Prospective aspects of preferential {001} facets of N,S-co-doped TiO₂ photocatalysts for visible-light-responsive photocatalytic activity

Abstract

In this report, we describe the synthesis of nitrogen and sulfur co-doped TiO₂ photocatalysts (NST) with preferential {001} facets by surfactant- and template-free OPM routes and crystallized through hydrothermal treatment. The precursor solution formed a coordination complex containing Ti–peroxo complex and chelating ligands (thiourea and urea were used as ligands for sulfur and nitrogen-donating sources, respectively). The effects of dopant concentration on the structural and morphological properties of the as-prepared NST photocatalysts were studied using X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), field emission scanning electron microscopy (FE-SEM), high-resolution transmission electron microscopy (HR-TEM), N₂-adsorption–desorption isotherms and defuse reflectance spectroscopy (DRS). The visible-light-driven photocatalytic activity of the as-prepared NST samples was tested for the photodegradation of organic compounds (i.e. rhodamine B and phenol). High-performance liquid chromatography (HPLC) and LC-MS spectroscopy were used to analyze the by-products from the photodegradation of RhB. Different scavengers were added to the photocatalysis system in order to identify the role of active species in the photodegradation of organic compounds on the surface of NST photocatalysts. It was found that hydroxyl radicals (˙OH) and photogenerated holes (h⁺) played an important role in the photodegradation of organic compounds under visible-light irradiation. The electronic and structural characterizations of the as-prepared NST samples proved the successful incorporation of dopant elements (i.e. N and S) into the crystal lattice of TiO₂ – which shifted the absorption edge shoulder from UV to the visible-light region, due to the bandgap transition. Another reason for the red shift of the absorption edge in the visible-light region is expected to be the formation of new energy levels near to the conduction bands, because of the incorporation of dopant elements (N and S) into the bandgap of the TiO₂ crystal lattice. More interestingly, the pristine NST-0 sample showed photocatalytic activity, expected to be due to the formation of a substrate–surface complexation, resulting in an absorption shift of TiO₂ into the visible light region due to the transfer of charges from the ligand (i.e. the attached dye molecule) to the titanium atoms. Moreover, the photocatalytic efficiency of the as-prepared NST samples was higher than those of the other samples. The prominent synergetic factors responsible for enhanced photocatalytic activity of doped NST samples include: (1) the presence of preferential exposed {001} facets of the anatase TiO₂ nanorods, and (2) the introduction of the shift in the absorption edge shoulder towards the visible-light region. The preferential {001} facets act as a reservoir for the photogenerated charge carriers (i.e. electron–hole pairs) and slow down their rate of recombination. The introduction of absorption shift facilitates the adsorption of organic compounds on the photocatalyst surface.

---

Introduction

Environmental problems cause serious concern worldwide and, in particular, those caused by the draining of pollutant fossil fuels into the water system is an alarming issue for ecosystems.^1,2 A growing concern has been raised to limit the damage to ecosystems from industrial drains. For this purpose, photocatalysis using heterogeneous semiconductors has been given weight as the most effective technique to solve the challenges of wastewater treatment.^1,3,4 In semiconductor engineering, titanium dioxide (TiO₂) has special importance due to its unique physical and chemical characteristics.^5,6 Unfortunately, the high value of its bandgap energy restricts its applications to the high-energy ultra-violet region, which constitutes only 4–5% of our solar spectrum.⁵ Therefore, efforts have been made to tailor the bandgap energy of TiO₂ and to utilize solar light for photocatalytic efficiency. Various modifications have been made to the crystal structure of TiO₂ including doping with metals and non-metals,^7–9 coupling with other semiconductor surfaces,^10–13 and dye sensitization.^14–16

By addressing different concerns and limitations in modifying the crystal structure of TiO₂ by other routes, it has been established that non-metal doping and, in particular, non-metal co-doping is the most effective, efficient, inexpensive and stable route to achieve visible-light-driven photocatalytic activity of TiO₂.^17–20 In particular, non-metal co-doping not only enhanced the absorbance edge of TiO₂ in the visible-light region but also facilitated the transition and/or separation of photogenerated charge carriers (i.e. electron–hole pairs).^{17,18,21–23} Traditional routes applied to the synthesis of N,S-co-doped TiO₂ materials are either the use of any S- and N-donating ligands (urea and thiourea, etc.) for the sol–gel process^24,25 or thermal annealing for TiO₂ crystallization in the presence of the above-mentioned modifiers.^26–28 The as-synthesized N,S-co-doped TiO₂ material showed prominent photocatalytic activity for the degradation of water pollutants under visible-light irradiation, but improvement is still needed for long-term utilization of their photocatalytic applications.²⁹ A general list of reported approaches includes the preparation of photocatalysts with a double-region-structure,³⁰ decorating the surface of TiO₂ with specific molecules,^31,32 and adjustment of pH to control the surface electric charge of TiO₂.³³ It was found that the photocatalytic efficiency of the as-prepared materials was not sufficient because the reported methods were not adequate to drive the visible-light activity of TiO₂.^29–32 To the best of our knowledge, no one has cited the preparation of N,S-co-doped TiO₂ nanorods using an OPM-assisted hydrothermal method.

While keeping these advantages in mind, we have developed a simple and more effective OPM-assisted hydrothermal route for the synthesis of N,S-co-doped TiO₂ photocatalysts, utilizing previous knowledge from our cited research work.^2,22,27 The resulting N,S-co-doped TiO₂ material showed enhanced visible-light absorption thresholds due to expected band-to-band excitation and lower bandgap energy values as compared to the pristine sample. The as-prepared N,S-co-doped TiO₂ nanorods showed good adsorption capacity and enhanced visible-light-driven photocatalytic activity for the degradation of organic compounds (i.e. rhodamine B and phenol). The as-prepared samples showed good recycling selectivity for the photodegradation of target pollutants. The enhanced photocatalytic activity of the as-prepared NST samples compared to the pristine sample has been studied by structural and morphological characterizations. Thus, this report may offer a promising, facile and environmentally benign approach to the design of highly efficient visible-light-activated photocatalyst material.

Experimental

Synthesis of photocatalysts

All reagents purchased were of analytical grade and were used as such without further purification. The N,S-co-doped TiO₂ photocatalysts were prepared by forming a basin solution of both doping ligands (i.e. thiourea and urea) in a NH₃/H₂O₂ mixture solution (molar ratio of 1 : 3). The concentration of doping ligands was controlled in accordance with the Ti molar ratio. At the same time, in another beaker, a Ti–peroxo precursor solution was prepared in the same molar ratio of NH₃/H₂O₂. To this Ti–peroxo precursor solution different amounts of doping ligand solution were added separately. Briefly described, metallic Ti (99.7%, Aldrich) of the appropriate amount was added to a beaker containing 80 mL mixture solution of (3 : 1 molar ratios) H₂O₂–NH₃ and the beaker was placed in an ice water bath. The mixture solution was left for a fixed time (2 hours), resulting in the complete dissolution of metallic Ti to obtain a clear yellow solution, confirming the formation of the Ti–peroxo complex. To this precursor solution, the basin solution containing an appropriate amount of doping ligands (thiourea and urea) was added to obtain a molar ratio of Ti : S/N (1 : 0.01, 1 : 0.02, and 1 : 0.03) and these were marked as NST-1, NST-2, and NST-3, respectively. A schematic pathway for the incorporation of dopant elements (N and S) in the crystal lattice of TiO₂ is presented in Fig. 1S.† Each resin solution was set for overnight stirring to obtain a precipitate. A pure TiO₂ resin solution was obtained without adding any chelating ligands, as described earlier,³⁴ and this was marked as NST-0. The resulting precipitate was vigorously stirred to remove any volatile solvents and was then annealed in a hydrothermal cell conducted under controlled temperature in a Teflon cylinder at 200 °C for 4 hours. All samples were prepared under the same conditions and the resulting crystalline material was quenched in an ice water bath and separated by centrifugation. The crystalline material was washed with doubly-distilled water until the pH was neutral and with isopropanol, and was dried at 60 °C for 8 hours.

Characterization

A Shimadzu XRD 6000 diffractometer was used for X-ray diffraction (XRD) analysis to determine the crystalline phase present in the samples. The instrument has a beam of Cu Kα (λ = 1.5406 Å) radiation with an exposure time of 1 s and a default scan angular pass of 0.02°. The well-known Scherrer's equation was used to estimate the lengths of crystallographic coherence and the full width at half maximum (FWHM) was estimated using pseudo-Voigt approximation from the deconvolution of each XRD peak.³⁵ The surface morphology and particle size of the as-prepared sample were calculated with the help of a JEOL JSM6701F microscope instrument for carrying out field emission scanning electron microscopy (FE-SEM) analysis and a JEM 2100 URP instrument was used for transmission electron microscopy (TEM). An FEI A Titan-Cube microscope (operated at 80 kV) with aberration-correction and equipped with a CETCOR of CEOS GmbH and a Cs corrector was used for high-resolution (HRTEM) studies. An EDS Thermo-Noran coupled with JEOL JEM 2100 which has a Si detector was used for Energy Dispersive X-ray Spectroscopy (EDS) to determine the chemical composition of all the samples. Diffuse reflectance UV-Vis spectra (DRS) were collected from a Cary 5G spectrophotometer to analyze the shift in absorption due to the incorporation of dopant impurities into the crystal lattice of TiO₂. The Kubelka–Munk function³⁶ was used to estimate the band gap energy of the as-prepared samples.

