The emergence of semiconductor-based photocatalysis has drawn increasing interest because of its universal applications in environmental remediation and solar energy utilization^[1,2]. As a visible-light responsive photocatalyst, WO₃ has been widely used with the advantages of non-toxic and stable chemical properties, photocorrosion resistance and photosensitivity^[3,4,5,6,7]. However, pristine photocatalyst WO₃ always exhibit fast electron- hole recombination for its inherent physicochemical properties under light irradiation. This phenomenon could result in difficulty in the utilization of photo-generated electrons to generate active species and increase the electron-hole recombination rate, which greatly limits its application in the field of photocatalysis.

To solve the problem mentioned above, noble metal loading is one of commonly used approaches to achieve better charge separation for efficient photocatalysis^[8,9,10,11]. For example, Xi, et al.^[11] reported the noble-metal particles grown on WO₃ could facilitate the photogenerated electron-hole separation process due to high electron conductivity of the metal particles. However, the high cost and limited abundance of noble metal hinder their wide application, non-noble metals were applied as alternative ones. The introduction of low valence copper has also been proved as a powerful method for improving photocatalytic activity^{[12,13,14,15]}.

Recently, much attention has been paid to promoting the photocatalytic activity of cube-like WO₃ nanostructure by coupling with noble metal or other semiconductors^[16,17,18]. Inspired by this background and current challenge, non-noble metal Cu decorated WO₃ nanocube was designed in this work at room temperature. The photocatalytic performance of the WO₃-Cu hybrid was evaluated, and the contribution to the photocatalysis enhancement was also discussed.

In synthesis of WO₃ nanocube, 3.0 g Na₂WO₄·2H₂O was firstly dissolved in 65 mL H₂O. Then, 15 mL HCl (37%) was added into the solution gradually under vigorous magnetic stirring for 30 min. The resulting suspension was transferred to a 100 mL Teflon-lined autoclave and maintained at 180 ℃ for 10 h. After cooling down to room temperature, the precipitant was centrifuged and washed thoroughly with deionized water and ethanol for several times, and finally was dried at 60 ℃ overnight^[17].

In fabrication of WO₃-Cu hybrid, 0.2 g WO₃ was firstly put into a round bottle containing 10 mL deionized water and 10 mL ethylene glycol. After sonication, 0.03 mmol CuSO₄·5H₂O was dissolved into the solution under sonication. Then, 10 mL 0.5 mol/L ascorbic acid aqueous solution was gradually dropped into the above solution and further stirred for 2 h. Finally, the solid product was centrifuged and washed thoroughly with deionized water and ethanol for several times, and was dried at 60 ℃ overnight. This sample was denoted as WO₃-Cu-0.03. When 0.08, 0.3, 1.0, 3.0 mmol CuSO₄·5H₂O was involved into the reaction system, the sample was denoted as WO₃-Cu-0.08, WO₃-Cu-0.3, WO₃-Cu-1.0, WO₃-Cu-3.0, respectively.

X-ray powder diffraction (XRD) was carried out on a Bruker D8 Advance diffractometer by using Cu Kα radiation at a scan rate of 10 (°)/min in the 2θ range from 10° to 80°. Scanning electron microscopy (SEM) image was conducted on a JSM 5510LV at the operating voltage of 5 kV. The energy-dispersive X-ray spectrum (EDX) analysis was carried out on an Oxford Instrument INCA with a scanning range from 0 to 20 kV. Transmission electron microscopy (TEM) images were visualized on a JEM-2000, using an accelerating voltage of 200 kV. The Electron Spin Resonance (ESR) was performed on a Hitachi ESR spectrometer (JES, FA200). The transient photocurrent density and electrochemical impedance spectra (EIS) were analyzed on a CHI660E electrochemical workstation with a standard three-electrode system according to previous study^[19].

