Introduction

With the rapid development of modern society and the industrialization process, water environmental pollution has become the focus of attention. It is imperative to find a suitable and effective treatment method. In the past few decades, the application of photocatalysis has attracted attention because it can be widely used in many fields, especially in the environment and energy fields [1, 2]. Photocatalysis is an effective treatment method for degradation of persistent organic trace pollutants because the photocatalyst is stimulated by effective light to generate photo-generated electrons and holes, thereby causing O2 and H2O to generate active oxygen species (ROS, such as ·OH, ·O2−/·HO2, H2O2,1O2) to degrade the pollutants [3]. Part of the research on reactive oxygen species can also inactivate some microorganisms.

Among many photocatalysts, TiO2 has been the most widely studied due to its non-toxicity, availability, and low price [4]. Since the water was first demonstrated to decompose into H2 on TiO2 photoanode [5], it is one of the most widely used photocatalysts at present due to the overall superior properties of TiO2, including availability, long-term stability, and non-toxicity [6]. However, TiO2 only can respond to about 4% of solar ultraviolet radiation with large band-gap energy (~ 3.2 eV) [7]. Also, the fast electron–hole recombination inherent in conventional TiO2 photocatalysts is an important factor affecting its low photocatalytic efficiency.

Recently, tungsten trioxide (WO3) has attracted attention due to its strong ability to degrade organic pollutants, high stability, long life, non-toxicity, low cost, and suitable band edges. Since it was first reported in 1976, extensive research has been conducted on the photocatalytic performance of WO3. Especially in the past ten years, a great deal of research has focused on WO3 and improving WO3 photocatalytic performance. In Fig. 1a, the rising trend indicates that the application of WO3 photocatalyst in the field of photocatalysis is increasing.

Figure 1
figure 1

a The Number of Journal Citation Reports (JCR) articles per year, as reported by Web of Science (http://apps.webofknowledge.com) from 2006 and updated to December 2019 retrieved via the keywords “WO3” and “photocatalysis” in the topic of papers. b The mechanism of WO3 and strategies for enhancing the photocatalytic activity of WO3 photocatalysts in wastewater

Full size image

Over the past several decades, photocatalysis has been the best procedure for wastewater treatment because of the ability of this method to perfectly mineralize the contaminants. In 1976, Butler et al. already reported that n-type tungsten trioxides were a great photocatalyst for water oxidation [8]. Because of the narrow band-gap (2.6–2.8 eV), nontoxicity, and strong adaptability of WO3, it has considered as a photocatalyst that can effectively degrade pollutants [9]. The valence band (VB) of WO3 was about 3.1 eV [10], which made WO3 have a strong oxidizing property. The mechanism of WO3 is shown in Fig. 1b.

The large specific surface area of WO3 nanostructure can increase the effective surface area of photocatalytic reaction. However, there are also materials with larger surface areas showing lower photocatalytic activity. For example, some scholars have pointed out that 500 nm WO3 nanoparticles obtained at 800 °C can induce more O2 precipitation than 30 nm nanoparticles obtained at 500 °C [11], which is attributed to the fact that larger particles have better crystallinity (resulting in smaller Eg), and thus counteracting the effect of small specific surface area [12]. In addition to the above effects, the range of photo-response values and the recombination speed of photogenerated carriers will affect the WO3 photocatalytic performance. Therefore, some strategies which can change the WO3 photo-response value and photogenerated carriers recombination efficiency are adopted to enhance the photocatalytic activity of WO3. So far, strategies for improving the photocatalytic performance of WO3 for contaminants, i.e., surface modification and control of morphology size, synthetic heterojunction, element doping, are being reported. We need a review to summarize the improvement of photocatalysts based on WO3 in wastewater treatment. In this review, the photocatalytic mechanism of WO3 and three aspects to enhance the photocatalytic activity of WO3 in water pollution are introduced. Finally, the research status and application challenges of WO3-based photocatalysts are briefly summarized.

Advantages and limitations of pure WO3 in photocatalysis

As a photocatalytic material, it is impossible to oxidize H2/H2O (relative to NHE (common hydrogen electrode)) and reduce H2O/O2 due to the positions of the conduction band (CB) and valence band (VB) of the WO3 semiconductor. These allow WO3 to effectively degrade many organic compounds, such as textile dyes and antibiotics [12,13,14], which can also inactivate some microorganisms [15]. Besides, WO3 has significant stability in acidic environments and is an excellent material for treating organic acid contaminated water [16]. And some advantages are mentioned in the previous section, such as high stability, long life, non-toxicity, low cost, suitable band edges, and so on.

The photocatalyst absorbs energy from optical radiation to generate hole and electron pairs. Then, holes and electrons directly react with pollutants or generate free radicals and degrade pollutants through oxidation and reduction reactions. Therefore, the position of the energy level between VB and CB and the band-gap of WO3 play a decisive role in the ability to oxidize and reduce pollutants. Figure 2 shows the photocatalytic mechanism of the WO3 semiconductor. The reactions are as follows:

$${\text{Photocatalyst }} + {\text{ hv }} \to {\text{ e}}^{ - } + {\text{ h}}^{ + }$$
(1)
$${\text{H}}_{{2}} {\text{O}} \to {\text{H}}^{ + } + {\text{OH}}^{ - }$$
(2)
$${\text{e}}^{ - } + {\text{O}}_{{2}} \to \cdot {\text{O}}_{{2}}^{ - }$$
(3)
$${\text{h}}^{ + } + {\text{OH}}^{ - } \to \cdot {\text{OH}}$$
(4)
$$\cdot {\text{OH}}/{\text{h}}^{ + } + {\text{R }}\left( {{\text{pollutant}}} \right) \to {\text{R}} * \left( {{\text{oxidized}}} \right)$$
(5)
$$\cdot {\text{O}}_{{2}}^{ - } + {\text{R}} \to {\text{O}}_{{2}} + {\text{R}} * * \left( {{\text{reduced}}} \right)$$
(6)
$${\text{R}} * \left( {{\text{oxidized}}} \right){\text{ or R}} * * \left( {{\text{reduced}}} \right) \to \left( {\text{final products}} \right) \to {\text{CO}}_{{2}} + {\text{ H}}_{{2}} {\text{O}}$$
(7)
Figure 2
figure 2

Mechanism of WO3 degradation of pollutants

Full size image

The free radicals generated are the key to the degradation of organic pollutants. However, four important effects limit the photocatalytic performance of tungsten trioxides for wastewater treatment: first of all, low visible light response ability (λ < 470 nm); secondly, low specific surface area; thirdly, fast recombination rate of photogenerated electron–hole pairs [17]; fourthly, many reactions in photocatalysts cannot happen due to the low CB position (low than −0.1 V vs NHE, PH = 7), for example single-electron reduction of O2 (O2 + e− → O2−(aq), −0.33 V) and reduction of H+ to H2 (2H+  + 2e− → H2, −0.41 V) [18]. Therefore, for the WO3 semiconductor without any improvement, its photocatalytic activity is relatively low. In the recent years, many strategies have been proposed, which can improve the activity of WO3 photocatalyst. Therefore, in Sect. 3, we will elaborate on the following points, such as surface modification and morphological size control, the formation of heterojunctions and the modification of other elements.

