Preparation and characterisation of manganese–TiO 2 nanocomposites for solar water splitting

Introduction

Depletion of non-renewable fossil fuels and global warming problem have made hydrogen an ideal future energy source due to its high energy content and zero emission of greenhouse gases. Currently, the majority of hydrogen demands are sourced from steam reforming of natural gas; however, this process does not mitigate dependence on fossil fuels and generates large amount of carbon dioxide. Alternatively, photocatalytic water splitting has been considered as one of the most promising approach for renewable hydrogen production, ever since the discovery of photoelectrochemical water splitting at a TiO₂ electrode by Fujishima and Honda, due to its environmental friendliness and potential contribution to future energy needs.^1,2 Various researches have shown that TiO₂ is an ideal photocatalyst due to its relatively high reactivity and chemical stability under the irradiation of light.^3–9 However, the application of TiO₂ is limited by the fast recombination rate of the photogenerated electron–hole pairs and the high band gap (3.2 eV), which diminishes its absorption in the visible light range of the solar spectrum.^10,11 To efficiently utilise solar energy, many efforts have been devoted to extending the spectral response of TiO₂ to the visible spectrum by dye sensitisation, metal complex sensitisation, coupling with a small band gap semiconductor and doping transitional metal ions or non-metal atoms, and so on.^12–19 Unlike most of the metal dopant, manganese doped TiO₂ has generated considerable interests as a photocatalyst, showing optical response in the visible region. Different methods were used to prepare Mn doped TiO₂, but most of these methods were multistage, time consuming and expensive, and their reproducibility was weak.^20–24 Therefore, a simple, facile and inexpensive method of synthesising Mn doped TiO₂ is necessary. Until now, very less research has been done to introduce manganese into the TiO₂ nanotube array by anodising process. To the best of our knowledge, report on the preparation of Mn–TiO₂ nanotube composite films by single step anodisation is lacking. In addition, to our knowledge, the quantity effect of manganese doped in TiO₂ nanotubes for H₂ generation in photocatalytic activity has not been reported. In the present study, we report a simple approach to fabricating Mn–TiO₂ nanotube composites by a single step anodising of titanium substrate in a non-aqueous bath containing dimethyl sulphoxide (DMSO) fluoride electrolyte and various amounts of potassium permanganate (KMnO₄). For the first time, we used KMnO₄ as the manganese source. The TiO₂ nanotube films with different amounts of manganese were obtained by controlling the concentration of KMnO₄ in anodising electrolyte. The morphology and structure were characterised by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX) and X-ray diffraction (XRD). Optical properties were investigated by UV–Vis diffuse reflectance spectra.

Experimental

All the chemicals used were of analytical grade without further purifying before experiment, and solutions were prepared with distilled water.

Mn–TiO₂ nanotube composites were synthesised by anodising of titanium in a proper electrolyte. Before anodising, a piece of titanium sheet (99.99% purity, 1 mm thick) was cut into desired dimension and the titanium electrodes were first mechanically polished with different emery type abrasive papers (with the following grades: 400, 800, 1000, 1500 and 2000), rinsed in a bath of distilled water and then chemically etched by immersing in a mixture of HF and HNO₃ acids for 30 s. The ratio of components HF/HNO₃/H₂O in the mixture was 1:4:5 in volume. The last step of pretreatment was rinsing with distilled water. The anodising process was carried out in an electrolytic cell using a titanium foil as anode and platinum foil with ∼14 cm² geometric areas as cathode. Anodising was performed in a solution of DMSO and HF, followed by the dissolution of different concentrations of potassium permanganate. The temperature of the electrolyte was maintained during the anodising process at (25 ± 1)°C. Anodising was performed for 6 h at a constant potential of 40 V using a controlled DC power supply source (ADAK, PS405). A schematic of the pretreatment method of titanium sheets and process of producing Mn–TiO₂ nanocomposite films on titanium is shown in Fig. 1. After anodising, the as formed samples were annealed in oxygen atmosphere at 430°C for 2 h (5°C min^− 1) to obtain crystalline samples that lead to composites films with varied weight loading percentage of manganese. In the present work, we compare the photocatalytic performance of bare TiO₂ and Mn–TiO₂ films with different quantity of manganese referred to as Mn–TiO₂–1, Mn–TiO₂–2 and Mn–TiO₂–3. Table 1 summarises the experimental conditions for four different samples.

OPEN IN VIEWER

1

Schematic presentation of pretreatment method of titanium sheets and process of producing Mn–TiO₂ nanocomposite films on titanium foils

OPEN IN VIEWER

Table 1

Experimental parameters for fabrication of different films

Samples	Anodising solution	Anodising condition
Bare TiO2	98 mL DMSO +2 mL HF	40 V for 6 h at room temperature
Mn–TiO2–1	97 mL DMSO +2 mL HF +1 mL H2O+0.2 g KMnO4	40 V for 6 h at room temperature
Mn–TiO2–2	97 mL DMSO +2 mL HF +1 mL H2O+0.4 g KMnO4	40 V for 6 h at room temperature
Mn–TiO2–3	97 mL DMSO +2 mL HF +1 mL H2O+0.6 g KMnO4	40 V for 6 h at room temperature

The surface morphologies of all samples were characterised by field emission SEM (FE-SEM, Hitachi S-4160, Japan), and the elemental composition was estimated by EDX. Elemental mapping was conducted by energy dispersive spectroscopy. The crystalline phases were identified by XRD with Equinox 3000 diffractometer (Inel, France). Diffraction patterns were recorded in the 2θ range from 20 to 80° at room temperature. The optical absorption of the samples was determined using a diffuse reflectance UV-visible spectrophotometer (JASCO V-570).

