Enantioselective carbenoid insertion into C(sp3)–H bonds

The enantioselective carbenoid insertion into C(sp3)–H bonds is an important tool for the synthesis of complex molecules due to the high control of enantioselectivity in the formation of stereogenic centers. This paper presents a brief review of the early issues, related mechanistic studies and recent applications on this chemistry area.


Introduction
One of the major challenges met in organic synthesis is the formation of carbon-carbon bonds, in particular in a stereoselective way. Nucleophilic substitution reactions, radical reactions, cross-coupling reactions and the Heck reaction are wellknown approaches available to this goal. These reactions are based on the polar characteristic of the carbon-halogen or carbon-pseudohalogen bonds, as a result of the electronegativity difference between these atoms. Despite the proven success of these transformations, they are limited to pre-functionalization of the chemical structure of interest with halogen atoms or pseudohalogen functional groups.
One approach that has been gaining increasing attention, by not requiring the presence of a strongly polarized chemical bond, is the C(sp 3 )-H bonds activation by carbenoids [1]. The enantioselective insertion of these organometallic species into these non-polarized bonds is a recent topic in the chemical literature, when compared to the first reports of carbenoid chemistry around the 1950s.
Carbene is a molecule bearing a functional group with a divalent neutral carbon. This structural framework results in the presence of a nonbonding electron pair that may adopt two electronic configurations: singlet and triplet ( Figure 1). A carbenoid is an organometallic complex where the carbene acts as a neutral ligand to a metal center. This ensures a greater stability of the carbene, allows the modulation of its reactivity, and controls the chemo-, regio-and stereoselectivity in reactions.  The activation of the C(sp 3 )-H bond needs an appropriate interaction between the carbenoid intermediate and the carbon atom of the C(sp 3 )-H. Depending on the electronic demand of the substituent attached to the carbene carbon atom, the insertion reaction can be more or less selective. Very electrophilic carbenoid intermediates, for example, display little regio-and stereoselectivity, favoring the occurrence of side reactions. A less electrophilic carbenoid intermediate, on the other hand, has a lower reactivity, but its regio-and stereoselectivity are better [2].
The electrophilicity of the carbenoid intermediate is related to the substituents present in its structure. Electron-withdrawing groups (EWGs) increase the electrophilicity of the carbon atom in the carbenoid and electron-donating groups (EDGs) act with the opposite effect. Due to these observations, a carbenoid intermediate can be divided in three different classes based on the electronic nature of the chemical groups attached to its structure: acceptor, donor/acceptor and acceptor/acceptor ( Figure 2) [2]. The modification of the substituents on the carbenoid intermediate can change its reactivity and hence the selectivity of the carbenoid reaction.
The most commonly used diazo compounds rely on the formation of a donor/acceptor carbenoid intermediate type. The EWG increases the electrophilicity and reactivity of the donor/ acceptor carbenoid, while an EDG increases its stability and selectivity [2].
Despite the importance of the electronic factors to the reactivity and selectivity of carbenoid intermediates, steric and conforma-tional effects are also determining factors for carbenoid chemistry. Steric as well as electronic factors and the chemical properties of the ligands around the metal center also determine significantly the type of insertion performed by the carbenoid intermediate. The complexes used for the formation of carbenoids in enantioselective insertion reactions must present a balance between steric and electronic factors, to promote the formation of a specific enantiomer.
The search for the best balance of these properties of the carbenoid intermediates was also sought through the use of different metals such as copper [3], rhodium [4], iron [5], ruthenium [6], iridium [7], osmium [8], and others. From these, copper and rhodium have been the most frequently used ones in carbenoid insertion reactions.
This work aims to review the chemical literature, since 2009 [31,32] until the end of 2015, concerning the development of catalytic systems able to promote enantioselective carbenoid insertion into C(sp 3 )-H bonds, the mechanistic aspects recently discovered to the known catalytic systems and the application of these synthetic tools to the organic synthesis of natural products.

