Two novel hierarchical homogeneous nanoarchitectures of TiO2 nanorods branched and P25-coated TiO2 nanotube arrays and their photocurrent performances

We report here for the first time the synthesis of two novel hierarchical homogeneous nanoarchitectures of TiO2 nanorods branched TiO2 nanotube arrays (BTs) and P25-coated TiO2 nanotube arrays (PCTs) using two-step method including electrochemical anodization and hydrothermal modification process. Then the photocurrent densities versus applied potentials of BTs, PCTs, and pure TiO2 nanotube arrays (TNTAs) were investigated as well. Interestingly, at -0.11 V and under the same illumination condition, the photocurrent densities of BTs and PCTs show more than 1.5 and 1 times higher than that of pure TNTAs, respectively, which can be mainly attributed to significant improvement of the light-absorbing and charge-harvesting efficiency resulting from both larger and rougher surface areas of BTs and PCTs. Furthermore, these dramatic improvements suggest that BTs and PCTs will achieve better photoelectric conversion efficiency and become the promising candidates for applications in DSSCs, sensors, and photocatalysis.

In this communication, we report for the first time the synthesis of two novel hierarchical homogeneous modification nanoarchitectures (i.e., P25-coated TNTAs, PCTs; and TiO 2 nanorods branched TNTAs, BTs) via two-step method of electrochemical anodization and hydrothermal modification approach. The main precursors of modification are the P25 (Degussa, Germany) and titanium(IV) isopropoxide (TTIP of 95%). Erenow, the optimized nanoarchitecture TNTAs (with bigger pore diameter, longer length, and larger space among tubes) have been prepared by electrochemical anodization method. Interestingly, the as-synthesized BTs and PCTs with beautiful morphologies show both larger and rougher surface area, and these properties result in dramatic improvement of light-absorbing and charge-harvesting efficiency, which has been shown through the UV-Vis diffuse reflectance spectroscopic spectra and photoelectrochemical performances in this article.

Experimental section Fabrication of optimum nanoarchitecture TNTAs
In this article, TNTAs were prepared using a typical anodization approach [13]. Briefly, the fabrication process of the optimum nanoarchitecture TNTAs with bigger pore diameter, larger space among tubes and longer length was described as follows, Titanium foil samples, about 200 μm × 2 cm × 3.5 cm (Purity≥99.6%, from ShengXin non-ferrous metal Co., LTD, Baoji, Shanxi, China) were cleaned with soap, acetone, and isopropanol before anodization. A two-electrode configuration was used for anodization, with Ti foil as the anode, and platinum foil as the cathode. A 99.7% pure Ti foil (0.2 mm thickness, 2 × 3 cm 2 ) was immersed in the electrolyte containing 0.35 wt% NH 4 F (85% Lactic Acid) and 10 vol.% DMSO (dimethyl sulphoxide: purity ≥99.0%) at a 45 V constant potential for 9 h. Thus we obtained the amorphous TNTAs, and then the as-prepared TNTAs were annealed at 400°C for 1.5 h for further use.

Synthesis of hierarchical homogeneous nanoarchitecture BTs
The BTs were obtained via a modification process of growing TiO 2 nanorods on the as-prepared TNTAs by conventional hydrothermal growth method. Briefly, the as-prepared TNTAs were immersed in a beaker with growth solution, this solution was consisted of 90 mL of 0.8 M HCl (36-38%) with constant stirring at 25°C for about 15 min. After that, 6 mL of TTIP of 95% as precursor was dropped (0.16 μL/s) in mixture solution, kept stirring for 1 h [7,32,33], and then the beaker was sealed and heated at 95°C for 9 h, with slight stirring maintained for the entire heating process to grow TiO 2 nanorods on the TNTAs. After the reaction, the reactant was cooled freely to room temperature and washed several times with ethanol and distilled water, and the as-prepared BTs were obtained. The BTs were finally achieved through annealing in a muffle furnace at 400°C for 2 h.

Fabrication of hierarchical homogeneous nanoarchitecture PCTs
We fabricated PCTs via a hydrothermal approach of coating P25 on the as-prepared TNTAs. About 0.4 g P25 (Degussa, Germany) was put into a beaker with 300 mL of distilled water, then they were mixed through vigorous magnetic stirring and ultrasonicating alternately at room temperature more than 5 times (about 10 min per time), After that, the mixed solution was kept state static more than 3 h, and then transferred into a Teflon-lined autoclave (80 mL), in which the asprepared TNTAs were suspended. The autoclave was sealed and heated at 80-120°C for 12 h to coat P25 on the TNTAs, and then it was cooled freely to room temperature and washed several times with distilled water, thus the as-prepared PCTs were obtained. Finally, the PCTs were fabricated after the as-prepared PCTs were annealed at 400°C for 2 h.

