Preparation of Improved p-n Junction NiO / TiO 2 Nanotubes for Solar-Energy-Driven Light Photocatalysis

Self-organized TiO 2 nanotubes (TNTs) with average inner diameter of 109 nm, wall thickness of 15 nm, and tube length of 7– 10 μm were loaded with nickel oxide (NiO) nanoparticles via incipient wet impregnation method. The molar concentration of Ni(NO 3 ) 2 ⋅6H 2 O aqueous solution varied in a range of 0.5M–2.5M. The samples were characterized for crystalline phase, morphology, topography, chemical composition, Raman shift, and UV-Vis diffusion reflection properties. The finding shows that the loading of NiO did not influence the morphology, structure, and crystalline phase of TNTs but it exhibited significant effect on crystallite size and optical absorption properties. Further, the solar-energy-driven the photocatalytic activity of NiO/TNTs and pure TNTs was evaluated by degrading methylene blue (MB). The results confirm that photocatalytic efficiency of NiO/TNTs is higher than that of TNTs.


Introduction
Self-organized and vertically oriented TiO 2 nanotubes (TNTs) are of great interest in photocatalytic applications due to their high surface-to-volume ratios, high surface area, good charge transport properties, and chemical stability [1].Few studies have demonstrated TNTs with improved properties compared to other forms of TiO 2 for various applications including photocatalytic degradation of dyes and organic compounds, treatment of gaseous pollutants, water splitting, carbon dioxide (CO 2 ) conversion to methane, and dye-sensitized solar cells [2][3][4][5][6][7].However, the photocatalytic efficiency of TNTs is limited by fast recombination of photogenerated electron and hole pairs and its low visible light utilization.The wide bandgap energy of TiO 2 (3.2 eV for anatase and 3.0 eV for rutile) limits its excitation within the UV range of spectrum.The availability of such spectrum is marginal, that is, ≤5% of the whole solar spectrum in comparison to the visible light spectrum [8].
Many studies have been devoted to the improvement of TiO 2 photocatalytic activity towards solar light absorption.
Quite a few research groups have investigated the utilization of NiO-doped TiO 2 as hydrogen evolution sites from photocatalytic water splitting [23,34].However, the studies on the solar-energy-driven photocatalytic activity

Preparation of TNTs.
All the chemical reagents were of analytical purity and were purchased from Sigma-Aldrich Chemical Co.Self-organized TNTs layers were fabricated on a Ti substrate (99.7%,Sigma-Aldrich) by electrochemical anodization in ethylene glycol (anhydrous, 99.8%) electrolyte containing 0.3 M ammonium fluoride (NH 4 F, 98%) and 2 vol % water (H 2 O) using graphite rod as the counter electrode with a potential of 50 V for 3 h.Ti substrates (20 mm × 30 mm × 0.25 mm) were ultrasonically cleaned with acetone and ethanol prior to anodization.The anodized samples were rinsed thoroughly with deionized water and then annealed at 450 ∘ C for 1 h.The annealed samples were sonicated in ethanol for 30 min to remove surface debris.

