Practical tetrafluoroethylene fragment installation through a coupling reaction of (1,1,2,2-tetrafluorobut-3-en-1-yl)zinc bromide with various electrophiles

Introduction

Recently, much attention has been paid to organic compounds containing a perfluoroalkylene unit, e.g., –(CF₂)_n–, in various fields, such as medicine and materials sciences [1-4], because incorporation of multiple fluorine atoms into organic substances causes dramatic alterations in the chemical and physical properties of substances, which may significantly enhance their potential functionality [5,6]. Notably, organofluorine compounds bearing a tetrafluoroethylene (–CF₂CF₂–) unit have attracted significant interest as a promising framework for various functional molecules. In the medicinal field, for example, Linclau and co-workers reported the first enantioselective synthesis and the intriguing biological activities of CF₂CF₂-containing pyranose and furanose derivatives (Figure 1a) [7-9]. Subsequently, Gouverneur et al. also developed novel CF₂CF₂-containing C-nucleosides (Figure 1b) [10]. Meanwhile, in the field of materials sciences, Kirsch et al. revealed that the incorporation of the CF₂CF₂ unit between two cyclohexane rings caused significant enhancement in thermal stability in a liquid crystalline phase (Figure 1c) [11]. In addition to the development of a convenient access to CF₂CF₂-containing pyranoses [12], our group also showed that tricyclic mesogens with a CF₂CF₂-containing carbocycle exhibited large negative dielectric anisotropies (Figure 1d), which would be promising for candidates for vertically aligned (VA)-type display materials [13-17]. Owing to such valuable applications, the development of a much more efficient synthetic protocol for CF₂CF₂-containing organic molecules has a high priority in various research areas.

A number of synthetic protocols for CF₂CF₂-containing molecules have been developed in the last half-decade. Ogoshi and co-workers reported an efficient synthetic protocol for CF₂CF₂-containing compounds via carbocupration or oxycupration of tetrafluoroethylene (TFE) [18-20]. Gouverneur et al. disclosed a Cu(I)-mediated cross-coupling reaction using ArCF₂CF₂SiMe₃ derivatives that produced the corresponding tetrafluoroethylenated compounds [21]. Beier’s group also described a broad range of tetrafluoroethylene compounds generated by the reductive coupling of in situ-formed RCF₂CF₂MgCl·LiCl with various electrophiles [22,23]. Our group revealed that commercially available 4-bromo-3,3,4,4-tetrafluorobut-1-ene (1) underwent a Cu(0)-mediated cross-coupling reaction with aromatic iodides in DMSO at 160 °C in a sealed-tube apparatus [24] or reductive couplings with various carbonyl compounds [25], leading to good yields of versatile CF₂CF₂-containing substances. However, many obstacles to practical synthesis remain, e.g., the use of 1,2-dibromotetrafluoroethane (Halon-2402) with ozone depletion and global warming potentials as a starting material [26], explosive and difficult-to-handle gaseous tetrafluoroethylene [18-20], thermally unstable polyfluoroalkylmetal species [8,27,28], etc. Therefore, the development of a more efficient synthetic protocol featuring easy-handling, simple preparation, and thermal stability is highly desirable.

Our strategy focused on the preparation of a thermally stable CF₂CF₂-containing metal species for which the “unreactive” form can be easily changed to the “reactive” form through chemical transformation. Out of a variety of organometallics, we selected an organozinc reagent that possesses higher thermal stability than the corresponding organolithium or -magnesium species due to the almost covalent C–Zn bond [29,30]. Moreover, organozinc reagents can be easily transformed to the “reactive” species, through a transmetallation process with a transition metal (e.g., Pd or Cu), which can efficiently construct a new C–C bond [30-32]. Thus, we aimed at the development of an efficient preparation of a thermally stable organozinc reagent, possessing a CF₂CF₂ fragment as a thermally stable tetrafluoroethylenating agent, and successive C–C bond formation to produce a wide variety of CF₂CF₂-containing molecules, and the results are described in this article (Scheme 1).