In a typical photocatalysis test, 0.02 g catalyst was added into a reactor containing CR solution (40 mL, 20 mg/L) and sonicated to get homogeneous solution. Then, it was stirred for 1 h in the dark to reach adsorption-desorption equilibrium. Subsequently, the reactor with WO₃-Cu catalyst and CR solution was exposed to simulated sun light irradiation by a 500 W Xe lamp with magnetically stirring. 3 mL of the suspensions at each irradiation time interval was collected and then centrifuged. The collected upper solution was analyzed by a Shimadzu UV2800 spectrophotometer.

Fig. 1(a) shows the photographs of the as-synthesized WO₃-Cu samples. It can be observed that the color gradually deepens with the increase of Cu content. As shown in Fig. 1(b), all diffraction peaks of the samples can be assigned with the standard pattern of WO₃ (JCPDS 72-677), indicating that the hybrid contained the main phase of WO₃. The symbol of ★ corresponds to the Cu species in the XRD patterns. With Cu amounts increasing, the diffraction peak of Cu occurred and the intensity became stronger. This result suggested that WO₃-Cu hybrid be successfully prepared through the ascorbic acid reduction method at room temperature.

As shown in Fig. 2(a), the UV-Vis diffuse reflectance spectra demonstrate that the WO₃ nanocube have absorption edge in the visible light region. Compared to pure WO₃ product, a red shift to higher wavelength in the absorption edge of WO₃-Cu hybrids was observed. Fig. 2(b) reveals the band gap of the pure WO₃ is about 2.49 eV. With the content of Cu increasing, the obtained WO₃-Cu sample showed a narrow band gap. This result further confirms that the WO₃-Cu hybrids are successfully synthesized. The valence band (E_VB) and conduction band (E_CB) edge of WO₃ sample at the point of zero charge can be calculated by the empirical equation E_VB= X-E^e+0.5E_g, where X is the electronegativity of the semiconductor, E^e is the energy of free electrons on the hydrogen scale (about 4.5 eV), E_g is the band gap energy of the semiconductor, and E_CB can be determined by E_CB=E_VB-E_g^{[20,21,22,23]}. Based on the above equation and the DRS spectra, the E_CB and E_VB of WO₃ sample were calculated to be 0.83 and 3.38 eV. Under the visible-light irradiation, because the holes in the VB of WO₃ (3.83 eV) locate at the potential positions lower than those of OH-/•OH couple (2.70 eV)^[24], which could oxidize OH^- or H₂O to form •OH and initiate the occur of the photocatalysis reaction.

As depicted in Fig. 3(a, b), the WO₃-Cu-1.0 hybrid exhibited the same morphology with pure WO₃. The EDX-spectrum (Fig. 3(c)) was also performed to analyze the compositions of the WO₃-Cu-1.0 hybrid. It illustrates that the sample is composed of W, O and Cu, which is in good agreement with the XRD result. The structure of the WO₃-Cu-1.0 hybrid was further confirmed by TEM images. As revealed in Fig. 3(d, e), the Cu particle was deposited on the surface of WO₃ nanocube in the hybrid. Fig. 3(e) displays the HRTEM image of an individual WO₃ nanocube decorated with Cu particle. The clear lattice fringes of 0.37 and 0.38 nm, which correspond to the (020) and (002) crystal planes of WO₃, respectively. On the surface of WO₃ nanocube, the lattice d-spacing of 0.21 nm assigned well with the (111) crystal plane of Cu. In addition, as shown in Fig. 3(g, i), Cu, O, W elements were found in terms of the selected area. This result also indicated that the hybrids of Cu particles deposited on WO₃ nanocubes were successfully obtained.

Fig. 4(a) shows the variation in Congo Red (CR) concentration (C/C₀) versus irradiation time upon different samples under simulated sun light irradiation. It was found that the WO₃-Cu hybrid displayed superior photocatalytic performance than pure WO₃, while the Congo Red cannot be removed by direct photolysis process. Furthermore, the WO₃-Cu-1.0 exhibited the best performance among the hybrids, and it could remove 85% Congo Red from the solution. Fig. 4(b) shows a pseudo- first-order model, ln(C₀/C_t)=kt (where k is the pseudo- first-order constant, t is the reaction time). To understand the reaction kinetics of the CR photodegradation, the k value was listed in Fig. 4(c). It was observed that the WO₃-Cu-1.0 had the highest k, indicating its good photocatalytic activity. In addition, Fig. 4(d) indicates that the WO₃-Cu-1.0 also owns good cycling photodegradation performance.

