Palladium-catalysed synthesis of triaryl(heteroaryl)methanes

Tetraarylmethane derivatives are desirable for a variety of applications, but difficult to access with modern C–C bond-forming reactions. Here we report a straightforward method for palladium-catalysed arylation of aryl(heteroaryl)methanes and diaryl(heteroaryl)methanes with aryl chlorides. This reaction enables introduction of various aryl groups to construct triaryl(heteroaryl)methanes via a C–H functionalization in good to excellent yield, and represents the first step towards a general transition metal catalysed synthesis of tetraarylmethanes.

T etraarylmethanes and related derivatives are important building blocks, with uses ranging from molecular devices [1][2][3][4] to porous organic frameworks [5][6][7] and applications from protein translocation detection 8 to drug delivery 9 . Furthermore, a recent study of 9,000 bioactive compounds by AstraZeneca to evaluate molecular space concluded that most biologically active compounds are linear or disk shaped, and very few are sphere-like molecules. They concluded that chemists should 'y expand the current arsenal of tools to access less populated space y' and 'y this may prove advantageous as the pharmaceutical industry ventures into new disease areas and new target classes which require different molecular shapes to bind and achieve the desired effect' 10 . Tetraarylmethanes are members of spherelike molecules that have not been widely explored due to synthetic difficulties.
Transition metal-catalysed cross-coupling reactions have emerged as an excellent method to construct C-C bonds. For example, they have been used with great success in the coupling of aryl halides to diarylmethanes to afford triarylmethanes [22][23][24][25] (Fig. 1c) 26 . The application of transition metal catalysts to the synthesis of tetraarylmethanes, however, has proven quite challenging. This may be due to difficulties associated with the transmetallation of the bulky triarylmethyl organometallic species. Thus, there is tremendous unrealized potential in the transition metal catalysed construction of tetraarylmethane derivatives.
We are only aware of a handful of metal catalysed syntheses of tetraarylmethane derivatives. In their synthesis of triarylmethanes, Yorimitsu, Oshima and co-workers noted formation of 8% yield of a tetraarylmethane (Fig. 1c). Despite the intriguing nature of this byproduct, subsequent reports have not been forthcoming. Ghosh and co-workers reported the nickel-catalysed reaction of carbon tetrachloride and PhMgCl to form a 3:2 ratio of triphenylmethyl chloride to tetraphenylmethane (Fig. 1d) 27 . These products were not isolated. Recently, the palladium-catalysed arylation of fluorenes by the teams of Wu and Song, Xie, and Huang was reported to give diarylfluorenes ( Fig. 1e) 28,29 . Fluorene is more acidic and less sterically demanding than other diarylmethane derivatives. In significant work, the team of Nambo and Crudden outlined the generation of triarylacetonitriles, Ar 3 C-CN, then assemble a 4th heteroaryl group by building on the nitrile (Fig. 1f) 30 .   Herein, we present a novel palladium-catalysed method for the synthesis of heteroaryl-substituted tetraarylmethane derivatives (Fig. 1g). This effort represents the first high-yielding transition metal catalysed preparation of tetraarylmethane derivatives. It enables the synthesis of a broad range of triaryl(heteroaryl)methanes, including those with four different aryl groups attached to the central carbon atom.

Results
Preliminary catalyst screening. Given that the only example of arylation of a triarylmethane derivative we are aware of is Yorimitsu, Oshima and co-workers' 8% yield of triphenyl(4pyridyl)methane (Fig. 1c), we chose this starting point for our reaction design. Instead of CsOH, we employed KOt-Bu, because of the greater solubility of this base. We reasoned that this would also allow us to lower the reaction temperature, and thus use toluene as the solvent. We initiated a search for a suitable catalyst for bis-arylation of 4-benzylpyridine (1a) with bromobenzene (2a) by examining phosphine ligands on microscale using 10 mmol of 1a (see Supplementary Table 1). Thus, 48 electronically diverse mono-(20 mol%) and bidentate phosphine ligands (10 mol%) were tested using Pd(OAc) 2 (10 mol%) and 4 equiv KOt-Bu. Reactions were conducted at 110°C for 12 h, cooled, and  ARTICLE quenched with water. The most promising microscale results, based on product to internal standard (4,4 0 -di-tert-butylbiphenyl) ratios, were with PCy 3 (L1) and cataCXium A (L2, Fig. 2).
Repeating the bis-arylation reaction in Fig. 2 using 4-benzylpyridine 1a and 4-bromotoluene 2b with PCy 3 (L1) and cataCXium A (L2) on laboratory scale (0.25 mmol) led to the desired product in 68 (PCy 3 ) and 77% AY (cataCXium A) after 12 h (AY ¼ assay yield, determined by 1 HNMR analysis, see the Methods section for details) ( Table 1, entries 1-2). With cataCXium A as the ligand, 4-iodotoluene 2c reacted with 4-benzylpyridine 1a to afford tetraarylmethane product in 76% isolated yield (entry 3). Considering that aryl chlorides are more abundant and less expensive than aryl bromides, we examined aryl chlorides. In the event, the reaction of 4-benzylpyridine 1a with 4-chlorotoluene 2d using the same conditions as above was conducted. Interestingly, the diarylated products were obtained in nearly identical assay yields with the aryl chlorides (66% PCy 3 and 74% cataCXium A, entries 4-5) and aryl bromides. Although cataCXium A exhibited better AY than PCy 3, we chose to first optimize the reaction with the more economical and readily available PCy 3 . As outlined below, more challenging substrates gave better results with cataCXium A. To begin with, the catalyst loading was examined. Changing the loading from 10 to 5 mol% resulted in no change in AY (entry 6). Further reducing the amount of catalyst to 2.5 mol%, however, resulted in a drop to 35% (entry 7).
To demonstrate the potential utility of our method, we performed a gram scale reaction of 4-benzylpyridine 1a with 4-chlorobenzophenone 2j. The desired product 4g was obtained in 79% yield (1.88 g, Fig. 3), demonstrating the reaction is scalable (see Supplementary Methods).
Kato and co-workers reported a liquid-crystalline bowl-shaped molecule that form columnar and micellar cubic structures, using triary(4-pyridyl)methane moieties as building blocks 18 . In order to apply our method, we synthesized the same compound ( Fig. 4) from the reaction product 4q (Table 3). It is noteworthy that Kato synthesized triary(4-pyridyl)methane was based  on Friedel-Crafts arylations from diaryl(4-pyridyl)methane with 46% yield, which is less than our method (91% yield, see Supplementary Methods).

