2.1 Materials and instruments

All reagents for synthesis and analysis were obtained commercially and of analytical grade and used without further purification. The elemental analysis (CHN) of compound was performed using a Carlo ERBA Model EA 1108 analyzer. Fourier transform infrared (FT-IR) spectroscopy was carried out by utilizing a Unicam Matson 1000 FT-IR spectrophotometer using KBr disks at room temperature. Thermogravimetric analysis (TGA) was performed on a Perkin Elmer TGA7 analyzer in a N₂ atmosphere with a heating rate of 5 °C min⁻¹. Powder X-ray diffraction (XRD) patterns were recorded by using a Rigaku D-max C III and X-ray diffractometer using Ni-filtered CuKα radiation. Inductively coupled plasma–mass spectroscopy (ICP-MS) was performed by using an ICP-MS HP 4500. The morphology and size of palladium nanoparticles in Pd/MnBDC were investigated by incorporating scanning electron microscopy (SEM, S-4160, Hitachi) and high-resolution transmission electron microscopy (HR-TEM, Philips CM30) at 300 kV. X-ray photoelectron spectroscopy (XPS) measurements were performed on a PerkinElmer PHI 5000CESCA system with a base pressure of 10⁻⁹ Torr. The products of the coupling reaction were determined and analyzed by using an HP Agilent 6890 gas chromatograph equipped with an HP-5 capillary column (phenylmethylsiloxane 30 m × 320 μm × 0.25 μm) and a flame-ionization detector.

2.2 Synthesis of the [Co₃(BDC)₃(DMF)₂(H₂O)₂]_n framework

The [Co₃(BDC)₃(DMF)₂(H₂O)₂]_n coordination polymer (CoBDC) was synthesized based on our previous study [24] by adding a solution of 1,4-benzenedicarboxylic acid (H₂BDC) (166 mg, 1 mmol) in 5 mL of DMF to CoCl₂·6H₂O (291 mg, 1 mmol) solution (DMF, 5 mL). The solution was heated for 48 h at 80 °C. After 2 days, purple crystals of [Co₃(BDC)₃(DMF)₂(H₂O)₂]_n (85% yield based on H₂BDC) were collected. Anal. Calcd: for Co₃C₃₀H₂₄N₂O₁₆: C, 42.42; H, 3.32; N, 3.30. Found: C, 42.63; H, 2.86; N, 3.28. FT-IR (KBr 4000–400 cm⁻¹): 3491 (br), 2931 (w), 1663 (s), 1619 (s), 1559 (s), 1418 (s), 1254 (w), 1115 (m), 855 (m), 813 (m), 760 (s), 693 (s), 655 (m), 468 (m).

2.3 Synthesis of the [Co₃(BDCNH₂)₃(DMF)₂(H₂O)₂]_n framework

A DMF solution (5 mL) of CoCl₂·6H₂O (291 mg, 1 mmol) was added to a solution of 2-amino-1,4-benzenedicarboxylic acid (H₂BDCNH₂) (182 mg, 1 mmol) in 5 mL of DMF in a small sample vessel. The solution was heated for 24 h at 90 °C. A light cream powder of [Co₃(BDC NH₂)₃(DMF)₂(H₂O)₂]_n (named CoBDCNH₂, 85% yield based on H₂BDCNH₂) was obtained as an amino-functionalized polymer. The crystals were washed with DMF (3 mL, 2 times) and dried at 80 °C. Anal. Calcd: for Co₃C₃₀H₃₃N₅O₁₆: C, 40.19; H, 3.71; N, 7.81. Found: C, 40.72; H, 3.70; N, 7.75. FT-IR (KBr 4000–400 cm⁻¹): 3491 (br), 2931 (w), 1663 (s), 1619 (s), 1559 (s), 1418 (s), 1254 (w), 1115 (m), 855 (m), 813 (m), 760 (s), 693 (s), 655 (m), 468 (m).

