Among aromatic carbonyl compounds, a,b-alkynones are important structural motifs in many bioactive molecules and intermediates for the synthesis of natural products and pharmaceuticals [1, 2, 3, 4]. In general, a,b-alkynones have been prepared by the coupling of alkyne organometallic reagents with acid chlorides in the presence of transition metal reagents [5, 6, 7, 8, 9]. Unfortunately, this methodology is limited with respect to functional group tolerance, substrate stability, reaction conditions (dry solvent under an inert atmosphere), atom economy, and environmental friendliness. Since they were first reported by Tanaka’s group [10] in 1981, palladium (Pd)- catalyzed carbonylative Sonogashira reactions have become an alternative approach to synthesize a,b-alkynones. In carbonylative Sonogashira reactions, CO acts as an important C1 building block to introduce carbonyl group into the parent molecules. Many researchers have modified and improved this methodology, including Wu et al. [11] for homogeneous carbonylative Sonogashira reactions and other groups [12, 13, 14] for heterogeneous ones. It is widely accepted that the involved ligands can greatly tailor the catalytic performance of homogenous Pd complex catalysts through electronic and steric effects [15, 16, 17, 18, 19]. However, these catalysts still suffer from the problems of difficult product separation, rapid activity loss, and non-recyclability [11, 20, 21].
Non-volatile ionic liquids (ILs) have been recognized as promising alternative solvents to immobilize homogenous palladium catalysts, preventing their leaching and deactivation [22]. ILs can be functionalized with phosphine ligands, and the resulting phosphine-functionalized ILs have been investigated as ionic complexes (catalysts) and used in combination with room-temperature ionic liquids (RTILs) to immobilize homogenous catalysts [23, 24, 25, 26, 27]. Because the melting points of phosphine-functionalized ILs are usually higher than 100℃, they are typically defined as ionic phosphines. It has also been found that when phosphines are introduced into ILs to obtain ionic phosphines, the positive charges of ILs with strong electron-withdrawing character dramatically influence the coordinating ability of the involved phosphines and in turn the properties of the corresponding transition metal complexes in terms of structure and catalytic activity [27, 28].
Herein, the neutral monophosphine L1 (1-(5’- diphenyl- phosphinothiazol-2’-yl)-imidazolyl) and its ionic counterpart L2 (1-(5’-diphenylphosphinothiazol-2’-yl)-3- methylimidazolium trifluoromethanesulfonate) reported by us previously [29] were complexed with PdCl2(CH3CN)2 to afford the neutral Pd(II) complex bis-[1-(5’-diphenylphosphinothiazol-2’-yl)- imidazolyl] dichloropalladium(II) (1A) and an ionic complex bis-[1-(5’-diphenylphosphinothiazol-2’-yl)-3-methylimidazolium] dichloropalladium(II) trifluoromethanesulfonate (2A) (Scheme 1). The effects of the positive charge in L2 on the coordination ability of the involved phosphine, the structure of the corresponding Pd complex, and the catalytic performance of the complexes in the carbonylative Sonogashira reaction are discussed (Scheme 2). In addition, 2A can be regarded as an ionic Pd complex that is compatible with RTILs, so its recyclability in 1-butyl-3-methylimidazolium hexafluorophosphate ([Bmim]PF6) is investigated.
L1 and L2 were first prepared according to our previously published method [29]. Other chemical reagents were purchased from Shanghai Aladdin Chemical Reagent Co. Ltd. and Shanghai Alfa Aesar, China, and used as received. Diethyl ether was distilled from sodium. DMF was dried over CaSO4 and distilled before use. Both diethyl ether and DMF were stored over 4Å molecular sieves under N2. 1H, 31P, and 13C NMR spectra were recorded on a Bruker ARX 400 spectrometer (400 MHz) at room temperature. 31P NMR spectra were referenced to 85% H3PO4 sealed in a capillary tube as an internal standard. CHN elemental analysis was performed on a Vario EL III Elemental Analyzer. The amount of Pd in the organic phase was quantified by inductively coupled plasma atomic emission spectrometry (ICP-AES) on an IRIS Intrepid II XSP instrument (Thermo Electron Corporation). Gas chromatography (GC) was performed on a Shimadzu-2014 equipped with a DM-Wax capillary column (30 m x 0.25 mm x 0.25 mm). GC-mass spectrometry (GC-MS) was recorded on an Agilent 6890 spectrometer equipped with an Agilent 5973 mass selective detector. Purification of the compounds was carried out by flash chromatography on silica gel.
