Inorganic materials, mainly silica and metal oxides (MOs), with long-range ordered nanostructures have found a wide range of significant applications.^[Reference Blanford, Yan, Schroden, Al-Daous and Stein^1–Reference Milliron, Buonsanti, Llordes and Helms^5] Two types of template methods are typically used to achieve the long-range ordering.^[Reference van Bommel, Friggeri and Shinkai^6–Reference Cheng, Tao, Liang and Chen^8] Soft template method uses surfactants or polymers which form micelles with certain nanoscale features in the liquid phase to guide the growth of inorganics, whereas hard template method (or nanocasting) uses hard templates such as anodic aluminum oxide (AAO) membranes, mesoporous silica, zeolites, and carbon, which possess well-defined voids, channels or other mesostructures, to synthesize inorganics with negative replicas of the template structures. Typical procedures involved in the hard template method include: (I) impregnation of metal- or silicon-containing precursors into a template; (II) nucleation and growth of inorganic nanostructures; (III) removal of the template by chemical treatment (e.g., treatment with HF or NaOH to remove silica template) or calcination (typically for removal of polymer and carbon templates). The hard template method has the advantage of creating a variety of long-range ordered inorganics with high crystallinity. For example, α-Fe₂O₃ with an ordered mesoporous structure and crystalline walls was synthesized using mesoporous silica KIT-6 as a hard template.^[Reference Jiao, Harrison, Jumas, Chadwick, Kockelmann and Bruce^9] Chiral nematic ordering was successfully transferred to free-standing nanocrystalline TiO₂ films using chiral nematic mesoporous silica films as hard templates.^[Reference Shopsowitz, Stahl, Hamad and MacLachlan^4]

Here, we demonstrate that free-standing graphene oxide (GO) film (or paper) can be used as a novel new type of templates for the syntheses of MO and reduced GO (rGO)/MO hybrid materials with ordered nanocrystalline MO particles. GO film possesses well-defined layers with a nano-ranged d-spacing (0.6–1.2 nm),^[Reference Dikin, Stankovich, Zimney, Piner, Dommett, Evmenenko, Nguyen and Ruoff^10, Reference Dreyer, Park, Bielawski and Ruoff^11] which templates the assembly of MO nanocrystals into two-dimensional (2D) sheets and 3D lamellar structures. Compared to nanocomposites with random phase distribution and orientation, 2D MO nanocrystal sheets have been demonstrated to increase the active surface area, improve photocatalytic activity and chemical stability,^[Reference Gunjakar, Kim, Kim, Kim and Hwang^12] and improve electrocatalytic activity.^[Reference Shin, Go, Kim, Lee, Lee and Hwang^13] The 3D lamellar rGO/MO hybrids synthesized using our method can potentially be used as efficient and durable electrocatalysts in energy conversion and storage as a result of the strong chemical attachment and electrical coupling between graphene and in-situ grown MOs.^[Reference Liang, Li, Wang and Dai^14] In addition, the synthesis approach presented here might lead to new strategies for achieving lamellar nanostructures in thin-film applications.^[Reference Bruns, Herman, Minuzzo, Lehrman and Stupp^15]

A schematic illustration of our synthesis process is shown in Fig. 1, which includes the infiltration of metal ions into a GO film (step I), nucleation and growth of MO crystallites in the film (step II), and removal of the GO film template by annealing in air (step III). Layered rGO/MO hybrids can also be prepared by annealing in argon instead of air during step III. 3D lamellar ZnO, α-Fe₂O₃, and rGO/ZnO hybrid were synthesized as examples to demonstrate this general approach. In addition, 2D assembly of MO nanocrystallites into microsheets was also observed.

Figure 1.

Schematic illustration of the GO film-templated synthesis process.

GO was synthesized via modified Hummers method,^[Reference Hummers and Offeman^16, Reference Gao and Tang^17] of which the detailed procedures can be found in supplementary materials (SM). In brief, graphite was intercalated and oxidized with KMnO₄, NaNO₃, and concentrated H₂SO₄, and exfoliated to single-layer GO by ultrasonication in water. GO platelets were deposited on freshly cleaved mica substrates by drop casting of a GO water/ethanol solution at a volume ratio of 1:9 and used for atomic force microscopy (AFM) characterization. AFM images of GO platelets [Fig. 2(a)] were collected using a Nanoscope MultiMode^TM AFM (Veeco) in the tapping mode at a scan rate of 1.5 Hz. The GO platelets were of irregular shapes and their sizes ranged from hundreds of nanometers to a few micrometers. The height profile across the indicated line showed that the thickness of GO platelets was ~1 nm (Supplemental Fig. S1, SM online), which is consistent with the literature reported thickness of single-layer GO (0.8–1.2 nm).^[Reference Stankovich, Dikin, Piner, Kohlhaas, Kleinhammes, Jia, Wu, Nguyen and Ruoff^18] Free-standing GO films were prepared by assembling GO platelets under a directional flow using vacuum filtration. Each GO film was prepared by filtering 4 ml of GO aqueous solution (2.0 mg/ml) using a 0.1 µm polycarbonate filter membrane (Whatman; diameter 25 mm) and air-dried. The GO films appeared uniform and possessed a brown color as shown in Fig. 2(b). Scanning electron microscopy (SEM) image of the cross-section of a GO film [Fig. 2(c)] showed well-defined lamellar structure resulting from the close stacking of GO platelets.

