Manipulating the Self-Assembling Process to Obtain Control over the

Department of Chemistry and Kanbar Laboratory for Nanomaterials, Bar-Ilan ... AdVanced Materials and Nanotechnology, Bar-Ilan UniVersity, Ramat-Gan ...
0 downloads 0 Views 1MB Size
Langmuir 2007, 23, 5971-5977

5971

Manipulating the Self-Assembling Process to Obtain Control over the Morphologies of Copper Oxide in Hydrothermal Synthesis and Creating Pores in the Oxide Architecture Ziyi Zhong,*,† Vivien Ng,† Jizhong Luo,† Siew-Pheng Teh,† Jaclyn Teo,† and Aharon Gedanken*,‡ Institute of Chemical and Engineering and Sciences, 1 Pesek Road, Jurong Island, Singapore 627833, and Department of Chemistry and Kanbar Laboratory for Nanomaterials, Bar-Ilan UniVersity Center for AdVanced Materials and Nanotechnology, Bar-Ilan UniVersity, Ramat-Gan 52900, Israel ReceiVed NoVember 16, 2006. In Final Form: March 28, 2007 Copper oxide with various morphologies was synthesized by the hydrolysis of Cu(ac)2 with urea under mild hydrothermal conditions. In the synthesis, a series of organic amines with one or two amine groups (monoamine and diamine), including isobutylamine, octylamine (OLA), dodecylamine, octadecylamine (monoamines), ethylenediamine dihydrochloride, and hexamethylenediamine (diamines), was used as the “structure-directing agent”. The monoamines led to the formation of one-dimensional (1D) aggregates of the copper oxide precursor particles (Pre-CuO), while the diamines led to the formation of two-dimensional (2D) aggregates. In both cases, the shorter carbon-chain amine molecules showed a stronger structure-directing function than that of the longer carbon-chain amine molecules. Next, in a series of syntheses, OLA was selected for further study, and the experimental parameters were systematically manipulated. When the hydrolysis was adjusted to a very slow rate by coupling the hydrolysis reaction with an esterification reaction, 1D aggregates of Pre-CuO were formed; when the hydrolysis rate was in the middle range, spherical Pre-CuO architectures composed of smaller linear aggregates were formed. However, under the high hydrolysis rates achieved by increasing the precipitation agent (urea) or by conducting the reaction at high temperatures (g120 °C), only Pre-CuO nanoparticles with a featureless morphology were formed. The formed spherical Pre-CuO architectures can be converted to a porous structure (CuOx) after removing the OLA molecules via calcination. Compared to the 1D and 2D aggregates, this porous architecture is highly thermally stable and did not collapse even after calcination at 500 °C. Preliminary results showed that the porous structure can be used both as a catalyst support and as a catalyst for the oxidation of CO at low temperatures.

Introduction The synthesis of inorganic materials with manipulated morphology, size, and textural structure is of great interest in materials science and technology because, in many cases, the properties of the synthesized/fabricated materials are dependent on these parameters.1,2 For example, in catalysis, porous catalysts usually have a larger surface area and a higher catalytic activity than nonporous catalysts. In the natural world, many highly sophisticated structures/entities are self-assembled from their building blocks. Over the last two decades, this self-assembly route has been extensively explored in supramolecular chemistry3 and materials synthesis.4 Recently, we succeeded in synthesizing various Au colloidal assemblages using this route.5 However, many self-assembly processes are quite complicated, and a variety of factors can have an impact on them. Both intrinsic and extrinsic factors (e.g., crystal structure and the growth environment) have a significant influence on the final morphology and structure of * Corresponding authors. E-mail: [email protected] (Z.Z.); [email protected] (A.G.). † Institute of Chemical and Engineering and Sciences. ‡ Bar-Ilan University. (1) Mann, S.; Ozin, G. A. Nature 1996, 382, 313. (2) Manna, L.; Scher, E. C.; Li, L. S.; Alivisatos, A. P. J. Am. Chem. Soc. 2002, 124, 7136. (3) Whitesides, G. M.; Grzybowski, B. Science 2002, 295, 2418. (4) Lehn, J. M. Science 2002, 295, 2400. (5) (a) Zhong, Z.; Patskovskyy, S.; Bouvrette, P.; Luong, J. H. T.; Gedanken, A. J. Phys. Chem. B 2004, 108, 4046. (b) Zhong, Z.; Subramanian, A. S.; Highfield, J.; Carpenter, K.; Gedanken, A. Chem.sEur. J. 2005, 14, 1126. (c) Zhong, Z.; Chen, F.; Subramanian, A. S.; Lin, J.; Highfield, J.; Gedanken, A. J. Mater. Chem. 2006, 16, 489-495.

