1. Introduction
Ullmann-type
N-arylations have long been employed for the synthesis of various molecules and intermediates, suitable for a large number of biological, medicinal, and pharmaceutical applications, and, therefore, they belong to the most important transition metal-catalyzed cross-coupling reactions
[1,2,3][1][2][3]. However, copper-catalyzed
N-arylations have often been found to require harsh reaction conditions and displayed a poor tolerance towards other functional groups
[4]. In the past decade, the efficiency of copper-catalyzed Ullmann reactions has been considerably improved by applying novel copper sources, specific ligands, and additives
[5,6,7,8][5][6][7][8]. Several mild and sustainable methods have been developed for these reactions, including the
N-arylations of indoles and other heterocycles
[9,10,11,12,13][9][10][11][12][13]. The reaction pathway suggested for the Ullmann-type C–N heterocoupling reaction is indicated in
Scheme 1. The reaction of an amine with the active Cu
I species is followed by the oxidative addition of an aromatic halide, producing an aryl complex, in which the oxidation state of copper has changed to Cu
III. The final step is reductive elimination, which releases the coupling product, containing a new C–N bond, and regenerates the active Cu
I species, which is able to participate in another catalytic cycle.
Scheme 1. The reaction mechanism proposed for the Ullmann-type N-arylation.
Although the homogeneous
N-arylations catalyzed by copper complexes afforded high product yields and selectivities, problems associated with the recovery and recycling of these complexes initiated further studies focused on the development of more efficient and recyclable heterogeneous catalysts. Novel synthesis procedures based on the immobilization of catalytically active metal species or organometallic complexes on various support materials provided reusable heterogeneous catalysts with improved performances
[14]. Whereas recent studies on the Ullmann homocoupling reactions clearly indicated the predominance of heterogeneous Pd catalysts
[15], C–N heterocoupling reactions have still been preferably conducted by using more readily available and less toxic Cu-based catalysts
[16,17,18,19,20,21][16][17][18][19][20][21].
The subject of the current review is the presentation of the recent progress achieved in the Ullmann-type N-arylation reactions performed in heterogeneous systems, including the utilization of Cu and Pd-based catalysts and magnetic nanoparticles.
2. Magnetic Catalysts
Heterogeneous catalysts containing magnetite (Fe
3O
4) nanoparticles have been the subject of considerable attention because of their low cost, enhanced stability, large surface area, low toxicity, and good biocompatibility. More importantly, these catalysts can be readily removed from the reaction mixtures with an external magnet, which considerably improves their recyclability
[69,70,71,72][22][23][24][25]. In a recent study, Yousefi et al. reported results obtained for the catalytic application of Cu
I species immobilized on polyvinyl-alcohol (PVA) coated magnetic Fe
3O
4 nanoparticles in Ullmann-type N-arylations
[73][26]. In the first step of the synthesis procedure, an aqueous dispersion of as-prepared Fe
3O
4 nanoparticles
[74][27] was subjected to ultrasonic treatment, followed by the addition of PVA, and then the reaction mixture was left under stirring for 24 h at 80 °C. The resulting solid, PVA/Fe
3O
4, was impregnated with an aqueous solution of CuCl at room temperature for another 24 h, which afforded the product, CuCI-PVA/Fe
3O
4, with a copper loading of 0.44 mmol g
−1. Structural characterization of the sample was performed by FT-IR spectroscopy, field emission scanning electron microscopy (FESEM), EDX, vibrating sample magnetometry (VSM), ICP, XRD, and TEM. According to XRD patterns, the crystalline phase of Fe
3O
4 was unaffected by the PVA coating, and FT-IR spectra indicated that the interaction of PVA with the magnetic nanoparticles took place via the surface hydroxyl groups. FESEM images revealed the formation of quasi-spherical Fe
3O
4 nanoparticles with a mean diameter of 10–20 nm, and EDX spectra confirmed that efficient coordination of the copper ions on the Fe
3O
4-PVA surface took place. The characteristic signals of the Cu
I species also appeared on the XPS spectra at the binding energies 932.3 and 952.2 eV, assigned to the Cu 2p
3/2 and Cu 2p
1/2 signals, respectively. The sample was tested as a catalyst in the C–N heterocoupling reactions of various heterocyclic amines with aryl halides. The results are indicated in
Scheme 2.
Scheme 2. Ullmann heterocoupling reactions catalyzed by CuCl-PVA/Fe3O4.
