Controllable growth of three-dimensional CdS nanoparticles on TiO2 nanotubes to enhance photocatalytic activity

Exploiting photocatalysts with characteristics of low cost, high reactivity and good recyclability is a great significance for environmental remediation and energy conversion. Herein, hollow TiO2 nanotubes were fabricated by a novel and efficient method via electrospinning and an impregnation calcination method. With the hydrothermal method, the CdS nanoparticles were modified on the surface and in walls of the TiO2 nanotubes. By changing the reaction conditions, the morphology of CdS nanoparticles presents a controllable three-dimensional (3D) structure. The morphology of the samples was characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The structure and components of samples were characterized by X-ray diffraction (XRD), energy dispersive X-ray analysis (EDX) and X-ray photoelectron spectroscopy (XPS). The light absorption efficiency was detected using UV-vis diffuse reflectance spectroscopy (DRS) and photoluminescence (PL). The photocatalytic properties were evaluated by degradation of methyl orange (MO) and photocatalytic hydrogen evolution under visible light irradiation. From the results, the TiO2/CdS nanotubes exhibit better photocatalytic activity than the pure TiO2. The synthetic mechanism of TiO2/CdS heterostructures and a possible photocatalytic mechanism based on the experimental results were proposed.


Introduction
With progress in the eld of nanotechnology, it has become possible to prepare semiconducting materials of nanocrystal structures with special shapes and dimensions, such as nano-bers, nanorods, nanoowers, nanospheres and nanotubes, [1][2][3][4][5] which have garnered extensive attention on account of their unique electronic, optical, and photoactive properties. [6][7][8][9] Meanwhile, they are widely utilized in the elds of gas sensors, optical devices, visible light detectors, and eld emission emitters, especially for photocatalysis. [10][11][12][13][14][15] Among copious inorganic semiconductor nanostructures, titanium oxide (TiO 2 ) has been extensively studied to date, with extensive research on its high surface-to-volume ratio, low cost, environmental friendliness and unique physical and chemical properties. [16][17][18][19] Recently, the application of TiO 2 as photocatalysts for degradation of toxic organic pollutants have attracted wide public attention, because the TiO 2 nanostructure and its composites can signicantly improve the photocatalytic performance. [20][21][22] The properties of nanomaterials mainly rely on their nanoscale and morphology, for nanoscale semiconductor to perform well as catalysts, they at least satisfy three characteristics: (1) high specic surface area, (2) have good stability and (3) the size and distribution should be uniform. [23][24][25] Therefore, TiO 2 nanotubes shows intrinsic advantages compared to traditional TiO 2 in terms of size and specic surface area, 26 design and fabricate TiO 2 based nanotubes for heterogeneous photocatalysis have become a hot research topic. Thus far, many synthesis methods such as electrodeposition method, sol-gel method, templateassisted and hydrothermal treatment, have been developed for fabricating TiO 2 nanotube. 18,19,27,28 However, these traditional method oen accompanied problems such as cumbersome operation, template hard removal, high production cost and poor structural stability, etc. 29 Meanwhile, the high recombination rate of photogenerated (PG) electron/hole pairs is another limits factor of photocatalysis. 30,31 Although reduce the size of nanoscale TiO 2 open a convenient way for improving photocatalytic efficiency, but it cannot tackle the fundamental limits issue. 32,33 By contrast, making TiO 2 doped with other elements, modied with noble metal and combined with other semiconductors to form heterostructures, which provide a effective means to inhibit the recombination of PG electron/hole pairs. [34][35][36][37][38] It is necessary to design new-type composite photocatalysts not only improve the separation of PG electron/hole pairs, but also extend the optical absorption of TiO 2 , thus to obtain a more efficient utilization of solar energy. 39,40 Cadmium sulde (CdS) is an excellent n-type semiconductor with an ideal band-gap energy of 2.4 eV, which could absorb the visible light radiation up to 520 nm, 41,42 has been widely used as a visible light response sensitizer. 34,43 By comparing the energy levels of TiO 2 with CdS, which could nd their analogical band structures are quite suitable to construct heterostructures and bring an effective separation of PG electron/hole pairs. 44,45 Therefore, that is expected to combination the TiO 2 with CdS form a heterostructure structure for overcoming the demerits that exist in TiO 2 and CdS simultaneously. 46 Herein, we described the fabrication of CdS decorated TiO 2 nanotubes via electrospinning combined with controlled calcination and hydrothermal process. By changing the synthesis condition, the density and size of CdS on the surface (inner wall) of the TiO 2 nanotubes can be controlled. The morphology and dimension of the CdS/TiO 2 nanotubes were observed by the SEM and TEM. The crystal phases, optical absorption properties, elemental and chemical state were explored by XRD, DRS and XPS. Due to the formation of CdS/TiO 2 heterojunction structure, the hybrid nanoparticles could reduce the recombination of PG electron-hole pairs, and the hollow nanotubes structure provides a larger specic surface area and more active sites, that the as-prepared CdS/TiO 2 nanotubes signicantly improved photocatalytic activity. Moreover, that may provide a new method for preparation of semiconductor heterostructure nanotubes.

