Nickel‐Catalyzed Anionic Cross‐Coupling Reaction of Lithium Sulfonimidoyl Alkylidene Carbenoids With Organolithiums

Abstract

The mechanistic platform for a novel nickel⁰‐catalyzed anionic cross‐coupling reaction (ACCR) of lithium sulfonimidoyl alkylidene carbenoids (metalloalkenyl sulfoximines) with organometallic reagents is reported herein, affording substituted alkenylmetals and lithium sulfinamides. The Ni⁰‐catalyzed ACCR of three different types of metalloalkenyl sulfoximines, including acyclic, axially chiral and exocyclic derivatives, with sp² organolithiums and sp² and sp³ Grignard reagents has been studied. The ACCR of metalloalkenyl sulfoximines with PhLi in the presence of the Ni⁰‐catalyst and precatalyst Ni(PPh₃)₂Cl₂ afforded alkenyllithiums, under inversion of configuration at the C atom and complete retention at the S atom. In a combination of experimental and DFT studies, we propose a catalytic cycle of the Ni⁰‐catalyzed ACCR of lithioalkenyl sulfoximines. Computational studies reveal two distinctive pathways of the ACCR, depending on whether a phosphine or 1,5‐cyclooctadiene (COD) is the ligand of the Ni atom. They rectify the underlying importance of forming the key Ni⁰‐vinylidene intermediate through an indispensable electron‐rich Ni⁰‐center coordinated by phosphine ligands. Fundamentally, we present a mechanistic study in controlling the diastereoselectivity of the alkenyllithium formation via the key lithium sulfinamide coordinated Ni⁰‐vinylidene complex, which consequently avoids an unselective formation of an alkylidene carbene Ni‐complex and ultimately racemic alkenyllithium.

The nickel⁰‐catalyzed anionic cross‐coupling reaction (ACCR) of lithium sulfonimidoyl alkylidene carbenoids (metalloalkenyl sulfoximines) with organometallic reagents affords substituted alkenylmetals and lithium sulfinamides. The Ni⁰‐catalyzed ACCR of three different types of metalloalkenyl sulfoximines, including acyclic, axially chiral, and exocyclic derivatives, with sp² organolithiums and sp² and sp³ Grignard reagents has been experimentally and theoretically studied.

Introduction

Myriads of studies of nickel and palladium have extensively expanded not only the research spectrum on organometallic catalysis but also their utilization in cross‐coupling reactions. Cross‐coupling reactions catalyzed by late transition metals such as Pd and Ni have been widely exploited for the introduction of various functional groups into unsaturated substrates such as alkenes and aromatic rings.1 Although chemists have shed much light on the nature of catalysis by interchanging transition metals that share isoelectronic properties to mimic their respective complex systems, Ni and Pd have flourished in many other pairs/trios of transition metals in producing fruitful insight to chemists.1d Anionic Pd complexes have proven to play a crucial role in cross‐coupling or Heck reactions through high catalytic activity.2 In a similar vein, anionic Ni‐ate complexes showed promise in overcoming the less efficient oxidative addition of alkyl halides as a Ni‐catalyzed cross‐coupling partner than that of aryl or alkenyl halide.3 Especially, Kambe described the formation of anionic Ni‐ate complexes from organometallic reagents, which show an enhanced nucleophilicity in the cleavage of the C(sp³)−X bond of alkyl halides.4 Cornella recently reported that highly reduced Ni‐precatalysts, which are stabilized by simple olefinic ligands, exhibit an increased catalytic activity in the Kumada–Corriu–Negishi (KCN) cross‐coupling.5 Despite these advances in anionic cross‐coupling reactions (ACCR) and pervasive usage of interchanging Ni and Pd to their respective complex systems, fundamental aspects of electron‐rich Ni‐ate complexes have not been well‐understood and studies in this field remain relatively scarce.

The ubiquitous use of Ni‐catalysts in modern chemistry and the increased interest they received in recent years instigated us to carefully examine the mechanism of the Ni‐catalyzed ACCR of metalloalkenyl sulfoximines with organolithium or Grignard reagents, yielding substituted alkenylmetals.6, 7 In these reactions, metalloalkenyl sulfoximines react under inversion of configuration at the C atom and complete retention at the S atom. Combining experimental and computational techniques, we found a new catalytic cycle that includes an anionic Ni⁰‐ate intermediate induced by organolithium that traverses a putative Ni⁰‐vinylidene intermediate through α‐elimination of the sulfonimidoyl group, which is a nucleofugal leaving group. In this paper, we disclose a complete and detailed mechanism of Ni‐catalyzed ACCR of metalloalkenyl sulfoximines.

Metallo alkylidene carbenoids 1, carrying an electrofugal and a nucleofugal leaving group at the Cα atom, are an intriguing class of compounds (Scheme 1 a).8, 9 Key reactivity features of 1 are electrophilic and nucleophilic substitution at the Cα atom, Fritsch–Buttenberg–Wiechell (FBW) rearrangement, α‐elimination, and formation of transition‐metal‐ate complexes, for example. Lithioalkenyl sulfoximines Li‐2 6, 7, 10, 11, 12 are a new group of non‐classical alkylidene carbenoids, displaying characteristic ambiphilic reactivity (Scheme 1 b). For example, although electrophilic substitution of Li‐2 gives alkenyl sulfoximines 3,6, 10, 12 nucleophilic substitution of Li‐2 with cuprates furnishes alkenyl cuprates 4 through 1,2‐metalate shift of higher‐order cuprates.13 Unlike the classical alkylidene carbenoids 1, Li‐2 carry a chiral nucleofuge, are generally stable in solution up to 0 °C, and do not undergo FBW rearrangement or nucleophilic substitution with organolithium reagents.14 According to experimental and theoretical studies, lithioalkenyl sulfoximines Li‐2, which are obtained by lithiation of alkenyl sulfoximines H‐2, are monomeric in etheral solution and dimeric in the crystal (bridging O−Li bonds), contain a fluxional C−Li bond and a Li atom being coordinated by the N atom or in case of a N‐sulfonyl group by one of its O atoms.15, 16

Scheme 1.

(a) Metallo alkylidene carbenoids. (b) Lithium sulfonimidoyl alkylidene carbenoids. (c) Ni‐catalyzed ACCR.