The photocatalytic activity of the as-prepared NST samples was tested for the photodegradation of organic pollutants (i.e. rhodamine B (RhB) and phenol) under visible-light irradiation. For this purpose, six Osram radiation lamps (Model = L18/10 – daylight intensity 18 W and maximum intensity at 440 nm) were used as the visible-light source with 25% of the total spectrum below maximum intensity. In detail, 10 mg of the as-prepared samples was dispersed into 20 mL of the 10 mg L⁻¹ of the organic compound solution in beakers and these were placed in a homemade photoreactor, maintained at 18 °C by a thermostatically controller. The resulting mixture solutions were placed in the dark overnight to reach an adsorption–desorption equilibrium prior to visible-light irradiation. The detailed description of the homemade photoreactor can be found in our recent publication.³⁷ UV-Vis spectroscopy (Shimadzu UV-1601PC) was used to monitor the photocatalytic oxidation reaction of the organic compounds for fixed intervals of visible-light irradiation. Different aliquots of the reaction solution were taken for fixed intervals of visible-light irradiation and were separated by centrifugation before performing the UV-Vis analysis. The formation of different by-products during the photocatalytic oxidation of RhB dye was monitored with the help of Mass Spectrometer (ESI-MS, Varian 310-MS) and high-pressure liquid chromatography instruments. The aliquots of the reaction solution were analyzed with the help of an Electrospray Ionization Spectrometer. A glass micro-syringe was used to inject the aliquots into the ESI source (20 mL min⁻¹ flow rate) and the spectra were obtained after stabilization of the equipment.

Additionally, different chemical scavengers were added to the NST-containing RhB solution systems to monitor the photodegradation pathway. For this purpose, silver nitrate (AgNO₃), dimethyl sulfoxide (DMSO), and sodium oxalate (SO) were added separately into the photocatalytic system as scavengers for conduction band (CB) electrons, ˙OH radicals, and photogenerated valence band (VB) holes (h⁺), respectively.³⁸ The photoluminescence (PL) technique was used to monitor the generation of hydroxyl radicals (˙OH) on the interface of photoilluminated photocatalyst/water in the presence of a probing molecule (i.e. terephthalic acid). This technique was initially described by Fujishima and Ishibashi.^39,40 The basic theme for the use of terephthalic acid (5 × 10⁻⁴ M of terephthalic acid in 2 × 10⁻³ M NaOH aqueous solution) as a probing molecule is its ability to produce a highly fluorescent product (2-hydroxyterephthalic acid) as a result of a rapid reaction with the freshly generated hydroxyl radicals in the solution.^41,42 The change in PL peak intensity signal gives an indication of the terephthalic acid hydroxylation reaction with the ˙OH radical at the interface of photocatalyst/water. The concentration of hydroxyl radicals generated is directly proportional to the PL intensity of the fluorescent product (2-hydroxyterephthalic acid) in water.^41,42 The experimental setup used for measurement of hydroxyl radical concentration is similar to the photocatalytic experiment. A Hitachi F-7000 fluorescence spectrophotometer was used to measure the PL spectra of the fluorescent product (2-hydroxyterephthalic acid). The mixture solution was exposed to visible-light for a fixed interval of time and then the PL intensity was measured.

Results and discussion

The EDS analysis was used to determine the concentration of doping elements in the as-prepared samples, where the NST-2 sample showed the presence of Ti, N, O and S elements, as shown in Fig. 1. The presence of minor peaks in the EDS spectrum for the copper element in the range of 1.0 keV is expected to arise from the TEM grid. The effects of dopant level on the crystallite size and phase structure of the as-synthesized samples were investigated with the help of X-ray diffraction analysis (XRD), as shown in Fig. 2. In particular, XRD patterns matched well for the TiO₂ anatase phase, for all samples regardless of the concentration of the dopant level [JCPDS no. 21-1272]. As described before, analytical techniques have proved the successful incorporation of dopant elements (N and S) into the TiO₂ crystal lattice, and the fact that any other phase peaks corresponding to either nitrogen oxide or sulfur oxide were absent in the XRD measurements revealed that dopant atoms are directly incorporated into the crystal lattice of TiO₂ without forming any extra impurity phase. Furthermore, it was observed that the intensity of the XRD peaks decreases with increasing dopant concentrations and the XRD diffraction peaks started to shrink due to a decrease in crystallinity degree and the formation of smaller crystallites of TiO₂ (as shown in Table 1). More interestingly, the effect of dopant level on the specific crystallographic direction of the TiO₂ crystal lattice was estimated using the well-known Scherrer equation,³⁵ as shown in Fig. 3. For this purpose, the ratio of calculated crystallite size was plotted against the molar ratio of Ti in differently doped samples. It was found that material anisotropies were detected for the growth of the ratio of crystallite size to exposed facets [004] which are parallel to [001], and similarly for [220] or [200] which are parallel to [110] and [100], respectively. Literature study shows that the presence of both directions [100] and [110] in the same plane is more probably due to the presence of a perpendicular plane in the [001] direction.⁴³ In particular, preferential growth of nanorods in the direction of [001] is expected due to the fact that the anatase TiO₂ crystalline lattice has more energetic facets in the planes of (001).^44,45 Thus, one can predict that the NST-2 sample will have twice the crystallite size in the [001] direction compared to the perpendicular direction. Another interesting fact that can be noted from Fig. 3 is that the ratio of crystallite size decreases with an increase in dopant concentration because of the inter-layer exchange of atoms in the crystal lattice of TiO₂. Thus, the lower concentration of dopants introduces anisotropic nanostructures parallel to the isotropic structure.

Fig. 1 The EDS spectrum of the NST-2 sample for better illustration in the marked region as shown in the inset of FE-SEM image.

Fig. 2 XRD patterns of the as-prepared samples with different dopant concentrations.

Table 1 Physical characteristics of pristine and N,S-co-doped TiO₂ samples with changing the molar ratio of dopant levels

Sample	Phase^a	Crystal size^b (nm)	SBET (m^2 g^−1)	Pore volume (cm^3 g^−1)	Average pore size (nm)	Porosity (%)	Relative crystallinity^c
a A stands for anatase phase.b Stands for the average crystallite size of anatase titania (TiO2) calculated using Scherrer equation from the broadening of the {101} diffraction peak.c Relative crystallinity of the anatase and calculated from the ratio of the relative intensity of the anatase {101} diffraction peak to the same diffraction peak for undoped sample. SBET, pore volume and pore size is calculated from the nitrogen adsorption–desorption isotherm as discussed in the Results and discussion section.
NST-0	A	23.7	55.3	0.41	34.7	53.7	1
NST-1	A	22.3	59.2	0.47	31.3	57.4	0.89
NST-2	A	21.8	63.7	0.52	27.4	59.2	0.85
NST-3	A	21.5	65.2	0.56	25.9	60.5	0.83

---

Fig. 3 Relationship between the anisotropy and dopant molar in the as-prepared samples.

Fig. 4 shows the FE-SEM images of the as-prepared NST samples. From the FE-SEM images, it was revealed that the pristine sample has the surface morphology of a nanorod shape. The increase in dopant concentration results in the evolution of an anisotropic nanostructure to form an isotropic structure, as depicted in the FE-SEM images. The structure of the as-prepared samples consisted of nanorods with a regular shape with an average length and diameter in the range of ca. 60–80 nm and ca. 20–30 nm, respectively. The elemental composition and surface mapping of the as-prepared samples are shown in Fig. 2S and Table 1S.† More interestingly, an increase in dopant concentration did not change the surface morphology of the as-prepared samples. The nanorod-shaped morphology of the samples was also revealed with the help of TEM analysis. Fig. 5(a) and (b) show the typical TEM images of the NST-0 and NST-2 samples – to observe the surface morphologies the samples were prepared by the OPM route via hydrothermal treatment (see Experimental section). The as-prepared samples have a nanorod shape with anisotropic structures. The length and diameter of the nanorods are in accord with the values calculated from FE-SEM analysis. From the HR-TEM images, it was revealed that the length of the nanorods decreases to a marginal value with an increase in dopant concentration, whereas a slight difference in the diameter values is also observed for the higher concentration of dopant level. This suggests that nanorods are made by the restructuring and fusion of different original close-contact TiO₂ crystalline lattices. These findings are in good agreement with the isotropic structure from the FE-SEM and XRD results. Fig. 5(c) and (d) show the HR-TEM and Fourier Transform (inset) images of the as-marked region, which reveal that the monocrystalline nature of the as-prepared samples is retained for the higher concentrations of dopant impurity. Close analysis of the HR-TEM and FT analyses revealed that no secondary phase was present and even the local agglomeration of dopant impurities did not happen (the formation of nitrogen oxide or sulfur oxide particles). Thus, these findings further support the earlier discussion that dopant atoms (N and S) are incorporated into the crystal lattice of TiO₂ without forming any extra impurity phase and is true even for the higher concentration of dopants. Similar findings were observed for the XRD results of all the as-prepared samples.

