A series of N-doped TiO2 nanocrystals with exposed facets was prepared successfully by sol-gel method for the first time. The physicochemical properties of these resultant photocatalysts were characterized by XRD, TEM, XPS, and DRS, and their photocatalytic performances were evaluated by photocatalytic decoloration of methylene blue solution under visible light ( nm) irradiation. The results showed that the N-doped TiO2 nanocrystals with exposed facets showed higher photocatalytic activity than P25. The enhanced photocatalytic performance can be attributed to synergistic effects of several factors, such as good crystallinity, better light response characteristic, and high reactivity of facets.

1. Introduction

TiO2 with exposed facets has attracted great attention in recent years because of its outstanding advantages in photocatalysis field [13], such as excellent thermodynamic stability, strong oxidizing power, and low cost. Many researchers have participated in optimizing its photocatalytic activity since the large percentage of reactive facets were prepared by one-pot hydrothermal method using TiF4 as precursor firstly [4]. It is widely believed that facets of anatase TiO2 provide more active sites for photocatalytic oxidation reaction [5, 6], so its photocatalytic activity can be improved significantly. However, the photocatalytic activity of TiO2 is still limited by narrow light absorption (only responsive to UV with wavelength below 387 nm due to its wide band-gap) and rapid recombination of photon-generated carriers. To promote the better performance of TiO2, anion doping, especially nitrogen doping, would be effective solution. Experiments and theories have proved that the doping of nitrogen can introduce a new impurity level in the band gap of TiO2, which provides a springboard for electron transition and promotes the visible-light response of TiO2 [79].

The popular approach to synthesize TiO2 with exposed facets is hydro(solvo)thermal method [10, 11], and anion doped TiO2 (e.g., N, C, S) with exposed facets has been synthesized successfully by hydro(solvo)thermal method in recent years [1214]. The resultant samples are usually with a larger size, submicron, and even several microns [15], which is disadvantaged to the photocatalytic activity of TiO2. Therefore, anion doped TiO2 with exposed facets and smaller size is still desirable. Lots of publications have proved that sol-gel is an ideal method to prepare uniform TiO2 powder with smaller size. Therefore, sol-gel method may be a good attempt to obtain anion doped TiO2 nanocrystals with exposed facets [16]. However, the sol-gel preparation of anion doped TiO2 nanocrystals with exposed facets has not been reported so far.

In this study, we prepared N-doped TiO2 nanocrystals with exposed facets by using sol-gel method successfully. The as-synthesized N-doped TiO2 nanocrystals with exposed facets showed higher photocatalytic activity than P25 in the degradation of methylene blue (MB) dye under visible-light illumination, and the influenced factors of photocatalytic activity were discussed in detail. To our knowledge, this is the first time to prepare N-doped TiO2 nanocrystals with exposed facets successfully by using sol-gel method, and the excellent visible-light activity has rarely been reported.

2. Experimental

2.1. Preparation of N-Doped TiO2 Nanocrystals with Exposed {001} Facets

N-doped TiO2 nanocrystals with exposed facets were synthesized by a sol-gel method [17]. Firstly, 6.85 mL of tetrabutyl titanate was dispersed in 15.5 mL of absolute ethanol under vigorous stirring for 30 min in a conical flask with cover (solution A). Secondly, a given amount of NH4F (the molar ratio of Ti to N was controlled at 1 : 4, 1 : 2, 1 : 1, 2 : 1), 7.75 mL of absolute ethanol, 8 mL of acetic acid, and 2.88 mL of pure water were mixed in a beaker (solution B). Then, solution B was dropwise added to solution A with vigorous magnetic agitation. Subsequently, the obtained mixtures were kept in incubator chamber at 90°C for 12 h. The obtained gel from the sol-gel reaction was dried at 80°C in an oven and then porphyrized into powders. The as-obtained samples were named S1, S2, S3, and S4, respectively. In order to improve the crystallinity, the four samples were calcined at 500°C in a furnace for 2 h to obtain the samples labeled as S1-500, S2-500, S3-500, and S4-500, respectively.

2.2. Characterization

Powder X-ray diffraction (XRD) patterns were recorded at room temperature with an X’pert diffractometer (PANalytical, Holland) with copper radiation to determine the crystalline structure of the samples. SEM images were obtained by a field emission scanning electron microscopy (S-4800, Hitachi, Japan) equipment and TEM images were carried out on a transmission electron microscope (JEM 2100, JEOL, Japan) to characterize morphology of the samples. XPS analyses were tested on an X-ray photoelectron spectrometer (ESCALAB250, Thermo Scientific, USA) with aluminum Ka radiation to analyze the chemical nature of surface elements. UV-visible diffuse reflectance spectra (DRS) were obtained with a UV-visible spectrophotometer (UV-2450, Shimadzu, Japan) and the baseline correction was done using a standard sample of BaSO4 to describe the absorption of light at different wavelengths.

