Nickel-Based Catalysts for the Selective Monoarylation of Dichloropyridines: Ligand Effects and Mechanistic Insights

Abstract

This report describes a detailed study of Ni phosphine catalysts for the Suzuki-Miyaura coupling of dichloropyridines with halogen-containing (hetero)aryl boronic acids. With most phosphine ligands these transformations afford mixtures of mono- and diarylated cross-coupling products as well as competing oligomerization of the boronic acid. However, a ligand screen revealed that PPh₂Me and PPh₃ afford high yield and selectivity for monoarylation over diarylation as well as minimal competing oligomerization of the boronic acid. Several key observations were made regarding the selectivity of these reactions, including: (1) phosphine ligands that afford high selectivity for monoarylation fall within a narrow range of Tolman cone angles (between 136° and 157°); (2) more electron-rich trialkylphosphines afford predominantly diarylated products, while less-electron rich di- and triarylphosphines favor monoarylation; (3) diarylation proceeds via intramolecular oxidative addition; and (4) the solvent (MeCN) plays a crucial role in achieving high monoarylation selectivity. Experimental and DFT studies suggest that all these data can be explained based on the reactivity of a key intermediate: a Ni⁰-π complex of the monoarylated product. With larger, more electron-rich trialkylphosphine ligands, this π complex undergoes intramolecular oxidative addition faster than ligand substitution by the MeCN solvent, leading to selective diarylation. In contrast, with relatively small di- and triarylphosphine ligands, associative ligand substitution by MeCN is competitive with oxidative addition, resulting in selective formation of monoarylated products. The generality of this method is demonstrated with a variety of dichloropyridines and chloro-substituted aryl boronic acids. Furthermore, the optimal ligand (PPh₂Me) and solvent (MeCN) are leveraged to achieve the Ni-catalyzed monoarylation of a broader set of dichloroarene substrates.

Graphical Abstract

Introduction

Dihalo(hetero)arenes, particularly dihalopyridines, have attracted considerable attention as electrophiles in transition metal-catalyzed cross-coupling reactions.¹ The selective monoarylation of these substrates affords monohalobiaryls, a common motif in bioactive molecules (Figure 1).^2,3 Furthermore, the presence of a second electrophilic site on the product can be leveraged to access multiple functionalized (hetero)arenes, which also appear in drugs and agrochemicals (Figure 1).⁴

Figure 1.

Bioactive molecules accessible from cross-coupling reactions of 2,x-dihalopyridines (x = 3, 4, or 5)

Palladium-catalyzed cross-coupling reactions of 2,x-dihalopyridines (x = 3, 4, or 5) with aryl nucleophiles have been well-studied, and most Pd catalysts exhibit high selectivity for monoarylated product A (Scheme 1a).^5,6 In contrast, analogous reactions with nickel catalysts typically afford mixtures of A and the diarylated product B (Scheme 1a).⁷ The Ni-catalyzed reactions are further complicated when the aryl nucleophile contains a halide substituent, as this often leads to competing oligomerization/polymerization processes.⁸ Our overall objective was to identify a Ni catalyst system that promotes the selective C2-arylation of dichloropyridines with halide-substituted arylboronic acids.

Scheme 1.

Arylation of dichloropyridines: (a) Precedented reactivity with Pd and Ni; (b) Selective monoarylation with Ni

Herein, we report that selectivity in these transformations is highly sensitive to two factors: (1) the structure of the phosphine ligand and (2) the reaction solvent. The combination of diphenylmethylphosphine (PPh₂Me) as ligand and acetonitrile (MeCN) as solvent proved optimal for achieving selective monoarylation of a variety of dichloropyridines and chloro-substituted aryl boronic acids (Scheme 1b). A combination of experimental and computational mechanistic studies provide insight into the origin of selectivity in this system. Finally, we demonstrate the generality of these findings (specifically the optimal ligand and solvent choice) for achieving selective monoarylation to other dichloroarene substrates.

