Fabrications of Hollow Nanocubes of Cu2O and Cu via Reductive Self

Jul 12, 2006 - For the first time, we demonstrate that nanostructured polyhedrons of functional materials with desired interiors can be synthesized in...
0 downloads 4 Views 613KB Size
Download PDF
Langmuir 2006, 22, 7369-7377

7369

Fabrications of Hollow Nanocubes of Cu2O and Cu via Reductive Self-Assembly of CuO Nanocrystals Joong Jiat Teo, Yu Chang, and Hua Chun Zeng* Department of Chemical and Biomolecular Engineering, Faculty of Engineering, National UniVersity of Singapore, 10 Kent Ridge Crescent, Singapore 119260 ReceiVed February 15, 2006. In Final Form: June 5, 2006 In this work, a template-free synthetic approach for generating single-crystalline hollow nanostructures has been described. Using the small optical band-gap cuprous oxide Cu2O as a model case, we demonstrate that, instead of normally known spherical aggregates, primary nanocrystalline particles can first self-aggregate into porous organized solids with a well-defined polyhedral shape according to the oriented attachment mechanism, during which chemical conversion can also be introduced. In contrast to the spherical aggregates, where the nanocrystallites are randomly joined together, the Cu2O nanocrystallites in the present case are well organized, maintaining a definite geometric shape and a global crystal symmetry. Due to the presence of intercrystallite space, hollowing and chemical conversion can also be carried out in order to create central space and change the chemical phase of nanostructured polyhedrons. It has been revealed that Ostwald ripening plays a key role in the solid evacuation process. Using this synthetic strategy, we have successfully prepared single-crystal-like Cu2O nanocubes and polycrystalline Cu nanocubes with hollow interiors. For the first time, we demonstrate that nanostructured polyhedrons of functional materials with desired interiors can be synthesized in solution via a combination of oriented attachment and Ostwald ripening processes.

Introduction Among many nanomaterials with distinct geometric shapes,1-62 spheres and cubes,1-54 including various polyhedrons (such as * To whom correspondence should be addressed. Tel: (65) 6516-2896. Fax: (65) 6779-1936. E-mail: [email protected]. (1) Caruso, F.; Caruso, R. A.; Mo¨hwald, H. Science 1998, 282, 1111-1114. (2) Go¨ltner, C. G. Angew. Chem., Int. Ed. Engl. 1999, 38, 3155-3156. (3) Scha¨rtl, W. AdV. Mater. 2000, 12, 1899-1908. (4) Zhong, Z.; Yin, Y.; Gates, B.; Xia, Y. N. AdV. Mater. 2000, 12, 206-209. (5) Caruso, F. AdV. Mater. 2001, 13, 11-22. (6) Zhong, C.-J.; Maye, M. M. AdV. Mater. 2001, 13, 1507-1511. (7) Kobayashi, S.; Hamasaki, N.; Suzuki, M.; Kimura, M.; Shirai, H.; Hanabusa, K. J. Am. Chem. Soc. 2002, 124, 6550-6551. (8) Dinsmore, A. D.; Hsu, M. F.; Nikolaides, M. G.; Marquez, M.; Bausch, A. R.; Weitz, D. A. Science 2002, 298, 1006-1008. (9) Kamata, K.; Lu, Y.; Xia, Y. N. J. Am. Chem. Soc. 2003, 125, 2384-2385. (10) Nakashima, T.; Kimizuka, N. J. Am. Chem. Soc. 2003, 125, 6386-6387. (11) Guo, C.-W.; Cao, Y.; Xie, S.-H.; Dai, W.-L.; Fan, K.-N. Chem. Commun. 2003, 700-701. (12) Sun, Y.; Mayers, B.; Xia, Y. N. AdV. Mater. 2003, 15, 641-646. (13) Bao, J.; Liang, Y.; Xu, Z.; Si, L. AdV. Mater. 2003, 15, 1832-1835. (14) Sobal, N. S.; Ebels, U.; Mo¨hwald, H.; Giersig, M. J. Phys. Chem. B 2003, 107, 7351-7354. (15) Collins, A. M.; Spickermann, C.; Mann, S. J. Mater. Chem. 2003, 13, 1112-1114. (16) Peng, Q.; Dong, Y.; Li, Y. Angew. Chem., Int. Ed. 2003, 42, 3027-3030. (17) Yang, Z.; Niu, Z.; Lu, Y.; Hu, Z.; Han, C. C. Angew. Chem., Int. Ed. 2003, 42, 1943-1945. (18) Park, S.; Lim, J.-H.; Chung, S.-W.; Mirkin, C. A. Science 2004, 303, 348-351. (19) Yin, Y.; Rioux, R. M.; Erdonmez, C. K.; Hughes, S.; Somorjai, G. A.; Alivisatos, A. P. Science 2004, 304, 711-714. (20) Gu, H.; Zheng, R.; Zhang, X.; Xu, B. J. Am. Chem. Soc. 2004, 126, 5664-5665. (21) Liu, B.; Zeng, H. C. J. Am. Chem. Soc. 2004, 126, 8124-8125. (22) Liu, B.; Zeng, H. C. J. Am. Chem. Soc. 2004, 126, 16744-16746. (23) Yang, H. G.; Zeng, H. C. Angew. Chem., Int. Ed. 2004, 43, 5206-5209. (24) Noble, P. F.; Cayre, O. J.; Alargova, R. G.; Velev, O. D.; Paunov, V. N. J. Am. Chem. Soc. 2004, 126, 8092-8093. (25) Yang, H. G.; Zeng, H. C. J. Phys. Chem. B 2004, 108, 3492-3495. (26) Liu, B.; Zeng, H. C. Small 2005, 1, 566-571. (27) Chang, Y.; Teo, J. J.; Zeng, H. C. Langmuir 2005, 21, 1074-1079. (28) Li, J.; Zeng, H. C. Angew. Chem., Int. Ed. 2005, 44, 4342-4345. (29) Santiago, A. Dalton Trans. 2005, 2209-2233. (30) Sun, Y. G.; Xia, Y. N. Science 2002, 298, 2176-2179. (31) Murphy, C. J. Science 2002, 298, 2139. (32) Lee, H. K.; Schulthess, T. C.; Brown, G.; Landau, D. P.; Sorge, K. D.; Thompson, J. R. J. Appl. Phys. 2003, 93, 7047-7049. (33) Sun, X. M.; Li, Y. D. Chem. Commun. 2003, 1768-1769.

