Suzuki‐Miyaura Cross‐Couplings: Juxtaposing the Transmetalation of Arylboronic Acids and Alkylboranes

Abstract

This report employs computational methods at unified levels of theory to systematically and comprehensively examine the respective transmetalation mechanisms underpinning classical palladium‐catalyzed C _sp ²–C _sp ² and C _sp ²–C _sp ³ Suzuki–Miyaura cross‐couplings. This approach enabled fundamental aspects of these transmetalation pathways to be directly juxtaposed, which allowed for subtle, but important, features underpinning the transmetalation mechanisms of archetypal arylboronic acid and B‐alkyl‐9‐borabicyclo[3.3.1]nonane nucleophiles to be highlighted and placed in a broader context.

This report employs computational methods at unified levels of theory to examine the respective transmetalation mechanisms underpinning classical palladium‐catalyzed C _sp ²–C _sp ² and C _sp ²–C _sp ³ Suzuki–Miyaura cross‐couplings. This approach enabled fundamental aspects of these transmetalation pathways involving archetypal arylboronic acid and B‐alkyl‐9‐borabicyclo[3.3.1]nonane nucleophiles to be directly juxtaposed.

1. Introduction

The venerable Suzuki–Miyaura (SM) reaction remains an exquisitely versatile, powerful, and reliable tool for the efficient formation of carbon–carbon bonds (Figure 1a).^{[ 1} , ^{2 ]} Palladium‐catalyzed C _sp ²–C _sp ² cross‐couplings of alkenyl‐ and arylboron nucleophiles (III) were first reported by Suzuki and Miyaura in 1979 and 1981.^{[ 3 ]} In 1986, these researchers disclosed the first palladium‐catalyzed C _sp ²–C _sp ³ cross‐couplings of alkylboron nucleophiles.^{[ 4 ]} The first palladium‐catalyzed C _sp ³–C _sp ³ cross‐couplings were reported by Suzuki and Miyaura in 1992.^{[ 5 ]} B‐alkyl‐9‐borabicyclo[3.3.1]nonane (B‐alkyl‐9‐BBN) coupling partners (VI), which featured in both of these pioneering C _sp ³ coupling studies,^{[ 4} , ^{5 ]} remain the most common alkylboron nucleophiles deployed in this chemistry.^{[ 2 ]}

Figure 1.

(A) Proposed mechanistic pathways operating in palladium‐catalyzed SM reactions of arylboronic acids (top) and B‐alkyl‐9‐BBN coupling partners (bottom); X = Cl, Br, I; L = monodentate tertiary phosphine (PR₃). (B) Transmetalation processes examined computationally in this study.

The SM catalytic cycle, consistent with many of the other fundamental palladium‐catalyzed C–C cross‐couplings, involves three sequential elementary steps: oxidative addition, transmetalation, and reductive elimination.^{[ 6 ]} It is generally acknowledged that transmetalation represents the most complex elementary step in the SM reaction.^{[ 7 ]} Accordingly, the mechanism of transmetalation in palladium‐catalyzed C _sp ²–C _sp ² SM cross‐couplings has been the subject of considerable study, employing both experimental and computational methods.^{[ 8} , ⁹ , ^{10 ]} It is proposed that following oxidative addition, arylpalladium(II) halide IIa reacts with arylboronate IIIa to form transient pretransmetalation species IV via path A (Figure 1A).^{[ 7} , ^{8 ]} This is termed the “boronate pathway”. Alternatively, arylpalladium(II) halide IIa reacts with hydroxide to form arylpalladium(II) hydroxide IIb. The latter intermediate may then react with the free arylboronic acid (IIIb) to generate species IV via path B. This is known as the “oxo‐palladium pathway”. Paths A and B converge at pretransmetalation species IV, from which it is proposed that transmetalation occurs via μ₂ ‐hydroxo‐bridged, four‐membered transition state V.^{[ 7} , ^{8 ]}

