N-Rich Doped Anatase TiO2 with Smart Defect Engineering as Efficient Photocatalysts for Acetaldehyde Degradation

Nitrogen (N) doping is an effective strategy for improving the solar-driven photocatalytic performance of anatase TiO2, but controllable methods for nitrogen-rich doping and associated defect engineering are highly desired. In this work, N-rich doped anatase TiO2 nanoparticles (4.2 at%) were successfully prepared via high-temperature nitridation based on thermally stable H3PO4-modified TiO2. Subsequently, the associated deep-energy-level defects such as oxygen vacancies and Ti3+ were successfully healed by smart photo-Fenton oxidation treatment. Under visible-light irradiation, the healed N-doped TiO2 exhibited a ~2-times higher activity of gas-phase acetaldehyde degradation than the non-treated one and even better than standard P25 TiO2 under UV-visible-light irradiation. The exceptional performance is attributed to the extended spectral response range from N-rich doping, the enhanced charge separation from hole capturing by N-doped species, and the healed defect levels with the proper thermodynamic ability for facilitating O2 reduction, depending on the results of ∙O2− radicals and defect measurement by electron spin resonance, X-ray photoelectron spectroscopy, atmosphere-controlled surface photovoltage spectra, etc. This work provides an easy and efficient strategy for the preparation of high-performance solar-driven TiO2 photocatalysts.


Introduction
Wide bandgap oxide photocatalysts (such as classical TiO 2 ) have displayed all-round activities because of the adequate potentials of the conduction and valence bands for redox [1][2][3][4][5][6]. However, their low solar-spectrum utilization (<5%), one of the most critical bottlenecks for practical application, has been identified [7]. Because O 2p primarily contributes to the valance band of a wide bandgap oxide, it is feasible to narrow the bandgap by elevating the valence band by doping anions with higher potential energy to increase light utilization [8][9][10]. Since oxygen and nitrogen have similar properties such as polarizability, electronegativity, and ionic radii, nitrogen (N) doping is perfectible in this aspect [11,12]. Meanwhile, N with long-pair electrons facilitates trapping holes for enhancing charge separation [13]. To date, extensive N doping techniques for improving photoactivities have been established. Most studies, on the other hand, obtained lowconcentration N doping results, which are mainly attributed to the high metal-oxygen bonding energy (for example, Ti-O = 346.4 kJ/mol) and imperfect methodologies [14][15][16]. The development of efficient strategies for preparing high-efficiency N-rich doping oxides is still a work in progress.
Among the developed strategies, nitridation in an ammonia atmosphere has been demonstrated to be the most efficient method for N doping [17]. To overcome the barrier of breaking the metal-oxide bonds, a high temperature is required for N-rich doping in general. The phase transformation of oxides, however, normally occurs at a high temperature. For example, more than 600 • C is required to prepare the N-rich doping anatase TiO 2 , but such a high temperature will cause phase transformation from the high-active anatase to the inferior-rutile one [18,19]. As a result, the thermal stability of anatase TiO 2 is crucial for Nrich doping at high temperatures. Fortunately, the strategy of phosphoric acid modification was previously developed to improve the thermal stability of anatase TiO 2 as high as 750 • C, making phosphate-modified anatase TiO 2 (PTO) an available substrate to prepare N-rich doped TiO 2 at a high temperature [20].
The charge difference between lattice O 2− and dopant N 3− in N-TiO 2 inevitably introduces some associated defects, such as oxygen vacancies (O v ) and Ti 3+ defects [21,22]. O v at shallow energy levels are usually demonstrated to trap the photogenerated electrons, which enhance charge separation and thermodynamically favor O 2 activation [23,24]. O v or the relative Ti 3+ at deep energy levels from a high N doping concentration can be either advantageous or disadvantageous, depending on the situation. In some cases, O v at deep states are intentionally pursued in cases of self-doped TiO 2, when the optical absorption at wavelengths is longer than 400 nm [25,26]. However, they can also have a net disadvantageous effect when they are not exploited by optical absorption, so that their overall effect as recombination centers is to hamper the charge mobility and thermodynamically reduce O 2 , resulting in decreased photocatalytic activity [27,28]. Usually, a high temperature promotes N-rich doping, but it also produces numerous O v and Ti 3+ defects at deep energy levels. Liu's group recently prepared a boron-doped TiO 2 as a precursor to achieve an N-rich doping by using boron ions to balance the charge difference so as to avoid Ti 3+ at deep energy levels [29]. However, TiB 2 as a raw material is unusual for a versatile application. Developing a controllable method for healing the adverse O v at deep energy levels for Nrich doped TiO 2 is of great significance to the high photocatalytic activity. As for the essence of O v at deep energy levels, Ti 3+ defects are typical and important species. Oxidizing Ti 3+ defects into Ti 4+ is feasible to heal O v at deep energy levels. It has been reported that Ti 3+ is an active site to achieve a photo-Fenton-treated reaction, resulting in the oxidation of Ti 3+ defects into Ti 4+ [30]. This interesting reaction is of great importance and opens our minds to controllably heal O v without introducing impurities. To date, there is no work referring to the utilization of the photo-Fenton-treated reaction to control the vacancies in N-TiO 2 .
In this work, N-rich doped anatase TiO 2 nanoparticles (4.2 at%) were successfully fabricated via high-temperature nitridation based on thermally stable H 3 PO 4 -modified TiO 2 . Subsequently, the associated deep-energy-level defects such as O v and Ti 3+ were successfully healed by a smart photo-Fenton-treated oxidation strategy. Under visible-light irradiation, the healed N-doped TiO 2 exhibited a~2-times higher activity of gas-phase acetaldehyde degradation than the non-treated one, and even better than the standard P25 TiO 2 under UV-visible-light irradiation. The high activity was attributed to the extended light response due to N-rich doping, the enhanced charge separation from hole capturing by N-doped species, and remaining O v after healing of deep-energy-evel defects with a proper thermodynamic ability for facilitating O 2 reduction. This work uncovers the N-rich doping in TiO 2 -photocatalyst-produced defects at various energy levels and their influence on charge separation and O 2 activation, as well as an efficient technique for engineering the defects for high-performance solar-driven TiO 2 photocatalysts.

