Marie
Genelot
a,
Nicolas
Villandier
a,
Anissa
Bendjeriou
b,
Patchareeporn
Jaithong
a,
Laurent
Djakovitch
*a and
Véronique
Dufaud‡
*b
aUniversité de Lyon, CNRS, UMR 5256, IRCELYON, Institut de recherches sur la catalyse et l'environnement de Lyon, 2 avenue Albert Einstein, F-69626 Villeurbanne, France. E-mail: laurent.djakovitch@ircelyon.univ-lyon1.fr; Fax: +33 472445399; Tel: +33 472445381
bUniversité de Lyon, CNRS, UMR 5182, Laboratoire de Chimie, Ecole Normale Supérieure de Lyon, 46 allée d’Italie, F-69364 Lyon Cedex 07, France. E-mail: vdufaud@ens-lyon.fr; Fax: +33 472728860; Tel: +33 472728860
First published on 22nd March 2012
A promising route for the double carbonylation of aryl iodide derivatives with secondary and primary amines to produce α-ketoamides is described using covalently immobilized palladium complexes on SBA-15 silica. Adequate adjustments of the different reaction parameters (temperature, CO pressure, nature of base, solvent, substrate…) to achieve optimal catalyst performance were made using PdCl2(PPh2)2@SBA-15 as catalytic system. High conversions (up to 80%) and excellent selectivities (up to 96%) for the double carbonylated α-ketoamide products were obtained using K2CO3 as base, MEK or DMF as solvent and a 1 mol% [Pd] catalyst. We also demonstrated that two other palladium hybrid mesoporous materials can be alternatively used, namely PdCl2(PCy2)2@SBA-15 and PdCl2(PNP)@SBA-15, without loss of activity and selectivity. Finally, catalyst recycling of PdCl2(PPh2)2@SBA-15 showed that the catalyst could be reused for up to 3 cycles without affecting catalyst performance.
Since its simultaneous discovery in 1982 by Ozawa, Yamamoto and co-workers14 and by Kobayashi and Tanaka,15 the palladium-catalysed double carbonylation reaction of aryl halides with secondary amines to prepare α-ketoamides has been the subject of numerous studies12b,16 (Scheme 1). This reaction is also a useful synthetic means to produce α-ketoacids17 and α-ketoesters18 by substitution of the amine by a water molecule or alcohol. While intensive studies on carbonylation reactions have contributed to establish the mechanism of the reaction,12b,18e,19 most of the works have been realised in the presence of homogeneous catalysts making the separation/purification and the recovery of expensive metal catalysts and ligands very difficult, or even impossible. These environmental and economic concerns associated with the contamination issues of biologically active compounds are particularly important when dealing with large scale-synthesis and industrial processes.
Scheme 1 Palladium-catalysed double carbonylation of aryl halide to α-ketoamides. |
Although few examples have been reported for monocarbonylation, reports on double carbonylation reaction catalysed by heterogeneous palladium materials are still rare. In 1997, Yan et al. used a silica-supported polytitazane–palladium (Ti–N–Pd) complex for the double carbonylation of phenyl halide in the presence of diethylamine.20 The authors have shown that there is an optimal temperature (100 °C) below which the conversions are slow and above which the selectivity for α-ketoamide decreases. Moreover, it was noticed that a high CO pressure and the use of polar solvent enhanced both the conversion and the selectivity to the α-ketoamide formation. The supported catalyst could be reused ten times without noticeable decrease in catalytic activity and selectivity.
In 2009, a catalytic system composed of Pd/C and a homogenous phosphine, the triphenylphosphine (PPh3), was used by Liu et al. to synthesize different α-ketoamides with yields ranging from 64 to 87% in tetrahydrofuran (THF) and using 1,4-diazabicyclo[2.2.2]octane (DABCO) as the base.21 However, when the Pd/C was recycled, only 17% yield for α-ketoamide was obtained. This result was ascribed to Pd leaching phenomena, the Pd/C being likely not the active species but the source of palladium in solution.
Finally, a commercial catalyst composed of a palladium complex [Pd(PPh3)4] immobilized in CatCart™ cartridges has been used to carry out the double carbonylation of iodobenzene in the presence of amines in a microfluidic-based flow reactor.22 With this methodology, the highest yields for α-ketoamides were obtained at 80 °C, using 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) as the base and primary amine as the nucleophile. However with this base, side reactions were observed as DBU could react with the acyl palladium complex formed in situ from the iodobenzene precursor and CO.
