Cross-coupling reactions used to construct biaryl compounds mainly involve Ar1X as an electrophile and Ar2M as a nucleophile.1 Recently, CH bond activation has been used, where Ar2H served as the nucleophile to react with Ar1X (direct arylation of arenes).2, 3 This strategy efficiently avoids the disposal of stoichiometric amounts of metal waste generated from the organometallic reagent, ArM, in the traditional coupling manner.2, 3 Various transition metals have been reported as efficient catalysts for this transformation, for example, Pd,4–11 Rh,12–15 Ru,16–18 Ir,19, 20 Cu,21–27 and other transition metals.28–30
The application of inexpensive, non-toxic, commercially available, and environmentally benign iron complexes as catalysts in chemical syntheses has attracted much attention.31, 32 Recently, iron has been utilized extensively as a catalyst to promote the “traditional” cross-coupling between R1X and R2M.33–41 Iron catalysts are also involved in many important transformations, such as Friedel–Crafts benzylation,42, 43 carbonylation,44 oxidation45, 46 and other processes.47–54
Of great interest are the recently developed oxidative coupling reactions of Ar1M with Ar2H to generate Ar1Ar2 products by employing Fe complexes as the catalysts. Nakamura and co-workers reported an elegant Fe-catalyzed oxidative coupling reaction between Ar2H, which contain directing groups, and diaryl zinc reagents.55, 56 Yu and co-workers explored the oxidative reaction between Ar2H and Ar1B(OH)2 using a stoichiometric amount of iron reagent.57 Many other oxidative coupling reactions, which involve CH activation using iron catalysts in the presence of stoichiometric amounts of oxidants, have also been reported.58–61 However, to the best of our knowledge, no example of Fe-catalyzed cross-coupling between Ar1X and Ar2H has been reported (Scheme 1). Herein, we report the development of novel Fe-catalyzed cross-coupling reactions between electrophilic Ar1X (X=I, Br, Cl) and unactivated Ar2H coupling partners.
Initially, we chose 4-bromoanisole and unactivated benzene as the model substrates for surveying reaction parameters in the model reaction (Table 1). Commonly used inorganic bases, such as Li2CO3, Na2CO3, K2CO3, Cs2CO3, and K3PO4 were employed in the reaction, however, no conversion was observed at 80 °C for 48 hours in the presence of the iron salt. When NaOtBu was employed, a trace amount of the direct arylation product was observed by GC/MS analysis. This outcome encouraged us to further examine the feasibility of this catalysis.
|Entry||[Fe] (mol %)||Ligand||Base||Benzene [mL]||Yield [%][b]|
When LiHMDS (3.0 equiv) was added as a base, the desired product 3 a (16 % yield) was obtained in the presence of FeCl3 (20 mol %; Table 1, entry 3). Meanwhile, FeCl2 and FeBr2 were less efficient catalysts under the same reaction conditions (Table 1, entries 1 and 2). The nature of the base and the ligand was essential to influence the yield of the product (Table 1, entries 4–14). In the presence of DMEDA (40 mol %), the direct arylation of benzene was achieved in 66% yield (Table 1, entry 9), and this result showed that this particular diamine ligand was more efficient than others. Notably, the amount of benzene has some effect on this reaction. The best result (75 % yield) was obtained when the reaction was carried out in the presence of 4 mL of benzene (90 equiv; compare Table 1, entries 9–11). The effects of catalyst loading was also examined, and 15 mol % was the best and yielded 3 a in 79 % yield (compare Table 1, entries 10, 12, and 13). When LiHMDS was replaced with NaOtBu or LiOtBu, the reactions only produced trace amounts or no desired product, respectively (Table 1, entries 15 and 16). When KOtBu was employed as the base, the desired product 3 a was obtained in 61 % yield (Table 1, entry 17). Moreover, the direct arylation of benzene was not observed at room temperature, and an increase to 60 °C led to poor conversion.
Control experiments revealed that no reaction was observed in the absence of FeCl3 (Table 1, entries 18 and 19). To eliminate the contaminants which might potentially affect the catalysis, a different batch of Fe complex was employed. When high purity FeCl3 (>99.99 %, from Aldrich) was used under the standard conditions (15 mol % of FeCl3, 30 mol % of DMEDA, 2.0 equiv of base), the yield remained unchanged (see the Supporting Information).62 This result suggested that the CH activation reaction was solely catalyzed by the Fe complex. When CuCl, CuBr, or CuI were tested as catalysts under the standard conditions, both conversion and yield became relatively low, thus indicating that Cu salts were much less effective catalysts in this reaction (for more details see the Supporting Information). These experiments further indicate that the Fe catalyst plays a crucial role in this aromatic CH transformation.