Review
Historic perspective on the carbenoid reaction insertion into X-H bonds One of the former works to address the chemistry of carbenoid was reported in 1952 by Peter Yates, although the author does not specifically use the term carbenoid, but -carbene-copper complex (Scheme 1) [3]. In his opinion, the copper catalyst promotes the decomposition of diazoketones to afford "free carbenes", the chemical intermediates responsible for the insertion reaction in X-H bonds (X = O, S, N, or C). The authors demonstrated that an insertion reaction in C(sp 3 )-H bonds only occurs with considerable yield when small amounts of copper powder or copper salts, such as CuSO 4 and CuCN, were employed. Even using the term "carbenoid", the work does not present the formation of a copper carbenoid intermediate. It only suggests an insertion reaction occurring through free carbenes with copper working only as a catalyst to promote the carbene formation.
In 1985, Taber and coworkers reported the synthesis of (+)-αcuparenone (8) through the construction of a five-membered ring prepared by an enantioselective carbenoid insertion into a C(sp 3 )-H bond (Scheme 3) [34]. To carry out the cyclization, the carbenoid was formed by the action of Rh 2 (OAc) 4 on the diazo compound 6. That intermediate intramolecularly inserted into the C(sp 3 )-H bond of the asymmetric carbon to yield ketoester 7 in 67% yield. This latter compound was converted to (+)-α-cuparenone (8) in 26% yield and 96% enantiomeric excess.  In the late 1980s, many studies have been published by Taber [35], Sonawane [36], Doyle [37] and their respective coworkers regarding the regiochemistry of carbenoid insertion into C(sp 3 )-H bonds, and also the steric and electronic factors related to this insertion.
The first example of an enantioselective carbenoid insertion reaction in chemical bonds catalyzed by chiral metal complexes was introduced in 1966 by Noyori and coworkers (Scheme 4) [38]. In addition to the novelty of the use of the chiral copper complex 11 for controlling the enantioselectivity, the authors proposed the participation of the copper carbenoid 13, formed from the reaction between the copper complex 11 and methyl diazoacetate 9 as active intermediate in the catalytic cycle of this transformation.
From the 1990s, the enantioselective carbenoid insertion into C(sp 3 )-H bonds starts to be better discussed in the literature. Ikegami and coworkers reported the enantioselective insertion of α-diazo-β-ketoesters into C(sp 3 )-H bonds catalyzed by rhodium carboxylate complexes in their homochiral form (Table 1) [39]. Modest enantiomeric excesses were provided by the three tested catalysts. The reactions carried out with complex 17a and 17b show very similar stereoselectivity, forming the R-enantiomer of compound 16 as the main product after decarboxylation reaction. The catalyst 17c showed opposite enantioselectivity when compared to the catalysts 17a and 17b, with the S-enantiomer formed as the major product.
The authors could observe the enantioselective formation of the lactones 20 with high enantiomeric excess ( Table 2). The carbenoid formed by (S)-18 favored the S configuration at the generated stereogenic center for most of the prepared lactones. The opposite preference, R configuration at the new stereogenic center of 20, was reported to the use of the enantiomeric rhodium complex (R)-18. When substrate 19f reacts under catalysis of rhodium(II) carboxamide complexes (R)-18 and (S)-18, the configuration of the new stereogenic center of 20f was reversed, probably due to the lack of the oxygen atom in the substituent R, as suggested by the authors.
In 1997, Davies and Hansen reported the intermolecular carbenoid insertion into C(sp 3 )-H catalyzed by rhodium complex (S)-23 with good to excellent enantioselective control (Scheme 5) [41]. The best results were observed when the reaction was carried out at room temperature. A wide range of substituent groups were evaluated at para position of the aryldiazoacetate aromatic ring. The cyclic hydrocarbon reagents, also used as solvent, were cyclopentane, cyclohexane and cycloheptane. Reagent catalyst yield (%) ee (%) configuration of 20 Two factors are noteworthy in this work. Unlike the carboxamide complexes (R)-18 and (S)-18 previously reported by Doyle and coworkers (Table 2), where the complexation of the chiral ligand to rhodium atoms occurs through the carboxamide group, in the new chiral catalyst (S)-23 the rhodium atoms are complexed to the chiral ligands by the carboxylate group, similar to those chiral complexes presented by Ikegami and coworkers (Table 1). Another important feature of this work is, unlike to the work that preceded it, that the new stereogenic center is formed on the carbenoid carbon coordinated to the The authors also reported in this work the insertion into the C(sp 3 )-H bond of tetrahydrofuran. This reaction showed good yield, regio-, diastereo-and enantioselectivity and represents the first example of the formation of a new stereogenic center out of the diazoacetate scaffold by an intermolecular carbenoid insertion into C(sp 3 )-H bond (Scheme 6).

