Water pollution causes great damage to ecosystems, human health, as well as the sustainable economic and social development because the pollutant complex, along with bacteria, cause difficulty in decontamination by conventional water treatment processes (Schwarzenbach et al., 2010; Wang and Yang, 2016). Hence, developing an effective and facile way to degrade pollutants has become an active area in environmental research. Recently, inorganic nanomaterials have attracted numerous attentions because of their controllable shapes and sizes, as well as their effective photocatalytic activities, such as those in metal oxide semiconductors (e.g. TiO₂, ZnO) or narrow band gap semiconductors (e.g. Ag₃PO₄) (Shao et al., 2019; Zhang et al., 2020).

Although these inorganic materials exhibit promising photocatalytic activities, there are still several problems that need to be overcome, for instance, relatively poor light-harvesting abilities in the visible region, the use of toxic or harmful chemicals, or poor charge separation and transport. Cuprous oxide (Cu₂O) is a promising metal oxide material in the application of photocatalysis because it is a p-type semiconductor (C _hole > C _electron) with a small band gap (E_g = 2.17 eV). Recently, numerous efforts have been devoted to synthesize Cu₂O with different morphologies such as nanowires, octahedra, cuboctahedra, and nanocubes (Hua et al., 2011; Deng et al., 2012; Hou et al., 2013; Hong et al., 2014). Among them, a hollow structure has intrinsic advantages in photocatalysis applications Xiao et al. (2019), such as enhancing light harvesting Li et al. (2007), Wang et al. (2012), Wu et al. (2013), Dinh et al. (2014), Qi et al. (2014), promoting the synergistic effects of light scattering and localized surface plasmon resonance (LSPR) Zhang et al. (2014), Shi et al. (2016), reducing charge recombination Marschall (2014), Li et al. (2015), and accelerating surface reactions due to a high surface area (Sun et al., 2003; Vaughn and Schaak, 2012; Wang et al., 2017). Furthermore, due to a large carbon sheet structure, graphene oxide (GO) was also introduced in a hybrid with Cu₂O to effectively increase the adsorption sites and improve the transfer of electrons between the materials for the inhibition of hole and electron recombination. The addition of carbon based 2D materials during the in-situ preparation of the nanoparticles may improve the photocatalytic performance effectively (Khan et al., 2015; Khan et al., 2016a; Khan et al., 2016b; Khan et al., 2017; Ahmed and Haider, 2018; Khan et al., 2018). Lee and his group synthesized Ag-Cu₂O together with graphene oxide for the enhanced photocatalytic performance (Sharma et al., 2018). SunilMeti et al. reported zinc oxide nanocomposites wrapped with reduced graphene oxide to enhance the photocatalytic activity (Meti et al., 2018).

Herein, the nontoxic and novel visible-light-driven GO-Cu₂O composite was used as an inspiring photocatalytic material to address the aforementioned problems in water pollution. In this work, aqueous solutions of copper salt, alkali, surfactant, and reductant were used to prepare hollow dodecahedral Cu₂O through an in-situ synthesis process. The GO sheet was added through an electrostatic reaction of negatively charged GO and Cu ions and leaving a final modified GO-Cu₂O hollow dodecahedral (GCD) structure. The solid cubic GO-Cu₂O structures were also fabricated by systematically changing the reductant amount for comparison. After the successful preparation of the crystal components, the structures were characterized and confirmed. The photocatalytic performance of the GO-Cu₂O in different dyes under visible light were investigated and compared, while the antibacterial performance was evaluated by a flat colony counting method and TEM. In general, the as-prepared GCD enhanced light-harvesting, separated the excited e⁻-h⁺ pairs, and promoted charge transfer. The increased reactive oxygen species generated from visible irradiation made the oxidation of the organic pollutant and elimination of the bacteria possible, as shown in Scheme 1.

With the addition of the NH₂OHHCl reductant during the in-situ process, the Cu(OH)₂ was the first nucleation seed and then the Cu₂O nanocrystals grew on the surface to form various morphologies. The higher concentration of the NH₂OHHCl increased the growth rate of the Cu₂O and changed the morphology from a cube to a dodecahedron, while the HCl from the reductant etched the Cu₂O to form the final hollow structure (Kuo and Huang, 2008).

The morphologies of the samples were characterized by SEM. Figure 1A showed that the synthesized CC possessed a cube morphology of about 100–500 nm in size. From the SEM image in Figure 1B, CD displayed a dodecahedron morphology, and the average diameter was about 300 nm, most of which were broken or full of holes. After the addition of GO to the samples, the ionic interaction between the Cu cations and GO⁻ anions modified the Cu₂O morphology. The SEM image of the GCC composites (Figure 1C) highlighted the presence of cube-like Cu₂O polyhedrons with a uniform size of less than 200 nm, most of which were wrapped by GO sheets. The addition of GO also initiated a minor change of the morphology. The GCD in Figure 1D illustrated that the dodecahedron-like particles have an average diameter of about 100–200 nm.