Fig. 4 FE-SEM images of the as-prepared samples: (a) & (b) NST-0 (pristine); (c) NST-1; (d) & (e) NST-2; (f) NST-3.

Fig. 5 TEM (a and c) and HRTEM (b and d) of the NST-0 and NST-2 samples. Inset shows the respective Fourier Transform (FT).

HR-TEM and FT results show that NST-0 nanorods have grown in the direction of [001] as the lattice spacing parallel to the exposed facets is ca. 0.235 nm from top to bottom, as discussed in the XRD Results and discussion section. Moreover, the isotropic nanostructures have the small crystallite size of nanorods and preferential facets of (101) for the NST-2 sample. One factor for this abnormal change in preferential facets is the change in the surface energy for different planes of crystallite.^46,47 The change in surface energy of crystalline planes is responsible for the predominant evolution of the nanorod morphology with the incorporation of dopant level. The average values of surface energy reported for the TiO₂ anatase phase are 0.53 J m⁻² for {100}, 0.44 J m⁻² for {101}, and 0.90 J m⁻² for {001}, respectively.^46,48,49 Thus, the preferentially exposed facet for the anatase phase of TiO₂ is high-energy {001} as expected from the cited values and is noted to be the most stable phase for thermal treatment. Conversely, HR-TEM analysis revealed that the incorporation of dopant atoms (N and S) into the TiO₂ crystal lattice preferred exposed (001) planes either without thermal annealing of the material for high temperature or without the post-calcination step. One can predict from these findings that the existence and incorporation of dopant atoms have a significant impact on the structure and morphology of the materials by either affecting the mechanism of crystal growth or changing the surface energy.^47,50 Therefore, our findings support the statement that dopant atoms (N and S) have played a significant role in the mechanism involved in crystal growth and controlling the morphology of the nanostructure materials.

Fig. 6 shows the nitrogen adsorption–desorption isotherms of the NST-0 and NST-2 samples, while the inset shows the pore diameter distribution of the pore volume of the as-prepared samples. Similar patterns were observed for the isotherms and pore size distributions of other samples (not shown here). The presence of large mesopores (i.e. a particle gap range between 2 and 50 nm) was indicated by the existence of type IV isotherms for the NST-2 sample with very narrow hysteresis loops close to unity at relative pressures. Moreover, type II isotherms were observed for nitrogen isotherms of the adsorption branch which steadily increases on approaching unity in relative pressure (P/P₀) due to the existence of mesopores (>100 nm) with a high external surface area. As shown in the inset of Fig. 6, the pore volume distribution curve is broad in the range of 20 to over 70 nm which further supports the presence of mesopores. From the agglomeration of nanoparticles resulting in the formation of these mesopores it is evident that single-crystals of nanorods have nanoporous channels.²² For photocatalytic applications, this nanoporous channel structure is reported to be useful for efficient and fast transfer of reactant and product molecule fragments.²² The specific surface area, porosity, and pore volume of the as-prepared samples are shown in Table 1. An increase in dopant concentrations results in a decrease in average pore size due to the conquest of the crystal growth mechanism.

Fig. 6 Nitrogen adsorption–desorption isotherm for NST-0 and NST-2, and inset shows the corresponding pore volume distribution curve for NST-2 sample.

The chemical composition and chemical state of the dopant elements (N and S) in the lattice of TiO₂ were identified with the help of XPS analysis. For this purpose, survey spectra of NST-0 and NST-2 samples are shown in Fig. 7(a). It was noted that the survey spectrum of pure TiO₂ has sharp photoelectron peaks for Ti, O and C elements corresponding to the binding energies at 459.2 (Ti 2p), 530.7 (O 1s) and 284.8 eV (C 1s). The nominal atomic composition for anatase TiO₂ was observed for the NST-0 sample from the XPS survey spectrum with an atomic ratio of Ti to O which is about 1 : 2 and is in good agreement. More interestingly, the adventitious carbon peak is reported to arise from the XPS instrument grid. Whereas, for the NST-2 sample, minor XPS photoelectron peaks corresponding to dopant elements (N and S) were present at binding energies of at 401 (N 1s) and 168.7 eV (S 2p), beside the main Ti, O, and C photoelectron peaks.

Fig. 7 (a) XPS survey spectra of NST-0 and NST-2 samples, respectively. (b) shows the high-resolution XPS spectra of C 1s NST-0 and NST-2 samples, respectively, and (c) and (d) show the photoelectron XPS peak in the region of N 1s and S 2p for NSTi-2 samples.

The high-resolution XPS spectra of the C 1s photoelectron peaks for NST-0 and NST-2 samples are shown in Fig. 7(b). It was found that the C 1s photoelectron peaks are similar in shape and appeared at the same binding energies. More importantly, the symmetrical C 1s peak at a binding energy of 284.8 eV is ascribed to the presence of methyl groups (organic compound) and is expected to arise from the XPS instrument itself. In particular, no clue was found for the presence of either anionic or interstitial C 1s peaks corresponding to the binding energies at 282 and 288 eV for C-doped TiO₂ material and thus eliminated the chance of carbon doping into the TiO₂ crystal lattice.⁵¹

The high-resolution XPS spectrum was taken from the surface of the NST-2 sample and is shown in Fig. 7(c) for the deconvolution the N 1s photoelectron region. The absence of a sharp N 1s peak at 400.6 eV is attributed to the fact that the adopted synthesis method is superior to a number of previously reported methods to obtain N,S-co-doped samples without performing the post-annealing treatment. It also supports the fact that nitrogen atoms are not present in the oxidized form of Ti–N–O in the crystal lattice of TiO₂.^52,53 Whereas the presence of a sharp XPS signal at BE of 397.0 eV for the N 1s peak is attributed to the anionic N-doping in the TiO₂ lattice forming O–Ti–N bonds because of the replacement of oxygen atoms by nitrogen atoms in the TiO₂ crystal lattice. Thus, interstitial nitrogen atom-doping did not take place and only anionic nitrogen-doping occurred. It is reported in the literature that preparation routes control the localization of nitrogen atoms in the crystal lattice of TiO₂ for N-doped TiO₂ samples.⁵⁴ Therefore, the form of nitrogen doping in the TiO₂ crystal lattice largely depends on the synthesis route and the source of N-doping reagents.⁵² More interestingly, oxygen-rich conditions favored interstitial N-doping and the absence of oxygen (anaerobic condition) or the presence of a reducing environment favor the substitution of oxygen atoms by nitrogen atoms in the TiO₂ crystal lattice. Thus, the absence of interstitial N-doped TiO₂ is not surprising due to the absence of anaerobic atmosphere in the OMP route.

The high-resolution XPS spectrum of S 2p photoelectrons for the NST-2 sample is shown in Fig. 7(d). It was observed that the XPS spectrum consists of two types of peak for S species in the region of S 2p. The presence of a prominent XPS peak at about 168.7 eV corresponds to the S⁶⁺ state of the S species in the TiO₂ crystal lattice. Whereas the trace of a minor peak at 163.8 eV corresponds to the presence of some sulfur atoms as anionic doping in the state of S²⁻ in the crystal lattice of TiO₂ due to the replacement of oxygen atoms. In detail, the cationic substitution of titanium atoms by sulfur atoms in the lattice is attributed to a rise in the XPS peak at 168.7 eV. On the contrary, the presence of a weak peak at 163.8 eV is attributed to the formation of some Ti–S bond.^55,56 Similar findings were reported by Umebayashi et al.⁵⁷ for anionic sulphur-doping, where titanium disulfide (TiS₂) was used as a starting material. It was reported that during the preparation process most of the sulfur atoms were oxidized from TiS₂ and remained as residual sulfur in the form of S²⁻. More importantly, the large ionic radius of S²⁻ (1.7 Å) as compared to O²⁻ (1.22 Å) make it difficult for the successful substitution of oxygen atoms by a sulfur atom in the lattice due to the requirement of large activation energy for the formation of a Ti–S bond. On the other hand, the lower value of the activation energy for the cationic substitution of Ti⁴⁺ by S⁶⁺ in the lattice makes it more feasible compared to the replacement of oxygen atoms by sulfur atoms.⁵⁵ The XPS findings revealed that the exact amount of dopant elements (N and S) in the NST-2 sample are in accord. Furthermore, the XPS results confirmed that an increase in dopant concentration results in an increase in their atomic percentage in the lattice TiO₂.

