Introduction

Titanium dioxide (TiO2) has been extensively investigated due to its chemical stability, low cost, catalytic properties, photocatalytic properties, environmental clean-up by organic compound mineralization1 and because it is a renewable energy generator used for efficient hydrogen production2,3,4.

TiO2 has four crystalline polymorphs, anatase, rutile, brookite and TiO2 (B) (biphasic based on anatase). Anatase and anatase biphasic TiO2 have been extensively investigated in photocatalysis5 under various forms and morphologies: microspheres, nanoflows, nanotrees and nanobelts6,7,8.

One of the objectives of photocatalysis is the efficient use and conversion of solar light, this is done by doping classical photocatalysts with noble metals (Au, Ag, Pt, etc.)7,9,10 or other oxides (RuO2)11,12. Also, understanding the principles of crystal growth is a major challenge for many researchers, surface science is an intensely studied field. In the last decade, the science of anatastic surface has attracted attention due to the special properties of the surface, the differences in reactivity and surface energy13. Yang to al.13 were the promoters of the TiO2 anatase crystal synthesis with a high percentage (47% by weight) of facets (001), using hydrogen fluoride. Nanobelts structures are considered 1D nanostructures. The dimensional study plays an important role in determining material properties and is a huge challenge for researchers. 1D nanostructures have been extensively investigated due to their distinctive properties related to 0D and 2D materials. 1D nanostructures show two important properties: fast electronic transport and effective load transfer14. Synthesis of 1D nanoparticles depends on several factors: chemical method, pH and temperature, while the catalytic activity depends mainly on the properties of the surface8. The increased interest in the controlled synthesis of the exposed face exposed TiO2 particles (001) is due to a high photocatalytic activity compared to (101) faces15. The study research of Zhao et al. have shown that surface energy of rutile TiO2 facets is γ (001) > γ (100) > γ (101) (0,90 J m-2 > 0,53 J m-2 > 0,44 J m-2)14. The chemical dissociation of the water molecules is energetically favored on the plane (001), and the hydroxyl radicals on the TiO2 surface react much more easily with the dissociated molecules of the organic compounds14.

In this paper, it was developed a facile sol-gel method without templates for preparing rutile.

RuxTi1-xO2 nanobelts crystals with controlled morphologies. The nanobelts have a layer structure which is beneficial for the introduction of heteroatoms, as an example Ru, because RuO2 adopts the rutile structure. Introduction of the metal ions into the crystalline structure of TiO2 nanobelts, expands the absorption edge to the visible light range to enhance, therefore we propose the introduction of ruthenium into the crystal for the extend of the absorption to the visible light range.

Results

The crystalline phases structures of rutile TiO2 and rutile RuxTi1-xO2 nanobelts are shown in Fig. 1. RuO2 shows a rutile structure and it has lattice parameters similar to those of TiO2 rutile, therefore no new peaks were observed in the diffraction spectrums of nanobelts structures, regardless of Ru doping concentration.

Figure 1
figure 1

XRD patterns of RuxTi1-xO2 nanobelts a(x = 0); b(x = 0.01); c(x = 0.07); d(x = 0.16).

Whatever of the Ru doping concentration, the typical diffraction peaks which appear at 2θ = 28°, 36°, 42°, 55°, 57°, 63°, 64°, 68°, 69°could be assigned as the (110); (101); (111); (211); (220); (002); (310); (301); (112) diffraction lines corresponding to the rutile phase. The crystalline phase composition and lattice parameters from rutile nanobelts, having a tetragonal phase with lattice parameters a = b from 4.587 to 4.595 Å, and c = 2.957 to 2.967 Å is obtained with very high crystalline quality. The crystalline phase composition and lattice parameters from Rietveld refinement are summarized in Table 1.

Table 1 The crystalline phase composition and lattice parameters from Rietveld refinement.

It should be noted that the larger lattice constants and d-spacing values are due to upon Ru doping. This is due to the larger ionic radius of Ru than that of Ti in the rutile, the structural relaxation follows the Vegard’s law16. The d-spacing value of the rutile sample for (110) plane is in agreement to the interatomic dimension observed at HRTEM microscopy. Figure 2 shows the HRTEM images of RuxTi1-xO2 (x = 0.16) nanobelts sample. The TEM images confirm the formation of one-dimensional nanobelts nanostructures. Rutile TiO2 nanobelts have very small thicknesses, widths of 6–20 nm and can be 100–200 nm in length.

Figure 2
figure 2

HRTEM images for RuxTi1-xO2 (x = 0.16) nanobelts.