Strategies for improving photocatalytic activity of WO3 for wastewater treatment

Surface modification and control of morphology size of WO3 for wastewater treatment

Surface modification and morphology control of photocatalysts are generally regarded as effective strategies to enhance the activity of photocatalysts because the crystal surface and morphology can be changed. Besides, the shape of the material will affect the photocatalytic activity of WO3 nanostructures to a certain extent [19]. Therefore, surface modification and morphology control of WO3 nanostructures are very important for efficient photocatalytic degradation of water pollutants.

WO3 one-dimensional structures and two-dimensional structures

One-dimensional (1-D) semiconductor structures can provide a direct path for photo-generated charges transfer, have small grain boundaries and thus own excellent charge transport properties. Since the scattering of free electrons is suppressed, the photocatalytic activity of one-dimensional nanomaterials can be improved compared to nanoparticles [20]. In the recent years, various one-dimensional WO3 nanostructures have been developed, such as nanorods [21,22,23], nanofibers [24], nanotubes [25], and nanowires [26]. Table 1 shows the effects of various WO3 one-dimensional structures and two-dimensional shapes on wastewater treatment. For example, nanofibers have excellent porosity, excellent mechanical and ideal chemistry property, which can improve the photocatalytic activity of WO3. It has been reported that WO3 nanofibers were prepared by electro-spinning, which degraded methylene blue twice as efficiently as WO3 particles [24]. Compared with nanofibers, nanotubes can provide higher surface area and more effective sites to degrade pollutants, so that pollutant molecules can be diffused quickly and effectively in nanotube structures. For example, Zhao et al. successfully synthesized WO3 nanotubes through template-free WCl6 urea-assisted alcoholysis and the SEM images are shown in Fig. 3a–b [27]. These WO3 NTs are monodisperse, with a diameter of about 300–1000 nm and a length of about 2–20 mm. They are composed of a single WO3 nanoparticle linear arrangement, and many self-supporting pores are formed due to incomplete aggregation of the nanoparticles. Experiment results show that the BET value of these WO3 NTs (25 m2g−1) is increased by 5.7 times, compared with the BET value of commercial WO3 particles (4.4 m2g−1), and that WO3 NTs can generate electron–holes in visible light with a wavelength greater than 400 nm. Compared with commercial WO3 particles, the prepared nanotubes have higher RhB degradation efficiency and better photocatalytic performance. The relationship between the initial decomposition rate of the two materials and the Pt loading concentration is shown in Fig. 3c. Because the tubular structure has larger effective surface area, higher charge carrier mobility, and wider light response range, the degradation activity of the nanotube is enhanced. Different shapes of the one-dimensional WO3 structure have their own advantages, and can be selected according to the purpose in order to degrade the pollutants efficiently and cheaply.

Table 1 Modification for enhancing the photocatalytic activity by using various shapes WO3 photocatalyst
Full size table
Table 2 Modification for enhancing the photocatalytic activity by using WO3-based photocatalyst
Full size table
Figure 3
figure 3

a and b SEM images with WO3 NTs. c Plots of initial decomposition rates versus Pt-loading concentration (adapted with permission from reference [27]. Copyright (2008) Wiley–VCH). d and e SEM images with different magnification of 2D-WO3. f Change of MO normalization concentration versus the exposure time under irradiation with the different photocatalysts (adapted with permission from reference [33], Copyright (2019) Elsevier)

Full size image

Unlike one-dimensional materials, two-dimensional materials are nano-flaky materials with flat surfaces and high aspect ratios, with an extremely small thickness and strong adhesion to substrates. Nanosheets [28], nanoplates [29] and films [30, 31] have been developed. Research has shown that by using WO3 nanosheet as the adsorbent, the saturated adsorption amount can reach MB 600 mg/g, higher than the normally activated carbon powder [32]. Due to the high adsorption capacity, the photocatalytic performance of the nanosheets is improved. Compared with ordinary two-dimensional nanosheets, ultrathin nanosheets have the advantages of greater specific surface area and richer active sites, which make WO3 ultrathin nanosheets present better catalytic performance. For example, Liang et al. synthesized a two-dimensional (2D) ultrathin WO3 nanosheet dominated by {002} crystal plan through a simple surfactant-induced self-assembly method and the SEM images of 2D-WO3 are shown in Fig. 3d–e [33]. The SEM diagram shows the ultra-thin nanosheet structure with a lateral dimension of hundreds of nanometers and a thickness of about 4.9 nm. The experimental results show that the degradation rate constant of MO by 2D-WO3 is 6.5 times higher than that of WO3 nanoparticles, showing strong degradation activity (Fig. 3f). The improvement in properties can be owned to the 2D-WO3 with unique structure, such as high reactivity {002} crystal surface percentage, high specific surface area, wide photo-response range, and high photogenerated electron–hole separation rate. It can be seen that WO3 two-dimensional materials show superiority in degrading pollutants.

Based on the synthesized one-dimensional and two-dimensional WO3 nanomaterials, the degradation efficiency of the photocatalyst can be further improved by other methods. Here, we only give a brief example by adding oxygen vacancies on the surface of the materials. For example, Wang and his colleagues synthesized uniformly distributed oxygen vacancies on the surface of WO3 nanorods through the hydrothermal method [21]. The performance of WO3 nanorods photocatalyst was demonstrated by comparing with the efficiency of alcohol oxidation to the corresponding ketone. The experimental results show that the activity of WO3 nanorods prepared by this method is greatly improved, through the adsorption of alcohol molecules with higher specific surface area, the rapid transfer of photogenerated electrons with smaller crystal size, and the surface oxygen vacancies as traps to capture photoelectrons, thus reducing the recombination of photoelectrons and holes. The author believes that the combined effect of the above three points improves the photocatalytic performance of WO3. Also, Wu et al. successfully synthesized WO3 ultra-thin nanosheets with oxygen vacancies on the surface [34]. The improvement of the material in photocatalysis is similar to the nanorods appealed. Some studies call the addition of oxygen vacancies as self-doped, which means the dope of oxygen vacancies on the surface of the material. For example, Wang and his colleagues used electrochemical methods to add oxygen holes to the surface of WO3/TiO2 heterojunction to improve the degradation of exhaust gas [35].