Synthesis of H 3 PO 4 -Modified Anatase TiO 2 Nanoparticles
Anatase TiO 2 nanoparticles were prepared by a common sol-hydrothermal method [20]. The typical procedures are described in the Supporting Information. For H 3 PO 4 modification, 0.5 g of as-prepared anatase TiO 2 nanoparticles was dispersed in 30 mL of H 3 PO 4 solution (0.1 M) under continuous stirring for 4 h. The precipitate was then centrifuged and washed once with absolute ethanol before being dried at 80 • C for 10 h and calcined at 450 • C for 1 h to yield H 3 PO 4 -modified anatase nanoparticles, denoted as PTO.

Synthesis of Nitrogen-Doped PTO
Nitrogen-doped PTO was prepared by heating PTO with pure NH 3 . In a typical experiment, 0.2 g of PTO was put into a quartz boat in a tubular furnace, and then, the temperature was elevated to 500-650 • C in steps of 5 • C/min in a NH 3 flow (50 mL/min). PTO was treated with NH 3 for 2 h before naturally cooling down to room temperature. The samples are denoted as NPTO-T, where -T represents the treatment temperature in ammonia. As a reference sample, NTO-500 was prepared by the treatment of TO with a pure NH 3 atmosphere at 500 • C.

Defects Healing of NPTO
Photo-Fenton-treated healing process: 0.2 g of NPTO-650 was placed in 30 mL of 30% H 2 O 2 solution and stirred under 300 W Xenon-light irradiation for 1-3 h at room temperature. The sample is denoted as R L-H 2 O 2 -t-NPTO-650, where -t represents the treatment time. R H 2 O 2 -t-NPTO-650 as reference samples were prepared by the same treatment process, except under dark condition.

Oxidation Healing Process
There was 0.2 g of NPTO-650 put into a quartz boat in a tubular furnace, and then, the temperature was elevated to 300-450 • C in steps of 5 • C/min in an O 2 flow (50 mL/min). NPTO-650 was treated with O 2 for 0.5-1.5 h before cooling down to room temperature naturally. The samples are denoted as R O 2 -t-T-NPTO-650, where -T-and -t represent the treatment temperature and time, respectively.

