Download PDF
Synlett 2024; 35(04): 405-411
DOI: 10.1055/a-2112-9552
account
11th Singapore International Chemistry Conference (SICC-11)

Enantioselective Palladium-Catalyzed Suzuki–Miyaura Reactions Enabled by Ionic Ligand–Substrate Interactions

Ivan K. W. On
,
Ye Zhu  ∗

We gratefully acknowledge the support from the Ministry of Education - Singapore (academic research fund MOE2019-T2-2-139). I.K.W.O thanks the National University of Singapore for a research scholarship.
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Abstract

Enzymes harness an array of noncovalent interactions to accomplish stereospecific transformations. Similarly, chemists have engineered chiral catalysts capable of eliciting noncovalent interactions for asymmetric synthesis. In this context, incorporating ionic groups into synthetic transition-metal catalysts represents a promising design element for enantioselective reactions by engaging electrostatic interactions between ligands and substrates. However, the nondirectional nature of ionic interactions presents a unique challenge in precise transmission of chirality. This account summarizes our recent work on developing phosphine ligands possessing nonligating ionic groups for exerting long-range stereocontrol in Suzuki–Miyaura reactions.

1 Introduction

2 Remote Quaternary Stereocenters

3 Mechanically Planar Chiral Rotaxanes

4 Atropo-enantioenriched Biaryls

5 Conclusions


#

Key words

stereoselectivity - cross-coupling - noncovalent interactions - quaternary stereocenters - rotaxanes - atropisomerism - mechanically planar chirality - ionic interactions
1

Introduction

Asymmetric transition-metal catalysis represents a versatile and irreplaceable tool in the arsenal of synthetic chemists. Chiral supporting ligands play a pivotal role in accomplishing enantiocontrol in widespread applications of asymmetric catalysis. Over the past decades, synthetic chemists have been actively exploring the structural diversity of ligands in order to improve reactivity and stereoselectivity. Traditionally, the design elements of chiral ligands have been largely centered on steric repulsion. The chiral ligand provides a stereochemically asymmetric environment, typically represented by ‘fences’ surrounding the metal center. In this way, stereochemical control is achieved by blocking all of the diastereomeric pathways that lead to the formation of the disfavored stereoisomers through repulsive interactions between the ligand and the substrate (Scheme [1], top left).

Zoom Image
Scheme 1 Ionic interactions in asymmetric transition-metal catalysis

As natural catalysts, enzymes harness an array of attractive noncovalent interactions to accomplish stereospecific transformations, by steering the substrates to the active sites, and orientating the reactive group to the catalytic sites. In recent years, interest in emulating enzymes by engaging secondary attractive interactions in asymmetric catalysis has experienced a renaissance.[1] This approach emphasizes on accelerating one of the diastereomeric pathways through the use of multiple noncovalent interactions between the substrate and the catalyst to achieve substrate recognition, pre-organization, and ultimately, stereoselective catalysis (Scheme [1], top right).

Ionic interactions are ubiquitous noncovalent interactions. The electrostatic attraction between oppositely charged species is instrumental in biological processes such as protein folding, ligand binding, and enzymatic processes. Ionic interactions are unique; the influence of an electric field can be felt over distances as far as 5 Å.[2] The range is nearly a two-fold increase in comparison to hydrogen bonds, which typically operate at a distance of approximately 3 Å.[3] Conceivably, harnessing electrostatic attractions as in the secondary coordination sphere of transition metals offers unprecedented opportunities in asymmetric catalysis.

However, the adoption of ionic interactions as secondary noncovalent interactions to transition-metal catalysis pales in comparison to that of hydrogen-bonding interactions.[4] Presumably, the challenges of applying ionic secondary interactions in enantioselective transition-metal catalysis can be attributed to the lack in directionality of ionic bonds.[4] The resultant substrate–catalyst pre-organization complexes could possess high levels of conformational freedom, thus rendering the energetic differentiation between the diastereomeric pathways ineffective.