For the reactions of iodo- and bromobenzenes, high product yields were obtained, irrespective of the electron donating or withdrawing character of their substituent. On the other hand, the transformations of chlorobenzenes afforded considerably lower yields of 55–60%. Recycling of the catalyst was investigated for the reaction of bromobenzene and morpholine. After completing the reaction, the catalyst was removed by an external magnet and reused. It was established that the catalyst remained efficient up to seven cycles (yield: 95–86%, avg. 91.6%, eight runs), and a hot filtration test confirmed that the CuCI-PVA/Fe3O4 sample may be considered a heterogeneous catalyst.
Eshghi et al. synthesized magnetic Cu nanorods and investigated their catalytic performance in the Ullmann heterocoupling reaction
[75][28]. Magnetic Fe
3O
4 particles, prepared by co-precipitation, were covered by a silica layer by applying the Stöber sol-gel method
[76][29]. The resulting solid, SiO
2/Fe
3O
4, was added to an ethanol solution of epibromohydrin (EP), and the mixture was subsequently stirred at 60 °C for 5 h. The dry precipitate was suspended in ethylenediamine (EN), followed by stirring at 60 °C for another 24 h. Further treatment of the product with another portion of epibromohydrin resulted in the formation of a solid material, EP/SiO
2/Fe
3O
4, containing a polydentate ligand bonded on the surface hydroxyl groups of the silica-coated magnetic nanoparticles
[77][30]. This was applied as the support material of Cu nanorods, produced from the ethanol solution of the precursor Cu(OAc)
2 under reflux conditions, followed by reduction with NaBH
4. The final product, Cu
0-EP/SiO
2/Fe
3O
4, was subjected to structural characterization. XRD patterns indicated that the crystal structure of Fe
3O
4 was retained after modification. According to TGA curves, the amount of the organic linker on the surface of the magnetic nanoparticles was 0.6 mmol g
−1. TEM images gave evidence that Cu nanorods of 10 nm were distributed on the surface of spherical SiO
2/Fe
3O
4 particles with an average diameter of 20 nm. The Cu loading of the product, 2.59 mmol g
−1, was determined by ICP analysis. The catalytic test reactions, the
N-arylations of various heterocycles with aryl halides, were performed by using 8 mol% of catalyst, K
2CO
3 as a base, and DMF as a solvent. The results are displayed in
Scheme 3.
Scheme 3. N-Arylations of heterocycles, catalyzed by Cu0-EP/SiO2/Fe3O4.
For the reactions of various heterocycles, including benzimidazole, indole, pyrazole, and 1,2,4-triazole, substantial product yields were obtained by varying the reaction time between 12–20 h. A recycling test was accomplished for the N-arylation of pyrazole with 4-methoxyiodobenzene, and no appreciable deactivation was observed (yield: 98–90%, avg. 94%, five runs). The Cu loading of the recovered catalyst was 1.82 mmol g−1 after the fifth cycle, indicating a low amount of leaching, and TEM images confirmed that no structural modification of the catalyst occurred under reaction conditions.
The preparation and the catalytic application of a hydrotalcite-based magnetic CuI nanocatalyst have been recently reported by Rajabzadeh and Khalifeh et al.
[78][31]. In the synthesis procedure, the surface of a magnetic hydrotalcite, HT/Fe
3O
4, was functionalized by 3-chloropropyltrimethoxysilane via covalent attachment, performed in toluene under reflux conditions
[79][32]. The resulting solid was treated by a tricationic ionic linker (TIL) obtained from epichlorohydrine and 1-methylimidazole. This linker was dissolved in ethanol, and the reaction mixture was kept at boiling temperature under stirring for 24 h. The dispersion of the modified magnetic hydrotalcite, TIL/HT/Fe
3O
4, and CuI in acetonitrile was subsequently heated at 50 °C under an Ar atmosphere for 8 h, which afforded the final product, 7.7 wt% CuI-TIL/HT/Fe
3O
4. The anchoring of TIL on the surface of the magnetic support material was revealed by FT-IR spectra. XRD analysis indicated the formation of Fe
3O
4 nanoparticles on the surface of hydrotalcite, and the presence of Cu
I in CuI-TIL/HT/Fe
3O
4 was also confirmed. According to FESEM and TEM images, the plate-like morphology of HT/Fe
3O
4 was maintained after functionalization and the immobilization of CuI. The catalytic properties of CuI-TIL/HT/Fe
3O
4 were investigated in the
N-arylations of various heterocycles with aryl halides, and the results are shown in
Scheme 4.
Scheme 4. Heterocoupling reactions of N-heterocycles with aryl halides, promoted by CuI-TIL/HT/Fe3O4.
