Reduction of Supported Noble-Metal Ions Using Glow Discharge Plasma

HZSM-5 were reduced using an Ar glow discharge plasma. Optical absorption spectra and X-ray photoelectron spectroscopy show that the supported metal i...
0 downloads 0 Views 533KB Size
11388

Langmuir 2006, 22, 11388-11394

Reduction of Supported Noble-Metal Ions Using Glow Discharge Plasma Ji-Jun Zou,† Yue-ping Zhang,‡ and Chang-Jun Liu*,† Key Laboratory for Green Chemical Technology, School of Chemical Engineering and Technology, and Department of Chemistry, Tianjin UniVersity, Tianjin 300072, P. R. China ReceiVed June 22, 2006. In Final Form: October 3, 2006 A novel plasma reduction method has been developed to reduce supported noble-metal ions without the use of any reducing chemicals. H2PtCl6, PdCl2, AgNO3, and HAuCl4 supported on nonporous TiO2 and porous γ-Al2O3 and HZSM-5 were reduced using an Ar glow discharge plasma. Optical absorption spectra and X-ray photoelectron spectroscopy show that the supported metal ions are completely reduced to metallic species. Transmission electron microscopy shows that the prepared metals are amorphous clusters and homogeneously distributed with nanoscale sizes. X-ray diffraction also confirms that the plasma-reduced metals exist as small crystallites or amorphous clusters. Thermal annealing of plasma-reduced samples at elevated temperature transforms the clusters into crystals with a slight increase in particle sizes, but the sizes are still smaller than those of H2-reduced metals. O2 glow discharge plasma can also reduce noble-metal ions, accompanied by production of a small amount of oxides. Plasma reduction is very promising for the preparation of metal nanoparticles and supported metal catalysts.

Introduction Supported noble metals have been extensively used as catalytic materials in numerous areas such as petroleum chemistry, elimination of automobile exhaust, hydrogenation of carbon monoxide, and electrochemical reactions. The shape and size of metal particles are crucial to their catalytic performance. Small, flat nanoparticles are generally favored because a higher fraction of metal atoms are exposed at the surface where they are accessible to reactant molecules and available for reactions.1-3 The properties of supported metals are closely related to the preparation method. Supported metals are traditionally prepared by impregnating a porous or nonporous support material with the solution of metal salts, followed by reduction in hydrogen at elevated temperature.4 The resulting metal particles are usually heterogeneously dispersed with a broad size distribution due to particle aggregations at elevated temperature. Many methods have been developed to control the size distribution of metal particles. For example, chemical reduction uses reducing chemicals to produce homogeneously distributed metal particles at ambient temperatures. However, utilization of chemicals leads to environmental concerns.5,6 Photocatalytic reduction uses UV irradiation to replace dangerous chemicals, but the support materials are limited to semiconductiors.7,8 In addition, these wet processes are complex and the amount of metal loaded is difficult to control. Although there are many methods to prepare supported metals such as ion exchange, photodeposition, sol-gel, and colloidal approaches, * To whom correspondence should be addressed. E-mail: ughg_cjl@ yahoo.com. † School of Chemical Engineering and Technology. ‡ Department of Chemistry. (1) Haruta, M. CATTECH 2002, 6, 102. (2) Browker, M.; James, D.; Stone, P.; Bennett, R.; Perkins, N.; Millard, L.; Greaves, J.; Dickinson, A. J. Catal. 2003, 217, 427. (3) Molina, L. M.; Hammer, B. Phys. ReV. B 2004, 69, 155424. (4) Schmidt, F. Appl. Catal., A 2001, 221, 15. (5) Mallick, K.; Witcomb, M. J.; Scurrell, M. S. Appl. Catal., A 2004, 259, 163. (6) Deng, J.-F.; Li, H.; Wang, W. Catal. Today 1999, 51, 113. (7) Zhang, F.; Guan, N.; Li, Y.; Zhang, X.; Chen, J.; Zeng, H. Langmuir 2003, 19, 8230. (8) Chan, S. C.; Barteau, M. A. Langmuir 2005, 21, 5588.