2.3. Photocatalytic Performance

The photocatalytic decomposition of methylene blue (MB) was carried out in a photo reaction system (as illustrated in our previous publication [18]). The visible light source was offered by a 1000 W Xe lamp equipped with a glass filter (removing the UV irradiation below 420 nm wavelength) positioned in the center of a water-cooled system. The reactive bottle is a 50 mL cylindrical vessel 4 cm away from the light source. In the bottom of the reactive bottle, a magnetic stirrer was rotated to achieve effective dispersion. The temperature of the reaction solution was maintained at 30 ± 0.5°C by cooling water. 50 mg of samples was added to 50 mL of 10 mg/L MB solution to form suspension. After the suspensions were stirred in the dark for 1 h to reach the adsorption-desorption equilibrium, the suspensions were irradiated with visible light and stirred continuously. At given time interval, 2 mL of suspension was taken out and immediately centrifuged to eliminate the solid particles. The absorbance of the filtrate was measured by a spectrophotometer at the maximum absorbance peak 665 nm.

3. Results and Discussion

3.1. XRD

The XRD patterns of four samples are displayed in Figure 1. It can be clearly seen from Figure 1(a) the variations of products with the different dosages of NH4F. The peaks at = 25.28°, 37.80°, 48.05°, 53.89°, 55.06°, and 62.69° belong to anatase TiO2 (reference no. 00-021-1272), and the peaks at = 14.08°, 23.57°, 27.56°, 28.37°, 33.59°, and 49.65° are corresponding to NH4TiOF3 (reference no. 00-052-1674). From the contrast of four samples, we can see all the peaks of S1 and S2 belong to anatase TiO2, indicating that only anatase TiO2 formed when the dosages of NH4F are relative low. With the increasing of the dosage of NH4F, new peaks corresponding to NH4TiOF3 appear in the XRD patterns of S3 and S4, indicating that both anatase TiO2 and NH4TiOF3 formed when the dosages of NH4F are relative higher, and the anatase TiO2 contents in S3 and S4 are 75.3% and 60.1%, respectively. Figure 2(b) presents the XRD patterns of the samples after calcination. Both of the two samples (S3 and S4) are transformed into anatase TiO2 because of the heating effect. Compared with the samples without calcination treatment, the diffraction peak intensities of anatase TiO2 are increased and the Full Width Half Maximums (FWHM) are obviously narrowed due to heat treatment, which demonstrates that the crystallization degree of anatase TiO2 is enhanced due to the calcination at 500°C for 2 h. The average crystal size of all samples was calculated using the Scherrer equation where is the average crystal size, is a constant (shape factor, about 0.89), is the X-ray wavelength, is the FWHM of the diffraction line, and is the diffraction angle. The calculated results are 16.89, 17.70, 37.73, and 25.26 nm for S1-500, S2-500, S3-500, and S4-500, respectively.

The above results can be explained by the following reasons. Because of the high hydrolytic activity of tetranbutyl titanate, it is necessary to slow down the speed of hydrolysis by adding depressor such as acetic acid. In the reaction, acetate is used as a ligand for titanium to control the rate of tetranbutyl titanate hydrolysis effectively; thus the formation of gel can be stable and controllable. Then, the coordinate polymers are transformed into TiO2 by condensation polymerization [16]. The followings are the reaction approaches: It is adverse to the production of TiO2 when the dosage of NH4F is excessive, and the reason can be ascribed to the following reactions:

3.2. TEM

Figure 2 shows the TEM images of the four calcined samples (the TEM images of uncalcined samples were not shown here due to their obscure facets). As shown in Figure 2(a), when the NH4F dosage is a quarter of titanium in amount, TiO2 nanocrystals with the size of 15~20 nm present square outline, which is different from the morphology of TiO2 nanocrystals prepared by traditional sol-gel method where crystal surface control agent (F) was absent [17], indicating the formation of nanocrystals with exposed facets [19]. In Figure 2(b), TiO2 nanocrystals show quite distinct square outline as the increasing of the dosage of NH4F, and the nanocrystals are about 20 nm in length, about 15 nm in thickness, and about 40% in the percentage of facets. When the NH4F dosage is equal to titanium in amount, nanocrystals appear in an irregular shape and some nanocrystals are aggregated to some extent, which can be obviously seen from the TEM image (Figure 2(c)). When the NH4F dosage is double that of the titanium, the shapes of crystals are more irregular, and the edges are obscure; the reason can be ascribed to the production of NH4TiOF3. That is, the anatase TiO2 produced in the sol-gel process can likely show facets after calcination. The inset in Figure 2(b) (HR-TEM of S2-500) clearly shows the continuous (004) atomic planes with a lattice spacing of 0.238 nm, corresponding to the planes of anatase TiO2 single crystals, which further confirms that the exposed crystal facet is facet.

