Abstract

A series of S/Gd3+-codoped TiO2 photocatalysts were synthesized by a modified sol-gel method. The materials were characterized by X-ray diffraction (XRD), Raman spectroscopy, Fourier transform infrared spectroscopy (FTIR), UV-visible diffuse reflectance spectroscopy, scanning electron microscopy (SEM)/energy-dispersive X-ray spectroscopy (EDX), and transmission electron microscopy (TEM)/energy-dispersive spectroscopy (EDS). Laboratory experiments with Indigo Carmine chosen as a model for organic pollutants were used to evaluate the photocatalytic performance of S/Gd3+-codoped TiO2 under visible-light with varying concentrations of Gd3+ ions in the host material. XRD and Raman results confirmed the existence of anatase phase TiO2 with particle size ranging from 5 to 12 nm. Codoping has exerted a great influence on the optical responses along with red shift in the absorption edge. S/Gd3+-codoped TiO2 showed significant visible-light induced photocatalytic activity towards Indigo Carmine dye compared with S-TiO2 or commercial TiO2. TiO2-S/Gd3+ (0.6% Gd3+) degraded the dye ( = 5.6 × 10−2 min−1) completely in 50 min.

1. Introduction

Recent advances in nanotechnology have shown significant interest in the study of semiconductor nanocrystals owing to their unique electronic and optical properties due to quantum confinement effect [15]. Semiconductors have attracted much attention in the past few decades due to their unique photocatalytic, magnetic, and optical properties. They have been widely used in solar energy applications, such as photovoltaic, photochemical, and photocatalytic remediation [6, 7]. Photocatalytic activities of semiconducting inorganic solids have attracted passionate research interest in the past few years, where their potential use as catalysts for photodegradation of toxic organic pollutants in water has been fully or partially investigated [810]. Semiconductors (such as TiO2, ZnO, Fe2O3, CdS, and ZnS), which are characterized by filled valence band and empty conduction band [11], are proved to be important materials due to the electronic structure of the metal atom in chemical combination.

Nowadays scientific and engineering interest in TiO2 semiconductor have received much attention owing to its high photocatalytic activity, strong oxidizing power, low cost, chemical and thermal stability, resistance to photocorrosion and nontoxicity, and its favorable optoelectronic property [12]. However, the wide band gap (anatase: 3.2 eV, rutile: 3.0 eV) limits its photocatalytic efficiency. Even though the anatase is known to be the most reactive phase of TiO2 than the rutile crystalline phase, it has low quantum yield for oxidation steps (~5%) as a result of rapid recombination of photogenerated electron-hole pairs [13]. The high intrinsic band gap of pure TiO2 operates effectively as a photocatalyst in the UV region of the electromagnetic spectrum. Thus, pure TiO2 is able to use only around 4% of terrestrial solar spectrum because of its wide band gap [14]. In order to harness the full potential of the sunlight in photocatalysis, various methods, such as sensitization [15, 16], composite semiconductor coupling [17, 18], and doping of TiO2 [19, 20], have been developed to overcome this challenge.

Nonmetal doping as an alternative for improving the visible light response of TiO2 has been employed [2123], and extensive research work has been done on synthesis of N-doped, C-doped, S-doped, and F-doped TiO2 [2427]. The doping of nonmetal could narrow the band gap by modifying the electronic structure around the conduction band of TiO2. Theoretical calculation has shown that band gap narrowing originates from the electronic perturbations caused by the change of lattice parameters and/or by the presence of the trapped states within conduction and valence bands of TiO2 [2834].

Many of the recent efforts and strategies revealed that codoping of TiO2 with a metal and nonmetal can result in the development of a highly efficient visible light active photocatalyst [3537]. A series of studies reported the characteristic behavior of visible-active metal-doped semiconductor photocatalysts, including noble metals [38], rare earth metals [39], and transition metals [40]. Several rare earth metal ions such as La3+, Eu3+, Nd3+, and Ce3+, when doped in TiO2, showed considerable shift of absorption towards visible region [4144]. Additionally, lanthanide ions have the aptitude to form complexes with a variety of organic pollutants by the interaction of functional groups with f-orbitals of the lanthanides. This could provide a way to concentrate organic pollutants at the semiconductor surface [45].