the products.6 Generally, growth speeds on different crystal facets are either different or can be adjusted by the selective functionalization of the surfaces of the small particles with some organic molecules. These molecules can adsorb on certain crystal facets, thus leading to a preferential growth/stacking along certain directions,7,8 or they can cross-link the small particles along certain directions.5 However, one prerequisite for obtaining control over the growth direction is that the nucleation and condensation process of the small particles must be well separated; otherwise, branching among the small particles will occur in all directions. In the hydrolysis of Ti isopropoxide, by slowing down the water supply via coupling it with an esterification reaction and by using octylamine (OLA) as a directing agent, we separated the hydrolysis and condensation steps of the highly active Ti isopropoxide and prepared one-dimensional (1D) and porous TiO2.9 Further manipulation of the growth kinetics led to the formation of macroporous TiO2 spheres.10 Recently, we extended this approach to a variety of inorganic oxides.11 However, besides the aforementioned factors, other experimental conditions, for example, reaction temperature, concentration, and different types of organic amines, should be explored. By analyzing and varying (6) (a) Liu, X. Y.; Boek, E. S.; Briels, W. J.; Bennema, P. Nature 1995, 374, 342. (b) Xu, J.; Xue, D. J. Phys. Chem. B 2006, 110, 7750. (7) Xia, Y.; Yang, P.; Sun, Y.; Wu, Y.; Mayers, B.; Gates, B.; Yin, Y.; Kim, F.; Han, F. AdV. Mater. 2003, 15, 353. (8) Chemseddine, A.; Moritz, T. Eur. J. Inorg. Chem. 1999, 235. (9) Zhong, Z.; Ang, T. P.; Luo, J.; Gan, H. C.; Gedanken, A. Chem. Mater. 2005, 17, 6814. (10) Zhong, Z.; Chen, F.; Ang, T. P.; Han, F. Y.; Lim, W.; Gedanken, A. Inorg. Chem. 2006, 45, 4619. (11) Zhong, Z.; Lin, M.; Ng, V.; Ng, G. X. B.; Foo, Y. L.; Gedanken, A. Chem. Mater. 2006, 18, 6031.

10.1021/la063344x CCC: $37.00 © 2007 American Chemical Society Published on Web 05/01/2007

5972 Langmuir, Vol. 23, No. 11, 2007

Zhong et al.

Figure 1. Influence of various organic amines on the morphology of the Pre-CuO product. In panels A, B, and C, monoamine molecules were used (isobutylamine was used in A and B, and dodecylamine was used in panel C), and in panels D, E, and F, diamine molecules were used (ethylenediamine dihydrochloride was used in D and E, and hexamethylenediamine was used in F). In all the experiments, the molar ratio of amine/Cu was 2:1, and the reaction temperature and time were 80 °C and 20 h, respectively.

