Catalytic transformations of unsaturated π-bond chemicals to sp³-riched building blocks have attracted increasing interest in the synthetic community [1–5]. Difunctionalization of alkenes constitute an efficient strategy to assemble complex valuable building blocks from simple units [6–9], and can be achieved by either transition-metal catalysis [10,11] or transition-metal-free catalysis [12–15]. These transformations highly reply on the application of styrene and its derivatives, predominantly due to the formation of benzylic stabilized intermediates reducing the possibility of other side-transformations [16]. Chain-walking catalysis provides extensive opportunity to access unique products in both alkene polymerizations [17] and alkene difunctionalizations [18–23]. However, the chain-walking transformations is less likely to occur in the reactions with styrene and its derivatives, due to the formed benzylic stabilized intermediates are less likely to occur β-H elimination (Fig. 1a) [16]. Furthermore, even β-H elimination occurs, following migratory insertion still favors to produce the thermodynamically stable benzylic intermediates. As a consequence, 1,2-addition products, rather than migratory 1,1-addition products, are always favored in the reaction of styrenes [24–28]. Compared with the reactions with unactivated alkenes [29–32], migratory bisfunctionalization of styrenes need overcoming the benzylic stabilization, therefore, is much more challenging.

On the other side, the state of the art of chain-walking stabilizing strategies include benzylic stabilization [33–38], allylic stabilization [39–44], boron stabilization [45–48] and remote coordinating-group stabilization [49–52]. Knowledge on the regioselectivity control between different stabilizing strategies would greatly push forward the development of synthetic methods involving chain-walking [53–59]. However, it is still rather limited to date.

As an extension of our continuous interest in migratory difunctionalization of alkenes [60–63], we speculated that if alkenyl electrophiles were chosen to study a migratory 1,1-carboboration of styrene, it would increase the chance of success. As illustrated in Fig. 1c, either the reaction was initiated by an alkenyl-metal species or a boryl-metal species, it would lead to the generation of an intermediate that has the possibility to afford the migratory 1,1-addition product: if the reaction was initialed by an alkenyl-metal species (Fig. 1c, left side) [64,65], the reaction would generate benzylic metal species, and following 1,2-metal chain-walking led to the formation of an allyl-metal species. Therefore, the reaction involves the competition between benzylic and allylic selectivity. Accordingly, if the reaction was initiated by a boryl-metal species (Fig. 1c, right side), it would involve a competition between α-aryl and α-boronate selectivity. Herein we disclose our success in the above speculations, with the development of a palladium-catalyzed 1,1-alkenylboration of styrenes (Fig. 1b). This success not only represents a new entry to the 1,1-regioselective difunctionalization of styrenes and provides a modular protocol for the synthesis of secondary allyl boronates from easily accessible materials [66–70], but also offers a guide for tackling regioselective issues between benzylic and allylic positions in chain-walking transformations. Particularly notable is that, the reaction involves a conjugated 1,3-diene intermediate, which followed by a unique regioselective migratory insertion. This phenomenon provides plenty of new opportunities for 1,3-diene functionalizations [71].

We commenced this study by optimization of model reaction of styrene (1a), alkenyl triflate (2a) and bis(pinacolato)diboron (B₂pin₂, 3). Nickel catalysts were initially assessed in this three-component reaction, but a mixture of 1,2-addition product 4a and 1,1-addition product 5a were always obtained with poor regioisomeric ratios (see Supporting information for details). Then we shifted our attention to palladium catalysts and the results were summarized in Table 1. Ligand evaluation demonstrated that both bisphosphine and monophosphine ligands were ineffective to this three-component reaction (see Supporting information for details). When phosphoric acid L1 [35] and bidentate N, P-ligand L2 were used, the three-component coupling products were obtained in poor yields with poor regioselectivity (entries 1 and 2). Bisnitrogen-based ligands (L3-L10) could improve the yield, but could not improve the regioselectivity (entries 3–10). To our delight, additive examination (entries 11 and 12) was found that the addition of chloride salts favored the formation of 1,1-addition product, leading to an excellent regioisomeric ratio (up to > 20:1) [40,72,73]. Notably, the effect could be extended to other ligands (entries 13 and 14). Finally, the reaction efficiency was further improved by increasing the reagent dosage, with 1,1-regioselective addition product 5a isolated in 81% yield (entry 15). Control experiments indicated that nickel salts were not competent catalyst under the same conditions (entry 16), the bisnitrogen ligand exerted an effect on the reactivity but not the regioselectivity (entry 17), and the base was indispensable for the success of this cascade reaction (entry 18). Notably, the 1,2-addition product 4a' was never detected in these reactions, demonstrating that this Pd-catalyzed carboboration reaction was initiated by the alkenyl electrophile part [64,74,75].