Compared with pure WO3, the photocatalytic activity of different shapes of WO3 has been improved to some extent, but the comparison between different shapes is not clear. In 2014, Farhadian et al. have prepared and characterized one-dimensional WO3 nanostructures and two-dimensional WO3 nanosheets, i.e., nanorods, nanosphere, and nanoplates, to study the photocatalytic performance of the shape on the degradation of RhB dye, as displayed in Fig. 4 [19]. In this experiment, the authors found that tetrahedral and cubic nanostructures (nanorods and nanoplates) had higher catalytic activity than spherical nanostructures (nanosphere) because they had more atoms at the edges and corners and these atoms showed higher catalytic activity. Therefore, nanoplates and nanorods had a stronger adsorption capacity than nanosphere. However, the band-gap of WO3 nanorods in this experiment was the same as the band-gap of TiO2, so it has a very low light response range in visible light. The results show that nanoplates have the highest degradation performance among the three shapes. Similarly, R. Narayanan et al. studied the effect of different shaped materials on photocatalytic activity in 2004, which was consistent with this conclusion. So we can see that different shapes of WO3 materials can affect the band-gap, specific surface area, and adsorption capacity. Through this experiment, it can be seen that different shapes of WO3 materials can also affect the activity of atoms.

Figure 4
figure 4

Schematic illustration of the shape effect on the photocatalytic activity of the WO3 nanostructures (adapted with permission from reference [19], Copyright (2015) Elsevier)

Full size image

Special morphology

Other different shapes can increase the photocatalytic activity of pure WO3 in water pollution treatment through measures such as increasing specific surface area, improving separation efficiency, and enhancing light response. Some scholars have reported that other shapes have been synthesized but mainly composed of WO3 nanoparticles, nanorods, and nanosheets. Hollow particles (Fig. 5a) [36] and flowers (Fig. 5b) [37] are composed of nanoparticles. The nanorods can form cylindrical stacks (Fig. 5c) [37] and a flower shape (Fig. 5d) [38]. A flower-like structure is formed by most nanosheets stacked closely together (Fig. 5e) [39], and a WO3·0.33H2O microsphere structure is formed by nanosheets stacked (Fig. 5f) [40]. The core layer is a dense structure composed of aggregated nanoparticles, and the shell layer is a layered structure composed of WO3 ultra-fine nanoplates [41]. Table 1 shows the effects of various WO3 special morphology shapes on wastewater treatment.

Figure 5
figure 5

SEM images of a WO3 hollow particles (adapted with permission from reference [36], Copyright (2008) Elsevier); b WO3 flower-like and c WO3 cylindrical stacks (adapted with permission from reference [37], adapted with permission from reference); d flower-like WO3·0.33H2O (adapted with permission from reference [38], Copyright (2014) American Chemical Society); e hierarchical WO3 structures (adapted with permission from reference [42], Copyright (2016) Elsevier) and f 3D hierarchical WO3·0.33H2O (adapted with permission from reference [40], Copyright (2017) Elsevier)

Full size image

For example, Xu and his colleagues successfully synthesized three-dimensional flower-like and wheel-like structures based on one-dimensional WO3 nanorods (Fig. 5d) [38]. Due to the high charge separation efficiency of wheel-shaped and flower-shaped WO3, the degradation activity of RhB was improved.

Because after irradiation with appropriate wavelength, the generated holes migrate to the surface along the potential slope generated by the bending of the band, and are captured by the H2O molecules adsorbed on the surface of WO3·0.33H2O, thereby generating hydroxyl radicals ·OH and reducing the photogenerated electron–hole recombination. The larger photo-response range and longer photo-carriers existing time make the material have better photocatalytic performance. Another example is the WO3 layered structure composed of WO3 nanosheets; Yao et al. synthesized WO3 nanosheets by a simple and surfactant-free hydrothermal method and combined them into WO3 layered structure (Fig. 5f) [42]. The effect of this material on the degradation of MB, ER and CR harmful organic dyes was studied under simulated sunlight. The results show that the layered WO3 nanostructures have more excellent water pollutant degradation efficiency. Besides, by changing different raw materials, different WO3 shapes can be obtained, thereby enhancing the degradation of pollutants by WO3. For example, the calcined acid-treated PbWO4 (sacrifice template) can obtain dendritic and spherical morphology, while SrWO4 can observe dumbbells [43].

In general, the larger specific surface area will have the greater the number of reaction sites. The synthesized spheres and flowers have a higher specific surface area, so the photocatalytic performance of pure WO3 is improved to a certain extent. The research and control of different morphologies show that the morphology of photocatalysts is very important in the development of its increased activity. Because these reactions are usually completed on the surface, it depends to a large extent on the morphology of the surface. Compared with general commercial WO3, the material composed of tiny single crystals has a larger surface area, higher light transmittance, and more active sites, and thus has higher photocatalytic activity. This provides an idea for the preparation of WO3 structures with stronger photocatalytic degradation ability in the future.

WO3 high-surface-energy facets

In some studies, the engineering of crystal planes of semiconductors has become an important strategy for improving the performance of photocatalysts by fine-tuning the properties of materials. The arrangement and coordination of surface atoms essentially determine the adsorption, desorption, and carrier transfer efficiency of the pollutants from the surface to the adsorbed molecules. As a result, the atomic structure on the surface of the photocatalyst has a great effect on the degradation of pollutants. The atomic arrangement and coordination on the catalyst surface change with the crystal plane changing in different directions [44]. From this, we can conclude that the crystal has a great influence on the photocatalyst. In the recent years, some scholars have synthesized WO3 materials with different crystal planes and studied their properties to compare the different crystals planes on the photocatalytic performance of WO3.

For example, Yang et al. synthesized three different WO3 materials and studied the degradation performance of sewage under ozone, namely monoclinic WO3 mainly exposed to {100}, {002} monoclinic WO3 and hexagonal WO3 mainly exposed to {100} surface, which is called M-100 (Fig. 6a–c), M-002 (Fig. 6d–f), H-100 (Fig. 6g–i) [45]. In general, a larger surface area is beneficial for photocatalytic processes. However, in this report, M-100 has a relatively low surface area, but it owns the highest catalytic degradation performance (Fig. 6j–k), which means that the high-surface-energy facets would influence the photocatalytic performance of the photocatalysts greatly. Because M-100 has the highest CB position, it is thermodynamically conducive to the electron capture of dissolved O3. Additionally, this study also shows that monoclinic crystals with the same crystal plane are more active than hexagonal crystals. For the same crystal, the {100} crystal plane can change the position of the CB, compared to the {002} crystal plane, leading to the enhancement of degradation activity. Similarly, in another report, the authors also showed that the electronic structure effect of the crystal face was caused by the different atomic structure configurations on the {002}, {020} and {200} planes, which lead to the shift of the CB position (Fig. 6l) [18]. In summary, the engineering of the crystal facet is to cause the CB edge to shift upwards and then optimize the degradation activity of WO3 on pollutants.