3.3. XPS

Figure 3(a) displays the XPS survey scan patterns of typical samples S2 and S2-500. Two significant differences can be observed by comparing the two samples. One of them is that the two peaks centered at 830.0 eV and 685.2 eV, which can be assigned to F (A) and F 1s, respectively, appear in uncalcined sample, and are removed due to calcination [20], indicating F groups anchored on the surface of S2 can be removed by calcinations at 500°C for 2 h. The other one is that the peak centered at about 400.0 eV is weakened, which can be ascribed to the removal of adsorbed nitrogen groups due to calcination. The small peak at about 400.0 eV in sample S2-500 can be ascribed to N 1s, implying that N-doping has been realized. As shown in Figure 3(b), two peaks can be obtained by fitting the N 1s spectrum of S2. The peak at 399.7 eV (peak 1) can be assigned to Ti-(N-O) bond [2124], and the peak at 401.4 eV (peak 2) is attributed to the interstitial N atoms in N-O bonds [25], indicating that some nitrogen groups are adsorbed on the surface or interspace of TiO2. In Figure 3(c), two peaks at 399.7 eV (peak 1) and 402.9 eV (peak 3) can be fitted from the N 1s spectrum of S2-500, which can be assigned to Ti-(N-O) bond and bond [23, 26], implying that nitrogen atoms are doped in the lattices of TiO2. The doping amount of nitrogen is 0.52% in atom.

3.4. DRS

Figure 4 gives the UV-Vis diffuse reflectance spectra of all samples. The optical absorption edges of S1, S2, S3, S4, and S1-500 locate at 420 nm, which is similar to pure TiO2 (Figure 4(a)). However, the optical absorption edges of S2-500, S3-500, and S4-500 shift to the lower-energy region (about 550 nm) due to the N-doping (Figure 4(b)). These phenomena can be explained by combining UV-Vis diffuse reflectance results and XPS analysis results. Before calcination, nitrogen groups are just adsorbed on the surface or interspace of TiO2, not doped into the lattices of TiO2, which cannot change the band gaps of S1, S2, S3, and S4. Therefore, the optical absorption edges of the four samples do not shift towards long wavelength. After calcination, the optical absorption edge of S1-500 still locates at 420 nm, which can be ascribed to the low NH4F dosage. The optical absorption edges of S2-500, S3-500, and S4-500 are shifted to the lower-energy region, implying that nitrogen atoms are doped into the lattices of TiO2 due to the heat-treatment at high temperature and enough NH4F dosage. It is widely accepted that the nitrogen incorporation into the crystal matrix of TiO2 modifies the electronic band structure of TiO2, leading to a new substitution N 2p band formed above O 2p valance band, which narrows the band gap of TiO2 and shifts optical absorption to the visible light region [8, 12, 27]. In conclusion, the shift of optical absorption edge can be ascribed to the formation of impurity states in the band gap due to N-doping, which is realized by calcination treatment.

3.5. Photocatalytic Performances

Figure 5 shows the photocatalytic performance of all samples (due to low photocatalytic efficiency, the photocatalytic performances of these uncalcined samples were not shown). For comparison, P25, a standard photocatalyst, was also tested at the same conditions. From the comparison of the as-prepared four samples, it can be concluded that the photocatalytic performance of S2-500 is the optimal, and its photocatalytic performances is superior to P25 to a large extent. The high photocatalytic performance of S2-500 can be attributed to its high crystallization, better light response characteristic, and high reactivity of facets.

3.5.1. Crystallization

Crystallization plays a significant role in the photocatalytic activity of TiO2. A better crystallization means the decrease of crystal defects, which are the recombination centers of photoinduced charges [28, 29]. Therefore, in a sense, the improvement in photocatalytic activity of TiO2 can be achieved by increasing the crystallization of TiO2 [30], while, heat treatment is one of the methods to improve the crystallization of TiO2 [31, 32]. In current work, the photocatalytic performances of calcined samples are higher than those of the corresponding uncalcined samples, which can be attributed to the fact that the calcined samples possess better crystallization than the corresponding uncalcined samples, which has been confirmed from their XRD patterns.

3.5.2. Light Response Characteristic

It is well known that light absorption characteristic of photocatalyst is an important factor influencing photocatalytic activity. The enhancement of absorbance in the UV-Vis region increases the number of photogenerated electrons and holes to participate in the photocatalytic reaction, which can enhance the photocatalytic activity of TiO2 [33, 34]. It has been confirmed from Figure 4 that the optical absorption edge of samples is shifted to 550 nm from 420 nm due to the calcination treatment, which is one of the reasons why the photocatalytic performances of calcined samples are higher than those of the corresponding uncalcined samples.

