English

• Development of highly efficient photocatalytic materials for water splitting or eliminating organic pollutants has been considered to be one of the important investigative fields [1-3]. TiO₂ is regarded as an ideal photocatalytic material, thanks to its non-toxicity, outstanding stability, and low cost [4, 5]. However, pure TiO₂ photocatalysts generally suffer from the fast recombination rate of photoexcited electrons and holes, and a large band gap energy (3.2 eV) that is excited by only ultraviolet light (~4% of sunlight) [6]. Therefore, it is still urgently needed to design and prepare novel TiO₂ based photocatalysts with enhanced separation efficiency of photoexcited electron/hole pairs and visible light response to meet the practical requirement.

  Recently, nitrogen (N) doped TiO₂ has attracted a lot of attention due to its narrowed band gap [7-9]. In the N-doped TiO₂, there are two states of N, namely substitutional and interstitial (Schemes S1a and b in Supporting information). In the substitutional case (Scheme S1a), corresponding to an oxygen (O) atom replaced by a N at a regular lattice site, an impurity energy level (E_i) in N-doped TiO₂ is introduced, which just above the valence band maximum [10]. While, in the case of interstitial position (Scheme S1b), i.e. directly bounding to lattice oxygen, leading to a slightly higher E_i in the gap [10]. This indicates that the electronic structure of N in the N-doped TiO₂ is part of the energy levels (N 2p) of TiO₂, which can effectively narrow its band gap. Moreover, N doping can simultaneously introduce oxygen vacancies (V_o) in the lattice [10-12]. Due to the presence of two excess electrons in N atom, two Ti⁴⁺ cations can be reduced to Ti³⁺ cations, leading to the formation of V_o and defect level (Ti³⁺ 3d, E_d). According to the previous studies, the formation of Ti³⁺ is found to be about 0.8 eV below the conduction band minimum, which can also narrow the band gap of TiO₂ [10, 13, 14]. However, the narrowed band gap is also a "double-edged sword". Although the doping of N can effectively improve the visible light response of the TiO₂, the photogenerated electrons and holes can recombine more easily due to the increased defect sites.

  It is noteworthy that the presence of N impurities and V_o can also decrease the Fermi level (E_F) of the TiO₂, resulting in the electron transfer from the pure TiO₂ to N-TiO₂ when combining them together [15]. After the equilibrium of E_F, the band structure will be bending and causing the formation of n-n⁺ heterojunction [16], which can facilitate the transfer of electrons and holes to the N-TiO₂. In this case, constructing a core-shell structured TiO₂@N-TiO₂ photocatalyst will be favorable to enhance the photocatalytic performance because the photogenerated electrons and holes can migrate from the core to the shell to trigger water splitting or organic oxidation. However, the conventional type-Ⅱ heterojunction may cause the retention of electrons or holes in the core, which will hinder their usage. On the other hand, the preparation of heteroatomic gradient-doped semiconductors to adjust the band structure has gained a lot of attention [6, 7, 17, 18], because the gradually changed heteroatomic doping level may form a bent band structure at the gradient-doped layer, which can facilitate the charge separation. Therefore, constructing core-shell TiO₂@N-TiO₂ heterojunction with gradient-doped N level could be a fantastic way to improve the photocatalytic performance, which is rarely reported before.

  Herein, a branched core-shell nanosphere composed of an anatase TiO₂ (a-TiO₂) core and a shell of gradient-doped N-TiO₂ nanobranches (a-TiO₂@N-TiO₂) is synthesized by an in situ doping process in the presence of NH₃·H₂O as N source for the first time. During the preparation process, a mixed crystal anatase-rutile TiO₂ (ar-TiO₂) nanosphere is first prepared by oxidizing Ti in H₂O₂ solution and then etched by NH₃·H₂O to form (NH₄)₂TiO₃ shell, which is further annealed in ambience to obtain the final product, a-TiO₂@N-TiO₂. The morphology of the samples could be modified by the amount of NH₃·H₂O. The optimized a-TiO₂@N-TiO₂ with an uniform size of ~700 nm, including a ~500 nm core and a ~100 nm shell composed of gradient-doped N-TiO₂ nanobranches (the doping level gradually increased from the shell bottom to the shell surface), which exhibit narrowed band gap to improve the visible light response and form a n-n⁺ heterojunction between the a-TiO₂ core and N-TiO₂ shell. Furthermore, the gradient-doped N should result in a bent band structure in the shell, which can also enhance the separation of photogenerated carriers, leading to a remarkably improved photocatalytic performance in both producing H₂ and degrading refractory organic pollutants.

