Highly Active and Selective Nanoalumina-Supported Wilkinson's

Jan 27, 2009 - G0/nano-Al2O3 and G1/nano-Al2O3 were used to tether Wilkinson's catalyst. (RhCl(PPh3)3) to produce heterogenized rhodium complex ...
4 downloads 0 Views 237KB Size
Download PDF
1824

Ind. Eng. Chem. Res. 2009, 48, 1824–1830

Highly Active and Selective Nanoalumina-Supported Wilkinson’s Catalysts for Hydroformylation of Styrene Peng Li, Warintorn Thitsartarn, and Sibudjing Kawi* Department of Chemical & Biomolecular Engineering, National UniVersity of Singapore, 10 Kent Ridge Crescent, Singapore 119260, Republic of Singapore

Uniform nano-Al2O3 particles were functionalized with surface amine ligands to form G0/nano-Al2O3 (where G stands for dendrimer generation) and then grafted with first generation PAMAM (polyamidoamine) dendrimers to form G1/nano-Al2O3. G0/nano-Al2O3 and G1/nano-Al2O3 were used to tether Wilkinson’s catalyst (RhCl(PPh3)3) to produce heterogenized rhodium complex catalysts, named respectively as Rh/G0/nanoAl2O3 and Rh/G1/nano-Al2O3. Both Rh/G0/nano-Al2O3 and Rh/G1/nano-Al2O3 catalysts show much higher activity and selectivity for hydroformylation of styrene than RhCl(PPh3)3 anchored on supports such as R-Al2O3, γ-Al2O3, and SBA-15. An increase of regioselectivity was observed from Rh/G0/nano-Al2O3 to Rh/G1/nanoAl2O3, showing the positive dendrimer effect on promoting the selectivity of styrene hydroformylation. 1. Introduction The development of versatile catalytic systems incorporating the advantages of both homogeneous and heterogeneous catalysis has been a primary goal for both academic and industrial research in the area of hydroformylation reaction.1 Various supports, such as polymer, silica, MCM-41, zeolite, and alumina have been employed to heterogenize rhodium complexes.2-7 Although major efforts have been directed in this area, unfortunately so far the immobilization of catalyst on any solid supports results in a significant loss of activity and selectivity.8,9 Recently two unique catalyst supports have been emerging. One is the nanoparticle which has been found to have potential applications in catalysis.10-13 The other is dendrimer which is known to be a unique nanoscale device and several dendritic catalysts with positive dendrimer effect for hydroformylation reactions have been reported.14-17 To incorporate the advantages of homogeneous and heterogeneous catalysis, silica has been employed as a support for growing PAMAM dendrimers from zero to fourth generation and the dendritic silica has been used to anchor rhodium complex catalysts.18 However, the performance of these dendritic catalysts for hydroformylation reaction is not directly proportional to the growth of dendrimer generation because of the structural imperfection within the silica-grafted dendrimer framework. The entanglement of arms in the grafted dendrimer, which may block the active sites as the generation increases, is responsible for the decrease in activity for the third and fourth generation-based catalysts as compared to the zeroth-generationbased catalyst in styrene hydroformylation.19 As the generation of dendrimer increases, the terminal active sites at the periphery are expected to increase at an exponential rate. However, longer arms bring more fractional growth and entanglement, and only the terminal groups at the periphery are useful for catalysis. If all the reactive terminal ends of the dendrimer molecule are cut and grafted on an inorganic nanoscale support, the resulting dendrimer-grafted nanoscale material not only harnesses the multiple active sites of dendrimer but also can be recycled more easily and economically. Therefore, nanoparticles appear to be the ideal support for grafting the terminal groups of dendrimer. In view of the large * To whom correspondence should be addressed. E-mail: chekawis@ nus.edu.sg. Tel.: 65-6516 6312. Fax: 65-6779 1936.

surface to volume ratio of nanoparticles as compared with those of the bigger particles, nanoparticles seem able to provide more efficient hosting of active catalytic sites per unit mass. Recent studies using nanoparticle-supported catalysts demonstrate excellent performances of these catalysts.20 Dendrimers supported on mesoporous SBA-15 support have also been recently reported to anchor rhodium complexes stably during catalytic reaction.21 Given with the prospect that a combination of nanoscale support with nanoscale dendrimer may provide large surface areas and multiple active sites to improve the catalytic performance, nano-Al2O3 is thus used in this study as the support for PAMAM dendrimers to produce dendritic rhodium catalysts. The resulting dendrimer-grafted nano-Al2O3supported rhodium catalysts have been used for styrene hydroformylation. This paper reports, for the first time, the employment of nanoscale alumina particles as supports which bring a profound promotional effect with the grafted dendrimer on the activity and regioselectivity of the catalyst for styrene hydroformylation. 2. Experimental Details 2.1. Chemicals. Aluminum nitrate, 3-aminopropyltriethoxysilane (APES), tetraethylorthsilicate (TEOS), methyl acrylate (MA), and ethylene diamine (EDA) were all purchased from Aldrich. Wilkinson’s catalyst (RhCl(PPh3)3) was purchased from Strem Chemicals, Inc. Reactions involving air-sensitive compounds were performed using a Schlenk line under a positive pressure of purified nitrogen. All the solvents were dried and distilled before use. 2.2. Synthesis of Nanoparticles of Al2O3. A 90.51 g portion of aluminum nitrate was added to.360 mL of ionized water in a 500 mL beaker and the mixture was stirred continuously using a magnetic stirrer. At the same time, aqueous ammonia was added to the solution slowly until the pH value was adjusted to 9.0. The slurry mixture was transferred into Teflon-lined autoclaves and heated in a furnace at 150 °C for 20 h. After heating, the slurry mixture was cooled and then separated by centrifuge and the collected solid was left to dry overnight at 80 °C. The solid product was then calcined at 550 °C for 12 h with a ramping rate of 10 °C/min. After calcination, the resulting nano-Al2O3 particles were stored in a glovebox for future use.