Results and Discussion

Our initial studies examined the Suzuki-Miyaura coupling of methyl-2,5-dichloropicolinate (1a) with (2-fluorophenyl)boronic acid to benchmark reactivity and selectivity. Ni(cod)₂ and PCy₃ were selected as the pre-catalyst and ligand, respectively, as this combination has been widely used for Ni-catalyzed Suzuki-Miyaura couplings, including those involving chloropyridine substrates.⁹ The reaction was examined in THF and MeCN, two solvents that are commonly used in related cross-coupling reactions.^{6a,6f,6g,9b–e,10} As shown in Scheme 2a, the reactions afford mixtures of monoarylated 2a, diarylated 3a, and homocoupled boronic acid (Ar-Ar). In both solvents,^9c diarylated 3a was the major product and was formed in comparable yield (28% versus 38%) and selectivity (3a : 2a = 3.1 : 1 versus 3.8 : 1). Overall, slightly higher reactivity was observed in MeCN (TON = 17 versus 13 in THF). As such, this was used as the solvent moving forward.

Scheme 2.

Product selectivity in Ni-catalyzed arylation of 1a with (a) simple aryl boronic acid and (b) chloro-containing aryl boronic acid^a,b

^aConditions: 1a (1.0 equiv, 0.1 mmol), ArB(OH)>₂ (1.1 equiv), Ni(cod)₂ (5 mol%), PCy₃ (20 mol%), K₂CO₃ (3.0 equiv), solvent (c = 0.2 M), 60 °C, 20 h. ^bYields determined by ¹⁹F NMR spectroscopy with trifluorotoluene as internal standard.

We next evaluated the Ni/PCy₃ catalyst in an analogous reaction with (4-chloro-2-fluoropheny)boronic acid, which contains a potentially reactive aryl chloride substituent. This resulted in a 2 : 1 mixture of diarylated 3j and the corresponding monoarylation product 2j in low yields (14% and 7%, respectively), with no homocoupling observed. Further analysis revealed that the boronic acid was completely consumed, and an insoluble solid was generated. These observations implicate the oligomerization of (4-chloro-2-fluorophenyl)boronic acid to form C.⁸

Having established the baseline reactivity of the Ni(cod)₂/PCy₃ catalyst, we sought to identify phosphine ligands that afford high selectivity and yield for the monoarylation product, while limiting the polymerization of chloro-containing aryl boronic acids. We hypothesized that the challenges of polymerization were related to that of the mono- versus diarylation selectivity (vide infra), and thus first focused on addressing the latter issue. A series of phosphine ligands with varied electronic (mono-, di- and trialkyl substituents) and steric (wide range of Tolman cone angles) properties were evaluated. These ligands were first tested in the Ni-catalyzed coupling of 1a with 2.2 equiv of (2-fluorophenyl)boronic acid to afford 2a and/or 3a. Table 1 shows the phosphines that afforded >4 : 1 selectivity for either 2a or 3a. Data for all of the ligands examined are available in Table S1.

Table 1.

Survey of monodentate phosphine ligands for the Ni-catalyzed arylation of 1a^a,b

entry	ligand	2a (%)	3a (%)	2a : 3a
1	PnBu3	2	23	1 : 12
2	PEt3	4	35	1 : 8.7
3	PtBu2Me	4	64	1 : 16
4	PiPr3	17	72	1 : 4.2
5	PPh3	62	8	7.7 : 1 0
6	PPh2Me	83	14	5.9 : 1 .
7	PPh2Et	55	5	11 : 1 .
8	PPh2iPr	39	3	13 : 1 .
9	PPh2Cy	33	2	17 : 1 .

^aReactions performed on 0.1 mmol scale (c = 0.2 M).

^bYields determined by ¹⁹F NMR spectroscopy with trifluorotoluene as internal standard.