  The detailed synthesis process of branched core-shell a-TiO₂@N-TiO₂ nanosphere is illustrated in the Supporting information and Fig. 1a. In the first stage, H₂O₂ reacts with the Ti atoms on the surface of Ti foil (Eq. 1), leading to the generation of mixed crystal ar-TiO₂ nanospheres (Figs. S1 and S2 in Supporting information). As the time increased, NH₃·H₂O reacts with the outer TiO₂ on the surface of ar-TiO₂ microspheres to generate (NH₄)₂TiO₃ (Eq. 2) [19]. The reactions could be described as follows:

  (1)

  (2)

  Figure 1

  

  Figure 1.  (a) Schematic illustration of formation of branched core-shell a-TiO₂@N-TiO₂ nanosphere. (b-d) SEM images of the a-TiO₂@N-TiO₂–4, (e–g) EDX mapping of a-TiO₂@N-TiO₂–4 confirming the presence of Ti, O and N. TEM (h and i) and HRTEM (j and k) images of the a-TiO₂@N-TiO₂–4. Insert curves in (i) are the line scanning EDX results of the core-shell nanosphere.

  DownLoad: Full-Size Img PowerPoint

  The reaction between Ti atoms and H₂O₂ forms a porous structure in the TiO₂ according to the Kirkendall effect (Fig. S2d) [19]. Meanwhile, NH₃·H₂O solutions corrode the TiO₂ on the surface, leading to the formation of core-shell structure with ar-TiO₂ as core and (NH₄)₂TiO₃ as shell. Under the effects of the H₂O₂ and NH₃·H₂O, the surface of TiO₂ particle is corroded into a branched structure. After calcination, the ar-TiO₂ can be converted into anatase TiO₂ [20], while the (NH₄)₂TiO₃ can be turned into N-TiO₂ [21], and finally the 3D branched core-shell a-TiO₂@N-TiO₂ nanospheres are synthesized.

  The morphologies of the as-prepared samples are first characterized by scanning electron microscopy (SEM). In the control experiments, various amount of NH₃·H₂O is used in the preparation process and different kinds of nanospheres are obtained (Figs. 1b–d and Fig. S3 in Supporting information) and denoted as a-TiO₂@N-TiO₂-x (x = 0, 2, 4, 6 and 8). However, the branched core-shell a-TiO₂@N-TiO₂ nanospheres can be synthesized with only 4 mL NH₃·H₂O (Figs. 1b-d), which display a structure of branched nanosphere with a uniform size distribution (Fig. 1b) and average diameter of ~700 nm (Fig. 1c). Fig. 1d depicts magnified SEM image of a single a-TiO₂@N-TiO₂–4 nanosphere. The surface of the nanosphere is composed of disorderly nanobranches with the length about ~50 nm, which significantly improves its surface area and the amount of active sites. The results of energy dispersive X-ray (EDX) mapping (Figs. 1e–g) confirm that the a-TiO₂@N-TiO₂ nanospheres consisted of Ti, O and N, demonstrating the existence of N. The X-Ray diffraction (XRD) plots (Fig. S4 in Supporting information) indicate that all these samples display highly crystallized structure with distinct characteristic peaks of anatase-phase TiO₂ (PDF #21–1272) [22], implying that the doping of N do not affect the crystalline phases of the TiO₂, and the ar-TiO₂ could be transferred into anatase TiO₂ completely after annealing.

