Oxygen vacancies enhance the photocatalytic deep oxidation of NO over an N-doped KNbO₃ catalyst

Abstract

N-Doped KNbO₃ (N-KNbO₃) was prepared through mechanical mixing, grinding and calcination of urea and KNbO₃. The photocatalytic NO oxidation performance was evaluated at room temperature and under visible light irradiation. Compared with the pristine KNbO₃, N-KNbO₃ not only has a better NO oxidation performance but also can effectively inhibit the formation of NO₂ (electron paramagnetic resonance). EPR tests showed that the doping of non-metallic element N created oxygen vacancies (OVs) on the surface of KNbO₃, which was beneficial to the formation of impurity energy levels to reduce the band gap. The in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) results indicated that NO acquired electrons around OVs to form N₂O₂⁻ intermediates, which were then oxidized by reactive oxygen species (˙OH, ˙O₂⁻ and ¹O₂). Moreover, visible light can promote the activation of NO and O₂ and also the formation of OVs. This work provided a new idea for the design and preparation of N-doped photocatalysts for NO photocatalytic oxidation.

---

1. Introduction

With the progress and development of human civilization, people have basically reached a consensus that environmental pollution (greenhouse effect, emission of toxic gases) needs to be rectified.^1–5 In addition to limiting the combustion of coal and oil from the source, the purification of pollutants at the back end is the key point. NO_x, as a common gas pollutant, is easily generated in automobile exhaust and chemical plants.^6–8 The massive discharge has caused great harm to the entire ecological environment and humans. To effectively remove NO_x, many technologies and methods have been explored and developed.^9,10 Among them, the photocatalytic oxidation of NO technology is considered to be the simplest, most effective and most environmentally friendly technology.^11,12

In recent years, perovskite oxides with unique physicochemical properties have attracted increasing attention from researchers.^13,14 KNbO₃ is a perovskite material with excellent chemical stability and has great potential in NO photocatalytic oxidation. However, KNbO₃ has a wide band gap (about 3.3 eV), which greatly limited the utilization of visible light and the efficiency of photocatalytic oxidation of NO.^15,16 It had been reported that the introduction of defects such as oxygen vacancies (OVs) can well solve the problem of low utilization of visible light by forming impurity energy levels.^17,18 Moreover, the existence of OVs was beneficial to the adsorption and activation of NO and O₂, thereby improving the photocatalytic oxidation performance of NO.^19,20

Ion doping has been used for more than 30 years as a common means of introducing OVs.^21,22 Both metal and non-metal doping usually accompanied the formation of OVs. It has been reported that X-g-C₃N₄ (X = Mg, Ca, Sr and Ba),²³ Fe–TiO₂,²⁴ and Ce–SnO₂²⁵ can successfully form OVs by introducing metal ions, thereby enhancing the oxidation performance of NO. Moreover, non-metal doping such as X-TiO₂ (X = C, N, B)^26–28 can also successfully introduce OVs into pristine samples. Among them, non-metallic N doping stands out due to its cheap experimental raw materials and simple experimental operations.

In this work, the non-metallic element N was successfully introduced into KNbO₃ through mechanical mixing, grinding and calcination of urea and KNbO₃. The catalytic NO oxidation performance was tested under simulated natural environment conditions. The results showed that the non-metallic element N doping can create new OVs on the surface of KNbO₃, which can promote the adsorption and activation of NO and O₂. Compared with pristine KNbO₃, N-KNbO₃ can generate more hydroxyl radicals, superoxide radicals and singlet oxygen under the excitation of visible light, and exhibited excellent photocatalytic activity and stability for NO oxidation. The possible mechanism of the photocatalytic oxidation of NO was revealed.

2. Experimental methods

2.1. Preparation methods

2.1.1. Synthesis of KNbO₃ catalysts. In a typical process, 33.60 g mol KOH was dissolved in 60 mL of deionized water to obtain a KOH solution (10 mol L⁻¹). After it cooled to room temperature, 1.06 g of commercial Nb₂O₅ was added to the KOH solution and stirred magnetically for 2 h. It was then transferred to a 100 mL Teflon-lined stainless-steel autoclave, heated in an oven at 180 °C for 12 h, then centrifuged and dried to obtain white KNbO₃ powder.

2.1.2. Synthesis of N-KNbO₃ catalysts. N-KNbO₃ was prepared by calcination after mechanical grinding. Specifically, 1 g KNbO₃ and a certain amount of urea were placed in a mortar and mechanically ground for 1 h, and then transferred to a crucible. The crucible was then transferred to a muffle furnace and heated to 425 °C at a heating rate of 2 °C min⁻¹, and finally calcined in air for 4 h to obtain N-KNbO₃. According to the amount of urea introduced, it is labelled X-N-KNbO₃ (X = 1.0, 1.5, 2.0), where X represents the mass ratio of urea and KNbO₃. Since 1.5-N-KNbO₃ had the best catalytic activity, N-KNbO₃ was referred to 1.5-N-KNbO₃ unless otherwise specified.

