Photocatalysis technology uses solar energy to transform toxic and harmful substances into non-toxic substances through a series of oxidation-reduction reactions. It has been widely investigated for its potential in applications for treatment of environmental pollutants, such as formaldehyde, toluene, oxalic acid, carbon dioxide, and nitrogen oxide [1-8]. There are three primary processes in photocatalytic reactions: (1) the generation of electron/hole pairs under light illumination; (2) the separation and migration of photo-stimulated electrons and holes; and (3) the photocatalytic redox reaction [9, 10]. Currently, the overall efficiency of photocatalytic conversion of typical photocatalysts is still unsatisfactory. Therefore, various strategies have been applied to improve photocatalytic performance, for instance, by doping with impurity atoms, constructing vacancies, designing heterostructures, and so on [11-13].
Bi2O2CO3 (BOC) is a typical "Aurivillius" phase in which a [Bi2O2]2+ layer and a [CO3]2– layer are integrated in an orthogonal manner, forming a two-dimensional structure [14]. The highly asymmetric internal structure of BOC is obviously beneficial for the formation of special planar BOC sheets [15, 16], and BOC is a candidate suitable for use in a variety of fields, such as healthcare, photocatalysis, humidity sensing, nonlinear optical applications, and energy storage in supercapacitors. Recently, the photocatalytic properties of BOC have drawn increased attention because BOC can be used as a semiconductor photocatalyst to efficiently degrade dye wastewater [17] and remove gaseous acetaldehyde [18]. However, pristine BOC strongly absorbs UV, which drives a high charge-carrier recombination rate, thereby limiting photocatalytic efficiency [19].
Recently, it was reported that photocatalytic nanomaterials can be enhanced by adding oxygen vacancies [20, 21]. In particular, surface oxygen vacancies could capture light-generated charges and transfer the captured electrons to the adsorbed species of catalyst, thus effectively preventing charge-carrier recombination. This would enhance the photocatalytic activity [22, 23]. On the other hand, by the introduction of surface oxygen vacancies, the optical response range could be expanded by lifting up the top of the valence band; thus reducing the band gap [24]. However, the potential mechanisms of the oxygen-vacancies-dependent photocatalytic activity and selectivity are not fully understood. Thus, constructing oxygen vacancies on the surface of BOC to enhance its photocatalytic performance, and understanding fully the surface oxygen-vacancies-dependent photocatalytic activity and selectivity are essential for developing a novel photocatalyst.
In this study, we introduced oxygen vacancies on the surface of pristine BOC by adding NaBH4 (the product is labeled OV-BOC) and found that the addition of NaBH4 could change the surface structure of BOC and produce more oxygen vacancies as reaction sites for activating reactants. The oxygen vacancies in BOC could decrease the forbidden band width, promote charge transfer and improve the photocatalytic performance of visible light. The results of electron spin resonance (ESR) spectra and density functional theory (DFT) calculation indicate that activation of O2 and H2O molecules is highly promoted and thus more reactive oxygen species (ROS) would be generated to participate in photocatalytic reactions. Therefore, the efficiency of NO removal should be greatly enhanced. More interestingly, the evolution process of intermediate products in the process of photocatalytic NO oxidation was monitored dynamically using in situ DRIFTS. The results imply that oxygen vacancies can also facilitate electron exchange between the intermediates and the surface OV in OV-BOC, leading them to be destroyed more easily by the active radicals. This is also beneficial for the transformation of NO into target products rather than toxic byproducts, so that the selectivity can be enormously promoted by the pathway: NO→NO2→NO3–. This work should provide new insights into the understanding of oxygen vacancies in reactant activation and in photocatalytic selectivity.
All chemicals used in this work were analytical grade and were used without further treatment. In a typical synthesis, a certain amount of Bi(NO3)3·5H2O (1.21 g) was dissolved in 100 mL of nitric acid solution (0.9 mol/L) and stirred for 30 min. Then, 15 mL of concentrated ammonia solution (25.0%) was added and a white precipitate was produced and stirred for 5 min. A CO2 gas stream (1.0 L/min) was introduced into the above suspension, which was stirred for 30 min. After the gas stream was stopped, the resulting precipitate was filtered, washed with water and ethanol four times, and then dried at 60 ℃ to get the final Bi2O2CO3 with no further treatment. The resulting product was labeled BOC.