  The detailed morphologies and microstructures of a-TiO₂@N-TiO₂–4 are further investigated by transmission electron microscopy (TEM). Fig. 1h shows a typical TEM image of the a-TiO₂@N-TiO₂–4 nanospheres, in which the nanospheres are well-dispersed without aggregation and the average diameter is ~700 nm. Further magnified TEM image (Fig. 1i) clearly reveals a core–shell structure with a core diameter of ~500 nm and a shell thickness of ~100 nm. Based on the EDX line scan profile (insert in Fig. 1i), the EDX intensity of N element is gradually increased from the bottom of the shell to the edge, suggesting a gradient-doped N in the shell, which could result in a bent band structure in the shell to facilitate the separation of photogenerated charges [7]. However, the Ti and O elements show higher intensity than N element and slightly reduced from the core center to the edge, which demonstrates a structure of localized a-TiO₂ core and N-TiO₂ shell. The shell is consisted of a large number of nanocrystals, which form the nanobranches (Fig. 1j). Thanks to the interlaced nanobranches on the surface of a-TiO₂@N-TiO₂–4, the sample displays an ultrahigh average pore-diameter of 18.44 nm with a Brunauer–Emmett–Teller (BET) surface area of 123 m²/g (Fig. S5 in Supporting information), so that much more accessible reaction sites can be exposed and the capture of the light can be improved. Fig. 1k shows a typical high-resolution TEM (HRTEM) image of a-TiO₂@N-TiO₂–4, which is taken on the edge of a single nanorods. The lattice spacing is 0.35 nm, corresponding to the (101) crystal plane of anatase TiO₂ [23]. Moreover, it should be noted that the lattice fringes become blurred and discontinuous at the edges of N-TiO₂, indicating that it is a defects-rich region [24]. This could be attributed to the formation of abundant V_o in the lattice of TiO₂ due to the high-ratio of doping N at the edge as mentioned above (insert in Fig. 1i). This amorphous N-TiO₂ may derive more reactive sites for water splitting [25, 26].

  X-ray photoelectron spectroscopy (XPS) measurement reveals the existence of Ti, O and N in the a-TiO₂@N-TiO₂–4 (Fig. 2a), which is consistent with the results of EDX. The high-resolution XPS spectrum of Ti 2p (Fig. 2b) of a-TiO₂@N-TiO₂–4 shows two evident peaks located at 458.5 and 464.0 eV, which are ascribed to Ti 2p_3/2 and Ti 2p_1/2, respectively, which are slightly red-shifted compared with that of a-TiO₂@N-TiO₂–0 [27]. This red-shift could be ascribed to the N doping in the nanobranches coating, because the XPS mainly detects the sample's surface [24]. Fig. 2c shows the XPS spectra for O 1s region of a-TiO₂@N-TiO₂–0 and a-TiO₂@N-TiO₂–4. It can be seen that the a-TiO₂@N-TiO₂–0 shows a broad peak at 530.0 eV, which can be ascribed to the Ti–O-Ti in anatase TiO₂, while a small peak at 531.9 eV is attributed to the hydroxyl groups of the chemisorbed water [28]. However, the O 1s peak of a-TiO₂@N-TiO₂–4 can be split as two red-shifted peaks at 529.9 eV and 531.7 eV, respectively, which could also be attributed to the formation of abundant V_o and amorphous TiO₂ due to the N doping. The a-TiO₂@N-TiO₂–4 shows a weak peak at 400 eV ascribed to N 1s (Fig. 2d), which is in accordance with the doped N in TiO₂ as reported before [29]. The XPS results further demonstrate the a-TiO₂@N-TiO₂ nanospheres are anatase phase, and N is successfully doped in the sample by the facile in situ doping process.

  Figure 2

  

  Figure 2.  (a) XPS spectrum of the a-TiO₂@N-TiO₂–4, (b) Ti 2p, (c) O 1s and (d) N 1s. (e) UV–vis, (f) enlarged UV–vis spectra (600–750 nm), (g) Tauc plot and (h) PL spectra of a-TiO₂@N-TiO₂-x (x = 0, 2, 4, 6 and 8).

