Mono- and double carbonylation of iodobenzene in the presence of reusable supported palladium catalysts

2.1.1 Preparation of CAT-1/1:

200 mg [BMIM][BF₄] and 0.02 mmol (20.7 mg) Pd₂(dba)₃.CHCl₃ were dissolved in a mixture of 2 ml acetonitrile and 2 ml THF. The mixture was stirred for 15 min at room temperature. Then, 550 mg silica [Kieselgel 60 (0.040–0.063 mm), Merck, pre-treated by heating for 5 h at 250°C] was added under stirring and the resulting mixture was stirred for 24 h. The solvents were removed in vacuo, and the catalyst was dried at 35°C in vacuo for 3 h and was stored under argon until use. Palladium content of the catalyst: 0.40% (determined by ICP).

2.1.2 Preparation of CAT-1/2:

200 mg [BMIM][BF₄] and 0.04 mmol (8.7 mg) Pd(OAc)₂ were dissolved in a mixture of 2 ml acetonitrile and 2 ml THF. The mixture was stirred for 15 min at room temperature. Then 550 mg silica [Kieselgel 60 (0.040–0.063 mm), Merck, pre-treated by heating for 5 h at 250°C] was added under stirring, and the resulting mixture was stirred for 24 h. The solvents were removed in vacuo, and the catalyst was dried at 35°C in vacuo for 3 h and was stored under argon until use. The palladium content of the catalyst: 0.49% (determined by ICP).

2.1.3 Preparation of CAT-1/3:

200 mg [BMIM][BF₄] and 0.04 mmol (15.3 mg) dichloro-bis(1-butyl-3-methylimidazol-2-ylidene)palladium(II), (Pd(BMIM)₂Cl₂), prepared by a known method [28]) were dissolved in a mixture of 2 ml acetonitrile and 2 ml THF. The mixture was stirred for 15 min at room temperature. Then 550 mg silica (Kieselgel 60 (0.040-0.063 mm), Merck, pre-treated by heating for 5 h at 250°C) was added under stirring, and the resulting mixture was stirred for 24 h. The solvents were removed in vacuo, and the catalyst was dried at 35°C in vacuo for 3 h and was stored under argon until use. Palladium content of the catalyst: 0.36% (determined by ICP).

2.1.4 Preparation of CAT-1/4:

Two hundred milligrams [BMIM][BF₄] and 0.04 mmol (46.2 mg) Pd(PPh₃)₄ were dissolved in a mixture of 2 ml acetonitrile and 2 ml THF. The mixture was stirred for 15 min at room temperature. Then 550 mg silica [Kieselgel 60 (0.040–0.063 mm), Merck, pre-treated by heating for 5 h at 250°C] was added under stirring and the resulting mixture was stirred for 24 h. The solvents were removed in vacuo, and the catalyst was dried at 35°C in vacuo for 3 h and was stored under argon until use. Palladium content of the catalyst: 0.55% (determined by ICP).

2.1.5 Preparation of CAT-1/5:

Two hundred milligrams [BMIM][BF₄], 0.04 mmol (46.2 mg) Pd(PPh₃)₄ and 0.08 mmol (21.0 mg) PPh₃ were dissolved in a mixture of 2 ml acetonitrile and 2 ml THF. The mixture was stirred for 15 min at room temperature. Then 550 mg silica (Kieselgel 60 (0.040-0.063 mm), Merck, pre-treated by heating for 5 h at 250°C) was added under stirring and the resulting mixture was stirred for 24 h. The solvents were removed in vacuo, and the catalyst was dried at 35°C in vacuo for 3 h and was stored under argon until use. Palladium content of the catalyst: 0.54% (determined by ICP).

2.1.6 Preparation of CAT-2/1:

The silica-supported ionic liquid phase SILP-2 was prepared as described previously [26].

