Synthesis of 2-arylpyridines by the Suzuki–Miyaura cross-coupling of PyFluor with hetero(aryl) boronic acids and esters

Abstract

The Suzuki–Miyaura cross-coupling of pyridine-2-sulfonyl fluoride (PyFluor) with hetero(aryl) boronic acids and pinacol boronic esters is reported. The reactions can be performed using Pd(dppf)Cl₂ as the catalyst, at temperatures between 65 and 100 °C and in the presence of water and oxygen. This transformation generates 2-arylpyridines in modest to good yields (5%–89%).

Introduction

Suzuki–Miyaura cross-coupling (SMC) has been extensively applied in the formation of C(sp²)–C(sp²) bonds,^1–5 which are present in a variety of pharmaceuticals,^6–8 agrochemicals,⁹ and new organic materials.¹⁰ During the last two decades, the exploration of new electrophilic partners for this transformation has drawn considerable interest. Carboxylic acid derivatives,^11–13 ethers,¹⁴ sulfones,^15,16 and thioesters¹⁷ are some of the functional groups that have been successfully activated and coupled with boronic acids. Sulfonyl chlorides are highly reactive and have also been shown to be competent reaction partners in the SMC,^18,19 but until recently, this reactivity had not been extended to sulfonyl fluorides. Although the S–F bond in sulfonyl fluorides has been activated using sulfur (VI) fluoride exchange (SuFEx) chemistry^20–23 for the construction of sulfonamides,^24–26 sulfonate esters,^27,28 and sulfones,^29–31 among other functional groups, there was no experimental evidence of a transition metal-catalyzed desulfonative transformation of sulfonyl fluorides. This changed in 2018, when Grygorenko et al. reported the first example of C–S bond activation in the cross-coupling of halide-substituted pyridyl–sulfonyl fluoride with (Scheme 1A).³² Based on those results, Moran et al. explored the boronic acid scope of this transformation, showing that strongly sigma donating and bulky phosphine ligand RuPhos (2-dicyclohexylphosphino-2′,6′-diisopropoxybiphenyl) was optimal for this methodology (Scheme 1B).³³ Additionally, it was reported that the transformation did not require the addition of base. Later, Ding et al. reported that in the presence of potassium phosphate, the scope of aromatic sulfonyl fluorides could be extended to electron-withdrawing and electron-donating substituted benzene groups as well as the pyridyl group. Furthermore, two aryl boronic esters were shown to be competent nucleophilic counterparts (Scheme 1C).³⁴ Due to our group(s) interest in transition metal-mediated activation of fluorinated molecules,^35–38 as well as the general reactivity of sulfonyl fluorides,^{21,24,25,27,30,39} we wanted to further explore the scope and optimized conditions for this transformation. We were especially interested in evaluating the scope of heteroarylboronic acids, to see whether this methodology could be used to generate heteroatom-rich aromatic molecules, which are relevant motifs in drug discovery.^40–43 Here, we report an extension of the use of Pd(dppf)₂Cl₂ as a catalyst for the SMC of sulfonyl fluorides and a variety of hetero(aryl) boronic acids and pinacol boronic esters under milder conditions. The results range from modest-to-good yields (5%–89%) and are highly dependent on the chemical structure of the boronic ester.

Scheme 1.

Precedent for the Suzuki–Miyaura cross-coupling of sulfonyl fluorides with boronic acids/esters.

