Environ. Sci. Technol. 2008, 42, 4902–4907
Hydrothermal Synthesis and Photocatalytic Activity of Zinc Oxide Hollow Spheres JIAGUO YU* AND XIAOXIAO YU State Key Laboratory of Advanced Technology for Material Synthesis and Processing, Wuhan University of Technology, Luoshi Road 122, Wuhan 430070, P.R. China
Received January 14, 2008. Revised manuscript received April 25, 2008. Accepted April 28, 2008.
ZnO hollow spheres with porous crystalline shells were onepot fabricated by hydrothermal treatment of glucose/ZnCl2 mixtures at 180 °C for 24 h, and then calcined at different temperatures for 4 h. The as-prepared samples were characterized by X-ray diffraction, scanning electron microscopy, transmission electron microscopy and nitrogen adsorption-desorption isotherms. The photocatalytic activity of the as-prepared samples was evaluated by photocatalytic decolorization of Rhodamine B aqueous solution at ambient temperature. The results indicated that the average crystallite size, shell thickness, specific surface areas, pore structures, and photocatalytic activity of ZnO hollow spheres could be controlled by varying the molar ratio of glucose to zinc ions (R). With increasing R, the photocatalytic activity increases and reaches a maximum value at R ) 15, which can be attributed to the combined effects of several factors such as specific surface area, the porous structure and the crystallite size. Further results show that hollow spheres can be more readily separated from the slurry system by filtration or sedimentation after photocatalytic reaction and reused than conventional powder photocatalyst. After many recycles for the photodegradation of RhB, the catalyst does not exhibit any great loss in activity, confirming ZnO hollow spheres is stability and not photocorroded. The prepared ZnO hollow spheres are also of great interest in solar cell, catalysis, separation technology, biomedical engineering, and nanotechnology.
Introduction In the past two decades, oxide semiconductor photocatalysis has attracted extensive attention due to its wide potential application in environmental protection procedures such as air purification, water disinfection, hazardous waste remediation, and water purification (1–3). Among various oxide semiconductor photocatalysis, TiO2 and ZnO have been recognized as the excellent materials for photocatalytic process due to their high photosensitivity, nontoxic nature, and large bandgap (4–7). Although TiO2 is universally considered as the most important photocatalyst, ZnO is also a suitable alternative to TiO2 due to their similar bandgap energy (3.2 eV) and its lower cost. Moreover, in certain case, larger quantum efficiency and higher photocatalytic activity than TiO2 have been reported (8, 9). However, the photocatalytic activity of zinc oxide should be further enhanced * Corresponding author fax: +86-27-87879468; e-mail: jiaguoyu@ yahoo.com. 4902
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 13, 2008
from the viewpoint of practical use and commerce. Therefore, various methods have been developed to reduce the e-/h+ recombination rate of ZnO in the photocatalytic processes (4, 5). In particular, it has been demonstrated that the photocatalytic activity of ZnO is strongly dependent on its crystallite size, specific surface area, morphologies and textures (10). Zinc oxide (ZnO), a II-VI compound semiconductor with a wide direct bandgap (3.37 eV) and a large exciton binding energy (60 meV) at room temperature, has attracted increasing interest over the past 10 years due to its specific electrical and optoelectronic properties and wide potential applications in luminescence, photocatalysts, surface acoustic wave filters, piezoelectric transducers and actuators, gas sensors, solar cells, and so on (11). Usually, the properties of nanomaterials depend on their size, morphology, and dimensionality. Various approaches have been employed for the preparation of various morphological ZnO nanostructures. These methods include hydrothermal precipitation, chemical vapor deposition, thermal evaporation, template-directed process, solution-phase synthesis, and so on. Until now, well-defined ZnO nanostructures with various morphologies such as nanowires and nanorods, nanocables and nanotubes, helical nanorods and columns, nanobelts and nanosheets, and complicated hierarchical ZnO nanostructures have been fabricated (12–15). Furthermore, fabrication of ZnO hollow spheres has also attracted a great deal of attention because of their low density, high surface area, good surface permeability as well as hollow textures (16, 17). And, it is expected that high photocatalytic activity and large light-harvesting efficiencies could be achieved using ZnO hollow spheres as photocatalysts. Up to now, the most important methods for hollow structures rely on the use of sacrificial templates, either hard or soft template, and the desired hollow interiors are generated upon the removal of templates by calcination or dissolution (18). Very recently, some novel methods for the fabrication of metal oxide hollow spheres have been developed. For example, Sun et al. (19, 20) have reported synthesis of various oxide hollow microspheres using carbonaceous polysaccharide microspheres prepared from saccharide solution as templates. Yin et al. have reported one-pot template-free synthesis for various hollow structures based on direct solid evacuation via the Kirkendall effect (21), Ostwald ripening (22) and chemically induced selftransformation (23). However, preparation of well-crystallized ZnO hollow microspheres with controllable surface morphology and shell thickness and high photocatalytic activity is still a great challenge (24, 25). In this study, ZnO hollow spheres with crystalline nanoporous shells are one-pot fabricated according to the method reported by Titirici et al. (20) and their photocatalytic activity and environment application are carefully investigated.