Preparation of NiO/
TNTs.An incipient wet impregnation method was adopted for the preparation of NiO/TNTs.The procedure is as follows.The drawn samples were dispensed back into the beaker after the measurement to allow further degradation.The control experiment was performed without a photocatalyst under identical conditions.All experiments were conducted under clear sky conditions at University of Malaya, Kuala Lumpur (latitude 101 ∘ 39  E and longitude 3 ∘ 7  N), between 11.00 AM and 6.30 PMA in April (2012).Solar light intensity was measured using LT Lutron LX-101 Lux meter of 1000 × 100 lx, and the average light intensity over the duration of clear sky weather condition was found to be 87940 lux (for sunlight AM 1.5, 100 mW/cm 2 corresponds to 120 000 lux) [35].the uniformity of NiO dispersion.Thus, the EDS analysis confirms the presence of Ni in TNTs.The detailed elemental composition is tabulated in Table 1.STEM measurements are carried out for distribution analysis of NiO nanoparticles throughout the TNTs surface.Figure 2(a) shows the STEM image of NiO nanoparticles along with size distribution (∼29.8-40.6 nm) inside the walls of TNTs.Bright-field and dark-field images (Figure S2) indicate the presence of nanosized NiO particles inside the hollow structure of nanotubes.No nanoparticles were detected on the top openings of nanotubes (Figure S3), instead they were located along the walls of the TNTs.HRTEM pattern of NiO/TNTs is obtained to verify the crystallization of nanotubes.Direct evidence of crystalline nature of NiO and TiO 2 is observed in Figure 2(b).The lattice spacing of two lattice planes with spacing of 0.35 nm and 0.48 nm, corresponding to the (1 0 1) plane of anatase TiO 2 (JCPDS no.21-1272) and (1 1 1) plane of NiO (JCPDS no.89-5881) is observed clearly.

Structural and Morphological Characterization
Figure 3(a) shows the XRD patterns of NiO/TNTs compared to that of TNTs.All diffraction peaks can be assigned as 100% anatase TiO 2 and Ti substrate, which reveals that there has been virtually no phase change in TiO 2 after the loading of NiO, irrespective of the concentration of NiO.The peaks of tetragonal TiO 2 anatase phase (JCPDS no.21-1272) appeared at 25.3 ∘ , 36.9 ∘ , 37.8 ∘ , 48.0 ∘ , 53.9 ∘ , 55.1 ∘ , 62.7 ∘ , 68.8 ∘ , and 75.0 ∘ , corresponding to (1 0 1), (1 0 3), (0 0 4), (2 0 0), (1 0 5), (2 1 1), (2 0 4), (1 1 6), and (2 1 5) crystal planes, respectively.The Ti peaks are also observed which belong to Ti substrate underneath the oxide nanostructure layer.No new peaks associated with crystalline phases of NiO were detected.This could be explained by the following reasons: (i) Ni cations (0.72 Å) are well substituted into the Ti cations (0.68 Å) in anatase TiO 2 lattice due to their similar ionic radii; (ii) the formation of NiO is below the XRD detection limit due to the presence of NiO in low concentration; (iii) well and uniform dispersion of NiO particles on TiO 2 surface [36,37].The lattice parameters and crystallite size of all samples are calculated using the formula stated below and are summarized in Table 2.The average crystallite sizes of TiO 2 anatase were calculated using Scherrer's equation: where  is the full width half maximum (FWHM) for the 2 peak,  is the shape factor taken as 0.89 for calculations,  is the wavelength of X-ray (0.154 nm), and  is the diffraction angle.The lattice parameters were measured using (1 0 1) and (2 0 0) in anatase crystal planes by using Bragg's equations: Formula for tetragonal system is as follows: where  (ℎ  ) is the distance between crystal planes of (ℎ  ),  is the X-ray wavelength,  is the diffraction angle of crystal plane (ℎ  ), (ℎ  ) is the crystal index, and , , and  are lattice parameters (in anatase form,  =  ̸ = ).The lattice parameters of all samples remain almost unchanged along and -axis, whereas the -axis parameter undergoes a minor increase with increasing NiO concentrations.The XRD results are correlated to XPS analysis in Figure 6(b).It reveals that a large fraction of Ni 2+ ions segregates as a separate NiO and as a major phase, while the remainder fraction is incorporated substitutionally in TiO 2 lattice.As evidence, the Ti 2p XPS spectra in Figure 6(b) could only detect the presence of Ti 4+ in TiO 2 , and there is no apparent peak broadening in the XRD patterns of NiO/TNTs (Figure 3(b)).The slight increase of -axis parameter with increasing NiO concentrations (Table 2) suggested that minor fraction of Ni 2+ replaces Ti 4+ .This was achieved due to similar ionic radius of both Ni 2+ (0.72 Å) and Ti 4+ (0.68 Å) [38].It is also speculated that Ni 2+ was not incorporated interstitially because the lattice constants did not remain constant for higher NiO concentrations.Upon the NiO loading to 2.5 M (Table 2), the crystallite size of NiO/TNTs (34.54 nm) is larger than that of TNTs (33.81 nm).The increase in crystallite size suggests that NiO loading does not lead to the suppression of TiO 2 crystal growth.