Photocatalytic Activity Evaluation
The photocatalytic activity was evaluated by gas-phase acetaldehyde degradation. In a typical process, 0.1 g of catalyst was uniformly coated on the surface of a 5 cm × 5 cm glass sheet, which was placed in a self-made stainless steel cell (Volume: 1 L) with a quartz window. Then, a gas mixture (800 ppm acetaldehyde, 20% O 2 , and about 80% N 2 ) was continuously passed through the reactor for 2 h, reaching adsorption equilibrium in the dark. The acetaldehyde concentration was determined with a gas chromatograph (Shanghai Kechuang, GC-2002, Shanghai, China) equipped with a flame ionization detector after reaction under different light irradiation (visible light: 150 W high-pressure Xenon lamp with a 420 nm cutoff filter; the incident integrated irradiance of the lamp on the sample surface was 58 mW/cm 2 ; single-wavelength light: 100W LED light with different wavelengths; the incident integrated irradiance of the lamp on the sample surface was 20 mW/cm 2 ). The power spectrum vs. wavelength of the Xe lamp and LED lamp can be seen in Figure S1.The cycle test was performed in the same way, and the reactor and photocatalyst were purged by pure N 2 before the repeated experiment.

Characterizations
The powder X-ray diffraction (XRD) analyses were recorded with a Bruker D8 Advance diffractometer (CuKα radiation, Billerica, MA, USA). The morphologies and element distribution of the photocatalysts were obtained by a scanning electron microscope (SEM, SIGMA-500, ZEISS, Jena, Germany) and transmission electron microscope (TEM, JEM-F200, JEOL, Showa, Tokyo, Japan) with energy dispersive X-ray spectroscopy (EDS). The UV-visible diffuse reflectance spectra (DRS) of the photocatalysts were investigated in the range of 220-800 nm by the Shimadzu UV-2700 model spectrophotometer (Suzhou, China). Fourier transform infrared (FT-IR) spectroscopy was performed using a Thermo Scientific Nicolet iS50 FT-IR spectrometer (Waltham, MA, USA). The Raman spectra were measured with a Renishaw inVia Confocal Raman spectrometer (Loucestershire, London, UK). X-ray photoelectron spectra (XPS) surveys were tested on a Kratos-AXIS ULTRA DLD (Manchester, UK), Aluminum (Mono) spectrometer. The electron spin resonance (ESR) and DMPO spin-trapping ESR analyses were carried out on a Bruker EMXPLUS model spectrometer (Billerica, MA, USA) at ambient temperature. The surface photovoltage spectra (SPS) were implemented on home-built equipment with a lock-in amplifier (SR830, Silicon Valley, CA, USA) synchronized with a light chopper (SR540, Silicon Valley, CA, USA).