We observed an interesting stereoselective C−C bond formation of metalloalkenyl sulfoximines using organolithium reagents RLi to form alkenyllithiums 5. It traverses through a Ni‐vinylidene intermediate, which is synthetically derived from either anionic or dianionic Ni⁰‐ate intermediates, in which the sulfonimidoyl group acted as a nucleofuge as described in Scheme 1 c. Of particular interest was the role of the anionic Ni⁰‐ate complex that leads to the Ni⁰‐vinylidene intermediate under reduction of the S(VI) atom of the sulfonimidoyl group to S(IV) and extrusion of PhS(O)N(Li)Me. Herein, the extremely electron‐rich nickel center is necessary for pushing the electron density to the α‐carbon to form a putative Ni⁰‐vinylidene intermediate. In cases of vinylidene‐mediated catalysis, there have been plenty of studies of metal–vinylidene complexes in reaction pathways.17 These studies generally demonstrate that the rearrangement of an acetylene to a vinylidene in the coordination sphere of a transition metal is a thermodynamically favorable process and henceforth has been widely employed for approaching the metal–vinylidene complex. In closer scrutiny, however, there have been very few examples of metal–vinylidene complexes involving Group 10 metals,18 because they have preferred modes of reactivity with alkynes that do not easily allow formation of metal–vinylidene intermediates during the reaction. Instead, Ni₂(vinylidenoid) intermediates have been accessed in vinylidene transfer reactions from dinuclear Ni‐catalysts and 1,1‐dichloroalkenes as vinylidene precursors.19 Herein, we demonstrate that metalloalkenyl sulfoximines can also lead to Ni⁰‐vinylidenes, in which the nickel center is highly reduced.

During synthetic studies of carbocyclic prostacyclin analogs,20 we made the intriguing observation that the exocyclic magnesioalkenyl sulfoximines (E/Z)‐M‐6, obtained through metalation of (E)‐H‐6,21 seemingly partake in a Ni‐catalyzed ACCR with PhMgBr, affording the phenyl‐substituted alkenylmagnesiums (E/Z)‐M‐7 (Scheme 2).6 Furthermore, the acyclic lithioalkenyl sulfoximines (E)‐Li‐8,11 prepared by lithiation of (Z)‐H‐8,11 were found to apparently participate in a Ni‐catalyzed ACCR with PhLi, furnishing alkenyllithiums (E/Z)‐Li‐9 as intermediates.7 In the absence of the Ni‐precatalyst no ACCR took place between the metalloalkenyl sulfoximines and PhM.

Scheme 2.

Nickel‐catalyzed anionic cross‐coupling reaction (ACCR) of metalloalkenyl sulfoximines wtih PhM (M=MgBr, Li).

Shortly after the publication of our preliminary observation with the exocyclic magnesioalkenyl sulfoximines (E/Z)‐Mg‐6, Kocieńsky reported that cyclic α‐lithioalkenyl ethers engage in a copper‐catalyzed ACCR with organolithiums.22 To date, Ni‐ and Pd‐catalyzed ACCRs of metallo alkylidene carbenoids with organometallic reagents have not been described with the exception of the examples depicted in Scheme 2, however. This prompted us to carry out a comprehensive study of the Ni‐catalyzed ACCR of metalloalkenyl sulfoximines M‐2 using experimental and computational techniques.

Results and Discussion

The magnesioalkenyl sulfoximines (E/Z,R)‐Mg‐6 (X=Br) (50:50 dr), which were obtained through metalation of the alkenyl sulfoximine (E,R)‐H‐6 23 with PhMgBr, engaged in an ACCR with PhMgBr (3 equiv) and Ni(dppp)Cl₂ (dppp=1,3‐bis(diphenylphosphino)propane) (6–10 mol %) in ether at 0 °C for 3 h and gave alkenylmagnesiums (E/Z)‐Mg‐7 a (X=Br), which after quenching with CF₃CO₂D furnished the deuterated alkenes (E/Z)‐D‐7 a (93 % D) in 80 % yield and 50:50 dr (Scheme 3 a). A similar ACCR of the lithioalkenyl sulfoximines (E/Z,R)‐Li‐6 (50:50 dr), which were prepared from (E,R)‐H‐6 upon treatment with MeLi, with PhLi (3 equiv) and Ni(dppp)Cl₂ (10 mol %) in diethyl ether at 0 °C for 3–5 h afforded alkenyllithiums (E/Z)‐Li‐7 a. Quenching of the alkenyllithiums with CF₃CO₂H gave alkenes (E)‐H‐7 a and (Z)‐H‐7 a in 80 % yield and 50:50 dr. The ACCR of (E/Z,R)‐Mg‐6 (X=Br) with ClMg(CH₂)₄OSitBuPh₂ 6 (4.3 equiv) in the presence of Ni(dppp)Cl₂ (6 mol %) in diethyl ether at 0 °C for 3 d furnished alkenylmagnesiums (E/Z)‐Mg‐7 b (X=Br, Cl). Quenching of the alkenylmagnesiums with CF₃CO₂H gave alkenes (E)‐H‐7 b and (Z)‐H‐7 b in 27 % yield and 50:50 dr.

Scheme 3.

(a) Ni‐catalyzed ACCR of exocyclic metalloalkenyl sulfoximines. (b) Ni‐catalyzed ACCR of acyclic (E)‐configured lithioalkenyl sulfoximines. (c) Proposed pathway through [1,5]‐retro‐Brook rearrangement of δ‐silyloxy alkenyllithiums. (d) Synthesis of acyclic and cyclic aryl‐substituted homoallylic alcohols, carrying a (Z)‐configured alkenylsilane group.