  DownLoad: Full-Size Img PowerPoint

  Generally, doping with non-metallic elements, including N, and formation of V_o could alter the electronic properties of TiO₂. We therefore explore the electronic band structure of a-TiO₂@N-TiO₂-x using UV-DRS spectra. As shown in Fig. 2e, the visible light absorption of the a-TiO₂@N-TiO₂-x is significantly enhanced with the increase of the N content, indicating the N dopant can effectively narrow the band gap of TiO₂ and make it active under visible light. In addition, a-TiO₂@N-TiO₂-x also exhibit another obviously weak absorption band in the range from 600 nm to 750 nm (Fig. 2f), suggesting there is not a homogeneous distribution of N doping and V_o in the bulk a-TiO₂@N-TiO₂-x [24], and this is consistent with the TEM results. The corresponding band structures of the a-TiO₂@N-TiO₂-x are also calculated and shown in Fig. 2g. According to the transformed Kubelka-Munk function αh_υ = A(h_υE_g)^1/2, a gradually red shift of the band gaps (3.35, 3.31, 3.27, 3.25, 3.1 eV) of a-TiO₂@N-TiO₂-x are observed with the amount of NH₃·H₂O increased, which is in accordance with their optical adsorption properties. It should be noted that, exception of the main band gaps, a-TiO₂@N-TiO₂-x (x = 2, 4, 6, and 8) have another weak band structures, which can be attributed to the existed impurity level and defect level [24, 30, 31]. In addition, the photoluminescence (PL) spectra are executed to evaluate the separation efficiency of photoexcited charge carriers in the samples. As displayed in Fig. 2h, a-TiO₂@N-TiO₂–4 shows the weakest PL intensity, indicating the fast electrons transfer between a-TiO₂ and N-TiO₂, which lead to the PL quenching and reduce the electron-hole recombination [32].

  The photocatalytic H₂ production performance is evaluated under air mass (AM) 1.5 irradiation with 20 vol% methanol as the sacrificial agent (Figs. 3a and b). As displayed in Fig. 3a, the photocatalytic activity of a-TiO₂@N-TiO₂-x is enhanced with increasing N content from a-TiO₂@N-TiO₂–0 to a-TiO₂@N-TiO₂–4, which should be attributed to the enhanced charge transfer at the photocatalyst/solution interface due to the arising of nanobranches, and improved separation of photogenerated electron/hole pairs in the bulk phase ascribed to the formation of gradient N doping. The a-TiO₂@N-TiO₂–4 possesses the highest photocatalytic activity with the H₂ production rate of 308.1 µmol g⁻¹ h⁻¹, which is superior to most of the TiO₂-based photocatalysts for H₂ production reported to date (Table S1 in Supporting information). However, a further increase of N content, the photocatalytic activity of a-TiO₂@N-TiO₂-x (x = 6 and 8) composites is decreased. This can be due to the excess O defects sites in a-TiO₂@N-TiO₂-x, which can also act as the recombination centers for the photogenerated electron/hole pairs [33]. Moreover, the superfluous NH₃·H₂O would destroy the branched core-shell morphology (Figs. S3d-g in Supporting information), which is adverse for the separation and transportation of the charges. The stability of the a-TiO₂@N-TiO₂–4 photocatalyst is further evaluated (Fig. 3b). There is no evident photocatalyst deactivation in continuous H₂ evolution measurements after 15 h operation, which demonstrates the excellent stability of a-TiO₂@N-TiO₂–4 for photocatalytic H₂ evolution.

  Figure 3

  

  Figure 3.  (a) Hydrogen evolution of the samples and (b) recycled hydrogen evolution for 15 h of a-TiO₂@N-TiO₂–4 under AM 1.5 irradiation with 20 vol% methanol as the sacrificial agent. (c) Removal ratio and (d) rate constant on photocatalytic degradation of TC by a-TiO₂@N-TiO₂-x under AM 1.5 illumination (TC = 10 ppm, T = 25 ℃). (e) Chopped J-t curves tested at 0.8 V vs. SCE under AM 1.5 illumination, and (f) EIS Nyquist plots of the a-TiO₂@N-TiO₂-x.