Figure 6
figure 6

FETEM images of M-100 (a), M-002 (d) and H-100 (g). SAED images of M-100 (b), M-002 (e) and H-100 (h). HRTEM images of M-100 (c), M-002 (f) and H-100 (i). Degradation of OA (j) and mineralization of cephalexin (k) in photocatalysis, ozonation and photocatalytic ozonation; and band structures of the prepared M-100, M-002 and H-100 (l) (adapted with permission from reference [45], Copyright (2018) Elsevier)

Full size image

Some scholars have also compared WO3 with different crystal phases [46]. Cubic WO3 (c-WO3) had strong adsorption on MB and increased with increasing temperature, reaching a maximum adsorption capacity of 35.95 mg/g. Monoclinic WO3 (m-WO3) had strong photocatalytic degradation of MB, and the degradation efficiency of MB was 100%, through the generation of photoinduced holes and hydroxyl (·OH) under 120 min of visible light irradiation. In future research, we can combine the changes of crystal plane and crystal phase to select the corresponding WO3 for different water quality to achieve different treatment requirements and purposes.

WO3-based heterojunction for wastewater treatment

In this section, we introduce the effect of changing the synthesis of heterojunctions based on WO3 on the degradation of pollutants. Some scholars believe that the photocatalytic performance of one-component photocatalyst is still affected by the high recombination rate of photogenerated electron–hole pairs [49]. Therefore, the formation of heterojunctions by combining binary or ternary semiconductors with a suitable band-gap has been considered to be an effective strategy to improve the performance of photocatalysts because they can simultaneously expand the absorption range of light and promote charge separation [50]. Here, this section describes the application of WO3 to wastewater treatment based on the synthesis of heterojunctions from different materials.

Preparing binary composites

Many studies have proved that synthetic heterojunction composites can reduce the photogenerated electron–hole recombination rate and improve the utilization of electrons and holes [51]. There are many reports on the synthesis of binary heterojunctions with WO3 to enhance the photocatalytic activity on wastewater, such as WO3/g–C3N4 [10, 50, 52, 53], WO3/TiO2 [54], WO3/BiO4 [55], WO3/grapheme [9, 13, 20, 56, 57], etc. In particular, WO3/g–C3N4 composites have been studied in detail by changing the formation of methods and conditions. WO3/g–C3N4 composite photocatalysts with different photocatalytic mechanisms have been successfully prepared, such as traditional type-II and Z-scheme heterojunctions. So far, most researches on forming heterojunction composites with WO3-based have been carried out in p–n heterojunctions, conventional heterojunctions, direct Z-scheme heterojunctions, and S-scheme heterojunctions. In addition to the above heterostructures, there are some special heterojunctions formed by the combination of graphene and semiconductors. The systematic improvement effect of the above mechanisms will be described below.

The p–n heterojunction is composed of n-type WO3 semiconductors and p-type semiconductor [58, 59]. In the p–n-type heterostructure, a WO3/BiOI heterojunction photocatalyst is taken as an example. Luo et al. proposed the photocatalytic mechanism of the WO3/BiOI heterojunction catalyst, as shown in Fig. 7 [60]. BiOI is a p-type semiconductor with Fermi level (Ef–p) near VB, while WO3 is an n-type semiconductor with Fermi level (Ef–p) near CB (Fig. 7a). After the p–n-type WO3/BiOI heterojunction is formed, electrons will be transferred from WO3 to BiOI, while holes will be transferred from BiOI to WO3. When the Fermi levels of the two reached equilibrium, the internal electric field was established at the interface due to the transfer of electrons. The internal electric field can also greatly promote the migration of photogenerated carriers and effectively reduce the recombination rate of photogenerated electron–hole pairs, thereby improving the performance of photocatalyst. As shown in Fig. 7b, the Fermi level (Ef–p) of BiOI moves upward along the interface, and the Fermi level (Ef–n) of WO3 moves upward along the interface so the migration of charge causes adjacent energy bands to occur bending. It can be found that the p–n structure greatly inhibits and slows down the recombination of photogenerated electrons and holes with the migration of charges and holes and the internal electric field, which can improve the photodegradation.

Figure 7
figure 7

a Schematic diagrams for energy bands of p-type BiOI and n-type WO3 before contact. b The formation of a p–n junction and its energy band diagram at equilibrium and transfer of photoinduced electrons from p-type BiOI to n-type WO3 under visible-light irradiation (adapted with permission from reference [60], Copyright (2015) Elsevier). c Proposed mechanism for the photodegradation of MB on WO3/g-C3N4 composites (adapted with permission from reference [52], Copyright (2015) The Royal Society of Chemistry). d Schematic diagram of Z-scheme photocatalytic mechanism of WO3 NS/g-C3N4 NS composite photocatalyst (adapted with permission from reference [10], Copyright (2019) Springer). e The work functions of g–C3N4 and WO3 before contact. f The internal electric field and band edge bending at the interface of WO3/g–C3N4 after contact. g The S-scheme charge transfer mechanism between WO3 and g–C3N4 under light irradiation (adapted with permission from reference [63], Copyright (2019) Elsevier)

Full size image

When two n-type semiconductors synthesize a heterojunction photocatalyst, and the two semiconductors have suitable energy bands. This mechanism is called the traditional type-II, where the electrons and holes generated by the semiconductor are transferred to the CB of semiconductor I and the VB of semiconductor II, and no electric field is generated inside. Currently reported WO3/g–C3N4 heterojunctions, WO3/BiVO4 heterojunctions [55], WO3/Ag3VO4 heterojunctions [59] and so on are all traditional type-II. In this type, we take WO3/g–C3N4 heterojunction photocatalyst as an example. Huang and his partners synthesized a WO3/g–C3N4 heterojunction through a simple calcination process in 2013 to degrade pollutants under visible light [52]. Figure 7c shows the mechanism of the WO3/g–C3N4 heterojunction. Since the potential of CB and VB of WO3 is higher than the potential of CB and VB of g–C3N4, the electrons generated on g–C3N4 are transferred to the CB of WO3, and the photogenerated holes on VB of WO3 are transferred to VB of g–C3N4. The formed electrons gather on the side of WO3, and the holes gather on the side of g–C3N4, which can reduce the electron–hole recombination, thereby promoting the photocatalytic degradation of MB by the material. This mechanism is similar to the p–n-type heterojunction, but no electric field is generated. Therefore, the WO3/g–C3N4 heterojunction showed higher photocatalytic degradation performance when compared to single component of pure WO3 and g–C3N4 photocatalysts.

