Earth-Abundant 3d Transition Metal Catalysts for Hydroalkoxylation and Hydroamination of Unactivated Alkenes

Abstract

This review summarizes the most noteworthy achievements in the field of C–O and C–N bond formation by hydroalkoxylation and hydroamination reactions on unactivated alkenes (including 1,2- and 1,3-dienes) promoted by earth-abundant 3d transition metal catalysts based on manganese, iron, cobalt, nickel, copper and zinc. The relevant literature from 2012 until early 2021 has been covered.

1. Introduction

The prevalence of ether and amine motifs in various areas of chemistry, such as bulk and fine chemicals, biological and pharmaceutical chemistry or catalysis, has driven the development of original and innovative methods for their efficient synthesis [1,2]. Among these methods, the most attractive and atom-economical approach towards these prevalent motifs is undoubtedly the addition of E–H (E = O, N) σ-bond onto unactivated alkenes, referred to as alkene hydrofunctionalization [3,4,5,6,7,8,9]. This route is also step-economical and benefits from relatively available and inexpensive starting materials. However, the inherent electronic repulsion between the olefin π-electrons and the nonbonding lone pair of the heteroatom renders this approach kinetically challenging in the absence of a catalyst. Control of the regioselectivity and enantioselectivity of the process to access (enantioenriched) key building blocks are also essential issues that need to be tackled for a broad synthetic use of the methodology. In light of these challenges, a variety of elegant activation strategies has been developed over the years to address some of these issues and extend the methodology potential to access various classes of O- and N-containing compounds. For several decades, the field has been dominated with successes by Group 1–4 elements and noble metals-based catalysts that operate through a two-electron reactivity by either amine activation (via deprotonation) or alkene activation (via metal coordination) [10,11,12,13,14,15,16]. The current booming trend for the use of earth-abundant, first-row late transition metal, metal-free and photoredox processes in catalysis has reshaped the field of alkene hydrofunctionalization giving birth to efficient strategies that not only rely on classical activation modes but also on novel radical-type reactivities [9,10,17,18,19]. For examples, photo-mediated E–H (E = O, N) bond addition on electron-rich alkenes wherein single-electron oxidation of one of the nucleophilic partners furnishes an electrophilic oxygen- [20], nitrogen- [18,21,22,23,24,25,26,27] or carbon- [28,29,30,31,32] centered radical cation that can be intercepted by the second nucleophilic partner has become a valuable tool for the construction of C–E (E = O, N) bond by an anti-Markovnikov hydrofunctionalization process. Designed chiral Brønsted acids with specific confined environments have successfully enabled the contemporary challenge of asymmetric cyclohydroalkoxylation of aliphatic alcohols tethered to 1,1-disubstituted alkenes [33,34].

This review will survey the most relevant achievements in the field of intra- and intermolecular hydroalkoxylation and hydroamination of unactivated alkenes (including 1,2- and 1,3-dienes) promoted by earth-abundant 3d transition metal catalysts from manganese to zinc. It is beyond the scope of this review to cover all the growing activity of alkene hydrofunctionalization and, in this context, the literature coverage has been restricted from 2012 until early 2021 albeit a few related reports that appeared earlier are mentioned. For earlier developments, the readers are invited to refer to previous reviews on the topic [6,7,8]. Relevant breakthroughs in the direct addition of electronically biased amines (such sulfonamides, amides, etc.) onto unactivated olefins will also be covered but the term hydroamidation will be preferred to refer to these compounds bearing electron-withdrawing substituents. The O–H bond formation reactions by hydration [35,36] or hydroacyloxylation [37,38,39] of unactivated alkenes are not included in this survey. The achievements are summarized according to the nature of the 3d transition metal and the type of hydrofunctionalization reaction.

2. Manganese

Despite the interesting features of being of low toxicity and eco-friendly, manganese remains unexploited for the development of catalytic hydroamination and hydroalkoxylation methodologies of unactivated alkenes [40]. To our knowledge, there is only one report of manganese-catalyzed hydroamination of unactivated alkenes [41,42]. This year, the group of Qu and Zhang reported an elegant visible-light-induced manganese-promoted N–F bond activation strategy to generate amidyl radicals from N-fluorinated sulfonamides (Figure 1). These electrophilic amidyl radicals can react with olefins in the presence of hydrosilane to afford a rich variety of N-sulfonamides pyrrolidines and functionalized linear N-sulfonamides by exo-cyclization and anti-Markovnikov addition respectively. The intermolecular process is restricted to styrene derivatives, 2-vinylthiophene and electron-rich alkenes, such as vinyl sulfides and vinyl ethers. Indeed, due to electronic matching problems, low reactivity is observed for alkyl-substituted alkenes. This protocol can also be applied to a two- and three-component carboamination of alkenes [41]. The authors proposed that a light-induced homolysis of the Mn–Mn bond in Mn₂(CO)₁₀ generates a metalloradical Mn(CO)₅ having the ability to abstract the F atom from the substrate and leading to the formation of the reactive amidyl radicals [43]. Subsequent regioselective addition of these N-centered radicals to the alkenes leads after silane-mediated HAT to the formation of the product. From mechanistic investigations, it is proposed that two competitive radical chain reactions, one of which including metallic species, complete the catalytic cycle.

3. Iron

3.1. Hydroalkoxylation

In the 2010s, intra- and intermolecular hydroalkoxylation of unactivated alkenes emerged using Lewis acid iron salts as catalysts [44]. Indeed, inexpensive and poorly toxic iron (III) chloride mainly associated with various additives, such as silver triflate or p-toluenesulfonic acid, afford effective Lewis acid catalysts for various hydroalkoxylation reactions involving alkene derivatives of various substitution patterns [45,46]. Nevertheless, in 2007, the Takaki group demonstrated that cationic iron complexes generated by the combination of FeCl₃ and AgOTf enable the efficient cyclohydroalkoxylation of various alkenes under mild conditions [47]. Several years later, Zhou et al. reported a Markovnikov selectivity intermolecular version involving aliphatic alcohols with styrene derivatives and 1-octene using the FeCl₃/TsOH combination [48]. Iron (III) chloride can also be supported on montmorillonite as natural inorganic material to provide a very effective heterogeneous and reusable catalyst for this process. In 2015, the supported Fe-MMT catalyst developed by Antoniotti et al. showed good activity and selectivity at 80 °C in dimethyl carbonate (DMC) for the cyclization of alkenols with various substitution patterns on the alkene moiety (Figure 2) [49,50]. This heterogeneous catalyst, which can be reused several times without loss of activity, displays higher activity and selectivity than those of reactions carried out under homogeneous catalysis. The system was demonstrated to proceed under truly heterogeneous catalysis and can be transferred to an intermolecular process despite lower efficiency.

In 2012, Kang et al. developed an efficient catalytic system for the exo-cyclohydroalkoxylation of trialkylated δ-allenyl alcohols observing isomerization or no isomerization of the product alkene moiety depending on the nature of the iron counterion and the operating conditions as illustrated in Figure 3 (top) [51]. This methodology is restricted to trialkylated allenols as mono- and dialkylated did not undergo cyclization under these conditions. The Fe(OTs)₃ system can also be applied to the cyclization of γ-allenyl alcohols with concomitant alkene isomerization (Figure 3, top) [52]. During the reactivity study of γ-cyclohexyl-substituted allenic alcohols by this catalytic system, it was shown that the temperature influences not only the isomerization of the hydroalkoxylation products but also the endo/exo-methylene selectivity.

