Synthesis of fluoroalkenes and fluoroenynes via cross-coupling reactions using novel multihalogenated vinyl ethers

Introduction

Fluoroalkenes are one of the important frameworks for a wide range of industrial fields. For example, they are used as a bioisostere of amide bonds in medicines and agrochemicals, and contribute to the synthesis of peptide medicines that are stable to enzymatic metabolism and possess high lipophilicity . In fact, several inhibitors of the β-site amyloyd β A4 precursor protein cleaving enzyme (BACE1), which is involved in the production of β-amyloid, and fluoroalkene analogs of dipeptidyl peptidase-4 inhibitors have previously been reported . These inhibitors possess higher drug efficacies than their parent compounds. Furthermore, fluoroalkenes can be utilized as feedstock for fluoropolymers. Teflon, which is a well-known fluoropolymer with excellent water-repellent and oleophobic properties, is synthesized by polymerizing a monomer called tetrafluoroethylene. As a consequence, convenient and diverse synthetic methods for fluoroalkenes have attracted considerably and become increasingly necessary in pharmaceutical and industrial fields.

Fluoroalkenes have been constructed in a variety of methods , and one of the methods is to make use of fluorine-containing building blocks. When using them as nucleophilic reagents , the reaction between anion species, such as fluorine-containing Horner–Wadsworth–Emmons reagents, and carbonyl compounds led to E-selective olefination (Scheme 1A) . On the other hand, some reactions with electrophilic fluorine-containing building blocks have been developed . Jubault and Poisson et al. reported SN2’ reactions of hydride or alcohols to electrophilic fluorine-containing alkenes gave the corresponding fluoroalkenes (Scheme 1B) . In recent years, many fluorine-containing coupling reagents have been developed. These reagents are easily being converted into multisubstituted fluoroalkenes through cross-coupling using palladium, nickel, copper, ruthenium, and manganese catalysts . Hosoya and Niwa et al. published the development of a dual-reactive fluorine-containing C2-unit, which was prepared from trifluoroethanol in two steps in 63% yield, allowed the convergent synthesis of fluoroalkenes (Scheme 1C) . We recently found multihalogenated vinyl ethers 1 could be obtained by the reaction of phenols with 2-bromo-2-chloro-1,1,1-trifluoroethane (halothane) in good yields (Scheme 1D) . Compound 1 has a unique structure possessing three types of halogen atoms, namely bromine, chlorine, and fluorine, and it would be expected to afford multisubstituted fluoroalkenes by installing various substituents to bromine or chlorine atoms as reported by Hosoya and Niwa et al. In this study, we investigated the synthesis of fluoroalkenes 2 or fluoroenynes 3 by Suzuki–Miyaura or Sonogashira cross-couplings with a key building block 1 (Scheme 1D).

Scheme 1: Synthesis of monofluoroalkenes using fluorine-containing building blocks.

Results and Discussion Optimization of the conditions of cross-coupling reactions

First, we optimized the conditions of the Suzuki–Miyaura cross-coupling in reference to the report by Yang et al. (Table 1) . Upon the treatment of multihalogenated vinyl ether 1a with phenylboronic acid 4a (1.3 equiv) and palladium diacetate (10 mol %) as a catalyst at 40 °C, Suzuki–Miyaura cross-coupling proceeded to produce fluoroalkene 2a in 50% yield (Table 1, entry 1). Increasing the amount of 4a to 2.0 equiv and decreasing the amount of palladium diacetate to 5 mol % improved the reaction yield (Table 1, entry 2). When the reaction mixture was heated to 60 °C or reflux conditions, 2a could be synthesized in 84% yield under reflux conditions (Table 1, entries 3 and 4). Next, we examined an effective catalyst for the cross-coupling. Reactions using palladium dichloride or bis(2,4-pentanedionato)palladium significantly reduced the yields of 2a (Table 1, entries 5 and 6, respectively). When an allylpalladium chloride dimer or bis(triphenylphosphine)palladium dichloride were used as catalyst, the reaction proceeded with the same yield as that in Table 1, entry 4 (entries 7 and 8). Utilizing palladium catalyst such as bis(triphenylphosphine)palladium dichloride, all these reactions could convert 1a into 2a in good yields (Table 1, entries 9–11). Cross-coupling with palladium bis(trifluoroacetate), which is more reactive than palladium diacetate, gave the corresponding product in high yield of 96% (Table 1, entry 12). Without the addition of triphenylphosphine, the reaction proceeded in only 12% yield (Table 1, entry 13). Thus, it was concluded that triphenylphosphine is necessary for Suzuki–Miyaura cross-coupling of 1 with 4 and that it is involved in the production of palladium(0).

Table 1: Optimization of reaction conditions for Suzuki–Miyaura cross-coupling using multihalogenated vinyl ether 1a.

Entry 4a (equiv) Pd cat. (mol %) Temp. (°C) Time (h) 2a (%)a 1 1.3 Pd(OAc)2 (10) 40 3.0 50 2 2.0 Pd(OAc)2 (5) 40 2.5 73 3 2.0 Pd(OAc)2 (5) 60 2.5 68 4 2.0 Pd(OAc)2 (5) reflux 3.5 84 5 2.0 PdCl2 (5) reflux 3.5 61 6 2.0 Pd(acac)2 (5) reflux 3.5 12 7 2.0 [Pd(allyl)Cl]2 (5) reflux 3.5 81 8 2.0 Pd(PPh3)2Cl2 (5) reflux 3.5 84 9 2.0 Pd[P(o-Tol)3]2Cl2 (5) reflux 3.5 89 10 2.0 Pd(MeCN)2Cl2 (5) reflux 3.5 93 11 2.0 Pd(PhCN)2Cl2 (5) reflux 3.5 92 12 2.0 Pd(OCOCF3)2 (5) reflux 3.5 96 13b 2.0 Pd(OCOCF3)2 (5) reflux 3.5 12

aIsolated yields; bno PPh3.