Discussion
Our research team has been interested in the catalytic functionalization of weakly acidic sp 3 -hybridized C-H bonds through a DCCP. These reactions involve a weakly acidic C-H of the substrate (pronucleophile) that is reversibly deprotonated under the reaction conditions. Subsequently, the nucleophile undergoes catalyst promoted arylation [33][34][35][36] or vinylation 37 . This method has been particularly successful with the generation of triarylmethanes [38][39][40] . Tetraarylmethane derivatives are challenging to efficiently prepare by both classical and state-ofthe-art methods. The breadth of their applications has outpaced chemists' ability to prepare them in a concise fashion. Crosscoupling methods represent an appealing approach to tetraarylmethane derivatives, but to date successful reports of such processes are still lacking. Outlined herein is a palladiumcatalysed DCCP for direct arylation of aryl(heteroaryl)methanes and diaryl(heteroaryl)methanes with aryl chlorides to provide triaryl(heteroaryl)methanes. Unlike traditional cross-coupling procedures, which employ prefunctionalized coupling partners, our approach relies on reversible deprotonation of the diarylmethane derivatives under the conditions used for the catalytic C-C bond forming reaction. Under our reaction conditions, a variety of triaryl(heteroaryl)methanes were prepared in good to excellent yields. This communication represents the first steps towards our goal of developing metal catalysed approaches for the construction of a wide range of tetraarylmethanes.

Methods
General procedure A. An oven-dried 8.0 ml reaction vial equipped with a stir bar was charged with 2-methylthiazole (7a, 0.50 mmol, 1.0 equiv) or 2-methylbenzoxazole (7b, 0.50 mmol, 1.0 equiv) and chlorobenzene (2e, 3.0 equiv) in a glove box under a nitrogen atmosphere at room temperature. A stock solution containing Pd(OAc) 2 (5.6 mg, 0.025 mmol, 5 mol%) and PCy 3 (14.0 mg, 0.05 mmol, 10 mol%) in dry o-xylene (2.5 ml). Then, NaOt-Bu (3.0 equiv) was added to the reaction mixture. The vial was capped, removed from the glove box, and heated to 130°C for 12 h with stirring. The reaction mixture was quenched with 0.5 ml of H 2 O, diluted with 10 ml of ethyl acetate, and filtered over a pad of MgSO 4 and silica. The pad was rinsed with ethyl acetate (10 Â 2 ml) and the combined solutions were concentrated in vacuo. The crude material was loaded onto a deactivated silica gel column and purified by flash chromatography to afford the desired products 8a and 8b in 81 and 49% yield, respectively.
General procedure B. An oven-dried 8 ml reaction vial equipped with a stir bar was charged with aryl(4-pyridyl)methanes (1, 0.10 mmol, 1.0 equiv) or diaryl(heteroaryl)methanes (5, 0.10 mmol, or 8, 0.20 mmol, 1.0 equiv) and aryl chlorides (2, 2.0-4.0 equiv) in a glove box under a nitrogen atmosphere at room temperature. A stock solution containing Pd(OAc) 2 (1.1 mg, 0.005 mmol, 5 mol%) and PCy 3 (2.8 mg, 0.01 mmol, 10 mol%) in dry THF or cataCXium A (3.6 mg, 0.01 mmol, 10 mol%) in dry 1,4-dioxane was taken up by syringe and added to the reaction vial under nitrogen. Then, NaOt-Bu (2.0-4.0 equiv) was added to the reaction mixture. The vial was capped, removed from the glove box, and heated to 100°C for 12 h with stirring. The reaction mixture was quenched with three drops of H 2 O, diluted with 3 ml of ethyl acetate, and filtered over a pad of MgSO 4 and silica. The pad was rinsed with ethyl acetate (3 Â 2 ml) and the combined solutions were concentrated in vacuo. The crude material was loaded onto a deactivated silica gel column and purified by flash chromatography to afford the desired products.
Data availability. The authors declare that the data supporting the findings of this study are available within the article and its Supplementary Information