1.1. Biaryl formation

Unsymmetrical biaryls are key structural motifs found in a broad range of biologically active compounds, including natural products, pharmaceuticals, functional materials, and agrochemicals [29, 30, 31]. Traditionally, their synthesis involves the transition-metal-catalyzed cross-coupling of an aryl halide with an organometallic reagent—such as arylboronic acids, arylzinc, stannane, or Grignard derivatives—typically mediated by palladium or nickel catalysts.

Pioneering work by Kharasch and coworkers has demonstrated that aryl Grignard reagents can undergo efficient homocoupling in the presence of catalytic amounts of first-row transition metal salts such as cobalt(II) chloride (CoCl₂) [32]. Beyond homocoupling, cobalt has also proven to be an effective catalyst for cross-coupling reactions between two different aryl partners. Notably, Nakamura et al. [33] and von Wangelin et al. [34] have independently showed that cobalt salts catalyze the cross-coupling between aryl Grignard reagents and aryl or heteroaryl halides, enabling the formation of unsymmetrical biaryls. This strategy has since been extended to include aryl tosylates as electrophilic partners by using a cobalt diisopropyl disubstituted bis(phosphino)pyridine pincer (^iPrPNP)-based catalyst [35].

In a complementary approach using commercially available precursors, cobalt precatalysts bearing N-heterocyclic carbene (NHC) ligands, have also been employed in Suzuki–Miyaura cross-coupling reactions of aryl chlorides and bromides with arylboronic pinacol esters, activated in situ with alkyllithium reagents [36].

To circumvent the need for stoichiometric organometallic reagents, the first cobalt-catalyzed eXEC was developed for biaryl synthesis. Initially, the method used aryl halides and 4-chloroquinolines as electrophilic partners (Scheme 2) [37]. This was later expanded to include couplings between two distinct aryl halides (Scheme 3) [38]. In both cases, the reactions were conducted in an undivided electrochemical cell equipped with an iron anode and required pyridine as a ligand to activate the cobalt catalyst.

The aromatic moiety involved in these transformations can accommodate a broad range of electron-donating and electron-withdrawing substituents. In 2008, the cobalt-catalyzed synthesis of biaryls via eXEC was successfully adapted to conventional chemical methods by changing some parameters (Scheme 4). These included reactions between two aryl halides or between an aryl halide and a heteroaryl halide with comparable reactivities [39]. In this protocol, manganese powder was employed as the reductant in a DMF/pyridine solvent system. The cobalt(II) bromide (CoBr₂) catalyst, ligated with triphenylphosphine (PPh₃), enabled the efficient formation of unsymmetrical biaryls in good yields. With these adapted reaction conditions, the proposed mechanism proceeds via a non-radical pathway.

We applied this strategy to the synthesis of 2-(4-tolyl)benzonitrile, a key intermediate in the preparation of sartans—widely used antihypertensive agents. More recently, we demonstrated that cobalt catalysis could be extended to inert C–CN bond activation, thereby enabling the challenging formation of biaryls from aryl halides and benzonitriles—an approach that poses a greater synthetic challenge (Scheme 5) [40]. In this system, the addition of AlMe₃ to the reaction medium enhanced the reactivity of the aryl–CN bond.

Combined DFT calculations and experimental studies revealed that the reaction proceeds through the in situ generation of two low-valent cobalt species: a Co(I) species that activates the aryl halide, and a Co(0) species that facilitates the cleavage of the aryl–CN bond. Consistent with earlier findings by Chatani et al. in rhodium catalysis [41], an intramolecular version of this transformation was also shown to be feasible, albeit requiring prolonged reaction times.

1.2. Formation of vinyl arenes

Although Heck et al. initially reported the formation of styrene in low yield from vinyl acetate, via palladium catalysis, the first efficient XEC between an aryl and a vinyl compound was achieved through electrosynthesis using a sacrificial anode and cobalt catalysis. This approach enabled, for the first time, the use of vinyl acetates as reagents, compounds that are readily accessible from abundant carbonyl precursors, inexpensive, stable, and environmentally benign.

Although being considered highly attractive alkenyl reagents, vinyl acetates are rarely employed due to their low intrinsic reactivity. Daves and Arai described the first direct Pd-catalyzed vinylation starting from vinyl acetate and iodo compounds. However, poor yields were obtained [42, 43]. Our group was the first to report a cobalt-catalyzed electrochemical vinylation of aryl halides using vinyl acetates (Scheme 6). This reaction was performed in an undivided cell equipped with an iron anode and a mixture of acetonitrile/pyridine [44]. This method afforded good to excellent yields, although it required a stoichiometric amount of bipyridine, which remains a notable limitation.

In 2005, we adapted this transformation to a conventional chemical method using manganese as the reductant, this time employing only a catalytic amount of bipyridine (Scheme 7) [45]. The proposed mechanism involves a six-membered transition state in which the aryl group adds to the more substituted carbon of the double bond.

Later, Shu et al. described several nickel-catalyzed XECs for C–C bond formation using alkenyl acetates as starting materials [46].

