An advancement in the synthesis of nano Pd@magnetic amine-Functionalized UiO-66-NH2 catalyst for cyanation and O-arylation reactions

The magnetic MOF-based catalytic system has been reported here to be an efficient catalyst for synthesis of benzonitriles and diarylethers of aryl halides under optimal conditions. The MOF catalyst was built based on magnetic nanoparticles and UiO-66-NH2 which further modified with 2,4,6-trichloro-1,3,5-triazine and 5-phenyl tetrazole at the same time and the catalyst structure was confirmed by various techniques. This new modification has been applied to increase anchoring palladium into the support. Furthermore, the products’ yields were obtained in good to excellent for all reactions under mild conditions which result from superior activity of the synthesized heterogeneous catalyst containing palladium. Also, the magnetic property of the MOF-based catalyst makes it easy to separate from reaction mediums and reuse in the next runs.


Results and discussion
Catalyst preparation. The magnetic amine-Functionalized UiO-66-NH 2 , a porous scaffold with large surface area, was synthesized to immobilize palladium nanoparticles. At first, the magnetic nanoparticles were functionalized by acrylic acid (AA) to prepare the surface for growth of UiO-66-NH 2 (Fig. 2). Subsequently, the amino groups of MOF were modified by 2,4,6-trichloro-1,3,5-triazine (TCT) and 5-phenyl tetrazole which reacted with K 2 PdCl 4 to provide an efficient heterogeneous catalyst (Fig. 3). Briefly, one chloride of TCT was substituted by amine groups of magnetic UiO-66-NH 2 support. Then, the remaining two chlorides of TCT were substituted by amine groups of 5-phenyl tetrazole. It was expected, the 3D network structure was formed by substitution of all three chlorides in TCT which could effectively immobilized palladium ions into the resulting support by coordination with NNN pincer-like groups.
Characterization of Pd/MOF Catalyst. FT-IR spectroscopy. The characteristic absorption bands at 3440, 1093, and 576 cm −1 are related to O-H stretching vibration Si-O and Fe-O stretching vibration, and 3-(trimethoxysilyl)propylmethacrylate (MPS) sharp peaks can be seen at 1714 and 1408 cm −1 for C=O and C=C stretching bonds which confirms the Fe 3 O 4 surface was coated successfully with MPS ( Fig. 4) 36 . The PAA illustrates characteristic absorption bands at 2941, 1718, 1460, and 1411 cm −1 assigned to the CH 2 stretching and bending modes, C=O and C-O stretching vibrations in COOH group, respectively 37 . The FT-IR spectrum of magnetic UiO-66 is shown the characteristic vibrational modes including 1573 cm −1 for COO− asymmetric mode, 1390 and 1433 cm −1 for COO− symmetric modes, 3361 cm −1 for NH 2 symmetric mode, 3467 cm −1 for NH 2 asymmetric mode, and 1259 cm −1 for C−N vibrational mode 17 . The magnetic UiO-66-NH 2 was modified by TCT and 5-phenyl tetrazole at the same time and the strong characteristic band of triazine ring skeleton is observed at around 1570 cm −138 and 5-phenyl tetrazole ring show C=N and N=N stretching vibrations at 1450 and 1530 cm −139 which these bands overlap with the peaks of the magnetic UiO-66-NH 2 . The peak at 3435 cm −1 shows two branches related to stretching mode of NH 2 groups which transformed to one broad peak related to NH groups after modification. Thus, it is a strong evidence for functionalization of the magnetic UiO-66-NH 2 .
XRD patterns. XRD patterns of UiO-66-NH 2 , magnetic UiO-66-NH 2 , magnetic amine-Functionalized UiO-66-NH 2 and Pd 0 @ magnetic amine-Functionalized UiO-66-NH 2 are presented in Fig. 5. In the patterns, all diffraction peaks are similar to UiO-66 pattern which reported by Wang et al. 16 . These patterns confirm successful synthesis of UiO-66-NH 2 and the peaks of Fe 3 O 4 was not apparent in the patterns owing to applying low amount of Fe 3 O 4 in this procedure. Therefore, the crystalline nature of the magnetic Zr-MOF is preserved after modification. In accordance with Fig. 5d, Pd 0 @ magnetic amine-Functionalized UiO-66-NH 2 shows the both peaks of UiO-66-NH 2 and Pd nanoparticles confirming immobilization of Pd nanoparticles into the magnetic amine-Functionalized UiO-66-NH 2 support.
