3.1. Crystal and phase analysis
The composition and crystallinity of the samples were measured by X-ray diffraction (XRD). Figure 1 shows the XRD spectra of pure Bi2O3, BiOIO3 and final products doped with different concentrations (X = 0,1,10) of Yb3+. The diffraction peak of BiOIO3 is consistent with that of standard card JCPDS#26-2019(Wu et al. 2018). Compared with the α-Bi2O3 standard card(JCPDS#41-1449), the characteristic diffraction peaks 2θ of prepared Bi2O3 are 25.7°,26.9°,27.4°,28°,33.2°,43.6°,52.9° and 55.5°, respectively. The reflection peaks corresponding to (002), (111), (120), (012), (200), (041), (231) and (222) of α-Bi2O3 monoclinic crystal indicate that the crystal form of the material is α-Bi2O3, without impurities. When Yb3+ was added to the YBB mixture at 0%, no additional wave peaks were observed in the image, indicating that no other crystal phase was generated. When the Yb3+ doping ratio is 1%, the X-ray diffraction patterns show that 2θ is equal to 27.9°. It belongs to the structural peak of tetrahedral β-Bi2O3 (JCPDS 27–0050), which corresponds to the (201) plane of the material, i.e. The coexistence phase structure of α-Bi2O3 and β-Bi2O3 exists in the prepared composite samples. When the concentration of Yb3+ increases to 10%, the phase transition of Bi2O3 can be clearly observed. The reason is that the radius of Bi3+ is 103pm, which is close to the radius of rare earth ions. Rare earth ions easily enter into the Bi2O3 lattice to replace some Bi3+, resulting in Bi2O3 phase transition. However, after the Bi2O3 doped rare earth ions and coupled with BiOIO3, no other impurity diffraction peaks appear, indicating that Bi2O3 does not cause new crystal orientation of BiOIO3. With the increase of the proportion of Bi2O3 in the composite, the diffraction peak intensity of BiOIO3 decreases gradually, which indicates that the coupling between Bi2O3 and BiOIO3 is successful.
3.2. Morphology and structure analysis
SEM and TEM analysis were performed on the prepared samples to observe the microstructure difference of pure samples BiOIO3, Bi2O3, and YBB composite photocatalyst. SEM diagrams of BiOIO3, Bi2O3, YBB-0, YBB-1, and YBB-10 photocatalysts are shown in Fig. 2 (a)-(e). From the SEM images of the pure samples BiOIO3 and Bi2O3 in Fig. 2 (a) and (b), it can be seen that BiOIO3 has a smooth surface and a neat edge, which is a layered structure with good crystallinity. Bi2O3 particles are peanut-like and evenly distributed, and a small amount of agglomeration is caused by the calcination process. In Fig. 2 (c), Bi2O3 was successfully loaded onto BiOIO3. After the introduction of 1% Yb3+, the morphology of Bi2O3 in Fig. 2 (d) changes. After the introduction of 10% Yb3+, the morphology of Bi2O3 in Fig. 2 (e) changes more obviously, the particle size increases, and the agglomeration phenomenon aggravates, which has a great relationship with the crystal transformation of Bi2O3, indicating that the doping of Yb3+ can cause the crystal transformation of Bi2O3, and the crystal transformation is also different with the doping amount.
It can be clearly seen from TEM images that the transparent layered structure is in great contact with the shadowed irregular particle structure. In order to further illustrate the bonding mode between different phases of BiOIO3 and Bi2O3, a high-resolution projection electron microscope (HRTEM) was placed on the contact surface of the material, as shown in Fig. 3 (b). It can be clearly seen from the figure that the lattice stripe spacing of BiOIO3 and Bi2O3 materials is 0.290, 0.319 and 0.407nm, which correspond to the (002) plane of BiOIO3, the (111) plane of β-Bi2O3 and the (020) plane of α-Bi2O3, respectively. It is consistent with the XRD characterization results, indicating that when 1% Yb3+ ions are doped, α-Bi2O3, β-Bi2O3 and BiOIO3 coexist and compound with each other to form a triple heterojunction. It provides strong evidence for exploring the transfer path of photogenerated electron hole pairs between semiconductors in the process of photocatalysis. In the sample element analysis shown in Fig. 3 (c), all elements are evenly distributed in the whole nanocomposite, which proves that the YBB sample has been successfully synthesized.