Intrigued by these observations, we synthesized highly efficient visible light active S/Gd3+-codoped TiO2 nanoparticles by modified sol-gel method. The preparation of S/Gd3+-codoped TiO2 was optimized by varying the amount of Gd3+ to obtain an effective photocatalyst. The catalysts exhibited a higher visible light photocatalytic activity towards the degradation of Indigo Carmine (IC) in aqueous solution.

2. Experimental Details

All chemicals used were of analytical grade and used without further treatment. Titanium (IV) isopropoxide, Ti(OC3H7)4 (97%), Indigo Carmine, gadolinium (III) nitrate hexahydrate (Gd(NO3)3·6H2O), titanium (IV) oxide (product number 634662, 99.5%, <100 nm), and 2-propanol, C3H8O (99.8%) were purchased from Sigma Aldrich, Germany. Thiourea, CS(NH3)2 (99%) was purchased from Hopkin and Williams Ltd., England.

2.1. Sample Preparation

S/Gd3+-codoped TiO2 was prepared by adding titanium isopropoxide (10 mL) to 2-propanol (50 mL) slowly and the mixture stirred for 30 min. A calculated amount of gadolinium (III) nitrate hexahydrate was dissolved in DI water (2 mL) and added to the mixture to give Gd : Ti of 0.2–1.0% and the mixture was further stirred for 1 h. Then thiourea (3.0 g), dissolved in DI water (5 mL), was added slowly with vigorous stirring for 2 h. The resulting mixture was dried in air in an oven at 100°C for 12 h and calcined at 500°C for 2 h. TiO2-S/Gd3+ (0.0% Gd3+) sample was prepared in a similar manner without gadolinium (III) nitrate hexahydrate.

2.2. Evaluation of Photocatalytic Activity

The photocatalytic activity of the materials was evaluated through a suspension of 100 mg of the catalyst in 100 mL of aqueous solution of Indigo Carmine (20 mg/L) that was kept under magnetic stirring and visible light irradiation filtered using dichroic UV filter ( nm). A 150 W tungsten filament (Eurolux), kept at a distance of 11 cm from the reaction vessel, was employed as a source of radiation. The samples were magnetically stirred in the dark for 1 h prior to illumination to allow for adsorption equilibrium. Aliquots of the suspension (5 mL) were withdrawn at periodic time intervals using disposable syringe and filtered through 0.4 μm PVDF membrane filter at 30 min intervals for 4 hrs. The concentration of the Indigo Carmine remaining after illumination in the supernatant solution was determined using a Shimadzu UV-2450 spectrophotometer at  nm. The photodegradation performance of the process was assessed in terms of decolorization efficiencies and kinetic studies.

2.3. Characterization

X-ray diffraction (XRD) measurements were performed on X-ray diffractometer (Rigaku Ultima IV) at 40 kV and 30 mA with Cu Kα radiation () and K-beta filter. Measurements were performed using a scintillation counter in the range of 5–90 deg at a speed of 2.0 deg/min. FT-IR spectra of the samples were recorded on PerkinElmer FT-IR spectrometer (Spectrum 100). Raman spectra of the samples were measured on PerkinElmer Raman microscope (Raman Micro 200). Optical properties were investigated using UV-vis absorption and UV-vis diffuse reflectance spectroscopy on a Shimadzu UV-2540 (Japan). BaSO4 was used as the reflectance standard. Scanning electron microscopy (SEM) studies were obtained on a TESCAN (Vega 3 XMU) to observe the surface morphology of the powders and high resolution transmission electron spectroscopy (HRTEM) analysis was performed on a JEOL field emission electron microscope (JEM-2100F) to observe the surface morphology, structure, and grain size of the nanoparticles. Energy-dispersive X-ray spectroscopy (EDX) attached to the SEM and energy-dispersive spectroscopy (EDS) attached to HRTEM were used to determine the surface elemental composition.