Morphology analysis. The morphology of the magnetic amine-Functionalized UiO-66-NH 2 are presented in FE-SEM images (Fig. 6a, b). The FE-SEM images illustrate the particles have a mean diameter 125 nm with cubic structure which are similar to pervious works 17 . In addition, the magnetic amine-Functionalized UiO-66-NH 2 are uniform without aggregation. TEM images of the prepared MOF show good agreement with other literatures and can confirm the Fe 3 O 4 core of the obtained UiO-66-NH 2 ( Fig. 6c-f).    7a) and TEM (Fig. 7b) images of Pd 0 @magnetic amine-Functionalized UiO-66-NH 2 catalyst are also provided. The size distribution of the catalyst was calculated based on its FESEM image and the mean particle size is around 135 nm (Fig. 7c).
The TEM images of Pd-catalyst have been shown in two magnifications 50 and 100 nm in which the presence of Pd nanoparticles into the polymeric network can be approved (Fig. 7b). The FESEM of recycled Pd 0 @magnetic amine-Functionalized UiO-66-NH 2 catalyst after six runs for cyanation reaction is presented in Fig. 7d and the small changes in the shape of recycled catalyst are observed. Also, the presence of Pd nanoparticles and other components are confirmed by Elemental mapping and EDX analysis (Fig. 8a, b).
Thermal properties. TGA analysis was performed under nitrogen atmosphere at a heating rate 10 °C min −1 to investigate thermal decomposition of samples. It is noteworthy, samples were dried overnight in a vacuum oven at 80 °C before analysis. The thermograms of magnetic UiO-66-NH 2 and magnetic amine-Functionalized UiO-66-NH 2 are demonstrated in Fig. 9. The first weight loss around 100 °C is related to degradation of water molecules trapping into MOF pores. Furthermore, the second is ascribed to decomposition of organic groups of samples such as H 2 BDC-NH 2 , TCT and 5-phenyl tetrazole. The difference between two curves shows the relative amounts of TCT and 5-phenyl tetrazole grafted to magnetic UiO-66-NH 2 which is about 26% (w/w).
Brunauer-Emmett-Teller (BET) surface area analysis. The N 2 adsorption-desorption data have been summarized in Table 1. The BET specific surface areas of magnetic amine-Functionalized UiO-66-NH 2 and Pd 0 @ magnetic amine-Functionalized UiO-66-NH 2 are 828 and 664 m 2 g −1 , respectively. In accordance with results, the presence of Pd nanoparticles are confirmed by decreasing BET specific surface area, total pure volume, and mean pore diameter data of Pd 0 @magnetic amine-Functionalized UiO-66-NH 2 compared with magnetic amine-Functionalized UiO-66-NH 2 [40][41][42] . Cynation over Pd/MOF catalyst. The catalytic activity of the MOF-based catalyst was investigated through C-CN coupling reaction, after conformation of the catalyst structure with some techniques. To achieve optimal conditions, different reaction parameters were screened involving various amounts of catalyst, bases, and solvents by bromobenzene and K 4 Fe(CN) 6 as a green cyanide source (Table 2). At first, bases including K 2 CO 3 , DABCO, KOH, KHCO 3 , and Et 3 N were tested in the presence of 2.0 mg of MOF based-catalyst and it was found that K 2 CO 3 and KOH can facilitate the reaction among them but the KOH was the best base and chosen for the reaction ( Table 2, entry 1-5). Subsequently, various solvents such as NMP, H 2 O, EtOH, DMSO, DMF, and Toluene were used and the effect of solvents was investigated on the reaction conversion ( Table 2, entry 5-10). Based on solvent screening, polar aprotic solvents such as NMP and DMF were more effective rather than polar protic solvents such as H 2 O (Table 2, entry 6), and NMP was more favored solvent (Table 2, entry 5). Then, the reaction was performed with 1.0, 2.0, 5.0 mg of MOF based-catalyst and catalyst without Pd nanoparticles under the same reaction conditions and also the model reaction was tested in the presence of 2.0 mg of Pd 0 @magnetic UiO-66-NH 2 and desired product was obtained in 62% yield ( Table 2, entry 17) compared with Pd 0 @magnetic  With having the best reaction parameters in hand, the applicability of the reported protocol was studied for versatile aryl halides bearing both electron-deficient and electron-rich functional groups to provide target products with good to excellent yields. The aryl halides were reacted with K 4 Fe(CN) 6 in the presence of KOH and 2.0 mg of MOF based-catalyst and NMP under 100 °C as optimal conditions as presented in Table 3. Generally, aryl iodides and aryl bromides with electron-deficient groups such as nitro groups in meta and para positions have shown excellent yields (Table 3, entry 7 and 16) and less yields were seen in products having electron-rich groups such as methyl groups (Table 3, entry 5 and 12). Also, in this case aryl chlorides have demonstrated good yields of benzonitriles which their reactions proceed with longer times. The aryl chloride bearing electron-deficient www.nature.com/scientificreports/ (Table 3, entry 21-23) have shown higher yields in comparison to electon-rich aryl chloride and they need higher temperature and longer reaction time to provide desired yield (Table 3, entry 24). Then, the heterocyclic compounds including 2-and 4-bromopyridine were examined and the 2-cyanopyridine (94%) and 4-cyanopyridine (91%) were obtained in excellent yields (Table 3, entry 18 and 19).