they can all be generally classified as chemical or photoreduction methods according to the types of reducing reagents used.5-11 Here we present a novel cold plasma reduction method that does not require utilization of any reducing chemicals. Cold plasmas have previously been used to prepare supported metal nanoparticles. In these processes, cold plasmas are often used to dissociate relatively unreactive chemicals into reducing agents or accelerate the transformation of chemicals to reducing agents, thus shortening the reduction time. Microwave irradiation and acoustic cavitation, especially the former, have been applied.12,13 H2 plasmas have excellent abilities in reducing metal ions due to the strong reducing nature of the H radicals or atoms that are produced in the plasma. For example, noble-metal ions supported on TiO2-gel films are reduced with low-temperature H2 plasma.14 A plasma reduction of metal catalysts using a H2containing dielectric-barrier discharge plasma has also been reported.15,16 Recently, H2PtCl6 solution was reduced to Pt nanoparticles by hydrogen atoms produced in an AC discharge plasma.17 H2 plasmas are clean and dry compared with chemical and photochemical reductions. However, they still require the presence of H2 as reducing chemicals. Ultrasound irradiation has also been used to produce metals nanoparticles by decomposing H2O to H‚ radicals that then acted as the reducing reagent,18 which is very promising because no reducing chemicals are added. However, the method to be introduced here is different from the already developed plasma techniques. (9) Armelao, L.; Barreca, D.; Bottaro, G.; Gasparrotto, A.; Tondello, E.; Ferroni, M.; Polizzi, S. Chem. Mater. 2004, 16, 333. (10) Bonini, M.; Bardi, U.; Berti, D.; Neto, C.; Baglioni, P. J. Phys. Chem. B 2002, 106, 6178. (11) Bjerneld, E. J.; Svedberg, F.; Ka¨ll, M. Nano Lett. 2003, 3, 593. (12) Komarneni, S.; Li, D.; Newalkar, B.; Katsuki, K.; Bhalla, A. S. Langmuir 2002, 18, 5959. (13) Boxall, D. L.; Deluga, G. A.; Kenik, E. A.; King, W. D.; Lukehart, C. M. Chem. Mater. 2001, 13, 891. (14) He, J.; Ichinose, I.; Kunitake, T.; Nakao, A. Langmuir 2002, 18, 10005. (15) Kim, S.-S.; Lee, H.; Na, B.-K.; Song, H.-K. Catal. Today 2004, 89, 193. (16) Song, H. K.; Lee, H.; Choi, J.-W.; Yang, G.-S.; Kim, S.-S.; Na, B.-K. U.S. Patent 2005/0032626 A1. (17) Koo, I. G.; Lee, M. S.; Shim, J. H.; Ahn, J. H.; Lee, W. M. J. Mater. Chem. 2005, 15, 4125. (18) Pol, V. G.; Grisaru, H.; Gedanken A. Langmuir 2005, 21, 3635.

10.1021/la061795b CCC: $33.50 © 2006 American Chemical Society Published on Web 11/17/2006

Reduction of Supported Noble-Metal Ions

Langmuir, Vol. 22, No. 26, 2006 11389

Figure 1. Preparation of support metals using plasma reduction.

Figure 2. Optical adsorption spectra for plasma-reduced metals: (a) 3.0% Pt/TiO2, (b) 2.0% Pd/TiO2, (c) 2.0% Ag/TiO2, (d) 2.0% Au/TiO2.

We previously utilized an Ar glow discharge plasma to treat supportedmetalsinordertoimprovetheircatalyticperformance.19-24 We found that some metal ions exist in their metallic states after plasma treatment, which suggests that plasma treatment can reduce metal ions. It has also been reported that Nb2O5 was partially reduced to Nb4+ and Nb2+ with microwave plasma treatment.25 These phenomena suggest that cold plasma may be a potential reduction method for metal ions as shown in Figure 1, which would be very attractive and promising because no reducing chemicals would be necessary. Compared with current reduction methods, plasma reduction would be much greener. In addition, plasma reduction is simple, easy to manipulate, and compatible with the commonly used impregnation processes. Thus, it is expected that plasma reduction might open a new door for the preparation of metal nanoparticles and supported metal catalysts. (19) Liu, C.-J.; Zou, J.-J.; Yu, K.-L.; Cheng, D.-G.; Zhan, J.; Ratanatawanate, C.; Jang, B. W.-L. Pure Appl. Chem. 2006, 78, 1227. (20) Zou, J.-J.; Liu, C.-J.; Zhang, Y.-P. Langmuir 2006, 22, 2334. (21) Liu, C.-J.; Yu, K.-L.; Zhu, X.-L.; Zhang, Y.-P.; He, F.; Eliasson, B. Appl. Catal., B 2004, 47, 95. (22) Zou, J.-J.; Liu, C.-J.; Yu, K.-L.; Cheng, D.-G.; Zhang, Y.-P.; He, F.; Du, H.-Y.; Cui, L. Chem. Phys. Lett. 2004, 400, 520. (23) Yu, K.-L.; Liu, C.-J.; Zhang, Y.-P.; He, F.; Zhu, X.-L.; Eliasson, B. Plasma Chem. Plasma Process. 2004, 24, 393. (24) Zou, J.-J.; Chen, C.; Liu, C.-J.; Zhang, Y.-P.; Han, Y.; Cui, L. Mater. Lett. 2005, 59, 3437. (25) Sugiyama, K.; Aana, G.; Shimada, T.; Ohkoshi, T.; Ushikubo, T. Surf. Coating Technol. 1999, 112, 76.