3.5.3. High Reactivity of {001} Facets

The surface energy of facets is higher than that of facets due to the 100% surface unsaturated atoms on facets while only 50% for the facets. Therefore, the facets are theoretically considered more reactive than and facets [35, 36], which is in agreement with most experimental results. Under Wu’s experimental conditions, the higher the percentage of facets, the higher the photooxidation reactivity, for the TiO2 crystals with the same size [6]. D’Arienzo et al. think that surfaces can be considered as oxidation sites with a significant role in the photooxidation of TiO2 [37]; the concentration of trapped holes (O centers) increases with increasing surface area and photoactivity in vacuum conditions. In our experiments, the high reactivity of S2-500 can due to its high percentage of facets and small size.

3.6. Factors Influencing the Photocatalytic Degradation Efficiencies

Sample S2-500 is used as the model catalyst to study the factors which influence the degradation efficiencies.

3.6.1. Effect of pH

The decoloration efficiencies of MB at different pH values are displayed in Figure 6. It shows that both the adsorption capacity and photocatalytic degradation efficiencies will rise as pH increases. The reasons can be interpreted from the following aspects.

Firstly, pH value may influence the surface electrical behavior of TiO2 [38]. In general, the more negative of zeta potential indicates that there are more hydroxides on the surface of TiO2. The point of zero charge (pzc) of S2-500 is at pH 4.8; thus, the TiO2 surface is positively charged when pH < 4.8 while it is negatively charged when pH > 4.8. It is believed that organic molecules can be adsorbed more easily onto the surface of samples with the increasing of pH.

Secondly, the changes of pH can affect the yield of radicals in the photocatalytic oxidation. The holes are considered as the major oxidation specie at low pH value whereas radicals are considered as the predominant specie at medium or high pH levels [39, 40]. It is known that radicals are easier to be generated on TiO2 surface in alkaline environment; hence the photocatalytic activity is enhanced.

Thirdly, the stability of organic pollutants can be transformed by pH value. It is reported that MB is more stable in acid solution than that in alkaline solution [41]. Therefore, MB molecules can be degraded more easily as pH increases.

3.6.2. Effect of Catalyst Concentration

Figure 7 shows that the effect of catalyst concentration on the photocatalytic performances. It clearly displays that the adsorption capacity can be improved as the increasing of catalyst concentration. However, the photocatalytic degradation rates will not increase when the concentration exceeds a certain value. Generally, the more the dosage of catalyst, the higher the photocatalytic rate. However, excess catalysts are not favorable to boost the reaction speed for spare catalysts will shut out visible light. The optimal catalyst concentration is 1.0 g·L−1 in the case of our experiment condition.

3.6.3. Effect of Initial Substrate Concentration

It can be observed from Figure 8 that the photocatalytic activity of the catalyst can be influenced by substrate concentration. The degradation efficiencies decline with the increasing of substrate concentration, which can be ascribed to the following two reasons. One cause is that the generation of radicals on the surface of catalysts will be reduced because the active sites will be covered by organic ions for high substrate concentrations. Another reason is the light-screening effect of high substrate concentration, especially for dye solution; a considerable amount of light may be absorbed by dye molecules rather than TiO2 particles, and thus it will reduce the efficiency due to the low production of radicals [42, 43].

3.6.4. Effect of Substrates

In order to confirm the universality of the catalyst, two more typical organics are used as target pollutants, namely, congo red (10 mg/L, CR) and diclofenac sodium (20 mg/L, DCF). Figure 9 shows the photocatalytic performance of S2-500 via degrading three different organic pollutants. It can be seen that the removal rates all exceed 90%. From the result, we could deduce that the synthesized nitrogen doped TiO2 with exposed facets possesses excellent photocatalytic performance in the degradation of organic pollutants.

4. Conclusions

In this study, nitrogen-doped TiO2 nanocrystals with exposed facets were prepared by sol-gel method firstly. The as-obtained N-doped TiO2 nanocrystals with exposed facets showed higher photocatalytic activity than P25 in the degradation of methylene blue dye under the visible-light illumination. The enhanced photocatalytic performance can be attributed to synergistic effects of several factors, such as good crystallinity, better light response characteristic, and high reactivity of facets.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publishing of this paper.


This work was supported by the Key Project of Science and Technology Plan of Fujian Province (no. 2012Y0066), Xiamen Distinguished Young Scholar Award (no. 3502Z20126011), and Xiamen Science & Technology Major Program (no. 3502Z20131018). The valuable comments of anonymous reviewers are greatly appreciated.