Cooperative Pd/Cu Catalysis for Alkene Arylboration: Opportunities for Divergent Reactivity

Abstract

A preeminent challenge in alkene difunctionalization is the control of regio-, diastereo-, and enantioselectivity. In this Perspective, a Pd/Cu-cooperative catalytic system for alkene arylboration is highlighted that allows for the controlled introduction of substituents. In particular, examples that allowed for divergent reactivity from a single substrate based on the tuning of catalysts and reaction conditions are emphasized.

Graphical Abstract

Cooperative catalysis involves the activation of two substrates by two different catalysts that subsequently undergo a reaction to achieve product formation.¹ This type of catalysis offers additional opportunities compared to single catalyst systems not only for the possibility of achieving a greater number of transformations in a single step but also due to the potential for each catalytic cycle to be tailored toward selective and divergent product outcomes. For this reason, among others, cooperative catalysis has emerged as a research area of intense interest.¹

Alkene difunctionalization is an important strategy for chemical synthesis because molecular complexity can be rapidly generated from simple alkene precursors.² However, an inherent challenge with alkene difunctionalization is the control of selectivity (regio-, diastereo-, and enantioselectivity) (Scheme 1A). Classical strategies rely on substrate design or directing groups to aid in the control of selectivity.³ However, an ideal scenario would be one in which each group could be programmatically added to the alkene with control of the regio-, diastereo- and enantioselectivity without substrate interference or assistance. Without question, this is a lofty goal in alkene functionalization; however, cooperative catalysis offers the opportunity for the precise introduction of substituents where each catalyst is responsible for a distinct aspect of selectivity (Scheme 1B). In the hypothetical scenario illustrated in Scheme 1B, Cat-1 and Cat-3 control the regioselectivity for the introduction of the first substituent, whereas Cat-2 and Cat-4 allow for stereoselective incorporation of the second group. Properly harnessed, this approach has great promise for starting with a single isomer of substrate to produce a multitude of divergent, difunctionalized analogs.

Scheme 1.

Alkene Functionalization

While many useful methods have been developed to achieve selective alkene difunctionalization, our lab has been interested in alkene carboboration,⁴ particularly via cooperative catalysis. These reactions are useful because in addition to the generation of a new C–C bond, a synthetically versatile C–B bond is formed.⁵ In this Perspective, we highlight examples in which Pd/Cu-cooperative catalysis has been used to achieve divergent selectivity from a single starting substrate in alkene arylboration reactions.^6,7

The inspiration and basis for many of the arylboration reactions highlighted in this Perspective originated from a seminal 2006 report from Sadighi and co-workers.⁸ In this paper, it was described that NHC–Cu–Bpin complexes can undergo syn-addition to alkenes to generate stereodefined alkyl–Cu complexes. In addition, as the reaction generally occurs with activated alkenes, such as styrene, the formation of a benzyl–Cu complex is heavily favored. Since these contributions, many research groups have investigated catalytic processes in which the generated alkyl–Cu complex is captured by an electrophile to generate either protoboration adducts or a myriad of difunctionalized products.^4,9

In 2014, Semba and Nakao¹¹ and our lab¹⁰ disclosed independent reports for the Pd/Cu-catalyzed arylboration of alkenyl arenes (Scheme 2A). In general, these reactions are proposed to proceed by addition of an in situ generated LCu–Bpin (Scheme 2, catalytic cycle, intermediate II) to an alkene to form an alkyl–Cu intermediate (III), which undergoes Pd-catalyzed cross-coupling (Scheme 2, Pd cycle).

Scheme 2.

Summary of Pd/Cu-Catalyzed Arylborations of Alkenes Developed to Date (2021)

Since the initial report from Semba and Nakao¹¹ and our lab,¹⁰ several contributions have been made in this area. With respect to advancing the arylboration of alkenyl arenes, our lab investigated the reactions of 1,2- and 1,1-disubstituted derivatives (Scheme 2B,C).¹² In addition, enantioselective variants emerged from our lab and that of Liao.^12b–d,13 The concepts learned in these studies have been extended to the reactions of alkenylBdan derivatives (Scheme 2D).^14,15 The examples with alkenyl bromides are particularly noteworthy as the products can be useful starting reagents for a double-allylation sequence.¹⁵

