A. Composition analysis

Typical XRD patterns of the samples are shown in Fig. 1. It is seen that P25 sample contains anatase and rutile, which are corresponding to JCPDS no 21–1272 and JCPDS no 21–1276 cards, respectively. And all the diffraction peaks of N-TiO₂, N-S-TiO₂, and pure TiO₂ can be indexed to the unique anatase phase, indicating that the doping with nitrogen into TiO₂ would not produce new phases. Furthermore, there are no nitrogen-originated XRD peaks existed in the samples, which is probably due to the low dopants content in the samples below the detection limit of the instrument. In addition, it is because the dopants were dispersed in the lattice of TiO₂. ^{Reference Cheng, Yu, Xing and Yang27}

FIG. 1.

Typical XRD patterns of pure TiO₂, N-TiO₂, N-S-TiO₂, and P25 samples.

Figure 2 presents a comparison on the Raman spectra of N-doped TiO₂ by two methods and pure TiO₂. For all the samples, the four peaks could be assigned to E_g(1) (144 cm⁻¹), B_1g(1) (396 cm⁻¹), A_1g + B_1g(2) (516 cm⁻¹) and E_g(2) (639 cm⁻¹) modes of anatase TiO₂ phase, respectively. In addition, there is no nitrogen peaks in the spectra, which is in agreement with the XRD results.

FIG. 2.

Raman spectra of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples.

BET N₂-sorption measurements can be used to investigate the surface areas, pore volumes, and average pore diameters of the samples. The recorded N₂ adsorption–desorption isotherms for pure TiO₂, N-TiO₂, and N-S-TiO₂ samples are shown in Fig. 3. N₂ adsorption–desorption isotherms include six types from I to VI, indicating the existence of micropore solids, nonhole, or macropore structure, the interaction of gas–solid on nonhole or macropore materials, mesoporous structure, the interaction of gas–solid on micropore and mesoporous materials, and the multilayer adsorption on surface of nonhole materials, respectively. The isotherms of all the three samples belong to the type IV with H2 hysteresis loop, indicating the existence of mesoporous structure according to the IUPAC classification. ^{Reference Xie, Bian, Yao, Hao and Wei28} The formation of mesoporous is due to the shrinking of crystals at high temperature, which leaves space between the crystals. In addition, the melt ions always left the space between the particles in mixed nitrates system. ^{Reference Sathiyanarayanan, Alimohammadi, Zhou and Fichthorn29}

FIG. 3.

N₂ adsorption–desorption isotherms of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples.

According to the measured results by BET method, the specific surface areas, pore sizes and pore volumes of the three samples were calculated, and the results are listed in Table I. The samples show variation in their surface areas, which are 107.2, 129.0, and 154.6 m²/g, respectively, for pure TiO₂, N-TiO₂, and N-S-TiO₂ samples. The specific surface areas of N-TiO₂ and N-S-TiO₂ samples are higher than that of the pure TiO₂, which may be due to the inhibition of urea to particle size growth of TiO₂ during the hydrothermal and calcination processes. The N-S-TiO₂ sample presents the highest surface area compared to the other two samples. This might be owing to the decomposition of the added nitrates dispersed in the samples at the calcination process, which would lead to the formation of large amount of pores in the samples. Moreover, the mixed nitrates could directly result in relative slow nucleation and growth rates owing to the low flowability and diffusivity, which might inhibit the particle size growth of TiO₂. Therefore, the N-S-TiO₂ sample with a larger specific surface area provided abundant reacting active sites and a mass of molecule transport channels that are conducive to the sufficient interaction between the reactive molecules and photocatalyst, thus improving photocatalytic performance.

TABLE I.

Structural properties of pure TiO₂, N-TiO₂ and N-S-TiO₂ samples.

^a The specific surface area was calculated using BET equation.

^b Pore volume.

^c Pore diameter.

The chemical structure of the prepared TiO₂ samples was examined by XPS. Typical XPS survey spectrum of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples is shown in Fig. 4(a). XPS survey spectrum of N-TiO₂ and N-S-TiO₂ sample reveals the existence of titanium, oxygen, and nitrogen, according to the peak positions of Ti 2p (464.2 eV), O 1s (531.6 eV) and N 1s (400 eV). In contrast, there is no N 1s peak existence in XPS survey spectrum of pure TiO₂. Therefore, it can be deduced that the doping of nitrogen into TiO₂ was successful.

FIG. 4.

Typical XPS results of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples: (a) survey spectrum, (b) Ti 2p peaks, (c) O 1s peaks, and (d) N 1s peaks for all the three samples.