Experimental Section Sample Preparation. All the reagents used in the experiments were in analytical grade (purchased from Shanghai Chemical Industrial Company) and used without further purification. ZnO hollow spheres were fabricated by a hydrothermal approach using ZnCl2 as a zinc source. In a typical synthesis, 15 g (75.5 mmol) of glucose and a certain amount of ZnCl2 were dissolved in 80 and 40 mL of distilled water under stirring, respectively. The above two solutions were mixed and the molar ratios of glucose to Zn2+ (R) varied from 0 to 20. After mixing, the pH of the mixed solutions was ca. 5.5. Then the reaction solution was transferred into a 200 mL Teflon-lined stainless steel autoclave, followed by hydro10.1021/es800036n CCC: $40.75
2008 American Chemical Society
Published on Web 05/30/2008
thermal treatment of the mixture at 180 °C for 24 h. After hydrothermal reaction, the black or puce precipitates were centrifuged, and then washed with distilled water and absolute alcohol five times. The washed precipitates were dried in a vacuum oven at 60 °C for 8 h and finally were calcined in air at 500 °C for 4 h. Characterization. Powder X-ray diffraction (XRD) patterns were obtained on a D/Max-RB X-ray diffractometer (Rigaku, Japan) using Cu KR radiation at a scan rate of 0.05 ° 2θ S1-. The average crystallite sizes were determined according to the Scherrer equation using the full-width half-maximum data after correcting the instrumental broadening. The relative crystallinity is evaluated from the relative intensity of the diffraction peak of wurtzite (101) plane using the sample obtained at R ) 15 as reference. Scanning electron microscopy (SEM) was performed with a JSM-5610LV microscope (JEOL, Japan) at an accelerating voltage of 20 kV and a S4800 field emission SEM (FESEM, Hitachi, Japan) at an accelerating voltage of 5 kV and linked with an Oxford Instruments X-ray analysis system. Transmission electron microscopy (TEM) analysis and selected area electron diffraction (SAED) were conducted using a JEM 2100F microscope at an accelerating voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) measurements were done on a VG ESCALAB MKII XPS system with Mg KR source and a charge neutralizer. All the binding energies were referenced to the C1s peak at 284.8 eV of the surface adventitious carbon. The Brunauer-Emmett-Teller (BET) surface area of the powders was analyzed by nitrogen adsorption in a Micromeritics ASAP 2020 nitrogen adsorption apparatus (U.S.). All the samples were degassed at 180 °C prior to nitrogen adsorption measurements. The BET surface area was determined by a multipoint BET method using the adsorption data in the relative pressure (P/P0) range of 0.05-0.3. A desorption isotherm was used to determine the pore size distribution by the Barret-Joyner-Halender (BJH) method, assuming a cylindrical pore model (26). The nitrogen adsorption volume at the relative pressure (P/P0) of 0.994 was used to determine the pore volume and average pore size. UV-vis diffused reflectance spectra of ZnO powders were obtained for the dry-pressed disk samples using a UV-vis spectrophotometer (UV-2550, Shimadzu, Japan). BaSO4 was used as a reflectance standard in a UV-vis diffuse reflectance experiment. Measurement of Photocatalytic Activity. The evaluation of photocatalytic activity of the prepared samples for the photocatalytic decolorization of Rhodamine B (RhB) aqueous solution was performed at ambient temperature, as reported in our previous studies (27). Experiments were as follows: 0.01 g of the prepared ZnO powder was dispersed in a 20 mL RhB aqueous solution with a concentration of 1 × 10-5 M in a rectangle cell (52 W × 155 L × 30 H mm). The solution was allowed to reach an adsorption-desorption equilibrium among the photocatalyst, RhB, and water before UV light irradiation. Experiment results indicated that the absorbed RhB was less than 10%. A 15 W 365 nm UV lamp (ColeParmer Instrument Co.) was used as a light source to trigger the photocatalytic reaction. The average light intensity striking on the surface of the reaction solution was about 112 µW cm-2, as measured by a UV meter (made in the photoelectric instrument factory of Beijing Normal University) with the peak intensity of 365 nm. The concentration of RhB was determined by an UV-visible spectrophotometer (UV-2550, Shimadzu, Japan). After UV irradiation for some time (20 min), the reaction solution was filtrated to measure the concentration change of RhB. To further determine the mineralization of RhB, changes in total organic carbon (TOC) were determined using a total organic carbon analyzer (model TOC-Aopllo 9000). The photocatalytic activity of Degussa P25 TiO2 powders (P25) was also measured as a reference.
