Facile synthesis of C, N-TiO2 nanorods via layered Ti 3 O 7 2 − -TMAH interlaminar bonding interaction and their enhanced catalytic performance

A green and efficient photocatalyst based on C and N co-doped titanium based nanorods (NRs) was prepared by facile hydrothermal synthesis and interlaminar bonding interaction between layered Ti 3 O 7 2 − and tetramethylammonium hydroxide (TMAH). TMAH as one of quaternary ammonium-based compounds was used as the doping source of C and N elements. The results showed that C and N co-doping did not affect the anatase crystal structure of TiO2 NRs. C, N-TiO2 NRs showed excellent photocatalytic activity for degrading methyl orange under simulated sunlight irradiation. In particular, the 20:1 TiO2 NRs catalyst showed the best catalytic performance, showing efficient photocatalytic rate of 90%. The photocatalytic reaction kinetics of undoped and doped catalysts was also studied. In addition, reactive species (RSs) experiments showed that hydroxyl radical (·OH), superoxide radical anions (·O2−) and photogenerated holes (h+) were major active species during the photocatalytic process. The mechanisms of the electrostatic attraction and potential photocatalytic degradation were proposed and discussed.


Introduction
More than 10 000 kinds of dyes per year, and moreover, the demand for printing and dyeing products has increased substantially in recent years [1][2][3]. However, the entry of dye wastewater into rivers and the food chain can have harmful effects on human health [4].
Photocatalysis has been widely employed as a green, non-secondary wastewater agent especially for dyeing wastewater treatment [5][6][7][8][9]. Photocatalytic nanomaterials have excellent degradation performance due to their high specific surface area [10][11][12]. In particular, TiO 2 nanomaterials appear to be a promising green catalyst due to its environmental safety, low cost and no secondary pollution [13][14][15][16]. However, low visible light utilization, the rapid recombination of electron-hole pairs and few surface reactive sites have obviously limited their practical applications [17].
In this respect, some non-metallic organic compounds (e.g. urea [18][19][20] and polyaniline [21]), noble metals [22,23] and semiconductor composite [24] can be used to modify the surface of TiO 2 nanoparticles to enhance their photocatalytic activities, thereby effectively inhibiting the recombination of photoinduced carriers and broadening the absorption wavelength. Bonding interaction is an effective strategy for modifying catalyst materials. For instance, Wang et al prepared a nitrogen-doped TiO 2 /graphene nanohybrid through self-assembly Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. of pyrene modified H 2 Ti 3 O 7 nanosheets and graphene via the π−π stacking interactions [25]. Ota et al reported that ethylenediamine-modified H 2 Ti 3 O 7 nanotube exhibit bonding interactions between the amines and OH groups on the surface of H 2 Ti 3 O 7 [26]. Due to the electronegativity of ions, the ions were adsorbed into the interlayer structure of nanotubes, leading to an improved photocatalytic performance of TiO 2 catalysts. However, interlaminar bonding interactions between layered titanates and TMAH (C-N doping source) has not been previously reported.
In this study, using TMAH as doping sources for C-N non-metallic elements, titanium-based nanorods were prepared and exhibited efficient photocatalytic degradation of dyes under simulated sunlight. This work can provide useful information and promote further interest in the preparation of C/N doped TiO 2 nanomaterials for photocatalytic degradation visible light irradiation.

Materials
All chemicals reagents were of analytical grade. P25 TiO 2 (80% anatase and 20% rutile crystalline phases) was supplied by Evonik Degussa GmbH. Ultrapure water with a resistance of 18.25 MΩ was provided by a water system (Milli-Q). NaOH, HCl (36%-38%), methanol and absolute alcohol were purchased from Chengdu Kelong Chemical. 25% aqueous solution of TMAH and Na 2 C 2 O 4 were purchased from Sinopharm Chemical Reagent. K 2 Cr 2 O 7 was provided by Macklin Chemical Reagent. Benzoquinone was purchased from Shandong Xiya Reagent. Methyl orange (MO) was purchased from Chengdu Kelong Chemical.