Platonic polyhedrons) derived from the cubic structure,29-54 are the two simplest forms, yet they possess the highest symmetries. One of the obvious geometric merits of this class of materials is their low resistivity under fluidic conditions, as they can be (34) Dumestre, F.; Chaudret, B.; Amiens, C.; Renaud, P.; Fejes, P. Science 2004, 303, 821-823. (35) Kim, F.; Connor, S.; Song, H.; Kuykendall, T.; Yang, P. D. Angew. Chem., Int. Ed. 2004, 43, 3673-3677. (36) Jin, R. C.; Egusa, S.; Scherer, N. F. J. Am. Chem. Soc. 2004, 126, 99009901. (37) Urban, J. J.; Ouyang, L.; Jo, M. H.; Wang, D. S.; Park, H. Nano Lett. 2004, 4, 1547-1550. (38) Yu, D. B.; Yam, V. W. W. J. Am. Chem. Soc. 2004, 126, 13200-13201. (39) Xu, R.; Zeng, H. C. Langmuir 2004, 20, 9780-9790. (40) Wiley, B.; Sun, Y. G.; Mayers, B.; Xia, Y. N. Chem. Euro. J. 2005, 11, 454-463. (41) Wang, W. Z.; Huang, J. Y.; Ren, Z. F. Langmuir 2005, 21, 751-754. (42) Yu, D. B.; Yam, V. W. W. J. Phys. Chem. B 2005, 109, 5497-5503. (43) Im, S. H.; Lee, Y. T.; Wiley, B.; Xia, Y. N. Angew. Chem., Int. Ed. 2005, 44, 2154-2157. (44) Stankic, S.; Muller, M.; Diwald, O.; Sterrer, M.; Knozinger, E.; Bernardi, J. Angew. Chem., Int. Ed. 2005, 44, 4917-4920. (45) Sterrer, M.; Berger, T.; Diwald, O.; Knozinger, E.; Sushko, P. V.; Shluger, A. L. J. Chem. Phys. 2005, 123, Art. No. 064714. (46) Feng, J.; Zeng, H. C. J. Phys. Chem. B 2005, 109, 17113-17119. (47) Desvaux, C.; Amiens, C.; Fejes, P.; Renaud, P.; Respaud, M.; Lecante, P.; Snoeck, E.; Chaudret, B. Nat. Mater. 2005, 4, 750-753. (48) Lu, W. G.; Fang, J. Y.; Ding, Y.; Wang, Z. L. J. Phys. Chem. B 2005, 109, 19219-19222. (49) Sherry, L. J.; Chang, S. H.; Schatz, G. C.; Van Duyne, R. P.; Wiley, B. J.; Xia, Y. N. Nano Lett. 2005, 5, 2034-2038. (50) Gou, L. F.; Murphy, C. J. Nano Lett. 2003, 3, 231-234. (51) Gou, L. F.; Murphy, C. J. J. Mater. Chem. 2004, 14, 735-738. (52) Li, X. D.; Gao, H. S.; Murphy, C. J.; Gou, L. F. Nano Lett. 2004, 4, 1903-1907. (53) Wang, Z. H.; Chen, X. Y.; Liu, J. W.; Mo, M. S.; Yang, L.; Qian, Y. T. Solid State Commun. 2004, 130, 585-589. (54) Chang, Y.; Zeng, H. C. Cryst. Growth Des. 2004, 4, 273-278. (55) Gao, P. X.; Wang, Z. L. J. Am. Chem. Soc. 2003, 125, 11299-11305. (56) Kong, X. Y.; Wang, Z. L. Nano Lett. 2003, 3, 1625-1631. (57) Kong, X. Y.; Ding, Y.; Yang, R.; Wang, Z. L. Science 2004, 303, 13481351. (58) Li, F.; Ding, Y.; Gao, P.; Xin, X.; Wang, Z. L. Angew. Chem., Int. Ed. 2004, 43, 5238-5242. (59) Yang, H. G.; Zeng, H. C. Angew. Chem., Int. Ed. 2004, 43, 5930-5933. (60) Yan, F.; Goedel, W. A. Angew. Chem., Int. Ed. 2005, 44, 2084-2088. (61) Cho, K.-S.; Talapin, D. V.; Gaschler, W.; Murray, C. B. J. Am. Chem. Soc. 2005, 127, 7140-7147. (62) B. Liu; Zeng, H. C. J. Am. Chem. Soc. 2005, 127, 18262-18268.