The Ti 2p core level for pure TiO₂, N-TiO₂, and N-S-TiO₂ samples is shown in Fig. 4(b). There are two strong peaks at the binding energies of 458.3 and 464.1 eV in Ti 2p region of N-S-TiO₂ sample, which can be attributed to the Ti 2p _3/2 and Ti 2p _1/2 spin-orbital split photoelectrons in the Ti⁴⁺ chemical state, respectively. ^{Reference Sun, Zhao, Zhou and Liu30} The splitting value of 5.8 eV between these bands is the characteristic of the Ti⁴⁺ species in anatase TiO₂. ^{Reference Wu and Tang31} The Ti 2p _3/2 core level can be deconvoluted into two peaks at 458.2 and 458.8 eV, which can be assigned to the Ti–N and Ti–O linkages of Ti 2p _3/2 orbit, respectively. In addition, the Ti 2p _1/2 core level can also be divided into two peaks at 463.2 and 464.2 eV, representing the presence of the Ti–N and Ti–O bonds in Ti 2p _1/2 orbit. The two peaks of pure TiO₂ at 458.7 and 464.5 eV are owing to the binding energies of Ti⁴⁺ 2p _3/2 and Ti⁴⁺ 2p _1/2, respectively. And there is no divided peak in the Ti 2p spectrum of pure TiO₂ sample. The Ti 2p _3/2 peak of the pure TiO₂ powder located at higher binding energy than that for the N-TiO₂ and N-S-TiO₂ samples. The corresponding peak of N-S-TiO₂ sample is obtained at lowest binding energy of 458.3 eV by comparison. This result indicates that the electronic interaction of Ti with anions in pure TiO₂, N-TiO₂, and N-S-TiO₂ samples is different. The partial Ti⁴⁺ might have electronic interaction with nitrogen when nitrogen was doped into TiO₂ lattice. The lower electronegativity of nitrogen compared to oxygen would result in a decrease of electron density around N anion, and thus the electron density on Ti increased. ^{Reference Chen and Burda32} This is why the binding energy of Ti 2p in N-S-TiO₂ sample decreased, which displays nitrogen is incorporated into the TiO₂ lattice.

Figure 4(c) shows the O 1s peaks of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples, which are all around the binding energy of 530 eV, indicating that O element exists mainly in the form of O²⁻. The O 1s spectrum consists of two peaks, which are around 530 and 532 eV, respectively. The strong peak around 530 eV can be attributed to the O–Ti–O lingkage of oxygen coordination, while the weak one around 532 eV may arise from the chemisorbed species from the ambient. ^{Reference Jagadale, Takale, Sonawane, Joshi and Patil33} The O 1s core level of N-S-TiO₂ sample is broader than that of pure TiO₂ and N-TiO₂ samples and has a shift to 529.5 eV, which might be due to the presence of oxygen and nitrogen in the same lattice units of TiO₂. ^{Reference Jagadale, Takale, Sonawane, Joshi and Patil33}

A comparison on the N 1s peaks of pure TiO₂, N-TiO₂, and N-S-TiO₂ is shown in Fig. 4(d). There is no peak in the N 1s spectrum of pure TiO₂ sample, while the other two samples both present a peak at 400 eV, which can be designated to the incorporation of nitrogen into TiO₂. The N 1s peak of N-S-TiO₂ sample is significantly stronger than that of N-TiO₂ sample. The nitrogen content of 2.392 at.% was detected in the N-S-TiO₂ sample, while that in the N-TiO₂ sample was only 1.160 at.%, which implies that the nitrogen content of the samples can be improved by the synergistic effect of urea and nitrates because of the addition of NaNO₃ and KNO₃ mixed nitrates compared with using urea as a single nitrogen source. And the use of two kinds of nitrates in this work also can increase the content of nitrogen of solid, because their different melting point and decomposition temperature can lead to the incorporation of nitrogen at different temperatures during the calcination process.