these experimental conditions systematically in this study, we synthesized a range of novel copper oxide frameworks, including spherical and porous structures, linear aggregates, and nanoparticles, resulting in a deep insight into how these factors influence the morphologies/structures of the final products. Copper oxides are rich in their crystal framework and structure, resulting in numerous investigations into their properties.12 More importantly, they have many applications in the semiconductor industry and in catalysis. The present work provides a simple and practical approach to obtaining a variety of copper oxide assemblages/structures. Their catalytic activity for the catalytic oxidation of CO at low temperatures is also explored. Experimental All chemicals were purchased from Aldrich (Sigma-Aldrich Pte. Ltd., Singapore) and used as received. In a typical synthesis, 1.0 g of copper acetate (Cu(ac)2) was dissolved in 10 mL of deionized (DI) H2O, after which 1.5 g of urea and 1.0 mL of OLA were added. The solution was stirred for several minutes and then transferred to an autoclave lined with a Teflon vessel. The autoclave was kept in an oven at 80 °C for 20 h. After the completion of the reaction, the precipitate was washed with acetone or anhydrous ethanol and dried in a vacuum oven at 100 °C. The Au/CuOx catalyst was prepared using our own developed method.5 First, 0.50 g of copper oxide was put in 10 mL of DI H2O, after which 3.0 mL of 0.01 M hydrogen tetrachloroaurate (III) (HAuCl4) and lysine were added. The pH value of the suspension was adjusted to 5-6 with 0.10 M of NaOH. The suspension was subjected to sonication for 20 s, during which freshly prepared NaBH4 was injected instantly. The suspension changed in color immediately from the previous light yellow to dark and was washed and centrifuged four times with DI H2O. The measurement of the catalytic oxidation of CO was carried out in a fixed-bed microreactor. Prior to the test, the catalyst was pretreated under air at 200 °C for 1 h. A reactant gas containing 1% CO under air was passed through the catalyst bed at a gas hourly space velocity of 15 000 h-1. The outlet gases were analyzed with an on-line gas chromatograph (Shimadzu-14B). (12) (a) Chang, Y.; Zeng, H. C. Cryst. Growth Des. 2004, 4, 273. (2) Chang, Y.; Zeng, H. C. Cryst. Growth Des. 2004, 4, 397. (c) Liu, B.; Zeng, H. C.; Chang, Y.; Zeng, H. C. J. Am. Chem. Soc. 2004, 126, 8124. (d) Chang, Y.; Lye, M. L.; Zeng, H. C. Langmuir 2005, 21, 3746. (e) Teo, J. J.; Chang, Y.; Zeng, H. C. Langmuir 2006, 22, 7369.

Figure 2. Influence of the molar ratio of OLA/Cu on the morphology of the Pre-CuO product: (A) 0:1; (B) 2:1. The shape and size of the copper oxide precursor particles (PreCuO) and CuOx samples and the Au/CuOx catalyst were observed with a transmission electron microscope (TEM, FEI-Philips Tecnai TF 20 S-twin, 200kV) and a scanning electron microscope (SEM, JEOL JSM-6700F field emission SEM). The powder X-ray diffraction (XRD) analysis was conducted on a Bruker D8 Advance (Bruker AXS, Inc., Madison, WI) X-ray diffractometer with CuKR1 radiation. Thermal analysis (thermogravimetric analysis/differential thermal analysis (TGA-DTA)) was performed on an SDT 2960 instrument (SDT 2960, TA Instruments, Madison, WI) under air flow. The Brunauer-Emmett-Teller (BET) surface area was measured at a liquid nitrogen temperature on a Quantachrome Autosorb-6B surface area and pore size analyzer (Quantachrome Instruments). X-ray photoelectron spectroscopy (XPS) spectra were recorded on an ESCALAB 250 spectrometer (Thermo Fisher Scientific, Inc.), and Al KR radiation was used as the X-ray source. The C1s peak at 284.5 eV was used as a reference for the calibration of the binding energy scale.

Results and Discussion The influence of various amines on the morphology/structures of the Pre-CuO oxide products is shown in Figure 1. With the use of the monoamines, 1D aggregates of the Pre-CuO particles

Synthesis of Various Morphologies of CuO

Langmuir, Vol. 23, No. 11, 2007 5973 Scheme 1. Schematic Illustration of the Formation of Pre-CuO with Various Morphologies/Structures.

Figure 3. Influence of the reaction temperature (T) and time (t) on the morphology/dimension of the Pre-CuO product. (A) T ) 80 °C and t ) 20 h; (B) T ) 120 °C and t ) 20 h. (C,D) T ) 80 °C and t ) 4 and 30 h, respectively. In these experiments, the molar ratio of OLA/Cu was fixed at 2:1.