Results and Discussion

We carried out the optimization of the reaction conditions for the preparation of 2-Zn by direct zinc insertion into the CF₂–Br bond of commercially available fluorinated substance 1. The results are summarized in Table 1.

Initially, we attempted zinc insertion using a typical protocol [33], i.e., treating 1 with 2.0 equiv of zinc powder, pre-activated with dilute HCl solution, in DMF at 0 °C for 0.5 h. Unfortunately, no desired zinc insertion occurred and substrate 1 was quantitatively recovered (Table 1, entry 1). Interestingly, when a zinc–silver couple Zn(Ag) [34] was employed instead of zinc powder, the desired zinc insertion took place very smoothly to form the desired (1,1,2,2-tetrafluorobut-3-en-1-yl)zinc bromide (2-Zn) in 86% NMR yield, along with a small amount of a reduction byproduct (2-H) (Table 1, entry 2). Reducing the amount of Zn(Ag) slightly retarded the zinc insertion, with only 50% NMR yield (Table 1, entry 3). Lowering the temperature to –30 °C also inhibited the zinc insertion, with quantitative recovery of 1 (Table 1, entry 4). The reaction was also attempted in various solvents, namely N,N-dimethylacetamide (DMA), N,N’-dimethylpropyleneurea (DMPU), N-methylpyrrolidin-2-one (NMP), and THF. In polar solvents (DMA, DMPU, and NMP), the reaction successfully produced the desired 2-Zn, although the yield was still unsatisfactory (22–58%, Table 1, entries 5–7). In the less polar THF, in contrast, complete recovery of 1 was observed (Table 1, entry 8). Nevertheless, after raising the temperature to 40 °C, directed zinc insertion into the CF₂–Br bond was achieved in THF, producing 2-Zn in 75% NMR yield (Table 1, entry 9). Eventually, the optimal conditions for the preparation of 2-Zn were determined to be that given in entry 2 (Table 1). It is noteworthy that 2-Zn can be prepared in DMF or THF on a large scale (approx. 40–50 mmol scale) without any problem. More remarkably, the prepared 2-Zn can be stored in DMF solution (ca. 0.70 M) for at least 1.5 years in the refrigerator [35].

In order to examine the stability of 2-Zn in more detail, we quantitatively evaluated the thermal stability of 2-Zn under various temperature conditions (Figure 2). After a given duration, the recovery yield of 2-Zn was determined by ¹⁹F NMR analysis using an internal reference.

Complete recovery of 2-Zn (0.70 M in DMF) was observed below 50 °C and gradual degradation occurred at 80 °C. It was revealed that 50% decomposition of 2-Zn was observed after 4 h stirring, and almost all 2-Zn was decomposed after 12 h stirring (96%). At 100 °C, 2-Zn was almost completely decomposed after being stirred for 2 h, with a recovery yield of only 8%. According to Beier’s report [22,23], tetrafluoroethylmagnesium species fully decompose within 50 min at −40 °C; CH₂=CHCF₂CF₂ZnBr (2-Zn) used in the present study was shown to be much more thermally stable than tetrafluoroethyllithium [12,25] and -magnesium species [22,23].

The synthetic uses of 2-Zn as a promising tetrafluorinating agent were tested in several reactions. First, we demonstrated a typical C–C bond formation reaction. Treating freshly prepared 2-Zn (ca. 0.70 M in DMF) with 5.0 equiv of iodobenzene (3a) in the presence of 10 mol % of CuI in DMF at 50 °C for 24 h resulted in limited formation (11% yield) of the cross-coupling product 4a (Table 2, entry 1). The Cu(I)-catalyzed cross-coupling reaction with 3a was proposed to take place via the following three key reaction steps [36]: (i) transmetallation from 2-Zn to the corresponding Cu(I) species, (ii) oxidative addition of a C_Ar–I bond to the Cu(I) atomic center to generate Cu(III) species, and (iii) reductive elimination of the product 4a along with the regeneration of the Cu(I) salt. When the initial transmetallation from 2-Zn to the reactive Cu(I) species proceeds much faster than the thermal decomposition of 2-Zn, the yield of 4a should be improved. Indeed, in the cross-coupling reaction carried out at 80 °C for 24 h, an enhanced yield of 4a was observed (Table 2, entry 2), while the reaction in THF did not give any trace of the product (Table 2, entry 3). Catalyst optimization (Table 2, entries 4–6) showed that Cu₂O led to the highest product yield (46%, Table 2, entry 6) [37]. After further exploring the reaction conditions (Table 2, entries 7–13) including catalyst loading, equivalents of 3a, additives, and reaction temperature, the best result (64% NMR yield) was observed with 6.0 equiv of 3a in the presence of 30 mol % of Cu₂O in DMF at 100 °C for 24 h (Table 2, entry 12) [38].