10.1021/ie800715k CCC: $40.75  2009 American Chemical Society Published on Web 01/27/2009

Ind. Eng. Chem. Res., Vol. 48, No. 4, 2009 1825

2.3. Synthesis of Nano-Al2O3-Supported Dendritic Catalysts. The functionalization of nano-Al2O3, R-Al2O3, γ-Al2O3, and SBA-15 with surface amine ligands, the grafting of PAMAM dendrimers on these functionalized supports, and the subsequent anchoring of RhCl(PPh3)3 on these supports follow those reported in the literature.22 A “Gn/support” symbol was used to designate these supported dendrimers as Gn signifies the generation of nth PAMAM dendrimer on the support. For example, G2/nano-Al2O3 means second generation PAMAM dendrimer supported on nano-Al2O3. The rhodium complex tethered on these nano-Al2O3, R-Al2O3, and γ-Al2O3 supports are named here as “Rh/Gn/support”. The synthesis of the SBA-15-based analogues were carried out based on the literature procedure.22 These SBA-15-supported catalysts are named here as WPSn, where “W” stands for Wilkinson’s catalyst (RhCl(PPh3)3), “P” for passivation of silanols outside the SBA-15 channels, “S” for SBA-15, and “n” for the generation of the PAMAM dendrimer of the catalyst. 2.4. Instrument. FE-SEM (field emission-scanning electron microscope) was used to characterize the synthesized nanoalumina and the dendrimer-grafted nanoalumina particles. The sample powders were coated with a layer of platinum before they were being analyzed by FE-SEM. The size and morphology of the synthesized supports could be determined using this method. In this study, the FE-SEM images were taken from JSM-6700F (JEOL) field emission scanning electron microscope. Gas chromatograph-flame ionization detection (Perkin-Elmer Auto System XL) was used to identify the reactant and product during or after reactions. The nonpolar capillary column used was HP-5MS (cross-linked 5% phenyl methyl siloxane) with a length of 30 m and a diameter of 0.25 mm. The hydroformylation reactions of styrene were carried out using 50 mL of 1-hexane solvent in a Parr reactor under 20 bar of CO/H2 gas mixture (CO:H2 ) 1:1) at 70 °C. 3. Results 3.1. FE-SEM Analysis. Figure 1 panels a, b, and c show the FE-SEM images of synthesized nano-Al2O3, surface aminefunctionalized nanoalumina (G0/nano-Al2O3), and first generation PAMAM grafted nanoalumina (G1/nano-Al2O3), respectively. It is obvious that these nanoparticles possess spherical morphology and their average diameters are around 50 nm. These nanoparticles have uniform crystal size and they do not show the tendency to agglomerate. The growth of PAMAM dendrimer on the surface of these nanoalumina particles does not cause the deformation of the spherical nanoparticles and does not disrupt the uniformity and scale of their crystal size. Hence, the synthesized nanoalumina-based PAMAM dendritic catalysts can be verified as nanoscale structured. 3.2. Comparison among Nano, r, γ-Alumina Supports. 3.2.1. Activity. Rh/G0/nano-Al2O3 and Rh/G1/nano-Al2O3 catalysts were applied to catalyze hydroformylation of styrene. To study the effect of nano support, R-alumina and γ-aluminasupported analogues were also synthesized and used for styrene hydroformylation for comparison. Figure 2 shows the catalytic results of Rh/G0/nano-Al2O3, Rh/G1/nano-Al2O3, Rh/G0/RAl2O3, Rh/G1/R-Al2O3, Rh/G0/γ-Al2O3, and Rh/G1/γ-Al2O3 catalysts as a function of reaction time on stream. It is obvious that both Rh/G0/nano-Al2O3 and Rh/G1/nano-Al2O3 catalysts are much more active than the R- and γ-alumina-supported analogues. Among all the alumina-supported rhodium catalysts, the Rh/G0/nano-Al2O3 catalyst achieved the highest conversion of styrene per rhodium atom within 2 h of reaction on stream, whereas the Rh/G0/R-Al2O3 catalyst exhibited the slowest

Figure 1. (a) FE-SEM image of nanoalumina support; (b) FE-SEM image of G0/nano-Al2O3; (c) FE-SEM image of G1/nano-Al2O3.