Results and discussion

We started to explore this transformation by searching for the optimal catalyst. Grygorenko et al. reported the use of either Pd(PPh₃)₄ or Pd(dppf)Cl₂ on the first reported examples of the SMC of halogenated pyridyl–sulfonyl fluorides.³² More recently, Moran and Ding explored the chemical space of monodentate phosphines in their respective optimizations,^33,34 which led us to test bidentate phosphine ligands (Scheme 2A). The use of PdCl₂, without additional ligands, did not generate any desired product and formed substantial amounts of palladium black. A control experiment using dppf without palladium did not generate the product either. Changing the bite angle (dppe, dppp, dppb, and dppf) did not correlate clearly with the yield observed. Then, we decided to explore the substitution in ferrocene phosphine ligands (Scheme 2B). Either decreasing (ⁱPr) or increasing (Cy, ^tBu) the steric hindrance on the ligand led to a dramatic drop in the yield of the transformation. The use of Pd(PPh₃)₄ and Pd(PPh₃)₄Cl₂ (Scheme 2C) resulted in a moderate yield, while nickel complexes were not effective precatalysts, as reported by Moran et al. (Scheme 2C). This initial catalyst screening resulted in the selection of Pd(dppf)Cl₂ as the catalyst for this study.

Scheme 2.

Catalyst optimization. (A) Exploration of the effect of phosphine bite angle. (B) Varying substituents on ferrocene phosphine ligands. (C) Using some common palladium and nickel catalysts. Reaction conditions: PyFluor (0.3 mmol), 2-thiopheneboronic acid pinacol ester (0.45 mmol), catalyst (0.03 mmol), ligand (0.03 mmol), Na₃PO₄ (0.9 mmol), and biphenyl (internal standard, 0.06 mmol); solvent: for the reactions in (A) dioxane 0.8 mL and water 0.2 mL. (B and C) Yields on parenthesis correspond to reactions performed on a mixture of 0.8 mL of dioxane and 0.2 mL of water. Yields outside parentheses correspond to reactions performed in dry and degassed dioxane. The reaction was performed in a capped 1 dram vial. Yield calculated using biphenyl as internal standard (HPLC/UV).

Next, we explored the solvent effects. A selection of the solvents tested can be found in Table 1, while a complete summary is available in the Supplementary material (Table S1). Toluene was found to be a less efficacious solvent for this reaction (entry 1), but interestingly adding water significantly increased the yield (entry 2). Dioxane fared better (entry 3), and again the addition of water was beneficial (entry 4). This led us to explore the optimal dioxane/water ratio (entries 11–18). We found the yield rose to 72% with 20% water, but further increase in the water content was detrimental. HPLC/UV analysis of the reaction mixtures for entries 16–18 showed that PyFluor hydrolyzes to the corresponding sulfonic acid sodium salt under the experimental conditions. Control experiments showed that this hydrolysis does not require a catalyst but requires the presence of base (see Table S1). The sulfonic acid salt was tested as a substrate instead of PyFluor, and the desired product was not detected. Consequently, we rationalize the decreased yield at higher water contents because of sacrificial substrate hydrolysis. Diglyme (entry 6) was comparable to dioxane, but DMF was found to be less effective. The use of acetonitrile and isopropanol (entries 7 and 8) led to acceptable yields, but overall, 20% water in dioxane was found to be the optimal solvent mixture for the reaction. The effect of bubbling N₂ through the reaction mixture versus performing the reaction under air was tested in entries 12–14. When the reaction was performed in a 0.5 dram vial, no significant difference was observed in the presence or absence of N₂ (entries 12 and 13). However, when a 3 dram vial was used, the reaction yield dropped 20% under an air atmosphere (entry 13), while 77% was obtained when N₂ was bubbled through the reaction mixture before heating (entry 14). These results show that this reaction can tolerate oxygen and water up to a certain proportion after which both become detrimental.

Table 1.

Solvent optimization for the synthesis of 1.

Entry	Solvent	Temperature (°C)	Yield (%)
1	Toluene	100	5
2	Toluene/H2O 4:1	100	40
3	Dioxane	100	39
4	Dioxane/H2O 4:1	100	74
5	DMF	100	22
6	Diglyme	100	42
7	MeCN	65	42
8	iPrOH	65	42
9	EtOH	65	11
10	THF	65	27
11	Dioxane/H2O 9:1	100	53
12a	Dioxane/H2O 4:1	100	72
13a,b	Dioxane/H2O 4:1	100	72
14c	Dioxane/H2O 4:1	100	52
15b,c	Dioxane/H2O 4:1	100	77
16	Dioxane/H2O 1:1	100	71
17	Dioxane/H2O 1:4	100	10
18	Dioxane/H2O 1:9	100	14

Note: Reaction conditions: PyFluor (0.3 mmol), 2-thiopheneboronic acid pinacol ester (0.45 mmol), Pd(dppf)Cl₂ (0.03 mmol), Na₃PO₄ (0.9 mmol), biphenyl (internal standard, 0.06 mmol), and solvent 1 mL in a closed 1 dram vial. Yield calculated using biphenyl as internal standard (HPLC/UV).