In the present work, we use a glow discharge to reduce supported noble-metal ions and then characterize the formed samples, attempting to develop a plasma reduction method for supported noble-metal ions. Experimental Section The glow discharge plasma setup is similar to that described previously.19-24 The glow discharge plasma was initiated using a dc high-voltage generator (Tianda Cutting and Welding Setup Inc. Ltd., China; alternatively, a TREK 20/20C high-voltage amplifier with HP 33120A signal generator has also been applied). The plasma chamber was a quartz tube (i.d. 35 mm) with two stainless steel electrodes (o.d. 30 mm). The gap between the electrodes was 170 mm. The chamber was evacuated to 100 Pa by a vacuum pump. Ar (99.9%) or O2 (99.0%) with a flow rate of 10 mL/min was introduced and served as the plasma-forming gas. The discharge voltage and current were measured using an oscilloscope (Tektronix TDS210) equipped with a high-voltage probe (Tektronix P6015A) and a current probe (Pearson Electronics 411). When the voltage reached 0.9-1.2 kV, bright glow discharge plasma was formed between the two electrodes, where the samples were placed for reduction. The current was in the range of 1-2 mA. The gas temperature of the plasma was slightly higher than the ambient temperature but less than 100 °C as measured by infrared imaging (Ircon, 100PHT). Four noble metals, Pt, Pd, Ag, and Au, supported on nonporous TiO2 and porous γ-Al2O3 and HZSM-5 were studied. The metal precursors including H2PtCl6, PdCl2, AgNO3, and HAuCl4 (99.5%

11390 Langmuir, Vol. 22, No. 26, 2006

Zou et al.

Figure 3. Deconvoluted Pt4f XPS spectra for plasma-reduced 2.0% Pt/TiO2: (a) 10-min reduction, (b) 30-min reduction, (c) 60-min reduction, (d) plot of degree of reduction vs time. purity, Tianjin Guangfu Institute of Fine Chemicals, China) and support materials were used as received. Metal ions were loaded on the support materials using conventional impregnation by an aqueous solution containing a given amount of the metal precursor. Then they were reduced using the glow discharge plasma. The reduction time was 60 min if not otherwise specified. Some of the plasmareduced samples were annealed in an Ar atmosphere at a given temperature for 2 h. Samples reduced with hydrogen at 300 °C for 2 h were also prepared for the purpose of comparison. UV-visible absorption spectra were recorded using a JASCO V-570 spectrometer with a BaSO4 plate as the reference. X-ray photoelectron spectroscopy (XPS) measurements were performed using a Perkin-Elmer PHI-1600 spectrometer with Mg KR (1253.6 eV) radiation. The analyzer band-pass energy was 187.7 eV for survey spectra and 29.3 eV for high-resolution spectra. The pressure in the detection chamber was 1.2 × 10-8 Torr. The binding energy was calibrated using the C1s peak (284.6 eV) of the surface carbon. The estimated standard deviation for binding energy was (0.2 eV. Transmission electron microscopy (TEM) observations were conducted using a Philips Tecnai G2 F20 electron microscope. X-ray diffraction (XRD) analysis was conducted using a Rigaku D/Max2500 V/PC diffractometer with Cu KR radiation (λ ) 0.154178 nm). The voltage was 40 kV, and the current was 150 mA.