3.3. Surface characterization
X-ray photoelectron spectroscopy (XPS) was used to characterize the valence and binding states of the elements in the samples. The XPS spectra of BiOIO3 nanosheets, Bi2O3 nanoparticles and YBB-1 composite are shown in Fig. 4. Figure 4 (a) shows the full XPS spectra of YBB-1. Bi, O, I and Yb elements were detected in the nanocomposites, indicating that the samples were successfully prepared, which further provided evidence for the coupling of BiOIO3 and Bi2O3. The characteristic peak of carbon element is due to detection error, and the C 1 peak with a binding energy of 284.6eV is used for calibration(Jia et al. 2021), so only the peaks of Bi, I, O, and Yb need to be considered. Figure 4 (b) - (e) extend the high resolution spectra of Bi 4f, I 3d, O 1s and Yb 4d. In Fig. 4 (b), the Bi 4f spectrum of YBB-1 has two characteristic peaks, Bi 4f7/2 and Bi 4f5/2, corresponding to 159.0eV and 164.3eV, respectively, which is consistent with the chemical state of Bi3+ in BiOIO3 and Bi2O3(Yang et al. 2019). Figure 4 (c) shows the pattern of I 3d. The dominant peaks at 623.7eV and 635.2eV correspond to the I 3d 5/2 and I 3d3/2 orbitals of I5+ in BiOIO3(Jia et al. 2020), while the peaks at 618.8eV and 630.5eV are attributed to the contribution of I−. It shows that part of I5+ is reduced to I− in the preparation process, which proves that there are interactions between BiOIO3 nanosheets and Bi2O3 nanoparticles. The O 1s spectrum can have three independent peaks at 529.3, 529.9 and 531.1 eV, respectively. The peaks at 529.3 and 529.9 eV denoted the lattice oxide in Bi − O and I − O bonds, respectively. The surface absorbed oxygen (− OH group and chemisorbed oxygen-containing species) corresponds to the peak located at 531.1 eV (Fig. 4d). The Yb3+ peak appeared at 184.1 eV (Fig. 4e). Compared with the original BiOIO3 and Bi2O3, the binding energies of Bi, I and O in YBB-1 move towards the direction of high binding energy, which indicates that BiOIO3-Bi2O3 indeed interacts at the composite interface.
The specific surface area, pore volume and pore size of a photocatalyst affect its catalytic performance, so the Brunauer-Emmett-Teller (BET) method was carried out. As shown in Fig. 5(a), the specific surface area of the sample was measured by the N2 adsorption-desorption isotherm method to verify the adsorption performance of the catalyst. All the isotherms belong to type IV and H3 hysteresis loops existed, verifying their mesoporous characteristics(Cheng et al. 2019), which could promote the photocatalytic oxidation reaction on the surface of a photocatalyst. The specific surface areas of BiOIO3, Bi2O3, YBB-0, YBB-1 and YBB-10 were calculated by BET method. As shown in Table 1, the specific surface area values were 20.14, 39.20, 11.38, 11.50 and 11.23m2/g, respectively. Obviously, the specific surface area of the composite sample decreases with the increase of the Yb3+ molar ratio, among which YBB-1 has the largest specific surface area in the composite sample, and the large specific surface area of the YBB-X composite can promote the photocatalytic reaction. The specific surface area of YBB-0 in the composite sample is the smallest and smaller than that of the pure sample, which can be attributed to the agglomeration of nanocomposites due to the calcination of the composite sample. The specific surface area may affect the catalytic oxidation performance, but it is not the decisive factor, so it needs to be further analyzed with other characterizations.