activity. Although the catalytic activity of Rh/G0/γ-Al2O3 is close to that of Rh/G1/nano-Al2O3, the catalytic activity of Rh/ G0/γ-Al2O3 still cannot be compared with those of Rh/G0/nanoAl2O3 even though these two types of catalysts differ only in the crystal size of their catalyst supports. It is also noted that, as the generation of PAMAM dendrimer goes up from zero to one, the nano and γ-alumina-supported catalysts show a reverse order of catalyst activity as compared to those of the R-alumina-supported ones. The catalyst activity decreases slightly from Rh/G0/nano-Al2O3 to Rh/G1/nano-Al2O3 and from Rh/G0/γ-Al2O3 to Rh/G1/γ-Al2O3, whereas the

1826 Ind. Eng. Chem. Res., Vol. 48, No. 4, 2009

Figure 2. Activity of hydroformylation of styrene catalyzed by nanoalumina, R-alumina, and γ-alumina-supported Wilkinson’s catalysts: [1/Rh] ) [mol of converted styrene/mol of rhodium in the supported catalyst].

Figure 3. Regioselectivity of hydroformylation of styrene catalyzed by nanoalumina, R-alumina, and γ-alumina-supported Wilkinson’s catalysts.

catalyst activity increases tremendously from Rh/G0/R-Al2O3 to Rh/G1/R-Al2O3. 3.2.2. Selectivity. Figure 3 compares the ratio of branch to linear aldehyde products as a function of styrene conversion over Rh/G0/nano-Al2O3, Rh/G1/nano-Al2O3, Rh/G0/R-Al2O3, Rh/G1/R-Al2O3, Rh/G0/γ-Al2O3, and Rh/G1/γ-Al2O3 catalysts. It is noted that both Rh/G0/nano-Al2O3 and Rh/G1/nano-Al2O3 catalysts show higher regioselectivity than the R- and γ-aluminasupported analogues. Among all the six alumina-supported catalysts, the Rh/G1/nano-Al2O3 catalyst shows the highest regioselectivity in producing the highest distribution toward 2-phenylpropionaldehyde product. The selectivities of these nanoalumina-supported catalysts are also consistent with their catalyst activities. 3.3. Comparison of Nanoalumina and SBA-15 Supports. 3.3.1. Activity. A comparison of the nanoalumina-supported PAMAM dendrimer-tethered Wilkinson’s catalysts with their non-nanoalumina based analogues show that the nanoscale supports produce significant promotional effects on both the catalyst’s activity and selectivity for styrene hydroformylation. Our recent reports on the catalytic performance of the passivated dendritic SBA-15-supported catalyst also show an enhanced catalytic performance of the heterogenized catalyst on the dendritic mesoporous SBA-15 support.21,22 Therefore a further comparison of the catalytic performance between these two robust catalyst supports for PAMAM dendrimer-tethered Wilkinson’s catalysts for hydroformylation reaction would be of interest and discussed below. Figure 4 comparatively shows the curves of the catalytic activity of dendritic nanoalumina-supported rhodium catalysts

Figure 4. Curves of the catalytic activity versus reaction time of nanoalumina and SBA-15-supported Wilkinson’s catalysts in hydroformylation of styrene.

Figure 5. Regioselectivity of hydroformylation of styrene catalyzed by nanoalumina and SBA-15-supported Wilkinson’s catalysts.

and passivated dendritic SBA-15-supported rhodium catalysts. The graph demonstrates that the Rh/G0/nano-Al2O3 and Rh/ G1/nano-Al2O3 catalysts are slightly more active than WPS1 (which stands for the Wilkinson’s catalyst anchored on the passivated SBA-15 grafted with first generation of dendrimer) and WPS0 (which stands for the Wilkinson’s catalyst anchored on the passivated SBA-15 grafted with zeroth generation of dendrimer), respectively.22 3.3.2. Selectivity. It has already been found that both the dendritic nanoalumina supports and the dendritic passivated SBA-15 supports have promotional effects on the activity of the catalysts for hydroformylation reaction of styrene.22 A comparison of the selectivity of rhodium catalysts anchored on these two novel dendritic supports can shed light on the properties of their nanostructured support and the effects of grafted dendrimer. Figure 5 comparatively shows the regioselectivity of Rh/G0/nano-Al2O3, Rh/G1/nano-Al2O3, WPS0, and WPS1 catalysts. It is clearly observed that the nanoaluminasupported catalysts are more selective than the SBA-15supported ones. 3.4. Exploration of Higher PAMAM Dendrimers on Nanoalumina. 3.4.1. Activity. Since Rh/G1/nano-Al2O3 shows higher regioselectivity than Rh/G0/nano-Al2O3, it is worthwhile to investigate the trends of both the activity and regioselectivity of these nanoalumina-supported rhodium catalysts tethered with higher generation of grafted PAMAM dendrimers. Therefore, higher PAMAM generation dendrimers were synthesized on these nanoalumina supports up to third generation. Wilkinson’s

Ind. Eng. Chem. Res., Vol. 48, No. 4, 2009 1827

Figure 6. Activity of hydroformylation of styrene catalyzed by nanoaluminasupported Wilkinson’s catalysts grafted with PAMAM dendrimer generation zero to three.