Experimental studies have demonstrated that both arylpalladium(II) halides IIa and hydroxides IIb are readily accessible in the presence of bases in organic solvent/water mixtures.^{[ 8} , ^{11 ]} Catalytic turnover is possible via either paths A or B in C _sp ²–C _sp ² SM cross‐couplings, and cases have been made for one pathway predominating over the other in the literature.^{[ 7} , ⁸ , ^{10 ]} However, differentiating between these pathways is particularly challenging. This is because the relative kinetic favorability of paths A or B is influenced by the specific reaction conditions employed and associated nuances. These include the identity and loading of the inorganic base; the nature of the solvent medium, such as the level of homogeneity and miscibility of organic/aqueous solvent mixtures, in addition to the pH; salt concentrations and associated effects (e.g., the common‐ion effect); and the reaction temperature. Taken together, these factors affect the speciation, availability/concentrations, and interplay of fundamental intermediates involved in SM transmetalation.^{[ 7} , ⁸ , ^{10 ]} It is also possible that the relative kinetic favorability of paths A or B may change over the course of the reaction.^{[ 7a ]} The forgoing issues reinforce the intrinsic difficulty drawing general conclusions regarding the relative kinetic favorability of paths A or B and the inherent challenges employing computational methods to model SM transmetalation mechanisms.^{[ 7} , ¹⁰ , ¹¹ , ^{12 ]} Accordingly, in their examination of SM transmetalation via density functional theory (DFT), Harvey and Yaman shrewdly note the importance of not interpreting results “too quantitively”.^{[ 10d ]}

The mechanism of transmetalation in palladium‐catalyzed C _sp ²–C _sp ³ SM cross‐couplings of alkylboron nucleophiles has been the subject of only three reports.^{[ 11} , ¹² , ^{13 ]} We completed the first computational study of this process in 2024.^{[ 12 ]} The mechanism of transmetalation in these systems has many similarities with transmetalation employing arylboronates, as shown in Figure 1A. It is generally accepted that transmetalation takes place from pretransmetalation species VII via μ₂ ‐hydroxo‐bridged, four‐membered transition state VIII, involving retention of configuration with respect to the carbon atom.^{[ 11} , ¹² , ^{13 ]} The speciation of B‐alkyl‐9‐BBN (VI) is governed by its high Lewis acidity. Specifically, experimental studies examining C _sp ²–C _sp ³ SM cross‐couplings of these alkylboron nucleophiles suggest that in the presence of hydroxide under catalytic conditions in THF/water mixtures, alkylborane VIb principally exists as boronate VIa.^{[ 11} , ¹² , ^{14 ]} Consequently, it is proposed that catalytic turnover predominates via path A under these conditions.^{[ 11} , ^{12 ]}

In this work, we employ computational methods at unified levels of theory to systematically and comprehensively compare and contrast fundamental aspects of the transmetalation of arylpalladium(II) species with arylboronates and alkylboranes for the first time (Figure 1B). This was the primary motivation for this study. For example, it allowed us to juxtapose the energetics of the respective reactions of the organoboron coupling partner with hydroxide, which has mechanistic implications (vide infra). We strived to integrate key findings in the mechanistic understanding of SM transmetalation that have emerged over the past decade as part of this commentary. Furthermore, we aimed to draw attention to subtle, but important, aspects that underpin SM transmetalation mechanisms and place these findings in a broader context.

2. DFT Methods

In 2014, a team led by Maseras and Ujaque comprehensively investigated transmetalation in palladium‐catalyzed C _sp ²–C _sp ² SM cross‐couplings of arylboronic acids employing the M06 functional.^{[ 10b ]} Additional calculations using PBE0 and B3LYP functionals were also carried out;^{[ 10b ]} and employing M06‐L, herein. Similar trends in energy barriers were obtained in a detailed study by Harvey and Yaman utilizing B3LYP‐D3BJ for optimizations reported in 2019.^{[ 10d ]} These seminal studies, and notable earlier work,^{[ 10 ]} laid the foundations for our examination of the transmetalation step in palladium‐catalyzed C _sp ²–C _sp ³ SM cross‐couplings of alkylboron nucleophiles in THF using M06‐L.^{[ 12} , ^{15 ]} This functional was selected given its recent outstanding record modelling organopalladium^{[ 16 ]} and associated systems,^{[ 17 ]} in addition to the demonstrated capacity of M06‐L to account for dispersion and solvent effects. All calculations (including optimizations) were undertaken in tetrahydrofuran, with optimizations including the SDD basis set with effective core potential for Pd and the 6–31G(d) basis set for other atoms, and single point calculation utilising the def2‐TZVP basis set and incorporating D3 computation to more fully account for dispersion (see Supporting Information for further details). We performed extensive benchmarking studies of optimizations/single point calculations, e.g. B3LYP‐D3/M06‐L‐D3, M06‐L/B3LYP‐D3BJ, and M06 functionals as part of this work, which supported our selection of M06‐L in this manifold.^{[ 12 ]}

In this present report, coordinates of species in the aforementioned phenylboronic acid system reported by Maseras and Ujaque^{[ 10b ]} were reoptimized employing M06‐L. This allowed us to systematically compare and contrast fundamental aspects of the transmetalation of PPh₃‐containing arylpalladium(II) species with archetypal arylboronic acids and alkylboranes in the presence of hydroxide and water in THF employing identical computational methods for the first time (Figure 1B and Supporting Information).