Result and Discussion Reduction Using Ar Plasma. During plasma treatment, the samples showed a significant change in color. Supported Pt and Pd turned from yellowish to gray, whereas Ag turned black, and Au looked a little pink. These color changes give evidence of

the change from metal ions to metallic particles. Figure 2 shows the absorption spectra for the plasma-reduced metals. Plasmareduced Pt/TiO2 and Pd/TiO2 exhibit enhanced absorption in the range of 380-800 nm. Such spectral changes have been attributed to formation of Pt and Pd nanoparticles.26,27 Plasma-reduced Ag/TiO2 exhibits an absorption peak around 480 nm that can be assigned to the surface plasmon resonance peak of the spatially confined electrons in Ag nanoparticles.28 Similarly, a surface plasmon resonance peak of Au nanoparticles is observed around 545 nm for the plasma-reduced Au.14 These absorption spectra indicate that metal nanoparticles are present after plasma reduction. Figure 3 shows the deconvoluted Pt4f XPS spectra of Pt/TiO2 as a function of reduction time. The binding energy of the metal ion is 76.3 and 72.9 eV for the Pt4f5/2 and Pt4f7/2 transitions, respectively. After 10 min of reduction, two new peaks with binding energies of 74.4 and 71.1 eV appear, which can be assigned to metallic Pt4f5/2 and Pt4f7/2, respectively.29,30 This suggests the existence of reduced metals. The percentage of reduced metal was calculated to be 45.7% based on the peak (26) Yonezawa, T.; Imamura, K.; Kimizuka, N. Langmuir 2001, 17, 4701. (27) Teranish, T.; Miyake, M. Chem. Mater. 1998, 10, 594. (28) Stathatos, E.; Lianos, P.; Falaras, P.; Siokou, A. Langmuir 2000, 16, 2398. (29) Yang, B.; Lu, Q.; Wang, Y.; Zhuang, L.; Lu, J.; Liu, P.; Wang, J.; Wang, R. Chem. Mater. 2003, 15, 3552. (30) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-ray photoelectron spectroscopy; Chastain, J., Ed.; Perkin-Elmer Corp.: Eden Prairie, MN, 1992.

Reduction of Supported Noble-Metal Ions

Langmuir, Vol. 22, No. 26, 2006 11391 Table 1. Chemical States of the Supported Metals Reduced with Ar Plasma samples 0.5% Pt/TiO2 2.0% Pt/TiO2 3.0% Pt/TiO2 2.0% Pd/TiO2 3.0% Pd/Al2O3 2.0% Pd/HZSM-5 2.0% Ag/TiO2 2.0% Au/TiO2

Figure 4. XPS spectra for plasma-reduced metals: (a) 3.0% Pd/ Al2O3, (b) 2.0% Ag/TiO2, (c) 2.0% Au/TiO2.

areas. As the reduction time increases, the peaks for the reduced metal grow while those for the metal ions shrink. When the reduction time was increased to 60 min, the peaks for the metal ion disappeared and only those of metal were present, indicating that the supported metal ions have been completely transferred into their metallic states. The percentage of metallic atoms is plotted as a function of reduction time. It is clear that the degree of reduction increases quickly as the reduction time increases. Sixty minutes is enough to completely reduce supported metal ions to their metallic states. Therefore, in the following experiments all plasma reductions were conducted for 60 min. Figure 4 shows the XPS spectra for the plasma-reduced Pd, Ag, and Au. Two symmetrical peaks are observed for each sample, and no other peaks can be deconvoluted. This indicates that these metals are in a single chemical state. Their binding energy