3. Results and Discussions

3.1. FT-IR Analysis

The FT-IR spectra provide information on the surface chemistry of the oxide nanoparticles which are usually affected by the hydration layers or organic species which are characterized by several bands as demonstrated by several studies. Figure 1 shows the FT-IR spectra of S-TiO2 and S/Gd3+-codoped TiO2 with different Gd3+ concentrations calcined at 500°C. The most common feature observed in all the spectra is the appearance of (i) broad bands below 1000 cm−1 assignable to the Ti–O–Ti crystal vibration [46] and (ii) those in the regions between 1620–1635 cm−1 and 3350–3450 cm−1 both due to bending vibrations of adsorbed water molecules and stretching vibrations [47] from the hydroxyl groups, respectively. The Ti–S peaks are the broad intense peaks below 1000 cm−1 [48]. The peak at around 2340 cm−1, which is dominant for the S-doped sample but absent in the commercial TiO2, may be due to out-of-phase stretching of the −N=C=O bond [49], which was left behind on the surface by incomplete decomposition of thiourea [50]. The shift to the larger wavenumbers (from ~665 to ~722 cm−1) and the sharpening of the Ti–O–Ti bands may be due to gadolinium doping. There is, however, no peak centered at 1389 cm−1 which is due to the bending vibrations of C–H bond in the catalyst [51] and no residual alkoxy peaks were observed indicating the absence of impurities in the samples after calcinations.

3.2. Raman Spectroscopic Analysis

Raman spectroscopy provides additional information about the anatase crystallinity. Factor group analysis shows six Raman active modes () for anatase TiO2 [52]. In this study, pronounced vibrational modes were observed for all the samples as shown in Figure 2. All the Raman spectra measured confirmed anatase phase of TiO2. The Raman modes can be assigned to the Raman spectra of the anatase crystal.49: ~143 (Eg), 197 (Eg), 398 (), 515 (), and 639 cm−1 (Eg). Almost all the peaks match quite well with those reported in the literature, confirming the formation of pure anatase phase. However, herein, only five pronounced bands appear probably due to overlapping of and bands. At the same time, the blue or red shift, accompanied with some bands, may be explained in terms of reduction in crystallite sizes in the doped TiO2 samples as noted earlier by others [53]. It is worth noting that reduction in the crystallite sizes has also been confirmed by the XRD measurements mentioned below. The highly intense Eg mode is due to scattering from the (110) face and mode from the (001). There was a shift in the position of the anatase Eg mode as a result of gadolinium doping from 637 cm−1 to 643 cm−1 for gadolinium-doped and -undoped S-TiO2, respectively. This shift may be ascribed to Gd2O3  Eg mode in the samples. There was no intense Raman band at ~333 cm−1 which is as a result of symmetry vibration of the cubic Gd2O3 indicating the partial replacement of Ti by Gd in the TiO2 crystal lattice. The sharp, narrow intense peaks show crystalline state and the crystallinity was not changed as a result of doping.

3.3. X-Ray Diffraction Study

The size and the crystalline phase formation were investigated by X-ray diffraction. The XRD measurements for S-TiO2 and S/Gd3+-codoped TiO2 are shown in Figure 3. All the samples show anatase phase. The diffraction patterns of the anatase S/Gd3+-codoped TiO2 and S-TiO2 powders were compared with JCPDS database. The peak positions and their relative intensities are consistent with the standard powder diffraction pattern of anatase TiO2. XRD patterns (Figure 3) exhibited strong diffraction peaks at 25° and 48°, indicating TiO2 in the anatase phase. All the peaks are in good agreement with the standard spectrum (JCPDS number 84-1286). The peaks at values of 25.3, 37.8, 48.1, 54.0, 55.1, 62.9, and 75.2 correspond to the (101), (004), (200), (105), (211), (204), and (215) planes, respectively. All these peaks depict the anatase signature peaks. The sharp, intense anatase peaks show an improvement in the degree of crystallinity with fewer lattice defects. The XRD patterns show that there was no second-phase peak and also that the doping was successful. Also, in S-TiO2 system, the S atoms may be attributed to be replaced some of the oxygen atoms in the TiO2 crystal lattice. The absence of any residual peak(s) explains the fact that there is no sulphur found on the surface of the nanocrystal. From the XRD results it can be concluded that sulphur is in a high dissolution state in the TiO2 lattice rather than an isolated species on the surface of TiO2. The presence of sulphur and gadolinium was, however, confirmed by elemental analysis using EDS and EDX.

