Portal de Periódicos da CAPES

Fluorous Biphasic Catalysis without Perfluorinated Solvents: Application to Pd-Mediated Suzuki and Sonogashira Couplings

2002; Wiley; Volume: 41; Issue: 23 Linguagem: Inglês

10.1002/1521-3773(20021202)41

1521-3773

Carl Christoph Tzschucke, Christian Märkert, Heiko Glatz, Willi Bannwarth,

Chemical Synthesis and Analysis

With and now without: Perfluoro-tagged catalysts immobilized on fluorous reversed-phase silica gel can be used for Suzuki and Sonogashira C–C coupling reactions (see scheme; R1=e.g., phenyl, 4-MeOC6H4, R2=e.g., 4-NO2C6H4, 4-MeCOC6H4, X=Br, I) without the need for perfluorinated solvents. After the reaction the products were isolated by decantation and the Pd catalysts were recovered and reused. In catalytic reactions easy handling of the catalyst together with its straightforward recovery and possible reuse remain an important topic. A widespread solution to reach these goals is the application of immobilized catalysts. Immobilization can be achieved by covalent attachment to organic polymers or inorganic support materials.1 Alternatively, catalysts can be adsorbed on silica gel2–4 or on reversed-phase silica gel.5 In some cases the immobilization has a beneficial effect on the catalyst's stability.6, 7 One profound advantage of such supported catalysts is the easy separation from the reaction product by filtration or decantation. Ideally, without further purification, there should be little or no contamination of the product with transition metals, which is important when the substances are to be tested in biological assays. A further aspect is the straightforward recovery and reuse of the catalyst. Generally, simplification of workup protocols and handling of small amounts of catalysts is especially important in multiparallel and automated synthesis in combinatorial chemistry.8, 9 Biphasic systems comprising perfluoro-tagged catalysts, which can be extracted out of organic products with fluorous solvents, have emerged as alternative tools for the separation, recovery, and reuse of catalysts as well as for simplified product isolation.10–15 This method, known as fluorous biphasic catalysis (FBC), has been applied to numerous catalytic reactions.8, 10, 16–31 Although perfluorinated solvents have certain advantages, they are expensive and environmentally persistent.32 Thus, it would be beneficial to perform separation and recovery of perfluoro-tagged catalysts without recourse to a perfluorinated solvent as a second liquid phase. Recently, the thermomorphic solution behavior of a fluorous phosphane was utilized in this context.33 While fluorous reversed-phase silica gel (FRPSG) has found application as the stationary phase for chromatographic separation of perfluoro-tagged compounds,34, 35 it has not been used to support catalysts. Our rational was to substitute fluorous solvents by FRPSG, thus gaining easy separation and at the same time retaining the advantages of soluble perfluoro-tagged catalysts. As a first example for this strategy, we report on the immobilization of perfluoro-tagged bis(triphenylphosphane)palladium complexes on FRPSG and their successful application to the Suzuki and Sonogashira CC coupling reactions without need for a fluorous solvent. The fluorous solid supports 11 and 21 were prepared by adapted standard procedures.36, 37 Palladium complexes 3 a–c1 were synthesized by our established procedures.30 To immobilize the complexes, FRPSG was added to a solution of the respective complex in diethyl ether and hexafluorobenzene and the solvent was evaporated. The immobilized precatalyst is an air-stable, free-flowing powder. For ease of handling, especially in separations, FRPSG with a coarse grain (100–300 μm particle size) was used. Silica gels 1 containing 0.1, 1.0, 10, and 100 mg complex per g FRPSG, respectively, were evaluated in the Suzuki cross-coupling of phenylboronic acid and para-nitrobromobenzene and in the Sonogashira