2.2. Analytical methods

X-ray diffraction (XRD) patterns of the samples were investigated on a powder X-ray diffractometer with Cu Kα₁ radiation (Bruker D8 Advance) operated at 40 kV and 40 mA. The surface morphologies of the samples were examined by scanning electron microscopy (SEM) (Hitachi S4800). SEM images were obtained on a Hitachi microscope with model SU8010. Transmission electron microscopy (TEM) and EDX mapping were carried out on a JEOL JEM-2010EX with a field emission gun at 200 kV. The Brunauer–Emmett–Teller (BET) specific surface area, pore volume and pore size were determined on a Micromeritics ASAP 2020 with nitrogen adsorption at 77 K. The optical properties of the samples were characterized using a Varian Cary 500 UV-vis diffuse reflectance spectrometer (DRS). The electron paramagnetic resonance (EPR) spectra were recorded with a Bruker A300 operating at the X-band microwave frequency (9.84 GHz). The capture agent DMPO was used to capture ˙OH and ˙O₂⁻ while TEMP was used to capture singlet oxygen (¹O₂). X-ray photoelectron spectroscopy (XPS) was performed using a Thermo ESCALAB 250XI with an Al-Kα X-ray source (1486.6 eV) with the C1s (284.8 eV) as a reference. The electrochemical characterization was carried out on an Autolab PGSTAT 204 with three electrode cells. The resulting FTO electrodes including the catalysts served as the working electrode. The platinum plate is used as the counter electrode and Ag/AgCl (saturated KCl solution) is used as the reference electrode. The photoluminescence (PL) emission spectra were obtained with an F-7000 FL spectrophotometer (Hitachi) with an excitation wavelength of 400 nm. NO₂⁻ and NO₃⁻ were detected by ion chromatography (Thermo Fisher-Dionex Aquion).

2.3. Catalytic activity evaluation

The reaction performance test of the catalytic oxidation of NO was carried out in an atmospheric continuous flow reactor, the same as our previous research.^24,25 400 mg of catalyst was charged into a self-made quartz reactor and the reaction temperature was maintained at 25 °C using a circulating water bath (Fig. S1†). High-purity air was used to provide O₂ for the reaction, NO with a concentration of 50 ppm was the reaction gas, and high-purity N₂ was the balance gas. The gas flow rate of the reaction gas NO was 20 mL min⁻¹, and the total gas flow rate of high-purity air was 80 mL min⁻¹ (the relative humidity of the reaction was controlled by adjusting the flow rate of wet air and dry air) (Fig. S2†). The light source was provided with a 300 W PLS-SXE300D xenon lamp (Beijing Perfectlight) with a visible light-reflectance filter. The light (420 nm < λ < 760 nm) can irradiate through the quartz glass reactor to the catalyst surface. The changes in NO concentration and the resulting changes in NO₂ were recorded by an NO_x analyzer (Thermo Scientific Model 42i). The calculation formulas for NO conversion and NO₂ selectivity were as follows:

NO conversion (%) = ([NO]_in − [NO]_out)/[NO]_in × 100% (1)

NO₂ selectivity (%) = [NO₂]_yield/([NO]_in − [NO]_out) × 100% (2)

2.4. In situ DRIFTS characterization for NO and O₂ adsorption

In situ DRIFTS characterization was performed on an infrared instrument (Bruker VERTEX 80). The spectra were recorded at a frequency of 64 scans in the wavenumber range (1000–4000 cm⁻¹) with a spectral resolution of 8 cm⁻¹. The sample background was collected after the samples were pretreated in Ar at 300 °C for 1 h and then cooled to room temperature. The sample was then exposed to the reaction gas to simulate actual reaction conditions and the data were collected. Visible light was introduced onto the sample surface through the ZnSe window after the adsorption was saturated.

3. Results and discussion

3.1. Catalyst characterization

3.1.1. XRD. X-ray powder diffraction can be used to analyze the crystal structure of the catalyst. Fig. 1 shows the XRD patterns of the pristine KNbO₃ and X-N-KNbO₃. The diffraction peaks 2θ = 22.05°, 31.05°, 45.07°, 50.81°, 56.03°, 65.47°, 70.41° and 74.93° are attributed to the (110), (111), (220), (221), (311), (222), (132) and (420) planes of orthogonal KNbO₃ (JCPDS no.: 71-0946).²⁹ Compared with KNbO₃, the XRD pattern of N-KNbO₃ had no obvious change, indicating that non-metallic N was uniformly distributed in the KNbO₃. Table S1† exhibits that KNbO₃ and X-N-KNbO₃ had similar lattice constants, indicating that the doping of N did not significantly destroy the crystal structure of KNbO₃. The digital photos of the samples are shown in Fig. S3.† Moreover, the specific surface area, pore volume and pore size did not change significantly after N doping (Fig. S4 and Table S2 in ESI† 6).

Fig. 1 XRD patterns of KNbO₃ and X-N-KNbO₃ (X = 1.0, 1.5, 2.0).

3.1.2. Morphology and microstructure. Fig. 2a and b are the SEM images of the KNbO₃ and N-KNbO₃. Both samples had a regular cubic structure with an edge length of about 720 nm, indicating that the doping of N did not change the KNbO₃ morphology. Fig. 2c and d exhibits the HRTEM images of KNbO₃ and N-KNbO₃. The lattice fringes with lattice spacing d = 0.407 nm were attributed to the (110) plane of KNbO₃.³⁰ Meanwhile, EDX mapping confirmed the existence of N, K, Nb, and O elements and demonstrated that the non-metallic element N was uniformly distributed on the KNbO₃ (Fig. S5†).