Using the optimal conditions for the reaction in entry 12, Table 2, various iodoarenes or -heteroarenes (3b–r) could be converted into the corresponding CF₂CF₂-substituted aromatic compounds 4b–r (Figure 3).

Aromatic iodides with an electron-donating group, such as OMe (3b) and Me (3c), at the para position on the benzene ring were successfully coupled with 2-Zn under the optimized conditions to form 4b and 4c in 43% and 36% isolated yields, respectively. p-Chloro- (3d) or p-trifluoromethyl-substituted iodobenzene (3e) could also participate in the coupling reaction, leading to the corresponding products 4d and 4e in 53% and 75% NMR (49% and 13% isolated) yields, respectively. Notably, iodoarenes with an ethoxycarbonyl (3f) or a nitro group (3g) at the para-position were successfully converted to the corresponding products 4f and 4g with good yields. Differences in the position of the substituents on the benzene ring had no effect on the coupling reaction; m-chloro- (3h) and o-chloroiodobenzene (3i) were transformed into their corresponding products 4h and 4i in 60% and 64% yields, respectively. Interestingly, the reaction using ethyl o-iodobenzoate (3j) quite efficiently proceeded to give the coupling product 4j almost quantitatively. Comparing the result from the reaction with the para-substituted analog 3f, the ester functionality at the ortho-position seems to significantly facilitate the formation of the coupling product. According to the previous reports by Jiang and co-workers [36], the oxygen atom in the ethoxycarbonyl group substituted at the ortho-position coordinates with the copper center to stabilize the Cu(III) intermediate, which facilitates the subsequent reductive elimination to form the desired coupling product. This acceleration effect of the ester functional group at the ortho-position led to a reduction in the chemical substances used. That is to say, the product 4j was obtained in a quantitative manner even when the reaction of 2-Zn was carried out with only a half equivalent of 3j. The same effect was also noted for iodoarenes having a labile functional group, such as formyl (3k), acetyl (3l), methoxymethyl (3m), or a nitro group (3n–q) at the ortho-position, giving rise to the corresponding products 4k–q with excellent efficiency. These results strongly suggest that 2-Zn can be successfully employed for the coupling reaction with an electrophile bearing a reactive functional group. Lastly, this synthetic protocol could be applied to prepare the CF₂CF₂-substituted heteroaromatic compound 4r from 3-iodopyridine (3r), demonstrating a promising pathway for constructing CF₂CF₂-containing heterocyclic compounds.

We also demonstrated the multigram preparation of CF₂CF₂-containing arenes through the present cross-coupling reaction, as shown in Scheme 2. Thus, treatment of 1.38 g (5.00 mmol) of ethyl o-iodobenzoate (3j) with 10.1 mmol of 2-Zn in the presence of 3 mmol of Cu₂O in DMF at 100 °C for 24 h gave 1.32 g (4.78 mmol) of the corresponding coupling product 4j (96% yield). Similarly, the reaction of 2-Zn with 1.49 g of o-iodonitrobenzene (3n) afforded 1.25 g (5.02 mmol) of the adduct 4n (84% yield). This achievement may lead to a significant contribution as a first scalable CF₂CF₂ fragment installation method.

