Cobalt supported on alumina as green catalyst for Biginelli reaction in mild conditions: effect of catalyst preparation method

2.1 Catalysts preparation

Cobalt supported on alumina with a loading (5–15% wt) was synthesized with different methods: IMP, SG and reverse ME. Co(NO₃)₂·6H₂O was used as the cobalt precursor. For samples prepared by the SG and ME methods, aluminum nitrate salt Al(NO₃)₃·9H₂O (Riedel-Haën, Hanover, Germany) was used as the precursor of Al₂O₃ support. In the IMP procedure, an aqueous solution of Co nitrate was added dropwise to γ-Al₂O₃ (Merck, Darmstadt, Germany) followed by drying at room temperature. In the SG method, 0.2 mol of citric acid was added to an aqueous solution constituted of 0.01 mol of Co(NO₃)₂ and 0.17 mol of Al(NO₃)₃. The solution was then stirred at 80°C until the gel formation. In the ME procedure, an aqueous solution (A) containing 0.17 mol Al(NO₃)₃.9H₂O, 0.01 mol Co(NO₃)₂.6H₂O, 300 ml cyclohexane, 62 ml butanol-1 and 0.15 mol cetyltrimethylammonium bromide (CTAB) surfactant was maintained under vigorous magnetic stirring for 1 h (reverse ME). A second aqueous solution (B) containing 80 ml ammonium hydroxide (2.05 M) and 300 ml cyclohexane, 62 ml butanol-1 and 0.15 mol CTAB surfactant was maintained under vigorous magnetic stirring for 1 h. Then, (A) and (B) solutions were mixed under vigorous magnetic stirring for 24 h. The resulting solid was filtered and washed repeatedly with methanol. The prepared solids by different methods were dried for 24 h at 110°C and calcined under air at 500°C for 4 h with a heating rate of 4°C/min.

The samples were noted as XCoAl-x, where X designates X%wt of Co and x the preparation process (IMP, SG and reverse ME).

2.2 Characterization

Nitrogen adsorption-desorption isotherms were obtained using a Micrometrics ASAP 2020 apparatus. Before physisorption measurements, samples were evacuated at 250°C. The Barrett, Joyner and Halenda (BJH) method was used to determine the mean pore size.

The real composition of the solids was determined by elemental analysis based on fluorescence X with an XSRS 3400 SIEMENS fluorescence spectrometer.

Powder XRD patterns of the catalysts were recorded on a SIEMENS D5000 powder diffractometer using Cu-Kα radiation. Crystalline phases were identified through Highscore plus software database.

Sample morphology was examined by both SEM and TEM using Philips XL30 ESEM and CM10 microscopes, respectively.

2.3 Catalytic test

Catalytic activity of cobalt supported on alumina was examined in the Biginelli reaction under mild and solvent free conditions.

In a 25 ml single-neck round bottom flask, a mixture of benzaldehyde (1) (2 mmol), ethyl acetoacetate (2) (2 mmol), urea (3) (3 mmol) and 0.2 g of catalyst was kept under regular stirring at 100°C for 1 h (Scheme 1). Completion of the reaction was monitored by thin layer chromatography. The reaction product was washed with boiled ethanol and then filtered to remove the catalyst. DHPM product was purified and recrystallized in hot ethanol, then placed in an ice bath to obtain a pure product. The isolated catalyst was dried at 100°C overnight and was reused in the next runs without further purification. Identification of DHPM is obtained via its melting temperature, infrared (IR) spectroscopic analysis and ¹H and ¹³C nuclear magnetic resonance (NMR).

Scheme 1:

Biginelli reaction over 10CoAl-x systems.

IR spectra were recorded on Fourier transform spectrometer Shimadzu FTIR-type 8400 and the NMR ones on a Bruker FT-NMR AVIIIHD500. Chemical shifts were expressed in parts per million (ppm) relative to Trimethylsilyl (δ=0 ppm) and coupling constant J in Hertz (Hz). NMR multiplicities are reported using the following abbreviations: br=broad, s=singlet, d=doublet, t=triplet, q=quartet, m=multiplet.

