Photocatalytic NO Removal by Ternary Composites Bi₁₂GeO₂₀/BiOCl/W₁₈O₄₉ Using a Waste Reutilization Strategy

Abstract

Heterojunction creation is demonstrated as an effective strategy to enhance the transfer and separation of charge carriers, which is beneficial for subsequent photocatalytic reactions. In this study, “sea urchin-like” W₁₈O₄₉ was in situ-grown on the surface of Bi₁₂GeO₂₀ through a hydrothermal process, and the released Cl⁻ anions tended to produce BiOCl simultaneously. Systematical characterizations confirmed the construction of ternary composites Bi₁₂GeO₂₀/BiOCl/W₁₈O₄₉ (GBW), in which Type I and Z-scheme models were integrated to promote charge carrier migration and separation by combining the structural merits of both models. Under UV–visible light, the catalytic performance of the as-synthesized samples was tested in terms of NO oxidation removal. Compared to pure Bi₁₂GeO₂₀, the composite GBW5 showed the highest NO photocatalytic removal efficiency of 42%, which was nearly four times that of pure Bi₁₂GeO₂₀. These improvements were mainly due to enhanced light absorption, suitable morphological features, effective separation of charge carriers, and the boosted generation of reactive species in the GBW series. This study paves the way for the construction of Bi₁₂GeO₂₀-based ternary composites using a comprehensive utilization of waste method and the employment of the composites for the photocatalytic removal of low concentrations of NO at the ppb level.

1. Introduction

Rapid industrialization and technological advancement are often accompanied by environmental pollution and energy consumption. After being emitted into the atmosphere, nitrogen oxides (NO_x, with NO as the major component) not only cause profound ecological damage but also pose threats to human health, leading to disasters such as acid rain and photochemical smog [1,2,3], as well as respiratory diseases upon inhalation [4,5]. NO at high concentrations readily oxidizes to the more toxic NO₂ in the atmosphere [1,6,7] and can be treated by conventional technologies [8,9,10]. However, these treatments suffer from high energy consumption, complicated equipment, and secondary pollution generation, and are, therefore, cost-ineffective for treating NO_x with low concentration and flow rate. In this context, photocatalysis emerges as a green, energy-efficient [11,12,13], and sustainable technology of research significance [14,15,16]. Under ambient conditions, it harnesses solar energy to generate reactive charge carriers and abundant radicals to eliminate toxic compounds [17]. Specifically, the oxidation of NO molecules occurs on the surface of catalysts to form ${\mathrm{N}\mathrm{O}}_2^{-}$ or ${\mathrm{N}\mathrm{O}}_3^{-}$ species that are converted to metabolic forms and further adsorbed by crops, which facilitates waste utilization and circulation [18,19,20]. Therefore, the fabrication of an efficient photocatalytic system to remove NO is an important challenge [21,22]. In particular, the preparation of catalysts via waste reuse strategies generally provides more comprehensive support for NOx removal research [23,24].

The semiconductor Bi₁₂GeO₂₀ exhibits unique optical and physicochemical properties, offering great potential for applications in photonic excitation, holographic imaging, optoelectronics, piezoelectrics, spatial light modulation, and photocatalysis [25,26,27]. However, its wide bandgap, slow charge kinetics, severe recombination of charge carriers, and poor visible light absorption greatly constrain its photocatalytic activity. Some modification strategies have been explored to enhance its photocatalytic capacity, such as morphological modulation [28], ion doping [29], carbonaceous material decoration [30], noble metal deposition [31], and heterojunction construction [32]. However, to the best of our knowledge, there are few reports on the modification of Bi₁₂GeO₂₀ for photocatalytic NO removal, suggesting considerable opportunities for further research in this area. The incorporation of another component with a narrow bandgap favors the absorption of visible light and redistribution of charge carriers in different phases, thus facilitating the enhancement of photocatalytic performance. In this context, tungsten oxide W₁₈O₄₉ is considered a fascinating candidate, possessing abundant oxygen vacancies and impressive absorption characteristics in the visible and near-infrared regions. In addition, the one-dimensional structure of W₁₈O₄₉ nanowires facilitates efficient long-distance electron transfer and sustained photocatalytic stability [33]. Moreover, W₁₈O₄₉ is able to generate high-energy hot electrons for N₂ reduction through the localized surface plasmon resonance effect [34]. Therefore, W₁₈O₄₉ with favorable physiochemical and optical merits has typically been selected to establish heterojunction composites for efficient NO removal.