Fig. 5(a) shows the influence of the addition of the scavengers, including methanol, isopropanol, CCl₄, and N₂, for h⁺, •OH, electron, and •O₂^- removal, respectively, on the related photocatalytic property. It was found that the CR photodegradation decreased with involving of methanol, isopropanol, or N₂, while the introduction of CCl₄ had no impact on the photocatalysis. It was proposed that the generated h⁺, •OH and •O₂^- were the predominant active species during the photocatalysis process. To exclude dye sensitization process, the colorless organic pollutants such as ciprofloxacin (CIP) antibiotic was selected for evaluating the photocatalytic activity of the WO₃-Cu-1.0 hybrid. As shown in Fig. 5(b), it is hard to remove the ciprofloxacin antibiotic through direct photolysis or in the presence of pure WO₃ product. However, when the WO₃-Cu-1.0 was used as the photocatalyst, about 75% of the ciprofloxacin was removed in 4.5 h. As a matter of fact, the Congo Red cannot be degraded under the light irradiation for 4 h. This result indicated that photocatalysis occurred during the Congo Red removal process. Fig. 5(c) shows the ESR spectra of DMPO•OH signal of WO₃-Cu-1.0 composite. It can be clearly observed the obvious characteristic peak of •OH for the WO₃-Cu-1.0 composite upon light irradiation, which further confirms the reactive oxygen species during the Congo Red photodegradation.

The separation efficiency of the electron-hole pairs on pure WO₃ and WO₃-Cu-1.0 hybrid was further studied by photocurrent and EIS experiments. As shown in Fig. 6(a, b), compared to WO₃ nanocube, the WO₃-Cu-1.0 hybrid had smaller arc radius and superior photocurrent. It reflected that the WO₃-Cu-1.0 hybrid exhibited rapid interfacial charge transfer and electron-hole separation efficiency, which contributed to its higher photocatalytic capability^{[25,26,27,28,29]}. On the basis of band gap structure of WO₃ and the scavengers trapping experiment result, a possible pathway for the photocatalytic degradation of CR with WO₃ photocatalyst was proposed as follows:

As shown in Fig. 6(c), under visible light illumination, the WO₃ was excited directly, and electrons (e_C^-_B) and holes (h_V⁺_B) were produced upon the WO₃. Subsequently, e_C^-_B reacted with the O₂ molecules on the surface to yield reactive oxygen radicals (•O₂^-), while h_V⁺_B was directly oxidized to generate •OH. Finally, the CR dye was degraded by the reactive oxygen radicals and/or h_V⁺_B^[30,31,32]. By anchoring Cu nanoparticles on the surface WO₃ of nanocube, the electrons moved to the Cu rapidly and improved the interfacial charge transfer. In other words, the photoinduced electron-hole pairs was separated effectively, and it would produce more reactive species to participate in the photodegradation process and thus promote efficiency for CR degradation over the WO₃-Cu-1.0 product.

In summary, cube-like WO₃-Cu hybrid was successfully fabricated by a facile room temperature method. The involving of Cu particle did not tailor the structure of WO₃ nanocubes, but had impact on its photocatalytic activity. Among the WO₃-Cu hybrid, WO₃-Cu-1.0 showed the highest efficiency towards Congo Red photodegradation. During the photocatalysis process, the generated holes and the •OH were the main active species. Based on photocurrent and EIS measurement, it was concluded that the enhancement of photocatalytic capability was mainly attributable to the higher charge transfer and lower electron- hole recombination of the WO₃-Cu hybrid.