2.4 Synthesis of supported palladium nanoparticles on the CoBDC polymer framework

The prepared CoBDC polymer (0.15 g) was added to an orange solution of PdCl₂ (0.04 g, 0.2 mmol) and NaCl (0.58 g, 1.0 mmol) in DMF (10 mL) under vigorous stirring. Upon adding hydrazine hydrate (4 mL, excess), the mixture immediately turned gray. After stirring for 15 min, the solid was isolated by centrifugation, washed with DMF and dried in an oven at 90 °C. Supported Pd nanoparticles on the Co-coordination polymer (Pd/CoBDC) was obtained as a gray powder. The Pd contents (9 wt %) in the samples were determined by ICP-MS. FT-IR (KBr 4000–400 cm⁻¹): 3475 (br), 3332 (br), 2935 (w), 1659 (w), 1560 (s), 1555 (m), 1419 (s), 1358 (m), 1142 (m), 855(m), 828 (w), 807 (w), 732 (s), 695 (s), 529 (w), 480 (w).

2.5 Synthesis of supported palladium nanoparticles on the CoBDCNH₂ polymer framework

The supported Pd-NPs on the amino-functionalized Co-coordination polymer (Pd/CoBDCNH₂) was obtained as a gray powder by the addition of an orange solution of PdCl₂ (0.04 g, 0.2 mmol) and NaCl (0.58 g, 1.0 mmol) in DMF (10 mL) to prepare CoBDCNH₂ (0.15 g) in the presence of hydrazine hydrate (4 mL, excess). After stirring for 15 min, the solid was isolated by centrifugation, washed with DMF and dried in an oven at 90 °C. The Pd contents (1 wt %) in the samples were determined by ICP-MS. FT-IR (KBr 4000–400 cm⁻¹): 3485 (br), 2930 (w), 1659 (s), 1620 (s), 1550 (m), 1420(s), 1250 (w), 1102 (m), 850 (m), 810 (m), 765 (s), 690 (s), 650 (m), 460 (m).

2.6 General procedure for Heck coupling reactions catalyzed by Pd/CoBDC and Pd/CoBDCNH₂ catalysts

The heterogeneous C–C coupling reaction was conducted in a two-necked round-bottom flask fitted to a condenser and magnetic stirrer and placed in a temperature-controlled oil bath. Typically, 1.1 mmol of the substrate was taken in 2 mL of solvent, followed by the addition of 13 mg Pd/CoBDC (0.11 mol % Pd) or 5 mg Pd/CoBDCNH₂ (0.05 mol % Pd) catalyst, 1 mmol iodobenzene and 1.5 mmol base. The flask was kept at 90 °C and stirred at appropriate times. The products from the reaction mixture were analyzed by gas chromatography and were identified by comparison to known standards.

3.1 Characterization of the polymer framework and supported Pd nanoparticles

The [Co₃(BDC)₃(DMF)₂(H₂O)₂]_n framework was synthesized and characterized based on our previous reported work [24]. The CoBDC polymer is made of Co–O₆ octahedral linked by H₂BDC ligands. The coordination polymer structure is a 2D-periodic framework formed by two-dimensional infinite straightforward chains through covalent interactions.

The CoBDCNH₂ framework was synthesized in the same way but the attempt to obtain its suitable single crystal was unsuccessful. Therefore, the amino-functionalized framework of CoBDCNH₂ was characterized based on solid powder.

The differences between the functional and nonfunctional polymers are very obvious in the FT-IR spectra of these polymers (Fig. 1). The finger point area for –NH₂ functional group is distinguishable in free ligands (Fig. 1a) and the Co-amino-functionalized polymer (Fig. 1b). These peaks appear at 3503 and 3394 cm⁻¹ in free ligands and at 3310 and 3149 cm⁻¹ in the amino-functionalized polymer, respectively. The existence of these peaks in the polymer indicates that the functional group remained unchanged after polymer formation. The slight shifts are due to the ortho effects of amino and carboxylate groups that happen after framework formation. The absence of these finger point peaks in the Co-nonfunctional polymer is the obvious difference between these two polymers (Fig. 1c).

Fig. 2 demonstrates the following coordination modes in H₂BDCNH₂ linker ligand from carboxylate groups to the metal centers that help to form a polymer framework.