FIGURE 1. XRD patterns of the products prepared with varying R: (a) 0, (b) 1, (c) 5, (d) 10, and (e) 15, and calcined at 500 °C for 4 h.
Results and Discussion XRD is used to investigate the changes of phase structure and crystallite size of the as-prepared ZnO powders before and after calcination. Figure 1 shows XRD patterns of the products obtained with varying R from 0 to 15 and calcined at 500 °C. All the diffraction peaks are in good agreement with those of the hexagonal wurtzite structure of ZnO (JCPDS card 36-1451). No other diffraction peaks are found, indicating that the products are pure ZnO. With increasing R, the intensities of diffraction peaks decrease, indicating that crystallite size decreases or crystallization becomes weak. This also suggests that glucose inhibits the crystallization of ZnO. Further investigation shows that in the presence of glucose, the as-prepared composite microspheres are amorphous before calcination (not shown here). On the contrary, in the absence of glucose, the as-prepared sample is crystalline ZnO before calcination. This further confirms that glucose prevents the crystallization of ZnO due to the fact that the metal precursors disperse in the hydrophilic shell of the carbon spheres as amorphous clusters before calcinations (20). According to the Debye-Scherrer formula, the average crystallite sizes of the samples are 77.0, 63.0, 27.3, 25.0, and 16.8 nm for R ) 0, 1, 5, 10, and 15, respectively. These results indicate that the higher the concentration of glucose, the smaller the average crystallite size of ZnO. The surface element composition of the as-prepared hollow spheres was also studied by XPS analysis. The XPS survey spectrum (not shown here) of the sample prepared at R ) 15 and calcined at 500 °C for 4 h indicate that no peaks of other elements except Zn, O, and C are observed. The carbon peak is attributed to the residual carbon from the sample and adventitious hydrocarbon from XPS instrument itself. The Zn2p3/2 peak at 1022 eV and the O1s peak at 531 eV can be assigned to Zn and O elements in ZnO, respectively. The high-resolution XPS spectrum of O1s region (Figure 2) displays that the O1s region can be fitted into two peaks. The main contribution (located at 530.8 eV) is attributed to the Zn-O in ZnO and the other peak at 532.1 eV can be ascribed to the OH on the surface of ZnO. Their atomic ratios are about 60 and 40%, respectively. The high hydroxyl content may be related to the destruction of Zn-O-Zn and the formation of Zn-OH on the surface of ZnO hollow spheres. Both XRD and XPS analysis indicate that the as-prepared hollow spheres are pure ZnO. Figure 3a shows SEM image of the as-prepared zinc-carbon composite microspheres obtained via the hydrothermal treatment before calcination. The average diameter of spheres is approximately 10 µm. Many spheres are aggregated and linked to each other and their surfaces are smooth. After calcination, diameters of spheres decrease drastically and VOL. 42, NO. 13, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4903
FIGURE 2. High-resolution XPS spectra of O1s region of the ZnO hollow spheres prepared at R )15 and calcined at 500 °C for 4 h.
FIGURE 4. TEM image (a), high magnification TEM image (b) and SAED pattern (inset in a) of the ZnO hollow spheres prepared at R )15 and calcined at 500 °C for 4 h.