The optical properties of the selected samples were studied with the help of normalized absorbance data. Generally speaking, non-metal-doping into the TiO₂ crystal lattice changes the light absorption pattern of TiO₂.^56,58 It was noted that incorporation of dopant atoms into the TiO₂ crystal lattice not only changes the nanostructure morphology and composition, but also changes the bandgap value of the nanorods. Fig. 8 shows the onset of the shift in absorbance band edge for the as-prepared NST samples. The inset of the Fig. 8 shows the bandgap plot using the Kubelka–Munk function from the indirect transition for the as-prepared NST samples.^59–61 In particular, the NST-2 sample shows two prominent features: (1) a new shift in the absorption shoulder to the visible wavelength region (400–500 nm) and (2) the absorbance enhanced in the range of 550–650 nm. The overall absorbance in the visible-light region increases with the increase in dopant concentration and is consistent with the appearance of a yellow color in the NST samples. It was reported that interstitial N-doping into the TiO₂ crystal lattice resulted in a narrowing of the valence band edge by creating some localized states due to the introduction of a visible-light absorption shoulder in the 380 to 500 nm wavelength range.⁵³ The synergistic effect of co-dopant elements (N and S) changed the TiO₂ electronic structure and enhanced the absorbance shoulder in the visible region i.e. 500–750 nm. Whereas nitrogen doping is directly responsible for the absorbance shoulder in the wavelength range of 380–500 nm. The calculated bandgap value for the NST-2 sample was in the range of 2.70 eV (460 nm). The incorporation of dopant atoms results in a shift of the bandgap edge towards the longer wavelength of the visible-light region. The expected transition by impurity levels is related to the shift in bandgap edge as reported earlier for the incorporation of nitrogen and sulfur atoms into the TiO₂ crystal lattice having a wider bandgap.^62,63 These findings further support the successful incorporation of nitrogen and sulfur atoms into the TiO₂ crystal lattice, which changes the crystallite size and electronic structure of the as-prepared NST samples.

Fig. 8 UV-visible diffuse reflectance spectra of selected NST samples with different dopant concentration, and inset shows the bandgap plots and the optical photos for transition of color change with dopant concentration levels.

The photocatalytic activity of the as-prepared NST samples was investigated for the degradation of organic compounds (i.e. rhodamine B (RhB) and phenol) in an aqueous solution beneath visible-light irradiation. In order to ensure the photocatalytic degradation of organic dyes over the as-prepared NST photocatalysts, the change in the organic compound concentration over the NST samples was measured under dark conditions, in the absence of visible-light irradiation. Fig. 3S† shows the change in RhB concentration over the as-prepared NST samples. It was found that the RhB concentration did not change to a significant amount under dark conditions. On the other hand, the RhB concentration decreased dramatically under visible-light irradiation. Therefore, the degradation of RhB molecules was performed over the surface of the NST photocatalysts under visible-light irradiation. The photocatalytic activity of the as-prepared NST samples was calculated by measuring the decrease in the RhB concentration over the different intervals of visible-light irradiation. The following eqn (1) was used to calculate the photodegradation rate constant:

where K_a, t, C_o and C are the rate constant, reaction time, initial and equilibrium concentrations of RhB, respectively. Fig. 9(a) and (b) show the photodegradation rate and the kinetic curves for the photodegradation of RhB over the as-prepared samples under visible-light irradiation. It was observed that the RhB degradation rate constant of NST-2 was higher than those of the other samples. The photocatalytic degradation pathway of RhB is divided into two types – surface reactions and bulk reactions in solution.⁶⁴ In detail, when the surface of the photocatalyst is irradiated by visible light, the photogenerated holes in the valence bands are excited and start to produce hydroxyl radicals by interacting with the surface-adsorbed water/hydroxyl groups, and promote the surface bulk reaction. The nascent hydroxyl radicals started the photooxidation reaction to degrade the RhB molecules in the aqueous solution. In a side reaction, photoinduced electrons participated in the surface reaction to produce hydroxyl radicals and provoked the de-ethylation/destruction of the RhB molecules under visible-light.

Fig. 9 (a) The photocatalytic degradation rate of RhB aqueous solution over as-prepared samples. (b) The photodegradation kinetic curve for NST-0 and NST-2 samples. (c) A comparative study of visible-light-driven photocatalytic activity for the degradation of phenol aqueous solution and (d) the apparent calculated values of rate constants.

The change in the shape of UV-Vis absorbance spectra helps to analyze the dominance of the photodegradation routes – either by surface or bulk reaction. In the case of the bulk solution reaction, the RhB molecules degraded without changing the position/shift of the RhB absorbance peak. Whereas the involvement of the surface reaction changes the position of the UV-Vis peak during the degradation of the RhB aqueous solution. The shift in absorbance spectra for the as-prepared NST samples was more prominent after the visible-light irradiation. The rate of surface reaction for the as-prepared NST-2 sample was much higher than for other samples. In particular, the rate of surface reaction for the as-prepared NST-2 photocatalysts is higher due to the occurrence of rapid surface reactions as compared to other samples, where the surface reactions occurred slowly. The UV-Vis absorbance spectra for RhB aqueous solution over the NST-2 sample started to decrease under visible-light irradiation. The decrease in absorbance intensity is due to the degradation of RhB molecules over the photocatalysts under visible-light irradiation via both fast bulk solution and surface reactions.

To eliminate the chance of a photosensitization process, the photocatalytic activity of the as-prepared samples was investigated for the degradation of a colorless organic compound (i.e. phenol) under visible-light irradiation. It was noted that no change in phenol concentration occurred over NST samples in dark conditions (without visible-light irradiation). Whereas a blank photocatalytic test (in the absence of photocatalysts) showed that no photodegradation of phenol took place without the addition of photocatalysts. Therefore, the presence of both NST photocatalysts and visible-light illumination is required for the successful photodegradation of phenol aqueous solution. It was found that photocatalytic oxidation reactions for the degradation of phenol aqueous solution take place on the surface of the photocatalysts under visible-light irradiation. The change in absorption of phenol aqueous solution over the as-prepared NST-2 photocatalyst under visible-light irradiation is shown in Fig. 4S.† The photocatalytic degradation of phenol and rate constants over different photocatalysts under visible-light irradiation are shown in Fig. 9(c) and (d). It was observed that the molar ratio of titanium to the dopant concentration has a significant effect on the photocatalytic efficiency of NST samples. More interestingly, an obvious photocatalytic activity was noted for the pristine TiO₂ sample under visible-light irradiation. Whereas it is theoretically established that anatase TiO₂ has a large bandgap of 3.2 eV and cannot be excited under visible-light irradiation. Therefore, it is very interesting to figure out the main reason for the photocatalytic activity of pristine TiO₂ towards organic compound degradation under visible-light irradiation. According to the literature, the surprising visible-light photocatalytic activity of TiO₂ is expected to arise from substrate–surface complexation between the photocatalyst (TiO₂) and an organic compound. The creation of substrate–surface complexation is involved in the transfer of charge from ligand-to-titanium atoms and extends the absorbance edge of titania in the visible-light region.^65,66 This extension of the absorbance edge in the visible-region is explained on the basis that the adsorption of organic compounds on the surface of TiO₂ is responsible for absorption in the visible region and TiO₂ itself did not absorb visible-light.^65,66 Hence, during the photocatalytic experiment, the adsorption of an organic compound on the surface of TiO₂ is responsible for shifting the absorption edge in the visible-light region due to an intramolecular transfer of charge from ligand to titanium atoms owing to the formation of a substrate–surface complexation. The photodegradation of organic compounds (i.e. phenol) over a TiO₂ photocatalyst under visible-light irradiation through substrate–surface complex mediation is presented in Fig. 10(a). Serpone et al.⁶⁶ described the surface–substrate complex formation through phenolate linkage during the adsorption of ligand on the surface of TiO₂ in the TiO₂/phenol system as:

≡Ti–OH + OH–Ph → ≡Ti–O–Ph + H₂O (2)

Fig. 10 (a) Schematic illustration for the mediated charge transfer pathway through surface–substrate complex formation and (b) shows mechanism involved for the photocatalytic degradation of organic compounds over NST samples under visible light irradiation.

The substrate–surface complex is excited upon visible-light irradiation through the transfer of charge from ligand to titanium. As a result, the photo-induced electrons started to shift to the TiO₂ conduction band. Afterwards, the adsorbed O₂ on the TiO₂ surface reacts with injected electrons to form superoxide radical anions (˙O²⁻). The superoxide radicals react further by protonation to form ˙OOH radicals which trapped the injected electrons and subsequently form H₂O₂ and ˙OH radicals. It is well-known that ˙OH radicals are the most active species for the degradation of organic compounds.^67,68 On the other hand, the photocatalytic activity of all the NST samples is much higher compared to the pristine sample under visible-light irradiation. The proposed mechanism involved for the higher photocatalytic activity of NST samples as compared to the pristine sample is shown in Fig. 10(b). Furthermore, the photocatalytic activity of NST samples increases with the increase in dopant concentrations and reaches a maximum value for the NST-2 sample. The photocatalytic efficiency of NST-2 is higher than for the other samples.