10.1021/la060439q CCC: $33.50 © 2006 American Chemical Society Published on Web 07/12/2006

7370 Langmuir, Vol. 22, No. 17, 2006

essentially considered as zero-dimensional (0D) entities when their size is trimmed down to the nanoscale regime.1-54 Another distinctive feature of these nanomaterials is their structural isotropy, which permits us to perform various architectural designs and superlattice assemblies using them as primary building blocks.1-54 When they are prepared with hollow interiors, furthermore, the nanomaterials represent a class of even more attractive functional materials though the complexity of the synthesis increases.1-54 For example, the hollow nanomaterials of this type have received increasing attention in recent years, owing to their potential applications across different technological fields, such as photonic crystals, host materials for intercalants, drug-delivery carriers, sensors, chemical reactors, etc.1-54 Other unique features for these isotropic hollow materials include their light weight, because of central hollow voids, and their large specific surface areas, owing to the presence of both interior and exterior surfaces.1-54 Concerning the creation of interior space for nanostructured materials, there have been two synthetic strategies available.1-3,8,18,19,25,30,59 The first one, template-assisted synthesis, provides an effective approach where hard templates such as polymeric and metallic cores or soft templates such as surfactant micelles or ionic solvents have been utilized.1-3,8,30 The second one, template-free synthesis, can also create interior space for nanomaterials through various physicochemical processes.18,19,25,59 For example, redox reactions and physical phenomena such as Ostwald ripening,25-28 Kirkendall effect,19,22 oriented attachment,21,59 and hydrophobic interactions18 have been employed respectively in creating interior spaces in liquid phase. It should be mentioned that most nanoproducts prepared in the latter attempts are polycrystalline and are limited only to the spherical morphology,18,19,21,25 except for some recent investigations on single-crystal metal oxide nanocubes (e.g., cuprous oxide Cu2O nanocubes)50-54 and polyhedrons with hollow interiors.59 Like many reported metal-oxide nanocubes,39,46 the cubic crystal morphology of Cu2O is inherited from Cu2O crystal symmetry by stabilizing {100}, {010}, and {001} planes.50-54 Nonetheless, the hollowing mechanism of Cu2O nanocubes as well as of other metal-oxide nanocubes in general has so far remained unknown. To exploit the general template-free approach in solution media, we had recently developed two methods for fabrications of nanostructures with interior voids.25,26,59 The first method, illustrated in Figure 1A, relies on a three-dimensional (3D) aggregation of primary nanocrystallites, followed by a solid evacuation through Ostwald ripening.25,26 The second method (Figure 1B), in which tiny nanocrystallites are gathered twodimensionally (2D), provides a direct means for construction of hollow structures through a plane-by-plane mechanism.59 The crystallites in this method must fulfill certain crystallographic requirements in order to undergo an assembling process called oriented attachment.59,63,64 It is noted that the hollow nanospheres produced from the first method are normally polycrystalline because of random arrangement of nanocrystallites during the initial aggregation;26 note that a spherical aggregate has the lowest system energy to hold the crystallites together. The nanocrystallites in the products formed from the second method, on the other hand, maintain a global crystallographic relationship, although localized disorientation can be observed.59 In this work, as depicted in Figure 1C, we will present a third type of methodic concept in this research area. As a first reported case, the present synthetic method is virtually a combination of (63) Penn, R. L.; Banfield, J. F. Science 1998, 281, 969. (64) Pacholski, C.; Kornowski, A.; Weller, H. Angew. Chem., Int. Ed. 2002, 41, 1188.

Teo et al.

Figure 1. Three different types of synthetic methods for generation of hollow nanostructures: (A) random aggregation of nanocrystallites and core hollowing via Ostwald ripening, resulting in polycrystalline nanospheres; (B) two-dimensional oriented attachment for formation of thin crystal planes and construction of hollow octahedra in a plane-by-plane manner; and (C) three-dimensional oriented attachment for solid nanocubes and creation of hollow interiors by Ostwald ripening. Hashed areas indicate the solid parts of nanostructures, and dark areas represent interior spaces.

both oriented attachment63,64 and Ostwald ripening25,26 processes. In the first stage, nanocrystallites undergo oriented attachment to form a well-defined geometrical structure (e.g., the cube). In the second stage, solid evacuation takes place in the central part of the shape-defined aggregate via Ostwald ripening, resulting in single-crystalline hollow nanocubes. The model oxide Cu2O (a p-type semiconductor) selected in the present work has been under extensive investigations in recent years.50-54 As mentioned earlier, in particular, single-crystalline hollow nanocubes of Cu2O have been reported in several syntheses.50-54 With this new fabrication method (Figure 1C), we are able to explain for the first-time mechanistic processes in the formation of single-crystal hollow nanocubes of Cu2O. The findings of the present work also shed light on the general formation mechanism for solution synthesis of transition-metal-oxide nanocubes with interior spaces. Experimental Section Materials Preparation. In a typical experiment, 30.0 mL of 0.005-0.010 M Cu(NO3)2‚3H2O solution [a certain quantity of Cu(NO3)2‚3H2O was dissolved in the organic solvent N,N-dimethylformamide (DMF) to make a stated concentration] and 0.10-0.60 mL of deionized water were mixed and sealed in a Teflon-lined stainless steel autoclave with 50 mL capacity and then heated to different temperatures (150, 160, 170, 180, 190, 200, and 210 °C) for different reaction times to form Cu2O nanoproducts. In some of the experiments, 0.004-0.020 g of NaNO3 salt was also added in the solution to adjust the morphology of Cu2O nanoproducts. Detailed information on the selection of experimental parameters can be found in the Supporting Information (SI-1). In addition to using DMFwater, Cu2O nanocubes were also synthesized with Cu2+ dissolved in DMF-ethanol cosolvents. Briefly, 30.0 mL of 0.005 or 0.010 M Cu(NO3)2‚3H2O in a DMF-ethanol mixed solvent (the volume ratio DMF to ethanol was changed from 29:1 to 4:26, note that there was a small amount of water inherited from the starting Cu(NO3)2‚3H2O salt; SI-2) were sealed in the above autoclave and then heated at 160