More recently, Rueping and co-workers have developed and used as a precatalyst the air and moisture stable tetraphenyl-substituted cyclopentadienone iron(0) tricarbonyl complex, easily prepared from tetraphenylcyclopentadienone and iron pentacarbonyl (Figure 4) [53]. This complex was efficient for the exo-cyclohydrofunctionalization of α-allenic tethered to secondary alcohols or N-Cbz protected amines providing the corresponding 2,3-dihydrofurans or 2,3-dihydropyrroles in good yields and high chemoselectivity. The authors presumed that the first step of the mechanism is the formation of the competent 16-electron iron species by trimethylamine N-oxide promoted decarbonylation of tricarbonyl precatalyst 1-Fe. Subsequent dual activation of the substrate by the metal-ligand bifunctional iron catalyst, through hydrogen-bonding between the C=O of the ligand and the O- or N-nucleophile and metal coordination of the allene, allows the C-heteroatom bond formation to occur. This step leads to the formation of an iron vinylidene intermediate, which isomerizes into a more stable complex by two consecutive hydrogen transfers implying the cyclopentadienone ligand. Protodemetalation of the more stable vinylidene intermediate leads to the liberation of the desired product. The reaction chemoselectivity in favor of 2,3-dihydrofuran is controlled by the formation of the more stable vinylidene iron complex.

In 2019, Marcyk and Cook described a very efficient method for intermolecular C–O and C–N bond formation by Markovnikov alkene hydrofunctionalization using a cationic iron(III) system (Figure 5) [54]. From various primary and secondary alcohols and mono-, di- and trisubstituted alkenes, they produced different ethers in moderate-to-good yields under very mild conditions using the air- and moisture-tolerant combination FeCl₃/AgSbF₆. In order to avoid deleterious substitution reactions leading to side products, primary alcohols were preferred. Furthermore, it is necessary to employ an excess of alkenes (3 equiv) to override the lower reactivity of alcohols with respect to sulfonamides. The study of the scope of the alcohol partner on cyclohexene indicates that the best results were obtained with alcohols substituted by an electronic deficient aromatic group such as 2-(4-nitrophenyl)-ethanol. This latter was then used with various cyclic and acyclic alkenes (Figure 5). It is worth noting that unsymmetric monoalkene 1-hexene affords a mixture of 2- and 3-substituted hydro-functionalized products likely due to a carbocation rearrangement. Additionally, trisubstituted alkenes gave superior yields with iron trichloride alone due to reduced competitive olefin dimerization.

3.2. Hydroamination

Until 2014, only iron (III) salts were reported in the literature as effective Lewis acid catalysts for intra or intermolecular C–N bond formation by hydroamidation of unactivated alkenes mainly from N-tosylamines, but regioselectivity issues were observed in some cases [55,56,57,58]. In 2012, the group of Kang extended the Fe(OTs)₃-based protocol developed for the cyclization of δ-allenyl alcohols to the intramolecular hydroamidation of δ-allenyl amines (Figure 3, bottom). In contrast to the hydroalkoxylation reaction, which was restricted to trialkylated substituted allene moiety, this extension was efficient for the cyclization of N-tosyl amines tethered to mono-, di- and trisubstituted alkenes [51]. However, the corresponding N-tosylpiperidines from trialkylated allenyl amines were obtained as an inseparable mixture of products resulting from double bond isomerization from exo-methylene to endo-methylene. A similar transformation was later applied by Wang in cascade reactions implying arylamines and propargylic alcohols catalyzed by FeCl₃·6H₂O [59]. In 2016, it was described a highly effective diastereoselective FeCl₃-catalyzed hydroamidation of α-substituted aminoalkenes derived from enantiopure starting materials, such as amino acids, and this methodology was applied to the enantiodivergent synthesis of (+)- and (−)-pyrrolidine 197B [60].

Recently, as for hydroalkoxylation, new developments have been described using iron for the hydroamidation of unactivated alkenes. In 2019, as mentioned above, Marcyk and Cook described the use of a combination of FeCl₃ and AgSbF₆ to functionalize mono-, di-, and trisubstituted olefins with a wide range of sulfonamide nucleophiles proceeding in very mild conditions with Markovnikov selectivity (Figure 5) [54]. No hydroamination product was observed with other electron-deficient protecting groups on the amine due to stronger binding to the cationic metal. This catalytic system was applied to a cascade process for the synthesis of tetrahydroisoquinolines [61]. The tetraphenyl-substituted cyclopentadienone iron(0) tricarbonyl complex was also efficiently used as a precatalyst for the endo-cyclohydroamidation of N-Cbz protected amines tethered to monoallenes (Figure 4, bottom) [53]. The unusual selectivity leading to cyclic enamines was explained by dual activation of the substrate by the iron-based metal−ligand catalyst.

Recently, C–N bond formation by hydroamination of unactivated alkenes with unbiased amines has also been described with iron(III) salts as Lewis acid catalysts. An interesting anti-Markovnikov hydroamination of vinylpyridines with azoles including diazoles and triazoles, catalyzed by FeCl₃ was recently reported (Figure 6, top) [62]. The reaction proceeded at 110 °C and gave moderate-to-excellent yields. The authors explained the unusual anti-Markovnikov selectivity by FeCl₃ activation of the N-atom of the vinylpyridine to form a zwitterionic intermediate, which is in equilibrium with one containing a positive charge on the terminal carbon of the vinyl group and subsequent conjugate addition of the azole.

Last year, Chen, Jiang and co-workers published a general catalytic system, composed of Fe(OTf)₃ and γ-cyclodextrin, for the Markovnikov selective hydroamination of styrenes (and norbornene) with carbazoles and carbazole-fused aromatics with high activity and selectivity under mild conditions (Figure 6, bottom) [63]. The scope of the alkenes and the carbazoles has been studied. The authors have shown that monosubstituted styrenes bearing electron-donating or electron-withdrawing groups, but also a bulky functional are well tolerated. Limitations appear with the use of larger alkenes than monosubstituted styrene such as 2-vinylnaphtalene. Concerning the use of aliphatic alkenes, norbornene also furnishes the hydroamination product under the standard conditions contrary to cyclohexene which remains inert. No limit in size of the carbazoles was observed, as various carbazole-fused aromatics provide the hydroamination products moderate-to-good yields. The authors proposed that bulky Fe(OTf)₃ coordinates at the narrower rim of cyclodextrins through hydrogen bonds with the primary hydroxyl groups to form in situ a highly active Fe(OTf)₃/cyclodextrine complex. The complex thus formed would then deprotonate the carbazole to generate the active iron species that would react with the styrene derivatives captured in the cyclodextrin cavity.

The first example of regioselective hydroamination of electronically unbiased amines catalyzed by iron complexes was described by the Hannedouche group in 2014 (Figure 7) [64]. In this work, the use of structurally defined C₂-symmetric low-coordinate β-diketiminatoiron(II)-alkyl complex 2a-Fe in the presence of cyclopentylamine as co-catalyst allows the formation of pyrrolidines and piperidines in good-to-excellent yields at 90 °C. This catalytic system requires aminoalkenes with a geminal disubstitution on the tether and does not proceed with 1,2-dialkylsubstituted alkenes. Mechanistic studies have shown that the activation pathway proceeds by initial deprotonation of the amine and involves a migratory 1,2-insertion of the alkene in an iron-amido bond of an isolable amido iron complex followed by a rate determining aminolysis step. It should be noted that the presence of cyclopentylamine, a non cyclizable primary amine as co-catalyst diminishes the formation of β-H elimination products but also reduces the reaction rate. To overcome this limitation and relying on experimental and computational mechanistic studies, the group subsequently developed the synthesis of a variety of well-defined C₁-symmetric β-diketiminatoiron(II)-alkyl complexes that differ by the nature of one aryl group on the N-atom and studied their reactivities [65]. It was shown that steric effects and/or the coordinating ability of an ortho-methoxy substituent on one N-aryl rings of the β-diketiminate affords a more active and selective catalyst 2b-Fe than the catalytic system C₂-symmetric 2a-Fe/cyclopentylamine (Figure 7) [66].

