Graphene Spheres-CuO Nanoflowers Composites for Use as a High Performance Photocatalyst

Abstract

Graphene spheres decorated with CuO nanoflowers were successfully synthesized via a spray-drying method. Scanning-electron and transmission-electron microscopy were used to confirm that the graphene formed spherical structures and that the CuO nanoflowers were well dispersed on the surface of these graphene spheres. These novel nanocomposites had an enhanced photocatalytic performance, achieving a 95.2% decomposition of methyl orange after 15 min in the presence of H₂O₂ when irradiated by visible light. These nanocomposites performed much better than CuO powders alone.

1. Introduction

Morphology is known to influence the photocatalytic performances of semiconductor oxides [1]. Structures with nanoflower morphologies have been shown to have unique properties that differ from those of mono-morphological structures, because of the combined benefits of their nanoscale building blocks. Materials with nanoflower structures show great potential for applications in photocatalysis [2]. However, the high recombination rate of photoinduced electron-hole pairs in semiconductor oxides is the main limitation of their photocatalytic activities [3].

Graphene, which is a flat monolayer of carbon atoms, has attracted widespread attention because of its large surface area and unique electronic properties [4]. To date, numerous studies have demonstrated that excellent photocatalytic performances can be achieved by assembling nanomaterials on graphene sheets, because of graphene's high electron transfer efficiency [5]. However, graphene tends to restack through van der Waals interactions, effectively eliminating its outstanding single-layer electrical properties [6]. Spray-drying techniques are preferred for processing graphene, because these techniques are fast and can be used in continuous processes [7]. Because solvent drops evaporate rapidly in the hot air streams used in this process, individual graphene sheets remain dispersed in the water and can form spherical graphene structures with decreased graphene aggregation [7, 8]. In previous work, graphene spheres were decorated with urchin-like Cu_xO (x = 1 or 2) for use as high-performance photocatalysts, in which the graphene spheres acted as co-catalysts to promote the separation and transfer of photo-generated electrons [8]. For the integration of nanoscale building blocks, CuO nanoflowers were anchored onto the graphene spheres to generate photocatalysts. However, to the best of our knowledge, such nanomaterials have not previously been studied.

In this work, graphene spheres decorated with CuO nanoflowers were successfully synthesized using a spray-drying method. These novel composites combined the merits of both of their components. The graphene spheres acted as a conductive substrate, while the CuO nanoflowers provided more catalytically active sites. Additionally, the approach presented in this work was simple and highly efficient for the generation of nanoparticles dispersed on graphene, as compared with solution-phase methods. Experimental results revealed that these novel nanomaterials exhibited excellent visible-light photocatalytic activities for the degradation of the dye methyl orange (MO).

2. Experimental Details

Graphene oxide (GO) was prepared using the Hummers' method [9]. Graphene/CuO (GR-CuO) was prepared by adding 50 mL of 20 g L⁻¹ Cu(OAC)₂ (∼98.5%, Aladdin) to a GO solution, which was then ultrasonicated for 30 min. Then, 25 mL of 0.2 g mL⁻¹ polyvinylpyrrolidone (PVP) was added to this solution. After 30 min of ultrasonication, this blended suspension was spray dried at 200 °C, producing a blue powder. This powder was then calcinated at 220 °C for 30 min to thermally decompose the Cu(OAC)₂ and sintered at 800 °C for 1 h to reduce the GO to graphene and decompose the PVP under an Ar atmosphere in a tube furnace. The resulting powder was then allowed to cool to 300 °C before being held in air for 1 h to oxidize the decomposed Cu(OAC)₂. Products were produced with different GR to CuO weight ratios, which are referred to here as CuO, 1wt.%GR-CuO, 2wt.%GR-CuO and 5wt.%GR-CuO.