can be formed, but their length decreases with an increase in the carbon chain of the monoamines. For example, when isobutylamine was used in the synthesis, the length of the formed 1D aggregates was in the range of 0.5-1.5 µm (Figure 1A,B), while, for dodecylamine (Figure 1C), the length of the 1D aggregates was in the range of 50-150 nm. It was seen in this sample (Figure 1C) that, although the 1D structures can be discerned, their edges were either unclear or still not fully developed. Perhaps when the carbon chain length is too long, the solubility of the monoamine becomes low in the aqueous solution. As a result, it is difficult for these molecules to adsorb on the surface of the Pre-CuO particles, thus weakening their “structure-guiding” function. When diamines are used in the synthesis, a similar phenomenon is observed. Shorter diamine molecules lead to the formation of linear aggregates (Figure 1D,E), while longer diamine molecules lead to the formation of a mixture of the small particles and non-fully developed 1D nanostructures (Figure 1F). As opposed to the monoamine molecules, diamine molecules facilitate the formation of two-dimensional (2D) aggregates because they can guide the assembly of the particle in two directions (Figure 1D,E). It is clear that, in both cases, the shorter carbon-chain amine molecules exhibit a stronger structure-guiding function than that of the longer-chain amine molecules. Therefore, in the following experiments, only OLA was selected for further studies. OLA molecules have a medium length carbon chain and are representative of the amine molecules. The morphology of the Pre-CuOx product is also highly dependent on the molar ratio of OLA/Cu and urea/Cu. Figure 2 shows the TEM images of Pre-CuO prepared at various OLA/ Cu ratios. In the absence of OLA molecules, the Pre-CuO product consists of diamond-shaped plates with sizes ranging from 50 to 200 nm. With the introduction of the OLA molecules in the synthesis, the 1D nanostructure gradually appears. When OLA/ Cu reached 0.5, about 10-20% of the product became the 1D structure, while the remainder were still diamond-shaped plates. When OLA/Cu reached 1:1, almost 100% of the product was the 1D structure. Figure 2B presents the Pre-CuO product prepared

(I) Monoamine molecules used as the structure-directing agent. (a) Pre-CuO nanoparticles form when the hydrolysis of Cu(ac)2 is fast under a high reaction temperature (see Figure 2B), or under a high urea/Cu ratio (g15; Supporting Information, SI-1). In addition, when the OLA/Cu ratio is high (g7), the Pre-CuO particles are fully covered with OLA molecules, thus blocking the branching among them (OLA molecules act as a capping agent). (b) Large spherical architectures of Pre-CuO composed of smaller linear aggregates are formed when the Rhydrolysis of Cu(ac)2 is in the middle range and OLA molecules partially function as a “structure-directing” agent. (c) 1D aggregates are formed when Rhydrolysis is very slow, as it is coupled with an esterification reaction for the water supply. The precipitation and condensation steps are well separated in this case, thus the role of the OLA molecules to guide the formation of the 1D structure is fully expressed. (d) Diamond-shaped plates are formed in the absence of the OLA molecules. The resulting Pre-CuO plates are well crystallized (Figure 1A). (II) Diamine molecules are used as the structure-directing agent. 2D aggregates can be formed with the guidance of the diamine molecules.

at OLA/Cu ) 2:1. The 1D structures bundle together to form an olive-like architecture. As proved in our previous reports,9,10,11 the OLA molecules act as a structure-directing agent that guides the growth of the particles along a certain direction. However, when the OLA/Cu is up to 7:1, only small particles are observed. The reason for this will be discussed later. In the synthesis, the role of urea is to cause the hydrolysis of Cu(ac)2 by releasing OH- groups. Without it, almost no precipitate was obtained. In the range of urea/Cu ) 1.7-5:1 and in the presence of OLA molecules (OLA/Cu ) 2), an olive-like architecture composed of 1D aggregates was obtained. However, when urea/Cu was increased to 15:1 (see the Supporting Information, SI-1), only small particles were obtained. In the following experiments, the ratio of OLA/Cu was kept at 2:1, and the ratio of urea/Cu was kept at 3:1, unless mentioned otherwise. The dimension and morphology of the Pre-CuO product can be further manipulated by varying reaction temperature and time.

5974 Langmuir, Vol. 23, No. 11, 2007

Zhong et al.

Figure 4. TEM images of the copper oxide: before (A,B) and after (C) the calcination at 300 °C. (D) SEM image of panel C. Mesopores were created after the calcination. The sample was prepared at 80 °C for 20 h with a molar ratio of OLA/Cu ) 2:1.