In 2014, an umpolung electrophilic amination was reported by Yang and co-workers as an alternative strategy to the classical hydroamination, similarly to the copper-hydride strategy reported by Buchwald, Hirano and Miura (vide infra) (Figure 8) [67]. After an optimization study, a combination of FeCl₂ and 2,6-diiminopyridine as precatalyst in the presence of four equivalents of cyclopentylmagnesium bromide as reducing agent showed its effectiveness for the regioselective Markovnikov hydroamination of a wide range of aromatic olefins and hydroxylamine esters as the electrophilic nitrogen source. The authors proposed that the catalytic cycle is based on in situ formation of a Fe–H species but unlike to the CuH methodology (vide infra), it is inefficient for the transformation of α- and β-methylstyrenes, p-Cl- and p-CF₃-styrenes and aliphatic terminal alkenes.

In 2015, the group of Baran described a practical formal hydroamination process based on iron hydride chemistry for the preparation of secondary (hetero) arylamines substituted by secondary or tertiary alkyl groups from nitro (hetero)arenes and mono- and polysubstituted alkenes (Figure 9, top) [68]. This system stems from their expertise on the H-atom transfer (HAT) promoted by Fe(III)–H for C–C bond radical formations from alkenes [69]. Using iron(III) acetylacetonate (30 mol%) and phenylsilane (2–3 equiv), compatible with air and moisture, a wide variety of hindered (hetero) aryl secondary amines were prepared in low-to-good yields and with full Markovnikov selectivity. In this process, [Feⁿ]–H species (generated in situ from iron (III) acetylacetonate and phenylsilane) [70] allows the formal single electron reduction of the nitro (hetero) arene and olefin to a nitroso (hetero) arene and an alkyl radical by regioselective HAT respectively. The formal hydroamination product is then obtained by attacking the alkyl radical on the nitroso (hetero) arene to lead to an O-centered radical species, which are subsequently involved in ethanolysis and SET [71]. A considerable advantage of this system over other C–N bond formation methods is its compatibility with a remarkable number of functional groups. Indeed, this methodology has a wide tolerance for amide, amine, alcohol, thioether, ketone, halide, nitrile, triflate and boronic acid. However, this protocol requires the use of an excess of alkene and is limited to nitro (hetero) arenes (excluding 2-nitropyridines, nitroimidazole and nitrophenyl having orthoesters or thiols or free alcohols) and aliphatic alkenes as substituted styrenes give low yields. A year later, the groups of Shenvi [72] and Thomas and Shaver [73,74] independently optimized the catalytic system developed by Baran by using either isopropoxy (phenyl) silane instead of phenylsilane, or an amine—iron(III) bis (phenolate) complex instead of iron(III) acetylacetonate. The catalyst loading could be reduced, and the reaction carried out at room temperature under these new conditions.

Last year, the Wang’s group extended the scope of these reductive C–N bond formations using a new system composed from a single half-sandwich iron(II) complex and (EtO)₃SiH for the synthesis of various secondary branched amines and indole derivatives by respectively inter or intramolecular reductive coupling (Figure 9, bottom) [75]. The authors showed that the complex 3a-Fe reacts with nitro (hetero)arenes to afford iron-nitrosoarene complex 3b-Fe, which can be isolated and allows the formation of arylamines from nitroarenes and alkenes in the presence of (EtO)₃SiH. In contrast to the Baran system which is low-yielding for arylarenes and required a reductive workup (Zn/HCl_aq), this catalytic system affords secondary amines in moderate-to-high yields from both aliphatic and aryl alkenes and in the presence of hydrosilane as the only reducing agent. For the intermolecular C–N bond coupling with styrene derivatives, electron-donating groups at the para position of the phenyl ring give better yields than electron-withdrawing, such as the cyano group, although the latter is compatible with the reaction. Nevertheless, aliphatic alkenes are generally less reactive than aromatic alkenes. As shown in Figure 9, a low catalyst loading (0.5 mol%) is enough to provide the indole products in very good yields. Fused tricyclic indole derivatives, 2,3-disubstituted indoles, 3-phenylindole but also indazole can be formed in good yields and functional group like OMe, F, Br and CF₃ are tolerated by the catalytic system.

4. Cobalt

4.1. Hydroalkoxylation

Cobalt catalysis has been widely studied for the development of hydrofunctionalization reactions of alkenes through the generation of alkyl radicals by metal-hydride hydrogen atom transfer (MHAT), which can subsequently be intercepted by heteroatom-based radicalophiles to create a C-heteroatom bond [76]. Mukaiyama and Isayama originally described the use of Co(acac)₂, with silane and molecular oxygen, for the hydration of olefin [77]. This metal was further exploited by Carreira for the introduction of various other functional groups (hydrazine, azide, cyanine, chlorine, oxime) using Co catalysts (such as 4a-Co, Figure 10) [78,79,80,81].

In 2013, with the aim of developing further hydrofunctionalization reaction (i.e., fluorination), Shigehisa, Hiroya and co-workers serendipitously found out that cobalt-Schiff base catalyst 4a-Co in the presence of N-fluorotrimethylpyridinium salts (Me₃NFPYBF₄ or Me₃NFPYOTf) and a hydrosilane reagent can promote the hydroalkoxylation of unactivated alkenes with alcohol solvents, such as MeOH or t-BuOH (Figure 10, top) or fluorinated alcohols [82,83]. The authors demonstrated the versatility of the reaction, able to tolerate a wide range of functionalities (silyl, acetal, esters, amides, nitro, tosylates, heterocycles, etc.) and applicable to mono-, di- and tri-substituted olefins. The postulated mechanism involves the formation of a cobalt(III)-fluoride complex followed by its transformation into Co(III) –H with the silane derivative. Then, the reaction of Co–H and alkene via an MHAT process affords the chemoselective generation of C-centered radical, which is further converted into a cation intermediate by a radical-polar crossover reaction facilitated by the N-fluorocollidinium salt and/or the cobalt-Schiff base catalyst. Trapping of the reactive cation by the nucleophilic alcohol gives the expected product (after deprotonation by 2,4,6-trimethylpyridine). A similar system was also used for the redox-switchable hydroalkoxylation of styrenes by incorporating a ferrocene ligand on the Co catalyst, able to modulate its reactivity in an on/off manner [84].

In 2016, Shigehisa and co-workers further exploited their system and reported intramolecular hydroalkoxylation of alkenols (Figure 10, bottom) [85]. Under similar conditions using catalyst 4a-Co, 5- and 6-membered rings were efficiently formed (70–97% yields) using aliphatic and benzylic alcohols (as well as carboxylic acids). However, phenol cyclization was found to be unsuitable for the reaction (<13% yield). For the synthesis of medium-ring products, such as 7–9-membered cyclic ethers and lactones, catalyst 4b-Co gave better results. As an example, an oxepane derivative product was obtained in 60% yield (vs. 30% using 4a-Co). Interestingly, when using protected alcohols bearing TBS (tert-butyldimethylsilyl), MOM (methoxy acetal) or Bn (benzyl) groups, deprotective cylizations were carried out directly affording the cyclic ethers. This was of particular interest for carrying efficiently phenolic cyclizations (2 examples: 70–90% yield using MOM groups) circumventing the reactivity issue observed with protecting-group free phenol substrates. More recently, this deprotective cyclization strategy was successfully exploited for the preparation of cyclic carbamates and isourea [86]. The mechanistic investigations revealed that cyclization should occur by simultaneous radical oxidation and nucleophilic trapping in a concerted transition state, leading to the formation of an oxonium intermediate followed by deprotection. In addition, the authors noted that incorporating chiral elements within the catalyst can lead to the observation of an enantioselectivity, which suggests that the cobalt should interfere within the C–O bond-forming transition state involving a cationic Co complex [84].