This phosphine screen uncovered several ligands (most notably PPh₂Me, entry 6) that afford high yield and >5 : 1 selectivity for monoarylation, even when using 2.2 equiv of the boronic acid. Further optimization revealed that the use of 10 mol % of PPh₂Me in combination with the commercially available, air-stable Ni^II precatalyst (PPh₂Me)₂Ni(o-tolyl)Cl¹¹ resulted in 12 : 1 selectivity and 92% yield of 2a (Scheme 3). Finally, when the loading of boronic acid was reduced to 1.1 equiv, 2a and 3a were formed in 88% and 5%, respectively.¹²

Scheme 3.

Air-stable Ni^II precatalyst for the arylation of 1a

The observed ligand effects appear to originate from a combination of steric and electronic factors. To better understand these, we compared the experimental selectivities and yields to various descriptors of phosphine ligands.¹³ Steric effects in this transformation are most clearly visualized by plotting the combined yields of 2a (blue) and 3a (grey) as a function of the Tolman cone angle (Figure 2a).^14,15 This plot shows that phosphines that afford high selectivity for monoarylation are generally clustered within a relatively narrow range of cone angles (~136° to 157°). At slightly higher or lower cone angles (±10° or more), the diarylated product predominates. Moreover, the overall yield of 2a + 3a decreases sharply with very large phosphines such as P^tBu₃ and P(o-tolyl)₃ (cone angle ≥ 170°). The latter effect is even better visualized by plotting the overall yield versus the minimum percent buried volume (min %V_bur) of the phosphine, which shows a clear break at %V_bur ~ 32% (Figure 2b). This value represents the cut-off between mono- and bis-ligation in Ni⁰-phosphine oxidative addition transition states,^13a suggesting that the mono-ligated Ni⁰ species are inactive for this cross-coupling.

Figure 2.

Analysis of arylation of 1a using steric parameters (a) Tolman cone angle (θ ) and (b) minimum percent buried volume (%V_bur). (c) Product distribution as a function of phosphine substitution pattern (Ar = aryl). Reaction conditions from Table 1. Showing monoarylated product (blue), and diarylated product (gray).

To visualize electronic effects, we plotted the combined yields of 2a and 3a versus various electronic parameters (Tolman electronic parameter, TEP; semiempirical electronic parameter, SEP; pK_b; phosphine substitution; Figure S4). In most of these cases, the parameters are only available for a subset of the ligands, making it difficult to draw definitive conclusions. Clustering the phosphines based on their substitution pattern [trialkyl, (dialkyl)aryl, (diaryl)alkyl, and triaryl] revealed the clearest trend (Figure 2c). The more electron rich and stronger sigma-donor trialkylphosphines generally afford good to excellent selectivity for the diarylated product 3a. In contrast, less electron rich and better π-accepting di- and triarylphosphines afford monoarylated 2a as the major product. The highest overall yield of and selectivity for 2a is observed with PPh₂Me and PPh₃, ligands that have cone angles in the optimal range (136° and 145°, respectively) along with di- or triaryl substitution.

To further rationalize the observed ligand effects on selectivity we considered the pathways to products 2a and 3a in more detail (Scheme 4). The formation of 2a is expected to proceed via a standard Suzuki coupling mechanism, involving initial oxidative addition at the more activated C(2)–Cl bond of 1a, transmetalation of the boronic acid to form intermediate I, C–C bond-forming reductive elimination to generate π-complex 2a-Ni⁰,¹⁶ and dissociation of 2a to release the product. In contrast, 3a could be formed via two distinct pathways starting from 𝜋-complex 2a-Ni⁰. In the first (Scheme 4a, Path A), the organic intermediate 2a could dissociate from Ni⁰ and subsequently re-engage with the catalyst to undergo C(5)–Cl oxidative addition. Alternatively, oxidative addition into the C(5)–Cl bond could occur in an intramolecular fashion at π-complex 2a-Ni⁰, without the release of free 2a (Scheme 4a, Path B).^9c

Scheme 4.