The traditional type-II heterojunction can improve the catalytic performance of the catalyst to a certain extent, but the reducibility of photo-generated electrons and the oxidizability of photo-generated holes will be reduced as the charge transfer between the semiconductors. If the semiconductor A and B are closely combined to form a heterojunction, which generates an intermediate electric field, so the heterojunction is called a Z-scheme heterojunction. This mechanism can significantly increase the space between electrons and holes, and retain its ability to redox [61]. Deng et al. successfully synthesized nanocomposite of Z-scheme WO3 nanosheet/g–C3N4 nanosheet by calcination methods, and studied its photocatalytic performance [10]. Results showed that in the WO3 NS/g–C3N4 NS composites with 20 wt% WO3 NS present best photocatalytic performance, and the main reason for the improvement of degradation performance is the participation of ⋅O−2, ⋅OH, and h+ in the reaction. Figure 7d shows the catalytic mechanism of Z-scheme WO3 NS/g–C3N4 NS composite photocatalyst, with photogenerated electrons migrating from WO3 NS to g–C3N4 NS, and photogenerated holes migrating to WO3 NS. Therefore, a reduction reaction occurs on g–C3N4 with a higher reduction potential to generate ⋅O−2, and an oxidation reaction occurs in WO3 with a higher oxidation potential to generate ⋅OH. The production of free radicals ⋅O−2 and ⋅OH optimizes the oxidation ability and the transmission efficiency of photogenerated electrons of Z-scheme WO3 NS/g–C3N4 NS. At present, various forms of WO3 and g–C3N4 have been successfully synthesized to form Z-scheme heterojunction photocatalysts, such as WO3 nanorods/g–C3N4 nanosheets [62], WO3 nanosheets/g–C3N4 nanosheet composites [10]. These complexes differently improved the photocatalytic activity of WO3/g–C3N4 in the degradation of water pollutants.

Based on p–n heterojunctions, conventional type II heterojunctions, and Z-scheme heterojunctions, related scholars have proposed a new concept of stepped (S-scheme) heterojunctions [64, 65]. The S-scheme heterojunction photocatalyst consists of n-type oxidation photocatalysts and n-type reduction photocatalysts. After the equilibrium is reached, the Fermi levels of the two semiconductors are at the same level, while the photogenerated electrons will be transferred to the oxidation photocatalysts and holes will be transferred to the reduction photocatalysts, and an internal electric field will be generated. Unlike other heterojunctions, as the S-scheme heterojunction, the electrons of CB in the oxidized photocatalytic with relatively useless and holes of VB in the reduced photocatalyst with relatively useless will recombine under the action of the internal electric field and eliminate, thereby retaining useful electrons and holes. Some scholars believe that the charge transfer path of the S-scheme heterojunction is similar to the "step" type, which has a strong redox capacity of space separation and photo-generated charge carriers [66]. Due to its unique structure and internal electric field, the S-scheme heterojunction can generate a large number of active materials to enhance the degradation efficiency of pollutants. In 2019, Fu et al. constructed a composite photocatalyst of S-scheme 2D/2D WO3/g–C3N4 heterojunction through electrostatic self-assembly methods [63]. Figure 7e-g shows the mechanism of the S-scheme WO3/g–C3N4 heterojunction. Generally, g–C3N4 is a reduction type photocatalyst with a small work function (4.18 eV) and a higher Fermi level. In contrast, WO3 is an oxidation-type photocatalyst with a large work function (6.23 eV) and a lower Fermi level (Fig. 6e). When g–C3N4 and WO3 are in close contact until the Fermi level is the same (Fig. 6f), g–C3N4 loses electrons and becomes positively charged, while WO3 gets electrons and becomes negatively charged at the interface. As a result, an internal electric field is generated at the interface and the band edges are bent of the two semiconductors, which can make some electrons from WO3 CB combine with holes from g–C3N4 VB. But this way can prevent electrons from g–C3N4 CB from combining with holes from WO3 VB, shown in Fig. 6g. In general, this heterojunction mechanism can recombine the relatively useless electrons and holes in the two semiconductors, while the useful electrons and holes are retained. It is because of this unique transfer process of electrons and holes that the 2D/2D WO3/g–C3N4 composites have strong oxidation and reduction ability, thereby improving the photocatalytic performance of composites. The mechanism has been shown to exhibit strong photocatalytic activity against water decomposition. Unfortunately, no research has been done on S-scheme WO3-based binary composite heterojunctions for wastewater. However, this mechanism shows great potential in terms of photocatalysis, and this photocatalytic mechanism will show greater potential in water pollution treatment in the future.

Within ordinary heterojunctions, Z-scheme and S-scheme heterojunctions, there are special heterojunctions formed by the combination of graphene (GR) and semiconductors. Graphene is a two-dimensional single-layer SP2 hybrid carbon atom with excellent charge transfer performance, high thermal conductivity, high surface area and hexagonal filled structure [67]. The ultra-high electrical conductivity and low conduction band potential (-0.08 V vs. SHE, pH = 0) of graphene allow photo-generated electrons to flow from the semiconductor to its surface, thereby reducing the compound photo-generated electrons-holes. In 2010, Zhang et al. first proposed the application of graphene in photocatalysis and they proved that the addition of graphene improved the degradation efficiency of MB in the composites [68]. Based on graphene, people also synthesize GO and RGO to synthesize composite materials with semiconductors. So far, many scholars have successfully prepared composite photocatalysts composed of WO3 and graphene with different morphological structures, such as WO3 nanoparticles [69], one-dimensional nanostructures [20, 70], two-dimensional nanosheets [13, 71] and so on, and all have proved the synthesized composites materials can significantly enhance the ability to degrade water pollution. Guo et al. synthesized WO3 nanoparticles on the graphene sheets by the sonochemical method [72]. Studies have shown that the amount of O2 precipitated from the water of WO3@GR composites with 40 wt% GR inside was twice that of pure WO3. The improvement in photocatalyst performance was the result of the joint action of WO3 nanoparticles and GR sheets, through enlarging the absorption range of visible light, enhancing the electron transport and promoting the separation of photogenerated charge carriers. The mechanism diagram is shown in Fig. 8a. As we all know, the application of one-dimensional single-crystal nanomaterials in photocatalysis is very important. Compared with nanoparticles, one-dimensional materials have smaller grain boundaries, which provide the path for photo-generated charges and inhibit free electron scattering, thereby having higher photocatalytic activity. For example, in 2012, An et al. synthesized WO3 nanorods on the surface of graphene through the hydrothermal method, and the mechanism shown in Fig. 8b [20]. Studies have shown that the degradation efficiency of rhodamine B 6G (RhB 6G) by WO3/graphene composites containing 3.5wt% graphene was 2.2 times than that of pure WO3 nanorods. They believe that the interaction between dyes and negatively charged groups in graphene can result in higher adsorption capacity of RhB 6G for WO3/graphene and improve migration of photo-generated carriers are important factors to improve the photocatalytic performance of WO3/graphene.