Fig. 2 SEM images and HRTEM images of KNbO₃ (a and c) and N-KNbO₃ (b and d). The insets in figure (c) and (d) are enlarged view of the red frame.

3.1.3. Optical absorption properties. The light absorption properties of the catalysts were tested by UV-vis diffuse reflectance absorption spectroscopy. As shown in Fig. 3, the absorption band edge of the KNbO₃ catalyst was about 390 nm, and the absorption band edge of N-KNbO₃ had a significant redshift so that the sample had obvious light absorption in the range of 400–500 nm. This was due to the fact that N doping introduced energy state N 2p above the valence band O 2p of KNbO₃, broadening the intrinsic absorption band.³¹ According to the Tauc plot empirical formula calculation (ESI† 5), the band gaps of the as-prepared KNbO₃, 1.0-N-KNbO₃, 1.5-N-KNbO₃ and 2.0-N-KNbO₃ were 3.3 eV, 2.95 eV, 2.83 eV and 2.76 eV (Fig. 3b), indicating that doping N could effectively reduce the band gap of KNbO₃.

Fig. 3 UV-vis DRS (a) and Tauc plots (b) of KNbO₃ and X-N-KNbO₃ (X = 1.0, 1.5, 2.0).

3.1.4. Photoelectrochemical performance test. The generation, separation, and transfer of photogenerated carriers was an important process in the photocatalytic reaction process, so photoelectrochemical tests were performed to study the photogenerated carrier transfer and separation capabilities of catalysts under visible light. Fig. 4a exhibits the transient photocurrent curves of KNbO₃ and N-KNbO₃ under multiple light-on–off cycles. The photocurrent of the N-KNbO₃ catalyst was significantly stronger than that of the KNbO₃ under the same conditions. The photocurrent of the KNbO₃ catalyst showed that the introduction of non-metallic element N can improve the separation efficiency of photogenerated electrons and holes.³² Moreover, Fig. 4b shows the electrochemical impedance spectra (EIS). N-KNbO₃ had a smaller radius of curvature, indicating that N-KNbO₃ had a smaller transport resistance, which was beneficial to improving the photogenerated charge carrier separation.³³

Fig. 4 Transient photocurrent (a) and electrochemical impedance spectra (b) of KNbO₃ and N-KNbO₃.

3.1.5. Fluorescence spectra. To further verify the separation ability of photogenerated carriers, fluorescence experiments were carried out. Fig. 5 shows the fluorescence spectra of KNbO₃ and N-KNbO₃. The peaks at 421 nm and 463 nm were ascribed to the strong blue emission of KNbO₃, which was due to the intrinsic radiative emission associated with the charge transfer of the oxygen atom to the central niobium atom of the NbO₆ octahedron.³⁴ Compared with the KNbO₃ catalyst, the peak intensity of N-KNbO₃ decreased to a certain extent, indicating that the doping of N was beneficial to inhibit the recombination of photogenerated carriers, which was consistent with the photocurrent and EIS experimental results.

Fig. 5 Fluorescence spectra of KNbO₃ and N-KNbO₃.

3.1.6. EPR. The OVs and reactive oxygen species (ROS) of KNbO₃ and N-KNbO₃ were further investigated by EPR characterization. As shown in Fig. 6a, the KNbO₃ had no EPR signal peak under dark conditions or under visible light while the signal peak of g = 2.004 for N-KNbO₃ can be attributed to the OVs under dark conditions (Fig. 6b). The signal peaks indicated that the introduction of non-metallic element N was beneficial to KNbO₃ to generate a large number of OVs. Meanwhile, the EPR signal peak of g = 2.004 was significantly enhanced under the irradiation of visible light, which confirmed that the irradiation of visible light was favorable for the further generation of OVs.³⁵

Fig. 6 Solid-phase EPR spectra of KNbO₃ (a) and N-KNbO₃ (b) in the dark and under visible light conditions.

Liquid-phase EPR tests were performed to characterize the ROS of the reaction. KNbO₃ and N-KNbO₃ had no EPR signal peaks of ˙OH (Fig. 7a). KNbO₃ and N-KNbO₃ showed signal peaks of ˙OH with a signal intensity ratio of 1 : 2 : 2 : 1 (numbered 1, 3, 5, and 7) under visible light, and the signal of the latter was significantly stronger than that of the former because the presence of OVs were more conducive to the production of ˙OH. ¹O₂ peaks also appeared for N-KNbO₃ with a signal intensity ratio of 1 : 1 : 1 (numbered 2, 4, and 6) under visible light, which was due to the cleavage and oxidation of the C–N bond by the reaction of DMPO and ¹O₂.³⁶ To further verify the existence of ¹O₂, TEMP was used to capture ¹O₂. As expected, ¹O₂ was successfully detected in N-KNbO₃ under visible light conditions.

Fig. 7 EPR spectra of ˙OH (a), ¹O₂ (b) and ˙O₂⁻ (c).