In the present contribution, we report on the applicability of covalently immobilized palladium complexes on SBA-15 silica to the double carbonylation reaction of aryl iodides with amines to afford α-ketoamides. The influence of different reaction parameters on activity and chemoselectivity (solvent, temperature, CO pressure, aryl halide and nucleophile nature) was systematically examined since adequate adjustments of these latter may be necessary to achieve optimal catalyst performance. Finally, considering catalyst stability and reusability of primal interest, a preliminary study of the performance of a [Pd]@SBA-15 type catalyst over several batch reaction cycles was also investigated.
Amines, bases, aryl iodides and solvents were obtained commercially and used without further purification. The palladium supported on active carbon Degussa type E101 NO/W comes from Sigma-Aldrich (Pd/C 4.7 wt% on dry basis, 52% water; denoted in Table 4 by PdII/C(ALD)) and palladium supported on active carbon (Pd/C 5.0 wt% on dry basis; denoted in Table 4 by Pd0/C(ALD)) comes also from Sigma-Aldrich. The palladium supported on active carbon type E105 CA/W is available at Evonik (Pd/C 5.0 wt%, 55% water; denoted hereafter PdII/C(EVO)). The palladium supported on zeolite NaY (denoted in Table 4 by PdII/NaY) and that on silica (denoted in Table 4 by PdII/SiO2) were prepared following the procedures described by Djakovitch et al.23 Analytical thin layer chromatography (TLC) was performed on Fluka Silica Gel 60 F254. All reactions were monitored by GC using dodecane as an internal standard. Response factors of the products with regards to dodecane were determined by calibration with known quantities of the corresponding pure substances. The experimental error was estimated to be Δrel = ±5%. GC analyses were performed on a HP 4890 chromatograph equipped with a FID detector, a HP 6890 autosampler and a HP-5 column (cross-linked 5% phenyl-methylsiloxane, 30 m × 0.25 mm id × 0.25 μm film thickness) with nitrogen as carrier gas and with the following program: 2 min at 60 °C followed by 25 °C min−1 ramp to 170 °C, then 2 min at 170 °C followed by 35 °C min−1 ramp to 240 °C, then 5 min at 240 °C followed by 30 °C min−1 ramp to 300 °C and then 8 min at this temperature. GC-MS analyses were performed on a Shimadzu GC–MS-QP2010S equipped with a Sulpelco SLB-5MS column (95% methylpolysiloxane + 5% phenylpolysiloxane, 30 m × 0.25 mm × 0.25 μm) with helium as carrier gas and with the same temperature program as for GC. Ionization was made by electronic impact at 70 eV. Purification of products was accomplished by flash chromatography performed at a pressure slightly greater than atmospheric pressure using silica (Macherey-Nagel Silica Gel 60, 230–400 mesh) with the indicated solvent system. Liquid NMR spectra were recorded on a BRUKER AC-250 spectrometer. All chemical shifts were measured relative to residual 1H or 13C NMR resonances in deuterated solvents: CDCl3, δ 7.26 ppm for 1H and 77.16 ppm for 13C. Coupling constants are expressed in Hertz (Hz). Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet). Melting points were determined on a Schorpp-Gerätetechnik MPM-H2 in open capillary tubes and are uncorrected. High Resolution Mass Spectrometry (HRMS) was carried out on a Brucker MicroTOF-QII mass spectrometer using electrospray ionization.
Scheme 2 Targeted palladium molecular precursors. |
Their covalent grafting onto organic-free SBA-15 silica was carried out in degassed toluene at 70 °C overnight leading to three Pd hybrid materials referred to as PdCl2(PPh2)2@SBA-15 (3.1 wt% [Pd]), PdCl2(PCy2)2@SBA-15 (1.2 wt% [Pd]) and PdCl2(PNP)@SBA-15 (3.5 wt% [Pd]) from the anchoring of, respectively, 1, 2 and 3 molecular precursors. The resulting hybrids were thoroughly characterised using molecular and surface techniques, which are presented in ESI†: X-ray powder diffraction (XRD) at small angles, elemental analysis, thermogravimetric analysis, multi-nuclear NMR spectroscopy and nitrogen sorptions. In all cases, well structured solids with hexagonally ordered mesophases were obtained with intact coordination spheres around the metal center as evidenced by XRD, carbon and phosphorus NMR and elemental analyses (ESI†, Fig. S1–S8 and Tables S1 and S2). In particular, the absence of free phosphine peaks in the 31P NMR spectra suggests that the Pd–P bond was unaffected by grafting.
