Low-temperature synthesis of high-ordered anatase TiO2 nanotube array films coated with exposed {001} nanofacets

High-ordered anatase TiO2 nanotube array films coated with exposed high-reactive {001} nanofacets were fabricated by a modified hydrothermal method using amorphous anodic TiO2 nanotube arrays (ATONAs) as starting materials. It was found that the reaction between gas phase HF and solid ATONAs played a key role in the transformation process from amorphous to anatase TiO2, and the TiO2 tubular structure kept unchanged during the surface modification with an exposed {001} facets up to 76.5%, which could be attributed to the low reaction temperature of 130 °C. Our study provided a novel route for the facile preparation of {001} facets exposed anatase TiO2.

TiO 2 nanotube array films with exposed {001} facets up to 76.5% was successfully prepared at reaction temperature as low as 130 °C. After the surface F − ions were removed by 2 h of 600 °C annealing 8,13,16 , the obtained F − free TiO 2 films with exposed {001} nanofacets showed much better photocatalytic activities than the original ATONAs for methyl orange (MO) degradation. Figure 1a,b shows the XRD patterns of the as-synthesized samples with different preparation conditions. For ATONAs, only diffraction peaks of titanium were detected, indicating that the as grown ATONAs have amorphous structure 17 . After hydrothermal treatment, broad diffraction peaks at 25.3°, 37.8°, 48.2°, 53.9° and 55.2° were observed, which can be indexed to the (101), (004), (200), (105) and (211) reflections of anatase TiO 2 , and the diffraction peak intensities at 25.3° and 37.8° increased with preparation time, indicating the nucleation and growth of anatase TiO 2 during 130 °C hydrothermal treatment (the XRD patterns of samples for 1.5 h and 2 h were shown in Fig. S1 in Supporting Information). Note that, the appearance of strong peak at 37.8°, which was attributed to (004) planes of anatase TiO 2 , suggested the presence of exposed {001} nanofacets 18 . Especially for H-16 and H-20 (Fig. 1b), the exposed {001} planes was estimated to be as high as 74.5% and 76.5%, respectively 19 . The EDS spectra in Fig. 1c indicated that the surface F − ions were completely removed by 600 °C annealing 8,13,16 . XPS measurement also confirmed the successful removal of the surface F − species, as shown in Fig. 1d. At the same time, diffraction peaks attributed to {101} planes sharpened, while the (004) peak at 37.8° weakened slightly, indicating that the loss of terminated F − preferred the exposal of {101} planes. The average grain size estimated from FWHM of diffraction peaks at 25.3° and 37.8° was 12.2 nm and 34.1 nm for H-16, and 30.6 nm and 29.3 nm for HT-16, respectively as shown in Fig. 1e. By the way, the post annealing also induced the formation of rutile nanoparticles, as a result, diffraction peaks of 27.5°, 36.1°, 39.3° and 54.3° appeared 20,21 . Figure 2 shows SEM images of samples prepared under different conditions. The inset of Fig. 2a is SEM image showing the morphology of the as-prepared ATONAs, which were highly ordered and were grown vertically to the titanium substrate. Figure 2a is the ATONAs after 0.5 h hydrothermal treatment, the tubular structure was maintained with small particles coated on the surface (~8 nm estimated from the FWHM of peak at 25.3°). Even the hydrothermal treatment time increased to 20 h, the tubular structure (d) XPS spectra of two samples before and after post annealing, which clearly showed that the surface Fions were removed by annealing. (e) Structural information of the synthesized samples. SH-101 and SH-004: the average particle sizes calculated from the FWHM of (101) and (004) peak for samples before post annealing; SHT-101 and SHT-004: the average particle sizes calculated from the FWHM of (101) and (004) peak for samples after annealing; PH-004 and PHT-004: the percentage of exposed {001} nanofacets before and after annealing. still kept unchanged as the cross section image shown in Fig. 2e. Amazingly, after 600 °C annealing, the surface coated nanoparticles grew up to be nanofacet like grains (insets of Fig. 2d,f ), and the tubular structure was still maintained as shown in Fig. 2b,d,f, which provided an extremely large surface area compared to the normal TiO 2 films (low-magnification SEM images were shown in Figs. S2 to S6). Figure 3a shows SEM image of nanoparticles with truncated bipyramidal or trapezoidal shapes and even some with cubic morphology of anatase TiO 2 coated on the surface of HT-2, which could be attributed to different degree of truncated octahedral. Figure 3b shows typical TEM image of some nanoparticles, which have regular facet-like geometry with exposed flat square 22 . All SEM and TEM images indicated the successful exposal of F − free (004) planes after annealing 8 . Figure 1e summarized the grain size and the evaluation of percentage of exposed {001} planes, according to early estimation method 19,23,24 , TiO 2 with exposed {001} facets up to 76.5% was successfully prepared. Figure 4a shows the HRTEM image of H-2 sample, nanoparticles with grain size in the orders of several nanometers could be observed, the crystal lattice with a spacing of 0.235 nm (004), and 0.35 nm (101) confirmed the anatase TiO 2 nanoparticles 24 . Moreover, the observed (004) planes are parallel to the particle surface, which presented direct evidence for the existence of {001} exposed nanofacets. Evidently, the amorphous ATONAs partially transformed into anatase TiO 2 with exposed {001} nanofacets by hydrothermal treatment at 130 °C. Figure 4b is a typical HRTEM image of HT-2, in this selected observation window, almost all grains had spacing of 0.235 nm (parallel to the grain surface); indicated the high percentage of exposed {001} planes. By the way, a rutile TiO 2 grain with a spacing of 0.229 nm