Undeterred, chemists have circumvented the hurdle posed by the inherent lack of directionality of ionic bonding and have innovated strategies that engage ionic substrate–ligand interactions in asymmetric transition-metal catalysis (Scheme [1], bottom). Recently, the first example of a directional ionic bond was achieved by Čorić and co-workers through the use of large, sterically congested nonpolar pockets to limit the exposed ionic surface to a single direction.[5] By the same underlying principle, it is possible to engage ionic substrate–ligand interactions by tethering a chiral ion (Z1) to a metal–ligand complex covalently (A). Importantly, the chiral ion (Z1) is embedded within a sterically demanding pocket, which imparts a high level of directionality to the ionic interactions with an ionic group (Z2) of the substrate. As the chiral information is encoded in the chiral ion (Z1), such a strategy is applicable if the prostereogenic reactive site (S1) is situated close to the ionic group (Z2), and in most cases, the enantio-differentiating step is actually taking place at the ionic group itself.

In addition, the ionic group of the chiral catalyst (Z3) can function as a substrate-binding site through electrostatic interactions with the ionic group of the substrate (Z4), thus steering and orientating the substrate to the catalytically active metal center (B). In comparison with (A), this approach could be applicable to establishing chirality through reaction at a position (S2) that is remote from the ionic group (Z4). Therefore, the chiral environment immediately surrounding the substrate–binding group (Z3) might not confer effective stereocontrol. Instead, the spatial arrangements between the metal catalytic center and ionic group (Z3), as defined by the chiral backbone of the ligand, become pivotal in effective stereocontrol.

Zoom Image
Scheme 2 A summary of bifunctional ionic chiral ligands over the years (not to scale)

The explicit use of ionic catalyst–substrate interactions to direct enantioselectivity was first reported by Yamaguchi[6] in 1990 and Ito[7] in 1992. This pioneering work led to increasing interest in the development of chiral ligands bearing nonligating ionic groups for this mode of catalysis (Scheme [2]). A wide range of metal-catalyzed enantioselective processes have been accomplished, through the contributions of Chen,[8] Ooi,[9] Peters,[10] Zhang,[11] Miller,[12] Guinchard,[13] Phipps,[14] and others[1] to this fast-growing field. This account summarizes our recent work on developing chiral dialkyl biarylphosphine ligands that possess nonligating ionic groups in order to exert long-range stereocontrol in Suzuki–Miyaura reactions.[15] [16] [17] [18]


# 2

Remote Quaternary Stereocenters

Creation of chirality at a prostereogenic center through a transformation at a remote position represents a distinct synthetic strategy that decouples the reaction sites and the chirality element. A breakthrough on a remote desymmetrizing Ullmann reaction was made in 2016 by Miller and co-workers (Scheme [3]A).[12a] The peptide-based copper catalyst pre-organizes the prochiral substrate through a combination of proximal metal coordination and distal cation–arene interactions (Scheme [3]B).

Zoom Image
Scheme 3 A desymmetrizing Ullmann coupling by Miller et al.

In 2018, Phipps and co-workers achieved high levels of site-selectivity in Pd-catalyzed cross-coupling reactions using sulfonated phosphine ligands (e.g., sSPhos).[19] It was proposed that the differentiation between meta- and para-aryl chlorides was made possible through ionic substrate–catalyst binding via bridging potassium cations, facilitating oxidative addition of the meta-aryl chloride to palladium (Scheme [4]).

Zoom Image
Scheme 4 Site-selective Suzuki–Miyaura coupling by Phipps et al.

We focused our research on engineering chiral catalysts by exploiting secondary substrate–ligand interactions to achieve stereoselective access to challenging chiral elements. In the context of catalytic desymmetrization, spatial separation of a prostereogenic center from either the reaction sites or the prochirality-revealing groups (i.e., points of difference between geminal substituents) presents a significant hurdle.[20] Therefore, we chose to develop an enantioselective catalytic transformation where both the enantiotopic reaction sites and the prochirality-revealing groups are distant from a prostereogenic quaternary carbon.