In 2018, using a similar Co–H mediated HAT/radical-polar crossover strategy, the group of Pronin reported a Co-Schiff base-catalyzed hydrofunctionalization of tertiary allylic alcohols that, under ligand control, selectively produces either the corresponding epoxides or semipinacol rearrangements products (Figure 11) [87]. Cyclohexanol (but not cycloheptanol or cyclopentanol) derivatives and 2-methylbut-3-en-2-ol form the corresponding epoxides in yields up to 69% when Co-salen 4c-Co is employed under the reactions conditions described in Figure 11 (top). Interestingly, switching the 1,2-diphenylethylenediamine backbone of the salen ligand by 2,3-dimethylbutane-2,3-diamine or replacing the p-tert-butyl substituents of the catalyst by nitro groups leads to a change in reaction outcome and allows the efficient formation of the corresponding ketone products from semipinacol rearrangement. This strong ligand-dependence of the hydrofunctionalization outcome and additional experiments suggest that the epoxidation reaction involves alkylcobalt(IV) complexes as electrophilic intermediates. Recently, the same research group has explored further the effect of the ligand structure on the enantioenrichment of the epoxidation reaction [88]. Extensive variation of the 1,2-diphenylethylenediamine fragment of the ligand led to the identification of dibenzofuran-containing Co-catalyst 4d-Co as the optimal catalyst affording 4-vinyltetrahydro-2H-pyran-4-ol in 69% yield and with 95% enantiomeric excess (ee) (Figure 11). In contrast, 4c-Co afforded the corresponding epoxide in 60% yield and with 15% ee. Good performance was also observed for various vinylcarbinols derivatives from 4-disubstituted cyclohexane, tetrahydropyran or piperidines (Figure 11). However, when applied to acyclic structures, no enrichment was obtained, while 5- or 7-membered ring-containing substrates did not afford the corresponding epoxides but led to the semipinacol rearrangement products. Mechanistically, the authors proposed the presence of cationic alkylcolbalt(IV) complexes in the enantio-determining step, where cation-π interactions between the radical cation of the Schiff base fragment and the biaryl substituents of the dibenzofuran motif, account for the superior asymmetric induction observed with 4d-Co [87].

Last year, Shigehisa and co-workers also explored the structural effect of chiral cobalt salen catalysts on their previously developed cyclohydroalkoxylation methodology of unactivated alkenes (Figure 11, bottom) [89]. They identified the chiral and bulky binaphtyl ligands as optimal for the reaction of gem-disubstituted alkenols bearing terminal olefin, leading to the formation of various 5-membered rings in good yields and with moderate-to-high enantioinduction. Surprisingly, the authors also demonstrated that the sense and degree of enantioselectivity highly depend on the steric hindrance of the silane derivative. Indeed, the cyclization of 2,2-diphenylpent-4-en-1-ol affords mostly the (S)-enantiomer and the (R)-enantiomer of the cyclic ether with EtSiH₂ and (Me₂SiH)₂O, respectively. This silane-controlled enantiodivergence was experimentally and theoretically investigated and suggests that two competing selective mechanisms are operating involving respectively an alkylcobalt(III) and (IV) intermediate in the selective-determining step. The relative contribution of a radical chain reaction to these steps appears to depend on the silane controlled-concentration of a diffused carbon-centered radical.

4.2. Hydroamination

In 2014, taking advantage of the cobalt-based catalytic system developed for alkene hydroalkoxylation (vide supra), the Shigehisa group reported the intramolecular hydroamidation of aminoalkenes catalyzed by the same catalytic system (Figure 12) [90]. The authors proved again the impressive substrate scope and functional group tolerance of the reaction, affording 3- to 7-membered ring derivatives in good to excellent yields from N-protected amines tethered to mono- and 1,1′-disubstituted alkenes. When competitive hydroamidation vs. hydroalkoxylation can occur, it is worth noting that hydroalkoxylation with free alcohol was favored. The selectivity can be switched with appropriated protecting group on the alcohol, e.g., an acetyl (Ac) group, whereas MOM or TMS resulted in a mixture of C–N and C–O bond formations. Notably, the suitable protecting groups on the nitrogen atom are highly electron-withdrawing groups (Ac, Bz, Cbz, Boc, Ns, Ts, TFA), whereas substrates with free or benzyl (Bn) protected amine are not tolerated. In this respect, the authors show that the Co catalyst system strongly activates the olefin to react with weakly nucleophilic nitrogen atoms. When applying the same methodology to the cyclization of alkenyl guanidines bearing typical N-protecting groups, various five-, six- and even seven- -to eight-membered rings guanidines were efficiently obtained [91]. These transformations likely operate via a cobalt- hydride hydrogen atom transfer and a radical-polar crossover similarly to the C–O bond formation methodologies described above.

Based on the idea that the facile cleavage of weak N–N single bonds could deliver the free amines, the group of Akai and Yahata exploited the ability of the cobalt-Schiff base/N-fluorotrimethylpyridinium salts/hydrosilane combination to promote the Markovnikov selective hydroamination of benzotriazoles as amine surrogates and terminal aliphatic leading to N²-alkylated benzotriazoles (Figure 13) [92]. In this study, racemic cobalt 4f-Co along with PhSiH₃ gave the best performance in terms of yields and regioselectivity. The substrates’ scope revealed the high chemoselectivity of the reaction with a wide range of functional groups (protected alcohols, silyls, esters, halogens) tolerated on both reaction partners. The reduction of the N–N bonds providing the free amines was demonstrated through zinc-mediated reduction in acidic media (Figure 13). Although the enantioinduction was poor using enantiopure (R,R)-4f-Co, this demonstrates that the catalyst interacts with the substrate during the C–N bond formation and offers great promise for the future development of a highly enantioselective Markovnikov cobalt-catalyzed hydroamination reaction. Based on control experiments, the authors postulated that the reaction mechanism proceeds in a similar way as the hydroalkoxylation, through the formation of a cationic alkylcobalt(IV) intermediate by a metal-hydride HAT and radical-polar crossover process. Subsequent S_N2-like displacement with silylated benzotriazole substrate affords the hydroamination product and regenerates the Co(II) catalyst. Notably, through this synthetic strategy, the authors also reported access to a large range of 2,5-disubstituted tetrazoles (18 examples, 43–83% yields) from the reaction of 5-substituted tetrazoles and unactivated aliphatic alkenes [93]. Using enantiopure cobalt catalysts, a promising asymmetric version toward enantioenriched 2,5-disubstituted tetrazoles with ee’s up to 66% was reported.

Based on the similar concept of promoting formal alkene hydroamination via MHAT and through an undirect pathway requiring a N–N bond cleavage step, the group of Lu recently reported a Co-catalyzed hydrohydrazidation of alkenes using diazo compounds in the presence of NNN-tridentate ligand and PhSiH₃ [94]. With a large substrate scope, the efficiency and the high tolerance towards various functional groups were well-established (42 examples, 50–95% yields). Coupled to a Zn-mediated N–N single bond reduction under acidic conditions followed by Boc-, Bz-protection or N-alkylation, a three steps protocol afforded various secondary amines (29 examples, 35–75% yields). Remarkably, when an enantiopure ligand was introduced, hydroamination products were obtained with good enantioselectivity (89:11 to 93:7 enantiomeric ratio (er)). It is worth noting that a similar protocol was successfully applied to sequential hydrogenation/hydrazidation of aliphatic terminal alkynes [95].