During composite preparation, the dissociation of reagents provides anions and cations, which may produce different components under suitable conditions. Such processes fully utilize all species in reagents and avoid introducing external ions, thus providing an opportunity to generate heterojunction composites in a green and cost-effective manner. For this purpose, the reagent WCl₆ was adopted to produce W₁₈O₄₉, and the released Cl⁻ anions react with bismuth cations in Bi₁₂GeO₂₀ lattices to generate BiOCl in situ. The obtained BiOCl, composed of {Bi₂O₂}²⁺ and Cl⁻ layers arranged in a cross pattern, possesses a unique structure that expands internal space and generates an intrinsic electric field. The combined effects of internal polarization due to intrinsic magnetism and the electronegativity differences among Bi, O, and Cl facilitate the effective separation of charge carriers, thereby enhancing photocatalytic activity under ultraviolet irradiation [35,36]. As a result, these components tend to form heterojunction structures in ternary composites in situ, by which charge carriers can be further regulated and separated and photocatalytic performance may be accordingly enhanced. Moreover, the combination of Type I and Z-scheme models facilitates the migration and separation of charge carriers with relatively strong redox potentials [37,38,39], which may produce more active species to efficiently photocatalytically remove NO at low concentrations.

In this study, a series of ternary composites Bi₁₂GeO₂₀/BiOCl/W₁₈O₄₉ were successfully fabricated using a hydrothermal method and comprehensively characterized by using various analytical techniques. Under UV–visible light irradiation, these GBW composites were able to photocatalytically remove NO at the ppb level. The enhanced photocatalytic performance was mainly related to suitable morphologies, enhanced light absorption, strengthened separation of charge carriers, and boosted generation of reactive species. According to experimental and analytical results, the photocatalysis mechanism was proposed to be the integration of Type I and Z-scheme models in obtained composites with sufficient structural stability.

2. Results and Discussion

2.1. Microstructural and Morphological Characterization

The phase composition and crystal structures of Bi₁₂GeO₂₀, W₁₈O₄₉, BiOCl, and GBW series were analyzed by X-ray diffraction (XRD), as shown in Figure 1A. The diffraction signals at 2θ = 21.3°, 24.7°, 27.7°, 30.4°, 32.9°, 35.3°, 37.5°, 39.6°, 41.7°, 43.6°, 45.5°, 49.1°, 52.5°, 54.1°, 55.7°, and 61.9° in the XRD pattern of Bi₁₂GeO₂₀ correspond well to the (211), (220), (310), (222), (321), (400), (411), (024), (332), (422), (134), (125), (035), (600), (532), and (631) crystal planes of tetragonal Bi₁₂GeO₂₀ (Joint Committee on Powder Diffraction Standards, JCPDS File NO. 77-0861) [27], respectively. The XRD pattern of W₁₈O₄₉ is in good accordance with the monoclinic phase (JCPDS File NO. 05–0392) [33]. However, diffraction peaks of W₁₈O₄₉ are unobservable in GBW composites, mainly due to their trace amount and uniform distribution. However, some extra peaks emerged in the composite GBW3 and became enhanced with an increase in WCl₆ addition. By comparison, these extra peaks are derived from the tetragonal matlockite BiOCl phase (JCPDS File NO. 06-0249) [35]. For further identification, NaCl was adopted as the Cl source to replace WCl₆ to produce a series of binary composites GB that clearly includes diffraction peaks from both components Bi₁₂GeO₂₀ and BiOCl, as shown in Figure S2. Therefore, during the synthesis of W₁₈O₄₉, the component BiOCl was simultaneously produced by combining Bi³⁺ cations on the surface of Bi₁₂GeO₂₀ and Cl⁻ anions from WCl₆, fully utilizing the different species contained. The other different structural parameters were investigated by using XRD patterns, which included lattice parameters, c/a ratio, density ρ (g/cm³), and unit cell volume ν (ų) [40,41]. For Bi₁₂GeO₂₀: a = 10.153, b = 10.153, c = 10.153, α = 90.00, β = 90.00, c/a = 1.000, volume ν = 1046.6, density ρ = 9.201. For W₁₈O₄₉: a = 18.318, b = 3.783, c = 14.028, α = 90.00, β = 115.21, c/a = 0.766, volume ν = 879.5, and density ρ = 7.727.