  DownLoad: Full-Size Img PowerPoint

  As one of the most active photocatalyst, the photogenerated holes of TiO₂ have enough oxidation potential (~3.1 V vs. Normal Hydrogen Electrode, NHE) to oxidize the absorbed hydroxyl into hydroxyl radicals and also directly oxidize most kinds of organics [34]. The photocatalytic performance is further evaluated by degrading tetracycline (TC) under AM 1.5 illumination (Figs. 3c and d). The results indicate that the a-TiO₂@N-TiO₂–4 shows the highest active in removing TC, which can remove almost 100% TC from the solution after 1 h operation with a rate constant (k) of ~0.093 min⁻¹ (Fig. S6 in Supporting information). The ranking of activity of these samples are the same as the photocatalytic H₂ evolution, which further demonstrates the optimized sample is a-TiO₂@N-TiO₂–4. The active species in the TC degradation process are studied using the trapping experiments. As shown in Fig. S7 (Supporting information), ^·OH and h⁺ are the main active species in the degradation of TC. In the photocatalytic degradation process, organics can be oxidized by the holes and the corresponding hydroxyl radicals [35], which means the photocatalytic activity is mainly determined by the separation efficiency of the photogenerated electron/hole pairs in the photocatalyst. So, the ranking of activity is the same in both photocatalytic water splitting and organic decomposition.

  The photocatalytic activity of a-TiO₂@N-TiO₂–4 is further investigated by removing various refractory organic pollutants (Table S2 in Supporting information), including organic dyes, phenols, antibiotics, and personal care products (PCPs). Excellent removal ratios are achieved for these model pollutants after 1 h or 2 h operation under AM 1.5 illumination. The removal ratios of the organic dyes are higher than 91% after 1 h operation even with an initial concentration of 20 ppm, which are much higher than that of others. This should be due to the relatively easy fading of organic dyes. The removal rates of phenols are relatively smaller than others, lower than 90% after 1 h operation with an initial concentration of 10 ppm, which should be attributed to their more stable molecular structures than other model pollutants. However, after 2 h operation, all these organic pollutants could be degraded over 96%, and the removal ratios of total organic carbon (TOC) are similar and reach to 65.8%−77.3% after 5 h operation, which further demonstrates the outstanding photocatalytic performance of a-TiO₂@N-TiO₂–4 in degrading organic pollutants.

  To confirm the enhanced charge separation efficiency in the branched core-shell a-TiO₂@N-TiO₂ nanospheres, chronoamperometry (J-t) and electrochemical impedance spectroscopy (EIS) measurements are carried out for all as prepared samples. As shown in Fig. 3e, all samples show transient photocurrent response when turn on the light, which further demonstrates all of them have photoactivity. A higher photocurrent response means a more efficient charge separation performance. The a-TiO₂@N-TiO₂–4 shows the highest photocurrent density compared with others, which indicates the optimized morphology of nanobranched core-shell nanosphere and gradient N doping significantly enhance the charge transfer and separation. Furthermore, the a-TiO₂@N-TiO₂–4 presents a representative EIS Nyquist plot with a much smaller arc radius than that of other samples (Fig. 3f), which means that the charge transfer resistance is remarkably decreased by the appropriate doping of N in a-TiO₂@N-TiO₂–4. In general, both J-t and ESI results reveal the high efficiency of a-TiO₂@N-TiO₂–4 in accelerating charge transfer and separation, and thus an excellent photocatalytic performance could be anticipated.