Fabrications of Hollow Nanocubes

Langmuir, Vol. 22, No. 17, 2006 7371

and 180 °C in order to produce Cu2O. In some experiments, NaNO3 (mostly 0.01 g) was also added as a mineralizer. In a two-step synthesis of Cu nanocubes, the autoclave was first heated at 180 °C for 6 h, and consecutively heated at 200 °C for 1-2 h, as detailed in SI-2. Materials Characterization. The crystallographic information of the prepared samples was established by powder X-ray diffraction (XRD, Shimadzu, model XRD-6000, Cu KR radiation λ ) 1.5406 Å). Morphological investigation was carried out with transmission electron microscopy and selected area electron diffraction (TEM/ SAED, Joel model JEM-2010, 200 kV; and HRTEM, Philips FEI model Tecnai-G2, 200 kV), scanning electron microscopy (SEM, Joel model JSM-5600LV), field emission scanning electron microscopy, and energy-dispersive X-ray spectroscopy (FESEM/EDX, Joel model JSM-6700LV). Surface analysis for the samples was also performed using X-ray photoelectron spectroscopy [XPS (also used for Auger electron spectroscopy (AES) measurement), Kratos Analytical model AXIS-Hsi] with a monochromated Al KR X-ray source (1486.6 eV).25,28

Results and Discussion Composition and Morphology of Products. A series of experiments was first carried out to narrow down some synthetic parameters for hollow-cube formation (SI-1). The crystal structure of all prepared Cu2O products was investigated with XRD method. All recorded XRD peaks can be assigned to the cubic symmetry of Cu2O (space group: Pn3hm; ao ) 4.26 Å; JCPDS no. 030892),50-54 confirming that all the products are phase-pure Cu 2O. The composition of the prepared Cu2O products was also checked with EDX technique, which shows that the atomic ratio of Cu to O is indeed close to 2:1 (e.g., 65.7:30.0, see SI-3) in our Cu2O samples. In the present synthesis, hollow nanocubes of Cu2O were formed with the addition of a suitable amount of water in DMF solution; water is a key to the control of cubic morphology. Figure 2 shows the morphology of prepared Cu2O hollow cubes at 200 °C for 6.5 h, with 0.50 mL of deionized water in 30 mL of 0.005 M of Cu2+ DMF solution (SI-1). It is observed that the Cu2O products formed are of cubic structures (Figure 1A,B), and there is a void space in the center of each cube (Figure 2C,D). The measured size distribution of the hollow cubes in Figure 2C,D is about 200 ( 30 nm, indicating that the hollow cubes are monodispersed in size. Quite interestingly, surfaces of these hollow cubes are not particularly smooth, suggesting that they might be made from an aggregation of smaller particles which will be addressed shortly. Apart from the TEM investigation, the hollow interior is also confirmed with the FESEM method. For instance, one small pinhole or two through central space can be easily observed on the edges or corners of these nanocubes (SI-4), although the inner void of hollow cubes is sealed rather completely. These pinholes may serve as exchange channels for chemical constituents inside and outside the cubes when employing them as nanoreactors or nanocontainers. Depending on the synthetic parameters, furthermore, hollow nanocubes with edge lengths smaller than 100 nm can also be prepared with the present method (Figure 2E,F). The crystal orientation and crystallinity of Cu2O hollow cubes were further studied with HRTEM/SAED methods. Figure 3 shows the electron diffraction pattern of a Cu2O hollow cube (Figure 3A,B). In this measurement, the electron beam was injected along the [001] direction and the spot array has a 4-fold symmetry that can be indexed with hk0 (i.e., [001] zone spots, in accordance to the extinction rule of electron diffraction of space group Pn3hm). As indicated by the clear diffraction spots, all of the hollow cubes are nearly single-crystalline, and bounded with six crystal planes of {100}, {010}, and {001}. A HRTEM image of the edge area of such a hollow cube is shown in Figure

Figure 2. FESEM images (A and B) and TEM images (C and D) of as-prepared Cu2O hollow nanocubes. Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.50 mL of H2O at 180 °C for 15 h. TEM images (E and F) of smaller Cu2O hollow nanocubes. Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.40 mL of H2O at 180 °C for 15 h.

Figure 3. (A and B) TEM image of a Cu2O hollow cube and its SAED pattern. Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.50 mL of H2O at 180 °C for 15 h. (C) HRTEM image of a Cu2O hollow cube. Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.50 mL of H2O at 200 °C for 6.5 h.

3. The crystallinity within a crystallite block is high, although there are intercrystallite mismatches (e.g., splitting of spots, Figure

7372 Langmuir, Vol. 22, No. 17, 2006

Teo et al.

Figure 4. Powder XRD patterns of nanoproducts synthesized with different reaction times. Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.50 mL of H2O at 200 °C for 1.5 to 7 h. Single-asterisk (*) denotes the CuO phase, whereas the doubleasterisk (**) represents the metallic Cu component; the remaining diffraction peaks are from Cu2O phase.