Solid state 29Si NMR spectroscopy provides further information about the nature of the link to the surface. The 29Si CP-MAS NMR spectra of the modified oxides all showed peaks in the spectral region associated with tertiary silicon atoms, from −40 to −65 ppm, which indicates unambiguously the presence of organosiloxane moieties. Furthermore, multiple peaks are obtained from each of the modified oxides, suggesting different degrees of linkage of the organosiloxane group with the silica surface. Specifically, all three modified oxides exhibit principally T1 and T2 sites, that is, an anchoring of the organic ligand via one or two Si–O–Si bonds. This suggests that both phosphine linkers in PdCl2(PPh2)2@SBA-15 and PdCl2(PCy2)2@SBA-15 hybrids were involved, at least, in one covalent bond with the support resulting in chelation through the silica surface. In addition, peaks assignable to Q2, Q3, and Q4 silicon sites related to the siliceous framework are also discernible in all three spectra.
Scheme 3 Model reaction studied: double carbonylation of iodobenzene with diethylamine. |
To improve this preliminary result, a study on the influence of the solvent was performed (Table 1). In all solvents tested, high selectivities, varying from 92 to 97%, have been obtained for the double carbonylation product (4). However, conversion of iodobenzene remained relatively low (57% in the best case) whether using polar solvents such as methyl tert-butyl ether (MTBE), 1,4-dioxane and ethyl acetate (EtOAc) (Table 1, entries 2–4) or aromatic solvents such as anisole and toluene (Table 1, entries 5 and 6). Although polar aprotic solvents such as dimethylformamide (DMF) and methylpyrrolidone (NMP) demonstrated greater conversion, up to 88% (Table 1, entries 8 and 9), methyl ethyl ketone (MEK) was chosen as solvent later in the project. Indeed, even if the conversion was lower than with DMF and NMP (Table 1, entry 10), still reaching 70% conversion of iodobenzene, MEK can be considered as a green solvent29 with lower toxicity. Moreover, MEK can be easily removed from the reaction medium by evaporation facilitating work-up and purification procedures.
Entry | Solvent | Conv. PhIb (%) | Select. 4/5b (%) | v max c/mmol h−1 gPd−1 |
---|---|---|---|---|
a Iodobenzene (1 mmol), diethylamine (4 mmol), PdCl2(PPh2)2@SBA-15 (1 mol% [Pd]), solvent (5 mL), CO pressure (40 bar), 60 °C, 22 h. b Determined by GC. c Determined at maximum of the slope of the kinetic curve. | ||||
1 | THF | 46 | 95 | 61.5 |
2 | MTBE | 49 | 92 | 38.5 |
3 | 1,4-Dioxane | 55 | 96 | 67.7 |
4 | EtOAc | 35 | 97 | 27.8 |
5 | Toluene | 57 | 95 | 32.9 |
6 | Anisole | 51 | 97 | 31.3 |
7 | MeCN | 65 | 97 | 94.0 |
8 | DMF | 88 | 97 | 55.7 |
9 | NMP | 87 | 96 | 48.3 |
10 | MEK | 70 | 96 | 57.1 |
The effects of bases on the double carbonylation of iodobenzene with diethylamine were examined using MEK as solvent and PdCl2(PPh2)2@SBA-15 (1 mol% [Pd]) as catalyst (Table 2). Organic bases such as pyridine, triethylamine, diisopropylamine and sodium acetate (NaOAc) showed moderate conversion of iodobenzene while keeping high selectivity for the double carbonylation product (Table 2, entries 1–4). In contrast, a very poor selectivity of 6% for (4) was achieved when using piperidine as a base (Table 2, entry 5). This result can be attributed to the nucleophilic properties of piperidine which reacts directly with iodobenzene thus reducing the selectivity of the carbonylation reaction. This effect can also explain, to some extent, the moderate