TABLE 1. Effects of R on the Physical Properties of the ZnO Samples
FIGURE 3. SEM images of the ZnO hollow spheres prepared at R )15 before calcination (a) and after calcination (b) at 500 °C for 4 h. are about 800 nm in size (see Figure 3b). This significant shrinkage (from about 10 to 0.8 µm in size) of the structure during calcination indicates that the zinc ions adsorbed loosely on the carbon spheres have transformed into the dense zinc oxide network in the shells of hollow spheres. Further observation reveals that their surfaces are rough and porous (inset in Figure 3b), suggesting that the shells of ZnO hollow spheres are composed of many small nanoparticles. To get more information about the hollow structures, the obtained hollow spheres were further investigated by TEM. Figure 4a shows a typical TEM image of ZnO hollow spheres, further confirming the hollow interiors clearly. The shell has a thickness of about 60 nm and an inherent porosity arise from aggregated nanoparticles (see Figure 4a and inset in Figure 3b). The corresponding SAED pattern (inset in Figure 4a) reveals that the hollow spheres are polycrystalline. High magnification TEM image (Figure 4b) shows clearly that the hollow sphere was composed of randomly aggregated nanocrystal particles with sizes of about 18 nm, which was in accord with the results of XRD (see Table 1). Further investigation shows that the morphology, shell thickness, specific surface areas, average crystallite sizes, and relative crystallinity of ZnO hollow spheres could be easily controlled by changing R. As can be seen from Table 1, with decreasing R from 15 to 0, the specific surface areas decrease from 63 to 7 m2/g. On the contrary, the average crystallite size and relative crystallinity increase. Figure 5 presents SEM images of the ZnO hollow spheres prepared at R ) 5 and 1. 4904
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 13, 2008
R
crystallite size (nm)
crystallinity
SBET (m2 g-1)
pore volume (cm3 g-1)
0 1 5 10 15
77.0 63.0 27.3 25.0 16.8
6.6 5.2 2.2 1.6 1
7 8 42 58 63
0.02 0.03 0.09 0.11 0.17
It can be seen that, at R ) 5 and 1, intact ZnO hollow spheres can be easily obtained due to its thicker shells. In addition, higher concentration of Zn2+ ions (at R ) 1) results in the formation of aggregated particles and a small amount of prismlike solid ZnO particles (not shown here) due to homogeneous nucleation of ZnO itself. SEM and TEM results further indicate that the thicknesses of hollow spheres are 100 and 150 nm for the samples prepared at R ) 5 and 1, respectively (not shown here). In addition, when R further increased to 20, no hollow spheres were obtained. This is due to the fact that the shell of hollow spheres was so thin that it easily collapsed during calcination. Hence, it can be inferred from the above results that the shell thickness of the hollow spheres increases with decreasing R. This can be explained by considering increasing amount of metal salts in the shells of as-prepared carbon-zinc composite microspheres. According to the results reported by Sun et al. (19, 20) the proposed formation mechanism at different R is schematically illustrated in Supporting Information (see Figure S1). The formation of ZnO hollow spheres probably involves three steps, that is, the formation of the carbon spheres involving the dehydration of the glucose first and subsequent carbonization of the organic compounds. The surface of the spheres is hydrophilic and contains a large amount of OH and CdO groups due to non- or just partially dehydrated
FIGURE 6. Adsorption changes of RhB aqueous solution at room temperature in the presence of ZnO hollow spheres obtained at R ) 15 under UV irradiation. conductor, the relation between the absorption coefficient (R) and photon energy (hν) can be written as R ) Bd(hv - Eg)1/2 ⁄ hv FIGURE 5. SEM images of the ZnO hollow spheres prepared at R ) 5 (a) and 1 (b) and calcined at 500 °C for 4 h. glucose (19). The secondary step is the embedding of the zinc precursors into the hydrophilic shell of as-formed carbon microspheres due to the fact that the functional groups in the surface layer can bind metal cations through coordination or electrostatic interactions. Finally, the removal of carbon cores, and densification and cross-link of incorporating metal cationic ions in the layer via heat-treatment results in the formation of ZnO hollow spheres. With increasing amount of metal precursors, more metal ions are incorporated into the layer, which leads to the formation of thicker shells. Therefore, it is not surprising that the thickness of shell can be easily controlled by changing R. The nitrogen adsorption/desorption isotherms (see Supporting Information Figure S2) and pore-size distributions (see Supporting Information Figure S3) of ZnO hollow spheres at different R display that all samples show the type IV isotherms with type H3 hysteresis loops according to Brunauer-Deming-Deming-Teller (BDDT) classification (26), indicating the presence of mesopores (2-50 nm). The observed hysteresis loops shifted to a high relative pressure P/P0 ≈ 1, suggesting the presence of large pores (>50 nm) (28). For the sample obtained at R ) 15, the pore-size distribution indicates a bimodal pore-size distribution with a small (