The chemical composition and surface valence states of Bi₁₂GeO₂₀ and composite GBW5 were analyzed using X-ray photoelectron spectroscopy (XPS). The full-scan spectrum in Figure 1B confirms the presence of Bi, Ge, and O in both samples. In addition, the spectrum of GBW5 shows a distinct signal for W, with an atomic percentage of tungsten measured at 5.3%, which is consistent with the theoretical value of 5%. This confirms the successful incorporation of tungsten into the composites. To further distinguish the valence states of each element, high-resolution (HR) XPS spectra of Bi 4f, Ge 3d, O 1s, W 4f, and Cl 2p are exhibited in Figure 1C–F and Figure S3. In Figure 1C, the peaks at 159.5 and 164.8 eV in Bi₁₂GeO₂₀ and 159.4 and 164.7 eV in composite GBW5, respectively, correspond to the Bi 4f_7/2 and Bi 4f_5/2 orbitals [42]. The Bi 5d and Ge 3d signals at 25.7 and 28.8 eV in Bi₁₂GeO₂₀ can be seen in Figure 1D, and slightly shift to 25.6 eV and 28.7 eV in composite GBW5, confirming the presence of Bi³⁺ and Ge⁴⁺ cations in both samples. As to the O 1s spectra in Figure 1E, signals positioned at 529.3, 530.3 eV, and 531.5 eV in Bi₁₂GeO₂₀ respectively attribute to Ge–O bonds, Bi–O bonds, and adsorbed surface oxygen species. In addition, Figure 1F displays characteristic signals at 35.9 and 37.9 eV that correspond to the W 4f_7/2 and W 4f_5/2 orbitals, respectively. Figure S3 shows the characteristic signals at 198.2 and 199.7 eV, corresponding to the Cl 2p_3/2 and Cl 2p_1/2 orbitals, respectively. Compared with Bi₁₂GeO₂₀, the signal peaks of Bi 4f and Ge 3d in the GBW5 composite have shifted downfield, indicating a certain degree of electron transfer and migration to Bi₁₂GeO₂₀ [43], which further benefits the photocatalytic processes [44].