The coupling products were formed with good yields, and the transformations of aryl halides containing both Br and I substituents were found to be selective for the C–I bond. Recycling studies performed for the reaction of p-methoxyiodobenzene with benzimidazole gave evidence that the catalyst could be used in six consecutive cycles without an appreciable decrease of activity (yield: 97–92%, avg. 95.8%, six runs). A hot filtration test also indicated the heterogeneous nature of the catalyst.
The preparation and the catalytic application of a magnetically recoverable Pd nanocatalyst have been reported by Ghorbani-Vaghei et al.
[80][33]. The synthesis procedure was based on the production of amidoxime-functionalized Fe
3O
4 particles, which were applied as the support material of the active Pd species. Magnetic Fe
3O
4 nanoparticles were obtained by chemical co-precipitation of Fe
III and Fe
II ions, followed by the addition of NH
4OH, which resulted in the formation of surface hydroxyl groups
[81][34]. After dispersing the Fe
3O
4 powder in toluene, triethoxyethylcyanide was added, and the reaction was completed at 100 °C for 48 h under an argon atmosphere. Subsequent treatment with an aqueous hydroxylamine solution produced amidoxime (AO) groups on the surface of the magnetic nanoparticles. The resulting solid (AO/Fe
3O
4) was dispersed in acetonitrile, followed by the addition of PdCl
2 and stirring at room temperature for 10 h and the subsequent reduction of the precursor with hydrazine hydrate. The final product, Pd
0-AO/Fe
3O
4, had a Pd loading of 1.83 wt%, obtained from ICP-AES and EDS. FESEM and TEM images revealed spherical morphologies for both the functionalized Fe
3O
4 and the Pd nanoparticles, for which the mean diameters were 10–15 and 3 nm, respectively. The catalytic performance of the Pd
0-AO/Fe
3O
4 sample was investigated in the Ullmann-type
N-arylations of indoles with iodobenzenes by using 0.1 mol% catalyst, triethylamine as a base and DMF as a solvent. The results are summarized in
Scheme 5.
Scheme 5. N-Arylations catalyzed by 1.83% Pd0-AO/Fe3O4.
The coupling reaction proved to be selective for the formation of the N-aryl-substituted product and proceeded with excellent yields for aryl iodides containing both electron-donating and electron-withdrawing substituents. Recycling of the Pd0-AO/Fe3O4 catalyst revealed satisfactory stability (yield: 96–82%, avg. 90%, seven runs). However, it should be noted that these data were obtained for the Suzuki reaction of 4-methyliodobenzene with phenylboronic acid, and no related data were disclosed for the N-arylations.
Another recyclable magnetic Pd nanocatalyst has been fabricated by Hajipour et al.
[82][35]. The magnetic Fe
3O
4 nanoparticles, prepared by co-precipitation
[80][33], were coated with a silica layer upon reaction with tetraethyl orthosilicate (TEOS). In the next step, 3-iodopropyltrimethoxysilane was added in a nitrogen atmosphere to the Fe
3O
4/SiO
2 material, dispersed in toluene, and the mixture was heated under reflux conditions for 24 h. The following treatment of the product with cysteine and K
2CO
3 was performed under similar conditions in acetonitrile. The formation of iodo-functionalized groups on the surface of the silica layer ensured the grafting of cysteine moieties on the magnetic nanoparticles. Finally, palladium acetate was immobilized on the magnetic support in ethanol, which resulted in the formation of the product, Pd
0-cysteine/Fe
3O
4. ICP analysis revealed that the Pd loading of the sample was 0.47 wt%. Experimental evidence for the immobilization of cysteine moieties on the surface of the magnetic nanoparticles was obtained by FT-IR spectroscopy. XRD spectra displayed characteristic signals corresponding to the silica-coated magnetic nanoparticles and the Pd
0 species, whereas TEM images indicated the formation of monodispersed, spherical Pd nanoparticles with an average diameter of 14 nm. The catalytic activity of Pd
0-cysteine/Fe
3O
4 was examined in the
N-arylations of aryl halides with various amines, and the coupling products were obtained with pronounced yields (
Scheme 6).
Scheme 6. C–N heterocoupling reactions catalyzed by Pd0-cysteine/Fe3O4.
A heterogeneous pathway was confirmed by a hot filtration test, indicating that no further reaction progress took place after the removal of the magnetic Pd catalyst, which was in accordance with the result of ICP analysis, indicating that the amount of dissolved Pd was 0.03 ppm.