The formation of intermediate by-products during the degradation of organic pollutants was monitored with the help of HPLC and LC-MS analysis. The HPLC analysis was performed during the photocatalytic degradation experiment of RhB aqueous solution over NST-2 photocatalysts for the different intervals of visible-light irradiation. Fig. 11(a) shows the matrix peaks at retention times of t_R1/4 = 8.5 min of RhB at 554 nm. The evidence for the photodegradation of RhB was obvious from the diminishing of the matrix peak with photocatalytic experiment time. The complete photodegradation of RhB into less-toxic fragments was achieved at the end of the photocatalytic experiment. In particular, no peak corresponding to the presence of any by-product was found – except for the RhB. The LC-MS analysis was performed to further confirm the preliminary degradation process of RhB over NST photocatalysts under visible-light irradiation. Fig. 5S† shows the LC-MS spectra for the pattern formation of different fragments/intermediates during the photocatalytic degradation of RhB under visible-light irradiation. The molecular ion peak for RhB started to decrease and diminished completely in the LC-MS analysis after the completion of the photocatalytic experiment. In detail, the intensity of the higher m/z value peaks started to decrease due to the degradation of RhB into non-toxic levels of inorganic products (i.e. CO₂, H₂O, etc.). Thus, this confirmed the complete degradation of RhB under visible-light irradiation.

Fig. 11 (a) Time dependent HPLC chromatogram for photodegradation of RHB over NST-2 sample at λ_max = 554 nm. (b) Effects of different scavengers addition on the photodegradation efficiency of the RhB over NST-2 photocatalysts.

Excellent reproducibility was observed for the NST-2 sample over 4 consecutive cycles of the photocatalytic experiments. Fig. 6S† shows the 4 cycles of photocatalytic degradation of RhB over the NST-2 sample under visible-light irradiation. The surface bulk solution reactions for the as-prepared samples were faster than for the reference samples. The high visible-light absorbance for the as-prepared photocatalysts was expected to bring about the rapid bulk solution reactions. The rate of surface reaction depends on the concentration of the doping elements in the as-prepared samples. It has been reported that, other than the doping element concentration in the TiO₂ crystal structure, the variation in the crystal structure, degree of crystallinity, specific surface area and the chemical states of the dopants on the TiO₂ surface have a considerable effect on the rate of surface reactions.⁶⁹

Fig. 7S(a)† shows the rate constant of ˙OH radical formation as a function of visible-light irradiation time. It was found that the rate constant for ˙OH radical formation follows the same trend as that observed for the photodegradation of organic pollutants. This shows that an indirect mechanism played a key role in the photocatalytic oxidation reactions. The detection of ˙OH was performed to confirm the photocatalytic activity of as-prepared samples under visible-light irradiation. Fig. 7S(b)† shows the change in PL spectra for the terephthalic acid solution as a function of visible-light irradiation time. It was noted that the PL intensity of terephthalic acid solution increases gradually with time. Whereas no change in PL intensity was noted in the absence of either NST photocatalysts or visible-light irradiation. Therefore, chemical reactions between the generated ˙OH radicals and the terephthalic acid are responsible for the rise in fluorescence intensity. The change in PL intensities is linear over the as-prepared samples for a given interval of time. The concentration of ˙OH radical formation is directly proportional to the visible-light irradiation time. The photogenerated holes possessed enough oxidation power to produce ˙OH radicals by interacting with the surface-adsorbed hydroxyl ions/water molecules. The photogenerated electrons in the conduction bands participated in the redox reactions to produce ˙OH radicals on the surface of NST photocatalysts.^70,71 It was observed that the PL intensity for the NST-2 sample was higher than for other samples i.e. higher amounts of ˙OH radical were produced over the NST-2 sample under visible-light irradiation. Therefore, higher photocatalytic activity was observed for the NST-2 sample. For pristine TiO₂, the intensity of the PL peak is very low, as expected due to the fact that TiO₂ is not activated under visible-light irradiation. As evidence, the concentration of generated ˙OH on the surface of photocatalysts played an active role in the photocatalytic efficiency of the as-prepared samples. Therefore, the more hydroxyl radicals produced, the higher the photocatalytic activity.

A detailed study was performed to characterize the mechanism involved in the photocatalytic process. The involvement of active species was studied in the photocatalytic degradation of the organic pollutant (RhB) over the NST-2 sample under visible-light irradiation – by adding different chemical scavengers into the reaction systems. For this purpose, AgNO₃ (a CB electron acceptor plus a strong oxidant), DMSO (an ˙OH scavenger) and sodium oxalate (a photogenerated holes (h⁺) near-VB scavenger) were added into the photocatalysis system.^38,72,73 Furthermore, the dissolved oxygen is responsible for inhibiting the recombination of charge carriers due to the formation of superoxide radicals (˙O²⁻) – which played an important role in some of the photocatalytic processes.⁷⁴ As a rational approach, if the photocatalytic process is performed by superoxide radicals – as an active species, the photocatalytic reaction rate would be reduced substantially with the addition of Ag⁺ into the reaction system.^72,73 On the contrary, the reduction of reaction rate with the addition of either DMSO or SO showed that ˙OH radicals and/or holes played active roles in the photodegradation process.^72,73,75 Fig. 11(b) shows that the addition of sodium oxalate (0.5 mmol L⁻¹) reduced the photodegradation efficiency of RhB from 78 to 48% as compared to the experiment in the absence of scavengers. This shows that h⁺ played an active role in the photodegradation of RhB aqueous solution under visible-light irradiation. Similarly, an excess of DMSO (5 mmol L⁻¹) was added into the reaction system to monitor the effect of ˙OH formation on the photodegradation processes. The excess DMSO captured all the ˙OH during the photodegradation reaction of RhB aqueous solution.⁷⁵ It was found that the photodegradation efficiency of RhB reduced greatly from 78 to 42% after 2 hours of visible-light irradiation. Thus, ˙OH played a more important part in the photodegradation of RhB as compared to the photogenerated holes. Overall, photogenerated h⁺ and ˙OH played an important role in the photodegradation of RhB aqueous solution over the NST photocatalysts under visible-light irradiation. These findings further support the results obtained for the formation of ˙OH from the terephthalic acid experiments. On the contrary, the addition of excess Ag⁺ (1 mmol L⁻¹) did not change the photodegradation efficiency of RhB to an obvious degree. This shows that O²⁻˙ played a minor part in the photocatalytic degradation of RhB aqueous solution under visible-light irradiation.

The synergetic effects of enhanced photocatalytic activity for NST samples are ascribed to the presence of red-shift of intense absorption edges in the visible-light region and the presence of highly reactive preferential {001} facets. As discussed earlier, the introduction of dopant elements into the TiO₂ lattice shift the absorption edge in the visible-light region. Thus, the NST samples were activated under visible-light irradiation and generated photo-induced charge carriers (i.e. electrons and holes pairs) to participate in the photocatalytic oxidation reactions. Furthermore, the presence of preferential highly reactive exposed {001} facets as compared to thermodynamically stable {101} facets dominates the active site on the surface of the photocatalysts for photocatalytic oxidation reactions.^76,77 As a result, the 1-D nanorods with preferential {001} facets showed enhanced visible-light photocatalytic activity due to more feasible adsorption of organic pollutants on the surface of the photocatalysts for degradation.⁷⁸ Moreover, most of the time anatase TiO₂ nanocrystals are prepared with preferential {101} facets and showed lower photocatalytic activity. Therefore, the higher photocatalytic activity of as-prepared NST nanorods was obtained with preferential exposed {001} facets under visible-light irradiation. More interestingly, the NST-2 nanoparticles with less dominant {001} facets prepared in another experiment showed lower photocatalytic activity under visible-light irradiation as compared to NST-2 nanorods having preferential exposed {001} facets. Thus, we can conclude from these findings that the morphology of TiO₂ played an important role in the photocatalytic performance of TiO₂ nanorods having preferential {001} facets under visible-light irradiation as compared to corresponding nanoparticles. In the end, a further increase in dopant element concentration results in a gradual decrease in photocatalytic activity. This gradual decrease in photocatalytic activity is expected due to the deterioration in the degree of crystallinity and creation of more defects sites which act as recombination centers in the NST samples (see Table 1).