Figure 2.

(a) AFM image of individual GO platelets. (b) Optical image of a free-standing GO film. (c) SEM image of the cross-section of a GO film.

For the synthesis of 3D-ordered zinc oxide, Zn-containing precursor solution was prepared by dissolving 1 g Zn(CH₃COO)₂·2H₂O in a solution of 25 ml ammonium hydroxide (28 wt%) and 50 ml deionized water, which was stirred vigorously for half an hour. A small volume (8 ml) of the precursor solution was then slowly filtered through a GO film. The filtration speed was controlled to be slow to ensure effective adsorption of Zn-containing precursors within the GO film. Then the wet film was fixed between two glass slides, soaked in the precursor solution in a 200 ml glass beaker and subjected to microwave irradiation in a home-use oven (1100 W, 2.45 GHz) for 2 min. Sandwiching the film with glass slides helped prevent the film from being torn apart by the boiling solution. After the microwave treatment, the film was taken out, washed thoroughly with water and ethanol to remove the surface-adsorbed byproducts, and dried in air. The remaining synthesis solution was filtered to collect free particles formed in the solution. Finally, the film was then calcined at 600 °C in air to remove GO and meanwhile increase the crystallinity of ZnO. A free-standing, white, and highly reflective ZnO film was obtained, as shown in Fig. 3(a). Energy-dispersive X-ray spectroscopy (EDX) analysis of the ZnO film confirmed that GO was removed at the calcination step and the film consisted of only ZnO (Supplemental Fig. S2, SM online). SEM images, cross-sectional views [Figs. 3(c) and 3(d)], and top view [Fig. 3(c) inset], of the ZnO film showed that the layering order of the GO film was successfully replicated in the ZnO film. By contrast, synthesis under the same conditions yielded flower-like ZnO particles [Fig. 3(b)] in the solution (not in the GO film).^[Reference Mazloumi, Mandal and Tang^19] SEM images and EDX data were collected using a LEO 1530 field-emission SEM equipped with an EDX attachment (Zeiss).

Figure 3.

(a) Optical image of a free-standing ZnO film. (b) SEM image of flower-like ZnO particles formed in solution (not in GO film). (c, d) Cross-sectional SEM images of a ZnO film at different magnifications. Inset in (c) is a top view SEM image. (e, f) Cross-sectional SEM images of an rGO/ZnO hybrid film at various magnifications.

Small pieces of ZnO films and powders of ZnO particles were used for X-ray diffraction (XRD) characterization, and the data were recorded with a Bruker D8-Advance diffractometer equipped with Cu-Kα radiation (λ = 1.5406 Å). XRD patterns showed that both the lamellar structured ZnO film and the flower-like particles were composed of wurtzite ZnO [Inorganic Crystal Structure Database (ICSD) No. 065121] (Fig. 4). No peaks from impurities were present, indicating the high ZnO purity. The XRD spectrum of the ZnO film showed a predominant (002) peak and much suppressed (100) and (101) peaks compared with the spectrum of the flower-like ZnO particles, suggesting that the crystallites in the ZnO film were preferentially oriented along the c-axis. Similar XRD results were observed for 3D perpendicularly oriented ZnO microrod array^[Reference Vayssieres, Keis, Lindquist and Hagfeldt^20] and ZnO nanowires grown vertically on ZnO films.^[Reference Tseng, Huang, Cheng, Lin, Liu and Chen^21] The relatively broadened diffraction peaks of the ZnO film with respect to those of ZnO flowers indicated the smaller grain size in the former. The grain size (D) in ZnO film was evaluated to be 13 nm using the Scherrer equation,

(1)$$D = K\lambda /\lpar \beta {\rm cos}\theta \rpar \comma \;$$

where K is a shape factor and taken as 0.9 here, λ is the wavelength of the X-ray, β is the full-width at half-maximum (FWHM) of the (002) peak in radians, and θ is the diffraction angle.^[Reference Cullity^22] The use of microwave irradiation for the growth of ZnO micro- and nanostructures is well known.^[Reference Cho, Jung and Lee^23, Reference Mazloumi, Shadmehr, Rangom, Nazar and Tang^24] However, the XRD pattern of the film right after microwave treatment suggested that ZnO nucleated on GO had poor crystallinity (Supplemental Fig. S3, SM online). Therefore, microwave irradiation might have promoted particle nucleation, while thermal annealing improved particle crystallinity and enlarged crystallite size.