Figure 8
figure 8

a The procedure of photocatalytic oxidation for the WO3@GR composite (adapted with permission from reference [72], Copyright (2012) The Royal Society of Chemistry). b Proposed photodegradation mechanism of RhB 6G over WO3 nanorods/graphene composites (adapted with permission from reference [20], Copyright (2012) The Royal Society of Chemistry)

Full size image

Preparing ternary composites

To date, many binary composites based on WO3 have been synthesized. However, studies have shown that ternary nanocomposites generally have wider visible light response range, lower recombination rate and higher interfacial charge transfer than binary nanocomposites [73,74,75,76]. Among the ternary composites of WO3-based, there are two kinds of semiconductors combined with electronic mediators to synthesis composite materials, and there are three kinds of semiconductors, which we will discuss separately.

In ternary Z-scheme heterojunctions, electron mediators are usually used as conductive materials to improve electron transfer, such as Ag [77,78,79], Au [80], carbon nanodots [81] and RGO [82]. Because the Fermi level of the electron mediator is between two semiconductors, the electron mediator and the two semiconductor materials form a Z-scheme photocatalytic mechanism. Here, take redox graphene (RGO) and metallic Ag nanoparticles as examples. For example, in 2018, Lu et al. successfully prepared Z-scheme WO3/RGO/g–C3N4 composite materials, in which RGO is an electron mediator [83]. The degradation mechanism is shown in Fig. 9a. RGO, as an electron mediator, changes the charge transfer pathway of the composite, which is different from the binary Z-scheme heterojunction. In the ternary Z-scheme WO3/RGO/g–C3N4 composite, the photogenerated electrons on CB of WO3 combined with the holes on VB of g–C3N4 through RGO in the interface between RGO and g–C3N4. Thus, useful electrons and holes are retained for generating the radicals ⋅O−2 and ⋅OH. Free radicals can participate in the oxidation reaction, thereby improving the performance of WO3/RGO/g–C3N4 photocatalyst. Also in 2019, Chen et al. successfully synthesized WO3 nanoplate/Ag/g–C3N4 nanosheet compound materials by solvent evaporation and in situ calcination [77]. The study showed that WO3/Ag/g–C3N4 can degrade RhB about 96.2% in 40 min, while the degradation efficiency of RhB by WO3/g–C3N4 is 58.2% under the same conditions. The synergistic effect of g–C3N4 nanosheets and WO3 nanoplates beneficial to enhance photocatalytic performance can provide conditions for the rapid transfer of photogenerated electrons and holes, and the electron mediator Ag nanoparticles are conducive to the charge transfer (Fig. 9b). Moreover, the Z-scheme heterostructure allows the composite to retain high redox capacity. Excellent photocatalytic activity, easy design and easy manufacturing are all advantages of binary Z-scheme composite photocatalyst. However, the composite materials have the disadvantages, such as low surface area, small response range of visible light, poor adsorption performance, and weak redox capacity. The electron mediator usually provides a close contact area between WO3 and other semiconductors, so that the electrons are better transferred, the photogenerated electron–hole pairs recombination is reduced, and the degradation efficiency is improved. This mechanism is similar to the S-scheme mechanism through the combination of relatively useless electrons and holes to leave useful electrons and holes, which use useful electrons and holes to generate free radicals to improve photocatalytic activity. At present, some scholars have proposed to form ternary S-scheme heterojunctions by doping electron mediators. For example, Pan et al. doped C in S-scheme 2D/2D WO3/g–C3N4 and studied the photocatalytic ability to degrade MB [64]. The degradation of MB can reach 92.4% in 60 min. When C as electron mediator, the heterojunction has high redox capacity, short charge transfer distance, and wide response range of visible light, which greatly improves photocatalytic ability.

Figure 9
figure 9

a Schematic illustration of the separation and transfer of photogenerated charges and the reactive species in the degradation process of Z-scheme photocatalyst g–C3N4/RGO/WO3 (left) and heterojunction-type photocatalyst g–C3N4/WO3 (right) (adapted with permission from reference [83], Copyright (2018) Elsevier). b Photocatalytic mechanism for WO3/Ag/CN composite under visible light irradiation (adapted with permission from reference [77], Copyright (2018) Elsevier). c Schematic diagram for the possible charge separation of Z-scheme WO3/g–C3N4/Bi2O3 (adapted with permission from reference [17], Copyright (2018) Elsevier). (d) Schematic diagram for the possible charge separation of flower-like BiOBr-WO3–Bi2WO6 ternary hybrid (adapted with permission from reference [88], Copyright (2015) Elsevier)

Full size image

The Z-scheme composites coupled with three kinds of semiconductors can produce more electrons and holes to improve the photocatalytic performance of the composites. Recently, scholars have successfully synthesized ternary Z-scheme complexes based on WO3, such as WO3/MoS2/g–C3N4 and WO3/g–C3N4/Bi2O3 84] presenting much better photocatalytic abilities than binary Z-scheme photocatalytic systems. For example, Jiang and his partners have synthesized the Z-scheme WO3/g–C3N4/Bi2O3 composite through a one-step co-calcination strategy and proved that the material has excellent photocatalytic performance [17]. The migration of electrons along the interface leads to accumulating in the CB of g–C3N4, while holes accumulate in the VB of WO3 and Bi2O3 shown in Fig. 9c. Therefore, the electrons in the CB of g–C3N4 can be captured by O2 to generate ⋅O−2, while the holes in VB of WO3 and Bi2O3 can oxidatively degrade TC or oxidize H2O to form ⋅OH radicals. Active free radicals then participated in the degradation of pollutants, which increased the degradation rate of TC by the composites.