A similar strategy was applied to the synthesis of stilbenes from β-bromostyrenes and aryl halides (Scheme 8) [47]. These compounds are of significant interest due to their presence in the manufacture of industrial dyes, dye lasers, optical brighteners, scintillator, other materials, and a wide range of biologically active molecules, including resveratrol and combretastatin. Stilbenes can be obtained via XEC between aryl halides and bromostyrenes using cobalt catalysis. However, halostyrenes (X = Cl or Br) are more reactive than vinyl acetates and must be added dropwise to prevent dimerization.

Aryl bromides bearing electron-donating or electron-withdrawing groups in the ortho, meta, or para positions afforded good yields. Moreover, this method proceeds with complete retention of the double bond configuration, in contrast to reactions catalyzed by nickel.

1.3. Formation of 1,3-dienes

1,3-Dienes are a prominent class of functional molecules with diverse applications in materials science, natural product synthesis, and pharmaceuticals. Recent years have seen significant progress in both their synthesis and utilization. Beyond classic methods such as the Wittig and Julia reactions, 1,3-dienes can be accessed through elegant transition-metal-catalyzed coupling strategies. This includes palladium-catalyzed Suzuki, Stille, and Heck reactions involving vinyl halides, as well as transformations using alkynes, allenes, and carbenes. Moreover, ruthenium-, iridium-, and palladium-catalyzed processes offer straightforward access to 1,3-dienes as well.

Nickel-catalyzed XECs developed by Gong and Shu et al. allow the formation of vinyl–vinyl frameworks from vinyl halides and vinyl triflates, or between two vinyl halides [48, 49]. While cobalt-catalyzed couplings using vinyl Grignard or vinyl zinc reagents with vinyl halides, triflates, or acetates have been reported, the only cobalt-catalyzed XEC between two vinyl compounds was described by Beng et al., involving α-bromo enones and α-bromo enamides for the synthesis of azepanes, piperidines, and benzotropones (Scheme 9) [50].

2.1. Coupling of aryl halides with allylic compounds

This initial transformation involving aryl halides and allyl acetates was reported by electrosynthesis. Unlike nickel catalysis, which required dropwise addition of reactive allyl chlorides or acetate [51], the cobalt system allowed the addition of allyl acetates at the beginning of the reaction, simplifying the procedure. The reaction was conducted in an undivided cell with an iron anode, using cobalt bromide as the catalyst in acetonitrile/pyridine (Scheme 10) [52].

Aryl bromides bearing electron-donating or -withdrawing groups in ortho, meta, or para positions were well tolerated. In contrast, aryl chlorides required electron-withdrawing substituents. Heteroaryl substrates such as bromothiophenes or chloroquinaldines were also compatible. Remarkably, unlike many other methods, the major product was the linear isomer. This electrochemical synthesis was patented in 2001 [53]. Due to the perceived complexity of electrochemical methods at the time, we later developed a conventional chemical variant. Although this reaction was the first XEC involving a non-noble metal, yields were lower with substituted allyl acetates compared to the electrochemical approach [54]. Depending on the nature of the halide on aryl moiety, conditions are different (Scheme 11).

2.2. Coupling of aryl halides with benzylic compounds

Cobalt catalysis also enables coupling of aryl halides with benzylic compounds. Prior to exploring XECs, Gosmini and Knochel jointly demonstrated that arylzinc [55] or benzylzinc [56] reagents could react with benzyl chlorides or aryl halides, respectively, under cobalt catalysis. However, these methods required the use of preformed organozinc species.

To overcome this limitation, a direct cobalt-catalyzed reductive arylation of benzyl chlorides with aryl halides was developed to synthesize diarylmethanes—key motifs in pharmaceuticals, agrochemicals, and supramolecular chemistry [57]. This protocol, which tolerates a wide range of functional groups, proceeds under mild conditions (rt to 35 °C) in acetonitrile without the need for ligands and uses pyridine as a cosolvent (Scheme 12). Heteroaryl bromides were also successfully employed.

Due to environmental concerns associated with halides, alternative electrophiles are being explored. Weix et al. reported a dual Ni/Co-catalyzed coupling of benzyl mesylates with aryl halides employing cobalt phthalocyanine as a radical precursor [58]. Inspired by this, our team demonstrated that benzyl mesylates—generated in situ from benzylic alcohols—could replace benzyl chlorides using simple cobalt bromide catalysis without ligands (Scheme 13) [59]. This method tolerates various functional groups on both benzyl and aryl moieties.

More recently, this strategy was extended to bench-stable benzyl sulfonium salts (Scheme 14) [60]. Under conditions similar to those optimized for mesylates, initial cross-coupling reaction yields were low but improved significantly upon addition of bipyridine. Mechanistic studies suggest that low-valent cobalt activates both aryl halides and C–S bonds via single-electron transfer, generating benzyl radicals.