Table 1
Physical properties and photocatalytic mercury removal efficiency of BiOIO3, Bi2O3, YBB-0, YBB-1 and YBB-10
Sample
|
BET Surface Area (m2/g)
|
Pore Volume (cm3/g)
|
Pore Size (nm)
|
YBB-0
|
11.3846
|
0.075207
|
14.0805
|
YBB-1
|
11.4965
|
0.080337
|
14.8904
|
YBB-10
|
11.2277
|
0.049413
|
17.6038
|
Bi2O3
|
39.1974
|
0.044527
|
4.5439
|
BiOIO3
|
20.1372
|
0.083724
|
16.6308
|
3.4. Optical properties
As we all know, the light absorption capacity of a photocatalyst is an important factor affecting its photocatalytic performance. Figure 5(b) depicts the UV-Vis DRS spectrums of the composite samples. The absorption edges of BiOIO3 and Bi2O3 nanoplates are about 400 nm and 468 nm, respectively, indicating that the response spectra of Bi2O3 and BiOIO3 ranged from ultraviolet to part of visible light. YBB-0, YBB-1 and YBB-10 have a wide absorption tail at 400 ~ 612 nm. The absorption boundaries of YBB-0, YBB-1 and YBB-10 are 488, 600 and 612 nm respectively. Compared with the pure sample, the absorption edge has a significant red shift, indicating that Bi2O3 and BiOIO3 are successfully compounded and the band gap energy is reduced. YBB-1 and YBB-10 have large absorption cross sections in the range of 850 ~ 1100 nm, overlapping with one of the maximum near-infrared energy distributions of sunlight. Among them, the absorption peak at 980nm is the highest, which just corresponds to the absorption of Yb3+, further proving that the sample preparation is successful(Negreira et al. 2013).
In order to further verify the optical properties of nanocomposites, the band gap energy (Eg) of photocatalysts was calculated.
The band gap energy (Eg) of the prepared samples can be estimated according to equations (2) and (3):
$${\text{E}}_{\text{g}}=\frac{1240}{{{\lambda }}_{\text{A}\text{b}\text{s}\text{o}\text{r}\text{p}.\text{E}\text{d}\text{g}\text{e}}} \left(2\right)$$
$${\alpha }\text{h}{\nu }=\text{A}{\left(\text{h}{\nu }-{\text{E}}_{\text{g}}\right)}^{\frac{\text{n}}{2}} \left(3\right)$$
where Eg, α, hν and A denote band gap, absorption coefficient, photon energy and constant, respectively. BiOIO3 is an indirect transition semiconductor, n = 4. Bi2O3 is a direct semiconductor, n = 2. The band gap (Eg) of BiOIO3 and Bi2O3 can be estimated from the relative photon energy diagram. The tangent intercept is Eg of BiOIO3 and Bi2O3, as shown in Fig. 5(c)-(d), with values of 3.0 eV and 2.85 eV, respectively. Although pure BiOIO3 is easy to transfer e− to the (001) plane along the Z-axis in an electric field, it still faces a large recombination rate of e− and h+. The recombination and generation of electron hole pairs cancel each other out, resulting in a reduction of the efficiency of the photocatalyst. However, the band gaps of the composites are lower than those of the two. From the subsequent DFT study, it can be seen that Yb3+ doping provides an impurity level for Bi2O3, thereby reducing the total band gap. Therefore, the higher reaction efficiency is due to their excellent separation ability of photogenerated electron hole pairs.Second, the impurity level reduces the total band gap and increases the optical absorption boundary.
Photoluminescence spectroscopy (PL) is a general method, that can be used to further analyze the optical properties of photocatalysts to characterize the separation ability of photogenerated electrons and holes. In the PL spectrum, the lower the intensity of the peak, the lower the recombination efficiency of electrons and holes. In this work, the F-4600 FL spectrophotometer is adopted and the excitation wavelength is 330nm and the wavelength range is 400-600nm. It can be seen from Fig. 10 (a) that the fluorescence intensity of YBB-0, YBB-1 and YBB-10 is significantly lower than that of pure BiOIO3 and Bi2O3, among which YBB-1 shows the lowest emission intensity, which indicates that YBB-1 has the lowest electron hole recombination efficiency. In addition, the peak value is around 455nm, which can be attributed to the formation of extra doped energy levels and triple heterojunction.
