Suzuki-Miyaura Cross-Coupling of Benzylic Bromides Under Microwave Conditions

Abstract

A procedure for benzylic Suzuki-Miyaura cross-coupling under microwave conditions has been developed. These conditions allowed for heterocyclic compounds to be coupled. Optimum conditions found were Pd(OAc)₂, JohnPhos as the catalyst and ligand, potassium carbonate as base, and DMF as the solvent. Using these conditions, a library of structurally diverse compounds was synthesized.

System x_c⁻ (Sx_c⁻) is a membrane transporter that functions as an obligate exchanger of L-glutamate (L-Glu) and cystine.¹ The high concentration of L-Glu within cells drives the export of L-Glu and the import of cystine in a 1:1 molar ratio.² Sx_c⁻ transporter has emerged as a very promising target for CNS disorders such as drug addiction,³ viral pathology,⁴ and brain tumor growth because of its role in oxidative protection⁵, blood brain barrier operation, neurotransmitter release, and synaptic organization.⁶ We recently reported a novel series of Sx_c⁻ inhibitors based on an isoxazole scaffold.⁷ Inhibition of Sx_c⁻ reduces the tumor cells ability to protect itself from oxidative stress.

To further explore structure-activity relationships of this series we sought to introduce an aromatic cycle at the benzylic C5 position of an isoxazole. One straightforward approach to introduce the aromatic ring at the benzylic C5 position of an isoxazole would be to use a Suzuki-Miyaura cross-coupling reaction. There are very few examples in the literature of Suzuki couplings at the benzylic position and no examples involving heteroaryl benzylic bromide. Benzylic carbonates have been coupled using [Pd]-DPPPent with K₂CO₃ and DMF.⁸ McLaughlin coupled benzylic phosphates with Pd(OAc)₂/PPh₃.⁹ Burns et al. coupled heteroarylboronic acids with benzylic carbons.¹⁰ Molander et al. developed protocol using potassium organotrifluoroborates as the nucleophilic boron.¹¹ Chowdhury et al.¹² and Ines et al.¹³ also used water in their protocols. Unfortunately, this route fails using previously described protocols because of the water sensitivity of ethyl 1,2-oxazole-3-carboxylate derivatives. Our previous attempts at this coupling have resulted in low yields in our series of compounds. We developed a protocol that is useful for the coupling of water sensitive compounds. Optimizing this coupling is necessary to create a library of Sx_c⁻ inhibitors. In this present paper, we report a method for the palladium- catalyzed benzylic arylation of isoxazole and benzoate derivatives with arylboronic acids.

The isoxazole was prepared using previously reported synthetic strategy.^14,15 In our initial attempt at the benzylic coupling, phenyl boronic acid was used, but the compound resulting from Suzuki coupling was only separable from the starting material by HPLC. In order to have a better monitoring of the reaction, 3-methoxybenzeneboronic acids and the isoxazole were used for the optimization of the Suzuki coupling under microwave conditions. Microwave was used instead of classical heating to shorten reaction time. Yields were comparable between conventional and microwave heating. Details on the experimental protocol used are mentioned in Table 1, and 2.

Table 1.

Condition screen for benzylic Suzuki-Miyaura Cross-Coupling Reactions under microwave conditions. ^a

Entry	Base	Solvent	Yield %
1	Cs2CO3	Dioxane	10
2b	Cs2CO3	Dioxane	10
3c	Cs2CO3	Dioxane	10
4	Cs2CO3	THF	10
5	Cs2CO3	DMF	15
6	NaHCO3	DMF	15
7	K2CO3	DMF	15
8d	K2CO3	DMF	15

^aReaction conditions: boronic acid (1.5 mmol), isoxazole bromide (1.0 mmol), base (3.0 mmol), Pd (PPh₃)₄ (5 mol %), solvent (2 mL), 20 minutes.

^bduration was 60 minutes.

^ctemperature 160°C.

^dAg₂O (2.5 mmol) was added.

Table 2.

Condition screen for benzylic Suzuki-Miyaura Cross-Coupling Reactions under microwave conditions. ^a

Entry	Catalyst System	Base	Temp (°C)	Yield %
9a	PdCl2(dppf)	K2CO3	140	15
10	PdCl2(dppf)	K2CO3	140	15
11	Pd(OAc)2, L1	K2CO3	140	25
12	Pd(OAc)2, L2	K2CO3	140	50.4d
13	Pd(OAc)2, L2	K2CO3	150	40
14	Pd(OAc)2, L2	K2CO3	130	44d
15	Pd(OAc)2, L2	K2CO3	80	10
16c	Pd(OAc)2, L2	K2CO3	80	15
17	Pd(OAc)2, L3	K2CO3	140	25
18	Pd(OAc)2, L2	Et3N	140	25
19	Pd(OAc)2, L3	Et3N	140	25

^aReaction conditions: boronic acid (1.5 mmol), bromide (1.0 mmol), base (3.0 mmol), catalyst (5 mol %), ligand (10 mol %), DMF (2 mL), 20 minutes.