Fig. 7c shows the EPR spectra of ˙O₂⁻. Both KNbO₃ and N-KNbO₃ had no signal peaks under dark conditions. However, N-KNbO₃ had a significant ˙O₂⁻ signal peak under visible light, which was due to the easy acquisition of electrons by O₂ molecules around OVs.³⁷

3.2. NO oxidation removal activity test

Fig. 8a shows the five consecutive cycle activity tests of the photocatalytic NO oxidation of KNbO₃ and X-N-KNbO₃. Photocatalytic oxidation of NO over KNbO₃ had a conversion of about 20% and can maintain excellent durability. The performance of X-N-KNbO₃ for the photocatalytic oxidation of NO was significantly improved, and especially 1.5-N-KNbO₃ exhibited the best performance under visible light. Specifically, the photocatalytic oxidation conversion of NO remained around 83% during the five consecutive light–dark cycles. It is worth noting that the N doping amount of 1.5-N-KNbO₃ was about 7 at% (Table S3 in ESI† 7). However, the performance decreased with the increase of the doping amount of non-metallic element N (2.0-N-KNbO₃). It may be that the introduction of too much non-metallic element N led to the formation of more defect energy levels in the sample, which acted as the recombination center of photogenerated carriers to prevent the transfer of electron–hole pairs to the reaction molecules, thereby inhibiting the photocatalytic NO oxidation performance.³⁸

Fig. 8 NO removal curves under visible light irradiation at relative humidity (RH = 50%) (a); NO_x species distribution (b) over KNbO₃ and X-N-KNbO₃ (X = 1.0, 1.5, 2.0).

In order to further explore the selectivity of the samples, the target products (NO₂⁻ and NO₃⁻) were detected by ion chromatography (Table S4 in ESI† 8). As shown in Fig. 8b, the introduction of N effectively suppressed the generation of the toxic by-product NO₂, showing higher NO photocatalytic oxidation selectivity. In addition, the photocatalytic oxidation products of KNbO₃ and X-N-KNbO₃ were mainly nitrite. Based on the above experimental results, it can be determined that the introduction of N created more active sites in the catalyst to promote the further conversion of NO₂.

3.3. The analysis of XPS

To further explore the chemical state of the catalysts and the electron density of the surface, XPS tests were performed on the as-prepared catalysts under different treatments: the fresh sample, reacted in the dark, and reacted under visible light irradiation. As shown in Fig. 9a, two characteristic peaks at 292.80 eV and 295.50 eV were assigned to K 2p_3/2 and K 2p_1/2 of KNbO₃, respectively.³⁹ Compared with K 2p of KNbO₃, the characteristic peak of K 2p of N-KNbO₃ was shifted by 0.5 eV to the lower binding energy, indicating that the introduction of non-metallic element N was beneficial to improving the electron density of K 2p. Besides, the electron binding energy of K 2p decreased after the dark reaction and the electron binding energy was further reduced after visible light illumination, which was due to the adsorption of the target products (NO₂⁻ and NO₃⁻) on K⁺.⁴⁰

Fig. 9 High-resolution XPS spectra of K 2p (a and b) and Nb 3d (c and d,) under different treatments over KNbO₃ and N-KNbO₃ samples.

Fig. 9c and d show the high-resolution spectra of Nb 3d of KNbO₃ and N-KNbO₃, respectively. The two characteristic peaks at 206.50 eV and 209.20 eV correspond to Nb 3d_5/2 and Nb 3d_3/2 of KNbO₃, respectively. Compared with Nb 3d of KNbO₃, the characteristic peak of Nb 3d of N-KNbO₃ was shifted by 0.2 eV to the lower binding energy, indicating that electrons migrate from N atoms to Nb atoms. Moreover, the binding energy of Nb 3d in N-KNbO₃ increased by 0.3 eV after the dark reaction, which indicated that the Nb atom acted as an electron donor in the dark reaction, transferring electrons to reactive molecules, which were beneficial to the adsorption and activation of NO and O₂.⁴¹ However, the binding energy decreased by 0.25 eV under visible light irradiation, which indicated that the photogenerated electrons generated by the irradiation can migrate to the Nb atoms to compensate for the electrons on the Nb atoms. Simultaneously, the characteristic peaks corresponding to Nb 3d of KNbO₃ did not change significantly before and after the reaction, which also indicated that KNbO₃ had a poor performance for the photocatalytic oxidation of NO.

The XPS high-resolution spectra of O 1s were fitted to three characteristic peaks (O_L, O_ad, and O_OH) in Fig. 10a and b.⁴² Compared with KNbO₃, the three characteristic peak changes corresponding to O 1s of N-KNbO₃ were consistent with those of Nb, which may be due to the doping of non-metallic N producing more OVs to enrich electrons. This was consistent with the EPR results. In addition, the electronic binding energy of the three characteristic peaks of O 1s of N-KNbO₃ increases by 0.3 eV in the dark reaction, and decreases by 0.2 eV under visible light illumination, which was consistent with the changing trend of Nb 3d. The three characteristic peaks of O 1s of KNbO₃ did not change significantly under the three conditions, indicating that the O atom remained stable during the catalytic reaction.

Fig. 10 High-resolution XPS spectra of O 1s (a and b) and N 1s (c) and relative content of O_ad (d) under different treatments over KNbO₃ and N-KNbO₃ samples.

Fig. 10c shows the XPS high-resolution spectra of N 1s. The electron binding energy was located at 398.40 eV, which belongs to the characteristic peak of N.⁴³ The electron binding energy increased after the dark reaction, and the electron binding energy decreased under the irradiation of visible light. This was consistent with the changing trend of Nb 3d and O 1s, indicating that they participated as a whole in the process of photocatalytic NO oxidation.