Then, the efficiency of electron hole separation of composite materials was further studied by using Electrochemical impedance (EIS) and photoelectric current response. As shown in Fig. 10 (b), YBB-1 has the smallest radius of arc, which indicates that the interface resistance is lower than that of other samples, which is conducive to the separation and migration of charge. Figure 10 (c) shows the transient photocurrent responses of BiOIO3, Bi2O3, YBB, YBB-1 and YBB-10, and the generally higher photocurrent density represents a better charge separation. The photocurrent response of the prepared sample is cycled four times under visible light. It can be seen that after turning on the light and turning off the light, the transient photocurrent of the sample increases and then decreases immediately, showing good photocurrent reproducibility, among which YBB-1 has the largest photocurrent density. It may be due to the joint action of rare earth ion doping and triple heterojunction. In conclusion, the electrochemical impedance and photocurrent response show that YBB-1 has the best photochemical properties.
3.5. Photocatalytic performance
The Hg0 removal experiment was carried out under the 24W LED light (k < 400 nm) of visible light. The experiment was divided into three stages and lasted 70min. In the first part of 15min, because the specific surface area of the catalyst changes greatly, the influence of adsorption must be considered. The second part is the most important part of the whole experiment. Open the visible light source for 40min to completely remove Hg0 pollutants through photocatalysis. In the last stage, the stabilization process is reached 15min after the lights are turned off. The overall photocatalytic reaction efficiency is shown in Fig. 6 (a) below. It can be seen from the figure that the mercury removal efficiency of YBB-0, YBB-1, YBB-10, BiOIO3 and Bi2O3 is 57.51%, 76.73%, 67.58%, 54.43% and 39.35%, respectively. All of them are much more efficient than the standard catalyst P25 reported before(Wu et al. 2018). The mercury removal efficiency of composite photocatalysis is higher than that of a single catalyst, which indicates that the two substances are successfully combined to form a heterojunction. The doping of Yb3+ makes Bi2O3 undergo phase transition. In YBB-0, only α-Bi2O3 and BiOIO3 compounds form a common single heterostructure, which can promote the separation of photogenerated electron hole pairs In YBB-1, α-Bi2O3 and β-Bi2O3 crystal phases coexist with bismuth trioxide, and the three combine to form a triple heterostructure, which constructs a more complex electronic transmission channel. However, α-Bi2O3 in YBB-10 is completely converted to β-Bi2O3, and a single heterostructure is restored, which reduces the photocatalytic efficiency. In addition, Fig. 6 (b) further describes the mercury removal efficiency of YBB-1 and YBB-0 photocatalysts under near-infrared radiation. The mercury removal efficiency of Yb3+ doped composite photocatalyst YBB-1 under infrared light is twice as high as that of Yb3+ doped composite photocatalyst YBB-0. This is due to the unique electronic transition characteristics of Yb3+, which broadens the spectral response range of the composite photocatalyst and improves the photocatalytic performance.
In order to further intuitively compare the reaction kinetics of Hg0 photocatalytic oxidation, the Langmuir Hinshelwood pseudo first order reaction model was studied(Negreira et al. 2013, Jia et al. 2020):
$$-\text{ln}\left(\frac{C}{{C}_{0}}\right)=kt \left(4\right)$$
where C0 and C represented inlet and outlet gas-phase Hg0 concentration (µg/m3), k and t represented pseudo-first-order rate constant (min− 1) and reaction time. Hence, as displayed in Fig. 6 (e), the reaction rate constant was determined to be 0.01428, 0.02430, 0.018773, 0.013098 and 0.008334 min− 1 respectively, for YBB-0, YBB-1, YBB-10, BiOIO3 and Bi2O3. The highest photocatalytic reaction rate of YBB-1 is about 3 times that of Bi2O3 and 2 times that of BiOIO3, which further proves that YBB-1 has better photocatalytic performance than other samples.