Mechanism of the carbenoid insertion into the C(sp 3 )-H bond
Nakamura [12] and Doyle [37] were the first to do important contributions to the comprehension of the mechanism of this catalytic cycle. In a simplified form, the mechanism of the carbenoid insertion into a C(sp 3 )-H bond can be represented as outlined in Scheme 7.
Nakamura et al. investigated the dirhodium tetracarboxylatemediated carbenoid insertion reaction into C(sp 3 )-H bonds in more detail using the relationship between the transition-state structures and their corresponding free energies obtained by DFT investigation (Scheme 8) [12]. The insertion step primarily consists in the formation of the metal carbenoid 29 by the interaction of the diazo compound 28 and the dirhodium complex 27. In sequence, the reaction proceeds through the transition state TS-30 to release N 2 , and yields the carbenoid 31. The divalent carbon attached to the rhodium atom starts to interact with the hydrogen of the C(sp 3 )-H bond of the compound 32 to form the van der Waals complex 33 which undergoes through

Copper-based chiral catalysts
In 2002, Müller and Boléa published a study evaluating the enantioselective insertion of copper carbenoids formed from phenyliodonium ylides and diazo compounds (Table 3) [9]. This work is particularly important because, at that time, the carbenoids derived from rhodium complexes were the most used for insertion reactions in C(sp 3 )-H bonds.  Comparing the results of Table 3, the same enantiomer was obtained mainly for both carbenoid precursors, ylide 42a and the diazo compound 42b. The authors suggested the formation of the same chiral copper carbenoid intermediate by the reaction of the in situ prepared chiral copper complexes with both 42a and 42b to provide the observed insertion products.
The reactions with phenyliodonium ylides 42a showed better ee when compared to that done with diazo compounds 42b. The authors attributed this observation to the large difference between the reaction temperatures, 0 °C to ylides versus 65 °C to diazo compounds. Higher temperatures increase the carbenoid formation rate by the chiral copper complexes as well as the carbene formation rate by direct decomposition of the precursors 42a and 42b. The competition between carbene insertion and chiral carbenoid insertion into C(sp 3 )-H bonds decreases the enantioselectivity of this transformation.
To confirm the copper carbenoid formation and its participation on the insertion reactions, the authors prepared the carbenoid precursors (R)-48a and (R)-48b and submit them to Rh 2 (OAc) 4 or Cu(hfa) 2 catalysis (Table 4). All reactions maintained the configuration of the asymmetric carbon where de insertion happened, independent to the carbenoid precursor and the catalyst, a strong evidence of the carbenoid intermediates formation.
The authors also performed the copper carbenoid insertion reaction to yield five-membered cyclic sulfones 54, under similar experimental conditions, in moderate yields and enantiomeric excesses of the trans stereoisomer (Table 6).
Independent to the size of the product, the authors emphasize the low dependence of the enantioselectivity with respect to the structural nature of the substrates where the lowest results are observed for the substrate 51h (R = CH 3 ). In contrast, the reaction times showed to be more dependent on the chemical structure of the substrates.
This research group has focused a lot of efforts to better comprehend the scope of this catalytic system, especially on features concerning the BARF salt effect [43,44] and electronic effects on the aromatic rings of the chiral ligands [45].