Another explanation for the enhanced photocatalytic activity of the as-prepared samples is based on the presence of dopant elements on the surface of NST photocatalysts – which efficiently enhanced the rate of surface reaction due to the enhanced interaction/adsorption of the organic dyes on the surface of the NST photocatalysts. The strong interaction between the dopant elements and the organic compounds promotes the adsorption and rate of the surface reaction of NST photocatalysts under visible-light irradiation. The photoinduced holes from the conduction band – i.e. those excited to the surface of the photocatalysts by visible-light irradiation – are adsorbed by the surface adsorbed oxygen molecules to produce hydroxyl radicals. The nascent hydroxyl radicals facilitate the surface reactions for the degradation of organic dyes on the surface of NST photocatalysts. The presence of delocalized states in the bandgap of the NST samples promotes the rate of bulk solution reaction by shifting the absorbance edge of the NST sample in the visible-light region. Thus, the as-prepared NST photocatalysts showed enhanced photocatalytic activity for the degradation of organic compounds under visible-light irradiation. The enhancement of dopant concentration after an ambient concentration reduced the photocatalytic efficiency due to the lower rate of surface and bulk solution reactions – as observed for the NST-3 sample.

Conclusion

A facile, low-temperature, surfactant- and template-free OPM route was developed to prepare nitrogen and sulfur co-doped TiO₂ nanorods with preferential {001} facets. The structural characterization of NST samples by ICP-AES, EDX, XPS and HR-TEM analysis confirmed that dopant elements (N and S) have been incorporated into the lattice of TiO₂ via cationic S-doping by replacing titanium atoms and interstitial N-doping by substituting oxygen atoms. The newly introduced sulfur and nitrogen atoms into the TiO₂ lattice formed new energy levels between the bandgap of TiO₂ that shifted the absorption edge shoulder in the visible-region. The electronic excitation from the valence band to the new-build impurity energy states facilitates the further transfer of electrons to the conduction bands by secondary excitation under visible-light irradiation and enhanced their availability on the surface of the photocatalysts. Another benefit of newly built energy states is that absorption of low-energy photons excites the electrons to these states and secondary excitation promotes them to the bottom level of CB for effective photo-response in the visible-light region. The formation of substrate–surface complexation results in the photocatalytic response of the pristine TiO₂ sample for the degradation of organic compounds under visible-light irradiation. This strange behaviour is expected to be due to the facile transfer of charge carriers through the ligand–titanium coordination that shifts the absorption edge in the visible region for the pristine TiO₂ sample. The as-prepared NST samples showed enhanced photocatalytic activity as compared to the pristine sample for the degradation of organic compounds under visible-light irradiation. This enhancement in photocatalytic activity is higher for the NST-2 sample as compared to other samples. The synergetic effect of morphology design is responsible for forming preferential {001} facets of nanorods. The introduction of dopant elements (N and S) produced the red shift in the absorption shoulder edge towards the visible-light region. Therefore, more photoinduced electron–hole pairs participated in the degradation of organic compounds under visible-light irradiation. The synergetic effect of having 1-D nanorods with highly reactive preferential {001} facets enhanced the adsorption of pollutant molecules on the surface of photocatalysts and results in enhanced visible-light photocatalytic efficiency. Thus, this report represents a more detailed investigation of the co-doping mechanism into the TiO₂ lattice – through insight into the chemical and physical enhancement mechanism – to explain the enhanced photocatalytic activity under visible-light irradiation. Thus, as-prepared N,S-co-doped TiO₂ nanorods are of great importance for practical applications, including optoelectronic devices, solar cells, sensors, photocatalytic H₂-production, and catalysis.

Acknowledgements

The authors would like to give special thanks to the TWAS-CNPq post-graduate program for financial support to the project. The authors would also like to thank Brazilian Nanotechnology National Laboratory LNNano for providing the equipment and technical support for doing XPS analyzes and Liec lab Ufscar for TEM analysis.