The acyclic (E)‐configured lithioalkenyl sulfoximine (E)‐Li‐8 a, which was obtained from alkenyl sulfoximine (Z)‐H‐8 a 11 through metalation with MeLi, engaged in an ACCR with PhLi (2 equiv) and Ni(PPh₃)₂Cl₂ (5 mol %) in diethyl ether at 0 °C. To our surprise, quenching of the reaction mixture with CF₃CO₂H furnished the (Z)‐configured alkenylsilane (Z)‐H‐10 a (≥98:2 dr) in 82 % yield instead of the corresponding alkene derived from protonation of alkenyllithium (Z)‐Li‐9 a or (E)‐Li‐9 a (Scheme 3 b). The (E)‐configured lithioalkenyl sulfoximine rac‐(E)‐Li‐8 b, which was obtained through metalation of (Z)‐H‐8 b 11 with MeLi, similarly engaged in an ACCR with PhLi and Ni(PPh₃)₂Cl₂ and afforded the (Z)‐configured alkenylsilane rac‐(Z)‐H‐10 b (≥98:2 dr) in 59 % yield (Scheme 3 c). The ACCR of (E)‐Li‐8 a and rac‐(E)‐Li‐8 b had initially given alkenyllithiums Li‐9 a and rac‐Li‐9 b either as (Z)‐ or (E)‐configured isomers or mixtures of both isomers. Regardless of the selectivity of the ACCR, the establishment of an equilibrium between the (E)‐ and (Z)‐isomers of alkenyllithiums Li‐9 a and rac‐Li‐9 b has to be assumed, because the α‐phenyl alkenyllithiums should have a low configurational stability under the reaction conditions.24 For example, (Z)‐(2‐isopropyl‐1‐phenylvinyl)lithium experienced under ACCR conditions a fast (Z,E)‐isomerization in ether at 0 °C (see the Supporting Information for details). From the equilibrium mixture of the isomeric alkenyllithiums the (E)‐configured isomers selectively reacted through migration of the silyl group from the O to the C atom under formation of the (Z)‐configured alkenylsilanes (Z)‐Li‐10 a and rac‐(Z)‐Li‐10 b. Silyl migration in the (Z)‐configured isomers should be prohibited because of steric reasons. The formation of alkenylsilanes (Z)‐Li‐10 a and rac‐(Z)‐Li‐10 b from the corresponding alkenyllithiums (E)‐Li‐9 a and rac‐(E)‐Li‐9 b are examples of a [1,5]‐retro‐Brook rearrangement involving sp²‐hybridized C atoms.25 Although examples of [1,4]‐retro‐Brook rearrangement have frequently been described,25, 26 those of [1,5]‐retro‐Brook rearrangement are scarce.25, 27 The alternative route to (Z)‐H‐10 a and rac‐(Z)‐H‐10 b starting with a [1,5]‐retro‐Brook rearrangement of (E)‐Li‐8 a and rac‐(E)‐Li‐8 b followed by a Ni‐catalyzed CCR of the corresponding α‐silyl alkenyl sulfoximines24 can be excluded, because of the inertness of the lithioalkenyl sulfoximines towards silyl migration. Ultimately, the Ni‐catalyzed ACCR of (E)‐Li‐8 with aryllithiums followed by a [1,5]‐retro Brook rearrangement of the intermediate alkenyllithiums provides a stereoselective route to acyclic homoallylic alcohols (Z)‐10, carrying a synthetically valuable (Z)‐alkenylsilane group28 (Scheme 3 d), from allylic sulfoximines and aldehydes. Because of the availability of the exocyclic lithioalkenyl sulfoximines (E)‐Li‐11,12g this route should also give access to the exocyclic silyl‐substituted homoallylic alcohols (Z)‐12.

Although we initially observed a Ni‐catalyzed ACCR of (E/Z,R)‐M‐6, its stereoselectivity remained unclear since we could not differentiate in diastereoselectivity between (E,R)‐M‐6 and (Z,R)‐M‐6 or (Z,S)‐M‐6 and (E,S)‐M‐6 as illustrated in Scheme 4 a. Although the exocyclic metalloalkenyl sulfoximines are configurationally stable at −70 °C in diethyl ether and diethyl ether/hexamethylphosphoramide (HMPA) and can be trapped with electrophiles, isomerization is fast at −35 °C, forming 1:1 mixtures of the diastereomers (see the Supporting Information for details). In addition, the ACCR of the acyclic lithioalkenyl sulfoximines (E)‐Li‐8 led to a C−C coupling and an intriguing [1,5]‐retro‐Brook rearrangement to form the (Z)‐alkenylsilanes as described in Schemes 2 and 3. However, the ACCR of the (Z)‐configured lithioalkenyl sulfoximines (Z)‐Li‐8 could not be studied, because of a fast and complete isomerization at −35 °C to the corresponding (E)‐configured diastereomers (E)‐Li‐8 (see the Supporting Information for details),12 as revealed by trapping experiments (Scheme 4 b). Thus, we had to employ a new class of metalloalkenyl sulfoximines as illustrated in Scheme 4 c, in order to investigate the mechanism and stereoselectivity of the ACCR.

Scheme 4.

(a) Isomerization of exocyclic metalloalkenyl sulfoximines, (b) acyclic metalloalkenyl sulfoximines, and (c) axially chiral lithioalkenyl sulfoximines.

The axially chiral lithioalkenyl (N‐methyl)sulfoximines (aS,S)‐Li‐13 a and (aR,S)‐Li‐13 a, which are accessible through metalation of the corresponding alkenyl sulfoximines (aS,S)‐H‐13 a and (aR,S)‐H‐13 a 23 with MeLi, exist as equilibrium mixtures in which the composition and the rate of isomerization are strongly solvent dependent. Although the dr of the lithioalkenyl (N‐methyl)sulfoximines (aS,S)‐Li‐13 a and (aR,S)‐Li‐13 a is 97:3 in diethyl ether at −70 °C (95:5 at 0 °C) and 97:3 in dimethoxyethane (DME) at −50 °C, it is 33:67 in THF at −70 °C to 0 °C. Lithioalkenyl sulfoximine (aS,S)‐Li‐13 a is configurationally stable in ether and DME at low temperatures for a prolonged period of time (see the Supporting Information for details). Isomerization of the lithioalkenyl sulfoximines is in THF significantly faster than in ether (see the Supporting Information for details). However, at low temperatures both lithioalkenyl sulfoximines (aS,S)‐Li‐13 a and (aR,S)‐Li‐13 a can be efficiently trapped with electrophiles in diethyl ether/HMPA or THF. Isomerization of the lithioalkenyl sulfoximines, which are expected to contain a C−Li bond and Li atom being coordinated by the N atom or N‐sulfonyl O atom, most likely involves a solvent‐assisted cleavage of the C−Li bond and formation of N,Li or O,Li contact ion pairs, which undergo inversion at the anionic C atom and rotation around the C−S bond followed by the formation of the C−Li bond to give the corresponding diastereomer.15 The unresolved issue within this dynamic scheme is the strong solvent dependency of the equilibrium between (N‐methyl)alkenyllithiums (aS,S)‐Li‐13 a and (aR,S)‐Li‐13 a and rate of isomerization. The (N‐silyl)sulfoximine (aR,S)‐Li‐13 c is configurationally stable at −70 °C in diethyl ether, but forms a 1:1 mixture of diastereomers (aR,S)‐Li‐13 c and (aS,S)‐Li‐13 c at 0 °C. The substituent at the N atom of the axially chiral lithioalkenyl sulfoximines exerts a strong influence on the rate of isomerization. In contrast to the (N‐methyl)‐ and (N‐silyl)sulfoximines, the (N‐sulfonyl)sulfoximine (aR,S)‐Li‐13 b experiences even at −70 °C in diethyl ether and THF a fast isomerization to give a mixture of (aR,S)‐Li‐13 b and (aS,S)‐Li‐13 b in 50:50 dr. The origin of the difference in configurational stability of the (N‐methyl)‐ and (N‐sulfonyl)alkenyllithiums is unknown.

