Langmuir 2005, 21, 2121-2123
2121
Pd Nanoparticle Aging and Its Implications in the Suzuki Cross-Coupling Reaction Jun Hu* and Yubiao Liu† Department of Chemistry, the University of Akron, Akron, Ohio 44325-3601 Received November 15, 2004. In Final Form: January 24, 2005 Examination of the catalysts recovered in the N,N-dihexylcarbodiimide-palladium nanoparticle composite catalyzed Suzuki cross-coupling reactions revealed that the metal nanoparticles transformed gradually from spherical-shape to larger needle-shaped crystals. Two types of Ostwald ripening processes were observed. One involves rapid aggregation of the incipient nanoparticle catalyst (2-5 nm) into blackberrylike assemblies (100-200 nm), which is accompanied with the much slower dissolution of small crystals or amorphous nanoparticles and the formation of larger needle-shaped crystals. The observed structural changes provided new insights into the durability of the polymer nanoparticle composite catalyst.
Polymer stabilized noble metal nanoparticles have attracted much attention recently as a new research direction in catalysis.1 The polymer matrixes serve as the supporting materials for keeping the nanoparticles from aggregation, provide the desired chemical interfaces between the nanoparticles and the reaction media, and facilitate the synthesis and reuse of the catalysts. The lifetime or durability of the nanoparticles in chemical reactions is of a fundamental interest in the applications of nanoparticle catalysts, including fuel cells, sensors, and “green chemistry” processes.2 We recently reported that poly(N,N-dihexylcarbodiimide) (PDHC), a helical backbone synthetic polymer,3 is useful for synthesizing and stabilizing nanoparticles of noble metals such as Au, Pt, and Pd (Scheme 1).4 We also reported the use of one of the composites, namely, poly(N,N-dihexylcarbodiimide)/palladium nanoparticle composite (PDHC-Pd), in the Suzuki cross-coupling reactions with conventional and microwave heating (Scheme 1).5-7 The catalyst was prepared by NaBH4 reduction of H2PdCl4 in a two-phase mixture of toluene and water with PDHC as the stabilizer.4 When the polymer was used in the in situ synthesis of noble metal nanoparticles, the kinetically formed nanoparticles in the polymer matrixes by rapid reduction of the metal salts were found to be nearly spherical by transmission electron microscopy (TEM), regardless of the cylindrical micelle morphology of the polymer and the lengths of the polymer chain.4,5 We also found that the same results could be obtained when * To whom correspondence should be addressed. E-mail: jhu@ uakron.edu. † Current address: Department of Polymer Science, the University of Akron, Akron, OH 44325. (1) (a) Mayer, A. B. R.; Mark, J. E. Macromol. Rep. 1996, A33, 451. (b) Toshima, N. Nanoscale Materials 2003, 79. (c) Herron, N.; Thorn, D. L. Adv. Mater. 1998, 10, 1173. (d) Li, Y.; Hong, X.; Collard, M. D. M.; El-Sayed, M. A. Org. Lett. 2000, 2, 2385. (e) Narayanan, R.; ElSayed, M. A. J. Am. Chem. Soc. 2003, 125, 8340. (f) Kim, S.-W.; Kim, M.; Lee, W. Y.; Hyeon, T. J. Am. Chem. Soc. 2002, 124, 7642. (2) Huang, J.; Jiang, T.; Gao, H.; Han, B.; Liu, Z.; Wu, W.; Chang, Y.; Zhao, G. Angew. Chem., Int. Ed. 2004, 43, 1397. (3) Shibayama, K.; Seidel, S. W.; Novak, B. M. Macromolecules 1997, 30, 3159. (4) Liu, Y.; Cheng, S. Z. D.; Wen, X.; Hu, J. Langmuir 2002, 18, 10500. (5) Liu, Y.; Khemtong, C.; Hu, J. Chem. Commun. 2004, 398. (6) (a) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457. (b) Suzuki, A. J. Organomet. Chem. 1999, 576, 147. (c) Wolfe, J. P.; Singer, R. A.; Yang, B. H.; Buchwald, S. L. J. Am. Chem. Soc. 1999, 121, 9550. (7) (a) Gedye, R.; Smith, F.; Westaway, K.; Ali, H.; Baldisera, L.; Laberge, L.; Rousell, J. Tetrahedron Lett. 1986, 27, 279. (b) Zhang, A.; Neumeyer, J. L. Org. Lett. 2003, 5, 201.