3.1 Catalysts characterization

BET surface area, pore and micropore volume, micropore area and pore size obtained for the samples are summarized in Table 1. It appears that the preparation method has a great influence on these parameters. The specific surface area, pore volume, micropore volume and micropore areas of 10CoAl-ME are much higher than those obtained for 10CoAl-IMP and 10CoAl-SG (329 against 51 and 11 m²/g, 47×10⁻² against 9×10⁻² and 1×10⁻² cm³·g⁻¹, 9.12×10⁻³ against 5.16×10⁻⁴ and 5.03×10⁻⁴ cm³·g⁻¹ and 27.02 against 2.22 and 1.44 m²/g, respectively). These differences can be related to the fact that the ME method favors the increase of specific surface area, pore volume, micropore volume and micropore areas of materials compared to other procedures, as reported in the literature [43]. Low specific surface area, low values of pore volume, micropore volume and micropore areas obtained with the IMP method were already underlined by some authors in the case of Co supported on alumina [44]. On the contrary, Pajonk [45] obtained higher surfaces parameters by the SG method. For all systems, the presence of hysteresis loops in nitrogen isotherms suggests the existence of mesopores with an average pore size of 72.3 Å, 57.6 Å and 42.5Å for 10CoAl-IMP, 10CoAl-ME and 10CoAl-SG, respectively. Their nitrogen adsorption isotherms could be ascribed to type IV (one example in Figure 1). The chemical analysis of samples is listed in Table 2.

Figure 1:

N₂ adsorption isotherms and BJH pore size distribution of 10CoAl-impregnation (IMP).

The obtained results were in good accordance with the expected ones for cobalt, alumina and oxygen. In this way, the Co_exp/Co_theo atomic ratios obtained were 1, 0.90 and 0.72 for IMP, SG and ME, respectively.

The XRD profiles of the cobalt supported on alumina prepared by different methods are shown in Figure 2. Similar XRD patterns were observed for both 10CoAl-ME and 10CoAl-IMP (Figure 2A) with intense lines assigned to the CoAl₂O₄ spinel phase, crystallized in a cubic structure (2θ=31°, 36°, 38°, 44°, 56°, 59° and 65°) and characteristic lines of the Co₃O₄ phase, crystallized in a cubic phase (2θ=31°, 3°8, 51°, 56°, 58°, 67°). NiAl₂O₄ spinel was already observed in the case of nickel supported on alumina [46] and the Co₃O₄ phase was also observed in the case of Co supported on Al₂O₃ prepared by IMP and ME [47], [48]. For the IMP cobalt sample, another peak at 2θ=39° was observed, which was attributed to CoO. It was noted that the observed peaks at 2θ=36°, 56°, 65° and 67° can be also assigned to Al₂O₃. The XRD pattern of 10CoAl-SG exhibits an amorphous phase (Figure 2B). It was underlined that the material crystallinity is sensitive to the calcination temperature [49], [50]; an increase in the latter can favor the oxide formation. Furthermore, a well-dispersed cobalt species on the support was observed when the system was amorphous [51].

Figure 2:

X-ray powder diffraction patterns of calcined materials at 500°C: (A) 10CoAl-impregnation (IMP) and 10CoAl-microemulsion (ME); (B) 10CoAl-sol-gel (SG).

SEM images of calcined 10CoAl-IMP, 10CoAl₋–SG and 10CoAl-ME at 500°C are represented in Figure 3. Micrographs of the materials are different. That of 10CoAl-IMP reveals a smoother surface on which particles are arranged fairly regularly with aggregates morphology, whereas 10CoAl-SG has a lamellar texture with a bright burst. With the ME method, a heterogeneous porous morphology was observed. These results are in accordance with those of BET and BJH surface parameters.

Figure 3:

Scanning electron microscopy (SEM) images of calcined materials at 500°C: 10CoAl-impregnation (IMP), 10CoAl-sol-gel (SG) and 10CoAl-microemulsion (ME).

The detailed microstructure of 10CoAl-ME was further investigated by TEM and the result is shown in Figure 4. The TEM micrograph reveals uniform distribution of cobalt nanoparticles over the alumina support, indicating good dispersion.

Figure 4:

Representative transmission electron microscopy (TEM) (A) and scanning electron microscopy (SEM) (B) images obtained for 10CoAl-microemulsion (ME) catalyst. Scale bars: (A) 200 nm; (B) 100 µm.

3.2 Biginelli reaction

The preliminary catalytic tests performed with Co(NO₃)₂·6H₂O or Al(NO₃)₃·9H₂O (entries 12 and 13) showed that the Biginelli reaction takes place in the homogenous phase and high DHPM yields were obtained (85% and 89%, respectively). Figures 5 and 6 show the ¹H and ¹³C NMR results.

Figure 5:

¹H nuclear magnetic resonance (NMR) of produced 3,4-dihydropyrimidin(1H)-one (DHPM) over 10CoAl-sol-gel (SG) system.

Figure 6:

¹³C nuclear magnetic resonance (NMR) of produced 3,4-dihydropyrimidin(1H)-one (DHPM) over 10CoAl-sol-gel (SG) system.

The catalytic performances of supported alumina cobalt systems prepared with different methods (SG, ME and IMP) were investigated for the 5-ethoxycarbonyl-4-phenyl-6-methyl-3,4-dihydropyridin-2(1H)-one (DHPM) synthesis under free solvent conditions from benzaldehyde, ethyl acetoacetate and urea with a ratio of 2:2:3, via the Biginelli reaction.