Characterization
The crystal structures of the as-synthesized samples were firstly determined by using a Bruker D8 advance X-ray diffractometer (XRD, Cu Kα radiation; λ = 1.5418 Å). Then the morphologies were observed by fieldemission scanning electron microscopy (FESEM, JOEL, JSM-6700F), and transmission electron microscopy (TEM and HRTEM, JEM-2010FEF; 200 kV). Photoelectrochemical experiments were carried out using a threeelectrode configuration (CH instruments, CHI 660C) with a Pt wire counter electrode, a reference saturated calomel electrode and a working electrode. The all samples used as working electrodes were illuminated with a 150~350 W adjustable xenon lamp (from Shanghai Lansheng Electronics Co., LTD., Model, XQ350W). The measured light irradiance was approximately 100 mW/cm 2 , and the scan rate was 100 mV/s

Results and discussion
In this study, the two-step method is used to synthesize the BTs and PCTs. The first step is the fabrication of the optimize nanoarchitecture TNTAs [36,37]. From Figure 1, it can be found that the TNTAs show very nice highly ordered, self-organized, and free-standing morphologies, and the optimize geometrical architectures (average external diameter, 350 nm; tube length, 3.5 μm; wall thickness, 10 nm; and space among tubes, 60 nm), and also show at least local single-crystalline status. These characterizations can be observed from the FESEM images of the top view and cross-section of the TNTAs shown in Figure 1a and the TEM, SAED, and HRTEM images in Figure 1b.
The second step is the synthesis of BTs and PCTs using hydrothermal modification method. In brief, they were obtained from growing branched TiO 2 nanorods and coating P25 on the pre-prepared TNTAs via hydrothermal modification process, the images of obtained BTs and PCTs are shown in Figures 2 and 3, respectively. As for the BTs, the mechanism of the formation of TiO 2 crystal nucleus and growth of the anisotropic 1D nanocrystalline TiO 2 nanorods, and their corresponding FESEM images are depicted in Figure 4. From schematic diagram of the morphologies evolution of the BTs and the FESEM images, it is clearly observed that more and more TiO 2 nanocrystal nucleus were firstly formed on the rough surfaces of the TiO 2 tubes with special bamboo structures, many rings and attached particles, these special structures and morphologies are the probable cause of crystal nucleus formed. And then the nucleus gradually grew up and became increasing TiO 2 nanorods along the backbones of the TiO 2 tubes, along with a small quantity of free-grown rods random adhered to the backbones of the tubes. Thus these TiO 2 nanorods made BTs have both larger and rougher surface area [7,34]. Furthermore, the same conclusion can also be confirmed by the top view FESEM images showed in Figure 2a, c, the cross-sectional view in Figure 2b, the TEM image of a individual branched TiO 2 nanotube in Figure 2d. And the insets in Figure 2d are the SAED pattern and the HRTEM images, which show the BTs are evident polycrystalline. Figure 3 is the characterization of another homogeneity nanostructure (the PCTs). Figure 3a is the top view FESEM image of the PCTs. A cross-sectional view in Figure 3b shows that the length of the tubes is the same as that of TNTAs (about 3.5 μm) and the P25 nanoparticles are densely grown on the whole surface (including inside and outside) of the TiO 2 tubes. And the top view of the PCTs with many attached P25 particles is clearly shown by the high-magnification FESEM image in Figure 3c. Meanwhile, Figure 3d shows the PCTs' TEM image, and its inset of the HRTEM image shows the (101) crystal facet and the 0.35 nm interplane distance of a typical anatase TiO 2 while the another inset of the SAED pattern shows that the PCTs are polycrystalline structure [24]. The growth mechanism of the PCTs is mainly dependent on the special structures and morphology of TNTAs, especially its bigger pore diameter, larger space among tubes, and rough surface. Moreover, annealing plays an important role in the process of transforming the P25 on the TiO 2 tube surface from attached state into crystallization state.
Otherwise, the X-ray diffraction (XRD) patterns in Figure 5a, b, c are also employed to characterize the properties of the obtained samples. We can find that the diffraction peaks of the samples (b, c) and the dominant diffraction peaks of the samples (a) match well with the crystal structure of the anatase TiO 2 phase (JCPDS 21-1272) [38] except for one peak of the Ti (101). The main reason can be attributed to thermal treatment temperature of no more than 400°C for 2 h. It is noteworthy that the two peaks [R (110) and R (211)] in Figure 5a just match with the crystal structure of the rutile TiO 2 nanorod (JCPDS no. 21-1276) [7,12], this comes from those rutile TiO2 nanorods grown on the TNTAs.