In 2019, in order to widen the application scope of this cobalt-catalyzed alkene hydrofunctionalization methodology, the Zhu research group proposed to use an additional catalytic oxidant for a better redox matching during the radical-polar crossover process that converts the organocobalt(III) species (resulting from metal-hydride HAT of alkene followed by cage collapse) into the cationic organocobalt(IV) intermediates [96,97]. By evaluating the electrochemical properties of 4a-Co and its cyclohexenediamine-derivative as well as a model alkylcobalt(III)-salen complex, it was hypothesized that the facile reductive quenching of the excited state of Ru(bpy)³⁺ by organocobalt(III) species should lead to the formation of the key Co(IV) intermediate and that subsequent nucleophilic trapping should give the desired Markovnikov product. The stoichiometric presence of a two-electron oxidant, such as N-fluoro pyridinium salts, should trigger the regeneration of Ru(III) and Co(III) species. This analytic-driven approach has allowed the efficient development of a novel methodology for the Markovnikov intermolecular hydrofunctionalization of vinyl(hetero)arenes and a variety of N-heterocycles (indazoles, imidazoles, pyrazoles, carbazoles) and sulfonamides by a light-mediated cobalt-ruthenium dual catalysis as illustrated in Figure 14. Mechanistic investigations support the postulated photochemical oxidation of alkylcobalt(III) species. This successful merger of a photoredox cycle with Co-mediated HAT should open the door for further interesting developments in the rich chemistry of CoH-mediated oxidative hydrofunctionalization reaction.

In 2018, with the aim of developing a Co-catalyzed direct hydroamination protocol with primary amines (and oxidant-free) and inspired by their previous Fe-catalyzed cyclohydroamination, [64] Ujaque, Lledós and Hannedouche developed a β-diketiminatocobalt(II) complex to promote the exo-cyclization of unprotected aminoalkenes under mild conditions (Figure 15) [98,99]. This report was the first example of cobalt-catalyzed hydroamination of primary amines. It is worth mentioning that contrary to the Fe catalyst [64], a phenyl ring at the terminal position of the alkene was necessary to prevent alkene isomerization as a dominant pathway. This protocol allows the formation and isolation of diverse gem-disubstituted pyrrolidines and piperidines in good-to-high yields. However, it was unsuccessful for the cyclization of amines tethered to trisubstituted alkenes or substrates unbiased toward cyclization but suitable for primary amines bearing dimethyl-substituted allene or alkyne functionality. Joint experimental and theoretical investigations of the reaction mechanism suggested that the reaction proceeds through a nucleophilic attack of the amido group of monomeric cobalt(II) amidoalkene-aminoalkene adduct intermediate to the non-coordinated pendant alkene as the rate-determining step; followed by a rapid proton transfer from the coordinated substrate, which results in the cyclized adduct. This novel stepwise non-insertive mechanism in the field of alkene hydroamination is distinct from the stepwise insertive mechanism encountered in the Fe-catalyzed hydroamination of primary amines tethered to olefins.

5. Nickel

5.1. Hydroalkoxylation

The first example of Ni-catalyzed alkene hydroalkoxylation was reported by Suisse and Sauthier et al. in 2013 (Figure 16) [100]. They observed that the hydroalkoxylation of 1,3-butadiene with methanol, ethanol or benzylic alcohol promoted by a Ni(acac)₂/1,4-bis(diphenylphosphino)butane (dppb) in the presence of NaBH₄ affords alkylbutenyl ethers with high selectivity. In 2016, the same group performed an interesting optimization of this catalytic system to further improve the selectivities and proposed the use of 1,2-bis(diphenylphosphinomethyl)benzene (dppmb) as an alternative to 1,4-bis(diphenylphosphino)butane (dppb) (Figure 16). Under optimized catalytic system, a large range of substituted dienes was performed but a decrease of reactivity was observed compared to butadiene. However, this study proved the efficiency of the catalytic system under mild conditions for a large range of primary alcohols including bio-sourced alcohols [101]. A joint experimental and computational mechanistic study of the dppb- and dppmb-ligated nickel-catalyzed hydroalkoxylation reaction of butadiene discards the involvement of Ni-H species and shows that the formation of Ni-allyl species are likely intermediates accessible from direct protonation of the ligated Ni(0)/butadiene species [102]. The methodology was further applied to the hydroalkoxylation of butadiene with glycerol to obtain glycerylbutenylethers (GBE) [103,104].

In 2019, Mazet and co-workers developed a highly regioselective Ni-catalyzed hydroalkoxylation of 2-(hetero)aryl-substituted 1,3-dienes with primary alcohols under mild conditions (Figure 17) [105]. An excellent Markovnikov selectivity up to >20:1 resulting from a 3,4-addition was achieved by using the achiral dicyclophexylphosphinooxazoline ligand (Cy-PHOX) (Figure 17). The optimized protocol provides an access to a broad range of structurally diverse allylic ethers and tolerates a number of sensitive functional groups, such as halogens, nitriles, esters or pyridine or N-methyl imidazole. Isotopic labeling experiments have shown that the transformation is reversible and expected to involve Ni-π-allyl species reminiscent of those observed in the related hydroamination reactions (vide infra). Encouraging results in the development of an asymmetric version of this process were observed with the use of a ferrocenyl-based enantiopure phosphinooxazoline ligand.

5.2. Hydroamination

Early studies have demonstrated the ability of Ni-based system to catalytically promote the hydroamination of 1,3-dienes [106,107,108]. In 2019, inspired by the work of Hartwig on cyclic dienes [105], the group of Mazet reported a regio- and enantioselective Ni-catalyzed intermolecular hydroamination of 2-substituted 1,3-dienes with primary and secondary amines under mild conditions using a catalytic combination of Ni(COD)₂ and enantiopure biphosphine (R,R)-BenzP* ligand and trifluoroethanol (TFE) as an additive (Figure 18, top) [109]. The method is broadly applicable, highly regio-, chemo-, and enantioselective for 2-(hetero)aryl-substituted 1,3-dienes and primary amines, and provides direct access to valuable enantioenriched allylic secondary amines by selective 3,4-addition. However, the reaction, which is compatible with a range of functional groups, appears more challenging for 2-alkyl-substituted 1,3-dienes or secondary amines leading to a significant decrease in yield and regioselectivity, despite a high enantioinduction. In-depth mechanistic investigations highlight that the C–N bond formation is rate-limiting and proceeds by an outer-sphere nucleophilic addition on a Ni-π-allyl species assigned as the catalytic resting state. In contrast to hydroalkoxylation described in Figure 17, this step is irreversible.

The same year, a novel biphosphine ligated nickel-catalyzed asymmetric hydroamination of monosubstituted 1,3-dienes using a Brønsted acid as cocatalyst was established by the group of Yin (Figure 18, bottom) [110]. By using a catalytic combination of Ni(COD)₂, enantiopure rigid ligand (S,S)-Me-DuPhos and phthalic acid as Brønsted acid, a wide series of primary and secondary amines could be selectively added to 4-aryl and 4-alkyl-substituted 1,3-dienes affording enantioenriched functionalized allylic amines in yields and with enantioselectivity up to 99%. The choice of chiral biphosphine ligand and the use of an appropriate Brønsted acid as cocatalyst were key to the success of this hydroamination reaction. A preliminary mechanistic study indicates that the C–N bond-forming step is, in this case reversible leading to racemization over time for the reaction of secondary amines but not for primary amines. The methodology has a good functional group tolerance and exhibits good chemoselectivity with respect to amines bearing two different nucleophilic sites but is unsuitable to the reaction of disubstituted 1,3-dienes.