transition peak Pt4f7/2 Pt4f7/2 Pt4f7/2 Pd3d5/2 Pd3d5/2 Pd3d5/2 Ag3d5/2 Au4f7/2

BE (eV) 70.7 70.7 70.8 334.6 334.6 334.8 367.7 83.4

transition peak Pt4f5/2 Pt4f5/2 Pt4f5/2 Pd3d3/2 Pd3d3/2 Pd3d3/2 Ag3d3/2 Au4f5/2

BE (eV) 74.1 74.0 74.1 340.0 339.6 340.2 373.5 87.1

and corresponding assignment are summarized in Table 1. According to reference data, they can be assigned to the electron transitions for metallic Pd, Ag, and Au.7,30,31 These results confirm that the supported metal ions have been reduced to their metallic states. The binding energy of the reduced metal is identical for various amounts of loading, and it does not vary when the metal is loaded on different support materials, suggesting that reduction may be an intrinsic nature of the plasma. The binding energy of the plasma-reduced metals is slightly lower than those for dispersed or bulk metals, probably because of the different microenvironments and structures of the reduced metals. Figure 5 shows the TEM images of the plasma-reduced platinum. Pt nanoparticles are homogeneously distributed on the surface of the support. The average particle size is 3.8 nm for 3.0% Pt/TiO2. The selected area diffraction (SAD) pattern shows three Debye-Scherrer rings corresponding to the (111), (200), and (220) faces of the fcc platinum lattice, confirming the existence of Pt particles. However, the lattice fringes of the metal nanoparticles are difficult to observe in the high-resolution image. These metal particles are composed of many small crystallites or amorphous clusters. Energy-dispersive X-ray (EDX) analysis shows that these clusters contain Pt without H2PtCl6 residue. Ti and O are from the support material. This confirms that the amorphous clusters are metallic platinum. Figure 6 shows TEM images of the plasma-reduced Pd, Ag, and Au. The SAD patterns of the supported Pd show the (111), (200), and (220) faces of the fcc palladium lattice. These nanoparticles are homogeneously distributed on the support materials. The mean size of Pd particles is 2.1 nm. Ag and Au nanoparticles exhibit slightly larger sizes of 6.1 and 6.7 nm, respectively. Structure and Crystallization of Plasma-Reduced Metals. Figure 7 shows the XRD patterns of the plasma-reduced metals. The diffraction patterns of the reduced metals are continuous and broad with weak intensities. This suggests that these nanoparticles are composed of amorphous clusters or small crystallites,32-34 which is consistent with the TEM observations. It is interesting because metal clusters often exhibit enhanced catalytic performance due to lattice defects and distortions. However, in some cases perfect metal crystals are more favorable. Thus, the plasma-reduced metals were crystallized by annealing the samples in Ar at a given temperature for 2 h. The XRD patterns of the crystallized samples are also shown in Figure 7. The diffraction peaks of the metals became stronger as the annealing temperature was increased. After being annealed at 500 °C, the diffraction patterns for the plasma-reduced metals (31) Arabatzis, I. M.; Stergiopoulos, T.; Andreeva, D.; Kitova, S.; Nephytides, D. G.; Falaras, P. J. Catal. 2003, 220, 127. (32) Lu, W.; Wang, B.; Wang, K.; Wang, X.; Hou, J. G. Langmuir 2003, 19, 5887. (33) Yu, X.; Wang, M.; Li, H. Appl. Catal., A 2000, 202, 17. (34) Fang, J.; Chen, X.; Liu, B.; Yan, S.; Qiao, M.; Li, H.; He, H.; Fan, K. J. Catal. 2005, 229, 97.

11392 Langmuir, Vol. 22, No. 26, 2006

Zou et al.

Figure 5. TEM characterizations of plasma-reduced 3.0% Pt/TiO2: (a) TEM image, (b) size distribution, (c) high-resolution TEM image, (d) EDX spectrum of cluster in c.

were equal to those of the H2-reduced counterparts. The crystallinity of the supported Pt and Pd was calculated using the intensities of the (111) and (200) peaks

crystallinity (%) ) [I(111) + I(200)]plasma-reduced/[I(111) + I(200)]H2-reduced × 100%

Figure 6. TEM images and size distribution of plasma-reduced metals: (a and b) 3.0% Pd/Al2O3, (c and d) 2.0% Ag/TiO2, (e and f) 2.0% Au/TiO2.

The results are presented in Figure 8. The crystallinity of the plasma-reduced metals is extremely low. Thermal annealing at an evaluated temperature can improve the crystallinity that can be as high as 100% when annealed at 500 °C. High-resolution TEM images of a crystallized Pt and Pd nanoparticle are also shown in Figure 8. Well-defined lattice fringes of Pt(111) and Pd(111) are observed, indicating that the particles are uniformly crystallized. The crystallinity of Ag and Au are difficult to determine because their peaks partially overlap with the support material. Thermal annealing at the evaluated temperature also induces aggregation of metal particles. The particle sizes determined by TEM are shown in Figure 9. Although annealing at a high temperature induces the undesirable increase of particle size, the particle sizes are still smaller than those of the conventionally H2-reduced counterpart, indicating that plasma reduction is more favorable than conventional reduction for enhancing the distribution of metal nanoparticles. Reduction Using O2 Plasma. To determine whether plasma reduction is related to the plasma-forming gas, we also used oxygen as the plasma-forming gas to treat the supported metal ions. Figure 10a shows the XPS spectra of the O2-plasma treated metals. No precursor-derived elements are detected, indicating that no metal precursors survive after plasma treatment. Thus, the supported metal ions are changed to other chemical states. The Pd3d XPS spectrum was deconvoluted to determine the chemical state of the O2-plasma-treated metals, as shown in Figure 10b. Each deconvoluted spectrum contains two pairs of transition peaks. The binding energies of the prominent pair are in good