The stability of the photocatalyst is an important factor affecting its practical application. The photocatalytic stability of YBB-1 samples was tested by the cyclic photocatalytic oxidation of Hg0. Each cyclic experiment was carried out under the same conditions. First, the sample was placed in the dark for 30min, then the quartz glass coated with YBB-1 sample was placed under the LED lamp for 90min, the LED lamp was turned off, and then the Hg0 concentration reached a stable value. About 30min later, the next cycle experiment was carried out in the same steps. After five cycles of experiments, it can be seen from Fig. 6 (c) that the mercury removal efficiency changed from 76.7–71.2%, only about 5% decreased. At the same time, the photocatalyst after five cycles was characterized by XRD. It can be seen from Fig. 6 (d) that the sample atlas has almost no change, indicating that the composite photocatalyst has good stability and chemical durability.
3.6. Reaction mechanism
3.6.1. Theoretical calculation
According to the above characterization of physical and chemical properties and the experiment of mercury removal by photocatalytic oxidation, it can be determined that YBB-1 has the best photocatalytic performance. The possible mechanism of improving photocatalytic activity was further explained by density functional theory (DFT) calculation. Figure 7 (a) - (b) shows the geometric optimization structure of BiOIO3 and α-Bi2O3. Band gap is the potential difference between the maximum value of the valence band (VB) and the minimum value of the conduction band (CB), which has an important impact on the photoelectric performance of catalysts(Jia et al. 2020). Therefore, we simulated the band gap structure of the catalyst by CASTEP calculation. In Fig. 7 (c)-(d), it is calculated that the band structures of BiOIO3 and α-Bi2O3 are 1.961 and 2.376 eV respectively, which is slightly different from the experimental calculation value. It can be attributed to the calculation error of the GGA functional that usually underestimates the band gap(Jia et al. 2020, Zhang et al. 2021). Simultaneously, the density of states (DOS) of BiOIO3 and α-Bi2O3 were calculated, and the results are shown in Fig. 7 (e)-(f).. It can be observed from Fig. 7 (e) that in the case of BiOIO3 bulk, the VB was predominately composed of O 2p states, and the bottom of CB was made up of Bi 6p states, O 2p states, and I 5p states. The results are consistent with previous literature on BiOIO3(Wang et al. 2013, Huang et al. 2014). In Fig. 7 (f), it can be seen that the VB is mainly derived from the O 2p orbitals in the case of α-Bi2O3 bulk, whereas the CB is mainly derived from the Bi 6p orbitals, which is consistent with the results of previous studies(Huang 2009, Guo et al. 2019).
Compared with pure Bi2O3, Yb3+ doping can effectively reduce the band gap of Bi2O3 by generating impurity energy levels, which may lead to more light absorption (Fig. 8a). It is consistent with the results of UV-Vis-NIR spectra in Fig. 7. From the detailed Yb3+ doped DOS spectrum shown in Fig. 8 (b), it can be seen that VBM still mainly comes from the O 2p orbital. The CBM is mainly composed of Bi 6p and Yb 4f orbitals, while the new impurity band is mainly contributed by Yb 4f orbitals and a small amount of O 2p orbitals. The results show that Yb3+ doping can induce O 2p orbitals to generate impurity energy levels in the band gap of Bi2O3, which can reduce the band gap and promote the separation of photogenerated electron hole (e−/h+) pairs, which is conducive to the photocatalytic performance.
