity of the insertion reaction make their use particularly attractive in target synthesis. The potential for this chemistry to be rendered catalytic through gold chemistry is a notable recent breakthrough and suggests that catalytic asymmetric variants may be possible.