References

1. X. Shen, C. Xu and L. Ye, Molecularly Imprinted Polymers for Clean Water: Analysis and Purification, Ind. Eng. Chem. Res., 2013, 52, 13890–13899,  DOI:10.1021/ie302623s.
2. S. A. Bakar and C. Ribeiro, Nitrogen-doped Titanium Dioxide: an overview of material design and dimensionality effect over modern applications, J. Photochem. Photobiol., C, 2016, 27, 1–29.
3. L. Bao, M. Meng, K. Sun, W. Li, D. Zhao and H. Li, et al., Selective adsorption and degradation of rhodamine B with modified titanium dioxide photocatalyst, J. Appl. Polym. Sci., 2014, 131, 40890.
4. S. A. Bakar and C. Ribeiro, A comparative run for visible-light-driven photocatalytic activity of anionic and cationic S-doped TiO₂ photocatalysts: A case study of possible sulfur doping through chemical protocol, J. Mol. Catal. A: Chem., 2016, 421, 1–15,  DOI:10.1016/j.molcata.2016.05.003.
5. N. Nishiyama, Y. Fujiwara, K. Adachi, K. Inumaru and S. Yamazaki, Preparation of porous metal-ion-doped titanium dioxide and the photocatalytic degradation of 4-chlorophenol under visible light irradiation, Appl. Catal., B, 2015, 176–177, 347–353,  DOI:10.1016/j.apcatb.2015.04.015.
6. J. Wen, X. Li, W. Liu, Y. Fang, J. Xie and Y. Xu, Photocatalysis fundamentals and surface modification of TiO₂ nanomaterials, Chin. J. Catal., 2015, 36, 2049–2070,  DOI:10.1016/S1872-2067(15)60999-8.
7. R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, Visible-light photocatalysis in nitrogen-doped titanium oxides, Science, 2001, 293, 269–271,  DOI:10.1126/science.1061051.
8. Z. He, Y. Yu, D. Wang, J. Tang, J. Chen and S. Song, Photocatalytic reduction of carbon dioxide using iodine-doped titanium dioxide with high exposed {001} facets under visible light, RSC Adv., 2016, 6, 23134–23140,  10.1039/c5ra26761g.
9. J.-W. Shi, C. Liu, C. He, J. Li, C. Xie and S. Yang, et al., Carbon-doped titania nanoplates with exposed {001} facets: facile synthesis, characterization and visible-light photocatalytic performance, RSC Adv., 2015, 5, 17667–17675,  10.1039/c4ra15824e.
10. C. Xue, T. Wang, G. Yang, B. Yang and S. Ding, A facile strategy for the synthesis of hierarchical TiO₂/CdS hollow sphere heterostructures with excellent visible light activity, J. Mater. Chem. A, 2014, 2, 7674,  10.1039/c4ta01190b.
11. S. Abu Bakar, S. Tajammul Hussain and M. Mazhar, CdTiO₃ thin films from an octa-nuclear bimetallic single source precursor by aerosol assisted chemical vapor deposition (AACVD), New J. Chem., 2012, 36, 1844,  10.1039/c2nj40275k.
12. S. T. Hussain, S. A. Bakar, B. Saima and B. Muhammad, Low temperature deposition of silver sulfide thin films by AACVD for gas sensor application, Appl. Surf. Sci., 2012, 258, 9610–9616,  DOI:10.1016/j.apsusc.2012.05.157.
13. S. A. Bakar, S. T. Hussain, S. A. Trimizi and M. N. Tahir, Crack free and spherical shape copper–zinc composite oxide thin films fabrications by AACVD, Polyhedron, 2013, 49, 138–144,  DOI:10.1016/j.poly.2012.09.040.
14. M. Pastore and F. De Angelis, First-Principles Modeling of a Dye-Sensitized TiO₂/IrO₂ Photoanode for Water Oxidation, J. Am. Chem. Soc., 2015, 137, 5798–5809,  DOI:10.1021/jacs.5b02128.
15. E. M. Samsudin, S. B. A. Hamid, J. C. Juan and W. J. Basirun, Influence of triblock copolymer (pluronic F127) on enhancing the physico-chemical properties and photocatalytic response of mesoporous TiO₂, Appl. Surf. Sci., 2015, 355, 959–968,  DOI:10.1016/j.apsusc.2015.07.178.
16. P. Zhou, J. Wu, W. Yu, G. Zhao, G. Fang and S. Cao, Vectorial doping-promoting charge transfer in anatase TiO₂ {001} surface, Appl. Surf. Sci., 2014, 319, 167–172,  DOI:10.1016/j.apsusc.2014.05.045.
17. T. Wang, X. Yan, S. Zhao, B. Lin, C. Xue and G. Yang, et al., A facile one-step synthesis of three-dimensionally ordered macroporous N-doped TiO₂ with ethanediamine as the nitrogen source, J. Mater. Chem. A, 2014, 2, 15611,  10.1039/c4ta01922a.
18. T. He, X. Guo, K. Zhang, Y. Feng and X. Wang, Synthesis and characterization of B–N co-doped mesoporous TiO₂ with enhanced photocatalytic activity, RSC Adv., 2014, 4, 5880,  10.1039/c3ra44046j.
19. J. Cheng, J. Chen, W. Lin, Y. Liu and Y. Kong, Improved visible light photocatalytic activity of fluorine and nitrogen co-doped TiO₂ with tunable nanoparticle size, Appl. Surf. Sci., 2015, 332, 573–580,  DOI:10.1016/j.apsusc.2015.01.218.
20. M. S. Akple, J. Low, Z. Qin, S. Wageh, A. A. Al-Ghamdi and J. Yu, et al., Nitrogen-doped TiO₂ microsheets with enhanced visible light photocatalytic activity for CO₂ reduction, Chin. J. Catal., 2015, 36, 2127–2134,  DOI:10.1016/S1872-2067(15)60989-5.
21. S. A. Bakar and C. Ribeiro, Low temperature synthesis of N-doped TiO₂ with rice-like morphology through peroxo assisted hydrothermal route: Materials characterization and photocatalytic properties, Appl. Surf. Sci., 2016, 377, 121–133,  DOI:10.1016/j.apsusc.2016.03.137.
22. S. A. Bakar, G. Byzynski and C. Ribeiro, Synergistic effect on the photocatalytic activity of N-doped TiO₂ nanorods synthesised by novel route with exposed (110) facet, J. Alloys Compd., 2016, 666, 38–49,  DOI:10.1016/j.jallcom.2016.01.112.
23. S. T. Hussain, S. A. Bakar, M. Mazhar and M. Zeller, Chloridotris[μ2-2-(dimethyl-amino)-ethano-lato]-μ3-hydroxido-tri-μ2-trifluoro-acetato-tetra-copper(II) tetra-hydro-furan solvate, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2010, 66, m814.
24. T. Ohno, M. Akiyoshi, T. Umebayashi, K. Asai, T. Mitsui and M. Matsumura, Preparation of S-doped TiO₂ photocatalysts and their photocatalytic activities under visible light, Appl. Catal., A, 2004, 265, 115–121,  DOI:10.1016/j.apcata.2004.01.007.
25. J. Wang, H. Li, H. Li and C. Zou, Mesoporous TiO_2−xA_y (A = N, S) as a visible-light-response photocatalyst, Solid State Sci., 2010, 12, 490–497,  DOI:10.1016/j.solidstatesciences.2009.12.013.
26. J. A. Rengifo-Herrera, K. Pierzchała, A. Sienkiewicz, L. Forró, J. Kiwi and J. E. Moser, et al., Synthesis, Characterization, and Photocatalytic Activities of Nanoparticulate N, S-Codoped TiO₂ Having Different Surface-to-Volume Ratios, J. Phys. Chem. C, 2010, 114, 2717–2723,  DOI:10.1021/jp910486f.
27. S. Abu Bakar and C. Ribeiro, An insight toward the photocatalytic activity of S doped 1-D TiO₂ nanorods prepared via novel route: As promising platform for environmental leap, J. Mol. Catal. A: Chem., 2016, 412, 78–92,  DOI:10.1016/j.molcata.2015.12.002.
28. S. Abu Bakar and C. Ribeiro, Rapid and morphology controlled synthesis of anionic S-doped TiO₂ photocatalysts for visible-light-driven photodegradation of organic pollutants, RSC Adv., 2016, 6, 36516–36527.
29. Q. Xiang, J. Yu and M. Jaroniec, Nitrogen and sulfur co-doped TiO₂ nanosheets with exposed {001} facets: synthesis, characterization and visible-light photocatalytic activity, Phys. Chem. Chem. Phys., 2011, 13, 4853–4861,  10.1039/c0cp01459a.
30. O. Ola and M. M. Maroto-Valer, Review of material design and reactor engineering on TiO₂ photocatalysis for CO₂ reduction, J. Photochem. Photobiol., C, 2015, 24, 16–42,  DOI:10.1016/j.jphotochemrev.2015.06.001.
31. T. S. Natarajan, H. C. Bajaj and R. J. Tayade, Preferential adsorption behavior of methylene blue dye onto surface hydroxyl group enriched TiO₂ nanotube and its photocatalytic regeneration, J. Colloid Interface Sci., 2014, 433, 104–114,  DOI:10.1016/j.jcis.2014.07.019.
32. R. Chalasani and S. Vasudevan, Cyclodextrin-Functionalized Fe₃O₄@TiO₂: Reusable, Magnetic Nanoparticles for Photocatalytic Degradation of Endocrine-Disrupting Chemicals in Water Supplies, ACS Nano, 2013, 7, 4093–4104,  DOI:10.1021/nn400287k.
33. S. Miyayama, K. Nishijima, T. Kamai, T. Chiyoya, T. Tsubota and T. Ohno, Photocatalytic selective oxidation of anionic compounds on TiO₂ photocatalysts modified with quaternary ammonium base groups, Sep. Purif. Technol., 2007, 58, 206–210,  DOI:10.1016/j.seppur.2007.07.030.
34. V. R. de Mendonça and C. Ribeiro, Influence of TiO₂ morphological parameters in dye photodegradation: A comparative study in peroxo-based synthesis, Appl. Catal., B, 2011, 105, 298–305,  DOI:10.1016/j.apcatb.2011.04.018.
35. B. D. Cullity and S. R. Stock, Elements of X-Ray Diffraction, Prentice Hall, 3rd edn, 2001.
36. K. Suzuki and K. Kijima, Optical Band Gap of Barium Titanate Nanoparticles Prepared by RF-plasma Chemical Vapor Deposition, Jpn. J. Appl. Phys., 2005, 44, 2081–2082,  DOI:10.1143/JJAP.44.2081.
37. O. F. Lopes, K. T. G. Carvalho, G. K. Macedo, V. R. de Mendonça, W. Avansi and C. Ribeiro, Synthesis of BiVO₄ via oxidant peroxo-method: insights into the photocatalytic performance and degradation mechanism of pollutants, New J. Chem., 2015, 39, 6231–6237,  10.1039/c5nj00984g.
38. Y. Li, J. Wang, H. Yao, L. Dang and Z. Li, Efficient decomposition of organic compounds and reaction mechanism with BiOI photocatalyst under visible light irradiation, J. Mol. Catal. A: Chem., 2011, 334, 116–122,  DOI:10.1016/j.molcata.2010.11.005.
39. K. Ishibashi, A. Fujishima, T. Watanabe and K. Hashimoto, Quantum yields of active oxidative species formed on TiO₂ photocatalyst, J. Photochem. Photobiol., A, 2000, 134, 139–142,  DOI:10.1016/S1010-6030(00)00264-1.