selectivity observed with 1,1,3,3-tetramethylguanidine (TMG) (Table 2, entry 6). On the other hand, superior conversions and selectivities were obtained with DBU and DABCO as bases (Table 2, entries 7 and 8). In general, inorganic bases such as K3PO4 and K2CO3 exhibited good conversions and selectivities except for Na2CO3 (Table 2, entries 9–12). Even though similar results in terms of conversion and selectivity were obtained with DBU and K2CO3, DBU was discarded for further study given that Skoda-Földes et al. have reported that DBU can react with the palladium catalyst to give by-product.22
Entry | Base | Conv. PhIb (%) | Select. 4/5b (%) | v max c/mmol h−1 gPd−1 |
---|---|---|---|---|
a Iodobenzene (1 mmol), diethylamine (2 mmol), base (2 mmol), PdCl2(PPh2)2@SBA-15 (1 mol% [Pd]), MEK (5 mL), CO pressure (40 bar), 60 °C, 22 h. b Determined by GC. c Determined at maximum of the slope of the kinetic curve. | ||||
1 | Pyridine | 52 | 94 | 57.8 |
2 | Triethylamine | 45 | 96 | 52.9 |
3 | Diisopropylamine | 53 | 92 | 58.9 |
4 | NaOAc | 50 | 95 | 83.1 |
5 | Piperidine | 86 | 6 | 68.9 |
6 | TMG | 89 | 81 | 48.2 |
7 | DBU | 79 | 96 | 67.1 |
8 | DABCO | 69 | 97 | 56.4 |
9 | K3PO4·H2O | 74 | 97 | 52.1 |
10 | K3PO4 | 71 | 97 | 65.8 |
11 | K2CO3 | 78 | 96 | 57.6 |
12 | Na2CO3 | 54 | 96 | 75.2 |
Further optimisation of reaction conditions was carried out by studying the impact of the temperature and CO pressure on the conversion and selectivity (Table 3). The conversion of iodobenzene was greatly affected by the reaction temperature. Indeed, the conversion increased from 13 to 100% when raising the temperature from 40 to 80 °C with a selectivity of up to 96% for α-ketoamides (Table 3, entries 1–5). Generally, under our reaction conditions, the higher CO pressure enhanced the conversion (Table 3, entries 6–8). Surprisingly, the conversion and selectivity for α-ketoamides decreased slightly when the CO pressure was raised from 40 to 50 bar (Table 3, entries 8 and 9), suggesting that there is a CO pressure which must not be exceeded.
Entry | Temp./°C | Pressure/bar | Conv. PhIb (%) | Select. 4/5b (%) | v max c/mmol h−1 gPd−1 |
---|---|---|---|---|---|
a Iodobenzene (1 mmol), diethylamine (2 mmol), K2CO3 (2 mmol), PdCl2(PPh2)2@SBA-15 (1 mol% [Pd]), MEK (5 mL), 22 h. b Determined by GC. c Determined at maximum of the slope of the kinetic curve. d Not determined due to low PhI conversions. | |||||
1 | 40 | 40 | 13 | 95 | ndd |
2 | 50 | 40 | 38 | 91 | ndd |
3 | 60 | 40 | 78 | 96 | 57.6 |
4 | 70 | 40 | 92 | 91 | 86.4 |
5 | 80 | 40 | 100 | 96 | 89.0 |
6 | 60 | 10 | 61 | 94 | 42.4 |
7 | 60 | 20 | 68 | 93 | 40.6 |
8 | 60 | 40 | 78 | 96 | 57.6 |
9 | 60 | 50 | 75 | 93 | 37.6 |
The influence of several heterogeneous catalytic systems on the double carbonylation of iodobenzene was also investigated using 1 mol% [Pd] (Table 4). In addition to PdCl2(PPh2)2@SBA-15, two other palladium complexes immobilized onto SBA-15 type silica, with various coordination spheres, were tested: the first, PdCl2(PCy2)2@SBA-15, which contains monodentate phosphine, provides a ligand system which is both more electron-rich and sterically hindered; the second, PdCl2(PNP)@SBA-15, which bears a chelating diphosphine linker, provides a stronger metal–ligand bonding which may affect the stability towards leaching. The performance of these silica supported palladium hybrid materials was compared to traditional ligand-free supported palladium using readily available catalysts such as Pd/C, Pd/NaY and Pd/SiO2.