The HRTEM image shows the (101) and (002) atomic planes with lattice spacings of 3.499 Å (1/r) and 4.255 Å (1/r), respectively17. The interfacial angle between these two crystalline facets was found to be 67.27° and it was determined by Fast-Fourier Transform (FFT) image (Fig. 2). Deviation with one degree may be due to ruthenium from the crystalline lattice. These results even if less than the theoretical value, reveal that the interfacial angle between rutile is in good agreement with the theoretical value of the angle between the (101) and (001) planes. HRTEM images suggest that the prepared sample behaved like a well-crystallized heterostructure nanobelts.

Rutile RuxTi1-xO2 nanobelts have an octahedral arrangement, where either Ti or Ru atoms prefer a coordination number of 6. The layered arrangements of octahedrons facilitate their growth in the (001) direction as nanobelt like geometry14.

The UV-Vis diffuse reflectance spectroscopy (DRS) spectra of all samples are shown in Fig. 3A. The DRS method is employed to determine the band gap energy for photocatalysts. The intense absorption feature in the range of 200 nm–408 nm is characteristic of the TiO2 and corresponds to the band gap energy of 3.2 eV as for rutile phase. The visible absorption band in the range 408–627 nm corresponds to the band gap energy from 2 to 3.2 eV and can be assigned to charge transfer transition of the donor (Ru4+ → Ru5+  + e, Ru3+ → Ru4+  + e) or acceptor (Ru4+ → Ru3+  + h+) type18,19.

Figure 3
figure 3

(A) UV-vis spectra of RuxTi1-xO2 nanobelts; (B) XPS survey spectrum of RuxTi1-xO2 (x = 0.16) nanobelts; (C) High resolution core XPS spectra of RuxTi1-xO2 nanobelts a(x = 0); b(x = 0.01); c(x = 0.07); d(x = 0.16); (D) Raman spectrum of RuxTi1-xO2 nanobelts.

This suggests that rutile RuxTi1-xO2 photocatalysts could be active in the visible light region. The specters UV-vis from rutile RuxTi1-xO2 photocatalysts were transformed to the absorption specters according to the Kubelka Munk theory: \({[{\boldsymbol{F}}({{\boldsymbol{R}}}_{\infty })\cdot {\boldsymbol{hv}}]}^{1/2}\) (Fig. 4). The optical band gap energies of the photocatalysts can be approximated from the plot of \({[{\boldsymbol{F}}({{\boldsymbol{R}}}_{\infty })\cdot {\boldsymbol{hv}}]}^{1/2}\) versus hν (photon energy) and were estimated from the intercept of the tangent with the abscissa axis yielding the band gap energies. It can be observed that the gap energies band is presented in Fig. 4 compared to the bulk rutile TiO2 of 3.2 eV. The bandgap energies decreased from 3.2 eV for rutile TiO2 to 2.55 eV, 2.68 eV 3.07 eV for RuxTi1-xO2 x = 0.16, x = 0.07 and x = 0,01 respectively.

Figure 4
figure 4

Determination of energy band gap of RuxTi1-xO2 nanobelts a (x = 0); b (x = 0.01); c (x = 0.07); d (x = 0.16).

XPS measurements have been performed to analyze the surface composition and oxidation states of Ti and Ru in the samples. The survey spectrum for the RuxTi1-xO2 (x = 0.16), with the highest ruthenium content (Fig. 3B), confirms the complete removal of chlorine in the samples TiO2 doped. The absence of the peaks in area 197.9 eV and 199.5 eV (Cl 2p3/2 and Cl 2p1/2) were observed. Figure 3C shows the XPS spectra of Ti 2p on the surface of each sample. The binding energies at 454.0 and 459.7 eV (Ti 2p3/2 and Ti 2p1/2) correspond to octahedral coordinated Ti4+ state19,20. The energy difference between the two peaks is 5,7 eV which is consistent with the energy difference between the level of spin-orbit splitting coupling effect. c (Fig. 3C). The substitution of Ru ions into the TiO2 lattice can induce electronic structure.

The Raman spectra for rutile RuxTi1-xO2 photocatalysts are displayed in Fig. 3D, where two bands features of tetragonal rutile TiO2 (space group D4h) at 445 and 610 cm−1 were assigned to Eg (planar O-O vibration) and A1g (Ti-O stretch) modes 20. The broad band at 235 cm−1 was attributed to the multiple photon scattering process. The absence of the peaks to features attributable to RuO2 from 528 cm−1 and 646 cm−1 corresponds to Eg and A1g modes21, was evident in all of the samples, confirming the lattice substitution of the RuO2 in the rutile TiO2. Ruthenium doping induces a low red-shifts with increasing Ru concentration in all of the samples, probably due to the formation of oxygen defects or Ru-O-Ti linkages (Fig. 3D).