The axially chiral metalloalkenyl sulfoximine (aS,S)‐Li‐13 a was considered to be a good stereochemical probe, because it carries diastereotopic methylene groups at the double bond and ACCR will afford enantiomeric alkenes. Moreover, the three types of metalloalkenyl sulfoximines, (aS,S)‐Li‐13 a, (aR,S)‐Li‐13 b, and (aR,S)‐Li‐13 c should also serve to explore the substrate specificity of the ACCR. Finally, the axially chiral metalloalkenyl sulfoximines, which carry sulfonimidoyl groups of different nucleofugacity and Lewis basicity,15, 29 were intended to be probes for the influence of the nucleofuge upon the ACCR.

The dependency of the ACCR of metalloalkenyl sulfoximines on a number of variables, including the Ni‐catalyst, Ni‐precatalyst, ligands of the Ni atom, sulfonimidoyl group and reaction conditions, was studied by using (aS,S)‐Li‐13 a as substrate. Generally, ACCR was run in diethyl ether, because of the high dr and the configurational stability of the lithioalkenyl sulfoximine in this solvent at low temperatures. Treatment of lithioalkenyl sulfoximine (aS,S)‐Li‐13 a of 96:4 dr in diethyl ether with Ni(COD)₂ 30 (5 mol %) and salt‐free PhLi (2 equiv)31 at −60 °C to −55 °C for 3 h followed by quenching the mixture with CF₃CO₂D afforded alkenes (aR)‐D‐14 a and (aS)‐D‐14 a with 97 % D in 82 % yield and 84:16 er (Table 1, entry 1). The er of (aR)‐D‐14 a and (aS)‐D‐14 a was determined by ¹H NMR spectroscopy in the presence of AgFOD (FOD=7,7‐dimethyl‐1,1,1,2,2,3,3‐heptafluoro‐octan‐4,6‐dionato) and Pr(tfc)₃ 32 (tfc=(3‐trifluoroacetyl‐d‐camphorato) and by GC on a chiral stationary phase. The absolute configuration of (aR)‐D‐14 a was determined based on a comparison of its chirotropic properties with those reported in the literature for (aS)‐H‐14 a and structurally closely related compounds.33 In addition to (aR)‐D‐14 a and (aS)‐D‐14 a sulfinamide (S)‐H‐15 34, 35 was isolated in 73 % yield and ≥98:2 er. Repetition of the above ACCR experiment with salt‐containing PhLi gave similar results.

Table 1.

Ni‐catalyzed ACCR of lithioalkenyl sulfoximine (aS,S)‐Li‐13 a using PhLi.

Entry	Catalyst/Precatalyst	Yield [%]	D [%]	er
1	Ni(COD)2	82	97	84:16
2	Ni(PPh3)2(COD)	93	[a]	85:15
3	Ni(COD)2/2PiPr3	88	95	83:17
4	Ni(COD)2/2P(OMe)3	94	[a]	82:18
5	Ni(PPh3)2Cl2	91	97	85:15
6[b]	Ni(acac)2	80	[a]	82:18
7	NiCl2/(+)‐DIOP	87	[a]	82:18
8[c]	Ni(PPh3)2Cl2	82	98	64:36[c]

Reaction conditions: (aS,S)‐Li‐13 a, catalyst/precatalyst (5 mol %), PhLi (2 equiv), 3 h, −65 °C to −50 °C. [a] Not determined. [b] 1.5 h. [c] Reaction was conducted with (aR,S)‐Li‐13 a for 4 h at −65 °C, and er represents the ratio of (aS)‐D‐14 a to (aR)‐D‐14 a.

Because of the difference in the 96:4 dr of (aS,S)‐Li‐13 a (equilibrium ratio) (see, Table S3, Supporting Information) and the 84:16 er of (aR)‐H‐14 a and (aS)‐H‐14 a found in the ACCR, control experiments with (aS,S)‐Li‐13 a (96:4 dr), Ni(COD)₂ and PhLi in diethyl ether were run at −65 °C to −55 °C. First, the reaction mixture was quenched after 1.5 h and the alkenyl sulfoximine (aS,S)‐H‐13 a was isolated, the dr of which was 92:8. In a second experiment, the reaction mixture was quenched after 15 min, and the alkenes (aR)‐H‐14 a and (aS)‐H‐14 a were isolated (85:15 er), showing that no racemization of (aR)‐Li‐14 a occurred during the reaction. Racemization of (aR)‐Li‐14 a in diethyl ether is slow. For example, (aR)‐Li‐14 a with 62:38 er had after being kept in ether at 0 °C for 6 h an er of 59:41.

To reveal the possible influence of phosphines, reaction of (aS,S)‐Li‐13 a (97:3 dr) with PhLi was conducted with in situ prepared Ni(PPh₃)₂(COD)36 (5 mol %) in diethyl ether at −60 to −55 °C for 3 h followed by quenching the mixture with CF₃CO₂D, which gave (aR)‐D‐14 a and (aS)‐D‐14 a in 85:15 er and 93 % yield (entry 2). In further experiments, the ligand for the Ni atom was further varied in the ACCRs of (aS,S)‐Li‐13 a (97:3 dr) with PhLi and Ni(COD)₂ (entries 3 and 4), and different Ni^II‐precatalysts were tested (entries 5–7). Complex Ni(PPh₃)₂Cl₂ proved to be an excellent precatalyst (entry 5), which is converted by PhLi to a Ni⁰‐complex.37 Alkenes (aR)‐H(D)‐14 a and (aS)‐H(D)‐14 a were isolated in 80–94 % yield and 82:18 to 85:15 er. ACCR of (aS,S)‐Li‐13 a (97:3 dr) with PhLi and Ni(PPh₃)₂Cl₂ in diethyl ether at 0 °C was less selective. After a reaction time of only 2 min (aR)‐D‐14 a and (aS)‐D‐14 a were isolated in 62:38 er and 84 % yield. Table 1 reveals that the catalyst, the precatalyst(s) and the ligands of the Ni atom have only a minor influence upon the ACCR. We also studied the ACCR of (aR,S)‐Li‐13 a, which is configurationally less stable than (aS,S)‐Li‐13 a. ACCR of (aR,S)‐Li‐9 a of approximately 95:5 dr with PhLi in ether in the presence of Ni(PPh₃)₂Cl₂ at −65 °C for 4 h followed by quenching the mixture with CF₃CO₂D gave alkenes (aS)‐D‐10 a and (aR)‐D‐10 a (98 % D) in 64:36 er and 82 % yield (entry 8). The selectivity of the ACCR of (aR,S)‐Li‐13 a is lower than that of diastereomer (aS,S)‐Li‐13 a. Responsible for the lower selectivity is most likely the competing isomerization of (aR,S)‐Li‐13 a to the more stable diastereomer (aS,S)‐Li‐13 a and its respective ACCR.