Figure 5. TEM image of the Pre-CuO product prepared at T ) 80 °C, t ) 10 h, and a molar ratio of OLA/Cu ) 2:1. The water needed for the hydrolysis reaction was slowly released from the esterification reaction between acetic acid (20 mL) and isopropanol (3 mL). It should be pointed out that, in this experiment, only a small sample was obtained for TEM measurement.

At a low temperature range between 60 and 100 °C, a bundled 1D structure was formed (Figure 3A). The diameter of these 1D structures was ca. 6-8 nm. When the reaction temperature exceeded 120 °C, only Pre-CuOx particles with an irregular shape and a size ranging from 30-50 nm were formed (Figure 3B). The aspect ratio of the Pre-CuO is highly dependent on reaction time. At a short reaction time (0.5-4 h), some short nanorods could be observed, but they were packed randomly (Figure 3C).

A further increase in the reaction time led to the formation of a more orderly packed architecture (Figure 3D), and the shape of the architecture evolved from a sphere-like to an olive-like structure. The length of the 1D structure in the architecture was increased to 100-200 nm (Figure 3D) from the previous 50100 nm (Figure 3C). However, in all the 1D structures, the diameter in a cross-section direction was almost the same, suggesting that the 1D structures were formed by assembling/

Synthesis of Various Morphologies of CuO

Langmuir, Vol. 23, No. 11, 2007 5975

Figure 7. The XRD patterns and the XPS spectrum of the copper oxide calcined at 300 °C for 1 h.

Figure 6. TGA-DTA profile and FTIR spectra of the 1D Pre-CuO product that was previously shown in Figure 1A.

reorganizing the small particles that were initially formed. This conclusion was further confirmed by our TEM observation under high resolution (Supporting Information, SI-2). It was clearly seen that the 1D structure was composed of small particles, and not a continuous crystal lattice (Supporting Information, SI-2B). This indicates that the small particles were formed first, but later were reorganized into 1D structures under the guidance of the OLA molecules. The cross-sectional view of one of the architectures (Figure 4D) indicates that the small particles repacked themselves linearly (1D structure). The variation in the aforementioned experimental parameters is attributed to manipulating the kinetics of the hydrolysis reaction and the condensation process so as to influence the final morphology and structure of the Pre-CuO product. There are three rates that can influence this: (1) the hydrolysis rate of Cu(ac)2 (Rhydrolysis), (2) the capping rate of the OLA molecules (Rcap), and (3) the condensation rate (Rc). When the reaction temperature or urea/Cu is high, Rhydrolysis is faster than Rcap. Most of the formed particles will branch together over a very short time (leading to a high Rc), and the guiding function from the OLA molecules will be lost. In the case of a high OLA/Cu ratio (g7:1), the formed Pre-CuO particles are capped densely with OLA molecules (Rcap is faster than Rc). In this case, all Pre-CuO crystal facets are blocked by the OLA molecules, and thus the branching (by forming a Cu-O-Cu bond) along certain crystal facets becomes difficult. As a result, the 1D structure cannot be formed. Therefore, Rhydrolysis, Rcap, and Rc should be carefully manipulated so as to achieve good control over the morphology and structure of the final product. This is illustrated in Scheme 1.