N-Rich Doping
Anatase TiO 2 (TO) and H 3 PO 4 -modified TiO 2 (PTO) were synthesized by using a common sol-hydrothermal method as described in our previous work [20]. The XRD patterns ( Figure S2a), SEM images ( Figure S2b), and DRS ( Figure S2c) illustrate that there was no change in the crystal phase (anatase), particle sizes (5~10 nm), and bandgap (3.2 eV) of TO before and after phosphoric acid modification. FT-IR spectra ( Figure S2d) showed a new absorption peak at 1010~1250 cm −1 in PTO, which can be attributed to PO 4 3− . As shown in Figure S1e, XPS confirmed the phosphate from the P 2p peak at 133.6 eV [31,32]. In addition, the SPS response of PTO in Figure S1f shows a higher signal than TO, indicating that H 3 PO 4 modification can improve the charge separation of TO. This is because the appropriate modified PO 4 3− can promote the oxygen adsorption and subsequently capture photogenerated electrons [33]. Based on the XRD patterns ( Figure S3a,b), the phase transformation from anatase to rutile appeared in TO after thermal treatment at 550 • C and was completely converted into the rutile phase until 650 • C. Compared with TO, PTO retained a pure anatase phase after thermal treatment at 650 • C, displaying high thermal stability. This was further confirmed by the DRS. The absorption edge of TO exhibited a red shift after 550 • C treatment due to the formation of rutile TiO 2 ( Figure S2c), but no change of PTO could be detected even after 650 • C. PTO with high thermal stability provides the feasibility of N-rich doped anatase TiO 2 .
Subsequently, the effects of temperature on the as-prepared N-doped PTO (NPTO-T) were investigated. The XRD patterns of NPTO-T and NTO-500 in Figure 1a reveal that both phases of NTO-500 and NPTO-T were anatase. TEM image (Figure 1b) illustrated the particle size of NPTO-650 to be approximately 5-10 nm. The anatase phase was further confirmed by the HR-TEM image (inset Figure 1b) according to the interplanar distances of (004) and (101) planes of anatase TiO 2 [34]. Moreover, the morphology of the as-prepared samples was characterized by SEM images ( Figure S4). There was no obvious difference between NTO-500 and NPTO-T. NTO-650 exhibited an obvious change in the dramatically increased particle size because of the growth of grains during the crystal-phase transition. The DRS in Figure 1c shows that NTO-500 only exhibited a weak shoulder-like visible-light absorption because of the comparatively low content of doped N at 500 • C. However, the light absorption of NPTO-T obviously enhanced with the increasing nitridation temperature. The bandgap edge of NPTO-650 surprisingly shifted to 700 nm. The corresponding photos in Figure 1c display the color change of samples from white (PTO) to pale yellow (NPTO-500), and finally, dark red (NPTO-650).
The N-doped species in the TiO 2 lattice were further studied by XPS. In the N 1s spectra (Figure 1d), all the samples presented a peak at 399.6 eV, which was ascribed to the interstitial N-like Ti-O-N [35]. The intensity increased along with the rising temperature, indicating the increased N content. Noticeably, an additional peak appeared at 395.6 eV in NPTO-600 and NPTO-650, which was ascribed to the substituted N for lattice oxygen by forming O-Ti-N bonds [36]. The content of substituted N in NPTO-650 was higher than that in NPTO-600. In addition, the XPS valence band spectra ( Figure S5a  The chemical state of Ti in different N-doped TiO2 was studied by XPS and electron spin resonance (ESR). In Figure 1e, NTO-500 has two peaks at 464.4 and 458.7 eV, respectively, which correspond to the typical Ti 4+ -O bond [37]. In comparison to NTO-500, there was a noticeable red-shift of Ti 2p in NPTO-T samples, which gradually improved with increasing temperature. NPTO-650 exhibited a new binding energy peak at 457.6 eV, which was ascribed to the Ti 3+ species [38]. By analyzing the previous literature [29,39], ESR (Figure 1f) consistently showed sharp signals at g = 1.996 in all The chemical state of Ti in different N-doped TiO 2 was studied by XPS and electron spin resonance (ESR). In Figure 1e, NTO-500 has two peaks at 464.4 and 458.7 eV, respectively, which correspond to the typical Ti 4+ -O bond [37]. In comparison to NTO-500, there was a noticeable red-shift of Ti 2p in NPTO-T samples, which gradually improved with increasing temperature. NPTO-650 exhibited a new binding energy peak at 457.6 eV, which was ascribed to the Ti 3+ species [38]. By analyzing the previous literature [29,39], ESR ( Figure 1f) consistently showed sharp signals at g = 1.996 in all samples, implying the defects associated with the N doping amount. The additional weak signal at g = 1.974 was attributed to the Ti 3+ defects in NPTO-650, which coincided with the XPS [29].
To clarify the effect of N doping and the associated defects on the photocatalytic performance, the photocatalytic activities of the as-prepared N-doped TiO 2 were tested for acetaldehyde degradation under visible-light irradiation (Figure 2a). All of the NPTO-T samples exhibited higher activities than NTO-500, while NPTO-600 showed the best photocatalytic performance. The degradation rate of NPTO-600 with good stability was 2.5-times higher than that of NTO-500 (Figure 2b). The radicals produced during the reaction, which are known to play a significant role in photodegradation, were studied by the ESR spectra [40]. Spin-trapping ESR for the generated superoxide radicals (·O 2 − ) and hydroxyl radicals (·OH) by 5,5-dimethyl-1-pyrroline N-oxide (DMPO) trapping was carried out. As shown in Figure 2c, all the samples exhibited four strong 1:1:1:1 signals assigned to DMPO-·O 2 − after visible-light irradiation for 5 min, verifying that the ·O 2 − radicals were efficiently produced. The peak intensity of NPTO-T was higher than that of NTO-500. NPTO-600 presented the highest intensity. This was further confirmed by the ·OH-related ESR (Figure 2d). All the samples exhibited four strong 1:2:2:1 signals assigned to DMPO-·OH after visible-light irradiation for 5 min, identifying that the ·OH radicals can also be produced in these samples under irradiation. In addition, the chargeseparation-related photocurrent densities ( Figure S5b) were consistent with the results of photocatalytic activities and radical-related ESR.
Nanomaterials 2022, 12, x FOR PEER REVIEW 6 of 13 samples, implying the defects associated with the N doping amount. The additional weak signal at g = 1.974 was attributed to the Ti 3+ defects in NPTO-650, which coincided with the XPS [29].
To clarify the effect of N doping and the associated defects on the photocatalytic performance, the photocatalytic activities of the as-prepared N-doped TiO2 were tested for acetaldehyde degradation under visible-light irradiation (Figure 2a). All of the NPTO-T samples exhibited higher activities than NTO-500, while NPTO-600 showed the best photocatalytic performance. The degradation rate of NPTO-600 with good stability was 2.5-times higher than that of NTO-500 (Figure 2b). The radicals produced during the reaction, which are known to play a significant role in photodegradation, were studied by the ESR spectra [40]. Spin-trapping ESR for the generated superoxide radicals (•O2 − ) and hydroxyl radicals (•OH) by 5,5-dimethyl-1-pyrroline N-oxide (DMPO) trapping was carried out. As shown in Figure 2c, all the samples exhibited four strong 1:1:1:1 signals assigned to DMPO-•O2 − after visible-light irradiation for 5 min, verifying that the •O2 − radicals were efficiently produced. The peak intensity of NPTO-T was higher than that of NTO-500. NPTO-600 presented the highest intensity. This was further confirmed by the •OH-related ESR (Figure 2d). All the samples exhibited four strong 1:2:2:1 signals assigned to DMPO-•OH after visible-light irradiation for 5 min, identifying that the •OH radicals can also be produced in these samples under irradiation. In addition, the charge-separation-related photocurrent densities ( Figure S5b) were consistent with the results of photocatalytic activities and radical-related ESR. A contradiction between the depressed photocatalytic performance and optimized N doping content of NPTO-650 (atomic content 4.22%, as shown in Table S1) may be A contradiction between the depressed photocatalytic performance and optimized N doping content of NPTO-650 (atomic content 4.22%, as shown in Table S1) may be attributed to the associated defects (deep-energy-level O v and Ti 3+ defects; Figure 1f) during the nitridation process. Those defects act as impurity energy levels under the conduction band to trap the photogenerated electrons and weaken the O 2 reduction ability [17]. Accordingly, eliminating the associated defects is expected to improve the photocatalytic performance of NPTO-650, since the N-rich doping greatly improved the visible-light absorption (Figure 1c).