Although great progress has been made in the Ni-catalyzed hydroamination of 1,3-dienes and 1,2-dienes [111], there are only a few reports on the development of nickel-based catalysts for the regioselective hydroamination of an unactivated terminal and more challenging internal alkenes [112]. In 2020, the group of Hong made a breakthrough in this domain by reporting an efficient Ni-H-based system having the ability to promote a proximal-selective hydroamination of unactivated alkenes tethered to a bidentate directing group (Figure 19) [113]. Indeed, the combination of Ni(acac)₂/(EtO)₂MeSiH in the presence of electrophilic hydroxylamine esters as amine transfer agent provides a useful synthetic method for the conversion of mono-, di- and trisubstituted alkenes tethered to amide-linked aminoquinoline or to amine-derived picolinamides to structurally diverse and functionalized β- and γ-amino acid building blocks in moderate to good yields. It is interesting to note that both (E) and (Z)-alkenes are compatible as well as mono- and dialkyl-substituted amine transfer agent. The strong bidentate coordination ability of the directing groups to the Ni–H species is crucial to promote and regiocontrol the challenging migratory-insertion of the Ni–H species into the substituted alkene. In-depth kinetic studies and DFT calculations suggest that this step is rate-limiting.

6. Copper

6.1. Hydroalkoxylation

In 2015, the group of Sawamura reported the cyclohydroalkoxylation of gem-disubstituted alkenols catalyzed by a mesitylcopper(I)/Xantphos (L₀) (Figure 21) system according to a different activation mode from other copper Lewis-acid catalysts previously reported [114,115,116] for alkene hydroalkoxylation (Figure 20) [117]. In this system, the basic copper (pre)catalyst activates the alcohol functionality of the substrate to form an alkoxycopper(I) species that can further insert into the terminal alkene and leads after homolysis to the desired cyclic ether. Ligand screening identified Xantphos (L₀) with a large bite angle as the most efficient ligand for the cyclization of 2,2-diphenyl-4-penten-1-ol. The protocol was reported for the hydroalkoxylation of primary and secondary gem-disubstituted alkenols, but not for substrates unbiased for cyclization andwas inefficient for the transformation of alcohol tethered to 1,2-disubstituted alkene. Preliminary studies on the development of an asymmetric version of this protocol revealed that, among the chiral ligand screened, enantiopur (R)-DTBM-SEGPHOS (L₃) (Figure 21) ligand could afford 2-methyl-4,4-diphenyltetrahydrofuran with 71% ee under slight reaction conditions modifications but in low yield. This report was the first example of metal-catalyzed enantioselective cyclohydroalkoxylation of unactivated alkenes.

Last year, inspired by their previous works on copper(II)-catalyzed enantioselective carboetherifications of alkenols [118,119,120,121], the group of Chemler reported the enantioselective cyclohydroalkoxylation of a range of aliphatic 4- and 5-alkenols catalyzed by chiral copper(II) bis(oxazoline) (Figure 22 [122]. In the presence of an oxidant and an H-atom donor, the Cu(OTf)₂/(S,S)-i-Pr-Box (L₇) or (S,S)-Ph-Box (L₈) system is indeed an efficient system to access enantioenriched α-chiral substituted tetrahydrofurans, phthalans, isochromans and morpholine from primary and tertiary alcohols tethered to the terminal, 1,1- and 1,2-disubstituted alkenes as illustrated in Figure 22. This system is limited to (E)-alkenes as (Z)-isomers are less selective, affords higher enantioinduction for tertiary alkenols than primary alkenols and is unreactive for phenols. Isotopic labeling experiments and previous works are in favor of a polar/radical mechanism implying cis-oxycupration of the alkene followed by homolysis of the subsequent σ-Cu–C bond and H-atom transfer. Oxidation of the Cu(I) to Cu(II) species completes the catalytical cycle.

6.2. Hydroamination

Earlier works in the field have highlighted the potential of copper catalysis to promote inter- and intramolecular hydroamination of unactivated alkenes and sulfonamides [123,124,125], or less common protecting group-free primary amines [126,127]. In 2017, the group of Monnier and Taillefer disclosed the first report of copper-catalyzed intermolecular hydroamination of terminal allenes and cyclic secondary amines or anilines using a ligand-free copper(II) triflate salt as precatalyst (Figure 23) [128,129,130]. This method is highly regio- and stereoselective affording exclusively the corresponding linear (E)-allylamines in moderate-to-excellent yields and is compatible with various functional groups (ethers, free tertiary alcohols, halogens, imides) on both partners. A joint experimental and theoretical mechanistic study rationalizes the reaction selectivity in favor of the linear (E)-allylamine product and supports a cationic Cu(I) species as the catalytically active species [131]. This study allows the extension of this protocol to the reaction of various functionalized allenamides, N-allenyl-carbamates, -(1,2)-azoles and -(1,3)-azoles under milder conditions using Cu(NCMe)₄(PF₆) [132] as precatalyst (Figure 23) [128,133]. More recently, it was demonstrated that the protocol using Cu(NCMe)₄(PF₆) was also efficient for the hydroamination of N-allenyl-sulfonamides and trifluoromethyl-containing α-allenyl-α-aminocarboxylates and α-allenyl-α-aminophosphonate with similar regio- and stereoselectivity [134,135].

In 2013, an astounding breakthrough based on an innovative strategy was independently but contemporaneously reported by the groups of Hirano, Miura [136] and Buchwald [137] for the highly regio- and enantioselective intermolecular hydroamination of arylalkenes with exclusive Markovnikov selectivity (Figure 24) [138,139]. This strategy exploits the well-known phosphine-ligated copper-hydride chemistry of reduction [140,141] and the electrophilic character of hydroxylamine esters [142] to propose an original umpolung amination approach for the C–N bond formation. Both reports disclose that the combination of a Cu(I) or (II) salt and a chiral biphosphine ligand in the presence of a reducing silane agent provides highly efficient catalytic systems for the regio- and enantioselective hydroamination of styrene derivatives and O-benzoylhydroxylamines, affording Markovnikov addition products in good-to-excellent yields with a similarly high level of enantioselectivity (Figure 24). The Miura and Hirano combination ((S,S)-Me-Duphos (L₁) or (R,R)-Ph-BPE (L₂) (Figure 21))/CuCl/diethoxymethylsilane/LiOtBu) can be applied to functionalized styrenes and trans-β-substituted styrenes but is unsuitable for cis-β-substituted styrenes and terminal alkenes, although it allows the conversion of strained aza and oxa-bicyclic alkenes [132,143]. The original Buchwald ((R)-DTBM-SEGPHOS (L₃) (Figure 21)/Cu(OAc)₂/HSiMe(OEt)₂) has a broader scope, providing the enantioenriched tertiary amines for functionalized styrenes, cis- and trans-β-substituted and β,β-disubstituted styrenes [133]. Furthermore, using the racemic version of this copper system, terminal and linear aliphatic olefins afford amines in high yields with anti-Markovnikov regioselectivity.