Inspired by the work of Miller[12a] and Phipps,[19] we explored an axially chiral 5′-phosphonate phosphine ligand L1 and its performance in catalytic remote desymmetrization reactions. We developed a synthetic route to access enantioenriched L1 starting from commercially available RuPhos. The resulting Pd catalyst system was then applied to desymmetrization of fluorene and xanthene scaffolds. The Suzuki–Miyaura coupling at one of the enantiotopic chloro groups establishes chirality at a remote quaternary center (Scheme [5]A).[15]

Zoom Image
Scheme 5 A desymmetrizing Suzuki–Miyaura reaction to establish remote quaternary centers

In this case, the lack of directionality of ionic interactions proved to be beneficial. The catalytic system is highly adaptive to various structural changes in the substrates. Notably, the desymmetrization proceeded stereoselectively on fluorene substrates when the length of the aliphatic linkers between the carboxylate group and the prochirality center and the reaction sites was extended (Scheme [5]B, 1–12). Contrary to our initial expectations, it is viable to achieve long-range enantioinduction by engaging ionic substrate–catalyst interactions at a remote position, even as far as 12 connecting atoms away from the reaction site. In addition, the use of sulfonates as the catalyst binding group provided satisfactory results, e.g., products 8 and 9. The results ruled out the possibility that the carboxylate group acts as a directing group through binding to Pd directly.

We were intrigued by the ability of the catalyst to both convey long-range enantiocontrol and adapt to the structural changes of the substrates, which contrasts with the fact that its ligand possesses a sole chiral element without sterically congested, deep chiral pockets. This unique feature prompted us to carry out control experiments to elucidate the roles of individual ionic components in this reaction using 3 as the model substrate (Scheme [5]C). First, masking the carboxylate group of 3 as an ester led to formation of a racemic product (Scheme [5]C, top, 3 vs 3-CO2Et). Replacing an ionic directing group with an amide as a hydrogen bond donor led to diminished enantioselectivity (Scheme [5]C, top, 3 vs 3-amide). The anionic phosphonate group of L1 was found to be equally critical. Intriguingly, when the phosphonate group of L1 was masked as a phosphonate ester, the reaction remained enantioselective (Scheme [5]C, middle, 3 vs 3-a). Plausibly the phosphonate ester is capable of engaging ion–dipole interactions with the potassium cation, albeit with reduced effectiveness. Removal of the phosphonate group rendered the reaction unselective (Scheme [5]C, middle, 3 vs 3b). Finally, we probed the role of the potassium cation through encapsulation using 18-crown-6 (Scheme [5]C, bottom). We observed a reduction in enantioselectivity, which was in parallel to the amount of 18-crown-6 added. These results cemented the fact that potassium ions are involved in the transmission of chirality from the chiral catalyst to the substrate. The collective participation of the carboxylate of the substrate, the phosphonate of the catalyst and the potassium cations in the selectivity-determining step of the desymmetrization reaction guided us to propose a stereochemical model as shown in Scheme [5]B (dashed box).


# 3

Mechanically Planar Chiral Rotaxanes

The previous study suggests that the long-range stereocontrol enabled by the ionic chiral catalyst is effective even when both the enantiotopic reaction sites and the anionic catalyst-binding groups are distant from a prostereogenic quaternary carbon. Therefore, the strategy could be applicable to the creation of other types of chiral elements beyond point chirality. In other words, the ionic chiral catalyst recognizes the spatial arrangements of the enantiotopic reaction sites and the anionic catalyst-binding groups, regardless of the types of prostereogenic elements that connect them. To put the ionic stereocontrol to a rigorous test, we turned our attention to mechanically planar chirality,[21] where these components of the molecules are interlocked through noncovalent mechanical bonds.

To this end, we further evolved the ionic catalyst systems to introduce an aromatic spacer between the phosphine backbone and the anionic groups (i.e., carboxylate) using a synthetic procedure that we developed for the preparation of 3′-aryl dialkyl biarylphosphine ligands.[22] In this way, the spatial arrangements between the metal-binding phosphorus atom and the substrate-binding carboxylate groups can be fine-turned by varying the substitution patterns of the aromatic spacer. Eventually, we achieved desymmetrization of prochiral rotaxanes using Pd–L2 as the optimal catalyst (Scheme [6]).[16]

Zoom Image
Scheme 6 Catalytic desymmetrization to give mechanically planar chiral rotaxanes

We propose that the biscarboxylate catalyst attracts the biscarboxylate capping group of the axle component of the rotaxane via bridging cations, which directs the Pd catalyst to recognize one of the enantiotopic aryl chlorides of the macrocyclic component of the rotaxane. Evidently, the nature of the ionic interactions is likely nondirectional as the anionic binding group of the catalyst is exposed in the open. Nevertheless, the ionic interactions are sufficient to distinguish the enantiotopic aryl chlorides through long-range stereochemical relay across the conformationally flexible mechanical interlocking bond.