The chemoselectivity of this protocol was tested by 1-chloro-2-iodobenzene and 1-chloro-4-iodobenzene which they converted to 2-chlorobenzonitrile and 4-chlorobenzonitrile with excellent yields, 94% and 96%, respectively (Table 3, entry 2 and 3). Moreover,1-bromo-4-iodobenzene and 1-bromo-4-chlorobenzene provided 4-bromobenzonitrile and 4-chlorobenzonitrile, 93% and 90%, respectively ( Table 3, Table 4. In accordance with Table 4, various bases such as K 2 CO 3 , K 3 PO 4 , KOH, NaHCO 3 , and Et 3 N were applied and the KOH was the best base (Table 4, entry 1-5). In the next step, solvents were screened and DMSO and H 2 O were the optimal solvents rather than DMF, NMP, CH 3 CN, and Toluene (Table 4, entry 3 and entry 6-10). But water was chosen as the optimal solvent because of its green nature. Afterwards, the reaction was tested at room temperature, 80 °C and 120 °C (Table 4, entry [11][12][13]. When the reaction took place at 80 °C instead of 100 °C, the desired product was acquired without remarkable change in the yield. To optimize amounts of the catalyst, different amounts of the MOF based-catalyst was used (1.0, 2.0, 5.0 mg, and none) and catalyst without Pd nanoparticles which the 2.0 mg of the catalyst was gave the best yield and use of 5.0 mg of the catalyst showed no significant effect on the yield (Table 4, entry [14][15][16][17]. Also, the model reaction was tested in the presence of 2.0 mg of Pd 0 @magnetic UiO-66-NH 2 and desired product was obtained in 67% yield (   (Table 5). To gain different O-arylated derivatives, several electron-rich and electron-deficient substrates were tested. Firstly, the scope of aryl halides was studied and Iodobenzene and bromobenzene illustrated higher activity in comparison with chlorobenzene because of lower polarizability of C-Cl bond related to oxidationaddition step of palladium insertion in reaction mechanism (Table 5). Subsequently, the electron-rich aryl halides such as methyl and methoxy groups on them showed good yields and electron-deficient aryl halides showed higher yields. In the next step, phenols were examined and electron-rich phenols with methyl and methoxy groups were provided diaryl ethers in excellent yields and electron-deficient phenols having nitro groups depicted lower yields because of decreasing the nucleophilicity of phenols. In the case of 2-nitrophenol which has greater steric hindrance compared with 4-nitrophenol, it was provided the expected products but in low yields. Also, 1-naphthol and 2-naphthol were tested and they generated the desired products in good yields under longer reaction times (See Supplementary information for 1 H and 13 C NMR spectral data).
Catalyst recycling. The reuseability of the catalyst was examined through optimized reaction conditions between iodobenzene and K 4 Fe(CN) 6 as model raw materials. After each run, the catalyst was collected by an  www.nature.com/scientificreports/ external magnetic field and the isolated catalyst was washed with methanol and water, dried completely, and applied for next run. This MOF-based catalyst was used over six successive runs and the isolated yields were shown in Table 6. The results confirm that this catalytic system remained still active during six runs of cyanation reaction without loss of catalytic activity. Also, the recycleability of the catalyst was tested for synthesis of diaryl ethers between iodobenzene and phenol as model reaction and the catalyst was reused over five successive runs based on the mentioned procedure (Table 7). After final runs, the loading amounts of Pd were investigated by ICP-OES analysis and they were 0.72 mmol g −1 for Cyanation and 0.71 for O-arylation.