A solution of 8.95 μmol (9.3 mg) Pd₂(dba)₃.CHCl₃ in 2 ml acetonitrile, and 2 ml THF was stirred for 15 min. Then, 500 mg of SILP-2 was added, and the resulting mixture was stirred for 24 h at room temperature. After evaporation of the solvents, the catalyst was dried at 35°C in vacuo for 3 h. Palladium content of the catalyst: 0.29% (determined by ICP).

2.1.7 Preparation of CAT-2/2:

A solution of 17.9 μmol (4 mg) Pd(OAc)₂ in 2 ml EtOH was stirred for 15 min. Then, 500 mg of SILP-2 was added, and the resulting mixture was stirred for 24 h at room temperature. The catalyst was filtered, washed with ethanol (2 ml) and diethylether (2 ml), and dried at 35°C in vacuo for 3 h. Palladium content of the catalyst: 0.77% (determined by ICP).

2.1.8 Preparation of CAT-2/3:

A solution of 17.9 μmol (3.2 mg) PdCl₂ in 2 ml EtOH was stirred for 15 min. Then, 500 mg of SILP-2 was added, and the resulting mixture was stirred for 24 h at room temperature. The catalyst was filtered, washed with ethanol (2 ml) and diethylether (2 ml), and dried at 35°C in vacuo for 3 h. Palladium content of the catalyst: 0.26% (determined by ICP).

2.1.9 Preparation of CAT-2/4:

A solution of 20 μmol (3.5 mg) PdCl₂ in 1 ml acetonitrile was refluxed in an inert atmosphere for 2 h. The solvent was evaporated in vacuo, and the residue was dissolved in 2 ml EtOH, and the mixture was stirred for 15 min. Then 400 mg of SILP-2 was added, and the resulting mixture was stirred for 24 h at room temperature. The catalyst was filtered, washed with ethanol (2 ml) and diethylether (2 ml) and dried at 35°C in vacuo for 3 h. Palladium content of the catalyst: 0.38% (determined by ICP).

2.1.10 Preparation of CAT-2/5:

A solution of 14.8 μmol (15.3 mg) Pd₂(dba)₃.CHCl₃ in 2 ml EtOH was stirred for 15 min. Then, 500 mg of SILP-2 was added, and the resulting mixture was stirred for 5 days at room temperature. The catalyst was filtered, washed with ethanol (2 ml), and diethylether (2 ml) and dried at 35°C in vacuo for 3 h. Palladium content of the catalyst: 0.44% (determined by ICP).

2.1.11 Preparation of Pd/silica:

A total of 0.02 mmol (20.7 mg) Pd₂(dba)₃.CHCl₃ was dissolved in a mixture of 2 ml acetonitrile, and 2 ml THF. The mixture was stirred for 15 min at room temperature. Then 550 mg silica [Kieselgel 60 (0.040-0.063 mm), Merck, pre-treated by heating for 5 h at 250°C] was added and the resulting mixture was stirred for 24 h. The solvents were removed in vacuo and the catalyst was dried at 35°C in vacuo for 3 h and was stored under argon until use. Palladium content of the catalyst: 0.55% (determined by ICP).

2.2.1 Catalytic reactions at atmospheric pressure:

In a typical experiment, a solution containing the palladium catalyst (with 3.6 μmol Pd-content) was placed in a Schlenk-tube. Under argon, 0.2 mmol (22.5 μl) iodobenzene (1), 0.5 mmol (44 μl) morpholine (2a), 0.25 mmol (35 μl) triethylamine, and 1 ml DMF was added, and the atmosphere was changed to carbon monoxide. The reaction was conducted for 3 h at 100°C. The reaction mixture was analysed by gas chromatography.