Figure 7. Regioselectivity of hydroformylation of styrene catalyzed by nanoalumina-supported Wilkinson’s catalysts grafted with PAMAM dendrimer generation zero to three.

catalysts tethered on these dendritic supports were then used to catalyze the hydroformylation of styrene under the same reaction conditions. Figure 6 shows the catalytic activities of Rh/G0/nano-Al2O3, Rh/G1/nano-Al2O3, Rh/G2/nano-Al2O3, and Rh/G3/nano-Al2O3 catalysts for styrene hydroformylation. It is noted that the catalytic activities of Rh/G2/nano-Al2O3 and Rh/G3/nano-Al2O3 catalysts fall in the middle of the activity of Rh/G0/nano-Al2O3 and Rh/G1/nano-Al2O3 catalysts. 3.4.2. Selectivity. As the Rh/G1/nano-Al2O3 catalyst shows higher regioselectivity than the Rh/G0/nano-Al2O3 catalyst, it is of great interest to investigate whether this increase of regioselectivity along with the increase of PAMAM dendrimer generation would continue and, if it is so, how far this synchronization may reach. Figure 7 shows the regioselectivity of Rh/G0/nano-Al2O3, Rh/G1/nano-Al2O3, Rh/G2/nano-Al2O3, and Rh/G3/nano-Al2O3 catalysts. The Rh/G0/nano-Al2O3 catalyst shows the lowest selectivity among these four catalysts. Similar to the activity curves, the selectivity of Rh/G2/nanoAl2O3 and Rh/G3/nano-Al2O3 catalysts fall in between the selectivity of Rh/G0/nano-Al2O3 and Rh/G1/nano-Al2O3 catalysts. 4. Discussions 4.1. Support Morphology. Figure 1 panels a-c show that nanoalumina supports can still be categorized as heterogeneous as they can be separated from the liquid phase through centrifuge. As the particle size falls in the nanoscale, the

nanoalumina supports possess large surface areas which are favorable for catalysis. In particular, their surface areas are larger than those of the conventional alumina supports.23 Since there are catalytic rhodium centers tethered at the outer surface of these nanoparticles, they can be more easily accessible than those tethered inside the mesopores of MCM-41, even though the latter also possess large surface areas.24,25 4.2. Effect of Nanoalumina Supports on the Catalytic Activity. Figures 2 and 4 show that the nanoalumina-supported catalysts exhibited the highest activity among all the tested catalysts. This observation is attributed to the effect of nanoalumina supports.26 Both the Rh/G0/nano-Al2O3 and Rh/G0/R-Al2O3 catalysts were functionalized with zeroth generation of PAMAM dendrimer and were synthesized using the same methods and steps from functionalization of surface ligands to tethering of Wilkinson’s catalysts, with the only difference attributed to the size of the support material. In the case of the R-alumina-based catalysts, their catalytic activity increases as the PAMAM generation increases from zero to one. The increased PAMAM generation brings greater homogeneity to the synthesized catalysts, thus enhancing the activity due to the elevated flexibility of dendrimer arms.27 However, for the nanoaluminasupported catalysts, the effect of nanosupports which are highly promotional to the catalyst activity seem somewhat obscured by the higher generation of PAMAM dendrimer.18,28,29 The Rh/G0/γ-Al2O3 catalyst is more active than the Rh/G1/ γ-Al2O3 catalyst, similar to that of the nanoalumina-supported ones where Rh/G0/nano-Al2O3 is more active than Rh/G1/nanoAl2O3. This phenomenon may be accounted for by the density of PAMAM dendrimer. Since both nanoalumina and γ-alumina are calcined under a relatively lower temperature than R-alumina, they possess more surface silanols than R-alumina. Therefore, the functionalized surface amine ligands are denser on the nanoalumina and γ-alumina supports than those on the R-alumina supports. When the PAMAM dendrimer increases from zeroth to first generation on the nanoalumina and γ-alumina supports, the supported dendrimers become more crowded than those grown on the R-alumina supports. The crowded dendrimers on the nanoalumina and γ-alumina supports tend to entangle and entwine among themselves, thus leading to the blocking of catalytically active rhodium centers by the entangled dendrimers on the nanoalumina and γ-alumina supports.30 For the Rh/G1/R-Al2O3 catalyst, the density of the first generation dendrimer grafted on the surface of R-Al2O3 is not so crowded to the point of triggering entanglement and blocking of active sites, thus showing a reverse order of the catalyst activity according to the generation of PAMAM dendrimer. For the comparison with SBA-15-supported catalysts (Figure 4), the small difference in activity between nanoalumina-based catalysts and SBA-15-based counterparts may be due to the diffusion resistance from the SBA-15 pores, where mass transfer becomes the controlling factor for the catalyst activity.31,32 Generally, the nanoalumina-supported catalysts are more active than their SBA-15-supported analogues. These two systems possess different order of activity according to the generation of grafted PAMAM dendrimer. Rh/G1/nano-Al2O3 is less active than Rh/G0/nano-Al2O3 for the dendritic nanoalumina-supported rhodium catalysts, whereas WPS1 is more active than WPS0 for the dendritic SBA-15-supported rhodium catalyst. This result indicates that the space inside the SBA-15 mesopore channels may provide a special microenvironment for PAMAM dendrimers to exhibit positive dendrimer effects.30,33 However, on