Determining diarylation mechanism: (a) Proposed pathways; (b) Competition experiment; (c) Proposed mechanism

^aReactions performed on 0.1 mmol scale (c = 0.2 M). ^bYields determined by ¹⁹F NMR using trifluorotoluene as internal standard.

To distinguish these possibilities, we conducted a competition study using equimolar quantities of dichloropyridine 1a, aryl(chloro)pyridine 2a’, and (2-fluorophenyl)boronic acid (Scheme 4b). This experiment was performed using P^tBu₂Me, which afforded the highest selectivity for the diarylated product in Table 1. All of the possible intermediates/products (2a, 2a’, 3a and 3a’) have distinct ¹⁹F NMR signals, enabling quantitative analysis of the product mixture by ¹⁹F NMR spectroscopy. As summarized in Scheme 4b, the major product after 20 h was 3a, derived from selective diarylation of 1a. Only traces of the monoarylated product 2a were detected. Furthermore, <5% of 2a’ underwent C(5)-arylation to produce 3a’. When the reaction was analyzed at shorter time points (1 h or 3 h, respectively), the yield of 3a was lower, but only traces of 2a were observed (see Supporting Information for complete details). Collectively, these data suggest that free 2a is not an intermediate en route to 3a. Instead, diarylation of 1a appears to involve two sequential Suzuki-Miyaura cross-coupling reactions without dissociation of the Ni⁰–π intermediate, 2a-Ni⁰

This proposed pathway is analogous to the mechanism of living catalyst-transfer polymerization (CTP) reactions, where the intramolecular oxidative addition of a Ni⁰–π intermediate enables a chain-growth pathway.⁸ Indeed, the oligomerization of (4-chloro-2-fluoropheny)boronic acid observed in Scheme 2b with PCy₃ as the ligand likely occurs via this pathway. In CTP polymerization, the steric and electronic properties of ligands have been shown to play a pivotal role in productive reactivity by controlling the rate of ligand displacement at the π–complex.^8b,c We hypothesize that the ligand effects in our system are similarly related to the relative reactivity of 2a-Ni⁰ towards ligand substitution (for example, with the coordinating solvent acetonitrile) versus intramolecular oxidative addition (Scheme 5). In terms of steric effects, relatively small cone angle ligands are expected to facilitate associative substitution of solvent at 2a-Ni⁰ to release mono-arylated product 2a. Increasing the cone angle of the phosphine (and hence steric crowding at the Ni⁰ center) is expected to slow associative substitution relative to intramolecular oxidative addition, thus favoring diarylation product 3a. The phosphine substitution effects can also be rationalized based on the reactivity of 2a-Ni⁰. All other things being equal, more electron rich, sigma-donating trialkylphosphines are expected to increase the relative rate of oxidative addition¹⁷ compared to their di- or triaryl counterparts, thus pushing selectivity towards diarylation.

Scheme 5.

Proposed ligand exchange/solvent coordination as selectivity step for monoarylation of 1a

To experimentally probe the proposed role of acetonitrile in displacing the monoarylated product from 2a-Ni⁰, we evaluated the impact of solvent on product distribution with PPh₂Me and PPh₃ (Table 2).¹⁸ Under our standard conditions, both ligands afford high selectivity for the monoarylated product 2a, with ratios of 18 : 1 for PPh₂Me and >20 : 1 for PPh₃ (Table 2, entries 1 and 4). Changing the solvent from MeCN¹⁹ to THF led to a decrease in selectivity with PPh₂Me, shifting the ratio to 10 : 1 (Table 2, entry 2), but still favoring monoarylation. More dramatically, a reversal of selectivity was observed with PPh₃ in THF, leading to diarylated product 3a in 43% yield and 1 : 8.6 selectivity (Table 2, entry 5).

Table 2.