Previously, alkenyl sulfoximines (E)‐H‐6 and (aS,S)‐H‐13 a were found to be subject to a highly stereoselective Ni⁰‐catalyzed and Lewis acid‐promoted cross‐coupling reaction with diorganozincs,6, 23, 38, 39 which most likely proceeds by a KCN catalytic cycle.3, 23 The adaption of this cycle to the ACCR involved (i) an oxidative addition of (aS,S)‐Li‐13 a to the Ni⁰‐catalyst to afford a Ni^II‐complex, (ii) a transmetalation with PhLi, and (iii) a reductive elimination to furnish alkenyllithium (aS)‐Li‐14 a as described in Scheme 5 with the dotted red line. Although the KCN cycle leads, under retention of configuration, to alkenyllithium (aS)‐Li‐14 a, the alkenyllithium obtained experimentally in ACCR had the (aR)‐configuration, under inversion of configuration. The observed difference in stereoselectivity led us to deduce that the KCN cycle is not applicable to the ACCR of metalloalkenyl sulfoximines. Moreover, our density functional theory (DFT) calculations indicate that the barrier of the oxidative addition pathway is 27.7 kcal mol⁻¹, which is too high for the given experimental conditions to be mechanistically relevant (see the Supporting Information for details). Therefore, we considered an unconventional and intriguing mechanism, as illustrated in Scheme 5. We imagined that the highly reduced nickel center of the complex, containing a C−Ni bond, may be compelled to form the Ni⁰‐vinylidene complex that is stabilized by lithium sulfinamide coordination and which we were able to locate in our DFT calculations. Our calculations suggest that the reaction may proceed further by coordination of the phenyl anion to the Ni‐center of the Ni⁰‐vinylidene to form a Ni^II‐intermediate with carbanionic character, which requires stabilization by two Li⁺. Finally, reductive elimination of the Ni^II‐intermediate extrudes alkenyllithium (aR)‐Li‐14 a.

Scheme 5.

Solid lines represent the newly proposed reaction pathway of the Ni‐catalyzed ACCR. Dotted lines represent the oxidative addition pathway, which is of opposite stereoselectivity (OA, oxidative addition; TM, transmetalation; RE, reductive elimination).

Figure 1 illustrates the computed free‐energy profile for the proposed mechanism. In our computer model, Ni(PMe₃)₂Cl₂ was used as the precatalyst and (aS,S)‐Li‐13 a was employed as representative substrate. Notably, the electron‐rich Ni‐center in the anionic Ni⁰‐ate intermediate A1 is stabilized by π‐back‐donation to the anti‐bonding orbital of the C=C moiety in the lithioalkenyl sulfoximine, leading to a notable lengthening of the C=C bond from 1.35 to 1.44 Å (see the Supporting Information for details). The carbenoid addition from A1 to the dianionic Ni⁰‐ate complex A10 at 25.1 kcal mol⁻¹ is energetically too high to be mechanistically relevant, as indicated in red. This finding is easy to understand given that in the putative intermediate A10, there is no appropriate electron acceptor that may allow for delocalizing the high electron density of the highly reduced nickel center. Of course, the phosphine ligands are notorious for being excellent π‐acceptors and, thus, we also tested if the arylphosphine ligand PPh₃ could act as an effective electron acceptor. Our calculations indicate that compared with the model ligand PMe₃, the PPh₃‐analogue of A10 is indeed lower, but still too high in energy at 20.6 kcal mol⁻¹ to be meaningful for the overall chemical reaction (see the Supporting Information for details).

Figure 1.

Free energy profile for stereoselective ACCR via lithium coordinated Ni⁰‐vinylidene intermediate. Black and blue traces represent the stereoselective and non‐stereoselective pathway, respectively.

We propose that the productive mechanistic pathway involves the neutral Ni⁰ intermediate A2 located at 4.2 kcal mol⁻¹, which can form the anionic Ni⁰‐ate intermediate A3 through a nucleophilic attack of the carbanion substrate mediated by Li⁺. We found that this process only requires the passage through the transition state A2‐TS at 16.3 kcal mol⁻¹. Subsequently, the transient intermediate A3 must undergo an α‐elimination via A3‐TS at 20.4 kcal mol⁻¹ to form intermediate A4. And again, the presence of Li⁺ is important for this step, as the barrier would be nearly 6 kcal mol⁻¹ higher in energy without it (see Supporting Information for details). Intriguingly, Li⁺ also plays a critical role in determining the stereoselectivity. Lithium sulfinamide‐coordinated Ni⁰‐vinylidene A4 is energetically preferred over the noncoordinated Ni⁰‐vinylidene A7, a finding which is ultimately responsible for the observed stereoselectivity. The geometry of A4 exposes only one possible site for the addition of a PhLi reagent to the Ni‐center, rendering intermediate A5, stabilized by two Li⁺ and located at −17.2 kcal mol⁻¹, the only plausible adduct. Formation of A5 with the opposite configuration at the chiral axis requires a prior isomerization of the Ni⁰‐vinylidene intermediate A4. Attempts to locate a transition state for the putative isomerization of A4 were unsuccessful, but the conversion of A4 to A5 having the aR configuration should be fast given the driving force of ΔG≈−17 kcal mol⁻¹ and should only be limited by diffusion and collision of A4 with PhLi.