With our optimized reaction conditions in hand, the scope of this direct arylation of benzene with various aryl halides was investigated (Table 2, Scheme 2, and Scheme 3). In general, the yields of the reactions with electron-rich aryl bromides were good (72 %–81 % yield; Table 2, entries 1–3, 5, 7, and 9) and were higher than those with the electron-deficient aryl bromides (Table 2, entries 11–13). 2-Naphthyl bromides 1 h and 1 j also underwent direct arylation smoothly and afforded the desired products in 70 % and 62 % yield, respectively (Table 2, entries 8 and 10). Aryl bromides with ortho-substituted groups led to lower yields (Table 2, entries 4 and 6).
|Entry||Aryl halides||Product||Yield [%][b]|
|1||1 a||3 a||81|
|2||1 b||3 b||77|
|3||1 c||3 c||73|
|4||1 d||3 d||45|
|5||1 e||3 e||73|
|6||1 f||3 f||37|
|7||1 g||3 g||72|
|8||1 h||3 h||70|
|9||1 i||3 i||72|
|10||1 j||3 j||62|
|11||1 k||3 k||63|
|12||1 l||3 l||51|
|13||1 m||3 m||45|
Aryl iodides containing electron-donating and electron-withdrawing groups reacted smoothly with benzene and gave the corresponding products in the range of yields (62 %–82 %; Scheme 2). Interestingly, in the direct arylation of benzene with aryl iodides reactions which employed KOtBu as the base gave better yields compared with LiHMDS (Scheme 2). Notably, no homo-coupling products from the corresponding aryl halides were observed when using either aryl bromides or aryl iodides.
In addition to aryl bromides and iodides, the applicability of more challenging aryl chlorides in this reaction was examined. The reaction of benzene chloride gave the desired coupling product (52 % yield), and the reactions of p-MeC6H4Cl and p-MeOC6H4Cl gave 34 % and 44 % yields, respectively (Scheme 3). Under the standard reaction conditions, ArCl bearing electron-withdrawing groups resulted in poor yields (<5 %; Scheme 3). Moreover, no reaction was observed using fluorobenzene as the electrophile under the same conditions.
Competitive experiments were carried out to probe the reactivity between the chloro and the bromo group. Aryl bromide was preferentially transformed, while the aryl chloride was tolerated at either the para or ortho position under the reaction conditions [Eq. (1) and (2)]. Both the reaction of p-ClC6H4Br and o-ClC6H4Br gave good yields, 72 % and 73 %, respectively.
When 1,4-dibromobenzene or 1,4-diiodobenzene was treated with benzene in the presence of base (3 equiv), 1,4-diphenylbenzene was produced in 43 % and 53 % yields, respectively (Scheme 4).
The Fe-catalyzed direct arylation of substituted arenes with aryl bromides was investigated, and the results are listed in Table 3. The reaction between anisole and bromobenzene gave 69 % combined yield of direct arylation product (three isomers, with a ratio of 64:24:12; Table 3, entry 2). Naphthalene could be arylated in moderate yields (65 % and 68 %, respectively; Table 3, entries 3 and 4), and produced a mixture of 1- and 2-substituted naphthalene derivatives. Direct arylation of toluene with 4-MeOC6H4Br afforded the direct arylated product in 39 % yield. Significantly, in the presence of Fe catalyst (15 mol %), ferrocene was successfully arylated with bromobenzene, and gave the corresponding monoarylated product (54 % yield; Table 3, entry 5).
|Entry||Aryl halides||Arene||Product||Yield [%][b]|
|1[c]||4 a||39 (o/m/p= 47:34:19)|
|2[d]||4 b||69 (o/m/p= 64:24:12)|
|3[e]||4 c||65 (α/β= 61:39)|
|4[e]||4 d||68 (α/β= 69:31)|
If the reaction proceeded through a benzyne pathway, when substituted aryl bromide such as 1 a was employed, two products would be generated. However, the reactions of various substituted aryl bromides with benzene all produced the sole corresponding direct arylation products (Table 2), which clearly ruled out the benzyne mechanism. The isotope effect of the reaction was examined. The reaction was conducted under the standard conditions using 4-bromoanisole (1 a) coupling with equimolar amounts of benzene and [D6]benzene (see the Supporting Information). The labeling experiment yielded 3 a and deuterated 3 a in the ratio of 5:3 (kH/kD=1.7).
In summary, we have successfully developed the first novel Fe-catalyzed direct arylation of unactivated arenes with a broad range of aryl halides, including aryl chlorides, bromides, and iodides, through CH bond activation to prepare biaryl compounds. The choice of the base and ligand are essential to the success of this Fe-catalyzed direct arylation of unactivated arenes. Further studies on the mechanism are actively underway, and will be reported in due course.