^aReaction performed in a 0.5 dram vial.

^bN₂ was bubbled through the reaction mixture during 15 min before heating.

^cReaction performed in a 3 dram vial.

To understand the role of water in this transformation, we tested the reactions of PyFluor with both 2-thiopheneboronic acid pinacol ester and 2-thiopheneboronic acid in dry and degassed dioxane, untreated dioxane, and dioxane with 20% water at different temperatures (Fig. 1). As seen by comparing the results of graphs (a) and (b), there is a bigger difference between the dry and 20% water conditions for the boronic ester than for the boronic acid. This result, in addition to a control reaction, which shows that 60% of the boronic ester hydrolyzes to the boronic acid in 20% water in dioxane at 100 °C (Figs. S1 and S2), suggests that water may be increasing the reaction efficiency by hydrolyzing the boronic ester in situ. As shown in the bottom graph, the boronic acids seems to be more reactive, leading to effective couplings at lower temperatures. The increased reactivities of boronic acids compared to pinacol boronic esters have been reported before.⁴⁴ An additional effect observed with water is the increasing solubility of Na₃PO₄ in the reaction mixture. Water can also alter the speciation of boronic acids and esters during the reaction, as well as influence the mechanism of transmetalation by enabling the formation of Pd–OH intermediates in the presence of a base.^44–47 The strong hydrogen bonding of the fluoride-leaving group to water⁴⁸ may also increase the reactivity of the sulfonyl fluoride, as proposed by Sharpless et al. in the context of SuFEx chemistry.²⁰

Fig. 1.

(a) Effect of temperature and water content in the yield of the reaction of PyFluor and 2-thiopheneboronic acid pinacol ester or (b) 2-thiophene boronic acid. Reaction conditions: PyFluor (0.3 mmol), 2-thiopheneboronic acid (or pinacol ester) (0.45 mmol), Pd(dppf)Cl₂ (0.03 mmol), Na₃PO₄ (0.9 mmol), and biphenyl (internal standard, 0.06 mmol). Solvent mixture (1mL in a closed 1 dram vial). “dry” dioxane was degassed by three cycles of freeze–pump–thaw and dried under molecular sieves (10% w/v) for 48 h before using it. “wet” dioxane was used as purchased. “20% water” refers to a solvent mixture of dioxane/water 4:1. Yield was calculated using biphenyl as internal standard (HPLC/UV). Yields are the average of two runs.

Additional experiments were carried out to test different bases in the reaction (Table 2). With a few exceptions (entries 3, 6, and 7), the yields were higher in the presence of excess water. We did not observe a clear trend in terms of the cation effect on the bases tested. For example, CsF (entry 7) was overall better than NaF (entry 4) and KF (entry 5), but Cs₂CO₃ (entry 15) was less effective than Na₂CO₃ (entry 10) and K₂CO₃ (entry 11). Another example is Li₂CO₃ (entry 9), which fared better than Na₂CO₃ and K₂CO₃ in 20% water, while Li₃PO₄ (entry 1) was substantially worse than Na₃PO₄ (entry 2) under the same conditions. Differential solubility of the salts in the dioxane/water mixture may influence yield, but solubility determinations are beyond the scope of this work. It is worth mentioning that the solubility of inorganic salts in organic solvents and their corresponding mixtures with water is an active field of study.^49,50 NaOH (entry 13) and KOH (entry 14) led to extensive PyFluor hydrolysis. Organic bases were less effective than inorganic ones (entries 15–17). Overall, Na₃PO₄ was the best base in the presence of 20% water.