Figure 2 illustrates the optimized structures of A4 and A5, intermediates featuring structural and energetic stabilization by Li⁺‐coordination. Ni⁰‐vinylidene A4 shows a Ni−C1 bond length of 1.74 Å, clearly indicating a double‐bond character. Relatively weak Ni−Li interaction is indicated by the calculated bond length of 2.58 Å. Similar metal–lithium interactions have been previously observed in lithiated Fe‐ Co‐, and Ni‐complexes.40 The short distance between the vinylidene carbon and the Li center of 2.26 Å suggests a strong interaction and is analogous to interactions previously reported for metal–carbene complexes stabilized by alkali metals.41 Taken together, this structural pattern demonstrates why the three‐center interaction of Ni−Li−C1 cannot be easily overcome to afford the free Ni⁰‐vinylidene intermediate A7. Moreover, the two Li⁺ interact covalently with C1 as shown in A5 with bond lengths of C1−Li1 (2.13 Å) and C1−Li2 (2.10 Å). The elongated bond length of Ni−C1 of 1.85 Å shows that the two Li⁺ stabilize C1, allowing it to formally remain a carbanion center instead of becoming part of a Fischer‐type metal–carbon double bond. Consequently, the Ni‐center of A5 has a formal oxidation state of Ni^II. Intermediate A5 easily undergoes reductive elimination to form the C−C bond of A6 with a barrier of only 7.6 kcal mol⁻¹, which is 3 kcal mol⁻¹ lower than the barrier of 10.5 kcal mol⁻¹ associated with the alternative transition state A8‐TS. Furthermore, transition state A5‐TS illustrates the impact of the second Li⁺. It facilitates the reductive elimination by temporarily increasing the positive charge of the Ni‐center without actively participating in the redox process. Finally, alkenyllithium (aR)‐Li‐14 a coordinated by (S)‐Li‐15 is extruded from A6 through ligand exchange with (aS,S)‐13 a to give A2, which can start a new catalytic cycle.

Figure 2.

Optimized structures of Ni‐complexes A4 and A5 (bond lengths are represented in Å, and hydrogen atoms have been omitted for clarity).

In summary, our computational study suggests an intriguing mechanistic role of Li⁺ and proposes that the stereoselectivity of ACCR is determined by the electrostatic stabilization of the putative Ni^II‐complex A5 through coordination of two Li⁺, which is in accordance with the experimental observations. To test this hypothesis, we examined the decreasing effect of the Li⁺ when DME is used as the solvent. It is well known that DME can chelate Li⁺ by formation of Li−O interactions and thereby give rise to the formation of solvent‐separated ion pairs.42 In DME treatment of (aS,S)‐Li‐13 a (97:3 dr) with PhLi and Ni(PPh₃)₂Cl₂ (5 mol %) at −60 to −55 °C for 3 h followed by quenching the mixture with MeI at −55 °C to 0 °C gave a mixture of the methylated alkenyl sulfoximines (aS,R)‐Me‐13 a and (aR,R)‐Me‐13 a in 50:50 dr in 74 % yield, but only traces of the methylated alkene derived from a methylation of alkenyllithium (aR)‐Li‐14 a (see below, Scheme 6). Formation of the mixture of the diastereomeric methylated alkenyl sulfoximines in 50:50 dr indicates that the Ni⁰‐vinylidene intermediate A7 and two different isomers of A7/Li‐sulfinamide (Scheme 6) exist in the reaction path and Li⁺ is chelated by DME instead of forming A4 (see the Supporting Information for details). Moreover, it is likely that the carbanion intermediate A4 is formed to afford the Ni^II intermediate A5 that undergoes reductive elimination and advances the C−C bond formation with PhLi. However, because of the reduced activity of Li⁺ it may not accelerate the formation of the Ni^II intermediate A5 from A7 or (aR)‐ and (aS)‐A7/Li‐sulfinamide intermediates, which might be generated because of the different reaction conditions with DME as solvent.

Scheme 6.

Attempted Ni‐catalyzed ACCR of lithioalkenyl sulfoximine (aS,S)‐Li‐13 a with PhLi in DME using PPh₃ as ligand.

The anionic Ni⁰‐ate mediated catalytic pathway was established with Ni(PPh₃)Cl₂ as precatalyst. However, we also obtained experimental results when employing Ni(PPh₃)₂(COD) as the precatalyst that were inconsistent with the aforementioned mechanism. Scheme 7 represents a control experiment with (aS,S)‐Li‐13 a (96:4 dr) and stoichiometric amounts of Ni(PPh₃)₂(COD)36 at −55 °C in diethyl ether/cyclohexene for 2 h in the absence of PhLi. The Ni‐complex did not cause a decomposition of the lithioalkenyl sulfoximine at low temperatures. Instead, the decomposition took place when the temperature was increased to 0 °C,43 indicating that the reaction pathway of A2→A3→A3‐TS does not occur when COD is present as innocent ligand. In this case, the C−S bond cleavage of the lithioalkenyl sulfoximine will be accelerated by the existence of the PhLi reagent at low temperature. Computational studies show that PhLi is not necessary for the Ni(PPh₃)₂Cl₂ system in which the dianionic Ni⁰ intermediate A10 is located at high energy of 25.1 kcal mol⁻¹, whereas in the Ni(COD)₂ system the dianionic Ni⁰ intermediate B3 (see below) has a relatively low energy. Thus, the use of olefinic ligands such as COD for the Ni‐catalyzed ACCR gave the same product with a similar yield (see Table 1) but to our surprise, the reaction path seemed to be different. Recently, formation of anionic Ni⁰ complexes has been described when COD was employed as innocent ligand.5 In these complexes, which display an intriguing catalytic activity, the highly reduced Ni⁰ atom is stabilized by π‐back‐donation to the C=C bond of the olefinic ligand. Given that we sensed for the reaction pathway of the ACCR with Ni(COD)₂ disparities in comparison with the aforementioned one based on Ni(PPh₃)₂Cl₂, we decided to further examine the elusive mechanism of the ACCR via dianionic Ni⁰‐ate intermediates with theoretical methods.

Scheme 7.

Control experiment of (aS,S)‐Li‐13 a with Ni(PPh₃)₂(COD) in the absence of PhLi.