Specifically, the nitrogen 1s core levels of N-TiO₂ and N-S-TiO₂ samples can be deconvoluted into two peaks, in which one peak appears at 399.4 eV for N-TiO₂ and at 398.9 eV for N-S-TiO₂, and another peak appears at 400.4 and 400.3 eV for N-TiO₂ and N-S-TiO₂ samples, respectively. In literature, it was reported that nitrogen from simple chemisorbed nitrogen or TiN should appear at ≤397.5 eV, and nitrogen in NO species must appear above 400 eV. ^{Reference Sugai, Watanabe, Kioka, Miki and Kawasaki34–Reference Rodriguez, Jirsak, Dvorak, Sambasivan and Fischer36} Here the peak at 398.9 and 399.4 eV (for N-TiO₂ and N-S-TiO₂ samples, respectively) can be attributed to the anionic N incorporated in TiO₂ in O–Ti–N linkages. The existence of a peak at or near 400.3 eV can be ascribed to the nitrogen in NO or oxidized nitrogen in the form of Ti–O–N or Ti–N–O linkages. ^{Reference Sathish, Viswanathan, Viswanath and Gopinath37} Moreover, the peak at or near 400.3 eV of N-S-TiO₂ sample is stronger than that of N-TiO₂ sample. From this fact, it can be deduced that the mixed nitrates method may enable TiO₂ surface containing more NO due to the synergistic effect of urea and nitrates, and the strong oxidation ability of NO is beneficial for the photocatalytic activity of N-S-doped TiO₂.

B. Morphological structure

Typical SEM images of pure TiO₂, N-TiO₂, N-S-TiO₂ samples are shown in Fig. 5. Although all the three samples have a large number of inhomogeneous blocks, the N-S-TiO₂ sample has a relatively small mean particle size. There are many irregular small grains adhering to the surface of the inhomogeneous blocks in all the three samples. Moreover, the particle surface of pure TiO₂ is rather smoother [see Fig. 5(a)], while there are more small grains in both N-TiO₂ [see Fig. 5(b)] and N-S-TiO₂ [see Fig. 5(c)] samples. And the N-S-TiO₂ sample has more uniform particle size and less large blocks. A high magnification SEM image of N-S-TiO₂ sample is shown in Fig. 5(d), where many nanoparticles with a diameter ranging from 10 nm to 5 μm are visible.

FIG. 5.

Typical SEM images of (a) pure TiO₂, (b) N-TiO₂, and (c) N-S-TiO₂ samples, and (d) high magnification SEM image of N-S-TiO₂ sample.

TEM and HRTEM were used to evaluate the particle size and crystalline structure of the three samples. Typical TEM and HRTEM images of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples are shown in Fig. 6. It is seen from this figure that all the three samples are composed of irregular nanoparticles. The nanoparticles are small grains adhering to the surface of the inhomogeneous blocks observed in SEM. The TEM image in combination with the inset shown in Fig. 6(a) indicates that the sample of pure TiO₂ has a particle size in the range of 13–23 nm. Figure 6(b) reveals the size distribution of N-TiO₂ sample ranging from 9 to 16 nm. From Fig. 6(c), it was calculated that the mean particle size of N-S-TiO₂ sample was about 7 nm. These results imply that the present method of nitrogen doping can reduce the average particle size of the samples, and the mixed nitrates method can even make the particle size become smaller. It corroborated the results of BET. In addition, the clear mesoporous structure composed of titania nanoparticles is easily found. The formation of anatase can produce grain boundaries. The grain boundary might inhibit the growth of TiO₂ grains of anatase phase as well as the degradation of mesostructure during calcinations. ^{Reference Yu, Zhang, Zheng and Zhao38}

FIG. 6.

TEM (a, b, c) and HRTEM (d, e, f) images of pure TiO₂, N-TiO₂, and N-S-TiO₂ samples, respectively. The insets in (a–c) show the particle size distribution of the three samples.

Figures 6(d)–6(f) displays the HRTEM micrographs of the nanoparticles in pure TiO₂, N-TiO₂, and N-S-TiO₂ samples, respectively. From the micrographs, the lattice spacing of each sample can be measured. The spacing of two neighboring parallel fringes is about 0.35 nm for all the three samples, which matches well with the d-spacing between adjacent (101) crystallographic planes of TiO₂ anatase phase, ^{Reference Selvam, Balachandran, Velmurugan and Swaminathan39} is consistent with the data of XRD. Additionally, the shape, edge, and lattice fringes of TiO₂ nanoparticles could be clearly observed, and the direction of lattice fringes is unidirectional, which indicates the N-S-TiO₂ photocatalyst with single-crystal-like structure and high crystallinity can be obtained by this method. It was reported that the photogenerated charge carriers transport pathways were affected sensitively by crystal boundary of photocatalysts. ^{Reference Li, Chen, Sun, Rao, Han, Hu and Zhou11} And it can be presumed that the N-S-TiO₂ nanoparticles are single-crystal-like structure that might be directly related to the effect of morphology modifier of nitrates in the mixed nitrates system. The mixed nitrates could directly result in suppression effect of intrinsic anisotropic crystal growth owing to the relative slow nucleation and growth rates. Compared to polycrystalline TiO₂, the single-crystal-like TiO₂ had lesser crystal boundary, which facilitated long-range electron transport and also reduced the number of defects acting usually as recombination centers of photogenerated charge carriers, thus effectively increasing the lifetime of electrons and holes.