In addition to ternary semiconductors coupled into Z-scheme composites, ternary semiconductors are coupled into cascade structures, in which electrons and holes migrate through the interface potential gradient in the ternary mixed-valence band. WO3/Bi2WO6/BiOBr, WO3/TiO2/CdS, WO3/Cu2O/BiVO4, and WO3/BiVO4/BiOCl have been reported [85,86,87]. For example, Zhu et al. successfully prepared flower-like WO3-BiOBr-Bi2WO6 ternary composites, in 2015 [88]. The experimental results showed that the composite material showed higher photocatalytic activity compared with the WO3-Bi2WO6 binary composite, and the degradation mechanism of RhB by this material is proposed (Fig. 9d). In Fig. 9d, it can be seen that the CB edge of BiOBr is located between Bi2WO6 and WO3. Therefore, the ternary composite can form a cascade structure, similar to the traditional type-II heterojunction. For the WO3-BiOBr-Bi2WO6 material, the electrons are accumulated in the CB of WO3 and the holes are accumulated in Bi2WO6 by the migration of charge carriers. So the electrons on the surface of WO3 and the hydroxyl radicals generated by the holes on the surface of Bi2WO6 can directly participate in the reaction to degrade organic pollutants. WO3-BiOBr-Bi2WO6 cascade structure has improved RhB degradation efficiency by high surface area, close interfacial contact, and differences in energy band positions. Besides, the cascade CB positioning of the ternary semiconductor will generate a built-in potential gradient, which can better promote the separation of photo-generated electrons and holes, thereby promoting electron transfer within the junction structure. This conclusion has been proved in other reports [86, 87].

From this, in the process of degradation of sewage with WO3 photocatalyst, Z-scheme ternary composites, and ternary cascade composites can enhance the degradation of organic compounds. In the recent years, many scholars have proposed that heterojunction composites have great potential in the degradation of water pollution. Therefore, the formation of Z-scheme (or S-scheme) heterojunctions can obtain the best performance by increasing the absorption range of visible light, increasing the specific surface area, promoting effective charge separation, strengthening interface contact and generating free radicals (Table 2).

Element modification of WO3 for wastewater treatment

At present, some scholars have used non-metal and metal to modify semiconductors for improving the photocatalytic degradation performance of WO3 on sewage. Because the modification of elements in photocatalyst can enhance the photocatalytic activity of photocatalysts by enhancing the separation rate of photogenerated electron–hole pairs and increasing the photo-response range of visible light. So far, Fe, Ni, Cu, Zn, Co, and other metal ion-doped WO3 composite materials and N, S, C, P, I, F and other non-metal ion-doped composite materials have been successfully synthesized. Co-doping of elements can promote the separation of photogenerated electrons and holes faster, which means that the co-doping of elements WO3 is more conducive to the improvement of photocatalytic performance. In this section, the photocatalytic degradation properties of transition metals, precious metals, rare earth metals, non-metals, and multi-element co-doped materials are studied. The results are summarized in Table 3.

Table 3 Element modification for enhancing the photocatalytic activity of WO3 photocatalyst
Full size table

Metal element doping

Co, Zn, Ni, Cu, and Fe transition metals have been studied to dope WO3 to widen the visible light response range of WO3 for improving the photocatalytic activity of WO3. Hameed and his colleagues studied the effect of Co, Zn, Ni, Cu-doped WO3 on photocatalysis [100]. Studies have shown that among transition metals, the doping of Ni has the greatest effect on the catalytic hydrogen production of WO3. When doped with 1.0% and 10.0% Ni, the photocatalytic oxygen generation efficiency of WO3 was 4 times and 19 times of the original. However, oxygen production and hydrogen production are different. By doping pure WO3 with 10% Fe, WO3 had the highest hydrogen production capacity, in which the hydrogen production was 7 times that of the original. The effect of doping different concentrations of Fe on the degradation of RhB by WO3 was also reported [101]. When Fe was doped at 5.25%, WO3 had the highest photocatalytic efficiency under visible light. About 93% of phenol was reduced in 240 min, and about 92% of RhB was degraded in 120 min by 5.25% Fe-doped WO3. When 5% Fe-doped WO3 showed the best photocatalytic performance in MB degradation, it could degrade about 95% of MB in 120 min under visible light irradiation [102]. Thereby, the transition metal-doped WO3 shows higher photocatalytic degradation ability when treating organic compounds in wastewater.

The rare earth metals in the periodic table include 17 elements, which can be used as dopants for WO3 semiconductors to degrade organic pollutants. In the recent years, some scholars have proposed that doping rare earth metals into WO3 can promote the concept of charge separation. Because the 4f orbit of the rare earth metal is not completely occupied, and the 5d orbit is empty, it can effectively capture electrons, which can effectively promote the separation of the photogenerated carrier, thereby playing an important role in doping for improving the performance of photocatalysis [103]. The effects of doping rare earth metals such as Gd 104], Dy [105], La [106], Eu [107], and Yb [108] on the photo-activity and photo-stability of WO3 have been studied. In the case of Dy-WO3, Dy3+ can provide electrons to the adsorbed oxygen and then convert it to Dy4+, thereby promoting the generation of superoxide radicals. Besides, Dy4+ can trap electrons in WO3 CB and inhibit photo-generated electron–hole recombination. WO3 may be partially consumed in aqueous solution, and Dy3+ doping will hinder this deactivation process [109]. Tahi et al. used a hydrothermal method to synthesize rare metal-modified WO3 composites [104]. Studies have shown that the doping of rare metals affects the grain size and specific surface area of the photocatalyst so that WO3 exhibits excellent photocatalytic performance during the degradation of harmful dyes. Among them, doping 2% Gd showed the most effective degradation performance of WO3, and the degradation efficiency of various pollutants could reach about 98%. Although doped rare earth metals are very expensive, they have great potential for improving the WO3 degradation of pollutants.

Non-metal element doping

The above-mentioned that the performance of the catalyst can be enhanced through doping metals, but some researchers have found that metal-doped semiconductors are not heat-resistant and may cause photoelectron-hole pair recombination [110]. Therefore, it is proposed that doping non-metals can also improve the wide band-gap and promote the separation of photogenerated carriers [111]. Some researchers have demonstrated that doping non-metals can enhance the performance of WO3 photocatalyst [110, 112,113,114]. For example, Chen et al. synthesized S-doped WO3 samples by the hydrothermal method [110]. Experimental results showed that compared with undoped WO3, S-WO3 samples had better photo-degradability, and the maximum MB removal efficiency of 5% S-WO3 samples was 78.7%. It is attributed to the lower bandgap energy, more oxygen vacancies in the surface lattice and the heterojunction formed by WS2 and WO3. Because the CB position of WS2 was higher than the CB position of WO3, which caused the generated electrons to be injected into the CB of WO3, resulting in effective charge separation. Therefore, non-metal-doped WO3 shows higher photocatalytic degradation ability than undoped WO3 when treating organic compounds in wastewater treatment.