According to the above analysis, in order to obtain a welldefined 1D structure of Pre-CuO, it is necessary to further slow down Rhydrolysis and separate the hydrolysis and condensation stages but magnify the “guiding” function of the OLA molecules. One experiment was designed for this purpose: the water used for the hydrolysis of Cu(ac)2 was released from the esterification reaction between acetic acid and isopropanol. By controlling the reaction temperature and the concentration of the acid and the alcohol, the speed of the water release can obviously be lowered, thus slowing down the hydrolysis of Cu(ac)2. This strategy was also used in our previous study to slow down the hydrolysis of Titanium isopropoxide.9,10 The TEM image of the prepared PreCuO is shown in Figure 5. Clearly, it forms a very good 1D bamboo-like structure, revealing a self-assembly mechanism for the small Pre-CuO particles. The existence of the OLA molecules in the Pre-CuO was confirmed by TGA-DTA and Fourier transform infrared (FTIR) results. The as-prepared 1D Pre-CuO was in its precursor form and contained a copious level of H2O and OLA molecules. From the TGA measurement (Figure 6A), the total weight loss was ca. 54% after calcination at 300 °C. As indicated by XRD results, a mixture of Cu2O and CuO was formed after calcination at 300 °C. Assuming that Cu(OH)2 is the 1D Pre-CuO product and the final product is Cu2O or CuO, the weight loss should be 26.6% and 18.5%, respectively, much less than 54%. Therefore, both hydrated H2O and OLA ligand molecules should exist in the 1D Pre-CuO, even after being washed thoroughly with acetone or ethanol. The derivative weight curve exhibits two sharp peaks at 191 and 223 °C and one small peak at 308 °C. Figure 6B shows the IR spectra of OLA and Pre-CuO that were calcined at various temperatures. In the spectra, the bands at 2927 and 2853 cm-1 are assigned to C-H vibrations, and the bands at 3367, 3290, and 1602 cm-1 are assigned to N-H vibrations.13 These C-H and N-H vibration bands are

5976 Langmuir, Vol. 23, No. 11, 2007

Zhong et al.

Figure 8. TEM images of a 1 wt % Au/CuOx catalyst.

Figure 9. Oxidation of CO in air.

marked by dotted lines in Figure 6B. Clearly the C-H and N-H vibrations are seen, even when the sample was treated at 200 °C, indicating that the OLA molecules were not removed after being washed with acetone and ethanol, and even after being calcined at 200 °C. Most probably there is chemical or hydrogen bonding between the OLA molecules and the copper hydroxide particles,14 similar to that in TiO2, where the amine adducts were formed.15 Therefore, the peaks at 191 and 223 °C in the DTA curve (Figure 6A) should be assigned to the loss of H2O and OLA molecules, respectively. The XRD pattern for the calcined sample at 300 °C is shown in Figure 7A. Surprisingly, the sample calcined at 300 °C for 1 h was a mixture of Cu2O and CuO. The measured XPS spectrum shown in Figure 7B has two strong satellites at 940 and 945 eV.16-18 These satellites are absent in the Cu2O and metallic Cu species.17,18 Therefore, it seems that the Cu2O phase only exists in the bulk, not on the surface, because XPS is basically a surfaceprobing technique. The Cu2O phase should be formed during the calcination process, as we did not detect the Cu+1 species by XPS for the uncalcined Pre-CuO (Supporting Information, SI3). In the calcination process, the OLA molecules were (13) (a) Socrates, G. Infrared Characteristic Group Frequencies, 2nd ed.; Wiley & Sons: New York, 1994; pp 74-79. (b) Thimmaiah, S.; Rajamathi, M.; Singh, N.; Bera, P.; Meldrum, F. C.; Chandrasekhar, N.; Seshadri, R. J. Mater. Chem. 2001, 11, 3215. (14) Li, Y.; Afzaal, M.; O’Brien, P. J. Mater. Chem. 2006, 16, 2175. (15) Fric, H.; Schubert, U. New J. Chem. 2005, 29, 232. (16) Poulston, S.; Parlett, P. M.; Stone, P.; Bowker, M. Surf. Interface Anal. 1996, 24, 811. (17) Rodriguez, J. A.; Hrbek, J. J. Vac. Sci. Technol., A 1994, 12, 2140. (18) Rodriguez, J. A.; Kim, J. Y.; Hanson, J. C.; Pe´rez, M.; Frenkel, A. I. Catal. Lett. 2003, 85, 247.

