Photocatalysis is regarded as the most promising technique for solving the problems of energy shortages and environmental pollution, which threaten the future development of human society. Photocatalytic technology has bright prospects in areas such as the generation of clean hydrogen by water splitting, CO2 reduction into useful organics, and organic synthesis under mild conditions. Furthermore, the photocatalytic decomposition of organic and/or inorganic pollutants into CO2, H2O and nontoxic products without secondary pollutants has been considered a potential method of environmental remediation [1-6].
Since Fujishima and co-workers reported TiO2 as a photocatalyst for H2 and O2 generation from water splitting [7], much attention has been paid to TiO2 and TiO2-based photocatalysts [8, 9]. However, with a wide band gap of 3.2 eV, TiO2 can only be excited under ultraviolet irradiation and it uses less than 4% of solar energy. Much effort has thus been made to improve the visible light absorption ability of TiO2; such efforts include noble metal loading, element doping and combining TiO2 with semiconductors having narrow band gaps. Meanwhile, achievements have been made in the development of novel photocatalysts driven by visible light. Among novel photocatalysts, bismuth-containing photocatalysts have received tremendous attention, and their advancement has been summarized systematically by Yu and co-workers [10]. They presented the types of bismuth-containing photocatalysts and discussed the effects of the preparation method, structure, morphology, and composition on photocatalytic properties. Despite much effort, a practically viable photocatalyst that is cheap and possesses sufficient efficiency and stability is yet to be found. Indeed, it is difficult to simultaneously fulfill the different metrics with a single material.
The recombination of photogenerated electrons and holes goes against the photocatalytic process. Since Li and coworkers first reported the high separation efficiency of charge carriers over heterojunction photocatalysts [11, 12], there has been much similar research [13-15]. It is generally regarded that the drawbacks of single-component photocatalysts can be eliminated through the synthesis of a heterogeneous photocatalyst of integrated functional components that combines the advantages of the individual components. In this article, we present a wide range of heterojunctions fabricated to enhance the performance of bismuth-containing photocatalysts. The heterojunctions are noble metal/Bi-containing composite, semiconductor/Bi-containing composite, heterojunctions between two Bi-containing photocatalysts, carbon-based material/ Bi- containing composite, and Bi metal/Bi-containing semiconductor. Finally, we briefly discuss the future development and potential applications of this class of photocatalysts.
With excellent conductivity and strong electron trapping ability, noble metals are used in photocatalysis for the separation and transfer of light-excited charge carriers. Under visible light, the transfer of charge carriers between the noble metal and support (semiconductor) can follow two pathways. In one case, the support of a wide band gap cannot be excited under visible-light irradiation but there is a transfer of electrons from the surface plasmon of the noble metal. In the other case, which is normally observed, there is a transfer of excited electrons from the semiconductor to noble metal. Furthermore, the noble metal of nanostructures can induce surface plasmon resonance (SPR), which can broaden the light absorption range of a composite photocatalyst [16]. In recent years, noble metals have been widely used as cocatalysts to enhance photocatalytic activity, especially in the areas of water splitting and the selective oxidation of organics.
As reported by Manna et al. [17], because of lattice similarity between Au and Bi2S3, a junction forms between the (101) plane of Au and (221) plane of Bi2S3. With the formation of heterojunctions, the light absorption is red-shifted and the SPR peak of Au broadens from ca. 520 to 560 nm (Fig. 1(a) and (b)). When photocatalytic reduction of methylene blue (MB) under visible-light irradiation (500-nm monochromatic light) is used as a probe reaction, the transfer of photogenerated electrons observed over the Au-Bi2S3 heterojunctions is much faster than that over Au or Bi2S3. The Au-Bi2S3 photocatalyst is typically nanostructured, and the improvement in photocatalytic activity can be attributed to the enhanced separation of charge carriers and the broadened light absorption range as a result of the noble metal plasmon (Fig. 1(c) and (d)). Owing to the fast electron transfer from the semiconductor to noble metal, Pt-coupled Bi2WO6 composite shows highly selective and efficient conversion of aromatic alcohols to their corresponding aldehydes under simulated sunlight and ambient conditions. It has been pointed out that, rather than there being a mechanism of direct oxidation involving ·O2− in the conduction band, the valance band holes are the main active sites for the two-step oxidation of alcohols, and the transfer of hot electrons to Pt can prevent the recombination of charge carriers efficiently [18].