  Based on the above analyses, we propose the possible energy levels and charge transfer pathways in the branched core-shell a-TiO₂@N-TiO₂ nanospheres and show them in Fig. 4. In the darkness, the E_F of N-TiO₂ would be lower than that of a-TiO₂ because the existed impurity level induced by N doping (left part of Fig. 4a) [14]. As the doping amount of N increasing from the bottom of the branched shell to the surface, the E_F should decrease gradually [16]. Therefore, the electrons in the a-TiO₂ would transfer to the N-TiO₂, which could lead to the equilibration of E_F and the band bending in the gradient N doped branch layers, forming a n-n⁺ heterojunction with unique band structure (right part of Fig. 4a). Furthermore, the migration of electrons could cause the formation of electric field between a-TiO₂ and N-TiO₂. In this situation, both photogenerated electrons and holes can be driven to the conduction band (CB) and valence band (VB) of N-TiO₂ nanobranches, respectively. In the presence of hole scavengers, the charge recombination will be effectively suppressed in the N-TiO₂ nanobranches, therefore the evolution of H₂ can be significantly enhanced by the reduction of electrons. When applying in photocatalytic wastewater treatment, the holes should oxidize the absorbed organics directly, or oxidize the absorbed –OH groups into hydroxyl radicals (^·OH) for further oxidizing organics efficiently, while the electrons should be captured by dissolved oxygen for oxygen reduction reaction (ORR) [34]. Obviously, this unique band structure and charge transfer model is very applicable in this core-shell heterojunction for hindering charge recombination, because both electrons and holes can be applied in reduction and oxidation reactions efficiently at the photocatalyst/solution interface (Fig. 4b). Furthermore, the ultrathin amorphous layer at the nanobranches′ surface (Fig. 1k) can enhance the absorption of substances and provide more active sites [17], therefore facilitating the charge transfer at the photocatalyst/solution interface. In addition, the nanobranches should cause the reflection of light, enhancing the light harvesting for producing more charges. Consequently, the branched core-shell a-TiO₂@N-TiO₂ nanospheres with gradient N doping show significantly improved photocatalytic activities in both solar water splitting and wastewater treatment.

  Figure 4

  

  Figure 4.  Schematic illustration of (a) the band structure evolution for the formation of homojunction and (b) photoexcited carrier transfer in the branched core-shell a-TiO₂/N-TiO₂ nanospheres with a gradient distribution of N element in the nanobranches.

  DownLoad: Full-Size Img PowerPoint

  In general, we synthesize a novel branched core-shell a-TiO₂@N-TiO₂ nanosphere with a gradient N doping in the nanobranches for highly efficient photocatalytic hydrogen generation and wastewater purification. The nanosphere is prepared by a facile wet chemical method, in which the amount of NH₃·H₂O plays an important role to control the nanosphere's morphology. The optimized amount is 4 mL in the given conditions, and the obtained nanospheres has uniform morphology with a mesoporous core of a-TiO₂ (~500 nm of diameter) and a shell of N-TiO₂ nanobranches with gradient N doping (gradually increased from the bottom to the top, ~100 nm of thickness) and a ultrathin amorphous coating. This optimized sample shows a H₂ production rate of 308.1 µmol h⁻¹ g⁻¹ under AM 1.5 illumination, which is about 6.7 times higher than that of the bare TiO₂ nanosphere. Furthermore, the a-TiO₂@N-TiO₂ nanospheres also exhibit high activity in removing various refractory organic pollutants. The outstanding photocatalytic performances of the a-TiO₂@N-TiO₂ nanosphere should be attributed to the following reasons: (1) the gradient N doping cause the formation of n-n⁺ heterojunction and electric field between a-TiO₂ and N-TiO₂, which drive photogenerated electrons and holes transferring to the N-TiO₂ nanobranches; (2) the ultrathin amorphous layer at the nanobranches' surface enhance the absorption of substances and provide more active sites; (3) the nanobranches cause the reflection of light, enhancing the light harvesting for producing more charges. This work provides a significant reference for the rational design of unique nanostructured photocatalyst used in hydrogen production and wastewater purification.

  Declaration of competing interest

  The authors declare no conflict of interest.

  Acknowledgments

  This research was supported by the National Natural Science Foundation of China (No. 52170083), the Excellent Youth Fund Project of Natural Science Foundation of Hunan Province (No. 2021JJ20007), and the Research Foundation of Education Bureau of Hunan Province, China (No. 21B0441).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.06.051.