The microstructures and morphologies of Bi₁₂GeO₂₀, W₁₈O₄₉, and composite GBW5 were thoroughly examined using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The tetragonal-phased Bi₁₂GeO₂₀ exhibits an accumulation of regular large and incomplete small triangular pyramids of several micrometers in Figure 2A and Figure S4A. As shown in Figure 2B and Figure S4B, the W₁₈O₄₉ presents sea urchin-like spheres that are accumulated by numerous nanosheets. The composite GBW5 shows the morphology of nanosheets from sea urchin-like spheres covering the surface of triangular pyramids in Figure 2C and Figure S4C, which reveals the coexistence of both Bi₁₂GeO₂₀ and W₁₈O₄₉ components and facilitates reflection of incident light and mass transfer of reactants and products, thus improving photocatalytic outcomes. Unfortunately, it is difficult to see any BiOCl, which generally presents as lamellar structures, similar to that of W₁₈O₄₉. To this end, the elemental distribution of composite GBW5 was measured at the selected area in Figure S4D and the presence of Bi, Ge, O, W, and Cl is obviously confirmed in Figure S4E–I, corroborating the successful generation of ternary heterojunction composites. Furthermore, the SEM image of composite GB5 in Figure 2D and Figure S4J shows triangular pyramids with a rough surface and some small flakes adhered to them, in contrast to that of Bi₁₂GeO₂₀. The produced BiOCl phase exists as small flakes or particles closely attached to the surface of the Bi₁₂GeO₂₀, which can be identified from the elemental mapping images shown in Figure S4K–O. TEM and HRTEM images were employed to further investigate the morphologies and microstructures of relevant samples. The TEM image of composite GBW5 in Figure 2E reveals a sea urchin-like morphology including a bulky structure covered with numerous nanosheets, which is in good accordance with the SEM image in Figure 2C. In addition, some small particles attached to the surface among the nanosheets are observed. This area is framed by a red elliptical ring and was the target for high-resolution TEM analysis. Significantly, three kinds of well-defined lattice fringes are observable in Figure 2F. Lattice fringes in the bulky phase with a lattice spacing of approximately 0.27 nm correspond to the (321) crystal planes of Bi₁₂GeO₂₀. In addition, lattice fringes on nanosheets with lattice spacing of approximately 0.37 nm belong to the (010) crystal planes of W₁₈O₄₉, and other lattice fringes with lattice spacing of 0.32 nm belong to (110) crystal planes of BiOCl. This close contact of three kinds of components indicates the formation of the target composites, which can steer and redistribute charge carriers to promote photocatalytic reactions [45].

Ultraviolet–visible diffuse reflectance spectroscopy (UV–Vis DRS) was utilized to assess the optical properties and bandgap structures of the semiconductors. BiOCl in white shows strong adsorption in the ultraviolet region because of the wide bandgap shown in Figure 3A. Pristine Bi₁₂GeO₂₀ in light yellow exhibits an intensive absorption ranging from 200 to 500 nm, indicating a moderate response in the visible region. However, W₁₈O₄₉ in navy blue demonstrates a significant absorption in both the ultraviolet and visible regions, mainly owing to the suitable band structure with the localized surface plasmon resonance effect. As the W₁₈O₄₉ content of the composites increases, the absorption edge of the GBW composite progressively shifts toward shorter wavelengths, which is primarily attributed to the formation of heterojunction structures. From Figure 3A, the adsorption in the near-infrared region of composites is unable to induce variations in the photocatalytic performance, mainly due to insufficient incident energy. Instead, other effects, such as the dosage of the introduced catalyst [46], greatly affect photocatalytic outcomes. Figure S5 shows the UV–Vis DRS spectra of composites synthesized using NaCl as the chlorine source. Significantly, the absorption edge of these composites remains almost unchanged.

According to the calculations in Figure S6A–C, it was found that Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl follow an indirect transition, with the value n as 4. The bandgap (E_g) of semiconductor-based samples can be computed from the equation αhν = A(hν − E_g)^n/2, where α, A, h, and ν represent the absorption coefficient, a constant, Planck’s constant, and optical frequency, respectively. As shown in Figure 3B,C, the values of E_g of Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl are determined to be 2.70, 2.76, and 3.26 eV, respectively. In Figure S7, the (ahv)^1/2 vs hv plots for the GBW series of samples reveal that the values of E_g gradually increase with increasing W₁₈O₄₉ content. Accordingly, the value of E_g of composite GBW5 is measured at 2.79 eV. It is well known that E_g refers to the energy required to excite electrons from the valence band (VB) to the conduction band (CB) in a semiconductor. The reduction in E_g in composite GBW5 facilitates visible light absorption, the generation of charge carriers, and further photocatalytic reactions.