Entry | Catalyst | Conv. PhIb (%) | Select. 4/5b (%) | v max c/mmol h−1 gPd−1 |
---|---|---|---|---|
a Iodobenzene (1 mmol), diethylamine (2 mmol), K2CO3 (2 mmol), catalyst (1 mol% [Pd]), MEK (5 mL), 60 °C, 40 bar, 22 h. b Determined by GC. c Determined at maximum of the slope of the kinetic curve. d Not determined due to low PhI conversions. | ||||
1 | PdCl2(PPh2)2@SBA-15 (3.1 wt% [Pd]) | 78 | 96 | 57.6 |
2 | PdCl2(PCy2)2@SBA-15 (1.2 wt% [Pd]) | 65 | 97 | 41.8 |
3 | PdCl2(PNP)@SBA-15 (3.8 wt% [Pd]) | 80 | 95 | 44.2 |
4 | PdII/C(EVO) (5 wt% [Pd]) | 15 | 95 | ndd |
5 | PdII/C(ALD) (4.7 wt% [Pd]) | 13 | 96 | ndd |
6 | Pd0/C(ALD) (5 wt% [Pd]) | 22 | 92 | ndd |
7 | PdII/NaY (5 wt% [Pd]) | 17 | 100 | ndd |
8 | PdII/SiO2 (10 wt% [Pd]) | 23 | 98 | ndd |
Globally, palladium containing mesostructured hybrid materials exhibited high conversion and excellent selectivity for the double carbonylation product (4). When the PdCl2(PCy2)2@SBA-15 catalyst was used, the conversion of iodobenzene decreased slightly from 78 to 65% (Table 4, entries 1 and 2), while in the presence of PdCl2(PNP)@SBA-15 similar conversion and selectivity were achieved (Table 4, entries 1 and 3). However, in all cases, the silica supported palladium complexes outperformed in activity the classical supported palladium particles which exhibited very poor conversion with a maximum of 23% despite a high selectivity (Table 4, entries 4–8).§
In order to get more insights on the PdCl2(PPh2)2@SBA-15 catalyst performances, we plot the evolution of the conversion and the selectivity toward the α-ketoamide versus the time for the carbonylative coupling reaction of iodobenzene with diethylamine under the optimized conditions (1 mmol iodobenzene, 2 mmol diethylamine, 2 mmol K2CO3, 1 mol% [Pd] PdCl2(PPh2)2@SBA-15, 60 °C, 40 bar, 5 mL MEK, 48 h). Fig. 1 shows that the selectivity decreased slightly when conversion and/or time increased: while 100% selectivity was reached at 40% conversion (i.e. after 10 h of reaction) it decreased to 85% when the conversion reached 90% (i.e. at 48 h run). On the other hand, the activity of the catalyst seems not affected, indeed this modification of selectivity occurs when the catalyst reaches its highest activity (i.e. 57.6 mmol h−1 gPd−1) that is achieved after an initiation period of 2.5 h. Moreover, almost quantitative conversion is achieved after 48 h giving 76% yield toward the expected α-ketoamide.
Fig. 1 Evolution of the conversion and the selectivity versus the time. |
Amines used in the double carbonylation have a considerable influence on both the yield and selectivity of the reaction. Of nine secondary amines tested, the more voluminous amines such as diisopropylamine and N,N-dibenzylamine were respectively either totally inactive (Table 5, entry 1) or weakly active giving a low yield of carbonylated α-ketoamide and amide products (Table 5, entry 2). High yields and selectivities for α-ketoamides were achieved with less sterically demanding secondary amines (Table 5, entries 3–8). Among these secondary amines, 1,2,3,4-tetrahydroisoquinoline (THIQ) showed the highest reactivity with an almost quantitative conversion of iodobenzene and a selectivity of 91% for the desired α-ketoamide. The reaction between iodobenzene and THIQ leads to the formation of two conformers, in a ratio 70/30, which explains the duplication of signals observed in 1H and 13C NMR. Two conformers (60/40) are also obtained with the N-benzylmethylamine (NMR analyses are given in the ESI†). Surprisingly, selectivity with TMG was totally reversed in favour of amide despite its high basicity (Table 5, entry 9).