The solid cubes and hollow dodecahedra were also characterized by TEM. The size of the CC particles coated with the GO sheets (Figure 2A) was estimated to be between 150–200 nm, which was consistent with the SEM image of the GCC particles in Figure 1C. Figure 2B was a typical TEM image of CD covered with a small quantity of GO sheets, from which the hollow structure and the presence of GO were clear.

After determining the composites’ morphologies by SEM and TEM, the XRD spectra were utilized to confirm the prepared crystal components. According to the JCPDs, the peaks at (110), (111), (200), (220), (311), and (222) were assigned to the standard cubic structure Cu₂O (No. 03–0898) (Wang et al., 2002; Zhang et al., 2007). No other peaks were detected, such as CuO and Cu cupric oxide, demonstrating the purity of the as-obtained products. There was no clear peak for GO observed in the XRD pattern, due to the small amount and low diffraction intensity. The presence of GO was confirmed by Raman and TEM. The (200) diffraction peak of the Cu₂O cube was stronger than the other diffraction peaks, indicating a high proportion of (100) facets. Similar results were also confirmed for the (220) diffraction peak of the rhombic dodecahedral Cu₂O nanocrystals, implying a high proportion of (110) crystal planes. Also, Raman spectroscopy was used to measure the vibrations of the sp²-hybridized carbon atoms for the confirmation of the presence of GO. Hence, it was observed from Figure 3B that the Raman spectra of the as-synthesized samples contained the D peak (1,343cm⁻¹, disorder-activated Raman mode) and G peak (1,588 cm⁻¹, sp² hybridized carbon) which were assigned to GO (Yang et al., 2015; Yang et al., 2009; Krishnamoorthy et al., 2012). Figure 3C shows the UV/vis spectra of the as-synthesized CC, CD, GCC, and GCD nanocrystals. Interestingly, after the introduction of GO, the absorption abilities of the GCC and GCD both increased because of the scattering effect of the GO. The inset figure was the Kubelka-Munk transformation of light energy versus energy to calculate the band gap (Yang et al., 2013a; Yang et al., 2013c; Sasca and Popa, 2013). As shown in the figure, the band gaps of the CC, CD, GCC, and GCD were approximately 1.16, 1.01, 0.79, and 0.33 eV respectively. Because of the enhanced light harvesting, slow photons, and the synergistic effect of light-harvesting and LSPR (Xiao et al., 2019), the band gaps of the hollow structures were narrower than the solid structures. After the addition of GO, the proposed band gap of the GO-Cu₂O composite shifted to lower energies compared with pure Cu₂O. This indicated that the addition of the GO generated an extra band that served to narrow the band gap of the Cu₂O for further enhancement of the photocatalytic and bactericidal properties (Yang et al., 2015). In the FT-IR (Figure 3D), the peaks at 1,625 and 1727 cm⁻¹ corresponded to the bending and stretching vibrations of O-H and C=O, in the COOH groups of GO sheets, respectively. Also, the C-O and -CH₃ stretching peaks, and the stretching vibrations of the quinoid ring/benzenoid ring from poly(o-anisidine) were around 1,250, 1,386, 1,559, and 1,506 cm⁻¹, respectively. Also, the Cu-O stretching vibration of the Cu₂O was found at 624 cm⁻¹. In general, the characterizations from XRD, Raman, UV/vis, and FT-IR spectroscopies confirmed the successful preparation of the nanocomposites.

The photocatalytic activities of the GO-Cu₂O (50 mg) for the absorption and degradation of various organic pollutants were determined when different photocatalytic dyes (MO-10ppm, RhB-10ppm, Phenol-60ppm) were chosen to react under visible light (Figure 4 and Supplementary Figure S1). Figure 4 1) and (b) demonstrate the degradation curves for methyl orange (MO) using different materials under visible light irradiation. Figure 4A displays the changes in degradation for MO using GCD. It was observed from Figure 4A that after the ultrasonic mixing of the solution for 30 min, the solution was absorbed to some degree. The absorption capacity of pure CD was about 80%, which was higher than the GCD composites. The inset in Figure 4A indicated the degradation efficiency of the GCD increased as the amount of GO. When 40 mg of GO was introduced into the pure Cu₂O, the corresponding composite showed the highest photocatalytic activities of near 100% within 60 min. However, further addition of GO led to a decrease in the photocatalytic activity of the composites. The possible reason was that too much GO may fully wrap the GCD to eradicate the light irradiation.