  
  
• 1. [1]

     Q.S. Zhang, Y.M. Li, H. Ren, et al., Chem. Eng. J. 408 (2021) 127267. doi: 10.1016/j.cej.2020.127267

  2. [2]

     H.D. Ji, J.R. Ni, D.Y. Zhao, et al., ACS EST Eng. 2 (2022) 1015–1038. doi: 10.1021/acsestengg.1c00451

  3. [3]

     Q.S. Zhang, Y. Xiao, Y.M. Li, et al., Chem. Eng. J. 393 (2020) 124681. doi: 10.1016/j.cej.2020.124681

  4. [4]

     Y.C. Zhou, P. Wang, H. Fu, et al., Chin. Chem. Lett. 31 (2020) 2645–2650. doi: 10.1016/j.cclet.2020.02.048

  5. [5]

     X.N. Liu, H.D. Ji, S. Li, et al., Chemosphere 233 (2019) 198–206. doi: 10.1016/j.chemosphere.2019.05.229

  6. [6]

     H.D. Ji, W. Liu, F.B. Sun, et al., Chem. Eng. J. 419 (2021) 129605. doi: 10.1016/j.cej.2021.129605

  7. [7]

     A. Mirzaei, M. Eddah, S. Roualdes, et al., Chem. Eng. J. 422 (2021) 130507. doi: 10.1016/j.cej.2021.130507

  8. [8]

     R. Asahi, T. Morikawa, T. Ohwaki, et al., Science 293 (2001) 269–271. doi: 10.1126/science.1061051

  9. [9]

     M. Esmat, H. El-Hosainy, R. Tahawy, et al., Appl. Catal. B 285 (2021) 119755. doi: 10.1016/j.apcatb.2020.119755

  10. [10]

      S. Livraghi, M.C. Paganini, E. Giamello, et al., J. Am. Chem. Soc. 128 (2006) 15666–15671. doi: 10.1021/ja064164c

  11. [11]

      R. Asahi, T. Morikawa, H. Irie, et al., Chem. Rev. 114 (2014) 9824–9852. doi: 10.1021/cr5000738

  12. [12]

      D.N. Zhang, X.Y. Ma, H.W. Zhang, et al., Mater. Today Energy 10 (2018) 132–140. doi: 10.1016/j.mtener.2018.08.018

  13. [13]

      A.K. See, R.A. Bartynski, J. Vac. Sci. Technol. A 10 (1992) 2591–2596. doi: 10.1116/1.578105

  14. [14]

      M.A. Henderson, W.S. Epling, C.H.F. Peden, et al., J. Phys. Chem. B 107 (2003) 534–545. doi: 10.1021/jp0262113

  15. [15]

      K. Wang, T. Peng, Z.M. Wang, et al., Appl. Catal. B 250 (2019) 89–98. doi: 10.1016/j.apcatb.2019.03.026

  16. [16]

      X.S. Huang, Z.M. Gao, P. Li, et al., J. Appl. Phys. 123 (2018) 084502. doi: 10.1063/1.5018772

  17. [17]

      C.X. Li, S.H. Dong, R. Tang, et al., Energy Environ. Sci. 11 (2018) 3201–3211. doi: 10.1039/C8EE01046C

  18. [18]

      J.Y. Kim, G.C. Liu, R.E.A. Ardhi, et al., Nano-Micro Lett. 14 (2022) 46. doi: 10.1007/s40820-021-00788-z

  19. [19]

      C.X. Wang, L.W. Yin, L.Y. Zhang, et al., Langmuir 26 (2010) 12841–12848. doi: 10.1021/la100910u

  20. [20]

      W.W. Fu, G.D. Li, Y. Wang, et al., Chem. Commun. 54 (2018) 58. doi: 10.1039/C7CC05750D

  21. [21]

      D.H. Chen, F.Z. Huang, L. Cao, et al., Chem. Eur. J. 43 (2012) 13762–13769.