Entry | Amine (pKb) | Conv. PhIb (%) | Yield of α-ketoamidesc (%) | Yield of amidesb (%) | Selec. 4/5b (%) |
---|---|---|---|---|---|
a Iodobenzene (1 mmol), amine (2 mmol), K2CO3 (2 mmol), PdCl2(PPh2)2@SBA-15 (1 mol% [Pd]), MEK (5 mL), 60 °C, 40 bar, 48 h. b Determined by GC. c Determined by GC (isolated yield). d Reaction carried out in DMF. | |||||
1 | Diisopropylamine (2.95) | 0 | 0 | 0 | 0 |
2 | N,N-Dibenzylamine (5.24) | 56 | 18 | 19 | 49 |
3 | Diethylamine (3.02) | 95 | 74 (71) | 13 | 85 |
4 | Dipropylamine (3.00) | 95 | 71 (69) | 14 | 85 |
5 | Piperidine (2.78) | 100 | 83 (81) | 8 | 90 |
6 | Morpholine (5.64) | 99 | 74 (69) | 13 | 85 |
7 | N-Benzylmethylamine (4.15) | 99 | 76 (74) | 7 | 84 |
8 | 1,2,3,4-Tetrahydroisoquinoline (4.70) | 99 | 84 (83) | 9 | 91 |
9 | TMG (0.04) | 100 | 0 | 96 (94) | 0 |
10 | Anilined (9.34) | 100 | 0 | 81 | 0 |
11 | Benzylamined (4.66) | 98 | 65 (63) | 20 | 76 |
12 | n-Butylamined (3.32) | 99 | 52 (50) | 19 | 73 |
The reactivity of primary amines was also studied. To avoid the reaction between the primary amines and MEK, DMF was used as solvent. Under our reaction conditions, weakly basic primary amines such as aniline did not provide double carbonylation products (Table 5, entry 10). In fact, only the amide was obtained with 81% yield. In contrast, in the presence of more basic primary amines such as benzylamine and n-butylamine, α-ketoamide was the major product with selectivity of 76 and 73% respectively. In the case of n-butylamine, the formation of the Schiff base product from the reaction of the primary amine with the keto group of the α-ketoamide produced during the course of the reaction was evidenced by GC with 28% yield. This side reaction has already been reported in homogenous catalysis12a,15 and explains the low yield for α-ketoamide obtained with n-butylamine. When benzylamine was used, this secondary reaction did not occur, benzylamine being less basic than n-butylamine.
The introduction of substituents into the aryl group is known to modify the reactivity of aryl halides. Table 6 shows the influence of both position and nature of different substituents on the double carbonylation reaction of phenyl iodide derivatives. It is worth noting that this study was carried out in the presence of 1,2,3,4-tetrahydroisoquinoline as nucleophile since this amine showed the best results in terms of activity and selectivity. The introduction of a methoxy group in the ortho position of iodobenzene strongly decreased the selectivity for α-ketoamide (Table 6, entries 1 and 2). This result can be ascribed to the existence of some interactions between the palladium and the methoxy group in the oxidative addition intermediate, which may reduce the coordination rate of the CO molecule. In contrast, when the methoxy group was placed in the meta or para position, the yields and selectivities increased significantly (Table 6, entries 3 and 4). Indeed, in these cases, the palladium and substituent cannot interact with each other, because they are too far apart. Moreover, as the methoxy group is an activating group, the selectivity with 3-iodoanisole and 4-iodoanisole was slightly better than that of iodobenzene. Introducing a methyl substituent at the ortho and para positions of iodobenzene resulted in an increase in selectivity for the α-ketoamide formation (Table 6, entries 1 vs. 5 and 6) although a slight decrease in conversion was observed with 2-iodotoluene (Table 6, entries 1 and 5). Finally, 1-iodonaphthalene readily underwent double carbonylation with full conversion and a selectivity of 90% for the double carbonylation product (Table 6, entries 1 and 7). All the products of this last study are obtained in the form of two conformers.