Figure 4.

XRD patterns of ZnO film (top) synthesized using GO film as a template and ZnO flower (bottom) synthesized without using the GO film template.

By thermal annealing in argon, instead of air, 3D layered and highly porous rGO/ZnO hybrid films could be obtained [Figs. 3(e) and 3(f)]. EDX data of the hybrid films confirmed the presence of ZnO (Supplemental Fig. S4, SM online). The rGO/ZnO layers were evidently more rippled, leaving mesopores between layers. The creation of the mesoporous structure might be attributed to the rapid generation of gaseous byproducts such as carbon dioxide by the thermal decomposition of the oxygen-functional groups on GO, which disrupted the compact stacking of GO platelets, similar to the reduction of GO films by laser irradiation.^[Reference El-Kady, Strong, Dubin and Kaner^25] The conversion of initially closely stacked GO sheets to well-exfoliated rGO sheets [Figs. 2(c) and 3(f)] by our method might be beneficial for electrochemical applications. XRD patterns of the hybrid (Supplemental Fig. S5, SM online) showed (100), (002), and (101) peaks, confirming the successful synthesis of ZnO. The shift of the (002) peak of graphitic materials from 12.5° in the GO film to 25.6° in the rGO/ZnO hybrid film indicated the thermal reduction of GO.^[Reference Moon, Lee, Ruoff and Lee^26] We also annealed the GO film alone (no ZnO) in argon, which showed similar shift of the (002) peak in the XRD patterns (Supplemental Fig. S5, SM online).

Transmission electron microscope (TEM) characterization was carried out to examine the organization of crystallites in the lamellar ZnO film and the domain size. Small pieces of ZnO films were soaked in ethanol and subject to mild sonication. Then the sample was deposited on a copper grid for TEM imaging (Philips CM10). Interestingly, 2D assembly of ZnO crystallites into microsheets was observed, which is templated by the 2D GO platelets as expected. The multi-layered ZnO sheets could be exfoliated to few-layered or single-layered sheets by sonication in ethanol as shown in Figs. 5(a) and 5(b). Some sheets were very flat, while others have wrinkles, as indicated by the arrow in Fig. 5(a). The wrinkles on the ZnO sheets were believed to be replicas of the wrinkles in the GO film which were formed during the film preparation process due to the high flexibility of GO platelets (Supplemental Fig. S6, SM online). High-resolution TEM image [Fig. 5(c)] showed that the crystallites were monodisperse with sizes about 15 nm, which is consistent with the calculated particle size using the (002) peak in the XRD spectrum. The compact stacking of GO layers, with narrow d-spacing (~1 nm), provided a unique environment for achieving small ZnO nanocrystals and assembling them into a monolayer. In comparison, ZnO crystals formed in solution under the same synthesis condition were a few 100 nm in size. The fusion of individual grains to make an interconnected framework was also observed after the sintering process.

Figure 5.

TEM images of ZnO nanocrystal sheets at various magnifications.

To demonstrate the versatility of our method, free-standing α-Fe₂O₃ film with lamellar order was also synthesized, as shown in Supplemental Figs. S7(a) and S7(b), SM online. Fe-containing precursors, made by dissolving Fe(NO₃)₃ in DI water, was impregnated in GO films using the same filtration method, and then GO was removed by calcination at 600 °C in air; meanwhile α-Fe₂O₃ was formed by thermal decomposition of Fe(NO₃)₃ during the calcination process.^[Reference Jiao, Harrison, Jumas, Chadwick, Kockelmann and Bruce^9, Reference Chen, Xu, Li and Gou^27] EDX analysis and XRD pattern confirmed the formation of α-Fe₂O₃ [Supplemental Figs. S7(d) and S7(e), SM online]. Similarly, TEM characterization showed the 2D assembly of α-Fe₂O₃ nanocrystallites into microsheets [Supplemental Fig. 7(c) online].

In summary, free-standing 3D lamellar ZnO and α-Fe₂O₃ structures were synthesized using GO films as templates. Such lamellar structures could also be exfoliated to obtain 2D assembled MO microsheets. It is anticipated that this method can be extended to the syntheses of other layered MOs, for example, TiO₂, SnO₂, and Co₃O₄, considering its facileness and the similarity between our method and other hard template methods reported in the literature.^[Reference Shopsowitz, Stahl, Hamad and MacLachlan^4, Reference Lee, Orilall, Warren, Kamperman, DiSalvo and Wiesner^28–Reference Li, Xu and Chen^31] The easiness in preparation, the large surface area of GO platelets for MO nucleation, and the nanometer-sized inter-layer spacing make GO films a unique type of templates for tightly controlling MO nanocrystal size, altering preferential crystal growth direction, and more importantly achieving 2D assembled and 3D lamellar MO and rGO/MO hybrids. Vast applications can be envisioned by applying such materials in optoelectronics, catalysis, sensing, energy conversion, and storage.