Reduction of Supported Noble-Metal Ions

Langmuir, Vol. 22, No. 26, 2006 11393

Figure 7. XRD patterns of (a) 3.0% Pt/TiO2, (b) 3.0% Pd/Al2O3, (c) 2.0% Ag/TiO2, and (d) 2.0% Au/TiO2: (i) plasma reduced. (ii) annealed at 200 °C. (iii) annealed at 300 °C. (iv) annealed at 500 °C. (v) H2-reduced at 300 °C.

Figure 9. Particle size of supported metals.

Figure 8. Crystallization of plasma-reduced metal particles: (a) crystallinity as a function of annealing temperature, (b) TEM image of Pt annealed at 500 °C, (c) TEM image of Pd annealed at 500 °C.

agreement with those for the reduced metal, confirming the existence of the reduced metal as the major component. The

minor pair is assigned to the oxidized metal. The possibility of the metal precursor is excluded because no other precursorderived elements are detected. Figure 11 shows the XRD patterns of the plasma-treated and annealed Pd. The diffraction patterns of the O2-plasma-treated sample are similar to those of the Ar plasma. Only diffraction peaks of metallic Pd are observed. No palladium oxide peaks are observed because the amount of oxides is below the detection limit of the XRD technique. Therefore, reduction is the dominant effect when supported metal ions are treated by O2 plasma. It has been shown that both Ar and O2 plasmas can effectively reduce supported metals ions to their metallic states. This process is independent of the amount of loading, the type of support

11394 Langmuir, Vol. 22, No. 26, 2006

Zou et al.

Figure 11. XRD patterns of plasma-reduced 2.0% Pd/HZSM-5 with annealing at 300 °C; the unmarked peaks belong to HZSM-5.

one characteristic for all kinds of plasmas, which is independent of the type plasma-forming gas. As the only reducing species in the plasma, these electrons most likely serve as the reducing agents. Thus, plasma reduction may be a direct transfer of electrons from the plasma to the metal ions. In O2 plasma, active oxygen atoms are also produced, which produce some oxides. Further investigations are necessary to clarify the mechanisms of plasma reduction.

Conclusions

Figure 10. XPS spectra for 2.0% Pd/HZSM-5 reduced with O2 plasma: (a) survey spectrum, (b) Pd3d spectrum.

materials, and the plasma-reforming gas, indicating that plasma reduction may be universal for noble-metal ions. This is totally different from traditional reduction methods because no conventional reducing agents are added or formed. When reduction occurs, there must be some reducing agents or species to serve as electron donors. An abundance of high-energy electrons is

The present work has presented a green plasma reduction method without the assistance of reducing chemicals. H2PtCl6, PdCl2, AgNO3, and HAuCl4 supported on porous and nonporous materials are reduced to their metallic states by an Ar glow discharge plasma. The degree of reduction increases with the increase of reduction time, and complete reduction is achieved in 60 min. The resulting metal nanoparticles are homogeneously distributed with nanoscale sizes. These nanoparticles are composed of amorphous clusters or small crystallites, but crystallization occurs when they are annealed at elevated temperatures. The metal particle size increases slightly after thermal annealing but is still smaller than the conventionally reduced sample. O2 plasma is as effective as Ar plasma but accompanied by a small amount of oxidation. This green plasma reduction may be used as a new reduction method for the preparation of highly dispersed metal nanoparticles or metalsupported catalysts. Acknowledgment. This work was supported by the National Natural Science Foundation of China (20225618) and the 973 Project (2005CB221406). The instrument support from ABB Switzerland and the Program for Changjiang Scholars and Innovative Research Team from the Ministry of Education of China are also appreciated. LA061795B