Besides Au, other noble metals such as Pt, Pd, Rh and Ag have been used to modify Bi-containing semiconductors to fabricate plasmonic photocatalysts [19-23]. Among them, Ag is relatively cheap and has hence received more attention. For example, Luo et al. [19] prepared Ag/Bi3TaO7 composite that had enhanced activity towards the degradation of tetracycline under visible light. They also ascribed the enhanced activity to the SPR effect of Ag and the enhanced separation of charge carriers at heterojunctions. Photoreduction is the most commonly used method for the fabrication of Ag-modified photocatalysts. However, the method is complicated by the two separated processes. Taking Ag/BiVO4 as an example, steps for the synthesis of Ag/BiVO4 composites involve a sol-gel procedure followed by a photoreduction process, and the as-prepared samples are not porous [24, 25]. Our group [26] fabricated hollow and porous Ag/BiVO4 composites employing a one-step hydrothermal method (Fig. 2(a) and (b)). Compared with those prepared by photoreduction, the composites prepared employing the one-step method show higher activity towards dye degradation. Besides the SPR effect of Ag and separation enhancement of charge carriers at the heterojunctions, it is deduced that the hollow and porous structure plays an important role in activity improvement as a result of enhanced light absorption efficiency (Fig. 2(c) and (d)).
Through the fabrication of Schottky junctions, the light absorption ability and the separation of charge carriers can be enhanced. However, owing to low interaction between the noble metal and semiconductor and the agglomeration of noble metals in the photocatalytic process, the reusability of plasmonic photocatalysts is unsatisfactory. In view of this and the high cost of noble metals, there is a need to fabricate heterojunctions of other kinds. In other words, the generation of heterojunctions between two semiconductors and/or between a semiconductor and another kind of material is of great importance.
Li and co-workers [11, 12] reported that the formation of surface junctions between anatase- and rutile-phase TiO2 as well as the formation of surface junctions between MoS2 and CdS can greatly enhance the photocatalytic activity by accelerating the separation of photogenerated electrons and holes. Much attention has since been paid to constructing composite photocatalysts that are rich in heterojunctions.
Owing to their strong visible-light absorption ability and relatively high oxidation ability, bismuth-containing photocatalysts have been widely combined with other semiconductors (such as TiO2 [27-33], ZnO [34], Ag3PO4 [35-37], CuS [38], WO3 [39], γ‐Fe2O3 [40, 41], CeO2 [42], CaFe2O4 [43] and Cu2O [44-47]) to form heterojunctions. In these composites, the bismuth-containing photocatalysts not only play an important role in preventing the recombination of charge carriers but also act as visible-light sensitizers. Taking TiO2 as an example, Naik and co-workers [28] reported that the Bi2O3 sensitization of TiO2-xNx led to an excellent light harvesting capability and good charge transfer, resulting in enhanced visible-light activity for the selective hydroxylation of phenol to hydroquinone. Enhanced performance was reported by Jiang et al. [30] (Bi2O3-TiO2), Lu et al. [31] (Bi2S3-TiO2) and Li et al. [29] (Bi2MoO6-TiO2) when a Bi-containing semiconductor was loaded on TiO2. Besides the formation of heterojunctions between Bi2O3 and TiO2, bidirectional doping can also enhance photocatalytic activity as reported by Li et al. [27]. They found that the light absorption edge of TiO2-NTs could be extended to 495 nm by modification with Bi species. Additionally, the valance band of Bi2O3 was lowered from 2.02 to 2.28 eV through interaction with the Ti species. Consequently, the combination of β-Bi2O3 with stake-structured TiO2-NTs resulted in much higher activity towards carbamazepine degradation under visible-light irradiation.