DFT calculations of the ACCR of (aS,S)‐Li‐13 a with PhLi in the presence of Ni(COD)₂ represented in Figure 3 began with the neutral Ni⁰ intermediate B1 that coordinates the lithioalkenyl sulfoximine. In the absence of PhLi, the carbenoid addition traversing the transition state B6‐TS is associated with a reaction barrier of 23.7 kcal mol⁻¹. This barrier suggests that the decomposition of the lithioalkenyl sulfoximine via Ni⁰‐vinylidenes B7 and B8 is not feasible at low temperatures, when the olefinic ligand is employed, as observed in the control experiment mentioned above. In contrast, the addition of PhLi to B1 leads to the anionic Ni⁰‐ate complex B2, which is stabilized by two olefinic ligands, COD and the lithioalkenyl sulfoximine. Next, the carbenoid addition to the Ni atom generates the dianionic Ni⁰‐ate complex B3 at 1.5 kcal mol⁻¹.

Figure 3.

Free‐energy profile for the stereoselective ACCR via dianionic Ni⁰‐ate intermediate. Black and red traces represent the reaction pathway via dianionic Ni⁰‐ate and anionic Ni⁰‐ate intermediates.

Optimized structures of B2 and B3 are shown in Figure 4. Although the anionic Ni⁰‐ate complex could be stabilized by the C=C bond of the lithioalkenyl sulfoximine for both ligands PMe₃ (A1) and COD (B2), only the dianionic Ni⁰‐ate complex B3, carrying COD as ligand, was located at relatively low energy. This differentiates the two distinct mechanisms of the Ni‐catalyzed ACCR with the different innocent ligands because the analogous dianionic Ni⁰‐ate A10 is associated with a high barrier. Electron‐rich [Ni⁰]⁻¹ in B2 engages in π‐backdonation to the C=C bond of the lithioalkenyl sulfoximine that displays an elongated C3−C4 bond (1.43 Å) as a consequence. This orbital interaction is also present in A1, in which the alkenyl double bond is found to be 1.44 Å (see the Supporting Information for details). On the way from B2 to the dianionic Ni⁰‐ate complex B3, the interaction between the Ni atom and C=C bond of the lithioalkenyl sulfoximine is resolved and instead the Ni−C3 σ‐bond is formed. The other olefinic ligand, COD, accepts an electron from the highly reduced [Ni⁰]²⁻ metal center, which is indicated by the elongated bond length of C1−C2 (1.45 Å) in B3, whereas there is no more proper electron acceptor in the PMe₃ coordinated system (A10). The proposed formation of the Ni⁰‐ate complexes B2 and B3 is supported by the synthesis of lithium and dilithium Ni⁰‐ate complexes from Ni⁰ complexes including Ni(COD)₂ and organolithiums.44 Furthermore, the highly electron‐rich [Ni⁰]²⁻ metal center of B3 can easily push some electron density to the alkenyl sulfoximine group, consequently transpiring the C−S bond cleavage with the reduction of sulfur from S(VI) to S(IV). The overall barrier associated with B3‐TS is 16.4 kcal mol⁻¹. The transformation of B3 to the Ni^II complex B4 is mediated by two Li⁺ in an exogenic process. Finally, reductive elimination of B4 via transition state B4‐TS affords alkenyllithium (aR)‐Li‐14 a, as observed when Ni(PPh₃)₂Cl₂ was used as precatalyst.

Figure 4.

Optimized structures of Ni‐complexes B2 and B3 (bond lengths are represented in Å, and hydrogen atoms have been omitted for clarity).

To gain information about the observed lack of a competing ACCR of the lithioalkenyl sulfoximine with alkenyllithium (aR)‐Li‐14 a, ACCRs of (aS,S)‐Li‐13 a (97:3 dr) with vinyllithium and α‐phenyl vinyllithium as model compounds for (aR)‐Li‐14 a were performed (Table 2, entries 1–4). ACCR of (aS,S)‐Li‐13 a (96:4 dr) with vinyllithium in the presence of Ni(PPh₃)₂Cl₂ (5 mol %) in diethyl ether was slow at low temperatures (entries 1 and 2). It occurred with reasonable rate only at 0 °C and gave after quenching the reaction mixture with CF₃CO₂D diene rac‐D‐14 b (83 % D) in 86 % yield (entry 3). A similar ACCR of (aS,S)‐Li‐13 a (96:4 dr) with α‐phenyl vinyllithium in the presence of Ni(PPh₃)₂Cl₂ (5 mol %) at 0 °C in diethyl ether for 1.5 h followed by treatment of the reaction mixture with CF₃CO₂D afforded diene rac‐D‐14 c (42 % D) in only 30 % yield (entry 4). The rate of the ACCR of (aS,S)‐Li‐13 a with the sp²‐organolithiums strongly decreases in the order PhLi≫vinyllithium>α‐phenyl vinyllithium. These data suggest that the ACCR of (aS,S)‐Li‐13 a with alkenyllithium (aR)‐Li‐14 a should be even slower than with α‐phenyl vinyllithium. Isolation of rac‐D‐14 b and rac‐D‐14 c, having a low D‐content, appears to be the result of a lithiation of the corresponding lithioallyl sulfoximine, resulting from a vinyl–allyl isomerization of (aS,S)‐Li‐13 a, by the dienyllithium under formation of the corresponding dilithioallyl sulfoximine and the protonated diene (see the Supporting Information for details). In contrast, attempts to achieve a Ni‐catalyzed ACCR of (aS,S)‐Li‐13 a with alkyllithiums, including MeLi, nBuLi, and LiCH₂SiMe₃, were unsuccessful.

Table 2.

Ni‐catalyzed ACCR based on different lithioalkenyl sulfoximines and alkenyllithiums.

Entry	Alkenyl sulfoximine	R′Li	T [°C]	t [h]	Yield [%]	D [%]	er
1	(aS,S)‐Li‐13 a	CH2=CHLi	−30	3.5	39[a]	41	50:50
2	(aS,S)‐Li‐13 a	CH2=CHLi	−20	3.5	68[b]	79	50:50
3	(aS,S)‐Li‐13 a	CH2=CHLi	0	1	86	83	50:50
4	(aS,S)‐Li‐13 a	CH2=C(Ph)Li	0	1.5	30[c]	42	50:50
5	(aR,S)‐Li‐13 b	PhLi	−75	1	–	–	–
	(aS,S)‐Li‐13 b
6	(aS,S)‐Li‐13 c	PhLi	−75	1	–	–	–
7	(aR,S)‐Li‐13 c	PhLi	−5	5	12	77	50:50
	(aS,S)‐Li‐13 c

[a] A mixture of (aS,S)‐D‐13 a and (aR,S)‐D‐13 a (70:30 dr) and the corresponding diastereomeric allylic sulfoximines in a ratio of 92:8 was recovered in 47 % yield. [b] A mixture of (aS,S)‐D‐13 a and (aR,S)‐D‐13 a (75:25 dr) and the corresponding diastereomeric allylic sulfoximines (Fully deuterated at the α‐position) in a ratio of 92:8 was recovered in 12 % yield. [c] Amount of unreacted sulfoximine was not determined.