The band structures of Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl were inferred through flat-band potentials (E_fb) on an electrochemical workstation. The Mott–Schottky plots of Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl are shown in Figure 3D–F, respectively. Usually, the type of semiconductor can be determined by analyzing the slope of the straight-line section, with a negative or positive slope indicating p-type or n-type characteristics, respectively. Clearly, the positive slopes in the linear regions in three samples suggest an n-type nature, and the slopes of Bi₁₂GeO₂₀ and BiOCl are smaller than that of W₁₈O₄₉, revealing the large charge carrier concentration and better photocatalytic performance [47]. The intersections with the x-axis and the straight-line section can allow values of E_fb for Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl to be approximated as −0.38, −0.04, and −0.51 V (vs. saturated calomel electrode, SCE), respectively. By using the correction of 0.24 V, E_fb values are calculated as −0.14, 0.20, and −0.27 V (vs. normal hydrogen electrode, NHE) for Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl, respectively. According to the characteristics of n-type semiconductors, their potentials of conduction band bottom (E_CB) are generally about 0.1–0.3 V higher than E_fb values. Herein, a median value of 0.2 V is chosen for calculation, and resultant values of E_CB of −0.34, 0, and −0.47 V are calculated for Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl, respectively. Considering E_g values for Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl, the potentials of the valence band top (E_VB) can be calculated to be 2.36, 2.76, and 2.79 V, respectively, using the formula E_CB = E_VB − E_g [48].

2.2. Photocatalytic NO Removal Performance

Photocatalytic NO removal efficiencies were measured by placing catalysts in a reactor with continuously flowing NO gas under UV–visible irradiation. To ensure the accuracy and reproducibility of results, each experiment was conducted at least three times in parallel to provide average data with error bars. Prior to the experiments, the effect of relative humidity in the reactor over NO removal was assessed in the presence of composite GBW5, as shown in Figure S8. Obviously, the NO removal efficiency at 50% relative humidity is approximately 20% higher than those at 25% and 75% relative humidity. Accordingly, 50% relative humidity was chosen for subsequent experiments.

The photocatalytic performance on NO removal by Bi₁₂GeO₂₀, W₁₈O₄₉, and the GBW series is depicted in Figure 4A. Both Bi₁₂GeO₂₀ and BiOCl exhibit weak removal efficiencies of around 10% after 30 min of light exposure, while W₁₈O₄₉ exhibits negligible catalytic activity under identical conditions. As for the GBW composites, their removal efficiencies increase greatly and achieve peak values after 5 min. In addition, the peak values can be maintained for up to 30 min, revealing the satisfactory stability of catalytic capacity. Moreover, with increasing WCl₆ addition, the photocatalytic performance of the GBW series first increases and then reduces. As a result, the composite GBW5 with a suitable composition exhibits the best removal efficiency of around 42%, which is significantly higher than those of pure Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl. The trend in photocatalytic removal of NO_x (including NO and NO₂) under UV–visible light for different samples shown in Figure S9A shows the same behavior as for the NO removal shown in Figure 4A. However, GBW composites cause a slight increase in NO₂ production in comparison to pure Bi₁₂GeO₂₀, W₁₈O₄₉, and BiOCl, as shown in Figure S9B, which is unexpected since NO₂ molecules are more toxic than NO. For a comprehensive evaluation of NO_x elimination, both the NO removal and NO₂ generation should be considered by using the DeNO_x index. The larger the DeNO_x index, the better the photocatalytic performance. Significantly, as shown in Figure 4B, the DeNO_x index of GBW5 is the largest among all values by the different candidates, revealing the best overall efficiency among all as-synthesized samples. In order to confirm the generation of nitrate (NO₃⁻) and nitrite (NO₂⁻) species during the photocatalytic processes, XPS spectra of composite GBW5 stirred for 1 h in an aqueous solution containing NaNO₃ and NaNO₂ and after five consecutive runs (150 min) were recorded. As shown in Figure S10, signals at 403.2 and 407.3 eV are both clearly visible in the spectra for GBW5 stirred for 1 h in an aqueous solution, assigned to NO₂⁻ and NO₃⁻ adsorbed on the surface, respectively. A distinct signal at 406.7 eV is observed in GBW5 after a reaction of 150 min, indicating the formation of NO₃⁻ species during the reaction. The absence of another signal indexed to NO₂⁻ is possibly attributed to its instability and ease of converting to stable NO₃⁻. Generally, the apparent reaction kinetics constant (K_app) of catalyst processes can be calculated from the pseudo-first-order kinetic as shown in Figure S11 and Figure 4C. As can be seen, the composite GBW5 has the largest K_app value at 67.83 × 10⁻³ min⁻¹, which is about 3.7 and 24 times those of Bi₁₂GeO₂₀ and W₁₈O₄₉, respectively. From the aforementioned experimental results, the GBW ternary heterojunction composites exhibit visibly improved photocatalytic capacity in comparison to the individual materials. As shown in Figure S12, the photocatalytic NO removal and NO₂ generation with the binary composites GB are slightly better than those with pure Bi₁₂GeO₂₀ and BiOCl but nowhere near as good as those with the GBW series, revealing the synergistic effect of integrating three components with the appropriate composition. On the whole, our catalyst system provides comparable and even better removal efficiencies in comparison to recent relevant reports, which are collected in Table S2 [49,50,51,52,53,54,55,56].

