The hydroformylation of alkenes represents one of the most widely used homogeneously catalyzed reactions in industry. The aldehyde products are valuable intermediates in the synthesis of bulk chemicals like alcohols, esters, and amines [1, 2, 3, 4, 5, 6]. Different catalysts have been developed based on cobalt or rhodium complexes [3, 7, 8, 9]. Rhodium complexes generally show much higher activity and aldehyde selectivity than cobalt complexes under mild conditions. However, until very recently, all commercial hydroformylation of long chain alkenes used cobalt-based catalysts because of the high cost of noble metals and the difficulty in the separation of thermally sensitive rhodium-based catalysts from the low volatility products [3]. Therefore, the immobilization of rhodium complexes on a solid support is an important issue for practical application because of the advantages of heterogeneous catalysis, such as easy purification of the product and recovering and recycling of the catalyst [10].
Several strategies have been developed for the immobilization of rhodium complexes, such as covalent tethering, electrostatic interaction, encapsulation, and so on [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]. The covalent tethering of homogeneous rhodium complexes on a chemically modified inorganic solid support containing donor functions often gives superior stability to those prepared by other methods. Rhodium complexes anchored on silica gel and mesoporous silica (SBA-15, MCM-41, or MCM-48) have been reported previously [10, 22]. Although the immobilized catalysts are easily recyclable, they often exhibited lower activity than their corresponding homogeneous catalysts due to the altered microenvironment of the metal complexes and mass transfer limitation in the solid catalysts. Tethering metal complexes in a network of an organic polymer support provides another approach for the immobilization of homogeneous catalysts [23, 24]. However, polymer supported catalysts often have the problem of swelling in an organic solvent. Polymer@silica composite materials, which combines the advantages of both an organic polymer (e.g., flexibility, ductility, and processibility) and an inorganic material (e.g., rigidity and thermal stability), have potential as an efficient solid support for the immobilization of metal complex catalyst [25, 26].
Here, we report the synthesis of Rh-PPh3- polymer@mesosilica composite catalysts by engineering the polymerization of divinylbenzene (DVB) and 4-vinyl- triphenylphosphine monomer in the nanopores of a mesoporous silica followed by coordination with Rh(acac)(CO)2 (acac = acetylacetonate) for the hydroformylation of 1-octene. The influence of polymer content, pore structure, and pore size of the support on the catalytic performance of the composite catalysts was investigated.
All chemicals were used as received unless otherwise stated. Pluronic P123 copolymer (EO20PO70EO20) and pluronic F127 (EO106PO70EO106) were purchased from Sigma Aldrich. Tetraethylorthosilicate (TEOS, AR) and hexadecyltrimethylammonium bromide (CTAB) were purchased from Shanghai Chemical Reagent Company of the Chinese Medicine Group. The metal precursor Rh(acac)(CO)2 was purchased from Energy Chemical Company. Divinylbenzene (DVB, 80%) was purchased from Aladdin Reagent Company. Toluene was distilled from sodium benzophenone ketyl under argon. 4-Vinyl-triphenylphosphine was synthesized according to the literature method [27]. SBA-15, MCM-41, and FDU-12 were synthesized according to a reported method [28, 29, 30].
DVB was purified on an alumina column to remove the polymerization inhibitors. The mesoporous silica materials FDU-12, SBA-15, and MCM-41 were used after vacuum degassing at 120 °C for 3 h. Typically, DVB (Mw = 130; 21 mg), 4- vinyl-triphenylphosphine (Mw = 288; 144 mg) monomer mixtures and 2,2′-azobisisobutyronitrile (AIBN, 3% relative to total vinyl group) were added to 1.0 g of activated mesoporous silica following the wet impregnation procedure using N,N- dimethylformamide (DMF) as solvent [31]. After impregnation with the solution, the sample was subjected to a freeze-vacuum- thaw procedure to remove air and reach equilibration at the same time to achieve a uniform distribution. Then, the sample was heated to perform the polymerization reaction at 90 °C for 20 h. The resulting samples were thoroughly washed with dichloromethane and ethanol several times to remove unreacted monomers and oligomers, followed by drying under vacuum. The samples prepared with FDU-12, SBA-15, and MCM-41 were denoted as PPh3-polymer@FDU-12, PPh3-polymer@SBA-15, and PPh3-polymer@MCM-41, respectively. For FDU-12 based composite materials, samples with different polymer contents were also prepared. The composite material synthesized with 1.0 g of FDU-12, 576 mg of 4-vinyl-triphenylphosphine, and 81 mg of DVB was denoted as L-PPh3-polymer@FDU-12, while that synthesized with 1.0 g of FDU-12, 288 mg of 4-vinyl- triphenylphosphine, and 41 mg of DVB was denoted as M-PPh3-polymer@FDU-12.