Under N2 atmosphere, L1 (0.13 g, 0.4 mmol) dissolved in dry CH2Cl2 (4 mL) was added to a solution of PdCl2(MeCN)2 (0.05 g, 0.2 mmol) in dry CH2Cl2 (8 mL). The mixture was stirred vigorously at room temperature for 3 h. The solvent was removed under vacuum using a rotary evaporator. The obtained yellow solid residue was washed with diethyl ether and dried under vacuum to give the product 1A in 97% yield. A sample suitable for single-crystal X-ray diffraction (XRD) analysis was obtained by slow volatilization of a CH2Cl2 solution containing 1A. 1H NMR (δ, CD2Cl2): 8.23 (s, 2H, NCHN), 7.55-7.75 (m, 24H), 7.21 (s, 2H); 13C NMR (δ, CD2Cl2): 164.43 (s, NCN), 151.17 (s), 135.69 (s), 134.25 (s), 131.70 (s), 131.05 (s), 128.67 (s), 128.23 (s), 121.70 (s), 117.62 (s); 31P NMR (δ, CD2Cl2): 10.80 (s, PPh2). CHN (%) elemental analysis for 1A (C36H28Cl2N6P2Pd1S2, 848.0): C 50.82, H 3.43, N 9.96 (Calcd. C 50.99, H 3.33, N 9.91).
Under N2 atmosphere, L2 (0.250 g, 0.50 mmol) dissolved in dry CH2Cl2 (3 mL) was added to a solution of PdCl2(MeCN)2 (0.065 g, 0.25 mmol) in dry CH2Cl2 (7 mL). The mixture was stirred vigorously at room temperature for 3 h. The collected yellow solid precipitate was washed with diethyl ether and dried under vacuum to give the product 2A with a yield of 86%. A sample suitable for single-crystal XRD analysis was obtained by slow volatilization of a CH2Cl2 solution containing 2A. 1H NMR (δ, CD2Cl2): 9.93 (s, 2H, NCHN), 8.39 (t, 2H, J=2 Hz), 8.04 (t, 2H, J=2 Hz), 7.77-7.83 (m, 10H), 7.65-7.69 (m, 4H), 7.58-7.63 (m, 8H), 4.23 (s, 6H, 2CH3); 13C NMR (δ, CD3CN): 160.80 (s, NCN), 150.61 (s), 135.80 (s), 134.33 (t, J=27 Hz), 132.54 (s), 132.19 (s), 128.99 (t, J=22.5 Hz), 127.60 (q, J=117.5 Hz, CF3), 125.22 (s), 122.40 (s), 120.73 (s), 36.77 (s, CH3); 31P NMR (δ, CD2Cl2): 11.8 (s, PPh2). CHN (%) elemental analysis of 2A (C40H34Cl2F6N6O6P2Pd1S4, 1176.2): C 40.76, H 2.97, N 7.19 (Calcd. C 40.84, H 2.91, N 7.14).
Phosphine selenides were prepared by the reactions of phosphines L1, L2, and PPh3 with 30 mg selenium (7.63% 77Se) with a molar ratio of 1:1 in CDCl3 at 50℃ for 24 h. The selenides were directly analyzed without isolation by a Bruker Avance 500 spectrometer at ambient temperature.
Intensity data for 1A and 2A were collected at 25℃ on a Bruker SMART APEX II diffractometer using graphite- monochromated Mo-Ka radiation (l=0.071073 nm). Data reduction included absorption correction by the multi-scan method. The structures were solved by direct methods and refined by full-matrix least-squares using SHELXS-97 (Sheldrick, 1990), with all non-hydrogen atoms refined anisotropically. Hydrogen atoms were added at their geometrically ideal positions and refined isotropically. The crystal data and refinement details are given in Table 1.
In a typical experiment, the isolated crystalline precatalyst 1A (or 2A, 0.005 mmol) was sequentially mixed with 3 mL of solvent (DMF or [Bmim]PF6 if required), iodobenzene (5 mmol), phenylacetylene (6 mmol), and Et3N (7.5 mmol). The obtained mixture was placed in a sealed Teflon-lined stainless steel autoclave, purged with syngas (CO, 1.0 MPa) and then stirred vigorously at the required temperature for the appointed time. Upon completion of the reaction, the mixture was cooled to room temperature and the pressure was carefully released. The reaction mixture was extracted with diethyl ether (5 mL x 3). The ether fractions were combined and then analyzed by GC to determine the conversion of PhI (1-dodecane as an internal standard) and the selectivity for the carbonylative products (normalization method). The structures of the carbonylative products were further confirmed by GC-MS.