Attempts to heterogeneous catalysis using chiral copper complexes were also done. Fraile et al reported, in 2011, the copper catalyst 60 for enantioselective insertion of carbenoid into O-heterocycles C(sp 3 )-H bonds (Table 7) [47,48]. The reaction was performed under homogeneous and heterogeneous conditions, with laponite as support for the catalyst. The reaction afforded moderate yields, diastereomeric ratio and enantioselec-tivity under both conditions. The supported catalytic system was reused over three cycles with no performance decrease. The same heterogeneous catalyst was also used to perform an enantioselective insertion of the carbenoid into benzylic C(sp 3 )-H bonds and similar results were observed [49].
In 2011, the same research group developed a new heterogeneous copper catalyst for carbenoid insertion into C(sp 3 )-H bonds [50]. The solid support was based on SiO 2 /Al 2 O 3 and, after addition of ligand 1c to the reaction media, the reaction afforded moderate yields, diastereomeric ratio and enantioselectivity. These catalysts were reused over three cycles with progressive yield and enantioselectivity decrease.

Iridium-based chiral catalysts
Most recently, chiral iridium complexes have been used as catalyst for insertion reactions in C(sp 3 )-H bonds. In 2009, Suematsu and Katsuki published the first study addressed to the  use of iridium-based chiral complexes as catalyst for the formation of carbenoid intermediates ( Figure 5) [51]. The authors conducted insertion reactions in C(sp 3 )-H bonds in a diastereoand enantioselective manner. For enantioselective insertion reactions the authors tested two specific iridium complexes, 61a and 61b.
The authors used a wide range of α-substituted α-diazoacetates for performing insertion into substrates such as tetrahydrofuran (24) and 1,4-cyclohexadiene (62). Using these two compounds, the authors evaluated the best conditions for carrying out the reaction. When the iridium-catalyzed insertion reaction of carbenoids into tetrahydrofurans C(sp 3 )-H bond was performed at room temperature, the authors reported the formation of dimers of α-substituted α-diazoacetates as the main products of this reaction. This issue was circumvented when low temper- atures, −50 °C, were used and the insertion reaction occurred with considerable yields and good enantiomeric excess (Table 8). According to the authors, the low temperature could reduce some type of steric strain on the transition state of the insertion reaction and avoid the dimer formation.
For the insertion reaction of iridium carbenoid into the 1,4cyclohexadiene (62) bis-allylic C(sp 3 )-H bond the authors observed the formation of two products, one resulting from iridium carbenoid insertion into the C(sp 3 )-H bond (63) and the other as a result of the cyclopropanation reaction (64, Table 9).
Both examples reported by Suematsu and Katsuki showed very good yields and excellent enantiomeric excesses of the products. This work is noteworthy because it is the first report in the literature of an enantioselective insertion of an iridium carbenoid into C(sp 3 )-H bonds.

Che and coworkers introduced the first porphyrin-based chiral iridium catalyst (−)-65 to insertion of carbenoids into C(sp 3 )-H
bonds [52]. The reaction with 1,4-cyclohexadiene was promoted by 1 mol % of the catalyst at low temperatures to affords the product in high yields and enantioselectivity (Scheme 12).
The same catalytic system was applied to carbenoid insertions into tetrahydrofuran C(sp 3 )-H bonds (Scheme 13). The reaction afforded the desired product in a regioselective way and high diastereoselectivity, ranging from 2.5:1 to >20:1, favoring the anti-product, in a complementary sense when compared to the results reported by Suematsu and Katsuki for iridium catalyst 61a (Table 8). Poor to excellent yields and high enantioselectivity were reported for the main product.
The chiral porphyrin-based iridium complex (−)-65 was also used by the same research group to catalyze the intramolecular carbenoid insertion into C(sp 3 )-H bonds and affords the synthe- In 2013, Davies, Blakey and coworkers reported a new iridium catalyst to perform a carbenoid insertion into the C(sp 3 )-H bond (Scheme 15) [54]. The reactions were performed at room temperature and low catalyst loading (0.5 mol %) to afford the desired product in high yield and enantioselectivity. To reduce the amount of cyclohexadiene, the reaction was also performed in trifluorotoluene, resulting in a yield decrease (93% when 1,4cyclohexadiene was solvent and reagent; 60% when 2,5 equivalents of cyclohexadiene and PhCF 3 as solvent where used) but with almost the same enantioselectivity.