XRD patterns were recorded on a Rigaku RINT D/Max-2500 powder diffraction system using Cu Kα radiation. Nitrogen physical adsorption measurement was carried out on Micrometritics ASAP 2020 volumetric adsorption analyzer. Before the measurement, the samples were outgassed at 120 °C for 5 h. The BET surface area was evaluated from the data in the relative pressure range p/p0 of 0.05 to 0.25. The total pore volume was estimated from the amount adsorbed at the p/p0 value of 0.99. The pore diameter was determined from the adsorption branch by the BJH method. Thermogravimetric analysis (TGA) was performed under an air atmosphere with a heating rate of 5 °C/min using a NETZSCH STA-449F3 thermogravimetric analyzer. Transmission electron microscopy (TEM) was performed using an FEI Tecnai G2 Spirit at an acceleration voltage of 120 kV. Scanning electron microscopy (SEM) was undertaken on a JEOL JSM-6360 scanning electron microscope operating at an acceleration voltage of 20-30 kV. FT-IR spectra were collected with a Nicolet Nexus 470 IR spectrometer.
The PPh3-polymer@mesosilica composite materials and rhodium precursor Rh(acac)(CO)2 (P/Rh = 5) were added to dichloromethane with stirring at room temperature. After 2 h, the solvent was removed under vacuum. The yellow solid material obtained was used directly for the catalytic reaction. Anhydrous toluene (3 mL), 1-octene (280.5 mg, 2.5 mmol), and dodecane were added to a test tube containing a desired amount of PPh3-polymer@mesosilica composite catalyst (5 μmol Rh) under nitrogen. The test tube was transferred into a stainless steel autoclave and sealed. After purging with CO/H2 (1/1) several times, the pressure was adjusted to 5 MPa and left with stirring at 80 °C in an oil bath. After reaction, the CO/H2 pressure was released, and the solid catalyst was separated by centrifugation. The conversion and aldehyde selectivity were analyzed by gas chromatography using an HP-5 capillary column (30 m × 0.32 mm × 0.25 mm). For the catalyst recycling, the solid catalyst obtained after centrifugation was used directly for the next catalytic reaction.
High quality host materials of FDU-12, SBA-15, and MCM-41 were employed for the insertion of the polymer in the nanopores. The PPh3 functionalized composite materials were straightforwardly constructed by in situ radical polymerization of DVB and 4-vinyl-triphenylphosphine monomers loaded inside the nanopores of the mesoporous silica. DVB was used as both monomer and cross linker to increase the cross linking degree, which is beneficial for the stability of the polymer in the nanopores. The monomers (DVB and 4-vinyl- triphenylphosphine) and initiator were introduced into the nanopores of the mesoporous silica by a wet impregnation method, followed by equilibration under reduced pressure in order to achieve a uniform distribution. Subsequently, the monomers adsorbed in the nanopores of mesoporous silicas were polymerized by heating to form the polymer layer. Finally, the PPh3-polymer@mesosilica composite materials were washed with dichloromethane and ethanol to remove unreacted monomers and polymers loosely adsorbed on the outer surface of the mesoporous silica.
As shown in Fig. 1, the XRD patterns of the parent SBA-15 and MCM-41 exhibited three diffraction peaks assigned to the (100), (110), and (200) faces, which are characteristics of mesoporous materials with a 2-D hexagonal mesostructure. The XRD pattern of FDU-12 showed an intense peak corresponding to the (111) reflection along with a shoulder peak assigned to the (311) reflection, showing that FDU-12 has the cubic Fm3m symmetry. The XRD characterization suggested that mesoporous silicas (SBA-15, FDU-12, and MCM-41) with different porous structure have been successfully synthesized. After incorporation of the polymer in the nanopores, the PPh3-polymer@mesosilica composite materials obtained exhibited similar reflection patterns to their parent silica materials, indicating that the periodic mesopore structure was retained.