1828 Ind. Eng. Chem. Res., Vol. 48, No. 4, 2009

the nanoalumina supports, the grafted first generation PAMAM dendrimer makes the catalytic rhodium center bulky, thus inhibiting some activity of the rhodium catalytic center during styrene hydroformylation.34 4.3. Effect of Nanoalumina Supports on the Catalytic Selectivity. Figures 3 and 5 show that the nanoaluminasupported catalysts exhibited the highest selectivity among all the tested catalysts. This result is attributed to the support effects from dendritic nanoalumina, possibly due to the better and more efficient contact between the reactants and catalytic centers in the nanosupported catalysts. Since nanoscale catalysts possess high surface areas, all the catalytic active centers are dispersed well at the support surfaces. This may be one major cause for the superior catalytic performances shown by the nanoaluminasupported catalysts. It is interesting to observe that the nanoalumina-supported catalysts show a reverse order of selectivity as compared with their order of activity. The selectivity of the catalysts from Rh/ G0/nano-Al2O3 to Rh/G1/nano-Al2O3 increases slightly, whereas the activity of the catalysts from Rh/G0/nano-Al2O3 to Rh/G1/ nano-Al2O3 decreases slightly. The decrease of catalyst activity with the increase of PAMAM dendrimer generationsyet maintaining relative constant selectivitysis also observed by Reek et al.35 This may be due to the bulkier catalytic center of Rh/G1/nano-Al2O3. Accordingly, the activity decreases due to the more steric hindrance.30 However, in the case of R-aluminabased catalysts, the selectivity changes only slightly as the PAMAM generation increases from zero to one. It should be noted that the nanoalumina particles possess much higher surface areas than the R-alumina particles. Therefore, the dendrimers grown on R-alumina tend to be more crowded and entangled with each other. The first generation of PAMAM dendrimer may bring more homogeneity to Rh/G1/ R-Al2O3; however, the elongated dendrimer arms may intermingle to some extent, thus negating the positive dendritic effect in enhancing selectivity,36 whereas the high surface areas of nanoalumina supports provide more space for the dendrimers to grow on them, thus mitigating the congestion of dendrimer arms. In contrast to the case of the nanoalumina-supported catalysts, the sequence of the selectivity of γ-alumina-supported catalysts is the same as that of their activity, that is, Rh/G0/γ-Al2O3 > Rh/G1/γ-Al2O3. The decrease of selectivity from Rh/G0/γ-Al2O3 to Rh/G1/γ-Al2O3 is due to the overcrowding of surface dendrimers which not only block some active sites but also distort the conformation of the catalytic center favoring the branch aldehyde product. This overcrowding stems from the initial density of silanols of γ-alumina.37 Despite of the high density of grafted dendrimers on the support surface, the high surface area of Rh/G1/nano-Al2O3 provides the microenvironment even more spatio-restraint than that of Rh/G0/nano-Al2O3, leading to the positive dendrimer effect on selectivity.31,33 Among the four support materials (SBA-15, nanoalumina, γ-alumina, and R-alumina), the first two possess large surface areas while the latter two do not, and the first two supported catalysts also exhibit the highest activity. This result indicates that a large surface area can provide more efficient contact between reactants and active sites, thus producing more active catalysts. However, both γ-alumina and R-alumina-supported catalysts demonstrate higher selectivity than their counterparts supported by SBA-15 despite the latter possessing larger surface areas. For example, Rh/G1/R-Al2O3 has considerably high selectivity (Figure 3) but shows low activity (Figure 2), indicating that selectivity may be influenced but not controlled

by surface area. In other words, the activity of a heterogenized catalyst can be enhanced much more by surface area than its regioselectivity. This result is in good agreement with the literature which reports that the regioselectivity is controlled by the conformation of the catalytic center.38 It is also noted that the nanoalumina-supported catalysts and SBA-15-supported counterparts have the same order of selectivity according to the PAMAM generation. The selectivity of these two-catalyst systems increases with the increase of the generation of grafted dendrimer. For the SBA-15-based catalysts, this phenomenon is explained by the positive dendrimer effects which are mainly attributed to the better dispersion of rhodium complexes and the special reaction environment provided by the support. For the nanoalumina-based catalysts, it seems that the combination of PAMAM G1 dendrimer and the nanoalumina support provides a highly spatioselective environment for producing branch aldehyde.33,39 Without the pore restriction, higher generation of PAMAM dendrimers can be grown on nanoalumina for further investigation. The results will be discussed in the next section. 4.4. Role of PAMAM Dendrimer Generation on the Nanoalumina-Supported Catalysts. Figure 6 shows that there is a slight increase of catalytic activity from Rh/G1/nano-Al2O3 to Rh/G2/nano-Al2O3 catalysts; however, the increase of the activity cannot continue from Rh/G2/nano-Al2O3 to Rh/G3/ nano-Al2O3 catalysts. This is because from Rh/G1/nano-Al2O3 to Rh/G2/nano-Al2O3 the elongated dendrimer arms bring more homogeneity to the catalyst, whereas from Rh/G2/nano-Al2O3 to Rh/G3/nano-Al2O3 this effect is neutralized by the more severe blocking of active catalytic sites with increasing dendrimer length.18,40 It appears that there is a promotional effect from the nano-Al2O3 support to the activity of the catalytic center in Rh/G0/nano-Al2O3. However, this promotional effect only functions within a very short distance (around 10 Å) from the nano-Al2O3 surface. Over this limit the promotional effect drops significantly. This result is in contrast to that observed on the PAMAM dendrimer grafted on the SiO2 support, in which case the increase of dendrimer generation can increase the activity of the catalyst up to second generation,19 as the mass transfer resistance may possibly become prevalent on the catalysts grafted with dendrimer higher than second generation. However, this promotional effect by the higher generation of PAMAM dendrimers is saturated at the first generation of PAMAM dendrimer as Rh/G1/nano-Al2O3, Rh/G2/nano-Al2O3, and Rh/G3/nano-Al2O3 catalysts show very similar regioselectivity toward the branch aldehyde product. As previously stated, the nanoalumina was synthesized under a relatively low temperature and thus it possesses a lot of silanol groups. Although nanoalumina has high surface areas, the exponential increase of the grafted PAMAM dendrimers at higher generation would cause the surface of the catalyst to be crowded with dendrimer arms.41 The increased bulkiness of catalytic center can enhance selectivity but can also block active sites.42 Therefore the selectivity cannot be infinitely increased with higher generations of dendrimers as there is a threshold which governs the density of the surface ligands. Below this threshold, positive dendrimer effects are observed since there is some free space which allows the multiple catalytic centers to cooperatively function.43-45 Whereas, above this threshold, a further growth of the generation of grafted PAMAM dendrimer is no longer helpful for higher selectivity since the reaction environment becomes too crowded. Therefore, among the dendritic nanoalumina tethered Wilkinson’s catalysts, Rh/G1/nano-Al2O3 offers the best regioselec-