The photocatalytic activity of bismuth-containing photocatalysts can also be enhanced by modification with other semiconductors. Ag3PO4, which has strong oxidation power, can achieve quantum efficiency up to 90% for O2 generation through water splitting under visible-light irradiation. As reported by Xu et al. [37], the monodispersion of Ag3PO4 on the surface of spherical Bi2MoO6 resulted in enhanced photocatalytic activity in the decolorization of MB and Rhodamine B (Rh-B). They ascribed the phenomena to the extension of the light absorption region and the efficient separation of charge carriers at the Ag3PO4/Bi2MoO6 interfaces. Recently, Zhang and co-workers [36] found that if there is a partial reduction of Ag+ to Ag0 at the surface of Ag3PO4/Bi2MoO6, then there is the formation of Ag/Ag3PO4/Bi2MoO6 having even higher activity than Ag3PO4/Bi2MoO6 owing to the strong electron trapping property of Ag. Similar enhancement was also observed over the Ag3PO4/BiOBr composite [36].
Recently, Li and co-workers pointed out that the selective deposition of co-catalysts on certain facets of polyhedron semiconductors could optimize photocatalytic activities [48]. They found that the selective loading of reduction and oxidation co-catalysts on (010) and (110) facets, respectively, produced composites that have much higher activity in both photocatalytic and photoelectrocatalytic water oxidation. Therefore, the combination of morphology engineering and heterojunction fabrication could be a practical approach for designing composite photocatalysts of high activity. Gong and co-workers [35] reported the selective deposition of Ag3PO4 on the (040) facets of BiVO4 and the as-resulted composite showed enhanced activity for dye degradation under visible light (Fig. 3). Most recently, Li et al. [49] fabricated a Z-scheme photocatalyst by decorating the (040) facet of BiVO4 with Ag@AgCl, which showed much higher activity than Ag/BiVO4 and BiVO4.
Photocatalytic processes are commonly divided into three steps: (1) light harvesting and generation of charge carriers; (2) charge carrier separation and transfer to the surface of catalysts; and (3) redox reaction on the surface of catalysts. To some extent, the efficient utilization of surface electrons and holes is also critical for photocatalytic efficiency. In our research [47], we fabricated the Cu2O/BiVO4 composite and used it as photocatalyst for the simultaneous oxidation of organic dyes and reduction of Cr(VI) under visible light (Fig. 4(a) and (b)). With the matching of the band-gap energy between Cu2O and BiOV4, there is an efficient separation of charge carriers in the catalyst (Fig. 4(c)). Therefore, the photocatalytic oxidation of organics on the valance band of Cu2O and the photocatalytic reduction of Cr(VI) on the conduction band of BiVO4 occur concurrently. This is the first report on the simultaneous removal of organic and inorganic pollutants from waste water under a neutral condition.
Different bismuth-containing semiconductors have similar layered structures, which makes the preparation of heterojunctions between two bismuth-containing photocatalysts easy. Many composite photocatalysts of this kind have been reported in recent years; e.g., Bi2O3-BiVO4 [50-52], Bi2O3-Bi2WO6 [53], Bi2O3/Bi2SiO5 [54], BiOBr/Bi24O31Br10 [55], BiOI/BiOBr and BiOCl/BiOI [56, 57], BiOCl/BiVO4 [58], BiOI/Bi2O2CO3 [59, 60], Bi2O3-Bi2S3 [61, 62], BiVO4/Bi2O2CO3 [63], and Bi2O2CO3/ Bi2MoO6 [64], BiOI/Bi2WO6 [65], β-Bi2O3-BiOI [66], BiOI/Bi2MoO6 [67] and BiVO4/Bi2WO6 [68].