2.2.2 Catalytic reactions at elevated pressure:

In a typical experiment, the catalyst (containing 3.6 or 6 μmol Pd) was placed in a stainless steel autoclave. Aryl iodides (1 and 6a-f) (0.2 mmol), the amine (2a-e) (0.5 mmol), 0.25 mmol (35 μl) or 0.7 mmol (100 μl) triethylamine and solvent (1 ml) were transferred into it under an inert atmosphere. It was charged with carbon monoxide (30 bar or 5 bar) and heated with stirring in an oil bath at 100°C. After cooling to room temperature, the liquid phase was removed with a syringe. The reaction mixture was analysed by gas chromatography and the catalyst was reused.

3.1.1 Aminocarbonylation in DMF

In case of catalysts obtained with physically adsorbed ionic liquid (SILP-1), a substantial loss of [BMIM][BF₄] (>95% of the original amount, based on the ¹H NMR analysis of the reaction mixtures) was observed in the first runs of aminocarbonylations carried out in DMF.

Interestingly, phosphine-free palladium-catalysts (CAT-1/1 – CAT-1/3) showed better performance than Pd/silica (Figure 2) especially concerning recyclability, despite the greater loss of palladium observed in the reactions of the former catalysts (Table 2, entries 1, 2, and 4). The beneficial effect of the ionic liquid might be due to a stabilisation of the initially formed palladium nanoparticles at the beginning of the reaction. Imidazolium ionic liquids are frequently used as template, stabiliser, and immobilising agent for catalytic transition-metal nanoparticles to prevent aggregation [35]. TEM image (Figure 3) of spent CAT-1/1 proves the presence of palladium nanoparticles and shows some residual ionic liquid despite the great loss of [BMIM][BF₄] in the first cycle.

Figure 2:

Selectivity and recyclability of catalysts immobilised on silica or SILP-1 in aminocarbonylation carried out in DMF [0.2 mmol 1, 0.5 mmol 2a, 0.25 mmol Et₃N, catalyst (with 3.6 μmol Pd-content), 1 ml DMF, 30 bar CO, 100°C].

Figure 3:

TEM images of spent CAT-1/1 (left) and CAT-2/1 (right) after two runs (with 8-h reaction time).

When comparing the performance of catalysts obtained from different precursors, [Pd₂(dba)₃.CHCl₃ (CAT-1/1), Pd(OAc)₂ (CAT-1/2) and Pd(BMIM)₂Cl₂ (CAT-1/3)], CAT-1/1 showed the lowest activity. The application of the Pd-NHC precursor led to a catalyst (CAT-1/3) with high activity but lower selectivity to the α-ketomide (3a). In addition, a considerable amount of side product 5 was formed when the catalyst was reused.

In our group, an enhanced activity of palladium catalysts with excess phosphine ligands, with a P/Pd ratio of 6–10, was observed before in aminocarbonylations using ionic liquids as solvents [36]. As a consequence, the efficiency of Pd(PPh₃)₄ was also tested (CAT-1/4). After initial high activity, a great decrease in product yields was observed. The latter could be reduced to some extent by an increase in the phosphine/palladium ratio (Figure 2, CAT-1/5). The loss of activity may be caused by the leaching of PPh₃ into the reaction mixture (90% of the original amount in the first run, according to ¹H NMR measurements). It should be mentioned however, that the great loss of ligand was not accompanied by an unusually high loss of palladium (Table 2, entry 5), that was found to be almost the same as in case of CAT-1/1. The great drop of activity after the first run may be explained by the formation of inactive palladium complexes after the loss of phosphine ligands.

All of the catalysts obtained from silica or SILP-1 showed high selectivity towards the α-ketoamide (3a), which even increased in the recycling experiments with the exception of CAT-1/3. A similar phenomenon was observed before by Li and Yuan [11] in double carbonylation with palladium supported on a cross-linked functional polymer.

A higher activity and better recyclability of the catalyst (CAT-2/1), prepared by a similar method but immobilised on the support with grafted ionic liquid (SILP-2), was observed both in experiments with shorter and longer reaction times (Figure 4). Selectivity towards the α-ketoamide (3a) was found to be a bit lower in the first five runs but it was improved during recycling and complete conversion of iodobenzene (1) to product 3a was observed in the 9^th and 10^th cycles (reaction time: 8 h). Better performance upon reuse may be partly due to a lower loss of palladium (Table 2, entries 6, 7) compared to CAT-1/1 and CAT-1/2.