coupling of phenylacetylene and para-nitrobromobenzene (Table 1). Entry Catalyst Catalyst Yield [%][a] loading of loading [mol%] FRPSG [mg g−1] Suzuki 1 100 1.5 >98 (>98, 92) 2 10 0.1 >98 (>98) 3 1 0.01 >98 (>98) 4 0.1 0.001 86 (45) Sonogashira 5 100 2 >98 (97, 71) 6 10 0.2 >98 (30, 0) 7 1 0.02 22 (0) 8 0.1 0.002 11 (0) In the Suzuki reaction (Scheme 1) complete conversions were obtained with catalyst loadings down to 0.01 mol % and the catalyst could be recycled (Table 1, entries 1–3). With 0.001 mol % of catalyst, the yield in the first run was 86 % and dropped to 45 % in the second run (Table 1, entry 4). This corresponds to a cumulated turnover number (TON) of 131 000. Suzuki reactions performed with 3 a–c immobilized on support 1 as catalysts. a) 0.001–1.5 mol % catalyst, DME, 2 M aq. Na2CO3, 80 °C. Further experiments were performed by using 10 mg Pd complex per g FRPSG, and a catalyst loading of 0.1 mol %. The three different complexes 3 a–c were used in the coupling of para-nitrobromobenzene and phenylboronic acid. Complexes 3 a and 3 b both have an ethylene spacer separating the perfluoroalkyl chain and the phenyl ring, the perfluoro tag is attached to the para or meta position, respectively. In complex 3 c an OCH2 spacer is used. Earlier experiments had shown that spacers between the perfluoroalkyl chain and the aromatic ring are advantageous for the stability of the catalysts.38 All three catalysts gave complete conversions and could be recycled without significant decrease of activity (Table 2). Catalyst leaching was assessed for the coupling of para-nitrobromobenzene and phenylboronic acid with 0.1 mol % of complex 3 a. Using support 1, a Pd content of 5.4 ppm in the crude organic product was determined by inductively coupled plasma mass spectrometry (ICP-MS),39 which corresponds to 1.8 % of the total Pd. In the inorganic residue, which consists mainly of Na2CO3, 0.2 ppm of Pd was found, which corresponds to less than 0.1 % of the Pd. Thus, less than 1.9 % of the catalyst was washed off from FRPSG 1. In an identical experiment using support 2 a Pd leaching of less than 1.6 % was determined. Accordingly, the nature of the fluorous solid support seems to influence the leaching only marginally. Entry Pd complex Yield [%][a] 1 3 a >98 (93, 93, 93) 2 3 b 96 (98, 93, 91) 3 3 c >98 (96, 95, 95) The catalyst 3 a on support 1 was tested with a number of different substrates (Scheme 2, Table 3). Yields were generally high for electron-deficient aryl bromides (Table 3, entries 1–6, and 11–15) and for aryl iodides (Table 3, entry 8). For electron-rich para-bromoanisole only 48 % conversion was achieved (Table 3, entry 7), but higher conversions were obtained with either higher catalyst loading (Table 3, entry 9) or additional phosphane ligand (Table 3, entry 10). Recycling was successful for very reactive substrates (Table 3, entries 1, 11, and 12). For all other substrates a significant decrease of catalyst activity was observed, the decrease being the greater, the less reactive the halide was. With 3-thienylboronic acid high conversions were obtained in the first run, but no conversion was found with recycled catalyst and only the aryl bromide was recovered (Table 3, entries 15 and 16). This complete loss of activity might be due to catalyst poisening by the thiophene. With cyclohexylboronic acid no conversion was observed (Table 3, entries 17 and 18). Suzuki reactions with different substrates performed with 3 a immobilized on support 1 as catalyst. X=Br, I. a) 0.1 mol % catalyst, DME, 2 M aq. Na2CO3, 80 °C, 15 h. Entry R1 R2 Yield [%][a] 1 Ph 4-NO2-C6H4 95 (97, 97) 2 Ph 4-CH3CO-C6H4 >98 (85, 14) 3 Ph 4-EtOOC-C6H4 93 (69, 4) 4 Ph 2-naphthyl 82 (17) 5 Ph 3,4-F2C6H3 78 (6, 0) 6 Ph 3,4,5-F3C6H2 83 (34, 2) 7 Ph 4-MeO-C6H4 48 (2) 8[b] Ph 4-MeO-C6H4 76, (75, 52) 9[c] Ph 4-MeO-C6H4 70, (10, 0) 10[d] Ph 4-MeO-C6H4 73 (4, 0) 11 4-MeO-C6H4 4-NO2-C6H4 94 (97, 94) 12 4-MeO-C6H4 4-CN-C6H4 94 (97, >98) 