Like the phase junction between anatase and rutile TiO2 reported by Li et al. [12], α-β phase heterojunctions on Bi2O3 nanowires were synthesized employing an in situ hydrothermal method [69, 70]. Owing to the difference between their valance bands and conduction bands, photogenerated charge carriers can transfer along the phase junction and thus be separated efficiently. Since it is commonly regarded that the structure and exposure of active facets are influential in photocatalytic processes, the construction of composite photocatalysts with novel structures has become a hot topic. As a typical p-type semiconductor, Bi2O3 is widely used to form p-n heterojunctions with n-type semiconductors. Our group [52], by using L-lysine as the structure directing agent and by controlling the Bi to V molar ratio of the precursor, prepared peanut-like BiVO4 and p-n Bi2O3-BiVO4 that are hollow and porous. Although the band-gap energies of Bi2O3 and BiVO4 cannot enhance the charge transfer directly, the Fermi level of Bi2O3 is close to the valance band while that of BiVO4 is close to the conduction band. As illustrated in Fig. 5, the Fermi levels have similar values when p-type Bi2O3 and n-type BiVO4 come into contact with each other. That is to say, with the formation of p-n junctions, the energy bands of Bi2O3 shift upward along the Fermi level while those of BiVO4 shift downward. Along the p-n junctions and promoted by the inner electric field, photogenerated electrons and holes can be separated efficiently and the composite thus shows enhanced photocatalytic activity for dye degradation under visible light. A novel photocatalyst of a one- dimensional Bi2O3 nanorod-Bi2WO6 nanosheet with p-n junctions was prepared by Peng et al. [53] employing a three-step method. Owing to its novel structure and the light absorption efficiency, the composite has high activity towards dye degradation and water splitting. It was reckoned that the formation of p-n junctions enhanced the separation of charge carriers, and the one-dimensional structure favors the transport of charges.
Among the reported bismuth-containing photocatalysts, Bi2S3 (1.3-1.7 eV) and BiOI (1.7-1.9 eV) have relatively narrow band gaps. They have potential for use in visible-light-driven photocatalysis but the fast recombination of photogenerated electrons and holes is an inherent problem. Nevertheless, they are good visible-light sensitizer and can be used as a component for the formation of heterojunctions. Bi2O2CO3 is a semiconductor with a wide band gap of about 3.55 eV and has good photocatalytic performance [71]. We combined BiOI with Bi2O2CO3 to form p-n heterojunctions [59]. As expected, the light absorption edge was red-shifted from 360 to 460 nm by introducing a trace amount of BiOI onto flower-like Bi2O2CO3. It is noted that the composite shows excellent activity and stability for the degradation of both a single dye and a mixture of dyes. Later, Cao et al. [60] prepared BiOI/Bi2O2CO3 composites with p-n junctions employing an in situ chemical etching method. Their results further confirmed that introducing BiOI onto Bi2O2CO3 results in broadening of the light absorption region and separation enhancement of charge carriers. Besides Bi2O2CO3, BiOI was also used to combine with visible-light- driven semiconductors such as Bi2WO6. As reported by Xiang et al. [65], the formation of p-n junctions between BiOI and Bi2WO6 not only greatly enhances the visible-light absorption ability but also improves the separation of charge carriers. Compared with pure Bi2WO6, pure BiOI, and other Bi2WO6/BiOI composites, the composite has improved photocatalytic performance under visible-light irradiation. Almost all (99.99%) of Pseudomonas Aeruginosa (P. aeruginosa), Escherichia Coli (E. coli) and Staphylococcus Aureus (S. aureus) were killed and the degradation efficiency of MB reached 100% within 60 min in the presence of 30%Bi2WO6/BiOI. Using Bi2S3 instead of BiOI, the as-prepared Bi2S3 -Bi2O2CO3 composite shows performance similar to that of BiOI/Bi2O2CO3 [72].