To further explore the changes in the relative content of O_ad during the reaction process, the calculation results are shown in Fig. 10d. During the reaction process, the relative content of O_ad in KNbO₃ did not change significantly, which also reflected the poor photocatalytic oxidation performance of KNbO₃ for NO. The relative content of O_ad of N-KNbO₃ after the dark reaction was reduced by about 5% compared with that of fresh N-KNbO₃, which was due to the adsorption of the target product NO₂⁻ or NO₃⁻ on the catalyst surface. Compared with the dark reaction, the O_ad of N-KNbO₃ increased under visible light irradiation, which may be due to the partial lattice oxygen being migrated out to generate more OVs under light irradiation. In addition, compared with KNbO₃, N-KNbO₃ had a significant increase in surface chemisorbed oxygen (O_ad), which was due to the introduction of N to generate more OVs.

3.4. In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS)

To further explore the reaction mechanism of this reaction, in situ DRIFTS characterization was performed. When the reaction gas was introduced, some infrared peaks appeared and the peaks were assigned as shown in Table 1. Fig. 11a and b show the spectra of the NO + O₂ co-adsorption of KNbO₃ under dark conditions and visible light, respectively. The peaks at 3705 cm⁻¹ and 3644 cm⁻¹ belong to the terminal hydroxyl and the bridging hydroxyl groups, respectively.^44,45 With the extension of the adsorption time, these two peaks were continuously consumed. Besides, infrared peaks at 1648 cm⁻¹, 1555 cm⁻¹, 1535 cm⁻¹, 1436 cm⁻¹, 1336 cm⁻¹ and 1254 cm⁻¹ were generated, which were attributed to the bending vibration peak of H₂O, bidentate nitrate (K⁺-NO₃⁻), bridge bidentate nitrite (K⁺-NO₂⁻), monodentate nitrite (K⁺-NO₂⁻), bridge bidentate nitrate (Nb⁵⁺-NO₃⁻), and bridge bidentate nitrite (Nb⁵⁺-NO₂⁻).⁴⁶ This indicated that NO mainly interacted with hydroxyl groups on the catalyst surface to form nitrate and nitrite under dark reaction conditions. When the adsorption was saturated, visible light was introduced. The bridge bidentate nitrate peaks and the bridge bidentate nitrite peaks at 1336 cm⁻¹ and 1254 cm⁻¹ increased continuously, indicating that visible light is beneficial to promoting the photocatalytic oxidation of NO.⁴⁸

Table 1 In situ DRIFTS peak assignments of KNbO₃ and N-KNbO₃

Wavenumber (cm^−1)	Assignment	Ref.
3705	–OH (terminal)	44
3644	–OH (bridging)	45
3540	NOH	48
1648	H2O (bending vibration)	24
1555	K^+-NO3^− (bidentate nitrates)	47
1535,1507	K^+-NO2^− (bridging bidentate nitrites)	43
1436	K^+-NO2^− (mondentate nitrites)	40
1348,1336	Nb^5+-NO3^− (bridging bidentate nitrates)	47
1260,1254	Nb^5+-NO2^− (bridging bidentate nitrites)	48
1224	Nb^5+-NO3^− (bridged bidentate nitrates)	49
1080	trans-N2O2^−	50

---

Fig. 11 In situ DRIFTS spectra of adsorbing NO + O₂ and the visible light reaction process over KNbO₃ samples (a and b) (D is equal to dark; L is equal to light).

Fig. 12 exhibits the in situ DRIFTS of the NO + O₂ co-adsorption of N-KNbO₃. The reactive molecules NO + O₂ were introduced under dark conditions. With the extension of the adsorption time, the terminal hydroxyl group at 3705 cm⁻¹ and the N–OH at 3540 cm⁻¹ decreased continuously.⁴⁸ The H₂O at 1648 cm⁻¹ and the bidentate nitrate (K⁺-NO₃⁻) at 1555 cm⁻¹, bridge bidentate nitrite (K⁺-NO₂⁻) at 1507 cm⁻¹, bridge bidentate nitrite (Nb⁵⁺-NO₂⁻) at 1348 cm⁻¹ and bridge bidentate nitrates (Nb⁵⁺-NO₃⁻) at 1260 cm⁻¹ were formed.⁴⁹ This was due to the activation of NO by hydroxyl groups on the catalyst surface, resulting in the formation of nitrate and nitrite accompanied by the generation of H₂O. The trans-N₂O₂⁻ at 1080 cm⁻¹ also decreased with the prolongation of the reaction time, which was due to the reaction with hydroxyl groups to generate nitrite (1260 cm⁻¹).⁵⁰ The peak at 1348 cm⁻¹ attributed to bridge bidentate nitrate decreased continuously until it disappeared, which was due to the conversion of nitrate to nitrite by the reaction gas NO (NO₃⁻ + NO → NO₂⁻).

Fig. 12 In situ DRIFTS spectra of adsorbing NO + O₂ and the visible light reaction process over N-KNbO₃ samples (a and b) (D is equal to dark; L is equal to light).