C. Photocatalytic properties

The UV–Vis absorption spectra of pure TiO₂, N-TiO₂, N-S-TiO₂, and P25 samples are presented in Fig. 7(a). The pure TiO₂ and P25 samples display an intense transition in the UV region mediated by the intrinsic band gap absorption of TiO₂, resulting from the electron transitions from the valence band (VB) to conduction band (CB) (O 2p → Ti 3d). ^{Reference Xiang, Yu and Jaroniec40,Reference Ong, Tan, Chai, Yong and Mohamed41} In addition, the absorption in ultraviolet spectrum of N-TiO₂ and N-S-TiO₂ samples is slightly stronger than that of the pure TiO₂ sample, which may be due to that the nitrogen doping can reduce the particle size and increase the specific surface area of the samples, thus improving their absorption capability on light and reducing the reflectance. After N was doped into the samples, a clear red shift was observed and the absorption increased significantly in the visible range. The introduction of nitrogen into the crystalline lattice of TiO₂ in the forms N-Ti-O or O-N-Ti linkages would modify the electronic band structure of TiO₂ leading to the formation of N-dopant energy levels above the O 2p valence band (VB), thus reducing the energy gap and the energy required for electronic transitions, resulting in the absorption of visible light. ^{Reference Selvam, Balachandran, Velmurugan and Swaminathan39} In other word, the incorporation of nitrogen into the TiO₂ lattice caused the absorption of visible light. In Fig. 7(a), a remarkable absorption was observed in the visible region between 400 and 600 nm for N-TiO₂ and N-S-TiO₂ samples, indicating that N atoms were successfully incorporated into the TiO₂ lattice, which is consistent with the results presented in Sec. III. A and III. B.

FIG. 7.

(a) UV–Vis absorption spectra of pure TiO₂, N-TiO₂, N-S-TiO₂, and P25 samples, and (b) their corresponding Kubelka–Munk curves.

The band gap energy of the samples can be evaluated with the Kubelka–Munk function. The respective Kubelka–Munk curves of pure TiO₂, N-TiO₂, N-S-TiO₂, and P25 samples are obtained via the transformation based on the (Ahν)^1/2 versus energy of absorbed light is shown in Fig. 7(b). The calculated band gap values of pure TiO₂, N-TiO₂, N-S-TiO₂, and P25 samples listed in Table II. It is seen that the band gap of N-TiO₂ and N-S-TiO₂ samples are narrower than that of pure TiO₂ and P25, indicating that there are some N-dopant energy levels due to the nitrogen dopants being introduced in the band gap of TiO₂, which is in accordance with the red shift shown in Fig. 7(a) and is beneficial to improving the photocatalytic efficiency of the obtained catalysts. Furthermore, the band gap of N-S-TiO₂ sample is slightly wider than that of the N-TiO₂. It is speculated that the mixed nitrates method did not make hybrid nitrogen increase, but led to the surface of TiO₂ containing more NO, as confirmed by XPS analysis. But the strong oxidizing ability of NO would be beneficial for improving the photocatalytic activity of catalysts. Meanwhile, the wider band gap of N-S-TiO₂ sample could enhance reduction ability of electrons and oxidizing ability of holes, finally resulting in improvement of photocatalytic activity.

TABLE II.

Band gap energy (E _g) of pure TiO₂, N-TiO₂, N-S-TiO₂ and P25 samples.