3B). The lattice fringe of d200 was measured at about 2.13 Å, which is in good agreement with 2.12 and 2.14 Å reported in the literature (JCPDS files of 03-0892 and 34-1354).50-54 Growth Process of Cu2O Hollow Cubes. The growth process of Cu2O hollow cubes was investigated in detail at different reaction time and temperature (see SI-1). Figure 4 displays a series of evolutional XRD patterns of the samples investigated. The solid products formed after 1.5-2.0 h of reactions are phase-pure CuO. With a longer reaction time of 2.5-3.0 h, the product becomes a mixture of CuO and Cu2O. This mixture is then converted into phase-pure Cu2O over a reaction time span of 3.5-6.5 h. However, when the time exceeds 7 h, metallic copper (Cu0)65 appears in the product as a secondary phase, indicating the reduction of CuO has been overdone at the time spent. Corresponding to the above phase evolution, the timedependent crystal morphology of the samples was reported in Figure 5. At a short reaction time of 1.5 h, TEM images show that CuO nanocrystallites self-aggregate to oval-shaped structures with the lengths of about 80-150 nm (Figure 5A,B); the CuO phase of these crystallites is also confirmed in the polycrystalline ED rings (inset, Figure 5A,B). TEM images of Figure 5C-F illustrate the morphology of product formed at an intermediate reaction time corresponding to the formation of the CuO and Cu2O mixture (2.5 h; also refer to Figure 4). It is interesting to note that partially converted Cu2O crystallites formed at this reaction time are aligned into square-like frames (Figure 5C), which lays down the base for a cubelike aggregate (Figure 5D). It is further noted that there are plenty of intercrystallite spaces present in these premature cubic structures. Two Cu2O nanocubes of this kind are displayed in Figure 5E,F, together with their small starting CuO crystallites. During the growth, the CuO crystallites are reduced (to Cu2O) and attached simultaneously, noting that the resulting Cu2O cubes at this stage are not compact, as indicated in the TEM image contrasts. After 3.5 h, initial CuO nanocrystallites have been significantly reduced, whereas the cubical shape of Cu2O has been largely established (Figure 5G,H). (65) Chang, Y.; Lye, M. L.; Zeng, H. C. Langmuir 2005, 21, 3746-3748.

Figure 5. TEM images of nanoproducts prepared at 200 °C: 1.5 h (CuO, A and B) and 2.5 h (CuO + Cu2O, C-F), 3.5 h (Cu2O + CuO, G and H), and 5.5 h (Cu2O, I and J). Other experimental parameters: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.50 mL of H2O. The XRD patterns of the above samples can be seen in Figure 4.

Over the next 2 h of reaction (at 5.5 h), hollowing and recrystallization take place with Ostwald ripening, through which a central space is created (Figure 5I,J). Since the solid cubic precursor has been determined to be Cu2O and the hollowing process does not produce any new chemical phases, the hollowing should be assigned to the Ostwald ripening mechanism.25,26 On the other hand, the Kirkendall effect refers to competitive atomic diffusions among two or more phases upon thermal treatment.19,22 Because we only have a single phase (i.e., Cu2O) to consider in the present case, the Kirkendall mechanism can be ruled out unambiguously. From the above complementary XRD/TEM analyses, it can be understood that the formation process of Cu2O hollow cubes composes the following three consecutive steps: (i) production of primary CuO nanocrystallites, (ii) reductive formation of loosely packed cubic aggregates of Cu2O from the preformed CuO crystallites, and (iii) evacuation of central crystallites and

Fabrications of Hollow Nanocubes

Langmuir, Vol. 22, No. 17, 2006 7373 Table 1. Binding Energies (eV) of Cu 2p3/2 of CuO and Cu2O and Their Relative Contents (in Parentheses) sample

CuO

Cu2O

1.5 h 2.5 h 3.5 h 6.5 h

933.8 (1.00) 933.8 (0.80) 933.5 (0.18) 933.5 (0.21)

932.7 (0.20) 932.5 (0.82) 932.5 (0.79)

Table 2. Binding Energies (eV) of O 1s of Different Chemical Species and Their Relative Contents (in Parentheses) sample 1.5 h 2.5 h 3.5 h 6.5 h

Figure 6. XPS spectra of Cu 2p3/2 and O 1s (top panels) and comparison of Cu 2p spectra of and Cu L3VV spectra (bottom panel) for nanoproducts synthesized at 200 °C for different reaction times (1.5 to 6.5 h). All signal intensities are in arbitrary units. Other experimental parameters: 30.0 mL of [Cu2+] (0.005 M in DMF) + 0.50 mL of H2O. The XRD patterns of these samples can be seen in Figure 4.

perfection of Cu2O hollow nanocubes through Ostwald ripening mechanism. Surface Compositional Analysis. On the basis of XRD/TEM/ SAED findings, we have known that the CuO crystallites are reduced and aggregated simultaneously to Cu2O nanocubes in the above formation process (2.5-3.5 h; Figures 4 and 5; step ii). To have a better understating on this growth mechanism, surface compositional analysis was further carried out. XPS spectra of Cu 2p3/2 measured for the samples over this transforming process are displayed in Figure 6, and the binding energies (BEs) of different chemical species are listed in Table 1. In general, Cu 2p3/2 spectra can be fitted into 2 peaks except for the sample at 1.5 h. The BEs at 933.8, 933.5, and 933.5 eV measured in our samples agree well with the reported values of the CuO surface phase in a range of 933.4-933.9 eV,66-73 and (66) Wagner, C.; Riggs, W.; Davis, L.; Moulder, J. Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corporation: Eden Prairie, MN, 1979.

CuO

Cu2O

529.2 (0.47) 529.3 (0.50) 530.7 (0.20) 529.1 (0.05) 530.0 (0.42) 530.1 (0.44)

OH- and CO32-

H2O

531.2 (0.41) 531.6 (0.23) 531.5 (0.36) 531.5 (0.38)

533.0 (0.12) 532.9 (0.07) 533.0 (0.18) 533.0 (0.18)