3. Overview

Following oxidative addition to form arylpalladium(II) halide 1,^{[ 18 ]} reactions with either aryl‐ or alkylboron coupling partners may be divided into four common stages: (i) ligand substitution; (ii) organopalladium boronate formation; (iii) PPh₃ dissociation; and (iv) transmetalation (Figure 2). We examine each of these fundamental stages in more detail in the subsequent sections. A conscious decision was made to avoid quantitative comparisons of activation energy barriers from our analysis in most cases. This was informed by the established intrinsic difficulty drawing general conclusions regarding the relative kinetic favorability of paths A or B in C _sp ²–C _sp ² SM cross‐couplings,^{[ 7} , ^{10 ]} in addition to the inherent challenges employing computational methods to model SM transmetalation mechanisms as discussed in detail by Harvey and Yaman.^{[ 10d ]}

Figure 2.

Energy profiles illustrating key species computed in path A (black) and path B (blue) for the reaction of [Pd(Ph)(PPh₃)₂Br] (1) with (a) PhB(OH)₂ (3) in the presence of hydroxide^{[ 10b ]} and (b) (9‐BBN)(CH₂)₂(o‐F‐C₆H₄) (5) in the presence of hydroxide,^{[ 12 ]} to afford diorganopalladium(II) intermediates tm_Ar and tm_Alk , respectively. Structures of species containing water clusters are idealized from Gaussview representations.^{[ 10} , ^{12 ]} Energies ΔG, in units of kcal/mol, are as reported for structures in (b),^{[ 12 ]} and recomputed using identical DFT protocols from coordinates reported for structures in (a).^{[ 10b ]} ΔG^ǂ values in black are indicated for transition states, adjacent to reported ΔG^ǂ values for the phenylboronic acid system computed using different DFT protocols in green.^{[ 10b ]} Selected interatomic distances are provided in Å; w = H₂O; M06‐L‐D3 (def2‐TZVP//M06‐L (6–31G(d), SDD).

Key structures in the formation of diorganopalladium(II) post‐transmetalation intermediates tm_Ar and tm_Alk are provided for both aryl and alkyl transfer processes in Figure 2. Identical DFT protocols were used for aryl and alkyl transfer, which involved the application of models for water as a cluster (H₂O)₃, hydration of hydroxide as [HO(H₂O)₃]⁻, bromide as [Br(H₂O)₃]⁻,^{[ 10} , ^{19 ]} and involvement of (H₂O)₃ clusters for TS‐OH_Ar , TS‐Br_Ar , and TS‐OH_Alk .^{[ 10b ]} Essentially identical structures were obtained, noting that the reported structures involved different basis sets and gas‐phase computation rather than in tetrahydrofuran. For example, the key transition states TS‐OH_Ar and TS‐Br_Ar (Figure 2A) exhibit Pd···Br distances within ∼0.1 Å, Pd···O within 0.3 Å, and O···Pd···Br angles within 3°. For ease of presentation in Figure 2, w₃ represents (H₂O)₃ fragments for transition states TS‐OH_Ar , TS‐Br_Ar , and TS‐OH_Alk . GaussView diagrams for these transition states and TS‐L_Alk are included in the Supporting Information.

Figure 2 illustrates the four common steps in the mechanisms for coupling aryl and alkyl nucleophiles, where each step is addressed in separate sections.

4. Ligand Substitution

In both cases, path B features substitution of bromide in [Pd(Ph)(PPh₃)₂(Br)] (1) by hydroxide via transition states TS‐OH_Ar and TS‐OH_Alk to form [Pd(Ph)(PPh₃)₂(OH)] (2) (Figure 2).^{[ 10} , ^{12 ]}

For the arylboronate system, path A involves the reaction of the arylboronate with [Pd(Ph)(PPh₃)₂(Br)] (1) to release bromide and generate pretransmetalation intermediate pre‐tm(L₂)_Ar directly. This contrasts markedly with the B‐alkyl‐9‐BBN system in which [Pd(Ph)(PPh₃)₂(Br)] (1) reacts with the alkylboronate to form species 8_Alk via the release of PPh₃.^{[ 12 ]} It is somewhat unexpected that the departure of PPh₃ takes place in preference to bromide.^{[ 11} , ^{12 ]} However, this is primarily governed by steric effects that prevent the formation of a viable transition state for bromide release.^{[ 12 ]} The transition state for PPh₃ loss (TS‐L_Alk ) also reflects significant steric effects, which are manifested in a very long Pd···P distance (3.190 Å).^{[ 12 ]} Another important feature throughout the sequence leading to the formation species 8_Alk is the presence of stabilizing hydrogen bonding interactions between the boronate OH proton and the bromide ligand. A similar interaction was found to be present in TS‐Br_Ar for the phenylboronate system.