^bAg₂O (2.5 mmol) was added.

^cduration was 60 minutes.

^dIsolated yield.

Conversion yields were determined by LCMS analysis. A low yield, approximately 10%, was achieved when coupling 3-methoxybenzeneboronic acid to the isoxazole using Pd(PPh₃)₄ as a catalyst, Cs₂CO₃ as a base, and with dioxane as a solvent.

To improve the yield of this reaction several modifications were explored. Increasing the time or the temperature did not improve the yield (Table 1, entry 1, 2, 3). We then studied the impact of solvent on the yield. Using DMF slightly increased the yield compared to dioxane and THF (Table 1, entry 5). Changes to the base had no effect on the yield. It has been proposed that silver oxide increases the rate of boron to palladium transfer step of the catalytic cycle.^16,17 Use of silver oxide (Table 1, entry 8) had no effect on the yield. We decided to use the pre-catalysts PdCl₂(dppf) because of its chelating biphosphine with a large “bite” angle¹⁸ even though β-hydride elimination cannot compete with reductive elimination in our case. The use of PdCl₂(dppf) had no effect on the yield. Several bulky, electron-rich Buchwald phosphine ligands were chosen. These ligands have been shown to accelerate the oxidative addition and reductive elimination processes due to their bulky electron-rich phosphines groups.^19,20 Use of these phosphine ligands and Pd(OAc)₂ as the catalyst resulted in the greatest increase of the yield (Table 2). The conditions that provided the greatest yield of 50% were L2 as shown in Table 2, entry 12.

Using Table 2, entry 12 protocol we coupled a small series of isoxazoles and benzoates to show the effects of electron donating and electron withdrawing groups, as well as the versatility this protocol (Table 3). The phenylboronic acid had the highest yield with 75% for the benzoate derivative and 69% for the isoxazole derivative. Electron withdrawing groups such as 4-trifluorobenzeneboronic acid had the lowest yields with 35% and 20%. As expected, better yields were obtained with electron donating groups like 4-methoxybenzene boronic acid. Although the 4-methoxybenzene boronic acid would be expected to have a much higher yield compared to the 3-methoxybenzeneboronic acid, yet the 4-methoxybenzeneboronic acid only had a slightly better yield then the 3-methoxybenzeneboronic acid. This can be attributed to the boronic acids being used as the 4-methoxybenzeneboronic acid was of much lower purity than the 3-methoxybenzeneboronic acid as determined by NMR analysis (data not shown). Comparable moderate to good yields were obtained in the benzyl bromide series.

Table 3.

Cross-Coupling of Benzylic Bromides with Arylboronic Acids. ^a

1 (50.4%)	2 (60.9%)
3 (68.9%)	4 (20.1%)
5 (63.0%)	6 (66.7%)
7 (74.6%)	8 (34.7%)

^aReaction conditions: boronic acid (1.5 mmol), bromide (1.0 mmol), K₂CO₃ (3.0 mmol), Pd(OAc)₂ (5 mol %), L2 (10 mol %), DMF (2 mL), 140°C, 20 minutes. Isolated yields are shown.

The current protocol is superior to two other routes to C5 aryl methyl isoxazoles we had previously investigated: (1) the analogous Stille coupling suffers from the limitation of handling the relatively unstable and toxic stannyl methylene isoxazoles²¹ and (2) nucleophilic addition of lithioalkyl isoxazoles to arene cyclopentadienyl iron cation complexes followed by oxidation²² proceeds in moderate overall yields²³.

To conclude, few examples of Suzuki couplings at a benzylic position and even less with heteroaromatic rings such as isoxazoles have been described in the literature. Our study provides a protocol for achieving benzylic Suzuki coupling with moderate to good yields. Our study also provides a method that is useful for the synthesis of water sensitive compounds using a catalyst system commercially available. Furthermore, our method is quick and versatile with these findings; the synthesis of new class of water sensitive isoxazoles was possible. Our study allowed us to obtain the desired compounds to explore the SARs in this series.

Figure 1.

Buchwald phosphine ligands