2.3. Coupling of aryl halides with non-activated alkyl halides

Earlier studies primarily utilized activated alkyl electrophiles such as benzyl or allyl derivatives. Jacobi von Wangelin et al. used iron or cobalt catalysis in combination with magnesium as a reductant to expand the scope to non-activated alkyl halides. However, the underlying mechanism involves the in situ formation of Grignard reagents, which inherently limits the functional group tolerance of the reaction [61].

Later, a more versatile system based on CoBr₂/iPr₂PhP/Mn in DMF/pyridine at 30 °C was developed (Scheme 15). This protocol tolerated functional groups and gave improved yields when using monodentate phosphine ligands, in contrast to bidentate nitrogen ligands. A slight excess of alkyl bromide was required, unlike the stoichiometric amounts typically sufficient for more reactive alkyl partners.

The mechanistic hypothesis is similar to other cobalt-catalyzed diarylmethane formation (Scheme 16). The initiating step is the reduction of the Co(II) precatalyst into the active low-valent Co(I) species by manganese. Subsequent oxidative addition to the aryl bromide forms an aryl Co(III) intermediate, which is reduced to arylcobalt(II). At the same time, the reduction of alkyl compound leads to the alkyl radical intermediate. Finally, the alkyl arylcobalt complex, generated from the combination of the alkyl radical and the arylcobalt complex, provides the cross-coupling product by reductive elimination together with the regeneration of the active species Co(I), closing the catalytic cycle.

2.4. Coupling of vinyl compounds with benzyl chloride

Beyond aryl partners, vinyl halides can also be coupled with benzyl chlorides. A cobalt/manganese system using CoBr₂(PPh₃)₂ was reported in acetonitrile with NaI as additive (Scheme 17). Electron-withdrawing groups required elevated temperatures (50 °C), while electron-donating groups allowed reactions at 0 °C to room temperature. Lipshutz et al. [62] previously described a similar transformation using palladium without organic solvents. More recently, Reisman and Cherney [63] developed an enantioselective nickel-catalyzed cross-coupling of secondary benzyl chlorides with styrenyl bromides. However, their method was not stereospecific: Z-alkenyl bromides yielded exclusively E-products. In contrast, the cobalt/manganese system preserved the stereochemistry of the starting material.

3.1. Coupling of alkyl halides with allyl acetates

While cobalt-catalyzed homocoupling of non-activated alkyl halides has been reported [65], heterocoupling has also been achieved—either between an activated and a non-activated alkyl electrophile, or between two activated alkyl species such as benzyl or allyl compounds.

In 2011, we reported the first reductive cobalt-catalyzed Csp³–Csp³ formation by describing the allylation of alkyl halides using either allyl acetates or allyl carbonates, depending on the nature of the alkyl halide (Scheme 18) [66]. The reaction was effective with primary, secondary, and tertiary alkyl substrates and was conducted in a mixture of acetonitrile and pyridine.

For reactive alkyl halides such as benzyl chlorides, alkyl iodides, or alkyl bromides bearing β-electron-withdrawing groups, more reactive allyl carbonates were preferred over acetates, and the reaction temperature was lowered to 50 °C (versus 80 °C). With substituted allyl compounds, the linear product was consistently favored. Experimental evidence supported the involvement of alkyl radical intermediates.

3.2. Coupling of activated alkyl compounds with allylic compounds

More recently, the cobalt-catalyzed reductive alkylation was extended to other activated alkyl electrophiles, including benzyl sulfonium salts and Katritzky pyridinium salts, and XEC of benzyl sulfonium salts with O-allyl electrophiles, using simple CoBr₂ and Mn as reductant in acetonitrile without ligands, was demonstrated (Scheme 19) [67]. This system activated both C–S bonds (from sulfonium salts) and C–O bonds (from allyl esters or ethers), enabling the synthesis of functionalized alkyl–olefinic compounds in moderate to good yields. Notably, these reactions involved two non-halogenated electrophiles—a first in cobalt-catalyzed C–C bond formation.

While nickel-catalyzed cross-coupling reactions with Katritzky salts had been described by Watson et al., Ni et al., and Rueping et al. [68, 69, 70], they typically involved alkyl metals or aryl halides. The only cobalt-catalyzed variant was reported by Kojima and Matsunaga et al. via cobalt/organophotoredox dual catalysis, but it was limited to secondary alkyl pyridinium salts [71]. A similar transformation using benzylic Katritzky salts was also reported [72]. This method offers a simple and efficient alternative for synthesizing allyl–benzyl derivatives, using CoBr₂, under mild conditions (Scheme 20). The reaction was also applicable to allyl alkyl ethers, as previously shown with benzyl sulfonium salts. Mechanistic studies suggest the formation of benzyl radicals from pyridinium or sulfonium salts, and π-allylcobalt intermediates from allyl compounds.

Benzyl sulfonium salts were also shown to couple with benzyl chlorides or alkyl iodides, yielding functionalized alkyl derivatives [73] (Scheme 21). Preliminary mechanistic investigations suggest the involvement of two radical species—one from the sulfonium salt and the other from the alkyl halide.