Methods
The precursor solution was prepared by adding 1.4 g PS to 5.0 mL DMF with magnetic stirring at 50 C. Aer PS dissolution in the DMF completely, the prepared solution was loaded into a syringe, which was connected to a 15 kV high voltage. The ow rate of the solution was controlled at 0.5 mL h À1 , and the distance from needle to the rotate acceptor was 20 cm. The whole process was kept with humidity below 20%. Then, the PS membrane were soaked in the BOT/ethanol (volume ratio 1 : 10) precursor solution for 5 min. Then, the composite TBOT/PS precursor bers were natural air dried. Finally, the precursor bers were calcined to 550 C with a rate of 2 C min À1 and keep 550 C for 2 hours, to burn the PS templates completely. Aer annealing, the hollow TiO 2 nanotubes were obtained.
Hydrothermal method was used for CdS nanoparticles modied. The as prepared TiO 2 nanotubes were transferred into a Teon-lined autoclave, which containing 30 mL solution of Cd(NO 3 ) 2 and thiourea, the concentration are equal and both volumes are 15 mL. Then, the Teon-lined autoclave was heated to 200 C and keep 8 hours. Finally, the CdS/TiO 2 nanotubes were collected and washed in DI water. For compare the effect of reactant concentration on morphology, change the concentration of Cd(NO 3 ) 2 solution to 5, 10, 20, 30 mmol L À1 and the corresponding CdS/TiO 2 nanotubes was named as TC1, TC2, TC3 and TC4, respectively ( Fig. 1).

Measurement of photocatalytic activity
Methyl orange (MO) was used as the target substrates for evaluating the photocatalytic activity. The photocatalytic reaction was carried out in a reactor with a circulating water system to maintain a constant temperature. The xenon lamp (Philips, 100 W) equipped with a 420 nm cut-off lter instead and the irradiation time were ranged from 5 to 200 min. Under the same conditions, all samples were placed in the quartz tube, which away from the irradiation source 5 cm. The degradation ratio was 40 mL MO solution (10 mg L À1 ) with 40 mg photocatalyst. Then, the suspension was stirred in the dark for 1 hour to establish adsorption and desorption equilibrium. During the photoreaction process, approximately 3 mL of the mixture was sampled in a constant time, and use centrifugation to separate the nanoparticles. The, using the spectrophotometer to determination the solution concentration. H 2 evolution reactions were proceed in a closed gas-circulation system. 50 mg samples were dispersed in 200 mL aqueous solution of Na 2 S (1.0 M) and Na 2 SO 3 (1.0 M). Aer removing the air of reaction system, the solution were irradiated by a 100 W xenon lamp equipped with a 420 nm cut-off lter, and the amount of H 2 were measured by gas chromatography (GC-9700, N 2 carrier).