Rhodium-based chiral catalysts
Since the pioneering reports by Ikegami [39], Doyle [40] and Davies [41] introducing their dirhodium chiral catalysts, these complexes have been the most frequently used and studied catalysts for enantioselective insertion of carbenoids into C(sp 3 )-H [55]. It is noteworthy the intensive contribution of the Davies research group which reported important works related to regioselectivity and stereoselectivity [56], and chemoselectivity [57] of this transformation.
In 2011, Davies et al reported a chiral rhodium complex based on a new cyclopropylcarboxylate ligand (Scheme 16) [58]. Among the various transformations promoted by this new catalyst we can find enantioselective carbenoid insertion into the endocyclic allylic C(sp 3 )-H bond of 71 followed by the Cope rearrangement and retro-Cope strategy previously described by the same research group [59]. The product was obtained in excellent yield, diastereo-and enantioselectivity.
Later, the same authors showed a new chiral rhodium complex (R)-74 based on an analogue cyclopropylcarboxylate ligand (Scheme 17) [60]. This new ligand favors the regiochemistry of rhodium carbenoid insertion into primary C(sp 3 )-H activated bonds even in the presence of activated secondary C(sp 3 )-H bonds. This preference stems from the greater volume of the ligand and the consequent greater steric strain in the transition state that leads to the minor insertion product at activated secondary C(sp 3 )-H bonds. Substrates with benzylic bond, allylic and α-oxygen C(sp 3 )-H were submitted to the new catalyst, under dichloromethane reflux, and led to the preferential formation of the insertion products into primary carbon, (from 5:1 to >20:1), high yields and enantioselectivity, higher than 90% ee, 88% ee and 64% ee, respectively. TCE heteroaryldiazoacetates were also successfully employed for the formation of the rhodium carbenoid insertion products in superior yields when compared to the reaction with methyl aryldiazoacetates. TCE aryldiazoacetates reduced significantly the carbene dimerization allowing the reduction of the TCE diazoacetate addition time from 1.5 hours to 5 seconds.
The deactivated aryl methyl ether 84 was also functionalized by the use of TCE aryldiazoacetates in combination with (R)-47 in Scheme 20: Sequential C-H functionalization approach to 2,3-dihydrobenzofurans.
good yield and excellent enantioselectivity. The reaction between 84 and the methyl aryldiazoacetate 21e afforded the insertion product in only 15% yield.
In 2013, Davies, Yo et al reported a new strategy to construct 2,3-dihydrobenzofurans based on a sequential enantioselective rhodium catalyzed carbenoid insertion into a C(sp 3 )-H bond followed by a palladium C(sp 2 )-H bond activation to build a new C-O bond (Scheme 20) [62]. A wide range of benzyl silyl ethers and diazo compounds were tested providing the desired 2,3-dihydrobenzofuran in good yields and excellent diastereoand enantioselectivity. Later, this strategy was further used by Davies, Zakarian and coworkers to access the total synthesis of (−)-maoecrystal V [63].
During this study, the authors observed an unexpected result when ortho-halosubstituted diazo compounds were used. Here the formation of a β-lactone by the carbenoid insertion into the C(sp 3 )-H bond of the alkyl substituent of the alkoxy moiety of the ester (Scheme 21). The authors decided to investigate this observation and reported a more detailed study concerning the synthesis of cis-disubstituted β-lactones in high yield, diastereoand enantioselectivity [64].
Total syntheses of 2,3-dihydrobenzofurans containing natural products have also been recently reported independently by Hashimoto [65,66] and Kan [67] based on an enantioselective intramolecular rhodium carbenoid insertion into C(sp 3 )-H bonds.
In 2012, Pavlyuk and coworkers performed the synthesis of azacycloalkenes by rhodium carbenoid insertion into C(sp 3 )-H bonds, and subsequent ring closing olefin metathesis (RCM) [68]. The insertion of the rhodium carbenoids derived from  vinyl diazoacetate into the C(sp 3 )-H bonds of the alkenylcarbamates 97a-d yields two reaction products (Table 10). The major one (99a-d) was the result of the cyclopropanation reaction of the double bond present in 97a-d. The minor product (100a-d) was the desired one, resulting from the insertion reaction on the C(sp 3 )-H bond α to the nitrogen atom.