A thermal study shows that both polymers have similar decomposition profiles (Fig. 3). Final destruction occurred at 330–350 °C for CoBDC and at 460–490 °C for the Co-BDCNH₂ polymer. It seems that the functionalized polymer has a higher thermal stability. The first major mass loss step of 19.56% at 380 °C in Co-BDCNH₂ polymer was attributed to the removal of 2 DMF and 2 water molecules per formula unit (calcd. 20.04%). Furthermore, upon further heating, there was no mass loss until the structure collapsed at 460–480 °C (weight loss 43.56%). These weight losses are in agreement with [Co₃(BDCNH₂)₃(DMF)₂(H₂O)₂]_n empirical formula. This supposed formula of the Co-BDCNH₂ polymer was further confirmed by elemental analysis.

The solution impregnation method was used to immobilize Pd-NPs on the functional and nonfunctional Co-coordination polymers (Pd/CoBDCNH₂ and Pd/CoBDC).

The TGA analysis of supported PdNPs on the CoBDC polymer (Fig. 3c) exhibited the same destruction profile as CoBDC but with a slow loss of weight. The weight loss started at 80 °C and continued to 160 °C, then reached 400 °C by a constant slope. Final collapse occurred at 420–470 °C. The weight loss in Pd/CoBDCNH₂ polymer started at 100–300 °C by gentle slope and continued to 340–450 °C by steep slope. The final decomposition of this polymer occurred at 530 °C–600 °C. It seems that the presence of active Pd-NPs on polymers is due to the simple degradation of organic moieties [25].

As shown in the XRD pattern, the presence of PdNPs had no influence on the structural characteristics of the polymers and there is no significant loss of crystallinity in framework diffraction patterns after PdNP loading (Fig. 4). The characteristic peaks of Pd (111) and Pd (200) at 2θ = 40.10 and 46.5°,respectively, are distinguishable.

The size and local composition of PdNPs were investigated by HR-TEM (Fig. 5). The images obviously show that the highly aggregated Pd-NPs are attached to the polymer surface (Fig. 5a,b). EDX spectroscopy confirmed the presence of aggregated Pd-NPs on the external surface of the CoBDC coordination polymer. As observed in Pd/MIL-53(Al) [4], this aggregation was predictable due to the nonfunctionalized polymer framework [4,23]. PdNPs were well dispersed on the surface of CoBDCNH₂ polymers (Fig. 5c and d). It seems that the functionalized polymer possesses an effective influence to prevent the aggregation of particles. Another difference between functionalized and nonfunctionalized polymers is the morphology. It was observed from the SEM images that the change in morphology occurs from nanospheres in the CoBDC polymer to nanosheets in the CoBDCNH₂ polymer due to functionalized linker ligands (Fig. 6). It is obvious that the morphology remains unchanged after the loading of PdNPs.

Using the Debye–Scherer equation, the average diameter of Pd-NPs was estimated to be 30 nm for Pd/CoBDC and 40 nm for Pd/CoBDCNH₂, which matched very well with TEM data as estimated by size distribution. The Pd contents in Pd/CoBDC and Pd/CoBDCNH₂ polymers were determined (9% and 1% wt, respectively) by inductively coupled plasma–mass spectroscopy (ICP-MS).

The X-ray photoelectron spectroscopy measurement (XPS) study demonstrated that the Pd species in polymers are present in the metallic state (Fig. 7). This is verified by the two peaks with a binding energies of 335.6 and 341.1 eV in Pd/CoBDC (Fig. 7a) and 336.3 and 341.9 eV in Pd/CoBDCNH₂ (Fig. 7b) in the 3d_5/2 and 3d_3/2 levels of Pd(0), respectively. Photoelectrons emitted from carboxyl groups, C 1s and O 1s core level in the 1,4-benzenedicarboxylate linker in Pd/CoBDC, generated peaks centered at 284 and 533 eV binding energy, respectively [23]. Photoelectrons emitted from carboxyl groups and an amine branch, C 1s, N 1s and O 1s core level in Pd/CoBDCNH₂, generated peaks centered at 285.5, 399.8 and 532 eV binding energy, respectively. The peaks at 788 and 800 eV were assigned to Co(2p_3/2) and Co(2p_1/2) in the Co-O₆ environment, respectively [26].