Using 13 as a model substrate, we successfully demonstrated the catalytic desymmetrization reaction with excellent enantioselectivity. As a major drawback, the yield of the reaction was low. Upon studying the mass balance, we observed formation of a hydrodechlorination product in 12% yield (13-H), which was as comparably enantioenriched as 13. Furthermore, we obtained 13% of a bis-arylated product (13-bis), and recovered 40% of the unconsumed starting rotaxane substrate.

Several mechanically planar chiral rotaxanes (13–17) have been prepared with good to excellent enantiopurities, which highlights the advantage of the desymmetrization strategy that decouples the catalytic reaction from the assembly of the interlocking bonds. Control experiments elucidated the critical role of ionic interactions between the catalyst and substrate, as void of ionic interactions through masking of the ionic group of the catalyst (16 vs 16-a) or the substrate (16 vs 16-b and 17 vs 17-a) led to diminished enantioselectivities. These results suggest that the intrinsic dissymmetry of the axle component as a result of the sulfonamide group is unlikely a determinant factor in the presence of anionic terminal capping groups, and that the ionic stereocontrol overrides its influence regardless of the orientation of the sulfonamide linkage (i.e., 16 and 17).


# 4

Atropo-enantioenriched Biaryls

Catalyst-controlled atropo-enantioselective Suzuki–Miyaura coupling has been broadly employed for the synthesis of chiral biaryls.[23] Its applicability has been demonstrated on 200-kilogram scale for the manufacturing of a BTK inhibitor.[24] Recently, we explored ionic substrate–catalyst interactions in atroposelective Suzuki–Miyaura couplings. Using carboxylate ligand L3, we accomplished the atropo-enantioselective synthesis of biaryls bearing phenols, trifluoroacetamides and carboxylic acids (Schemes 7A and 7B).[17] The nondirectional, long-range nature of ionic bonds has enabled us to forge chiral axes not only at the ortho positions, but also at the remote meta positions of the ionic group of the substrate.

Zoom Image
Scheme 7 Atroposelective Suzuki–Miyaura coupling

In the desymmetrizing strategy, most critical is the ability of the ionic chiral catalyst to confer stereocontrol over the oxidative addition step, which is likely selectivity-determining. In the case of atropo-enantioselective coupling, besides oxidative addition and transmetalation, the reductive elimination step is known to play an important role in the stereochemical outcome. We hypothesized that in both ortho stereocontrol and meta stereocontrol, the ionic ligand–substrate interactions persist during the reductive elimination step irrespective of the positioning of the ionic groups (Scheme [7]C). The consistency between the absolute stereochemistry of various types of coupling products supported this proposal.

Further investigation of the meta-ionic-interactions-directed atroposelective coupling revealed the 1,2-synergism and 1,4-antagonism between the meta ionic interactions and the inherent steric effect of ortho substituents (Scheme [7]D). First, the use of a neutral aryl halide resulted in 18 in essentially racemic form; the presence of meta-hydroxy substrates improved the stereochemical outcomes in both cases for 19 and 20. Analysis of the absolute stereochemistry of the products showed that the ionic interactions of the 1-phenol induced opposite stereoselectivity to the 4-methyl group at the ortho position of the new biaryl bond. Therefore, the meta ionic interactions override the steric preference by the 4-methyl group, even though the latter is closer to the site of catalytic bond formation in 19. On the other hand, the meta ionic interactions have a synergistic relationship with the ortho 2-methyl group, thus reinforcing the stereochemistry in 20. The results showed that the meta ionic stereocontrol is the dominant stereo-controlling factor, regardless of the relative positioning of the ortho substituents.