Figure 4:

Selectivity and recyclability of catalysts immobilised on SILP-2 in aminocarbonylation carried out in DMF [0.2 mmol 1, 0.5 mmol 2a, 0.25 mmol Et₃N, catalyst (with 3.6 μmol Pd-content], 1 ml DMF, 30 bar CO, 100°C).

Also, an increase in the reaction time resulted in an excellent recyclability and led to a 100% yield of product 3a, even in the 10th run. A better conversion can be achieved in longer reaction times, and this can be explained by a dissolution/reprecipitation mechanism proposed by Köhler [37]. Even if soluble palladium complexes or palladium nanoparticles are involved in the catalytic cycle, they can return to the surface of the support in the form of nanoparticles after the reaction is completed. The reprecipitation of palladium is supported by the smaller leaching of the catalysts in longer reactions (Table 2). The higher activity of the recycled catalyst (CAT-2/1) in the longer experiments can also be proved. Mixtures obtained after a 1-h reaction time in the second run with spent catalysts of the 3-h and 8-h experiments showed 38% and 47% conversions, respectively (Table 3, entries 1, 2). A smaller difference in the results was observed in the second runs of a similar experiment performed in the presence of CAT-2/5 in acetonitrile (entries 3,4).

The TEM image of the spent catalyst again shows the presence of palladium nanoparticles (Figure 3). To obtain information about the homogeneous or heterogeneous nature of the catalytic reaction, hot filtration and mercury poisoning tests were carried out in the reaction of CAT-2/1 (Table 4). The catalytic mixture was filtered after 30 min. One half of the mixture was heated further in a CO atmosphere, and the other half was treated similarly but in the presence of mercury.

According to the results, a low conversion of iodobenzene (1) could be observed after the removal of the heterogeneous catalyst, and the reaction could be completely stopped by the addition of mercury (Table 4, entries 3, 4). This shows that a small amount of catalytically active palladium nanoparticles had leached to the reaction mixture.

For comparison, a similar experiment was carried out with catalyst CAT-1/3 obtained from a Pd-NHC precursor (Table 4, entries 1, 2). Although in this case some further reaction could be observed in the filtrate, even in the presence of mercury, conversion was much lower than without the addition of the catalyst poison. This supports the presence of some complexes and nanoparticles in the reaction mixture.

According to previous reports, although the presence of palladium-carbene complexes was observed in 1,3-dialkylimidazolium-chloride solutions of Pd(II) salts, ionic liquids with [BF₄]^- or [PF₆]^- anions facilitated the formation of palladium nanoparticles [38]. In addition, palladium nanoparticles were shown to be formed in ionic liquids under thermal decomposition conditions without the addition of a reducing agent, starting from Pd(dba)₂ and Pd(OAc)_2, as well as from palladium-carbene complexes [39]. Therefore, although the role of palladium-carbene complexes as catalysts cannot be excluded, they may also serve as reservoirs for catalytically active palladium nanoparticles under carbonylation conditions.

Catalysts obtained from SILP-2 in EtOH (CAT-2/2 – CAT-2/5) exerted high activity in the conversion of iodobenzene. Acceptable results were obtained upon recycling, but a considerable amount of side-product 5 was formed in DMF in each case.