Considering that Bi2O3 and Bi2S3 are similar in structure but incommensurable in solubility (Ksp: 4 × 10−31 and 1 × 10−97, respectively), we developed an etching and regrowth process for the preparation of hollow and branched Bi2O3-Bi2S3 composite [62]. The two-step hydrothermal preparation method involves no organic solvent and/or structure directing agent (Fig. 6). First, Bi2O3 microtubes are synthesized employing a simple hydrothermal method and Bi(NO3)3·5H2O as the precursor, and the microtubes are then treated in a S-containing solution to achieve Bi2O3-Bi2S3. By varying the kind of S sources, the structure of Bi2S3 on the surface of Bi2O3 changes but is always single-crystalline in nature. Because of the sensitization of Bi2S3 and the branched structure, there is a broadening of the light absorption region and enhancement of light absorption over the composites. Therefore, the visible-light-excited photocurrent response of the composite is much higher than that of the components. Consequently, there is efficient separation of charge carriers at the heterojunctions and high photocatalytic activity for dye degradation. The separation efficiency of charge carriers in the composite is higher than that in Bi2S3 or Bi2O3 as confirmed by room-temperature photoluminescence analysis.
Besides binary heterostructured composites, there are ternary heterostructured composites [73-76]. For example, Zou et al. [73] prepared ternary heterostructured TiO2-Bi2MoO6- Bi3.64Mo0.36O6.55 employing a one-step solvothermal method. The photocatalytic activities of all the individual components and the binary composites were also investigated. As expected, the ternary composite performed much better than the others. The authors ascribed this to the efficient separation of charge carriers in the ternary composites, and proposed a possible process for the transfer of charges.
For both binary and ternary heterojunctions, charge separation along the junctions is enhanced and the transfer direction of electrons and holes locked owing to the presence of the Schottky barrier and the inner electric field. According to the basic principle of photocatalysis, if the photogenerated charges can be transported and separated as soon as they are generated, there is enhanced photocatalytic activity.
The use of graphene or reduced graphene oxide is a hot topic owing to their unique properties such as an extremely high specific surface area, high conductivity, super electron mobility and good electron trapping ability. Most recently, these carbon materials were employed as a structure-directing agent, photogenerated electron transfer channel, light absorption sensitizer and reactant adsorbent in photocatalysis [77]. Tu et al. [78] synthesized BiOBr-graphene composite employing an in situ method and graphene as the precursor. After combination with graphene, there was 40-nm red shift on the light absorption edge, an approximately 20% increase in dye (RhB) adsorption (compared with that of BiOBr), and a 2-fold increase in the degradation rate (according to the degradation rate constant). Gupta et al. [79] found that the modification of Bi2Ti2O7 with a proper amount of graphene not only enhances the photocatalytic activity by enhancing light absorption and charge carrier separation but also allows the formation of a structure of Bi2Ti2O7/graphene having high stability. Works of this kind on the preparation of Bi2Fe4O9/RGO [80-82], BiVO4-RGO [83], Bi2O3-RGO [84], BiPO4-GO [85], RGO-BiPO4 [86] and BiOI-GO [87] were reported in recent years. However, for all these composites, the RGO or graphene amount has to be optimized with respect to the amount of Bi-containing semiconductor. Too high a graphene amount results in the active sites of photocatalysts being covered, reducing the photocatalytic activity.
Because of their unique electrical and optical properties, polymers (especially conductive polymers) are widely combined with semiconductors for the construction of efficient photocatalysts [88, 89]. Taking polypyrrole (PPy) as an example, the polymer has high absorption coefficients in the visible-light region, high mobility of charge carriers and excellent stability [90, 91]. Most importantly, unlike graphene and other carbon-based materials, PPy is also an efficient electron donor and good hole transporter upon visible-light irradiation. Recently, Zhang et al. [92] reported a PPy/Bi2WO6 composite that has much higher photocatalytic activity for phenol degradation under simulated sunlight irradiation (Fig. 7(a)). The high photocatalytic activity could be explained by the good contact of PPy with Bi2WO6 that facilitates the separation of charge carriers, promoting the transfer of holes to the PPy/solution interface at a rate much higher than that for the self-oxidation of Bi2WO6 (Fig. 7(b)).