Table 2.

Base optimization for the synthesis of 1.

Entry	Base	Yield (%)-drya	Yield (%)-not dryb	Yield (%)-dioxane/water 4:1
1	Li3PO4	0	4	19
2	Na3PO4	39	41	75
3	K3PO4	51	17	11
4	NaF	14	4	37
5	KF	51	24	54
6	KHF2	76	25	26
7	CsF	69	8	58
8	NBu4F	ND	12	14
9	Li2CO3	0	4	54
10	Na2CO3	33	7	35
11	K2CO3	26	19	25
12	Cs2CO3	15	9	10
13	NaOH	NP	36	5
14	KOH	NP	7	18
15	Et3N	4	9	9
16	Pyridine	0	4	10
17	DBU	NP	6	8
18	None	-	-	10

Note: Reaction conditions: PyFluor (0.3 mmol), 2-thiopheneboronic acid pinacol ester (0.45 mmol), Pd(dppf)Cl₂ (0.03 mmol), base (0.9 mmol), biphenyl (internal standard, 0.06 mmol), and solvent mixture (1 mL in a closed 1 dram vial).

^aDioxane was degassed by three cycles of freeze–pump–thaw and dried under molecular sieves (10% v/w) for 48 h before using it.

^bSolvent was used as purchased. Yield calculated using biphenyl as internal standard (HPLC/UV). NP = not performed.

The effect of using Lewis acids or bases as additives was then studied. We hypothesized that a Lewis acid could further polarize the S–F bond, facilitating the oxidative addition of PyFluor.⁵¹ Entries 1–3 in Table 3 show that the presence of Lewis acidic additives was detrimental, specially in dry dioxane. We then moved to test DBU (entry 4), as it has been applied as a nucleophilic additive in the S–F bond activation in SuFEx reactions.^21,52 Even sub stoichiometric amounts of this base were detrimental to the reaction. The addition of halide salts has been reported to be beneficial in some cross-couplings. It was proposed that the halides coordinated to the palladium center, forming an ate complex with high reactivity towards the oxidative addition of aryl halides.⁵³ Furthermore, TBAB (NBu₄Br) has been shown to stabilize palladium nanoparticles.^54,55 In our case, we observed a drastic yield reduction with one equivalent of either NBu₄Cl or NBu₄Br (entries 5 and 6).

Table 3.

Exploration of additives in the synthesis of X1.

Entry	Additive	Drya	Wetb	Dioxane/water 4:1
1	Ca(NTf2)2	4	24	60
2	B(OMe)3	8	31	54
3	LiBF4	-	-	65
4	DBU	7	—	71
5	NBu4Cl	0	—	22
6	NBu4Br	0	—	30

Note: Reaction conditions: PyFluor (0.3 mmol), 2-thiopheneboronic acid pinacol ester (0.45 mmol), Pd(dppf)Cl₂ (0.03 mmol), Na₃PO₄ (0.9 mmol), additive (0.3 mmol), biphenyl (internal standard, 0.06 mmol), and solvent mixture (1 mL in a closed 1 dram vial).

^aDioxane was degassed by three cycles of freeze–pump–thaw and dried under molecular sieves (10% v/w) for 48 h before using it.

^bSolvent was used as purchased. Yield calculated using biphenyl as internal standard (HPLC/UV).

With the optimized conditions at hand, we explored the scope of aryl boronic acid pinacol esters (Scheme 3). The presence of a sulfur (1) or oxygen (4) α to the boronic ester functionality was found to be beneficial. When the heteroatom was in the β position, the yield dropped substantially (2 and 5). Moving from thiophene to thiazole (3) essentially shut down reactivity, potentially because of catalyst poisoning. Modest yields were obtained using 3- and 4-pyridyl boronic esters (6 and 7). A methylpyrazole moiety was not tolerated (8), while a 5-pyrimidyl boronic ester successfully reacted to obtain modest yields (9). 4-cyano (10) and 4-methoxy phenyl boronic acids (11) also led to modest yields.