Defect Healing
It is generally accepted that the O v and Ti 3+ defects can be healed by the oxidation process. For comparison, a common calcination process under an O 2 atmosphere at 400 • C was used to treat NPTO-650. The optimized R O 2 -NPTO-650 showed extended light absorption ( Figure S6a,c) and high photocatalytic activity for acetaldehyde degradation (Figure S6b,d). However, the doped N easily escapes at such a high temperature. A more effective and simpler photo-Fenton-treated reaction was therefore performed. Based on the DRS analysis ( Figure S7a DMPO spin-trapping ESR spectra were applied to verify the photo-Fenton-treated healing strategy. In Figure 3d, it is not possible to detect any signal in the system with the presence of H2O2 and DMPO, even with additional PTO samples in the dark. However, the NPTO-650 induced H2O2 reduction to produce •OH, which was trapped by the DMPO in dark (D-NPTO-650) to generate DMPO-•OH with a weak signal. This signal was dramatically enhanced under visible-light irradiation (L-NPTO-650). However, RL-H2O2-NPTO-650 cannot induce the H2O2 reduction reaction under the dark condition (D-RL-H2O2-NPTO-650), which may be attributed to the absence of Ti 3+ , as confirmed in Figure 3b,c. Accordingly, this smart photo-Fenton-treated healing strategy can be described as the following equation: H2O2 + Ti 3+ = Ti 4+ + OH − + •OH.