Since the first report on the use of graphite-like carbon nitride (g-C3N4, Eg ≈ 2.7 eV) for hydrogen production from water splitting under visible-light irradiation [93], the use of C3N4 in photocatalysis has attracted great attention. g-C3N4 materials can be synthesized from a simple precursor via a series of polycondensation reactions without any metal involvement. Unlike other metal-containing photocatalysts, g-C3N4 is a sustainable and environmentally benign organic semiconductor. It is thermally and photochemically stable owing to the high degree of polymerization and π-conjugated structure [94-98]. However, g-C3N4 by itself suffers from the high recombination of photogenerated electrons and holes. To improve its photocatalytic activity, g-C3N4 has to be combined with a semiconductor for the construction of heterostructures. In recent years, there have been many reports on the combination of g-C3N4 with bismuth-containing photocatalysts; e.g., g-C3N4/BiVO4 [99], g-C3N4-BiOCl [100], g-C3N4-Bi2WO6[101], g-C3N4-BiOBr [102], g-C3N4/BiFeO3 [103], g-C3N4-Bi2O2CO3 [104] and β-Bi2O3- g-C3N4 [105]. All of these composites have higher activity than their components. Among them, BiVO4 decorated with g-C3N4 nano-islands (5-10 nm) has enhanced activity towards MB degradation and the authors ascribed the good performance to factors such as the full exposure of reaction sites, separation of redox reaction sites, and high separation rate of charge carriers at the heterojunctions [99]. Later by first-principle calculation, Zhang et al. [106] found that g-C3N4/BiVO4 composites, especially those with heterojunctions between the (001) facet of g-C3N4 and the (010) facet of BiVO4, showed enhanced visible-light-driven activity.
Our group has fabricated heterojunctions between g-C3N4 and Bi2O2CO3 and used the composite for the degradation of organic dyes under visible light [104]. We found that the synergism of heterojunctions and dye sensitization results in enhanced catalytic activity (Fig. 8). Interestingly, when the g-C3N4-Bi2O2CO3 composite was calcined at 400 ℃, the β-Bi2O3-g-C3N4 composite formed [105]. It is noted that the calcination of Bi2O2CO3 at 400 ℃ in the absence of g-C3N4 does not result in the generation of pure-phase β-Bi2O3, indicating the unique role of g-C3N4 in directing the phase structure of Bi2O3 during calcination. This is the first report that g-C3N4 can act as a phase regulating agent. Furthermore, the β-Bi2O3- g-C3N4 composite shows high photocatalytic activity even for the degradation of 2, 4-dichlorophenol, which is known to have no visible-light absorption ability.
It should be noted that the combination of semiconductors affects the band-gap energy (i.e., redox abilities) of the composite, but the band gaps of the individual components do not changed. That is to say, the natural physical property of a semiconductor is unaffected by the combination of semiconductors.
In recent years, there has been much work on the doping of semiconductors using metals or nonmetals. The doping of heteroatoms broadens the light absorption ability and enhances structure stability during the photocatalytic process.
Many reports revealed that the doping of metal ions into bismuth-containing photocatalysts resulted in enhanced activity [107-113]. Furthermore, the doping of metal ions also stabilizes the composites. For example, Wang and co-workers [114] reported that the doping of titanium ions into β-Bi2O3 could prevent the conversion to Bi2O2CO3 in the photocatalytic process. Colon and co-workers [115] then reported that, with the doping of Er3+ into the lattice of BiVO4 even in a trace amount, the fabrication of heterojunctions between monoclinic (m-BiVO4) and tetragonal (t-BiVO4) BiVO4 could be achieved in a controlled manner. As the amount of Er doping increased from 0 to 10 at%, there was a gradual transfer of m-BiVO4 to t-BiVO4. At 0.75 at% Er doping, the as-prepared Er-m-t-BiVO4 showed the highest photocatalytic activity for MB degradation and O2 evolution from water splitting (Fig. 9) [115, 116]. Additionally, Zhou et al. [117] reported that the doping of Er3+ into Bi2MoO6 affects the final morphology of Bi2MoO6 and greatly enhances the activity for dye and phenol degradation.
Besides doping with a single kind of ion, doping with more than one kind of ion has been studied. The synergistic effect of doping with two or more kinds of ions was also investigated. For example, Nussbaum et al. [118] reported the co-doping of BiOCl with Fe and Nb ions. It was observed that even at low contents of Fe and Nb doping, the composite showed much higher activity than the components and the composites doped with a single kind of ion. A similar phenomenon was reported by Zhou et al. [119], who used W and Mo ions to co-dope BiVO4.