13 4-MeO-C6H4 4-CH3CO-C6H4 85 (>98, 29) 14 4-MeO-C6H4 4-EtOOC-C6H4 72 (56, 19) 15 3-thienyl 4-NO2-C6H4 89 (0) 16 3-thienyl 4-CH3CO-C6H4 61 (0) 17 Cy 4-NO2-C6H4 0 (0) 18 Cy 4-CH3CO-C6H4 0 (0) In the Sonogashira reaction (Scheme 3) with 2 mol % catalyst high yields were obtained for three successive experiments (Table 1, entry 5). With 0.2 mol % catalyst conversion was still complete, but the yield dropped significantly when the catalyst was reused (Table 1, entry 6). With lower catalyst loadings, conversions were not complete within 14 h and the recovered FRPSG showed no catalytic activity (Table 1, entries 7 and 8). Sonogashira reactions performed with complex 3 a immobilized on support 1 as catalyst. a) 0.002–2 mol % catalyst, CuI, DME, nBu2NH, 100 °C, 14 h. In summary, we could demonstrate the immobilization of perfluoro-tagged palladium catalysts on FRPSG, and their use for Suzuki and Sonogashira cross-coupling reactions. The catalysts showed activities comparable to those found in liquid–liquid FBC. The catalysts were separated from the product by simple decantation. Leaching was as low as 1.9 % and 1.6 % for 1 and 2, respectively, and the recovered catalyst could be reused in several instances. Because of the dilution of the catalyst with FRPSG, very small amounts of catalysts could be easily and precisely handled. In contrast to conventional FBC approaches, no fluorous solvent was needed for the reaction and the isolation and recovery of the catalyst. An advantage of our strategy compared to conventional covalent catalyst immobilization is that the same support can be used for different catalysts, without the need for a separate linker unit. Before attachment, the catalyst can be characterized by usual methods. It is also conceivable to adjust the reaction conditions in such a way that the catalyst is detached from the FRPSG during the reaction and re-attached during workup. Further investigations to extend this immobilization strategy to other catalytic systems are currently underway. General procedure for Suzuki cross-coupling reactions: A 50 mL Schlenk tube was charged with FRPSG supported catalyst (100 mg), evacuated, and refilled with argon (3×). Stock solutions of the aryl halide (0.3 M in dimethoxyethane (DME), 1.0 mL, 0.3 mmol), the boronic acid (0.33 M in DME (with 4-methoxyphenylboronic acid methanol was used instead), 1.0 mL, 0.33 mmol) and Na2CO3 (2 M in water, 1.0 mL, 2.0 mmol) were added. The tube was sealed with a screw cap and shaken at 80 °C for 15 h. The reaction mixture was cooled to 0 °C and the liquid phase was removed under argon with a pipette. The FRPSG was washed with DME (2×2 mL), water (2×2 mL), and DME (2×2 mL). The combined liquid phases were diluted with water (40 mL) and brine (20 mL) and were extracted with tert-butyl methyl ether (3×20 mL). The combined extracts were concentrated in vacuo, the residue was take up in diethyl ether (2 mL), put on a plug of neutral alumina (3 mL, activity 2–3) and eluted with diethyl ether (∼14 mL). Evaporation of the solvent gave the product. Yields were determined by 1H NMR integration against a known amount of 1,2-dibromoethane. The immobilized catalyst was reused as such in further experiments. Determination of catalyst leaching: A 100 mL Schlenk flask was charged with FRPSG supported catalyst (500 mg, 10 mg Pd complex per g FRPSG, 1.48 μmol), 4-bromonitrobenzene (303 mg, 1.50 mmol), and phenylboronic acid (205 mg, 1.68 mmol), evacuated and refilled with argon (3×). DME (10 mL) and Na2CO3 (2 M in water, 5.0 mL, 10.0 mmol) were added. The flask was shaken under argon at 80 °C for 15 h, and then the reaction mixture was cooled to room temperature and filtered. The residue was washed with DME (2×10 mL), water (2×10 mL), and again with DME (2×10 mL), and the organic and aqueous filtrates were collected separately. The solvents were removed in vacuo, the resulting solids were powdered, and the Pd content determined by ICP-MS.