Synthesis of H 2 Ti 3 O 7 template
The synthetic process of H 2 Ti 3 O 7 nanotubes (NTs) was given as follows [27]. 22 g of NaOH was dissolved into 55 ml of ultrapure water to obtain solution. 1 g of P25 TiO 2 NPs was added into the solution. Then, the slurry was stirred vigorously at room temperature for 3 h. The resulting slurry was poured into a reactor, in which was heated at 150°C for 36 h. Thereafter, the obtained precipitates were washed with anhydrous alcohol and ultrapure water. Subsequently, 0.1 M of hydrochloric acid was added to the above obtained product and then vigorously stirred for 4 h to exchange Na + ions into H + . After the reaction, the titanate nanotubes were rinsed with ultrapure water to make neutral. Then, the precipitates were dried at 80°C for 10 h. Finally, H 2 Ti 3 O 7 template was obtained, which was porphyrized by an agate mortar.
2.3. Synthesis of C, N-TiO 2 nanorods C, N-TiO 2 nanorods were prepared via facile hydrothermal synthesis and interlaminar bonding interaction method by using TMAH as the C and N sources. Briefly, H 2 Ti 3 O 7 NTs and 25% TMAH aqueous solution were added into 100 ml ultrapure water (mol ratios of Ti and N element were kept at 20:0.5, 20:1, 20:5, 20:10 and 20:15, respectively). The mixture was stirred for 6 h. The precipitates were rinsed with ultrapure water, which were dried at 80°C for 10 h and then calcined at 600°C for 5 h. Finally, the required samples were synthesized.

Photocatalytic reactions
25 mg of catalyst was dispersed in 50 ml MO aqueous solution (15 mg l −1 ). After an adsorption-desorption equilibrium for 30 min, the MO aqueous solution was irradiated under simulated sunlight. During the photocatalytic process, samples were taken at intervals. The slurry was put into the centrifugal tube 4 ml at a rate of 12 000 rpm min −1 for 3 min to remove the particles. The maximum absorbance of the sample was measured by UV-vis spectrophotometer. Finally, the percentage removal efficiency (%) were calculated as the following equations.  This phenomenon demonstrated that the N doping process did not change the crystal structure of anatase phase, which indicated considerable stability of the TiO 2 sample. It is noteworthy that the (101) lattice plane gradually shifted to the high angle region with the increase of N content ( figure 1(b)). Numerous studies showed that the doping of nitrogen would cause the shift of the diffraction peak of TiO 2 [28,29]. Therefore, the reason is that N doping can cause lattice distortion, which lead to changes in the lattice parameters of TiO 2 . Moreover, the 20:1 TiO 2 NRs showed a better crystallinity than that of 20:5 and 20:15 TiO 2 NRs. The reason was that a large amount of TMAH in the precursors can retard the formation of Ti-O-Ti network and decrease the crystallinity of TiO 2 . Figure 2 showed an SEM image of the agglomerated TiO 2 nanorods. The TiO 2 nanorods had an excellent one-dimensional rod-like shape. As observed, the diameter and length of TiO 2 nanorods were observed in the range of 15-20 nm and 50-200 nm, respectively. The length of the nanorods varied greatly due to the fracture of the nanorods during the calcination process. The EDS of 20: 1 TiO 2 NRs shown in figure 3 showed the presence of Ti, O, C, and N elements. The presence of Pt was attributed to the sample holder of the EDS.
In order to further study the changes of chemical bonds in modified catalysts, the elemental composition and superficial chemistry of C, N-TiO 2 NRs were characterized by x-ray photoelectron spectroscopy (XPS). As indicated in figure 4(a), the C 1s consisted of four characteristic peaks. And the binding energies at 284.6, 285.2, 286.0 and 288.4 eV were vested in C-C, C-N, C-OH and C=O bonds, respectively, indicating that the catalyst contained C and N elements [30]. As shown in figure 4(b), the N 1s spectrum included two secondary peaks,  which were located at 399.6 and 400.5 eV, corresponding to the N-Ti-O and N-C, respectively. The characteristic peak at 402.36 eV was related to the fragments of C and N molecules, which might be the residual organic molecule fragments of quaternary ammonium [10,29,30]. Moreover, in figure 4(c), the O 1s displayed three peaks at 529.5, 529.9 and 530.8 eV corresponding to Ti 4+ -O, Ti-OH and N-Ti-O bonds, respectively. It showed that N element was doped into the TiO 2 lattice in the form of replacing oxygen atom. Ti-OH bonds on the surface of TiO 2 promoted the formation of ·OH during photocatalytic reaction [29,30]. The Ti 2p spectrum ( figure 4(d)) was split into four peaks at 458.2, 458.6, 463.7 and 464.4 eV which belonged to Ti 3+ 2p 3/2 , Ti 4+ 2p 3/2 , Ti 3+ 2p 1/2 and Ti 4+ 2p 1/2 bonds, respectively [10,29,30]. In addition, the atomic percentage of the catalyst elements obtained by XPS is shown in table 1.