In order to introduce the oxygen vacancies into the Bi2O2CO3, appropriate amounts of PVP (1.0 g) and BOC (2 mmol) were mixed with 100 mL of H2O in a 250 mL beaker and stirred for 20 min. Next, 30 mL of 70 mmol/L NaBH4 solution was prepared and gradually poured into the beaker, stirring for 1 h and aged for 1 h. The resulting precipitate was filtered, washed with water and ethanol four times and dried at 40 ℃ to get the final product, labeled OV-BOC. To illustrate the formation mechanism of OV-BOC via NaBH4 treatment, a surface molecular structure model and corresponding chemical equation are provided in Scheme S1 (Supporting Information).
The crystal phases of the product samples were analyzed using X-ray diffraction (XRD) with Cu Kα radiation (model D/max RA, Rigaku Co., Japan). X-ray photoelectron spectroscopy (XPS) with Al Kα X-rays (Thermo ESCALAB 250, U.S.A.) was used to investigate the surface properties. The morphology and structure of the samples were examined using scanning electron microscopy (SEM, model JSM-6490, JEOL, Japan) and transmission electron microscopy (TEM, JEM-2010, Japan). The UV-vis diffuse-reflectance spectrometry (UV-vis DRS) spectra were obtained for the dry-pressed disk samples using a scanning UV-vis spectrophotometer (UV2550, Shimadzu, Japan) equipped with an integrating sphere assembly, and using 100% BaSO4 as the reflectance sample. Photoluminescence (PL) studies (F-7000, HITACHI, Japan) were conducted to investigate the optical properties of the samples. Electron spin resonance (ESR) of radicals spin-trapped by 5, 5-dimethyl-1-pyrroline N-oxide (DMPO) was recorded on a JES FA200 spectrometer. Samples for ESR measurement were prepared by mixing the samples in a 40 mmol/L DMPO solution tank (aqueous dispersion for DMPO-•OH and methanol dispersion for DMPO-•O2–) and irradiating with visible light. Electron paramagnetic resonance (EPR) measurements were carried out on a Bruker ESP 500 spectrometer.
The photocatalytic activity was investigated using the removal ratio of NO at parts per billion proportions in a continuous-flow reactor at ambient temperature. The volume of the rectangular reactor, which was made of polymeric glass and covered with Saint-Glass, was 4.5 L (30 cm × 15 cm × 10 cm). A 150 W commercial tungsten halogen lamp was placed vertically outside the reactor. A UV cutoff filter (420 nm) was utilized to remove the UV light in the light beam. For each test, 0.20 g of the prepared sample was dispersed in distilled water (50 mL) in a beaker via ultrasonic treatment for dispersion and then coated onto two glass dishes (12.0 cm diameter). Each coated dish was dried at 60 ℃ to remove the water and then cooled to room temperature before the test.
The NO gas was obtained from a compressed gas cylinder at a concentration of 100 ppm; then the NO was diluted to about 600 ppb by the air stream. The desired relative humidity level of the NO flow was kept at 50% by passing the zero air streams through a humidification chamber. The gas streams were premixed completely using a gas blender, and the flow rates of the air stream and NO were maintained at 2.4 L/min and 24 mL/min, respectively, by a mass flow controller. After the adsorption-desorption equilibrium was achieved, the lamp was turned on. The concentration of NO was continuously measured using a NOx analyzer (Thermo Environmental Instruments Inc., Model 42c-TL). The removal ratio (η) of NO was calculated as η = (1 − C/C0) × 100%, where C and C0 are the concentrations of NO in the outlet steam and the feed stream, respectively. The generation ratio (φ) of NO2 was calculated as φ = C1/(C0–C) × 100%, where C1 is the concentration of NO2 in the outlet steam.