Solvent effect on product selectivity^a,b

entry	PR3	solvent	additive (equiv)	2a (%)	3a (%)	2a : 3a
1	PPh2Me	MeCN	---	88	5	18 : 1
2		THF	---	90	9	10 : 1
3		THF	MeCN (0.5)	85	5	17 : 1
4	PPh3	MeCN	---	53	1	53 : 1
5		THF	---	5	43	1 : 8.6
6		THF	MeCN (0.5)	51	21	2.4 : 1
7		THF	MeCN (1)	76	24	3.2 : 1
8		THF	MeCN (2)	80	11	7.2 : 1
9		THF	sulfolane (1)	4	46	1 : 11

^aReactions performed on 0.1 mmol scale (c = 0.2 M).

^bYields determined by ¹⁹F NMR using trifluorotoluene as internal standard. [Ni] = (PR₃)₂Ni(o-tolyl)Cl.

We next titrated various equivalents of MeCN into the THF reaction mixtures, with the hypothesis that this would serve to displace the monoarylated product from 2a-Ni⁰ and thus shift selectivity back towards 2a. Indeed, the addition of just 0.5 equiv of MeCN restored high (17 : 1) selectivity with PPh₂Me (Table 2, entry 3). With PPh₃, the selectivity responded in a dose-dependent manner, with increasing formation of 2a as more MeCN was added (Table 2, entries 6–8). Finally, to rule out effects related to solvent polarity changes, we examined the impact of highly polar yet weakly coordinating sulfolane as an additive.²⁰ The reaction with PPh₃ afforded 3a as the major product, with comparable yield and selectivity as in THF alone. This result indicates that the role of MeCN is not simply to increase the polarity of the reaction mixture. Instead, it implicates MeCN serving as a coordinating ligand to displace the monoarylated product from 2a-Ni⁰.

Density functional theory (DFT) calculations were conducted to probe the reactivity of the Ni⁰ π-complex,²¹ using PPh₃ as a representative ligand that favors monoarylation (Figure 3) and PⁱPr₃ as a representative ligand that favors diarylation (Figure 4). These symmetrical phosphines were selected to simplify the calculations by minimizing the number of accessible ligand conformers. The pi-bound substrate for the calculations was 2r, which contains a phenyl group at the 2-position. Thermodynamic quantities were calculated at 60 °C, applying corrections for concentrations consistent with the conditions in Table 1 (see SI for details).^22,23 In the PPh₃ system, intramolecular oxidative addition (3-TS-PPh₃) and associative displacement of Ni⁰ by MeCN (2-TS-PPh₃) are energetically comparable (ΔG^‡ = 17.5 and 16.7 kcal/mol) at the level of theory used. The difference in these activation barriers is 0.8 kcal/mol favoring diarylation, whereas the experimental selectivity with PPh₃ corresponds to ΔΔG^‡ ~1.4 kcal/mol favoring monoarylation. Overall, the computed ΔΔG^‡ differs from experiment by 2.2 kcal/mol, an error value within range of what is typical for DFT. Most importantly, the calculations implicate a ligand displacement pathway that is energetically competitive with oxidative addition. An alternative pathway for release of 2r involving direct dissociation to generate 2r and 14e^– Ni(PPh₃)₂ is much less likely due to the high free energy of the coordinatively unsaturated nickel(0) fragment (26.1 kcal/mol). Thus, overall the calculations are consistent with the involvement of MeCN in displacing Ni⁰ from the monoarylated product, rather than direct dissociation of 2r from Ni(PPh₃)₂.

Figure 3.

Computed reaction free energy diagram for pathways leading to mono- and diarylation of 1a using PPh₃.

Figure 4.

Computed reaction free energy diagram for pathways leading to mono- and diarylation of 1a using PⁱPr₃.