Scheme 3.

Substrate exploration using hetero(aryl) boronic esters. Reaction conditions: PyFluor (0.3 mmol), boronic acid pinacol ester (0.45 mmol), Pd(dppf)Cl₂ (0.03 mmol), Na₃PO₄ (0.9 mmol), dioxane 0.8 mL, and H₂O 0.2 mL. The reactionwas performed in capped 1 dram vial. Yields were calculated by ¹H NMR using 1,3,5-trimethoxybenzene as an internal standard.

We then tested the transformation using boronic acids (Scheme 4). Although Fig. 1 shows that the best yield was obtained at 100 °C in dry dioxane, we were interested in evaluating whether our system was active enough to drop the temperatures reported by Moran (130 °C) and Ding (120 °C) to 65 °C.^33,34 Although there was a 31% drop in yield, (1) could still be obtained at 51% yield at 65 °C. Analogously as shown in Scheme 3, 2-furanboronic acid led to a good yield of product (4) compared to the rest of the substrates. In the case of (10), the product was obtained in a 10% yield at 65 °C. Increasing the temperature to 100 °C led to an 11% increase in yield. Adding three equivalents of the boronic acid did not improve the outcome, but the yield rose to 35% when 20% water was added at 100 °C. The yield of (11) was significantly higher than (10). In this case, rising the temperature led to a negligible increase in yield, while the addition of three equivalents of boronic acid at 100 °C led to a 66% yield. Adding 20% water at 100 °C was detrimental, leading to significant PyFluor hydrolysis. The difference between the yields of (10) and (11) in dry dioxane and 20% water at 100 °C showed that the effect of water in the reactions with boronic acids may be beneficial for some substrates, while unfavorable for others. The transformation tolerated a tertiary amine (12) and an unprotected hydroxy group (13), although resulting in a low yield. Halogens were also tolerated, generating products (16) and (17) in a moderate yield. Except for (13), boronic acids with electron-donating groups or slightly electron-withdrawing groups led to better yields than more electron poor boronic acids (18–21). In these cases, there was a substantial amount of unreacted PyFluor left after the reaction. Increasing the steric hindrance in the boronic acid (22) did not reduce the yield compared to (15). Finally, 4-methylbenzenesulfonyl fluoride, 4-chlorobenzenesulfonyl fluoride, and 4-cyanobenzenesulfonyl fluoride were tested as coupling partners for 2-thiopheneboronic acid under the conditions described in Scheme 4. No products could be detected by GC/MS or ¹H NMR spectroscopy. In all cases the starting sulfonyl fluoride was the mayor component of the crude reaction mixture, reflecting the higher reactivity of PyFluor.

Scheme 4.

Substrate exploration using hetero(aryl) boronic acids. Reaction conditions: PyFluor (0.3mmol), boronic acid (0.45 mmol), Pd(dppf)Cl₂ (0.03 mmol), Na₃PO₄ (0.9 mmol), and dioxane 1.0 mL. The reaction was performed in capped 1 dram vial. Yields where calculated by ¹H NMR using 1,3,5-trimethoxybenzene as an internal standard. ∗Reaction performed at 100 °C. †Three equivalents of boronic acid at 100 °C. ‡Reaction performed at 100 °C with 20% water.

Conclusions

In summary, Pd(dppf)Cl₂ is a competent catalyst for the SMC of PyFluor and hetero(aryl) boronic acids and pinacol esters. The exploration of reaction conditions showed that the addition of 20% water was beneficial when pinacol esters of boronic acids were used as transmetalation partners, while it was substrate dependant in the case of boronic acids. Electron-rich boronic acids were favored over electron-poor ones. The presence of a sulfur or oxygen α to the boronic ester functionality seemed to be beneficial. This methodology tolerates water and oxygen to a certain extent.

Data availability

Data generated or analyzed during this study are provided in the supplementary material.