This strategy has enabled us to synthesize valuable ortho-phenols and trifluoroacetamides 21–24 (Scheme [7]E). For meta-stereocontrolled coupling, an array of chiral carboxylic acids (25–28) was obtained with good enantioselectivities. Importantly, due to the use of ionic interactions as a key element for enantioinduction, we have decoupled the reliance of using sterically well-differentiated ortho substituents to facilitate enantioselectivity. As such, sterically similar/or identical ortho-substituted substrates such as 21 and 27 can be accessed. Furthermore, this methodology is also applicable in bis-Suzuki–Miyaura coupling to forge double chiral axes; in these cases, the ionic groups do not interfere with each other nor do they compete with the inherent steric effects, and thus good enantioselectivities and diastereoselectivities can be obtained (e.g., 24 and 28).

Furthermore, we have utilized ionic catalyst–substrate interactions to serve a dual role in a reaction, i.e., mediating both stereoselectivity and site-selectivity (Scheme [8]).[18] Using 3,4-dichloroarenes bearing 1-ionizable groups, functionalized biaryl chlorides can be prepared readily with good atropo-enantioselectivity and with modest site-selectivity for coupling at the 3-position. The results illustrate the versatility of ionic-interactions-directed selectivity control, especially when the two types of selectivity are likely set at different elementary steps of the catalytic cycle.

Zoom Image
Scheme 8 Site-selective and atropo-enantioselective Suzuki–Miyaura couplings

# 5

Conclusions

We have been exploring long-range stereocontrol by engaging ionic substrate–ligand interactions at a distal position. To this end, we have developed desymmetrizing Suzuki–Miyaura coupling and atropo-enantioselective Suzuki–Miyaura coupling through engineering anionic axially chiral dialkyl biarylphosphine ligands. The strong ionic substrate–ligand interactions contribute to the exquisite preorganization of substrates and chiral catalysts, thus enabling the transmission of chirality across a long distance. In addition, the nondirectional nature of ionic interactions can be considered an advantage, particularly when targeting highly adaptative catalysts that can accommodate diverse structural changes of substrates.

The use of distal ionic interactions as an element for ligand design has opened up new opportunities in the realm of asymmetric catalysis. Our research has elicited interest,[23] [25] yet the application of ionic chiral catalysts for remote site-selectivity and stereoselectivity control is still in its infancy. Mechanistic insights such as (1) understanding of the intricate networks of ionic interactions involving multiple anions and cations within the catalyst–substrate preorganized complexes, particularly when hydration of ions is in play,[26] (2) decoding how the chiral information of the catalysts is transferred to the substrates despite the perceived lack of directionality of ionic interactions,[27] and (3) the orientated electronic effects of the nonligating charged ionic species on transition-metal catalysis,[28] will be invaluable for future development. We hope that this account will stimulate the pursuit of ionic chiral catalysts and unleash the potential of ionic stereocontrol.


#
#

Conflict of Interest

The authors declare no conflict of interest.


Corresponding Author

Ye Zhu
Department of Chemistry, Faculty of Science, National University of Singapore
3 Science Drive, 3, Singapore 117543
Singapore   

Publication History

Received: 16 May 2023

Accepted after revision: 19 June 2023

Accepted Manuscript online:
19 June 2023

Article published online:
04 August 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Ionic interactions in asymmetric transition-metal catalysis
Zoom Image
Scheme 2 A summary of bifunctional ionic chiral ligands over the years (not to scale)
Zoom Image
Scheme 3 A desymmetrizing Ullmann coupling by Miller et al.
Zoom Image
Scheme 4 Site-selective Suzuki–Miyaura coupling by Phipps et al.
Zoom Image
Scheme 5 A desymmetrizing Suzuki–Miyaura reaction to establish remote quaternary centers
Zoom Image
Scheme 6 Catalytic desymmetrization to give mechanically planar chiral rotaxanes
Zoom Image
Scheme 7 Atroposelective Suzuki–Miyaura coupling
Zoom Image
Scheme 8 Site-selective and atropo-enantioselective Suzuki–Miyaura couplings