The DFT calculations with PⁱPr₃ as the ligand show very different results (Figure 4). Here, the barrier for oxidative addition at 2r-Ni⁰-PⁱPr₃ is significantly lower than that with PPh₃ (ΔG^‡ = 10.3 versus 16.7 kcal/mol, respectively), while the lowest energy pathway for ligand displacement has a significantly higher ΔG^‡ of 19.5 kcal/mol (compared to 17.5 kcal/mol for PPh₃). Furthermore, the calculations indicate that this ligand displacement takes place by a different mechanism than for the PPh₃ analogue, involving (i) initial dissociation of one PⁱPr₃ ligand, (ii) coordination of MeCN to form (PⁱPr₃)(MeCN)Ni⁰-2r, and (iii) reaction of this new π-complex with PⁱPr₃ to release 2r. In contrast, a transition structure involving associative substitution at the (PⁱPr₃)₂Ni⁰–π intermediate (analogous to 2-TS in Figure 3) could not be located. Unlike Ni(PPh₃)₂, direct dissociation of Ni(PⁱPr₃)₂ may be feasible based on the calculated thermodynamics of this step, but we were unable to find a transition state for this pathway. Overall, these results suggest that a combination of faster oxidative addition and slower release of 2r from Ni⁰ result in a preference for diarylation product 3r with PⁱPr₃.

We next evaluated the scope of this transformation with respect to the dichloropyridine (Table 3) and aryl boronic acid (Table 4) components. Using PPh₂Me as the ligand, a variety of 2,3- and 2,5-dichloropyridines underwent selective monoarylation at the 2-position, affording products 2a-h with high selectivity. Notably, 2,4-dichloropyridine was an exception, resulting in an intractable mixture of mono- and diarylated products (see Supporting Information for details about low performing substrates).

Table 3.

Selective Ni-catalyzed monoarylation of dichloropyridine derivatives with (2-fluorophenyl)boronic acid^a

^aReactions performed on a 0.3 mmol scale (c = 0.2 M). Combined ¹⁹F NMR yields of mono- (2) and diarylated (3) products reported with the respective ratio (2:3).

^bIsolation of 1.0 mmol scale reaction.

^c1.5 equiv of ArB(OH)₂ were used. Isolated yield of monoarylated product (2) in parenthesis. [Ni] = (PPh₂Me)₂Ni(o-tolyl)Cl.

Table 4.

Selective Ni-catalyzed monoarylation of 1a with various chloro-containing (aryl)boronic acids^a

^aIsolated yields of reactions performed on a 0.3 mmol scale (c = 0.2 M).

^b1.5 equiv of ArB(OH)₂ were used.

^cFrom N-Boc indole. [Ni] = (PPh₂Me)₂Ni(o-tolyl)Cl.

Boronic acids bearing chloride substituents at the meta and para positions also reacted in high yields, with no detectable oligomerization or over-arylation.²⁴ Again, the combination of PPh₂Me as a ligand with MeCN as a solvent is believed to accelerate displacement of the initial Ni⁰–π intermediate, thus limiting CTP-like polymerization of the boronic acids. For instance, monoarylated product 2j was obtained in 75% yield with our Ni/PPh₂Me catalyst system (Table 4), while the Ni/PCy₃ initial results afforded just 7%, largely due to competing oligomerization of the boronic acid (Scheme 2b). Other chloride-substituted (hetero)aryl boronic acids were also compatible (2i-2q), albeit affording slightly lower yields. Overall, a range of functionalities were well-tolerated, including ester, nitrile, amino, methoxy and fluoride.

The investigations above all focus on one specific class of electrophiles: 2,x-dichloropyridines. Thus, a key outstanding question is whether the pyridine moiety is crucial for obtaining high monoarylation selectivity (since the nitrogen of the substrate could potentially play a key role in displacing the π-complex). To address this question, we undertook a final set of studies to evaluate the generality of the observed ligand and solvent effects with other dichloroarenes.²⁵ We first examined the cross-coupling of (2-fluorophenyl)boronic acid with 1,3-dichlorobenzene (1s) using Ni(cod)₂ as the catalyst (Table 5). With PPh₂Me as ligand and MeCN as solvent, the monoarylated product 2s was obtained in 4 : 1 selectivity (Table 5, entry 1). Changing the solvent to THF under otherwise identical conditions afforded an erosion of selectivity to 2 : 1 (entry 2). Furthermore, moving to PCy₃ as a ligand and THF as the solvent led to 1 : 8 selectivity for the diarylated product (entry 4). This selectivity increased to 1 : 13 when the stoichiometry of the boronic acid was increased from 1.1 to 2.2 equiv (compare entries 4 and 6). Overall, the results with this substrate show nearly identical ligand and solvent effects to those with 2,x-dichloropyridines, indicating the generality of these results.