In situ DRIFTS measurements were conducted using the Tensor Ⅱ FT-IR spectrometer (Bruker) equipped with an in situ diffuse reflectance cell (Harrick). Photocatalysts were put into the reaction cell. First, the He gas (100 mL/min) was used to remove the residual hydrocarbons, H2O, and CO2 at 300 ℃. The real-time FT-IR spectrum after ventilation was utilized as the background. Then, the reaction mixtures (50 mL/min NO, 50 mL/min O2) were introduced into the cell. The NO adsorption to the catalysts was carried out for 20 min. Next, photocatalysts were illuminated using a visible light source (MUA-210) for 1 h. The real-time FT-IR spectra were detected every eight minutes. Meanwhile the gas fluxes were kept the same (50 mL/min NO, 50 mL/min O2). Finally, FT-IR spectra were recorded every two minutes with the same gas fluxes after turning off the light. The IR scanning range was 4000–600 cm–1. The intervals 2250–2000 cm–1 and 1120–840 cm–1 were analyzed to present the photocatalytic oxidation process happening on the catalysts.
Spin-polarized DFT-D2 calculations were conducted using the "Vienna ab initio simulation package" (VASP5.4), applying a generalized gradient correlation functional [25]. A plane-wave basis set with cut-off energy 450 eV within the framework of the projector-augmented wave method was employed [26]. The Gaussian smearing width was set to 0.2 eV. The Brillouin zone was sampled with a 3×3×1 Monkhorst Pack grid. All atoms were converged to 0.01 eV −1. Hybrid functionals based on the Heyd-Scuseria-Ernzerhof (HSE06) method [27] were applied to estimate the exact band structures. The adsorption energy (Eads) is defined as
Eads = Etot – (Es + Emol)
where Etot, Es, and Emol depict the total energy of the adsorption complex, the samples (pure BOC or OV-BOC), and the isolated molecule, respectively.
We established two models of oxygen vacancy (OV) to find the optimal one (Table 1). Two models of OV-BOC were used for comparison of the total energy (Etot) of the different OV positions in BOC. The DFT results show that the total energy of Position 2 (−583.72619 eV) is lower than the total energy of Position 1 (−580.27379 eV), which suggests that Position 2 is more stable. Then, we chose the model of Position 2 to study the effects of oxygen vacancy in BOC.
The pristine BOC and OV-BOC samples were characterized by XRD patterns in Fig. 1a. The peaks of BOC are in accordance with typical tetragonal BOC (JCPDS card No. 41-4188) and no other peaks appeared, indicating the high purity of the as-prepared Bi2O2CO3. Notably, the XRD pattern of the BOC sample shows a preferential exposed {001} crystallographic plane supported by the intensified (002), (004), and (006) peaks [28]. In addition, the main peaks of OV-BOC are nearly identical to those of BOC, but the intensities of the peaks are obviously lower, demonstrating that the crystallinity of BOC is reduced after creation of oxygen vacancies on the surface.
The XPS survey spectra of both samples showed that peaks of the three elements Bi-4f, O-1s, and C-1s can all be detected (Fig. 1b). The high resolution C-1s spectrum of BOC (Fig. 1c) reveals two peaks at 289.0 and 284.8 eV, which are assigned to the C–O bond and adventitious hydrocarbon, respectively [29]. The two apparent characteristic peaks for Bi-4f located at 159.1 and 164.5 eV are attributed to Bi-4f-7/2 and Bi-4f-5/2 in BOC (Fig. 1d) [30, 31]. The high resolution XPS spectra of Bi-4f in OV-BOC and OV-BOC-Sputter (after the 20 nm sputtering) in Fig. 1c show that the Bi-4f-7/2 and Bi-4f-5/2 peaks move in the direction of lower binding energy in comparison to those of BOC. Moreover, the peaks of O-1s at 530.3, 531.1, and 532 eV of OV-BOC and OV-BOC-Sputter also shifted to lower binding energy compared to those of BOC (Fig. 1e) [32, 33]. The shifting of Bi-4f and O-1s should be ascribed to the existence of the oxygen vacancies, which made it easier to enrich electrons and transmit electrons to the Bi and O atoms in the [Bi2O2]2+ layers. Furthermore, the etching spectra of the OV-BOC surface (Fig. 1f) reveal less O-atom exposure on the surface, which may also suggest the formation of oxygen vacancies on OV-BOC.