The as-prepared products were characterized using scanning electron microscopy (SEM, S4800), transmission electron microscopy (TEM, JEM-2100F), powder X-ray diffraction (XRD, D5000), Raman spectroscopy (J-Y, T6400) and X-ray photoelectron spectroscopy (XPS, K-Alpha 1063), and total organic carbon (TOC) was monitored with a TOC analyser (Apollo 9000, Terkmar-Dohrmann, USA).

Electrochemical impedance spectroscopy (EIS) was performed with a CHI660B electrochemical workstation. EIS measurements were carried out in 1 M Na₂SO₄ using a three-electrode system, which consisted of a platinum foil electrode as the counter electrode and a saturated calomel electrode (SCE) as the reference electrode. EIS measurements were recorded at 0.5 V with an alternating current (AC) voltage amplitude of 5 mV over the frequency range of 1 MHz to 5 mHz.

The Brunauer-Emmett-Teller (BET) specific surface areas and porosities of the samples were evaluated on the basis of nitrogen adsorption isotherms, which were measured at −196 °C using a gas adsorption apparatus (ASAP 2020, Micromeritics, USA).

The photocatalytic degradation of MO was measured at an ambient temperature in the presence of H₂O₂. The asprepared products (20 mg) were dispersed in 300 mL of a 20 mg L⁻¹ MO solution to which 2 mL of H₂O₂ had been added. It was placed beside the beaker at a distance of 20 cm. After stirring for 30 min in the dark, a 500 W Xe arc lamp was used to irradiate the sample. Samples of this mixture (5 mL) were withdrawn every 3 min and immediately centrifuged to separate the suspended solids. The absorbance of the supernatant was measured at 464 nm using ultraviolet-visible (UV-vis) spectrophotometry.

3. Results and Discussion

The XRD patterns of graphene, CuO and 2wt.%GR-CuO are shown in Fig. 1a. The XRD pattern of graphene had a major peak at 2θ = 25°. The as-prepared CuO and 2wt.%GR-CuO nanocomposites produced similar XRD patterns, and their diffraction peaks at 32.5°, 35.5°, 38.7°, 48.8°, 53.3°, 58.4°, 61.5°, 66.3°, 68.0° and 75.1° corresponded to the (110), (002), (111), (-202), (020), (202), (-113), (-311), (113) and (-222) crystal planes of CuO, respectively (JCPDS No. 41-0254). However, 2wt.%GR-CuO had an additional peak at 25°, which corresponded to graphene, and a peak at 42.5°, which corresponded to Cu₂O (JCPDS No. 78-2076). These results suggested that a small amount of CuO nanoparticles were reduced to Cu₂O by graphene during the thermal treatment [8].

Figure 1.

(a) XRD pattern, (b) Raman spectra, (c) Survey spectra, (d) Cu2p region XPS spectrum, (e) EIS spectra, (f) Nitrogen adsorption-desorption isotherm of the as-synthesized products

Raman spectroscopy is an important tool for the characterization of carbon-based materials. As shown in Fig. 1b, GO displayed a Raman shift in the D (1323 cm⁻¹) and G (1592 cm⁻¹) bands, arising from the disruption of GO's symmetrical hexagonal graphitic lattice and the in-plane stretching motion of symmetrical sp² C–C bonds, respectively. The D/G intensity ratio for GO was 1.08. The GR-CuO nanocomposites had Raman peaks in similar positions. However, the D/G intensity ratios of 1wt%.GR-CuO, 2wt.%GR-CuO and 5wt.%GR-CuO increased to 1.18, 1.39 and 1.31, respectively. These increased D/G ratios indicated a partial reduction of GO to graphene [10], which was also visually observed by the colour change of the product from brownish for GO to black for graphene.