In order to optimize the reaction parameters, several tests were carried out under various conditions. Obtained catalytic results are summarized in Table 3. The Biginelli reaction, only in the γ-Al₂O₃ support presence led to a low DHPM yield (40%) after 1 h of reaction time (Table 3, entry 1).

The effect of cobalt content on the support has been investigated using the IMP preparation method. The catalytic results showed that a cobalt amount increase from 5 wt% up to 10 wt% led to DHPM yield increase from 50% to 56%, while with 15 wt% of cobalt content, the DHPM yield was only 57% (entries 2–4, Table 3). From these observations, the other catalytic tests were carried out with 10 wt% of cobalt. It is important to note that the performance of 10CoAl-x catalysts depended on the preparation method. In this sense, DHPM yields decreased in the following order: SG (71%)>ME (62%)>IMP (56%) (entries 3, 5, and 6). These observations have already been reported by other authors on Co/Al₂O₃ systems [44], [52], [53], [54], [55], [56]. Therefore, the low catalytic activity of Co/Al₂O₃ prepared by IMP compared to systems prepared by the SG and ME techniques can be related to the strong interaction between Co particles and Al₂O₃ support reducing the number of active Co sites and the difficulty to obtain homogenous cobalt particles distribution by conventional IMP [44]. Moreover, it was observed that the reagent nature, operation conditions (pH…) and synthesis method used can influence the distribution and the active species nature [53]. A pH modification can alter the particle size [57]. The citric acid use in the case of the IMP method favors a “fully” dispersed active species [58].

The influence of the reaction time (45 min, 60 min, 75 min and 90 min) on DHPM yield, on the 10CoAl-SG system was studied. The increase of reaction time led to higher DHPM yield (71%) (entries 7, 5, 8 and 9). The results showed that DHPM yield increases with time and reaches a maximum of 71% after 60 min. From these results, the reaction time was fixed to 1 h.

The influence of catalyst weight (0.05 g, 0.1 g and 0.2 g) was also investigated with the 10CoAl-SG system. DHPM yield increased from 63% up to 71% with the increase of catalyst weight from 0.05 g up to 0.2 g (entries 11, 10 and 5). It is important to note that up to 0.2 g, a bad agitation was observed following to the formation of a compact reagent mixture.

The results compiled in Table 3 also show that when the reactant ratio (benzaldehyde:ethyl acetoacetate:urea) changed from 1:1:1.5 to 2:2:3, the DHPM yield increased from 48% up to 71% for a catalyst mass of 0.2 g (entries 5c and 5b).

The reusability of both 10CoAl-SG and 10CoAl-ME was investigated under optimal operating conditions (catalyst: 0.2 g, reaction temperature: 100°C, reaction time: 1 h). Used samples were recovered by filtration, washed with ultra-pure water, dried overnight at 100°C and reused in the next runs without further purification. DHPM yields are given in Table 4. After three reaction cycles, DHPM yield decreased by 6% and 0% in the presence of 10CoAl-SG and 10CoAl-ME, respectively.

This study showed that the 10CoAl-SG system is the most active for the Biginelli reaction. However, the high activity obtained is difficult to correlate to surface parameters because the catalytic sample studied exhibits low BET area surface, low pore volume and low micropore volume and different morphology compared to 10CoAl-IMP and 10CoAl-ME systems.

Physical data for 5-ethoxycarbonyl-4-phenyl-6-methyl-3,4-dihydropyridin-2(1H)-one (DHPM). Mp 201–205°C. IR (KBr): ν_max=3245 cm⁻¹, 3100 cm⁻¹, 2960 cm⁻¹, 1720 cm⁻¹, 1630 cm⁻¹, 1540 cm⁻¹, 1490 cm⁻¹. ¹H NMR (Figure 5) (500 MHz, DMSO-d₆): δ=9.19 (s, 1H, NH), 7.73 (s, 1H, NH), 7.34 (s, 5H, C₆H₅), 5.15 (s, 1H, J=2.87 Hz, CH), 4.01 (q, J=6.5 Hz, 2H, OCH₂CH₃), 2.51 (s, 3H, CH₃), 1.11 (t, J=6.5 Hz, 3H, OCH₂CH₃) ppm. ¹³C NMR (Figure 6) (125 MHz, DMSO-d₆): δ=165.7 (COOEt), 152.5 (C2), 148.8 (C6), [145.3, 128.8, 127.7, 126.7] C arom, 99.7 (C5), 59.6 (OCH₂CH₃), 54.4 (C4), 18.2 (CH₃), 14.5 (OCH₂CH₃) ppm.