During the development of alkenyl arene arylboration, the reactions of dienes emerged. For example, our lab reported on the arylboration of 1- and 2-substituted 1,3-dienes (Scheme 2E).¹⁶ More recently, Liao et al. developed enantioselective reactions of enynes to generate chiral allenes (Scheme 2F).¹⁷

The move away from the reaction of conjugated alkenes has been met with resistance. However, our lab has been able to develop arylboration reactions of strained alkenes such as norbornene^12b and cyclobutene¹⁸ derivatives (Scheme 2G). While reactions of unactivated alkenes are particularly challenging, Semba et al. have made a notable advance (Scheme 2H).¹⁹ While not highlighted in this review, recent developments in Ni catalysis have led to the development of a general process for 1,2-arylboration of unactivated alkenes.^20,21

Among all the methods illustrated in Scheme 2, the control of selectivity has been achieved in some context, thus partly fulfilling the project goal outlined in Scheme 1. However, several examples outlined in Scheme 2 (highlighted in green text) have taken steps toward the controlled, divergent introduction of substituents from a conserved starting substrate.

DIASTEREODIVERGENCE

After the publication of the initial report for the arylboration of vinylarenes, our attention then focused toward arylboration of 1,2-disubstituted alkenyl arenes to allow for the synthesis of molecules with two adjacent stereogenic centers.^12a Key to development of this type of reaction was identifying a Pd catalyst that selectively allows for either syn- or anti-transmetalation of the stereodefined alkyl–Cu complex. The evaluation of various Pd catalysts led to the finding that [Pd–RuPhos]²² in THF allowed for the synthesis of the syn-diastereomer, whereas [Pd–Pi–Bu₃] in toluene led to the synthesis of the anti-diastereomer (Scheme 3A,B). The observed stereodivergence is consistent across a range of substrates. Some of these concepts have been extended to enantioselective and diastereodivergent reactions of alkenyl arenes and alkenylBdan derivatives.^12b,15

Scheme 3.

Arylboration of Alkenyl Arenes

Mechanism studies confirmed that the stereodivergence occurs at the transmetalation stage of the reaction (Scheme 3C). More specifically, the alkyl–Cu complex could be generated as an intermediate and then subjected to the Pd-catalyzed cross-coupling conditions. In each case, the expected stereoisomer was observed as the major product. For the reaction promoted by [Pd–RuPhos], a four-centered, stereo-retentive transmetalation is proposed (Scheme 3C, top).²³ In the case of the reactions promoted by [Pd–Pi–Bu₃], (Bpin)₂ and NaOt-amyl were found to be necessary to observe high diastereoselectivity. Therefore, it is proposed that the addition of a Lewis base is necessary for the stereoinvertive transmetalation via an S_E2 type process (Scheme 3C, bottom). Similar stereoinvertive S_E2 transmetalations invoke Lewis bases to facilitate the process.²⁴ Finally, it should be noted that solvent plays an important role in both transformations; however, its role is unclear.

REGIODIVERGENCE (1,2 vs 1,4)

The arylboration of 1,3-dienes has also been extensively explored.¹⁶ The development of these reactions is complicated for several reasons: (1) the addition of the NHC–Cu–Bpin complex can generate η¹-1,2-, η¹-1,4-, or η³-π-allyl–Cu complexes, (2) the transmetalation can occur through either a direct or an allylic process,²⁵ and (3) the reductive elimination can occur at one of the two sites.

With these challenges in mind, our group set out to develop a regioselective arylboration of 1-substituted dienes (Scheme 4).^16b From these investigations, it was uncovered that the 1,2-addition product was favored with IPr–CuCl in combination with [Pd–PAd₂n-Bu],²⁶ whereas the 1,4-product was formed using SIMes–CuCl in conjunction with [Pd–JackiePhos]²² (Scheme 4A,B).

Scheme 4.