In 2015, Hashimoto et al reported the synthesis of methyl 2-vinyltetrahydropyran-3-carboxylates (104) by an enantioselective rhodium carbenoid insertion into C(sp 3 )-H bond strategy [69]. The desired product was obtained in very good yield and excellent diastereo-and enantioselectivity favoring the cis isomer (Scheme 22). An interesting work was reported by Che and coworkers concerning the first rhodium porphyrin-based catalyst for enantioselective carbenoid insertion into C(sp 3 )-H bonds [70]. The reaction with acyclic alkanes showed regioselectivity in favor of the formation of the insertion product into primary carbons in modest stereoselectivity (Scheme 23). The preference for the reaction in less hindered carbon was attributed to high steric demand required by the chiral ligand in the transition state of the carbenoid insertion step in the C(sp 3 )-H bond.
Cyclic alkanes were also tested with yields ranging from 64-80% and enantioselectivities between 88 and 92% ee. The reaction with cyclohexane was conducted on a gram scale and, after 10 hours of reaction, 2.88 g (73% yield) were obtained of the carbenoid insertion product in 91% ee. Cyclohexene showed high regioselectivity for the carbenoid insertion of the allylic C(sp 3 )-H bond, 43% yield, 71% ee and a 60:40 diastereoisomeric ratio. The formation of the cyclopropanation product was also observed in 14% yield. Ethylbenzene (110) was used and also showed high regioselectivity favoring the carbenoid insertion into benzylic C(sp 3 )-H bonds (Scheme 24). The diastereoisomers 111 and 112 were obtained in 45% yield and 16%, respectively, and moderate stereoselectivity was observed in both products.
Few examples of attempts to develop heterogeneous catalytic systems based on chiral rhodium complexes were also done. In 2010, Hashimoto and coworkers reported the synthesis of a highly robust polymer-supported chiral dirhodium(II) complex [71]. The chiral part of the catalyst was based on the N-phthaloyl-(S)-amino rhodium carboxylate (S)-17d. Two substrates were submitted to the enantioselective carbenoid insertion into the C(sp 3 )-H bond in toluene as solvent and at temperatures ranging from −78 °C to −60 °C. The desired products were obtained in yields up to 80% and enantioselectivity up to 90%, even after 15 recycles of the catalyst.
Jones, Davies and coworkers also recently published a new heterogeneous catalytic cycle base on homogeneous catalyst (S)-23 [72]. The chiral scaffold was covalently supported on silica surface through an organic linker and was tested in a tandem enantioselective carbenoid insertion into C(sp 3 )-H bond/Cope rearrangement. The reactions afforded good yields and excellent enantioselectivity. The recycling of the catalyst was evaluated in a cyclopropanation reaction and no significant decrease on its performance could be observed after five runs.

Conclusion
The efforts focused on the development of the enantioselective insertion of carbenoids into C(sp 3 )-H bonds have provided a wide range of catalytic systems to the chemical community able to perform this transformation and to introduce new C-C bonds in a enantiocontrolled way. The chiral rhodium catalysts are the state of art of this synthetic tool. However, rhodium is an expensive metal and increases the cost of the chemical process despite the low catalyst loads found in literature.
Despite the good results presented until today, the use of the reported chiral iridium catalysts is even more expensive than the use of other metals such as rhodium, copper and ruthenium, for example. Efforts should be directed toward the development of simpler ligands specially those based on inexpensive chiral building blocks like amino acids and sugars.
The examples of works focused on copper-based catalysts are growing in number and quality. Among the well-known metals able to react with diazo compounds to afford carbenoid intermediates, copper is inexpensive and has a wide range of wellestablished chiral ligands able to be tested and to inspire the rational design of new ligands.
For all metals commonly used in this transformation, more efforts should be focused towards the development of new and robust heterogeneous catalytic systems. This strategy can reduce the costs related to metals like rhodium or iridium and can also contribute to reduce the disposal of these metals in the environment.