Discussion
In order to clarify the mechanism of charge separation in the optimized sample RL-H2O2-NPTO-650, atmosphere-controlled surface photovoltage spectra (AC-SPS) were measured. As shown in Figure 4a, RL-H2O2-NPTO-650 displayed the highest response under a N2 atmosphere, which decreased with the increasing content of O2 in the atmosphere. Because the doped N species in RL-H2O2-NPTO-650 can capture the photogenerated holes, the photogenerated electrons have an easier time diffusing to the surface of the material, which resulted in the SPS response [13,33]. Moreover, the SPS results were further supported by the photoelectrochemical measurement ( Figure S11a).

Discussion
In order to clarify the mechanism of charge separation in the optimized sample R L-H 2 O 2 -NPTO-650, atmosphere-controlled surface photovoltage spectra (AC-SPS) were measured. As shown in Figure 4a, R L-H 2 O 2 -NPTO-650 displayed the highest response under a N 2 atmosphere, which decreased with the increasing content of O 2 in the atmosphere. Because the doped N species in R L-H 2 O 2 -NPTO-650 can capture the photogenerated holes, the photogenerated electrons have an easier time diffusing to the surface of the material, which resulted in the SPS response [13,33]. Moreover, the SPS results were further supported by the photoelectrochemical measurement ( Figure S11a).
Based on the ESR results, both ·O 2 ( Figure 4b) and ·OH ( Figure S11b) production from R L-H 2 O 2 -NPTO-650 were the highest among all samples, implying the most enhanced charge separation. Photocatalytic acetaldehyde degradation under different light irradiation was measured. As shown in Figure 4c, the R L-H 2 O 2 -NPTO-650 sample displayed the highest activity, which was twice as high as the NPTO-650 sample under visible-light irradiation for 1 h. Meanwhile, the R L-H 2 O 2 -NPTO-650 sample exhibited a good stability for acetaldehyde degradation ( Figure S11c). The activities of R L-H 2 O 2 -NPTO-650 and NPTO-650 under LED light with different single wavelength irradiations were investigated (Figure 4d). One can see that the trend of single-wavelength irradiating activity matches well with the optical absorption of the photocatalysts, suggesting that the activity was indeed induced by the excitation of the sample. The R L-H 2 O 2 -NPTO-650 exhibited activities when irradiated with light from the ultraviolet to visible regions. Even under 590 nm single-wavelength-light irradiation, it also displayed obvious activity. Acetaldehyde can be almost completely degraded after 15 min under UV-visible-light irradiation over R L-H 2 O 2 -NPTO-650, as shown in Figure S11d, which is better than the commercial P25 TiO 2 . As shown in Figure S12a, the rate constant k of the samples was in the order of R L-H 2 O 2 -NPTO-650 > R H 2 O 2 -NPTO-650 > R O 2 -NPTO-650 > NPTO-650. Compared with previous work, our sample showed very good activity, as shown Table S2. The in suit FT-IR spectra were carried out to study the mechanism. Both in dark and light irradiation, the peaks of the intermediates during the degradation of acetaldehyde were observed in R L-H 2 O 2 -NPTO-650, indicating a more efficient adsorption and degradation capacity ( Figure S12b). Besides, the photo-Fentontreated healing strategy was also used for the treatment of NPTO-600. DRS ( Figure S13a) indicated the bandgap of R L-H 2 O 2 -NPTO-650 to be smaller than R L-H 2 O 2 -NPTO-600 due to the N-rich doping. There was no obvious difference in the defect-related light absorption between them. Accordingly, the photocatalytic activity of NPTO-600 can also be improved after the healing process, but it was still lower than that of R L-H 2 O 2 -NPTO-650 ( Figure S13b). Based on the ESR results, both •O2 (Figure 4b) and •OH ( Figure S11b) production from RL-H2O2-NPTO-650 were the highest among all samples, implying the most enhanced charge separation. Photocatalytic acetaldehyde degradation under different light irradiation was measured. As shown in Figure 4c, the RL-H2O2-NPTO-650 sample displayed the highest activity, which was twice as high as the NPTO-650 sample under visible-light irradiation for 1h. Meanwhile, the RL-H2O2-NPTO-650 sample exhibited a good stability for acetaldehyde degradation ( Figure S11c). The activities of RL-H2O2-NPTO-650 and NPTO-650 under LED light with different single wavelength irradiations were investigated (Figure 4d). One can see that the trend of single-wavelength irradiating activity matches well with the optical absorption of the photocatalysts, suggesting that the activity was indeed induced by the excitation of the sample. The RL-H2O2-NPTO-650 exhibited activities when irradiated with light from the ultraviolet to visible regions. Even under 590 nm single-wavelength-light irradiation, it also displayed obvious activity. Acetaldehyde can be almost completely degraded after 15 min under UV-visible-light irradiation over RL-H2O2-NPTO-650, as shown in Figure S11d, which is better than the commercial P25 TiO2. As shown in Figure S12a, the rate constant k of the samples was in the order of RL-H2O2-NPTO-650 > RH2O2-NPTO-650 > RO2-NPTO-650 > Based on the above discussion, a detailed mechanism is illustrated in Figure 5. Thanks to the high thermal stability of H 3 PO 4 -modified anatase TiO 2 , N-rich doped TiO 2 with an excellent light absorption property was successfully prepared. However, the associated deep-energy-level defects (Ti 3+ and partial O v ) were also formed along with the hightemperature nitridation process, which acted as the photogenerated electron acceptor and was adverse to O 2 activation. As for this, the associated deep-energy-level defects were effectively healed, but appropriately doped N species and shallow-energy-level defects were retained by the smart photo-Fenton-treated healing strategy. After that, doped N species and shallow-energy-level defects can capture the photogenerated h + and e − , respectively, to enhance charge separation and promote the subsequent reaction, so as to perform a good photocatalytic activity of acetaldehyde degradation. the high-temperature nitridation process, which acted as the photogenerated electron acceptor and was adverse to O2 activation. As for this, the associated deep-energy-level defects were effectively healed, but appropriately doped N species and shallow-energylevel defects were retained by the smart photo-Fenton-treated healing strategy. After that, doped N species and shallow-energy-level defects can capture the photogenerated h + and e − , respectively, to enhance charge separation and promote the subsequent reaction, so as to perform a good photocatalytic activity of acetaldehyde degradation.