Synthesis of co-doped NRs via interlaminar bonding interaction
TEM was used to characterize the morphology and microstructure of H 2 Ti 3 O 7 NTs (figures 5 and 6). As shown in figure 5(a), typical nanotubes of H 2 Ti 3 O 7 from several tens to hundreds of nanometers were observed in the as-prepared H 2 Ti 3 O 7 sample. Structures of 3-5 layers were also observed on the tube wall with the thickness of each layer of ∼0.8 nm, as shown in figure 5(b). The outer diameter ranged in 5-15 nm and inner diameters in   3-10 nm (figure 6). The above results revealed that the layered structure provided a large specific surface area, which can offer adsorption sites in the intermediate layer, thereby improving the photocatalytic efficiency. Figure 7 illustrated the interionic adsorption process between tetramethylammonium cations and layered H 2 Ti 3 O 7 nanotubes. In the synthesis process, tetramethylammonium cations were attracted by layered H 2 Ti 3 O 7 nanotubes with a negative charge through electrostatic interaction. The interlayer spacing was 0.7-0.8 nm, which was in agreement with previous report [27]. The tube wall consisted of 3-5 layered structures. The surface negative charge was provided by 2 structures offered adsorption sites to the TMAH cation. Tetramethylammonium cations of ∼0.3 nm size can enter the space between the two layers. Therefore, the N and C elements from tetramethylammonium cations were distributed uniformly in the layered H 2 Ti 3 O 7 nanotubes. Figure 8 showed the UV-vis diffuse-reflectance spectra of the catalysts. Pure TiO 2 NRs only showed light absorption within the UV region at an absorption edge of 388 nm. As expected, the C, N-TiO 2 showed an enhanced visible light absorption, that the absorption edge showed a slight red shift to 411 nm, as shown in figure 8. This indicated that the band gap was narrowed by the non-metal element doping. Based on the above results, the estimated bandgaps of pure TiO 2 NRs and C, N-TiO 2 NRs were 3.20 eV and 3.02 eV, respectively.  In the presence of a 20: 1 TiO 2 NRs catalyst in an aqueous solution at pH 7.0, the photocatalytic of MO was studied by UV-visible absorption spectra, as shown in figure 9(a). In the UV-vis spectra, obvious absorption peak change was observed at 464 nm. The degradation rate of MO in the presence of TiO 2 NRs and C, N-TiO 2 NRs are given in figure 9(b). Compared with pure TiO 2 NRs, C, N-TiO 2 NRs showed a slight adsorption of methyl orange. Under simulated sunlight irradiation for 150 min, 20:1 TiO 2 nanorods showed the highest photocatalytic activity, reaching 90% degradation efficiency.