Conclusion
In summary, the presented article described the preparation and application of the palladium MOF-based catalyst. The magnetic catalyst includes UiO-66-NH 2 which has been modified with 2,4,6-trichloro-1,3,5-triazine and 5-phenyl tetrazole to support palladium nanoparticles. The shape and morphology of the modified UiO-66-NH 2 was confirmed by FESEM and TEM analysis and they corresponded with pervious literatures. This catalytic system has shown very efficient activity for both syntheses of cyanoarenes and diaryl ethers in mild reaction conditions with good to excellent yields.

Experimental section
Catalyst preparation. Preparation of the magnetic nanoparticles. Immobilization of acrylic acid on Fe 3 O 4 @SiO 2 Microspheres. Firstly, the Fe 3 O 4 @SiO 2 nanoparticles were prepared through co-precipitation method and treated with 3-(trimethoxysilyl)propylmethacrylate (MPS) as reported in previous works 43 . Afterwards, the surface of Fe 3 O 4 @SiO 2 @MPS was functionalized with acrylic acid (AA): 3.0 ml of acrylic acid was added to 0.50 g of Fe 3 O 4 @SiO 2 @MPS in 20 ml deionized water. Then, the flask was charged with 10 mg of AIBN after degassed under N 2 atmosphere and refluxed for 24h. Then, the obtained magnetite nanoparticles were collected by an external magnet and washed with deionized water/methanol three times, and dried in a vacuum oven at 60 °C for 12h to provide the magnetic PAAs.
Preparation of the magnetic UiO-66-NH 2 . The magnetic UiO-66-NH 2 was synthesized based on literature reported by Wang et al. 16 . In a round bottom flask, 0.  Preparation of the Pd 0 @ magnetic amine-Functionalized UiO-66-NH 2 . A round bottom flask was charged by the magnetic UiO-66-NH 2 (1.0 g) and dry THF (20 ml). Then, 2,4,6-trichloro-1,3,5-triazine (TCT: 10 mmol) at 0 °C with stirring bar for 7 h. Afterwards, the 14 mmol of K 2 CO 3 and 5-phenyl tetrazole (20 mmol) was added to flask and stirred at room temperature. After 4h, the flask was equipped with condenser and refluxed at 50 °C for 24h. the final solid sample was separated and washed with water/methanol three times and dried in a vacuum oven at 60 °C for 12 h. In the end, the 0.2 g of final support was added to the saturated K 2 PdCl 4 solution and stirred at room temperature for 24 h and then, Pd (II) was reduced to Pd (0) with aim of NaBH 4 (15 mg). The Pd-catalyst was separated easily by an external magnet, washed with water/methanol three times and dried under reduced pressure. Based on ICP-OES analysis, the loading of Pd 0 was found 0.78 mmol g −1 for fresh catalyst.
Preparation of the Pd 0 @ UiO-66-NH 2 . A round bottom flask was charged by the magnetic UiO-66-NH 2 (0.2 g) and saturated K 2 PdCl 4 solution and then stirred at room temperature for 24 h. In the end, Pd (II) was reduced to Pd (0) with aim of NaBH 4 (15 mg). The Pd 0 @UiO-66-NH 2 was separated easily by an external magnet, washed with water/methanol three times and dried under reduced pressure. Based on ICP-OES analysis, the loading of Pd 0 was found 0.51 mmol g −1 for fresh catalyst.
Catalytic performance for cyanation. The experiments were performed in a vessel containing aryl halide (1.0 mmol), potassium hexacyanoferrate(II) (0.2 mmol), potassium hydroxide (1.2 mmol), NMP (2.0 ml) and MOF-based catalyst (2.0 mol). The vessel was equipped with stirrer bar and temperature was increased from room temperature to 100 °C slowly. The reaction was monitored until completed (TLC, EtOAc: n-hexane, 1:5). Then, the mixture was diluted by EtOAc and water. The organic phase was with brine and dried with Na 2 SO 4 . The organic layers were mixed, purified by column chromatography, and confirmed by 1 HNMR and 13 CNMR.  www.nature.com/scientificreports/ 80 °C slowly. The reaction was monitored until completed (TLC, EtOAc: n-hexane, 1:10). Then, the mixture was diluted by EtOAc and water. The organic phase was with brine and dried with Na 2 SO 4 . The organic layers were mixed, purified by column chromatography, and confirmed by 1 HNMR and 13 CNMR.