N2H4 was used as the reducing agent in the two-phase synthesis or in toluene alone. Polymers with metal binding basic functionalities such as poly(4-vinyl pyridine)8 and poly(N-vinyl-2-pyrrolidone) have been used to prepare narrowly dispersed noble metal nanoparticles.9 The polymers usually form polyeletrolyte species in the reaction media and, therefore, are electrostatically charged. Aggregation of metal colloidal particles supported in the polymer matrixes is minimized by the steric hindrance and electrostatic repulsion. Most likely, PDHC stabilizes the metal nanoparticles by a similar mechanism. It is an amphiphilic material with a hydrophilic and basic backbone. In typical reaction media, the backbone is hydrated to form a polyelectrolyte while the side chains are soluble in the organic solvent. The utility of the catalyst is shown here in the microwave heated Suzuki coupling reaction between phenylboronic acid and bromobenzene (eq 1). Excellent conversions were achieved in about 40 min of microwave irradiation. The composite is used as a colloidal solution, and it can be conveniently recovered by precipitation and filtration. The recovered catalyst showed retention of most of the catalytic activities, and up to five reuses of the recovered catalyst were demonstrated in a simulated batch reaction process (Figure 1). With 40 min of microwave heating in each cycle, we observed lower yields for each of the consecutive reaction cycles, indicating a small yet significant decrease in the catalytic activity of the recycled catalyst. No catalytic activity was detected from the filtrate solution. Pd leaching out of the polymer into the reaction solution is, therefore, negligible.
To investigate the structure/activity properties of the PDHC-Pd nanoparticle catalyst, particularly the effect of aging of the nanoparticles, we recovered the catalysts (8) Roucoux, A.; Schulz, J.; Patin, H. Chem. Rev. 2002, 102, 3757. (9) Teranishi, T.; Kiyokawa, I.; Miyake, M. Adv. Mater. 1998, 10, 596.
10.1021/la0471902 CCC: $30.25 © 2005 American Chemical Society Published on Web 02/12/2005
2122
Langmuir, Vol. 21, No. 6, 2005
Letters
Scheme 1. Synthesis of the Polymer/Pd Nanoparticle Catalyst and the Catalyst Aging in the Reactiona
a
Subjects are not drawn to scale.
Figure 1. Isolated yields of Suzuki coupling reactions between phenylboronic acid and bromobenzene versus number of recycles of the catalyst under microwave heating at 100 °C for 40 min.
and examined morphological changes of the Pd nanoparticles with TEM (Figure 2). Image 1 in Figure 2 shows the catalyst recovered after a single batch reaction for 40 min under microwave heating. The nanoparticles aggregated into nanosized “blackberries”, as a result of the Ostwald ripening similar to what was reported by Narayanan et al.1e Image 4 shows the catalyst recovered after four batch reactions (4 × 40 min). Interestingly, the nearly spherical and amorphous aggregated particles in images 1 and 2 (Figure 2) became irregular and started to display sharp crystalline edges. The polydispersity of the nanoparticles also increased. Image 5 shows the recovered catalyst after five batch reactions (5 × 40 min) with microwave heating. Images 4 and 5 indicated a second type of ripening process of the nanoparticle in the reaction. In this second type of ripening process, atomic rearrangement occurred and the nanoparticles transformed from microcrystalline/amorphous into needle-shaped nanocrystals. The observed two types of ripening processes were also evident from the observed product yields in each catalyst recycle, which showed a steep decline in the catalytic activity of the composite in the first and second runs. The decline in the catalytic activity seemed to level off afterward (Figure 1). The catalytic activity of noble metal nanoparticles is wellknown to be size-dependent.10 The above observation indicated that the crystal morphology of the catalyst may also affect the reactivity under the reaction conditions. The above experimental results, particularly the TEM observations, prompted us to propose the following mech(10) Schmid, G. Chem. Rev. 1992, 92, 1709.