When the adsorption was saturated and visible light was introduced, the terminal hydroxyl group (3705 cm⁻¹) and trans-N₂O₂⁻ (1080 cm⁻¹) were continuously consumed, and the 1260 cm⁻¹ peak attributed to bridge bidentate nitrite was continuously increased. A new peak appears at 1224 cm⁻¹, which was attributed to bridge bidentate nitrates (Nb⁵⁺-NO₃⁻). This was due to the fact that visible light can promote the conversion of surface hydroxyl groups, H₂O and O₂ into ROS: ˙OH, ˙O₂⁻ and ¹O₂, thereby promoting the deconversion of NO and N₂O₂⁻ into nitrate and nitrite. Moreover, the N–OH at 3540 cm⁻¹ decreased continuously under visible light illumination and showed a slight blue shift, which were due to the adsorption of NO on OVs and the interaction with surface hydroxyl groups. A large number of ROS further reacted with N–OH to generate nitrate or nitrite under light conditions.

3.5. Mechanism of NO removal over N-KNbO₃

1. The adsorption process of NO over N-KNbO₃ in the dark. (1) The NO reacted with surface hydroxyl groups:

NO-K⁺ + 2OH⁻ (3705 cm⁻¹) → K⁺-NO₂⁻ (1507 cm⁻¹) + H₂O (1655 cm⁻¹) (1-1)

NO-K⁺ + 2OH⁻ (3644 cm⁻¹) → K⁺-NO₃⁻ (1555 cm⁻¹) + H₂O (1-2)

NO-Nb⁵⁺ + 2OH⁻ (3705 cm⁻¹) → Nb⁵⁺-NO₂⁻ (1260 cm⁻¹) + H₂O (1-3)

NO-Nb⁵⁺ + 2OH⁻ (3644 cm⁻¹) → Nb⁵⁺-NO₃⁻ (1348 cm⁻¹) + H₂O (1-4)

(2) A certain amount of bridge bidentate nitrate continues to interact with NO to obtain bridge bidentate nitrite:

Nb⁵⁺-NO₃⁻ (1348 cm⁻¹) + NO → Nb⁵⁺-NO₂⁻ (1260 cm⁻¹) (1-5)

(3) NO gained electrons at the OVs to form an intermediate species N₂O₄²⁻

2NO + e⁻ → N₂O₂⁻ (1080 cm⁻¹) (1-6)

2. The photocatalytic oxidation of NO over N-KNbO₃. (1) Photogenerated electrons and holes were generated under visible light:

N-KNbO₃ + hν → h⁺ + e⁻ (2-1)

(2) The photogenerated holes reacted with surface hydroxyl groups and water to generate hydroxyl radicals:

h⁺ + OH⁻ → ˙OH (2-2)

h⁺ + H₂O → ˙OH + H⁺ (2-3)

(3) The existence of OVs was conducive to the easier adsorption of O₂, thereby capturing photogenerated electrons to generate superoxide radicals, and some superoxide radicals continue to react with holes to generate singlet oxygen:

O₂ + e⁻ → ˙O₂⁻ (2-4)

˙O₂⁻ + h⁺→¹O₂ (2-5)

(4) Visible light irradiation enriched the electron density on the sample surface to promote the generation of OVs, which was the reason for the good stability of the samples:

N-KNbO₃ + e⁻ → N-KNbO₃-OVs (2-6)

(5) The generated ROS (˙OH, ˙O₂⁻ and ¹O₂) further interacted with NO or intermediate N₂O₂⁻ to generate nitrate or nitrite:

NO-Nb⁵⁺ + ˙OH/¹O₂ → Nb⁵⁺-NO₃⁻ (1224 cm⁻¹) + H (2-7)

NO-Nb⁵⁺ + ˙O₂⁻ → Nb⁵⁺-NO₃⁻ (1224 cm⁻¹) (2-8)

N₂O₂⁻ + ˙OH/¹O₂ → 2Nb⁵⁺-NO₂⁻ (1260 cm⁻¹) + H⁺ (2-9)

N₂O₂⁻ + ˙O₂⁻ → 2Nb⁵⁺-NO₂⁻ (1260 cm⁻¹) (2-10)

(6) Some NO₂ reacted with ROS to form the target product nitrate or nitrite

NO₂-Nb⁵⁺ + ˙OH/¹O₂ → Nb⁵⁺-NO₃⁻ (1224 cm⁻¹) + H⁺ (2-11)

NO₂-Nb⁵⁺ + ˙O₂⁻ → Nb⁵⁺-NO₃⁻ (1224 cm⁻¹) (2-12)

NO₂-K⁺ + ˙OH/¹O₂ → K⁺-NO₃⁻ (1555 cm⁻¹) + H⁺ (2-13)

NO₂-K⁺ + ˙O₂⁻ → K⁺-NO₃⁻ (1555 cm⁻¹) (2-14)

The proposed process of the NO + O₂ reaction over N-KNbO₃ under visible light irradiation is described in Fig. 13. NO was adsorbed & activated at OVs and then the formed N₂O₂⁻ intermediates reacted with the ROS induced by visible light irradiation into NO₂⁻ and NO₃⁻. It is worth noting that the ability to generate free radicals (O₂⁻, ¹O₂) was a key factor for the enhanced N-KNbO₃ activity.

Fig. 13 The proposed process of the NO + O₂ reaction over N-KNbO₃ under visible light irradiation. NO was adsorbed & activated at OVs and then the formed N₂O₂⁻ intermediates reacted with the ROS induced by visible light irradiation into NO₂⁻ and NO₃⁻.