the BEs at 932.7, 932.5, and 932.5 eV are also close to the literature data of Cu2O at 932.4 or 932.5 eV.66,67,74,75 Therefore, the evolution of peak areas in these samples reveals a gradual transformation from CuO to Cu2O (Table 1). Through this surface analysis, more quantitative information can now be acquired. The starting Cu2+ in DMF solution initially forms pure CuO crystallites at 1.5 h. When the reaction time increases to 2.5 h, the product is a mixture of CuO and Cu2O, and the surface molar ratio of CuO to Cu2O is ca. 4:1 (Table 1). When the reaction time is longer than 3.5 h, the main phase of products becomes Cu2O, and the molar ratio of CuO to Cu2O is changed to approximately 1:4 (Table 1). According to the XRD patterns of these samples (Figure 4), however, the formed bulk products are all pure Cu2O if the reaction time is longer than 3.5 h. Therefore, it is deduced that a small fraction of surface Cu2O was oxidized to CuO during the sample drying and handling under normal ambient conditions. Figure 6 also displays the corresponding O 1s spectra. As tabulated in Table 2, the lowest BEs of O 1s (the lattice O2-) are in the range of 529.1-529.3 eV for the samples at 1.5, 2.5, and 3.5 h, which is consistent with the literature data of CuO (529.5 eV).70,71 The second peaks at 530.0-530.7 eV indicate the formation of Cu2O phase.66,67,74,75 It should be mentioned that the lattice oxygen O2- for the CuO phase is not detectable in the sample prepared at 6.5 h due to its small quantity. Auger electron spectra of Cu L3VV are also useful in distinguishing oxidation states between Cu+ and Cu2+. A comparison between the Cu 2p and Cu L3VV spectra is thus made in Figure 6. For the sample prepared at 1.5 h, the Cu 2p3/2 BE of 933.8 eV and Cu L3VV KE of 918.2 eV agree well with the reported values of 933.7 and 918.1 eV of the CuO phase.66,67 For the sample prepared at 2.5 h, the Cu 2p3/2 spectrum has two peaks at 933.8 and 932.7 eV, respectively, which correspond to surface phases of CuO and Cu2O (in a smaller quantity, Table 1). In this agreement, the position of the KE in the Cu L3VV spectrum for this sample is around 918.1 eV for CuO,66 noting (67) Espinu¨s, J. P.; Morales, J.; Barranco, A.; Caballero, A.; Holgado J. P.; Gonza´lez-Elipe, A. R. J. Phys. Chem. B 2002, 106, 6921-6929. (68) Zhu, J.; Chen, H.; Liu, H.; Yang, X.; Lu, L.; Wang, X. Mater. Sci. Eng. A 2004, 384, 172-176. (69) Wang, W.; Liu, Z.; Liu, Y.; Xu, C.; Zheng, C.; Wang, G. Appl. Phys. A 2003, 76, 417-420. (70) Wang, W.; Zhan, Y.; Wang, X.; Liu, Y.; Zheng, C.; Wang, G. Mater. Res. Bull. 2002, 37, 1093-1100. (71) Xu, J. F.; Ji, W.; Shen, Z. X.; Tang, S. H.; Ye, X. R.; Jia, D. Z.; Xin, X. Q. J. Solid State Chem. 1999, 147, 516-519. (72) Wang, H.; Xu, J. Z.; Zhu, J. J.; Chen, H. Y. J. Cryst. Growth 2002, 244, 88-94. (73) Brookshier, M. A.; Chusuei, C. C.; Goodman, D. W. Langmuir 1999, 15, 2043-2046. (74) Fernando, C. A. N.; de Silva, P. H. C.; Wethasinha, S. K.; Dharmadasa, I. M.; Delsol, T.; Simmonds, M. C. Renewable Energy 2002, 26, 521-529. (75) Wang, W.; Wang, G.; Wang, X.; Zhan, Y.; Liu, Y.; Zheng, C. AdV. Mater. 2002, 14, 67-69.

7374 Langmuir, Vol. 22, No. 17, 2006

Teo et al.

Figure 7. Surface models for the Cu2O (100) crystal planes: (A) Cu-cation terminated plane (Cu in orange), and (B) O2--anion terminated plane (O2- in blue and dark blue). Reductive formation of cubic structures: (C) under a low water content condition, the formed Cu2O crystallites are smaller and the surfaces of Cu2O (100) are rougher, leading to more mismatches among the crystallites and a lower packing density for the central void formation (via Ostwald ripening), and (D) with a higher content of water, the formed Cu2O crystallites are larger and surface of Cu2O (100) are smoother, resulting in a better oriented attachment among the crystallites and a denser packing for the final crystal cubes. CuO crystallites attached to the forming Cu2O cubes (in gray) are represented with darker rectangular blocks.

that there is a significant peak broadening due to the transient nature of this sample. When the reaction time is prolonged to 3.5 and 6.5 h, the BEs are 932.5 and 932.5 eV and the KEs are 917.1 and 917.1 eV, respectively. These values once again indicate that the two samples are primarily in Cu2O phase. Effects of Water on Cu2O Morphology. The redox chemical process for the formation of the Cu2O phase in the DMF-water solvent system had been proposed in our previous work,27 where DMF solvent acts as a weak reducing agent in synthesis and a small amount of water is incorporated into the synthesis through the starting reagent Cu(NO3)2‚3H2O:

HCON(CH3)2 + H2O f HCOOH + NH(CH3)2

(1)

Cu2+ + H2O + 2 NH(CH3)2 f CuO(s) + 2 NH2(CH3)2+ (2) 2CuO(s) + HCOOH f Cu2O(s) + H2O + CO2

(3)

Instead of polycrystalline Cu2O hollow nanospheres,27 surprisingly, single-crystal Cu2O hollow nanocubes could be obtained with an additional amount of water (which was deliberately added in the present work, SI-1). The significant alternation in crystal morphology can be attributed to a gain in stabilization of the {100}, {010}, and {001} crystal planes with more abundant water in the present work. Figure 7A,B shows two possible surface terminations of the Cu2O (100) plane. Whether a (100) surface is terminated with Cu+ cations or with O2- anions, there must be local unbalanced charged domains (e.g., atomic steps) on its surface. Therefore, surface adsorptions with ions in opposite charges and interaction with polar molecules will compensate the surface charges and thus stabilize the crystal planes. Consistent with this postulation, the stabilizing effect is indeed obtained when additional water is added to the reactions, owing to more charged species generated in the products (e.g., NH2(CH3)2+ in eq 2) as well as more polar solvent molecules (e.g., water) can interact with the Cu2O (100) surfaces via direct charge compensation or through formation of different hydrogen bonding.