Ind. Eng. Chem. Res., Vol. 48, No. 4, 2009 1829

tivity for hydroformylation of styrene. Any higher generation of dendrimer for the purpose of increasing regioselectivity would only waste chemicals and time required for the synthesis of the catalyst. However, higher generation PAMAM dendrimers may bring about higher selectivity if they are anchored on nanoalumina supports smaller than the ones used in this study since the catalyst supports with smaller particle size would possess higher surface to volume ratios. 5. Conclusions Nanoalumina supports grafted with PAMAM dendrimers from zeroth to third generation have been successfully synthesized and used to tether Wilkinson’s catalyst to catalyze styrene hydroformylation. This method takes both advantages of the high surface area of nanoscale material, which is desirable for a heterogeneous catalyst support, and the unique properties of PAMAM dendrimers, which not only can tether catalytic metal complexes but also can bring positive dendrimer effects on the activity and selectivity of the resulting catalysts. This beneficial combination of PAMAM dendrimers and nanoalumina support has been shown for the first time in this study. Dendritic nanoalumina-supported rhodium catalysts exhibit superior catalytic performances, in terms of both activity and selectivity, to dendritic R-alumina, dendritic γ-alumina, or dendritic SBA-15-supported rhodium catalysts. Among all the supported Wilkinson’s catalysts, Rh/G0/nano-Al2O3 shows the highest activity while Rh/G1/nano-Al2O3 shows the highest regioselectivity. The high surface area of nanoalumina is found to be the main reason accounting for the high activity shown by Rh/G0/nano-Al2O3. Higher generations of PAMAM dendrimer would cause the entanglement of dendrimer arms and thus decrease the activity of the catalytic rhodium center. Although Rh/G1/nano-Al2O3 showed higher regioselectivity than Rh/G0/nano-Al2O3, Rh/G2/nano-Al2O3 and Rh/G3/nanoAl2O3 failed to increase the regioselectivity higher, possibly due to the crowdedness in the catalyst structure caused by the longer dendrimer arms of higher dendrimer generation. Acknowledgment The financial support from a research project (RP-279000232112) funded by the National University of Singapore and the Ministry of Education of Singapore is greatly appreciated. Literature Cited (1) Lenarda, M.; Storaro, L.; Ganzerla, R. Review Hydroformylation of Simple Olefins Catalyzed by Metals and Clusters Supported on Unfunctionalized Inorganic Carriers. J. Mol. Catal., A 1996, 111, 203. (2) Nozaki, K.; Itoi, Y.; Shibahara, F.; Shirakawa, E.; Ohta, T.; Takaya, H.; Hiyama, T. Asymmetric Hydroformylation of Olefins in a Highly CrossLinked Polymer Matrix. J. Am. Chem. Soc. 1998, 120, 4051. (3) Foley, H. C.; DeCanio, S. J.; Tau, K. D.; Chao, K. J.; Onuferko, J. H.; Dybowski, C.; Gates, B. C. Surface Organometallic Chemistry: Reactivity of Silica-bound Rhodium Allyl Complexes and the Genesis of Highly Dispersed Supported Rhodium Catalysts. J. Am. Chem. Soc. 1983, 105, 3074. (4) Reynhardt, J. P. K.; Yang, Y.; Sayari, A.; Alper, H. Periodic Mesoporous Silica-Supported Recyclable Rhodium-Complexed Dendrimer Catalysts. Chem. Mater. 2004, 16, 4095. (5) Mukhopadhyay, K.; Chaudhari, R. V. Heterogenized HRh(CO) (PPh3) 3 on Zeolite Y Using Phosphotungstic Acid as Tethering Agent: A Novel Hydroformylation Catalyst. J. Catal. 2003, 213, 73. (6) Alini, S.; Bottino, A.; Capannelli, G.; Comite, A.; Pananelli, S. Preparation and Characterisation of Rh/Al2O3 Catalysts and Their Application in the Adiponitrile Partial Hydrogenation and Styrene Hydroformylation. Appl. Catal., A 2005, 292, 105.