The electronic state of the final 2wt.%GR-CuO product was characterized using XPS. According to Fig. 1c, peaks associated with Cu 2p, O 1s and C 1s were observed for the 2wt.%GR-CuO nanocomposite. Fig. 1d shows the Cu 2p spectra of products. The Cu2p spectrum exhibited a Cu 2p1/2 peak at 954.08 eV and a Cu 2p3/2 peak at 933.78 eV, which are characteristics of Cu²⁺ in CuO. Strong shake-up satellite peaks were also observed at 942.33 eV with an overlapping series at 962.38 eV, further confirming the presence of Cu(II) on the surface [11]. These observations suggested that 2wt.%-CuO nanocomposites were successfully synthesized.

EIS is a powerful diagnostic technique for the investigation of new materials and electrodes. As shown in Fig. 1e, the impedance plots of the GR-CuO nanocomposites had smaller radii than those of the CuO electrode, indicating that the high conductivity of graphene decreased the resistance of the GR-CuO nanocomposites. These decreased resistances facilitated electron transfer between CuO and graphene. Ultimately, this substantial decrease in charge-transfer resistance effectively prevented the recombination of electrons and holes in the nanocomposites.

Fig. 1f shows the nitrogen adsorption-desorption isotherms' curves. The BET surface areas of the CuO, 1wt.%GR-CuO, 2wt.%GR-CuO and 5wt.%GR-CuO were 35.6 m.g⁻¹, 57.4 m.g⁻¹, 61.2 m².g⁻¹ and 73.8 m².g⁻¹, which were much larger than those of the pure CuO nanoparticles (23.1m².g⁻¹) [12] and the graphene/CuO nanocomposites (53.2 m².g⁻¹) [13]. These high BET surface areas resulted from the large surface areas of the CuO nanorods that made up the CuO nanoflowers, and likely provided more photocatalytic reaction centres.

Fig. 2 shows the morphologies of the samples and the effect of graphene on the samples' microscopic structures. The 2wt.%GR-CuO sample consisted of many monodispersed graphene spheres with rough surfaces (Fig. 2a). According to the image in Fig. 2b, uniform and discrete nanoparticles covered the surfaces of the graphene spheres without forming aggregates, indicating that a strong link was formed between the nanoparticles and the graphene spheres. TEM (Fig. 2c) indicated that the graphene was densely decorated with many small nanoparticles. A higher resolution image of these nanoparticles (Fig. 2d) revealed that these nanoparticles had flower-like structures and that each of the flower-like structures was composed of multiple smaller nanorods. The lwt.%GR-CuO and 5wt.%GR-CuO samples were also composed of similar graphene spheres. However, the 1wt.%GR-CuO contained more nanoparticles, which were spread uniformly on some surfaces of the graphene spheres (Fig. 2e, f) while other large areas lacked nanoparticles and resembled crumpled silk veil waves (Fig. 2g, h). In general, spray drying successfully produced composites of graphene spheres with nanoparticles on their surfaces.

Figure 2.

(a, b) SEM (c, d) TEM image of 2wt.%GR-CuO, (e, f) SEM image of lwt.%GR-CuO, (g, h) SEM image of 5wt.%GR-CuO

To determine the effect of PVP on the formation of these nanocomposites, experiments were performed using the same chemical reagents in the absence and presence of PVP. The products produced in the absence of PVP are shown in Fig. 3a, b. While graphene still formed spherical structures without PVP, the nanoparticles became aggregated and formed large particles. With PVP added to the synthesis solution, only CuO nanoflowers were formed and almost no large CuO aggregates were observed, as shown in Fig. 2d. Clearly, PVP played an important role in the successful formation of CuO nanoflowers

Figure 3.

