The chemistry of π-allyl palladium complexes has become one of the most successful areas of organic synthesis, as exemplified by the Tsuji-Trost reaction ^[1]. In contrast, reactions involving the η³-benzyl palladium complexes generated from benzylic derivatives are much less common because they require the dearomatization of an aryl system. Despite the difficulties associated with the preparation of η³-benzyl palladium complexes, these systems are considered to be useful intermediates in organic synthesis, because they can be used as special η³-allyl-palladium species. In this way, η³-benzyl palladium intermediates have been successfully substituted at their benzylic, ortho- and para-positions with a wide variety of different nucleophiles, including carbon ^{[2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]}, nitrogen ^{[16, 26, 27]}, oxygen ^{[16, 28]} and sulfur-based nucleophiles ^{[16, 29]} (Fig. 1).

Fig. 1. η3‐Benzyl palladium intermediates and possible products.

η³-Benzyl palladium intermediates can be generated by the oxidative addition reactions of a wide range of benzylic derivatives, including carbonates ^{[3, 4, 5, 6, 7, 28, 29, 30, 31]}, phosphates ^{[8, 9, 10]}, carboxylates ^{[11, 12, 13, 14, 15]} and even fluorides ^[16], to Pd(0). Benzyl chlorides can also be treated as common precursors for the formation of η³-benzyl palladium species because they are readily available with various commercial suppliers and stable ^{[17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]}. Several elegant palladium-catalyzed benzylation processes have been developed during the course of the last decade. In 2001, Yamamoto group ^[17] reported a facile palladium-catalyzed allylative dearomatization reaction involving the reaction of benzyl chloride with allyltributyltin at the para-position. Following on from this work, Bao group ^{[19, 20, 21, 22, 23, 24, 25, 26, 27]} reported their research towards the development of a series of reactions involving η³-benzyl palladium intermediates. Despite these developments, there have been very few studies reported in the literature pertaining to the reaction of electron-rich arenes as nucleophiles with η³-benzyl palladium intermediates ^{[32, 33]}. In this study, it was envisioned that indole could be used as an effective nucleophile to attack benzylic systems and afford the corresponding triarylmethane compounds bearing an indole fragment, which could have numerous potential applications in organic synthesis ^{[34, 35, 36, 37, 38, 39, 40]}. It is noteworthy, however, that the use of indole in this way could also result in nucleophilic attack at the ortho- or para-position of the naphthyl ring, which would result in the formation of the corresponding dearomatized product through a catalytic dearomatization process ^{[41, 42, 43, 44]}. Herein, we report the results of our recent study towards the Pd-catalyzed benzylation of indoles.

The reaction of 1-(chloro(phenyl)methyl)-naphthalene (1a) with indole (2a) was initially selected as a model reaction to allow for the optimization of the reaction conditions. Several different variables were evaluated, including the solvent, base and reaction time, in terms of their impact on the yield of the reaction, and the results are summarized in Table 1. Pleasingly, the reaction proceeded smoothly when it was conducted in CH₃CN at room temperature in the presence of 10 mol % Pd(PPh₃)₄ and 2.0 equivalents of Cs₂CO₃ to afford the triarylmethane product 3a as a single regioisomer in 89% yield (Table 1, entry 1). Furthermore, none of the dearomatized product resulting from the nucleophilic attack of the indole on the naphthalene ring was observed. Encouraged by this preliminary result, we proceeded to evaluate the effects of several other solvents (i.e., toluene, 1,4-dioxane, Et₂O, DCM and THF) and bases (i.e., Cs₂CO₃, Li₂CO₃, Na₂CO₃, K₂CO₃, Et₃N and DBU) on the yield of the reaction (Table 1, entries 2-11). The highest yield of all of the conditions tested was achieved using a combination of THF with Cs₂CO₃, which provided the desired product in 99% yield (Table 1, entry 6). It is noteworthy that the reaction proceeded in only 60% yield when it was conducted without base (Table 1, entry 12). Furthermore, the reaction did not give any of the desired product when it was conducted in the absence of Pd(PPh₃)₄, which suggested that this reaction was proceeding via an η³-benzyl palladium intermediate (Table 1, entry 13). Decreasing the amounts of catalyst and indole led to a reduction in the yield of the desired product (Table 1, entries 14-16). Based on these results, the optimized conditions were determined to be the following: 5 mol % Pd (PPh₃)₄, 2.0 equiv of Cs₂CO₃ and 2.0 equiv of indole in THF at room temperature.