The average crystalline size was estimated from Scherrer’s equation on the anatase (, 37.8, and 48.1°) diffraction peaks (the most intense peaks for each sample): where is the crystal size of the catalyst, is the X-ray wavelength (1.54056 Ǻ), is the full width at maximum (FWHM) of the diffraction peak (radian), is a constant (0.9), and is the diffraction angle at the maximum [54]. The average crystalline size of S- (, 0.4, 0.6, 0.8, and 1.0) was calculated to be between 5 and 12 nm.

3.4. TEM and SEM Study

HRTEM and SEM were used to observe the uniformity, morphology, and microstructure of the as-prepared S-TiO2 and S/Gd3+-codoped TiO2 nanoparticles calcined at 500°C. Figures 4(a) and 4(b) show a typical HRTEM and SEM image of the prepared TiO2-S/Gd3+ (0.6% Gd3+) photocatalysts, respectively. The particles are small and nearly spherical in shape. All the samples prepared show regular morphology, regardless of S/Gd ratio. The uniform morphology, with uniform size (5–12 nm), is due to improvement in the sol-gel process which resulted in the mobilization of Ti and O homogeneity in the TiO2 crystalline structure.

The elemental composition of the prepared samples was estimated by EDX and EDS analyses. Figures 5(a) and 5(b) show the EDX and EDS spectra of TiO2-S/Gd3+ (0.6% Gd3+), respectively. Both spectra of the codoped TiO2 confirm the presence of Ti, Gd, O, and S. The spectra indicate that the main components are Ti and O with low contents of Gd and S. This may be due to the formation of S/Gd3+-codoped TiO2.

3.5. UV-Visible Study

The light absorption characteristics of TiO2 usually change after doping with nonmetal doping [2123] and metal/nonmetal codoping [3537]. Figures 6 and 7 show a comparison of the UV-vis absorption edge and UV-vis diffuse reflectance of commercial TiO2, S-TiO2, and S/Gd3+-codoped TiO2 (with different concentrations of Gd3+), respectively. The absorption spectrum of TiO2 consists of a single broad intense absorption around 400 nm due to the charge transfer from the valence band (mainly formed by 2p orbitals of the oxide anions) to the corresponding conduction band (mainly formed by 3d t2g orbitals of Ti4+ cation [55]. The onset of the absorption edge for pure TiO2 at Ca. 385 nm is consistent with intrinsic band gap absorption of pure TiO2. However, in comparison with the bulk TiO2, the TiO2-S (and N from incomplete decomposition of thiourea, shown in Figure 1) shows two prominent features: (i) the appearance of a new absorption shoulder around 390–500 nm and (ii) the greatly enhanced absorbance in the range of 550–750 nm. The overall visible-light absorbance increased with Gd3+ doping. Such red shift in absorbance in the visible-light range is partly due to the yellow characteristic colour of the S-TiO2 and the synergistic effect of S/Gd3+ codoping. In addition, in order to obtain a synergistic effect between S-TiO2 and S/Gd3+-codoped TiO2, it is important to control the addition ratios of Gd, which lead to an optimum photocatalytic activity. It was obvious that the TiO2-S/Gd3+ (0.6% Gd3+) nanoparticles exhibited the best visible-light absorption with improved band gap.