(a) SEM (b) TEM of 2wt.%GR-CuO without PVP

As a demonstration of a potential application, the photocatalytic performance of GR-CuO was evaluated for the degradation of MO, which was chosen as a representative organic dye. The degradation ratio was defined as Ce/C₀ where C₀ and Ce were the initial concentration of MO and the concentration of MO at a given time, respectively. Fig. 4a shows MO solutions after their adsorption-desorption equilibria were reached in the dark in the presence of the different catalysts. The images revealed that the abilities of the catalysts to adsorb MO were limited (less than 20%). However, H₂O₂ alone degraded MO only a very little when irradiated with visible light, as shown in Fig. 4b. Only the combined use of H₂O₂ and the catalysts produced a significant improvement in degradation efficiency (previous research [7] has shown that CuO catalysts without H₂O₂ under visible light produces little degradation of MO). As shown in Fig. 4c, the CuO, lwt.%GR-CuO, 2wt.%GR-CuO and 5wt.%GR-CuO samples degraded by approximately 74.3%, 90.3%, 95.2% and 78.8% of MO after 15 min of visible-light irradiation in the presence of H₂O₂. Clearly, 2wt.%GR-CuO had made the best use of visible light. Fig. 4d shows typical time-dependent UV-vis absorption spectra of MO solutions during photodegradation in the presence of 2wt.%GR-CuO. Over time, the absorption peak at 464 nm, which corresponded to the presence of MO, decreased and nearly disappeared after 15 min. The inset in Fig. 4d shows a significant decrease of the initial TOC content in the solution after 15 min, further proving the cleavage of the MO dye molecules. These results indicated that MO was successfully degraded in the presence of 2wt.%GR-CuO under visible-light irradiation. The superior photocatalytic performances of the GR-CuO nanocomposites may have resulted from the phenomena described in the following. First, the substantially enhanced specific surface areas of the graphene spheres provided an ideal substrate for the loading of nanoparticles without forming aggregates [14]. The graphene spheres were also highly electronically conductive and acted as electron transporters, allowing the photo-excited electrons in the composites to be quickly transferred from CuO to GR, and eventually enabling the enhanced photocatalytic activities of the composites [15]. It should be noted that the graphene content influenced the extent to which the MO degradation reaction was enhanced. The photocatalytic activities of the GR-CuO systems were enhanced when the amount of graphene added was increased. This behaviour was ascribed to graphene's ability to effectively inhibit the recombination of photo-excited electron-hole pairs in the nanocomposites. However, an increase in the nanocomposites' graphene contents above 2wt% decreased the transfer of electrons from the excitation state of the CuO nanoflowers, lessening the photocatalytic activity of these materials. Finally, the CuO nanoflowers consisted of nanorods that possessed large specific surface areas and provided more photocatalytically active reaction sites. This increase in the reactive surface area likely enhanced the photocatalytic performances of the GR-CuO nanocomposites.

Figure 4.

Photocatalytic degradation efficiency of MO (a) with catalyst and without H₂O₂ in the dark, (b) with H₂O₂ and without catalyst under visible light, (c) different catalysts with H₂O₂ under visible light, (d) UV-vis absorption spectra of MO solution with 2wt.%GR-CuO at different time intervals (inset figure in (d) shows the change of TOC with time)

Sheme 1.

Proposed scheme of photodegradation mechanism

The proposed photodegradation mechanism of the GR-CuO nanocomposites is shown in Scheme 1. When irradiated by visible light, valence band (VB) electrons in the CuO nanoflowers were excited in the conduction band (CB). However, these charge carriers easily recombined, leading to a low photocatalytic performance. When the nanoparticles were tightly anchored to the surfaces of the graphene spheres, the graphene served as an electron transporter, which efficiently transported those photo-generated electrons and improved electron-hole separation. Simultaneously, active radicals (such as •OH) generated from the reduction of H₂O₂ by photo-generated electrons directly decomposed MO into CO₂ and H₂O.

4. Conclusions

In this study, CuO nanoflowers anchored to graphene spheres were prepared using a spray-drying method. The photocatalytic performances of the GR-CuO nanocomposites were increased considerably compared to that of CuO alone, based on the rate of MO degradation under irradiation by visible light. This work not only demonstrated the potential of graphene as a support for nanoparticle-based photocatalysts but also highlighted more generally the potential applications of graphene-based materials in other fields.