3.1.2 Aminocarbonylation in acetonitrile

The efficiency of another polar solvent, acetonitrile, was also investigated in reactions with some of the catalysts immobilised either on SILP-1 (CAT-1/1, CAT-1/5) or SILP-2 (CAT-2/3, CAT-2/4, CAT-2/5). Good conversion of iodobenzene (1) was achieved with the first group of catalysts in the first runs, but unfortunately, a great loss of activity was observed upon recycling, even carrying out the carbonylation for a longer reaction time. CAT-2/3, CAT-2/4 and CAT-2/5 showed lower activity in acetonitrile than in DMF (Figure 5) but compounds 3a and 4a were the only products in the reaction mixtures, and the catalysts could be reused with only a small loss of activity. When the reaction time was increased to 8 h, the conversion of iodobenzene dropped below 90%, but only in the 9th run in the presence of CAT-2/5. Contrary to the reactions in DMF, there is no appreciable change in the selectivity of aminocarbonylation in the recycling experiments. As another advantage, the amount of leached palladium is considerably lower here than in case of reactions carried out in DMF (Table 2, entry 8).

Figure 5:

Selectivity and recyclability of catalysts in aminocarbonylation carried out in acetonitrile [0.2 mmol 1, 0.5 mmol 2a, 0.25 mmol Et₃N, catalyst (with 3.6 μmol Pd-content), 1 ml acetonitrile, 30 bar CO, 100°C].

Hot filtration and mercury poisoning tests (Table 4, entries 5, 6) showed that the conversion of iodobenzene (1) was very slow after the removal of the heterogeneous catalyst, and it could be completely stopped by the addition of mercury. These results are in good accordance with the small amount of leached palladium detected during aminocarbonylation of this catalyst (Table 2, entry 8).

3.1.3 Aminocarbonylation in toluene

Some previous reports show the excellent activity of immobilised palladium catalysts in toluene [8, 9, 13, 19, 22], so the possibility of the use of this solvent was also taken into account. As it was observed before in the case of CAT-1/1 and CAT-2/1 [26], the use of an apolar solvent reduced catalytic efficiency considerably in all of the catalysts investigated (Figure 6). At the same time, under these conditions, the main product was amide 4a despite the use of a high CO pressure (30 bar). This probably can be explained by the fact that, although the formation of amides is thought to occur via neutral complexes, ionic species are involved in the catalytic cycle of double carbonylation. As a consequence, polar solvents may enhance the rate of the latter reaction by stabilisation of the ionic intermediates in the presence of apolar ones, and monocarbonylation is favoured.

Figure 6:

Selectivity and recyclability of catalysts in aminocarbonylation carried out in toluene [0.2 mmol 1a, 0.5 mmol 2a, 0.25 mmol Et₃N, catalyst (with 3.6 μmol Pd-content), 1 ml toluene, 30 bar CO, 100°C].

Despite the fact that no loss of ionic liquid was observed, even when catalysts were prepared with SILP-1, recyclability is poor with the exception of CAT-2/3, where there is only a moderate loss of activity in the first three runs.

3.1.4 Aminocarbonylation at lower pressure

Aminocarbonylation of iodobenzene (1) with morpholine (2a) was attempted using lower carbon monoxide pressure [5 bar (Figure 7) and 1 bar (Figure 8)]. A good selectivity towards the α-ketoamide (3a) was observed at 5 bar, and double carbonylation was found to take place even under atmospheric conditions. The catalysts could be recycled with some loss of activity when moderate pressure was used, but a great drop of activity was observed in the atmospheric reactions when catalyst recycling was attempted. The formation of side-product 5 could be detected in reaction mixtures obtained with catalysts prepared in EtOH.

Figure 7:

Selectivity and recyclability of catalysts in aminocarbonylation carried out in DMF at 5 bar [0.2 mmol 1, 0.5 mmol 2a, 0.25 mmol Et₃N, catalyst (with 3.6 μmol Pd-content), 1 ml DMF, 5 bar CO, 100°C].

Figure 8:

Selectivity and recyclability of catalysts in aminocarbonylation carried out in DMF at atmospheric pressure [0.2 mmol 1, 0.5 mmol 2a, 0.25 mmol Et₃N, catalyst (with 3.6 μmol Pd-content), 1 ml DMF, 1 bar CO, 100°C].