To confirm further the successful creation of oxygen vacancies in BOC via the addition of NaBH4, low-temperature solid-state electron paramagnetic resonance (EPR), a sensitive and efficient technique, was applied to detect the oxygen vacancies. As shown in Fig. 1f, OV-BOC shows the obvious paramagnetic resonance signal of oxygen vacancies at a g value of 1.995 in the dark, in contrast to that of pure BOC. This provides solid evidence for the existence of oxygen vacancies in OV-BOC [34].
SEM and TEM were employed to survey the morphology and micro-structure of pristine BOC and OV-BOC (Fig. S1 and Fig. 2). Based on the SEM observations (Fig. S1), it can be observed that both BOC and OV-BOC consist of large scale nanoplates that might be formed owing to the internal layered crystal structure of Bi2O2CO3 with weak Van der Waals interactions along the [001] direction [35]. Furthermore, TEM images of BOC and OV-BOC further confirm the nanoplate morphology in Fig. 2a and 2c, respectively. Interestingly, the OV-BOC reveals obviously the regular contour of the nanosheets. The HRTEM images of pristine BOC (Fig. 2b) show that the fringe spacing of 0.68 nm could match well the (002) atomic planes of tetragonal Bi2O2CO3, consistent with the results from the XRD patterns (Fig. 1a). The single-crystal characteristics and the corresponding {001} plane exposure of BOC are demonstrated by the selected-area electron diffraction (SAED) pattern in Fig. 2b. The angle between the typical planes are in line with theoretical values [36]. The HRTEM image (Fig. 2d) exhibits fringe spacing of 0.273 nm and an angle of 90°, which well match the (110) atomic planes of tetragonal Bi2O2CO3, consistent with the results from XRD patterns (Fig. 1a). On the basis of these results, the wide surface of OV-BOC can be unambiguously identified as the {001} crystallographic plane [37, 38]. Moreover, it is noteworthy that the edge of the OV-BOC nanosheet was somewhat damaged, which may demonstrate the existence of oxygen vacancies on the surface of OV-BOC [39].
The photocatalytic activity of the prepared samples was evaluated via photo-oxidation of NO in a continuous-flow reactor under visible light irradiation. As shown in Fig. 3a, the photocatalytic performance of OV-BOC (NO removal ratio of 50.2%) far exceeds that of pristine BOC (about 10.0%), which indicates that the construction of oxygen vacancies could effectively enhance the photocatalytic performance of BOC. For photocatalytic NO oxidation removal, the generation of toxic NO2 intermediates should be controlled. Notably, NO2 is almost controlled for OV-BOC, in comparison with the high fraction of NO2 generation (roughly 80.9%). This indicates that oxygen vacancies could promote the transformation of NO into target products instead of toxic byproducts (NO2). To detect further the final products after photocatalytic reaction, we washed the used catalyst after 8 h reaction using 100 mL of distilled water. The nitrate in the resultant solution was then analyzed via ion chromatography. The result reveals that the concentration of nitrate is 5.57 g/L (0.0089 mol nitrate in 100 mL water). Within 8 h, 0.0117 mol of NO was introduced into the reactor. Accordingly, the conversion ratio of NO to NO3– was 76.1%.
More specifically, the OV-BOC exhibits a broadened visible light response, as shown in the Fig. 3b, compared to that of pristine BOC. Moreover, both the photoluminescence (Fig. 3c) and photovoltage spectra (Fig. 3d) demonstrated the enhanced electron/hole separation efficiency of OV-BOC compared to that of pristine BOC. The forbidden band width Eg can be obtained by means of (αhv)1/2 of the BOC and OV-BOC samples [40]. The band gap of OV-BOC (2.53 eV) is less than BOC (3.15 eV), suggesting that the introduction of oxygen vacancies could reduce the band gap of BOC in Fig. S2a.