After identification of the optimal reaction conditions, we next shifted our attention to investigate the generality of this carboboration reaction, and the results have been illustrated in Scheme 1. A variety of substituted styrenes were first assessed, and all could afford the corresponding allylboronic esters in moderate to good yields with good regioisomeric ratios (5a-5l). Notably, the electronic property of the substituents did not affect the regioselectivity (comparing 5b and 5c). Then we examined the alkenyl electrophile partner. Both conjugated and nonconjugated alkenyl triflates were able to smoothly participate in this reaction to furnish the carboboration products (5m-5y). Good to excellent regioselectivity could be always expected from the conjugated alkenyl triflates (5m-5q). But regioisomers derived from the allyl-palladium(Ⅱ) intermediates were isolated from the nonconjugated alkenyl coupling partner (5r-5w), which is also consistent with the reaction was initiated by the electrophile part. Good yields and good regioselectivities were obtained when either complex electrophiles or complex olefins were used (5x-5aa). It is noteworthy that, acrylate esters could also produce the migratory carboboration products under the same reaction conditions (5ab-5ad). The asymmetric version of this 1,1-carboboration reaction was also investigated, and a preliminary experimental result showed that, when a chiral PyrBox ligand L7 was used, the enantioenriched products were obtained in same level of yields with a good enantiomeric excess (Fig. 2a).

To further explore the synthetic potential of this method, follow-up transformations of a representative product was conducted. As outlined in Fig. 2b, the allyl boronate 5r underwent homologation and following oxidation to afford homoallyl alcohol 6, and reacted with benzaldehyde to deliver alcohol 7.

To shed light on the catalytic cycle of this palladium-catalyzed reaction, a terminal deuterium-labeled alkene d2–1b was synthesized and examined under the standard reaction conditions. The corresponding 1,1-alkenylboration product d2–5b was isolated in 53% yield, with deuterium observed at the both α- and β-position with exquisite stereoselectivity (Fig. 2c). This result suggests that the formation of the target product involves a 1,2-Pd chain-walking process, but not a cascade of Heck and reductive Heck reactions [76].

Finally, we proposed a catalytic cycle for this palladium-catalyzed three-component reactions. As illustrated in Fig. 2d, a L10-ligated palladium(0) catalyst (I) is formed and initialed the reaction by oxidative addition with an alkenyl triflate. Following styrene migratory insertion into the resulting alkenyl-Pd(Ⅱ) species (II) to furnish a benzyl-Pd(Ⅱ) species (III). Subsequent β-hydride elimination leads to the formation of conjugated 1,3-diene coordinated Pd(Ⅱ)-H intermediate (IV). Following migratory insertion selectively delivers a more reactive allyl-Pd(Ⅱ) intermediate (V), and then transmetallation with B₂pin₂ leads to the formation of an allyl-Pd(Ⅱ)-Bpin species (VI). Finally, reductive elimination delivers the carboboration product (5) and regenerates the Pd(0) catalyst (I).

In conclusion, we have developed a palladium-catalyzed carboboration of styrenes with alkenyl triflate and B₂pin₂ as coupling partners. This reaction features good 1,1-regioselectivity and broad substrate scope. The addition of chloride salt plays a crucial role in the improvement of reactivity and regioselectivity in this reaction. This study not only provides a protocol for the synthesis of secondary allyl boronic esters from readily accessible starting materials, but also lays a foundation for the solution of site-selective issues of chain-walking transformations. Further mechanistic details are still ongoing in our laboratory.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by grants from the National Natural Science Foundation of China (No. 22122107) and the Fundamental Research Funds for Central Universities (No. 2042021kf0190).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.108087.