The surface work functions of BiOIO3(010)(Di Liberto et al. 2021), α-Bi2O3(001)(Chen et al. 2020) and β-Bi2O3(111)(Tian et al. 2021) planes were calculated by CASTEP using density functional theory, so as to reveal the interface transfer between them more deeply(Fu et al. 2019). The work functions of BiOIO3(010), α-Bi2O3(001) and β-Bi2O3(111) are calculated according to the following equation:
$${\Phi }={E}_{vac}-{E}_{f} \left(5\right)$$
Where Evac is the electrostatic potential at the vacuum level and Ef is the fermi energy(Wei et al. 2019). The electrostatic potentials of BiOIO3(010), α-Bi2O3 (001), and β-Bi2O3 (111) planes are shown in Fig. 8 (c)-(e). We calculated and attained the work functions of BiOIO3(010), α-Bi2O3(001) and β-Bi2O3(111) planes, which are 5.842, 3.818 and 3.29eV, respectively. This indicates that there is charge transport at the interface between the three semiconductors. The charge can be transferred from BiOIO3 with a higher work function to Bi2O3 with a lower work function. At the same time, between different crystal phases of Bi2O3, the charge is transferred from α-Bi2O3 with a higher work function to β-Bi2O3 with a lower work function. Finally, the Fermi energy levels will be equal(Xu et al. 2018). The high work function results in the loss of electrons at the interface to obtain holes, which are positively charged. The low work function electrons lose holes and are negatively charged. Therefore, the interface heterojunction of BiOIO3 and α-Bi2O3, the interface heterojunction of BiOIO3 and β-Bi2O3, and the interface homojunction of α-Bi2O3 and β-Bi2O3 form three progressive internal electric fields respectively, accelerating the separation and transmission of charge carriers(Fu et al. 2019). Due to the loss of electrons, the energy band edges of BiOIO3 and α-Bi2O3 at the interface bend upward. Due to the accumulation of electrons, the energy band edges of β-Bi2O3 at the interface bend downward, which further effectively inhibits the recombination of photogenerated carriers(Wang et al. 2020).In conclusion, the band bending and built-in electric field at the interface of BiOIO3, α-Bi2O3 and β-Bi2O3 promote charge transfer, and maintain the high reducibility of β-Bi2O3 and the high oxidation capacity of BiOIO3(Fu et al. 2019).
3.6.2. Charge transfer
The electron spin resonance (ESR) technology of DMPO was used to test the reactivity of YBB-1 under dark and visible light conditions, and to study the substances that play a important role in mercury removal under visible light. As shown in Fig. 9 (a), •O2− and •OH ESR signals are not generated in the dark. A group of six spectral peaks were observed under visible light, which were specific signal peaks of DMPO-•O2−. This means that O2 on the catalyst surface is reduced to •O2− by electrons. Similarly, by dispersing the photocatalyst in water, a strong DMPO-•OH signal of 1:2:2:1 can be observed, indicating that •OH is formed on the surface of the composite catalyst. Therefore, the above experimental results show that •O2− and •OH are mainly responsible for the photocatalytic oxidation of gaseous Hg0.
In addition, the flat band potential (Efb) of BiOIO3 and Bi2O3 is further measured by using the Mott-Schottky curve. The positive slope of the Mott-Schottky curve in Fig. 9 (c)-(d) indicates that BiOIO3 and Bi2O3 are both n-type semiconductors. It can be seen from the figure that the flat band potential (Efb) of BiOIO3 and Bi2O3 are − 0.69 eV and − 0.46eV, respectively. Because the value calculated by the formula is close to the conduction band value, and it is often equal to the conduction band potential (ECB) of the material, so the CB of BiOIO3 and Bi2O3 are − 0.69eV and − 0.46eV respectively. According to formula (6), the VB values of BiOIO3 and Bi2O3 are 2.31 eV and 2.39 eV respectively.