Conclusions
In summary, N-rich doped anatase TiO2 was successfully fabricated by hightemperature nitridation under a NH3 atmosphere, depending on the H3PO4-modified TiO2. The doped N was demonstrated to extend visible-light absorption and trap photogenerated h + to enhance charge separation. The associated deep-energy-evel defects (Ov and Ti 3+ ) were successfully healed by a smart photo-Fenton-treated healing strategy to improve O2 activation and enhance charge separation. The visible-light activity for gas-phase acetaldehyde degradation of the healed N-doped TiO2 was twice as high as the untreated one. It also displayed better activity as compared to P25 TiO2 under UV-visible-light irradiation. This study sheds light on the importance of N-rich doping and modulating the associated defects for developing high-efficiency photocatalysts.
Supplementary Materials: The following Supporting Information can be downloaded at: www.mdpi.com/xxx/s1, Table S1: Nitrogen atomic content of NTO-500 and NPTO-T samples.; Table S2: Comparison of our work with other previous work about N-doped-TiO2-based photocatalysts for acetaldehyde degradation; Figure

Conclusions
In summary, N-rich doped anatase TiO 2 was successfully fabricated by high-temperature nitridation under a NH 3 atmosphere, depending on the H 3 PO 4 -modified TiO 2 . The doped N was demonstrated to extend visible-light absorption and trap photogenerated h + to enhance charge separation. The associated deep-energy-evel defects (O v and Ti 3+ ) were successfully healed by a smart photo-Fenton-treated healing strategy to improve O 2 activation and enhance charge separation. The visible-light activity for gas-phase acetaldehyde degradation of the healed N-doped TiO 2 was twice as high as the untreated one. It also displayed better activity as compared to P25 TiO 2 under UV-visible-light irradiation. This study sheds light on the importance of N-rich doping and modulating the associated defects for developing high-efficiency photocatalysts.