To probe further the change of band gap, the electrochemical Mott-Schottky measurements were used to obtain the energy-band potential of the samples (Fig. S2b). The positive slope of the C–2-V curves represents the n-type character of synthesized BOC and OV-BOC [41]. Meanwhile, the flat potential of OV-BOC and BOC obtained by extrapolation of the Mott-Schottky plot is roughly −0.79 and −0.70 V, respectively, versus the saturated calomel electrode (SCE). The latter is equivalent to −0.55 and −0.46 V, respectively, versus the normal hydrogen electrode (NHE). As far as we know, the flat band potential (quasi Fermi level) is 0.1 V lower than the conduction band minimum for n-type semiconductors [42]. Therefore, the conduction band (CB) minimum of OV-BOC and BOC is −0.65 and −0.56 V, respectively. The valence bands (VB) of OV-BOC and BOC were determined to be 1.88 and 2.59 V, respectively.
To study further the durability of the OV during the photoreaction, we tested the photo-oxidation of NO over OV-BOC in the continuous-flow reactor under visible light irradiation for 8 h (labeled OV-BOC-8h). Note that the NO removal ratio decreased from 47.0% to 21.0% (Fig. S3a). This reduction was ascribed to the decreased number of oxygen vacancies under prolonged visible light illumination, and to the accumulation of nitrates (NO3–) on the surface of the photocatalyst. The decreased visible-light response (Fig. S3b) and ESR signal (Fig. S3c) of OV-BOC-8h certify that the oxygen vacancies are partially consumed and thus lead to decreased photocatalytic activity. As shown in Fig. S3d–f, the peaks of Bi-4f-7/2 and Bi-4f-5/2 of OV-BOC-8h shifted positively compared to the peaks of OV-BOC (Fig. S3d). This indicates that the oxygen vacancies have been partially consumed via reaction with O2.
The generation of active radicals is a prerequisite to initiation of the photo-oxidation reactions [43]. The reactive oxygen species (ROS) were detected via DMPO spin-trapping ESR measuring. The electron (e–) signal intensity of OV-BOC is obviously decreased under visible light illumination in comparison to that of pristine BOC in Fig. 4a, which indicates that the photo-generated electrons (e–) are consumed to activate reactants (O2) and induce the generation of ROS (•O2–). Apparently, the signals of super oxygen (•O2–), hydroxyl radicals (•OH), and singlet oxygen (1O2) were observed under visible light (Fig. 4b–d) [44]. Interestingly, all the ROS signal intensities of OV-BOC vastly surpass those of BOC, which further demonstrates that oxygen vacancies can promote the transport of carriers and thus facilitate the generation of ROS. Moreover, the ESR signals of •O2– are weaker than those of •OH and 1O2, which may be attributed to the conversion of •O2– following the pathway •O2–→1O2 (oxidation) or •O2–→H2O2→•OH (reduction) [45, 46]. The generation of •OH on OV-BOC follows a more direct way: H2O→•OH, in which the H2O is oxidized by the photo-generated holes (h+), which is further certified via the results of the DFT method in Fig. 6. In addition, the •OH signal intensity of OV-BOC exceeds that of pristine BOC, which demonstrates that oxygen vacancies can drive the separation of electrons/holes, and thus stimulate the generation of •OH.
Hence, Fig. 5 illustrates the band structure and photocatalytic NO removal process on OV-BOC. It can be seen that the oxygen vacancies in OV-BOC could reduce the width of the band gap, increase the charge separation efficiency, and activate the reactants to generate abundant reactive oxygen species (ROS).