To be further convinced of the importance of the present reaction, we carried out an additional transformation of the coupling adduct 4n to a CF₂CF₂-containing π-conjugates molecule, a fluorinated tolane derivative, applicable to promising functional materials, such as light-emitting and liquid-crystalline materials (Scheme 3) [39-44].

Thus, CF₂CF₂ group containing 4n with a nitro group at the ortho-position of the aromatic ring was effectively prepared from 2-Zn with the present protocol. 4n was smoothly converted to the corresponding aniline derivative 4s in 87% isolated yield, after exposure to a reductive environment using Fe powder and NH₄Cl. Subsequent Sandmeyer reaction of 4s took place very nicely to afford the corresponding CF₂CF₂-substituted iodobenzene derivative 4t in 67% isolated yield. Then, 4t underwent Pd(0)-catalyzed Sonogashira cross-coupling reaction with phenylacetylene, producing the corresponding tolane derivative 4u with a CF₂CF₂ fragment in good yield (30% overall yield from 2-Zn). Consequently, 2-Zn is found to be a powerful tetrafluoroethylenating agent for producing a broad range of organic molecules with a CF₂CF₂ unit.

Next, the Cu(I)-catalyzed cross-coupling reaction of 2-Zn with benzoyl chloride (5a) as a coupling partner was investigated (Table 3).

Treating 2-Zn (ca. 0.70 M DMF solution) with 2.4 equiv of 5a in the presence of CuI (30 mol %) in DMF at 100 °C for 20 h enabled the formation of the corresponding cross-coupling reaction product 6a in 67% yield (Table 3, entry 1). Optimization of the Cu(I) catalyst for the benzoylation reaction (Table 3, entries 2–4) revealed that Cu₂O was the most efficient for producing 6a (81% by NMR, Table 3, entry 4). To further improve the reaction efficiency, three different additives (TMEDA, phen, and 2,2’-bipyridyl (bpy)) were tested. The first two additives were found to be ineffective for the present reaction (Table 3, entries 5 and 6), whereas bpy led to the quantitative formation of 6a (Table 3, entry 7). The facilitative effect of bpy as an additive made it possible to convert 5a to 6a even at ambient temperature (Table 3, entry 8). Additionally, to utilize the present reaction in an environmentally benign protocol, we conducted the reaction with a decreased amount of 5a. Unfortunately, this slightly lowered the yield of 6a (Table 3, entry 9).

With the optimized conditions (Table 3, entry 8), various kinds of acid chlorides 5b–k could be converted to the corresponding CF₂CF₂-substituted products 6b–k (Figure 4).

Benzoyl chloride derivatives with an electron-donating (5b–d) and an electron-withdrawing group (5e,f) gave rise to the corresponding coupling products 6b–f in 50–98% isolated yields. However, the reaction using nitro-substituted substrate 5g was quite slow under the same conditions. Slight modification of the reaction conditions, namely increasing the amount of 5g and raising the reaction temperature, significantly improved the yield of 6g. Acid chlorides with a heteroaromatic moiety, e.g., 5h and 5i, could also readily participate in the coupling reaction, leading to 6h and 6i in 83% and 87% isolated yields, respectively. Additionally, cinnamoyl chloride (5j) and n-octanoyl chloride (5k) were also suitable electrophiles to form the respective products 6j and 6k in high-to-excellent yields.

Conclusion

We developed a novel and practical tetrafluoroethylenating agent, viz. CH₂=CHCF₂CF₂ZnBr (2-Zn), which can be prepared in large-scale and stored for at least 1.5 years in the refrigerator without decomposition. 2-Zn could be successfully transformed into a broad range of CF₂CF₂-containing molecules with good-to-excellent efficiency. Considering that our previous study found that the vinyl moiety in the coupling product could be a useful molecular building block [14,45], 2-Zn presented here should be a suitable and valuable tetrafluoroethylenating agent for preparing various CF₂CF₂-containing molecules, thereby providing a powerful and practical synthetic tool in organofluorine chemistry.