On the basis of the above observations and structural analyses, we conclude that both of the BTs and PCTs can provide larger and rougher surface areas than the TNTAs compared with the arrays of same geometrical size and quantity [7,34,35]. As a result, this larger and rougher surface areas are favorable to improve lightabsorbing and charge-harvesting efficiency and to absorb more dye for better photoelectric conversion efficiency and better applications such as photocatalysis, sensors, etc. Moreover, it is also found that the growth length and density of the TiO 2 nanorods of the BTs can be readily controlled by adjusting the growth time and the concentration of growth solution, and that the density of the coated P25 particles can also be controlled through changing the coating time and the concentration of coating solution. Figure 6 shows the UV-Vis diffuse reflectance spectra of three samples (TNTAs, PCTs, and BTs) and Ti foil. Comparing to the UV-Vis absorption spectrum of the TNTAs, the absorption edges of the samples (PCTs and BTs) displayed appreciable shifts (BTs is a little bit larger than PCTs) to visible region revealing some decreases in their band gaps. This conclusion is mainly consistent with the above discussions and the previous studies [39][40][41]. Simultaneously, it can also be found  that the absorption intensity of each sample (TNTAs, PCTs, and BTs) is gradually increasing after their absorption peaks. The cause for this effect mainly comes from absorption effect of the annealed (400°C, 2 h) Ti foil substrate to visible (see the inset in Figure 6). Otherwise, the general UV-Vis absorption spectra only reflect the intrinsic optical property for the bulk of a solid. However, the actual absorption spectrum of a photocatalyst is an overlapping result of intrinsic and extrinsic absorption bands [42].
Furthermore, Figure 7 clearly shows the comparison curves of the photocurrent densities versus applied potentials for three different TiO 2 photoanodes (TNTAs, BTs, and PCTs) under Xe lamp irradiation (100 mW/cm 2 ) in 1 M KOH electrolyte [43]. It can be observed that the values of the photocurrent densities of BTs and PCTs are dramatically greater than that of TNTAs. At -0.11 V and under the same illumination conditions, the photocurrent density of BTs shows more than 1.5 times higher than that of TNTAs while PCTs versus TNTAs is more than 1 times higher. These experimental results are well consistent with the effect from above UV-Vis diffuse reflectance spectra. They suggest that the BTs and PCTs used as photoanodes can harvest more solar light and more photogenerated charge than that of the TNTAs with the same geometrical structure. In addition, the photocurrent densities of the BTs and PCTs also show a steeper increase when their applied potentials are over -0.7 V. Thus as for the BTs and PCTs, e --h + pairs induced by photon absorption are split more readily compared with the TNTAs. The conclusion mainly results from the fact that more incident photons are absorbed on the electrode with larger and rougher space area [44].

Conclusion
In summary, we have reported here the fabrication of two novel hierarchical homogeneous nanoarchitectures of BTs and PCTs with larger and rougher surface areas via facile hydrothermal modification process. Based on the investigation of the photocurrent densities versus applied potential, the photocurrent density of BTs, at -0.11 V and under the same illumination conditions, shows more than 1.5 times higher than that of TNTAs while PCTs versus TNTAs is more than 1 times higher. On the basis of the results and discussion, we conclude that the dramatically improved photocurrent densities of the BTs and PCTs used as photoanodes are mainly due to their better incident photons and photogenerated chargeharvesting capability compared to TNTAs resulting from their further enhanced and rough surface areas. As a result, our study will also provide a new approach in conformating hierarchical homogeneity nanostructure materials and presenting two kinds of promising candidates for applications in DSSCs, sensors, and photocatalysis.   JH and FW participated in the investigation of the photocurrent performances. XT and JP participated in the design and idea of the study. All authors read and approved the final manuscript.