The photocatalytic properties of the rutile RuxTi1-xO2 nanobelts have been evaluated by photocatalytic degradation of methylene blue (MB) under sunlight at the 40000 lx intensity. The photocatalytic results are displayed in Fig. 5A. It was observed that the photodegradation process of methylene blue took place faster in the presence of rutile doped with ruthenium compared with undoped rutile. The increase of ruthenium content gave a decrease of the band gap that can be correlated with the increase absorption to the visible light range, and respectively enhance of the photocatalytic activity. It was observed that the photodegradation process of methylene blue took place faster in the presence of rutile RuxTi1-xO2 (x = 0.16) nanobelts compared to the other synthesized photocatalysts. It can be observed that after irradiation for 90 min, about 90% of the methylene blue has been degraded for the Ru0.16Ti0.84O2 nanobelts which are above the results of the rutile TiO2 (74 wt.%).

Figure 5
figure 5

(A) The C/C0 versus time curves of photocatalytic degradation of MB; (B) Photocatalytic performances of RuxTi1-xO2 nanobelts a(x = 0); b(x = 0.01); c(x = 0.07); d(x = 0.16).

The kinetic curves of sunlight MB degradation over each catalyst are depicted in Fig. 5B and show that the degradation of MB dye follows pseudo first-order kinetics law, lnC/C0 = kappt, where kapp is the pseudo-first order constant rate.

The constant rate kapp, was determined by plotting the lnC/C0 versus irradiation time. It can be observed that with increasing of the ruthenium content the constant rate kapp rises from 0.0125 min−1 to 0.0227 min−1 (Table 2).

Table 2 The rate constant kapp of RuxTi1-xO2 nanobelts.

On the basis of the experimental results, a possible mechanism of the enhanced photocatalytic activity over the rutile RuxTi1-xO2 is proposed in Fig. 6.

Figure 6
figure 6

Mechanism of RuxTi1-xO2 nanobelts photocatalysis.

The enhanced photocatalytic activity caused by ruthenium doping and to the ability of Ru to capture the photogenerated holes on the TiO2 (valence band). The conduction band can be calculated by using the empirical equation22:

Ecb = Evb − Eg and Evb of semiconductor were estimated by the following equation:

Evb = X − Ee + 1/2Eg, where Ecb is the conduction band, Evb is the valence band, X the geometric mean of the Mulliken, Ee is the energy of free electrons on the hydrogen scale (~4.5 eV) and Eg is the bandgap value of semiconductor, respectively. The values of Ecb, Evb are displayed in Table 3.

Table 3 The values Ecb, Evb of RuxTi1-xO2 nanobelts.

It can be observed that hydrogen peroxide and peroxide radicals can form. The photocatalytic process is shown in Fig. 6. Excitaded electrons under sunlight jumped from VB (valence band) to CB (conduction band) and generate charge carriers (electron-hole pairs). Due to the nanobelts structure, the electron-hole pairs are moved to the photocatalyst surface, where electrons and holes are involved in redox reactions on the surface. The photocatalytic degradation efficiency can be due to the reactive (001) facets, these having a strong ability to dissociate water molecules to form hydrogen peroxide and peroxide radicals, contributors in the photo-oxidation process16.

Discussions

RuxTi1-xO2 nanobelts were synthesized by the sol-gel method by doping TiO2 with RuO2. The reaction was carried out in HCl medium (c > 10 mol/l), which facilitated the formation of nanobelts nanostructures. RuO2 has been chosen for doping since both RuO2 and TiO2 have the same crystalline crystal structure, the crystalline parameters have almost equal values, coordinate number 6 for both Ti and Ru. From the diffraction spectrum, the rutile phase is observed for all nanobelts with high crystallinity. Also, doping TiO2 with ruthenium has a double role: decrease the bandgap and facilitating light from the visible field in the photocatalysis process, as well as to inhibit the recombination of the electron-hole pairs.

UV-vis absorption spectra were transformed using the Kubelka Munk function and the values of the bandgap energy for all samples with which the values of conduction bands and valence bands were calculated. In both the UV-vis and the transformed Kubelka Munk spectra, there are two types of transitions: interband between the 2p orbitals of the oxygen and the orbitals of the ruthenium and an interband type d-d between the orbitals of Ru (1.65–1.81 eV)23,24,25. TiO2 is considered a n-type semiconductor. There is a decrease in the bandgap from 3.17 to 2.55 eV as the content of Ru concentration increases in the samples, there is also an increase in the conductive band value, respectively, a decrease of valence value of nanobelts as the increase in doping with Ru. This favors the splitting of water and the formation of oxidoreduction species. Corroborated with the shape of nanostructures with surface exposure (001) facets showed the performance of efficient photocatalysts in the degradation of organic compounds. Doping with Ru induces a stress in the crystal of RuxTi1-xO2, which is revealed by an angle deviation of 68.3° to 67.3° for the (001) facets. Crystal stress is also noticeable in Raman spectra by easily moving to redshifts the bands assigned to Eg and A1g corresponding to the rutile phase of TiO2. The XPS spectrum shows a slight (low) displacement of the corresponding Ti 2p3/2 and Ti 2p1/2 blue-shifts spectra due to ruthenium doping. The high performance of the photocatalysts is due both to the nanobelts form and to the ruthenium doping and the visible light access to the photodegradation process. The superoxide and hydroxyl radicals obtained on the photocatalyst surface favor the mineralization reaction of the organic dye by in CO2 and H2O.