Furthermore, the energy band structures of KNbO₃ and N-KNbO₃ (Fig. S6 and S7†) and the photocatalytic oxidation mechanism of NO on KNbO₃ (ESI† 10) are provided in the ESI.†

4. Conclusions

A series of N-KNbO₃ was prepared by mechanical mixing, grinding and calcination of urea and KNbO₃, and their photocatalytic NO oxidation performance was tested under simulated natural environment conditions. Combined with the activity and a series of characterization, the relevant mechanism of NO photocatalytic oxidation was explored. The conclusions were as follows:

(I) Compared with KNbO₃, the photocatalytic NO oxidation performance of N-KNbO₃ was significantly improved. It had good stability in five cycles and the production of toxic by-product NO₂ decreased.

(II) The introduction of the non-metal element N can create new oxygen vacancies in the catalyst KNbO₃ to promote the adsorption and activation of the reaction molecule O₂. Besides, the impurity energy level formed by OVs can narrow the band gap and promote the effective utilization of visible light.

(III) The method of mechanically grinding urea with samples followed by calcination is a simple and effective method for the N-doping of samples.

This work provided a new idea for the design and preparation of N-doped photocatalysts for NO photocatalytic oxidation.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (no. 21872030) and the Key Program of Qingyuan Innovation Laboratory (Grant No. 00121001).