Raman Spectra.
Raman spectra of TNTs and NiO/TNTs are depicted in Figure 4(a).2.5 M NiO/TNTs sample yields four distinct Raman peaks at 145 (  ), 399 ( 1 ), 519 ( 1 +  1 ), and 639 cm −1 (  ) with slight broadening, as compared with the other two samples.These are directly attributable to the anatase phase.The broadening of Raman spectra at 145 cm −1 (Figure 4(b)) implies the breakdown of long-range translational crystal symmetry, owing to the substitution of Ni ions into the TiO 2 lattice [39].The absence of NiO or other oxidation state of Ni related modes in the Raman spectra is in good agreement with XRD results.

UV-Vis Diffuse
Reflectance Spectra.The prior challenge in improving the properties of titania is to shift the absorption spectrum of TiO 2 towards the visible region for efficient solar light photons harvesting.The diffuse reflectance spectra (DRS) of different concentrations of NiO/TNTs and TNTs samples are shown in Figure 5.All samples exhibit an absorption band lower than 380 nm (UV region) due to the charge transfer from O 2p valence band to Ti 3d conduction band [40].
It can be seen that NiO/TNTs (0.5 M) did not show a significant red-shifted absorption edge towards visible region due to its marginal increase of NiO concentrations compared to that of TNTs.Meanwhile, the further increasing of NiO concentrations up to 1.5 M and 2.5 M results in a drastic shift towards the visible region.Sample NiO/TNTs (2.5 M) shows a distinct hump between 450 and 515 nm which is due to the crystal field splitting of 3d 8 orbital and charge transfer from Ni 2+ to Ti 4+ , respectively [41].In the junction region of NiO/TNTs, the overlap of conduction band of 3d level in Ti 4+ with d-level of Ni 2+ enables charge transfer transitions between electrons in d-level of Ni 2+ and the conduction band of TiO 2 .It also decreases the energy gap between Ti 3d and O 2p states of TiO 2 [41,42] and facilitates absorption in visible spectrum.This reveals the good contact in between NiO and TiO 2 in consequence of the interdispersion of the two oxides [32].observed at 458.4 eV (Ti 2p 3/2 ) and 464.1 eV (Ti 2p 1/2 ), and both correspond to Ti 4+ in TiO 2 [43].
3.5.Photocatalytic Activity NiO/TNTs.Figure 7(a) shows the solar-light-induced photocatalytic activity of the NiO/TNTs.The initial concentration ( 0 ) is the MB concentration after adsorption-desorption equilibrium.Irrespective of NiO concentrations from 50.8 to 51.7% of dye is removed under dark conditions (30 min).The good adsorption capability of NiO/TNTs is attributed to the large surface area, which enables the MB molecules to diffuse freely inside NiO/TNTs [45].43.25% of the MB was removed in the control experiment.The photocatalytic reactions for all samples followed pseudo-first-order reaction kinetics, which is expressed by where  is the first-order reaction constant,  0 and  are the initial and the final concentrations of MB dye, respectively.The kinetic plot and photocatalytic results are shown in Figure 7(b) and Table 3, respectively.All NiO/TNTs samples show higher degradation efficiency than TNTs.The loading of NiO in TNTs resulted in a doubled reaction rate (0.004 min −1 ) with 86% of MB removal than that of pure TNTs ( = 0.002 min −1 ) with 68% MB removal.In the initial, NiO/TNTs (2.5 M) shows higher reaction rate than NiO/TNTs samples with lower concentrations (0.5 M and 1.5 M).This could be contributed to the distinct hump observed in the diffuse reflectance spectra (Figure 5) of NiO/TNTs (2.5 M).However, the solarlight-induced activities of all NiO/TNTs samples start to slow   down and reach an identical MB degradation (86%) towards the end of reaction.The increase in degradation efficiency of NiO/TNTs is mainly due to the efficient electron-hole pairs separation and visible light absorption.The presence of NiO develops a p-n junction to separate electron-hole pairs effectively, while the visible-light property is due to charge transfer transition from the electron donor levels formed by the 3d orbitals of substituted Ni 2+ to the conduction bands of TiO 2 [46].Further, Ni 2 O 3 has a dark colour, and it facilitates the absorption of visible light [17].The other contributing factor is the presence of Ni 2 O 3 which creates Ni 2+ vacancies in NiO which can lower the electrical resistance of NiO [44].
A competitive adsorption on the active sites between the reactant and the intermediate products reduces the degradation rate towards the end of reaction [47].Hence, the accessibility of reactant to the active sites is affected resulting in a nonsignificant difference of degradation efficiency for all NiO/TNTs samples when reactions end.In addition, the similar Ni content in 0.5 M NiO/TNTs (0.8 wt%) and 1.5 M NiO/TNTs (1 wt%) could also result in a nonsignificant difference of degradation rate for both samples.Overall, it can be concluded that the loading of NiO effectively improve solarlight-induced photoactivity.Though NiO/TNTs with higher NiO concentration possess better visible-light absorption properties, it has no direct effect on the enhancement of photocatalytic activity (Figure 5).