Figure 8 shows a plot of the Kubelka-Munk function, , versus wavelength based on the following Kubelka-Munk equation: where reflectance, . The plot shows a significant red shift in the absorption coefficient of the S/Gd3+-codoped photocatalyst. Figure 9 contains Tauc [56] plots constructed from the Kubelka-Munk data (obtained from the UV-visible diffuse reflectance) for the photocatalysts as well as for the commercial benchmark sample. The band gap energies for the various samples were calculated by plotting the values of versus the photon energy (). The band gap values are estimated from a plot of versus . The indirect band gap values are summarized in Table 1. The number depends on the nature of the electronic transition and is 1 for a direct and 1/2 for an indirect band gap for semiconductors with a crystalline structure [56].

Extrapolation of this line to the photon energy axis (Figure 9) gives the semiconductor band gap, which is a key indicator of its visible light efficiency. Commercial TiO2 also shows an absorption edge that fits well to the energy axis. Comparing the band gap of commercial TiO2 and the S/Gd3+-codoped samples, the latter shows significant reduction in the band gap with gadolinium doping. The increase in the band gap with gadolinium doping may be due to the dominance of the d-f transitions over the sp-d transitions.

3.6. Photocatalytic Activity

The visible-light activity of the as-prepared nanoparticles was evaluated for the degradation of organic pollutants in aqueous solution, using Indigo Carmine (IC) dye as the model pollutant. The optical absorption peak at 610 nm was chosen to monitor the photodegradation process. The UV-visible spectral change of 20 ppm IC as a function of irradiation time during the course of degradation by TiO2-S/Gd3+ (0.6% Gd3+) is shown in Figure 10.

With time increasing from 0 to 50 min, the characteristic absorption band at peaks of 610 nm decreased gradually, indicating that the IC was gradually photodegraded by the catalyst. The IC was completely removed from the solution after 50 min. The percentage removal for commercial TiO2, TiO2-S, and S/Gd3+-codoped TiO2 with different Gd3+ contents is shown in Table 1. The linear relationship of versus time (Figure 11) shows that the photocatalytic degradation of IC follows the pseudo-first-order kinetics: where is the normalized IC concentration, is the reaction time, and is the apparent reaction rate constant.

The value for (Table 1, Figure 12) for the TiO2-S/Gd3+ (0.6% Gd3+) (the most efficient photocatalyst) is  min−1 which is about nine (9) times that of the commercial TiO2,  min−1, further demonstrating that the S/Gd3+-codoped TiO2 exhibits high photocatalytic efficiency. The high photocatalytic performance of the codoped TiO2 could be assigned to a number of factors: (i) high visible-light absorption of the catalyst due to reduced band gap, (ii) the delayed electron-hole recombination due to trapping of electrons in the conduction band by Gd3+ orbital, (iii) efficient transfer of the charge carriers, and (iv) effective utilization of the charge carries by the reactants, as demonstrated by the UV-visible and UV-visible DRS analyses (Figures 6 and 7).

4. Conclusions

A series of S/Gd3+-codoped TiO2 nanoparticles were synthesized by a modified sol-gel method. The samples were characterized by various spectroscopic and analytical techniques. XRD and Raman analysis confirm the formation of pure crystalline anatase phase TiO2. The UV-vis and UV-vis DRS spectral analyses indicated that S/Gd3+-codoping causes a red shift in the absorption band, resulting in the reduction in band gaps. The synergic effect of codoping with S and Gd3+ is evident in the photocatalytic performance of the catalyst. The S/Gd3+ systems are very effective visible-light active photocatalysts for the degradation of Indigo Carmine. Highest photocatalytic activity was observed for the TiO2-S/Gd3+ (0.6% Gd3+) sample. The enhanced photocatalytic activity was mainly attributed to the small crystallite size, intense light absorption in the visible region, and narrow band gap energy.

Conflict of Interests

The authors declare that they have no conflict of interests regarding the publication of this paper.

Acknowledgments

The authors gratefully acknowledge the financial support from the Faculty of Science, University of Johannesburg and the National Research Fund (NRF) South Africa, and Nanotechnology and Applications Centre, University of Allahabad, Allahabad, India. The authors also wish to thank Dr. H Mittal and Dr. V. Parashar, University of Johannesburg, for their technical support.