Detailed mechanistic studies conducted on the C–N bond formation of styrenes and O-benzoyl-N,N-dialkyl-hydroxylamine [144], and completed by computational investigations [145] proposed that the reaction proceeds by (1) in situ formation of catalytically competent and monomeric L₃Cu(I) –H (in a putative equilibrium with its dimer or higher-order species or aggregates) from the ligand L₃, Cu(OAc)₂ and the hydrosilane reagent (2) irreversible, enantio-determining and regioselective alkene 2,1-migratory insertion into the Cu–H σ-bond (3) subsequent, stereoretentive electrophilic amination of the ensuing insertive copper species by the amine reagent which leads to C–N bond formation, liberation of hydroamination product and generation of L₃Cu(I)–OBz as the catalyst resting state (4) rate-determining transmetalation between L₃Cu(I)–OBz and the hydrosilane regenerates L₃Cu(I)–H. Formation of a less sterically congested insertive copper species (resulting from migratory-insertion) and electronic stabilization by the adjacent aryl group may account for the anti-Markovnikov and Markovnikov regioselectivity noticed for linear olefins and styrene derivatives respectively. For terminal (and internal) olefin, the hydrocupration step is likely rate-limiting. As suggested by DFT calculations, the superior efficiency of the bulky DTBM-SEGPHOS (L₃) ligand over other biphosphine ligand originates primarily from a better stabilization of the hydrocupration transition state due to greater superior dispersion interactions from the tert-butyl substituent of the phosphine and the alkyl substituent of the olefin [146]. The understanding of the substrate-ligand interactions was exploited to increase the activity of the CuH catalyst through the design of SEGFAST ligand (L₄) (Figure 21) [147]. Indeed, this ligand affords a 62-fold rate increase compared to L₃ in the formal hydroamination of terminal aliphatic olefins and O-benzoyl-N,N-dibenzyl-hydroxylamine. Both 3,5-i-C₃F₇ and 3,5-t-Bu-4-MeOC₆H₂ substituents of the ligand are essential to reach the appropriate balance between catalyst reactivity and stability.

Over the years, this biphosphine CuH-methodology has been impressively extended to a plethora of class of unactivated and less activated alkenes besides styrenes and linear terminal alkenes including a range of various trans-1,2-and 1,1′-disubstituted alkenes, cyclic cis-1,2- disubstituted and even strained alkenes as summarized in Figure 25 and Figure 26, delivering the chiral tertiary amines at an impressively high level of enantioinduction. It is indeed efficient for the regio- and enantiocontrolled hydroamination of vinylsilanes [148], 1,1-disubstitued alkenes [149], symmetric [150] or unsymmetrical (E)-1,2-dialkylsubstituted alkenes [151] (including the feedstock olefin, (E)-2-butene [146]), alkenyl 1,8-diaminonaphthyl boronates [152], vinylphosphine boranes [153], 1-trifluoromethylalkenes [154], enamines [155], N-protected γ-substituted allylic amines [156], unprotected allylic alcohols [157] delivering a range of structurally diverse β-chiral, α- and β-functionalized tertiary amines in moderate-to-high enantiopurity as privileged building blocks (Figure 25). Although minimal alteration of the initial reaction conditions using the DTBM-SEGPHOS ligand (L₃) as privileged ancillary ligand is usually required, structural optimization of the amine transfer agent, hydrosilane, ligand and substrate or use of additional monodentate phosphine ligand (PPh₃, P(p-tolyl₎₃), proton source (tBuOH) or specific external bases (LiOtBu, CsOAc) was needed in some cases to broaden the scope to new compound families and/or overall improve the process efficiency. For instance, the use of an electron-enriched amine transfer agent (such as p-NMe₂C₆H₄CO₂NR¹R²) instead of the original BzONR¹R² was crucial for the productive enantioselective transformation of the challenging unactivated internal alkenes [146], vinylphosphine boranes [149], enamines [151], allylic amines and alcohols [152,153]. This electronic effect may result in a better ability to restore the competent active Cu(I)–H in the rate-determing transmetation step for the catalytic cycle and a dismished non-productive deterioration of the electrophilic amine by the catalyst. In another example, the judicious choice of the ligand (R)-L₅ (Figure 21) and additional base CsOAc was key to suppress the undesired β-F elimination side-reaction from the alkylcopper intermediate for the formation of enantiopur α-trifluoromethylamines of great interest in medicinal chemistry [150]. Additionally, the nature of the protecting group on the nitrogen or oxygen functionality attached at the (homo)allylic position of the internal alkenes significantly improves the reaction efficiency by either enabling the hydrocupration, preventing the detrimental β-H elimination or electronically controlling the regioselectivity [147,152]. In some cases, the addition of a stoichiometric amount of t-BuOH and/or catalytic amount of PPh₃ to the reaction medium could enhance the reaction rate and improve the yield of C–N bond formation product, as is the case in the formation of enantioenriched N-arylamines [158], by suppression undesired reduction pathways [147,152]. This methodology could also be impressively extended to the asymmetric synthesis of three and five-membered N-heterocycles, such as aziridines or α-(hetero)arylpyrrolidines or six-to nine membered benzo-fused analogues by intramolecular cyclization of hydroxylamine ester derivatives (not shown in Figure 25) [159,160].

Although most of the studies on internal alkenes were pursued on (E)-isomers [161], this protocol was recently successfully extended to cis-1,2-disubstituted alkenes known as challenging and problematic alkene partners [162]. Indeed, the CuH-methodology with (R)-DTBM-SEGPHOS (L₃) or (R,R)-Ph-BPE (L₂) ligand (Figure 21) was efficiently applied to the regio- and enantioselective hydroamination of various substituted 2H-chromenes [163], 1,2-dihydroquinolines [164] and benzofurans [165] to access a series of 4-aminochromanes, 4-aminotetrahydroquinolines and N,N-dibenzylaminophenols respectively as important key building blocks (Figure 26). The latter products arise from CuH-promoted ring-opening of the benzofuran scaffold and subsequent regio- and enantio-controlled hydroamination of the resulting arylalkene. Key to the success of these applications was the fine-tuning of the electrophilic amine partner or the use of additional phosphine ligand. Strained alkenes are also a class of alkenes that have been successfully converted into useful N-based compounds (Figure 26). Indeed, CuH-catalyzed regioselective ring-opening hydroamination of 2-aryl and -alkyl-substituted methylenecyclopropanes and O-benzoylhydroxylamines afford the corresponding homoallylic amines via ring cleavage at the more congested position [166]. Ligand screening reveals a notable ligand effect and leads to the identification of CF₃-dppbz (L₆) (Figure 21) as the more efficient bidentate phosphine for this process (Figure 26). In general, unactivated trisubstituted alkenes are an extremely challenging class of substrates for CuH-promoted hydroamination reaction. A strategy to achieve such reactivity is to rely on the release of the ring strain energy of stained trisubstituted olefins. Early last year, the group of Buchwald reported the diastereo- and enantioselective CuH-catalyzed hydroamination of 1-aryl and alkyl-substituted cyclobutenes, and 1-phenyl, 1-alkyl, 1-silyl substituted cyclopropenes leading to the C–N bond formation with different regioselectivity according to the substitution and ring size (Figure 26) [167]. Markovnikov regioselectivity was noticed for 1-arylcyclobutenes while anti-Markovnikov regioselectivity was observed for all other substrates.

The DTBM-SEGPHOS-or Ph-BPE-based CuH-protocol is not restricted to the synthesis of a range of structurally diverse optically active tertiary amines as described above. It has also been applicable to the preparation of various enantioenriched secondary (aryl)amines and amides and primary amines with high efficiency via structural alteration of the amine transfer agent and optimization of the reaction conditions (Figure 27). The use of the electron-rich ester p-NMe₂C₆H₄CO₂NHR³ [168,169], 1,2-benzisoxazole [170] and substituted 2,1-benzisoxazoles [171] as modified electrophilic amine partners gives straightforward access to secondary amines, primary amines and secondary arylamines tethered to a benzylic alcohol respectively in moderate-to-good yields and high enantioinduction. For the reaction with benzisoxazole derivatives, a work-up is needed to liberate the amine functionality. Readily available 1,4,2-dioxazol-5-ones could also be employed as electrophilic amidating reagents to provide a direct access to amides from (hetereo)arylalkenes with high level of enantiopurity (Figure 27) [172,173,174].