(7) Alvila, L.; Pakkanen, T. A.; Pakkanen, T. T.; Krause, O. Catalytic Hydroformylation of 1-Hexene over CO2Rh2(CO) 12 Supported on Inorganic Carrier Materials. J. Mol. Catal., A 1992, 75, 333. (8) Moffat, A. J. New Oxo Chemistry Via Solid Poplymer-Cobalt Carbonyl Complexes. J. Catal. 1970, 18, 193. (9) Lu, S. M.; Alper, H. Hydroformylation Reactions with Recyclable Rhodium-Complexed Dendrimers on a Resin. J. Am. Chem. Soc. 2003, 125, 13126. (10) Borrelli, N. F.; Hall, D. W.; Holland, H. J.; Smith, D. W. Quantum Confinement Effects of Semiconducting Microcrystallites in Glass. J. Appl. Phys. 1987, 61, 5399. (11) Gates, B. C. Supported Metal Clusters: Synthesis, Structure, and Catalysis. Chem. ReV. 1995, 95, 511. (12) Wong, E. W.; Sheehan, P. E.; Lieber, C. M. Nanobeam Mechanics: Elasticity, Strength, and Toughness of Nanorods and Nanotubes. Science 1997, 277, 1971. (13) Shen, S. C.; Hidajat, K.; Yu, L. E.; Kawi, S. Novel Nanocrystalline Ga-Al-Zn Complex oxide: Catalyst for Simultaneous Treatment of NPAC and Lean NOx. Catal. Today 2004, 98, 387. (14) Tour, J. M. Conjugated Macromolecules of Precise Length and Constitution. Organic Synthesis for the Construction of Nanoarchitectures. Chem. ReV. 1996, 96, 537. (15) Frechet J. M. J.; Tomalia D. A., Dendrimers and Other Dendritic Polymers; John Wiley & Sons: New York, 2001. (16) Chow, H. F.; Wan, C. W. Multicenter Homogeneous Dendritic Catalysts: The Higher the Generation, the Better the Reactivity and Selectivity?-AComparative Study of the Catalytic Efficiency of Dendrimeric [1,1-Binaphthalene]-2,2’-diol- Derived Catalysts. HelV. Chim. Acta 2002, 85, 3444. (17) Reek, J. N. H.; Arevalo, S.; Van Heerbeeken, R.; Van Leeuwen, P. W. N. M. Dendrimers in Catalysis, Advances in Catalysis. AdV. Catal. 2006, 49, 71. (18) Vogtle, F. Dendrimers III: Design, Dimension, Function; Topics in Current Chemistry, Vol 212; Springer: Berlin, 2001; p 3. (19) Bourque, S. C.; Maltais, F.; Xiao, W. J.; Tardif, O.; Alper, H.; Arya, P.; Manzer, L. E. Hydroformylation Reactions with RhodiumComplexed Dendrimers on Silica. J. Am. Chem. Soc. 1999, 121, 3035. (20) Hu, A.; Gordon, T. Y.; Lin, W. Magnetically Recoverable Chiral Catalysts Immobilized on Magnetite Nanoparticles for Asymmetric Hydrogenation of Aromatic Ketones. J. Am. Chem. Soc. 2005, 127, 12486. (21) Li, P.; Kawi, S. SBA-15-based Polyamidoamine Dendrimer Tethered Wilkinson’s Rhodium Complex for Hydroformylation of Styrene. J. Catal. 2008, 257, 23. (22) Li, P.; Kawi, S. Dendritic SBA-15 Supported Wilkinson’s Catalyst for Hydroformylation of Styrene. Catal. Today 2008, 131, 61. (23) Rong, M. Z.; Ji, Q. L.; Zhang, M. Q.; Friedrich, K. Graft Polymerization of Vinyl Monomers onto Nanosized Alumina Particles. Eur. Polym. J. 2002, 38, 1573. (24) Michal, K.; Jaroniec, M. Relations between Pore Structure Parameters and Their Implications for Characterization of MCM-41 Using Gas Adsorption and X-ray Diffraction. Chem. Mater. 1999, 11, 492. (25) Shephard, D. S.; Zhou, W. Z.; Maschmeyer, T.; Matters, J. M.; Roper, C. L.; Parsons, S.; Johnson, B. F. G.; Duer, M. J. Site-Directed Surface Derivatization of MCM-41: Use of High-Resolution Transmission Electron Microscopy and Molecular Recognition for Determining the Position of Functionality within Mesoporous Materials. Angew. Chem., Int. Ed. Engl. 1998, 37, 2719. (26) Shen, S. C.; Hidajat, K.; Yu, L. E.; Kawi, S. Simple Hydrothermal Synthesis of Nanostructured and Nanorod Zn-Al Complex Oxides as Novel Nanocatalysts. AdV. Mater. 2004, 16, 541. (27) Arena, C. G.; Drago, D.; Faraone, F. Hydroformylation of Styrene and 1-Octene Catalyzed by Binuclear and Oligomer Rhodium(I) Complexes Containing the Bis-p-Phosphinito Ligands[(p-Ph2POC6H4) 2X](X ) O, CMe2, S). J. Mol. Catal., A 1999, 144, 379. (28) Chow, H. F.; Mak, C. C. Dendritic Bis(ozazoline)copper(II) Catalysts. 2.1 Synthesis, Reactivity, and Substrate Selectivity. J. Org. Chem. 1997, 62, 5116. (29) Nithyanandhan, J.; Jayaraman, N. Synthesis and Reactivity Profiles of Phosphinated Poly(Alkyl Aryl Ether) Dendrimers. Tetrahedron 2005, 61, 11184. (30) Pruett, R. L.; Smith, J. A. A Low-Pressure System for Producing Normal Aldehydes by Hydroformylation of R Olefins. J. Org. Chem. 1969, 34, 327. (31) Che, S. N.; Lund, K.; Tatsumi, T.; Iijima, S.; Joo, S. H.; Ryoo, R.; Terasaki, O. Direct Observation of 3D Mesoporous Structure by Scanning Electron Microscopy (SEM): SBA-15 Silica and CMK-5 Carbon. Angew. Chem., Int. Ed. 2003, 42, 2182.