Photocatalytic activity
The results indicated that small amount of C-N co-doping accelerated the photocatalytic oxidation reaction. Doping of non-metallic elements can change the valence band level and improve visible light response, which was consistent with the obtained XPS spectrum [31,32]. Doping a certain amount of C or N into the TiO 2 lattice, the sample showed better photocatalytic activity. It can be observed in the O1s spectrum that as the doping amount increased, the N-Ti-O content also increased. However, the photocatalytic activity of the catalyst of 20: 0.5 was the weakest compared with the catalysts of other ratios. Because the doping amount of the N element in the 20: 0.5 catalyst was too low, the effect of the doping element was reduced. Moreover, when the doping amount of C and N exceeded the substitution ability of O in TiO 2 , redundant non-metallic atoms would remain on the surface of the doped TiO 2 nanorods. The C1s energy spectrum proved that no C-Ti bond was formed, but carbon layers covering the surface of the catalyst was generated. In this case, much active sites of photocatalyst would be shielded by excess C/N atoms, so that the photocatalytic activity was suppressed (figure S1 is available online at stacks.iop.org/MRX/7/025022/mmedia).
In order to better compare the photocatalytic rate of the above catalysts, reaction kinetics of degradation of MO were showed in figure 9(c). The fitting curves were obtained from the reaction of pseudo first order model, that the kinetic expression was −ln(C/C 0 )=kt [33,34]. As can be seen, the 20:1 TiO 2 NRs catalyst presented the highest apparent rate constant (k=1.46×10 −2 min −1 ), which was almost 7.9 and 8.3 times higher than those of the doped TiO 2 nanorods and undoped TiO 2 nanorods, respectively. Table S1 listed the comparison of the catalytic rate of C, N-TiO 2 NRs with other catalysts. Non-metallic of C and N elements doping can introduce impurity levels around at O 2p band level. It can inhibit the recombination of photon-generated carriers, which can effectively increase the electrons and holes used in the radical reaction [30,[35][36][37]. However, excessive N element doping would inhibit the catalytic rate, as shown in figure S2.
During the photocatalytic degradation of organic pollutants, various reactive species (RSs) have different functions when C, N-TiO 2 was used as catalyst. Methanol (1:15/V:V), benzoquinone (50 μM), Na 2 C 2 O 4 (10 mM), and K 2 Cr 2 O 7 (2 mM) were used to scavenge ·OH, ·O 2 −, h + and e − , respectively [33,38]. It should be pointed out through the addition performed discrepant inhibitory effects on the photocatalytic degradation of MO ( figure 9(d)). When no scavenger was added, 15 mg l −1 MO can be degraded within 150 min and the photocatalytic efficiency reached 90%. When K 2 Cr 2 O 7 was added, the photodegradation efficiency decreased slightly, which indicated that e − was not the main active species. Nevertheless, methanol, Na 2 C 2 O 4 and benzoquinone recorded obvious inhibitory effects during the degradation of MO with inhibition rate of 75.3%, 79.7% and 95.5%, respectively. In the catalytic reaction, the undoped TiO 2 showed the same active radical, as shown in figure S3. No other reactive groups were produced. Therefore, these results demonstrated that ·OH, h + and ·O 2 − were the major reactive species toward the degradation of MO, while e − showed a minor effect.
Based on the above dates, the potential photodegradable mechanisms were demonstrated in scheme 1. Under simulated sunlight, the valence band (VB) electrons of C, N-TiO 2 NRs were excited to the conduction band (CB), leading to the generation of h + and e − . According to the relatively strong electron affinity of N atoms in free tetramethylammonium cations, electrons in the CB can be easily transferred to tetramethylammonium cations, so that inhibited the recombination of e − /h + pairs. Subsequently, the e − contacted with H 2 O or O 2 in solution, accompanied by a series of free radical chain reaction, leading to the generation of ·O 2 −, that a fraction of ·O 2 −reacted with H + to generate ·OH. Moreover, a portion of h + reacted with H 2 O molecule, resulting in the generation of ·OH. Consequently, the production of RSs (·OH, ·O 2 − and h + ) would attack MO, leading to its efficient photodegradation. Thus, C, N-TiO 2 NRs showed excellent photocatalytic degradation performance.

Conclusion
In summary, novel carbon-nitrogen co-doped titanium based nanorods were prepared by facile hydrothermal synthesis and interlaminar bonding interaction. Tetramethylammonium cations were electrostatically attracted by layered H 2 Ti 3 O 7 nanotubes with a negative surface charge. Tetramethylammonium cations entered into the space between two layers of nanotubes. Therefore, N and C elements from tetramethylammonium cations were distributed uniformly in layered H 2 Ti 3 O 7 nanotubes. The as-synthesized C, N-TiO 2 nanorods had a length of 50-200 nm and a diameter of 15-20 nm. From the elemental analysis of the SEM microdomain, it was concluded that the catalyst contained C and N elements. Moreover, the C, N-TiO 2 nanorods exhibited better photocatalytic properties for degrading MO under simulated sunlight irradiation. The 20:1 TiO 2 NRs catalyst presented the highest apparent rate constant, which was almost 8.3 times higher than that of undoped TiO 2 Scheme 1. Schematic diagram of the possible photocatalytic process of C, N-TiO 2 NRs under simulated sunlight.
nanorods. Reactive species (RSs) scavenging experiments revealed that ·OH, h + and ·O 2 −were the major reaction species in the degradation of MO. Hence, our results demonstrated that interlaminar bonding interaction was a promising technology for the facile synthesis of non-metal doped titanium-based nanomaterials.