New technologies for carbon-carbon bond formation have a profound impact on fields such as organic synthesis. Among these, constructing C(sp³)-C(sp³) bonds from alkenes is particularly challenging yet highly significant. In 2014, the Baran group developed an iron-catalyzed reductive cross-coupling between electron-rich and electron-deficient alkenes, and the Melchiorre group recently reported a similar photoredox-catalyzed reaction; however, the products were obtained as racemates. Transition-metal-catalyzed hydrodimerization of alkenes provides a new option for C(sp³)-C(sp³) bond formation. The groups of Gong Hegui and Shu Wei independently developed head-to-head hydrodimerization of electronically neutral terminal alkenes and head-to-tail cross-hydrodimerization of electron-rich with electronically neutral alkenes, respectively. Nevertheless, the asymmetric C(sp³)-C(sp³) coupling of electron-deficient alkenes remains a significant challenge. Recently, the team of Prof. Wei Shu (Southern University of Science and Technology) and Prof. Meiwan Chen (University of Macau) achieved the asymmetric cross-hydrodimerization between vinyl phosphine sulfides and unactivated alkenes, providing access to α-chiral phosphine compounds. This reaction offers excellent control over both head-to-tail regioselectivity and enantioselectivity, while avoiding the use of alkyl electrophiles and nucleophiles (Figure 1).

The authors initially explored the coupling of various electron-deficient alkenes with unactivated alkenes and found that vinyl phosphine sulfides could undergo chemoselective and regioselective coupling with unactivated alkenes. Subsequently, using vinyl phosphine sulfide (1a) and 1-octene (2a) to evaluate reaction conditions, it was determined that at room temperature, with NiBr₂•dme (10 mol%) as catalyst, a pyridine-oxazolidine ligand (L1, 12 mol%) as ligand, 3-bromocyclohexene (OX1) as oxidant, in a system containing trimethoxysilane (3.0 equiv) and acetone (0.2 M), the head-to-tail cross-hydrodimerization product 3a​ was obtained in 63% yield, 90% ee, and 7:1 rr. Further testing of allylic bromides revealed that acyclic allylic bromides could improve the regioselectivity of hydrometalation, with OX5​ proving optimal, affording 3a​ in 64% yield, 89% ee, and 18:1 rr. During ligand evaluation, the authors discovered that the pyridine ligand with a 6-methyl substitution (L1) most significantly promoted the reaction, markedly enhancing both efficiency and selectivity. The addition of a phenolate (0.5 equiv) improved enantioselectivity via anion exchange. After optimization, 3a​ was obtained in 68% yield, 91% ee, and 15:1 rr. Control experiments confirmed that the oxidant is indispensable for the reaction, as no target product was formed in the absence of OX5​ (Figure 2).

After establishing the optimal conditions, the authors investigated the substrate scope of the nickel-catalyzed head-to-tail asymmetric alkyl-alkyl cross-coupling between unsaturated phosphine sulfides and unactivated alkenes (Figure 3). Regarding unactivated alkenes, terminal alkenes with different chain lengths (3a, 3b) yielded the target products in 61% and 59% yields with 91% and 88% ee, respectively. Substrates containing alkyl chloride or bromide groups (3c, 3d) reacted in 61% and 58% yields with ee values of 85% and 92%, with the halogens serving as handles for further modification. α-Branched alkenes (3e, 3f) were obtained in 72% and 61% yields with 88% and 89% ee, and regioselectivity rr >20:1. The reaction also tolerated various functional groups including ethers, esters, and boronic esters (3g–3z), and was applicable to 1,1-disubstituted alkenes (3aa). The absolute configuration of product 3f​ was determined as Sby X-ray crystallography. The reaction was successfully applied to the late-stage modification of complex alkenes derived from coumarin and drug molecules (3ab–3ae). For the vinyl phosphine sulfide substrates, diphenyl and 2-naphthyl derivatives (4a, 4b) were obtained in 49% and 41% yields with ee >91%. Aryl-substituted substrates bearing electron-withdrawing or electron-donating groups (4c–4h), heteroaryl substrates (4i, 4j), and internal alkenyl substrates (4k, 4l) also performed excellently, with ee values up to 98%. In contrast, a vinyl phosphine oxide substrate showed poor reactivity, yielding only 7% product, indicating that coordination of the sulfur atom to nickel is crucial for both reaction activity and regioselectivity.