The transfer and segregation efficiency of photogenerated charge carriers can be assessed using electrochemical impedance spectroscopy (EIS) and transient photocurrent measurements, which directly relate to the photocatalytic performance of the catalysts. The EIS Nyquist plots demonstrate the interfacial impedance between electrode and electrolyte, where a smaller radius means a weaker charge transfer resistance. The Nyquist plots of Bi₁₂GeO₂₀, W₁₈O₄₉, BiOCl, and composite GBW5 are shown in Figure 4D. Clearly, the arc radii of relevant samples change in an order of GBW5 < BiOCl < Bi₁₂GeO₂₀ < W₁₈O₄₉ with the corresponding fitted data collected in Table S1, which is in good agreement with the trend in photocatalytic capacity. The intensity of the transient photocurrent indicates the transfer rate of photogenerated carriers under illumination. As shown in Figure 4E, composite GBW5 demonstrates the highest steady-state photocurrent response, far exceeding those of other samples, revealing the highly efficient migration and segregation efficiency of ternary heterojunction systems arising from the integration of the structural merits of the three components.

Five repeated photocatalytic experiments were conducted to estimate the recycling capability and structural stability of the as-synthesized composites. The first four experiments were consecutively carried out without interruption, the used GBW5 was then washed three times each with water and ethanol before being subjected to the fifth cycle. As shown in Figure 4F, photocatalytic performance remains high with only a slight reduction after four cycles and can be restored in the fifth cycle after simply washing. During photocatalytic processes, NO₂ generation remains approximately invariant, as shown in Figure 4G. These results confirm the impressive reusability of the composite. In addition, the XRD patterns of GBW5 before and after five cycles are identical, suggesting good maintenance of the composite and crystal structure, which indicates the sufficient structural stability of these robust samples.

By conducting capture experiments, the reactive species that contribute to photocatalytic processes can be identified. As depicted in Figure 5A, the photocatalytic removal efficiency of GBW5 is greatly inhibited after the addition of reagents KI, PBQ, TBA, or K₂CrO₇, confirming the crucial roles of species h⁺, e⁻, ·O₂⁻, and ·OH. The generation of ·OH radicals can be directly detected by fluorescence spectroscopy (PL). After light exposure for 30 min, intensive bands centered at 425 nm appear, as shown in Figure 5B, verifying the formation of ·OH radicals in the presence of both Bi₁₂GeO₂₀ and GBW5. Moreover, the fluorescence signal intensity of GBW5 is higher than that of Bi₁₂GeO₂₀, indicating the boosted generation of ·OH radicals due to the strong oxidative capability of these ternary composites. To further detect the formation ·OH and ·O₂⁻ radicals, electron spin resonance (ESR) measurements were performed. The typical signals of DMPO-·OH adducts in the spectra of both Bi₁₂GeO₂₀ and composite GBW5 are visible in Figure 5C. Neither Bi₁₂GeO₂₀ nor GBW5 causes the generation of ·OH radicals in the dark. In contrast, evident ESR signals in the spectra of both samples can be found after 10 min of light irradiation, revealing the possible formation of ·OH radicals upon illumination. Moreover, the stronger ESR signals generated with GBW5 compared to those generated with Bi₁₂GeO₂₀ suggest the strengthened production of ·OH radicals because of efficient regulation of interfacial charge carriers in ternary composites. Similar results are found for the appearance of DMPO-·O₂⁻ signals in methanol under light illumination in Figure 5D, revealing that more ·O⁻ radicals can be produced by GBW5 than by Bi₁₂GeO₂₀. Although the GBW5 and Bi₁₂GeO₂₀ catalytic systems are both able to induce the production of ·OH and ·O₂⁻ radicals, the synergistic effect represented by the ternary composites is favorable for the formation of sufficient radicals, thus enhancing photocatalytic NO removal performance [48].