anism for the PDHC-Pd catalyzed Suzuki coupling reactions and the aging process of the nanoparticles (Scheme 2). The Suzuki coupling reaction is considered highly corrosive, that is, involving both Pd(II) and Pd(0) species in the catalytic cycle.5,7 The reactants also have a high amount of ionic species. The initial chemical absorption of the arylhalide on the palladium(0) nanoparticle surface is believed to lead to the oxidative insertion which can etch away the Pd atom and form a discrete Pd(II) complex. The fact that we observed only crosscoupling rather than homo-coupling reactions supports the existence of the proposed discrete Pd(II) intermediate.5 The transmetalation step most likely occurs in a discrete Pd species similar to that in the homogeneous catalysis.7 During the reductive elimination process, the discrete Pd(II) complex is reduced back to Pd(0). There are two possibilities for regenerating Pd(0): the first is the formation of discrete Pd(0) atoms in the polymer support, and the second is the deposition of the Pd(0) atoms back onto the Pd nanoparticles. Our experimental results indicated that the second pathway is dominant, although the first possibility is not ruled out completely. The TEM observation of the nanoparticles becoming needle-shaped crystals after repeated uses in the reaction supported the repeated catalytic cycles between the discrete Pd(II) species and the solid-state Pd(0). Murphy and co-workers have shown that controlling nanoparticle shape is feasible in a seeded growth method for generating elongated crystalline Au nanorods.11 The repeated cycles of etching and deposition of the Pd(0) on the nanoparticle surface mimics the slow and thermodynamically controlled crystallization process. The shape of the resulting crystals seemed to suggest specific interactions between the polymer and the resulting nanoparticles or discrete species in the catalytic cycles. Without surface-selective ligands or surfactants, the metal (111) surface usually has the lowest energy and the resulting thermodynamic-controlled nanocrystals should maximize this surface by the formation of truncated polyhedrons. The initial aggregation of the nanoparticles is most likely due to the large increase in the ionic strength when the reactants were mixed. It is well-known that the zeta potentials of the colloidal particles are weakened by the high dielectric and ionic strength of the media. We did not observe such aggregation for the metal nanoparticle (including Au, Pd, and Pt) in the polymer composites for (11) Jana, N. R.; Gearheart, L.; Murphy, C. J.; J. Phys. Chem. B 2001, 105, 4065.
Letters
Langmuir, Vol. 21, No. 6, 2005 2123
Figure 2. TEM images of the recovered PDHC-Pd catalyst after 1, 2, 4, and 5 uses in the microwave-heated Suzuki cross-coupling reaction. Scheme 2. Proposed Mechanism of the Suzuki Cross-Coupling Catalytic Cycle
several months when the materials are dry or in toluene (anhydrous) as it was reported previously.4 Control experiments by heating the nanoparticle composite in the same solvent mixture alone did not produce the nanocrystal morphology changes observed in the above Suzuki coupling reaction. Polymer matrixes used for stabilizing the nanoparticles are not sufficiently strong transition metal ligands, and polymer metal complexes are usually insignificant in the reaction mixture compared to metal nanoparticles. Because the nanoparticle-catalyzed reactions are usually 1-3 orders of magnitude lower in TONs compared to those catalyzed by the corresponding homogeneous metal catalysts and larger nanoparticles are catalytically inactive,5,12 it is difficult to rule out the possibility that the observed (12) Li, J.-H.; Liu, W.-J. Org. Lett. 2004, 6, 2809.
catalytic activity is due to the formation of a small amount, yet extremely active discrete Pd(0) complexes in the reaction mixture. PDHC are considered weaker coordinative ligands than the prevailing polymers such as poly(vinylpyridine) used in Suzuki coupling previously.8 For the reactive intermediates in the catalytic processes, the weaker coordination results in higher activity of the catalysts. For the nanoparticles, weaker coordinative ligands are less corrosive, and, therefore, the slower redistribution of the nanoparticles to the larger size occurs in the reactions. In summary, we observed for the first time in a transition metal nanoparticles catalyzed C-C bond formation reaction that (1) small metal nanoparticles aggregated rapidly during the initial phase of the catalytic reactions; (2) small nanoparticles and the corresponding amorphous aggregates dissolved slowly, and large crystalline nanoparticles were formed when further aggregation of the nanoparticles was prohibited; and (3) the selection of the polymer ligand for the catalyst is critical for achieving optimum activity and durability of the catalysts by preventing aggregation. The observation provided new critical experimental parameters for design and optimization of durable nanoparticle catalysts. Acknowledgment. This work was supported by the U.S. National Science Foundation through the NER program (DMR0210508). J.H. thanks the University of Akron Research Foundation for a startup grant and a faculty research fellowship. We thank Prof. S. Z. D. Cheng and Dr. A. Jin at the Polymer Institute of the University of Akron for help with the TEM analysis. Supporting Information Available: The original TEM images of the nanoparticle catalysts recovered in the Suzuki coupling reactions and in the control experiment. This material is available free of charge via the Internet at http://pubs.acs.org. LA0471902