In the recycling experiments, the remaining slurry containing 2A, [Bmim]PF6, and the formed ammonium salt (Et3N×HI) after ether extraction was used directly without further treatment for the next run unless otherwise specified. In specified cases, the formed ammonium salt (Et3N×HI) was removed by washing the slurry with water (3 mL). The remaining IL phase was then collected for reuse. Because of the stoichiometric consumption of the base, Et3N (7.5 mmol) was added to each reaction. All manipulations were conducted in air.
1,3-Diphenylprop-2-yn-1-one (Table 3, entry 1). 1H NMR (δ, CDCl3): 8.23 (d, 2H, J=7.2 Hz), 7.65-7.68 (m, 2H), 7.59-7.63 (m, 1H), 7.45-7.52 (m, 3H), 7.38-7.42 (m, 2H); 13C NMR (δ, CDCl3): 178.08 (s, C=O), 136,89 (s), 134.18 (s), 133.12 (s), 130.85 (s), 129.62 (s), 128.73 (s), 128.67 (s), 120.14 (s), 93.16 (s), 86.91 (s).
3-Phenyl-1-(o-tolyl)prop-2-yn-1-one (Table 3, entry 2). 1H NMR (δ, CDCl3): 8.31 (d, 2H, J=8.8 Hz), 7.65-7.67 (m, 2H), 7.35-7.49 (m, 5H), 7.28 (d, 1H, J=7.2 Hz), 2.68 (s, 3H, CH3); 13C NMR (δ, CDCl3): 179.78 (s, C=O), 140.54 (s), 140.54 (s), 135.70 (s), 133.24 (s), 132.96 (s), 132.21 (s), 130.63 (s), 128.67 (s), 125.92 (s), 120.35 (s), 91.87 (s), 88.38 (s), 22.02 (s, CH3).
3-Phenyl-1-(m-tolyl)prop-2-yn-1-one (Table 3, entry 3). 1H NMR (δ, CDCl3): 8.04 (t, 2H, J=6.8 Hz), 7.68-7.70 (m, 2H), 7.39-7.51 (m, 5H), 2.45 (s, 3H, CH3); 13C NMR (δ, CDCl3): 178.28 (s, C=O), 138.53 (s), 136.91 (s), 135.03 (s), 133.09 (s), 130.79 (s), 129.81 (s), 128.71 (s), 128.55 (s), 127.16 (s), 120.22 (s), 92.91 (s), 87.03 (s), 21.38 (s, CH3).
3-Phenyl-1-(p-tolyl)prop-2-yn-1-one (Table 3, entry 4). 1H NMR (δ, CDCl3): 8.12 (d, 2H, J=8 Hz), 7.67-7.70 (m, 2H), 7.47-7.51 (m, 1H), 7.40-7.44 (m, 2H), 7.32 (d, 1H, J=8 Hz), 2.45 (s, 3H, CH3); 13C NMR (δ, CDCl3): 177.78 (s, C=O), 145.26 (s), 134.61 (s), 133.07 (s), 130.72 (s), 129.75 (s), 129.38 (s), 128.69 (s), 120.27 (s), 92.64 (s), 86.94 (s), 21.89 (s, CH3).
1-(4-Methoxyphenyl)-3-phenylprop-2-yn-1-one (Table 3, entry 5). 1H NMR (δ, CDCl3): 8.18-8.22 (m, 2H), 7.67-7.69 (m, 2H), 7.40-7.52 (m, 3H), 6.97-7.01 (m, 2H), 3.91 (s, 3H, CH3); 13C NMR (δ, CDCl3): 176.63 (s, C=O), 164.50 (s), 132.99 (s), 131.93 (s), 130.62 (s), 130.22 (s), 128.68 (s), 120.36 (s), 113.91 (s), 92.24 (s), 86.93(s), 55.64 (s, CH3).
1-(2-Methoxyphenyl)-3-phenylprop-2-yn-1-one (Table 3, entry 6). 1H NMR (δ, CDCl3): 8.10 (dd, 1H, J1=2 Hz, J2=7.6 Hz), 7.62-7.65 (m, 2H), 7.53-7.57 (m, 1H), 7.38-7.48 (m, 3H), 7.01-7.08 (m, 2H), 3.97 (s, 3H, CH3); 13C NMR (δ, CDCl3): 176.81 (s, C=O), 159.84 (s), 135.11 (s), 133.00 (s), 132.74 (s), 130.51 (s), 128.63 (s), 126.66 (s), 120.67 (s), 120.33 (s), 112.20 (s), 91.65(s), 89.19 (s), 55.96 (s, CH3).