To understand better the generation of ROS from the activation of O2 and H2O by oxygen vacancies at the molecular level, DFT methods were utilized to simulate the adsorption activation of O2 and H2O on the oxygen vacancy sites. In Fig. 6, the yellow and blue regions represent electron depletion and accumulation in the charge difference distribution calculation, respectively. The total charge (Δq) of O2 and H2O molecules were calculated by the Bader method. As shown in Fig. 6a and 6b, the total charge (Δq) of the adsorbed O2 molecule on the surface of OV-BOC is negative. The O2 molecule is mainly located in the blue region in comparison with that of BOC, which indicates that O2 molecules are inclined to capture electrons from oxygen vacancies in OV-BOC. This further certified the enhanced activation of O2 molecules via the oxygen vacancies. This corresponds to the increasing ESR signals of •O2– and 1O2 in Fig. 6b and 6d. Most importantly, the total charge (Δq) of H2O is in the order: BOC (0.018 e) < OV-BOC (0.095 e). The H2O molecule on OV-BOC tends to occupy the oxygen vacancy sites and lose electrons to the surrounding atoms. This implies that the H2O molecules on OV-BOC are prone to lose electrons and inject them into the photogenerated holes (h+) around the oxygen vacancies. In this way, the H2O is activated to form •OH radicals as follows: H2O→•OH, which is in accordance with the obviously enhanced ESR signals of •OH in Fig. 6c. Therefore, the surface oxygen vacancies could not only function as activation sites for reactants, but could also prevent charge-carrier recombination.
The in situ DRIFTS, which can dynamically monitor the adsorbed reaction intermediates and products in a time sequence, was subsequently carried out to intuitively reveal the conversion pathways and reaction mechanism for photocatalytic NO oxidation (as shown in Fig. 7). Table S1 summarizes the possible assignments of the observed IR bands. The adsorption bands (NO) at 1767 and 1787 cm–1 can be observed in the adsorption equilibrium (Ads. Equil.) line on BOC in the dark (Fig. 7c) [47]. Simultaneously, the adsorption bands at 1057 and 1042 cm–1 appeared as well in the adsorption process (Fig. 7a) [48], which could be assigned to nitrates species. In addition, the intermediate (NO2) band at 1745 cm–1 was also found in Fig. 7c [49, 50]. Furthermore, as shown in the adsorption equilibrium (Ads. Equil.) line on OV-BOC (in Fig. 7b and 7d), we can see the adsorption bands at 1767 (NO), 1745 (NO2), and 1057 (nitrates) cm–1.
Interestingly, the IR bands of BOC and OV-BOC gradually changed under visible light irradiation. As shown in the reaction process of BOC (Fig. 7a and 7c), the NO adsorption bands at 1787 and 1767 cm–1 weaken gradually after turning on the visible light. The new bands at 2155 and 1220 cm–1 [47, 51] are observable and positively developed, and can be assigned to NO+ and nitrates, respectively. Moreover, the intensity of the bands at 1745 cm–1 (NO2) also gradually increases under visible light irradiation. The transformation process on BOC (NO→NO2) under visible light irradiation can be described by the following equations (1–8):
During the reaction process of OV-BOC (Fig. 7b and 7d), the NO adsorption bands at 1767 cm–1 are negatively developed, but the development of the NO2 band at 1745 cm–1 and nitrate bands at 1270, 1113, 1084, and 1057 cm–1 are positive [49]. In addition, the intensity of the band at 1190 cm–1 [50, 52], assigned to NO–, can be observed in Fig. 7b. Furthermore, in contrast to the BOC reaction process, the intensity of the NO2 band of OV-BOC (Fig. 7d) increases faster, but weakens after 8 min and more new nitrate bands appear at 1270 [53], 1113, and 1084 cm–1. This might be closely related to the strong adsorption activation function of oxygen vacancies, and thus should promote the generation of ROS to participate in oxidation-reduction reactions, corresponding to the results of the DMPO spin-trapping ESR measuring (Fig. 4) and DFT methods (Fig. 6). The final transformations on OV-BOC (NO→NO2→NO3–) under visible light irradiation can be described by equations (9–17) [54].