The stationary points ‘1 + [PhB(OH)₃.3 w]⁻’ and ‘1 + [w₃.HO‐BR₂R_F]⁻’ are the lowest energy points in Figures 2A,B, respectively, and are, thus, reference energies for ΔG^ǂ values. It is evident that alkylboronate formation (−22.0 kcal/mol) is much more exergonic than arylboronate formation (−15.2 kcal/mol), which accords with the greater Lewis acidity of B‐alkyl‐9‐BBN boranes relative to arylboronic acids. Specifically, the latter molecules feature π‐donation contributions via B─C and B─O bonding for the trigonal boron center, which are not available in the former species. The highly exergonic reaction of the B‐alkyl‐9‐BBN with hydroxide is consistent with experimental observations indicating that, in the presence of hydroxide, alkylboranes VIb principally exist as boronates VIa (Figure 1A).^{[ 11} , ^{12 ]}

5. Organopalladium Boronate Formation: Convergence of Paths A and B

In both cases, path B involves the direct reaction of [Pd(Ph)(PPh₃)₂(OH)] (2) with the free boronic acid/B‐alkyl‐9‐BBN to form pretransmetalation intermediates pre‐tm(L₂)_Ar and pre‐tm(L₂)_Alk , respectively (Figure 2).^{[ 10} , ^{12 ]} These pretransmetalation species correspond to respective models IV and VII shown in Figure 1.

For the arylboronate system, path A delivers arylpalladium(II) boronate pre‐tm(L₂)_Ar directly from TS‐Br_Ar via the loss of bromide. For the alkylboronate system, dissociation of bromide from boronate 8_Alk in path A is followed by isomerization leading to formation of pre‐tm(L₂)_Alk . While this process also involves the formal loss of the bromide ligand, this sequence is more complicated as PPh₃ dissociation must take place prior to bromide loss (as noted earlier).

6. PPh₃ Dissociation

Transmetalation from three‐coordinate monophosphine complexes is kinetically favored over direct transmetalation from pre‐tm(L₂)_Ar ^{[ 8c,d,g , 10b,d ]} and pre‐tm(L₂)_Alk ^{[ 12 ]} in each case (Figure 2). For the arylboronate system, dissociation via TS‐L_Ar is followed by isomerization to give pre‐tm(L)_Ar . This accords with experimental observations suggesting that phosphine dissociation is a kinetic requirement for transmetalation in C _sp ²–C _sp ² SM cross‐couplings.^{[ 8c,d ]} The presence of coordinatively unsaturated, electrophilic organopalladium(II) species are crucial to facilitating transmetalation in both cases.^{[ 8} , ^{12 ]} For the alkylboronate system, dissociation of PPh₃ is also facile as species 7_Alk is lower energy than preceding transition state TS‐isom_Alk , which provides access to pre‐tm(L)_Alk . It should be noted that species 7_Alk is accessible directly from intermediate 10_Alk , and this is in competition with the formation of pre‐tm(L₂)_Ar in the presence of free PPh₃, which is highly exergonic.^{[ 20 ]}

7. Transmetalation

In both cases, the cis‐diorganopalladium(II) configurations within TS‐tm_Ar and TS‐tm_Alk lead to more facile C‐transfer than the equivalent trans‐configurations (Figure 2).^{[ 8} , ¹⁰ , ^{12 ]} These respective transmetalation processes deliver post‐transmetalation intermediates tm_Ar and tm_Alk , from which the ultimate formation of cross‐coupled products 4 and 6 readily occurs. The results of experimental and theoretical KIE studies employing ¹³C NMR spectroscopy to examine reactions of arylboronates were found to be consistent with transmetalation operating via TS‐tm_Ar .^{[ 8g ]} Previous DFT studies of SM cross‐couplings of arylboronic acids have suggested that the increased steric bulk of supporting phosphine ligands raises the transmetalation barrier.^{[ 10h ]} It was observed that this barrier was lowered with increasing π‐acceptor capacity of the phosphine.^{[ 10h ]}

8. Summary and Conclusions

This study employed computational methods at unified levels of theory (M06‐L) to systematically and comprehensively examine the respective transmetalation mechanisms that underpin classical palladium‐catalyzed C _sp ²–C _sp ² and C _sp ²–C _sp ³ SM cross‐couplings (Figure 2). Following oxidative addition to form arylpalladium(II) halide 1, reactions with either aryl‐ or alkylboron coupling partners may be divided into four common stages: (i) ligand substitution; (ii) organopalladium boronate formation, at which point paths A and B converge; (iii) PPh₃ dissociation; and (iv) transmetalation. Key observations are summarized below.