Based on the experimental and analytical results described above, a probable mechanism for photocatalytic NO removal using the GBW composites is proposed in Figure 6. As described, W₁₈O₄₉ and BiOCl phases are generated on the surface of Bi₁₂GeO₂₀, thus the relative position of the three components with relevant band structures is exhibited on the left side of the diagram. The production of a BiOCl phase on the surface of Bi₁₂GeO₂₀ through the comprehensive utilization of “waste” Cl anions may form a heterojunction structure that regulates interfacial charge carriers. In addition, the presence of W₁₈O₄₉ nanosheets is capable of enhancing visible light absorption and promoting the separation of charge carriers. As a result, a synergistic effect is formed at the interface between the three components. Under ultraviolet–visible light, the three components are excited and distribute electrons to the CB and holes to the VB. BiOCl possesses a more negative CB and positive VB than Bi₁₂GeO₂₀, thus, electrons and holes from BiOCl tend to thermodynamically transfer to Bi₁₂GeO₂₀. Because of the faster transfer rate of e⁻ than h⁺, more electrons accumulate in the CB of Bi₁₂GeO₂₀ as revealed by XPS analyses and more holes remain in the VB of BiOCl, thus causing efficient segregation efficiency through a Type I scheme. Subsequently, holes in the VB of Bi₁₂GeO₂₀ can recombine with electrons on Bi₁₂GeO₂₀ and holes on W₁₈O₄₉ through a Z-scheme model. Accordingly, electrons easily convert to ·O₂⁻ radical because of the more negative CB potential of Bi₁₂GeO₂₀ (0.34 V) than the potential (O₂⁻/·O₂⁻, −0.33 V), and holes readily convert to ·OH radicals because of more positive VB potentials of BiOCl (2.79 V) and W₁₈O₄₉ (2.76 V) than potentials (OH⁻/·OH, 1.99 V) and (H₂O/·OH, 2.27 V), which can be confirmed by the entrapping experiments, PL spectra, and ESR spectra in Figure 5. As a result, the combination of both Type I and Z-scheme transfer models effectively suppresses the recombination of photogenerated charge carriers, prolongs the lifetimes of charge carriers, and enhances the light-responsive capability, thus facilitating the improvement of photocatalytic NO removal efficiencies of the GBW ternary heterojunction composites.

3. Materials and Methods

3.1. Catalysts Fabrication

Detailed information on the reagents used in this study can be found in the Supplementary Materials. The original Bi₁₂GeO₂₀ was simply prepared by a solvothermal method. Typically, Bi(NO₃)₃·5H₂O (2.91 g, 6 mmol) was dispersed in a NaOH solution (5 M, 35 mL), and the mixture was stirred magnetically until the color turned to light yellow. Subsequently, GeO₂ (0.05 g, 0.5 mmol) and cetyltrimethylammonium bromide (0.25 g, 0.7 mmol) were added to the solution to make a mixture that was vigorously stirred for 30 min. The resultant mixture was then transferred to a Teflon-lined autoclave. After heating at 180 °C for 12 h, the suspension was centrifuged, washed three times with water and ethanol, and then dried at 60 °C for 8 h to obtain pristine Bi₁₂GeO₂₀. W₁₈O₄₉ was synthesized using WCl₆ as the tungsten source and anhydrous ethanol as the solvent through a solvothermal method [33].