Nonmetal (most commonly carbon) doping is usually employed to narrow the band gap of semiconductors and thus improve the visible-light response, just as ever reported to enhance the visible-light absorption of TiO2 by C doping [120]. Doping Bi-containing photocatalysts with carbon also produces such an effect. Yin et al. [121] fabricated C-doped BiOV4 using a one-step method. They reported that due to the presence of C in the lattice of BiVO4, there was high activity towards dye degradation and water splitting for O2 evolution under visible-light irradiation. The main reason for this was the extension of the light absorption edge from 480 to 700 nm. Homoplastically, the carbon doping of Bi2O3 and BiVO4 can greatly enhance visible-light-driven activity [122, 123].
Bismuth is a semimetal that has a very small band overlap and possesses an interesting SPR property [124]. Given the high price of noble metals, inexpensive Bi with its unique properties can be used as an alternative in the SPR photocatalytic process. The synthesis of Bi-modified photocatalysts has aroused intense interest [125-127]. Most recently, Dong and co-workers [128] used Bi metal as a co-catalyst on the surface of Bi2O2CO3 microspheres and the composites showed enhanced activity for NO removal under visible-light irradiation at different intensities. They demonstrated that the enhancement of photocatalytic activity of Bi2O2CO3 could be ascribed to the SPR effect of Bi and to the fact that Bi metal on the Bi2O2CO3 surface can prolong the lifetime of charge carriers (Fig. 10). The phenomenon is in agreement with the SPR effect of a noble metal. Unfortunately, the stability of the as-prepared samples was not investigated. Bi metal is commonly known to be unstable under atmospheric conditions because it readily oxidizes into Bi2O3.
With unique physicochemical properties, bismuth- containing photocatalysts are endowed with excellent visible-light responses. Unfortunately, the photocatalytic activity is still unsatisfactory owing to the recombination of photogenerated charge carriers. To improve photocatalytic activity, strategies such as loading with a noble metal, the generation of heterojunctions through combination with a semiconductor, modification with carbon-based materials, doping with heteroatoms and modification with Bi metal have been adopted. Although significant progress has been made, further efforts are required to further advance the use of this class of photocatalysts (especially those with heterojunctions). Three aspects are of special relevance.
(1) Throughout the literature, the application of bismuth-containing photocatalysts with heterojunctions mainly focused on the degradation of organic pollutants, and only a few studies focused on water splitting and CO2 reduction. The generation of Z-scheme heterojunctions between bismuth-containing photocatalysts and semiconductors of more negative CB could be promising for the preparation of efficient visible-light-driven composites for the generation of H2 from water splitting and fuel production from CO2 reduction.
(2) Owing to the positive VB edges, bismuth-containing photocatalysts commonly show relatively high activity for oxidation reactions. In our previous study, Bi2WO6 showed high activity and selectivity for the oxidation of saturated alpha-carbon (i.e., C-H bonds) in aromatic alkanes at room temperature under solvent-free conditions [129]. After adjusting the band-gap energy of Bi2WO6 through combination with semiconductors, the as-prepared composites could be used for the selective oxidation of organics to produce fine chemicals. Photocatalytic organic synthesis conducted under mild conditions has a great potential owing to the green energy and mild conditions involved.
(3) The design and fabrication of nonstoichiometric bismuth-containing photocatalysts is a potential method of adjusting the band-gap structures and constructing defects as well as oxygen vacancies. Importantly, it is envisaged that, besides the enrichment of active sites, the presence of defects and oxygen vacancies on the surface of a catalyst is essential for the enhanced separation of charge carriers. Additionally, the fabrication of nonstoichiometric bismuth-containing materials could be a feasible way of further enhancing the visible-light-driven activity of this class of photocatalysts.
2016,
Vol. 37 Issue (6): 780-791 
PDF (21840 KB)
2016,
Vol. 37 Issue (6): 780-791 PDF (21840 KB)