Figure 8. Cu2O morphological changes with water: (A) hollow cubes: with 0.30 mL of H2O at 170 °C for 26 h, (B) large cubes: with 0.50 mL of H2O at 170 °C for 26 h, (C) hollow cubes: with 0.50 mL of H2O at 190 °C for 11 h, (D) large cubes: with 0.60 mL of H2O at 190 °C for 11 h, (E) hollow cubes: with 0.50 mL of H2O at 200 °C for 6.5 h, (F) large cubes: with 0.60 mL of H2O at 200 °C for 6.5 h, (G) hollow cubes: with 0.40 mL of H2O at 210 °C for 5.5 h, and (H) large cubes: with 0.50 mL of H2O at 210 °C for 5.5 h. A same amount of 30.0 mL of [Cu2+] (0.005 M in DMF) was used in each of the above experiments.

To demonstrate the effect of water, Figure 8 shows some TEM/SEM images of Cu2O hollow cubes and large cubes prepared at different reaction temperatures and times. Figure 8A,B shows the samples prepared at 170 °C for 26 h with 0.30 and 0.50 mL of H2O (see SI-1). In these experiments, the volume of water was only varied from 0.30 to 0.50 mL, but the Cu2O crystal morphology was changed from small hollow nanocubes to large solid microcubes, accompanied by a drastic size change from 0.15 to 3-6 µm. Furthermore, in Figure 8C,D are the SEM images of samples prepared at a higher temperature 190 °C for 11 h with 0.50 and 0.60 mL of H2O (SI-1). With a difference of 0.10 mL in water (0.50 versus 0.60 mL), the average size of cubes increased from about 0.2 to 0.8 µm. When the reaction temperature was raised to 200 °C, very large cubes also began to form with 0.60 mL of water in the solution (see Figure 8F). When the temperature was increased to 210 °C, an amount of 0.50 mL of deionized water was sufficient to generate the large cubes, owing to the fast reaction rate at the even higher temperature (Figure 8G,H). On the basis of these results, we observe that, with the additional deionized water in the starting solution, the crystal morphology of Cu2O changes from hollow nanocubes to large solid microcubes. The optimized water volume for the formation of

Fabrications of Hollow Nanocubes

Langmuir, Vol. 22, No. 17, 2006 7375

Figure 9. Crystal morphology of Cu2O nanocubes (synthesized at 180 °C for 6 h): (A) detailed view on a crystal cube (note that there is a lighter center), (B) the SAED pattern of (A), and (C) general crystal morphology in a large scale. (D) Powder XRD patterns of nanoproducts (synthesized at 180 °C for 2-6 h), and nanoproducts prepared with a two-step method at 180 °C for 6 h and consecutively at 200 °C for 1 to 2 h. Other experimental parameters used for these samples: 30.0 mL of [Cu2+] solution (0.005 M in DMF:EtOH ) 10:20 mL/mL). Unmarked diffraction peaks are from Cu2O phase, and single-asterisk (*) denotes CuO phase while the double-asterisk (**) represents metallic Cu phase.

Cu2O hollow cubes is around 0.40-0.50 mL under our general experimental conditions (180-200 °C, SI-1). With a lower concentration of copper ions at a higher temperature, the amount of water used becomes even more critical in controlling crystal morphology (SI-5). The optimal amount of water is further addressed in Figure 7. With a limited amount of water, for example, the resultant Cu2O crystallites are generally small and the surface of Cu2O (100) may not be entirely smooth, which may lead to the presence of more intercrystallite voids for the development of a hollow interior (Figure 7C). With a larger amount of water present, on the other hand, bigger and smoother surfaces of Cu2O (100) can be attained, which favors a faster reaction and formation of more compact crystallite aggregates, resulting in large solid Cu2O microcubes (Figure 7D). Effects of Ethanol on Cu2O Morphology. To investigate the effort of solvent on the stability of the Cu2O (100) surface, weaker polar molecules such as ethanol were also examined in the present study (SI-2). Figure 9A-C shows the TEM images and SAED pattern of the Cu2O nanocubes prepared with ethanol, and their crystallographic structure is also confirmed with the XRD method (Figure 9D, SI-6). The cubes formed under this condition are uniform, and the edge length of the cubes is smaller than 100 nm. Similar to Figure 2B, the spot array shows a 4-fold axis that can be indexed with hk0, also indicating a cubic crystal-symmetry for Cu2O nanocubes. Without adding ethanol to the synthesis, in contrast, only polycrystalline Cu2O spheres could be formed.27 The stabilizing effect of ethanol on the Cu2O (100) surface has thus been demonstrated. XRD investigation of Figure 9D indicates that the product formed at 2 h is a mixture of CuO and Cu2O. When the reaction time is prolonged to 3-6 h, pure Cu2O products are obtained. Based on the results of Figures 9D and 10 (SI-2 and SI-6), it can be concluded that the synthetic process using DMF-ethanol is very similar to that using DMF-water, i.e., Cu2+ initially forms CuO nanocrystallites in the basic solutions and then the resultant nanocrystallites are reduced and aggregate to Cu2O cubes. Our present process is different from a reported process, in which the Cu2+ salts in alkaline solution are reduced with hydrazine to intermediates of CuOH and then transfer to Cu2O by the thermal decomposition.50 However, when the CuO suspensions in an

Figure 10. TEM/SAED characterization for the mixture of CuO and Cu2O: (A, C, E, and F) TEM images of mixture; (B) SAED pattern of (A); (D) SAED pattern of (C); (G) SAED pattern of (F). Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF: EtOH ) 10:20 mL/mL) was heated at 180 °C for 2 h.

aqueous solution were used as the precursor to synthesize Cu2O particles at 30 °C, the morphology of Cu2O is uniform polycrystalline spheres with an average diameter of about 0.27 µm.76 It is apparent that high temperature and polarity of solvent used in our investigation are very important synthetic parameters for the formation of Cu2O single-crystal-like nanocubes. From (76) Muramatsu, A.; Sugimoto, T. J. Colloid Interface Sci. 1997, 189, 167173.