  22. [22]

      Q.Y. Zeng, S. Chang, A. Beyhaqi, et al., Nano Energy 67 (2020) 104237. doi: 10.1016/j.nanoen.2019.104237

  23. [23]

      M.H. Wu, Y.P. Gao, Y. Hu, et al., Chin. Chem. Lett. 31 (2020) 897–902. doi: 10.1016/j.cclet.2019.07.039

  24. [24]

      X.Q. Hao, Y.C. Wang, J. Zhou, et al., Appl. Catal. B 221 (2018) 302–311. doi: 10.1016/j.apcatb.2017.09.006

  25. [25]

      X. Hu, C. Li, J. Song, et al., J. Colloid Interf. Sci. 574 (2020) 61–73. doi: 10.1016/j.jcis.2020.04.035

  26. [26]

      C.T. Feng, P.F. Wu, Q.L. Li, et al., New J. Chem. 46 (2022) 1731–1740. doi: 10.1039/D1NJ05917C

  27. [27]

      J.H. Park, S. Kim, A.J. Bard, Nano Lett. 6 (2006) 24–28. doi: 10.1021/nl051807y

  28. [28]

      W.C. Wang, Z.X. Yang, G.Y. Li, et al., J. Alloys Compd. 890 (2021) 161895.

  29. [29]

      S. Khan, T.L. Ruwer, N. Khan, et al., J. Mater. Chem. A 9 (2021) 12214. doi: 10.1039/D0TA11494D

  30. [30]

      X.Q. Hao, Z.W. Cui, J. Zhou, et al., Nano Energy 52 (2018) 105–116. doi: 10.1016/j.nanoen.2018.07.043

  31. [31]

      S.M. Zhu, X.J. Qian, D.P. Lan, et al., Appl. Catal. B 269 (2020) 118806. doi: 10.1016/j.apcatb.2020.118806

  32. [32]

      S. Sun, P. Gao, Y. Yang, et al., ACS Appl. Mater. Interfaces 8 (2016) 18126–18131. doi: 10.1021/acsami.6b05244

  33. [33]

      M. Kong, Y.Z. Li, X. Chen, et al., J. Am. Chem. Soc. 133 (2011) 16414–16417. doi: 10.1021/ja207826q

  34. [34]

      Q.Y. Zeng, S. Chang, A. Beyhaqi, et al., J. Hazard. Mater. 394 (2020) 121–425.

  35. [35]

      Q.Y. Zeng, S. Chang, M. Wang, et al., Chin. Chem. Lett. 32 (2021) 2212–2216. doi: 10.1016/j.cclet.2020.12.062

• 

  Figure 1  (a) Schematic illustration of formation of branched core-shell a-TiO₂@N-TiO₂ nanosphere. (b-d) SEM images of the a-TiO₂@N-TiO₂–4, (e–g) EDX mapping of a-TiO₂@N-TiO₂–4 confirming the presence of Ti, O and N. TEM (h and i) and HRTEM (j and k) images of the a-TiO₂@N-TiO₂–4. Insert curves in (i) are the line scanning EDX results of the core-shell nanosphere.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) XPS spectrum of the a-TiO₂@N-TiO₂–4, (b) Ti 2p, (c) O 1s and (d) N 1s. (e) UV–vis, (f) enlarged UV–vis spectra (600–750 nm), (g) Tauc plot and (h) PL spectra of a-TiO₂@N-TiO₂-x (x = 0, 2, 4, 6 and 8).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Hydrogen evolution of the samples and (b) recycled hydrogen evolution for 15 h of a-TiO₂@N-TiO₂–4 under AM 1.5 irradiation with 20 vol% methanol as the sacrificial agent. (c) Removal ratio and (d) rate constant on photocatalytic degradation of TC by a-TiO₂@N-TiO₂-x under AM 1.5 illumination (TC = 10 ppm, T = 25 ℃). (e) Chopped J-t curves tested at 0.8 V vs. SCE under AM 1.5 illumination, and (f) EIS Nyquist plots of the a-TiO₂@N-TiO₂-x.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Schematic illustration of (a) the band structure evolution for the formation of homojunction and (b) photoexcited carrier transfer in the branched core-shell a-TiO₂/N-TiO₂ nanospheres with a gradient distribution of N element in the nanobranches.

  下载: 全尺寸图片 幻灯片
•