8.1. The Influence of Lewis Acidity on Boron Speciation and Associated Reactivity

We found that alkylboronate formation (ΔG^ǂ = −22.0 kcal/mol) is much more exergonic than arylboronate formation (ΔG^ǂ = −15.2 kcal/mol), which is consistent with the greater Lewis acidity of the former species. The stationary points ‘1 + [PhB(OH)₃.3 w]−’ and ‘1 + [w₃.HO‐BR₂R_F]−’ are the lowest energy points in Figures 2A,B, respectively. The differences in the Lewis acidity of arylboronic acids and B‐alkyl‐9‐BBN coupling partners influence their predominant speciation under the reaction conditions. Experimental findings suggest that, while both free arylboronic acid IIIb and arylboronate IIIa should be readily available in the presence of inorganic bases in organic solvent/water mixtures,^{[ 7} , ^{8 ]} alkylboranes VIb are likely to principally exist as boronates VIa (Figure 1A).^{[ 7} , ¹¹ , ^{12 ]} Accordingly, catalytic turnover is most likely restricted to path A in palladium‐catalyzed C _sp ²–C _sp ³ SM reactions with B‐alkyl‐9‐BBN boranes under these conditions. In comparison, catalytic turnover via either paths A or B is likely to be viable in palladium‐catalyzed C _sp ²–C _sp ² SM cross‐couplings in the presence of base in organic solvent/water mixtures. Cases have been made for one transmetalation pathway predominating over the other in C _sp ²–C _sp ² couplings, and this topic remains the subject of debate.^{[ 7} , ⁸ , ^{10 ]} Harvey and Yaman provide erudite, balanced, and comprehensive consideration of these aspects and associated mechanistic nuances.^{[ 10d ]}

8.2. Oxo‐Palladium Transmetalation Pathway (Path B)

Transmetalation path B features substitution of bromide in [Pd(Ph)(PPh₃)₂(Br)] (1) by hydroxide to generate [Pd(Ph)(PPh₃)₂(OH)] (2).^{[ 18 ]} In both cases, the latter species react with the free boronic acid/B‐alkyl‐9‐BBN to form respective pretransmetalation intermediates pre‐tm(L₂)_Ar and pre‐tm(L₂)_Alk . Transmetalation from three‐coordinate monophosphine complexes pre‐tm(L)_Ar and pre‐tm(L)_Alk is a kinetic requirement in each case. This is consistent with experimental results.^{[ 8c,d,g ]}

8.3. Boronate Transmetalation Pathway (Path A)

For the arylboronate system, transmetalation path A involves the reaction of the arylboronate with [Pd(Ph)(PPh₃)₂(Br)] (1),^{[ 18 ]} which releases bromide to generate pretransmetalation intermediate pre‐tm(L₂)_Ar directly (Figure 2). This is in stark contrast to the B‐alkyl‐9‐BBN system. In this case, path A features the reaction of arylpalladium(II) bromide (1) with the alkylboronate to form species 8_Alk via the loss of PPh₃ to accommodate the bulkier coupling partner.^{[ 12 ]} This accords with the limited examples of SM cross‐couplings of secondary/tertiary B‐alkyl‐9‐BBN nucleophiles.^{[ 2 ]} To our knowledge, reported palladium‐catalyzed SM reactions involving such systems are restricted to B‐cyclopropyl‐, B‐isopropyl‐, and B‐tert‐butyl‐9‐BBN coupling partners.^{[ 21 ]} The presence of stabilizing hydrogen bonding interactions between the boronate OH proton and the bromide ligand in TS‐Br_Ar , TS‐L_alk , and 8_Alk is also notable.^{[ 12 ]}

Our examination of transmetalation processes operating in palladium‐catalyzed SM reactions identifies fundamental mechanistic aspects common to classical C _sp ²–C _sp ² and C _sp ²–C _sp ³ cross‐couplings, while also highlighting subtle factors that differentiate (and influence) transmetalation in these systems.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting Information

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.