Table 5.

Exploring product selectivity in Ni-catalyzed arylation of 1,3-dichlorobenzene^a,b

entry	PR3	solvent	ArB(OH)2 (equiv)	2s (%)	3s (%)	2s : 3s
1	PPh2Me	MeCN	1.1	63	16	4 : 1
2		THF	1.1	50	24	2 : 1
3	PCy3	MeCN	1.1	6	36	1 : 6
4		THF	1.1	6	48	1 : 8
5		MeCN	2.2	9	63	1 : 7
6		THF	2.2	7	90	1 : 13

^aReactions performed on 0.1 mmol scale (c = 0.2 M).

^bYields determined by ¹⁹F NMR using trifluorotoluene as internal standard.

Finally, we scaled up the coupling of 1s and its regioisomers [1,2-dichlorobezene (1t) and 1,4-dichlorobenzene (1u)] under the optimal monoarylation conditions. As summarized in Table 6, the monoarylated products 2s-u were formed with good to excellent levels of selectivity and were isolated in 56%, 80%, and 39% yield, respectively.

Table 6.

Ni-catalyzed monoarylation of dichlorobenzene derivatives with (2-fluorophenyl)boronic acid^a

^aReactions performed on a 0.3 mmol scale (c = 0.2 M). Combined ¹⁹F NMR yields of mono- (2) and diarylated (3) products reported with the respective ratio (2:3). Isolated yield of monoarylated product (2) in parenthesis. [Ni] = (PPh₂Me)₂Ni(o-tolyl)Cl.

In summary, a Ni phosphine catalyst has been identified for the selective Suzuki-Miyaura coupling of dichloropyridine derivatives with halogen-containing (hetero)aryl boronic acids. A ligand screen showed that PPh₂Me is the optimal ligand for achieving high selectivity and yield for monoarylation with a wide range of dichloropyridine and boronic acid substrates. Several key observations were made regarding the selectivity of these reactions, including: (1) more electron-rich trialkylphosphines afford diarylation as the major product, while less-electron rich di- and triarylphosphines favor monoarylation; (2) the di- and triarylphosphines lie in a narrow range of Tolman cone angles (between 136° and 157°); (3) diarylation proceeds via intramolecular oxidative addition; and (4) the solvent (MeCN) plays a crucial role in achieving high monoarylation selectivity. Experimental and DFT studies suggest that these effects arise from the reactivity of a key intermediate: the Ni⁰-π complex of the mono-arylated product. With larger, more electron-rich trialkylphosphines, this π-complex undergoes intramolecular oxidative addition faster than ligand substitution by solvent, leading to selective diarylation. In contrast, with relatively small di- and triarylphosphine ligands, associative ligand substitution by MeCN is competitive with oxidative addition, resulting in selective formation of the monoarylated product. Finally, we demonstrated that these ligand and solvent effects can be applied to other dihaloarenes, affording selective monoarylation of dichlorobenzene derivatives. Overall, these studies demonstrate how the interplay of substrate, phosphine ligand, and solvent can impact selectivity in the Ni-catalyzed Suzuki-Miyaura couplings of polyfunctional substrates. Moving forward, we anticipate that these results will inform the development and optimization of other Ni-catalyzed C–C coupling reactions to form both small molecules and oligomers/polymers.

ABBREVIATIONS

SMCC

Suzuki-Miyaura cross coupling

TON

turn over number

OA

oxidative addition

TM

transmetallation

RE

reductive elimination

MeCN

acetonitrile

CTP

catalyst-transfer polymerization