Arylboration of 1,3-Dienes

The mechanistic investigation revealed that the addition of the Bpin unit occurs at the terminal position and the Cu center is located at the 1-position, indicative of a formal 1,4-borylcupration (Scheme 4C). In addition, the Z-alkene was the major product from the reaction. While the allyl–Cu complex was observed with the reaction of both SIMes-CuCl and IPr–CuCl, it was only stable when isolated from the latter. Thus, the treatment of the complex derived from IPr–CuCl to the Pd-catalyzed cross-coupling conditions led to the formation of the observed 1,2-addition product. While an allylic transmetalation is likely, it is not clear at this time whether an open-or closed-transition state is operative. In the case of the complex derived from SIMes–CuCl, while it could be observed by ¹H NMR, it proved to be too unstable to carry out the analogous stoichiometric studies. However, given that the observed 1,4-addition product contains a Z-alkene and the generated alkyl–Cu complex is of the same configuration, they are likely connected on the reaction coordinate via the allyl–Pd complex.

REGIODIVERGENCE (1,2 vs 1,1)

The 1,2-arylboration of 1,1-disubstituted alkenyl arenes is challenging due to the formation of a sterically congested C–C bond. Despite this challenge, our lab was able to develop this process to allow for the synthesis of a quaternary carbon.^12c During our studies, we uncovered an unexpected 1,1-arylboration reaction and viewed this as a unique opportunity to tailor the system for regiodivergent outcomes. Whereas several Pd catalysts allowed for the synthesis of the 1,2-arylboration product (e.g., [Pd–APhos]),²⁷ the 1,1-arylboration required the use of [Pd–PCy₃] (Scheme 5A,B).

Scheme 5.

Arylboration of 1-Substituted Alkenyl Arenes

On the basis of mechanistic studies, the regiodivergence occurs after the formation of the alkyl–Cu complex during the transmetalation with Pd. In the case of the 1,2-arylboration, the generated alkyl–Cu complex undergoes transmetalation and rapid reductive elimination. In the case of the 1,1-arylboration, after transmetalation, the reductive elimination is slow and a β-hydride elimination/reinsertion process becomes competitive (Scheme 5C). It is proposed that the rapid reductive elimination reaction for the 1,2-arylboration occurs due to the steric pressure induced by APhos. In the case of the 1,1-arylboration, mechanistic studies revealed that a phosphine-free Pd complex is likely operative. In the absence of large phosphine-based ligands, reductive elimination is slow and, therefore, β-hydride elimination dominates to provide eventual access to the 1,1-arylboration products.

CONCLUSIONS AND FUTURE PERSPECTIVE

As seen with the highlighted examples, Pd/Cu-catalyzed arylboration reactions can be tuned to achieve divergent reaction outcomes in several different contexts. Despite such successes, many challenges remain within the field: (1) The mechanistic understanding needs to improve. While in most of these Pd/Cu cooperative catalysis systems the basic premise and timing of the steps can be elucidated due to the stability of the alkyl–Cu complexes (allowing for their stoichiometric generation and subsequent study with the Pd catalytic cycle), the delineation of ligand and/or substrate effects is much more difficult. This, in turn, makes it quite difficult to predict and design Pd/Cu systems with a specific outcome in mind (i.e., predicting which exact system will provide a 1,2-arylboration product featuring a stereoinvertive transmetalation).Therefore, to aid in the development of future systems, we think that ligand parametrization and/or multivariate analyses that incorporate substrate effects will prove useful.²⁸ (2) Some of the systems outlined in Scheme 2 result in the formation of one isomer. While undoubtedly important, it would be valuable to be able to tune the catalysts or conditions to achieve divergent reactivity, for example, probing whether a catalyst can be identified to generate the 1,2-arylboration adduct of a 1,3-enyne (see Scheme 2F). (3) Most of the systems that have been identified for divergent reactivity rely on tuning of the Pd catalysts. This is typically because, with activated alkenes, the initial borylcupration occurs along the path that allows for the generation of a stable π-benzyl or π-allyl complex. Additional efforts need to be placed on identifying systems where the Cu catalyst is tuned to achieve catalyst-controlled regioselective borylcupration. For example, in the case of the arylboration of cyclobutenes (Scheme 2G), perhaps a Cu catalyst could be identified that would ultimately result in the formation of the other regioisomer.⁷ (4) Finally, there is a need to increase substrate scope. While there have been recent advances involving the reaction of unactivated alkenes and heterocyclic motifs, more work needs to be done to expand the utility of this system in these areas. Additionally, secondary or tertiary alkyl electrophiles remain a formidable challenge for Pd/Cu cooperative catalysis. Innovations in the above areas, as well as others, will continue to aid in achieving the goal of selective alkene functionalization and the synthesis of complex molecules from simple substrates.