The ternary composites were fabricated by a post-synthesis protocol, where the as-synthesized Bi₁₂GeO₂₀ was added during the preparation of W₁₈O₄₉. Typically, the required amount of WCl₆ (0.065 g, 0.108 g, 0.151 g) was dissolved in anhydrous ethanol (40 mL) to produce a mixture that was stirred until the yellow solution turned colorless. Afterward, Bi₁₂GeO₂₀ (1 g) was added and the resultant suspension was magnetically stirred for 30 min, followed by ultrasonic treatment for another 30 min. The mixture was then transferred to a Teflon-lined autoclave and heated at 160 °C for 12 h. After cooling to room temperature, the precipitate was collected by centrifugation, washed three times with water and alcohol, and dried in a vacuum oven at 60 °C for 6 h. These obtained composites were labeled GBW3, GBW5, and GBW7 according to the mass ratio of elemental W to Bi₁₂GeO₂₀. During the aforementioned procedure, BiOCl was also synthesized by a solvothermal method as follows. Bi(NO₃)₃·5H₂O (0.97 g, 2 mmol) in ethylene glycol (76 mL) and NaCl (0.234 g, 4 mmol) in water (5 mL) were mixed to gain a mixture that was stirred for 30 min and then left to stand at 140 °C for 5 h to supply BiOCl. Binary composites Bi₁₂GeO₂₀/BiOCl were synthesized using a similar procedure to that of the ternary composites, except that WCl₆ was replaced by NaCl as the chlorine source, based on moles of Cl element. The syntheszied composites were labeled GB3, GB5, and GB7. The schematic diagram of the synthetic procedure of the as-synthesized samples is shown in Figure S1.

The microstructural analyses and relevant instruments are listed in the Supplementary Materials for clarity.

3.2. Photocatalytic NO Removal and Entrapping Experiments

The specific procedure for the photocatalytic NO removal and recycling experiments are provided in the Supplementary Materials. Entrapping experiments were conducted to identify the reactive species that played roles during photocatalytic NO removal. To that end, a reagent such as KI (0.1 g), tert-butanol (TBA, 0.1 mM), p-benzoquinone (PBQ, 0.1 g), or potassium dichromate (K₂CrO₇, 0.1 g) was introduced into the photocatalytic systems to trap holes (h⁺), hydroxyl radicals (∙OH), superoxide radicals (∙O₂⁻), or e⁻, respectively. The procedure was the same as that used for photocatalytic NO removal. The active species ∙OH and ∙O₂⁻ captured by 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethyl-4-piperidone hydrochloride (TEMP) were detected using electron spin resonance (ESR) spectroscopy. ∙OH radicals entrapped by terephthalic acid were detected using an F96Pro spectrophotometer (Shanghai Lengguang Technology Co., Ltd., Shanghai, China).

4. Conclusions

In this investigation, the integration of “sea urchin-like” W₁₈O₄₉ and BiOCl by comprehensively utilizing all ions in precursor WCl₆ was successfully completed on the surface of Bi₁₂GeO₂₀, leading to the generation of ternary composites, designated GBW. These as-prepared composites were thoroughly analyzed by various techniques and subjected to photocatalytic NO removal evaluation. Consistent with expectations, these ternary composites showed enhanced photocatalytic performance under UV–visible irradiation. Among all tested samples, the composite GBW5 showed the highest NO photocatalytic removal efficiency of 42%, which was nearly four times that of pure Bi₁₂GeO₂₀. In addition, it possessed the largest K_app value and DeNOx index, which was mainly attributed to enhanced light absorption, suitable morphological features, effective separation of charge carriers, and boosted generation of reactive species. Based on the analytical and experimental results, the robust GBW composites underwent a combination of both Type I and Z-scheme transfer, by which charge carriers could efficiently migrate and be segregated. This study paves the way for the construction of Bi₁₂GeO₂₀-based ternary composites using a comprehensive utilization of waste matter and their employment for the photocatalytic removal of low concentrations of NO at the ppb level.