3.2 Mizoroki–Heck coupling reaction of iodobenzene and terminal alkenes in the presence of supported Pd-NPs

The coupling of C–C bonds constitutes significant reactions in organic chemistry that can lead to a diverse range of derivatives. Therefore, the catalytic activity of supported PdNPs on the coordination polymers was applied to the Mizoroki–Heck coupling reaction. In order to find the suitable reaction conditions to achieve the maximum yield of the Mizoroki–Heck reaction, conditions such as solvent, base and reaction temperature were studied by choosing styrene and iodobenzene as a model reaction (Table 1). The excellent yields of products were obtained in dimethylacetamide (DMA) (entries 1–5) even at a lower temperature (entry 14). Among the different organic and inorganic bases, the highest conversion of coupling reactions was observed in the presence of Et₃N base (entries 3, 8, 11–13). A controlled experiment indicated that no cross-coupling product was observed in the absence of the catalyst or base (entries 15, 16).

The coupling between iodobenzene and various terminal alkenes has been investigated under the above-optimized reaction conditions, Et₃N as the base and DMA as the solvent at 90 °C (Table 2). The results show that in the presence of supported PdNPs, alkenes with either electron-withdrawing or electron-donating substituents react with iodobenzene, rapidly forming the corresponding coupled products with good yields and excellent selectivity to E-isomeric products (Table 2). The conversion and selectivity of products show that the coupling reactions catalyzed by both catalysts proceed through the same mechanism proposed in the literature [23].

The recyclability of the supported Pd-NPs catalyst was investigated in the reaction of styrene with iodobenzene using Et₃N/DMA at 90 °C for 9 h in the presence of Pd/CoBDC (13 mg, 0.11 mol % Pd) and Pd/CoBDCNH₂ (5 mg, 0.05 mol % Pd). The solid catalyst was easily isolated by centrifugation after the first catalytic reaction and recovered by being washed with the solvent and dried at 80 °C. Afterwards, it was used for the next run under the same reaction conditions. The catalytic activity was found to remain almost unchanged even after four cycles, without significant degradation in activity and selectivity to trans-stilbene as a major E-product (Table 3).

The catalytic results show that supported PdNPs on polymers are active catalysts in the C–C coupling reaction. For C-C coupling reaction of iodobenzene and terminal olefins, amount of 13 mg (0.11 mol % Pd) from Pd/CoBDC catalyst and 5 mg (0.05 mol % Pd) from Pd/CoBDCNH2 catalyst, were used. So, from a catalytic point of view, the Pd/CoBDCNH₂ is superior in terms of the amount of the catalyst. It seems that highly dispersed and low loading PdNPs on CoBDCNH₂ (1%) are efficient than aggregated and high loading Pd-NPs on CoBDC (9%). This fact demonstrates the functionality effect of the polymer on catalyst efficiency.

Since some of the active sites in the solid catalyst could dissolve into the solution during the catalytic reaction, the amount of palladium in solution after each catalytic run was determined by ICP-MS to find whether palladium leaching occurs during the reaction. Therefore, after the first catalytic reaction of iodobenzene and styrene by Pd/CoBDC and Pd/CoBDCNH₂, the heterogeneous catalyst was simply separated from the reaction mixture by centrifugation and the content of palladium in solution was determined by ICP-MS (Table 3). The leaching test after each catalytic run revealed that the amount of palladium leached from the catalyst to solution is very negligible. Moreover, fresh reagents were added to the solution and stirred at 90 °C. Only less than 6% conversion was observed after 24 h for both catalysts. Therefore, there is no contribution of catalytically active palladium leached species in solution.

In order to further confirm that the reaction was indeed catalyzed by supported PdNPs instead of free Pd-NPs in styrene and iodobenzene coupling reaction, a hot filtration test was carried out by stopping the reaction after 3 h. The reaction reached 25 and 32% conversion for Pd/CoBDC and Pd/CoBDCNH_2, respectively. Then the catalyst was separated from the reaction mixture and reaction was continued. The conversion was increased by 6 and 7%, respectively, after 24 h. These results confirmed that the coupling reaction could only continue in the presence of the solid catalyst, and there was no contribution from the homogeneous catalysis of leached active species in the liquid phase.

The XRD patterns of catalysts after the reaction displayed the same diffraction profile and relatively similar intensity that indicated that the catalysts retained a well-crystallized structure after the catalytic reaction (Fig. 4c,f).