To validate the synthetic potential, the authors conducted a gram-scale reaction (1.0 mmol scale) (Figure 4A), successfully obtaining product 3a​ in 53% yield, 89% ee, and 15:1 rr. Reduction of the phosphine sulfide efficiently provided α-branched chiral monophosphine ligands, offering a new synthetic route for this class of difficult-to-access ligands. A series of control experiments were performed to probe the reaction mechanism: In deuterium labeling experiments (Figure 4B), d₂-unsaturated phosphine sulfide (1a-D) reacted with 4-allylanisole to give product 7​ (51% yield, 92% ee) without deuterium scrambling, indicating that nickel-hydride insertion into the unsaturated phosphine sulfide generates a secondary alkyl-nickel species with regioselectivity and irreversibility. Reaction of d₂-4-allylanisole (2z-D) with 1a​ afforded product 8​ (41% yield, 90% ee), also without deuterium scrambling, confirming the regioselectivity and irreversibility of nickel-hydride insertion into the unactivated alkene. Kinetic analysis (Figure 4C) showed a first-order dependence on the catalyst and zero-order dependence on the unsaturated phosphine sulfide, unactivated alkene, base, silane, additive, and oxidant, suggesting that catalyst generation might be the rate-determining step. Investigation of the oxidant fate (Figure 4D) showed that besides 26% recovery of the oxidant, homodimerization product 9​ (23% yield) and radical addition product 10​ (20% yield) were also formed.

Based on these findings, a plausible reaction mechanism was proposed, as illustrated in Figure 5. The reaction is initiated by the formation of a ligated Ni(I)-Br species (A), which is converted to a Ni(I)-H species (B) by the action of silane and base. B​ undergoes regioselective migratory hydrometalation with the unsaturated phosphine sulfide (1), generating an alkyl Ni(I) species (C). C​ undergoes single-electron oxidative addition with oxidant OX5​ to afford an alkyl Ni(II) intermediate (D). In the presence of base, D​ undergoes a second transmetalation with silane to yield an alkyl Ni(II)-H intermediate (E). E​ then engages in regio- and enantioselective hydrometalation with the unactivated alkene (2) to form a dialkyl Ni(II) species (F). F​ is subsequently oxidized by OX5​ via single-electron transfer to generate a dialkyl Ni(III) intermediate (G). Finally, reductive elimination from G​ produces the alkyl-alkyl cross-coupling product (3) and releases Ni(I), thus completing the catalytic cycle.

This study developed a nickel-catalyzed asymmetric cross-hydrodimerization between vinyl phosphine sulfides and unactivated alkenes. Through a redox-neutral sequence, this reaction achieves the head-to-tail asymmetric alkyl-alkyl cross-coupling of neutral unactivated alkenes with electron-deficient alkenes, generating aliphatic phosphine compounds bearing an α-chiral center that are difficult to access by other methods. The key to success lies in the compatibility of the oxidant and reductant, coupled with exquisite control over the chemo-, regio-, and enantioselectivity for both alkene partners. The methodology demonstrates potential in the late-stage derivatization of natural products and the synthesis of chiral monophosphine ligands.

Authors: Jian-Yu Zou, Fang-Li Xing, Quan-Xing Zi, Hai-Wu Du, Meiwan Chen, Wei Shu

Title: Asymmetric alkyl-alkyl coupling between electron-deficient and unactivated alkenes to access α-chiral phosphines by Ni catalysis

Journal: Science Advances

DOI: 10.1126/sciadv.adv657