Elements co-doping

In addition to single-element doping, multi-element co-doping has also been used to increase the efficiency of WO3 photocatalysts [115]. Multi-element co-doped WO3 composites have been successfully synthesized and proved to have good photocatalytic properties, such as Zn-Cu co-doped WO3 [116], Nb-F co-doped WO3 [117], and so on. Here, we take I-P co-doped WO3 as an example. Tijani et al. prepared the I-P elements co-doped WO3 nanoparticles and studied the photocatalytic performance of the material to degrade the local wastewater contaminated with dyes [118]. The found that spherical WO3 nanoparticles can be completely transformed into rods and bamboo bundles with different doping percent of I and P elements, and the relative images are shown in Fig. 10. For I-P co-doped WO3, it can be observed that the relative morphology can be transferred from spherical symmetry into rod-like structure by adjusting the doping content of different I and P element. The reasons may refer to the fact that the nature of the dopant vis-a-vis atomic weight and ionic size can make great influence on the morphology of the target materials. Compared with the undoped WO3 with spherical and cubic structure, both I− and P+3 dopants can be treated as the structure directing agents for the formation of a less compacted rod and hexagonal nanostructures. The detail morphology adjustment mechanism can be explained as follows: the formation of the nanorods liked WO3 with 10% iodine or phosphorus mainly originated from the side-by-side alignment, which caused by the high lateral capillary forces; the formation of bamboo-like or nanorods bundles liked I-P co-doped WO3 refers to the oriented attachment, which caused by the reduction of the surface energy due to the synergetic effect between I and P element [118]. As for the photocatalytic performance, I-P co-doped WO3 nanocomposites can degrade 93.4% TOC and 95.14% COD, and show the highest photocatalytic activity compared with single-doped and undoped WO3. I− and P+3 occupied the oxygen vacancies in WO3 nanoparticles, but through the synergy between the two dopants, the crystal size was reduced and the surface area was increased. As a result, some I and P diffused on the surface of WO3, which may cause surface defects, thereby improving the degradation of printing and dyeing wastewater. Also, local internal electric field determined by both I and P can make the rapid separation of photogenerated carriers to improve the degradation efficiency. In general, this material had many advantages, such as higher specific surface area, smaller band-gap energy, good crystallinity, wider visible light response range, and lower photo-generated electron–hole recombination rate. It can be obtained that the presence of co-dopants can further improve the photocatalytic degradation ability of WO3 doped with a single element on wastewater.

Figure 10
figure 10

HR-SEM images for a undoped WO3, b 2% I-doped WO3, c 2% P-doped WO3 and d 2% P and I co-doped WO3 (adapted with permission from reference [118], Copyright (2019) Elsevier)

Full size image

Deposition of noble metal

At present, the research on the deposit of WO3, such as Au, Pt, Ag, Pd and so on, shows that the degradation performance of WO3 can be improved by improving charge transfer, increasing electron traps and reducing band-gap energy. It has been reported that noble metal-deposited WO3 nanoparticles improve photocatalytic activity by adjusting the Fermi level balance between noble metal and WO3 photocatalyst to reduce band-gap energy [119], and suppress electron–hole recombination [120]. As early as 2010, scholars reported that the deposition of Pt can improve the performance of WO3 photocatalyst [121]. The effect of WO3 with different shapes of Pt deposited on ethylene under visible light was also studied. Studies have shown that the order of photoactivity is Pt/nanocubes > Pt/nanoparticles > Pt/nanobundles [122]. WO3 nanocube had the best photo-degradability due to its unique geometry. The presence of Pt deposits improved the photo-activity of nanoparticles and nanocubes, which was attributed to the ability of Pt deposits to promote the multi-electron reduction of O2. Also, some scholars have compared the photocatalytic performance of Pt loaded with different shapes of WO3 [123]. According to the photocatalytic evaluation results of Pt-loaded samples, the sequence of the most active sample was not significantly different from that of the unloaded sample. This meant that the morphological structure of WO3 on the photocatalytic degradation ability was greater than that of the supported co-catalyst. At the same time, the size of noble metal particles also affects the performance of WO3. The deposition of Au nanoparticles was not conducive to improving the activity of WO3 to degrade pollutants. Because the size of Au nanoparticles was too large, and most of the surface of WO3 was covered by gold particles, which prevented the incident light from reaching the surface and made this part of the catalyst in an inactive state during the reaction [124]. The effect of different concentrations of deposited Ag on the degradation efficiency of sulfanilamide (SAM) was also studied [125]. Within a certain range, the degradation efficiency increased with the increase in Ag concentration. Ag nanoparticles as electron capture centers during WO3 degradation SAM process can improve the separation of photogenerated electrons. Moreover, WO3/Ag composites could also deactivate Escherichia Coli and Bacillus subtilis under visible light. The antibacterial effect can be attributed to synergistic effect among Ag, Ag+, and antibacterial of WO3/Ag composite. Therefore, the photocatalytic degradation ability of noble metal-deposited WO3 is higher than that of WO3 without noble metal deposition when treating organic compounds in wastewater. And it has great potential in treating wastewater contaminated by pathogens.

Conclusions and future prospective

Many studies have shown that WO3 is a promising photocatalyst for water pollution treatment by responding to visible light because of its highly adjustable performance and excellent performance in removing persistent organic micro-pollutants and some microorganisms in complex water performance. Through a variety of improvement measures, scholars have synthesized WO3-based materials with large specific surface area and charge separation ability, which have good photocatalytic performance, economic feasibility, sustainability and durability. We have described and compared the different forms of WO3, synthetic binary or ternary heterojunctions, and other elements doped in different forms of WO3. By modifying WO3, the photocatalytic performance of WO3 is improved to a certain extent. Studies have also shown that although doping a certain element promotes the photocatalytic performance of WO3, the morphological structure of WO3 has a greater effect on degradation wastewater ability than the supported cocatalyst. Therefore, there is still much space for improving the performance of WO3 photocatalyst.

  1. (1)

    Surface modification and morphology control have been demonstrated that can improve the photocatalytic activity of WO3 photocatalysts. But more simple and efficient method with low cost is necessary to be developed for its realistic employment.

  2. (2)

    The combination of WO3 with another semiconductor for the formation of heterojunction is a promising way to promote the photocatalytic performance. However, whether binary or ternary, or more complex composites of WO3-based photocatalysts, the deep micro-scale photocatalytic mechanism analysis is still a challenge and the integration of experiment and computational could be a good entry point.

  3. (3)

    Many publications have been focused on element doping and modification to enhance the photocatalytic properties of WO3. In most cases, noble elements or metals have been utilized, which increases the cost of materials and hinders its practical application. So highly efficient cheap metal elements or non-metallic elements modification method are urgently needed.

  4. (4)

    Surface defect or vacancy is a very novel and effective methods in the study of other photocatalysts. However, about WO3, relative studies are scarce, so some interests should be focused on this field of WO3, which would be very efficient in further improving the photocatalytic activity of WO3-based materials.