3-Phenyl-1-(4-(trifluoromethyl)phenyl)prop-2-yn-1-one
(Table 3, entry 7). 1H NMR (δ, CDCl3): 8.33 (d, 2H, J=8 Hz), 7.79 (d, 2H, J=8 Hz), 7.69-7.71 (m, 2H), 7.50-7.54 (m, 1H), 7.42-7.46 (m, 2H); 13C NMR (δ, CDCl3): 176.76 (s, C=O), 139.35 (s), 133.17 (q, CF3, J=130 Hz), 133.25 (s), 131.25 (s), 128.83 (s), 125.73 (q, J=14.8 Hz), 124.91 (s), 122.20 (s), 119.64 (s), 94.52 (s), 86.58 (s).
1-(3-Nitrophenyl)-3-phenylprop-2-yn-1-one (Table 3, entry 9). 1H NMR (δ, CDCl3): 9.06 (t, 1H, J=2 Hz), 8.48-8.54 (m, 2H), 7.72-7.78 (m, 3H), 7.53-7.57 (m, 1H), 7.45-7.49 (m, 2H); 13C NMR (δ, CDCl3): 175.48 (s, C=O), 148.48 (s), 138.13 (s), 134.62 (s), 133.38 (s), 131.48 (s), 129.98 (s), 128.89 (s), 128.18 (s), 124.50 (s), 119.39 (s), 95.33 (s), 86.25 (s).
3-Phenyl-1-(thiophen-2-yl)prop-2-yn-1-one (Table 3, entry 10). 1H NMR (δ, CDCl3): 8.04 (d, 1H, J=3.5 Hz), 7.75 (d, 1H, J=4.5 Hz), 7.69 (d, 2H, J=7 Hz), 7.51 (t, 1H, J=7 Hz), 7.45 (t, 2H, J=7 Hz), 7.22 (t, 1H, J=4.5 Hz); 13C NMR (δ, CDCl3): 169.80 (s, C=O), 144.97 (s), 135.26 (s), 135.10 (s), 133.06 (s), 130.88 (s), 128.73 (s), 128.37 (s), 119.95 (s), 91.75 (s), 86.53 (s).
1-Nitro-4-(phenylethynyl)benzene (Table 3, entry 11). 1H NMR (δ, CDCl3): 8.20 (d, 2H, J=8.8 Hz), 7.65 (d, 2H, J=8.8 Hz), 7.54-7.58 (m, 2H), 7.37-7.40 (m, 3H); 13C NMR (δ, CDCl3): 147.00 (s, C=O), 132.28 (s), 131.87 (s), 130.27 (s), 129.31 (s), 128.57 (s), 123.65 (s), 122.13 (s), 94.74 (s), 87.59(s).
It was believed that the introduction of a positive charge with strong electron-withdrawing ability into L1 to afford L2 would change the coordination ability of the corresponding phosphines. According to the method reported in Refs. [30, 31, 32], the values of 1JSe-P in the 77Se isotopomer of the corresponding phosphine selenides in 31P NMR spectra can be used to evaluate the s-donor ability of a phosphine. As shown in Fig. 1, after replacement of one phenyl group in PPh3 (1JSe-P=729 Hz) with thiazolylimidazolyl, the phosphine obtained using L1 exhibited an increased value of 1JSe-P (744 Hz). This indicates an increase in the p-acceptor ability (i.e., weaker s-donor ability) originating from the relatively stronger electron-withdrawing ability of thiazolylimidazolyl than that of a phenyl group. Compared with L1 and PPh3, the highest 1JSe-P value (768 Hz) was observed for L2 because of the intense electron-withdrawing effect resulting from the positively charged thiazolylimidazolium ring on the phosphine fragment.
The different coordinating ability of L1 and L2 affected the structural parameters of the corresponding Pd complexes 1A and 2A, as depicted in Fig. 2. Both 1A and 2A possessed typical square-planar geometry, which is structurally similar to trans-PdCl2(PPh3)2 [33]. The Pd(II) (d8) center, lying at the center of inversion, was coordinated by two chloride ions and two phosphines in trans positions. However, 1A possessed distorted square-planar geometry, whereas ideal square-planar geometry with complete anti-symmetry was observed for 2A because the repulsive force of the two positive charges made the two pendant thiazolylimidazolium groups move apart as far as possible. The bond distances of Pd-P in 1A (0.2329(2), and 0.2326(2) nm) are slightly shorter than that in 2A (0.23316(1) nm), indicating the higher structural stability of 1A compared with 2A. In 1A, the interplanar dihedral angles (q) between the thiazolyl imidazolyl rings are 1.671(3)° and 14.685(3)° in comparison with those in 2A (both 13.303(2)°), which further indicates the torsional configuration of 1A.