Here, we analyze further the reasons for the different reaction processes on OV-BOC and BOC. First, NO transforms into the intermediate forms NO– and NO+ on the surface of OV-BOC and BOC, respectively, under illumination, which is closely related to the surface structure. The oxygen vacancies in OV-BOC can collect photo-generated electrons and then NO could capture the electrons and transform into NO–. The NO adsorbed on BOC gives electrons to fill in holes and thus NO is converted to NO+. Notably, the N–O bond energy of several typical oxynitride and nitrogen oxygen ions follows the sequence NO+ > NO > NO– according to the molecular orbital theory. Because the bond energy of NO– is smaller, its N–O bond is more easily broken and further oxidized to form final products (nitrates) by ROS in photocatalytic reactions on OV-BOC. However, the high N–O bond energy of NO+ and insufficient generation of ROS on BOC, lead to the incomplete oxidation of NO and generation of a toxic byproduct (NO2).
In order to demonstrate in detail the enhancement of reactant (intermediate) activation via creation of oxygen vacancies via photocatalysis, DFT calculations were further carried out to analyze the adsorption energies on the active site (Fig. 8). Negative adsorption energy means the adsorption process releases heat, but the positive one is adverse. As shown in Fig. 8, the adsorption energy of NO and NO2 for both BOC and OV-BOC changes from negative to positive values, which indicates that the process needs external energy to be triggered. Apparently, the energy barrier preventing change of NO to NO2 for OV-BOC (0.6 eV) is lower than that of BOC (1.5 eV). This suggests that the existence of oxygen vacancies could reduce the energy barrier and thus promote reactant activation under illumination by visible light. Moreover, the adsorption energy of the NO2 → NO3– reaction for pristine BOC and OV-BOC tends to increase, which indicates the improved stability of NO3– on OV-BOC. The energy of the exothermic reaction process is favorable. Interestingly, the exothermic quantity of this process for OV-BOC (6.9 eV) is higher than that of pristine BOC (5.9 eV). This suggests that the NO2 → NO3– reaction of OV-BOC proceeds more easily because of the increased adsorption energy, which also corresponds well to the new nitrate bands during reaction on OV-BOC (Fig. 7b).
In comparison to that on the pristine BOC, the N–O bond length is apparently enlarged from 1.19 to 1.24 during NO2 adsorption on OV-BOC, and from 1.26 to 1.40 during NO3– adsorption on OV-BOC. This results in a weakened Coulomb interaction of gas molecules on the catalyst surface, thus leading to change in the energy adsorbed from 0.7 to 0.2 eV for NO2 and from −5.2 to −6.7 eV for NO3–. Interestingly, the N–O bond length in the NO adsorption of OV-BOC is slightly longer than that of BOC, which corresponds well to the appearance of NO+ on BOC and NO– on OV-BOC. This is because the N–O bond energy of NO+ is greater than that of NO–. These results imply that all reactants and intermediates can be more easily activated via the oxygen vacancies on OV-BOC. This is beneficial for the transformation of NO to target products rather than to toxic byproducts. Above all, the adsorption state and energy indicate that the strengthened activation of reactants and intermediates could promote the overall reaction efficiency and thus enormously boost the photocatalytic selectivity.
Bi2O2CO3 (BOC) nanosheets with oxygen vacancies were prepared successfully via a facile method. The OV-BOC displayed strengthened photocatalytic activity for NO removal compared to that of primary BOC under visible light irradiation. The oxygen vacancies could modify the band gap of BOC by forming defects at an intermediate level between the conduction band (CB) and valence band (VB). In addition, oxygen vacancies are beneficial for carrier separation and transportation, and for facilitating the activation of O2 and H2O molecules to generate abundant active species. More importantly, the oxygen vacancies can promote electron exchange between the reaction intermediates and the surface OV in OV-BOC, making them fully oxidized by the active radicals and transforming them into target products rather than toxic byproducts. This study provides a new approach for enhancing photocatalytic activity and selectivity, and also offers new insights for understanding the gas phase photocatalytic reaction mechanism.
2019,
Vol. 40 Issue (5): 620-630 DOI: S1872-2067(19)63279-1 
PDF
2019,
Vol. 40 Issue (5): 620-630 DOI: S1872-2067(19)63279-1 PDF