1830 Ind. Eng. Chem. Res., Vol. 48, No. 4, 2009 (32) Dufaud, V.; Davis, M. E. Design of Heterogeneous Catalysts via Multiple Active Site Positioning in Organic-Inorganic Hybrid Materials. J. Am. Chem. Soc. 2003, 125, 9403. (33) Zhang, Y.; Zhang, H. B.; Lin, G. D.; Chen, P.; Yuan, Y. Z.; Tsai, K. R. Preparation, Characterization and Catalytic Hydroformylation Properties of Carbon Nanotubes-Supported Rh-Phosphine Catalyst. Appl. Catal., A 1999, 187, 213. (34) Evans, D.; Osborn, J. A.; Wilkinson, G. Hydroformylation of Alkenes by Use of Rhodium Complex Catalysts. J. Chem. Soc. A 1968, 12, 3133. (35) Reek, J. N. H.; De Groot, D.; Oosterom, G. E.; Kamer, P C J.; Van Leeuwen, P. W. N. M. Phosphine-Functionalized Dendrimers for Transition-Metal Catalysis. C. R. Chim. 2003, 6, 1061. (36) Jiang, Y.; Gao, Q. Heterogeneous Hydrogenation Catalyses over Recyclable Pd(0) Nanoparticle Catalysts Stabilized by PAMAM-SBA-15 Organic-Inorganic Hybrid Composites. J. Am. Chem. Soc. 2006, 128, 716. (37) Bu, J.; Li, R.; Quah, C. W.; Carpenter, K. J. Propagation of PAMAM Dendrons on Silica Gel: A Study on the Reaction Kinetics. Macromolecules 2004, 37, 6687. (38) Strohmeier, W.; Michel, M. Kinetics and Product Distribution in Homogeneous Hydroformylation of 1-Hexene with RhHCO(PPh3) 3 without Solvent. Z. Phys. Chem. (Wiesbaden) 1981, 124, 23. (39) Abu-Reziq, R.; Alper, H.; Wang, D. S.; Post, M. K. Metal Supported on Dendronized Magnetic Nanoparticles: Highly Selective Hydroformylation Catalysts. J. Am. Chem. Soc. 2006, 128, 5279.

(40) Dvornic, P. R.; Claire, H. T.; Keinath, S. E.; Hill, E. J. OrganicInorganic Polyamidoamine (PAMAM) Dendrimer-Polyhedral Oligosilsesquioxane (POSS) Nanohybrids. Macromolecules 2004, 37, 7818. (41) Feuerbacher N.; Vo¨gtle F. Dendrimers, Springer: Berlin, Heidelberg, Germany, 1998; Chapter 1, p 3. (42) Caroline, R. N.; Nathalie, L. B.; Laurent, P.; Clement, J. C.; Des Abbayes, H. Rhodium-Catalyzed Hydroformylation of Styrene at Low Temperature Using Potentially Hemilabile Phosphite-Phosphonate Ligands. Tetrahedron Lett. 2001, 42, 643. (43) Arya, P.; Rao, N. V.; Singkhonrat, J. A. Divergent, Solid-Phase Approach to Dendritic Ligands on Beads. Heterogeneous Catalysis for Hydroformylation Reactions. J. Org. Chem. 2000, 65, 1881. (44) Reetz, M. T.; Lohmer, G.; Schwickardi, R. Synthesis and Catalytic Activity of Dendritic Diphosphane Metal Complexes. Angew. Chem., Int. Ed. Engl. 1997, 36, 1526. (45) Breinbauer, R.; Jacobsen, E. N. Cooperative Asymmetric Catalysis with Dendrimeric [Co(salen)] Complexes. Angew. Chem., Int. Ed. 2000, 39, 3604.

ReceiVed for reView May 2, 2008 ReVised manuscript receiVed September 11, 2008 Accepted September 18, 2008 IE800715K