Entry | Phenyl iodide | Conv. PhIb (%) | Yield of α-ketoamidesc (%) | Yield of amidesc (%) | Selec. 4/5b (%) |
---|---|---|---|---|---|
a Phenyl iodide (1 mmol), 1,2,3,4-tetrahydroisoquinoline (2 mmol), K2CO3 (2 mmol), PdCl2(PPh2)2@SBA-15 (1 mol% [Pd]), MEK (5 mL), 60 °C, 40 bar, 48 h. b Determined by GC. c Determined by GC (isolated yield). | |||||
1 | Iodobenzene | 99 | 82 (81) | 8 | 91 |
2 | 2-Iodoanisole | 92 | 62 (58) | 23 | 73 |
3 | 3-Iodoanisole | 100 | 85 (86) | 7 | 93 |
4 | 4-Iodoanisole | 99 | 82 (80) | 3 | 97 |
5 | 2-Iodotoluene | 96 | 83 (84) | 2 | 98 |
6 | 4-Iodotoluene | 99 | 83 (83) | 5 | 95 |
7 | 1-Iodonaphthalene | 100 | 82 (77) | 9 | 90 |
The remarkable activity of the PdCl2(PPh2)2@SBA-15 catalyst made the study of its recycling particularly interesting (Fig. 2). The recycling was examined for the double carbonylation reaction of iodobenzene with diethylamine under optimised reaction conditions (1 mmol iodobenzene, 2 mmol diethylamine, 2 mmol K2CO3, 1 mol% [Pd] PdCl2(PPh2)2@SBA-15, 60 °C, 40 bar, 5 mL MEK, 48 h). The procedure was performed as follows: at the completion of the 1st run (48 hours) using a fresh catalyst, the solid was separated by centrifugation from the reaction mixture, washed three times with the reaction solvent (MEK) and then engaged in a new catalytic cycle under the same reaction conditions. This procedure was repeated up to a total of 3 runs. The results shown in Fig. 2 revealed that PdCl2(PPh2)2@SBA-15 exhibited high recyclability giving for each run close conversions of iodobenzene of 91, 94, 96% for, respectively, the first, the second and the third run. However, from Fig. 2, it is observed that the catalyst activity increases slightly from the first run to the others as the rate at the highest catalyst activity increases from 57.2 mmol h−1 gPd−1 to 103.4 and 131.5 for the second and third run, respectively. Moreover, the initiation period is also reduced from 2.5 h for the first run down to 1.5 h then 0.5 h for the second and third run respectively. This is more probably related to modifications of the catalyst during the first run as observed elsewhere.§ These modifications led to the formation of small well dispersed palladium particles through the well documented dissolution/redeposition phenomenon. Upon recycling, higher palladium dispersion can be reached leading thus to higher activity and shorter initiation period.30
Fig. 2 Reuse of PdCl2(PPh2)2@SBA-15 for the double carbonylation reaction of iodobenzene with diethylamine. |
Additionally, it was observed that both the yield and selectivity toward the α-ketoamide remained stable over the different cycles.
These results showed that PdCl2(PPh2)2@SBA-15 is not only an efficient and active catalyst, but also exhibits relatively high stability for the double carbonylation reaction.
Footnotes |
† Electronic supplementary information (ESI) available: Palladium catalysts characterization, reaction products characterization after double carbonylation of aryl iodides. See DOI: 10.1039/c2cy00516f |
‡ Present address: Université de Lyon, CNRS, UMR 5265, Laboratoire de Chimie, Catalyse et Procédés de Polymérisation, CPE Lyon, 43, boulevard du 11 Novembre 1918, 69622 Villeurbanne Cedex, France. |
§ (1) Additional experiments were done in order to compare the efficiency of the heterogeneous catalyst (PdCl2(PPh2)2@SBA-15) versus the homogeneous precursor used to synthesise the catalytic material i.e.Cl2Pd[PPh2CH2CH2Si(OEt)3]2 (see ESI†, Table S3). (2) As a personal communication from the authors. While the used catalysts were not fully characterized in the case of double carbonylation reactions, characterizations were performed in the case of monocarbonylation reactions. We assume that the same observations could be made in the present case given that close experimental conditions were used. Generally, no structural modification of the support was observed after the first run nor several runs (up to 5 runs) of the catalyst, as observed from XRD and TEM experiments. Thus the hexagonal structure of SBA-15 remains stable. On the other hand, small Pd particles can be observed depending on the nature of the grafted precursor. This was the case with PdCl2(PPh2)2@SBA-15 and PdCl2(PCy2)2@SBA-15. These particles are formed through the leaching phenomenon that was observed in the case of carbonylative coupling reactions of iodobenzene with various C- and N-nucleophiles. Hot filtration experiments demonstrated that leached palladium was responsible for the catalytic activity observed, as expected. |
This journal is © The Royal Society of Chemistry 2012 |