Notes and references

1. L. Han, S. Cai, M. Gao, J. Y. Hasegawa, P. Wang, J. Zhang, L. Shi and D. Zhang, Chem. Rev., 2019, 119, 10916–10976.
2. J. Zhao, Y. Xia, M. Li, S. Li, W. Li and S. Zhang, Environ. Sci. Technol., 2016, 50, 8705–8712.
3. Y. Duan, Y. Wang, L. Gan, J. Meng, Y. Feng, K. Wang, K. Zhou, C. Wang, X. Han and X. Zhou, Adv. Energy Mater., 2021, 11, 2004001.
4. J. Li, Z. Zhang, W. Cui, H. Wang, W. Cen, G. Johnson, G. Jiang, S. Zhang and F. Dong, ACS Catal., 2018, 8, 8376–8385.
5. L. Wang, M. Xia, H. Wang, K. Huang, C. Qian, C. T. Maravelias and G. A. Ozin, Joule, 2018, 2, 1055–1074.
6. Y. Zhang, Z. Liang, G. Zhang, J. Liu, Y. Wang, Y. Zhao, G. Li and Y. Lv, Catal. Sci. Technol., 2022, 12, 4871–4883.
7. Q. Yang, X. Wang, X. Wang, Q. Li, L. Li, W. Yang, X. Chu, H. Liu, J. Men, Y. Peng, Y. Ma and J. Li, ACS Catal., 2021, 11, 14507–14520.
8. K. Kobayashi, R. Atsumi, Y. Manaka, H. Matsumoto and T. Nanba, Catal. Sci. Technol., 2019, 9, 2898–2905.
9. A. Filtschew, P. Beato, S. B. Rasmussen and C. Hess, Phys. Chem. Chem. Phys., 2021, 23, 1874–1887.
10. L. Han, S. Cai, M. Gao, J. Y. Hasegawa, P. Wang, J. Zhang, L. Shi and D. Zhang, Chem. Rev., 2019, 119, 10916–10976.
11. W. Dong, W. Zhang, Y. Cui, H. Sun, Z. Wu Huang and F. Dong, Catal. Sci. Technol., 2017, 7, 1324–1332.
12. H. Wang, Y. Sun, G. Jiang, Y. Zhang, H. Huang, Z. Wu, S. C. Lee and F. Dong, Environ. Sci. Technol., 2018, 52, 1479–1487.
13. C. Huang, Y. Zhu, X. Wang, X. Liu, J. Wang and T. Zhang, J. Catal., 2017, 347, 9–20.
14. Y. Huang, Y. Liang, Y. Rao, D. Zhu, J. J. Cao, Z. Shen, W. Ho and S. C. Lee, Environ. Sci. Technol., 2017, 51, 2924–2933.
15. J. He, F. Gao, H. Wang, F. Liu, J. Lin, B. Wang, C. Liu, F. Huang, Z. Lin and M. Wang, Environ. Sci.: Nano, 2022, 9, 1952–1960.
16. W. Tang, X. Xue, C. Yuan, J. Chen, L. Miao, Q. Feng, J. Xu, G. Rao, J. Wang, C. Zhou and Y. Guo, J. Mater. Chem. A, 2022, 10, 7238–7250.
17. A. Sarkar and G. G. Khan, Nanoscale, 2019, 11, 3414–3444.
18. Q. Ji, L. Bi, J. Zhang, H. Cao and X. S. Zhao, Energy Environ. Sci., 2020, 13, 1408–1428.
19. X. Pan, M. Q. Yang, X. Fu, N. Zhang and Y. J. Xu, Nanoscale, 2013, 5, 3601–3614.
20. G. Cheng, X. Tan, X. Song, X. Chen, W. Dai, R. Yuan and X. Fu, Appl. Catal., B, 2019, 251, 130–142.
21. H. Shao, X. Bai, H. Cui, G. Pan, P. Jing, S. Qu, J. Zhu, Y. Zhai, B. Dong and H. Song, Nanoscale, 2018, 10, 1023–1029.
22. M. Hu, S. Du, B. Dong, Y. Qi, Z. Feng and F. Zhang, J. Mater. Chem. A, 2022, 10, 16541–16546.
23. M. Zhou, G. Dong, J. Ma, F. Dong, C. Wang and J. Sun, Appl. Catal., B, 2020, 273, 119007.
24. G. Cheng, X. Liu, X. Song, X. Chen, W. Dai, R. Yuan and X. Fu, Appl. Catal., B, 2020, 277, 119196.
25. X. Song, G. Qin, G. Cheng, W. Jiang, X. Chen, W. Dai and X. Fu, Appl. Catal., B, 2021, 284, 119761.
26. F. Dong, S. Guo, H. Wang, X. Li and Z. Wu, J. Phys. Chem. C, 2011, 115, 13285–13292.
27. K. Wang, T. Peng, Z. Wang, H. Wang, X. Chen, W. Dai and X. Fu, Appl. Catal., B, 2019, 250, 89–98.
28. X. Wang, K. Wang, H. Wang, Z. Wang, X. Chen, W. Dai and X. Fu, Phys. Chem. Chem. Phys., 2021, 23, 186–195.
29. R. Wang, Y. Zhu, Y. Qiu, C.-F. Leung, J. He, G. Liu and T.-C. Lau, Chem. Eng. J., 2013, 226, 123–130.
30. D. Yu, Z. Liu, J. Zhang, S. Li, Z. Zhao, L. Zhu, W. Liu, Y. Lin, H. Liu and Z. Zhang, Nano Energy, 2019, 58, 695–705.
31. A. I. Kontos, A. G. Kontos, Y. S. Raptis and P. Falaras, Phys. Status Solidi RRL, 2008, 2, 83–85.
32. S. S. Wong, M. J. Hülsey, H. An and N. Yan, Catal. Sci. Technol., 2022, 12, 5217–5228.
33. A. Liu, J. Hu, J. He, X. Huang, N. Hu, Y. Li, Q. Huang, S. Guo, X. Liu, Z. Yang, Y. Zhou and Z. Zou, Catal. Sci. Technol., 2021, 11, 7963–7972.
34. J. Yu, Z. Chen, L. Zeng, Y. Ma, Z. Feng, Y. Wu, H. Lin, L. Zhao and Y. He, Sol. Energy Mater. Sol. Cells, 2018, 179, 45–56.
35. X. Tan, G. Cheng, X. Song, X. Chen, W. Dai and X. Fu, Catal. Sci. Technol., 2019, 9, 3637–3646.
36. Z. Wang, X. Wang, H. Wang, X. Chen, W. Dai and X. Fu, Appl. Catal., B, 2020, 277, 119169.
37. H. Shang, S. Huang, H. Li, M. Li, S. Zhao, J. Wang, Z. Ai and L. Zhang, Chem. Eng. J., 2020, 386, 124047.
38. J. He, J. Ye, K. Ge, J. Cao, C. Fu, Z. Li, Y. Zhang and Y. Yang, Catal. Sci. Technol., 2020, 10, 4809–4819.
39. J. Xu, C. Liu, J. Niu and M. Chen, Sep. Purif. Technol., 2020, 230, 115861.
40. X. Song, W. Jiang, Z. Cai, X. Yue, X. Chen, W. Dai and X. Fu, Chem. Eng. J., 2022, 444, 136709.
41. Z. Wang, X. Ming, X. Wang, H. Wang, X. Chen, W. Dai, Y. Yan and X. Fu, Appl. Surf. Sci., 2022, 592, 153246.
42. K. Zhao, J. Qi, H. Yin, Z. Wang, S. Zhao, X. Ma, J. Wan, L. Chang, Y. Gao, R. Yu and Z. Tang, J. Mater. Chem. A, 2015, 3, 20465–20470.
43. X. Tan, G. Qin, G. Cheng, X. Song, X. Chen, W. Dai and X. Fu, Catal. Sci. Technol., 2020, 10, 6923.
44. K. Yang, Y. Li, K. Huang, X. Chen, X. Fu and W. Dai, Int. J. Hydrogen Energy, 2014, 39, 18312–18325.
45. Z. Wang, Y. Huang, M. Chen, X. Shi, Y. Zhang, J. Cao, W. Ho and S. C. Lee, ACS Appl. Mater. Interfaces, 2019, 11, 10651–10662.
46. M. Kantcheva, J. Catal., 2004, 223, 352–363.
47. W. C. Huo, X. A. Dong, J. Y. Li, M. Liu, X. Y. Liu, Y. X. Zhang and F. Dong, Chem. Eng. J., 2019, 361, 129–138.
48. R. V. Mikhaylov, A. A. Lisachenko, B. N. Shelimov, V. B. Kazansky, G. Martra and S. Coluccia, J. Phys. Chem. C, 2013, 117, 10345–10352.
49. W. Zhang, X. Shi, Z. Yan, Y. Shan, Y. Zhu, Y. Yu and H. He, ACS Catal., 2021, 11, 9825–9836.
50. C. Deng, J. Qian, C. Yu, Y. Yi, P. Zhang, W. Li, L. Dong, B. Li and M. Fan, RSC Adv., 2016, 6, 113630–113647.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2cy01656g

---