In conclusion, the rutile RuxTi1-xO2 nanobelts with exposure of (001) facets were successfully synthesized by a simple sol-gel method chemical route using hydrochloric acid and ethanol as capping and stabilizing agents. The superior photocatalytic activity of rutile nanobelts can be assigned to the band gap energy reduction from 3.2 to 2.55 eV and due to its small size, high surface area and of exposed highly reactive (001) facets. Also, the architecture of the heterostructure with exposure (001) facets favors the separation of photogenerated electron-hole pairs and inhibit the recombination of the electron–hole pairs resulting in the increase of the number of main superoxide and hydroxyl radicals.

A useful application of green technology can be the utilization of the synthesized photocatalysts in the remediation of the environment by decomposition of organic compounds under the sunlight.

Methods

Materials

The chemicals used in this work were of analytical reagent. Titanium n-butoxide Ti(OBu)4, ruthenium chloride (RuCl3xH2O), hydrochloric acid (HCl), methylene blue (MB) and NH4OH were purchased from Sigma-Aldrich. All solutions were prepared with distilled water.

Preparation of RuxTi1-xO2 nanobelts

The 1D (one-dimension) RuxTi1-xO2 nanobelts were synthesized via a sol-gel method described by Nguyen-Phan at al. 20. A typical synthesis implies mixing at room temperature for 30 minutes, of 14 mL of titanium n-butoxide and 14 mL of hydrochloric acid (35 wt.%). Then a desirable amount of RuCl3xH2O was added into the solution. The mixture was stirred for 12 h at 110 °C. Finally, the resulting precipitate was washed with aqueous 0.1 M NH4OH solution (20 ml), and with distilled water. After drying at 70 °C overnight, the products were calcinated in the air at 850 °C for 3 hours. The samples were denoted as RuxTi1-xO2, where x represented the nominal doping dosage of ruthenium (x = 0.01, 0.07 and 0.16). The TiO2 rutile sample was synthesized by a similar method without adding RuCl3xH2O precursor and labeled as TiO2.

Characterization

X-Ray diffraction (XRD) of the samples was analyzed at ambient temperature on a Bruker D8 Advance diffractometer using the characteristic Kα radiation of copper at a voltage of 40 kV and a current of 40 mA. XRD patterns were collected in the 2θ range between 5° and 80°.

The UV-vis diffuse reflectance spectra of rutile RuxTi1-xO2 were obtained by using a Jasco UV-Vis V-550 spectrophotometer in the wavelength range from 200 to 900 nm with an integrating sphere assembly. The sample was diluted with MgO (ratio 1:6) and then mechanically mixed. The UV-vis absorption was transformed according to the Kubelka Munk function, \(F({R}_{\infty })\), for infinite thick samples. The sample surface elements and their oxidation states were analyzed by Thermo Scientific K-Alpha X-ray Photoelectron Spectrometer (XPS) system with Al K-alfa radiation. The Raman spectra of samples were registered using a DXR Raman Microscope from Thermo Scientific. The morphology of the samples was characterized by transmission electron microscope Tecnai™ G2 F20 TWIN Cryo-TEM, FEI Company™, through bright field (BF-TEM) and scanning transmission electron microscopy analyses. The TEM was operated under an acceleration voltage of 200 kV. A small drop of well-dispersed sample, ultrasonicated for 5 minutes, was put on a carbon film copper grid and then, visualized on TEM.

Photocatalytic experiments

The photocatalytic studies of the rutile RuxTi1-xO2 and rutile TiO2 composites were evaluated via the degradation of MB under sunlight with a light intensity I = 40000 lx. For photocatalysis analysis, 0,1 g of the photocatalyst was suspended in a beaker with 50 ml aqueous of MB C0 = 10−4 mol/l. Before exposure to sunlight, the reaction systems were kept in the dark for 30 min to reach absorption-desorption equilibrium between RuxTi1-xO2 photocatalysts and MB solution. The absorption intensity of MB was monitoring by using a UV-vis spectrophotometer Jasco UV-vis V-540. At 30 min intervals, 2 ml solution was collected and it was measured absorption at 665 nm (sample is recovered).

The degradation efficiency (DC) of MB can be calculated via the formula:

Degradation efficiency \((DC \% )=\frac{{c}_{o}-c}{{c}_{o}}\), where C0 and C are concentrations of MB initial respectively measured at every 30 min. The photodegradation of MB follows pseudo-first order kinetics.