Morphology of electrospun nanotubes TiO 2 /CdS
The morphology of the nanotubes were characterized by SEM. Fig. 3A show the SEM image of TiO 2 , which are composed by regular and uniform nanotubes with diameter of approximately 700 nm and the uniformity was above 80%, demonstrating the TBOT/ethanol solution permeate into the surface of the PS template bers successful. From the illustration of Fig. 3A, we can observe the high magnication SEM of a TiO 2 nanotube that present the wall thickness of the nanotube about 100 nm and the inwall of the nanotubes are rough. Fig. 3B clearly present the SEM image of core-shell structure TiO 2 nanotubes, which was obtained by replacing PS/DMF solution with PAN/ DMF solution (PAN: 13 wt%) and other operations were the same. The core might be carbon bers which produced by the PAN incomplete calcination. The SEM of four kinds of CdS/TiO 2 hybrid showed in the Fig. 3C-F with low and high magnications. Compared to the pure TiO 2 nanotubes, the CdS/TiO 2 nanotubes could observed amount of nanoparticles deposited on the surface and inwall, and the density of nanoparticles increase with the concentration of Cd 2+ ions. These results show that the morphology of CdS modied on the nanotubes was related to the solution concentration. Notably, further increase of Cd 2+ , the CdS nanoparticles grown together to form a pinecone shaped 3D structure as shown in Fig. 3F. The morphology and microstructure of as-prepared photocatalysts also investigated by TEM. The low magnication TEM image (Fig. 4B) further veries that CdS nanoparticles were well covered the surface and inwall of TiO 2 nanotubes, compared to the TEM of TiO 2 nanotubes (Fig. 4A). These nanoparticles with sizes of 5-50 nm are distributed randomly on the TiO 2 . The high resolution TEM image (Fig. 4D) clearly reveals an intimate interface between the TiO 2 and CdS in the composite to form heterojunctions, which is favorable for the electron transfer in photocatalytic activity. Also demonstrates the CdS/TiO 2 photocatalyst obtained by hydrothermal method form a heterostructure rather than a mixture of TiO 2 and CdS. Meanwhile, the Fig. 2 The XRD patterns of the as-prepared photocatalysts.  high resolution TEM are carried out to survey the lattice fringes of CdS/TiO 2 nanotubes, the lattice spacing are 0.35 nm and 0.33 nm, which are coincident with the (101) spacing of anatase TiO 2 and (111) of CdS phase. The legible fringe also suggests the high degree of crystallinity corresponds with the result of XRD results. The Fig. 4E illustrates the selected area electron diffraction (SAED) pattern of TiO 2 /CdS, which presentation a bright diffraction rings assigned to high crystallization and polycrystalline structure. The Fig. 4F was SAED pattern of pure CdS nanoparticles on the surface of TC3 means the obtained CdS particles were pure single crystals.

XPS analysis
XPS analyze could reveal the information about the surface composition and chemical state. As shown in Fig. 5A, CdS/TiO 2 hybrid (TC2) is mainly composed of ve elements, which is Ti, O, Cd, S and C, The peak of C 1s can be ascribed to the hydrocarbon from the instrument itself. The high resolution spectra of Ti 2p were exhibited in the Fig. 5B. The Ti 2p spectra shows the peaks at 458.3 eV and 464.1 eV for Ti 2p 3/2 and Ti 2p 1/2 of TiO 2 , respectively. The distance between the two peaks is 5.8 eV, which is identical to that of the neat TiO 2 indicating that Ti ion is in the form of Ti 4+ , high resolution spectrum of Cd 3d (Fig. 4C) displayed characteristic peaks of 404.5 eV and 411.4 eV, which are corresponding to the Cd 3d 5/2 and Cd 3d 3/2 states of Cd 2+ in CdS, respectively. Consequently, XPS analysis further determines the formation of TiO 2 and CdS in the sample, which is agreement with XRD and TEM result.

UV-vis diffuse reection spectra and PL spectra
UV-vis DRS are tested to determine the optical properties of the photocatalysts. From the results of DRS (Fig. 6A) the reection spectra of TiO 2 and CdS/TiO 2 heterostructure in the region spanning from 230 to 800 nm were showed. Compared to the TiO 2 nanotubes, the TC2 had a lower reectivity in the range of 240-380 nm, especially in the ultraviolet light range. Aer the modied of CdS, all CdS/TiO 2 nanotubes present a remarkable light adsorption into the visible light region, which may due to the excellent visible-light photoresponse of CdS nanoparticles. Notably, the adsorption edges of CdS/TiO 2 nanotubes gradually broaden with the mount of modied CdS nanoparticles increasing, indicated the high responsiveness of the photocatalyst to visible light. Fig. 6B is the forbidden bandwidth obtained by the catalyst according to the Kubelka-Munk formula. Among them, the forbidden band widths of TiO 2 and four CdS/TiO 2 heterostructures (TC1, TC2, TC3 and TC4) are 3.17 eV, 2.17 eV, 2.08 eV, 1.92 eV and 2.01 eV. It can be seen from the comparison with the TiO 2 tubes that the forbidden bandwidth of the CdS/TiO 2 heterostructure is signicantly reduced. Among them, TC3 has the lowest forbidden bandwidth. We can speculate that this sample has the best catalytic performance, and the catalytic experiments proved that the prediction was correct. (The detailed content of the edge position is Fig. S1. †)

Photoluminescence analysis
PL spectroscopy is correlation with the separation and recombination of PG electrons/holes. As shown in Fig. 7, the PL spectra of the TiO 2 nanotubes showed a strong emission peak at 390 nm (excitation wavelength: 370 nm). However, the PL peak intensity of CdS/TiO 2 nanotubes was signicantly decreased  with modication amount of CdS nanoparticles increased. The key factor in the formation of this phenomenon is the defects in the crystal structure of TiO 2 , which act as traps for capturing the photo-excited electrons and thus, inhibit the recombination of the electrons and holes pairs.