Discussion
Normally, anatase TiO 2 grains are dominated by {101} facets rather than {001} facets, the latter is more active but with high surface energy [25][26][27][28][29] . When the surface was terminated with F − , the expose of {001} is energetically preferable to {101}, and this was proven theoretically and experimentally by using hydrothermal method 8-10 , by which the Ti precursors were directly immersed into F − /HF containing solutions 30 . A dissolution-precipitation or dissolution-recrystallization process was proposed to describe the transformation from {101} to {001} facets [31][32][33] . Here we modified the hydrothermal method by separating the ATONAs from the acid HF solution 30 , accordingly, the reaction between gas phase HF and solid ATONAs becomes the only route for the transformation from amorphous to anatase TiO 2 . From the XRD spectra shown in Fig. 1, a low temperature of 130 °C was enough to induce this transformation. This could be attributed to the use of amorphous TiO 2 as starting materials, which reduced the activation energy for atom rearrangement. The tubular structure of ATONAs was kept unchanged after hydrothermal treatment, which provided additional evidence that this transformation could not be explained through dissolution-precipitation or dissolution-recrystallization process, which destroyed the initial tubular structure [31][32][33] . The driving force for this transformation could be the surface energy change caused by F − termination, and {001} nanofacets were preferred finally [8][9][10][11][12][13][14] . From literature, the solid state transformation via atomic rearrangement in amorphous matrix is frequently proposed 34,35 . And the synthesis of anatase TiO 2 with exposed {001} facets by heat treatment (450 °C) using amorphous ATONAs as starting materials was achieved 13,16 . Here we demonstrated that a solid-state phase transformation from TiO 2 {101} to {001} facets could be achieved at a low temperature of 130 °C.

Conclusions
In summary, anatase TiO 2 nanotube array films with exposed {001} nanofacets were successfully prepared by a modified low-temperature hydrothermal method at 130 °C. The novel reaction route between gas phase HF and solid ATONAs was demonstrated which played a key role for the solid state transformation process from amorphous to anatase TiO 2 , after the surface dangling bonds were terminated with F − ions, {001} facets became energetically favored. F − free {001} facets with high reactivity could be achieved via 600 °C post annealing. The prepared anatase TiO 2 nanotube array films with exposed {001} nanofacets exhibited enhanced photocatalytic activity for methyl orange (MO) degradation under ultraviolet light (UV), which could be attributed to the improvement of charge separation derived from the synergy effect between {001} and {101} facets.

Methods
Preparation of well-aligned anodic TiO 2 nanotubes (ATONAs). All reagents (Sinopharm, analytical grade, Shanghai, China) were used without further purification. Well-aligned ATONAs on a Ti substrate were fabricated via traditional electrochemical anodization 37 . Briefly, the anodic growth was conducted using a home-made two electrode electrochemical cell under a constant voltage of 40 V at room temperature, titanium plates (Ti, 90 mm × 40 mm × 0.2 mm, 99.9% purity, Baoji, Shanxi, China) were used as working electrode and a graphite plate as the counter electrode, the electrolyte was composed of 3 wt% NH 4 F and 0.5 v% H 2 O dissolved in the ethylene glycol. The obtained well-aligned ATONAs were rinsed with ethylene glycol and ethanol to remove the residual electrolyte solution followed by drying at 80 °C for later use (inset of Fig. 2a).
Exposing of {001} nanofacets. The well-aligned ATONAs were treated at 130 °C in an autoclave.
In a typical experiment, 6 ml of hydrofluoric acid solution with pH = 3 was transferred to a Teflon-lined autoclave (capacity: 25 mL), then the as-synthesized ATONAs , which were cut into 2 cm × 2 cm dices, were fixed by Teflon holders 1 cm above the solution to avoid dipping the ATONAs into the solution. Then, the autoclave was sealed and heated at 130 °C for 0.5 h to 20 h. After the autoclave was cooled down to room temperature, the products were washed with ethylene glycol, ethanol and deionized water, and then dried in an oven at 80 °C for 0.5 h, the samples were labeled as H-t (where t stood for the hydrothermal time). Then, samples were heated at 600 °C for 2 h in air with a heating rate of 2°/min to obtain {001} nanofacets exposed films, and the final samples were labeled as HT-t.
Characterization. X-ray diffraction (XRD) patterns were recorded on a Philips XD-98 X-ray diffractometer with Cu Kα radiation (λ = 0.15406 nm). The morphology of the samples was characterized by A ZEISS ULTRA 55 scanning electron microscopy (SEM) equipped with energy dispersive spectrometer (EDS) and transmission microscopy (TEM, FEI Tecnai G 2 F20 S-TWIN, FEI Inc., America). X-ray photoelectron spectroscopy (XPS) data were obtained using a Thermo ESCALAB 250Xi (Thermo Fisher Scientific), the X-ray source was an Al Kα radiation, and all binding energies were referenced to the 284.8 eV C1s. Photocatalytic activity. The photocatalytic activities were evaluated by degrading the methyl orange (MO, 0.15 mg/L) as the model organic pollutants in aqueous solutions, three plates of {001} exposed TiO 2 films (size: 2 cm × 2 cm, TiO 2 weight: ca. 1.8 mg) were dipped in 30 mL of MO solution and then were irradiated by a 125 W mercury lamp which irradiated the light of 365 nm. Before irradiation, the system was put in a darkroom for 0.5 h with magnetic stirring to ensure adsorption and desorption equilibrium between the samples and organic molecules. The photodegradation experiments were performed in an open quartz vessel under continuously stirring. During degradation, 2 mL of MO were taken every 30 min and the concentration of MO was measured by UV-3600 spectrophotometer (Shimadzu, Japan) at 463 nm 38 . The influence of sampling was compensated in degradation efficiency calculation.