Next, the reaction conditions for the Sonogashira cross-coupling were optimized (Table 2). On the basis of a previous study by Thorand , we performed the reaction between fluorine-containing vinyl ether 1a and 1.05 equiv of trimethylsilylacetylene (5a) to afford the corresponding enyne 3a in 55% yield (Table 2, entry 1). Cross-coupling utilizing a palladium(II) catalysts containing phosphine ligands produced low yields of 3a (Table 2, entries 2 and 3). In the case of palladium(II), which produced good yields of the Suzuki–Miyaura cross-coupling products, only a small amount of 3a was obtained (Table 2, entries 4–8). In particular, when the allylpalladium dichloride dimer was used, Sonogashira coupling hardly proceeded at all, and the starting ether 1a was recovered in an 83% yield (Table 2, entry 6). Zero-valent tetrakis(triphenylphosphine)palladium and tris(dibenzylideneacetone)dipalladium allowed the reaction to undergo in 37% or 23% yields, respectively (Table 2, entries 9 and 10). In entry 11, Table 2, we selected bis(triphenylphosphine)palladium as an effective catalyst, but increase of 5a to 1.5 equiv did not improve the reaction yield. Diluting the reaction concentration from 0.83 M to 0.2 M achieved to give 3a in a 63% yield (Table 2, entry 12). Increasing the amount of palladium catalyst to 4 mol % led to the conversion of 1a into 3a in 77% yield (Table 2, entry 13). In addition, using 2.0 equiv of 5a gave 3a in high 80% yield (Table 2, entry 14).

Table 2: Optimization of reaction conditions for Sonogashira cross-coupling using multihalogenated vinyl ether 1a.

Entry 5a (equiv) Pd cat. (mol %) Time (h) 3a (%)a 1 1.05 Pd(PPh3)2Cl2 (2) 6.5 55 2 1.05 Pd[P(o-Tol)3]2Cl2 (2) 24 14 3 1.05 Pd(PCy3)2Cl2 (2) 24 0 4 1.05 Pd(OAc)2 (2) 20 6 5 1.05 PdCl2 (2) 24 13 6b 1.05 [Pd(allyl)Cl]2 (2) 18.5 0 7 1.05 Pd(MeCN)2Cl2 (2) 25.5 13 8 1.05 Pd(PhCN)2Cl2 (2) 24 10 9 1.05 Pd(PPh3)4 (2) 21.5 37 10 1.05 Pd2(dba)3 (2) 24 23 11 1.5 Pd(PPh3)2Cl2 (2) 19 53 12c 1.5 Pd(PPh3)2Cl2 (2) 17 63 13c 1.5 Pd(PPh3)2Cl2 (4) 18 77 14c 2.0 Pd(PPh3)2Cl2 (4) 19 80

aIsolated yields; bRecovery of 1a was 83% yield; cTHF (0.2 M) was used.

Based on these results, we determined entry 13 in Table 1 and entry 14 in Table 2 as the optimum reaction conditions. We used 1 as a mixture of diastereomers (diastereomer ratio = 1:1) for cross-coupling, and the corresponding compounds 2 and 3 were obtained as mixtures of diastereomers in a certain ratio as estimated by proton and fluorine NMR spectroscopy.

Substrate scope for cross-coupling reactions

The substrate scope was investigated using various boronic acids 4 and alkynes 5 in cross-coupling reactions using 1 (Table 3 and Table 4). p-Tolylboronic acid 4b provided 2b quantitatively, whereas m- and o-tolylboronic acids 4c and 4d produced 2c and 2d in low yields because the methyl group was positioned near the reaction site (Table 3, entries 1–3). Introduction of 3,4-methylenedioxyphenyl (4e) or p-fluorophenyl groups (4f) to 1a proceeded in high yields (Table 3, entries 4 and 5). Boronic acids with carbonyl groups such as acetyl, ester or formyl moieties in para position (4g–i) underwent the cross-coupling in 76, 96 or 77% yields (Table 3, entries 6–8). The reaction between 1a and 4j, which contains an electron-withdrawing nitro group, afforded 2j in 88% yield (Table 3, entry 9). Although p-hydroxyphenylboronic acid (4k) gave 2k in only 9% yield, m-aminophenylboronic acid (4l) provided 2l in high yield (Table 3, entries 10 and 11). We predicted that the product yield would decrease because 2k is labile in column chromatography. Utilizing a boronic acid bearing an n-butyl group as a primary alkyl group (4m), the cross-coupling did not proceed due to β-elimination (Table 3, entry 12). In contrast, the reaction with cyclopropylboronic acid (4n) achieved to give 2n in a 71% yield (Table 3, entry 13). When thiopheneboronic acid 4o was used as a coupling partner, the thiophene ring could be installed on 1a in a comparatively low yield of 31% (Table 3, entry 14). In addition, we investigated the substrate scope of 1 in the Suzuki–Miyaura cross-coupling. The reaction of 1b or 1c, which had a m-methoxy or p-nitro group on the benzene ring, with 4a proceeded smoothly to furnish 2p or 2q in good yieds (Table 3, entries 15 and 16). A phenyl group could be introduced into 1d possessing an ester moiety in moderate yield, whereas the cross-coupling between 1e, derived from m-aminophenol, and 4a proceeded in only 15% yield (Table 3, entries 17 and 18).