To further understand the photocatalytic process, the visible light photocatalytic experiments shown in Figures 4B–D were investigated using MO, RhB, and phenol as target dyes. As shown in Figure 4B, the curve of MO using the GCD was the lowest, whereas the best performance for degrading MO and phenol (Figures 4B, D) was the pure Cu₂O cube. The efficiencies of absorption and degradation of MO, RhB, and phenol (Figures 4B–D) using the GCD composite were approximately 100%, 75%, and 5% respectively, within 120 min. While the efficiencies of MO, RhB, and phenol using GCC were 50%, 60%, and 6%. Hence, it was suggested that the GCD composite exhibited excellent photocatalytic performance for colored organic dyes. However, the performance of colorless dyes was very poor. This demonstrated that the photocatalytic effect of the GO-Cu₂O nanoparticles was mainly due to the adsorption of the colored dyes instead of degradation of organic dyes, which was the limitation of the synthesized photocatalytic materials.

Using the experimental results, a possible photocatalytic mechanism was deduced. Generally, the mechanism of the GO-Cu₂O catalytic activity was similar to other photocatalytic materials. The mechanism consisted of the absorption and degradation of the dyes and the absorption was greater than the degradation in the photocatalytic performance. The possible photocatalytic absorption mechanism was in large part due to the structure of the Cu₂O. Due to the advantage of a hollow structure, the GCD nanocomposites had the highest absorption. In addition, GO sheets offered more active adsorption sites, which also improved the adsorption of dyes. On the other hand, the photocatalytic reaction was initiated under the irradiation of visible light, leading to the separation of electron-hole pairs in the Cu₂O. The separated electrons were then excited and moved from the VB to the CB, leaving the holes in the VB. The reactive holes at VB and the reactive oxygen species (ROS) generated through reaction of O₂ and H₂O can degrade the dyes. Moreover, GO transferred electrons from the Cu₂O to the GO sheets which maintained the stability of the material. However, too much GO entangled the Cu₂O, especially the dodecahedral Cu₂O, and prevented it from absorbing the colored dyes.

After the investigation of the photocatalytic performance of GCD, the flat colony counting method was also introduced to examine the specific bactericidal effect (Dahle et al., 2004; Yang et al., 2015; Choudhry, 2016). As shown in Figure 5, the bacterial number was calculated after the addition of GCD at 0, 24, and 48 h. Specifically, the original concentration of the pathogenic bacteria was approximately 1*10⁷ CFU/ml. After the incubation of the bacteria for a period of 24 and 48 h without the GCD, an obvious increase was observed in the E. coli and S. typhi at 1*10^7.5 CFU/ml. The amount of S. aureous and P. aeruginosa remained the same. However, with the treatment of GCD at 24 h, the number of bacteria decreased from 1*10^7.5 to 1*10^4.5 CFU/ml with an antibacterial rate of more than 99.9%. After the incubation time was prolonged to 48 h, the number of the S. aureous was reduced to 1*10^2.5 CFU/ml, and the specific bactericidal rate for all the bacteria was summarized in Table 1. Considering the bactericidal experiment, it was concluded that the prepared GCD had a broad-spectrum antibacterial activity toward pathogenic bacteria.

To discuss the mechanism of the bactericidal effect, the morphological changes of E. coli and S. aureus were investigated by TEM before/after the addition of GCD. Figure 6 (a,b) showed the original E. coli rod-like structures with a size of 1 μm × 0.5 μm, while the S. aureus has a spherical structure with a diameter of 0.5 μm in the absence of GCD. However, after the treatment of samples that involved the light irradiation of the GCD, ROS was generated to interact with the membrane, making the cytoplasm flow out of the bacteria, and finally killed them. Also, other research considered that the presence of Cu ions in the GCD reacted with the oxygen to produce ROS through the Fenton reaction for further photocatalytic and bactericidal performance (Touati, 2000; El Saeed et al., 2016; Shen et al., 2020). Generally, GCD acted as a kind of bifunctional material for applications in photocatalysis and bactericide.

In summary, an effective in-situ synthesis to produce different morphologies of GO-Cu₂O was demonstrated. After characterizing with SEM, TEM, XRD, Raman spectroscopy, UV/vis spectroscopy, and FT-IR spectroscopy, the degradation performance of Cu₂O and GO-Cu₂O for different dyes under visible light was measured. Antibacterial experiments were also investigated against pathogenic bacteria. The presence of GO along with the hollow structure created a synergistic effect that increased the photo harvesting and facilitated the electron transfer to generate more ROS for the enhancement of the photocatalytic and bactericidal performances. Herein, this work offers new insights into the facile synthesis of GO-based nanocomposites for the applications of photocatalytic degradation and sterilization of wastewater pollutants using visible light.