The textural properties of the mesoporous silica and the corresponding PPh3-polymer@mesosilica composite materials were characterized by N2 adsorption experiment. The results are summarized in Fig. 2 and Table 1. As can be seen in Fig. 2, all of the samples exhibited a type IV N2 adsorption isotherm, showing that all the samples possess mesoporous pore structure. The PPh3-polymer@mesosilica composite materials showed adsorption curves and hysteresis loops similar to their parent supports, suggesting that the porous structure was not changed after polymer inclusion, which was consistent with the XRD results. As the polymer content increased, the H2 hysteresis loop of the FDU-12 based composite materials became less defined, probably due to the occupation by the polymer in the pores. After the inclusion of polymer in the nanopores of the mesoporous silica, the BET surface area, pore volume, and pore diameter decreased slightly (Table 1). For example, the BET surface area, pore volume, and pore size of PPh3- polymer@SBA-15 and PPh3-polymer@MCM-41 decreased from 672 m2/g, 1.24 cm3/g, and 8.2 nm to 561 m2/g, 0.84 cm3/g, and 8.0 nm, and 932 m2/g, 0.80 cm3/g, and 2.3 nm to 783 m2/g, 0.69 cm3/g, and 2.0 nm, respectively. As the polymer content increased from 14.0% to 33.9%, the BET surface area, pore volume, and pore size of the FDU-12 based composite materials decreased from 543 m2/g, 0.54 cm3/g, and 11.6 nm to 375 m2/g, 0.33 cm3/g, and 8.2 nm, respectively. The reduction in the textural parameters was probably due to the introduction of polymer into the pores. The N2 adsorption experiment results demonstrated that the PPh3-polymer@silica composite materials obtained possess high surface area and large pore volume and pore size, which would be beneficial for the exposure of active site and diffusion of the substrates.
TEM measurement was also performed. As shown in Fig. 3, all the samples have a uniform, well defined mesoporous structure. Bulk polymer formation was not found on the external surface of the silica particles, demonstrating that most of the monomers were polymerized inside the silica mesopores [31]. The SEM images showed that both MCM-41 and PPh3- polymer@MCM-41 have a spherical morphology with a particle size of 100-200 nm. No bulk polymers were found in the SEM of PPh3-polymer@MCM-41, which further demonstrated that most of the monomers were polymerized inside the silica mesopores.
Fig. 4 shows FT-IR spectra of silica FDU-12 and the PPh3-polymer@mesosilica composite materials. All the materials displayed the general characteristic bands of pure silica at 3456 and 1085 cm-1for ν(O-H) and ω(Si-O), respectively. The weak bands at 2850-3000 cm-1 in the composite materials were assigned to the C-H stretching vibration of phenyl group. The bands at 1400-1500 cm-1 were attributed to the C=C vibration of the phenyl group. The bands at 650-770 cm-1 were assigned to the vibration of the phenyl ring. All these observations demonstrated the successful incorporation of the polymer within the mesoporous silica.
The quantitative determination of the incorporated polymer was made by TGA (Fig. 5 and Table 1). The mass loss below 200 °C corresponded to adsorbed water. There was a little increase of mass at 230-300 °C, which was attributed to the oxidation of the element P. The mass loss between 200 and 800 °C demonstrated that more than 95% of the monomers were converted to a stable polymer.
After coordination of the metal precursor Rh(acac)(CO)2, the composite catalyst was obtained. The catalytic performance of the solid Rh-PPh3-polymer@mesosilica composite catalyst was tested by the hydroformylation of 1-octene. The results are summarized in Table 2. The homogeneous metal precursor Rh(acac)(CO)2 catalyzed the reaction very fast, but the aldehyde selectivity for the aldehyde products was poor (80%, entry 1). After adding some PPh3 ligands into the reaction solution, the selectivity was increased to 99% (n/b = 73/27, entry 2). By tethering the metal precursor to the PPh3 functionalized polymer, 54% conversion and 98% aldehyde selectivity (entry 3) were obtained within 9 h. The activity was enhanced by tethering the metal precursor to the PPh3-polymer@mesosilica composite materials, which catalyze the reaction to the desired aldehyde products smoothly with high selectivity (92%-96%). For example, with Rh-PPh3-polymer@FDU-12 as the catalyst, the substrate was fully converted within 9 h with 94% aldehyde selectivity (entry 6). The increased activity was related to the large surface area and pore volume of the catalyst, which are beneficial for the exposure of active sites and mass transport of substrates.