Degradation Mechanism.
The degradation mechanism of MB over NiO/TNTs and the effect of p-n-junction are illustrated in Figure 8.In presence of solar irradiation, the electrons are excited from valence band of TiO 2 to the conduction band.As indicated in XRD and XPS results, a major part of Ni 2+ ions segregates as separate NiO nanoparticles on the surface of TNTs, whereas the remainder portion is incorporated substitutionally in TiO 2 lattice.When segregated NiO nanoparticles and TiO 2 integrate, a p-n-junction will be formed between p-type NiO (p-NiO) and n-type TiO 2 (n-TiO 2 ).At the equilibrium, the negative charge and  positive charge will be formed in p-NiO region and n-TiO 2 region, respectively.When the p-n junction is irradiated by photons, the photogenerated holes flow to valence band of NiO nanoparticles (negative field), while the electrons flow to conduction band of TiO 2 (positive field) [31,33].Therefore, the electrons on the TiO

Conclusions
Self-organized and vertically oriented TNTs were successfully loaded with NiO nanoparticles.XRD and XPS results reveal that a large part of Ni 2+ ions segregate as a separate NiO and as a major phase, and the remainding part is incorporated substitutionally in TiO 2 lattice.Both the TNTs and NiO/TNTs photocatalysts possess large surface area which facilitates diffusion of MB molecules inside channel of samples and hence enhances its adsorption capability.Solarenergy-induced photocatalytic activity showed the degradation rate of MB is independent of NiO concentration.Overall, the NiO/TNTs samples exhibited higher degradation efficiency compared to that of TNTs.

Figure 7 :
Figure 7: (a) Photocatalytic degradation of MB over TNTs samples with varying NiO concentrations under solar light irradiation and (b) kinetic plot of TNTs samples with varying NiO concentrations.

Table 1 :
Experimental conditions and elemental composition of the prepared composite samples.
Methylene blue (MB) dye was chosen as the model pollutant to evaluate the photocatalytic activity of the pure and NiOloaded TNTs.Direct solar light was employed to illustrate the possibility of solar light utilization.

Table 2 :
Lattice parameters for TNTs and NiO/TNTs samples with different concentrations.

Table 3 :
Photodegradation of MB dye for TNTs and NiO/TNTs samples under solar light irradiation.Apparent rate constant deduced from linear fitting of ln( 0 /) versus reaction time.
a After reaction for 7.5 h.b c The control was the photolysis of MB dye.