decomposed, and some gas molecules such as CO were produced. These gas molecules could reduce Pre-CuO and CuO. Meanwhile, on the surface, the Cu2O phase was oxidized into CuO because it was exposed to oxygen during calcination. To confirm this, longer calcination was conducted at 300 °C for 4 h, leading to the formation of a pure CuO phase. The removal of the OLA molecules from the Pre-CuO architectures is expected to produce mesopores in CuOx, as in the case of TiO2.9,19 Figure 4C,D shows TEM and SEM images of the calcined architectures at 300 °C. Compared to the sample without calcination, some newly created pores are observed after the calcination, while the spherical architecture is still maintained. The measured BET surface area after calcination was ca. 140 m2/g. Here it should be emphasized that this spherical architecture has a high thermal stability and is stable even after calcination at 500 °C. Meanwhile, the aforementioned 1D and 2D aggregates were not thermally stable and collapsed after calcination at 300 °C. Potentially, this highly thermally stable and porous nanostructure has applications in catalysis. Next we used this porous and thermally stable spherical architecture as a catalyst and as a catalyst support. Supported Au catalysts have been under intense examination since they were reported on by Haruta et al.20 for their extraordinarily high activity for the oxidation of CO at low temperatures. Recently, it was reported that copper oxide itself has a good activity for this reaction.21 Therefore, we prepared a Au/CuO oxide catalyst, and one of the TEM images for this catalyst is shown in Figure 8. The Au particles deposited on the CuOx support are well controlled in the 2-5 nm range. Figure 9 shows that the porous CuOx exhibits good activity for CO oxidation, and its temperature for the 100% conversion of CO is 145 °C. This good catalytic activity is probably related to the reversible oxygen storage and release of CuO via the chemical reaction of CuO T Cu2O. After being deposited with the Au particles, the catalytic activity is further improved, and a 100% conversion of CO is achieved at ca. 100 °C. This activity is comparable to that of 1% Au/TiO2-P25, but lower than that of 1% Au/(mesoporous TiO2).9 To summarize, we have prepared copper oxide with various morphologies by the hydrolysis of Cu(ac)2 with urea under mild hydrothermal conditions. In the synthesis, various monoamines and diamines are used as the structure-guiding agent. Monoamines usually lead to the formation of 1D aggregates, but the length (19) Wang, Y.; Tang, X.; Yin, L.; Huang, W.; Hacohen, Y. R.; Gedanken, A. AdV. Mater. 2000, 12, 1183. (20) (a) Haruta, M.; Yamada, N.; Kobayashi, T.; Lijima, S. J. Catal. 1989, 115, 301. (b) Valden, M.; Lai, X.; Goodman, D. W. Science 1998, 281, 1647. (21) Pillai, U. R.; Deevi, S. Appl. Catal., B: EnViron. 2006, 64, 146.

Synthesis of Various Morphologies of CuO

of the aggregates decreases with an increase in the carbon chain length of the monoamines. Diamines usually cause the formation of 2D aggregates because their two amine groups can independently direct the assembly of the Pre-CuO particles. OLA was then chosen for further study. In a series of syntheses, the experimental parameters are systematically manipulated so as to reveal the guiding function of the OLA molecules. The morphology of the final copper oxide is determined by the relative Rhydrolysis of Cu(ac)2 and Rc against the Rcap. When Rhydrolysis is adjusted to a very slow rate by coupling the hydrolysis reaction with an esterification reaction, 1D aggregates of the Pre-CuO particles are formed because the “assembling” is fully guided by the OLA molecules. When Rhydrolysis is in the middle, the spherical architectures composed of smaller linear aggregates are formed because the guiding function of the OLA molecules can still be partially expressed. However, under high Rhydrolysis that is realized by increasing the reaction temperature or the concentration of urea, only small particles of Pre-CuO are obtained. In the latter case, the OLA molecules probably cannot cover the surface of the small copper particles before they are

Langmuir, Vol. 23, No. 11, 2007 5977

branched. After removing the OLA molecules via calcination, the spherical Pre-CuO architectures can be converted into a porous structure. Compared to the 1D and 2D aggregates, this porous architecture is highly thermally stable, and can be used as a catalyst support or a catalyst directly. Preliminary results showed that the porous structure is a good catalyst or support for the oxidation of CO at low temperatures. Acknowledgment. The collaborative research was supported by the Agency for Science, Technology and Research in Singapore (A-STAR, ICES04-414001). Z.Z. thanks Ms. Zhan Wang for measurement of the XPS spectra, and Drs. Keith Carpenter and P. K. Wong for their support of this project. Supporting Information Available: TEM images of (SI-1) CuO after the addition of urea and (SI-2) spherical Pre-CuO aggregates, and (SI-3)XPS spectra of Pre-CuO and commercial CuO. This material is available free of charge via the Internet at http://pubs.acs.org. LA063344X