Photocatalytic activity and recycled photocatalytic degradation performance
To identify the photocatalytic of the as-prepared samples, MO solution degradation were tested under UV and visible irradiation. The photocatalytic effect was evaluated by C 0 /C t (C 0 nd C t were the initial concentration and the concentration of MO at time t, respectively). From the Fig. 8A, shows the concentration change of MO in the process of photodegradation, aer irradiated by UV light for 30 min, the photocatalytic degradation efficiencies of the TC2 nanotubes are 95.3% slightly better than the TC3 (83.2%), which is much higher than the TC4 (72.5%) and pure TiO 2 nanotubes (37.8%). The photodegradation of MO without catalysis almost negligible. However, as shown in Fig. 8B, under visible light irradiation (l > 420 nm) for 200 minutes the TC3 shows the excellent visible light catalysis (95.2%) similar to the TC4 (94.7%), which CdS load looks like pine-cone shaped and much higher than TC2 (82.6%), TC1 (61.2%), and pure TiO 2 nanotubes (0.09%), which was consistent with the results of UV-vis DRS. In addition, we also investigated the photocatalytic stability of TiO 2 /CdS heterostructure (TC3) under visible light irradiation. The results indicated that the hybrids exhibited good stability aer being reused for ve times (Fig. 8D).
Meanwhile, the photocatalytic activities of the nanotubes were evaluated by H 2 production experiment under the visible light irradiation in aqueous solution (containing 1.0 M Na 2 SO 3 and 1.0 M Na 2 S as sacricial reagents). Fig. 8C shows the H 2 evolution rate by the TiO 2 and TiO 2 /CdS nanotubes, and there is no appreciable H 2 was observed when TiO 2 nanotubes was tested. In contrast, the TC1 exhibited a slightly enhanced H 2 evolution rate of 22.35 mmol g À1 h À1 . With the increase of CdS, the H 2 evolution rate are increased. The H 2 evolution rate of TC2 photocatalyst is 64.26 mmol g À1 h À1 and TC3 (118.73 mmol g À1 h À1 ) reaches a maximum H 2 evolution rate, demonstrates that is effective to employ CdS as cocatalyst for improving efficiency of TiO 2 photocatalytic activity. Further increases the amount of modication of CdS, the H 2 generation rate (TC4 112.86 mol g À1 h À1 ) showed a slight decrease. This might be due to the less exposed active sites of TiO 2 /CdS resulting from the excess CdS shielding.

Mechanism for the enhancement of photocatalytic activity
The principle of photocatalytic decomposition over CdS/TiO 2 heterostructure is delineated in Fig. 9. Under the visible-light irradiation, the charge carriers were rstly generated from CdS owing to its narrow energy band gap. The photoexcited electrons on the valence band (VB) of CdS can move to their own conduction band (CB). However, the high recombination of electrons (e À ) and holes (h + ) of CdS limits its photocatalytic activity. When the TiO 2 and CdS form a heterostructure, the PG e À on CdS nanoparticles can easily transfer to the CB of TiO 2 through the interface as a result of the lower conduction edge Fig. 7 The photoluminescence spectra of the as-prepared photocatalysts.  potential of TiO 2 than that of CdS. Meanwhile, the h + remain on the VB of CdS. Therefore, the TiO 2 acts as a temporary e À trap to decrease the recombination of PG e À /h + . Consequently, the PG h + on the VB of CdS can react with the surface adsorbed H 2 O to generate $OH. The e À on the CB of TiO 2 could reduce the adsorbed O 2 to form $O 2 À . Organic dye (MO) are eventually oxidized by these highly active species to CO 2 and H 2 O products. The major reaction steps are revealed as follows: TiO 2 /CdS + hv / TiO 2 /CdS (e À + h + ) TiO 2 /CdS (e À + h + ) / TiO 2 (e À ) + CdS (h + )

Conclusions
In summary, hollow TiO 2 nanotubes were obtained using electrospinning combined with an impregnation calcination method. CdS nanoparticles were modied on the surface and inner wall of TiO 2 nanotubes via hydrothermal process. By changing the reaction conditions, the size and density of CdS on the surface and inner wall of the TiO 2 nanotubes can be controlled. Owing to the formation of heterojunction structure, the hybrid nanoparticles could enhanced separation of e À /h + pairs and reduce the recombination of charge carriers, and the hollow nanotubes structure provides a larger specic surface area and more active sites, that the CdS/TiO 2 nanotubes showed remarkably photocatalytic activity than pure TiO 2 . Moreover, that may provide a new method for preparation of heterostructure nanotubes.

Conflicts of interest
No conicts of interest in this article.