The influence of polymer content on the catalytic performance of the Rh-PPh3-polymer@FDU-12 composite catalysts was investigated (entries 4-6). On increasing the polymer content from 14.0% to 33.9%, the activity of composite catalyst decreased significantly. Within 9 h, the conversion decreased from 98% to 69%, while the selectivity was maintained almost the same (94%-96%). The lower activity at high polymer content was due to the decreased surface area and pore volume, which would cause a decreased diffusion rate of reactants and products during the catalytic process.
The combination of high surface area with a control of the pore size and pore connectivity of the silica support has a significant influence on the activity of the Rh-PPh3- polymer@mesosilica composite catalysts. All the solid catalysts with different silica supports can efficiently convert 1-octene with almost full conversion and excellent selectivity to the desired aldehyde products (Table 2, entries 6-8). To investigate the catalytic difference between the various supports, the reaction kinetic plots were also recorded. As shown in Fig. 6, as the reaction time increased, the conversion of 1-octene increased. All the Rh-PPh3-polymer@mesosilica composite catalysts show better catalytic activity than the pure polymer catalyst. The TOF of the solid catalysts decreased in the order of Rh-PPh3-polymer@MCM-41 > Rh-PPh3-polymer@SBA-15 > Rh-PPh3-polymer@FDU-12 > Rh-PPh3-polymer, which was the same as the order of their BET surface area. Their aldehyde selectivity was slightly different, and the Rh-PPh3- polymer@FDU-12 composite catalyst showed a little higher selectivity than the others (range from 92% to 94%). The above results indicated that Rh-PPh3-polymer@mesosilica composite catalysts with high surface area and pore volume benefit the catalytic performance while the pore size and pore connectivity may influence their activity and selectivity. After the reaction, the rhodium leached amount from the solid catalyst was measured, and very little metal loss was found. The leached amount of Rh-PPh3-polymer@mesosilica composite catalysts (range from 0.04% to 0.24%) was lower than the pure polymer catalyst (0.56%), which demonstrated that silica around the polymer catalyst is beneficial for decreasing metal leaching.
The influence of the reaction conditions, such as temperature, H2/CO pressure, and solvent, was also investigated (Table 2, entries 9-14). On increasing the reaction temperature to 100 °C, the substrate was fully converted within 6 h. In contrast, on decreasing the temperature to 50 °C, the conversion only reached 22% after 12 h. On varying the CO/H2 pressure, the catalyst showed the similar tendency that a higher pressure resulted in higher activity. With 1,4-dioxide as the reaction medium, the reaction activity was comparable to toluene, but the product selectivity decreased (Table 2, entry 9). Notably, the composite solid catalyst also catalyzed the reaction in water medium under similar reaction conditions. After reaction for 12 h, 74% conversion and 85% selectivity were obtained (Table 2, entry 10).
The recyclability of the solid catalyst was tested by using Rh-PPh3-polymer@FDU-12 as a model catalyst for the hydroformylation of 1-octene in toluene (Fig. 7). The solid catalyst can be reused more than 15 times without loss of activity. The solid catalyst after recycling was characterized by N2 adsorption and TG to investigate its stability. The BET surface area and pore volume decreased to 358 m2/g and 0.36 cm3/g, respectively. The polymer content was almost the same as with the catalyst before reaction, which demonstrated that the composite catalyst was stable. However, the aldehyde selectivity decreased to 58% after 15 recycle, which was much lower than with the fresh catalyst. The TEM image of the solid catalyst after recycling 15 times showed the existence of rhodium nanoparticles on the support (Fig. 8). This may be the main reason for the decreased selectivity.
We demonstrated the successful immobilization of rhodium complexes in PPh3-polymer@mesosilica composite materials for the hydroformylation of 1-octene. Under optimized conditions, the solid materials exhibited high surface area, large pore size, and large pore volume, which are beneficial for the exposure of active sites and diffusion of substrate. The solid catalysts efficiently catalyzed the hydroformylation of 1-octene with good yield (99%) and aldehyde selectivity (95%). Efficient solid catalysts can be prepared by the polymerization of the ligands in the nanopores of a mesoporous silica.
2015, Vol. 36 Issue (2): 168-174
PDF (20259 KB)
2015, Vol. 36 Issue (2): 168-174
PDF (20259 KB)