Because of their different structures, the catalytic performance of the Pd complexes was also varied dramatically. The carbonylative coupling of phenylacetylene with iodobenzene under CuI-free conditions was selected as a model reaction to evaluate the catalytic performance of 1A and 2A; the results are summarized in Table 2. DMF was selected as the optimal solvent after screening DMF, THF, toluene, and acetone over 1A. With the same Pd concentration (0.1 mol%), 1A exhibited higher activity (turnover frequency (TOF) of 840 h−1) than that of 2A and PdCl2(PPh3)2 under mild reaction conditions (90℃, 1 h, and p(CO)=1.0 MPa) (entries 1 vs. 2 and 3). When the reaction temperature was increased from 90 to 120℃, a very high TOF of 3560 h−1 was obtained over 1A when Pd concentration was decreased to 0.05 mol% (entry 4). Under the same conditions, a relatively lower TOF of 2960 h−1 was obtained over 2A (entry 5). As observed in Fig. 1, the phosphine fragments in L1, L2, and PPh3 exhibited the s-donor ability (electron-rich character) in the order of PPh3 (1JSe-P=729 Hz) > L1 (1JSe-P=744 Hz) > L2 (1JSe-P=768 Hz). However, the catalytic efficiency over A1, A2, and PdCl2(PPh3)2 displayed the order: A1 (TOF=840 h−1) > PdCl2(PPh3)2 (TOF=770 h−1) > A2 (TOF=71 h−1), which was not consistently correlated to the s-donor ability of the involved phosphines. It was noted that, compared with PPh3, L1 contains additional coordination sites at S- and N-donor atoms besides that of the common phosphine fragment. Therefore, the available N,S-donors of the weak coordinating ligands in 1A are believed to be able to temporarily associate with Pd center to provide additional protection for the active Pd catalyst against deactivation and then dissociate to provide an unsaturated site for the subsequent insertion of the reaction substrate, which gave rise to the high activity of 1A. As for 2A and PdCl2(PPh3)2, the absence of N-donor atoms meant that the active Pd catalyst was unable to be protected effectively, leading to the depressed activity of these catalysts compared with that of 1A.
The scope of the reaction catalyzed by 1A was then examined (Table 3). The electronic effect and steric hindrance of the aryl substituents of the aryl iodides obviously affected the reaction rate. Because of the competing Sonogashira reaction, the corresponding side-products of aryl-substituted internal acetylenes were found universally but in very low yields. However, when 1-bromo-4-nitrobenzene was used in parallel, the carbonylative Sonogashira coupling barely proceeded. Conversely, when CO was replaced with N2, the Sonogashira cross-coupling occurred smoothly under the same reaction conditions. This suggests that the active Pd0 species formed in N2 atmosphere exhibited higher oxidative addition ability than that of the Pd0-CO species formed in CO atmosphere toward the substrates, especially aryl bromides with low reactivity.
Ionic complex 2A could be used as a catalyst with the RTIL solvent [Bmim]PF6, which immobilized the Pd complex catalyst to aid recovery and recycling (Table 4). Because of the mass-transfer limitation in the organic-IL biphasic reaction system, the reaction time was prolonged to 1.5 h during the recycling experiments. Ionic complex 2A was tightly locked in the IL phase; there was no loss of Pd detected by ICP-AES analysis of the combined organic phase after eight runs. The accumulated slurry of ammonium salts in [Bmim]PF6 formed after scavenging acid by Et3N could be removed by water to leave the clear IL phase for the next run because of the hydrophobicity of [Bmim]PF6 (entries 3 and 6 in Table 4). However, the TOFs of the 2A-[Bmim]PF6 system gradually decreased because of the unavoidable deactivation and mechanical loss of the catalyst after multiple reaction cycles.
After the introduction of positive charges, Pd complex 2A exhibited not only changes in complex configuration and structural stability, but also relatively lower catalytic efficiency toward the carbonylative Sonogashira reaction (TOF=2960 h−1). Compared with PdCl2(PPh3)2 and 2A with structural similarity, 1A ligated by L1 containing three types of the coordinating sites (P, N, and S donor atoms) exhibited the highest catalytic performance (TOF=3560 h−1), indicating the importance of the weak, reversible N- and S-ligation to the Pd center through intermolecular coordination in protecting the active Pd catalyst. It was also found that 2A in [Bmim]PF6 could be recycled eight times without detectable metal leaching.