In further experiments with axially chiral lithioalkenyl sulfoximines, we studied the influence of the substituent at the N atom upon the ACCR. In stark contrast to the (N‐methyl)sulfoximine (aS,S)‐Li‐13 a, the (N‐sulfonyl)sulfoximines (aR,S)‐Li‐13 b and (aS,S)‐Li‐13 b (50:50 dr), which were synthesized through metalation of alkenyl sulfoximine (aR,S)‐H‐13 b 23 with nBuLi, did not undergo an ACCR with PhLi at −75 °C in diethyl ether for 1 h in the presence of Ni(PPh₃)₂Cl₂ (5 mol %) (Table 2, entry 5). Quenching the reaction mixture with CF₃CO₂D led to a recovery of the deuterated sulfoximines (aR,S)‐D‐13 b and (aS,S)‐D‐13 b (95 % D) in 50:50 dr. Under these conditions, already 30 % of the (N‐methyl)sulfoximine (aS,S)‐Li‐13 a had been converted to alkenyllithium (aR,S)‐Li‐14 a (see above). Correspondingly, no ACCR was observed between the (N‐silyl)sulfoximine (aR,S)‐Li‐13 c, which was obtained from alkenyl sulfoximine (aR,S)‐H‐13 c on reaction with MeLi, and PhLi in the presence of Ni(PPh₃)₂Cl₂ (5 mol %) at −75 °C (entry 6). After the temperature of the reaction mixture was raised to −5 °C and the mixture quenched with CF₃CO₂D, alkene rac‐D‐14 a (77 % D) was isolated in only 12 % yield. Alkenyl sulfoximines (aR,S)‐D‐13 c and (aS,S)‐D‐13 c (98 % D) were recovered in 73 % yield and 50:50 dr (entry 7).

So far the phenyl‐substituted alkenyllithiums (aS,S)‐Li‐14 a and (aR,S)‐Li‐14 a were trapped by protonation and/or deuteriation. It was of interest to see, however, whether the alkenyllithiums can also be intercepted with other electrophiles. Therefore, experiments were carried out by using methyl iodide, dibromoethane, and isobutyraldehyde as trapping reagents (Scheme 8).

Scheme 8.

Trapping of axially chiral alkenyllithiums with electrophiles.

First, the ACCR of (aS,S)‐Li‐13 a (97:3 dr) with PhLi was run in the presence of Ni(PPh₃)₂Cl₂ or Ni(PPh₃)₂(COD)36 (5 mol %). Treatment of thus obtained alkenyllithiums (aR)‐Li‐14 a and (aS)‐Li‐14 a (85:15 er) with MeI at −55 °C to 0 °C afforded the disubstituted axially chiral alkenes (aR)‐14 d and (aS)‐14 d in 82 % yield and 85:15 er. Determination of the er was done by GC on a chiral stationary phase and by ¹H NMR spectroscopy in the presence of AgFOD and Pr(tfc)₃.32 The absolute configuration of (aR)‐14 d was assigned based on its chirotropic properties in comparison with those of (aR)‐H‐14 a. Addition of a cold (−55 °C) solution of (aR)‐Li‐14 a and (aS)‐Li‐14 a in diethyl ether to a solution of 1,2‐dibromoethane in ether at 0 °C gave the axially chiral bromoalkenes (aS)‐14 e and (aR)‐14 e in 68 % yield and 78:22 er. Determination of the er was done by ¹H NMR spectroscopy in the presence of AgFOD and Pr(tfc)₃.32 The absolute configuration of (aR)‐14 e was assigned based on its chiroptical properties in comparison with those of (aR)‐H‐14 a. Trapping of (aR)‐Li‐14 a and (aS)‐Li‐14 a in diethyl ether at −50 °C to 0 °C with iPrCHO furnished a mixture of diastereomeric alcohols, (aS,R)‐14 f and (aS,S)‐14 f, in a ratio of 54:46 in 79 % yield, which were separated by medium‐pressure liquid chromatography (MPLC). The configuration of the stereogenic center of the alcohols was not determined. The minor diastereomer had an er of 84:16 as determined by ¹H NMR spectroscopy in the presence of AgFOD and Pr(tfc)₃.32 The configuration of the chiral axis of (aS,R)‐14 f and (aS,S)‐14 f was assigned in analogy to that of (aR)‐14 a.

Conclusions

The ACCR of metalloalkenyl sulfoximines with PhLi in the presence of the Ni⁰‐catalyst yielded alkenyllithiums and lithium sulfinamide under inversion of configuration at the C atom and complete retention at the S atom. Utilizing NiCl₂ as the precatalyst at optimal conditions of −60 °C, COD and phosphine ligands or Ni(COD)₂ as catalyst proved to be ideal in attaining a favorable yield and er. In search of establishing the first ACCR with nickel, we discovered an interesting feature in which the oxidative addition step does not initiate the ACCR catalytic cycle, further supported by computational studies. Moreover, in conjunction with DFT calculations, we were able to differentiate the distinct pathways with respect to the two different innocent ligands PPh₃ and COD for the reaction system. Fundamentally, the role of Li⁺ is vital. On the one hand, it stabilizes the electron‐rich metal center in the Ni⁰‐intermediates and on the other hand it participates in controlling the reaction barriers that lead to the key Ni⁰‐vinylidene and dianionic Ni⁰‐ate intermediates of the ACCR pathways. The mechanism of the Ni catalyzed ACCR of lithioalkenyl sulfoximines is distinctly different from that of the Cu catalyzed ACCR of lithioalkenyl ethers8c, 8d, 22 and Cu mediated ACCR of litthioalkenyl sulfoximines,13 which has been postulated as key step a 1,2‐metalate shift8 of higher‐order cuprates under displacement of the nucleofuge at the sp² C atom.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

I. Erdelmeier, J. Won, S. Park, J. Decker, G. Bülow, M.-H. Baik, H.-J. Gais, Chem. Eur. J. 2020, 26, 2914.