40. K. Ishibashi, A. Fujishima, T. Watanabe and K. Hashimoto, Detection of active oxidative species in TiO₂ photocatalysis using the fluorescence technique, Electrochem. Commun., 2000, 2, 207–210,  DOI:10.1016/S1388-2481(00)00006-0.
41. Q. Xiang, J. Yu, B. Cheng and H. C. Ong, Microwave-hydrothermal preparation and visible-light photoactivity of plasmonic photocatalyst Ag–TiO₂ nanocomposite hollow spheres, Chem.–Asian J., 2010, 5, 1466–1474,  DOI:10.1002/asia.200900695.
42. J. Yu, W. Wang, B. Cheng and B.-L. Su, Enhancement of Photocatalytic Activity of Mesoporous TiO₂ Powders by Hydrothermal Surface Fluorination Treatment, J. Phys. Chem. C, 2009, 113, 6743–6750,  DOI:10.1021/jp900136q.
43. U. Diebold, The surface science of titanium dioxide, Surf. Sci. Rep., 2003, 48, 53–229,  DOI:10.1016/S0167-5729(02)00100-0.
44. M. Lazzeri, A. Vittadini and A. Selloni, Structure and energetics of stoichiometric TiO₂ anatase surfaces, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 155409.
45. E. V. Sukhorukov, G. Burkard and D. Loss, Noise of a quantum dot system in the cotunneling regime, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 125315,  DOI:10.1103/PhysRevB.63.125315.
46. Y. Dai, C. M. Cobley, J. Zeng, Y. Sun and Y. Xia, Synthesis of anatase TiO₂ nanocrystals with exposed {001} facets, Nano Lett., 2009, 9, 2455–2459,  DOI:10.1021/nl901181n.
47. M. Alfredsson, F. Corà, D. P. Dobson, J. Davy, J. P. Brodholt and S. C. Parker, et al., Dopant control over the crystal morphology of ceramic materials, Surf. Sci., 2007, 601, 4793–4800,  DOI:10.1016/j.susc.2007.07.025.
48. M. Brandbyge, J.-L. Mozos, P. Ordejón, J. Taylor and K. Stokbro, Density-functional method for nonequilibrium electron transport, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 65, 165401,  DOI:10.1103/PhysRevB.65.165401.
49. X. Han, Q. Kuang, M. Jin, Z. Xie and L. Zheng, Synthesis of titania nanosheets with a high percentage of exposed (001) facets and related photocatalytic properties, J. Am. Chem. Soc., 2009, 131, 3152–3153,  DOI:10.1021/ja8092373.
50. D. G. Stroppa, L. F. Zagonel, L. A. Montoro, E. R. Leite and A. J. Ramirez, High-Resolution Scanning Transmission Electron Microscopy (HRSTEM) Techniques: High-Resolution Imaging and Spectroscopy Side by Side, ChemPhysChem, 2012, 13, 437–443,  DOI:10.1002/cphc.201100729.
51. E. M. Neville, M. J. Mattle, D. Loughrey, B. Rajesh, M. Rahman and J. M. D. MacElroy, et al., Carbon-Doped TiO₂ and Carbon, Tungsten-Codoped TiO₂ through Sol–Gel Processes in the Presence of Melamine Borate: Reflections through Photocatalysis, J. Phys. Chem. C, 2012, 116, 16511–16521,  DOI:10.1021/jp303645p.
52. J. A. Rengifo-Herrera, E. Mielczarski, J. Mielczarski, N. C. Castillo, J. Kiwi and C. Pulgarin, Escherichia coli inactivation by N, S co-doped commercial TiO₂ powders under UV and visible light, Appl. Catal., B, 2008, 84, 448–456,  DOI:10.1016/j.apcatb.2008.04.030.
53. X. Chen and C. Burda, The Electronic Origin of the Visible-Light Absorption Properties of C-, N- and S-Doped TiO₂ Nanomaterials, J. Am. Chem. Soc., 2008, 130, 5018–5019,  DOI:10.1021/ja711023z.
54. Z. Zhai, C. Hu, X. Yang, L. Zhang, C. Liu and Y. Fan, et al., Nitrogen-doped mesoporous nanohybrids of TiO₂ nanoparticles and HTiNbO₅ nanosheets with a high visible-light photocatalytic activity and a good biocompatibility, J. Mater. Chem., 2012, 22, 19122,  10.1039/c2jm32338a.
55. J. C. Yu, W. Ho, J. Yu, H. Yip, P. K. Wong and J. Zhao, Efficient Visible-Light-Induced Photocatalytic Disinfection on Sulfur-Doped Nanocrystalline Titania, Environ. Sci. Technol., 2005, 39, 1175–1179,  DOI:10.1021/es035374h.
56. M. Zhou and J. Yu, Preparation and enhanced daylight-induced photocatalytic activity of C,N,S-tridoped titanium dioxide powders, J. Hazard. Mater., 2008, 152, 1229–1236,  DOI:10.1016/j.jhazmat.2007.07.113.
57. T. Umebayashi, T. Yamaki, S. Tanaka and K. Asai, Visible Light-Induced Degradation of Methylene Blue on S-doped TiO₂, Chem. Lett., 2003, 32, 330–331,  DOI:10.1246/cl.2003.330.
58. J. Yu, M. Zhou, B. Cheng and X. Zhao, Preparation, characterization and photocatalytic activity of in situ N,S-codoped TiO₂ powders, J. Mol. Catal. A: Chem., 2006, 246, 176–184,  DOI:10.1016/j.molcata.2005.10.034.
59. P. Periyat, N. Leyland, D. E. McCormack, J. Colreavy, D. Corr and S. C. Pillai, Rapid microwave synthesis of mesoporous TiO₂ for electrochromic displays, J. Mater. Chem., 2010, 20, 3650,  10.1039/b924341k.
60. Z. Zheng, B. Huang, X. Qin, X. Zhang and Y. Dai, Strategic synthesis of hierarchical TiO₂ microspheres with enhanced photocatalytic activity, Chemistry, 2010, 16, 11266–11270,  DOI:10.1002/chem.201001280.
61. H.-C. Lin, Using normal approximation on testing and determining sample size for c pk, J. Chin. Inst. Ind. Eng., 2006, 23, 1–11,  DOI:10.1080/10170660609508991.
62. H. Li, J. Wang, H. Li, S. Yin and T. Sato, Photocatalytic activity of (sulfur, nitrogen)-codoped mesoporous TiO₂ thin films, Res. Chem. Intermed., 2010, 36, 27–37,  DOI:10.1007/s11164-010-0111-z.
63. P. Periyat, D. E. McCormack, S. J. Hinder and S. C. Pillai, One-Pot Synthesis of Anionic (Nitrogen) and Cationic (Sulfur) Codoped High-Temperature Stable, Visible Light Active, Anatase Photocatalysts, J. Phys. Chem. C, 2009, 113, 3246–3253,  DOI:10.1021/jp808444y.
64. T. Wu, G. Liu, J. Zhao, H. Hidaka and N. Serpone, Photoassisted Degradation of Dye Pollutants. V. Self-Photosensitized Oxidative Transformation of Rhodamine B under Visible Light Irradiation in Aqueous TiO₂ Dispersions, J. Phys. Chem. B, 1998, 102, 5845–5851,  DOI:10.1021/jp980922c.
65. S. Kim and W. Choi, Visible-Light-Induced Photocatalytic Degradation of 4-Chlorophenol and Phenolic Compounds in Aqueous Suspension of Pure Titania: Demonstrating the Existence of a Surface-Complex-Mediated Path, J. Phys. Chem. B, 2005, 109, 5143–5149,  DOI:10.1021/jp045806q.
66. N. Serpone and E. Pelizzetti, Homogeneous and Heterogeneous Photocatalysis, Springer, Dordrecht, Netherlands, 1986,  DOI:10.1007/978-94-009-4642-2.
67. Y. A. Shaban, M. A. El Sayed, A. A. El Maradny, R. K. Al Farawati and M. I. Al Zobidi, Photocatalytic degradation of phenol in natural seawater using visible light active carbon modified (CM)-n-TiO₂ nanoparticles under UV light and natural sunlight illuminations, Chemosphere, 2013, 91, 307–313,  DOI:10.1016/j.chemosphere.2012.11.035.
68. R. Su, R. Tiruvalam, Q. He, N. Dimitratos, L. Kesavan and C. Hammond, et al., Promotion of Phenol Photodecomposition over TiO₂ Using Au, Pd, and Au–Pd Nanoparticles, ACS Nano, 2012, 6, 6284–6292,  DOI:10.1021/nn301718v.
69. M. J. Muñoz-Batista, A. Kubacka and M. Fernández-García, Effective Enhancement of TiO₂ Photocatalysis by Synergistic Interaction of Surface Species: From Promoters to Co-catalysts, ACS Catal., 2014, 4, 4277–4288,  DOI:10.1021/cs501408u.
70. X. Yu, J. Yu, B. Cheng and M. Jaroniec, Synthesis of hierarchical flower-like AlOOH and TiO₂/AlOOH superstructures and their enhanced photocatalytic properties, J. Phys. Chem. C, 2009, 113, 17527–17535,  DOI:10.1021/jp906992r.
71. J. Yu and L. Shi, One-pot hydrothermal synthesis and enhanced photocatalytic activity of trifluoroacetic acid modified TiO₂ hollow microspheres, J. Mol. Catal. A: Chem., 2010, 326, 8–14,  DOI:10.1016/j.molcata.2010.04.016.
72. M. Yin, Z. Li, J. Kou and Z. Zou, Mechanism Investigation of Visible Light-Induced Degradation in a Heterogeneous TiO₂/Eosin Y/Rhodamine B System, Environ. Sci. Technol., 2009, 43, 8361–8366,  DOI:10.1021/es902011h.
73. J. Cao, B. Xu, B. Luo, H. Lin and S. Chen, Novel BiOI/BiOBr heterojunction photocatalysts with enhanced visible light photocatalytic properties, Catal. Commun., 2011, 13, 63–68,  DOI:10.1016/j.catcom.2011.06.019.
74. Y. Zhao, C. Eley, J. Hu, J. S. Foord, L. Ye and H. He, et al., Shape-Dependent Acidity and Photocatalytic Activity of Nb₂O₅ Nanocrystals with an Active TT (001) Surface, Angew. Chem., Int. Ed., 2012, 51, 3846–3849,  DOI:10.1002/anie.201108580.
75. W. Y. Teoh, J. A. Scott and R. Amal, Progress in Heterogeneous Photocatalysis: From Classical Radical Chemistry to Engineering Nanomaterials and Solar Reactors, J. Phys. Chem. Lett., 2012, 3, 629–639,  DOI:10.1021/jz3000646.
76. H. G. Yang, C. H. Sun, S. Z. Qiao, J. Zou, G. Liu and S. C. Smith, et al., Anatase TiO₂ single crystals with a large percentage of reactive facets, Nature, 2008, 453, 638–641,  DOI:10.1038/nature06964.
77. J. Yu, Q. Xiang, J. Ran and S. Mann, One-step hydrothermal fabrication and photocatalytic activity of surface-fluorinated TiO₂ hollow microspheres and tabular anatase single micro-crystals with high-energy facets, CrystEngComm, 2010, 12, 872,  10.1039/b914385h.
78. J. Yu, J. Fan and K. Lv, Anatase TiO₂ nanosheets with exposed (001) facets: improved photoelectric conversion efficiency in dye-sensitized solar cells, Nanoscale, 2010, 2, 2144,  10.1039/c0nr00427h.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16881g

---