$${E}_{CB}={E}_{VB}-Eg \left(6\right)$$
Based on the above analysis, YBB-1 has good photocatalytic activity for Hg0 oxidation. Figure 10 shows a possible mechanism for enhancing photocatalytic activity. It can be seen from the figure that the separation and transfer of photogenerated electrons and holes may involve the following processes: because of the doping of rare earth ion Yb3+, Bi2O3 has two crystal phases: α-Bi2O3 and β-Bi2O3. According to the previous calculation, it can be seen that the conduction band position sequence is BiOIO3 > α-Bi2O3 > β-Bi2O3. Therefore, as shown in the figure, When BiOIO3 contacts with crystal α in Bi2O3, the h+ generated after BiOIO3 absorbs light energy can be left on VB, and then the e− on the CB of BiOIO3 transfers to the CB of β-Bi2O3 through α-Bi2O3, while the e− excited by visible light on the CB of α-Bi2O3 also transfers to adjacent β-Bi2O3. Similarly, When BiOIO3 contacts with crystal β in Bi2O3, e− is lost on the CB of BiOIO3, and e− is obtained from β-Bi2O3. Similarly, h+ moves to the VB of BiOIO3 in the opposite path. Based on the construction of its complex electron transport orbit, the recombination rate of photogenerated electrons and holes is greatly hindered, even almost no recombination. Since O2/·O2− (-0.28eV) is greater than e− (-0.46eV), the photogenerated electrons on the CB of Bi2O3 can easily reduce O2 to strong oxidizing ·O2− ions. Since ·OH /OH− (+ 1.99eV) is less than h+ (+ 2.31 eV), h+ can oxidize H2O and OH− to ·OH, and these active oxygen clusters (·O2− and ·OH) can oxidize Hg0 to HgO. The generated HgO can be well adsorbed on the catalyst surface for further removal.
In conclusion, YBB-1 has the best photocatalytic performance. First of all, Yb3+ doping Bi2O3 makes Bi2O3 phase transition. BiOIO3, α-Bi2O3 and β-Bi2O3 coexist and double compound to form a triple heterostructure, which constructs a unique electron transport orbit and promotes the separation of photogenerated electrons and hole pairs. Secondly, Yb3+ doping generates an impurity level, which reduces the band gap of Bi2O3. At the same time, Yb3+ has unique electronic orbital transition characteristics(Zhou et al. 2020), effectively broadening the spectral response of composite photocatalysis. The Hg0 removal efficiency of YBB-1 is up to 76.73%. The specific reaction process is as follows:
$$2{\text{F}}_{5/2}\left(\text{Y}\text{b}\right)+2{\text{F}}_{5/2}\left(\text{Y}\text{b}\right)\leftrightarrow 2{\text{F}}_{7/2}\left(\text{Y}\text{b}\right)+ 2{\text{F}}_{7/2}\left(\text{Y}\text{b}\right)+\text{h}\text{v} \left(7\right)$$
$$\text{h}\text{v}(\text{B}\text{i}\text{O}\text{I}{\text{O}}_{3}/{\text{Y}\text{b}}^{3+}-{\text{B}\text{i}}_{2}{\text{O}}_{3}) \to {\text{e}}^{-}+{\text{h}}^{+} \left(8\right)$$
$${\text{H}}_{2}\text{O}\leftrightarrow {\text{H}}^{+}+{\text{O}\text{H}}^{-} \left(9\right)$$
$${\text{H}}_{2}\text{O}+{\text{h}}^{+}\to ·\text{O}\text{H}+{\text{H}}^{+} \left(10\right)$$
$${\text{O}\text{H}}^{-}+{\text{h}}^{+}\to ·\text{O}\text{H} \left(11\right)$$
$${\text{O}}_{2}+{\text{e}}^{-}\to ·{\text{O}}_{2}^{-} \left(12\right)$$
$${\text{H}\text{g}}^{0}+2{\text{h}}^{+}+2{\text{O}\text{H}}^{-}\to \text{H}\text{g}\text{O}+{\text{H}}_{2}\text{O} \left(13\right)$$
$$2·{\text{O}}_{2}^{-}+3{\text{H}\text{g}}^{0}+2{\text{H}}^{+} \to 3\text{H}\text{g}\text{O}+ {\text{H}}_{2}\text{O} \left(14\right)$$
$${\text{H}\text{g}}^{0}+2·\text{O}\text{H}\to \text{H}\text{g}\text{O}+{\text{H}}_{2}\text{O} \left(15\right)$$