The photocatalytic activities of the pure TiO₂, N-TiO₂, and N-S-TiO₂ samples are evaluated by the degradation of MB versus irradiation time under visible light and the results are shown in Fig. 8(a). A blank study without catalyst was used as a background check. And P25 catalyst was also studied under the same conditions as a comparison. It can be seen that there is a slight self-degradation of MB solution without catalysts under visible light after 150 min, and the degradation efficiency of MB is only 11%. In contrast, the great degradation is clearly observed when pure TiO₂, N-TiO₂, and N-S-TiO₂ catalysts are present. The N-S-TiO₂ sample has much better photocatalytic activity than that of N-TiO₂ sample. The N-S-TiO₂ sample takes 60 min to degrade 83% MB, presenting the highest efficiency in all the applied catalysts, while the N-TiO₂ sample degrade only 22% MB for the same time, which is even lower than that by P25 (degrading 49% of MB). In addition, pure TiO₂ sample shows very weak photocatalytic activity. The linear relationship between the reaction time versus −ln(C/C ₀) is shown in Fig. 8(b), which indicates that the degradation of MB follows the first-order kinetics. The values of the rate constants can be calculated from the slope of the lines between −ln(C/C ₀) and the reaction time, which are 0.00024, 0.0032, 0.013, and 0.013 min⁻¹ for the reactions with pure TiO₂, N-TiO₂, N-S-TiO₂ and P25 samples. The degradation rate of N-S-TiO₂ sample is the same as that of P25 sample, which indicated that the N-S-TiO₂ sample have the best photocatalytic activity. In addition, the photostability of N-S-TiO₂ sample was also studied under the same conditions. ^{Reference Niu, Cheng and Cao42,Reference Cheng, Chen, Lin, Liu and Kong43} As shown in Fig. 8(c), N-S-TiO₂ sample still kept a nice activity and showed a good stability on the degradation of MB under visible light after 3 successive runs.

FIG. 8.

(a) Degradation curves of MB by the catalysis of pure TiO₂, N-TiO₂, N-S-TiO₂, and P25 samples under visible light irradiation, (b) their corresponding plots of −ln(C/C ₀), and (c) the degradation of MB on N-S-TiO₂ sample under visible light for 3 runs.

The investigation of the reactive oxidation species in photocatalytic process is very important to the explication of the photocatalytic mechanism. To trap the reactive oxidation species in experiments, under the same experiment condition for each photocatalytic test on MB by the N-S-TiO₂ samples, 1 mM TBA and BQ, as scavengers of hydroxyl radicals $\left( { \cdot {\rm{OH}}} \right)$ and superoxide radicals $\left( { \cdot {{\rm{O}}^{{\rm{2}} - }}} \right)$ , were added into the reaction system, respectively. ^{Reference Huang, Liu, Chen, Zhang, Zhang and Wang23–Reference Yan, Chen, Zou and Wang26} Fig. 9 shows the degradation curves of MB by the catalysis of N-S-TiO₂ sample in the presence of TBA or BQ under visible light irradiation. It can be seen that the photocatalytic activity of the N-S-TiO₂ sample was decreased in the presence of TBA as $\cdot {\rm{OH}}$ scavenger, and the inhibition efficiencies for degradation of MB is about 12.5%. However, the photodegradation of MB over the N-S-TiO₂ sample was almost not affected by the addition of 1 mM BQ as quencher of $\cdot {{\rm{O}}^{{\rm{2}} - }}$ . Thus, it could be speculated that $\cdot {\rm{OH}}$ serves as the main reactive oxidation species for the photodegradation of MB over N-S-TiO₂ sample under sunlight irradiation.

FIG. 9.

Photodegradation curves of MB by the catalysis of N-S-TiO₂ sample in the presence of TBA or BQ scavengers under visible light irradiation.

On the basis of the above discussion and results, a possible photocatalysis mechanism for the visible light photocatalytic reaction over N-S-TiO₂ photocatalyst is illustrated in Fig. 10. The gradient in the energy band represents mixing of the N 2p and O 2p atomic orbitals. ^{Reference Naik, Moon, Kim and Park44,Reference Wen, Li, Liu, Fang, Xie and Xu45} Nitrogen doping can create intra-band-gap states (N 2p) which is located above the valence band edges (O 2p), which increases the driving force of electron injected from N 2p energy level and accelerates the reductive process of interfacial electron transfer. ^{Reference Zhang, Wu, Xing, Leghari and Sajjad46,Reference Li, Liu, Luo, Li, Huang, Li, Fang, Xu and Zhu47} Therefore, the N-S-TiO₂ sample resulted in the best response to visible-light and the red-shift due to the effect of N dopant. In addition, the N-S-TiO₂ sample could be activated by the visible light, thus more electrons and holes are generated and participated in the photocatalytic redox reactions. And there are a lot of NO on the surface of the TiO₂, the strong oxidation ability of NO is beneficial for the photocatalytic activity of N-S-TiO₂. The major reactions in this photocatalytic mechanism are summarized by the following steps:

and

During the photocatalytic reaction, MB ions adsorbed on the active surface sites of N-S-TiO₂ photocatalyst were immediately oxidized by ‧OH radicals and NO.

FIG. 10.

Mechanism for photocatalytic degradation of MB over N-S-TiO₂ photocatalyst under visible light irradiation.