Photocatalytic water splitting to generate hydrogen was carried out in a three-branch quartz cell. A 200 W xenon lamp was used as the light source. The luminous intensity of the xenon lamp was 100 mW cm^− 2. The Mn–TiO₂ and bare TiO₂ sample was fixed in the cell containing 1.0 M NaOH (pH 13.6) solution facing the lamp. Hydrogen evolution was measured for 360 min, and H₂ gas was collected using the water displacement technique. A schematic diagram of the experimental set-up for photocatalytic water splitting was shown in Fig. 1. A Pt coil served as the cathode. The cathode was inserted into a burette where the hydrogen was collected via electrolyte displacement. The volume of hydrogen was measured by directly reading the variation of the electrolyte level in the burette for various times.

Results and discussion

Figure 2 shows top view FE-SEM images of Mn–TiO₂–1, Mn–TiO₂–2 and Mn–TiO₂–3 respectively, which clearly shows formation of films on the surface of titanium. In Fig. 2a and b, different samples displayed nanotube arrays, in which the surface of them was open. In Fig. 2a, highly ordered and vertically aligned nanotubes were formed on the Ti substrate. The average pore diameter of these nanotubes as calculated from SEM images is 70–100 nm, and wall thickness is 20–30 nm. In Fig. 2b, a combined structure of nanotubes and nanopores was formed. It seems that a porous structure is emerging on the surface of the nanotubes. In Fig. 2c, no nanotubes formed. Porous films were formed instead (Mn–TiO₂–3). It can be said that when the potassium permanganate concentration in anodising solution increased to 4 g L^− 1 (Fig. 2c), the formed nanotube array becomes very non-uniform, implying that an appropriate concentration of dopant is important for the structure of nanotube arrays. From the cross-sectional view of Mn–TiO₂–1 sample in Fig. 2d and e, it can be seen that the formed nanotubes are parallel aligned, opened on top and have a length in the range of 7–8 μm.

OPEN IN VIEWER

2

SEM top view images of Mn–TiO₂ with different magnification: a Mn–TiO₂–1; b Mn–TiO₂–2; c Mn–TiO₂–3; d, e cross-sectional image of Mn–TiO₂–1 with different magnifications

The XRD patterns of Mn–TiO₂ film are shown in Fig. 3. Diffractions that are attributable to anatase TiO₂ are clearly observed in the calcinated sample. The Ti peaks were due to the titanium substrate. XRD pattern of the synthesised Mn–TiO₂ shows the diffraction peaks of both MnO₂ and TiO₂, and two diffraction peaks of MnO₂ are found in this sample.

OPEN IN VIEWER

3

XRD patterns of Mn–TiO₂ sample: a as prepared; b annealed at 430°C

The elemental compositions of the as prepared Mn–TiO₂ sample were studied by EDX and energy dispersive X-ray mapping, and the results are displayed in Fig. 4. The data confirm that the samples are composed of Ti, Mn and O. The occurrence of traces of contaminants such as the fluorine, sulphur and potassium from precursors is also observed. The elemental maps demonstrated that O, Ti and manganese are homogeneously distributed on the surface of the sample.

OPEN IN VIEWER

4

EDX spectra and EDX mapping of MnTNT sample, indicating spatial distribution of different components on surface of MnTNT sample

The optical properties of different samples were studied. Figure 5 shows the optical band gap energy of bare TiO₂ and Mn–TiO₂ samples annealed at 430°C respectively. The reflectance data, reported as F(R) values, have been obtained by application of the Kubelka–Munk algorithm. The band gaps of the samples have been deduced from the Tauc plot. Figure 5 is the plot of [F(R)hυ]½ versus photon energy. The extrapolation of [F(R)hυ]½ to the abscissa at zero F(R) provides the band gap energy as ∼3.22 eV for bare TiO₂ and the band gap energies of ∼ 3.0, 2.90 and 2.72 eV for Mn–TiO₂–1, Mn–TiO₂–2 and Mn–TiO₂–3 respectively. It can be seen that compared with bare TiO₂, all of the Mn–TiO₂ samples exhibit a red shift of absorption edge and band gap decrease with the increase of manganese concentration. These results indicate that the doping of Mn in TiO₂ films decreases the band gap.

OPEN IN VIEWER

5

Plot of versus hυ employed to calculate band gap value of different samples

The primary purpose of this work is to investigate the activities of TiO₂ modified with different amounts of Mn through water splitting over these samples. Figure 6 shows the amount of hydrogen generated as a function of time using bare TiO₂ and Mn–TiO₂ samples under Xe light illumination. Control experiments indicated that no appreciable hydrogen production was detected in the absence of either light irradiation or photocatalyst, suggesting that hydrogen was produced by photocatalytic reactions on samples. In all catalysts, hydrogen evolves steadily over extended periods of time. It can be found that Mn–TiO₂–1 sample performed the highest photocatalytic activity under light irradiation. The total amount of H₂ that evolved on the Mn–TiO₂–1 sample was 96 μL cm^− 2 after 6 h, which is ∼1.3 times higher than that on the sample Mn–TiO₂–2 (74 μL cm^− 2), 2.1 times higher than that on the sample Mn–TiO₂–3 (45 μL cm^− 2) and 5 times higher than that on the bare TiO₂ sample (19 μL cm^− 2).

OPEN IN VIEWER

6

H₂ production by water splitting over different samples under irradiation

Conclusion

The major conclusions from the present work are as follows:

Undoped TiO₂ and Mn doped TiO₂ films were successfully synthesised using single step anodisation method.