Table 1 Optimization of the reaction conditions a Entry Solvent Base Time (h) Yield b (%) 1 CH3CN Cs2CO3 15 89 2 toluene Cs2CO3 3.0 91 3 1,4-dioxane Cs2CO3 25 65 4 Et2O Cs2CO3 25 70 5 DCM Cs2CO3 4.0 97 6 THF Cs2CO3 2.0 99 7 THF Li2CO3 2.5 94 8 THF Na2CO3 2.5 88 9 THF K2CO3 2.0 80 10 THF Et3N 30 85 11 THF DBU 20 65 12 THF — 2.0 60 13 c THF Cs2CO3 12 — 14 d THF Cs2CO3 24 83 15 e THF Cs2CO3 30 75 16 f THF Cs2CO3 4.0 85 a Reaction conditions: 1a (0.5 mmol), 2a (1.0 mmol), base (1.0 mmol) and Pd(PPh3)4 (0.025 mmol) in solvent (3 mL) at room temperature under an argon atmosphere. b Isolated yield. c No Pd(PPh3)4. d 1.5 equiv of 2a and Cs2CO3. e 1.0 equiv of 2a and Cs2CO3. f Pd(PPh3)4 (2.5 mol %).

Table 1 Optimization of the reaction conditions a

With the optimized conditions in hand, we proceeded to evaluate the scope of this reaction using a variety of different substituted indoles and benzyl chloride-type compounds. The results of these experiments are summarized in Table 2. The electronic properties of the indole ring had no discernible impact on the outcome of the reaction, with indoles 2 bearing an electron-donating (i.e., 7-Me or 5-MeO) or electron- withdrawing group (i.e., 4-Br, 5-Br, 6-F, 6-Cl or 6-Br) proceeded smoothly under the optimized condition to give the corresponding benzylated products in 90%-99% yields (Table 2, entries 1-8). N-Methyl indole also reacted smoothly to give benzylated product 3i in an excellent yield (Table 2, entry 9). Furthermore, the optimized conditions appeared to be general for benzyl chlorides 1 bearing a variety of different substituents on the phenyl ring, with the corresponding benzylated products being obtained in 91%-98% yields (Table 2, entries 10-13). Notably, 2-(chloro(phenyl)-methyl)-naphthalene was also well tolerated under the optimized conditions (Table 2, entry 14). In contrast, however, only trace amounts of the benzylated product 3o were observed following the reaction of chlorodiphenylmethane with indole, which indicated that the η³-naphthyl palladium intermediate had been formed preferentially (Table 2, entry 15). The structure of compound 3l was unambiguously determined by X-ray crystallography (Fig. 2).

Fig. 2.. 2. X-ray crystal structure of 3l.

Table 2 Pd-catalyzed benzylation of indoles a. Entry 3, R1, R2, R3 Time (h) Yield b (%) 1 3a, H, H, H 2.0 99 2 3b, H, 7-Me, H 5.0 90 3 3c, H, 5-MeO, H 2.0 91 4 3d, H, 4-Br, H 3.0 93 5 3e, H, 5-Br, H 2.0 96 6 3f, H, 6-Br, H 3.0 94 7 3g, H, 6-F, H 2.5 90 8 3h, H, 6-Cl, H 4.0 96 9 3i, H, H, Me 3.0 99 10 3j, 4-tBu, H, H 2.0 98 11 3k, 4-Me, H, H 2.0 98 12 3l, 4-F, H, H 2.0 91 13 3m, 4-Br, H, H 2.5 94 14 3.0 98 15 20 trace a Reaction conditions: 1 (0.5 mmol), 2 (1.0 mmol), Pd(PPh3)4 (0.025 mol) and Cs2CO3 (1.0 mmol) in THF (3 mL) at room temperature under Ar atmosphere. b Isolated yield.

Table 2 Pd-catalyzed benzylation of indoles a.

The asymmetric benzylation of indole was also investigated using a chiral ligand. Unfortunately, however, the use of the (R,R)-DACH-phenyl Trost ligand (L1) or (R)-BINAP (L2) led to a racemic mixture of 3a, albeit in excellent yield (Scheme 1).

Scheme 1. Attempted asymmetric benzylation of indole.

Based on these results, we proposed a catalytic cycle for this transformation, which is shown in Scheme 2. Oxidative addition of 1-(chloro(phenyl)methyl)naphthalene (1a) to Pd(0) would generate the η³-naphthyl palladium intermediate I, which would be in equilibrium the intermediate II. Intermediate I would be preferentially attacked by the indole at the benzylic position to afford the benzylated product 3a, together with the regeneration of the Pd(0) catalyst.

Scheme 2. Proposed catalytic cycle.

In summary, we have developed an efficient palladium-catalyzed method for the completely regioselective benzylation of indoles to give a variety of triarylmethanes in good yields under mild conditions.