This efficient umpolung amination strategy was further applied to the development of asymmetric N- and C3-selective alkylation methods of indoles (Figure 28) [175]. In these methods, indoles derivatives were used as electrophiles to circumvent the natural C3-alkylation of nucleophilic indoles. It was noted that the N-to C3 regioselectivity of the CuH-catalyzed alkylation of indoles by styrene derivatives was ligand-controlled affording either enantioenriched N-alkylated indoles when (R)-DTBM-SEGPHOS was used as supporting ligand or C3-alkylated indoles when (S,S)-Ph-BPE was used (Figure 28) [170,176]. The employ of bulkier N-(2,4,6-trimethylbenzoyl)indole as electrophile is also important to slow down the reductive cleavage of N–O bond by the CuH species and allows the formation of the C–N bond. As a general trend, the DTBM-SEGPHOS-promoted N-alkylation reaction is more efficient than the Ph-BDE-promoted C3-alkylation in terms of yield, regioselectivity and enantioinduction as illustrated in Figure 28. Theoretical calculations rationalized the ligand-controlled regiodivergence observed in this transformation.

The broad generality, functional group tolerance, predictability and catalyst-control of this biphosphine-based CuH-methodology have paved the way for its successful integration into various auto-tandem catalytic process, providing access to numerous classes of amines, such as α-, γ-, or δ-chiral amines, β-amino acids or enamines from allylic derivatives [177], α, β-unsaturated carbonyl compounds [178,179], or alkynes [180,181], with high levels of regio-, stereo- and chemoselectivity. The synthetic utility of this methodology has also been demonstrated in late-stage preparation of numerous pharmaceutically relevant compounds [146,153,164,165,175].

7. Zinc

7.1. Hydroalkoxylation

Very few zinc-based systems have been reported in the literature to catalyze the formation of C–O bond by hydroalkoxylation. Indeed, until 2019, to our knowledge, only Zn(OTf)₂ was reported by the group of Hii for the cyclohydroalkylation of γ–allenols. Whereas the more common 5-exo-trig selectivity was generally observed with various systems like AgOTf, Au and Pt catalysts for this transformation, in an interesting way, the unusual 6-exo-dig cyclization occurs mostly when the reaction is catalyzed by Zn(OTf)₂ [182]. In 2019, Chein et al. reported two efficient and mild catalytic Lewis acid-assisted Brønsted acid systems ZnI₂/TsOH and Zn(OTf)₂/TsOH for various intramolecular alkene hydrofunctionalization reactions, such as hydroalkoxylation, hydrocarboxylation, hydroamination, hydroamidation, hydroarylation and polyene cyclizations [183]. Although only three examples of formation of five- and six-membered cyclic ethers by hydroalkoxylation were reported, this protocol illustrates the potential of zinc-derived catalytic system to form C–O bond by this methodology.

7.2. Hydroamination

Various research groups have studied the potential of zinc in catalysis for hydroamination reactions [184,185]. The first examples of alkene cyclohydroamination catalyzed by zinc derivatives were reported by the Roesky group in 2005 using [N-isopropyl-2-(isopropylamino)troponiminato] methylzinc complex in the presence of [PhNMe₂H][B(C₆F₅)₄] as Brønsted acid co-catalyst [186]. Several modulations of the steric and electronic properties of the aminotroponimate skeleton were further attempted to improve the catalyst activity and stability [180,187,188]. Inspired by this original work, Mandal and co-workers reported well-defined alkyl zinc complexes 6a-Zn and 6b-Zn bearing a symmetrical N-isopropyl and N-cyclohexyl-substituted phenalenyl-based ligand respectively (Figure 29) [189,190]. These complexes, in the presence of [PhNMe₂H][B(C₆F₅)₄] co-catalyst, display similar catalytic activity to that of Roesky’s [N-cyclohexyl-2-(cyclohexylamino)troponiminato] methylzinc complex for the conversion of functionalized primary and secondary aminoalkenes into five- and six-membered rings. As previously observed, the rate is improved in the presence of the Brønsted acid co-catalyst and altered by the nature of the ligand N-substituent.

Li and co-workers described a very effective Zn(CF₃SO₃)₂-catalyzed domino intermolecular hydroamination-ring cleavage of 2,5-dihydrofuran with aniline derivatives (Figure 30) [191]. The homoallylic β-amino alcohols are obtained in moderate yields under mild reaction conditions. The authors propose that the Lewis acid Zn(CF₃SO₃)₂ induces the formation of anilinium triflate which can transfer a proton to 2,5-dihydrofuran leading to the formation of a carbocation. The carbocation is then attacked by the aniline derivative. At the same time, activated by the zinc triflate, the C–O bond of the alkene partner is cleaved by transfer of a triflate anion coming from the metal. Finally, formal elimination of triflic acid leads to the formation of the product terminal double bond.

More recently, the group of Nembenna prepared and fully characterized a zinc complex bearing bulky guanidinato and amido ligands [192]. This complex was very efficient for the intramolecular hydroamination for primary and secondary N-benzyl amines tethered to mono- and disubstituted alkenes. Reactions were carried out at 80 °C with 5 mol% of zinc guanidinate complex and gave the corresponding pyrrolidines and piperidines (14 examples) with almost total conversion in less than 4 h.

The first and sole demonstration of enantioselective Zn-promoted alkene hydroamination was done by the Hultzsch group in 2009 using a bis(organozinc) complex supported by a chiral L-proline modified diamidobinaphthyl ligand [193]. Although moderate enantioinduction were noticed for the cyclized products (<29% ee), the proof-of-principle was clearly established [194].

8. Conclusions

This review has put forward the most relevant advances of the last decade in the addition of E–H (E = O, N) σ–bond onto unactivated alkenes, 1,2- and 1,3-dienes promoted by manganese, iron, cobalt, nickel, copper and zinc catalysis. It is unquestionable that earth-abundant 3d metals have played a major role in the latest developments of metal-catalyzed alkene hydrofunctionalization as witnessed by the burgeoning activity highlighted herein. The inherent properties of these 3d elements have been a strong leverage for the emergence of alternative and complementary strategies that not only reckon on classical two-electron reactivities but also on original one-electron reactivities. These novel strategies have led to some outstanding breakthroughs in terms of scope, control of (stereo)selectivities and applicability. The advent of the chiral biphosphine ligated copper-hydride chemistry with electrophilic amines has revolutionized the field and has brought the hydroamination methodology to a broad synthetic applicability that has never been reached by any other metal-catalyzed hydrofunctionalization process. The growing and fruitful exploitation of cobalt-hydride mediated HAT/radical polar crossover strategies for hydroamination and hydroalkoxylation, as well as iron-hydride mediated HAT for hydroamination, have also successfully reshaped the area via a controlled and metal-promoted radical-type reactivity. The former strategy has even tackled the challenging enantioselective alkene hydroalkoxylation, that has been poorly exploited by 3d late transition metal catalysis. However, these ground-breaking strategies are not without shortcomings as they usually require the need of a substoichiometric reducing agent and/or oxidant and so diverge from the original concept of step- and atom efficiency of direct alkene hydrofunctionalization. Nevertheless, significant advances in this direction have also been reached by iron, cobalt and copper catalysts via more classical activation pathways. Worth noting is also the rebirth of nickel catalysis that delivers efficient catalysts for the 1,3-diene hydrofunctionalization with high level of enantioinduction in hydroamination. Most of the 3d metal catalytic systems reported herein afford Markovnikov selectivity and only some of them display the opposite selectivity. This highlights the challenging issue of selectivity control as Markovnikov and anti-Markovnikov adducts are both compounds of interest. All these thriving advances have elegantly enriched the portfolio of strategies to construct C–O and C–N bond by earth-abundant 3d late transition metal-catalyzed hydrofunctionalization and will undoubtedly guide the development of fascinating synthetic chemistry in the near future.