7376 Langmuir, Vol. 22, No. 17, 2006

Teo et al.

Figure 11. TEM/SAED characterization for metallic Cu hollow nanocubes prepared by the two-step method: (A and B) TEM images; and (C) SAED pattern measured for (B). (D) Detailed XPS spectra of O 1s and Cu 2p3/2 for the Cu hollow cubes, and (E) Overall XPS spectrum of Cu 2p. Signal intensities in (D) and (E) are in arbitrary units. Experimental conditions: 30.0 mL of [Cu2+] (0.005 M in DMF:EtOH ) 10:20 mL/mL) was heated at 180 °C for 6 h and consecutively at 200 °C for 2 h. XRD pattern of this sample can be found in Figure 9D. Table 3. Binding Energies (eV) of O 1s and Cu 2p3/2 of Different Chemical Species and Their Relative Contents (in Parentheses) element

CuO

Cu

OH- and CO32-

O 1s 530.4 (0.35) 531.6 (0.39) Cu 2p3/2 934.3 (0.47) 932.8 (0.53) -

H2O 532.9 (0.26) -

our synthesis results (SI-2 and SI-6), we further know that the ratio of DMF to ethanol plays an important role on the morphology of Cu2O products (Figures 9 and 10). Formation of Cu Nanocubes. In this work, the formed Cu2O nanocubes were also used as solid precursors to fabricate metallic Cu hollow nanostructures. Owing to the weak reducing power of DMF, it takes a long time to reduce Cu2+ to Cu2O and to Cu at low temperatures (e.g., 20 h is needed at 180 °C; SI-2). To overcome this process difficulty, a two-step method was devised (SI-2): the experiments were conducted at 180 °C for 6 h and consecutively at 200 °C for 1-2 h, while the volume ratio of DMF to ethanol was still kept at 10:20. The Cu products were also confirmed by XRD method; Figure 9D shows an XRD pattern (6 + 2 h) of the samples prepared with this method. In Figure 11A,B, the hollow interior of Cu nanoproducts is confirmed, noting that the formed Cu preserves the shape of the pristine Cu2O nanocubes, and the sharp SAED ring pattern in Figure 11C reveals high crystallinity of this metal product. Figure 11D shows the XPS spectra of Cu 2p3/2 and O 1s for the air-dried Cu hollow cubes prepared with the above two-step method. The BE positions of each component are detailed in Table 3. The Cu 2p3/2 spectrum can also be fitted into two peaks at 934.3 and 932.8 eV, respectively. The peak at 934.3 eV belongs to the CuO, although it is higher than the reported energy of CuO (933.7 eV).66 However, it has been reported that when the component of surface CuO is small, the BE shifts from 933.6 eV to a higher value of 934.4 eV.67 This may suggest that a thin layer of CuO forms on the outermost surface of the sample. In Figure 11E, a weak split at 944.4 eV is the Cu 2p3/2 shake-up

peak typical for CuO,73 which indeed indicates the above oxidation. However, the Cu 2p3/2 BEs of Cu2O and Cu0 are close to each other; the reported values are 932.5 and 932.6 eV, respectively.66 Thus the peak at 932.8 eV cannot be confirmed only by the Cu 2p3/2 alone. The O 1s spectrum in Figure 11D can be fitted into three peaks. The high BE component at 530.4 eV is due to the presence of CuO on the surface of Cu hollow particles.70,71 The next component at 531.6 eV can be attributed to the presence of a hydroxyl group and/or carbonate group.39 The final peak at 532.9 eV is assigned to molecular water adsorbed on the surface.39 From the O 1s spectrum, we can deduce that there exists no Cu2O on the outermost surface of the sample. Combining the Cu 2p3/2 with O 1s results, it can be concluded that the BE of Cu 2p3/2 at 932.8 eV belongs to metallic copper (Cu0) and only a thin layer of CuO exists on the outermost surface of the sample, though Cu2O might exist as an underneath layer below the surface CuO, which may not be detectable with XPS.

Conclusions In summary, we have developed a simple hydrothermal method to prepare single-crystalline metal oxide and polycrystalline metal nanocubes with interior spaces. The chemical processes can be divided into the following steps: the Cu2+ ions are consecutively converted to oxide phases of CuO and Cu2O and then to metallic Cu under our reducing reaction environment. Instead of forming polycrystalline Cu2O nanospheres, more specifically, primary CuO nanocrystalline particles can self-aggregate first into porous Cu2O nanocubes with a well-defined cubic structure according to oriented attachment mechanism, in which the Cu2O nanocrystallites are orderly attached each other, maintaining a global cubic crystal-symmetry. Due to the presence of intercrystallite space in the Cu2O aggregates, solid hollowing can be further conducted, which leads to the formation of single-crystal-like nanocubes with central voids. It has been revealed that Ostwald ripening plays a key role in this solid evacuation process. Using this synthetic strategy, polycrystalline Cu nanocubes with hollow

Fabrications of Hollow Nanocubes

interiors can also be obtained. On the basis of the current findings, it has been demonstrated that nanostructured polyhedrons of functional materials with desired interiors can be synthesized in solution via wet chemical approaches using combined oriented attachment and Ostwald ripening processes.

Langmuir, Vol. 22, No. 17, 2006 7377

Supporting Information Available: Detailed information on the selection of experimental parameters. This material is available free of charge via the Internet at http://pubs.acs.org. LA060439Q